diff --git "a/out/checkpoint-17000/trainer_state.json" "b/out/checkpoint-17000/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/out/checkpoint-17000/trainer_state.json"
@@ -0,0 +1,119169 @@
+{
+  "best_metric": 2.3730249404907227,
+  "best_model_checkpoint": "./out/checkpoint-17000",
+  "epoch": 1.3719635219110644,
+  "eval_steps": 1000,
+  "global_step": 17000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 8.07037365830038e-05,
+      "grad_norm": 0.8911969065666199,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 2.6759,
+      "step": 1
+    },
+    {
+      "epoch": 0.0001614074731660076,
+      "grad_norm": 0.8724873661994934,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 2.7001,
+      "step": 2
+    },
+    {
+      "epoch": 0.00024211120974901139,
+      "grad_norm": 0.9050428867340088,
+      "learning_rate": 6e-06,
+      "loss": 2.6291,
+      "step": 3
+    },
+    {
+      "epoch": 0.0003228149463320152,
+      "grad_norm": 0.9249712824821472,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 2.7174,
+      "step": 4
+    },
+    {
+      "epoch": 0.000403518682915019,
+      "grad_norm": 0.9102846384048462,
+      "learning_rate": 1e-05,
+      "loss": 2.6831,
+      "step": 5
+    },
+    {
+      "epoch": 0.00048422241949802277,
+      "grad_norm": 0.9129141569137573,
+      "learning_rate": 1.2e-05,
+      "loss": 2.684,
+      "step": 6
+    },
+    {
+      "epoch": 0.0005649261560810266,
+      "grad_norm": 0.8648065328598022,
+      "learning_rate": 1.4000000000000001e-05,
+      "loss": 2.6488,
+      "step": 7
+    },
+    {
+      "epoch": 0.0006456298926640304,
+      "grad_norm": 0.8677545785903931,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 2.7143,
+      "step": 8
+    },
+    {
+      "epoch": 0.0007263336292470342,
+      "grad_norm": 0.919029712677002,
+      "learning_rate": 1.8e-05,
+      "loss": 2.631,
+      "step": 9
+    },
+    {
+      "epoch": 0.000807037365830038,
+      "grad_norm": 0.9289683103561401,
+      "learning_rate": 2e-05,
+      "loss": 2.6564,
+      "step": 10
+    },
+    {
+      "epoch": 0.0008877411024130417,
+      "grad_norm": 0.8810267448425293,
+      "learning_rate": 2.2000000000000003e-05,
+      "loss": 2.6395,
+      "step": 11
+    },
+    {
+      "epoch": 0.0009684448389960455,
+      "grad_norm": 0.8185754418373108,
+      "learning_rate": 2.4e-05,
+      "loss": 2.6871,
+      "step": 12
+    },
+    {
+      "epoch": 0.0010491485755790492,
+      "grad_norm": 0.9476913213729858,
+      "learning_rate": 2.6000000000000002e-05,
+      "loss": 2.7011,
+      "step": 13
+    },
+    {
+      "epoch": 0.0011298523121620531,
+      "grad_norm": 0.9616057872772217,
+      "learning_rate": 2.8000000000000003e-05,
+      "loss": 2.7373,
+      "step": 14
+    },
+    {
+      "epoch": 0.0012105560487450568,
+      "grad_norm": 0.9429686665534973,
+      "learning_rate": 3e-05,
+      "loss": 2.7556,
+      "step": 15
+    },
+    {
+      "epoch": 0.0012912597853280607,
+      "grad_norm": 1.0331422090530396,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 2.7756,
+      "step": 16
+    },
+    {
+      "epoch": 0.0013719635219110644,
+      "grad_norm": 0.906057596206665,
+      "learning_rate": 3.4000000000000007e-05,
+      "loss": 2.7053,
+      "step": 17
+    },
+    {
+      "epoch": 0.0014526672584940683,
+      "grad_norm": 0.8677626252174377,
+      "learning_rate": 3.6e-05,
+      "loss": 2.7012,
+      "step": 18
+    },
+    {
+      "epoch": 0.001533370995077072,
+      "grad_norm": 0.9378079175949097,
+      "learning_rate": 3.8e-05,
+      "loss": 2.6786,
+      "step": 19
+    },
+    {
+      "epoch": 0.001614074731660076,
+      "grad_norm": 1.0333882570266724,
+      "learning_rate": 4e-05,
+      "loss": 2.689,
+      "step": 20
+    },
+    {
+      "epoch": 0.0016947784682430796,
+      "grad_norm": 0.9435378909111023,
+      "learning_rate": 4.2e-05,
+      "loss": 2.7084,
+      "step": 21
+    },
+    {
+      "epoch": 0.0017754822048260835,
+      "grad_norm": 0.9530225396156311,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 2.7039,
+      "step": 22
+    },
+    {
+      "epoch": 0.0018561859414090872,
+      "grad_norm": 1.0154749155044556,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 2.6623,
+      "step": 23
+    },
+    {
+      "epoch": 0.001936889677992091,
+      "grad_norm": 1.0341671705245972,
+      "learning_rate": 4.8e-05,
+      "loss": 2.7072,
+      "step": 24
+    },
+    {
+      "epoch": 0.002017593414575095,
+      "grad_norm": 0.9185739159584045,
+      "learning_rate": 5e-05,
+      "loss": 2.6595,
+      "step": 25
+    },
+    {
+      "epoch": 0.0020982971511580985,
+      "grad_norm": 1.060390591621399,
+      "learning_rate": 5.2000000000000004e-05,
+      "loss": 2.7045,
+      "step": 26
+    },
+    {
+      "epoch": 0.0021790008877411024,
+      "grad_norm": 0.9720118641853333,
+      "learning_rate": 5.4000000000000005e-05,
+      "loss": 2.6513,
+      "step": 27
+    },
+    {
+      "epoch": 0.0022597046243241063,
+      "grad_norm": 0.9426784515380859,
+      "learning_rate": 5.6000000000000006e-05,
+      "loss": 2.6541,
+      "step": 28
+    },
+    {
+      "epoch": 0.00234040836090711,
+      "grad_norm": 0.9736170768737793,
+      "learning_rate": 5.8e-05,
+      "loss": 2.7324,
+      "step": 29
+    },
+    {
+      "epoch": 0.0024211120974901136,
+      "grad_norm": 0.9831354022026062,
+      "learning_rate": 6e-05,
+      "loss": 2.6651,
+      "step": 30
+    },
+    {
+      "epoch": 0.0025018158340731175,
+      "grad_norm": 1.0222605466842651,
+      "learning_rate": 6.2e-05,
+      "loss": 2.7375,
+      "step": 31
+    },
+    {
+      "epoch": 0.0025825195706561214,
+      "grad_norm": 0.9182235598564148,
+      "learning_rate": 6.400000000000001e-05,
+      "loss": 2.7142,
+      "step": 32
+    },
+    {
+      "epoch": 0.0026632233072391254,
+      "grad_norm": 1.0200958251953125,
+      "learning_rate": 6.6e-05,
+      "loss": 2.6785,
+      "step": 33
+    },
+    {
+      "epoch": 0.002743927043822129,
+      "grad_norm": 1.0153381824493408,
+      "learning_rate": 6.800000000000001e-05,
+      "loss": 2.6737,
+      "step": 34
+    },
+    {
+      "epoch": 0.0028246307804051327,
+      "grad_norm": 0.8998087644577026,
+      "learning_rate": 7e-05,
+      "loss": 2.7594,
+      "step": 35
+    },
+    {
+      "epoch": 0.0029053345169881366,
+      "grad_norm": 0.9005621671676636,
+      "learning_rate": 7.2e-05,
+      "loss": 2.713,
+      "step": 36
+    },
+    {
+      "epoch": 0.0029860382535711405,
+      "grad_norm": 1.0165663957595825,
+      "learning_rate": 7.4e-05,
+      "loss": 2.7197,
+      "step": 37
+    },
+    {
+      "epoch": 0.003066741990154144,
+      "grad_norm": 1.0011894702911377,
+      "learning_rate": 7.6e-05,
+      "loss": 2.6315,
+      "step": 38
+    },
+    {
+      "epoch": 0.003147445726737148,
+      "grad_norm": 1.141209602355957,
+      "learning_rate": 7.800000000000001e-05,
+      "loss": 2.7249,
+      "step": 39
+    },
+    {
+      "epoch": 0.003228149463320152,
+      "grad_norm": 0.9114719033241272,
+      "learning_rate": 8e-05,
+      "loss": 2.7039,
+      "step": 40
+    },
+    {
+      "epoch": 0.0033088531999031557,
+      "grad_norm": 1.0193392038345337,
+      "learning_rate": 8.2e-05,
+      "loss": 2.6501,
+      "step": 41
+    },
+    {
+      "epoch": 0.003389556936486159,
+      "grad_norm": 0.9458270072937012,
+      "learning_rate": 8.4e-05,
+      "loss": 2.725,
+      "step": 42
+    },
+    {
+      "epoch": 0.003470260673069163,
+      "grad_norm": 0.9667492508888245,
+      "learning_rate": 8.6e-05,
+      "loss": 2.7232,
+      "step": 43
+    },
+    {
+      "epoch": 0.003550964409652167,
+      "grad_norm": 0.9987972378730774,
+      "learning_rate": 8.800000000000001e-05,
+      "loss": 2.6554,
+      "step": 44
+    },
+    {
+      "epoch": 0.003631668146235171,
+      "grad_norm": 1.0166393518447876,
+      "learning_rate": 9e-05,
+      "loss": 2.7291,
+      "step": 45
+    },
+    {
+      "epoch": 0.0037123718828181744,
+      "grad_norm": 0.9557009935379028,
+      "learning_rate": 9.200000000000001e-05,
+      "loss": 2.7194,
+      "step": 46
+    },
+    {
+      "epoch": 0.0037930756194011783,
+      "grad_norm": 0.9575492143630981,
+      "learning_rate": 9.4e-05,
+      "loss": 2.6671,
+      "step": 47
+    },
+    {
+      "epoch": 0.003873779355984182,
+      "grad_norm": 0.9614555239677429,
+      "learning_rate": 9.6e-05,
+      "loss": 2.6865,
+      "step": 48
+    },
+    {
+      "epoch": 0.003954483092567186,
+      "grad_norm": 0.9245515465736389,
+      "learning_rate": 9.8e-05,
+      "loss": 2.7821,
+      "step": 49
+    },
+    {
+      "epoch": 0.00403518682915019,
+      "grad_norm": 0.9756044745445251,
+      "learning_rate": 0.0001,
+      "loss": 2.7608,
+      "step": 50
+    },
+    {
+      "epoch": 0.0041158905657331935,
+      "grad_norm": 0.95787513256073,
+      "learning_rate": 0.00010200000000000001,
+      "loss": 2.6458,
+      "step": 51
+    },
+    {
+      "epoch": 0.004196594302316197,
+      "grad_norm": 1.0102490186691284,
+      "learning_rate": 0.00010400000000000001,
+      "loss": 2.7835,
+      "step": 52
+    },
+    {
+      "epoch": 0.004277298038899201,
+      "grad_norm": 0.9676176309585571,
+      "learning_rate": 0.00010600000000000002,
+      "loss": 2.702,
+      "step": 53
+    },
+    {
+      "epoch": 0.004358001775482205,
+      "grad_norm": 0.9724096655845642,
+      "learning_rate": 0.00010800000000000001,
+      "loss": 2.714,
+      "step": 54
+    },
+    {
+      "epoch": 0.004438705512065208,
+      "grad_norm": 0.9482994675636292,
+      "learning_rate": 0.00011000000000000002,
+      "loss": 2.8069,
+      "step": 55
+    },
+    {
+      "epoch": 0.0045194092486482125,
+      "grad_norm": 0.9886480569839478,
+      "learning_rate": 0.00011200000000000001,
+      "loss": 2.7468,
+      "step": 56
+    },
+    {
+      "epoch": 0.004600112985231216,
+      "grad_norm": 0.9696247577667236,
+      "learning_rate": 0.00011399999999999999,
+      "loss": 2.7486,
+      "step": 57
+    },
+    {
+      "epoch": 0.00468081672181422,
+      "grad_norm": 1.0638912916183472,
+      "learning_rate": 0.000116,
+      "loss": 2.7747,
+      "step": 58
+    },
+    {
+      "epoch": 0.004761520458397224,
+      "grad_norm": 1.016483187675476,
+      "learning_rate": 0.000118,
+      "loss": 2.6925,
+      "step": 59
+    },
+    {
+      "epoch": 0.004842224194980227,
+      "grad_norm": 1.0298779010772705,
+      "learning_rate": 0.00012,
+      "loss": 2.7487,
+      "step": 60
+    },
+    {
+      "epoch": 0.004922927931563232,
+      "grad_norm": 1.1082268953323364,
+      "learning_rate": 0.000122,
+      "loss": 2.7697,
+      "step": 61
+    },
+    {
+      "epoch": 0.005003631668146235,
+      "grad_norm": 0.9202101826667786,
+      "learning_rate": 0.000124,
+      "loss": 2.7429,
+      "step": 62
+    },
+    {
+      "epoch": 0.0050843354047292386,
+      "grad_norm": 1.0140503644943237,
+      "learning_rate": 0.000126,
+      "loss": 2.7492,
+      "step": 63
+    },
+    {
+      "epoch": 0.005165039141312243,
+      "grad_norm": 1.0689163208007812,
+      "learning_rate": 0.00012800000000000002,
+      "loss": 2.7353,
+      "step": 64
+    },
+    {
+      "epoch": 0.005245742877895246,
+      "grad_norm": 0.9947141408920288,
+      "learning_rate": 0.00013000000000000002,
+      "loss": 2.7385,
+      "step": 65
+    },
+    {
+      "epoch": 0.005326446614478251,
+      "grad_norm": 1.2034410238265991,
+      "learning_rate": 0.000132,
+      "loss": 2.7632,
+      "step": 66
+    },
+    {
+      "epoch": 0.005407150351061254,
+      "grad_norm": 0.9450412392616272,
+      "learning_rate": 0.000134,
+      "loss": 2.7547,
+      "step": 67
+    },
+    {
+      "epoch": 0.005487854087644258,
+      "grad_norm": 1.1818269491195679,
+      "learning_rate": 0.00013600000000000003,
+      "loss": 2.7663,
+      "step": 68
+    },
+    {
+      "epoch": 0.005568557824227262,
+      "grad_norm": 1.003347396850586,
+      "learning_rate": 0.000138,
+      "loss": 2.7299,
+      "step": 69
+    },
+    {
+      "epoch": 0.0056492615608102655,
+      "grad_norm": 1.0105760097503662,
+      "learning_rate": 0.00014,
+      "loss": 2.7261,
+      "step": 70
+    },
+    {
+      "epoch": 0.005729965297393269,
+      "grad_norm": 0.9459090232849121,
+      "learning_rate": 0.000142,
+      "loss": 2.7237,
+      "step": 71
+    },
+    {
+      "epoch": 0.005810669033976273,
+      "grad_norm": 0.9716219305992126,
+      "learning_rate": 0.000144,
+      "loss": 2.8175,
+      "step": 72
+    },
+    {
+      "epoch": 0.005891372770559277,
+      "grad_norm": 0.9968419075012207,
+      "learning_rate": 0.000146,
+      "loss": 2.7828,
+      "step": 73
+    },
+    {
+      "epoch": 0.005972076507142281,
+      "grad_norm": 1.099680781364441,
+      "learning_rate": 0.000148,
+      "loss": 2.7111,
+      "step": 74
+    },
+    {
+      "epoch": 0.0060527802437252845,
+      "grad_norm": 1.004846453666687,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 2.7508,
+      "step": 75
+    },
+    {
+      "epoch": 0.006133483980308288,
+      "grad_norm": 1.0568128824234009,
+      "learning_rate": 0.000152,
+      "loss": 2.7341,
+      "step": 76
+    },
+    {
+      "epoch": 0.006214187716891292,
+      "grad_norm": 0.9871000051498413,
+      "learning_rate": 0.000154,
+      "loss": 2.7831,
+      "step": 77
+    },
+    {
+      "epoch": 0.006294891453474296,
+      "grad_norm": 1.005947232246399,
+      "learning_rate": 0.00015600000000000002,
+      "loss": 2.6798,
+      "step": 78
+    },
+    {
+      "epoch": 0.006375595190057299,
+      "grad_norm": 0.9984713792800903,
+      "learning_rate": 0.00015800000000000002,
+      "loss": 2.8126,
+      "step": 79
+    },
+    {
+      "epoch": 0.006456298926640304,
+      "grad_norm": 0.9805751442909241,
+      "learning_rate": 0.00016,
+      "loss": 2.7826,
+      "step": 80
+    },
+    {
+      "epoch": 0.006537002663223307,
+      "grad_norm": 1.02998685836792,
+      "learning_rate": 0.000162,
+      "loss": 2.7636,
+      "step": 81
+    },
+    {
+      "epoch": 0.006617706399806311,
+      "grad_norm": 1.0790135860443115,
+      "learning_rate": 0.000164,
+      "loss": 2.7809,
+      "step": 82
+    },
+    {
+      "epoch": 0.006698410136389315,
+      "grad_norm": 1.1058307886123657,
+      "learning_rate": 0.000166,
+      "loss": 2.787,
+      "step": 83
+    },
+    {
+      "epoch": 0.006779113872972318,
+      "grad_norm": 1.0199624300003052,
+      "learning_rate": 0.000168,
+      "loss": 2.7171,
+      "step": 84
+    },
+    {
+      "epoch": 0.006859817609555323,
+      "grad_norm": 1.006494402885437,
+      "learning_rate": 0.00017,
+      "loss": 2.7791,
+      "step": 85
+    },
+    {
+      "epoch": 0.006940521346138326,
+      "grad_norm": 0.9672449827194214,
+      "learning_rate": 0.000172,
+      "loss": 2.6929,
+      "step": 86
+    },
+    {
+      "epoch": 0.00702122508272133,
+      "grad_norm": 0.9747781157493591,
+      "learning_rate": 0.000174,
+      "loss": 2.7676,
+      "step": 87
+    },
+    {
+      "epoch": 0.007101928819304334,
+      "grad_norm": 0.9193839430809021,
+      "learning_rate": 0.00017600000000000002,
+      "loss": 2.7124,
+      "step": 88
+    },
+    {
+      "epoch": 0.0071826325558873375,
+      "grad_norm": 1.078499436378479,
+      "learning_rate": 0.00017800000000000002,
+      "loss": 2.8018,
+      "step": 89
+    },
+    {
+      "epoch": 0.007263336292470342,
+      "grad_norm": 1.070957899093628,
+      "learning_rate": 0.00018,
+      "loss": 2.7889,
+      "step": 90
+    },
+    {
+      "epoch": 0.007344040029053345,
+      "grad_norm": 1.160942554473877,
+      "learning_rate": 0.000182,
+      "loss": 2.8026,
+      "step": 91
+    },
+    {
+      "epoch": 0.007424743765636349,
+      "grad_norm": 0.9988501071929932,
+      "learning_rate": 0.00018400000000000003,
+      "loss": 2.7746,
+      "step": 92
+    },
+    {
+      "epoch": 0.007505447502219353,
+      "grad_norm": 1.0882319211959839,
+      "learning_rate": 0.00018600000000000002,
+      "loss": 2.8105,
+      "step": 93
+    },
+    {
+      "epoch": 0.0075861512388023565,
+      "grad_norm": 1.1882357597351074,
+      "learning_rate": 0.000188,
+      "loss": 2.8294,
+      "step": 94
+    },
+    {
+      "epoch": 0.00766685497538536,
+      "grad_norm": 1.0761829614639282,
+      "learning_rate": 0.00019,
+      "loss": 2.7846,
+      "step": 95
+    },
+    {
+      "epoch": 0.007747558711968364,
+      "grad_norm": 1.0665982961654663,
+      "learning_rate": 0.000192,
+      "loss": 2.8542,
+      "step": 96
+    },
+    {
+      "epoch": 0.007828262448551369,
+      "grad_norm": 1.206127405166626,
+      "learning_rate": 0.000194,
+      "loss": 2.7711,
+      "step": 97
+    },
+    {
+      "epoch": 0.007908966185134371,
+      "grad_norm": 1.095150113105774,
+      "learning_rate": 0.000196,
+      "loss": 2.732,
+      "step": 98
+    },
+    {
+      "epoch": 0.007989669921717376,
+      "grad_norm": 1.118348240852356,
+      "learning_rate": 0.00019800000000000002,
+      "loss": 2.7736,
+      "step": 99
+    },
+    {
+      "epoch": 0.00807037365830038,
+      "grad_norm": 1.0646461248397827,
+      "learning_rate": 0.0002,
+      "loss": 2.8584,
+      "step": 100
+    },
+    {
+      "epoch": 0.008151077394883383,
+      "grad_norm": 1.0387661457061768,
+      "learning_rate": 0.0001999999987538693,
+      "loss": 2.7961,
+      "step": 101
+    },
+    {
+      "epoch": 0.008231781131466387,
+      "grad_norm": 1.1905474662780762,
+      "learning_rate": 0.00019999999501547723,
+      "loss": 2.8615,
+      "step": 102
+    },
+    {
+      "epoch": 0.008312484868049391,
+      "grad_norm": 0.9630722999572754,
+      "learning_rate": 0.0001999999887848239,
+      "loss": 2.8076,
+      "step": 103
+    },
+    {
+      "epoch": 0.008393188604632394,
+      "grad_norm": 1.1034537553787231,
+      "learning_rate": 0.00019999998006190942,
+      "loss": 2.8402,
+      "step": 104
+    },
+    {
+      "epoch": 0.008473892341215398,
+      "grad_norm": 1.0679295063018799,
+      "learning_rate": 0.00019999996884673403,
+      "loss": 2.7948,
+      "step": 105
+    },
+    {
+      "epoch": 0.008554596077798403,
+      "grad_norm": 1.0108860731124878,
+      "learning_rate": 0.00019999995513929802,
+      "loss": 2.7996,
+      "step": 106
+    },
+    {
+      "epoch": 0.008635299814381405,
+      "grad_norm": 1.3762084245681763,
+      "learning_rate": 0.0001999999389396017,
+      "loss": 2.8023,
+      "step": 107
+    },
+    {
+      "epoch": 0.00871600355096441,
+      "grad_norm": 1.1320533752441406,
+      "learning_rate": 0.00019999992024764555,
+      "loss": 2.793,
+      "step": 108
+    },
+    {
+      "epoch": 0.008796707287547414,
+      "grad_norm": 1.1752389669418335,
+      "learning_rate": 0.00019999989906342998,
+      "loss": 2.8274,
+      "step": 109
+    },
+    {
+      "epoch": 0.008877411024130416,
+      "grad_norm": 1.2734956741333008,
+      "learning_rate": 0.00019999987538695552,
+      "loss": 2.8017,
+      "step": 110
+    },
+    {
+      "epoch": 0.00895811476071342,
+      "grad_norm": 1.3703055381774902,
+      "learning_rate": 0.00019999984921822273,
+      "loss": 2.8699,
+      "step": 111
+    },
+    {
+      "epoch": 0.009038818497296425,
+      "grad_norm": 1.0079127550125122,
+      "learning_rate": 0.0001999998205572323,
+      "loss": 2.8845,
+      "step": 112
+    },
+    {
+      "epoch": 0.00911952223387943,
+      "grad_norm": 1.28025484085083,
+      "learning_rate": 0.000199999789403985,
+      "loss": 2.8636,
+      "step": 113
+    },
+    {
+      "epoch": 0.009200225970462432,
+      "grad_norm": 1.1057093143463135,
+      "learning_rate": 0.00019999975575848148,
+      "loss": 2.8484,
+      "step": 114
+    },
+    {
+      "epoch": 0.009280929707045436,
+      "grad_norm": 1.0874677896499634,
+      "learning_rate": 0.00019999971962072265,
+      "loss": 2.7314,
+      "step": 115
+    },
+    {
+      "epoch": 0.00936163344362844,
+      "grad_norm": 1.0909658670425415,
+      "learning_rate": 0.00019999968099070943,
+      "loss": 2.7827,
+      "step": 116
+    },
+    {
+      "epoch": 0.009442337180211443,
+      "grad_norm": 1.0881624221801758,
+      "learning_rate": 0.00019999963986844273,
+      "loss": 2.827,
+      "step": 117
+    },
+    {
+      "epoch": 0.009523040916794448,
+      "grad_norm": 1.2498180866241455,
+      "learning_rate": 0.00019999959625392362,
+      "loss": 2.8695,
+      "step": 118
+    },
+    {
+      "epoch": 0.009603744653377452,
+      "grad_norm": 1.1344549655914307,
+      "learning_rate": 0.00019999955014715317,
+      "loss": 2.8079,
+      "step": 119
+    },
+    {
+      "epoch": 0.009684448389960455,
+      "grad_norm": 1.032563328742981,
+      "learning_rate": 0.00019999950154813253,
+      "loss": 2.7787,
+      "step": 120
+    },
+    {
+      "epoch": 0.009765152126543459,
+      "grad_norm": 0.9630110263824463,
+      "learning_rate": 0.0001999994504568629,
+      "loss": 2.8103,
+      "step": 121
+    },
+    {
+      "epoch": 0.009845855863126463,
+      "grad_norm": 1.0418641567230225,
+      "learning_rate": 0.0001999993968733456,
+      "loss": 2.8679,
+      "step": 122
+    },
+    {
+      "epoch": 0.009926559599709466,
+      "grad_norm": 0.9797310829162598,
+      "learning_rate": 0.00019999934079758188,
+      "loss": 2.7792,
+      "step": 123
+    },
+    {
+      "epoch": 0.01000726333629247,
+      "grad_norm": 1.0494028329849243,
+      "learning_rate": 0.00019999928222957323,
+      "loss": 2.8007,
+      "step": 124
+    },
+    {
+      "epoch": 0.010087967072875475,
+      "grad_norm": 1.1570640802383423,
+      "learning_rate": 0.00019999922116932105,
+      "loss": 2.8331,
+      "step": 125
+    },
+    {
+      "epoch": 0.010168670809458477,
+      "grad_norm": 1.2753098011016846,
+      "learning_rate": 0.00019999915761682684,
+      "loss": 2.8533,
+      "step": 126
+    },
+    {
+      "epoch": 0.010249374546041481,
+      "grad_norm": 0.9804013967514038,
+      "learning_rate": 0.00019999909157209227,
+      "loss": 2.841,
+      "step": 127
+    },
+    {
+      "epoch": 0.010330078282624486,
+      "grad_norm": 1.320839285850525,
+      "learning_rate": 0.00019999902303511892,
+      "loss": 2.8738,
+      "step": 128
+    },
+    {
+      "epoch": 0.01041078201920749,
+      "grad_norm": 1.1105059385299683,
+      "learning_rate": 0.0001999989520059085,
+      "loss": 2.8458,
+      "step": 129
+    },
+    {
+      "epoch": 0.010491485755790493,
+      "grad_norm": 1.2869762182235718,
+      "learning_rate": 0.0001999988784844628,
+      "loss": 2.7951,
+      "step": 130
+    },
+    {
+      "epoch": 0.010572189492373497,
+      "grad_norm": 1.1609153747558594,
+      "learning_rate": 0.00019999880247078368,
+      "loss": 2.8147,
+      "step": 131
+    },
+    {
+      "epoch": 0.010652893228956501,
+      "grad_norm": 1.066728115081787,
+      "learning_rate": 0.00019999872396487297,
+      "loss": 2.863,
+      "step": 132
+    },
+    {
+      "epoch": 0.010733596965539504,
+      "grad_norm": 1.2868720293045044,
+      "learning_rate": 0.0001999986429667327,
+      "loss": 2.7765,
+      "step": 133
+    },
+    {
+      "epoch": 0.010814300702122508,
+      "grad_norm": 1.0064955949783325,
+      "learning_rate": 0.00019999855947636485,
+      "loss": 2.7834,
+      "step": 134
+    },
+    {
+      "epoch": 0.010895004438705513,
+      "grad_norm": 1.146589756011963,
+      "learning_rate": 0.00019999847349377143,
+      "loss": 2.7966,
+      "step": 135
+    },
+    {
+      "epoch": 0.010975708175288515,
+      "grad_norm": 0.9831073880195618,
+      "learning_rate": 0.0001999983850189547,
+      "loss": 2.8877,
+      "step": 136
+    },
+    {
+      "epoch": 0.01105641191187152,
+      "grad_norm": 1.1690322160720825,
+      "learning_rate": 0.0001999982940519168,
+      "loss": 2.8514,
+      "step": 137
+    },
+    {
+      "epoch": 0.011137115648454524,
+      "grad_norm": 1.0014944076538086,
+      "learning_rate": 0.00019999820059266003,
+      "loss": 2.7846,
+      "step": 138
+    },
+    {
+      "epoch": 0.011217819385037527,
+      "grad_norm": 0.9581566452980042,
+      "learning_rate": 0.0001999981046411867,
+      "loss": 2.7907,
+      "step": 139
+    },
+    {
+      "epoch": 0.011298523121620531,
+      "grad_norm": 1.1300675868988037,
+      "learning_rate": 0.00019999800619749922,
+      "loss": 2.8099,
+      "step": 140
+    },
+    {
+      "epoch": 0.011379226858203535,
+      "grad_norm": 0.9845526814460754,
+      "learning_rate": 0.0001999979052616,
+      "loss": 2.8607,
+      "step": 141
+    },
+    {
+      "epoch": 0.011459930594786538,
+      "grad_norm": 1.0781387090682983,
+      "learning_rate": 0.0001999978018334916,
+      "loss": 2.831,
+      "step": 142
+    },
+    {
+      "epoch": 0.011540634331369542,
+      "grad_norm": 1.1142648458480835,
+      "learning_rate": 0.00019999769591317658,
+      "loss": 2.9194,
+      "step": 143
+    },
+    {
+      "epoch": 0.011621338067952547,
+      "grad_norm": 0.9972650408744812,
+      "learning_rate": 0.00019999758750065757,
+      "loss": 2.8253,
+      "step": 144
+    },
+    {
+      "epoch": 0.01170204180453555,
+      "grad_norm": 1.040738582611084,
+      "learning_rate": 0.0001999974765959373,
+      "loss": 2.7378,
+      "step": 145
+    },
+    {
+      "epoch": 0.011782745541118553,
+      "grad_norm": 0.9824327826499939,
+      "learning_rate": 0.00019999736319901848,
+      "loss": 2.8263,
+      "step": 146
+    },
+    {
+      "epoch": 0.011863449277701558,
+      "grad_norm": 1.0531679391860962,
+      "learning_rate": 0.00019999724730990402,
+      "loss": 2.7975,
+      "step": 147
+    },
+    {
+      "epoch": 0.011944153014284562,
+      "grad_norm": 1.0699561834335327,
+      "learning_rate": 0.0001999971289285967,
+      "loss": 2.8199,
+      "step": 148
+    },
+    {
+      "epoch": 0.012024856750867565,
+      "grad_norm": 1.0203633308410645,
+      "learning_rate": 0.0001999970080550996,
+      "loss": 2.8479,
+      "step": 149
+    },
+    {
+      "epoch": 0.012105560487450569,
+      "grad_norm": 1.035589575767517,
+      "learning_rate": 0.00019999688468941564,
+      "loss": 2.8263,
+      "step": 150
+    },
+    {
+      "epoch": 0.012186264224033573,
+      "grad_norm": 0.9706670641899109,
+      "learning_rate": 0.00019999675883154792,
+      "loss": 2.8324,
+      "step": 151
+    },
+    {
+      "epoch": 0.012266967960616576,
+      "grad_norm": 1.1565446853637695,
+      "learning_rate": 0.00019999663048149958,
+      "loss": 2.8098,
+      "step": 152
+    },
+    {
+      "epoch": 0.01234767169719958,
+      "grad_norm": 1.025796890258789,
+      "learning_rate": 0.0001999964996392738,
+      "loss": 2.7906,
+      "step": 153
+    },
+    {
+      "epoch": 0.012428375433782585,
+      "grad_norm": 1.117438554763794,
+      "learning_rate": 0.00019999636630487386,
+      "loss": 2.8276,
+      "step": 154
+    },
+    {
+      "epoch": 0.012509079170365587,
+      "grad_norm": 1.025159478187561,
+      "learning_rate": 0.00019999623047830308,
+      "loss": 2.8089,
+      "step": 155
+    },
+    {
+      "epoch": 0.012589782906948592,
+      "grad_norm": 1.007582664489746,
+      "learning_rate": 0.00019999609215956487,
+      "loss": 2.8147,
+      "step": 156
+    },
+    {
+      "epoch": 0.012670486643531596,
+      "grad_norm": 1.0504885911941528,
+      "learning_rate": 0.0001999959513486626,
+      "loss": 2.8329,
+      "step": 157
+    },
+    {
+      "epoch": 0.012751190380114599,
+      "grad_norm": 0.918382465839386,
+      "learning_rate": 0.00019999580804559987,
+      "loss": 2.878,
+      "step": 158
+    },
+    {
+      "epoch": 0.012831894116697603,
+      "grad_norm": 0.9397236704826355,
+      "learning_rate": 0.0001999956622503802,
+      "loss": 2.8254,
+      "step": 159
+    },
+    {
+      "epoch": 0.012912597853280607,
+      "grad_norm": 0.9985697269439697,
+      "learning_rate": 0.00019999551396300723,
+      "loss": 2.8417,
+      "step": 160
+    },
+    {
+      "epoch": 0.01299330158986361,
+      "grad_norm": 0.9866878390312195,
+      "learning_rate": 0.00019999536318348465,
+      "loss": 2.7524,
+      "step": 161
+    },
+    {
+      "epoch": 0.013074005326446614,
+      "grad_norm": 1.0707440376281738,
+      "learning_rate": 0.00019999520991181627,
+      "loss": 2.8171,
+      "step": 162
+    },
+    {
+      "epoch": 0.013154709063029619,
+      "grad_norm": 0.9359755516052246,
+      "learning_rate": 0.00019999505414800583,
+      "loss": 2.8463,
+      "step": 163
+    },
+    {
+      "epoch": 0.013235412799612623,
+      "grad_norm": 1.056647777557373,
+      "learning_rate": 0.00019999489589205726,
+      "loss": 2.8602,
+      "step": 164
+    },
+    {
+      "epoch": 0.013316116536195625,
+      "grad_norm": 0.975370466709137,
+      "learning_rate": 0.0001999947351439745,
+      "loss": 2.8292,
+      "step": 165
+    },
+    {
+      "epoch": 0.01339682027277863,
+      "grad_norm": 0.9241237044334412,
+      "learning_rate": 0.00019999457190376157,
+      "loss": 2.7827,
+      "step": 166
+    },
+    {
+      "epoch": 0.013477524009361634,
+      "grad_norm": 0.9478302001953125,
+      "learning_rate": 0.00019999440617142247,
+      "loss": 2.7708,
+      "step": 167
+    },
+    {
+      "epoch": 0.013558227745944637,
+      "grad_norm": 0.9804863333702087,
+      "learning_rate": 0.00019999423794696142,
+      "loss": 2.7696,
+      "step": 168
+    },
+    {
+      "epoch": 0.013638931482527641,
+      "grad_norm": 0.9764013886451721,
+      "learning_rate": 0.00019999406723038255,
+      "loss": 2.8521,
+      "step": 169
+    },
+    {
+      "epoch": 0.013719635219110645,
+      "grad_norm": 1.026532769203186,
+      "learning_rate": 0.00019999389402169016,
+      "loss": 2.8507,
+      "step": 170
+    },
+    {
+      "epoch": 0.013800338955693648,
+      "grad_norm": 0.9983204007148743,
+      "learning_rate": 0.00019999371832088854,
+      "loss": 2.8761,
+      "step": 171
+    },
+    {
+      "epoch": 0.013881042692276652,
+      "grad_norm": 0.9914593696594238,
+      "learning_rate": 0.00019999354012798206,
+      "loss": 2.8723,
+      "step": 172
+    },
+    {
+      "epoch": 0.013961746428859657,
+      "grad_norm": 1.066962718963623,
+      "learning_rate": 0.00019999335944297517,
+      "loss": 2.8635,
+      "step": 173
+    },
+    {
+      "epoch": 0.01404245016544266,
+      "grad_norm": 1.0848973989486694,
+      "learning_rate": 0.0001999931762658724,
+      "loss": 2.8645,
+      "step": 174
+    },
+    {
+      "epoch": 0.014123153902025664,
+      "grad_norm": 1.0245702266693115,
+      "learning_rate": 0.0001999929905966783,
+      "loss": 2.8463,
+      "step": 175
+    },
+    {
+      "epoch": 0.014203857638608668,
+      "grad_norm": 1.2363669872283936,
+      "learning_rate": 0.00019999280243539747,
+      "loss": 2.8345,
+      "step": 176
+    },
+    {
+      "epoch": 0.01428456137519167,
+      "grad_norm": 1.0224756002426147,
+      "learning_rate": 0.0001999926117820346,
+      "loss": 2.8309,
+      "step": 177
+    },
+    {
+      "epoch": 0.014365265111774675,
+      "grad_norm": 1.0882402658462524,
+      "learning_rate": 0.0001999924186365945,
+      "loss": 2.8619,
+      "step": 178
+    },
+    {
+      "epoch": 0.01444596884835768,
+      "grad_norm": 1.0384254455566406,
+      "learning_rate": 0.00019999222299908192,
+      "loss": 2.8477,
+      "step": 179
+    },
+    {
+      "epoch": 0.014526672584940684,
+      "grad_norm": 0.9662587642669678,
+      "learning_rate": 0.00019999202486950177,
+      "loss": 2.8087,
+      "step": 180
+    },
+    {
+      "epoch": 0.014607376321523686,
+      "grad_norm": 0.9086892604827881,
+      "learning_rate": 0.000199991824247859,
+      "loss": 2.7688,
+      "step": 181
+    },
+    {
+      "epoch": 0.01468808005810669,
+      "grad_norm": 1.004185676574707,
+      "learning_rate": 0.00019999162113415854,
+      "loss": 2.8237,
+      "step": 182
+    },
+    {
+      "epoch": 0.014768783794689695,
+      "grad_norm": 0.997965395450592,
+      "learning_rate": 0.00019999141552840552,
+      "loss": 2.8228,
+      "step": 183
+    },
+    {
+      "epoch": 0.014849487531272697,
+      "grad_norm": 0.9844975471496582,
+      "learning_rate": 0.00019999120743060503,
+      "loss": 2.8582,
+      "step": 184
+    },
+    {
+      "epoch": 0.014930191267855702,
+      "grad_norm": 1.0531272888183594,
+      "learning_rate": 0.00019999099684076232,
+      "loss": 2.8571,
+      "step": 185
+    },
+    {
+      "epoch": 0.015010895004438706,
+      "grad_norm": 1.1178920269012451,
+      "learning_rate": 0.00019999078375888257,
+      "loss": 2.85,
+      "step": 186
+    },
+    {
+      "epoch": 0.015091598741021709,
+      "grad_norm": 1.0773903131484985,
+      "learning_rate": 0.0001999905681849711,
+      "loss": 2.826,
+      "step": 187
+    },
+    {
+      "epoch": 0.015172302477604713,
+      "grad_norm": 1.1573486328125,
+      "learning_rate": 0.00019999035011903325,
+      "loss": 2.8866,
+      "step": 188
+    },
+    {
+      "epoch": 0.015253006214187717,
+      "grad_norm": 1.0401980876922607,
+      "learning_rate": 0.00019999012956107456,
+      "loss": 2.788,
+      "step": 189
+    },
+    {
+      "epoch": 0.01533370995077072,
+      "grad_norm": 1.0150686502456665,
+      "learning_rate": 0.00019998990651110045,
+      "loss": 2.8542,
+      "step": 190
+    },
+    {
+      "epoch": 0.015414413687353724,
+      "grad_norm": 1.1902797222137451,
+      "learning_rate": 0.0001999896809691165,
+      "loss": 2.9209,
+      "step": 191
+    },
+    {
+      "epoch": 0.015495117423936729,
+      "grad_norm": 1.0177555084228516,
+      "learning_rate": 0.0001999894529351283,
+      "loss": 2.7852,
+      "step": 192
+    },
+    {
+      "epoch": 0.015575821160519731,
+      "grad_norm": 1.062322974205017,
+      "learning_rate": 0.00019998922240914159,
+      "loss": 2.8328,
+      "step": 193
+    },
+    {
+      "epoch": 0.015656524897102737,
+      "grad_norm": 1.0937334299087524,
+      "learning_rate": 0.00019998898939116205,
+      "loss": 2.8069,
+      "step": 194
+    },
+    {
+      "epoch": 0.015737228633685738,
+      "grad_norm": 0.9553198218345642,
+      "learning_rate": 0.00019998875388119554,
+      "loss": 2.8402,
+      "step": 195
+    },
+    {
+      "epoch": 0.015817932370268743,
+      "grad_norm": 1.1802356243133545,
+      "learning_rate": 0.0001999885158792479,
+      "loss": 2.945,
+      "step": 196
+    },
+    {
+      "epoch": 0.015898636106851747,
+      "grad_norm": 1.160346269607544,
+      "learning_rate": 0.0001999882753853251,
+      "loss": 2.8341,
+      "step": 197
+    },
+    {
+      "epoch": 0.01597933984343475,
+      "grad_norm": 1.0379278659820557,
+      "learning_rate": 0.00019998803239943305,
+      "loss": 2.898,
+      "step": 198
+    },
+    {
+      "epoch": 0.016060043580017756,
+      "grad_norm": 1.2022395133972168,
+      "learning_rate": 0.00019998778692157792,
+      "loss": 2.8302,
+      "step": 199
+    },
+    {
+      "epoch": 0.01614074731660076,
+      "grad_norm": 1.057017207145691,
+      "learning_rate": 0.00019998753895176575,
+      "loss": 2.8474,
+      "step": 200
+    },
+    {
+      "epoch": 0.01622145105318376,
+      "grad_norm": 0.9299072027206421,
+      "learning_rate": 0.00019998728849000271,
+      "loss": 2.8266,
+      "step": 201
+    },
+    {
+      "epoch": 0.016302154789766765,
+      "grad_norm": 1.0296592712402344,
+      "learning_rate": 0.00019998703553629512,
+      "loss": 2.8106,
+      "step": 202
+    },
+    {
+      "epoch": 0.01638285852634977,
+      "grad_norm": 0.9641671180725098,
+      "learning_rate": 0.0001999867800906492,
+      "loss": 2.8089,
+      "step": 203
+    },
+    {
+      "epoch": 0.016463562262932774,
+      "grad_norm": 0.9951125383377075,
+      "learning_rate": 0.00019998652215307136,
+      "loss": 2.813,
+      "step": 204
+    },
+    {
+      "epoch": 0.016544265999515778,
+      "grad_norm": 1.0089969635009766,
+      "learning_rate": 0.00019998626172356804,
+      "loss": 2.8021,
+      "step": 205
+    },
+    {
+      "epoch": 0.016624969736098782,
+      "grad_norm": 0.9916231632232666,
+      "learning_rate": 0.00019998599880214566,
+      "loss": 2.8455,
+      "step": 206
+    },
+    {
+      "epoch": 0.016705673472681787,
+      "grad_norm": 0.9612492322921753,
+      "learning_rate": 0.00019998573338881088,
+      "loss": 2.8653,
+      "step": 207
+    },
+    {
+      "epoch": 0.016786377209264788,
+      "grad_norm": 0.984578013420105,
+      "learning_rate": 0.00019998546548357022,
+      "loss": 2.8359,
+      "step": 208
+    },
+    {
+      "epoch": 0.016867080945847792,
+      "grad_norm": 0.9457565546035767,
+      "learning_rate": 0.0001999851950864304,
+      "loss": 2.8507,
+      "step": 209
+    },
+    {
+      "epoch": 0.016947784682430796,
+      "grad_norm": 1.0219026803970337,
+      "learning_rate": 0.00019998492219739817,
+      "loss": 2.8326,
+      "step": 210
+    },
+    {
+      "epoch": 0.0170284884190138,
+      "grad_norm": 0.971570611000061,
+      "learning_rate": 0.00019998464681648032,
+      "loss": 2.8079,
+      "step": 211
+    },
+    {
+      "epoch": 0.017109192155596805,
+      "grad_norm": 0.9731320738792419,
+      "learning_rate": 0.00019998436894368368,
+      "loss": 2.8536,
+      "step": 212
+    },
+    {
+      "epoch": 0.01718989589217981,
+      "grad_norm": 1.0519105195999146,
+      "learning_rate": 0.00019998408857901525,
+      "loss": 2.8589,
+      "step": 213
+    },
+    {
+      "epoch": 0.01727059962876281,
+      "grad_norm": 0.9725883603096008,
+      "learning_rate": 0.00019998380572248194,
+      "loss": 2.7937,
+      "step": 214
+    },
+    {
+      "epoch": 0.017351303365345815,
+      "grad_norm": 1.0397064685821533,
+      "learning_rate": 0.00019998352037409084,
+      "loss": 2.9145,
+      "step": 215
+    },
+    {
+      "epoch": 0.01743200710192882,
+      "grad_norm": 0.9094852209091187,
+      "learning_rate": 0.00019998323253384904,
+      "loss": 2.7692,
+      "step": 216
+    },
+    {
+      "epoch": 0.017512710838511823,
+      "grad_norm": 0.941646158695221,
+      "learning_rate": 0.00019998294220176374,
+      "loss": 2.7975,
+      "step": 217
+    },
+    {
+      "epoch": 0.017593414575094828,
+      "grad_norm": 0.9939892888069153,
+      "learning_rate": 0.00019998264937784216,
+      "loss": 2.8421,
+      "step": 218
+    },
+    {
+      "epoch": 0.017674118311677832,
+      "grad_norm": 0.8985795378684998,
+      "learning_rate": 0.0001999823540620916,
+      "loss": 2.8146,
+      "step": 219
+    },
+    {
+      "epoch": 0.017754822048260833,
+      "grad_norm": 1.0436078310012817,
+      "learning_rate": 0.00019998205625451943,
+      "loss": 2.8416,
+      "step": 220
+    },
+    {
+      "epoch": 0.017835525784843837,
+      "grad_norm": 0.9941675066947937,
+      "learning_rate": 0.00019998175595513305,
+      "loss": 2.8723,
+      "step": 221
+    },
+    {
+      "epoch": 0.01791622952142684,
+      "grad_norm": 0.9203903675079346,
+      "learning_rate": 0.00019998145316393995,
+      "loss": 2.7791,
+      "step": 222
+    },
+    {
+      "epoch": 0.017996933258009846,
+      "grad_norm": 0.9325969815254211,
+      "learning_rate": 0.00019998114788094768,
+      "loss": 2.8664,
+      "step": 223
+    },
+    {
+      "epoch": 0.01807763699459285,
+      "grad_norm": 0.9483599662780762,
+      "learning_rate": 0.00019998084010616388,
+      "loss": 2.7782,
+      "step": 224
+    },
+    {
+      "epoch": 0.018158340731175854,
+      "grad_norm": 0.9555078744888306,
+      "learning_rate": 0.00019998052983959615,
+      "loss": 2.7771,
+      "step": 225
+    },
+    {
+      "epoch": 0.01823904446775886,
+      "grad_norm": 0.9452421069145203,
+      "learning_rate": 0.00019998021708125233,
+      "loss": 2.8878,
+      "step": 226
+    },
+    {
+      "epoch": 0.01831974820434186,
+      "grad_norm": 0.9784894585609436,
+      "learning_rate": 0.00019997990183114007,
+      "loss": 2.8382,
+      "step": 227
+    },
+    {
+      "epoch": 0.018400451940924864,
+      "grad_norm": 1.0844931602478027,
+      "learning_rate": 0.00019997958408926735,
+      "loss": 2.8015,
+      "step": 228
+    },
+    {
+      "epoch": 0.01848115567750787,
+      "grad_norm": 1.0416710376739502,
+      "learning_rate": 0.00019997926385564207,
+      "loss": 2.8364,
+      "step": 229
+    },
+    {
+      "epoch": 0.018561859414090873,
+      "grad_norm": 0.9213813543319702,
+      "learning_rate": 0.00019997894113027215,
+      "loss": 2.8489,
+      "step": 230
+    },
+    {
+      "epoch": 0.018642563150673877,
+      "grad_norm": 1.0186388492584229,
+      "learning_rate": 0.00019997861591316567,
+      "loss": 2.914,
+      "step": 231
+    },
+    {
+      "epoch": 0.01872326688725688,
+      "grad_norm": 1.0032236576080322,
+      "learning_rate": 0.00019997828820433072,
+      "loss": 2.8733,
+      "step": 232
+    },
+    {
+      "epoch": 0.018803970623839882,
+      "grad_norm": 0.9783569574356079,
+      "learning_rate": 0.0001999779580037755,
+      "loss": 2.851,
+      "step": 233
+    },
+    {
+      "epoch": 0.018884674360422887,
+      "grad_norm": 0.8471441268920898,
+      "learning_rate": 0.00019997762531150825,
+      "loss": 2.7923,
+      "step": 234
+    },
+    {
+      "epoch": 0.01896537809700589,
+      "grad_norm": 0.8912937641143799,
+      "learning_rate": 0.00019997729012753717,
+      "loss": 2.8725,
+      "step": 235
+    },
+    {
+      "epoch": 0.019046081833588895,
+      "grad_norm": 1.2453325986862183,
+      "learning_rate": 0.00019997695245187075,
+      "loss": 2.9292,
+      "step": 236
+    },
+    {
+      "epoch": 0.0191267855701719,
+      "grad_norm": 0.8870908617973328,
+      "learning_rate": 0.0001999766122845173,
+      "loss": 2.8008,
+      "step": 237
+    },
+    {
+      "epoch": 0.019207489306754904,
+      "grad_norm": 1.0679768323898315,
+      "learning_rate": 0.0001999762696254853,
+      "loss": 2.8919,
+      "step": 238
+    },
+    {
+      "epoch": 0.01928819304333791,
+      "grad_norm": 0.9769917130470276,
+      "learning_rate": 0.00019997592447478337,
+      "loss": 2.7937,
+      "step": 239
+    },
+    {
+      "epoch": 0.01936889677992091,
+      "grad_norm": 1.066183090209961,
+      "learning_rate": 0.00019997557683242004,
+      "loss": 2.8375,
+      "step": 240
+    },
+    {
+      "epoch": 0.019449600516503913,
+      "grad_norm": 0.9834103584289551,
+      "learning_rate": 0.000199975226698404,
+      "loss": 2.8577,
+      "step": 241
+    },
+    {
+      "epoch": 0.019530304253086918,
+      "grad_norm": 1.102211833000183,
+      "learning_rate": 0.00019997487407274396,
+      "loss": 2.8466,
+      "step": 242
+    },
+    {
+      "epoch": 0.019611007989669922,
+      "grad_norm": 0.9936226606369019,
+      "learning_rate": 0.00019997451895544872,
+      "loss": 2.7729,
+      "step": 243
+    },
+    {
+      "epoch": 0.019691711726252926,
+      "grad_norm": 1.0995992422103882,
+      "learning_rate": 0.00019997416134652713,
+      "loss": 2.8425,
+      "step": 244
+    },
+    {
+      "epoch": 0.01977241546283593,
+      "grad_norm": 0.94181889295578,
+      "learning_rate": 0.00019997380124598814,
+      "loss": 2.8495,
+      "step": 245
+    },
+    {
+      "epoch": 0.01985311919941893,
+      "grad_norm": 0.9791487455368042,
+      "learning_rate": 0.00019997343865384067,
+      "loss": 2.8919,
+      "step": 246
+    },
+    {
+      "epoch": 0.019933822936001936,
+      "grad_norm": 0.9173399209976196,
+      "learning_rate": 0.00019997307357009375,
+      "loss": 2.8593,
+      "step": 247
+    },
+    {
+      "epoch": 0.02001452667258494,
+      "grad_norm": 0.9675281047821045,
+      "learning_rate": 0.00019997270599475653,
+      "loss": 2.8226,
+      "step": 248
+    },
+    {
+      "epoch": 0.020095230409167945,
+      "grad_norm": 0.8928244113922119,
+      "learning_rate": 0.00019997233592783812,
+      "loss": 2.8296,
+      "step": 249
+    },
+    {
+      "epoch": 0.02017593414575095,
+      "grad_norm": 0.928601861000061,
+      "learning_rate": 0.0001999719633693478,
+      "loss": 2.8399,
+      "step": 250
+    },
+    {
+      "epoch": 0.020256637882333953,
+      "grad_norm": 0.9378123879432678,
+      "learning_rate": 0.00019997158831929482,
+      "loss": 2.8711,
+      "step": 251
+    },
+    {
+      "epoch": 0.020337341618916954,
+      "grad_norm": 0.9041047692298889,
+      "learning_rate": 0.00019997121077768853,
+      "loss": 2.8338,
+      "step": 252
+    },
+    {
+      "epoch": 0.02041804535549996,
+      "grad_norm": 0.9673274755477905,
+      "learning_rate": 0.00019997083074453832,
+      "loss": 2.8556,
+      "step": 253
+    },
+    {
+      "epoch": 0.020498749092082963,
+      "grad_norm": 0.9204083681106567,
+      "learning_rate": 0.0001999704482198537,
+      "loss": 2.7954,
+      "step": 254
+    },
+    {
+      "epoch": 0.020579452828665967,
+      "grad_norm": 0.9267606735229492,
+      "learning_rate": 0.00019997006320364417,
+      "loss": 2.8656,
+      "step": 255
+    },
+    {
+      "epoch": 0.02066015656524897,
+      "grad_norm": 0.9562919735908508,
+      "learning_rate": 0.00019996967569591936,
+      "loss": 2.8406,
+      "step": 256
+    },
+    {
+      "epoch": 0.020740860301831976,
+      "grad_norm": 0.9065950512886047,
+      "learning_rate": 0.0001999692856966889,
+      "loss": 2.7856,
+      "step": 257
+    },
+    {
+      "epoch": 0.02082156403841498,
+      "grad_norm": 0.9136463403701782,
+      "learning_rate": 0.0001999688932059625,
+      "loss": 2.8083,
+      "step": 258
+    },
+    {
+      "epoch": 0.02090226777499798,
+      "grad_norm": 0.9785570502281189,
+      "learning_rate": 0.00019996849822374998,
+      "loss": 2.7984,
+      "step": 259
+    },
+    {
+      "epoch": 0.020982971511580985,
+      "grad_norm": 0.9549168348312378,
+      "learning_rate": 0.00019996810075006117,
+      "loss": 2.8048,
+      "step": 260
+    },
+    {
+      "epoch": 0.02106367524816399,
+      "grad_norm": 0.8923975825309753,
+      "learning_rate": 0.00019996770078490594,
+      "loss": 2.8559,
+      "step": 261
+    },
+    {
+      "epoch": 0.021144378984746994,
+      "grad_norm": 0.9516206383705139,
+      "learning_rate": 0.0001999672983282943,
+      "loss": 2.9171,
+      "step": 262
+    },
+    {
+      "epoch": 0.02122508272133,
+      "grad_norm": 0.9101666808128357,
+      "learning_rate": 0.0001999668933802363,
+      "loss": 2.8746,
+      "step": 263
+    },
+    {
+      "epoch": 0.021305786457913003,
+      "grad_norm": 0.9081267714500427,
+      "learning_rate": 0.00019996648594074195,
+      "loss": 2.8637,
+      "step": 264
+    },
+    {
+      "epoch": 0.021386490194496004,
+      "grad_norm": 1.0048178434371948,
+      "learning_rate": 0.0001999660760098215,
+      "loss": 2.8783,
+      "step": 265
+    },
+    {
+      "epoch": 0.021467193931079008,
+      "grad_norm": 0.9625924229621887,
+      "learning_rate": 0.0001999656635874851,
+      "loss": 2.8226,
+      "step": 266
+    },
+    {
+      "epoch": 0.021547897667662012,
+      "grad_norm": 0.9911805391311646,
+      "learning_rate": 0.00019996524867374306,
+      "loss": 2.8135,
+      "step": 267
+    },
+    {
+      "epoch": 0.021628601404245017,
+      "grad_norm": 0.8920134902000427,
+      "learning_rate": 0.00019996483126860572,
+      "loss": 2.7934,
+      "step": 268
+    },
+    {
+      "epoch": 0.02170930514082802,
+      "grad_norm": 1.0806514024734497,
+      "learning_rate": 0.00019996441137208346,
+      "loss": 2.8435,
+      "step": 269
+    },
+    {
+      "epoch": 0.021790008877411025,
+      "grad_norm": 0.9426547884941101,
+      "learning_rate": 0.00019996398898418675,
+      "loss": 2.7919,
+      "step": 270
+    },
+    {
+      "epoch": 0.021870712613994026,
+      "grad_norm": 0.9893020987510681,
+      "learning_rate": 0.00019996356410492615,
+      "loss": 2.8616,
+      "step": 271
+    },
+    {
+      "epoch": 0.02195141635057703,
+      "grad_norm": 1.0196046829223633,
+      "learning_rate": 0.00019996313673431218,
+      "loss": 2.8101,
+      "step": 272
+    },
+    {
+      "epoch": 0.022032120087160035,
+      "grad_norm": 0.9556699991226196,
+      "learning_rate": 0.00019996270687235558,
+      "loss": 2.8669,
+      "step": 273
+    },
+    {
+      "epoch": 0.02211282382374304,
+      "grad_norm": 0.8985902667045593,
+      "learning_rate": 0.00019996227451906702,
+      "loss": 2.8078,
+      "step": 274
+    },
+    {
+      "epoch": 0.022193527560326044,
+      "grad_norm": 1.0198246240615845,
+      "learning_rate": 0.00019996183967445726,
+      "loss": 2.8314,
+      "step": 275
+    },
+    {
+      "epoch": 0.022274231296909048,
+      "grad_norm": 0.9360179901123047,
+      "learning_rate": 0.00019996140233853715,
+      "loss": 2.7969,
+      "step": 276
+    },
+    {
+      "epoch": 0.022354935033492052,
+      "grad_norm": 1.0250160694122314,
+      "learning_rate": 0.00019996096251131759,
+      "loss": 2.7897,
+      "step": 277
+    },
+    {
+      "epoch": 0.022435638770075053,
+      "grad_norm": 0.934582531452179,
+      "learning_rate": 0.00019996052019280954,
+      "loss": 2.8667,
+      "step": 278
+    },
+    {
+      "epoch": 0.022516342506658057,
+      "grad_norm": 0.9394461512565613,
+      "learning_rate": 0.00019996007538302407,
+      "loss": 2.7681,
+      "step": 279
+    },
+    {
+      "epoch": 0.022597046243241062,
+      "grad_norm": 0.9468861222267151,
+      "learning_rate": 0.00019995962808197216,
+      "loss": 2.7709,
+      "step": 280
+    },
+    {
+      "epoch": 0.022677749979824066,
+      "grad_norm": 0.9798515439033508,
+      "learning_rate": 0.00019995917828966506,
+      "loss": 2.8274,
+      "step": 281
+    },
+    {
+      "epoch": 0.02275845371640707,
+      "grad_norm": 1.0403941869735718,
+      "learning_rate": 0.00019995872600611395,
+      "loss": 2.8897,
+      "step": 282
+    },
+    {
+      "epoch": 0.022839157452990075,
+      "grad_norm": 0.9795030951499939,
+      "learning_rate": 0.00019995827123133006,
+      "loss": 2.8792,
+      "step": 283
+    },
+    {
+      "epoch": 0.022919861189573076,
+      "grad_norm": 0.9162538647651672,
+      "learning_rate": 0.00019995781396532479,
+      "loss": 2.8339,
+      "step": 284
+    },
+    {
+      "epoch": 0.02300056492615608,
+      "grad_norm": 1.0864707231521606,
+      "learning_rate": 0.00019995735420810947,
+      "loss": 2.8599,
+      "step": 285
+    },
+    {
+      "epoch": 0.023081268662739084,
+      "grad_norm": 0.9181776642799377,
+      "learning_rate": 0.0001999568919596956,
+      "loss": 2.8736,
+      "step": 286
+    },
+    {
+      "epoch": 0.02316197239932209,
+      "grad_norm": 0.8880531191825867,
+      "learning_rate": 0.00019995642722009472,
+      "loss": 2.8215,
+      "step": 287
+    },
+    {
+      "epoch": 0.023242676135905093,
+      "grad_norm": 0.9287240505218506,
+      "learning_rate": 0.00019995595998931835,
+      "loss": 2.844,
+      "step": 288
+    },
+    {
+      "epoch": 0.023323379872488097,
+      "grad_norm": 0.886894941329956,
+      "learning_rate": 0.0001999554902673782,
+      "loss": 2.8319,
+      "step": 289
+    },
+    {
+      "epoch": 0.0234040836090711,
+      "grad_norm": 0.9564458131790161,
+      "learning_rate": 0.0001999550180542859,
+      "loss": 2.8126,
+      "step": 290
+    },
+    {
+      "epoch": 0.023484787345654103,
+      "grad_norm": 0.8745970726013184,
+      "learning_rate": 0.00019995454335005334,
+      "loss": 2.8344,
+      "step": 291
+    },
+    {
+      "epoch": 0.023565491082237107,
+      "grad_norm": 1.0343137979507446,
+      "learning_rate": 0.00019995406615469217,
+      "loss": 2.8498,
+      "step": 292
+    },
+    {
+      "epoch": 0.02364619481882011,
+      "grad_norm": 0.9951575994491577,
+      "learning_rate": 0.0001999535864682145,
+      "loss": 2.8655,
+      "step": 293
+    },
+    {
+      "epoch": 0.023726898555403116,
+      "grad_norm": 0.8457592725753784,
+      "learning_rate": 0.0001999531042906321,
+      "loss": 2.8189,
+      "step": 294
+    },
+    {
+      "epoch": 0.02380760229198612,
+      "grad_norm": 0.9126954674720764,
+      "learning_rate": 0.00019995261962195708,
+      "loss": 2.8272,
+      "step": 295
+    },
+    {
+      "epoch": 0.023888306028569124,
+      "grad_norm": 1.0171937942504883,
+      "learning_rate": 0.0001999521324622015,
+      "loss": 2.869,
+      "step": 296
+    },
+    {
+      "epoch": 0.023969009765152125,
+      "grad_norm": 0.9887226223945618,
+      "learning_rate": 0.00019995164281137753,
+      "loss": 2.7643,
+      "step": 297
+    },
+    {
+      "epoch": 0.02404971350173513,
+      "grad_norm": 1.4240798950195312,
+      "learning_rate": 0.00019995115066949733,
+      "loss": 2.8332,
+      "step": 298
+    },
+    {
+      "epoch": 0.024130417238318134,
+      "grad_norm": 0.9856921434402466,
+      "learning_rate": 0.00019995065603657316,
+      "loss": 2.8283,
+      "step": 299
+    },
+    {
+      "epoch": 0.024211120974901138,
+      "grad_norm": 0.997164785861969,
+      "learning_rate": 0.0001999501589126174,
+      "loss": 2.9164,
+      "step": 300
+    },
+    {
+      "epoch": 0.024291824711484142,
+      "grad_norm": 1.6480412483215332,
+      "learning_rate": 0.00019994965929764238,
+      "loss": 2.8941,
+      "step": 301
+    },
+    {
+      "epoch": 0.024372528448067147,
+      "grad_norm": 1.1590758562088013,
+      "learning_rate": 0.0001999491571916606,
+      "loss": 2.8127,
+      "step": 302
+    },
+    {
+      "epoch": 0.024453232184650148,
+      "grad_norm": 1.1228376626968384,
+      "learning_rate": 0.00019994865259468454,
+      "loss": 2.8439,
+      "step": 303
+    },
+    {
+      "epoch": 0.024533935921233152,
+      "grad_norm": 1.0426349639892578,
+      "learning_rate": 0.0001999481455067268,
+      "loss": 2.8671,
+      "step": 304
+    },
+    {
+      "epoch": 0.024614639657816156,
+      "grad_norm": 1.0911917686462402,
+      "learning_rate": 0.00019994763592779996,
+      "loss": 2.8297,
+      "step": 305
+    },
+    {
+      "epoch": 0.02469534339439916,
+      "grad_norm": 1.0493195056915283,
+      "learning_rate": 0.00019994712385791683,
+      "loss": 2.7996,
+      "step": 306
+    },
+    {
+      "epoch": 0.024776047130982165,
+      "grad_norm": 0.9275023341178894,
+      "learning_rate": 0.00019994660929709008,
+      "loss": 2.7949,
+      "step": 307
+    },
+    {
+      "epoch": 0.02485675086756517,
+      "grad_norm": 1.1074799299240112,
+      "learning_rate": 0.00019994609224533255,
+      "loss": 2.8364,
+      "step": 308
+    },
+    {
+      "epoch": 0.024937454604148174,
+      "grad_norm": 0.9189429879188538,
+      "learning_rate": 0.00019994557270265717,
+      "loss": 2.8293,
+      "step": 309
+    },
+    {
+      "epoch": 0.025018158340731175,
+      "grad_norm": 0.9577780961990356,
+      "learning_rate": 0.00019994505066907683,
+      "loss": 2.8295,
+      "step": 310
+    },
+    {
+      "epoch": 0.02509886207731418,
+      "grad_norm": 1.0707277059555054,
+      "learning_rate": 0.0001999445261446046,
+      "loss": 2.795,
+      "step": 311
+    },
+    {
+      "epoch": 0.025179565813897183,
+      "grad_norm": 0.9211257696151733,
+      "learning_rate": 0.0001999439991292535,
+      "loss": 2.8355,
+      "step": 312
+    },
+    {
+      "epoch": 0.025260269550480188,
+      "grad_norm": 0.987779438495636,
+      "learning_rate": 0.00019994346962303667,
+      "loss": 2.8175,
+      "step": 313
+    },
+    {
+      "epoch": 0.025340973287063192,
+      "grad_norm": 0.9317128658294678,
+      "learning_rate": 0.00019994293762596734,
+      "loss": 2.8205,
+      "step": 314
+    },
+    {
+      "epoch": 0.025421677023646196,
+      "grad_norm": 0.8989154100418091,
+      "learning_rate": 0.00019994240313805873,
+      "loss": 2.8257,
+      "step": 315
+    },
+    {
+      "epoch": 0.025502380760229197,
+      "grad_norm": 0.8391042351722717,
+      "learning_rate": 0.00019994186615932423,
+      "loss": 2.8105,
+      "step": 316
+    },
+    {
+      "epoch": 0.0255830844968122,
+      "grad_norm": 0.8908089995384216,
+      "learning_rate": 0.00019994132668977715,
+      "loss": 2.7894,
+      "step": 317
+    },
+    {
+      "epoch": 0.025663788233395206,
+      "grad_norm": 0.8666881322860718,
+      "learning_rate": 0.00019994078472943097,
+      "loss": 2.7934,
+      "step": 318
+    },
+    {
+      "epoch": 0.02574449196997821,
+      "grad_norm": 0.8834616541862488,
+      "learning_rate": 0.00019994024027829914,
+      "loss": 2.8166,
+      "step": 319
+    },
+    {
+      "epoch": 0.025825195706561214,
+      "grad_norm": 0.9831370115280151,
+      "learning_rate": 0.00019993969333639532,
+      "loss": 2.889,
+      "step": 320
+    },
+    {
+      "epoch": 0.02590589944314422,
+      "grad_norm": 0.9171644449234009,
+      "learning_rate": 0.00019993914390373308,
+      "loss": 2.8582,
+      "step": 321
+    },
+    {
+      "epoch": 0.02598660317972722,
+      "grad_norm": 0.9624861478805542,
+      "learning_rate": 0.00019993859198032615,
+      "loss": 2.8574,
+      "step": 322
+    },
+    {
+      "epoch": 0.026067306916310224,
+      "grad_norm": 0.8826586008071899,
+      "learning_rate": 0.00019993803756618826,
+      "loss": 2.8544,
+      "step": 323
+    },
+    {
+      "epoch": 0.02614801065289323,
+      "grad_norm": 0.9286447763442993,
+      "learning_rate": 0.0001999374806613332,
+      "loss": 2.7937,
+      "step": 324
+    },
+    {
+      "epoch": 0.026228714389476233,
+      "grad_norm": 0.9901685118675232,
+      "learning_rate": 0.00019993692126577493,
+      "loss": 2.7654,
+      "step": 325
+    },
+    {
+      "epoch": 0.026309418126059237,
+      "grad_norm": 0.9624341130256653,
+      "learning_rate": 0.00019993635937952734,
+      "loss": 2.8804,
+      "step": 326
+    },
+    {
+      "epoch": 0.02639012186264224,
+      "grad_norm": 0.8867596387863159,
+      "learning_rate": 0.0001999357950026044,
+      "loss": 2.8254,
+      "step": 327
+    },
+    {
+      "epoch": 0.026470825599225246,
+      "grad_norm": 0.9243817925453186,
+      "learning_rate": 0.00019993522813502022,
+      "loss": 2.8177,
+      "step": 328
+    },
+    {
+      "epoch": 0.026551529335808247,
+      "grad_norm": 0.9322247505187988,
+      "learning_rate": 0.00019993465877678895,
+      "loss": 2.9023,
+      "step": 329
+    },
+    {
+      "epoch": 0.02663223307239125,
+      "grad_norm": 0.8768174648284912,
+      "learning_rate": 0.00019993408692792474,
+      "loss": 2.8184,
+      "step": 330
+    },
+    {
+      "epoch": 0.026712936808974255,
+      "grad_norm": 0.9436870813369751,
+      "learning_rate": 0.00019993351258844184,
+      "loss": 2.8319,
+      "step": 331
+    },
+    {
+      "epoch": 0.02679364054555726,
+      "grad_norm": 0.9970327019691467,
+      "learning_rate": 0.0001999329357583546,
+      "loss": 2.7946,
+      "step": 332
+    },
+    {
+      "epoch": 0.026874344282140264,
+      "grad_norm": 0.9100088477134705,
+      "learning_rate": 0.00019993235643767736,
+      "loss": 2.782,
+      "step": 333
+    },
+    {
+      "epoch": 0.02695504801872327,
+      "grad_norm": 0.9693402051925659,
+      "learning_rate": 0.00019993177462642456,
+      "loss": 2.8182,
+      "step": 334
+    },
+    {
+      "epoch": 0.02703575175530627,
+      "grad_norm": 0.8761965036392212,
+      "learning_rate": 0.00019993119032461073,
+      "loss": 2.8058,
+      "step": 335
+    },
+    {
+      "epoch": 0.027116455491889273,
+      "grad_norm": 1.0699270963668823,
+      "learning_rate": 0.00019993060353225043,
+      "loss": 2.9211,
+      "step": 336
+    },
+    {
+      "epoch": 0.027197159228472278,
+      "grad_norm": 1.0094172954559326,
+      "learning_rate": 0.00019993001424935822,
+      "loss": 2.8837,
+      "step": 337
+    },
+    {
+      "epoch": 0.027277862965055282,
+      "grad_norm": 0.9683573842048645,
+      "learning_rate": 0.00019992942247594887,
+      "loss": 2.8523,
+      "step": 338
+    },
+    {
+      "epoch": 0.027358566701638286,
+      "grad_norm": 1.3243813514709473,
+      "learning_rate": 0.00019992882821203708,
+      "loss": 2.7891,
+      "step": 339
+    },
+    {
+      "epoch": 0.02743927043822129,
+      "grad_norm": 1.0227056741714478,
+      "learning_rate": 0.0001999282314576377,
+      "loss": 2.8396,
+      "step": 340
+    },
+    {
+      "epoch": 0.027519974174804295,
+      "grad_norm": 1.03257417678833,
+      "learning_rate": 0.00019992763221276556,
+      "loss": 2.824,
+      "step": 341
+    },
+    {
+      "epoch": 0.027600677911387296,
+      "grad_norm": 0.86456698179245,
+      "learning_rate": 0.00019992703047743562,
+      "loss": 2.8006,
+      "step": 342
+    },
+    {
+      "epoch": 0.0276813816479703,
+      "grad_norm": 0.965339720249176,
+      "learning_rate": 0.00019992642625166286,
+      "loss": 2.8658,
+      "step": 343
+    },
+    {
+      "epoch": 0.027762085384553305,
+      "grad_norm": 1.0028942823410034,
+      "learning_rate": 0.00019992581953546236,
+      "loss": 2.8311,
+      "step": 344
+    },
+    {
+      "epoch": 0.02784278912113631,
+      "grad_norm": 0.984307050704956,
+      "learning_rate": 0.0001999252103288492,
+      "loss": 2.8748,
+      "step": 345
+    },
+    {
+      "epoch": 0.027923492857719313,
+      "grad_norm": 0.9405032396316528,
+      "learning_rate": 0.00019992459863183858,
+      "loss": 2.8371,
+      "step": 346
+    },
+    {
+      "epoch": 0.028004196594302318,
+      "grad_norm": 0.9867002367973328,
+      "learning_rate": 0.0001999239844444458,
+      "loss": 2.7914,
+      "step": 347
+    },
+    {
+      "epoch": 0.02808490033088532,
+      "grad_norm": 0.9224951267242432,
+      "learning_rate": 0.00019992336776668613,
+      "loss": 2.7986,
+      "step": 348
+    },
+    {
+      "epoch": 0.028165604067468323,
+      "grad_norm": 1.002838134765625,
+      "learning_rate": 0.0001999227485985749,
+      "loss": 2.8207,
+      "step": 349
+    },
+    {
+      "epoch": 0.028246307804051327,
+      "grad_norm": 0.8922045826911926,
+      "learning_rate": 0.00019992212694012757,
+      "loss": 2.8264,
+      "step": 350
+    },
+    {
+      "epoch": 0.02832701154063433,
+      "grad_norm": 1.0860323905944824,
+      "learning_rate": 0.00019992150279135964,
+      "loss": 2.8778,
+      "step": 351
+    },
+    {
+      "epoch": 0.028407715277217336,
+      "grad_norm": 1.0995604991912842,
+      "learning_rate": 0.0001999208761522867,
+      "loss": 2.8599,
+      "step": 352
+    },
+    {
+      "epoch": 0.02848841901380034,
+      "grad_norm": 0.8741658926010132,
+      "learning_rate": 0.0001999202470229243,
+      "loss": 2.7757,
+      "step": 353
+    },
+    {
+      "epoch": 0.02856912275038334,
+      "grad_norm": 0.9142587184906006,
+      "learning_rate": 0.00019991961540328815,
+      "loss": 2.8235,
+      "step": 354
+    },
+    {
+      "epoch": 0.028649826486966345,
+      "grad_norm": 1.0000953674316406,
+      "learning_rate": 0.000199918981293394,
+      "loss": 2.8,
+      "step": 355
+    },
+    {
+      "epoch": 0.02873053022354935,
+      "grad_norm": 0.9416046738624573,
+      "learning_rate": 0.00019991834469325763,
+      "loss": 2.7941,
+      "step": 356
+    },
+    {
+      "epoch": 0.028811233960132354,
+      "grad_norm": 0.9135935306549072,
+      "learning_rate": 0.00019991770560289496,
+      "loss": 2.8315,
+      "step": 357
+    },
+    {
+      "epoch": 0.02889193769671536,
+      "grad_norm": 0.8867244124412537,
+      "learning_rate": 0.00019991706402232184,
+      "loss": 2.8649,
+      "step": 358
+    },
+    {
+      "epoch": 0.028972641433298363,
+      "grad_norm": 0.9360243678092957,
+      "learning_rate": 0.00019991641995155431,
+      "loss": 2.7556,
+      "step": 359
+    },
+    {
+      "epoch": 0.029053345169881367,
+      "grad_norm": 0.8903766870498657,
+      "learning_rate": 0.00019991577339060842,
+      "loss": 2.8379,
+      "step": 360
+    },
+    {
+      "epoch": 0.029134048906464368,
+      "grad_norm": 1.0178784132003784,
+      "learning_rate": 0.00019991512433950023,
+      "loss": 2.8045,
+      "step": 361
+    },
+    {
+      "epoch": 0.029214752643047372,
+      "grad_norm": 0.9318631887435913,
+      "learning_rate": 0.000199914472798246,
+      "loss": 2.823,
+      "step": 362
+    },
+    {
+      "epoch": 0.029295456379630377,
+      "grad_norm": 0.9384647011756897,
+      "learning_rate": 0.00019991381876686195,
+      "loss": 2.9379,
+      "step": 363
+    },
+    {
+      "epoch": 0.02937616011621338,
+      "grad_norm": 0.9318633675575256,
+      "learning_rate": 0.00019991316224536433,
+      "loss": 2.8222,
+      "step": 364
+    },
+    {
+      "epoch": 0.029456863852796385,
+      "grad_norm": 0.8653938174247742,
+      "learning_rate": 0.00019991250323376952,
+      "loss": 2.8447,
+      "step": 365
+    },
+    {
+      "epoch": 0.02953756758937939,
+      "grad_norm": 0.8997991681098938,
+      "learning_rate": 0.00019991184173209398,
+      "loss": 2.8523,
+      "step": 366
+    },
+    {
+      "epoch": 0.02961827132596239,
+      "grad_norm": 0.8587092161178589,
+      "learning_rate": 0.00019991117774035416,
+      "loss": 2.8141,
+      "step": 367
+    },
+    {
+      "epoch": 0.029698975062545395,
+      "grad_norm": 0.8740741014480591,
+      "learning_rate": 0.00019991051125856663,
+      "loss": 2.7487,
+      "step": 368
+    },
+    {
+      "epoch": 0.0297796787991284,
+      "grad_norm": 0.9099416732788086,
+      "learning_rate": 0.00019990984228674798,
+      "loss": 2.834,
+      "step": 369
+    },
+    {
+      "epoch": 0.029860382535711404,
+      "grad_norm": 0.8675365447998047,
+      "learning_rate": 0.0001999091708249149,
+      "loss": 2.8259,
+      "step": 370
+    },
+    {
+      "epoch": 0.029941086272294408,
+      "grad_norm": 1.0141092538833618,
+      "learning_rate": 0.00019990849687308412,
+      "loss": 2.8369,
+      "step": 371
+    },
+    {
+      "epoch": 0.030021790008877412,
+      "grad_norm": 0.849155604839325,
+      "learning_rate": 0.00019990782043127243,
+      "loss": 2.7505,
+      "step": 372
+    },
+    {
+      "epoch": 0.030102493745460413,
+      "grad_norm": 1.073754072189331,
+      "learning_rate": 0.0001999071414994967,
+      "loss": 2.8939,
+      "step": 373
+    },
+    {
+      "epoch": 0.030183197482043417,
+      "grad_norm": 0.8615279197692871,
+      "learning_rate": 0.00019990646007777383,
+      "loss": 2.7662,
+      "step": 374
+    },
+    {
+      "epoch": 0.030263901218626422,
+      "grad_norm": 0.8803398609161377,
+      "learning_rate": 0.0001999057761661208,
+      "loss": 2.7992,
+      "step": 375
+    },
+    {
+      "epoch": 0.030344604955209426,
+      "grad_norm": 0.8901834487915039,
+      "learning_rate": 0.00019990508976455473,
+      "loss": 2.8222,
+      "step": 376
+    },
+    {
+      "epoch": 0.03042530869179243,
+      "grad_norm": 0.9443284869194031,
+      "learning_rate": 0.00019990440087309263,
+      "loss": 2.8326,
+      "step": 377
+    },
+    {
+      "epoch": 0.030506012428375435,
+      "grad_norm": 0.9122868180274963,
+      "learning_rate": 0.0001999037094917517,
+      "loss": 2.7653,
+      "step": 378
+    },
+    {
+      "epoch": 0.03058671616495844,
+      "grad_norm": 0.8764635920524597,
+      "learning_rate": 0.0001999030156205492,
+      "loss": 2.7813,
+      "step": 379
+    },
+    {
+      "epoch": 0.03066741990154144,
+      "grad_norm": 0.8466865420341492,
+      "learning_rate": 0.0001999023192595024,
+      "loss": 2.8338,
+      "step": 380
+    },
+    {
+      "epoch": 0.030748123638124444,
+      "grad_norm": 0.8833961486816406,
+      "learning_rate": 0.00019990162040862863,
+      "loss": 2.78,
+      "step": 381
+    },
+    {
+      "epoch": 0.03082882737470745,
+      "grad_norm": 1.0298357009887695,
+      "learning_rate": 0.00019990091906794537,
+      "loss": 2.8059,
+      "step": 382
+    },
+    {
+      "epoch": 0.030909531111290453,
+      "grad_norm": 0.8651318550109863,
+      "learning_rate": 0.00019990021523747005,
+      "loss": 2.8608,
+      "step": 383
+    },
+    {
+      "epoch": 0.030990234847873457,
+      "grad_norm": 1.0262864828109741,
+      "learning_rate": 0.0001998995089172202,
+      "loss": 2.8226,
+      "step": 384
+    },
+    {
+      "epoch": 0.03107093858445646,
+      "grad_norm": 0.9266276955604553,
+      "learning_rate": 0.00019989880010721348,
+      "loss": 2.9414,
+      "step": 385
+    },
+    {
+      "epoch": 0.031151642321039463,
+      "grad_norm": 0.8762117028236389,
+      "learning_rate": 0.00019989808880746749,
+      "loss": 2.8023,
+      "step": 386
+    },
+    {
+      "epoch": 0.031232346057622467,
+      "grad_norm": 0.8531816601753235,
+      "learning_rate": 0.00019989737501800004,
+      "loss": 2.777,
+      "step": 387
+    },
+    {
+      "epoch": 0.031313049794205475,
+      "grad_norm": 0.8999545574188232,
+      "learning_rate": 0.0001998966587388288,
+      "loss": 2.8656,
+      "step": 388
+    },
+    {
+      "epoch": 0.03139375353078847,
+      "grad_norm": 0.932248055934906,
+      "learning_rate": 0.00019989593996997177,
+      "loss": 2.8212,
+      "step": 389
+    },
+    {
+      "epoch": 0.031474457267371476,
+      "grad_norm": 0.9059134125709534,
+      "learning_rate": 0.00019989521871144672,
+      "loss": 2.7945,
+      "step": 390
+    },
+    {
+      "epoch": 0.03155516100395448,
+      "grad_norm": 0.9323028922080994,
+      "learning_rate": 0.00019989449496327172,
+      "loss": 2.8338,
+      "step": 391
+    },
+    {
+      "epoch": 0.031635864740537485,
+      "grad_norm": 0.9141251444816589,
+      "learning_rate": 0.0001998937687254648,
+      "loss": 2.7935,
+      "step": 392
+    },
+    {
+      "epoch": 0.03171656847712049,
+      "grad_norm": 1.0026880502700806,
+      "learning_rate": 0.000199893039998044,
+      "loss": 2.8811,
+      "step": 393
+    },
+    {
+      "epoch": 0.031797272213703494,
+      "grad_norm": 1.0178622007369995,
+      "learning_rate": 0.00019989230878102756,
+      "loss": 2.9003,
+      "step": 394
+    },
+    {
+      "epoch": 0.0318779759502865,
+      "grad_norm": 0.9111912846565247,
+      "learning_rate": 0.00019989157507443363,
+      "loss": 2.8399,
+      "step": 395
+    },
+    {
+      "epoch": 0.0319586796868695,
+      "grad_norm": 1.054563283920288,
+      "learning_rate": 0.00019989083887828052,
+      "loss": 2.9088,
+      "step": 396
+    },
+    {
+      "epoch": 0.03203938342345251,
+      "grad_norm": 0.9459816217422485,
+      "learning_rate": 0.00019989010019258663,
+      "loss": 2.805,
+      "step": 397
+    },
+    {
+      "epoch": 0.03212008716003551,
+      "grad_norm": 1.0139873027801514,
+      "learning_rate": 0.00019988935901737033,
+      "loss": 2.8452,
+      "step": 398
+    },
+    {
+      "epoch": 0.032200790896618516,
+      "grad_norm": 0.986325204372406,
+      "learning_rate": 0.00019988861535265006,
+      "loss": 2.8311,
+      "step": 399
+    },
+    {
+      "epoch": 0.03228149463320152,
+      "grad_norm": 0.9565223455429077,
+      "learning_rate": 0.00019988786919844436,
+      "loss": 2.7766,
+      "step": 400
+    },
+    {
+      "epoch": 0.032362198369784524,
+      "grad_norm": 0.8901559710502625,
+      "learning_rate": 0.0001998871205547719,
+      "loss": 2.7966,
+      "step": 401
+    },
+    {
+      "epoch": 0.03244290210636752,
+      "grad_norm": 1.0959528684616089,
+      "learning_rate": 0.00019988636942165123,
+      "loss": 2.8377,
+      "step": 402
+    },
+    {
+      "epoch": 0.032523605842950526,
+      "grad_norm": 1.0768988132476807,
+      "learning_rate": 0.00019988561579910118,
+      "loss": 2.8267,
+      "step": 403
+    },
+    {
+      "epoch": 0.03260430957953353,
+      "grad_norm": 0.9563855528831482,
+      "learning_rate": 0.00019988485968714048,
+      "loss": 2.8459,
+      "step": 404
+    },
+    {
+      "epoch": 0.032685013316116535,
+      "grad_norm": 0.930927038192749,
+      "learning_rate": 0.00019988410108578796,
+      "loss": 2.8053,
+      "step": 405
+    },
+    {
+      "epoch": 0.03276571705269954,
+      "grad_norm": 1.0658363103866577,
+      "learning_rate": 0.00019988333999506255,
+      "loss": 2.8512,
+      "step": 406
+    },
+    {
+      "epoch": 0.03284642078928254,
+      "grad_norm": 0.9258090257644653,
+      "learning_rate": 0.0001998825764149832,
+      "loss": 2.8541,
+      "step": 407
+    },
+    {
+      "epoch": 0.03292712452586555,
+      "grad_norm": 1.18158757686615,
+      "learning_rate": 0.00019988181034556895,
+      "loss": 2.8838,
+      "step": 408
+    },
+    {
+      "epoch": 0.03300782826244855,
+      "grad_norm": 0.9506754875183105,
+      "learning_rate": 0.00019988104178683891,
+      "loss": 2.7733,
+      "step": 409
+    },
+    {
+      "epoch": 0.033088531999031556,
+      "grad_norm": 0.9559460282325745,
+      "learning_rate": 0.0001998802707388122,
+      "loss": 2.9259,
+      "step": 410
+    },
+    {
+      "epoch": 0.03316923573561456,
+      "grad_norm": 0.9322298765182495,
+      "learning_rate": 0.00019987949720150808,
+      "loss": 2.8318,
+      "step": 411
+    },
+    {
+      "epoch": 0.033249939472197565,
+      "grad_norm": 0.9226691722869873,
+      "learning_rate": 0.00019987872117494576,
+      "loss": 2.9063,
+      "step": 412
+    },
+    {
+      "epoch": 0.03333064320878057,
+      "grad_norm": 1.0543674230575562,
+      "learning_rate": 0.00019987794265914464,
+      "loss": 2.7877,
+      "step": 413
+    },
+    {
+      "epoch": 0.033411346945363574,
+      "grad_norm": 0.989986002445221,
+      "learning_rate": 0.00019987716165412408,
+      "loss": 2.8354,
+      "step": 414
+    },
+    {
+      "epoch": 0.03349205068194657,
+      "grad_norm": 0.8703451752662659,
+      "learning_rate": 0.0001998763781599036,
+      "loss": 2.8127,
+      "step": 415
+    },
+    {
+      "epoch": 0.033572754418529575,
+      "grad_norm": 0.974943220615387,
+      "learning_rate": 0.0001998755921765027,
+      "loss": 2.9272,
+      "step": 416
+    },
+    {
+      "epoch": 0.03365345815511258,
+      "grad_norm": 0.8714169859886169,
+      "learning_rate": 0.000199874803703941,
+      "loss": 2.8027,
+      "step": 417
+    },
+    {
+      "epoch": 0.033734161891695584,
+      "grad_norm": 0.9251161217689514,
+      "learning_rate": 0.00019987401274223804,
+      "loss": 2.8186,
+      "step": 418
+    },
+    {
+      "epoch": 0.03381486562827859,
+      "grad_norm": 0.9657236933708191,
+      "learning_rate": 0.00019987321929141366,
+      "loss": 2.8297,
+      "step": 419
+    },
+    {
+      "epoch": 0.03389556936486159,
+      "grad_norm": 0.9022002816200256,
+      "learning_rate": 0.00019987242335148757,
+      "loss": 2.881,
+      "step": 420
+    },
+    {
+      "epoch": 0.0339762731014446,
+      "grad_norm": 0.9479621052742004,
+      "learning_rate": 0.0001998716249224796,
+      "loss": 2.8288,
+      "step": 421
+    },
+    {
+      "epoch": 0.0340569768380276,
+      "grad_norm": 0.9458955526351929,
+      "learning_rate": 0.00019987082400440968,
+      "loss": 2.8861,
+      "step": 422
+    },
+    {
+      "epoch": 0.034137680574610606,
+      "grad_norm": 0.9444572329521179,
+      "learning_rate": 0.0001998700205972978,
+      "loss": 2.8877,
+      "step": 423
+    },
+    {
+      "epoch": 0.03421838431119361,
+      "grad_norm": 0.9263925552368164,
+      "learning_rate": 0.00019986921470116392,
+      "loss": 2.8028,
+      "step": 424
+    },
+    {
+      "epoch": 0.034299088047776614,
+      "grad_norm": 1.0690566301345825,
+      "learning_rate": 0.00019986840631602812,
+      "loss": 2.882,
+      "step": 425
+    },
+    {
+      "epoch": 0.03437979178435962,
+      "grad_norm": 0.8999007940292358,
+      "learning_rate": 0.0001998675954419106,
+      "loss": 2.8179,
+      "step": 426
+    },
+    {
+      "epoch": 0.03446049552094262,
+      "grad_norm": 0.894395112991333,
+      "learning_rate": 0.00019986678207883153,
+      "loss": 2.814,
+      "step": 427
+    },
+    {
+      "epoch": 0.03454119925752562,
+      "grad_norm": 0.8621550798416138,
+      "learning_rate": 0.00019986596622681123,
+      "loss": 2.7584,
+      "step": 428
+    },
+    {
+      "epoch": 0.034621902994108625,
+      "grad_norm": 0.9452527165412903,
+      "learning_rate": 0.00019986514788587,
+      "loss": 2.8949,
+      "step": 429
+    },
+    {
+      "epoch": 0.03470260673069163,
+      "grad_norm": 0.8973272442817688,
+      "learning_rate": 0.0001998643270560282,
+      "loss": 2.868,
+      "step": 430
+    },
+    {
+      "epoch": 0.034783310467274633,
+      "grad_norm": 0.9887418150901794,
+      "learning_rate": 0.00019986350373730634,
+      "loss": 2.8009,
+      "step": 431
+    },
+    {
+      "epoch": 0.03486401420385764,
+      "grad_norm": 0.9449994564056396,
+      "learning_rate": 0.0001998626779297249,
+      "loss": 2.8305,
+      "step": 432
+    },
+    {
+      "epoch": 0.03494471794044064,
+      "grad_norm": 1.052871823310852,
+      "learning_rate": 0.0001998618496333045,
+      "loss": 2.8136,
+      "step": 433
+    },
+    {
+      "epoch": 0.035025421677023647,
+      "grad_norm": 0.9600724577903748,
+      "learning_rate": 0.00019986101884806576,
+      "loss": 2.7857,
+      "step": 434
+    },
+    {
+      "epoch": 0.03510612541360665,
+      "grad_norm": 0.874043345451355,
+      "learning_rate": 0.00019986018557402942,
+      "loss": 2.8524,
+      "step": 435
+    },
+    {
+      "epoch": 0.035186829150189655,
+      "grad_norm": 0.9810616374015808,
+      "learning_rate": 0.0001998593498112162,
+      "loss": 2.7506,
+      "step": 436
+    },
+    {
+      "epoch": 0.03526753288677266,
+      "grad_norm": 0.9163016080856323,
+      "learning_rate": 0.00019985851155964693,
+      "loss": 2.798,
+      "step": 437
+    },
+    {
+      "epoch": 0.035348236623355664,
+      "grad_norm": 1.0688380002975464,
+      "learning_rate": 0.00019985767081934252,
+      "loss": 2.8916,
+      "step": 438
+    },
+    {
+      "epoch": 0.03542894035993867,
+      "grad_norm": 0.925020158290863,
+      "learning_rate": 0.00019985682759032393,
+      "loss": 2.8017,
+      "step": 439
+    },
+    {
+      "epoch": 0.035509644096521666,
+      "grad_norm": 0.9429430961608887,
+      "learning_rate": 0.0001998559818726122,
+      "loss": 2.837,
+      "step": 440
+    },
+    {
+      "epoch": 0.03559034783310467,
+      "grad_norm": 0.9135627150535583,
+      "learning_rate": 0.00019985513366622832,
+      "loss": 2.8423,
+      "step": 441
+    },
+    {
+      "epoch": 0.035671051569687674,
+      "grad_norm": 0.9218924045562744,
+      "learning_rate": 0.00019985428297119353,
+      "loss": 2.854,
+      "step": 442
+    },
+    {
+      "epoch": 0.03575175530627068,
+      "grad_norm": 0.9307878613471985,
+      "learning_rate": 0.00019985342978752897,
+      "loss": 2.8591,
+      "step": 443
+    },
+    {
+      "epoch": 0.03583245904285368,
+      "grad_norm": 0.935394287109375,
+      "learning_rate": 0.00019985257411525592,
+      "loss": 2.8388,
+      "step": 444
+    },
+    {
+      "epoch": 0.03591316277943669,
+      "grad_norm": 0.890959620475769,
+      "learning_rate": 0.0001998517159543957,
+      "loss": 2.78,
+      "step": 445
+    },
+    {
+      "epoch": 0.03599386651601969,
+      "grad_norm": 1.110924482345581,
+      "learning_rate": 0.0001998508553049697,
+      "loss": 2.8117,
+      "step": 446
+    },
+    {
+      "epoch": 0.036074570252602696,
+      "grad_norm": 0.8774176239967346,
+      "learning_rate": 0.0001998499921669994,
+      "loss": 2.8368,
+      "step": 447
+    },
+    {
+      "epoch": 0.0361552739891857,
+      "grad_norm": 0.9766948819160461,
+      "learning_rate": 0.00019984912654050625,
+      "loss": 2.764,
+      "step": 448
+    },
+    {
+      "epoch": 0.036235977725768705,
+      "grad_norm": 1.1439398527145386,
+      "learning_rate": 0.00019984825842551187,
+      "loss": 2.84,
+      "step": 449
+    },
+    {
+      "epoch": 0.03631668146235171,
+      "grad_norm": 0.8995118737220764,
+      "learning_rate": 0.0001998473878220379,
+      "loss": 2.834,
+      "step": 450
+    },
+    {
+      "epoch": 0.03639738519893471,
+      "grad_norm": 0.9810060858726501,
+      "learning_rate": 0.000199846514730106,
+      "loss": 2.9338,
+      "step": 451
+    },
+    {
+      "epoch": 0.03647808893551772,
+      "grad_norm": 1.0862053632736206,
+      "learning_rate": 0.00019984563914973795,
+      "loss": 2.837,
+      "step": 452
+    },
+    {
+      "epoch": 0.036558792672100715,
+      "grad_norm": 0.9456702470779419,
+      "learning_rate": 0.0001998447610809556,
+      "loss": 2.7664,
+      "step": 453
+    },
+    {
+      "epoch": 0.03663949640868372,
+      "grad_norm": 1.0714432001113892,
+      "learning_rate": 0.0001998438805237808,
+      "loss": 2.8339,
+      "step": 454
+    },
+    {
+      "epoch": 0.036720200145266724,
+      "grad_norm": 0.89134281873703,
+      "learning_rate": 0.00019984299747823547,
+      "loss": 2.7818,
+      "step": 455
+    },
+    {
+      "epoch": 0.03680090388184973,
+      "grad_norm": 0.869742214679718,
+      "learning_rate": 0.0001998421119443417,
+      "loss": 2.7916,
+      "step": 456
+    },
+    {
+      "epoch": 0.03688160761843273,
+      "grad_norm": 0.9307265281677246,
+      "learning_rate": 0.00019984122392212149,
+      "loss": 2.8485,
+      "step": 457
+    },
+    {
+      "epoch": 0.03696231135501574,
+      "grad_norm": 0.900215744972229,
+      "learning_rate": 0.00019984033341159698,
+      "loss": 2.8536,
+      "step": 458
+    },
+    {
+      "epoch": 0.03704301509159874,
+      "grad_norm": 0.8679699897766113,
+      "learning_rate": 0.00019983944041279038,
+      "loss": 2.8344,
+      "step": 459
+    },
+    {
+      "epoch": 0.037123718828181745,
+      "grad_norm": 0.9540488719940186,
+      "learning_rate": 0.00019983854492572394,
+      "loss": 2.873,
+      "step": 460
+    },
+    {
+      "epoch": 0.03720442256476475,
+      "grad_norm": 0.8697962760925293,
+      "learning_rate": 0.00019983764695042,
+      "loss": 2.8122,
+      "step": 461
+    },
+    {
+      "epoch": 0.037285126301347754,
+      "grad_norm": 0.9534483551979065,
+      "learning_rate": 0.0001998367464869009,
+      "loss": 2.8842,
+      "step": 462
+    },
+    {
+      "epoch": 0.03736583003793076,
+      "grad_norm": 0.8402275443077087,
+      "learning_rate": 0.00019983584353518911,
+      "loss": 2.8135,
+      "step": 463
+    },
+    {
+      "epoch": 0.03744653377451376,
+      "grad_norm": 0.8226146697998047,
+      "learning_rate": 0.0001998349380953071,
+      "loss": 2.8036,
+      "step": 464
+    },
+    {
+      "epoch": 0.03752723751109677,
+      "grad_norm": 0.9292199611663818,
+      "learning_rate": 0.0001998340301672775,
+      "loss": 2.7887,
+      "step": 465
+    },
+    {
+      "epoch": 0.037607941247679764,
+      "grad_norm": 0.9035555124282837,
+      "learning_rate": 0.0001998331197511229,
+      "loss": 2.7851,
+      "step": 466
+    },
+    {
+      "epoch": 0.03768864498426277,
+      "grad_norm": 0.9411706328392029,
+      "learning_rate": 0.00019983220684686596,
+      "loss": 2.7782,
+      "step": 467
+    },
+    {
+      "epoch": 0.03776934872084577,
+      "grad_norm": 0.9867696166038513,
+      "learning_rate": 0.0001998312914545295,
+      "loss": 2.8125,
+      "step": 468
+    },
+    {
+      "epoch": 0.03785005245742878,
+      "grad_norm": 0.9683675169944763,
+      "learning_rate": 0.00019983037357413624,
+      "loss": 2.8325,
+      "step": 469
+    },
+    {
+      "epoch": 0.03793075619401178,
+      "grad_norm": 0.963941752910614,
+      "learning_rate": 0.00019982945320570913,
+      "loss": 2.8281,
+      "step": 470
+    },
+    {
+      "epoch": 0.038011459930594786,
+      "grad_norm": 0.9812459349632263,
+      "learning_rate": 0.0001998285303492711,
+      "loss": 2.765,
+      "step": 471
+    },
+    {
+      "epoch": 0.03809216366717779,
+      "grad_norm": 0.9681405425071716,
+      "learning_rate": 0.00019982760500484516,
+      "loss": 2.8882,
+      "step": 472
+    },
+    {
+      "epoch": 0.038172867403760795,
+      "grad_norm": 0.8983948826789856,
+      "learning_rate": 0.00019982667717245432,
+      "loss": 2.8182,
+      "step": 473
+    },
+    {
+      "epoch": 0.0382535711403438,
+      "grad_norm": 0.9875261783599854,
+      "learning_rate": 0.00019982574685212178,
+      "loss": 2.8072,
+      "step": 474
+    },
+    {
+      "epoch": 0.038334274876926804,
+      "grad_norm": 0.8889442086219788,
+      "learning_rate": 0.00019982481404387064,
+      "loss": 2.8635,
+      "step": 475
+    },
+    {
+      "epoch": 0.03841497861350981,
+      "grad_norm": 0.8904242515563965,
+      "learning_rate": 0.00019982387874772418,
+      "loss": 2.829,
+      "step": 476
+    },
+    {
+      "epoch": 0.03849568235009281,
+      "grad_norm": 1.0182000398635864,
+      "learning_rate": 0.00019982294096370574,
+      "loss": 2.8552,
+      "step": 477
+    },
+    {
+      "epoch": 0.03857638608667582,
+      "grad_norm": 0.9867151975631714,
+      "learning_rate": 0.00019982200069183867,
+      "loss": 2.8201,
+      "step": 478
+    },
+    {
+      "epoch": 0.038657089823258814,
+      "grad_norm": 0.9785345196723938,
+      "learning_rate": 0.0001998210579321464,
+      "loss": 2.8652,
+      "step": 479
+    },
+    {
+      "epoch": 0.03873779355984182,
+      "grad_norm": 0.9696915149688721,
+      "learning_rate": 0.00019982011268465243,
+      "loss": 2.8276,
+      "step": 480
+    },
+    {
+      "epoch": 0.03881849729642482,
+      "grad_norm": 0.9257470965385437,
+      "learning_rate": 0.00019981916494938033,
+      "loss": 2.8321,
+      "step": 481
+    },
+    {
+      "epoch": 0.03889920103300783,
+      "grad_norm": 0.9394895434379578,
+      "learning_rate": 0.00019981821472635369,
+      "loss": 2.8747,
+      "step": 482
+    },
+    {
+      "epoch": 0.03897990476959083,
+      "grad_norm": 0.9888504147529602,
+      "learning_rate": 0.00019981726201559626,
+      "loss": 2.8201,
+      "step": 483
+    },
+    {
+      "epoch": 0.039060608506173836,
+      "grad_norm": 0.8957003951072693,
+      "learning_rate": 0.0001998163068171317,
+      "loss": 2.8255,
+      "step": 484
+    },
+    {
+      "epoch": 0.03914131224275684,
+      "grad_norm": 0.9792008996009827,
+      "learning_rate": 0.00019981534913098383,
+      "loss": 2.7985,
+      "step": 485
+    },
+    {
+      "epoch": 0.039222015979339844,
+      "grad_norm": 0.8689060211181641,
+      "learning_rate": 0.00019981438895717656,
+      "loss": 2.7945,
+      "step": 486
+    },
+    {
+      "epoch": 0.03930271971592285,
+      "grad_norm": 0.9932593703269958,
+      "learning_rate": 0.0001998134262957338,
+      "loss": 2.9041,
+      "step": 487
+    },
+    {
+      "epoch": 0.03938342345250585,
+      "grad_norm": 0.8496069312095642,
+      "learning_rate": 0.00019981246114667955,
+      "loss": 2.8433,
+      "step": 488
+    },
+    {
+      "epoch": 0.03946412718908886,
+      "grad_norm": 0.8484126925468445,
+      "learning_rate": 0.00019981149351003786,
+      "loss": 2.7872,
+      "step": 489
+    },
+    {
+      "epoch": 0.03954483092567186,
+      "grad_norm": 0.9208858013153076,
+      "learning_rate": 0.00019981052338583283,
+      "loss": 2.7776,
+      "step": 490
+    },
+    {
+      "epoch": 0.03962553466225486,
+      "grad_norm": 0.9305418729782104,
+      "learning_rate": 0.00019980955077408865,
+      "loss": 2.7851,
+      "step": 491
+    },
+    {
+      "epoch": 0.03970623839883786,
+      "grad_norm": 0.9803212881088257,
+      "learning_rate": 0.00019980857567482955,
+      "loss": 2.8469,
+      "step": 492
+    },
+    {
+      "epoch": 0.03978694213542087,
+      "grad_norm": 0.9165790677070618,
+      "learning_rate": 0.00019980759808807985,
+      "loss": 2.8513,
+      "step": 493
+    },
+    {
+      "epoch": 0.03986764587200387,
+      "grad_norm": 0.9153794050216675,
+      "learning_rate": 0.00019980661801386393,
+      "loss": 2.8322,
+      "step": 494
+    },
+    {
+      "epoch": 0.039948349608586876,
+      "grad_norm": 0.89347904920578,
+      "learning_rate": 0.00019980563545220616,
+      "loss": 2.8316,
+      "step": 495
+    },
+    {
+      "epoch": 0.04002905334516988,
+      "grad_norm": 0.9882236123085022,
+      "learning_rate": 0.00019980465040313105,
+      "loss": 2.7471,
+      "step": 496
+    },
+    {
+      "epoch": 0.040109757081752885,
+      "grad_norm": 0.9391099810600281,
+      "learning_rate": 0.00019980366286666322,
+      "loss": 2.8182,
+      "step": 497
+    },
+    {
+      "epoch": 0.04019046081833589,
+      "grad_norm": 1.0155293941497803,
+      "learning_rate": 0.00019980267284282717,
+      "loss": 2.8721,
+      "step": 498
+    },
+    {
+      "epoch": 0.040271164554918894,
+      "grad_norm": 0.9952930212020874,
+      "learning_rate": 0.00019980168033164765,
+      "loss": 2.8538,
+      "step": 499
+    },
+    {
+      "epoch": 0.0403518682915019,
+      "grad_norm": 0.8385666608810425,
+      "learning_rate": 0.00019980068533314934,
+      "loss": 2.8242,
+      "step": 500
+    },
+    {
+      "epoch": 0.0404325720280849,
+      "grad_norm": 0.8747559785842896,
+      "learning_rate": 0.0001997996878473571,
+      "loss": 2.7908,
+      "step": 501
+    },
+    {
+      "epoch": 0.04051327576466791,
+      "grad_norm": 0.9267926216125488,
+      "learning_rate": 0.00019979868787429575,
+      "loss": 2.8359,
+      "step": 502
+    },
+    {
+      "epoch": 0.04059397950125091,
+      "grad_norm": 0.8194155693054199,
+      "learning_rate": 0.00019979768541399022,
+      "loss": 2.8161,
+      "step": 503
+    },
+    {
+      "epoch": 0.04067468323783391,
+      "grad_norm": 0.8923258185386658,
+      "learning_rate": 0.00019979668046646548,
+      "loss": 2.7547,
+      "step": 504
+    },
+    {
+      "epoch": 0.04075538697441691,
+      "grad_norm": 0.8965646028518677,
+      "learning_rate": 0.00019979567303174663,
+      "loss": 2.8432,
+      "step": 505
+    },
+    {
+      "epoch": 0.04083609071099992,
+      "grad_norm": 0.814481794834137,
+      "learning_rate": 0.0001997946631098587,
+      "loss": 2.8327,
+      "step": 506
+    },
+    {
+      "epoch": 0.04091679444758292,
+      "grad_norm": 0.8806928396224976,
+      "learning_rate": 0.00019979365070082694,
+      "loss": 2.8573,
+      "step": 507
+    },
+    {
+      "epoch": 0.040997498184165926,
+      "grad_norm": 0.8546919822692871,
+      "learning_rate": 0.00019979263580467653,
+      "loss": 2.8618,
+      "step": 508
+    },
+    {
+      "epoch": 0.04107820192074893,
+      "grad_norm": 0.8557277321815491,
+      "learning_rate": 0.00019979161842143274,
+      "loss": 2.8454,
+      "step": 509
+    },
+    {
+      "epoch": 0.041158905657331935,
+      "grad_norm": 0.9153180122375488,
+      "learning_rate": 0.00019979059855112098,
+      "loss": 2.8027,
+      "step": 510
+    },
+    {
+      "epoch": 0.04123960939391494,
+      "grad_norm": 0.8616741895675659,
+      "learning_rate": 0.00019978957619376666,
+      "loss": 2.7628,
+      "step": 511
+    },
+    {
+      "epoch": 0.04132031313049794,
+      "grad_norm": 0.8777137398719788,
+      "learning_rate": 0.00019978855134939524,
+      "loss": 2.8443,
+      "step": 512
+    },
+    {
+      "epoch": 0.04140101686708095,
+      "grad_norm": 0.852100133895874,
+      "learning_rate": 0.0001997875240180323,
+      "loss": 2.8125,
+      "step": 513
+    },
+    {
+      "epoch": 0.04148172060366395,
+      "grad_norm": 0.8470742702484131,
+      "learning_rate": 0.00019978649419970338,
+      "loss": 2.8139,
+      "step": 514
+    },
+    {
+      "epoch": 0.041562424340246956,
+      "grad_norm": 0.8890305161476135,
+      "learning_rate": 0.0001997854618944342,
+      "loss": 2.8633,
+      "step": 515
+    },
+    {
+      "epoch": 0.04164312807682996,
+      "grad_norm": 0.8893599510192871,
+      "learning_rate": 0.00019978442710225043,
+      "loss": 2.8066,
+      "step": 516
+    },
+    {
+      "epoch": 0.04172383181341296,
+      "grad_norm": 0.9093891382217407,
+      "learning_rate": 0.00019978338982317792,
+      "loss": 2.8026,
+      "step": 517
+    },
+    {
+      "epoch": 0.04180453554999596,
+      "grad_norm": 0.9775434136390686,
+      "learning_rate": 0.00019978235005724252,
+      "loss": 2.849,
+      "step": 518
+    },
+    {
+      "epoch": 0.04188523928657897,
+      "grad_norm": 1.0014091730117798,
+      "learning_rate": 0.00019978130780447012,
+      "loss": 2.8572,
+      "step": 519
+    },
+    {
+      "epoch": 0.04196594302316197,
+      "grad_norm": 0.8487632870674133,
+      "learning_rate": 0.00019978026306488668,
+      "loss": 2.7611,
+      "step": 520
+    },
+    {
+      "epoch": 0.042046646759744975,
+      "grad_norm": 0.86592698097229,
+      "learning_rate": 0.00019977921583851825,
+      "loss": 2.7616,
+      "step": 521
+    },
+    {
+      "epoch": 0.04212735049632798,
+      "grad_norm": 1.0285916328430176,
+      "learning_rate": 0.00019977816612539093,
+      "loss": 2.8049,
+      "step": 522
+    },
+    {
+      "epoch": 0.042208054232910984,
+      "grad_norm": 0.9716495871543884,
+      "learning_rate": 0.00019977711392553092,
+      "loss": 2.8459,
+      "step": 523
+    },
+    {
+      "epoch": 0.04228875796949399,
+      "grad_norm": 0.8842264413833618,
+      "learning_rate": 0.0001997760592389644,
+      "loss": 2.7934,
+      "step": 524
+    },
+    {
+      "epoch": 0.04236946170607699,
+      "grad_norm": 0.8839964866638184,
+      "learning_rate": 0.00019977500206571765,
+      "loss": 2.8135,
+      "step": 525
+    },
+    {
+      "epoch": 0.04245016544266,
+      "grad_norm": 0.870331346988678,
+      "learning_rate": 0.00019977394240581705,
+      "loss": 2.8684,
+      "step": 526
+    },
+    {
+      "epoch": 0.042530869179243,
+      "grad_norm": 0.8844720125198364,
+      "learning_rate": 0.000199772880259289,
+      "loss": 2.7867,
+      "step": 527
+    },
+    {
+      "epoch": 0.042611572915826006,
+      "grad_norm": 0.9353455901145935,
+      "learning_rate": 0.00019977181562615994,
+      "loss": 2.8051,
+      "step": 528
+    },
+    {
+      "epoch": 0.04269227665240901,
+      "grad_norm": 0.9530816078186035,
+      "learning_rate": 0.00019977074850645646,
+      "loss": 2.7915,
+      "step": 529
+    },
+    {
+      "epoch": 0.04277298038899201,
+      "grad_norm": 0.8984190821647644,
+      "learning_rate": 0.00019976967890020507,
+      "loss": 2.7957,
+      "step": 530
+    },
+    {
+      "epoch": 0.04285368412557501,
+      "grad_norm": 0.9146613478660583,
+      "learning_rate": 0.00019976860680743252,
+      "loss": 2.9053,
+      "step": 531
+    },
+    {
+      "epoch": 0.042934387862158016,
+      "grad_norm": 0.9228026866912842,
+      "learning_rate": 0.0001997675322281655,
+      "loss": 2.8578,
+      "step": 532
+    },
+    {
+      "epoch": 0.04301509159874102,
+      "grad_norm": 0.8266343474388123,
+      "learning_rate": 0.0001997664551624308,
+      "loss": 2.7393,
+      "step": 533
+    },
+    {
+      "epoch": 0.043095795335324025,
+      "grad_norm": 0.9197628498077393,
+      "learning_rate": 0.0001997653756102552,
+      "loss": 2.8828,
+      "step": 534
+    },
+    {
+      "epoch": 0.04317649907190703,
+      "grad_norm": 0.9145991802215576,
+      "learning_rate": 0.00019976429357166566,
+      "loss": 2.7767,
+      "step": 535
+    },
+    {
+      "epoch": 0.04325720280849003,
+      "grad_norm": 0.9123281240463257,
+      "learning_rate": 0.00019976320904668913,
+      "loss": 2.7993,
+      "step": 536
+    },
+    {
+      "epoch": 0.04333790654507304,
+      "grad_norm": 0.8597636818885803,
+      "learning_rate": 0.00019976212203535266,
+      "loss": 2.8148,
+      "step": 537
+    },
+    {
+      "epoch": 0.04341861028165604,
+      "grad_norm": 0.8963296413421631,
+      "learning_rate": 0.00019976103253768334,
+      "loss": 2.7722,
+      "step": 538
+    },
+    {
+      "epoch": 0.043499314018239046,
+      "grad_norm": 0.9480688571929932,
+      "learning_rate": 0.0001997599405537083,
+      "loss": 2.8038,
+      "step": 539
+    },
+    {
+      "epoch": 0.04358001775482205,
+      "grad_norm": 0.8115736842155457,
+      "learning_rate": 0.00019975884608345476,
+      "loss": 2.8069,
+      "step": 540
+    },
+    {
+      "epoch": 0.043660721491405055,
+      "grad_norm": 0.9642506837844849,
+      "learning_rate": 0.00019975774912695,
+      "loss": 2.8703,
+      "step": 541
+    },
+    {
+      "epoch": 0.04374142522798805,
+      "grad_norm": 0.9638697504997253,
+      "learning_rate": 0.0001997566496842214,
+      "loss": 2.8223,
+      "step": 542
+    },
+    {
+      "epoch": 0.04382212896457106,
+      "grad_norm": 0.9478490352630615,
+      "learning_rate": 0.00019975554775529628,
+      "loss": 2.8164,
+      "step": 543
+    },
+    {
+      "epoch": 0.04390283270115406,
+      "grad_norm": 1.1771583557128906,
+      "learning_rate": 0.00019975444334020215,
+      "loss": 2.7969,
+      "step": 544
+    },
+    {
+      "epoch": 0.043983536437737066,
+      "grad_norm": 0.9597339034080505,
+      "learning_rate": 0.00019975333643896655,
+      "loss": 2.8025,
+      "step": 545
+    },
+    {
+      "epoch": 0.04406424017432007,
+      "grad_norm": 0.981595516204834,
+      "learning_rate": 0.00019975222705161704,
+      "loss": 2.7994,
+      "step": 546
+    },
+    {
+      "epoch": 0.044144943910903074,
+      "grad_norm": 0.9581133723258972,
+      "learning_rate": 0.00019975111517818127,
+      "loss": 2.802,
+      "step": 547
+    },
+    {
+      "epoch": 0.04422564764748608,
+      "grad_norm": 0.8643878698348999,
+      "learning_rate": 0.00019975000081868697,
+      "loss": 2.7958,
+      "step": 548
+    },
+    {
+      "epoch": 0.04430635138406908,
+      "grad_norm": 1.2188652753829956,
+      "learning_rate": 0.0001997488839731619,
+      "loss": 2.8786,
+      "step": 549
+    },
+    {
+      "epoch": 0.04438705512065209,
+      "grad_norm": 0.9138071537017822,
+      "learning_rate": 0.00019974776464163387,
+      "loss": 2.809,
+      "step": 550
+    },
+    {
+      "epoch": 0.04446775885723509,
+      "grad_norm": 0.9604587554931641,
+      "learning_rate": 0.00019974664282413083,
+      "loss": 2.8009,
+      "step": 551
+    },
+    {
+      "epoch": 0.044548462593818096,
+      "grad_norm": 1.0271116495132446,
+      "learning_rate": 0.00019974551852068072,
+      "loss": 2.8689,
+      "step": 552
+    },
+    {
+      "epoch": 0.0446291663304011,
+      "grad_norm": 0.9330877065658569,
+      "learning_rate": 0.00019974439173131155,
+      "loss": 2.7613,
+      "step": 553
+    },
+    {
+      "epoch": 0.044709870066984105,
+      "grad_norm": 0.9549325108528137,
+      "learning_rate": 0.00019974326245605136,
+      "loss": 2.8314,
+      "step": 554
+    },
+    {
+      "epoch": 0.0447905738035671,
+      "grad_norm": 0.8928439021110535,
+      "learning_rate": 0.00019974213069492836,
+      "loss": 2.8097,
+      "step": 555
+    },
+    {
+      "epoch": 0.044871277540150106,
+      "grad_norm": 0.8705076575279236,
+      "learning_rate": 0.00019974099644797075,
+      "loss": 2.8112,
+      "step": 556
+    },
+    {
+      "epoch": 0.04495198127673311,
+      "grad_norm": 0.988345742225647,
+      "learning_rate": 0.00019973985971520676,
+      "loss": 2.7648,
+      "step": 557
+    },
+    {
+      "epoch": 0.045032685013316115,
+      "grad_norm": 0.9161957502365112,
+      "learning_rate": 0.00019973872049666475,
+      "loss": 2.8691,
+      "step": 558
+    },
+    {
+      "epoch": 0.04511338874989912,
+      "grad_norm": 0.8404076099395752,
+      "learning_rate": 0.00019973757879237312,
+      "loss": 2.7708,
+      "step": 559
+    },
+    {
+      "epoch": 0.045194092486482124,
+      "grad_norm": 1.05247962474823,
+      "learning_rate": 0.0001997364346023603,
+      "loss": 2.8638,
+      "step": 560
+    },
+    {
+      "epoch": 0.04527479622306513,
+      "grad_norm": 0.9235066175460815,
+      "learning_rate": 0.00019973528792665483,
+      "loss": 2.7876,
+      "step": 561
+    },
+    {
+      "epoch": 0.04535549995964813,
+      "grad_norm": 1.220075249671936,
+      "learning_rate": 0.00019973413876528526,
+      "loss": 2.8563,
+      "step": 562
+    },
+    {
+      "epoch": 0.04543620369623114,
+      "grad_norm": 0.9098384976387024,
+      "learning_rate": 0.00019973298711828025,
+      "loss": 2.8427,
+      "step": 563
+    },
+    {
+      "epoch": 0.04551690743281414,
+      "grad_norm": 0.8792217969894409,
+      "learning_rate": 0.00019973183298566848,
+      "loss": 2.8673,
+      "step": 564
+    },
+    {
+      "epoch": 0.045597611169397145,
+      "grad_norm": 0.9895235896110535,
+      "learning_rate": 0.00019973067636747875,
+      "loss": 2.8262,
+      "step": 565
+    },
+    {
+      "epoch": 0.04567831490598015,
+      "grad_norm": 0.9191479086875916,
+      "learning_rate": 0.00019972951726373984,
+      "loss": 2.8005,
+      "step": 566
+    },
+    {
+      "epoch": 0.045759018642563154,
+      "grad_norm": 0.9631491899490356,
+      "learning_rate": 0.0001997283556744807,
+      "loss": 2.8438,
+      "step": 567
+    },
+    {
+      "epoch": 0.04583972237914615,
+      "grad_norm": 0.8302746415138245,
+      "learning_rate": 0.00019972719159973024,
+      "loss": 2.8221,
+      "step": 568
+    },
+    {
+      "epoch": 0.045920426115729156,
+      "grad_norm": 0.8238534927368164,
+      "learning_rate": 0.00019972602503951748,
+      "loss": 2.7674,
+      "step": 569
+    },
+    {
+      "epoch": 0.04600112985231216,
+      "grad_norm": 0.9675811529159546,
+      "learning_rate": 0.00019972485599387146,
+      "loss": 2.8457,
+      "step": 570
+    },
+    {
+      "epoch": 0.046081833588895164,
+      "grad_norm": 0.8663914203643799,
+      "learning_rate": 0.00019972368446282134,
+      "loss": 2.7851,
+      "step": 571
+    },
+    {
+      "epoch": 0.04616253732547817,
+      "grad_norm": 0.9904592633247375,
+      "learning_rate": 0.00019972251044639636,
+      "loss": 2.8792,
+      "step": 572
+    },
+    {
+      "epoch": 0.04624324106206117,
+      "grad_norm": 0.907600462436676,
+      "learning_rate": 0.0001997213339446257,
+      "loss": 2.7991,
+      "step": 573
+    },
+    {
+      "epoch": 0.04632394479864418,
+      "grad_norm": 0.871362566947937,
+      "learning_rate": 0.00019972015495753876,
+      "loss": 2.7959,
+      "step": 574
+    },
+    {
+      "epoch": 0.04640464853522718,
+      "grad_norm": 0.9664937853813171,
+      "learning_rate": 0.00019971897348516486,
+      "loss": 2.7847,
+      "step": 575
+    },
+    {
+      "epoch": 0.046485352271810186,
+      "grad_norm": 1.0670619010925293,
+      "learning_rate": 0.0001997177895275335,
+      "loss": 2.8864,
+      "step": 576
+    },
+    {
+      "epoch": 0.04656605600839319,
+      "grad_norm": 0.9281025528907776,
+      "learning_rate": 0.00019971660308467414,
+      "loss": 2.8568,
+      "step": 577
+    },
+    {
+      "epoch": 0.046646759744976195,
+      "grad_norm": 0.8964822888374329,
+      "learning_rate": 0.00019971541415661639,
+      "loss": 2.7246,
+      "step": 578
+    },
+    {
+      "epoch": 0.0467274634815592,
+      "grad_norm": 0.8921917676925659,
+      "learning_rate": 0.00019971422274338985,
+      "loss": 2.8513,
+      "step": 579
+    },
+    {
+      "epoch": 0.0468081672181422,
+      "grad_norm": 0.9550159573554993,
+      "learning_rate": 0.0001997130288450242,
+      "loss": 2.7615,
+      "step": 580
+    },
+    {
+      "epoch": 0.0468888709547252,
+      "grad_norm": 0.9330170154571533,
+      "learning_rate": 0.00019971183246154925,
+      "loss": 2.9017,
+      "step": 581
+    },
+    {
+      "epoch": 0.046969574691308205,
+      "grad_norm": 0.9125271439552307,
+      "learning_rate": 0.00019971063359299477,
+      "loss": 2.8263,
+      "step": 582
+    },
+    {
+      "epoch": 0.04705027842789121,
+      "grad_norm": 1.0005927085876465,
+      "learning_rate": 0.00019970943223939066,
+      "loss": 2.8371,
+      "step": 583
+    },
+    {
+      "epoch": 0.047130982164474214,
+      "grad_norm": 1.0333613157272339,
+      "learning_rate": 0.00019970822840076685,
+      "loss": 2.8275,
+      "step": 584
+    },
+    {
+      "epoch": 0.04721168590105722,
+      "grad_norm": 0.8684708476066589,
+      "learning_rate": 0.00019970702207715334,
+      "loss": 2.8343,
+      "step": 585
+    },
+    {
+      "epoch": 0.04729238963764022,
+      "grad_norm": 1.1112761497497559,
+      "learning_rate": 0.00019970581326858025,
+      "loss": 2.9012,
+      "step": 586
+    },
+    {
+      "epoch": 0.04737309337422323,
+      "grad_norm": 1.0187962055206299,
+      "learning_rate": 0.00019970460197507763,
+      "loss": 2.8423,
+      "step": 587
+    },
+    {
+      "epoch": 0.04745379711080623,
+      "grad_norm": 0.9802024960517883,
+      "learning_rate": 0.00019970338819667567,
+      "loss": 2.867,
+      "step": 588
+    },
+    {
+      "epoch": 0.047534500847389236,
+      "grad_norm": 0.9825551509857178,
+      "learning_rate": 0.00019970217193340467,
+      "loss": 2.8359,
+      "step": 589
+    },
+    {
+      "epoch": 0.04761520458397224,
+      "grad_norm": 1.1399210691452026,
+      "learning_rate": 0.00019970095318529494,
+      "loss": 2.8356,
+      "step": 590
+    },
+    {
+      "epoch": 0.047695908320555244,
+      "grad_norm": 1.0373995304107666,
+      "learning_rate": 0.00019969973195237684,
+      "loss": 2.8005,
+      "step": 591
+    },
+    {
+      "epoch": 0.04777661205713825,
+      "grad_norm": 1.133596420288086,
+      "learning_rate": 0.00019969850823468077,
+      "loss": 2.8778,
+      "step": 592
+    },
+    {
+      "epoch": 0.047857315793721246,
+      "grad_norm": 1.0187327861785889,
+      "learning_rate": 0.00019969728203223728,
+      "loss": 2.8291,
+      "step": 593
+    },
+    {
+      "epoch": 0.04793801953030425,
+      "grad_norm": 1.0588128566741943,
+      "learning_rate": 0.00019969605334507688,
+      "loss": 2.9396,
+      "step": 594
+    },
+    {
+      "epoch": 0.048018723266887255,
+      "grad_norm": 0.8783230781555176,
+      "learning_rate": 0.00019969482217323026,
+      "loss": 2.8076,
+      "step": 595
+    },
+    {
+      "epoch": 0.04809942700347026,
+      "grad_norm": 1.0500195026397705,
+      "learning_rate": 0.00019969358851672805,
+      "loss": 2.9099,
+      "step": 596
+    },
+    {
+      "epoch": 0.04818013074005326,
+      "grad_norm": 0.9523593187332153,
+      "learning_rate": 0.000199692352375601,
+      "loss": 2.7448,
+      "step": 597
+    },
+    {
+      "epoch": 0.04826083447663627,
+      "grad_norm": 1.0008500814437866,
+      "learning_rate": 0.00019969111374987995,
+      "loss": 2.8212,
+      "step": 598
+    },
+    {
+      "epoch": 0.04834153821321927,
+      "grad_norm": 0.8992626070976257,
+      "learning_rate": 0.00019968987263959575,
+      "loss": 2.8698,
+      "step": 599
+    },
+    {
+      "epoch": 0.048422241949802276,
+      "grad_norm": 0.9914852380752563,
+      "learning_rate": 0.00019968862904477935,
+      "loss": 2.8221,
+      "step": 600
+    },
+    {
+      "epoch": 0.04850294568638528,
+      "grad_norm": 0.9633241295814514,
+      "learning_rate": 0.00019968738296546168,
+      "loss": 2.8835,
+      "step": 601
+    },
+    {
+      "epoch": 0.048583649422968285,
+      "grad_norm": 1.055831789970398,
+      "learning_rate": 0.00019968613440167387,
+      "loss": 2.8781,
+      "step": 602
+    },
+    {
+      "epoch": 0.04866435315955129,
+      "grad_norm": 0.913856029510498,
+      "learning_rate": 0.000199684883353447,
+      "loss": 2.7863,
+      "step": 603
+    },
+    {
+      "epoch": 0.048745056896134294,
+      "grad_norm": 0.8429243564605713,
+      "learning_rate": 0.00019968362982081226,
+      "loss": 2.7753,
+      "step": 604
+    },
+    {
+      "epoch": 0.0488257606327173,
+      "grad_norm": 0.9324761629104614,
+      "learning_rate": 0.0001996823738038009,
+      "loss": 2.8058,
+      "step": 605
+    },
+    {
+      "epoch": 0.048906464369300295,
+      "grad_norm": 1.0004981756210327,
+      "learning_rate": 0.0001996811153024442,
+      "loss": 2.8537,
+      "step": 606
+    },
+    {
+      "epoch": 0.0489871681058833,
+      "grad_norm": 0.9438043236732483,
+      "learning_rate": 0.00019967985431677354,
+      "loss": 2.8828,
+      "step": 607
+    },
+    {
+      "epoch": 0.049067871842466304,
+      "grad_norm": 0.9359340071678162,
+      "learning_rate": 0.00019967859084682034,
+      "loss": 2.8149,
+      "step": 608
+    },
+    {
+      "epoch": 0.04914857557904931,
+      "grad_norm": 1.0400227308273315,
+      "learning_rate": 0.00019967732489261609,
+      "loss": 2.8489,
+      "step": 609
+    },
+    {
+      "epoch": 0.04922927931563231,
+      "grad_norm": 0.8978031277656555,
+      "learning_rate": 0.00019967605645419237,
+      "loss": 2.8599,
+      "step": 610
+    },
+    {
+      "epoch": 0.04930998305221532,
+      "grad_norm": 0.9982689619064331,
+      "learning_rate": 0.00019967478553158073,
+      "loss": 2.9024,
+      "step": 611
+    },
+    {
+      "epoch": 0.04939068678879832,
+      "grad_norm": 1.0695222616195679,
+      "learning_rate": 0.00019967351212481292,
+      "loss": 2.8483,
+      "step": 612
+    },
+    {
+      "epoch": 0.049471390525381326,
+      "grad_norm": 1.0615525245666504,
+      "learning_rate": 0.0001996722362339206,
+      "loss": 2.806,
+      "step": 613
+    },
+    {
+      "epoch": 0.04955209426196433,
+      "grad_norm": 0.9624890089035034,
+      "learning_rate": 0.0001996709578589356,
+      "loss": 2.8641,
+      "step": 614
+    },
+    {
+      "epoch": 0.049632797998547334,
+      "grad_norm": 0.9156595468521118,
+      "learning_rate": 0.00019966967699988985,
+      "loss": 2.7991,
+      "step": 615
+    },
+    {
+      "epoch": 0.04971350173513034,
+      "grad_norm": 0.8687645196914673,
+      "learning_rate": 0.00019966839365681517,
+      "loss": 2.774,
+      "step": 616
+    },
+    {
+      "epoch": 0.04979420547171334,
+      "grad_norm": 0.9175437688827515,
+      "learning_rate": 0.00019966710782974359,
+      "loss": 2.8064,
+      "step": 617
+    },
+    {
+      "epoch": 0.04987490920829635,
+      "grad_norm": 0.8897463083267212,
+      "learning_rate": 0.00019966581951870715,
+      "loss": 2.8487,
+      "step": 618
+    },
+    {
+      "epoch": 0.049955612944879345,
+      "grad_norm": 0.8908397555351257,
+      "learning_rate": 0.00019966452872373795,
+      "loss": 2.8523,
+      "step": 619
+    },
+    {
+      "epoch": 0.05003631668146235,
+      "grad_norm": 0.95484858751297,
+      "learning_rate": 0.00019966323544486818,
+      "loss": 2.8471,
+      "step": 620
+    },
+    {
+      "epoch": 0.050117020418045354,
+      "grad_norm": 0.9995831251144409,
+      "learning_rate": 0.00019966193968213008,
+      "loss": 2.8341,
+      "step": 621
+    },
+    {
+      "epoch": 0.05019772415462836,
+      "grad_norm": 0.8731706142425537,
+      "learning_rate": 0.00019966064143555587,
+      "loss": 2.8491,
+      "step": 622
+    },
+    {
+      "epoch": 0.05027842789121136,
+      "grad_norm": 0.9213298559188843,
+      "learning_rate": 0.000199659340705178,
+      "loss": 2.8256,
+      "step": 623
+    },
+    {
+      "epoch": 0.050359131627794367,
+      "grad_norm": 0.9565179347991943,
+      "learning_rate": 0.00019965803749102885,
+      "loss": 2.8177,
+      "step": 624
+    },
+    {
+      "epoch": 0.05043983536437737,
+      "grad_norm": 1.0076881647109985,
+      "learning_rate": 0.00019965673179314086,
+      "loss": 2.7812,
+      "step": 625
+    },
+    {
+      "epoch": 0.050520539100960375,
+      "grad_norm": 0.989647388458252,
+      "learning_rate": 0.00019965542361154666,
+      "loss": 2.9226,
+      "step": 626
+    },
+    {
+      "epoch": 0.05060124283754338,
+      "grad_norm": 0.9671580791473389,
+      "learning_rate": 0.00019965411294627878,
+      "loss": 2.8204,
+      "step": 627
+    },
+    {
+      "epoch": 0.050681946574126384,
+      "grad_norm": 0.9275986552238464,
+      "learning_rate": 0.00019965279979736989,
+      "loss": 2.8481,
+      "step": 628
+    },
+    {
+      "epoch": 0.05076265031070939,
+      "grad_norm": 0.9949543476104736,
+      "learning_rate": 0.00019965148416485273,
+      "loss": 2.8606,
+      "step": 629
+    },
+    {
+      "epoch": 0.05084335404729239,
+      "grad_norm": 0.9506482481956482,
+      "learning_rate": 0.0001996501660487601,
+      "loss": 2.8088,
+      "step": 630
+    },
+    {
+      "epoch": 0.0509240577838754,
+      "grad_norm": 0.9147887229919434,
+      "learning_rate": 0.00019964884544912488,
+      "loss": 2.7997,
+      "step": 631
+    },
+    {
+      "epoch": 0.051004761520458394,
+      "grad_norm": 0.8964840769767761,
+      "learning_rate": 0.00019964752236597993,
+      "loss": 2.8342,
+      "step": 632
+    },
+    {
+      "epoch": 0.0510854652570414,
+      "grad_norm": 0.931811511516571,
+      "learning_rate": 0.00019964619679935824,
+      "loss": 2.8229,
+      "step": 633
+    },
+    {
+      "epoch": 0.0511661689936244,
+      "grad_norm": 0.8634423017501831,
+      "learning_rate": 0.00019964486874929282,
+      "loss": 2.803,
+      "step": 634
+    },
+    {
+      "epoch": 0.05124687273020741,
+      "grad_norm": 0.892223596572876,
+      "learning_rate": 0.00019964353821581683,
+      "loss": 2.802,
+      "step": 635
+    },
+    {
+      "epoch": 0.05132757646679041,
+      "grad_norm": 0.8373630046844482,
+      "learning_rate": 0.00019964220519896338,
+      "loss": 2.7693,
+      "step": 636
+    },
+    {
+      "epoch": 0.051408280203373416,
+      "grad_norm": 0.8729730248451233,
+      "learning_rate": 0.0001996408696987657,
+      "loss": 2.8467,
+      "step": 637
+    },
+    {
+      "epoch": 0.05148898393995642,
+      "grad_norm": 0.8994413614273071,
+      "learning_rate": 0.0001996395317152571,
+      "loss": 2.8837,
+      "step": 638
+    },
+    {
+      "epoch": 0.051569687676539425,
+      "grad_norm": 0.9146113395690918,
+      "learning_rate": 0.0001996381912484709,
+      "loss": 2.8189,
+      "step": 639
+    },
+    {
+      "epoch": 0.05165039141312243,
+      "grad_norm": 0.9330562353134155,
+      "learning_rate": 0.00019963684829844052,
+      "loss": 2.7873,
+      "step": 640
+    },
+    {
+      "epoch": 0.05173109514970543,
+      "grad_norm": 0.9076224565505981,
+      "learning_rate": 0.00019963550286519944,
+      "loss": 2.802,
+      "step": 641
+    },
+    {
+      "epoch": 0.05181179888628844,
+      "grad_norm": 0.9580704569816589,
+      "learning_rate": 0.00019963415494878115,
+      "loss": 2.8173,
+      "step": 642
+    },
+    {
+      "epoch": 0.05189250262287144,
+      "grad_norm": 0.9291248917579651,
+      "learning_rate": 0.00019963280454921928,
+      "loss": 2.7866,
+      "step": 643
+    },
+    {
+      "epoch": 0.05197320635945444,
+      "grad_norm": 0.9815296530723572,
+      "learning_rate": 0.0001996314516665475,
+      "loss": 2.7903,
+      "step": 644
+    },
+    {
+      "epoch": 0.052053910096037444,
+      "grad_norm": 0.9461820721626282,
+      "learning_rate": 0.00019963009630079949,
+      "loss": 2.7854,
+      "step": 645
+    },
+    {
+      "epoch": 0.05213461383262045,
+      "grad_norm": 0.9660771489143372,
+      "learning_rate": 0.00019962873845200908,
+      "loss": 2.9187,
+      "step": 646
+    },
+    {
+      "epoch": 0.05221531756920345,
+      "grad_norm": 0.8987802863121033,
+      "learning_rate": 0.00019962737812021002,
+      "loss": 2.8854,
+      "step": 647
+    },
+    {
+      "epoch": 0.05229602130578646,
+      "grad_norm": 0.9810429215431213,
+      "learning_rate": 0.0001996260153054363,
+      "loss": 2.8974,
+      "step": 648
+    },
+    {
+      "epoch": 0.05237672504236946,
+      "grad_norm": 0.8185738325119019,
+      "learning_rate": 0.00019962465000772183,
+      "loss": 2.797,
+      "step": 649
+    },
+    {
+      "epoch": 0.052457428778952465,
+      "grad_norm": 0.8976237773895264,
+      "learning_rate": 0.0001996232822271007,
+      "loss": 2.8557,
+      "step": 650
+    },
+    {
+      "epoch": 0.05253813251553547,
+      "grad_norm": 0.8591496348381042,
+      "learning_rate": 0.0001996219119636069,
+      "loss": 2.8521,
+      "step": 651
+    },
+    {
+      "epoch": 0.052618836252118474,
+      "grad_norm": 0.8907031416893005,
+      "learning_rate": 0.00019962053921727472,
+      "loss": 2.8117,
+      "step": 652
+    },
+    {
+      "epoch": 0.05269953998870148,
+      "grad_norm": 0.9034241437911987,
+      "learning_rate": 0.00019961916398813823,
+      "loss": 2.741,
+      "step": 653
+    },
+    {
+      "epoch": 0.05278024372528448,
+      "grad_norm": 0.8284802436828613,
+      "learning_rate": 0.00019961778627623176,
+      "loss": 2.776,
+      "step": 654
+    },
+    {
+      "epoch": 0.05286094746186749,
+      "grad_norm": 0.8459529876708984,
+      "learning_rate": 0.00019961640608158967,
+      "loss": 2.8027,
+      "step": 655
+    },
+    {
+      "epoch": 0.05294165119845049,
+      "grad_norm": 0.9720042943954468,
+      "learning_rate": 0.00019961502340424636,
+      "loss": 2.9086,
+      "step": 656
+    },
+    {
+      "epoch": 0.05302235493503349,
+      "grad_norm": 0.8581427335739136,
+      "learning_rate": 0.00019961363824423626,
+      "loss": 2.8347,
+      "step": 657
+    },
+    {
+      "epoch": 0.05310305867161649,
+      "grad_norm": 0.9545331597328186,
+      "learning_rate": 0.00019961225060159386,
+      "loss": 2.828,
+      "step": 658
+    },
+    {
+      "epoch": 0.0531837624081995,
+      "grad_norm": 1.0303562879562378,
+      "learning_rate": 0.00019961086047635385,
+      "loss": 2.8461,
+      "step": 659
+    },
+    {
+      "epoch": 0.0532644661447825,
+      "grad_norm": 0.86605304479599,
+      "learning_rate": 0.0001996094678685508,
+      "loss": 2.8355,
+      "step": 660
+    },
+    {
+      "epoch": 0.053345169881365506,
+      "grad_norm": 0.8146334886550903,
+      "learning_rate": 0.0001996080727782194,
+      "loss": 2.8638,
+      "step": 661
+    },
+    {
+      "epoch": 0.05342587361794851,
+      "grad_norm": 0.9434560537338257,
+      "learning_rate": 0.00019960667520539446,
+      "loss": 2.8196,
+      "step": 662
+    },
+    {
+      "epoch": 0.053506577354531515,
+      "grad_norm": 0.9362602829933167,
+      "learning_rate": 0.00019960527515011084,
+      "loss": 2.8452,
+      "step": 663
+    },
+    {
+      "epoch": 0.05358728109111452,
+      "grad_norm": 0.828713059425354,
+      "learning_rate": 0.00019960387261240334,
+      "loss": 2.8079,
+      "step": 664
+    },
+    {
+      "epoch": 0.053667984827697524,
+      "grad_norm": 0.8610214591026306,
+      "learning_rate": 0.00019960246759230697,
+      "loss": 2.8197,
+      "step": 665
+    },
+    {
+      "epoch": 0.05374868856428053,
+      "grad_norm": 0.8913124799728394,
+      "learning_rate": 0.00019960106008985674,
+      "loss": 2.8392,
+      "step": 666
+    },
+    {
+      "epoch": 0.05382939230086353,
+      "grad_norm": 0.8109759092330933,
+      "learning_rate": 0.00019959965010508778,
+      "loss": 2.7961,
+      "step": 667
+    },
+    {
+      "epoch": 0.05391009603744654,
+      "grad_norm": 0.8714832663536072,
+      "learning_rate": 0.00019959823763803514,
+      "loss": 2.7984,
+      "step": 668
+    },
+    {
+      "epoch": 0.05399079977402954,
+      "grad_norm": 0.9008125066757202,
+      "learning_rate": 0.00019959682268873408,
+      "loss": 2.8319,
+      "step": 669
+    },
+    {
+      "epoch": 0.05407150351061254,
+      "grad_norm": 0.8718584775924683,
+      "learning_rate": 0.00019959540525721985,
+      "loss": 2.7973,
+      "step": 670
+    },
+    {
+      "epoch": 0.05415220724719554,
+      "grad_norm": 0.8666327595710754,
+      "learning_rate": 0.00019959398534352774,
+      "loss": 2.8296,
+      "step": 671
+    },
+    {
+      "epoch": 0.05423291098377855,
+      "grad_norm": 0.9755229949951172,
+      "learning_rate": 0.00019959256294769322,
+      "loss": 2.8358,
+      "step": 672
+    },
+    {
+      "epoch": 0.05431361472036155,
+      "grad_norm": 1.193708062171936,
+      "learning_rate": 0.0001995911380697517,
+      "loss": 2.7672,
+      "step": 673
+    },
+    {
+      "epoch": 0.054394318456944556,
+      "grad_norm": 0.9104088544845581,
+      "learning_rate": 0.00019958971070973866,
+      "loss": 2.8389,
+      "step": 674
+    },
+    {
+      "epoch": 0.05447502219352756,
+      "grad_norm": 0.9266251921653748,
+      "learning_rate": 0.0001995882808676897,
+      "loss": 2.8226,
+      "step": 675
+    },
+    {
+      "epoch": 0.054555725930110564,
+      "grad_norm": 1.1161282062530518,
+      "learning_rate": 0.00019958684854364046,
+      "loss": 2.8236,
+      "step": 676
+    },
+    {
+      "epoch": 0.05463642966669357,
+      "grad_norm": 0.9200586080551147,
+      "learning_rate": 0.00019958541373762666,
+      "loss": 2.8074,
+      "step": 677
+    },
+    {
+      "epoch": 0.05471713340327657,
+      "grad_norm": 1.0372560024261475,
+      "learning_rate": 0.000199583976449684,
+      "loss": 2.815,
+      "step": 678
+    },
+    {
+      "epoch": 0.05479783713985958,
+      "grad_norm": 0.8822301030158997,
+      "learning_rate": 0.0001995825366798483,
+      "loss": 2.7985,
+      "step": 679
+    },
+    {
+      "epoch": 0.05487854087644258,
+      "grad_norm": 0.9226076006889343,
+      "learning_rate": 0.00019958109442815553,
+      "loss": 2.7649,
+      "step": 680
+    },
+    {
+      "epoch": 0.054959244613025586,
+      "grad_norm": 0.8769479990005493,
+      "learning_rate": 0.00019957964969464156,
+      "loss": 2.8483,
+      "step": 681
+    },
+    {
+      "epoch": 0.05503994834960859,
+      "grad_norm": 0.8601027727127075,
+      "learning_rate": 0.0001995782024793424,
+      "loss": 2.8072,
+      "step": 682
+    },
+    {
+      "epoch": 0.05512065208619159,
+      "grad_norm": 0.9684911370277405,
+      "learning_rate": 0.00019957675278229416,
+      "loss": 2.8693,
+      "step": 683
+    },
+    {
+      "epoch": 0.05520135582277459,
+      "grad_norm": 0.9119890928268433,
+      "learning_rate": 0.00019957530060353294,
+      "loss": 2.853,
+      "step": 684
+    },
+    {
+      "epoch": 0.055282059559357596,
+      "grad_norm": 0.9588247537612915,
+      "learning_rate": 0.0001995738459430949,
+      "loss": 2.8435,
+      "step": 685
+    },
+    {
+      "epoch": 0.0553627632959406,
+      "grad_norm": 0.8317441940307617,
+      "learning_rate": 0.00019957238880101636,
+      "loss": 2.8208,
+      "step": 686
+    },
+    {
+      "epoch": 0.055443467032523605,
+      "grad_norm": 0.92695152759552,
+      "learning_rate": 0.00019957092917733361,
+      "loss": 2.8378,
+      "step": 687
+    },
+    {
+      "epoch": 0.05552417076910661,
+      "grad_norm": 0.8908315300941467,
+      "learning_rate": 0.00019956946707208305,
+      "loss": 2.8041,
+      "step": 688
+    },
+    {
+      "epoch": 0.055604874505689614,
+      "grad_norm": 0.9787055253982544,
+      "learning_rate": 0.00019956800248530107,
+      "loss": 2.8604,
+      "step": 689
+    },
+    {
+      "epoch": 0.05568557824227262,
+      "grad_norm": 0.8707631826400757,
+      "learning_rate": 0.00019956653541702415,
+      "loss": 2.7763,
+      "step": 690
+    },
+    {
+      "epoch": 0.05576628197885562,
+      "grad_norm": 1.0059715509414673,
+      "learning_rate": 0.00019956506586728896,
+      "loss": 2.8267,
+      "step": 691
+    },
+    {
+      "epoch": 0.05584698571543863,
+      "grad_norm": 0.88490891456604,
+      "learning_rate": 0.00019956359383613203,
+      "loss": 2.8278,
+      "step": 692
+    },
+    {
+      "epoch": 0.05592768945202163,
+      "grad_norm": 0.9527923464775085,
+      "learning_rate": 0.00019956211932359007,
+      "loss": 2.8251,
+      "step": 693
+    },
+    {
+      "epoch": 0.056008393188604635,
+      "grad_norm": 0.9612617492675781,
+      "learning_rate": 0.00019956064232969987,
+      "loss": 2.8148,
+      "step": 694
+    },
+    {
+      "epoch": 0.05608909692518763,
+      "grad_norm": 0.9261285066604614,
+      "learning_rate": 0.0001995591628544982,
+      "loss": 2.8176,
+      "step": 695
+    },
+    {
+      "epoch": 0.05616980066177064,
+      "grad_norm": 0.9766250252723694,
+      "learning_rate": 0.0001995576808980219,
+      "loss": 2.7968,
+      "step": 696
+    },
+    {
+      "epoch": 0.05625050439835364,
+      "grad_norm": 0.9287495017051697,
+      "learning_rate": 0.00019955619646030802,
+      "loss": 2.7679,
+      "step": 697
+    },
+    {
+      "epoch": 0.056331208134936646,
+      "grad_norm": 0.9182924032211304,
+      "learning_rate": 0.00019955470954139345,
+      "loss": 2.8295,
+      "step": 698
+    },
+    {
+      "epoch": 0.05641191187151965,
+      "grad_norm": 0.8650663495063782,
+      "learning_rate": 0.00019955322014131524,
+      "loss": 2.7928,
+      "step": 699
+    },
+    {
+      "epoch": 0.056492615608102655,
+      "grad_norm": 0.9543934464454651,
+      "learning_rate": 0.00019955172826011062,
+      "loss": 2.8049,
+      "step": 700
+    },
+    {
+      "epoch": 0.05657331934468566,
+      "grad_norm": 0.9060636162757874,
+      "learning_rate": 0.00019955023389781664,
+      "loss": 2.871,
+      "step": 701
+    },
+    {
+      "epoch": 0.05665402308126866,
+      "grad_norm": 0.9824137091636658,
+      "learning_rate": 0.00019954873705447065,
+      "loss": 2.816,
+      "step": 702
+    },
+    {
+      "epoch": 0.05673472681785167,
+      "grad_norm": 0.8831053972244263,
+      "learning_rate": 0.00019954723773010988,
+      "loss": 2.8207,
+      "step": 703
+    },
+    {
+      "epoch": 0.05681543055443467,
+      "grad_norm": 0.9603390693664551,
+      "learning_rate": 0.00019954573592477173,
+      "loss": 2.831,
+      "step": 704
+    },
+    {
+      "epoch": 0.056896134291017676,
+      "grad_norm": 0.911556601524353,
+      "learning_rate": 0.00019954423163849364,
+      "loss": 2.7679,
+      "step": 705
+    },
+    {
+      "epoch": 0.05697683802760068,
+      "grad_norm": 0.8558745384216309,
+      "learning_rate": 0.00019954272487131305,
+      "loss": 2.7934,
+      "step": 706
+    },
+    {
+      "epoch": 0.057057541764183685,
+      "grad_norm": 1.0175282955169678,
+      "learning_rate": 0.00019954121562326758,
+      "loss": 2.905,
+      "step": 707
+    },
+    {
+      "epoch": 0.05713824550076668,
+      "grad_norm": 0.9480875730514526,
+      "learning_rate": 0.00019953970389439483,
+      "loss": 2.85,
+      "step": 708
+    },
+    {
+      "epoch": 0.05721894923734969,
+      "grad_norm": 0.9271003603935242,
+      "learning_rate": 0.0001995381896847324,
+      "loss": 2.8237,
+      "step": 709
+    },
+    {
+      "epoch": 0.05729965297393269,
+      "grad_norm": 0.8439653515815735,
+      "learning_rate": 0.00019953667299431815,
+      "loss": 2.821,
+      "step": 710
+    },
+    {
+      "epoch": 0.057380356710515695,
+      "grad_norm": 0.9750552177429199,
+      "learning_rate": 0.0001995351538231898,
+      "loss": 2.8613,
+      "step": 711
+    },
+    {
+      "epoch": 0.0574610604470987,
+      "grad_norm": 0.9409266710281372,
+      "learning_rate": 0.0001995336321713852,
+      "loss": 2.7876,
+      "step": 712
+    },
+    {
+      "epoch": 0.057541764183681704,
+      "grad_norm": 0.811138927936554,
+      "learning_rate": 0.00019953210803894233,
+      "loss": 2.7957,
+      "step": 713
+    },
+    {
+      "epoch": 0.05762246792026471,
+      "grad_norm": 0.9504825472831726,
+      "learning_rate": 0.00019953058142589916,
+      "loss": 2.8536,
+      "step": 714
+    },
+    {
+      "epoch": 0.05770317165684771,
+      "grad_norm": 0.8183554410934448,
+      "learning_rate": 0.00019952905233229368,
+      "loss": 2.7697,
+      "step": 715
+    },
+    {
+      "epoch": 0.05778387539343072,
+      "grad_norm": 1.1146113872528076,
+      "learning_rate": 0.0001995275207581641,
+      "loss": 2.8629,
+      "step": 716
+    },
+    {
+      "epoch": 0.05786457913001372,
+      "grad_norm": 0.8797986507415771,
+      "learning_rate": 0.00019952598670354852,
+      "loss": 2.7962,
+      "step": 717
+    },
+    {
+      "epoch": 0.057945282866596726,
+      "grad_norm": 0.8771101832389832,
+      "learning_rate": 0.00019952445016848517,
+      "loss": 2.8323,
+      "step": 718
+    },
+    {
+      "epoch": 0.05802598660317973,
+      "grad_norm": 0.9003355503082275,
+      "learning_rate": 0.00019952291115301235,
+      "loss": 2.777,
+      "step": 719
+    },
+    {
+      "epoch": 0.058106690339762734,
+      "grad_norm": 0.846125602722168,
+      "learning_rate": 0.00019952136965716846,
+      "loss": 2.7875,
+      "step": 720
+    },
+    {
+      "epoch": 0.05818739407634573,
+      "grad_norm": 0.908833920955658,
+      "learning_rate": 0.00019951982568099187,
+      "loss": 2.7975,
+      "step": 721
+    },
+    {
+      "epoch": 0.058268097812928736,
+      "grad_norm": 0.8616230487823486,
+      "learning_rate": 0.00019951827922452106,
+      "loss": 2.7486,
+      "step": 722
+    },
+    {
+      "epoch": 0.05834880154951174,
+      "grad_norm": 0.8791850805282593,
+      "learning_rate": 0.00019951673028779462,
+      "loss": 2.8301,
+      "step": 723
+    },
+    {
+      "epoch": 0.058429505286094745,
+      "grad_norm": 0.9437321424484253,
+      "learning_rate": 0.00019951517887085112,
+      "loss": 2.7956,
+      "step": 724
+    },
+    {
+      "epoch": 0.05851020902267775,
+      "grad_norm": 0.9263394474983215,
+      "learning_rate": 0.00019951362497372922,
+      "loss": 2.867,
+      "step": 725
+    },
+    {
+      "epoch": 0.05859091275926075,
+      "grad_norm": 0.9442462921142578,
+      "learning_rate": 0.00019951206859646764,
+      "loss": 2.8447,
+      "step": 726
+    },
+    {
+      "epoch": 0.05867161649584376,
+      "grad_norm": 0.9286711812019348,
+      "learning_rate": 0.0001995105097391052,
+      "loss": 2.7588,
+      "step": 727
+    },
+    {
+      "epoch": 0.05875232023242676,
+      "grad_norm": 0.9338774085044861,
+      "learning_rate": 0.00019950894840168072,
+      "loss": 2.7394,
+      "step": 728
+    },
+    {
+      "epoch": 0.058833023969009766,
+      "grad_norm": 0.8880760073661804,
+      "learning_rate": 0.00019950738458423314,
+      "loss": 2.7949,
+      "step": 729
+    },
+    {
+      "epoch": 0.05891372770559277,
+      "grad_norm": 1.0091183185577393,
+      "learning_rate": 0.00019950581828680143,
+      "loss": 2.8633,
+      "step": 730
+    },
+    {
+      "epoch": 0.058994431442175775,
+      "grad_norm": 0.8657729625701904,
+      "learning_rate": 0.0001995042495094246,
+      "loss": 2.8649,
+      "step": 731
+    },
+    {
+      "epoch": 0.05907513517875878,
+      "grad_norm": 1.0084047317504883,
+      "learning_rate": 0.00019950267825214176,
+      "loss": 2.8422,
+      "step": 732
+    },
+    {
+      "epoch": 0.059155838915341784,
+      "grad_norm": 0.9096506237983704,
+      "learning_rate": 0.00019950110451499208,
+      "loss": 2.7908,
+      "step": 733
+    },
+    {
+      "epoch": 0.05923654265192478,
+      "grad_norm": 1.1338937282562256,
+      "learning_rate": 0.0001994995282980148,
+      "loss": 2.8093,
+      "step": 734
+    },
+    {
+      "epoch": 0.059317246388507786,
+      "grad_norm": 0.8813811540603638,
+      "learning_rate": 0.00019949794960124915,
+      "loss": 2.8866,
+      "step": 735
+    },
+    {
+      "epoch": 0.05939795012509079,
+      "grad_norm": 0.8457592129707336,
+      "learning_rate": 0.00019949636842473453,
+      "loss": 2.7744,
+      "step": 736
+    },
+    {
+      "epoch": 0.059478653861673794,
+      "grad_norm": 0.8731856346130371,
+      "learning_rate": 0.0001994947847685103,
+      "loss": 2.7822,
+      "step": 737
+    },
+    {
+      "epoch": 0.0595593575982568,
+      "grad_norm": 0.8915185332298279,
+      "learning_rate": 0.00019949319863261597,
+      "loss": 2.773,
+      "step": 738
+    },
+    {
+      "epoch": 0.0596400613348398,
+      "grad_norm": 0.9478987455368042,
+      "learning_rate": 0.00019949161001709106,
+      "loss": 2.8462,
+      "step": 739
+    },
+    {
+      "epoch": 0.05972076507142281,
+      "grad_norm": 0.8903716206550598,
+      "learning_rate": 0.00019949001892197515,
+      "loss": 2.7741,
+      "step": 740
+    },
+    {
+      "epoch": 0.05980146880800581,
+      "grad_norm": 0.8870117664337158,
+      "learning_rate": 0.00019948842534730786,
+      "loss": 2.8255,
+      "step": 741
+    },
+    {
+      "epoch": 0.059882172544588816,
+      "grad_norm": 1.0766080617904663,
+      "learning_rate": 0.00019948682929312898,
+      "loss": 2.8865,
+      "step": 742
+    },
+    {
+      "epoch": 0.05996287628117182,
+      "grad_norm": 0.846447229385376,
+      "learning_rate": 0.00019948523075947824,
+      "loss": 2.8441,
+      "step": 743
+    },
+    {
+      "epoch": 0.060043580017754825,
+      "grad_norm": 0.9847991466522217,
+      "learning_rate": 0.00019948362974639552,
+      "loss": 2.8099,
+      "step": 744
+    },
+    {
+      "epoch": 0.06012428375433783,
+      "grad_norm": 0.9170514941215515,
+      "learning_rate": 0.00019948202625392068,
+      "loss": 2.8797,
+      "step": 745
+    },
+    {
+      "epoch": 0.060204987490920826,
+      "grad_norm": 0.8564898371696472,
+      "learning_rate": 0.0001994804202820937,
+      "loss": 2.7993,
+      "step": 746
+    },
+    {
+      "epoch": 0.06028569122750383,
+      "grad_norm": 0.8527392148971558,
+      "learning_rate": 0.00019947881183095457,
+      "loss": 2.7816,
+      "step": 747
+    },
+    {
+      "epoch": 0.060366394964086835,
+      "grad_norm": 0.9170876145362854,
+      "learning_rate": 0.00019947720090054342,
+      "loss": 2.8031,
+      "step": 748
+    },
+    {
+      "epoch": 0.06044709870066984,
+      "grad_norm": 0.8891414403915405,
+      "learning_rate": 0.0001994755874909004,
+      "loss": 2.8072,
+      "step": 749
+    },
+    {
+      "epoch": 0.060527802437252844,
+      "grad_norm": 0.8853670358657837,
+      "learning_rate": 0.0001994739716020657,
+      "loss": 2.8857,
+      "step": 750
+    },
+    {
+      "epoch": 0.06060850617383585,
+      "grad_norm": 0.9011211395263672,
+      "learning_rate": 0.0001994723532340796,
+      "loss": 2.8519,
+      "step": 751
+    },
+    {
+      "epoch": 0.06068920991041885,
+      "grad_norm": 0.8843330144882202,
+      "learning_rate": 0.00019947073238698243,
+      "loss": 2.7882,
+      "step": 752
+    },
+    {
+      "epoch": 0.06076991364700186,
+      "grad_norm": 0.8712944984436035,
+      "learning_rate": 0.00019946910906081463,
+      "loss": 2.791,
+      "step": 753
+    },
+    {
+      "epoch": 0.06085061738358486,
+      "grad_norm": 0.8296090364456177,
+      "learning_rate": 0.00019946748325561656,
+      "loss": 2.8073,
+      "step": 754
+    },
+    {
+      "epoch": 0.060931321120167865,
+      "grad_norm": 0.9239117503166199,
+      "learning_rate": 0.00019946585497142885,
+      "loss": 2.8209,
+      "step": 755
+    },
+    {
+      "epoch": 0.06101202485675087,
+      "grad_norm": 0.8885170221328735,
+      "learning_rate": 0.000199464224208292,
+      "loss": 2.8391,
+      "step": 756
+    },
+    {
+      "epoch": 0.061092728593333874,
+      "grad_norm": 0.933720588684082,
+      "learning_rate": 0.0001994625909662467,
+      "loss": 2.7635,
+      "step": 757
+    },
+    {
+      "epoch": 0.06117343232991688,
+      "grad_norm": 0.9751253724098206,
+      "learning_rate": 0.00019946095524533362,
+      "loss": 2.7933,
+      "step": 758
+    },
+    {
+      "epoch": 0.061254136066499876,
+      "grad_norm": 0.9469670057296753,
+      "learning_rate": 0.00019945931704559353,
+      "loss": 2.7652,
+      "step": 759
+    },
+    {
+      "epoch": 0.06133483980308288,
+      "grad_norm": 0.8559684157371521,
+      "learning_rate": 0.00019945767636706728,
+      "loss": 2.8258,
+      "step": 760
+    },
+    {
+      "epoch": 0.061415543539665884,
+      "grad_norm": 1.021478295326233,
+      "learning_rate": 0.00019945603320979574,
+      "loss": 2.8047,
+      "step": 761
+    },
+    {
+      "epoch": 0.06149624727624889,
+      "grad_norm": 0.8421681523323059,
+      "learning_rate": 0.00019945438757381986,
+      "loss": 2.8233,
+      "step": 762
+    },
+    {
+      "epoch": 0.06157695101283189,
+      "grad_norm": 0.900654137134552,
+      "learning_rate": 0.0001994527394591807,
+      "loss": 2.7591,
+      "step": 763
+    },
+    {
+      "epoch": 0.0616576547494149,
+      "grad_norm": 0.878300666809082,
+      "learning_rate": 0.0001994510888659193,
+      "loss": 2.715,
+      "step": 764
+    },
+    {
+      "epoch": 0.0617383584859979,
+      "grad_norm": 0.9170855283737183,
+      "learning_rate": 0.00019944943579407678,
+      "loss": 2.8604,
+      "step": 765
+    },
+    {
+      "epoch": 0.061819062222580906,
+      "grad_norm": 0.8532859683036804,
+      "learning_rate": 0.00019944778024369434,
+      "loss": 2.8124,
+      "step": 766
+    },
+    {
+      "epoch": 0.06189976595916391,
+      "grad_norm": 0.8549049496650696,
+      "learning_rate": 0.00019944612221481332,
+      "loss": 2.8066,
+      "step": 767
+    },
+    {
+      "epoch": 0.061980469695746915,
+      "grad_norm": 0.9602857828140259,
+      "learning_rate": 0.00019944446170747492,
+      "loss": 2.8424,
+      "step": 768
+    },
+    {
+      "epoch": 0.06206117343232992,
+      "grad_norm": 0.910953164100647,
+      "learning_rate": 0.0001994427987217206,
+      "loss": 2.8093,
+      "step": 769
+    },
+    {
+      "epoch": 0.06214187716891292,
+      "grad_norm": 0.8536386489868164,
+      "learning_rate": 0.0001994411332575918,
+      "loss": 2.802,
+      "step": 770
+    },
+    {
+      "epoch": 0.06222258090549593,
+      "grad_norm": 0.9166232347488403,
+      "learning_rate": 0.00019943946531513,
+      "loss": 2.783,
+      "step": 771
+    },
+    {
+      "epoch": 0.062303284642078925,
+      "grad_norm": 0.9954056739807129,
+      "learning_rate": 0.00019943779489437678,
+      "loss": 2.8198,
+      "step": 772
+    },
+    {
+      "epoch": 0.06238398837866193,
+      "grad_norm": 0.8527171015739441,
+      "learning_rate": 0.0001994361219953738,
+      "loss": 2.8159,
+      "step": 773
+    },
+    {
+      "epoch": 0.062464692115244934,
+      "grad_norm": 0.8951592445373535,
+      "learning_rate": 0.00019943444661816274,
+      "loss": 2.7969,
+      "step": 774
+    },
+    {
+      "epoch": 0.06254539585182795,
+      "grad_norm": 0.9348207116127014,
+      "learning_rate": 0.00019943276876278532,
+      "loss": 2.8403,
+      "step": 775
+    },
+    {
+      "epoch": 0.06262609958841095,
+      "grad_norm": 0.866318941116333,
+      "learning_rate": 0.00019943108842928342,
+      "loss": 2.7886,
+      "step": 776
+    },
+    {
+      "epoch": 0.06270680332499395,
+      "grad_norm": 0.8571285605430603,
+      "learning_rate": 0.00019942940561769884,
+      "loss": 2.771,
+      "step": 777
+    },
+    {
+      "epoch": 0.06278750706157694,
+      "grad_norm": 0.8384295105934143,
+      "learning_rate": 0.00019942772032807357,
+      "loss": 2.7885,
+      "step": 778
+    },
+    {
+      "epoch": 0.06286821079815995,
+      "grad_norm": 0.9934808611869812,
+      "learning_rate": 0.00019942603256044961,
+      "loss": 2.8399,
+      "step": 779
+    },
+    {
+      "epoch": 0.06294891453474295,
+      "grad_norm": 0.8275915384292603,
+      "learning_rate": 0.00019942434231486902,
+      "loss": 2.8983,
+      "step": 780
+    },
+    {
+      "epoch": 0.06302961827132596,
+      "grad_norm": 0.9073596000671387,
+      "learning_rate": 0.0001994226495913739,
+      "loss": 2.7886,
+      "step": 781
+    },
+    {
+      "epoch": 0.06311032200790896,
+      "grad_norm": 0.9091461300849915,
+      "learning_rate": 0.00019942095439000646,
+      "loss": 2.814,
+      "step": 782
+    },
+    {
+      "epoch": 0.06319102574449197,
+      "grad_norm": 0.9356934428215027,
+      "learning_rate": 0.000199419256710809,
+      "loss": 2.8238,
+      "step": 783
+    },
+    {
+      "epoch": 0.06327172948107497,
+      "grad_norm": 0.883514940738678,
+      "learning_rate": 0.00019941755655382374,
+      "loss": 2.7912,
+      "step": 784
+    },
+    {
+      "epoch": 0.06335243321765797,
+      "grad_norm": 0.8770506381988525,
+      "learning_rate": 0.00019941585391909308,
+      "loss": 2.7774,
+      "step": 785
+    },
+    {
+      "epoch": 0.06343313695424098,
+      "grad_norm": 0.8891726136207581,
+      "learning_rate": 0.00019941414880665948,
+      "loss": 2.7975,
+      "step": 786
+    },
+    {
+      "epoch": 0.06351384069082398,
+      "grad_norm": 0.9280585050582886,
+      "learning_rate": 0.00019941244121656545,
+      "loss": 2.9468,
+      "step": 787
+    },
+    {
+      "epoch": 0.06359454442740699,
+      "grad_norm": 0.8545510768890381,
+      "learning_rate": 0.00019941073114885347,
+      "loss": 2.8165,
+      "step": 788
+    },
+    {
+      "epoch": 0.06367524816398999,
+      "grad_norm": 0.8631312847137451,
+      "learning_rate": 0.0001994090186035662,
+      "loss": 2.7955,
+      "step": 789
+    },
+    {
+      "epoch": 0.063755951900573,
+      "grad_norm": 0.8883851170539856,
+      "learning_rate": 0.00019940730358074634,
+      "loss": 2.7828,
+      "step": 790
+    },
+    {
+      "epoch": 0.063836655637156,
+      "grad_norm": 0.8421074748039246,
+      "learning_rate": 0.00019940558608043664,
+      "loss": 2.7999,
+      "step": 791
+    },
+    {
+      "epoch": 0.063917359373739,
+      "grad_norm": 0.918134868144989,
+      "learning_rate": 0.0001994038661026799,
+      "loss": 2.7888,
+      "step": 792
+    },
+    {
+      "epoch": 0.06399806311032201,
+      "grad_norm": 0.8513637781143188,
+      "learning_rate": 0.00019940214364751896,
+      "loss": 2.7719,
+      "step": 793
+    },
+    {
+      "epoch": 0.06407876684690501,
+      "grad_norm": 0.9181898236274719,
+      "learning_rate": 0.00019940041871499675,
+      "loss": 2.8345,
+      "step": 794
+    },
+    {
+      "epoch": 0.06415947058348802,
+      "grad_norm": 0.8129134774208069,
+      "learning_rate": 0.00019939869130515626,
+      "loss": 2.7316,
+      "step": 795
+    },
+    {
+      "epoch": 0.06424017432007102,
+      "grad_norm": 0.8782191872596741,
+      "learning_rate": 0.00019939696141804057,
+      "loss": 2.7852,
+      "step": 796
+    },
+    {
+      "epoch": 0.06432087805665403,
+      "grad_norm": 0.9064851403236389,
+      "learning_rate": 0.00019939522905369276,
+      "loss": 2.8105,
+      "step": 797
+    },
+    {
+      "epoch": 0.06440158179323703,
+      "grad_norm": 0.9888454675674438,
+      "learning_rate": 0.00019939349421215603,
+      "loss": 2.8496,
+      "step": 798
+    },
+    {
+      "epoch": 0.06448228552982004,
+      "grad_norm": 0.8717427253723145,
+      "learning_rate": 0.0001993917568934736,
+      "loss": 2.8227,
+      "step": 799
+    },
+    {
+      "epoch": 0.06456298926640304,
+      "grad_norm": 0.922980010509491,
+      "learning_rate": 0.0001993900170976888,
+      "loss": 2.8571,
+      "step": 800
+    },
+    {
+      "epoch": 0.06464369300298604,
+      "grad_norm": 0.8311850428581238,
+      "learning_rate": 0.00019938827482484492,
+      "loss": 2.7905,
+      "step": 801
+    },
+    {
+      "epoch": 0.06472439673956905,
+      "grad_norm": 0.9274900555610657,
+      "learning_rate": 0.0001993865300749855,
+      "loss": 2.8526,
+      "step": 802
+    },
+    {
+      "epoch": 0.06480510047615205,
+      "grad_norm": 0.9072165489196777,
+      "learning_rate": 0.00019938478284815388,
+      "loss": 2.8384,
+      "step": 803
+    },
+    {
+      "epoch": 0.06488580421273504,
+      "grad_norm": 0.854099452495575,
+      "learning_rate": 0.0001993830331443937,
+      "loss": 2.8459,
+      "step": 804
+    },
+    {
+      "epoch": 0.06496650794931805,
+      "grad_norm": 0.824126660823822,
+      "learning_rate": 0.00019938128096374854,
+      "loss": 2.7845,
+      "step": 805
+    },
+    {
+      "epoch": 0.06504721168590105,
+      "grad_norm": 0.8570442795753479,
+      "learning_rate": 0.0001993795263062621,
+      "loss": 2.8446,
+      "step": 806
+    },
+    {
+      "epoch": 0.06512791542248406,
+      "grad_norm": 0.8998628854751587,
+      "learning_rate": 0.00019937776917197805,
+      "loss": 2.8604,
+      "step": 807
+    },
+    {
+      "epoch": 0.06520861915906706,
+      "grad_norm": 0.9189189076423645,
+      "learning_rate": 0.00019937600956094023,
+      "loss": 2.7866,
+      "step": 808
+    },
+    {
+      "epoch": 0.06528932289565006,
+      "grad_norm": 0.9471604824066162,
+      "learning_rate": 0.00019937424747319248,
+      "loss": 2.7619,
+      "step": 809
+    },
+    {
+      "epoch": 0.06537002663223307,
+      "grad_norm": 0.8507755994796753,
+      "learning_rate": 0.00019937248290877874,
+      "loss": 2.8259,
+      "step": 810
+    },
+    {
+      "epoch": 0.06545073036881607,
+      "grad_norm": 0.8800963759422302,
+      "learning_rate": 0.00019937071586774292,
+      "loss": 2.827,
+      "step": 811
+    },
+    {
+      "epoch": 0.06553143410539908,
+      "grad_norm": 0.8851124048233032,
+      "learning_rate": 0.00019936894635012915,
+      "loss": 2.793,
+      "step": 812
+    },
+    {
+      "epoch": 0.06561213784198208,
+      "grad_norm": 0.88127601146698,
+      "learning_rate": 0.00019936717435598144,
+      "loss": 2.8885,
+      "step": 813
+    },
+    {
+      "epoch": 0.06569284157856509,
+      "grad_norm": 0.9115073084831238,
+      "learning_rate": 0.000199365399885344,
+      "loss": 2.8278,
+      "step": 814
+    },
+    {
+      "epoch": 0.06577354531514809,
+      "grad_norm": 0.8722662925720215,
+      "learning_rate": 0.00019936362293826107,
+      "loss": 2.8125,
+      "step": 815
+    },
+    {
+      "epoch": 0.0658542490517311,
+      "grad_norm": 0.8332365155220032,
+      "learning_rate": 0.0001993618435147769,
+      "loss": 2.7682,
+      "step": 816
+    },
+    {
+      "epoch": 0.0659349527883141,
+      "grad_norm": 0.9524003863334656,
+      "learning_rate": 0.0001993600616149359,
+      "loss": 2.8166,
+      "step": 817
+    },
+    {
+      "epoch": 0.0660156565248971,
+      "grad_norm": 0.8402767181396484,
+      "learning_rate": 0.0001993582772387824,
+      "loss": 2.8192,
+      "step": 818
+    },
+    {
+      "epoch": 0.06609636026148011,
+      "grad_norm": 0.8589913249015808,
+      "learning_rate": 0.0001993564903863609,
+      "loss": 2.7785,
+      "step": 819
+    },
+    {
+      "epoch": 0.06617706399806311,
+      "grad_norm": 1.034550428390503,
+      "learning_rate": 0.00019935470105771598,
+      "loss": 2.8407,
+      "step": 820
+    },
+    {
+      "epoch": 0.06625776773464612,
+      "grad_norm": 0.856490969657898,
+      "learning_rate": 0.0001993529092528921,
+      "loss": 2.794,
+      "step": 821
+    },
+    {
+      "epoch": 0.06633847147122912,
+      "grad_norm": 0.897498369216919,
+      "learning_rate": 0.0001993511149719341,
+      "loss": 2.7959,
+      "step": 822
+    },
+    {
+      "epoch": 0.06641917520781213,
+      "grad_norm": 0.8495277166366577,
+      "learning_rate": 0.00019934931821488658,
+      "loss": 2.783,
+      "step": 823
+    },
+    {
+      "epoch": 0.06649987894439513,
+      "grad_norm": 0.8362239599227905,
+      "learning_rate": 0.00019934751898179436,
+      "loss": 2.8628,
+      "step": 824
+    },
+    {
+      "epoch": 0.06658058268097813,
+      "grad_norm": 0.8702061176300049,
+      "learning_rate": 0.00019934571727270225,
+      "loss": 2.7878,
+      "step": 825
+    },
+    {
+      "epoch": 0.06666128641756114,
+      "grad_norm": 0.8341560363769531,
+      "learning_rate": 0.0001993439130876552,
+      "loss": 2.7345,
+      "step": 826
+    },
+    {
+      "epoch": 0.06674199015414414,
+      "grad_norm": 0.880181074142456,
+      "learning_rate": 0.00019934210642669813,
+      "loss": 2.7789,
+      "step": 827
+    },
+    {
+      "epoch": 0.06682269389072715,
+      "grad_norm": 0.9088126420974731,
+      "learning_rate": 0.00019934029728987607,
+      "loss": 2.7893,
+      "step": 828
+    },
+    {
+      "epoch": 0.06690339762731014,
+      "grad_norm": 0.8087106347084045,
+      "learning_rate": 0.00019933848567723416,
+      "loss": 2.7967,
+      "step": 829
+    },
+    {
+      "epoch": 0.06698410136389314,
+      "grad_norm": 0.8970876336097717,
+      "learning_rate": 0.00019933667158881745,
+      "loss": 2.8837,
+      "step": 830
+    },
+    {
+      "epoch": 0.06706480510047615,
+      "grad_norm": 0.9344804883003235,
+      "learning_rate": 0.00019933485502467128,
+      "loss": 2.7754,
+      "step": 831
+    },
+    {
+      "epoch": 0.06714550883705915,
+      "grad_norm": 0.8119301795959473,
+      "learning_rate": 0.00019933303598484084,
+      "loss": 2.7919,
+      "step": 832
+    },
+    {
+      "epoch": 0.06722621257364216,
+      "grad_norm": 0.9370681047439575,
+      "learning_rate": 0.00019933121446937148,
+      "loss": 2.8011,
+      "step": 833
+    },
+    {
+      "epoch": 0.06730691631022516,
+      "grad_norm": 0.8358973264694214,
+      "learning_rate": 0.00019932939047830858,
+      "loss": 2.8339,
+      "step": 834
+    },
+    {
+      "epoch": 0.06738762004680816,
+      "grad_norm": 0.8565972447395325,
+      "learning_rate": 0.00019932756401169765,
+      "loss": 2.8269,
+      "step": 835
+    },
+    {
+      "epoch": 0.06746832378339117,
+      "grad_norm": 0.8405514359474182,
+      "learning_rate": 0.00019932573506958417,
+      "loss": 2.7621,
+      "step": 836
+    },
+    {
+      "epoch": 0.06754902751997417,
+      "grad_norm": 0.8217617869377136,
+      "learning_rate": 0.00019932390365201373,
+      "loss": 2.8363,
+      "step": 837
+    },
+    {
+      "epoch": 0.06762973125655718,
+      "grad_norm": 0.9121438264846802,
+      "learning_rate": 0.00019932206975903198,
+      "loss": 2.8033,
+      "step": 838
+    },
+    {
+      "epoch": 0.06771043499314018,
+      "grad_norm": 0.9113054871559143,
+      "learning_rate": 0.00019932023339068464,
+      "loss": 2.8696,
+      "step": 839
+    },
+    {
+      "epoch": 0.06779113872972319,
+      "grad_norm": 0.8638293743133545,
+      "learning_rate": 0.00019931839454701743,
+      "loss": 2.8008,
+      "step": 840
+    },
+    {
+      "epoch": 0.06787184246630619,
+      "grad_norm": 0.862932562828064,
+      "learning_rate": 0.0001993165532280762,
+      "loss": 2.8092,
+      "step": 841
+    },
+    {
+      "epoch": 0.0679525462028892,
+      "grad_norm": 0.9089607000350952,
+      "learning_rate": 0.00019931470943390685,
+      "loss": 2.8921,
+      "step": 842
+    },
+    {
+      "epoch": 0.0680332499394722,
+      "grad_norm": 0.9233555793762207,
+      "learning_rate": 0.00019931286316455537,
+      "loss": 2.9025,
+      "step": 843
+    },
+    {
+      "epoch": 0.0681139536760552,
+      "grad_norm": 0.9403017163276672,
+      "learning_rate": 0.0001993110144200677,
+      "loss": 2.7875,
+      "step": 844
+    },
+    {
+      "epoch": 0.06819465741263821,
+      "grad_norm": 0.9194290637969971,
+      "learning_rate": 0.00019930916320048996,
+      "loss": 2.8254,
+      "step": 845
+    },
+    {
+      "epoch": 0.06827536114922121,
+      "grad_norm": 0.8238688111305237,
+      "learning_rate": 0.00019930730950586828,
+      "loss": 2.82,
+      "step": 846
+    },
+    {
+      "epoch": 0.06835606488580422,
+      "grad_norm": 0.8560660481452942,
+      "learning_rate": 0.00019930545333624885,
+      "loss": 2.8516,
+      "step": 847
+    },
+    {
+      "epoch": 0.06843676862238722,
+      "grad_norm": 0.9127222895622253,
+      "learning_rate": 0.0001993035946916779,
+      "loss": 2.7674,
+      "step": 848
+    },
+    {
+      "epoch": 0.06851747235897022,
+      "grad_norm": 0.8679420948028564,
+      "learning_rate": 0.00019930173357220182,
+      "loss": 2.777,
+      "step": 849
+    },
+    {
+      "epoch": 0.06859817609555323,
+      "grad_norm": 0.9686945676803589,
+      "learning_rate": 0.00019929986997786699,
+      "loss": 2.7841,
+      "step": 850
+    },
+    {
+      "epoch": 0.06867887983213623,
+      "grad_norm": 0.8366333246231079,
+      "learning_rate": 0.00019929800390871977,
+      "loss": 2.7993,
+      "step": 851
+    },
+    {
+      "epoch": 0.06875958356871924,
+      "grad_norm": 0.8374585509300232,
+      "learning_rate": 0.00019929613536480675,
+      "loss": 2.7545,
+      "step": 852
+    },
+    {
+      "epoch": 0.06884028730530224,
+      "grad_norm": 0.9843763709068298,
+      "learning_rate": 0.00019929426434617451,
+      "loss": 2.8118,
+      "step": 853
+    },
+    {
+      "epoch": 0.06892099104188525,
+      "grad_norm": 0.8093454241752625,
+      "learning_rate": 0.0001992923908528696,
+      "loss": 2.7301,
+      "step": 854
+    },
+    {
+      "epoch": 0.06900169477846824,
+      "grad_norm": 0.8374418020248413,
+      "learning_rate": 0.00019929051488493877,
+      "loss": 2.7745,
+      "step": 855
+    },
+    {
+      "epoch": 0.06908239851505124,
+      "grad_norm": 0.869965136051178,
+      "learning_rate": 0.00019928863644242875,
+      "loss": 2.7637,
+      "step": 856
+    },
+    {
+      "epoch": 0.06916310225163425,
+      "grad_norm": 0.9280590415000916,
+      "learning_rate": 0.00019928675552538638,
+      "loss": 2.7792,
+      "step": 857
+    },
+    {
+      "epoch": 0.06924380598821725,
+      "grad_norm": 0.8624193668365479,
+      "learning_rate": 0.00019928487213385852,
+      "loss": 2.7755,
+      "step": 858
+    },
+    {
+      "epoch": 0.06932450972480025,
+      "grad_norm": 0.8379972577095032,
+      "learning_rate": 0.00019928298626789212,
+      "loss": 2.8563,
+      "step": 859
+    },
+    {
+      "epoch": 0.06940521346138326,
+      "grad_norm": 0.9272914528846741,
+      "learning_rate": 0.00019928109792753418,
+      "loss": 2.836,
+      "step": 860
+    },
+    {
+      "epoch": 0.06948591719796626,
+      "grad_norm": 0.9239040613174438,
+      "learning_rate": 0.00019927920711283175,
+      "loss": 2.7999,
+      "step": 861
+    },
+    {
+      "epoch": 0.06956662093454927,
+      "grad_norm": 0.9125113487243652,
+      "learning_rate": 0.00019927731382383195,
+      "loss": 2.8494,
+      "step": 862
+    },
+    {
+      "epoch": 0.06964732467113227,
+      "grad_norm": 0.8782855868339539,
+      "learning_rate": 0.00019927541806058198,
+      "loss": 2.767,
+      "step": 863
+    },
+    {
+      "epoch": 0.06972802840771528,
+      "grad_norm": 0.8815447092056274,
+      "learning_rate": 0.00019927351982312907,
+      "loss": 2.7877,
+      "step": 864
+    },
+    {
+      "epoch": 0.06980873214429828,
+      "grad_norm": 0.8555476069450378,
+      "learning_rate": 0.00019927161911152056,
+      "loss": 2.8057,
+      "step": 865
+    },
+    {
+      "epoch": 0.06988943588088128,
+      "grad_norm": 0.8562924265861511,
+      "learning_rate": 0.00019926971592580382,
+      "loss": 2.8049,
+      "step": 866
+    },
+    {
+      "epoch": 0.06997013961746429,
+      "grad_norm": 0.846503734588623,
+      "learning_rate": 0.00019926781026602625,
+      "loss": 2.8545,
+      "step": 867
+    },
+    {
+      "epoch": 0.07005084335404729,
+      "grad_norm": 0.8439623713493347,
+      "learning_rate": 0.00019926590213223535,
+      "loss": 2.7451,
+      "step": 868
+    },
+    {
+      "epoch": 0.0701315470906303,
+      "grad_norm": 0.8471730351448059,
+      "learning_rate": 0.00019926399152447868,
+      "loss": 2.7879,
+      "step": 869
+    },
+    {
+      "epoch": 0.0702122508272133,
+      "grad_norm": 0.8721400499343872,
+      "learning_rate": 0.00019926207844280387,
+      "loss": 2.8594,
+      "step": 870
+    },
+    {
+      "epoch": 0.0702929545637963,
+      "grad_norm": 0.8110925555229187,
+      "learning_rate": 0.0001992601628872586,
+      "loss": 2.7789,
+      "step": 871
+    },
+    {
+      "epoch": 0.07037365830037931,
+      "grad_norm": 0.9593119025230408,
+      "learning_rate": 0.0001992582448578906,
+      "loss": 2.8792,
+      "step": 872
+    },
+    {
+      "epoch": 0.07045436203696231,
+      "grad_norm": 0.8553354144096375,
+      "learning_rate": 0.00019925632435474765,
+      "loss": 2.8056,
+      "step": 873
+    },
+    {
+      "epoch": 0.07053506577354532,
+      "grad_norm": 0.8062612414360046,
+      "learning_rate": 0.00019925440137787768,
+      "loss": 2.7762,
+      "step": 874
+    },
+    {
+      "epoch": 0.07061576951012832,
+      "grad_norm": 0.8264921307563782,
+      "learning_rate": 0.00019925247592732858,
+      "loss": 2.8435,
+      "step": 875
+    },
+    {
+      "epoch": 0.07069647324671133,
+      "grad_norm": 0.7770401835441589,
+      "learning_rate": 0.00019925054800314828,
+      "loss": 2.7846,
+      "step": 876
+    },
+    {
+      "epoch": 0.07077717698329433,
+      "grad_norm": 0.8426765203475952,
+      "learning_rate": 0.0001992486176053849,
+      "loss": 2.782,
+      "step": 877
+    },
+    {
+      "epoch": 0.07085788071987734,
+      "grad_norm": 0.855330228805542,
+      "learning_rate": 0.00019924668473408655,
+      "loss": 2.8051,
+      "step": 878
+    },
+    {
+      "epoch": 0.07093858445646034,
+      "grad_norm": 0.8762049674987793,
+      "learning_rate": 0.00019924474938930135,
+      "loss": 2.7634,
+      "step": 879
+    },
+    {
+      "epoch": 0.07101928819304333,
+      "grad_norm": 0.9226812124252319,
+      "learning_rate": 0.0001992428115710776,
+      "loss": 2.8342,
+      "step": 880
+    },
+    {
+      "epoch": 0.07109999192962634,
+      "grad_norm": 0.9031660556793213,
+      "learning_rate": 0.00019924087127946353,
+      "loss": 2.7953,
+      "step": 881
+    },
+    {
+      "epoch": 0.07118069566620934,
+      "grad_norm": 1.0151792764663696,
+      "learning_rate": 0.00019923892851450757,
+      "loss": 2.8225,
+      "step": 882
+    },
+    {
+      "epoch": 0.07126139940279234,
+      "grad_norm": 0.9805678725242615,
+      "learning_rate": 0.00019923698327625806,
+      "loss": 2.7727,
+      "step": 883
+    },
+    {
+      "epoch": 0.07134210313937535,
+      "grad_norm": 0.8831729888916016,
+      "learning_rate": 0.00019923503556476356,
+      "loss": 2.7682,
+      "step": 884
+    },
+    {
+      "epoch": 0.07142280687595835,
+      "grad_norm": 1.0311404466629028,
+      "learning_rate": 0.00019923308538007253,
+      "loss": 2.8422,
+      "step": 885
+    },
+    {
+      "epoch": 0.07150351061254136,
+      "grad_norm": 0.8143388628959656,
+      "learning_rate": 0.0001992311327222336,
+      "loss": 2.7876,
+      "step": 886
+    },
+    {
+      "epoch": 0.07158421434912436,
+      "grad_norm": 0.877017617225647,
+      "learning_rate": 0.00019922917759129552,
+      "loss": 2.7486,
+      "step": 887
+    },
+    {
+      "epoch": 0.07166491808570737,
+      "grad_norm": 0.930646538734436,
+      "learning_rate": 0.0001992272199873069,
+      "loss": 2.8022,
+      "step": 888
+    },
+    {
+      "epoch": 0.07174562182229037,
+      "grad_norm": 0.934753954410553,
+      "learning_rate": 0.00019922525991031655,
+      "loss": 2.8485,
+      "step": 889
+    },
+    {
+      "epoch": 0.07182632555887337,
+      "grad_norm": 0.9564220905303955,
+      "learning_rate": 0.00019922329736037339,
+      "loss": 2.761,
+      "step": 890
+    },
+    {
+      "epoch": 0.07190702929545638,
+      "grad_norm": 0.9457311630249023,
+      "learning_rate": 0.00019922133233752626,
+      "loss": 2.8279,
+      "step": 891
+    },
+    {
+      "epoch": 0.07198773303203938,
+      "grad_norm": 0.9385658502578735,
+      "learning_rate": 0.0001992193648418242,
+      "loss": 2.8222,
+      "step": 892
+    },
+    {
+      "epoch": 0.07206843676862239,
+      "grad_norm": 1.0157524347305298,
+      "learning_rate": 0.00019921739487331616,
+      "loss": 2.9166,
+      "step": 893
+    },
+    {
+      "epoch": 0.07214914050520539,
+      "grad_norm": 0.9143860340118408,
+      "learning_rate": 0.00019921542243205132,
+      "loss": 2.8139,
+      "step": 894
+    },
+    {
+      "epoch": 0.0722298442417884,
+      "grad_norm": 0.8769320249557495,
+      "learning_rate": 0.00019921344751807878,
+      "loss": 2.8023,
+      "step": 895
+    },
+    {
+      "epoch": 0.0723105479783714,
+      "grad_norm": 0.9647517204284668,
+      "learning_rate": 0.0001992114701314478,
+      "loss": 2.8872,
+      "step": 896
+    },
+    {
+      "epoch": 0.0723912517149544,
+      "grad_norm": 1.025978446006775,
+      "learning_rate": 0.00019920949027220762,
+      "loss": 2.837,
+      "step": 897
+    },
+    {
+      "epoch": 0.07247195545153741,
+      "grad_norm": 0.8848521113395691,
+      "learning_rate": 0.0001992075079404076,
+      "loss": 2.7498,
+      "step": 898
+    },
+    {
+      "epoch": 0.07255265918812041,
+      "grad_norm": 0.9395595788955688,
+      "learning_rate": 0.0001992055231360972,
+      "loss": 2.8752,
+      "step": 899
+    },
+    {
+      "epoch": 0.07263336292470342,
+      "grad_norm": 0.8711572885513306,
+      "learning_rate": 0.00019920353585932578,
+      "loss": 2.8608,
+      "step": 900
+    },
+    {
+      "epoch": 0.07271406666128642,
+      "grad_norm": 0.8606846332550049,
+      "learning_rate": 0.00019920154611014295,
+      "loss": 2.829,
+      "step": 901
+    },
+    {
+      "epoch": 0.07279477039786943,
+      "grad_norm": 0.859354555606842,
+      "learning_rate": 0.0001991995538885983,
+      "loss": 2.8102,
+      "step": 902
+    },
+    {
+      "epoch": 0.07287547413445243,
+      "grad_norm": 0.9063243865966797,
+      "learning_rate": 0.00019919755919474143,
+      "loss": 2.8509,
+      "step": 903
+    },
+    {
+      "epoch": 0.07295617787103544,
+      "grad_norm": 0.8321940898895264,
+      "learning_rate": 0.00019919556202862207,
+      "loss": 2.796,
+      "step": 904
+    },
+    {
+      "epoch": 0.07303688160761844,
+      "grad_norm": 0.8875191807746887,
+      "learning_rate": 0.00019919356239029003,
+      "loss": 2.8672,
+      "step": 905
+    },
+    {
+      "epoch": 0.07311758534420143,
+      "grad_norm": 0.9028071165084839,
+      "learning_rate": 0.0001991915602797951,
+      "loss": 2.8926,
+      "step": 906
+    },
+    {
+      "epoch": 0.07319828908078443,
+      "grad_norm": 0.9449291825294495,
+      "learning_rate": 0.0001991895556971872,
+      "loss": 2.8159,
+      "step": 907
+    },
+    {
+      "epoch": 0.07327899281736744,
+      "grad_norm": 0.871576189994812,
+      "learning_rate": 0.0001991875486425163,
+      "loss": 2.8162,
+      "step": 908
+    },
+    {
+      "epoch": 0.07335969655395044,
+      "grad_norm": 0.818423330783844,
+      "learning_rate": 0.0001991855391158324,
+      "loss": 2.8882,
+      "step": 909
+    },
+    {
+      "epoch": 0.07344040029053345,
+      "grad_norm": 0.8802343606948853,
+      "learning_rate": 0.0001991835271171856,
+      "loss": 2.8245,
+      "step": 910
+    },
+    {
+      "epoch": 0.07352110402711645,
+      "grad_norm": 0.916023313999176,
+      "learning_rate": 0.000199181512646626,
+      "loss": 2.8966,
+      "step": 911
+    },
+    {
+      "epoch": 0.07360180776369946,
+      "grad_norm": 1.0663317441940308,
+      "learning_rate": 0.0001991794957042039,
+      "loss": 2.7736,
+      "step": 912
+    },
+    {
+      "epoch": 0.07368251150028246,
+      "grad_norm": 0.9212445616722107,
+      "learning_rate": 0.00019917747628996947,
+      "loss": 2.7924,
+      "step": 913
+    },
+    {
+      "epoch": 0.07376321523686546,
+      "grad_norm": 0.9785256385803223,
+      "learning_rate": 0.00019917545440397308,
+      "loss": 2.8021,
+      "step": 914
+    },
+    {
+      "epoch": 0.07384391897344847,
+      "grad_norm": 0.8510444760322571,
+      "learning_rate": 0.00019917343004626514,
+      "loss": 2.7991,
+      "step": 915
+    },
+    {
+      "epoch": 0.07392462271003147,
+      "grad_norm": 0.8967106342315674,
+      "learning_rate": 0.0001991714032168961,
+      "loss": 2.8838,
+      "step": 916
+    },
+    {
+      "epoch": 0.07400532644661448,
+      "grad_norm": 0.8940563797950745,
+      "learning_rate": 0.0001991693739159164,
+      "loss": 2.8124,
+      "step": 917
+    },
+    {
+      "epoch": 0.07408603018319748,
+      "grad_norm": 0.9270479679107666,
+      "learning_rate": 0.0001991673421433767,
+      "loss": 2.7627,
+      "step": 918
+    },
+    {
+      "epoch": 0.07416673391978049,
+      "grad_norm": 0.905805230140686,
+      "learning_rate": 0.0001991653078993276,
+      "loss": 2.781,
+      "step": 919
+    },
+    {
+      "epoch": 0.07424743765636349,
+      "grad_norm": 0.9295129179954529,
+      "learning_rate": 0.00019916327118381982,
+      "loss": 2.8332,
+      "step": 920
+    },
+    {
+      "epoch": 0.0743281413929465,
+      "grad_norm": 0.863331139087677,
+      "learning_rate": 0.00019916123199690408,
+      "loss": 2.8489,
+      "step": 921
+    },
+    {
+      "epoch": 0.0744088451295295,
+      "grad_norm": 0.9966896772384644,
+      "learning_rate": 0.00019915919033863127,
+      "loss": 2.9107,
+      "step": 922
+    },
+    {
+      "epoch": 0.0744895488661125,
+      "grad_norm": 0.8921390771865845,
+      "learning_rate": 0.00019915714620905218,
+      "loss": 2.7668,
+      "step": 923
+    },
+    {
+      "epoch": 0.07457025260269551,
+      "grad_norm": 0.9378434419631958,
+      "learning_rate": 0.00019915509960821782,
+      "loss": 2.8305,
+      "step": 924
+    },
+    {
+      "epoch": 0.07465095633927851,
+      "grad_norm": 1.0351817607879639,
+      "learning_rate": 0.0001991530505361792,
+      "loss": 2.9412,
+      "step": 925
+    },
+    {
+      "epoch": 0.07473166007586152,
+      "grad_norm": 0.7995476722717285,
+      "learning_rate": 0.0001991509989929874,
+      "loss": 2.7872,
+      "step": 926
+    },
+    {
+      "epoch": 0.07481236381244452,
+      "grad_norm": 0.858830988407135,
+      "learning_rate": 0.0001991489449786935,
+      "loss": 2.7775,
+      "step": 927
+    },
+    {
+      "epoch": 0.07489306754902753,
+      "grad_norm": 1.1254682540893555,
+      "learning_rate": 0.00019914688849334867,
+      "loss": 2.7913,
+      "step": 928
+    },
+    {
+      "epoch": 0.07497377128561053,
+      "grad_norm": 0.9475330710411072,
+      "learning_rate": 0.00019914482953700428,
+      "loss": 2.7945,
+      "step": 929
+    },
+    {
+      "epoch": 0.07505447502219353,
+      "grad_norm": 0.8427290916442871,
+      "learning_rate": 0.00019914276810971152,
+      "loss": 2.8297,
+      "step": 930
+    },
+    {
+      "epoch": 0.07513517875877652,
+      "grad_norm": 0.9308956265449524,
+      "learning_rate": 0.00019914070421152183,
+      "loss": 2.8534,
+      "step": 931
+    },
+    {
+      "epoch": 0.07521588249535953,
+      "grad_norm": 0.9264787435531616,
+      "learning_rate": 0.00019913863784248664,
+      "loss": 2.7959,
+      "step": 932
+    },
+    {
+      "epoch": 0.07529658623194253,
+      "grad_norm": 0.8432087302207947,
+      "learning_rate": 0.00019913656900265742,
+      "loss": 2.8479,
+      "step": 933
+    },
+    {
+      "epoch": 0.07537728996852554,
+      "grad_norm": 0.8237274885177612,
+      "learning_rate": 0.0001991344976920858,
+      "loss": 2.782,
+      "step": 934
+    },
+    {
+      "epoch": 0.07545799370510854,
+      "grad_norm": 0.8143243789672852,
+      "learning_rate": 0.0001991324239108233,
+      "loss": 2.7567,
+      "step": 935
+    },
+    {
+      "epoch": 0.07553869744169155,
+      "grad_norm": 0.8824434280395508,
+      "learning_rate": 0.0001991303476589217,
+      "loss": 2.7971,
+      "step": 936
+    },
+    {
+      "epoch": 0.07561940117827455,
+      "grad_norm": 0.8202407360076904,
+      "learning_rate": 0.00019912826893643272,
+      "loss": 2.7825,
+      "step": 937
+    },
+    {
+      "epoch": 0.07570010491485755,
+      "grad_norm": 0.8001337647438049,
+      "learning_rate": 0.00019912618774340813,
+      "loss": 2.8294,
+      "step": 938
+    },
+    {
+      "epoch": 0.07578080865144056,
+      "grad_norm": 0.8875572085380554,
+      "learning_rate": 0.00019912410407989982,
+      "loss": 2.8013,
+      "step": 939
+    },
+    {
+      "epoch": 0.07586151238802356,
+      "grad_norm": 0.8676280379295349,
+      "learning_rate": 0.0001991220179459597,
+      "loss": 2.767,
+      "step": 940
+    },
+    {
+      "epoch": 0.07594221612460657,
+      "grad_norm": 0.9767136573791504,
+      "learning_rate": 0.00019911992934163982,
+      "loss": 2.8315,
+      "step": 941
+    },
+    {
+      "epoch": 0.07602291986118957,
+      "grad_norm": 0.8690733909606934,
+      "learning_rate": 0.0001991178382669922,
+      "loss": 2.8042,
+      "step": 942
+    },
+    {
+      "epoch": 0.07610362359777258,
+      "grad_norm": 0.862978458404541,
+      "learning_rate": 0.00019911574472206893,
+      "loss": 2.8243,
+      "step": 943
+    },
+    {
+      "epoch": 0.07618432733435558,
+      "grad_norm": 0.9116127490997314,
+      "learning_rate": 0.00019911364870692225,
+      "loss": 2.7377,
+      "step": 944
+    },
+    {
+      "epoch": 0.07626503107093859,
+      "grad_norm": 0.8765420317649841,
+      "learning_rate": 0.00019911155022160433,
+      "loss": 2.7673,
+      "step": 945
+    },
+    {
+      "epoch": 0.07634573480752159,
+      "grad_norm": 0.8229342699050903,
+      "learning_rate": 0.0001991094492661675,
+      "loss": 2.7749,
+      "step": 946
+    },
+    {
+      "epoch": 0.0764264385441046,
+      "grad_norm": 0.8340098261833191,
+      "learning_rate": 0.00019910734584066412,
+      "loss": 2.7871,
+      "step": 947
+    },
+    {
+      "epoch": 0.0765071422806876,
+      "grad_norm": 0.8116940259933472,
+      "learning_rate": 0.0001991052399451466,
+      "loss": 2.8202,
+      "step": 948
+    },
+    {
+      "epoch": 0.0765878460172706,
+      "grad_norm": 0.8730412721633911,
+      "learning_rate": 0.00019910313157966747,
+      "loss": 2.8661,
+      "step": 949
+    },
+    {
+      "epoch": 0.07666854975385361,
+      "grad_norm": 0.8272213339805603,
+      "learning_rate": 0.0001991010207442792,
+      "loss": 2.8352,
+      "step": 950
+    },
+    {
+      "epoch": 0.07674925349043661,
+      "grad_norm": 0.8586944937705994,
+      "learning_rate": 0.0001990989074390345,
+      "loss": 2.8018,
+      "step": 951
+    },
+    {
+      "epoch": 0.07682995722701962,
+      "grad_norm": 0.81830894947052,
+      "learning_rate": 0.00019909679166398592,
+      "loss": 2.8154,
+      "step": 952
+    },
+    {
+      "epoch": 0.07691066096360262,
+      "grad_norm": 0.8158484101295471,
+      "learning_rate": 0.00019909467341918627,
+      "loss": 2.7618,
+      "step": 953
+    },
+    {
+      "epoch": 0.07699136470018562,
+      "grad_norm": 0.816834032535553,
+      "learning_rate": 0.00019909255270468833,
+      "loss": 2.8125,
+      "step": 954
+    },
+    {
+      "epoch": 0.07707206843676863,
+      "grad_norm": 0.944790780544281,
+      "learning_rate": 0.00019909042952054496,
+      "loss": 2.8054,
+      "step": 955
+    },
+    {
+      "epoch": 0.07715277217335163,
+      "grad_norm": 0.9281302690505981,
+      "learning_rate": 0.00019908830386680904,
+      "loss": 2.8724,
+      "step": 956
+    },
+    {
+      "epoch": 0.07723347590993462,
+      "grad_norm": 0.8850300908088684,
+      "learning_rate": 0.00019908617574353356,
+      "loss": 2.7906,
+      "step": 957
+    },
+    {
+      "epoch": 0.07731417964651763,
+      "grad_norm": 0.8997938632965088,
+      "learning_rate": 0.00019908404515077158,
+      "loss": 2.7814,
+      "step": 958
+    },
+    {
+      "epoch": 0.07739488338310063,
+      "grad_norm": 0.8814194798469543,
+      "learning_rate": 0.0001990819120885762,
+      "loss": 2.7423,
+      "step": 959
+    },
+    {
+      "epoch": 0.07747558711968364,
+      "grad_norm": 0.8759928345680237,
+      "learning_rate": 0.00019907977655700054,
+      "loss": 2.7803,
+      "step": 960
+    },
+    {
+      "epoch": 0.07755629085626664,
+      "grad_norm": 0.8439476490020752,
+      "learning_rate": 0.00019907763855609787,
+      "loss": 2.8277,
+      "step": 961
+    },
+    {
+      "epoch": 0.07763699459284965,
+      "grad_norm": 0.8745121955871582,
+      "learning_rate": 0.00019907549808592144,
+      "loss": 2.8152,
+      "step": 962
+    },
+    {
+      "epoch": 0.07771769832943265,
+      "grad_norm": 1.0439598560333252,
+      "learning_rate": 0.00019907335514652465,
+      "loss": 2.7882,
+      "step": 963
+    },
+    {
+      "epoch": 0.07779840206601565,
+      "grad_norm": 0.9516503810882568,
+      "learning_rate": 0.00019907120973796082,
+      "loss": 2.8555,
+      "step": 964
+    },
+    {
+      "epoch": 0.07787910580259866,
+      "grad_norm": 0.928717315196991,
+      "learning_rate": 0.0001990690618602835,
+      "loss": 2.8214,
+      "step": 965
+    },
+    {
+      "epoch": 0.07795980953918166,
+      "grad_norm": 0.7923071384429932,
+      "learning_rate": 0.00019906691151354617,
+      "loss": 2.8153,
+      "step": 966
+    },
+    {
+      "epoch": 0.07804051327576467,
+      "grad_norm": 0.8783324956893921,
+      "learning_rate": 0.00019906475869780246,
+      "loss": 2.7691,
+      "step": 967
+    },
+    {
+      "epoch": 0.07812121701234767,
+      "grad_norm": 0.8974801301956177,
+      "learning_rate": 0.000199062603413106,
+      "loss": 2.8156,
+      "step": 968
+    },
+    {
+      "epoch": 0.07820192074893068,
+      "grad_norm": 0.9304391741752625,
+      "learning_rate": 0.00019906044565951052,
+      "loss": 2.8489,
+      "step": 969
+    },
+    {
+      "epoch": 0.07828262448551368,
+      "grad_norm": 0.8351098895072937,
+      "learning_rate": 0.00019905828543706976,
+      "loss": 2.7744,
+      "step": 970
+    },
+    {
+      "epoch": 0.07836332822209668,
+      "grad_norm": 0.8634265065193176,
+      "learning_rate": 0.0001990561227458376,
+      "loss": 2.8193,
+      "step": 971
+    },
+    {
+      "epoch": 0.07844403195867969,
+      "grad_norm": 0.8969653248786926,
+      "learning_rate": 0.00019905395758586792,
+      "loss": 2.7548,
+      "step": 972
+    },
+    {
+      "epoch": 0.07852473569526269,
+      "grad_norm": 0.8964852094650269,
+      "learning_rate": 0.0001990517899572147,
+      "loss": 2.8037,
+      "step": 973
+    },
+    {
+      "epoch": 0.0786054394318457,
+      "grad_norm": 0.8567596077919006,
+      "learning_rate": 0.00019904961985993196,
+      "loss": 2.7942,
+      "step": 974
+    },
+    {
+      "epoch": 0.0786861431684287,
+      "grad_norm": 0.8275273442268372,
+      "learning_rate": 0.00019904744729407374,
+      "loss": 2.8359,
+      "step": 975
+    },
+    {
+      "epoch": 0.0787668469050117,
+      "grad_norm": 0.9458810091018677,
+      "learning_rate": 0.00019904527225969424,
+      "loss": 2.8354,
+      "step": 976
+    },
+    {
+      "epoch": 0.07884755064159471,
+      "grad_norm": 0.8690593838691711,
+      "learning_rate": 0.00019904309475684767,
+      "loss": 2.7894,
+      "step": 977
+    },
+    {
+      "epoch": 0.07892825437817771,
+      "grad_norm": 0.810279130935669,
+      "learning_rate": 0.00019904091478558823,
+      "loss": 2.7939,
+      "step": 978
+    },
+    {
+      "epoch": 0.07900895811476072,
+      "grad_norm": 0.8779012560844421,
+      "learning_rate": 0.0001990387323459703,
+      "loss": 2.7551,
+      "step": 979
+    },
+    {
+      "epoch": 0.07908966185134372,
+      "grad_norm": 0.7936381101608276,
+      "learning_rate": 0.00019903654743804833,
+      "loss": 2.814,
+      "step": 980
+    },
+    {
+      "epoch": 0.07917036558792673,
+      "grad_norm": 0.9567989110946655,
+      "learning_rate": 0.00019903436006187667,
+      "loss": 2.7715,
+      "step": 981
+    },
+    {
+      "epoch": 0.07925106932450972,
+      "grad_norm": 0.9250255823135376,
+      "learning_rate": 0.00019903217021750987,
+      "loss": 2.8967,
+      "step": 982
+    },
+    {
+      "epoch": 0.07933177306109272,
+      "grad_norm": 0.8342804312705994,
+      "learning_rate": 0.00019902997790500256,
+      "loss": 2.7728,
+      "step": 983
+    },
+    {
+      "epoch": 0.07941247679767573,
+      "grad_norm": 0.8321473598480225,
+      "learning_rate": 0.00019902778312440932,
+      "loss": 2.8479,
+      "step": 984
+    },
+    {
+      "epoch": 0.07949318053425873,
+      "grad_norm": 0.894727885723114,
+      "learning_rate": 0.00019902558587578484,
+      "loss": 2.8211,
+      "step": 985
+    },
+    {
+      "epoch": 0.07957388427084174,
+      "grad_norm": 0.8093457221984863,
+      "learning_rate": 0.0001990233861591839,
+      "loss": 2.7481,
+      "step": 986
+    },
+    {
+      "epoch": 0.07965458800742474,
+      "grad_norm": 0.8626284599304199,
+      "learning_rate": 0.00019902118397466132,
+      "loss": 2.8368,
+      "step": 987
+    },
+    {
+      "epoch": 0.07973529174400774,
+      "grad_norm": 0.799648642539978,
+      "learning_rate": 0.00019901897932227204,
+      "loss": 2.8713,
+      "step": 988
+    },
+    {
+      "epoch": 0.07981599548059075,
+      "grad_norm": 0.9658265709877014,
+      "learning_rate": 0.00019901677220207092,
+      "loss": 2.7284,
+      "step": 989
+    },
+    {
+      "epoch": 0.07989669921717375,
+      "grad_norm": 0.877299427986145,
+      "learning_rate": 0.00019901456261411303,
+      "loss": 2.7916,
+      "step": 990
+    },
+    {
+      "epoch": 0.07997740295375676,
+      "grad_norm": 0.926450252532959,
+      "learning_rate": 0.00019901235055845337,
+      "loss": 2.8207,
+      "step": 991
+    },
+    {
+      "epoch": 0.08005810669033976,
+      "grad_norm": 0.8858455419540405,
+      "learning_rate": 0.00019901013603514716,
+      "loss": 2.795,
+      "step": 992
+    },
+    {
+      "epoch": 0.08013881042692277,
+      "grad_norm": 0.8619922995567322,
+      "learning_rate": 0.0001990079190442495,
+      "loss": 2.8163,
+      "step": 993
+    },
+    {
+      "epoch": 0.08021951416350577,
+      "grad_norm": 0.859200656414032,
+      "learning_rate": 0.00019900569958581572,
+      "loss": 2.7715,
+      "step": 994
+    },
+    {
+      "epoch": 0.08030021790008877,
+      "grad_norm": 0.8346282839775085,
+      "learning_rate": 0.0001990034776599011,
+      "loss": 2.8312,
+      "step": 995
+    },
+    {
+      "epoch": 0.08038092163667178,
+      "grad_norm": 0.9188725352287292,
+      "learning_rate": 0.00019900125326656102,
+      "loss": 2.799,
+      "step": 996
+    },
+    {
+      "epoch": 0.08046162537325478,
+      "grad_norm": 0.8548648953437805,
+      "learning_rate": 0.00019899902640585092,
+      "loss": 2.7778,
+      "step": 997
+    },
+    {
+      "epoch": 0.08054232910983779,
+      "grad_norm": 0.8883183002471924,
+      "learning_rate": 0.00019899679707782624,
+      "loss": 2.809,
+      "step": 998
+    },
+    {
+      "epoch": 0.08062303284642079,
+      "grad_norm": 0.8915852308273315,
+      "learning_rate": 0.00019899456528254267,
+      "loss": 2.8309,
+      "step": 999
+    },
+    {
+      "epoch": 0.0807037365830038,
+      "grad_norm": 0.8092094659805298,
+      "learning_rate": 0.00019899233102005573,
+      "loss": 2.7753,
+      "step": 1000
+    },
+    {
+      "epoch": 0.0807037365830038,
+      "eval_loss": 2.7104671001434326,
+      "eval_runtime": 773.7354,
+      "eval_samples_per_second": 3.386,
+      "eval_steps_per_second": 0.565,
+      "step": 1000
+    },
+    {
+      "epoch": 0.0807844403195868,
+      "grad_norm": 0.8744900226593018,
+      "learning_rate": 0.00019899009429042114,
+      "loss": 2.7948,
+      "step": 1001
+    },
+    {
+      "epoch": 0.0808651440561698,
+      "grad_norm": 0.8749974370002747,
+      "learning_rate": 0.0001989878550936946,
+      "loss": 2.7609,
+      "step": 1002
+    },
+    {
+      "epoch": 0.08094584779275281,
+      "grad_norm": 0.8622820377349854,
+      "learning_rate": 0.000198985613429932,
+      "loss": 2.8023,
+      "step": 1003
+    },
+    {
+      "epoch": 0.08102655152933581,
+      "grad_norm": 0.9404367208480835,
+      "learning_rate": 0.00019898336929918915,
+      "loss": 2.7992,
+      "step": 1004
+    },
+    {
+      "epoch": 0.08110725526591882,
+      "grad_norm": 0.8846708536148071,
+      "learning_rate": 0.000198981122701522,
+      "loss": 2.8084,
+      "step": 1005
+    },
+    {
+      "epoch": 0.08118795900250182,
+      "grad_norm": 0.8105908036231995,
+      "learning_rate": 0.0001989788736369865,
+      "loss": 2.8504,
+      "step": 1006
+    },
+    {
+      "epoch": 0.08126866273908483,
+      "grad_norm": 1.0107187032699585,
+      "learning_rate": 0.0001989766221056388,
+      "loss": 2.7935,
+      "step": 1007
+    },
+    {
+      "epoch": 0.08134936647566782,
+      "grad_norm": 0.7825451493263245,
+      "learning_rate": 0.0001989743681075349,
+      "loss": 2.8024,
+      "step": 1008
+    },
+    {
+      "epoch": 0.08143007021225082,
+      "grad_norm": 0.8478613495826721,
+      "learning_rate": 0.000198972111642731,
+      "loss": 2.8645,
+      "step": 1009
+    },
+    {
+      "epoch": 0.08151077394883383,
+      "grad_norm": 0.8432144522666931,
+      "learning_rate": 0.0001989698527112834,
+      "loss": 2.8469,
+      "step": 1010
+    },
+    {
+      "epoch": 0.08159147768541683,
+      "grad_norm": 0.8147936463356018,
+      "learning_rate": 0.00019896759131324835,
+      "loss": 2.7799,
+      "step": 1011
+    },
+    {
+      "epoch": 0.08167218142199983,
+      "grad_norm": 0.8446993827819824,
+      "learning_rate": 0.00019896532744868224,
+      "loss": 2.7685,
+      "step": 1012
+    },
+    {
+      "epoch": 0.08175288515858284,
+      "grad_norm": 0.7635807394981384,
+      "learning_rate": 0.00019896306111764146,
+      "loss": 2.7823,
+      "step": 1013
+    },
+    {
+      "epoch": 0.08183358889516584,
+      "grad_norm": 0.8272855877876282,
+      "learning_rate": 0.00019896079232018253,
+      "loss": 2.7877,
+      "step": 1014
+    },
+    {
+      "epoch": 0.08191429263174885,
+      "grad_norm": 0.8079700469970703,
+      "learning_rate": 0.00019895852105636193,
+      "loss": 2.7849,
+      "step": 1015
+    },
+    {
+      "epoch": 0.08199499636833185,
+      "grad_norm": 0.8518063426017761,
+      "learning_rate": 0.0001989562473262363,
+      "loss": 2.8622,
+      "step": 1016
+    },
+    {
+      "epoch": 0.08207570010491486,
+      "grad_norm": 0.8646622896194458,
+      "learning_rate": 0.00019895397112986235,
+      "loss": 2.8224,
+      "step": 1017
+    },
+    {
+      "epoch": 0.08215640384149786,
+      "grad_norm": 0.8764398097991943,
+      "learning_rate": 0.00019895169246729672,
+      "loss": 2.938,
+      "step": 1018
+    },
+    {
+      "epoch": 0.08223710757808086,
+      "grad_norm": 0.8304057717323303,
+      "learning_rate": 0.0001989494113385963,
+      "loss": 2.7586,
+      "step": 1019
+    },
+    {
+      "epoch": 0.08231781131466387,
+      "grad_norm": 0.8569272756576538,
+      "learning_rate": 0.00019894712774381787,
+      "loss": 2.7803,
+      "step": 1020
+    },
+    {
+      "epoch": 0.08239851505124687,
+      "grad_norm": 0.8788578510284424,
+      "learning_rate": 0.00019894484168301836,
+      "loss": 2.8138,
+      "step": 1021
+    },
+    {
+      "epoch": 0.08247921878782988,
+      "grad_norm": 0.9113569855690002,
+      "learning_rate": 0.0001989425531562548,
+      "loss": 2.8023,
+      "step": 1022
+    },
+    {
+      "epoch": 0.08255992252441288,
+      "grad_norm": 0.8630590438842773,
+      "learning_rate": 0.00019894026216358413,
+      "loss": 2.791,
+      "step": 1023
+    },
+    {
+      "epoch": 0.08264062626099589,
+      "grad_norm": 0.8691157698631287,
+      "learning_rate": 0.00019893796870506348,
+      "loss": 2.811,
+      "step": 1024
+    },
+    {
+      "epoch": 0.08272132999757889,
+      "grad_norm": 0.9078284502029419,
+      "learning_rate": 0.00019893567278075007,
+      "loss": 2.8282,
+      "step": 1025
+    },
+    {
+      "epoch": 0.0828020337341619,
+      "grad_norm": 0.867511510848999,
+      "learning_rate": 0.00019893337439070105,
+      "loss": 2.7862,
+      "step": 1026
+    },
+    {
+      "epoch": 0.0828827374707449,
+      "grad_norm": 0.8016698360443115,
+      "learning_rate": 0.00019893107353497372,
+      "loss": 2.8083,
+      "step": 1027
+    },
+    {
+      "epoch": 0.0829634412073279,
+      "grad_norm": 0.8583545684814453,
+      "learning_rate": 0.00019892877021362543,
+      "loss": 2.8041,
+      "step": 1028
+    },
+    {
+      "epoch": 0.08304414494391091,
+      "grad_norm": 0.8302493691444397,
+      "learning_rate": 0.0001989264644267136,
+      "loss": 2.7866,
+      "step": 1029
+    },
+    {
+      "epoch": 0.08312484868049391,
+      "grad_norm": 0.9628411531448364,
+      "learning_rate": 0.00019892415617429567,
+      "loss": 2.8187,
+      "step": 1030
+    },
+    {
+      "epoch": 0.08320555241707692,
+      "grad_norm": 0.874840259552002,
+      "learning_rate": 0.0001989218454564292,
+      "loss": 2.7475,
+      "step": 1031
+    },
+    {
+      "epoch": 0.08328625615365992,
+      "grad_norm": 0.8641294836997986,
+      "learning_rate": 0.0001989195322731717,
+      "loss": 2.7795,
+      "step": 1032
+    },
+    {
+      "epoch": 0.08336695989024291,
+      "grad_norm": 0.8219757080078125,
+      "learning_rate": 0.0001989172166245809,
+      "loss": 2.7683,
+      "step": 1033
+    },
+    {
+      "epoch": 0.08344766362682592,
+      "grad_norm": 0.7905694246292114,
+      "learning_rate": 0.00019891489851071455,
+      "loss": 2.7668,
+      "step": 1034
+    },
+    {
+      "epoch": 0.08352836736340892,
+      "grad_norm": 0.8180816173553467,
+      "learning_rate": 0.0001989125779316303,
+      "loss": 2.7661,
+      "step": 1035
+    },
+    {
+      "epoch": 0.08360907109999192,
+      "grad_norm": 0.8337293267250061,
+      "learning_rate": 0.00019891025488738605,
+      "loss": 2.7823,
+      "step": 1036
+    },
+    {
+      "epoch": 0.08368977483657493,
+      "grad_norm": 0.9673140048980713,
+      "learning_rate": 0.00019890792937803973,
+      "loss": 2.8164,
+      "step": 1037
+    },
+    {
+      "epoch": 0.08377047857315793,
+      "grad_norm": 0.8810501098632812,
+      "learning_rate": 0.00019890560140364922,
+      "loss": 2.7904,
+      "step": 1038
+    },
+    {
+      "epoch": 0.08385118230974094,
+      "grad_norm": 0.9507614374160767,
+      "learning_rate": 0.0001989032709642726,
+      "loss": 2.7928,
+      "step": 1039
+    },
+    {
+      "epoch": 0.08393188604632394,
+      "grad_norm": 0.953738808631897,
+      "learning_rate": 0.00019890093805996793,
+      "loss": 2.7922,
+      "step": 1040
+    },
+    {
+      "epoch": 0.08401258978290695,
+      "grad_norm": 0.8079931139945984,
+      "learning_rate": 0.00019889860269079336,
+      "loss": 2.7909,
+      "step": 1041
+    },
+    {
+      "epoch": 0.08409329351948995,
+      "grad_norm": 1.0330647230148315,
+      "learning_rate": 0.0001988962648568071,
+      "loss": 2.7526,
+      "step": 1042
+    },
+    {
+      "epoch": 0.08417399725607295,
+      "grad_norm": 0.8988988399505615,
+      "learning_rate": 0.00019889392455806738,
+      "loss": 2.7471,
+      "step": 1043
+    },
+    {
+      "epoch": 0.08425470099265596,
+      "grad_norm": 0.7986348271369934,
+      "learning_rate": 0.00019889158179463255,
+      "loss": 2.7208,
+      "step": 1044
+    },
+    {
+      "epoch": 0.08433540472923896,
+      "grad_norm": 0.9231631755828857,
+      "learning_rate": 0.000198889236566561,
+      "loss": 2.7953,
+      "step": 1045
+    },
+    {
+      "epoch": 0.08441610846582197,
+      "grad_norm": 0.8438155055046082,
+      "learning_rate": 0.00019888688887391117,
+      "loss": 2.8006,
+      "step": 1046
+    },
+    {
+      "epoch": 0.08449681220240497,
+      "grad_norm": 0.8915219306945801,
+      "learning_rate": 0.0001988845387167416,
+      "loss": 2.8184,
+      "step": 1047
+    },
+    {
+      "epoch": 0.08457751593898798,
+      "grad_norm": 0.924401581287384,
+      "learning_rate": 0.0001988821860951108,
+      "loss": 2.8411,
+      "step": 1048
+    },
+    {
+      "epoch": 0.08465821967557098,
+      "grad_norm": 0.8144630193710327,
+      "learning_rate": 0.00019887983100907745,
+      "loss": 2.8258,
+      "step": 1049
+    },
+    {
+      "epoch": 0.08473892341215399,
+      "grad_norm": 0.9974459409713745,
+      "learning_rate": 0.00019887747345870028,
+      "loss": 2.7567,
+      "step": 1050
+    },
+    {
+      "epoch": 0.08481962714873699,
+      "grad_norm": 0.944526195526123,
+      "learning_rate": 0.00019887511344403796,
+      "loss": 2.8657,
+      "step": 1051
+    },
+    {
+      "epoch": 0.08490033088532,
+      "grad_norm": 0.8204831480979919,
+      "learning_rate": 0.00019887275096514936,
+      "loss": 2.8054,
+      "step": 1052
+    },
+    {
+      "epoch": 0.084981034621903,
+      "grad_norm": 0.8855900168418884,
+      "learning_rate": 0.00019887038602209336,
+      "loss": 2.8019,
+      "step": 1053
+    },
+    {
+      "epoch": 0.085061738358486,
+      "grad_norm": 0.9025108814239502,
+      "learning_rate": 0.0001988680186149289,
+      "loss": 2.7934,
+      "step": 1054
+    },
+    {
+      "epoch": 0.08514244209506901,
+      "grad_norm": 0.8486441373825073,
+      "learning_rate": 0.00019886564874371494,
+      "loss": 2.809,
+      "step": 1055
+    },
+    {
+      "epoch": 0.08522314583165201,
+      "grad_norm": 0.778364896774292,
+      "learning_rate": 0.00019886327640851058,
+      "loss": 2.7783,
+      "step": 1056
+    },
+    {
+      "epoch": 0.08530384956823502,
+      "grad_norm": 0.8515299558639526,
+      "learning_rate": 0.00019886090160937497,
+      "loss": 2.8122,
+      "step": 1057
+    },
+    {
+      "epoch": 0.08538455330481802,
+      "grad_norm": 0.8466131091117859,
+      "learning_rate": 0.00019885852434636724,
+      "loss": 2.7798,
+      "step": 1058
+    },
+    {
+      "epoch": 0.08546525704140101,
+      "grad_norm": 0.8856541514396667,
+      "learning_rate": 0.00019885614461954667,
+      "loss": 2.8033,
+      "step": 1059
+    },
+    {
+      "epoch": 0.08554596077798401,
+      "grad_norm": 0.8853924870491028,
+      "learning_rate": 0.00019885376242897258,
+      "loss": 2.8368,
+      "step": 1060
+    },
+    {
+      "epoch": 0.08562666451456702,
+      "grad_norm": 0.7858660221099854,
+      "learning_rate": 0.0001988513777747043,
+      "loss": 2.7806,
+      "step": 1061
+    },
+    {
+      "epoch": 0.08570736825115002,
+      "grad_norm": 0.8601513504981995,
+      "learning_rate": 0.0001988489906568013,
+      "loss": 2.8434,
+      "step": 1062
+    },
+    {
+      "epoch": 0.08578807198773303,
+      "grad_norm": 0.9126001596450806,
+      "learning_rate": 0.00019884660107532306,
+      "loss": 2.8469,
+      "step": 1063
+    },
+    {
+      "epoch": 0.08586877572431603,
+      "grad_norm": 0.9016061425209045,
+      "learning_rate": 0.00019884420903032912,
+      "loss": 2.7907,
+      "step": 1064
+    },
+    {
+      "epoch": 0.08594947946089904,
+      "grad_norm": 0.9134494066238403,
+      "learning_rate": 0.00019884181452187915,
+      "loss": 2.8426,
+      "step": 1065
+    },
+    {
+      "epoch": 0.08603018319748204,
+      "grad_norm": 0.8891138434410095,
+      "learning_rate": 0.00019883941755003272,
+      "loss": 2.8092,
+      "step": 1066
+    },
+    {
+      "epoch": 0.08611088693406505,
+      "grad_norm": 0.822884202003479,
+      "learning_rate": 0.0001988370181148497,
+      "loss": 2.8454,
+      "step": 1067
+    },
+    {
+      "epoch": 0.08619159067064805,
+      "grad_norm": 0.8341901898384094,
+      "learning_rate": 0.0001988346162163898,
+      "loss": 2.8027,
+      "step": 1068
+    },
+    {
+      "epoch": 0.08627229440723105,
+      "grad_norm": 0.8653229475021362,
+      "learning_rate": 0.00019883221185471291,
+      "loss": 2.7487,
+      "step": 1069
+    },
+    {
+      "epoch": 0.08635299814381406,
+      "grad_norm": 0.8065966367721558,
+      "learning_rate": 0.00019882980502987894,
+      "loss": 2.7847,
+      "step": 1070
+    },
+    {
+      "epoch": 0.08643370188039706,
+      "grad_norm": 0.9106903076171875,
+      "learning_rate": 0.0001988273957419479,
+      "loss": 2.7962,
+      "step": 1071
+    },
+    {
+      "epoch": 0.08651440561698007,
+      "grad_norm": 0.953815221786499,
+      "learning_rate": 0.0001988249839909798,
+      "loss": 2.8168,
+      "step": 1072
+    },
+    {
+      "epoch": 0.08659510935356307,
+      "grad_norm": 0.8642842173576355,
+      "learning_rate": 0.00019882256977703477,
+      "loss": 2.8205,
+      "step": 1073
+    },
+    {
+      "epoch": 0.08667581309014608,
+      "grad_norm": 0.8500350117683411,
+      "learning_rate": 0.000198820153100173,
+      "loss": 2.8798,
+      "step": 1074
+    },
+    {
+      "epoch": 0.08675651682672908,
+      "grad_norm": 0.9212989807128906,
+      "learning_rate": 0.00019881773396045467,
+      "loss": 2.8088,
+      "step": 1075
+    },
+    {
+      "epoch": 0.08683722056331208,
+      "grad_norm": 0.8897970914840698,
+      "learning_rate": 0.0001988153123579401,
+      "loss": 2.7983,
+      "step": 1076
+    },
+    {
+      "epoch": 0.08691792429989509,
+      "grad_norm": 0.7942636609077454,
+      "learning_rate": 0.00019881288829268968,
+      "loss": 2.7711,
+      "step": 1077
+    },
+    {
+      "epoch": 0.08699862803647809,
+      "grad_norm": 0.8286700248718262,
+      "learning_rate": 0.00019881046176476374,
+      "loss": 2.7995,
+      "step": 1078
+    },
+    {
+      "epoch": 0.0870793317730611,
+      "grad_norm": 0.9436343908309937,
+      "learning_rate": 0.00019880803277422281,
+      "loss": 2.8399,
+      "step": 1079
+    },
+    {
+      "epoch": 0.0871600355096441,
+      "grad_norm": 0.9592518210411072,
+      "learning_rate": 0.00019880560132112742,
+      "loss": 2.7888,
+      "step": 1080
+    },
+    {
+      "epoch": 0.0872407392462271,
+      "grad_norm": 0.8956589698791504,
+      "learning_rate": 0.00019880316740553816,
+      "loss": 2.7635,
+      "step": 1081
+    },
+    {
+      "epoch": 0.08732144298281011,
+      "grad_norm": 1.055312156677246,
+      "learning_rate": 0.00019880073102751574,
+      "loss": 2.7778,
+      "step": 1082
+    },
+    {
+      "epoch": 0.08740214671939311,
+      "grad_norm": 0.783273458480835,
+      "learning_rate": 0.00019879829218712075,
+      "loss": 2.735,
+      "step": 1083
+    },
+    {
+      "epoch": 0.0874828504559761,
+      "grad_norm": 0.8315421938896179,
+      "learning_rate": 0.00019879585088441413,
+      "loss": 2.7973,
+      "step": 1084
+    },
+    {
+      "epoch": 0.08756355419255911,
+      "grad_norm": 0.9550945162773132,
+      "learning_rate": 0.00019879340711945662,
+      "loss": 2.8083,
+      "step": 1085
+    },
+    {
+      "epoch": 0.08764425792914211,
+      "grad_norm": 0.9579277634620667,
+      "learning_rate": 0.00019879096089230915,
+      "loss": 2.7411,
+      "step": 1086
+    },
+    {
+      "epoch": 0.08772496166572512,
+      "grad_norm": 0.8602219223976135,
+      "learning_rate": 0.0001987885122030327,
+      "loss": 2.7461,
+      "step": 1087
+    },
+    {
+      "epoch": 0.08780566540230812,
+      "grad_norm": 0.9749068021774292,
+      "learning_rate": 0.00019878606105168829,
+      "loss": 2.7701,
+      "step": 1088
+    },
+    {
+      "epoch": 0.08788636913889113,
+      "grad_norm": 0.8128982186317444,
+      "learning_rate": 0.00019878360743833703,
+      "loss": 2.7949,
+      "step": 1089
+    },
+    {
+      "epoch": 0.08796707287547413,
+      "grad_norm": 0.9177080988883972,
+      "learning_rate": 0.00019878115136304003,
+      "loss": 2.7471,
+      "step": 1090
+    },
+    {
+      "epoch": 0.08804777661205714,
+      "grad_norm": 0.9052132368087769,
+      "learning_rate": 0.0001987786928258585,
+      "loss": 2.8356,
+      "step": 1091
+    },
+    {
+      "epoch": 0.08812848034864014,
+      "grad_norm": 0.8972994089126587,
+      "learning_rate": 0.00019877623182685378,
+      "loss": 2.8304,
+      "step": 1092
+    },
+    {
+      "epoch": 0.08820918408522314,
+      "grad_norm": 0.861251950263977,
+      "learning_rate": 0.0001987737683660871,
+      "loss": 2.8436,
+      "step": 1093
+    },
+    {
+      "epoch": 0.08828988782180615,
+      "grad_norm": 0.9139869809150696,
+      "learning_rate": 0.00019877130244361996,
+      "loss": 2.7583,
+      "step": 1094
+    },
+    {
+      "epoch": 0.08837059155838915,
+      "grad_norm": 0.8441170454025269,
+      "learning_rate": 0.00019876883405951377,
+      "loss": 2.7508,
+      "step": 1095
+    },
+    {
+      "epoch": 0.08845129529497216,
+      "grad_norm": 0.8624769449234009,
+      "learning_rate": 0.00019876636321383004,
+      "loss": 2.8003,
+      "step": 1096
+    },
+    {
+      "epoch": 0.08853199903155516,
+      "grad_norm": 0.9033877849578857,
+      "learning_rate": 0.00019876388990663037,
+      "loss": 2.7934,
+      "step": 1097
+    },
+    {
+      "epoch": 0.08861270276813817,
+      "grad_norm": 0.9492632746696472,
+      "learning_rate": 0.0001987614141379764,
+      "loss": 2.7852,
+      "step": 1098
+    },
+    {
+      "epoch": 0.08869340650472117,
+      "grad_norm": 0.9004682302474976,
+      "learning_rate": 0.00019875893590792982,
+      "loss": 2.7518,
+      "step": 1099
+    },
+    {
+      "epoch": 0.08877411024130417,
+      "grad_norm": 0.8352272510528564,
+      "learning_rate": 0.0001987564552165524,
+      "loss": 2.8035,
+      "step": 1100
+    },
+    {
+      "epoch": 0.08885481397788718,
+      "grad_norm": 0.8488562107086182,
+      "learning_rate": 0.00019875397206390593,
+      "loss": 2.7672,
+      "step": 1101
+    },
+    {
+      "epoch": 0.08893551771447018,
+      "grad_norm": 0.9450985193252563,
+      "learning_rate": 0.00019875148645005238,
+      "loss": 2.7558,
+      "step": 1102
+    },
+    {
+      "epoch": 0.08901622145105319,
+      "grad_norm": 0.9203561544418335,
+      "learning_rate": 0.0001987489983750536,
+      "loss": 2.7983,
+      "step": 1103
+    },
+    {
+      "epoch": 0.08909692518763619,
+      "grad_norm": 0.8761897087097168,
+      "learning_rate": 0.0001987465078389717,
+      "loss": 2.7536,
+      "step": 1104
+    },
+    {
+      "epoch": 0.0891776289242192,
+      "grad_norm": 0.9064637422561646,
+      "learning_rate": 0.00019874401484186867,
+      "loss": 2.8104,
+      "step": 1105
+    },
+    {
+      "epoch": 0.0892583326608022,
+      "grad_norm": 0.8394999504089355,
+      "learning_rate": 0.00019874151938380666,
+      "loss": 2.7459,
+      "step": 1106
+    },
+    {
+      "epoch": 0.0893390363973852,
+      "grad_norm": 0.8782099485397339,
+      "learning_rate": 0.00019873902146484785,
+      "loss": 2.8675,
+      "step": 1107
+    },
+    {
+      "epoch": 0.08941974013396821,
+      "grad_norm": 0.8564850091934204,
+      "learning_rate": 0.00019873652108505458,
+      "loss": 2.8561,
+      "step": 1108
+    },
+    {
+      "epoch": 0.08950044387055121,
+      "grad_norm": 0.8343809843063354,
+      "learning_rate": 0.0001987340182444891,
+      "loss": 2.8406,
+      "step": 1109
+    },
+    {
+      "epoch": 0.0895811476071342,
+      "grad_norm": 1.096273422241211,
+      "learning_rate": 0.00019873151294321376,
+      "loss": 2.8264,
+      "step": 1110
+    },
+    {
+      "epoch": 0.08966185134371721,
+      "grad_norm": 0.8654618263244629,
+      "learning_rate": 0.00019872900518129103,
+      "loss": 2.7956,
+      "step": 1111
+    },
+    {
+      "epoch": 0.08974255508030021,
+      "grad_norm": 0.8868138194084167,
+      "learning_rate": 0.00019872649495878344,
+      "loss": 2.8028,
+      "step": 1112
+    },
+    {
+      "epoch": 0.08982325881688322,
+      "grad_norm": 0.8139104843139648,
+      "learning_rate": 0.00019872398227575348,
+      "loss": 2.7502,
+      "step": 1113
+    },
+    {
+      "epoch": 0.08990396255346622,
+      "grad_norm": 0.8277762532234192,
+      "learning_rate": 0.00019872146713226384,
+      "loss": 2.7913,
+      "step": 1114
+    },
+    {
+      "epoch": 0.08998466629004923,
+      "grad_norm": 0.8470397591590881,
+      "learning_rate": 0.00019871894952837717,
+      "loss": 2.7982,
+      "step": 1115
+    },
+    {
+      "epoch": 0.09006537002663223,
+      "grad_norm": 0.8424760103225708,
+      "learning_rate": 0.00019871642946415625,
+      "loss": 2.8067,
+      "step": 1116
+    },
+    {
+      "epoch": 0.09014607376321523,
+      "grad_norm": 0.8253894448280334,
+      "learning_rate": 0.00019871390693966382,
+      "loss": 2.8339,
+      "step": 1117
+    },
+    {
+      "epoch": 0.09022677749979824,
+      "grad_norm": 0.8120691776275635,
+      "learning_rate": 0.00019871138195496282,
+      "loss": 2.7938,
+      "step": 1118
+    },
+    {
+      "epoch": 0.09030748123638124,
+      "grad_norm": 0.920189619064331,
+      "learning_rate": 0.00019870885451011617,
+      "loss": 2.8083,
+      "step": 1119
+    },
+    {
+      "epoch": 0.09038818497296425,
+      "grad_norm": 0.8990969657897949,
+      "learning_rate": 0.0001987063246051868,
+      "loss": 2.7481,
+      "step": 1120
+    },
+    {
+      "epoch": 0.09046888870954725,
+      "grad_norm": 0.8280801773071289,
+      "learning_rate": 0.0001987037922402378,
+      "loss": 2.8536,
+      "step": 1121
+    },
+    {
+      "epoch": 0.09054959244613026,
+      "grad_norm": 0.8510503768920898,
+      "learning_rate": 0.0001987012574153323,
+      "loss": 2.758,
+      "step": 1122
+    },
+    {
+      "epoch": 0.09063029618271326,
+      "grad_norm": 0.9103946685791016,
+      "learning_rate": 0.00019869872013053344,
+      "loss": 2.7594,
+      "step": 1123
+    },
+    {
+      "epoch": 0.09071099991929626,
+      "grad_norm": 0.804916262626648,
+      "learning_rate": 0.00019869618038590448,
+      "loss": 2.7489,
+      "step": 1124
+    },
+    {
+      "epoch": 0.09079170365587927,
+      "grad_norm": 0.7542802095413208,
+      "learning_rate": 0.00019869363818150867,
+      "loss": 2.76,
+      "step": 1125
+    },
+    {
+      "epoch": 0.09087240739246227,
+      "grad_norm": 0.7725108861923218,
+      "learning_rate": 0.00019869109351740947,
+      "loss": 2.8124,
+      "step": 1126
+    },
+    {
+      "epoch": 0.09095311112904528,
+      "grad_norm": 0.8533692955970764,
+      "learning_rate": 0.0001986885463936702,
+      "loss": 2.8499,
+      "step": 1127
+    },
+    {
+      "epoch": 0.09103381486562828,
+      "grad_norm": 0.8351541757583618,
+      "learning_rate": 0.0001986859968103544,
+      "loss": 2.8075,
+      "step": 1128
+    },
+    {
+      "epoch": 0.09111451860221129,
+      "grad_norm": 0.8780044913291931,
+      "learning_rate": 0.0001986834447675256,
+      "loss": 2.7587,
+      "step": 1129
+    },
+    {
+      "epoch": 0.09119522233879429,
+      "grad_norm": 0.9587519764900208,
+      "learning_rate": 0.00019868089026524736,
+      "loss": 2.8069,
+      "step": 1130
+    },
+    {
+      "epoch": 0.0912759260753773,
+      "grad_norm": 0.8285651206970215,
+      "learning_rate": 0.00019867833330358342,
+      "loss": 2.8209,
+      "step": 1131
+    },
+    {
+      "epoch": 0.0913566298119603,
+      "grad_norm": 0.8589211106300354,
+      "learning_rate": 0.00019867577388259745,
+      "loss": 2.8144,
+      "step": 1132
+    },
+    {
+      "epoch": 0.0914373335485433,
+      "grad_norm": 0.8740364909172058,
+      "learning_rate": 0.00019867321200235324,
+      "loss": 2.858,
+      "step": 1133
+    },
+    {
+      "epoch": 0.09151803728512631,
+      "grad_norm": 0.8368108868598938,
+      "learning_rate": 0.00019867064766291467,
+      "loss": 2.7997,
+      "step": 1134
+    },
+    {
+      "epoch": 0.0915987410217093,
+      "grad_norm": 0.8243690133094788,
+      "learning_rate": 0.00019866808086434564,
+      "loss": 2.7925,
+      "step": 1135
+    },
+    {
+      "epoch": 0.0916794447582923,
+      "grad_norm": 0.8296996355056763,
+      "learning_rate": 0.0001986655116067101,
+      "loss": 2.7953,
+      "step": 1136
+    },
+    {
+      "epoch": 0.09176014849487531,
+      "grad_norm": 0.9255942702293396,
+      "learning_rate": 0.0001986629398900721,
+      "loss": 2.844,
+      "step": 1137
+    },
+    {
+      "epoch": 0.09184085223145831,
+      "grad_norm": 0.7498174905776978,
+      "learning_rate": 0.00019866036571449574,
+      "loss": 2.7372,
+      "step": 1138
+    },
+    {
+      "epoch": 0.09192155596804132,
+      "grad_norm": 0.8170139193534851,
+      "learning_rate": 0.00019865778908004513,
+      "loss": 2.7656,
+      "step": 1139
+    },
+    {
+      "epoch": 0.09200225970462432,
+      "grad_norm": 0.8858106732368469,
+      "learning_rate": 0.00019865520998678458,
+      "loss": 2.7657,
+      "step": 1140
+    },
+    {
+      "epoch": 0.09208296344120732,
+      "grad_norm": 0.8789847493171692,
+      "learning_rate": 0.00019865262843477826,
+      "loss": 2.8419,
+      "step": 1141
+    },
+    {
+      "epoch": 0.09216366717779033,
+      "grad_norm": 0.8433314561843872,
+      "learning_rate": 0.00019865004442409058,
+      "loss": 2.7981,
+      "step": 1142
+    },
+    {
+      "epoch": 0.09224437091437333,
+      "grad_norm": 0.8822595477104187,
+      "learning_rate": 0.0001986474579547859,
+      "loss": 2.8368,
+      "step": 1143
+    },
+    {
+      "epoch": 0.09232507465095634,
+      "grad_norm": 0.9067013263702393,
+      "learning_rate": 0.00019864486902692872,
+      "loss": 2.7807,
+      "step": 1144
+    },
+    {
+      "epoch": 0.09240577838753934,
+      "grad_norm": 0.9551558494567871,
+      "learning_rate": 0.00019864227764058355,
+      "loss": 2.7617,
+      "step": 1145
+    },
+    {
+      "epoch": 0.09248648212412235,
+      "grad_norm": 0.8337206244468689,
+      "learning_rate": 0.00019863968379581494,
+      "loss": 2.8289,
+      "step": 1146
+    },
+    {
+      "epoch": 0.09256718586070535,
+      "grad_norm": 0.952702522277832,
+      "learning_rate": 0.0001986370874926876,
+      "loss": 2.8508,
+      "step": 1147
+    },
+    {
+      "epoch": 0.09264788959728835,
+      "grad_norm": 0.8586699366569519,
+      "learning_rate": 0.00019863448873126615,
+      "loss": 2.8784,
+      "step": 1148
+    },
+    {
+      "epoch": 0.09272859333387136,
+      "grad_norm": 0.7625309228897095,
+      "learning_rate": 0.00019863188751161544,
+      "loss": 2.7936,
+      "step": 1149
+    },
+    {
+      "epoch": 0.09280929707045436,
+      "grad_norm": 0.8912700414657593,
+      "learning_rate": 0.0001986292838338003,
+      "loss": 2.8745,
+      "step": 1150
+    },
+    {
+      "epoch": 0.09289000080703737,
+      "grad_norm": 0.8618904948234558,
+      "learning_rate": 0.00019862667769788553,
+      "loss": 2.8086,
+      "step": 1151
+    },
+    {
+      "epoch": 0.09297070454362037,
+      "grad_norm": 1.0013352632522583,
+      "learning_rate": 0.00019862406910393617,
+      "loss": 2.8211,
+      "step": 1152
+    },
+    {
+      "epoch": 0.09305140828020338,
+      "grad_norm": 0.7922475337982178,
+      "learning_rate": 0.0001986214580520172,
+      "loss": 2.7668,
+      "step": 1153
+    },
+    {
+      "epoch": 0.09313211201678638,
+      "grad_norm": 0.9490330815315247,
+      "learning_rate": 0.00019861884454219365,
+      "loss": 2.7571,
+      "step": 1154
+    },
+    {
+      "epoch": 0.09321281575336939,
+      "grad_norm": 0.8780270218849182,
+      "learning_rate": 0.00019861622857453076,
+      "loss": 2.7598,
+      "step": 1155
+    },
+    {
+      "epoch": 0.09329351948995239,
+      "grad_norm": 0.9220066070556641,
+      "learning_rate": 0.00019861361014909365,
+      "loss": 2.7609,
+      "step": 1156
+    },
+    {
+      "epoch": 0.0933742232265354,
+      "grad_norm": 0.8299020528793335,
+      "learning_rate": 0.0001986109892659476,
+      "loss": 2.8655,
+      "step": 1157
+    },
+    {
+      "epoch": 0.0934549269631184,
+      "grad_norm": 0.9700348377227783,
+      "learning_rate": 0.0001986083659251579,
+      "loss": 2.8597,
+      "step": 1158
+    },
+    {
+      "epoch": 0.0935356306997014,
+      "grad_norm": 0.8820784687995911,
+      "learning_rate": 0.00019860574012679001,
+      "loss": 2.8776,
+      "step": 1159
+    },
+    {
+      "epoch": 0.0936163344362844,
+      "grad_norm": 0.8134172558784485,
+      "learning_rate": 0.0001986031118709093,
+      "loss": 2.8163,
+      "step": 1160
+    },
+    {
+      "epoch": 0.0936970381728674,
+      "grad_norm": 0.885974109172821,
+      "learning_rate": 0.00019860048115758123,
+      "loss": 2.752,
+      "step": 1161
+    },
+    {
+      "epoch": 0.0937777419094504,
+      "grad_norm": 0.9650186896324158,
+      "learning_rate": 0.0001985978479868715,
+      "loss": 2.7587,
+      "step": 1162
+    },
+    {
+      "epoch": 0.0938584456460334,
+      "grad_norm": 0.8550445437431335,
+      "learning_rate": 0.00019859521235884563,
+      "loss": 2.7887,
+      "step": 1163
+    },
+    {
+      "epoch": 0.09393914938261641,
+      "grad_norm": 0.9686560034751892,
+      "learning_rate": 0.00019859257427356933,
+      "loss": 2.7974,
+      "step": 1164
+    },
+    {
+      "epoch": 0.09401985311919941,
+      "grad_norm": 0.9185387492179871,
+      "learning_rate": 0.00019858993373110837,
+      "loss": 2.7933,
+      "step": 1165
+    },
+    {
+      "epoch": 0.09410055685578242,
+      "grad_norm": 0.9549610018730164,
+      "learning_rate": 0.00019858729073152852,
+      "loss": 2.7698,
+      "step": 1166
+    },
+    {
+      "epoch": 0.09418126059236542,
+      "grad_norm": 1.0523492097854614,
+      "learning_rate": 0.0001985846452748957,
+      "loss": 2.7215,
+      "step": 1167
+    },
+    {
+      "epoch": 0.09426196432894843,
+      "grad_norm": 0.8551118969917297,
+      "learning_rate": 0.00019858199736127582,
+      "loss": 2.805,
+      "step": 1168
+    },
+    {
+      "epoch": 0.09434266806553143,
+      "grad_norm": 1.021374225616455,
+      "learning_rate": 0.0001985793469907349,
+      "loss": 2.794,
+      "step": 1169
+    },
+    {
+      "epoch": 0.09442337180211444,
+      "grad_norm": 0.8745501041412354,
+      "learning_rate": 0.0001985766941633389,
+      "loss": 2.7793,
+      "step": 1170
+    },
+    {
+      "epoch": 0.09450407553869744,
+      "grad_norm": 0.7426434755325317,
+      "learning_rate": 0.00019857403887915402,
+      "loss": 2.7808,
+      "step": 1171
+    },
+    {
+      "epoch": 0.09458477927528045,
+      "grad_norm": 0.9183726906776428,
+      "learning_rate": 0.0001985713811382464,
+      "loss": 2.8001,
+      "step": 1172
+    },
+    {
+      "epoch": 0.09466548301186345,
+      "grad_norm": 0.8136709928512573,
+      "learning_rate": 0.00019856872094068233,
+      "loss": 2.7394,
+      "step": 1173
+    },
+    {
+      "epoch": 0.09474618674844645,
+      "grad_norm": 0.9399348497390747,
+      "learning_rate": 0.00019856605828652807,
+      "loss": 2.7733,
+      "step": 1174
+    },
+    {
+      "epoch": 0.09482689048502946,
+      "grad_norm": 0.8233176469802856,
+      "learning_rate": 0.00019856339317584997,
+      "loss": 2.7672,
+      "step": 1175
+    },
+    {
+      "epoch": 0.09490759422161246,
+      "grad_norm": 0.9157048463821411,
+      "learning_rate": 0.00019856072560871447,
+      "loss": 2.7992,
+      "step": 1176
+    },
+    {
+      "epoch": 0.09498829795819547,
+      "grad_norm": 0.8729545474052429,
+      "learning_rate": 0.00019855805558518803,
+      "loss": 2.749,
+      "step": 1177
+    },
+    {
+      "epoch": 0.09506900169477847,
+      "grad_norm": 0.8592300415039062,
+      "learning_rate": 0.00019855538310533722,
+      "loss": 2.7257,
+      "step": 1178
+    },
+    {
+      "epoch": 0.09514970543136148,
+      "grad_norm": 0.8470803499221802,
+      "learning_rate": 0.00019855270816922867,
+      "loss": 2.7479,
+      "step": 1179
+    },
+    {
+      "epoch": 0.09523040916794448,
+      "grad_norm": 0.8538667559623718,
+      "learning_rate": 0.00019855003077692897,
+      "loss": 2.7576,
+      "step": 1180
+    },
+    {
+      "epoch": 0.09531111290452748,
+      "grad_norm": 0.8890984654426575,
+      "learning_rate": 0.0001985473509285049,
+      "loss": 2.7961,
+      "step": 1181
+    },
+    {
+      "epoch": 0.09539181664111049,
+      "grad_norm": 0.7769411206245422,
+      "learning_rate": 0.00019854466862402324,
+      "loss": 2.8087,
+      "step": 1182
+    },
+    {
+      "epoch": 0.09547252037769349,
+      "grad_norm": 0.8892520666122437,
+      "learning_rate": 0.00019854198386355085,
+      "loss": 2.7935,
+      "step": 1183
+    },
+    {
+      "epoch": 0.0955532241142765,
+      "grad_norm": 0.8675585389137268,
+      "learning_rate": 0.00019853929664715464,
+      "loss": 2.833,
+      "step": 1184
+    },
+    {
+      "epoch": 0.0956339278508595,
+      "grad_norm": 0.8053853511810303,
+      "learning_rate": 0.00019853660697490154,
+      "loss": 2.8002,
+      "step": 1185
+    },
+    {
+      "epoch": 0.09571463158744249,
+      "grad_norm": 0.9237198829650879,
+      "learning_rate": 0.00019853391484685865,
+      "loss": 2.8281,
+      "step": 1186
+    },
+    {
+      "epoch": 0.0957953353240255,
+      "grad_norm": 0.8432926535606384,
+      "learning_rate": 0.000198531220263093,
+      "loss": 2.8131,
+      "step": 1187
+    },
+    {
+      "epoch": 0.0958760390606085,
+      "grad_norm": 0.796380341053009,
+      "learning_rate": 0.0001985285232236718,
+      "loss": 2.753,
+      "step": 1188
+    },
+    {
+      "epoch": 0.0959567427971915,
+      "grad_norm": 0.9183037281036377,
+      "learning_rate": 0.00019852582372866225,
+      "loss": 2.7625,
+      "step": 1189
+    },
+    {
+      "epoch": 0.09603744653377451,
+      "grad_norm": 0.8194435238838196,
+      "learning_rate": 0.0001985231217781316,
+      "loss": 2.7906,
+      "step": 1190
+    },
+    {
+      "epoch": 0.09611815027035751,
+      "grad_norm": 0.8430871367454529,
+      "learning_rate": 0.00019852041737214725,
+      "loss": 2.8457,
+      "step": 1191
+    },
+    {
+      "epoch": 0.09619885400694052,
+      "grad_norm": 1.0237345695495605,
+      "learning_rate": 0.0001985177105107765,
+      "loss": 2.789,
+      "step": 1192
+    },
+    {
+      "epoch": 0.09627955774352352,
+      "grad_norm": 0.8721581101417542,
+      "learning_rate": 0.00019851500119408692,
+      "loss": 2.7187,
+      "step": 1193
+    },
+    {
+      "epoch": 0.09636026148010653,
+      "grad_norm": 0.8089142441749573,
+      "learning_rate": 0.00019851228942214603,
+      "loss": 2.7544,
+      "step": 1194
+    },
+    {
+      "epoch": 0.09644096521668953,
+      "grad_norm": 1.1076842546463013,
+      "learning_rate": 0.0001985095751950213,
+      "loss": 2.7859,
+      "step": 1195
+    },
+    {
+      "epoch": 0.09652166895327254,
+      "grad_norm": 0.84585040807724,
+      "learning_rate": 0.0001985068585127805,
+      "loss": 2.8005,
+      "step": 1196
+    },
+    {
+      "epoch": 0.09660237268985554,
+      "grad_norm": 0.8231167197227478,
+      "learning_rate": 0.00019850413937549127,
+      "loss": 2.8561,
+      "step": 1197
+    },
+    {
+      "epoch": 0.09668307642643854,
+      "grad_norm": 1.0028103590011597,
+      "learning_rate": 0.00019850141778322136,
+      "loss": 2.8049,
+      "step": 1198
+    },
+    {
+      "epoch": 0.09676378016302155,
+      "grad_norm": 0.8575148582458496,
+      "learning_rate": 0.0001984986937360387,
+      "loss": 2.7723,
+      "step": 1199
+    },
+    {
+      "epoch": 0.09684448389960455,
+      "grad_norm": 0.8567116260528564,
+      "learning_rate": 0.00019849596723401107,
+      "loss": 2.7418,
+      "step": 1200
+    },
+    {
+      "epoch": 0.09692518763618756,
+      "grad_norm": 1.1159218549728394,
+      "learning_rate": 0.00019849323827720645,
+      "loss": 2.8352,
+      "step": 1201
+    },
+    {
+      "epoch": 0.09700589137277056,
+      "grad_norm": 0.849656879901886,
+      "learning_rate": 0.0001984905068656929,
+      "loss": 2.7875,
+      "step": 1202
+    },
+    {
+      "epoch": 0.09708659510935357,
+      "grad_norm": 0.8479150533676147,
+      "learning_rate": 0.00019848777299953847,
+      "loss": 2.7828,
+      "step": 1203
+    },
+    {
+      "epoch": 0.09716729884593657,
+      "grad_norm": 0.9143954515457153,
+      "learning_rate": 0.00019848503667881125,
+      "loss": 2.7978,
+      "step": 1204
+    },
+    {
+      "epoch": 0.09724800258251957,
+      "grad_norm": 0.8162297010421753,
+      "learning_rate": 0.0001984822979035795,
+      "loss": 2.7621,
+      "step": 1205
+    },
+    {
+      "epoch": 0.09732870631910258,
+      "grad_norm": 0.8625509142875671,
+      "learning_rate": 0.00019847955667391144,
+      "loss": 2.7484,
+      "step": 1206
+    },
+    {
+      "epoch": 0.09740941005568558,
+      "grad_norm": 0.8485168218612671,
+      "learning_rate": 0.00019847681298987543,
+      "loss": 2.7599,
+      "step": 1207
+    },
+    {
+      "epoch": 0.09749011379226859,
+      "grad_norm": 0.8962678909301758,
+      "learning_rate": 0.00019847406685153976,
+      "loss": 2.7753,
+      "step": 1208
+    },
+    {
+      "epoch": 0.09757081752885159,
+      "grad_norm": 0.8890791535377502,
+      "learning_rate": 0.00019847131825897297,
+      "loss": 2.7635,
+      "step": 1209
+    },
+    {
+      "epoch": 0.0976515212654346,
+      "grad_norm": 0.8461710810661316,
+      "learning_rate": 0.00019846856721224355,
+      "loss": 2.796,
+      "step": 1210
+    },
+    {
+      "epoch": 0.0977322250020176,
+      "grad_norm": 0.912738025188446,
+      "learning_rate": 0.00019846581371141996,
+      "loss": 2.7889,
+      "step": 1211
+    },
+    {
+      "epoch": 0.09781292873860059,
+      "grad_norm": 0.8530749082565308,
+      "learning_rate": 0.00019846305775657097,
+      "loss": 2.8298,
+      "step": 1212
+    },
+    {
+      "epoch": 0.0978936324751836,
+      "grad_norm": 0.8890148401260376,
+      "learning_rate": 0.00019846029934776516,
+      "loss": 2.7491,
+      "step": 1213
+    },
+    {
+      "epoch": 0.0979743362117666,
+      "grad_norm": 0.8936887979507446,
+      "learning_rate": 0.0001984575384850713,
+      "loss": 2.7759,
+      "step": 1214
+    },
+    {
+      "epoch": 0.0980550399483496,
+      "grad_norm": 0.7811321020126343,
+      "learning_rate": 0.00019845477516855823,
+      "loss": 2.8126,
+      "step": 1215
+    },
+    {
+      "epoch": 0.09813574368493261,
+      "grad_norm": 0.8751768469810486,
+      "learning_rate": 0.00019845200939829484,
+      "loss": 2.792,
+      "step": 1216
+    },
+    {
+      "epoch": 0.09821644742151561,
+      "grad_norm": 0.8749501705169678,
+      "learning_rate": 0.00019844924117434998,
+      "loss": 2.7818,
+      "step": 1217
+    },
+    {
+      "epoch": 0.09829715115809862,
+      "grad_norm": 0.8130955100059509,
+      "learning_rate": 0.0001984464704967927,
+      "loss": 2.8581,
+      "step": 1218
+    },
+    {
+      "epoch": 0.09837785489468162,
+      "grad_norm": 0.8158220648765564,
+      "learning_rate": 0.00019844369736569196,
+      "loss": 2.7704,
+      "step": 1219
+    },
+    {
+      "epoch": 0.09845855863126463,
+      "grad_norm": 0.9351849555969238,
+      "learning_rate": 0.00019844092178111702,
+      "loss": 2.7857,
+      "step": 1220
+    },
+    {
+      "epoch": 0.09853926236784763,
+      "grad_norm": 0.8373914957046509,
+      "learning_rate": 0.00019843814374313697,
+      "loss": 2.8217,
+      "step": 1221
+    },
+    {
+      "epoch": 0.09861996610443063,
+      "grad_norm": 0.8919960856437683,
+      "learning_rate": 0.00019843536325182104,
+      "loss": 2.7914,
+      "step": 1222
+    },
+    {
+      "epoch": 0.09870066984101364,
+      "grad_norm": 0.9994316697120667,
+      "learning_rate": 0.00019843258030723858,
+      "loss": 2.7981,
+      "step": 1223
+    },
+    {
+      "epoch": 0.09878137357759664,
+      "grad_norm": 0.8144915699958801,
+      "learning_rate": 0.0001984297949094589,
+      "loss": 2.811,
+      "step": 1224
+    },
+    {
+      "epoch": 0.09886207731417965,
+      "grad_norm": 0.8957876563072205,
+      "learning_rate": 0.0001984270070585514,
+      "loss": 2.7752,
+      "step": 1225
+    },
+    {
+      "epoch": 0.09894278105076265,
+      "grad_norm": 0.9426520466804504,
+      "learning_rate": 0.0001984242167545856,
+      "loss": 2.8139,
+      "step": 1226
+    },
+    {
+      "epoch": 0.09902348478734566,
+      "grad_norm": 0.888769268989563,
+      "learning_rate": 0.00019842142399763106,
+      "loss": 2.8305,
+      "step": 1227
+    },
+    {
+      "epoch": 0.09910418852392866,
+      "grad_norm": 0.9497748613357544,
+      "learning_rate": 0.00019841862878775736,
+      "loss": 2.748,
+      "step": 1228
+    },
+    {
+      "epoch": 0.09918489226051166,
+      "grad_norm": 0.8715065717697144,
+      "learning_rate": 0.00019841583112503416,
+      "loss": 2.7794,
+      "step": 1229
+    },
+    {
+      "epoch": 0.09926559599709467,
+      "grad_norm": 0.875599205493927,
+      "learning_rate": 0.00019841303100953116,
+      "loss": 2.8016,
+      "step": 1230
+    },
+    {
+      "epoch": 0.09934629973367767,
+      "grad_norm": 0.8631919622421265,
+      "learning_rate": 0.0001984102284413182,
+      "loss": 2.8239,
+      "step": 1231
+    },
+    {
+      "epoch": 0.09942700347026068,
+      "grad_norm": 0.9028074741363525,
+      "learning_rate": 0.0001984074234204651,
+      "loss": 2.8372,
+      "step": 1232
+    },
+    {
+      "epoch": 0.09950770720684368,
+      "grad_norm": 0.890933096408844,
+      "learning_rate": 0.00019840461594704175,
+      "loss": 2.799,
+      "step": 1233
+    },
+    {
+      "epoch": 0.09958841094342669,
+      "grad_norm": 0.9626480340957642,
+      "learning_rate": 0.00019840180602111816,
+      "loss": 2.8207,
+      "step": 1234
+    },
+    {
+      "epoch": 0.09966911468000969,
+      "grad_norm": 0.798394501209259,
+      "learning_rate": 0.00019839899364276433,
+      "loss": 2.7784,
+      "step": 1235
+    },
+    {
+      "epoch": 0.0997498184165927,
+      "grad_norm": 0.8246447443962097,
+      "learning_rate": 0.00019839617881205036,
+      "loss": 2.8193,
+      "step": 1236
+    },
+    {
+      "epoch": 0.09983052215317569,
+      "grad_norm": 0.8315989375114441,
+      "learning_rate": 0.0001983933615290464,
+      "loss": 2.8036,
+      "step": 1237
+    },
+    {
+      "epoch": 0.09991122588975869,
+      "grad_norm": 0.8889075517654419,
+      "learning_rate": 0.00019839054179382267,
+      "loss": 2.7606,
+      "step": 1238
+    },
+    {
+      "epoch": 0.0999919296263417,
+      "grad_norm": 0.7558645009994507,
+      "learning_rate": 0.00019838771960644942,
+      "loss": 2.7666,
+      "step": 1239
+    },
+    {
+      "epoch": 0.1000726333629247,
+      "grad_norm": 0.8876601457595825,
+      "learning_rate": 0.00019838489496699704,
+      "loss": 2.8778,
+      "step": 1240
+    },
+    {
+      "epoch": 0.1001533370995077,
+      "grad_norm": 0.8609516620635986,
+      "learning_rate": 0.00019838206787553588,
+      "loss": 2.8189,
+      "step": 1241
+    },
+    {
+      "epoch": 0.10023404083609071,
+      "grad_norm": 0.8521148562431335,
+      "learning_rate": 0.00019837923833213644,
+      "loss": 2.8159,
+      "step": 1242
+    },
+    {
+      "epoch": 0.10031474457267371,
+      "grad_norm": 0.9155359268188477,
+      "learning_rate": 0.0001983764063368692,
+      "loss": 2.8351,
+      "step": 1243
+    },
+    {
+      "epoch": 0.10039544830925672,
+      "grad_norm": 0.8595378398895264,
+      "learning_rate": 0.00019837357188980475,
+      "loss": 2.8447,
+      "step": 1244
+    },
+    {
+      "epoch": 0.10047615204583972,
+      "grad_norm": 0.900244951248169,
+      "learning_rate": 0.00019837073499101373,
+      "loss": 2.8646,
+      "step": 1245
+    },
+    {
+      "epoch": 0.10055685578242272,
+      "grad_norm": 0.8404260277748108,
+      "learning_rate": 0.00019836789564056689,
+      "loss": 2.7824,
+      "step": 1246
+    },
+    {
+      "epoch": 0.10063755951900573,
+      "grad_norm": 0.8776196241378784,
+      "learning_rate": 0.0001983650538385349,
+      "loss": 2.8045,
+      "step": 1247
+    },
+    {
+      "epoch": 0.10071826325558873,
+      "grad_norm": 0.8889327049255371,
+      "learning_rate": 0.00019836220958498868,
+      "loss": 2.7967,
+      "step": 1248
+    },
+    {
+      "epoch": 0.10079896699217174,
+      "grad_norm": 0.8905191421508789,
+      "learning_rate": 0.00019835936287999906,
+      "loss": 2.8167,
+      "step": 1249
+    },
+    {
+      "epoch": 0.10087967072875474,
+      "grad_norm": 0.839970052242279,
+      "learning_rate": 0.000198356513723637,
+      "loss": 2.8643,
+      "step": 1250
+    },
+    {
+      "epoch": 0.10096037446533775,
+      "grad_norm": 0.7989531755447388,
+      "learning_rate": 0.00019835366211597353,
+      "loss": 2.8493,
+      "step": 1251
+    },
+    {
+      "epoch": 0.10104107820192075,
+      "grad_norm": 0.7960095405578613,
+      "learning_rate": 0.0001983508080570797,
+      "loss": 2.7377,
+      "step": 1252
+    },
+    {
+      "epoch": 0.10112178193850375,
+      "grad_norm": 0.7989903092384338,
+      "learning_rate": 0.00019834795154702661,
+      "loss": 2.7409,
+      "step": 1253
+    },
+    {
+      "epoch": 0.10120248567508676,
+      "grad_norm": 0.8557813167572021,
+      "learning_rate": 0.0001983450925858855,
+      "loss": 2.7945,
+      "step": 1254
+    },
+    {
+      "epoch": 0.10128318941166976,
+      "grad_norm": 0.948357880115509,
+      "learning_rate": 0.0001983422311737276,
+      "loss": 2.826,
+      "step": 1255
+    },
+    {
+      "epoch": 0.10136389314825277,
+      "grad_norm": 0.8356020450592041,
+      "learning_rate": 0.00019833936731062423,
+      "loss": 2.8157,
+      "step": 1256
+    },
+    {
+      "epoch": 0.10144459688483577,
+      "grad_norm": 0.8199872970581055,
+      "learning_rate": 0.00019833650099664678,
+      "loss": 2.7273,
+      "step": 1257
+    },
+    {
+      "epoch": 0.10152530062141878,
+      "grad_norm": 0.8178466558456421,
+      "learning_rate": 0.00019833363223186669,
+      "loss": 2.7513,
+      "step": 1258
+    },
+    {
+      "epoch": 0.10160600435800178,
+      "grad_norm": 0.8165889978408813,
+      "learning_rate": 0.00019833076101635538,
+      "loss": 2.7689,
+      "step": 1259
+    },
+    {
+      "epoch": 0.10168670809458479,
+      "grad_norm": 0.8240275979042053,
+      "learning_rate": 0.0001983278873501845,
+      "loss": 2.7477,
+      "step": 1260
+    },
+    {
+      "epoch": 0.10176741183116779,
+      "grad_norm": 0.8470584750175476,
+      "learning_rate": 0.00019832501123342563,
+      "loss": 2.7414,
+      "step": 1261
+    },
+    {
+      "epoch": 0.1018481155677508,
+      "grad_norm": 0.819063663482666,
+      "learning_rate": 0.00019832213266615046,
+      "loss": 2.7335,
+      "step": 1262
+    },
+    {
+      "epoch": 0.10192881930433378,
+      "grad_norm": 0.8045673370361328,
+      "learning_rate": 0.00019831925164843071,
+      "loss": 2.8141,
+      "step": 1263
+    },
+    {
+      "epoch": 0.10200952304091679,
+      "grad_norm": 0.7827214598655701,
+      "learning_rate": 0.00019831636818033824,
+      "loss": 2.7549,
+      "step": 1264
+    },
+    {
+      "epoch": 0.10209022677749979,
+      "grad_norm": 0.9596436619758606,
+      "learning_rate": 0.00019831348226194485,
+      "loss": 2.7327,
+      "step": 1265
+    },
+    {
+      "epoch": 0.1021709305140828,
+      "grad_norm": 0.826909601688385,
+      "learning_rate": 0.0001983105938933225,
+      "loss": 2.7166,
+      "step": 1266
+    },
+    {
+      "epoch": 0.1022516342506658,
+      "grad_norm": 0.8060985207557678,
+      "learning_rate": 0.00019830770307454313,
+      "loss": 2.7514,
+      "step": 1267
+    },
+    {
+      "epoch": 0.1023323379872488,
+      "grad_norm": 0.8257390856742859,
+      "learning_rate": 0.00019830480980567887,
+      "loss": 2.77,
+      "step": 1268
+    },
+    {
+      "epoch": 0.10241304172383181,
+      "grad_norm": 0.844406008720398,
+      "learning_rate": 0.00019830191408680173,
+      "loss": 2.8548,
+      "step": 1269
+    },
+    {
+      "epoch": 0.10249374546041481,
+      "grad_norm": 0.84171462059021,
+      "learning_rate": 0.00019829901591798398,
+      "loss": 2.7404,
+      "step": 1270
+    },
+    {
+      "epoch": 0.10257444919699782,
+      "grad_norm": 0.8084118962287903,
+      "learning_rate": 0.00019829611529929774,
+      "loss": 2.8078,
+      "step": 1271
+    },
+    {
+      "epoch": 0.10265515293358082,
+      "grad_norm": 0.8273561000823975,
+      "learning_rate": 0.00019829321223081538,
+      "loss": 2.787,
+      "step": 1272
+    },
+    {
+      "epoch": 0.10273585667016383,
+      "grad_norm": 0.799098551273346,
+      "learning_rate": 0.00019829030671260925,
+      "loss": 2.7563,
+      "step": 1273
+    },
+    {
+      "epoch": 0.10281656040674683,
+      "grad_norm": 0.885866105556488,
+      "learning_rate": 0.00019828739874475172,
+      "loss": 2.7313,
+      "step": 1274
+    },
+    {
+      "epoch": 0.10289726414332984,
+      "grad_norm": 0.7702760696411133,
+      "learning_rate": 0.00019828448832731529,
+      "loss": 2.7919,
+      "step": 1275
+    },
+    {
+      "epoch": 0.10297796787991284,
+      "grad_norm": 0.7577444911003113,
+      "learning_rate": 0.0001982815754603725,
+      "loss": 2.7149,
+      "step": 1276
+    },
+    {
+      "epoch": 0.10305867161649584,
+      "grad_norm": 0.8439713716506958,
+      "learning_rate": 0.00019827866014399592,
+      "loss": 2.7881,
+      "step": 1277
+    },
+    {
+      "epoch": 0.10313937535307885,
+      "grad_norm": 0.8504937291145325,
+      "learning_rate": 0.00019827574237825827,
+      "loss": 2.7611,
+      "step": 1278
+    },
+    {
+      "epoch": 0.10322007908966185,
+      "grad_norm": 0.7775665521621704,
+      "learning_rate": 0.00019827282216323218,
+      "loss": 2.7312,
+      "step": 1279
+    },
+    {
+      "epoch": 0.10330078282624486,
+      "grad_norm": 0.8671591281890869,
+      "learning_rate": 0.00019826989949899048,
+      "loss": 2.836,
+      "step": 1280
+    },
+    {
+      "epoch": 0.10338148656282786,
+      "grad_norm": 0.9308713674545288,
+      "learning_rate": 0.00019826697438560603,
+      "loss": 2.7494,
+      "step": 1281
+    },
+    {
+      "epoch": 0.10346219029941087,
+      "grad_norm": 0.9145268797874451,
+      "learning_rate": 0.0001982640468231517,
+      "loss": 2.8054,
+      "step": 1282
+    },
+    {
+      "epoch": 0.10354289403599387,
+      "grad_norm": 0.8150805234909058,
+      "learning_rate": 0.00019826111681170043,
+      "loss": 2.7879,
+      "step": 1283
+    },
+    {
+      "epoch": 0.10362359777257688,
+      "grad_norm": 0.8576685786247253,
+      "learning_rate": 0.00019825818435132531,
+      "loss": 2.8184,
+      "step": 1284
+    },
+    {
+      "epoch": 0.10370430150915988,
+      "grad_norm": 0.8838599920272827,
+      "learning_rate": 0.00019825524944209937,
+      "loss": 2.7838,
+      "step": 1285
+    },
+    {
+      "epoch": 0.10378500524574288,
+      "grad_norm": 0.9119304418563843,
+      "learning_rate": 0.00019825231208409576,
+      "loss": 2.8392,
+      "step": 1286
+    },
+    {
+      "epoch": 0.10386570898232589,
+      "grad_norm": 0.8112398982048035,
+      "learning_rate": 0.00019824937227738771,
+      "loss": 2.7844,
+      "step": 1287
+    },
+    {
+      "epoch": 0.10394641271890888,
+      "grad_norm": 0.8714308738708496,
+      "learning_rate": 0.00019824643002204847,
+      "loss": 2.7765,
+      "step": 1288
+    },
+    {
+      "epoch": 0.10402711645549188,
+      "grad_norm": 0.8733358979225159,
+      "learning_rate": 0.00019824348531815138,
+      "loss": 2.771,
+      "step": 1289
+    },
+    {
+      "epoch": 0.10410782019207489,
+      "grad_norm": 0.8218281269073486,
+      "learning_rate": 0.00019824053816576981,
+      "loss": 2.8099,
+      "step": 1290
+    },
+    {
+      "epoch": 0.10418852392865789,
+      "grad_norm": 0.8647308945655823,
+      "learning_rate": 0.00019823758856497725,
+      "loss": 2.7738,
+      "step": 1291
+    },
+    {
+      "epoch": 0.1042692276652409,
+      "grad_norm": 0.8358582854270935,
+      "learning_rate": 0.00019823463651584718,
+      "loss": 2.8021,
+      "step": 1292
+    },
+    {
+      "epoch": 0.1043499314018239,
+      "grad_norm": 0.7943673133850098,
+      "learning_rate": 0.00019823168201845318,
+      "loss": 2.8293,
+      "step": 1293
+    },
+    {
+      "epoch": 0.1044306351384069,
+      "grad_norm": 0.8501425981521606,
+      "learning_rate": 0.0001982287250728689,
+      "loss": 2.7701,
+      "step": 1294
+    },
+    {
+      "epoch": 0.10451133887498991,
+      "grad_norm": 0.8503665328025818,
+      "learning_rate": 0.00019822576567916797,
+      "loss": 2.7881,
+      "step": 1295
+    },
+    {
+      "epoch": 0.10459204261157291,
+      "grad_norm": 0.9687628149986267,
+      "learning_rate": 0.0001982228038374242,
+      "loss": 2.7623,
+      "step": 1296
+    },
+    {
+      "epoch": 0.10467274634815592,
+      "grad_norm": 0.8034376502037048,
+      "learning_rate": 0.00019821983954771146,
+      "loss": 2.8072,
+      "step": 1297
+    },
+    {
+      "epoch": 0.10475345008473892,
+      "grad_norm": 0.817135214805603,
+      "learning_rate": 0.00019821687281010352,
+      "loss": 2.7572,
+      "step": 1298
+    },
+    {
+      "epoch": 0.10483415382132193,
+      "grad_norm": 0.7961457371711731,
+      "learning_rate": 0.0001982139036246744,
+      "loss": 2.8405,
+      "step": 1299
+    },
+    {
+      "epoch": 0.10491485755790493,
+      "grad_norm": 0.7572407722473145,
+      "learning_rate": 0.00019821093199149804,
+      "loss": 2.7495,
+      "step": 1300
+    },
+    {
+      "epoch": 0.10499556129448794,
+      "grad_norm": 0.7990664839744568,
+      "learning_rate": 0.00019820795791064856,
+      "loss": 2.7567,
+      "step": 1301
+    },
+    {
+      "epoch": 0.10507626503107094,
+      "grad_norm": 0.8197236061096191,
+      "learning_rate": 0.0001982049813822,
+      "loss": 2.7807,
+      "step": 1302
+    },
+    {
+      "epoch": 0.10515696876765394,
+      "grad_norm": 0.9491304159164429,
+      "learning_rate": 0.00019820200240622664,
+      "loss": 2.8531,
+      "step": 1303
+    },
+    {
+      "epoch": 0.10523767250423695,
+      "grad_norm": 0.8143845200538635,
+      "learning_rate": 0.00019819902098280268,
+      "loss": 2.7542,
+      "step": 1304
+    },
+    {
+      "epoch": 0.10531837624081995,
+      "grad_norm": 0.9055941104888916,
+      "learning_rate": 0.0001981960371120024,
+      "loss": 2.863,
+      "step": 1305
+    },
+    {
+      "epoch": 0.10539907997740296,
+      "grad_norm": 0.7804721593856812,
+      "learning_rate": 0.0001981930507939002,
+      "loss": 2.8213,
+      "step": 1306
+    },
+    {
+      "epoch": 0.10547978371398596,
+      "grad_norm": 0.8375318050384521,
+      "learning_rate": 0.00019819006202857046,
+      "loss": 2.8222,
+      "step": 1307
+    },
+    {
+      "epoch": 0.10556048745056897,
+      "grad_norm": 0.9145569801330566,
+      "learning_rate": 0.00019818707081608773,
+      "loss": 2.805,
+      "step": 1308
+    },
+    {
+      "epoch": 0.10564119118715197,
+      "grad_norm": 0.7899324893951416,
+      "learning_rate": 0.00019818407715652654,
+      "loss": 2.8246,
+      "step": 1309
+    },
+    {
+      "epoch": 0.10572189492373497,
+      "grad_norm": 0.7843480110168457,
+      "learning_rate": 0.0001981810810499615,
+      "loss": 2.7909,
+      "step": 1310
+    },
+    {
+      "epoch": 0.10580259866031798,
+      "grad_norm": 0.8071008920669556,
+      "learning_rate": 0.00019817808249646723,
+      "loss": 2.7434,
+      "step": 1311
+    },
+    {
+      "epoch": 0.10588330239690098,
+      "grad_norm": 0.8682011961936951,
+      "learning_rate": 0.0001981750814961185,
+      "loss": 2.8387,
+      "step": 1312
+    },
+    {
+      "epoch": 0.10596400613348399,
+      "grad_norm": 0.7501091361045837,
+      "learning_rate": 0.0001981720780489902,
+      "loss": 2.7633,
+      "step": 1313
+    },
+    {
+      "epoch": 0.10604470987006698,
+      "grad_norm": 0.9259567856788635,
+      "learning_rate": 0.000198169072155157,
+      "loss": 2.8309,
+      "step": 1314
+    },
+    {
+      "epoch": 0.10612541360664998,
+      "grad_norm": 0.8018674254417419,
+      "learning_rate": 0.00019816606381469393,
+      "loss": 2.8647,
+      "step": 1315
+    },
+    {
+      "epoch": 0.10620611734323299,
+      "grad_norm": 0.8218088746070862,
+      "learning_rate": 0.00019816305302767595,
+      "loss": 2.823,
+      "step": 1316
+    },
+    {
+      "epoch": 0.10628682107981599,
+      "grad_norm": 0.812125027179718,
+      "learning_rate": 0.00019816003979417808,
+      "loss": 2.7216,
+      "step": 1317
+    },
+    {
+      "epoch": 0.106367524816399,
+      "grad_norm": 0.787407636642456,
+      "learning_rate": 0.0001981570241142754,
+      "loss": 2.7639,
+      "step": 1318
+    },
+    {
+      "epoch": 0.106448228552982,
+      "grad_norm": 0.7982528805732727,
+      "learning_rate": 0.00019815400598804312,
+      "loss": 2.8597,
+      "step": 1319
+    },
+    {
+      "epoch": 0.106528932289565,
+      "grad_norm": 0.8490404486656189,
+      "learning_rate": 0.00019815098541555646,
+      "loss": 2.7947,
+      "step": 1320
+    },
+    {
+      "epoch": 0.10660963602614801,
+      "grad_norm": 0.8743172883987427,
+      "learning_rate": 0.00019814796239689064,
+      "loss": 2.8674,
+      "step": 1321
+    },
+    {
+      "epoch": 0.10669033976273101,
+      "grad_norm": 0.8338125348091125,
+      "learning_rate": 0.00019814493693212106,
+      "loss": 2.781,
+      "step": 1322
+    },
+    {
+      "epoch": 0.10677104349931402,
+      "grad_norm": 0.871516764163971,
+      "learning_rate": 0.00019814190902132307,
+      "loss": 2.8742,
+      "step": 1323
+    },
+    {
+      "epoch": 0.10685174723589702,
+      "grad_norm": 0.8935555815696716,
+      "learning_rate": 0.00019813887866457216,
+      "loss": 2.7991,
+      "step": 1324
+    },
+    {
+      "epoch": 0.10693245097248003,
+      "grad_norm": 0.840067446231842,
+      "learning_rate": 0.00019813584586194388,
+      "loss": 2.7922,
+      "step": 1325
+    },
+    {
+      "epoch": 0.10701315470906303,
+      "grad_norm": 0.7919262647628784,
+      "learning_rate": 0.0001981328106135138,
+      "loss": 2.7912,
+      "step": 1326
+    },
+    {
+      "epoch": 0.10709385844564603,
+      "grad_norm": 0.7974550127983093,
+      "learning_rate": 0.00019812977291935752,
+      "loss": 2.8497,
+      "step": 1327
+    },
+    {
+      "epoch": 0.10717456218222904,
+      "grad_norm": 0.9126157164573669,
+      "learning_rate": 0.00019812673277955082,
+      "loss": 2.7698,
+      "step": 1328
+    },
+    {
+      "epoch": 0.10725526591881204,
+      "grad_norm": 0.8329752683639526,
+      "learning_rate": 0.0001981236901941694,
+      "loss": 2.8366,
+      "step": 1329
+    },
+    {
+      "epoch": 0.10733596965539505,
+      "grad_norm": 0.8313524127006531,
+      "learning_rate": 0.00019812064516328915,
+      "loss": 2.6863,
+      "step": 1330
+    },
+    {
+      "epoch": 0.10741667339197805,
+      "grad_norm": 0.8917783498764038,
+      "learning_rate": 0.0001981175976869859,
+      "loss": 2.7817,
+      "step": 1331
+    },
+    {
+      "epoch": 0.10749737712856106,
+      "grad_norm": 0.8370450735092163,
+      "learning_rate": 0.00019811454776533566,
+      "loss": 2.837,
+      "step": 1332
+    },
+    {
+      "epoch": 0.10757808086514406,
+      "grad_norm": 0.8415676355361938,
+      "learning_rate": 0.00019811149539841443,
+      "loss": 2.7399,
+      "step": 1333
+    },
+    {
+      "epoch": 0.10765878460172706,
+      "grad_norm": 0.8576632142066956,
+      "learning_rate": 0.00019810844058629825,
+      "loss": 2.7747,
+      "step": 1334
+    },
+    {
+      "epoch": 0.10773948833831007,
+      "grad_norm": 0.8943549394607544,
+      "learning_rate": 0.00019810538332906328,
+      "loss": 2.7368,
+      "step": 1335
+    },
+    {
+      "epoch": 0.10782019207489307,
+      "grad_norm": 0.8878718018531799,
+      "learning_rate": 0.00019810232362678568,
+      "loss": 2.7907,
+      "step": 1336
+    },
+    {
+      "epoch": 0.10790089581147608,
+      "grad_norm": 0.8131409287452698,
+      "learning_rate": 0.00019809926147954174,
+      "loss": 2.7782,
+      "step": 1337
+    },
+    {
+      "epoch": 0.10798159954805908,
+      "grad_norm": 0.8733747005462646,
+      "learning_rate": 0.0001980961968874078,
+      "loss": 2.8552,
+      "step": 1338
+    },
+    {
+      "epoch": 0.10806230328464207,
+      "grad_norm": 0.8997320532798767,
+      "learning_rate": 0.0001980931298504602,
+      "loss": 2.8452,
+      "step": 1339
+    },
+    {
+      "epoch": 0.10814300702122508,
+      "grad_norm": 0.8400282263755798,
+      "learning_rate": 0.00019809006036877538,
+      "loss": 2.786,
+      "step": 1340
+    },
+    {
+      "epoch": 0.10822371075780808,
+      "grad_norm": 0.8173925280570984,
+      "learning_rate": 0.00019808698844242983,
+      "loss": 2.8363,
+      "step": 1341
+    },
+    {
+      "epoch": 0.10830441449439109,
+      "grad_norm": 0.872278094291687,
+      "learning_rate": 0.00019808391407150015,
+      "loss": 2.7789,
+      "step": 1342
+    },
+    {
+      "epoch": 0.10838511823097409,
+      "grad_norm": 0.8939952254295349,
+      "learning_rate": 0.00019808083725606293,
+      "loss": 2.7453,
+      "step": 1343
+    },
+    {
+      "epoch": 0.1084658219675571,
+      "grad_norm": 0.8351218104362488,
+      "learning_rate": 0.00019807775799619484,
+      "loss": 2.8004,
+      "step": 1344
+    },
+    {
+      "epoch": 0.1085465257041401,
+      "grad_norm": 0.8381102681159973,
+      "learning_rate": 0.00019807467629197266,
+      "loss": 2.8155,
+      "step": 1345
+    },
+    {
+      "epoch": 0.1086272294407231,
+      "grad_norm": 0.869458019733429,
+      "learning_rate": 0.00019807159214347317,
+      "loss": 2.8219,
+      "step": 1346
+    },
+    {
+      "epoch": 0.10870793317730611,
+      "grad_norm": 0.8251017928123474,
+      "learning_rate": 0.00019806850555077326,
+      "loss": 2.7978,
+      "step": 1347
+    },
+    {
+      "epoch": 0.10878863691388911,
+      "grad_norm": 0.8056492209434509,
+      "learning_rate": 0.0001980654165139498,
+      "loss": 2.7994,
+      "step": 1348
+    },
+    {
+      "epoch": 0.10886934065047212,
+      "grad_norm": 0.9566174745559692,
+      "learning_rate": 0.00019806232503307984,
+      "loss": 2.794,
+      "step": 1349
+    },
+    {
+      "epoch": 0.10895004438705512,
+      "grad_norm": 0.7891408801078796,
+      "learning_rate": 0.0001980592311082404,
+      "loss": 2.7134,
+      "step": 1350
+    },
+    {
+      "epoch": 0.10903074812363812,
+      "grad_norm": 0.8894741535186768,
+      "learning_rate": 0.00019805613473950862,
+      "loss": 2.7829,
+      "step": 1351
+    },
+    {
+      "epoch": 0.10911145186022113,
+      "grad_norm": 0.893086850643158,
+      "learning_rate": 0.0001980530359269616,
+      "loss": 2.7475,
+      "step": 1352
+    },
+    {
+      "epoch": 0.10919215559680413,
+      "grad_norm": 0.8758537173271179,
+      "learning_rate": 0.00019804993467067666,
+      "loss": 2.8715,
+      "step": 1353
+    },
+    {
+      "epoch": 0.10927285933338714,
+      "grad_norm": 0.9304648041725159,
+      "learning_rate": 0.00019804683097073098,
+      "loss": 2.8051,
+      "step": 1354
+    },
+    {
+      "epoch": 0.10935356306997014,
+      "grad_norm": 0.8465876579284668,
+      "learning_rate": 0.00019804372482720202,
+      "loss": 2.7879,
+      "step": 1355
+    },
+    {
+      "epoch": 0.10943426680655315,
+      "grad_norm": 0.8485612273216248,
+      "learning_rate": 0.00019804061624016713,
+      "loss": 2.7783,
+      "step": 1356
+    },
+    {
+      "epoch": 0.10951497054313615,
+      "grad_norm": 0.835630476474762,
+      "learning_rate": 0.0001980375052097038,
+      "loss": 2.8116,
+      "step": 1357
+    },
+    {
+      "epoch": 0.10959567427971915,
+      "grad_norm": 0.8404836058616638,
+      "learning_rate": 0.00019803439173588956,
+      "loss": 2.8257,
+      "step": 1358
+    },
+    {
+      "epoch": 0.10967637801630216,
+      "grad_norm": 0.8048505783081055,
+      "learning_rate": 0.00019803127581880206,
+      "loss": 2.7762,
+      "step": 1359
+    },
+    {
+      "epoch": 0.10975708175288516,
+      "grad_norm": 0.8481776118278503,
+      "learning_rate": 0.00019802815745851885,
+      "loss": 2.8243,
+      "step": 1360
+    },
+    {
+      "epoch": 0.10983778548946817,
+      "grad_norm": 0.8565996885299683,
+      "learning_rate": 0.00019802503665511775,
+      "loss": 2.7958,
+      "step": 1361
+    },
+    {
+      "epoch": 0.10991848922605117,
+      "grad_norm": 0.8867515921592712,
+      "learning_rate": 0.0001980219134086765,
+      "loss": 2.7973,
+      "step": 1362
+    },
+    {
+      "epoch": 0.10999919296263418,
+      "grad_norm": 0.8459765911102295,
+      "learning_rate": 0.0001980187877192729,
+      "loss": 2.848,
+      "step": 1363
+    },
+    {
+      "epoch": 0.11007989669921718,
+      "grad_norm": 0.7929832339286804,
+      "learning_rate": 0.0001980156595869849,
+      "loss": 2.8583,
+      "step": 1364
+    },
+    {
+      "epoch": 0.11016060043580017,
+      "grad_norm": 0.8475651741027832,
+      "learning_rate": 0.00019801252901189043,
+      "loss": 2.8436,
+      "step": 1365
+    },
+    {
+      "epoch": 0.11024130417238318,
+      "grad_norm": 0.8545576333999634,
+      "learning_rate": 0.00019800939599406755,
+      "loss": 2.7457,
+      "step": 1366
+    },
+    {
+      "epoch": 0.11032200790896618,
+      "grad_norm": 1.0093715190887451,
+      "learning_rate": 0.00019800626053359435,
+      "loss": 2.8198,
+      "step": 1367
+    },
+    {
+      "epoch": 0.11040271164554918,
+      "grad_norm": 0.8728145956993103,
+      "learning_rate": 0.0001980031226305489,
+      "loss": 2.7794,
+      "step": 1368
+    },
+    {
+      "epoch": 0.11048341538213219,
+      "grad_norm": 0.8538581728935242,
+      "learning_rate": 0.00019799998228500946,
+      "loss": 2.8018,
+      "step": 1369
+    },
+    {
+      "epoch": 0.11056411911871519,
+      "grad_norm": 0.9452785849571228,
+      "learning_rate": 0.00019799683949705432,
+      "loss": 2.8173,
+      "step": 1370
+    },
+    {
+      "epoch": 0.1106448228552982,
+      "grad_norm": 0.806508481502533,
+      "learning_rate": 0.00019799369426676174,
+      "loss": 2.8192,
+      "step": 1371
+    },
+    {
+      "epoch": 0.1107255265918812,
+      "grad_norm": 0.8952856063842773,
+      "learning_rate": 0.00019799054659421018,
+      "loss": 2.8072,
+      "step": 1372
+    },
+    {
+      "epoch": 0.1108062303284642,
+      "grad_norm": 0.8863561749458313,
+      "learning_rate": 0.00019798739647947802,
+      "loss": 2.7836,
+      "step": 1373
+    },
+    {
+      "epoch": 0.11088693406504721,
+      "grad_norm": 0.8544357419013977,
+      "learning_rate": 0.00019798424392264378,
+      "loss": 2.7714,
+      "step": 1374
+    },
+    {
+      "epoch": 0.11096763780163021,
+      "grad_norm": 0.807546854019165,
+      "learning_rate": 0.00019798108892378607,
+      "loss": 2.7635,
+      "step": 1375
+    },
+    {
+      "epoch": 0.11104834153821322,
+      "grad_norm": 0.8198233246803284,
+      "learning_rate": 0.0001979779314829835,
+      "loss": 2.8253,
+      "step": 1376
+    },
+    {
+      "epoch": 0.11112904527479622,
+      "grad_norm": 0.9268671870231628,
+      "learning_rate": 0.00019797477160031477,
+      "loss": 2.8007,
+      "step": 1377
+    },
+    {
+      "epoch": 0.11120974901137923,
+      "grad_norm": 0.8547680974006653,
+      "learning_rate": 0.0001979716092758586,
+      "loss": 2.7749,
+      "step": 1378
+    },
+    {
+      "epoch": 0.11129045274796223,
+      "grad_norm": 0.8052394390106201,
+      "learning_rate": 0.00019796844450969384,
+      "loss": 2.763,
+      "step": 1379
+    },
+    {
+      "epoch": 0.11137115648454524,
+      "grad_norm": 0.8291144371032715,
+      "learning_rate": 0.00019796527730189936,
+      "loss": 2.8053,
+      "step": 1380
+    },
+    {
+      "epoch": 0.11145186022112824,
+      "grad_norm": 0.8114006519317627,
+      "learning_rate": 0.00019796210765255404,
+      "loss": 2.8047,
+      "step": 1381
+    },
+    {
+      "epoch": 0.11153256395771124,
+      "grad_norm": 0.9326293468475342,
+      "learning_rate": 0.00019795893556173697,
+      "loss": 2.8199,
+      "step": 1382
+    },
+    {
+      "epoch": 0.11161326769429425,
+      "grad_norm": 0.7702555656433105,
+      "learning_rate": 0.00019795576102952714,
+      "loss": 2.7909,
+      "step": 1383
+    },
+    {
+      "epoch": 0.11169397143087725,
+      "grad_norm": 0.8115492463111877,
+      "learning_rate": 0.0001979525840560037,
+      "loss": 2.748,
+      "step": 1384
+    },
+    {
+      "epoch": 0.11177467516746026,
+      "grad_norm": 0.8926187753677368,
+      "learning_rate": 0.0001979494046412458,
+      "loss": 2.7791,
+      "step": 1385
+    },
+    {
+      "epoch": 0.11185537890404326,
+      "grad_norm": 0.8549754023551941,
+      "learning_rate": 0.0001979462227853327,
+      "loss": 2.7989,
+      "step": 1386
+    },
+    {
+      "epoch": 0.11193608264062627,
+      "grad_norm": 0.8625262975692749,
+      "learning_rate": 0.0001979430384883437,
+      "loss": 2.7202,
+      "step": 1387
+    },
+    {
+      "epoch": 0.11201678637720927,
+      "grad_norm": 0.8134698867797852,
+      "learning_rate": 0.00019793985175035813,
+      "loss": 2.8008,
+      "step": 1388
+    },
+    {
+      "epoch": 0.11209749011379228,
+      "grad_norm": 0.8546617031097412,
+      "learning_rate": 0.00019793666257145547,
+      "loss": 2.8076,
+      "step": 1389
+    },
+    {
+      "epoch": 0.11217819385037527,
+      "grad_norm": 0.8003748059272766,
+      "learning_rate": 0.00019793347095171514,
+      "loss": 2.826,
+      "step": 1390
+    },
+    {
+      "epoch": 0.11225889758695827,
+      "grad_norm": 0.8116614818572998,
+      "learning_rate": 0.00019793027689121674,
+      "loss": 2.7096,
+      "step": 1391
+    },
+    {
+      "epoch": 0.11233960132354127,
+      "grad_norm": 0.7785829901695251,
+      "learning_rate": 0.00019792708039003984,
+      "loss": 2.748,
+      "step": 1392
+    },
+    {
+      "epoch": 0.11242030506012428,
+      "grad_norm": 0.7999277710914612,
+      "learning_rate": 0.0001979238814482641,
+      "loss": 2.7671,
+      "step": 1393
+    },
+    {
+      "epoch": 0.11250100879670728,
+      "grad_norm": 0.8862190842628479,
+      "learning_rate": 0.00019792068006596925,
+      "loss": 2.8484,
+      "step": 1394
+    },
+    {
+      "epoch": 0.11258171253329029,
+      "grad_norm": 0.8747627139091492,
+      "learning_rate": 0.00019791747624323512,
+      "loss": 2.7477,
+      "step": 1395
+    },
+    {
+      "epoch": 0.11266241626987329,
+      "grad_norm": 0.8280831575393677,
+      "learning_rate": 0.0001979142699801415,
+      "loss": 2.87,
+      "step": 1396
+    },
+    {
+      "epoch": 0.1127431200064563,
+      "grad_norm": 0.8069074153900146,
+      "learning_rate": 0.00019791106127676832,
+      "loss": 2.7724,
+      "step": 1397
+    },
+    {
+      "epoch": 0.1128238237430393,
+      "grad_norm": 0.8253301382064819,
+      "learning_rate": 0.00019790785013319557,
+      "loss": 2.7351,
+      "step": 1398
+    },
+    {
+      "epoch": 0.1129045274796223,
+      "grad_norm": 0.8298853635787964,
+      "learning_rate": 0.00019790463654950323,
+      "loss": 2.7709,
+      "step": 1399
+    },
+    {
+      "epoch": 0.11298523121620531,
+      "grad_norm": 0.7796407341957092,
+      "learning_rate": 0.0001979014205257715,
+      "loss": 2.7766,
+      "step": 1400
+    },
+    {
+      "epoch": 0.11306593495278831,
+      "grad_norm": 0.8922166228294373,
+      "learning_rate": 0.00019789820206208037,
+      "loss": 2.8473,
+      "step": 1401
+    },
+    {
+      "epoch": 0.11314663868937132,
+      "grad_norm": 0.7763219475746155,
+      "learning_rate": 0.00019789498115851015,
+      "loss": 2.8629,
+      "step": 1402
+    },
+    {
+      "epoch": 0.11322734242595432,
+      "grad_norm": 0.8679928779602051,
+      "learning_rate": 0.0001978917578151411,
+      "loss": 2.8017,
+      "step": 1403
+    },
+    {
+      "epoch": 0.11330804616253733,
+      "grad_norm": 0.8491933941841125,
+      "learning_rate": 0.00019788853203205357,
+      "loss": 2.7156,
+      "step": 1404
+    },
+    {
+      "epoch": 0.11338874989912033,
+      "grad_norm": 0.8271194696426392,
+      "learning_rate": 0.00019788530380932792,
+      "loss": 2.7892,
+      "step": 1405
+    },
+    {
+      "epoch": 0.11346945363570334,
+      "grad_norm": 0.9224163293838501,
+      "learning_rate": 0.00019788207314704463,
+      "loss": 2.7824,
+      "step": 1406
+    },
+    {
+      "epoch": 0.11355015737228634,
+      "grad_norm": 0.7662777900695801,
+      "learning_rate": 0.00019787884004528422,
+      "loss": 2.7364,
+      "step": 1407
+    },
+    {
+      "epoch": 0.11363086110886934,
+      "grad_norm": 0.8750362396240234,
+      "learning_rate": 0.00019787560450412728,
+      "loss": 2.7546,
+      "step": 1408
+    },
+    {
+      "epoch": 0.11371156484545235,
+      "grad_norm": 0.9158821105957031,
+      "learning_rate": 0.0001978723665236544,
+      "loss": 2.8304,
+      "step": 1409
+    },
+    {
+      "epoch": 0.11379226858203535,
+      "grad_norm": 0.8291050791740417,
+      "learning_rate": 0.0001978691261039463,
+      "loss": 2.758,
+      "step": 1410
+    },
+    {
+      "epoch": 0.11387297231861836,
+      "grad_norm": 0.801886796951294,
+      "learning_rate": 0.00019786588324508374,
+      "loss": 2.7805,
+      "step": 1411
+    },
+    {
+      "epoch": 0.11395367605520136,
+      "grad_norm": 0.8140222430229187,
+      "learning_rate": 0.00019786263794714757,
+      "loss": 2.8155,
+      "step": 1412
+    },
+    {
+      "epoch": 0.11403437979178437,
+      "grad_norm": 0.7747580409049988,
+      "learning_rate": 0.00019785939021021865,
+      "loss": 2.778,
+      "step": 1413
+    },
+    {
+      "epoch": 0.11411508352836737,
+      "grad_norm": 0.8954138159751892,
+      "learning_rate": 0.0001978561400343779,
+      "loss": 2.7756,
+      "step": 1414
+    },
+    {
+      "epoch": 0.11419578726495037,
+      "grad_norm": 0.9038921594619751,
+      "learning_rate": 0.00019785288741970634,
+      "loss": 2.7181,
+      "step": 1415
+    },
+    {
+      "epoch": 0.11427649100153336,
+      "grad_norm": 0.8284393548965454,
+      "learning_rate": 0.000197849632366285,
+      "loss": 2.7467,
+      "step": 1416
+    },
+    {
+      "epoch": 0.11435719473811637,
+      "grad_norm": 0.8996441960334778,
+      "learning_rate": 0.00019784637487419514,
+      "loss": 2.7918,
+      "step": 1417
+    },
+    {
+      "epoch": 0.11443789847469937,
+      "grad_norm": 0.9868448376655579,
+      "learning_rate": 0.00019784311494351777,
+      "loss": 2.7687,
+      "step": 1418
+    },
+    {
+      "epoch": 0.11451860221128238,
+      "grad_norm": 0.8491402864456177,
+      "learning_rate": 0.0001978398525743342,
+      "loss": 2.8492,
+      "step": 1419
+    },
+    {
+      "epoch": 0.11459930594786538,
+      "grad_norm": 1.06125807762146,
+      "learning_rate": 0.0001978365877667258,
+      "loss": 2.8041,
+      "step": 1420
+    },
+    {
+      "epoch": 0.11468000968444839,
+      "grad_norm": 0.8194011449813843,
+      "learning_rate": 0.00019783332052077386,
+      "loss": 2.7109,
+      "step": 1421
+    },
+    {
+      "epoch": 0.11476071342103139,
+      "grad_norm": 0.972620964050293,
+      "learning_rate": 0.00019783005083655984,
+      "loss": 2.8107,
+      "step": 1422
+    },
+    {
+      "epoch": 0.1148414171576144,
+      "grad_norm": 0.925410270690918,
+      "learning_rate": 0.0001978267787141652,
+      "loss": 2.7603,
+      "step": 1423
+    },
+    {
+      "epoch": 0.1149221208941974,
+      "grad_norm": 0.920156717300415,
+      "learning_rate": 0.00019782350415367152,
+      "loss": 2.7644,
+      "step": 1424
+    },
+    {
+      "epoch": 0.1150028246307804,
+      "grad_norm": 0.8617576360702515,
+      "learning_rate": 0.00019782022715516043,
+      "loss": 2.769,
+      "step": 1425
+    },
+    {
+      "epoch": 0.11508352836736341,
+      "grad_norm": 1.0987342596054077,
+      "learning_rate": 0.00019781694771871356,
+      "loss": 2.8224,
+      "step": 1426
+    },
+    {
+      "epoch": 0.11516423210394641,
+      "grad_norm": 0.8418076634407043,
+      "learning_rate": 0.00019781366584441264,
+      "loss": 2.7947,
+      "step": 1427
+    },
+    {
+      "epoch": 0.11524493584052942,
+      "grad_norm": 0.8010901808738708,
+      "learning_rate": 0.0001978103815323395,
+      "loss": 2.733,
+      "step": 1428
+    },
+    {
+      "epoch": 0.11532563957711242,
+      "grad_norm": 0.8649042844772339,
+      "learning_rate": 0.00019780709478257598,
+      "loss": 2.7681,
+      "step": 1429
+    },
+    {
+      "epoch": 0.11540634331369543,
+      "grad_norm": 0.7728127837181091,
+      "learning_rate": 0.00019780380559520397,
+      "loss": 2.7795,
+      "step": 1430
+    },
+    {
+      "epoch": 0.11548704705027843,
+      "grad_norm": 0.7770940065383911,
+      "learning_rate": 0.00019780051397030545,
+      "loss": 2.743,
+      "step": 1431
+    },
+    {
+      "epoch": 0.11556775078686143,
+      "grad_norm": 0.8341890573501587,
+      "learning_rate": 0.0001977972199079625,
+      "loss": 2.8047,
+      "step": 1432
+    },
+    {
+      "epoch": 0.11564845452344444,
+      "grad_norm": 0.7894187569618225,
+      "learning_rate": 0.00019779392340825717,
+      "loss": 2.7757,
+      "step": 1433
+    },
+    {
+      "epoch": 0.11572915826002744,
+      "grad_norm": 0.8002873063087463,
+      "learning_rate": 0.00019779062447127164,
+      "loss": 2.7816,
+      "step": 1434
+    },
+    {
+      "epoch": 0.11580986199661045,
+      "grad_norm": 0.8256075978279114,
+      "learning_rate": 0.0001977873230970881,
+      "loss": 2.7839,
+      "step": 1435
+    },
+    {
+      "epoch": 0.11589056573319345,
+      "grad_norm": 0.8695322871208191,
+      "learning_rate": 0.0001977840192857889,
+      "loss": 2.746,
+      "step": 1436
+    },
+    {
+      "epoch": 0.11597126946977646,
+      "grad_norm": 0.767425537109375,
+      "learning_rate": 0.00019778071303745628,
+      "loss": 2.797,
+      "step": 1437
+    },
+    {
+      "epoch": 0.11605197320635946,
+      "grad_norm": 0.8263241052627563,
+      "learning_rate": 0.0001977774043521727,
+      "loss": 2.7702,
+      "step": 1438
+    },
+    {
+      "epoch": 0.11613267694294246,
+      "grad_norm": 0.8108638525009155,
+      "learning_rate": 0.0001977740932300206,
+      "loss": 2.6981,
+      "step": 1439
+    },
+    {
+      "epoch": 0.11621338067952547,
+      "grad_norm": 0.7945007681846619,
+      "learning_rate": 0.00019777077967108255,
+      "loss": 2.7357,
+      "step": 1440
+    },
+    {
+      "epoch": 0.11629408441610846,
+      "grad_norm": 0.8480326533317566,
+      "learning_rate": 0.00019776746367544107,
+      "loss": 2.8563,
+      "step": 1441
+    },
+    {
+      "epoch": 0.11637478815269146,
+      "grad_norm": 0.8202071785926819,
+      "learning_rate": 0.00019776414524317882,
+      "loss": 2.7955,
+      "step": 1442
+    },
+    {
+      "epoch": 0.11645549188927447,
+      "grad_norm": 0.8202874660491943,
+      "learning_rate": 0.00019776082437437852,
+      "loss": 2.765,
+      "step": 1443
+    },
+    {
+      "epoch": 0.11653619562585747,
+      "grad_norm": 0.8053051829338074,
+      "learning_rate": 0.00019775750106912294,
+      "loss": 2.6866,
+      "step": 1444
+    },
+    {
+      "epoch": 0.11661689936244048,
+      "grad_norm": 0.831968367099762,
+      "learning_rate": 0.00019775417532749486,
+      "loss": 2.7022,
+      "step": 1445
+    },
+    {
+      "epoch": 0.11669760309902348,
+      "grad_norm": 0.8903129696846008,
+      "learning_rate": 0.00019775084714957725,
+      "loss": 2.7308,
+      "step": 1446
+    },
+    {
+      "epoch": 0.11677830683560649,
+      "grad_norm": 0.8178622722625732,
+      "learning_rate": 0.000197747516535453,
+      "loss": 2.7446,
+      "step": 1447
+    },
+    {
+      "epoch": 0.11685901057218949,
+      "grad_norm": 0.8270576596260071,
+      "learning_rate": 0.00019774418348520508,
+      "loss": 2.7716,
+      "step": 1448
+    },
+    {
+      "epoch": 0.1169397143087725,
+      "grad_norm": 0.7965807914733887,
+      "learning_rate": 0.00019774084799891662,
+      "loss": 2.7305,
+      "step": 1449
+    },
+    {
+      "epoch": 0.1170204180453555,
+      "grad_norm": 0.8499472737312317,
+      "learning_rate": 0.00019773751007667073,
+      "loss": 2.7584,
+      "step": 1450
+    },
+    {
+      "epoch": 0.1171011217819385,
+      "grad_norm": 0.8961663842201233,
+      "learning_rate": 0.0001977341697185506,
+      "loss": 2.7729,
+      "step": 1451
+    },
+    {
+      "epoch": 0.1171818255185215,
+      "grad_norm": 1.0203527212142944,
+      "learning_rate": 0.0001977308269246395,
+      "loss": 2.727,
+      "step": 1452
+    },
+    {
+      "epoch": 0.11726252925510451,
+      "grad_norm": 0.953289806842804,
+      "learning_rate": 0.0001977274816950207,
+      "loss": 2.8158,
+      "step": 1453
+    },
+    {
+      "epoch": 0.11734323299168752,
+      "grad_norm": 1.0064597129821777,
+      "learning_rate": 0.0001977241340297776,
+      "loss": 2.8743,
+      "step": 1454
+    },
+    {
+      "epoch": 0.11742393672827052,
+      "grad_norm": 0.8541988730430603,
+      "learning_rate": 0.00019772078392899363,
+      "loss": 2.8532,
+      "step": 1455
+    },
+    {
+      "epoch": 0.11750464046485352,
+      "grad_norm": 0.8351433873176575,
+      "learning_rate": 0.00019771743139275228,
+      "loss": 2.7749,
+      "step": 1456
+    },
+    {
+      "epoch": 0.11758534420143653,
+      "grad_norm": 0.9555812478065491,
+      "learning_rate": 0.00019771407642113712,
+      "loss": 2.7408,
+      "step": 1457
+    },
+    {
+      "epoch": 0.11766604793801953,
+      "grad_norm": 0.7943894267082214,
+      "learning_rate": 0.0001977107190142317,
+      "loss": 2.7265,
+      "step": 1458
+    },
+    {
+      "epoch": 0.11774675167460254,
+      "grad_norm": 0.8636460900306702,
+      "learning_rate": 0.0001977073591721198,
+      "loss": 2.8178,
+      "step": 1459
+    },
+    {
+      "epoch": 0.11782745541118554,
+      "grad_norm": 0.8673834800720215,
+      "learning_rate": 0.00019770399689488506,
+      "loss": 2.7928,
+      "step": 1460
+    },
+    {
+      "epoch": 0.11790815914776855,
+      "grad_norm": 0.9463722705841064,
+      "learning_rate": 0.00019770063218261133,
+      "loss": 2.7448,
+      "step": 1461
+    },
+    {
+      "epoch": 0.11798886288435155,
+      "grad_norm": 0.8429726362228394,
+      "learning_rate": 0.00019769726503538246,
+      "loss": 2.7564,
+      "step": 1462
+    },
+    {
+      "epoch": 0.11806956662093455,
+      "grad_norm": 0.9412201642990112,
+      "learning_rate": 0.00019769389545328236,
+      "loss": 2.793,
+      "step": 1463
+    },
+    {
+      "epoch": 0.11815027035751756,
+      "grad_norm": 0.9112111926078796,
+      "learning_rate": 0.000197690523436395,
+      "loss": 2.7787,
+      "step": 1464
+    },
+    {
+      "epoch": 0.11823097409410056,
+      "grad_norm": 0.8417023420333862,
+      "learning_rate": 0.00019768714898480444,
+      "loss": 2.7654,
+      "step": 1465
+    },
+    {
+      "epoch": 0.11831167783068357,
+      "grad_norm": 0.8275290727615356,
+      "learning_rate": 0.00019768377209859476,
+      "loss": 2.7914,
+      "step": 1466
+    },
+    {
+      "epoch": 0.11839238156726656,
+      "grad_norm": 0.8113142848014832,
+      "learning_rate": 0.00019768039277785017,
+      "loss": 2.7516,
+      "step": 1467
+    },
+    {
+      "epoch": 0.11847308530384956,
+      "grad_norm": 0.8655288219451904,
+      "learning_rate": 0.0001976770110226548,
+      "loss": 2.8158,
+      "step": 1468
+    },
+    {
+      "epoch": 0.11855378904043257,
+      "grad_norm": 0.8063547611236572,
+      "learning_rate": 0.000197673626833093,
+      "loss": 2.7624,
+      "step": 1469
+    },
+    {
+      "epoch": 0.11863449277701557,
+      "grad_norm": 0.843772292137146,
+      "learning_rate": 0.00019767024020924908,
+      "loss": 2.86,
+      "step": 1470
+    },
+    {
+      "epoch": 0.11871519651359858,
+      "grad_norm": 0.7942481637001038,
+      "learning_rate": 0.0001976668511512075,
+      "loss": 2.758,
+      "step": 1471
+    },
+    {
+      "epoch": 0.11879590025018158,
+      "grad_norm": 0.841275155544281,
+      "learning_rate": 0.00019766345965905268,
+      "loss": 2.8014,
+      "step": 1472
+    },
+    {
+      "epoch": 0.11887660398676458,
+      "grad_norm": 0.8003600835800171,
+      "learning_rate": 0.00019766006573286915,
+      "loss": 2.7829,
+      "step": 1473
+    },
+    {
+      "epoch": 0.11895730772334759,
+      "grad_norm": 0.8437239527702332,
+      "learning_rate": 0.00019765666937274147,
+      "loss": 2.7706,
+      "step": 1474
+    },
+    {
+      "epoch": 0.11903801145993059,
+      "grad_norm": 0.8118240833282471,
+      "learning_rate": 0.00019765327057875433,
+      "loss": 2.8185,
+      "step": 1475
+    },
+    {
+      "epoch": 0.1191187151965136,
+      "grad_norm": 0.8051649928092957,
+      "learning_rate": 0.00019764986935099244,
+      "loss": 2.7676,
+      "step": 1476
+    },
+    {
+      "epoch": 0.1191994189330966,
+      "grad_norm": 0.7786862850189209,
+      "learning_rate": 0.00019764646568954053,
+      "loss": 2.8069,
+      "step": 1477
+    },
+    {
+      "epoch": 0.1192801226696796,
+      "grad_norm": 0.8199592232704163,
+      "learning_rate": 0.0001976430595944834,
+      "loss": 2.7718,
+      "step": 1478
+    },
+    {
+      "epoch": 0.11936082640626261,
+      "grad_norm": 0.8696652054786682,
+      "learning_rate": 0.00019763965106590604,
+      "loss": 2.7682,
+      "step": 1479
+    },
+    {
+      "epoch": 0.11944153014284561,
+      "grad_norm": 0.7993931174278259,
+      "learning_rate": 0.00019763624010389334,
+      "loss": 2.7607,
+      "step": 1480
+    },
+    {
+      "epoch": 0.11952223387942862,
+      "grad_norm": 0.8107055425643921,
+      "learning_rate": 0.0001976328267085303,
+      "loss": 2.7885,
+      "step": 1481
+    },
+    {
+      "epoch": 0.11960293761601162,
+      "grad_norm": 0.8189423084259033,
+      "learning_rate": 0.000197629410879902,
+      "loss": 2.7332,
+      "step": 1482
+    },
+    {
+      "epoch": 0.11968364135259463,
+      "grad_norm": 0.9134814143180847,
+      "learning_rate": 0.0001976259926180936,
+      "loss": 2.7691,
+      "step": 1483
+    },
+    {
+      "epoch": 0.11976434508917763,
+      "grad_norm": 0.8642883896827698,
+      "learning_rate": 0.00019762257192319023,
+      "loss": 2.7876,
+      "step": 1484
+    },
+    {
+      "epoch": 0.11984504882576064,
+      "grad_norm": 0.7411352396011353,
+      "learning_rate": 0.0001976191487952772,
+      "loss": 2.7577,
+      "step": 1485
+    },
+    {
+      "epoch": 0.11992575256234364,
+      "grad_norm": 0.7741669416427612,
+      "learning_rate": 0.00019761572323443978,
+      "loss": 2.8005,
+      "step": 1486
+    },
+    {
+      "epoch": 0.12000645629892664,
+      "grad_norm": 0.8195405602455139,
+      "learning_rate": 0.0001976122952407634,
+      "loss": 2.7421,
+      "step": 1487
+    },
+    {
+      "epoch": 0.12008716003550965,
+      "grad_norm": 0.8355886936187744,
+      "learning_rate": 0.00019760886481433345,
+      "loss": 2.8156,
+      "step": 1488
+    },
+    {
+      "epoch": 0.12016786377209265,
+      "grad_norm": 0.8321093916893005,
+      "learning_rate": 0.00019760543195523542,
+      "loss": 2.7261,
+      "step": 1489
+    },
+    {
+      "epoch": 0.12024856750867566,
+      "grad_norm": 0.7792446613311768,
+      "learning_rate": 0.0001976019966635549,
+      "loss": 2.7319,
+      "step": 1490
+    },
+    {
+      "epoch": 0.12032927124525866,
+      "grad_norm": 0.770535409450531,
+      "learning_rate": 0.00019759855893937748,
+      "loss": 2.7727,
+      "step": 1491
+    },
+    {
+      "epoch": 0.12040997498184165,
+      "grad_norm": 0.8168532252311707,
+      "learning_rate": 0.00019759511878278887,
+      "loss": 2.7763,
+      "step": 1492
+    },
+    {
+      "epoch": 0.12049067871842466,
+      "grad_norm": 0.8395755290985107,
+      "learning_rate": 0.00019759167619387476,
+      "loss": 2.8382,
+      "step": 1493
+    },
+    {
+      "epoch": 0.12057138245500766,
+      "grad_norm": 0.8682762384414673,
+      "learning_rate": 0.00019758823117272097,
+      "loss": 2.8056,
+      "step": 1494
+    },
+    {
+      "epoch": 0.12065208619159067,
+      "grad_norm": 0.815192699432373,
+      "learning_rate": 0.00019758478371941337,
+      "loss": 2.7602,
+      "step": 1495
+    },
+    {
+      "epoch": 0.12073278992817367,
+      "grad_norm": 0.7919273376464844,
+      "learning_rate": 0.00019758133383403786,
+      "loss": 2.7989,
+      "step": 1496
+    },
+    {
+      "epoch": 0.12081349366475667,
+      "grad_norm": 1.004387378692627,
+      "learning_rate": 0.00019757788151668045,
+      "loss": 2.7765,
+      "step": 1497
+    },
+    {
+      "epoch": 0.12089419740133968,
+      "grad_norm": 1.0032062530517578,
+      "learning_rate": 0.00019757442676742715,
+      "loss": 2.7751,
+      "step": 1498
+    },
+    {
+      "epoch": 0.12097490113792268,
+      "grad_norm": 0.8797723054885864,
+      "learning_rate": 0.00019757096958636407,
+      "loss": 2.7798,
+      "step": 1499
+    },
+    {
+      "epoch": 0.12105560487450569,
+      "grad_norm": 0.9239820241928101,
+      "learning_rate": 0.0001975675099735774,
+      "loss": 2.7976,
+      "step": 1500
+    },
+    {
+      "epoch": 0.12113630861108869,
+      "grad_norm": 0.9903601408004761,
+      "learning_rate": 0.00019756404792915328,
+      "loss": 2.7891,
+      "step": 1501
+    },
+    {
+      "epoch": 0.1212170123476717,
+      "grad_norm": 0.8402895331382751,
+      "learning_rate": 0.0001975605834531781,
+      "loss": 2.8037,
+      "step": 1502
+    },
+    {
+      "epoch": 0.1212977160842547,
+      "grad_norm": 0.8986102342605591,
+      "learning_rate": 0.00019755711654573813,
+      "loss": 2.8375,
+      "step": 1503
+    },
+    {
+      "epoch": 0.1213784198208377,
+      "grad_norm": 0.8795471787452698,
+      "learning_rate": 0.0001975536472069198,
+      "loss": 2.7916,
+      "step": 1504
+    },
+    {
+      "epoch": 0.12145912355742071,
+      "grad_norm": 0.866278350353241,
+      "learning_rate": 0.00019755017543680962,
+      "loss": 2.7884,
+      "step": 1505
+    },
+    {
+      "epoch": 0.12153982729400371,
+      "grad_norm": 0.7877952456474304,
+      "learning_rate": 0.00019754670123549398,
+      "loss": 2.7659,
+      "step": 1506
+    },
+    {
+      "epoch": 0.12162053103058672,
+      "grad_norm": 0.857155978679657,
+      "learning_rate": 0.00019754322460305962,
+      "loss": 2.8029,
+      "step": 1507
+    },
+    {
+      "epoch": 0.12170123476716972,
+      "grad_norm": 0.8323284387588501,
+      "learning_rate": 0.00019753974553959314,
+      "loss": 2.7764,
+      "step": 1508
+    },
+    {
+      "epoch": 0.12178193850375273,
+      "grad_norm": 0.8557485938072205,
+      "learning_rate": 0.00019753626404518117,
+      "loss": 2.7448,
+      "step": 1509
+    },
+    {
+      "epoch": 0.12186264224033573,
+      "grad_norm": 0.8026818037033081,
+      "learning_rate": 0.00019753278011991058,
+      "loss": 2.7323,
+      "step": 1510
+    },
+    {
+      "epoch": 0.12194334597691874,
+      "grad_norm": 0.8578904271125793,
+      "learning_rate": 0.00019752929376386816,
+      "loss": 2.759,
+      "step": 1511
+    },
+    {
+      "epoch": 0.12202404971350174,
+      "grad_norm": 0.8617175221443176,
+      "learning_rate": 0.00019752580497714076,
+      "loss": 2.7641,
+      "step": 1512
+    },
+    {
+      "epoch": 0.12210475345008474,
+      "grad_norm": 0.8261943459510803,
+      "learning_rate": 0.00019752231375981538,
+      "loss": 2.7554,
+      "step": 1513
+    },
+    {
+      "epoch": 0.12218545718666775,
+      "grad_norm": 0.9984099268913269,
+      "learning_rate": 0.00019751882011197902,
+      "loss": 2.763,
+      "step": 1514
+    },
+    {
+      "epoch": 0.12226616092325075,
+      "grad_norm": 0.8014064431190491,
+      "learning_rate": 0.00019751532403371874,
+      "loss": 2.8083,
+      "step": 1515
+    },
+    {
+      "epoch": 0.12234686465983376,
+      "grad_norm": 0.9276653528213501,
+      "learning_rate": 0.0001975118255251217,
+      "loss": 2.8055,
+      "step": 1516
+    },
+    {
+      "epoch": 0.12242756839641676,
+      "grad_norm": 0.9365193843841553,
+      "learning_rate": 0.00019750832458627503,
+      "loss": 2.7397,
+      "step": 1517
+    },
+    {
+      "epoch": 0.12250827213299975,
+      "grad_norm": 0.8952646851539612,
+      "learning_rate": 0.00019750482121726605,
+      "loss": 2.8305,
+      "step": 1518
+    },
+    {
+      "epoch": 0.12258897586958276,
+      "grad_norm": 0.8395531177520752,
+      "learning_rate": 0.00019750131541818204,
+      "loss": 2.7852,
+      "step": 1519
+    },
+    {
+      "epoch": 0.12266967960616576,
+      "grad_norm": 0.8123572468757629,
+      "learning_rate": 0.0001974978071891104,
+      "loss": 2.831,
+      "step": 1520
+    },
+    {
+      "epoch": 0.12275038334274876,
+      "grad_norm": 0.8716141581535339,
+      "learning_rate": 0.00019749429653013851,
+      "loss": 2.8012,
+      "step": 1521
+    },
+    {
+      "epoch": 0.12283108707933177,
+      "grad_norm": 0.7848379611968994,
+      "learning_rate": 0.0001974907834413539,
+      "loss": 2.7812,
+      "step": 1522
+    },
+    {
+      "epoch": 0.12291179081591477,
+      "grad_norm": 0.834072470664978,
+      "learning_rate": 0.00019748726792284414,
+      "loss": 2.7442,
+      "step": 1523
+    },
+    {
+      "epoch": 0.12299249455249778,
+      "grad_norm": 0.8377225399017334,
+      "learning_rate": 0.0001974837499746968,
+      "loss": 2.7967,
+      "step": 1524
+    },
+    {
+      "epoch": 0.12307319828908078,
+      "grad_norm": 0.8809494376182556,
+      "learning_rate": 0.0001974802295969996,
+      "loss": 2.8042,
+      "step": 1525
+    },
+    {
+      "epoch": 0.12315390202566379,
+      "grad_norm": 0.8504741787910461,
+      "learning_rate": 0.00019747670678984028,
+      "loss": 2.7909,
+      "step": 1526
+    },
+    {
+      "epoch": 0.12323460576224679,
+      "grad_norm": 0.9444355368614197,
+      "learning_rate": 0.00019747318155330663,
+      "loss": 2.8567,
+      "step": 1527
+    },
+    {
+      "epoch": 0.1233153094988298,
+      "grad_norm": 0.859166145324707,
+      "learning_rate": 0.00019746965388748645,
+      "loss": 2.8305,
+      "step": 1528
+    },
+    {
+      "epoch": 0.1233960132354128,
+      "grad_norm": 0.8431086540222168,
+      "learning_rate": 0.00019746612379246777,
+      "loss": 2.7799,
+      "step": 1529
+    },
+    {
+      "epoch": 0.1234767169719958,
+      "grad_norm": 0.8872438669204712,
+      "learning_rate": 0.00019746259126833846,
+      "loss": 2.8413,
+      "step": 1530
+    },
+    {
+      "epoch": 0.12355742070857881,
+      "grad_norm": 0.8698925375938416,
+      "learning_rate": 0.0001974590563151866,
+      "loss": 2.8446,
+      "step": 1531
+    },
+    {
+      "epoch": 0.12363812444516181,
+      "grad_norm": 0.8926429152488708,
+      "learning_rate": 0.0001974555189331003,
+      "loss": 2.7859,
+      "step": 1532
+    },
+    {
+      "epoch": 0.12371882818174482,
+      "grad_norm": 0.8089048862457275,
+      "learning_rate": 0.00019745197912216775,
+      "loss": 2.7985,
+      "step": 1533
+    },
+    {
+      "epoch": 0.12379953191832782,
+      "grad_norm": 0.8180400729179382,
+      "learning_rate": 0.0001974484368824771,
+      "loss": 2.7587,
+      "step": 1534
+    },
+    {
+      "epoch": 0.12388023565491083,
+      "grad_norm": 0.9584212303161621,
+      "learning_rate": 0.00019744489221411668,
+      "loss": 2.766,
+      "step": 1535
+    },
+    {
+      "epoch": 0.12396093939149383,
+      "grad_norm": 0.8425920009613037,
+      "learning_rate": 0.00019744134511717485,
+      "loss": 2.8125,
+      "step": 1536
+    },
+    {
+      "epoch": 0.12404164312807683,
+      "grad_norm": 0.9109299182891846,
+      "learning_rate": 0.00019743779559173996,
+      "loss": 2.8613,
+      "step": 1537
+    },
+    {
+      "epoch": 0.12412234686465984,
+      "grad_norm": 0.8840214610099792,
+      "learning_rate": 0.0001974342436379005,
+      "loss": 2.7603,
+      "step": 1538
+    },
+    {
+      "epoch": 0.12420305060124284,
+      "grad_norm": 0.8128962516784668,
+      "learning_rate": 0.00019743068925574502,
+      "loss": 2.7593,
+      "step": 1539
+    },
+    {
+      "epoch": 0.12428375433782585,
+      "grad_norm": 0.8150052428245544,
+      "learning_rate": 0.00019742713244536204,
+      "loss": 2.8099,
+      "step": 1540
+    },
+    {
+      "epoch": 0.12436445807440885,
+      "grad_norm": 0.8442968130111694,
+      "learning_rate": 0.00019742357320684027,
+      "loss": 2.7746,
+      "step": 1541
+    },
+    {
+      "epoch": 0.12444516181099186,
+      "grad_norm": 0.9347402453422546,
+      "learning_rate": 0.00019742001154026838,
+      "loss": 2.8247,
+      "step": 1542
+    },
+    {
+      "epoch": 0.12452586554757485,
+      "grad_norm": 0.8305966854095459,
+      "learning_rate": 0.00019741644744573512,
+      "loss": 2.7398,
+      "step": 1543
+    },
+    {
+      "epoch": 0.12460656928415785,
+      "grad_norm": 0.8811129927635193,
+      "learning_rate": 0.00019741288092332935,
+      "loss": 2.8014,
+      "step": 1544
+    },
+    {
+      "epoch": 0.12468727302074085,
+      "grad_norm": 1.0287303924560547,
+      "learning_rate": 0.00019740931197313996,
+      "loss": 2.8449,
+      "step": 1545
+    },
+    {
+      "epoch": 0.12476797675732386,
+      "grad_norm": 0.8499771356582642,
+      "learning_rate": 0.00019740574059525588,
+      "loss": 2.7845,
+      "step": 1546
+    },
+    {
+      "epoch": 0.12484868049390686,
+      "grad_norm": 0.8110969066619873,
+      "learning_rate": 0.00019740216678976614,
+      "loss": 2.7565,
+      "step": 1547
+    },
+    {
+      "epoch": 0.12492938423048987,
+      "grad_norm": 0.8530771136283875,
+      "learning_rate": 0.00019739859055675977,
+      "loss": 2.8098,
+      "step": 1548
+    },
+    {
+      "epoch": 0.12501008796707289,
+      "grad_norm": 0.8483901619911194,
+      "learning_rate": 0.00019739501189632591,
+      "loss": 2.812,
+      "step": 1549
+    },
+    {
+      "epoch": 0.1250907917036559,
+      "grad_norm": 0.7894467711448669,
+      "learning_rate": 0.00019739143080855378,
+      "loss": 2.8576,
+      "step": 1550
+    },
+    {
+      "epoch": 0.1251714954402389,
+      "grad_norm": 0.8270247578620911,
+      "learning_rate": 0.0001973878472935326,
+      "loss": 2.7613,
+      "step": 1551
+    },
+    {
+      "epoch": 0.1252521991768219,
+      "grad_norm": 0.8496212959289551,
+      "learning_rate": 0.00019738426135135174,
+      "loss": 2.8375,
+      "step": 1552
+    },
+    {
+      "epoch": 0.1253329029134049,
+      "grad_norm": 0.8465524911880493,
+      "learning_rate": 0.00019738067298210045,
+      "loss": 2.8023,
+      "step": 1553
+    },
+    {
+      "epoch": 0.1254136066499879,
+      "grad_norm": 0.7843824028968811,
+      "learning_rate": 0.00019737708218586826,
+      "loss": 2.7424,
+      "step": 1554
+    },
+    {
+      "epoch": 0.1254943103865709,
+      "grad_norm": 0.8310040235519409,
+      "learning_rate": 0.00019737348896274462,
+      "loss": 2.7608,
+      "step": 1555
+    },
+    {
+      "epoch": 0.1255750141231539,
+      "grad_norm": 0.7895017266273499,
+      "learning_rate": 0.00019736989331281914,
+      "loss": 2.7549,
+      "step": 1556
+    },
+    {
+      "epoch": 0.1256557178597369,
+      "grad_norm": 0.8140431642532349,
+      "learning_rate": 0.00019736629523618138,
+      "loss": 2.802,
+      "step": 1557
+    },
+    {
+      "epoch": 0.1257364215963199,
+      "grad_norm": 0.8026889562606812,
+      "learning_rate": 0.000197362694732921,
+      "loss": 2.7758,
+      "step": 1558
+    },
+    {
+      "epoch": 0.1258171253329029,
+      "grad_norm": 0.8018048405647278,
+      "learning_rate": 0.0001973590918031278,
+      "loss": 2.7729,
+      "step": 1559
+    },
+    {
+      "epoch": 0.1258978290694859,
+      "grad_norm": 0.8394612073898315,
+      "learning_rate": 0.00019735548644689147,
+      "loss": 2.7692,
+      "step": 1560
+    },
+    {
+      "epoch": 0.1259785328060689,
+      "grad_norm": 0.819804310798645,
+      "learning_rate": 0.00019735187866430198,
+      "loss": 2.6933,
+      "step": 1561
+    },
+    {
+      "epoch": 0.12605923654265191,
+      "grad_norm": 0.8094257116317749,
+      "learning_rate": 0.0001973482684554492,
+      "loss": 2.7722,
+      "step": 1562
+    },
+    {
+      "epoch": 0.12613994027923492,
+      "grad_norm": 0.8647315502166748,
+      "learning_rate": 0.00019734465582042305,
+      "loss": 2.787,
+      "step": 1563
+    },
+    {
+      "epoch": 0.12622064401581792,
+      "grad_norm": 0.8439335823059082,
+      "learning_rate": 0.00019734104075931367,
+      "loss": 2.8,
+      "step": 1564
+    },
+    {
+      "epoch": 0.12630134775240093,
+      "grad_norm": 0.852480947971344,
+      "learning_rate": 0.00019733742327221105,
+      "loss": 2.8656,
+      "step": 1565
+    },
+    {
+      "epoch": 0.12638205148898393,
+      "grad_norm": 0.813846230506897,
+      "learning_rate": 0.00019733380335920542,
+      "loss": 2.7733,
+      "step": 1566
+    },
+    {
+      "epoch": 0.12646275522556694,
+      "grad_norm": 0.7860896587371826,
+      "learning_rate": 0.00019733018102038698,
+      "loss": 2.8201,
+      "step": 1567
+    },
+    {
+      "epoch": 0.12654345896214994,
+      "grad_norm": 0.7857748866081238,
+      "learning_rate": 0.00019732655625584602,
+      "loss": 2.8726,
+      "step": 1568
+    },
+    {
+      "epoch": 0.12662416269873294,
+      "grad_norm": 0.8152899146080017,
+      "learning_rate": 0.00019732292906567286,
+      "loss": 2.7738,
+      "step": 1569
+    },
+    {
+      "epoch": 0.12670486643531595,
+      "grad_norm": 0.8281696438789368,
+      "learning_rate": 0.00019731929944995788,
+      "loss": 2.7966,
+      "step": 1570
+    },
+    {
+      "epoch": 0.12678557017189895,
+      "grad_norm": 0.8070773482322693,
+      "learning_rate": 0.00019731566740879158,
+      "loss": 2.6988,
+      "step": 1571
+    },
+    {
+      "epoch": 0.12686627390848196,
+      "grad_norm": 0.7859680652618408,
+      "learning_rate": 0.00019731203294226445,
+      "loss": 2.7241,
+      "step": 1572
+    },
+    {
+      "epoch": 0.12694697764506496,
+      "grad_norm": 0.7753982543945312,
+      "learning_rate": 0.0001973083960504671,
+      "loss": 2.7621,
+      "step": 1573
+    },
+    {
+      "epoch": 0.12702768138164797,
+      "grad_norm": 0.8063471913337708,
+      "learning_rate": 0.00019730475673349014,
+      "loss": 2.7298,
+      "step": 1574
+    },
+    {
+      "epoch": 0.12710838511823097,
+      "grad_norm": 0.7943962812423706,
+      "learning_rate": 0.0001973011149914243,
+      "loss": 2.7714,
+      "step": 1575
+    },
+    {
+      "epoch": 0.12718908885481398,
+      "grad_norm": 0.8297483325004578,
+      "learning_rate": 0.00019729747082436033,
+      "loss": 2.7743,
+      "step": 1576
+    },
+    {
+      "epoch": 0.12726979259139698,
+      "grad_norm": 0.8728111386299133,
+      "learning_rate": 0.000197293824232389,
+      "loss": 2.8251,
+      "step": 1577
+    },
+    {
+      "epoch": 0.12735049632797998,
+      "grad_norm": 0.8762480020523071,
+      "learning_rate": 0.00019729017521560128,
+      "loss": 2.8036,
+      "step": 1578
+    },
+    {
+      "epoch": 0.127431200064563,
+      "grad_norm": 0.9266185164451599,
+      "learning_rate": 0.00019728652377408806,
+      "loss": 2.7335,
+      "step": 1579
+    },
+    {
+      "epoch": 0.127511903801146,
+      "grad_norm": 0.9289839267730713,
+      "learning_rate": 0.00019728286990794037,
+      "loss": 2.7715,
+      "step": 1580
+    },
+    {
+      "epoch": 0.127592607537729,
+      "grad_norm": 0.8811823725700378,
+      "learning_rate": 0.0001972792136172493,
+      "loss": 2.7389,
+      "step": 1581
+    },
+    {
+      "epoch": 0.127673311274312,
+      "grad_norm": 0.8174294233322144,
+      "learning_rate": 0.00019727555490210588,
+      "loss": 2.7483,
+      "step": 1582
+    },
+    {
+      "epoch": 0.127754015010895,
+      "grad_norm": 0.8254107236862183,
+      "learning_rate": 0.00019727189376260137,
+      "loss": 2.7897,
+      "step": 1583
+    },
+    {
+      "epoch": 0.127834718747478,
+      "grad_norm": 0.8478763699531555,
+      "learning_rate": 0.000197268230198827,
+      "loss": 2.7394,
+      "step": 1584
+    },
+    {
+      "epoch": 0.12791542248406101,
+      "grad_norm": 0.8356192111968994,
+      "learning_rate": 0.00019726456421087404,
+      "loss": 2.7518,
+      "step": 1585
+    },
+    {
+      "epoch": 0.12799612622064402,
+      "grad_norm": 0.8523107767105103,
+      "learning_rate": 0.00019726089579883392,
+      "loss": 2.7893,
+      "step": 1586
+    },
+    {
+      "epoch": 0.12807682995722702,
+      "grad_norm": 0.9048579931259155,
+      "learning_rate": 0.00019725722496279804,
+      "loss": 2.7488,
+      "step": 1587
+    },
+    {
+      "epoch": 0.12815753369381003,
+      "grad_norm": 0.8242251873016357,
+      "learning_rate": 0.00019725355170285787,
+      "loss": 2.7544,
+      "step": 1588
+    },
+    {
+      "epoch": 0.12823823743039303,
+      "grad_norm": 0.8343983888626099,
+      "learning_rate": 0.00019724987601910497,
+      "loss": 2.7317,
+      "step": 1589
+    },
+    {
+      "epoch": 0.12831894116697604,
+      "grad_norm": 0.8084509372711182,
+      "learning_rate": 0.00019724619791163095,
+      "loss": 2.7822,
+      "step": 1590
+    },
+    {
+      "epoch": 0.12839964490355904,
+      "grad_norm": 0.8397380113601685,
+      "learning_rate": 0.00019724251738052745,
+      "loss": 2.8188,
+      "step": 1591
+    },
+    {
+      "epoch": 0.12848034864014204,
+      "grad_norm": 0.8558558821678162,
+      "learning_rate": 0.00019723883442588624,
+      "loss": 2.7623,
+      "step": 1592
+    },
+    {
+      "epoch": 0.12856105237672505,
+      "grad_norm": 0.7602639198303223,
+      "learning_rate": 0.0001972351490477991,
+      "loss": 2.7932,
+      "step": 1593
+    },
+    {
+      "epoch": 0.12864175611330805,
+      "grad_norm": 0.8379851579666138,
+      "learning_rate": 0.00019723146124635786,
+      "loss": 2.8296,
+      "step": 1594
+    },
+    {
+      "epoch": 0.12872245984989106,
+      "grad_norm": 0.8454548716545105,
+      "learning_rate": 0.00019722777102165444,
+      "loss": 2.8192,
+      "step": 1595
+    },
+    {
+      "epoch": 0.12880316358647406,
+      "grad_norm": 0.8344082832336426,
+      "learning_rate": 0.0001972240783737808,
+      "loss": 2.7628,
+      "step": 1596
+    },
+    {
+      "epoch": 0.12888386732305707,
+      "grad_norm": 0.809093713760376,
+      "learning_rate": 0.000197220383302829,
+      "loss": 2.8055,
+      "step": 1597
+    },
+    {
+      "epoch": 0.12896457105964007,
+      "grad_norm": 0.7909694910049438,
+      "learning_rate": 0.0001972166858088911,
+      "loss": 2.7292,
+      "step": 1598
+    },
+    {
+      "epoch": 0.12904527479622308,
+      "grad_norm": 0.8350280523300171,
+      "learning_rate": 0.00019721298589205928,
+      "loss": 2.7671,
+      "step": 1599
+    },
+    {
+      "epoch": 0.12912597853280608,
+      "grad_norm": 0.7857616543769836,
+      "learning_rate": 0.00019720928355242568,
+      "loss": 2.729,
+      "step": 1600
+    },
+    {
+      "epoch": 0.12920668226938908,
+      "grad_norm": 0.7899746298789978,
+      "learning_rate": 0.0001972055787900827,
+      "loss": 2.8023,
+      "step": 1601
+    },
+    {
+      "epoch": 0.1292873860059721,
+      "grad_norm": 0.8604246377944946,
+      "learning_rate": 0.00019720187160512256,
+      "loss": 2.749,
+      "step": 1602
+    },
+    {
+      "epoch": 0.1293680897425551,
+      "grad_norm": 0.8517864942550659,
+      "learning_rate": 0.0001971981619976377,
+      "loss": 2.7203,
+      "step": 1603
+    },
+    {
+      "epoch": 0.1294487934791381,
+      "grad_norm": 0.8860471248626709,
+      "learning_rate": 0.00019719444996772056,
+      "loss": 2.7372,
+      "step": 1604
+    },
+    {
+      "epoch": 0.1295294972157211,
+      "grad_norm": 0.8355888724327087,
+      "learning_rate": 0.00019719073551546367,
+      "loss": 2.7284,
+      "step": 1605
+    },
+    {
+      "epoch": 0.1296102009523041,
+      "grad_norm": 0.7998479604721069,
+      "learning_rate": 0.00019718701864095955,
+      "loss": 2.7726,
+      "step": 1606
+    },
+    {
+      "epoch": 0.12969090468888708,
+      "grad_norm": 0.8564549088478088,
+      "learning_rate": 0.00019718329934430092,
+      "loss": 2.7334,
+      "step": 1607
+    },
+    {
+      "epoch": 0.1297716084254701,
+      "grad_norm": 0.8594443798065186,
+      "learning_rate": 0.00019717957762558044,
+      "loss": 2.7865,
+      "step": 1608
+    },
+    {
+      "epoch": 0.1298523121620531,
+      "grad_norm": 0.804553210735321,
+      "learning_rate": 0.00019717585348489082,
+      "loss": 2.8094,
+      "step": 1609
+    },
+    {
+      "epoch": 0.1299330158986361,
+      "grad_norm": 0.7892553806304932,
+      "learning_rate": 0.0001971721269223249,
+      "loss": 2.7969,
+      "step": 1610
+    },
+    {
+      "epoch": 0.1300137196352191,
+      "grad_norm": 0.8703331351280212,
+      "learning_rate": 0.0001971683979379756,
+      "loss": 2.8192,
+      "step": 1611
+    },
+    {
+      "epoch": 0.1300944233718021,
+      "grad_norm": 0.8176589012145996,
+      "learning_rate": 0.00019716466653193582,
+      "loss": 2.7902,
+      "step": 1612
+    },
+    {
+      "epoch": 0.1301751271083851,
+      "grad_norm": 0.8305137157440186,
+      "learning_rate": 0.00019716093270429855,
+      "loss": 2.8202,
+      "step": 1613
+    },
+    {
+      "epoch": 0.1302558308449681,
+      "grad_norm": 0.8261505365371704,
+      "learning_rate": 0.00019715719645515688,
+      "loss": 2.7905,
+      "step": 1614
+    },
+    {
+      "epoch": 0.13033653458155112,
+      "grad_norm": 0.9465535879135132,
+      "learning_rate": 0.00019715345778460389,
+      "loss": 2.7965,
+      "step": 1615
+    },
+    {
+      "epoch": 0.13041723831813412,
+      "grad_norm": 0.8847100138664246,
+      "learning_rate": 0.00019714971669273275,
+      "loss": 2.8177,
+      "step": 1616
+    },
+    {
+      "epoch": 0.13049794205471713,
+      "grad_norm": 0.9768328666687012,
+      "learning_rate": 0.0001971459731796367,
+      "loss": 2.7668,
+      "step": 1617
+    },
+    {
+      "epoch": 0.13057864579130013,
+      "grad_norm": 0.7498586177825928,
+      "learning_rate": 0.0001971422272454091,
+      "loss": 2.761,
+      "step": 1618
+    },
+    {
+      "epoch": 0.13065934952788313,
+      "grad_norm": 1.0455373525619507,
+      "learning_rate": 0.00019713847889014325,
+      "loss": 2.7652,
+      "step": 1619
+    },
+    {
+      "epoch": 0.13074005326446614,
+      "grad_norm": 0.8484631180763245,
+      "learning_rate": 0.00019713472811393258,
+      "loss": 2.7858,
+      "step": 1620
+    },
+    {
+      "epoch": 0.13082075700104914,
+      "grad_norm": 0.8190686702728271,
+      "learning_rate": 0.00019713097491687057,
+      "loss": 2.7217,
+      "step": 1621
+    },
+    {
+      "epoch": 0.13090146073763215,
+      "grad_norm": 0.8866000175476074,
+      "learning_rate": 0.00019712721929905077,
+      "loss": 2.7868,
+      "step": 1622
+    },
+    {
+      "epoch": 0.13098216447421515,
+      "grad_norm": 0.8026713132858276,
+      "learning_rate": 0.00019712346126056677,
+      "loss": 2.7276,
+      "step": 1623
+    },
+    {
+      "epoch": 0.13106286821079816,
+      "grad_norm": 0.8306462168693542,
+      "learning_rate": 0.00019711970080151225,
+      "loss": 2.7747,
+      "step": 1624
+    },
+    {
+      "epoch": 0.13114357194738116,
+      "grad_norm": 0.8276618123054504,
+      "learning_rate": 0.0001971159379219809,
+      "loss": 2.7146,
+      "step": 1625
+    },
+    {
+      "epoch": 0.13122427568396416,
+      "grad_norm": 0.9749011993408203,
+      "learning_rate": 0.00019711217262206648,
+      "loss": 2.8731,
+      "step": 1626
+    },
+    {
+      "epoch": 0.13130497942054717,
+      "grad_norm": 0.828484058380127,
+      "learning_rate": 0.00019710840490186292,
+      "loss": 2.803,
+      "step": 1627
+    },
+    {
+      "epoch": 0.13138568315713017,
+      "grad_norm": 0.8095957636833191,
+      "learning_rate": 0.00019710463476146402,
+      "loss": 2.7751,
+      "step": 1628
+    },
+    {
+      "epoch": 0.13146638689371318,
+      "grad_norm": 0.8731853365898132,
+      "learning_rate": 0.0001971008622009638,
+      "loss": 2.8274,
+      "step": 1629
+    },
+    {
+      "epoch": 0.13154709063029618,
+      "grad_norm": 0.8180200457572937,
+      "learning_rate": 0.00019709708722045628,
+      "loss": 2.813,
+      "step": 1630
+    },
+    {
+      "epoch": 0.13162779436687919,
+      "grad_norm": 0.7740067839622498,
+      "learning_rate": 0.00019709330982003553,
+      "loss": 2.7319,
+      "step": 1631
+    },
+    {
+      "epoch": 0.1317084981034622,
+      "grad_norm": 0.8439326882362366,
+      "learning_rate": 0.0001970895299997957,
+      "loss": 2.8182,
+      "step": 1632
+    },
+    {
+      "epoch": 0.1317892018400452,
+      "grad_norm": 0.8254802823066711,
+      "learning_rate": 0.000197085747759831,
+      "loss": 2.7874,
+      "step": 1633
+    },
+    {
+      "epoch": 0.1318699055766282,
+      "grad_norm": 0.8128175139427185,
+      "learning_rate": 0.00019708196310023562,
+      "loss": 2.8125,
+      "step": 1634
+    },
+    {
+      "epoch": 0.1319506093132112,
+      "grad_norm": 0.8664820790290833,
+      "learning_rate": 0.00019707817602110402,
+      "loss": 2.8446,
+      "step": 1635
+    },
+    {
+      "epoch": 0.1320313130497942,
+      "grad_norm": 0.8101332783699036,
+      "learning_rate": 0.00019707438652253044,
+      "loss": 2.8027,
+      "step": 1636
+    },
+    {
+      "epoch": 0.1321120167863772,
+      "grad_norm": 0.8296725153923035,
+      "learning_rate": 0.00019707059460460945,
+      "loss": 2.7677,
+      "step": 1637
+    },
+    {
+      "epoch": 0.13219272052296022,
+      "grad_norm": 0.7321150898933411,
+      "learning_rate": 0.0001970668002674355,
+      "loss": 2.6991,
+      "step": 1638
+    },
+    {
+      "epoch": 0.13227342425954322,
+      "grad_norm": 0.8321375250816345,
+      "learning_rate": 0.0001970630035111031,
+      "loss": 2.6948,
+      "step": 1639
+    },
+    {
+      "epoch": 0.13235412799612623,
+      "grad_norm": 0.7622714042663574,
+      "learning_rate": 0.00019705920433570694,
+      "loss": 2.6957,
+      "step": 1640
+    },
+    {
+      "epoch": 0.13243483173270923,
+      "grad_norm": 0.8413416147232056,
+      "learning_rate": 0.00019705540274134173,
+      "loss": 2.7277,
+      "step": 1641
+    },
+    {
+      "epoch": 0.13251553546929223,
+      "grad_norm": 0.8798941373825073,
+      "learning_rate": 0.00019705159872810218,
+      "loss": 2.7699,
+      "step": 1642
+    },
+    {
+      "epoch": 0.13259623920587524,
+      "grad_norm": 0.788287341594696,
+      "learning_rate": 0.00019704779229608304,
+      "loss": 2.7933,
+      "step": 1643
+    },
+    {
+      "epoch": 0.13267694294245824,
+      "grad_norm": 0.8547430634498596,
+      "learning_rate": 0.00019704398344537927,
+      "loss": 2.7706,
+      "step": 1644
+    },
+    {
+      "epoch": 0.13275764667904125,
+      "grad_norm": 0.8474008440971375,
+      "learning_rate": 0.00019704017217608575,
+      "loss": 2.8005,
+      "step": 1645
+    },
+    {
+      "epoch": 0.13283835041562425,
+      "grad_norm": 0.8636945486068726,
+      "learning_rate": 0.00019703635848829747,
+      "loss": 2.8241,
+      "step": 1646
+    },
+    {
+      "epoch": 0.13291905415220726,
+      "grad_norm": 0.8158168792724609,
+      "learning_rate": 0.00019703254238210947,
+      "loss": 2.7576,
+      "step": 1647
+    },
+    {
+      "epoch": 0.13299975788879026,
+      "grad_norm": 0.8420887589454651,
+      "learning_rate": 0.0001970287238576169,
+      "loss": 2.7677,
+      "step": 1648
+    },
+    {
+      "epoch": 0.13308046162537326,
+      "grad_norm": 0.7910059690475464,
+      "learning_rate": 0.00019702490291491486,
+      "loss": 2.7807,
+      "step": 1649
+    },
+    {
+      "epoch": 0.13316116536195627,
+      "grad_norm": 0.8308143615722656,
+      "learning_rate": 0.00019702107955409863,
+      "loss": 2.7698,
+      "step": 1650
+    },
+    {
+      "epoch": 0.13324186909853927,
+      "grad_norm": 0.8215764760971069,
+      "learning_rate": 0.00019701725377526349,
+      "loss": 2.8263,
+      "step": 1651
+    },
+    {
+      "epoch": 0.13332257283512228,
+      "grad_norm": 0.8780504465103149,
+      "learning_rate": 0.00019701342557850476,
+      "loss": 2.8032,
+      "step": 1652
+    },
+    {
+      "epoch": 0.13340327657170528,
+      "grad_norm": 0.8125136494636536,
+      "learning_rate": 0.0001970095949639179,
+      "loss": 2.8317,
+      "step": 1653
+    },
+    {
+      "epoch": 0.13348398030828829,
+      "grad_norm": 0.8170902132987976,
+      "learning_rate": 0.00019700576193159831,
+      "loss": 2.7528,
+      "step": 1654
+    },
+    {
+      "epoch": 0.1335646840448713,
+      "grad_norm": 0.8318637013435364,
+      "learning_rate": 0.00019700192648164157,
+      "loss": 2.7963,
+      "step": 1655
+    },
+    {
+      "epoch": 0.1336453877814543,
+      "grad_norm": 0.8445270657539368,
+      "learning_rate": 0.00019699808861414327,
+      "loss": 2.772,
+      "step": 1656
+    },
+    {
+      "epoch": 0.1337260915180373,
+      "grad_norm": 0.7908959984779358,
+      "learning_rate": 0.00019699424832919906,
+      "loss": 2.7528,
+      "step": 1657
+    },
+    {
+      "epoch": 0.13380679525462028,
+      "grad_norm": 0.8153900504112244,
+      "learning_rate": 0.00019699040562690462,
+      "loss": 2.7643,
+      "step": 1658
+    },
+    {
+      "epoch": 0.13388749899120328,
+      "grad_norm": 0.86302250623703,
+      "learning_rate": 0.0001969865605073557,
+      "loss": 2.8037,
+      "step": 1659
+    },
+    {
+      "epoch": 0.13396820272778628,
+      "grad_norm": 0.8373419046401978,
+      "learning_rate": 0.0001969827129706482,
+      "loss": 2.7647,
+      "step": 1660
+    },
+    {
+      "epoch": 0.1340489064643693,
+      "grad_norm": 0.8166481852531433,
+      "learning_rate": 0.00019697886301687798,
+      "loss": 2.8333,
+      "step": 1661
+    },
+    {
+      "epoch": 0.1341296102009523,
+      "grad_norm": 0.7807812094688416,
+      "learning_rate": 0.00019697501064614098,
+      "loss": 2.7495,
+      "step": 1662
+    },
+    {
+      "epoch": 0.1342103139375353,
+      "grad_norm": 0.8375338315963745,
+      "learning_rate": 0.00019697115585853324,
+      "loss": 2.7518,
+      "step": 1663
+    },
+    {
+      "epoch": 0.1342910176741183,
+      "grad_norm": 0.7392182350158691,
+      "learning_rate": 0.00019696729865415077,
+      "loss": 2.758,
+      "step": 1664
+    },
+    {
+      "epoch": 0.1343717214107013,
+      "grad_norm": 0.8041971921920776,
+      "learning_rate": 0.00019696343903308978,
+      "loss": 2.7485,
+      "step": 1665
+    },
+    {
+      "epoch": 0.1344524251472843,
+      "grad_norm": 0.789310097694397,
+      "learning_rate": 0.00019695957699544643,
+      "loss": 2.8179,
+      "step": 1666
+    },
+    {
+      "epoch": 0.13453312888386731,
+      "grad_norm": 0.7643609642982483,
+      "learning_rate": 0.00019695571254131693,
+      "loss": 2.7791,
+      "step": 1667
+    },
+    {
+      "epoch": 0.13461383262045032,
+      "grad_norm": 0.8284661769866943,
+      "learning_rate": 0.00019695184567079766,
+      "loss": 2.717,
+      "step": 1668
+    },
+    {
+      "epoch": 0.13469453635703332,
+      "grad_norm": 0.7620903253555298,
+      "learning_rate": 0.00019694797638398494,
+      "loss": 2.7808,
+      "step": 1669
+    },
+    {
+      "epoch": 0.13477524009361633,
+      "grad_norm": 0.9123913645744324,
+      "learning_rate": 0.00019694410468097524,
+      "loss": 2.7648,
+      "step": 1670
+    },
+    {
+      "epoch": 0.13485594383019933,
+      "grad_norm": 0.735518217086792,
+      "learning_rate": 0.000196940230561865,
+      "loss": 2.7653,
+      "step": 1671
+    },
+    {
+      "epoch": 0.13493664756678234,
+      "grad_norm": 0.8363413214683533,
+      "learning_rate": 0.00019693635402675085,
+      "loss": 2.766,
+      "step": 1672
+    },
+    {
+      "epoch": 0.13501735130336534,
+      "grad_norm": 0.8206491470336914,
+      "learning_rate": 0.00019693247507572936,
+      "loss": 2.7829,
+      "step": 1673
+    },
+    {
+      "epoch": 0.13509805503994834,
+      "grad_norm": 0.7726099491119385,
+      "learning_rate": 0.0001969285937088972,
+      "loss": 2.7381,
+      "step": 1674
+    },
+    {
+      "epoch": 0.13517875877653135,
+      "grad_norm": 0.8970316052436829,
+      "learning_rate": 0.0001969247099263511,
+      "loss": 2.7836,
+      "step": 1675
+    },
+    {
+      "epoch": 0.13525946251311435,
+      "grad_norm": 0.7966172099113464,
+      "learning_rate": 0.00019692082372818788,
+      "loss": 2.7135,
+      "step": 1676
+    },
+    {
+      "epoch": 0.13534016624969736,
+      "grad_norm": 0.8583024740219116,
+      "learning_rate": 0.00019691693511450438,
+      "loss": 2.7908,
+      "step": 1677
+    },
+    {
+      "epoch": 0.13542086998628036,
+      "grad_norm": 0.9430457353591919,
+      "learning_rate": 0.0001969130440853975,
+      "loss": 2.7311,
+      "step": 1678
+    },
+    {
+      "epoch": 0.13550157372286337,
+      "grad_norm": 0.8066009879112244,
+      "learning_rate": 0.00019690915064096424,
+      "loss": 2.7039,
+      "step": 1679
+    },
+    {
+      "epoch": 0.13558227745944637,
+      "grad_norm": 1.0169655084609985,
+      "learning_rate": 0.0001969052547813016,
+      "loss": 2.7832,
+      "step": 1680
+    },
+    {
+      "epoch": 0.13566298119602938,
+      "grad_norm": 0.8606080412864685,
+      "learning_rate": 0.00019690135650650672,
+      "loss": 2.751,
+      "step": 1681
+    },
+    {
+      "epoch": 0.13574368493261238,
+      "grad_norm": 0.8625333905220032,
+      "learning_rate": 0.00019689745581667674,
+      "loss": 2.761,
+      "step": 1682
+    },
+    {
+      "epoch": 0.13582438866919538,
+      "grad_norm": 0.9304285645484924,
+      "learning_rate": 0.00019689355271190886,
+      "loss": 2.7566,
+      "step": 1683
+    },
+    {
+      "epoch": 0.1359050924057784,
+      "grad_norm": 0.793397068977356,
+      "learning_rate": 0.00019688964719230035,
+      "loss": 2.7648,
+      "step": 1684
+    },
+    {
+      "epoch": 0.1359857961423614,
+      "grad_norm": 0.8496749401092529,
+      "learning_rate": 0.00019688573925794858,
+      "loss": 2.7461,
+      "step": 1685
+    },
+    {
+      "epoch": 0.1360664998789444,
+      "grad_norm": 0.7807914018630981,
+      "learning_rate": 0.0001968818289089509,
+      "loss": 2.8266,
+      "step": 1686
+    },
+    {
+      "epoch": 0.1361472036155274,
+      "grad_norm": 0.8186607956886292,
+      "learning_rate": 0.0001968779161454048,
+      "loss": 2.8447,
+      "step": 1687
+    },
+    {
+      "epoch": 0.1362279073521104,
+      "grad_norm": 0.8007118701934814,
+      "learning_rate": 0.0001968740009674078,
+      "loss": 2.7888,
+      "step": 1688
+    },
+    {
+      "epoch": 0.1363086110886934,
+      "grad_norm": 0.8735570311546326,
+      "learning_rate": 0.00019687008337505749,
+      "loss": 2.7152,
+      "step": 1689
+    },
+    {
+      "epoch": 0.13638931482527641,
+      "grad_norm": 0.8546476364135742,
+      "learning_rate": 0.00019686616336845144,
+      "loss": 2.8113,
+      "step": 1690
+    },
+    {
+      "epoch": 0.13647001856185942,
+      "grad_norm": 0.9156736135482788,
+      "learning_rate": 0.0001968622409476874,
+      "loss": 2.7561,
+      "step": 1691
+    },
+    {
+      "epoch": 0.13655072229844242,
+      "grad_norm": 0.8091925382614136,
+      "learning_rate": 0.0001968583161128631,
+      "loss": 2.7384,
+      "step": 1692
+    },
+    {
+      "epoch": 0.13663142603502543,
+      "grad_norm": 0.7871039509773254,
+      "learning_rate": 0.0001968543888640764,
+      "loss": 2.7138,
+      "step": 1693
+    },
+    {
+      "epoch": 0.13671212977160843,
+      "grad_norm": 0.9537062048912048,
+      "learning_rate": 0.00019685045920142516,
+      "loss": 2.7726,
+      "step": 1694
+    },
+    {
+      "epoch": 0.13679283350819144,
+      "grad_norm": 0.8663280010223389,
+      "learning_rate": 0.00019684652712500728,
+      "loss": 2.7509,
+      "step": 1695
+    },
+    {
+      "epoch": 0.13687353724477444,
+      "grad_norm": 0.8717214465141296,
+      "learning_rate": 0.0001968425926349208,
+      "loss": 2.791,
+      "step": 1696
+    },
+    {
+      "epoch": 0.13695424098135744,
+      "grad_norm": 0.8942584991455078,
+      "learning_rate": 0.00019683865573126374,
+      "loss": 2.77,
+      "step": 1697
+    },
+    {
+      "epoch": 0.13703494471794045,
+      "grad_norm": 0.8243421316146851,
+      "learning_rate": 0.00019683471641413424,
+      "loss": 2.8063,
+      "step": 1698
+    },
+    {
+      "epoch": 0.13711564845452345,
+      "grad_norm": 0.8618699908256531,
+      "learning_rate": 0.0001968307746836305,
+      "loss": 2.6872,
+      "step": 1699
+    },
+    {
+      "epoch": 0.13719635219110646,
+      "grad_norm": 0.7931695580482483,
+      "learning_rate": 0.00019682683053985072,
+      "loss": 2.7495,
+      "step": 1700
+    },
+    {
+      "epoch": 0.13727705592768946,
+      "grad_norm": 0.7549482583999634,
+      "learning_rate": 0.00019682288398289324,
+      "loss": 2.7543,
+      "step": 1701
+    },
+    {
+      "epoch": 0.13735775966427247,
+      "grad_norm": 0.7953789234161377,
+      "learning_rate": 0.00019681893501285636,
+      "loss": 2.6895,
+      "step": 1702
+    },
+    {
+      "epoch": 0.13743846340085547,
+      "grad_norm": 0.7916574478149414,
+      "learning_rate": 0.00019681498362983857,
+      "loss": 2.819,
+      "step": 1703
+    },
+    {
+      "epoch": 0.13751916713743847,
+      "grad_norm": 0.7986735105514526,
+      "learning_rate": 0.0001968110298339383,
+      "loss": 2.8062,
+      "step": 1704
+    },
+    {
+      "epoch": 0.13759987087402148,
+      "grad_norm": 0.8601658940315247,
+      "learning_rate": 0.00019680707362525407,
+      "loss": 2.7625,
+      "step": 1705
+    },
+    {
+      "epoch": 0.13768057461060448,
+      "grad_norm": 0.8888362050056458,
+      "learning_rate": 0.00019680311500388454,
+      "loss": 2.7747,
+      "step": 1706
+    },
+    {
+      "epoch": 0.1377612783471875,
+      "grad_norm": 0.7762896418571472,
+      "learning_rate": 0.00019679915396992833,
+      "loss": 2.7959,
+      "step": 1707
+    },
+    {
+      "epoch": 0.1378419820837705,
+      "grad_norm": 0.8942253589630127,
+      "learning_rate": 0.00019679519052348416,
+      "loss": 2.7717,
+      "step": 1708
+    },
+    {
+      "epoch": 0.13792268582035347,
+      "grad_norm": 0.8388909697532654,
+      "learning_rate": 0.00019679122466465082,
+      "loss": 2.7448,
+      "step": 1709
+    },
+    {
+      "epoch": 0.13800338955693647,
+      "grad_norm": 0.8826024532318115,
+      "learning_rate": 0.00019678725639352712,
+      "loss": 2.7307,
+      "step": 1710
+    },
+    {
+      "epoch": 0.13808409329351948,
+      "grad_norm": 0.8972313404083252,
+      "learning_rate": 0.00019678328571021204,
+      "loss": 2.7619,
+      "step": 1711
+    },
+    {
+      "epoch": 0.13816479703010248,
+      "grad_norm": 0.9373044371604919,
+      "learning_rate": 0.00019677931261480444,
+      "loss": 2.7664,
+      "step": 1712
+    },
+    {
+      "epoch": 0.1382455007666855,
+      "grad_norm": 0.8060994148254395,
+      "learning_rate": 0.00019677533710740343,
+      "loss": 2.7707,
+      "step": 1713
+    },
+    {
+      "epoch": 0.1383262045032685,
+      "grad_norm": 0.8324100971221924,
+      "learning_rate": 0.000196771359188108,
+      "loss": 2.8249,
+      "step": 1714
+    },
+    {
+      "epoch": 0.1384069082398515,
+      "grad_norm": 0.879176676273346,
+      "learning_rate": 0.00019676737885701738,
+      "loss": 2.7767,
+      "step": 1715
+    },
+    {
+      "epoch": 0.1384876119764345,
+      "grad_norm": 0.8823966979980469,
+      "learning_rate": 0.0001967633961142307,
+      "loss": 2.791,
+      "step": 1716
+    },
+    {
+      "epoch": 0.1385683157130175,
+      "grad_norm": 0.8176039457321167,
+      "learning_rate": 0.00019675941095984728,
+      "loss": 2.8225,
+      "step": 1717
+    },
+    {
+      "epoch": 0.1386490194496005,
+      "grad_norm": 0.8005076050758362,
+      "learning_rate": 0.00019675542339396635,
+      "loss": 2.8175,
+      "step": 1718
+    },
+    {
+      "epoch": 0.1387297231861835,
+      "grad_norm": 0.800854504108429,
+      "learning_rate": 0.0001967514334166874,
+      "loss": 2.8226,
+      "step": 1719
+    },
+    {
+      "epoch": 0.13881042692276652,
+      "grad_norm": 0.7941261529922485,
+      "learning_rate": 0.00019674744102810978,
+      "loss": 2.7488,
+      "step": 1720
+    },
+    {
+      "epoch": 0.13889113065934952,
+      "grad_norm": 0.7955947518348694,
+      "learning_rate": 0.00019674344622833302,
+      "loss": 2.7749,
+      "step": 1721
+    },
+    {
+      "epoch": 0.13897183439593253,
+      "grad_norm": 0.8353856205940247,
+      "learning_rate": 0.00019673944901745674,
+      "loss": 2.7982,
+      "step": 1722
+    },
+    {
+      "epoch": 0.13905253813251553,
+      "grad_norm": 0.8711503744125366,
+      "learning_rate": 0.00019673544939558047,
+      "loss": 2.8007,
+      "step": 1723
+    },
+    {
+      "epoch": 0.13913324186909853,
+      "grad_norm": 0.8525274991989136,
+      "learning_rate": 0.00019673144736280396,
+      "loss": 2.7423,
+      "step": 1724
+    },
+    {
+      "epoch": 0.13921394560568154,
+      "grad_norm": 0.8143991231918335,
+      "learning_rate": 0.0001967274429192269,
+      "loss": 2.7752,
+      "step": 1725
+    },
+    {
+      "epoch": 0.13929464934226454,
+      "grad_norm": 0.8508228063583374,
+      "learning_rate": 0.00019672343606494912,
+      "loss": 2.7422,
+      "step": 1726
+    },
+    {
+      "epoch": 0.13937535307884755,
+      "grad_norm": 0.8320932984352112,
+      "learning_rate": 0.0001967194268000705,
+      "loss": 2.7598,
+      "step": 1727
+    },
+    {
+      "epoch": 0.13945605681543055,
+      "grad_norm": 0.8233908414840698,
+      "learning_rate": 0.00019671541512469092,
+      "loss": 2.7834,
+      "step": 1728
+    },
+    {
+      "epoch": 0.13953676055201356,
+      "grad_norm": 0.8097162246704102,
+      "learning_rate": 0.00019671140103891038,
+      "loss": 2.7856,
+      "step": 1729
+    },
+    {
+      "epoch": 0.13961746428859656,
+      "grad_norm": 0.9043141007423401,
+      "learning_rate": 0.0001967073845428289,
+      "loss": 2.8047,
+      "step": 1730
+    },
+    {
+      "epoch": 0.13969816802517956,
+      "grad_norm": 0.9118517637252808,
+      "learning_rate": 0.00019670336563654662,
+      "loss": 2.789,
+      "step": 1731
+    },
+    {
+      "epoch": 0.13977887176176257,
+      "grad_norm": 0.8016074895858765,
+      "learning_rate": 0.00019669934432016368,
+      "loss": 2.7506,
+      "step": 1732
+    },
+    {
+      "epoch": 0.13985957549834557,
+      "grad_norm": 0.8376848697662354,
+      "learning_rate": 0.0001966953205937803,
+      "loss": 2.7832,
+      "step": 1733
+    },
+    {
+      "epoch": 0.13994027923492858,
+      "grad_norm": 0.8511834144592285,
+      "learning_rate": 0.0001966912944574968,
+      "loss": 2.7564,
+      "step": 1734
+    },
+    {
+      "epoch": 0.14002098297151158,
+      "grad_norm": 0.7796351909637451,
+      "learning_rate": 0.00019668726591141344,
+      "loss": 2.7489,
+      "step": 1735
+    },
+    {
+      "epoch": 0.14010168670809459,
+      "grad_norm": 0.8204767107963562,
+      "learning_rate": 0.00019668323495563068,
+      "loss": 2.7634,
+      "step": 1736
+    },
+    {
+      "epoch": 0.1401823904446776,
+      "grad_norm": 0.9049975872039795,
+      "learning_rate": 0.000196679201590249,
+      "loss": 2.7863,
+      "step": 1737
+    },
+    {
+      "epoch": 0.1402630941812606,
+      "grad_norm": 0.7473673224449158,
+      "learning_rate": 0.0001966751658153689,
+      "loss": 2.7557,
+      "step": 1738
+    },
+    {
+      "epoch": 0.1403437979178436,
+      "grad_norm": 0.7765525579452515,
+      "learning_rate": 0.0001966711276310909,
+      "loss": 2.7865,
+      "step": 1739
+    },
+    {
+      "epoch": 0.1404245016544266,
+      "grad_norm": 0.8766517043113708,
+      "learning_rate": 0.00019666708703751576,
+      "loss": 2.7873,
+      "step": 1740
+    },
+    {
+      "epoch": 0.1405052053910096,
+      "grad_norm": 0.8351505994796753,
+      "learning_rate": 0.00019666304403474408,
+      "loss": 2.7355,
+      "step": 1741
+    },
+    {
+      "epoch": 0.1405859091275926,
+      "grad_norm": 0.7612324953079224,
+      "learning_rate": 0.00019665899862287667,
+      "loss": 2.7608,
+      "step": 1742
+    },
+    {
+      "epoch": 0.14066661286417562,
+      "grad_norm": 0.894249439239502,
+      "learning_rate": 0.00019665495080201434,
+      "loss": 2.7469,
+      "step": 1743
+    },
+    {
+      "epoch": 0.14074731660075862,
+      "grad_norm": 0.8528907895088196,
+      "learning_rate": 0.00019665090057225803,
+      "loss": 2.773,
+      "step": 1744
+    },
+    {
+      "epoch": 0.14082802033734163,
+      "grad_norm": 0.7718498706817627,
+      "learning_rate": 0.00019664684793370855,
+      "loss": 2.8045,
+      "step": 1745
+    },
+    {
+      "epoch": 0.14090872407392463,
+      "grad_norm": 0.8013718128204346,
+      "learning_rate": 0.00019664279288646706,
+      "loss": 2.7665,
+      "step": 1746
+    },
+    {
+      "epoch": 0.14098942781050763,
+      "grad_norm": 0.828803539276123,
+      "learning_rate": 0.00019663873543063448,
+      "loss": 2.7846,
+      "step": 1747
+    },
+    {
+      "epoch": 0.14107013154709064,
+      "grad_norm": 0.8349393606185913,
+      "learning_rate": 0.00019663467556631204,
+      "loss": 2.7405,
+      "step": 1748
+    },
+    {
+      "epoch": 0.14115083528367364,
+      "grad_norm": 0.8273345232009888,
+      "learning_rate": 0.00019663061329360085,
+      "loss": 2.7578,
+      "step": 1749
+    },
+    {
+      "epoch": 0.14123153902025665,
+      "grad_norm": 0.7989444136619568,
+      "learning_rate": 0.0001966265486126022,
+      "loss": 2.739,
+      "step": 1750
+    },
+    {
+      "epoch": 0.14131224275683965,
+      "grad_norm": 0.8690519332885742,
+      "learning_rate": 0.00019662248152341736,
+      "loss": 2.7566,
+      "step": 1751
+    },
+    {
+      "epoch": 0.14139294649342266,
+      "grad_norm": 0.8453623056411743,
+      "learning_rate": 0.0001966184120261477,
+      "loss": 2.8572,
+      "step": 1752
+    },
+    {
+      "epoch": 0.14147365023000566,
+      "grad_norm": 0.8396254777908325,
+      "learning_rate": 0.00019661434012089468,
+      "loss": 2.786,
+      "step": 1753
+    },
+    {
+      "epoch": 0.14155435396658866,
+      "grad_norm": 0.7643738389015198,
+      "learning_rate": 0.00019661026580775973,
+      "loss": 2.8193,
+      "step": 1754
+    },
+    {
+      "epoch": 0.14163505770317167,
+      "grad_norm": 0.8124154806137085,
+      "learning_rate": 0.00019660618908684443,
+      "loss": 2.7754,
+      "step": 1755
+    },
+    {
+      "epoch": 0.14171576143975467,
+      "grad_norm": 0.8620683550834656,
+      "learning_rate": 0.00019660210995825036,
+      "loss": 2.7827,
+      "step": 1756
+    },
+    {
+      "epoch": 0.14179646517633768,
+      "grad_norm": 0.8241196274757385,
+      "learning_rate": 0.0001965980284220792,
+      "loss": 2.7573,
+      "step": 1757
+    },
+    {
+      "epoch": 0.14187716891292068,
+      "grad_norm": 0.8264089822769165,
+      "learning_rate": 0.00019659394447843262,
+      "loss": 2.8214,
+      "step": 1758
+    },
+    {
+      "epoch": 0.14195787264950369,
+      "grad_norm": 0.9129722118377686,
+      "learning_rate": 0.00019658985812741247,
+      "loss": 2.7962,
+      "step": 1759
+    },
+    {
+      "epoch": 0.14203857638608666,
+      "grad_norm": 0.7976365089416504,
+      "learning_rate": 0.00019658576936912057,
+      "loss": 2.7534,
+      "step": 1760
+    },
+    {
+      "epoch": 0.14211928012266967,
+      "grad_norm": 0.7587228417396545,
+      "learning_rate": 0.00019658167820365882,
+      "loss": 2.7083,
+      "step": 1761
+    },
+    {
+      "epoch": 0.14219998385925267,
+      "grad_norm": 0.757882833480835,
+      "learning_rate": 0.00019657758463112918,
+      "loss": 2.7135,
+      "step": 1762
+    },
+    {
+      "epoch": 0.14228068759583568,
+      "grad_norm": 0.8541501760482788,
+      "learning_rate": 0.00019657348865163369,
+      "loss": 2.7833,
+      "step": 1763
+    },
+    {
+      "epoch": 0.14236139133241868,
+      "grad_norm": 0.7708966135978699,
+      "learning_rate": 0.00019656939026527442,
+      "loss": 2.7128,
+      "step": 1764
+    },
+    {
+      "epoch": 0.14244209506900168,
+      "grad_norm": 0.8733000159263611,
+      "learning_rate": 0.00019656528947215347,
+      "loss": 2.7597,
+      "step": 1765
+    },
+    {
+      "epoch": 0.1425227988055847,
+      "grad_norm": 0.7913360595703125,
+      "learning_rate": 0.0001965611862723731,
+      "loss": 2.7681,
+      "step": 1766
+    },
+    {
+      "epoch": 0.1426035025421677,
+      "grad_norm": 0.8692380785942078,
+      "learning_rate": 0.00019655708066603555,
+      "loss": 2.7587,
+      "step": 1767
+    },
+    {
+      "epoch": 0.1426842062787507,
+      "grad_norm": 0.8231006860733032,
+      "learning_rate": 0.00019655297265324317,
+      "loss": 2.772,
+      "step": 1768
+    },
+    {
+      "epoch": 0.1427649100153337,
+      "grad_norm": 0.7373722791671753,
+      "learning_rate": 0.0001965488622340983,
+      "loss": 2.7875,
+      "step": 1769
+    },
+    {
+      "epoch": 0.1428456137519167,
+      "grad_norm": 0.8614751696586609,
+      "learning_rate": 0.0001965447494087034,
+      "loss": 2.7962,
+      "step": 1770
+    },
+    {
+      "epoch": 0.1429263174884997,
+      "grad_norm": 0.8336494565010071,
+      "learning_rate": 0.000196540634177161,
+      "loss": 2.7072,
+      "step": 1771
+    },
+    {
+      "epoch": 0.14300702122508271,
+      "grad_norm": 0.844292163848877,
+      "learning_rate": 0.00019653651653957362,
+      "loss": 2.8043,
+      "step": 1772
+    },
+    {
+      "epoch": 0.14308772496166572,
+      "grad_norm": 0.7366824150085449,
+      "learning_rate": 0.0001965323964960439,
+      "loss": 2.7296,
+      "step": 1773
+    },
+    {
+      "epoch": 0.14316842869824872,
+      "grad_norm": 0.75767982006073,
+      "learning_rate": 0.0001965282740466745,
+      "loss": 2.7946,
+      "step": 1774
+    },
+    {
+      "epoch": 0.14324913243483173,
+      "grad_norm": 0.8361382484436035,
+      "learning_rate": 0.00019652414919156823,
+      "loss": 2.7232,
+      "step": 1775
+    },
+    {
+      "epoch": 0.14332983617141473,
+      "grad_norm": 0.8473719358444214,
+      "learning_rate": 0.0001965200219308278,
+      "loss": 2.774,
+      "step": 1776
+    },
+    {
+      "epoch": 0.14341053990799774,
+      "grad_norm": 0.7446423172950745,
+      "learning_rate": 0.00019651589226455613,
+      "loss": 2.7439,
+      "step": 1777
+    },
+    {
+      "epoch": 0.14349124364458074,
+      "grad_norm": 0.8332851529121399,
+      "learning_rate": 0.00019651176019285616,
+      "loss": 2.7891,
+      "step": 1778
+    },
+    {
+      "epoch": 0.14357194738116374,
+      "grad_norm": 0.885313868522644,
+      "learning_rate": 0.0001965076257158308,
+      "loss": 2.7677,
+      "step": 1779
+    },
+    {
+      "epoch": 0.14365265111774675,
+      "grad_norm": 0.8506965637207031,
+      "learning_rate": 0.00019650348883358315,
+      "loss": 2.8112,
+      "step": 1780
+    },
+    {
+      "epoch": 0.14373335485432975,
+      "grad_norm": 0.8415799736976624,
+      "learning_rate": 0.0001964993495462163,
+      "loss": 2.8242,
+      "step": 1781
+    },
+    {
+      "epoch": 0.14381405859091276,
+      "grad_norm": 0.8501513004302979,
+      "learning_rate": 0.00019649520785383338,
+      "loss": 2.8352,
+      "step": 1782
+    },
+    {
+      "epoch": 0.14389476232749576,
+      "grad_norm": 0.7839778065681458,
+      "learning_rate": 0.00019649106375653767,
+      "loss": 2.7194,
+      "step": 1783
+    },
+    {
+      "epoch": 0.14397546606407877,
+      "grad_norm": 0.8013346195220947,
+      "learning_rate": 0.00019648691725443243,
+      "loss": 2.7665,
+      "step": 1784
+    },
+    {
+      "epoch": 0.14405616980066177,
+      "grad_norm": 1.0338317155838013,
+      "learning_rate": 0.00019648276834762095,
+      "loss": 2.8599,
+      "step": 1785
+    },
+    {
+      "epoch": 0.14413687353724478,
+      "grad_norm": 0.898417592048645,
+      "learning_rate": 0.0001964786170362067,
+      "loss": 2.7192,
+      "step": 1786
+    },
+    {
+      "epoch": 0.14421757727382778,
+      "grad_norm": 0.8876320123672485,
+      "learning_rate": 0.00019647446332029313,
+      "loss": 2.7722,
+      "step": 1787
+    },
+    {
+      "epoch": 0.14429828101041078,
+      "grad_norm": 0.819461464881897,
+      "learning_rate": 0.00019647030719998373,
+      "loss": 2.7698,
+      "step": 1788
+    },
+    {
+      "epoch": 0.1443789847469938,
+      "grad_norm": 0.848380446434021,
+      "learning_rate": 0.0001964661486753821,
+      "loss": 2.7894,
+      "step": 1789
+    },
+    {
+      "epoch": 0.1444596884835768,
+      "grad_norm": 0.8343753814697266,
+      "learning_rate": 0.0001964619877465919,
+      "loss": 2.699,
+      "step": 1790
+    },
+    {
+      "epoch": 0.1445403922201598,
+      "grad_norm": 0.8718340396881104,
+      "learning_rate": 0.0001964578244137168,
+      "loss": 2.7313,
+      "step": 1791
+    },
+    {
+      "epoch": 0.1446210959567428,
+      "grad_norm": 0.866122841835022,
+      "learning_rate": 0.00019645365867686056,
+      "loss": 2.7112,
+      "step": 1792
+    },
+    {
+      "epoch": 0.1447017996933258,
+      "grad_norm": 0.8351789712905884,
+      "learning_rate": 0.000196449490536127,
+      "loss": 2.7765,
+      "step": 1793
+    },
+    {
+      "epoch": 0.1447825034299088,
+      "grad_norm": 0.8628408312797546,
+      "learning_rate": 0.00019644531999162004,
+      "loss": 2.7375,
+      "step": 1794
+    },
+    {
+      "epoch": 0.14486320716649181,
+      "grad_norm": 0.8414484858512878,
+      "learning_rate": 0.00019644114704344358,
+      "loss": 2.7502,
+      "step": 1795
+    },
+    {
+      "epoch": 0.14494391090307482,
+      "grad_norm": 0.9092586636543274,
+      "learning_rate": 0.00019643697169170166,
+      "loss": 2.7714,
+      "step": 1796
+    },
+    {
+      "epoch": 0.14502461463965782,
+      "grad_norm": 0.8458060622215271,
+      "learning_rate": 0.0001964327939364983,
+      "loss": 2.8376,
+      "step": 1797
+    },
+    {
+      "epoch": 0.14510531837624083,
+      "grad_norm": 0.8150759935379028,
+      "learning_rate": 0.00019642861377793764,
+      "loss": 2.7147,
+      "step": 1798
+    },
+    {
+      "epoch": 0.14518602211282383,
+      "grad_norm": 0.9008790850639343,
+      "learning_rate": 0.00019642443121612387,
+      "loss": 2.7786,
+      "step": 1799
+    },
+    {
+      "epoch": 0.14526672584940684,
+      "grad_norm": 0.848671555519104,
+      "learning_rate": 0.00019642024625116117,
+      "loss": 2.7813,
+      "step": 1800
+    },
+    {
+      "epoch": 0.14534742958598984,
+      "grad_norm": 0.8035007119178772,
+      "learning_rate": 0.00019641605888315393,
+      "loss": 2.7988,
+      "step": 1801
+    },
+    {
+      "epoch": 0.14542813332257284,
+      "grad_norm": 0.8210242390632629,
+      "learning_rate": 0.00019641186911220645,
+      "loss": 2.8451,
+      "step": 1802
+    },
+    {
+      "epoch": 0.14550883705915585,
+      "grad_norm": 0.8852066397666931,
+      "learning_rate": 0.00019640767693842318,
+      "loss": 2.7492,
+      "step": 1803
+    },
+    {
+      "epoch": 0.14558954079573885,
+      "grad_norm": 0.8421196937561035,
+      "learning_rate": 0.0001964034823619086,
+      "loss": 2.759,
+      "step": 1804
+    },
+    {
+      "epoch": 0.14567024453232186,
+      "grad_norm": 0.8166298866271973,
+      "learning_rate": 0.00019639928538276724,
+      "loss": 2.7942,
+      "step": 1805
+    },
+    {
+      "epoch": 0.14575094826890486,
+      "grad_norm": 0.8502809405326843,
+      "learning_rate": 0.00019639508600110368,
+      "loss": 2.7829,
+      "step": 1806
+    },
+    {
+      "epoch": 0.14583165200548787,
+      "grad_norm": 0.8371078372001648,
+      "learning_rate": 0.0001963908842170226,
+      "loss": 2.7168,
+      "step": 1807
+    },
+    {
+      "epoch": 0.14591235574207087,
+      "grad_norm": 0.8148230910301208,
+      "learning_rate": 0.0001963866800306287,
+      "loss": 2.7706,
+      "step": 1808
+    },
+    {
+      "epoch": 0.14599305947865387,
+      "grad_norm": 0.8984564542770386,
+      "learning_rate": 0.0001963824734420268,
+      "loss": 2.7761,
+      "step": 1809
+    },
+    {
+      "epoch": 0.14607376321523688,
+      "grad_norm": 0.9357183575630188,
+      "learning_rate": 0.00019637826445132172,
+      "loss": 2.7738,
+      "step": 1810
+    },
+    {
+      "epoch": 0.14615446695181986,
+      "grad_norm": 0.8545449376106262,
+      "learning_rate": 0.00019637405305861834,
+      "loss": 2.772,
+      "step": 1811
+    },
+    {
+      "epoch": 0.14623517068840286,
+      "grad_norm": 1.1674948930740356,
+      "learning_rate": 0.00019636983926402165,
+      "loss": 2.8988,
+      "step": 1812
+    },
+    {
+      "epoch": 0.14631587442498586,
+      "grad_norm": 0.7875451445579529,
+      "learning_rate": 0.00019636562306763665,
+      "loss": 2.7053,
+      "step": 1813
+    },
+    {
+      "epoch": 0.14639657816156887,
+      "grad_norm": 0.8980962038040161,
+      "learning_rate": 0.0001963614044695684,
+      "loss": 2.7731,
+      "step": 1814
+    },
+    {
+      "epoch": 0.14647728189815187,
+      "grad_norm": 0.8403381705284119,
+      "learning_rate": 0.00019635718346992207,
+      "loss": 2.8555,
+      "step": 1815
+    },
+    {
+      "epoch": 0.14655798563473488,
+      "grad_norm": 0.8736433982849121,
+      "learning_rate": 0.00019635296006880284,
+      "loss": 2.7918,
+      "step": 1816
+    },
+    {
+      "epoch": 0.14663868937131788,
+      "grad_norm": 0.8604151606559753,
+      "learning_rate": 0.000196348734266316,
+      "loss": 2.7493,
+      "step": 1817
+    },
+    {
+      "epoch": 0.1467193931079009,
+      "grad_norm": 0.8329424262046814,
+      "learning_rate": 0.00019634450606256681,
+      "loss": 2.7348,
+      "step": 1818
+    },
+    {
+      "epoch": 0.1468000968444839,
+      "grad_norm": 0.9835913181304932,
+      "learning_rate": 0.0001963402754576607,
+      "loss": 2.7651,
+      "step": 1819
+    },
+    {
+      "epoch": 0.1468808005810669,
+      "grad_norm": 0.7968378067016602,
+      "learning_rate": 0.0001963360424517031,
+      "loss": 2.7672,
+      "step": 1820
+    },
+    {
+      "epoch": 0.1469615043176499,
+      "grad_norm": 0.8012512922286987,
+      "learning_rate": 0.00019633180704479948,
+      "loss": 2.8022,
+      "step": 1821
+    },
+    {
+      "epoch": 0.1470422080542329,
+      "grad_norm": 0.7656376957893372,
+      "learning_rate": 0.0001963275692370554,
+      "loss": 2.7561,
+      "step": 1822
+    },
+    {
+      "epoch": 0.1471229117908159,
+      "grad_norm": 0.8030453324317932,
+      "learning_rate": 0.00019632332902857656,
+      "loss": 2.8048,
+      "step": 1823
+    },
+    {
+      "epoch": 0.1472036155273989,
+      "grad_norm": 0.8050903677940369,
+      "learning_rate": 0.0001963190864194685,
+      "loss": 2.7846,
+      "step": 1824
+    },
+    {
+      "epoch": 0.14728431926398192,
+      "grad_norm": 0.8001886606216431,
+      "learning_rate": 0.00019631484140983705,
+      "loss": 2.7382,
+      "step": 1825
+    },
+    {
+      "epoch": 0.14736502300056492,
+      "grad_norm": 0.8589862585067749,
+      "learning_rate": 0.00019631059399978796,
+      "loss": 2.8376,
+      "step": 1826
+    },
+    {
+      "epoch": 0.14744572673714793,
+      "grad_norm": 0.86325603723526,
+      "learning_rate": 0.00019630634418942714,
+      "loss": 2.7643,
+      "step": 1827
+    },
+    {
+      "epoch": 0.14752643047373093,
+      "grad_norm": 0.7893280386924744,
+      "learning_rate": 0.00019630209197886046,
+      "loss": 2.713,
+      "step": 1828
+    },
+    {
+      "epoch": 0.14760713421031393,
+      "grad_norm": 0.8890528082847595,
+      "learning_rate": 0.00019629783736819394,
+      "loss": 2.7435,
+      "step": 1829
+    },
+    {
+      "epoch": 0.14768783794689694,
+      "grad_norm": 0.794924795627594,
+      "learning_rate": 0.00019629358035753357,
+      "loss": 2.7703,
+      "step": 1830
+    },
+    {
+      "epoch": 0.14776854168347994,
+      "grad_norm": 0.7712973952293396,
+      "learning_rate": 0.00019628932094698545,
+      "loss": 2.7487,
+      "step": 1831
+    },
+    {
+      "epoch": 0.14784924542006295,
+      "grad_norm": 0.7810670137405396,
+      "learning_rate": 0.00019628505913665576,
+      "loss": 2.7687,
+      "step": 1832
+    },
+    {
+      "epoch": 0.14792994915664595,
+      "grad_norm": 0.8331059813499451,
+      "learning_rate": 0.0001962807949266507,
+      "loss": 2.7166,
+      "step": 1833
+    },
+    {
+      "epoch": 0.14801065289322896,
+      "grad_norm": 0.8983452916145325,
+      "learning_rate": 0.00019627652831707656,
+      "loss": 2.8096,
+      "step": 1834
+    },
+    {
+      "epoch": 0.14809135662981196,
+      "grad_norm": 0.8387179374694824,
+      "learning_rate": 0.00019627225930803963,
+      "loss": 2.8252,
+      "step": 1835
+    },
+    {
+      "epoch": 0.14817206036639496,
+      "grad_norm": 0.8619294762611389,
+      "learning_rate": 0.0001962679878996464,
+      "loss": 2.7623,
+      "step": 1836
+    },
+    {
+      "epoch": 0.14825276410297797,
+      "grad_norm": 0.8195026516914368,
+      "learning_rate": 0.0001962637140920032,
+      "loss": 2.7295,
+      "step": 1837
+    },
+    {
+      "epoch": 0.14833346783956097,
+      "grad_norm": 0.806216835975647,
+      "learning_rate": 0.00019625943788521664,
+      "loss": 2.7184,
+      "step": 1838
+    },
+    {
+      "epoch": 0.14841417157614398,
+      "grad_norm": 0.7758379578590393,
+      "learning_rate": 0.00019625515927939327,
+      "loss": 2.7675,
+      "step": 1839
+    },
+    {
+      "epoch": 0.14849487531272698,
+      "grad_norm": 0.7617168426513672,
+      "learning_rate": 0.0001962508782746397,
+      "loss": 2.8041,
+      "step": 1840
+    },
+    {
+      "epoch": 0.14857557904930999,
+      "grad_norm": 0.9630066156387329,
+      "learning_rate": 0.00019624659487106264,
+      "loss": 2.814,
+      "step": 1841
+    },
+    {
+      "epoch": 0.148656282785893,
+      "grad_norm": 0.7656112313270569,
+      "learning_rate": 0.00019624230906876888,
+      "loss": 2.7564,
+      "step": 1842
+    },
+    {
+      "epoch": 0.148736986522476,
+      "grad_norm": 0.9394779801368713,
+      "learning_rate": 0.0001962380208678652,
+      "loss": 2.7958,
+      "step": 1843
+    },
+    {
+      "epoch": 0.148817690259059,
+      "grad_norm": 0.7647004127502441,
+      "learning_rate": 0.00019623373026845842,
+      "loss": 2.72,
+      "step": 1844
+    },
+    {
+      "epoch": 0.148898393995642,
+      "grad_norm": 0.809079647064209,
+      "learning_rate": 0.00019622943727065555,
+      "loss": 2.7732,
+      "step": 1845
+    },
+    {
+      "epoch": 0.148979097732225,
+      "grad_norm": 0.8241337537765503,
+      "learning_rate": 0.00019622514187456357,
+      "loss": 2.759,
+      "step": 1846
+    },
+    {
+      "epoch": 0.149059801468808,
+      "grad_norm": 0.8979619145393372,
+      "learning_rate": 0.00019622084408028948,
+      "loss": 2.8307,
+      "step": 1847
+    },
+    {
+      "epoch": 0.14914050520539102,
+      "grad_norm": 0.8058865666389465,
+      "learning_rate": 0.00019621654388794047,
+      "loss": 2.807,
+      "step": 1848
+    },
+    {
+      "epoch": 0.14922120894197402,
+      "grad_norm": 0.81967693567276,
+      "learning_rate": 0.00019621224129762364,
+      "loss": 2.7762,
+      "step": 1849
+    },
+    {
+      "epoch": 0.14930191267855702,
+      "grad_norm": 0.7385755777359009,
+      "learning_rate": 0.0001962079363094463,
+      "loss": 2.7854,
+      "step": 1850
+    },
+    {
+      "epoch": 0.14938261641514003,
+      "grad_norm": 0.8585657477378845,
+      "learning_rate": 0.00019620362892351566,
+      "loss": 2.7781,
+      "step": 1851
+    },
+    {
+      "epoch": 0.14946332015172303,
+      "grad_norm": 0.8328986763954163,
+      "learning_rate": 0.00019619931913993912,
+      "loss": 2.8245,
+      "step": 1852
+    },
+    {
+      "epoch": 0.14954402388830604,
+      "grad_norm": 0.749727189540863,
+      "learning_rate": 0.0001961950069588241,
+      "loss": 2.8049,
+      "step": 1853
+    },
+    {
+      "epoch": 0.14962472762488904,
+      "grad_norm": 0.7886502742767334,
+      "learning_rate": 0.00019619069238027803,
+      "loss": 2.7521,
+      "step": 1854
+    },
+    {
+      "epoch": 0.14970543136147205,
+      "grad_norm": 0.816137433052063,
+      "learning_rate": 0.00019618637540440848,
+      "loss": 2.8383,
+      "step": 1855
+    },
+    {
+      "epoch": 0.14978613509805505,
+      "grad_norm": 0.80442214012146,
+      "learning_rate": 0.000196182056031323,
+      "loss": 2.7227,
+      "step": 1856
+    },
+    {
+      "epoch": 0.14986683883463806,
+      "grad_norm": 0.7605221271514893,
+      "learning_rate": 0.00019617773426112924,
+      "loss": 2.7494,
+      "step": 1857
+    },
+    {
+      "epoch": 0.14994754257122106,
+      "grad_norm": 0.8745137453079224,
+      "learning_rate": 0.00019617341009393497,
+      "loss": 2.6978,
+      "step": 1858
+    },
+    {
+      "epoch": 0.15002824630780406,
+      "grad_norm": 0.8151741623878479,
+      "learning_rate": 0.00019616908352984789,
+      "loss": 2.7817,
+      "step": 1859
+    },
+    {
+      "epoch": 0.15010895004438707,
+      "grad_norm": 0.773876428604126,
+      "learning_rate": 0.0001961647545689759,
+      "loss": 2.812,
+      "step": 1860
+    },
+    {
+      "epoch": 0.15018965378097007,
+      "grad_norm": 0.8216966390609741,
+      "learning_rate": 0.00019616042321142683,
+      "loss": 2.8181,
+      "step": 1861
+    },
+    {
+      "epoch": 0.15027035751755305,
+      "grad_norm": 0.8097409605979919,
+      "learning_rate": 0.00019615608945730862,
+      "loss": 2.8336,
+      "step": 1862
+    },
+    {
+      "epoch": 0.15035106125413605,
+      "grad_norm": 0.8085697293281555,
+      "learning_rate": 0.00019615175330672932,
+      "loss": 2.8176,
+      "step": 1863
+    },
+    {
+      "epoch": 0.15043176499071906,
+      "grad_norm": 0.7658133506774902,
+      "learning_rate": 0.00019614741475979701,
+      "loss": 2.7543,
+      "step": 1864
+    },
+    {
+      "epoch": 0.15051246872730206,
+      "grad_norm": 0.7193909883499146,
+      "learning_rate": 0.00019614307381661978,
+      "loss": 2.7475,
+      "step": 1865
+    },
+    {
+      "epoch": 0.15059317246388507,
+      "grad_norm": 0.835608959197998,
+      "learning_rate": 0.0001961387304773058,
+      "loss": 2.8017,
+      "step": 1866
+    },
+    {
+      "epoch": 0.15067387620046807,
+      "grad_norm": 0.7898489832878113,
+      "learning_rate": 0.0001961343847419634,
+      "loss": 2.7613,
+      "step": 1867
+    },
+    {
+      "epoch": 0.15075457993705108,
+      "grad_norm": 0.8031982183456421,
+      "learning_rate": 0.0001961300366107008,
+      "loss": 2.7442,
+      "step": 1868
+    },
+    {
+      "epoch": 0.15083528367363408,
+      "grad_norm": 0.8427363634109497,
+      "learning_rate": 0.00019612568608362642,
+      "loss": 2.8095,
+      "step": 1869
+    },
+    {
+      "epoch": 0.15091598741021708,
+      "grad_norm": 0.8282802700996399,
+      "learning_rate": 0.00019612133316084863,
+      "loss": 2.7216,
+      "step": 1870
+    },
+    {
+      "epoch": 0.1509966911468001,
+      "grad_norm": 0.7799758911132812,
+      "learning_rate": 0.000196116977842476,
+      "loss": 2.793,
+      "step": 1871
+    },
+    {
+      "epoch": 0.1510773948833831,
+      "grad_norm": 0.8151525259017944,
+      "learning_rate": 0.00019611262012861702,
+      "loss": 2.7641,
+      "step": 1872
+    },
+    {
+      "epoch": 0.1511580986199661,
+      "grad_norm": 0.7926812767982483,
+      "learning_rate": 0.0001961082600193803,
+      "loss": 2.7523,
+      "step": 1873
+    },
+    {
+      "epoch": 0.1512388023565491,
+      "grad_norm": 0.8737135529518127,
+      "learning_rate": 0.0001961038975148745,
+      "loss": 2.7965,
+      "step": 1874
+    },
+    {
+      "epoch": 0.1513195060931321,
+      "grad_norm": 0.7948090434074402,
+      "learning_rate": 0.00019609953261520837,
+      "loss": 2.7737,
+      "step": 1875
+    },
+    {
+      "epoch": 0.1514002098297151,
+      "grad_norm": 0.8161277770996094,
+      "learning_rate": 0.0001960951653204907,
+      "loss": 2.7423,
+      "step": 1876
+    },
+    {
+      "epoch": 0.15148091356629811,
+      "grad_norm": 0.8904973864555359,
+      "learning_rate": 0.00019609079563083026,
+      "loss": 2.7066,
+      "step": 1877
+    },
+    {
+      "epoch": 0.15156161730288112,
+      "grad_norm": 0.8107061982154846,
+      "learning_rate": 0.00019608642354633604,
+      "loss": 2.7939,
+      "step": 1878
+    },
+    {
+      "epoch": 0.15164232103946412,
+      "grad_norm": 0.8410987854003906,
+      "learning_rate": 0.00019608204906711694,
+      "loss": 2.7521,
+      "step": 1879
+    },
+    {
+      "epoch": 0.15172302477604713,
+      "grad_norm": 0.8336483836174011,
+      "learning_rate": 0.0001960776721932821,
+      "loss": 2.7613,
+      "step": 1880
+    },
+    {
+      "epoch": 0.15180372851263013,
+      "grad_norm": 0.730549156665802,
+      "learning_rate": 0.00019607329292494044,
+      "loss": 2.8019,
+      "step": 1881
+    },
+    {
+      "epoch": 0.15188443224921314,
+      "grad_norm": 0.7543070912361145,
+      "learning_rate": 0.0001960689112622012,
+      "loss": 2.6907,
+      "step": 1882
+    },
+    {
+      "epoch": 0.15196513598579614,
+      "grad_norm": 0.848414421081543,
+      "learning_rate": 0.00019606452720517359,
+      "loss": 2.7278,
+      "step": 1883
+    },
+    {
+      "epoch": 0.15204583972237914,
+      "grad_norm": 0.8331718444824219,
+      "learning_rate": 0.00019606014075396682,
+      "loss": 2.6994,
+      "step": 1884
+    },
+    {
+      "epoch": 0.15212654345896215,
+      "grad_norm": 0.9192764759063721,
+      "learning_rate": 0.00019605575190869025,
+      "loss": 2.7095,
+      "step": 1885
+    },
+    {
+      "epoch": 0.15220724719554515,
+      "grad_norm": 0.8377116322517395,
+      "learning_rate": 0.00019605136066945324,
+      "loss": 2.7925,
+      "step": 1886
+    },
+    {
+      "epoch": 0.15228795093212816,
+      "grad_norm": 0.7302869558334351,
+      "learning_rate": 0.00019604696703636525,
+      "loss": 2.7286,
+      "step": 1887
+    },
+    {
+      "epoch": 0.15236865466871116,
+      "grad_norm": 0.7972438335418701,
+      "learning_rate": 0.00019604257100953577,
+      "loss": 2.7732,
+      "step": 1888
+    },
+    {
+      "epoch": 0.15244935840529417,
+      "grad_norm": 1.0350826978683472,
+      "learning_rate": 0.00019603817258907435,
+      "loss": 2.8211,
+      "step": 1889
+    },
+    {
+      "epoch": 0.15253006214187717,
+      "grad_norm": 0.782755970954895,
+      "learning_rate": 0.00019603377177509067,
+      "loss": 2.8489,
+      "step": 1890
+    },
+    {
+      "epoch": 0.15261076587846018,
+      "grad_norm": 0.9072603583335876,
+      "learning_rate": 0.0001960293685676943,
+      "loss": 2.7764,
+      "step": 1891
+    },
+    {
+      "epoch": 0.15269146961504318,
+      "grad_norm": 0.7878704071044922,
+      "learning_rate": 0.0001960249629669951,
+      "loss": 2.7494,
+      "step": 1892
+    },
+    {
+      "epoch": 0.15277217335162618,
+      "grad_norm": 0.8770418167114258,
+      "learning_rate": 0.00019602055497310278,
+      "loss": 2.7318,
+      "step": 1893
+    },
+    {
+      "epoch": 0.1528528770882092,
+      "grad_norm": 0.8004975914955139,
+      "learning_rate": 0.00019601614458612723,
+      "loss": 2.7272,
+      "step": 1894
+    },
+    {
+      "epoch": 0.1529335808247922,
+      "grad_norm": 0.8511070013046265,
+      "learning_rate": 0.00019601173180617835,
+      "loss": 2.7876,
+      "step": 1895
+    },
+    {
+      "epoch": 0.1530142845613752,
+      "grad_norm": 0.7946128845214844,
+      "learning_rate": 0.00019600731663336617,
+      "loss": 2.7435,
+      "step": 1896
+    },
+    {
+      "epoch": 0.1530949882979582,
+      "grad_norm": 0.8155317902565002,
+      "learning_rate": 0.00019600289906780067,
+      "loss": 2.7642,
+      "step": 1897
+    },
+    {
+      "epoch": 0.1531756920345412,
+      "grad_norm": 0.8086098432540894,
+      "learning_rate": 0.000195998479109592,
+      "loss": 2.7358,
+      "step": 1898
+    },
+    {
+      "epoch": 0.1532563957711242,
+      "grad_norm": 0.8698278665542603,
+      "learning_rate": 0.00019599405675885026,
+      "loss": 2.725,
+      "step": 1899
+    },
+    {
+      "epoch": 0.15333709950770721,
+      "grad_norm": 0.8756006360054016,
+      "learning_rate": 0.00019598963201568573,
+      "loss": 2.7209,
+      "step": 1900
+    },
+    {
+      "epoch": 0.15341780324429022,
+      "grad_norm": 0.7984628081321716,
+      "learning_rate": 0.0001959852048802086,
+      "loss": 2.7685,
+      "step": 1901
+    },
+    {
+      "epoch": 0.15349850698087322,
+      "grad_norm": 0.8244056105613708,
+      "learning_rate": 0.0001959807753525293,
+      "loss": 2.7692,
+      "step": 1902
+    },
+    {
+      "epoch": 0.15357921071745623,
+      "grad_norm": 0.8577731251716614,
+      "learning_rate": 0.00019597634343275814,
+      "loss": 2.7571,
+      "step": 1903
+    },
+    {
+      "epoch": 0.15365991445403923,
+      "grad_norm": 0.8410975933074951,
+      "learning_rate": 0.00019597190912100566,
+      "loss": 2.7862,
+      "step": 1904
+    },
+    {
+      "epoch": 0.15374061819062224,
+      "grad_norm": 0.9094158411026001,
+      "learning_rate": 0.0001959674724173823,
+      "loss": 2.7655,
+      "step": 1905
+    },
+    {
+      "epoch": 0.15382132192720524,
+      "grad_norm": 0.8375208973884583,
+      "learning_rate": 0.00019596303332199868,
+      "loss": 2.8129,
+      "step": 1906
+    },
+    {
+      "epoch": 0.15390202566378824,
+      "grad_norm": 0.8335977792739868,
+      "learning_rate": 0.00019595859183496543,
+      "loss": 2.7835,
+      "step": 1907
+    },
+    {
+      "epoch": 0.15398272940037125,
+      "grad_norm": 0.7973531484603882,
+      "learning_rate": 0.0001959541479563932,
+      "loss": 2.7785,
+      "step": 1908
+    },
+    {
+      "epoch": 0.15406343313695425,
+      "grad_norm": 0.7808824181556702,
+      "learning_rate": 0.0001959497016863928,
+      "loss": 2.7862,
+      "step": 1909
+    },
+    {
+      "epoch": 0.15414413687353726,
+      "grad_norm": 0.853824257850647,
+      "learning_rate": 0.00019594525302507504,
+      "loss": 2.6721,
+      "step": 1910
+    },
+    {
+      "epoch": 0.15422484061012026,
+      "grad_norm": 0.8589324355125427,
+      "learning_rate": 0.00019594080197255073,
+      "loss": 2.7948,
+      "step": 1911
+    },
+    {
+      "epoch": 0.15430554434670327,
+      "grad_norm": 0.7951898574829102,
+      "learning_rate": 0.00019593634852893086,
+      "loss": 2.7903,
+      "step": 1912
+    },
+    {
+      "epoch": 0.15438624808328624,
+      "grad_norm": 0.8333349227905273,
+      "learning_rate": 0.0001959318926943264,
+      "loss": 2.8073,
+      "step": 1913
+    },
+    {
+      "epoch": 0.15446695181986925,
+      "grad_norm": 0.8552380800247192,
+      "learning_rate": 0.0001959274344688484,
+      "loss": 2.8199,
+      "step": 1914
+    },
+    {
+      "epoch": 0.15454765555645225,
+      "grad_norm": 0.8356214165687561,
+      "learning_rate": 0.000195922973852608,
+      "loss": 2.7985,
+      "step": 1915
+    },
+    {
+      "epoch": 0.15462835929303526,
+      "grad_norm": 0.7167248725891113,
+      "learning_rate": 0.00019591851084571634,
+      "loss": 2.6802,
+      "step": 1916
+    },
+    {
+      "epoch": 0.15470906302961826,
+      "grad_norm": 0.7980726361274719,
+      "learning_rate": 0.00019591404544828464,
+      "loss": 2.692,
+      "step": 1917
+    },
+    {
+      "epoch": 0.15478976676620126,
+      "grad_norm": 0.7766004800796509,
+      "learning_rate": 0.00019590957766042424,
+      "loss": 2.7219,
+      "step": 1918
+    },
+    {
+      "epoch": 0.15487047050278427,
+      "grad_norm": 0.828852653503418,
+      "learning_rate": 0.0001959051074822464,
+      "loss": 2.7369,
+      "step": 1919
+    },
+    {
+      "epoch": 0.15495117423936727,
+      "grad_norm": 0.7818129062652588,
+      "learning_rate": 0.0001959006349138626,
+      "loss": 2.7778,
+      "step": 1920
+    },
+    {
+      "epoch": 0.15503187797595028,
+      "grad_norm": 0.8428593873977661,
+      "learning_rate": 0.00019589615995538432,
+      "loss": 2.8257,
+      "step": 1921
+    },
+    {
+      "epoch": 0.15511258171253328,
+      "grad_norm": 0.8756616115570068,
+      "learning_rate": 0.00019589168260692307,
+      "loss": 2.7692,
+      "step": 1922
+    },
+    {
+      "epoch": 0.15519328544911629,
+      "grad_norm": 0.7802519202232361,
+      "learning_rate": 0.0001958872028685904,
+      "loss": 2.7811,
+      "step": 1923
+    },
+    {
+      "epoch": 0.1552739891856993,
+      "grad_norm": 0.7787032723426819,
+      "learning_rate": 0.00019588272074049797,
+      "loss": 2.7546,
+      "step": 1924
+    },
+    {
+      "epoch": 0.1553546929222823,
+      "grad_norm": 0.848479151725769,
+      "learning_rate": 0.0001958782362227575,
+      "loss": 2.7759,
+      "step": 1925
+    },
+    {
+      "epoch": 0.1554353966588653,
+      "grad_norm": 0.8331353664398193,
+      "learning_rate": 0.00019587374931548076,
+      "loss": 2.7881,
+      "step": 1926
+    },
+    {
+      "epoch": 0.1555161003954483,
+      "grad_norm": 0.8646424412727356,
+      "learning_rate": 0.00019586926001877958,
+      "loss": 2.8059,
+      "step": 1927
+    },
+    {
+      "epoch": 0.1555968041320313,
+      "grad_norm": 0.912253737449646,
+      "learning_rate": 0.00019586476833276584,
+      "loss": 2.7446,
+      "step": 1928
+    },
+    {
+      "epoch": 0.1556775078686143,
+      "grad_norm": 0.9256471395492554,
+      "learning_rate": 0.00019586027425755147,
+      "loss": 2.8,
+      "step": 1929
+    },
+    {
+      "epoch": 0.15575821160519732,
+      "grad_norm": 1.0984607934951782,
+      "learning_rate": 0.0001958557777932485,
+      "loss": 2.7759,
+      "step": 1930
+    },
+    {
+      "epoch": 0.15583891534178032,
+      "grad_norm": 0.8736081123352051,
+      "learning_rate": 0.00019585127893996895,
+      "loss": 2.7464,
+      "step": 1931
+    },
+    {
+      "epoch": 0.15591961907836333,
+      "grad_norm": 0.932538628578186,
+      "learning_rate": 0.00019584677769782498,
+      "loss": 2.7874,
+      "step": 1932
+    },
+    {
+      "epoch": 0.15600032281494633,
+      "grad_norm": 0.9742087125778198,
+      "learning_rate": 0.0001958422740669288,
+      "loss": 2.7727,
+      "step": 1933
+    },
+    {
+      "epoch": 0.15608102655152933,
+      "grad_norm": 0.8975874781608582,
+      "learning_rate": 0.00019583776804739256,
+      "loss": 2.7812,
+      "step": 1934
+    },
+    {
+      "epoch": 0.15616173028811234,
+      "grad_norm": 0.9380232691764832,
+      "learning_rate": 0.00019583325963932864,
+      "loss": 2.7284,
+      "step": 1935
+    },
+    {
+      "epoch": 0.15624243402469534,
+      "grad_norm": 0.8332872986793518,
+      "learning_rate": 0.00019582874884284938,
+      "loss": 2.7792,
+      "step": 1936
+    },
+    {
+      "epoch": 0.15632313776127835,
+      "grad_norm": 1.0017194747924805,
+      "learning_rate": 0.0001958242356580672,
+      "loss": 2.7187,
+      "step": 1937
+    },
+    {
+      "epoch": 0.15640384149786135,
+      "grad_norm": 0.9433515667915344,
+      "learning_rate": 0.0001958197200850946,
+      "loss": 2.8394,
+      "step": 1938
+    },
+    {
+      "epoch": 0.15648454523444436,
+      "grad_norm": 0.8781030178070068,
+      "learning_rate": 0.00019581520212404407,
+      "loss": 2.7667,
+      "step": 1939
+    },
+    {
+      "epoch": 0.15656524897102736,
+      "grad_norm": 0.895656168460846,
+      "learning_rate": 0.00019581068177502826,
+      "loss": 2.799,
+      "step": 1940
+    },
+    {
+      "epoch": 0.15664595270761036,
+      "grad_norm": 0.8336960673332214,
+      "learning_rate": 0.0001958061590381598,
+      "loss": 2.8152,
+      "step": 1941
+    },
+    {
+      "epoch": 0.15672665644419337,
+      "grad_norm": 0.9184536337852478,
+      "learning_rate": 0.00019580163391355143,
+      "loss": 2.7746,
+      "step": 1942
+    },
+    {
+      "epoch": 0.15680736018077637,
+      "grad_norm": 0.8564908504486084,
+      "learning_rate": 0.00019579710640131587,
+      "loss": 2.7674,
+      "step": 1943
+    },
+    {
+      "epoch": 0.15688806391735938,
+      "grad_norm": 0.7491608262062073,
+      "learning_rate": 0.00019579257650156605,
+      "loss": 2.7665,
+      "step": 1944
+    },
+    {
+      "epoch": 0.15696876765394238,
+      "grad_norm": 0.9165031313896179,
+      "learning_rate": 0.00019578804421441478,
+      "loss": 2.7343,
+      "step": 1945
+    },
+    {
+      "epoch": 0.15704947139052539,
+      "grad_norm": 0.8413978815078735,
+      "learning_rate": 0.00019578350953997512,
+      "loss": 2.7503,
+      "step": 1946
+    },
+    {
+      "epoch": 0.1571301751271084,
+      "grad_norm": 0.7820419073104858,
+      "learning_rate": 0.00019577897247835993,
+      "loss": 2.7535,
+      "step": 1947
+    },
+    {
+      "epoch": 0.1572108788636914,
+      "grad_norm": 0.8134996294975281,
+      "learning_rate": 0.00019577443302968246,
+      "loss": 2.7504,
+      "step": 1948
+    },
+    {
+      "epoch": 0.1572915826002744,
+      "grad_norm": 0.8201301097869873,
+      "learning_rate": 0.00019576989119405574,
+      "loss": 2.6927,
+      "step": 1949
+    },
+    {
+      "epoch": 0.1573722863368574,
+      "grad_norm": 0.8343217372894287,
+      "learning_rate": 0.00019576534697159296,
+      "loss": 2.7742,
+      "step": 1950
+    },
+    {
+      "epoch": 0.1574529900734404,
+      "grad_norm": 0.8161751627922058,
+      "learning_rate": 0.0001957608003624074,
+      "loss": 2.8236,
+      "step": 1951
+    },
+    {
+      "epoch": 0.1575336938100234,
+      "grad_norm": 0.8626808524131775,
+      "learning_rate": 0.00019575625136661242,
+      "loss": 2.7305,
+      "step": 1952
+    },
+    {
+      "epoch": 0.15761439754660642,
+      "grad_norm": 0.8238986730575562,
+      "learning_rate": 0.0001957516999843213,
+      "loss": 2.7641,
+      "step": 1953
+    },
+    {
+      "epoch": 0.15769510128318942,
+      "grad_norm": 0.7806095480918884,
+      "learning_rate": 0.00019574714621564755,
+      "loss": 2.7155,
+      "step": 1954
+    },
+    {
+      "epoch": 0.15777580501977242,
+      "grad_norm": 0.8137761950492859,
+      "learning_rate": 0.0001957425900607046,
+      "loss": 2.7529,
+      "step": 1955
+    },
+    {
+      "epoch": 0.15785650875635543,
+      "grad_norm": 0.8383988738059998,
+      "learning_rate": 0.00019573803151960606,
+      "loss": 2.7726,
+      "step": 1956
+    },
+    {
+      "epoch": 0.15793721249293843,
+      "grad_norm": 0.8734413385391235,
+      "learning_rate": 0.00019573347059246549,
+      "loss": 2.8563,
+      "step": 1957
+    },
+    {
+      "epoch": 0.15801791622952144,
+      "grad_norm": 0.8018438816070557,
+      "learning_rate": 0.0001957289072793966,
+      "loss": 2.8031,
+      "step": 1958
+    },
+    {
+      "epoch": 0.15809861996610444,
+      "grad_norm": 0.8175764083862305,
+      "learning_rate": 0.0001957243415805131,
+      "loss": 2.7824,
+      "step": 1959
+    },
+    {
+      "epoch": 0.15817932370268745,
+      "grad_norm": 0.7642164826393127,
+      "learning_rate": 0.00019571977349592878,
+      "loss": 2.7666,
+      "step": 1960
+    },
+    {
+      "epoch": 0.15826002743927045,
+      "grad_norm": 0.7584841847419739,
+      "learning_rate": 0.0001957152030257575,
+      "loss": 2.7211,
+      "step": 1961
+    },
+    {
+      "epoch": 0.15834073117585346,
+      "grad_norm": 0.822610080242157,
+      "learning_rate": 0.00019571063017011312,
+      "loss": 2.7025,
+      "step": 1962
+    },
+    {
+      "epoch": 0.15842143491243646,
+      "grad_norm": 0.7553817629814148,
+      "learning_rate": 0.00019570605492910968,
+      "loss": 2.8122,
+      "step": 1963
+    },
+    {
+      "epoch": 0.15850213864901944,
+      "grad_norm": 0.7224497199058533,
+      "learning_rate": 0.0001957014773028612,
+      "loss": 2.7613,
+      "step": 1964
+    },
+    {
+      "epoch": 0.15858284238560244,
+      "grad_norm": 0.8563623428344727,
+      "learning_rate": 0.00019569689729148168,
+      "loss": 2.8005,
+      "step": 1965
+    },
+    {
+      "epoch": 0.15866354612218544,
+      "grad_norm": 0.7665508389472961,
+      "learning_rate": 0.00019569231489508537,
+      "loss": 2.7387,
+      "step": 1966
+    },
+    {
+      "epoch": 0.15874424985876845,
+      "grad_norm": 0.7788479328155518,
+      "learning_rate": 0.0001956877301137864,
+      "loss": 2.7229,
+      "step": 1967
+    },
+    {
+      "epoch": 0.15882495359535145,
+      "grad_norm": 0.7326748371124268,
+      "learning_rate": 0.00019568314294769908,
+      "loss": 2.7728,
+      "step": 1968
+    },
+    {
+      "epoch": 0.15890565733193446,
+      "grad_norm": 0.790492594242096,
+      "learning_rate": 0.00019567855339693772,
+      "loss": 2.7809,
+      "step": 1969
+    },
+    {
+      "epoch": 0.15898636106851746,
+      "grad_norm": 0.8026898503303528,
+      "learning_rate": 0.0001956739614616167,
+      "loss": 2.7267,
+      "step": 1970
+    },
+    {
+      "epoch": 0.15906706480510047,
+      "grad_norm": 0.7963770627975464,
+      "learning_rate": 0.00019566936714185046,
+      "loss": 2.7161,
+      "step": 1971
+    },
+    {
+      "epoch": 0.15914776854168347,
+      "grad_norm": 0.7708200216293335,
+      "learning_rate": 0.00019566477043775354,
+      "loss": 2.7223,
+      "step": 1972
+    },
+    {
+      "epoch": 0.15922847227826648,
+      "grad_norm": 0.8036624789237976,
+      "learning_rate": 0.00019566017134944042,
+      "loss": 2.7644,
+      "step": 1973
+    },
+    {
+      "epoch": 0.15930917601484948,
+      "grad_norm": 0.8221341967582703,
+      "learning_rate": 0.00019565556987702581,
+      "loss": 2.7629,
+      "step": 1974
+    },
+    {
+      "epoch": 0.15938987975143248,
+      "grad_norm": 0.7685462832450867,
+      "learning_rate": 0.00019565096602062435,
+      "loss": 2.8016,
+      "step": 1975
+    },
+    {
+      "epoch": 0.1594705834880155,
+      "grad_norm": 0.8173574209213257,
+      "learning_rate": 0.00019564635978035075,
+      "loss": 2.761,
+      "step": 1976
+    },
+    {
+      "epoch": 0.1595512872245985,
+      "grad_norm": 0.7567519545555115,
+      "learning_rate": 0.00019564175115631988,
+      "loss": 2.7794,
+      "step": 1977
+    },
+    {
+      "epoch": 0.1596319909611815,
+      "grad_norm": 0.8754587173461914,
+      "learning_rate": 0.00019563714014864654,
+      "loss": 2.7769,
+      "step": 1978
+    },
+    {
+      "epoch": 0.1597126946977645,
+      "grad_norm": 0.753871738910675,
+      "learning_rate": 0.00019563252675744569,
+      "loss": 2.7489,
+      "step": 1979
+    },
+    {
+      "epoch": 0.1597933984343475,
+      "grad_norm": 0.777103841304779,
+      "learning_rate": 0.00019562791098283225,
+      "loss": 2.7667,
+      "step": 1980
+    },
+    {
+      "epoch": 0.1598741021709305,
+      "grad_norm": 0.8227293491363525,
+      "learning_rate": 0.00019562329282492131,
+      "loss": 2.7904,
+      "step": 1981
+    },
+    {
+      "epoch": 0.15995480590751351,
+      "grad_norm": 0.7595541477203369,
+      "learning_rate": 0.00019561867228382797,
+      "loss": 2.7654,
+      "step": 1982
+    },
+    {
+      "epoch": 0.16003550964409652,
+      "grad_norm": 0.8330550789833069,
+      "learning_rate": 0.00019561404935966733,
+      "loss": 2.7533,
+      "step": 1983
+    },
+    {
+      "epoch": 0.16011621338067952,
+      "grad_norm": 0.8213297128677368,
+      "learning_rate": 0.0001956094240525547,
+      "loss": 2.8103,
+      "step": 1984
+    },
+    {
+      "epoch": 0.16019691711726253,
+      "grad_norm": 0.8046056628227234,
+      "learning_rate": 0.00019560479636260527,
+      "loss": 2.7666,
+      "step": 1985
+    },
+    {
+      "epoch": 0.16027762085384553,
+      "grad_norm": 0.7886037230491638,
+      "learning_rate": 0.0001956001662899344,
+      "loss": 2.7066,
+      "step": 1986
+    },
+    {
+      "epoch": 0.16035832459042854,
+      "grad_norm": 0.8300043940544128,
+      "learning_rate": 0.00019559553383465748,
+      "loss": 2.7617,
+      "step": 1987
+    },
+    {
+      "epoch": 0.16043902832701154,
+      "grad_norm": 0.7963815331459045,
+      "learning_rate": 0.00019559089899688994,
+      "loss": 2.6891,
+      "step": 1988
+    },
+    {
+      "epoch": 0.16051973206359454,
+      "grad_norm": 0.7794002294540405,
+      "learning_rate": 0.00019558626177674734,
+      "loss": 2.8012,
+      "step": 1989
+    },
+    {
+      "epoch": 0.16060043580017755,
+      "grad_norm": 0.8345863819122314,
+      "learning_rate": 0.00019558162217434526,
+      "loss": 2.7715,
+      "step": 1990
+    },
+    {
+      "epoch": 0.16068113953676055,
+      "grad_norm": 0.8883393406867981,
+      "learning_rate": 0.00019557698018979927,
+      "loss": 2.7863,
+      "step": 1991
+    },
+    {
+      "epoch": 0.16076184327334356,
+      "grad_norm": 0.8069450259208679,
+      "learning_rate": 0.0001955723358232251,
+      "loss": 2.759,
+      "step": 1992
+    },
+    {
+      "epoch": 0.16084254700992656,
+      "grad_norm": 0.9014191031455994,
+      "learning_rate": 0.00019556768907473852,
+      "loss": 2.711,
+      "step": 1993
+    },
+    {
+      "epoch": 0.16092325074650957,
+      "grad_norm": 0.8429470658302307,
+      "learning_rate": 0.0001955630399444553,
+      "loss": 2.6936,
+      "step": 1994
+    },
+    {
+      "epoch": 0.16100395448309257,
+      "grad_norm": 0.7859500050544739,
+      "learning_rate": 0.00019555838843249128,
+      "loss": 2.7343,
+      "step": 1995
+    },
+    {
+      "epoch": 0.16108465821967557,
+      "grad_norm": 0.8068249821662903,
+      "learning_rate": 0.00019555373453896245,
+      "loss": 2.7492,
+      "step": 1996
+    },
+    {
+      "epoch": 0.16116536195625858,
+      "grad_norm": 0.8194023370742798,
+      "learning_rate": 0.00019554907826398478,
+      "loss": 2.7265,
+      "step": 1997
+    },
+    {
+      "epoch": 0.16124606569284158,
+      "grad_norm": 0.8139404654502869,
+      "learning_rate": 0.00019554441960767434,
+      "loss": 2.7311,
+      "step": 1998
+    },
+    {
+      "epoch": 0.1613267694294246,
+      "grad_norm": 0.8210673928260803,
+      "learning_rate": 0.00019553975857014718,
+      "loss": 2.7095,
+      "step": 1999
+    },
+    {
+      "epoch": 0.1614074731660076,
+      "grad_norm": 0.8615561723709106,
+      "learning_rate": 0.0001955350951515195,
+      "loss": 2.7458,
+      "step": 2000
+    },
+    {
+      "epoch": 0.1614074731660076,
+      "eval_loss": 2.6739437580108643,
+      "eval_runtime": 813.8274,
+      "eval_samples_per_second": 3.219,
+      "eval_steps_per_second": 0.537,
+      "step": 2000
+    },
+    {
+      "epoch": 0.1614881769025906,
+      "grad_norm": 0.8945594429969788,
+      "learning_rate": 0.0001955304293519075,
+      "loss": 2.776,
+      "step": 2001
+    },
+    {
+      "epoch": 0.1615688806391736,
+      "grad_norm": 0.7943438291549683,
+      "learning_rate": 0.00019552576117142748,
+      "loss": 2.7484,
+      "step": 2002
+    },
+    {
+      "epoch": 0.1616495843757566,
+      "grad_norm": 0.8264374136924744,
+      "learning_rate": 0.00019552109061019582,
+      "loss": 2.7725,
+      "step": 2003
+    },
+    {
+      "epoch": 0.1617302881123396,
+      "grad_norm": 0.7591681480407715,
+      "learning_rate": 0.00019551641766832887,
+      "loss": 2.7217,
+      "step": 2004
+    },
+    {
+      "epoch": 0.16181099184892261,
+      "grad_norm": 0.8275293707847595,
+      "learning_rate": 0.0001955117423459431,
+      "loss": 2.7279,
+      "step": 2005
+    },
+    {
+      "epoch": 0.16189169558550562,
+      "grad_norm": 0.8109650611877441,
+      "learning_rate": 0.00019550706464315504,
+      "loss": 2.8111,
+      "step": 2006
+    },
+    {
+      "epoch": 0.16197239932208862,
+      "grad_norm": 0.8710397481918335,
+      "learning_rate": 0.00019550238456008127,
+      "loss": 2.7166,
+      "step": 2007
+    },
+    {
+      "epoch": 0.16205310305867163,
+      "grad_norm": 0.8569270968437195,
+      "learning_rate": 0.00019549770209683845,
+      "loss": 2.7739,
+      "step": 2008
+    },
+    {
+      "epoch": 0.16213380679525463,
+      "grad_norm": 0.7927817702293396,
+      "learning_rate": 0.00019549301725354325,
+      "loss": 2.7154,
+      "step": 2009
+    },
+    {
+      "epoch": 0.16221451053183764,
+      "grad_norm": 0.7576590776443481,
+      "learning_rate": 0.00019548833003031244,
+      "loss": 2.7276,
+      "step": 2010
+    },
+    {
+      "epoch": 0.16229521426842064,
+      "grad_norm": 0.8092780709266663,
+      "learning_rate": 0.00019548364042726283,
+      "loss": 2.7494,
+      "step": 2011
+    },
+    {
+      "epoch": 0.16237591800500364,
+      "grad_norm": 0.7643424868583679,
+      "learning_rate": 0.0001954789484445113,
+      "loss": 2.7877,
+      "step": 2012
+    },
+    {
+      "epoch": 0.16245662174158665,
+      "grad_norm": 0.8235166072845459,
+      "learning_rate": 0.0001954742540821748,
+      "loss": 2.7884,
+      "step": 2013
+    },
+    {
+      "epoch": 0.16253732547816965,
+      "grad_norm": 0.9297853708267212,
+      "learning_rate": 0.00019546955734037034,
+      "loss": 2.765,
+      "step": 2014
+    },
+    {
+      "epoch": 0.16261802921475263,
+      "grad_norm": 0.7778275609016418,
+      "learning_rate": 0.0001954648582192149,
+      "loss": 2.7178,
+      "step": 2015
+    },
+    {
+      "epoch": 0.16269873295133563,
+      "grad_norm": 0.8767017126083374,
+      "learning_rate": 0.00019546015671882566,
+      "loss": 2.8254,
+      "step": 2016
+    },
+    {
+      "epoch": 0.16277943668791864,
+      "grad_norm": 0.7870603203773499,
+      "learning_rate": 0.0001954554528393198,
+      "loss": 2.797,
+      "step": 2017
+    },
+    {
+      "epoch": 0.16286014042450164,
+      "grad_norm": 0.8112391233444214,
+      "learning_rate": 0.00019545074658081454,
+      "loss": 2.8562,
+      "step": 2018
+    },
+    {
+      "epoch": 0.16294084416108465,
+      "grad_norm": 0.8216677308082581,
+      "learning_rate": 0.00019544603794342713,
+      "loss": 2.7894,
+      "step": 2019
+    },
+    {
+      "epoch": 0.16302154789766765,
+      "grad_norm": 0.8445515632629395,
+      "learning_rate": 0.00019544132692727497,
+      "loss": 2.8618,
+      "step": 2020
+    },
+    {
+      "epoch": 0.16310225163425066,
+      "grad_norm": 0.8275444507598877,
+      "learning_rate": 0.00019543661353247548,
+      "loss": 2.8087,
+      "step": 2021
+    },
+    {
+      "epoch": 0.16318295537083366,
+      "grad_norm": 0.8142833709716797,
+      "learning_rate": 0.00019543189775914608,
+      "loss": 2.8075,
+      "step": 2022
+    },
+    {
+      "epoch": 0.16326365910741666,
+      "grad_norm": 0.8182976245880127,
+      "learning_rate": 0.0001954271796074043,
+      "loss": 2.8312,
+      "step": 2023
+    },
+    {
+      "epoch": 0.16334436284399967,
+      "grad_norm": 0.7629228234291077,
+      "learning_rate": 0.0001954224590773678,
+      "loss": 2.7191,
+      "step": 2024
+    },
+    {
+      "epoch": 0.16342506658058267,
+      "grad_norm": 0.8630000948905945,
+      "learning_rate": 0.00019541773616915418,
+      "loss": 2.8013,
+      "step": 2025
+    },
+    {
+      "epoch": 0.16350577031716568,
+      "grad_norm": 0.8917906880378723,
+      "learning_rate": 0.00019541301088288115,
+      "loss": 2.7573,
+      "step": 2026
+    },
+    {
+      "epoch": 0.16358647405374868,
+      "grad_norm": 0.8641694188117981,
+      "learning_rate": 0.00019540828321866648,
+      "loss": 2.7509,
+      "step": 2027
+    },
+    {
+      "epoch": 0.16366717779033169,
+      "grad_norm": 0.7687639594078064,
+      "learning_rate": 0.00019540355317662798,
+      "loss": 2.7266,
+      "step": 2028
+    },
+    {
+      "epoch": 0.1637478815269147,
+      "grad_norm": 0.7870400547981262,
+      "learning_rate": 0.00019539882075688355,
+      "loss": 2.8217,
+      "step": 2029
+    },
+    {
+      "epoch": 0.1638285852634977,
+      "grad_norm": 0.9373054504394531,
+      "learning_rate": 0.0001953940859595511,
+      "loss": 2.7562,
+      "step": 2030
+    },
+    {
+      "epoch": 0.1639092890000807,
+      "grad_norm": 0.7941255569458008,
+      "learning_rate": 0.00019538934878474872,
+      "loss": 2.7553,
+      "step": 2031
+    },
+    {
+      "epoch": 0.1639899927366637,
+      "grad_norm": 0.735977053642273,
+      "learning_rate": 0.00019538460923259438,
+      "loss": 2.7058,
+      "step": 2032
+    },
+    {
+      "epoch": 0.1640706964732467,
+      "grad_norm": 0.7812782526016235,
+      "learning_rate": 0.00019537986730320625,
+      "loss": 2.7885,
+      "step": 2033
+    },
+    {
+      "epoch": 0.1641514002098297,
+      "grad_norm": 1.1534128189086914,
+      "learning_rate": 0.0001953751229967025,
+      "loss": 2.7139,
+      "step": 2034
+    },
+    {
+      "epoch": 0.16423210394641272,
+      "grad_norm": 0.9139814972877502,
+      "learning_rate": 0.00019537037631320135,
+      "loss": 2.7869,
+      "step": 2035
+    },
+    {
+      "epoch": 0.16431280768299572,
+      "grad_norm": 0.8330421447753906,
+      "learning_rate": 0.00019536562725282116,
+      "loss": 2.7491,
+      "step": 2036
+    },
+    {
+      "epoch": 0.16439351141957873,
+      "grad_norm": 0.9040594696998596,
+      "learning_rate": 0.00019536087581568026,
+      "loss": 2.7637,
+      "step": 2037
+    },
+    {
+      "epoch": 0.16447421515616173,
+      "grad_norm": 0.9158666729927063,
+      "learning_rate": 0.00019535612200189705,
+      "loss": 2.7709,
+      "step": 2038
+    },
+    {
+      "epoch": 0.16455491889274473,
+      "grad_norm": 0.8668088912963867,
+      "learning_rate": 0.00019535136581158997,
+      "loss": 2.7994,
+      "step": 2039
+    },
+    {
+      "epoch": 0.16463562262932774,
+      "grad_norm": 0.9179345369338989,
+      "learning_rate": 0.00019534660724487764,
+      "loss": 2.747,
+      "step": 2040
+    },
+    {
+      "epoch": 0.16471632636591074,
+      "grad_norm": 0.9690881967544556,
+      "learning_rate": 0.00019534184630187862,
+      "loss": 2.742,
+      "step": 2041
+    },
+    {
+      "epoch": 0.16479703010249375,
+      "grad_norm": 0.8478729724884033,
+      "learning_rate": 0.00019533708298271157,
+      "loss": 2.7824,
+      "step": 2042
+    },
+    {
+      "epoch": 0.16487773383907675,
+      "grad_norm": 0.8286584615707397,
+      "learning_rate": 0.00019533231728749518,
+      "loss": 2.7263,
+      "step": 2043
+    },
+    {
+      "epoch": 0.16495843757565976,
+      "grad_norm": 0.8095324039459229,
+      "learning_rate": 0.00019532754921634826,
+      "loss": 2.7845,
+      "step": 2044
+    },
+    {
+      "epoch": 0.16503914131224276,
+      "grad_norm": 0.9552872776985168,
+      "learning_rate": 0.0001953227787693896,
+      "loss": 2.7676,
+      "step": 2045
+    },
+    {
+      "epoch": 0.16511984504882576,
+      "grad_norm": 1.021515130996704,
+      "learning_rate": 0.00019531800594673815,
+      "loss": 2.784,
+      "step": 2046
+    },
+    {
+      "epoch": 0.16520054878540877,
+      "grad_norm": 0.7847293019294739,
+      "learning_rate": 0.00019531323074851276,
+      "loss": 2.7319,
+      "step": 2047
+    },
+    {
+      "epoch": 0.16528125252199177,
+      "grad_norm": 0.7803899049758911,
+      "learning_rate": 0.0001953084531748326,
+      "loss": 2.8321,
+      "step": 2048
+    },
+    {
+      "epoch": 0.16536195625857478,
+      "grad_norm": 0.8687692880630493,
+      "learning_rate": 0.0001953036732258166,
+      "loss": 2.763,
+      "step": 2049
+    },
+    {
+      "epoch": 0.16544265999515778,
+      "grad_norm": 0.8212031126022339,
+      "learning_rate": 0.00019529889090158392,
+      "loss": 2.7262,
+      "step": 2050
+    },
+    {
+      "epoch": 0.16552336373174079,
+      "grad_norm": 0.8460689187049866,
+      "learning_rate": 0.0001952941062022538,
+      "loss": 2.8018,
+      "step": 2051
+    },
+    {
+      "epoch": 0.1656040674683238,
+      "grad_norm": 0.9189361929893494,
+      "learning_rate": 0.00019528931912794547,
+      "loss": 2.8079,
+      "step": 2052
+    },
+    {
+      "epoch": 0.1656847712049068,
+      "grad_norm": 0.9529987573623657,
+      "learning_rate": 0.00019528452967877816,
+      "loss": 2.8015,
+      "step": 2053
+    },
+    {
+      "epoch": 0.1657654749414898,
+      "grad_norm": 0.8468493223190308,
+      "learning_rate": 0.00019527973785487133,
+      "loss": 2.8013,
+      "step": 2054
+    },
+    {
+      "epoch": 0.1658461786780728,
+      "grad_norm": 0.8150945901870728,
+      "learning_rate": 0.00019527494365634436,
+      "loss": 2.7975,
+      "step": 2055
+    },
+    {
+      "epoch": 0.1659268824146558,
+      "grad_norm": 0.814942479133606,
+      "learning_rate": 0.00019527014708331674,
+      "loss": 2.7503,
+      "step": 2056
+    },
+    {
+      "epoch": 0.1660075861512388,
+      "grad_norm": 0.7841517329216003,
+      "learning_rate": 0.000195265348135908,
+      "loss": 2.7921,
+      "step": 2057
+    },
+    {
+      "epoch": 0.16608828988782182,
+      "grad_norm": 0.7603738903999329,
+      "learning_rate": 0.0001952605468142378,
+      "loss": 2.7658,
+      "step": 2058
+    },
+    {
+      "epoch": 0.16616899362440482,
+      "grad_norm": 0.8460882902145386,
+      "learning_rate": 0.00019525574311842574,
+      "loss": 2.7644,
+      "step": 2059
+    },
+    {
+      "epoch": 0.16624969736098782,
+      "grad_norm": 0.8633555173873901,
+      "learning_rate": 0.00019525093704859156,
+      "loss": 2.7956,
+      "step": 2060
+    },
+    {
+      "epoch": 0.16633040109757083,
+      "grad_norm": 0.7700977325439453,
+      "learning_rate": 0.00019524612860485503,
+      "loss": 2.7103,
+      "step": 2061
+    },
+    {
+      "epoch": 0.16641110483415383,
+      "grad_norm": 0.888770580291748,
+      "learning_rate": 0.00019524131778733602,
+      "loss": 2.7325,
+      "step": 2062
+    },
+    {
+      "epoch": 0.16649180857073684,
+      "grad_norm": 0.8338149189949036,
+      "learning_rate": 0.00019523650459615438,
+      "loss": 2.7533,
+      "step": 2063
+    },
+    {
+      "epoch": 0.16657251230731984,
+      "grad_norm": 0.7723987698554993,
+      "learning_rate": 0.0001952316890314301,
+      "loss": 2.7316,
+      "step": 2064
+    },
+    {
+      "epoch": 0.16665321604390285,
+      "grad_norm": 0.8952934145927429,
+      "learning_rate": 0.0001952268710932832,
+      "loss": 2.7825,
+      "step": 2065
+    },
+    {
+      "epoch": 0.16673391978048582,
+      "grad_norm": 0.8201496601104736,
+      "learning_rate": 0.00019522205078183378,
+      "loss": 2.7162,
+      "step": 2066
+    },
+    {
+      "epoch": 0.16681462351706883,
+      "grad_norm": 0.7733781337738037,
+      "learning_rate": 0.00019521722809720188,
+      "loss": 2.7834,
+      "step": 2067
+    },
+    {
+      "epoch": 0.16689532725365183,
+      "grad_norm": 0.8285118937492371,
+      "learning_rate": 0.0001952124030395078,
+      "loss": 2.8475,
+      "step": 2068
+    },
+    {
+      "epoch": 0.16697603099023484,
+      "grad_norm": 0.84097820520401,
+      "learning_rate": 0.00019520757560887174,
+      "loss": 2.784,
+      "step": 2069
+    },
+    {
+      "epoch": 0.16705673472681784,
+      "grad_norm": 0.7336563467979431,
+      "learning_rate": 0.000195202745805414,
+      "loss": 2.7663,
+      "step": 2070
+    },
+    {
+      "epoch": 0.16713743846340084,
+      "grad_norm": 0.8359388113021851,
+      "learning_rate": 0.000195197913629255,
+      "loss": 2.7931,
+      "step": 2071
+    },
+    {
+      "epoch": 0.16721814219998385,
+      "grad_norm": 0.8272559642791748,
+      "learning_rate": 0.0001951930790805151,
+      "loss": 2.8578,
+      "step": 2072
+    },
+    {
+      "epoch": 0.16729884593656685,
+      "grad_norm": 0.7970743179321289,
+      "learning_rate": 0.00019518824215931487,
+      "loss": 2.8148,
+      "step": 2073
+    },
+    {
+      "epoch": 0.16737954967314986,
+      "grad_norm": 0.856200098991394,
+      "learning_rate": 0.00019518340286577482,
+      "loss": 2.8067,
+      "step": 2074
+    },
+    {
+      "epoch": 0.16746025340973286,
+      "grad_norm": 0.7581893801689148,
+      "learning_rate": 0.00019517856120001556,
+      "loss": 2.7339,
+      "step": 2075
+    },
+    {
+      "epoch": 0.16754095714631587,
+      "grad_norm": 0.8488386869430542,
+      "learning_rate": 0.00019517371716215774,
+      "loss": 2.7332,
+      "step": 2076
+    },
+    {
+      "epoch": 0.16762166088289887,
+      "grad_norm": 0.7488275170326233,
+      "learning_rate": 0.00019516887075232212,
+      "loss": 2.7734,
+      "step": 2077
+    },
+    {
+      "epoch": 0.16770236461948188,
+      "grad_norm": 0.9173932075500488,
+      "learning_rate": 0.00019516402197062945,
+      "loss": 2.7792,
+      "step": 2078
+    },
+    {
+      "epoch": 0.16778306835606488,
+      "grad_norm": 0.8200702667236328,
+      "learning_rate": 0.0001951591708172006,
+      "loss": 2.8046,
+      "step": 2079
+    },
+    {
+      "epoch": 0.16786377209264788,
+      "grad_norm": 0.8270781636238098,
+      "learning_rate": 0.00019515431729215642,
+      "loss": 2.7467,
+      "step": 2080
+    },
+    {
+      "epoch": 0.1679444758292309,
+      "grad_norm": 0.8660609722137451,
+      "learning_rate": 0.00019514946139561799,
+      "loss": 2.8169,
+      "step": 2081
+    },
+    {
+      "epoch": 0.1680251795658139,
+      "grad_norm": 0.78753262758255,
+      "learning_rate": 0.0001951446031277062,
+      "loss": 2.7388,
+      "step": 2082
+    },
+    {
+      "epoch": 0.1681058833023969,
+      "grad_norm": 0.791593074798584,
+      "learning_rate": 0.00019513974248854224,
+      "loss": 2.8776,
+      "step": 2083
+    },
+    {
+      "epoch": 0.1681865870389799,
+      "grad_norm": 0.7883535623550415,
+      "learning_rate": 0.0001951348794782472,
+      "loss": 2.78,
+      "step": 2084
+    },
+    {
+      "epoch": 0.1682672907755629,
+      "grad_norm": 0.7877013087272644,
+      "learning_rate": 0.00019513001409694224,
+      "loss": 2.7559,
+      "step": 2085
+    },
+    {
+      "epoch": 0.1683479945121459,
+      "grad_norm": 0.8838450908660889,
+      "learning_rate": 0.00019512514634474864,
+      "loss": 2.7489,
+      "step": 2086
+    },
+    {
+      "epoch": 0.16842869824872891,
+      "grad_norm": 0.7751588821411133,
+      "learning_rate": 0.00019512027622178775,
+      "loss": 2.6832,
+      "step": 2087
+    },
+    {
+      "epoch": 0.16850940198531192,
+      "grad_norm": 0.90345299243927,
+      "learning_rate": 0.00019511540372818095,
+      "loss": 2.8189,
+      "step": 2088
+    },
+    {
+      "epoch": 0.16859010572189492,
+      "grad_norm": 0.7820938229560852,
+      "learning_rate": 0.00019511052886404966,
+      "loss": 2.7655,
+      "step": 2089
+    },
+    {
+      "epoch": 0.16867080945847793,
+      "grad_norm": 0.8250375986099243,
+      "learning_rate": 0.00019510565162951537,
+      "loss": 2.7866,
+      "step": 2090
+    },
+    {
+      "epoch": 0.16875151319506093,
+      "grad_norm": 0.8063845634460449,
+      "learning_rate": 0.00019510077202469962,
+      "loss": 2.7774,
+      "step": 2091
+    },
+    {
+      "epoch": 0.16883221693164394,
+      "grad_norm": 0.7627965807914734,
+      "learning_rate": 0.00019509589004972403,
+      "loss": 2.7201,
+      "step": 2092
+    },
+    {
+      "epoch": 0.16891292066822694,
+      "grad_norm": 0.8392470479011536,
+      "learning_rate": 0.00019509100570471027,
+      "loss": 2.7613,
+      "step": 2093
+    },
+    {
+      "epoch": 0.16899362440480994,
+      "grad_norm": 0.7807552814483643,
+      "learning_rate": 0.0001950861189897801,
+      "loss": 2.7451,
+      "step": 2094
+    },
+    {
+      "epoch": 0.16907432814139295,
+      "grad_norm": 0.7829259634017944,
+      "learning_rate": 0.00019508122990505528,
+      "loss": 2.7128,
+      "step": 2095
+    },
+    {
+      "epoch": 0.16915503187797595,
+      "grad_norm": 0.7793046832084656,
+      "learning_rate": 0.00019507633845065766,
+      "loss": 2.7849,
+      "step": 2096
+    },
+    {
+      "epoch": 0.16923573561455896,
+      "grad_norm": 0.869752824306488,
+      "learning_rate": 0.00019507144462670915,
+      "loss": 2.7882,
+      "step": 2097
+    },
+    {
+      "epoch": 0.16931643935114196,
+      "grad_norm": 0.7550783753395081,
+      "learning_rate": 0.00019506654843333174,
+      "loss": 2.7211,
+      "step": 2098
+    },
+    {
+      "epoch": 0.16939714308772497,
+      "grad_norm": 0.8364891409873962,
+      "learning_rate": 0.0001950616498706474,
+      "loss": 2.7171,
+      "step": 2099
+    },
+    {
+      "epoch": 0.16947784682430797,
+      "grad_norm": 0.8026537299156189,
+      "learning_rate": 0.0001950567489387783,
+      "loss": 2.8362,
+      "step": 2100
+    },
+    {
+      "epoch": 0.16955855056089097,
+      "grad_norm": 0.8073398470878601,
+      "learning_rate": 0.00019505184563784652,
+      "loss": 2.7635,
+      "step": 2101
+    },
+    {
+      "epoch": 0.16963925429747398,
+      "grad_norm": 0.8168368935585022,
+      "learning_rate": 0.00019504693996797424,
+      "loss": 2.7553,
+      "step": 2102
+    },
+    {
+      "epoch": 0.16971995803405698,
+      "grad_norm": 0.7933681011199951,
+      "learning_rate": 0.0001950420319292838,
+      "loss": 2.7887,
+      "step": 2103
+    },
+    {
+      "epoch": 0.16980066177064,
+      "grad_norm": 0.8326540589332581,
+      "learning_rate": 0.00019503712152189748,
+      "loss": 2.7844,
+      "step": 2104
+    },
+    {
+      "epoch": 0.169881365507223,
+      "grad_norm": 0.8357202410697937,
+      "learning_rate": 0.00019503220874593765,
+      "loss": 2.7744,
+      "step": 2105
+    },
+    {
+      "epoch": 0.169962069243806,
+      "grad_norm": 0.8541022539138794,
+      "learning_rate": 0.00019502729360152676,
+      "loss": 2.7867,
+      "step": 2106
+    },
+    {
+      "epoch": 0.170042772980389,
+      "grad_norm": 0.8338841795921326,
+      "learning_rate": 0.0001950223760887873,
+      "loss": 2.7208,
+      "step": 2107
+    },
+    {
+      "epoch": 0.170123476716972,
+      "grad_norm": 0.8824255466461182,
+      "learning_rate": 0.00019501745620784187,
+      "loss": 2.7658,
+      "step": 2108
+    },
+    {
+      "epoch": 0.170204180453555,
+      "grad_norm": 0.7710463404655457,
+      "learning_rate": 0.00019501253395881306,
+      "loss": 2.7167,
+      "step": 2109
+    },
+    {
+      "epoch": 0.17028488419013801,
+      "grad_norm": 0.7740076184272766,
+      "learning_rate": 0.0001950076093418235,
+      "loss": 2.7251,
+      "step": 2110
+    },
+    {
+      "epoch": 0.17036558792672102,
+      "grad_norm": 0.8258434534072876,
+      "learning_rate": 0.00019500268235699597,
+      "loss": 2.7533,
+      "step": 2111
+    },
+    {
+      "epoch": 0.17044629166330402,
+      "grad_norm": 0.8347997069358826,
+      "learning_rate": 0.00019499775300445326,
+      "loss": 2.7372,
+      "step": 2112
+    },
+    {
+      "epoch": 0.17052699539988703,
+      "grad_norm": 0.8246529698371887,
+      "learning_rate": 0.00019499282128431823,
+      "loss": 2.7458,
+      "step": 2113
+    },
+    {
+      "epoch": 0.17060769913647003,
+      "grad_norm": 0.8510704040527344,
+      "learning_rate": 0.00019498788719671378,
+      "loss": 2.8144,
+      "step": 2114
+    },
+    {
+      "epoch": 0.17068840287305304,
+      "grad_norm": 0.7793454527854919,
+      "learning_rate": 0.00019498295074176286,
+      "loss": 2.7927,
+      "step": 2115
+    },
+    {
+      "epoch": 0.17076910660963604,
+      "grad_norm": 0.7888665199279785,
+      "learning_rate": 0.00019497801191958853,
+      "loss": 2.7156,
+      "step": 2116
+    },
+    {
+      "epoch": 0.17084981034621902,
+      "grad_norm": 0.8502812385559082,
+      "learning_rate": 0.00019497307073031386,
+      "loss": 2.7906,
+      "step": 2117
+    },
+    {
+      "epoch": 0.17093051408280202,
+      "grad_norm": 0.8376502990722656,
+      "learning_rate": 0.00019496812717406203,
+      "loss": 2.7354,
+      "step": 2118
+    },
+    {
+      "epoch": 0.17101121781938503,
+      "grad_norm": 0.7974401116371155,
+      "learning_rate": 0.0001949631812509562,
+      "loss": 2.7755,
+      "step": 2119
+    },
+    {
+      "epoch": 0.17109192155596803,
+      "grad_norm": 0.7760190963745117,
+      "learning_rate": 0.00019495823296111965,
+      "loss": 2.7694,
+      "step": 2120
+    },
+    {
+      "epoch": 0.17117262529255103,
+      "grad_norm": 0.7721701860427856,
+      "learning_rate": 0.00019495328230467575,
+      "loss": 2.7474,
+      "step": 2121
+    },
+    {
+      "epoch": 0.17125332902913404,
+      "grad_norm": 0.7360577583312988,
+      "learning_rate": 0.0001949483292817478,
+      "loss": 2.8044,
+      "step": 2122
+    },
+    {
+      "epoch": 0.17133403276571704,
+      "grad_norm": 0.7536107301712036,
+      "learning_rate": 0.0001949433738924593,
+      "loss": 2.8165,
+      "step": 2123
+    },
+    {
+      "epoch": 0.17141473650230005,
+      "grad_norm": 0.7668276429176331,
+      "learning_rate": 0.00019493841613693375,
+      "loss": 2.7964,
+      "step": 2124
+    },
+    {
+      "epoch": 0.17149544023888305,
+      "grad_norm": 0.8323161602020264,
+      "learning_rate": 0.0001949334560152947,
+      "loss": 2.7395,
+      "step": 2125
+    },
+    {
+      "epoch": 0.17157614397546606,
+      "grad_norm": 0.8132179975509644,
+      "learning_rate": 0.00019492849352766576,
+      "loss": 2.7511,
+      "step": 2126
+    },
+    {
+      "epoch": 0.17165684771204906,
+      "grad_norm": 0.7806998491287231,
+      "learning_rate": 0.0001949235286741706,
+      "loss": 2.7649,
+      "step": 2127
+    },
+    {
+      "epoch": 0.17173755144863206,
+      "grad_norm": 0.8315939903259277,
+      "learning_rate": 0.00019491856145493298,
+      "loss": 2.7742,
+      "step": 2128
+    },
+    {
+      "epoch": 0.17181825518521507,
+      "grad_norm": 0.8368063569068909,
+      "learning_rate": 0.00019491359187007672,
+      "loss": 2.7667,
+      "step": 2129
+    },
+    {
+      "epoch": 0.17189895892179807,
+      "grad_norm": 0.9183431267738342,
+      "learning_rate": 0.0001949086199197256,
+      "loss": 2.7444,
+      "step": 2130
+    },
+    {
+      "epoch": 0.17197966265838108,
+      "grad_norm": 0.7824065089225769,
+      "learning_rate": 0.0001949036456040036,
+      "loss": 2.7455,
+      "step": 2131
+    },
+    {
+      "epoch": 0.17206036639496408,
+      "grad_norm": 0.777974009513855,
+      "learning_rate": 0.00019489866892303468,
+      "loss": 2.7466,
+      "step": 2132
+    },
+    {
+      "epoch": 0.17214107013154709,
+      "grad_norm": 0.8068816065788269,
+      "learning_rate": 0.00019489368987694286,
+      "loss": 2.7081,
+      "step": 2133
+    },
+    {
+      "epoch": 0.1722217738681301,
+      "grad_norm": 0.8757622838020325,
+      "learning_rate": 0.00019488870846585222,
+      "loss": 2.8005,
+      "step": 2134
+    },
+    {
+      "epoch": 0.1723024776047131,
+      "grad_norm": 0.7967162728309631,
+      "learning_rate": 0.00019488372468988693,
+      "loss": 2.7737,
+      "step": 2135
+    },
+    {
+      "epoch": 0.1723831813412961,
+      "grad_norm": 0.7700283527374268,
+      "learning_rate": 0.00019487873854917117,
+      "loss": 2.7431,
+      "step": 2136
+    },
+    {
+      "epoch": 0.1724638850778791,
+      "grad_norm": 0.8259130716323853,
+      "learning_rate": 0.00019487375004382927,
+      "loss": 2.7635,
+      "step": 2137
+    },
+    {
+      "epoch": 0.1725445888144621,
+      "grad_norm": 0.8253815770149231,
+      "learning_rate": 0.0001948687591739855,
+      "loss": 2.7046,
+      "step": 2138
+    },
+    {
+      "epoch": 0.1726252925510451,
+      "grad_norm": 0.8087987303733826,
+      "learning_rate": 0.00019486376593976426,
+      "loss": 2.7728,
+      "step": 2139
+    },
+    {
+      "epoch": 0.17270599628762812,
+      "grad_norm": 0.8437588214874268,
+      "learning_rate": 0.00019485877034128998,
+      "loss": 2.7606,
+      "step": 2140
+    },
+    {
+      "epoch": 0.17278670002421112,
+      "grad_norm": 0.8416075110435486,
+      "learning_rate": 0.00019485377237868723,
+      "loss": 2.7396,
+      "step": 2141
+    },
+    {
+      "epoch": 0.17286740376079412,
+      "grad_norm": 0.784275472164154,
+      "learning_rate": 0.00019484877205208046,
+      "loss": 2.766,
+      "step": 2142
+    },
+    {
+      "epoch": 0.17294810749737713,
+      "grad_norm": 0.8082472681999207,
+      "learning_rate": 0.0001948437693615944,
+      "loss": 2.8,
+      "step": 2143
+    },
+    {
+      "epoch": 0.17302881123396013,
+      "grad_norm": 0.8904329538345337,
+      "learning_rate": 0.00019483876430735365,
+      "loss": 2.6579,
+      "step": 2144
+    },
+    {
+      "epoch": 0.17310951497054314,
+      "grad_norm": 0.7864851355552673,
+      "learning_rate": 0.000194833756889483,
+      "loss": 2.8231,
+      "step": 2145
+    },
+    {
+      "epoch": 0.17319021870712614,
+      "grad_norm": 0.7445049285888672,
+      "learning_rate": 0.00019482874710810723,
+      "loss": 2.7498,
+      "step": 2146
+    },
+    {
+      "epoch": 0.17327092244370915,
+      "grad_norm": 0.8266116380691528,
+      "learning_rate": 0.00019482373496335117,
+      "loss": 2.7152,
+      "step": 2147
+    },
+    {
+      "epoch": 0.17335162618029215,
+      "grad_norm": 0.7712300419807434,
+      "learning_rate": 0.0001948187204553398,
+      "loss": 2.7751,
+      "step": 2148
+    },
+    {
+      "epoch": 0.17343232991687516,
+      "grad_norm": 0.7472708225250244,
+      "learning_rate": 0.00019481370358419807,
+      "loss": 2.7397,
+      "step": 2149
+    },
+    {
+      "epoch": 0.17351303365345816,
+      "grad_norm": 0.763454020023346,
+      "learning_rate": 0.00019480868435005095,
+      "loss": 2.7174,
+      "step": 2150
+    },
+    {
+      "epoch": 0.17359373739004116,
+      "grad_norm": 0.8187674283981323,
+      "learning_rate": 0.00019480366275302362,
+      "loss": 2.8424,
+      "step": 2151
+    },
+    {
+      "epoch": 0.17367444112662417,
+      "grad_norm": 0.8183228373527527,
+      "learning_rate": 0.0001947986387932412,
+      "loss": 2.7351,
+      "step": 2152
+    },
+    {
+      "epoch": 0.17375514486320717,
+      "grad_norm": 0.807231605052948,
+      "learning_rate": 0.00019479361247082884,
+      "loss": 2.8054,
+      "step": 2153
+    },
+    {
+      "epoch": 0.17383584859979018,
+      "grad_norm": 0.8383626341819763,
+      "learning_rate": 0.00019478858378591194,
+      "loss": 2.7181,
+      "step": 2154
+    },
+    {
+      "epoch": 0.17391655233637318,
+      "grad_norm": 0.8330298662185669,
+      "learning_rate": 0.0001947835527386157,
+      "loss": 2.748,
+      "step": 2155
+    },
+    {
+      "epoch": 0.17399725607295619,
+      "grad_norm": 0.8433073163032532,
+      "learning_rate": 0.0001947785193290656,
+      "loss": 2.8115,
+      "step": 2156
+    },
+    {
+      "epoch": 0.1740779598095392,
+      "grad_norm": 0.8873384594917297,
+      "learning_rate": 0.000194773483557387,
+      "loss": 2.8288,
+      "step": 2157
+    },
+    {
+      "epoch": 0.1741586635461222,
+      "grad_norm": 0.8399423360824585,
+      "learning_rate": 0.00019476844542370546,
+      "loss": 2.7514,
+      "step": 2158
+    },
+    {
+      "epoch": 0.1742393672827052,
+      "grad_norm": 0.7808830738067627,
+      "learning_rate": 0.00019476340492814655,
+      "loss": 2.7003,
+      "step": 2159
+    },
+    {
+      "epoch": 0.1743200710192882,
+      "grad_norm": 0.8268750905990601,
+      "learning_rate": 0.00019475836207083589,
+      "loss": 2.7961,
+      "step": 2160
+    },
+    {
+      "epoch": 0.1744007747558712,
+      "grad_norm": 0.9144260883331299,
+      "learning_rate": 0.0001947533168518991,
+      "loss": 2.769,
+      "step": 2161
+    },
+    {
+      "epoch": 0.1744814784924542,
+      "grad_norm": 0.8409113883972168,
+      "learning_rate": 0.000194748269271462,
+      "loss": 2.8004,
+      "step": 2162
+    },
+    {
+      "epoch": 0.17456218222903722,
+      "grad_norm": 0.8747037649154663,
+      "learning_rate": 0.00019474321932965035,
+      "loss": 2.7602,
+      "step": 2163
+    },
+    {
+      "epoch": 0.17464288596562022,
+      "grad_norm": 0.8582575917243958,
+      "learning_rate": 0.00019473816702659,
+      "loss": 2.7292,
+      "step": 2164
+    },
+    {
+      "epoch": 0.17472358970220322,
+      "grad_norm": 0.7402843832969666,
+      "learning_rate": 0.0001947331123624069,
+      "loss": 2.7287,
+      "step": 2165
+    },
+    {
+      "epoch": 0.17480429343878623,
+      "grad_norm": 0.8019410967826843,
+      "learning_rate": 0.000194728055337227,
+      "loss": 2.7451,
+      "step": 2166
+    },
+    {
+      "epoch": 0.17488499717536923,
+      "grad_norm": 0.9137046337127686,
+      "learning_rate": 0.0001947229959511763,
+      "loss": 2.808,
+      "step": 2167
+    },
+    {
+      "epoch": 0.1749657009119522,
+      "grad_norm": 0.7539177536964417,
+      "learning_rate": 0.000194717934204381,
+      "loss": 2.7031,
+      "step": 2168
+    },
+    {
+      "epoch": 0.17504640464853521,
+      "grad_norm": 0.8611089587211609,
+      "learning_rate": 0.00019471287009696715,
+      "loss": 2.8751,
+      "step": 2169
+    },
+    {
+      "epoch": 0.17512710838511822,
+      "grad_norm": 0.906134843826294,
+      "learning_rate": 0.000194707803629061,
+      "loss": 2.9163,
+      "step": 2170
+    },
+    {
+      "epoch": 0.17520781212170122,
+      "grad_norm": 0.8066667318344116,
+      "learning_rate": 0.00019470273480078879,
+      "loss": 2.7549,
+      "step": 2171
+    },
+    {
+      "epoch": 0.17528851585828423,
+      "grad_norm": 0.7962325215339661,
+      "learning_rate": 0.00019469766361227692,
+      "loss": 2.7964,
+      "step": 2172
+    },
+    {
+      "epoch": 0.17536921959486723,
+      "grad_norm": 0.7802287340164185,
+      "learning_rate": 0.0001946925900636517,
+      "loss": 2.7022,
+      "step": 2173
+    },
+    {
+      "epoch": 0.17544992333145024,
+      "grad_norm": 0.783478319644928,
+      "learning_rate": 0.0001946875141550396,
+      "loss": 2.7798,
+      "step": 2174
+    },
+    {
+      "epoch": 0.17553062706803324,
+      "grad_norm": 0.8006815314292908,
+      "learning_rate": 0.00019468243588656713,
+      "loss": 2.7345,
+      "step": 2175
+    },
+    {
+      "epoch": 0.17561133080461624,
+      "grad_norm": 0.7566428184509277,
+      "learning_rate": 0.00019467735525836085,
+      "loss": 2.7822,
+      "step": 2176
+    },
+    {
+      "epoch": 0.17569203454119925,
+      "grad_norm": 0.772282600402832,
+      "learning_rate": 0.0001946722722705474,
+      "loss": 2.7346,
+      "step": 2177
+    },
+    {
+      "epoch": 0.17577273827778225,
+      "grad_norm": 0.7808345556259155,
+      "learning_rate": 0.00019466718692325347,
+      "loss": 2.755,
+      "step": 2178
+    },
+    {
+      "epoch": 0.17585344201436526,
+      "grad_norm": 0.8150362372398376,
+      "learning_rate": 0.00019466209921660576,
+      "loss": 2.7691,
+      "step": 2179
+    },
+    {
+      "epoch": 0.17593414575094826,
+      "grad_norm": 0.7952939867973328,
+      "learning_rate": 0.0001946570091507311,
+      "loss": 2.8175,
+      "step": 2180
+    },
+    {
+      "epoch": 0.17601484948753127,
+      "grad_norm": 0.8211334347724915,
+      "learning_rate": 0.00019465191672575634,
+      "loss": 2.7561,
+      "step": 2181
+    },
+    {
+      "epoch": 0.17609555322411427,
+      "grad_norm": 0.7726178765296936,
+      "learning_rate": 0.00019464682194180838,
+      "loss": 2.7435,
+      "step": 2182
+    },
+    {
+      "epoch": 0.17617625696069728,
+      "grad_norm": 0.7614372372627258,
+      "learning_rate": 0.00019464172479901422,
+      "loss": 2.7301,
+      "step": 2183
+    },
+    {
+      "epoch": 0.17625696069728028,
+      "grad_norm": 0.7818898558616638,
+      "learning_rate": 0.00019463662529750083,
+      "loss": 2.6964,
+      "step": 2184
+    },
+    {
+      "epoch": 0.17633766443386328,
+      "grad_norm": 0.7849796414375305,
+      "learning_rate": 0.0001946315234373954,
+      "loss": 2.7431,
+      "step": 2185
+    },
+    {
+      "epoch": 0.1764183681704463,
+      "grad_norm": 0.7939459085464478,
+      "learning_rate": 0.00019462641921882506,
+      "loss": 2.7126,
+      "step": 2186
+    },
+    {
+      "epoch": 0.1764990719070293,
+      "grad_norm": 0.8391629457473755,
+      "learning_rate": 0.00019462131264191696,
+      "loss": 2.8394,
+      "step": 2187
+    },
+    {
+      "epoch": 0.1765797756436123,
+      "grad_norm": 0.7548067569732666,
+      "learning_rate": 0.0001946162037067984,
+      "loss": 2.7315,
+      "step": 2188
+    },
+    {
+      "epoch": 0.1766604793801953,
+      "grad_norm": 0.8278634548187256,
+      "learning_rate": 0.00019461109241359674,
+      "loss": 2.8298,
+      "step": 2189
+    },
+    {
+      "epoch": 0.1767411831167783,
+      "grad_norm": 0.8275949954986572,
+      "learning_rate": 0.00019460597876243933,
+      "loss": 2.8072,
+      "step": 2190
+    },
+    {
+      "epoch": 0.1768218868533613,
+      "grad_norm": 0.7720363140106201,
+      "learning_rate": 0.00019460086275345363,
+      "loss": 2.7478,
+      "step": 2191
+    },
+    {
+      "epoch": 0.17690259058994431,
+      "grad_norm": 0.7795925140380859,
+      "learning_rate": 0.00019459574438676714,
+      "loss": 2.7633,
+      "step": 2192
+    },
+    {
+      "epoch": 0.17698329432652732,
+      "grad_norm": 0.7722043991088867,
+      "learning_rate": 0.00019459062366250743,
+      "loss": 2.8001,
+      "step": 2193
+    },
+    {
+      "epoch": 0.17706399806311032,
+      "grad_norm": 0.8560587763786316,
+      "learning_rate": 0.00019458550058080212,
+      "loss": 2.7494,
+      "step": 2194
+    },
+    {
+      "epoch": 0.17714470179969333,
+      "grad_norm": 0.7473754286766052,
+      "learning_rate": 0.00019458037514177886,
+      "loss": 2.7112,
+      "step": 2195
+    },
+    {
+      "epoch": 0.17722540553627633,
+      "grad_norm": 0.7625827789306641,
+      "learning_rate": 0.00019457524734556542,
+      "loss": 2.7496,
+      "step": 2196
+    },
+    {
+      "epoch": 0.17730610927285934,
+      "grad_norm": 0.7809351682662964,
+      "learning_rate": 0.00019457011719228962,
+      "loss": 2.7764,
+      "step": 2197
+    },
+    {
+      "epoch": 0.17738681300944234,
+      "grad_norm": 0.7846190333366394,
+      "learning_rate": 0.00019456498468207927,
+      "loss": 2.7189,
+      "step": 2198
+    },
+    {
+      "epoch": 0.17746751674602534,
+      "grad_norm": 0.7919551134109497,
+      "learning_rate": 0.0001945598498150623,
+      "loss": 2.7798,
+      "step": 2199
+    },
+    {
+      "epoch": 0.17754822048260835,
+      "grad_norm": 0.796183705329895,
+      "learning_rate": 0.0001945547125913667,
+      "loss": 2.7498,
+      "step": 2200
+    },
+    {
+      "epoch": 0.17762892421919135,
+      "grad_norm": 0.791668176651001,
+      "learning_rate": 0.0001945495730111205,
+      "loss": 2.7638,
+      "step": 2201
+    },
+    {
+      "epoch": 0.17770962795577436,
+      "grad_norm": 0.8303191661834717,
+      "learning_rate": 0.0001945444310744518,
+      "loss": 2.8079,
+      "step": 2202
+    },
+    {
+      "epoch": 0.17779033169235736,
+      "grad_norm": 0.8245917558670044,
+      "learning_rate": 0.00019453928678148872,
+      "loss": 2.7222,
+      "step": 2203
+    },
+    {
+      "epoch": 0.17787103542894037,
+      "grad_norm": 0.793456494808197,
+      "learning_rate": 0.0001945341401323595,
+      "loss": 2.8532,
+      "step": 2204
+    },
+    {
+      "epoch": 0.17795173916552337,
+      "grad_norm": 0.7574856877326965,
+      "learning_rate": 0.00019452899112719235,
+      "loss": 2.7361,
+      "step": 2205
+    },
+    {
+      "epoch": 0.17803244290210637,
+      "grad_norm": 0.7748556733131409,
+      "learning_rate": 0.0001945238397661157,
+      "loss": 2.7423,
+      "step": 2206
+    },
+    {
+      "epoch": 0.17811314663868938,
+      "grad_norm": 0.8973588347434998,
+      "learning_rate": 0.00019451868604925782,
+      "loss": 2.7604,
+      "step": 2207
+    },
+    {
+      "epoch": 0.17819385037527238,
+      "grad_norm": 0.7613589763641357,
+      "learning_rate": 0.00019451352997674722,
+      "loss": 2.7168,
+      "step": 2208
+    },
+    {
+      "epoch": 0.1782745541118554,
+      "grad_norm": 0.8152763247489929,
+      "learning_rate": 0.00019450837154871243,
+      "loss": 2.7904,
+      "step": 2209
+    },
+    {
+      "epoch": 0.1783552578484384,
+      "grad_norm": 0.8115083575248718,
+      "learning_rate": 0.00019450321076528194,
+      "loss": 2.7595,
+      "step": 2210
+    },
+    {
+      "epoch": 0.1784359615850214,
+      "grad_norm": 0.772665798664093,
+      "learning_rate": 0.00019449804762658438,
+      "loss": 2.7125,
+      "step": 2211
+    },
+    {
+      "epoch": 0.1785166653216044,
+      "grad_norm": 0.8002723455429077,
+      "learning_rate": 0.0001944928821327485,
+      "loss": 2.8121,
+      "step": 2212
+    },
+    {
+      "epoch": 0.1785973690581874,
+      "grad_norm": 0.8354858160018921,
+      "learning_rate": 0.00019448771428390296,
+      "loss": 2.8662,
+      "step": 2213
+    },
+    {
+      "epoch": 0.1786780727947704,
+      "grad_norm": 0.7799130082130432,
+      "learning_rate": 0.0001944825440801766,
+      "loss": 2.7247,
+      "step": 2214
+    },
+    {
+      "epoch": 0.1787587765313534,
+      "grad_norm": 0.810265302658081,
+      "learning_rate": 0.00019447737152169828,
+      "loss": 2.7095,
+      "step": 2215
+    },
+    {
+      "epoch": 0.17883948026793642,
+      "grad_norm": 0.8305599093437195,
+      "learning_rate": 0.00019447219660859687,
+      "loss": 2.7448,
+      "step": 2216
+    },
+    {
+      "epoch": 0.17892018400451942,
+      "grad_norm": 0.7899554371833801,
+      "learning_rate": 0.00019446701934100138,
+      "loss": 2.7295,
+      "step": 2217
+    },
+    {
+      "epoch": 0.17900088774110243,
+      "grad_norm": 0.7675672173500061,
+      "learning_rate": 0.00019446183971904082,
+      "loss": 2.7236,
+      "step": 2218
+    },
+    {
+      "epoch": 0.1790815914776854,
+      "grad_norm": 0.8717279434204102,
+      "learning_rate": 0.0001944566577428443,
+      "loss": 2.8044,
+      "step": 2219
+    },
+    {
+      "epoch": 0.1791622952142684,
+      "grad_norm": 0.8151431679725647,
+      "learning_rate": 0.00019445147341254094,
+      "loss": 2.7753,
+      "step": 2220
+    },
+    {
+      "epoch": 0.1792429989508514,
+      "grad_norm": 0.8481619358062744,
+      "learning_rate": 0.00019444628672825998,
+      "loss": 2.7954,
+      "step": 2221
+    },
+    {
+      "epoch": 0.17932370268743442,
+      "grad_norm": 0.8133199214935303,
+      "learning_rate": 0.00019444109769013065,
+      "loss": 2.7235,
+      "step": 2222
+    },
+    {
+      "epoch": 0.17940440642401742,
+      "grad_norm": 0.8250097036361694,
+      "learning_rate": 0.00019443590629828232,
+      "loss": 2.8352,
+      "step": 2223
+    },
+    {
+      "epoch": 0.17948511016060043,
+      "grad_norm": 0.8279787302017212,
+      "learning_rate": 0.00019443071255284433,
+      "loss": 2.7513,
+      "step": 2224
+    },
+    {
+      "epoch": 0.17956581389718343,
+      "grad_norm": 0.7781538963317871,
+      "learning_rate": 0.00019442551645394612,
+      "loss": 2.7239,
+      "step": 2225
+    },
+    {
+      "epoch": 0.17964651763376643,
+      "grad_norm": 0.7718615531921387,
+      "learning_rate": 0.00019442031800171727,
+      "loss": 2.7387,
+      "step": 2226
+    },
+    {
+      "epoch": 0.17972722137034944,
+      "grad_norm": 0.7704512476921082,
+      "learning_rate": 0.00019441511719628724,
+      "loss": 2.792,
+      "step": 2227
+    },
+    {
+      "epoch": 0.17980792510693244,
+      "grad_norm": 0.8290835618972778,
+      "learning_rate": 0.00019440991403778566,
+      "loss": 2.7745,
+      "step": 2228
+    },
+    {
+      "epoch": 0.17988862884351545,
+      "grad_norm": 0.8408392667770386,
+      "learning_rate": 0.00019440470852634227,
+      "loss": 2.7688,
+      "step": 2229
+    },
+    {
+      "epoch": 0.17996933258009845,
+      "grad_norm": 0.8503465056419373,
+      "learning_rate": 0.00019439950066208676,
+      "loss": 2.6747,
+      "step": 2230
+    },
+    {
+      "epoch": 0.18005003631668146,
+      "grad_norm": 0.8213364481925964,
+      "learning_rate": 0.0001943942904451489,
+      "loss": 2.7212,
+      "step": 2231
+    },
+    {
+      "epoch": 0.18013074005326446,
+      "grad_norm": 0.8511209487915039,
+      "learning_rate": 0.0001943890778756586,
+      "loss": 2.701,
+      "step": 2232
+    },
+    {
+      "epoch": 0.18021144378984746,
+      "grad_norm": 0.8034417033195496,
+      "learning_rate": 0.00019438386295374577,
+      "loss": 2.7029,
+      "step": 2233
+    },
+    {
+      "epoch": 0.18029214752643047,
+      "grad_norm": 0.7603715658187866,
+      "learning_rate": 0.0001943786456795403,
+      "loss": 2.7201,
+      "step": 2234
+    },
+    {
+      "epoch": 0.18037285126301347,
+      "grad_norm": 0.9210647940635681,
+      "learning_rate": 0.0001943734260531723,
+      "loss": 2.7847,
+      "step": 2235
+    },
+    {
+      "epoch": 0.18045355499959648,
+      "grad_norm": 0.7429665923118591,
+      "learning_rate": 0.00019436820407477186,
+      "loss": 2.7493,
+      "step": 2236
+    },
+    {
+      "epoch": 0.18053425873617948,
+      "grad_norm": 0.8290510773658752,
+      "learning_rate": 0.00019436297974446905,
+      "loss": 2.7711,
+      "step": 2237
+    },
+    {
+      "epoch": 0.18061496247276249,
+      "grad_norm": 0.7593570947647095,
+      "learning_rate": 0.0001943577530623941,
+      "loss": 2.7539,
+      "step": 2238
+    },
+    {
+      "epoch": 0.1806956662093455,
+      "grad_norm": 0.8222225308418274,
+      "learning_rate": 0.00019435252402867734,
+      "loss": 2.7703,
+      "step": 2239
+    },
+    {
+      "epoch": 0.1807763699459285,
+      "grad_norm": 0.8280842900276184,
+      "learning_rate": 0.00019434729264344898,
+      "loss": 2.7966,
+      "step": 2240
+    },
+    {
+      "epoch": 0.1808570736825115,
+      "grad_norm": 0.8258495926856995,
+      "learning_rate": 0.00019434205890683952,
+      "loss": 2.759,
+      "step": 2241
+    },
+    {
+      "epoch": 0.1809377774190945,
+      "grad_norm": 0.8294420838356018,
+      "learning_rate": 0.00019433682281897932,
+      "loss": 2.6996,
+      "step": 2242
+    },
+    {
+      "epoch": 0.1810184811556775,
+      "grad_norm": 0.8258811235427856,
+      "learning_rate": 0.0001943315843799989,
+      "loss": 2.774,
+      "step": 2243
+    },
+    {
+      "epoch": 0.1810991848922605,
+      "grad_norm": 0.8035838007926941,
+      "learning_rate": 0.0001943263435900288,
+      "loss": 2.7806,
+      "step": 2244
+    },
+    {
+      "epoch": 0.18117988862884352,
+      "grad_norm": 0.7900332808494568,
+      "learning_rate": 0.00019432110044919964,
+      "loss": 2.7462,
+      "step": 2245
+    },
+    {
+      "epoch": 0.18126059236542652,
+      "grad_norm": 0.8126730918884277,
+      "learning_rate": 0.00019431585495764212,
+      "loss": 2.6913,
+      "step": 2246
+    },
+    {
+      "epoch": 0.18134129610200952,
+      "grad_norm": 0.8411321043968201,
+      "learning_rate": 0.00019431060711548695,
+      "loss": 2.7503,
+      "step": 2247
+    },
+    {
+      "epoch": 0.18142199983859253,
+      "grad_norm": 0.7712867856025696,
+      "learning_rate": 0.0001943053569228649,
+      "loss": 2.7703,
+      "step": 2248
+    },
+    {
+      "epoch": 0.18150270357517553,
+      "grad_norm": 0.9093566536903381,
+      "learning_rate": 0.00019430010437990688,
+      "loss": 2.7838,
+      "step": 2249
+    },
+    {
+      "epoch": 0.18158340731175854,
+      "grad_norm": 0.8184913396835327,
+      "learning_rate": 0.00019429484948674372,
+      "loss": 2.8167,
+      "step": 2250
+    },
+    {
+      "epoch": 0.18166411104834154,
+      "grad_norm": 0.7215915322303772,
+      "learning_rate": 0.00019428959224350643,
+      "loss": 2.739,
+      "step": 2251
+    },
+    {
+      "epoch": 0.18174481478492455,
+      "grad_norm": 0.7842726111412048,
+      "learning_rate": 0.000194284332650326,
+      "loss": 2.8547,
+      "step": 2252
+    },
+    {
+      "epoch": 0.18182551852150755,
+      "grad_norm": 0.7758263349533081,
+      "learning_rate": 0.00019427907070733357,
+      "loss": 2.7746,
+      "step": 2253
+    },
+    {
+      "epoch": 0.18190622225809056,
+      "grad_norm": 0.7710500359535217,
+      "learning_rate": 0.00019427380641466027,
+      "loss": 2.7415,
+      "step": 2254
+    },
+    {
+      "epoch": 0.18198692599467356,
+      "grad_norm": 0.8233851194381714,
+      "learning_rate": 0.00019426853977243724,
+      "loss": 2.7471,
+      "step": 2255
+    },
+    {
+      "epoch": 0.18206762973125656,
+      "grad_norm": 0.7856284379959106,
+      "learning_rate": 0.00019426327078079578,
+      "loss": 2.6892,
+      "step": 2256
+    },
+    {
+      "epoch": 0.18214833346783957,
+      "grad_norm": 0.7978290915489197,
+      "learning_rate": 0.00019425799943986722,
+      "loss": 2.7346,
+      "step": 2257
+    },
+    {
+      "epoch": 0.18222903720442257,
+      "grad_norm": 0.8339362740516663,
+      "learning_rate": 0.00019425272574978293,
+      "loss": 2.7403,
+      "step": 2258
+    },
+    {
+      "epoch": 0.18230974094100558,
+      "grad_norm": 0.8035171031951904,
+      "learning_rate": 0.0001942474497106743,
+      "loss": 2.7444,
+      "step": 2259
+    },
+    {
+      "epoch": 0.18239044467758858,
+      "grad_norm": 0.7950475811958313,
+      "learning_rate": 0.0001942421713226729,
+      "loss": 2.7218,
+      "step": 2260
+    },
+    {
+      "epoch": 0.18247114841417159,
+      "grad_norm": 0.8439741730690002,
+      "learning_rate": 0.00019423689058591022,
+      "loss": 2.7498,
+      "step": 2261
+    },
+    {
+      "epoch": 0.1825518521507546,
+      "grad_norm": 0.8585919737815857,
+      "learning_rate": 0.00019423160750051789,
+      "loss": 2.7459,
+      "step": 2262
+    },
+    {
+      "epoch": 0.1826325558873376,
+      "grad_norm": 0.857276201248169,
+      "learning_rate": 0.00019422632206662755,
+      "loss": 2.8404,
+      "step": 2263
+    },
+    {
+      "epoch": 0.1827132596239206,
+      "grad_norm": 0.7692707777023315,
+      "learning_rate": 0.000194221034284371,
+      "loss": 2.8069,
+      "step": 2264
+    },
+    {
+      "epoch": 0.1827939633605036,
+      "grad_norm": 0.9107782244682312,
+      "learning_rate": 0.00019421574415387998,
+      "loss": 2.7554,
+      "step": 2265
+    },
+    {
+      "epoch": 0.1828746670970866,
+      "grad_norm": 0.763300895690918,
+      "learning_rate": 0.00019421045167528628,
+      "loss": 2.8031,
+      "step": 2266
+    },
+    {
+      "epoch": 0.1829553708336696,
+      "grad_norm": 0.8625530004501343,
+      "learning_rate": 0.0001942051568487219,
+      "loss": 2.7622,
+      "step": 2267
+    },
+    {
+      "epoch": 0.18303607457025262,
+      "grad_norm": 0.8483080863952637,
+      "learning_rate": 0.00019419985967431875,
+      "loss": 2.7726,
+      "step": 2268
+    },
+    {
+      "epoch": 0.18311677830683562,
+      "grad_norm": 0.8295309543609619,
+      "learning_rate": 0.00019419456015220884,
+      "loss": 2.7676,
+      "step": 2269
+    },
+    {
+      "epoch": 0.1831974820434186,
+      "grad_norm": 0.812976062297821,
+      "learning_rate": 0.0001941892582825243,
+      "loss": 2.745,
+      "step": 2270
+    },
+    {
+      "epoch": 0.1832781857800016,
+      "grad_norm": 0.799846351146698,
+      "learning_rate": 0.00019418395406539717,
+      "loss": 2.7474,
+      "step": 2271
+    },
+    {
+      "epoch": 0.1833588895165846,
+      "grad_norm": 0.7825174331665039,
+      "learning_rate": 0.00019417864750095976,
+      "loss": 2.7982,
+      "step": 2272
+    },
+    {
+      "epoch": 0.1834395932531676,
+      "grad_norm": 0.8331060409545898,
+      "learning_rate": 0.00019417333858934424,
+      "loss": 2.7279,
+      "step": 2273
+    },
+    {
+      "epoch": 0.18352029698975061,
+      "grad_norm": 0.8579809665679932,
+      "learning_rate": 0.00019416802733068295,
+      "loss": 2.7425,
+      "step": 2274
+    },
+    {
+      "epoch": 0.18360100072633362,
+      "grad_norm": 0.8643589019775391,
+      "learning_rate": 0.0001941627137251083,
+      "loss": 2.7369,
+      "step": 2275
+    },
+    {
+      "epoch": 0.18368170446291662,
+      "grad_norm": 0.9086846113204956,
+      "learning_rate": 0.00019415739777275265,
+      "loss": 2.7681,
+      "step": 2276
+    },
+    {
+      "epoch": 0.18376240819949963,
+      "grad_norm": 0.8442896604537964,
+      "learning_rate": 0.00019415207947374853,
+      "loss": 2.7733,
+      "step": 2277
+    },
+    {
+      "epoch": 0.18384311193608263,
+      "grad_norm": 0.7858592867851257,
+      "learning_rate": 0.00019414675882822846,
+      "loss": 2.7726,
+      "step": 2278
+    },
+    {
+      "epoch": 0.18392381567266564,
+      "grad_norm": 0.8191118240356445,
+      "learning_rate": 0.00019414143583632503,
+      "loss": 2.8142,
+      "step": 2279
+    },
+    {
+      "epoch": 0.18400451940924864,
+      "grad_norm": 0.8093815445899963,
+      "learning_rate": 0.00019413611049817097,
+      "loss": 2.7068,
+      "step": 2280
+    },
+    {
+      "epoch": 0.18408522314583164,
+      "grad_norm": 0.80247563123703,
+      "learning_rate": 0.00019413078281389895,
+      "loss": 2.7459,
+      "step": 2281
+    },
+    {
+      "epoch": 0.18416592688241465,
+      "grad_norm": 0.8200877904891968,
+      "learning_rate": 0.00019412545278364176,
+      "loss": 2.6963,
+      "step": 2282
+    },
+    {
+      "epoch": 0.18424663061899765,
+      "grad_norm": 0.870662271976471,
+      "learning_rate": 0.00019412012040753224,
+      "loss": 2.8636,
+      "step": 2283
+    },
+    {
+      "epoch": 0.18432733435558066,
+      "grad_norm": 0.7626601457595825,
+      "learning_rate": 0.00019411478568570332,
+      "loss": 2.8082,
+      "step": 2284
+    },
+    {
+      "epoch": 0.18440803809216366,
+      "grad_norm": 0.7492787837982178,
+      "learning_rate": 0.00019410944861828787,
+      "loss": 2.7231,
+      "step": 2285
+    },
+    {
+      "epoch": 0.18448874182874667,
+      "grad_norm": 0.8172419667243958,
+      "learning_rate": 0.000194104109205419,
+      "loss": 2.7054,
+      "step": 2286
+    },
+    {
+      "epoch": 0.18456944556532967,
+      "grad_norm": 0.7749670147895813,
+      "learning_rate": 0.0001940987674472297,
+      "loss": 2.6907,
+      "step": 2287
+    },
+    {
+      "epoch": 0.18465014930191267,
+      "grad_norm": 0.8855465054512024,
+      "learning_rate": 0.00019409342334385316,
+      "loss": 2.7439,
+      "step": 2288
+    },
+    {
+      "epoch": 0.18473085303849568,
+      "grad_norm": 0.8066419363021851,
+      "learning_rate": 0.00019408807689542257,
+      "loss": 2.7126,
+      "step": 2289
+    },
+    {
+      "epoch": 0.18481155677507868,
+      "grad_norm": 0.7759004235267639,
+      "learning_rate": 0.00019408272810207114,
+      "loss": 2.7207,
+      "step": 2290
+    },
+    {
+      "epoch": 0.1848922605116617,
+      "grad_norm": 0.8593513369560242,
+      "learning_rate": 0.00019407737696393215,
+      "loss": 2.7375,
+      "step": 2291
+    },
+    {
+      "epoch": 0.1849729642482447,
+      "grad_norm": 0.8154759407043457,
+      "learning_rate": 0.00019407202348113904,
+      "loss": 2.7608,
+      "step": 2292
+    },
+    {
+      "epoch": 0.1850536679848277,
+      "grad_norm": 0.7912892699241638,
+      "learning_rate": 0.0001940666676538252,
+      "loss": 2.7886,
+      "step": 2293
+    },
+    {
+      "epoch": 0.1851343717214107,
+      "grad_norm": 0.9184576272964478,
+      "learning_rate": 0.0001940613094821241,
+      "loss": 2.7867,
+      "step": 2294
+    },
+    {
+      "epoch": 0.1852150754579937,
+      "grad_norm": 0.8114588856697083,
+      "learning_rate": 0.0001940559489661693,
+      "loss": 2.8105,
+      "step": 2295
+    },
+    {
+      "epoch": 0.1852957791945767,
+      "grad_norm": 0.7681595683097839,
+      "learning_rate": 0.00019405058610609438,
+      "loss": 2.7707,
+      "step": 2296
+    },
+    {
+      "epoch": 0.18537648293115971,
+      "grad_norm": 0.7719643712043762,
+      "learning_rate": 0.000194045220902033,
+      "loss": 2.6767,
+      "step": 2297
+    },
+    {
+      "epoch": 0.18545718666774272,
+      "grad_norm": 0.7602487206459045,
+      "learning_rate": 0.00019403985335411888,
+      "loss": 2.7698,
+      "step": 2298
+    },
+    {
+      "epoch": 0.18553789040432572,
+      "grad_norm": 0.8044554591178894,
+      "learning_rate": 0.00019403448346248578,
+      "loss": 2.7578,
+      "step": 2299
+    },
+    {
+      "epoch": 0.18561859414090873,
+      "grad_norm": 0.7830328345298767,
+      "learning_rate": 0.00019402911122726757,
+      "loss": 2.7113,
+      "step": 2300
+    },
+    {
+      "epoch": 0.18569929787749173,
+      "grad_norm": 0.7793100476264954,
+      "learning_rate": 0.0001940237366485981,
+      "loss": 2.7388,
+      "step": 2301
+    },
+    {
+      "epoch": 0.18578000161407474,
+      "grad_norm": 0.9127374887466431,
+      "learning_rate": 0.00019401835972661133,
+      "loss": 2.7459,
+      "step": 2302
+    },
+    {
+      "epoch": 0.18586070535065774,
+      "grad_norm": 0.8007177114486694,
+      "learning_rate": 0.00019401298046144128,
+      "loss": 2.776,
+      "step": 2303
+    },
+    {
+      "epoch": 0.18594140908724074,
+      "grad_norm": 0.7384614944458008,
+      "learning_rate": 0.000194007598853222,
+      "loss": 2.6819,
+      "step": 2304
+    },
+    {
+      "epoch": 0.18602211282382375,
+      "grad_norm": 0.798909068107605,
+      "learning_rate": 0.0001940022149020876,
+      "loss": 2.7218,
+      "step": 2305
+    },
+    {
+      "epoch": 0.18610281656040675,
+      "grad_norm": 0.8388963341712952,
+      "learning_rate": 0.0001939968286081723,
+      "loss": 2.8248,
+      "step": 2306
+    },
+    {
+      "epoch": 0.18618352029698976,
+      "grad_norm": 0.8411754369735718,
+      "learning_rate": 0.0001939914399716103,
+      "loss": 2.7575,
+      "step": 2307
+    },
+    {
+      "epoch": 0.18626422403357276,
+      "grad_norm": 0.7936103343963623,
+      "learning_rate": 0.00019398604899253594,
+      "loss": 2.7488,
+      "step": 2308
+    },
+    {
+      "epoch": 0.18634492777015577,
+      "grad_norm": 0.7913734912872314,
+      "learning_rate": 0.00019398065567108357,
+      "loss": 2.7963,
+      "step": 2309
+    },
+    {
+      "epoch": 0.18642563150673877,
+      "grad_norm": 0.8341575860977173,
+      "learning_rate": 0.00019397526000738754,
+      "loss": 2.7698,
+      "step": 2310
+    },
+    {
+      "epoch": 0.18650633524332177,
+      "grad_norm": 0.8323128819465637,
+      "learning_rate": 0.00019396986200158244,
+      "loss": 2.7218,
+      "step": 2311
+    },
+    {
+      "epoch": 0.18658703897990478,
+      "grad_norm": 0.748073160648346,
+      "learning_rate": 0.0001939644616538027,
+      "loss": 2.7798,
+      "step": 2312
+    },
+    {
+      "epoch": 0.18666774271648778,
+      "grad_norm": 0.8166958689689636,
+      "learning_rate": 0.00019395905896418296,
+      "loss": 2.661,
+      "step": 2313
+    },
+    {
+      "epoch": 0.1867484464530708,
+      "grad_norm": 0.796791672706604,
+      "learning_rate": 0.00019395365393285786,
+      "loss": 2.7297,
+      "step": 2314
+    },
+    {
+      "epoch": 0.1868291501896538,
+      "grad_norm": 0.7851170897483826,
+      "learning_rate": 0.0001939482465599621,
+      "loss": 2.7798,
+      "step": 2315
+    },
+    {
+      "epoch": 0.1869098539262368,
+      "grad_norm": 0.7545836567878723,
+      "learning_rate": 0.00019394283684563045,
+      "loss": 2.7327,
+      "step": 2316
+    },
+    {
+      "epoch": 0.1869905576628198,
+      "grad_norm": 0.8100360631942749,
+      "learning_rate": 0.00019393742478999776,
+      "loss": 2.7901,
+      "step": 2317
+    },
+    {
+      "epoch": 0.1870712613994028,
+      "grad_norm": 0.7874314785003662,
+      "learning_rate": 0.00019393201039319887,
+      "loss": 2.7597,
+      "step": 2318
+    },
+    {
+      "epoch": 0.1871519651359858,
+      "grad_norm": 0.7698730826377869,
+      "learning_rate": 0.00019392659365536876,
+      "loss": 2.7327,
+      "step": 2319
+    },
+    {
+      "epoch": 0.1872326688725688,
+      "grad_norm": 0.7417994141578674,
+      "learning_rate": 0.0001939211745766424,
+      "loss": 2.7413,
+      "step": 2320
+    },
+    {
+      "epoch": 0.1873133726091518,
+      "grad_norm": 0.7823258638381958,
+      "learning_rate": 0.00019391575315715485,
+      "loss": 2.7577,
+      "step": 2321
+    },
+    {
+      "epoch": 0.1873940763457348,
+      "grad_norm": 0.82382732629776,
+      "learning_rate": 0.00019391032939704124,
+      "loss": 2.7769,
+      "step": 2322
+    },
+    {
+      "epoch": 0.1874747800823178,
+      "grad_norm": 0.8405026197433472,
+      "learning_rate": 0.0001939049032964367,
+      "loss": 2.8402,
+      "step": 2323
+    },
+    {
+      "epoch": 0.1875554838189008,
+      "grad_norm": 0.8307906985282898,
+      "learning_rate": 0.00019389947485547654,
+      "loss": 2.7642,
+      "step": 2324
+    },
+    {
+      "epoch": 0.1876361875554838,
+      "grad_norm": 0.8618248701095581,
+      "learning_rate": 0.000193894044074296,
+      "loss": 2.7853,
+      "step": 2325
+    },
+    {
+      "epoch": 0.1877168912920668,
+      "grad_norm": 0.8040831685066223,
+      "learning_rate": 0.00019388861095303046,
+      "loss": 2.7467,
+      "step": 2326
+    },
+    {
+      "epoch": 0.18779759502864982,
+      "grad_norm": 0.7723637223243713,
+      "learning_rate": 0.0001938831754918153,
+      "loss": 2.7222,
+      "step": 2327
+    },
+    {
+      "epoch": 0.18787829876523282,
+      "grad_norm": 0.8189084529876709,
+      "learning_rate": 0.000193877737690786,
+      "loss": 2.7857,
+      "step": 2328
+    },
+    {
+      "epoch": 0.18795900250181583,
+      "grad_norm": 0.8335791826248169,
+      "learning_rate": 0.00019387229755007805,
+      "loss": 2.6997,
+      "step": 2329
+    },
+    {
+      "epoch": 0.18803970623839883,
+      "grad_norm": 0.7732782959938049,
+      "learning_rate": 0.00019386685506982707,
+      "loss": 2.7155,
+      "step": 2330
+    },
+    {
+      "epoch": 0.18812040997498183,
+      "grad_norm": 0.8262906670570374,
+      "learning_rate": 0.0001938614102501687,
+      "loss": 2.7638,
+      "step": 2331
+    },
+    {
+      "epoch": 0.18820111371156484,
+      "grad_norm": 0.7969058156013489,
+      "learning_rate": 0.00019385596309123862,
+      "loss": 2.7363,
+      "step": 2332
+    },
+    {
+      "epoch": 0.18828181744814784,
+      "grad_norm": 0.7834853529930115,
+      "learning_rate": 0.0001938505135931726,
+      "loss": 2.7205,
+      "step": 2333
+    },
+    {
+      "epoch": 0.18836252118473085,
+      "grad_norm": 0.748481810092926,
+      "learning_rate": 0.00019384506175610647,
+      "loss": 2.7759,
+      "step": 2334
+    },
+    {
+      "epoch": 0.18844322492131385,
+      "grad_norm": 0.8137786984443665,
+      "learning_rate": 0.00019383960758017604,
+      "loss": 2.828,
+      "step": 2335
+    },
+    {
+      "epoch": 0.18852392865789686,
+      "grad_norm": 0.8065745234489441,
+      "learning_rate": 0.00019383415106551734,
+      "loss": 2.7408,
+      "step": 2336
+    },
+    {
+      "epoch": 0.18860463239447986,
+      "grad_norm": 0.768643856048584,
+      "learning_rate": 0.0001938286922122663,
+      "loss": 2.6503,
+      "step": 2337
+    },
+    {
+      "epoch": 0.18868533613106286,
+      "grad_norm": 0.7677921652793884,
+      "learning_rate": 0.00019382323102055897,
+      "loss": 2.7088,
+      "step": 2338
+    },
+    {
+      "epoch": 0.18876603986764587,
+      "grad_norm": 0.7648717164993286,
+      "learning_rate": 0.0001938177674905315,
+      "loss": 2.7015,
+      "step": 2339
+    },
+    {
+      "epoch": 0.18884674360422887,
+      "grad_norm": 0.7517116665840149,
+      "learning_rate": 0.00019381230162231997,
+      "loss": 2.7095,
+      "step": 2340
+    },
+    {
+      "epoch": 0.18892744734081188,
+      "grad_norm": 0.8147841691970825,
+      "learning_rate": 0.00019380683341606067,
+      "loss": 2.8563,
+      "step": 2341
+    },
+    {
+      "epoch": 0.18900815107739488,
+      "grad_norm": 0.7849822640419006,
+      "learning_rate": 0.00019380136287188988,
+      "loss": 2.7432,
+      "step": 2342
+    },
+    {
+      "epoch": 0.18908885481397789,
+      "grad_norm": 0.813811719417572,
+      "learning_rate": 0.0001937958899899439,
+      "loss": 2.7419,
+      "step": 2343
+    },
+    {
+      "epoch": 0.1891695585505609,
+      "grad_norm": 0.8142707943916321,
+      "learning_rate": 0.00019379041477035923,
+      "loss": 2.7658,
+      "step": 2344
+    },
+    {
+      "epoch": 0.1892502622871439,
+      "grad_norm": 0.7594506740570068,
+      "learning_rate": 0.00019378493721327217,
+      "loss": 2.7298,
+      "step": 2345
+    },
+    {
+      "epoch": 0.1893309660237269,
+      "grad_norm": 0.8374232053756714,
+      "learning_rate": 0.00019377945731881936,
+      "loss": 2.8112,
+      "step": 2346
+    },
+    {
+      "epoch": 0.1894116697603099,
+      "grad_norm": 0.783608615398407,
+      "learning_rate": 0.00019377397508713734,
+      "loss": 2.8168,
+      "step": 2347
+    },
+    {
+      "epoch": 0.1894923734968929,
+      "grad_norm": 0.720214307308197,
+      "learning_rate": 0.0001937684905183627,
+      "loss": 2.7516,
+      "step": 2348
+    },
+    {
+      "epoch": 0.1895730772334759,
+      "grad_norm": 0.7939600944519043,
+      "learning_rate": 0.0001937630036126322,
+      "loss": 2.7609,
+      "step": 2349
+    },
+    {
+      "epoch": 0.18965378097005892,
+      "grad_norm": 0.787315309047699,
+      "learning_rate": 0.00019375751437008252,
+      "loss": 2.758,
+      "step": 2350
+    },
+    {
+      "epoch": 0.18973448470664192,
+      "grad_norm": 0.7862411141395569,
+      "learning_rate": 0.00019375202279085053,
+      "loss": 2.6866,
+      "step": 2351
+    },
+    {
+      "epoch": 0.18981518844322492,
+      "grad_norm": 0.8651136159896851,
+      "learning_rate": 0.000193746528875073,
+      "loss": 2.7488,
+      "step": 2352
+    },
+    {
+      "epoch": 0.18989589217980793,
+      "grad_norm": 0.8150602579116821,
+      "learning_rate": 0.00019374103262288696,
+      "loss": 2.7417,
+      "step": 2353
+    },
+    {
+      "epoch": 0.18997659591639093,
+      "grad_norm": 0.9053540229797363,
+      "learning_rate": 0.00019373553403442934,
+      "loss": 2.7587,
+      "step": 2354
+    },
+    {
+      "epoch": 0.19005729965297394,
+      "grad_norm": 0.8775703310966492,
+      "learning_rate": 0.0001937300331098372,
+      "loss": 2.733,
+      "step": 2355
+    },
+    {
+      "epoch": 0.19013800338955694,
+      "grad_norm": 0.7714357972145081,
+      "learning_rate": 0.0001937245298492476,
+      "loss": 2.7595,
+      "step": 2356
+    },
+    {
+      "epoch": 0.19021870712613995,
+      "grad_norm": 0.8648017048835754,
+      "learning_rate": 0.0001937190242527977,
+      "loss": 2.7944,
+      "step": 2357
+    },
+    {
+      "epoch": 0.19029941086272295,
+      "grad_norm": 0.9367388486862183,
+      "learning_rate": 0.00019371351632062477,
+      "loss": 2.7902,
+      "step": 2358
+    },
+    {
+      "epoch": 0.19038011459930596,
+      "grad_norm": 0.8116368651390076,
+      "learning_rate": 0.00019370800605286604,
+      "loss": 2.7291,
+      "step": 2359
+    },
+    {
+      "epoch": 0.19046081833588896,
+      "grad_norm": 0.7892753481864929,
+      "learning_rate": 0.00019370249344965882,
+      "loss": 2.8192,
+      "step": 2360
+    },
+    {
+      "epoch": 0.19054152207247196,
+      "grad_norm": 0.8109372854232788,
+      "learning_rate": 0.00019369697851114056,
+      "loss": 2.6982,
+      "step": 2361
+    },
+    {
+      "epoch": 0.19062222580905497,
+      "grad_norm": 0.8756314516067505,
+      "learning_rate": 0.00019369146123744864,
+      "loss": 2.744,
+      "step": 2362
+    },
+    {
+      "epoch": 0.19070292954563797,
+      "grad_norm": 0.7400399446487427,
+      "learning_rate": 0.00019368594162872058,
+      "loss": 2.7328,
+      "step": 2363
+    },
+    {
+      "epoch": 0.19078363328222098,
+      "grad_norm": 0.8223158717155457,
+      "learning_rate": 0.000193680419685094,
+      "loss": 2.7614,
+      "step": 2364
+    },
+    {
+      "epoch": 0.19086433701880398,
+      "grad_norm": 0.7350139617919922,
+      "learning_rate": 0.00019367489540670645,
+      "loss": 2.7074,
+      "step": 2365
+    },
+    {
+      "epoch": 0.19094504075538699,
+      "grad_norm": 0.7915631532669067,
+      "learning_rate": 0.00019366936879369563,
+      "loss": 2.7835,
+      "step": 2366
+    },
+    {
+      "epoch": 0.19102574449197,
+      "grad_norm": 0.7765628099441528,
+      "learning_rate": 0.00019366383984619932,
+      "loss": 2.765,
+      "step": 2367
+    },
+    {
+      "epoch": 0.191106448228553,
+      "grad_norm": 0.8127059936523438,
+      "learning_rate": 0.00019365830856435525,
+      "loss": 2.7753,
+      "step": 2368
+    },
+    {
+      "epoch": 0.191187151965136,
+      "grad_norm": 0.8652897477149963,
+      "learning_rate": 0.0001936527749483013,
+      "loss": 2.7137,
+      "step": 2369
+    },
+    {
+      "epoch": 0.191267855701719,
+      "grad_norm": 0.8086774945259094,
+      "learning_rate": 0.00019364723899817541,
+      "loss": 2.7209,
+      "step": 2370
+    },
+    {
+      "epoch": 0.191348559438302,
+      "grad_norm": 0.7965098023414612,
+      "learning_rate": 0.00019364170071411554,
+      "loss": 2.786,
+      "step": 2371
+    },
+    {
+      "epoch": 0.19142926317488498,
+      "grad_norm": 0.7954064607620239,
+      "learning_rate": 0.00019363616009625967,
+      "loss": 2.7508,
+      "step": 2372
+    },
+    {
+      "epoch": 0.191509966911468,
+      "grad_norm": 0.7835928201675415,
+      "learning_rate": 0.00019363061714474595,
+      "loss": 2.7423,
+      "step": 2373
+    },
+    {
+      "epoch": 0.191590670648051,
+      "grad_norm": 0.8720580339431763,
+      "learning_rate": 0.0001936250718597125,
+      "loss": 2.7877,
+      "step": 2374
+    },
+    {
+      "epoch": 0.191671374384634,
+      "grad_norm": 0.836066484451294,
+      "learning_rate": 0.00019361952424129747,
+      "loss": 2.8456,
+      "step": 2375
+    },
+    {
+      "epoch": 0.191752078121217,
+      "grad_norm": 0.793666660785675,
+      "learning_rate": 0.00019361397428963923,
+      "loss": 2.786,
+      "step": 2376
+    },
+    {
+      "epoch": 0.1918327818578,
+      "grad_norm": 0.8573217391967773,
+      "learning_rate": 0.000193608422004876,
+      "loss": 2.7569,
+      "step": 2377
+    },
+    {
+      "epoch": 0.191913485594383,
+      "grad_norm": 0.81243896484375,
+      "learning_rate": 0.00019360286738714623,
+      "loss": 2.771,
+      "step": 2378
+    },
+    {
+      "epoch": 0.19199418933096601,
+      "grad_norm": 0.7449626326560974,
+      "learning_rate": 0.00019359731043658832,
+      "loss": 2.7479,
+      "step": 2379
+    },
+    {
+      "epoch": 0.19207489306754902,
+      "grad_norm": 0.8124165534973145,
+      "learning_rate": 0.00019359175115334076,
+      "loss": 2.7602,
+      "step": 2380
+    },
+    {
+      "epoch": 0.19215559680413202,
+      "grad_norm": 0.7786986827850342,
+      "learning_rate": 0.00019358618953754211,
+      "loss": 2.6926,
+      "step": 2381
+    },
+    {
+      "epoch": 0.19223630054071503,
+      "grad_norm": 0.7987258434295654,
+      "learning_rate": 0.000193580625589331,
+      "loss": 2.7573,
+      "step": 2382
+    },
+    {
+      "epoch": 0.19231700427729803,
+      "grad_norm": 0.8236463665962219,
+      "learning_rate": 0.00019357505930884606,
+      "loss": 2.6755,
+      "step": 2383
+    },
+    {
+      "epoch": 0.19239770801388104,
+      "grad_norm": 0.8285779356956482,
+      "learning_rate": 0.00019356949069622602,
+      "loss": 2.7658,
+      "step": 2384
+    },
+    {
+      "epoch": 0.19247841175046404,
+      "grad_norm": 0.7823960781097412,
+      "learning_rate": 0.0001935639197516097,
+      "loss": 2.7404,
+      "step": 2385
+    },
+    {
+      "epoch": 0.19255911548704704,
+      "grad_norm": 0.968638002872467,
+      "learning_rate": 0.00019355834647513591,
+      "loss": 2.7836,
+      "step": 2386
+    },
+    {
+      "epoch": 0.19263981922363005,
+      "grad_norm": 0.8170328736305237,
+      "learning_rate": 0.00019355277086694357,
+      "loss": 2.7816,
+      "step": 2387
+    },
+    {
+      "epoch": 0.19272052296021305,
+      "grad_norm": 0.8342583179473877,
+      "learning_rate": 0.00019354719292717163,
+      "loss": 2.8204,
+      "step": 2388
+    },
+    {
+      "epoch": 0.19280122669679606,
+      "grad_norm": 0.8160435557365417,
+      "learning_rate": 0.0001935416126559591,
+      "loss": 2.6938,
+      "step": 2389
+    },
+    {
+      "epoch": 0.19288193043337906,
+      "grad_norm": 0.7888174653053284,
+      "learning_rate": 0.00019353603005344504,
+      "loss": 2.6804,
+      "step": 2390
+    },
+    {
+      "epoch": 0.19296263416996207,
+      "grad_norm": 0.8389205932617188,
+      "learning_rate": 0.00019353044511976865,
+      "loss": 2.7571,
+      "step": 2391
+    },
+    {
+      "epoch": 0.19304333790654507,
+      "grad_norm": 0.7920562028884888,
+      "learning_rate": 0.00019352485785506906,
+      "loss": 2.7174,
+      "step": 2392
+    },
+    {
+      "epoch": 0.19312404164312807,
+      "grad_norm": 0.7853459715843201,
+      "learning_rate": 0.00019351926825948555,
+      "loss": 2.7626,
+      "step": 2393
+    },
+    {
+      "epoch": 0.19320474537971108,
+      "grad_norm": 0.9109459519386292,
+      "learning_rate": 0.0001935136763331574,
+      "loss": 2.7568,
+      "step": 2394
+    },
+    {
+      "epoch": 0.19328544911629408,
+      "grad_norm": 0.7983853816986084,
+      "learning_rate": 0.00019350808207622397,
+      "loss": 2.7412,
+      "step": 2395
+    },
+    {
+      "epoch": 0.1933661528528771,
+      "grad_norm": 0.7416854500770569,
+      "learning_rate": 0.00019350248548882472,
+      "loss": 2.7335,
+      "step": 2396
+    },
+    {
+      "epoch": 0.1934468565894601,
+      "grad_norm": 0.7305171489715576,
+      "learning_rate": 0.0001934968865710991,
+      "loss": 2.7295,
+      "step": 2397
+    },
+    {
+      "epoch": 0.1935275603260431,
+      "grad_norm": 0.7717033624649048,
+      "learning_rate": 0.0001934912853231867,
+      "loss": 2.7568,
+      "step": 2398
+    },
+    {
+      "epoch": 0.1936082640626261,
+      "grad_norm": 0.7833831906318665,
+      "learning_rate": 0.00019348568174522705,
+      "loss": 2.736,
+      "step": 2399
+    },
+    {
+      "epoch": 0.1936889677992091,
+      "grad_norm": 0.872831404209137,
+      "learning_rate": 0.00019348007583735983,
+      "loss": 2.7719,
+      "step": 2400
+    },
+    {
+      "epoch": 0.1937696715357921,
+      "grad_norm": 0.8389193415641785,
+      "learning_rate": 0.0001934744675997248,
+      "loss": 2.7572,
+      "step": 2401
+    },
+    {
+      "epoch": 0.19385037527237511,
+      "grad_norm": 0.8442249298095703,
+      "learning_rate": 0.00019346885703246165,
+      "loss": 2.8117,
+      "step": 2402
+    },
+    {
+      "epoch": 0.19393107900895812,
+      "grad_norm": 0.8451170325279236,
+      "learning_rate": 0.00019346324413571027,
+      "loss": 2.7216,
+      "step": 2403
+    },
+    {
+      "epoch": 0.19401178274554112,
+      "grad_norm": 0.898529052734375,
+      "learning_rate": 0.00019345762890961052,
+      "loss": 2.8119,
+      "step": 2404
+    },
+    {
+      "epoch": 0.19409248648212413,
+      "grad_norm": 0.8302313685417175,
+      "learning_rate": 0.00019345201135430236,
+      "loss": 2.76,
+      "step": 2405
+    },
+    {
+      "epoch": 0.19417319021870713,
+      "grad_norm": 0.8975207209587097,
+      "learning_rate": 0.00019344639146992582,
+      "loss": 2.8043,
+      "step": 2406
+    },
+    {
+      "epoch": 0.19425389395529014,
+      "grad_norm": 0.8972581028938293,
+      "learning_rate": 0.0001934407692566209,
+      "loss": 2.7487,
+      "step": 2407
+    },
+    {
+      "epoch": 0.19433459769187314,
+      "grad_norm": 0.8311447501182556,
+      "learning_rate": 0.00019343514471452776,
+      "loss": 2.7653,
+      "step": 2408
+    },
+    {
+      "epoch": 0.19441530142845614,
+      "grad_norm": 0.8336243033409119,
+      "learning_rate": 0.0001934295178437866,
+      "loss": 2.753,
+      "step": 2409
+    },
+    {
+      "epoch": 0.19449600516503915,
+      "grad_norm": 0.8339207172393799,
+      "learning_rate": 0.0001934238886445376,
+      "loss": 2.7643,
+      "step": 2410
+    },
+    {
+      "epoch": 0.19457670890162215,
+      "grad_norm": 0.906074583530426,
+      "learning_rate": 0.0001934182571169211,
+      "loss": 2.7777,
+      "step": 2411
+    },
+    {
+      "epoch": 0.19465741263820516,
+      "grad_norm": 0.8759943246841431,
+      "learning_rate": 0.00019341262326107742,
+      "loss": 2.77,
+      "step": 2412
+    },
+    {
+      "epoch": 0.19473811637478816,
+      "grad_norm": 0.8399369716644287,
+      "learning_rate": 0.00019340698707714699,
+      "loss": 2.752,
+      "step": 2413
+    },
+    {
+      "epoch": 0.19481882011137117,
+      "grad_norm": 0.8551808595657349,
+      "learning_rate": 0.00019340134856527026,
+      "loss": 2.6727,
+      "step": 2414
+    },
+    {
+      "epoch": 0.19489952384795417,
+      "grad_norm": 0.7660732865333557,
+      "learning_rate": 0.00019339570772558778,
+      "loss": 2.7491,
+      "step": 2415
+    },
+    {
+      "epoch": 0.19498022758453717,
+      "grad_norm": 0.8257685303688049,
+      "learning_rate": 0.00019339006455824015,
+      "loss": 2.7584,
+      "step": 2416
+    },
+    {
+      "epoch": 0.19506093132112018,
+      "grad_norm": 0.797275960445404,
+      "learning_rate": 0.00019338441906336794,
+      "loss": 2.7051,
+      "step": 2417
+    },
+    {
+      "epoch": 0.19514163505770318,
+      "grad_norm": 0.8311913013458252,
+      "learning_rate": 0.00019337877124111193,
+      "loss": 2.8084,
+      "step": 2418
+    },
+    {
+      "epoch": 0.1952223387942862,
+      "grad_norm": 0.7995893359184265,
+      "learning_rate": 0.0001933731210916128,
+      "loss": 2.7556,
+      "step": 2419
+    },
+    {
+      "epoch": 0.1953030425308692,
+      "grad_norm": 0.792850136756897,
+      "learning_rate": 0.00019336746861501147,
+      "loss": 2.7289,
+      "step": 2420
+    },
+    {
+      "epoch": 0.1953837462674522,
+      "grad_norm": 0.8058848977088928,
+      "learning_rate": 0.00019336181381144873,
+      "loss": 2.7394,
+      "step": 2421
+    },
+    {
+      "epoch": 0.1954644500040352,
+      "grad_norm": 0.8267124891281128,
+      "learning_rate": 0.00019335615668106555,
+      "loss": 2.771,
+      "step": 2422
+    },
+    {
+      "epoch": 0.19554515374061818,
+      "grad_norm": 0.7641060948371887,
+      "learning_rate": 0.00019335049722400292,
+      "loss": 2.7311,
+      "step": 2423
+    },
+    {
+      "epoch": 0.19562585747720118,
+      "grad_norm": 0.8023245930671692,
+      "learning_rate": 0.00019334483544040186,
+      "loss": 2.7658,
+      "step": 2424
+    },
+    {
+      "epoch": 0.19570656121378419,
+      "grad_norm": 0.8341927528381348,
+      "learning_rate": 0.00019333917133040348,
+      "loss": 2.7476,
+      "step": 2425
+    },
+    {
+      "epoch": 0.1957872649503672,
+      "grad_norm": 0.7985726594924927,
+      "learning_rate": 0.000193333504894149,
+      "loss": 2.7362,
+      "step": 2426
+    },
+    {
+      "epoch": 0.1958679686869502,
+      "grad_norm": 0.7267594933509827,
+      "learning_rate": 0.0001933278361317796,
+      "loss": 2.6875,
+      "step": 2427
+    },
+    {
+      "epoch": 0.1959486724235332,
+      "grad_norm": 0.8292990326881409,
+      "learning_rate": 0.00019332216504343652,
+      "loss": 2.7619,
+      "step": 2428
+    },
+    {
+      "epoch": 0.1960293761601162,
+      "grad_norm": 0.7549588680267334,
+      "learning_rate": 0.00019331649162926116,
+      "loss": 2.7385,
+      "step": 2429
+    },
+    {
+      "epoch": 0.1961100798966992,
+      "grad_norm": 0.7688446640968323,
+      "learning_rate": 0.0001933108158893949,
+      "loss": 2.7544,
+      "step": 2430
+    },
+    {
+      "epoch": 0.1961907836332822,
+      "grad_norm": 0.8168436884880066,
+      "learning_rate": 0.00019330513782397918,
+      "loss": 2.8013,
+      "step": 2431
+    },
+    {
+      "epoch": 0.19627148736986522,
+      "grad_norm": 0.8405759334564209,
+      "learning_rate": 0.00019329945743315556,
+      "loss": 2.7299,
+      "step": 2432
+    },
+    {
+      "epoch": 0.19635219110644822,
+      "grad_norm": 0.79430091381073,
+      "learning_rate": 0.00019329377471706554,
+      "loss": 2.7293,
+      "step": 2433
+    },
+    {
+      "epoch": 0.19643289484303122,
+      "grad_norm": 0.8428656458854675,
+      "learning_rate": 0.0001932880896758508,
+      "loss": 2.8211,
+      "step": 2434
+    },
+    {
+      "epoch": 0.19651359857961423,
+      "grad_norm": 0.7883139252662659,
+      "learning_rate": 0.00019328240230965298,
+      "loss": 2.6943,
+      "step": 2435
+    },
+    {
+      "epoch": 0.19659430231619723,
+      "grad_norm": 0.7539335489273071,
+      "learning_rate": 0.00019327671261861387,
+      "loss": 2.6926,
+      "step": 2436
+    },
+    {
+      "epoch": 0.19667500605278024,
+      "grad_norm": 0.9986057281494141,
+      "learning_rate": 0.00019327102060287524,
+      "loss": 2.7851,
+      "step": 2437
+    },
+    {
+      "epoch": 0.19675570978936324,
+      "grad_norm": 0.7716113924980164,
+      "learning_rate": 0.000193265326262579,
+      "loss": 2.752,
+      "step": 2438
+    },
+    {
+      "epoch": 0.19683641352594625,
+      "grad_norm": 0.9134296774864197,
+      "learning_rate": 0.000193259629597867,
+      "loss": 2.7698,
+      "step": 2439
+    },
+    {
+      "epoch": 0.19691711726252925,
+      "grad_norm": 0.7966345548629761,
+      "learning_rate": 0.00019325393060888124,
+      "loss": 2.7839,
+      "step": 2440
+    },
+    {
+      "epoch": 0.19699782099911226,
+      "grad_norm": 0.8051251173019409,
+      "learning_rate": 0.0001932482292957638,
+      "loss": 2.7322,
+      "step": 2441
+    },
+    {
+      "epoch": 0.19707852473569526,
+      "grad_norm": 0.843169629573822,
+      "learning_rate": 0.0001932425256586567,
+      "loss": 2.8263,
+      "step": 2442
+    },
+    {
+      "epoch": 0.19715922847227826,
+      "grad_norm": 0.7552370429039001,
+      "learning_rate": 0.00019323681969770213,
+      "loss": 2.7342,
+      "step": 2443
+    },
+    {
+      "epoch": 0.19723993220886127,
+      "grad_norm": 0.844473123550415,
+      "learning_rate": 0.0001932311114130423,
+      "loss": 2.776,
+      "step": 2444
+    },
+    {
+      "epoch": 0.19732063594544427,
+      "grad_norm": 0.8002473711967468,
+      "learning_rate": 0.00019322540080481945,
+      "loss": 2.7382,
+      "step": 2445
+    },
+    {
+      "epoch": 0.19740133968202728,
+      "grad_norm": 0.8564329147338867,
+      "learning_rate": 0.00019321968787317594,
+      "loss": 2.7592,
+      "step": 2446
+    },
+    {
+      "epoch": 0.19748204341861028,
+      "grad_norm": 0.7853825688362122,
+      "learning_rate": 0.00019321397261825408,
+      "loss": 2.7101,
+      "step": 2447
+    },
+    {
+      "epoch": 0.19756274715519329,
+      "grad_norm": 0.8482939004898071,
+      "learning_rate": 0.0001932082550401964,
+      "loss": 2.7891,
+      "step": 2448
+    },
+    {
+      "epoch": 0.1976434508917763,
+      "grad_norm": 0.8361770510673523,
+      "learning_rate": 0.00019320253513914536,
+      "loss": 2.7341,
+      "step": 2449
+    },
+    {
+      "epoch": 0.1977241546283593,
+      "grad_norm": 0.7814618945121765,
+      "learning_rate": 0.0001931968129152435,
+      "loss": 2.771,
+      "step": 2450
+    },
+    {
+      "epoch": 0.1978048583649423,
+      "grad_norm": 0.7588146924972534,
+      "learning_rate": 0.00019319108836863343,
+      "loss": 2.7577,
+      "step": 2451
+    },
+    {
+      "epoch": 0.1978855621015253,
+      "grad_norm": 0.9184895157814026,
+      "learning_rate": 0.00019318536149945785,
+      "loss": 2.7711,
+      "step": 2452
+    },
+    {
+      "epoch": 0.1979662658381083,
+      "grad_norm": 0.8454298973083496,
+      "learning_rate": 0.00019317963230785947,
+      "loss": 2.7748,
+      "step": 2453
+    },
+    {
+      "epoch": 0.1980469695746913,
+      "grad_norm": 0.7662420868873596,
+      "learning_rate": 0.0001931739007939811,
+      "loss": 2.7704,
+      "step": 2454
+    },
+    {
+      "epoch": 0.19812767331127432,
+      "grad_norm": 0.837888777256012,
+      "learning_rate": 0.0001931681669579655,
+      "loss": 2.7613,
+      "step": 2455
+    },
+    {
+      "epoch": 0.19820837704785732,
+      "grad_norm": 0.7835226058959961,
+      "learning_rate": 0.0001931624307999557,
+      "loss": 2.6888,
+      "step": 2456
+    },
+    {
+      "epoch": 0.19828908078444032,
+      "grad_norm": 0.8491464257240295,
+      "learning_rate": 0.00019315669232009456,
+      "loss": 2.7521,
+      "step": 2457
+    },
+    {
+      "epoch": 0.19836978452102333,
+      "grad_norm": 0.7590088248252869,
+      "learning_rate": 0.00019315095151852516,
+      "loss": 2.7441,
+      "step": 2458
+    },
+    {
+      "epoch": 0.19845048825760633,
+      "grad_norm": 0.9316127300262451,
+      "learning_rate": 0.00019314520839539052,
+      "loss": 2.786,
+      "step": 2459
+    },
+    {
+      "epoch": 0.19853119199418934,
+      "grad_norm": 0.7819615006446838,
+      "learning_rate": 0.0001931394629508338,
+      "loss": 2.7003,
+      "step": 2460
+    },
+    {
+      "epoch": 0.19861189573077234,
+      "grad_norm": 0.7675932049751282,
+      "learning_rate": 0.0001931337151849982,
+      "loss": 2.7065,
+      "step": 2461
+    },
+    {
+      "epoch": 0.19869259946735535,
+      "grad_norm": 0.7797678112983704,
+      "learning_rate": 0.000193127965098027,
+      "loss": 2.7605,
+      "step": 2462
+    },
+    {
+      "epoch": 0.19877330320393835,
+      "grad_norm": 0.789544403553009,
+      "learning_rate": 0.00019312221269006345,
+      "loss": 2.7913,
+      "step": 2463
+    },
+    {
+      "epoch": 0.19885400694052136,
+      "grad_norm": 0.9594957232475281,
+      "learning_rate": 0.00019311645796125094,
+      "loss": 2.785,
+      "step": 2464
+    },
+    {
+      "epoch": 0.19893471067710436,
+      "grad_norm": 0.8154739141464233,
+      "learning_rate": 0.00019311070091173287,
+      "loss": 2.6716,
+      "step": 2465
+    },
+    {
+      "epoch": 0.19901541441368736,
+      "grad_norm": 0.9042142629623413,
+      "learning_rate": 0.00019310494154165274,
+      "loss": 2.734,
+      "step": 2466
+    },
+    {
+      "epoch": 0.19909611815027037,
+      "grad_norm": 0.7803483605384827,
+      "learning_rate": 0.0001930991798511541,
+      "loss": 2.7052,
+      "step": 2467
+    },
+    {
+      "epoch": 0.19917682188685337,
+      "grad_norm": 0.7917614579200745,
+      "learning_rate": 0.00019309341584038055,
+      "loss": 2.728,
+      "step": 2468
+    },
+    {
+      "epoch": 0.19925752562343638,
+      "grad_norm": 0.8295063376426697,
+      "learning_rate": 0.00019308764950947568,
+      "loss": 2.7496,
+      "step": 2469
+    },
+    {
+      "epoch": 0.19933822936001938,
+      "grad_norm": 0.790831983089447,
+      "learning_rate": 0.0001930818808585833,
+      "loss": 2.7356,
+      "step": 2470
+    },
+    {
+      "epoch": 0.19941893309660239,
+      "grad_norm": 0.8527843952178955,
+      "learning_rate": 0.0001930761098878471,
+      "loss": 2.718,
+      "step": 2471
+    },
+    {
+      "epoch": 0.1994996368331854,
+      "grad_norm": 0.8518494367599487,
+      "learning_rate": 0.00019307033659741096,
+      "loss": 2.7189,
+      "step": 2472
+    },
+    {
+      "epoch": 0.1995803405697684,
+      "grad_norm": 0.8027220368385315,
+      "learning_rate": 0.00019306456098741872,
+      "loss": 2.7272,
+      "step": 2473
+    },
+    {
+      "epoch": 0.19966104430635137,
+      "grad_norm": 0.7516468167304993,
+      "learning_rate": 0.00019305878305801434,
+      "loss": 2.798,
+      "step": 2474
+    },
+    {
+      "epoch": 0.19974174804293438,
+      "grad_norm": 0.7676397562026978,
+      "learning_rate": 0.00019305300280934187,
+      "loss": 2.8076,
+      "step": 2475
+    },
+    {
+      "epoch": 0.19982245177951738,
+      "grad_norm": 0.8237762451171875,
+      "learning_rate": 0.00019304722024154528,
+      "loss": 2.6998,
+      "step": 2476
+    },
+    {
+      "epoch": 0.19990315551610038,
+      "grad_norm": 0.8397759199142456,
+      "learning_rate": 0.0001930414353547688,
+      "loss": 2.806,
+      "step": 2477
+    },
+    {
+      "epoch": 0.1999838592526834,
+      "grad_norm": 0.8911117911338806,
+      "learning_rate": 0.00019303564814915645,
+      "loss": 2.7566,
+      "step": 2478
+    },
+    {
+      "epoch": 0.2000645629892664,
+      "grad_norm": 0.765404999256134,
+      "learning_rate": 0.00019302985862485264,
+      "loss": 2.7363,
+      "step": 2479
+    },
+    {
+      "epoch": 0.2001452667258494,
+      "grad_norm": 0.7898589372634888,
+      "learning_rate": 0.0001930240667820015,
+      "loss": 2.7007,
+      "step": 2480
+    },
+    {
+      "epoch": 0.2002259704624324,
+      "grad_norm": 0.7581521272659302,
+      "learning_rate": 0.0001930182726207475,
+      "loss": 2.7508,
+      "step": 2481
+    },
+    {
+      "epoch": 0.2003066741990154,
+      "grad_norm": 0.8179795742034912,
+      "learning_rate": 0.00019301247614123495,
+      "loss": 2.7327,
+      "step": 2482
+    },
+    {
+      "epoch": 0.2003873779355984,
+      "grad_norm": 0.8103611469268799,
+      "learning_rate": 0.00019300667734360838,
+      "loss": 2.7869,
+      "step": 2483
+    },
+    {
+      "epoch": 0.20046808167218141,
+      "grad_norm": 0.7368054389953613,
+      "learning_rate": 0.0001930008762280123,
+      "loss": 2.73,
+      "step": 2484
+    },
+    {
+      "epoch": 0.20054878540876442,
+      "grad_norm": 0.7679662108421326,
+      "learning_rate": 0.00019299507279459127,
+      "loss": 2.7905,
+      "step": 2485
+    },
+    {
+      "epoch": 0.20062948914534742,
+      "grad_norm": 0.7783839702606201,
+      "learning_rate": 0.0001929892670434899,
+      "loss": 2.6816,
+      "step": 2486
+    },
+    {
+      "epoch": 0.20071019288193043,
+      "grad_norm": 0.7575809359550476,
+      "learning_rate": 0.00019298345897485298,
+      "loss": 2.7351,
+      "step": 2487
+    },
+    {
+      "epoch": 0.20079089661851343,
+      "grad_norm": 0.7674959301948547,
+      "learning_rate": 0.00019297764858882514,
+      "loss": 2.7682,
+      "step": 2488
+    },
+    {
+      "epoch": 0.20087160035509644,
+      "grad_norm": 0.7972592115402222,
+      "learning_rate": 0.00019297183588555127,
+      "loss": 2.782,
+      "step": 2489
+    },
+    {
+      "epoch": 0.20095230409167944,
+      "grad_norm": 0.8417105674743652,
+      "learning_rate": 0.00019296602086517624,
+      "loss": 2.8173,
+      "step": 2490
+    },
+    {
+      "epoch": 0.20103300782826244,
+      "grad_norm": 0.7194239497184753,
+      "learning_rate": 0.00019296020352784496,
+      "loss": 2.7735,
+      "step": 2491
+    },
+    {
+      "epoch": 0.20111371156484545,
+      "grad_norm": 0.801895022392273,
+      "learning_rate": 0.00019295438387370237,
+      "loss": 2.7018,
+      "step": 2492
+    },
+    {
+      "epoch": 0.20119441530142845,
+      "grad_norm": 0.900943398475647,
+      "learning_rate": 0.0001929485619028936,
+      "loss": 2.77,
+      "step": 2493
+    },
+    {
+      "epoch": 0.20127511903801146,
+      "grad_norm": 0.7882106304168701,
+      "learning_rate": 0.00019294273761556366,
+      "loss": 2.7195,
+      "step": 2494
+    },
+    {
+      "epoch": 0.20135582277459446,
+      "grad_norm": 0.7471950054168701,
+      "learning_rate": 0.00019293691101185775,
+      "loss": 2.7346,
+      "step": 2495
+    },
+    {
+      "epoch": 0.20143652651117747,
+      "grad_norm": 0.7498352527618408,
+      "learning_rate": 0.00019293108209192104,
+      "loss": 2.7255,
+      "step": 2496
+    },
+    {
+      "epoch": 0.20151723024776047,
+      "grad_norm": 0.8233164548873901,
+      "learning_rate": 0.0001929252508558989,
+      "loss": 2.8253,
+      "step": 2497
+    },
+    {
+      "epoch": 0.20159793398434347,
+      "grad_norm": 0.7533289790153503,
+      "learning_rate": 0.00019291941730393658,
+      "loss": 2.7487,
+      "step": 2498
+    },
+    {
+      "epoch": 0.20167863772092648,
+      "grad_norm": 0.7372691035270691,
+      "learning_rate": 0.0001929135814361795,
+      "loss": 2.6799,
+      "step": 2499
+    },
+    {
+      "epoch": 0.20175934145750948,
+      "grad_norm": 0.7760890126228333,
+      "learning_rate": 0.00019290774325277305,
+      "loss": 2.8366,
+      "step": 2500
+    },
+    {
+      "epoch": 0.2018400451940925,
+      "grad_norm": 0.7653746008872986,
+      "learning_rate": 0.0001929019027538628,
+      "loss": 2.7413,
+      "step": 2501
+    },
+    {
+      "epoch": 0.2019207489306755,
+      "grad_norm": 0.7364951372146606,
+      "learning_rate": 0.0001928960599395943,
+      "loss": 2.7405,
+      "step": 2502
+    },
+    {
+      "epoch": 0.2020014526672585,
+      "grad_norm": 0.8317872285842896,
+      "learning_rate": 0.00019289021481011314,
+      "loss": 2.7186,
+      "step": 2503
+    },
+    {
+      "epoch": 0.2020821564038415,
+      "grad_norm": 0.8325691223144531,
+      "learning_rate": 0.00019288436736556502,
+      "loss": 2.7305,
+      "step": 2504
+    },
+    {
+      "epoch": 0.2021628601404245,
+      "grad_norm": 0.7674683332443237,
+      "learning_rate": 0.00019287851760609566,
+      "loss": 2.7171,
+      "step": 2505
+    },
+    {
+      "epoch": 0.2022435638770075,
+      "grad_norm": 0.8043155074119568,
+      "learning_rate": 0.00019287266553185084,
+      "loss": 2.7425,
+      "step": 2506
+    },
+    {
+      "epoch": 0.2023242676135905,
+      "grad_norm": 0.8522058725357056,
+      "learning_rate": 0.00019286681114297642,
+      "loss": 2.7764,
+      "step": 2507
+    },
+    {
+      "epoch": 0.20240497135017352,
+      "grad_norm": 0.7700086236000061,
+      "learning_rate": 0.00019286095443961832,
+      "loss": 2.7499,
+      "step": 2508
+    },
+    {
+      "epoch": 0.20248567508675652,
+      "grad_norm": 0.8078013062477112,
+      "learning_rate": 0.0001928550954219225,
+      "loss": 2.7863,
+      "step": 2509
+    },
+    {
+      "epoch": 0.20256637882333953,
+      "grad_norm": 0.7431712746620178,
+      "learning_rate": 0.00019284923409003496,
+      "loss": 2.8296,
+      "step": 2510
+    },
+    {
+      "epoch": 0.20264708255992253,
+      "grad_norm": 0.753754734992981,
+      "learning_rate": 0.00019284337044410182,
+      "loss": 2.722,
+      "step": 2511
+    },
+    {
+      "epoch": 0.20272778629650554,
+      "grad_norm": 0.8117631077766418,
+      "learning_rate": 0.00019283750448426918,
+      "loss": 2.7718,
+      "step": 2512
+    },
+    {
+      "epoch": 0.20280849003308854,
+      "grad_norm": 0.9149020910263062,
+      "learning_rate": 0.00019283163621068325,
+      "loss": 2.7416,
+      "step": 2513
+    },
+    {
+      "epoch": 0.20288919376967154,
+      "grad_norm": 0.8240262866020203,
+      "learning_rate": 0.0001928257656234903,
+      "loss": 2.811,
+      "step": 2514
+    },
+    {
+      "epoch": 0.20296989750625455,
+      "grad_norm": 0.7394035458564758,
+      "learning_rate": 0.00019281989272283657,
+      "loss": 2.7345,
+      "step": 2515
+    },
+    {
+      "epoch": 0.20305060124283755,
+      "grad_norm": 0.7827345132827759,
+      "learning_rate": 0.00019281401750886854,
+      "loss": 2.7955,
+      "step": 2516
+    },
+    {
+      "epoch": 0.20313130497942056,
+      "grad_norm": 0.7482333183288574,
+      "learning_rate": 0.00019280813998173252,
+      "loss": 2.6963,
+      "step": 2517
+    },
+    {
+      "epoch": 0.20321200871600356,
+      "grad_norm": 0.8187180757522583,
+      "learning_rate": 0.00019280226014157509,
+      "loss": 2.7413,
+      "step": 2518
+    },
+    {
+      "epoch": 0.20329271245258657,
+      "grad_norm": 0.7708666920661926,
+      "learning_rate": 0.00019279637798854274,
+      "loss": 2.7636,
+      "step": 2519
+    },
+    {
+      "epoch": 0.20337341618916957,
+      "grad_norm": 0.7414180040359497,
+      "learning_rate": 0.00019279049352278208,
+      "loss": 2.7321,
+      "step": 2520
+    },
+    {
+      "epoch": 0.20345411992575257,
+      "grad_norm": 0.8172248601913452,
+      "learning_rate": 0.00019278460674443975,
+      "loss": 2.8026,
+      "step": 2521
+    },
+    {
+      "epoch": 0.20353482366233558,
+      "grad_norm": 0.7463089227676392,
+      "learning_rate": 0.0001927787176536625,
+      "loss": 2.74,
+      "step": 2522
+    },
+    {
+      "epoch": 0.20361552739891858,
+      "grad_norm": 0.7684210538864136,
+      "learning_rate": 0.00019277282625059704,
+      "loss": 2.782,
+      "step": 2523
+    },
+    {
+      "epoch": 0.2036962311355016,
+      "grad_norm": 0.9246797561645508,
+      "learning_rate": 0.00019276693253539027,
+      "loss": 2.8546,
+      "step": 2524
+    },
+    {
+      "epoch": 0.20377693487208456,
+      "grad_norm": 0.753753125667572,
+      "learning_rate": 0.00019276103650818906,
+      "loss": 2.7422,
+      "step": 2525
+    },
+    {
+      "epoch": 0.20385763860866757,
+      "grad_norm": 0.7461897134780884,
+      "learning_rate": 0.00019275513816914032,
+      "loss": 2.7575,
+      "step": 2526
+    },
+    {
+      "epoch": 0.20393834234525057,
+      "grad_norm": 0.7555257081985474,
+      "learning_rate": 0.00019274923751839106,
+      "loss": 2.7423,
+      "step": 2527
+    },
+    {
+      "epoch": 0.20401904608183358,
+      "grad_norm": 0.7628511786460876,
+      "learning_rate": 0.00019274333455608837,
+      "loss": 2.7386,
+      "step": 2528
+    },
+    {
+      "epoch": 0.20409974981841658,
+      "grad_norm": 0.7529371976852417,
+      "learning_rate": 0.00019273742928237937,
+      "loss": 2.6852,
+      "step": 2529
+    },
+    {
+      "epoch": 0.20418045355499959,
+      "grad_norm": 0.7466779351234436,
+      "learning_rate": 0.00019273152169741118,
+      "loss": 2.6996,
+      "step": 2530
+    },
+    {
+      "epoch": 0.2042611572915826,
+      "grad_norm": 0.7916153073310852,
+      "learning_rate": 0.0001927256118013311,
+      "loss": 2.7644,
+      "step": 2531
+    },
+    {
+      "epoch": 0.2043418610281656,
+      "grad_norm": 0.7662972211837769,
+      "learning_rate": 0.00019271969959428636,
+      "loss": 2.7497,
+      "step": 2532
+    },
+    {
+      "epoch": 0.2044225647647486,
+      "grad_norm": 0.8244680166244507,
+      "learning_rate": 0.00019271378507642432,
+      "loss": 2.7598,
+      "step": 2533
+    },
+    {
+      "epoch": 0.2045032685013316,
+      "grad_norm": 0.7721532583236694,
+      "learning_rate": 0.00019270786824789244,
+      "loss": 2.7303,
+      "step": 2534
+    },
+    {
+      "epoch": 0.2045839722379146,
+      "grad_norm": 0.7598209381103516,
+      "learning_rate": 0.0001927019491088381,
+      "loss": 2.734,
+      "step": 2535
+    },
+    {
+      "epoch": 0.2046646759744976,
+      "grad_norm": 0.7778685092926025,
+      "learning_rate": 0.00019269602765940887,
+      "loss": 2.7113,
+      "step": 2536
+    },
+    {
+      "epoch": 0.20474537971108062,
+      "grad_norm": 0.7447141408920288,
+      "learning_rate": 0.00019269010389975235,
+      "loss": 2.7205,
+      "step": 2537
+    },
+    {
+      "epoch": 0.20482608344766362,
+      "grad_norm": 0.8066664338111877,
+      "learning_rate": 0.00019268417783001613,
+      "loss": 2.7637,
+      "step": 2538
+    },
+    {
+      "epoch": 0.20490678718424662,
+      "grad_norm": 0.7055318355560303,
+      "learning_rate": 0.00019267824945034794,
+      "loss": 2.6936,
+      "step": 2539
+    },
+    {
+      "epoch": 0.20498749092082963,
+      "grad_norm": 0.832647979259491,
+      "learning_rate": 0.0001926723187608955,
+      "loss": 2.7423,
+      "step": 2540
+    },
+    {
+      "epoch": 0.20506819465741263,
+      "grad_norm": 0.7316983938217163,
+      "learning_rate": 0.0001926663857618066,
+      "loss": 2.7136,
+      "step": 2541
+    },
+    {
+      "epoch": 0.20514889839399564,
+      "grad_norm": 0.8115554451942444,
+      "learning_rate": 0.00019266045045322915,
+      "loss": 2.6964,
+      "step": 2542
+    },
+    {
+      "epoch": 0.20522960213057864,
+      "grad_norm": 0.802573025226593,
+      "learning_rate": 0.00019265451283531108,
+      "loss": 2.7989,
+      "step": 2543
+    },
+    {
+      "epoch": 0.20531030586716165,
+      "grad_norm": 0.7073348164558411,
+      "learning_rate": 0.00019264857290820033,
+      "loss": 2.7399,
+      "step": 2544
+    },
+    {
+      "epoch": 0.20539100960374465,
+      "grad_norm": 0.7749258279800415,
+      "learning_rate": 0.00019264263067204495,
+      "loss": 2.7321,
+      "step": 2545
+    },
+    {
+      "epoch": 0.20547171334032766,
+      "grad_norm": 0.7473557591438293,
+      "learning_rate": 0.00019263668612699305,
+      "loss": 2.7774,
+      "step": 2546
+    },
+    {
+      "epoch": 0.20555241707691066,
+      "grad_norm": 0.8073423504829407,
+      "learning_rate": 0.0001926307392731928,
+      "loss": 2.7429,
+      "step": 2547
+    },
+    {
+      "epoch": 0.20563312081349366,
+      "grad_norm": 0.9106586575508118,
+      "learning_rate": 0.00019262479011079235,
+      "loss": 2.7972,
+      "step": 2548
+    },
+    {
+      "epoch": 0.20571382455007667,
+      "grad_norm": 0.7975970506668091,
+      "learning_rate": 0.00019261883863994002,
+      "loss": 2.7561,
+      "step": 2549
+    },
+    {
+      "epoch": 0.20579452828665967,
+      "grad_norm": 0.8967030048370361,
+      "learning_rate": 0.00019261288486078414,
+      "loss": 2.7368,
+      "step": 2550
+    },
+    {
+      "epoch": 0.20587523202324268,
+      "grad_norm": 0.7157345414161682,
+      "learning_rate": 0.00019260692877347304,
+      "loss": 2.7329,
+      "step": 2551
+    },
+    {
+      "epoch": 0.20595593575982568,
+      "grad_norm": 0.8758620619773865,
+      "learning_rate": 0.00019260097037815524,
+      "loss": 2.7522,
+      "step": 2552
+    },
+    {
+      "epoch": 0.20603663949640869,
+      "grad_norm": 0.7948124408721924,
+      "learning_rate": 0.00019259500967497916,
+      "loss": 2.7675,
+      "step": 2553
+    },
+    {
+      "epoch": 0.2061173432329917,
+      "grad_norm": 0.8233941197395325,
+      "learning_rate": 0.00019258904666409344,
+      "loss": 2.7728,
+      "step": 2554
+    },
+    {
+      "epoch": 0.2061980469695747,
+      "grad_norm": 0.8084299564361572,
+      "learning_rate": 0.0001925830813456466,
+      "loss": 2.7728,
+      "step": 2555
+    },
+    {
+      "epoch": 0.2062787507061577,
+      "grad_norm": 0.8004557490348816,
+      "learning_rate": 0.00019257711371978737,
+      "loss": 2.7783,
+      "step": 2556
+    },
+    {
+      "epoch": 0.2063594544427407,
+      "grad_norm": 0.7999755144119263,
+      "learning_rate": 0.0001925711437866645,
+      "loss": 2.7632,
+      "step": 2557
+    },
+    {
+      "epoch": 0.2064401581793237,
+      "grad_norm": 0.7317264080047607,
+      "learning_rate": 0.0001925651715464267,
+      "loss": 2.7101,
+      "step": 2558
+    },
+    {
+      "epoch": 0.2065208619159067,
+      "grad_norm": 0.7906385660171509,
+      "learning_rate": 0.00019255919699922287,
+      "loss": 2.7258,
+      "step": 2559
+    },
+    {
+      "epoch": 0.20660156565248972,
+      "grad_norm": 0.7932917475700378,
+      "learning_rate": 0.0001925532201452019,
+      "loss": 2.7714,
+      "step": 2560
+    },
+    {
+      "epoch": 0.20668226938907272,
+      "grad_norm": 0.8039286732673645,
+      "learning_rate": 0.00019254724098451275,
+      "loss": 2.7469,
+      "step": 2561
+    },
+    {
+      "epoch": 0.20676297312565572,
+      "grad_norm": 0.79400634765625,
+      "learning_rate": 0.00019254125951730444,
+      "loss": 2.7499,
+      "step": 2562
+    },
+    {
+      "epoch": 0.20684367686223873,
+      "grad_norm": 0.8072263598442078,
+      "learning_rate": 0.00019253527574372603,
+      "loss": 2.7805,
+      "step": 2563
+    },
+    {
+      "epoch": 0.20692438059882173,
+      "grad_norm": 0.7117579579353333,
+      "learning_rate": 0.00019252928966392667,
+      "loss": 2.7321,
+      "step": 2564
+    },
+    {
+      "epoch": 0.20700508433540474,
+      "grad_norm": 0.7080324292182922,
+      "learning_rate": 0.00019252330127805554,
+      "loss": 2.7225,
+      "step": 2565
+    },
+    {
+      "epoch": 0.20708578807198774,
+      "grad_norm": 0.7276670336723328,
+      "learning_rate": 0.00019251731058626186,
+      "loss": 2.7592,
+      "step": 2566
+    },
+    {
+      "epoch": 0.20716649180857075,
+      "grad_norm": 0.8030811548233032,
+      "learning_rate": 0.00019251131758869495,
+      "loss": 2.7184,
+      "step": 2567
+    },
+    {
+      "epoch": 0.20724719554515375,
+      "grad_norm": 0.7808283567428589,
+      "learning_rate": 0.0001925053222855042,
+      "loss": 2.7504,
+      "step": 2568
+    },
+    {
+      "epoch": 0.20732789928173675,
+      "grad_norm": 0.783225953578949,
+      "learning_rate": 0.00019249932467683902,
+      "loss": 2.7125,
+      "step": 2569
+    },
+    {
+      "epoch": 0.20740860301831976,
+      "grad_norm": 0.7440134286880493,
+      "learning_rate": 0.00019249332476284887,
+      "loss": 2.7938,
+      "step": 2570
+    },
+    {
+      "epoch": 0.20748930675490276,
+      "grad_norm": 0.8729553818702698,
+      "learning_rate": 0.00019248732254368328,
+      "loss": 2.8338,
+      "step": 2571
+    },
+    {
+      "epoch": 0.20757001049148577,
+      "grad_norm": 0.8170497417449951,
+      "learning_rate": 0.0001924813180194918,
+      "loss": 2.7254,
+      "step": 2572
+    },
+    {
+      "epoch": 0.20765071422806877,
+      "grad_norm": 0.733220100402832,
+      "learning_rate": 0.00019247531119042418,
+      "loss": 2.6401,
+      "step": 2573
+    },
+    {
+      "epoch": 0.20773141796465178,
+      "grad_norm": 0.7247937917709351,
+      "learning_rate": 0.00019246930205663008,
+      "loss": 2.736,
+      "step": 2574
+    },
+    {
+      "epoch": 0.20781212170123478,
+      "grad_norm": 0.7880212068557739,
+      "learning_rate": 0.00019246329061825925,
+      "loss": 2.7173,
+      "step": 2575
+    },
+    {
+      "epoch": 0.20789282543781776,
+      "grad_norm": 0.820808470249176,
+      "learning_rate": 0.00019245727687546149,
+      "loss": 2.7331,
+      "step": 2576
+    },
+    {
+      "epoch": 0.20797352917440076,
+      "grad_norm": 0.8605412840843201,
+      "learning_rate": 0.00019245126082838673,
+      "loss": 2.761,
+      "step": 2577
+    },
+    {
+      "epoch": 0.20805423291098377,
+      "grad_norm": 0.763506293296814,
+      "learning_rate": 0.00019244524247718486,
+      "loss": 2.7053,
+      "step": 2578
+    },
+    {
+      "epoch": 0.20813493664756677,
+      "grad_norm": 0.8428114652633667,
+      "learning_rate": 0.00019243922182200592,
+      "loss": 2.724,
+      "step": 2579
+    },
+    {
+      "epoch": 0.20821564038414977,
+      "grad_norm": 0.821986734867096,
+      "learning_rate": 0.0001924331988629999,
+      "loss": 2.7615,
+      "step": 2580
+    },
+    {
+      "epoch": 0.20829634412073278,
+      "grad_norm": 0.8177430629730225,
+      "learning_rate": 0.00019242717360031693,
+      "loss": 2.7012,
+      "step": 2581
+    },
+    {
+      "epoch": 0.20837704785731578,
+      "grad_norm": 0.7584180235862732,
+      "learning_rate": 0.00019242114603410724,
+      "loss": 2.7372,
+      "step": 2582
+    },
+    {
+      "epoch": 0.2084577515938988,
+      "grad_norm": 0.9384645223617554,
+      "learning_rate": 0.00019241511616452096,
+      "loss": 2.695,
+      "step": 2583
+    },
+    {
+      "epoch": 0.2085384553304818,
+      "grad_norm": 0.8518964648246765,
+      "learning_rate": 0.00019240908399170844,
+      "loss": 2.8216,
+      "step": 2584
+    },
+    {
+      "epoch": 0.2086191590670648,
+      "grad_norm": 0.9082949161529541,
+      "learning_rate": 0.00019240304951581995,
+      "loss": 2.777,
+      "step": 2585
+    },
+    {
+      "epoch": 0.2086998628036478,
+      "grad_norm": 0.7906371355056763,
+      "learning_rate": 0.00019239701273700597,
+      "loss": 2.7083,
+      "step": 2586
+    },
+    {
+      "epoch": 0.2087805665402308,
+      "grad_norm": 0.7711954712867737,
+      "learning_rate": 0.00019239097365541686,
+      "loss": 2.6907,
+      "step": 2587
+    },
+    {
+      "epoch": 0.2088612702768138,
+      "grad_norm": 0.8155506253242493,
+      "learning_rate": 0.0001923849322712032,
+      "loss": 2.7602,
+      "step": 2588
+    },
+    {
+      "epoch": 0.20894197401339681,
+      "grad_norm": 0.8843441009521484,
+      "learning_rate": 0.0001923788885845155,
+      "loss": 2.7525,
+      "step": 2589
+    },
+    {
+      "epoch": 0.20902267774997982,
+      "grad_norm": 0.7336379289627075,
+      "learning_rate": 0.00019237284259550444,
+      "loss": 2.731,
+      "step": 2590
+    },
+    {
+      "epoch": 0.20910338148656282,
+      "grad_norm": 0.8261263370513916,
+      "learning_rate": 0.00019236679430432066,
+      "loss": 2.6493,
+      "step": 2591
+    },
+    {
+      "epoch": 0.20918408522314583,
+      "grad_norm": 0.7716216444969177,
+      "learning_rate": 0.00019236074371111497,
+      "loss": 2.7775,
+      "step": 2592
+    },
+    {
+      "epoch": 0.20926478895972883,
+      "grad_norm": 0.8390100598335266,
+      "learning_rate": 0.00019235469081603808,
+      "loss": 2.7532,
+      "step": 2593
+    },
+    {
+      "epoch": 0.20934549269631184,
+      "grad_norm": 0.8388446569442749,
+      "learning_rate": 0.00019234863561924087,
+      "loss": 2.8171,
+      "step": 2594
+    },
+    {
+      "epoch": 0.20942619643289484,
+      "grad_norm": 0.8003209829330444,
+      "learning_rate": 0.00019234257812087425,
+      "loss": 2.7385,
+      "step": 2595
+    },
+    {
+      "epoch": 0.20950690016947784,
+      "grad_norm": 0.8008458018302917,
+      "learning_rate": 0.00019233651832108918,
+      "loss": 2.7366,
+      "step": 2596
+    },
+    {
+      "epoch": 0.20958760390606085,
+      "grad_norm": 0.7701897025108337,
+      "learning_rate": 0.00019233045622003676,
+      "loss": 2.69,
+      "step": 2597
+    },
+    {
+      "epoch": 0.20966830764264385,
+      "grad_norm": 0.8106730580329895,
+      "learning_rate": 0.00019232439181786796,
+      "loss": 2.6911,
+      "step": 2598
+    },
+    {
+      "epoch": 0.20974901137922686,
+      "grad_norm": 0.9580766558647156,
+      "learning_rate": 0.00019231832511473401,
+      "loss": 2.7663,
+      "step": 2599
+    },
+    {
+      "epoch": 0.20982971511580986,
+      "grad_norm": 0.7851876616477966,
+      "learning_rate": 0.0001923122561107861,
+      "loss": 2.7632,
+      "step": 2600
+    },
+    {
+      "epoch": 0.20991041885239287,
+      "grad_norm": 0.8160942196846008,
+      "learning_rate": 0.0001923061848061754,
+      "loss": 2.8533,
+      "step": 2601
+    },
+    {
+      "epoch": 0.20999112258897587,
+      "grad_norm": 0.8540663719177246,
+      "learning_rate": 0.00019230011120105334,
+      "loss": 2.7083,
+      "step": 2602
+    },
+    {
+      "epoch": 0.21007182632555887,
+      "grad_norm": 0.8273833394050598,
+      "learning_rate": 0.0001922940352955712,
+      "loss": 2.7916,
+      "step": 2603
+    },
+    {
+      "epoch": 0.21015253006214188,
+      "grad_norm": 0.8394255638122559,
+      "learning_rate": 0.00019228795708988046,
+      "loss": 2.8561,
+      "step": 2604
+    },
+    {
+      "epoch": 0.21023323379872488,
+      "grad_norm": 0.8291410803794861,
+      "learning_rate": 0.00019228187658413258,
+      "loss": 2.7462,
+      "step": 2605
+    },
+    {
+      "epoch": 0.2103139375353079,
+      "grad_norm": 0.7984235286712646,
+      "learning_rate": 0.00019227579377847912,
+      "loss": 2.7459,
+      "step": 2606
+    },
+    {
+      "epoch": 0.2103946412718909,
+      "grad_norm": 0.8343340158462524,
+      "learning_rate": 0.00019226970867307163,
+      "loss": 2.6963,
+      "step": 2607
+    },
+    {
+      "epoch": 0.2104753450084739,
+      "grad_norm": 0.6982808709144592,
+      "learning_rate": 0.00019226362126806184,
+      "loss": 2.7333,
+      "step": 2608
+    },
+    {
+      "epoch": 0.2105560487450569,
+      "grad_norm": 0.8039572834968567,
+      "learning_rate": 0.0001922575315636014,
+      "loss": 2.7253,
+      "step": 2609
+    },
+    {
+      "epoch": 0.2106367524816399,
+      "grad_norm": 0.8708705902099609,
+      "learning_rate": 0.00019225143955984214,
+      "loss": 2.7555,
+      "step": 2610
+    },
+    {
+      "epoch": 0.2107174562182229,
+      "grad_norm": 0.8773347735404968,
+      "learning_rate": 0.00019224534525693585,
+      "loss": 2.7598,
+      "step": 2611
+    },
+    {
+      "epoch": 0.2107981599548059,
+      "grad_norm": 0.8151054978370667,
+      "learning_rate": 0.0001922392486550344,
+      "loss": 2.7398,
+      "step": 2612
+    },
+    {
+      "epoch": 0.21087886369138892,
+      "grad_norm": 0.7922329306602478,
+      "learning_rate": 0.0001922331497542898,
+      "loss": 2.7296,
+      "step": 2613
+    },
+    {
+      "epoch": 0.21095956742797192,
+      "grad_norm": 0.7536506652832031,
+      "learning_rate": 0.00019222704855485396,
+      "loss": 2.7897,
+      "step": 2614
+    },
+    {
+      "epoch": 0.21104027116455493,
+      "grad_norm": 0.7539274096488953,
+      "learning_rate": 0.000192220945056879,
+      "loss": 2.7809,
+      "step": 2615
+    },
+    {
+      "epoch": 0.21112097490113793,
+      "grad_norm": 0.7737646698951721,
+      "learning_rate": 0.00019221483926051705,
+      "loss": 2.7195,
+      "step": 2616
+    },
+    {
+      "epoch": 0.21120167863772094,
+      "grad_norm": 0.7421913743019104,
+      "learning_rate": 0.00019220873116592024,
+      "loss": 2.6817,
+      "step": 2617
+    },
+    {
+      "epoch": 0.21128238237430394,
+      "grad_norm": 0.7872927784919739,
+      "learning_rate": 0.0001922026207732408,
+      "loss": 2.7379,
+      "step": 2618
+    },
+    {
+      "epoch": 0.21136308611088694,
+      "grad_norm": 0.7950671315193176,
+      "learning_rate": 0.00019219650808263104,
+      "loss": 2.7135,
+      "step": 2619
+    },
+    {
+      "epoch": 0.21144378984746995,
+      "grad_norm": 0.7711792588233948,
+      "learning_rate": 0.0001921903930942433,
+      "loss": 2.7021,
+      "step": 2620
+    },
+    {
+      "epoch": 0.21152449358405295,
+      "grad_norm": 0.9030743837356567,
+      "learning_rate": 0.00019218427580822996,
+      "loss": 2.8083,
+      "step": 2621
+    },
+    {
+      "epoch": 0.21160519732063596,
+      "grad_norm": 0.8191907405853271,
+      "learning_rate": 0.0001921781562247435,
+      "loss": 2.6998,
+      "step": 2622
+    },
+    {
+      "epoch": 0.21168590105721896,
+      "grad_norm": 0.7883538603782654,
+      "learning_rate": 0.00019217203434393644,
+      "loss": 2.7573,
+      "step": 2623
+    },
+    {
+      "epoch": 0.21176660479380197,
+      "grad_norm": 0.7565868496894836,
+      "learning_rate": 0.00019216591016596134,
+      "loss": 2.7725,
+      "step": 2624
+    },
+    {
+      "epoch": 0.21184730853038497,
+      "grad_norm": 0.8579828143119812,
+      "learning_rate": 0.00019215978369097086,
+      "loss": 2.7529,
+      "step": 2625
+    },
+    {
+      "epoch": 0.21192801226696797,
+      "grad_norm": 0.7835422158241272,
+      "learning_rate": 0.0001921536549191176,
+      "loss": 2.6926,
+      "step": 2626
+    },
+    {
+      "epoch": 0.21200871600355095,
+      "grad_norm": 0.8041907548904419,
+      "learning_rate": 0.00019214752385055442,
+      "loss": 2.7541,
+      "step": 2627
+    },
+    {
+      "epoch": 0.21208941974013396,
+      "grad_norm": 0.7754014730453491,
+      "learning_rate": 0.00019214139048543406,
+      "loss": 2.6807,
+      "step": 2628
+    },
+    {
+      "epoch": 0.21217012347671696,
+      "grad_norm": 0.8222344517707825,
+      "learning_rate": 0.00019213525482390936,
+      "loss": 2.7339,
+      "step": 2629
+    },
+    {
+      "epoch": 0.21225082721329996,
+      "grad_norm": 0.8083673715591431,
+      "learning_rate": 0.0001921291168661333,
+      "loss": 2.739,
+      "step": 2630
+    },
+    {
+      "epoch": 0.21233153094988297,
+      "grad_norm": 0.8039100766181946,
+      "learning_rate": 0.0001921229766122588,
+      "loss": 2.7372,
+      "step": 2631
+    },
+    {
+      "epoch": 0.21241223468646597,
+      "grad_norm": 0.7513072490692139,
+      "learning_rate": 0.00019211683406243892,
+      "loss": 2.7284,
+      "step": 2632
+    },
+    {
+      "epoch": 0.21249293842304898,
+      "grad_norm": 0.7653890252113342,
+      "learning_rate": 0.00019211068921682673,
+      "loss": 2.6911,
+      "step": 2633
+    },
+    {
+      "epoch": 0.21257364215963198,
+      "grad_norm": 0.7210217714309692,
+      "learning_rate": 0.00019210454207557542,
+      "loss": 2.6989,
+      "step": 2634
+    },
+    {
+      "epoch": 0.21265434589621499,
+      "grad_norm": 0.7389202117919922,
+      "learning_rate": 0.00019209839263883814,
+      "loss": 2.7016,
+      "step": 2635
+    },
+    {
+      "epoch": 0.212735049632798,
+      "grad_norm": 0.8069031238555908,
+      "learning_rate": 0.00019209224090676813,
+      "loss": 2.8213,
+      "step": 2636
+    },
+    {
+      "epoch": 0.212815753369381,
+      "grad_norm": 0.8019161224365234,
+      "learning_rate": 0.00019208608687951877,
+      "loss": 2.7413,
+      "step": 2637
+    },
+    {
+      "epoch": 0.212896457105964,
+      "grad_norm": 0.775572657585144,
+      "learning_rate": 0.00019207993055724343,
+      "loss": 2.7016,
+      "step": 2638
+    },
+    {
+      "epoch": 0.212977160842547,
+      "grad_norm": 0.7482941746711731,
+      "learning_rate": 0.0001920737719400955,
+      "loss": 2.7991,
+      "step": 2639
+    },
+    {
+      "epoch": 0.21305786457913,
+      "grad_norm": 0.8467636704444885,
+      "learning_rate": 0.0001920676110282285,
+      "loss": 2.7401,
+      "step": 2640
+    },
+    {
+      "epoch": 0.213138568315713,
+      "grad_norm": 0.8726305365562439,
+      "learning_rate": 0.00019206144782179597,
+      "loss": 2.7599,
+      "step": 2641
+    },
+    {
+      "epoch": 0.21321927205229602,
+      "grad_norm": 0.740527868270874,
+      "learning_rate": 0.00019205528232095148,
+      "loss": 2.7326,
+      "step": 2642
+    },
+    {
+      "epoch": 0.21329997578887902,
+      "grad_norm": 0.7932354211807251,
+      "learning_rate": 0.00019204911452584873,
+      "loss": 2.7873,
+      "step": 2643
+    },
+    {
+      "epoch": 0.21338067952546202,
+      "grad_norm": 0.7994125485420227,
+      "learning_rate": 0.00019204294443664143,
+      "loss": 2.7305,
+      "step": 2644
+    },
+    {
+      "epoch": 0.21346138326204503,
+      "grad_norm": 0.880557656288147,
+      "learning_rate": 0.00019203677205348338,
+      "loss": 2.7295,
+      "step": 2645
+    },
+    {
+      "epoch": 0.21354208699862803,
+      "grad_norm": 0.8269557952880859,
+      "learning_rate": 0.00019203059737652836,
+      "loss": 2.765,
+      "step": 2646
+    },
+    {
+      "epoch": 0.21362279073521104,
+      "grad_norm": 0.8732784986495972,
+      "learning_rate": 0.00019202442040593026,
+      "loss": 2.6742,
+      "step": 2647
+    },
+    {
+      "epoch": 0.21370349447179404,
+      "grad_norm": 0.7921704649925232,
+      "learning_rate": 0.0001920182411418431,
+      "loss": 2.7144,
+      "step": 2648
+    },
+    {
+      "epoch": 0.21378419820837705,
+      "grad_norm": 0.8097628355026245,
+      "learning_rate": 0.00019201205958442082,
+      "loss": 2.7513,
+      "step": 2649
+    },
+    {
+      "epoch": 0.21386490194496005,
+      "grad_norm": 0.8230542540550232,
+      "learning_rate": 0.00019200587573381744,
+      "loss": 2.7648,
+      "step": 2650
+    },
+    {
+      "epoch": 0.21394560568154306,
+      "grad_norm": 0.7719153761863708,
+      "learning_rate": 0.0001919996895901872,
+      "loss": 2.7637,
+      "step": 2651
+    },
+    {
+      "epoch": 0.21402630941812606,
+      "grad_norm": 0.9022669792175293,
+      "learning_rate": 0.00019199350115368415,
+      "loss": 2.7707,
+      "step": 2652
+    },
+    {
+      "epoch": 0.21410701315470906,
+      "grad_norm": 0.8111257553100586,
+      "learning_rate": 0.00019198731042446263,
+      "loss": 2.7423,
+      "step": 2653
+    },
+    {
+      "epoch": 0.21418771689129207,
+      "grad_norm": 0.7534981966018677,
+      "learning_rate": 0.00019198111740267683,
+      "loss": 2.7474,
+      "step": 2654
+    },
+    {
+      "epoch": 0.21426842062787507,
+      "grad_norm": 0.761411190032959,
+      "learning_rate": 0.00019197492208848117,
+      "loss": 2.7541,
+      "step": 2655
+    },
+    {
+      "epoch": 0.21434912436445808,
+      "grad_norm": 0.8076324462890625,
+      "learning_rate": 0.00019196872448203002,
+      "loss": 2.7198,
+      "step": 2656
+    },
+    {
+      "epoch": 0.21442982810104108,
+      "grad_norm": 0.7987746000289917,
+      "learning_rate": 0.00019196252458347784,
+      "loss": 2.7164,
+      "step": 2657
+    },
+    {
+      "epoch": 0.21451053183762409,
+      "grad_norm": 0.7581545114517212,
+      "learning_rate": 0.0001919563223929792,
+      "loss": 2.6837,
+      "step": 2658
+    },
+    {
+      "epoch": 0.2145912355742071,
+      "grad_norm": 0.8773601055145264,
+      "learning_rate": 0.00019195011791068857,
+      "loss": 2.8248,
+      "step": 2659
+    },
+    {
+      "epoch": 0.2146719393107901,
+      "grad_norm": 0.7027503252029419,
+      "learning_rate": 0.00019194391113676066,
+      "loss": 2.6726,
+      "step": 2660
+    },
+    {
+      "epoch": 0.2147526430473731,
+      "grad_norm": 0.8650866746902466,
+      "learning_rate": 0.00019193770207135015,
+      "loss": 2.7348,
+      "step": 2661
+    },
+    {
+      "epoch": 0.2148333467839561,
+      "grad_norm": 0.8521862030029297,
+      "learning_rate": 0.0001919314907146118,
+      "loss": 2.7409,
+      "step": 2662
+    },
+    {
+      "epoch": 0.2149140505205391,
+      "grad_norm": 0.8098535537719727,
+      "learning_rate": 0.00019192527706670033,
+      "loss": 2.7615,
+      "step": 2663
+    },
+    {
+      "epoch": 0.2149947542571221,
+      "grad_norm": 0.7396193146705627,
+      "learning_rate": 0.0001919190611277707,
+      "loss": 2.7191,
+      "step": 2664
+    },
+    {
+      "epoch": 0.21507545799370512,
+      "grad_norm": 0.8245799541473389,
+      "learning_rate": 0.00019191284289797776,
+      "loss": 2.7429,
+      "step": 2665
+    },
+    {
+      "epoch": 0.21515616173028812,
+      "grad_norm": 0.791646420955658,
+      "learning_rate": 0.00019190662237747656,
+      "loss": 2.7197,
+      "step": 2666
+    },
+    {
+      "epoch": 0.21523686546687112,
+      "grad_norm": 0.7850802540779114,
+      "learning_rate": 0.00019190039956642205,
+      "loss": 2.7353,
+      "step": 2667
+    },
+    {
+      "epoch": 0.21531756920345413,
+      "grad_norm": 0.7657971978187561,
+      "learning_rate": 0.00019189417446496937,
+      "loss": 2.7083,
+      "step": 2668
+    },
+    {
+      "epoch": 0.21539827294003713,
+      "grad_norm": 0.7704403400421143,
+      "learning_rate": 0.00019188794707327363,
+      "loss": 2.7813,
+      "step": 2669
+    },
+    {
+      "epoch": 0.21547897667662014,
+      "grad_norm": 0.7345917224884033,
+      "learning_rate": 0.00019188171739149005,
+      "loss": 2.7098,
+      "step": 2670
+    },
+    {
+      "epoch": 0.21555968041320314,
+      "grad_norm": 0.728831946849823,
+      "learning_rate": 0.00019187548541977392,
+      "loss": 2.6745,
+      "step": 2671
+    },
+    {
+      "epoch": 0.21564038414978615,
+      "grad_norm": 0.8079627156257629,
+      "learning_rate": 0.0001918692511582805,
+      "loss": 2.6427,
+      "step": 2672
+    },
+    {
+      "epoch": 0.21572108788636915,
+      "grad_norm": 0.766808032989502,
+      "learning_rate": 0.0001918630146071652,
+      "loss": 2.6956,
+      "step": 2673
+    },
+    {
+      "epoch": 0.21580179162295215,
+      "grad_norm": 0.7555391192436218,
+      "learning_rate": 0.00019185677576658345,
+      "loss": 2.6499,
+      "step": 2674
+    },
+    {
+      "epoch": 0.21588249535953516,
+      "grad_norm": 0.7740229964256287,
+      "learning_rate": 0.00019185053463669074,
+      "loss": 2.7685,
+      "step": 2675
+    },
+    {
+      "epoch": 0.21596319909611816,
+      "grad_norm": 0.8272803425788879,
+      "learning_rate": 0.00019184429121764257,
+      "loss": 2.7272,
+      "step": 2676
+    },
+    {
+      "epoch": 0.21604390283270117,
+      "grad_norm": 0.870625376701355,
+      "learning_rate": 0.00019183804550959463,
+      "loss": 2.7509,
+      "step": 2677
+    },
+    {
+      "epoch": 0.21612460656928414,
+      "grad_norm": 0.8021238446235657,
+      "learning_rate": 0.0001918317975127025,
+      "loss": 2.7058,
+      "step": 2678
+    },
+    {
+      "epoch": 0.21620531030586715,
+      "grad_norm": 0.729918897151947,
+      "learning_rate": 0.00019182554722712192,
+      "loss": 2.6145,
+      "step": 2679
+    },
+    {
+      "epoch": 0.21628601404245015,
+      "grad_norm": 0.7658380270004272,
+      "learning_rate": 0.00019181929465300867,
+      "loss": 2.712,
+      "step": 2680
+    },
+    {
+      "epoch": 0.21636671777903316,
+      "grad_norm": 0.7702174186706543,
+      "learning_rate": 0.00019181303979051858,
+      "loss": 2.8257,
+      "step": 2681
+    },
+    {
+      "epoch": 0.21644742151561616,
+      "grad_norm": 0.7782231569290161,
+      "learning_rate": 0.00019180678263980755,
+      "loss": 2.8226,
+      "step": 2682
+    },
+    {
+      "epoch": 0.21652812525219917,
+      "grad_norm": 0.7448495626449585,
+      "learning_rate": 0.0001918005232010315,
+      "loss": 2.7877,
+      "step": 2683
+    },
+    {
+      "epoch": 0.21660882898878217,
+      "grad_norm": 0.7273527979850769,
+      "learning_rate": 0.00019179426147434647,
+      "loss": 2.7169,
+      "step": 2684
+    },
+    {
+      "epoch": 0.21668953272536517,
+      "grad_norm": 0.7730992436408997,
+      "learning_rate": 0.00019178799745990846,
+      "loss": 2.717,
+      "step": 2685
+    },
+    {
+      "epoch": 0.21677023646194818,
+      "grad_norm": 0.7709231376647949,
+      "learning_rate": 0.0001917817311578736,
+      "loss": 2.7676,
+      "step": 2686
+    },
+    {
+      "epoch": 0.21685094019853118,
+      "grad_norm": 0.7825181484222412,
+      "learning_rate": 0.00019177546256839812,
+      "loss": 2.7473,
+      "step": 2687
+    },
+    {
+      "epoch": 0.2169316439351142,
+      "grad_norm": 0.8133581280708313,
+      "learning_rate": 0.0001917691916916382,
+      "loss": 2.7242,
+      "step": 2688
+    },
+    {
+      "epoch": 0.2170123476716972,
+      "grad_norm": 0.7833015322685242,
+      "learning_rate": 0.00019176291852775011,
+      "loss": 2.8128,
+      "step": 2689
+    },
+    {
+      "epoch": 0.2170930514082802,
+      "grad_norm": 0.7423487305641174,
+      "learning_rate": 0.00019175664307689028,
+      "loss": 2.6999,
+      "step": 2690
+    },
+    {
+      "epoch": 0.2171737551448632,
+      "grad_norm": 0.7881289124488831,
+      "learning_rate": 0.000191750365339215,
+      "loss": 2.7349,
+      "step": 2691
+    },
+    {
+      "epoch": 0.2172544588814462,
+      "grad_norm": 0.8316197395324707,
+      "learning_rate": 0.00019174408531488077,
+      "loss": 2.7654,
+      "step": 2692
+    },
+    {
+      "epoch": 0.2173351626180292,
+      "grad_norm": 0.7589917778968811,
+      "learning_rate": 0.00019173780300404413,
+      "loss": 2.6815,
+      "step": 2693
+    },
+    {
+      "epoch": 0.21741586635461221,
+      "grad_norm": 0.7752439975738525,
+      "learning_rate": 0.00019173151840686163,
+      "loss": 2.7804,
+      "step": 2694
+    },
+    {
+      "epoch": 0.21749657009119522,
+      "grad_norm": 0.8156552910804749,
+      "learning_rate": 0.0001917252315234899,
+      "loss": 2.7325,
+      "step": 2695
+    },
+    {
+      "epoch": 0.21757727382777822,
+      "grad_norm": 0.8886982798576355,
+      "learning_rate": 0.00019171894235408564,
+      "loss": 2.7257,
+      "step": 2696
+    },
+    {
+      "epoch": 0.21765797756436123,
+      "grad_norm": 0.8270704746246338,
+      "learning_rate": 0.00019171265089880558,
+      "loss": 2.7357,
+      "step": 2697
+    },
+    {
+      "epoch": 0.21773868130094423,
+      "grad_norm": 0.807700514793396,
+      "learning_rate": 0.00019170635715780651,
+      "loss": 2.7488,
+      "step": 2698
+    },
+    {
+      "epoch": 0.21781938503752724,
+      "grad_norm": 0.8195288181304932,
+      "learning_rate": 0.00019170006113124533,
+      "loss": 2.7048,
+      "step": 2699
+    },
+    {
+      "epoch": 0.21790008877411024,
+      "grad_norm": 0.817097008228302,
+      "learning_rate": 0.00019169376281927888,
+      "loss": 2.7148,
+      "step": 2700
+    },
+    {
+      "epoch": 0.21798079251069324,
+      "grad_norm": 0.8415588140487671,
+      "learning_rate": 0.0001916874622220642,
+      "loss": 2.7376,
+      "step": 2701
+    },
+    {
+      "epoch": 0.21806149624727625,
+      "grad_norm": 0.8004198670387268,
+      "learning_rate": 0.00019168115933975826,
+      "loss": 2.7145,
+      "step": 2702
+    },
+    {
+      "epoch": 0.21814219998385925,
+      "grad_norm": 0.8167368769645691,
+      "learning_rate": 0.0001916748541725182,
+      "loss": 2.6923,
+      "step": 2703
+    },
+    {
+      "epoch": 0.21822290372044226,
+      "grad_norm": 0.8877980709075928,
+      "learning_rate": 0.0001916685467205011,
+      "loss": 2.8232,
+      "step": 2704
+    },
+    {
+      "epoch": 0.21830360745702526,
+      "grad_norm": 0.7835622429847717,
+      "learning_rate": 0.00019166223698386422,
+      "loss": 2.7797,
+      "step": 2705
+    },
+    {
+      "epoch": 0.21838431119360827,
+      "grad_norm": 0.8023552894592285,
+      "learning_rate": 0.00019165592496276477,
+      "loss": 2.6697,
+      "step": 2706
+    },
+    {
+      "epoch": 0.21846501493019127,
+      "grad_norm": 0.8549069166183472,
+      "learning_rate": 0.00019164961065736008,
+      "loss": 2.729,
+      "step": 2707
+    },
+    {
+      "epoch": 0.21854571866677427,
+      "grad_norm": 0.8561950325965881,
+      "learning_rate": 0.00019164329406780753,
+      "loss": 2.772,
+      "step": 2708
+    },
+    {
+      "epoch": 0.21862642240335728,
+      "grad_norm": 0.6979276537895203,
+      "learning_rate": 0.00019163697519426453,
+      "loss": 2.7195,
+      "step": 2709
+    },
+    {
+      "epoch": 0.21870712613994028,
+      "grad_norm": 0.7659175395965576,
+      "learning_rate": 0.00019163065403688856,
+      "loss": 2.7742,
+      "step": 2710
+    },
+    {
+      "epoch": 0.2187878298765233,
+      "grad_norm": 0.8621466755867004,
+      "learning_rate": 0.00019162433059583718,
+      "loss": 2.721,
+      "step": 2711
+    },
+    {
+      "epoch": 0.2188685336131063,
+      "grad_norm": 0.8086833357810974,
+      "learning_rate": 0.00019161800487126795,
+      "loss": 2.7356,
+      "step": 2712
+    },
+    {
+      "epoch": 0.2189492373496893,
+      "grad_norm": 0.816215455532074,
+      "learning_rate": 0.00019161167686333855,
+      "loss": 2.7159,
+      "step": 2713
+    },
+    {
+      "epoch": 0.2190299410862723,
+      "grad_norm": 0.9180822968482971,
+      "learning_rate": 0.0001916053465722067,
+      "loss": 2.7162,
+      "step": 2714
+    },
+    {
+      "epoch": 0.2191106448228553,
+      "grad_norm": 0.7547199130058289,
+      "learning_rate": 0.00019159901399803014,
+      "loss": 2.7338,
+      "step": 2715
+    },
+    {
+      "epoch": 0.2191913485594383,
+      "grad_norm": 0.7380769848823547,
+      "learning_rate": 0.00019159267914096675,
+      "loss": 2.7149,
+      "step": 2716
+    },
+    {
+      "epoch": 0.2192720522960213,
+      "grad_norm": 0.7242285013198853,
+      "learning_rate": 0.00019158634200117433,
+      "loss": 2.724,
+      "step": 2717
+    },
+    {
+      "epoch": 0.21935275603260432,
+      "grad_norm": 0.8400316834449768,
+      "learning_rate": 0.00019158000257881087,
+      "loss": 2.7528,
+      "step": 2718
+    },
+    {
+      "epoch": 0.21943345976918732,
+      "grad_norm": 0.8437172770500183,
+      "learning_rate": 0.00019157366087403435,
+      "loss": 2.7872,
+      "step": 2719
+    },
+    {
+      "epoch": 0.21951416350577033,
+      "grad_norm": 0.7428301572799683,
+      "learning_rate": 0.00019156731688700282,
+      "loss": 2.6831,
+      "step": 2720
+    },
+    {
+      "epoch": 0.21959486724235333,
+      "grad_norm": 0.7589641213417053,
+      "learning_rate": 0.00019156097061787445,
+      "loss": 2.7105,
+      "step": 2721
+    },
+    {
+      "epoch": 0.21967557097893634,
+      "grad_norm": 0.7607305645942688,
+      "learning_rate": 0.00019155462206680727,
+      "loss": 2.7913,
+      "step": 2722
+    },
+    {
+      "epoch": 0.21975627471551934,
+      "grad_norm": 0.7455689311027527,
+      "learning_rate": 0.00019154827123395963,
+      "loss": 2.6321,
+      "step": 2723
+    },
+    {
+      "epoch": 0.21983697845210234,
+      "grad_norm": 0.7860318422317505,
+      "learning_rate": 0.00019154191811948974,
+      "loss": 2.7907,
+      "step": 2724
+    },
+    {
+      "epoch": 0.21991768218868535,
+      "grad_norm": 0.8101385235786438,
+      "learning_rate": 0.00019153556272355596,
+      "loss": 2.7682,
+      "step": 2725
+    },
+    {
+      "epoch": 0.21999838592526835,
+      "grad_norm": 0.7437283396720886,
+      "learning_rate": 0.00019152920504631667,
+      "loss": 2.7271,
+      "step": 2726
+    },
+    {
+      "epoch": 0.22007908966185136,
+      "grad_norm": 0.7390851974487305,
+      "learning_rate": 0.00019152284508793034,
+      "loss": 2.7492,
+      "step": 2727
+    },
+    {
+      "epoch": 0.22015979339843436,
+      "grad_norm": 0.9074966311454773,
+      "learning_rate": 0.0001915164828485555,
+      "loss": 2.8076,
+      "step": 2728
+    },
+    {
+      "epoch": 0.22024049713501734,
+      "grad_norm": 0.7644218802452087,
+      "learning_rate": 0.00019151011832835063,
+      "loss": 2.7238,
+      "step": 2729
+    },
+    {
+      "epoch": 0.22032120087160034,
+      "grad_norm": 0.823567807674408,
+      "learning_rate": 0.0001915037515274744,
+      "loss": 2.7701,
+      "step": 2730
+    },
+    {
+      "epoch": 0.22040190460818335,
+      "grad_norm": 0.7601858377456665,
+      "learning_rate": 0.00019149738244608552,
+      "loss": 2.6981,
+      "step": 2731
+    },
+    {
+      "epoch": 0.22048260834476635,
+      "grad_norm": 0.8242961764335632,
+      "learning_rate": 0.00019149101108434269,
+      "loss": 2.6916,
+      "step": 2732
+    },
+    {
+      "epoch": 0.22056331208134936,
+      "grad_norm": 0.7970656156539917,
+      "learning_rate": 0.0001914846374424047,
+      "loss": 2.7858,
+      "step": 2733
+    },
+    {
+      "epoch": 0.22064401581793236,
+      "grad_norm": 0.7844050526618958,
+      "learning_rate": 0.0001914782615204304,
+      "loss": 2.6782,
+      "step": 2734
+    },
+    {
+      "epoch": 0.22072471955451536,
+      "grad_norm": 0.7965044975280762,
+      "learning_rate": 0.00019147188331857868,
+      "loss": 2.7563,
+      "step": 2735
+    },
+    {
+      "epoch": 0.22080542329109837,
+      "grad_norm": 0.8189071416854858,
+      "learning_rate": 0.00019146550283700856,
+      "loss": 2.7587,
+      "step": 2736
+    },
+    {
+      "epoch": 0.22088612702768137,
+      "grad_norm": 0.7610960602760315,
+      "learning_rate": 0.00019145912007587898,
+      "loss": 2.663,
+      "step": 2737
+    },
+    {
+      "epoch": 0.22096683076426438,
+      "grad_norm": 0.7642313838005066,
+      "learning_rate": 0.00019145273503534907,
+      "loss": 2.78,
+      "step": 2738
+    },
+    {
+      "epoch": 0.22104753450084738,
+      "grad_norm": 0.7699539065361023,
+      "learning_rate": 0.0001914463477155779,
+      "loss": 2.7429,
+      "step": 2739
+    },
+    {
+      "epoch": 0.22112823823743039,
+      "grad_norm": 0.7674413919448853,
+      "learning_rate": 0.00019143995811672477,
+      "loss": 2.7048,
+      "step": 2740
+    },
+    {
+      "epoch": 0.2212089419740134,
+      "grad_norm": 0.7871866226196289,
+      "learning_rate": 0.00019143356623894882,
+      "loss": 2.7769,
+      "step": 2741
+    },
+    {
+      "epoch": 0.2212896457105964,
+      "grad_norm": 0.8453468680381775,
+      "learning_rate": 0.00019142717208240937,
+      "loss": 2.7677,
+      "step": 2742
+    },
+    {
+      "epoch": 0.2213703494471794,
+      "grad_norm": 0.8050780892372131,
+      "learning_rate": 0.00019142077564726582,
+      "loss": 2.7809,
+      "step": 2743
+    },
+    {
+      "epoch": 0.2214510531837624,
+      "grad_norm": 0.811287522315979,
+      "learning_rate": 0.0001914143769336776,
+      "loss": 2.7201,
+      "step": 2744
+    },
+    {
+      "epoch": 0.2215317569203454,
+      "grad_norm": 0.823106050491333,
+      "learning_rate": 0.00019140797594180412,
+      "loss": 2.7371,
+      "step": 2745
+    },
+    {
+      "epoch": 0.2216124606569284,
+      "grad_norm": 0.778126060962677,
+      "learning_rate": 0.0001914015726718049,
+      "loss": 2.6925,
+      "step": 2746
+    },
+    {
+      "epoch": 0.22169316439351142,
+      "grad_norm": 0.8240278959274292,
+      "learning_rate": 0.0001913951671238396,
+      "loss": 2.7227,
+      "step": 2747
+    },
+    {
+      "epoch": 0.22177386813009442,
+      "grad_norm": 0.8061805963516235,
+      "learning_rate": 0.0001913887592980678,
+      "loss": 2.7092,
+      "step": 2748
+    },
+    {
+      "epoch": 0.22185457186667742,
+      "grad_norm": 0.9111800789833069,
+      "learning_rate": 0.00019138234919464925,
+      "loss": 2.7364,
+      "step": 2749
+    },
+    {
+      "epoch": 0.22193527560326043,
+      "grad_norm": 0.8154863715171814,
+      "learning_rate": 0.0001913759368137437,
+      "loss": 2.6983,
+      "step": 2750
+    },
+    {
+      "epoch": 0.22201597933984343,
+      "grad_norm": 0.8547734022140503,
+      "learning_rate": 0.0001913695221555109,
+      "loss": 2.7016,
+      "step": 2751
+    },
+    {
+      "epoch": 0.22209668307642644,
+      "grad_norm": 0.7488531470298767,
+      "learning_rate": 0.00019136310522011079,
+      "loss": 2.6641,
+      "step": 2752
+    },
+    {
+      "epoch": 0.22217738681300944,
+      "grad_norm": 0.9118027091026306,
+      "learning_rate": 0.00019135668600770326,
+      "loss": 2.6965,
+      "step": 2753
+    },
+    {
+      "epoch": 0.22225809054959245,
+      "grad_norm": 0.7629117369651794,
+      "learning_rate": 0.00019135026451844834,
+      "loss": 2.7836,
+      "step": 2754
+    },
+    {
+      "epoch": 0.22233879428617545,
+      "grad_norm": 0.8081222176551819,
+      "learning_rate": 0.000191343840752506,
+      "loss": 2.7339,
+      "step": 2755
+    },
+    {
+      "epoch": 0.22241949802275846,
+      "grad_norm": 0.9143899083137512,
+      "learning_rate": 0.00019133741471003636,
+      "loss": 2.7051,
+      "step": 2756
+    },
+    {
+      "epoch": 0.22250020175934146,
+      "grad_norm": 0.8096790909767151,
+      "learning_rate": 0.00019133098639119962,
+      "loss": 2.6884,
+      "step": 2757
+    },
+    {
+      "epoch": 0.22258090549592446,
+      "grad_norm": 0.7959297895431519,
+      "learning_rate": 0.00019132455579615597,
+      "loss": 2.7127,
+      "step": 2758
+    },
+    {
+      "epoch": 0.22266160923250747,
+      "grad_norm": 0.7111356854438782,
+      "learning_rate": 0.00019131812292506563,
+      "loss": 2.7418,
+      "step": 2759
+    },
+    {
+      "epoch": 0.22274231296909047,
+      "grad_norm": 0.7584012150764465,
+      "learning_rate": 0.00019131168777808898,
+      "loss": 2.6705,
+      "step": 2760
+    },
+    {
+      "epoch": 0.22282301670567348,
+      "grad_norm": 0.7646663784980774,
+      "learning_rate": 0.0001913052503553864,
+      "loss": 2.7166,
+      "step": 2761
+    },
+    {
+      "epoch": 0.22290372044225648,
+      "grad_norm": 0.7643954157829285,
+      "learning_rate": 0.00019129881065711827,
+      "loss": 2.7967,
+      "step": 2762
+    },
+    {
+      "epoch": 0.22298442417883949,
+      "grad_norm": 0.7591429948806763,
+      "learning_rate": 0.0001912923686834451,
+      "loss": 2.6611,
+      "step": 2763
+    },
+    {
+      "epoch": 0.2230651279154225,
+      "grad_norm": 0.7182386517524719,
+      "learning_rate": 0.00019128592443452749,
+      "loss": 2.6808,
+      "step": 2764
+    },
+    {
+      "epoch": 0.2231458316520055,
+      "grad_norm": 0.7689648270606995,
+      "learning_rate": 0.00019127947791052602,
+      "loss": 2.7288,
+      "step": 2765
+    },
+    {
+      "epoch": 0.2232265353885885,
+      "grad_norm": 0.7851321697235107,
+      "learning_rate": 0.00019127302911160136,
+      "loss": 2.7227,
+      "step": 2766
+    },
+    {
+      "epoch": 0.2233072391251715,
+      "grad_norm": 0.8419411182403564,
+      "learning_rate": 0.00019126657803791424,
+      "loss": 2.7397,
+      "step": 2767
+    },
+    {
+      "epoch": 0.2233879428617545,
+      "grad_norm": 0.7657596468925476,
+      "learning_rate": 0.0001912601246896254,
+      "loss": 2.7223,
+      "step": 2768
+    },
+    {
+      "epoch": 0.2234686465983375,
+      "grad_norm": 0.8033619523048401,
+      "learning_rate": 0.00019125366906689567,
+      "loss": 2.7256,
+      "step": 2769
+    },
+    {
+      "epoch": 0.22354935033492052,
+      "grad_norm": 0.7784682512283325,
+      "learning_rate": 0.00019124721116988601,
+      "loss": 2.7692,
+      "step": 2770
+    },
+    {
+      "epoch": 0.22363005407150352,
+      "grad_norm": 0.7842707633972168,
+      "learning_rate": 0.00019124075099875731,
+      "loss": 2.7707,
+      "step": 2771
+    },
+    {
+      "epoch": 0.22371075780808652,
+      "grad_norm": 0.7864845395088196,
+      "learning_rate": 0.0001912342885536706,
+      "loss": 2.6912,
+      "step": 2772
+    },
+    {
+      "epoch": 0.22379146154466953,
+      "grad_norm": 0.8544312715530396,
+      "learning_rate": 0.0001912278238347869,
+      "loss": 2.8345,
+      "step": 2773
+    },
+    {
+      "epoch": 0.22387216528125253,
+      "grad_norm": 0.7210882306098938,
+      "learning_rate": 0.0001912213568422674,
+      "loss": 2.6933,
+      "step": 2774
+    },
+    {
+      "epoch": 0.22395286901783554,
+      "grad_norm": 0.8877022862434387,
+      "learning_rate": 0.00019121488757627318,
+      "loss": 2.7583,
+      "step": 2775
+    },
+    {
+      "epoch": 0.22403357275441854,
+      "grad_norm": 0.902886688709259,
+      "learning_rate": 0.00019120841603696554,
+      "loss": 2.8,
+      "step": 2776
+    },
+    {
+      "epoch": 0.22411427649100155,
+      "grad_norm": 0.771294355392456,
+      "learning_rate": 0.0001912019422245058,
+      "loss": 2.7712,
+      "step": 2777
+    },
+    {
+      "epoch": 0.22419498022758455,
+      "grad_norm": 0.7973463535308838,
+      "learning_rate": 0.0001911954661390552,
+      "loss": 2.7368,
+      "step": 2778
+    },
+    {
+      "epoch": 0.22427568396416755,
+      "grad_norm": 0.776836633682251,
+      "learning_rate": 0.00019118898778077524,
+      "loss": 2.7126,
+      "step": 2779
+    },
+    {
+      "epoch": 0.22435638770075053,
+      "grad_norm": 0.8286641240119934,
+      "learning_rate": 0.00019118250714982731,
+      "loss": 2.7148,
+      "step": 2780
+    },
+    {
+      "epoch": 0.22443709143733354,
+      "grad_norm": 0.7848700284957886,
+      "learning_rate": 0.00019117602424637294,
+      "loss": 2.7284,
+      "step": 2781
+    },
+    {
+      "epoch": 0.22451779517391654,
+      "grad_norm": 0.7658216953277588,
+      "learning_rate": 0.0001911695390705737,
+      "loss": 2.7186,
+      "step": 2782
+    },
+    {
+      "epoch": 0.22459849891049954,
+      "grad_norm": 0.7596792578697205,
+      "learning_rate": 0.00019116305162259124,
+      "loss": 2.6854,
+      "step": 2783
+    },
+    {
+      "epoch": 0.22467920264708255,
+      "grad_norm": 0.7901157140731812,
+      "learning_rate": 0.00019115656190258726,
+      "loss": 2.7347,
+      "step": 2784
+    },
+    {
+      "epoch": 0.22475990638366555,
+      "grad_norm": 0.7499287128448486,
+      "learning_rate": 0.00019115006991072346,
+      "loss": 2.7219,
+      "step": 2785
+    },
+    {
+      "epoch": 0.22484061012024856,
+      "grad_norm": 0.7427374124526978,
+      "learning_rate": 0.00019114357564716162,
+      "loss": 2.7147,
+      "step": 2786
+    },
+    {
+      "epoch": 0.22492131385683156,
+      "grad_norm": 0.8305855393409729,
+      "learning_rate": 0.00019113707911206363,
+      "loss": 2.7587,
+      "step": 2787
+    },
+    {
+      "epoch": 0.22500201759341457,
+      "grad_norm": 0.8266459703445435,
+      "learning_rate": 0.00019113058030559142,
+      "loss": 2.7275,
+      "step": 2788
+    },
+    {
+      "epoch": 0.22508272132999757,
+      "grad_norm": 0.7338323593139648,
+      "learning_rate": 0.0001911240792279069,
+      "loss": 2.762,
+      "step": 2789
+    },
+    {
+      "epoch": 0.22516342506658057,
+      "grad_norm": 0.7653434872627258,
+      "learning_rate": 0.00019111757587917216,
+      "loss": 2.6715,
+      "step": 2790
+    },
+    {
+      "epoch": 0.22524412880316358,
+      "grad_norm": 0.76301509141922,
+      "learning_rate": 0.00019111107025954923,
+      "loss": 2.698,
+      "step": 2791
+    },
+    {
+      "epoch": 0.22532483253974658,
+      "grad_norm": 0.7810547947883606,
+      "learning_rate": 0.00019110456236920024,
+      "loss": 2.7295,
+      "step": 2792
+    },
+    {
+      "epoch": 0.2254055362763296,
+      "grad_norm": 0.7885214686393738,
+      "learning_rate": 0.00019109805220828742,
+      "loss": 2.7724,
+      "step": 2793
+    },
+    {
+      "epoch": 0.2254862400129126,
+      "grad_norm": 0.8087031841278076,
+      "learning_rate": 0.00019109153977697301,
+      "loss": 2.7888,
+      "step": 2794
+    },
+    {
+      "epoch": 0.2255669437494956,
+      "grad_norm": 0.795101523399353,
+      "learning_rate": 0.00019108502507541933,
+      "loss": 2.6815,
+      "step": 2795
+    },
+    {
+      "epoch": 0.2256476474860786,
+      "grad_norm": 0.8337482213973999,
+      "learning_rate": 0.0001910785081037887,
+      "loss": 2.8192,
+      "step": 2796
+    },
+    {
+      "epoch": 0.2257283512226616,
+      "grad_norm": 0.8357288241386414,
+      "learning_rate": 0.00019107198886224357,
+      "loss": 2.7867,
+      "step": 2797
+    },
+    {
+      "epoch": 0.2258090549592446,
+      "grad_norm": 0.80678391456604,
+      "learning_rate": 0.00019106546735094644,
+      "loss": 2.7313,
+      "step": 2798
+    },
+    {
+      "epoch": 0.2258897586958276,
+      "grad_norm": 0.7481401562690735,
+      "learning_rate": 0.00019105894357005979,
+      "loss": 2.7073,
+      "step": 2799
+    },
+    {
+      "epoch": 0.22597046243241062,
+      "grad_norm": 0.8025074005126953,
+      "learning_rate": 0.00019105241751974622,
+      "loss": 2.6922,
+      "step": 2800
+    },
+    {
+      "epoch": 0.22605116616899362,
+      "grad_norm": 0.7308986186981201,
+      "learning_rate": 0.00019104588920016842,
+      "loss": 2.7511,
+      "step": 2801
+    },
+    {
+      "epoch": 0.22613186990557663,
+      "grad_norm": 0.7727689146995544,
+      "learning_rate": 0.00019103935861148905,
+      "loss": 2.707,
+      "step": 2802
+    },
+    {
+      "epoch": 0.22621257364215963,
+      "grad_norm": 0.8611076474189758,
+      "learning_rate": 0.0001910328257538709,
+      "loss": 2.8494,
+      "step": 2803
+    },
+    {
+      "epoch": 0.22629327737874264,
+      "grad_norm": 0.8487605452537537,
+      "learning_rate": 0.00019102629062747677,
+      "loss": 2.7698,
+      "step": 2804
+    },
+    {
+      "epoch": 0.22637398111532564,
+      "grad_norm": 0.7495502233505249,
+      "learning_rate": 0.00019101975323246952,
+      "loss": 2.7091,
+      "step": 2805
+    },
+    {
+      "epoch": 0.22645468485190864,
+      "grad_norm": 0.7334234118461609,
+      "learning_rate": 0.0001910132135690121,
+      "loss": 2.7375,
+      "step": 2806
+    },
+    {
+      "epoch": 0.22653538858849165,
+      "grad_norm": 0.879912257194519,
+      "learning_rate": 0.00019100667163726747,
+      "loss": 2.7278,
+      "step": 2807
+    },
+    {
+      "epoch": 0.22661609232507465,
+      "grad_norm": 0.8087306618690491,
+      "learning_rate": 0.0001910001274373987,
+      "loss": 2.8065,
+      "step": 2808
+    },
+    {
+      "epoch": 0.22669679606165766,
+      "grad_norm": 0.7548169493675232,
+      "learning_rate": 0.00019099358096956887,
+      "loss": 2.7235,
+      "step": 2809
+    },
+    {
+      "epoch": 0.22677749979824066,
+      "grad_norm": 0.7505785822868347,
+      "learning_rate": 0.00019098703223394118,
+      "loss": 2.6633,
+      "step": 2810
+    },
+    {
+      "epoch": 0.22685820353482367,
+      "grad_norm": 0.829075813293457,
+      "learning_rate": 0.00019098048123067875,
+      "loss": 2.7389,
+      "step": 2811
+    },
+    {
+      "epoch": 0.22693890727140667,
+      "grad_norm": 0.7731673121452332,
+      "learning_rate": 0.00019097392795994493,
+      "loss": 2.7639,
+      "step": 2812
+    },
+    {
+      "epoch": 0.22701961100798967,
+      "grad_norm": 0.7389004826545715,
+      "learning_rate": 0.00019096737242190303,
+      "loss": 2.717,
+      "step": 2813
+    },
+    {
+      "epoch": 0.22710031474457268,
+      "grad_norm": 0.7520460486412048,
+      "learning_rate": 0.0001909608146167164,
+      "loss": 2.7203,
+      "step": 2814
+    },
+    {
+      "epoch": 0.22718101848115568,
+      "grad_norm": 0.7272354364395142,
+      "learning_rate": 0.00019095425454454849,
+      "loss": 2.7306,
+      "step": 2815
+    },
+    {
+      "epoch": 0.2272617222177387,
+      "grad_norm": 0.7593528032302856,
+      "learning_rate": 0.00019094769220556282,
+      "loss": 2.7565,
+      "step": 2816
+    },
+    {
+      "epoch": 0.2273424259543217,
+      "grad_norm": 0.7312695384025574,
+      "learning_rate": 0.0001909411275999229,
+      "loss": 2.744,
+      "step": 2817
+    },
+    {
+      "epoch": 0.2274231296909047,
+      "grad_norm": 0.7483308911323547,
+      "learning_rate": 0.00019093456072779238,
+      "loss": 2.7938,
+      "step": 2818
+    },
+    {
+      "epoch": 0.2275038334274877,
+      "grad_norm": 0.8515620231628418,
+      "learning_rate": 0.00019092799158933486,
+      "loss": 2.7392,
+      "step": 2819
+    },
+    {
+      "epoch": 0.2275845371640707,
+      "grad_norm": 0.7119776606559753,
+      "learning_rate": 0.00019092142018471415,
+      "loss": 2.6985,
+      "step": 2820
+    },
+    {
+      "epoch": 0.2276652409006537,
+      "grad_norm": 0.7549445033073425,
+      "learning_rate": 0.00019091484651409394,
+      "loss": 2.7621,
+      "step": 2821
+    },
+    {
+      "epoch": 0.2277459446372367,
+      "grad_norm": 0.8728097081184387,
+      "learning_rate": 0.00019090827057763814,
+      "loss": 2.8321,
+      "step": 2822
+    },
+    {
+      "epoch": 0.22782664837381972,
+      "grad_norm": 0.755043089389801,
+      "learning_rate": 0.00019090169237551057,
+      "loss": 2.7341,
+      "step": 2823
+    },
+    {
+      "epoch": 0.22790735211040272,
+      "grad_norm": 0.7949401140213013,
+      "learning_rate": 0.00019089511190787523,
+      "loss": 2.7646,
+      "step": 2824
+    },
+    {
+      "epoch": 0.22798805584698573,
+      "grad_norm": 0.8027622103691101,
+      "learning_rate": 0.00019088852917489607,
+      "loss": 2.7606,
+      "step": 2825
+    },
+    {
+      "epoch": 0.22806875958356873,
+      "grad_norm": 0.8609418869018555,
+      "learning_rate": 0.0001908819441767372,
+      "loss": 2.7433,
+      "step": 2826
+    },
+    {
+      "epoch": 0.22814946332015174,
+      "grad_norm": 0.8021805882453918,
+      "learning_rate": 0.00019087535691356271,
+      "loss": 2.7723,
+      "step": 2827
+    },
+    {
+      "epoch": 0.22823016705673474,
+      "grad_norm": 0.8104252219200134,
+      "learning_rate": 0.00019086876738553675,
+      "loss": 2.7229,
+      "step": 2828
+    },
+    {
+      "epoch": 0.22831087079331774,
+      "grad_norm": 0.8714433908462524,
+      "learning_rate": 0.00019086217559282362,
+      "loss": 2.75,
+      "step": 2829
+    },
+    {
+      "epoch": 0.22839157452990075,
+      "grad_norm": 0.7598714828491211,
+      "learning_rate": 0.0001908555815355875,
+      "loss": 2.6979,
+      "step": 2830
+    },
+    {
+      "epoch": 0.22847227826648372,
+      "grad_norm": 0.859708309173584,
+      "learning_rate": 0.00019084898521399283,
+      "loss": 2.7863,
+      "step": 2831
+    },
+    {
+      "epoch": 0.22855298200306673,
+      "grad_norm": 0.7798011302947998,
+      "learning_rate": 0.00019084238662820397,
+      "loss": 2.7623,
+      "step": 2832
+    },
+    {
+      "epoch": 0.22863368573964973,
+      "grad_norm": 0.7869576811790466,
+      "learning_rate": 0.00019083578577838535,
+      "loss": 2.7341,
+      "step": 2833
+    },
+    {
+      "epoch": 0.22871438947623274,
+      "grad_norm": 0.7486738562583923,
+      "learning_rate": 0.0001908291826647015,
+      "loss": 2.7615,
+      "step": 2834
+    },
+    {
+      "epoch": 0.22879509321281574,
+      "grad_norm": 0.8270190954208374,
+      "learning_rate": 0.00019082257728731704,
+      "loss": 2.7515,
+      "step": 2835
+    },
+    {
+      "epoch": 0.22887579694939875,
+      "grad_norm": 0.9060254693031311,
+      "learning_rate": 0.00019081596964639648,
+      "loss": 2.874,
+      "step": 2836
+    },
+    {
+      "epoch": 0.22895650068598175,
+      "grad_norm": 0.7802320122718811,
+      "learning_rate": 0.00019080935974210458,
+      "loss": 2.7224,
+      "step": 2837
+    },
+    {
+      "epoch": 0.22903720442256476,
+      "grad_norm": 0.9513018131256104,
+      "learning_rate": 0.00019080274757460607,
+      "loss": 2.7168,
+      "step": 2838
+    },
+    {
+      "epoch": 0.22911790815914776,
+      "grad_norm": 0.7139711976051331,
+      "learning_rate": 0.0001907961331440657,
+      "loss": 2.676,
+      "step": 2839
+    },
+    {
+      "epoch": 0.22919861189573076,
+      "grad_norm": 0.8635632395744324,
+      "learning_rate": 0.00019078951645064838,
+      "loss": 2.6979,
+      "step": 2840
+    },
+    {
+      "epoch": 0.22927931563231377,
+      "grad_norm": 0.8823218941688538,
+      "learning_rate": 0.000190782897494519,
+      "loss": 2.7345,
+      "step": 2841
+    },
+    {
+      "epoch": 0.22936001936889677,
+      "grad_norm": 0.8139359354972839,
+      "learning_rate": 0.00019077627627584246,
+      "loss": 2.6988,
+      "step": 2842
+    },
+    {
+      "epoch": 0.22944072310547978,
+      "grad_norm": 0.8935994505882263,
+      "learning_rate": 0.00019076965279478383,
+      "loss": 2.7706,
+      "step": 2843
+    },
+    {
+      "epoch": 0.22952142684206278,
+      "grad_norm": 0.8362705111503601,
+      "learning_rate": 0.00019076302705150816,
+      "loss": 2.7593,
+      "step": 2844
+    },
+    {
+      "epoch": 0.22960213057864579,
+      "grad_norm": 0.7534157633781433,
+      "learning_rate": 0.00019075639904618066,
+      "loss": 2.7501,
+      "step": 2845
+    },
+    {
+      "epoch": 0.2296828343152288,
+      "grad_norm": 0.8826640248298645,
+      "learning_rate": 0.00019074976877896642,
+      "loss": 2.7758,
+      "step": 2846
+    },
+    {
+      "epoch": 0.2297635380518118,
+      "grad_norm": 0.8395571112632751,
+      "learning_rate": 0.0001907431362500307,
+      "loss": 2.7625,
+      "step": 2847
+    },
+    {
+      "epoch": 0.2298442417883948,
+      "grad_norm": 0.7927684783935547,
+      "learning_rate": 0.00019073650145953885,
+      "loss": 2.7392,
+      "step": 2848
+    },
+    {
+      "epoch": 0.2299249455249778,
+      "grad_norm": 0.823208749294281,
+      "learning_rate": 0.00019072986440765618,
+      "loss": 2.7259,
+      "step": 2849
+    },
+    {
+      "epoch": 0.2300056492615608,
+      "grad_norm": 0.889416515827179,
+      "learning_rate": 0.00019072322509454815,
+      "loss": 2.7539,
+      "step": 2850
+    },
+    {
+      "epoch": 0.2300863529981438,
+      "grad_norm": 0.7957748770713806,
+      "learning_rate": 0.0001907165835203802,
+      "loss": 2.7756,
+      "step": 2851
+    },
+    {
+      "epoch": 0.23016705673472682,
+      "grad_norm": 0.7924029231071472,
+      "learning_rate": 0.00019070993968531782,
+      "loss": 2.7439,
+      "step": 2852
+    },
+    {
+      "epoch": 0.23024776047130982,
+      "grad_norm": 0.7811052799224854,
+      "learning_rate": 0.0001907032935895266,
+      "loss": 2.7479,
+      "step": 2853
+    },
+    {
+      "epoch": 0.23032846420789282,
+      "grad_norm": 0.7973877191543579,
+      "learning_rate": 0.00019069664523317225,
+      "loss": 2.7502,
+      "step": 2854
+    },
+    {
+      "epoch": 0.23040916794447583,
+      "grad_norm": 0.7524267435073853,
+      "learning_rate": 0.0001906899946164204,
+      "loss": 2.75,
+      "step": 2855
+    },
+    {
+      "epoch": 0.23048987168105883,
+      "grad_norm": 0.7594791054725647,
+      "learning_rate": 0.00019068334173943683,
+      "loss": 2.6534,
+      "step": 2856
+    },
+    {
+      "epoch": 0.23057057541764184,
+      "grad_norm": 0.7253785729408264,
+      "learning_rate": 0.00019067668660238733,
+      "loss": 2.7246,
+      "step": 2857
+    },
+    {
+      "epoch": 0.23065127915422484,
+      "grad_norm": 0.788737416267395,
+      "learning_rate": 0.00019067002920543775,
+      "loss": 2.757,
+      "step": 2858
+    },
+    {
+      "epoch": 0.23073198289080785,
+      "grad_norm": 0.7577618956565857,
+      "learning_rate": 0.00019066336954875403,
+      "loss": 2.674,
+      "step": 2859
+    },
+    {
+      "epoch": 0.23081268662739085,
+      "grad_norm": 0.7682929635047913,
+      "learning_rate": 0.0001906567076325022,
+      "loss": 2.8193,
+      "step": 2860
+    },
+    {
+      "epoch": 0.23089339036397385,
+      "grad_norm": 0.7742112874984741,
+      "learning_rate": 0.00019065004345684817,
+      "loss": 2.6969,
+      "step": 2861
+    },
+    {
+      "epoch": 0.23097409410055686,
+      "grad_norm": 0.7981678247451782,
+      "learning_rate": 0.00019064337702195814,
+      "loss": 2.7681,
+      "step": 2862
+    },
+    {
+      "epoch": 0.23105479783713986,
+      "grad_norm": 0.7608500123023987,
+      "learning_rate": 0.00019063670832799817,
+      "loss": 2.7459,
+      "step": 2863
+    },
+    {
+      "epoch": 0.23113550157372287,
+      "grad_norm": 0.7563463449478149,
+      "learning_rate": 0.00019063003737513455,
+      "loss": 2.7678,
+      "step": 2864
+    },
+    {
+      "epoch": 0.23121620531030587,
+      "grad_norm": 0.7915034890174866,
+      "learning_rate": 0.00019062336416353343,
+      "loss": 2.7577,
+      "step": 2865
+    },
+    {
+      "epoch": 0.23129690904688888,
+      "grad_norm": 0.7229592204093933,
+      "learning_rate": 0.00019061668869336122,
+      "loss": 2.7308,
+      "step": 2866
+    },
+    {
+      "epoch": 0.23137761278347188,
+      "grad_norm": 0.7910905480384827,
+      "learning_rate": 0.00019061001096478425,
+      "loss": 2.7571,
+      "step": 2867
+    },
+    {
+      "epoch": 0.23145831652005489,
+      "grad_norm": 0.8474656939506531,
+      "learning_rate": 0.00019060333097796895,
+      "loss": 2.7011,
+      "step": 2868
+    },
+    {
+      "epoch": 0.2315390202566379,
+      "grad_norm": 0.8005419373512268,
+      "learning_rate": 0.00019059664873308178,
+      "loss": 2.7441,
+      "step": 2869
+    },
+    {
+      "epoch": 0.2316197239932209,
+      "grad_norm": 0.7728021740913391,
+      "learning_rate": 0.00019058996423028935,
+      "loss": 2.7753,
+      "step": 2870
+    },
+    {
+      "epoch": 0.2317004277298039,
+      "grad_norm": 0.7338094115257263,
+      "learning_rate": 0.00019058327746975816,
+      "loss": 2.7009,
+      "step": 2871
+    },
+    {
+      "epoch": 0.2317811314663869,
+      "grad_norm": 0.7746245265007019,
+      "learning_rate": 0.00019057658845165494,
+      "loss": 2.6938,
+      "step": 2872
+    },
+    {
+      "epoch": 0.2318618352029699,
+      "grad_norm": 0.7474356293678284,
+      "learning_rate": 0.00019056989717614636,
+      "loss": 2.7161,
+      "step": 2873
+    },
+    {
+      "epoch": 0.2319425389395529,
+      "grad_norm": 0.9540585279464722,
+      "learning_rate": 0.00019056320364339917,
+      "loss": 2.7753,
+      "step": 2874
+    },
+    {
+      "epoch": 0.23202324267613592,
+      "grad_norm": 0.799726665019989,
+      "learning_rate": 0.00019055650785358024,
+      "loss": 2.7301,
+      "step": 2875
+    },
+    {
+      "epoch": 0.23210394641271892,
+      "grad_norm": 0.8087828159332275,
+      "learning_rate": 0.0001905498098068564,
+      "loss": 2.7305,
+      "step": 2876
+    },
+    {
+      "epoch": 0.23218465014930192,
+      "grad_norm": 0.8177600502967834,
+      "learning_rate": 0.00019054310950339457,
+      "loss": 2.7462,
+      "step": 2877
+    },
+    {
+      "epoch": 0.23226535388588493,
+      "grad_norm": 0.7106238603591919,
+      "learning_rate": 0.00019053640694336181,
+      "loss": 2.7183,
+      "step": 2878
+    },
+    {
+      "epoch": 0.23234605762246793,
+      "grad_norm": 0.884185791015625,
+      "learning_rate": 0.00019052970212692514,
+      "loss": 2.7549,
+      "step": 2879
+    },
+    {
+      "epoch": 0.23242676135905094,
+      "grad_norm": 0.7532132267951965,
+      "learning_rate": 0.00019052299505425163,
+      "loss": 2.7524,
+      "step": 2880
+    },
+    {
+      "epoch": 0.23250746509563394,
+      "grad_norm": 0.7295021414756775,
+      "learning_rate": 0.00019051628572550842,
+      "loss": 2.6928,
+      "step": 2881
+    },
+    {
+      "epoch": 0.23258816883221692,
+      "grad_norm": 0.8475896716117859,
+      "learning_rate": 0.00019050957414086278,
+      "loss": 2.7138,
+      "step": 2882
+    },
+    {
+      "epoch": 0.23266887256879992,
+      "grad_norm": 0.7219378352165222,
+      "learning_rate": 0.00019050286030048198,
+      "loss": 2.7034,
+      "step": 2883
+    },
+    {
+      "epoch": 0.23274957630538293,
+      "grad_norm": 0.8410176634788513,
+      "learning_rate": 0.0001904961442045333,
+      "loss": 2.7413,
+      "step": 2884
+    },
+    {
+      "epoch": 0.23283028004196593,
+      "grad_norm": 0.7792301177978516,
+      "learning_rate": 0.00019048942585318414,
+      "loss": 2.6771,
+      "step": 2885
+    },
+    {
+      "epoch": 0.23291098377854894,
+      "grad_norm": 0.7457073926925659,
+      "learning_rate": 0.00019048270524660196,
+      "loss": 2.7325,
+      "step": 2886
+    },
+    {
+      "epoch": 0.23299168751513194,
+      "grad_norm": 0.8258858323097229,
+      "learning_rate": 0.00019047598238495424,
+      "loss": 2.7434,
+      "step": 2887
+    },
+    {
+      "epoch": 0.23307239125171494,
+      "grad_norm": 0.8188657164573669,
+      "learning_rate": 0.00019046925726840853,
+      "loss": 2.732,
+      "step": 2888
+    },
+    {
+      "epoch": 0.23315309498829795,
+      "grad_norm": 0.8084142208099365,
+      "learning_rate": 0.00019046252989713246,
+      "loss": 2.7537,
+      "step": 2889
+    },
+    {
+      "epoch": 0.23323379872488095,
+      "grad_norm": 0.75553297996521,
+      "learning_rate": 0.00019045580027129364,
+      "loss": 2.6685,
+      "step": 2890
+    },
+    {
+      "epoch": 0.23331450246146396,
+      "grad_norm": 0.8145995736122131,
+      "learning_rate": 0.00019044906839105986,
+      "loss": 2.7654,
+      "step": 2891
+    },
+    {
+      "epoch": 0.23339520619804696,
+      "grad_norm": 0.8433949947357178,
+      "learning_rate": 0.0001904423342565988,
+      "loss": 2.7713,
+      "step": 2892
+    },
+    {
+      "epoch": 0.23347590993462997,
+      "grad_norm": 0.7826054096221924,
+      "learning_rate": 0.0001904355978680784,
+      "loss": 2.7108,
+      "step": 2893
+    },
+    {
+      "epoch": 0.23355661367121297,
+      "grad_norm": 0.7281686663627625,
+      "learning_rate": 0.0001904288592256665,
+      "loss": 2.7606,
+      "step": 2894
+    },
+    {
+      "epoch": 0.23363731740779597,
+      "grad_norm": 0.8282813429832458,
+      "learning_rate": 0.00019042211832953103,
+      "loss": 2.6662,
+      "step": 2895
+    },
+    {
+      "epoch": 0.23371802114437898,
+      "grad_norm": 0.8227263689041138,
+      "learning_rate": 0.00019041537517984,
+      "loss": 2.7493,
+      "step": 2896
+    },
+    {
+      "epoch": 0.23379872488096198,
+      "grad_norm": 0.839350700378418,
+      "learning_rate": 0.0001904086297767615,
+      "loss": 2.7258,
+      "step": 2897
+    },
+    {
+      "epoch": 0.233879428617545,
+      "grad_norm": 0.713231086730957,
+      "learning_rate": 0.00019040188212046357,
+      "loss": 2.6722,
+      "step": 2898
+    },
+    {
+      "epoch": 0.233960132354128,
+      "grad_norm": 0.8314552903175354,
+      "learning_rate": 0.00019039513221111447,
+      "loss": 2.8509,
+      "step": 2899
+    },
+    {
+      "epoch": 0.234040836090711,
+      "grad_norm": 0.8885688781738281,
+      "learning_rate": 0.0001903883800488824,
+      "loss": 2.7608,
+      "step": 2900
+    },
+    {
+      "epoch": 0.234121539827294,
+      "grad_norm": 0.755308210849762,
+      "learning_rate": 0.00019038162563393555,
+      "loss": 2.7065,
+      "step": 2901
+    },
+    {
+      "epoch": 0.234202243563877,
+      "grad_norm": 0.7436641454696655,
+      "learning_rate": 0.00019037486896644236,
+      "loss": 2.6865,
+      "step": 2902
+    },
+    {
+      "epoch": 0.23428294730046,
+      "grad_norm": 0.7861987948417664,
+      "learning_rate": 0.0001903681100465712,
+      "loss": 2.7238,
+      "step": 2903
+    },
+    {
+      "epoch": 0.234363651037043,
+      "grad_norm": 0.7481045126914978,
+      "learning_rate": 0.0001903613488744905,
+      "loss": 2.7038,
+      "step": 2904
+    },
+    {
+      "epoch": 0.23444435477362602,
+      "grad_norm": 0.790765106678009,
+      "learning_rate": 0.0001903545854503688,
+      "loss": 2.6865,
+      "step": 2905
+    },
+    {
+      "epoch": 0.23452505851020902,
+      "grad_norm": 0.8594793677330017,
+      "learning_rate": 0.0001903478197743746,
+      "loss": 2.7324,
+      "step": 2906
+    },
+    {
+      "epoch": 0.23460576224679203,
+      "grad_norm": 0.7504310011863708,
+      "learning_rate": 0.00019034105184667662,
+      "loss": 2.6535,
+      "step": 2907
+    },
+    {
+      "epoch": 0.23468646598337503,
+      "grad_norm": 0.7824578881263733,
+      "learning_rate": 0.00019033428166744342,
+      "loss": 2.7113,
+      "step": 2908
+    },
+    {
+      "epoch": 0.23476716971995804,
+      "grad_norm": 0.7766899466514587,
+      "learning_rate": 0.0001903275092368438,
+      "loss": 2.6907,
+      "step": 2909
+    },
+    {
+      "epoch": 0.23484787345654104,
+      "grad_norm": 0.8082600235939026,
+      "learning_rate": 0.00019032073455504657,
+      "loss": 2.6781,
+      "step": 2910
+    },
+    {
+      "epoch": 0.23492857719312404,
+      "grad_norm": 0.7790517210960388,
+      "learning_rate": 0.0001903139576222205,
+      "loss": 2.7277,
+      "step": 2911
+    },
+    {
+      "epoch": 0.23500928092970705,
+      "grad_norm": 0.7449578046798706,
+      "learning_rate": 0.00019030717843853453,
+      "loss": 2.7078,
+      "step": 2912
+    },
+    {
+      "epoch": 0.23508998466629005,
+      "grad_norm": 0.7931632399559021,
+      "learning_rate": 0.0001903003970041576,
+      "loss": 2.7165,
+      "step": 2913
+    },
+    {
+      "epoch": 0.23517068840287306,
+      "grad_norm": 0.7970653176307678,
+      "learning_rate": 0.00019029361331925873,
+      "loss": 2.7993,
+      "step": 2914
+    },
+    {
+      "epoch": 0.23525139213945606,
+      "grad_norm": 0.8497335314750671,
+      "learning_rate": 0.00019028682738400697,
+      "loss": 2.7564,
+      "step": 2915
+    },
+    {
+      "epoch": 0.23533209587603907,
+      "grad_norm": 0.7840128540992737,
+      "learning_rate": 0.0001902800391985715,
+      "loss": 2.7546,
+      "step": 2916
+    },
+    {
+      "epoch": 0.23541279961262207,
+      "grad_norm": 0.8237372636795044,
+      "learning_rate": 0.00019027324876312146,
+      "loss": 2.7507,
+      "step": 2917
+    },
+    {
+      "epoch": 0.23549350334920507,
+      "grad_norm": 0.8445321917533875,
+      "learning_rate": 0.00019026645607782603,
+      "loss": 2.7287,
+      "step": 2918
+    },
+    {
+      "epoch": 0.23557420708578808,
+      "grad_norm": 0.8380417227745056,
+      "learning_rate": 0.0001902596611428546,
+      "loss": 2.7778,
+      "step": 2919
+    },
+    {
+      "epoch": 0.23565491082237108,
+      "grad_norm": 0.7989064455032349,
+      "learning_rate": 0.00019025286395837646,
+      "loss": 2.7254,
+      "step": 2920
+    },
+    {
+      "epoch": 0.2357356145589541,
+      "grad_norm": 0.8223496079444885,
+      "learning_rate": 0.00019024606452456102,
+      "loss": 2.7028,
+      "step": 2921
+    },
+    {
+      "epoch": 0.2358163182955371,
+      "grad_norm": 0.8090229630470276,
+      "learning_rate": 0.00019023926284157775,
+      "loss": 2.6911,
+      "step": 2922
+    },
+    {
+      "epoch": 0.2358970220321201,
+      "grad_norm": 0.7556560635566711,
+      "learning_rate": 0.00019023245890959615,
+      "loss": 2.7183,
+      "step": 2923
+    },
+    {
+      "epoch": 0.2359777257687031,
+      "grad_norm": 0.7907983660697937,
+      "learning_rate": 0.00019022565272878582,
+      "loss": 2.6805,
+      "step": 2924
+    },
+    {
+      "epoch": 0.2360584295052861,
+      "grad_norm": 0.9404142498970032,
+      "learning_rate": 0.0001902188442993164,
+      "loss": 2.8081,
+      "step": 2925
+    },
+    {
+      "epoch": 0.2361391332418691,
+      "grad_norm": 0.8349069952964783,
+      "learning_rate": 0.0001902120336213575,
+      "loss": 2.8329,
+      "step": 2926
+    },
+    {
+      "epoch": 0.2362198369784521,
+      "grad_norm": 0.8557522892951965,
+      "learning_rate": 0.00019020522069507892,
+      "loss": 2.704,
+      "step": 2927
+    },
+    {
+      "epoch": 0.23630054071503512,
+      "grad_norm": 0.7557278275489807,
+      "learning_rate": 0.00019019840552065044,
+      "loss": 2.7071,
+      "step": 2928
+    },
+    {
+      "epoch": 0.23638124445161812,
+      "grad_norm": 0.8810723423957825,
+      "learning_rate": 0.00019019158809824193,
+      "loss": 2.7535,
+      "step": 2929
+    },
+    {
+      "epoch": 0.23646194818820113,
+      "grad_norm": 0.7845562100410461,
+      "learning_rate": 0.00019018476842802326,
+      "loss": 2.7254,
+      "step": 2930
+    },
+    {
+      "epoch": 0.23654265192478413,
+      "grad_norm": 0.7566044926643372,
+      "learning_rate": 0.00019017794651016444,
+      "loss": 2.7295,
+      "step": 2931
+    },
+    {
+      "epoch": 0.23662335566136714,
+      "grad_norm": 0.8083382248878479,
+      "learning_rate": 0.00019017112234483545,
+      "loss": 2.7305,
+      "step": 2932
+    },
+    {
+      "epoch": 0.2367040593979501,
+      "grad_norm": 0.7924187183380127,
+      "learning_rate": 0.00019016429593220638,
+      "loss": 2.7659,
+      "step": 2933
+    },
+    {
+      "epoch": 0.23678476313453312,
+      "grad_norm": 0.8400307297706604,
+      "learning_rate": 0.00019015746727244737,
+      "loss": 2.7293,
+      "step": 2934
+    },
+    {
+      "epoch": 0.23686546687111612,
+      "grad_norm": 0.6931199431419373,
+      "learning_rate": 0.0001901506363657286,
+      "loss": 2.7189,
+      "step": 2935
+    },
+    {
+      "epoch": 0.23694617060769912,
+      "grad_norm": 0.8263585567474365,
+      "learning_rate": 0.0001901438032122203,
+      "loss": 2.7368,
+      "step": 2936
+    },
+    {
+      "epoch": 0.23702687434428213,
+      "grad_norm": 0.8001893162727356,
+      "learning_rate": 0.0001901369678120928,
+      "loss": 2.7793,
+      "step": 2937
+    },
+    {
+      "epoch": 0.23710757808086513,
+      "grad_norm": 0.7724235653877258,
+      "learning_rate": 0.00019013013016551644,
+      "loss": 2.717,
+      "step": 2938
+    },
+    {
+      "epoch": 0.23718828181744814,
+      "grad_norm": 0.7617147564888,
+      "learning_rate": 0.00019012329027266164,
+      "loss": 2.7275,
+      "step": 2939
+    },
+    {
+      "epoch": 0.23726898555403114,
+      "grad_norm": 0.80738765001297,
+      "learning_rate": 0.00019011644813369884,
+      "loss": 2.7444,
+      "step": 2940
+    },
+    {
+      "epoch": 0.23734968929061415,
+      "grad_norm": 0.7885528802871704,
+      "learning_rate": 0.00019010960374879861,
+      "loss": 2.7377,
+      "step": 2941
+    },
+    {
+      "epoch": 0.23743039302719715,
+      "grad_norm": 0.720268964767456,
+      "learning_rate": 0.00019010275711813147,
+      "loss": 2.6897,
+      "step": 2942
+    },
+    {
+      "epoch": 0.23751109676378016,
+      "grad_norm": 0.7532111406326294,
+      "learning_rate": 0.00019009590824186815,
+      "loss": 2.8117,
+      "step": 2943
+    },
+    {
+      "epoch": 0.23759180050036316,
+      "grad_norm": 0.780777633190155,
+      "learning_rate": 0.00019008905712017925,
+      "loss": 2.7565,
+      "step": 2944
+    },
+    {
+      "epoch": 0.23767250423694616,
+      "grad_norm": 0.8721919059753418,
+      "learning_rate": 0.00019008220375323553,
+      "loss": 2.801,
+      "step": 2945
+    },
+    {
+      "epoch": 0.23775320797352917,
+      "grad_norm": 0.8258914947509766,
+      "learning_rate": 0.00019007534814120786,
+      "loss": 2.7696,
+      "step": 2946
+    },
+    {
+      "epoch": 0.23783391171011217,
+      "grad_norm": 0.7292730808258057,
+      "learning_rate": 0.00019006849028426704,
+      "loss": 2.7512,
+      "step": 2947
+    },
+    {
+      "epoch": 0.23791461544669518,
+      "grad_norm": 0.7789164185523987,
+      "learning_rate": 0.00019006163018258398,
+      "loss": 2.7489,
+      "step": 2948
+    },
+    {
+      "epoch": 0.23799531918327818,
+      "grad_norm": 0.8049725294113159,
+      "learning_rate": 0.00019005476783632967,
+      "loss": 2.672,
+      "step": 2949
+    },
+    {
+      "epoch": 0.23807602291986119,
+      "grad_norm": 0.7440119981765747,
+      "learning_rate": 0.00019004790324567519,
+      "loss": 2.7208,
+      "step": 2950
+    },
+    {
+      "epoch": 0.2381567266564442,
+      "grad_norm": 0.7695925235748291,
+      "learning_rate": 0.00019004103641079154,
+      "loss": 2.7816,
+      "step": 2951
+    },
+    {
+      "epoch": 0.2382374303930272,
+      "grad_norm": 0.7623234391212463,
+      "learning_rate": 0.00019003416733184988,
+      "loss": 2.7034,
+      "step": 2952
+    },
+    {
+      "epoch": 0.2383181341296102,
+      "grad_norm": 0.8136502504348755,
+      "learning_rate": 0.00019002729600902141,
+      "loss": 2.7638,
+      "step": 2953
+    },
+    {
+      "epoch": 0.2383988378661932,
+      "grad_norm": 0.7813066840171814,
+      "learning_rate": 0.00019002042244247743,
+      "loss": 2.7606,
+      "step": 2954
+    },
+    {
+      "epoch": 0.2384795416027762,
+      "grad_norm": 0.7863059043884277,
+      "learning_rate": 0.0001900135466323892,
+      "loss": 2.7219,
+      "step": 2955
+    },
+    {
+      "epoch": 0.2385602453393592,
+      "grad_norm": 0.8712359070777893,
+      "learning_rate": 0.00019000666857892806,
+      "loss": 2.7485,
+      "step": 2956
+    },
+    {
+      "epoch": 0.23864094907594222,
+      "grad_norm": 0.8130611777305603,
+      "learning_rate": 0.00018999978828226547,
+      "loss": 2.7195,
+      "step": 2957
+    },
+    {
+      "epoch": 0.23872165281252522,
+      "grad_norm": 0.759503960609436,
+      "learning_rate": 0.00018999290574257292,
+      "loss": 2.6856,
+      "step": 2958
+    },
+    {
+      "epoch": 0.23880235654910822,
+      "grad_norm": 0.7490882277488708,
+      "learning_rate": 0.0001899860209600219,
+      "loss": 2.7587,
+      "step": 2959
+    },
+    {
+      "epoch": 0.23888306028569123,
+      "grad_norm": 0.8111297488212585,
+      "learning_rate": 0.000189979133934784,
+      "loss": 2.7688,
+      "step": 2960
+    },
+    {
+      "epoch": 0.23896376402227423,
+      "grad_norm": 0.844894289970398,
+      "learning_rate": 0.0001899722446670309,
+      "loss": 2.7706,
+      "step": 2961
+    },
+    {
+      "epoch": 0.23904446775885724,
+      "grad_norm": 0.7875459790229797,
+      "learning_rate": 0.00018996535315693423,
+      "loss": 2.7535,
+      "step": 2962
+    },
+    {
+      "epoch": 0.23912517149544024,
+      "grad_norm": 0.7768518328666687,
+      "learning_rate": 0.0001899584594046658,
+      "loss": 2.7268,
+      "step": 2963
+    },
+    {
+      "epoch": 0.23920587523202325,
+      "grad_norm": 0.8645716309547424,
+      "learning_rate": 0.00018995156341039744,
+      "loss": 2.7856,
+      "step": 2964
+    },
+    {
+      "epoch": 0.23928657896860625,
+      "grad_norm": 0.7816600799560547,
+      "learning_rate": 0.00018994466517430097,
+      "loss": 2.757,
+      "step": 2965
+    },
+    {
+      "epoch": 0.23936728270518925,
+      "grad_norm": 0.7967644333839417,
+      "learning_rate": 0.00018993776469654832,
+      "loss": 2.7021,
+      "step": 2966
+    },
+    {
+      "epoch": 0.23944798644177226,
+      "grad_norm": 0.800589919090271,
+      "learning_rate": 0.00018993086197731146,
+      "loss": 2.6838,
+      "step": 2967
+    },
+    {
+      "epoch": 0.23952869017835526,
+      "grad_norm": 0.7658529281616211,
+      "learning_rate": 0.00018992395701676246,
+      "loss": 2.6992,
+      "step": 2968
+    },
+    {
+      "epoch": 0.23960939391493827,
+      "grad_norm": 0.848456621170044,
+      "learning_rate": 0.00018991704981507338,
+      "loss": 2.7249,
+      "step": 2969
+    },
+    {
+      "epoch": 0.23969009765152127,
+      "grad_norm": 0.7365427017211914,
+      "learning_rate": 0.00018991014037241638,
+      "loss": 2.7044,
+      "step": 2970
+    },
+    {
+      "epoch": 0.23977080138810428,
+      "grad_norm": 0.8026351928710938,
+      "learning_rate": 0.00018990322868896365,
+      "loss": 2.7409,
+      "step": 2971
+    },
+    {
+      "epoch": 0.23985150512468728,
+      "grad_norm": 0.788646936416626,
+      "learning_rate": 0.00018989631476488744,
+      "loss": 2.7331,
+      "step": 2972
+    },
+    {
+      "epoch": 0.23993220886127029,
+      "grad_norm": 0.8388644456863403,
+      "learning_rate": 0.00018988939860036007,
+      "loss": 2.7478,
+      "step": 2973
+    },
+    {
+      "epoch": 0.2400129125978533,
+      "grad_norm": 0.7479026913642883,
+      "learning_rate": 0.00018988248019555394,
+      "loss": 2.7248,
+      "step": 2974
+    },
+    {
+      "epoch": 0.2400936163344363,
+      "grad_norm": 0.7313364744186401,
+      "learning_rate": 0.00018987555955064144,
+      "loss": 2.7323,
+      "step": 2975
+    },
+    {
+      "epoch": 0.2401743200710193,
+      "grad_norm": 0.7858260273933411,
+      "learning_rate": 0.00018986863666579505,
+      "loss": 2.6845,
+      "step": 2976
+    },
+    {
+      "epoch": 0.2402550238076023,
+      "grad_norm": 0.8090949654579163,
+      "learning_rate": 0.00018986171154118732,
+      "loss": 2.8094,
+      "step": 2977
+    },
+    {
+      "epoch": 0.2403357275441853,
+      "grad_norm": 0.7917135953903198,
+      "learning_rate": 0.00018985478417699085,
+      "loss": 2.7106,
+      "step": 2978
+    },
+    {
+      "epoch": 0.2404164312807683,
+      "grad_norm": 0.8192126154899597,
+      "learning_rate": 0.00018984785457337825,
+      "loss": 2.7729,
+      "step": 2979
+    },
+    {
+      "epoch": 0.24049713501735132,
+      "grad_norm": 0.797922670841217,
+      "learning_rate": 0.00018984092273052226,
+      "loss": 2.7747,
+      "step": 2980
+    },
+    {
+      "epoch": 0.24057783875393432,
+      "grad_norm": 0.9050948023796082,
+      "learning_rate": 0.00018983398864859564,
+      "loss": 2.7453,
+      "step": 2981
+    },
+    {
+      "epoch": 0.24065854249051732,
+      "grad_norm": 0.7827617526054382,
+      "learning_rate": 0.0001898270523277712,
+      "loss": 2.7371,
+      "step": 2982
+    },
+    {
+      "epoch": 0.24073924622710033,
+      "grad_norm": 0.7530156373977661,
+      "learning_rate": 0.0001898201137682218,
+      "loss": 2.7397,
+      "step": 2983
+    },
+    {
+      "epoch": 0.2408199499636833,
+      "grad_norm": 0.7989545464515686,
+      "learning_rate": 0.00018981317297012034,
+      "loss": 2.7532,
+      "step": 2984
+    },
+    {
+      "epoch": 0.2409006537002663,
+      "grad_norm": 0.7501168847084045,
+      "learning_rate": 0.00018980622993363988,
+      "loss": 2.7395,
+      "step": 2985
+    },
+    {
+      "epoch": 0.2409813574368493,
+      "grad_norm": 0.8073468208312988,
+      "learning_rate": 0.0001897992846589534,
+      "loss": 2.7673,
+      "step": 2986
+    },
+    {
+      "epoch": 0.24106206117343232,
+      "grad_norm": 0.9155512452125549,
+      "learning_rate": 0.00018979233714623401,
+      "loss": 2.6608,
+      "step": 2987
+    },
+    {
+      "epoch": 0.24114276491001532,
+      "grad_norm": 0.7461311221122742,
+      "learning_rate": 0.00018978538739565485,
+      "loss": 2.7657,
+      "step": 2988
+    },
+    {
+      "epoch": 0.24122346864659833,
+      "grad_norm": 0.8011443018913269,
+      "learning_rate": 0.00018977843540738914,
+      "loss": 2.7363,
+      "step": 2989
+    },
+    {
+      "epoch": 0.24130417238318133,
+      "grad_norm": 0.7602998614311218,
+      "learning_rate": 0.0001897714811816101,
+      "loss": 2.7285,
+      "step": 2990
+    },
+    {
+      "epoch": 0.24138487611976434,
+      "grad_norm": 0.8283531069755554,
+      "learning_rate": 0.00018976452471849116,
+      "loss": 2.7614,
+      "step": 2991
+    },
+    {
+      "epoch": 0.24146557985634734,
+      "grad_norm": 0.7358889579772949,
+      "learning_rate": 0.00018975756601820556,
+      "loss": 2.7429,
+      "step": 2992
+    },
+    {
+      "epoch": 0.24154628359293034,
+      "grad_norm": 0.7749240398406982,
+      "learning_rate": 0.0001897506050809268,
+      "loss": 2.6884,
+      "step": 2993
+    },
+    {
+      "epoch": 0.24162698732951335,
+      "grad_norm": 0.7529963254928589,
+      "learning_rate": 0.00018974364190682837,
+      "loss": 2.7619,
+      "step": 2994
+    },
+    {
+      "epoch": 0.24170769106609635,
+      "grad_norm": 0.7946054935455322,
+      "learning_rate": 0.00018973667649608376,
+      "loss": 2.7403,
+      "step": 2995
+    },
+    {
+      "epoch": 0.24178839480267936,
+      "grad_norm": 0.735870897769928,
+      "learning_rate": 0.0001897297088488666,
+      "loss": 2.7158,
+      "step": 2996
+    },
+    {
+      "epoch": 0.24186909853926236,
+      "grad_norm": 0.8409188985824585,
+      "learning_rate": 0.00018972273896535055,
+      "loss": 2.768,
+      "step": 2997
+    },
+    {
+      "epoch": 0.24194980227584537,
+      "grad_norm": 0.8351938724517822,
+      "learning_rate": 0.0001897157668457093,
+      "loss": 2.7548,
+      "step": 2998
+    },
+    {
+      "epoch": 0.24203050601242837,
+      "grad_norm": 0.8339046239852905,
+      "learning_rate": 0.00018970879249011663,
+      "loss": 2.7842,
+      "step": 2999
+    },
+    {
+      "epoch": 0.24211120974901137,
+      "grad_norm": 0.8092730641365051,
+      "learning_rate": 0.00018970181589874637,
+      "loss": 2.7141,
+      "step": 3000
+    },
+    {
+      "epoch": 0.24211120974901137,
+      "eval_loss": 2.643277406692505,
+      "eval_runtime": 784.7512,
+      "eval_samples_per_second": 3.339,
+      "eval_steps_per_second": 0.557,
+      "step": 3000
+    },
+    {
+      "epoch": 0.24219191348559438,
+      "grad_norm": 0.8014447093009949,
+      "learning_rate": 0.00018969483707177235,
+      "loss": 2.7341,
+      "step": 3001
+    },
+    {
+      "epoch": 0.24227261722217738,
+      "grad_norm": 0.744153618812561,
+      "learning_rate": 0.00018968785600936855,
+      "loss": 2.678,
+      "step": 3002
+    },
+    {
+      "epoch": 0.2423533209587604,
+      "grad_norm": 0.7264240384101868,
+      "learning_rate": 0.0001896808727117089,
+      "loss": 2.7321,
+      "step": 3003
+    },
+    {
+      "epoch": 0.2424340246953434,
+      "grad_norm": 0.8214067220687866,
+      "learning_rate": 0.00018967388717896748,
+      "loss": 2.7311,
+      "step": 3004
+    },
+    {
+      "epoch": 0.2425147284319264,
+      "grad_norm": 0.7871330976486206,
+      "learning_rate": 0.00018966689941131838,
+      "loss": 2.7184,
+      "step": 3005
+    },
+    {
+      "epoch": 0.2425954321685094,
+      "grad_norm": 0.7301360964775085,
+      "learning_rate": 0.00018965990940893575,
+      "loss": 2.7039,
+      "step": 3006
+    },
+    {
+      "epoch": 0.2426761359050924,
+      "grad_norm": 0.8290385603904724,
+      "learning_rate": 0.00018965291717199382,
+      "loss": 2.7848,
+      "step": 3007
+    },
+    {
+      "epoch": 0.2427568396416754,
+      "grad_norm": 0.7465909123420715,
+      "learning_rate": 0.00018964592270066683,
+      "loss": 2.7271,
+      "step": 3008
+    },
+    {
+      "epoch": 0.2428375433782584,
+      "grad_norm": 0.7992933988571167,
+      "learning_rate": 0.00018963892599512913,
+      "loss": 2.7749,
+      "step": 3009
+    },
+    {
+      "epoch": 0.24291824711484142,
+      "grad_norm": 0.7879100441932678,
+      "learning_rate": 0.00018963192705555507,
+      "loss": 2.6844,
+      "step": 3010
+    },
+    {
+      "epoch": 0.24299895085142442,
+      "grad_norm": 0.7895401120185852,
+      "learning_rate": 0.00018962492588211905,
+      "loss": 2.725,
+      "step": 3011
+    },
+    {
+      "epoch": 0.24307965458800743,
+      "grad_norm": 0.7699374556541443,
+      "learning_rate": 0.00018961792247499564,
+      "loss": 2.7408,
+      "step": 3012
+    },
+    {
+      "epoch": 0.24316035832459043,
+      "grad_norm": 0.828372597694397,
+      "learning_rate": 0.0001896109168343593,
+      "loss": 2.7527,
+      "step": 3013
+    },
+    {
+      "epoch": 0.24324106206117344,
+      "grad_norm": 0.7611951231956482,
+      "learning_rate": 0.0001896039089603847,
+      "loss": 2.7294,
+      "step": 3014
+    },
+    {
+      "epoch": 0.24332176579775644,
+      "grad_norm": 0.8214892148971558,
+      "learning_rate": 0.00018959689885324646,
+      "loss": 2.6931,
+      "step": 3015
+    },
+    {
+      "epoch": 0.24340246953433944,
+      "grad_norm": 0.7472538352012634,
+      "learning_rate": 0.00018958988651311928,
+      "loss": 2.7316,
+      "step": 3016
+    },
+    {
+      "epoch": 0.24348317327092245,
+      "grad_norm": 0.7574933171272278,
+      "learning_rate": 0.00018958287194017795,
+      "loss": 2.7764,
+      "step": 3017
+    },
+    {
+      "epoch": 0.24356387700750545,
+      "grad_norm": 0.739152729511261,
+      "learning_rate": 0.00018957585513459723,
+      "loss": 2.7949,
+      "step": 3018
+    },
+    {
+      "epoch": 0.24364458074408846,
+      "grad_norm": 0.824097752571106,
+      "learning_rate": 0.00018956883609655208,
+      "loss": 2.6612,
+      "step": 3019
+    },
+    {
+      "epoch": 0.24372528448067146,
+      "grad_norm": 0.7891144156455994,
+      "learning_rate": 0.00018956181482621744,
+      "loss": 2.7139,
+      "step": 3020
+    },
+    {
+      "epoch": 0.24380598821725447,
+      "grad_norm": 0.7364415526390076,
+      "learning_rate": 0.0001895547913237682,
+      "loss": 2.6984,
+      "step": 3021
+    },
+    {
+      "epoch": 0.24388669195383747,
+      "grad_norm": 0.7631362080574036,
+      "learning_rate": 0.0001895477655893795,
+      "loss": 2.7015,
+      "step": 3022
+    },
+    {
+      "epoch": 0.24396739569042047,
+      "grad_norm": 0.780541181564331,
+      "learning_rate": 0.00018954073762322637,
+      "loss": 2.7716,
+      "step": 3023
+    },
+    {
+      "epoch": 0.24404809942700348,
+      "grad_norm": 0.7877349853515625,
+      "learning_rate": 0.00018953370742548403,
+      "loss": 2.6654,
+      "step": 3024
+    },
+    {
+      "epoch": 0.24412880316358648,
+      "grad_norm": 0.7786216139793396,
+      "learning_rate": 0.00018952667499632763,
+      "loss": 2.7491,
+      "step": 3025
+    },
+    {
+      "epoch": 0.2442095069001695,
+      "grad_norm": 0.8207663893699646,
+      "learning_rate": 0.00018951964033593247,
+      "loss": 2.7212,
+      "step": 3026
+    },
+    {
+      "epoch": 0.2442902106367525,
+      "grad_norm": 0.8271831274032593,
+      "learning_rate": 0.00018951260344447386,
+      "loss": 2.7456,
+      "step": 3027
+    },
+    {
+      "epoch": 0.2443709143733355,
+      "grad_norm": 0.7610505819320679,
+      "learning_rate": 0.00018950556432212722,
+      "loss": 2.7472,
+      "step": 3028
+    },
+    {
+      "epoch": 0.2444516181099185,
+      "grad_norm": 0.7521701455116272,
+      "learning_rate": 0.00018949852296906792,
+      "loss": 2.7263,
+      "step": 3029
+    },
+    {
+      "epoch": 0.2445323218465015,
+      "grad_norm": 0.7518337965011597,
+      "learning_rate": 0.00018949147938547144,
+      "loss": 2.7069,
+      "step": 3030
+    },
+    {
+      "epoch": 0.2446130255830845,
+      "grad_norm": 0.7823107838630676,
+      "learning_rate": 0.00018948443357151343,
+      "loss": 2.7858,
+      "step": 3031
+    },
+    {
+      "epoch": 0.2446937293196675,
+      "grad_norm": 0.733132004737854,
+      "learning_rate": 0.00018947738552736938,
+      "loss": 2.7194,
+      "step": 3032
+    },
+    {
+      "epoch": 0.24477443305625052,
+      "grad_norm": 0.7756488919258118,
+      "learning_rate": 0.00018947033525321501,
+      "loss": 2.7299,
+      "step": 3033
+    },
+    {
+      "epoch": 0.24485513679283352,
+      "grad_norm": 0.7971112728118896,
+      "learning_rate": 0.00018946328274922598,
+      "loss": 2.7474,
+      "step": 3034
+    },
+    {
+      "epoch": 0.2449358405294165,
+      "grad_norm": 0.7871260643005371,
+      "learning_rate": 0.0001894562280155781,
+      "loss": 2.6994,
+      "step": 3035
+    },
+    {
+      "epoch": 0.2450165442659995,
+      "grad_norm": 0.7431116104125977,
+      "learning_rate": 0.00018944917105244717,
+      "loss": 2.6834,
+      "step": 3036
+    },
+    {
+      "epoch": 0.2450972480025825,
+      "grad_norm": 0.7372273206710815,
+      "learning_rate": 0.00018944211186000906,
+      "loss": 2.6988,
+      "step": 3037
+    },
+    {
+      "epoch": 0.2451779517391655,
+      "grad_norm": 0.8161508440971375,
+      "learning_rate": 0.00018943505043843975,
+      "loss": 2.7595,
+      "step": 3038
+    },
+    {
+      "epoch": 0.24525865547574852,
+      "grad_norm": 0.8062586784362793,
+      "learning_rate": 0.00018942798678791518,
+      "loss": 2.6893,
+      "step": 3039
+    },
+    {
+      "epoch": 0.24533935921233152,
+      "grad_norm": 0.824023425579071,
+      "learning_rate": 0.0001894209209086114,
+      "loss": 2.7188,
+      "step": 3040
+    },
+    {
+      "epoch": 0.24542006294891452,
+      "grad_norm": 0.740466833114624,
+      "learning_rate": 0.00018941385280070455,
+      "loss": 2.674,
+      "step": 3041
+    },
+    {
+      "epoch": 0.24550076668549753,
+      "grad_norm": 0.8543577194213867,
+      "learning_rate": 0.00018940678246437073,
+      "loss": 2.7423,
+      "step": 3042
+    },
+    {
+      "epoch": 0.24558147042208053,
+      "grad_norm": 0.7059324979782104,
+      "learning_rate": 0.0001893997098997862,
+      "loss": 2.6669,
+      "step": 3043
+    },
+    {
+      "epoch": 0.24566217415866354,
+      "grad_norm": 0.7739956974983215,
+      "learning_rate": 0.00018939263510712721,
+      "loss": 2.7118,
+      "step": 3044
+    },
+    {
+      "epoch": 0.24574287789524654,
+      "grad_norm": 0.7701205611228943,
+      "learning_rate": 0.00018938555808657007,
+      "loss": 2.7653,
+      "step": 3045
+    },
+    {
+      "epoch": 0.24582358163182955,
+      "grad_norm": 0.7243000864982605,
+      "learning_rate": 0.00018937847883829115,
+      "loss": 2.6789,
+      "step": 3046
+    },
+    {
+      "epoch": 0.24590428536841255,
+      "grad_norm": 0.7645598649978638,
+      "learning_rate": 0.00018937139736246693,
+      "loss": 2.7108,
+      "step": 3047
+    },
+    {
+      "epoch": 0.24598498910499556,
+      "grad_norm": 0.7544745802879333,
+      "learning_rate": 0.00018936431365927385,
+      "loss": 2.6958,
+      "step": 3048
+    },
+    {
+      "epoch": 0.24606569284157856,
+      "grad_norm": 0.709282398223877,
+      "learning_rate": 0.00018935722772888848,
+      "loss": 2.6728,
+      "step": 3049
+    },
+    {
+      "epoch": 0.24614639657816156,
+      "grad_norm": 0.7524243593215942,
+      "learning_rate": 0.00018935013957148742,
+      "loss": 2.7283,
+      "step": 3050
+    },
+    {
+      "epoch": 0.24622710031474457,
+      "grad_norm": 0.7959655523300171,
+      "learning_rate": 0.0001893430491872473,
+      "loss": 2.7384,
+      "step": 3051
+    },
+    {
+      "epoch": 0.24630780405132757,
+      "grad_norm": 0.7252553105354309,
+      "learning_rate": 0.00018933595657634486,
+      "loss": 2.7226,
+      "step": 3052
+    },
+    {
+      "epoch": 0.24638850778791058,
+      "grad_norm": 0.7387316226959229,
+      "learning_rate": 0.00018932886173895686,
+      "loss": 2.7546,
+      "step": 3053
+    },
+    {
+      "epoch": 0.24646921152449358,
+      "grad_norm": 0.804856538772583,
+      "learning_rate": 0.0001893217646752601,
+      "loss": 2.7321,
+      "step": 3054
+    },
+    {
+      "epoch": 0.24654991526107659,
+      "grad_norm": 0.6929069757461548,
+      "learning_rate": 0.0001893146653854315,
+      "loss": 2.6735,
+      "step": 3055
+    },
+    {
+      "epoch": 0.2466306189976596,
+      "grad_norm": 0.7076159715652466,
+      "learning_rate": 0.00018930756386964794,
+      "loss": 2.7368,
+      "step": 3056
+    },
+    {
+      "epoch": 0.2467113227342426,
+      "grad_norm": 0.7522851228713989,
+      "learning_rate": 0.00018930046012808648,
+      "loss": 2.7448,
+      "step": 3057
+    },
+    {
+      "epoch": 0.2467920264708256,
+      "grad_norm": 0.8347200155258179,
+      "learning_rate": 0.00018929335416092408,
+      "loss": 2.6837,
+      "step": 3058
+    },
+    {
+      "epoch": 0.2468727302074086,
+      "grad_norm": 0.737503707408905,
+      "learning_rate": 0.00018928624596833786,
+      "loss": 2.693,
+      "step": 3059
+    },
+    {
+      "epoch": 0.2469534339439916,
+      "grad_norm": 0.7836787104606628,
+      "learning_rate": 0.00018927913555050503,
+      "loss": 2.7335,
+      "step": 3060
+    },
+    {
+      "epoch": 0.2470341376805746,
+      "grad_norm": 0.7823840975761414,
+      "learning_rate": 0.00018927202290760278,
+      "loss": 2.6736,
+      "step": 3061
+    },
+    {
+      "epoch": 0.24711484141715762,
+      "grad_norm": 0.7894529700279236,
+      "learning_rate": 0.00018926490803980833,
+      "loss": 2.7112,
+      "step": 3062
+    },
+    {
+      "epoch": 0.24719554515374062,
+      "grad_norm": 0.8289024829864502,
+      "learning_rate": 0.000189257790947299,
+      "loss": 2.7667,
+      "step": 3063
+    },
+    {
+      "epoch": 0.24727624889032362,
+      "grad_norm": 0.70560222864151,
+      "learning_rate": 0.00018925067163025227,
+      "loss": 2.6946,
+      "step": 3064
+    },
+    {
+      "epoch": 0.24735695262690663,
+      "grad_norm": 0.6954196095466614,
+      "learning_rate": 0.00018924355008884548,
+      "loss": 2.7237,
+      "step": 3065
+    },
+    {
+      "epoch": 0.24743765636348963,
+      "grad_norm": 0.7975121736526489,
+      "learning_rate": 0.0001892364263232561,
+      "loss": 2.6392,
+      "step": 3066
+    },
+    {
+      "epoch": 0.24751836010007264,
+      "grad_norm": 0.777350902557373,
+      "learning_rate": 0.00018922930033366174,
+      "loss": 2.7284,
+      "step": 3067
+    },
+    {
+      "epoch": 0.24759906383665564,
+      "grad_norm": 0.738240659236908,
+      "learning_rate": 0.00018922217212023995,
+      "loss": 2.6884,
+      "step": 3068
+    },
+    {
+      "epoch": 0.24767976757323865,
+      "grad_norm": 0.8077268600463867,
+      "learning_rate": 0.0001892150416831684,
+      "loss": 2.7205,
+      "step": 3069
+    },
+    {
+      "epoch": 0.24776047130982165,
+      "grad_norm": 0.8108188509941101,
+      "learning_rate": 0.00018920790902262483,
+      "loss": 2.7592,
+      "step": 3070
+    },
+    {
+      "epoch": 0.24784117504640465,
+      "grad_norm": 0.7842642664909363,
+      "learning_rate": 0.00018920077413878695,
+      "loss": 2.7474,
+      "step": 3071
+    },
+    {
+      "epoch": 0.24792187878298766,
+      "grad_norm": 0.7644543051719666,
+      "learning_rate": 0.0001891936370318326,
+      "loss": 2.7179,
+      "step": 3072
+    },
+    {
+      "epoch": 0.24800258251957066,
+      "grad_norm": 0.7761854529380798,
+      "learning_rate": 0.00018918649770193965,
+      "loss": 2.71,
+      "step": 3073
+    },
+    {
+      "epoch": 0.24808328625615367,
+      "grad_norm": 0.7724074125289917,
+      "learning_rate": 0.00018917935614928607,
+      "loss": 2.7359,
+      "step": 3074
+    },
+    {
+      "epoch": 0.24816398999273667,
+      "grad_norm": 0.7360609173774719,
+      "learning_rate": 0.0001891722123740498,
+      "loss": 2.7342,
+      "step": 3075
+    },
+    {
+      "epoch": 0.24824469372931968,
+      "grad_norm": 0.757561206817627,
+      "learning_rate": 0.00018916506637640894,
+      "loss": 2.7647,
+      "step": 3076
+    },
+    {
+      "epoch": 0.24832539746590268,
+      "grad_norm": 0.7180947065353394,
+      "learning_rate": 0.00018915791815654148,
+      "loss": 2.6771,
+      "step": 3077
+    },
+    {
+      "epoch": 0.24840610120248569,
+      "grad_norm": 0.7219653129577637,
+      "learning_rate": 0.0001891507677146257,
+      "loss": 2.7772,
+      "step": 3078
+    },
+    {
+      "epoch": 0.2484868049390687,
+      "grad_norm": 0.749113917350769,
+      "learning_rate": 0.0001891436150508397,
+      "loss": 2.6996,
+      "step": 3079
+    },
+    {
+      "epoch": 0.2485675086756517,
+      "grad_norm": 0.766180157661438,
+      "learning_rate": 0.00018913646016536183,
+      "loss": 2.7896,
+      "step": 3080
+    },
+    {
+      "epoch": 0.2486482124122347,
+      "grad_norm": 0.7672411799430847,
+      "learning_rate": 0.00018912930305837032,
+      "loss": 2.7307,
+      "step": 3081
+    },
+    {
+      "epoch": 0.2487289161488177,
+      "grad_norm": 0.7639018297195435,
+      "learning_rate": 0.00018912214373004364,
+      "loss": 2.6569,
+      "step": 3082
+    },
+    {
+      "epoch": 0.2488096198854007,
+      "grad_norm": 0.8935483694076538,
+      "learning_rate": 0.00018911498218056013,
+      "loss": 2.6897,
+      "step": 3083
+    },
+    {
+      "epoch": 0.2488903236219837,
+      "grad_norm": 0.8506368398666382,
+      "learning_rate": 0.00018910781841009836,
+      "loss": 2.778,
+      "step": 3084
+    },
+    {
+      "epoch": 0.24897102735856672,
+      "grad_norm": 0.8026999235153198,
+      "learning_rate": 0.0001891006524188368,
+      "loss": 2.7799,
+      "step": 3085
+    },
+    {
+      "epoch": 0.2490517310951497,
+      "grad_norm": 0.784637987613678,
+      "learning_rate": 0.00018909348420695406,
+      "loss": 2.673,
+      "step": 3086
+    },
+    {
+      "epoch": 0.2491324348317327,
+      "grad_norm": 0.8949337601661682,
+      "learning_rate": 0.00018908631377462882,
+      "loss": 2.7726,
+      "step": 3087
+    },
+    {
+      "epoch": 0.2492131385683157,
+      "grad_norm": 0.73841792345047,
+      "learning_rate": 0.00018907914112203974,
+      "loss": 2.7403,
+      "step": 3088
+    },
+    {
+      "epoch": 0.2492938423048987,
+      "grad_norm": 0.7305924296379089,
+      "learning_rate": 0.00018907196624936564,
+      "loss": 2.6713,
+      "step": 3089
+    },
+    {
+      "epoch": 0.2493745460414817,
+      "grad_norm": 0.7707394361495972,
+      "learning_rate": 0.0001890647891567853,
+      "loss": 2.7306,
+      "step": 3090
+    },
+    {
+      "epoch": 0.2494552497780647,
+      "grad_norm": 0.8691473603248596,
+      "learning_rate": 0.00018905760984447759,
+      "loss": 2.6775,
+      "step": 3091
+    },
+    {
+      "epoch": 0.24953595351464772,
+      "grad_norm": 0.7466028332710266,
+      "learning_rate": 0.00018905042831262144,
+      "loss": 2.7196,
+      "step": 3092
+    },
+    {
+      "epoch": 0.24961665725123072,
+      "grad_norm": 0.7785150408744812,
+      "learning_rate": 0.0001890432445613958,
+      "loss": 2.7099,
+      "step": 3093
+    },
+    {
+      "epoch": 0.24969736098781373,
+      "grad_norm": 0.7775028347969055,
+      "learning_rate": 0.0001890360585909798,
+      "loss": 2.698,
+      "step": 3094
+    },
+    {
+      "epoch": 0.24977806472439673,
+      "grad_norm": 0.829257071018219,
+      "learning_rate": 0.00018902887040155245,
+      "loss": 2.711,
+      "step": 3095
+    },
+    {
+      "epoch": 0.24985876846097974,
+      "grad_norm": 0.8492234945297241,
+      "learning_rate": 0.00018902167999329295,
+      "loss": 2.7164,
+      "step": 3096
+    },
+    {
+      "epoch": 0.24993947219756274,
+      "grad_norm": 0.7332174777984619,
+      "learning_rate": 0.00018901448736638045,
+      "loss": 2.6925,
+      "step": 3097
+    },
+    {
+      "epoch": 0.25002017593414577,
+      "grad_norm": 0.7494251728057861,
+      "learning_rate": 0.00018900729252099426,
+      "loss": 2.6899,
+      "step": 3098
+    },
+    {
+      "epoch": 0.25010087967072875,
+      "grad_norm": 0.7760747075080872,
+      "learning_rate": 0.00018900009545731367,
+      "loss": 2.6626,
+      "step": 3099
+    },
+    {
+      "epoch": 0.2501815834073118,
+      "grad_norm": 0.7270001173019409,
+      "learning_rate": 0.00018899289617551804,
+      "loss": 2.7338,
+      "step": 3100
+    },
+    {
+      "epoch": 0.25026228714389476,
+      "grad_norm": 0.7832693457603455,
+      "learning_rate": 0.0001889856946757868,
+      "loss": 2.6668,
+      "step": 3101
+    },
+    {
+      "epoch": 0.2503429908804778,
+      "grad_norm": 0.8833239674568176,
+      "learning_rate": 0.00018897849095829945,
+      "loss": 2.7219,
+      "step": 3102
+    },
+    {
+      "epoch": 0.25042369461706077,
+      "grad_norm": 0.8144814372062683,
+      "learning_rate": 0.0001889712850232355,
+      "loss": 2.724,
+      "step": 3103
+    },
+    {
+      "epoch": 0.2505043983536438,
+      "grad_norm": 0.9466180801391602,
+      "learning_rate": 0.0001889640768707746,
+      "loss": 2.7499,
+      "step": 3104
+    },
+    {
+      "epoch": 0.2505851020902268,
+      "grad_norm": 0.926292359828949,
+      "learning_rate": 0.00018895686650109632,
+      "loss": 2.7391,
+      "step": 3105
+    },
+    {
+      "epoch": 0.2506658058268098,
+      "grad_norm": 0.8214002251625061,
+      "learning_rate": 0.00018894965391438038,
+      "loss": 2.7546,
+      "step": 3106
+    },
+    {
+      "epoch": 0.2507465095633928,
+      "grad_norm": 0.9021030068397522,
+      "learning_rate": 0.00018894243911080655,
+      "loss": 2.7188,
+      "step": 3107
+    },
+    {
+      "epoch": 0.2508272132999758,
+      "grad_norm": 0.778366208076477,
+      "learning_rate": 0.00018893522209055465,
+      "loss": 2.7852,
+      "step": 3108
+    },
+    {
+      "epoch": 0.2509079170365588,
+      "grad_norm": 0.8780209422111511,
+      "learning_rate": 0.00018892800285380456,
+      "loss": 2.7344,
+      "step": 3109
+    },
+    {
+      "epoch": 0.2509886207731418,
+      "grad_norm": 0.7581839561462402,
+      "learning_rate": 0.00018892078140073614,
+      "loss": 2.6697,
+      "step": 3110
+    },
+    {
+      "epoch": 0.2510693245097248,
+      "grad_norm": 0.7818635702133179,
+      "learning_rate": 0.00018891355773152944,
+      "loss": 2.6969,
+      "step": 3111
+    },
+    {
+      "epoch": 0.2511500282463078,
+      "grad_norm": 0.7528424859046936,
+      "learning_rate": 0.0001889063318463644,
+      "loss": 2.7359,
+      "step": 3112
+    },
+    {
+      "epoch": 0.2512307319828908,
+      "grad_norm": 0.8274288773536682,
+      "learning_rate": 0.0001888991037454212,
+      "loss": 2.7124,
+      "step": 3113
+    },
+    {
+      "epoch": 0.2513114357194738,
+      "grad_norm": 0.7186813354492188,
+      "learning_rate": 0.00018889187342888,
+      "loss": 2.7037,
+      "step": 3114
+    },
+    {
+      "epoch": 0.2513921394560568,
+      "grad_norm": 0.7458071112632751,
+      "learning_rate": 0.00018888464089692088,
+      "loss": 2.7178,
+      "step": 3115
+    },
+    {
+      "epoch": 0.2514728431926398,
+      "grad_norm": 0.7814257740974426,
+      "learning_rate": 0.00018887740614972418,
+      "loss": 2.7554,
+      "step": 3116
+    },
+    {
+      "epoch": 0.2515535469292228,
+      "grad_norm": 0.7706831097602844,
+      "learning_rate": 0.0001888701691874702,
+      "loss": 2.7441,
+      "step": 3117
+    },
+    {
+      "epoch": 0.2516342506658058,
+      "grad_norm": 0.8177775740623474,
+      "learning_rate": 0.0001888629300103393,
+      "loss": 2.7257,
+      "step": 3118
+    },
+    {
+      "epoch": 0.25171495440238884,
+      "grad_norm": 0.791097104549408,
+      "learning_rate": 0.00018885568861851188,
+      "loss": 2.6937,
+      "step": 3119
+    },
+    {
+      "epoch": 0.2517956581389718,
+      "grad_norm": 0.7521430850028992,
+      "learning_rate": 0.00018884844501216845,
+      "loss": 2.7723,
+      "step": 3120
+    },
+    {
+      "epoch": 0.25187636187555484,
+      "grad_norm": 0.8119359016418457,
+      "learning_rate": 0.00018884119919148948,
+      "loss": 2.7573,
+      "step": 3121
+    },
+    {
+      "epoch": 0.2519570656121378,
+      "grad_norm": 0.7579830288887024,
+      "learning_rate": 0.00018883395115665562,
+      "loss": 2.6943,
+      "step": 3122
+    },
+    {
+      "epoch": 0.25203776934872085,
+      "grad_norm": 0.7718791365623474,
+      "learning_rate": 0.00018882670090784748,
+      "loss": 2.6911,
+      "step": 3123
+    },
+    {
+      "epoch": 0.25211847308530383,
+      "grad_norm": 0.7718087434768677,
+      "learning_rate": 0.00018881944844524576,
+      "loss": 2.7505,
+      "step": 3124
+    },
+    {
+      "epoch": 0.25219917682188686,
+      "grad_norm": 0.7696875333786011,
+      "learning_rate": 0.0001888121937690312,
+      "loss": 2.7272,
+      "step": 3125
+    },
+    {
+      "epoch": 0.25227988055846984,
+      "grad_norm": 0.8082131743431091,
+      "learning_rate": 0.00018880493687938464,
+      "loss": 2.6677,
+      "step": 3126
+    },
+    {
+      "epoch": 0.25236058429505287,
+      "grad_norm": 0.857224702835083,
+      "learning_rate": 0.00018879767777648686,
+      "loss": 2.7237,
+      "step": 3127
+    },
+    {
+      "epoch": 0.25244128803163585,
+      "grad_norm": 0.8135749697685242,
+      "learning_rate": 0.00018879041646051886,
+      "loss": 2.7298,
+      "step": 3128
+    },
+    {
+      "epoch": 0.2525219917682189,
+      "grad_norm": 0.7772457003593445,
+      "learning_rate": 0.0001887831529316616,
+      "loss": 2.7723,
+      "step": 3129
+    },
+    {
+      "epoch": 0.25260269550480186,
+      "grad_norm": 0.795555055141449,
+      "learning_rate": 0.00018877588719009607,
+      "loss": 2.7207,
+      "step": 3130
+    },
+    {
+      "epoch": 0.2526833992413849,
+      "grad_norm": 0.7677939534187317,
+      "learning_rate": 0.00018876861923600337,
+      "loss": 2.6649,
+      "step": 3131
+    },
+    {
+      "epoch": 0.25276410297796786,
+      "grad_norm": 0.7706151008605957,
+      "learning_rate": 0.00018876134906956464,
+      "loss": 2.7154,
+      "step": 3132
+    },
+    {
+      "epoch": 0.2528448067145509,
+      "grad_norm": 0.8230584859848022,
+      "learning_rate": 0.00018875407669096105,
+      "loss": 2.7871,
+      "step": 3133
+    },
+    {
+      "epoch": 0.2529255104511339,
+      "grad_norm": 0.7037158608436584,
+      "learning_rate": 0.0001887468021003739,
+      "loss": 2.669,
+      "step": 3134
+    },
+    {
+      "epoch": 0.2530062141877169,
+      "grad_norm": 0.8485400080680847,
+      "learning_rate": 0.00018873952529798441,
+      "loss": 2.7517,
+      "step": 3135
+    },
+    {
+      "epoch": 0.2530869179242999,
+      "grad_norm": 0.7803399562835693,
+      "learning_rate": 0.000188732246283974,
+      "loss": 2.6987,
+      "step": 3136
+    },
+    {
+      "epoch": 0.2531676216608829,
+      "grad_norm": 0.7884016633033752,
+      "learning_rate": 0.0001887249650585241,
+      "loss": 2.7348,
+      "step": 3137
+    },
+    {
+      "epoch": 0.2532483253974659,
+      "grad_norm": 0.7794530987739563,
+      "learning_rate": 0.0001887176816218161,
+      "loss": 2.6934,
+      "step": 3138
+    },
+    {
+      "epoch": 0.2533290291340489,
+      "grad_norm": 0.7905173301696777,
+      "learning_rate": 0.00018871039597403156,
+      "loss": 2.714,
+      "step": 3139
+    },
+    {
+      "epoch": 0.2534097328706319,
+      "grad_norm": 0.7857949137687683,
+      "learning_rate": 0.0001887031081153521,
+      "loss": 2.7591,
+      "step": 3140
+    },
+    {
+      "epoch": 0.25349043660721493,
+      "grad_norm": 0.8602419495582581,
+      "learning_rate": 0.00018869581804595927,
+      "loss": 2.7819,
+      "step": 3141
+    },
+    {
+      "epoch": 0.2535711403437979,
+      "grad_norm": 0.7845202088356018,
+      "learning_rate": 0.00018868852576603483,
+      "loss": 2.6796,
+      "step": 3142
+    },
+    {
+      "epoch": 0.25365184408038094,
+      "grad_norm": 0.7600612640380859,
+      "learning_rate": 0.00018868123127576048,
+      "loss": 2.6785,
+      "step": 3143
+    },
+    {
+      "epoch": 0.2537325478169639,
+      "grad_norm": 0.7731521725654602,
+      "learning_rate": 0.000188673934575318,
+      "loss": 2.7435,
+      "step": 3144
+    },
+    {
+      "epoch": 0.25381325155354695,
+      "grad_norm": 0.8214225172996521,
+      "learning_rate": 0.0001886666356648893,
+      "loss": 2.7264,
+      "step": 3145
+    },
+    {
+      "epoch": 0.2538939552901299,
+      "grad_norm": 0.7623010277748108,
+      "learning_rate": 0.00018865933454465628,
+      "loss": 2.73,
+      "step": 3146
+    },
+    {
+      "epoch": 0.25397465902671296,
+      "grad_norm": 0.7864633798599243,
+      "learning_rate": 0.00018865203121480088,
+      "loss": 2.7654,
+      "step": 3147
+    },
+    {
+      "epoch": 0.25405536276329593,
+      "grad_norm": 0.7654051780700684,
+      "learning_rate": 0.0001886447256755051,
+      "loss": 2.7171,
+      "step": 3148
+    },
+    {
+      "epoch": 0.25413606649987897,
+      "grad_norm": 0.8045486211776733,
+      "learning_rate": 0.0001886374179269511,
+      "loss": 2.7385,
+      "step": 3149
+    },
+    {
+      "epoch": 0.25421677023646194,
+      "grad_norm": 0.8504971861839294,
+      "learning_rate": 0.0001886301079693209,
+      "loss": 2.6719,
+      "step": 3150
+    },
+    {
+      "epoch": 0.254297473973045,
+      "grad_norm": 0.771538496017456,
+      "learning_rate": 0.0001886227958027967,
+      "loss": 2.6707,
+      "step": 3151
+    },
+    {
+      "epoch": 0.25437817770962795,
+      "grad_norm": 0.8472220301628113,
+      "learning_rate": 0.0001886154814275608,
+      "loss": 2.7201,
+      "step": 3152
+    },
+    {
+      "epoch": 0.254458881446211,
+      "grad_norm": 0.7639158368110657,
+      "learning_rate": 0.00018860816484379545,
+      "loss": 2.76,
+      "step": 3153
+    },
+    {
+      "epoch": 0.25453958518279396,
+      "grad_norm": 0.8042064905166626,
+      "learning_rate": 0.000188600846051683,
+      "loss": 2.6862,
+      "step": 3154
+    },
+    {
+      "epoch": 0.254620288919377,
+      "grad_norm": 0.7481087446212769,
+      "learning_rate": 0.0001885935250514059,
+      "loss": 2.7394,
+      "step": 3155
+    },
+    {
+      "epoch": 0.25470099265595997,
+      "grad_norm": 0.7826097011566162,
+      "learning_rate": 0.00018858620184314653,
+      "loss": 2.596,
+      "step": 3156
+    },
+    {
+      "epoch": 0.254781696392543,
+      "grad_norm": 0.7477610111236572,
+      "learning_rate": 0.00018857887642708743,
+      "loss": 2.7385,
+      "step": 3157
+    },
+    {
+      "epoch": 0.254862400129126,
+      "grad_norm": 0.7347466945648193,
+      "learning_rate": 0.00018857154880341122,
+      "loss": 2.722,
+      "step": 3158
+    },
+    {
+      "epoch": 0.254943103865709,
+      "grad_norm": 0.7853806018829346,
+      "learning_rate": 0.00018856421897230048,
+      "loss": 2.7675,
+      "step": 3159
+    },
+    {
+      "epoch": 0.255023807602292,
+      "grad_norm": 0.7497034072875977,
+      "learning_rate": 0.0001885568869339379,
+      "loss": 2.6882,
+      "step": 3160
+    },
+    {
+      "epoch": 0.255104511338875,
+      "grad_norm": 0.7932263612747192,
+      "learning_rate": 0.0001885495526885062,
+      "loss": 2.7938,
+      "step": 3161
+    },
+    {
+      "epoch": 0.255185215075458,
+      "grad_norm": 0.7776823043823242,
+      "learning_rate": 0.00018854221623618815,
+      "loss": 2.6955,
+      "step": 3162
+    },
+    {
+      "epoch": 0.25526591881204097,
+      "grad_norm": 0.7564878463745117,
+      "learning_rate": 0.00018853487757716666,
+      "loss": 2.7644,
+      "step": 3163
+    },
+    {
+      "epoch": 0.255346622548624,
+      "grad_norm": 0.836270809173584,
+      "learning_rate": 0.00018852753671162454,
+      "loss": 2.7119,
+      "step": 3164
+    },
+    {
+      "epoch": 0.255427326285207,
+      "grad_norm": 0.7540388703346252,
+      "learning_rate": 0.00018852019363974485,
+      "loss": 2.797,
+      "step": 3165
+    },
+    {
+      "epoch": 0.25550803002179,
+      "grad_norm": 0.7943860292434692,
+      "learning_rate": 0.0001885128483617105,
+      "loss": 2.7973,
+      "step": 3166
+    },
+    {
+      "epoch": 0.255588733758373,
+      "grad_norm": 0.7743831276893616,
+      "learning_rate": 0.00018850550087770463,
+      "loss": 2.7403,
+      "step": 3167
+    },
+    {
+      "epoch": 0.255669437494956,
+      "grad_norm": 0.7593801021575928,
+      "learning_rate": 0.00018849815118791028,
+      "loss": 2.7203,
+      "step": 3168
+    },
+    {
+      "epoch": 0.255750141231539,
+      "grad_norm": 0.7663586139678955,
+      "learning_rate": 0.00018849079929251068,
+      "loss": 2.7481,
+      "step": 3169
+    },
+    {
+      "epoch": 0.25583084496812203,
+      "grad_norm": 0.7218170166015625,
+      "learning_rate": 0.00018848344519168905,
+      "loss": 2.6698,
+      "step": 3170
+    },
+    {
+      "epoch": 0.255911548704705,
+      "grad_norm": 0.8374441266059875,
+      "learning_rate": 0.00018847608888562868,
+      "loss": 2.8121,
+      "step": 3171
+    },
+    {
+      "epoch": 0.25599225244128804,
+      "grad_norm": 0.7488373517990112,
+      "learning_rate": 0.00018846873037451286,
+      "loss": 2.6871,
+      "step": 3172
+    },
+    {
+      "epoch": 0.256072956177871,
+      "grad_norm": 0.7513325810432434,
+      "learning_rate": 0.00018846136965852505,
+      "loss": 2.6924,
+      "step": 3173
+    },
+    {
+      "epoch": 0.25615365991445405,
+      "grad_norm": 0.7467690706253052,
+      "learning_rate": 0.00018845400673784865,
+      "loss": 2.714,
+      "step": 3174
+    },
+    {
+      "epoch": 0.256234363651037,
+      "grad_norm": 0.7717954516410828,
+      "learning_rate": 0.0001884466416126672,
+      "loss": 2.6679,
+      "step": 3175
+    },
+    {
+      "epoch": 0.25631506738762005,
+      "grad_norm": 0.7086547613143921,
+      "learning_rate": 0.0001884392742831642,
+      "loss": 2.7046,
+      "step": 3176
+    },
+    {
+      "epoch": 0.25639577112420303,
+      "grad_norm": 0.7024885416030884,
+      "learning_rate": 0.00018843190474952337,
+      "loss": 2.6724,
+      "step": 3177
+    },
+    {
+      "epoch": 0.25647647486078606,
+      "grad_norm": 0.8376390933990479,
+      "learning_rate": 0.00018842453301192827,
+      "loss": 2.7818,
+      "step": 3178
+    },
+    {
+      "epoch": 0.25655717859736904,
+      "grad_norm": 0.8190221190452576,
+      "learning_rate": 0.00018841715907056265,
+      "loss": 2.7455,
+      "step": 3179
+    },
+    {
+      "epoch": 0.25663788233395207,
+      "grad_norm": 0.8029047846794128,
+      "learning_rate": 0.0001884097829256103,
+      "loss": 2.7102,
+      "step": 3180
+    },
+    {
+      "epoch": 0.25671858607053505,
+      "grad_norm": 0.7467923760414124,
+      "learning_rate": 0.00018840240457725508,
+      "loss": 2.7051,
+      "step": 3181
+    },
+    {
+      "epoch": 0.2567992898071181,
+      "grad_norm": 0.7850394248962402,
+      "learning_rate": 0.00018839502402568086,
+      "loss": 2.6826,
+      "step": 3182
+    },
+    {
+      "epoch": 0.25687999354370106,
+      "grad_norm": 0.7144927978515625,
+      "learning_rate": 0.00018838764127107155,
+      "loss": 2.6694,
+      "step": 3183
+    },
+    {
+      "epoch": 0.2569606972802841,
+      "grad_norm": 0.7580311894416809,
+      "learning_rate": 0.0001883802563136112,
+      "loss": 2.7191,
+      "step": 3184
+    },
+    {
+      "epoch": 0.25704140101686707,
+      "grad_norm": 0.7366482615470886,
+      "learning_rate": 0.0001883728691534838,
+      "loss": 2.7175,
+      "step": 3185
+    },
+    {
+      "epoch": 0.2571221047534501,
+      "grad_norm": 0.6961715817451477,
+      "learning_rate": 0.0001883654797908735,
+      "loss": 2.7705,
+      "step": 3186
+    },
+    {
+      "epoch": 0.2572028084900331,
+      "grad_norm": 0.7473716735839844,
+      "learning_rate": 0.00018835808822596445,
+      "loss": 2.707,
+      "step": 3187
+    },
+    {
+      "epoch": 0.2572835122266161,
+      "grad_norm": 0.8376151919364929,
+      "learning_rate": 0.00018835069445894087,
+      "loss": 2.7424,
+      "step": 3188
+    },
+    {
+      "epoch": 0.2573642159631991,
+      "grad_norm": 0.7950237393379211,
+      "learning_rate": 0.00018834329848998706,
+      "loss": 2.7593,
+      "step": 3189
+    },
+    {
+      "epoch": 0.2574449196997821,
+      "grad_norm": 0.7637122869491577,
+      "learning_rate": 0.0001883359003192873,
+      "loss": 2.6708,
+      "step": 3190
+    },
+    {
+      "epoch": 0.2575256234363651,
+      "grad_norm": 0.709516704082489,
+      "learning_rate": 0.00018832849994702597,
+      "loss": 2.6988,
+      "step": 3191
+    },
+    {
+      "epoch": 0.2576063271729481,
+      "grad_norm": 0.7465435266494751,
+      "learning_rate": 0.00018832109737338757,
+      "loss": 2.7183,
+      "step": 3192
+    },
+    {
+      "epoch": 0.2576870309095311,
+      "grad_norm": 0.7619186043739319,
+      "learning_rate": 0.00018831369259855653,
+      "loss": 2.6833,
+      "step": 3193
+    },
+    {
+      "epoch": 0.25776773464611413,
+      "grad_norm": 0.7501961588859558,
+      "learning_rate": 0.0001883062856227174,
+      "loss": 2.725,
+      "step": 3194
+    },
+    {
+      "epoch": 0.2578484383826971,
+      "grad_norm": 0.7720133066177368,
+      "learning_rate": 0.00018829887644605483,
+      "loss": 2.7988,
+      "step": 3195
+    },
+    {
+      "epoch": 0.25792914211928014,
+      "grad_norm": 0.7253942489624023,
+      "learning_rate": 0.00018829146506875344,
+      "loss": 2.6999,
+      "step": 3196
+    },
+    {
+      "epoch": 0.2580098458558631,
+      "grad_norm": 0.7759599685668945,
+      "learning_rate": 0.00018828405149099792,
+      "loss": 2.6831,
+      "step": 3197
+    },
+    {
+      "epoch": 0.25809054959244615,
+      "grad_norm": 0.7250547409057617,
+      "learning_rate": 0.0001882766357129731,
+      "loss": 2.6742,
+      "step": 3198
+    },
+    {
+      "epoch": 0.2581712533290291,
+      "grad_norm": 0.7565183043479919,
+      "learning_rate": 0.00018826921773486372,
+      "loss": 2.6777,
+      "step": 3199
+    },
+    {
+      "epoch": 0.25825195706561216,
+      "grad_norm": 0.7183675169944763,
+      "learning_rate": 0.0001882617975568547,
+      "loss": 2.6743,
+      "step": 3200
+    },
+    {
+      "epoch": 0.25833266080219514,
+      "grad_norm": 0.7021663784980774,
+      "learning_rate": 0.00018825437517913098,
+      "loss": 2.727,
+      "step": 3201
+    },
+    {
+      "epoch": 0.25841336453877817,
+      "grad_norm": 0.7406932711601257,
+      "learning_rate": 0.00018824695060187753,
+      "loss": 2.7448,
+      "step": 3202
+    },
+    {
+      "epoch": 0.25849406827536114,
+      "grad_norm": 0.7766773104667664,
+      "learning_rate": 0.0001882395238252794,
+      "loss": 2.69,
+      "step": 3203
+    },
+    {
+      "epoch": 0.2585747720119442,
+      "grad_norm": 0.7483372688293457,
+      "learning_rate": 0.00018823209484952164,
+      "loss": 2.6611,
+      "step": 3204
+    },
+    {
+      "epoch": 0.25865547574852715,
+      "grad_norm": 0.781831681728363,
+      "learning_rate": 0.0001882246636747895,
+      "loss": 2.7292,
+      "step": 3205
+    },
+    {
+      "epoch": 0.2587361794851102,
+      "grad_norm": 0.7188203930854797,
+      "learning_rate": 0.00018821723030126806,
+      "loss": 2.718,
+      "step": 3206
+    },
+    {
+      "epoch": 0.25881688322169316,
+      "grad_norm": 0.7332054972648621,
+      "learning_rate": 0.00018820979472914263,
+      "loss": 2.6492,
+      "step": 3207
+    },
+    {
+      "epoch": 0.2588975869582762,
+      "grad_norm": 0.7044041156768799,
+      "learning_rate": 0.00018820235695859858,
+      "loss": 2.7047,
+      "step": 3208
+    },
+    {
+      "epoch": 0.25897829069485917,
+      "grad_norm": 0.8651862740516663,
+      "learning_rate": 0.00018819491698982121,
+      "loss": 2.6301,
+      "step": 3209
+    },
+    {
+      "epoch": 0.2590589944314422,
+      "grad_norm": 0.8118106126785278,
+      "learning_rate": 0.00018818747482299598,
+      "loss": 2.6522,
+      "step": 3210
+    },
+    {
+      "epoch": 0.2591396981680252,
+      "grad_norm": 0.7239218354225159,
+      "learning_rate": 0.00018818003045830832,
+      "loss": 2.7058,
+      "step": 3211
+    },
+    {
+      "epoch": 0.2592204019046082,
+      "grad_norm": 0.8557687997817993,
+      "learning_rate": 0.00018817258389594382,
+      "loss": 2.7125,
+      "step": 3212
+    },
+    {
+      "epoch": 0.2593011056411912,
+      "grad_norm": 0.7685148119926453,
+      "learning_rate": 0.00018816513513608801,
+      "loss": 2.7516,
+      "step": 3213
+    },
+    {
+      "epoch": 0.25938180937777416,
+      "grad_norm": 0.7497698664665222,
+      "learning_rate": 0.00018815768417892664,
+      "loss": 2.6536,
+      "step": 3214
+    },
+    {
+      "epoch": 0.2594625131143572,
+      "grad_norm": 0.7041923403739929,
+      "learning_rate": 0.0001881502310246453,
+      "loss": 2.7031,
+      "step": 3215
+    },
+    {
+      "epoch": 0.2595432168509402,
+      "grad_norm": 0.7815428376197815,
+      "learning_rate": 0.00018814277567342976,
+      "loss": 2.7291,
+      "step": 3216
+    },
+    {
+      "epoch": 0.2596239205875232,
+      "grad_norm": 0.7285065650939941,
+      "learning_rate": 0.00018813531812546583,
+      "loss": 2.7712,
+      "step": 3217
+    },
+    {
+      "epoch": 0.2597046243241062,
+      "grad_norm": 0.7606547474861145,
+      "learning_rate": 0.0001881278583809394,
+      "loss": 2.6714,
+      "step": 3218
+    },
+    {
+      "epoch": 0.2597853280606892,
+      "grad_norm": 0.7166680097579956,
+      "learning_rate": 0.00018812039644003638,
+      "loss": 2.7147,
+      "step": 3219
+    },
+    {
+      "epoch": 0.2598660317972722,
+      "grad_norm": 0.8977978229522705,
+      "learning_rate": 0.0001881129323029427,
+      "loss": 2.7743,
+      "step": 3220
+    },
+    {
+      "epoch": 0.2599467355338552,
+      "grad_norm": 0.7447277307510376,
+      "learning_rate": 0.00018810546596984446,
+      "loss": 2.7049,
+      "step": 3221
+    },
+    {
+      "epoch": 0.2600274392704382,
+      "grad_norm": 0.7343515157699585,
+      "learning_rate": 0.00018809799744092768,
+      "loss": 2.6999,
+      "step": 3222
+    },
+    {
+      "epoch": 0.26010814300702123,
+      "grad_norm": 0.7303341627120972,
+      "learning_rate": 0.00018809052671637852,
+      "loss": 2.7222,
+      "step": 3223
+    },
+    {
+      "epoch": 0.2601888467436042,
+      "grad_norm": 0.7412950396537781,
+      "learning_rate": 0.00018808305379638314,
+      "loss": 2.6957,
+      "step": 3224
+    },
+    {
+      "epoch": 0.26026955048018724,
+      "grad_norm": 0.7495343089103699,
+      "learning_rate": 0.00018807557868112781,
+      "loss": 2.7123,
+      "step": 3225
+    },
+    {
+      "epoch": 0.2603502542167702,
+      "grad_norm": 0.8137524724006653,
+      "learning_rate": 0.00018806810137079886,
+      "loss": 2.7191,
+      "step": 3226
+    },
+    {
+      "epoch": 0.26043095795335325,
+      "grad_norm": 0.786374568939209,
+      "learning_rate": 0.0001880606218655826,
+      "loss": 2.7237,
+      "step": 3227
+    },
+    {
+      "epoch": 0.2605116616899362,
+      "grad_norm": 0.9969484806060791,
+      "learning_rate": 0.00018805314016566543,
+      "loss": 2.7603,
+      "step": 3228
+    },
+    {
+      "epoch": 0.26059236542651926,
+      "grad_norm": 0.8132432103157043,
+      "learning_rate": 0.00018804565627123386,
+      "loss": 2.6807,
+      "step": 3229
+    },
+    {
+      "epoch": 0.26067306916310223,
+      "grad_norm": 0.7604904174804688,
+      "learning_rate": 0.00018803817018247436,
+      "loss": 2.7105,
+      "step": 3230
+    },
+    {
+      "epoch": 0.26075377289968527,
+      "grad_norm": 0.743505597114563,
+      "learning_rate": 0.00018803068189957354,
+      "loss": 2.7152,
+      "step": 3231
+    },
+    {
+      "epoch": 0.26083447663626824,
+      "grad_norm": 0.7780006527900696,
+      "learning_rate": 0.000188023191422718,
+      "loss": 2.7043,
+      "step": 3232
+    },
+    {
+      "epoch": 0.2609151803728513,
+      "grad_norm": 0.7683089375495911,
+      "learning_rate": 0.00018801569875209447,
+      "loss": 2.7033,
+      "step": 3233
+    },
+    {
+      "epoch": 0.26099588410943425,
+      "grad_norm": 0.7540118098258972,
+      "learning_rate": 0.0001880082038878896,
+      "loss": 2.7121,
+      "step": 3234
+    },
+    {
+      "epoch": 0.2610765878460173,
+      "grad_norm": 0.7509592771530151,
+      "learning_rate": 0.00018800070683029025,
+      "loss": 2.6575,
+      "step": 3235
+    },
+    {
+      "epoch": 0.26115729158260026,
+      "grad_norm": 0.8015461564064026,
+      "learning_rate": 0.00018799320757948327,
+      "loss": 2.6956,
+      "step": 3236
+    },
+    {
+      "epoch": 0.2612379953191833,
+      "grad_norm": 0.7586383819580078,
+      "learning_rate": 0.00018798570613565553,
+      "loss": 2.6719,
+      "step": 3237
+    },
+    {
+      "epoch": 0.26131869905576627,
+      "grad_norm": 0.7833155989646912,
+      "learning_rate": 0.000187978202498994,
+      "loss": 2.7317,
+      "step": 3238
+    },
+    {
+      "epoch": 0.2613994027923493,
+      "grad_norm": 0.7976018786430359,
+      "learning_rate": 0.00018797069666968565,
+      "loss": 2.7514,
+      "step": 3239
+    },
+    {
+      "epoch": 0.2614801065289323,
+      "grad_norm": 0.8388968706130981,
+      "learning_rate": 0.00018796318864791763,
+      "loss": 2.6845,
+      "step": 3240
+    },
+    {
+      "epoch": 0.2615608102655153,
+      "grad_norm": 0.8082842230796814,
+      "learning_rate": 0.00018795567843387701,
+      "loss": 2.7204,
+      "step": 3241
+    },
+    {
+      "epoch": 0.2616415140020983,
+      "grad_norm": 0.7514800429344177,
+      "learning_rate": 0.00018794816602775094,
+      "loss": 2.7117,
+      "step": 3242
+    },
+    {
+      "epoch": 0.2617222177386813,
+      "grad_norm": 0.8676564693450928,
+      "learning_rate": 0.00018794065142972664,
+      "loss": 2.6596,
+      "step": 3243
+    },
+    {
+      "epoch": 0.2618029214752643,
+      "grad_norm": 0.7449865341186523,
+      "learning_rate": 0.0001879331346399915,
+      "loss": 2.7089,
+      "step": 3244
+    },
+    {
+      "epoch": 0.2618836252118473,
+      "grad_norm": 0.8020811676979065,
+      "learning_rate": 0.00018792561565873274,
+      "loss": 2.7293,
+      "step": 3245
+    },
+    {
+      "epoch": 0.2619643289484303,
+      "grad_norm": 0.7961642146110535,
+      "learning_rate": 0.00018791809448613783,
+      "loss": 2.7269,
+      "step": 3246
+    },
+    {
+      "epoch": 0.26204503268501333,
+      "grad_norm": 0.7842351198196411,
+      "learning_rate": 0.00018791057112239415,
+      "loss": 2.6773,
+      "step": 3247
+    },
+    {
+      "epoch": 0.2621257364215963,
+      "grad_norm": 0.7494246959686279,
+      "learning_rate": 0.00018790304556768925,
+      "loss": 2.7317,
+      "step": 3248
+    },
+    {
+      "epoch": 0.26220644015817934,
+      "grad_norm": 0.7822836637496948,
+      "learning_rate": 0.0001878955178222107,
+      "loss": 2.6834,
+      "step": 3249
+    },
+    {
+      "epoch": 0.2622871438947623,
+      "grad_norm": 0.8432494401931763,
+      "learning_rate": 0.00018788798788614607,
+      "loss": 2.7048,
+      "step": 3250
+    },
+    {
+      "epoch": 0.26236784763134535,
+      "grad_norm": 0.9599446058273315,
+      "learning_rate": 0.000187880455759683,
+      "loss": 2.7793,
+      "step": 3251
+    },
+    {
+      "epoch": 0.26244855136792833,
+      "grad_norm": 0.8097226023674011,
+      "learning_rate": 0.00018787292144300928,
+      "loss": 2.7177,
+      "step": 3252
+    },
+    {
+      "epoch": 0.26252925510451136,
+      "grad_norm": 0.8423499464988708,
+      "learning_rate": 0.00018786538493631265,
+      "loss": 2.7265,
+      "step": 3253
+    },
+    {
+      "epoch": 0.26260995884109434,
+      "grad_norm": 0.7388847470283508,
+      "learning_rate": 0.00018785784623978095,
+      "loss": 2.6778,
+      "step": 3254
+    },
+    {
+      "epoch": 0.26269066257767737,
+      "grad_norm": 0.766368567943573,
+      "learning_rate": 0.0001878503053536021,
+      "loss": 2.654,
+      "step": 3255
+    },
+    {
+      "epoch": 0.26277136631426035,
+      "grad_norm": 0.8181266188621521,
+      "learning_rate": 0.00018784276227796394,
+      "loss": 2.7568,
+      "step": 3256
+    },
+    {
+      "epoch": 0.2628520700508434,
+      "grad_norm": 0.8235312104225159,
+      "learning_rate": 0.00018783521701305452,
+      "loss": 2.7317,
+      "step": 3257
+    },
+    {
+      "epoch": 0.26293277378742635,
+      "grad_norm": 0.7103183269500732,
+      "learning_rate": 0.00018782766955906195,
+      "loss": 2.6919,
+      "step": 3258
+    },
+    {
+      "epoch": 0.2630134775240094,
+      "grad_norm": 0.7202538251876831,
+      "learning_rate": 0.0001878201199161742,
+      "loss": 2.7179,
+      "step": 3259
+    },
+    {
+      "epoch": 0.26309418126059236,
+      "grad_norm": 0.8402286171913147,
+      "learning_rate": 0.00018781256808457952,
+      "loss": 2.7789,
+      "step": 3260
+    },
+    {
+      "epoch": 0.2631748849971754,
+      "grad_norm": 0.8136829137802124,
+      "learning_rate": 0.00018780501406446613,
+      "loss": 2.6872,
+      "step": 3261
+    },
+    {
+      "epoch": 0.26325558873375837,
+      "grad_norm": 0.8017000555992126,
+      "learning_rate": 0.00018779745785602224,
+      "loss": 2.7527,
+      "step": 3262
+    },
+    {
+      "epoch": 0.2633362924703414,
+      "grad_norm": 0.7880774140357971,
+      "learning_rate": 0.00018778989945943619,
+      "loss": 2.7348,
+      "step": 3263
+    },
+    {
+      "epoch": 0.2634169962069244,
+      "grad_norm": 0.7402438521385193,
+      "learning_rate": 0.00018778233887489635,
+      "loss": 2.6946,
+      "step": 3264
+    },
+    {
+      "epoch": 0.26349769994350736,
+      "grad_norm": 0.7450907230377197,
+      "learning_rate": 0.0001877747761025912,
+      "loss": 2.7502,
+      "step": 3265
+    },
+    {
+      "epoch": 0.2635784036800904,
+      "grad_norm": 0.7504056692123413,
+      "learning_rate": 0.00018776721114270917,
+      "loss": 2.832,
+      "step": 3266
+    },
+    {
+      "epoch": 0.26365910741667337,
+      "grad_norm": 0.7710226774215698,
+      "learning_rate": 0.00018775964399543878,
+      "loss": 2.6895,
+      "step": 3267
+    },
+    {
+      "epoch": 0.2637398111532564,
+      "grad_norm": 0.769927978515625,
+      "learning_rate": 0.00018775207466096867,
+      "loss": 2.6801,
+      "step": 3268
+    },
+    {
+      "epoch": 0.2638205148898394,
+      "grad_norm": 0.7210869193077087,
+      "learning_rate": 0.0001877445031394875,
+      "loss": 2.6966,
+      "step": 3269
+    },
+    {
+      "epoch": 0.2639012186264224,
+      "grad_norm": 0.7731119990348816,
+      "learning_rate": 0.00018773692943118393,
+      "loss": 2.6965,
+      "step": 3270
+    },
+    {
+      "epoch": 0.2639819223630054,
+      "grad_norm": 0.7539728283882141,
+      "learning_rate": 0.00018772935353624672,
+      "loss": 2.753,
+      "step": 3271
+    },
+    {
+      "epoch": 0.2640626260995884,
+      "grad_norm": 0.7993821501731873,
+      "learning_rate": 0.00018772177545486472,
+      "loss": 2.7177,
+      "step": 3272
+    },
+    {
+      "epoch": 0.2641433298361714,
+      "grad_norm": 0.7880005240440369,
+      "learning_rate": 0.00018771419518722672,
+      "loss": 2.6854,
+      "step": 3273
+    },
+    {
+      "epoch": 0.2642240335727544,
+      "grad_norm": 0.8079188466072083,
+      "learning_rate": 0.0001877066127335217,
+      "loss": 2.734,
+      "step": 3274
+    },
+    {
+      "epoch": 0.2643047373093374,
+      "grad_norm": 0.8241428732872009,
+      "learning_rate": 0.00018769902809393865,
+      "loss": 2.7156,
+      "step": 3275
+    },
+    {
+      "epoch": 0.26438544104592043,
+      "grad_norm": 0.8007158041000366,
+      "learning_rate": 0.00018769144126866657,
+      "loss": 2.693,
+      "step": 3276
+    },
+    {
+      "epoch": 0.2644661447825034,
+      "grad_norm": 0.8360451459884644,
+      "learning_rate": 0.00018768385225789456,
+      "loss": 2.6919,
+      "step": 3277
+    },
+    {
+      "epoch": 0.26454684851908644,
+      "grad_norm": 0.7596627473831177,
+      "learning_rate": 0.00018767626106181172,
+      "loss": 2.7861,
+      "step": 3278
+    },
+    {
+      "epoch": 0.2646275522556694,
+      "grad_norm": 0.7469248175621033,
+      "learning_rate": 0.00018766866768060727,
+      "loss": 2.7305,
+      "step": 3279
+    },
+    {
+      "epoch": 0.26470825599225245,
+      "grad_norm": 0.7103936076164246,
+      "learning_rate": 0.00018766107211447045,
+      "loss": 2.6456,
+      "step": 3280
+    },
+    {
+      "epoch": 0.2647889597288354,
+      "grad_norm": 0.7595266103744507,
+      "learning_rate": 0.00018765347436359056,
+      "loss": 2.7235,
+      "step": 3281
+    },
+    {
+      "epoch": 0.26486966346541846,
+      "grad_norm": 0.786648154258728,
+      "learning_rate": 0.00018764587442815698,
+      "loss": 2.7182,
+      "step": 3282
+    },
+    {
+      "epoch": 0.26495036720200144,
+      "grad_norm": 0.7152618169784546,
+      "learning_rate": 0.00018763827230835908,
+      "loss": 2.6842,
+      "step": 3283
+    },
+    {
+      "epoch": 0.26503107093858447,
+      "grad_norm": 0.89169842004776,
+      "learning_rate": 0.00018763066800438636,
+      "loss": 2.7661,
+      "step": 3284
+    },
+    {
+      "epoch": 0.26511177467516744,
+      "grad_norm": 0.8148171305656433,
+      "learning_rate": 0.00018762306151642833,
+      "loss": 2.7264,
+      "step": 3285
+    },
+    {
+      "epoch": 0.2651924784117505,
+      "grad_norm": 0.8070533871650696,
+      "learning_rate": 0.00018761545284467454,
+      "loss": 2.7425,
+      "step": 3286
+    },
+    {
+      "epoch": 0.26527318214833345,
+      "grad_norm": 0.8536118268966675,
+      "learning_rate": 0.00018760784198931465,
+      "loss": 2.702,
+      "step": 3287
+    },
+    {
+      "epoch": 0.2653538858849165,
+      "grad_norm": 0.7422329783439636,
+      "learning_rate": 0.00018760022895053833,
+      "loss": 2.6913,
+      "step": 3288
+    },
+    {
+      "epoch": 0.26543458962149946,
+      "grad_norm": 0.7415527105331421,
+      "learning_rate": 0.0001875926137285353,
+      "loss": 2.6472,
+      "step": 3289
+    },
+    {
+      "epoch": 0.2655152933580825,
+      "grad_norm": 0.8432031273841858,
+      "learning_rate": 0.00018758499632349538,
+      "loss": 2.7506,
+      "step": 3290
+    },
+    {
+      "epoch": 0.26559599709466547,
+      "grad_norm": 0.8113259077072144,
+      "learning_rate": 0.0001875773767356084,
+      "loss": 2.6866,
+      "step": 3291
+    },
+    {
+      "epoch": 0.2656767008312485,
+      "grad_norm": 0.7898122668266296,
+      "learning_rate": 0.00018756975496506424,
+      "loss": 2.6516,
+      "step": 3292
+    },
+    {
+      "epoch": 0.2657574045678315,
+      "grad_norm": 0.7627275586128235,
+      "learning_rate": 0.0001875621310120529,
+      "loss": 2.7065,
+      "step": 3293
+    },
+    {
+      "epoch": 0.2658381083044145,
+      "grad_norm": 0.8227291107177734,
+      "learning_rate": 0.00018755450487676435,
+      "loss": 2.7614,
+      "step": 3294
+    },
+    {
+      "epoch": 0.2659188120409975,
+      "grad_norm": 0.8162109851837158,
+      "learning_rate": 0.00018754687655938868,
+      "loss": 2.7924,
+      "step": 3295
+    },
+    {
+      "epoch": 0.2659995157775805,
+      "grad_norm": 0.7231846451759338,
+      "learning_rate": 0.00018753924606011602,
+      "loss": 2.7505,
+      "step": 3296
+    },
+    {
+      "epoch": 0.2660802195141635,
+      "grad_norm": 0.8635944724082947,
+      "learning_rate": 0.00018753161337913647,
+      "loss": 2.7505,
+      "step": 3297
+    },
+    {
+      "epoch": 0.26616092325074653,
+      "grad_norm": 0.8131890892982483,
+      "learning_rate": 0.00018752397851664031,
+      "loss": 2.7872,
+      "step": 3298
+    },
+    {
+      "epoch": 0.2662416269873295,
+      "grad_norm": 0.7336695790290833,
+      "learning_rate": 0.00018751634147281786,
+      "loss": 2.7517,
+      "step": 3299
+    },
+    {
+      "epoch": 0.26632233072391254,
+      "grad_norm": 0.7541754841804504,
+      "learning_rate": 0.00018750870224785939,
+      "loss": 2.7807,
+      "step": 3300
+    },
+    {
+      "epoch": 0.2664030344604955,
+      "grad_norm": 0.9347110390663147,
+      "learning_rate": 0.0001875010608419553,
+      "loss": 2.6954,
+      "step": 3301
+    },
+    {
+      "epoch": 0.26648373819707855,
+      "grad_norm": 0.7591213583946228,
+      "learning_rate": 0.00018749341725529604,
+      "loss": 2.7019,
+      "step": 3302
+    },
+    {
+      "epoch": 0.2665644419336615,
+      "grad_norm": 0.811527669429779,
+      "learning_rate": 0.00018748577148807211,
+      "loss": 2.7123,
+      "step": 3303
+    },
+    {
+      "epoch": 0.26664514567024455,
+      "grad_norm": 0.7419980764389038,
+      "learning_rate": 0.00018747812354047408,
+      "loss": 2.7383,
+      "step": 3304
+    },
+    {
+      "epoch": 0.26672584940682753,
+      "grad_norm": 0.7801192402839661,
+      "learning_rate": 0.00018747047341269256,
+      "loss": 2.7245,
+      "step": 3305
+    },
+    {
+      "epoch": 0.26680655314341056,
+      "grad_norm": 0.7392756938934326,
+      "learning_rate": 0.00018746282110491816,
+      "loss": 2.6992,
+      "step": 3306
+    },
+    {
+      "epoch": 0.26688725687999354,
+      "grad_norm": 0.7085927724838257,
+      "learning_rate": 0.00018745516661734161,
+      "loss": 2.739,
+      "step": 3307
+    },
+    {
+      "epoch": 0.26696796061657657,
+      "grad_norm": 0.7218676209449768,
+      "learning_rate": 0.00018744750995015373,
+      "loss": 2.7091,
+      "step": 3308
+    },
+    {
+      "epoch": 0.26704866435315955,
+      "grad_norm": 0.847872257232666,
+      "learning_rate": 0.0001874398511035453,
+      "loss": 2.699,
+      "step": 3309
+    },
+    {
+      "epoch": 0.2671293680897426,
+      "grad_norm": 0.8280770778656006,
+      "learning_rate": 0.00018743219007770723,
+      "loss": 2.763,
+      "step": 3310
+    },
+    {
+      "epoch": 0.26721007182632556,
+      "grad_norm": 0.7271165251731873,
+      "learning_rate": 0.0001874245268728304,
+      "loss": 2.7219,
+      "step": 3311
+    },
+    {
+      "epoch": 0.2672907755629086,
+      "grad_norm": 0.7342363595962524,
+      "learning_rate": 0.00018741686148910586,
+      "loss": 2.6765,
+      "step": 3312
+    },
+    {
+      "epoch": 0.26737147929949157,
+      "grad_norm": 0.7260174751281738,
+      "learning_rate": 0.0001874091939267246,
+      "loss": 2.7003,
+      "step": 3313
+    },
+    {
+      "epoch": 0.2674521830360746,
+      "grad_norm": 0.742494523525238,
+      "learning_rate": 0.00018740152418587775,
+      "loss": 2.7371,
+      "step": 3314
+    },
+    {
+      "epoch": 0.2675328867726576,
+      "grad_norm": 0.7238131165504456,
+      "learning_rate": 0.00018739385226675646,
+      "loss": 2.7486,
+      "step": 3315
+    },
+    {
+      "epoch": 0.26761359050924055,
+      "grad_norm": 0.7329363226890564,
+      "learning_rate": 0.0001873861781695519,
+      "loss": 2.6414,
+      "step": 3316
+    },
+    {
+      "epoch": 0.2676942942458236,
+      "grad_norm": 0.7078117728233337,
+      "learning_rate": 0.00018737850189445534,
+      "loss": 2.7271,
+      "step": 3317
+    },
+    {
+      "epoch": 0.26777499798240656,
+      "grad_norm": 0.7945309281349182,
+      "learning_rate": 0.00018737082344165814,
+      "loss": 2.7323,
+      "step": 3318
+    },
+    {
+      "epoch": 0.2678557017189896,
+      "grad_norm": 0.7510890364646912,
+      "learning_rate": 0.0001873631428113516,
+      "loss": 2.6563,
+      "step": 3319
+    },
+    {
+      "epoch": 0.26793640545557257,
+      "grad_norm": 0.7790820002555847,
+      "learning_rate": 0.0001873554600037272,
+      "loss": 2.7445,
+      "step": 3320
+    },
+    {
+      "epoch": 0.2680171091921556,
+      "grad_norm": 0.7689393162727356,
+      "learning_rate": 0.00018734777501897636,
+      "loss": 2.669,
+      "step": 3321
+    },
+    {
+      "epoch": 0.2680978129287386,
+      "grad_norm": 0.8227118253707886,
+      "learning_rate": 0.00018734008785729065,
+      "loss": 2.7279,
+      "step": 3322
+    },
+    {
+      "epoch": 0.2681785166653216,
+      "grad_norm": 0.7551290392875671,
+      "learning_rate": 0.00018733239851886162,
+      "loss": 2.6864,
+      "step": 3323
+    },
+    {
+      "epoch": 0.2682592204019046,
+      "grad_norm": 0.8572004437446594,
+      "learning_rate": 0.00018732470700388097,
+      "loss": 2.8159,
+      "step": 3324
+    },
+    {
+      "epoch": 0.2683399241384876,
+      "grad_norm": 0.7509044408798218,
+      "learning_rate": 0.00018731701331254033,
+      "loss": 2.7698,
+      "step": 3325
+    },
+    {
+      "epoch": 0.2684206278750706,
+      "grad_norm": 0.8474129438400269,
+      "learning_rate": 0.00018730931744503148,
+      "loss": 2.6745,
+      "step": 3326
+    },
+    {
+      "epoch": 0.2685013316116536,
+      "grad_norm": 0.8310953378677368,
+      "learning_rate": 0.00018730161940154618,
+      "loss": 2.712,
+      "step": 3327
+    },
+    {
+      "epoch": 0.2685820353482366,
+      "grad_norm": 0.8820717334747314,
+      "learning_rate": 0.00018729391918227632,
+      "loss": 2.7776,
+      "step": 3328
+    },
+    {
+      "epoch": 0.26866273908481964,
+      "grad_norm": 0.8827663064002991,
+      "learning_rate": 0.00018728621678741384,
+      "loss": 2.7115,
+      "step": 3329
+    },
+    {
+      "epoch": 0.2687434428214026,
+      "grad_norm": 0.7896323800086975,
+      "learning_rate": 0.00018727851221715064,
+      "loss": 2.6799,
+      "step": 3330
+    },
+    {
+      "epoch": 0.26882414655798564,
+      "grad_norm": 0.7775614261627197,
+      "learning_rate": 0.0001872708054716788,
+      "loss": 2.7021,
+      "step": 3331
+    },
+    {
+      "epoch": 0.2689048502945686,
+      "grad_norm": 0.8150187134742737,
+      "learning_rate": 0.0001872630965511903,
+      "loss": 2.679,
+      "step": 3332
+    },
+    {
+      "epoch": 0.26898555403115165,
+      "grad_norm": 0.7821844220161438,
+      "learning_rate": 0.00018725538545587736,
+      "loss": 2.7067,
+      "step": 3333
+    },
+    {
+      "epoch": 0.26906625776773463,
+      "grad_norm": 0.8390234112739563,
+      "learning_rate": 0.00018724767218593216,
+      "loss": 2.7133,
+      "step": 3334
+    },
+    {
+      "epoch": 0.26914696150431766,
+      "grad_norm": 0.8150694370269775,
+      "learning_rate": 0.00018723995674154687,
+      "loss": 2.7022,
+      "step": 3335
+    },
+    {
+      "epoch": 0.26922766524090064,
+      "grad_norm": 0.7473872900009155,
+      "learning_rate": 0.0001872322391229138,
+      "loss": 2.7268,
+      "step": 3336
+    },
+    {
+      "epoch": 0.26930836897748367,
+      "grad_norm": 0.7591951489448547,
+      "learning_rate": 0.0001872245193302253,
+      "loss": 2.7516,
+      "step": 3337
+    },
+    {
+      "epoch": 0.26938907271406665,
+      "grad_norm": 0.7914662957191467,
+      "learning_rate": 0.00018721679736367382,
+      "loss": 2.6613,
+      "step": 3338
+    },
+    {
+      "epoch": 0.2694697764506497,
+      "grad_norm": 0.7823428511619568,
+      "learning_rate": 0.00018720907322345172,
+      "loss": 2.6661,
+      "step": 3339
+    },
+    {
+      "epoch": 0.26955048018723266,
+      "grad_norm": 0.8428264260292053,
+      "learning_rate": 0.00018720134690975156,
+      "loss": 2.672,
+      "step": 3340
+    },
+    {
+      "epoch": 0.2696311839238157,
+      "grad_norm": 0.71320641040802,
+      "learning_rate": 0.00018719361842276587,
+      "loss": 2.7326,
+      "step": 3341
+    },
+    {
+      "epoch": 0.26971188766039866,
+      "grad_norm": 0.7972821593284607,
+      "learning_rate": 0.00018718588776268731,
+      "loss": 2.7182,
+      "step": 3342
+    },
+    {
+      "epoch": 0.2697925913969817,
+      "grad_norm": 0.7924500107765198,
+      "learning_rate": 0.0001871781549297085,
+      "loss": 2.7308,
+      "step": 3343
+    },
+    {
+      "epoch": 0.2698732951335647,
+      "grad_norm": 0.7668356895446777,
+      "learning_rate": 0.0001871704199240222,
+      "loss": 2.678,
+      "step": 3344
+    },
+    {
+      "epoch": 0.2699539988701477,
+      "grad_norm": 0.866973876953125,
+      "learning_rate": 0.00018716268274582114,
+      "loss": 2.7802,
+      "step": 3345
+    },
+    {
+      "epoch": 0.2700347026067307,
+      "grad_norm": 0.7709557414054871,
+      "learning_rate": 0.0001871549433952982,
+      "loss": 2.7418,
+      "step": 3346
+    },
+    {
+      "epoch": 0.2701154063433137,
+      "grad_norm": 0.7707573771476746,
+      "learning_rate": 0.00018714720187264626,
+      "loss": 2.7486,
+      "step": 3347
+    },
+    {
+      "epoch": 0.2701961100798967,
+      "grad_norm": 0.8007768392562866,
+      "learning_rate": 0.00018713945817805822,
+      "loss": 2.7106,
+      "step": 3348
+    },
+    {
+      "epoch": 0.2702768138164797,
+      "grad_norm": 0.7239583134651184,
+      "learning_rate": 0.0001871317123117271,
+      "loss": 2.7209,
+      "step": 3349
+    },
+    {
+      "epoch": 0.2703575175530627,
+      "grad_norm": 0.775104820728302,
+      "learning_rate": 0.00018712396427384594,
+      "loss": 2.6503,
+      "step": 3350
+    },
+    {
+      "epoch": 0.27043822128964573,
+      "grad_norm": 0.7492741346359253,
+      "learning_rate": 0.0001871162140646079,
+      "loss": 2.699,
+      "step": 3351
+    },
+    {
+      "epoch": 0.2705189250262287,
+      "grad_norm": 0.7550846338272095,
+      "learning_rate": 0.00018710846168420604,
+      "loss": 2.7458,
+      "step": 3352
+    },
+    {
+      "epoch": 0.27059962876281174,
+      "grad_norm": 0.807996928691864,
+      "learning_rate": 0.0001871007071328336,
+      "loss": 2.7604,
+      "step": 3353
+    },
+    {
+      "epoch": 0.2706803324993947,
+      "grad_norm": 0.7381845116615295,
+      "learning_rate": 0.00018709295041068386,
+      "loss": 2.6833,
+      "step": 3354
+    },
+    {
+      "epoch": 0.27076103623597775,
+      "grad_norm": 0.7542420625686646,
+      "learning_rate": 0.00018708519151795016,
+      "loss": 2.6462,
+      "step": 3355
+    },
+    {
+      "epoch": 0.2708417399725607,
+      "grad_norm": 0.7675846219062805,
+      "learning_rate": 0.00018707743045482582,
+      "loss": 2.7068,
+      "step": 3356
+    },
+    {
+      "epoch": 0.27092244370914376,
+      "grad_norm": 0.7437357902526855,
+      "learning_rate": 0.0001870696672215043,
+      "loss": 2.73,
+      "step": 3357
+    },
+    {
+      "epoch": 0.27100314744572673,
+      "grad_norm": 0.7880852222442627,
+      "learning_rate": 0.00018706190181817903,
+      "loss": 2.759,
+      "step": 3358
+    },
+    {
+      "epoch": 0.27108385118230977,
+      "grad_norm": 0.7403178811073303,
+      "learning_rate": 0.00018705413424504363,
+      "loss": 2.7538,
+      "step": 3359
+    },
+    {
+      "epoch": 0.27116455491889274,
+      "grad_norm": 0.7601225972175598,
+      "learning_rate": 0.00018704636450229164,
+      "loss": 2.7331,
+      "step": 3360
+    },
+    {
+      "epoch": 0.2712452586554758,
+      "grad_norm": 0.7810701727867126,
+      "learning_rate": 0.0001870385925901167,
+      "loss": 2.7736,
+      "step": 3361
+    },
+    {
+      "epoch": 0.27132596239205875,
+      "grad_norm": 0.8934530019760132,
+      "learning_rate": 0.0001870308185087125,
+      "loss": 2.7214,
+      "step": 3362
+    },
+    {
+      "epoch": 0.2714066661286418,
+      "grad_norm": 0.7468441128730774,
+      "learning_rate": 0.0001870230422582728,
+      "loss": 2.6957,
+      "step": 3363
+    },
+    {
+      "epoch": 0.27148736986522476,
+      "grad_norm": 0.7643293142318726,
+      "learning_rate": 0.00018701526383899144,
+      "loss": 2.6773,
+      "step": 3364
+    },
+    {
+      "epoch": 0.2715680736018078,
+      "grad_norm": 0.7602033615112305,
+      "learning_rate": 0.0001870074832510622,
+      "loss": 2.7095,
+      "step": 3365
+    },
+    {
+      "epoch": 0.27164877733839077,
+      "grad_norm": 0.772065281867981,
+      "learning_rate": 0.00018699970049467908,
+      "loss": 2.6753,
+      "step": 3366
+    },
+    {
+      "epoch": 0.27172948107497374,
+      "grad_norm": 0.7718359231948853,
+      "learning_rate": 0.00018699191557003598,
+      "loss": 2.6857,
+      "step": 3367
+    },
+    {
+      "epoch": 0.2718101848115568,
+      "grad_norm": 0.8207093477249146,
+      "learning_rate": 0.00018698412847732693,
+      "loss": 2.7549,
+      "step": 3368
+    },
+    {
+      "epoch": 0.27189088854813975,
+      "grad_norm": 0.7393590807914734,
+      "learning_rate": 0.00018697633921674605,
+      "loss": 2.6884,
+      "step": 3369
+    },
+    {
+      "epoch": 0.2719715922847228,
+      "grad_norm": 0.7955869436264038,
+      "learning_rate": 0.0001869685477884874,
+      "loss": 2.708,
+      "step": 3370
+    },
+    {
+      "epoch": 0.27205229602130576,
+      "grad_norm": 0.7392188906669617,
+      "learning_rate": 0.00018696075419274527,
+      "loss": 2.717,
+      "step": 3371
+    },
+    {
+      "epoch": 0.2721329997578888,
+      "grad_norm": 0.800204873085022,
+      "learning_rate": 0.00018695295842971376,
+      "loss": 2.7184,
+      "step": 3372
+    },
+    {
+      "epoch": 0.27221370349447177,
+      "grad_norm": 0.8195740580558777,
+      "learning_rate": 0.00018694516049958725,
+      "loss": 2.6865,
+      "step": 3373
+    },
+    {
+      "epoch": 0.2722944072310548,
+      "grad_norm": 0.8617578148841858,
+      "learning_rate": 0.00018693736040256007,
+      "loss": 2.7098,
+      "step": 3374
+    },
+    {
+      "epoch": 0.2723751109676378,
+      "grad_norm": 0.8184413909912109,
+      "learning_rate": 0.00018692955813882662,
+      "loss": 2.7449,
+      "step": 3375
+    },
+    {
+      "epoch": 0.2724558147042208,
+      "grad_norm": 0.990275502204895,
+      "learning_rate": 0.00018692175370858133,
+      "loss": 2.7891,
+      "step": 3376
+    },
+    {
+      "epoch": 0.2725365184408038,
+      "grad_norm": 0.7857810854911804,
+      "learning_rate": 0.0001869139471120187,
+      "loss": 2.6884,
+      "step": 3377
+    },
+    {
+      "epoch": 0.2726172221773868,
+      "grad_norm": 0.8040915131568909,
+      "learning_rate": 0.00018690613834933335,
+      "loss": 2.7047,
+      "step": 3378
+    },
+    {
+      "epoch": 0.2726979259139698,
+      "grad_norm": 0.7512348294258118,
+      "learning_rate": 0.00018689832742071983,
+      "loss": 2.6898,
+      "step": 3379
+    },
+    {
+      "epoch": 0.27277862965055283,
+      "grad_norm": 0.6781859397888184,
+      "learning_rate": 0.00018689051432637288,
+      "loss": 2.6396,
+      "step": 3380
+    },
+    {
+      "epoch": 0.2728593333871358,
+      "grad_norm": 0.7858247756958008,
+      "learning_rate": 0.00018688269906648716,
+      "loss": 2.6785,
+      "step": 3381
+    },
+    {
+      "epoch": 0.27294003712371884,
+      "grad_norm": 0.7342140674591064,
+      "learning_rate": 0.00018687488164125744,
+      "loss": 2.6778,
+      "step": 3382
+    },
+    {
+      "epoch": 0.2730207408603018,
+      "grad_norm": 0.8113372921943665,
+      "learning_rate": 0.00018686706205087858,
+      "loss": 2.6982,
+      "step": 3383
+    },
+    {
+      "epoch": 0.27310144459688485,
+      "grad_norm": 0.7904205918312073,
+      "learning_rate": 0.0001868592402955455,
+      "loss": 2.7891,
+      "step": 3384
+    },
+    {
+      "epoch": 0.2731821483334678,
+      "grad_norm": 0.7274135947227478,
+      "learning_rate": 0.00018685141637545308,
+      "loss": 2.6908,
+      "step": 3385
+    },
+    {
+      "epoch": 0.27326285207005085,
+      "grad_norm": 0.7675744295120239,
+      "learning_rate": 0.0001868435902907963,
+      "loss": 2.6987,
+      "step": 3386
+    },
+    {
+      "epoch": 0.27334355580663383,
+      "grad_norm": 0.8085030913352966,
+      "learning_rate": 0.00018683576204177026,
+      "loss": 2.7798,
+      "step": 3387
+    },
+    {
+      "epoch": 0.27342425954321686,
+      "grad_norm": 0.7498135566711426,
+      "learning_rate": 0.00018682793162857006,
+      "loss": 2.7216,
+      "step": 3388
+    },
+    {
+      "epoch": 0.27350496327979984,
+      "grad_norm": 0.900741696357727,
+      "learning_rate": 0.0001868200990513908,
+      "loss": 2.6871,
+      "step": 3389
+    },
+    {
+      "epoch": 0.27358566701638287,
+      "grad_norm": 0.7948571443557739,
+      "learning_rate": 0.00018681226431042772,
+      "loss": 2.6985,
+      "step": 3390
+    },
+    {
+      "epoch": 0.27366637075296585,
+      "grad_norm": 0.8739100098609924,
+      "learning_rate": 0.00018680442740587612,
+      "loss": 2.6922,
+      "step": 3391
+    },
+    {
+      "epoch": 0.2737470744895489,
+      "grad_norm": 0.730084240436554,
+      "learning_rate": 0.00018679658833793125,
+      "loss": 2.7029,
+      "step": 3392
+    },
+    {
+      "epoch": 0.27382777822613186,
+      "grad_norm": 0.7560603022575378,
+      "learning_rate": 0.00018678874710678853,
+      "loss": 2.7429,
+      "step": 3393
+    },
+    {
+      "epoch": 0.2739084819627149,
+      "grad_norm": 0.8331460356712341,
+      "learning_rate": 0.00018678090371264334,
+      "loss": 2.7157,
+      "step": 3394
+    },
+    {
+      "epoch": 0.27398918569929787,
+      "grad_norm": 0.8070168495178223,
+      "learning_rate": 0.00018677305815569122,
+      "loss": 2.7629,
+      "step": 3395
+    },
+    {
+      "epoch": 0.2740698894358809,
+      "grad_norm": 0.7922534346580505,
+      "learning_rate": 0.00018676521043612762,
+      "loss": 2.7159,
+      "step": 3396
+    },
+    {
+      "epoch": 0.2741505931724639,
+      "grad_norm": 0.7838901281356812,
+      "learning_rate": 0.0001867573605541482,
+      "loss": 2.6721,
+      "step": 3397
+    },
+    {
+      "epoch": 0.2742312969090469,
+      "grad_norm": 0.8912512063980103,
+      "learning_rate": 0.00018674950850994856,
+      "loss": 2.7243,
+      "step": 3398
+    },
+    {
+      "epoch": 0.2743120006456299,
+      "grad_norm": 0.7205448150634766,
+      "learning_rate": 0.0001867416543037244,
+      "loss": 2.7152,
+      "step": 3399
+    },
+    {
+      "epoch": 0.2743927043822129,
+      "grad_norm": 0.6992877721786499,
+      "learning_rate": 0.00018673379793567146,
+      "loss": 2.7183,
+      "step": 3400
+    },
+    {
+      "epoch": 0.2744734081187959,
+      "grad_norm": 0.8009448051452637,
+      "learning_rate": 0.00018672593940598556,
+      "loss": 2.715,
+      "step": 3401
+    },
+    {
+      "epoch": 0.2745541118553789,
+      "grad_norm": 0.7812647819519043,
+      "learning_rate": 0.0001867180787148626,
+      "loss": 2.7579,
+      "step": 3402
+    },
+    {
+      "epoch": 0.2746348155919619,
+      "grad_norm": 0.7300555109977722,
+      "learning_rate": 0.00018671021586249835,
+      "loss": 2.694,
+      "step": 3403
+    },
+    {
+      "epoch": 0.27471551932854493,
+      "grad_norm": 0.8082736134529114,
+      "learning_rate": 0.00018670235084908887,
+      "loss": 2.768,
+      "step": 3404
+    },
+    {
+      "epoch": 0.2747962230651279,
+      "grad_norm": 0.7729581594467163,
+      "learning_rate": 0.0001866944836748302,
+      "loss": 2.7256,
+      "step": 3405
+    },
+    {
+      "epoch": 0.27487692680171094,
+      "grad_norm": 0.8113458752632141,
+      "learning_rate": 0.00018668661433991835,
+      "loss": 2.6692,
+      "step": 3406
+    },
+    {
+      "epoch": 0.2749576305382939,
+      "grad_norm": 0.7757337689399719,
+      "learning_rate": 0.00018667874284454948,
+      "loss": 2.6769,
+      "step": 3407
+    },
+    {
+      "epoch": 0.27503833427487695,
+      "grad_norm": 0.7896093726158142,
+      "learning_rate": 0.00018667086918891976,
+      "loss": 2.7118,
+      "step": 3408
+    },
+    {
+      "epoch": 0.2751190380114599,
+      "grad_norm": 0.7764071822166443,
+      "learning_rate": 0.00018666299337322543,
+      "loss": 2.7284,
+      "step": 3409
+    },
+    {
+      "epoch": 0.27519974174804296,
+      "grad_norm": 0.794815182685852,
+      "learning_rate": 0.00018665511539766273,
+      "loss": 2.7232,
+      "step": 3410
+    },
+    {
+      "epoch": 0.27528044548462594,
+      "grad_norm": 0.8134122490882874,
+      "learning_rate": 0.0001866472352624281,
+      "loss": 2.7023,
+      "step": 3411
+    },
+    {
+      "epoch": 0.27536114922120897,
+      "grad_norm": 0.7654025554656982,
+      "learning_rate": 0.00018663935296771782,
+      "loss": 2.7002,
+      "step": 3412
+    },
+    {
+      "epoch": 0.27544185295779194,
+      "grad_norm": 0.6930806636810303,
+      "learning_rate": 0.0001866314685137284,
+      "loss": 2.6764,
+      "step": 3413
+    },
+    {
+      "epoch": 0.275522556694375,
+      "grad_norm": 0.7535184621810913,
+      "learning_rate": 0.00018662358190065631,
+      "loss": 2.6657,
+      "step": 3414
+    },
+    {
+      "epoch": 0.27560326043095795,
+      "grad_norm": 0.7775620818138123,
+      "learning_rate": 0.00018661569312869816,
+      "loss": 2.6931,
+      "step": 3415
+    },
+    {
+      "epoch": 0.275683964167541,
+      "grad_norm": 0.7209072113037109,
+      "learning_rate": 0.00018660780219805048,
+      "loss": 2.7293,
+      "step": 3416
+    },
+    {
+      "epoch": 0.27576466790412396,
+      "grad_norm": 0.7182055711746216,
+      "learning_rate": 0.00018659990910891,
+      "loss": 2.6561,
+      "step": 3417
+    },
+    {
+      "epoch": 0.27584537164070694,
+      "grad_norm": 0.7130969166755676,
+      "learning_rate": 0.00018659201386147338,
+      "loss": 2.7156,
+      "step": 3418
+    },
+    {
+      "epoch": 0.27592607537728997,
+      "grad_norm": 0.7296265959739685,
+      "learning_rate": 0.00018658411645593745,
+      "loss": 2.6894,
+      "step": 3419
+    },
+    {
+      "epoch": 0.27600677911387295,
+      "grad_norm": 0.7707972526550293,
+      "learning_rate": 0.000186576216892499,
+      "loss": 2.7528,
+      "step": 3420
+    },
+    {
+      "epoch": 0.276087482850456,
+      "grad_norm": 0.6945170164108276,
+      "learning_rate": 0.0001865683151713549,
+      "loss": 2.6762,
+      "step": 3421
+    },
+    {
+      "epoch": 0.27616818658703896,
+      "grad_norm": 0.7664114236831665,
+      "learning_rate": 0.0001865604112927021,
+      "loss": 2.7212,
+      "step": 3422
+    },
+    {
+      "epoch": 0.276248890323622,
+      "grad_norm": 0.6950399875640869,
+      "learning_rate": 0.0001865525052567376,
+      "loss": 2.7035,
+      "step": 3423
+    },
+    {
+      "epoch": 0.27632959406020496,
+      "grad_norm": 0.7307506799697876,
+      "learning_rate": 0.00018654459706365838,
+      "loss": 2.7296,
+      "step": 3424
+    },
+    {
+      "epoch": 0.276410297796788,
+      "grad_norm": 0.720912516117096,
+      "learning_rate": 0.0001865366867136616,
+      "loss": 2.6884,
+      "step": 3425
+    },
+    {
+      "epoch": 0.276491001533371,
+      "grad_norm": 0.7581072449684143,
+      "learning_rate": 0.00018652877420694436,
+      "loss": 2.705,
+      "step": 3426
+    },
+    {
+      "epoch": 0.276571705269954,
+      "grad_norm": 0.7473136186599731,
+      "learning_rate": 0.0001865208595437039,
+      "loss": 2.7316,
+      "step": 3427
+    },
+    {
+      "epoch": 0.276652409006537,
+      "grad_norm": 0.7272855639457703,
+      "learning_rate": 0.00018651294272413745,
+      "loss": 2.6834,
+      "step": 3428
+    },
+    {
+      "epoch": 0.27673311274312,
+      "grad_norm": 0.7046366930007935,
+      "learning_rate": 0.0001865050237484423,
+      "loss": 2.6491,
+      "step": 3429
+    },
+    {
+      "epoch": 0.276813816479703,
+      "grad_norm": 0.7521376609802246,
+      "learning_rate": 0.00018649710261681586,
+      "loss": 2.708,
+      "step": 3430
+    },
+    {
+      "epoch": 0.276894520216286,
+      "grad_norm": 0.7372453808784485,
+      "learning_rate": 0.0001864891793294555,
+      "loss": 2.682,
+      "step": 3431
+    },
+    {
+      "epoch": 0.276975223952869,
+      "grad_norm": 0.7381749749183655,
+      "learning_rate": 0.0001864812538865587,
+      "loss": 2.7526,
+      "step": 3432
+    },
+    {
+      "epoch": 0.27705592768945203,
+      "grad_norm": 0.7891514301300049,
+      "learning_rate": 0.00018647332628832298,
+      "loss": 2.6904,
+      "step": 3433
+    },
+    {
+      "epoch": 0.277136631426035,
+      "grad_norm": 0.7942724823951721,
+      "learning_rate": 0.00018646539653494596,
+      "loss": 2.7873,
+      "step": 3434
+    },
+    {
+      "epoch": 0.27721733516261804,
+      "grad_norm": 0.7365398406982422,
+      "learning_rate": 0.0001864574646266252,
+      "loss": 2.6684,
+      "step": 3435
+    },
+    {
+      "epoch": 0.277298038899201,
+      "grad_norm": 0.7802249193191528,
+      "learning_rate": 0.00018644953056355846,
+      "loss": 2.7152,
+      "step": 3436
+    },
+    {
+      "epoch": 0.27737874263578405,
+      "grad_norm": 0.7801448106765747,
+      "learning_rate": 0.0001864415943459434,
+      "loss": 2.7034,
+      "step": 3437
+    },
+    {
+      "epoch": 0.277459446372367,
+      "grad_norm": 0.7722738981246948,
+      "learning_rate": 0.00018643365597397786,
+      "loss": 2.7135,
+      "step": 3438
+    },
+    {
+      "epoch": 0.27754015010895006,
+      "grad_norm": 0.7847445011138916,
+      "learning_rate": 0.00018642571544785967,
+      "loss": 2.6999,
+      "step": 3439
+    },
+    {
+      "epoch": 0.27762085384553303,
+      "grad_norm": 0.7226125597953796,
+      "learning_rate": 0.00018641777276778675,
+      "loss": 2.7613,
+      "step": 3440
+    },
+    {
+      "epoch": 0.27770155758211607,
+      "grad_norm": 0.713188111782074,
+      "learning_rate": 0.000186409827933957,
+      "loss": 2.6953,
+      "step": 3441
+    },
+    {
+      "epoch": 0.27778226131869904,
+      "grad_norm": 0.7308298349380493,
+      "learning_rate": 0.0001864018809465685,
+      "loss": 2.7045,
+      "step": 3442
+    },
+    {
+      "epoch": 0.2778629650552821,
+      "grad_norm": 0.7606719732284546,
+      "learning_rate": 0.00018639393180581925,
+      "loss": 2.7883,
+      "step": 3443
+    },
+    {
+      "epoch": 0.27794366879186505,
+      "grad_norm": 0.7583296895027161,
+      "learning_rate": 0.00018638598051190738,
+      "loss": 2.6734,
+      "step": 3444
+    },
+    {
+      "epoch": 0.2780243725284481,
+      "grad_norm": 0.7147012948989868,
+      "learning_rate": 0.00018637802706503108,
+      "loss": 2.7223,
+      "step": 3445
+    },
+    {
+      "epoch": 0.27810507626503106,
+      "grad_norm": 0.7812997102737427,
+      "learning_rate": 0.00018637007146538853,
+      "loss": 2.7277,
+      "step": 3446
+    },
+    {
+      "epoch": 0.2781857800016141,
+      "grad_norm": 0.7460772395133972,
+      "learning_rate": 0.000186362113713178,
+      "loss": 2.6875,
+      "step": 3447
+    },
+    {
+      "epoch": 0.27826648373819707,
+      "grad_norm": 0.7359143495559692,
+      "learning_rate": 0.0001863541538085979,
+      "loss": 2.7122,
+      "step": 3448
+    },
+    {
+      "epoch": 0.2783471874747801,
+      "grad_norm": 0.7122978568077087,
+      "learning_rate": 0.00018634619175184655,
+      "loss": 2.6381,
+      "step": 3449
+    },
+    {
+      "epoch": 0.2784278912113631,
+      "grad_norm": 0.6965885758399963,
+      "learning_rate": 0.00018633822754312234,
+      "loss": 2.6957,
+      "step": 3450
+    },
+    {
+      "epoch": 0.2785085949479461,
+      "grad_norm": 0.7737082242965698,
+      "learning_rate": 0.00018633026118262385,
+      "loss": 2.7579,
+      "step": 3451
+    },
+    {
+      "epoch": 0.2785892986845291,
+      "grad_norm": 0.6925420165061951,
+      "learning_rate": 0.00018632229267054958,
+      "loss": 2.6226,
+      "step": 3452
+    },
+    {
+      "epoch": 0.2786700024211121,
+      "grad_norm": 0.7496356964111328,
+      "learning_rate": 0.0001863143220070981,
+      "loss": 2.7059,
+      "step": 3453
+    },
+    {
+      "epoch": 0.2787507061576951,
+      "grad_norm": 0.7066817283630371,
+      "learning_rate": 0.0001863063491924681,
+      "loss": 2.681,
+      "step": 3454
+    },
+    {
+      "epoch": 0.2788314098942781,
+      "grad_norm": 0.8143237829208374,
+      "learning_rate": 0.0001862983742268583,
+      "loss": 2.6698,
+      "step": 3455
+    },
+    {
+      "epoch": 0.2789121136308611,
+      "grad_norm": 0.7518483996391296,
+      "learning_rate": 0.00018629039711046737,
+      "loss": 2.7041,
+      "step": 3456
+    },
+    {
+      "epoch": 0.27899281736744413,
+      "grad_norm": 0.8756366968154907,
+      "learning_rate": 0.00018628241784349422,
+      "loss": 2.7547,
+      "step": 3457
+    },
+    {
+      "epoch": 0.2790735211040271,
+      "grad_norm": 0.8709446787834167,
+      "learning_rate": 0.0001862744364261377,
+      "loss": 2.7068,
+      "step": 3458
+    },
+    {
+      "epoch": 0.27915422484061014,
+      "grad_norm": 0.8121913075447083,
+      "learning_rate": 0.00018626645285859666,
+      "loss": 2.673,
+      "step": 3459
+    },
+    {
+      "epoch": 0.2792349285771931,
+      "grad_norm": 0.7685909271240234,
+      "learning_rate": 0.00018625846714107012,
+      "loss": 2.7389,
+      "step": 3460
+    },
+    {
+      "epoch": 0.27931563231377615,
+      "grad_norm": 0.7098073363304138,
+      "learning_rate": 0.0001862504792737571,
+      "loss": 2.6942,
+      "step": 3461
+    },
+    {
+      "epoch": 0.27939633605035913,
+      "grad_norm": 0.7718049883842468,
+      "learning_rate": 0.00018624248925685666,
+      "loss": 2.7359,
+      "step": 3462
+    },
+    {
+      "epoch": 0.27947703978694216,
+      "grad_norm": 0.7912909984588623,
+      "learning_rate": 0.00018623449709056797,
+      "loss": 2.6658,
+      "step": 3463
+    },
+    {
+      "epoch": 0.27955774352352514,
+      "grad_norm": 0.7255454659461975,
+      "learning_rate": 0.0001862265027750902,
+      "loss": 2.771,
+      "step": 3464
+    },
+    {
+      "epoch": 0.27963844726010817,
+      "grad_norm": 0.7542218565940857,
+      "learning_rate": 0.00018621850631062254,
+      "loss": 2.6741,
+      "step": 3465
+    },
+    {
+      "epoch": 0.27971915099669115,
+      "grad_norm": 0.8386052846908569,
+      "learning_rate": 0.00018621050769736437,
+      "loss": 2.67,
+      "step": 3466
+    },
+    {
+      "epoch": 0.2797998547332742,
+      "grad_norm": 0.8563781976699829,
+      "learning_rate": 0.00018620250693551495,
+      "loss": 2.7461,
+      "step": 3467
+    },
+    {
+      "epoch": 0.27988055846985715,
+      "grad_norm": 0.7490699291229248,
+      "learning_rate": 0.00018619450402527376,
+      "loss": 2.6863,
+      "step": 3468
+    },
+    {
+      "epoch": 0.27996126220644013,
+      "grad_norm": 0.8008999824523926,
+      "learning_rate": 0.00018618649896684017,
+      "loss": 2.7769,
+      "step": 3469
+    },
+    {
+      "epoch": 0.28004196594302316,
+      "grad_norm": 0.7678235769271851,
+      "learning_rate": 0.00018617849176041378,
+      "loss": 2.7237,
+      "step": 3470
+    },
+    {
+      "epoch": 0.28012266967960614,
+      "grad_norm": 0.8774877786636353,
+      "learning_rate": 0.00018617048240619408,
+      "loss": 2.7502,
+      "step": 3471
+    },
+    {
+      "epoch": 0.28020337341618917,
+      "grad_norm": 0.8150283098220825,
+      "learning_rate": 0.00018616247090438073,
+      "loss": 2.6941,
+      "step": 3472
+    },
+    {
+      "epoch": 0.28028407715277215,
+      "grad_norm": 0.7330089807510376,
+      "learning_rate": 0.00018615445725517332,
+      "loss": 2.7002,
+      "step": 3473
+    },
+    {
+      "epoch": 0.2803647808893552,
+      "grad_norm": 0.748275101184845,
+      "learning_rate": 0.00018614644145877168,
+      "loss": 2.6996,
+      "step": 3474
+    },
+    {
+      "epoch": 0.28044548462593816,
+      "grad_norm": 0.7718296647071838,
+      "learning_rate": 0.0001861384235153755,
+      "loss": 2.7333,
+      "step": 3475
+    },
+    {
+      "epoch": 0.2805261883625212,
+      "grad_norm": 0.7751123309135437,
+      "learning_rate": 0.00018613040342518465,
+      "loss": 2.7362,
+      "step": 3476
+    },
+    {
+      "epoch": 0.28060689209910417,
+      "grad_norm": 0.70979243516922,
+      "learning_rate": 0.000186122381188399,
+      "loss": 2.6651,
+      "step": 3477
+    },
+    {
+      "epoch": 0.2806875958356872,
+      "grad_norm": 0.9607138633728027,
+      "learning_rate": 0.00018611435680521848,
+      "loss": 2.7779,
+      "step": 3478
+    },
+    {
+      "epoch": 0.2807682995722702,
+      "grad_norm": 0.709671676158905,
+      "learning_rate": 0.0001861063302758431,
+      "loss": 2.6994,
+      "step": 3479
+    },
+    {
+      "epoch": 0.2808490033088532,
+      "grad_norm": 0.8765757083892822,
+      "learning_rate": 0.00018609830160047283,
+      "loss": 2.7107,
+      "step": 3480
+    },
+    {
+      "epoch": 0.2809297070454362,
+      "grad_norm": 0.7996764183044434,
+      "learning_rate": 0.0001860902707793079,
+      "loss": 2.7921,
+      "step": 3481
+    },
+    {
+      "epoch": 0.2810104107820192,
+      "grad_norm": 0.7094513177871704,
+      "learning_rate": 0.0001860822378125483,
+      "loss": 2.7211,
+      "step": 3482
+    },
+    {
+      "epoch": 0.2810911145186022,
+      "grad_norm": 0.8068607449531555,
+      "learning_rate": 0.0001860742027003944,
+      "loss": 2.675,
+      "step": 3483
+    },
+    {
+      "epoch": 0.2811718182551852,
+      "grad_norm": 0.7737938165664673,
+      "learning_rate": 0.00018606616544304628,
+      "loss": 2.7538,
+      "step": 3484
+    },
+    {
+      "epoch": 0.2812525219917682,
+      "grad_norm": 0.7979975342750549,
+      "learning_rate": 0.0001860581260407044,
+      "loss": 2.7894,
+      "step": 3485
+    },
+    {
+      "epoch": 0.28133322572835123,
+      "grad_norm": 0.7671655416488647,
+      "learning_rate": 0.00018605008449356904,
+      "loss": 2.7097,
+      "step": 3486
+    },
+    {
+      "epoch": 0.2814139294649342,
+      "grad_norm": 0.7284159064292908,
+      "learning_rate": 0.00018604204080184062,
+      "loss": 2.7447,
+      "step": 3487
+    },
+    {
+      "epoch": 0.28149463320151724,
+      "grad_norm": 0.7425351142883301,
+      "learning_rate": 0.00018603399496571968,
+      "loss": 2.7302,
+      "step": 3488
+    },
+    {
+      "epoch": 0.2815753369381002,
+      "grad_norm": 0.7709810733795166,
+      "learning_rate": 0.00018602594698540663,
+      "loss": 2.6979,
+      "step": 3489
+    },
+    {
+      "epoch": 0.28165604067468325,
+      "grad_norm": 0.744628369808197,
+      "learning_rate": 0.00018601789686110214,
+      "loss": 2.7279,
+      "step": 3490
+    },
+    {
+      "epoch": 0.2817367444112662,
+      "grad_norm": 0.7679976224899292,
+      "learning_rate": 0.00018600984459300678,
+      "loss": 2.6862,
+      "step": 3491
+    },
+    {
+      "epoch": 0.28181744814784926,
+      "grad_norm": 0.7923497557640076,
+      "learning_rate": 0.0001860017901813213,
+      "loss": 2.6975,
+      "step": 3492
+    },
+    {
+      "epoch": 0.28189815188443224,
+      "grad_norm": 0.7896692156791687,
+      "learning_rate": 0.00018599373362624636,
+      "loss": 2.7052,
+      "step": 3493
+    },
+    {
+      "epoch": 0.28197885562101527,
+      "grad_norm": 0.7913276553153992,
+      "learning_rate": 0.00018598567492798284,
+      "loss": 2.7233,
+      "step": 3494
+    },
+    {
+      "epoch": 0.28205955935759824,
+      "grad_norm": 0.7385257482528687,
+      "learning_rate": 0.00018597761408673146,
+      "loss": 2.7616,
+      "step": 3495
+    },
+    {
+      "epoch": 0.2821402630941813,
+      "grad_norm": 0.7181909084320068,
+      "learning_rate": 0.00018596955110269323,
+      "loss": 2.718,
+      "step": 3496
+    },
+    {
+      "epoch": 0.28222096683076425,
+      "grad_norm": 0.8313151597976685,
+      "learning_rate": 0.00018596148597606907,
+      "loss": 2.6775,
+      "step": 3497
+    },
+    {
+      "epoch": 0.2823016705673473,
+      "grad_norm": 0.7235481142997742,
+      "learning_rate": 0.00018595341870705995,
+      "loss": 2.7085,
+      "step": 3498
+    },
+    {
+      "epoch": 0.28238237430393026,
+      "grad_norm": 0.7092145085334778,
+      "learning_rate": 0.00018594534929586697,
+      "loss": 2.7167,
+      "step": 3499
+    },
+    {
+      "epoch": 0.2824630780405133,
+      "grad_norm": 0.7929207682609558,
+      "learning_rate": 0.0001859372777426912,
+      "loss": 2.663,
+      "step": 3500
+    },
+    {
+      "epoch": 0.28254378177709627,
+      "grad_norm": 0.7488871216773987,
+      "learning_rate": 0.00018592920404773383,
+      "loss": 2.7911,
+      "step": 3501
+    },
+    {
+      "epoch": 0.2826244855136793,
+      "grad_norm": 0.8230419158935547,
+      "learning_rate": 0.0001859211282111961,
+      "loss": 2.754,
+      "step": 3502
+    },
+    {
+      "epoch": 0.2827051892502623,
+      "grad_norm": 0.731971025466919,
+      "learning_rate": 0.00018591305023327924,
+      "loss": 2.7142,
+      "step": 3503
+    },
+    {
+      "epoch": 0.2827858929868453,
+      "grad_norm": 0.8159881234169006,
+      "learning_rate": 0.00018590497011418457,
+      "loss": 2.7046,
+      "step": 3504
+    },
+    {
+      "epoch": 0.2828665967234283,
+      "grad_norm": 0.750266432762146,
+      "learning_rate": 0.0001858968878541135,
+      "loss": 2.6951,
+      "step": 3505
+    },
+    {
+      "epoch": 0.2829473004600113,
+      "grad_norm": 0.7750049233436584,
+      "learning_rate": 0.00018588880345326748,
+      "loss": 2.6958,
+      "step": 3506
+    },
+    {
+      "epoch": 0.2830280041965943,
+      "grad_norm": 0.8559218049049377,
+      "learning_rate": 0.00018588071691184795,
+      "loss": 2.7205,
+      "step": 3507
+    },
+    {
+      "epoch": 0.28310870793317733,
+      "grad_norm": 0.7334830164909363,
+      "learning_rate": 0.00018587262823005642,
+      "loss": 2.7134,
+      "step": 3508
+    },
+    {
+      "epoch": 0.2831894116697603,
+      "grad_norm": 0.8749497532844543,
+      "learning_rate": 0.00018586453740809456,
+      "loss": 2.6811,
+      "step": 3509
+    },
+    {
+      "epoch": 0.28327011540634334,
+      "grad_norm": 0.8800753355026245,
+      "learning_rate": 0.00018585644444616396,
+      "loss": 2.7427,
+      "step": 3510
+    },
+    {
+      "epoch": 0.2833508191429263,
+      "grad_norm": 0.8666185736656189,
+      "learning_rate": 0.00018584834934446632,
+      "loss": 2.6828,
+      "step": 3511
+    },
+    {
+      "epoch": 0.28343152287950935,
+      "grad_norm": 0.7451635003089905,
+      "learning_rate": 0.00018584025210320343,
+      "loss": 2.6784,
+      "step": 3512
+    },
+    {
+      "epoch": 0.2835122266160923,
+      "grad_norm": 0.8512656688690186,
+      "learning_rate": 0.00018583215272257708,
+      "loss": 2.7762,
+      "step": 3513
+    },
+    {
+      "epoch": 0.28359293035267535,
+      "grad_norm": 0.9298297166824341,
+      "learning_rate": 0.00018582405120278907,
+      "loss": 2.7714,
+      "step": 3514
+    },
+    {
+      "epoch": 0.28367363408925833,
+      "grad_norm": 0.7968065738677979,
+      "learning_rate": 0.0001858159475440414,
+      "loss": 2.7286,
+      "step": 3515
+    },
+    {
+      "epoch": 0.28375433782584136,
+      "grad_norm": 0.7381564378738403,
+      "learning_rate": 0.00018580784174653596,
+      "loss": 2.6697,
+      "step": 3516
+    },
+    {
+      "epoch": 0.28383504156242434,
+      "grad_norm": 0.8199222683906555,
+      "learning_rate": 0.00018579973381047481,
+      "loss": 2.7463,
+      "step": 3517
+    },
+    {
+      "epoch": 0.28391574529900737,
+      "grad_norm": 0.8022071123123169,
+      "learning_rate": 0.00018579162373606002,
+      "loss": 2.6898,
+      "step": 3518
+    },
+    {
+      "epoch": 0.28399644903559035,
+      "grad_norm": 0.7899700999259949,
+      "learning_rate": 0.0001857835115234937,
+      "loss": 2.7074,
+      "step": 3519
+    },
+    {
+      "epoch": 0.2840771527721733,
+      "grad_norm": 0.7237183451652527,
+      "learning_rate": 0.00018577539717297805,
+      "loss": 2.6699,
+      "step": 3520
+    },
+    {
+      "epoch": 0.28415785650875636,
+      "grad_norm": 0.7627314329147339,
+      "learning_rate": 0.00018576728068471526,
+      "loss": 2.7745,
+      "step": 3521
+    },
+    {
+      "epoch": 0.28423856024533933,
+      "grad_norm": 0.7301654815673828,
+      "learning_rate": 0.00018575916205890766,
+      "loss": 2.7191,
+      "step": 3522
+    },
+    {
+      "epoch": 0.28431926398192237,
+      "grad_norm": 0.7441647052764893,
+      "learning_rate": 0.00018575104129575753,
+      "loss": 2.7529,
+      "step": 3523
+    },
+    {
+      "epoch": 0.28439996771850534,
+      "grad_norm": 0.7715914249420166,
+      "learning_rate": 0.0001857429183954673,
+      "loss": 2.6893,
+      "step": 3524
+    },
+    {
+      "epoch": 0.2844806714550884,
+      "grad_norm": 0.7464057207107544,
+      "learning_rate": 0.00018573479335823944,
+      "loss": 2.7169,
+      "step": 3525
+    },
+    {
+      "epoch": 0.28456137519167135,
+      "grad_norm": 0.753198504447937,
+      "learning_rate": 0.00018572666618427638,
+      "loss": 2.7144,
+      "step": 3526
+    },
+    {
+      "epoch": 0.2846420789282544,
+      "grad_norm": 0.7681953310966492,
+      "learning_rate": 0.00018571853687378073,
+      "loss": 2.709,
+      "step": 3527
+    },
+    {
+      "epoch": 0.28472278266483736,
+      "grad_norm": 0.7591876983642578,
+      "learning_rate": 0.0001857104054269551,
+      "loss": 2.7519,
+      "step": 3528
+    },
+    {
+      "epoch": 0.2848034864014204,
+      "grad_norm": 0.7417709827423096,
+      "learning_rate": 0.00018570227184400205,
+      "loss": 2.6756,
+      "step": 3529
+    },
+    {
+      "epoch": 0.28488419013800337,
+      "grad_norm": 0.7641329169273376,
+      "learning_rate": 0.0001856941361251244,
+      "loss": 2.6614,
+      "step": 3530
+    },
+    {
+      "epoch": 0.2849648938745864,
+      "grad_norm": 0.7813490033149719,
+      "learning_rate": 0.0001856859982705249,
+      "loss": 2.7145,
+      "step": 3531
+    },
+    {
+      "epoch": 0.2850455976111694,
+      "grad_norm": 0.7777202129364014,
+      "learning_rate": 0.00018567785828040628,
+      "loss": 2.7015,
+      "step": 3532
+    },
+    {
+      "epoch": 0.2851263013477524,
+      "grad_norm": 0.7647144794464111,
+      "learning_rate": 0.0001856697161549715,
+      "loss": 2.7311,
+      "step": 3533
+    },
+    {
+      "epoch": 0.2852070050843354,
+      "grad_norm": 0.7477256655693054,
+      "learning_rate": 0.00018566157189442342,
+      "loss": 2.6832,
+      "step": 3534
+    },
+    {
+      "epoch": 0.2852877088209184,
+      "grad_norm": 0.7037049531936646,
+      "learning_rate": 0.00018565342549896506,
+      "loss": 2.6942,
+      "step": 3535
+    },
+    {
+      "epoch": 0.2853684125575014,
+      "grad_norm": 0.7309197783470154,
+      "learning_rate": 0.00018564527696879945,
+      "loss": 2.6797,
+      "step": 3536
+    },
+    {
+      "epoch": 0.2854491162940844,
+      "grad_norm": 0.798075795173645,
+      "learning_rate": 0.00018563712630412967,
+      "loss": 2.6926,
+      "step": 3537
+    },
+    {
+      "epoch": 0.2855298200306674,
+      "grad_norm": 0.7831682562828064,
+      "learning_rate": 0.0001856289735051588,
+      "loss": 2.7537,
+      "step": 3538
+    },
+    {
+      "epoch": 0.28561052376725043,
+      "grad_norm": 0.7983096241950989,
+      "learning_rate": 0.0001856208185720901,
+      "loss": 2.7037,
+      "step": 3539
+    },
+    {
+      "epoch": 0.2856912275038334,
+      "grad_norm": 0.7250573635101318,
+      "learning_rate": 0.00018561266150512678,
+      "loss": 2.7282,
+      "step": 3540
+    },
+    {
+      "epoch": 0.28577193124041644,
+      "grad_norm": 0.7800211906433105,
+      "learning_rate": 0.00018560450230447218,
+      "loss": 2.6541,
+      "step": 3541
+    },
+    {
+      "epoch": 0.2858526349769994,
+      "grad_norm": 0.7624209523200989,
+      "learning_rate": 0.00018559634097032953,
+      "loss": 2.7041,
+      "step": 3542
+    },
+    {
+      "epoch": 0.28593333871358245,
+      "grad_norm": 0.7212036848068237,
+      "learning_rate": 0.0001855881775029024,
+      "loss": 2.7287,
+      "step": 3543
+    },
+    {
+      "epoch": 0.28601404245016543,
+      "grad_norm": 0.7774164080619812,
+      "learning_rate": 0.00018558001190239408,
+      "loss": 2.6515,
+      "step": 3544
+    },
+    {
+      "epoch": 0.28609474618674846,
+      "grad_norm": 0.7169588208198547,
+      "learning_rate": 0.0001855718441690082,
+      "loss": 2.7111,
+      "step": 3545
+    },
+    {
+      "epoch": 0.28617544992333144,
+      "grad_norm": 0.7473909258842468,
+      "learning_rate": 0.00018556367430294827,
+      "loss": 2.7405,
+      "step": 3546
+    },
+    {
+      "epoch": 0.28625615365991447,
+      "grad_norm": 0.7213929295539856,
+      "learning_rate": 0.0001855555023044179,
+      "loss": 2.7336,
+      "step": 3547
+    },
+    {
+      "epoch": 0.28633685739649745,
+      "grad_norm": 0.701816201210022,
+      "learning_rate": 0.00018554732817362078,
+      "loss": 2.721,
+      "step": 3548
+    },
+    {
+      "epoch": 0.2864175611330805,
+      "grad_norm": 0.8158134818077087,
+      "learning_rate": 0.00018553915191076064,
+      "loss": 2.6979,
+      "step": 3549
+    },
+    {
+      "epoch": 0.28649826486966345,
+      "grad_norm": 0.7303084135055542,
+      "learning_rate": 0.00018553097351604118,
+      "loss": 2.6734,
+      "step": 3550
+    },
+    {
+      "epoch": 0.2865789686062465,
+      "grad_norm": 0.8140435814857483,
+      "learning_rate": 0.00018552279298966634,
+      "loss": 2.6832,
+      "step": 3551
+    },
+    {
+      "epoch": 0.28665967234282946,
+      "grad_norm": 0.7024678587913513,
+      "learning_rate": 0.00018551461033183988,
+      "loss": 2.7118,
+      "step": 3552
+    },
+    {
+      "epoch": 0.2867403760794125,
+      "grad_norm": 0.7277806401252747,
+      "learning_rate": 0.00018550642554276582,
+      "loss": 2.6362,
+      "step": 3553
+    },
+    {
+      "epoch": 0.28682107981599547,
+      "grad_norm": 0.8376575112342834,
+      "learning_rate": 0.00018549823862264812,
+      "loss": 2.744,
+      "step": 3554
+    },
+    {
+      "epoch": 0.2869017835525785,
+      "grad_norm": 0.712195098400116,
+      "learning_rate": 0.00018549004957169082,
+      "loss": 2.6715,
+      "step": 3555
+    },
+    {
+      "epoch": 0.2869824872891615,
+      "grad_norm": 0.7511523962020874,
+      "learning_rate": 0.00018548185839009805,
+      "loss": 2.7655,
+      "step": 3556
+    },
+    {
+      "epoch": 0.2870631910257445,
+      "grad_norm": 0.7397211790084839,
+      "learning_rate": 0.00018547366507807388,
+      "loss": 2.6813,
+      "step": 3557
+    },
+    {
+      "epoch": 0.2871438947623275,
+      "grad_norm": 0.6926341652870178,
+      "learning_rate": 0.00018546546963582253,
+      "loss": 2.6477,
+      "step": 3558
+    },
+    {
+      "epoch": 0.2872245984989105,
+      "grad_norm": 0.7776244878768921,
+      "learning_rate": 0.00018545727206354827,
+      "loss": 2.6979,
+      "step": 3559
+    },
+    {
+      "epoch": 0.2873053022354935,
+      "grad_norm": 0.7639400959014893,
+      "learning_rate": 0.00018544907236145542,
+      "loss": 2.6913,
+      "step": 3560
+    },
+    {
+      "epoch": 0.28738600597207653,
+      "grad_norm": 0.7738329768180847,
+      "learning_rate": 0.0001854408705297483,
+      "loss": 2.7231,
+      "step": 3561
+    },
+    {
+      "epoch": 0.2874667097086595,
+      "grad_norm": 0.7182422876358032,
+      "learning_rate": 0.00018543266656863137,
+      "loss": 2.718,
+      "step": 3562
+    },
+    {
+      "epoch": 0.28754741344524254,
+      "grad_norm": 0.7257261276245117,
+      "learning_rate": 0.00018542446047830903,
+      "loss": 2.7354,
+      "step": 3563
+    },
+    {
+      "epoch": 0.2876281171818255,
+      "grad_norm": 0.7761391997337341,
+      "learning_rate": 0.00018541625225898588,
+      "loss": 2.705,
+      "step": 3564
+    },
+    {
+      "epoch": 0.28770882091840855,
+      "grad_norm": 0.9272314310073853,
+      "learning_rate": 0.0001854080419108664,
+      "loss": 2.7278,
+      "step": 3565
+    },
+    {
+      "epoch": 0.2877895246549915,
+      "grad_norm": 0.7622589468955994,
+      "learning_rate": 0.00018539982943415527,
+      "loss": 2.7224,
+      "step": 3566
+    },
+    {
+      "epoch": 0.28787022839157456,
+      "grad_norm": 0.725349485874176,
+      "learning_rate": 0.0001853916148290572,
+      "loss": 2.6782,
+      "step": 3567
+    },
+    {
+      "epoch": 0.28795093212815753,
+      "grad_norm": 0.776242733001709,
+      "learning_rate": 0.0001853833980957768,
+      "loss": 2.6467,
+      "step": 3568
+    },
+    {
+      "epoch": 0.28803163586474057,
+      "grad_norm": 0.8461112976074219,
+      "learning_rate": 0.00018537517923451896,
+      "loss": 2.6763,
+      "step": 3569
+    },
+    {
+      "epoch": 0.28811233960132354,
+      "grad_norm": 0.8161221742630005,
+      "learning_rate": 0.00018536695824548848,
+      "loss": 2.7057,
+      "step": 3570
+    },
+    {
+      "epoch": 0.2881930433379065,
+      "grad_norm": 0.7404211759567261,
+      "learning_rate": 0.00018535873512889024,
+      "loss": 2.7083,
+      "step": 3571
+    },
+    {
+      "epoch": 0.28827374707448955,
+      "grad_norm": 0.831042468547821,
+      "learning_rate": 0.00018535050988492918,
+      "loss": 2.6121,
+      "step": 3572
+    },
+    {
+      "epoch": 0.2883544508110725,
+      "grad_norm": 0.7286352515220642,
+      "learning_rate": 0.00018534228251381035,
+      "loss": 2.7165,
+      "step": 3573
+    },
+    {
+      "epoch": 0.28843515454765556,
+      "grad_norm": 0.7951883673667908,
+      "learning_rate": 0.00018533405301573872,
+      "loss": 2.6794,
+      "step": 3574
+    },
+    {
+      "epoch": 0.28851585828423854,
+      "grad_norm": 0.7431079149246216,
+      "learning_rate": 0.00018532582139091944,
+      "loss": 2.6758,
+      "step": 3575
+    },
+    {
+      "epoch": 0.28859656202082157,
+      "grad_norm": 0.7408809065818787,
+      "learning_rate": 0.0001853175876395576,
+      "loss": 2.6901,
+      "step": 3576
+    },
+    {
+      "epoch": 0.28867726575740454,
+      "grad_norm": 0.7428708672523499,
+      "learning_rate": 0.00018530935176185848,
+      "loss": 2.6679,
+      "step": 3577
+    },
+    {
+      "epoch": 0.2887579694939876,
+      "grad_norm": 0.7670302987098694,
+      "learning_rate": 0.00018530111375802735,
+      "loss": 2.7306,
+      "step": 3578
+    },
+    {
+      "epoch": 0.28883867323057055,
+      "grad_norm": 0.7582474946975708,
+      "learning_rate": 0.00018529287362826943,
+      "loss": 2.7715,
+      "step": 3579
+    },
+    {
+      "epoch": 0.2889193769671536,
+      "grad_norm": 0.750973105430603,
+      "learning_rate": 0.0001852846313727902,
+      "loss": 2.7147,
+      "step": 3580
+    },
+    {
+      "epoch": 0.28900008070373656,
+      "grad_norm": 0.771854043006897,
+      "learning_rate": 0.00018527638699179498,
+      "loss": 2.6874,
+      "step": 3581
+    },
+    {
+      "epoch": 0.2890807844403196,
+      "grad_norm": 0.785469651222229,
+      "learning_rate": 0.00018526814048548928,
+      "loss": 2.6858,
+      "step": 3582
+    },
+    {
+      "epoch": 0.28916148817690257,
+      "grad_norm": 0.7601101398468018,
+      "learning_rate": 0.00018525989185407864,
+      "loss": 2.6927,
+      "step": 3583
+    },
+    {
+      "epoch": 0.2892421919134856,
+      "grad_norm": 0.7313411831855774,
+      "learning_rate": 0.00018525164109776861,
+      "loss": 2.6813,
+      "step": 3584
+    },
+    {
+      "epoch": 0.2893228956500686,
+      "grad_norm": 0.7471718192100525,
+      "learning_rate": 0.00018524338821676483,
+      "loss": 2.6791,
+      "step": 3585
+    },
+    {
+      "epoch": 0.2894035993866516,
+      "grad_norm": 0.7615204453468323,
+      "learning_rate": 0.00018523513321127302,
+      "loss": 2.7767,
+      "step": 3586
+    },
+    {
+      "epoch": 0.2894843031232346,
+      "grad_norm": 0.766793966293335,
+      "learning_rate": 0.00018522687608149886,
+      "loss": 2.664,
+      "step": 3587
+    },
+    {
+      "epoch": 0.2895650068598176,
+      "grad_norm": 0.7897932529449463,
+      "learning_rate": 0.00018521861682764816,
+      "loss": 2.7148,
+      "step": 3588
+    },
+    {
+      "epoch": 0.2896457105964006,
+      "grad_norm": 0.7366818785667419,
+      "learning_rate": 0.00018521035544992679,
+      "loss": 2.69,
+      "step": 3589
+    },
+    {
+      "epoch": 0.28972641433298363,
+      "grad_norm": 0.7503829598426819,
+      "learning_rate": 0.00018520209194854058,
+      "loss": 2.7141,
+      "step": 3590
+    },
+    {
+      "epoch": 0.2898071180695666,
+      "grad_norm": 0.8064351081848145,
+      "learning_rate": 0.00018519382632369556,
+      "loss": 2.6738,
+      "step": 3591
+    },
+    {
+      "epoch": 0.28988782180614964,
+      "grad_norm": 0.7364048361778259,
+      "learning_rate": 0.00018518555857559768,
+      "loss": 2.6731,
+      "step": 3592
+    },
+    {
+      "epoch": 0.2899685255427326,
+      "grad_norm": 0.7065430283546448,
+      "learning_rate": 0.00018517728870445297,
+      "loss": 2.7314,
+      "step": 3593
+    },
+    {
+      "epoch": 0.29004922927931565,
+      "grad_norm": 0.8233428001403809,
+      "learning_rate": 0.0001851690167104676,
+      "loss": 2.727,
+      "step": 3594
+    },
+    {
+      "epoch": 0.2901299330158986,
+      "grad_norm": 0.7563758492469788,
+      "learning_rate": 0.00018516074259384768,
+      "loss": 2.665,
+      "step": 3595
+    },
+    {
+      "epoch": 0.29021063675248165,
+      "grad_norm": 0.7451249361038208,
+      "learning_rate": 0.00018515246635479943,
+      "loss": 2.7686,
+      "step": 3596
+    },
+    {
+      "epoch": 0.29029134048906463,
+      "grad_norm": 0.7374305725097656,
+      "learning_rate": 0.00018514418799352918,
+      "loss": 2.6466,
+      "step": 3597
+    },
+    {
+      "epoch": 0.29037204422564766,
+      "grad_norm": 0.7596983909606934,
+      "learning_rate": 0.00018513590751024315,
+      "loss": 2.6763,
+      "step": 3598
+    },
+    {
+      "epoch": 0.29045274796223064,
+      "grad_norm": 0.7808190584182739,
+      "learning_rate": 0.0001851276249051478,
+      "loss": 2.7362,
+      "step": 3599
+    },
+    {
+      "epoch": 0.29053345169881367,
+      "grad_norm": 0.765785276889801,
+      "learning_rate": 0.00018511934017844948,
+      "loss": 2.7049,
+      "step": 3600
+    },
+    {
+      "epoch": 0.29061415543539665,
+      "grad_norm": 0.7503563165664673,
+      "learning_rate": 0.0001851110533303547,
+      "loss": 2.6262,
+      "step": 3601
+    },
+    {
+      "epoch": 0.2906948591719797,
+      "grad_norm": 0.7287782430648804,
+      "learning_rate": 0.00018510276436107,
+      "loss": 2.7076,
+      "step": 3602
+    },
+    {
+      "epoch": 0.29077556290856266,
+      "grad_norm": 0.7748721837997437,
+      "learning_rate": 0.00018509447327080193,
+      "loss": 2.6945,
+      "step": 3603
+    },
+    {
+      "epoch": 0.2908562666451457,
+      "grad_norm": 0.7482423186302185,
+      "learning_rate": 0.00018508618005975714,
+      "loss": 2.7326,
+      "step": 3604
+    },
+    {
+      "epoch": 0.29093697038172867,
+      "grad_norm": 0.7708765864372253,
+      "learning_rate": 0.00018507788472814238,
+      "loss": 2.7602,
+      "step": 3605
+    },
+    {
+      "epoch": 0.2910176741183117,
+      "grad_norm": 0.7308060526847839,
+      "learning_rate": 0.0001850695872761643,
+      "loss": 2.6735,
+      "step": 3606
+    },
+    {
+      "epoch": 0.2910983778548947,
+      "grad_norm": 0.7512951493263245,
+      "learning_rate": 0.00018506128770402972,
+      "loss": 2.6877,
+      "step": 3607
+    },
+    {
+      "epoch": 0.2911790815914777,
+      "grad_norm": 0.6806616187095642,
+      "learning_rate": 0.00018505298601194552,
+      "loss": 2.6689,
+      "step": 3608
+    },
+    {
+      "epoch": 0.2912597853280607,
+      "grad_norm": 0.7825661301612854,
+      "learning_rate": 0.00018504468220011857,
+      "loss": 2.7108,
+      "step": 3609
+    },
+    {
+      "epoch": 0.2913404890646437,
+      "grad_norm": 0.8243381977081299,
+      "learning_rate": 0.00018503637626875584,
+      "loss": 2.6789,
+      "step": 3610
+    },
+    {
+      "epoch": 0.2914211928012267,
+      "grad_norm": 0.745012640953064,
+      "learning_rate": 0.00018502806821806429,
+      "loss": 2.7658,
+      "step": 3611
+    },
+    {
+      "epoch": 0.2915018965378097,
+      "grad_norm": 0.7091341018676758,
+      "learning_rate": 0.00018501975804825104,
+      "loss": 2.7046,
+      "step": 3612
+    },
+    {
+      "epoch": 0.2915826002743927,
+      "grad_norm": 0.729026734828949,
+      "learning_rate": 0.0001850114457595232,
+      "loss": 2.6692,
+      "step": 3613
+    },
+    {
+      "epoch": 0.29166330401097573,
+      "grad_norm": 0.8098071813583374,
+      "learning_rate": 0.00018500313135208786,
+      "loss": 2.712,
+      "step": 3614
+    },
+    {
+      "epoch": 0.2917440077475587,
+      "grad_norm": 0.7387483716011047,
+      "learning_rate": 0.0001849948148261523,
+      "loss": 2.6705,
+      "step": 3615
+    },
+    {
+      "epoch": 0.29182471148414174,
+      "grad_norm": 0.7904576659202576,
+      "learning_rate": 0.0001849864961819238,
+      "loss": 2.5969,
+      "step": 3616
+    },
+    {
+      "epoch": 0.2919054152207247,
+      "grad_norm": 0.7560681700706482,
+      "learning_rate": 0.00018497817541960964,
+      "loss": 2.6971,
+      "step": 3617
+    },
+    {
+      "epoch": 0.29198611895730775,
+      "grad_norm": 0.8488430976867676,
+      "learning_rate": 0.00018496985253941723,
+      "loss": 2.7367,
+      "step": 3618
+    },
+    {
+      "epoch": 0.2920668226938907,
+      "grad_norm": 0.7641268372535706,
+      "learning_rate": 0.00018496152754155399,
+      "loss": 2.6948,
+      "step": 3619
+    },
+    {
+      "epoch": 0.29214752643047376,
+      "grad_norm": 0.7219721674919128,
+      "learning_rate": 0.00018495320042622736,
+      "loss": 2.7225,
+      "step": 3620
+    },
+    {
+      "epoch": 0.29222823016705674,
+      "grad_norm": 0.7583872675895691,
+      "learning_rate": 0.00018494487119364493,
+      "loss": 2.7335,
+      "step": 3621
+    },
+    {
+      "epoch": 0.2923089339036397,
+      "grad_norm": 0.7771418690681458,
+      "learning_rate": 0.00018493653984401424,
+      "loss": 2.6712,
+      "step": 3622
+    },
+    {
+      "epoch": 0.29238963764022274,
+      "grad_norm": 0.7537891268730164,
+      "learning_rate": 0.00018492820637754296,
+      "loss": 2.7282,
+      "step": 3623
+    },
+    {
+      "epoch": 0.2924703413768057,
+      "grad_norm": 0.7334226965904236,
+      "learning_rate": 0.00018491987079443875,
+      "loss": 2.7072,
+      "step": 3624
+    },
+    {
+      "epoch": 0.29255104511338875,
+      "grad_norm": 0.7768076658248901,
+      "learning_rate": 0.00018491153309490942,
+      "loss": 2.7176,
+      "step": 3625
+    },
+    {
+      "epoch": 0.29263174884997173,
+      "grad_norm": 0.6831281185150146,
+      "learning_rate": 0.0001849031932791627,
+      "loss": 2.6982,
+      "step": 3626
+    },
+    {
+      "epoch": 0.29271245258655476,
+      "grad_norm": 0.7150557637214661,
+      "learning_rate": 0.00018489485134740648,
+      "loss": 2.7325,
+      "step": 3627
+    },
+    {
+      "epoch": 0.29279315632313774,
+      "grad_norm": 0.782667338848114,
+      "learning_rate": 0.00018488650729984863,
+      "loss": 2.7146,
+      "step": 3628
+    },
+    {
+      "epoch": 0.29287386005972077,
+      "grad_norm": 0.7718524932861328,
+      "learning_rate": 0.0001848781611366971,
+      "loss": 2.746,
+      "step": 3629
+    },
+    {
+      "epoch": 0.29295456379630375,
+      "grad_norm": 0.7066439390182495,
+      "learning_rate": 0.00018486981285815998,
+      "loss": 2.7497,
+      "step": 3630
+    },
+    {
+      "epoch": 0.2930352675328868,
+      "grad_norm": 0.7705665826797485,
+      "learning_rate": 0.00018486146246444522,
+      "loss": 2.6448,
+      "step": 3631
+    },
+    {
+      "epoch": 0.29311597126946976,
+      "grad_norm": 0.7334863543510437,
+      "learning_rate": 0.000184853109955761,
+      "loss": 2.6931,
+      "step": 3632
+    },
+    {
+      "epoch": 0.2931966750060528,
+      "grad_norm": 0.7903133630752563,
+      "learning_rate": 0.0001848447553323155,
+      "loss": 2.6954,
+      "step": 3633
+    },
+    {
+      "epoch": 0.29327737874263576,
+      "grad_norm": 0.6821191310882568,
+      "learning_rate": 0.00018483639859431689,
+      "loss": 2.6165,
+      "step": 3634
+    },
+    {
+      "epoch": 0.2933580824792188,
+      "grad_norm": 0.7187811136245728,
+      "learning_rate": 0.00018482803974197344,
+      "loss": 2.6387,
+      "step": 3635
+    },
+    {
+      "epoch": 0.2934387862158018,
+      "grad_norm": 0.7429843544960022,
+      "learning_rate": 0.00018481967877549354,
+      "loss": 2.6848,
+      "step": 3636
+    },
+    {
+      "epoch": 0.2935194899523848,
+      "grad_norm": 0.7431524395942688,
+      "learning_rate": 0.0001848113156950855,
+      "loss": 2.7044,
+      "step": 3637
+    },
+    {
+      "epoch": 0.2936001936889678,
+      "grad_norm": 0.7008687853813171,
+      "learning_rate": 0.00018480295050095778,
+      "loss": 2.6922,
+      "step": 3638
+    },
+    {
+      "epoch": 0.2936808974255508,
+      "grad_norm": 0.7106652855873108,
+      "learning_rate": 0.00018479458319331884,
+      "loss": 2.6845,
+      "step": 3639
+    },
+    {
+      "epoch": 0.2937616011621338,
+      "grad_norm": 0.7288951873779297,
+      "learning_rate": 0.00018478621377237723,
+      "loss": 2.7017,
+      "step": 3640
+    },
+    {
+      "epoch": 0.2938423048987168,
+      "grad_norm": 0.7228607535362244,
+      "learning_rate": 0.00018477784223834155,
+      "loss": 2.7449,
+      "step": 3641
+    },
+    {
+      "epoch": 0.2939230086352998,
+      "grad_norm": 0.7180825471878052,
+      "learning_rate": 0.00018476946859142043,
+      "loss": 2.7291,
+      "step": 3642
+    },
+    {
+      "epoch": 0.29400371237188283,
+      "grad_norm": 0.7854947447776794,
+      "learning_rate": 0.00018476109283182258,
+      "loss": 2.7619,
+      "step": 3643
+    },
+    {
+      "epoch": 0.2940844161084658,
+      "grad_norm": 0.7871318459510803,
+      "learning_rate": 0.00018475271495975673,
+      "loss": 2.6695,
+      "step": 3644
+    },
+    {
+      "epoch": 0.29416511984504884,
+      "grad_norm": 0.7813127636909485,
+      "learning_rate": 0.00018474433497543165,
+      "loss": 2.735,
+      "step": 3645
+    },
+    {
+      "epoch": 0.2942458235816318,
+      "grad_norm": 0.7835291028022766,
+      "learning_rate": 0.00018473595287905623,
+      "loss": 2.7336,
+      "step": 3646
+    },
+    {
+      "epoch": 0.29432652731821485,
+      "grad_norm": 0.6970148682594299,
+      "learning_rate": 0.00018472756867083935,
+      "loss": 2.6912,
+      "step": 3647
+    },
+    {
+      "epoch": 0.2944072310547978,
+      "grad_norm": 0.7968462109565735,
+      "learning_rate": 0.00018471918235098998,
+      "loss": 2.6889,
+      "step": 3648
+    },
+    {
+      "epoch": 0.29448793479138086,
+      "grad_norm": 0.7011313438415527,
+      "learning_rate": 0.00018471079391971714,
+      "loss": 2.6989,
+      "step": 3649
+    },
+    {
+      "epoch": 0.29456863852796383,
+      "grad_norm": 0.8047335743904114,
+      "learning_rate": 0.00018470240337722991,
+      "loss": 2.6827,
+      "step": 3650
+    },
+    {
+      "epoch": 0.29464934226454687,
+      "grad_norm": 0.7446332573890686,
+      "learning_rate": 0.00018469401072373733,
+      "loss": 2.7089,
+      "step": 3651
+    },
+    {
+      "epoch": 0.29473004600112984,
+      "grad_norm": 0.7610359191894531,
+      "learning_rate": 0.00018468561595944862,
+      "loss": 2.6766,
+      "step": 3652
+    },
+    {
+      "epoch": 0.2948107497377129,
+      "grad_norm": 0.7705755233764648,
+      "learning_rate": 0.000184677219084573,
+      "loss": 2.7445,
+      "step": 3653
+    },
+    {
+      "epoch": 0.29489145347429585,
+      "grad_norm": 0.7466446757316589,
+      "learning_rate": 0.00018466882009931973,
+      "loss": 2.726,
+      "step": 3654
+    },
+    {
+      "epoch": 0.2949721572108789,
+      "grad_norm": 0.7912059426307678,
+      "learning_rate": 0.00018466041900389813,
+      "loss": 2.6865,
+      "step": 3655
+    },
+    {
+      "epoch": 0.29505286094746186,
+      "grad_norm": 0.722588837146759,
+      "learning_rate": 0.00018465201579851757,
+      "loss": 2.7039,
+      "step": 3656
+    },
+    {
+      "epoch": 0.2951335646840449,
+      "grad_norm": 0.739311933517456,
+      "learning_rate": 0.00018464361048338752,
+      "loss": 2.6991,
+      "step": 3657
+    },
+    {
+      "epoch": 0.29521426842062787,
+      "grad_norm": 0.7784128785133362,
+      "learning_rate": 0.00018463520305871743,
+      "loss": 2.753,
+      "step": 3658
+    },
+    {
+      "epoch": 0.2952949721572109,
+      "grad_norm": 0.8261777758598328,
+      "learning_rate": 0.00018462679352471682,
+      "loss": 2.7257,
+      "step": 3659
+    },
+    {
+      "epoch": 0.2953756758937939,
+      "grad_norm": 0.7510927319526672,
+      "learning_rate": 0.0001846183818815953,
+      "loss": 2.6981,
+      "step": 3660
+    },
+    {
+      "epoch": 0.2954563796303769,
+      "grad_norm": 0.7403035163879395,
+      "learning_rate": 0.00018460996812956254,
+      "loss": 2.744,
+      "step": 3661
+    },
+    {
+      "epoch": 0.2955370833669599,
+      "grad_norm": 0.7927733063697815,
+      "learning_rate": 0.00018460155226882817,
+      "loss": 2.6304,
+      "step": 3662
+    },
+    {
+      "epoch": 0.2956177871035429,
+      "grad_norm": 0.7923495769500732,
+      "learning_rate": 0.000184593134299602,
+      "loss": 2.7882,
+      "step": 3663
+    },
+    {
+      "epoch": 0.2956984908401259,
+      "grad_norm": 0.7639210224151611,
+      "learning_rate": 0.00018458471422209377,
+      "loss": 2.7171,
+      "step": 3664
+    },
+    {
+      "epoch": 0.2957791945767089,
+      "grad_norm": 0.736652672290802,
+      "learning_rate": 0.00018457629203651337,
+      "loss": 2.7479,
+      "step": 3665
+    },
+    {
+      "epoch": 0.2958598983132919,
+      "grad_norm": 0.7718610763549805,
+      "learning_rate": 0.00018456786774307066,
+      "loss": 2.7135,
+      "step": 3666
+    },
+    {
+      "epoch": 0.29594060204987493,
+      "grad_norm": 0.7711780071258545,
+      "learning_rate": 0.00018455944134197565,
+      "loss": 2.6867,
+      "step": 3667
+    },
+    {
+      "epoch": 0.2960213057864579,
+      "grad_norm": 0.7202491760253906,
+      "learning_rate": 0.0001845510128334383,
+      "loss": 2.6657,
+      "step": 3668
+    },
+    {
+      "epoch": 0.29610200952304094,
+      "grad_norm": 0.8155657649040222,
+      "learning_rate": 0.00018454258221766869,
+      "loss": 2.7342,
+      "step": 3669
+    },
+    {
+      "epoch": 0.2961827132596239,
+      "grad_norm": 0.7972069382667542,
+      "learning_rate": 0.00018453414949487696,
+      "loss": 2.7351,
+      "step": 3670
+    },
+    {
+      "epoch": 0.29626341699620695,
+      "grad_norm": 0.8645625710487366,
+      "learning_rate": 0.00018452571466527325,
+      "loss": 2.6778,
+      "step": 3671
+    },
+    {
+      "epoch": 0.29634412073278993,
+      "grad_norm": 0.7410334944725037,
+      "learning_rate": 0.00018451727772906775,
+      "loss": 2.7228,
+      "step": 3672
+    },
+    {
+      "epoch": 0.2964248244693729,
+      "grad_norm": 0.7845733165740967,
+      "learning_rate": 0.0001845088386864708,
+      "loss": 2.7068,
+      "step": 3673
+    },
+    {
+      "epoch": 0.29650552820595594,
+      "grad_norm": 0.7709881067276001,
+      "learning_rate": 0.00018450039753769266,
+      "loss": 2.676,
+      "step": 3674
+    },
+    {
+      "epoch": 0.2965862319425389,
+      "grad_norm": 0.7214749455451965,
+      "learning_rate": 0.00018449195428294371,
+      "loss": 2.6488,
+      "step": 3675
+    },
+    {
+      "epoch": 0.29666693567912195,
+      "grad_norm": 0.7467561960220337,
+      "learning_rate": 0.00018448350892243443,
+      "loss": 2.7262,
+      "step": 3676
+    },
+    {
+      "epoch": 0.2967476394157049,
+      "grad_norm": 0.8412678241729736,
+      "learning_rate": 0.00018447506145637522,
+      "loss": 2.7898,
+      "step": 3677
+    },
+    {
+      "epoch": 0.29682834315228795,
+      "grad_norm": 0.7130109071731567,
+      "learning_rate": 0.00018446661188497668,
+      "loss": 2.7344,
+      "step": 3678
+    },
+    {
+      "epoch": 0.29690904688887093,
+      "grad_norm": 0.7807374000549316,
+      "learning_rate": 0.00018445816020844937,
+      "loss": 2.7198,
+      "step": 3679
+    },
+    {
+      "epoch": 0.29698975062545396,
+      "grad_norm": 0.8497760891914368,
+      "learning_rate": 0.00018444970642700394,
+      "loss": 2.7479,
+      "step": 3680
+    },
+    {
+      "epoch": 0.29707045436203694,
+      "grad_norm": 0.6827178001403809,
+      "learning_rate": 0.0001844412505408511,
+      "loss": 2.727,
+      "step": 3681
+    },
+    {
+      "epoch": 0.29715115809861997,
+      "grad_norm": 0.8063304424285889,
+      "learning_rate": 0.00018443279255020152,
+      "loss": 2.7896,
+      "step": 3682
+    },
+    {
+      "epoch": 0.29723186183520295,
+      "grad_norm": 0.7759353518486023,
+      "learning_rate": 0.00018442433245526604,
+      "loss": 2.7014,
+      "step": 3683
+    },
+    {
+      "epoch": 0.297312565571786,
+      "grad_norm": 0.7380958199501038,
+      "learning_rate": 0.00018441587025625554,
+      "loss": 2.6665,
+      "step": 3684
+    },
+    {
+      "epoch": 0.29739326930836896,
+      "grad_norm": 0.7623556852340698,
+      "learning_rate": 0.00018440740595338087,
+      "loss": 2.6955,
+      "step": 3685
+    },
+    {
+      "epoch": 0.297473973044952,
+      "grad_norm": 0.8204537630081177,
+      "learning_rate": 0.000184398939546853,
+      "loss": 2.6854,
+      "step": 3686
+    },
+    {
+      "epoch": 0.29755467678153497,
+      "grad_norm": 0.7346726655960083,
+      "learning_rate": 0.00018439047103688293,
+      "loss": 2.6664,
+      "step": 3687
+    },
+    {
+      "epoch": 0.297635380518118,
+      "grad_norm": 0.777860701084137,
+      "learning_rate": 0.00018438200042368173,
+      "loss": 2.6423,
+      "step": 3688
+    },
+    {
+      "epoch": 0.297716084254701,
+      "grad_norm": 0.7331553101539612,
+      "learning_rate": 0.00018437352770746054,
+      "loss": 2.6137,
+      "step": 3689
+    },
+    {
+      "epoch": 0.297796787991284,
+      "grad_norm": 0.7634466290473938,
+      "learning_rate": 0.00018436505288843043,
+      "loss": 2.7266,
+      "step": 3690
+    },
+    {
+      "epoch": 0.297877491727867,
+      "grad_norm": 0.8151016235351562,
+      "learning_rate": 0.00018435657596680268,
+      "loss": 2.7373,
+      "step": 3691
+    },
+    {
+      "epoch": 0.29795819546445,
+      "grad_norm": 0.7806773781776428,
+      "learning_rate": 0.00018434809694278857,
+      "loss": 2.7011,
+      "step": 3692
+    },
+    {
+      "epoch": 0.298038899201033,
+      "grad_norm": 0.7575243711471558,
+      "learning_rate": 0.00018433961581659935,
+      "loss": 2.6601,
+      "step": 3693
+    },
+    {
+      "epoch": 0.298119602937616,
+      "grad_norm": 0.7527276873588562,
+      "learning_rate": 0.00018433113258844647,
+      "loss": 2.6864,
+      "step": 3694
+    },
+    {
+      "epoch": 0.298200306674199,
+      "grad_norm": 0.8024318218231201,
+      "learning_rate": 0.0001843226472585413,
+      "loss": 2.728,
+      "step": 3695
+    },
+    {
+      "epoch": 0.29828101041078203,
+      "grad_norm": 0.7549982666969299,
+      "learning_rate": 0.0001843141598270954,
+      "loss": 2.6834,
+      "step": 3696
+    },
+    {
+      "epoch": 0.298361714147365,
+      "grad_norm": 0.7699971199035645,
+      "learning_rate": 0.0001843056702943202,
+      "loss": 2.7209,
+      "step": 3697
+    },
+    {
+      "epoch": 0.29844241788394804,
+      "grad_norm": 0.823842465877533,
+      "learning_rate": 0.0001842971786604273,
+      "loss": 2.6924,
+      "step": 3698
+    },
+    {
+      "epoch": 0.298523121620531,
+      "grad_norm": 0.7645791172981262,
+      "learning_rate": 0.00018428868492562837,
+      "loss": 2.6821,
+      "step": 3699
+    },
+    {
+      "epoch": 0.29860382535711405,
+      "grad_norm": 0.7530989050865173,
+      "learning_rate": 0.00018428018909013506,
+      "loss": 2.7592,
+      "step": 3700
+    },
+    {
+      "epoch": 0.298684529093697,
+      "grad_norm": 0.7958168387413025,
+      "learning_rate": 0.00018427169115415914,
+      "loss": 2.6925,
+      "step": 3701
+    },
+    {
+      "epoch": 0.29876523283028006,
+      "grad_norm": 0.7777522802352905,
+      "learning_rate": 0.00018426319111791242,
+      "loss": 2.6757,
+      "step": 3702
+    },
+    {
+      "epoch": 0.29884593656686304,
+      "grad_norm": 0.7418079972267151,
+      "learning_rate": 0.00018425468898160667,
+      "loss": 2.6445,
+      "step": 3703
+    },
+    {
+      "epoch": 0.29892664030344607,
+      "grad_norm": 0.7591132521629333,
+      "learning_rate": 0.00018424618474545382,
+      "loss": 2.7157,
+      "step": 3704
+    },
+    {
+      "epoch": 0.29900734404002904,
+      "grad_norm": 0.7591627836227417,
+      "learning_rate": 0.00018423767840966586,
+      "loss": 2.6691,
+      "step": 3705
+    },
+    {
+      "epoch": 0.2990880477766121,
+      "grad_norm": 0.7934779524803162,
+      "learning_rate": 0.00018422916997445476,
+      "loss": 2.7262,
+      "step": 3706
+    },
+    {
+      "epoch": 0.29916875151319505,
+      "grad_norm": 0.7964254021644592,
+      "learning_rate": 0.00018422065944003252,
+      "loss": 2.6196,
+      "step": 3707
+    },
+    {
+      "epoch": 0.2992494552497781,
+      "grad_norm": 0.7448374032974243,
+      "learning_rate": 0.0001842121468066113,
+      "loss": 2.6732,
+      "step": 3708
+    },
+    {
+      "epoch": 0.29933015898636106,
+      "grad_norm": 0.7813000679016113,
+      "learning_rate": 0.00018420363207440329,
+      "loss": 2.6978,
+      "step": 3709
+    },
+    {
+      "epoch": 0.2994108627229441,
+      "grad_norm": 0.7760851979255676,
+      "learning_rate": 0.00018419511524362064,
+      "loss": 2.7466,
+      "step": 3710
+    },
+    {
+      "epoch": 0.29949156645952707,
+      "grad_norm": 0.7786797881126404,
+      "learning_rate": 0.00018418659631447564,
+      "loss": 2.7044,
+      "step": 3711
+    },
+    {
+      "epoch": 0.2995722701961101,
+      "grad_norm": 0.7860158085823059,
+      "learning_rate": 0.00018417807528718055,
+      "loss": 2.6587,
+      "step": 3712
+    },
+    {
+      "epoch": 0.2996529739326931,
+      "grad_norm": 0.8327339291572571,
+      "learning_rate": 0.0001841695521619478,
+      "loss": 2.7112,
+      "step": 3713
+    },
+    {
+      "epoch": 0.2997336776692761,
+      "grad_norm": 0.7535735368728638,
+      "learning_rate": 0.00018416102693898982,
+      "loss": 2.726,
+      "step": 3714
+    },
+    {
+      "epoch": 0.2998143814058591,
+      "grad_norm": 0.7781090140342712,
+      "learning_rate": 0.000184152499618519,
+      "loss": 2.7238,
+      "step": 3715
+    },
+    {
+      "epoch": 0.2998950851424421,
+      "grad_norm": 0.7700545191764832,
+      "learning_rate": 0.00018414397020074795,
+      "loss": 2.7081,
+      "step": 3716
+    },
+    {
+      "epoch": 0.2999757888790251,
+      "grad_norm": 0.7578303217887878,
+      "learning_rate": 0.0001841354386858892,
+      "loss": 2.6591,
+      "step": 3717
+    },
+    {
+      "epoch": 0.30005649261560813,
+      "grad_norm": 0.7506501078605652,
+      "learning_rate": 0.00018412690507415538,
+      "loss": 2.6551,
+      "step": 3718
+    },
+    {
+      "epoch": 0.3001371963521911,
+      "grad_norm": 0.7869547009468079,
+      "learning_rate": 0.00018411836936575918,
+      "loss": 2.7169,
+      "step": 3719
+    },
+    {
+      "epoch": 0.30021790008877414,
+      "grad_norm": 0.7547428607940674,
+      "learning_rate": 0.00018410983156091332,
+      "loss": 2.7498,
+      "step": 3720
+    },
+    {
+      "epoch": 0.3002986038253571,
+      "grad_norm": 0.7829383015632629,
+      "learning_rate": 0.0001841012916598306,
+      "loss": 2.6885,
+      "step": 3721
+    },
+    {
+      "epoch": 0.30037930756194015,
+      "grad_norm": 0.8469082117080688,
+      "learning_rate": 0.00018409274966272386,
+      "loss": 2.7594,
+      "step": 3722
+    },
+    {
+      "epoch": 0.3004600112985231,
+      "grad_norm": 0.7690171599388123,
+      "learning_rate": 0.00018408420556980596,
+      "loss": 2.7892,
+      "step": 3723
+    },
+    {
+      "epoch": 0.3005407150351061,
+      "grad_norm": 0.7295899987220764,
+      "learning_rate": 0.00018407565938128987,
+      "loss": 2.7023,
+      "step": 3724
+    },
+    {
+      "epoch": 0.30062141877168913,
+      "grad_norm": 0.7249528169631958,
+      "learning_rate": 0.00018406711109738856,
+      "loss": 2.7135,
+      "step": 3725
+    },
+    {
+      "epoch": 0.3007021225082721,
+      "grad_norm": 0.7237234711647034,
+      "learning_rate": 0.0001840585607183151,
+      "loss": 2.6117,
+      "step": 3726
+    },
+    {
+      "epoch": 0.30078282624485514,
+      "grad_norm": 0.7426557540893555,
+      "learning_rate": 0.00018405000824428256,
+      "loss": 2.7202,
+      "step": 3727
+    },
+    {
+      "epoch": 0.3008635299814381,
+      "grad_norm": 0.7572938799858093,
+      "learning_rate": 0.00018404145367550414,
+      "loss": 2.7373,
+      "step": 3728
+    },
+    {
+      "epoch": 0.30094423371802115,
+      "grad_norm": 0.7198675274848938,
+      "learning_rate": 0.00018403289701219295,
+      "loss": 2.6675,
+      "step": 3729
+    },
+    {
+      "epoch": 0.3010249374546041,
+      "grad_norm": 0.722532331943512,
+      "learning_rate": 0.00018402433825456235,
+      "loss": 2.6933,
+      "step": 3730
+    },
+    {
+      "epoch": 0.30110564119118716,
+      "grad_norm": 0.7621530890464783,
+      "learning_rate": 0.0001840157774028256,
+      "loss": 2.6951,
+      "step": 3731
+    },
+    {
+      "epoch": 0.30118634492777013,
+      "grad_norm": 0.7435615062713623,
+      "learning_rate": 0.00018400721445719604,
+      "loss": 2.7323,
+      "step": 3732
+    },
+    {
+      "epoch": 0.30126704866435317,
+      "grad_norm": 0.7233619689941406,
+      "learning_rate": 0.00018399864941788708,
+      "loss": 2.6789,
+      "step": 3733
+    },
+    {
+      "epoch": 0.30134775240093614,
+      "grad_norm": 0.7421496510505676,
+      "learning_rate": 0.00018399008228511224,
+      "loss": 2.72,
+      "step": 3734
+    },
+    {
+      "epoch": 0.3014284561375192,
+      "grad_norm": 0.7250909805297852,
+      "learning_rate": 0.000183981513059085,
+      "loss": 2.6717,
+      "step": 3735
+    },
+    {
+      "epoch": 0.30150915987410215,
+      "grad_norm": 0.7642899751663208,
+      "learning_rate": 0.0001839729417400189,
+      "loss": 2.6823,
+      "step": 3736
+    },
+    {
+      "epoch": 0.3015898636106852,
+      "grad_norm": 0.7434508204460144,
+      "learning_rate": 0.00018396436832812758,
+      "loss": 2.6441,
+      "step": 3737
+    },
+    {
+      "epoch": 0.30167056734726816,
+      "grad_norm": 0.7163311839103699,
+      "learning_rate": 0.00018395579282362473,
+      "loss": 2.6736,
+      "step": 3738
+    },
+    {
+      "epoch": 0.3017512710838512,
+      "grad_norm": 0.6936792731285095,
+      "learning_rate": 0.00018394721522672404,
+      "loss": 2.6792,
+      "step": 3739
+    },
+    {
+      "epoch": 0.30183197482043417,
+      "grad_norm": 0.7791975736618042,
+      "learning_rate": 0.0001839386355376393,
+      "loss": 2.653,
+      "step": 3740
+    },
+    {
+      "epoch": 0.3019126785570172,
+      "grad_norm": 0.7902694940567017,
+      "learning_rate": 0.00018393005375658437,
+      "loss": 2.7448,
+      "step": 3741
+    },
+    {
+      "epoch": 0.3019933822936002,
+      "grad_norm": 0.7405624389648438,
+      "learning_rate": 0.0001839214698837731,
+      "loss": 2.6977,
+      "step": 3742
+    },
+    {
+      "epoch": 0.3020740860301832,
+      "grad_norm": 0.8033632040023804,
+      "learning_rate": 0.00018391288391941943,
+      "loss": 2.7468,
+      "step": 3743
+    },
+    {
+      "epoch": 0.3021547897667662,
+      "grad_norm": 0.8148884177207947,
+      "learning_rate": 0.00018390429586373735,
+      "loss": 2.6992,
+      "step": 3744
+    },
+    {
+      "epoch": 0.3022354935033492,
+      "grad_norm": 0.7633625268936157,
+      "learning_rate": 0.00018389570571694089,
+      "loss": 2.6604,
+      "step": 3745
+    },
+    {
+      "epoch": 0.3023161972399322,
+      "grad_norm": 0.8687180876731873,
+      "learning_rate": 0.00018388711347924413,
+      "loss": 2.6808,
+      "step": 3746
+    },
+    {
+      "epoch": 0.3023969009765152,
+      "grad_norm": 0.6974104046821594,
+      "learning_rate": 0.0001838785191508612,
+      "loss": 2.7613,
+      "step": 3747
+    },
+    {
+      "epoch": 0.3024776047130982,
+      "grad_norm": 0.7919288873672485,
+      "learning_rate": 0.00018386992273200633,
+      "loss": 2.664,
+      "step": 3748
+    },
+    {
+      "epoch": 0.30255830844968123,
+      "grad_norm": 0.7708829045295715,
+      "learning_rate": 0.00018386132422289374,
+      "loss": 2.7703,
+      "step": 3749
+    },
+    {
+      "epoch": 0.3026390121862642,
+      "grad_norm": 0.7099813222885132,
+      "learning_rate": 0.00018385272362373775,
+      "loss": 2.6485,
+      "step": 3750
+    },
+    {
+      "epoch": 0.30271971592284724,
+      "grad_norm": 0.7629622220993042,
+      "learning_rate": 0.0001838441209347527,
+      "loss": 2.7339,
+      "step": 3751
+    },
+    {
+      "epoch": 0.3028004196594302,
+      "grad_norm": 0.727275550365448,
+      "learning_rate": 0.00018383551615615295,
+      "loss": 2.7194,
+      "step": 3752
+    },
+    {
+      "epoch": 0.30288112339601325,
+      "grad_norm": 0.7158832550048828,
+      "learning_rate": 0.00018382690928815302,
+      "loss": 2.6698,
+      "step": 3753
+    },
+    {
+      "epoch": 0.30296182713259623,
+      "grad_norm": 0.8075565099716187,
+      "learning_rate": 0.00018381830033096735,
+      "loss": 2.7198,
+      "step": 3754
+    },
+    {
+      "epoch": 0.30304253086917926,
+      "grad_norm": 0.7949094176292419,
+      "learning_rate": 0.00018380968928481057,
+      "loss": 2.7048,
+      "step": 3755
+    },
+    {
+      "epoch": 0.30312323460576224,
+      "grad_norm": 0.7009503841400146,
+      "learning_rate": 0.00018380107614989724,
+      "loss": 2.709,
+      "step": 3756
+    },
+    {
+      "epoch": 0.30320393834234527,
+      "grad_norm": 0.668574869632721,
+      "learning_rate": 0.00018379246092644204,
+      "loss": 2.6515,
+      "step": 3757
+    },
+    {
+      "epoch": 0.30328464207892825,
+      "grad_norm": 0.7470806241035461,
+      "learning_rate": 0.00018378384361465968,
+      "loss": 2.7577,
+      "step": 3758
+    },
+    {
+      "epoch": 0.3033653458155113,
+      "grad_norm": 0.7529913783073425,
+      "learning_rate": 0.0001837752242147649,
+      "loss": 2.7189,
+      "step": 3759
+    },
+    {
+      "epoch": 0.30344604955209425,
+      "grad_norm": 0.7373302578926086,
+      "learning_rate": 0.00018376660272697258,
+      "loss": 2.7197,
+      "step": 3760
+    },
+    {
+      "epoch": 0.3035267532886773,
+      "grad_norm": 0.7650466561317444,
+      "learning_rate": 0.0001837579791514975,
+      "loss": 2.6613,
+      "step": 3761
+    },
+    {
+      "epoch": 0.30360745702526026,
+      "grad_norm": 0.775209903717041,
+      "learning_rate": 0.00018374935348855468,
+      "loss": 2.6454,
+      "step": 3762
+    },
+    {
+      "epoch": 0.3036881607618433,
+      "grad_norm": 0.7049290537834167,
+      "learning_rate": 0.00018374072573835903,
+      "loss": 2.6663,
+      "step": 3763
+    },
+    {
+      "epoch": 0.30376886449842627,
+      "grad_norm": 0.7060630917549133,
+      "learning_rate": 0.0001837320959011256,
+      "loss": 2.6908,
+      "step": 3764
+    },
+    {
+      "epoch": 0.3038495682350093,
+      "grad_norm": 0.7561464905738831,
+      "learning_rate": 0.00018372346397706944,
+      "loss": 2.673,
+      "step": 3765
+    },
+    {
+      "epoch": 0.3039302719715923,
+      "grad_norm": 0.7293568849563599,
+      "learning_rate": 0.0001837148299664057,
+      "loss": 2.6431,
+      "step": 3766
+    },
+    {
+      "epoch": 0.3040109757081753,
+      "grad_norm": 0.8460379838943481,
+      "learning_rate": 0.00018370619386934962,
+      "loss": 2.7493,
+      "step": 3767
+    },
+    {
+      "epoch": 0.3040916794447583,
+      "grad_norm": 0.8136082291603088,
+      "learning_rate": 0.00018369755568611632,
+      "loss": 2.7298,
+      "step": 3768
+    },
+    {
+      "epoch": 0.3041723831813413,
+      "grad_norm": 0.6916636824607849,
+      "learning_rate": 0.00018368891541692116,
+      "loss": 2.7173,
+      "step": 3769
+    },
+    {
+      "epoch": 0.3042530869179243,
+      "grad_norm": 0.7547643780708313,
+      "learning_rate": 0.0001836802730619795,
+      "loss": 2.6343,
+      "step": 3770
+    },
+    {
+      "epoch": 0.30433379065450733,
+      "grad_norm": 0.7439205050468445,
+      "learning_rate": 0.00018367162862150665,
+      "loss": 2.6627,
+      "step": 3771
+    },
+    {
+      "epoch": 0.3044144943910903,
+      "grad_norm": 0.7781087756156921,
+      "learning_rate": 0.0001836629820957181,
+      "loss": 2.7223,
+      "step": 3772
+    },
+    {
+      "epoch": 0.30449519812767334,
+      "grad_norm": 0.7876880764961243,
+      "learning_rate": 0.00018365433348482935,
+      "loss": 2.7139,
+      "step": 3773
+    },
+    {
+      "epoch": 0.3045759018642563,
+      "grad_norm": 0.7571346163749695,
+      "learning_rate": 0.00018364568278905595,
+      "loss": 2.6939,
+      "step": 3774
+    },
+    {
+      "epoch": 0.3046566056008393,
+      "grad_norm": 0.9011813402175903,
+      "learning_rate": 0.00018363703000861346,
+      "loss": 2.7516,
+      "step": 3775
+    },
+    {
+      "epoch": 0.3047373093374223,
+      "grad_norm": 0.7809761762619019,
+      "learning_rate": 0.00018362837514371755,
+      "loss": 2.7587,
+      "step": 3776
+    },
+    {
+      "epoch": 0.3048180130740053,
+      "grad_norm": 0.7486867308616638,
+      "learning_rate": 0.00018361971819458393,
+      "loss": 2.6617,
+      "step": 3777
+    },
+    {
+      "epoch": 0.30489871681058833,
+      "grad_norm": 0.7434267401695251,
+      "learning_rate": 0.00018361105916142836,
+      "loss": 2.7328,
+      "step": 3778
+    },
+    {
+      "epoch": 0.3049794205471713,
+      "grad_norm": 0.7895822525024414,
+      "learning_rate": 0.0001836023980444666,
+      "loss": 2.7038,
+      "step": 3779
+    },
+    {
+      "epoch": 0.30506012428375434,
+      "grad_norm": 0.7329267263412476,
+      "learning_rate": 0.00018359373484391458,
+      "loss": 2.6533,
+      "step": 3780
+    },
+    {
+      "epoch": 0.3051408280203373,
+      "grad_norm": 0.7578477263450623,
+      "learning_rate": 0.00018358506955998817,
+      "loss": 2.723,
+      "step": 3781
+    },
+    {
+      "epoch": 0.30522153175692035,
+      "grad_norm": 0.7174215316772461,
+      "learning_rate": 0.0001835764021929033,
+      "loss": 2.7665,
+      "step": 3782
+    },
+    {
+      "epoch": 0.3053022354935033,
+      "grad_norm": 0.7261673808097839,
+      "learning_rate": 0.00018356773274287605,
+      "loss": 2.7239,
+      "step": 3783
+    },
+    {
+      "epoch": 0.30538293923008636,
+      "grad_norm": 0.7550768852233887,
+      "learning_rate": 0.00018355906121012244,
+      "loss": 2.6952,
+      "step": 3784
+    },
+    {
+      "epoch": 0.30546364296666934,
+      "grad_norm": 0.7805373668670654,
+      "learning_rate": 0.0001835503875948586,
+      "loss": 2.6453,
+      "step": 3785
+    },
+    {
+      "epoch": 0.30554434670325237,
+      "grad_norm": 0.7753674983978271,
+      "learning_rate": 0.0001835417118973007,
+      "loss": 2.7188,
+      "step": 3786
+    },
+    {
+      "epoch": 0.30562505043983534,
+      "grad_norm": 0.719774603843689,
+      "learning_rate": 0.00018353303411766496,
+      "loss": 2.69,
+      "step": 3787
+    },
+    {
+      "epoch": 0.3057057541764184,
+      "grad_norm": 0.786780059337616,
+      "learning_rate": 0.00018352435425616763,
+      "loss": 2.7015,
+      "step": 3788
+    },
+    {
+      "epoch": 0.30578645791300135,
+      "grad_norm": 0.7481613159179688,
+      "learning_rate": 0.00018351567231302508,
+      "loss": 2.6267,
+      "step": 3789
+    },
+    {
+      "epoch": 0.3058671616495844,
+      "grad_norm": 0.8138384222984314,
+      "learning_rate": 0.00018350698828845365,
+      "loss": 2.7301,
+      "step": 3790
+    },
+    {
+      "epoch": 0.30594786538616736,
+      "grad_norm": 0.7911081314086914,
+      "learning_rate": 0.00018349830218266982,
+      "loss": 2.6661,
+      "step": 3791
+    },
+    {
+      "epoch": 0.3060285691227504,
+      "grad_norm": 0.763179361820221,
+      "learning_rate": 0.00018348961399588997,
+      "loss": 2.6509,
+      "step": 3792
+    },
+    {
+      "epoch": 0.30610927285933337,
+      "grad_norm": 0.8214982748031616,
+      "learning_rate": 0.00018348092372833072,
+      "loss": 2.6951,
+      "step": 3793
+    },
+    {
+      "epoch": 0.3061899765959164,
+      "grad_norm": 0.7271003127098083,
+      "learning_rate": 0.00018347223138020865,
+      "loss": 2.7227,
+      "step": 3794
+    },
+    {
+      "epoch": 0.3062706803324994,
+      "grad_norm": 0.7727730870246887,
+      "learning_rate": 0.00018346353695174037,
+      "loss": 2.721,
+      "step": 3795
+    },
+    {
+      "epoch": 0.3063513840690824,
+      "grad_norm": 0.844895601272583,
+      "learning_rate": 0.00018345484044314257,
+      "loss": 2.6757,
+      "step": 3796
+    },
+    {
+      "epoch": 0.3064320878056654,
+      "grad_norm": 0.7409898638725281,
+      "learning_rate": 0.00018344614185463197,
+      "loss": 2.6798,
+      "step": 3797
+    },
+    {
+      "epoch": 0.3065127915422484,
+      "grad_norm": 0.8284425139427185,
+      "learning_rate": 0.00018343744118642542,
+      "loss": 2.7573,
+      "step": 3798
+    },
+    {
+      "epoch": 0.3065934952788314,
+      "grad_norm": 0.7535427808761597,
+      "learning_rate": 0.00018342873843873973,
+      "loss": 2.7026,
+      "step": 3799
+    },
+    {
+      "epoch": 0.30667419901541443,
+      "grad_norm": 0.8013898730278015,
+      "learning_rate": 0.00018342003361179176,
+      "loss": 2.7331,
+      "step": 3800
+    },
+    {
+      "epoch": 0.3067549027519974,
+      "grad_norm": 0.7458386421203613,
+      "learning_rate": 0.0001834113267057985,
+      "loss": 2.6976,
+      "step": 3801
+    },
+    {
+      "epoch": 0.30683560648858044,
+      "grad_norm": 0.8333673477172852,
+      "learning_rate": 0.00018340261772097695,
+      "loss": 2.7064,
+      "step": 3802
+    },
+    {
+      "epoch": 0.3069163102251634,
+      "grad_norm": 0.7273485064506531,
+      "learning_rate": 0.00018339390665754414,
+      "loss": 2.6619,
+      "step": 3803
+    },
+    {
+      "epoch": 0.30699701396174645,
+      "grad_norm": 0.8199014067649841,
+      "learning_rate": 0.0001833851935157172,
+      "loss": 2.654,
+      "step": 3804
+    },
+    {
+      "epoch": 0.3070777176983294,
+      "grad_norm": 0.780197024345398,
+      "learning_rate": 0.00018337647829571324,
+      "loss": 2.6814,
+      "step": 3805
+    },
+    {
+      "epoch": 0.30715842143491245,
+      "grad_norm": 0.7214049100875854,
+      "learning_rate": 0.0001833677609977495,
+      "loss": 2.709,
+      "step": 3806
+    },
+    {
+      "epoch": 0.30723912517149543,
+      "grad_norm": 0.7680457830429077,
+      "learning_rate": 0.00018335904162204326,
+      "loss": 2.6628,
+      "step": 3807
+    },
+    {
+      "epoch": 0.30731982890807846,
+      "grad_norm": 0.760728120803833,
+      "learning_rate": 0.00018335032016881178,
+      "loss": 2.7005,
+      "step": 3808
+    },
+    {
+      "epoch": 0.30740053264466144,
+      "grad_norm": 0.7631687521934509,
+      "learning_rate": 0.00018334159663827243,
+      "loss": 2.7012,
+      "step": 3809
+    },
+    {
+      "epoch": 0.30748123638124447,
+      "grad_norm": 0.7515785694122314,
+      "learning_rate": 0.00018333287103064266,
+      "loss": 2.7062,
+      "step": 3810
+    },
+    {
+      "epoch": 0.30756194011782745,
+      "grad_norm": 0.804500162601471,
+      "learning_rate": 0.00018332414334613987,
+      "loss": 2.7888,
+      "step": 3811
+    },
+    {
+      "epoch": 0.3076426438544105,
+      "grad_norm": 0.7551451325416565,
+      "learning_rate": 0.00018331541358498164,
+      "loss": 2.6345,
+      "step": 3812
+    },
+    {
+      "epoch": 0.30772334759099346,
+      "grad_norm": 0.7342958450317383,
+      "learning_rate": 0.0001833066817473855,
+      "loss": 2.6601,
+      "step": 3813
+    },
+    {
+      "epoch": 0.3078040513275765,
+      "grad_norm": 0.8059296607971191,
+      "learning_rate": 0.0001832979478335691,
+      "loss": 2.7694,
+      "step": 3814
+    },
+    {
+      "epoch": 0.30788475506415947,
+      "grad_norm": 0.7037352919578552,
+      "learning_rate": 0.0001832892118437501,
+      "loss": 2.6788,
+      "step": 3815
+    },
+    {
+      "epoch": 0.3079654588007425,
+      "grad_norm": 0.759509801864624,
+      "learning_rate": 0.0001832804737781462,
+      "loss": 2.7115,
+      "step": 3816
+    },
+    {
+      "epoch": 0.3080461625373255,
+      "grad_norm": 0.7911720871925354,
+      "learning_rate": 0.00018327173363697524,
+      "loss": 2.6676,
+      "step": 3817
+    },
+    {
+      "epoch": 0.3081268662739085,
+      "grad_norm": 0.7592991590499878,
+      "learning_rate": 0.00018326299142045496,
+      "loss": 2.7245,
+      "step": 3818
+    },
+    {
+      "epoch": 0.3082075700104915,
+      "grad_norm": 0.7620227932929993,
+      "learning_rate": 0.00018325424712880333,
+      "loss": 2.7224,
+      "step": 3819
+    },
+    {
+      "epoch": 0.3082882737470745,
+      "grad_norm": 0.7834638953208923,
+      "learning_rate": 0.0001832455007622382,
+      "loss": 2.7469,
+      "step": 3820
+    },
+    {
+      "epoch": 0.3083689774836575,
+      "grad_norm": 0.7765992879867554,
+      "learning_rate": 0.00018323675232097757,
+      "loss": 2.7193,
+      "step": 3821
+    },
+    {
+      "epoch": 0.3084496812202405,
+      "grad_norm": 0.7334728837013245,
+      "learning_rate": 0.00018322800180523949,
+      "loss": 2.667,
+      "step": 3822
+    },
+    {
+      "epoch": 0.3085303849568235,
+      "grad_norm": 0.7674607634544373,
+      "learning_rate": 0.00018321924921524207,
+      "loss": 2.6479,
+      "step": 3823
+    },
+    {
+      "epoch": 0.30861108869340653,
+      "grad_norm": 0.7616469860076904,
+      "learning_rate": 0.0001832104945512034,
+      "loss": 2.6535,
+      "step": 3824
+    },
+    {
+      "epoch": 0.3086917924299895,
+      "grad_norm": 0.7693164944648743,
+      "learning_rate": 0.00018320173781334172,
+      "loss": 2.7616,
+      "step": 3825
+    },
+    {
+      "epoch": 0.3087724961665725,
+      "grad_norm": 0.7099221348762512,
+      "learning_rate": 0.0001831929790018752,
+      "loss": 2.6729,
+      "step": 3826
+    },
+    {
+      "epoch": 0.3088531999031555,
+      "grad_norm": 0.7389346957206726,
+      "learning_rate": 0.00018318421811702222,
+      "loss": 2.6396,
+      "step": 3827
+    },
+    {
+      "epoch": 0.3089339036397385,
+      "grad_norm": 0.8302628397941589,
+      "learning_rate": 0.00018317545515900106,
+      "loss": 2.6786,
+      "step": 3828
+    },
+    {
+      "epoch": 0.3090146073763215,
+      "grad_norm": 0.7441998720169067,
+      "learning_rate": 0.00018316669012803015,
+      "loss": 2.6769,
+      "step": 3829
+    },
+    {
+      "epoch": 0.3090953111129045,
+      "grad_norm": 0.8454675674438477,
+      "learning_rate": 0.00018315792302432788,
+      "loss": 2.7275,
+      "step": 3830
+    },
+    {
+      "epoch": 0.30917601484948753,
+      "grad_norm": 0.8129739761352539,
+      "learning_rate": 0.00018314915384811282,
+      "loss": 2.7603,
+      "step": 3831
+    },
+    {
+      "epoch": 0.3092567185860705,
+      "grad_norm": 0.7525617480278015,
+      "learning_rate": 0.00018314038259960349,
+      "loss": 2.7156,
+      "step": 3832
+    },
+    {
+      "epoch": 0.30933742232265354,
+      "grad_norm": 0.7319022417068481,
+      "learning_rate": 0.0001831316092790185,
+      "loss": 2.676,
+      "step": 3833
+    },
+    {
+      "epoch": 0.3094181260592365,
+      "grad_norm": 0.7767768502235413,
+      "learning_rate": 0.00018312283388657646,
+      "loss": 2.7022,
+      "step": 3834
+    },
+    {
+      "epoch": 0.30949882979581955,
+      "grad_norm": 0.709293007850647,
+      "learning_rate": 0.00018311405642249616,
+      "loss": 2.6241,
+      "step": 3835
+    },
+    {
+      "epoch": 0.30957953353240253,
+      "grad_norm": 0.715360701084137,
+      "learning_rate": 0.0001831052768869963,
+      "loss": 2.6777,
+      "step": 3836
+    },
+    {
+      "epoch": 0.30966023726898556,
+      "grad_norm": 0.7361319065093994,
+      "learning_rate": 0.0001830964952802957,
+      "loss": 2.6539,
+      "step": 3837
+    },
+    {
+      "epoch": 0.30974094100556854,
+      "grad_norm": 0.7243087291717529,
+      "learning_rate": 0.0001830877116026132,
+      "loss": 2.7506,
+      "step": 3838
+    },
+    {
+      "epoch": 0.30982164474215157,
+      "grad_norm": 0.7361106872558594,
+      "learning_rate": 0.00018307892585416776,
+      "loss": 2.697,
+      "step": 3839
+    },
+    {
+      "epoch": 0.30990234847873455,
+      "grad_norm": 0.7541893720626831,
+      "learning_rate": 0.00018307013803517833,
+      "loss": 2.694,
+      "step": 3840
+    },
+    {
+      "epoch": 0.3099830522153176,
+      "grad_norm": 0.7235575914382935,
+      "learning_rate": 0.00018306134814586388,
+      "loss": 2.6711,
+      "step": 3841
+    },
+    {
+      "epoch": 0.31006375595190055,
+      "grad_norm": 0.7868196368217468,
+      "learning_rate": 0.00018305255618644354,
+      "loss": 2.7177,
+      "step": 3842
+    },
+    {
+      "epoch": 0.3101444596884836,
+      "grad_norm": 0.8074443340301514,
+      "learning_rate": 0.00018304376215713637,
+      "loss": 2.7293,
+      "step": 3843
+    },
+    {
+      "epoch": 0.31022516342506656,
+      "grad_norm": 0.6993385553359985,
+      "learning_rate": 0.00018303496605816158,
+      "loss": 2.6942,
+      "step": 3844
+    },
+    {
+      "epoch": 0.3103058671616496,
+      "grad_norm": 0.7272824645042419,
+      "learning_rate": 0.00018302616788973839,
+      "loss": 2.7093,
+      "step": 3845
+    },
+    {
+      "epoch": 0.31038657089823257,
+      "grad_norm": 0.7496963143348694,
+      "learning_rate": 0.00018301736765208605,
+      "loss": 2.7096,
+      "step": 3846
+    },
+    {
+      "epoch": 0.3104672746348156,
+      "grad_norm": 0.7407644987106323,
+      "learning_rate": 0.00018300856534542387,
+      "loss": 2.6956,
+      "step": 3847
+    },
+    {
+      "epoch": 0.3105479783713986,
+      "grad_norm": 0.742382287979126,
+      "learning_rate": 0.00018299976096997132,
+      "loss": 2.6744,
+      "step": 3848
+    },
+    {
+      "epoch": 0.3106286821079816,
+      "grad_norm": 0.7314567565917969,
+      "learning_rate": 0.0001829909545259477,
+      "loss": 2.7544,
+      "step": 3849
+    },
+    {
+      "epoch": 0.3107093858445646,
+      "grad_norm": 0.7550896406173706,
+      "learning_rate": 0.0001829821460135726,
+      "loss": 2.714,
+      "step": 3850
+    },
+    {
+      "epoch": 0.3107900895811476,
+      "grad_norm": 0.7496031522750854,
+      "learning_rate": 0.00018297333543306548,
+      "loss": 2.6718,
+      "step": 3851
+    },
+    {
+      "epoch": 0.3108707933177306,
+      "grad_norm": 0.7600073218345642,
+      "learning_rate": 0.00018296452278464596,
+      "loss": 2.7141,
+      "step": 3852
+    },
+    {
+      "epoch": 0.31095149705431363,
+      "grad_norm": 0.7242388129234314,
+      "learning_rate": 0.00018295570806853366,
+      "loss": 2.7407,
+      "step": 3853
+    },
+    {
+      "epoch": 0.3110322007908966,
+      "grad_norm": 0.723874568939209,
+      "learning_rate": 0.00018294689128494824,
+      "loss": 2.7253,
+      "step": 3854
+    },
+    {
+      "epoch": 0.31111290452747964,
+      "grad_norm": 0.7902834415435791,
+      "learning_rate": 0.00018293807243410947,
+      "loss": 2.7118,
+      "step": 3855
+    },
+    {
+      "epoch": 0.3111936082640626,
+      "grad_norm": 0.7676794528961182,
+      "learning_rate": 0.00018292925151623717,
+      "loss": 2.684,
+      "step": 3856
+    },
+    {
+      "epoch": 0.31127431200064565,
+      "grad_norm": 0.767431378364563,
+      "learning_rate": 0.0001829204285315511,
+      "loss": 2.6936,
+      "step": 3857
+    },
+    {
+      "epoch": 0.3113550157372286,
+      "grad_norm": 0.7802234888076782,
+      "learning_rate": 0.00018291160348027122,
+      "loss": 2.7181,
+      "step": 3858
+    },
+    {
+      "epoch": 0.31143571947381166,
+      "grad_norm": 0.7823610305786133,
+      "learning_rate": 0.00018290277636261743,
+      "loss": 2.7014,
+      "step": 3859
+    },
+    {
+      "epoch": 0.31151642321039463,
+      "grad_norm": 0.8199869394302368,
+      "learning_rate": 0.00018289394717880978,
+      "loss": 2.73,
+      "step": 3860
+    },
+    {
+      "epoch": 0.31159712694697766,
+      "grad_norm": 0.7725761532783508,
+      "learning_rate": 0.00018288511592906822,
+      "loss": 2.6978,
+      "step": 3861
+    },
+    {
+      "epoch": 0.31167783068356064,
+      "grad_norm": 0.752034068107605,
+      "learning_rate": 0.00018287628261361296,
+      "loss": 2.6635,
+      "step": 3862
+    },
+    {
+      "epoch": 0.3117585344201437,
+      "grad_norm": 0.7961714267730713,
+      "learning_rate": 0.0001828674472326641,
+      "loss": 2.7047,
+      "step": 3863
+    },
+    {
+      "epoch": 0.31183923815672665,
+      "grad_norm": 0.7413069605827332,
+      "learning_rate": 0.00018285860978644182,
+      "loss": 2.6872,
+      "step": 3864
+    },
+    {
+      "epoch": 0.3119199418933097,
+      "grad_norm": 0.8943146467208862,
+      "learning_rate": 0.00018284977027516636,
+      "loss": 2.7611,
+      "step": 3865
+    },
+    {
+      "epoch": 0.31200064562989266,
+      "grad_norm": 0.7663856744766235,
+      "learning_rate": 0.0001828409286990581,
+      "loss": 2.7541,
+      "step": 3866
+    },
+    {
+      "epoch": 0.3120813493664757,
+      "grad_norm": 0.7557348608970642,
+      "learning_rate": 0.00018283208505833731,
+      "loss": 2.6633,
+      "step": 3867
+    },
+    {
+      "epoch": 0.31216205310305867,
+      "grad_norm": 0.7690094113349915,
+      "learning_rate": 0.00018282323935322445,
+      "loss": 2.7117,
+      "step": 3868
+    },
+    {
+      "epoch": 0.3122427568396417,
+      "grad_norm": 0.8059033751487732,
+      "learning_rate": 0.00018281439158393997,
+      "loss": 2.6743,
+      "step": 3869
+    },
+    {
+      "epoch": 0.3123234605762247,
+      "grad_norm": 0.7877150774002075,
+      "learning_rate": 0.00018280554175070438,
+      "loss": 2.6546,
+      "step": 3870
+    },
+    {
+      "epoch": 0.3124041643128077,
+      "grad_norm": 0.799670934677124,
+      "learning_rate": 0.0001827966898537382,
+      "loss": 2.7184,
+      "step": 3871
+    },
+    {
+      "epoch": 0.3124848680493907,
+      "grad_norm": 0.8353915214538574,
+      "learning_rate": 0.0001827878358932621,
+      "loss": 2.7235,
+      "step": 3872
+    },
+    {
+      "epoch": 0.3125655717859737,
+      "grad_norm": 0.7954776883125305,
+      "learning_rate": 0.00018277897986949672,
+      "loss": 2.5992,
+      "step": 3873
+    },
+    {
+      "epoch": 0.3126462755225567,
+      "grad_norm": 0.7959856986999512,
+      "learning_rate": 0.00018277012178266277,
+      "loss": 2.6877,
+      "step": 3874
+    },
+    {
+      "epoch": 0.3127269792591397,
+      "grad_norm": 0.8220208883285522,
+      "learning_rate": 0.00018276126163298102,
+      "loss": 2.6891,
+      "step": 3875
+    },
+    {
+      "epoch": 0.3128076829957227,
+      "grad_norm": 0.7827965021133423,
+      "learning_rate": 0.0001827523994206723,
+      "loss": 2.7271,
+      "step": 3876
+    },
+    {
+      "epoch": 0.3128883867323057,
+      "grad_norm": 0.764369010925293,
+      "learning_rate": 0.00018274353514595746,
+      "loss": 2.6661,
+      "step": 3877
+    },
+    {
+      "epoch": 0.3129690904688887,
+      "grad_norm": 0.7440944314002991,
+      "learning_rate": 0.00018273466880905744,
+      "loss": 2.6621,
+      "step": 3878
+    },
+    {
+      "epoch": 0.3130497942054717,
+      "grad_norm": 0.8544813394546509,
+      "learning_rate": 0.00018272580041019319,
+      "loss": 2.7168,
+      "step": 3879
+    },
+    {
+      "epoch": 0.3131304979420547,
+      "grad_norm": 0.7232592701911926,
+      "learning_rate": 0.00018271692994958577,
+      "loss": 2.6666,
+      "step": 3880
+    },
+    {
+      "epoch": 0.3132112016786377,
+      "grad_norm": 0.750525712966919,
+      "learning_rate": 0.00018270805742745617,
+      "loss": 2.6984,
+      "step": 3881
+    },
+    {
+      "epoch": 0.31329190541522073,
+      "grad_norm": 0.8195550441741943,
+      "learning_rate": 0.00018269918284402565,
+      "loss": 2.7183,
+      "step": 3882
+    },
+    {
+      "epoch": 0.3133726091518037,
+      "grad_norm": 0.7695632576942444,
+      "learning_rate": 0.0001826903061995153,
+      "loss": 2.7092,
+      "step": 3883
+    },
+    {
+      "epoch": 0.31345331288838674,
+      "grad_norm": 0.7631582617759705,
+      "learning_rate": 0.0001826814274941463,
+      "loss": 2.7061,
+      "step": 3884
+    },
+    {
+      "epoch": 0.3135340166249697,
+      "grad_norm": 0.8318471908569336,
+      "learning_rate": 0.0001826725467281401,
+      "loss": 2.694,
+      "step": 3885
+    },
+    {
+      "epoch": 0.31361472036155275,
+      "grad_norm": 0.7313492298126221,
+      "learning_rate": 0.00018266366390171784,
+      "loss": 2.6729,
+      "step": 3886
+    },
+    {
+      "epoch": 0.3136954240981357,
+      "grad_norm": 0.7508631944656372,
+      "learning_rate": 0.00018265477901510105,
+      "loss": 2.731,
+      "step": 3887
+    },
+    {
+      "epoch": 0.31377612783471875,
+      "grad_norm": 0.8106402158737183,
+      "learning_rate": 0.00018264589206851107,
+      "loss": 2.7113,
+      "step": 3888
+    },
+    {
+      "epoch": 0.31385683157130173,
+      "grad_norm": 0.771542489528656,
+      "learning_rate": 0.00018263700306216945,
+      "loss": 2.644,
+      "step": 3889
+    },
+    {
+      "epoch": 0.31393753530788476,
+      "grad_norm": 0.812441885471344,
+      "learning_rate": 0.00018262811199629768,
+      "loss": 2.6889,
+      "step": 3890
+    },
+    {
+      "epoch": 0.31401823904446774,
+      "grad_norm": 0.8231199979782104,
+      "learning_rate": 0.00018261921887111738,
+      "loss": 2.6466,
+      "step": 3891
+    },
+    {
+      "epoch": 0.31409894278105077,
+      "grad_norm": 0.7492454051971436,
+      "learning_rate": 0.00018261032368685012,
+      "loss": 2.6693,
+      "step": 3892
+    },
+    {
+      "epoch": 0.31417964651763375,
+      "grad_norm": 0.7651814222335815,
+      "learning_rate": 0.00018260142644371772,
+      "loss": 2.6569,
+      "step": 3893
+    },
+    {
+      "epoch": 0.3142603502542168,
+      "grad_norm": 0.7504465579986572,
+      "learning_rate": 0.0001825925271419418,
+      "loss": 2.684,
+      "step": 3894
+    },
+    {
+      "epoch": 0.31434105399079976,
+      "grad_norm": 0.749650239944458,
+      "learning_rate": 0.00018258362578174424,
+      "loss": 2.6482,
+      "step": 3895
+    },
+    {
+      "epoch": 0.3144217577273828,
+      "grad_norm": 0.8445256352424622,
+      "learning_rate": 0.00018257472236334686,
+      "loss": 2.727,
+      "step": 3896
+    },
+    {
+      "epoch": 0.31450246146396577,
+      "grad_norm": 0.7628257870674133,
+      "learning_rate": 0.0001825658168869715,
+      "loss": 2.7314,
+      "step": 3897
+    },
+    {
+      "epoch": 0.3145831652005488,
+      "grad_norm": 0.7738446593284607,
+      "learning_rate": 0.00018255690935284019,
+      "loss": 2.7478,
+      "step": 3898
+    },
+    {
+      "epoch": 0.3146638689371318,
+      "grad_norm": 0.7578958868980408,
+      "learning_rate": 0.00018254799976117486,
+      "loss": 2.6922,
+      "step": 3899
+    },
+    {
+      "epoch": 0.3147445726737148,
+      "grad_norm": 0.8367362022399902,
+      "learning_rate": 0.00018253908811219764,
+      "loss": 2.7347,
+      "step": 3900
+    },
+    {
+      "epoch": 0.3148252764102978,
+      "grad_norm": 0.7530354857444763,
+      "learning_rate": 0.00018253017440613057,
+      "loss": 2.7151,
+      "step": 3901
+    },
+    {
+      "epoch": 0.3149059801468808,
+      "grad_norm": 0.7168053388595581,
+      "learning_rate": 0.00018252125864319578,
+      "loss": 2.7072,
+      "step": 3902
+    },
+    {
+      "epoch": 0.3149866838834638,
+      "grad_norm": 0.7480056285858154,
+      "learning_rate": 0.00018251234082361555,
+      "loss": 2.6489,
+      "step": 3903
+    },
+    {
+      "epoch": 0.3150673876200468,
+      "grad_norm": 0.8563880324363708,
+      "learning_rate": 0.0001825034209476121,
+      "loss": 2.7384,
+      "step": 3904
+    },
+    {
+      "epoch": 0.3151480913566298,
+      "grad_norm": 0.7959346771240234,
+      "learning_rate": 0.0001824944990154077,
+      "loss": 2.631,
+      "step": 3905
+    },
+    {
+      "epoch": 0.31522879509321283,
+      "grad_norm": 0.7385980486869812,
+      "learning_rate": 0.00018248557502722476,
+      "loss": 2.7394,
+      "step": 3906
+    },
+    {
+      "epoch": 0.3153094988297958,
+      "grad_norm": 0.7682650685310364,
+      "learning_rate": 0.00018247664898328567,
+      "loss": 2.7327,
+      "step": 3907
+    },
+    {
+      "epoch": 0.31539020256637884,
+      "grad_norm": 0.7720316648483276,
+      "learning_rate": 0.0001824677208838129,
+      "loss": 2.6442,
+      "step": 3908
+    },
+    {
+      "epoch": 0.3154709063029618,
+      "grad_norm": 0.7927379608154297,
+      "learning_rate": 0.00018245879072902895,
+      "loss": 2.7738,
+      "step": 3909
+    },
+    {
+      "epoch": 0.31555161003954485,
+      "grad_norm": 0.7506012916564941,
+      "learning_rate": 0.00018244985851915637,
+      "loss": 2.6825,
+      "step": 3910
+    },
+    {
+      "epoch": 0.3156323137761278,
+      "grad_norm": 0.6996353268623352,
+      "learning_rate": 0.00018244092425441781,
+      "loss": 2.6783,
+      "step": 3911
+    },
+    {
+      "epoch": 0.31571301751271086,
+      "grad_norm": 0.8039344549179077,
+      "learning_rate": 0.00018243198793503588,
+      "loss": 2.7628,
+      "step": 3912
+    },
+    {
+      "epoch": 0.31579372124929384,
+      "grad_norm": 0.7890963554382324,
+      "learning_rate": 0.0001824230495612334,
+      "loss": 2.7512,
+      "step": 3913
+    },
+    {
+      "epoch": 0.31587442498587687,
+      "grad_norm": 0.7470870614051819,
+      "learning_rate": 0.00018241410913323301,
+      "loss": 2.7058,
+      "step": 3914
+    },
+    {
+      "epoch": 0.31595512872245984,
+      "grad_norm": 0.7056336402893066,
+      "learning_rate": 0.0001824051666512576,
+      "loss": 2.6091,
+      "step": 3915
+    },
+    {
+      "epoch": 0.3160358324590429,
+      "grad_norm": 0.7818490862846375,
+      "learning_rate": 0.00018239622211553002,
+      "loss": 2.7509,
+      "step": 3916
+    },
+    {
+      "epoch": 0.31611653619562585,
+      "grad_norm": 0.7590607404708862,
+      "learning_rate": 0.0001823872755262732,
+      "loss": 2.7238,
+      "step": 3917
+    },
+    {
+      "epoch": 0.3161972399322089,
+      "grad_norm": 0.7157841920852661,
+      "learning_rate": 0.00018237832688371014,
+      "loss": 2.6639,
+      "step": 3918
+    },
+    {
+      "epoch": 0.31627794366879186,
+      "grad_norm": 0.7515804171562195,
+      "learning_rate": 0.00018236937618806382,
+      "loss": 2.6973,
+      "step": 3919
+    },
+    {
+      "epoch": 0.3163586474053749,
+      "grad_norm": 0.6691949963569641,
+      "learning_rate": 0.00018236042343955733,
+      "loss": 2.727,
+      "step": 3920
+    },
+    {
+      "epoch": 0.31643935114195787,
+      "grad_norm": 0.8122327327728271,
+      "learning_rate": 0.0001823514686384138,
+      "loss": 2.7513,
+      "step": 3921
+    },
+    {
+      "epoch": 0.3165200548785409,
+      "grad_norm": 0.7813653349876404,
+      "learning_rate": 0.0001823425117848564,
+      "loss": 2.7037,
+      "step": 3922
+    },
+    {
+      "epoch": 0.3166007586151239,
+      "grad_norm": 0.6869354844093323,
+      "learning_rate": 0.00018233355287910834,
+      "loss": 2.693,
+      "step": 3923
+    },
+    {
+      "epoch": 0.3166814623517069,
+      "grad_norm": 0.7773037552833557,
+      "learning_rate": 0.00018232459192139296,
+      "loss": 2.687,
+      "step": 3924
+    },
+    {
+      "epoch": 0.3167621660882899,
+      "grad_norm": 0.7644256949424744,
+      "learning_rate": 0.00018231562891193352,
+      "loss": 2.6753,
+      "step": 3925
+    },
+    {
+      "epoch": 0.3168428698248729,
+      "grad_norm": 0.8427005410194397,
+      "learning_rate": 0.00018230666385095343,
+      "loss": 2.6641,
+      "step": 3926
+    },
+    {
+      "epoch": 0.3169235735614559,
+      "grad_norm": 0.7194599509239197,
+      "learning_rate": 0.0001822976967386761,
+      "loss": 2.7091,
+      "step": 3927
+    },
+    {
+      "epoch": 0.3170042772980389,
+      "grad_norm": 0.7710655331611633,
+      "learning_rate": 0.00018228872757532512,
+      "loss": 2.6938,
+      "step": 3928
+    },
+    {
+      "epoch": 0.3170849810346219,
+      "grad_norm": 0.8003759980201721,
+      "learning_rate": 0.0001822797563611239,
+      "loss": 2.7019,
+      "step": 3929
+    },
+    {
+      "epoch": 0.3171656847712049,
+      "grad_norm": 0.7960470914840698,
+      "learning_rate": 0.00018227078309629606,
+      "loss": 2.661,
+      "step": 3930
+    },
+    {
+      "epoch": 0.3172463885077879,
+      "grad_norm": 0.7731126546859741,
+      "learning_rate": 0.00018226180778106526,
+      "loss": 2.7023,
+      "step": 3931
+    },
+    {
+      "epoch": 0.3173270922443709,
+      "grad_norm": 0.7561383843421936,
+      "learning_rate": 0.00018225283041565515,
+      "loss": 2.6768,
+      "step": 3932
+    },
+    {
+      "epoch": 0.3174077959809539,
+      "grad_norm": 0.7578409910202026,
+      "learning_rate": 0.0001822438510002895,
+      "loss": 2.7145,
+      "step": 3933
+    },
+    {
+      "epoch": 0.3174884997175369,
+      "grad_norm": 0.7901952862739563,
+      "learning_rate": 0.00018223486953519214,
+      "loss": 2.7121,
+      "step": 3934
+    },
+    {
+      "epoch": 0.31756920345411993,
+      "grad_norm": 0.82305908203125,
+      "learning_rate": 0.0001822258860205868,
+      "loss": 2.7553,
+      "step": 3935
+    },
+    {
+      "epoch": 0.3176499071907029,
+      "grad_norm": 0.748055636882782,
+      "learning_rate": 0.0001822169004566975,
+      "loss": 2.7236,
+      "step": 3936
+    },
+    {
+      "epoch": 0.31773061092728594,
+      "grad_norm": 0.7981358766555786,
+      "learning_rate": 0.0001822079128437481,
+      "loss": 2.7444,
+      "step": 3937
+    },
+    {
+      "epoch": 0.3178113146638689,
+      "grad_norm": 0.7938945889472961,
+      "learning_rate": 0.0001821989231819626,
+      "loss": 2.7512,
+      "step": 3938
+    },
+    {
+      "epoch": 0.31789201840045195,
+      "grad_norm": 0.7250397205352783,
+      "learning_rate": 0.0001821899314715651,
+      "loss": 2.6843,
+      "step": 3939
+    },
+    {
+      "epoch": 0.3179727221370349,
+      "grad_norm": 0.8844723701477051,
+      "learning_rate": 0.00018218093771277965,
+      "loss": 2.6295,
+      "step": 3940
+    },
+    {
+      "epoch": 0.31805342587361796,
+      "grad_norm": 0.7545698881149292,
+      "learning_rate": 0.0001821719419058304,
+      "loss": 2.7478,
+      "step": 3941
+    },
+    {
+      "epoch": 0.31813412961020093,
+      "grad_norm": 0.7254738807678223,
+      "learning_rate": 0.00018216294405094157,
+      "loss": 2.665,
+      "step": 3942
+    },
+    {
+      "epoch": 0.31821483334678397,
+      "grad_norm": 0.7664754390716553,
+      "learning_rate": 0.00018215394414833737,
+      "loss": 2.7431,
+      "step": 3943
+    },
+    {
+      "epoch": 0.31829553708336694,
+      "grad_norm": 0.8250303864479065,
+      "learning_rate": 0.00018214494219824217,
+      "loss": 2.6957,
+      "step": 3944
+    },
+    {
+      "epoch": 0.31837624081995,
+      "grad_norm": 0.7425532341003418,
+      "learning_rate": 0.00018213593820088026,
+      "loss": 2.666,
+      "step": 3945
+    },
+    {
+      "epoch": 0.31845694455653295,
+      "grad_norm": 0.6943121552467346,
+      "learning_rate": 0.00018212693215647604,
+      "loss": 2.716,
+      "step": 3946
+    },
+    {
+      "epoch": 0.318537648293116,
+      "grad_norm": 0.732829213142395,
+      "learning_rate": 0.00018211792406525403,
+      "loss": 2.6557,
+      "step": 3947
+    },
+    {
+      "epoch": 0.31861835202969896,
+      "grad_norm": 0.7666537165641785,
+      "learning_rate": 0.00018210891392743866,
+      "loss": 2.7275,
+      "step": 3948
+    },
+    {
+      "epoch": 0.318699055766282,
+      "grad_norm": 0.7652621865272522,
+      "learning_rate": 0.00018209990174325455,
+      "loss": 2.6372,
+      "step": 3949
+    },
+    {
+      "epoch": 0.31877975950286497,
+      "grad_norm": 0.7416055202484131,
+      "learning_rate": 0.00018209088751292626,
+      "loss": 2.6688,
+      "step": 3950
+    },
+    {
+      "epoch": 0.318860463239448,
+      "grad_norm": 0.7504609227180481,
+      "learning_rate": 0.00018208187123667848,
+      "loss": 2.6912,
+      "step": 3951
+    },
+    {
+      "epoch": 0.318941166976031,
+      "grad_norm": 0.7308809757232666,
+      "learning_rate": 0.00018207285291473588,
+      "loss": 2.7272,
+      "step": 3952
+    },
+    {
+      "epoch": 0.319021870712614,
+      "grad_norm": 0.8031618595123291,
+      "learning_rate": 0.00018206383254732326,
+      "loss": 2.7354,
+      "step": 3953
+    },
+    {
+      "epoch": 0.319102574449197,
+      "grad_norm": 0.81386798620224,
+      "learning_rate": 0.00018205481013466542,
+      "loss": 2.676,
+      "step": 3954
+    },
+    {
+      "epoch": 0.31918327818578,
+      "grad_norm": 0.7845911383628845,
+      "learning_rate": 0.0001820457856769872,
+      "loss": 2.7094,
+      "step": 3955
+    },
+    {
+      "epoch": 0.319263981922363,
+      "grad_norm": 0.7189298272132874,
+      "learning_rate": 0.00018203675917451357,
+      "loss": 2.6764,
+      "step": 3956
+    },
+    {
+      "epoch": 0.319344685658946,
+      "grad_norm": 0.8253228664398193,
+      "learning_rate": 0.00018202773062746944,
+      "loss": 2.6805,
+      "step": 3957
+    },
+    {
+      "epoch": 0.319425389395529,
+      "grad_norm": 0.7965289950370789,
+      "learning_rate": 0.0001820187000360798,
+      "loss": 2.7148,
+      "step": 3958
+    },
+    {
+      "epoch": 0.31950609313211203,
+      "grad_norm": 0.7505398988723755,
+      "learning_rate": 0.0001820096674005698,
+      "loss": 2.6732,
+      "step": 3959
+    },
+    {
+      "epoch": 0.319586796868695,
+      "grad_norm": 0.7554877400398254,
+      "learning_rate": 0.0001820006327211645,
+      "loss": 2.7467,
+      "step": 3960
+    },
+    {
+      "epoch": 0.31966750060527804,
+      "grad_norm": 0.7836194038391113,
+      "learning_rate": 0.00018199159599808907,
+      "loss": 2.7252,
+      "step": 3961
+    },
+    {
+      "epoch": 0.319748204341861,
+      "grad_norm": 0.7967261672019958,
+      "learning_rate": 0.00018198255723156877,
+      "loss": 2.6814,
+      "step": 3962
+    },
+    {
+      "epoch": 0.31982890807844405,
+      "grad_norm": 0.7411713600158691,
+      "learning_rate": 0.00018197351642182882,
+      "loss": 2.6928,
+      "step": 3963
+    },
+    {
+      "epoch": 0.31990961181502703,
+      "grad_norm": 0.6961422562599182,
+      "learning_rate": 0.00018196447356909454,
+      "loss": 2.6651,
+      "step": 3964
+    },
+    {
+      "epoch": 0.31999031555161006,
+      "grad_norm": 0.7245771884918213,
+      "learning_rate": 0.00018195542867359134,
+      "loss": 2.6726,
+      "step": 3965
+    },
+    {
+      "epoch": 0.32007101928819304,
+      "grad_norm": 0.784654974937439,
+      "learning_rate": 0.00018194638173554462,
+      "loss": 2.6829,
+      "step": 3966
+    },
+    {
+      "epoch": 0.32015172302477607,
+      "grad_norm": 0.7373329997062683,
+      "learning_rate": 0.00018193733275517985,
+      "loss": 2.6481,
+      "step": 3967
+    },
+    {
+      "epoch": 0.32023242676135905,
+      "grad_norm": 0.7878682613372803,
+      "learning_rate": 0.00018192828173272258,
+      "loss": 2.6701,
+      "step": 3968
+    },
+    {
+      "epoch": 0.3203131304979421,
+      "grad_norm": 0.759676992893219,
+      "learning_rate": 0.00018191922866839835,
+      "loss": 2.7218,
+      "step": 3969
+    },
+    {
+      "epoch": 0.32039383423452505,
+      "grad_norm": 0.7923088669776917,
+      "learning_rate": 0.00018191017356243282,
+      "loss": 2.6841,
+      "step": 3970
+    },
+    {
+      "epoch": 0.3204745379711081,
+      "grad_norm": 0.7084882855415344,
+      "learning_rate": 0.00018190111641505164,
+      "loss": 2.7167,
+      "step": 3971
+    },
+    {
+      "epoch": 0.32055524170769106,
+      "grad_norm": 0.7166235446929932,
+      "learning_rate": 0.00018189205722648054,
+      "loss": 2.6647,
+      "step": 3972
+    },
+    {
+      "epoch": 0.3206359454442741,
+      "grad_norm": 0.7997722029685974,
+      "learning_rate": 0.0001818829959969453,
+      "loss": 2.7199,
+      "step": 3973
+    },
+    {
+      "epoch": 0.32071664918085707,
+      "grad_norm": 0.8309516310691833,
+      "learning_rate": 0.0001818739327266718,
+      "loss": 2.8006,
+      "step": 3974
+    },
+    {
+      "epoch": 0.3207973529174401,
+      "grad_norm": 0.7164002656936646,
+      "learning_rate": 0.00018186486741588582,
+      "loss": 2.6258,
+      "step": 3975
+    },
+    {
+      "epoch": 0.3208780566540231,
+      "grad_norm": 0.7715865969657898,
+      "learning_rate": 0.0001818558000648134,
+      "loss": 2.7034,
+      "step": 3976
+    },
+    {
+      "epoch": 0.3209587603906061,
+      "grad_norm": 0.7806593775749207,
+      "learning_rate": 0.0001818467306736804,
+      "loss": 2.6758,
+      "step": 3977
+    },
+    {
+      "epoch": 0.3210394641271891,
+      "grad_norm": 0.8026594519615173,
+      "learning_rate": 0.00018183765924271298,
+      "loss": 2.6976,
+      "step": 3978
+    },
+    {
+      "epoch": 0.32112016786377207,
+      "grad_norm": 0.7971245050430298,
+      "learning_rate": 0.00018182858577213716,
+      "loss": 2.7312,
+      "step": 3979
+    },
+    {
+      "epoch": 0.3212008716003551,
+      "grad_norm": 0.7347297072410583,
+      "learning_rate": 0.00018181951026217908,
+      "loss": 2.6664,
+      "step": 3980
+    },
+    {
+      "epoch": 0.3212815753369381,
+      "grad_norm": 0.7929779291152954,
+      "learning_rate": 0.0001818104327130649,
+      "loss": 2.6603,
+      "step": 3981
+    },
+    {
+      "epoch": 0.3213622790735211,
+      "grad_norm": 0.7465224862098694,
+      "learning_rate": 0.00018180135312502089,
+      "loss": 2.6566,
+      "step": 3982
+    },
+    {
+      "epoch": 0.3214429828101041,
+      "grad_norm": 0.7114695906639099,
+      "learning_rate": 0.00018179227149827334,
+      "loss": 2.6492,
+      "step": 3983
+    },
+    {
+      "epoch": 0.3215236865466871,
+      "grad_norm": 0.7179337739944458,
+      "learning_rate": 0.00018178318783304857,
+      "loss": 2.6778,
+      "step": 3984
+    },
+    {
+      "epoch": 0.3216043902832701,
+      "grad_norm": 0.7182629704475403,
+      "learning_rate": 0.000181774102129573,
+      "loss": 2.7057,
+      "step": 3985
+    },
+    {
+      "epoch": 0.3216850940198531,
+      "grad_norm": 0.7383119463920593,
+      "learning_rate": 0.000181765014388073,
+      "loss": 2.6633,
+      "step": 3986
+    },
+    {
+      "epoch": 0.3217657977564361,
+      "grad_norm": 0.7340527176856995,
+      "learning_rate": 0.00018175592460877512,
+      "loss": 2.6838,
+      "step": 3987
+    },
+    {
+      "epoch": 0.32184650149301913,
+      "grad_norm": 0.7934359312057495,
+      "learning_rate": 0.00018174683279190593,
+      "loss": 2.6795,
+      "step": 3988
+    },
+    {
+      "epoch": 0.3219272052296021,
+      "grad_norm": 0.6960840821266174,
+      "learning_rate": 0.00018173773893769192,
+      "loss": 2.6669,
+      "step": 3989
+    },
+    {
+      "epoch": 0.32200790896618514,
+      "grad_norm": 0.7513574361801147,
+      "learning_rate": 0.00018172864304635985,
+      "loss": 2.6744,
+      "step": 3990
+    },
+    {
+      "epoch": 0.3220886127027681,
+      "grad_norm": 0.7516636848449707,
+      "learning_rate": 0.00018171954511813629,
+      "loss": 2.6652,
+      "step": 3991
+    },
+    {
+      "epoch": 0.32216931643935115,
+      "grad_norm": 0.7817716002464294,
+      "learning_rate": 0.00018171044515324808,
+      "loss": 2.6671,
+      "step": 3992
+    },
+    {
+      "epoch": 0.3222500201759341,
+      "grad_norm": 0.6859925389289856,
+      "learning_rate": 0.000181701343151922,
+      "loss": 2.6984,
+      "step": 3993
+    },
+    {
+      "epoch": 0.32233072391251716,
+      "grad_norm": 0.7669627666473389,
+      "learning_rate": 0.00018169223911438485,
+      "loss": 2.7102,
+      "step": 3994
+    },
+    {
+      "epoch": 0.32241142764910014,
+      "grad_norm": 0.784724235534668,
+      "learning_rate": 0.00018168313304086357,
+      "loss": 2.7413,
+      "step": 3995
+    },
+    {
+      "epoch": 0.32249213138568317,
+      "grad_norm": 0.7341497540473938,
+      "learning_rate": 0.00018167402493158509,
+      "loss": 2.706,
+      "step": 3996
+    },
+    {
+      "epoch": 0.32257283512226614,
+      "grad_norm": 0.7975730299949646,
+      "learning_rate": 0.00018166491478677641,
+      "loss": 2.6896,
+      "step": 3997
+    },
+    {
+      "epoch": 0.3226535388588492,
+      "grad_norm": 0.8138537406921387,
+      "learning_rate": 0.00018165580260666458,
+      "loss": 2.6986,
+      "step": 3998
+    },
+    {
+      "epoch": 0.32273424259543215,
+      "grad_norm": 0.6734997034072876,
+      "learning_rate": 0.0001816466883914767,
+      "loss": 2.6686,
+      "step": 3999
+    },
+    {
+      "epoch": 0.3228149463320152,
+      "grad_norm": 0.7742779850959778,
+      "learning_rate": 0.00018163757214143992,
+      "loss": 2.7222,
+      "step": 4000
+    },
+    {
+      "epoch": 0.3228149463320152,
+      "eval_loss": 2.615234375,
+      "eval_runtime": 783.0394,
+      "eval_samples_per_second": 3.346,
+      "eval_steps_per_second": 0.558,
+      "step": 4000
+    },
+    {
+      "epoch": 0.32289565006859816,
+      "grad_norm": 0.7654715180397034,
+      "learning_rate": 0.00018162845385678145,
+      "loss": 2.7016,
+      "step": 4001
+    },
+    {
+      "epoch": 0.3229763538051812,
+      "grad_norm": 0.8698763251304626,
+      "learning_rate": 0.0001816193335377285,
+      "loss": 2.6709,
+      "step": 4002
+    },
+    {
+      "epoch": 0.32305705754176417,
+      "grad_norm": 0.758056640625,
+      "learning_rate": 0.00018161021118450843,
+      "loss": 2.7277,
+      "step": 4003
+    },
+    {
+      "epoch": 0.3231377612783472,
+      "grad_norm": 0.7462654113769531,
+      "learning_rate": 0.00018160108679734856,
+      "loss": 2.623,
+      "step": 4004
+    },
+    {
+      "epoch": 0.3232184650149302,
+      "grad_norm": 0.7274953722953796,
+      "learning_rate": 0.00018159196037647628,
+      "loss": 2.6875,
+      "step": 4005
+    },
+    {
+      "epoch": 0.3232991687515132,
+      "grad_norm": 0.7737346887588501,
+      "learning_rate": 0.0001815828319221191,
+      "loss": 2.6967,
+      "step": 4006
+    },
+    {
+      "epoch": 0.3233798724880962,
+      "grad_norm": 0.7793172001838684,
+      "learning_rate": 0.00018157370143450448,
+      "loss": 2.724,
+      "step": 4007
+    },
+    {
+      "epoch": 0.3234605762246792,
+      "grad_norm": 0.7791805863380432,
+      "learning_rate": 0.00018156456891385995,
+      "loss": 2.6653,
+      "step": 4008
+    },
+    {
+      "epoch": 0.3235412799612622,
+      "grad_norm": 0.7225624918937683,
+      "learning_rate": 0.0001815554343604132,
+      "loss": 2.745,
+      "step": 4009
+    },
+    {
+      "epoch": 0.32362198369784523,
+      "grad_norm": 0.6958494782447815,
+      "learning_rate": 0.0001815462977743918,
+      "loss": 2.6856,
+      "step": 4010
+    },
+    {
+      "epoch": 0.3237026874344282,
+      "grad_norm": 0.7572030425071716,
+      "learning_rate": 0.0001815371591560235,
+      "loss": 2.7053,
+      "step": 4011
+    },
+    {
+      "epoch": 0.32378339117101124,
+      "grad_norm": 0.7133952975273132,
+      "learning_rate": 0.00018152801850553605,
+      "loss": 2.6984,
+      "step": 4012
+    },
+    {
+      "epoch": 0.3238640949075942,
+      "grad_norm": 0.7598705291748047,
+      "learning_rate": 0.00018151887582315728,
+      "loss": 2.6632,
+      "step": 4013
+    },
+    {
+      "epoch": 0.32394479864417725,
+      "grad_norm": 0.7670698165893555,
+      "learning_rate": 0.00018150973110911503,
+      "loss": 2.7035,
+      "step": 4014
+    },
+    {
+      "epoch": 0.3240255023807602,
+      "grad_norm": 0.7547060251235962,
+      "learning_rate": 0.00018150058436363723,
+      "loss": 2.6531,
+      "step": 4015
+    },
+    {
+      "epoch": 0.32410620611734325,
+      "grad_norm": 0.7943035364151001,
+      "learning_rate": 0.00018149143558695178,
+      "loss": 2.766,
+      "step": 4016
+    },
+    {
+      "epoch": 0.32418690985392623,
+      "grad_norm": 0.864356517791748,
+      "learning_rate": 0.00018148228477928675,
+      "loss": 2.7134,
+      "step": 4017
+    },
+    {
+      "epoch": 0.32426761359050926,
+      "grad_norm": 0.7773902416229248,
+      "learning_rate": 0.00018147313194087018,
+      "loss": 2.6948,
+      "step": 4018
+    },
+    {
+      "epoch": 0.32434831732709224,
+      "grad_norm": 0.839131772518158,
+      "learning_rate": 0.0001814639770719302,
+      "loss": 2.7393,
+      "step": 4019
+    },
+    {
+      "epoch": 0.32442902106367527,
+      "grad_norm": 0.807837963104248,
+      "learning_rate": 0.00018145482017269498,
+      "loss": 2.7835,
+      "step": 4020
+    },
+    {
+      "epoch": 0.32450972480025825,
+      "grad_norm": 0.7133228182792664,
+      "learning_rate": 0.00018144566124339272,
+      "loss": 2.6859,
+      "step": 4021
+    },
+    {
+      "epoch": 0.3245904285368413,
+      "grad_norm": 0.8450621962547302,
+      "learning_rate": 0.00018143650028425162,
+      "loss": 2.7548,
+      "step": 4022
+    },
+    {
+      "epoch": 0.32467113227342426,
+      "grad_norm": 0.8594980835914612,
+      "learning_rate": 0.00018142733729550013,
+      "loss": 2.6636,
+      "step": 4023
+    },
+    {
+      "epoch": 0.3247518360100073,
+      "grad_norm": 0.7134621739387512,
+      "learning_rate": 0.0001814181722773665,
+      "loss": 2.6501,
+      "step": 4024
+    },
+    {
+      "epoch": 0.32483253974659027,
+      "grad_norm": 0.8630430698394775,
+      "learning_rate": 0.0001814090052300792,
+      "loss": 2.6994,
+      "step": 4025
+    },
+    {
+      "epoch": 0.3249132434831733,
+      "grad_norm": 0.7044873237609863,
+      "learning_rate": 0.00018139983615386666,
+      "loss": 2.6603,
+      "step": 4026
+    },
+    {
+      "epoch": 0.3249939472197563,
+      "grad_norm": 0.6896052360534668,
+      "learning_rate": 0.00018139066504895744,
+      "loss": 2.6649,
+      "step": 4027
+    },
+    {
+      "epoch": 0.3250746509563393,
+      "grad_norm": 0.802855372428894,
+      "learning_rate": 0.00018138149191558012,
+      "loss": 2.7067,
+      "step": 4028
+    },
+    {
+      "epoch": 0.3251553546929223,
+      "grad_norm": 0.7555437088012695,
+      "learning_rate": 0.00018137231675396324,
+      "loss": 2.6471,
+      "step": 4029
+    },
+    {
+      "epoch": 0.32523605842950526,
+      "grad_norm": 0.6846967339515686,
+      "learning_rate": 0.00018136313956433552,
+      "loss": 2.6774,
+      "step": 4030
+    },
+    {
+      "epoch": 0.3253167621660883,
+      "grad_norm": 0.7435858249664307,
+      "learning_rate": 0.0001813539603469257,
+      "loss": 2.7135,
+      "step": 4031
+    },
+    {
+      "epoch": 0.32539746590267127,
+      "grad_norm": 0.7669098377227783,
+      "learning_rate": 0.00018134477910196253,
+      "loss": 2.7014,
+      "step": 4032
+    },
+    {
+      "epoch": 0.3254781696392543,
+      "grad_norm": 0.7797521352767944,
+      "learning_rate": 0.00018133559582967482,
+      "loss": 2.7229,
+      "step": 4033
+    },
+    {
+      "epoch": 0.3255588733758373,
+      "grad_norm": 0.7377886176109314,
+      "learning_rate": 0.00018132641053029142,
+      "loss": 2.7196,
+      "step": 4034
+    },
+    {
+      "epoch": 0.3256395771124203,
+      "grad_norm": 0.7387986779212952,
+      "learning_rate": 0.0001813172232040413,
+      "loss": 2.687,
+      "step": 4035
+    },
+    {
+      "epoch": 0.3257202808490033,
+      "grad_norm": 0.7276624441146851,
+      "learning_rate": 0.0001813080338511534,
+      "loss": 2.6954,
+      "step": 4036
+    },
+    {
+      "epoch": 0.3258009845855863,
+      "grad_norm": 0.7929670214653015,
+      "learning_rate": 0.00018129884247185683,
+      "loss": 2.7431,
+      "step": 4037
+    },
+    {
+      "epoch": 0.3258816883221693,
+      "grad_norm": 0.7896441221237183,
+      "learning_rate": 0.0001812896490663805,
+      "loss": 2.6823,
+      "step": 4038
+    },
+    {
+      "epoch": 0.3259623920587523,
+      "grad_norm": 0.8642957210540771,
+      "learning_rate": 0.00018128045363495368,
+      "loss": 2.7334,
+      "step": 4039
+    },
+    {
+      "epoch": 0.3260430957953353,
+      "grad_norm": 0.7156081795692444,
+      "learning_rate": 0.00018127125617780542,
+      "loss": 2.6886,
+      "step": 4040
+    },
+    {
+      "epoch": 0.32612379953191833,
+      "grad_norm": 0.8260853290557861,
+      "learning_rate": 0.00018126205669516507,
+      "loss": 2.6802,
+      "step": 4041
+    },
+    {
+      "epoch": 0.3262045032685013,
+      "grad_norm": 0.6853542327880859,
+      "learning_rate": 0.00018125285518726182,
+      "loss": 2.6392,
+      "step": 4042
+    },
+    {
+      "epoch": 0.32628520700508434,
+      "grad_norm": 0.7574017643928528,
+      "learning_rate": 0.00018124365165432505,
+      "loss": 2.7412,
+      "step": 4043
+    },
+    {
+      "epoch": 0.3263659107416673,
+      "grad_norm": 0.8656191825866699,
+      "learning_rate": 0.00018123444609658408,
+      "loss": 2.6903,
+      "step": 4044
+    },
+    {
+      "epoch": 0.32644661447825035,
+      "grad_norm": 0.7443257570266724,
+      "learning_rate": 0.00018122523851426837,
+      "loss": 2.682,
+      "step": 4045
+    },
+    {
+      "epoch": 0.32652731821483333,
+      "grad_norm": 0.7222229242324829,
+      "learning_rate": 0.0001812160289076074,
+      "loss": 2.6196,
+      "step": 4046
+    },
+    {
+      "epoch": 0.32660802195141636,
+      "grad_norm": 0.8531985878944397,
+      "learning_rate": 0.00018120681727683066,
+      "loss": 2.6777,
+      "step": 4047
+    },
+    {
+      "epoch": 0.32668872568799934,
+      "grad_norm": 0.7380290627479553,
+      "learning_rate": 0.0001811976036221678,
+      "loss": 2.6847,
+      "step": 4048
+    },
+    {
+      "epoch": 0.32676942942458237,
+      "grad_norm": 0.7250707149505615,
+      "learning_rate": 0.00018118838794384837,
+      "loss": 2.6846,
+      "step": 4049
+    },
+    {
+      "epoch": 0.32685013316116535,
+      "grad_norm": 0.763504147529602,
+      "learning_rate": 0.00018117917024210208,
+      "loss": 2.69,
+      "step": 4050
+    },
+    {
+      "epoch": 0.3269308368977484,
+      "grad_norm": 0.7740737795829773,
+      "learning_rate": 0.00018116995051715867,
+      "loss": 2.6945,
+      "step": 4051
+    },
+    {
+      "epoch": 0.32701154063433135,
+      "grad_norm": 0.7777624726295471,
+      "learning_rate": 0.00018116072876924792,
+      "loss": 2.6918,
+      "step": 4052
+    },
+    {
+      "epoch": 0.3270922443709144,
+      "grad_norm": 0.7957910895347595,
+      "learning_rate": 0.0001811515049985997,
+      "loss": 2.7237,
+      "step": 4053
+    },
+    {
+      "epoch": 0.32717294810749736,
+      "grad_norm": 0.7828991413116455,
+      "learning_rate": 0.00018114227920544375,
+      "loss": 2.7008,
+      "step": 4054
+    },
+    {
+      "epoch": 0.3272536518440804,
+      "grad_norm": 0.6695161461830139,
+      "learning_rate": 0.00018113305139001016,
+      "loss": 2.7311,
+      "step": 4055
+    },
+    {
+      "epoch": 0.32733435558066337,
+      "grad_norm": 0.7693436145782471,
+      "learning_rate": 0.00018112382155252883,
+      "loss": 2.7102,
+      "step": 4056
+    },
+    {
+      "epoch": 0.3274150593172464,
+      "grad_norm": 0.7520042657852173,
+      "learning_rate": 0.0001811145896932298,
+      "loss": 2.6455,
+      "step": 4057
+    },
+    {
+      "epoch": 0.3274957630538294,
+      "grad_norm": 0.786834716796875,
+      "learning_rate": 0.00018110535581234317,
+      "loss": 2.6965,
+      "step": 4058
+    },
+    {
+      "epoch": 0.3275764667904124,
+      "grad_norm": 0.742001473903656,
+      "learning_rate": 0.00018109611991009905,
+      "loss": 2.7341,
+      "step": 4059
+    },
+    {
+      "epoch": 0.3276571705269954,
+      "grad_norm": 0.813522219657898,
+      "learning_rate": 0.00018108688198672766,
+      "loss": 2.8116,
+      "step": 4060
+    },
+    {
+      "epoch": 0.3277378742635784,
+      "grad_norm": 0.7611314058303833,
+      "learning_rate": 0.00018107764204245916,
+      "loss": 2.6741,
+      "step": 4061
+    },
+    {
+      "epoch": 0.3278185780001614,
+      "grad_norm": 0.7285993695259094,
+      "learning_rate": 0.00018106840007752392,
+      "loss": 2.671,
+      "step": 4062
+    },
+    {
+      "epoch": 0.32789928173674443,
+      "grad_norm": 0.773151695728302,
+      "learning_rate": 0.0001810591560921522,
+      "loss": 2.7106,
+      "step": 4063
+    },
+    {
+      "epoch": 0.3279799854733274,
+      "grad_norm": 0.7448920011520386,
+      "learning_rate": 0.00018104991008657445,
+      "loss": 2.7176,
+      "step": 4064
+    },
+    {
+      "epoch": 0.32806068920991044,
+      "grad_norm": 0.7088467478752136,
+      "learning_rate": 0.0001810406620610211,
+      "loss": 2.7085,
+      "step": 4065
+    },
+    {
+      "epoch": 0.3281413929464934,
+      "grad_norm": 0.7507789731025696,
+      "learning_rate": 0.00018103141201572255,
+      "loss": 2.7361,
+      "step": 4066
+    },
+    {
+      "epoch": 0.32822209668307645,
+      "grad_norm": 0.7065643072128296,
+      "learning_rate": 0.00018102215995090943,
+      "loss": 2.6573,
+      "step": 4067
+    },
+    {
+      "epoch": 0.3283028004196594,
+      "grad_norm": 0.6888713836669922,
+      "learning_rate": 0.0001810129058668123,
+      "loss": 2.6699,
+      "step": 4068
+    },
+    {
+      "epoch": 0.32838350415624246,
+      "grad_norm": 0.736347496509552,
+      "learning_rate": 0.00018100364976366174,
+      "loss": 2.7089,
+      "step": 4069
+    },
+    {
+      "epoch": 0.32846420789282543,
+      "grad_norm": 0.6854562759399414,
+      "learning_rate": 0.0001809943916416885,
+      "loss": 2.7051,
+      "step": 4070
+    },
+    {
+      "epoch": 0.32854491162940846,
+      "grad_norm": 0.7481048107147217,
+      "learning_rate": 0.0001809851315011233,
+      "loss": 2.7428,
+      "step": 4071
+    },
+    {
+      "epoch": 0.32862561536599144,
+      "grad_norm": 0.7600961923599243,
+      "learning_rate": 0.0001809758693421969,
+      "loss": 2.7153,
+      "step": 4072
+    },
+    {
+      "epoch": 0.3287063191025745,
+      "grad_norm": 0.7545063495635986,
+      "learning_rate": 0.00018096660516514024,
+      "loss": 2.6736,
+      "step": 4073
+    },
+    {
+      "epoch": 0.32878702283915745,
+      "grad_norm": 0.7967175841331482,
+      "learning_rate": 0.0001809573389701841,
+      "loss": 2.6711,
+      "step": 4074
+    },
+    {
+      "epoch": 0.3288677265757405,
+      "grad_norm": 0.7115446925163269,
+      "learning_rate": 0.00018094807075755943,
+      "loss": 2.6761,
+      "step": 4075
+    },
+    {
+      "epoch": 0.32894843031232346,
+      "grad_norm": 0.8230876326560974,
+      "learning_rate": 0.00018093880052749725,
+      "loss": 2.6749,
+      "step": 4076
+    },
+    {
+      "epoch": 0.3290291340489065,
+      "grad_norm": 0.8549706935882568,
+      "learning_rate": 0.00018092952828022856,
+      "loss": 2.7084,
+      "step": 4077
+    },
+    {
+      "epoch": 0.32910983778548947,
+      "grad_norm": 0.7379534244537354,
+      "learning_rate": 0.00018092025401598448,
+      "loss": 2.7241,
+      "step": 4078
+    },
+    {
+      "epoch": 0.3291905415220725,
+      "grad_norm": 0.7659998536109924,
+      "learning_rate": 0.00018091097773499616,
+      "loss": 2.7108,
+      "step": 4079
+    },
+    {
+      "epoch": 0.3292712452586555,
+      "grad_norm": 0.8074536323547363,
+      "learning_rate": 0.00018090169943749476,
+      "loss": 2.676,
+      "step": 4080
+    },
+    {
+      "epoch": 0.32935194899523845,
+      "grad_norm": 0.7588536143302917,
+      "learning_rate": 0.00018089241912371153,
+      "loss": 2.639,
+      "step": 4081
+    },
+    {
+      "epoch": 0.3294326527318215,
+      "grad_norm": 0.7510811686515808,
+      "learning_rate": 0.00018088313679387775,
+      "loss": 2.6722,
+      "step": 4082
+    },
+    {
+      "epoch": 0.32951335646840446,
+      "grad_norm": 0.7538900971412659,
+      "learning_rate": 0.0001808738524482248,
+      "loss": 2.6917,
+      "step": 4083
+    },
+    {
+      "epoch": 0.3295940602049875,
+      "grad_norm": 0.8071155548095703,
+      "learning_rate": 0.00018086456608698402,
+      "loss": 2.6964,
+      "step": 4084
+    },
+    {
+      "epoch": 0.32967476394157047,
+      "grad_norm": 0.7778098583221436,
+      "learning_rate": 0.00018085527771038686,
+      "loss": 2.7301,
+      "step": 4085
+    },
+    {
+      "epoch": 0.3297554676781535,
+      "grad_norm": 0.7717564702033997,
+      "learning_rate": 0.00018084598731866485,
+      "loss": 2.7484,
+      "step": 4086
+    },
+    {
+      "epoch": 0.3298361714147365,
+      "grad_norm": 0.7361736297607422,
+      "learning_rate": 0.00018083669491204948,
+      "loss": 2.6299,
+      "step": 4087
+    },
+    {
+      "epoch": 0.3299168751513195,
+      "grad_norm": 0.736681342124939,
+      "learning_rate": 0.00018082740049077238,
+      "loss": 2.7521,
+      "step": 4088
+    },
+    {
+      "epoch": 0.3299975788879025,
+      "grad_norm": 0.8011857867240906,
+      "learning_rate": 0.00018081810405506517,
+      "loss": 2.724,
+      "step": 4089
+    },
+    {
+      "epoch": 0.3300782826244855,
+      "grad_norm": 0.7741932272911072,
+      "learning_rate": 0.00018080880560515956,
+      "loss": 2.6766,
+      "step": 4090
+    },
+    {
+      "epoch": 0.3301589863610685,
+      "grad_norm": 0.7321778535842896,
+      "learning_rate": 0.00018079950514128724,
+      "loss": 2.6614,
+      "step": 4091
+    },
+    {
+      "epoch": 0.33023969009765153,
+      "grad_norm": 0.7916514277458191,
+      "learning_rate": 0.00018079020266368006,
+      "loss": 2.7177,
+      "step": 4092
+    },
+    {
+      "epoch": 0.3303203938342345,
+      "grad_norm": 0.7961388826370239,
+      "learning_rate": 0.00018078089817256986,
+      "loss": 2.6671,
+      "step": 4093
+    },
+    {
+      "epoch": 0.33040109757081754,
+      "grad_norm": 0.7167038321495056,
+      "learning_rate": 0.0001807715916681885,
+      "loss": 2.6989,
+      "step": 4094
+    },
+    {
+      "epoch": 0.3304818013074005,
+      "grad_norm": 0.6924864649772644,
+      "learning_rate": 0.00018076228315076794,
+      "loss": 2.6484,
+      "step": 4095
+    },
+    {
+      "epoch": 0.33056250504398355,
+      "grad_norm": 0.777881383895874,
+      "learning_rate": 0.00018075297262054013,
+      "loss": 2.6498,
+      "step": 4096
+    },
+    {
+      "epoch": 0.3306432087805665,
+      "grad_norm": 0.7878376841545105,
+      "learning_rate": 0.0001807436600777372,
+      "loss": 2.7745,
+      "step": 4097
+    },
+    {
+      "epoch": 0.33072391251714955,
+      "grad_norm": 0.8418465256690979,
+      "learning_rate": 0.0001807343455225912,
+      "loss": 2.7195,
+      "step": 4098
+    },
+    {
+      "epoch": 0.33080461625373253,
+      "grad_norm": 0.7780830264091492,
+      "learning_rate": 0.00018072502895533424,
+      "loss": 2.6652,
+      "step": 4099
+    },
+    {
+      "epoch": 0.33088531999031556,
+      "grad_norm": 0.7102445960044861,
+      "learning_rate": 0.00018071571037619853,
+      "loss": 2.6618,
+      "step": 4100
+    },
+    {
+      "epoch": 0.33096602372689854,
+      "grad_norm": 0.7028098106384277,
+      "learning_rate": 0.00018070638978541633,
+      "loss": 2.7114,
+      "step": 4101
+    },
+    {
+      "epoch": 0.33104672746348157,
+      "grad_norm": 0.7529525756835938,
+      "learning_rate": 0.00018069706718321996,
+      "loss": 2.7231,
+      "step": 4102
+    },
+    {
+      "epoch": 0.33112743120006455,
+      "grad_norm": 0.7404564023017883,
+      "learning_rate": 0.0001806877425698417,
+      "loss": 2.6564,
+      "step": 4103
+    },
+    {
+      "epoch": 0.3312081349366476,
+      "grad_norm": 0.7725130319595337,
+      "learning_rate": 0.00018067841594551401,
+      "loss": 2.677,
+      "step": 4104
+    },
+    {
+      "epoch": 0.33128883867323056,
+      "grad_norm": 0.7616425156593323,
+      "learning_rate": 0.00018066908731046927,
+      "loss": 2.6586,
+      "step": 4105
+    },
+    {
+      "epoch": 0.3313695424098136,
+      "grad_norm": 0.7318183779716492,
+      "learning_rate": 0.00018065975666494002,
+      "loss": 2.6624,
+      "step": 4106
+    },
+    {
+      "epoch": 0.33145024614639657,
+      "grad_norm": 0.7012802958488464,
+      "learning_rate": 0.00018065042400915878,
+      "loss": 2.6663,
+      "step": 4107
+    },
+    {
+      "epoch": 0.3315309498829796,
+      "grad_norm": 0.815226674079895,
+      "learning_rate": 0.00018064108934335814,
+      "loss": 2.7248,
+      "step": 4108
+    },
+    {
+      "epoch": 0.3316116536195626,
+      "grad_norm": 0.68972247838974,
+      "learning_rate": 0.00018063175266777077,
+      "loss": 2.6961,
+      "step": 4109
+    },
+    {
+      "epoch": 0.3316923573561456,
+      "grad_norm": 0.7563794255256653,
+      "learning_rate": 0.00018062241398262937,
+      "loss": 2.6526,
+      "step": 4110
+    },
+    {
+      "epoch": 0.3317730610927286,
+      "grad_norm": 0.7878836989402771,
+      "learning_rate": 0.00018061307328816662,
+      "loss": 2.7316,
+      "step": 4111
+    },
+    {
+      "epoch": 0.3318537648293116,
+      "grad_norm": 0.7189129590988159,
+      "learning_rate": 0.00018060373058461537,
+      "loss": 2.6577,
+      "step": 4112
+    },
+    {
+      "epoch": 0.3319344685658946,
+      "grad_norm": 0.7517561912536621,
+      "learning_rate": 0.00018059438587220847,
+      "loss": 2.668,
+      "step": 4113
+    },
+    {
+      "epoch": 0.3320151723024776,
+      "grad_norm": 0.7602595686912537,
+      "learning_rate": 0.00018058503915117878,
+      "loss": 2.6741,
+      "step": 4114
+    },
+    {
+      "epoch": 0.3320958760390606,
+      "grad_norm": 0.7702187299728394,
+      "learning_rate": 0.00018057569042175927,
+      "loss": 2.7082,
+      "step": 4115
+    },
+    {
+      "epoch": 0.33217657977564363,
+      "grad_norm": 0.7289660573005676,
+      "learning_rate": 0.00018056633968418294,
+      "loss": 2.6728,
+      "step": 4116
+    },
+    {
+      "epoch": 0.3322572835122266,
+      "grad_norm": 0.6936683654785156,
+      "learning_rate": 0.0001805569869386828,
+      "loss": 2.6735,
+      "step": 4117
+    },
+    {
+      "epoch": 0.33233798724880964,
+      "grad_norm": 0.7128138542175293,
+      "learning_rate": 0.000180547632185492,
+      "loss": 2.646,
+      "step": 4118
+    },
+    {
+      "epoch": 0.3324186909853926,
+      "grad_norm": 0.7234248518943787,
+      "learning_rate": 0.00018053827542484363,
+      "loss": 2.6497,
+      "step": 4119
+    },
+    {
+      "epoch": 0.33249939472197565,
+      "grad_norm": 0.7084202170372009,
+      "learning_rate": 0.0001805289166569709,
+      "loss": 2.6328,
+      "step": 4120
+    },
+    {
+      "epoch": 0.3325800984585586,
+      "grad_norm": 0.8068051934242249,
+      "learning_rate": 0.00018051955588210708,
+      "loss": 2.6576,
+      "step": 4121
+    },
+    {
+      "epoch": 0.33266080219514166,
+      "grad_norm": 0.787680447101593,
+      "learning_rate": 0.00018051019310048544,
+      "loss": 2.7091,
+      "step": 4122
+    },
+    {
+      "epoch": 0.33274150593172463,
+      "grad_norm": 0.698946475982666,
+      "learning_rate": 0.00018050082831233931,
+      "loss": 2.6657,
+      "step": 4123
+    },
+    {
+      "epoch": 0.33282220966830767,
+      "grad_norm": 0.7946122288703918,
+      "learning_rate": 0.00018049146151790215,
+      "loss": 2.6981,
+      "step": 4124
+    },
+    {
+      "epoch": 0.33290291340489064,
+      "grad_norm": 0.8025123476982117,
+      "learning_rate": 0.00018048209271740736,
+      "loss": 2.6878,
+      "step": 4125
+    },
+    {
+      "epoch": 0.3329836171414737,
+      "grad_norm": 0.7493376135826111,
+      "learning_rate": 0.0001804727219110884,
+      "loss": 2.6556,
+      "step": 4126
+    },
+    {
+      "epoch": 0.33306432087805665,
+      "grad_norm": 0.7143186926841736,
+      "learning_rate": 0.00018046334909917886,
+      "loss": 2.6879,
+      "step": 4127
+    },
+    {
+      "epoch": 0.3331450246146397,
+      "grad_norm": 0.7375641465187073,
+      "learning_rate": 0.00018045397428191235,
+      "loss": 2.6817,
+      "step": 4128
+    },
+    {
+      "epoch": 0.33322572835122266,
+      "grad_norm": 0.7201291918754578,
+      "learning_rate": 0.00018044459745952248,
+      "loss": 2.6765,
+      "step": 4129
+    },
+    {
+      "epoch": 0.3333064320878057,
+      "grad_norm": 0.7924519777297974,
+      "learning_rate": 0.00018043521863224296,
+      "loss": 2.7748,
+      "step": 4130
+    },
+    {
+      "epoch": 0.33338713582438867,
+      "grad_norm": 0.7773354053497314,
+      "learning_rate": 0.00018042583780030752,
+      "loss": 2.6839,
+      "step": 4131
+    },
+    {
+      "epoch": 0.33346783956097165,
+      "grad_norm": 0.7527397274971008,
+      "learning_rate": 0.00018041645496394998,
+      "loss": 2.6749,
+      "step": 4132
+    },
+    {
+      "epoch": 0.3335485432975547,
+      "grad_norm": 0.7329208254814148,
+      "learning_rate": 0.00018040707012340418,
+      "loss": 2.7535,
+      "step": 4133
+    },
+    {
+      "epoch": 0.33362924703413765,
+      "grad_norm": 0.7637773752212524,
+      "learning_rate": 0.00018039768327890397,
+      "loss": 2.632,
+      "step": 4134
+    },
+    {
+      "epoch": 0.3337099507707207,
+      "grad_norm": 0.823623776435852,
+      "learning_rate": 0.00018038829443068333,
+      "loss": 2.7122,
+      "step": 4135
+    },
+    {
+      "epoch": 0.33379065450730366,
+      "grad_norm": 0.8040826916694641,
+      "learning_rate": 0.00018037890357897632,
+      "loss": 2.7197,
+      "step": 4136
+    },
+    {
+      "epoch": 0.3338713582438867,
+      "grad_norm": 0.7483998537063599,
+      "learning_rate": 0.00018036951072401686,
+      "loss": 2.6535,
+      "step": 4137
+    },
+    {
+      "epoch": 0.33395206198046967,
+      "grad_norm": 0.8141106367111206,
+      "learning_rate": 0.00018036011586603914,
+      "loss": 2.7127,
+      "step": 4138
+    },
+    {
+      "epoch": 0.3340327657170527,
+      "grad_norm": 0.7226041555404663,
+      "learning_rate": 0.00018035071900527724,
+      "loss": 2.6846,
+      "step": 4139
+    },
+    {
+      "epoch": 0.3341134694536357,
+      "grad_norm": 0.7624794840812683,
+      "learning_rate": 0.00018034132014196541,
+      "loss": 2.6725,
+      "step": 4140
+    },
+    {
+      "epoch": 0.3341941731902187,
+      "grad_norm": 0.7299962043762207,
+      "learning_rate": 0.00018033191927633785,
+      "loss": 2.6728,
+      "step": 4141
+    },
+    {
+      "epoch": 0.3342748769268017,
+      "grad_norm": 0.7920462489128113,
+      "learning_rate": 0.0001803225164086289,
+      "loss": 2.6544,
+      "step": 4142
+    },
+    {
+      "epoch": 0.3343555806633847,
+      "grad_norm": 0.7469778656959534,
+      "learning_rate": 0.00018031311153907282,
+      "loss": 2.7356,
+      "step": 4143
+    },
+    {
+      "epoch": 0.3344362843999677,
+      "grad_norm": 0.8831696510314941,
+      "learning_rate": 0.0001803037046679041,
+      "loss": 2.6584,
+      "step": 4144
+    },
+    {
+      "epoch": 0.33451698813655073,
+      "grad_norm": 0.8047679662704468,
+      "learning_rate": 0.00018029429579535715,
+      "loss": 2.6213,
+      "step": 4145
+    },
+    {
+      "epoch": 0.3345976918731337,
+      "grad_norm": 0.7109517455101013,
+      "learning_rate": 0.00018028488492166645,
+      "loss": 2.6622,
+      "step": 4146
+    },
+    {
+      "epoch": 0.33467839560971674,
+      "grad_norm": 0.7240141034126282,
+      "learning_rate": 0.0001802754720470665,
+      "loss": 2.6794,
+      "step": 4147
+    },
+    {
+      "epoch": 0.3347590993462997,
+      "grad_norm": 0.7292990684509277,
+      "learning_rate": 0.000180266057171792,
+      "loss": 2.6079,
+      "step": 4148
+    },
+    {
+      "epoch": 0.33483980308288275,
+      "grad_norm": 0.8055328130722046,
+      "learning_rate": 0.00018025664029607756,
+      "loss": 2.7044,
+      "step": 4149
+    },
+    {
+      "epoch": 0.3349205068194657,
+      "grad_norm": 0.8348979949951172,
+      "learning_rate": 0.00018024722142015781,
+      "loss": 2.6757,
+      "step": 4150
+    },
+    {
+      "epoch": 0.33500121055604876,
+      "grad_norm": 0.7797044515609741,
+      "learning_rate": 0.00018023780054426754,
+      "loss": 2.7125,
+      "step": 4151
+    },
+    {
+      "epoch": 0.33508191429263173,
+      "grad_norm": 0.802442729473114,
+      "learning_rate": 0.00018022837766864153,
+      "loss": 2.7121,
+      "step": 4152
+    },
+    {
+      "epoch": 0.33516261802921476,
+      "grad_norm": 0.7248829007148743,
+      "learning_rate": 0.00018021895279351463,
+      "loss": 2.7344,
+      "step": 4153
+    },
+    {
+      "epoch": 0.33524332176579774,
+      "grad_norm": 0.7458582520484924,
+      "learning_rate": 0.00018020952591912175,
+      "loss": 2.665,
+      "step": 4154
+    },
+    {
+      "epoch": 0.3353240255023808,
+      "grad_norm": 0.8153703808784485,
+      "learning_rate": 0.0001802000970456978,
+      "loss": 2.7416,
+      "step": 4155
+    },
+    {
+      "epoch": 0.33540472923896375,
+      "grad_norm": 0.7583708763122559,
+      "learning_rate": 0.00018019066617347779,
+      "loss": 2.7002,
+      "step": 4156
+    },
+    {
+      "epoch": 0.3354854329755468,
+      "grad_norm": 0.7522469162940979,
+      "learning_rate": 0.00018018123330269678,
+      "loss": 2.7196,
+      "step": 4157
+    },
+    {
+      "epoch": 0.33556613671212976,
+      "grad_norm": 0.7386923432350159,
+      "learning_rate": 0.00018017179843358983,
+      "loss": 2.6947,
+      "step": 4158
+    },
+    {
+      "epoch": 0.3356468404487128,
+      "grad_norm": 0.7366231083869934,
+      "learning_rate": 0.00018016236156639205,
+      "loss": 2.7377,
+      "step": 4159
+    },
+    {
+      "epoch": 0.33572754418529577,
+      "grad_norm": 0.7727232575416565,
+      "learning_rate": 0.00018015292270133872,
+      "loss": 2.7566,
+      "step": 4160
+    },
+    {
+      "epoch": 0.3358082479218788,
+      "grad_norm": 0.6781843304634094,
+      "learning_rate": 0.000180143481838665,
+      "loss": 2.6796,
+      "step": 4161
+    },
+    {
+      "epoch": 0.3358889516584618,
+      "grad_norm": 0.7036039233207703,
+      "learning_rate": 0.00018013403897860624,
+      "loss": 2.7012,
+      "step": 4162
+    },
+    {
+      "epoch": 0.3359696553950448,
+      "grad_norm": 0.8252625465393066,
+      "learning_rate": 0.00018012459412139776,
+      "loss": 2.6613,
+      "step": 4163
+    },
+    {
+      "epoch": 0.3360503591316278,
+      "grad_norm": 0.6924486756324768,
+      "learning_rate": 0.00018011514726727493,
+      "loss": 2.6425,
+      "step": 4164
+    },
+    {
+      "epoch": 0.3361310628682108,
+      "grad_norm": 0.7735962271690369,
+      "learning_rate": 0.0001801056984164732,
+      "loss": 2.7235,
+      "step": 4165
+    },
+    {
+      "epoch": 0.3362117666047938,
+      "grad_norm": 0.7439951300621033,
+      "learning_rate": 0.0001800962475692281,
+      "loss": 2.7428,
+      "step": 4166
+    },
+    {
+      "epoch": 0.3362924703413768,
+      "grad_norm": 0.6830539107322693,
+      "learning_rate": 0.0001800867947257751,
+      "loss": 2.5907,
+      "step": 4167
+    },
+    {
+      "epoch": 0.3363731740779598,
+      "grad_norm": 0.8355144262313843,
+      "learning_rate": 0.00018007733988634986,
+      "loss": 2.6978,
+      "step": 4168
+    },
+    {
+      "epoch": 0.33645387781454283,
+      "grad_norm": 0.6880978941917419,
+      "learning_rate": 0.00018006788305118798,
+      "loss": 2.6934,
+      "step": 4169
+    },
+    {
+      "epoch": 0.3365345815511258,
+      "grad_norm": 0.762709379196167,
+      "learning_rate": 0.0001800584242205251,
+      "loss": 2.684,
+      "step": 4170
+    },
+    {
+      "epoch": 0.33661528528770884,
+      "grad_norm": 0.7543070912361145,
+      "learning_rate": 0.0001800489633945971,
+      "loss": 2.6857,
+      "step": 4171
+    },
+    {
+      "epoch": 0.3366959890242918,
+      "grad_norm": 0.787651777267456,
+      "learning_rate": 0.00018003950057363964,
+      "loss": 2.6979,
+      "step": 4172
+    },
+    {
+      "epoch": 0.33677669276087485,
+      "grad_norm": 0.7831481099128723,
+      "learning_rate": 0.00018003003575788856,
+      "loss": 2.7158,
+      "step": 4173
+    },
+    {
+      "epoch": 0.33685739649745783,
+      "grad_norm": 0.844904363155365,
+      "learning_rate": 0.00018002056894757986,
+      "loss": 2.6459,
+      "step": 4174
+    },
+    {
+      "epoch": 0.33693810023404086,
+      "grad_norm": 0.7529420852661133,
+      "learning_rate": 0.00018001110014294937,
+      "loss": 2.685,
+      "step": 4175
+    },
+    {
+      "epoch": 0.33701880397062384,
+      "grad_norm": 0.776719868183136,
+      "learning_rate": 0.0001800016293442331,
+      "loss": 2.6353,
+      "step": 4176
+    },
+    {
+      "epoch": 0.33709950770720687,
+      "grad_norm": 0.7988671660423279,
+      "learning_rate": 0.00017999215655166716,
+      "loss": 2.7241,
+      "step": 4177
+    },
+    {
+      "epoch": 0.33718021144378985,
+      "grad_norm": 0.7190617918968201,
+      "learning_rate": 0.00017998268176548752,
+      "loss": 2.7278,
+      "step": 4178
+    },
+    {
+      "epoch": 0.3372609151803729,
+      "grad_norm": 0.8337060809135437,
+      "learning_rate": 0.0001799732049859304,
+      "loss": 2.7059,
+      "step": 4179
+    },
+    {
+      "epoch": 0.33734161891695585,
+      "grad_norm": 0.7547435164451599,
+      "learning_rate": 0.0001799637262132319,
+      "loss": 2.7782,
+      "step": 4180
+    },
+    {
+      "epoch": 0.3374223226535389,
+      "grad_norm": 0.8067883253097534,
+      "learning_rate": 0.0001799542454476284,
+      "loss": 2.7978,
+      "step": 4181
+    },
+    {
+      "epoch": 0.33750302639012186,
+      "grad_norm": 0.7451581358909607,
+      "learning_rate": 0.00017994476268935609,
+      "loss": 2.6931,
+      "step": 4182
+    },
+    {
+      "epoch": 0.33758373012670484,
+      "grad_norm": 0.7521898746490479,
+      "learning_rate": 0.00017993527793865125,
+      "loss": 2.6939,
+      "step": 4183
+    },
+    {
+      "epoch": 0.33766443386328787,
+      "grad_norm": 0.7608996033668518,
+      "learning_rate": 0.0001799257911957504,
+      "loss": 2.715,
+      "step": 4184
+    },
+    {
+      "epoch": 0.33774513759987085,
+      "grad_norm": 0.7459948658943176,
+      "learning_rate": 0.00017991630246088987,
+      "loss": 2.6951,
+      "step": 4185
+    },
+    {
+      "epoch": 0.3378258413364539,
+      "grad_norm": 0.7549717426300049,
+      "learning_rate": 0.00017990681173430618,
+      "loss": 2.7353,
+      "step": 4186
+    },
+    {
+      "epoch": 0.33790654507303686,
+      "grad_norm": 0.7234344482421875,
+      "learning_rate": 0.0001798973190162359,
+      "loss": 2.6491,
+      "step": 4187
+    },
+    {
+      "epoch": 0.3379872488096199,
+      "grad_norm": 0.7652330994606018,
+      "learning_rate": 0.00017988782430691553,
+      "loss": 2.765,
+      "step": 4188
+    },
+    {
+      "epoch": 0.33806795254620287,
+      "grad_norm": 0.742953360080719,
+      "learning_rate": 0.00017987832760658177,
+      "loss": 2.7079,
+      "step": 4189
+    },
+    {
+      "epoch": 0.3381486562827859,
+      "grad_norm": 0.7440767288208008,
+      "learning_rate": 0.00017986882891547125,
+      "loss": 2.6751,
+      "step": 4190
+    },
+    {
+      "epoch": 0.3382293600193689,
+      "grad_norm": 0.7141925096511841,
+      "learning_rate": 0.00017985932823382078,
+      "loss": 2.6249,
+      "step": 4191
+    },
+    {
+      "epoch": 0.3383100637559519,
+      "grad_norm": 0.7200489044189453,
+      "learning_rate": 0.00017984982556186707,
+      "loss": 2.6811,
+      "step": 4192
+    },
+    {
+      "epoch": 0.3383907674925349,
+      "grad_norm": 0.7677409648895264,
+      "learning_rate": 0.00017984032089984696,
+      "loss": 2.6641,
+      "step": 4193
+    },
+    {
+      "epoch": 0.3384714712291179,
+      "grad_norm": 0.7386545538902283,
+      "learning_rate": 0.00017983081424799741,
+      "loss": 2.6504,
+      "step": 4194
+    },
+    {
+      "epoch": 0.3385521749657009,
+      "grad_norm": 0.7528583407402039,
+      "learning_rate": 0.00017982130560655526,
+      "loss": 2.6422,
+      "step": 4195
+    },
+    {
+      "epoch": 0.3386328787022839,
+      "grad_norm": 0.7339407801628113,
+      "learning_rate": 0.0001798117949757575,
+      "loss": 2.7047,
+      "step": 4196
+    },
+    {
+      "epoch": 0.3387135824388669,
+      "grad_norm": 0.7655882239341736,
+      "learning_rate": 0.00017980228235584117,
+      "loss": 2.7644,
+      "step": 4197
+    },
+    {
+      "epoch": 0.33879428617544993,
+      "grad_norm": 0.7602109909057617,
+      "learning_rate": 0.00017979276774704342,
+      "loss": 2.697,
+      "step": 4198
+    },
+    {
+      "epoch": 0.3388749899120329,
+      "grad_norm": 0.7188911437988281,
+      "learning_rate": 0.00017978325114960126,
+      "loss": 2.7147,
+      "step": 4199
+    },
+    {
+      "epoch": 0.33895569364861594,
+      "grad_norm": 0.7672597765922546,
+      "learning_rate": 0.00017977373256375194,
+      "loss": 2.6558,
+      "step": 4200
+    },
+    {
+      "epoch": 0.3390363973851989,
+      "grad_norm": 0.784187912940979,
+      "learning_rate": 0.0001797642119897327,
+      "loss": 2.7005,
+      "step": 4201
+    },
+    {
+      "epoch": 0.33911710112178195,
+      "grad_norm": 0.7359703779220581,
+      "learning_rate": 0.00017975468942778075,
+      "loss": 2.6578,
+      "step": 4202
+    },
+    {
+      "epoch": 0.3391978048583649,
+      "grad_norm": 0.7776080965995789,
+      "learning_rate": 0.00017974516487813345,
+      "loss": 2.6747,
+      "step": 4203
+    },
+    {
+      "epoch": 0.33927850859494796,
+      "grad_norm": 0.6934135556221008,
+      "learning_rate": 0.00017973563834102824,
+      "loss": 2.6335,
+      "step": 4204
+    },
+    {
+      "epoch": 0.33935921233153094,
+      "grad_norm": 0.7715818881988525,
+      "learning_rate": 0.00017972610981670245,
+      "loss": 2.6062,
+      "step": 4205
+    },
+    {
+      "epoch": 0.33943991606811397,
+      "grad_norm": 0.7466367483139038,
+      "learning_rate": 0.0001797165793053936,
+      "loss": 2.7243,
+      "step": 4206
+    },
+    {
+      "epoch": 0.33952061980469694,
+      "grad_norm": 0.7485085129737854,
+      "learning_rate": 0.00017970704680733926,
+      "loss": 2.6603,
+      "step": 4207
+    },
+    {
+      "epoch": 0.33960132354128,
+      "grad_norm": 0.7365782856941223,
+      "learning_rate": 0.0001796975123227769,
+      "loss": 2.7179,
+      "step": 4208
+    },
+    {
+      "epoch": 0.33968202727786295,
+      "grad_norm": 0.8405506014823914,
+      "learning_rate": 0.00017968797585194422,
+      "loss": 2.7413,
+      "step": 4209
+    },
+    {
+      "epoch": 0.339762731014446,
+      "grad_norm": 0.8227888941764832,
+      "learning_rate": 0.00017967843739507888,
+      "loss": 2.6814,
+      "step": 4210
+    },
+    {
+      "epoch": 0.33984343475102896,
+      "grad_norm": 0.8247283697128296,
+      "learning_rate": 0.0001796688969524186,
+      "loss": 2.6802,
+      "step": 4211
+    },
+    {
+      "epoch": 0.339924138487612,
+      "grad_norm": 0.7639476656913757,
+      "learning_rate": 0.00017965935452420116,
+      "loss": 2.7422,
+      "step": 4212
+    },
+    {
+      "epoch": 0.34000484222419497,
+      "grad_norm": 0.7846776247024536,
+      "learning_rate": 0.00017964981011066436,
+      "loss": 2.7443,
+      "step": 4213
+    },
+    {
+      "epoch": 0.340085545960778,
+      "grad_norm": 0.7593334913253784,
+      "learning_rate": 0.00017964026371204608,
+      "loss": 2.7179,
+      "step": 4214
+    },
+    {
+      "epoch": 0.340166249697361,
+      "grad_norm": 0.7878177165985107,
+      "learning_rate": 0.00017963071532858425,
+      "loss": 2.7118,
+      "step": 4215
+    },
+    {
+      "epoch": 0.340246953433944,
+      "grad_norm": 0.7728220224380493,
+      "learning_rate": 0.00017962116496051685,
+      "loss": 2.6646,
+      "step": 4216
+    },
+    {
+      "epoch": 0.340327657170527,
+      "grad_norm": 0.8419308066368103,
+      "learning_rate": 0.00017961161260808187,
+      "loss": 2.7829,
+      "step": 4217
+    },
+    {
+      "epoch": 0.34040836090711,
+      "grad_norm": 0.7066153883934021,
+      "learning_rate": 0.0001796020582715174,
+      "loss": 2.6498,
+      "step": 4218
+    },
+    {
+      "epoch": 0.340489064643693,
+      "grad_norm": 0.7976264953613281,
+      "learning_rate": 0.00017959250195106156,
+      "loss": 2.7496,
+      "step": 4219
+    },
+    {
+      "epoch": 0.34056976838027603,
+      "grad_norm": 0.736595630645752,
+      "learning_rate": 0.0001795829436469525,
+      "loss": 2.6497,
+      "step": 4220
+    },
+    {
+      "epoch": 0.340650472116859,
+      "grad_norm": 0.818550705909729,
+      "learning_rate": 0.0001795733833594285,
+      "loss": 2.6793,
+      "step": 4221
+    },
+    {
+      "epoch": 0.34073117585344204,
+      "grad_norm": 0.7712778449058533,
+      "learning_rate": 0.00017956382108872773,
+      "loss": 2.6215,
+      "step": 4222
+    },
+    {
+      "epoch": 0.340811879590025,
+      "grad_norm": 0.746306300163269,
+      "learning_rate": 0.00017955425683508858,
+      "loss": 2.7372,
+      "step": 4223
+    },
+    {
+      "epoch": 0.34089258332660805,
+      "grad_norm": 0.7269306778907776,
+      "learning_rate": 0.00017954469059874937,
+      "loss": 2.6438,
+      "step": 4224
+    },
+    {
+      "epoch": 0.340973287063191,
+      "grad_norm": 0.7426211833953857,
+      "learning_rate": 0.00017953512237994855,
+      "loss": 2.6539,
+      "step": 4225
+    },
+    {
+      "epoch": 0.34105399079977405,
+      "grad_norm": 0.7269948124885559,
+      "learning_rate": 0.0001795255521789246,
+      "loss": 2.6833,
+      "step": 4226
+    },
+    {
+      "epoch": 0.34113469453635703,
+      "grad_norm": 0.7279343605041504,
+      "learning_rate": 0.00017951597999591598,
+      "loss": 2.7011,
+      "step": 4227
+    },
+    {
+      "epoch": 0.34121539827294006,
+      "grad_norm": 0.7554663419723511,
+      "learning_rate": 0.0001795064058311613,
+      "loss": 2.7036,
+      "step": 4228
+    },
+    {
+      "epoch": 0.34129610200952304,
+      "grad_norm": 0.7516502141952515,
+      "learning_rate": 0.00017949682968489912,
+      "loss": 2.6699,
+      "step": 4229
+    },
+    {
+      "epoch": 0.34137680574610607,
+      "grad_norm": 0.7931745052337646,
+      "learning_rate": 0.00017948725155736818,
+      "loss": 2.6655,
+      "step": 4230
+    },
+    {
+      "epoch": 0.34145750948268905,
+      "grad_norm": 0.6981344223022461,
+      "learning_rate": 0.0001794776714488071,
+      "loss": 2.6987,
+      "step": 4231
+    },
+    {
+      "epoch": 0.3415382132192721,
+      "grad_norm": 0.7513911724090576,
+      "learning_rate": 0.00017946808935945474,
+      "loss": 2.6985,
+      "step": 4232
+    },
+    {
+      "epoch": 0.34161891695585506,
+      "grad_norm": 0.7373185753822327,
+      "learning_rate": 0.00017945850528954983,
+      "loss": 2.7269,
+      "step": 4233
+    },
+    {
+      "epoch": 0.34169962069243803,
+      "grad_norm": 0.6990259289741516,
+      "learning_rate": 0.0001794489192393313,
+      "loss": 2.6763,
+      "step": 4234
+    },
+    {
+      "epoch": 0.34178032442902107,
+      "grad_norm": 0.7661817669868469,
+      "learning_rate": 0.00017943933120903797,
+      "loss": 2.7057,
+      "step": 4235
+    },
+    {
+      "epoch": 0.34186102816560404,
+      "grad_norm": 0.7570027112960815,
+      "learning_rate": 0.0001794297411989089,
+      "loss": 2.7358,
+      "step": 4236
+    },
+    {
+      "epoch": 0.3419417319021871,
+      "grad_norm": 0.7751824855804443,
+      "learning_rate": 0.000179420149209183,
+      "loss": 2.6771,
+      "step": 4237
+    },
+    {
+      "epoch": 0.34202243563877005,
+      "grad_norm": 0.8028360605239868,
+      "learning_rate": 0.0001794105552400994,
+      "loss": 2.6399,
+      "step": 4238
+    },
+    {
+      "epoch": 0.3421031393753531,
+      "grad_norm": 0.7398171424865723,
+      "learning_rate": 0.00017940095929189716,
+      "loss": 2.6532,
+      "step": 4239
+    },
+    {
+      "epoch": 0.34218384311193606,
+      "grad_norm": 0.8300225138664246,
+      "learning_rate": 0.0001793913613648155,
+      "loss": 2.6798,
+      "step": 4240
+    },
+    {
+      "epoch": 0.3422645468485191,
+      "grad_norm": 0.7501145005226135,
+      "learning_rate": 0.00017938176145909356,
+      "loss": 2.7132,
+      "step": 4241
+    },
+    {
+      "epoch": 0.34234525058510207,
+      "grad_norm": 0.7178483605384827,
+      "learning_rate": 0.00017937215957497063,
+      "loss": 2.7172,
+      "step": 4242
+    },
+    {
+      "epoch": 0.3424259543216851,
+      "grad_norm": 0.7207306027412415,
+      "learning_rate": 0.00017936255571268599,
+      "loss": 2.629,
+      "step": 4243
+    },
+    {
+      "epoch": 0.3425066580582681,
+      "grad_norm": 0.7339839935302734,
+      "learning_rate": 0.00017935294987247899,
+      "loss": 2.6262,
+      "step": 4244
+    },
+    {
+      "epoch": 0.3425873617948511,
+      "grad_norm": 0.6977292895317078,
+      "learning_rate": 0.00017934334205458907,
+      "loss": 2.6949,
+      "step": 4245
+    },
+    {
+      "epoch": 0.3426680655314341,
+      "grad_norm": 0.7368096113204956,
+      "learning_rate": 0.00017933373225925564,
+      "loss": 2.681,
+      "step": 4246
+    },
+    {
+      "epoch": 0.3427487692680171,
+      "grad_norm": 0.7234459519386292,
+      "learning_rate": 0.00017932412048671825,
+      "loss": 2.6891,
+      "step": 4247
+    },
+    {
+      "epoch": 0.3428294730046001,
+      "grad_norm": 0.7659995555877686,
+      "learning_rate": 0.00017931450673721642,
+      "loss": 2.7394,
+      "step": 4248
+    },
+    {
+      "epoch": 0.3429101767411831,
+      "grad_norm": 0.7799893617630005,
+      "learning_rate": 0.00017930489101098974,
+      "loss": 2.7707,
+      "step": 4249
+    },
+    {
+      "epoch": 0.3429908804777661,
+      "grad_norm": 0.7063946723937988,
+      "learning_rate": 0.00017929527330827786,
+      "loss": 2.6573,
+      "step": 4250
+    },
+    {
+      "epoch": 0.34307158421434913,
+      "grad_norm": 0.7090561389923096,
+      "learning_rate": 0.0001792856536293205,
+      "loss": 2.7095,
+      "step": 4251
+    },
+    {
+      "epoch": 0.3431522879509321,
+      "grad_norm": 0.8020029067993164,
+      "learning_rate": 0.0001792760319743574,
+      "loss": 2.6905,
+      "step": 4252
+    },
+    {
+      "epoch": 0.34323299168751514,
+      "grad_norm": 0.7221484780311584,
+      "learning_rate": 0.00017926640834362836,
+      "loss": 2.6853,
+      "step": 4253
+    },
+    {
+      "epoch": 0.3433136954240981,
+      "grad_norm": 0.7102623581886292,
+      "learning_rate": 0.00017925678273737324,
+      "loss": 2.6821,
+      "step": 4254
+    },
+    {
+      "epoch": 0.34339439916068115,
+      "grad_norm": 0.7702807784080505,
+      "learning_rate": 0.00017924715515583187,
+      "loss": 2.6986,
+      "step": 4255
+    },
+    {
+      "epoch": 0.34347510289726413,
+      "grad_norm": 0.7938152551651001,
+      "learning_rate": 0.00017923752559924425,
+      "loss": 2.7162,
+      "step": 4256
+    },
+    {
+      "epoch": 0.34355580663384716,
+      "grad_norm": 0.7340937852859497,
+      "learning_rate": 0.00017922789406785036,
+      "loss": 2.6904,
+      "step": 4257
+    },
+    {
+      "epoch": 0.34363651037043014,
+      "grad_norm": 0.7010839581489563,
+      "learning_rate": 0.00017921826056189026,
+      "loss": 2.6969,
+      "step": 4258
+    },
+    {
+      "epoch": 0.34371721410701317,
+      "grad_norm": 0.758178174495697,
+      "learning_rate": 0.00017920862508160403,
+      "loss": 2.6391,
+      "step": 4259
+    },
+    {
+      "epoch": 0.34379791784359615,
+      "grad_norm": 0.7861726880073547,
+      "learning_rate": 0.0001791989876272318,
+      "loss": 2.7088,
+      "step": 4260
+    },
+    {
+      "epoch": 0.3438786215801792,
+      "grad_norm": 0.6764364242553711,
+      "learning_rate": 0.00017918934819901377,
+      "loss": 2.6221,
+      "step": 4261
+    },
+    {
+      "epoch": 0.34395932531676215,
+      "grad_norm": 0.76728355884552,
+      "learning_rate": 0.00017917970679719018,
+      "loss": 2.6854,
+      "step": 4262
+    },
+    {
+      "epoch": 0.3440400290533452,
+      "grad_norm": 0.7161166071891785,
+      "learning_rate": 0.00017917006342200133,
+      "loss": 2.7048,
+      "step": 4263
+    },
+    {
+      "epoch": 0.34412073278992816,
+      "grad_norm": 0.7182073593139648,
+      "learning_rate": 0.00017916041807368753,
+      "loss": 2.7559,
+      "step": 4264
+    },
+    {
+      "epoch": 0.3442014365265112,
+      "grad_norm": 0.832258403301239,
+      "learning_rate": 0.0001791507707524892,
+      "loss": 2.6743,
+      "step": 4265
+    },
+    {
+      "epoch": 0.34428214026309417,
+      "grad_norm": 0.7048495411872864,
+      "learning_rate": 0.00017914112145864675,
+      "loss": 2.693,
+      "step": 4266
+    },
+    {
+      "epoch": 0.3443628439996772,
+      "grad_norm": 0.7475518584251404,
+      "learning_rate": 0.00017913147019240068,
+      "loss": 2.6881,
+      "step": 4267
+    },
+    {
+      "epoch": 0.3444435477362602,
+      "grad_norm": 0.72830730676651,
+      "learning_rate": 0.00017912181695399154,
+      "loss": 2.659,
+      "step": 4268
+    },
+    {
+      "epoch": 0.3445242514728432,
+      "grad_norm": 0.7183662056922913,
+      "learning_rate": 0.00017911216174365988,
+      "loss": 2.6611,
+      "step": 4269
+    },
+    {
+      "epoch": 0.3446049552094262,
+      "grad_norm": 0.7487103343009949,
+      "learning_rate": 0.0001791025045616463,
+      "loss": 2.6518,
+      "step": 4270
+    },
+    {
+      "epoch": 0.3446856589460092,
+      "grad_norm": 0.7733812928199768,
+      "learning_rate": 0.0001790928454081916,
+      "loss": 2.6359,
+      "step": 4271
+    },
+    {
+      "epoch": 0.3447663626825922,
+      "grad_norm": 0.7774991393089294,
+      "learning_rate": 0.00017908318428353642,
+      "loss": 2.6654,
+      "step": 4272
+    },
+    {
+      "epoch": 0.34484706641917523,
+      "grad_norm": 0.6882895827293396,
+      "learning_rate": 0.00017907352118792157,
+      "loss": 2.686,
+      "step": 4273
+    },
+    {
+      "epoch": 0.3449277701557582,
+      "grad_norm": 0.7571535110473633,
+      "learning_rate": 0.00017906385612158785,
+      "loss": 2.7108,
+      "step": 4274
+    },
+    {
+      "epoch": 0.34500847389234124,
+      "grad_norm": 0.7324517369270325,
+      "learning_rate": 0.00017905418908477615,
+      "loss": 2.6663,
+      "step": 4275
+    },
+    {
+      "epoch": 0.3450891776289242,
+      "grad_norm": 0.7476221919059753,
+      "learning_rate": 0.00017904452007772744,
+      "loss": 2.7202,
+      "step": 4276
+    },
+    {
+      "epoch": 0.34516988136550725,
+      "grad_norm": 0.7648386359214783,
+      "learning_rate": 0.00017903484910068268,
+      "loss": 2.6759,
+      "step": 4277
+    },
+    {
+      "epoch": 0.3452505851020902,
+      "grad_norm": 0.7375434637069702,
+      "learning_rate": 0.00017902517615388282,
+      "loss": 2.6603,
+      "step": 4278
+    },
+    {
+      "epoch": 0.34533128883867326,
+      "grad_norm": 0.7248519062995911,
+      "learning_rate": 0.00017901550123756906,
+      "loss": 2.7147,
+      "step": 4279
+    },
+    {
+      "epoch": 0.34541199257525623,
+      "grad_norm": 0.7264916896820068,
+      "learning_rate": 0.0001790058243519824,
+      "loss": 2.6992,
+      "step": 4280
+    },
+    {
+      "epoch": 0.34549269631183926,
+      "grad_norm": 0.8370026350021362,
+      "learning_rate": 0.0001789961454973641,
+      "loss": 2.7114,
+      "step": 4281
+    },
+    {
+      "epoch": 0.34557340004842224,
+      "grad_norm": 0.72071373462677,
+      "learning_rate": 0.00017898646467395538,
+      "loss": 2.6957,
+      "step": 4282
+    },
+    {
+      "epoch": 0.3456541037850053,
+      "grad_norm": 0.7355397343635559,
+      "learning_rate": 0.0001789767818819975,
+      "loss": 2.6744,
+      "step": 4283
+    },
+    {
+      "epoch": 0.34573480752158825,
+      "grad_norm": 0.734756588935852,
+      "learning_rate": 0.00017896709712173173,
+      "loss": 2.726,
+      "step": 4284
+    },
+    {
+      "epoch": 0.3458155112581712,
+      "grad_norm": 0.7890543341636658,
+      "learning_rate": 0.00017895741039339945,
+      "loss": 2.6726,
+      "step": 4285
+    },
+    {
+      "epoch": 0.34589621499475426,
+      "grad_norm": 0.7768735885620117,
+      "learning_rate": 0.00017894772169724216,
+      "loss": 2.7617,
+      "step": 4286
+    },
+    {
+      "epoch": 0.34597691873133724,
+      "grad_norm": 0.7306547164916992,
+      "learning_rate": 0.00017893803103350125,
+      "loss": 2.6253,
+      "step": 4287
+    },
+    {
+      "epoch": 0.34605762246792027,
+      "grad_norm": 0.767066478729248,
+      "learning_rate": 0.00017892833840241828,
+      "loss": 2.6522,
+      "step": 4288
+    },
+    {
+      "epoch": 0.34613832620450324,
+      "grad_norm": 0.7018097639083862,
+      "learning_rate": 0.00017891864380423477,
+      "loss": 2.7111,
+      "step": 4289
+    },
+    {
+      "epoch": 0.3462190299410863,
+      "grad_norm": 0.7305615544319153,
+      "learning_rate": 0.00017890894723919236,
+      "loss": 2.6924,
+      "step": 4290
+    },
+    {
+      "epoch": 0.34629973367766925,
+      "grad_norm": 0.7588002681732178,
+      "learning_rate": 0.00017889924870753275,
+      "loss": 2.6952,
+      "step": 4291
+    },
+    {
+      "epoch": 0.3463804374142523,
+      "grad_norm": 0.7162861824035645,
+      "learning_rate": 0.0001788895482094976,
+      "loss": 2.6239,
+      "step": 4292
+    },
+    {
+      "epoch": 0.34646114115083526,
+      "grad_norm": 0.7494024634361267,
+      "learning_rate": 0.00017887984574532868,
+      "loss": 2.6763,
+      "step": 4293
+    },
+    {
+      "epoch": 0.3465418448874183,
+      "grad_norm": 0.7100037336349487,
+      "learning_rate": 0.0001788701413152678,
+      "loss": 2.6378,
+      "step": 4294
+    },
+    {
+      "epoch": 0.34662254862400127,
+      "grad_norm": 0.7316900491714478,
+      "learning_rate": 0.00017886043491955684,
+      "loss": 2.7001,
+      "step": 4295
+    },
+    {
+      "epoch": 0.3467032523605843,
+      "grad_norm": 0.8467028737068176,
+      "learning_rate": 0.00017885072655843772,
+      "loss": 2.7536,
+      "step": 4296
+    },
+    {
+      "epoch": 0.3467839560971673,
+      "grad_norm": 0.7248796820640564,
+      "learning_rate": 0.00017884101623215237,
+      "loss": 2.6956,
+      "step": 4297
+    },
+    {
+      "epoch": 0.3468646598337503,
+      "grad_norm": 0.7183107137680054,
+      "learning_rate": 0.0001788313039409428,
+      "loss": 2.743,
+      "step": 4298
+    },
+    {
+      "epoch": 0.3469453635703333,
+      "grad_norm": 0.6835163831710815,
+      "learning_rate": 0.00017882158968505105,
+      "loss": 2.7016,
+      "step": 4299
+    },
+    {
+      "epoch": 0.3470260673069163,
+      "grad_norm": 0.7973365783691406,
+      "learning_rate": 0.00017881187346471925,
+      "loss": 2.6927,
+      "step": 4300
+    },
+    {
+      "epoch": 0.3471067710434993,
+      "grad_norm": 0.700040876865387,
+      "learning_rate": 0.00017880215528018954,
+      "loss": 2.6961,
+      "step": 4301
+    },
+    {
+      "epoch": 0.34718747478008233,
+      "grad_norm": 0.8180583119392395,
+      "learning_rate": 0.00017879243513170415,
+      "loss": 2.642,
+      "step": 4302
+    },
+    {
+      "epoch": 0.3472681785166653,
+      "grad_norm": 0.7134599685668945,
+      "learning_rate": 0.0001787827130195053,
+      "loss": 2.6901,
+      "step": 4303
+    },
+    {
+      "epoch": 0.34734888225324834,
+      "grad_norm": 0.767998218536377,
+      "learning_rate": 0.0001787729889438353,
+      "loss": 2.6472,
+      "step": 4304
+    },
+    {
+      "epoch": 0.3474295859898313,
+      "grad_norm": 0.7260780930519104,
+      "learning_rate": 0.0001787632629049365,
+      "loss": 2.6791,
+      "step": 4305
+    },
+    {
+      "epoch": 0.34751028972641435,
+      "grad_norm": 0.6918236613273621,
+      "learning_rate": 0.00017875353490305132,
+      "loss": 2.6596,
+      "step": 4306
+    },
+    {
+      "epoch": 0.3475909934629973,
+      "grad_norm": 0.7734197974205017,
+      "learning_rate": 0.00017874380493842216,
+      "loss": 2.6402,
+      "step": 4307
+    },
+    {
+      "epoch": 0.34767169719958035,
+      "grad_norm": 0.7051037549972534,
+      "learning_rate": 0.00017873407301129154,
+      "loss": 2.7517,
+      "step": 4308
+    },
+    {
+      "epoch": 0.34775240093616333,
+      "grad_norm": 0.7026919722557068,
+      "learning_rate": 0.00017872433912190203,
+      "loss": 2.7058,
+      "step": 4309
+    },
+    {
+      "epoch": 0.34783310467274636,
+      "grad_norm": 0.7248546481132507,
+      "learning_rate": 0.00017871460327049618,
+      "loss": 2.666,
+      "step": 4310
+    },
+    {
+      "epoch": 0.34791380840932934,
+      "grad_norm": 0.7348842620849609,
+      "learning_rate": 0.0001787048654573167,
+      "loss": 2.7712,
+      "step": 4311
+    },
+    {
+      "epoch": 0.34799451214591237,
+      "grad_norm": 0.7923693656921387,
+      "learning_rate": 0.00017869512568260618,
+      "loss": 2.6469,
+      "step": 4312
+    },
+    {
+      "epoch": 0.34807521588249535,
+      "grad_norm": 0.7604066729545593,
+      "learning_rate": 0.00017868538394660743,
+      "loss": 2.7152,
+      "step": 4313
+    },
+    {
+      "epoch": 0.3481559196190784,
+      "grad_norm": 0.6811137795448303,
+      "learning_rate": 0.00017867564024956324,
+      "loss": 2.715,
+      "step": 4314
+    },
+    {
+      "epoch": 0.34823662335566136,
+      "grad_norm": 0.7292799353599548,
+      "learning_rate": 0.00017866589459171643,
+      "loss": 2.6374,
+      "step": 4315
+    },
+    {
+      "epoch": 0.3483173270922444,
+      "grad_norm": 0.6961250901222229,
+      "learning_rate": 0.0001786561469733099,
+      "loss": 2.6592,
+      "step": 4316
+    },
+    {
+      "epoch": 0.34839803082882737,
+      "grad_norm": 0.7447086572647095,
+      "learning_rate": 0.00017864639739458658,
+      "loss": 2.6965,
+      "step": 4317
+    },
+    {
+      "epoch": 0.3484787345654104,
+      "grad_norm": 0.7107378244400024,
+      "learning_rate": 0.00017863664585578942,
+      "loss": 2.7057,
+      "step": 4318
+    },
+    {
+      "epoch": 0.3485594383019934,
+      "grad_norm": 0.7372235655784607,
+      "learning_rate": 0.00017862689235716153,
+      "loss": 2.6289,
+      "step": 4319
+    },
+    {
+      "epoch": 0.3486401420385764,
+      "grad_norm": 0.7360481023788452,
+      "learning_rate": 0.00017861713689894593,
+      "loss": 2.7208,
+      "step": 4320
+    },
+    {
+      "epoch": 0.3487208457751594,
+      "grad_norm": 0.7378106713294983,
+      "learning_rate": 0.00017860737948138575,
+      "loss": 2.6836,
+      "step": 4321
+    },
+    {
+      "epoch": 0.3488015495117424,
+      "grad_norm": 0.7110548615455627,
+      "learning_rate": 0.00017859762010472423,
+      "loss": 2.6941,
+      "step": 4322
+    },
+    {
+      "epoch": 0.3488822532483254,
+      "grad_norm": 0.7419706583023071,
+      "learning_rate": 0.00017858785876920455,
+      "loss": 2.6591,
+      "step": 4323
+    },
+    {
+      "epoch": 0.3489629569849084,
+      "grad_norm": 0.7759542465209961,
+      "learning_rate": 0.00017857809547506997,
+      "loss": 2.6966,
+      "step": 4324
+    },
+    {
+      "epoch": 0.3490436607214914,
+      "grad_norm": 0.7894207239151001,
+      "learning_rate": 0.0001785683302225639,
+      "loss": 2.7298,
+      "step": 4325
+    },
+    {
+      "epoch": 0.34912436445807443,
+      "grad_norm": 0.7342399954795837,
+      "learning_rate": 0.0001785585630119296,
+      "loss": 2.6998,
+      "step": 4326
+    },
+    {
+      "epoch": 0.3492050681946574,
+      "grad_norm": 0.8684173822402954,
+      "learning_rate": 0.0001785487938434106,
+      "loss": 2.7179,
+      "step": 4327
+    },
+    {
+      "epoch": 0.34928577193124044,
+      "grad_norm": 0.7557523846626282,
+      "learning_rate": 0.00017853902271725033,
+      "loss": 2.7081,
+      "step": 4328
+    },
+    {
+      "epoch": 0.3493664756678234,
+      "grad_norm": 0.7910173535346985,
+      "learning_rate": 0.0001785292496336923,
+      "loss": 2.718,
+      "step": 4329
+    },
+    {
+      "epoch": 0.34944717940440645,
+      "grad_norm": 0.7878917455673218,
+      "learning_rate": 0.00017851947459298007,
+      "loss": 2.674,
+      "step": 4330
+    },
+    {
+      "epoch": 0.3495278831409894,
+      "grad_norm": 0.7290656566619873,
+      "learning_rate": 0.0001785096975953573,
+      "loss": 2.6962,
+      "step": 4331
+    },
+    {
+      "epoch": 0.34960858687757246,
+      "grad_norm": 0.8465737104415894,
+      "learning_rate": 0.00017849991864106763,
+      "loss": 2.6793,
+      "step": 4332
+    },
+    {
+      "epoch": 0.34968929061415543,
+      "grad_norm": 0.7183132171630859,
+      "learning_rate": 0.0001784901377303548,
+      "loss": 2.6902,
+      "step": 4333
+    },
+    {
+      "epoch": 0.34976999435073847,
+      "grad_norm": 0.7535461783409119,
+      "learning_rate": 0.00017848035486346255,
+      "loss": 2.7153,
+      "step": 4334
+    },
+    {
+      "epoch": 0.34985069808732144,
+      "grad_norm": 0.778734028339386,
+      "learning_rate": 0.0001784705700406347,
+      "loss": 2.6316,
+      "step": 4335
+    },
+    {
+      "epoch": 0.3499314018239044,
+      "grad_norm": 0.6937401294708252,
+      "learning_rate": 0.00017846078326211516,
+      "loss": 2.6902,
+      "step": 4336
+    },
+    {
+      "epoch": 0.35001210556048745,
+      "grad_norm": 0.7450751066207886,
+      "learning_rate": 0.00017845099452814774,
+      "loss": 2.6898,
+      "step": 4337
+    },
+    {
+      "epoch": 0.35009280929707043,
+      "grad_norm": 0.7535614967346191,
+      "learning_rate": 0.0001784412038389765,
+      "loss": 2.6969,
+      "step": 4338
+    },
+    {
+      "epoch": 0.35017351303365346,
+      "grad_norm": 0.6971385478973389,
+      "learning_rate": 0.00017843141119484543,
+      "loss": 2.6517,
+      "step": 4339
+    },
+    {
+      "epoch": 0.35025421677023644,
+      "grad_norm": 0.7233202457427979,
+      "learning_rate": 0.00017842161659599858,
+      "loss": 2.7332,
+      "step": 4340
+    },
+    {
+      "epoch": 0.35033492050681947,
+      "grad_norm": 0.7870340347290039,
+      "learning_rate": 0.00017841182004268,
+      "loss": 2.6485,
+      "step": 4341
+    },
+    {
+      "epoch": 0.35041562424340245,
+      "grad_norm": 0.7387053966522217,
+      "learning_rate": 0.0001784020215351339,
+      "loss": 2.6945,
+      "step": 4342
+    },
+    {
+      "epoch": 0.3504963279799855,
+      "grad_norm": 0.8357887268066406,
+      "learning_rate": 0.00017839222107360453,
+      "loss": 2.703,
+      "step": 4343
+    },
+    {
+      "epoch": 0.35057703171656845,
+      "grad_norm": 0.7197332978248596,
+      "learning_rate": 0.000178382418658336,
+      "loss": 2.6649,
+      "step": 4344
+    },
+    {
+      "epoch": 0.3506577354531515,
+      "grad_norm": 0.7416980862617493,
+      "learning_rate": 0.0001783726142895728,
+      "loss": 2.7393,
+      "step": 4345
+    },
+    {
+      "epoch": 0.35073843918973446,
+      "grad_norm": 0.6807832717895508,
+      "learning_rate": 0.00017836280796755912,
+      "loss": 2.6619,
+      "step": 4346
+    },
+    {
+      "epoch": 0.3508191429263175,
+      "grad_norm": 0.6858795285224915,
+      "learning_rate": 0.00017835299969253945,
+      "loss": 2.6266,
+      "step": 4347
+    },
+    {
+      "epoch": 0.35089984666290047,
+      "grad_norm": 0.8432363867759705,
+      "learning_rate": 0.0001783431894647582,
+      "loss": 2.6534,
+      "step": 4348
+    },
+    {
+      "epoch": 0.3509805503994835,
+      "grad_norm": 0.7240749001502991,
+      "learning_rate": 0.0001783333772844599,
+      "loss": 2.6851,
+      "step": 4349
+    },
+    {
+      "epoch": 0.3510612541360665,
+      "grad_norm": 0.7814531326293945,
+      "learning_rate": 0.00017832356315188906,
+      "loss": 2.7085,
+      "step": 4350
+    },
+    {
+      "epoch": 0.3511419578726495,
+      "grad_norm": 0.6989716291427612,
+      "learning_rate": 0.00017831374706729026,
+      "loss": 2.6674,
+      "step": 4351
+    },
+    {
+      "epoch": 0.3512226616092325,
+      "grad_norm": 0.7118446230888367,
+      "learning_rate": 0.0001783039290309082,
+      "loss": 2.6837,
+      "step": 4352
+    },
+    {
+      "epoch": 0.3513033653458155,
+      "grad_norm": 0.7641892433166504,
+      "learning_rate": 0.00017829410904298754,
+      "loss": 2.6415,
+      "step": 4353
+    },
+    {
+      "epoch": 0.3513840690823985,
+      "grad_norm": 0.6975794434547424,
+      "learning_rate": 0.000178284287103773,
+      "loss": 2.6679,
+      "step": 4354
+    },
+    {
+      "epoch": 0.35146477281898153,
+      "grad_norm": 0.7192546725273132,
+      "learning_rate": 0.00017827446321350943,
+      "loss": 2.6539,
+      "step": 4355
+    },
+    {
+      "epoch": 0.3515454765555645,
+      "grad_norm": 0.8749549388885498,
+      "learning_rate": 0.00017826463737244155,
+      "loss": 2.7254,
+      "step": 4356
+    },
+    {
+      "epoch": 0.35162618029214754,
+      "grad_norm": 0.8509732484817505,
+      "learning_rate": 0.0001782548095808144,
+      "loss": 2.7679,
+      "step": 4357
+    },
+    {
+      "epoch": 0.3517068840287305,
+      "grad_norm": 0.7647901773452759,
+      "learning_rate": 0.00017824497983887278,
+      "loss": 2.7049,
+      "step": 4358
+    },
+    {
+      "epoch": 0.35178758776531355,
+      "grad_norm": 0.7551973462104797,
+      "learning_rate": 0.00017823514814686178,
+      "loss": 2.7086,
+      "step": 4359
+    },
+    {
+      "epoch": 0.3518682915018965,
+      "grad_norm": 0.730140209197998,
+      "learning_rate": 0.00017822531450502633,
+      "loss": 2.6334,
+      "step": 4360
+    },
+    {
+      "epoch": 0.35194899523847956,
+      "grad_norm": 0.8210160136222839,
+      "learning_rate": 0.00017821547891361158,
+      "loss": 2.7248,
+      "step": 4361
+    },
+    {
+      "epoch": 0.35202969897506253,
+      "grad_norm": 0.761972963809967,
+      "learning_rate": 0.00017820564137286264,
+      "loss": 2.6502,
+      "step": 4362
+    },
+    {
+      "epoch": 0.35211040271164556,
+      "grad_norm": 0.7564061284065247,
+      "learning_rate": 0.00017819580188302466,
+      "loss": 2.6795,
+      "step": 4363
+    },
+    {
+      "epoch": 0.35219110644822854,
+      "grad_norm": 0.7382947206497192,
+      "learning_rate": 0.00017818596044434293,
+      "loss": 2.6754,
+      "step": 4364
+    },
+    {
+      "epoch": 0.3522718101848116,
+      "grad_norm": 0.737194836139679,
+      "learning_rate": 0.00017817611705706266,
+      "loss": 2.7098,
+      "step": 4365
+    },
+    {
+      "epoch": 0.35235251392139455,
+      "grad_norm": 0.7183281779289246,
+      "learning_rate": 0.0001781662717214292,
+      "loss": 2.6528,
+      "step": 4366
+    },
+    {
+      "epoch": 0.3524332176579776,
+      "grad_norm": 0.7785990238189697,
+      "learning_rate": 0.00017815642443768794,
+      "loss": 2.6419,
+      "step": 4367
+    },
+    {
+      "epoch": 0.35251392139456056,
+      "grad_norm": 0.7114452719688416,
+      "learning_rate": 0.00017814657520608427,
+      "loss": 2.7088,
+      "step": 4368
+    },
+    {
+      "epoch": 0.3525946251311436,
+      "grad_norm": 0.746969997882843,
+      "learning_rate": 0.00017813672402686365,
+      "loss": 2.7199,
+      "step": 4369
+    },
+    {
+      "epoch": 0.35267532886772657,
+      "grad_norm": 0.7700605988502502,
+      "learning_rate": 0.00017812687090027165,
+      "loss": 2.6713,
+      "step": 4370
+    },
+    {
+      "epoch": 0.3527560326043096,
+      "grad_norm": 0.7733504772186279,
+      "learning_rate": 0.0001781170158265538,
+      "loss": 2.6916,
+      "step": 4371
+    },
+    {
+      "epoch": 0.3528367363408926,
+      "grad_norm": 0.7769689559936523,
+      "learning_rate": 0.00017810715880595566,
+      "loss": 2.7787,
+      "step": 4372
+    },
+    {
+      "epoch": 0.3529174400774756,
+      "grad_norm": 0.7538996934890747,
+      "learning_rate": 0.000178097299838723,
+      "loss": 2.6964,
+      "step": 4373
+    },
+    {
+      "epoch": 0.3529981438140586,
+      "grad_norm": 0.7777890563011169,
+      "learning_rate": 0.00017808743892510146,
+      "loss": 2.6882,
+      "step": 4374
+    },
+    {
+      "epoch": 0.3530788475506416,
+      "grad_norm": 0.8331751823425293,
+      "learning_rate": 0.00017807757606533683,
+      "loss": 2.7113,
+      "step": 4375
+    },
+    {
+      "epoch": 0.3531595512872246,
+      "grad_norm": 0.8039207458496094,
+      "learning_rate": 0.00017806771125967492,
+      "loss": 2.6694,
+      "step": 4376
+    },
+    {
+      "epoch": 0.3532402550238076,
+      "grad_norm": 0.7727575898170471,
+      "learning_rate": 0.00017805784450836154,
+      "loss": 2.6639,
+      "step": 4377
+    },
+    {
+      "epoch": 0.3533209587603906,
+      "grad_norm": 0.8247967958450317,
+      "learning_rate": 0.00017804797581164264,
+      "loss": 2.6539,
+      "step": 4378
+    },
+    {
+      "epoch": 0.35340166249697363,
+      "grad_norm": 0.7574009299278259,
+      "learning_rate": 0.0001780381051697642,
+      "loss": 2.7163,
+      "step": 4379
+    },
+    {
+      "epoch": 0.3534823662335566,
+      "grad_norm": 0.7304368615150452,
+      "learning_rate": 0.0001780282325829721,
+      "loss": 2.5759,
+      "step": 4380
+    },
+    {
+      "epoch": 0.35356306997013964,
+      "grad_norm": 0.7133963704109192,
+      "learning_rate": 0.00017801835805151257,
+      "loss": 2.7008,
+      "step": 4381
+    },
+    {
+      "epoch": 0.3536437737067226,
+      "grad_norm": 0.7525407075881958,
+      "learning_rate": 0.00017800848157563157,
+      "loss": 2.6785,
+      "step": 4382
+    },
+    {
+      "epoch": 0.35372447744330565,
+      "grad_norm": 0.7306779623031616,
+      "learning_rate": 0.00017799860315557528,
+      "loss": 2.6454,
+      "step": 4383
+    },
+    {
+      "epoch": 0.35380518117988863,
+      "grad_norm": 0.6657043695449829,
+      "learning_rate": 0.00017798872279158994,
+      "loss": 2.708,
+      "step": 4384
+    },
+    {
+      "epoch": 0.35388588491647166,
+      "grad_norm": 0.7655978202819824,
+      "learning_rate": 0.00017797884048392177,
+      "loss": 2.727,
+      "step": 4385
+    },
+    {
+      "epoch": 0.35396658865305464,
+      "grad_norm": 0.6802939176559448,
+      "learning_rate": 0.00017796895623281702,
+      "loss": 2.659,
+      "step": 4386
+    },
+    {
+      "epoch": 0.3540472923896376,
+      "grad_norm": 0.7191160917282104,
+      "learning_rate": 0.00017795907003852207,
+      "loss": 2.6335,
+      "step": 4387
+    },
+    {
+      "epoch": 0.35412799612622065,
+      "grad_norm": 0.7771886587142944,
+      "learning_rate": 0.00017794918190128337,
+      "loss": 2.6658,
+      "step": 4388
+    },
+    {
+      "epoch": 0.3542086998628036,
+      "grad_norm": 0.7133512496948242,
+      "learning_rate": 0.00017793929182134723,
+      "loss": 2.6701,
+      "step": 4389
+    },
+    {
+      "epoch": 0.35428940359938665,
+      "grad_norm": 0.7795221209526062,
+      "learning_rate": 0.00017792939979896022,
+      "loss": 2.6932,
+      "step": 4390
+    },
+    {
+      "epoch": 0.35437010733596963,
+      "grad_norm": 0.726767897605896,
+      "learning_rate": 0.00017791950583436887,
+      "loss": 2.676,
+      "step": 4391
+    },
+    {
+      "epoch": 0.35445081107255266,
+      "grad_norm": 0.7447288632392883,
+      "learning_rate": 0.00017790960992781972,
+      "loss": 2.7195,
+      "step": 4392
+    },
+    {
+      "epoch": 0.35453151480913564,
+      "grad_norm": 0.8053649663925171,
+      "learning_rate": 0.0001778997120795595,
+      "loss": 2.6851,
+      "step": 4393
+    },
+    {
+      "epoch": 0.35461221854571867,
+      "grad_norm": 0.7258884906768799,
+      "learning_rate": 0.00017788981228983474,
+      "loss": 2.6819,
+      "step": 4394
+    },
+    {
+      "epoch": 0.35469292228230165,
+      "grad_norm": 0.7279395461082458,
+      "learning_rate": 0.0001778799105588923,
+      "loss": 2.6954,
+      "step": 4395
+    },
+    {
+      "epoch": 0.3547736260188847,
+      "grad_norm": 0.7372962236404419,
+      "learning_rate": 0.0001778700068869789,
+      "loss": 2.7049,
+      "step": 4396
+    },
+    {
+      "epoch": 0.35485432975546766,
+      "grad_norm": 0.712003767490387,
+      "learning_rate": 0.00017786010127434135,
+      "loss": 2.7413,
+      "step": 4397
+    },
+    {
+      "epoch": 0.3549350334920507,
+      "grad_norm": 0.7487424612045288,
+      "learning_rate": 0.0001778501937212266,
+      "loss": 2.7231,
+      "step": 4398
+    },
+    {
+      "epoch": 0.35501573722863367,
+      "grad_norm": 0.73053377866745,
+      "learning_rate": 0.00017784028422788146,
+      "loss": 2.7029,
+      "step": 4399
+    },
+    {
+      "epoch": 0.3550964409652167,
+      "grad_norm": 0.697062611579895,
+      "learning_rate": 0.00017783037279455298,
+      "loss": 2.7139,
+      "step": 4400
+    },
+    {
+      "epoch": 0.3551771447017997,
+      "grad_norm": 0.7750880718231201,
+      "learning_rate": 0.00017782045942148819,
+      "loss": 2.6601,
+      "step": 4401
+    },
+    {
+      "epoch": 0.3552578484383827,
+      "grad_norm": 0.7124977111816406,
+      "learning_rate": 0.00017781054410893413,
+      "loss": 2.6119,
+      "step": 4402
+    },
+    {
+      "epoch": 0.3553385521749657,
+      "grad_norm": 0.7773111462593079,
+      "learning_rate": 0.00017780062685713785,
+      "loss": 2.7181,
+      "step": 4403
+    },
+    {
+      "epoch": 0.3554192559115487,
+      "grad_norm": 0.7282142639160156,
+      "learning_rate": 0.00017779070766634663,
+      "loss": 2.7141,
+      "step": 4404
+    },
+    {
+      "epoch": 0.3554999596481317,
+      "grad_norm": 0.8578598499298096,
+      "learning_rate": 0.0001777807865368076,
+      "loss": 2.7628,
+      "step": 4405
+    },
+    {
+      "epoch": 0.3555806633847147,
+      "grad_norm": 0.7126399874687195,
+      "learning_rate": 0.00017777086346876809,
+      "loss": 2.6914,
+      "step": 4406
+    },
+    {
+      "epoch": 0.3556613671212977,
+      "grad_norm": 0.8026365637779236,
+      "learning_rate": 0.00017776093846247533,
+      "loss": 2.7059,
+      "step": 4407
+    },
+    {
+      "epoch": 0.35574207085788073,
+      "grad_norm": 0.7839884161949158,
+      "learning_rate": 0.0001777510115181767,
+      "loss": 2.7265,
+      "step": 4408
+    },
+    {
+      "epoch": 0.3558227745944637,
+      "grad_norm": 0.7498767971992493,
+      "learning_rate": 0.00017774108263611966,
+      "loss": 2.7201,
+      "step": 4409
+    },
+    {
+      "epoch": 0.35590347833104674,
+      "grad_norm": 0.6996301412582397,
+      "learning_rate": 0.0001777311518165516,
+      "loss": 2.6271,
+      "step": 4410
+    },
+    {
+      "epoch": 0.3559841820676297,
+      "grad_norm": 0.7721461057662964,
+      "learning_rate": 0.00017772121905972003,
+      "loss": 2.6739,
+      "step": 4411
+    },
+    {
+      "epoch": 0.35606488580421275,
+      "grad_norm": 0.8018803000450134,
+      "learning_rate": 0.00017771128436587256,
+      "loss": 2.7092,
+      "step": 4412
+    },
+    {
+      "epoch": 0.3561455895407957,
+      "grad_norm": 0.7185639142990112,
+      "learning_rate": 0.0001777013477352567,
+      "loss": 2.6996,
+      "step": 4413
+    },
+    {
+      "epoch": 0.35622629327737876,
+      "grad_norm": 0.7218519449234009,
+      "learning_rate": 0.0001776914091681202,
+      "loss": 2.6555,
+      "step": 4414
+    },
+    {
+      "epoch": 0.35630699701396173,
+      "grad_norm": 0.7234479188919067,
+      "learning_rate": 0.00017768146866471062,
+      "loss": 2.6762,
+      "step": 4415
+    },
+    {
+      "epoch": 0.35638770075054477,
+      "grad_norm": 0.6723350286483765,
+      "learning_rate": 0.00017767152622527582,
+      "loss": 2.6272,
+      "step": 4416
+    },
+    {
+      "epoch": 0.35646840448712774,
+      "grad_norm": 0.7281947731971741,
+      "learning_rate": 0.00017766158185006356,
+      "loss": 2.7216,
+      "step": 4417
+    },
+    {
+      "epoch": 0.3565491082237108,
+      "grad_norm": 0.8350874781608582,
+      "learning_rate": 0.00017765163553932166,
+      "loss": 2.6619,
+      "step": 4418
+    },
+    {
+      "epoch": 0.35662981196029375,
+      "grad_norm": 0.7454007267951965,
+      "learning_rate": 0.00017764168729329801,
+      "loss": 2.6623,
+      "step": 4419
+    },
+    {
+      "epoch": 0.3567105156968768,
+      "grad_norm": 0.7419041395187378,
+      "learning_rate": 0.00017763173711224058,
+      "loss": 2.6773,
+      "step": 4420
+    },
+    {
+      "epoch": 0.35679121943345976,
+      "grad_norm": 0.7965987920761108,
+      "learning_rate": 0.0001776217849963973,
+      "loss": 2.6426,
+      "step": 4421
+    },
+    {
+      "epoch": 0.3568719231700428,
+      "grad_norm": 0.7093302607536316,
+      "learning_rate": 0.00017761183094601622,
+      "loss": 2.6745,
+      "step": 4422
+    },
+    {
+      "epoch": 0.35695262690662577,
+      "grad_norm": 0.7937216758728027,
+      "learning_rate": 0.00017760187496134548,
+      "loss": 2.7275,
+      "step": 4423
+    },
+    {
+      "epoch": 0.3570333306432088,
+      "grad_norm": 0.9185259938240051,
+      "learning_rate": 0.00017759191704263313,
+      "loss": 2.7055,
+      "step": 4424
+    },
+    {
+      "epoch": 0.3571140343797918,
+      "grad_norm": 0.7365124821662903,
+      "learning_rate": 0.00017758195719012743,
+      "loss": 2.6504,
+      "step": 4425
+    },
+    {
+      "epoch": 0.3571947381163748,
+      "grad_norm": 0.6992416977882385,
+      "learning_rate": 0.0001775719954040765,
+      "loss": 2.6684,
+      "step": 4426
+    },
+    {
+      "epoch": 0.3572754418529578,
+      "grad_norm": 0.7742372751235962,
+      "learning_rate": 0.00017756203168472866,
+      "loss": 2.6877,
+      "step": 4427
+    },
+    {
+      "epoch": 0.3573561455895408,
+      "grad_norm": 0.7448472380638123,
+      "learning_rate": 0.0001775520660323323,
+      "loss": 2.7027,
+      "step": 4428
+    },
+    {
+      "epoch": 0.3574368493261238,
+      "grad_norm": 0.7201915979385376,
+      "learning_rate": 0.00017754209844713569,
+      "loss": 2.7046,
+      "step": 4429
+    },
+    {
+      "epoch": 0.3575175530627068,
+      "grad_norm": 0.6675081253051758,
+      "learning_rate": 0.0001775321289293873,
+      "loss": 2.6503,
+      "step": 4430
+    },
+    {
+      "epoch": 0.3575982567992898,
+      "grad_norm": 0.7252706289291382,
+      "learning_rate": 0.0001775221574793356,
+      "loss": 2.6053,
+      "step": 4431
+    },
+    {
+      "epoch": 0.35767896053587284,
+      "grad_norm": 0.7134702801704407,
+      "learning_rate": 0.00017751218409722906,
+      "loss": 2.6857,
+      "step": 4432
+    },
+    {
+      "epoch": 0.3577596642724558,
+      "grad_norm": 0.7074102163314819,
+      "learning_rate": 0.0001775022087833163,
+      "loss": 2.6871,
+      "step": 4433
+    },
+    {
+      "epoch": 0.35784036800903885,
+      "grad_norm": 0.693520724773407,
+      "learning_rate": 0.00017749223153784588,
+      "loss": 2.6629,
+      "step": 4434
+    },
+    {
+      "epoch": 0.3579210717456218,
+      "grad_norm": 0.6933221817016602,
+      "learning_rate": 0.0001774822523610665,
+      "loss": 2.6793,
+      "step": 4435
+    },
+    {
+      "epoch": 0.35800177548220485,
+      "grad_norm": 0.75307297706604,
+      "learning_rate": 0.00017747227125322685,
+      "loss": 2.7012,
+      "step": 4436
+    },
+    {
+      "epoch": 0.35808247921878783,
+      "grad_norm": 0.7732915282249451,
+      "learning_rate": 0.0001774622882145757,
+      "loss": 2.6908,
+      "step": 4437
+    },
+    {
+      "epoch": 0.3581631829553708,
+      "grad_norm": 0.7067054510116577,
+      "learning_rate": 0.0001774523032453618,
+      "loss": 2.7494,
+      "step": 4438
+    },
+    {
+      "epoch": 0.35824388669195384,
+      "grad_norm": 0.7412838935852051,
+      "learning_rate": 0.00017744231634583406,
+      "loss": 2.6734,
+      "step": 4439
+    },
+    {
+      "epoch": 0.3583245904285368,
+      "grad_norm": 0.7663930654525757,
+      "learning_rate": 0.00017743232751624136,
+      "loss": 2.6952,
+      "step": 4440
+    },
+    {
+      "epoch": 0.35840529416511985,
+      "grad_norm": 0.70650714635849,
+      "learning_rate": 0.00017742233675683268,
+      "loss": 2.6806,
+      "step": 4441
+    },
+    {
+      "epoch": 0.3584859979017028,
+      "grad_norm": 0.698310375213623,
+      "learning_rate": 0.00017741234406785692,
+      "loss": 2.6471,
+      "step": 4442
+    },
+    {
+      "epoch": 0.35856670163828586,
+      "grad_norm": 0.7274026274681091,
+      "learning_rate": 0.00017740234944956323,
+      "loss": 2.6688,
+      "step": 4443
+    },
+    {
+      "epoch": 0.35864740537486883,
+      "grad_norm": 0.6944074034690857,
+      "learning_rate": 0.00017739235290220067,
+      "loss": 2.6954,
+      "step": 4444
+    },
+    {
+      "epoch": 0.35872810911145186,
+      "grad_norm": 0.841995358467102,
+      "learning_rate": 0.00017738235442601834,
+      "loss": 2.7169,
+      "step": 4445
+    },
+    {
+      "epoch": 0.35880881284803484,
+      "grad_norm": 0.74863201379776,
+      "learning_rate": 0.00017737235402126545,
+      "loss": 2.6534,
+      "step": 4446
+    },
+    {
+      "epoch": 0.3588895165846179,
+      "grad_norm": 0.7260422110557556,
+      "learning_rate": 0.00017736235168819126,
+      "loss": 2.6266,
+      "step": 4447
+    },
+    {
+      "epoch": 0.35897022032120085,
+      "grad_norm": 0.7450951337814331,
+      "learning_rate": 0.00017735234742704504,
+      "loss": 2.7328,
+      "step": 4448
+    },
+    {
+      "epoch": 0.3590509240577839,
+      "grad_norm": 0.6942493319511414,
+      "learning_rate": 0.00017734234123807614,
+      "loss": 2.7219,
+      "step": 4449
+    },
+    {
+      "epoch": 0.35913162779436686,
+      "grad_norm": 0.7676761746406555,
+      "learning_rate": 0.00017733233312153393,
+      "loss": 2.6594,
+      "step": 4450
+    },
+    {
+      "epoch": 0.3592123315309499,
+      "grad_norm": 0.7446104288101196,
+      "learning_rate": 0.00017732232307766778,
+      "loss": 2.6877,
+      "step": 4451
+    },
+    {
+      "epoch": 0.35929303526753287,
+      "grad_norm": 0.7551130056381226,
+      "learning_rate": 0.00017731231110672727,
+      "loss": 2.672,
+      "step": 4452
+    },
+    {
+      "epoch": 0.3593737390041159,
+      "grad_norm": 0.6876464486122131,
+      "learning_rate": 0.00017730229720896182,
+      "loss": 2.6658,
+      "step": 4453
+    },
+    {
+      "epoch": 0.3594544427406989,
+      "grad_norm": 0.6992844343185425,
+      "learning_rate": 0.00017729228138462107,
+      "loss": 2.6805,
+      "step": 4454
+    },
+    {
+      "epoch": 0.3595351464772819,
+      "grad_norm": 0.8437497615814209,
+      "learning_rate": 0.00017728226363395466,
+      "loss": 2.6884,
+      "step": 4455
+    },
+    {
+      "epoch": 0.3596158502138649,
+      "grad_norm": 0.7669322490692139,
+      "learning_rate": 0.00017727224395721217,
+      "loss": 2.6432,
+      "step": 4456
+    },
+    {
+      "epoch": 0.3596965539504479,
+      "grad_norm": 0.7613428831100464,
+      "learning_rate": 0.0001772622223546434,
+      "loss": 2.6124,
+      "step": 4457
+    },
+    {
+      "epoch": 0.3597772576870309,
+      "grad_norm": 0.719932496547699,
+      "learning_rate": 0.00017725219882649807,
+      "loss": 2.6623,
+      "step": 4458
+    },
+    {
+      "epoch": 0.3598579614236139,
+      "grad_norm": 0.7650800347328186,
+      "learning_rate": 0.000177242173373026,
+      "loss": 2.7551,
+      "step": 4459
+    },
+    {
+      "epoch": 0.3599386651601969,
+      "grad_norm": 0.7423754930496216,
+      "learning_rate": 0.0001772321459944771,
+      "loss": 2.7375,
+      "step": 4460
+    },
+    {
+      "epoch": 0.36001936889677993,
+      "grad_norm": 0.7602835297584534,
+      "learning_rate": 0.0001772221166911012,
+      "loss": 2.7086,
+      "step": 4461
+    },
+    {
+      "epoch": 0.3601000726333629,
+      "grad_norm": 0.7246943712234497,
+      "learning_rate": 0.00017721208546314827,
+      "loss": 2.7068,
+      "step": 4462
+    },
+    {
+      "epoch": 0.36018077636994594,
+      "grad_norm": 0.715965211391449,
+      "learning_rate": 0.00017720205231086837,
+      "loss": 2.689,
+      "step": 4463
+    },
+    {
+      "epoch": 0.3602614801065289,
+      "grad_norm": 0.7696218490600586,
+      "learning_rate": 0.00017719201723451151,
+      "loss": 2.611,
+      "step": 4464
+    },
+    {
+      "epoch": 0.36034218384311195,
+      "grad_norm": 0.7599236369132996,
+      "learning_rate": 0.00017718198023432779,
+      "loss": 2.6504,
+      "step": 4465
+    },
+    {
+      "epoch": 0.36042288757969493,
+      "grad_norm": 0.7674956321716309,
+      "learning_rate": 0.0001771719413105674,
+      "loss": 2.7559,
+      "step": 4466
+    },
+    {
+      "epoch": 0.36050359131627796,
+      "grad_norm": 0.7263289093971252,
+      "learning_rate": 0.00017716190046348045,
+      "loss": 2.6822,
+      "step": 4467
+    },
+    {
+      "epoch": 0.36058429505286094,
+      "grad_norm": 0.7564195990562439,
+      "learning_rate": 0.0001771518576933173,
+      "loss": 2.7319,
+      "step": 4468
+    },
+    {
+      "epoch": 0.36066499878944397,
+      "grad_norm": 0.7291253805160522,
+      "learning_rate": 0.00017714181300032813,
+      "loss": 2.704,
+      "step": 4469
+    },
+    {
+      "epoch": 0.36074570252602695,
+      "grad_norm": 0.7354169487953186,
+      "learning_rate": 0.00017713176638476332,
+      "loss": 2.6344,
+      "step": 4470
+    },
+    {
+      "epoch": 0.36082640626261,
+      "grad_norm": 0.7104110717773438,
+      "learning_rate": 0.0001771217178468733,
+      "loss": 2.665,
+      "step": 4471
+    },
+    {
+      "epoch": 0.36090710999919295,
+      "grad_norm": 0.6913934350013733,
+      "learning_rate": 0.00017711166738690847,
+      "loss": 2.6674,
+      "step": 4472
+    },
+    {
+      "epoch": 0.360987813735776,
+      "grad_norm": 0.7999634742736816,
+      "learning_rate": 0.0001771016150051193,
+      "loss": 2.6847,
+      "step": 4473
+    },
+    {
+      "epoch": 0.36106851747235896,
+      "grad_norm": 0.7878915667533875,
+      "learning_rate": 0.00017709156070175634,
+      "loss": 2.7125,
+      "step": 4474
+    },
+    {
+      "epoch": 0.361149221208942,
+      "grad_norm": 0.7145688533782959,
+      "learning_rate": 0.00017708150447707017,
+      "loss": 2.6863,
+      "step": 4475
+    },
+    {
+      "epoch": 0.36122992494552497,
+      "grad_norm": 0.7518604397773743,
+      "learning_rate": 0.00017707144633131143,
+      "loss": 2.6616,
+      "step": 4476
+    },
+    {
+      "epoch": 0.361310628682108,
+      "grad_norm": 0.735634982585907,
+      "learning_rate": 0.0001770613862647308,
+      "loss": 2.6315,
+      "step": 4477
+    },
+    {
+      "epoch": 0.361391332418691,
+      "grad_norm": 0.7925180196762085,
+      "learning_rate": 0.00017705132427757895,
+      "loss": 2.6951,
+      "step": 4478
+    },
+    {
+      "epoch": 0.361472036155274,
+      "grad_norm": 0.6949547529220581,
+      "learning_rate": 0.00017704126037010667,
+      "loss": 2.6934,
+      "step": 4479
+    },
+    {
+      "epoch": 0.361552739891857,
+      "grad_norm": 0.7233577966690063,
+      "learning_rate": 0.00017703119454256483,
+      "loss": 2.6773,
+      "step": 4480
+    },
+    {
+      "epoch": 0.36163344362844,
+      "grad_norm": 0.7303269505500793,
+      "learning_rate": 0.00017702112679520424,
+      "loss": 2.6351,
+      "step": 4481
+    },
+    {
+      "epoch": 0.361714147365023,
+      "grad_norm": 0.7620660066604614,
+      "learning_rate": 0.00017701105712827583,
+      "loss": 2.6748,
+      "step": 4482
+    },
+    {
+      "epoch": 0.36179485110160603,
+      "grad_norm": 0.7744965553283691,
+      "learning_rate": 0.00017700098554203057,
+      "loss": 2.7013,
+      "step": 4483
+    },
+    {
+      "epoch": 0.361875554838189,
+      "grad_norm": 0.8017357587814331,
+      "learning_rate": 0.00017699091203671947,
+      "loss": 2.7273,
+      "step": 4484
+    },
+    {
+      "epoch": 0.36195625857477204,
+      "grad_norm": 0.8014432191848755,
+      "learning_rate": 0.0001769808366125936,
+      "loss": 2.6864,
+      "step": 4485
+    },
+    {
+      "epoch": 0.362036962311355,
+      "grad_norm": 0.6914888620376587,
+      "learning_rate": 0.00017697075926990406,
+      "loss": 2.6851,
+      "step": 4486
+    },
+    {
+      "epoch": 0.36211766604793805,
+      "grad_norm": 0.7472698092460632,
+      "learning_rate": 0.00017696068000890196,
+      "loss": 2.695,
+      "step": 4487
+    },
+    {
+      "epoch": 0.362198369784521,
+      "grad_norm": 0.7506285309791565,
+      "learning_rate": 0.00017695059882983855,
+      "loss": 2.7055,
+      "step": 4488
+    },
+    {
+      "epoch": 0.362279073521104,
+      "grad_norm": 0.7501141428947449,
+      "learning_rate": 0.00017694051573296507,
+      "loss": 2.7109,
+      "step": 4489
+    },
+    {
+      "epoch": 0.36235977725768703,
+      "grad_norm": 0.6654670834541321,
+      "learning_rate": 0.00017693043071853284,
+      "loss": 2.6165,
+      "step": 4490
+    },
+    {
+      "epoch": 0.36244048099427,
+      "grad_norm": 0.7894664406776428,
+      "learning_rate": 0.00017692034378679315,
+      "loss": 2.7274,
+      "step": 4491
+    },
+    {
+      "epoch": 0.36252118473085304,
+      "grad_norm": 0.7206711173057556,
+      "learning_rate": 0.00017691025493799743,
+      "loss": 2.7047,
+      "step": 4492
+    },
+    {
+      "epoch": 0.362601888467436,
+      "grad_norm": 0.7656282186508179,
+      "learning_rate": 0.00017690016417239708,
+      "loss": 2.696,
+      "step": 4493
+    },
+    {
+      "epoch": 0.36268259220401905,
+      "grad_norm": 0.7357437610626221,
+      "learning_rate": 0.00017689007149024362,
+      "loss": 2.7279,
+      "step": 4494
+    },
+    {
+      "epoch": 0.362763295940602,
+      "grad_norm": 0.7262146472930908,
+      "learning_rate": 0.00017687997689178864,
+      "loss": 2.6964,
+      "step": 4495
+    },
+    {
+      "epoch": 0.36284399967718506,
+      "grad_norm": 0.7839891910552979,
+      "learning_rate": 0.00017686988037728365,
+      "loss": 2.651,
+      "step": 4496
+    },
+    {
+      "epoch": 0.36292470341376803,
+      "grad_norm": 0.7150306105613708,
+      "learning_rate": 0.00017685978194698028,
+      "loss": 2.6481,
+      "step": 4497
+    },
+    {
+      "epoch": 0.36300540715035107,
+      "grad_norm": 0.7144685387611389,
+      "learning_rate": 0.00017684968160113025,
+      "loss": 2.7169,
+      "step": 4498
+    },
+    {
+      "epoch": 0.36308611088693404,
+      "grad_norm": 0.7593061327934265,
+      "learning_rate": 0.00017683957933998525,
+      "loss": 2.7543,
+      "step": 4499
+    },
+    {
+      "epoch": 0.3631668146235171,
+      "grad_norm": 0.7301446199417114,
+      "learning_rate": 0.00017682947516379707,
+      "loss": 2.6806,
+      "step": 4500
+    },
+    {
+      "epoch": 0.36324751836010005,
+      "grad_norm": 0.7314243316650391,
+      "learning_rate": 0.00017681936907281757,
+      "loss": 2.7227,
+      "step": 4501
+    },
+    {
+      "epoch": 0.3633282220966831,
+      "grad_norm": 0.7695817351341248,
+      "learning_rate": 0.00017680926106729852,
+      "loss": 2.7229,
+      "step": 4502
+    },
+    {
+      "epoch": 0.36340892583326606,
+      "grad_norm": 0.6885762810707092,
+      "learning_rate": 0.00017679915114749198,
+      "loss": 2.7246,
+      "step": 4503
+    },
+    {
+      "epoch": 0.3634896295698491,
+      "grad_norm": 0.6893608570098877,
+      "learning_rate": 0.0001767890393136498,
+      "loss": 2.6572,
+      "step": 4504
+    },
+    {
+      "epoch": 0.36357033330643207,
+      "grad_norm": 0.7011978626251221,
+      "learning_rate": 0.00017677892556602402,
+      "loss": 2.6775,
+      "step": 4505
+    },
+    {
+      "epoch": 0.3636510370430151,
+      "grad_norm": 0.6693406105041504,
+      "learning_rate": 0.00017676880990486672,
+      "loss": 2.6183,
+      "step": 4506
+    },
+    {
+      "epoch": 0.3637317407795981,
+      "grad_norm": 0.7023048996925354,
+      "learning_rate": 0.00017675869233043002,
+      "loss": 2.6772,
+      "step": 4507
+    },
+    {
+      "epoch": 0.3638124445161811,
+      "grad_norm": 0.6903806328773499,
+      "learning_rate": 0.00017674857284296605,
+      "loss": 2.6486,
+      "step": 4508
+    },
+    {
+      "epoch": 0.3638931482527641,
+      "grad_norm": 0.6799258589744568,
+      "learning_rate": 0.000176738451442727,
+      "loss": 2.6305,
+      "step": 4509
+    },
+    {
+      "epoch": 0.3639738519893471,
+      "grad_norm": 0.7935682535171509,
+      "learning_rate": 0.00017672832812996517,
+      "loss": 2.7365,
+      "step": 4510
+    },
+    {
+      "epoch": 0.3640545557259301,
+      "grad_norm": 0.7593684196472168,
+      "learning_rate": 0.00017671820290493284,
+      "loss": 2.7029,
+      "step": 4511
+    },
+    {
+      "epoch": 0.36413525946251313,
+      "grad_norm": 0.7185288667678833,
+      "learning_rate": 0.00017670807576788234,
+      "loss": 2.6646,
+      "step": 4512
+    },
+    {
+      "epoch": 0.3642159631990961,
+      "grad_norm": 0.7260291576385498,
+      "learning_rate": 0.00017669794671906606,
+      "loss": 2.6615,
+      "step": 4513
+    },
+    {
+      "epoch": 0.36429666693567914,
+      "grad_norm": 0.6933417916297913,
+      "learning_rate": 0.00017668781575873646,
+      "loss": 2.6678,
+      "step": 4514
+    },
+    {
+      "epoch": 0.3643773706722621,
+      "grad_norm": 0.7657343149185181,
+      "learning_rate": 0.00017667768288714603,
+      "loss": 2.7155,
+      "step": 4515
+    },
+    {
+      "epoch": 0.36445807440884515,
+      "grad_norm": 0.7326949834823608,
+      "learning_rate": 0.0001766675481045473,
+      "loss": 2.732,
+      "step": 4516
+    },
+    {
+      "epoch": 0.3645387781454281,
+      "grad_norm": 0.7370324730873108,
+      "learning_rate": 0.0001766574114111929,
+      "loss": 2.6124,
+      "step": 4517
+    },
+    {
+      "epoch": 0.36461948188201115,
+      "grad_norm": 0.7280072569847107,
+      "learning_rate": 0.00017664727280733536,
+      "loss": 2.6793,
+      "step": 4518
+    },
+    {
+      "epoch": 0.36470018561859413,
+      "grad_norm": 0.7174237370491028,
+      "learning_rate": 0.00017663713229322748,
+      "loss": 2.629,
+      "step": 4519
+    },
+    {
+      "epoch": 0.36478088935517716,
+      "grad_norm": 0.6660771369934082,
+      "learning_rate": 0.0001766269898691219,
+      "loss": 2.6862,
+      "step": 4520
+    },
+    {
+      "epoch": 0.36486159309176014,
+      "grad_norm": 0.7024446725845337,
+      "learning_rate": 0.00017661684553527143,
+      "loss": 2.6602,
+      "step": 4521
+    },
+    {
+      "epoch": 0.36494229682834317,
+      "grad_norm": 0.7419618964195251,
+      "learning_rate": 0.0001766066992919289,
+      "loss": 2.6904,
+      "step": 4522
+    },
+    {
+      "epoch": 0.36502300056492615,
+      "grad_norm": 0.7425804138183594,
+      "learning_rate": 0.00017659655113934716,
+      "loss": 2.7312,
+      "step": 4523
+    },
+    {
+      "epoch": 0.3651037043015092,
+      "grad_norm": 0.7117013931274414,
+      "learning_rate": 0.00017658640107777915,
+      "loss": 2.6411,
+      "step": 4524
+    },
+    {
+      "epoch": 0.36518440803809216,
+      "grad_norm": 0.719613254070282,
+      "learning_rate": 0.00017657624910747782,
+      "loss": 2.6799,
+      "step": 4525
+    },
+    {
+      "epoch": 0.3652651117746752,
+      "grad_norm": 0.7654159665107727,
+      "learning_rate": 0.0001765660952286962,
+      "loss": 2.6675,
+      "step": 4526
+    },
+    {
+      "epoch": 0.36534581551125817,
+      "grad_norm": 0.7111814022064209,
+      "learning_rate": 0.00017655593944168734,
+      "loss": 2.6717,
+      "step": 4527
+    },
+    {
+      "epoch": 0.3654265192478412,
+      "grad_norm": 0.7494712471961975,
+      "learning_rate": 0.00017654578174670436,
+      "loss": 2.7181,
+      "step": 4528
+    },
+    {
+      "epoch": 0.3655072229844242,
+      "grad_norm": 0.8062291145324707,
+      "learning_rate": 0.0001765356221440004,
+      "loss": 2.6563,
+      "step": 4529
+    },
+    {
+      "epoch": 0.3655879267210072,
+      "grad_norm": 0.7923303842544556,
+      "learning_rate": 0.00017652546063382866,
+      "loss": 2.6295,
+      "step": 4530
+    },
+    {
+      "epoch": 0.3656686304575902,
+      "grad_norm": 0.7417340278625488,
+      "learning_rate": 0.00017651529721644238,
+      "loss": 2.6727,
+      "step": 4531
+    },
+    {
+      "epoch": 0.3657493341941732,
+      "grad_norm": 0.7326166033744812,
+      "learning_rate": 0.0001765051318920949,
+      "loss": 2.702,
+      "step": 4532
+    },
+    {
+      "epoch": 0.3658300379307562,
+      "grad_norm": 0.8133745193481445,
+      "learning_rate": 0.00017649496466103957,
+      "loss": 2.7157,
+      "step": 4533
+    },
+    {
+      "epoch": 0.3659107416673392,
+      "grad_norm": 0.710502564907074,
+      "learning_rate": 0.00017648479552352973,
+      "loss": 2.6668,
+      "step": 4534
+    },
+    {
+      "epoch": 0.3659914454039222,
+      "grad_norm": 0.6947012543678284,
+      "learning_rate": 0.00017647462447981885,
+      "loss": 2.6865,
+      "step": 4535
+    },
+    {
+      "epoch": 0.36607214914050523,
+      "grad_norm": 0.8432720899581909,
+      "learning_rate": 0.0001764644515301604,
+      "loss": 2.6226,
+      "step": 4536
+    },
+    {
+      "epoch": 0.3661528528770882,
+      "grad_norm": 0.7321269512176514,
+      "learning_rate": 0.00017645427667480802,
+      "loss": 2.662,
+      "step": 4537
+    },
+    {
+      "epoch": 0.36623355661367124,
+      "grad_norm": 0.8099743723869324,
+      "learning_rate": 0.00017644409991401515,
+      "loss": 2.6853,
+      "step": 4538
+    },
+    {
+      "epoch": 0.3663142603502542,
+      "grad_norm": 0.6885355114936829,
+      "learning_rate": 0.0001764339212480355,
+      "loss": 2.6672,
+      "step": 4539
+    },
+    {
+      "epoch": 0.3663949640868372,
+      "grad_norm": 0.911396324634552,
+      "learning_rate": 0.00017642374067712276,
+      "loss": 2.5778,
+      "step": 4540
+    },
+    {
+      "epoch": 0.3664756678234202,
+      "grad_norm": 0.7461941838264465,
+      "learning_rate": 0.0001764135582015306,
+      "loss": 2.6629,
+      "step": 4541
+    },
+    {
+      "epoch": 0.3665563715600032,
+      "grad_norm": 0.772741436958313,
+      "learning_rate": 0.0001764033738215128,
+      "loss": 2.725,
+      "step": 4542
+    },
+    {
+      "epoch": 0.36663707529658623,
+      "grad_norm": 0.7256152629852295,
+      "learning_rate": 0.0001763931875373232,
+      "loss": 2.6439,
+      "step": 4543
+    },
+    {
+      "epoch": 0.3667177790331692,
+      "grad_norm": 0.8089167475700378,
+      "learning_rate": 0.0001763829993492157,
+      "loss": 2.5972,
+      "step": 4544
+    },
+    {
+      "epoch": 0.36679848276975224,
+      "grad_norm": 0.7115232944488525,
+      "learning_rate": 0.0001763728092574442,
+      "loss": 2.633,
+      "step": 4545
+    },
+    {
+      "epoch": 0.3668791865063352,
+      "grad_norm": 0.7189347147941589,
+      "learning_rate": 0.00017636261726226266,
+      "loss": 2.619,
+      "step": 4546
+    },
+    {
+      "epoch": 0.36695989024291825,
+      "grad_norm": 0.7667742967605591,
+      "learning_rate": 0.00017635242336392506,
+      "loss": 2.667,
+      "step": 4547
+    },
+    {
+      "epoch": 0.36704059397950123,
+      "grad_norm": 0.7982457876205444,
+      "learning_rate": 0.00017634222756268545,
+      "loss": 2.6667,
+      "step": 4548
+    },
+    {
+      "epoch": 0.36712129771608426,
+      "grad_norm": 0.7465574145317078,
+      "learning_rate": 0.00017633202985879804,
+      "loss": 2.6436,
+      "step": 4549
+    },
+    {
+      "epoch": 0.36720200145266724,
+      "grad_norm": 0.7297804951667786,
+      "learning_rate": 0.00017632183025251686,
+      "loss": 2.6464,
+      "step": 4550
+    },
+    {
+      "epoch": 0.36728270518925027,
+      "grad_norm": 0.6885054111480713,
+      "learning_rate": 0.0001763116287440962,
+      "loss": 2.6742,
+      "step": 4551
+    },
+    {
+      "epoch": 0.36736340892583325,
+      "grad_norm": 0.7341574430465698,
+      "learning_rate": 0.00017630142533379023,
+      "loss": 2.6688,
+      "step": 4552
+    },
+    {
+      "epoch": 0.3674441126624163,
+      "grad_norm": 0.8565430045127869,
+      "learning_rate": 0.0001762912200218533,
+      "loss": 2.6889,
+      "step": 4553
+    },
+    {
+      "epoch": 0.36752481639899925,
+      "grad_norm": 0.7509489059448242,
+      "learning_rate": 0.00017628101280853974,
+      "loss": 2.6177,
+      "step": 4554
+    },
+    {
+      "epoch": 0.3676055201355823,
+      "grad_norm": 0.8128334879875183,
+      "learning_rate": 0.00017627080369410396,
+      "loss": 2.7301,
+      "step": 4555
+    },
+    {
+      "epoch": 0.36768622387216526,
+      "grad_norm": 0.7511637210845947,
+      "learning_rate": 0.00017626059267880035,
+      "loss": 2.7327,
+      "step": 4556
+    },
+    {
+      "epoch": 0.3677669276087483,
+      "grad_norm": 0.8350822925567627,
+      "learning_rate": 0.00017625037976288347,
+      "loss": 2.6073,
+      "step": 4557
+    },
+    {
+      "epoch": 0.36784763134533127,
+      "grad_norm": 0.7743313312530518,
+      "learning_rate": 0.00017624016494660776,
+      "loss": 2.7055,
+      "step": 4558
+    },
+    {
+      "epoch": 0.3679283350819143,
+      "grad_norm": 0.8196439146995544,
+      "learning_rate": 0.00017622994823022787,
+      "loss": 2.6565,
+      "step": 4559
+    },
+    {
+      "epoch": 0.3680090388184973,
+      "grad_norm": 0.7223393321037292,
+      "learning_rate": 0.00017621972961399837,
+      "loss": 2.68,
+      "step": 4560
+    },
+    {
+      "epoch": 0.3680897425550803,
+      "grad_norm": 0.7215418219566345,
+      "learning_rate": 0.000176209509098174,
+      "loss": 2.6627,
+      "step": 4561
+    },
+    {
+      "epoch": 0.3681704462916633,
+      "grad_norm": 0.8050473928451538,
+      "learning_rate": 0.00017619928668300946,
+      "loss": 2.5802,
+      "step": 4562
+    },
+    {
+      "epoch": 0.3682511500282463,
+      "grad_norm": 0.7452750205993652,
+      "learning_rate": 0.00017618906236875948,
+      "loss": 2.6524,
+      "step": 4563
+    },
+    {
+      "epoch": 0.3683318537648293,
+      "grad_norm": 0.7950742244720459,
+      "learning_rate": 0.00017617883615567888,
+      "loss": 2.6371,
+      "step": 4564
+    },
+    {
+      "epoch": 0.36841255750141233,
+      "grad_norm": 0.7185397744178772,
+      "learning_rate": 0.00017616860804402261,
+      "loss": 2.6531,
+      "step": 4565
+    },
+    {
+      "epoch": 0.3684932612379953,
+      "grad_norm": 0.7480553388595581,
+      "learning_rate": 0.0001761583780340455,
+      "loss": 2.6727,
+      "step": 4566
+    },
+    {
+      "epoch": 0.36857396497457834,
+      "grad_norm": 0.7740724086761475,
+      "learning_rate": 0.00017614814612600251,
+      "loss": 2.6095,
+      "step": 4567
+    },
+    {
+      "epoch": 0.3686546687111613,
+      "grad_norm": 0.9159810543060303,
+      "learning_rate": 0.00017613791232014866,
+      "loss": 2.7039,
+      "step": 4568
+    },
+    {
+      "epoch": 0.36873537244774435,
+      "grad_norm": 0.7478305697441101,
+      "learning_rate": 0.00017612767661673905,
+      "loss": 2.6307,
+      "step": 4569
+    },
+    {
+      "epoch": 0.3688160761843273,
+      "grad_norm": 0.9154726266860962,
+      "learning_rate": 0.00017611743901602874,
+      "loss": 2.675,
+      "step": 4570
+    },
+    {
+      "epoch": 0.36889677992091036,
+      "grad_norm": 0.7903287410736084,
+      "learning_rate": 0.0001761071995182728,
+      "loss": 2.6938,
+      "step": 4571
+    },
+    {
+      "epoch": 0.36897748365749333,
+      "grad_norm": 0.7919119596481323,
+      "learning_rate": 0.0001760969581237266,
+      "loss": 2.7092,
+      "step": 4572
+    },
+    {
+      "epoch": 0.36905818739407636,
+      "grad_norm": 0.8052253723144531,
+      "learning_rate": 0.00017608671483264522,
+      "loss": 2.6914,
+      "step": 4573
+    },
+    {
+      "epoch": 0.36913889113065934,
+      "grad_norm": 0.7660435438156128,
+      "learning_rate": 0.00017607646964528403,
+      "loss": 2.674,
+      "step": 4574
+    },
+    {
+      "epoch": 0.3692195948672424,
+      "grad_norm": 0.8554383516311646,
+      "learning_rate": 0.00017606622256189836,
+      "loss": 2.6792,
+      "step": 4575
+    },
+    {
+      "epoch": 0.36930029860382535,
+      "grad_norm": 0.7719140648841858,
+      "learning_rate": 0.00017605597358274358,
+      "loss": 2.6836,
+      "step": 4576
+    },
+    {
+      "epoch": 0.3693810023404084,
+      "grad_norm": 0.733068585395813,
+      "learning_rate": 0.00017604572270807513,
+      "loss": 2.6496,
+      "step": 4577
+    },
+    {
+      "epoch": 0.36946170607699136,
+      "grad_norm": 0.7622445225715637,
+      "learning_rate": 0.00017603546993814849,
+      "loss": 2.7097,
+      "step": 4578
+    },
+    {
+      "epoch": 0.3695424098135744,
+      "grad_norm": 0.7326679825782776,
+      "learning_rate": 0.00017602521527321913,
+      "loss": 2.6786,
+      "step": 4579
+    },
+    {
+      "epoch": 0.36962311355015737,
+      "grad_norm": 0.7579432129859924,
+      "learning_rate": 0.00017601495871354272,
+      "loss": 2.6618,
+      "step": 4580
+    },
+    {
+      "epoch": 0.3697038172867404,
+      "grad_norm": 0.8812715411186218,
+      "learning_rate": 0.00017600470025937485,
+      "loss": 2.6942,
+      "step": 4581
+    },
+    {
+      "epoch": 0.3697845210233234,
+      "grad_norm": 0.7230449318885803,
+      "learning_rate": 0.00017599443991097116,
+      "loss": 2.6374,
+      "step": 4582
+    },
+    {
+      "epoch": 0.3698652247599064,
+      "grad_norm": 0.8347739577293396,
+      "learning_rate": 0.00017598417766858735,
+      "loss": 2.6653,
+      "step": 4583
+    },
+    {
+      "epoch": 0.3699459284964894,
+      "grad_norm": 0.7826598882675171,
+      "learning_rate": 0.0001759739135324792,
+      "loss": 2.6342,
+      "step": 4584
+    },
+    {
+      "epoch": 0.3700266322330724,
+      "grad_norm": 0.749060332775116,
+      "learning_rate": 0.00017596364750290254,
+      "loss": 2.7256,
+      "step": 4585
+    },
+    {
+      "epoch": 0.3701073359696554,
+      "grad_norm": 0.7470815181732178,
+      "learning_rate": 0.00017595337958011323,
+      "loss": 2.6485,
+      "step": 4586
+    },
+    {
+      "epoch": 0.3701880397062384,
+      "grad_norm": 0.7251530289649963,
+      "learning_rate": 0.00017594310976436716,
+      "loss": 2.6613,
+      "step": 4587
+    },
+    {
+      "epoch": 0.3702687434428214,
+      "grad_norm": 0.7143718004226685,
+      "learning_rate": 0.00017593283805592027,
+      "loss": 2.6101,
+      "step": 4588
+    },
+    {
+      "epoch": 0.37034944717940443,
+      "grad_norm": 0.7378203272819519,
+      "learning_rate": 0.00017592256445502855,
+      "loss": 2.6735,
+      "step": 4589
+    },
+    {
+      "epoch": 0.3704301509159874,
+      "grad_norm": 0.7193629741668701,
+      "learning_rate": 0.00017591228896194808,
+      "loss": 2.719,
+      "step": 4590
+    },
+    {
+      "epoch": 0.3705108546525704,
+      "grad_norm": 0.7377258539199829,
+      "learning_rate": 0.00017590201157693494,
+      "loss": 2.6789,
+      "step": 4591
+    },
+    {
+      "epoch": 0.3705915583891534,
+      "grad_norm": 0.7468351721763611,
+      "learning_rate": 0.00017589173230024522,
+      "loss": 2.6389,
+      "step": 4592
+    },
+    {
+      "epoch": 0.3706722621257364,
+      "grad_norm": 0.7612246870994568,
+      "learning_rate": 0.0001758814511321352,
+      "loss": 2.7045,
+      "step": 4593
+    },
+    {
+      "epoch": 0.37075296586231943,
+      "grad_norm": 0.7603838443756104,
+      "learning_rate": 0.00017587116807286102,
+      "loss": 2.7323,
+      "step": 4594
+    },
+    {
+      "epoch": 0.3708336695989024,
+      "grad_norm": 0.7436477541923523,
+      "learning_rate": 0.000175860883122679,
+      "loss": 2.7331,
+      "step": 4595
+    },
+    {
+      "epoch": 0.37091437333548544,
+      "grad_norm": 0.7004369497299194,
+      "learning_rate": 0.0001758505962818455,
+      "loss": 2.6418,
+      "step": 4596
+    },
+    {
+      "epoch": 0.3709950770720684,
+      "grad_norm": 0.711980938911438,
+      "learning_rate": 0.00017584030755061683,
+      "loss": 2.6184,
+      "step": 4597
+    },
+    {
+      "epoch": 0.37107578080865145,
+      "grad_norm": 0.6999367475509644,
+      "learning_rate": 0.0001758300169292495,
+      "loss": 2.6584,
+      "step": 4598
+    },
+    {
+      "epoch": 0.3711564845452344,
+      "grad_norm": 0.6755785942077637,
+      "learning_rate": 0.0001758197244179999,
+      "loss": 2.664,
+      "step": 4599
+    },
+    {
+      "epoch": 0.37123718828181745,
+      "grad_norm": 0.7174055576324463,
+      "learning_rate": 0.00017580943001712455,
+      "loss": 2.6821,
+      "step": 4600
+    },
+    {
+      "epoch": 0.37131789201840043,
+      "grad_norm": 0.8218933343887329,
+      "learning_rate": 0.00017579913372688005,
+      "loss": 2.6355,
+      "step": 4601
+    },
+    {
+      "epoch": 0.37139859575498346,
+      "grad_norm": 0.7417960166931152,
+      "learning_rate": 0.000175788835547523,
+      "loss": 2.7226,
+      "step": 4602
+    },
+    {
+      "epoch": 0.37147929949156644,
+      "grad_norm": 0.824421763420105,
+      "learning_rate": 0.00017577853547931006,
+      "loss": 2.6526,
+      "step": 4603
+    },
+    {
+      "epoch": 0.37156000322814947,
+      "grad_norm": 0.7391949892044067,
+      "learning_rate": 0.00017576823352249794,
+      "loss": 2.6702,
+      "step": 4604
+    },
+    {
+      "epoch": 0.37164070696473245,
+      "grad_norm": 0.7890247106552124,
+      "learning_rate": 0.00017575792967734337,
+      "loss": 2.7281,
+      "step": 4605
+    },
+    {
+      "epoch": 0.3717214107013155,
+      "grad_norm": 0.785527765750885,
+      "learning_rate": 0.00017574762394410317,
+      "loss": 2.6728,
+      "step": 4606
+    },
+    {
+      "epoch": 0.37180211443789846,
+      "grad_norm": 0.7195863127708435,
+      "learning_rate": 0.00017573731632303415,
+      "loss": 2.6329,
+      "step": 4607
+    },
+    {
+      "epoch": 0.3718828181744815,
+      "grad_norm": 0.7896780371665955,
+      "learning_rate": 0.0001757270068143932,
+      "loss": 2.6776,
+      "step": 4608
+    },
+    {
+      "epoch": 0.37196352191106447,
+      "grad_norm": 0.7568275332450867,
+      "learning_rate": 0.00017571669541843735,
+      "loss": 2.6668,
+      "step": 4609
+    },
+    {
+      "epoch": 0.3720442256476475,
+      "grad_norm": 0.7923939228057861,
+      "learning_rate": 0.00017570638213542348,
+      "loss": 2.7033,
+      "step": 4610
+    },
+    {
+      "epoch": 0.3721249293842305,
+      "grad_norm": 0.7586569786071777,
+      "learning_rate": 0.00017569606696560868,
+      "loss": 2.7286,
+      "step": 4611
+    },
+    {
+      "epoch": 0.3722056331208135,
+      "grad_norm": 0.8222009539604187,
+      "learning_rate": 0.00017568574990925004,
+      "loss": 2.6448,
+      "step": 4612
+    },
+    {
+      "epoch": 0.3722863368573965,
+      "grad_norm": 0.7144019603729248,
+      "learning_rate": 0.00017567543096660466,
+      "loss": 2.6671,
+      "step": 4613
+    },
+    {
+      "epoch": 0.3723670405939795,
+      "grad_norm": 0.7602240443229675,
+      "learning_rate": 0.00017566511013792973,
+      "loss": 2.6492,
+      "step": 4614
+    },
+    {
+      "epoch": 0.3724477443305625,
+      "grad_norm": 0.7949689626693726,
+      "learning_rate": 0.00017565478742348245,
+      "loss": 2.7002,
+      "step": 4615
+    },
+    {
+      "epoch": 0.3725284480671455,
+      "grad_norm": 0.6922519207000732,
+      "learning_rate": 0.00017564446282352012,
+      "loss": 2.6917,
+      "step": 4616
+    },
+    {
+      "epoch": 0.3726091518037285,
+      "grad_norm": 0.7382915616035461,
+      "learning_rate": 0.0001756341363383,
+      "loss": 2.6375,
+      "step": 4617
+    },
+    {
+      "epoch": 0.37268985554031153,
+      "grad_norm": 0.7511888742446899,
+      "learning_rate": 0.00017562380796807956,
+      "loss": 2.6823,
+      "step": 4618
+    },
+    {
+      "epoch": 0.3727705592768945,
+      "grad_norm": 0.7273457646369934,
+      "learning_rate": 0.00017561347771311608,
+      "loss": 2.6124,
+      "step": 4619
+    },
+    {
+      "epoch": 0.37285126301347754,
+      "grad_norm": 0.689440131187439,
+      "learning_rate": 0.0001756031455736671,
+      "loss": 2.6931,
+      "step": 4620
+    },
+    {
+      "epoch": 0.3729319667500605,
+      "grad_norm": 0.7755659222602844,
+      "learning_rate": 0.00017559281154999013,
+      "loss": 2.6273,
+      "step": 4621
+    },
+    {
+      "epoch": 0.37301267048664355,
+      "grad_norm": 0.6940193176269531,
+      "learning_rate": 0.00017558247564234265,
+      "loss": 2.641,
+      "step": 4622
+    },
+    {
+      "epoch": 0.3730933742232265,
+      "grad_norm": 0.7387529015541077,
+      "learning_rate": 0.00017557213785098232,
+      "loss": 2.7229,
+      "step": 4623
+    },
+    {
+      "epoch": 0.37317407795980956,
+      "grad_norm": 0.6807727217674255,
+      "learning_rate": 0.00017556179817616678,
+      "loss": 2.6469,
+      "step": 4624
+    },
+    {
+      "epoch": 0.37325478169639253,
+      "grad_norm": 0.7203819751739502,
+      "learning_rate": 0.0001755514566181537,
+      "loss": 2.6239,
+      "step": 4625
+    },
+    {
+      "epoch": 0.37333548543297557,
+      "grad_norm": 0.9345876574516296,
+      "learning_rate": 0.0001755411131772008,
+      "loss": 2.7154,
+      "step": 4626
+    },
+    {
+      "epoch": 0.37341618916955854,
+      "grad_norm": 0.6787357330322266,
+      "learning_rate": 0.00017553076785356594,
+      "loss": 2.6374,
+      "step": 4627
+    },
+    {
+      "epoch": 0.3734968929061416,
+      "grad_norm": 0.7153670191764832,
+      "learning_rate": 0.0001755204206475069,
+      "loss": 2.6734,
+      "step": 4628
+    },
+    {
+      "epoch": 0.37357759664272455,
+      "grad_norm": 0.736464262008667,
+      "learning_rate": 0.00017551007155928154,
+      "loss": 2.7241,
+      "step": 4629
+    },
+    {
+      "epoch": 0.3736583003793076,
+      "grad_norm": 0.7134939432144165,
+      "learning_rate": 0.0001754997205891478,
+      "loss": 2.682,
+      "step": 4630
+    },
+    {
+      "epoch": 0.37373900411589056,
+      "grad_norm": 0.7071199417114258,
+      "learning_rate": 0.0001754893677373637,
+      "loss": 2.7361,
+      "step": 4631
+    },
+    {
+      "epoch": 0.3738197078524736,
+      "grad_norm": 0.7040621638298035,
+      "learning_rate": 0.00017547901300418722,
+      "loss": 2.7031,
+      "step": 4632
+    },
+    {
+      "epoch": 0.37390041158905657,
+      "grad_norm": 0.7179287075996399,
+      "learning_rate": 0.00017546865638987642,
+      "loss": 2.6755,
+      "step": 4633
+    },
+    {
+      "epoch": 0.3739811153256396,
+      "grad_norm": 0.7579259276390076,
+      "learning_rate": 0.00017545829789468944,
+      "loss": 2.6514,
+      "step": 4634
+    },
+    {
+      "epoch": 0.3740618190622226,
+      "grad_norm": 0.7825835347175598,
+      "learning_rate": 0.0001754479375188844,
+      "loss": 2.6876,
+      "step": 4635
+    },
+    {
+      "epoch": 0.3741425227988056,
+      "grad_norm": 0.7913421988487244,
+      "learning_rate": 0.00017543757526271956,
+      "loss": 2.7153,
+      "step": 4636
+    },
+    {
+      "epoch": 0.3742232265353886,
+      "grad_norm": 0.7766042947769165,
+      "learning_rate": 0.00017542721112645313,
+      "loss": 2.645,
+      "step": 4637
+    },
+    {
+      "epoch": 0.3743039302719716,
+      "grad_norm": 0.7363953590393066,
+      "learning_rate": 0.00017541684511034343,
+      "loss": 2.6376,
+      "step": 4638
+    },
+    {
+      "epoch": 0.3743846340085546,
+      "grad_norm": 0.6928617358207703,
+      "learning_rate": 0.00017540647721464881,
+      "loss": 2.6882,
+      "step": 4639
+    },
+    {
+      "epoch": 0.3744653377451376,
+      "grad_norm": 0.7832257747650146,
+      "learning_rate": 0.0001753961074396277,
+      "loss": 2.7305,
+      "step": 4640
+    },
+    {
+      "epoch": 0.3745460414817206,
+      "grad_norm": 0.7180350422859192,
+      "learning_rate": 0.00017538573578553844,
+      "loss": 2.6783,
+      "step": 4641
+    },
+    {
+      "epoch": 0.3746267452183036,
+      "grad_norm": 0.718209981918335,
+      "learning_rate": 0.00017537536225263964,
+      "loss": 2.6961,
+      "step": 4642
+    },
+    {
+      "epoch": 0.3747074489548866,
+      "grad_norm": 0.7056655287742615,
+      "learning_rate": 0.00017536498684118975,
+      "loss": 2.7096,
+      "step": 4643
+    },
+    {
+      "epoch": 0.3747881526914696,
+      "grad_norm": 0.8004828691482544,
+      "learning_rate": 0.0001753546095514474,
+      "loss": 2.7168,
+      "step": 4644
+    },
+    {
+      "epoch": 0.3748688564280526,
+      "grad_norm": 0.7630821466445923,
+      "learning_rate": 0.0001753442303836712,
+      "loss": 2.7091,
+      "step": 4645
+    },
+    {
+      "epoch": 0.3749495601646356,
+      "grad_norm": 0.7539668083190918,
+      "learning_rate": 0.0001753338493381198,
+      "loss": 2.651,
+      "step": 4646
+    },
+    {
+      "epoch": 0.37503026390121863,
+      "grad_norm": 0.7243319749832153,
+      "learning_rate": 0.000175323466415052,
+      "loss": 2.6765,
+      "step": 4647
+    },
+    {
+      "epoch": 0.3751109676378016,
+      "grad_norm": 0.8906281590461731,
+      "learning_rate": 0.00017531308161472647,
+      "loss": 2.5938,
+      "step": 4648
+    },
+    {
+      "epoch": 0.37519167137438464,
+      "grad_norm": 0.787966251373291,
+      "learning_rate": 0.0001753026949374021,
+      "loss": 2.6011,
+      "step": 4649
+    },
+    {
+      "epoch": 0.3752723751109676,
+      "grad_norm": 0.7763915061950684,
+      "learning_rate": 0.00017529230638333772,
+      "loss": 2.7197,
+      "step": 4650
+    },
+    {
+      "epoch": 0.37535307884755065,
+      "grad_norm": 0.7717103362083435,
+      "learning_rate": 0.00017528191595279224,
+      "loss": 2.6605,
+      "step": 4651
+    },
+    {
+      "epoch": 0.3754337825841336,
+      "grad_norm": 0.7340055108070374,
+      "learning_rate": 0.00017527152364602464,
+      "loss": 2.6856,
+      "step": 4652
+    },
+    {
+      "epoch": 0.37551448632071666,
+      "grad_norm": 0.7805169820785522,
+      "learning_rate": 0.0001752611294632939,
+      "loss": 2.7088,
+      "step": 4653
+    },
+    {
+      "epoch": 0.37559519005729963,
+      "grad_norm": 0.7894891500473022,
+      "learning_rate": 0.00017525073340485912,
+      "loss": 2.6691,
+      "step": 4654
+    },
+    {
+      "epoch": 0.37567589379388266,
+      "grad_norm": 0.7627872824668884,
+      "learning_rate": 0.0001752403354709793,
+      "loss": 2.6536,
+      "step": 4655
+    },
+    {
+      "epoch": 0.37575659753046564,
+      "grad_norm": 0.8097225427627563,
+      "learning_rate": 0.00017522993566191367,
+      "loss": 2.7108,
+      "step": 4656
+    },
+    {
+      "epoch": 0.3758373012670487,
+      "grad_norm": 0.834449827671051,
+      "learning_rate": 0.00017521953397792137,
+      "loss": 2.7565,
+      "step": 4657
+    },
+    {
+      "epoch": 0.37591800500363165,
+      "grad_norm": 0.7924147844314575,
+      "learning_rate": 0.00017520913041926166,
+      "loss": 2.7101,
+      "step": 4658
+    },
+    {
+      "epoch": 0.3759987087402147,
+      "grad_norm": 0.7407249808311462,
+      "learning_rate": 0.00017519872498619385,
+      "loss": 2.6501,
+      "step": 4659
+    },
+    {
+      "epoch": 0.37607941247679766,
+      "grad_norm": 0.7251791954040527,
+      "learning_rate": 0.0001751883176789772,
+      "loss": 2.6786,
+      "step": 4660
+    },
+    {
+      "epoch": 0.3761601162133807,
+      "grad_norm": 0.7120431661605835,
+      "learning_rate": 0.00017517790849787116,
+      "loss": 2.7244,
+      "step": 4661
+    },
+    {
+      "epoch": 0.37624081994996367,
+      "grad_norm": 0.724836528301239,
+      "learning_rate": 0.00017516749744313513,
+      "loss": 2.7099,
+      "step": 4662
+    },
+    {
+      "epoch": 0.3763215236865467,
+      "grad_norm": 0.7788939476013184,
+      "learning_rate": 0.00017515708451502855,
+      "loss": 2.6206,
+      "step": 4663
+    },
+    {
+      "epoch": 0.3764022274231297,
+      "grad_norm": 0.7518914341926575,
+      "learning_rate": 0.00017514666971381099,
+      "loss": 2.7505,
+      "step": 4664
+    },
+    {
+      "epoch": 0.3764829311597127,
+      "grad_norm": 0.8004730939865112,
+      "learning_rate": 0.00017513625303974194,
+      "loss": 2.6119,
+      "step": 4665
+    },
+    {
+      "epoch": 0.3765636348962957,
+      "grad_norm": 0.7661109566688538,
+      "learning_rate": 0.00017512583449308107,
+      "loss": 2.724,
+      "step": 4666
+    },
+    {
+      "epoch": 0.3766443386328787,
+      "grad_norm": 0.7669692635536194,
+      "learning_rate": 0.00017511541407408805,
+      "loss": 2.7109,
+      "step": 4667
+    },
+    {
+      "epoch": 0.3767250423694617,
+      "grad_norm": 0.738608181476593,
+      "learning_rate": 0.00017510499178302253,
+      "loss": 2.6642,
+      "step": 4668
+    },
+    {
+      "epoch": 0.3768057461060447,
+      "grad_norm": 0.7194661498069763,
+      "learning_rate": 0.00017509456762014432,
+      "loss": 2.6906,
+      "step": 4669
+    },
+    {
+      "epoch": 0.3768864498426277,
+      "grad_norm": 0.7025040984153748,
+      "learning_rate": 0.00017508414158571314,
+      "loss": 2.6596,
+      "step": 4670
+    },
+    {
+      "epoch": 0.37696715357921073,
+      "grad_norm": 0.7756575345993042,
+      "learning_rate": 0.00017507371367998892,
+      "loss": 2.7114,
+      "step": 4671
+    },
+    {
+      "epoch": 0.3770478573157937,
+      "grad_norm": 0.834966778755188,
+      "learning_rate": 0.00017506328390323148,
+      "loss": 2.7554,
+      "step": 4672
+    },
+    {
+      "epoch": 0.37712856105237674,
+      "grad_norm": 0.6997280120849609,
+      "learning_rate": 0.0001750528522557008,
+      "loss": 2.6285,
+      "step": 4673
+    },
+    {
+      "epoch": 0.3772092647889597,
+      "grad_norm": 0.7101716995239258,
+      "learning_rate": 0.0001750424187376569,
+      "loss": 2.6465,
+      "step": 4674
+    },
+    {
+      "epoch": 0.37728996852554275,
+      "grad_norm": 0.6577222347259521,
+      "learning_rate": 0.0001750319833493597,
+      "loss": 2.6372,
+      "step": 4675
+    },
+    {
+      "epoch": 0.37737067226212573,
+      "grad_norm": 0.7402529120445251,
+      "learning_rate": 0.00017502154609106937,
+      "loss": 2.6464,
+      "step": 4676
+    },
+    {
+      "epoch": 0.37745137599870876,
+      "grad_norm": 0.6858490705490112,
+      "learning_rate": 0.00017501110696304596,
+      "loss": 2.6141,
+      "step": 4677
+    },
+    {
+      "epoch": 0.37753207973529174,
+      "grad_norm": 0.729468822479248,
+      "learning_rate": 0.0001750006659655497,
+      "loss": 2.6671,
+      "step": 4678
+    },
+    {
+      "epoch": 0.37761278347187477,
+      "grad_norm": 0.7197559475898743,
+      "learning_rate": 0.0001749902230988408,
+      "loss": 2.6462,
+      "step": 4679
+    },
+    {
+      "epoch": 0.37769348720845775,
+      "grad_norm": 0.7171144485473633,
+      "learning_rate": 0.00017497977836317957,
+      "loss": 2.6427,
+      "step": 4680
+    },
+    {
+      "epoch": 0.3777741909450408,
+      "grad_norm": 0.7423805594444275,
+      "learning_rate": 0.00017496933175882617,
+      "loss": 2.662,
+      "step": 4681
+    },
+    {
+      "epoch": 0.37785489468162375,
+      "grad_norm": 0.7498061060905457,
+      "learning_rate": 0.0001749588832860411,
+      "loss": 2.6243,
+      "step": 4682
+    },
+    {
+      "epoch": 0.3779355984182068,
+      "grad_norm": 0.7706165909767151,
+      "learning_rate": 0.0001749484329450847,
+      "loss": 2.6928,
+      "step": 4683
+    },
+    {
+      "epoch": 0.37801630215478976,
+      "grad_norm": 0.723363995552063,
+      "learning_rate": 0.00017493798073621745,
+      "loss": 2.6787,
+      "step": 4684
+    },
+    {
+      "epoch": 0.3780970058913728,
+      "grad_norm": 0.7444875836372375,
+      "learning_rate": 0.00017492752665969983,
+      "loss": 2.6789,
+      "step": 4685
+    },
+    {
+      "epoch": 0.37817770962795577,
+      "grad_norm": 0.6946491599082947,
+      "learning_rate": 0.00017491707071579237,
+      "loss": 2.6761,
+      "step": 4686
+    },
+    {
+      "epoch": 0.3782584133645388,
+      "grad_norm": 0.7171412706375122,
+      "learning_rate": 0.00017490661290475568,
+      "loss": 2.6788,
+      "step": 4687
+    },
+    {
+      "epoch": 0.3783391171011218,
+      "grad_norm": 0.7503272891044617,
+      "learning_rate": 0.00017489615322685038,
+      "loss": 2.7057,
+      "step": 4688
+    },
+    {
+      "epoch": 0.3784198208377048,
+      "grad_norm": 0.7458747625350952,
+      "learning_rate": 0.00017488569168233714,
+      "loss": 2.6857,
+      "step": 4689
+    },
+    {
+      "epoch": 0.3785005245742878,
+      "grad_norm": 0.7030516266822815,
+      "learning_rate": 0.0001748752282714768,
+      "loss": 2.6522,
+      "step": 4690
+    },
+    {
+      "epoch": 0.3785812283108708,
+      "grad_norm": 0.7717545628547668,
+      "learning_rate": 0.00017486476299452994,
+      "loss": 2.6527,
+      "step": 4691
+    },
+    {
+      "epoch": 0.3786619320474538,
+      "grad_norm": 0.6788322925567627,
+      "learning_rate": 0.0001748542958517575,
+      "loss": 2.6362,
+      "step": 4692
+    },
+    {
+      "epoch": 0.3787426357840368,
+      "grad_norm": 0.8518630266189575,
+      "learning_rate": 0.0001748438268434204,
+      "loss": 2.6812,
+      "step": 4693
+    },
+    {
+      "epoch": 0.3788233395206198,
+      "grad_norm": 0.7167141437530518,
+      "learning_rate": 0.00017483335596977945,
+      "loss": 2.6414,
+      "step": 4694
+    },
+    {
+      "epoch": 0.3789040432572028,
+      "grad_norm": 0.7748053073883057,
+      "learning_rate": 0.00017482288323109567,
+      "loss": 2.7291,
+      "step": 4695
+    },
+    {
+      "epoch": 0.3789847469937858,
+      "grad_norm": 0.7203041911125183,
+      "learning_rate": 0.00017481240862763002,
+      "loss": 2.6957,
+      "step": 4696
+    },
+    {
+      "epoch": 0.3790654507303688,
+      "grad_norm": 0.7973119020462036,
+      "learning_rate": 0.00017480193215964362,
+      "loss": 2.7456,
+      "step": 4697
+    },
+    {
+      "epoch": 0.3791461544669518,
+      "grad_norm": 0.7851223945617676,
+      "learning_rate": 0.00017479145382739755,
+      "loss": 2.6525,
+      "step": 4698
+    },
+    {
+      "epoch": 0.3792268582035348,
+      "grad_norm": 0.7012068629264832,
+      "learning_rate": 0.0001747809736311529,
+      "loss": 2.6662,
+      "step": 4699
+    },
+    {
+      "epoch": 0.37930756194011783,
+      "grad_norm": 0.7266128659248352,
+      "learning_rate": 0.00017477049157117093,
+      "loss": 2.5853,
+      "step": 4700
+    },
+    {
+      "epoch": 0.3793882656767008,
+      "grad_norm": 0.7264416217803955,
+      "learning_rate": 0.00017476000764771285,
+      "loss": 2.6972,
+      "step": 4701
+    },
+    {
+      "epoch": 0.37946896941328384,
+      "grad_norm": 0.797709047794342,
+      "learning_rate": 0.00017474952186103995,
+      "loss": 2.6997,
+      "step": 4702
+    },
+    {
+      "epoch": 0.3795496731498668,
+      "grad_norm": 0.7552568912506104,
+      "learning_rate": 0.00017473903421141358,
+      "loss": 2.7178,
+      "step": 4703
+    },
+    {
+      "epoch": 0.37963037688644985,
+      "grad_norm": 0.7611108422279358,
+      "learning_rate": 0.0001747285446990951,
+      "loss": 2.6997,
+      "step": 4704
+    },
+    {
+      "epoch": 0.3797110806230328,
+      "grad_norm": 0.8081753253936768,
+      "learning_rate": 0.00017471805332434595,
+      "loss": 2.7242,
+      "step": 4705
+    },
+    {
+      "epoch": 0.37979178435961586,
+      "grad_norm": 0.728301465511322,
+      "learning_rate": 0.0001747075600874276,
+      "loss": 2.5885,
+      "step": 4706
+    },
+    {
+      "epoch": 0.37987248809619883,
+      "grad_norm": 0.7548539638519287,
+      "learning_rate": 0.00017469706498860155,
+      "loss": 2.7038,
+      "step": 4707
+    },
+    {
+      "epoch": 0.37995319183278187,
+      "grad_norm": 0.7054354548454285,
+      "learning_rate": 0.00017468656802812938,
+      "loss": 2.6566,
+      "step": 4708
+    },
+    {
+      "epoch": 0.38003389556936484,
+      "grad_norm": 0.7231585383415222,
+      "learning_rate": 0.0001746760692062727,
+      "loss": 2.6564,
+      "step": 4709
+    },
+    {
+      "epoch": 0.3801145993059479,
+      "grad_norm": 0.6931934952735901,
+      "learning_rate": 0.00017466556852329318,
+      "loss": 2.6403,
+      "step": 4710
+    },
+    {
+      "epoch": 0.38019530304253085,
+      "grad_norm": 0.7882393598556519,
+      "learning_rate": 0.00017465506597945255,
+      "loss": 2.6337,
+      "step": 4711
+    },
+    {
+      "epoch": 0.3802760067791139,
+      "grad_norm": 0.7015109658241272,
+      "learning_rate": 0.0001746445615750125,
+      "loss": 2.6742,
+      "step": 4712
+    },
+    {
+      "epoch": 0.38035671051569686,
+      "grad_norm": 0.7653505802154541,
+      "learning_rate": 0.0001746340553102348,
+      "loss": 2.6742,
+      "step": 4713
+    },
+    {
+      "epoch": 0.3804374142522799,
+      "grad_norm": 0.7166270613670349,
+      "learning_rate": 0.0001746235471853814,
+      "loss": 2.5995,
+      "step": 4714
+    },
+    {
+      "epoch": 0.38051811798886287,
+      "grad_norm": 0.7612236738204956,
+      "learning_rate": 0.0001746130372007141,
+      "loss": 2.7595,
+      "step": 4715
+    },
+    {
+      "epoch": 0.3805988217254459,
+      "grad_norm": 0.6783852577209473,
+      "learning_rate": 0.00017460252535649493,
+      "loss": 2.6156,
+      "step": 4716
+    },
+    {
+      "epoch": 0.3806795254620289,
+      "grad_norm": 0.7495827078819275,
+      "learning_rate": 0.00017459201165298578,
+      "loss": 2.6847,
+      "step": 4717
+    },
+    {
+      "epoch": 0.3807602291986119,
+      "grad_norm": 0.814798891544342,
+      "learning_rate": 0.0001745814960904487,
+      "loss": 2.6211,
+      "step": 4718
+    },
+    {
+      "epoch": 0.3808409329351949,
+      "grad_norm": 0.7541367411613464,
+      "learning_rate": 0.0001745709786691458,
+      "loss": 2.6214,
+      "step": 4719
+    },
+    {
+      "epoch": 0.3809216366717779,
+      "grad_norm": 0.7065702676773071,
+      "learning_rate": 0.00017456045938933921,
+      "loss": 2.6699,
+      "step": 4720
+    },
+    {
+      "epoch": 0.3810023404083609,
+      "grad_norm": 0.751960813999176,
+      "learning_rate": 0.000174549938251291,
+      "loss": 2.6085,
+      "step": 4721
+    },
+    {
+      "epoch": 0.3810830441449439,
+      "grad_norm": 0.72068190574646,
+      "learning_rate": 0.00017453941525526353,
+      "loss": 2.6201,
+      "step": 4722
+    },
+    {
+      "epoch": 0.3811637478815269,
+      "grad_norm": 0.7201167941093445,
+      "learning_rate": 0.00017452889040151892,
+      "loss": 2.6775,
+      "step": 4723
+    },
+    {
+      "epoch": 0.38124445161810994,
+      "grad_norm": 0.7904958128929138,
+      "learning_rate": 0.00017451836369031956,
+      "loss": 2.7217,
+      "step": 4724
+    },
+    {
+      "epoch": 0.3813251553546929,
+      "grad_norm": 0.7096366882324219,
+      "learning_rate": 0.0001745078351219278,
+      "loss": 2.7004,
+      "step": 4725
+    },
+    {
+      "epoch": 0.38140585909127594,
+      "grad_norm": 0.6812441945075989,
+      "learning_rate": 0.00017449730469660602,
+      "loss": 2.6555,
+      "step": 4726
+    },
+    {
+      "epoch": 0.3814865628278589,
+      "grad_norm": 0.8037428855895996,
+      "learning_rate": 0.00017448677241461665,
+      "loss": 2.7094,
+      "step": 4727
+    },
+    {
+      "epoch": 0.38156726656444195,
+      "grad_norm": 0.7282679677009583,
+      "learning_rate": 0.00017447623827622223,
+      "loss": 2.6699,
+      "step": 4728
+    },
+    {
+      "epoch": 0.38164797030102493,
+      "grad_norm": 0.745705783367157,
+      "learning_rate": 0.00017446570228168523,
+      "loss": 2.6098,
+      "step": 4729
+    },
+    {
+      "epoch": 0.38172867403760796,
+      "grad_norm": 0.7098714113235474,
+      "learning_rate": 0.00017445516443126828,
+      "loss": 2.6628,
+      "step": 4730
+    },
+    {
+      "epoch": 0.38180937777419094,
+      "grad_norm": 0.7376620769500732,
+      "learning_rate": 0.00017444462472523405,
+      "loss": 2.7086,
+      "step": 4731
+    },
+    {
+      "epoch": 0.38189008151077397,
+      "grad_norm": 0.717800498008728,
+      "learning_rate": 0.00017443408316384512,
+      "loss": 2.6582,
+      "step": 4732
+    },
+    {
+      "epoch": 0.38197078524735695,
+      "grad_norm": 0.7061530947685242,
+      "learning_rate": 0.00017442353974736428,
+      "loss": 2.6817,
+      "step": 4733
+    },
+    {
+      "epoch": 0.38205148898394,
+      "grad_norm": 0.744667112827301,
+      "learning_rate": 0.0001744129944760543,
+      "loss": 2.6649,
+      "step": 4734
+    },
+    {
+      "epoch": 0.38213219272052296,
+      "grad_norm": 0.7302529215812683,
+      "learning_rate": 0.00017440244735017797,
+      "loss": 2.7313,
+      "step": 4735
+    },
+    {
+      "epoch": 0.382212896457106,
+      "grad_norm": 0.6845258474349976,
+      "learning_rate": 0.00017439189836999816,
+      "loss": 2.637,
+      "step": 4736
+    },
+    {
+      "epoch": 0.38229360019368896,
+      "grad_norm": 0.7060490250587463,
+      "learning_rate": 0.0001743813475357778,
+      "loss": 2.6674,
+      "step": 4737
+    },
+    {
+      "epoch": 0.382374303930272,
+      "grad_norm": 0.7146841287612915,
+      "learning_rate": 0.00017437079484777977,
+      "loss": 2.6607,
+      "step": 4738
+    },
+    {
+      "epoch": 0.382455007666855,
+      "grad_norm": 0.7107662558555603,
+      "learning_rate": 0.00017436024030626719,
+      "loss": 2.6777,
+      "step": 4739
+    },
+    {
+      "epoch": 0.382535711403438,
+      "grad_norm": 0.7356777191162109,
+      "learning_rate": 0.00017434968391150303,
+      "loss": 2.5801,
+      "step": 4740
+    },
+    {
+      "epoch": 0.382616415140021,
+      "grad_norm": 0.6839054226875305,
+      "learning_rate": 0.00017433912566375037,
+      "loss": 2.6319,
+      "step": 4741
+    },
+    {
+      "epoch": 0.382697118876604,
+      "grad_norm": 0.7049627900123596,
+      "learning_rate": 0.00017432856556327236,
+      "loss": 2.741,
+      "step": 4742
+    },
+    {
+      "epoch": 0.382777822613187,
+      "grad_norm": 0.7926551103591919,
+      "learning_rate": 0.00017431800361033224,
+      "loss": 2.64,
+      "step": 4743
+    },
+    {
+      "epoch": 0.38285852634976997,
+      "grad_norm": 0.734272301197052,
+      "learning_rate": 0.0001743074398051932,
+      "loss": 2.6575,
+      "step": 4744
+    },
+    {
+      "epoch": 0.382939230086353,
+      "grad_norm": 0.6959543824195862,
+      "learning_rate": 0.00017429687414811847,
+      "loss": 2.664,
+      "step": 4745
+    },
+    {
+      "epoch": 0.383019933822936,
+      "grad_norm": 0.7258255481719971,
+      "learning_rate": 0.00017428630663937148,
+      "loss": 2.6597,
+      "step": 4746
+    },
+    {
+      "epoch": 0.383100637559519,
+      "grad_norm": 0.8067473769187927,
+      "learning_rate": 0.0001742757372792155,
+      "loss": 2.6798,
+      "step": 4747
+    },
+    {
+      "epoch": 0.383181341296102,
+      "grad_norm": 0.7000626921653748,
+      "learning_rate": 0.000174265166067914,
+      "loss": 2.6561,
+      "step": 4748
+    },
+    {
+      "epoch": 0.383262045032685,
+      "grad_norm": 0.818914532661438,
+      "learning_rate": 0.00017425459300573045,
+      "loss": 2.6491,
+      "step": 4749
+    },
+    {
+      "epoch": 0.383342748769268,
+      "grad_norm": 0.7060543298721313,
+      "learning_rate": 0.00017424401809292833,
+      "loss": 2.6825,
+      "step": 4750
+    },
+    {
+      "epoch": 0.383423452505851,
+      "grad_norm": 0.893488883972168,
+      "learning_rate": 0.0001742334413297712,
+      "loss": 2.7201,
+      "step": 4751
+    },
+    {
+      "epoch": 0.383504156242434,
+      "grad_norm": 0.8131078481674194,
+      "learning_rate": 0.00017422286271652265,
+      "loss": 2.7828,
+      "step": 4752
+    },
+    {
+      "epoch": 0.38358485997901703,
+      "grad_norm": 0.7735587954521179,
+      "learning_rate": 0.00017421228225344634,
+      "loss": 2.6489,
+      "step": 4753
+    },
+    {
+      "epoch": 0.3836655637156,
+      "grad_norm": 0.713800311088562,
+      "learning_rate": 0.000174201699940806,
+      "loss": 2.6686,
+      "step": 4754
+    },
+    {
+      "epoch": 0.38374626745218304,
+      "grad_norm": 0.8246580362319946,
+      "learning_rate": 0.00017419111577886528,
+      "loss": 2.6771,
+      "step": 4755
+    },
+    {
+      "epoch": 0.383826971188766,
+      "grad_norm": 0.694542646408081,
+      "learning_rate": 0.00017418052976788805,
+      "loss": 2.6632,
+      "step": 4756
+    },
+    {
+      "epoch": 0.38390767492534905,
+      "grad_norm": 0.7200453281402588,
+      "learning_rate": 0.0001741699419081381,
+      "loss": 2.6386,
+      "step": 4757
+    },
+    {
+      "epoch": 0.38398837866193203,
+      "grad_norm": 0.7002073526382446,
+      "learning_rate": 0.00017415935219987933,
+      "loss": 2.6399,
+      "step": 4758
+    },
+    {
+      "epoch": 0.38406908239851506,
+      "grad_norm": 0.7056967616081238,
+      "learning_rate": 0.00017414876064337565,
+      "loss": 2.7048,
+      "step": 4759
+    },
+    {
+      "epoch": 0.38414978613509804,
+      "grad_norm": 0.7406448721885681,
+      "learning_rate": 0.000174138167238891,
+      "loss": 2.6256,
+      "step": 4760
+    },
+    {
+      "epoch": 0.38423048987168107,
+      "grad_norm": 0.7280529737472534,
+      "learning_rate": 0.00017412757198668945,
+      "loss": 2.6393,
+      "step": 4761
+    },
+    {
+      "epoch": 0.38431119360826405,
+      "grad_norm": 0.7626908421516418,
+      "learning_rate": 0.00017411697488703502,
+      "loss": 2.6717,
+      "step": 4762
+    },
+    {
+      "epoch": 0.3843918973448471,
+      "grad_norm": 0.716345489025116,
+      "learning_rate": 0.00017410637594019184,
+      "loss": 2.6457,
+      "step": 4763
+    },
+    {
+      "epoch": 0.38447260108143005,
+      "grad_norm": 0.8825077414512634,
+      "learning_rate": 0.00017409577514642405,
+      "loss": 2.7042,
+      "step": 4764
+    },
+    {
+      "epoch": 0.3845533048180131,
+      "grad_norm": 0.7301186919212341,
+      "learning_rate": 0.00017408517250599585,
+      "loss": 2.7065,
+      "step": 4765
+    },
+    {
+      "epoch": 0.38463400855459606,
+      "grad_norm": 0.8235788345336914,
+      "learning_rate": 0.0001740745680191715,
+      "loss": 2.6315,
+      "step": 4766
+    },
+    {
+      "epoch": 0.3847147122911791,
+      "grad_norm": 0.7355515956878662,
+      "learning_rate": 0.00017406396168621527,
+      "loss": 2.6939,
+      "step": 4767
+    },
+    {
+      "epoch": 0.38479541602776207,
+      "grad_norm": 0.6781682372093201,
+      "learning_rate": 0.0001740533535073915,
+      "loss": 2.6071,
+      "step": 4768
+    },
+    {
+      "epoch": 0.3848761197643451,
+      "grad_norm": 0.801191508769989,
+      "learning_rate": 0.0001740427434829646,
+      "loss": 2.6635,
+      "step": 4769
+    },
+    {
+      "epoch": 0.3849568235009281,
+      "grad_norm": 0.759682297706604,
+      "learning_rate": 0.00017403213161319903,
+      "loss": 2.6823,
+      "step": 4770
+    },
+    {
+      "epoch": 0.3850375272375111,
+      "grad_norm": 0.806498110294342,
+      "learning_rate": 0.00017402151789835916,
+      "loss": 2.7111,
+      "step": 4771
+    },
+    {
+      "epoch": 0.3851182309740941,
+      "grad_norm": 0.7677996158599854,
+      "learning_rate": 0.00017401090233870958,
+      "loss": 2.6701,
+      "step": 4772
+    },
+    {
+      "epoch": 0.3851989347106771,
+      "grad_norm": 0.7449933290481567,
+      "learning_rate": 0.00017400028493451487,
+      "loss": 2.7037,
+      "step": 4773
+    },
+    {
+      "epoch": 0.3852796384472601,
+      "grad_norm": 0.7506107091903687,
+      "learning_rate": 0.0001739896656860396,
+      "loss": 2.6587,
+      "step": 4774
+    },
+    {
+      "epoch": 0.38536034218384313,
+      "grad_norm": 0.8781036734580994,
+      "learning_rate": 0.00017397904459354844,
+      "loss": 2.7634,
+      "step": 4775
+    },
+    {
+      "epoch": 0.3854410459204261,
+      "grad_norm": 0.7067514657974243,
+      "learning_rate": 0.0001739684216573061,
+      "loss": 2.638,
+      "step": 4776
+    },
+    {
+      "epoch": 0.38552174965700914,
+      "grad_norm": 0.7742886543273926,
+      "learning_rate": 0.00017395779687757735,
+      "loss": 2.7043,
+      "step": 4777
+    },
+    {
+      "epoch": 0.3856024533935921,
+      "grad_norm": 0.7348291277885437,
+      "learning_rate": 0.00017394717025462697,
+      "loss": 2.7404,
+      "step": 4778
+    },
+    {
+      "epoch": 0.38568315713017515,
+      "grad_norm": 0.7449346780776978,
+      "learning_rate": 0.00017393654178871984,
+      "loss": 2.631,
+      "step": 4779
+    },
+    {
+      "epoch": 0.3857638608667581,
+      "grad_norm": 0.7191200256347656,
+      "learning_rate": 0.00017392591148012078,
+      "loss": 2.6776,
+      "step": 4780
+    },
+    {
+      "epoch": 0.38584456460334116,
+      "grad_norm": 0.7055533528327942,
+      "learning_rate": 0.00017391527932909476,
+      "loss": 2.6219,
+      "step": 4781
+    },
+    {
+      "epoch": 0.38592526833992413,
+      "grad_norm": 0.73755943775177,
+      "learning_rate": 0.0001739046453359068,
+      "loss": 2.6692,
+      "step": 4782
+    },
+    {
+      "epoch": 0.38600597207650716,
+      "grad_norm": 0.7469369769096375,
+      "learning_rate": 0.00017389400950082185,
+      "loss": 2.6572,
+      "step": 4783
+    },
+    {
+      "epoch": 0.38608667581309014,
+      "grad_norm": 0.7552534341812134,
+      "learning_rate": 0.00017388337182410504,
+      "loss": 2.6853,
+      "step": 4784
+    },
+    {
+      "epoch": 0.3861673795496732,
+      "grad_norm": 0.7453532814979553,
+      "learning_rate": 0.00017387273230602145,
+      "loss": 2.6601,
+      "step": 4785
+    },
+    {
+      "epoch": 0.38624808328625615,
+      "grad_norm": 0.7259301543235779,
+      "learning_rate": 0.0001738620909468363,
+      "loss": 2.6997,
+      "step": 4786
+    },
+    {
+      "epoch": 0.3863287870228392,
+      "grad_norm": 0.6970019936561584,
+      "learning_rate": 0.00017385144774681476,
+      "loss": 2.7497,
+      "step": 4787
+    },
+    {
+      "epoch": 0.38640949075942216,
+      "grad_norm": 0.7172032594680786,
+      "learning_rate": 0.00017384080270622208,
+      "loss": 2.7182,
+      "step": 4788
+    },
+    {
+      "epoch": 0.3864901944960052,
+      "grad_norm": 0.7184371948242188,
+      "learning_rate": 0.00017383015582532357,
+      "loss": 2.6358,
+      "step": 4789
+    },
+    {
+      "epoch": 0.38657089823258817,
+      "grad_norm": 0.7302096486091614,
+      "learning_rate": 0.00017381950710438458,
+      "loss": 2.6066,
+      "step": 4790
+    },
+    {
+      "epoch": 0.3866516019691712,
+      "grad_norm": 0.7043540477752686,
+      "learning_rate": 0.00017380885654367053,
+      "loss": 2.699,
+      "step": 4791
+    },
+    {
+      "epoch": 0.3867323057057542,
+      "grad_norm": 0.6919732689857483,
+      "learning_rate": 0.0001737982041434468,
+      "loss": 2.6025,
+      "step": 4792
+    },
+    {
+      "epoch": 0.3868130094423372,
+      "grad_norm": 0.7277705669403076,
+      "learning_rate": 0.00017378754990397894,
+      "loss": 2.6764,
+      "step": 4793
+    },
+    {
+      "epoch": 0.3868937131789202,
+      "grad_norm": 0.7546190619468689,
+      "learning_rate": 0.00017377689382553247,
+      "loss": 2.5865,
+      "step": 4794
+    },
+    {
+      "epoch": 0.38697441691550316,
+      "grad_norm": 0.7636401653289795,
+      "learning_rate": 0.00017376623590837294,
+      "loss": 2.6488,
+      "step": 4795
+    },
+    {
+      "epoch": 0.3870551206520862,
+      "grad_norm": 0.6945658922195435,
+      "learning_rate": 0.00017375557615276595,
+      "loss": 2.6739,
+      "step": 4796
+    },
+    {
+      "epoch": 0.38713582438866917,
+      "grad_norm": 0.7503637075424194,
+      "learning_rate": 0.00017374491455897722,
+      "loss": 2.6854,
+      "step": 4797
+    },
+    {
+      "epoch": 0.3872165281252522,
+      "grad_norm": 0.7457373142242432,
+      "learning_rate": 0.00017373425112727247,
+      "loss": 2.6659,
+      "step": 4798
+    },
+    {
+      "epoch": 0.3872972318618352,
+      "grad_norm": 0.7742534875869751,
+      "learning_rate": 0.0001737235858579174,
+      "loss": 2.6461,
+      "step": 4799
+    },
+    {
+      "epoch": 0.3873779355984182,
+      "grad_norm": 0.7397909760475159,
+      "learning_rate": 0.0001737129187511779,
+      "loss": 2.6779,
+      "step": 4800
+    },
+    {
+      "epoch": 0.3874586393350012,
+      "grad_norm": 0.7922031879425049,
+      "learning_rate": 0.00017370224980731974,
+      "loss": 2.6417,
+      "step": 4801
+    },
+    {
+      "epoch": 0.3875393430715842,
+      "grad_norm": 0.8503968715667725,
+      "learning_rate": 0.00017369157902660887,
+      "loss": 2.7063,
+      "step": 4802
+    },
+    {
+      "epoch": 0.3876200468081672,
+      "grad_norm": 0.7143701314926147,
+      "learning_rate": 0.00017368090640931125,
+      "loss": 2.6152,
+      "step": 4803
+    },
+    {
+      "epoch": 0.38770075054475023,
+      "grad_norm": 0.8016753196716309,
+      "learning_rate": 0.0001736702319556928,
+      "loss": 2.6005,
+      "step": 4804
+    },
+    {
+      "epoch": 0.3877814542813332,
+      "grad_norm": 0.7329538464546204,
+      "learning_rate": 0.00017365955566601962,
+      "loss": 2.6027,
+      "step": 4805
+    },
+    {
+      "epoch": 0.38786215801791624,
+      "grad_norm": 0.7005148530006409,
+      "learning_rate": 0.00017364887754055773,
+      "loss": 2.6585,
+      "step": 4806
+    },
+    {
+      "epoch": 0.3879428617544992,
+      "grad_norm": 0.7092769145965576,
+      "learning_rate": 0.00017363819757957333,
+      "loss": 2.6763,
+      "step": 4807
+    },
+    {
+      "epoch": 0.38802356549108225,
+      "grad_norm": 0.7475202679634094,
+      "learning_rate": 0.0001736275157833325,
+      "loss": 2.5969,
+      "step": 4808
+    },
+    {
+      "epoch": 0.3881042692276652,
+      "grad_norm": 0.822496235370636,
+      "learning_rate": 0.0001736168321521016,
+      "loss": 2.6758,
+      "step": 4809
+    },
+    {
+      "epoch": 0.38818497296424825,
+      "grad_norm": 0.7756842374801636,
+      "learning_rate": 0.0001736061466861467,
+      "loss": 2.6676,
+      "step": 4810
+    },
+    {
+      "epoch": 0.38826567670083123,
+      "grad_norm": 0.7192497849464417,
+      "learning_rate": 0.00017359545938573428,
+      "loss": 2.7045,
+      "step": 4811
+    },
+    {
+      "epoch": 0.38834638043741426,
+      "grad_norm": 0.7064149379730225,
+      "learning_rate": 0.00017358477025113063,
+      "loss": 2.6169,
+      "step": 4812
+    },
+    {
+      "epoch": 0.38842708417399724,
+      "grad_norm": 0.7297258973121643,
+      "learning_rate": 0.00017357407928260215,
+      "loss": 2.612,
+      "step": 4813
+    },
+    {
+      "epoch": 0.38850778791058027,
+      "grad_norm": 0.7011935114860535,
+      "learning_rate": 0.00017356338648041528,
+      "loss": 2.6507,
+      "step": 4814
+    },
+    {
+      "epoch": 0.38858849164716325,
+      "grad_norm": 0.7647256255149841,
+      "learning_rate": 0.00017355269184483651,
+      "loss": 2.6838,
+      "step": 4815
+    },
+    {
+      "epoch": 0.3886691953837463,
+      "grad_norm": 0.690182089805603,
+      "learning_rate": 0.0001735419953761324,
+      "loss": 2.6996,
+      "step": 4816
+    },
+    {
+      "epoch": 0.38874989912032926,
+      "grad_norm": 0.7142173647880554,
+      "learning_rate": 0.00017353129707456955,
+      "loss": 2.6705,
+      "step": 4817
+    },
+    {
+      "epoch": 0.3888306028569123,
+      "grad_norm": 0.801369309425354,
+      "learning_rate": 0.00017352059694041456,
+      "loss": 2.7002,
+      "step": 4818
+    },
+    {
+      "epoch": 0.38891130659349527,
+      "grad_norm": 0.7021649479866028,
+      "learning_rate": 0.0001735098949739341,
+      "loss": 2.7042,
+      "step": 4819
+    },
+    {
+      "epoch": 0.3889920103300783,
+      "grad_norm": 0.6802586317062378,
+      "learning_rate": 0.00017349919117539488,
+      "loss": 2.7186,
+      "step": 4820
+    },
+    {
+      "epoch": 0.3890727140666613,
+      "grad_norm": 0.7723212838172913,
+      "learning_rate": 0.0001734884855450637,
+      "loss": 2.608,
+      "step": 4821
+    },
+    {
+      "epoch": 0.3891534178032443,
+      "grad_norm": 0.7037193179130554,
+      "learning_rate": 0.00017347777808320735,
+      "loss": 2.6198,
+      "step": 4822
+    },
+    {
+      "epoch": 0.3892341215398273,
+      "grad_norm": 0.7172731161117554,
+      "learning_rate": 0.00017346706879009272,
+      "loss": 2.7037,
+      "step": 4823
+    },
+    {
+      "epoch": 0.3893148252764103,
+      "grad_norm": 0.7421539425849915,
+      "learning_rate": 0.00017345635766598667,
+      "loss": 2.6619,
+      "step": 4824
+    },
+    {
+      "epoch": 0.3893955290129933,
+      "grad_norm": 0.7587071061134338,
+      "learning_rate": 0.0001734456447111562,
+      "loss": 2.6229,
+      "step": 4825
+    },
+    {
+      "epoch": 0.3894762327495763,
+      "grad_norm": 0.6981459259986877,
+      "learning_rate": 0.00017343492992586822,
+      "loss": 2.5927,
+      "step": 4826
+    },
+    {
+      "epoch": 0.3895569364861593,
+      "grad_norm": 0.7628491520881653,
+      "learning_rate": 0.00017342421331038987,
+      "loss": 2.7047,
+      "step": 4827
+    },
+    {
+      "epoch": 0.38963764022274233,
+      "grad_norm": 0.8005064129829407,
+      "learning_rate": 0.00017341349486498818,
+      "loss": 2.6918,
+      "step": 4828
+    },
+    {
+      "epoch": 0.3897183439593253,
+      "grad_norm": 0.7756431102752686,
+      "learning_rate": 0.0001734027745899303,
+      "loss": 2.6621,
+      "step": 4829
+    },
+    {
+      "epoch": 0.38979904769590834,
+      "grad_norm": 0.7317833304405212,
+      "learning_rate": 0.00017339205248548338,
+      "loss": 2.7134,
+      "step": 4830
+    },
+    {
+      "epoch": 0.3898797514324913,
+      "grad_norm": 0.7293959259986877,
+      "learning_rate": 0.0001733813285519147,
+      "loss": 2.6865,
+      "step": 4831
+    },
+    {
+      "epoch": 0.38996045516907435,
+      "grad_norm": 0.7120299935340881,
+      "learning_rate": 0.00017337060278949147,
+      "loss": 2.6915,
+      "step": 4832
+    },
+    {
+      "epoch": 0.3900411589056573,
+      "grad_norm": 0.7255397439002991,
+      "learning_rate": 0.00017335987519848103,
+      "loss": 2.6671,
+      "step": 4833
+    },
+    {
+      "epoch": 0.39012186264224036,
+      "grad_norm": 0.7849408388137817,
+      "learning_rate": 0.0001733491457791507,
+      "loss": 2.6301,
+      "step": 4834
+    },
+    {
+      "epoch": 0.39020256637882333,
+      "grad_norm": 0.6998472809791565,
+      "learning_rate": 0.00017333841453176797,
+      "loss": 2.6587,
+      "step": 4835
+    },
+    {
+      "epoch": 0.39028327011540637,
+      "grad_norm": 0.7530023455619812,
+      "learning_rate": 0.00017332768145660024,
+      "loss": 2.7011,
+      "step": 4836
+    },
+    {
+      "epoch": 0.39036397385198934,
+      "grad_norm": 0.7251207828521729,
+      "learning_rate": 0.00017331694655391497,
+      "loss": 2.6416,
+      "step": 4837
+    },
+    {
+      "epoch": 0.3904446775885724,
+      "grad_norm": 0.7016854882240295,
+      "learning_rate": 0.00017330620982397975,
+      "loss": 2.7224,
+      "step": 4838
+    },
+    {
+      "epoch": 0.39052538132515535,
+      "grad_norm": 0.7253310084342957,
+      "learning_rate": 0.00017329547126706217,
+      "loss": 2.6747,
+      "step": 4839
+    },
+    {
+      "epoch": 0.3906060850617384,
+      "grad_norm": 0.7114601731300354,
+      "learning_rate": 0.00017328473088342987,
+      "loss": 2.6654,
+      "step": 4840
+    },
+    {
+      "epoch": 0.39068678879832136,
+      "grad_norm": 0.7773289680480957,
+      "learning_rate": 0.00017327398867335048,
+      "loss": 2.6625,
+      "step": 4841
+    },
+    {
+      "epoch": 0.3907674925349044,
+      "grad_norm": 0.7541868686676025,
+      "learning_rate": 0.00017326324463709175,
+      "loss": 2.667,
+      "step": 4842
+    },
+    {
+      "epoch": 0.39084819627148737,
+      "grad_norm": 0.8095890283584595,
+      "learning_rate": 0.00017325249877492147,
+      "loss": 2.706,
+      "step": 4843
+    },
+    {
+      "epoch": 0.3909289000080704,
+      "grad_norm": 0.7019474506378174,
+      "learning_rate": 0.00017324175108710742,
+      "loss": 2.6125,
+      "step": 4844
+    },
+    {
+      "epoch": 0.3910096037446534,
+      "grad_norm": 0.7055396437644958,
+      "learning_rate": 0.00017323100157391746,
+      "loss": 2.6373,
+      "step": 4845
+    },
+    {
+      "epoch": 0.39109030748123635,
+      "grad_norm": 0.7332476377487183,
+      "learning_rate": 0.00017322025023561955,
+      "loss": 2.6559,
+      "step": 4846
+    },
+    {
+      "epoch": 0.3911710112178194,
+      "grad_norm": 0.7740387916564941,
+      "learning_rate": 0.00017320949707248158,
+      "loss": 2.7341,
+      "step": 4847
+    },
+    {
+      "epoch": 0.39125171495440236,
+      "grad_norm": 0.7371044754981995,
+      "learning_rate": 0.0001731987420847716,
+      "loss": 2.7318,
+      "step": 4848
+    },
+    {
+      "epoch": 0.3913324186909854,
+      "grad_norm": 0.7897786498069763,
+      "learning_rate": 0.00017318798527275758,
+      "loss": 2.6759,
+      "step": 4849
+    },
+    {
+      "epoch": 0.39141312242756837,
+      "grad_norm": 0.7149896621704102,
+      "learning_rate": 0.0001731772266367077,
+      "loss": 2.7097,
+      "step": 4850
+    },
+    {
+      "epoch": 0.3914938261641514,
+      "grad_norm": 0.7824358344078064,
+      "learning_rate": 0.00017316646617689002,
+      "loss": 2.6376,
+      "step": 4851
+    },
+    {
+      "epoch": 0.3915745299007344,
+      "grad_norm": 0.7704496383666992,
+      "learning_rate": 0.00017315570389357272,
+      "loss": 2.6539,
+      "step": 4852
+    },
+    {
+      "epoch": 0.3916552336373174,
+      "grad_norm": 0.7489706873893738,
+      "learning_rate": 0.00017314493978702407,
+      "loss": 2.6716,
+      "step": 4853
+    },
+    {
+      "epoch": 0.3917359373739004,
+      "grad_norm": 0.7368690967559814,
+      "learning_rate": 0.00017313417385751234,
+      "loss": 2.7171,
+      "step": 4854
+    },
+    {
+      "epoch": 0.3918166411104834,
+      "grad_norm": 0.7215858697891235,
+      "learning_rate": 0.00017312340610530579,
+      "loss": 2.6306,
+      "step": 4855
+    },
+    {
+      "epoch": 0.3918973448470664,
+      "grad_norm": 0.7622217535972595,
+      "learning_rate": 0.00017311263653067285,
+      "loss": 2.6089,
+      "step": 4856
+    },
+    {
+      "epoch": 0.39197804858364943,
+      "grad_norm": 0.7317889332771301,
+      "learning_rate": 0.00017310186513388185,
+      "loss": 2.6831,
+      "step": 4857
+    },
+    {
+      "epoch": 0.3920587523202324,
+      "grad_norm": 0.894185483455658,
+      "learning_rate": 0.0001730910919152013,
+      "loss": 2.684,
+      "step": 4858
+    },
+    {
+      "epoch": 0.39213945605681544,
+      "grad_norm": 0.7313157916069031,
+      "learning_rate": 0.00017308031687489968,
+      "loss": 2.6465,
+      "step": 4859
+    },
+    {
+      "epoch": 0.3922201597933984,
+      "grad_norm": 0.7765825390815735,
+      "learning_rate": 0.00017306954001324552,
+      "loss": 2.6526,
+      "step": 4860
+    },
+    {
+      "epoch": 0.39230086352998145,
+      "grad_norm": 0.7171424031257629,
+      "learning_rate": 0.00017305876133050742,
+      "loss": 2.6212,
+      "step": 4861
+    },
+    {
+      "epoch": 0.3923815672665644,
+      "grad_norm": 0.7215112447738647,
+      "learning_rate": 0.000173047980826954,
+      "loss": 2.6329,
+      "step": 4862
+    },
+    {
+      "epoch": 0.39246227100314746,
+      "grad_norm": 0.7393578886985779,
+      "learning_rate": 0.00017303719850285396,
+      "loss": 2.7264,
+      "step": 4863
+    },
+    {
+      "epoch": 0.39254297473973043,
+      "grad_norm": 0.7620136737823486,
+      "learning_rate": 0.00017302641435847603,
+      "loss": 2.6686,
+      "step": 4864
+    },
+    {
+      "epoch": 0.39262367847631346,
+      "grad_norm": 0.7290963530540466,
+      "learning_rate": 0.00017301562839408893,
+      "loss": 2.578,
+      "step": 4865
+    },
+    {
+      "epoch": 0.39270438221289644,
+      "grad_norm": 0.6978541612625122,
+      "learning_rate": 0.00017300484060996153,
+      "loss": 2.6783,
+      "step": 4866
+    },
+    {
+      "epoch": 0.3927850859494795,
+      "grad_norm": 0.7212007641792297,
+      "learning_rate": 0.00017299405100636264,
+      "loss": 2.6282,
+      "step": 4867
+    },
+    {
+      "epoch": 0.39286578968606245,
+      "grad_norm": 0.757324755191803,
+      "learning_rate": 0.0001729832595835612,
+      "loss": 2.6933,
+      "step": 4868
+    },
+    {
+      "epoch": 0.3929464934226455,
+      "grad_norm": 0.7052869200706482,
+      "learning_rate": 0.00017297246634182618,
+      "loss": 2.7152,
+      "step": 4869
+    },
+    {
+      "epoch": 0.39302719715922846,
+      "grad_norm": 0.7326259016990662,
+      "learning_rate": 0.0001729616712814265,
+      "loss": 2.6792,
+      "step": 4870
+    },
+    {
+      "epoch": 0.3931079008958115,
+      "grad_norm": 0.7540302276611328,
+      "learning_rate": 0.00017295087440263128,
+      "loss": 2.6621,
+      "step": 4871
+    },
+    {
+      "epoch": 0.39318860463239447,
+      "grad_norm": 0.765454888343811,
+      "learning_rate": 0.00017294007570570956,
+      "loss": 2.7049,
+      "step": 4872
+    },
+    {
+      "epoch": 0.3932693083689775,
+      "grad_norm": 0.7303065061569214,
+      "learning_rate": 0.0001729292751909305,
+      "loss": 2.6867,
+      "step": 4873
+    },
+    {
+      "epoch": 0.3933500121055605,
+      "grad_norm": 0.7049854397773743,
+      "learning_rate": 0.00017291847285856325,
+      "loss": 2.7052,
+      "step": 4874
+    },
+    {
+      "epoch": 0.3934307158421435,
+      "grad_norm": 0.7199053764343262,
+      "learning_rate": 0.00017290766870887704,
+      "loss": 2.7195,
+      "step": 4875
+    },
+    {
+      "epoch": 0.3935114195787265,
+      "grad_norm": 0.7536180019378662,
+      "learning_rate": 0.00017289686274214118,
+      "loss": 2.6861,
+      "step": 4876
+    },
+    {
+      "epoch": 0.3935921233153095,
+      "grad_norm": 0.7295238971710205,
+      "learning_rate": 0.00017288605495862492,
+      "loss": 2.6684,
+      "step": 4877
+    },
+    {
+      "epoch": 0.3936728270518925,
+      "grad_norm": 0.7575719952583313,
+      "learning_rate": 0.00017287524535859763,
+      "loss": 2.6439,
+      "step": 4878
+    },
+    {
+      "epoch": 0.3937535307884755,
+      "grad_norm": 0.678909182548523,
+      "learning_rate": 0.00017286443394232874,
+      "loss": 2.6562,
+      "step": 4879
+    },
+    {
+      "epoch": 0.3938342345250585,
+      "grad_norm": 0.6908892393112183,
+      "learning_rate": 0.00017285362071008768,
+      "loss": 2.6364,
+      "step": 4880
+    },
+    {
+      "epoch": 0.39391493826164153,
+      "grad_norm": 0.7414079904556274,
+      "learning_rate": 0.00017284280566214397,
+      "loss": 2.5872,
+      "step": 4881
+    },
+    {
+      "epoch": 0.3939956419982245,
+      "grad_norm": 0.6824749112129211,
+      "learning_rate": 0.0001728319887987671,
+      "loss": 2.641,
+      "step": 4882
+    },
+    {
+      "epoch": 0.39407634573480754,
+      "grad_norm": 0.6908513903617859,
+      "learning_rate": 0.0001728211701202267,
+      "loss": 2.6977,
+      "step": 4883
+    },
+    {
+      "epoch": 0.3941570494713905,
+      "grad_norm": 0.7214735746383667,
+      "learning_rate": 0.0001728103496267924,
+      "loss": 2.5826,
+      "step": 4884
+    },
+    {
+      "epoch": 0.39423775320797355,
+      "grad_norm": 0.812781572341919,
+      "learning_rate": 0.00017279952731873385,
+      "loss": 2.6806,
+      "step": 4885
+    },
+    {
+      "epoch": 0.39431845694455653,
+      "grad_norm": 0.7610746026039124,
+      "learning_rate": 0.00017278870319632078,
+      "loss": 2.6046,
+      "step": 4886
+    },
+    {
+      "epoch": 0.39439916068113956,
+      "grad_norm": 0.7151652574539185,
+      "learning_rate": 0.00017277787725982293,
+      "loss": 2.6543,
+      "step": 4887
+    },
+    {
+      "epoch": 0.39447986441772254,
+      "grad_norm": 0.7293612360954285,
+      "learning_rate": 0.00017276704950951017,
+      "loss": 2.6384,
+      "step": 4888
+    },
+    {
+      "epoch": 0.39456056815430557,
+      "grad_norm": 0.8138254284858704,
+      "learning_rate": 0.00017275621994565233,
+      "loss": 2.7208,
+      "step": 4889
+    },
+    {
+      "epoch": 0.39464127189088855,
+      "grad_norm": 0.7557196021080017,
+      "learning_rate": 0.00017274538856851924,
+      "loss": 2.6571,
+      "step": 4890
+    },
+    {
+      "epoch": 0.3947219756274716,
+      "grad_norm": 0.7297266721725464,
+      "learning_rate": 0.00017273455537838097,
+      "loss": 2.6222,
+      "step": 4891
+    },
+    {
+      "epoch": 0.39480267936405455,
+      "grad_norm": 0.7838431596755981,
+      "learning_rate": 0.00017272372037550743,
+      "loss": 2.782,
+      "step": 4892
+    },
+    {
+      "epoch": 0.3948833831006376,
+      "grad_norm": 0.7799673676490784,
+      "learning_rate": 0.00017271288356016866,
+      "loss": 2.6658,
+      "step": 4893
+    },
+    {
+      "epoch": 0.39496408683722056,
+      "grad_norm": 0.8495545387268066,
+      "learning_rate": 0.0001727020449326348,
+      "loss": 2.6552,
+      "step": 4894
+    },
+    {
+      "epoch": 0.3950447905738036,
+      "grad_norm": 0.7317770719528198,
+      "learning_rate": 0.00017269120449317588,
+      "loss": 2.6616,
+      "step": 4895
+    },
+    {
+      "epoch": 0.39512549431038657,
+      "grad_norm": 0.7518885731697083,
+      "learning_rate": 0.00017268036224206217,
+      "loss": 2.6864,
+      "step": 4896
+    },
+    {
+      "epoch": 0.39520619804696955,
+      "grad_norm": 0.83487468957901,
+      "learning_rate": 0.00017266951817956382,
+      "loss": 2.7535,
+      "step": 4897
+    },
+    {
+      "epoch": 0.3952869017835526,
+      "grad_norm": 0.7440658211708069,
+      "learning_rate": 0.00017265867230595113,
+      "loss": 2.6584,
+      "step": 4898
+    },
+    {
+      "epoch": 0.39536760552013556,
+      "grad_norm": 0.7060485482215881,
+      "learning_rate": 0.00017264782462149438,
+      "loss": 2.6892,
+      "step": 4899
+    },
+    {
+      "epoch": 0.3954483092567186,
+      "grad_norm": 0.8410428166389465,
+      "learning_rate": 0.00017263697512646394,
+      "loss": 2.6425,
+      "step": 4900
+    },
+    {
+      "epoch": 0.39552901299330157,
+      "grad_norm": 0.757046639919281,
+      "learning_rate": 0.0001726261238211302,
+      "loss": 2.6159,
+      "step": 4901
+    },
+    {
+      "epoch": 0.3956097167298846,
+      "grad_norm": 0.7288908958435059,
+      "learning_rate": 0.00017261527070576365,
+      "loss": 2.6753,
+      "step": 4902
+    },
+    {
+      "epoch": 0.3956904204664676,
+      "grad_norm": 0.8194541335105896,
+      "learning_rate": 0.0001726044157806347,
+      "loss": 2.6673,
+      "step": 4903
+    },
+    {
+      "epoch": 0.3957711242030506,
+      "grad_norm": 0.7957740426063538,
+      "learning_rate": 0.00017259355904601393,
+      "loss": 2.6662,
+      "step": 4904
+    },
+    {
+      "epoch": 0.3958518279396336,
+      "grad_norm": 0.8790122270584106,
+      "learning_rate": 0.0001725827005021719,
+      "loss": 2.7513,
+      "step": 4905
+    },
+    {
+      "epoch": 0.3959325316762166,
+      "grad_norm": 0.7674984335899353,
+      "learning_rate": 0.00017257184014937924,
+      "loss": 2.6375,
+      "step": 4906
+    },
+    {
+      "epoch": 0.3960132354127996,
+      "grad_norm": 0.7250992655754089,
+      "learning_rate": 0.00017256097798790663,
+      "loss": 2.63,
+      "step": 4907
+    },
+    {
+      "epoch": 0.3960939391493826,
+      "grad_norm": 0.8578312397003174,
+      "learning_rate": 0.00017255011401802475,
+      "loss": 2.702,
+      "step": 4908
+    },
+    {
+      "epoch": 0.3961746428859656,
+      "grad_norm": 0.7365253567695618,
+      "learning_rate": 0.00017253924824000438,
+      "loss": 2.6156,
+      "step": 4909
+    },
+    {
+      "epoch": 0.39625534662254863,
+      "grad_norm": 0.7148925065994263,
+      "learning_rate": 0.00017252838065411633,
+      "loss": 2.6658,
+      "step": 4910
+    },
+    {
+      "epoch": 0.3963360503591316,
+      "grad_norm": 0.7517829537391663,
+      "learning_rate": 0.00017251751126063148,
+      "loss": 2.6347,
+      "step": 4911
+    },
+    {
+      "epoch": 0.39641675409571464,
+      "grad_norm": 0.7880864143371582,
+      "learning_rate": 0.00017250664005982066,
+      "loss": 2.7045,
+      "step": 4912
+    },
+    {
+      "epoch": 0.3964974578322976,
+      "grad_norm": 0.7460693120956421,
+      "learning_rate": 0.00017249576705195482,
+      "loss": 2.6976,
+      "step": 4913
+    },
+    {
+      "epoch": 0.39657816156888065,
+      "grad_norm": 0.7179895043373108,
+      "learning_rate": 0.00017248489223730496,
+      "loss": 2.6366,
+      "step": 4914
+    },
+    {
+      "epoch": 0.3966588653054636,
+      "grad_norm": 0.7737421989440918,
+      "learning_rate": 0.00017247401561614213,
+      "loss": 2.7116,
+      "step": 4915
+    },
+    {
+      "epoch": 0.39673956904204666,
+      "grad_norm": 0.8561483025550842,
+      "learning_rate": 0.0001724631371887374,
+      "loss": 2.6591,
+      "step": 4916
+    },
+    {
+      "epoch": 0.39682027277862963,
+      "grad_norm": 0.7616356611251831,
+      "learning_rate": 0.00017245225695536182,
+      "loss": 2.6436,
+      "step": 4917
+    },
+    {
+      "epoch": 0.39690097651521267,
+      "grad_norm": 0.7754645943641663,
+      "learning_rate": 0.0001724413749162866,
+      "loss": 2.6699,
+      "step": 4918
+    },
+    {
+      "epoch": 0.39698168025179564,
+      "grad_norm": 0.800165593624115,
+      "learning_rate": 0.000172430491071783,
+      "loss": 2.7155,
+      "step": 4919
+    },
+    {
+      "epoch": 0.3970623839883787,
+      "grad_norm": 0.8448799848556519,
+      "learning_rate": 0.00017241960542212223,
+      "loss": 2.6991,
+      "step": 4920
+    },
+    {
+      "epoch": 0.39714308772496165,
+      "grad_norm": 0.7106496095657349,
+      "learning_rate": 0.00017240871796757556,
+      "loss": 2.628,
+      "step": 4921
+    },
+    {
+      "epoch": 0.3972237914615447,
+      "grad_norm": 0.7332959175109863,
+      "learning_rate": 0.00017239782870841436,
+      "loss": 2.6159,
+      "step": 4922
+    },
+    {
+      "epoch": 0.39730449519812766,
+      "grad_norm": 0.7573551535606384,
+      "learning_rate": 0.00017238693764491002,
+      "loss": 2.67,
+      "step": 4923
+    },
+    {
+      "epoch": 0.3973851989347107,
+      "grad_norm": 0.7833136320114136,
+      "learning_rate": 0.00017237604477733399,
+      "loss": 2.7276,
+      "step": 4924
+    },
+    {
+      "epoch": 0.39746590267129367,
+      "grad_norm": 0.7233073711395264,
+      "learning_rate": 0.00017236515010595773,
+      "loss": 2.6654,
+      "step": 4925
+    },
+    {
+      "epoch": 0.3975466064078767,
+      "grad_norm": 0.7920324206352234,
+      "learning_rate": 0.00017235425363105273,
+      "loss": 2.7611,
+      "step": 4926
+    },
+    {
+      "epoch": 0.3976273101444597,
+      "grad_norm": 0.7096883058547974,
+      "learning_rate": 0.00017234335535289063,
+      "loss": 2.687,
+      "step": 4927
+    },
+    {
+      "epoch": 0.3977080138810427,
+      "grad_norm": 0.7231960296630859,
+      "learning_rate": 0.000172332455271743,
+      "loss": 2.6441,
+      "step": 4928
+    },
+    {
+      "epoch": 0.3977887176176257,
+      "grad_norm": 0.7852105498313904,
+      "learning_rate": 0.00017232155338788146,
+      "loss": 2.5948,
+      "step": 4929
+    },
+    {
+      "epoch": 0.3978694213542087,
+      "grad_norm": 0.788789689540863,
+      "learning_rate": 0.0001723106497015778,
+      "loss": 2.6797,
+      "step": 4930
+    },
+    {
+      "epoch": 0.3979501250907917,
+      "grad_norm": 0.7082793116569519,
+      "learning_rate": 0.00017229974421310377,
+      "loss": 2.6787,
+      "step": 4931
+    },
+    {
+      "epoch": 0.3980308288273747,
+      "grad_norm": 0.8157992362976074,
+      "learning_rate": 0.00017228883692273106,
+      "loss": 2.6367,
+      "step": 4932
+    },
+    {
+      "epoch": 0.3981115325639577,
+      "grad_norm": 0.7576673030853271,
+      "learning_rate": 0.00017227792783073157,
+      "loss": 2.6826,
+      "step": 4933
+    },
+    {
+      "epoch": 0.39819223630054074,
+      "grad_norm": 0.7225388884544373,
+      "learning_rate": 0.00017226701693737718,
+      "loss": 2.668,
+      "step": 4934
+    },
+    {
+      "epoch": 0.3982729400371237,
+      "grad_norm": 0.7029562592506409,
+      "learning_rate": 0.00017225610424293985,
+      "loss": 2.6613,
+      "step": 4935
+    },
+    {
+      "epoch": 0.39835364377370674,
+      "grad_norm": 0.73081374168396,
+      "learning_rate": 0.0001722451897476915,
+      "loss": 2.6378,
+      "step": 4936
+    },
+    {
+      "epoch": 0.3984343475102897,
+      "grad_norm": 0.744008481502533,
+      "learning_rate": 0.0001722342734519042,
+      "loss": 2.6501,
+      "step": 4937
+    },
+    {
+      "epoch": 0.39851505124687275,
+      "grad_norm": 0.7482618093490601,
+      "learning_rate": 0.00017222335535584996,
+      "loss": 2.7287,
+      "step": 4938
+    },
+    {
+      "epoch": 0.39859575498345573,
+      "grad_norm": 0.6487892866134644,
+      "learning_rate": 0.00017221243545980093,
+      "loss": 2.6417,
+      "step": 4939
+    },
+    {
+      "epoch": 0.39867645872003876,
+      "grad_norm": 0.7894789576530457,
+      "learning_rate": 0.00017220151376402923,
+      "loss": 2.7431,
+      "step": 4940
+    },
+    {
+      "epoch": 0.39875716245662174,
+      "grad_norm": 0.8232294321060181,
+      "learning_rate": 0.00017219059026880708,
+      "loss": 2.6824,
+      "step": 4941
+    },
+    {
+      "epoch": 0.39883786619320477,
+      "grad_norm": 0.6844691634178162,
+      "learning_rate": 0.00017217966497440668,
+      "loss": 2.6294,
+      "step": 4942
+    },
+    {
+      "epoch": 0.39891856992978775,
+      "grad_norm": 0.7245259881019592,
+      "learning_rate": 0.00017216873788110037,
+      "loss": 2.6815,
+      "step": 4943
+    },
+    {
+      "epoch": 0.3989992736663708,
+      "grad_norm": 0.7197226881980896,
+      "learning_rate": 0.00017215780898916045,
+      "loss": 2.725,
+      "step": 4944
+    },
+    {
+      "epoch": 0.39907997740295376,
+      "grad_norm": 0.8391285538673401,
+      "learning_rate": 0.00017214687829885934,
+      "loss": 2.6724,
+      "step": 4945
+    },
+    {
+      "epoch": 0.3991606811395368,
+      "grad_norm": 0.7357564568519592,
+      "learning_rate": 0.00017213594581046938,
+      "loss": 2.7052,
+      "step": 4946
+    },
+    {
+      "epoch": 0.39924138487611976,
+      "grad_norm": 0.7611483931541443,
+      "learning_rate": 0.00017212501152426312,
+      "loss": 2.7214,
+      "step": 4947
+    },
+    {
+      "epoch": 0.39932208861270274,
+      "grad_norm": 0.7314950227737427,
+      "learning_rate": 0.00017211407544051306,
+      "loss": 2.6594,
+      "step": 4948
+    },
+    {
+      "epoch": 0.3994027923492858,
+      "grad_norm": 0.774131178855896,
+      "learning_rate": 0.00017210313755949169,
+      "loss": 2.6812,
+      "step": 4949
+    },
+    {
+      "epoch": 0.39948349608586875,
+      "grad_norm": 0.707003116607666,
+      "learning_rate": 0.00017209219788147167,
+      "loss": 2.7334,
+      "step": 4950
+    },
+    {
+      "epoch": 0.3995641998224518,
+      "grad_norm": 0.8179643154144287,
+      "learning_rate": 0.0001720812564067256,
+      "loss": 2.6554,
+      "step": 4951
+    },
+    {
+      "epoch": 0.39964490355903476,
+      "grad_norm": 0.6572005152702332,
+      "learning_rate": 0.00017207031313552621,
+      "loss": 2.6423,
+      "step": 4952
+    },
+    {
+      "epoch": 0.3997256072956178,
+      "grad_norm": 0.7663072943687439,
+      "learning_rate": 0.00017205936806814623,
+      "loss": 2.689,
+      "step": 4953
+    },
+    {
+      "epoch": 0.39980631103220077,
+      "grad_norm": 0.7351107001304626,
+      "learning_rate": 0.00017204842120485846,
+      "loss": 2.631,
+      "step": 4954
+    },
+    {
+      "epoch": 0.3998870147687838,
+      "grad_norm": 0.7754253149032593,
+      "learning_rate": 0.00017203747254593564,
+      "loss": 2.6371,
+      "step": 4955
+    },
+    {
+      "epoch": 0.3999677185053668,
+      "grad_norm": 0.7471042275428772,
+      "learning_rate": 0.00017202652209165074,
+      "loss": 2.6542,
+      "step": 4956
+    },
+    {
+      "epoch": 0.4000484222419498,
+      "grad_norm": 0.7357343435287476,
+      "learning_rate": 0.00017201556984227664,
+      "loss": 2.6226,
+      "step": 4957
+    },
+    {
+      "epoch": 0.4001291259785328,
+      "grad_norm": 0.8096252679824829,
+      "learning_rate": 0.00017200461579808626,
+      "loss": 2.6458,
+      "step": 4958
+    },
+    {
+      "epoch": 0.4002098297151158,
+      "grad_norm": 0.7622970938682556,
+      "learning_rate": 0.0001719936599593526,
+      "loss": 2.7129,
+      "step": 4959
+    },
+    {
+      "epoch": 0.4002905334516988,
+      "grad_norm": 0.7374953627586365,
+      "learning_rate": 0.00017198270232634882,
+      "loss": 2.696,
+      "step": 4960
+    },
+    {
+      "epoch": 0.4003712371882818,
+      "grad_norm": 0.7897924184799194,
+      "learning_rate": 0.00017197174289934787,
+      "loss": 2.7508,
+      "step": 4961
+    },
+    {
+      "epoch": 0.4004519409248648,
+      "grad_norm": 0.7047984004020691,
+      "learning_rate": 0.00017196078167862298,
+      "loss": 2.6733,
+      "step": 4962
+    },
+    {
+      "epoch": 0.40053264466144783,
+      "grad_norm": 0.7866294980049133,
+      "learning_rate": 0.0001719498186644473,
+      "loss": 2.694,
+      "step": 4963
+    },
+    {
+      "epoch": 0.4006133483980308,
+      "grad_norm": 0.739923894405365,
+      "learning_rate": 0.00017193885385709409,
+      "loss": 2.7125,
+      "step": 4964
+    },
+    {
+      "epoch": 0.40069405213461384,
+      "grad_norm": 0.7506374716758728,
+      "learning_rate": 0.00017192788725683652,
+      "loss": 2.627,
+      "step": 4965
+    },
+    {
+      "epoch": 0.4007747558711968,
+      "grad_norm": 0.6591607928276062,
+      "learning_rate": 0.00017191691886394802,
+      "loss": 2.6723,
+      "step": 4966
+    },
+    {
+      "epoch": 0.40085545960777985,
+      "grad_norm": 0.7748788595199585,
+      "learning_rate": 0.00017190594867870192,
+      "loss": 2.6486,
+      "step": 4967
+    },
+    {
+      "epoch": 0.40093616334436283,
+      "grad_norm": 0.7518232464790344,
+      "learning_rate": 0.0001718949767013716,
+      "loss": 2.6879,
+      "step": 4968
+    },
+    {
+      "epoch": 0.40101686708094586,
+      "grad_norm": 0.7360039949417114,
+      "learning_rate": 0.00017188400293223052,
+      "loss": 2.6506,
+      "step": 4969
+    },
+    {
+      "epoch": 0.40109757081752884,
+      "grad_norm": 0.7217130064964294,
+      "learning_rate": 0.0001718730273715522,
+      "loss": 2.6263,
+      "step": 4970
+    },
+    {
+      "epoch": 0.40117827455411187,
+      "grad_norm": 0.7246078252792358,
+      "learning_rate": 0.00017186205001961015,
+      "loss": 2.6222,
+      "step": 4971
+    },
+    {
+      "epoch": 0.40125897829069485,
+      "grad_norm": 0.7566879391670227,
+      "learning_rate": 0.00017185107087667794,
+      "loss": 2.7003,
+      "step": 4972
+    },
+    {
+      "epoch": 0.4013396820272779,
+      "grad_norm": 0.7881271243095398,
+      "learning_rate": 0.00017184008994302924,
+      "loss": 2.6463,
+      "step": 4973
+    },
+    {
+      "epoch": 0.40142038576386085,
+      "grad_norm": 0.7307420372962952,
+      "learning_rate": 0.00017182910721893775,
+      "loss": 2.667,
+      "step": 4974
+    },
+    {
+      "epoch": 0.4015010895004439,
+      "grad_norm": 0.7088132500648499,
+      "learning_rate": 0.00017181812270467708,
+      "loss": 2.6073,
+      "step": 4975
+    },
+    {
+      "epoch": 0.40158179323702686,
+      "grad_norm": 0.7839647531509399,
+      "learning_rate": 0.0001718071364005211,
+      "loss": 2.6594,
+      "step": 4976
+    },
+    {
+      "epoch": 0.4016624969736099,
+      "grad_norm": 0.7472013235092163,
+      "learning_rate": 0.00017179614830674353,
+      "loss": 2.737,
+      "step": 4977
+    },
+    {
+      "epoch": 0.40174320071019287,
+      "grad_norm": 0.7241616249084473,
+      "learning_rate": 0.0001717851584236183,
+      "loss": 2.6615,
+      "step": 4978
+    },
+    {
+      "epoch": 0.4018239044467759,
+      "grad_norm": 0.7918941378593445,
+      "learning_rate": 0.00017177416675141929,
+      "loss": 2.6774,
+      "step": 4979
+    },
+    {
+      "epoch": 0.4019046081833589,
+      "grad_norm": 0.801003098487854,
+      "learning_rate": 0.00017176317329042039,
+      "loss": 2.6749,
+      "step": 4980
+    },
+    {
+      "epoch": 0.4019853119199419,
+      "grad_norm": 0.7556802034378052,
+      "learning_rate": 0.00017175217804089564,
+      "loss": 2.6197,
+      "step": 4981
+    },
+    {
+      "epoch": 0.4020660156565249,
+      "grad_norm": 0.7539604902267456,
+      "learning_rate": 0.00017174118100311904,
+      "loss": 2.6222,
+      "step": 4982
+    },
+    {
+      "epoch": 0.4021467193931079,
+      "grad_norm": 0.741436243057251,
+      "learning_rate": 0.0001717301821773647,
+      "loss": 2.6471,
+      "step": 4983
+    },
+    {
+      "epoch": 0.4022274231296909,
+      "grad_norm": 0.7449339628219604,
+      "learning_rate": 0.0001717191815639067,
+      "loss": 2.6448,
+      "step": 4984
+    },
+    {
+      "epoch": 0.40230812686627393,
+      "grad_norm": 0.7771497964859009,
+      "learning_rate": 0.0001717081791630192,
+      "loss": 2.673,
+      "step": 4985
+    },
+    {
+      "epoch": 0.4023888306028569,
+      "grad_norm": 0.6916669607162476,
+      "learning_rate": 0.00017169717497497646,
+      "loss": 2.6025,
+      "step": 4986
+    },
+    {
+      "epoch": 0.40246953433943994,
+      "grad_norm": 0.7373276948928833,
+      "learning_rate": 0.0001716861690000527,
+      "loss": 2.6783,
+      "step": 4987
+    },
+    {
+      "epoch": 0.4025502380760229,
+      "grad_norm": 0.7756158709526062,
+      "learning_rate": 0.0001716751612385222,
+      "loss": 2.7296,
+      "step": 4988
+    },
+    {
+      "epoch": 0.40263094181260595,
+      "grad_norm": 0.7725681066513062,
+      "learning_rate": 0.00017166415169065933,
+      "loss": 2.7169,
+      "step": 4989
+    },
+    {
+      "epoch": 0.4027116455491889,
+      "grad_norm": 0.7165024280548096,
+      "learning_rate": 0.00017165314035673846,
+      "loss": 2.677,
+      "step": 4990
+    },
+    {
+      "epoch": 0.40279234928577196,
+      "grad_norm": 0.8888981938362122,
+      "learning_rate": 0.00017164212723703404,
+      "loss": 2.7694,
+      "step": 4991
+    },
+    {
+      "epoch": 0.40287305302235493,
+      "grad_norm": 0.7439224720001221,
+      "learning_rate": 0.00017163111233182052,
+      "loss": 2.674,
+      "step": 4992
+    },
+    {
+      "epoch": 0.40295375675893796,
+      "grad_norm": 0.6948431730270386,
+      "learning_rate": 0.00017162009564137244,
+      "loss": 2.6595,
+      "step": 4993
+    },
+    {
+      "epoch": 0.40303446049552094,
+      "grad_norm": 0.7274380922317505,
+      "learning_rate": 0.00017160907716596438,
+      "loss": 2.649,
+      "step": 4994
+    },
+    {
+      "epoch": 0.403115164232104,
+      "grad_norm": 0.7127148509025574,
+      "learning_rate": 0.0001715980569058709,
+      "loss": 2.6883,
+      "step": 4995
+    },
+    {
+      "epoch": 0.40319586796868695,
+      "grad_norm": 0.7129155993461609,
+      "learning_rate": 0.00017158703486136668,
+      "loss": 2.6516,
+      "step": 4996
+    },
+    {
+      "epoch": 0.40327657170527,
+      "grad_norm": 0.7848126292228699,
+      "learning_rate": 0.00017157601103272646,
+      "loss": 2.6778,
+      "step": 4997
+    },
+    {
+      "epoch": 0.40335727544185296,
+      "grad_norm": 0.752268373966217,
+      "learning_rate": 0.0001715649854202249,
+      "loss": 2.7228,
+      "step": 4998
+    },
+    {
+      "epoch": 0.40343797917843593,
+      "grad_norm": 0.7750338912010193,
+      "learning_rate": 0.00017155395802413684,
+      "loss": 2.6338,
+      "step": 4999
+    },
+    {
+      "epoch": 0.40351868291501897,
+      "grad_norm": 0.7165457010269165,
+      "learning_rate": 0.00017154292884473713,
+      "loss": 2.6195,
+      "step": 5000
+    },
+    {
+      "epoch": 0.40351868291501897,
+      "eval_loss": 2.585501194000244,
+      "eval_runtime": 901.8519,
+      "eval_samples_per_second": 2.905,
+      "eval_steps_per_second": 0.485,
+      "step": 5000
+    },
+    {
+      "epoch": 0.40359938665160194,
+      "grad_norm": 0.8118943572044373,
+      "learning_rate": 0.00017153189788230062,
+      "loss": 2.6649,
+      "step": 5001
+    },
+    {
+      "epoch": 0.403680090388185,
+      "grad_norm": 0.722984790802002,
+      "learning_rate": 0.00017152086513710221,
+      "loss": 2.6929,
+      "step": 5002
+    },
+    {
+      "epoch": 0.40376079412476795,
+      "grad_norm": 0.700690507888794,
+      "learning_rate": 0.00017150983060941686,
+      "loss": 2.6368,
+      "step": 5003
+    },
+    {
+      "epoch": 0.403841497861351,
+      "grad_norm": 0.7331504225730896,
+      "learning_rate": 0.00017149879429951965,
+      "loss": 2.6826,
+      "step": 5004
+    },
+    {
+      "epoch": 0.40392220159793396,
+      "grad_norm": 0.7312643527984619,
+      "learning_rate": 0.00017148775620768553,
+      "loss": 2.6279,
+      "step": 5005
+    },
+    {
+      "epoch": 0.404002905334517,
+      "grad_norm": 0.7488462924957275,
+      "learning_rate": 0.00017147671633418972,
+      "loss": 2.6711,
+      "step": 5006
+    },
+    {
+      "epoch": 0.40408360907109997,
+      "grad_norm": 0.8620340824127197,
+      "learning_rate": 0.00017146567467930725,
+      "loss": 2.6637,
+      "step": 5007
+    },
+    {
+      "epoch": 0.404164312807683,
+      "grad_norm": 0.683907151222229,
+      "learning_rate": 0.00017145463124331335,
+      "loss": 2.6331,
+      "step": 5008
+    },
+    {
+      "epoch": 0.404245016544266,
+      "grad_norm": 0.7389389276504517,
+      "learning_rate": 0.0001714435860264833,
+      "loss": 2.7232,
+      "step": 5009
+    },
+    {
+      "epoch": 0.404325720280849,
+      "grad_norm": 0.7456515431404114,
+      "learning_rate": 0.00017143253902909228,
+      "loss": 2.6363,
+      "step": 5010
+    },
+    {
+      "epoch": 0.404406424017432,
+      "grad_norm": 0.7044962644577026,
+      "learning_rate": 0.0001714214902514157,
+      "loss": 2.6672,
+      "step": 5011
+    },
+    {
+      "epoch": 0.404487127754015,
+      "grad_norm": 0.7410328984260559,
+      "learning_rate": 0.00017141043969372887,
+      "loss": 2.6059,
+      "step": 5012
+    },
+    {
+      "epoch": 0.404567831490598,
+      "grad_norm": 0.6697140336036682,
+      "learning_rate": 0.00017139938735630722,
+      "loss": 2.7151,
+      "step": 5013
+    },
+    {
+      "epoch": 0.404648535227181,
+      "grad_norm": 0.746675431728363,
+      "learning_rate": 0.00017138833323942617,
+      "loss": 2.6792,
+      "step": 5014
+    },
+    {
+      "epoch": 0.404729238963764,
+      "grad_norm": 0.7724997401237488,
+      "learning_rate": 0.00017137727734336129,
+      "loss": 2.6234,
+      "step": 5015
+    },
+    {
+      "epoch": 0.40480994270034704,
+      "grad_norm": 0.8014429211616516,
+      "learning_rate": 0.00017136621966838805,
+      "loss": 2.6795,
+      "step": 5016
+    },
+    {
+      "epoch": 0.40489064643693,
+      "grad_norm": 0.6900430917739868,
+      "learning_rate": 0.00017135516021478205,
+      "loss": 2.7127,
+      "step": 5017
+    },
+    {
+      "epoch": 0.40497135017351304,
+      "grad_norm": 0.6648666858673096,
+      "learning_rate": 0.00017134409898281896,
+      "loss": 2.6564,
+      "step": 5018
+    },
+    {
+      "epoch": 0.405052053910096,
+      "grad_norm": 0.7054181098937988,
+      "learning_rate": 0.00017133303597277442,
+      "loss": 2.6652,
+      "step": 5019
+    },
+    {
+      "epoch": 0.40513275764667905,
+      "grad_norm": 0.6847733855247498,
+      "learning_rate": 0.00017132197118492414,
+      "loss": 2.6997,
+      "step": 5020
+    },
+    {
+      "epoch": 0.40521346138326203,
+      "grad_norm": 0.7047749757766724,
+      "learning_rate": 0.00017131090461954392,
+      "loss": 2.6752,
+      "step": 5021
+    },
+    {
+      "epoch": 0.40529416511984506,
+      "grad_norm": 0.7549976706504822,
+      "learning_rate": 0.00017129983627690957,
+      "loss": 2.6736,
+      "step": 5022
+    },
+    {
+      "epoch": 0.40537486885642804,
+      "grad_norm": 0.7436367273330688,
+      "learning_rate": 0.00017128876615729686,
+      "loss": 2.7189,
+      "step": 5023
+    },
+    {
+      "epoch": 0.40545557259301107,
+      "grad_norm": 0.6515071988105774,
+      "learning_rate": 0.00017127769426098177,
+      "loss": 2.6422,
+      "step": 5024
+    },
+    {
+      "epoch": 0.40553627632959405,
+      "grad_norm": 0.6960858702659607,
+      "learning_rate": 0.00017126662058824024,
+      "loss": 2.6619,
+      "step": 5025
+    },
+    {
+      "epoch": 0.4056169800661771,
+      "grad_norm": 0.8075968623161316,
+      "learning_rate": 0.0001712555451393482,
+      "loss": 2.6678,
+      "step": 5026
+    },
+    {
+      "epoch": 0.40569768380276006,
+      "grad_norm": 0.6864624619483948,
+      "learning_rate": 0.00017124446791458176,
+      "loss": 2.6331,
+      "step": 5027
+    },
+    {
+      "epoch": 0.4057783875393431,
+      "grad_norm": 0.7218763828277588,
+      "learning_rate": 0.0001712333889142169,
+      "loss": 2.6316,
+      "step": 5028
+    },
+    {
+      "epoch": 0.40585909127592606,
+      "grad_norm": 0.7024715542793274,
+      "learning_rate": 0.0001712223081385298,
+      "loss": 2.623,
+      "step": 5029
+    },
+    {
+      "epoch": 0.4059397950125091,
+      "grad_norm": 0.6681575775146484,
+      "learning_rate": 0.0001712112255877966,
+      "loss": 2.6786,
+      "step": 5030
+    },
+    {
+      "epoch": 0.4060204987490921,
+      "grad_norm": 0.7249817848205566,
+      "learning_rate": 0.0001712001412622935,
+      "loss": 2.6179,
+      "step": 5031
+    },
+    {
+      "epoch": 0.4061012024856751,
+      "grad_norm": 0.7178316116333008,
+      "learning_rate": 0.00017118905516229677,
+      "loss": 2.696,
+      "step": 5032
+    },
+    {
+      "epoch": 0.4061819062222581,
+      "grad_norm": 0.7838767766952515,
+      "learning_rate": 0.0001711779672880827,
+      "loss": 2.6881,
+      "step": 5033
+    },
+    {
+      "epoch": 0.4062626099588411,
+      "grad_norm": 0.799937903881073,
+      "learning_rate": 0.0001711668776399276,
+      "loss": 2.7587,
+      "step": 5034
+    },
+    {
+      "epoch": 0.4063433136954241,
+      "grad_norm": 0.7622246146202087,
+      "learning_rate": 0.0001711557862181079,
+      "loss": 2.6621,
+      "step": 5035
+    },
+    {
+      "epoch": 0.4064240174320071,
+      "grad_norm": 0.7158814072608948,
+      "learning_rate": 0.00017114469302290003,
+      "loss": 2.6421,
+      "step": 5036
+    },
+    {
+      "epoch": 0.4065047211685901,
+      "grad_norm": 0.7913404107093811,
+      "learning_rate": 0.0001711335980545804,
+      "loss": 2.6323,
+      "step": 5037
+    },
+    {
+      "epoch": 0.40658542490517313,
+      "grad_norm": 0.718325138092041,
+      "learning_rate": 0.00017112250131342556,
+      "loss": 2.6171,
+      "step": 5038
+    },
+    {
+      "epoch": 0.4066661286417561,
+      "grad_norm": 0.7793646454811096,
+      "learning_rate": 0.0001711114027997121,
+      "loss": 2.7494,
+      "step": 5039
+    },
+    {
+      "epoch": 0.40674683237833914,
+      "grad_norm": 0.7774816155433655,
+      "learning_rate": 0.00017110030251371656,
+      "loss": 2.5534,
+      "step": 5040
+    },
+    {
+      "epoch": 0.4068275361149221,
+      "grad_norm": 0.8547549247741699,
+      "learning_rate": 0.00017108920045571564,
+      "loss": 2.7155,
+      "step": 5041
+    },
+    {
+      "epoch": 0.40690823985150515,
+      "grad_norm": 0.7685851454734802,
+      "learning_rate": 0.000171078096625986,
+      "loss": 2.6109,
+      "step": 5042
+    },
+    {
+      "epoch": 0.4069889435880881,
+      "grad_norm": 0.7953611016273499,
+      "learning_rate": 0.00017106699102480445,
+      "loss": 2.7034,
+      "step": 5043
+    },
+    {
+      "epoch": 0.40706964732467116,
+      "grad_norm": 0.7550730109214783,
+      "learning_rate": 0.00017105588365244764,
+      "loss": 2.7026,
+      "step": 5044
+    },
+    {
+      "epoch": 0.40715035106125413,
+      "grad_norm": 0.7036548256874084,
+      "learning_rate": 0.0001710447745091925,
+      "loss": 2.6246,
+      "step": 5045
+    },
+    {
+      "epoch": 0.40723105479783717,
+      "grad_norm": 0.7154512405395508,
+      "learning_rate": 0.00017103366359531586,
+      "loss": 2.6592,
+      "step": 5046
+    },
+    {
+      "epoch": 0.40731175853442014,
+      "grad_norm": 0.7773932218551636,
+      "learning_rate": 0.00017102255091109463,
+      "loss": 2.6458,
+      "step": 5047
+    },
+    {
+      "epoch": 0.4073924622710032,
+      "grad_norm": 0.7458996176719666,
+      "learning_rate": 0.0001710114364568058,
+      "loss": 2.643,
+      "step": 5048
+    },
+    {
+      "epoch": 0.40747316600758615,
+      "grad_norm": 0.7465376257896423,
+      "learning_rate": 0.00017100032023272633,
+      "loss": 2.6677,
+      "step": 5049
+    },
+    {
+      "epoch": 0.40755386974416913,
+      "grad_norm": 0.7340850830078125,
+      "learning_rate": 0.0001709892022391333,
+      "loss": 2.6372,
+      "step": 5050
+    },
+    {
+      "epoch": 0.40763457348075216,
+      "grad_norm": 0.7189164757728577,
+      "learning_rate": 0.00017097808247630377,
+      "loss": 2.6524,
+      "step": 5051
+    },
+    {
+      "epoch": 0.40771527721733514,
+      "grad_norm": 0.6954184174537659,
+      "learning_rate": 0.0001709669609445149,
+      "loss": 2.7383,
+      "step": 5052
+    },
+    {
+      "epoch": 0.40779598095391817,
+      "grad_norm": 0.736409604549408,
+      "learning_rate": 0.00017095583764404384,
+      "loss": 2.6424,
+      "step": 5053
+    },
+    {
+      "epoch": 0.40787668469050115,
+      "grad_norm": 0.6773545742034912,
+      "learning_rate": 0.0001709447125751678,
+      "loss": 2.6557,
+      "step": 5054
+    },
+    {
+      "epoch": 0.4079573884270842,
+      "grad_norm": 0.718748927116394,
+      "learning_rate": 0.00017093358573816412,
+      "loss": 2.6884,
+      "step": 5055
+    },
+    {
+      "epoch": 0.40803809216366715,
+      "grad_norm": 0.8276848793029785,
+      "learning_rate": 0.00017092245713331002,
+      "loss": 2.6642,
+      "step": 5056
+    },
+    {
+      "epoch": 0.4081187959002502,
+      "grad_norm": 0.7694761157035828,
+      "learning_rate": 0.00017091132676088294,
+      "loss": 2.644,
+      "step": 5057
+    },
+    {
+      "epoch": 0.40819949963683316,
+      "grad_norm": 0.766724705696106,
+      "learning_rate": 0.0001709001946211602,
+      "loss": 2.6918,
+      "step": 5058
+    },
+    {
+      "epoch": 0.4082802033734162,
+      "grad_norm": 0.7067074775695801,
+      "learning_rate": 0.00017088906071441927,
+      "loss": 2.7228,
+      "step": 5059
+    },
+    {
+      "epoch": 0.40836090710999917,
+      "grad_norm": 0.7216899991035461,
+      "learning_rate": 0.00017087792504093767,
+      "loss": 2.7068,
+      "step": 5060
+    },
+    {
+      "epoch": 0.4084416108465822,
+      "grad_norm": 0.6728984713554382,
+      "learning_rate": 0.00017086678760099287,
+      "loss": 2.686,
+      "step": 5061
+    },
+    {
+      "epoch": 0.4085223145831652,
+      "grad_norm": 0.7546882033348083,
+      "learning_rate": 0.0001708556483948625,
+      "loss": 2.6907,
+      "step": 5062
+    },
+    {
+      "epoch": 0.4086030183197482,
+      "grad_norm": 0.7471179962158203,
+      "learning_rate": 0.00017084450742282416,
+      "loss": 2.6857,
+      "step": 5063
+    },
+    {
+      "epoch": 0.4086837220563312,
+      "grad_norm": 0.7879743576049805,
+      "learning_rate": 0.00017083336468515548,
+      "loss": 2.7224,
+      "step": 5064
+    },
+    {
+      "epoch": 0.4087644257929142,
+      "grad_norm": 0.691343367099762,
+      "learning_rate": 0.00017082222018213422,
+      "loss": 2.6561,
+      "step": 5065
+    },
+    {
+      "epoch": 0.4088451295294972,
+      "grad_norm": 0.7497386336326599,
+      "learning_rate": 0.00017081107391403805,
+      "loss": 2.6317,
+      "step": 5066
+    },
+    {
+      "epoch": 0.40892583326608023,
+      "grad_norm": 0.6846269965171814,
+      "learning_rate": 0.00017079992588114485,
+      "loss": 2.6522,
+      "step": 5067
+    },
+    {
+      "epoch": 0.4090065370026632,
+      "grad_norm": 0.7312905192375183,
+      "learning_rate": 0.0001707887760837324,
+      "loss": 2.588,
+      "step": 5068
+    },
+    {
+      "epoch": 0.40908724073924624,
+      "grad_norm": 0.6966867446899414,
+      "learning_rate": 0.00017077762452207866,
+      "loss": 2.6316,
+      "step": 5069
+    },
+    {
+      "epoch": 0.4091679444758292,
+      "grad_norm": 0.6882073283195496,
+      "learning_rate": 0.00017076647119646147,
+      "loss": 2.6977,
+      "step": 5070
+    },
+    {
+      "epoch": 0.40924864821241225,
+      "grad_norm": 0.7392483949661255,
+      "learning_rate": 0.00017075531610715884,
+      "loss": 2.6768,
+      "step": 5071
+    },
+    {
+      "epoch": 0.4093293519489952,
+      "grad_norm": 0.7311073541641235,
+      "learning_rate": 0.00017074415925444876,
+      "loss": 2.6628,
+      "step": 5072
+    },
+    {
+      "epoch": 0.40941005568557826,
+      "grad_norm": 0.6769934296607971,
+      "learning_rate": 0.00017073300063860934,
+      "loss": 2.6438,
+      "step": 5073
+    },
+    {
+      "epoch": 0.40949075942216123,
+      "grad_norm": 0.736456573009491,
+      "learning_rate": 0.00017072184025991862,
+      "loss": 2.6151,
+      "step": 5074
+    },
+    {
+      "epoch": 0.40957146315874426,
+      "grad_norm": 0.7026283740997314,
+      "learning_rate": 0.00017071067811865476,
+      "loss": 2.6726,
+      "step": 5075
+    },
+    {
+      "epoch": 0.40965216689532724,
+      "grad_norm": 0.6825234293937683,
+      "learning_rate": 0.00017069951421509597,
+      "loss": 2.6795,
+      "step": 5076
+    },
+    {
+      "epoch": 0.4097328706319103,
+      "grad_norm": 0.7243828773498535,
+      "learning_rate": 0.0001706883485495205,
+      "loss": 2.687,
+      "step": 5077
+    },
+    {
+      "epoch": 0.40981357436849325,
+      "grad_norm": 0.7300469875335693,
+      "learning_rate": 0.00017067718112220658,
+      "loss": 2.6268,
+      "step": 5078
+    },
+    {
+      "epoch": 0.4098942781050763,
+      "grad_norm": 0.698095440864563,
+      "learning_rate": 0.00017066601193343255,
+      "loss": 2.6461,
+      "step": 5079
+    },
+    {
+      "epoch": 0.40997498184165926,
+      "grad_norm": 0.7318777441978455,
+      "learning_rate": 0.00017065484098347677,
+      "loss": 2.6817,
+      "step": 5080
+    },
+    {
+      "epoch": 0.4100556855782423,
+      "grad_norm": 0.7681582570075989,
+      "learning_rate": 0.00017064366827261772,
+      "loss": 2.7309,
+      "step": 5081
+    },
+    {
+      "epoch": 0.41013638931482527,
+      "grad_norm": 0.7690179944038391,
+      "learning_rate": 0.0001706324938011337,
+      "loss": 2.6292,
+      "step": 5082
+    },
+    {
+      "epoch": 0.4102170930514083,
+      "grad_norm": 0.6745284199714661,
+      "learning_rate": 0.00017062131756930338,
+      "loss": 2.7133,
+      "step": 5083
+    },
+    {
+      "epoch": 0.4102977967879913,
+      "grad_norm": 0.7524279952049255,
+      "learning_rate": 0.00017061013957740518,
+      "loss": 2.6237,
+      "step": 5084
+    },
+    {
+      "epoch": 0.4103785005245743,
+      "grad_norm": 0.7813692092895508,
+      "learning_rate": 0.00017059895982571773,
+      "loss": 2.6953,
+      "step": 5085
+    },
+    {
+      "epoch": 0.4104592042611573,
+      "grad_norm": 0.7128829956054688,
+      "learning_rate": 0.00017058777831451967,
+      "loss": 2.6771,
+      "step": 5086
+    },
+    {
+      "epoch": 0.4105399079977403,
+      "grad_norm": 0.7249834537506104,
+      "learning_rate": 0.00017057659504408963,
+      "loss": 2.6376,
+      "step": 5087
+    },
+    {
+      "epoch": 0.4106206117343233,
+      "grad_norm": 0.7742593288421631,
+      "learning_rate": 0.00017056541001470637,
+      "loss": 2.6227,
+      "step": 5088
+    },
+    {
+      "epoch": 0.4107013154709063,
+      "grad_norm": 0.6994228959083557,
+      "learning_rate": 0.00017055422322664863,
+      "loss": 2.6573,
+      "step": 5089
+    },
+    {
+      "epoch": 0.4107820192074893,
+      "grad_norm": 0.7144249081611633,
+      "learning_rate": 0.00017054303468019518,
+      "loss": 2.6602,
+      "step": 5090
+    },
+    {
+      "epoch": 0.41086272294407233,
+      "grad_norm": 0.7695099711418152,
+      "learning_rate": 0.00017053184437562497,
+      "loss": 2.6516,
+      "step": 5091
+    },
+    {
+      "epoch": 0.4109434266806553,
+      "grad_norm": 0.7610031962394714,
+      "learning_rate": 0.00017052065231321678,
+      "loss": 2.6963,
+      "step": 5092
+    },
+    {
+      "epoch": 0.41102413041723834,
+      "grad_norm": 0.7117859721183777,
+      "learning_rate": 0.0001705094584932496,
+      "loss": 2.6954,
+      "step": 5093
+    },
+    {
+      "epoch": 0.4111048341538213,
+      "grad_norm": 0.7891486287117004,
+      "learning_rate": 0.00017049826291600244,
+      "loss": 2.7265,
+      "step": 5094
+    },
+    {
+      "epoch": 0.41118553789040435,
+      "grad_norm": 0.7347370386123657,
+      "learning_rate": 0.00017048706558175423,
+      "loss": 2.658,
+      "step": 5095
+    },
+    {
+      "epoch": 0.41126624162698733,
+      "grad_norm": 0.7541289925575256,
+      "learning_rate": 0.00017047586649078414,
+      "loss": 2.6596,
+      "step": 5096
+    },
+    {
+      "epoch": 0.41134694536357036,
+      "grad_norm": 0.7471255660057068,
+      "learning_rate": 0.00017046466564337118,
+      "loss": 2.7008,
+      "step": 5097
+    },
+    {
+      "epoch": 0.41142764910015334,
+      "grad_norm": 0.7566937208175659,
+      "learning_rate": 0.00017045346303979457,
+      "loss": 2.7006,
+      "step": 5098
+    },
+    {
+      "epoch": 0.41150835283673637,
+      "grad_norm": 0.6991304159164429,
+      "learning_rate": 0.00017044225868033353,
+      "loss": 2.6846,
+      "step": 5099
+    },
+    {
+      "epoch": 0.41158905657331935,
+      "grad_norm": 0.7286314368247986,
+      "learning_rate": 0.00017043105256526724,
+      "loss": 2.6219,
+      "step": 5100
+    },
+    {
+      "epoch": 0.4116697603099023,
+      "grad_norm": 0.6953727006912231,
+      "learning_rate": 0.000170419844694875,
+      "loss": 2.6093,
+      "step": 5101
+    },
+    {
+      "epoch": 0.41175046404648535,
+      "grad_norm": 0.6942756772041321,
+      "learning_rate": 0.00017040863506943615,
+      "loss": 2.6399,
+      "step": 5102
+    },
+    {
+      "epoch": 0.41183116778306833,
+      "grad_norm": 0.7513531446456909,
+      "learning_rate": 0.00017039742368923005,
+      "loss": 2.6187,
+      "step": 5103
+    },
+    {
+      "epoch": 0.41191187151965136,
+      "grad_norm": 0.7530633211135864,
+      "learning_rate": 0.00017038621055453617,
+      "loss": 2.6124,
+      "step": 5104
+    },
+    {
+      "epoch": 0.41199257525623434,
+      "grad_norm": 0.7487555146217346,
+      "learning_rate": 0.00017037499566563392,
+      "loss": 2.6331,
+      "step": 5105
+    },
+    {
+      "epoch": 0.41207327899281737,
+      "grad_norm": 0.7641858458518982,
+      "learning_rate": 0.00017036377902280282,
+      "loss": 2.6875,
+      "step": 5106
+    },
+    {
+      "epoch": 0.41215398272940035,
+      "grad_norm": 0.6962767839431763,
+      "learning_rate": 0.0001703525606263224,
+      "loss": 2.6538,
+      "step": 5107
+    },
+    {
+      "epoch": 0.4122346864659834,
+      "grad_norm": 0.8183409571647644,
+      "learning_rate": 0.0001703413404764723,
+      "loss": 2.6204,
+      "step": 5108
+    },
+    {
+      "epoch": 0.41231539020256636,
+      "grad_norm": 0.7029808759689331,
+      "learning_rate": 0.00017033011857353207,
+      "loss": 2.6369,
+      "step": 5109
+    },
+    {
+      "epoch": 0.4123960939391494,
+      "grad_norm": 0.7171663045883179,
+      "learning_rate": 0.00017031889491778149,
+      "loss": 2.6211,
+      "step": 5110
+    },
+    {
+      "epoch": 0.41247679767573237,
+      "grad_norm": 0.7456090450286865,
+      "learning_rate": 0.0001703076695095002,
+      "loss": 2.6574,
+      "step": 5111
+    },
+    {
+      "epoch": 0.4125575014123154,
+      "grad_norm": 0.7468575239181519,
+      "learning_rate": 0.000170296442348968,
+      "loss": 2.598,
+      "step": 5112
+    },
+    {
+      "epoch": 0.4126382051488984,
+      "grad_norm": 0.7106603384017944,
+      "learning_rate": 0.0001702852134364647,
+      "loss": 2.6577,
+      "step": 5113
+    },
+    {
+      "epoch": 0.4127189088854814,
+      "grad_norm": 0.7788330912590027,
+      "learning_rate": 0.00017027398277227017,
+      "loss": 2.6797,
+      "step": 5114
+    },
+    {
+      "epoch": 0.4127996126220644,
+      "grad_norm": 0.7794120907783508,
+      "learning_rate": 0.00017026275035666427,
+      "loss": 2.5834,
+      "step": 5115
+    },
+    {
+      "epoch": 0.4128803163586474,
+      "grad_norm": 0.7270684838294983,
+      "learning_rate": 0.00017025151618992702,
+      "loss": 2.7153,
+      "step": 5116
+    },
+    {
+      "epoch": 0.4129610200952304,
+      "grad_norm": 0.8169006109237671,
+      "learning_rate": 0.00017024028027233827,
+      "loss": 2.6786,
+      "step": 5117
+    },
+    {
+      "epoch": 0.4130417238318134,
+      "grad_norm": 0.8053112626075745,
+      "learning_rate": 0.00017022904260417815,
+      "loss": 2.6456,
+      "step": 5118
+    },
+    {
+      "epoch": 0.4131224275683964,
+      "grad_norm": 0.7646365165710449,
+      "learning_rate": 0.0001702178031857267,
+      "loss": 2.6784,
+      "step": 5119
+    },
+    {
+      "epoch": 0.41320313130497943,
+      "grad_norm": 0.7878902554512024,
+      "learning_rate": 0.00017020656201726406,
+      "loss": 2.66,
+      "step": 5120
+    },
+    {
+      "epoch": 0.4132838350415624,
+      "grad_norm": 0.8602383732795715,
+      "learning_rate": 0.00017019531909907037,
+      "loss": 2.7018,
+      "step": 5121
+    },
+    {
+      "epoch": 0.41336453877814544,
+      "grad_norm": 0.801092267036438,
+      "learning_rate": 0.00017018407443142585,
+      "loss": 2.7728,
+      "step": 5122
+    },
+    {
+      "epoch": 0.4134452425147284,
+      "grad_norm": 0.7372604012489319,
+      "learning_rate": 0.00017017282801461074,
+      "loss": 2.6588,
+      "step": 5123
+    },
+    {
+      "epoch": 0.41352594625131145,
+      "grad_norm": 0.7553830146789551,
+      "learning_rate": 0.0001701615798489053,
+      "loss": 2.6844,
+      "step": 5124
+    },
+    {
+      "epoch": 0.4136066499878944,
+      "grad_norm": 0.7699872255325317,
+      "learning_rate": 0.0001701503299345899,
+      "loss": 2.6523,
+      "step": 5125
+    },
+    {
+      "epoch": 0.41368735372447746,
+      "grad_norm": 0.7087047696113586,
+      "learning_rate": 0.0001701390782719449,
+      "loss": 2.6785,
+      "step": 5126
+    },
+    {
+      "epoch": 0.41376805746106043,
+      "grad_norm": 0.7835792303085327,
+      "learning_rate": 0.0001701278248612507,
+      "loss": 2.7064,
+      "step": 5127
+    },
+    {
+      "epoch": 0.41384876119764347,
+      "grad_norm": 0.7833154201507568,
+      "learning_rate": 0.0001701165697027878,
+      "loss": 2.6552,
+      "step": 5128
+    },
+    {
+      "epoch": 0.41392946493422644,
+      "grad_norm": 0.8240615725517273,
+      "learning_rate": 0.0001701053127968367,
+      "loss": 2.7074,
+      "step": 5129
+    },
+    {
+      "epoch": 0.4140101686708095,
+      "grad_norm": 0.7612149119377136,
+      "learning_rate": 0.0001700940541436779,
+      "loss": 2.7484,
+      "step": 5130
+    },
+    {
+      "epoch": 0.41409087240739245,
+      "grad_norm": 0.7795391082763672,
+      "learning_rate": 0.00017008279374359212,
+      "loss": 2.6022,
+      "step": 5131
+    },
+    {
+      "epoch": 0.4141715761439755,
+      "grad_norm": 0.7714587450027466,
+      "learning_rate": 0.00017007153159685992,
+      "loss": 2.6529,
+      "step": 5132
+    },
+    {
+      "epoch": 0.41425227988055846,
+      "grad_norm": 0.7821317911148071,
+      "learning_rate": 0.00017006026770376194,
+      "loss": 2.6356,
+      "step": 5133
+    },
+    {
+      "epoch": 0.4143329836171415,
+      "grad_norm": 0.7300596833229065,
+      "learning_rate": 0.00017004900206457897,
+      "loss": 2.6552,
+      "step": 5134
+    },
+    {
+      "epoch": 0.41441368735372447,
+      "grad_norm": 0.780505359172821,
+      "learning_rate": 0.00017003773467959174,
+      "loss": 2.675,
+      "step": 5135
+    },
+    {
+      "epoch": 0.4144943910903075,
+      "grad_norm": 0.7107391357421875,
+      "learning_rate": 0.00017002646554908107,
+      "loss": 2.7096,
+      "step": 5136
+    },
+    {
+      "epoch": 0.4145750948268905,
+      "grad_norm": 0.7358834743499756,
+      "learning_rate": 0.0001700151946733279,
+      "loss": 2.6619,
+      "step": 5137
+    },
+    {
+      "epoch": 0.4146557985634735,
+      "grad_norm": 0.7573859095573425,
+      "learning_rate": 0.00017000392205261298,
+      "loss": 2.6234,
+      "step": 5138
+    },
+    {
+      "epoch": 0.4147365023000565,
+      "grad_norm": 0.7032024264335632,
+      "learning_rate": 0.00016999264768721738,
+      "loss": 2.6096,
+      "step": 5139
+    },
+    {
+      "epoch": 0.4148172060366395,
+      "grad_norm": 0.743813693523407,
+      "learning_rate": 0.00016998137157742203,
+      "loss": 2.6782,
+      "step": 5140
+    },
+    {
+      "epoch": 0.4148979097732225,
+      "grad_norm": 0.8861347436904907,
+      "learning_rate": 0.00016997009372350793,
+      "loss": 2.6645,
+      "step": 5141
+    },
+    {
+      "epoch": 0.4149786135098055,
+      "grad_norm": 0.7598684430122375,
+      "learning_rate": 0.00016995881412575623,
+      "loss": 2.649,
+      "step": 5142
+    },
+    {
+      "epoch": 0.4150593172463885,
+      "grad_norm": 0.7535565495491028,
+      "learning_rate": 0.00016994753278444798,
+      "loss": 2.6449,
+      "step": 5143
+    },
+    {
+      "epoch": 0.41514002098297154,
+      "grad_norm": 0.7073138356208801,
+      "learning_rate": 0.0001699362496998644,
+      "loss": 2.6253,
+      "step": 5144
+    },
+    {
+      "epoch": 0.4152207247195545,
+      "grad_norm": 0.7161526679992676,
+      "learning_rate": 0.00016992496487228662,
+      "loss": 2.6623,
+      "step": 5145
+    },
+    {
+      "epoch": 0.41530142845613754,
+      "grad_norm": 0.8284714818000793,
+      "learning_rate": 0.00016991367830199595,
+      "loss": 2.7363,
+      "step": 5146
+    },
+    {
+      "epoch": 0.4153821321927205,
+      "grad_norm": 0.7127673625946045,
+      "learning_rate": 0.0001699023899892737,
+      "loss": 2.6274,
+      "step": 5147
+    },
+    {
+      "epoch": 0.41546283592930355,
+      "grad_norm": 0.7496370673179626,
+      "learning_rate": 0.00016989109993440112,
+      "loss": 2.6364,
+      "step": 5148
+    },
+    {
+      "epoch": 0.41554353966588653,
+      "grad_norm": 0.7616143822669983,
+      "learning_rate": 0.00016987980813765963,
+      "loss": 2.7225,
+      "step": 5149
+    },
+    {
+      "epoch": 0.41562424340246956,
+      "grad_norm": 0.6935909986495972,
+      "learning_rate": 0.00016986851459933067,
+      "loss": 2.6109,
+      "step": 5150
+    },
+    {
+      "epoch": 0.41570494713905254,
+      "grad_norm": 0.721023678779602,
+      "learning_rate": 0.00016985721931969566,
+      "loss": 2.6993,
+      "step": 5151
+    },
+    {
+      "epoch": 0.4157856508756355,
+      "grad_norm": 0.8216699361801147,
+      "learning_rate": 0.00016984592229903617,
+      "loss": 2.6512,
+      "step": 5152
+    },
+    {
+      "epoch": 0.41586635461221855,
+      "grad_norm": 0.7425234913825989,
+      "learning_rate": 0.00016983462353763372,
+      "loss": 2.5903,
+      "step": 5153
+    },
+    {
+      "epoch": 0.4159470583488015,
+      "grad_norm": 0.7292542457580566,
+      "learning_rate": 0.00016982332303576986,
+      "loss": 2.692,
+      "step": 5154
+    },
+    {
+      "epoch": 0.41602776208538456,
+      "grad_norm": 0.7466831803321838,
+      "learning_rate": 0.0001698120207937263,
+      "loss": 2.7145,
+      "step": 5155
+    },
+    {
+      "epoch": 0.41610846582196753,
+      "grad_norm": 0.7271949648857117,
+      "learning_rate": 0.00016980071681178471,
+      "loss": 2.655,
+      "step": 5156
+    },
+    {
+      "epoch": 0.41618916955855056,
+      "grad_norm": 0.7505547404289246,
+      "learning_rate": 0.00016978941109022677,
+      "loss": 2.7167,
+      "step": 5157
+    },
+    {
+      "epoch": 0.41626987329513354,
+      "grad_norm": 0.7307172417640686,
+      "learning_rate": 0.00016977810362933427,
+      "loss": 2.6735,
+      "step": 5158
+    },
+    {
+      "epoch": 0.4163505770317166,
+      "grad_norm": 0.7839170098304749,
+      "learning_rate": 0.00016976679442938904,
+      "loss": 2.6818,
+      "step": 5159
+    },
+    {
+      "epoch": 0.41643128076829955,
+      "grad_norm": 0.7131803631782532,
+      "learning_rate": 0.00016975548349067293,
+      "loss": 2.6921,
+      "step": 5160
+    },
+    {
+      "epoch": 0.4165119845048826,
+      "grad_norm": 0.8129798173904419,
+      "learning_rate": 0.0001697441708134678,
+      "loss": 2.6682,
+      "step": 5161
+    },
+    {
+      "epoch": 0.41659268824146556,
+      "grad_norm": 0.7634746432304382,
+      "learning_rate": 0.00016973285639805563,
+      "loss": 2.6684,
+      "step": 5162
+    },
+    {
+      "epoch": 0.4166733919780486,
+      "grad_norm": 0.7367348074913025,
+      "learning_rate": 0.0001697215402447184,
+      "loss": 2.6424,
+      "step": 5163
+    },
+    {
+      "epoch": 0.41675409571463157,
+      "grad_norm": 0.7235338687896729,
+      "learning_rate": 0.00016971022235373815,
+      "loss": 2.6817,
+      "step": 5164
+    },
+    {
+      "epoch": 0.4168347994512146,
+      "grad_norm": 0.7764291763305664,
+      "learning_rate": 0.0001696989027253969,
+      "loss": 2.6477,
+      "step": 5165
+    },
+    {
+      "epoch": 0.4169155031877976,
+      "grad_norm": 0.8207562565803528,
+      "learning_rate": 0.00016968758135997683,
+      "loss": 2.6408,
+      "step": 5166
+    },
+    {
+      "epoch": 0.4169962069243806,
+      "grad_norm": 0.7291484475135803,
+      "learning_rate": 0.00016967625825776005,
+      "loss": 2.6233,
+      "step": 5167
+    },
+    {
+      "epoch": 0.4170769106609636,
+      "grad_norm": 0.7060603499412537,
+      "learning_rate": 0.0001696649334190288,
+      "loss": 2.6204,
+      "step": 5168
+    },
+    {
+      "epoch": 0.4171576143975466,
+      "grad_norm": 0.7058241963386536,
+      "learning_rate": 0.00016965360684406528,
+      "loss": 2.6212,
+      "step": 5169
+    },
+    {
+      "epoch": 0.4172383181341296,
+      "grad_norm": 0.8248410224914551,
+      "learning_rate": 0.00016964227853315177,
+      "loss": 2.6688,
+      "step": 5170
+    },
+    {
+      "epoch": 0.4173190218707126,
+      "grad_norm": 0.7287606596946716,
+      "learning_rate": 0.0001696309484865707,
+      "loss": 2.6201,
+      "step": 5171
+    },
+    {
+      "epoch": 0.4173997256072956,
+      "grad_norm": 0.7214288115501404,
+      "learning_rate": 0.00016961961670460433,
+      "loss": 2.682,
+      "step": 5172
+    },
+    {
+      "epoch": 0.41748042934387863,
+      "grad_norm": 0.7133594155311584,
+      "learning_rate": 0.00016960828318753516,
+      "loss": 2.7167,
+      "step": 5173
+    },
+    {
+      "epoch": 0.4175611330804616,
+      "grad_norm": 0.6935842633247375,
+      "learning_rate": 0.00016959694793564558,
+      "loss": 2.6134,
+      "step": 5174
+    },
+    {
+      "epoch": 0.41764183681704464,
+      "grad_norm": 0.6863382458686829,
+      "learning_rate": 0.00016958561094921815,
+      "loss": 2.6396,
+      "step": 5175
+    },
+    {
+      "epoch": 0.4177225405536276,
+      "grad_norm": 0.7659433484077454,
+      "learning_rate": 0.0001695742722285354,
+      "loss": 2.6926,
+      "step": 5176
+    },
+    {
+      "epoch": 0.41780324429021065,
+      "grad_norm": 0.6997129917144775,
+      "learning_rate": 0.00016956293177387992,
+      "loss": 2.6983,
+      "step": 5177
+    },
+    {
+      "epoch": 0.41788394802679363,
+      "grad_norm": 0.6784526705741882,
+      "learning_rate": 0.00016955158958553433,
+      "loss": 2.6961,
+      "step": 5178
+    },
+    {
+      "epoch": 0.41796465176337666,
+      "grad_norm": 0.8227884769439697,
+      "learning_rate": 0.00016954024566378132,
+      "loss": 2.7008,
+      "step": 5179
+    },
+    {
+      "epoch": 0.41804535549995964,
+      "grad_norm": 0.7733054757118225,
+      "learning_rate": 0.0001695289000089036,
+      "loss": 2.6615,
+      "step": 5180
+    },
+    {
+      "epoch": 0.41812605923654267,
+      "grad_norm": 0.7077545523643494,
+      "learning_rate": 0.00016951755262118394,
+      "loss": 2.6388,
+      "step": 5181
+    },
+    {
+      "epoch": 0.41820676297312565,
+      "grad_norm": 0.7962050437927246,
+      "learning_rate": 0.00016950620350090513,
+      "loss": 2.7063,
+      "step": 5182
+    },
+    {
+      "epoch": 0.4182874667097087,
+      "grad_norm": 0.6950554847717285,
+      "learning_rate": 0.00016949485264835005,
+      "loss": 2.7076,
+      "step": 5183
+    },
+    {
+      "epoch": 0.41836817044629165,
+      "grad_norm": 0.8546960949897766,
+      "learning_rate": 0.00016948350006380162,
+      "loss": 2.6533,
+      "step": 5184
+    },
+    {
+      "epoch": 0.4184488741828747,
+      "grad_norm": 0.7469324469566345,
+      "learning_rate": 0.00016947214574754272,
+      "loss": 2.5884,
+      "step": 5185
+    },
+    {
+      "epoch": 0.41852957791945766,
+      "grad_norm": 0.7125554084777832,
+      "learning_rate": 0.0001694607896998563,
+      "loss": 2.6448,
+      "step": 5186
+    },
+    {
+      "epoch": 0.4186102816560407,
+      "grad_norm": 0.6998329758644104,
+      "learning_rate": 0.00016944943192102549,
+      "loss": 2.5569,
+      "step": 5187
+    },
+    {
+      "epoch": 0.41869098539262367,
+      "grad_norm": 0.9046749472618103,
+      "learning_rate": 0.00016943807241133328,
+      "loss": 2.7701,
+      "step": 5188
+    },
+    {
+      "epoch": 0.4187716891292067,
+      "grad_norm": 0.7842074036598206,
+      "learning_rate": 0.00016942671117106274,
+      "loss": 2.7124,
+      "step": 5189
+    },
+    {
+      "epoch": 0.4188523928657897,
+      "grad_norm": 0.7625874280929565,
+      "learning_rate": 0.00016941534820049713,
+      "loss": 2.6626,
+      "step": 5190
+    },
+    {
+      "epoch": 0.4189330966023727,
+      "grad_norm": 0.7006461024284363,
+      "learning_rate": 0.00016940398349991957,
+      "loss": 2.6283,
+      "step": 5191
+    },
+    {
+      "epoch": 0.4190138003389557,
+      "grad_norm": 0.7081875205039978,
+      "learning_rate": 0.00016939261706961332,
+      "loss": 2.69,
+      "step": 5192
+    },
+    {
+      "epoch": 0.4190945040755387,
+      "grad_norm": 0.7554503083229065,
+      "learning_rate": 0.00016938124890986166,
+      "loss": 2.641,
+      "step": 5193
+    },
+    {
+      "epoch": 0.4191752078121217,
+      "grad_norm": 0.7478535175323486,
+      "learning_rate": 0.0001693698790209479,
+      "loss": 2.7035,
+      "step": 5194
+    },
+    {
+      "epoch": 0.41925591154870473,
+      "grad_norm": 0.7323064208030701,
+      "learning_rate": 0.00016935850740315545,
+      "loss": 2.6713,
+      "step": 5195
+    },
+    {
+      "epoch": 0.4193366152852877,
+      "grad_norm": 0.8011505007743835,
+      "learning_rate": 0.00016934713405676764,
+      "loss": 2.6413,
+      "step": 5196
+    },
+    {
+      "epoch": 0.41941731902187074,
+      "grad_norm": 0.768851637840271,
+      "learning_rate": 0.00016933575898206804,
+      "loss": 2.6147,
+      "step": 5197
+    },
+    {
+      "epoch": 0.4194980227584537,
+      "grad_norm": 0.7255160808563232,
+      "learning_rate": 0.00016932438217934006,
+      "loss": 2.6093,
+      "step": 5198
+    },
+    {
+      "epoch": 0.41957872649503675,
+      "grad_norm": 0.7431769967079163,
+      "learning_rate": 0.00016931300364886722,
+      "loss": 2.6658,
+      "step": 5199
+    },
+    {
+      "epoch": 0.4196594302316197,
+      "grad_norm": 0.7532122731208801,
+      "learning_rate": 0.00016930162339093318,
+      "loss": 2.6371,
+      "step": 5200
+    },
+    {
+      "epoch": 0.41974013396820276,
+      "grad_norm": 0.7253943681716919,
+      "learning_rate": 0.00016929024140582152,
+      "loss": 2.6365,
+      "step": 5201
+    },
+    {
+      "epoch": 0.41982083770478573,
+      "grad_norm": 0.7323265075683594,
+      "learning_rate": 0.00016927885769381593,
+      "loss": 2.7096,
+      "step": 5202
+    },
+    {
+      "epoch": 0.4199015414413687,
+      "grad_norm": 0.7340009808540344,
+      "learning_rate": 0.00016926747225520008,
+      "loss": 2.6983,
+      "step": 5203
+    },
+    {
+      "epoch": 0.41998224517795174,
+      "grad_norm": 0.838706374168396,
+      "learning_rate": 0.00016925608509025776,
+      "loss": 2.7098,
+      "step": 5204
+    },
+    {
+      "epoch": 0.4200629489145347,
+      "grad_norm": 0.7320838570594788,
+      "learning_rate": 0.0001692446961992728,
+      "loss": 2.6767,
+      "step": 5205
+    },
+    {
+      "epoch": 0.42014365265111775,
+      "grad_norm": 0.7275335192680359,
+      "learning_rate": 0.00016923330558252898,
+      "loss": 2.6754,
+      "step": 5206
+    },
+    {
+      "epoch": 0.4202243563877007,
+      "grad_norm": 0.7572353482246399,
+      "learning_rate": 0.00016922191324031017,
+      "loss": 2.7076,
+      "step": 5207
+    },
+    {
+      "epoch": 0.42030506012428376,
+      "grad_norm": 0.7991098165512085,
+      "learning_rate": 0.0001692105191729004,
+      "loss": 2.7281,
+      "step": 5208
+    },
+    {
+      "epoch": 0.42038576386086673,
+      "grad_norm": 0.70769202709198,
+      "learning_rate": 0.00016919912338058356,
+      "loss": 2.684,
+      "step": 5209
+    },
+    {
+      "epoch": 0.42046646759744977,
+      "grad_norm": 0.6895349621772766,
+      "learning_rate": 0.0001691877258636436,
+      "loss": 2.6723,
+      "step": 5210
+    },
+    {
+      "epoch": 0.42054717133403274,
+      "grad_norm": 0.7368944883346558,
+      "learning_rate": 0.00016917632662236476,
+      "loss": 2.601,
+      "step": 5211
+    },
+    {
+      "epoch": 0.4206278750706158,
+      "grad_norm": 0.7122060060501099,
+      "learning_rate": 0.00016916492565703097,
+      "loss": 2.703,
+      "step": 5212
+    },
+    {
+      "epoch": 0.42070857880719875,
+      "grad_norm": 0.735251784324646,
+      "learning_rate": 0.00016915352296792646,
+      "loss": 2.7715,
+      "step": 5213
+    },
+    {
+      "epoch": 0.4207892825437818,
+      "grad_norm": 0.7686039805412292,
+      "learning_rate": 0.00016914211855533536,
+      "loss": 2.6935,
+      "step": 5214
+    },
+    {
+      "epoch": 0.42086998628036476,
+      "grad_norm": 0.8457472920417786,
+      "learning_rate": 0.00016913071241954195,
+      "loss": 2.6535,
+      "step": 5215
+    },
+    {
+      "epoch": 0.4209506900169478,
+      "grad_norm": 0.6913465261459351,
+      "learning_rate": 0.00016911930456083046,
+      "loss": 2.6453,
+      "step": 5216
+    },
+    {
+      "epoch": 0.42103139375353077,
+      "grad_norm": 0.6939878463745117,
+      "learning_rate": 0.00016910789497948524,
+      "loss": 2.6483,
+      "step": 5217
+    },
+    {
+      "epoch": 0.4211120974901138,
+      "grad_norm": 0.7240888476371765,
+      "learning_rate": 0.00016909648367579062,
+      "loss": 2.6649,
+      "step": 5218
+    },
+    {
+      "epoch": 0.4211928012266968,
+      "grad_norm": 0.7570972442626953,
+      "learning_rate": 0.00016908507065003102,
+      "loss": 2.6633,
+      "step": 5219
+    },
+    {
+      "epoch": 0.4212735049632798,
+      "grad_norm": 0.72161465883255,
+      "learning_rate": 0.00016907365590249082,
+      "loss": 2.6999,
+      "step": 5220
+    },
+    {
+      "epoch": 0.4213542086998628,
+      "grad_norm": 0.7818038463592529,
+      "learning_rate": 0.00016906223943345458,
+      "loss": 2.6478,
+      "step": 5221
+    },
+    {
+      "epoch": 0.4214349124364458,
+      "grad_norm": 0.7292464971542358,
+      "learning_rate": 0.00016905082124320684,
+      "loss": 2.6725,
+      "step": 5222
+    },
+    {
+      "epoch": 0.4215156161730288,
+      "grad_norm": 0.7612937092781067,
+      "learning_rate": 0.0001690394013320321,
+      "loss": 2.6474,
+      "step": 5223
+    },
+    {
+      "epoch": 0.4215963199096118,
+      "grad_norm": 0.7325131297111511,
+      "learning_rate": 0.000169027979700215,
+      "loss": 2.6525,
+      "step": 5224
+    },
+    {
+      "epoch": 0.4216770236461948,
+      "grad_norm": 0.7736644148826599,
+      "learning_rate": 0.00016901655634804022,
+      "loss": 2.662,
+      "step": 5225
+    },
+    {
+      "epoch": 0.42175772738277784,
+      "grad_norm": 0.758522629737854,
+      "learning_rate": 0.00016900513127579244,
+      "loss": 2.6558,
+      "step": 5226
+    },
+    {
+      "epoch": 0.4218384311193608,
+      "grad_norm": 0.7559491991996765,
+      "learning_rate": 0.00016899370448375642,
+      "loss": 2.7361,
+      "step": 5227
+    },
+    {
+      "epoch": 0.42191913485594384,
+      "grad_norm": 0.7791146039962769,
+      "learning_rate": 0.00016898227597221692,
+      "loss": 2.6739,
+      "step": 5228
+    },
+    {
+      "epoch": 0.4219998385925268,
+      "grad_norm": 0.7280717492103577,
+      "learning_rate": 0.00016897084574145878,
+      "loss": 2.6316,
+      "step": 5229
+    },
+    {
+      "epoch": 0.42208054232910985,
+      "grad_norm": 0.7455596327781677,
+      "learning_rate": 0.0001689594137917669,
+      "loss": 2.7244,
+      "step": 5230
+    },
+    {
+      "epoch": 0.42216124606569283,
+      "grad_norm": 0.7965813875198364,
+      "learning_rate": 0.00016894798012342613,
+      "loss": 2.6757,
+      "step": 5231
+    },
+    {
+      "epoch": 0.42224194980227586,
+      "grad_norm": 0.6740596294403076,
+      "learning_rate": 0.00016893654473672148,
+      "loss": 2.631,
+      "step": 5232
+    },
+    {
+      "epoch": 0.42232265353885884,
+      "grad_norm": 0.695105254650116,
+      "learning_rate": 0.00016892510763193795,
+      "loss": 2.6563,
+      "step": 5233
+    },
+    {
+      "epoch": 0.42240335727544187,
+      "grad_norm": 0.7623865008354187,
+      "learning_rate": 0.00016891366880936051,
+      "loss": 2.6738,
+      "step": 5234
+    },
+    {
+      "epoch": 0.42248406101202485,
+      "grad_norm": 0.7545912265777588,
+      "learning_rate": 0.00016890222826927435,
+      "loss": 2.6949,
+      "step": 5235
+    },
+    {
+      "epoch": 0.4225647647486079,
+      "grad_norm": 0.7280749678611755,
+      "learning_rate": 0.00016889078601196452,
+      "loss": 2.6571,
+      "step": 5236
+    },
+    {
+      "epoch": 0.42264546848519086,
+      "grad_norm": 0.6624523401260376,
+      "learning_rate": 0.00016887934203771625,
+      "loss": 2.6854,
+      "step": 5237
+    },
+    {
+      "epoch": 0.4227261722217739,
+      "grad_norm": 0.7835487127304077,
+      "learning_rate": 0.0001688678963468147,
+      "loss": 2.6437,
+      "step": 5238
+    },
+    {
+      "epoch": 0.42280687595835686,
+      "grad_norm": 0.7384940981864929,
+      "learning_rate": 0.00016885644893954518,
+      "loss": 2.6584,
+      "step": 5239
+    },
+    {
+      "epoch": 0.4228875796949399,
+      "grad_norm": 0.8227531313896179,
+      "learning_rate": 0.00016884499981619292,
+      "loss": 2.673,
+      "step": 5240
+    },
+    {
+      "epoch": 0.4229682834315229,
+      "grad_norm": 0.7442220449447632,
+      "learning_rate": 0.00016883354897704334,
+      "loss": 2.6729,
+      "step": 5241
+    },
+    {
+      "epoch": 0.4230489871681059,
+      "grad_norm": 0.7182636857032776,
+      "learning_rate": 0.00016882209642238175,
+      "loss": 2.6833,
+      "step": 5242
+    },
+    {
+      "epoch": 0.4231296909046889,
+      "grad_norm": 0.7061870098114014,
+      "learning_rate": 0.00016881064215249362,
+      "loss": 2.6696,
+      "step": 5243
+    },
+    {
+      "epoch": 0.4232103946412719,
+      "grad_norm": 0.6792885065078735,
+      "learning_rate": 0.00016879918616766445,
+      "loss": 2.6805,
+      "step": 5244
+    },
+    {
+      "epoch": 0.4232910983778549,
+      "grad_norm": 0.7439807057380676,
+      "learning_rate": 0.00016878772846817968,
+      "loss": 2.6522,
+      "step": 5245
+    },
+    {
+      "epoch": 0.4233718021144379,
+      "grad_norm": 0.7078969478607178,
+      "learning_rate": 0.00016877626905432492,
+      "loss": 2.6549,
+      "step": 5246
+    },
+    {
+      "epoch": 0.4234525058510209,
+      "grad_norm": 0.7103868126869202,
+      "learning_rate": 0.00016876480792638577,
+      "loss": 2.6812,
+      "step": 5247
+    },
+    {
+      "epoch": 0.42353320958760393,
+      "grad_norm": 0.7224452495574951,
+      "learning_rate": 0.00016875334508464782,
+      "loss": 2.6657,
+      "step": 5248
+    },
+    {
+      "epoch": 0.4236139133241869,
+      "grad_norm": 0.6885106563568115,
+      "learning_rate": 0.00016874188052939682,
+      "loss": 2.6421,
+      "step": 5249
+    },
+    {
+      "epoch": 0.42369461706076994,
+      "grad_norm": 0.6736720204353333,
+      "learning_rate": 0.00016873041426091845,
+      "loss": 2.6717,
+      "step": 5250
+    },
+    {
+      "epoch": 0.4237753207973529,
+      "grad_norm": 0.7597963809967041,
+      "learning_rate": 0.00016871894627949846,
+      "loss": 2.6231,
+      "step": 5251
+    },
+    {
+      "epoch": 0.42385602453393595,
+      "grad_norm": 0.8295687437057495,
+      "learning_rate": 0.00016870747658542275,
+      "loss": 2.6631,
+      "step": 5252
+    },
+    {
+      "epoch": 0.4239367282705189,
+      "grad_norm": 0.6750548481941223,
+      "learning_rate": 0.0001686960051789771,
+      "loss": 2.6997,
+      "step": 5253
+    },
+    {
+      "epoch": 0.4240174320071019,
+      "grad_norm": 0.7229160666465759,
+      "learning_rate": 0.0001686845320604474,
+      "loss": 2.6525,
+      "step": 5254
+    },
+    {
+      "epoch": 0.42409813574368493,
+      "grad_norm": 0.8318623900413513,
+      "learning_rate": 0.00016867305723011967,
+      "loss": 2.7774,
+      "step": 5255
+    },
+    {
+      "epoch": 0.4241788394802679,
+      "grad_norm": 0.8391026854515076,
+      "learning_rate": 0.00016866158068827979,
+      "loss": 2.6712,
+      "step": 5256
+    },
+    {
+      "epoch": 0.42425954321685094,
+      "grad_norm": 0.691146969795227,
+      "learning_rate": 0.00016865010243521388,
+      "loss": 2.6459,
+      "step": 5257
+    },
+    {
+      "epoch": 0.4243402469534339,
+      "grad_norm": 0.7223602533340454,
+      "learning_rate": 0.00016863862247120794,
+      "loss": 2.6675,
+      "step": 5258
+    },
+    {
+      "epoch": 0.42442095069001695,
+      "grad_norm": 0.8400631546974182,
+      "learning_rate": 0.0001686271407965481,
+      "loss": 2.6978,
+      "step": 5259
+    },
+    {
+      "epoch": 0.42450165442659993,
+      "grad_norm": 0.737684965133667,
+      "learning_rate": 0.0001686156574115205,
+      "loss": 2.6992,
+      "step": 5260
+    },
+    {
+      "epoch": 0.42458235816318296,
+      "grad_norm": 0.7511717677116394,
+      "learning_rate": 0.0001686041723164114,
+      "loss": 2.6947,
+      "step": 5261
+    },
+    {
+      "epoch": 0.42466306189976594,
+      "grad_norm": 0.7434492707252502,
+      "learning_rate": 0.00016859268551150698,
+      "loss": 2.7353,
+      "step": 5262
+    },
+    {
+      "epoch": 0.42474376563634897,
+      "grad_norm": 0.746609628200531,
+      "learning_rate": 0.00016858119699709353,
+      "loss": 2.7519,
+      "step": 5263
+    },
+    {
+      "epoch": 0.42482446937293195,
+      "grad_norm": 0.7709949612617493,
+      "learning_rate": 0.0001685697067734574,
+      "loss": 2.7018,
+      "step": 5264
+    },
+    {
+      "epoch": 0.424905173109515,
+      "grad_norm": 0.7496309876441956,
+      "learning_rate": 0.00016855821484088488,
+      "loss": 2.6761,
+      "step": 5265
+    },
+    {
+      "epoch": 0.42498587684609795,
+      "grad_norm": 0.7071252465248108,
+      "learning_rate": 0.00016854672119966243,
+      "loss": 2.6762,
+      "step": 5266
+    },
+    {
+      "epoch": 0.425066580582681,
+      "grad_norm": 0.7991356253623962,
+      "learning_rate": 0.00016853522585007658,
+      "loss": 2.6134,
+      "step": 5267
+    },
+    {
+      "epoch": 0.42514728431926396,
+      "grad_norm": 0.8194605708122253,
+      "learning_rate": 0.0001685237287924137,
+      "loss": 2.6601,
+      "step": 5268
+    },
+    {
+      "epoch": 0.425227988055847,
+      "grad_norm": 0.7451688051223755,
+      "learning_rate": 0.00016851223002696037,
+      "loss": 2.6631,
+      "step": 5269
+    },
+    {
+      "epoch": 0.42530869179242997,
+      "grad_norm": 0.7220263481140137,
+      "learning_rate": 0.0001685007295540032,
+      "loss": 2.6631,
+      "step": 5270
+    },
+    {
+      "epoch": 0.425389395529013,
+      "grad_norm": 0.7268854975700378,
+      "learning_rate": 0.00016848922737382874,
+      "loss": 2.6752,
+      "step": 5271
+    },
+    {
+      "epoch": 0.425470099265596,
+      "grad_norm": 0.8841642141342163,
+      "learning_rate": 0.00016847772348672378,
+      "loss": 2.7153,
+      "step": 5272
+    },
+    {
+      "epoch": 0.425550803002179,
+      "grad_norm": 0.7725942134857178,
+      "learning_rate": 0.00016846621789297489,
+      "loss": 2.6726,
+      "step": 5273
+    },
+    {
+      "epoch": 0.425631506738762,
+      "grad_norm": 0.7179448008537292,
+      "learning_rate": 0.00016845471059286887,
+      "loss": 2.6659,
+      "step": 5274
+    },
+    {
+      "epoch": 0.425712210475345,
+      "grad_norm": 0.7630325555801392,
+      "learning_rate": 0.00016844320158669257,
+      "loss": 2.7133,
+      "step": 5275
+    },
+    {
+      "epoch": 0.425792914211928,
+      "grad_norm": 0.7349739670753479,
+      "learning_rate": 0.00016843169087473272,
+      "loss": 2.6397,
+      "step": 5276
+    },
+    {
+      "epoch": 0.42587361794851103,
+      "grad_norm": 0.7670298218727112,
+      "learning_rate": 0.00016842017845727626,
+      "loss": 2.6485,
+      "step": 5277
+    },
+    {
+      "epoch": 0.425954321685094,
+      "grad_norm": 0.692095160484314,
+      "learning_rate": 0.00016840866433461013,
+      "loss": 2.6058,
+      "step": 5278
+    },
+    {
+      "epoch": 0.42603502542167704,
+      "grad_norm": 0.6888624429702759,
+      "learning_rate": 0.00016839714850702125,
+      "loss": 2.5757,
+      "step": 5279
+    },
+    {
+      "epoch": 0.42611572915826,
+      "grad_norm": 0.6816484332084656,
+      "learning_rate": 0.00016838563097479664,
+      "loss": 2.6656,
+      "step": 5280
+    },
+    {
+      "epoch": 0.42619643289484305,
+      "grad_norm": 0.7778486609458923,
+      "learning_rate": 0.00016837411173822333,
+      "loss": 2.6738,
+      "step": 5281
+    },
+    {
+      "epoch": 0.426277136631426,
+      "grad_norm": 0.73436439037323,
+      "learning_rate": 0.00016836259079758845,
+      "loss": 2.6346,
+      "step": 5282
+    },
+    {
+      "epoch": 0.42635784036800906,
+      "grad_norm": 0.673528254032135,
+      "learning_rate": 0.00016835106815317908,
+      "loss": 2.6636,
+      "step": 5283
+    },
+    {
+      "epoch": 0.42643854410459203,
+      "grad_norm": 0.6892737150192261,
+      "learning_rate": 0.00016833954380528242,
+      "loss": 2.6723,
+      "step": 5284
+    },
+    {
+      "epoch": 0.42651924784117506,
+      "grad_norm": 0.7404607534408569,
+      "learning_rate": 0.00016832801775418571,
+      "loss": 2.6751,
+      "step": 5285
+    },
+    {
+      "epoch": 0.42659995157775804,
+      "grad_norm": 0.7040587663650513,
+      "learning_rate": 0.00016831649000017618,
+      "loss": 2.6079,
+      "step": 5286
+    },
+    {
+      "epoch": 0.4266806553143411,
+      "grad_norm": 0.7295164465904236,
+      "learning_rate": 0.00016830496054354112,
+      "loss": 2.5928,
+      "step": 5287
+    },
+    {
+      "epoch": 0.42676135905092405,
+      "grad_norm": 0.7269962430000305,
+      "learning_rate": 0.00016829342938456788,
+      "loss": 2.6648,
+      "step": 5288
+    },
+    {
+      "epoch": 0.4268420627875071,
+      "grad_norm": 0.7296550273895264,
+      "learning_rate": 0.0001682818965235439,
+      "loss": 2.6814,
+      "step": 5289
+    },
+    {
+      "epoch": 0.42692276652409006,
+      "grad_norm": 0.8376085758209229,
+      "learning_rate": 0.00016827036196075655,
+      "loss": 2.702,
+      "step": 5290
+    },
+    {
+      "epoch": 0.4270034702606731,
+      "grad_norm": 0.7461032271385193,
+      "learning_rate": 0.00016825882569649332,
+      "loss": 2.6959,
+      "step": 5291
+    },
+    {
+      "epoch": 0.42708417399725607,
+      "grad_norm": 0.7218661308288574,
+      "learning_rate": 0.00016824728773104171,
+      "loss": 2.7182,
+      "step": 5292
+    },
+    {
+      "epoch": 0.4271648777338391,
+      "grad_norm": 0.7012860774993896,
+      "learning_rate": 0.00016823574806468933,
+      "loss": 2.6989,
+      "step": 5293
+    },
+    {
+      "epoch": 0.4272455814704221,
+      "grad_norm": 0.7039482593536377,
+      "learning_rate": 0.0001682242066977237,
+      "loss": 2.6153,
+      "step": 5294
+    },
+    {
+      "epoch": 0.4273262852070051,
+      "grad_norm": 0.8783851861953735,
+      "learning_rate": 0.0001682126636304325,
+      "loss": 2.7174,
+      "step": 5295
+    },
+    {
+      "epoch": 0.4274069889435881,
+      "grad_norm": 0.7266566157341003,
+      "learning_rate": 0.00016820111886310343,
+      "loss": 2.6571,
+      "step": 5296
+    },
+    {
+      "epoch": 0.4274876926801711,
+      "grad_norm": 0.7512212991714478,
+      "learning_rate": 0.0001681895723960242,
+      "loss": 2.6802,
+      "step": 5297
+    },
+    {
+      "epoch": 0.4275683964167541,
+      "grad_norm": 0.7786974310874939,
+      "learning_rate": 0.00016817802422948254,
+      "loss": 2.6514,
+      "step": 5298
+    },
+    {
+      "epoch": 0.4276491001533371,
+      "grad_norm": 0.7454531788825989,
+      "learning_rate": 0.00016816647436376634,
+      "loss": 2.6508,
+      "step": 5299
+    },
+    {
+      "epoch": 0.4277298038899201,
+      "grad_norm": 0.7542992830276489,
+      "learning_rate": 0.0001681549227991634,
+      "loss": 2.6455,
+      "step": 5300
+    },
+    {
+      "epoch": 0.42781050762650313,
+      "grad_norm": 0.7405722141265869,
+      "learning_rate": 0.0001681433695359616,
+      "loss": 2.6505,
+      "step": 5301
+    },
+    {
+      "epoch": 0.4278912113630861,
+      "grad_norm": 0.7120002508163452,
+      "learning_rate": 0.00016813181457444896,
+      "loss": 2.6652,
+      "step": 5302
+    },
+    {
+      "epoch": 0.42797191509966914,
+      "grad_norm": 0.7645997405052185,
+      "learning_rate": 0.00016812025791491334,
+      "loss": 2.6456,
+      "step": 5303
+    },
+    {
+      "epoch": 0.4280526188362521,
+      "grad_norm": 0.7214465141296387,
+      "learning_rate": 0.00016810869955764286,
+      "loss": 2.6261,
+      "step": 5304
+    },
+    {
+      "epoch": 0.4281333225728351,
+      "grad_norm": 0.7653367519378662,
+      "learning_rate": 0.00016809713950292551,
+      "loss": 2.7295,
+      "step": 5305
+    },
+    {
+      "epoch": 0.4282140263094181,
+      "grad_norm": 0.6798970103263855,
+      "learning_rate": 0.0001680855777510495,
+      "loss": 2.6549,
+      "step": 5306
+    },
+    {
+      "epoch": 0.4282947300460011,
+      "grad_norm": 0.7693684101104736,
+      "learning_rate": 0.00016807401430230288,
+      "loss": 2.7001,
+      "step": 5307
+    },
+    {
+      "epoch": 0.42837543378258414,
+      "grad_norm": 0.6962063312530518,
+      "learning_rate": 0.00016806244915697384,
+      "loss": 2.6582,
+      "step": 5308
+    },
+    {
+      "epoch": 0.4284561375191671,
+      "grad_norm": 0.7526959776878357,
+      "learning_rate": 0.00016805088231535068,
+      "loss": 2.7204,
+      "step": 5309
+    },
+    {
+      "epoch": 0.42853684125575014,
+      "grad_norm": 0.7403820753097534,
+      "learning_rate": 0.0001680393137777217,
+      "loss": 2.6505,
+      "step": 5310
+    },
+    {
+      "epoch": 0.4286175449923331,
+      "grad_norm": 0.7056909799575806,
+      "learning_rate": 0.00016802774354437506,
+      "loss": 2.5981,
+      "step": 5311
+    },
+    {
+      "epoch": 0.42869824872891615,
+      "grad_norm": 0.6756439805030823,
+      "learning_rate": 0.0001680161716155993,
+      "loss": 2.6845,
+      "step": 5312
+    },
+    {
+      "epoch": 0.42877895246549913,
+      "grad_norm": 0.7634297013282776,
+      "learning_rate": 0.0001680045979916827,
+      "loss": 2.6399,
+      "step": 5313
+    },
+    {
+      "epoch": 0.42885965620208216,
+      "grad_norm": 0.6793022751808167,
+      "learning_rate": 0.0001679930226729138,
+      "loss": 2.6808,
+      "step": 5314
+    },
+    {
+      "epoch": 0.42894035993866514,
+      "grad_norm": 0.7692369222640991,
+      "learning_rate": 0.00016798144565958103,
+      "loss": 2.673,
+      "step": 5315
+    },
+    {
+      "epoch": 0.42902106367524817,
+      "grad_norm": 0.668798565864563,
+      "learning_rate": 0.00016796986695197293,
+      "loss": 2.6465,
+      "step": 5316
+    },
+    {
+      "epoch": 0.42910176741183115,
+      "grad_norm": 0.719160795211792,
+      "learning_rate": 0.00016795828655037805,
+      "loss": 2.5876,
+      "step": 5317
+    },
+    {
+      "epoch": 0.4291824711484142,
+      "grad_norm": 0.7352864742279053,
+      "learning_rate": 0.000167946704455085,
+      "loss": 2.625,
+      "step": 5318
+    },
+    {
+      "epoch": 0.42926317488499716,
+      "grad_norm": 0.7103392481803894,
+      "learning_rate": 0.00016793512066638254,
+      "loss": 2.602,
+      "step": 5319
+    },
+    {
+      "epoch": 0.4293438786215802,
+      "grad_norm": 0.7005727291107178,
+      "learning_rate": 0.0001679235351845592,
+      "loss": 2.6723,
+      "step": 5320
+    },
+    {
+      "epoch": 0.42942458235816316,
+      "grad_norm": 0.7686243653297424,
+      "learning_rate": 0.00016791194800990387,
+      "loss": 2.693,
+      "step": 5321
+    },
+    {
+      "epoch": 0.4295052860947462,
+      "grad_norm": 0.7026933431625366,
+      "learning_rate": 0.00016790035914270526,
+      "loss": 2.6334,
+      "step": 5322
+    },
+    {
+      "epoch": 0.4295859898313292,
+      "grad_norm": 0.748938262462616,
+      "learning_rate": 0.0001678887685832522,
+      "loss": 2.6757,
+      "step": 5323
+    },
+    {
+      "epoch": 0.4296666935679122,
+      "grad_norm": 0.7753568887710571,
+      "learning_rate": 0.00016787717633183355,
+      "loss": 2.6782,
+      "step": 5324
+    },
+    {
+      "epoch": 0.4297473973044952,
+      "grad_norm": 0.7605767846107483,
+      "learning_rate": 0.00016786558238873823,
+      "loss": 2.6822,
+      "step": 5325
+    },
+    {
+      "epoch": 0.4298281010410782,
+      "grad_norm": 0.7516531348228455,
+      "learning_rate": 0.00016785398675425524,
+      "loss": 2.6802,
+      "step": 5326
+    },
+    {
+      "epoch": 0.4299088047776612,
+      "grad_norm": 0.7551677227020264,
+      "learning_rate": 0.0001678423894286735,
+      "loss": 2.6509,
+      "step": 5327
+    },
+    {
+      "epoch": 0.4299895085142442,
+      "grad_norm": 0.765364944934845,
+      "learning_rate": 0.00016783079041228206,
+      "loss": 2.6552,
+      "step": 5328
+    },
+    {
+      "epoch": 0.4300702122508272,
+      "grad_norm": 0.7016649842262268,
+      "learning_rate": 0.00016781918970537002,
+      "loss": 2.6861,
+      "step": 5329
+    },
+    {
+      "epoch": 0.43015091598741023,
+      "grad_norm": 0.7266311645507812,
+      "learning_rate": 0.0001678075873082265,
+      "loss": 2.7064,
+      "step": 5330
+    },
+    {
+      "epoch": 0.4302316197239932,
+      "grad_norm": 0.7414532899856567,
+      "learning_rate": 0.00016779598322114064,
+      "loss": 2.6273,
+      "step": 5331
+    },
+    {
+      "epoch": 0.43031232346057624,
+      "grad_norm": 0.7032443881034851,
+      "learning_rate": 0.00016778437744440167,
+      "loss": 2.6577,
+      "step": 5332
+    },
+    {
+      "epoch": 0.4303930271971592,
+      "grad_norm": 0.7150338888168335,
+      "learning_rate": 0.00016777276997829882,
+      "loss": 2.6586,
+      "step": 5333
+    },
+    {
+      "epoch": 0.43047373093374225,
+      "grad_norm": 0.6893971562385559,
+      "learning_rate": 0.0001677611608231214,
+      "loss": 2.6713,
+      "step": 5334
+    },
+    {
+      "epoch": 0.4305544346703252,
+      "grad_norm": 0.861935555934906,
+      "learning_rate": 0.00016774954997915867,
+      "loss": 2.7037,
+      "step": 5335
+    },
+    {
+      "epoch": 0.43063513840690826,
+      "grad_norm": 0.7140138745307922,
+      "learning_rate": 0.00016773793744670012,
+      "loss": 2.6684,
+      "step": 5336
+    },
+    {
+      "epoch": 0.43071584214349123,
+      "grad_norm": 0.7245929837226868,
+      "learning_rate": 0.00016772632322603506,
+      "loss": 2.6349,
+      "step": 5337
+    },
+    {
+      "epoch": 0.43079654588007427,
+      "grad_norm": 0.7216203808784485,
+      "learning_rate": 0.000167714707317453,
+      "loss": 2.6338,
+      "step": 5338
+    },
+    {
+      "epoch": 0.43087724961665724,
+      "grad_norm": 0.7076452374458313,
+      "learning_rate": 0.00016770308972124343,
+      "loss": 2.6614,
+      "step": 5339
+    },
+    {
+      "epoch": 0.4309579533532403,
+      "grad_norm": 0.7392035722732544,
+      "learning_rate": 0.00016769147043769586,
+      "loss": 2.6697,
+      "step": 5340
+    },
+    {
+      "epoch": 0.43103865708982325,
+      "grad_norm": 0.7235357761383057,
+      "learning_rate": 0.00016767984946709994,
+      "loss": 2.6664,
+      "step": 5341
+    },
+    {
+      "epoch": 0.4311193608264063,
+      "grad_norm": 0.6985526084899902,
+      "learning_rate": 0.00016766822680974524,
+      "loss": 2.6157,
+      "step": 5342
+    },
+    {
+      "epoch": 0.43120006456298926,
+      "grad_norm": 0.769963264465332,
+      "learning_rate": 0.0001676566024659214,
+      "loss": 2.6096,
+      "step": 5343
+    },
+    {
+      "epoch": 0.4312807682995723,
+      "grad_norm": 0.7504093050956726,
+      "learning_rate": 0.00016764497643591823,
+      "loss": 2.5795,
+      "step": 5344
+    },
+    {
+      "epoch": 0.43136147203615527,
+      "grad_norm": 0.7193379402160645,
+      "learning_rate": 0.0001676333487200254,
+      "loss": 2.6158,
+      "step": 5345
+    },
+    {
+      "epoch": 0.4314421757727383,
+      "grad_norm": 0.777357280254364,
+      "learning_rate": 0.00016762171931853273,
+      "loss": 2.6388,
+      "step": 5346
+    },
+    {
+      "epoch": 0.4315228795093213,
+      "grad_norm": 0.8590179085731506,
+      "learning_rate": 0.00016761008823173003,
+      "loss": 2.6597,
+      "step": 5347
+    },
+    {
+      "epoch": 0.4316035832459043,
+      "grad_norm": 0.7040170431137085,
+      "learning_rate": 0.0001675984554599072,
+      "loss": 2.6447,
+      "step": 5348
+    },
+    {
+      "epoch": 0.4316842869824873,
+      "grad_norm": 0.7682301998138428,
+      "learning_rate": 0.00016758682100335417,
+      "loss": 2.6738,
+      "step": 5349
+    },
+    {
+      "epoch": 0.4317649907190703,
+      "grad_norm": 0.8342414498329163,
+      "learning_rate": 0.00016757518486236087,
+      "loss": 2.7058,
+      "step": 5350
+    },
+    {
+      "epoch": 0.4318456944556533,
+      "grad_norm": 0.7410600781440735,
+      "learning_rate": 0.00016756354703721736,
+      "loss": 2.6597,
+      "step": 5351
+    },
+    {
+      "epoch": 0.4319263981922363,
+      "grad_norm": 0.7633174061775208,
+      "learning_rate": 0.00016755190752821363,
+      "loss": 2.6461,
+      "step": 5352
+    },
+    {
+      "epoch": 0.4320071019288193,
+      "grad_norm": 0.7855150103569031,
+      "learning_rate": 0.00016754026633563973,
+      "loss": 2.6556,
+      "step": 5353
+    },
+    {
+      "epoch": 0.43208780566540234,
+      "grad_norm": 0.7197602391242981,
+      "learning_rate": 0.00016752862345978587,
+      "loss": 2.6511,
+      "step": 5354
+    },
+    {
+      "epoch": 0.4321685094019853,
+      "grad_norm": 0.7748876810073853,
+      "learning_rate": 0.00016751697890094223,
+      "loss": 2.7,
+      "step": 5355
+    },
+    {
+      "epoch": 0.4322492131385683,
+      "grad_norm": 0.7457308173179626,
+      "learning_rate": 0.00016750533265939895,
+      "loss": 2.6934,
+      "step": 5356
+    },
+    {
+      "epoch": 0.4323299168751513,
+      "grad_norm": 0.8003394603729248,
+      "learning_rate": 0.00016749368473544633,
+      "loss": 2.6273,
+      "step": 5357
+    },
+    {
+      "epoch": 0.4324106206117343,
+      "grad_norm": 0.7163615822792053,
+      "learning_rate": 0.00016748203512937464,
+      "loss": 2.6605,
+      "step": 5358
+    },
+    {
+      "epoch": 0.43249132434831733,
+      "grad_norm": 0.6859120726585388,
+      "learning_rate": 0.00016747038384147422,
+      "loss": 2.6748,
+      "step": 5359
+    },
+    {
+      "epoch": 0.4325720280849003,
+      "grad_norm": 0.7169440984725952,
+      "learning_rate": 0.0001674587308720355,
+      "loss": 2.6674,
+      "step": 5360
+    },
+    {
+      "epoch": 0.43265273182148334,
+      "grad_norm": 0.7762351036071777,
+      "learning_rate": 0.00016744707622134888,
+      "loss": 2.6673,
+      "step": 5361
+    },
+    {
+      "epoch": 0.4327334355580663,
+      "grad_norm": 0.7169542908668518,
+      "learning_rate": 0.0001674354198897048,
+      "loss": 2.7341,
+      "step": 5362
+    },
+    {
+      "epoch": 0.43281413929464935,
+      "grad_norm": 0.7903403043746948,
+      "learning_rate": 0.00016742376187739376,
+      "loss": 2.6019,
+      "step": 5363
+    },
+    {
+      "epoch": 0.4328948430312323,
+      "grad_norm": 0.8395403027534485,
+      "learning_rate": 0.00016741210218470634,
+      "loss": 2.6519,
+      "step": 5364
+    },
+    {
+      "epoch": 0.43297554676781536,
+      "grad_norm": 0.7521546483039856,
+      "learning_rate": 0.0001674004408119331,
+      "loss": 2.6067,
+      "step": 5365
+    },
+    {
+      "epoch": 0.43305625050439833,
+      "grad_norm": 0.7186779975891113,
+      "learning_rate": 0.0001673887777593647,
+      "loss": 2.6435,
+      "step": 5366
+    },
+    {
+      "epoch": 0.43313695424098136,
+      "grad_norm": 0.7362968921661377,
+      "learning_rate": 0.0001673771130272918,
+      "loss": 2.6031,
+      "step": 5367
+    },
+    {
+      "epoch": 0.43321765797756434,
+      "grad_norm": 0.8033537864685059,
+      "learning_rate": 0.0001673654466160051,
+      "loss": 2.7234,
+      "step": 5368
+    },
+    {
+      "epoch": 0.4332983617141474,
+      "grad_norm": 0.7109711766242981,
+      "learning_rate": 0.0001673537785257954,
+      "loss": 2.6621,
+      "step": 5369
+    },
+    {
+      "epoch": 0.43337906545073035,
+      "grad_norm": 0.7499226927757263,
+      "learning_rate": 0.0001673421087569535,
+      "loss": 2.706,
+      "step": 5370
+    },
+    {
+      "epoch": 0.4334597691873134,
+      "grad_norm": 0.7192875146865845,
+      "learning_rate": 0.00016733043730977017,
+      "loss": 2.6053,
+      "step": 5371
+    },
+    {
+      "epoch": 0.43354047292389636,
+      "grad_norm": 0.6939374208450317,
+      "learning_rate": 0.00016731876418453636,
+      "loss": 2.6621,
+      "step": 5372
+    },
+    {
+      "epoch": 0.4336211766604794,
+      "grad_norm": 0.720741331577301,
+      "learning_rate": 0.00016730708938154297,
+      "loss": 2.6358,
+      "step": 5373
+    },
+    {
+      "epoch": 0.43370188039706237,
+      "grad_norm": 0.6979780793190002,
+      "learning_rate": 0.00016729541290108095,
+      "loss": 2.6162,
+      "step": 5374
+    },
+    {
+      "epoch": 0.4337825841336454,
+      "grad_norm": 0.8014200925827026,
+      "learning_rate": 0.00016728373474344136,
+      "loss": 2.6255,
+      "step": 5375
+    },
+    {
+      "epoch": 0.4338632878702284,
+      "grad_norm": 0.7780057787895203,
+      "learning_rate": 0.0001672720549089152,
+      "loss": 2.6257,
+      "step": 5376
+    },
+    {
+      "epoch": 0.4339439916068114,
+      "grad_norm": 0.7111102938652039,
+      "learning_rate": 0.00016726037339779358,
+      "loss": 2.6384,
+      "step": 5377
+    },
+    {
+      "epoch": 0.4340246953433944,
+      "grad_norm": 0.7077106833457947,
+      "learning_rate": 0.00016724869021036764,
+      "loss": 2.6293,
+      "step": 5378
+    },
+    {
+      "epoch": 0.4341053990799774,
+      "grad_norm": 0.8328250646591187,
+      "learning_rate": 0.00016723700534692853,
+      "loss": 2.6186,
+      "step": 5379
+    },
+    {
+      "epoch": 0.4341861028165604,
+      "grad_norm": 0.6942149996757507,
+      "learning_rate": 0.00016722531880776752,
+      "loss": 2.6032,
+      "step": 5380
+    },
+    {
+      "epoch": 0.4342668065531434,
+      "grad_norm": 0.7180305123329163,
+      "learning_rate": 0.00016721363059317583,
+      "loss": 2.6166,
+      "step": 5381
+    },
+    {
+      "epoch": 0.4343475102897264,
+      "grad_norm": 0.8093443512916565,
+      "learning_rate": 0.00016720194070344476,
+      "loss": 2.6596,
+      "step": 5382
+    },
+    {
+      "epoch": 0.43442821402630943,
+      "grad_norm": 0.7337743043899536,
+      "learning_rate": 0.00016719024913886568,
+      "loss": 2.6137,
+      "step": 5383
+    },
+    {
+      "epoch": 0.4345089177628924,
+      "grad_norm": 0.7590384483337402,
+      "learning_rate": 0.00016717855589972993,
+      "loss": 2.6541,
+      "step": 5384
+    },
+    {
+      "epoch": 0.43458962149947544,
+      "grad_norm": 0.6945257186889648,
+      "learning_rate": 0.00016716686098632898,
+      "loss": 2.686,
+      "step": 5385
+    },
+    {
+      "epoch": 0.4346703252360584,
+      "grad_norm": 0.7175764441490173,
+      "learning_rate": 0.00016715516439895424,
+      "loss": 2.6081,
+      "step": 5386
+    },
+    {
+      "epoch": 0.43475102897264145,
+      "grad_norm": 0.7287259697914124,
+      "learning_rate": 0.00016714346613789732,
+      "loss": 2.6462,
+      "step": 5387
+    },
+    {
+      "epoch": 0.43483173270922443,
+      "grad_norm": 0.6864096522331238,
+      "learning_rate": 0.00016713176620344964,
+      "loss": 2.7104,
+      "step": 5388
+    },
+    {
+      "epoch": 0.43491243644580746,
+      "grad_norm": 0.6554383039474487,
+      "learning_rate": 0.00016712006459590289,
+      "loss": 2.6153,
+      "step": 5389
+    },
+    {
+      "epoch": 0.43499314018239044,
+      "grad_norm": 0.6415165662765503,
+      "learning_rate": 0.00016710836131554867,
+      "loss": 2.6198,
+      "step": 5390
+    },
+    {
+      "epoch": 0.43507384391897347,
+      "grad_norm": 0.6998475193977356,
+      "learning_rate": 0.00016709665636267869,
+      "loss": 2.6774,
+      "step": 5391
+    },
+    {
+      "epoch": 0.43515454765555645,
+      "grad_norm": 0.7437679171562195,
+      "learning_rate": 0.00016708494973758465,
+      "loss": 2.6176,
+      "step": 5392
+    },
+    {
+      "epoch": 0.4352352513921395,
+      "grad_norm": 0.6898311376571655,
+      "learning_rate": 0.00016707324144055825,
+      "loss": 2.6194,
+      "step": 5393
+    },
+    {
+      "epoch": 0.43531595512872245,
+      "grad_norm": 0.7536425590515137,
+      "learning_rate": 0.00016706153147189138,
+      "loss": 2.672,
+      "step": 5394
+    },
+    {
+      "epoch": 0.4353966588653055,
+      "grad_norm": 0.7576118111610413,
+      "learning_rate": 0.00016704981983187581,
+      "loss": 2.6473,
+      "step": 5395
+    },
+    {
+      "epoch": 0.43547736260188846,
+      "grad_norm": 0.7452495098114014,
+      "learning_rate": 0.00016703810652080349,
+      "loss": 2.6487,
+      "step": 5396
+    },
+    {
+      "epoch": 0.4355580663384715,
+      "grad_norm": 0.7817744612693787,
+      "learning_rate": 0.0001670263915389663,
+      "loss": 2.61,
+      "step": 5397
+    },
+    {
+      "epoch": 0.43563877007505447,
+      "grad_norm": 0.7195492386817932,
+      "learning_rate": 0.00016701467488665624,
+      "loss": 2.6745,
+      "step": 5398
+    },
+    {
+      "epoch": 0.4357194738116375,
+      "grad_norm": 0.7703930735588074,
+      "learning_rate": 0.0001670029565641653,
+      "loss": 2.7196,
+      "step": 5399
+    },
+    {
+      "epoch": 0.4358001775482205,
+      "grad_norm": 0.6859520673751831,
+      "learning_rate": 0.00016699123657178553,
+      "loss": 2.6317,
+      "step": 5400
+    },
+    {
+      "epoch": 0.4358808812848035,
+      "grad_norm": 0.7380268573760986,
+      "learning_rate": 0.00016697951490980903,
+      "loss": 2.6008,
+      "step": 5401
+    },
+    {
+      "epoch": 0.4359615850213865,
+      "grad_norm": 0.7903439402580261,
+      "learning_rate": 0.00016696779157852792,
+      "loss": 2.6411,
+      "step": 5402
+    },
+    {
+      "epoch": 0.4360422887579695,
+      "grad_norm": 0.7022606134414673,
+      "learning_rate": 0.0001669560665782344,
+      "loss": 2.6153,
+      "step": 5403
+    },
+    {
+      "epoch": 0.4361229924945525,
+      "grad_norm": 0.8196203112602234,
+      "learning_rate": 0.00016694433990922068,
+      "loss": 2.6128,
+      "step": 5404
+    },
+    {
+      "epoch": 0.43620369623113553,
+      "grad_norm": 0.7342696189880371,
+      "learning_rate": 0.000166932611571779,
+      "loss": 2.6802,
+      "step": 5405
+    },
+    {
+      "epoch": 0.4362843999677185,
+      "grad_norm": 0.7475131154060364,
+      "learning_rate": 0.0001669208815662017,
+      "loss": 2.6106,
+      "step": 5406
+    },
+    {
+      "epoch": 0.4363651037043015,
+      "grad_norm": 0.7067655324935913,
+      "learning_rate": 0.00016690914989278107,
+      "loss": 2.6362,
+      "step": 5407
+    },
+    {
+      "epoch": 0.4364458074408845,
+      "grad_norm": 0.7550163865089417,
+      "learning_rate": 0.00016689741655180956,
+      "loss": 2.6256,
+      "step": 5408
+    },
+    {
+      "epoch": 0.4365265111774675,
+      "grad_norm": 0.7341828346252441,
+      "learning_rate": 0.00016688568154357952,
+      "loss": 2.6912,
+      "step": 5409
+    },
+    {
+      "epoch": 0.4366072149140505,
+      "grad_norm": 0.7501869201660156,
+      "learning_rate": 0.00016687394486838349,
+      "loss": 2.7122,
+      "step": 5410
+    },
+    {
+      "epoch": 0.4366879186506335,
+      "grad_norm": 0.7041562795639038,
+      "learning_rate": 0.00016686220652651392,
+      "loss": 2.6755,
+      "step": 5411
+    },
+    {
+      "epoch": 0.43676862238721653,
+      "grad_norm": 0.7218217253684998,
+      "learning_rate": 0.00016685046651826338,
+      "loss": 2.693,
+      "step": 5412
+    },
+    {
+      "epoch": 0.4368493261237995,
+      "grad_norm": 0.6880577206611633,
+      "learning_rate": 0.00016683872484392448,
+      "loss": 2.638,
+      "step": 5413
+    },
+    {
+      "epoch": 0.43693002986038254,
+      "grad_norm": 0.6864475607872009,
+      "learning_rate": 0.0001668269815037898,
+      "loss": 2.6497,
+      "step": 5414
+    },
+    {
+      "epoch": 0.4370107335969655,
+      "grad_norm": 0.7326167821884155,
+      "learning_rate": 0.00016681523649815212,
+      "loss": 2.6858,
+      "step": 5415
+    },
+    {
+      "epoch": 0.43709143733354855,
+      "grad_norm": 0.6773428320884705,
+      "learning_rate": 0.00016680348982730405,
+      "loss": 2.6489,
+      "step": 5416
+    },
+    {
+      "epoch": 0.4371721410701315,
+      "grad_norm": 0.7117835283279419,
+      "learning_rate": 0.00016679174149153837,
+      "loss": 2.6607,
+      "step": 5417
+    },
+    {
+      "epoch": 0.43725284480671456,
+      "grad_norm": 0.7268334031105042,
+      "learning_rate": 0.00016677999149114793,
+      "loss": 2.703,
+      "step": 5418
+    },
+    {
+      "epoch": 0.43733354854329753,
+      "grad_norm": 0.7672972679138184,
+      "learning_rate": 0.00016676823982642554,
+      "loss": 2.5803,
+      "step": 5419
+    },
+    {
+      "epoch": 0.43741425227988057,
+      "grad_norm": 0.6966733932495117,
+      "learning_rate": 0.00016675648649766407,
+      "loss": 2.6149,
+      "step": 5420
+    },
+    {
+      "epoch": 0.43749495601646354,
+      "grad_norm": 0.752896249294281,
+      "learning_rate": 0.00016674473150515644,
+      "loss": 2.7108,
+      "step": 5421
+    },
+    {
+      "epoch": 0.4375756597530466,
+      "grad_norm": 0.7094796895980835,
+      "learning_rate": 0.00016673297484919565,
+      "loss": 2.6989,
+      "step": 5422
+    },
+    {
+      "epoch": 0.43765636348962955,
+      "grad_norm": 0.7631612420082092,
+      "learning_rate": 0.00016672121653007465,
+      "loss": 2.6673,
+      "step": 5423
+    },
+    {
+      "epoch": 0.4377370672262126,
+      "grad_norm": 0.7083843946456909,
+      "learning_rate": 0.00016670945654808655,
+      "loss": 2.6529,
+      "step": 5424
+    },
+    {
+      "epoch": 0.43781777096279556,
+      "grad_norm": 0.7291569709777832,
+      "learning_rate": 0.0001666976949035244,
+      "loss": 2.633,
+      "step": 5425
+    },
+    {
+      "epoch": 0.4378984746993786,
+      "grad_norm": 0.8351448774337769,
+      "learning_rate": 0.00016668593159668138,
+      "loss": 2.5993,
+      "step": 5426
+    },
+    {
+      "epoch": 0.43797917843596157,
+      "grad_norm": 0.7339642643928528,
+      "learning_rate": 0.00016667416662785058,
+      "loss": 2.6486,
+      "step": 5427
+    },
+    {
+      "epoch": 0.4380598821725446,
+      "grad_norm": 0.7257512211799622,
+      "learning_rate": 0.00016666239999732526,
+      "loss": 2.6453,
+      "step": 5428
+    },
+    {
+      "epoch": 0.4381405859091276,
+      "grad_norm": 0.7282476425170898,
+      "learning_rate": 0.00016665063170539872,
+      "loss": 2.6654,
+      "step": 5429
+    },
+    {
+      "epoch": 0.4382212896457106,
+      "grad_norm": 0.726685643196106,
+      "learning_rate": 0.00016663886175236417,
+      "loss": 2.65,
+      "step": 5430
+    },
+    {
+      "epoch": 0.4383019933822936,
+      "grad_norm": 0.7478880286216736,
+      "learning_rate": 0.000166627090138515,
+      "loss": 2.623,
+      "step": 5431
+    },
+    {
+      "epoch": 0.4383826971188766,
+      "grad_norm": 0.7624948024749756,
+      "learning_rate": 0.00016661531686414457,
+      "loss": 2.6438,
+      "step": 5432
+    },
+    {
+      "epoch": 0.4384634008554596,
+      "grad_norm": 0.8098936676979065,
+      "learning_rate": 0.00016660354192954633,
+      "loss": 2.6226,
+      "step": 5433
+    },
+    {
+      "epoch": 0.4385441045920426,
+      "grad_norm": 0.7305725812911987,
+      "learning_rate": 0.0001665917653350137,
+      "loss": 2.6425,
+      "step": 5434
+    },
+    {
+      "epoch": 0.4386248083286256,
+      "grad_norm": 0.7064421772956848,
+      "learning_rate": 0.00016657998708084027,
+      "loss": 2.6069,
+      "step": 5435
+    },
+    {
+      "epoch": 0.43870551206520864,
+      "grad_norm": 0.8279524445533752,
+      "learning_rate": 0.00016656820716731945,
+      "loss": 2.6609,
+      "step": 5436
+    },
+    {
+      "epoch": 0.4387862158017916,
+      "grad_norm": 0.742659866809845,
+      "learning_rate": 0.00016655642559474488,
+      "loss": 2.64,
+      "step": 5437
+    },
+    {
+      "epoch": 0.43886691953837464,
+      "grad_norm": 0.757780909538269,
+      "learning_rate": 0.00016654464236341026,
+      "loss": 2.6546,
+      "step": 5438
+    },
+    {
+      "epoch": 0.4389476232749576,
+      "grad_norm": 0.7439742684364319,
+      "learning_rate": 0.00016653285747360918,
+      "loss": 2.6717,
+      "step": 5439
+    },
+    {
+      "epoch": 0.43902832701154065,
+      "grad_norm": 0.7529581189155579,
+      "learning_rate": 0.0001665210709256354,
+      "loss": 2.6204,
+      "step": 5440
+    },
+    {
+      "epoch": 0.43910903074812363,
+      "grad_norm": 0.7224153876304626,
+      "learning_rate": 0.00016650928271978258,
+      "loss": 2.6417,
+      "step": 5441
+    },
+    {
+      "epoch": 0.43918973448470666,
+      "grad_norm": 0.6792185306549072,
+      "learning_rate": 0.00016649749285634462,
+      "loss": 2.6382,
+      "step": 5442
+    },
+    {
+      "epoch": 0.43927043822128964,
+      "grad_norm": 0.6887058019638062,
+      "learning_rate": 0.00016648570133561533,
+      "loss": 2.6302,
+      "step": 5443
+    },
+    {
+      "epoch": 0.43935114195787267,
+      "grad_norm": 0.7373671531677246,
+      "learning_rate": 0.00016647390815788853,
+      "loss": 2.625,
+      "step": 5444
+    },
+    {
+      "epoch": 0.43943184569445565,
+      "grad_norm": 0.7595719695091248,
+      "learning_rate": 0.0001664621133234582,
+      "loss": 2.6444,
+      "step": 5445
+    },
+    {
+      "epoch": 0.4395125494310387,
+      "grad_norm": 0.7331473231315613,
+      "learning_rate": 0.00016645031683261825,
+      "loss": 2.6308,
+      "step": 5446
+    },
+    {
+      "epoch": 0.43959325316762166,
+      "grad_norm": 0.7724922895431519,
+      "learning_rate": 0.0001664385186856627,
+      "loss": 2.6646,
+      "step": 5447
+    },
+    {
+      "epoch": 0.4396739569042047,
+      "grad_norm": 0.6960163712501526,
+      "learning_rate": 0.00016642671888288563,
+      "loss": 2.6196,
+      "step": 5448
+    },
+    {
+      "epoch": 0.43975466064078766,
+      "grad_norm": 0.6769189834594727,
+      "learning_rate": 0.00016641491742458103,
+      "loss": 2.6558,
+      "step": 5449
+    },
+    {
+      "epoch": 0.4398353643773707,
+      "grad_norm": 0.7435783743858337,
+      "learning_rate": 0.0001664031143110431,
+      "loss": 2.6717,
+      "step": 5450
+    },
+    {
+      "epoch": 0.4399160681139537,
+      "grad_norm": 0.7234118580818176,
+      "learning_rate": 0.00016639130954256603,
+      "loss": 2.6549,
+      "step": 5451
+    },
+    {
+      "epoch": 0.4399967718505367,
+      "grad_norm": 0.720825731754303,
+      "learning_rate": 0.00016637950311944392,
+      "loss": 2.6098,
+      "step": 5452
+    },
+    {
+      "epoch": 0.4400774755871197,
+      "grad_norm": 0.6977505087852478,
+      "learning_rate": 0.0001663676950419711,
+      "loss": 2.6351,
+      "step": 5453
+    },
+    {
+      "epoch": 0.4401581793237027,
+      "grad_norm": 0.6959076523780823,
+      "learning_rate": 0.00016635588531044185,
+      "loss": 2.6918,
+      "step": 5454
+    },
+    {
+      "epoch": 0.4402388830602857,
+      "grad_norm": 0.7022189497947693,
+      "learning_rate": 0.00016634407392515044,
+      "loss": 2.6218,
+      "step": 5455
+    },
+    {
+      "epoch": 0.4403195867968687,
+      "grad_norm": 0.7147775292396545,
+      "learning_rate": 0.0001663322608863913,
+      "loss": 2.6966,
+      "step": 5456
+    },
+    {
+      "epoch": 0.4404002905334517,
+      "grad_norm": 0.7592755556106567,
+      "learning_rate": 0.00016632044619445882,
+      "loss": 2.6326,
+      "step": 5457
+    },
+    {
+      "epoch": 0.4404809942700347,
+      "grad_norm": 0.6914302110671997,
+      "learning_rate": 0.00016630862984964745,
+      "loss": 2.603,
+      "step": 5458
+    },
+    {
+      "epoch": 0.4405616980066177,
+      "grad_norm": 0.7735368609428406,
+      "learning_rate": 0.0001662968118522517,
+      "loss": 2.6666,
+      "step": 5459
+    },
+    {
+      "epoch": 0.4406424017432007,
+      "grad_norm": 0.7175899744033813,
+      "learning_rate": 0.00016628499220256612,
+      "loss": 2.666,
+      "step": 5460
+    },
+    {
+      "epoch": 0.4407231054797837,
+      "grad_norm": 0.6735796332359314,
+      "learning_rate": 0.00016627317090088523,
+      "loss": 2.6451,
+      "step": 5461
+    },
+    {
+      "epoch": 0.4408038092163667,
+      "grad_norm": 0.72022545337677,
+      "learning_rate": 0.0001662613479475037,
+      "loss": 2.6295,
+      "step": 5462
+    },
+    {
+      "epoch": 0.4408845129529497,
+      "grad_norm": 0.7084751725196838,
+      "learning_rate": 0.00016624952334271616,
+      "loss": 2.6633,
+      "step": 5463
+    },
+    {
+      "epoch": 0.4409652166895327,
+      "grad_norm": 0.7399250864982605,
+      "learning_rate": 0.00016623769708681735,
+      "loss": 2.6076,
+      "step": 5464
+    },
+    {
+      "epoch": 0.44104592042611573,
+      "grad_norm": 0.6904892325401306,
+      "learning_rate": 0.00016622586918010193,
+      "loss": 2.6799,
+      "step": 5465
+    },
+    {
+      "epoch": 0.4411266241626987,
+      "grad_norm": 0.7419006824493408,
+      "learning_rate": 0.00016621403962286478,
+      "loss": 2.65,
+      "step": 5466
+    },
+    {
+      "epoch": 0.44120732789928174,
+      "grad_norm": 0.7201282978057861,
+      "learning_rate": 0.00016620220841540064,
+      "loss": 2.6769,
+      "step": 5467
+    },
+    {
+      "epoch": 0.4412880316358647,
+      "grad_norm": 0.7223218679428101,
+      "learning_rate": 0.00016619037555800443,
+      "loss": 2.6342,
+      "step": 5468
+    },
+    {
+      "epoch": 0.44136873537244775,
+      "grad_norm": 0.7517585754394531,
+      "learning_rate": 0.00016617854105097104,
+      "loss": 2.6103,
+      "step": 5469
+    },
+    {
+      "epoch": 0.44144943910903073,
+      "grad_norm": 0.6765139698982239,
+      "learning_rate": 0.0001661667048945954,
+      "loss": 2.624,
+      "step": 5470
+    },
+    {
+      "epoch": 0.44153014284561376,
+      "grad_norm": 0.7197677493095398,
+      "learning_rate": 0.00016615486708917255,
+      "loss": 2.5786,
+      "step": 5471
+    },
+    {
+      "epoch": 0.44161084658219674,
+      "grad_norm": 0.7196774482727051,
+      "learning_rate": 0.00016614302763499742,
+      "loss": 2.6147,
+      "step": 5472
+    },
+    {
+      "epoch": 0.44169155031877977,
+      "grad_norm": 0.7210293412208557,
+      "learning_rate": 0.00016613118653236518,
+      "loss": 2.6526,
+      "step": 5473
+    },
+    {
+      "epoch": 0.44177225405536275,
+      "grad_norm": 0.6870129108428955,
+      "learning_rate": 0.00016611934378157092,
+      "loss": 2.665,
+      "step": 5474
+    },
+    {
+      "epoch": 0.4418529577919458,
+      "grad_norm": 0.6925365328788757,
+      "learning_rate": 0.00016610749938290975,
+      "loss": 2.5734,
+      "step": 5475
+    },
+    {
+      "epoch": 0.44193366152852875,
+      "grad_norm": 0.7399131655693054,
+      "learning_rate": 0.0001660956533366769,
+      "loss": 2.6935,
+      "step": 5476
+    },
+    {
+      "epoch": 0.4420143652651118,
+      "grad_norm": 0.7348966002464294,
+      "learning_rate": 0.00016608380564316758,
+      "loss": 2.6788,
+      "step": 5477
+    },
+    {
+      "epoch": 0.44209506900169476,
+      "grad_norm": 0.7597334980964661,
+      "learning_rate": 0.00016607195630267708,
+      "loss": 2.6732,
+      "step": 5478
+    },
+    {
+      "epoch": 0.4421757727382778,
+      "grad_norm": 0.6847043037414551,
+      "learning_rate": 0.00016606010531550072,
+      "loss": 2.6475,
+      "step": 5479
+    },
+    {
+      "epoch": 0.44225647647486077,
+      "grad_norm": 0.7065151929855347,
+      "learning_rate": 0.00016604825268193388,
+      "loss": 2.6674,
+      "step": 5480
+    },
+    {
+      "epoch": 0.4423371802114438,
+      "grad_norm": 0.7102208137512207,
+      "learning_rate": 0.0001660363984022719,
+      "loss": 2.6723,
+      "step": 5481
+    },
+    {
+      "epoch": 0.4424178839480268,
+      "grad_norm": 0.6912767887115479,
+      "learning_rate": 0.00016602454247681024,
+      "loss": 2.628,
+      "step": 5482
+    },
+    {
+      "epoch": 0.4424985876846098,
+      "grad_norm": 0.7265123128890991,
+      "learning_rate": 0.0001660126849058444,
+      "loss": 2.5935,
+      "step": 5483
+    },
+    {
+      "epoch": 0.4425792914211928,
+      "grad_norm": 0.8177923560142517,
+      "learning_rate": 0.0001660008256896699,
+      "loss": 2.6402,
+      "step": 5484
+    },
+    {
+      "epoch": 0.4426599951577758,
+      "grad_norm": 0.7196556925773621,
+      "learning_rate": 0.00016598896482858231,
+      "loss": 2.6939,
+      "step": 5485
+    },
+    {
+      "epoch": 0.4427406988943588,
+      "grad_norm": 0.7459850907325745,
+      "learning_rate": 0.0001659771023228772,
+      "loss": 2.6343,
+      "step": 5486
+    },
+    {
+      "epoch": 0.44282140263094183,
+      "grad_norm": 0.7399095892906189,
+      "learning_rate": 0.00016596523817285024,
+      "loss": 2.6139,
+      "step": 5487
+    },
+    {
+      "epoch": 0.4429021063675248,
+      "grad_norm": 0.7517558336257935,
+      "learning_rate": 0.0001659533723787971,
+      "loss": 2.6609,
+      "step": 5488
+    },
+    {
+      "epoch": 0.44298281010410784,
+      "grad_norm": 0.7073537707328796,
+      "learning_rate": 0.00016594150494101355,
+      "loss": 2.6326,
+      "step": 5489
+    },
+    {
+      "epoch": 0.4430635138406908,
+      "grad_norm": 0.7414752244949341,
+      "learning_rate": 0.0001659296358597953,
+      "loss": 2.6759,
+      "step": 5490
+    },
+    {
+      "epoch": 0.44314421757727385,
+      "grad_norm": 0.7636380195617676,
+      "learning_rate": 0.0001659177651354382,
+      "loss": 2.5743,
+      "step": 5491
+    },
+    {
+      "epoch": 0.4432249213138568,
+      "grad_norm": 0.6839539408683777,
+      "learning_rate": 0.00016590589276823804,
+      "loss": 2.631,
+      "step": 5492
+    },
+    {
+      "epoch": 0.44330562505043986,
+      "grad_norm": 0.8057516813278198,
+      "learning_rate": 0.0001658940187584908,
+      "loss": 2.6916,
+      "step": 5493
+    },
+    {
+      "epoch": 0.44338632878702283,
+      "grad_norm": 0.7479767799377441,
+      "learning_rate": 0.00016588214310649232,
+      "loss": 2.6811,
+      "step": 5494
+    },
+    {
+      "epoch": 0.44346703252360586,
+      "grad_norm": 0.7854729294776917,
+      "learning_rate": 0.00016587026581253866,
+      "loss": 2.6746,
+      "step": 5495
+    },
+    {
+      "epoch": 0.44354773626018884,
+      "grad_norm": 0.7782836556434631,
+      "learning_rate": 0.00016585838687692577,
+      "loss": 2.61,
+      "step": 5496
+    },
+    {
+      "epoch": 0.4436284399967719,
+      "grad_norm": 0.7047034502029419,
+      "learning_rate": 0.00016584650629994968,
+      "loss": 2.6573,
+      "step": 5497
+    },
+    {
+      "epoch": 0.44370914373335485,
+      "grad_norm": 0.7398735880851746,
+      "learning_rate": 0.0001658346240819066,
+      "loss": 2.6338,
+      "step": 5498
+    },
+    {
+      "epoch": 0.4437898474699379,
+      "grad_norm": 0.7243468165397644,
+      "learning_rate": 0.00016582274022309258,
+      "loss": 2.5898,
+      "step": 5499
+    },
+    {
+      "epoch": 0.44387055120652086,
+      "grad_norm": 0.7415906190872192,
+      "learning_rate": 0.00016581085472380376,
+      "loss": 2.5893,
+      "step": 5500
+    },
+    {
+      "epoch": 0.4439512549431039,
+      "grad_norm": 0.6935107707977295,
+      "learning_rate": 0.00016579896758433645,
+      "loss": 2.6704,
+      "step": 5501
+    },
+    {
+      "epoch": 0.44403195867968687,
+      "grad_norm": 0.7188034653663635,
+      "learning_rate": 0.00016578707880498685,
+      "loss": 2.643,
+      "step": 5502
+    },
+    {
+      "epoch": 0.4441126624162699,
+      "grad_norm": 0.6697022914886475,
+      "learning_rate": 0.0001657751883860513,
+      "loss": 2.6313,
+      "step": 5503
+    },
+    {
+      "epoch": 0.4441933661528529,
+      "grad_norm": 0.760154664516449,
+      "learning_rate": 0.00016576329632782613,
+      "loss": 2.6604,
+      "step": 5504
+    },
+    {
+      "epoch": 0.4442740698894359,
+      "grad_norm": 0.6883447170257568,
+      "learning_rate": 0.00016575140263060765,
+      "loss": 2.64,
+      "step": 5505
+    },
+    {
+      "epoch": 0.4443547736260189,
+      "grad_norm": 0.8628804683685303,
+      "learning_rate": 0.0001657395072946924,
+      "loss": 2.6651,
+      "step": 5506
+    },
+    {
+      "epoch": 0.4444354773626019,
+      "grad_norm": 0.7125170230865479,
+      "learning_rate": 0.0001657276103203768,
+      "loss": 2.7132,
+      "step": 5507
+    },
+    {
+      "epoch": 0.4445161810991849,
+      "grad_norm": 0.6965304613113403,
+      "learning_rate": 0.00016571571170795725,
+      "loss": 2.7109,
+      "step": 5508
+    },
+    {
+      "epoch": 0.44459688483576787,
+      "grad_norm": 0.720327615737915,
+      "learning_rate": 0.00016570381145773042,
+      "loss": 2.6323,
+      "step": 5509
+    },
+    {
+      "epoch": 0.4446775885723509,
+      "grad_norm": 0.7097898125648499,
+      "learning_rate": 0.00016569190956999287,
+      "loss": 2.6461,
+      "step": 5510
+    },
+    {
+      "epoch": 0.4447582923089339,
+      "grad_norm": 0.7142884731292725,
+      "learning_rate": 0.0001656800060450412,
+      "loss": 2.6894,
+      "step": 5511
+    },
+    {
+      "epoch": 0.4448389960455169,
+      "grad_norm": 0.6992002725601196,
+      "learning_rate": 0.0001656681008831721,
+      "loss": 2.6116,
+      "step": 5512
+    },
+    {
+      "epoch": 0.4449196997820999,
+      "grad_norm": 0.763841450214386,
+      "learning_rate": 0.00016565619408468227,
+      "loss": 2.6441,
+      "step": 5513
+    },
+    {
+      "epoch": 0.4450004035186829,
+      "grad_norm": 0.6958404183387756,
+      "learning_rate": 0.00016564428564986848,
+      "loss": 2.5751,
+      "step": 5514
+    },
+    {
+      "epoch": 0.4450811072552659,
+      "grad_norm": 0.8804046511650085,
+      "learning_rate": 0.00016563237557902744,
+      "loss": 2.6353,
+      "step": 5515
+    },
+    {
+      "epoch": 0.4451618109918489,
+      "grad_norm": 0.744864821434021,
+      "learning_rate": 0.00016562046387245608,
+      "loss": 2.6887,
+      "step": 5516
+    },
+    {
+      "epoch": 0.4452425147284319,
+      "grad_norm": 0.7627978920936584,
+      "learning_rate": 0.0001656085505304512,
+      "loss": 2.6347,
+      "step": 5517
+    },
+    {
+      "epoch": 0.44532321846501494,
+      "grad_norm": 0.7728918194770813,
+      "learning_rate": 0.00016559663555330975,
+      "loss": 2.6344,
+      "step": 5518
+    },
+    {
+      "epoch": 0.4454039222015979,
+      "grad_norm": 0.7853842377662659,
+      "learning_rate": 0.00016558471894132865,
+      "loss": 2.7239,
+      "step": 5519
+    },
+    {
+      "epoch": 0.44548462593818094,
+      "grad_norm": 0.7981860041618347,
+      "learning_rate": 0.00016557280069480495,
+      "loss": 2.66,
+      "step": 5520
+    },
+    {
+      "epoch": 0.4455653296747639,
+      "grad_norm": 0.7555295825004578,
+      "learning_rate": 0.0001655608808140356,
+      "loss": 2.6636,
+      "step": 5521
+    },
+    {
+      "epoch": 0.44564603341134695,
+      "grad_norm": 0.6893854141235352,
+      "learning_rate": 0.00016554895929931778,
+      "loss": 2.5999,
+      "step": 5522
+    },
+    {
+      "epoch": 0.44572673714792993,
+      "grad_norm": 0.7740506529808044,
+      "learning_rate": 0.0001655370361509485,
+      "loss": 2.6308,
+      "step": 5523
+    },
+    {
+      "epoch": 0.44580744088451296,
+      "grad_norm": 0.6956021785736084,
+      "learning_rate": 0.00016552511136922498,
+      "loss": 2.6376,
+      "step": 5524
+    },
+    {
+      "epoch": 0.44588814462109594,
+      "grad_norm": 0.7408841252326965,
+      "learning_rate": 0.00016551318495444445,
+      "loss": 2.6644,
+      "step": 5525
+    },
+    {
+      "epoch": 0.44596884835767897,
+      "grad_norm": 0.7715663313865662,
+      "learning_rate": 0.000165501256906904,
+      "loss": 2.6791,
+      "step": 5526
+    },
+    {
+      "epoch": 0.44604955209426195,
+      "grad_norm": 0.6880629062652588,
+      "learning_rate": 0.0001654893272269011,
+      "loss": 2.7209,
+      "step": 5527
+    },
+    {
+      "epoch": 0.446130255830845,
+      "grad_norm": 0.6765853762626648,
+      "learning_rate": 0.0001654773959147329,
+      "loss": 2.6548,
+      "step": 5528
+    },
+    {
+      "epoch": 0.44621095956742796,
+      "grad_norm": 0.739248514175415,
+      "learning_rate": 0.00016546546297069688,
+      "loss": 2.69,
+      "step": 5529
+    },
+    {
+      "epoch": 0.446291663304011,
+      "grad_norm": 0.7655714750289917,
+      "learning_rate": 0.00016545352839509038,
+      "loss": 2.6238,
+      "step": 5530
+    },
+    {
+      "epoch": 0.44637236704059396,
+      "grad_norm": 0.706068217754364,
+      "learning_rate": 0.00016544159218821088,
+      "loss": 2.6528,
+      "step": 5531
+    },
+    {
+      "epoch": 0.446453070777177,
+      "grad_norm": 0.7411316633224487,
+      "learning_rate": 0.00016542965435035578,
+      "loss": 2.7034,
+      "step": 5532
+    },
+    {
+      "epoch": 0.44653377451376,
+      "grad_norm": 0.6550690531730652,
+      "learning_rate": 0.0001654177148818227,
+      "loss": 2.6388,
+      "step": 5533
+    },
+    {
+      "epoch": 0.446614478250343,
+      "grad_norm": 0.7151147127151489,
+      "learning_rate": 0.00016540577378290915,
+      "loss": 2.7382,
+      "step": 5534
+    },
+    {
+      "epoch": 0.446695181986926,
+      "grad_norm": 0.7343939542770386,
+      "learning_rate": 0.00016539383105391276,
+      "loss": 2.6316,
+      "step": 5535
+    },
+    {
+      "epoch": 0.446775885723509,
+      "grad_norm": 0.702036440372467,
+      "learning_rate": 0.00016538188669513115,
+      "loss": 2.6465,
+      "step": 5536
+    },
+    {
+      "epoch": 0.446856589460092,
+      "grad_norm": 0.7212840914726257,
+      "learning_rate": 0.00016536994070686197,
+      "loss": 2.6471,
+      "step": 5537
+    },
+    {
+      "epoch": 0.446937293196675,
+      "grad_norm": 0.7345479130744934,
+      "learning_rate": 0.00016535799308940304,
+      "loss": 2.6746,
+      "step": 5538
+    },
+    {
+      "epoch": 0.447017996933258,
+      "grad_norm": 0.7447341084480286,
+      "learning_rate": 0.00016534604384305207,
+      "loss": 2.6487,
+      "step": 5539
+    },
+    {
+      "epoch": 0.44709870066984103,
+      "grad_norm": 0.6865687370300293,
+      "learning_rate": 0.00016533409296810687,
+      "loss": 2.6202,
+      "step": 5540
+    },
+    {
+      "epoch": 0.447179404406424,
+      "grad_norm": 0.8210769891738892,
+      "learning_rate": 0.0001653221404648653,
+      "loss": 2.7155,
+      "step": 5541
+    },
+    {
+      "epoch": 0.44726010814300704,
+      "grad_norm": 0.7768925428390503,
+      "learning_rate": 0.0001653101863336252,
+      "loss": 2.6011,
+      "step": 5542
+    },
+    {
+      "epoch": 0.44734081187959,
+      "grad_norm": 0.7160049080848694,
+      "learning_rate": 0.00016529823057468456,
+      "loss": 2.6541,
+      "step": 5543
+    },
+    {
+      "epoch": 0.44742151561617305,
+      "grad_norm": 0.7386900782585144,
+      "learning_rate": 0.00016528627318834134,
+      "loss": 2.6586,
+      "step": 5544
+    },
+    {
+      "epoch": 0.447502219352756,
+      "grad_norm": 0.7415460348129272,
+      "learning_rate": 0.0001652743141748935,
+      "loss": 2.7032,
+      "step": 5545
+    },
+    {
+      "epoch": 0.44758292308933906,
+      "grad_norm": 0.8483054637908936,
+      "learning_rate": 0.00016526235353463912,
+      "loss": 2.6145,
+      "step": 5546
+    },
+    {
+      "epoch": 0.44766362682592203,
+      "grad_norm": 0.7428778409957886,
+      "learning_rate": 0.00016525039126787629,
+      "loss": 2.7005,
+      "step": 5547
+    },
+    {
+      "epoch": 0.44774433056250507,
+      "grad_norm": 0.7214285731315613,
+      "learning_rate": 0.00016523842737490316,
+      "loss": 2.6267,
+      "step": 5548
+    },
+    {
+      "epoch": 0.44782503429908804,
+      "grad_norm": 0.6753950715065002,
+      "learning_rate": 0.0001652264618560179,
+      "loss": 2.6732,
+      "step": 5549
+    },
+    {
+      "epoch": 0.4479057380356711,
+      "grad_norm": 0.6969403028488159,
+      "learning_rate": 0.00016521449471151867,
+      "loss": 2.6218,
+      "step": 5550
+    },
+    {
+      "epoch": 0.44798644177225405,
+      "grad_norm": 0.7562664151191711,
+      "learning_rate": 0.00016520252594170377,
+      "loss": 2.69,
+      "step": 5551
+    },
+    {
+      "epoch": 0.4480671455088371,
+      "grad_norm": 0.6831937432289124,
+      "learning_rate": 0.0001651905555468715,
+      "loss": 2.709,
+      "step": 5552
+    },
+    {
+      "epoch": 0.44814784924542006,
+      "grad_norm": 0.6753427386283875,
+      "learning_rate": 0.00016517858352732017,
+      "loss": 2.5852,
+      "step": 5553
+    },
+    {
+      "epoch": 0.4482285529820031,
+      "grad_norm": 0.7573871612548828,
+      "learning_rate": 0.00016516660988334815,
+      "loss": 2.6187,
+      "step": 5554
+    },
+    {
+      "epoch": 0.44830925671858607,
+      "grad_norm": 0.6424254775047302,
+      "learning_rate": 0.00016515463461525383,
+      "loss": 2.6411,
+      "step": 5555
+    },
+    {
+      "epoch": 0.4483899604551691,
+      "grad_norm": 0.7460073232650757,
+      "learning_rate": 0.0001651426577233358,
+      "loss": 2.6239,
+      "step": 5556
+    },
+    {
+      "epoch": 0.4484706641917521,
+      "grad_norm": 0.6980866193771362,
+      "learning_rate": 0.0001651306792078924,
+      "loss": 2.605,
+      "step": 5557
+    },
+    {
+      "epoch": 0.4485513679283351,
+      "grad_norm": 0.7376009225845337,
+      "learning_rate": 0.00016511869906922217,
+      "loss": 2.7114,
+      "step": 5558
+    },
+    {
+      "epoch": 0.4486320716649181,
+      "grad_norm": 0.7227364778518677,
+      "learning_rate": 0.0001651067173076238,
+      "loss": 2.6212,
+      "step": 5559
+    },
+    {
+      "epoch": 0.44871277540150106,
+      "grad_norm": 0.8989635705947876,
+      "learning_rate": 0.00016509473392339584,
+      "loss": 2.671,
+      "step": 5560
+    },
+    {
+      "epoch": 0.4487934791380841,
+      "grad_norm": 0.7273553609848022,
+      "learning_rate": 0.0001650827489168369,
+      "loss": 2.6556,
+      "step": 5561
+    },
+    {
+      "epoch": 0.44887418287466707,
+      "grad_norm": 0.839439868927002,
+      "learning_rate": 0.00016507076228824578,
+      "loss": 2.6959,
+      "step": 5562
+    },
+    {
+      "epoch": 0.4489548866112501,
+      "grad_norm": 0.6912770867347717,
+      "learning_rate": 0.00016505877403792115,
+      "loss": 2.6709,
+      "step": 5563
+    },
+    {
+      "epoch": 0.4490355903478331,
+      "grad_norm": 0.7850949168205261,
+      "learning_rate": 0.00016504678416616182,
+      "loss": 2.7257,
+      "step": 5564
+    },
+    {
+      "epoch": 0.4491162940844161,
+      "grad_norm": 0.7768355011940002,
+      "learning_rate": 0.0001650347926732666,
+      "loss": 2.5939,
+      "step": 5565
+    },
+    {
+      "epoch": 0.4491969978209991,
+      "grad_norm": 0.6518398523330688,
+      "learning_rate": 0.0001650227995595343,
+      "loss": 2.6589,
+      "step": 5566
+    },
+    {
+      "epoch": 0.4492777015575821,
+      "grad_norm": 0.6855975389480591,
+      "learning_rate": 0.0001650108048252639,
+      "loss": 2.6372,
+      "step": 5567
+    },
+    {
+      "epoch": 0.4493584052941651,
+      "grad_norm": 0.7176938056945801,
+      "learning_rate": 0.0001649988084707543,
+      "loss": 2.6506,
+      "step": 5568
+    },
+    {
+      "epoch": 0.44943910903074813,
+      "grad_norm": 0.735335648059845,
+      "learning_rate": 0.00016498681049630448,
+      "loss": 2.608,
+      "step": 5569
+    },
+    {
+      "epoch": 0.4495198127673311,
+      "grad_norm": 0.6862306594848633,
+      "learning_rate": 0.00016497481090221346,
+      "loss": 2.5982,
+      "step": 5570
+    },
+    {
+      "epoch": 0.44960051650391414,
+      "grad_norm": 0.7213380336761475,
+      "learning_rate": 0.0001649628096887803,
+      "loss": 2.6457,
+      "step": 5571
+    },
+    {
+      "epoch": 0.4496812202404971,
+      "grad_norm": 0.7118985652923584,
+      "learning_rate": 0.0001649508068563041,
+      "loss": 2.6321,
+      "step": 5572
+    },
+    {
+      "epoch": 0.44976192397708015,
+      "grad_norm": 0.7663396596908569,
+      "learning_rate": 0.00016493880240508405,
+      "loss": 2.5865,
+      "step": 5573
+    },
+    {
+      "epoch": 0.4498426277136631,
+      "grad_norm": 0.6854543089866638,
+      "learning_rate": 0.00016492679633541926,
+      "loss": 2.6536,
+      "step": 5574
+    },
+    {
+      "epoch": 0.44992333145024616,
+      "grad_norm": 0.7071701884269714,
+      "learning_rate": 0.000164914788647609,
+      "loss": 2.6149,
+      "step": 5575
+    },
+    {
+      "epoch": 0.45000403518682913,
+      "grad_norm": 0.7610478401184082,
+      "learning_rate": 0.00016490277934195252,
+      "loss": 2.6326,
+      "step": 5576
+    },
+    {
+      "epoch": 0.45008473892341216,
+      "grad_norm": 0.7117596864700317,
+      "learning_rate": 0.0001648907684187491,
+      "loss": 2.6938,
+      "step": 5577
+    },
+    {
+      "epoch": 0.45016544265999514,
+      "grad_norm": 0.6980494856834412,
+      "learning_rate": 0.00016487875587829813,
+      "loss": 2.6798,
+      "step": 5578
+    },
+    {
+      "epoch": 0.4502461463965782,
+      "grad_norm": 0.7957972288131714,
+      "learning_rate": 0.00016486674172089898,
+      "loss": 2.6029,
+      "step": 5579
+    },
+    {
+      "epoch": 0.45032685013316115,
+      "grad_norm": 0.7258082032203674,
+      "learning_rate": 0.00016485472594685103,
+      "loss": 2.6785,
+      "step": 5580
+    },
+    {
+      "epoch": 0.4504075538697442,
+      "grad_norm": 0.7402041554450989,
+      "learning_rate": 0.0001648427085564538,
+      "loss": 2.6263,
+      "step": 5581
+    },
+    {
+      "epoch": 0.45048825760632716,
+      "grad_norm": 0.6943814158439636,
+      "learning_rate": 0.00016483068955000673,
+      "loss": 2.6761,
+      "step": 5582
+    },
+    {
+      "epoch": 0.4505689613429102,
+      "grad_norm": 0.8021644353866577,
+      "learning_rate": 0.00016481866892780947,
+      "loss": 2.6376,
+      "step": 5583
+    },
+    {
+      "epoch": 0.45064966507949317,
+      "grad_norm": 0.7748533487319946,
+      "learning_rate": 0.0001648066466901615,
+      "loss": 2.7465,
+      "step": 5584
+    },
+    {
+      "epoch": 0.4507303688160762,
+      "grad_norm": 0.7432222366333008,
+      "learning_rate": 0.00016479462283736248,
+      "loss": 2.6368,
+      "step": 5585
+    },
+    {
+      "epoch": 0.4508110725526592,
+      "grad_norm": 0.7835286259651184,
+      "learning_rate": 0.00016478259736971214,
+      "loss": 2.6449,
+      "step": 5586
+    },
+    {
+      "epoch": 0.4508917762892422,
+      "grad_norm": 0.7372995018959045,
+      "learning_rate": 0.00016477057028751007,
+      "loss": 2.6091,
+      "step": 5587
+    },
+    {
+      "epoch": 0.4509724800258252,
+      "grad_norm": 0.8230665326118469,
+      "learning_rate": 0.0001647585415910561,
+      "loss": 2.6345,
+      "step": 5588
+    },
+    {
+      "epoch": 0.4510531837624082,
+      "grad_norm": 0.7490825057029724,
+      "learning_rate": 0.00016474651128065002,
+      "loss": 2.5996,
+      "step": 5589
+    },
+    {
+      "epoch": 0.4511338874989912,
+      "grad_norm": 0.7950569987297058,
+      "learning_rate": 0.00016473447935659157,
+      "loss": 2.7109,
+      "step": 5590
+    },
+    {
+      "epoch": 0.4512145912355742,
+      "grad_norm": 0.7648342251777649,
+      "learning_rate": 0.00016472244581918074,
+      "loss": 2.6268,
+      "step": 5591
+    },
+    {
+      "epoch": 0.4512952949721572,
+      "grad_norm": 0.726828396320343,
+      "learning_rate": 0.00016471041066871733,
+      "loss": 2.5959,
+      "step": 5592
+    },
+    {
+      "epoch": 0.45137599870874023,
+      "grad_norm": 0.7855841517448425,
+      "learning_rate": 0.00016469837390550133,
+      "loss": 2.6671,
+      "step": 5593
+    },
+    {
+      "epoch": 0.4514567024453232,
+      "grad_norm": 0.6858882904052734,
+      "learning_rate": 0.00016468633552983275,
+      "loss": 2.6003,
+      "step": 5594
+    },
+    {
+      "epoch": 0.45153740618190624,
+      "grad_norm": 0.710926353931427,
+      "learning_rate": 0.0001646742955420116,
+      "loss": 2.6049,
+      "step": 5595
+    },
+    {
+      "epoch": 0.4516181099184892,
+      "grad_norm": 0.8359978199005127,
+      "learning_rate": 0.0001646622539423379,
+      "loss": 2.6636,
+      "step": 5596
+    },
+    {
+      "epoch": 0.45169881365507225,
+      "grad_norm": 0.7628041505813599,
+      "learning_rate": 0.00016465021073111186,
+      "loss": 2.6586,
+      "step": 5597
+    },
+    {
+      "epoch": 0.4517795173916552,
+      "grad_norm": 0.7723419666290283,
+      "learning_rate": 0.00016463816590863356,
+      "loss": 2.6213,
+      "step": 5598
+    },
+    {
+      "epoch": 0.45186022112823826,
+      "grad_norm": 0.7210986018180847,
+      "learning_rate": 0.0001646261194752032,
+      "loss": 2.6674,
+      "step": 5599
+    },
+    {
+      "epoch": 0.45194092486482124,
+      "grad_norm": 0.7665949463844299,
+      "learning_rate": 0.00016461407143112097,
+      "loss": 2.68,
+      "step": 5600
+    },
+    {
+      "epoch": 0.45202162860140427,
+      "grad_norm": 0.7225117087364197,
+      "learning_rate": 0.00016460202177668722,
+      "loss": 2.6473,
+      "step": 5601
+    },
+    {
+      "epoch": 0.45210233233798724,
+      "grad_norm": 0.6831738948822021,
+      "learning_rate": 0.0001645899705122022,
+      "loss": 2.6863,
+      "step": 5602
+    },
+    {
+      "epoch": 0.4521830360745703,
+      "grad_norm": 0.7006321549415588,
+      "learning_rate": 0.00016457791763796627,
+      "loss": 2.6242,
+      "step": 5603
+    },
+    {
+      "epoch": 0.45226373981115325,
+      "grad_norm": 0.7245663404464722,
+      "learning_rate": 0.00016456586315427983,
+      "loss": 2.6201,
+      "step": 5604
+    },
+    {
+      "epoch": 0.4523444435477363,
+      "grad_norm": 0.7444287538528442,
+      "learning_rate": 0.00016455380706144332,
+      "loss": 2.6684,
+      "step": 5605
+    },
+    {
+      "epoch": 0.45242514728431926,
+      "grad_norm": 0.6562673449516296,
+      "learning_rate": 0.00016454174935975714,
+      "loss": 2.5912,
+      "step": 5606
+    },
+    {
+      "epoch": 0.4525058510209023,
+      "grad_norm": 0.6494336724281311,
+      "learning_rate": 0.0001645296900495219,
+      "loss": 2.6245,
+      "step": 5607
+    },
+    {
+      "epoch": 0.45258655475748527,
+      "grad_norm": 0.6968161463737488,
+      "learning_rate": 0.0001645176291310381,
+      "loss": 2.6494,
+      "step": 5608
+    },
+    {
+      "epoch": 0.4526672584940683,
+      "grad_norm": 0.7351142764091492,
+      "learning_rate": 0.00016450556660460632,
+      "loss": 2.574,
+      "step": 5609
+    },
+    {
+      "epoch": 0.4527479622306513,
+      "grad_norm": 0.7522323131561279,
+      "learning_rate": 0.0001644935024705272,
+      "loss": 2.6512,
+      "step": 5610
+    },
+    {
+      "epoch": 0.45282866596723426,
+      "grad_norm": 0.6744225025177002,
+      "learning_rate": 0.0001644814367291014,
+      "loss": 2.6288,
+      "step": 5611
+    },
+    {
+      "epoch": 0.4529093697038173,
+      "grad_norm": 0.6933234333992004,
+      "learning_rate": 0.00016446936938062967,
+      "loss": 2.6076,
+      "step": 5612
+    },
+    {
+      "epoch": 0.45299007344040026,
+      "grad_norm": 0.7101204991340637,
+      "learning_rate": 0.00016445730042541272,
+      "loss": 2.6322,
+      "step": 5613
+    },
+    {
+      "epoch": 0.4530707771769833,
+      "grad_norm": 0.7647581696510315,
+      "learning_rate": 0.00016444522986375134,
+      "loss": 2.7021,
+      "step": 5614
+    },
+    {
+      "epoch": 0.4531514809135663,
+      "grad_norm": 0.7028820514678955,
+      "learning_rate": 0.00016443315769594635,
+      "loss": 2.6171,
+      "step": 5615
+    },
+    {
+      "epoch": 0.4532321846501493,
+      "grad_norm": 0.6933851838111877,
+      "learning_rate": 0.00016442108392229868,
+      "loss": 2.6119,
+      "step": 5616
+    },
+    {
+      "epoch": 0.4533128883867323,
+      "grad_norm": 0.7218462824821472,
+      "learning_rate": 0.0001644090085431092,
+      "loss": 2.6661,
+      "step": 5617
+    },
+    {
+      "epoch": 0.4533935921233153,
+      "grad_norm": 0.7390525341033936,
+      "learning_rate": 0.00016439693155867883,
+      "loss": 2.7084,
+      "step": 5618
+    },
+    {
+      "epoch": 0.4534742958598983,
+      "grad_norm": 0.734136164188385,
+      "learning_rate": 0.0001643848529693086,
+      "loss": 2.6896,
+      "step": 5619
+    },
+    {
+      "epoch": 0.4535549995964813,
+      "grad_norm": 0.8082060813903809,
+      "learning_rate": 0.00016437277277529954,
+      "loss": 2.5828,
+      "step": 5620
+    },
+    {
+      "epoch": 0.4536357033330643,
+      "grad_norm": 0.695988655090332,
+      "learning_rate": 0.0001643606909769527,
+      "loss": 2.6383,
+      "step": 5621
+    },
+    {
+      "epoch": 0.45371640706964733,
+      "grad_norm": 0.7415786385536194,
+      "learning_rate": 0.00016434860757456922,
+      "loss": 2.6388,
+      "step": 5622
+    },
+    {
+      "epoch": 0.4537971108062303,
+      "grad_norm": 0.7378649115562439,
+      "learning_rate": 0.0001643365225684502,
+      "loss": 2.6534,
+      "step": 5623
+    },
+    {
+      "epoch": 0.45387781454281334,
+      "grad_norm": 0.7686129808425903,
+      "learning_rate": 0.0001643244359588969,
+      "loss": 2.6637,
+      "step": 5624
+    },
+    {
+      "epoch": 0.4539585182793963,
+      "grad_norm": 0.7305558323860168,
+      "learning_rate": 0.00016431234774621047,
+      "loss": 2.6525,
+      "step": 5625
+    },
+    {
+      "epoch": 0.45403922201597935,
+      "grad_norm": 0.7994235157966614,
+      "learning_rate": 0.00016430025793069225,
+      "loss": 2.6316,
+      "step": 5626
+    },
+    {
+      "epoch": 0.4541199257525623,
+      "grad_norm": 0.6945801377296448,
+      "learning_rate": 0.0001642881665126435,
+      "loss": 2.6367,
+      "step": 5627
+    },
+    {
+      "epoch": 0.45420062948914536,
+      "grad_norm": 0.6855447292327881,
+      "learning_rate": 0.00016427607349236558,
+      "loss": 2.6317,
+      "step": 5628
+    },
+    {
+      "epoch": 0.45428133322572833,
+      "grad_norm": 0.6961888670921326,
+      "learning_rate": 0.00016426397887015992,
+      "loss": 2.6477,
+      "step": 5629
+    },
+    {
+      "epoch": 0.45436203696231137,
+      "grad_norm": 0.7531994581222534,
+      "learning_rate": 0.0001642518826463279,
+      "loss": 2.7219,
+      "step": 5630
+    },
+    {
+      "epoch": 0.45444274069889434,
+      "grad_norm": 0.7442335486412048,
+      "learning_rate": 0.00016423978482117102,
+      "loss": 2.706,
+      "step": 5631
+    },
+    {
+      "epoch": 0.4545234444354774,
+      "grad_norm": 0.7075700759887695,
+      "learning_rate": 0.00016422768539499076,
+      "loss": 2.6481,
+      "step": 5632
+    },
+    {
+      "epoch": 0.45460414817206035,
+      "grad_norm": 0.7831876873970032,
+      "learning_rate": 0.0001642155843680887,
+      "loss": 2.616,
+      "step": 5633
+    },
+    {
+      "epoch": 0.4546848519086434,
+      "grad_norm": 0.7514604926109314,
+      "learning_rate": 0.00016420348174076642,
+      "loss": 2.6282,
+      "step": 5634
+    },
+    {
+      "epoch": 0.45476555564522636,
+      "grad_norm": 0.7136685252189636,
+      "learning_rate": 0.0001641913775133255,
+      "loss": 2.6764,
+      "step": 5635
+    },
+    {
+      "epoch": 0.4548462593818094,
+      "grad_norm": 0.7406740784645081,
+      "learning_rate": 0.00016417927168606771,
+      "loss": 2.6126,
+      "step": 5636
+    },
+    {
+      "epoch": 0.45492696311839237,
+      "grad_norm": 0.7257869839668274,
+      "learning_rate": 0.0001641671642592947,
+      "loss": 2.6035,
+      "step": 5637
+    },
+    {
+      "epoch": 0.4550076668549754,
+      "grad_norm": 0.8378798961639404,
+      "learning_rate": 0.00016415505523330822,
+      "loss": 2.6657,
+      "step": 5638
+    },
+    {
+      "epoch": 0.4550883705915584,
+      "grad_norm": 0.7218836545944214,
+      "learning_rate": 0.00016414294460841003,
+      "loss": 2.6209,
+      "step": 5639
+    },
+    {
+      "epoch": 0.4551690743281414,
+      "grad_norm": 0.7792766690254211,
+      "learning_rate": 0.00016413083238490204,
+      "loss": 2.7208,
+      "step": 5640
+    },
+    {
+      "epoch": 0.4552497780647244,
+      "grad_norm": 0.7800823450088501,
+      "learning_rate": 0.000164118718563086,
+      "loss": 2.6351,
+      "step": 5641
+    },
+    {
+      "epoch": 0.4553304818013074,
+      "grad_norm": 0.7593275904655457,
+      "learning_rate": 0.00016410660314326395,
+      "loss": 2.7025,
+      "step": 5642
+    },
+    {
+      "epoch": 0.4554111855378904,
+      "grad_norm": 0.7561587691307068,
+      "learning_rate": 0.00016409448612573772,
+      "loss": 2.6188,
+      "step": 5643
+    },
+    {
+      "epoch": 0.4554918892744734,
+      "grad_norm": 0.7674516439437866,
+      "learning_rate": 0.00016408236751080937,
+      "loss": 2.629,
+      "step": 5644
+    },
+    {
+      "epoch": 0.4555725930110564,
+      "grad_norm": 0.7112495303153992,
+      "learning_rate": 0.00016407024729878095,
+      "loss": 2.6261,
+      "step": 5645
+    },
+    {
+      "epoch": 0.45565329674763944,
+      "grad_norm": 0.6861695647239685,
+      "learning_rate": 0.00016405812548995444,
+      "loss": 2.6984,
+      "step": 5646
+    },
+    {
+      "epoch": 0.4557340004842224,
+      "grad_norm": 0.7711648941040039,
+      "learning_rate": 0.000164046002084632,
+      "loss": 2.6839,
+      "step": 5647
+    },
+    {
+      "epoch": 0.45581470422080544,
+      "grad_norm": 0.6862967014312744,
+      "learning_rate": 0.00016403387708311578,
+      "loss": 2.5964,
+      "step": 5648
+    },
+    {
+      "epoch": 0.4558954079573884,
+      "grad_norm": 0.707374632358551,
+      "learning_rate": 0.00016402175048570793,
+      "loss": 2.6191,
+      "step": 5649
+    },
+    {
+      "epoch": 0.45597611169397145,
+      "grad_norm": 0.7980892658233643,
+      "learning_rate": 0.00016400962229271072,
+      "loss": 2.6288,
+      "step": 5650
+    },
+    {
+      "epoch": 0.45605681543055443,
+      "grad_norm": 0.686187744140625,
+      "learning_rate": 0.0001639974925044264,
+      "loss": 2.6277,
+      "step": 5651
+    },
+    {
+      "epoch": 0.45613751916713746,
+      "grad_norm": 0.6970425844192505,
+      "learning_rate": 0.0001639853611211573,
+      "loss": 2.5726,
+      "step": 5652
+    },
+    {
+      "epoch": 0.45621822290372044,
+      "grad_norm": 0.701500415802002,
+      "learning_rate": 0.00016397322814320573,
+      "loss": 2.6275,
+      "step": 5653
+    },
+    {
+      "epoch": 0.45629892664030347,
+      "grad_norm": 0.8432207107543945,
+      "learning_rate": 0.00016396109357087407,
+      "loss": 2.6185,
+      "step": 5654
+    },
+    {
+      "epoch": 0.45637963037688645,
+      "grad_norm": 0.7049770951271057,
+      "learning_rate": 0.00016394895740446476,
+      "loss": 2.674,
+      "step": 5655
+    },
+    {
+      "epoch": 0.4564603341134695,
+      "grad_norm": 0.7068646550178528,
+      "learning_rate": 0.00016393681964428026,
+      "loss": 2.6072,
+      "step": 5656
+    },
+    {
+      "epoch": 0.45654103785005246,
+      "grad_norm": 0.7698760032653809,
+      "learning_rate": 0.00016392468029062312,
+      "loss": 2.6547,
+      "step": 5657
+    },
+    {
+      "epoch": 0.4566217415866355,
+      "grad_norm": 0.7381031513214111,
+      "learning_rate": 0.00016391253934379583,
+      "loss": 2.6125,
+      "step": 5658
+    },
+    {
+      "epoch": 0.45670244532321846,
+      "grad_norm": 0.7367781400680542,
+      "learning_rate": 0.00016390039680410097,
+      "loss": 2.6763,
+      "step": 5659
+    },
+    {
+      "epoch": 0.4567831490598015,
+      "grad_norm": 0.7416272759437561,
+      "learning_rate": 0.00016388825267184121,
+      "loss": 2.7059,
+      "step": 5660
+    },
+    {
+      "epoch": 0.4568638527963845,
+      "grad_norm": 0.6933416724205017,
+      "learning_rate": 0.0001638761069473192,
+      "loss": 2.6028,
+      "step": 5661
+    },
+    {
+      "epoch": 0.45694455653296745,
+      "grad_norm": 0.7311314940452576,
+      "learning_rate": 0.00016386395963083756,
+      "loss": 2.6266,
+      "step": 5662
+    },
+    {
+      "epoch": 0.4570252602695505,
+      "grad_norm": 0.7172734141349792,
+      "learning_rate": 0.00016385181072269917,
+      "loss": 2.6754,
+      "step": 5663
+    },
+    {
+      "epoch": 0.45710596400613346,
+      "grad_norm": 0.7286428213119507,
+      "learning_rate": 0.00016383966022320671,
+      "loss": 2.6637,
+      "step": 5664
+    },
+    {
+      "epoch": 0.4571866677427165,
+      "grad_norm": 0.7296474575996399,
+      "learning_rate": 0.00016382750813266308,
+      "loss": 2.6655,
+      "step": 5665
+    },
+    {
+      "epoch": 0.45726737147929947,
+      "grad_norm": 0.6929224133491516,
+      "learning_rate": 0.00016381535445137105,
+      "loss": 2.6376,
+      "step": 5666
+    },
+    {
+      "epoch": 0.4573480752158825,
+      "grad_norm": 0.7012765407562256,
+      "learning_rate": 0.0001638031991796336,
+      "loss": 2.6222,
+      "step": 5667
+    },
+    {
+      "epoch": 0.4574287789524655,
+      "grad_norm": 0.7360745668411255,
+      "learning_rate": 0.00016379104231775368,
+      "loss": 2.6304,
+      "step": 5668
+    },
+    {
+      "epoch": 0.4575094826890485,
+      "grad_norm": 0.7276801466941833,
+      "learning_rate": 0.00016377888386603419,
+      "loss": 2.7046,
+      "step": 5669
+    },
+    {
+      "epoch": 0.4575901864256315,
+      "grad_norm": 0.688432514667511,
+      "learning_rate": 0.0001637667238247782,
+      "loss": 2.6598,
+      "step": 5670
+    },
+    {
+      "epoch": 0.4576708901622145,
+      "grad_norm": 0.6874414682388306,
+      "learning_rate": 0.00016375456219428877,
+      "loss": 2.7,
+      "step": 5671
+    },
+    {
+      "epoch": 0.4577515938987975,
+      "grad_norm": 0.711091160774231,
+      "learning_rate": 0.000163742398974869,
+      "loss": 2.6063,
+      "step": 5672
+    },
+    {
+      "epoch": 0.4578322976353805,
+      "grad_norm": 0.7131791710853577,
+      "learning_rate": 0.000163730234166822,
+      "loss": 2.5948,
+      "step": 5673
+    },
+    {
+      "epoch": 0.4579130013719635,
+      "grad_norm": 0.7166630625724792,
+      "learning_rate": 0.000163718067770451,
+      "loss": 2.6488,
+      "step": 5674
+    },
+    {
+      "epoch": 0.45799370510854653,
+      "grad_norm": 0.7285952568054199,
+      "learning_rate": 0.00016370589978605916,
+      "loss": 2.6445,
+      "step": 5675
+    },
+    {
+      "epoch": 0.4580744088451295,
+      "grad_norm": 0.728050172328949,
+      "learning_rate": 0.0001636937302139498,
+      "loss": 2.5425,
+      "step": 5676
+    },
+    {
+      "epoch": 0.45815511258171254,
+      "grad_norm": 0.7196047902107239,
+      "learning_rate": 0.00016368155905442615,
+      "loss": 2.7426,
+      "step": 5677
+    },
+    {
+      "epoch": 0.4582358163182955,
+      "grad_norm": 0.6844602823257446,
+      "learning_rate": 0.0001636693863077916,
+      "loss": 2.6157,
+      "step": 5678
+    },
+    {
+      "epoch": 0.45831652005487855,
+      "grad_norm": 0.7375781536102295,
+      "learning_rate": 0.0001636572119743495,
+      "loss": 2.7069,
+      "step": 5679
+    },
+    {
+      "epoch": 0.4583972237914615,
+      "grad_norm": 0.7667750120162964,
+      "learning_rate": 0.0001636450360544033,
+      "loss": 2.6589,
+      "step": 5680
+    },
+    {
+      "epoch": 0.45847792752804456,
+      "grad_norm": 0.6569861173629761,
+      "learning_rate": 0.00016363285854825642,
+      "loss": 2.6197,
+      "step": 5681
+    },
+    {
+      "epoch": 0.45855863126462754,
+      "grad_norm": 0.7177335023880005,
+      "learning_rate": 0.00016362067945621239,
+      "loss": 2.6104,
+      "step": 5682
+    },
+    {
+      "epoch": 0.45863933500121057,
+      "grad_norm": 0.7260481715202332,
+      "learning_rate": 0.00016360849877857469,
+      "loss": 2.6435,
+      "step": 5683
+    },
+    {
+      "epoch": 0.45872003873779355,
+      "grad_norm": 0.7083989381790161,
+      "learning_rate": 0.00016359631651564693,
+      "loss": 2.6366,
+      "step": 5684
+    },
+    {
+      "epoch": 0.4588007424743766,
+      "grad_norm": 0.6417020559310913,
+      "learning_rate": 0.00016358413266773271,
+      "loss": 2.6311,
+      "step": 5685
+    },
+    {
+      "epoch": 0.45888144621095955,
+      "grad_norm": 0.737856924533844,
+      "learning_rate": 0.0001635719472351357,
+      "loss": 2.6647,
+      "step": 5686
+    },
+    {
+      "epoch": 0.4589621499475426,
+      "grad_norm": 0.6774190068244934,
+      "learning_rate": 0.0001635597602181596,
+      "loss": 2.6366,
+      "step": 5687
+    },
+    {
+      "epoch": 0.45904285368412556,
+      "grad_norm": 0.6480480432510376,
+      "learning_rate": 0.0001635475716171081,
+      "loss": 2.6501,
+      "step": 5688
+    },
+    {
+      "epoch": 0.4591235574207086,
+      "grad_norm": 0.7886860370635986,
+      "learning_rate": 0.0001635353814322851,
+      "loss": 2.7239,
+      "step": 5689
+    },
+    {
+      "epoch": 0.45920426115729157,
+      "grad_norm": 0.7579021453857422,
+      "learning_rate": 0.0001635231896639942,
+      "loss": 2.6155,
+      "step": 5690
+    },
+    {
+      "epoch": 0.4592849648938746,
+      "grad_norm": 0.6853809356689453,
+      "learning_rate": 0.0001635109963125394,
+      "loss": 2.5933,
+      "step": 5691
+    },
+    {
+      "epoch": 0.4593656686304576,
+      "grad_norm": 0.661342978477478,
+      "learning_rate": 0.00016349880137822456,
+      "loss": 2.6277,
+      "step": 5692
+    },
+    {
+      "epoch": 0.4594463723670406,
+      "grad_norm": 0.6795682311058044,
+      "learning_rate": 0.0001634866048613536,
+      "loss": 2.6221,
+      "step": 5693
+    },
+    {
+      "epoch": 0.4595270761036236,
+      "grad_norm": 0.7375383377075195,
+      "learning_rate": 0.00016347440676223047,
+      "loss": 2.6082,
+      "step": 5694
+    },
+    {
+      "epoch": 0.4596077798402066,
+      "grad_norm": 0.7565153241157532,
+      "learning_rate": 0.0001634622070811592,
+      "loss": 2.6615,
+      "step": 5695
+    },
+    {
+      "epoch": 0.4596884835767896,
+      "grad_norm": 0.6869745254516602,
+      "learning_rate": 0.00016345000581844386,
+      "loss": 2.6172,
+      "step": 5696
+    },
+    {
+      "epoch": 0.45976918731337263,
+      "grad_norm": 0.7192853689193726,
+      "learning_rate": 0.0001634378029743885,
+      "loss": 2.6324,
+      "step": 5697
+    },
+    {
+      "epoch": 0.4598498910499556,
+      "grad_norm": 0.6919218301773071,
+      "learning_rate": 0.00016342559854929726,
+      "loss": 2.5965,
+      "step": 5698
+    },
+    {
+      "epoch": 0.45993059478653864,
+      "grad_norm": 0.6715282797813416,
+      "learning_rate": 0.00016341339254347432,
+      "loss": 2.6225,
+      "step": 5699
+    },
+    {
+      "epoch": 0.4600112985231216,
+      "grad_norm": 0.6768380999565125,
+      "learning_rate": 0.00016340118495722388,
+      "loss": 2.6376,
+      "step": 5700
+    },
+    {
+      "epoch": 0.46009200225970465,
+      "grad_norm": 0.6898325681686401,
+      "learning_rate": 0.00016338897579085018,
+      "loss": 2.667,
+      "step": 5701
+    },
+    {
+      "epoch": 0.4601727059962876,
+      "grad_norm": 0.7171810865402222,
+      "learning_rate": 0.00016337676504465747,
+      "loss": 2.678,
+      "step": 5702
+    },
+    {
+      "epoch": 0.46025340973287066,
+      "grad_norm": 0.7050724029541016,
+      "learning_rate": 0.00016336455271895016,
+      "loss": 2.619,
+      "step": 5703
+    },
+    {
+      "epoch": 0.46033411346945363,
+      "grad_norm": 0.8287240862846375,
+      "learning_rate": 0.00016335233881403248,
+      "loss": 2.71,
+      "step": 5704
+    },
+    {
+      "epoch": 0.46041481720603666,
+      "grad_norm": 0.6880568861961365,
+      "learning_rate": 0.000163340123330209,
+      "loss": 2.6516,
+      "step": 5705
+    },
+    {
+      "epoch": 0.46049552094261964,
+      "grad_norm": 0.7222896218299866,
+      "learning_rate": 0.00016332790626778402,
+      "loss": 2.5899,
+      "step": 5706
+    },
+    {
+      "epoch": 0.4605762246792027,
+      "grad_norm": 0.7707448601722717,
+      "learning_rate": 0.00016331568762706207,
+      "loss": 2.6116,
+      "step": 5707
+    },
+    {
+      "epoch": 0.46065692841578565,
+      "grad_norm": 0.7780653834342957,
+      "learning_rate": 0.0001633034674083477,
+      "loss": 2.6072,
+      "step": 5708
+    },
+    {
+      "epoch": 0.4607376321523687,
+      "grad_norm": 0.7551524639129639,
+      "learning_rate": 0.00016329124561194545,
+      "loss": 2.548,
+      "step": 5709
+    },
+    {
+      "epoch": 0.46081833588895166,
+      "grad_norm": 0.9312284588813782,
+      "learning_rate": 0.0001632790222381599,
+      "loss": 2.6557,
+      "step": 5710
+    },
+    {
+      "epoch": 0.4608990396255347,
+      "grad_norm": 0.7404753565788269,
+      "learning_rate": 0.0001632667972872957,
+      "loss": 2.6889,
+      "step": 5711
+    },
+    {
+      "epoch": 0.46097974336211767,
+      "grad_norm": 0.7423726916313171,
+      "learning_rate": 0.00016325457075965752,
+      "loss": 2.6265,
+      "step": 5712
+    },
+    {
+      "epoch": 0.46106044709870064,
+      "grad_norm": 1.0683187246322632,
+      "learning_rate": 0.0001632423426555501,
+      "loss": 2.6827,
+      "step": 5713
+    },
+    {
+      "epoch": 0.4611411508352837,
+      "grad_norm": 0.7204160094261169,
+      "learning_rate": 0.0001632301129752782,
+      "loss": 2.702,
+      "step": 5714
+    },
+    {
+      "epoch": 0.46122185457186665,
+      "grad_norm": 0.7591153383255005,
+      "learning_rate": 0.0001632178817191466,
+      "loss": 2.6031,
+      "step": 5715
+    },
+    {
+      "epoch": 0.4613025583084497,
+      "grad_norm": 0.8147456645965576,
+      "learning_rate": 0.00016320564888746013,
+      "loss": 2.6117,
+      "step": 5716
+    },
+    {
+      "epoch": 0.46138326204503266,
+      "grad_norm": 0.7880246639251709,
+      "learning_rate": 0.00016319341448052364,
+      "loss": 2.5896,
+      "step": 5717
+    },
+    {
+      "epoch": 0.4614639657816157,
+      "grad_norm": 0.6875137686729431,
+      "learning_rate": 0.00016318117849864206,
+      "loss": 2.6258,
+      "step": 5718
+    },
+    {
+      "epoch": 0.46154466951819867,
+      "grad_norm": 0.7197960615158081,
+      "learning_rate": 0.00016316894094212044,
+      "loss": 2.6656,
+      "step": 5719
+    },
+    {
+      "epoch": 0.4616253732547817,
+      "grad_norm": 0.7049540281295776,
+      "learning_rate": 0.0001631567018112636,
+      "loss": 2.6698,
+      "step": 5720
+    },
+    {
+      "epoch": 0.4617060769913647,
+      "grad_norm": 0.7128825783729553,
+      "learning_rate": 0.00016314446110637668,
+      "loss": 2.6552,
+      "step": 5721
+    },
+    {
+      "epoch": 0.4617867807279477,
+      "grad_norm": 0.7956201434135437,
+      "learning_rate": 0.00016313221882776477,
+      "loss": 2.6747,
+      "step": 5722
+    },
+    {
+      "epoch": 0.4618674844645307,
+      "grad_norm": 0.7598347663879395,
+      "learning_rate": 0.0001631199749757329,
+      "loss": 2.6187,
+      "step": 5723
+    },
+    {
+      "epoch": 0.4619481882011137,
+      "grad_norm": 0.6587582230567932,
+      "learning_rate": 0.00016310772955058627,
+      "loss": 2.596,
+      "step": 5724
+    },
+    {
+      "epoch": 0.4620288919376967,
+      "grad_norm": 0.700136125087738,
+      "learning_rate": 0.00016309548255263003,
+      "loss": 2.6527,
+      "step": 5725
+    },
+    {
+      "epoch": 0.4621095956742797,
+      "grad_norm": 0.7246582508087158,
+      "learning_rate": 0.00016308323398216945,
+      "loss": 2.6577,
+      "step": 5726
+    },
+    {
+      "epoch": 0.4621902994108627,
+      "grad_norm": 0.6951557993888855,
+      "learning_rate": 0.00016307098383950977,
+      "loss": 2.5816,
+      "step": 5727
+    },
+    {
+      "epoch": 0.46227100314744574,
+      "grad_norm": 0.7109191417694092,
+      "learning_rate": 0.0001630587321249563,
+      "loss": 2.6586,
+      "step": 5728
+    },
+    {
+      "epoch": 0.4623517068840287,
+      "grad_norm": 0.7357863783836365,
+      "learning_rate": 0.0001630464788388144,
+      "loss": 2.691,
+      "step": 5729
+    },
+    {
+      "epoch": 0.46243241062061174,
+      "grad_norm": 0.7916350960731506,
+      "learning_rate": 0.00016303422398138945,
+      "loss": 2.6584,
+      "step": 5730
+    },
+    {
+      "epoch": 0.4625131143571947,
+      "grad_norm": 0.6543231010437012,
+      "learning_rate": 0.00016302196755298685,
+      "loss": 2.6482,
+      "step": 5731
+    },
+    {
+      "epoch": 0.46259381809377775,
+      "grad_norm": 0.6978787183761597,
+      "learning_rate": 0.00016300970955391208,
+      "loss": 2.5956,
+      "step": 5732
+    },
+    {
+      "epoch": 0.46267452183036073,
+      "grad_norm": 0.7301886677742004,
+      "learning_rate": 0.00016299744998447065,
+      "loss": 2.6178,
+      "step": 5733
+    },
+    {
+      "epoch": 0.46275522556694376,
+      "grad_norm": 0.7381030321121216,
+      "learning_rate": 0.00016298518884496808,
+      "loss": 2.6712,
+      "step": 5734
+    },
+    {
+      "epoch": 0.46283592930352674,
+      "grad_norm": 0.7769027948379517,
+      "learning_rate": 0.00016297292613570995,
+      "loss": 2.6082,
+      "step": 5735
+    },
+    {
+      "epoch": 0.46291663304010977,
+      "grad_norm": 0.7698354721069336,
+      "learning_rate": 0.0001629606618570019,
+      "loss": 2.6543,
+      "step": 5736
+    },
+    {
+      "epoch": 0.46299733677669275,
+      "grad_norm": 0.7001554369926453,
+      "learning_rate": 0.00016294839600914957,
+      "loss": 2.6174,
+      "step": 5737
+    },
+    {
+      "epoch": 0.4630780405132758,
+      "grad_norm": 0.7589300274848938,
+      "learning_rate": 0.00016293612859245868,
+      "loss": 2.6338,
+      "step": 5738
+    },
+    {
+      "epoch": 0.46315874424985876,
+      "grad_norm": 0.7083945274353027,
+      "learning_rate": 0.00016292385960723493,
+      "loss": 2.6793,
+      "step": 5739
+    },
+    {
+      "epoch": 0.4632394479864418,
+      "grad_norm": 0.739439845085144,
+      "learning_rate": 0.00016291158905378412,
+      "loss": 2.7335,
+      "step": 5740
+    },
+    {
+      "epoch": 0.46332015172302476,
+      "grad_norm": 0.6868166923522949,
+      "learning_rate": 0.00016289931693241205,
+      "loss": 2.6139,
+      "step": 5741
+    },
+    {
+      "epoch": 0.4634008554596078,
+      "grad_norm": 0.7385871410369873,
+      "learning_rate": 0.0001628870432434246,
+      "loss": 2.6783,
+      "step": 5742
+    },
+    {
+      "epoch": 0.4634815591961908,
+      "grad_norm": 0.7227835655212402,
+      "learning_rate": 0.00016287476798712764,
+      "loss": 2.6732,
+      "step": 5743
+    },
+    {
+      "epoch": 0.4635622629327738,
+      "grad_norm": 0.6662411689758301,
+      "learning_rate": 0.00016286249116382709,
+      "loss": 2.6645,
+      "step": 5744
+    },
+    {
+      "epoch": 0.4636429666693568,
+      "grad_norm": 0.8110263347625732,
+      "learning_rate": 0.00016285021277382894,
+      "loss": 2.6448,
+      "step": 5745
+    },
+    {
+      "epoch": 0.4637236704059398,
+      "grad_norm": 0.7419269680976868,
+      "learning_rate": 0.0001628379328174392,
+      "loss": 2.7286,
+      "step": 5746
+    },
+    {
+      "epoch": 0.4638043741425228,
+      "grad_norm": 0.6518125534057617,
+      "learning_rate": 0.0001628256512949639,
+      "loss": 2.6545,
+      "step": 5747
+    },
+    {
+      "epoch": 0.4638850778791058,
+      "grad_norm": 0.6816060543060303,
+      "learning_rate": 0.00016281336820670917,
+      "loss": 2.6167,
+      "step": 5748
+    },
+    {
+      "epoch": 0.4639657816156888,
+      "grad_norm": 0.6537362337112427,
+      "learning_rate": 0.0001628010835529811,
+      "loss": 2.6522,
+      "step": 5749
+    },
+    {
+      "epoch": 0.46404648535227183,
+      "grad_norm": 0.6720992922782898,
+      "learning_rate": 0.00016278879733408585,
+      "loss": 2.6028,
+      "step": 5750
+    },
+    {
+      "epoch": 0.4641271890888548,
+      "grad_norm": 0.6778908371925354,
+      "learning_rate": 0.00016277650955032967,
+      "loss": 2.5591,
+      "step": 5751
+    },
+    {
+      "epoch": 0.46420789282543784,
+      "grad_norm": 0.6908471584320068,
+      "learning_rate": 0.0001627642202020187,
+      "loss": 2.6574,
+      "step": 5752
+    },
+    {
+      "epoch": 0.4642885965620208,
+      "grad_norm": 0.7034298181533813,
+      "learning_rate": 0.00016275192928945936,
+      "loss": 2.657,
+      "step": 5753
+    },
+    {
+      "epoch": 0.46436930029860385,
+      "grad_norm": 0.7245952486991882,
+      "learning_rate": 0.0001627396368129579,
+      "loss": 2.6572,
+      "step": 5754
+    },
+    {
+      "epoch": 0.4644500040351868,
+      "grad_norm": 0.6764482855796814,
+      "learning_rate": 0.0001627273427728207,
+      "loss": 2.6576,
+      "step": 5755
+    },
+    {
+      "epoch": 0.46453070777176986,
+      "grad_norm": 0.7074379920959473,
+      "learning_rate": 0.0001627150471693541,
+      "loss": 2.614,
+      "step": 5756
+    },
+    {
+      "epoch": 0.46461141150835283,
+      "grad_norm": 0.7292052507400513,
+      "learning_rate": 0.0001627027500028646,
+      "loss": 2.673,
+      "step": 5757
+    },
+    {
+      "epoch": 0.46469211524493587,
+      "grad_norm": 0.7554025650024414,
+      "learning_rate": 0.0001626904512736587,
+      "loss": 2.5919,
+      "step": 5758
+    },
+    {
+      "epoch": 0.46477281898151884,
+      "grad_norm": 0.6829606890678406,
+      "learning_rate": 0.00016267815098204284,
+      "loss": 2.7206,
+      "step": 5759
+    },
+    {
+      "epoch": 0.4648535227181019,
+      "grad_norm": 0.7201548218727112,
+      "learning_rate": 0.00016266584912832363,
+      "loss": 2.6651,
+      "step": 5760
+    },
+    {
+      "epoch": 0.46493422645468485,
+      "grad_norm": 0.6889227628707886,
+      "learning_rate": 0.00016265354571280764,
+      "loss": 2.6776,
+      "step": 5761
+    },
+    {
+      "epoch": 0.4650149301912679,
+      "grad_norm": 0.7286190986633301,
+      "learning_rate": 0.00016264124073580156,
+      "loss": 2.591,
+      "step": 5762
+    },
+    {
+      "epoch": 0.46509563392785086,
+      "grad_norm": 0.7222036123275757,
+      "learning_rate": 0.00016262893419761196,
+      "loss": 2.6422,
+      "step": 5763
+    },
+    {
+      "epoch": 0.46517633766443384,
+      "grad_norm": 0.6822768449783325,
+      "learning_rate": 0.00016261662609854562,
+      "loss": 2.6126,
+      "step": 5764
+    },
+    {
+      "epoch": 0.46525704140101687,
+      "grad_norm": 0.7263356447219849,
+      "learning_rate": 0.00016260431643890929,
+      "loss": 2.6304,
+      "step": 5765
+    },
+    {
+      "epoch": 0.46533774513759985,
+      "grad_norm": 0.7152180075645447,
+      "learning_rate": 0.00016259200521900972,
+      "loss": 2.6489,
+      "step": 5766
+    },
+    {
+      "epoch": 0.4654184488741829,
+      "grad_norm": 0.6988116502761841,
+      "learning_rate": 0.00016257969243915378,
+      "loss": 2.6151,
+      "step": 5767
+    },
+    {
+      "epoch": 0.46549915261076585,
+      "grad_norm": 0.7131790518760681,
+      "learning_rate": 0.00016256737809964831,
+      "loss": 2.6284,
+      "step": 5768
+    },
+    {
+      "epoch": 0.4655798563473489,
+      "grad_norm": 0.674196183681488,
+      "learning_rate": 0.00016255506220080025,
+      "loss": 2.5815,
+      "step": 5769
+    },
+    {
+      "epoch": 0.46566056008393186,
+      "grad_norm": 0.7166198492050171,
+      "learning_rate": 0.0001625427447429165,
+      "loss": 2.6594,
+      "step": 5770
+    },
+    {
+      "epoch": 0.4657412638205149,
+      "grad_norm": 0.6997127532958984,
+      "learning_rate": 0.00016253042572630407,
+      "loss": 2.6502,
+      "step": 5771
+    },
+    {
+      "epoch": 0.46582196755709787,
+      "grad_norm": 0.7761591076850891,
+      "learning_rate": 0.00016251810515126994,
+      "loss": 2.624,
+      "step": 5772
+    },
+    {
+      "epoch": 0.4659026712936809,
+      "grad_norm": 0.7038728594779968,
+      "learning_rate": 0.00016250578301812125,
+      "loss": 2.6096,
+      "step": 5773
+    },
+    {
+      "epoch": 0.4659833750302639,
+      "grad_norm": 0.7080080509185791,
+      "learning_rate": 0.00016249345932716505,
+      "loss": 2.6196,
+      "step": 5774
+    },
+    {
+      "epoch": 0.4660640787668469,
+      "grad_norm": 0.7461444735527039,
+      "learning_rate": 0.00016248113407870847,
+      "loss": 2.65,
+      "step": 5775
+    },
+    {
+      "epoch": 0.4661447825034299,
+      "grad_norm": 0.7914463877677917,
+      "learning_rate": 0.00016246880727305868,
+      "loss": 2.6539,
+      "step": 5776
+    },
+    {
+      "epoch": 0.4662254862400129,
+      "grad_norm": 0.7067776918411255,
+      "learning_rate": 0.00016245647891052295,
+      "loss": 2.72,
+      "step": 5777
+    },
+    {
+      "epoch": 0.4663061899765959,
+      "grad_norm": 0.7190818190574646,
+      "learning_rate": 0.00016244414899140852,
+      "loss": 2.7029,
+      "step": 5778
+    },
+    {
+      "epoch": 0.46638689371317893,
+      "grad_norm": 0.6740003824234009,
+      "learning_rate": 0.00016243181751602261,
+      "loss": 2.6404,
+      "step": 5779
+    },
+    {
+      "epoch": 0.4664675974497619,
+      "grad_norm": 0.7942661643028259,
+      "learning_rate": 0.00016241948448467267,
+      "loss": 2.6333,
+      "step": 5780
+    },
+    {
+      "epoch": 0.46654830118634494,
+      "grad_norm": 0.6415690183639526,
+      "learning_rate": 0.00016240714989766597,
+      "loss": 2.6354,
+      "step": 5781
+    },
+    {
+      "epoch": 0.4666290049229279,
+      "grad_norm": 0.7287769913673401,
+      "learning_rate": 0.00016239481375530997,
+      "loss": 2.6721,
+      "step": 5782
+    },
+    {
+      "epoch": 0.46670970865951095,
+      "grad_norm": 0.8197699189186096,
+      "learning_rate": 0.00016238247605791212,
+      "loss": 2.7577,
+      "step": 5783
+    },
+    {
+      "epoch": 0.4667904123960939,
+      "grad_norm": 0.8182012438774109,
+      "learning_rate": 0.0001623701368057799,
+      "loss": 2.6475,
+      "step": 5784
+    },
+    {
+      "epoch": 0.46687111613267696,
+      "grad_norm": 0.6974665522575378,
+      "learning_rate": 0.00016235779599922082,
+      "loss": 2.5897,
+      "step": 5785
+    },
+    {
+      "epoch": 0.46695181986925993,
+      "grad_norm": 0.7156379222869873,
+      "learning_rate": 0.00016234545363854247,
+      "loss": 2.5981,
+      "step": 5786
+    },
+    {
+      "epoch": 0.46703252360584296,
+      "grad_norm": 0.6875364780426025,
+      "learning_rate": 0.0001623331097240524,
+      "loss": 2.6333,
+      "step": 5787
+    },
+    {
+      "epoch": 0.46711322734242594,
+      "grad_norm": 0.7222917675971985,
+      "learning_rate": 0.00016232076425605835,
+      "loss": 2.5865,
+      "step": 5788
+    },
+    {
+      "epoch": 0.467193931079009,
+      "grad_norm": 0.7224915027618408,
+      "learning_rate": 0.00016230841723486792,
+      "loss": 2.667,
+      "step": 5789
+    },
+    {
+      "epoch": 0.46727463481559195,
+      "grad_norm": 0.7125402688980103,
+      "learning_rate": 0.00016229606866078887,
+      "loss": 2.6548,
+      "step": 5790
+    },
+    {
+      "epoch": 0.467355338552175,
+      "grad_norm": 0.6866132616996765,
+      "learning_rate": 0.00016228371853412894,
+      "loss": 2.6381,
+      "step": 5791
+    },
+    {
+      "epoch": 0.46743604228875796,
+      "grad_norm": 0.7573552131652832,
+      "learning_rate": 0.00016227136685519593,
+      "loss": 2.6766,
+      "step": 5792
+    },
+    {
+      "epoch": 0.467516746025341,
+      "grad_norm": 0.7565932273864746,
+      "learning_rate": 0.00016225901362429767,
+      "loss": 2.5965,
+      "step": 5793
+    },
+    {
+      "epoch": 0.46759744976192397,
+      "grad_norm": 0.7279250621795654,
+      "learning_rate": 0.00016224665884174207,
+      "loss": 2.6599,
+      "step": 5794
+    },
+    {
+      "epoch": 0.467678153498507,
+      "grad_norm": 0.7501276731491089,
+      "learning_rate": 0.000162234302507837,
+      "loss": 2.636,
+      "step": 5795
+    },
+    {
+      "epoch": 0.46775885723509,
+      "grad_norm": 0.7823930978775024,
+      "learning_rate": 0.00016222194462289042,
+      "loss": 2.6277,
+      "step": 5796
+    },
+    {
+      "epoch": 0.467839560971673,
+      "grad_norm": 0.7168415784835815,
+      "learning_rate": 0.00016220958518721034,
+      "loss": 2.6868,
+      "step": 5797
+    },
+    {
+      "epoch": 0.467920264708256,
+      "grad_norm": 0.7468454241752625,
+      "learning_rate": 0.00016219722420110478,
+      "loss": 2.7209,
+      "step": 5798
+    },
+    {
+      "epoch": 0.468000968444839,
+      "grad_norm": 0.6915228962898254,
+      "learning_rate": 0.0001621848616648818,
+      "loss": 2.6356,
+      "step": 5799
+    },
+    {
+      "epoch": 0.468081672181422,
+      "grad_norm": 0.7731573581695557,
+      "learning_rate": 0.00016217249757884955,
+      "loss": 2.6396,
+      "step": 5800
+    },
+    {
+      "epoch": 0.468162375918005,
+      "grad_norm": 0.6579388380050659,
+      "learning_rate": 0.0001621601319433161,
+      "loss": 2.6077,
+      "step": 5801
+    },
+    {
+      "epoch": 0.468243079654588,
+      "grad_norm": 0.7136246562004089,
+      "learning_rate": 0.00016214776475858967,
+      "loss": 2.6602,
+      "step": 5802
+    },
+    {
+      "epoch": 0.46832378339117103,
+      "grad_norm": 0.6929461359977722,
+      "learning_rate": 0.0001621353960249785,
+      "loss": 2.6851,
+      "step": 5803
+    },
+    {
+      "epoch": 0.468404487127754,
+      "grad_norm": 0.8001779913902283,
+      "learning_rate": 0.00016212302574279087,
+      "loss": 2.6577,
+      "step": 5804
+    },
+    {
+      "epoch": 0.46848519086433704,
+      "grad_norm": 0.7637671828269958,
+      "learning_rate": 0.00016211065391233498,
+      "loss": 2.6923,
+      "step": 5805
+    },
+    {
+      "epoch": 0.46856589460092,
+      "grad_norm": 0.6879906058311462,
+      "learning_rate": 0.0001620982805339193,
+      "loss": 2.6555,
+      "step": 5806
+    },
+    {
+      "epoch": 0.46864659833750305,
+      "grad_norm": 0.7731223702430725,
+      "learning_rate": 0.0001620859056078521,
+      "loss": 2.6301,
+      "step": 5807
+    },
+    {
+      "epoch": 0.468727302074086,
+      "grad_norm": 0.7351491451263428,
+      "learning_rate": 0.00016207352913444185,
+      "loss": 2.6154,
+      "step": 5808
+    },
+    {
+      "epoch": 0.46880800581066906,
+      "grad_norm": 0.716314435005188,
+      "learning_rate": 0.000162061151113997,
+      "loss": 2.6294,
+      "step": 5809
+    },
+    {
+      "epoch": 0.46888870954725204,
+      "grad_norm": 0.6974702477455139,
+      "learning_rate": 0.00016204877154682605,
+      "loss": 2.6046,
+      "step": 5810
+    },
+    {
+      "epoch": 0.46896941328383507,
+      "grad_norm": 0.7456035614013672,
+      "learning_rate": 0.00016203639043323745,
+      "loss": 2.6308,
+      "step": 5811
+    },
+    {
+      "epoch": 0.46905011702041804,
+      "grad_norm": 0.7198047637939453,
+      "learning_rate": 0.0001620240077735399,
+      "loss": 2.6303,
+      "step": 5812
+    },
+    {
+      "epoch": 0.4691308207570011,
+      "grad_norm": 0.7098269462585449,
+      "learning_rate": 0.00016201162356804192,
+      "loss": 2.6352,
+      "step": 5813
+    },
+    {
+      "epoch": 0.46921152449358405,
+      "grad_norm": 0.7060410976409912,
+      "learning_rate": 0.0001619992378170522,
+      "loss": 2.6489,
+      "step": 5814
+    },
+    {
+      "epoch": 0.46929222823016703,
+      "grad_norm": 0.7126092314720154,
+      "learning_rate": 0.0001619868505208794,
+      "loss": 2.66,
+      "step": 5815
+    },
+    {
+      "epoch": 0.46937293196675006,
+      "grad_norm": 0.7391123175621033,
+      "learning_rate": 0.00016197446167983223,
+      "loss": 2.6066,
+      "step": 5816
+    },
+    {
+      "epoch": 0.46945363570333304,
+      "grad_norm": 0.7282211780548096,
+      "learning_rate": 0.0001619620712942195,
+      "loss": 2.6422,
+      "step": 5817
+    },
+    {
+      "epoch": 0.46953433943991607,
+      "grad_norm": 0.7581801414489746,
+      "learning_rate": 0.00016194967936434998,
+      "loss": 2.702,
+      "step": 5818
+    },
+    {
+      "epoch": 0.46961504317649905,
+      "grad_norm": 0.6649011373519897,
+      "learning_rate": 0.00016193728589053248,
+      "loss": 2.6235,
+      "step": 5819
+    },
+    {
+      "epoch": 0.4696957469130821,
+      "grad_norm": 0.720312237739563,
+      "learning_rate": 0.00016192489087307592,
+      "loss": 2.5961,
+      "step": 5820
+    },
+    {
+      "epoch": 0.46977645064966506,
+      "grad_norm": 0.72076016664505,
+      "learning_rate": 0.0001619124943122892,
+      "loss": 2.6793,
+      "step": 5821
+    },
+    {
+      "epoch": 0.4698571543862481,
+      "grad_norm": 0.6695740818977356,
+      "learning_rate": 0.0001619000962084813,
+      "loss": 2.6325,
+      "step": 5822
+    },
+    {
+      "epoch": 0.46993785812283106,
+      "grad_norm": 0.7678804993629456,
+      "learning_rate": 0.0001618876965619612,
+      "loss": 2.7473,
+      "step": 5823
+    },
+    {
+      "epoch": 0.4700185618594141,
+      "grad_norm": 0.782349169254303,
+      "learning_rate": 0.00016187529537303792,
+      "loss": 2.6139,
+      "step": 5824
+    },
+    {
+      "epoch": 0.4700992655959971,
+      "grad_norm": 0.6906631588935852,
+      "learning_rate": 0.00016186289264202052,
+      "loss": 2.6529,
+      "step": 5825
+    },
+    {
+      "epoch": 0.4701799693325801,
+      "grad_norm": 0.732947051525116,
+      "learning_rate": 0.00016185048836921814,
+      "loss": 2.6416,
+      "step": 5826
+    },
+    {
+      "epoch": 0.4702606730691631,
+      "grad_norm": 0.8306718468666077,
+      "learning_rate": 0.0001618380825549399,
+      "loss": 2.6566,
+      "step": 5827
+    },
+    {
+      "epoch": 0.4703413768057461,
+      "grad_norm": 0.725764811038971,
+      "learning_rate": 0.00016182567519949502,
+      "loss": 2.6664,
+      "step": 5828
+    },
+    {
+      "epoch": 0.4704220805423291,
+      "grad_norm": 0.7301872372627258,
+      "learning_rate": 0.00016181326630319268,
+      "loss": 2.6666,
+      "step": 5829
+    },
+    {
+      "epoch": 0.4705027842789121,
+      "grad_norm": 0.7297122478485107,
+      "learning_rate": 0.00016180085586634216,
+      "loss": 2.6415,
+      "step": 5830
+    },
+    {
+      "epoch": 0.4705834880154951,
+      "grad_norm": 0.7445664405822754,
+      "learning_rate": 0.00016178844388925278,
+      "loss": 2.6112,
+      "step": 5831
+    },
+    {
+      "epoch": 0.47066419175207813,
+      "grad_norm": 0.7787267565727234,
+      "learning_rate": 0.00016177603037223384,
+      "loss": 2.6452,
+      "step": 5832
+    },
+    {
+      "epoch": 0.4707448954886611,
+      "grad_norm": 0.7386903762817383,
+      "learning_rate": 0.00016176361531559474,
+      "loss": 2.6919,
+      "step": 5833
+    },
+    {
+      "epoch": 0.47082559922524414,
+      "grad_norm": 0.7991776466369629,
+      "learning_rate": 0.0001617511987196449,
+      "loss": 2.6728,
+      "step": 5834
+    },
+    {
+      "epoch": 0.4709063029618271,
+      "grad_norm": 0.7196263670921326,
+      "learning_rate": 0.00016173878058469375,
+      "loss": 2.6008,
+      "step": 5835
+    },
+    {
+      "epoch": 0.47098700669841015,
+      "grad_norm": 0.6773477792739868,
+      "learning_rate": 0.00016172636091105086,
+      "loss": 2.6184,
+      "step": 5836
+    },
+    {
+      "epoch": 0.4710677104349931,
+      "grad_norm": 0.7238345742225647,
+      "learning_rate": 0.00016171393969902567,
+      "loss": 2.6221,
+      "step": 5837
+    },
+    {
+      "epoch": 0.47114841417157616,
+      "grad_norm": 0.702104926109314,
+      "learning_rate": 0.00016170151694892777,
+      "loss": 2.5909,
+      "step": 5838
+    },
+    {
+      "epoch": 0.47122911790815913,
+      "grad_norm": 0.7571590542793274,
+      "learning_rate": 0.00016168909266106677,
+      "loss": 2.6044,
+      "step": 5839
+    },
+    {
+      "epoch": 0.47130982164474217,
+      "grad_norm": 0.7408227324485779,
+      "learning_rate": 0.00016167666683575234,
+      "loss": 2.5771,
+      "step": 5840
+    },
+    {
+      "epoch": 0.47139052538132514,
+      "grad_norm": 0.6760764122009277,
+      "learning_rate": 0.00016166423947329414,
+      "loss": 2.6202,
+      "step": 5841
+    },
+    {
+      "epoch": 0.4714712291179082,
+      "grad_norm": 0.7085632681846619,
+      "learning_rate": 0.00016165181057400192,
+      "loss": 2.5887,
+      "step": 5842
+    },
+    {
+      "epoch": 0.47155193285449115,
+      "grad_norm": 0.7298943400382996,
+      "learning_rate": 0.00016163938013818538,
+      "loss": 2.609,
+      "step": 5843
+    },
+    {
+      "epoch": 0.4716326365910742,
+      "grad_norm": 0.7591157555580139,
+      "learning_rate": 0.0001616269481661544,
+      "loss": 2.6582,
+      "step": 5844
+    },
+    {
+      "epoch": 0.47171334032765716,
+      "grad_norm": 0.6727088093757629,
+      "learning_rate": 0.00016161451465821877,
+      "loss": 2.6289,
+      "step": 5845
+    },
+    {
+      "epoch": 0.4717940440642402,
+      "grad_norm": 0.6782706379890442,
+      "learning_rate": 0.00016160207961468835,
+      "loss": 2.6875,
+      "step": 5846
+    },
+    {
+      "epoch": 0.47187474780082317,
+      "grad_norm": 0.6839444041252136,
+      "learning_rate": 0.00016158964303587313,
+      "loss": 2.5687,
+      "step": 5847
+    },
+    {
+      "epoch": 0.4719554515374062,
+      "grad_norm": 0.7565997838973999,
+      "learning_rate": 0.00016157720492208295,
+      "loss": 2.6855,
+      "step": 5848
+    },
+    {
+      "epoch": 0.4720361552739892,
+      "grad_norm": 0.7286611199378967,
+      "learning_rate": 0.0001615647652736279,
+      "loss": 2.5906,
+      "step": 5849
+    },
+    {
+      "epoch": 0.4721168590105722,
+      "grad_norm": 0.7503396272659302,
+      "learning_rate": 0.00016155232409081793,
+      "loss": 2.6419,
+      "step": 5850
+    },
+    {
+      "epoch": 0.4721975627471552,
+      "grad_norm": 0.6924198865890503,
+      "learning_rate": 0.00016153988137396317,
+      "loss": 2.661,
+      "step": 5851
+    },
+    {
+      "epoch": 0.4722782664837382,
+      "grad_norm": 0.7731672525405884,
+      "learning_rate": 0.0001615274371233737,
+      "loss": 2.6993,
+      "step": 5852
+    },
+    {
+      "epoch": 0.4723589702203212,
+      "grad_norm": 0.7422799468040466,
+      "learning_rate": 0.00016151499133935964,
+      "loss": 2.6134,
+      "step": 5853
+    },
+    {
+      "epoch": 0.4724396739569042,
+      "grad_norm": 0.6924546957015991,
+      "learning_rate": 0.0001615025440222312,
+      "loss": 2.672,
+      "step": 5854
+    },
+    {
+      "epoch": 0.4725203776934872,
+      "grad_norm": 0.7205976843833923,
+      "learning_rate": 0.00016149009517229862,
+      "loss": 2.6722,
+      "step": 5855
+    },
+    {
+      "epoch": 0.47260108143007024,
+      "grad_norm": 0.6898519992828369,
+      "learning_rate": 0.0001614776447898721,
+      "loss": 2.6474,
+      "step": 5856
+    },
+    {
+      "epoch": 0.4726817851666532,
+      "grad_norm": 0.7512481212615967,
+      "learning_rate": 0.00016146519287526197,
+      "loss": 2.7413,
+      "step": 5857
+    },
+    {
+      "epoch": 0.47276248890323624,
+      "grad_norm": 0.6734220385551453,
+      "learning_rate": 0.0001614527394287786,
+      "loss": 2.6114,
+      "step": 5858
+    },
+    {
+      "epoch": 0.4728431926398192,
+      "grad_norm": 0.6745339632034302,
+      "learning_rate": 0.00016144028445073228,
+      "loss": 2.6039,
+      "step": 5859
+    },
+    {
+      "epoch": 0.47292389637640225,
+      "grad_norm": 0.7463086843490601,
+      "learning_rate": 0.0001614278279414335,
+      "loss": 2.6109,
+      "step": 5860
+    },
+    {
+      "epoch": 0.47300460011298523,
+      "grad_norm": 0.7203261256217957,
+      "learning_rate": 0.00016141536990119264,
+      "loss": 2.651,
+      "step": 5861
+    },
+    {
+      "epoch": 0.47308530384956826,
+      "grad_norm": 0.7718746066093445,
+      "learning_rate": 0.00016140291033032024,
+      "loss": 2.6953,
+      "step": 5862
+    },
+    {
+      "epoch": 0.47316600758615124,
+      "grad_norm": 0.7854858040809631,
+      "learning_rate": 0.0001613904492291268,
+      "loss": 2.5941,
+      "step": 5863
+    },
+    {
+      "epoch": 0.47324671132273427,
+      "grad_norm": 0.7218664288520813,
+      "learning_rate": 0.0001613779865979229,
+      "loss": 2.6447,
+      "step": 5864
+    },
+    {
+      "epoch": 0.47332741505931725,
+      "grad_norm": 0.7479045987129211,
+      "learning_rate": 0.0001613655224370191,
+      "loss": 2.6662,
+      "step": 5865
+    },
+    {
+      "epoch": 0.4734081187959002,
+      "grad_norm": 0.7335021495819092,
+      "learning_rate": 0.00016135305674672612,
+      "loss": 2.6283,
+      "step": 5866
+    },
+    {
+      "epoch": 0.47348882253248326,
+      "grad_norm": 0.7650331258773804,
+      "learning_rate": 0.00016134058952735453,
+      "loss": 2.7168,
+      "step": 5867
+    },
+    {
+      "epoch": 0.47356952626906623,
+      "grad_norm": 0.733383297920227,
+      "learning_rate": 0.00016132812077921513,
+      "loss": 2.6352,
+      "step": 5868
+    },
+    {
+      "epoch": 0.47365023000564926,
+      "grad_norm": 1.3944146633148193,
+      "learning_rate": 0.00016131565050261866,
+      "loss": 2.7518,
+      "step": 5869
+    },
+    {
+      "epoch": 0.47373093374223224,
+      "grad_norm": 0.746112585067749,
+      "learning_rate": 0.0001613031786978759,
+      "loss": 2.6253,
+      "step": 5870
+    },
+    {
+      "epoch": 0.4738116374788153,
+      "grad_norm": 0.9859737753868103,
+      "learning_rate": 0.00016129070536529766,
+      "loss": 2.6682,
+      "step": 5871
+    },
+    {
+      "epoch": 0.47389234121539825,
+      "grad_norm": 0.7358877062797546,
+      "learning_rate": 0.00016127823050519484,
+      "loss": 2.6712,
+      "step": 5872
+    },
+    {
+      "epoch": 0.4739730449519813,
+      "grad_norm": 0.7379923462867737,
+      "learning_rate": 0.0001612657541178783,
+      "loss": 2.6268,
+      "step": 5873
+    },
+    {
+      "epoch": 0.47405374868856426,
+      "grad_norm": 0.7671005725860596,
+      "learning_rate": 0.00016125327620365907,
+      "loss": 2.6127,
+      "step": 5874
+    },
+    {
+      "epoch": 0.4741344524251473,
+      "grad_norm": 0.8007156252861023,
+      "learning_rate": 0.00016124079676284805,
+      "loss": 2.6173,
+      "step": 5875
+    },
+    {
+      "epoch": 0.47421515616173027,
+      "grad_norm": 0.7930500507354736,
+      "learning_rate": 0.00016122831579575627,
+      "loss": 2.589,
+      "step": 5876
+    },
+    {
+      "epoch": 0.4742958598983133,
+      "grad_norm": 0.788006603717804,
+      "learning_rate": 0.00016121583330269484,
+      "loss": 2.6731,
+      "step": 5877
+    },
+    {
+      "epoch": 0.4743765636348963,
+      "grad_norm": 0.742148220539093,
+      "learning_rate": 0.00016120334928397483,
+      "loss": 2.674,
+      "step": 5878
+    },
+    {
+      "epoch": 0.4744572673714793,
+      "grad_norm": 0.6823038458824158,
+      "learning_rate": 0.00016119086373990736,
+      "loss": 2.6153,
+      "step": 5879
+    },
+    {
+      "epoch": 0.4745379711080623,
+      "grad_norm": 0.7542331218719482,
+      "learning_rate": 0.00016117837667080356,
+      "loss": 2.6739,
+      "step": 5880
+    },
+    {
+      "epoch": 0.4746186748446453,
+      "grad_norm": 0.8163543343544006,
+      "learning_rate": 0.00016116588807697476,
+      "loss": 2.6558,
+      "step": 5881
+    },
+    {
+      "epoch": 0.4746993785812283,
+      "grad_norm": 0.7528213858604431,
+      "learning_rate": 0.0001611533979587321,
+      "loss": 2.6243,
+      "step": 5882
+    },
+    {
+      "epoch": 0.4747800823178113,
+      "grad_norm": 0.7476626038551331,
+      "learning_rate": 0.00016114090631638695,
+      "loss": 2.5984,
+      "step": 5883
+    },
+    {
+      "epoch": 0.4748607860543943,
+      "grad_norm": 0.7436621785163879,
+      "learning_rate": 0.00016112841315025055,
+      "loss": 2.6118,
+      "step": 5884
+    },
+    {
+      "epoch": 0.47494148979097733,
+      "grad_norm": 0.8024004101753235,
+      "learning_rate": 0.0001611159184606343,
+      "loss": 2.6926,
+      "step": 5885
+    },
+    {
+      "epoch": 0.4750221935275603,
+      "grad_norm": 0.7475626468658447,
+      "learning_rate": 0.00016110342224784962,
+      "loss": 2.6175,
+      "step": 5886
+    },
+    {
+      "epoch": 0.47510289726414334,
+      "grad_norm": 0.7900637984275818,
+      "learning_rate": 0.00016109092451220796,
+      "loss": 2.6503,
+      "step": 5887
+    },
+    {
+      "epoch": 0.4751836010007263,
+      "grad_norm": 0.6988356113433838,
+      "learning_rate": 0.00016107842525402074,
+      "loss": 2.6494,
+      "step": 5888
+    },
+    {
+      "epoch": 0.47526430473730935,
+      "grad_norm": 1.0214186906814575,
+      "learning_rate": 0.00016106592447359948,
+      "loss": 2.6476,
+      "step": 5889
+    },
+    {
+      "epoch": 0.4753450084738923,
+      "grad_norm": 0.741527795791626,
+      "learning_rate": 0.00016105342217125578,
+      "loss": 2.6054,
+      "step": 5890
+    },
+    {
+      "epoch": 0.47542571221047536,
+      "grad_norm": 0.7196603417396545,
+      "learning_rate": 0.0001610409183473012,
+      "loss": 2.6146,
+      "step": 5891
+    },
+    {
+      "epoch": 0.47550641594705834,
+      "grad_norm": 0.8130923509597778,
+      "learning_rate": 0.00016102841300204737,
+      "loss": 2.6505,
+      "step": 5892
+    },
+    {
+      "epoch": 0.47558711968364137,
+      "grad_norm": 0.7929537892341614,
+      "learning_rate": 0.00016101590613580596,
+      "loss": 2.6725,
+      "step": 5893
+    },
+    {
+      "epoch": 0.47566782342022434,
+      "grad_norm": 0.7149303555488586,
+      "learning_rate": 0.00016100339774888865,
+      "loss": 2.6272,
+      "step": 5894
+    },
+    {
+      "epoch": 0.4757485271568074,
+      "grad_norm": 0.7242792248725891,
+      "learning_rate": 0.00016099088784160724,
+      "loss": 2.5948,
+      "step": 5895
+    },
+    {
+      "epoch": 0.47582923089339035,
+      "grad_norm": 0.7571540474891663,
+      "learning_rate": 0.00016097837641427346,
+      "loss": 2.689,
+      "step": 5896
+    },
+    {
+      "epoch": 0.4759099346299734,
+      "grad_norm": 0.7402021288871765,
+      "learning_rate": 0.00016096586346719916,
+      "loss": 2.7035,
+      "step": 5897
+    },
+    {
+      "epoch": 0.47599063836655636,
+      "grad_norm": 0.7195574045181274,
+      "learning_rate": 0.00016095334900069613,
+      "loss": 2.5862,
+      "step": 5898
+    },
+    {
+      "epoch": 0.4760713421031394,
+      "grad_norm": 0.7677412033081055,
+      "learning_rate": 0.00016094083301507634,
+      "loss": 2.6715,
+      "step": 5899
+    },
+    {
+      "epoch": 0.47615204583972237,
+      "grad_norm": 0.7131708860397339,
+      "learning_rate": 0.0001609283155106517,
+      "loss": 2.6555,
+      "step": 5900
+    },
+    {
+      "epoch": 0.4762327495763054,
+      "grad_norm": 0.6774055361747742,
+      "learning_rate": 0.00016091579648773414,
+      "loss": 2.621,
+      "step": 5901
+    },
+    {
+      "epoch": 0.4763134533128884,
+      "grad_norm": 0.6873257160186768,
+      "learning_rate": 0.00016090327594663571,
+      "loss": 2.6719,
+      "step": 5902
+    },
+    {
+      "epoch": 0.4763941570494714,
+      "grad_norm": 0.8004229068756104,
+      "learning_rate": 0.00016089075388766845,
+      "loss": 2.6926,
+      "step": 5903
+    },
+    {
+      "epoch": 0.4764748607860544,
+      "grad_norm": 0.7196173667907715,
+      "learning_rate": 0.00016087823031114438,
+      "loss": 2.6032,
+      "step": 5904
+    },
+    {
+      "epoch": 0.4765555645226374,
+      "grad_norm": 0.7665518522262573,
+      "learning_rate": 0.00016086570521737573,
+      "loss": 2.6359,
+      "step": 5905
+    },
+    {
+      "epoch": 0.4766362682592204,
+      "grad_norm": 0.7240240573883057,
+      "learning_rate": 0.0001608531786066746,
+      "loss": 2.6489,
+      "step": 5906
+    },
+    {
+      "epoch": 0.47671697199580343,
+      "grad_norm": 0.7603839039802551,
+      "learning_rate": 0.00016084065047935317,
+      "loss": 2.6064,
+      "step": 5907
+    },
+    {
+      "epoch": 0.4767976757323864,
+      "grad_norm": 0.7394058704376221,
+      "learning_rate": 0.0001608281208357237,
+      "loss": 2.6643,
+      "step": 5908
+    },
+    {
+      "epoch": 0.47687837946896944,
+      "grad_norm": 0.7183148860931396,
+      "learning_rate": 0.00016081558967609845,
+      "loss": 2.56,
+      "step": 5909
+    },
+    {
+      "epoch": 0.4769590832055524,
+      "grad_norm": 0.7181926965713501,
+      "learning_rate": 0.00016080305700078972,
+      "loss": 2.6665,
+      "step": 5910
+    },
+    {
+      "epoch": 0.47703978694213545,
+      "grad_norm": 0.7634081840515137,
+      "learning_rate": 0.00016079052281010988,
+      "loss": 2.7076,
+      "step": 5911
+    },
+    {
+      "epoch": 0.4771204906787184,
+      "grad_norm": 0.7928739190101624,
+      "learning_rate": 0.0001607779871043713,
+      "loss": 2.6512,
+      "step": 5912
+    },
+    {
+      "epoch": 0.47720119441530146,
+      "grad_norm": 0.7192893028259277,
+      "learning_rate": 0.00016076544988388643,
+      "loss": 2.6453,
+      "step": 5913
+    },
+    {
+      "epoch": 0.47728189815188443,
+      "grad_norm": 0.7171720862388611,
+      "learning_rate": 0.00016075291114896767,
+      "loss": 2.6501,
+      "step": 5914
+    },
+    {
+      "epoch": 0.47736260188846746,
+      "grad_norm": 0.6787160038948059,
+      "learning_rate": 0.00016074037089992756,
+      "loss": 2.6566,
+      "step": 5915
+    },
+    {
+      "epoch": 0.47744330562505044,
+      "grad_norm": 0.8118634819984436,
+      "learning_rate": 0.00016072782913707868,
+      "loss": 2.6635,
+      "step": 5916
+    },
+    {
+      "epoch": 0.4775240093616334,
+      "grad_norm": 0.7188509702682495,
+      "learning_rate": 0.0001607152858607335,
+      "loss": 2.6899,
+      "step": 5917
+    },
+    {
+      "epoch": 0.47760471309821645,
+      "grad_norm": 0.6742647290229797,
+      "learning_rate": 0.00016070274107120468,
+      "loss": 2.6221,
+      "step": 5918
+    },
+    {
+      "epoch": 0.4776854168347994,
+      "grad_norm": 0.7274083495140076,
+      "learning_rate": 0.00016069019476880488,
+      "loss": 2.6588,
+      "step": 5919
+    },
+    {
+      "epoch": 0.47776612057138246,
+      "grad_norm": 0.6984386444091797,
+      "learning_rate": 0.00016067764695384682,
+      "loss": 2.6376,
+      "step": 5920
+    },
+    {
+      "epoch": 0.47784682430796543,
+      "grad_norm": 0.7260883450508118,
+      "learning_rate": 0.00016066509762664315,
+      "loss": 2.6623,
+      "step": 5921
+    },
+    {
+      "epoch": 0.47792752804454847,
+      "grad_norm": 0.7540579438209534,
+      "learning_rate": 0.00016065254678750666,
+      "loss": 2.695,
+      "step": 5922
+    },
+    {
+      "epoch": 0.47800823178113144,
+      "grad_norm": 0.7032651305198669,
+      "learning_rate": 0.00016063999443675017,
+      "loss": 2.6791,
+      "step": 5923
+    },
+    {
+      "epoch": 0.4780889355177145,
+      "grad_norm": 0.682842493057251,
+      "learning_rate": 0.0001606274405746865,
+      "loss": 2.6198,
+      "step": 5924
+    },
+    {
+      "epoch": 0.47816963925429745,
+      "grad_norm": 0.6843859553337097,
+      "learning_rate": 0.00016061488520162853,
+      "loss": 2.6432,
+      "step": 5925
+    },
+    {
+      "epoch": 0.4782503429908805,
+      "grad_norm": 0.652119517326355,
+      "learning_rate": 0.00016060232831788918,
+      "loss": 2.6461,
+      "step": 5926
+    },
+    {
+      "epoch": 0.47833104672746346,
+      "grad_norm": 0.6986887454986572,
+      "learning_rate": 0.0001605897699237814,
+      "loss": 2.5885,
+      "step": 5927
+    },
+    {
+      "epoch": 0.4784117504640465,
+      "grad_norm": 0.7156725525856018,
+      "learning_rate": 0.00016057721001961817,
+      "loss": 2.6526,
+      "step": 5928
+    },
+    {
+      "epoch": 0.47849245420062947,
+      "grad_norm": 0.7367579936981201,
+      "learning_rate": 0.0001605646486057125,
+      "loss": 2.5842,
+      "step": 5929
+    },
+    {
+      "epoch": 0.4785731579372125,
+      "grad_norm": 0.7059770822525024,
+      "learning_rate": 0.00016055208568237746,
+      "loss": 2.617,
+      "step": 5930
+    },
+    {
+      "epoch": 0.4786538616737955,
+      "grad_norm": 0.7225117087364197,
+      "learning_rate": 0.00016053952124992619,
+      "loss": 2.6499,
+      "step": 5931
+    },
+    {
+      "epoch": 0.4787345654103785,
+      "grad_norm": 0.7027475237846375,
+      "learning_rate": 0.00016052695530867177,
+      "loss": 2.5934,
+      "step": 5932
+    },
+    {
+      "epoch": 0.4788152691469615,
+      "grad_norm": 0.7031852602958679,
+      "learning_rate": 0.00016051438785892743,
+      "loss": 2.5947,
+      "step": 5933
+    },
+    {
+      "epoch": 0.4788959728835445,
+      "grad_norm": 0.6731768846511841,
+      "learning_rate": 0.00016050181890100635,
+      "loss": 2.6811,
+      "step": 5934
+    },
+    {
+      "epoch": 0.4789766766201275,
+      "grad_norm": 0.7120038866996765,
+      "learning_rate": 0.0001604892484352218,
+      "loss": 2.6625,
+      "step": 5935
+    },
+    {
+      "epoch": 0.4790573803567105,
+      "grad_norm": 0.6895150542259216,
+      "learning_rate": 0.00016047667646188702,
+      "loss": 2.6784,
+      "step": 5936
+    },
+    {
+      "epoch": 0.4791380840932935,
+      "grad_norm": 0.7080708742141724,
+      "learning_rate": 0.0001604641029813154,
+      "loss": 2.6491,
+      "step": 5937
+    },
+    {
+      "epoch": 0.47921878782987654,
+      "grad_norm": 0.6522819399833679,
+      "learning_rate": 0.00016045152799382025,
+      "loss": 2.6113,
+      "step": 5938
+    },
+    {
+      "epoch": 0.4792994915664595,
+      "grad_norm": 0.6988112926483154,
+      "learning_rate": 0.00016043895149971506,
+      "loss": 2.6892,
+      "step": 5939
+    },
+    {
+      "epoch": 0.47938019530304254,
+      "grad_norm": 0.7545368671417236,
+      "learning_rate": 0.00016042637349931318,
+      "loss": 2.6872,
+      "step": 5940
+    },
+    {
+      "epoch": 0.4794608990396255,
+      "grad_norm": 0.7083707451820374,
+      "learning_rate": 0.0001604137939929281,
+      "loss": 2.6726,
+      "step": 5941
+    },
+    {
+      "epoch": 0.47954160277620855,
+      "grad_norm": 0.8198027014732361,
+      "learning_rate": 0.00016040121298087337,
+      "loss": 2.647,
+      "step": 5942
+    },
+    {
+      "epoch": 0.47962230651279153,
+      "grad_norm": 0.7296201586723328,
+      "learning_rate": 0.00016038863046346252,
+      "loss": 2.7122,
+      "step": 5943
+    },
+    {
+      "epoch": 0.47970301024937456,
+      "grad_norm": 0.7262474298477173,
+      "learning_rate": 0.00016037604644100913,
+      "loss": 2.6903,
+      "step": 5944
+    },
+    {
+      "epoch": 0.47978371398595754,
+      "grad_norm": 0.8010182976722717,
+      "learning_rate": 0.00016036346091382686,
+      "loss": 2.6942,
+      "step": 5945
+    },
+    {
+      "epoch": 0.47986441772254057,
+      "grad_norm": 0.7227098345756531,
+      "learning_rate": 0.00016035087388222932,
+      "loss": 2.6661,
+      "step": 5946
+    },
+    {
+      "epoch": 0.47994512145912355,
+      "grad_norm": 0.7374662756919861,
+      "learning_rate": 0.00016033828534653028,
+      "loss": 2.6233,
+      "step": 5947
+    },
+    {
+      "epoch": 0.4800258251957066,
+      "grad_norm": 0.7139650583267212,
+      "learning_rate": 0.00016032569530704342,
+      "loss": 2.5859,
+      "step": 5948
+    },
+    {
+      "epoch": 0.48010652893228956,
+      "grad_norm": 0.7067660689353943,
+      "learning_rate": 0.00016031310376408254,
+      "loss": 2.6677,
+      "step": 5949
+    },
+    {
+      "epoch": 0.4801872326688726,
+      "grad_norm": 0.694715142250061,
+      "learning_rate": 0.00016030051071796146,
+      "loss": 2.6415,
+      "step": 5950
+    },
+    {
+      "epoch": 0.48026793640545556,
+      "grad_norm": 0.728918194770813,
+      "learning_rate": 0.00016028791616899403,
+      "loss": 2.6274,
+      "step": 5951
+    },
+    {
+      "epoch": 0.4803486401420386,
+      "grad_norm": 0.699846088886261,
+      "learning_rate": 0.00016027532011749412,
+      "loss": 2.6613,
+      "step": 5952
+    },
+    {
+      "epoch": 0.4804293438786216,
+      "grad_norm": 0.7177432179450989,
+      "learning_rate": 0.0001602627225637757,
+      "loss": 2.6107,
+      "step": 5953
+    },
+    {
+      "epoch": 0.4805100476152046,
+      "grad_norm": 0.7502370476722717,
+      "learning_rate": 0.00016025012350815267,
+      "loss": 2.6534,
+      "step": 5954
+    },
+    {
+      "epoch": 0.4805907513517876,
+      "grad_norm": 0.7730218172073364,
+      "learning_rate": 0.0001602375229509391,
+      "loss": 2.7037,
+      "step": 5955
+    },
+    {
+      "epoch": 0.4806714550883706,
+      "grad_norm": 0.7046666145324707,
+      "learning_rate": 0.00016022492089244898,
+      "loss": 2.6336,
+      "step": 5956
+    },
+    {
+      "epoch": 0.4807521588249536,
+      "grad_norm": 0.7991104125976562,
+      "learning_rate": 0.0001602123173329964,
+      "loss": 2.7024,
+      "step": 5957
+    },
+    {
+      "epoch": 0.4808328625615366,
+      "grad_norm": 0.7056288123130798,
+      "learning_rate": 0.00016019971227289548,
+      "loss": 2.6088,
+      "step": 5958
+    },
+    {
+      "epoch": 0.4809135662981196,
+      "grad_norm": 0.7277925610542297,
+      "learning_rate": 0.00016018710571246038,
+      "loss": 2.6245,
+      "step": 5959
+    },
+    {
+      "epoch": 0.48099427003470263,
+      "grad_norm": 0.7545790672302246,
+      "learning_rate": 0.00016017449765200526,
+      "loss": 2.6076,
+      "step": 5960
+    },
+    {
+      "epoch": 0.4810749737712856,
+      "grad_norm": 0.7106321454048157,
+      "learning_rate": 0.00016016188809184434,
+      "loss": 2.5561,
+      "step": 5961
+    },
+    {
+      "epoch": 0.48115567750786864,
+      "grad_norm": 0.7464704513549805,
+      "learning_rate": 0.0001601492770322919,
+      "loss": 2.6336,
+      "step": 5962
+    },
+    {
+      "epoch": 0.4812363812444516,
+      "grad_norm": 0.7531768083572388,
+      "learning_rate": 0.00016013666447366228,
+      "loss": 2.6236,
+      "step": 5963
+    },
+    {
+      "epoch": 0.48131708498103465,
+      "grad_norm": 0.7412876486778259,
+      "learning_rate": 0.00016012405041626978,
+      "loss": 2.6309,
+      "step": 5964
+    },
+    {
+      "epoch": 0.4813977887176176,
+      "grad_norm": 0.7030940055847168,
+      "learning_rate": 0.00016011143486042878,
+      "loss": 2.6252,
+      "step": 5965
+    },
+    {
+      "epoch": 0.48147849245420066,
+      "grad_norm": 0.7932302951812744,
+      "learning_rate": 0.00016009881780645367,
+      "loss": 2.6797,
+      "step": 5966
+    },
+    {
+      "epoch": 0.48155919619078363,
+      "grad_norm": 0.7366262078285217,
+      "learning_rate": 0.00016008619925465893,
+      "loss": 2.6616,
+      "step": 5967
+    },
+    {
+      "epoch": 0.4816398999273666,
+      "grad_norm": 0.6938421130180359,
+      "learning_rate": 0.00016007357920535902,
+      "loss": 2.6888,
+      "step": 5968
+    },
+    {
+      "epoch": 0.48172060366394964,
+      "grad_norm": 0.7560005784034729,
+      "learning_rate": 0.00016006095765886853,
+      "loss": 2.6044,
+      "step": 5969
+    },
+    {
+      "epoch": 0.4818013074005326,
+      "grad_norm": 0.7330430150032043,
+      "learning_rate": 0.0001600483346155019,
+      "loss": 2.7023,
+      "step": 5970
+    },
+    {
+      "epoch": 0.48188201113711565,
+      "grad_norm": 0.7257955074310303,
+      "learning_rate": 0.00016003571007557388,
+      "loss": 2.6763,
+      "step": 5971
+    },
+    {
+      "epoch": 0.4819627148736986,
+      "grad_norm": 0.704187273979187,
+      "learning_rate": 0.000160023084039399,
+      "loss": 2.6229,
+      "step": 5972
+    },
+    {
+      "epoch": 0.48204341861028166,
+      "grad_norm": 0.7014813423156738,
+      "learning_rate": 0.00016001045650729196,
+      "loss": 2.6207,
+      "step": 5973
+    },
+    {
+      "epoch": 0.48212412234686464,
+      "grad_norm": 0.8039405941963196,
+      "learning_rate": 0.00015999782747956747,
+      "loss": 2.6198,
+      "step": 5974
+    },
+    {
+      "epoch": 0.48220482608344767,
+      "grad_norm": 0.7114945650100708,
+      "learning_rate": 0.0001599851969565403,
+      "loss": 2.6154,
+      "step": 5975
+    },
+    {
+      "epoch": 0.48228552982003065,
+      "grad_norm": 0.7603329420089722,
+      "learning_rate": 0.00015997256493852517,
+      "loss": 2.6217,
+      "step": 5976
+    },
+    {
+      "epoch": 0.4823662335566137,
+      "grad_norm": 0.7773346900939941,
+      "learning_rate": 0.000159959931425837,
+      "loss": 2.7054,
+      "step": 5977
+    },
+    {
+      "epoch": 0.48244693729319665,
+      "grad_norm": 0.8022029399871826,
+      "learning_rate": 0.0001599472964187906,
+      "loss": 2.6844,
+      "step": 5978
+    },
+    {
+      "epoch": 0.4825276410297797,
+      "grad_norm": 0.7384541630744934,
+      "learning_rate": 0.00015993465991770087,
+      "loss": 2.6516,
+      "step": 5979
+    },
+    {
+      "epoch": 0.48260834476636266,
+      "grad_norm": 0.6993509531021118,
+      "learning_rate": 0.00015992202192288273,
+      "loss": 2.6837,
+      "step": 5980
+    },
+    {
+      "epoch": 0.4826890485029457,
+      "grad_norm": 0.7430509328842163,
+      "learning_rate": 0.00015990938243465116,
+      "loss": 2.6717,
+      "step": 5981
+    },
+    {
+      "epoch": 0.48276975223952867,
+      "grad_norm": 0.7544847726821899,
+      "learning_rate": 0.0001598967414533212,
+      "loss": 2.6573,
+      "step": 5982
+    },
+    {
+      "epoch": 0.4828504559761117,
+      "grad_norm": 0.736955463886261,
+      "learning_rate": 0.00015988409897920786,
+      "loss": 2.6865,
+      "step": 5983
+    },
+    {
+      "epoch": 0.4829311597126947,
+      "grad_norm": 0.7771684527397156,
+      "learning_rate": 0.00015987145501262622,
+      "loss": 2.6173,
+      "step": 5984
+    },
+    {
+      "epoch": 0.4830118634492777,
+      "grad_norm": 0.7504391670227051,
+      "learning_rate": 0.00015985880955389143,
+      "loss": 2.6218,
+      "step": 5985
+    },
+    {
+      "epoch": 0.4830925671858607,
+      "grad_norm": 0.7025442123413086,
+      "learning_rate": 0.00015984616260331861,
+      "loss": 2.6107,
+      "step": 5986
+    },
+    {
+      "epoch": 0.4831732709224437,
+      "grad_norm": 0.6906485557556152,
+      "learning_rate": 0.000159833514161223,
+      "loss": 2.633,
+      "step": 5987
+    },
+    {
+      "epoch": 0.4832539746590267,
+      "grad_norm": 0.7771004438400269,
+      "learning_rate": 0.00015982086422791983,
+      "loss": 2.5956,
+      "step": 5988
+    },
+    {
+      "epoch": 0.48333467839560973,
+      "grad_norm": 0.6927372813224792,
+      "learning_rate": 0.00015980821280372432,
+      "loss": 2.5984,
+      "step": 5989
+    },
+    {
+      "epoch": 0.4834153821321927,
+      "grad_norm": 0.7196357846260071,
+      "learning_rate": 0.00015979555988895184,
+      "loss": 2.6386,
+      "step": 5990
+    },
+    {
+      "epoch": 0.48349608586877574,
+      "grad_norm": 0.7601087689399719,
+      "learning_rate": 0.0001597829054839177,
+      "loss": 2.6707,
+      "step": 5991
+    },
+    {
+      "epoch": 0.4835767896053587,
+      "grad_norm": 0.7783588767051697,
+      "learning_rate": 0.00015977024958893722,
+      "loss": 2.5815,
+      "step": 5992
+    },
+    {
+      "epoch": 0.48365749334194175,
+      "grad_norm": 0.7651833891868591,
+      "learning_rate": 0.00015975759220432592,
+      "loss": 2.6235,
+      "step": 5993
+    },
+    {
+      "epoch": 0.4837381970785247,
+      "grad_norm": 0.7158511877059937,
+      "learning_rate": 0.0001597449333303992,
+      "loss": 2.6813,
+      "step": 5994
+    },
+    {
+      "epoch": 0.48381890081510776,
+      "grad_norm": 0.7411341667175293,
+      "learning_rate": 0.0001597322729674726,
+      "loss": 2.7231,
+      "step": 5995
+    },
+    {
+      "epoch": 0.48389960455169073,
+      "grad_norm": 0.7168158292770386,
+      "learning_rate": 0.0001597196111158616,
+      "loss": 2.6408,
+      "step": 5996
+    },
+    {
+      "epoch": 0.48398030828827376,
+      "grad_norm": 0.7603393793106079,
+      "learning_rate": 0.00015970694777588175,
+      "loss": 2.7821,
+      "step": 5997
+    },
+    {
+      "epoch": 0.48406101202485674,
+      "grad_norm": 0.7298564910888672,
+      "learning_rate": 0.0001596942829478487,
+      "loss": 2.6828,
+      "step": 5998
+    },
+    {
+      "epoch": 0.4841417157614398,
+      "grad_norm": 0.7850572466850281,
+      "learning_rate": 0.0001596816166320781,
+      "loss": 2.6191,
+      "step": 5999
+    },
+    {
+      "epoch": 0.48422241949802275,
+      "grad_norm": 0.7697601914405823,
+      "learning_rate": 0.00015966894882888562,
+      "loss": 2.6768,
+      "step": 6000
+    },
+    {
+      "epoch": 0.48422241949802275,
+      "eval_loss": 2.5610127449035645,
+      "eval_runtime": 760.0481,
+      "eval_samples_per_second": 3.447,
+      "eval_steps_per_second": 0.575,
+      "step": 6000
+    },
+    {
+      "epoch": 0.4843031232346058,
+      "grad_norm": 0.7212432026863098,
+      "learning_rate": 0.00015965627953858693,
+      "loss": 2.5967,
+      "step": 6001
+    },
+    {
+      "epoch": 0.48438382697118876,
+      "grad_norm": 0.7629631757736206,
+      "learning_rate": 0.0001596436087614978,
+      "loss": 2.7005,
+      "step": 6002
+    },
+    {
+      "epoch": 0.4844645307077718,
+      "grad_norm": 0.7154754400253296,
+      "learning_rate": 0.00015963093649793404,
+      "loss": 2.6909,
+      "step": 6003
+    },
+    {
+      "epoch": 0.48454523444435477,
+      "grad_norm": 0.7365279793739319,
+      "learning_rate": 0.00015961826274821147,
+      "loss": 2.6268,
+      "step": 6004
+    },
+    {
+      "epoch": 0.4846259381809378,
+      "grad_norm": 0.8114632964134216,
+      "learning_rate": 0.00015960558751264596,
+      "loss": 2.6647,
+      "step": 6005
+    },
+    {
+      "epoch": 0.4847066419175208,
+      "grad_norm": 0.7411556243896484,
+      "learning_rate": 0.00015959291079155338,
+      "loss": 2.6378,
+      "step": 6006
+    },
+    {
+      "epoch": 0.4847873456541038,
+      "grad_norm": 0.7137390375137329,
+      "learning_rate": 0.00015958023258524968,
+      "loss": 2.6454,
+      "step": 6007
+    },
+    {
+      "epoch": 0.4848680493906868,
+      "grad_norm": 0.7477054595947266,
+      "learning_rate": 0.00015956755289405088,
+      "loss": 2.6463,
+      "step": 6008
+    },
+    {
+      "epoch": 0.4849487531272698,
+      "grad_norm": 0.7198071479797363,
+      "learning_rate": 0.0001595548717182729,
+      "loss": 2.6537,
+      "step": 6009
+    },
+    {
+      "epoch": 0.4850294568638528,
+      "grad_norm": 0.6697781085968018,
+      "learning_rate": 0.00015954218905823186,
+      "loss": 2.7018,
+      "step": 6010
+    },
+    {
+      "epoch": 0.4851101606004358,
+      "grad_norm": 0.7577201724052429,
+      "learning_rate": 0.00015952950491424382,
+      "loss": 2.6531,
+      "step": 6011
+    },
+    {
+      "epoch": 0.4851908643370188,
+      "grad_norm": 0.6852774024009705,
+      "learning_rate": 0.0001595168192866249,
+      "loss": 2.5819,
+      "step": 6012
+    },
+    {
+      "epoch": 0.48527156807360183,
+      "grad_norm": 0.7116097807884216,
+      "learning_rate": 0.0001595041321756913,
+      "loss": 2.5691,
+      "step": 6013
+    },
+    {
+      "epoch": 0.4853522718101848,
+      "grad_norm": 0.7478477954864502,
+      "learning_rate": 0.00015949144358175916,
+      "loss": 2.6658,
+      "step": 6014
+    },
+    {
+      "epoch": 0.48543297554676784,
+      "grad_norm": 0.816969633102417,
+      "learning_rate": 0.0001594787535051447,
+      "loss": 2.6709,
+      "step": 6015
+    },
+    {
+      "epoch": 0.4855136792833508,
+      "grad_norm": 0.6953164339065552,
+      "learning_rate": 0.00015946606194616427,
+      "loss": 2.6139,
+      "step": 6016
+    },
+    {
+      "epoch": 0.48559438301993385,
+      "grad_norm": 0.6698834300041199,
+      "learning_rate": 0.0001594533689051341,
+      "loss": 2.574,
+      "step": 6017
+    },
+    {
+      "epoch": 0.4856750867565168,
+      "grad_norm": 0.7686784267425537,
+      "learning_rate": 0.0001594406743823706,
+      "loss": 2.6271,
+      "step": 6018
+    },
+    {
+      "epoch": 0.4857557904930998,
+      "grad_norm": 0.7713280916213989,
+      "learning_rate": 0.00015942797837819009,
+      "loss": 2.6682,
+      "step": 6019
+    },
+    {
+      "epoch": 0.48583649422968284,
+      "grad_norm": 0.8102596998214722,
+      "learning_rate": 0.00015941528089290902,
+      "loss": 2.6771,
+      "step": 6020
+    },
+    {
+      "epoch": 0.4859171979662658,
+      "grad_norm": 0.7140331864356995,
+      "learning_rate": 0.00015940258192684382,
+      "loss": 2.6267,
+      "step": 6021
+    },
+    {
+      "epoch": 0.48599790170284884,
+      "grad_norm": 0.7057615518569946,
+      "learning_rate": 0.000159389881480311,
+      "loss": 2.6011,
+      "step": 6022
+    },
+    {
+      "epoch": 0.4860786054394318,
+      "grad_norm": 0.7106850147247314,
+      "learning_rate": 0.0001593771795536271,
+      "loss": 2.6681,
+      "step": 6023
+    },
+    {
+      "epoch": 0.48615930917601485,
+      "grad_norm": 0.7618210315704346,
+      "learning_rate": 0.00015936447614710867,
+      "loss": 2.6545,
+      "step": 6024
+    },
+    {
+      "epoch": 0.48624001291259783,
+      "grad_norm": 0.7577608227729797,
+      "learning_rate": 0.00015935177126107233,
+      "loss": 2.6479,
+      "step": 6025
+    },
+    {
+      "epoch": 0.48632071664918086,
+      "grad_norm": 0.758745551109314,
+      "learning_rate": 0.00015933906489583468,
+      "loss": 2.7057,
+      "step": 6026
+    },
+    {
+      "epoch": 0.48640142038576384,
+      "grad_norm": 0.785906970500946,
+      "learning_rate": 0.00015932635705171241,
+      "loss": 2.7081,
+      "step": 6027
+    },
+    {
+      "epoch": 0.48648212412234687,
+      "grad_norm": 0.6744558215141296,
+      "learning_rate": 0.00015931364772902228,
+      "loss": 2.6438,
+      "step": 6028
+    },
+    {
+      "epoch": 0.48656282785892985,
+      "grad_norm": 0.7451377511024475,
+      "learning_rate": 0.00015930093692808099,
+      "loss": 2.6509,
+      "step": 6029
+    },
+    {
+      "epoch": 0.4866435315955129,
+      "grad_norm": 0.6590149402618408,
+      "learning_rate": 0.0001592882246492053,
+      "loss": 2.5683,
+      "step": 6030
+    },
+    {
+      "epoch": 0.48672423533209586,
+      "grad_norm": 0.7433840036392212,
+      "learning_rate": 0.0001592755108927121,
+      "loss": 2.6647,
+      "step": 6031
+    },
+    {
+      "epoch": 0.4868049390686789,
+      "grad_norm": 0.876806378364563,
+      "learning_rate": 0.00015926279565891822,
+      "loss": 2.6482,
+      "step": 6032
+    },
+    {
+      "epoch": 0.48688564280526186,
+      "grad_norm": 0.7495005130767822,
+      "learning_rate": 0.00015925007894814058,
+      "loss": 2.6346,
+      "step": 6033
+    },
+    {
+      "epoch": 0.4869663465418449,
+      "grad_norm": 0.7005730271339417,
+      "learning_rate": 0.00015923736076069604,
+      "loss": 2.6241,
+      "step": 6034
+    },
+    {
+      "epoch": 0.4870470502784279,
+      "grad_norm": 0.664098858833313,
+      "learning_rate": 0.00015922464109690166,
+      "loss": 2.6281,
+      "step": 6035
+    },
+    {
+      "epoch": 0.4871277540150109,
+      "grad_norm": 0.7482514977455139,
+      "learning_rate": 0.00015921191995707442,
+      "loss": 2.5764,
+      "step": 6036
+    },
+    {
+      "epoch": 0.4872084577515939,
+      "grad_norm": 0.7450351715087891,
+      "learning_rate": 0.0001591991973415313,
+      "loss": 2.6433,
+      "step": 6037
+    },
+    {
+      "epoch": 0.4872891614881769,
+      "grad_norm": 0.6738519072532654,
+      "learning_rate": 0.00015918647325058948,
+      "loss": 2.6688,
+      "step": 6038
+    },
+    {
+      "epoch": 0.4873698652247599,
+      "grad_norm": 0.7999960780143738,
+      "learning_rate": 0.000159173747684566,
+      "loss": 2.6309,
+      "step": 6039
+    },
+    {
+      "epoch": 0.4874505689613429,
+      "grad_norm": 0.7249687910079956,
+      "learning_rate": 0.00015916102064377806,
+      "loss": 2.5808,
+      "step": 6040
+    },
+    {
+      "epoch": 0.4875312726979259,
+      "grad_norm": 0.7014601826667786,
+      "learning_rate": 0.00015914829212854286,
+      "loss": 2.6646,
+      "step": 6041
+    },
+    {
+      "epoch": 0.48761197643450893,
+      "grad_norm": 0.7091174721717834,
+      "learning_rate": 0.00015913556213917757,
+      "loss": 2.6576,
+      "step": 6042
+    },
+    {
+      "epoch": 0.4876926801710919,
+      "grad_norm": 0.6949019432067871,
+      "learning_rate": 0.00015912283067599952,
+      "loss": 2.5883,
+      "step": 6043
+    },
+    {
+      "epoch": 0.48777338390767494,
+      "grad_norm": 0.6990448236465454,
+      "learning_rate": 0.00015911009773932598,
+      "loss": 2.6413,
+      "step": 6044
+    },
+    {
+      "epoch": 0.4878540876442579,
+      "grad_norm": 0.7106831073760986,
+      "learning_rate": 0.00015909736332947425,
+      "loss": 2.6122,
+      "step": 6045
+    },
+    {
+      "epoch": 0.48793479138084095,
+      "grad_norm": 0.7052395343780518,
+      "learning_rate": 0.00015908462744676177,
+      "loss": 2.572,
+      "step": 6046
+    },
+    {
+      "epoch": 0.4880154951174239,
+      "grad_norm": 0.7250158190727234,
+      "learning_rate": 0.00015907189009150592,
+      "loss": 2.6582,
+      "step": 6047
+    },
+    {
+      "epoch": 0.48809619885400696,
+      "grad_norm": 0.7213590145111084,
+      "learning_rate": 0.00015905915126402414,
+      "loss": 2.7025,
+      "step": 6048
+    },
+    {
+      "epoch": 0.48817690259058993,
+      "grad_norm": 0.7136254906654358,
+      "learning_rate": 0.00015904641096463394,
+      "loss": 2.6823,
+      "step": 6049
+    },
+    {
+      "epoch": 0.48825760632717297,
+      "grad_norm": 0.7163361310958862,
+      "learning_rate": 0.00015903366919365282,
+      "loss": 2.6642,
+      "step": 6050
+    },
+    {
+      "epoch": 0.48833831006375594,
+      "grad_norm": 0.6842724680900574,
+      "learning_rate": 0.00015902092595139838,
+      "loss": 2.6599,
+      "step": 6051
+    },
+    {
+      "epoch": 0.488419013800339,
+      "grad_norm": 0.7426519393920898,
+      "learning_rate": 0.0001590081812381882,
+      "loss": 2.6271,
+      "step": 6052
+    },
+    {
+      "epoch": 0.48849971753692195,
+      "grad_norm": 0.7415586709976196,
+      "learning_rate": 0.00015899543505433985,
+      "loss": 2.6105,
+      "step": 6053
+    },
+    {
+      "epoch": 0.488580421273505,
+      "grad_norm": 0.7286739945411682,
+      "learning_rate": 0.00015898268740017105,
+      "loss": 2.6304,
+      "step": 6054
+    },
+    {
+      "epoch": 0.48866112501008796,
+      "grad_norm": 0.6898483633995056,
+      "learning_rate": 0.00015896993827599947,
+      "loss": 2.6237,
+      "step": 6055
+    },
+    {
+      "epoch": 0.488741828746671,
+      "grad_norm": 0.7020056247711182,
+      "learning_rate": 0.00015895718768214293,
+      "loss": 2.6166,
+      "step": 6056
+    },
+    {
+      "epoch": 0.48882253248325397,
+      "grad_norm": 0.7145286798477173,
+      "learning_rate": 0.00015894443561891914,
+      "loss": 2.6729,
+      "step": 6057
+    },
+    {
+      "epoch": 0.488903236219837,
+      "grad_norm": 0.6888289451599121,
+      "learning_rate": 0.00015893168208664594,
+      "loss": 2.6154,
+      "step": 6058
+    },
+    {
+      "epoch": 0.48898393995642,
+      "grad_norm": 0.6929970383644104,
+      "learning_rate": 0.00015891892708564116,
+      "loss": 2.6748,
+      "step": 6059
+    },
+    {
+      "epoch": 0.489064643693003,
+      "grad_norm": 0.679853618144989,
+      "learning_rate": 0.0001589061706162227,
+      "loss": 2.605,
+      "step": 6060
+    },
+    {
+      "epoch": 0.489145347429586,
+      "grad_norm": 0.71812504529953,
+      "learning_rate": 0.0001588934126787085,
+      "loss": 2.7249,
+      "step": 6061
+    },
+    {
+      "epoch": 0.489226051166169,
+      "grad_norm": 0.7083466053009033,
+      "learning_rate": 0.00015888065327341648,
+      "loss": 2.5986,
+      "step": 6062
+    },
+    {
+      "epoch": 0.489306754902752,
+      "grad_norm": 0.7476792931556702,
+      "learning_rate": 0.00015886789240066466,
+      "loss": 2.5942,
+      "step": 6063
+    },
+    {
+      "epoch": 0.489387458639335,
+      "grad_norm": 0.7197855114936829,
+      "learning_rate": 0.00015885513006077114,
+      "loss": 2.6198,
+      "step": 6064
+    },
+    {
+      "epoch": 0.489468162375918,
+      "grad_norm": 0.6678233742713928,
+      "learning_rate": 0.00015884236625405385,
+      "loss": 2.5793,
+      "step": 6065
+    },
+    {
+      "epoch": 0.48954886611250104,
+      "grad_norm": 0.7371037602424622,
+      "learning_rate": 0.00015882960098083105,
+      "loss": 2.6231,
+      "step": 6066
+    },
+    {
+      "epoch": 0.489629569849084,
+      "grad_norm": 0.7087417244911194,
+      "learning_rate": 0.00015881683424142078,
+      "loss": 2.6483,
+      "step": 6067
+    },
+    {
+      "epoch": 0.48971027358566704,
+      "grad_norm": 0.7300292253494263,
+      "learning_rate": 0.00015880406603614126,
+      "loss": 2.6778,
+      "step": 6068
+    },
+    {
+      "epoch": 0.48979097732225,
+      "grad_norm": 0.8347866535186768,
+      "learning_rate": 0.0001587912963653107,
+      "loss": 2.554,
+      "step": 6069
+    },
+    {
+      "epoch": 0.489871681058833,
+      "grad_norm": 0.7717794179916382,
+      "learning_rate": 0.00015877852522924732,
+      "loss": 2.6904,
+      "step": 6070
+    },
+    {
+      "epoch": 0.48995238479541603,
+      "grad_norm": 0.6960952281951904,
+      "learning_rate": 0.00015876575262826944,
+      "loss": 2.6059,
+      "step": 6071
+    },
+    {
+      "epoch": 0.490033088531999,
+      "grad_norm": 0.7316592931747437,
+      "learning_rate": 0.00015875297856269543,
+      "loss": 2.6685,
+      "step": 6072
+    },
+    {
+      "epoch": 0.49011379226858204,
+      "grad_norm": 0.6775457859039307,
+      "learning_rate": 0.00015874020303284362,
+      "loss": 2.6232,
+      "step": 6073
+    },
+    {
+      "epoch": 0.490194496005165,
+      "grad_norm": 0.7741925120353699,
+      "learning_rate": 0.00015872742603903237,
+      "loss": 2.6767,
+      "step": 6074
+    },
+    {
+      "epoch": 0.49027519974174805,
+      "grad_norm": 0.857490599155426,
+      "learning_rate": 0.00015871464758158017,
+      "loss": 2.6649,
+      "step": 6075
+    },
+    {
+      "epoch": 0.490355903478331,
+      "grad_norm": 0.7474274039268494,
+      "learning_rate": 0.00015870186766080545,
+      "loss": 2.6926,
+      "step": 6076
+    },
+    {
+      "epoch": 0.49043660721491406,
+      "grad_norm": 0.7266567945480347,
+      "learning_rate": 0.00015868908627702675,
+      "loss": 2.5919,
+      "step": 6077
+    },
+    {
+      "epoch": 0.49051731095149703,
+      "grad_norm": 0.7247830629348755,
+      "learning_rate": 0.0001586763034305626,
+      "loss": 2.6158,
+      "step": 6078
+    },
+    {
+      "epoch": 0.49059801468808006,
+      "grad_norm": 0.7654951214790344,
+      "learning_rate": 0.00015866351912173157,
+      "loss": 2.7236,
+      "step": 6079
+    },
+    {
+      "epoch": 0.49067871842466304,
+      "grad_norm": 0.732431948184967,
+      "learning_rate": 0.00015865073335085236,
+      "loss": 2.6349,
+      "step": 6080
+    },
+    {
+      "epoch": 0.4907594221612461,
+      "grad_norm": 0.7240673303604126,
+      "learning_rate": 0.0001586379461182435,
+      "loss": 2.6282,
+      "step": 6081
+    },
+    {
+      "epoch": 0.49084012589782905,
+      "grad_norm": 0.767473042011261,
+      "learning_rate": 0.00015862515742422374,
+      "loss": 2.6939,
+      "step": 6082
+    },
+    {
+      "epoch": 0.4909208296344121,
+      "grad_norm": 0.6977359056472778,
+      "learning_rate": 0.00015861236726911183,
+      "loss": 2.6591,
+      "step": 6083
+    },
+    {
+      "epoch": 0.49100153337099506,
+      "grad_norm": 0.7676639556884766,
+      "learning_rate": 0.00015859957565322655,
+      "loss": 2.6189,
+      "step": 6084
+    },
+    {
+      "epoch": 0.4910822371075781,
+      "grad_norm": 0.7157976031303406,
+      "learning_rate": 0.0001585867825768866,
+      "loss": 2.644,
+      "step": 6085
+    },
+    {
+      "epoch": 0.49116294084416107,
+      "grad_norm": 0.7080803513526917,
+      "learning_rate": 0.0001585739880404109,
+      "loss": 2.6099,
+      "step": 6086
+    },
+    {
+      "epoch": 0.4912436445807441,
+      "grad_norm": 0.7109760046005249,
+      "learning_rate": 0.0001585611920441183,
+      "loss": 2.7087,
+      "step": 6087
+    },
+    {
+      "epoch": 0.4913243483173271,
+      "grad_norm": 0.7274255156517029,
+      "learning_rate": 0.00015854839458832772,
+      "loss": 2.6394,
+      "step": 6088
+    },
+    {
+      "epoch": 0.4914050520539101,
+      "grad_norm": 0.7407883405685425,
+      "learning_rate": 0.00015853559567335812,
+      "loss": 2.6729,
+      "step": 6089
+    },
+    {
+      "epoch": 0.4914857557904931,
+      "grad_norm": 0.6879885196685791,
+      "learning_rate": 0.00015852279529952843,
+      "loss": 2.5971,
+      "step": 6090
+    },
+    {
+      "epoch": 0.4915664595270761,
+      "grad_norm": 0.7678415179252625,
+      "learning_rate": 0.00015850999346715772,
+      "loss": 2.6606,
+      "step": 6091
+    },
+    {
+      "epoch": 0.4916471632636591,
+      "grad_norm": 0.7108608484268188,
+      "learning_rate": 0.00015849719017656504,
+      "loss": 2.6494,
+      "step": 6092
+    },
+    {
+      "epoch": 0.4917278670002421,
+      "grad_norm": 0.7238833904266357,
+      "learning_rate": 0.00015848438542806945,
+      "loss": 2.6742,
+      "step": 6093
+    },
+    {
+      "epoch": 0.4918085707368251,
+      "grad_norm": 0.7316902279853821,
+      "learning_rate": 0.0001584715792219901,
+      "loss": 2.6757,
+      "step": 6094
+    },
+    {
+      "epoch": 0.49188927447340813,
+      "grad_norm": 0.7339446544647217,
+      "learning_rate": 0.00015845877155864612,
+      "loss": 2.607,
+      "step": 6095
+    },
+    {
+      "epoch": 0.4919699782099911,
+      "grad_norm": 0.6931337714195251,
+      "learning_rate": 0.0001584459624383568,
+      "loss": 2.6203,
+      "step": 6096
+    },
+    {
+      "epoch": 0.49205068194657414,
+      "grad_norm": 0.734229326248169,
+      "learning_rate": 0.00015843315186144126,
+      "loss": 2.646,
+      "step": 6097
+    },
+    {
+      "epoch": 0.4921313856831571,
+      "grad_norm": 0.7764919400215149,
+      "learning_rate": 0.00015842033982821883,
+      "loss": 2.6698,
+      "step": 6098
+    },
+    {
+      "epoch": 0.49221208941974015,
+      "grad_norm": 0.7707986235618591,
+      "learning_rate": 0.00015840752633900887,
+      "loss": 2.6995,
+      "step": 6099
+    },
+    {
+      "epoch": 0.4922927931563231,
+      "grad_norm": 0.7321949601173401,
+      "learning_rate": 0.00015839471139413066,
+      "loss": 2.6517,
+      "step": 6100
+    },
+    {
+      "epoch": 0.49237349689290616,
+      "grad_norm": 0.7087488770484924,
+      "learning_rate": 0.00015838189499390353,
+      "loss": 2.6153,
+      "step": 6101
+    },
+    {
+      "epoch": 0.49245420062948914,
+      "grad_norm": 0.7300730347633362,
+      "learning_rate": 0.00015836907713864706,
+      "loss": 2.5868,
+      "step": 6102
+    },
+    {
+      "epoch": 0.49253490436607217,
+      "grad_norm": 0.8476536273956299,
+      "learning_rate": 0.00015835625782868054,
+      "loss": 2.7158,
+      "step": 6103
+    },
+    {
+      "epoch": 0.49261560810265514,
+      "grad_norm": 0.8062012791633606,
+      "learning_rate": 0.0001583434370643236,
+      "loss": 2.6896,
+      "step": 6104
+    },
+    {
+      "epoch": 0.4926963118392382,
+      "grad_norm": 0.7336686849594116,
+      "learning_rate": 0.00015833061484589562,
+      "loss": 2.6416,
+      "step": 6105
+    },
+    {
+      "epoch": 0.49277701557582115,
+      "grad_norm": 0.6976929306983948,
+      "learning_rate": 0.00015831779117371627,
+      "loss": 2.6279,
+      "step": 6106
+    },
+    {
+      "epoch": 0.4928577193124042,
+      "grad_norm": 0.7262609601020813,
+      "learning_rate": 0.00015830496604810513,
+      "loss": 2.6144,
+      "step": 6107
+    },
+    {
+      "epoch": 0.49293842304898716,
+      "grad_norm": 0.7274572253227234,
+      "learning_rate": 0.00015829213946938183,
+      "loss": 2.7409,
+      "step": 6108
+    },
+    {
+      "epoch": 0.4930191267855702,
+      "grad_norm": 0.7438454031944275,
+      "learning_rate": 0.000158279311437866,
+      "loss": 2.5928,
+      "step": 6109
+    },
+    {
+      "epoch": 0.49309983052215317,
+      "grad_norm": 0.6885421872138977,
+      "learning_rate": 0.00015826648195387742,
+      "loss": 2.6659,
+      "step": 6110
+    },
+    {
+      "epoch": 0.4931805342587362,
+      "grad_norm": 0.6781450510025024,
+      "learning_rate": 0.0001582536510177358,
+      "loss": 2.6068,
+      "step": 6111
+    },
+    {
+      "epoch": 0.4932612379953192,
+      "grad_norm": 0.7618128657341003,
+      "learning_rate": 0.0001582408186297609,
+      "loss": 2.6705,
+      "step": 6112
+    },
+    {
+      "epoch": 0.4933419417319022,
+      "grad_norm": 0.7011203765869141,
+      "learning_rate": 0.00015822798479027256,
+      "loss": 2.596,
+      "step": 6113
+    },
+    {
+      "epoch": 0.4934226454684852,
+      "grad_norm": 0.7727806568145752,
+      "learning_rate": 0.00015821514949959065,
+      "loss": 2.6458,
+      "step": 6114
+    },
+    {
+      "epoch": 0.4935033492050682,
+      "grad_norm": 0.7318129539489746,
+      "learning_rate": 0.00015820231275803502,
+      "loss": 2.6009,
+      "step": 6115
+    },
+    {
+      "epoch": 0.4935840529416512,
+      "grad_norm": 0.6836227178573608,
+      "learning_rate": 0.00015818947456592563,
+      "loss": 2.6311,
+      "step": 6116
+    },
+    {
+      "epoch": 0.49366475667823423,
+      "grad_norm": 0.7657275199890137,
+      "learning_rate": 0.0001581766349235824,
+      "loss": 2.6079,
+      "step": 6117
+    },
+    {
+      "epoch": 0.4937454604148172,
+      "grad_norm": 0.74736487865448,
+      "learning_rate": 0.0001581637938313254,
+      "loss": 2.6752,
+      "step": 6118
+    },
+    {
+      "epoch": 0.49382616415140024,
+      "grad_norm": 0.716708242893219,
+      "learning_rate": 0.00015815095128947454,
+      "loss": 2.5896,
+      "step": 6119
+    },
+    {
+      "epoch": 0.4939068678879832,
+      "grad_norm": 0.740727424621582,
+      "learning_rate": 0.00015813810729835002,
+      "loss": 2.6528,
+      "step": 6120
+    },
+    {
+      "epoch": 0.4939875716245662,
+      "grad_norm": 0.6746687293052673,
+      "learning_rate": 0.0001581252618582719,
+      "loss": 2.6438,
+      "step": 6121
+    },
+    {
+      "epoch": 0.4940682753611492,
+      "grad_norm": 0.7547900080680847,
+      "learning_rate": 0.00015811241496956028,
+      "loss": 2.631,
+      "step": 6122
+    },
+    {
+      "epoch": 0.4941489790977322,
+      "grad_norm": 0.7500903606414795,
+      "learning_rate": 0.0001580995666325354,
+      "loss": 2.7039,
+      "step": 6123
+    },
+    {
+      "epoch": 0.49422968283431523,
+      "grad_norm": 0.7692849636077881,
+      "learning_rate": 0.00015808671684751743,
+      "loss": 2.5922,
+      "step": 6124
+    },
+    {
+      "epoch": 0.4943103865708982,
+      "grad_norm": 0.6964236497879028,
+      "learning_rate": 0.00015807386561482662,
+      "loss": 2.6239,
+      "step": 6125
+    },
+    {
+      "epoch": 0.49439109030748124,
+      "grad_norm": 0.7094165086746216,
+      "learning_rate": 0.0001580610129347833,
+      "loss": 2.6239,
+      "step": 6126
+    },
+    {
+      "epoch": 0.4944717940440642,
+      "grad_norm": 0.7579131126403809,
+      "learning_rate": 0.00015804815880770775,
+      "loss": 2.6654,
+      "step": 6127
+    },
+    {
+      "epoch": 0.49455249778064725,
+      "grad_norm": 0.7687693238258362,
+      "learning_rate": 0.00015803530323392034,
+      "loss": 2.6557,
+      "step": 6128
+    },
+    {
+      "epoch": 0.4946332015172302,
+      "grad_norm": 0.6913540363311768,
+      "learning_rate": 0.0001580224462137415,
+      "loss": 2.6299,
+      "step": 6129
+    },
+    {
+      "epoch": 0.49471390525381326,
+      "grad_norm": 0.7574129700660706,
+      "learning_rate": 0.0001580095877474916,
+      "loss": 2.6327,
+      "step": 6130
+    },
+    {
+      "epoch": 0.49479460899039623,
+      "grad_norm": 0.6834598183631897,
+      "learning_rate": 0.0001579967278354911,
+      "loss": 2.6402,
+      "step": 6131
+    },
+    {
+      "epoch": 0.49487531272697927,
+      "grad_norm": 0.7872750163078308,
+      "learning_rate": 0.00015798386647806057,
+      "loss": 2.6647,
+      "step": 6132
+    },
+    {
+      "epoch": 0.49495601646356224,
+      "grad_norm": 0.705211341381073,
+      "learning_rate": 0.00015797100367552055,
+      "loss": 2.6288,
+      "step": 6133
+    },
+    {
+      "epoch": 0.4950367202001453,
+      "grad_norm": 0.7302640080451965,
+      "learning_rate": 0.00015795813942819155,
+      "loss": 2.6683,
+      "step": 6134
+    },
+    {
+      "epoch": 0.49511742393672825,
+      "grad_norm": 0.7522360682487488,
+      "learning_rate": 0.0001579452737363942,
+      "loss": 2.5885,
+      "step": 6135
+    },
+    {
+      "epoch": 0.4951981276733113,
+      "grad_norm": 0.657376229763031,
+      "learning_rate": 0.0001579324066004492,
+      "loss": 2.5775,
+      "step": 6136
+    },
+    {
+      "epoch": 0.49527883140989426,
+      "grad_norm": 0.7539556622505188,
+      "learning_rate": 0.00015791953802067715,
+      "loss": 2.6236,
+      "step": 6137
+    },
+    {
+      "epoch": 0.4953595351464773,
+      "grad_norm": 0.7090374827384949,
+      "learning_rate": 0.00015790666799739883,
+      "loss": 2.5845,
+      "step": 6138
+    },
+    {
+      "epoch": 0.49544023888306027,
+      "grad_norm": 0.6883948445320129,
+      "learning_rate": 0.00015789379653093497,
+      "loss": 2.6621,
+      "step": 6139
+    },
+    {
+      "epoch": 0.4955209426196433,
+      "grad_norm": 0.7466424107551575,
+      "learning_rate": 0.00015788092362160633,
+      "loss": 2.6289,
+      "step": 6140
+    },
+    {
+      "epoch": 0.4956016463562263,
+      "grad_norm": 0.7424437403678894,
+      "learning_rate": 0.00015786804926973383,
+      "loss": 2.6405,
+      "step": 6141
+    },
+    {
+      "epoch": 0.4956823500928093,
+      "grad_norm": 0.7227851748466492,
+      "learning_rate": 0.00015785517347563822,
+      "loss": 2.6537,
+      "step": 6142
+    },
+    {
+      "epoch": 0.4957630538293923,
+      "grad_norm": 0.7548653483390808,
+      "learning_rate": 0.00015784229623964048,
+      "loss": 2.7377,
+      "step": 6143
+    },
+    {
+      "epoch": 0.4958437575659753,
+      "grad_norm": 0.7086976170539856,
+      "learning_rate": 0.00015782941756206152,
+      "loss": 2.6194,
+      "step": 6144
+    },
+    {
+      "epoch": 0.4959244613025583,
+      "grad_norm": 0.6605533957481384,
+      "learning_rate": 0.0001578165374432223,
+      "loss": 2.6265,
+      "step": 6145
+    },
+    {
+      "epoch": 0.4960051650391413,
+      "grad_norm": 0.7187899947166443,
+      "learning_rate": 0.00015780365588344384,
+      "loss": 2.5639,
+      "step": 6146
+    },
+    {
+      "epoch": 0.4960858687757243,
+      "grad_norm": 0.7014074921607971,
+      "learning_rate": 0.00015779077288304716,
+      "loss": 2.6011,
+      "step": 6147
+    },
+    {
+      "epoch": 0.49616657251230734,
+      "grad_norm": 0.7463840842247009,
+      "learning_rate": 0.00015777788844235335,
+      "loss": 2.6059,
+      "step": 6148
+    },
+    {
+      "epoch": 0.4962472762488903,
+      "grad_norm": 0.8022417426109314,
+      "learning_rate": 0.00015776500256168356,
+      "loss": 2.6011,
+      "step": 6149
+    },
+    {
+      "epoch": 0.49632797998547334,
+      "grad_norm": 0.7140083909034729,
+      "learning_rate": 0.0001577521152413589,
+      "loss": 2.6891,
+      "step": 6150
+    },
+    {
+      "epoch": 0.4964086837220563,
+      "grad_norm": 0.7266198992729187,
+      "learning_rate": 0.00015773922648170053,
+      "loss": 2.6561,
+      "step": 6151
+    },
+    {
+      "epoch": 0.49648938745863935,
+      "grad_norm": 0.7241406440734863,
+      "learning_rate": 0.0001577263362830297,
+      "loss": 2.6835,
+      "step": 6152
+    },
+    {
+      "epoch": 0.49657009119522233,
+      "grad_norm": 0.7422344088554382,
+      "learning_rate": 0.0001577134446456677,
+      "loss": 2.6039,
+      "step": 6153
+    },
+    {
+      "epoch": 0.49665079493180536,
+      "grad_norm": 0.8764764666557312,
+      "learning_rate": 0.0001577005515699358,
+      "loss": 2.68,
+      "step": 6154
+    },
+    {
+      "epoch": 0.49673149866838834,
+      "grad_norm": 0.7224323749542236,
+      "learning_rate": 0.0001576876570561553,
+      "loss": 2.5824,
+      "step": 6155
+    },
+    {
+      "epoch": 0.49681220240497137,
+      "grad_norm": 0.7601075172424316,
+      "learning_rate": 0.00015767476110464758,
+      "loss": 2.7124,
+      "step": 6156
+    },
+    {
+      "epoch": 0.49689290614155435,
+      "grad_norm": 0.7425428628921509,
+      "learning_rate": 0.0001576618637157341,
+      "loss": 2.5913,
+      "step": 6157
+    },
+    {
+      "epoch": 0.4969736098781374,
+      "grad_norm": 0.721969723701477,
+      "learning_rate": 0.0001576489648897362,
+      "loss": 2.6482,
+      "step": 6158
+    },
+    {
+      "epoch": 0.49705431361472036,
+      "grad_norm": 0.8142126798629761,
+      "learning_rate": 0.00015763606462697544,
+      "loss": 2.6231,
+      "step": 6159
+    },
+    {
+      "epoch": 0.4971350173513034,
+      "grad_norm": 0.6636359691619873,
+      "learning_rate": 0.00015762316292777326,
+      "loss": 2.6388,
+      "step": 6160
+    },
+    {
+      "epoch": 0.49721572108788636,
+      "grad_norm": 0.7093132734298706,
+      "learning_rate": 0.00015761025979245123,
+      "loss": 2.6562,
+      "step": 6161
+    },
+    {
+      "epoch": 0.4972964248244694,
+      "grad_norm": 0.7130851745605469,
+      "learning_rate": 0.00015759735522133094,
+      "loss": 2.6856,
+      "step": 6162
+    },
+    {
+      "epoch": 0.4973771285610524,
+      "grad_norm": 0.7303292155265808,
+      "learning_rate": 0.000157584449214734,
+      "loss": 2.6077,
+      "step": 6163
+    },
+    {
+      "epoch": 0.4974578322976354,
+      "grad_norm": 0.6742258071899414,
+      "learning_rate": 0.00015757154177298204,
+      "loss": 2.6644,
+      "step": 6164
+    },
+    {
+      "epoch": 0.4975385360342184,
+      "grad_norm": 0.6882894039154053,
+      "learning_rate": 0.00015755863289639677,
+      "loss": 2.6462,
+      "step": 6165
+    },
+    {
+      "epoch": 0.4976192397708014,
+      "grad_norm": 0.7882276773452759,
+      "learning_rate": 0.00015754572258529993,
+      "loss": 2.6509,
+      "step": 6166
+    },
+    {
+      "epoch": 0.4976999435073844,
+      "grad_norm": 0.7163859009742737,
+      "learning_rate": 0.00015753281084001324,
+      "loss": 2.627,
+      "step": 6167
+    },
+    {
+      "epoch": 0.4977806472439674,
+      "grad_norm": 0.7194411158561707,
+      "learning_rate": 0.0001575198976608585,
+      "loss": 2.6798,
+      "step": 6168
+    },
+    {
+      "epoch": 0.4978613509805504,
+      "grad_norm": 0.7233198881149292,
+      "learning_rate": 0.0001575069830481576,
+      "loss": 2.6616,
+      "step": 6169
+    },
+    {
+      "epoch": 0.49794205471713343,
+      "grad_norm": 0.7246997952461243,
+      "learning_rate": 0.00015749406700223231,
+      "loss": 2.6262,
+      "step": 6170
+    },
+    {
+      "epoch": 0.4980227584537164,
+      "grad_norm": 0.7509368658065796,
+      "learning_rate": 0.00015748114952340457,
+      "loss": 2.6148,
+      "step": 6171
+    },
+    {
+      "epoch": 0.4981034621902994,
+      "grad_norm": 0.7079075574874878,
+      "learning_rate": 0.00015746823061199637,
+      "loss": 2.6712,
+      "step": 6172
+    },
+    {
+      "epoch": 0.4981841659268824,
+      "grad_norm": 0.6821560859680176,
+      "learning_rate": 0.0001574553102683296,
+      "loss": 2.6253,
+      "step": 6173
+    },
+    {
+      "epoch": 0.4982648696634654,
+      "grad_norm": 0.7623000741004944,
+      "learning_rate": 0.00015744238849272634,
+      "loss": 2.6252,
+      "step": 6174
+    },
+    {
+      "epoch": 0.4983455734000484,
+      "grad_norm": 0.709434449672699,
+      "learning_rate": 0.00015742946528550858,
+      "loss": 2.555,
+      "step": 6175
+    },
+    {
+      "epoch": 0.4984262771366314,
+      "grad_norm": 0.7277799844741821,
+      "learning_rate": 0.00015741654064699846,
+      "loss": 2.6551,
+      "step": 6176
+    },
+    {
+      "epoch": 0.49850698087321443,
+      "grad_norm": 0.7208690643310547,
+      "learning_rate": 0.00015740361457751802,
+      "loss": 2.6747,
+      "step": 6177
+    },
+    {
+      "epoch": 0.4985876846097974,
+      "grad_norm": 0.8458136916160583,
+      "learning_rate": 0.00015739068707738946,
+      "loss": 2.6551,
+      "step": 6178
+    },
+    {
+      "epoch": 0.49866838834638044,
+      "grad_norm": 0.7718539834022522,
+      "learning_rate": 0.00015737775814693498,
+      "loss": 2.6246,
+      "step": 6179
+    },
+    {
+      "epoch": 0.4987490920829634,
+      "grad_norm": 0.6982735395431519,
+      "learning_rate": 0.00015736482778647674,
+      "loss": 2.5726,
+      "step": 6180
+    },
+    {
+      "epoch": 0.49882979581954645,
+      "grad_norm": 0.6759411692619324,
+      "learning_rate": 0.00015735189599633707,
+      "loss": 2.6603,
+      "step": 6181
+    },
+    {
+      "epoch": 0.4989104995561294,
+      "grad_norm": 0.7016656994819641,
+      "learning_rate": 0.0001573389627768382,
+      "loss": 2.6045,
+      "step": 6182
+    },
+    {
+      "epoch": 0.49899120329271246,
+      "grad_norm": 0.7170618176460266,
+      "learning_rate": 0.00015732602812830253,
+      "loss": 2.6419,
+      "step": 6183
+    },
+    {
+      "epoch": 0.49907190702929544,
+      "grad_norm": 0.6963300704956055,
+      "learning_rate": 0.00015731309205105237,
+      "loss": 2.6377,
+      "step": 6184
+    },
+    {
+      "epoch": 0.49915261076587847,
+      "grad_norm": 0.7437995672225952,
+      "learning_rate": 0.00015730015454541014,
+      "loss": 2.7013,
+      "step": 6185
+    },
+    {
+      "epoch": 0.49923331450246144,
+      "grad_norm": 0.6846518516540527,
+      "learning_rate": 0.00015728721561169827,
+      "loss": 2.5526,
+      "step": 6186
+    },
+    {
+      "epoch": 0.4993140182390445,
+      "grad_norm": 0.7343618273735046,
+      "learning_rate": 0.00015727427525023924,
+      "loss": 2.6567,
+      "step": 6187
+    },
+    {
+      "epoch": 0.49939472197562745,
+      "grad_norm": 0.6947566270828247,
+      "learning_rate": 0.00015726133346135554,
+      "loss": 2.6642,
+      "step": 6188
+    },
+    {
+      "epoch": 0.4994754257122105,
+      "grad_norm": 0.7402610778808594,
+      "learning_rate": 0.00015724839024536976,
+      "loss": 2.6964,
+      "step": 6189
+    },
+    {
+      "epoch": 0.49955612944879346,
+      "grad_norm": 0.7318306565284729,
+      "learning_rate": 0.00015723544560260444,
+      "loss": 2.5864,
+      "step": 6190
+    },
+    {
+      "epoch": 0.4996368331853765,
+      "grad_norm": 0.752216100692749,
+      "learning_rate": 0.00015722249953338215,
+      "loss": 2.6357,
+      "step": 6191
+    },
+    {
+      "epoch": 0.49971753692195947,
+      "grad_norm": 0.70283442735672,
+      "learning_rate": 0.00015720955203802565,
+      "loss": 2.5892,
+      "step": 6192
+    },
+    {
+      "epoch": 0.4997982406585425,
+      "grad_norm": 0.7457823753356934,
+      "learning_rate": 0.00015719660311685755,
+      "loss": 2.6663,
+      "step": 6193
+    },
+    {
+      "epoch": 0.4998789443951255,
+      "grad_norm": 0.7296229600906372,
+      "learning_rate": 0.00015718365277020058,
+      "loss": 2.6238,
+      "step": 6194
+    },
+    {
+      "epoch": 0.4999596481317085,
+      "grad_norm": 0.6963346004486084,
+      "learning_rate": 0.0001571707009983775,
+      "loss": 2.6303,
+      "step": 6195
+    },
+    {
+      "epoch": 0.5000403518682915,
+      "grad_norm": 0.7074694633483887,
+      "learning_rate": 0.0001571577478017111,
+      "loss": 2.6077,
+      "step": 6196
+    },
+    {
+      "epoch": 0.5001210556048745,
+      "grad_norm": 0.7826260328292847,
+      "learning_rate": 0.00015714479318052423,
+      "loss": 2.6668,
+      "step": 6197
+    },
+    {
+      "epoch": 0.5002017593414575,
+      "grad_norm": 0.6908758282661438,
+      "learning_rate": 0.00015713183713513974,
+      "loss": 2.6195,
+      "step": 6198
+    },
+    {
+      "epoch": 0.5002824630780405,
+      "grad_norm": 0.7571602463722229,
+      "learning_rate": 0.0001571188796658805,
+      "loss": 2.6546,
+      "step": 6199
+    },
+    {
+      "epoch": 0.5003631668146236,
+      "grad_norm": 0.7359431385993958,
+      "learning_rate": 0.0001571059207730695,
+      "loss": 2.5792,
+      "step": 6200
+    },
+    {
+      "epoch": 0.5004438705512065,
+      "grad_norm": 0.6886340379714966,
+      "learning_rate": 0.00015709296045702967,
+      "loss": 2.6099,
+      "step": 6201
+    },
+    {
+      "epoch": 0.5005245742877895,
+      "grad_norm": 0.6900473833084106,
+      "learning_rate": 0.000157079998718084,
+      "loss": 2.6461,
+      "step": 6202
+    },
+    {
+      "epoch": 0.5006052780243725,
+      "grad_norm": 0.66212397813797,
+      "learning_rate": 0.00015706703555655555,
+      "loss": 2.6178,
+      "step": 6203
+    },
+    {
+      "epoch": 0.5006859817609556,
+      "grad_norm": 0.7666565179824829,
+      "learning_rate": 0.00015705407097276744,
+      "loss": 2.7097,
+      "step": 6204
+    },
+    {
+      "epoch": 0.5007666854975386,
+      "grad_norm": 0.7294591069221497,
+      "learning_rate": 0.0001570411049670427,
+      "loss": 2.5995,
+      "step": 6205
+    },
+    {
+      "epoch": 0.5008473892341215,
+      "grad_norm": 0.7279765009880066,
+      "learning_rate": 0.00015702813753970453,
+      "loss": 2.5554,
+      "step": 6206
+    },
+    {
+      "epoch": 0.5009280929707045,
+      "grad_norm": 0.7174742817878723,
+      "learning_rate": 0.0001570151686910761,
+      "loss": 2.6523,
+      "step": 6207
+    },
+    {
+      "epoch": 0.5010087967072876,
+      "grad_norm": 0.67017662525177,
+      "learning_rate": 0.00015700219842148063,
+      "loss": 2.5613,
+      "step": 6208
+    },
+    {
+      "epoch": 0.5010895004438706,
+      "grad_norm": 0.7000258564949036,
+      "learning_rate": 0.00015698922673124138,
+      "loss": 2.5658,
+      "step": 6209
+    },
+    {
+      "epoch": 0.5011702041804535,
+      "grad_norm": 0.6894544363021851,
+      "learning_rate": 0.00015697625362068164,
+      "loss": 2.6925,
+      "step": 6210
+    },
+    {
+      "epoch": 0.5012509079170365,
+      "grad_norm": 0.6742957234382629,
+      "learning_rate": 0.00015696327909012466,
+      "loss": 2.6429,
+      "step": 6211
+    },
+    {
+      "epoch": 0.5013316116536196,
+      "grad_norm": 0.7039656639099121,
+      "learning_rate": 0.0001569503031398939,
+      "loss": 2.6313,
+      "step": 6212
+    },
+    {
+      "epoch": 0.5014123153902026,
+      "grad_norm": 0.720003604888916,
+      "learning_rate": 0.00015693732577031272,
+      "loss": 2.6207,
+      "step": 6213
+    },
+    {
+      "epoch": 0.5014930191267856,
+      "grad_norm": 0.8611499071121216,
+      "learning_rate": 0.00015692434698170456,
+      "loss": 2.6855,
+      "step": 6214
+    },
+    {
+      "epoch": 0.5015737228633685,
+      "grad_norm": 0.6664702296257019,
+      "learning_rate": 0.00015691136677439284,
+      "loss": 2.6174,
+      "step": 6215
+    },
+    {
+      "epoch": 0.5016544265999516,
+      "grad_norm": 0.7258509993553162,
+      "learning_rate": 0.00015689838514870111,
+      "loss": 2.6558,
+      "step": 6216
+    },
+    {
+      "epoch": 0.5017351303365346,
+      "grad_norm": 0.6972211599349976,
+      "learning_rate": 0.0001568854021049529,
+      "loss": 2.5913,
+      "step": 6217
+    },
+    {
+      "epoch": 0.5018158340731176,
+      "grad_norm": 0.7927280068397522,
+      "learning_rate": 0.00015687241764347177,
+      "loss": 2.6466,
+      "step": 6218
+    },
+    {
+      "epoch": 0.5018965378097006,
+      "grad_norm": 0.7044646143913269,
+      "learning_rate": 0.00015685943176458128,
+      "loss": 2.6195,
+      "step": 6219
+    },
+    {
+      "epoch": 0.5019772415462836,
+      "grad_norm": 0.6935598254203796,
+      "learning_rate": 0.00015684644446860516,
+      "loss": 2.6486,
+      "step": 6220
+    },
+    {
+      "epoch": 0.5020579452828666,
+      "grad_norm": 0.7965792417526245,
+      "learning_rate": 0.00015683345575586704,
+      "loss": 2.6265,
+      "step": 6221
+    },
+    {
+      "epoch": 0.5021386490194496,
+      "grad_norm": 0.727053701877594,
+      "learning_rate": 0.00015682046562669064,
+      "loss": 2.6714,
+      "step": 6222
+    },
+    {
+      "epoch": 0.5022193527560326,
+      "grad_norm": 0.7919184565544128,
+      "learning_rate": 0.0001568074740813997,
+      "loss": 2.7115,
+      "step": 6223
+    },
+    {
+      "epoch": 0.5023000564926156,
+      "grad_norm": 0.7724714279174805,
+      "learning_rate": 0.00015679448112031801,
+      "loss": 2.6636,
+      "step": 6224
+    },
+    {
+      "epoch": 0.5023807602291986,
+      "grad_norm": 0.6893701553344727,
+      "learning_rate": 0.0001567814867437694,
+      "loss": 2.6562,
+      "step": 6225
+    },
+    {
+      "epoch": 0.5024614639657816,
+      "grad_norm": 0.7089633345603943,
+      "learning_rate": 0.00015676849095207769,
+      "loss": 2.6125,
+      "step": 6226
+    },
+    {
+      "epoch": 0.5025421677023646,
+      "grad_norm": 0.7620012760162354,
+      "learning_rate": 0.00015675549374556682,
+      "loss": 2.6935,
+      "step": 6227
+    },
+    {
+      "epoch": 0.5026228714389476,
+      "grad_norm": 0.7293741703033447,
+      "learning_rate": 0.00015674249512456065,
+      "loss": 2.66,
+      "step": 6228
+    },
+    {
+      "epoch": 0.5027035751755307,
+      "grad_norm": 0.7366519570350647,
+      "learning_rate": 0.00015672949508938318,
+      "loss": 2.5968,
+      "step": 6229
+    },
+    {
+      "epoch": 0.5027842789121136,
+      "grad_norm": 0.6646310091018677,
+      "learning_rate": 0.00015671649364035846,
+      "loss": 2.5751,
+      "step": 6230
+    },
+    {
+      "epoch": 0.5028649826486966,
+      "grad_norm": 0.6682632565498352,
+      "learning_rate": 0.00015670349077781038,
+      "loss": 2.5902,
+      "step": 6231
+    },
+    {
+      "epoch": 0.5029456863852796,
+      "grad_norm": 0.7327528595924377,
+      "learning_rate": 0.00015669048650206313,
+      "loss": 2.6487,
+      "step": 6232
+    },
+    {
+      "epoch": 0.5030263901218627,
+      "grad_norm": 0.7114281058311462,
+      "learning_rate": 0.00015667748081344074,
+      "loss": 2.5779,
+      "step": 6233
+    },
+    {
+      "epoch": 0.5031070938584457,
+      "grad_norm": 0.7908105850219727,
+      "learning_rate": 0.00015666447371226737,
+      "loss": 2.6099,
+      "step": 6234
+    },
+    {
+      "epoch": 0.5031877975950286,
+      "grad_norm": 0.7823575139045715,
+      "learning_rate": 0.00015665146519886725,
+      "loss": 2.6339,
+      "step": 6235
+    },
+    {
+      "epoch": 0.5032685013316116,
+      "grad_norm": 0.7404836416244507,
+      "learning_rate": 0.00015663845527356447,
+      "loss": 2.6035,
+      "step": 6236
+    },
+    {
+      "epoch": 0.5033492050681947,
+      "grad_norm": 0.7448995113372803,
+      "learning_rate": 0.00015662544393668334,
+      "loss": 2.6566,
+      "step": 6237
+    },
+    {
+      "epoch": 0.5034299088047777,
+      "grad_norm": 0.7209747433662415,
+      "learning_rate": 0.00015661243118854815,
+      "loss": 2.682,
+      "step": 6238
+    },
+    {
+      "epoch": 0.5035106125413606,
+      "grad_norm": 0.691759467124939,
+      "learning_rate": 0.00015659941702948315,
+      "loss": 2.6435,
+      "step": 6239
+    },
+    {
+      "epoch": 0.5035913162779436,
+      "grad_norm": 0.7646063566207886,
+      "learning_rate": 0.00015658640145981275,
+      "loss": 2.591,
+      "step": 6240
+    },
+    {
+      "epoch": 0.5036720200145267,
+      "grad_norm": 0.8319387435913086,
+      "learning_rate": 0.00015657338447986133,
+      "loss": 2.5937,
+      "step": 6241
+    },
+    {
+      "epoch": 0.5037527237511097,
+      "grad_norm": 0.729193389415741,
+      "learning_rate": 0.00015656036608995323,
+      "loss": 2.651,
+      "step": 6242
+    },
+    {
+      "epoch": 0.5038334274876927,
+      "grad_norm": 0.720098614692688,
+      "learning_rate": 0.000156547346290413,
+      "loss": 2.681,
+      "step": 6243
+    },
+    {
+      "epoch": 0.5039141312242756,
+      "grad_norm": 0.7172541618347168,
+      "learning_rate": 0.00015653432508156508,
+      "loss": 2.5906,
+      "step": 6244
+    },
+    {
+      "epoch": 0.5039948349608587,
+      "grad_norm": 0.7352481484413147,
+      "learning_rate": 0.00015652130246373398,
+      "loss": 2.6376,
+      "step": 6245
+    },
+    {
+      "epoch": 0.5040755386974417,
+      "grad_norm": 0.6664925813674927,
+      "learning_rate": 0.0001565082784372443,
+      "loss": 2.706,
+      "step": 6246
+    },
+    {
+      "epoch": 0.5041562424340247,
+      "grad_norm": 0.7292987704277039,
+      "learning_rate": 0.0001564952530024206,
+      "loss": 2.6149,
+      "step": 6247
+    },
+    {
+      "epoch": 0.5042369461706077,
+      "grad_norm": 0.6904531121253967,
+      "learning_rate": 0.00015648222615958747,
+      "loss": 2.579,
+      "step": 6248
+    },
+    {
+      "epoch": 0.5043176499071907,
+      "grad_norm": 0.7385311722755432,
+      "learning_rate": 0.00015646919790906965,
+      "loss": 2.6137,
+      "step": 6249
+    },
+    {
+      "epoch": 0.5043983536437737,
+      "grad_norm": 0.7869507074356079,
+      "learning_rate": 0.0001564561682511918,
+      "loss": 2.6831,
+      "step": 6250
+    },
+    {
+      "epoch": 0.5044790573803567,
+      "grad_norm": 0.723680317401886,
+      "learning_rate": 0.00015644313718627867,
+      "loss": 2.6083,
+      "step": 6251
+    },
+    {
+      "epoch": 0.5045597611169397,
+      "grad_norm": 0.7029969692230225,
+      "learning_rate": 0.00015643010471465502,
+      "loss": 2.6462,
+      "step": 6252
+    },
+    {
+      "epoch": 0.5046404648535228,
+      "grad_norm": 0.818975031375885,
+      "learning_rate": 0.00015641707083664566,
+      "loss": 2.6393,
+      "step": 6253
+    },
+    {
+      "epoch": 0.5047211685901057,
+      "grad_norm": 0.7237667441368103,
+      "learning_rate": 0.0001564040355525754,
+      "loss": 2.5995,
+      "step": 6254
+    },
+    {
+      "epoch": 0.5048018723266887,
+      "grad_norm": 0.8613824248313904,
+      "learning_rate": 0.00015639099886276912,
+      "loss": 2.748,
+      "step": 6255
+    },
+    {
+      "epoch": 0.5048825760632717,
+      "grad_norm": 0.6802194118499756,
+      "learning_rate": 0.00015637796076755178,
+      "loss": 2.6393,
+      "step": 6256
+    },
+    {
+      "epoch": 0.5049632797998548,
+      "grad_norm": 0.7816255688667297,
+      "learning_rate": 0.00015636492126724823,
+      "loss": 2.6218,
+      "step": 6257
+    },
+    {
+      "epoch": 0.5050439835364378,
+      "grad_norm": 0.7443990707397461,
+      "learning_rate": 0.00015635188036218356,
+      "loss": 2.6181,
+      "step": 6258
+    },
+    {
+      "epoch": 0.5051246872730207,
+      "grad_norm": 0.7869458794593811,
+      "learning_rate": 0.0001563388380526827,
+      "loss": 2.6641,
+      "step": 6259
+    },
+    {
+      "epoch": 0.5052053910096037,
+      "grad_norm": 0.7423158288002014,
+      "learning_rate": 0.00015632579433907072,
+      "loss": 2.5849,
+      "step": 6260
+    },
+    {
+      "epoch": 0.5052860947461868,
+      "grad_norm": 0.7888280153274536,
+      "learning_rate": 0.00015631274922167272,
+      "loss": 2.7095,
+      "step": 6261
+    },
+    {
+      "epoch": 0.5053667984827698,
+      "grad_norm": 0.7053405046463013,
+      "learning_rate": 0.0001562997027008138,
+      "loss": 2.5747,
+      "step": 6262
+    },
+    {
+      "epoch": 0.5054475022193528,
+      "grad_norm": 0.7930825352668762,
+      "learning_rate": 0.0001562866547768191,
+      "loss": 2.6359,
+      "step": 6263
+    },
+    {
+      "epoch": 0.5055282059559357,
+      "grad_norm": 0.7431469559669495,
+      "learning_rate": 0.0001562736054500139,
+      "loss": 2.6167,
+      "step": 6264
+    },
+    {
+      "epoch": 0.5056089096925188,
+      "grad_norm": 0.8395694494247437,
+      "learning_rate": 0.00015626055472072324,
+      "loss": 2.7217,
+      "step": 6265
+    },
+    {
+      "epoch": 0.5056896134291018,
+      "grad_norm": 0.7318898439407349,
+      "learning_rate": 0.0001562475025892726,
+      "loss": 2.6866,
+      "step": 6266
+    },
+    {
+      "epoch": 0.5057703171656848,
+      "grad_norm": 0.7487025856971741,
+      "learning_rate": 0.0001562344490559871,
+      "loss": 2.7206,
+      "step": 6267
+    },
+    {
+      "epoch": 0.5058510209022677,
+      "grad_norm": 0.8187269568443298,
+      "learning_rate": 0.00015622139412119212,
+      "loss": 2.658,
+      "step": 6268
+    },
+    {
+      "epoch": 0.5059317246388508,
+      "grad_norm": 0.6714495420455933,
+      "learning_rate": 0.00015620833778521307,
+      "loss": 2.6182,
+      "step": 6269
+    },
+    {
+      "epoch": 0.5060124283754338,
+      "grad_norm": 0.7556246519088745,
+      "learning_rate": 0.00015619528004837528,
+      "loss": 2.6502,
+      "step": 6270
+    },
+    {
+      "epoch": 0.5060931321120168,
+      "grad_norm": 0.6989960074424744,
+      "learning_rate": 0.00015618222091100424,
+      "loss": 2.6031,
+      "step": 6271
+    },
+    {
+      "epoch": 0.5061738358485998,
+      "grad_norm": 0.7002139091491699,
+      "learning_rate": 0.0001561691603734254,
+      "loss": 2.6563,
+      "step": 6272
+    },
+    {
+      "epoch": 0.5062545395851827,
+      "grad_norm": 0.7064816355705261,
+      "learning_rate": 0.00015615609843596423,
+      "loss": 2.6482,
+      "step": 6273
+    },
+    {
+      "epoch": 0.5063352433217658,
+      "grad_norm": 0.6971433162689209,
+      "learning_rate": 0.00015614303509894634,
+      "loss": 2.6522,
+      "step": 6274
+    },
+    {
+      "epoch": 0.5064159470583488,
+      "grad_norm": 0.6982942223548889,
+      "learning_rate": 0.0001561299703626972,
+      "loss": 2.6477,
+      "step": 6275
+    },
+    {
+      "epoch": 0.5064966507949318,
+      "grad_norm": 0.7219811081886292,
+      "learning_rate": 0.0001561169042275425,
+      "loss": 2.6514,
+      "step": 6276
+    },
+    {
+      "epoch": 0.5065773545315148,
+      "grad_norm": 0.7391932010650635,
+      "learning_rate": 0.00015610383669380787,
+      "loss": 2.698,
+      "step": 6277
+    },
+    {
+      "epoch": 0.5066580582680978,
+      "grad_norm": 0.7852853536605835,
+      "learning_rate": 0.00015609076776181894,
+      "loss": 2.6281,
+      "step": 6278
+    },
+    {
+      "epoch": 0.5067387620046808,
+      "grad_norm": 0.7435647249221802,
+      "learning_rate": 0.00015607769743190147,
+      "loss": 2.6403,
+      "step": 6279
+    },
+    {
+      "epoch": 0.5068194657412638,
+      "grad_norm": 0.7300949096679688,
+      "learning_rate": 0.00015606462570438119,
+      "loss": 2.6125,
+      "step": 6280
+    },
+    {
+      "epoch": 0.5069001694778468,
+      "grad_norm": 0.7081549167633057,
+      "learning_rate": 0.00015605155257958388,
+      "loss": 2.6192,
+      "step": 6281
+    },
+    {
+      "epoch": 0.5069808732144299,
+      "grad_norm": 0.709020733833313,
+      "learning_rate": 0.00015603847805783537,
+      "loss": 2.6745,
+      "step": 6282
+    },
+    {
+      "epoch": 0.5070615769510128,
+      "grad_norm": 0.691684901714325,
+      "learning_rate": 0.0001560254021394615,
+      "loss": 2.5638,
+      "step": 6283
+    },
+    {
+      "epoch": 0.5071422806875958,
+      "grad_norm": 0.8338537812232971,
+      "learning_rate": 0.00015601232482478813,
+      "loss": 2.5835,
+      "step": 6284
+    },
+    {
+      "epoch": 0.5072229844241788,
+      "grad_norm": 0.659436047077179,
+      "learning_rate": 0.00015599924611414126,
+      "loss": 2.601,
+      "step": 6285
+    },
+    {
+      "epoch": 0.5073036881607619,
+      "grad_norm": 0.72590172290802,
+      "learning_rate": 0.00015598616600784676,
+      "loss": 2.602,
+      "step": 6286
+    },
+    {
+      "epoch": 0.5073843918973449,
+      "grad_norm": 0.6704443693161011,
+      "learning_rate": 0.00015597308450623066,
+      "loss": 2.5703,
+      "step": 6287
+    },
+    {
+      "epoch": 0.5074650956339278,
+      "grad_norm": 0.7298632264137268,
+      "learning_rate": 0.00015596000160961898,
+      "loss": 2.6859,
+      "step": 6288
+    },
+    {
+      "epoch": 0.5075457993705108,
+      "grad_norm": 0.6900345087051392,
+      "learning_rate": 0.00015594691731833776,
+      "loss": 2.6264,
+      "step": 6289
+    },
+    {
+      "epoch": 0.5076265031070939,
+      "grad_norm": 0.6705992221832275,
+      "learning_rate": 0.0001559338316327131,
+      "loss": 2.6135,
+      "step": 6290
+    },
+    {
+      "epoch": 0.5077072068436769,
+      "grad_norm": 0.691545307636261,
+      "learning_rate": 0.0001559207445530712,
+      "loss": 2.6538,
+      "step": 6291
+    },
+    {
+      "epoch": 0.5077879105802598,
+      "grad_norm": 0.6579985618591309,
+      "learning_rate": 0.00015590765607973811,
+      "loss": 2.6224,
+      "step": 6292
+    },
+    {
+      "epoch": 0.5078686143168428,
+      "grad_norm": 0.6938790678977966,
+      "learning_rate": 0.00015589456621304014,
+      "loss": 2.5932,
+      "step": 6293
+    },
+    {
+      "epoch": 0.5079493180534259,
+      "grad_norm": 0.7421671748161316,
+      "learning_rate": 0.00015588147495330346,
+      "loss": 2.7098,
+      "step": 6294
+    },
+    {
+      "epoch": 0.5080300217900089,
+      "grad_norm": 0.7076674699783325,
+      "learning_rate": 0.0001558683823008543,
+      "loss": 2.664,
+      "step": 6295
+    },
+    {
+      "epoch": 0.5081107255265919,
+      "grad_norm": 0.6829726696014404,
+      "learning_rate": 0.00015585528825601906,
+      "loss": 2.6029,
+      "step": 6296
+    },
+    {
+      "epoch": 0.5081914292631748,
+      "grad_norm": 0.6968080401420593,
+      "learning_rate": 0.000155842192819124,
+      "loss": 2.6256,
+      "step": 6297
+    },
+    {
+      "epoch": 0.5082721329997579,
+      "grad_norm": 0.7453410625457764,
+      "learning_rate": 0.00015582909599049554,
+      "loss": 2.6577,
+      "step": 6298
+    },
+    {
+      "epoch": 0.5083528367363409,
+      "grad_norm": 0.6603519916534424,
+      "learning_rate": 0.00015581599777046007,
+      "loss": 2.6066,
+      "step": 6299
+    },
+    {
+      "epoch": 0.5084335404729239,
+      "grad_norm": 0.7096173763275146,
+      "learning_rate": 0.00015580289815934401,
+      "loss": 2.5488,
+      "step": 6300
+    },
+    {
+      "epoch": 0.5085142442095069,
+      "grad_norm": 0.799298107624054,
+      "learning_rate": 0.0001557897971574739,
+      "loss": 2.6021,
+      "step": 6301
+    },
+    {
+      "epoch": 0.50859494794609,
+      "grad_norm": 0.6820314526557922,
+      "learning_rate": 0.00015577669476517618,
+      "loss": 2.6276,
+      "step": 6302
+    },
+    {
+      "epoch": 0.5086756516826729,
+      "grad_norm": 0.7119347453117371,
+      "learning_rate": 0.00015576359098277742,
+      "loss": 2.6627,
+      "step": 6303
+    },
+    {
+      "epoch": 0.5087563554192559,
+      "grad_norm": 0.7638720273971558,
+      "learning_rate": 0.00015575048581060422,
+      "loss": 2.6824,
+      "step": 6304
+    },
+    {
+      "epoch": 0.5088370591558389,
+      "grad_norm": 0.7360339164733887,
+      "learning_rate": 0.00015573737924898316,
+      "loss": 2.5805,
+      "step": 6305
+    },
+    {
+      "epoch": 0.508917762892422,
+      "grad_norm": 0.7220984697341919,
+      "learning_rate": 0.00015572427129824091,
+      "loss": 2.6374,
+      "step": 6306
+    },
+    {
+      "epoch": 0.5089984666290049,
+      "grad_norm": 0.670964777469635,
+      "learning_rate": 0.00015571116195870418,
+      "loss": 2.6371,
+      "step": 6307
+    },
+    {
+      "epoch": 0.5090791703655879,
+      "grad_norm": 0.7826075553894043,
+      "learning_rate": 0.00015569805123069968,
+      "loss": 2.7666,
+      "step": 6308
+    },
+    {
+      "epoch": 0.5091598741021709,
+      "grad_norm": 0.7691593766212463,
+      "learning_rate": 0.00015568493911455412,
+      "loss": 2.6242,
+      "step": 6309
+    },
+    {
+      "epoch": 0.509240577838754,
+      "grad_norm": 0.714500367641449,
+      "learning_rate": 0.0001556718256105943,
+      "loss": 2.6551,
+      "step": 6310
+    },
+    {
+      "epoch": 0.509321281575337,
+      "grad_norm": 0.7634009718894958,
+      "learning_rate": 0.00015565871071914706,
+      "loss": 2.7069,
+      "step": 6311
+    },
+    {
+      "epoch": 0.5094019853119199,
+      "grad_norm": 0.7134168148040771,
+      "learning_rate": 0.00015564559444053926,
+      "loss": 2.5816,
+      "step": 6312
+    },
+    {
+      "epoch": 0.5094826890485029,
+      "grad_norm": 0.6548121571540833,
+      "learning_rate": 0.0001556324767750978,
+      "loss": 2.6192,
+      "step": 6313
+    },
+    {
+      "epoch": 0.509563392785086,
+      "grad_norm": 0.7244428992271423,
+      "learning_rate": 0.0001556193577231496,
+      "loss": 2.6072,
+      "step": 6314
+    },
+    {
+      "epoch": 0.509644096521669,
+      "grad_norm": 0.6976662278175354,
+      "learning_rate": 0.0001556062372850216,
+      "loss": 2.6148,
+      "step": 6315
+    },
+    {
+      "epoch": 0.509724800258252,
+      "grad_norm": 0.772726833820343,
+      "learning_rate": 0.00015559311546104083,
+      "loss": 2.6458,
+      "step": 6316
+    },
+    {
+      "epoch": 0.5098055039948349,
+      "grad_norm": 0.7976188659667969,
+      "learning_rate": 0.00015557999225153428,
+      "loss": 2.6772,
+      "step": 6317
+    },
+    {
+      "epoch": 0.509886207731418,
+      "grad_norm": 0.6458039283752441,
+      "learning_rate": 0.00015556686765682903,
+      "loss": 2.6143,
+      "step": 6318
+    },
+    {
+      "epoch": 0.509966911468001,
+      "grad_norm": 0.7295405268669128,
+      "learning_rate": 0.0001555537416772522,
+      "loss": 2.5919,
+      "step": 6319
+    },
+    {
+      "epoch": 0.510047615204584,
+      "grad_norm": 0.657978355884552,
+      "learning_rate": 0.00015554061431313093,
+      "loss": 2.6245,
+      "step": 6320
+    },
+    {
+      "epoch": 0.510128318941167,
+      "grad_norm": 0.6726922392845154,
+      "learning_rate": 0.00015552748556479232,
+      "loss": 2.6207,
+      "step": 6321
+    },
+    {
+      "epoch": 0.51020902267775,
+      "grad_norm": 0.7954673767089844,
+      "learning_rate": 0.00015551435543256363,
+      "loss": 2.7177,
+      "step": 6322
+    },
+    {
+      "epoch": 0.510289726414333,
+      "grad_norm": 0.7186735272407532,
+      "learning_rate": 0.00015550122391677211,
+      "loss": 2.5953,
+      "step": 6323
+    },
+    {
+      "epoch": 0.510370430150916,
+      "grad_norm": 0.7835420966148376,
+      "learning_rate": 0.00015548809101774498,
+      "loss": 2.7039,
+      "step": 6324
+    },
+    {
+      "epoch": 0.510451133887499,
+      "grad_norm": 0.6966592073440552,
+      "learning_rate": 0.00015547495673580962,
+      "loss": 2.6287,
+      "step": 6325
+    },
+    {
+      "epoch": 0.5105318376240819,
+      "grad_norm": 0.6676180362701416,
+      "learning_rate": 0.00015546182107129328,
+      "loss": 2.638,
+      "step": 6326
+    },
+    {
+      "epoch": 0.510612541360665,
+      "grad_norm": 0.7285657525062561,
+      "learning_rate": 0.0001554486840245234,
+      "loss": 2.6661,
+      "step": 6327
+    },
+    {
+      "epoch": 0.510693245097248,
+      "grad_norm": 0.6453657150268555,
+      "learning_rate": 0.00015543554559582735,
+      "loss": 2.715,
+      "step": 6328
+    },
+    {
+      "epoch": 0.510773948833831,
+      "grad_norm": 0.7364684343338013,
+      "learning_rate": 0.0001554224057855326,
+      "loss": 2.6475,
+      "step": 6329
+    },
+    {
+      "epoch": 0.510854652570414,
+      "grad_norm": 0.670894980430603,
+      "learning_rate": 0.00015540926459396665,
+      "loss": 2.6091,
+      "step": 6330
+    },
+    {
+      "epoch": 0.510935356306997,
+      "grad_norm": 0.6750168204307556,
+      "learning_rate": 0.00015539612202145696,
+      "loss": 2.6473,
+      "step": 6331
+    },
+    {
+      "epoch": 0.51101606004358,
+      "grad_norm": 0.6552454233169556,
+      "learning_rate": 0.0001553829780683311,
+      "loss": 2.6158,
+      "step": 6332
+    },
+    {
+      "epoch": 0.511096763780163,
+      "grad_norm": 0.7387828230857849,
+      "learning_rate": 0.00015536983273491668,
+      "loss": 2.6219,
+      "step": 6333
+    },
+    {
+      "epoch": 0.511177467516746,
+      "grad_norm": 0.6993975639343262,
+      "learning_rate": 0.00015535668602154127,
+      "loss": 2.6446,
+      "step": 6334
+    },
+    {
+      "epoch": 0.5112581712533291,
+      "grad_norm": 0.6491217613220215,
+      "learning_rate": 0.00015534353792853254,
+      "loss": 2.6404,
+      "step": 6335
+    },
+    {
+      "epoch": 0.511338874989912,
+      "grad_norm": 0.7165521383285522,
+      "learning_rate": 0.0001553303884562182,
+      "loss": 2.6339,
+      "step": 6336
+    },
+    {
+      "epoch": 0.511419578726495,
+      "grad_norm": 0.7363756895065308,
+      "learning_rate": 0.0001553172376049259,
+      "loss": 2.6411,
+      "step": 6337
+    },
+    {
+      "epoch": 0.511500282463078,
+      "grad_norm": 0.7148438096046448,
+      "learning_rate": 0.00015530408537498347,
+      "loss": 2.5617,
+      "step": 6338
+    },
+    {
+      "epoch": 0.5115809861996611,
+      "grad_norm": 0.7140451669692993,
+      "learning_rate": 0.00015529093176671864,
+      "loss": 2.5898,
+      "step": 6339
+    },
+    {
+      "epoch": 0.5116616899362441,
+      "grad_norm": 0.7799252271652222,
+      "learning_rate": 0.00015527777678045926,
+      "loss": 2.6176,
+      "step": 6340
+    },
+    {
+      "epoch": 0.511742393672827,
+      "grad_norm": 0.7292928099632263,
+      "learning_rate": 0.00015526462041653323,
+      "loss": 2.6722,
+      "step": 6341
+    },
+    {
+      "epoch": 0.51182309740941,
+      "grad_norm": 0.6986904740333557,
+      "learning_rate": 0.00015525146267526837,
+      "loss": 2.6154,
+      "step": 6342
+    },
+    {
+      "epoch": 0.5119038011459931,
+      "grad_norm": 0.7239612936973572,
+      "learning_rate": 0.00015523830355699262,
+      "loss": 2.5664,
+      "step": 6343
+    },
+    {
+      "epoch": 0.5119845048825761,
+      "grad_norm": 0.6805121898651123,
+      "learning_rate": 0.00015522514306203395,
+      "loss": 2.6204,
+      "step": 6344
+    },
+    {
+      "epoch": 0.512065208619159,
+      "grad_norm": 0.7036689519882202,
+      "learning_rate": 0.00015521198119072035,
+      "loss": 2.6211,
+      "step": 6345
+    },
+    {
+      "epoch": 0.512145912355742,
+      "grad_norm": 0.7155849933624268,
+      "learning_rate": 0.00015519881794337988,
+      "loss": 2.6074,
+      "step": 6346
+    },
+    {
+      "epoch": 0.5122266160923251,
+      "grad_norm": 0.7183938026428223,
+      "learning_rate": 0.00015518565332034057,
+      "loss": 2.6148,
+      "step": 6347
+    },
+    {
+      "epoch": 0.5123073198289081,
+      "grad_norm": 0.7053570747375488,
+      "learning_rate": 0.0001551724873219305,
+      "loss": 2.6476,
+      "step": 6348
+    },
+    {
+      "epoch": 0.5123880235654911,
+      "grad_norm": 0.714846670627594,
+      "learning_rate": 0.00015515931994847785,
+      "loss": 2.5728,
+      "step": 6349
+    },
+    {
+      "epoch": 0.512468727302074,
+      "grad_norm": 0.7504729628562927,
+      "learning_rate": 0.00015514615120031076,
+      "loss": 2.6415,
+      "step": 6350
+    },
+    {
+      "epoch": 0.5125494310386571,
+      "grad_norm": 0.6940335035324097,
+      "learning_rate": 0.0001551329810777574,
+      "loss": 2.6115,
+      "step": 6351
+    },
+    {
+      "epoch": 0.5126301347752401,
+      "grad_norm": 0.7166119813919067,
+      "learning_rate": 0.00015511980958114608,
+      "loss": 2.6284,
+      "step": 6352
+    },
+    {
+      "epoch": 0.5127108385118231,
+      "grad_norm": 0.7787839770317078,
+      "learning_rate": 0.00015510663671080497,
+      "loss": 2.6385,
+      "step": 6353
+    },
+    {
+      "epoch": 0.5127915422484061,
+      "grad_norm": 0.7298412322998047,
+      "learning_rate": 0.00015509346246706245,
+      "loss": 2.629,
+      "step": 6354
+    },
+    {
+      "epoch": 0.5128722459849892,
+      "grad_norm": 0.7918897271156311,
+      "learning_rate": 0.00015508028685024683,
+      "loss": 2.6777,
+      "step": 6355
+    },
+    {
+      "epoch": 0.5129529497215721,
+      "grad_norm": 0.6867843866348267,
+      "learning_rate": 0.00015506710986068646,
+      "loss": 2.6101,
+      "step": 6356
+    },
+    {
+      "epoch": 0.5130336534581551,
+      "grad_norm": 0.716468870639801,
+      "learning_rate": 0.00015505393149870978,
+      "loss": 2.6558,
+      "step": 6357
+    },
+    {
+      "epoch": 0.5131143571947381,
+      "grad_norm": 0.6704092621803284,
+      "learning_rate": 0.0001550407517646452,
+      "loss": 2.6128,
+      "step": 6358
+    },
+    {
+      "epoch": 0.5131950609313212,
+      "grad_norm": 0.820716381072998,
+      "learning_rate": 0.00015502757065882124,
+      "loss": 2.6052,
+      "step": 6359
+    },
+    {
+      "epoch": 0.5132757646679041,
+      "grad_norm": 0.7328094840049744,
+      "learning_rate": 0.00015501438818156635,
+      "loss": 2.6399,
+      "step": 6360
+    },
+    {
+      "epoch": 0.5133564684044871,
+      "grad_norm": 0.6602808833122253,
+      "learning_rate": 0.00015500120433320911,
+      "loss": 2.5509,
+      "step": 6361
+    },
+    {
+      "epoch": 0.5134371721410701,
+      "grad_norm": 0.7013166546821594,
+      "learning_rate": 0.00015498801911407805,
+      "loss": 2.6439,
+      "step": 6362
+    },
+    {
+      "epoch": 0.5135178758776532,
+      "grad_norm": 0.7415499091148376,
+      "learning_rate": 0.00015497483252450186,
+      "loss": 2.575,
+      "step": 6363
+    },
+    {
+      "epoch": 0.5135985796142362,
+      "grad_norm": 0.7262336015701294,
+      "learning_rate": 0.00015496164456480912,
+      "loss": 2.6815,
+      "step": 6364
+    },
+    {
+      "epoch": 0.5136792833508191,
+      "grad_norm": 0.7353699803352356,
+      "learning_rate": 0.0001549484552353285,
+      "loss": 2.6172,
+      "step": 6365
+    },
+    {
+      "epoch": 0.5137599870874021,
+      "grad_norm": 0.7005086541175842,
+      "learning_rate": 0.00015493526453638879,
+      "loss": 2.5945,
+      "step": 6366
+    },
+    {
+      "epoch": 0.5138406908239852,
+      "grad_norm": 0.7469770908355713,
+      "learning_rate": 0.00015492207246831864,
+      "loss": 2.6797,
+      "step": 6367
+    },
+    {
+      "epoch": 0.5139213945605682,
+      "grad_norm": 0.6768934726715088,
+      "learning_rate": 0.00015490887903144693,
+      "loss": 2.6369,
+      "step": 6368
+    },
+    {
+      "epoch": 0.5140020982971512,
+      "grad_norm": 0.7625820636749268,
+      "learning_rate": 0.00015489568422610237,
+      "loss": 2.6182,
+      "step": 6369
+    },
+    {
+      "epoch": 0.5140828020337341,
+      "grad_norm": 0.749351978302002,
+      "learning_rate": 0.00015488248805261388,
+      "loss": 2.6066,
+      "step": 6370
+    },
+    {
+      "epoch": 0.5141635057703172,
+      "grad_norm": 0.8369480967521667,
+      "learning_rate": 0.00015486929051131032,
+      "loss": 2.7627,
+      "step": 6371
+    },
+    {
+      "epoch": 0.5142442095069002,
+      "grad_norm": 0.6482037305831909,
+      "learning_rate": 0.0001548560916025206,
+      "loss": 2.609,
+      "step": 6372
+    },
+    {
+      "epoch": 0.5143249132434832,
+      "grad_norm": 0.6801851391792297,
+      "learning_rate": 0.0001548428913265737,
+      "loss": 2.5878,
+      "step": 6373
+    },
+    {
+      "epoch": 0.5144056169800661,
+      "grad_norm": 0.744926929473877,
+      "learning_rate": 0.0001548296896837986,
+      "loss": 2.6569,
+      "step": 6374
+    },
+    {
+      "epoch": 0.5144863207166491,
+      "grad_norm": 0.6862614750862122,
+      "learning_rate": 0.00015481648667452425,
+      "loss": 2.5626,
+      "step": 6375
+    },
+    {
+      "epoch": 0.5145670244532322,
+      "grad_norm": 0.7186449766159058,
+      "learning_rate": 0.0001548032822990798,
+      "loss": 2.6783,
+      "step": 6376
+    },
+    {
+      "epoch": 0.5146477281898152,
+      "grad_norm": 0.699715256690979,
+      "learning_rate": 0.0001547900765577943,
+      "loss": 2.6709,
+      "step": 6377
+    },
+    {
+      "epoch": 0.5147284319263982,
+      "grad_norm": 0.7272205352783203,
+      "learning_rate": 0.00015477686945099687,
+      "loss": 2.6076,
+      "step": 6378
+    },
+    {
+      "epoch": 0.5148091356629811,
+      "grad_norm": 0.7667459845542908,
+      "learning_rate": 0.00015476366097901667,
+      "loss": 2.6541,
+      "step": 6379
+    },
+    {
+      "epoch": 0.5148898393995642,
+      "grad_norm": 0.6538121700286865,
+      "learning_rate": 0.00015475045114218285,
+      "loss": 2.5806,
+      "step": 6380
+    },
+    {
+      "epoch": 0.5149705431361472,
+      "grad_norm": 0.7388994097709656,
+      "learning_rate": 0.00015473723994082473,
+      "loss": 2.6293,
+      "step": 6381
+    },
+    {
+      "epoch": 0.5150512468727302,
+      "grad_norm": 0.7044215202331543,
+      "learning_rate": 0.00015472402737527142,
+      "loss": 2.5755,
+      "step": 6382
+    },
+    {
+      "epoch": 0.5151319506093132,
+      "grad_norm": 0.6807994246482849,
+      "learning_rate": 0.00015471081344585236,
+      "loss": 2.6493,
+      "step": 6383
+    },
+    {
+      "epoch": 0.5152126543458962,
+      "grad_norm": 0.676278293132782,
+      "learning_rate": 0.00015469759815289681,
+      "loss": 2.6319,
+      "step": 6384
+    },
+    {
+      "epoch": 0.5152933580824792,
+      "grad_norm": 0.7515453696250916,
+      "learning_rate": 0.00015468438149673412,
+      "loss": 2.6415,
+      "step": 6385
+    },
+    {
+      "epoch": 0.5153740618190622,
+      "grad_norm": 0.8694239854812622,
+      "learning_rate": 0.0001546711634776937,
+      "loss": 2.5818,
+      "step": 6386
+    },
+    {
+      "epoch": 0.5154547655556452,
+      "grad_norm": 0.717090368270874,
+      "learning_rate": 0.000154657944096105,
+      "loss": 2.7132,
+      "step": 6387
+    },
+    {
+      "epoch": 0.5155354692922283,
+      "grad_norm": 0.7098804116249084,
+      "learning_rate": 0.00015464472335229742,
+      "loss": 2.564,
+      "step": 6388
+    },
+    {
+      "epoch": 0.5156161730288112,
+      "grad_norm": 0.6879690289497375,
+      "learning_rate": 0.0001546315012466005,
+      "loss": 2.6094,
+      "step": 6389
+    },
+    {
+      "epoch": 0.5156968767653942,
+      "grad_norm": 0.7110763788223267,
+      "learning_rate": 0.00015461827777934377,
+      "loss": 2.5982,
+      "step": 6390
+    },
+    {
+      "epoch": 0.5157775805019772,
+      "grad_norm": 0.7168039679527283,
+      "learning_rate": 0.00015460505295085677,
+      "loss": 2.5451,
+      "step": 6391
+    },
+    {
+      "epoch": 0.5158582842385603,
+      "grad_norm": 0.7059877514839172,
+      "learning_rate": 0.00015459182676146914,
+      "loss": 2.6655,
+      "step": 6392
+    },
+    {
+      "epoch": 0.5159389879751433,
+      "grad_norm": 0.7278143763542175,
+      "learning_rate": 0.00015457859921151043,
+      "loss": 2.6587,
+      "step": 6393
+    },
+    {
+      "epoch": 0.5160196917117262,
+      "grad_norm": 0.7301023602485657,
+      "learning_rate": 0.0001545653703013104,
+      "loss": 2.7672,
+      "step": 6394
+    },
+    {
+      "epoch": 0.5161003954483092,
+      "grad_norm": 0.6933302283287048,
+      "learning_rate": 0.0001545521400311987,
+      "loss": 2.5924,
+      "step": 6395
+    },
+    {
+      "epoch": 0.5161810991848923,
+      "grad_norm": 0.7074775099754333,
+      "learning_rate": 0.00015453890840150508,
+      "loss": 2.6663,
+      "step": 6396
+    },
+    {
+      "epoch": 0.5162618029214753,
+      "grad_norm": 0.7069801092147827,
+      "learning_rate": 0.00015452567541255924,
+      "loss": 2.6791,
+      "step": 6397
+    },
+    {
+      "epoch": 0.5163425066580583,
+      "grad_norm": 0.6586462259292603,
+      "learning_rate": 0.00015451244106469108,
+      "loss": 2.6368,
+      "step": 6398
+    },
+    {
+      "epoch": 0.5164232103946412,
+      "grad_norm": 0.6862531900405884,
+      "learning_rate": 0.00015449920535823042,
+      "loss": 2.7099,
+      "step": 6399
+    },
+    {
+      "epoch": 0.5165039141312243,
+      "grad_norm": 0.7177795767784119,
+      "learning_rate": 0.00015448596829350706,
+      "loss": 2.5921,
+      "step": 6400
+    },
+    {
+      "epoch": 0.5165846178678073,
+      "grad_norm": 0.6936569213867188,
+      "learning_rate": 0.00015447272987085094,
+      "loss": 2.5739,
+      "step": 6401
+    },
+    {
+      "epoch": 0.5166653216043903,
+      "grad_norm": 0.7394363284111023,
+      "learning_rate": 0.00015445949009059202,
+      "loss": 2.5941,
+      "step": 6402
+    },
+    {
+      "epoch": 0.5167460253409732,
+      "grad_norm": 0.6713366508483887,
+      "learning_rate": 0.00015444624895306027,
+      "loss": 2.574,
+      "step": 6403
+    },
+    {
+      "epoch": 0.5168267290775563,
+      "grad_norm": 0.679128885269165,
+      "learning_rate": 0.0001544330064585856,
+      "loss": 2.6422,
+      "step": 6404
+    },
+    {
+      "epoch": 0.5169074328141393,
+      "grad_norm": 0.6803367137908936,
+      "learning_rate": 0.0001544197626074982,
+      "loss": 2.6503,
+      "step": 6405
+    },
+    {
+      "epoch": 0.5169881365507223,
+      "grad_norm": 0.8009794354438782,
+      "learning_rate": 0.000154406517400128,
+      "loss": 2.6434,
+      "step": 6406
+    },
+    {
+      "epoch": 0.5170688402873053,
+      "grad_norm": 0.7292529344558716,
+      "learning_rate": 0.00015439327083680517,
+      "loss": 2.6333,
+      "step": 6407
+    },
+    {
+      "epoch": 0.5171495440238884,
+      "grad_norm": 0.67046719789505,
+      "learning_rate": 0.00015438002291785988,
+      "loss": 2.5791,
+      "step": 6408
+    },
+    {
+      "epoch": 0.5172302477604713,
+      "grad_norm": 0.755501925945282,
+      "learning_rate": 0.00015436677364362225,
+      "loss": 2.5558,
+      "step": 6409
+    },
+    {
+      "epoch": 0.5173109514970543,
+      "grad_norm": 0.6957115530967712,
+      "learning_rate": 0.0001543535230144225,
+      "loss": 2.5839,
+      "step": 6410
+    },
+    {
+      "epoch": 0.5173916552336373,
+      "grad_norm": 0.6629074215888977,
+      "learning_rate": 0.0001543402710305909,
+      "loss": 2.6529,
+      "step": 6411
+    },
+    {
+      "epoch": 0.5174723589702204,
+      "grad_norm": 0.6647019386291504,
+      "learning_rate": 0.00015432701769245766,
+      "loss": 2.589,
+      "step": 6412
+    },
+    {
+      "epoch": 0.5175530627068033,
+      "grad_norm": 0.6472512483596802,
+      "learning_rate": 0.00015431376300035316,
+      "loss": 2.6184,
+      "step": 6413
+    },
+    {
+      "epoch": 0.5176337664433863,
+      "grad_norm": 0.6900136470794678,
+      "learning_rate": 0.0001543005069546077,
+      "loss": 2.7029,
+      "step": 6414
+    },
+    {
+      "epoch": 0.5177144701799693,
+      "grad_norm": 0.7702177166938782,
+      "learning_rate": 0.00015428724955555165,
+      "loss": 2.6189,
+      "step": 6415
+    },
+    {
+      "epoch": 0.5177951739165524,
+      "grad_norm": 0.641655445098877,
+      "learning_rate": 0.00015427399080351545,
+      "loss": 2.6486,
+      "step": 6416
+    },
+    {
+      "epoch": 0.5178758776531354,
+      "grad_norm": 0.6826485991477966,
+      "learning_rate": 0.00015426073069882952,
+      "loss": 2.6105,
+      "step": 6417
+    },
+    {
+      "epoch": 0.5179565813897183,
+      "grad_norm": 0.749812662601471,
+      "learning_rate": 0.00015424746924182434,
+      "loss": 2.5644,
+      "step": 6418
+    },
+    {
+      "epoch": 0.5180372851263013,
+      "grad_norm": 0.6737890243530273,
+      "learning_rate": 0.0001542342064328304,
+      "loss": 2.686,
+      "step": 6419
+    },
+    {
+      "epoch": 0.5181179888628844,
+      "grad_norm": 0.7131822109222412,
+      "learning_rate": 0.0001542209422721783,
+      "loss": 2.697,
+      "step": 6420
+    },
+    {
+      "epoch": 0.5181986925994674,
+      "grad_norm": 0.7543746829032898,
+      "learning_rate": 0.0001542076767601986,
+      "loss": 2.6349,
+      "step": 6421
+    },
+    {
+      "epoch": 0.5182793963360504,
+      "grad_norm": 0.7589309215545654,
+      "learning_rate": 0.00015419440989722184,
+      "loss": 2.63,
+      "step": 6422
+    },
+    {
+      "epoch": 0.5183601000726333,
+      "grad_norm": 0.7036365866661072,
+      "learning_rate": 0.00015418114168357872,
+      "loss": 2.605,
+      "step": 6423
+    },
+    {
+      "epoch": 0.5184408038092164,
+      "grad_norm": 0.733161985874176,
+      "learning_rate": 0.00015416787211959998,
+      "loss": 2.6708,
+      "step": 6424
+    },
+    {
+      "epoch": 0.5185215075457994,
+      "grad_norm": 0.6928101181983948,
+      "learning_rate": 0.00015415460120561623,
+      "loss": 2.6549,
+      "step": 6425
+    },
+    {
+      "epoch": 0.5186022112823824,
+      "grad_norm": 0.6557250022888184,
+      "learning_rate": 0.00015414132894195825,
+      "loss": 2.6185,
+      "step": 6426
+    },
+    {
+      "epoch": 0.5186829150189654,
+      "grad_norm": 0.7236297726631165,
+      "learning_rate": 0.00015412805532895684,
+      "loss": 2.6185,
+      "step": 6427
+    },
+    {
+      "epoch": 0.5187636187555483,
+      "grad_norm": 0.7194060683250427,
+      "learning_rate": 0.0001541147803669428,
+      "loss": 2.6123,
+      "step": 6428
+    },
+    {
+      "epoch": 0.5188443224921314,
+      "grad_norm": 0.7077342867851257,
+      "learning_rate": 0.00015410150405624696,
+      "loss": 2.6628,
+      "step": 6429
+    },
+    {
+      "epoch": 0.5189250262287144,
+      "grad_norm": 0.7036150693893433,
+      "learning_rate": 0.00015408822639720023,
+      "loss": 2.5966,
+      "step": 6430
+    },
+    {
+      "epoch": 0.5190057299652974,
+      "grad_norm": 0.7047349810600281,
+      "learning_rate": 0.00015407494739013352,
+      "loss": 2.6626,
+      "step": 6431
+    },
+    {
+      "epoch": 0.5190864337018803,
+      "grad_norm": 0.7537584900856018,
+      "learning_rate": 0.00015406166703537777,
+      "loss": 2.6452,
+      "step": 6432
+    },
+    {
+      "epoch": 0.5191671374384634,
+      "grad_norm": 0.7944707870483398,
+      "learning_rate": 0.00015404838533326394,
+      "loss": 2.6834,
+      "step": 6433
+    },
+    {
+      "epoch": 0.5192478411750464,
+      "grad_norm": 0.8602458238601685,
+      "learning_rate": 0.00015403510228412305,
+      "loss": 2.6238,
+      "step": 6434
+    },
+    {
+      "epoch": 0.5193285449116294,
+      "grad_norm": 0.7181896567344666,
+      "learning_rate": 0.0001540218178882862,
+      "loss": 2.652,
+      "step": 6435
+    },
+    {
+      "epoch": 0.5194092486482124,
+      "grad_norm": 0.7470960021018982,
+      "learning_rate": 0.0001540085321460844,
+      "loss": 2.6703,
+      "step": 6436
+    },
+    {
+      "epoch": 0.5194899523847955,
+      "grad_norm": 0.8249944448471069,
+      "learning_rate": 0.00015399524505784883,
+      "loss": 2.5945,
+      "step": 6437
+    },
+    {
+      "epoch": 0.5195706561213784,
+      "grad_norm": 0.7332444190979004,
+      "learning_rate": 0.00015398195662391057,
+      "loss": 2.6472,
+      "step": 6438
+    },
+    {
+      "epoch": 0.5196513598579614,
+      "grad_norm": 0.7727739810943604,
+      "learning_rate": 0.0001539686668446009,
+      "loss": 2.6276,
+      "step": 6439
+    },
+    {
+      "epoch": 0.5197320635945444,
+      "grad_norm": 0.7161617279052734,
+      "learning_rate": 0.00015395537572025094,
+      "loss": 2.624,
+      "step": 6440
+    },
+    {
+      "epoch": 0.5198127673311275,
+      "grad_norm": 0.7657529711723328,
+      "learning_rate": 0.00015394208325119198,
+      "loss": 2.6604,
+      "step": 6441
+    },
+    {
+      "epoch": 0.5198934710677104,
+      "grad_norm": 0.732904314994812,
+      "learning_rate": 0.00015392878943775527,
+      "loss": 2.6334,
+      "step": 6442
+    },
+    {
+      "epoch": 0.5199741748042934,
+      "grad_norm": 0.7058991193771362,
+      "learning_rate": 0.0001539154942802722,
+      "loss": 2.5936,
+      "step": 6443
+    },
+    {
+      "epoch": 0.5200548785408764,
+      "grad_norm": 0.7328821420669556,
+      "learning_rate": 0.00015390219777907405,
+      "loss": 2.5969,
+      "step": 6444
+    },
+    {
+      "epoch": 0.5201355822774595,
+      "grad_norm": 0.7899969220161438,
+      "learning_rate": 0.00015388889993449224,
+      "loss": 2.5856,
+      "step": 6445
+    },
+    {
+      "epoch": 0.5202162860140425,
+      "grad_norm": 0.6963860392570496,
+      "learning_rate": 0.00015387560074685817,
+      "loss": 2.6139,
+      "step": 6446
+    },
+    {
+      "epoch": 0.5202969897506254,
+      "grad_norm": 0.812053918838501,
+      "learning_rate": 0.00015386230021650327,
+      "loss": 2.716,
+      "step": 6447
+    },
+    {
+      "epoch": 0.5203776934872084,
+      "grad_norm": 0.766781210899353,
+      "learning_rate": 0.0001538489983437591,
+      "loss": 2.6509,
+      "step": 6448
+    },
+    {
+      "epoch": 0.5204583972237915,
+      "grad_norm": 0.6877299547195435,
+      "learning_rate": 0.00015383569512895712,
+      "loss": 2.6076,
+      "step": 6449
+    },
+    {
+      "epoch": 0.5205391009603745,
+      "grad_norm": 0.7009176015853882,
+      "learning_rate": 0.00015382239057242888,
+      "loss": 2.608,
+      "step": 6450
+    },
+    {
+      "epoch": 0.5206198046969575,
+      "grad_norm": 0.7187578678131104,
+      "learning_rate": 0.000153809084674506,
+      "loss": 2.5946,
+      "step": 6451
+    },
+    {
+      "epoch": 0.5207005084335404,
+      "grad_norm": 0.7242687344551086,
+      "learning_rate": 0.00015379577743552001,
+      "loss": 2.6752,
+      "step": 6452
+    },
+    {
+      "epoch": 0.5207812121701235,
+      "grad_norm": 0.7668174505233765,
+      "learning_rate": 0.00015378246885580266,
+      "loss": 2.6694,
+      "step": 6453
+    },
+    {
+      "epoch": 0.5208619159067065,
+      "grad_norm": 0.7676039338111877,
+      "learning_rate": 0.00015376915893568557,
+      "loss": 2.6379,
+      "step": 6454
+    },
+    {
+      "epoch": 0.5209426196432895,
+      "grad_norm": 0.7394412159919739,
+      "learning_rate": 0.00015375584767550053,
+      "loss": 2.6046,
+      "step": 6455
+    },
+    {
+      "epoch": 0.5210233233798724,
+      "grad_norm": 0.7246636748313904,
+      "learning_rate": 0.00015374253507557923,
+      "loss": 2.592,
+      "step": 6456
+    },
+    {
+      "epoch": 0.5211040271164555,
+      "grad_norm": 0.7121255993843079,
+      "learning_rate": 0.00015372922113625345,
+      "loss": 2.634,
+      "step": 6457
+    },
+    {
+      "epoch": 0.5211847308530385,
+      "grad_norm": 0.7378345131874084,
+      "learning_rate": 0.00015371590585785505,
+      "loss": 2.5753,
+      "step": 6458
+    },
+    {
+      "epoch": 0.5212654345896215,
+      "grad_norm": 0.6682030558586121,
+      "learning_rate": 0.00015370258924071587,
+      "loss": 2.6305,
+      "step": 6459
+    },
+    {
+      "epoch": 0.5213461383262045,
+      "grad_norm": 0.7164177894592285,
+      "learning_rate": 0.00015368927128516776,
+      "loss": 2.7188,
+      "step": 6460
+    },
+    {
+      "epoch": 0.5214268420627876,
+      "grad_norm": 0.7341115474700928,
+      "learning_rate": 0.00015367595199154273,
+      "loss": 2.6204,
+      "step": 6461
+    },
+    {
+      "epoch": 0.5215075457993705,
+      "grad_norm": 0.6781840920448303,
+      "learning_rate": 0.00015366263136017258,
+      "loss": 2.6104,
+      "step": 6462
+    },
+    {
+      "epoch": 0.5215882495359535,
+      "grad_norm": 0.7029077410697937,
+      "learning_rate": 0.0001536493093913894,
+      "loss": 2.6055,
+      "step": 6463
+    },
+    {
+      "epoch": 0.5216689532725365,
+      "grad_norm": 0.6958553194999695,
+      "learning_rate": 0.00015363598608552522,
+      "loss": 2.5991,
+      "step": 6464
+    },
+    {
+      "epoch": 0.5217496570091196,
+      "grad_norm": 0.6919750571250916,
+      "learning_rate": 0.00015362266144291207,
+      "loss": 2.6022,
+      "step": 6465
+    },
+    {
+      "epoch": 0.5218303607457025,
+      "grad_norm": 0.6980622410774231,
+      "learning_rate": 0.000153609335463882,
+      "loss": 2.6289,
+      "step": 6466
+    },
+    {
+      "epoch": 0.5219110644822855,
+      "grad_norm": 0.7468248009681702,
+      "learning_rate": 0.00015359600814876715,
+      "loss": 2.6327,
+      "step": 6467
+    },
+    {
+      "epoch": 0.5219917682188685,
+      "grad_norm": 0.7183729410171509,
+      "learning_rate": 0.00015358267949789966,
+      "loss": 2.6389,
+      "step": 6468
+    },
+    {
+      "epoch": 0.5220724719554516,
+      "grad_norm": 0.6558868885040283,
+      "learning_rate": 0.00015356934951161178,
+      "loss": 2.6261,
+      "step": 6469
+    },
+    {
+      "epoch": 0.5221531756920346,
+      "grad_norm": 0.8000216484069824,
+      "learning_rate": 0.00015355601819023562,
+      "loss": 2.6908,
+      "step": 6470
+    },
+    {
+      "epoch": 0.5222338794286175,
+      "grad_norm": 0.775056004524231,
+      "learning_rate": 0.00015354268553410355,
+      "loss": 2.6763,
+      "step": 6471
+    },
+    {
+      "epoch": 0.5223145831652005,
+      "grad_norm": 0.7345123291015625,
+      "learning_rate": 0.00015352935154354776,
+      "loss": 2.582,
+      "step": 6472
+    },
+    {
+      "epoch": 0.5223952869017836,
+      "grad_norm": 0.731311023235321,
+      "learning_rate": 0.0001535160162189006,
+      "loss": 2.6519,
+      "step": 6473
+    },
+    {
+      "epoch": 0.5224759906383666,
+      "grad_norm": 0.6481007933616638,
+      "learning_rate": 0.00015350267956049443,
+      "loss": 2.5695,
+      "step": 6474
+    },
+    {
+      "epoch": 0.5225566943749496,
+      "grad_norm": 0.7698814868927002,
+      "learning_rate": 0.00015348934156866163,
+      "loss": 2.5732,
+      "step": 6475
+    },
+    {
+      "epoch": 0.5226373981115325,
+      "grad_norm": 0.7404680848121643,
+      "learning_rate": 0.00015347600224373462,
+      "loss": 2.5826,
+      "step": 6476
+    },
+    {
+      "epoch": 0.5227181018481155,
+      "grad_norm": 0.6965613961219788,
+      "learning_rate": 0.00015346266158604584,
+      "loss": 2.6069,
+      "step": 6477
+    },
+    {
+      "epoch": 0.5227988055846986,
+      "grad_norm": 0.6611152291297913,
+      "learning_rate": 0.00015344931959592777,
+      "loss": 2.4937,
+      "step": 6478
+    },
+    {
+      "epoch": 0.5228795093212816,
+      "grad_norm": 0.7418150305747986,
+      "learning_rate": 0.00015343597627371296,
+      "loss": 2.5747,
+      "step": 6479
+    },
+    {
+      "epoch": 0.5229602130578646,
+      "grad_norm": 0.6847610473632812,
+      "learning_rate": 0.00015342263161973393,
+      "loss": 2.5906,
+      "step": 6480
+    },
+    {
+      "epoch": 0.5230409167944475,
+      "grad_norm": 0.7054881453514099,
+      "learning_rate": 0.00015340928563432326,
+      "loss": 2.5914,
+      "step": 6481
+    },
+    {
+      "epoch": 0.5231216205310306,
+      "grad_norm": 0.6918888092041016,
+      "learning_rate": 0.0001533959383178136,
+      "loss": 2.6412,
+      "step": 6482
+    },
+    {
+      "epoch": 0.5232023242676136,
+      "grad_norm": 0.7232856154441833,
+      "learning_rate": 0.00015338258967053755,
+      "loss": 2.6364,
+      "step": 6483
+    },
+    {
+      "epoch": 0.5232830280041966,
+      "grad_norm": 0.7345031499862671,
+      "learning_rate": 0.00015336923969282786,
+      "loss": 2.6649,
+      "step": 6484
+    },
+    {
+      "epoch": 0.5233637317407795,
+      "grad_norm": 0.7644383907318115,
+      "learning_rate": 0.0001533558883850172,
+      "loss": 2.6949,
+      "step": 6485
+    },
+    {
+      "epoch": 0.5234444354773626,
+      "grad_norm": 0.6532372832298279,
+      "learning_rate": 0.0001533425357474383,
+      "loss": 2.5915,
+      "step": 6486
+    },
+    {
+      "epoch": 0.5235251392139456,
+      "grad_norm": 0.7089118361473083,
+      "learning_rate": 0.000153329181780424,
+      "loss": 2.6446,
+      "step": 6487
+    },
+    {
+      "epoch": 0.5236058429505286,
+      "grad_norm": 0.6966068148612976,
+      "learning_rate": 0.00015331582648430705,
+      "loss": 2.6764,
+      "step": 6488
+    },
+    {
+      "epoch": 0.5236865466871116,
+      "grad_norm": 0.7130835056304932,
+      "learning_rate": 0.00015330246985942035,
+      "loss": 2.6279,
+      "step": 6489
+    },
+    {
+      "epoch": 0.5237672504236947,
+      "grad_norm": 0.729727029800415,
+      "learning_rate": 0.00015328911190609678,
+      "loss": 2.612,
+      "step": 6490
+    },
+    {
+      "epoch": 0.5238479541602776,
+      "grad_norm": 0.6804213523864746,
+      "learning_rate": 0.0001532757526246692,
+      "loss": 2.6113,
+      "step": 6491
+    },
+    {
+      "epoch": 0.5239286578968606,
+      "grad_norm": 0.7324437499046326,
+      "learning_rate": 0.0001532623920154707,
+      "loss": 2.6054,
+      "step": 6492
+    },
+    {
+      "epoch": 0.5240093616334436,
+      "grad_norm": 0.6166699528694153,
+      "learning_rate": 0.00015324903007883406,
+      "loss": 2.5822,
+      "step": 6493
+    },
+    {
+      "epoch": 0.5240900653700267,
+      "grad_norm": 0.7339944839477539,
+      "learning_rate": 0.00015323566681509242,
+      "loss": 2.6204,
+      "step": 6494
+    },
+    {
+      "epoch": 0.5241707691066096,
+      "grad_norm": 0.7267727255821228,
+      "learning_rate": 0.00015322230222457886,
+      "loss": 2.6094,
+      "step": 6495
+    },
+    {
+      "epoch": 0.5242514728431926,
+      "grad_norm": 0.6417120695114136,
+      "learning_rate": 0.00015320893630762635,
+      "loss": 2.6044,
+      "step": 6496
+    },
+    {
+      "epoch": 0.5243321765797756,
+      "grad_norm": 0.7092922329902649,
+      "learning_rate": 0.00015319556906456808,
+      "loss": 2.6428,
+      "step": 6497
+    },
+    {
+      "epoch": 0.5244128803163587,
+      "grad_norm": 0.7482922673225403,
+      "learning_rate": 0.00015318220049573714,
+      "loss": 2.6025,
+      "step": 6498
+    },
+    {
+      "epoch": 0.5244935840529417,
+      "grad_norm": 0.691925048828125,
+      "learning_rate": 0.00015316883060146675,
+      "loss": 2.6308,
+      "step": 6499
+    },
+    {
+      "epoch": 0.5245742877895246,
+      "grad_norm": 0.7084488272666931,
+      "learning_rate": 0.00015315545938209015,
+      "loss": 2.6535,
+      "step": 6500
+    },
+    {
+      "epoch": 0.5246549915261076,
+      "grad_norm": 0.7182802557945251,
+      "learning_rate": 0.00015314208683794056,
+      "loss": 2.6045,
+      "step": 6501
+    },
+    {
+      "epoch": 0.5247356952626907,
+      "grad_norm": 0.7043096423149109,
+      "learning_rate": 0.00015312871296935122,
+      "loss": 2.6465,
+      "step": 6502
+    },
+    {
+      "epoch": 0.5248163989992737,
+      "grad_norm": 0.7679466009140015,
+      "learning_rate": 0.00015311533777665547,
+      "loss": 2.6624,
+      "step": 6503
+    },
+    {
+      "epoch": 0.5248971027358567,
+      "grad_norm": 0.6825870275497437,
+      "learning_rate": 0.00015310196126018668,
+      "loss": 2.5548,
+      "step": 6504
+    },
+    {
+      "epoch": 0.5249778064724396,
+      "grad_norm": 0.7364058494567871,
+      "learning_rate": 0.00015308858342027816,
+      "loss": 2.6495,
+      "step": 6505
+    },
+    {
+      "epoch": 0.5250585102090227,
+      "grad_norm": 0.7333239316940308,
+      "learning_rate": 0.00015307520425726341,
+      "loss": 2.5835,
+      "step": 6506
+    },
+    {
+      "epoch": 0.5251392139456057,
+      "grad_norm": 0.7479620575904846,
+      "learning_rate": 0.00015306182377147583,
+      "loss": 2.6065,
+      "step": 6507
+    },
+    {
+      "epoch": 0.5252199176821887,
+      "grad_norm": 0.7347591519355774,
+      "learning_rate": 0.00015304844196324888,
+      "loss": 2.6624,
+      "step": 6508
+    },
+    {
+      "epoch": 0.5253006214187717,
+      "grad_norm": 0.6879193782806396,
+      "learning_rate": 0.0001530350588329161,
+      "loss": 2.6598,
+      "step": 6509
+    },
+    {
+      "epoch": 0.5253813251553547,
+      "grad_norm": 0.7841597199440002,
+      "learning_rate": 0.000153021674380811,
+      "loss": 2.53,
+      "step": 6510
+    },
+    {
+      "epoch": 0.5254620288919377,
+      "grad_norm": 0.7916845679283142,
+      "learning_rate": 0.0001530082886072672,
+      "loss": 2.6995,
+      "step": 6511
+    },
+    {
+      "epoch": 0.5255427326285207,
+      "grad_norm": 0.7066318988800049,
+      "learning_rate": 0.0001529949015126183,
+      "loss": 2.58,
+      "step": 6512
+    },
+    {
+      "epoch": 0.5256234363651037,
+      "grad_norm": 0.6871134638786316,
+      "learning_rate": 0.00015298151309719787,
+      "loss": 2.6095,
+      "step": 6513
+    },
+    {
+      "epoch": 0.5257041401016868,
+      "grad_norm": 0.7479702830314636,
+      "learning_rate": 0.00015296812336133963,
+      "loss": 2.608,
+      "step": 6514
+    },
+    {
+      "epoch": 0.5257848438382697,
+      "grad_norm": 0.6772119402885437,
+      "learning_rate": 0.00015295473230537735,
+      "loss": 2.5679,
+      "step": 6515
+    },
+    {
+      "epoch": 0.5258655475748527,
+      "grad_norm": 0.7365416884422302,
+      "learning_rate": 0.0001529413399296447,
+      "loss": 2.6722,
+      "step": 6516
+    },
+    {
+      "epoch": 0.5259462513114357,
+      "grad_norm": 0.7538040280342102,
+      "learning_rate": 0.00015292794623447545,
+      "loss": 2.5562,
+      "step": 6517
+    },
+    {
+      "epoch": 0.5260269550480188,
+      "grad_norm": 0.7471820712089539,
+      "learning_rate": 0.00015291455122020344,
+      "loss": 2.7079,
+      "step": 6518
+    },
+    {
+      "epoch": 0.5261076587846018,
+      "grad_norm": 0.7605932354927063,
+      "learning_rate": 0.00015290115488716247,
+      "loss": 2.6696,
+      "step": 6519
+    },
+    {
+      "epoch": 0.5261883625211847,
+      "grad_norm": 0.7081854939460754,
+      "learning_rate": 0.00015288775723568647,
+      "loss": 2.6502,
+      "step": 6520
+    },
+    {
+      "epoch": 0.5262690662577677,
+      "grad_norm": 0.7236372828483582,
+      "learning_rate": 0.0001528743582661093,
+      "loss": 2.662,
+      "step": 6521
+    },
+    {
+      "epoch": 0.5263497699943508,
+      "grad_norm": 0.6710047721862793,
+      "learning_rate": 0.0001528609579787649,
+      "loss": 2.5947,
+      "step": 6522
+    },
+    {
+      "epoch": 0.5264304737309338,
+      "grad_norm": 0.709381103515625,
+      "learning_rate": 0.00015284755637398726,
+      "loss": 2.5922,
+      "step": 6523
+    },
+    {
+      "epoch": 0.5265111774675167,
+      "grad_norm": 0.7029775381088257,
+      "learning_rate": 0.00015283415345211033,
+      "loss": 2.6777,
+      "step": 6524
+    },
+    {
+      "epoch": 0.5265918812040997,
+      "grad_norm": 0.7250857949256897,
+      "learning_rate": 0.00015282074921346825,
+      "loss": 2.6027,
+      "step": 6525
+    },
+    {
+      "epoch": 0.5266725849406828,
+      "grad_norm": 0.7192760705947876,
+      "learning_rate": 0.00015280734365839498,
+      "loss": 2.6544,
+      "step": 6526
+    },
+    {
+      "epoch": 0.5267532886772658,
+      "grad_norm": 0.693583071231842,
+      "learning_rate": 0.0001527939367872247,
+      "loss": 2.6302,
+      "step": 6527
+    },
+    {
+      "epoch": 0.5268339924138488,
+      "grad_norm": 0.7031428217887878,
+      "learning_rate": 0.00015278052860029145,
+      "loss": 2.6944,
+      "step": 6528
+    },
+    {
+      "epoch": 0.5269146961504317,
+      "grad_norm": 0.6986895799636841,
+      "learning_rate": 0.00015276711909792949,
+      "loss": 2.6595,
+      "step": 6529
+    },
+    {
+      "epoch": 0.5269953998870147,
+      "grad_norm": 0.7375979423522949,
+      "learning_rate": 0.000152753708280473,
+      "loss": 2.6839,
+      "step": 6530
+    },
+    {
+      "epoch": 0.5270761036235978,
+      "grad_norm": 0.7126755714416504,
+      "learning_rate": 0.0001527402961482562,
+      "loss": 2.5597,
+      "step": 6531
+    },
+    {
+      "epoch": 0.5271568073601808,
+      "grad_norm": 0.6631070971488953,
+      "learning_rate": 0.00015272688270161338,
+      "loss": 2.5566,
+      "step": 6532
+    },
+    {
+      "epoch": 0.5272375110967638,
+      "grad_norm": 0.6896609663963318,
+      "learning_rate": 0.00015271346794087874,
+      "loss": 2.5801,
+      "step": 6533
+    },
+    {
+      "epoch": 0.5273182148333467,
+      "grad_norm": 0.7437502145767212,
+      "learning_rate": 0.00015270005186638673,
+      "loss": 2.6572,
+      "step": 6534
+    },
+    {
+      "epoch": 0.5273989185699298,
+      "grad_norm": 0.7013052701950073,
+      "learning_rate": 0.00015268663447847166,
+      "loss": 2.621,
+      "step": 6535
+    },
+    {
+      "epoch": 0.5274796223065128,
+      "grad_norm": 0.7161773443222046,
+      "learning_rate": 0.00015267321577746795,
+      "loss": 2.5989,
+      "step": 6536
+    },
+    {
+      "epoch": 0.5275603260430958,
+      "grad_norm": 0.7654534578323364,
+      "learning_rate": 0.00015265979576371,
+      "loss": 2.6338,
+      "step": 6537
+    },
+    {
+      "epoch": 0.5276410297796787,
+      "grad_norm": 0.694646954536438,
+      "learning_rate": 0.0001526463744375323,
+      "loss": 2.6036,
+      "step": 6538
+    },
+    {
+      "epoch": 0.5277217335162618,
+      "grad_norm": 0.6594679355621338,
+      "learning_rate": 0.0001526329517992693,
+      "loss": 2.6256,
+      "step": 6539
+    },
+    {
+      "epoch": 0.5278024372528448,
+      "grad_norm": 0.6424389481544495,
+      "learning_rate": 0.00015261952784925557,
+      "loss": 2.6389,
+      "step": 6540
+    },
+    {
+      "epoch": 0.5278831409894278,
+      "grad_norm": 0.7465235590934753,
+      "learning_rate": 0.0001526061025878257,
+      "loss": 2.5449,
+      "step": 6541
+    },
+    {
+      "epoch": 0.5279638447260108,
+      "grad_norm": 0.6900132298469543,
+      "learning_rate": 0.0001525926760153142,
+      "loss": 2.5597,
+      "step": 6542
+    },
+    {
+      "epoch": 0.5280445484625939,
+      "grad_norm": 0.7505282163619995,
+      "learning_rate": 0.00015257924813205572,
+      "loss": 2.6526,
+      "step": 6543
+    },
+    {
+      "epoch": 0.5281252521991768,
+      "grad_norm": 0.72642582654953,
+      "learning_rate": 0.00015256581893838495,
+      "loss": 2.6593,
+      "step": 6544
+    },
+    {
+      "epoch": 0.5282059559357598,
+      "grad_norm": 0.6901132464408875,
+      "learning_rate": 0.00015255238843463656,
+      "loss": 2.6726,
+      "step": 6545
+    },
+    {
+      "epoch": 0.5282866596723428,
+      "grad_norm": 0.7741395831108093,
+      "learning_rate": 0.0001525389566211453,
+      "loss": 2.5929,
+      "step": 6546
+    },
+    {
+      "epoch": 0.5283673634089259,
+      "grad_norm": 0.7282403111457825,
+      "learning_rate": 0.00015252552349824585,
+      "loss": 2.5696,
+      "step": 6547
+    },
+    {
+      "epoch": 0.5284480671455088,
+      "grad_norm": 0.7421764731407166,
+      "learning_rate": 0.0001525120890662731,
+      "loss": 2.5593,
+      "step": 6548
+    },
+    {
+      "epoch": 0.5285287708820918,
+      "grad_norm": 0.6830468773841858,
+      "learning_rate": 0.00015249865332556182,
+      "loss": 2.6396,
+      "step": 6549
+    },
+    {
+      "epoch": 0.5286094746186748,
+      "grad_norm": 0.6758440732955933,
+      "learning_rate": 0.00015248521627644684,
+      "loss": 2.5375,
+      "step": 6550
+    },
+    {
+      "epoch": 0.5286901783552579,
+      "grad_norm": 0.6897253394126892,
+      "learning_rate": 0.00015247177791926308,
+      "loss": 2.6148,
+      "step": 6551
+    },
+    {
+      "epoch": 0.5287708820918409,
+      "grad_norm": 0.6391426920890808,
+      "learning_rate": 0.00015245833825434547,
+      "loss": 2.5563,
+      "step": 6552
+    },
+    {
+      "epoch": 0.5288515858284238,
+      "grad_norm": 0.7213610410690308,
+      "learning_rate": 0.00015244489728202893,
+      "loss": 2.6158,
+      "step": 6553
+    },
+    {
+      "epoch": 0.5289322895650068,
+      "grad_norm": 0.6678160429000854,
+      "learning_rate": 0.00015243145500264845,
+      "loss": 2.6177,
+      "step": 6554
+    },
+    {
+      "epoch": 0.5290129933015899,
+      "grad_norm": 0.7041724324226379,
+      "learning_rate": 0.00015241801141653905,
+      "loss": 2.6504,
+      "step": 6555
+    },
+    {
+      "epoch": 0.5290936970381729,
+      "grad_norm": 0.6551648378372192,
+      "learning_rate": 0.0001524045665240358,
+      "loss": 2.577,
+      "step": 6556
+    },
+    {
+      "epoch": 0.5291744007747559,
+      "grad_norm": 0.7190412878990173,
+      "learning_rate": 0.00015239112032547377,
+      "loss": 2.596,
+      "step": 6557
+    },
+    {
+      "epoch": 0.5292551045113388,
+      "grad_norm": 0.6936302781105042,
+      "learning_rate": 0.00015237767282118807,
+      "loss": 2.6551,
+      "step": 6558
+    },
+    {
+      "epoch": 0.5293358082479219,
+      "grad_norm": 0.6901839971542358,
+      "learning_rate": 0.0001523642240115138,
+      "loss": 2.6263,
+      "step": 6559
+    },
+    {
+      "epoch": 0.5294165119845049,
+      "grad_norm": 0.6905068159103394,
+      "learning_rate": 0.00015235077389678624,
+      "loss": 2.6323,
+      "step": 6560
+    },
+    {
+      "epoch": 0.5294972157210879,
+      "grad_norm": 0.7495188117027283,
+      "learning_rate": 0.00015233732247734057,
+      "loss": 2.6243,
+      "step": 6561
+    },
+    {
+      "epoch": 0.5295779194576709,
+      "grad_norm": 0.6758708357810974,
+      "learning_rate": 0.00015232386975351197,
+      "loss": 2.6184,
+      "step": 6562
+    },
+    {
+      "epoch": 0.5296586231942539,
+      "grad_norm": 0.6443266868591309,
+      "learning_rate": 0.00015231041572563573,
+      "loss": 2.6543,
+      "step": 6563
+    },
+    {
+      "epoch": 0.5297393269308369,
+      "grad_norm": 0.7384275794029236,
+      "learning_rate": 0.00015229696039404723,
+      "loss": 2.6117,
+      "step": 6564
+    },
+    {
+      "epoch": 0.5298200306674199,
+      "grad_norm": 0.6873897314071655,
+      "learning_rate": 0.00015228350375908178,
+      "loss": 2.5689,
+      "step": 6565
+    },
+    {
+      "epoch": 0.5299007344040029,
+      "grad_norm": 0.6715645790100098,
+      "learning_rate": 0.00015227004582107472,
+      "loss": 2.5943,
+      "step": 6566
+    },
+    {
+      "epoch": 0.529981438140586,
+      "grad_norm": 0.6814208030700684,
+      "learning_rate": 0.00015225658658036151,
+      "loss": 2.5562,
+      "step": 6567
+    },
+    {
+      "epoch": 0.5300621418771689,
+      "grad_norm": 0.6942310929298401,
+      "learning_rate": 0.00015224312603727755,
+      "loss": 2.5902,
+      "step": 6568
+    },
+    {
+      "epoch": 0.5301428456137519,
+      "grad_norm": 0.6856299042701721,
+      "learning_rate": 0.0001522296641921583,
+      "loss": 2.6115,
+      "step": 6569
+    },
+    {
+      "epoch": 0.5302235493503349,
+      "grad_norm": 0.870833694934845,
+      "learning_rate": 0.0001522162010453393,
+      "loss": 2.7492,
+      "step": 6570
+    },
+    {
+      "epoch": 0.530304253086918,
+      "grad_norm": 0.6796989440917969,
+      "learning_rate": 0.0001522027365971561,
+      "loss": 2.6957,
+      "step": 6571
+    },
+    {
+      "epoch": 0.530384956823501,
+      "grad_norm": 0.7043026685714722,
+      "learning_rate": 0.00015218927084794423,
+      "loss": 2.604,
+      "step": 6572
+    },
+    {
+      "epoch": 0.5304656605600839,
+      "grad_norm": 0.7533933520317078,
+      "learning_rate": 0.00015217580379803933,
+      "loss": 2.6271,
+      "step": 6573
+    },
+    {
+      "epoch": 0.5305463642966669,
+      "grad_norm": 0.7526697516441345,
+      "learning_rate": 0.000152162335447777,
+      "loss": 2.553,
+      "step": 6574
+    },
+    {
+      "epoch": 0.53062706803325,
+      "grad_norm": 0.6942071318626404,
+      "learning_rate": 0.00015214886579749284,
+      "loss": 2.7206,
+      "step": 6575
+    },
+    {
+      "epoch": 0.530707771769833,
+      "grad_norm": 0.7133236527442932,
+      "learning_rate": 0.00015213539484752273,
+      "loss": 2.6545,
+      "step": 6576
+    },
+    {
+      "epoch": 0.530788475506416,
+      "grad_norm": 0.7229849696159363,
+      "learning_rate": 0.00015212192259820222,
+      "loss": 2.6647,
+      "step": 6577
+    },
+    {
+      "epoch": 0.5308691792429989,
+      "grad_norm": 0.7142449617385864,
+      "learning_rate": 0.0001521084490498672,
+      "loss": 2.5777,
+      "step": 6578
+    },
+    {
+      "epoch": 0.5309498829795819,
+      "grad_norm": 0.6950247287750244,
+      "learning_rate": 0.00015209497420285342,
+      "loss": 2.6159,
+      "step": 6579
+    },
+    {
+      "epoch": 0.531030586716165,
+      "grad_norm": 0.7492622137069702,
+      "learning_rate": 0.00015208149805749668,
+      "loss": 2.6927,
+      "step": 6580
+    },
+    {
+      "epoch": 0.531111290452748,
+      "grad_norm": 0.7618215084075928,
+      "learning_rate": 0.00015206802061413287,
+      "loss": 2.5831,
+      "step": 6581
+    },
+    {
+      "epoch": 0.5311919941893309,
+      "grad_norm": 0.7448660731315613,
+      "learning_rate": 0.0001520545418730979,
+      "loss": 2.6123,
+      "step": 6582
+    },
+    {
+      "epoch": 0.5312726979259139,
+      "grad_norm": 0.7450618147850037,
+      "learning_rate": 0.00015204106183472766,
+      "loss": 2.5768,
+      "step": 6583
+    },
+    {
+      "epoch": 0.531353401662497,
+      "grad_norm": 0.7426019310951233,
+      "learning_rate": 0.0001520275804993581,
+      "loss": 2.603,
+      "step": 6584
+    },
+    {
+      "epoch": 0.53143410539908,
+      "grad_norm": 0.7503333687782288,
+      "learning_rate": 0.00015201409786732526,
+      "loss": 2.6159,
+      "step": 6585
+    },
+    {
+      "epoch": 0.531514809135663,
+      "grad_norm": 0.6944373846054077,
+      "learning_rate": 0.00015200061393896513,
+      "loss": 2.5201,
+      "step": 6586
+    },
+    {
+      "epoch": 0.5315955128722459,
+      "grad_norm": 0.6958110332489014,
+      "learning_rate": 0.00015198712871461375,
+      "loss": 2.5592,
+      "step": 6587
+    },
+    {
+      "epoch": 0.531676216608829,
+      "grad_norm": 0.7838244438171387,
+      "learning_rate": 0.00015197364219460727,
+      "loss": 2.6663,
+      "step": 6588
+    },
+    {
+      "epoch": 0.531756920345412,
+      "grad_norm": 0.754338800907135,
+      "learning_rate": 0.00015196015437928174,
+      "loss": 2.6183,
+      "step": 6589
+    },
+    {
+      "epoch": 0.531837624081995,
+      "grad_norm": 0.7394337058067322,
+      "learning_rate": 0.00015194666526897332,
+      "loss": 2.5622,
+      "step": 6590
+    },
+    {
+      "epoch": 0.531918327818578,
+      "grad_norm": 0.7352069020271301,
+      "learning_rate": 0.00015193317486401824,
+      "loss": 2.6173,
+      "step": 6591
+    },
+    {
+      "epoch": 0.531999031555161,
+      "grad_norm": 0.6318944096565247,
+      "learning_rate": 0.00015191968316475267,
+      "loss": 2.6159,
+      "step": 6592
+    },
+    {
+      "epoch": 0.532079735291744,
+      "grad_norm": 0.7071281671524048,
+      "learning_rate": 0.00015190619017151291,
+      "loss": 2.633,
+      "step": 6593
+    },
+    {
+      "epoch": 0.532160439028327,
+      "grad_norm": 0.7762585282325745,
+      "learning_rate": 0.00015189269588463517,
+      "loss": 2.6445,
+      "step": 6594
+    },
+    {
+      "epoch": 0.53224114276491,
+      "grad_norm": 0.7979930639266968,
+      "learning_rate": 0.0001518792003044558,
+      "loss": 2.5825,
+      "step": 6595
+    },
+    {
+      "epoch": 0.5323218465014931,
+      "grad_norm": 0.7355580925941467,
+      "learning_rate": 0.00015186570343131114,
+      "loss": 2.6197,
+      "step": 6596
+    },
+    {
+      "epoch": 0.532402550238076,
+      "grad_norm": 0.7286938428878784,
+      "learning_rate": 0.0001518522052655376,
+      "loss": 2.6385,
+      "step": 6597
+    },
+    {
+      "epoch": 0.532483253974659,
+      "grad_norm": 0.689143180847168,
+      "learning_rate": 0.00015183870580747156,
+      "loss": 2.6593,
+      "step": 6598
+    },
+    {
+      "epoch": 0.532563957711242,
+      "grad_norm": 0.714746356010437,
+      "learning_rate": 0.00015182520505744945,
+      "loss": 2.6059,
+      "step": 6599
+    },
+    {
+      "epoch": 0.5326446614478251,
+      "grad_norm": 0.8055040240287781,
+      "learning_rate": 0.00015181170301580777,
+      "loss": 2.6983,
+      "step": 6600
+    },
+    {
+      "epoch": 0.532725365184408,
+      "grad_norm": 0.7104170918464661,
+      "learning_rate": 0.00015179819968288297,
+      "loss": 2.6578,
+      "step": 6601
+    },
+    {
+      "epoch": 0.532806068920991,
+      "grad_norm": 0.7175524830818176,
+      "learning_rate": 0.0001517846950590117,
+      "loss": 2.6263,
+      "step": 6602
+    },
+    {
+      "epoch": 0.532886772657574,
+      "grad_norm": 0.6755492091178894,
+      "learning_rate": 0.00015177118914453042,
+      "loss": 2.5752,
+      "step": 6603
+    },
+    {
+      "epoch": 0.5329674763941571,
+      "grad_norm": 0.7020289897918701,
+      "learning_rate": 0.00015175768193977578,
+      "loss": 2.6186,
+      "step": 6604
+    },
+    {
+      "epoch": 0.5330481801307401,
+      "grad_norm": 0.7550958395004272,
+      "learning_rate": 0.0001517441734450844,
+      "loss": 2.628,
+      "step": 6605
+    },
+    {
+      "epoch": 0.533128883867323,
+      "grad_norm": 0.6697603464126587,
+      "learning_rate": 0.00015173066366079297,
+      "loss": 2.6433,
+      "step": 6606
+    },
+    {
+      "epoch": 0.533209587603906,
+      "grad_norm": 0.715372622013092,
+      "learning_rate": 0.0001517171525872382,
+      "loss": 2.6022,
+      "step": 6607
+    },
+    {
+      "epoch": 0.5332902913404891,
+      "grad_norm": 0.7081933617591858,
+      "learning_rate": 0.00015170364022475675,
+      "loss": 2.675,
+      "step": 6608
+    },
+    {
+      "epoch": 0.5333709950770721,
+      "grad_norm": 0.7074152231216431,
+      "learning_rate": 0.00015169012657368546,
+      "loss": 2.6637,
+      "step": 6609
+    },
+    {
+      "epoch": 0.5334516988136551,
+      "grad_norm": 0.6692848801612854,
+      "learning_rate": 0.00015167661163436108,
+      "loss": 2.5855,
+      "step": 6610
+    },
+    {
+      "epoch": 0.533532402550238,
+      "grad_norm": 0.7307556867599487,
+      "learning_rate": 0.00015166309540712048,
+      "loss": 2.6105,
+      "step": 6611
+    },
+    {
+      "epoch": 0.5336131062868211,
+      "grad_norm": 0.7026669383049011,
+      "learning_rate": 0.00015164957789230048,
+      "loss": 2.6656,
+      "step": 6612
+    },
+    {
+      "epoch": 0.5336938100234041,
+      "grad_norm": 0.6579706072807312,
+      "learning_rate": 0.000151636059090238,
+      "loss": 2.6456,
+      "step": 6613
+    },
+    {
+      "epoch": 0.5337745137599871,
+      "grad_norm": 0.6854498386383057,
+      "learning_rate": 0.00015162253900126993,
+      "loss": 2.5969,
+      "step": 6614
+    },
+    {
+      "epoch": 0.5338552174965701,
+      "grad_norm": 0.7542434334754944,
+      "learning_rate": 0.00015160901762573323,
+      "loss": 2.6333,
+      "step": 6615
+    },
+    {
+      "epoch": 0.5339359212331531,
+      "grad_norm": 0.6795105934143066,
+      "learning_rate": 0.0001515954949639649,
+      "loss": 2.6268,
+      "step": 6616
+    },
+    {
+      "epoch": 0.5340166249697361,
+      "grad_norm": 0.6395254135131836,
+      "learning_rate": 0.000151581971016302,
+      "loss": 2.5684,
+      "step": 6617
+    },
+    {
+      "epoch": 0.5340973287063191,
+      "grad_norm": 0.7069850564002991,
+      "learning_rate": 0.00015156844578308155,
+      "loss": 2.64,
+      "step": 6618
+    },
+    {
+      "epoch": 0.5341780324429021,
+      "grad_norm": 0.6779203414916992,
+      "learning_rate": 0.0001515549192646406,
+      "loss": 2.6255,
+      "step": 6619
+    },
+    {
+      "epoch": 0.5342587361794852,
+      "grad_norm": 0.6403560638427734,
+      "learning_rate": 0.00015154139146131632,
+      "loss": 2.611,
+      "step": 6620
+    },
+    {
+      "epoch": 0.5343394399160681,
+      "grad_norm": 0.7532669901847839,
+      "learning_rate": 0.00015152786237344583,
+      "loss": 2.5641,
+      "step": 6621
+    },
+    {
+      "epoch": 0.5344201436526511,
+      "grad_norm": 0.6827573776245117,
+      "learning_rate": 0.00015151433200136629,
+      "loss": 2.6096,
+      "step": 6622
+    },
+    {
+      "epoch": 0.5345008473892341,
+      "grad_norm": 0.6691904067993164,
+      "learning_rate": 0.000151500800345415,
+      "loss": 2.6602,
+      "step": 6623
+    },
+    {
+      "epoch": 0.5345815511258172,
+      "grad_norm": 0.7288634777069092,
+      "learning_rate": 0.00015148726740592906,
+      "loss": 2.6468,
+      "step": 6624
+    },
+    {
+      "epoch": 0.5346622548624002,
+      "grad_norm": 0.7087839245796204,
+      "learning_rate": 0.00015147373318324586,
+      "loss": 2.5795,
+      "step": 6625
+    },
+    {
+      "epoch": 0.5347429585989831,
+      "grad_norm": 0.6618373394012451,
+      "learning_rate": 0.00015146019767770267,
+      "loss": 2.638,
+      "step": 6626
+    },
+    {
+      "epoch": 0.5348236623355661,
+      "grad_norm": 0.7384989857673645,
+      "learning_rate": 0.00015144666088963684,
+      "loss": 2.6104,
+      "step": 6627
+    },
+    {
+      "epoch": 0.5349043660721492,
+      "grad_norm": 0.6662275195121765,
+      "learning_rate": 0.00015143312281938576,
+      "loss": 2.6174,
+      "step": 6628
+    },
+    {
+      "epoch": 0.5349850698087322,
+      "grad_norm": 0.6617184281349182,
+      "learning_rate": 0.0001514195834672868,
+      "loss": 2.6154,
+      "step": 6629
+    },
+    {
+      "epoch": 0.5350657735453151,
+      "grad_norm": 0.7173622846603394,
+      "learning_rate": 0.0001514060428336774,
+      "loss": 2.5741,
+      "step": 6630
+    },
+    {
+      "epoch": 0.5351464772818981,
+      "grad_norm": 0.7773584127426147,
+      "learning_rate": 0.00015139250091889502,
+      "loss": 2.6333,
+      "step": 6631
+    },
+    {
+      "epoch": 0.5352271810184811,
+      "grad_norm": 0.7255204916000366,
+      "learning_rate": 0.0001513789577232772,
+      "loss": 2.5459,
+      "step": 6632
+    },
+    {
+      "epoch": 0.5353078847550642,
+      "grad_norm": 0.7308403849601746,
+      "learning_rate": 0.00015136541324716144,
+      "loss": 2.5934,
+      "step": 6633
+    },
+    {
+      "epoch": 0.5353885884916472,
+      "grad_norm": 0.699367880821228,
+      "learning_rate": 0.0001513518674908853,
+      "loss": 2.6797,
+      "step": 6634
+    },
+    {
+      "epoch": 0.5354692922282301,
+      "grad_norm": 0.7236449718475342,
+      "learning_rate": 0.0001513383204547864,
+      "loss": 2.6289,
+      "step": 6635
+    },
+    {
+      "epoch": 0.5355499959648131,
+      "grad_norm": 0.6860557794570923,
+      "learning_rate": 0.00015132477213920234,
+      "loss": 2.6736,
+      "step": 6636
+    },
+    {
+      "epoch": 0.5356306997013962,
+      "grad_norm": 0.6724153161048889,
+      "learning_rate": 0.00015131122254447084,
+      "loss": 2.5581,
+      "step": 6637
+    },
+    {
+      "epoch": 0.5357114034379792,
+      "grad_norm": 0.6818630695343018,
+      "learning_rate": 0.00015129767167092949,
+      "loss": 2.5979,
+      "step": 6638
+    },
+    {
+      "epoch": 0.5357921071745622,
+      "grad_norm": 0.6956631541252136,
+      "learning_rate": 0.00015128411951891607,
+      "loss": 2.6116,
+      "step": 6639
+    },
+    {
+      "epoch": 0.5358728109111451,
+      "grad_norm": 0.6698076128959656,
+      "learning_rate": 0.00015127056608876837,
+      "loss": 2.65,
+      "step": 6640
+    },
+    {
+      "epoch": 0.5359535146477282,
+      "grad_norm": 0.7763264179229736,
+      "learning_rate": 0.00015125701138082415,
+      "loss": 2.6164,
+      "step": 6641
+    },
+    {
+      "epoch": 0.5360342183843112,
+      "grad_norm": 0.7148340940475464,
+      "learning_rate": 0.00015124345539542118,
+      "loss": 2.6467,
+      "step": 6642
+    },
+    {
+      "epoch": 0.5361149221208942,
+      "grad_norm": 0.7350041270256042,
+      "learning_rate": 0.00015122989813289733,
+      "loss": 2.6477,
+      "step": 6643
+    },
+    {
+      "epoch": 0.5361956258574772,
+      "grad_norm": 0.6993441581726074,
+      "learning_rate": 0.00015121633959359055,
+      "loss": 2.7526,
+      "step": 6644
+    },
+    {
+      "epoch": 0.5362763295940602,
+      "grad_norm": 0.6828470826148987,
+      "learning_rate": 0.00015120277977783873,
+      "loss": 2.6439,
+      "step": 6645
+    },
+    {
+      "epoch": 0.5363570333306432,
+      "grad_norm": 0.7076796889305115,
+      "learning_rate": 0.0001511892186859797,
+      "loss": 2.6375,
+      "step": 6646
+    },
+    {
+      "epoch": 0.5364377370672262,
+      "grad_norm": 0.6830769777297974,
+      "learning_rate": 0.0001511756563183516,
+      "loss": 2.6052,
+      "step": 6647
+    },
+    {
+      "epoch": 0.5365184408038092,
+      "grad_norm": 0.6482179760932922,
+      "learning_rate": 0.00015116209267529237,
+      "loss": 2.6251,
+      "step": 6648
+    },
+    {
+      "epoch": 0.5365991445403923,
+      "grad_norm": 0.6687620878219604,
+      "learning_rate": 0.00015114852775714,
+      "loss": 2.659,
+      "step": 6649
+    },
+    {
+      "epoch": 0.5366798482769752,
+      "grad_norm": 0.734108030796051,
+      "learning_rate": 0.0001511349615642327,
+      "loss": 2.6542,
+      "step": 6650
+    },
+    {
+      "epoch": 0.5367605520135582,
+      "grad_norm": 0.7092111706733704,
+      "learning_rate": 0.00015112139409690842,
+      "loss": 2.6228,
+      "step": 6651
+    },
+    {
+      "epoch": 0.5368412557501412,
+      "grad_norm": 0.6544996500015259,
+      "learning_rate": 0.0001511078253555054,
+      "loss": 2.5661,
+      "step": 6652
+    },
+    {
+      "epoch": 0.5369219594867243,
+      "grad_norm": 0.7012531161308289,
+      "learning_rate": 0.00015109425534036176,
+      "loss": 2.6447,
+      "step": 6653
+    },
+    {
+      "epoch": 0.5370026632233073,
+      "grad_norm": 0.6813335418701172,
+      "learning_rate": 0.0001510806840518157,
+      "loss": 2.5723,
+      "step": 6654
+    },
+    {
+      "epoch": 0.5370833669598902,
+      "grad_norm": 0.6711288094520569,
+      "learning_rate": 0.0001510671114902055,
+      "loss": 2.6096,
+      "step": 6655
+    },
+    {
+      "epoch": 0.5371640706964732,
+      "grad_norm": 0.721866250038147,
+      "learning_rate": 0.00015105353765586935,
+      "loss": 2.6167,
+      "step": 6656
+    },
+    {
+      "epoch": 0.5372447744330563,
+      "grad_norm": 0.8140639066696167,
+      "learning_rate": 0.00015103996254914562,
+      "loss": 2.5768,
+      "step": 6657
+    },
+    {
+      "epoch": 0.5373254781696393,
+      "grad_norm": 0.6859177947044373,
+      "learning_rate": 0.0001510263861703726,
+      "loss": 2.5638,
+      "step": 6658
+    },
+    {
+      "epoch": 0.5374061819062222,
+      "grad_norm": 0.7254204154014587,
+      "learning_rate": 0.00015101280851988864,
+      "loss": 2.5855,
+      "step": 6659
+    },
+    {
+      "epoch": 0.5374868856428052,
+      "grad_norm": 0.7181829810142517,
+      "learning_rate": 0.00015099922959803218,
+      "loss": 2.5358,
+      "step": 6660
+    },
+    {
+      "epoch": 0.5375675893793883,
+      "grad_norm": 0.7092663645744324,
+      "learning_rate": 0.00015098564940514155,
+      "loss": 2.679,
+      "step": 6661
+    },
+    {
+      "epoch": 0.5376482931159713,
+      "grad_norm": 0.7126225233078003,
+      "learning_rate": 0.00015097206794155527,
+      "loss": 2.6167,
+      "step": 6662
+    },
+    {
+      "epoch": 0.5377289968525543,
+      "grad_norm": 0.7469925880432129,
+      "learning_rate": 0.00015095848520761186,
+      "loss": 2.5906,
+      "step": 6663
+    },
+    {
+      "epoch": 0.5378097005891372,
+      "grad_norm": 0.6911186575889587,
+      "learning_rate": 0.00015094490120364973,
+      "loss": 2.6488,
+      "step": 6664
+    },
+    {
+      "epoch": 0.5378904043257203,
+      "grad_norm": 0.6579635143280029,
+      "learning_rate": 0.00015093131593000753,
+      "loss": 2.5894,
+      "step": 6665
+    },
+    {
+      "epoch": 0.5379711080623033,
+      "grad_norm": 0.7107242345809937,
+      "learning_rate": 0.00015091772938702377,
+      "loss": 2.6568,
+      "step": 6666
+    },
+    {
+      "epoch": 0.5380518117988863,
+      "grad_norm": 0.6845428943634033,
+      "learning_rate": 0.00015090414157503714,
+      "loss": 2.5697,
+      "step": 6667
+    },
+    {
+      "epoch": 0.5381325155354693,
+      "grad_norm": 0.6713212132453918,
+      "learning_rate": 0.00015089055249438622,
+      "loss": 2.5747,
+      "step": 6668
+    },
+    {
+      "epoch": 0.5382132192720523,
+      "grad_norm": 0.7091513276100159,
+      "learning_rate": 0.0001508769621454097,
+      "loss": 2.6765,
+      "step": 6669
+    },
+    {
+      "epoch": 0.5382939230086353,
+      "grad_norm": 0.7403436899185181,
+      "learning_rate": 0.00015086337052844627,
+      "loss": 2.6841,
+      "step": 6670
+    },
+    {
+      "epoch": 0.5383746267452183,
+      "grad_norm": 0.6745626330375671,
+      "learning_rate": 0.0001508497776438347,
+      "loss": 2.6436,
+      "step": 6671
+    },
+    {
+      "epoch": 0.5384553304818013,
+      "grad_norm": 0.7491294145584106,
+      "learning_rate": 0.00015083618349191372,
+      "loss": 2.6376,
+      "step": 6672
+    },
+    {
+      "epoch": 0.5385360342183844,
+      "grad_norm": 0.719761848449707,
+      "learning_rate": 0.00015082258807302222,
+      "loss": 2.5885,
+      "step": 6673
+    },
+    {
+      "epoch": 0.5386167379549673,
+      "grad_norm": 0.7302667498588562,
+      "learning_rate": 0.00015080899138749895,
+      "loss": 2.7019,
+      "step": 6674
+    },
+    {
+      "epoch": 0.5386974416915503,
+      "grad_norm": 0.7640584111213684,
+      "learning_rate": 0.0001507953934356828,
+      "loss": 2.6404,
+      "step": 6675
+    },
+    {
+      "epoch": 0.5387781454281333,
+      "grad_norm": 0.699515700340271,
+      "learning_rate": 0.0001507817942179127,
+      "loss": 2.6407,
+      "step": 6676
+    },
+    {
+      "epoch": 0.5388588491647164,
+      "grad_norm": 0.7305224537849426,
+      "learning_rate": 0.00015076819373452746,
+      "loss": 2.5994,
+      "step": 6677
+    },
+    {
+      "epoch": 0.5389395529012994,
+      "grad_norm": 0.7125952243804932,
+      "learning_rate": 0.00015075459198586616,
+      "loss": 2.6472,
+      "step": 6678
+    },
+    {
+      "epoch": 0.5390202566378823,
+      "grad_norm": 0.7077293395996094,
+      "learning_rate": 0.00015074098897226778,
+      "loss": 2.6168,
+      "step": 6679
+    },
+    {
+      "epoch": 0.5391009603744653,
+      "grad_norm": 0.6713843941688538,
+      "learning_rate": 0.00015072738469407127,
+      "loss": 2.5736,
+      "step": 6680
+    },
+    {
+      "epoch": 0.5391816641110483,
+      "grad_norm": 0.7101294994354248,
+      "learning_rate": 0.00015071377915161578,
+      "loss": 2.6994,
+      "step": 6681
+    },
+    {
+      "epoch": 0.5392623678476314,
+      "grad_norm": 0.7132740020751953,
+      "learning_rate": 0.00015070017234524032,
+      "loss": 2.586,
+      "step": 6682
+    },
+    {
+      "epoch": 0.5393430715842144,
+      "grad_norm": 0.7043401598930359,
+      "learning_rate": 0.00015068656427528402,
+      "loss": 2.6025,
+      "step": 6683
+    },
+    {
+      "epoch": 0.5394237753207973,
+      "grad_norm": 0.6831551194190979,
+      "learning_rate": 0.00015067295494208607,
+      "loss": 2.6183,
+      "step": 6684
+    },
+    {
+      "epoch": 0.5395044790573803,
+      "grad_norm": 0.7066370844841003,
+      "learning_rate": 0.0001506593443459856,
+      "loss": 2.6467,
+      "step": 6685
+    },
+    {
+      "epoch": 0.5395851827939634,
+      "grad_norm": 0.7908033132553101,
+      "learning_rate": 0.0001506457324873219,
+      "loss": 2.6929,
+      "step": 6686
+    },
+    {
+      "epoch": 0.5396658865305464,
+      "grad_norm": 0.7186774611473083,
+      "learning_rate": 0.00015063211936643407,
+      "loss": 2.5841,
+      "step": 6687
+    },
+    {
+      "epoch": 0.5397465902671293,
+      "grad_norm": 0.6634512543678284,
+      "learning_rate": 0.0001506185049836615,
+      "loss": 2.5517,
+      "step": 6688
+    },
+    {
+      "epoch": 0.5398272940037123,
+      "grad_norm": 0.734406590461731,
+      "learning_rate": 0.00015060488933934353,
+      "loss": 2.6317,
+      "step": 6689
+    },
+    {
+      "epoch": 0.5399079977402954,
+      "grad_norm": 0.7754772305488586,
+      "learning_rate": 0.00015059127243381937,
+      "loss": 2.6885,
+      "step": 6690
+    },
+    {
+      "epoch": 0.5399887014768784,
+      "grad_norm": 0.7636603713035583,
+      "learning_rate": 0.00015057765426742848,
+      "loss": 2.5767,
+      "step": 6691
+    },
+    {
+      "epoch": 0.5400694052134614,
+      "grad_norm": 0.6621577143669128,
+      "learning_rate": 0.00015056403484051017,
+      "loss": 2.5905,
+      "step": 6692
+    },
+    {
+      "epoch": 0.5401501089500443,
+      "grad_norm": 0.7605881094932556,
+      "learning_rate": 0.00015055041415340404,
+      "loss": 2.6166,
+      "step": 6693
+    },
+    {
+      "epoch": 0.5402308126866274,
+      "grad_norm": 0.7603485584259033,
+      "learning_rate": 0.0001505367922064494,
+      "loss": 2.6123,
+      "step": 6694
+    },
+    {
+      "epoch": 0.5403115164232104,
+      "grad_norm": 0.7021469473838806,
+      "learning_rate": 0.0001505231689999858,
+      "loss": 2.6754,
+      "step": 6695
+    },
+    {
+      "epoch": 0.5403922201597934,
+      "grad_norm": 0.7291955947875977,
+      "learning_rate": 0.00015050954453435273,
+      "loss": 2.6393,
+      "step": 6696
+    },
+    {
+      "epoch": 0.5404729238963764,
+      "grad_norm": 0.6658700704574585,
+      "learning_rate": 0.00015049591880988977,
+      "loss": 2.5888,
+      "step": 6697
+    },
+    {
+      "epoch": 0.5405536276329594,
+      "grad_norm": 0.7080146074295044,
+      "learning_rate": 0.00015048229182693657,
+      "loss": 2.6318,
+      "step": 6698
+    },
+    {
+      "epoch": 0.5406343313695424,
+      "grad_norm": 0.7440849542617798,
+      "learning_rate": 0.00015046866358583267,
+      "loss": 2.596,
+      "step": 6699
+    },
+    {
+      "epoch": 0.5407150351061254,
+      "grad_norm": 0.886578381061554,
+      "learning_rate": 0.00015045503408691775,
+      "loss": 2.6479,
+      "step": 6700
+    },
+    {
+      "epoch": 0.5407957388427084,
+      "grad_norm": 0.7221408486366272,
+      "learning_rate": 0.00015044140333053148,
+      "loss": 2.625,
+      "step": 6701
+    },
+    {
+      "epoch": 0.5408764425792915,
+      "grad_norm": 0.7193209528923035,
+      "learning_rate": 0.0001504277713170136,
+      "loss": 2.6044,
+      "step": 6702
+    },
+    {
+      "epoch": 0.5409571463158744,
+      "grad_norm": 0.7139819860458374,
+      "learning_rate": 0.00015041413804670384,
+      "loss": 2.5572,
+      "step": 6703
+    },
+    {
+      "epoch": 0.5410378500524574,
+      "grad_norm": 0.728875994682312,
+      "learning_rate": 0.00015040050351994196,
+      "loss": 2.6373,
+      "step": 6704
+    },
+    {
+      "epoch": 0.5411185537890404,
+      "grad_norm": 0.6794858574867249,
+      "learning_rate": 0.0001503868677370678,
+      "loss": 2.6265,
+      "step": 6705
+    },
+    {
+      "epoch": 0.5411992575256235,
+      "grad_norm": 0.6874774098396301,
+      "learning_rate": 0.00015037323069842117,
+      "loss": 2.6146,
+      "step": 6706
+    },
+    {
+      "epoch": 0.5412799612622065,
+      "grad_norm": 0.7064409255981445,
+      "learning_rate": 0.00015035959240434197,
+      "loss": 2.6126,
+      "step": 6707
+    },
+    {
+      "epoch": 0.5413606649987894,
+      "grad_norm": 0.7212977409362793,
+      "learning_rate": 0.00015034595285517006,
+      "loss": 2.6836,
+      "step": 6708
+    },
+    {
+      "epoch": 0.5414413687353724,
+      "grad_norm": 0.7826492190361023,
+      "learning_rate": 0.0001503323120512454,
+      "loss": 2.6648,
+      "step": 6709
+    },
+    {
+      "epoch": 0.5415220724719555,
+      "grad_norm": 0.7228415608406067,
+      "learning_rate": 0.000150318669992908,
+      "loss": 2.5734,
+      "step": 6710
+    },
+    {
+      "epoch": 0.5416027762085385,
+      "grad_norm": 0.6929590702056885,
+      "learning_rate": 0.00015030502668049778,
+      "loss": 2.6023,
+      "step": 6711
+    },
+    {
+      "epoch": 0.5416834799451214,
+      "grad_norm": 0.679990291595459,
+      "learning_rate": 0.0001502913821143548,
+      "loss": 2.5867,
+      "step": 6712
+    },
+    {
+      "epoch": 0.5417641836817044,
+      "grad_norm": 0.7324180603027344,
+      "learning_rate": 0.00015027773629481907,
+      "loss": 2.5722,
+      "step": 6713
+    },
+    {
+      "epoch": 0.5418448874182875,
+      "grad_norm": 0.686826765537262,
+      "learning_rate": 0.00015026408922223078,
+      "loss": 2.6138,
+      "step": 6714
+    },
+    {
+      "epoch": 0.5419255911548705,
+      "grad_norm": 0.7045193314552307,
+      "learning_rate": 0.00015025044089693,
+      "loss": 2.619,
+      "step": 6715
+    },
+    {
+      "epoch": 0.5420062948914535,
+      "grad_norm": 0.6839936375617981,
+      "learning_rate": 0.00015023679131925683,
+      "loss": 2.5778,
+      "step": 6716
+    },
+    {
+      "epoch": 0.5420869986280364,
+      "grad_norm": 0.7613961696624756,
+      "learning_rate": 0.00015022314048955153,
+      "loss": 2.6262,
+      "step": 6717
+    },
+    {
+      "epoch": 0.5421677023646195,
+      "grad_norm": 0.7867478728294373,
+      "learning_rate": 0.00015020948840815428,
+      "loss": 2.6576,
+      "step": 6718
+    },
+    {
+      "epoch": 0.5422484061012025,
+      "grad_norm": 0.7371038794517517,
+      "learning_rate": 0.0001501958350754053,
+      "loss": 2.6495,
+      "step": 6719
+    },
+    {
+      "epoch": 0.5423291098377855,
+      "grad_norm": 0.7146512269973755,
+      "learning_rate": 0.00015018218049164494,
+      "loss": 2.6514,
+      "step": 6720
+    },
+    {
+      "epoch": 0.5424098135743685,
+      "grad_norm": 0.7507650256156921,
+      "learning_rate": 0.00015016852465721346,
+      "loss": 2.6509,
+      "step": 6721
+    },
+    {
+      "epoch": 0.5424905173109515,
+      "grad_norm": 0.6786547303199768,
+      "learning_rate": 0.0001501548675724512,
+      "loss": 2.5983,
+      "step": 6722
+    },
+    {
+      "epoch": 0.5425712210475345,
+      "grad_norm": 0.7077932357788086,
+      "learning_rate": 0.0001501412092376985,
+      "loss": 2.622,
+      "step": 6723
+    },
+    {
+      "epoch": 0.5426519247841175,
+      "grad_norm": 0.7191271781921387,
+      "learning_rate": 0.00015012754965329584,
+      "loss": 2.6632,
+      "step": 6724
+    },
+    {
+      "epoch": 0.5427326285207005,
+      "grad_norm": 0.6785906553268433,
+      "learning_rate": 0.00015011388881958356,
+      "loss": 2.6312,
+      "step": 6725
+    },
+    {
+      "epoch": 0.5428133322572836,
+      "grad_norm": 0.6880263090133667,
+      "learning_rate": 0.00015010022673690222,
+      "loss": 2.5951,
+      "step": 6726
+    },
+    {
+      "epoch": 0.5428940359938665,
+      "grad_norm": 0.7769095301628113,
+      "learning_rate": 0.0001500865634055923,
+      "loss": 2.5503,
+      "step": 6727
+    },
+    {
+      "epoch": 0.5429747397304495,
+      "grad_norm": 0.6847476959228516,
+      "learning_rate": 0.0001500728988259942,
+      "loss": 2.6824,
+      "step": 6728
+    },
+    {
+      "epoch": 0.5430554434670325,
+      "grad_norm": 0.6829310059547424,
+      "learning_rate": 0.00015005923299844863,
+      "loss": 2.5683,
+      "step": 6729
+    },
+    {
+      "epoch": 0.5431361472036156,
+      "grad_norm": 0.7436082363128662,
+      "learning_rate": 0.0001500455659232961,
+      "loss": 2.6165,
+      "step": 6730
+    },
+    {
+      "epoch": 0.5432168509401986,
+      "grad_norm": 0.7876375913619995,
+      "learning_rate": 0.00015003189760087724,
+      "loss": 2.6203,
+      "step": 6731
+    },
+    {
+      "epoch": 0.5432975546767815,
+      "grad_norm": 0.6869253516197205,
+      "learning_rate": 0.0001500182280315327,
+      "loss": 2.6136,
+      "step": 6732
+    },
+    {
+      "epoch": 0.5433782584133645,
+      "grad_norm": 0.7179432511329651,
+      "learning_rate": 0.00015000455721560316,
+      "loss": 2.6049,
+      "step": 6733
+    },
+    {
+      "epoch": 0.5434589621499475,
+      "grad_norm": 0.7286917567253113,
+      "learning_rate": 0.00014999088515342939,
+      "loss": 2.5704,
+      "step": 6734
+    },
+    {
+      "epoch": 0.5435396658865306,
+      "grad_norm": 0.6841779351234436,
+      "learning_rate": 0.00014997721184535206,
+      "loss": 2.6095,
+      "step": 6735
+    },
+    {
+      "epoch": 0.5436203696231136,
+      "grad_norm": 0.7661791443824768,
+      "learning_rate": 0.00014996353729171196,
+      "loss": 2.6193,
+      "step": 6736
+    },
+    {
+      "epoch": 0.5437010733596965,
+      "grad_norm": 0.7365885376930237,
+      "learning_rate": 0.0001499498614928499,
+      "loss": 2.586,
+      "step": 6737
+    },
+    {
+      "epoch": 0.5437817770962795,
+      "grad_norm": 0.7423815131187439,
+      "learning_rate": 0.00014993618444910674,
+      "loss": 2.6199,
+      "step": 6738
+    },
+    {
+      "epoch": 0.5438624808328626,
+      "grad_norm": 0.7667781114578247,
+      "learning_rate": 0.0001499225061608233,
+      "loss": 2.6584,
+      "step": 6739
+    },
+    {
+      "epoch": 0.5439431845694456,
+      "grad_norm": 0.7148830890655518,
+      "learning_rate": 0.00014990882662834057,
+      "loss": 2.7172,
+      "step": 6740
+    },
+    {
+      "epoch": 0.5440238883060285,
+      "grad_norm": 0.7206205725669861,
+      "learning_rate": 0.00014989514585199936,
+      "loss": 2.5682,
+      "step": 6741
+    },
+    {
+      "epoch": 0.5441045920426115,
+      "grad_norm": 0.7306448221206665,
+      "learning_rate": 0.0001498814638321407,
+      "loss": 2.6724,
+      "step": 6742
+    },
+    {
+      "epoch": 0.5441852957791946,
+      "grad_norm": 0.7058824896812439,
+      "learning_rate": 0.00014986778056910556,
+      "loss": 2.6573,
+      "step": 6743
+    },
+    {
+      "epoch": 0.5442659995157776,
+      "grad_norm": 0.770588755607605,
+      "learning_rate": 0.000149854096063235,
+      "loss": 2.658,
+      "step": 6744
+    },
+    {
+      "epoch": 0.5443467032523606,
+      "grad_norm": 0.8283931612968445,
+      "learning_rate": 0.00014984041031487001,
+      "loss": 2.6624,
+      "step": 6745
+    },
+    {
+      "epoch": 0.5444274069889435,
+      "grad_norm": 0.6814693808555603,
+      "learning_rate": 0.00014982672332435176,
+      "loss": 2.5835,
+      "step": 6746
+    },
+    {
+      "epoch": 0.5445081107255266,
+      "grad_norm": 0.7059363722801208,
+      "learning_rate": 0.00014981303509202127,
+      "loss": 2.5977,
+      "step": 6747
+    },
+    {
+      "epoch": 0.5445888144621096,
+      "grad_norm": 0.6678106188774109,
+      "learning_rate": 0.00014979934561821975,
+      "loss": 2.6479,
+      "step": 6748
+    },
+    {
+      "epoch": 0.5446695181986926,
+      "grad_norm": 0.8167592883110046,
+      "learning_rate": 0.00014978565490328835,
+      "loss": 2.6529,
+      "step": 6749
+    },
+    {
+      "epoch": 0.5447502219352756,
+      "grad_norm": 0.807209849357605,
+      "learning_rate": 0.00014977196294756832,
+      "loss": 2.6546,
+      "step": 6750
+    },
+    {
+      "epoch": 0.5448309256718586,
+      "grad_norm": 0.7099517583847046,
+      "learning_rate": 0.00014975826975140085,
+      "loss": 2.6178,
+      "step": 6751
+    },
+    {
+      "epoch": 0.5449116294084416,
+      "grad_norm": 0.7900758981704712,
+      "learning_rate": 0.0001497445753151272,
+      "loss": 2.586,
+      "step": 6752
+    },
+    {
+      "epoch": 0.5449923331450246,
+      "grad_norm": 0.6826134920120239,
+      "learning_rate": 0.00014973087963908875,
+      "loss": 2.5914,
+      "step": 6753
+    },
+    {
+      "epoch": 0.5450730368816076,
+      "grad_norm": 0.7383863925933838,
+      "learning_rate": 0.0001497171827236268,
+      "loss": 2.6357,
+      "step": 6754
+    },
+    {
+      "epoch": 0.5451537406181907,
+      "grad_norm": 0.7208051085472107,
+      "learning_rate": 0.0001497034845690826,
+      "loss": 2.5435,
+      "step": 6755
+    },
+    {
+      "epoch": 0.5452344443547736,
+      "grad_norm": 0.680794894695282,
+      "learning_rate": 0.00014968978517579772,
+      "loss": 2.5691,
+      "step": 6756
+    },
+    {
+      "epoch": 0.5453151480913566,
+      "grad_norm": 0.680759847164154,
+      "learning_rate": 0.00014967608454411347,
+      "loss": 2.5761,
+      "step": 6757
+    },
+    {
+      "epoch": 0.5453958518279396,
+      "grad_norm": 0.719634473323822,
+      "learning_rate": 0.00014966238267437134,
+      "loss": 2.637,
+      "step": 6758
+    },
+    {
+      "epoch": 0.5454765555645227,
+      "grad_norm": 0.777302086353302,
+      "learning_rate": 0.0001496486795669128,
+      "loss": 2.6457,
+      "step": 6759
+    },
+    {
+      "epoch": 0.5455572593011057,
+      "grad_norm": 0.6875059604644775,
+      "learning_rate": 0.0001496349752220794,
+      "loss": 2.6116,
+      "step": 6760
+    },
+    {
+      "epoch": 0.5456379630376886,
+      "grad_norm": 0.6884258985519409,
+      "learning_rate": 0.0001496212696402127,
+      "loss": 2.5863,
+      "step": 6761
+    },
+    {
+      "epoch": 0.5457186667742716,
+      "grad_norm": 0.6667922139167786,
+      "learning_rate": 0.00014960756282165422,
+      "loss": 2.5892,
+      "step": 6762
+    },
+    {
+      "epoch": 0.5457993705108547,
+      "grad_norm": 0.6712725162506104,
+      "learning_rate": 0.00014959385476674559,
+      "loss": 2.5478,
+      "step": 6763
+    },
+    {
+      "epoch": 0.5458800742474377,
+      "grad_norm": 0.6803874969482422,
+      "learning_rate": 0.00014958014547582845,
+      "loss": 2.5785,
+      "step": 6764
+    },
+    {
+      "epoch": 0.5459607779840207,
+      "grad_norm": 0.6975811123847961,
+      "learning_rate": 0.0001495664349492445,
+      "loss": 2.5765,
+      "step": 6765
+    },
+    {
+      "epoch": 0.5460414817206036,
+      "grad_norm": 0.7676273584365845,
+      "learning_rate": 0.00014955272318733544,
+      "loss": 2.634,
+      "step": 6766
+    },
+    {
+      "epoch": 0.5461221854571867,
+      "grad_norm": 0.7044547200202942,
+      "learning_rate": 0.000149539010190443,
+      "loss": 2.646,
+      "step": 6767
+    },
+    {
+      "epoch": 0.5462028891937697,
+      "grad_norm": 0.7453166842460632,
+      "learning_rate": 0.00014952529595890887,
+      "loss": 2.6137,
+      "step": 6768
+    },
+    {
+      "epoch": 0.5462835929303527,
+      "grad_norm": 0.7281681299209595,
+      "learning_rate": 0.00014951158049307493,
+      "loss": 2.6558,
+      "step": 6769
+    },
+    {
+      "epoch": 0.5463642966669356,
+      "grad_norm": 0.7131047248840332,
+      "learning_rate": 0.00014949786379328298,
+      "loss": 2.6441,
+      "step": 6770
+    },
+    {
+      "epoch": 0.5464450004035187,
+      "grad_norm": 0.7072219848632812,
+      "learning_rate": 0.00014948414585987487,
+      "loss": 2.5861,
+      "step": 6771
+    },
+    {
+      "epoch": 0.5465257041401017,
+      "grad_norm": 0.7270335555076599,
+      "learning_rate": 0.00014947042669319252,
+      "loss": 2.6703,
+      "step": 6772
+    },
+    {
+      "epoch": 0.5466064078766847,
+      "grad_norm": 0.7314150929450989,
+      "learning_rate": 0.0001494567062935778,
+      "loss": 2.6101,
+      "step": 6773
+    },
+    {
+      "epoch": 0.5466871116132677,
+      "grad_norm": 0.8168460130691528,
+      "learning_rate": 0.00014944298466137266,
+      "loss": 2.662,
+      "step": 6774
+    },
+    {
+      "epoch": 0.5467678153498507,
+      "grad_norm": 0.7338390350341797,
+      "learning_rate": 0.00014942926179691913,
+      "loss": 2.6481,
+      "step": 6775
+    },
+    {
+      "epoch": 0.5468485190864337,
+      "grad_norm": 0.7065639495849609,
+      "learning_rate": 0.00014941553770055917,
+      "loss": 2.6192,
+      "step": 6776
+    },
+    {
+      "epoch": 0.5469292228230167,
+      "grad_norm": 0.7675396203994751,
+      "learning_rate": 0.00014940181237263483,
+      "loss": 2.5828,
+      "step": 6777
+    },
+    {
+      "epoch": 0.5470099265595997,
+      "grad_norm": 0.7085692286491394,
+      "learning_rate": 0.0001493880858134882,
+      "loss": 2.5815,
+      "step": 6778
+    },
+    {
+      "epoch": 0.5470906302961828,
+      "grad_norm": 0.757591187953949,
+      "learning_rate": 0.00014937435802346135,
+      "loss": 2.691,
+      "step": 6779
+    },
+    {
+      "epoch": 0.5471713340327657,
+      "grad_norm": 0.7299168705940247,
+      "learning_rate": 0.00014936062900289647,
+      "loss": 2.6246,
+      "step": 6780
+    },
+    {
+      "epoch": 0.5472520377693487,
+      "grad_norm": 0.693692684173584,
+      "learning_rate": 0.00014934689875213564,
+      "loss": 2.6149,
+      "step": 6781
+    },
+    {
+      "epoch": 0.5473327415059317,
+      "grad_norm": 0.733657956123352,
+      "learning_rate": 0.00014933316727152113,
+      "loss": 2.582,
+      "step": 6782
+    },
+    {
+      "epoch": 0.5474134452425147,
+      "grad_norm": 0.6881953477859497,
+      "learning_rate": 0.00014931943456139514,
+      "loss": 2.6023,
+      "step": 6783
+    },
+    {
+      "epoch": 0.5474941489790978,
+      "grad_norm": 0.7102411985397339,
+      "learning_rate": 0.00014930570062209988,
+      "loss": 2.6296,
+      "step": 6784
+    },
+    {
+      "epoch": 0.5475748527156807,
+      "grad_norm": 0.7263364791870117,
+      "learning_rate": 0.00014929196545397771,
+      "loss": 2.6414,
+      "step": 6785
+    },
+    {
+      "epoch": 0.5476555564522637,
+      "grad_norm": 0.7239066958427429,
+      "learning_rate": 0.00014927822905737092,
+      "loss": 2.6174,
+      "step": 6786
+    },
+    {
+      "epoch": 0.5477362601888467,
+      "grad_norm": 0.6909911632537842,
+      "learning_rate": 0.0001492644914326218,
+      "loss": 2.6036,
+      "step": 6787
+    },
+    {
+      "epoch": 0.5478169639254298,
+      "grad_norm": 0.719693124294281,
+      "learning_rate": 0.00014925075258007283,
+      "loss": 2.6507,
+      "step": 6788
+    },
+    {
+      "epoch": 0.5478976676620128,
+      "grad_norm": 0.7722225785255432,
+      "learning_rate": 0.0001492370125000663,
+      "loss": 2.6268,
+      "step": 6789
+    },
+    {
+      "epoch": 0.5479783713985957,
+      "grad_norm": 0.7456568479537964,
+      "learning_rate": 0.00014922327119294476,
+      "loss": 2.6426,
+      "step": 6790
+    },
+    {
+      "epoch": 0.5480590751351787,
+      "grad_norm": 0.7430242300033569,
+      "learning_rate": 0.00014920952865905062,
+      "loss": 2.6632,
+      "step": 6791
+    },
+    {
+      "epoch": 0.5481397788717618,
+      "grad_norm": 0.7363260388374329,
+      "learning_rate": 0.0001491957848987264,
+      "loss": 2.6021,
+      "step": 6792
+    },
+    {
+      "epoch": 0.5482204826083448,
+      "grad_norm": 0.6903972029685974,
+      "learning_rate": 0.00014918203991231462,
+      "loss": 2.6086,
+      "step": 6793
+    },
+    {
+      "epoch": 0.5483011863449277,
+      "grad_norm": 0.6765161752700806,
+      "learning_rate": 0.00014916829370015781,
+      "loss": 2.5806,
+      "step": 6794
+    },
+    {
+      "epoch": 0.5483818900815107,
+      "grad_norm": 0.7533403635025024,
+      "learning_rate": 0.0001491545462625986,
+      "loss": 2.6351,
+      "step": 6795
+    },
+    {
+      "epoch": 0.5484625938180938,
+      "grad_norm": 0.6841829419136047,
+      "learning_rate": 0.00014914079759997963,
+      "loss": 2.606,
+      "step": 6796
+    },
+    {
+      "epoch": 0.5485432975546768,
+      "grad_norm": 0.7671411037445068,
+      "learning_rate": 0.00014912704771264353,
+      "loss": 2.6645,
+      "step": 6797
+    },
+    {
+      "epoch": 0.5486240012912598,
+      "grad_norm": 0.7218797206878662,
+      "learning_rate": 0.00014911329660093295,
+      "loss": 2.6302,
+      "step": 6798
+    },
+    {
+      "epoch": 0.5487047050278427,
+      "grad_norm": 0.7269994020462036,
+      "learning_rate": 0.00014909954426519067,
+      "loss": 2.6261,
+      "step": 6799
+    },
+    {
+      "epoch": 0.5487854087644258,
+      "grad_norm": 0.765353262424469,
+      "learning_rate": 0.00014908579070575936,
+      "loss": 2.5787,
+      "step": 6800
+    },
+    {
+      "epoch": 0.5488661125010088,
+      "grad_norm": 0.6503065228462219,
+      "learning_rate": 0.00014907203592298189,
+      "loss": 2.6404,
+      "step": 6801
+    },
+    {
+      "epoch": 0.5489468162375918,
+      "grad_norm": 0.6869633197784424,
+      "learning_rate": 0.00014905827991720097,
+      "loss": 2.6463,
+      "step": 6802
+    },
+    {
+      "epoch": 0.5490275199741748,
+      "grad_norm": 0.7221426963806152,
+      "learning_rate": 0.00014904452268875947,
+      "loss": 2.6686,
+      "step": 6803
+    },
+    {
+      "epoch": 0.5491082237107578,
+      "grad_norm": 0.6781399250030518,
+      "learning_rate": 0.00014903076423800028,
+      "loss": 2.6274,
+      "step": 6804
+    },
+    {
+      "epoch": 0.5491889274473408,
+      "grad_norm": 0.7451084852218628,
+      "learning_rate": 0.00014901700456526626,
+      "loss": 2.6449,
+      "step": 6805
+    },
+    {
+      "epoch": 0.5492696311839238,
+      "grad_norm": 0.7159574627876282,
+      "learning_rate": 0.0001490032436709004,
+      "loss": 2.6664,
+      "step": 6806
+    },
+    {
+      "epoch": 0.5493503349205068,
+      "grad_norm": 0.724039614200592,
+      "learning_rate": 0.00014898948155524558,
+      "loss": 2.5816,
+      "step": 6807
+    },
+    {
+      "epoch": 0.5494310386570899,
+      "grad_norm": 0.7194633483886719,
+      "learning_rate": 0.0001489757182186448,
+      "loss": 2.5625,
+      "step": 6808
+    },
+    {
+      "epoch": 0.5495117423936728,
+      "grad_norm": 0.704133927822113,
+      "learning_rate": 0.0001489619536614411,
+      "loss": 2.6295,
+      "step": 6809
+    },
+    {
+      "epoch": 0.5495924461302558,
+      "grad_norm": 0.6717158555984497,
+      "learning_rate": 0.00014894818788397757,
+      "loss": 2.6168,
+      "step": 6810
+    },
+    {
+      "epoch": 0.5496731498668388,
+      "grad_norm": 0.7096573710441589,
+      "learning_rate": 0.0001489344208865972,
+      "loss": 2.6316,
+      "step": 6811
+    },
+    {
+      "epoch": 0.5497538536034219,
+      "grad_norm": 0.6383458375930786,
+      "learning_rate": 0.00014892065266964316,
+      "loss": 2.5577,
+      "step": 6812
+    },
+    {
+      "epoch": 0.5498345573400049,
+      "grad_norm": 0.7606377601623535,
+      "learning_rate": 0.0001489068832334586,
+      "loss": 2.7078,
+      "step": 6813
+    },
+    {
+      "epoch": 0.5499152610765878,
+      "grad_norm": 0.649162232875824,
+      "learning_rate": 0.00014889311257838665,
+      "loss": 2.6023,
+      "step": 6814
+    },
+    {
+      "epoch": 0.5499959648131708,
+      "grad_norm": 0.6445025205612183,
+      "learning_rate": 0.00014887934070477053,
+      "loss": 2.6,
+      "step": 6815
+    },
+    {
+      "epoch": 0.5500766685497539,
+      "grad_norm": 0.6873729825019836,
+      "learning_rate": 0.00014886556761295342,
+      "loss": 2.6398,
+      "step": 6816
+    },
+    {
+      "epoch": 0.5501573722863369,
+      "grad_norm": 0.7814947366714478,
+      "learning_rate": 0.0001488517933032787,
+      "loss": 2.5803,
+      "step": 6817
+    },
+    {
+      "epoch": 0.5502380760229199,
+      "grad_norm": 0.7140909433364868,
+      "learning_rate": 0.00014883801777608953,
+      "loss": 2.6051,
+      "step": 6818
+    },
+    {
+      "epoch": 0.5503187797595028,
+      "grad_norm": 0.7326326370239258,
+      "learning_rate": 0.00014882424103172936,
+      "loss": 2.6123,
+      "step": 6819
+    },
+    {
+      "epoch": 0.5503994834960859,
+      "grad_norm": 0.7093667387962341,
+      "learning_rate": 0.00014881046307054142,
+      "loss": 2.6527,
+      "step": 6820
+    },
+    {
+      "epoch": 0.5504801872326689,
+      "grad_norm": 0.6877567768096924,
+      "learning_rate": 0.00014879668389286915,
+      "loss": 2.6086,
+      "step": 6821
+    },
+    {
+      "epoch": 0.5505608909692519,
+      "grad_norm": 0.7095615863800049,
+      "learning_rate": 0.000148782903499056,
+      "loss": 2.6469,
+      "step": 6822
+    },
+    {
+      "epoch": 0.5506415947058348,
+      "grad_norm": 0.6931191086769104,
+      "learning_rate": 0.00014876912188944535,
+      "loss": 2.6842,
+      "step": 6823
+    },
+    {
+      "epoch": 0.5507222984424179,
+      "grad_norm": 0.7016414403915405,
+      "learning_rate": 0.00014875533906438072,
+      "loss": 2.5753,
+      "step": 6824
+    },
+    {
+      "epoch": 0.5508030021790009,
+      "grad_norm": 0.6813814640045166,
+      "learning_rate": 0.00014874155502420558,
+      "loss": 2.5739,
+      "step": 6825
+    },
+    {
+      "epoch": 0.5508837059155839,
+      "grad_norm": 0.7068608403205872,
+      "learning_rate": 0.00014872776976926347,
+      "loss": 2.6325,
+      "step": 6826
+    },
+    {
+      "epoch": 0.5509644096521669,
+      "grad_norm": 0.6978127360343933,
+      "learning_rate": 0.00014871398329989796,
+      "loss": 2.5614,
+      "step": 6827
+    },
+    {
+      "epoch": 0.55104511338875,
+      "grad_norm": 0.6923051476478577,
+      "learning_rate": 0.00014870019561645265,
+      "loss": 2.6075,
+      "step": 6828
+    },
+    {
+      "epoch": 0.5511258171253329,
+      "grad_norm": 0.6708533763885498,
+      "learning_rate": 0.00014868640671927117,
+      "loss": 2.5883,
+      "step": 6829
+    },
+    {
+      "epoch": 0.5512065208619159,
+      "grad_norm": 0.7679650783538818,
+      "learning_rate": 0.00014867261660869713,
+      "loss": 2.6105,
+      "step": 6830
+    },
+    {
+      "epoch": 0.5512872245984989,
+      "grad_norm": 0.7080917358398438,
+      "learning_rate": 0.0001486588252850743,
+      "loss": 2.5855,
+      "step": 6831
+    },
+    {
+      "epoch": 0.551367928335082,
+      "grad_norm": 0.7218755483627319,
+      "learning_rate": 0.00014864503274874635,
+      "loss": 2.5872,
+      "step": 6832
+    },
+    {
+      "epoch": 0.551448632071665,
+      "grad_norm": 0.689038872718811,
+      "learning_rate": 0.000148631239000057,
+      "loss": 2.5902,
+      "step": 6833
+    },
+    {
+      "epoch": 0.5515293358082479,
+      "grad_norm": 0.6810954213142395,
+      "learning_rate": 0.00014861744403935005,
+      "loss": 2.5938,
+      "step": 6834
+    },
+    {
+      "epoch": 0.5516100395448309,
+      "grad_norm": 0.7509457468986511,
+      "learning_rate": 0.00014860364786696933,
+      "loss": 2.593,
+      "step": 6835
+    },
+    {
+      "epoch": 0.5516907432814139,
+      "grad_norm": 0.739536702632904,
+      "learning_rate": 0.00014858985048325863,
+      "loss": 2.6668,
+      "step": 6836
+    },
+    {
+      "epoch": 0.551771447017997,
+      "grad_norm": 0.661829948425293,
+      "learning_rate": 0.00014857605188856184,
+      "loss": 2.6407,
+      "step": 6837
+    },
+    {
+      "epoch": 0.5518521507545799,
+      "grad_norm": 0.6869735717773438,
+      "learning_rate": 0.00014856225208322287,
+      "loss": 2.535,
+      "step": 6838
+    },
+    {
+      "epoch": 0.5519328544911629,
+      "grad_norm": 0.6724792122840881,
+      "learning_rate": 0.00014854845106758563,
+      "loss": 2.5629,
+      "step": 6839
+    },
+    {
+      "epoch": 0.5520135582277459,
+      "grad_norm": 0.7066503763198853,
+      "learning_rate": 0.00014853464884199407,
+      "loss": 2.6002,
+      "step": 6840
+    },
+    {
+      "epoch": 0.552094261964329,
+      "grad_norm": 0.7354215979576111,
+      "learning_rate": 0.0001485208454067922,
+      "loss": 2.6032,
+      "step": 6841
+    },
+    {
+      "epoch": 0.552174965700912,
+      "grad_norm": 0.8124571442604065,
+      "learning_rate": 0.00014850704076232405,
+      "loss": 2.5884,
+      "step": 6842
+    },
+    {
+      "epoch": 0.5522556694374949,
+      "grad_norm": 0.6941336393356323,
+      "learning_rate": 0.00014849323490893364,
+      "loss": 2.6461,
+      "step": 6843
+    },
+    {
+      "epoch": 0.5523363731740779,
+      "grad_norm": 0.6848790049552917,
+      "learning_rate": 0.00014847942784696505,
+      "loss": 2.6098,
+      "step": 6844
+    },
+    {
+      "epoch": 0.552417076910661,
+      "grad_norm": 0.6688000559806824,
+      "learning_rate": 0.00014846561957676237,
+      "loss": 2.6115,
+      "step": 6845
+    },
+    {
+      "epoch": 0.552497780647244,
+      "grad_norm": 0.6647306084632874,
+      "learning_rate": 0.00014845181009866975,
+      "loss": 2.597,
+      "step": 6846
+    },
+    {
+      "epoch": 0.552578484383827,
+      "grad_norm": 0.7277785539627075,
+      "learning_rate": 0.0001484379994130314,
+      "loss": 2.6223,
+      "step": 6847
+    },
+    {
+      "epoch": 0.5526591881204099,
+      "grad_norm": 0.6623761057853699,
+      "learning_rate": 0.00014842418752019146,
+      "loss": 2.5657,
+      "step": 6848
+    },
+    {
+      "epoch": 0.552739891856993,
+      "grad_norm": 0.7207754254341125,
+      "learning_rate": 0.00014841037442049423,
+      "loss": 2.5711,
+      "step": 6849
+    },
+    {
+      "epoch": 0.552820595593576,
+      "grad_norm": 0.6963560581207275,
+      "learning_rate": 0.00014839656011428389,
+      "loss": 2.6078,
+      "step": 6850
+    },
+    {
+      "epoch": 0.552901299330159,
+      "grad_norm": 0.6875078678131104,
+      "learning_rate": 0.00014838274460190475,
+      "loss": 2.6109,
+      "step": 6851
+    },
+    {
+      "epoch": 0.552982003066742,
+      "grad_norm": 0.7049943804740906,
+      "learning_rate": 0.00014836892788370118,
+      "loss": 2.5755,
+      "step": 6852
+    },
+    {
+      "epoch": 0.553062706803325,
+      "grad_norm": 0.6941191554069519,
+      "learning_rate": 0.00014835510996001744,
+      "loss": 2.6694,
+      "step": 6853
+    },
+    {
+      "epoch": 0.553143410539908,
+      "grad_norm": 0.7589484453201294,
+      "learning_rate": 0.000148341290831198,
+      "loss": 2.5677,
+      "step": 6854
+    },
+    {
+      "epoch": 0.553224114276491,
+      "grad_norm": 0.6594784259796143,
+      "learning_rate": 0.00014832747049758723,
+      "loss": 2.6209,
+      "step": 6855
+    },
+    {
+      "epoch": 0.553304818013074,
+      "grad_norm": 0.726598858833313,
+      "learning_rate": 0.00014831364895952952,
+      "loss": 2.6492,
+      "step": 6856
+    },
+    {
+      "epoch": 0.553385521749657,
+      "grad_norm": 0.6668030023574829,
+      "learning_rate": 0.0001482998262173694,
+      "loss": 2.6057,
+      "step": 6857
+    },
+    {
+      "epoch": 0.55346622548624,
+      "grad_norm": 0.7698997855186462,
+      "learning_rate": 0.0001482860022714514,
+      "loss": 2.6215,
+      "step": 6858
+    },
+    {
+      "epoch": 0.553546929222823,
+      "grad_norm": 0.6805251836776733,
+      "learning_rate": 0.00014827217712211997,
+      "loss": 2.5855,
+      "step": 6859
+    },
+    {
+      "epoch": 0.553627632959406,
+      "grad_norm": 0.8481020331382751,
+      "learning_rate": 0.00014825835076971968,
+      "loss": 2.6218,
+      "step": 6860
+    },
+    {
+      "epoch": 0.5537083366959891,
+      "grad_norm": 0.6801722645759583,
+      "learning_rate": 0.00014824452321459517,
+      "loss": 2.5998,
+      "step": 6861
+    },
+    {
+      "epoch": 0.553789040432572,
+      "grad_norm": 0.7174597978591919,
+      "learning_rate": 0.00014823069445709104,
+      "loss": 2.5782,
+      "step": 6862
+    },
+    {
+      "epoch": 0.553869744169155,
+      "grad_norm": 0.7607117891311646,
+      "learning_rate": 0.0001482168644975519,
+      "loss": 2.6492,
+      "step": 6863
+    },
+    {
+      "epoch": 0.553950447905738,
+      "grad_norm": 0.7554265856742859,
+      "learning_rate": 0.00014820303333632246,
+      "loss": 2.6511,
+      "step": 6864
+    },
+    {
+      "epoch": 0.5540311516423211,
+      "grad_norm": 0.7520260214805603,
+      "learning_rate": 0.00014818920097374745,
+      "loss": 2.6258,
+      "step": 6865
+    },
+    {
+      "epoch": 0.5541118553789041,
+      "grad_norm": 0.7897995114326477,
+      "learning_rate": 0.00014817536741017152,
+      "loss": 2.6153,
+      "step": 6866
+    },
+    {
+      "epoch": 0.554192559115487,
+      "grad_norm": 0.7444615960121155,
+      "learning_rate": 0.00014816153264593957,
+      "loss": 2.5892,
+      "step": 6867
+    },
+    {
+      "epoch": 0.55427326285207,
+      "grad_norm": 0.6593222618103027,
+      "learning_rate": 0.0001481476966813963,
+      "loss": 2.6048,
+      "step": 6868
+    },
+    {
+      "epoch": 0.5543539665886531,
+      "grad_norm": 0.7517102360725403,
+      "learning_rate": 0.0001481338595168866,
+      "loss": 2.6496,
+      "step": 6869
+    },
+    {
+      "epoch": 0.5544346703252361,
+      "grad_norm": 0.7314056754112244,
+      "learning_rate": 0.00014812002115275529,
+      "loss": 2.6009,
+      "step": 6870
+    },
+    {
+      "epoch": 0.554515374061819,
+      "grad_norm": 0.6718037724494934,
+      "learning_rate": 0.00014810618158934722,
+      "loss": 2.6279,
+      "step": 6871
+    },
+    {
+      "epoch": 0.554596077798402,
+      "grad_norm": 0.6853529810905457,
+      "learning_rate": 0.00014809234082700735,
+      "loss": 2.6562,
+      "step": 6872
+    },
+    {
+      "epoch": 0.5546767815349851,
+      "grad_norm": 0.713599443435669,
+      "learning_rate": 0.0001480784988660807,
+      "loss": 2.5783,
+      "step": 6873
+    },
+    {
+      "epoch": 0.5547574852715681,
+      "grad_norm": 0.6820243000984192,
+      "learning_rate": 0.00014806465570691213,
+      "loss": 2.5753,
+      "step": 6874
+    },
+    {
+      "epoch": 0.5548381890081511,
+      "grad_norm": 0.6999152302742004,
+      "learning_rate": 0.00014805081134984673,
+      "loss": 2.5839,
+      "step": 6875
+    },
+    {
+      "epoch": 0.554918892744734,
+      "grad_norm": 0.7145923376083374,
+      "learning_rate": 0.00014803696579522948,
+      "loss": 2.6153,
+      "step": 6876
+    },
+    {
+      "epoch": 0.5549995964813171,
+      "grad_norm": 0.7569223046302795,
+      "learning_rate": 0.00014802311904340548,
+      "loss": 2.5879,
+      "step": 6877
+    },
+    {
+      "epoch": 0.5550803002179001,
+      "grad_norm": 0.6977131962776184,
+      "learning_rate": 0.00014800927109471983,
+      "loss": 2.6587,
+      "step": 6878
+    },
+    {
+      "epoch": 0.5551610039544831,
+      "grad_norm": 0.6693562865257263,
+      "learning_rate": 0.00014799542194951764,
+      "loss": 2.6271,
+      "step": 6879
+    },
+    {
+      "epoch": 0.5552417076910661,
+      "grad_norm": 0.6937456130981445,
+      "learning_rate": 0.00014798157160814406,
+      "loss": 2.6213,
+      "step": 6880
+    },
+    {
+      "epoch": 0.5553224114276492,
+      "grad_norm": 0.761538565158844,
+      "learning_rate": 0.0001479677200709443,
+      "loss": 2.6053,
+      "step": 6881
+    },
+    {
+      "epoch": 0.5554031151642321,
+      "grad_norm": 0.707457959651947,
+      "learning_rate": 0.00014795386733826356,
+      "loss": 2.5763,
+      "step": 6882
+    },
+    {
+      "epoch": 0.5554838189008151,
+      "grad_norm": 0.7323198318481445,
+      "learning_rate": 0.0001479400134104471,
+      "loss": 2.6899,
+      "step": 6883
+    },
+    {
+      "epoch": 0.5555645226373981,
+      "grad_norm": 0.7181541323661804,
+      "learning_rate": 0.0001479261582878402,
+      "loss": 2.5743,
+      "step": 6884
+    },
+    {
+      "epoch": 0.5556452263739811,
+      "grad_norm": 0.7683241367340088,
+      "learning_rate": 0.00014791230197078813,
+      "loss": 2.5295,
+      "step": 6885
+    },
+    {
+      "epoch": 0.5557259301105641,
+      "grad_norm": 0.7248150706291199,
+      "learning_rate": 0.00014789844445963626,
+      "loss": 2.6131,
+      "step": 6886
+    },
+    {
+      "epoch": 0.5558066338471471,
+      "grad_norm": 0.6868402361869812,
+      "learning_rate": 0.00014788458575472997,
+      "loss": 2.6182,
+      "step": 6887
+    },
+    {
+      "epoch": 0.5558873375837301,
+      "grad_norm": 0.6995798945426941,
+      "learning_rate": 0.0001478707258564146,
+      "loss": 2.5969,
+      "step": 6888
+    },
+    {
+      "epoch": 0.5559680413203131,
+      "grad_norm": 0.6912558078765869,
+      "learning_rate": 0.00014785686476503565,
+      "loss": 2.6264,
+      "step": 6889
+    },
+    {
+      "epoch": 0.5560487450568962,
+      "grad_norm": 0.7485123872756958,
+      "learning_rate": 0.00014784300248093848,
+      "loss": 2.6036,
+      "step": 6890
+    },
+    {
+      "epoch": 0.5561294487934791,
+      "grad_norm": 0.7150819897651672,
+      "learning_rate": 0.00014782913900446864,
+      "loss": 2.5807,
+      "step": 6891
+    },
+    {
+      "epoch": 0.5562101525300621,
+      "grad_norm": 0.6715224385261536,
+      "learning_rate": 0.00014781527433597167,
+      "loss": 2.6164,
+      "step": 6892
+    },
+    {
+      "epoch": 0.5562908562666451,
+      "grad_norm": 0.6951256394386292,
+      "learning_rate": 0.000147801408475793,
+      "loss": 2.6106,
+      "step": 6893
+    },
+    {
+      "epoch": 0.5563715600032282,
+      "grad_norm": 0.7296997904777527,
+      "learning_rate": 0.00014778754142427832,
+      "loss": 2.6182,
+      "step": 6894
+    },
+    {
+      "epoch": 0.5564522637398112,
+      "grad_norm": 0.7484713196754456,
+      "learning_rate": 0.0001477736731817732,
+      "loss": 2.6384,
+      "step": 6895
+    },
+    {
+      "epoch": 0.5565329674763941,
+      "grad_norm": 0.6967526078224182,
+      "learning_rate": 0.00014775980374862326,
+      "loss": 2.5889,
+      "step": 6896
+    },
+    {
+      "epoch": 0.5566136712129771,
+      "grad_norm": 0.7004885077476501,
+      "learning_rate": 0.00014774593312517415,
+      "loss": 2.6549,
+      "step": 6897
+    },
+    {
+      "epoch": 0.5566943749495602,
+      "grad_norm": 0.7069302201271057,
+      "learning_rate": 0.00014773206131177158,
+      "loss": 2.6408,
+      "step": 6898
+    },
+    {
+      "epoch": 0.5567750786861432,
+      "grad_norm": 0.7048566341400146,
+      "learning_rate": 0.00014771818830876127,
+      "loss": 2.5909,
+      "step": 6899
+    },
+    {
+      "epoch": 0.5568557824227262,
+      "grad_norm": 0.7386630773544312,
+      "learning_rate": 0.00014770431411648897,
+      "loss": 2.6402,
+      "step": 6900
+    },
+    {
+      "epoch": 0.5569364861593091,
+      "grad_norm": 0.7244876027107239,
+      "learning_rate": 0.00014769043873530047,
+      "loss": 2.5548,
+      "step": 6901
+    },
+    {
+      "epoch": 0.5570171898958922,
+      "grad_norm": 0.6820651888847351,
+      "learning_rate": 0.00014767656216554156,
+      "loss": 2.682,
+      "step": 6902
+    },
+    {
+      "epoch": 0.5570978936324752,
+      "grad_norm": 0.7281784415245056,
+      "learning_rate": 0.00014766268440755812,
+      "loss": 2.622,
+      "step": 6903
+    },
+    {
+      "epoch": 0.5571785973690582,
+      "grad_norm": 0.6525030136108398,
+      "learning_rate": 0.00014764880546169594,
+      "loss": 2.5809,
+      "step": 6904
+    },
+    {
+      "epoch": 0.5572593011056411,
+      "grad_norm": 0.6735210418701172,
+      "learning_rate": 0.00014763492532830102,
+      "loss": 2.6645,
+      "step": 6905
+    },
+    {
+      "epoch": 0.5573400048422242,
+      "grad_norm": 0.674700140953064,
+      "learning_rate": 0.00014762104400771922,
+      "loss": 2.6466,
+      "step": 6906
+    },
+    {
+      "epoch": 0.5574207085788072,
+      "grad_norm": 0.7570134401321411,
+      "learning_rate": 0.00014760716150029652,
+      "loss": 2.57,
+      "step": 6907
+    },
+    {
+      "epoch": 0.5575014123153902,
+      "grad_norm": 0.6532449722290039,
+      "learning_rate": 0.00014759327780637893,
+      "loss": 2.6207,
+      "step": 6908
+    },
+    {
+      "epoch": 0.5575821160519732,
+      "grad_norm": 0.7697737812995911,
+      "learning_rate": 0.00014757939292631242,
+      "loss": 2.5846,
+      "step": 6909
+    },
+    {
+      "epoch": 0.5576628197885563,
+      "grad_norm": 0.6750194430351257,
+      "learning_rate": 0.00014756550686044308,
+      "loss": 2.6421,
+      "step": 6910
+    },
+    {
+      "epoch": 0.5577435235251392,
+      "grad_norm": 0.7357683777809143,
+      "learning_rate": 0.00014755161960911697,
+      "loss": 2.6173,
+      "step": 6911
+    },
+    {
+      "epoch": 0.5578242272617222,
+      "grad_norm": 0.6812090277671814,
+      "learning_rate": 0.0001475377311726802,
+      "loss": 2.5556,
+      "step": 6912
+    },
+    {
+      "epoch": 0.5579049309983052,
+      "grad_norm": 0.7633040547370911,
+      "learning_rate": 0.00014752384155147888,
+      "loss": 2.6505,
+      "step": 6913
+    },
+    {
+      "epoch": 0.5579856347348883,
+      "grad_norm": 0.7426417469978333,
+      "learning_rate": 0.00014750995074585922,
+      "loss": 2.5575,
+      "step": 6914
+    },
+    {
+      "epoch": 0.5580663384714712,
+      "grad_norm": 0.6926711201667786,
+      "learning_rate": 0.00014749605875616744,
+      "loss": 2.5751,
+      "step": 6915
+    },
+    {
+      "epoch": 0.5581470422080542,
+      "grad_norm": 0.70630943775177,
+      "learning_rate": 0.00014748216558274966,
+      "loss": 2.6228,
+      "step": 6916
+    },
+    {
+      "epoch": 0.5582277459446372,
+      "grad_norm": 0.7183346748352051,
+      "learning_rate": 0.0001474682712259522,
+      "loss": 2.5704,
+      "step": 6917
+    },
+    {
+      "epoch": 0.5583084496812203,
+      "grad_norm": 0.7622792720794678,
+      "learning_rate": 0.00014745437568612136,
+      "loss": 2.6031,
+      "step": 6918
+    },
+    {
+      "epoch": 0.5583891534178033,
+      "grad_norm": 0.6967802047729492,
+      "learning_rate": 0.00014744047896360344,
+      "loss": 2.6031,
+      "step": 6919
+    },
+    {
+      "epoch": 0.5584698571543862,
+      "grad_norm": 0.7827191948890686,
+      "learning_rate": 0.00014742658105874475,
+      "loss": 2.5427,
+      "step": 6920
+    },
+    {
+      "epoch": 0.5585505608909692,
+      "grad_norm": 0.6865705847740173,
+      "learning_rate": 0.0001474126819718917,
+      "loss": 2.6514,
+      "step": 6921
+    },
+    {
+      "epoch": 0.5586312646275523,
+      "grad_norm": 0.7181665897369385,
+      "learning_rate": 0.0001473987817033906,
+      "loss": 2.613,
+      "step": 6922
+    },
+    {
+      "epoch": 0.5587119683641353,
+      "grad_norm": 0.7198463082313538,
+      "learning_rate": 0.00014738488025358806,
+      "loss": 2.6423,
+      "step": 6923
+    },
+    {
+      "epoch": 0.5587926721007183,
+      "grad_norm": 0.773078441619873,
+      "learning_rate": 0.00014737097762283042,
+      "loss": 2.5946,
+      "step": 6924
+    },
+    {
+      "epoch": 0.5588733758373012,
+      "grad_norm": 0.7732799649238586,
+      "learning_rate": 0.00014735707381146416,
+      "loss": 2.6778,
+      "step": 6925
+    },
+    {
+      "epoch": 0.5589540795738843,
+      "grad_norm": 0.7639997601509094,
+      "learning_rate": 0.00014734316881983585,
+      "loss": 2.6064,
+      "step": 6926
+    },
+    {
+      "epoch": 0.5590347833104673,
+      "grad_norm": 0.7912085652351379,
+      "learning_rate": 0.00014732926264829198,
+      "loss": 2.5765,
+      "step": 6927
+    },
+    {
+      "epoch": 0.5591154870470503,
+      "grad_norm": 0.7460121512413025,
+      "learning_rate": 0.0001473153552971792,
+      "loss": 2.6724,
+      "step": 6928
+    },
+    {
+      "epoch": 0.5591961907836333,
+      "grad_norm": 0.6853603720664978,
+      "learning_rate": 0.00014730144676684408,
+      "loss": 2.5846,
+      "step": 6929
+    },
+    {
+      "epoch": 0.5592768945202163,
+      "grad_norm": 0.7368159294128418,
+      "learning_rate": 0.00014728753705763324,
+      "loss": 2.6626,
+      "step": 6930
+    },
+    {
+      "epoch": 0.5593575982567993,
+      "grad_norm": 0.6888907551765442,
+      "learning_rate": 0.0001472736261698934,
+      "loss": 2.6169,
+      "step": 6931
+    },
+    {
+      "epoch": 0.5594383019933823,
+      "grad_norm": 0.6978163719177246,
+      "learning_rate": 0.0001472597141039712,
+      "loss": 2.6367,
+      "step": 6932
+    },
+    {
+      "epoch": 0.5595190057299653,
+      "grad_norm": 0.7829774618148804,
+      "learning_rate": 0.00014724580086021335,
+      "loss": 2.5983,
+      "step": 6933
+    },
+    {
+      "epoch": 0.5595997094665484,
+      "grad_norm": 0.7872018218040466,
+      "learning_rate": 0.0001472318864389667,
+      "loss": 2.5418,
+      "step": 6934
+    },
+    {
+      "epoch": 0.5596804132031313,
+      "grad_norm": 0.6994973421096802,
+      "learning_rate": 0.00014721797084057793,
+      "loss": 2.6062,
+      "step": 6935
+    },
+    {
+      "epoch": 0.5597611169397143,
+      "grad_norm": 0.7281144857406616,
+      "learning_rate": 0.00014720405406539394,
+      "loss": 2.573,
+      "step": 6936
+    },
+    {
+      "epoch": 0.5598418206762973,
+      "grad_norm": 0.713513970375061,
+      "learning_rate": 0.0001471901361137615,
+      "loss": 2.6589,
+      "step": 6937
+    },
+    {
+      "epoch": 0.5599225244128803,
+      "grad_norm": 0.7752750515937805,
+      "learning_rate": 0.00014717621698602754,
+      "loss": 2.6478,
+      "step": 6938
+    },
+    {
+      "epoch": 0.5600032281494634,
+      "grad_norm": 0.6876000165939331,
+      "learning_rate": 0.00014716229668253889,
+      "loss": 2.6092,
+      "step": 6939
+    },
+    {
+      "epoch": 0.5600839318860463,
+      "grad_norm": 0.6371028423309326,
+      "learning_rate": 0.00014714837520364256,
+      "loss": 2.606,
+      "step": 6940
+    },
+    {
+      "epoch": 0.5601646356226293,
+      "grad_norm": 0.6488915085792542,
+      "learning_rate": 0.00014713445254968546,
+      "loss": 2.5769,
+      "step": 6941
+    },
+    {
+      "epoch": 0.5602453393592123,
+      "grad_norm": 0.7286413908004761,
+      "learning_rate": 0.00014712052872101458,
+      "loss": 2.6267,
+      "step": 6942
+    },
+    {
+      "epoch": 0.5603260430957954,
+      "grad_norm": 0.6863759160041809,
+      "learning_rate": 0.00014710660371797696,
+      "loss": 2.641,
+      "step": 6943
+    },
+    {
+      "epoch": 0.5604067468323783,
+      "grad_norm": 0.706900417804718,
+      "learning_rate": 0.00014709267754091964,
+      "loss": 2.6344,
+      "step": 6944
+    },
+    {
+      "epoch": 0.5604874505689613,
+      "grad_norm": 0.6462892293930054,
+      "learning_rate": 0.0001470787501901897,
+      "loss": 2.5561,
+      "step": 6945
+    },
+    {
+      "epoch": 0.5605681543055443,
+      "grad_norm": 0.7342472076416016,
+      "learning_rate": 0.00014706482166613425,
+      "loss": 2.583,
+      "step": 6946
+    },
+    {
+      "epoch": 0.5606488580421274,
+      "grad_norm": 0.7132803797721863,
+      "learning_rate": 0.00014705089196910038,
+      "loss": 2.558,
+      "step": 6947
+    },
+    {
+      "epoch": 0.5607295617787104,
+      "grad_norm": 0.7709125876426697,
+      "learning_rate": 0.00014703696109943533,
+      "loss": 2.6165,
+      "step": 6948
+    },
+    {
+      "epoch": 0.5608102655152933,
+      "grad_norm": 0.7108885645866394,
+      "learning_rate": 0.00014702302905748619,
+      "loss": 2.5788,
+      "step": 6949
+    },
+    {
+      "epoch": 0.5608909692518763,
+      "grad_norm": 0.7295591235160828,
+      "learning_rate": 0.0001470090958436003,
+      "loss": 2.6526,
+      "step": 6950
+    },
+    {
+      "epoch": 0.5609716729884594,
+      "grad_norm": 0.7235364317893982,
+      "learning_rate": 0.00014699516145812486,
+      "loss": 2.604,
+      "step": 6951
+    },
+    {
+      "epoch": 0.5610523767250424,
+      "grad_norm": 0.6723269820213318,
+      "learning_rate": 0.00014698122590140714,
+      "loss": 2.5838,
+      "step": 6952
+    },
+    {
+      "epoch": 0.5611330804616254,
+      "grad_norm": 0.7022266983985901,
+      "learning_rate": 0.00014696728917379447,
+      "loss": 2.6086,
+      "step": 6953
+    },
+    {
+      "epoch": 0.5612137841982083,
+      "grad_norm": 0.6923824548721313,
+      "learning_rate": 0.00014695335127563414,
+      "loss": 2.6678,
+      "step": 6954
+    },
+    {
+      "epoch": 0.5612944879347914,
+      "grad_norm": 0.6909339427947998,
+      "learning_rate": 0.0001469394122072736,
+      "loss": 2.6397,
+      "step": 6955
+    },
+    {
+      "epoch": 0.5613751916713744,
+      "grad_norm": 0.710299015045166,
+      "learning_rate": 0.00014692547196906022,
+      "loss": 2.5973,
+      "step": 6956
+    },
+    {
+      "epoch": 0.5614558954079574,
+      "grad_norm": 0.7141178250312805,
+      "learning_rate": 0.00014691153056134136,
+      "loss": 2.6111,
+      "step": 6957
+    },
+    {
+      "epoch": 0.5615365991445403,
+      "grad_norm": 0.6994750499725342,
+      "learning_rate": 0.00014689758798446456,
+      "loss": 2.6498,
+      "step": 6958
+    },
+    {
+      "epoch": 0.5616173028811234,
+      "grad_norm": 0.6951611638069153,
+      "learning_rate": 0.00014688364423877726,
+      "loss": 2.6208,
+      "step": 6959
+    },
+    {
+      "epoch": 0.5616980066177064,
+      "grad_norm": 0.6610642075538635,
+      "learning_rate": 0.000146869699324627,
+      "loss": 2.5725,
+      "step": 6960
+    },
+    {
+      "epoch": 0.5617787103542894,
+      "grad_norm": 0.6771267056465149,
+      "learning_rate": 0.00014685575324236135,
+      "loss": 2.6336,
+      "step": 6961
+    },
+    {
+      "epoch": 0.5618594140908724,
+      "grad_norm": 0.7431008815765381,
+      "learning_rate": 0.0001468418059923278,
+      "loss": 2.6782,
+      "step": 6962
+    },
+    {
+      "epoch": 0.5619401178274555,
+      "grad_norm": 0.7399705648422241,
+      "learning_rate": 0.000146827857574874,
+      "loss": 2.6212,
+      "step": 6963
+    },
+    {
+      "epoch": 0.5620208215640384,
+      "grad_norm": 0.7237067222595215,
+      "learning_rate": 0.00014681390799034763,
+      "loss": 2.6261,
+      "step": 6964
+    },
+    {
+      "epoch": 0.5621015253006214,
+      "grad_norm": 0.7033257484436035,
+      "learning_rate": 0.00014679995723909623,
+      "loss": 2.6912,
+      "step": 6965
+    },
+    {
+      "epoch": 0.5621822290372044,
+      "grad_norm": 0.6953759789466858,
+      "learning_rate": 0.00014678600532146762,
+      "loss": 2.6022,
+      "step": 6966
+    },
+    {
+      "epoch": 0.5622629327737875,
+      "grad_norm": 0.8338057994842529,
+      "learning_rate": 0.0001467720522378094,
+      "loss": 2.595,
+      "step": 6967
+    },
+    {
+      "epoch": 0.5623436365103704,
+      "grad_norm": 0.6506100296974182,
+      "learning_rate": 0.00014675809798846942,
+      "loss": 2.6033,
+      "step": 6968
+    },
+    {
+      "epoch": 0.5624243402469534,
+      "grad_norm": 0.7122468948364258,
+      "learning_rate": 0.0001467441425737954,
+      "loss": 2.56,
+      "step": 6969
+    },
+    {
+      "epoch": 0.5625050439835364,
+      "grad_norm": 0.7012680172920227,
+      "learning_rate": 0.00014673018599413516,
+      "loss": 2.6052,
+      "step": 6970
+    },
+    {
+      "epoch": 0.5625857477201195,
+      "grad_norm": 0.668187141418457,
+      "learning_rate": 0.00014671622824983653,
+      "loss": 2.6675,
+      "step": 6971
+    },
+    {
+      "epoch": 0.5626664514567025,
+      "grad_norm": 0.7259203791618347,
+      "learning_rate": 0.00014670226934124738,
+      "loss": 2.5977,
+      "step": 6972
+    },
+    {
+      "epoch": 0.5627471551932854,
+      "grad_norm": 0.6705875396728516,
+      "learning_rate": 0.00014668830926871555,
+      "loss": 2.649,
+      "step": 6973
+    },
+    {
+      "epoch": 0.5628278589298684,
+      "grad_norm": 0.682731568813324,
+      "learning_rate": 0.00014667434803258906,
+      "loss": 2.6084,
+      "step": 6974
+    },
+    {
+      "epoch": 0.5629085626664515,
+      "grad_norm": 0.7061700224876404,
+      "learning_rate": 0.00014666038563321577,
+      "loss": 2.6256,
+      "step": 6975
+    },
+    {
+      "epoch": 0.5629892664030345,
+      "grad_norm": 0.6839977502822876,
+      "learning_rate": 0.00014664642207094374,
+      "loss": 2.6342,
+      "step": 6976
+    },
+    {
+      "epoch": 0.5630699701396175,
+      "grad_norm": 0.7376503348350525,
+      "learning_rate": 0.00014663245734612094,
+      "loss": 2.6001,
+      "step": 6977
+    },
+    {
+      "epoch": 0.5631506738762004,
+      "grad_norm": 0.6901546716690063,
+      "learning_rate": 0.0001466184914590954,
+      "loss": 2.6715,
+      "step": 6978
+    },
+    {
+      "epoch": 0.5632313776127835,
+      "grad_norm": 0.816223680973053,
+      "learning_rate": 0.00014660452441021512,
+      "loss": 2.6407,
+      "step": 6979
+    },
+    {
+      "epoch": 0.5633120813493665,
+      "grad_norm": 0.6904644966125488,
+      "learning_rate": 0.00014659055619982835,
+      "loss": 2.5543,
+      "step": 6980
+    },
+    {
+      "epoch": 0.5633927850859495,
+      "grad_norm": 0.6784235239028931,
+      "learning_rate": 0.0001465765868282831,
+      "loss": 2.6184,
+      "step": 6981
+    },
+    {
+      "epoch": 0.5634734888225325,
+      "grad_norm": 0.7689006328582764,
+      "learning_rate": 0.00014656261629592755,
+      "loss": 2.644,
+      "step": 6982
+    },
+    {
+      "epoch": 0.5635541925591155,
+      "grad_norm": 0.7608775496482849,
+      "learning_rate": 0.0001465486446031099,
+      "loss": 2.5952,
+      "step": 6983
+    },
+    {
+      "epoch": 0.5636348962956985,
+      "grad_norm": 0.7266525626182556,
+      "learning_rate": 0.00014653467175017833,
+      "loss": 2.6479,
+      "step": 6984
+    },
+    {
+      "epoch": 0.5637156000322815,
+      "grad_norm": 0.6907477974891663,
+      "learning_rate": 0.00014652069773748113,
+      "loss": 2.5825,
+      "step": 6985
+    },
+    {
+      "epoch": 0.5637963037688645,
+      "grad_norm": 0.7790403366088867,
+      "learning_rate": 0.00014650672256536648,
+      "loss": 2.5948,
+      "step": 6986
+    },
+    {
+      "epoch": 0.5638770075054474,
+      "grad_norm": 0.7072858214378357,
+      "learning_rate": 0.00014649274623418278,
+      "loss": 2.6017,
+      "step": 6987
+    },
+    {
+      "epoch": 0.5639577112420305,
+      "grad_norm": 0.7140414118766785,
+      "learning_rate": 0.0001464787687442783,
+      "loss": 2.5709,
+      "step": 6988
+    },
+    {
+      "epoch": 0.5640384149786135,
+      "grad_norm": 0.857783317565918,
+      "learning_rate": 0.00014646479009600139,
+      "loss": 2.7049,
+      "step": 6989
+    },
+    {
+      "epoch": 0.5641191187151965,
+      "grad_norm": 0.7599344253540039,
+      "learning_rate": 0.00014645081028970047,
+      "loss": 2.6369,
+      "step": 6990
+    },
+    {
+      "epoch": 0.5641998224517795,
+      "grad_norm": 0.7286150455474854,
+      "learning_rate": 0.00014643682932572393,
+      "loss": 2.6238,
+      "step": 6991
+    },
+    {
+      "epoch": 0.5642805261883626,
+      "grad_norm": 0.7095075249671936,
+      "learning_rate": 0.0001464228472044202,
+      "loss": 2.5924,
+      "step": 6992
+    },
+    {
+      "epoch": 0.5643612299249455,
+      "grad_norm": 0.7583668828010559,
+      "learning_rate": 0.0001464088639261378,
+      "loss": 2.6098,
+      "step": 6993
+    },
+    {
+      "epoch": 0.5644419336615285,
+      "grad_norm": 0.7393970489501953,
+      "learning_rate": 0.00014639487949122515,
+      "loss": 2.6036,
+      "step": 6994
+    },
+    {
+      "epoch": 0.5645226373981115,
+      "grad_norm": 0.6789388656616211,
+      "learning_rate": 0.00014638089390003086,
+      "loss": 2.642,
+      "step": 6995
+    },
+    {
+      "epoch": 0.5646033411346946,
+      "grad_norm": 0.8021289706230164,
+      "learning_rate": 0.00014636690715290346,
+      "loss": 2.6851,
+      "step": 6996
+    },
+    {
+      "epoch": 0.5646840448712775,
+      "grad_norm": 0.6931039094924927,
+      "learning_rate": 0.00014635291925019152,
+      "loss": 2.6358,
+      "step": 6997
+    },
+    {
+      "epoch": 0.5647647486078605,
+      "grad_norm": 0.7356590032577515,
+      "learning_rate": 0.00014633893019224366,
+      "loss": 2.5661,
+      "step": 6998
+    },
+    {
+      "epoch": 0.5648454523444435,
+      "grad_norm": 0.6777941584587097,
+      "learning_rate": 0.0001463249399794085,
+      "loss": 2.5578,
+      "step": 6999
+    },
+    {
+      "epoch": 0.5649261560810266,
+      "grad_norm": 0.7163615822792053,
+      "learning_rate": 0.0001463109486120348,
+      "loss": 2.5582,
+      "step": 7000
+    },
+    {
+      "epoch": 0.5649261560810266,
+      "eval_loss": 2.5298855304718018,
+      "eval_runtime": 757.774,
+      "eval_samples_per_second": 3.457,
+      "eval_steps_per_second": 0.577,
+      "step": 7000
+    },
+    {
+      "epoch": 0.5650068598176096,
+      "grad_norm": 0.7175148129463196,
+      "learning_rate": 0.0001462969560904712,
+      "loss": 2.568,
+      "step": 7001
+    },
+    {
+      "epoch": 0.5650875635541925,
+      "grad_norm": 0.6998937129974365,
+      "learning_rate": 0.00014628296241506636,
+      "loss": 2.6347,
+      "step": 7002
+    },
+    {
+      "epoch": 0.5651682672907755,
+      "grad_norm": 0.8140312433242798,
+      "learning_rate": 0.00014626896758616916,
+      "loss": 2.6566,
+      "step": 7003
+    },
+    {
+      "epoch": 0.5652489710273586,
+      "grad_norm": 0.7218164205551147,
+      "learning_rate": 0.00014625497160412833,
+      "loss": 2.5693,
+      "step": 7004
+    },
+    {
+      "epoch": 0.5653296747639416,
+      "grad_norm": 0.6974074244499207,
+      "learning_rate": 0.0001462409744692927,
+      "loss": 2.6084,
+      "step": 7005
+    },
+    {
+      "epoch": 0.5654103785005246,
+      "grad_norm": 0.7475053071975708,
+      "learning_rate": 0.00014622697618201113,
+      "loss": 2.6534,
+      "step": 7006
+    },
+    {
+      "epoch": 0.5654910822371075,
+      "grad_norm": 0.6768492460250854,
+      "learning_rate": 0.00014621297674263247,
+      "loss": 2.585,
+      "step": 7007
+    },
+    {
+      "epoch": 0.5655717859736906,
+      "grad_norm": 0.7023029923439026,
+      "learning_rate": 0.0001461989761515056,
+      "loss": 2.6219,
+      "step": 7008
+    },
+    {
+      "epoch": 0.5656524897102736,
+      "grad_norm": 0.7248445749282837,
+      "learning_rate": 0.0001461849744089795,
+      "loss": 2.6382,
+      "step": 7009
+    },
+    {
+      "epoch": 0.5657331934468566,
+      "grad_norm": 0.6961148381233215,
+      "learning_rate": 0.00014617097151540308,
+      "loss": 2.7184,
+      "step": 7010
+    },
+    {
+      "epoch": 0.5658138971834396,
+      "grad_norm": 0.6649057269096375,
+      "learning_rate": 0.0001461569674711254,
+      "loss": 2.6059,
+      "step": 7011
+    },
+    {
+      "epoch": 0.5658946009200226,
+      "grad_norm": 0.7451788783073425,
+      "learning_rate": 0.00014614296227649542,
+      "loss": 2.5697,
+      "step": 7012
+    },
+    {
+      "epoch": 0.5659753046566056,
+      "grad_norm": 0.6880216598510742,
+      "learning_rate": 0.0001461289559318622,
+      "loss": 2.5785,
+      "step": 7013
+    },
+    {
+      "epoch": 0.5660560083931886,
+      "grad_norm": 0.7505971789360046,
+      "learning_rate": 0.00014611494843757482,
+      "loss": 2.5479,
+      "step": 7014
+    },
+    {
+      "epoch": 0.5661367121297716,
+      "grad_norm": 0.745914876461029,
+      "learning_rate": 0.00014610093979398235,
+      "loss": 2.6367,
+      "step": 7015
+    },
+    {
+      "epoch": 0.5662174158663547,
+      "grad_norm": 0.6758660674095154,
+      "learning_rate": 0.000146086930001434,
+      "loss": 2.5673,
+      "step": 7016
+    },
+    {
+      "epoch": 0.5662981196029376,
+      "grad_norm": 0.7114273309707642,
+      "learning_rate": 0.00014607291906027886,
+      "loss": 2.6188,
+      "step": 7017
+    },
+    {
+      "epoch": 0.5663788233395206,
+      "grad_norm": 0.6791165471076965,
+      "learning_rate": 0.00014605890697086613,
+      "loss": 2.6197,
+      "step": 7018
+    },
+    {
+      "epoch": 0.5664595270761036,
+      "grad_norm": 0.6948217153549194,
+      "learning_rate": 0.00014604489373354503,
+      "loss": 2.5996,
+      "step": 7019
+    },
+    {
+      "epoch": 0.5665402308126867,
+      "grad_norm": 0.6993576884269714,
+      "learning_rate": 0.00014603087934866483,
+      "loss": 2.565,
+      "step": 7020
+    },
+    {
+      "epoch": 0.5666209345492697,
+      "grad_norm": 0.6936905384063721,
+      "learning_rate": 0.0001460168638165748,
+      "loss": 2.6524,
+      "step": 7021
+    },
+    {
+      "epoch": 0.5667016382858526,
+      "grad_norm": 0.6810741424560547,
+      "learning_rate": 0.00014600284713762424,
+      "loss": 2.6519,
+      "step": 7022
+    },
+    {
+      "epoch": 0.5667823420224356,
+      "grad_norm": 0.7540227770805359,
+      "learning_rate": 0.00014598882931216245,
+      "loss": 2.659,
+      "step": 7023
+    },
+    {
+      "epoch": 0.5668630457590187,
+      "grad_norm": 0.6520613431930542,
+      "learning_rate": 0.0001459748103405388,
+      "loss": 2.5341,
+      "step": 7024
+    },
+    {
+      "epoch": 0.5669437494956017,
+      "grad_norm": 0.7159109711647034,
+      "learning_rate": 0.00014596079022310277,
+      "loss": 2.6548,
+      "step": 7025
+    },
+    {
+      "epoch": 0.5670244532321846,
+      "grad_norm": 0.803284227848053,
+      "learning_rate": 0.00014594676896020366,
+      "loss": 2.705,
+      "step": 7026
+    },
+    {
+      "epoch": 0.5671051569687676,
+      "grad_norm": 0.7069976925849915,
+      "learning_rate": 0.00014593274655219095,
+      "loss": 2.5733,
+      "step": 7027
+    },
+    {
+      "epoch": 0.5671858607053507,
+      "grad_norm": 0.7085167169570923,
+      "learning_rate": 0.00014591872299941417,
+      "loss": 2.6247,
+      "step": 7028
+    },
+    {
+      "epoch": 0.5672665644419337,
+      "grad_norm": 0.6748499274253845,
+      "learning_rate": 0.00014590469830222272,
+      "loss": 2.6446,
+      "step": 7029
+    },
+    {
+      "epoch": 0.5673472681785167,
+      "grad_norm": 0.6885821223258972,
+      "learning_rate": 0.00014589067246096623,
+      "loss": 2.5879,
+      "step": 7030
+    },
+    {
+      "epoch": 0.5674279719150996,
+      "grad_norm": 0.7220324277877808,
+      "learning_rate": 0.0001458766454759942,
+      "loss": 2.6249,
+      "step": 7031
+    },
+    {
+      "epoch": 0.5675086756516827,
+      "grad_norm": 0.6712783575057983,
+      "learning_rate": 0.00014586261734765628,
+      "loss": 2.5971,
+      "step": 7032
+    },
+    {
+      "epoch": 0.5675893793882657,
+      "grad_norm": 0.6582161784172058,
+      "learning_rate": 0.00014584858807630203,
+      "loss": 2.6224,
+      "step": 7033
+    },
+    {
+      "epoch": 0.5676700831248487,
+      "grad_norm": 0.6699219346046448,
+      "learning_rate": 0.0001458345576622811,
+      "loss": 2.5926,
+      "step": 7034
+    },
+    {
+      "epoch": 0.5677507868614317,
+      "grad_norm": 0.6508033871650696,
+      "learning_rate": 0.0001458205261059432,
+      "loss": 2.6311,
+      "step": 7035
+    },
+    {
+      "epoch": 0.5678314905980147,
+      "grad_norm": 0.7551338076591492,
+      "learning_rate": 0.00014580649340763802,
+      "loss": 2.5729,
+      "step": 7036
+    },
+    {
+      "epoch": 0.5679121943345977,
+      "grad_norm": 0.6875829100608826,
+      "learning_rate": 0.00014579245956771527,
+      "loss": 2.6253,
+      "step": 7037
+    },
+    {
+      "epoch": 0.5679928980711807,
+      "grad_norm": 0.698204517364502,
+      "learning_rate": 0.00014577842458652474,
+      "loss": 2.6218,
+      "step": 7038
+    },
+    {
+      "epoch": 0.5680736018077637,
+      "grad_norm": 0.8258630037307739,
+      "learning_rate": 0.00014576438846441615,
+      "loss": 2.6307,
+      "step": 7039
+    },
+    {
+      "epoch": 0.5681543055443466,
+      "grad_norm": 0.753105878829956,
+      "learning_rate": 0.00014575035120173942,
+      "loss": 2.5664,
+      "step": 7040
+    },
+    {
+      "epoch": 0.5682350092809297,
+      "grad_norm": 0.6999726295471191,
+      "learning_rate": 0.00014573631279884435,
+      "loss": 2.6857,
+      "step": 7041
+    },
+    {
+      "epoch": 0.5683157130175127,
+      "grad_norm": 0.6484847068786621,
+      "learning_rate": 0.00014572227325608078,
+      "loss": 2.6068,
+      "step": 7042
+    },
+    {
+      "epoch": 0.5683964167540957,
+      "grad_norm": 0.7098011374473572,
+      "learning_rate": 0.00014570823257379866,
+      "loss": 2.6591,
+      "step": 7043
+    },
+    {
+      "epoch": 0.5684771204906787,
+      "grad_norm": 0.8304192423820496,
+      "learning_rate": 0.0001456941907523479,
+      "loss": 2.6582,
+      "step": 7044
+    },
+    {
+      "epoch": 0.5685578242272618,
+      "grad_norm": 0.763214111328125,
+      "learning_rate": 0.00014568014779207844,
+      "loss": 2.6605,
+      "step": 7045
+    },
+    {
+      "epoch": 0.5686385279638447,
+      "grad_norm": 0.6805880665779114,
+      "learning_rate": 0.00014566610369334032,
+      "loss": 2.6362,
+      "step": 7046
+    },
+    {
+      "epoch": 0.5687192317004277,
+      "grad_norm": 0.6753434538841248,
+      "learning_rate": 0.00014565205845648352,
+      "loss": 2.6352,
+      "step": 7047
+    },
+    {
+      "epoch": 0.5687999354370107,
+      "grad_norm": 0.7065438032150269,
+      "learning_rate": 0.00014563801208185807,
+      "loss": 2.5975,
+      "step": 7048
+    },
+    {
+      "epoch": 0.5688806391735938,
+      "grad_norm": 0.6863527894020081,
+      "learning_rate": 0.00014562396456981407,
+      "loss": 2.576,
+      "step": 7049
+    },
+    {
+      "epoch": 0.5689613429101767,
+      "grad_norm": 0.7344440817832947,
+      "learning_rate": 0.00014560991592070158,
+      "loss": 2.5933,
+      "step": 7050
+    },
+    {
+      "epoch": 0.5690420466467597,
+      "grad_norm": 0.699992835521698,
+      "learning_rate": 0.00014559586613487082,
+      "loss": 2.6161,
+      "step": 7051
+    },
+    {
+      "epoch": 0.5691227503833427,
+      "grad_norm": 0.7287258505821228,
+      "learning_rate": 0.00014558181521267185,
+      "loss": 2.665,
+      "step": 7052
+    },
+    {
+      "epoch": 0.5692034541199258,
+      "grad_norm": 0.7304692268371582,
+      "learning_rate": 0.0001455677631544549,
+      "loss": 2.5696,
+      "step": 7053
+    },
+    {
+      "epoch": 0.5692841578565088,
+      "grad_norm": 0.6556086540222168,
+      "learning_rate": 0.00014555370996057016,
+      "loss": 2.6405,
+      "step": 7054
+    },
+    {
+      "epoch": 0.5693648615930917,
+      "grad_norm": 0.6796221137046814,
+      "learning_rate": 0.0001455396556313679,
+      "loss": 2.6475,
+      "step": 7055
+    },
+    {
+      "epoch": 0.5694455653296747,
+      "grad_norm": 0.7067505717277527,
+      "learning_rate": 0.00014552560016719838,
+      "loss": 2.6344,
+      "step": 7056
+    },
+    {
+      "epoch": 0.5695262690662578,
+      "grad_norm": 0.7108997106552124,
+      "learning_rate": 0.00014551154356841193,
+      "loss": 2.6543,
+      "step": 7057
+    },
+    {
+      "epoch": 0.5696069728028408,
+      "grad_norm": 0.7296212911605835,
+      "learning_rate": 0.0001454974858353588,
+      "loss": 2.6152,
+      "step": 7058
+    },
+    {
+      "epoch": 0.5696876765394238,
+      "grad_norm": 0.7329154014587402,
+      "learning_rate": 0.00014548342696838943,
+      "loss": 2.6338,
+      "step": 7059
+    },
+    {
+      "epoch": 0.5697683802760067,
+      "grad_norm": 0.6880258321762085,
+      "learning_rate": 0.00014546936696785412,
+      "loss": 2.5834,
+      "step": 7060
+    },
+    {
+      "epoch": 0.5698490840125898,
+      "grad_norm": 0.7140741348266602,
+      "learning_rate": 0.00014545530583410336,
+      "loss": 2.6361,
+      "step": 7061
+    },
+    {
+      "epoch": 0.5699297877491728,
+      "grad_norm": 0.6419476866722107,
+      "learning_rate": 0.00014544124356748755,
+      "loss": 2.4982,
+      "step": 7062
+    },
+    {
+      "epoch": 0.5700104914857558,
+      "grad_norm": 0.6934036612510681,
+      "learning_rate": 0.00014542718016835718,
+      "loss": 2.5748,
+      "step": 7063
+    },
+    {
+      "epoch": 0.5700911952223388,
+      "grad_norm": 0.721663236618042,
+      "learning_rate": 0.0001454131156370627,
+      "loss": 2.5419,
+      "step": 7064
+    },
+    {
+      "epoch": 0.5701718989589218,
+      "grad_norm": 0.734062671661377,
+      "learning_rate": 0.00014539904997395468,
+      "loss": 2.6288,
+      "step": 7065
+    },
+    {
+      "epoch": 0.5702526026955048,
+      "grad_norm": 0.7927694320678711,
+      "learning_rate": 0.00014538498317938367,
+      "loss": 2.6331,
+      "step": 7066
+    },
+    {
+      "epoch": 0.5703333064320878,
+      "grad_norm": 0.715929388999939,
+      "learning_rate": 0.00014537091525370025,
+      "loss": 2.6333,
+      "step": 7067
+    },
+    {
+      "epoch": 0.5704140101686708,
+      "grad_norm": 0.772230327129364,
+      "learning_rate": 0.00014535684619725498,
+      "loss": 2.6019,
+      "step": 7068
+    },
+    {
+      "epoch": 0.5704947139052539,
+      "grad_norm": 0.7277318239212036,
+      "learning_rate": 0.0001453427760103986,
+      "loss": 2.6062,
+      "step": 7069
+    },
+    {
+      "epoch": 0.5705754176418368,
+      "grad_norm": 0.6708227396011353,
+      "learning_rate": 0.00014532870469348164,
+      "loss": 2.6613,
+      "step": 7070
+    },
+    {
+      "epoch": 0.5706561213784198,
+      "grad_norm": 0.7507323622703552,
+      "learning_rate": 0.0001453146322468549,
+      "loss": 2.6456,
+      "step": 7071
+    },
+    {
+      "epoch": 0.5707368251150028,
+      "grad_norm": 0.6864063739776611,
+      "learning_rate": 0.00014530055867086912,
+      "loss": 2.6361,
+      "step": 7072
+    },
+    {
+      "epoch": 0.5708175288515859,
+      "grad_norm": 0.6805310249328613,
+      "learning_rate": 0.00014528648396587498,
+      "loss": 2.6088,
+      "step": 7073
+    },
+    {
+      "epoch": 0.5708982325881689,
+      "grad_norm": 0.7946523427963257,
+      "learning_rate": 0.00014527240813222325,
+      "loss": 2.6533,
+      "step": 7074
+    },
+    {
+      "epoch": 0.5709789363247518,
+      "grad_norm": 0.6814306974411011,
+      "learning_rate": 0.00014525833117026474,
+      "loss": 2.6478,
+      "step": 7075
+    },
+    {
+      "epoch": 0.5710596400613348,
+      "grad_norm": 0.749664843082428,
+      "learning_rate": 0.00014524425308035034,
+      "loss": 2.6296,
+      "step": 7076
+    },
+    {
+      "epoch": 0.5711403437979179,
+      "grad_norm": 0.6774656772613525,
+      "learning_rate": 0.00014523017386283091,
+      "loss": 2.5867,
+      "step": 7077
+    },
+    {
+      "epoch": 0.5712210475345009,
+      "grad_norm": 0.7331634163856506,
+      "learning_rate": 0.00014521609351805733,
+      "loss": 2.6484,
+      "step": 7078
+    },
+    {
+      "epoch": 0.5713017512710838,
+      "grad_norm": 0.7076910734176636,
+      "learning_rate": 0.00014520201204638045,
+      "loss": 2.6464,
+      "step": 7079
+    },
+    {
+      "epoch": 0.5713824550076668,
+      "grad_norm": 0.74099200963974,
+      "learning_rate": 0.00014518792944815127,
+      "loss": 2.6304,
+      "step": 7080
+    },
+    {
+      "epoch": 0.5714631587442499,
+      "grad_norm": 0.6673823595046997,
+      "learning_rate": 0.00014517384572372078,
+      "loss": 2.5903,
+      "step": 7081
+    },
+    {
+      "epoch": 0.5715438624808329,
+      "grad_norm": 0.6872609257698059,
+      "learning_rate": 0.00014515976087343997,
+      "loss": 2.6189,
+      "step": 7082
+    },
+    {
+      "epoch": 0.5716245662174159,
+      "grad_norm": 0.7363224625587463,
+      "learning_rate": 0.0001451456748976599,
+      "loss": 2.5845,
+      "step": 7083
+    },
+    {
+      "epoch": 0.5717052699539988,
+      "grad_norm": 0.7672157287597656,
+      "learning_rate": 0.00014513158779673157,
+      "loss": 2.6331,
+      "step": 7084
+    },
+    {
+      "epoch": 0.5717859736905819,
+      "grad_norm": 0.661195695400238,
+      "learning_rate": 0.00014511749957100612,
+      "loss": 2.5827,
+      "step": 7085
+    },
+    {
+      "epoch": 0.5718666774271649,
+      "grad_norm": 0.8034788370132446,
+      "learning_rate": 0.0001451034102208346,
+      "loss": 2.6209,
+      "step": 7086
+    },
+    {
+      "epoch": 0.5719473811637479,
+      "grad_norm": 0.7318302392959595,
+      "learning_rate": 0.00014508931974656822,
+      "loss": 2.5898,
+      "step": 7087
+    },
+    {
+      "epoch": 0.5720280849003309,
+      "grad_norm": 0.7334744930267334,
+      "learning_rate": 0.00014507522814855814,
+      "loss": 2.5893,
+      "step": 7088
+    },
+    {
+      "epoch": 0.5721087886369138,
+      "grad_norm": 0.783051609992981,
+      "learning_rate": 0.00014506113542715553,
+      "loss": 2.6284,
+      "step": 7089
+    },
+    {
+      "epoch": 0.5721894923734969,
+      "grad_norm": 0.7319497466087341,
+      "learning_rate": 0.00014504704158271165,
+      "loss": 2.5705,
+      "step": 7090
+    },
+    {
+      "epoch": 0.5722701961100799,
+      "grad_norm": 0.7886925935745239,
+      "learning_rate": 0.00014503294661557772,
+      "loss": 2.641,
+      "step": 7091
+    },
+    {
+      "epoch": 0.5723508998466629,
+      "grad_norm": 0.6882795691490173,
+      "learning_rate": 0.00014501885052610502,
+      "loss": 2.5714,
+      "step": 7092
+    },
+    {
+      "epoch": 0.5724316035832459,
+      "grad_norm": 0.7089235186576843,
+      "learning_rate": 0.00014500475331464494,
+      "loss": 2.6073,
+      "step": 7093
+    },
+    {
+      "epoch": 0.5725123073198289,
+      "grad_norm": 0.7261029481887817,
+      "learning_rate": 0.00014499065498154874,
+      "loss": 2.5595,
+      "step": 7094
+    },
+    {
+      "epoch": 0.5725930110564119,
+      "grad_norm": 0.7625105977058411,
+      "learning_rate": 0.0001449765555271678,
+      "loss": 2.5978,
+      "step": 7095
+    },
+    {
+      "epoch": 0.5726737147929949,
+      "grad_norm": 0.7853986024856567,
+      "learning_rate": 0.00014496245495185353,
+      "loss": 2.6378,
+      "step": 7096
+    },
+    {
+      "epoch": 0.5727544185295779,
+      "grad_norm": 0.8070923686027527,
+      "learning_rate": 0.00014494835325595736,
+      "loss": 2.7062,
+      "step": 7097
+    },
+    {
+      "epoch": 0.572835122266161,
+      "grad_norm": 0.7074965834617615,
+      "learning_rate": 0.00014493425043983073,
+      "loss": 2.5177,
+      "step": 7098
+    },
+    {
+      "epoch": 0.5729158260027439,
+      "grad_norm": 0.6890520453453064,
+      "learning_rate": 0.00014492014650382512,
+      "loss": 2.6058,
+      "step": 7099
+    },
+    {
+      "epoch": 0.5729965297393269,
+      "grad_norm": 0.6979860067367554,
+      "learning_rate": 0.00014490604144829202,
+      "loss": 2.5274,
+      "step": 7100
+    },
+    {
+      "epoch": 0.5730772334759099,
+      "grad_norm": 0.7972229719161987,
+      "learning_rate": 0.000144891935273583,
+      "loss": 2.6369,
+      "step": 7101
+    },
+    {
+      "epoch": 0.573157937212493,
+      "grad_norm": 0.6994345188140869,
+      "learning_rate": 0.0001448778279800496,
+      "loss": 2.5975,
+      "step": 7102
+    },
+    {
+      "epoch": 0.573238640949076,
+      "grad_norm": 0.7943929433822632,
+      "learning_rate": 0.0001448637195680434,
+      "loss": 2.6317,
+      "step": 7103
+    },
+    {
+      "epoch": 0.5733193446856589,
+      "grad_norm": 0.6975306272506714,
+      "learning_rate": 0.00014484961003791605,
+      "loss": 2.6264,
+      "step": 7104
+    },
+    {
+      "epoch": 0.5734000484222419,
+      "grad_norm": 0.6889060735702515,
+      "learning_rate": 0.00014483549939001917,
+      "loss": 2.5974,
+      "step": 7105
+    },
+    {
+      "epoch": 0.573480752158825,
+      "grad_norm": 0.7372777462005615,
+      "learning_rate": 0.00014482138762470444,
+      "loss": 2.5851,
+      "step": 7106
+    },
+    {
+      "epoch": 0.573561455895408,
+      "grad_norm": 0.7045157551765442,
+      "learning_rate": 0.00014480727474232362,
+      "loss": 2.6451,
+      "step": 7107
+    },
+    {
+      "epoch": 0.5736421596319909,
+      "grad_norm": 0.6974517107009888,
+      "learning_rate": 0.00014479316074322832,
+      "loss": 2.6796,
+      "step": 7108
+    },
+    {
+      "epoch": 0.5737228633685739,
+      "grad_norm": 0.7328097224235535,
+      "learning_rate": 0.00014477904562777038,
+      "loss": 2.5923,
+      "step": 7109
+    },
+    {
+      "epoch": 0.573803567105157,
+      "grad_norm": 0.7288877964019775,
+      "learning_rate": 0.0001447649293963016,
+      "loss": 2.6012,
+      "step": 7110
+    },
+    {
+      "epoch": 0.57388427084174,
+      "grad_norm": 0.7054389119148254,
+      "learning_rate": 0.00014475081204917372,
+      "loss": 2.6666,
+      "step": 7111
+    },
+    {
+      "epoch": 0.573964974578323,
+      "grad_norm": 0.7447949647903442,
+      "learning_rate": 0.00014473669358673865,
+      "loss": 2.6093,
+      "step": 7112
+    },
+    {
+      "epoch": 0.5740456783149059,
+      "grad_norm": 0.6431592106819153,
+      "learning_rate": 0.0001447225740093482,
+      "loss": 2.6242,
+      "step": 7113
+    },
+    {
+      "epoch": 0.574126382051489,
+      "grad_norm": 0.7096747756004333,
+      "learning_rate": 0.00014470845331735434,
+      "loss": 2.6297,
+      "step": 7114
+    },
+    {
+      "epoch": 0.574207085788072,
+      "grad_norm": 0.6918880939483643,
+      "learning_rate": 0.00014469433151110894,
+      "loss": 2.5849,
+      "step": 7115
+    },
+    {
+      "epoch": 0.574287789524655,
+      "grad_norm": 0.6617783308029175,
+      "learning_rate": 0.00014468020859096395,
+      "loss": 2.5972,
+      "step": 7116
+    },
+    {
+      "epoch": 0.574368493261238,
+      "grad_norm": 0.6525121927261353,
+      "learning_rate": 0.0001446660845572714,
+      "loss": 2.5888,
+      "step": 7117
+    },
+    {
+      "epoch": 0.574449196997821,
+      "grad_norm": 0.7024720907211304,
+      "learning_rate": 0.00014465195941038326,
+      "loss": 2.6135,
+      "step": 7118
+    },
+    {
+      "epoch": 0.574529900734404,
+      "grad_norm": 0.7660520672798157,
+      "learning_rate": 0.00014463783315065153,
+      "loss": 2.5837,
+      "step": 7119
+    },
+    {
+      "epoch": 0.574610604470987,
+      "grad_norm": 0.8206443190574646,
+      "learning_rate": 0.00014462370577842838,
+      "loss": 2.6749,
+      "step": 7120
+    },
+    {
+      "epoch": 0.57469130820757,
+      "grad_norm": 0.7176216840744019,
+      "learning_rate": 0.00014460957729406577,
+      "loss": 2.5814,
+      "step": 7121
+    },
+    {
+      "epoch": 0.5747720119441531,
+      "grad_norm": 0.7867588400840759,
+      "learning_rate": 0.0001445954476979159,
+      "loss": 2.5697,
+      "step": 7122
+    },
+    {
+      "epoch": 0.574852715680736,
+      "grad_norm": 0.7150471806526184,
+      "learning_rate": 0.0001445813169903309,
+      "loss": 2.5689,
+      "step": 7123
+    },
+    {
+      "epoch": 0.574933419417319,
+      "grad_norm": 0.7082479596138,
+      "learning_rate": 0.00014456718517166296,
+      "loss": 2.6081,
+      "step": 7124
+    },
+    {
+      "epoch": 0.575014123153902,
+      "grad_norm": 0.7207253575325012,
+      "learning_rate": 0.00014455305224226426,
+      "loss": 2.6573,
+      "step": 7125
+    },
+    {
+      "epoch": 0.5750948268904851,
+      "grad_norm": 0.7451751232147217,
+      "learning_rate": 0.00014453891820248704,
+      "loss": 2.6057,
+      "step": 7126
+    },
+    {
+      "epoch": 0.575175530627068,
+      "grad_norm": 0.7030230164527893,
+      "learning_rate": 0.0001445247830526835,
+      "loss": 2.6122,
+      "step": 7127
+    },
+    {
+      "epoch": 0.575256234363651,
+      "grad_norm": 0.7233754396438599,
+      "learning_rate": 0.00014451064679320605,
+      "loss": 2.5937,
+      "step": 7128
+    },
+    {
+      "epoch": 0.575336938100234,
+      "grad_norm": 0.6943942904472351,
+      "learning_rate": 0.0001444965094244069,
+      "loss": 2.6327,
+      "step": 7129
+    },
+    {
+      "epoch": 0.5754176418368171,
+      "grad_norm": 0.682056725025177,
+      "learning_rate": 0.00014448237094663843,
+      "loss": 2.6212,
+      "step": 7130
+    },
+    {
+      "epoch": 0.5754983455734001,
+      "grad_norm": 0.7424136400222778,
+      "learning_rate": 0.00014446823136025298,
+      "loss": 2.6031,
+      "step": 7131
+    },
+    {
+      "epoch": 0.575579049309983,
+      "grad_norm": 0.7464002370834351,
+      "learning_rate": 0.00014445409066560298,
+      "loss": 2.6363,
+      "step": 7132
+    },
+    {
+      "epoch": 0.575659753046566,
+      "grad_norm": 0.7137650847434998,
+      "learning_rate": 0.00014443994886304085,
+      "loss": 2.5343,
+      "step": 7133
+    },
+    {
+      "epoch": 0.5757404567831491,
+      "grad_norm": 0.6744158864021301,
+      "learning_rate": 0.00014442580595291901,
+      "loss": 2.6463,
+      "step": 7134
+    },
+    {
+      "epoch": 0.5758211605197321,
+      "grad_norm": 0.6947084069252014,
+      "learning_rate": 0.00014441166193558991,
+      "loss": 2.6074,
+      "step": 7135
+    },
+    {
+      "epoch": 0.5759018642563151,
+      "grad_norm": 0.6981585621833801,
+      "learning_rate": 0.00014439751681140616,
+      "loss": 2.6257,
+      "step": 7136
+    },
+    {
+      "epoch": 0.575982567992898,
+      "grad_norm": 0.6800102591514587,
+      "learning_rate": 0.00014438337058072023,
+      "loss": 2.6447,
+      "step": 7137
+    },
+    {
+      "epoch": 0.5760632717294811,
+      "grad_norm": 0.6952316164970398,
+      "learning_rate": 0.00014436922324388465,
+      "loss": 2.5739,
+      "step": 7138
+    },
+    {
+      "epoch": 0.5761439754660641,
+      "grad_norm": 0.709170937538147,
+      "learning_rate": 0.0001443550748012521,
+      "loss": 2.5918,
+      "step": 7139
+    },
+    {
+      "epoch": 0.5762246792026471,
+      "grad_norm": 0.7677363157272339,
+      "learning_rate": 0.00014434092525317512,
+      "loss": 2.6322,
+      "step": 7140
+    },
+    {
+      "epoch": 0.5763053829392301,
+      "grad_norm": 0.6730263233184814,
+      "learning_rate": 0.00014432677460000636,
+      "loss": 2.6764,
+      "step": 7141
+    },
+    {
+      "epoch": 0.576386086675813,
+      "grad_norm": 0.6782239675521851,
+      "learning_rate": 0.0001443126228420985,
+      "loss": 2.5208,
+      "step": 7142
+    },
+    {
+      "epoch": 0.5764667904123961,
+      "grad_norm": 0.7737600207328796,
+      "learning_rate": 0.00014429846997980424,
+      "loss": 2.6964,
+      "step": 7143
+    },
+    {
+      "epoch": 0.5765474941489791,
+      "grad_norm": 0.7456403374671936,
+      "learning_rate": 0.00014428431601347635,
+      "loss": 2.6163,
+      "step": 7144
+    },
+    {
+      "epoch": 0.5766281978855621,
+      "grad_norm": 0.7824606895446777,
+      "learning_rate": 0.00014427016094346754,
+      "loss": 2.6499,
+      "step": 7145
+    },
+    {
+      "epoch": 0.576708901622145,
+      "grad_norm": 0.7233635187149048,
+      "learning_rate": 0.00014425600477013055,
+      "loss": 2.6064,
+      "step": 7146
+    },
+    {
+      "epoch": 0.5767896053587281,
+      "grad_norm": 0.7008275389671326,
+      "learning_rate": 0.00014424184749381824,
+      "loss": 2.5585,
+      "step": 7147
+    },
+    {
+      "epoch": 0.5768703090953111,
+      "grad_norm": 0.6817710995674133,
+      "learning_rate": 0.00014422768911488346,
+      "loss": 2.6215,
+      "step": 7148
+    },
+    {
+      "epoch": 0.5769510128318941,
+      "grad_norm": 0.6860779523849487,
+      "learning_rate": 0.00014421352963367906,
+      "loss": 2.5877,
+      "step": 7149
+    },
+    {
+      "epoch": 0.5770317165684771,
+      "grad_norm": 0.732865035533905,
+      "learning_rate": 0.00014419936905055793,
+      "loss": 2.5704,
+      "step": 7150
+    },
+    {
+      "epoch": 0.5771124203050602,
+      "grad_norm": 0.6992458701133728,
+      "learning_rate": 0.00014418520736587297,
+      "loss": 2.6654,
+      "step": 7151
+    },
+    {
+      "epoch": 0.5771931240416431,
+      "grad_norm": 0.6865053176879883,
+      "learning_rate": 0.00014417104457997715,
+      "loss": 2.6389,
+      "step": 7152
+    },
+    {
+      "epoch": 0.5772738277782261,
+      "grad_norm": 0.7652727365493774,
+      "learning_rate": 0.00014415688069322345,
+      "loss": 2.6478,
+      "step": 7153
+    },
+    {
+      "epoch": 0.5773545315148091,
+      "grad_norm": 0.708692193031311,
+      "learning_rate": 0.0001441427157059648,
+      "loss": 2.6065,
+      "step": 7154
+    },
+    {
+      "epoch": 0.5774352352513922,
+      "grad_norm": 0.7549232244491577,
+      "learning_rate": 0.00014412854961855435,
+      "loss": 2.6484,
+      "step": 7155
+    },
+    {
+      "epoch": 0.5775159389879752,
+      "grad_norm": 0.6410655975341797,
+      "learning_rate": 0.00014411438243134506,
+      "loss": 2.6061,
+      "step": 7156
+    },
+    {
+      "epoch": 0.5775966427245581,
+      "grad_norm": 0.7711724042892456,
+      "learning_rate": 0.00014410021414469005,
+      "loss": 2.628,
+      "step": 7157
+    },
+    {
+      "epoch": 0.5776773464611411,
+      "grad_norm": 0.6723695993423462,
+      "learning_rate": 0.0001440860447589424,
+      "loss": 2.6214,
+      "step": 7158
+    },
+    {
+      "epoch": 0.5777580501977242,
+      "grad_norm": 0.7359206676483154,
+      "learning_rate": 0.0001440718742744553,
+      "loss": 2.6157,
+      "step": 7159
+    },
+    {
+      "epoch": 0.5778387539343072,
+      "grad_norm": 0.7320525050163269,
+      "learning_rate": 0.0001440577026915819,
+      "loss": 2.6081,
+      "step": 7160
+    },
+    {
+      "epoch": 0.5779194576708901,
+      "grad_norm": 0.7728561162948608,
+      "learning_rate": 0.00014404353001067535,
+      "loss": 2.5989,
+      "step": 7161
+    },
+    {
+      "epoch": 0.5780001614074731,
+      "grad_norm": 0.7380329370498657,
+      "learning_rate": 0.0001440293562320889,
+      "loss": 2.6337,
+      "step": 7162
+    },
+    {
+      "epoch": 0.5780808651440562,
+      "grad_norm": 0.667789876461029,
+      "learning_rate": 0.00014401518135617581,
+      "loss": 2.6324,
+      "step": 7163
+    },
+    {
+      "epoch": 0.5781615688806392,
+      "grad_norm": 0.6907219886779785,
+      "learning_rate": 0.00014400100538328935,
+      "loss": 2.5897,
+      "step": 7164
+    },
+    {
+      "epoch": 0.5782422726172222,
+      "grad_norm": 0.9051530957221985,
+      "learning_rate": 0.00014398682831378283,
+      "loss": 2.6895,
+      "step": 7165
+    },
+    {
+      "epoch": 0.5783229763538051,
+      "grad_norm": 0.7189533114433289,
+      "learning_rate": 0.00014397265014800956,
+      "loss": 2.5948,
+      "step": 7166
+    },
+    {
+      "epoch": 0.5784036800903882,
+      "grad_norm": 0.7003059983253479,
+      "learning_rate": 0.00014395847088632285,
+      "loss": 2.5814,
+      "step": 7167
+    },
+    {
+      "epoch": 0.5784843838269712,
+      "grad_norm": 0.8083534240722656,
+      "learning_rate": 0.0001439442905290762,
+      "loss": 2.6131,
+      "step": 7168
+    },
+    {
+      "epoch": 0.5785650875635542,
+      "grad_norm": 0.7068585157394409,
+      "learning_rate": 0.0001439301090766229,
+      "loss": 2.6027,
+      "step": 7169
+    },
+    {
+      "epoch": 0.5786457913001372,
+      "grad_norm": 0.7010494470596313,
+      "learning_rate": 0.00014391592652931653,
+      "loss": 2.5296,
+      "step": 7170
+    },
+    {
+      "epoch": 0.5787264950367202,
+      "grad_norm": 0.7577467560768127,
+      "learning_rate": 0.00014390174288751045,
+      "loss": 2.6347,
+      "step": 7171
+    },
+    {
+      "epoch": 0.5788071987733032,
+      "grad_norm": 0.643799364566803,
+      "learning_rate": 0.00014388755815155813,
+      "loss": 2.6152,
+      "step": 7172
+    },
+    {
+      "epoch": 0.5788879025098862,
+      "grad_norm": 0.740352988243103,
+      "learning_rate": 0.00014387337232181315,
+      "loss": 2.6123,
+      "step": 7173
+    },
+    {
+      "epoch": 0.5789686062464692,
+      "grad_norm": 0.7309309840202332,
+      "learning_rate": 0.00014385918539862907,
+      "loss": 2.6072,
+      "step": 7174
+    },
+    {
+      "epoch": 0.5790493099830523,
+      "grad_norm": 0.7237016558647156,
+      "learning_rate": 0.00014384499738235941,
+      "loss": 2.6375,
+      "step": 7175
+    },
+    {
+      "epoch": 0.5791300137196352,
+      "grad_norm": 0.6600970029830933,
+      "learning_rate": 0.00014383080827335784,
+      "loss": 2.5285,
+      "step": 7176
+    },
+    {
+      "epoch": 0.5792107174562182,
+      "grad_norm": 0.6822233200073242,
+      "learning_rate": 0.00014381661807197794,
+      "loss": 2.5497,
+      "step": 7177
+    },
+    {
+      "epoch": 0.5792914211928012,
+      "grad_norm": 0.6990383863449097,
+      "learning_rate": 0.00014380242677857337,
+      "loss": 2.6283,
+      "step": 7178
+    },
+    {
+      "epoch": 0.5793721249293843,
+      "grad_norm": 0.64422208070755,
+      "learning_rate": 0.00014378823439349783,
+      "loss": 2.5762,
+      "step": 7179
+    },
+    {
+      "epoch": 0.5794528286659673,
+      "grad_norm": 0.63804692029953,
+      "learning_rate": 0.00014377404091710501,
+      "loss": 2.5523,
+      "step": 7180
+    },
+    {
+      "epoch": 0.5795335324025502,
+      "grad_norm": 0.6978863477706909,
+      "learning_rate": 0.0001437598463497487,
+      "loss": 2.5089,
+      "step": 7181
+    },
+    {
+      "epoch": 0.5796142361391332,
+      "grad_norm": 0.7091087698936462,
+      "learning_rate": 0.00014374565069178257,
+      "loss": 2.7005,
+      "step": 7182
+    },
+    {
+      "epoch": 0.5796949398757163,
+      "grad_norm": 0.683659553527832,
+      "learning_rate": 0.00014373145394356053,
+      "loss": 2.5988,
+      "step": 7183
+    },
+    {
+      "epoch": 0.5797756436122993,
+      "grad_norm": 0.7352960705757141,
+      "learning_rate": 0.00014371725610543633,
+      "loss": 2.5671,
+      "step": 7184
+    },
+    {
+      "epoch": 0.5798563473488823,
+      "grad_norm": 0.6951913237571716,
+      "learning_rate": 0.00014370305717776382,
+      "loss": 2.5917,
+      "step": 7185
+    },
+    {
+      "epoch": 0.5799370510854652,
+      "grad_norm": 0.6644465923309326,
+      "learning_rate": 0.0001436888571608969,
+      "loss": 2.5954,
+      "step": 7186
+    },
+    {
+      "epoch": 0.5800177548220483,
+      "grad_norm": 0.7406458258628845,
+      "learning_rate": 0.00014367465605518942,
+      "loss": 2.6369,
+      "step": 7187
+    },
+    {
+      "epoch": 0.5800984585586313,
+      "grad_norm": 0.6724697351455688,
+      "learning_rate": 0.00014366045386099535,
+      "loss": 2.6227,
+      "step": 7188
+    },
+    {
+      "epoch": 0.5801791622952143,
+      "grad_norm": 0.6804977059364319,
+      "learning_rate": 0.00014364625057866867,
+      "loss": 2.6445,
+      "step": 7189
+    },
+    {
+      "epoch": 0.5802598660317972,
+      "grad_norm": 0.7020019888877869,
+      "learning_rate": 0.00014363204620856335,
+      "loss": 2.6733,
+      "step": 7190
+    },
+    {
+      "epoch": 0.5803405697683802,
+      "grad_norm": 0.6458491086959839,
+      "learning_rate": 0.00014361784075103332,
+      "loss": 2.572,
+      "step": 7191
+    },
+    {
+      "epoch": 0.5804212735049633,
+      "grad_norm": 0.7078056335449219,
+      "learning_rate": 0.00014360363420643272,
+      "loss": 2.7032,
+      "step": 7192
+    },
+    {
+      "epoch": 0.5805019772415463,
+      "grad_norm": 0.6367471814155579,
+      "learning_rate": 0.00014358942657511557,
+      "loss": 2.5369,
+      "step": 7193
+    },
+    {
+      "epoch": 0.5805826809781293,
+      "grad_norm": 0.7311955094337463,
+      "learning_rate": 0.00014357521785743596,
+      "loss": 2.6513,
+      "step": 7194
+    },
+    {
+      "epoch": 0.5806633847147122,
+      "grad_norm": 0.6957442164421082,
+      "learning_rate": 0.00014356100805374805,
+      "loss": 2.6512,
+      "step": 7195
+    },
+    {
+      "epoch": 0.5807440884512953,
+      "grad_norm": 0.7026693224906921,
+      "learning_rate": 0.0001435467971644059,
+      "loss": 2.6049,
+      "step": 7196
+    },
+    {
+      "epoch": 0.5808247921878783,
+      "grad_norm": 0.7337697744369507,
+      "learning_rate": 0.00014353258518976376,
+      "loss": 2.5516,
+      "step": 7197
+    },
+    {
+      "epoch": 0.5809054959244613,
+      "grad_norm": 0.6891856789588928,
+      "learning_rate": 0.00014351837213017577,
+      "loss": 2.5894,
+      "step": 7198
+    },
+    {
+      "epoch": 0.5809861996610443,
+      "grad_norm": 0.6710659265518188,
+      "learning_rate": 0.0001435041579859962,
+      "loss": 2.596,
+      "step": 7199
+    },
+    {
+      "epoch": 0.5810669033976273,
+      "grad_norm": 0.7637245059013367,
+      "learning_rate": 0.00014348994275757931,
+      "loss": 2.6278,
+      "step": 7200
+    },
+    {
+      "epoch": 0.5811476071342103,
+      "grad_norm": 0.7558664679527283,
+      "learning_rate": 0.00014347572644527934,
+      "loss": 2.6917,
+      "step": 7201
+    },
+    {
+      "epoch": 0.5812283108707933,
+      "grad_norm": 0.7254986763000488,
+      "learning_rate": 0.00014346150904945065,
+      "loss": 2.6161,
+      "step": 7202
+    },
+    {
+      "epoch": 0.5813090146073763,
+      "grad_norm": 0.7177211046218872,
+      "learning_rate": 0.00014344729057044753,
+      "loss": 2.555,
+      "step": 7203
+    },
+    {
+      "epoch": 0.5813897183439594,
+      "grad_norm": 0.6408729553222656,
+      "learning_rate": 0.00014343307100862432,
+      "loss": 2.6071,
+      "step": 7204
+    },
+    {
+      "epoch": 0.5814704220805423,
+      "grad_norm": 0.7399997711181641,
+      "learning_rate": 0.0001434188503643355,
+      "loss": 2.6013,
+      "step": 7205
+    },
+    {
+      "epoch": 0.5815511258171253,
+      "grad_norm": 0.7796236276626587,
+      "learning_rate": 0.00014340462863793543,
+      "loss": 2.603,
+      "step": 7206
+    },
+    {
+      "epoch": 0.5816318295537083,
+      "grad_norm": 0.7420137524604797,
+      "learning_rate": 0.00014339040582977855,
+      "loss": 2.5858,
+      "step": 7207
+    },
+    {
+      "epoch": 0.5817125332902914,
+      "grad_norm": 0.738042414188385,
+      "learning_rate": 0.00014337618194021928,
+      "loss": 2.592,
+      "step": 7208
+    },
+    {
+      "epoch": 0.5817932370268744,
+      "grad_norm": 0.6910614371299744,
+      "learning_rate": 0.00014336195696961222,
+      "loss": 2.6448,
+      "step": 7209
+    },
+    {
+      "epoch": 0.5818739407634573,
+      "grad_norm": 0.7838915586471558,
+      "learning_rate": 0.00014334773091831185,
+      "loss": 2.6257,
+      "step": 7210
+    },
+    {
+      "epoch": 0.5819546445000403,
+      "grad_norm": 0.7362141013145447,
+      "learning_rate": 0.0001433335037866727,
+      "loss": 2.6505,
+      "step": 7211
+    },
+    {
+      "epoch": 0.5820353482366234,
+      "grad_norm": 0.6892269253730774,
+      "learning_rate": 0.00014331927557504934,
+      "loss": 2.6518,
+      "step": 7212
+    },
+    {
+      "epoch": 0.5821160519732064,
+      "grad_norm": 0.7444556951522827,
+      "learning_rate": 0.0001433050462837964,
+      "loss": 2.6785,
+      "step": 7213
+    },
+    {
+      "epoch": 0.5821967557097893,
+      "grad_norm": 0.6948450207710266,
+      "learning_rate": 0.00014329081591326853,
+      "loss": 2.5753,
+      "step": 7214
+    },
+    {
+      "epoch": 0.5822774594463723,
+      "grad_norm": 0.713741660118103,
+      "learning_rate": 0.00014327658446382032,
+      "loss": 2.6425,
+      "step": 7215
+    },
+    {
+      "epoch": 0.5823581631829554,
+      "grad_norm": 0.7352245450019836,
+      "learning_rate": 0.00014326235193580657,
+      "loss": 2.6859,
+      "step": 7216
+    },
+    {
+      "epoch": 0.5824388669195384,
+      "grad_norm": 0.7151867151260376,
+      "learning_rate": 0.00014324811832958187,
+      "loss": 2.6106,
+      "step": 7217
+    },
+    {
+      "epoch": 0.5825195706561214,
+      "grad_norm": 0.7003469467163086,
+      "learning_rate": 0.000143233883645501,
+      "loss": 2.618,
+      "step": 7218
+    },
+    {
+      "epoch": 0.5826002743927043,
+      "grad_norm": 0.7139034867286682,
+      "learning_rate": 0.00014321964788391878,
+      "loss": 2.5772,
+      "step": 7219
+    },
+    {
+      "epoch": 0.5826809781292874,
+      "grad_norm": 0.6368305683135986,
+      "learning_rate": 0.00014320541104518992,
+      "loss": 2.5259,
+      "step": 7220
+    },
+    {
+      "epoch": 0.5827616818658704,
+      "grad_norm": 0.6921548247337341,
+      "learning_rate": 0.0001431911731296693,
+      "loss": 2.6403,
+      "step": 7221
+    },
+    {
+      "epoch": 0.5828423856024534,
+      "grad_norm": 0.6995570659637451,
+      "learning_rate": 0.00014317693413771175,
+      "loss": 2.6172,
+      "step": 7222
+    },
+    {
+      "epoch": 0.5829230893390364,
+      "grad_norm": 0.7557246088981628,
+      "learning_rate": 0.0001431626940696721,
+      "loss": 2.6347,
+      "step": 7223
+    },
+    {
+      "epoch": 0.5830037930756194,
+      "grad_norm": 0.6912205219268799,
+      "learning_rate": 0.00014314845292590528,
+      "loss": 2.5958,
+      "step": 7224
+    },
+    {
+      "epoch": 0.5830844968122024,
+      "grad_norm": 0.6896184682846069,
+      "learning_rate": 0.00014313421070676625,
+      "loss": 2.569,
+      "step": 7225
+    },
+    {
+      "epoch": 0.5831652005487854,
+      "grad_norm": 0.6900814771652222,
+      "learning_rate": 0.00014311996741260994,
+      "loss": 2.5466,
+      "step": 7226
+    },
+    {
+      "epoch": 0.5832459042853684,
+      "grad_norm": 0.7319771647453308,
+      "learning_rate": 0.00014310572304379132,
+      "loss": 2.6181,
+      "step": 7227
+    },
+    {
+      "epoch": 0.5833266080219515,
+      "grad_norm": 0.728138267993927,
+      "learning_rate": 0.0001430914776006654,
+      "loss": 2.6644,
+      "step": 7228
+    },
+    {
+      "epoch": 0.5834073117585344,
+      "grad_norm": 0.7361802458763123,
+      "learning_rate": 0.0001430772310835872,
+      "loss": 2.6079,
+      "step": 7229
+    },
+    {
+      "epoch": 0.5834880154951174,
+      "grad_norm": 0.6893376708030701,
+      "learning_rate": 0.00014306298349291182,
+      "loss": 2.5615,
+      "step": 7230
+    },
+    {
+      "epoch": 0.5835687192317004,
+      "grad_norm": 0.6661401987075806,
+      "learning_rate": 0.00014304873482899431,
+      "loss": 2.6028,
+      "step": 7231
+    },
+    {
+      "epoch": 0.5836494229682835,
+      "grad_norm": 0.6571504473686218,
+      "learning_rate": 0.0001430344850921898,
+      "loss": 2.5553,
+      "step": 7232
+    },
+    {
+      "epoch": 0.5837301267048665,
+      "grad_norm": 0.6878423690795898,
+      "learning_rate": 0.00014302023428285342,
+      "loss": 2.5336,
+      "step": 7233
+    },
+    {
+      "epoch": 0.5838108304414494,
+      "grad_norm": 0.768117368221283,
+      "learning_rate": 0.00014300598240134035,
+      "loss": 2.6036,
+      "step": 7234
+    },
+    {
+      "epoch": 0.5838915341780324,
+      "grad_norm": 0.6876625418663025,
+      "learning_rate": 0.0001429917294480058,
+      "loss": 2.6314,
+      "step": 7235
+    },
+    {
+      "epoch": 0.5839722379146155,
+      "grad_norm": 0.7146790027618408,
+      "learning_rate": 0.00014297747542320495,
+      "loss": 2.6029,
+      "step": 7236
+    },
+    {
+      "epoch": 0.5840529416511985,
+      "grad_norm": 0.7032392024993896,
+      "learning_rate": 0.00014296322032729308,
+      "loss": 2.6163,
+      "step": 7237
+    },
+    {
+      "epoch": 0.5841336453877815,
+      "grad_norm": 0.7323551177978516,
+      "learning_rate": 0.00014294896416062544,
+      "loss": 2.6706,
+      "step": 7238
+    },
+    {
+      "epoch": 0.5842143491243644,
+      "grad_norm": 0.7647258639335632,
+      "learning_rate": 0.00014293470692355734,
+      "loss": 2.6744,
+      "step": 7239
+    },
+    {
+      "epoch": 0.5842950528609475,
+      "grad_norm": 0.6824506521224976,
+      "learning_rate": 0.00014292044861644414,
+      "loss": 2.579,
+      "step": 7240
+    },
+    {
+      "epoch": 0.5843757565975305,
+      "grad_norm": 0.7553619742393494,
+      "learning_rate": 0.00014290618923964115,
+      "loss": 2.6196,
+      "step": 7241
+    },
+    {
+      "epoch": 0.5844564603341135,
+      "grad_norm": 0.6872109770774841,
+      "learning_rate": 0.00014289192879350375,
+      "loss": 2.555,
+      "step": 7242
+    },
+    {
+      "epoch": 0.5845371640706964,
+      "grad_norm": 0.664658784866333,
+      "learning_rate": 0.00014287766727838735,
+      "loss": 2.5781,
+      "step": 7243
+    },
+    {
+      "epoch": 0.5846178678072794,
+      "grad_norm": 0.6709543466567993,
+      "learning_rate": 0.00014286340469464744,
+      "loss": 2.6022,
+      "step": 7244
+    },
+    {
+      "epoch": 0.5846985715438625,
+      "grad_norm": 0.7236210107803345,
+      "learning_rate": 0.00014284914104263941,
+      "loss": 2.5609,
+      "step": 7245
+    },
+    {
+      "epoch": 0.5847792752804455,
+      "grad_norm": 0.6751740574836731,
+      "learning_rate": 0.0001428348763227188,
+      "loss": 2.5792,
+      "step": 7246
+    },
+    {
+      "epoch": 0.5848599790170285,
+      "grad_norm": 0.6684607267379761,
+      "learning_rate": 0.0001428206105352411,
+      "loss": 2.5705,
+      "step": 7247
+    },
+    {
+      "epoch": 0.5849406827536114,
+      "grad_norm": 0.6876732707023621,
+      "learning_rate": 0.00014280634368056186,
+      "loss": 2.6576,
+      "step": 7248
+    },
+    {
+      "epoch": 0.5850213864901945,
+      "grad_norm": 0.758637547492981,
+      "learning_rate": 0.0001427920757590366,
+      "loss": 2.6215,
+      "step": 7249
+    },
+    {
+      "epoch": 0.5851020902267775,
+      "grad_norm": 0.6839025020599365,
+      "learning_rate": 0.00014277780677102097,
+      "loss": 2.5898,
+      "step": 7250
+    },
+    {
+      "epoch": 0.5851827939633605,
+      "grad_norm": 0.6912671327590942,
+      "learning_rate": 0.00014276353671687056,
+      "loss": 2.5879,
+      "step": 7251
+    },
+    {
+      "epoch": 0.5852634976999435,
+      "grad_norm": 0.6727048754692078,
+      "learning_rate": 0.00014274926559694107,
+      "loss": 2.5501,
+      "step": 7252
+    },
+    {
+      "epoch": 0.5853442014365265,
+      "grad_norm": 0.7031945586204529,
+      "learning_rate": 0.00014273499341158812,
+      "loss": 2.625,
+      "step": 7253
+    },
+    {
+      "epoch": 0.5854249051731095,
+      "grad_norm": 0.6886943578720093,
+      "learning_rate": 0.0001427207201611674,
+      "loss": 2.6141,
+      "step": 7254
+    },
+    {
+      "epoch": 0.5855056089096925,
+      "grad_norm": 0.7906915545463562,
+      "learning_rate": 0.00014270644584603466,
+      "loss": 2.7189,
+      "step": 7255
+    },
+    {
+      "epoch": 0.5855863126462755,
+      "grad_norm": 0.6873704195022583,
+      "learning_rate": 0.00014269217046654567,
+      "loss": 2.6031,
+      "step": 7256
+    },
+    {
+      "epoch": 0.5856670163828586,
+      "grad_norm": 0.6655381321907043,
+      "learning_rate": 0.00014267789402305618,
+      "loss": 2.5747,
+      "step": 7257
+    },
+    {
+      "epoch": 0.5857477201194415,
+      "grad_norm": 0.6655673384666443,
+      "learning_rate": 0.00014266361651592204,
+      "loss": 2.625,
+      "step": 7258
+    },
+    {
+      "epoch": 0.5858284238560245,
+      "grad_norm": 0.6752866506576538,
+      "learning_rate": 0.00014264933794549901,
+      "loss": 2.5914,
+      "step": 7259
+    },
+    {
+      "epoch": 0.5859091275926075,
+      "grad_norm": 0.6680975556373596,
+      "learning_rate": 0.00014263505831214302,
+      "loss": 2.5572,
+      "step": 7260
+    },
+    {
+      "epoch": 0.5859898313291906,
+      "grad_norm": 0.6873607039451599,
+      "learning_rate": 0.00014262077761620994,
+      "loss": 2.6696,
+      "step": 7261
+    },
+    {
+      "epoch": 0.5860705350657736,
+      "grad_norm": 0.6745384335517883,
+      "learning_rate": 0.00014260649585805566,
+      "loss": 2.5738,
+      "step": 7262
+    },
+    {
+      "epoch": 0.5861512388023565,
+      "grad_norm": 0.6524637937545776,
+      "learning_rate": 0.0001425922130380361,
+      "loss": 2.6209,
+      "step": 7263
+    },
+    {
+      "epoch": 0.5862319425389395,
+      "grad_norm": 0.6729850172996521,
+      "learning_rate": 0.00014257792915650728,
+      "loss": 2.652,
+      "step": 7264
+    },
+    {
+      "epoch": 0.5863126462755226,
+      "grad_norm": 0.6713503003120422,
+      "learning_rate": 0.00014256364421382514,
+      "loss": 2.5658,
+      "step": 7265
+    },
+    {
+      "epoch": 0.5863933500121056,
+      "grad_norm": 0.6835616827011108,
+      "learning_rate": 0.00014254935821034575,
+      "loss": 2.5535,
+      "step": 7266
+    },
+    {
+      "epoch": 0.5864740537486886,
+      "grad_norm": 0.7425376176834106,
+      "learning_rate": 0.00014253507114642515,
+      "loss": 2.6369,
+      "step": 7267
+    },
+    {
+      "epoch": 0.5865547574852715,
+      "grad_norm": 0.6788069605827332,
+      "learning_rate": 0.00014252078302241932,
+      "loss": 2.601,
+      "step": 7268
+    },
+    {
+      "epoch": 0.5866354612218546,
+      "grad_norm": 0.6828538179397583,
+      "learning_rate": 0.0001425064938386845,
+      "loss": 2.5861,
+      "step": 7269
+    },
+    {
+      "epoch": 0.5867161649584376,
+      "grad_norm": 0.6763372421264648,
+      "learning_rate": 0.0001424922035955767,
+      "loss": 2.6035,
+      "step": 7270
+    },
+    {
+      "epoch": 0.5867968686950206,
+      "grad_norm": 0.6517930626869202,
+      "learning_rate": 0.0001424779122934521,
+      "loss": 2.5564,
+      "step": 7271
+    },
+    {
+      "epoch": 0.5868775724316035,
+      "grad_norm": 0.6633113622665405,
+      "learning_rate": 0.00014246361993266692,
+      "loss": 2.6163,
+      "step": 7272
+    },
+    {
+      "epoch": 0.5869582761681866,
+      "grad_norm": 0.684822678565979,
+      "learning_rate": 0.00014244932651357733,
+      "loss": 2.6057,
+      "step": 7273
+    },
+    {
+      "epoch": 0.5870389799047696,
+      "grad_norm": 0.7679704427719116,
+      "learning_rate": 0.00014243503203653952,
+      "loss": 2.6522,
+      "step": 7274
+    },
+    {
+      "epoch": 0.5871196836413526,
+      "grad_norm": 0.6834188103675842,
+      "learning_rate": 0.00014242073650190984,
+      "loss": 2.652,
+      "step": 7275
+    },
+    {
+      "epoch": 0.5872003873779356,
+      "grad_norm": 0.6903846859931946,
+      "learning_rate": 0.00014240643991004449,
+      "loss": 2.5894,
+      "step": 7276
+    },
+    {
+      "epoch": 0.5872810911145186,
+      "grad_norm": 0.7060866951942444,
+      "learning_rate": 0.0001423921422612998,
+      "loss": 2.5994,
+      "step": 7277
+    },
+    {
+      "epoch": 0.5873617948511016,
+      "grad_norm": 0.6646741628646851,
+      "learning_rate": 0.0001423778435560321,
+      "loss": 2.6432,
+      "step": 7278
+    },
+    {
+      "epoch": 0.5874424985876846,
+      "grad_norm": 0.6930218935012817,
+      "learning_rate": 0.0001423635437945978,
+      "loss": 2.6233,
+      "step": 7279
+    },
+    {
+      "epoch": 0.5875232023242676,
+      "grad_norm": 0.6914143562316895,
+      "learning_rate": 0.00014234924297735322,
+      "loss": 2.6143,
+      "step": 7280
+    },
+    {
+      "epoch": 0.5876039060608507,
+      "grad_norm": 0.7351366281509399,
+      "learning_rate": 0.0001423349411046548,
+      "loss": 2.6323,
+      "step": 7281
+    },
+    {
+      "epoch": 0.5876846097974336,
+      "grad_norm": 0.6813770532608032,
+      "learning_rate": 0.000142320638176859,
+      "loss": 2.5964,
+      "step": 7282
+    },
+    {
+      "epoch": 0.5877653135340166,
+      "grad_norm": 0.7049702405929565,
+      "learning_rate": 0.00014230633419432226,
+      "loss": 2.6284,
+      "step": 7283
+    },
+    {
+      "epoch": 0.5878460172705996,
+      "grad_norm": 0.7140446901321411,
+      "learning_rate": 0.00014229202915740107,
+      "loss": 2.6113,
+      "step": 7284
+    },
+    {
+      "epoch": 0.5879267210071827,
+      "grad_norm": 0.696588933467865,
+      "learning_rate": 0.00014227772306645196,
+      "loss": 2.6384,
+      "step": 7285
+    },
+    {
+      "epoch": 0.5880074247437657,
+      "grad_norm": 0.6800615787506104,
+      "learning_rate": 0.0001422634159218315,
+      "loss": 2.5743,
+      "step": 7286
+    },
+    {
+      "epoch": 0.5880881284803486,
+      "grad_norm": 0.7586596608161926,
+      "learning_rate": 0.00014224910772389624,
+      "loss": 2.6504,
+      "step": 7287
+    },
+    {
+      "epoch": 0.5881688322169316,
+      "grad_norm": 0.73286372423172,
+      "learning_rate": 0.00014223479847300278,
+      "loss": 2.6026,
+      "step": 7288
+    },
+    {
+      "epoch": 0.5882495359535147,
+      "grad_norm": 0.6808766722679138,
+      "learning_rate": 0.00014222048816950772,
+      "loss": 2.5822,
+      "step": 7289
+    },
+    {
+      "epoch": 0.5883302396900977,
+      "grad_norm": 0.7424919009208679,
+      "learning_rate": 0.0001422061768137677,
+      "loss": 2.6474,
+      "step": 7290
+    },
+    {
+      "epoch": 0.5884109434266807,
+      "grad_norm": 0.658183753490448,
+      "learning_rate": 0.00014219186440613948,
+      "loss": 2.6051,
+      "step": 7291
+    },
+    {
+      "epoch": 0.5884916471632636,
+      "grad_norm": 0.6693006157875061,
+      "learning_rate": 0.0001421775509469797,
+      "loss": 2.5774,
+      "step": 7292
+    },
+    {
+      "epoch": 0.5885723508998466,
+      "grad_norm": 0.7298646569252014,
+      "learning_rate": 0.00014216323643664508,
+      "loss": 2.5688,
+      "step": 7293
+    },
+    {
+      "epoch": 0.5886530546364297,
+      "grad_norm": 0.6665881276130676,
+      "learning_rate": 0.00014214892087549238,
+      "loss": 2.608,
+      "step": 7294
+    },
+    {
+      "epoch": 0.5887337583730127,
+      "grad_norm": 0.7220060229301453,
+      "learning_rate": 0.00014213460426387841,
+      "loss": 2.6078,
+      "step": 7295
+    },
+    {
+      "epoch": 0.5888144621095956,
+      "grad_norm": 0.6693970561027527,
+      "learning_rate": 0.00014212028660215997,
+      "loss": 2.597,
+      "step": 7296
+    },
+    {
+      "epoch": 0.5888951658461786,
+      "grad_norm": 0.682331919670105,
+      "learning_rate": 0.00014210596789069387,
+      "loss": 2.5752,
+      "step": 7297
+    },
+    {
+      "epoch": 0.5889758695827617,
+      "grad_norm": 0.7586890459060669,
+      "learning_rate": 0.000142091648129837,
+      "loss": 2.6878,
+      "step": 7298
+    },
+    {
+      "epoch": 0.5890565733193447,
+      "grad_norm": 0.6740901470184326,
+      "learning_rate": 0.00014207732731994624,
+      "loss": 2.6083,
+      "step": 7299
+    },
+    {
+      "epoch": 0.5891372770559277,
+      "grad_norm": 0.6959021091461182,
+      "learning_rate": 0.00014206300546137842,
+      "loss": 2.5765,
+      "step": 7300
+    },
+    {
+      "epoch": 0.5892179807925106,
+      "grad_norm": 0.7446078658103943,
+      "learning_rate": 0.0001420486825544906,
+      "loss": 2.662,
+      "step": 7301
+    },
+    {
+      "epoch": 0.5892986845290937,
+      "grad_norm": 0.7418847680091858,
+      "learning_rate": 0.0001420343585996397,
+      "loss": 2.6606,
+      "step": 7302
+    },
+    {
+      "epoch": 0.5893793882656767,
+      "grad_norm": 0.7185709476470947,
+      "learning_rate": 0.00014202003359718273,
+      "loss": 2.563,
+      "step": 7303
+    },
+    {
+      "epoch": 0.5894600920022597,
+      "grad_norm": 0.6960515379905701,
+      "learning_rate": 0.00014200570754747664,
+      "loss": 2.6182,
+      "step": 7304
+    },
+    {
+      "epoch": 0.5895407957388427,
+      "grad_norm": 0.6589705348014832,
+      "learning_rate": 0.00014199138045087849,
+      "loss": 2.6714,
+      "step": 7305
+    },
+    {
+      "epoch": 0.5896214994754257,
+      "grad_norm": 0.7027507424354553,
+      "learning_rate": 0.00014197705230774543,
+      "loss": 2.6145,
+      "step": 7306
+    },
+    {
+      "epoch": 0.5897022032120087,
+      "grad_norm": 0.6761246919631958,
+      "learning_rate": 0.00014196272311843447,
+      "loss": 2.5688,
+      "step": 7307
+    },
+    {
+      "epoch": 0.5897829069485917,
+      "grad_norm": 0.6618059277534485,
+      "learning_rate": 0.00014194839288330277,
+      "loss": 2.6194,
+      "step": 7308
+    },
+    {
+      "epoch": 0.5898636106851747,
+      "grad_norm": 0.7182614803314209,
+      "learning_rate": 0.00014193406160270747,
+      "loss": 2.5452,
+      "step": 7309
+    },
+    {
+      "epoch": 0.5899443144217578,
+      "grad_norm": 0.6830565333366394,
+      "learning_rate": 0.0001419197292770057,
+      "loss": 2.5728,
+      "step": 7310
+    },
+    {
+      "epoch": 0.5900250181583407,
+      "grad_norm": 0.6744499802589417,
+      "learning_rate": 0.00014190539590655475,
+      "loss": 2.5736,
+      "step": 7311
+    },
+    {
+      "epoch": 0.5901057218949237,
+      "grad_norm": 0.7177874445915222,
+      "learning_rate": 0.00014189106149171176,
+      "loss": 2.6271,
+      "step": 7312
+    },
+    {
+      "epoch": 0.5901864256315067,
+      "grad_norm": 0.6770105361938477,
+      "learning_rate": 0.000141876726032834,
+      "loss": 2.5924,
+      "step": 7313
+    },
+    {
+      "epoch": 0.5902671293680898,
+      "grad_norm": 0.7295818328857422,
+      "learning_rate": 0.0001418623895302788,
+      "loss": 2.644,
+      "step": 7314
+    },
+    {
+      "epoch": 0.5903478331046728,
+      "grad_norm": 0.7244859933853149,
+      "learning_rate": 0.00014184805198440338,
+      "loss": 2.5892,
+      "step": 7315
+    },
+    {
+      "epoch": 0.5904285368412557,
+      "grad_norm": 0.7067728638648987,
+      "learning_rate": 0.00014183371339556512,
+      "loss": 2.5985,
+      "step": 7316
+    },
+    {
+      "epoch": 0.5905092405778387,
+      "grad_norm": 0.6732490062713623,
+      "learning_rate": 0.0001418193737641214,
+      "loss": 2.5771,
+      "step": 7317
+    },
+    {
+      "epoch": 0.5905899443144218,
+      "grad_norm": 0.7087544202804565,
+      "learning_rate": 0.00014180503309042957,
+      "loss": 2.6373,
+      "step": 7318
+    },
+    {
+      "epoch": 0.5906706480510048,
+      "grad_norm": 0.772174596786499,
+      "learning_rate": 0.00014179069137484703,
+      "loss": 2.6262,
+      "step": 7319
+    },
+    {
+      "epoch": 0.5907513517875878,
+      "grad_norm": 0.6855718493461609,
+      "learning_rate": 0.00014177634861773118,
+      "loss": 2.6268,
+      "step": 7320
+    },
+    {
+      "epoch": 0.5908320555241707,
+      "grad_norm": 0.7168720364570618,
+      "learning_rate": 0.00014176200481943953,
+      "loss": 2.5892,
+      "step": 7321
+    },
+    {
+      "epoch": 0.5909127592607538,
+      "grad_norm": 0.7126333713531494,
+      "learning_rate": 0.0001417476599803296,
+      "loss": 2.6079,
+      "step": 7322
+    },
+    {
+      "epoch": 0.5909934629973368,
+      "grad_norm": 0.7451913952827454,
+      "learning_rate": 0.0001417333141007588,
+      "loss": 2.635,
+      "step": 7323
+    },
+    {
+      "epoch": 0.5910741667339198,
+      "grad_norm": 0.7405436038970947,
+      "learning_rate": 0.00014171896718108475,
+      "loss": 2.6014,
+      "step": 7324
+    },
+    {
+      "epoch": 0.5911548704705027,
+      "grad_norm": 0.7583999037742615,
+      "learning_rate": 0.00014170461922166498,
+      "loss": 2.6815,
+      "step": 7325
+    },
+    {
+      "epoch": 0.5912355742070858,
+      "grad_norm": 0.6653509140014648,
+      "learning_rate": 0.00014169027022285706,
+      "loss": 2.6153,
+      "step": 7326
+    },
+    {
+      "epoch": 0.5913162779436688,
+      "grad_norm": 0.7145548462867737,
+      "learning_rate": 0.00014167592018501864,
+      "loss": 2.6022,
+      "step": 7327
+    },
+    {
+      "epoch": 0.5913969816802518,
+      "grad_norm": 0.6996089816093445,
+      "learning_rate": 0.00014166156910850737,
+      "loss": 2.6586,
+      "step": 7328
+    },
+    {
+      "epoch": 0.5914776854168348,
+      "grad_norm": 0.735653281211853,
+      "learning_rate": 0.0001416472169936809,
+      "loss": 2.6084,
+      "step": 7329
+    },
+    {
+      "epoch": 0.5915583891534179,
+      "grad_norm": 0.695036768913269,
+      "learning_rate": 0.00014163286384089686,
+      "loss": 2.5058,
+      "step": 7330
+    },
+    {
+      "epoch": 0.5916390928900008,
+      "grad_norm": 0.9014756679534912,
+      "learning_rate": 0.00014161850965051307,
+      "loss": 2.5991,
+      "step": 7331
+    },
+    {
+      "epoch": 0.5917197966265838,
+      "grad_norm": 0.7079846858978271,
+      "learning_rate": 0.0001416041544228872,
+      "loss": 2.6067,
+      "step": 7332
+    },
+    {
+      "epoch": 0.5918005003631668,
+      "grad_norm": 0.7681204080581665,
+      "learning_rate": 0.00014158979815837705,
+      "loss": 2.5414,
+      "step": 7333
+    },
+    {
+      "epoch": 0.5918812040997499,
+      "grad_norm": 0.6501670479774475,
+      "learning_rate": 0.00014157544085734042,
+      "loss": 2.617,
+      "step": 7334
+    },
+    {
+      "epoch": 0.5919619078363328,
+      "grad_norm": 0.7573496103286743,
+      "learning_rate": 0.00014156108252013513,
+      "loss": 2.6341,
+      "step": 7335
+    },
+    {
+      "epoch": 0.5920426115729158,
+      "grad_norm": 0.6865558624267578,
+      "learning_rate": 0.00014154672314711903,
+      "loss": 2.6229,
+      "step": 7336
+    },
+    {
+      "epoch": 0.5921233153094988,
+      "grad_norm": 0.6859166622161865,
+      "learning_rate": 0.00014153236273864995,
+      "loss": 2.6149,
+      "step": 7337
+    },
+    {
+      "epoch": 0.5922040190460819,
+      "grad_norm": 0.7603647112846375,
+      "learning_rate": 0.00014151800129508585,
+      "loss": 2.5645,
+      "step": 7338
+    },
+    {
+      "epoch": 0.5922847227826649,
+      "grad_norm": 0.6740217208862305,
+      "learning_rate": 0.00014150363881678464,
+      "loss": 2.5883,
+      "step": 7339
+    },
+    {
+      "epoch": 0.5923654265192478,
+      "grad_norm": 0.6412263512611389,
+      "learning_rate": 0.00014148927530410426,
+      "loss": 2.576,
+      "step": 7340
+    },
+    {
+      "epoch": 0.5924461302558308,
+      "grad_norm": 0.669834315776825,
+      "learning_rate": 0.00014147491075740265,
+      "loss": 2.542,
+      "step": 7341
+    },
+    {
+      "epoch": 0.5925268339924139,
+      "grad_norm": 0.720024049282074,
+      "learning_rate": 0.00014146054517703786,
+      "loss": 2.6491,
+      "step": 7342
+    },
+    {
+      "epoch": 0.5926075377289969,
+      "grad_norm": 0.7191612720489502,
+      "learning_rate": 0.00014144617856336794,
+      "loss": 2.5933,
+      "step": 7343
+    },
+    {
+      "epoch": 0.5926882414655799,
+      "grad_norm": 0.7012050747871399,
+      "learning_rate": 0.00014143181091675087,
+      "loss": 2.5253,
+      "step": 7344
+    },
+    {
+      "epoch": 0.5927689452021628,
+      "grad_norm": 0.7825081944465637,
+      "learning_rate": 0.00014141744223754478,
+      "loss": 2.6225,
+      "step": 7345
+    },
+    {
+      "epoch": 0.5928496489387458,
+      "grad_norm": 0.6699295043945312,
+      "learning_rate": 0.00014140307252610775,
+      "loss": 2.5893,
+      "step": 7346
+    },
+    {
+      "epoch": 0.5929303526753289,
+      "grad_norm": 0.6668846011161804,
+      "learning_rate": 0.00014138870178279794,
+      "loss": 2.5944,
+      "step": 7347
+    },
+    {
+      "epoch": 0.5930110564119119,
+      "grad_norm": 0.7681072950363159,
+      "learning_rate": 0.0001413743300079735,
+      "loss": 2.5715,
+      "step": 7348
+    },
+    {
+      "epoch": 0.5930917601484949,
+      "grad_norm": 0.653075635433197,
+      "learning_rate": 0.00014135995720199258,
+      "loss": 2.5924,
+      "step": 7349
+    },
+    {
+      "epoch": 0.5931724638850778,
+      "grad_norm": 0.6807504892349243,
+      "learning_rate": 0.00014134558336521342,
+      "loss": 2.5395,
+      "step": 7350
+    },
+    {
+      "epoch": 0.5932531676216609,
+      "grad_norm": 0.681175708770752,
+      "learning_rate": 0.00014133120849799423,
+      "loss": 2.5401,
+      "step": 7351
+    },
+    {
+      "epoch": 0.5933338713582439,
+      "grad_norm": 0.7159900665283203,
+      "learning_rate": 0.0001413168326006933,
+      "loss": 2.5684,
+      "step": 7352
+    },
+    {
+      "epoch": 0.5934145750948269,
+      "grad_norm": 0.6517181992530823,
+      "learning_rate": 0.00014130245567366888,
+      "loss": 2.5887,
+      "step": 7353
+    },
+    {
+      "epoch": 0.5934952788314098,
+      "grad_norm": 0.6982731223106384,
+      "learning_rate": 0.00014128807771727936,
+      "loss": 2.5707,
+      "step": 7354
+    },
+    {
+      "epoch": 0.5935759825679929,
+      "grad_norm": 0.7003650069236755,
+      "learning_rate": 0.00014127369873188296,
+      "loss": 2.6415,
+      "step": 7355
+    },
+    {
+      "epoch": 0.5936566863045759,
+      "grad_norm": 0.7408339977264404,
+      "learning_rate": 0.0001412593187178381,
+      "loss": 2.5655,
+      "step": 7356
+    },
+    {
+      "epoch": 0.5937373900411589,
+      "grad_norm": 0.717218279838562,
+      "learning_rate": 0.00014124493767550317,
+      "loss": 2.586,
+      "step": 7357
+    },
+    {
+      "epoch": 0.5938180937777419,
+      "grad_norm": 0.6723458766937256,
+      "learning_rate": 0.00014123055560523657,
+      "loss": 2.593,
+      "step": 7358
+    },
+    {
+      "epoch": 0.593898797514325,
+      "grad_norm": 0.6861262321472168,
+      "learning_rate": 0.00014121617250739677,
+      "loss": 2.612,
+      "step": 7359
+    },
+    {
+      "epoch": 0.5939795012509079,
+      "grad_norm": 0.6811453104019165,
+      "learning_rate": 0.00014120178838234222,
+      "loss": 2.5708,
+      "step": 7360
+    },
+    {
+      "epoch": 0.5940602049874909,
+      "grad_norm": 0.6249656677246094,
+      "learning_rate": 0.00014118740323043136,
+      "loss": 2.5604,
+      "step": 7361
+    },
+    {
+      "epoch": 0.5941409087240739,
+      "grad_norm": 0.7671588659286499,
+      "learning_rate": 0.00014117301705202274,
+      "loss": 2.547,
+      "step": 7362
+    },
+    {
+      "epoch": 0.594221612460657,
+      "grad_norm": 0.6856057643890381,
+      "learning_rate": 0.00014115862984747496,
+      "loss": 2.6108,
+      "step": 7363
+    },
+    {
+      "epoch": 0.5943023161972399,
+      "grad_norm": 0.692331850528717,
+      "learning_rate": 0.0001411442416171465,
+      "loss": 2.6347,
+      "step": 7364
+    },
+    {
+      "epoch": 0.5943830199338229,
+      "grad_norm": 0.7256516814231873,
+      "learning_rate": 0.000141129852361396,
+      "loss": 2.6098,
+      "step": 7365
+    },
+    {
+      "epoch": 0.5944637236704059,
+      "grad_norm": 0.7522590160369873,
+      "learning_rate": 0.00014111546208058203,
+      "loss": 2.5688,
+      "step": 7366
+    },
+    {
+      "epoch": 0.594544427406989,
+      "grad_norm": 0.6915806531906128,
+      "learning_rate": 0.0001411010707750633,
+      "loss": 2.5899,
+      "step": 7367
+    },
+    {
+      "epoch": 0.594625131143572,
+      "grad_norm": 0.7355465292930603,
+      "learning_rate": 0.00014108667844519844,
+      "loss": 2.5212,
+      "step": 7368
+    },
+    {
+      "epoch": 0.5947058348801549,
+      "grad_norm": 0.731002926826477,
+      "learning_rate": 0.00014107228509134615,
+      "loss": 2.6369,
+      "step": 7369
+    },
+    {
+      "epoch": 0.5947865386167379,
+      "grad_norm": 0.6764423251152039,
+      "learning_rate": 0.0001410578907138652,
+      "loss": 2.6012,
+      "step": 7370
+    },
+    {
+      "epoch": 0.594867242353321,
+      "grad_norm": 0.7466071844100952,
+      "learning_rate": 0.0001410434953131142,
+      "loss": 2.5822,
+      "step": 7371
+    },
+    {
+      "epoch": 0.594947946089904,
+      "grad_norm": 0.7276137471199036,
+      "learning_rate": 0.00014102909888945205,
+      "loss": 2.6055,
+      "step": 7372
+    },
+    {
+      "epoch": 0.595028649826487,
+      "grad_norm": 0.7411746978759766,
+      "learning_rate": 0.00014101470144323752,
+      "loss": 2.6489,
+      "step": 7373
+    },
+    {
+      "epoch": 0.5951093535630699,
+      "grad_norm": 0.7511908411979675,
+      "learning_rate": 0.0001410003029748294,
+      "loss": 2.6268,
+      "step": 7374
+    },
+    {
+      "epoch": 0.595190057299653,
+      "grad_norm": 0.6623562574386597,
+      "learning_rate": 0.0001409859034845866,
+      "loss": 2.58,
+      "step": 7375
+    },
+    {
+      "epoch": 0.595270761036236,
+      "grad_norm": 0.6948572397232056,
+      "learning_rate": 0.00014097150297286785,
+      "loss": 2.5811,
+      "step": 7376
+    },
+    {
+      "epoch": 0.595351464772819,
+      "grad_norm": 0.6836786270141602,
+      "learning_rate": 0.0001409571014400322,
+      "loss": 2.5861,
+      "step": 7377
+    },
+    {
+      "epoch": 0.595432168509402,
+      "grad_norm": 0.6644341945648193,
+      "learning_rate": 0.00014094269888643854,
+      "loss": 2.6339,
+      "step": 7378
+    },
+    {
+      "epoch": 0.595512872245985,
+      "grad_norm": 0.6434289813041687,
+      "learning_rate": 0.0001409282953124458,
+      "loss": 2.4897,
+      "step": 7379
+    },
+    {
+      "epoch": 0.595593575982568,
+      "grad_norm": 0.6745082139968872,
+      "learning_rate": 0.0001409138907184129,
+      "loss": 2.522,
+      "step": 7380
+    },
+    {
+      "epoch": 0.595674279719151,
+      "grad_norm": 0.725321352481842,
+      "learning_rate": 0.0001408994851046989,
+      "loss": 2.5711,
+      "step": 7381
+    },
+    {
+      "epoch": 0.595754983455734,
+      "grad_norm": 0.7485500574111938,
+      "learning_rate": 0.00014088507847166283,
+      "loss": 2.6095,
+      "step": 7382
+    },
+    {
+      "epoch": 0.595835687192317,
+      "grad_norm": 0.721125602722168,
+      "learning_rate": 0.00014087067081966376,
+      "loss": 2.6762,
+      "step": 7383
+    },
+    {
+      "epoch": 0.5959163909289,
+      "grad_norm": 0.7099901437759399,
+      "learning_rate": 0.00014085626214906073,
+      "loss": 2.5667,
+      "step": 7384
+    },
+    {
+      "epoch": 0.595997094665483,
+      "grad_norm": 0.6889060139656067,
+      "learning_rate": 0.00014084185246021283,
+      "loss": 2.6723,
+      "step": 7385
+    },
+    {
+      "epoch": 0.596077798402066,
+      "grad_norm": 0.735698938369751,
+      "learning_rate": 0.00014082744175347923,
+      "loss": 2.6434,
+      "step": 7386
+    },
+    {
+      "epoch": 0.5961585021386491,
+      "grad_norm": 0.7603070735931396,
+      "learning_rate": 0.00014081303002921902,
+      "loss": 2.665,
+      "step": 7387
+    },
+    {
+      "epoch": 0.596239205875232,
+      "grad_norm": 0.6786355376243591,
+      "learning_rate": 0.00014079861728779141,
+      "loss": 2.5842,
+      "step": 7388
+    },
+    {
+      "epoch": 0.596319909611815,
+      "grad_norm": 0.6693331003189087,
+      "learning_rate": 0.00014078420352955565,
+      "loss": 2.6211,
+      "step": 7389
+    },
+    {
+      "epoch": 0.596400613348398,
+      "grad_norm": 0.74013751745224,
+      "learning_rate": 0.0001407697887548709,
+      "loss": 2.5886,
+      "step": 7390
+    },
+    {
+      "epoch": 0.5964813170849811,
+      "grad_norm": 0.739507257938385,
+      "learning_rate": 0.00014075537296409646,
+      "loss": 2.607,
+      "step": 7391
+    },
+    {
+      "epoch": 0.5965620208215641,
+      "grad_norm": 0.7121848464012146,
+      "learning_rate": 0.00014074095615759156,
+      "loss": 2.6052,
+      "step": 7392
+    },
+    {
+      "epoch": 0.596642724558147,
+      "grad_norm": 0.7526760697364807,
+      "learning_rate": 0.00014072653833571556,
+      "loss": 2.6051,
+      "step": 7393
+    },
+    {
+      "epoch": 0.59672342829473,
+      "grad_norm": 0.7867496609687805,
+      "learning_rate": 0.00014071211949882777,
+      "loss": 2.6228,
+      "step": 7394
+    },
+    {
+      "epoch": 0.596804132031313,
+      "grad_norm": 0.7527757883071899,
+      "learning_rate": 0.00014069769964728752,
+      "loss": 2.6793,
+      "step": 7395
+    },
+    {
+      "epoch": 0.5968848357678961,
+      "grad_norm": 0.7096899747848511,
+      "learning_rate": 0.00014068327878145423,
+      "loss": 2.5207,
+      "step": 7396
+    },
+    {
+      "epoch": 0.5969655395044791,
+      "grad_norm": 0.6863983869552612,
+      "learning_rate": 0.00014066885690168726,
+      "loss": 2.7059,
+      "step": 7397
+    },
+    {
+      "epoch": 0.597046243241062,
+      "grad_norm": 0.7782251834869385,
+      "learning_rate": 0.0001406544340083461,
+      "loss": 2.6232,
+      "step": 7398
+    },
+    {
+      "epoch": 0.597126946977645,
+      "grad_norm": 0.6944136619567871,
+      "learning_rate": 0.00014064001010179013,
+      "loss": 2.6134,
+      "step": 7399
+    },
+    {
+      "epoch": 0.5972076507142281,
+      "grad_norm": 0.7629704475402832,
+      "learning_rate": 0.00014062558518237892,
+      "loss": 2.5358,
+      "step": 7400
+    },
+    {
+      "epoch": 0.5972883544508111,
+      "grad_norm": 0.6922330260276794,
+      "learning_rate": 0.0001406111592504719,
+      "loss": 2.5457,
+      "step": 7401
+    },
+    {
+      "epoch": 0.597369058187394,
+      "grad_norm": 0.6992952227592468,
+      "learning_rate": 0.00014059673230642865,
+      "loss": 2.6241,
+      "step": 7402
+    },
+    {
+      "epoch": 0.597449761923977,
+      "grad_norm": 0.6587642431259155,
+      "learning_rate": 0.0001405823043506087,
+      "loss": 2.5867,
+      "step": 7403
+    },
+    {
+      "epoch": 0.5975304656605601,
+      "grad_norm": 0.6993013024330139,
+      "learning_rate": 0.00014056787538337164,
+      "loss": 2.6194,
+      "step": 7404
+    },
+    {
+      "epoch": 0.5976111693971431,
+      "grad_norm": 0.7605414986610413,
+      "learning_rate": 0.0001405534454050771,
+      "loss": 2.607,
+      "step": 7405
+    },
+    {
+      "epoch": 0.5976918731337261,
+      "grad_norm": 0.6624562740325928,
+      "learning_rate": 0.00014053901441608466,
+      "loss": 2.5962,
+      "step": 7406
+    },
+    {
+      "epoch": 0.597772576870309,
+      "grad_norm": 0.7432621717453003,
+      "learning_rate": 0.000140524582416754,
+      "loss": 2.6434,
+      "step": 7407
+    },
+    {
+      "epoch": 0.5978532806068921,
+      "grad_norm": 0.7184053659439087,
+      "learning_rate": 0.00014051014940744488,
+      "loss": 2.6139,
+      "step": 7408
+    },
+    {
+      "epoch": 0.5979339843434751,
+      "grad_norm": 0.7567455768585205,
+      "learning_rate": 0.00014049571538851687,
+      "loss": 2.5788,
+      "step": 7409
+    },
+    {
+      "epoch": 0.5980146880800581,
+      "grad_norm": 0.6759883761405945,
+      "learning_rate": 0.00014048128036032984,
+      "loss": 2.5584,
+      "step": 7410
+    },
+    {
+      "epoch": 0.5980953918166411,
+      "grad_norm": 0.7607424855232239,
+      "learning_rate": 0.00014046684432324343,
+      "loss": 2.5675,
+      "step": 7411
+    },
+    {
+      "epoch": 0.5981760955532242,
+      "grad_norm": 0.7134036421775818,
+      "learning_rate": 0.00014045240727761748,
+      "loss": 2.6805,
+      "step": 7412
+    },
+    {
+      "epoch": 0.5982567992898071,
+      "grad_norm": 0.6996984481811523,
+      "learning_rate": 0.00014043796922381184,
+      "loss": 2.5874,
+      "step": 7413
+    },
+    {
+      "epoch": 0.5983375030263901,
+      "grad_norm": 0.7098252177238464,
+      "learning_rate": 0.00014042353016218627,
+      "loss": 2.5895,
+      "step": 7414
+    },
+    {
+      "epoch": 0.5984182067629731,
+      "grad_norm": 0.7160520553588867,
+      "learning_rate": 0.00014040909009310068,
+      "loss": 2.6042,
+      "step": 7415
+    },
+    {
+      "epoch": 0.5984989104995562,
+      "grad_norm": 0.6727281212806702,
+      "learning_rate": 0.00014039464901691493,
+      "loss": 2.5356,
+      "step": 7416
+    },
+    {
+      "epoch": 0.5985796142361391,
+      "grad_norm": 0.7052881717681885,
+      "learning_rate": 0.00014038020693398891,
+      "loss": 2.6093,
+      "step": 7417
+    },
+    {
+      "epoch": 0.5986603179727221,
+      "grad_norm": 0.7151781916618347,
+      "learning_rate": 0.00014036576384468262,
+      "loss": 2.5776,
+      "step": 7418
+    },
+    {
+      "epoch": 0.5987410217093051,
+      "grad_norm": 0.7376574873924255,
+      "learning_rate": 0.0001403513197493559,
+      "loss": 2.6246,
+      "step": 7419
+    },
+    {
+      "epoch": 0.5988217254458882,
+      "grad_norm": 0.6882135272026062,
+      "learning_rate": 0.00014033687464836892,
+      "loss": 2.6028,
+      "step": 7420
+    },
+    {
+      "epoch": 0.5989024291824712,
+      "grad_norm": 0.6603999137878418,
+      "learning_rate": 0.00014032242854208153,
+      "loss": 2.5897,
+      "step": 7421
+    },
+    {
+      "epoch": 0.5989831329190541,
+      "grad_norm": 0.7001559734344482,
+      "learning_rate": 0.0001403079814308538,
+      "loss": 2.6033,
+      "step": 7422
+    },
+    {
+      "epoch": 0.5990638366556371,
+      "grad_norm": 0.7184363603591919,
+      "learning_rate": 0.00014029353331504582,
+      "loss": 2.7464,
+      "step": 7423
+    },
+    {
+      "epoch": 0.5991445403922202,
+      "grad_norm": 0.6794769167900085,
+      "learning_rate": 0.00014027908419501767,
+      "loss": 2.569,
+      "step": 7424
+    },
+    {
+      "epoch": 0.5992252441288032,
+      "grad_norm": 0.6846041083335876,
+      "learning_rate": 0.00014026463407112942,
+      "loss": 2.5995,
+      "step": 7425
+    },
+    {
+      "epoch": 0.5993059478653862,
+      "grad_norm": 0.6539658308029175,
+      "learning_rate": 0.00014025018294374129,
+      "loss": 2.5749,
+      "step": 7426
+    },
+    {
+      "epoch": 0.5993866516019691,
+      "grad_norm": 0.6572301983833313,
+      "learning_rate": 0.00014023573081321336,
+      "loss": 2.5312,
+      "step": 7427
+    },
+    {
+      "epoch": 0.5994673553385522,
+      "grad_norm": 0.7010765671730042,
+      "learning_rate": 0.00014022127767990581,
+      "loss": 2.5088,
+      "step": 7428
+    },
+    {
+      "epoch": 0.5995480590751352,
+      "grad_norm": 0.7193396091461182,
+      "learning_rate": 0.0001402068235441789,
+      "loss": 2.6193,
+      "step": 7429
+    },
+    {
+      "epoch": 0.5996287628117182,
+      "grad_norm": 0.6928533315658569,
+      "learning_rate": 0.00014019236840639288,
+      "loss": 2.6149,
+      "step": 7430
+    },
+    {
+      "epoch": 0.5997094665483012,
+      "grad_norm": 0.743658185005188,
+      "learning_rate": 0.00014017791226690794,
+      "loss": 2.5466,
+      "step": 7431
+    },
+    {
+      "epoch": 0.5997901702848842,
+      "grad_norm": 0.752082347869873,
+      "learning_rate": 0.0001401634551260844,
+      "loss": 2.6605,
+      "step": 7432
+    },
+    {
+      "epoch": 0.5998708740214672,
+      "grad_norm": 0.7280415296554565,
+      "learning_rate": 0.00014014899698428255,
+      "loss": 2.6128,
+      "step": 7433
+    },
+    {
+      "epoch": 0.5999515777580502,
+      "grad_norm": 0.7037710547447205,
+      "learning_rate": 0.0001401345378418628,
+      "loss": 2.6157,
+      "step": 7434
+    },
+    {
+      "epoch": 0.6000322814946332,
+      "grad_norm": 0.6984395980834961,
+      "learning_rate": 0.00014012007769918542,
+      "loss": 2.5579,
+      "step": 7435
+    },
+    {
+      "epoch": 0.6001129852312163,
+      "grad_norm": 0.6853601336479187,
+      "learning_rate": 0.00014010561655661085,
+      "loss": 2.6316,
+      "step": 7436
+    },
+    {
+      "epoch": 0.6001936889677992,
+      "grad_norm": 0.7551750540733337,
+      "learning_rate": 0.00014009115441449948,
+      "loss": 2.6671,
+      "step": 7437
+    },
+    {
+      "epoch": 0.6002743927043822,
+      "grad_norm": 0.7680155038833618,
+      "learning_rate": 0.0001400766912732117,
+      "loss": 2.6301,
+      "step": 7438
+    },
+    {
+      "epoch": 0.6003550964409652,
+      "grad_norm": 0.6757175922393799,
+      "learning_rate": 0.00014006222713310807,
+      "loss": 2.5584,
+      "step": 7439
+    },
+    {
+      "epoch": 0.6004358001775483,
+      "grad_norm": 0.6636163592338562,
+      "learning_rate": 0.00014004776199454897,
+      "loss": 2.5437,
+      "step": 7440
+    },
+    {
+      "epoch": 0.6005165039141312,
+      "grad_norm": 0.7317774891853333,
+      "learning_rate": 0.00014003329585789498,
+      "loss": 2.594,
+      "step": 7441
+    },
+    {
+      "epoch": 0.6005972076507142,
+      "grad_norm": 0.6903451681137085,
+      "learning_rate": 0.0001400188287235066,
+      "loss": 2.6175,
+      "step": 7442
+    },
+    {
+      "epoch": 0.6006779113872972,
+      "grad_norm": 0.7137858867645264,
+      "learning_rate": 0.00014000436059174437,
+      "loss": 2.6411,
+      "step": 7443
+    },
+    {
+      "epoch": 0.6007586151238803,
+      "grad_norm": 0.7124149203300476,
+      "learning_rate": 0.00013998989146296893,
+      "loss": 2.6562,
+      "step": 7444
+    },
+    {
+      "epoch": 0.6008393188604633,
+      "grad_norm": 0.7518175840377808,
+      "learning_rate": 0.00013997542133754087,
+      "loss": 2.6213,
+      "step": 7445
+    },
+    {
+      "epoch": 0.6009200225970462,
+      "grad_norm": 0.6843053698539734,
+      "learning_rate": 0.0001399609502158208,
+      "loss": 2.6099,
+      "step": 7446
+    },
+    {
+      "epoch": 0.6010007263336292,
+      "grad_norm": 0.6668025255203247,
+      "learning_rate": 0.0001399464780981694,
+      "loss": 2.609,
+      "step": 7447
+    },
+    {
+      "epoch": 0.6010814300702122,
+      "grad_norm": 0.6849119067192078,
+      "learning_rate": 0.00013993200498494735,
+      "loss": 2.6097,
+      "step": 7448
+    },
+    {
+      "epoch": 0.6011621338067953,
+      "grad_norm": 0.7767381072044373,
+      "learning_rate": 0.0001399175308765153,
+      "loss": 2.6351,
+      "step": 7449
+    },
+    {
+      "epoch": 0.6012428375433783,
+      "grad_norm": 0.6630256772041321,
+      "learning_rate": 0.0001399030557732341,
+      "loss": 2.5924,
+      "step": 7450
+    },
+    {
+      "epoch": 0.6013235412799612,
+      "grad_norm": 0.6918755769729614,
+      "learning_rate": 0.00013988857967546444,
+      "loss": 2.6205,
+      "step": 7451
+    },
+    {
+      "epoch": 0.6014042450165442,
+      "grad_norm": 0.7179181575775146,
+      "learning_rate": 0.00013987410258356708,
+      "loss": 2.5971,
+      "step": 7452
+    },
+    {
+      "epoch": 0.6014849487531273,
+      "grad_norm": 0.7233672738075256,
+      "learning_rate": 0.00013985962449790284,
+      "loss": 2.595,
+      "step": 7453
+    },
+    {
+      "epoch": 0.6015656524897103,
+      "grad_norm": 0.6861593127250671,
+      "learning_rate": 0.0001398451454188326,
+      "loss": 2.6127,
+      "step": 7454
+    },
+    {
+      "epoch": 0.6016463562262933,
+      "grad_norm": 0.6818981170654297,
+      "learning_rate": 0.00013983066534671714,
+      "loss": 2.5923,
+      "step": 7455
+    },
+    {
+      "epoch": 0.6017270599628762,
+      "grad_norm": 0.700036346912384,
+      "learning_rate": 0.0001398161842819174,
+      "loss": 2.5474,
+      "step": 7456
+    },
+    {
+      "epoch": 0.6018077636994593,
+      "grad_norm": 0.6884824633598328,
+      "learning_rate": 0.00013980170222479426,
+      "loss": 2.6041,
+      "step": 7457
+    },
+    {
+      "epoch": 0.6018884674360423,
+      "grad_norm": 0.6745120286941528,
+      "learning_rate": 0.00013978721917570866,
+      "loss": 2.6638,
+      "step": 7458
+    },
+    {
+      "epoch": 0.6019691711726253,
+      "grad_norm": 0.6886256337165833,
+      "learning_rate": 0.00013977273513502157,
+      "loss": 2.5733,
+      "step": 7459
+    },
+    {
+      "epoch": 0.6020498749092082,
+      "grad_norm": 0.7220930457115173,
+      "learning_rate": 0.00013975825010309394,
+      "loss": 2.5739,
+      "step": 7460
+    },
+    {
+      "epoch": 0.6021305786457913,
+      "grad_norm": 0.7281780242919922,
+      "learning_rate": 0.0001397437640802868,
+      "loss": 2.5646,
+      "step": 7461
+    },
+    {
+      "epoch": 0.6022112823823743,
+      "grad_norm": 0.7316896915435791,
+      "learning_rate": 0.00013972927706696115,
+      "loss": 2.6532,
+      "step": 7462
+    },
+    {
+      "epoch": 0.6022919861189573,
+      "grad_norm": 0.6288646459579468,
+      "learning_rate": 0.00013971478906347806,
+      "loss": 2.5753,
+      "step": 7463
+    },
+    {
+      "epoch": 0.6023726898555403,
+      "grad_norm": 0.7110145688056946,
+      "learning_rate": 0.00013970030007019862,
+      "loss": 2.6421,
+      "step": 7464
+    },
+    {
+      "epoch": 0.6024533935921234,
+      "grad_norm": 0.7437754273414612,
+      "learning_rate": 0.00013968581008748393,
+      "loss": 2.585,
+      "step": 7465
+    },
+    {
+      "epoch": 0.6025340973287063,
+      "grad_norm": 0.6839718222618103,
+      "learning_rate": 0.00013967131911569514,
+      "loss": 2.6249,
+      "step": 7466
+    },
+    {
+      "epoch": 0.6026148010652893,
+      "grad_norm": 0.7358397841453552,
+      "learning_rate": 0.00013965682715519332,
+      "loss": 2.597,
+      "step": 7467
+    },
+    {
+      "epoch": 0.6026955048018723,
+      "grad_norm": 0.673651397228241,
+      "learning_rate": 0.00013964233420633973,
+      "loss": 2.6111,
+      "step": 7468
+    },
+    {
+      "epoch": 0.6027762085384554,
+      "grad_norm": 0.7390083074569702,
+      "learning_rate": 0.00013962784026949553,
+      "loss": 2.6131,
+      "step": 7469
+    },
+    {
+      "epoch": 0.6028569122750383,
+      "grad_norm": 0.6902220249176025,
+      "learning_rate": 0.00013961334534502197,
+      "loss": 2.6116,
+      "step": 7470
+    },
+    {
+      "epoch": 0.6029376160116213,
+      "grad_norm": 0.6946651935577393,
+      "learning_rate": 0.00013959884943328033,
+      "loss": 2.6307,
+      "step": 7471
+    },
+    {
+      "epoch": 0.6030183197482043,
+      "grad_norm": 0.7277294993400574,
+      "learning_rate": 0.00013958435253463183,
+      "loss": 2.6065,
+      "step": 7472
+    },
+    {
+      "epoch": 0.6030990234847874,
+      "grad_norm": 0.743833601474762,
+      "learning_rate": 0.00013956985464943776,
+      "loss": 2.6644,
+      "step": 7473
+    },
+    {
+      "epoch": 0.6031797272213704,
+      "grad_norm": 0.6480288505554199,
+      "learning_rate": 0.0001395553557780595,
+      "loss": 2.5386,
+      "step": 7474
+    },
+    {
+      "epoch": 0.6032604309579533,
+      "grad_norm": 0.799443781375885,
+      "learning_rate": 0.00013954085592085834,
+      "loss": 2.5653,
+      "step": 7475
+    },
+    {
+      "epoch": 0.6033411346945363,
+      "grad_norm": 0.6790705323219299,
+      "learning_rate": 0.00013952635507819575,
+      "loss": 2.6229,
+      "step": 7476
+    },
+    {
+      "epoch": 0.6034218384311194,
+      "grad_norm": 0.6871588826179504,
+      "learning_rate": 0.00013951185325043302,
+      "loss": 2.6514,
+      "step": 7477
+    },
+    {
+      "epoch": 0.6035025421677024,
+      "grad_norm": 0.7236921787261963,
+      "learning_rate": 0.00013949735043793164,
+      "loss": 2.5931,
+      "step": 7478
+    },
+    {
+      "epoch": 0.6035832459042854,
+      "grad_norm": 0.6888518929481506,
+      "learning_rate": 0.00013948284664105305,
+      "loss": 2.6408,
+      "step": 7479
+    },
+    {
+      "epoch": 0.6036639496408683,
+      "grad_norm": 0.7292625904083252,
+      "learning_rate": 0.00013946834186015868,
+      "loss": 2.5829,
+      "step": 7480
+    },
+    {
+      "epoch": 0.6037446533774514,
+      "grad_norm": 0.6755293607711792,
+      "learning_rate": 0.00013945383609561009,
+      "loss": 2.5917,
+      "step": 7481
+    },
+    {
+      "epoch": 0.6038253571140344,
+      "grad_norm": 0.6808032989501953,
+      "learning_rate": 0.00013943932934776877,
+      "loss": 2.6103,
+      "step": 7482
+    },
+    {
+      "epoch": 0.6039060608506174,
+      "grad_norm": 0.747173547744751,
+      "learning_rate": 0.00013942482161699625,
+      "loss": 2.624,
+      "step": 7483
+    },
+    {
+      "epoch": 0.6039867645872004,
+      "grad_norm": 0.7265594005584717,
+      "learning_rate": 0.00013941031290365413,
+      "loss": 2.5672,
+      "step": 7484
+    },
+    {
+      "epoch": 0.6040674683237834,
+      "grad_norm": 0.6434060335159302,
+      "learning_rate": 0.000139395803208104,
+      "loss": 2.5885,
+      "step": 7485
+    },
+    {
+      "epoch": 0.6041481720603664,
+      "grad_norm": 0.7148730754852295,
+      "learning_rate": 0.00013938129253070747,
+      "loss": 2.6466,
+      "step": 7486
+    },
+    {
+      "epoch": 0.6042288757969494,
+      "grad_norm": 0.7724708318710327,
+      "learning_rate": 0.00013936678087182616,
+      "loss": 2.6364,
+      "step": 7487
+    },
+    {
+      "epoch": 0.6043095795335324,
+      "grad_norm": 0.6886702179908752,
+      "learning_rate": 0.0001393522682318218,
+      "loss": 2.5844,
+      "step": 7488
+    },
+    {
+      "epoch": 0.6043902832701155,
+      "grad_norm": 0.6501082181930542,
+      "learning_rate": 0.00013933775461105603,
+      "loss": 2.5767,
+      "step": 7489
+    },
+    {
+      "epoch": 0.6044709870066984,
+      "grad_norm": 0.7333959341049194,
+      "learning_rate": 0.00013932324000989058,
+      "loss": 2.5735,
+      "step": 7490
+    },
+    {
+      "epoch": 0.6045516907432814,
+      "grad_norm": 0.7057361602783203,
+      "learning_rate": 0.00013930872442868722,
+      "loss": 2.627,
+      "step": 7491
+    },
+    {
+      "epoch": 0.6046323944798644,
+      "grad_norm": 0.705078661441803,
+      "learning_rate": 0.00013929420786780767,
+      "loss": 2.6012,
+      "step": 7492
+    },
+    {
+      "epoch": 0.6047130982164475,
+      "grad_norm": 0.7192156314849854,
+      "learning_rate": 0.00013927969032761378,
+      "loss": 2.5594,
+      "step": 7493
+    },
+    {
+      "epoch": 0.6047938019530305,
+      "grad_norm": 0.703116774559021,
+      "learning_rate": 0.00013926517180846726,
+      "loss": 2.6099,
+      "step": 7494
+    },
+    {
+      "epoch": 0.6048745056896134,
+      "grad_norm": 0.6970264315605164,
+      "learning_rate": 0.00013925065231073006,
+      "loss": 2.5832,
+      "step": 7495
+    },
+    {
+      "epoch": 0.6049552094261964,
+      "grad_norm": 0.7308031320571899,
+      "learning_rate": 0.00013923613183476402,
+      "loss": 2.586,
+      "step": 7496
+    },
+    {
+      "epoch": 0.6050359131627794,
+      "grad_norm": 0.7212777137756348,
+      "learning_rate": 0.00013922161038093097,
+      "loss": 2.6374,
+      "step": 7497
+    },
+    {
+      "epoch": 0.6051166168993625,
+      "grad_norm": 0.6644641757011414,
+      "learning_rate": 0.0001392070879495929,
+      "loss": 2.5226,
+      "step": 7498
+    },
+    {
+      "epoch": 0.6051973206359454,
+      "grad_norm": 0.6683016419410706,
+      "learning_rate": 0.0001391925645411117,
+      "loss": 2.5279,
+      "step": 7499
+    },
+    {
+      "epoch": 0.6052780243725284,
+      "grad_norm": 0.7341439127922058,
+      "learning_rate": 0.00013917804015584932,
+      "loss": 2.5995,
+      "step": 7500
+    },
+    {
+      "epoch": 0.6053587281091114,
+      "grad_norm": 0.753942608833313,
+      "learning_rate": 0.0001391635147941678,
+      "loss": 2.5706,
+      "step": 7501
+    },
+    {
+      "epoch": 0.6054394318456945,
+      "grad_norm": 0.7541958093643188,
+      "learning_rate": 0.00013914898845642908,
+      "loss": 2.6365,
+      "step": 7502
+    },
+    {
+      "epoch": 0.6055201355822775,
+      "grad_norm": 0.6583349108695984,
+      "learning_rate": 0.00013913446114299528,
+      "loss": 2.534,
+      "step": 7503
+    },
+    {
+      "epoch": 0.6056008393188604,
+      "grad_norm": 0.6545756459236145,
+      "learning_rate": 0.00013911993285422835,
+      "loss": 2.5443,
+      "step": 7504
+    },
+    {
+      "epoch": 0.6056815430554434,
+      "grad_norm": 0.8290210366249084,
+      "learning_rate": 0.00013910540359049045,
+      "loss": 2.6196,
+      "step": 7505
+    },
+    {
+      "epoch": 0.6057622467920265,
+      "grad_norm": 0.7032577395439148,
+      "learning_rate": 0.0001390908733521437,
+      "loss": 2.6575,
+      "step": 7506
+    },
+    {
+      "epoch": 0.6058429505286095,
+      "grad_norm": 0.7018071413040161,
+      "learning_rate": 0.0001390763421395502,
+      "loss": 2.6272,
+      "step": 7507
+    },
+    {
+      "epoch": 0.6059236542651925,
+      "grad_norm": 0.6288552284240723,
+      "learning_rate": 0.00013906180995307206,
+      "loss": 2.5295,
+      "step": 7508
+    },
+    {
+      "epoch": 0.6060043580017754,
+      "grad_norm": 0.7013774514198303,
+      "learning_rate": 0.00013904727679307153,
+      "loss": 2.5669,
+      "step": 7509
+    },
+    {
+      "epoch": 0.6060850617383585,
+      "grad_norm": 0.6811630129814148,
+      "learning_rate": 0.00013903274265991082,
+      "loss": 2.5827,
+      "step": 7510
+    },
+    {
+      "epoch": 0.6061657654749415,
+      "grad_norm": 0.6690269112586975,
+      "learning_rate": 0.0001390182075539521,
+      "loss": 2.5947,
+      "step": 7511
+    },
+    {
+      "epoch": 0.6062464692115245,
+      "grad_norm": 0.6946289539337158,
+      "learning_rate": 0.00013900367147555768,
+      "loss": 2.59,
+      "step": 7512
+    },
+    {
+      "epoch": 0.6063271729481075,
+      "grad_norm": 0.7302843332290649,
+      "learning_rate": 0.0001389891344250898,
+      "loss": 2.5994,
+      "step": 7513
+    },
+    {
+      "epoch": 0.6064078766846905,
+      "grad_norm": 0.7462306022644043,
+      "learning_rate": 0.00013897459640291074,
+      "loss": 2.5983,
+      "step": 7514
+    },
+    {
+      "epoch": 0.6064885804212735,
+      "grad_norm": 0.6948123574256897,
+      "learning_rate": 0.0001389600574093829,
+      "loss": 2.5737,
+      "step": 7515
+    },
+    {
+      "epoch": 0.6065692841578565,
+      "grad_norm": 0.6897372007369995,
+      "learning_rate": 0.00013894551744486857,
+      "loss": 2.607,
+      "step": 7516
+    },
+    {
+      "epoch": 0.6066499878944395,
+      "grad_norm": 0.6808069348335266,
+      "learning_rate": 0.00013893097650973015,
+      "loss": 2.5712,
+      "step": 7517
+    },
+    {
+      "epoch": 0.6067306916310226,
+      "grad_norm": 0.7000731229782104,
+      "learning_rate": 0.00013891643460433,
+      "loss": 2.5654,
+      "step": 7518
+    },
+    {
+      "epoch": 0.6068113953676055,
+      "grad_norm": 0.7197545766830444,
+      "learning_rate": 0.0001389018917290306,
+      "loss": 2.5705,
+      "step": 7519
+    },
+    {
+      "epoch": 0.6068920991041885,
+      "grad_norm": 0.7001069188117981,
+      "learning_rate": 0.00013888734788419433,
+      "loss": 2.5934,
+      "step": 7520
+    },
+    {
+      "epoch": 0.6069728028407715,
+      "grad_norm": 0.7480459213256836,
+      "learning_rate": 0.00013887280307018377,
+      "loss": 2.5211,
+      "step": 7521
+    },
+    {
+      "epoch": 0.6070535065773546,
+      "grad_norm": 0.6913945078849792,
+      "learning_rate": 0.00013885825728736132,
+      "loss": 2.6013,
+      "step": 7522
+    },
+    {
+      "epoch": 0.6071342103139376,
+      "grad_norm": 0.6527336239814758,
+      "learning_rate": 0.00013884371053608948,
+      "loss": 2.5901,
+      "step": 7523
+    },
+    {
+      "epoch": 0.6072149140505205,
+      "grad_norm": 0.6897335052490234,
+      "learning_rate": 0.00013882916281673086,
+      "loss": 2.5389,
+      "step": 7524
+    },
+    {
+      "epoch": 0.6072956177871035,
+      "grad_norm": 0.7159501910209656,
+      "learning_rate": 0.00013881461412964798,
+      "loss": 2.5399,
+      "step": 7525
+    },
+    {
+      "epoch": 0.6073763215236866,
+      "grad_norm": 0.6744364500045776,
+      "learning_rate": 0.00013880006447520346,
+      "loss": 2.5658,
+      "step": 7526
+    },
+    {
+      "epoch": 0.6074570252602696,
+      "grad_norm": 0.819950520992279,
+      "learning_rate": 0.00013878551385375994,
+      "loss": 2.6143,
+      "step": 7527
+    },
+    {
+      "epoch": 0.6075377289968525,
+      "grad_norm": 0.744293212890625,
+      "learning_rate": 0.00013877096226568,
+      "loss": 2.6565,
+      "step": 7528
+    },
+    {
+      "epoch": 0.6076184327334355,
+      "grad_norm": 0.7121254205703735,
+      "learning_rate": 0.00013875640971132636,
+      "loss": 2.6151,
+      "step": 7529
+    },
+    {
+      "epoch": 0.6076991364700186,
+      "grad_norm": 0.7616204023361206,
+      "learning_rate": 0.00013874185619106163,
+      "loss": 2.6395,
+      "step": 7530
+    },
+    {
+      "epoch": 0.6077798402066016,
+      "grad_norm": 0.7481076121330261,
+      "learning_rate": 0.0001387273017052486,
+      "loss": 2.597,
+      "step": 7531
+    },
+    {
+      "epoch": 0.6078605439431846,
+      "grad_norm": 0.6660816073417664,
+      "learning_rate": 0.00013871274625425,
+      "loss": 2.5696,
+      "step": 7532
+    },
+    {
+      "epoch": 0.6079412476797675,
+      "grad_norm": 0.7491411566734314,
+      "learning_rate": 0.00013869818983842854,
+      "loss": 2.552,
+      "step": 7533
+    },
+    {
+      "epoch": 0.6080219514163506,
+      "grad_norm": 0.7130792140960693,
+      "learning_rate": 0.00013868363245814704,
+      "loss": 2.5959,
+      "step": 7534
+    },
+    {
+      "epoch": 0.6081026551529336,
+      "grad_norm": 0.7157341241836548,
+      "learning_rate": 0.00013866907411376827,
+      "loss": 2.5598,
+      "step": 7535
+    },
+    {
+      "epoch": 0.6081833588895166,
+      "grad_norm": 0.7750656008720398,
+      "learning_rate": 0.00013865451480565513,
+      "loss": 2.6217,
+      "step": 7536
+    },
+    {
+      "epoch": 0.6082640626260996,
+      "grad_norm": 0.6915080547332764,
+      "learning_rate": 0.00013863995453417043,
+      "loss": 2.6211,
+      "step": 7537
+    },
+    {
+      "epoch": 0.6083447663626826,
+      "grad_norm": 0.7245940566062927,
+      "learning_rate": 0.00013862539329967706,
+      "loss": 2.5619,
+      "step": 7538
+    },
+    {
+      "epoch": 0.6084254700992656,
+      "grad_norm": 0.8884119391441345,
+      "learning_rate": 0.0001386108311025379,
+      "loss": 2.6349,
+      "step": 7539
+    },
+    {
+      "epoch": 0.6085061738358486,
+      "grad_norm": 0.7889477610588074,
+      "learning_rate": 0.0001385962679431159,
+      "loss": 2.6169,
+      "step": 7540
+    },
+    {
+      "epoch": 0.6085868775724316,
+      "grad_norm": 0.7187505960464478,
+      "learning_rate": 0.00013858170382177403,
+      "loss": 2.5582,
+      "step": 7541
+    },
+    {
+      "epoch": 0.6086675813090147,
+      "grad_norm": 0.7502198219299316,
+      "learning_rate": 0.00013856713873887526,
+      "loss": 2.5418,
+      "step": 7542
+    },
+    {
+      "epoch": 0.6087482850455976,
+      "grad_norm": 0.797704815864563,
+      "learning_rate": 0.00013855257269478256,
+      "loss": 2.5764,
+      "step": 7543
+    },
+    {
+      "epoch": 0.6088289887821806,
+      "grad_norm": 0.7651431560516357,
+      "learning_rate": 0.00013853800568985896,
+      "loss": 2.5995,
+      "step": 7544
+    },
+    {
+      "epoch": 0.6089096925187636,
+      "grad_norm": 0.7048482298851013,
+      "learning_rate": 0.00013852343772446753,
+      "loss": 2.5656,
+      "step": 7545
+    },
+    {
+      "epoch": 0.6089903962553467,
+      "grad_norm": 0.7252251505851746,
+      "learning_rate": 0.00013850886879897135,
+      "loss": 2.6509,
+      "step": 7546
+    },
+    {
+      "epoch": 0.6090710999919297,
+      "grad_norm": 0.7220067381858826,
+      "learning_rate": 0.00013849429891373344,
+      "loss": 2.5558,
+      "step": 7547
+    },
+    {
+      "epoch": 0.6091518037285126,
+      "grad_norm": 0.7672600746154785,
+      "learning_rate": 0.000138479728069117,
+      "loss": 2.5682,
+      "step": 7548
+    },
+    {
+      "epoch": 0.6092325074650956,
+      "grad_norm": 0.7753601670265198,
+      "learning_rate": 0.0001384651562654852,
+      "loss": 2.6459,
+      "step": 7549
+    },
+    {
+      "epoch": 0.6093132112016786,
+      "grad_norm": 0.7346559166908264,
+      "learning_rate": 0.00013845058350320108,
+      "loss": 2.5988,
+      "step": 7550
+    },
+    {
+      "epoch": 0.6093939149382617,
+      "grad_norm": 0.7386072874069214,
+      "learning_rate": 0.00013843600978262797,
+      "loss": 2.6366,
+      "step": 7551
+    },
+    {
+      "epoch": 0.6094746186748446,
+      "grad_norm": 0.7114188075065613,
+      "learning_rate": 0.00013842143510412898,
+      "loss": 2.5515,
+      "step": 7552
+    },
+    {
+      "epoch": 0.6095553224114276,
+      "grad_norm": 0.6836373209953308,
+      "learning_rate": 0.00013840685946806742,
+      "loss": 2.6301,
+      "step": 7553
+    },
+    {
+      "epoch": 0.6096360261480106,
+      "grad_norm": 0.7548927068710327,
+      "learning_rate": 0.00013839228287480652,
+      "loss": 2.6508,
+      "step": 7554
+    },
+    {
+      "epoch": 0.6097167298845937,
+      "grad_norm": 0.6931679248809814,
+      "learning_rate": 0.00013837770532470957,
+      "loss": 2.5535,
+      "step": 7555
+    },
+    {
+      "epoch": 0.6097974336211767,
+      "grad_norm": 0.7621145248413086,
+      "learning_rate": 0.00013836312681813988,
+      "loss": 2.6831,
+      "step": 7556
+    },
+    {
+      "epoch": 0.6098781373577596,
+      "grad_norm": 0.6735427975654602,
+      "learning_rate": 0.00013834854735546079,
+      "loss": 2.5338,
+      "step": 7557
+    },
+    {
+      "epoch": 0.6099588410943426,
+      "grad_norm": 0.7157600522041321,
+      "learning_rate": 0.00013833396693703565,
+      "loss": 2.5713,
+      "step": 7558
+    },
+    {
+      "epoch": 0.6100395448309257,
+      "grad_norm": 0.718032956123352,
+      "learning_rate": 0.00013831938556322789,
+      "loss": 2.5625,
+      "step": 7559
+    },
+    {
+      "epoch": 0.6101202485675087,
+      "grad_norm": 0.7290309071540833,
+      "learning_rate": 0.0001383048032344008,
+      "loss": 2.5956,
+      "step": 7560
+    },
+    {
+      "epoch": 0.6102009523040917,
+      "grad_norm": 0.675470769405365,
+      "learning_rate": 0.00013829021995091792,
+      "loss": 2.6053,
+      "step": 7561
+    },
+    {
+      "epoch": 0.6102816560406746,
+      "grad_norm": 0.7348767518997192,
+      "learning_rate": 0.00013827563571314268,
+      "loss": 2.6174,
+      "step": 7562
+    },
+    {
+      "epoch": 0.6103623597772577,
+      "grad_norm": 0.64495849609375,
+      "learning_rate": 0.00013826105052143852,
+      "loss": 2.5923,
+      "step": 7563
+    },
+    {
+      "epoch": 0.6104430635138407,
+      "grad_norm": 0.7379264235496521,
+      "learning_rate": 0.000138246464376169,
+      "loss": 2.6438,
+      "step": 7564
+    },
+    {
+      "epoch": 0.6105237672504237,
+      "grad_norm": 0.7802134156227112,
+      "learning_rate": 0.00013823187727769756,
+      "loss": 2.5884,
+      "step": 7565
+    },
+    {
+      "epoch": 0.6106044709870067,
+      "grad_norm": 0.6907222867012024,
+      "learning_rate": 0.00013821728922638782,
+      "loss": 2.596,
+      "step": 7566
+    },
+    {
+      "epoch": 0.6106851747235897,
+      "grad_norm": 0.6924182176589966,
+      "learning_rate": 0.00013820270022260335,
+      "loss": 2.5631,
+      "step": 7567
+    },
+    {
+      "epoch": 0.6107658784601727,
+      "grad_norm": 0.729258120059967,
+      "learning_rate": 0.0001381881102667077,
+      "loss": 2.5761,
+      "step": 7568
+    },
+    {
+      "epoch": 0.6108465821967557,
+      "grad_norm": 0.7141425013542175,
+      "learning_rate": 0.00013817351935906455,
+      "loss": 2.6214,
+      "step": 7569
+    },
+    {
+      "epoch": 0.6109272859333387,
+      "grad_norm": 0.7564505338668823,
+      "learning_rate": 0.00013815892750003748,
+      "loss": 2.6338,
+      "step": 7570
+    },
+    {
+      "epoch": 0.6110079896699218,
+      "grad_norm": 0.674705982208252,
+      "learning_rate": 0.00013814433468999022,
+      "loss": 2.5604,
+      "step": 7571
+    },
+    {
+      "epoch": 0.6110886934065047,
+      "grad_norm": 0.6956657767295837,
+      "learning_rate": 0.00013812974092928642,
+      "loss": 2.5805,
+      "step": 7572
+    },
+    {
+      "epoch": 0.6111693971430877,
+      "grad_norm": 0.7393823862075806,
+      "learning_rate": 0.0001381151462182898,
+      "loss": 2.6312,
+      "step": 7573
+    },
+    {
+      "epoch": 0.6112501008796707,
+      "grad_norm": 0.7048184275627136,
+      "learning_rate": 0.00013810055055736407,
+      "loss": 2.5948,
+      "step": 7574
+    },
+    {
+      "epoch": 0.6113308046162538,
+      "grad_norm": 0.748798668384552,
+      "learning_rate": 0.0001380859539468731,
+      "loss": 2.5815,
+      "step": 7575
+    },
+    {
+      "epoch": 0.6114115083528368,
+      "grad_norm": 0.7146531343460083,
+      "learning_rate": 0.00013807135638718048,
+      "loss": 2.5803,
+      "step": 7576
+    },
+    {
+      "epoch": 0.6114922120894197,
+      "grad_norm": 0.6883770823478699,
+      "learning_rate": 0.00013805675787865025,
+      "loss": 2.6005,
+      "step": 7577
+    },
+    {
+      "epoch": 0.6115729158260027,
+      "grad_norm": 0.7808375358581543,
+      "learning_rate": 0.0001380421584216461,
+      "loss": 2.6539,
+      "step": 7578
+    },
+    {
+      "epoch": 0.6116536195625858,
+      "grad_norm": 0.6919417977333069,
+      "learning_rate": 0.00013802755801653192,
+      "loss": 2.5812,
+      "step": 7579
+    },
+    {
+      "epoch": 0.6117343232991688,
+      "grad_norm": 0.6651085615158081,
+      "learning_rate": 0.0001380129566636716,
+      "loss": 2.5952,
+      "step": 7580
+    },
+    {
+      "epoch": 0.6118150270357517,
+      "grad_norm": 0.7806586623191833,
+      "learning_rate": 0.00013799835436342897,
+      "loss": 2.6509,
+      "step": 7581
+    },
+    {
+      "epoch": 0.6118957307723347,
+      "grad_norm": 0.6522969007492065,
+      "learning_rate": 0.0001379837511161681,
+      "loss": 2.606,
+      "step": 7582
+    },
+    {
+      "epoch": 0.6119764345089178,
+      "grad_norm": 0.7566540837287903,
+      "learning_rate": 0.0001379691469222528,
+      "loss": 2.6625,
+      "step": 7583
+    },
+    {
+      "epoch": 0.6120571382455008,
+      "grad_norm": 0.7126421928405762,
+      "learning_rate": 0.00013795454178204715,
+      "loss": 2.6396,
+      "step": 7584
+    },
+    {
+      "epoch": 0.6121378419820838,
+      "grad_norm": 0.6534276008605957,
+      "learning_rate": 0.0001379399356959151,
+      "loss": 2.5841,
+      "step": 7585
+    },
+    {
+      "epoch": 0.6122185457186667,
+      "grad_norm": 0.7663385272026062,
+      "learning_rate": 0.00013792532866422065,
+      "loss": 2.6685,
+      "step": 7586
+    },
+    {
+      "epoch": 0.6122992494552498,
+      "grad_norm": 0.6971656084060669,
+      "learning_rate": 0.0001379107206873279,
+      "loss": 2.6036,
+      "step": 7587
+    },
+    {
+      "epoch": 0.6123799531918328,
+      "grad_norm": 0.6807122230529785,
+      "learning_rate": 0.00013789611176560088,
+      "loss": 2.6499,
+      "step": 7588
+    },
+    {
+      "epoch": 0.6124606569284158,
+      "grad_norm": 0.6712431311607361,
+      "learning_rate": 0.0001378815018994037,
+      "loss": 2.6725,
+      "step": 7589
+    },
+    {
+      "epoch": 0.6125413606649988,
+      "grad_norm": 0.6986604928970337,
+      "learning_rate": 0.00013786689108910045,
+      "loss": 2.6159,
+      "step": 7590
+    },
+    {
+      "epoch": 0.6126220644015818,
+      "grad_norm": 0.7004108428955078,
+      "learning_rate": 0.0001378522793350553,
+      "loss": 2.5743,
+      "step": 7591
+    },
+    {
+      "epoch": 0.6127027681381648,
+      "grad_norm": 0.6782098412513733,
+      "learning_rate": 0.00013783766663763239,
+      "loss": 2.5776,
+      "step": 7592
+    },
+    {
+      "epoch": 0.6127834718747478,
+      "grad_norm": 0.6697036027908325,
+      "learning_rate": 0.00013782305299719593,
+      "loss": 2.6195,
+      "step": 7593
+    },
+    {
+      "epoch": 0.6128641756113308,
+      "grad_norm": 0.6894395351409912,
+      "learning_rate": 0.00013780843841411014,
+      "loss": 2.662,
+      "step": 7594
+    },
+    {
+      "epoch": 0.6129448793479139,
+      "grad_norm": 0.6775636672973633,
+      "learning_rate": 0.00013779382288873918,
+      "loss": 2.6083,
+      "step": 7595
+    },
+    {
+      "epoch": 0.6130255830844968,
+      "grad_norm": 0.7143577337265015,
+      "learning_rate": 0.00013777920642144738,
+      "loss": 2.581,
+      "step": 7596
+    },
+    {
+      "epoch": 0.6131062868210798,
+      "grad_norm": 0.6143797636032104,
+      "learning_rate": 0.00013776458901259905,
+      "loss": 2.541,
+      "step": 7597
+    },
+    {
+      "epoch": 0.6131869905576628,
+      "grad_norm": 0.7003727555274963,
+      "learning_rate": 0.00013774997066255839,
+      "loss": 2.5748,
+      "step": 7598
+    },
+    {
+      "epoch": 0.6132676942942458,
+      "grad_norm": 0.6796504259109497,
+      "learning_rate": 0.0001377353513716898,
+      "loss": 2.596,
+      "step": 7599
+    },
+    {
+      "epoch": 0.6133483980308289,
+      "grad_norm": 0.7011274695396423,
+      "learning_rate": 0.00013772073114035762,
+      "loss": 2.5318,
+      "step": 7600
+    },
+    {
+      "epoch": 0.6134291017674118,
+      "grad_norm": 0.6584382057189941,
+      "learning_rate": 0.0001377061099689262,
+      "loss": 2.5793,
+      "step": 7601
+    },
+    {
+      "epoch": 0.6135098055039948,
+      "grad_norm": 0.6586211919784546,
+      "learning_rate": 0.00013769148785775995,
+      "loss": 2.5969,
+      "step": 7602
+    },
+    {
+      "epoch": 0.6135905092405778,
+      "grad_norm": 0.7187132835388184,
+      "learning_rate": 0.0001376768648072233,
+      "loss": 2.6407,
+      "step": 7603
+    },
+    {
+      "epoch": 0.6136712129771609,
+      "grad_norm": 0.7394679188728333,
+      "learning_rate": 0.00013766224081768072,
+      "loss": 2.5959,
+      "step": 7604
+    },
+    {
+      "epoch": 0.6137519167137439,
+      "grad_norm": 0.6802375912666321,
+      "learning_rate": 0.00013764761588949665,
+      "loss": 2.5956,
+      "step": 7605
+    },
+    {
+      "epoch": 0.6138326204503268,
+      "grad_norm": 0.6949049234390259,
+      "learning_rate": 0.00013763299002303553,
+      "loss": 2.556,
+      "step": 7606
+    },
+    {
+      "epoch": 0.6139133241869098,
+      "grad_norm": 0.7406589388847351,
+      "learning_rate": 0.00013761836321866196,
+      "loss": 2.5495,
+      "step": 7607
+    },
+    {
+      "epoch": 0.6139940279234929,
+      "grad_norm": 0.742499053478241,
+      "learning_rate": 0.0001376037354767404,
+      "loss": 2.589,
+      "step": 7608
+    },
+    {
+      "epoch": 0.6140747316600759,
+      "grad_norm": 0.7669157385826111,
+      "learning_rate": 0.00013758910679763551,
+      "loss": 2.576,
+      "step": 7609
+    },
+    {
+      "epoch": 0.6141554353966588,
+      "grad_norm": 0.6506752967834473,
+      "learning_rate": 0.00013757447718171182,
+      "loss": 2.5792,
+      "step": 7610
+    },
+    {
+      "epoch": 0.6142361391332418,
+      "grad_norm": 0.698514461517334,
+      "learning_rate": 0.00013755984662933393,
+      "loss": 2.5809,
+      "step": 7611
+    },
+    {
+      "epoch": 0.6143168428698249,
+      "grad_norm": 0.6541082262992859,
+      "learning_rate": 0.00013754521514086645,
+      "loss": 2.5755,
+      "step": 7612
+    },
+    {
+      "epoch": 0.6143975466064079,
+      "grad_norm": 0.6619362235069275,
+      "learning_rate": 0.0001375305827166741,
+      "loss": 2.5886,
+      "step": 7613
+    },
+    {
+      "epoch": 0.6144782503429909,
+      "grad_norm": 0.7205569744110107,
+      "learning_rate": 0.00013751594935712148,
+      "loss": 2.6293,
+      "step": 7614
+    },
+    {
+      "epoch": 0.6145589540795738,
+      "grad_norm": 0.7382494211196899,
+      "learning_rate": 0.00013750131506257339,
+      "loss": 2.6977,
+      "step": 7615
+    },
+    {
+      "epoch": 0.6146396578161569,
+      "grad_norm": 0.7492627501487732,
+      "learning_rate": 0.00013748667983339444,
+      "loss": 2.6492,
+      "step": 7616
+    },
+    {
+      "epoch": 0.6147203615527399,
+      "grad_norm": 0.6627328991889954,
+      "learning_rate": 0.00013747204366994947,
+      "loss": 2.5458,
+      "step": 7617
+    },
+    {
+      "epoch": 0.6148010652893229,
+      "grad_norm": 0.7039626836776733,
+      "learning_rate": 0.00013745740657260323,
+      "loss": 2.6578,
+      "step": 7618
+    },
+    {
+      "epoch": 0.6148817690259059,
+      "grad_norm": 0.6999295353889465,
+      "learning_rate": 0.00013744276854172046,
+      "loss": 2.6189,
+      "step": 7619
+    },
+    {
+      "epoch": 0.6149624727624889,
+      "grad_norm": 0.7604365348815918,
+      "learning_rate": 0.00013742812957766607,
+      "loss": 2.5344,
+      "step": 7620
+    },
+    {
+      "epoch": 0.6150431764990719,
+      "grad_norm": 0.6860831379890442,
+      "learning_rate": 0.0001374134896808048,
+      "loss": 2.6309,
+      "step": 7621
+    },
+    {
+      "epoch": 0.6151238802356549,
+      "grad_norm": 0.6628854274749756,
+      "learning_rate": 0.0001373988488515016,
+      "loss": 2.6339,
+      "step": 7622
+    },
+    {
+      "epoch": 0.6152045839722379,
+      "grad_norm": 0.7112562656402588,
+      "learning_rate": 0.00013738420709012134,
+      "loss": 2.6064,
+      "step": 7623
+    },
+    {
+      "epoch": 0.615285287708821,
+      "grad_norm": 0.7068392634391785,
+      "learning_rate": 0.0001373695643970289,
+      "loss": 2.624,
+      "step": 7624
+    },
+    {
+      "epoch": 0.6153659914454039,
+      "grad_norm": 0.6534786224365234,
+      "learning_rate": 0.00013735492077258924,
+      "loss": 2.5582,
+      "step": 7625
+    },
+    {
+      "epoch": 0.6154466951819869,
+      "grad_norm": 0.7433418035507202,
+      "learning_rate": 0.00013734027621716729,
+      "loss": 2.5803,
+      "step": 7626
+    },
+    {
+      "epoch": 0.6155273989185699,
+      "grad_norm": 0.7172532081604004,
+      "learning_rate": 0.00013732563073112804,
+      "loss": 2.5906,
+      "step": 7627
+    },
+    {
+      "epoch": 0.615608102655153,
+      "grad_norm": 0.6712297201156616,
+      "learning_rate": 0.00013731098431483653,
+      "loss": 2.5597,
+      "step": 7628
+    },
+    {
+      "epoch": 0.615688806391736,
+      "grad_norm": 0.7079061269760132,
+      "learning_rate": 0.00013729633696865775,
+      "loss": 2.5538,
+      "step": 7629
+    },
+    {
+      "epoch": 0.6157695101283189,
+      "grad_norm": 0.6968971490859985,
+      "learning_rate": 0.00013728168869295678,
+      "loss": 2.6429,
+      "step": 7630
+    },
+    {
+      "epoch": 0.6158502138649019,
+      "grad_norm": 0.7123236060142517,
+      "learning_rate": 0.00013726703948809864,
+      "loss": 2.5607,
+      "step": 7631
+    },
+    {
+      "epoch": 0.615930917601485,
+      "grad_norm": 0.6441208124160767,
+      "learning_rate": 0.00013725238935444843,
+      "loss": 2.6176,
+      "step": 7632
+    },
+    {
+      "epoch": 0.616011621338068,
+      "grad_norm": 0.7145917415618896,
+      "learning_rate": 0.00013723773829237137,
+      "loss": 2.5698,
+      "step": 7633
+    },
+    {
+      "epoch": 0.616092325074651,
+      "grad_norm": 0.6397334337234497,
+      "learning_rate": 0.00013722308630223252,
+      "loss": 2.596,
+      "step": 7634
+    },
+    {
+      "epoch": 0.6161730288112339,
+      "grad_norm": 0.6372843980789185,
+      "learning_rate": 0.00013720843338439702,
+      "loss": 2.5679,
+      "step": 7635
+    },
+    {
+      "epoch": 0.616253732547817,
+      "grad_norm": 0.707842230796814,
+      "learning_rate": 0.00013719377953923012,
+      "loss": 2.6296,
+      "step": 7636
+    },
+    {
+      "epoch": 0.6163344362844,
+      "grad_norm": 0.6629409193992615,
+      "learning_rate": 0.000137179124767097,
+      "loss": 2.542,
+      "step": 7637
+    },
+    {
+      "epoch": 0.616415140020983,
+      "grad_norm": 0.753646194934845,
+      "learning_rate": 0.00013716446906836288,
+      "loss": 2.5741,
+      "step": 7638
+    },
+    {
+      "epoch": 0.6164958437575659,
+      "grad_norm": 0.6409948468208313,
+      "learning_rate": 0.0001371498124433931,
+      "loss": 2.6723,
+      "step": 7639
+    },
+    {
+      "epoch": 0.616576547494149,
+      "grad_norm": 0.6489264965057373,
+      "learning_rate": 0.0001371351548925528,
+      "loss": 2.5806,
+      "step": 7640
+    },
+    {
+      "epoch": 0.616657251230732,
+      "grad_norm": 0.6857934594154358,
+      "learning_rate": 0.00013712049641620745,
+      "loss": 2.6406,
+      "step": 7641
+    },
+    {
+      "epoch": 0.616737954967315,
+      "grad_norm": 0.6754183769226074,
+      "learning_rate": 0.00013710583701472226,
+      "loss": 2.5576,
+      "step": 7642
+    },
+    {
+      "epoch": 0.616818658703898,
+      "grad_norm": 0.7083800435066223,
+      "learning_rate": 0.0001370911766884626,
+      "loss": 2.5747,
+      "step": 7643
+    },
+    {
+      "epoch": 0.616899362440481,
+      "grad_norm": 0.7281948924064636,
+      "learning_rate": 0.0001370765154377939,
+      "loss": 2.5627,
+      "step": 7644
+    },
+    {
+      "epoch": 0.616980066177064,
+      "grad_norm": 0.655414342880249,
+      "learning_rate": 0.00013706185326308148,
+      "loss": 2.5897,
+      "step": 7645
+    },
+    {
+      "epoch": 0.617060769913647,
+      "grad_norm": 0.6771859526634216,
+      "learning_rate": 0.0001370471901646908,
+      "loss": 2.5761,
+      "step": 7646
+    },
+    {
+      "epoch": 0.61714147365023,
+      "grad_norm": 0.6813557147979736,
+      "learning_rate": 0.00013703252614298732,
+      "loss": 2.5807,
+      "step": 7647
+    },
+    {
+      "epoch": 0.6172221773868131,
+      "grad_norm": 0.6948046684265137,
+      "learning_rate": 0.00013701786119833646,
+      "loss": 2.586,
+      "step": 7648
+    },
+    {
+      "epoch": 0.617302881123396,
+      "grad_norm": 0.643455982208252,
+      "learning_rate": 0.00013700319533110377,
+      "loss": 2.592,
+      "step": 7649
+    },
+    {
+      "epoch": 0.617383584859979,
+      "grad_norm": 0.7292457818984985,
+      "learning_rate": 0.0001369885285416547,
+      "loss": 2.6396,
+      "step": 7650
+    },
+    {
+      "epoch": 0.617464288596562,
+      "grad_norm": 0.642902672290802,
+      "learning_rate": 0.00013697386083035478,
+      "loss": 2.6115,
+      "step": 7651
+    },
+    {
+      "epoch": 0.617544992333145,
+      "grad_norm": 0.6536445021629333,
+      "learning_rate": 0.00013695919219756966,
+      "loss": 2.5406,
+      "step": 7652
+    },
+    {
+      "epoch": 0.6176256960697281,
+      "grad_norm": 0.6643723249435425,
+      "learning_rate": 0.0001369445226436648,
+      "loss": 2.6188,
+      "step": 7653
+    },
+    {
+      "epoch": 0.617706399806311,
+      "grad_norm": 0.6481621265411377,
+      "learning_rate": 0.00013692985216900592,
+      "loss": 2.5489,
+      "step": 7654
+    },
+    {
+      "epoch": 0.617787103542894,
+      "grad_norm": 0.6828036904335022,
+      "learning_rate": 0.00013691518077395856,
+      "loss": 2.5114,
+      "step": 7655
+    },
+    {
+      "epoch": 0.617867807279477,
+      "grad_norm": 0.6802895665168762,
+      "learning_rate": 0.00013690050845888838,
+      "loss": 2.5973,
+      "step": 7656
+    },
+    {
+      "epoch": 0.6179485110160601,
+      "grad_norm": 0.6980829238891602,
+      "learning_rate": 0.00013688583522416107,
+      "loss": 2.6032,
+      "step": 7657
+    },
+    {
+      "epoch": 0.618029214752643,
+      "grad_norm": 0.7157626748085022,
+      "learning_rate": 0.00013687116107014236,
+      "loss": 2.5552,
+      "step": 7658
+    },
+    {
+      "epoch": 0.618109918489226,
+      "grad_norm": 0.69700688123703,
+      "learning_rate": 0.00013685648599719792,
+      "loss": 2.5988,
+      "step": 7659
+    },
+    {
+      "epoch": 0.618190622225809,
+      "grad_norm": 0.6859539151191711,
+      "learning_rate": 0.0001368418100056935,
+      "loss": 2.6268,
+      "step": 7660
+    },
+    {
+      "epoch": 0.6182713259623921,
+      "grad_norm": 0.6812828183174133,
+      "learning_rate": 0.00013682713309599487,
+      "loss": 2.6002,
+      "step": 7661
+    },
+    {
+      "epoch": 0.6183520296989751,
+      "grad_norm": 0.6461766362190247,
+      "learning_rate": 0.00013681245526846783,
+      "loss": 2.6064,
+      "step": 7662
+    },
+    {
+      "epoch": 0.618432733435558,
+      "grad_norm": 0.7198306322097778,
+      "learning_rate": 0.00013679777652347814,
+      "loss": 2.6012,
+      "step": 7663
+    },
+    {
+      "epoch": 0.618513437172141,
+      "grad_norm": 0.7367191910743713,
+      "learning_rate": 0.00013678309686139168,
+      "loss": 2.6661,
+      "step": 7664
+    },
+    {
+      "epoch": 0.6185941409087241,
+      "grad_norm": 0.6975768804550171,
+      "learning_rate": 0.0001367684162825743,
+      "loss": 2.6394,
+      "step": 7665
+    },
+    {
+      "epoch": 0.6186748446453071,
+      "grad_norm": 0.7545140385627747,
+      "learning_rate": 0.0001367537347873919,
+      "loss": 2.624,
+      "step": 7666
+    },
+    {
+      "epoch": 0.6187555483818901,
+      "grad_norm": 0.6683520674705505,
+      "learning_rate": 0.0001367390523762103,
+      "loss": 2.6345,
+      "step": 7667
+    },
+    {
+      "epoch": 0.618836252118473,
+      "grad_norm": 0.6964975595474243,
+      "learning_rate": 0.00013672436904939552,
+      "loss": 2.591,
+      "step": 7668
+    },
+    {
+      "epoch": 0.6189169558550561,
+      "grad_norm": 0.7033975124359131,
+      "learning_rate": 0.00013670968480731344,
+      "loss": 2.566,
+      "step": 7669
+    },
+    {
+      "epoch": 0.6189976595916391,
+      "grad_norm": 0.706136167049408,
+      "learning_rate": 0.00013669499965033007,
+      "loss": 2.6073,
+      "step": 7670
+    },
+    {
+      "epoch": 0.6190783633282221,
+      "grad_norm": 0.7146300673484802,
+      "learning_rate": 0.0001366803135788114,
+      "loss": 2.6602,
+      "step": 7671
+    },
+    {
+      "epoch": 0.6191590670648051,
+      "grad_norm": 0.7603063583374023,
+      "learning_rate": 0.00013666562659312342,
+      "loss": 2.5286,
+      "step": 7672
+    },
+    {
+      "epoch": 0.6192397708013881,
+      "grad_norm": 0.744955837726593,
+      "learning_rate": 0.00013665093869363217,
+      "loss": 2.5678,
+      "step": 7673
+    },
+    {
+      "epoch": 0.6193204745379711,
+      "grad_norm": 0.7548620104789734,
+      "learning_rate": 0.00013663624988070373,
+      "loss": 2.6081,
+      "step": 7674
+    },
+    {
+      "epoch": 0.6194011782745541,
+      "grad_norm": 0.7367276549339294,
+      "learning_rate": 0.0001366215601547042,
+      "loss": 2.5559,
+      "step": 7675
+    },
+    {
+      "epoch": 0.6194818820111371,
+      "grad_norm": 0.7243839502334595,
+      "learning_rate": 0.00013660686951599962,
+      "loss": 2.5545,
+      "step": 7676
+    },
+    {
+      "epoch": 0.6195625857477202,
+      "grad_norm": 0.7595756649971008,
+      "learning_rate": 0.00013659217796495616,
+      "loss": 2.6547,
+      "step": 7677
+    },
+    {
+      "epoch": 0.6196432894843031,
+      "grad_norm": 0.7566717863082886,
+      "learning_rate": 0.00013657748550193998,
+      "loss": 2.6521,
+      "step": 7678
+    },
+    {
+      "epoch": 0.6197239932208861,
+      "grad_norm": 0.8441942930221558,
+      "learning_rate": 0.00013656279212731728,
+      "loss": 2.6325,
+      "step": 7679
+    },
+    {
+      "epoch": 0.6198046969574691,
+      "grad_norm": 0.7481170296669006,
+      "learning_rate": 0.00013654809784145418,
+      "loss": 2.6037,
+      "step": 7680
+    },
+    {
+      "epoch": 0.6198854006940522,
+      "grad_norm": 0.6626241207122803,
+      "learning_rate": 0.00013653340264471695,
+      "loss": 2.6028,
+      "step": 7681
+    },
+    {
+      "epoch": 0.6199661044306352,
+      "grad_norm": 0.7658020853996277,
+      "learning_rate": 0.00013651870653747186,
+      "loss": 2.5553,
+      "step": 7682
+    },
+    {
+      "epoch": 0.6200468081672181,
+      "grad_norm": 0.8218126893043518,
+      "learning_rate": 0.0001365040095200851,
+      "loss": 2.5661,
+      "step": 7683
+    },
+    {
+      "epoch": 0.6201275119038011,
+      "grad_norm": 0.6481068134307861,
+      "learning_rate": 0.00013648931159292304,
+      "loss": 2.5675,
+      "step": 7684
+    },
+    {
+      "epoch": 0.6202082156403842,
+      "grad_norm": 0.7529950141906738,
+      "learning_rate": 0.0001364746127563519,
+      "loss": 2.6137,
+      "step": 7685
+    },
+    {
+      "epoch": 0.6202889193769672,
+      "grad_norm": 0.7133232355117798,
+      "learning_rate": 0.00013645991301073816,
+      "loss": 2.6004,
+      "step": 7686
+    },
+    {
+      "epoch": 0.6203696231135502,
+      "grad_norm": 0.7809340953826904,
+      "learning_rate": 0.000136445212356448,
+      "loss": 2.6317,
+      "step": 7687
+    },
+    {
+      "epoch": 0.6204503268501331,
+      "grad_norm": 0.7106895446777344,
+      "learning_rate": 0.00013643051079384789,
+      "loss": 2.6086,
+      "step": 7688
+    },
+    {
+      "epoch": 0.6205310305867162,
+      "grad_norm": 0.6960744261741638,
+      "learning_rate": 0.00013641580832330423,
+      "loss": 2.5554,
+      "step": 7689
+    },
+    {
+      "epoch": 0.6206117343232992,
+      "grad_norm": 0.7078820466995239,
+      "learning_rate": 0.00013640110494518343,
+      "loss": 2.5902,
+      "step": 7690
+    },
+    {
+      "epoch": 0.6206924380598822,
+      "grad_norm": 0.7150746583938599,
+      "learning_rate": 0.00013638640065985195,
+      "loss": 2.5947,
+      "step": 7691
+    },
+    {
+      "epoch": 0.6207731417964651,
+      "grad_norm": 0.7507869601249695,
+      "learning_rate": 0.00013637169546767625,
+      "loss": 2.559,
+      "step": 7692
+    },
+    {
+      "epoch": 0.6208538455330482,
+      "grad_norm": 0.7453179359436035,
+      "learning_rate": 0.00013635698936902282,
+      "loss": 2.5612,
+      "step": 7693
+    },
+    {
+      "epoch": 0.6209345492696312,
+      "grad_norm": 0.7174177765846252,
+      "learning_rate": 0.00013634228236425816,
+      "loss": 2.6221,
+      "step": 7694
+    },
+    {
+      "epoch": 0.6210152530062142,
+      "grad_norm": 0.7394092679023743,
+      "learning_rate": 0.00013632757445374884,
+      "loss": 2.6045,
+      "step": 7695
+    },
+    {
+      "epoch": 0.6210959567427972,
+      "grad_norm": 0.7346367239952087,
+      "learning_rate": 0.0001363128656378614,
+      "loss": 2.677,
+      "step": 7696
+    },
+    {
+      "epoch": 0.6211766604793802,
+      "grad_norm": 0.6697696447372437,
+      "learning_rate": 0.00013629815591696245,
+      "loss": 2.5741,
+      "step": 7697
+    },
+    {
+      "epoch": 0.6212573642159632,
+      "grad_norm": 0.6993793845176697,
+      "learning_rate": 0.00013628344529141852,
+      "loss": 2.5206,
+      "step": 7698
+    },
+    {
+      "epoch": 0.6213380679525462,
+      "grad_norm": 0.6946697235107422,
+      "learning_rate": 0.00013626873376159631,
+      "loss": 2.6046,
+      "step": 7699
+    },
+    {
+      "epoch": 0.6214187716891292,
+      "grad_norm": 0.7641928195953369,
+      "learning_rate": 0.00013625402132786248,
+      "loss": 2.5459,
+      "step": 7700
+    },
+    {
+      "epoch": 0.6214994754257122,
+      "grad_norm": 0.6513504981994629,
+      "learning_rate": 0.00013623930799058363,
+      "loss": 2.6137,
+      "step": 7701
+    },
+    {
+      "epoch": 0.6215801791622952,
+      "grad_norm": 0.6745209097862244,
+      "learning_rate": 0.00013622459375012651,
+      "loss": 2.5285,
+      "step": 7702
+    },
+    {
+      "epoch": 0.6216608828988782,
+      "grad_norm": 0.7162348628044128,
+      "learning_rate": 0.0001362098786068578,
+      "loss": 2.6224,
+      "step": 7703
+    },
+    {
+      "epoch": 0.6217415866354612,
+      "grad_norm": 0.7387436032295227,
+      "learning_rate": 0.00013619516256114427,
+      "loss": 2.6216,
+      "step": 7704
+    },
+    {
+      "epoch": 0.6218222903720442,
+      "grad_norm": 0.764955461025238,
+      "learning_rate": 0.00013618044561335268,
+      "loss": 2.612,
+      "step": 7705
+    },
+    {
+      "epoch": 0.6219029941086273,
+      "grad_norm": 0.6492719054222107,
+      "learning_rate": 0.00013616572776384983,
+      "loss": 2.5532,
+      "step": 7706
+    },
+    {
+      "epoch": 0.6219836978452102,
+      "grad_norm": 0.6870293617248535,
+      "learning_rate": 0.0001361510090130025,
+      "loss": 2.5705,
+      "step": 7707
+    },
+    {
+      "epoch": 0.6220644015817932,
+      "grad_norm": 0.6899540424346924,
+      "learning_rate": 0.0001361362893611775,
+      "loss": 2.5768,
+      "step": 7708
+    },
+    {
+      "epoch": 0.6221451053183762,
+      "grad_norm": 0.658941924571991,
+      "learning_rate": 0.0001361215688087417,
+      "loss": 2.5664,
+      "step": 7709
+    },
+    {
+      "epoch": 0.6222258090549593,
+      "grad_norm": 0.6875531673431396,
+      "learning_rate": 0.000136106847356062,
+      "loss": 2.6128,
+      "step": 7710
+    },
+    {
+      "epoch": 0.6223065127915423,
+      "grad_norm": 0.657073974609375,
+      "learning_rate": 0.0001360921250035053,
+      "loss": 2.6449,
+      "step": 7711
+    },
+    {
+      "epoch": 0.6223872165281252,
+      "grad_norm": 0.7051201462745667,
+      "learning_rate": 0.00013607740175143848,
+      "loss": 2.5925,
+      "step": 7712
+    },
+    {
+      "epoch": 0.6224679202647082,
+      "grad_norm": 0.702877938747406,
+      "learning_rate": 0.0001360626776002285,
+      "loss": 2.5338,
+      "step": 7713
+    },
+    {
+      "epoch": 0.6225486240012913,
+      "grad_norm": 0.650935709476471,
+      "learning_rate": 0.00013604795255024233,
+      "loss": 2.5799,
+      "step": 7714
+    },
+    {
+      "epoch": 0.6226293277378743,
+      "grad_norm": 0.7035139203071594,
+      "learning_rate": 0.00013603322660184694,
+      "loss": 2.5476,
+      "step": 7715
+    },
+    {
+      "epoch": 0.6227100314744572,
+      "grad_norm": 0.6549977660179138,
+      "learning_rate": 0.0001360184997554094,
+      "loss": 2.6117,
+      "step": 7716
+    },
+    {
+      "epoch": 0.6227907352110402,
+      "grad_norm": 0.6882792115211487,
+      "learning_rate": 0.00013600377201129662,
+      "loss": 2.53,
+      "step": 7717
+    },
+    {
+      "epoch": 0.6228714389476233,
+      "grad_norm": 0.7390840649604797,
+      "learning_rate": 0.0001359890433698758,
+      "loss": 2.6345,
+      "step": 7718
+    },
+    {
+      "epoch": 0.6229521426842063,
+      "grad_norm": 0.7577612400054932,
+      "learning_rate": 0.00013597431383151386,
+      "loss": 2.6386,
+      "step": 7719
+    },
+    {
+      "epoch": 0.6230328464207893,
+      "grad_norm": 0.6818724870681763,
+      "learning_rate": 0.00013595958339657804,
+      "loss": 2.5806,
+      "step": 7720
+    },
+    {
+      "epoch": 0.6231135501573722,
+      "grad_norm": 0.6954349279403687,
+      "learning_rate": 0.0001359448520654354,
+      "loss": 2.5913,
+      "step": 7721
+    },
+    {
+      "epoch": 0.6231942538939553,
+      "grad_norm": 0.7976544499397278,
+      "learning_rate": 0.00013593011983845308,
+      "loss": 2.5686,
+      "step": 7722
+    },
+    {
+      "epoch": 0.6232749576305383,
+      "grad_norm": 0.7362754940986633,
+      "learning_rate": 0.00013591538671599824,
+      "loss": 2.5596,
+      "step": 7723
+    },
+    {
+      "epoch": 0.6233556613671213,
+      "grad_norm": 0.6842390298843384,
+      "learning_rate": 0.00013590065269843805,
+      "loss": 2.5793,
+      "step": 7724
+    },
+    {
+      "epoch": 0.6234363651037043,
+      "grad_norm": 0.6816275715827942,
+      "learning_rate": 0.0001358859177861398,
+      "loss": 2.5948,
+      "step": 7725
+    },
+    {
+      "epoch": 0.6235170688402873,
+      "grad_norm": 0.6892915964126587,
+      "learning_rate": 0.00013587118197947066,
+      "loss": 2.6287,
+      "step": 7726
+    },
+    {
+      "epoch": 0.6235977725768703,
+      "grad_norm": 0.6851752996444702,
+      "learning_rate": 0.00013585644527879792,
+      "loss": 2.5781,
+      "step": 7727
+    },
+    {
+      "epoch": 0.6236784763134533,
+      "grad_norm": 0.7022164463996887,
+      "learning_rate": 0.00013584170768448877,
+      "loss": 2.5856,
+      "step": 7728
+    },
+    {
+      "epoch": 0.6237591800500363,
+      "grad_norm": 0.6752299070358276,
+      "learning_rate": 0.0001358269691969106,
+      "loss": 2.6042,
+      "step": 7729
+    },
+    {
+      "epoch": 0.6238398837866194,
+      "grad_norm": 0.6861466765403748,
+      "learning_rate": 0.00013581222981643074,
+      "loss": 2.5887,
+      "step": 7730
+    },
+    {
+      "epoch": 0.6239205875232023,
+      "grad_norm": 0.7147940397262573,
+      "learning_rate": 0.00013579748954341647,
+      "loss": 2.5796,
+      "step": 7731
+    },
+    {
+      "epoch": 0.6240012912597853,
+      "grad_norm": 0.6704726219177246,
+      "learning_rate": 0.0001357827483782352,
+      "loss": 2.6027,
+      "step": 7732
+    },
+    {
+      "epoch": 0.6240819949963683,
+      "grad_norm": 0.6984317898750305,
+      "learning_rate": 0.0001357680063212543,
+      "loss": 2.635,
+      "step": 7733
+    },
+    {
+      "epoch": 0.6241626987329514,
+      "grad_norm": 0.6205787658691406,
+      "learning_rate": 0.00013575326337284115,
+      "loss": 2.5715,
+      "step": 7734
+    },
+    {
+      "epoch": 0.6242434024695344,
+      "grad_norm": 0.7214726805686951,
+      "learning_rate": 0.00013573851953336326,
+      "loss": 2.5605,
+      "step": 7735
+    },
+    {
+      "epoch": 0.6243241062061173,
+      "grad_norm": 0.6716169714927673,
+      "learning_rate": 0.000135723774803188,
+      "loss": 2.6766,
+      "step": 7736
+    },
+    {
+      "epoch": 0.6244048099427003,
+      "grad_norm": 0.6446832418441772,
+      "learning_rate": 0.00013570902918268293,
+      "loss": 2.5629,
+      "step": 7737
+    },
+    {
+      "epoch": 0.6244855136792834,
+      "grad_norm": 0.6721374988555908,
+      "learning_rate": 0.0001356942826722155,
+      "loss": 2.6093,
+      "step": 7738
+    },
+    {
+      "epoch": 0.6245662174158664,
+      "grad_norm": 0.7430365681648254,
+      "learning_rate": 0.0001356795352721532,
+      "loss": 2.5966,
+      "step": 7739
+    },
+    {
+      "epoch": 0.6246469211524494,
+      "grad_norm": 0.6787518858909607,
+      "learning_rate": 0.00013566478698286366,
+      "loss": 2.5519,
+      "step": 7740
+    },
+    {
+      "epoch": 0.6247276248890323,
+      "grad_norm": 0.6340047121047974,
+      "learning_rate": 0.0001356500378047144,
+      "loss": 2.5181,
+      "step": 7741
+    },
+    {
+      "epoch": 0.6248083286256154,
+      "grad_norm": 0.7559040188789368,
+      "learning_rate": 0.000135635287738073,
+      "loss": 2.6068,
+      "step": 7742
+    },
+    {
+      "epoch": 0.6248890323621984,
+      "grad_norm": 0.6819902062416077,
+      "learning_rate": 0.00013562053678330707,
+      "loss": 2.5754,
+      "step": 7743
+    },
+    {
+      "epoch": 0.6249697360987814,
+      "grad_norm": 0.6463500261306763,
+      "learning_rate": 0.00013560578494078423,
+      "loss": 2.5915,
+      "step": 7744
+    },
+    {
+      "epoch": 0.6250504398353643,
+      "grad_norm": 0.7510617971420288,
+      "learning_rate": 0.0001355910322108722,
+      "loss": 2.5738,
+      "step": 7745
+    },
+    {
+      "epoch": 0.6251311435719474,
+      "grad_norm": 0.75312739610672,
+      "learning_rate": 0.00013557627859393855,
+      "loss": 2.5938,
+      "step": 7746
+    },
+    {
+      "epoch": 0.6252118473085304,
+      "grad_norm": 0.7784396409988403,
+      "learning_rate": 0.0001355615240903511,
+      "loss": 2.6634,
+      "step": 7747
+    },
+    {
+      "epoch": 0.6252925510451134,
+      "grad_norm": 0.7174746990203857,
+      "learning_rate": 0.00013554676870047752,
+      "loss": 2.5973,
+      "step": 7748
+    },
+    {
+      "epoch": 0.6253732547816964,
+      "grad_norm": 0.6854952573776245,
+      "learning_rate": 0.0001355320124246855,
+      "loss": 2.5397,
+      "step": 7749
+    },
+    {
+      "epoch": 0.6254539585182795,
+      "grad_norm": 0.6584961414337158,
+      "learning_rate": 0.00013551725526334284,
+      "loss": 2.5574,
+      "step": 7750
+    },
+    {
+      "epoch": 0.6255346622548624,
+      "grad_norm": 0.7067389488220215,
+      "learning_rate": 0.00013550249721681738,
+      "loss": 2.5524,
+      "step": 7751
+    },
+    {
+      "epoch": 0.6256153659914454,
+      "grad_norm": 0.6923872232437134,
+      "learning_rate": 0.00013548773828547686,
+      "loss": 2.5651,
+      "step": 7752
+    },
+    {
+      "epoch": 0.6256960697280284,
+      "grad_norm": 0.6612355709075928,
+      "learning_rate": 0.00013547297846968915,
+      "loss": 2.6075,
+      "step": 7753
+    },
+    {
+      "epoch": 0.6257767734646114,
+      "grad_norm": 0.6762828826904297,
+      "learning_rate": 0.00013545821776982206,
+      "loss": 2.6136,
+      "step": 7754
+    },
+    {
+      "epoch": 0.6258574772011944,
+      "grad_norm": 0.6940783858299255,
+      "learning_rate": 0.0001354434561862435,
+      "loss": 2.5566,
+      "step": 7755
+    },
+    {
+      "epoch": 0.6259381809377774,
+      "grad_norm": 0.7874250411987305,
+      "learning_rate": 0.0001354286937193214,
+      "loss": 2.6732,
+      "step": 7756
+    },
+    {
+      "epoch": 0.6260188846743604,
+      "grad_norm": 0.6974111795425415,
+      "learning_rate": 0.0001354139303694236,
+      "loss": 2.5455,
+      "step": 7757
+    },
+    {
+      "epoch": 0.6260995884109434,
+      "grad_norm": 0.6710802316665649,
+      "learning_rate": 0.0001353991661369181,
+      "loss": 2.5608,
+      "step": 7758
+    },
+    {
+      "epoch": 0.6261802921475265,
+      "grad_norm": 0.681635320186615,
+      "learning_rate": 0.00013538440102217286,
+      "loss": 2.6107,
+      "step": 7759
+    },
+    {
+      "epoch": 0.6262609958841094,
+      "grad_norm": 0.7229577898979187,
+      "learning_rate": 0.0001353696350255558,
+      "loss": 2.5936,
+      "step": 7760
+    },
+    {
+      "epoch": 0.6263416996206924,
+      "grad_norm": 0.6909681558609009,
+      "learning_rate": 0.00013535486814743504,
+      "loss": 2.5521,
+      "step": 7761
+    },
+    {
+      "epoch": 0.6264224033572754,
+      "grad_norm": 0.7003746032714844,
+      "learning_rate": 0.0001353401003881785,
+      "loss": 2.5606,
+      "step": 7762
+    },
+    {
+      "epoch": 0.6265031070938585,
+      "grad_norm": 0.6883233785629272,
+      "learning_rate": 0.0001353253317481543,
+      "loss": 2.5971,
+      "step": 7763
+    },
+    {
+      "epoch": 0.6265838108304415,
+      "grad_norm": 0.7382355332374573,
+      "learning_rate": 0.0001353105622277305,
+      "loss": 2.5449,
+      "step": 7764
+    },
+    {
+      "epoch": 0.6266645145670244,
+      "grad_norm": 0.7090556621551514,
+      "learning_rate": 0.00013529579182727515,
+      "loss": 2.5988,
+      "step": 7765
+    },
+    {
+      "epoch": 0.6267452183036074,
+      "grad_norm": 0.6842581629753113,
+      "learning_rate": 0.00013528102054715643,
+      "loss": 2.6214,
+      "step": 7766
+    },
+    {
+      "epoch": 0.6268259220401905,
+      "grad_norm": 0.6969670653343201,
+      "learning_rate": 0.00013526624838774246,
+      "loss": 2.5443,
+      "step": 7767
+    },
+    {
+      "epoch": 0.6269066257767735,
+      "grad_norm": 0.7244827151298523,
+      "learning_rate": 0.00013525147534940138,
+      "loss": 2.5967,
+      "step": 7768
+    },
+    {
+      "epoch": 0.6269873295133565,
+      "grad_norm": 0.7022162675857544,
+      "learning_rate": 0.0001352367014325014,
+      "loss": 2.599,
+      "step": 7769
+    },
+    {
+      "epoch": 0.6270680332499394,
+      "grad_norm": 0.7065250873565674,
+      "learning_rate": 0.00013522192663741067,
+      "loss": 2.6105,
+      "step": 7770
+    },
+    {
+      "epoch": 0.6271487369865225,
+      "grad_norm": 0.6690711975097656,
+      "learning_rate": 0.0001352071509644975,
+      "loss": 2.55,
+      "step": 7771
+    },
+    {
+      "epoch": 0.6272294407231055,
+      "grad_norm": 0.6405982971191406,
+      "learning_rate": 0.00013519237441413011,
+      "loss": 2.6078,
+      "step": 7772
+    },
+    {
+      "epoch": 0.6273101444596885,
+      "grad_norm": 0.7340127229690552,
+      "learning_rate": 0.00013517759698667672,
+      "loss": 2.6244,
+      "step": 7773
+    },
+    {
+      "epoch": 0.6273908481962714,
+      "grad_norm": 0.6609435677528381,
+      "learning_rate": 0.00013516281868250566,
+      "loss": 2.5746,
+      "step": 7774
+    },
+    {
+      "epoch": 0.6274715519328545,
+      "grad_norm": 0.6681997179985046,
+      "learning_rate": 0.00013514803950198523,
+      "loss": 2.6181,
+      "step": 7775
+    },
+    {
+      "epoch": 0.6275522556694375,
+      "grad_norm": 0.7120032906532288,
+      "learning_rate": 0.0001351332594454838,
+      "loss": 2.6018,
+      "step": 7776
+    },
+    {
+      "epoch": 0.6276329594060205,
+      "grad_norm": 0.6618601679801941,
+      "learning_rate": 0.0001351184785133697,
+      "loss": 2.5342,
+      "step": 7777
+    },
+    {
+      "epoch": 0.6277136631426035,
+      "grad_norm": 0.7250192165374756,
+      "learning_rate": 0.00013510369670601132,
+      "loss": 2.5795,
+      "step": 7778
+    },
+    {
+      "epoch": 0.6277943668791865,
+      "grad_norm": 0.7918543219566345,
+      "learning_rate": 0.00013508891402377708,
+      "loss": 2.6544,
+      "step": 7779
+    },
+    {
+      "epoch": 0.6278750706157695,
+      "grad_norm": 0.678895890712738,
+      "learning_rate": 0.00013507413046703534,
+      "loss": 2.5937,
+      "step": 7780
+    },
+    {
+      "epoch": 0.6279557743523525,
+      "grad_norm": 0.7336576581001282,
+      "learning_rate": 0.00013505934603615457,
+      "loss": 2.598,
+      "step": 7781
+    },
+    {
+      "epoch": 0.6280364780889355,
+      "grad_norm": 0.6891419291496277,
+      "learning_rate": 0.00013504456073150332,
+      "loss": 2.5063,
+      "step": 7782
+    },
+    {
+      "epoch": 0.6281171818255186,
+      "grad_norm": 0.7949386835098267,
+      "learning_rate": 0.00013502977455344997,
+      "loss": 2.5703,
+      "step": 7783
+    },
+    {
+      "epoch": 0.6281978855621015,
+      "grad_norm": 0.7917985320091248,
+      "learning_rate": 0.00013501498750236306,
+      "loss": 2.639,
+      "step": 7784
+    },
+    {
+      "epoch": 0.6282785892986845,
+      "grad_norm": 0.7387086749076843,
+      "learning_rate": 0.00013500019957861113,
+      "loss": 2.5864,
+      "step": 7785
+    },
+    {
+      "epoch": 0.6283592930352675,
+      "grad_norm": 0.7189435958862305,
+      "learning_rate": 0.00013498541078256273,
+      "loss": 2.5627,
+      "step": 7786
+    },
+    {
+      "epoch": 0.6284399967718506,
+      "grad_norm": 0.6709900498390198,
+      "learning_rate": 0.00013497062111458646,
+      "loss": 2.5973,
+      "step": 7787
+    },
+    {
+      "epoch": 0.6285207005084336,
+      "grad_norm": 0.6925386190414429,
+      "learning_rate": 0.0001349558305750509,
+      "loss": 2.615,
+      "step": 7788
+    },
+    {
+      "epoch": 0.6286014042450165,
+      "grad_norm": 0.7191932201385498,
+      "learning_rate": 0.00013494103916432466,
+      "loss": 2.576,
+      "step": 7789
+    },
+    {
+      "epoch": 0.6286821079815995,
+      "grad_norm": 0.6798804402351379,
+      "learning_rate": 0.00013492624688277638,
+      "loss": 2.5661,
+      "step": 7790
+    },
+    {
+      "epoch": 0.6287628117181826,
+      "grad_norm": 0.6514562964439392,
+      "learning_rate": 0.00013491145373077475,
+      "loss": 2.6135,
+      "step": 7791
+    },
+    {
+      "epoch": 0.6288435154547656,
+      "grad_norm": 0.7345223426818848,
+      "learning_rate": 0.00013489665970868838,
+      "loss": 2.6015,
+      "step": 7792
+    },
+    {
+      "epoch": 0.6289242191913486,
+      "grad_norm": 0.7102675437927246,
+      "learning_rate": 0.0001348818648168861,
+      "loss": 2.5545,
+      "step": 7793
+    },
+    {
+      "epoch": 0.6290049229279315,
+      "grad_norm": 0.7151654362678528,
+      "learning_rate": 0.0001348670690557365,
+      "loss": 2.6464,
+      "step": 7794
+    },
+    {
+      "epoch": 0.6290856266645146,
+      "grad_norm": 0.7344057559967041,
+      "learning_rate": 0.00013485227242560844,
+      "loss": 2.6777,
+      "step": 7795
+    },
+    {
+      "epoch": 0.6291663304010976,
+      "grad_norm": 0.6622766852378845,
+      "learning_rate": 0.00013483747492687065,
+      "loss": 2.5713,
+      "step": 7796
+    },
+    {
+      "epoch": 0.6292470341376806,
+      "grad_norm": 0.6899346709251404,
+      "learning_rate": 0.0001348226765598919,
+      "loss": 2.5188,
+      "step": 7797
+    },
+    {
+      "epoch": 0.6293277378742635,
+      "grad_norm": 0.6711421012878418,
+      "learning_rate": 0.000134807877325041,
+      "loss": 2.5603,
+      "step": 7798
+    },
+    {
+      "epoch": 0.6294084416108466,
+      "grad_norm": 0.6973204016685486,
+      "learning_rate": 0.00013479307722268687,
+      "loss": 2.6621,
+      "step": 7799
+    },
+    {
+      "epoch": 0.6294891453474296,
+      "grad_norm": 0.7782350778579712,
+      "learning_rate": 0.00013477827625319824,
+      "loss": 2.5929,
+      "step": 7800
+    },
+    {
+      "epoch": 0.6295698490840126,
+      "grad_norm": 0.8703733682632446,
+      "learning_rate": 0.0001347634744169441,
+      "loss": 2.6884,
+      "step": 7801
+    },
+    {
+      "epoch": 0.6296505528205956,
+      "grad_norm": 0.7196036577224731,
+      "learning_rate": 0.00013474867171429326,
+      "loss": 2.6002,
+      "step": 7802
+    },
+    {
+      "epoch": 0.6297312565571785,
+      "grad_norm": 0.7224054932594299,
+      "learning_rate": 0.00013473386814561475,
+      "loss": 2.6007,
+      "step": 7803
+    },
+    {
+      "epoch": 0.6298119602937616,
+      "grad_norm": 0.7615752816200256,
+      "learning_rate": 0.00013471906371127743,
+      "loss": 2.6459,
+      "step": 7804
+    },
+    {
+      "epoch": 0.6298926640303446,
+      "grad_norm": 0.7189914584159851,
+      "learning_rate": 0.00013470425841165024,
+      "loss": 2.5692,
+      "step": 7805
+    },
+    {
+      "epoch": 0.6299733677669276,
+      "grad_norm": 0.7101845741271973,
+      "learning_rate": 0.00013468945224710225,
+      "loss": 2.5776,
+      "step": 7806
+    },
+    {
+      "epoch": 0.6300540715035106,
+      "grad_norm": 0.6860305666923523,
+      "learning_rate": 0.00013467464521800244,
+      "loss": 2.5567,
+      "step": 7807
+    },
+    {
+      "epoch": 0.6301347752400936,
+      "grad_norm": 0.7003797292709351,
+      "learning_rate": 0.0001346598373247198,
+      "loss": 2.6444,
+      "step": 7808
+    },
+    {
+      "epoch": 0.6302154789766766,
+      "grad_norm": 0.6341832876205444,
+      "learning_rate": 0.00013464502856762344,
+      "loss": 2.5475,
+      "step": 7809
+    },
+    {
+      "epoch": 0.6302961827132596,
+      "grad_norm": 0.6255922317504883,
+      "learning_rate": 0.00013463021894708242,
+      "loss": 2.5875,
+      "step": 7810
+    },
+    {
+      "epoch": 0.6303768864498426,
+      "grad_norm": 0.7136420607566833,
+      "learning_rate": 0.00013461540846346575,
+      "loss": 2.5708,
+      "step": 7811
+    },
+    {
+      "epoch": 0.6304575901864257,
+      "grad_norm": 0.7164542078971863,
+      "learning_rate": 0.00013460059711714267,
+      "loss": 2.4975,
+      "step": 7812
+    },
+    {
+      "epoch": 0.6305382939230086,
+      "grad_norm": 0.7667872905731201,
+      "learning_rate": 0.00013458578490848226,
+      "loss": 2.6124,
+      "step": 7813
+    },
+    {
+      "epoch": 0.6306189976595916,
+      "grad_norm": 0.6631812453269958,
+      "learning_rate": 0.0001345709718378537,
+      "loss": 2.5318,
+      "step": 7814
+    },
+    {
+      "epoch": 0.6306997013961746,
+      "grad_norm": 0.696864664554596,
+      "learning_rate": 0.0001345561579056261,
+      "loss": 2.6171,
+      "step": 7815
+    },
+    {
+      "epoch": 0.6307804051327577,
+      "grad_norm": 0.7368598580360413,
+      "learning_rate": 0.00013454134311216873,
+      "loss": 2.5734,
+      "step": 7816
+    },
+    {
+      "epoch": 0.6308611088693407,
+      "grad_norm": 0.7279712557792664,
+      "learning_rate": 0.00013452652745785083,
+      "loss": 2.6231,
+      "step": 7817
+    },
+    {
+      "epoch": 0.6309418126059236,
+      "grad_norm": 0.8070993423461914,
+      "learning_rate": 0.00013451171094304158,
+      "loss": 2.5486,
+      "step": 7818
+    },
+    {
+      "epoch": 0.6310225163425066,
+      "grad_norm": 0.7522621750831604,
+      "learning_rate": 0.0001344968935681103,
+      "loss": 2.5576,
+      "step": 7819
+    },
+    {
+      "epoch": 0.6311032200790897,
+      "grad_norm": 0.8185423612594604,
+      "learning_rate": 0.00013448207533342624,
+      "loss": 2.6068,
+      "step": 7820
+    },
+    {
+      "epoch": 0.6311839238156727,
+      "grad_norm": 0.7542584538459778,
+      "learning_rate": 0.0001344672562393587,
+      "loss": 2.643,
+      "step": 7821
+    },
+    {
+      "epoch": 0.6312646275522557,
+      "grad_norm": 0.7892276644706726,
+      "learning_rate": 0.00013445243628627712,
+      "loss": 2.6211,
+      "step": 7822
+    },
+    {
+      "epoch": 0.6313453312888386,
+      "grad_norm": 0.7216602563858032,
+      "learning_rate": 0.00013443761547455072,
+      "loss": 2.5725,
+      "step": 7823
+    },
+    {
+      "epoch": 0.6314260350254217,
+      "grad_norm": 0.6750743985176086,
+      "learning_rate": 0.0001344227938045489,
+      "loss": 2.5319,
+      "step": 7824
+    },
+    {
+      "epoch": 0.6315067387620047,
+      "grad_norm": 0.6711540222167969,
+      "learning_rate": 0.0001344079712766411,
+      "loss": 2.5957,
+      "step": 7825
+    },
+    {
+      "epoch": 0.6315874424985877,
+      "grad_norm": 0.6923524737358093,
+      "learning_rate": 0.00013439314789119667,
+      "loss": 2.6084,
+      "step": 7826
+    },
+    {
+      "epoch": 0.6316681462351706,
+      "grad_norm": 0.6859166026115417,
+      "learning_rate": 0.00013437832364858517,
+      "loss": 2.5608,
+      "step": 7827
+    },
+    {
+      "epoch": 0.6317488499717537,
+      "grad_norm": 0.7340966463088989,
+      "learning_rate": 0.0001343634985491759,
+      "loss": 2.531,
+      "step": 7828
+    },
+    {
+      "epoch": 0.6318295537083367,
+      "grad_norm": 0.7374520301818848,
+      "learning_rate": 0.00013434867259333848,
+      "loss": 2.5972,
+      "step": 7829
+    },
+    {
+      "epoch": 0.6319102574449197,
+      "grad_norm": 0.7252814769744873,
+      "learning_rate": 0.00013433384578144232,
+      "loss": 2.5874,
+      "step": 7830
+    },
+    {
+      "epoch": 0.6319909611815027,
+      "grad_norm": 0.7000489830970764,
+      "learning_rate": 0.000134319018113857,
+      "loss": 2.6137,
+      "step": 7831
+    },
+    {
+      "epoch": 0.6320716649180858,
+      "grad_norm": 0.805981457233429,
+      "learning_rate": 0.00013430418959095198,
+      "loss": 2.5581,
+      "step": 7832
+    },
+    {
+      "epoch": 0.6321523686546687,
+      "grad_norm": 0.7459721565246582,
+      "learning_rate": 0.00013428936021309693,
+      "loss": 2.5284,
+      "step": 7833
+    },
+    {
+      "epoch": 0.6322330723912517,
+      "grad_norm": 0.749794065952301,
+      "learning_rate": 0.00013427452998066136,
+      "loss": 2.5927,
+      "step": 7834
+    },
+    {
+      "epoch": 0.6323137761278347,
+      "grad_norm": 0.6925346255302429,
+      "learning_rate": 0.00013425969889401494,
+      "loss": 2.5703,
+      "step": 7835
+    },
+    {
+      "epoch": 0.6323944798644178,
+      "grad_norm": 0.6647117137908936,
+      "learning_rate": 0.00013424486695352728,
+      "loss": 2.5649,
+      "step": 7836
+    },
+    {
+      "epoch": 0.6324751836010007,
+      "grad_norm": 0.7358147501945496,
+      "learning_rate": 0.00013423003415956796,
+      "loss": 2.6122,
+      "step": 7837
+    },
+    {
+      "epoch": 0.6325558873375837,
+      "grad_norm": 0.7798088788986206,
+      "learning_rate": 0.00013421520051250675,
+      "loss": 2.5805,
+      "step": 7838
+    },
+    {
+      "epoch": 0.6326365910741667,
+      "grad_norm": 0.7108271718025208,
+      "learning_rate": 0.00013420036601271334,
+      "loss": 2.5457,
+      "step": 7839
+    },
+    {
+      "epoch": 0.6327172948107498,
+      "grad_norm": 0.7108528017997742,
+      "learning_rate": 0.00013418553066055734,
+      "loss": 2.6313,
+      "step": 7840
+    },
+    {
+      "epoch": 0.6327979985473328,
+      "grad_norm": 0.7325249910354614,
+      "learning_rate": 0.00013417069445640858,
+      "loss": 2.5598,
+      "step": 7841
+    },
+    {
+      "epoch": 0.6328787022839157,
+      "grad_norm": 0.6861844062805176,
+      "learning_rate": 0.0001341558574006368,
+      "loss": 2.5899,
+      "step": 7842
+    },
+    {
+      "epoch": 0.6329594060204987,
+      "grad_norm": 0.7576130628585815,
+      "learning_rate": 0.00013414101949361175,
+      "loss": 2.6077,
+      "step": 7843
+    },
+    {
+      "epoch": 0.6330401097570818,
+      "grad_norm": 0.7756128907203674,
+      "learning_rate": 0.0001341261807357033,
+      "loss": 2.6111,
+      "step": 7844
+    },
+    {
+      "epoch": 0.6331208134936648,
+      "grad_norm": 0.7131127715110779,
+      "learning_rate": 0.00013411134112728114,
+      "loss": 2.5227,
+      "step": 7845
+    },
+    {
+      "epoch": 0.6332015172302478,
+      "grad_norm": 0.6517898440361023,
+      "learning_rate": 0.00013409650066871525,
+      "loss": 2.5825,
+      "step": 7846
+    },
+    {
+      "epoch": 0.6332822209668307,
+      "grad_norm": 0.8452722430229187,
+      "learning_rate": 0.0001340816593603754,
+      "loss": 2.6037,
+      "step": 7847
+    },
+    {
+      "epoch": 0.6333629247034138,
+      "grad_norm": 0.7421110272407532,
+      "learning_rate": 0.00013406681720263153,
+      "loss": 2.5684,
+      "step": 7848
+    },
+    {
+      "epoch": 0.6334436284399968,
+      "grad_norm": 0.695139467716217,
+      "learning_rate": 0.0001340519741958535,
+      "loss": 2.5648,
+      "step": 7849
+    },
+    {
+      "epoch": 0.6335243321765798,
+      "grad_norm": 0.7780016660690308,
+      "learning_rate": 0.0001340371303404113,
+      "loss": 2.6849,
+      "step": 7850
+    },
+    {
+      "epoch": 0.6336050359131628,
+      "grad_norm": 0.7276864051818848,
+      "learning_rate": 0.00013402228563667482,
+      "loss": 2.6198,
+      "step": 7851
+    },
+    {
+      "epoch": 0.6336857396497458,
+      "grad_norm": 0.7566827535629272,
+      "learning_rate": 0.00013400744008501404,
+      "loss": 2.5803,
+      "step": 7852
+    },
+    {
+      "epoch": 0.6337664433863288,
+      "grad_norm": 0.7933458089828491,
+      "learning_rate": 0.00013399259368579894,
+      "loss": 2.6029,
+      "step": 7853
+    },
+    {
+      "epoch": 0.6338471471229118,
+      "grad_norm": 0.6849822402000427,
+      "learning_rate": 0.00013397774643939957,
+      "loss": 2.5454,
+      "step": 7854
+    },
+    {
+      "epoch": 0.6339278508594948,
+      "grad_norm": 0.7054651379585266,
+      "learning_rate": 0.00013396289834618594,
+      "loss": 2.5905,
+      "step": 7855
+    },
+    {
+      "epoch": 0.6340085545960777,
+      "grad_norm": 0.7036863565444946,
+      "learning_rate": 0.00013394804940652813,
+      "loss": 2.6342,
+      "step": 7856
+    },
+    {
+      "epoch": 0.6340892583326608,
+      "grad_norm": 0.7101735472679138,
+      "learning_rate": 0.00013393319962079614,
+      "loss": 2.6402,
+      "step": 7857
+    },
+    {
+      "epoch": 0.6341699620692438,
+      "grad_norm": 0.7053956389427185,
+      "learning_rate": 0.0001339183489893601,
+      "loss": 2.5841,
+      "step": 7858
+    },
+    {
+      "epoch": 0.6342506658058268,
+      "grad_norm": 0.7734887003898621,
+      "learning_rate": 0.0001339034975125902,
+      "loss": 2.652,
+      "step": 7859
+    },
+    {
+      "epoch": 0.6343313695424098,
+      "grad_norm": 0.6714119911193848,
+      "learning_rate": 0.0001338886451908565,
+      "loss": 2.5927,
+      "step": 7860
+    },
+    {
+      "epoch": 0.6344120732789928,
+      "grad_norm": 0.6580910682678223,
+      "learning_rate": 0.00013387379202452917,
+      "loss": 2.6114,
+      "step": 7861
+    },
+    {
+      "epoch": 0.6344927770155758,
+      "grad_norm": 0.6810200214385986,
+      "learning_rate": 0.00013385893801397836,
+      "loss": 2.5616,
+      "step": 7862
+    },
+    {
+      "epoch": 0.6345734807521588,
+      "grad_norm": 0.6989572048187256,
+      "learning_rate": 0.00013384408315957432,
+      "loss": 2.5954,
+      "step": 7863
+    },
+    {
+      "epoch": 0.6346541844887418,
+      "grad_norm": 0.7033671736717224,
+      "learning_rate": 0.00013382922746168728,
+      "loss": 2.6015,
+      "step": 7864
+    },
+    {
+      "epoch": 0.6347348882253249,
+      "grad_norm": 0.6873033046722412,
+      "learning_rate": 0.0001338143709206875,
+      "loss": 2.562,
+      "step": 7865
+    },
+    {
+      "epoch": 0.6348155919619078,
+      "grad_norm": 0.7361463904380798,
+      "learning_rate": 0.00013379951353694513,
+      "loss": 2.6175,
+      "step": 7866
+    },
+    {
+      "epoch": 0.6348962956984908,
+      "grad_norm": 0.7623226046562195,
+      "learning_rate": 0.00013378465531083055,
+      "loss": 2.7342,
+      "step": 7867
+    },
+    {
+      "epoch": 0.6349769994350738,
+      "grad_norm": 0.7427035570144653,
+      "learning_rate": 0.0001337697962427141,
+      "loss": 2.5468,
+      "step": 7868
+    },
+    {
+      "epoch": 0.6350577031716569,
+      "grad_norm": 0.6865772008895874,
+      "learning_rate": 0.00013375493633296598,
+      "loss": 2.6112,
+      "step": 7869
+    },
+    {
+      "epoch": 0.6351384069082399,
+      "grad_norm": 0.663567304611206,
+      "learning_rate": 0.00013374007558195666,
+      "loss": 2.5896,
+      "step": 7870
+    },
+    {
+      "epoch": 0.6352191106448228,
+      "grad_norm": 0.6804360151290894,
+      "learning_rate": 0.00013372521399005643,
+      "loss": 2.58,
+      "step": 7871
+    },
+    {
+      "epoch": 0.6352998143814058,
+      "grad_norm": 0.6755216121673584,
+      "learning_rate": 0.0001337103515576357,
+      "loss": 2.5593,
+      "step": 7872
+    },
+    {
+      "epoch": 0.6353805181179889,
+      "grad_norm": 0.8148807883262634,
+      "learning_rate": 0.00013369548828506491,
+      "loss": 2.6473,
+      "step": 7873
+    },
+    {
+      "epoch": 0.6354612218545719,
+      "grad_norm": 0.713009774684906,
+      "learning_rate": 0.00013368062417271447,
+      "loss": 2.6002,
+      "step": 7874
+    },
+    {
+      "epoch": 0.6355419255911549,
+      "grad_norm": 0.6390172839164734,
+      "learning_rate": 0.00013366575922095484,
+      "loss": 2.5794,
+      "step": 7875
+    },
+    {
+      "epoch": 0.6356226293277378,
+      "grad_norm": 0.7228195667266846,
+      "learning_rate": 0.00013365089343015649,
+      "loss": 2.6051,
+      "step": 7876
+    },
+    {
+      "epoch": 0.6357033330643209,
+      "grad_norm": 0.7563474178314209,
+      "learning_rate": 0.00013363602680068986,
+      "loss": 2.6308,
+      "step": 7877
+    },
+    {
+      "epoch": 0.6357840368009039,
+      "grad_norm": 0.7366798520088196,
+      "learning_rate": 0.00013362115933292557,
+      "loss": 2.5589,
+      "step": 7878
+    },
+    {
+      "epoch": 0.6358647405374869,
+      "grad_norm": 0.7137070894241333,
+      "learning_rate": 0.00013360629102723409,
+      "loss": 2.6428,
+      "step": 7879
+    },
+    {
+      "epoch": 0.6359454442740698,
+      "grad_norm": 0.6799132823944092,
+      "learning_rate": 0.000133591421883986,
+      "loss": 2.5549,
+      "step": 7880
+    },
+    {
+      "epoch": 0.6360261480106529,
+      "grad_norm": 0.7031344771385193,
+      "learning_rate": 0.00013357655190355188,
+      "loss": 2.6298,
+      "step": 7881
+    },
+    {
+      "epoch": 0.6361068517472359,
+      "grad_norm": 0.7441670298576355,
+      "learning_rate": 0.00013356168108630227,
+      "loss": 2.5844,
+      "step": 7882
+    },
+    {
+      "epoch": 0.6361875554838189,
+      "grad_norm": 0.7281978726387024,
+      "learning_rate": 0.00013354680943260784,
+      "loss": 2.5773,
+      "step": 7883
+    },
+    {
+      "epoch": 0.6362682592204019,
+      "grad_norm": 0.6969650983810425,
+      "learning_rate": 0.00013353193694283928,
+      "loss": 2.6156,
+      "step": 7884
+    },
+    {
+      "epoch": 0.636348962956985,
+      "grad_norm": 0.6668435335159302,
+      "learning_rate": 0.00013351706361736714,
+      "loss": 2.6328,
+      "step": 7885
+    },
+    {
+      "epoch": 0.6364296666935679,
+      "grad_norm": 0.6909573078155518,
+      "learning_rate": 0.0001335021894565622,
+      "loss": 2.5772,
+      "step": 7886
+    },
+    {
+      "epoch": 0.6365103704301509,
+      "grad_norm": 0.6740022897720337,
+      "learning_rate": 0.0001334873144607951,
+      "loss": 2.6435,
+      "step": 7887
+    },
+    {
+      "epoch": 0.6365910741667339,
+      "grad_norm": 0.7203185558319092,
+      "learning_rate": 0.0001334724386304366,
+      "loss": 2.5401,
+      "step": 7888
+    },
+    {
+      "epoch": 0.636671777903317,
+      "grad_norm": 0.7343020439147949,
+      "learning_rate": 0.0001334575619658574,
+      "loss": 2.5811,
+      "step": 7889
+    },
+    {
+      "epoch": 0.6367524816399,
+      "grad_norm": 0.6941348314285278,
+      "learning_rate": 0.00013344268446742835,
+      "loss": 2.6267,
+      "step": 7890
+    },
+    {
+      "epoch": 0.6368331853764829,
+      "grad_norm": 0.6983792185783386,
+      "learning_rate": 0.00013342780613552016,
+      "loss": 2.533,
+      "step": 7891
+    },
+    {
+      "epoch": 0.6369138891130659,
+      "grad_norm": 0.7093533277511597,
+      "learning_rate": 0.00013341292697050365,
+      "loss": 2.6616,
+      "step": 7892
+    },
+    {
+      "epoch": 0.636994592849649,
+      "grad_norm": 0.7377648949623108,
+      "learning_rate": 0.00013339804697274965,
+      "loss": 2.6032,
+      "step": 7893
+    },
+    {
+      "epoch": 0.637075296586232,
+      "grad_norm": 0.6669821739196777,
+      "learning_rate": 0.00013338316614262903,
+      "loss": 2.6082,
+      "step": 7894
+    },
+    {
+      "epoch": 0.6371560003228149,
+      "grad_norm": 0.6665576100349426,
+      "learning_rate": 0.00013336828448051263,
+      "loss": 2.6114,
+      "step": 7895
+    },
+    {
+      "epoch": 0.6372367040593979,
+      "grad_norm": 0.6893584132194519,
+      "learning_rate": 0.0001333534019867714,
+      "loss": 2.5886,
+      "step": 7896
+    },
+    {
+      "epoch": 0.637317407795981,
+      "grad_norm": 0.7651494741439819,
+      "learning_rate": 0.00013333851866177617,
+      "loss": 2.5622,
+      "step": 7897
+    },
+    {
+      "epoch": 0.637398111532564,
+      "grad_norm": 0.8124055862426758,
+      "learning_rate": 0.00013332363450589788,
+      "loss": 2.6036,
+      "step": 7898
+    },
+    {
+      "epoch": 0.637478815269147,
+      "grad_norm": 0.7394436597824097,
+      "learning_rate": 0.00013330874951950755,
+      "loss": 2.6214,
+      "step": 7899
+    },
+    {
+      "epoch": 0.6375595190057299,
+      "grad_norm": 0.6279659867286682,
+      "learning_rate": 0.00013329386370297615,
+      "loss": 2.5652,
+      "step": 7900
+    },
+    {
+      "epoch": 0.637640222742313,
+      "grad_norm": 0.7289649248123169,
+      "learning_rate": 0.00013327897705667455,
+      "loss": 2.5628,
+      "step": 7901
+    },
+    {
+      "epoch": 0.637720926478896,
+      "grad_norm": 0.7267701625823975,
+      "learning_rate": 0.0001332640895809739,
+      "loss": 2.5475,
+      "step": 7902
+    },
+    {
+      "epoch": 0.637801630215479,
+      "grad_norm": 0.7470490336418152,
+      "learning_rate": 0.00013324920127624515,
+      "loss": 2.5054,
+      "step": 7903
+    },
+    {
+      "epoch": 0.637882333952062,
+      "grad_norm": 0.6963294148445129,
+      "learning_rate": 0.00013323431214285944,
+      "loss": 2.5992,
+      "step": 7904
+    },
+    {
+      "epoch": 0.6379630376886449,
+      "grad_norm": 0.6993808746337891,
+      "learning_rate": 0.00013321942218118778,
+      "loss": 2.6044,
+      "step": 7905
+    },
+    {
+      "epoch": 0.638043741425228,
+      "grad_norm": 0.6620917916297913,
+      "learning_rate": 0.00013320453139160126,
+      "loss": 2.5278,
+      "step": 7906
+    },
+    {
+      "epoch": 0.638124445161811,
+      "grad_norm": 0.6535444855690002,
+      "learning_rate": 0.00013318963977447106,
+      "loss": 2.6069,
+      "step": 7907
+    },
+    {
+      "epoch": 0.638205148898394,
+      "grad_norm": 0.6913008689880371,
+      "learning_rate": 0.00013317474733016824,
+      "loss": 2.5271,
+      "step": 7908
+    },
+    {
+      "epoch": 0.638285852634977,
+      "grad_norm": 0.6760269403457642,
+      "learning_rate": 0.000133159854059064,
+      "loss": 2.7029,
+      "step": 7909
+    },
+    {
+      "epoch": 0.63836655637156,
+      "grad_norm": 0.7026536464691162,
+      "learning_rate": 0.0001331449599615295,
+      "loss": 2.592,
+      "step": 7910
+    },
+    {
+      "epoch": 0.638447260108143,
+      "grad_norm": 0.7935923933982849,
+      "learning_rate": 0.000133130065037936,
+      "loss": 2.5674,
+      "step": 7911
+    },
+    {
+      "epoch": 0.638527963844726,
+      "grad_norm": 0.694675087928772,
+      "learning_rate": 0.00013311516928865466,
+      "loss": 2.6727,
+      "step": 7912
+    },
+    {
+      "epoch": 0.638608667581309,
+      "grad_norm": 0.7378186583518982,
+      "learning_rate": 0.00013310027271405672,
+      "loss": 2.5691,
+      "step": 7913
+    },
+    {
+      "epoch": 0.638689371317892,
+      "grad_norm": 0.7684193849563599,
+      "learning_rate": 0.00013308537531451345,
+      "loss": 2.5796,
+      "step": 7914
+    },
+    {
+      "epoch": 0.638770075054475,
+      "grad_norm": 0.6881510019302368,
+      "learning_rate": 0.00013307047709039619,
+      "loss": 2.6,
+      "step": 7915
+    },
+    {
+      "epoch": 0.638850778791058,
+      "grad_norm": 0.7341364026069641,
+      "learning_rate": 0.00013305557804207618,
+      "loss": 2.622,
+      "step": 7916
+    },
+    {
+      "epoch": 0.638931482527641,
+      "grad_norm": 0.7620663642883301,
+      "learning_rate": 0.00013304067816992474,
+      "loss": 2.5571,
+      "step": 7917
+    },
+    {
+      "epoch": 0.6390121862642241,
+      "grad_norm": 0.6929789781570435,
+      "learning_rate": 0.00013302577747431322,
+      "loss": 2.6204,
+      "step": 7918
+    },
+    {
+      "epoch": 0.639092890000807,
+      "grad_norm": 0.6942943334579468,
+      "learning_rate": 0.000133010875955613,
+      "loss": 2.6737,
+      "step": 7919
+    },
+    {
+      "epoch": 0.63917359373739,
+      "grad_norm": 0.69537752866745,
+      "learning_rate": 0.0001329959736141955,
+      "loss": 2.6105,
+      "step": 7920
+    },
+    {
+      "epoch": 0.639254297473973,
+      "grad_norm": 0.6690821051597595,
+      "learning_rate": 0.00013298107045043203,
+      "loss": 2.6279,
+      "step": 7921
+    },
+    {
+      "epoch": 0.6393350012105561,
+      "grad_norm": 0.7748103141784668,
+      "learning_rate": 0.00013296616646469412,
+      "loss": 2.6307,
+      "step": 7922
+    },
+    {
+      "epoch": 0.6394157049471391,
+      "grad_norm": 0.7509558200836182,
+      "learning_rate": 0.00013295126165735311,
+      "loss": 2.6388,
+      "step": 7923
+    },
+    {
+      "epoch": 0.639496408683722,
+      "grad_norm": 0.7641764283180237,
+      "learning_rate": 0.0001329363560287806,
+      "loss": 2.5819,
+      "step": 7924
+    },
+    {
+      "epoch": 0.639577112420305,
+      "grad_norm": 0.6912327408790588,
+      "learning_rate": 0.00013292144957934794,
+      "loss": 2.5588,
+      "step": 7925
+    },
+    {
+      "epoch": 0.6396578161568881,
+      "grad_norm": 0.7568803429603577,
+      "learning_rate": 0.0001329065423094267,
+      "loss": 2.5627,
+      "step": 7926
+    },
+    {
+      "epoch": 0.6397385198934711,
+      "grad_norm": 0.7272306084632874,
+      "learning_rate": 0.00013289163421938843,
+      "loss": 2.6101,
+      "step": 7927
+    },
+    {
+      "epoch": 0.6398192236300541,
+      "grad_norm": 0.6965963840484619,
+      "learning_rate": 0.00013287672530960465,
+      "loss": 2.5967,
+      "step": 7928
+    },
+    {
+      "epoch": 0.639899927366637,
+      "grad_norm": 0.7729843854904175,
+      "learning_rate": 0.00013286181558044694,
+      "loss": 2.6222,
+      "step": 7929
+    },
+    {
+      "epoch": 0.6399806311032201,
+      "grad_norm": 0.6876606941223145,
+      "learning_rate": 0.00013284690503228687,
+      "loss": 2.6162,
+      "step": 7930
+    },
+    {
+      "epoch": 0.6400613348398031,
+      "grad_norm": 0.7555204629898071,
+      "learning_rate": 0.0001328319936654961,
+      "loss": 2.588,
+      "step": 7931
+    },
+    {
+      "epoch": 0.6401420385763861,
+      "grad_norm": 0.7324720621109009,
+      "learning_rate": 0.0001328170814804462,
+      "loss": 2.6111,
+      "step": 7932
+    },
+    {
+      "epoch": 0.640222742312969,
+      "grad_norm": 0.6802392601966858,
+      "learning_rate": 0.0001328021684775088,
+      "loss": 2.5955,
+      "step": 7933
+    },
+    {
+      "epoch": 0.6403034460495521,
+      "grad_norm": 0.7564330697059631,
+      "learning_rate": 0.00013278725465705568,
+      "loss": 2.5355,
+      "step": 7934
+    },
+    {
+      "epoch": 0.6403841497861351,
+      "grad_norm": 0.6916235089302063,
+      "learning_rate": 0.00013277234001945844,
+      "loss": 2.6037,
+      "step": 7935
+    },
+    {
+      "epoch": 0.6404648535227181,
+      "grad_norm": 0.688819169998169,
+      "learning_rate": 0.00013275742456508885,
+      "loss": 2.5626,
+      "step": 7936
+    },
+    {
+      "epoch": 0.6405455572593011,
+      "grad_norm": 0.6647922992706299,
+      "learning_rate": 0.0001327425082943186,
+      "loss": 2.6166,
+      "step": 7937
+    },
+    {
+      "epoch": 0.6406262609958842,
+      "grad_norm": 0.6792626976966858,
+      "learning_rate": 0.00013272759120751943,
+      "loss": 2.6206,
+      "step": 7938
+    },
+    {
+      "epoch": 0.6407069647324671,
+      "grad_norm": 0.6482827663421631,
+      "learning_rate": 0.00013271267330506312,
+      "loss": 2.5558,
+      "step": 7939
+    },
+    {
+      "epoch": 0.6407876684690501,
+      "grad_norm": 0.6628372073173523,
+      "learning_rate": 0.0001326977545873215,
+      "loss": 2.5904,
+      "step": 7940
+    },
+    {
+      "epoch": 0.6408683722056331,
+      "grad_norm": 0.7168916463851929,
+      "learning_rate": 0.00013268283505466635,
+      "loss": 2.5189,
+      "step": 7941
+    },
+    {
+      "epoch": 0.6409490759422162,
+      "grad_norm": 0.6691678762435913,
+      "learning_rate": 0.00013266791470746957,
+      "loss": 2.608,
+      "step": 7942
+    },
+    {
+      "epoch": 0.6410297796787991,
+      "grad_norm": 0.6850359439849854,
+      "learning_rate": 0.00013265299354610292,
+      "loss": 2.5929,
+      "step": 7943
+    },
+    {
+      "epoch": 0.6411104834153821,
+      "grad_norm": 0.6807669401168823,
+      "learning_rate": 0.0001326380715709383,
+      "loss": 2.6016,
+      "step": 7944
+    },
+    {
+      "epoch": 0.6411911871519651,
+      "grad_norm": 0.6450446844100952,
+      "learning_rate": 0.00013262314878234767,
+      "loss": 2.6129,
+      "step": 7945
+    },
+    {
+      "epoch": 0.6412718908885482,
+      "grad_norm": 0.679115355014801,
+      "learning_rate": 0.00013260822518070285,
+      "loss": 2.6049,
+      "step": 7946
+    },
+    {
+      "epoch": 0.6413525946251312,
+      "grad_norm": 0.7082008123397827,
+      "learning_rate": 0.00013259330076637583,
+      "loss": 2.5673,
+      "step": 7947
+    },
+    {
+      "epoch": 0.6414332983617141,
+      "grad_norm": 0.7357851266860962,
+      "learning_rate": 0.00013257837553973855,
+      "loss": 2.6118,
+      "step": 7948
+    },
+    {
+      "epoch": 0.6415140020982971,
+      "grad_norm": 0.687035083770752,
+      "learning_rate": 0.000132563449501163,
+      "loss": 2.5359,
+      "step": 7949
+    },
+    {
+      "epoch": 0.6415947058348802,
+      "grad_norm": 0.6950698494911194,
+      "learning_rate": 0.00013254852265102117,
+      "loss": 2.5527,
+      "step": 7950
+    },
+    {
+      "epoch": 0.6416754095714632,
+      "grad_norm": 0.6878959536552429,
+      "learning_rate": 0.00013253359498968507,
+      "loss": 2.611,
+      "step": 7951
+    },
+    {
+      "epoch": 0.6417561133080462,
+      "grad_norm": 0.7224605083465576,
+      "learning_rate": 0.00013251866651752675,
+      "loss": 2.5459,
+      "step": 7952
+    },
+    {
+      "epoch": 0.6418368170446291,
+      "grad_norm": 0.7299731969833374,
+      "learning_rate": 0.00013250373723491826,
+      "loss": 2.5651,
+      "step": 7953
+    },
+    {
+      "epoch": 0.6419175207812122,
+      "grad_norm": 0.7663037776947021,
+      "learning_rate": 0.00013248880714223163,
+      "loss": 2.6073,
+      "step": 7954
+    },
+    {
+      "epoch": 0.6419982245177952,
+      "grad_norm": 0.6532007455825806,
+      "learning_rate": 0.00013247387623983902,
+      "loss": 2.6087,
+      "step": 7955
+    },
+    {
+      "epoch": 0.6420789282543782,
+      "grad_norm": 0.7520449757575989,
+      "learning_rate": 0.00013245894452811255,
+      "loss": 2.5998,
+      "step": 7956
+    },
+    {
+      "epoch": 0.6421596319909612,
+      "grad_norm": 0.7196050882339478,
+      "learning_rate": 0.0001324440120074243,
+      "loss": 2.6448,
+      "step": 7957
+    },
+    {
+      "epoch": 0.6422403357275441,
+      "grad_norm": 0.7093806862831116,
+      "learning_rate": 0.0001324290786781465,
+      "loss": 2.5935,
+      "step": 7958
+    },
+    {
+      "epoch": 0.6423210394641272,
+      "grad_norm": 0.695541501045227,
+      "learning_rate": 0.00013241414454065125,
+      "loss": 2.5872,
+      "step": 7959
+    },
+    {
+      "epoch": 0.6424017432007102,
+      "grad_norm": 0.6763006448745728,
+      "learning_rate": 0.0001323992095953108,
+      "loss": 2.572,
+      "step": 7960
+    },
+    {
+      "epoch": 0.6424824469372932,
+      "grad_norm": 0.6403522491455078,
+      "learning_rate": 0.00013238427384249738,
+      "loss": 2.6137,
+      "step": 7961
+    },
+    {
+      "epoch": 0.6425631506738761,
+      "grad_norm": 0.6647571325302124,
+      "learning_rate": 0.00013236933728258315,
+      "loss": 2.5904,
+      "step": 7962
+    },
+    {
+      "epoch": 0.6426438544104592,
+      "grad_norm": 0.6931071877479553,
+      "learning_rate": 0.0001323543999159405,
+      "loss": 2.6085,
+      "step": 7963
+    },
+    {
+      "epoch": 0.6427245581470422,
+      "grad_norm": 0.6899439096450806,
+      "learning_rate": 0.00013233946174294155,
+      "loss": 2.5555,
+      "step": 7964
+    },
+    {
+      "epoch": 0.6428052618836252,
+      "grad_norm": 0.6564984321594238,
+      "learning_rate": 0.0001323245227639587,
+      "loss": 2.576,
+      "step": 7965
+    },
+    {
+      "epoch": 0.6428859656202082,
+      "grad_norm": 0.7427607774734497,
+      "learning_rate": 0.00013230958297936427,
+      "loss": 2.6178,
+      "step": 7966
+    },
+    {
+      "epoch": 0.6429666693567913,
+      "grad_norm": 0.6884508728981018,
+      "learning_rate": 0.00013229464238953054,
+      "loss": 2.6519,
+      "step": 7967
+    },
+    {
+      "epoch": 0.6430473730933742,
+      "grad_norm": 0.692442774772644,
+      "learning_rate": 0.00013227970099482993,
+      "loss": 2.5784,
+      "step": 7968
+    },
+    {
+      "epoch": 0.6431280768299572,
+      "grad_norm": 0.6637876629829407,
+      "learning_rate": 0.00013226475879563477,
+      "loss": 2.5785,
+      "step": 7969
+    },
+    {
+      "epoch": 0.6432087805665402,
+      "grad_norm": 0.6844972372055054,
+      "learning_rate": 0.0001322498157923175,
+      "loss": 2.5745,
+      "step": 7970
+    },
+    {
+      "epoch": 0.6432894843031233,
+      "grad_norm": 0.7259756922721863,
+      "learning_rate": 0.0001322348719852505,
+      "loss": 2.5696,
+      "step": 7971
+    },
+    {
+      "epoch": 0.6433701880397062,
+      "grad_norm": 0.6719023585319519,
+      "learning_rate": 0.00013221992737480625,
+      "loss": 2.6049,
+      "step": 7972
+    },
+    {
+      "epoch": 0.6434508917762892,
+      "grad_norm": 0.7160155773162842,
+      "learning_rate": 0.00013220498196135717,
+      "loss": 2.572,
+      "step": 7973
+    },
+    {
+      "epoch": 0.6435315955128722,
+      "grad_norm": 0.6920225620269775,
+      "learning_rate": 0.00013219003574527576,
+      "loss": 2.6576,
+      "step": 7974
+    },
+    {
+      "epoch": 0.6436122992494553,
+      "grad_norm": 0.698518693447113,
+      "learning_rate": 0.0001321750887269345,
+      "loss": 2.6074,
+      "step": 7975
+    },
+    {
+      "epoch": 0.6436930029860383,
+      "grad_norm": 0.7607932090759277,
+      "learning_rate": 0.00013216014090670594,
+      "loss": 2.6173,
+      "step": 7976
+    },
+    {
+      "epoch": 0.6437737067226212,
+      "grad_norm": 0.8130847811698914,
+      "learning_rate": 0.0001321451922849626,
+      "loss": 2.6023,
+      "step": 7977
+    },
+    {
+      "epoch": 0.6438544104592042,
+      "grad_norm": 0.676675021648407,
+      "learning_rate": 0.00013213024286207702,
+      "loss": 2.6174,
+      "step": 7978
+    },
+    {
+      "epoch": 0.6439351141957873,
+      "grad_norm": 0.7018851041793823,
+      "learning_rate": 0.00013211529263842183,
+      "loss": 2.5713,
+      "step": 7979
+    },
+    {
+      "epoch": 0.6440158179323703,
+      "grad_norm": 0.796097457408905,
+      "learning_rate": 0.00013210034161436954,
+      "loss": 2.5937,
+      "step": 7980
+    },
+    {
+      "epoch": 0.6440965216689533,
+      "grad_norm": 0.7118527293205261,
+      "learning_rate": 0.0001320853897902929,
+      "loss": 2.5721,
+      "step": 7981
+    },
+    {
+      "epoch": 0.6441772254055362,
+      "grad_norm": 0.7282249331474304,
+      "learning_rate": 0.00013207043716656445,
+      "loss": 2.5975,
+      "step": 7982
+    },
+    {
+      "epoch": 0.6442579291421193,
+      "grad_norm": 0.6710900664329529,
+      "learning_rate": 0.00013205548374355686,
+      "loss": 2.5809,
+      "step": 7983
+    },
+    {
+      "epoch": 0.6443386328787023,
+      "grad_norm": 0.7045658230781555,
+      "learning_rate": 0.00013204052952164278,
+      "loss": 2.5715,
+      "step": 7984
+    },
+    {
+      "epoch": 0.6444193366152853,
+      "grad_norm": 0.719507098197937,
+      "learning_rate": 0.00013202557450119504,
+      "loss": 2.5948,
+      "step": 7985
+    },
+    {
+      "epoch": 0.6445000403518683,
+      "grad_norm": 0.7603922486305237,
+      "learning_rate": 0.0001320106186825862,
+      "loss": 2.6176,
+      "step": 7986
+    },
+    {
+      "epoch": 0.6445807440884513,
+      "grad_norm": 0.7057444453239441,
+      "learning_rate": 0.0001319956620661891,
+      "loss": 2.5905,
+      "step": 7987
+    },
+    {
+      "epoch": 0.6446614478250343,
+      "grad_norm": 0.7884874939918518,
+      "learning_rate": 0.00013198070465237645,
+      "loss": 2.5892,
+      "step": 7988
+    },
+    {
+      "epoch": 0.6447421515616173,
+      "grad_norm": 0.6932834386825562,
+      "learning_rate": 0.00013196574644152103,
+      "loss": 2.6032,
+      "step": 7989
+    },
+    {
+      "epoch": 0.6448228552982003,
+      "grad_norm": 0.7361180186271667,
+      "learning_rate": 0.00013195078743399568,
+      "loss": 2.5877,
+      "step": 7990
+    },
+    {
+      "epoch": 0.6449035590347834,
+      "grad_norm": 0.6843615174293518,
+      "learning_rate": 0.00013193582763017315,
+      "loss": 2.5804,
+      "step": 7991
+    },
+    {
+      "epoch": 0.6449842627713663,
+      "grad_norm": 0.7592078447341919,
+      "learning_rate": 0.00013192086703042635,
+      "loss": 2.6464,
+      "step": 7992
+    },
+    {
+      "epoch": 0.6450649665079493,
+      "grad_norm": 0.7362154126167297,
+      "learning_rate": 0.0001319059056351281,
+      "loss": 2.6154,
+      "step": 7993
+    },
+    {
+      "epoch": 0.6451456702445323,
+      "grad_norm": 0.6721758246421814,
+      "learning_rate": 0.00013189094344465125,
+      "loss": 2.5735,
+      "step": 7994
+    },
+    {
+      "epoch": 0.6452263739811154,
+      "grad_norm": 0.6221550107002258,
+      "learning_rate": 0.00013187598045936874,
+      "loss": 2.5612,
+      "step": 7995
+    },
+    {
+      "epoch": 0.6453070777176984,
+      "grad_norm": 0.7225528359413147,
+      "learning_rate": 0.00013186101667965344,
+      "loss": 2.6263,
+      "step": 7996
+    },
+    {
+      "epoch": 0.6453877814542813,
+      "grad_norm": 0.7599418759346008,
+      "learning_rate": 0.00013184605210587837,
+      "loss": 2.5814,
+      "step": 7997
+    },
+    {
+      "epoch": 0.6454684851908643,
+      "grad_norm": 0.6778777837753296,
+      "learning_rate": 0.00013183108673841642,
+      "loss": 2.6158,
+      "step": 7998
+    },
+    {
+      "epoch": 0.6455491889274474,
+      "grad_norm": 0.6860963106155396,
+      "learning_rate": 0.00013181612057764058,
+      "loss": 2.6207,
+      "step": 7999
+    },
+    {
+      "epoch": 0.6456298926640304,
+      "grad_norm": 0.6615182757377625,
+      "learning_rate": 0.00013180115362392382,
+      "loss": 2.5571,
+      "step": 8000
+    },
+    {
+      "epoch": 0.6456298926640304,
+      "eval_loss": 2.5128066539764404,
+      "eval_runtime": 754.3655,
+      "eval_samples_per_second": 3.473,
+      "eval_steps_per_second": 0.579,
+      "step": 8000
+    },
+    {
+      "epoch": 0.6457105964006133,
+      "grad_norm": 0.688169538974762,
+      "learning_rate": 0.0001317861858776392,
+      "loss": 2.6513,
+      "step": 8001
+    },
+    {
+      "epoch": 0.6457913001371963,
+      "grad_norm": 0.6726182103157043,
+      "learning_rate": 0.00013177121733915975,
+      "loss": 2.5909,
+      "step": 8002
+    },
+    {
+      "epoch": 0.6458720038737794,
+      "grad_norm": 0.7348085641860962,
+      "learning_rate": 0.00013175624800885853,
+      "loss": 2.577,
+      "step": 8003
+    },
+    {
+      "epoch": 0.6459527076103624,
+      "grad_norm": 0.677435040473938,
+      "learning_rate": 0.00013174127788710856,
+      "loss": 2.5056,
+      "step": 8004
+    },
+    {
+      "epoch": 0.6460334113469454,
+      "grad_norm": 0.6864951848983765,
+      "learning_rate": 0.000131726306974283,
+      "loss": 2.5733,
+      "step": 8005
+    },
+    {
+      "epoch": 0.6461141150835283,
+      "grad_norm": 0.7070075869560242,
+      "learning_rate": 0.0001317113352707549,
+      "loss": 2.5359,
+      "step": 8006
+    },
+    {
+      "epoch": 0.6461948188201113,
+      "grad_norm": 0.7065049409866333,
+      "learning_rate": 0.00013169636277689746,
+      "loss": 2.6261,
+      "step": 8007
+    },
+    {
+      "epoch": 0.6462755225566944,
+      "grad_norm": 0.6691577434539795,
+      "learning_rate": 0.0001316813894930838,
+      "loss": 2.6015,
+      "step": 8008
+    },
+    {
+      "epoch": 0.6463562262932774,
+      "grad_norm": 0.6754019260406494,
+      "learning_rate": 0.0001316664154196871,
+      "loss": 2.5954,
+      "step": 8009
+    },
+    {
+      "epoch": 0.6464369300298604,
+      "grad_norm": 0.6172776818275452,
+      "learning_rate": 0.00013165144055708055,
+      "loss": 2.5599,
+      "step": 8010
+    },
+    {
+      "epoch": 0.6465176337664433,
+      "grad_norm": 0.6778094172477722,
+      "learning_rate": 0.00013163646490563737,
+      "loss": 2.5407,
+      "step": 8011
+    },
+    {
+      "epoch": 0.6465983375030264,
+      "grad_norm": 0.7363924980163574,
+      "learning_rate": 0.00013162148846573076,
+      "loss": 2.6075,
+      "step": 8012
+    },
+    {
+      "epoch": 0.6466790412396094,
+      "grad_norm": 0.6662711501121521,
+      "learning_rate": 0.00013160651123773404,
+      "loss": 2.5611,
+      "step": 8013
+    },
+    {
+      "epoch": 0.6467597449761924,
+      "grad_norm": 0.699670135974884,
+      "learning_rate": 0.00013159153322202043,
+      "loss": 2.5612,
+      "step": 8014
+    },
+    {
+      "epoch": 0.6468404487127754,
+      "grad_norm": 0.7382899522781372,
+      "learning_rate": 0.0001315765544189632,
+      "loss": 2.6017,
+      "step": 8015
+    },
+    {
+      "epoch": 0.6469211524493584,
+      "grad_norm": 0.7624868154525757,
+      "learning_rate": 0.0001315615748289357,
+      "loss": 2.6174,
+      "step": 8016
+    },
+    {
+      "epoch": 0.6470018561859414,
+      "grad_norm": 0.704622745513916,
+      "learning_rate": 0.00013154659445231129,
+      "loss": 2.5367,
+      "step": 8017
+    },
+    {
+      "epoch": 0.6470825599225244,
+      "grad_norm": 0.7117413878440857,
+      "learning_rate": 0.00013153161328946324,
+      "loss": 2.5958,
+      "step": 8018
+    },
+    {
+      "epoch": 0.6471632636591074,
+      "grad_norm": 0.6825408339500427,
+      "learning_rate": 0.00013151663134076497,
+      "loss": 2.5118,
+      "step": 8019
+    },
+    {
+      "epoch": 0.6472439673956905,
+      "grad_norm": 0.6732384562492371,
+      "learning_rate": 0.00013150164860658986,
+      "loss": 2.6312,
+      "step": 8020
+    },
+    {
+      "epoch": 0.6473246711322734,
+      "grad_norm": 0.712812602519989,
+      "learning_rate": 0.00013148666508731134,
+      "loss": 2.576,
+      "step": 8021
+    },
+    {
+      "epoch": 0.6474053748688564,
+      "grad_norm": 0.8128857612609863,
+      "learning_rate": 0.0001314716807833028,
+      "loss": 2.5333,
+      "step": 8022
+    },
+    {
+      "epoch": 0.6474860786054394,
+      "grad_norm": 0.7817162275314331,
+      "learning_rate": 0.00013145669569493773,
+      "loss": 2.6835,
+      "step": 8023
+    },
+    {
+      "epoch": 0.6475667823420225,
+      "grad_norm": 0.7164301872253418,
+      "learning_rate": 0.00013144170982258956,
+      "loss": 2.5573,
+      "step": 8024
+    },
+    {
+      "epoch": 0.6476474860786054,
+      "grad_norm": 0.67625892162323,
+      "learning_rate": 0.00013142672316663177,
+      "loss": 2.5976,
+      "step": 8025
+    },
+    {
+      "epoch": 0.6477281898151884,
+      "grad_norm": 0.6919494867324829,
+      "learning_rate": 0.0001314117357274379,
+      "loss": 2.6179,
+      "step": 8026
+    },
+    {
+      "epoch": 0.6478088935517714,
+      "grad_norm": 0.6787464618682861,
+      "learning_rate": 0.0001313967475053815,
+      "loss": 2.5405,
+      "step": 8027
+    },
+    {
+      "epoch": 0.6478895972883545,
+      "grad_norm": 0.6305621862411499,
+      "learning_rate": 0.00013138175850083605,
+      "loss": 2.6016,
+      "step": 8028
+    },
+    {
+      "epoch": 0.6479703010249375,
+      "grad_norm": 0.7456182837486267,
+      "learning_rate": 0.00013136676871417516,
+      "loss": 2.6091,
+      "step": 8029
+    },
+    {
+      "epoch": 0.6480510047615204,
+      "grad_norm": 0.7047890424728394,
+      "learning_rate": 0.00013135177814577238,
+      "loss": 2.6108,
+      "step": 8030
+    },
+    {
+      "epoch": 0.6481317084981034,
+      "grad_norm": 0.7509389519691467,
+      "learning_rate": 0.00013133678679600133,
+      "loss": 2.6396,
+      "step": 8031
+    },
+    {
+      "epoch": 0.6482124122346865,
+      "grad_norm": 0.63836270570755,
+      "learning_rate": 0.00013132179466523566,
+      "loss": 2.5759,
+      "step": 8032
+    },
+    {
+      "epoch": 0.6482931159712695,
+      "grad_norm": 0.6994885206222534,
+      "learning_rate": 0.000131306801753849,
+      "loss": 2.61,
+      "step": 8033
+    },
+    {
+      "epoch": 0.6483738197078525,
+      "grad_norm": 0.6762083768844604,
+      "learning_rate": 0.00013129180806221497,
+      "loss": 2.5431,
+      "step": 8034
+    },
+    {
+      "epoch": 0.6484545234444354,
+      "grad_norm": 0.6890944242477417,
+      "learning_rate": 0.0001312768135907073,
+      "loss": 2.5922,
+      "step": 8035
+    },
+    {
+      "epoch": 0.6485352271810185,
+      "grad_norm": 0.7409473061561584,
+      "learning_rate": 0.0001312618183396997,
+      "loss": 2.6132,
+      "step": 8036
+    },
+    {
+      "epoch": 0.6486159309176015,
+      "grad_norm": 0.6660643815994263,
+      "learning_rate": 0.00013124682230956585,
+      "loss": 2.5816,
+      "step": 8037
+    },
+    {
+      "epoch": 0.6486966346541845,
+      "grad_norm": 0.714235246181488,
+      "learning_rate": 0.0001312318255006795,
+      "loss": 2.5613,
+      "step": 8038
+    },
+    {
+      "epoch": 0.6487773383907675,
+      "grad_norm": 0.6568472385406494,
+      "learning_rate": 0.00013121682791341442,
+      "loss": 2.6382,
+      "step": 8039
+    },
+    {
+      "epoch": 0.6488580421273505,
+      "grad_norm": 0.6874251961708069,
+      "learning_rate": 0.00013120182954814438,
+      "loss": 2.593,
+      "step": 8040
+    },
+    {
+      "epoch": 0.6489387458639335,
+      "grad_norm": 0.7620158791542053,
+      "learning_rate": 0.0001311868304052432,
+      "loss": 2.589,
+      "step": 8041
+    },
+    {
+      "epoch": 0.6490194496005165,
+      "grad_norm": 0.6755926609039307,
+      "learning_rate": 0.00013117183048508467,
+      "loss": 2.5876,
+      "step": 8042
+    },
+    {
+      "epoch": 0.6491001533370995,
+      "grad_norm": 0.6952808499336243,
+      "learning_rate": 0.00013115682978804264,
+      "loss": 2.5909,
+      "step": 8043
+    },
+    {
+      "epoch": 0.6491808570736826,
+      "grad_norm": 0.6599535346031189,
+      "learning_rate": 0.00013114182831449098,
+      "loss": 2.6031,
+      "step": 8044
+    },
+    {
+      "epoch": 0.6492615608102655,
+      "grad_norm": 0.7816598415374756,
+      "learning_rate": 0.00013112682606480355,
+      "loss": 2.5633,
+      "step": 8045
+    },
+    {
+      "epoch": 0.6493422645468485,
+      "grad_norm": 0.7188639640808105,
+      "learning_rate": 0.00013111182303935425,
+      "loss": 2.6292,
+      "step": 8046
+    },
+    {
+      "epoch": 0.6494229682834315,
+      "grad_norm": 0.7131505608558655,
+      "learning_rate": 0.00013109681923851698,
+      "loss": 2.5729,
+      "step": 8047
+    },
+    {
+      "epoch": 0.6495036720200146,
+      "grad_norm": 0.7466408014297485,
+      "learning_rate": 0.00013108181466266568,
+      "loss": 2.5742,
+      "step": 8048
+    },
+    {
+      "epoch": 0.6495843757565976,
+      "grad_norm": 0.6707943677902222,
+      "learning_rate": 0.00013106680931217437,
+      "loss": 2.5506,
+      "step": 8049
+    },
+    {
+      "epoch": 0.6496650794931805,
+      "grad_norm": 0.6913424730300903,
+      "learning_rate": 0.0001310518031874169,
+      "loss": 2.5639,
+      "step": 8050
+    },
+    {
+      "epoch": 0.6497457832297635,
+      "grad_norm": 0.8261755704879761,
+      "learning_rate": 0.00013103679628876733,
+      "loss": 2.601,
+      "step": 8051
+    },
+    {
+      "epoch": 0.6498264869663466,
+      "grad_norm": 0.7410566806793213,
+      "learning_rate": 0.0001310217886165997,
+      "loss": 2.5326,
+      "step": 8052
+    },
+    {
+      "epoch": 0.6499071907029296,
+      "grad_norm": 0.7032365202903748,
+      "learning_rate": 0.00013100678017128798,
+      "loss": 2.5907,
+      "step": 8053
+    },
+    {
+      "epoch": 0.6499878944395125,
+      "grad_norm": 0.7074568271636963,
+      "learning_rate": 0.00013099177095320626,
+      "loss": 2.6193,
+      "step": 8054
+    },
+    {
+      "epoch": 0.6500685981760955,
+      "grad_norm": 0.7754546999931335,
+      "learning_rate": 0.00013097676096272855,
+      "loss": 2.5832,
+      "step": 8055
+    },
+    {
+      "epoch": 0.6501493019126786,
+      "grad_norm": 0.7475717663764954,
+      "learning_rate": 0.00013096175020022903,
+      "loss": 2.6233,
+      "step": 8056
+    },
+    {
+      "epoch": 0.6502300056492616,
+      "grad_norm": 0.7863949537277222,
+      "learning_rate": 0.00013094673866608173,
+      "loss": 2.5745,
+      "step": 8057
+    },
+    {
+      "epoch": 0.6503107093858446,
+      "grad_norm": 0.69294673204422,
+      "learning_rate": 0.0001309317263606608,
+      "loss": 2.5982,
+      "step": 8058
+    },
+    {
+      "epoch": 0.6503914131224275,
+      "grad_norm": 0.7096135020256042,
+      "learning_rate": 0.00013091671328434046,
+      "loss": 2.5944,
+      "step": 8059
+    },
+    {
+      "epoch": 0.6504721168590105,
+      "grad_norm": 0.7001097202301025,
+      "learning_rate": 0.00013090169943749476,
+      "loss": 2.5435,
+      "step": 8060
+    },
+    {
+      "epoch": 0.6505528205955936,
+      "grad_norm": 0.7522539496421814,
+      "learning_rate": 0.00013088668482049792,
+      "loss": 2.5843,
+      "step": 8061
+    },
+    {
+      "epoch": 0.6506335243321766,
+      "grad_norm": 0.6675420999526978,
+      "learning_rate": 0.00013087166943372418,
+      "loss": 2.5623,
+      "step": 8062
+    },
+    {
+      "epoch": 0.6507142280687596,
+      "grad_norm": 0.7779181599617004,
+      "learning_rate": 0.00013085665327754772,
+      "loss": 2.6087,
+      "step": 8063
+    },
+    {
+      "epoch": 0.6507949318053425,
+      "grad_norm": 0.7385239005088806,
+      "learning_rate": 0.00013084163635234284,
+      "loss": 2.5725,
+      "step": 8064
+    },
+    {
+      "epoch": 0.6508756355419256,
+      "grad_norm": 0.6966612339019775,
+      "learning_rate": 0.00013082661865848375,
+      "loss": 2.5745,
+      "step": 8065
+    },
+    {
+      "epoch": 0.6509563392785086,
+      "grad_norm": 0.7098337411880493,
+      "learning_rate": 0.00013081160019634468,
+      "loss": 2.5461,
+      "step": 8066
+    },
+    {
+      "epoch": 0.6510370430150916,
+      "grad_norm": 0.6514503359794617,
+      "learning_rate": 0.00013079658096630002,
+      "loss": 2.5869,
+      "step": 8067
+    },
+    {
+      "epoch": 0.6511177467516746,
+      "grad_norm": 0.680422306060791,
+      "learning_rate": 0.0001307815609687241,
+      "loss": 2.6316,
+      "step": 8068
+    },
+    {
+      "epoch": 0.6511984504882576,
+      "grad_norm": 0.6892665028572083,
+      "learning_rate": 0.00013076654020399117,
+      "loss": 2.5862,
+      "step": 8069
+    },
+    {
+      "epoch": 0.6512791542248406,
+      "grad_norm": 0.7605568170547485,
+      "learning_rate": 0.00013075151867247568,
+      "loss": 2.5342,
+      "step": 8070
+    },
+    {
+      "epoch": 0.6513598579614236,
+      "grad_norm": 0.7571204900741577,
+      "learning_rate": 0.00013073649637455192,
+      "loss": 2.5762,
+      "step": 8071
+    },
+    {
+      "epoch": 0.6514405616980066,
+      "grad_norm": 0.6910812258720398,
+      "learning_rate": 0.00013072147331059431,
+      "loss": 2.6635,
+      "step": 8072
+    },
+    {
+      "epoch": 0.6515212654345897,
+      "grad_norm": 0.765559196472168,
+      "learning_rate": 0.00013070644948097733,
+      "loss": 2.5885,
+      "step": 8073
+    },
+    {
+      "epoch": 0.6516019691711726,
+      "grad_norm": 0.7533665299415588,
+      "learning_rate": 0.00013069142488607532,
+      "loss": 2.6545,
+      "step": 8074
+    },
+    {
+      "epoch": 0.6516826729077556,
+      "grad_norm": 0.685089647769928,
+      "learning_rate": 0.0001306763995262628,
+      "loss": 2.5955,
+      "step": 8075
+    },
+    {
+      "epoch": 0.6517633766443386,
+      "grad_norm": 0.7280653715133667,
+      "learning_rate": 0.00013066137340191422,
+      "loss": 2.5548,
+      "step": 8076
+    },
+    {
+      "epoch": 0.6518440803809217,
+      "grad_norm": 0.6881482601165771,
+      "learning_rate": 0.00013064634651340404,
+      "loss": 2.6143,
+      "step": 8077
+    },
+    {
+      "epoch": 0.6519247841175047,
+      "grad_norm": 0.6878265142440796,
+      "learning_rate": 0.0001306313188611068,
+      "loss": 2.5681,
+      "step": 8078
+    },
+    {
+      "epoch": 0.6520054878540876,
+      "grad_norm": 0.685238242149353,
+      "learning_rate": 0.00013061629044539702,
+      "loss": 2.5517,
+      "step": 8079
+    },
+    {
+      "epoch": 0.6520861915906706,
+      "grad_norm": 0.6689820885658264,
+      "learning_rate": 0.00013060126126664928,
+      "loss": 2.6201,
+      "step": 8080
+    },
+    {
+      "epoch": 0.6521668953272537,
+      "grad_norm": 0.7128999829292297,
+      "learning_rate": 0.00013058623132523807,
+      "loss": 2.5829,
+      "step": 8081
+    },
+    {
+      "epoch": 0.6522475990638367,
+      "grad_norm": 0.6835216879844666,
+      "learning_rate": 0.00013057120062153805,
+      "loss": 2.6312,
+      "step": 8082
+    },
+    {
+      "epoch": 0.6523283028004196,
+      "grad_norm": 0.7140012383460999,
+      "learning_rate": 0.00013055616915592382,
+      "loss": 2.6148,
+      "step": 8083
+    },
+    {
+      "epoch": 0.6524090065370026,
+      "grad_norm": 0.7378252148628235,
+      "learning_rate": 0.00013054113692876994,
+      "loss": 2.5805,
+      "step": 8084
+    },
+    {
+      "epoch": 0.6524897102735857,
+      "grad_norm": 0.7569258213043213,
+      "learning_rate": 0.0001305261039404511,
+      "loss": 2.6088,
+      "step": 8085
+    },
+    {
+      "epoch": 0.6525704140101687,
+      "grad_norm": 0.6909007430076599,
+      "learning_rate": 0.00013051107019134195,
+      "loss": 2.5285,
+      "step": 8086
+    },
+    {
+      "epoch": 0.6526511177467517,
+      "grad_norm": 0.6785587072372437,
+      "learning_rate": 0.0001304960356818172,
+      "loss": 2.5527,
+      "step": 8087
+    },
+    {
+      "epoch": 0.6527318214833346,
+      "grad_norm": 0.7058801054954529,
+      "learning_rate": 0.0001304810004122515,
+      "loss": 2.6789,
+      "step": 8088
+    },
+    {
+      "epoch": 0.6528125252199177,
+      "grad_norm": 0.6920512318611145,
+      "learning_rate": 0.0001304659643830196,
+      "loss": 2.5748,
+      "step": 8089
+    },
+    {
+      "epoch": 0.6528932289565007,
+      "grad_norm": 0.6829244494438171,
+      "learning_rate": 0.00013045092759449625,
+      "loss": 2.5389,
+      "step": 8090
+    },
+    {
+      "epoch": 0.6529739326930837,
+      "grad_norm": 0.6942421793937683,
+      "learning_rate": 0.00013043589004705614,
+      "loss": 2.5851,
+      "step": 8091
+    },
+    {
+      "epoch": 0.6530546364296667,
+      "grad_norm": 0.6473072171211243,
+      "learning_rate": 0.0001304208517410741,
+      "loss": 2.56,
+      "step": 8092
+    },
+    {
+      "epoch": 0.6531353401662497,
+      "grad_norm": 0.6692056655883789,
+      "learning_rate": 0.00013040581267692494,
+      "loss": 2.5977,
+      "step": 8093
+    },
+    {
+      "epoch": 0.6532160439028327,
+      "grad_norm": 0.6918915510177612,
+      "learning_rate": 0.00013039077285498344,
+      "loss": 2.551,
+      "step": 8094
+    },
+    {
+      "epoch": 0.6532967476394157,
+      "grad_norm": 0.7432852387428284,
+      "learning_rate": 0.00013037573227562443,
+      "loss": 2.5537,
+      "step": 8095
+    },
+    {
+      "epoch": 0.6533774513759987,
+      "grad_norm": 0.6737081408500671,
+      "learning_rate": 0.0001303606909392228,
+      "loss": 2.5947,
+      "step": 8096
+    },
+    {
+      "epoch": 0.6534581551125818,
+      "grad_norm": 0.6810599565505981,
+      "learning_rate": 0.0001303456488461533,
+      "loss": 2.5704,
+      "step": 8097
+    },
+    {
+      "epoch": 0.6535388588491647,
+      "grad_norm": 0.675240159034729,
+      "learning_rate": 0.00013033060599679098,
+      "loss": 2.591,
+      "step": 8098
+    },
+    {
+      "epoch": 0.6536195625857477,
+      "grad_norm": 0.6888695359230042,
+      "learning_rate": 0.00013031556239151066,
+      "loss": 2.5403,
+      "step": 8099
+    },
+    {
+      "epoch": 0.6537002663223307,
+      "grad_norm": 0.7154796719551086,
+      "learning_rate": 0.00013030051803068727,
+      "loss": 2.5654,
+      "step": 8100
+    },
+    {
+      "epoch": 0.6537809700589138,
+      "grad_norm": 0.6655243635177612,
+      "learning_rate": 0.0001302854729146958,
+      "loss": 2.5867,
+      "step": 8101
+    },
+    {
+      "epoch": 0.6538616737954968,
+      "grad_norm": 0.7070788145065308,
+      "learning_rate": 0.00013027042704391115,
+      "loss": 2.5593,
+      "step": 8102
+    },
+    {
+      "epoch": 0.6539423775320797,
+      "grad_norm": 0.7071834206581116,
+      "learning_rate": 0.0001302553804187083,
+      "loss": 2.536,
+      "step": 8103
+    },
+    {
+      "epoch": 0.6540230812686627,
+      "grad_norm": 0.7086542248725891,
+      "learning_rate": 0.00013024033303946233,
+      "loss": 2.5644,
+      "step": 8104
+    },
+    {
+      "epoch": 0.6541037850052458,
+      "grad_norm": 0.6714556813240051,
+      "learning_rate": 0.00013022528490654818,
+      "loss": 2.5167,
+      "step": 8105
+    },
+    {
+      "epoch": 0.6541844887418288,
+      "grad_norm": 0.6905114054679871,
+      "learning_rate": 0.00013021023602034095,
+      "loss": 2.5227,
+      "step": 8106
+    },
+    {
+      "epoch": 0.6542651924784118,
+      "grad_norm": 0.7050586342811584,
+      "learning_rate": 0.00013019518638121563,
+      "loss": 2.5725,
+      "step": 8107
+    },
+    {
+      "epoch": 0.6543458962149947,
+      "grad_norm": 0.6940500736236572,
+      "learning_rate": 0.00013018013598954737,
+      "loss": 2.5912,
+      "step": 8108
+    },
+    {
+      "epoch": 0.6544265999515777,
+      "grad_norm": 0.7136965990066528,
+      "learning_rate": 0.00013016508484571122,
+      "loss": 2.6101,
+      "step": 8109
+    },
+    {
+      "epoch": 0.6545073036881608,
+      "grad_norm": 0.7205774188041687,
+      "learning_rate": 0.0001301500329500823,
+      "loss": 2.5869,
+      "step": 8110
+    },
+    {
+      "epoch": 0.6545880074247438,
+      "grad_norm": 0.6831154823303223,
+      "learning_rate": 0.00013013498030303575,
+      "loss": 2.5309,
+      "step": 8111
+    },
+    {
+      "epoch": 0.6546687111613267,
+      "grad_norm": 0.6778538823127747,
+      "learning_rate": 0.0001301199269049467,
+      "loss": 2.6297,
+      "step": 8112
+    },
+    {
+      "epoch": 0.6547494148979097,
+      "grad_norm": 0.705055832862854,
+      "learning_rate": 0.00013010487275619034,
+      "loss": 2.6188,
+      "step": 8113
+    },
+    {
+      "epoch": 0.6548301186344928,
+      "grad_norm": 0.6927980780601501,
+      "learning_rate": 0.00013008981785714188,
+      "loss": 2.5744,
+      "step": 8114
+    },
+    {
+      "epoch": 0.6549108223710758,
+      "grad_norm": 0.7070884108543396,
+      "learning_rate": 0.0001300747622081765,
+      "loss": 2.618,
+      "step": 8115
+    },
+    {
+      "epoch": 0.6549915261076588,
+      "grad_norm": 0.723479688167572,
+      "learning_rate": 0.0001300597058096694,
+      "loss": 2.5928,
+      "step": 8116
+    },
+    {
+      "epoch": 0.6550722298442417,
+      "grad_norm": 0.6689562201499939,
+      "learning_rate": 0.00013004464866199587,
+      "loss": 2.5592,
+      "step": 8117
+    },
+    {
+      "epoch": 0.6551529335808248,
+      "grad_norm": 0.6685079336166382,
+      "learning_rate": 0.00013002959076553115,
+      "loss": 2.558,
+      "step": 8118
+    },
+    {
+      "epoch": 0.6552336373174078,
+      "grad_norm": 0.678105890750885,
+      "learning_rate": 0.00013001453212065057,
+      "loss": 2.6176,
+      "step": 8119
+    },
+    {
+      "epoch": 0.6553143410539908,
+      "grad_norm": 0.7355597019195557,
+      "learning_rate": 0.00012999947272772933,
+      "loss": 2.6293,
+      "step": 8120
+    },
+    {
+      "epoch": 0.6553950447905738,
+      "grad_norm": 0.735862672328949,
+      "learning_rate": 0.00012998441258714284,
+      "loss": 2.635,
+      "step": 8121
+    },
+    {
+      "epoch": 0.6554757485271568,
+      "grad_norm": 0.6766025424003601,
+      "learning_rate": 0.0001299693516992664,
+      "loss": 2.5829,
+      "step": 8122
+    },
+    {
+      "epoch": 0.6555564522637398,
+      "grad_norm": 0.6701885461807251,
+      "learning_rate": 0.00012995429006447542,
+      "loss": 2.5996,
+      "step": 8123
+    },
+    {
+      "epoch": 0.6556371560003228,
+      "grad_norm": 0.6814082264900208,
+      "learning_rate": 0.00012993922768314518,
+      "loss": 2.5906,
+      "step": 8124
+    },
+    {
+      "epoch": 0.6557178597369058,
+      "grad_norm": 0.7104958295822144,
+      "learning_rate": 0.00012992416455565113,
+      "loss": 2.6708,
+      "step": 8125
+    },
+    {
+      "epoch": 0.6557985634734889,
+      "grad_norm": 0.6451221108436584,
+      "learning_rate": 0.0001299091006823687,
+      "loss": 2.5512,
+      "step": 8126
+    },
+    {
+      "epoch": 0.6558792672100718,
+      "grad_norm": 0.6736068725585938,
+      "learning_rate": 0.0001298940360636733,
+      "loss": 2.5839,
+      "step": 8127
+    },
+    {
+      "epoch": 0.6559599709466548,
+      "grad_norm": 0.6873149871826172,
+      "learning_rate": 0.00012987897069994031,
+      "loss": 2.5804,
+      "step": 8128
+    },
+    {
+      "epoch": 0.6560406746832378,
+      "grad_norm": 0.6937728524208069,
+      "learning_rate": 0.00012986390459154533,
+      "loss": 2.5648,
+      "step": 8129
+    },
+    {
+      "epoch": 0.6561213784198209,
+      "grad_norm": 0.7109464406967163,
+      "learning_rate": 0.00012984883773886377,
+      "loss": 2.6132,
+      "step": 8130
+    },
+    {
+      "epoch": 0.6562020821564039,
+      "grad_norm": 0.7134159803390503,
+      "learning_rate": 0.00012983377014227115,
+      "loss": 2.6029,
+      "step": 8131
+    },
+    {
+      "epoch": 0.6562827858929868,
+      "grad_norm": 0.6788110733032227,
+      "learning_rate": 0.000129818701802143,
+      "loss": 2.6344,
+      "step": 8132
+    },
+    {
+      "epoch": 0.6563634896295698,
+      "grad_norm": 0.6798231601715088,
+      "learning_rate": 0.00012980363271885483,
+      "loss": 2.5758,
+      "step": 8133
+    },
+    {
+      "epoch": 0.6564441933661529,
+      "grad_norm": 0.6586930155754089,
+      "learning_rate": 0.00012978856289278226,
+      "loss": 2.5918,
+      "step": 8134
+    },
+    {
+      "epoch": 0.6565248971027359,
+      "grad_norm": 0.6614218950271606,
+      "learning_rate": 0.0001297734923243008,
+      "loss": 2.5777,
+      "step": 8135
+    },
+    {
+      "epoch": 0.6566056008393188,
+      "grad_norm": 0.6874340176582336,
+      "learning_rate": 0.0001297584210137861,
+      "loss": 2.5528,
+      "step": 8136
+    },
+    {
+      "epoch": 0.6566863045759018,
+      "grad_norm": 0.6972174048423767,
+      "learning_rate": 0.00012974334896161376,
+      "loss": 2.6551,
+      "step": 8137
+    },
+    {
+      "epoch": 0.6567670083124849,
+      "grad_norm": 0.7414106726646423,
+      "learning_rate": 0.0001297282761681594,
+      "loss": 2.5719,
+      "step": 8138
+    },
+    {
+      "epoch": 0.6568477120490679,
+      "grad_norm": 0.6678279042243958,
+      "learning_rate": 0.00012971320263379868,
+      "loss": 2.555,
+      "step": 8139
+    },
+    {
+      "epoch": 0.6569284157856509,
+      "grad_norm": 0.692149817943573,
+      "learning_rate": 0.0001296981283589073,
+      "loss": 2.5991,
+      "step": 8140
+    },
+    {
+      "epoch": 0.6570091195222338,
+      "grad_norm": 0.6937025189399719,
+      "learning_rate": 0.00012968305334386094,
+      "loss": 2.5635,
+      "step": 8141
+    },
+    {
+      "epoch": 0.6570898232588169,
+      "grad_norm": 0.6250358819961548,
+      "learning_rate": 0.00012966797758903528,
+      "loss": 2.55,
+      "step": 8142
+    },
+    {
+      "epoch": 0.6571705269953999,
+      "grad_norm": 0.7388221025466919,
+      "learning_rate": 0.00012965290109480607,
+      "loss": 2.5307,
+      "step": 8143
+    },
+    {
+      "epoch": 0.6572512307319829,
+      "grad_norm": 0.7165891528129578,
+      "learning_rate": 0.00012963782386154904,
+      "loss": 2.5482,
+      "step": 8144
+    },
+    {
+      "epoch": 0.6573319344685659,
+      "grad_norm": 0.7605282068252563,
+      "learning_rate": 0.00012962274588963996,
+      "loss": 2.5839,
+      "step": 8145
+    },
+    {
+      "epoch": 0.657412638205149,
+      "grad_norm": 0.7259613275527954,
+      "learning_rate": 0.00012960766717945465,
+      "loss": 2.5612,
+      "step": 8146
+    },
+    {
+      "epoch": 0.6574933419417319,
+      "grad_norm": 0.7301480770111084,
+      "learning_rate": 0.00012959258773136885,
+      "loss": 2.5365,
+      "step": 8147
+    },
+    {
+      "epoch": 0.6575740456783149,
+      "grad_norm": 0.6800966262817383,
+      "learning_rate": 0.0001295775075457584,
+      "loss": 2.5663,
+      "step": 8148
+    },
+    {
+      "epoch": 0.6576547494148979,
+      "grad_norm": 0.6968960165977478,
+      "learning_rate": 0.0001295624266229992,
+      "loss": 2.5626,
+      "step": 8149
+    },
+    {
+      "epoch": 0.657735453151481,
+      "grad_norm": 0.9044952392578125,
+      "learning_rate": 0.00012954734496346704,
+      "loss": 2.6479,
+      "step": 8150
+    },
+    {
+      "epoch": 0.6578161568880639,
+      "grad_norm": 0.6955156922340393,
+      "learning_rate": 0.00012953226256753777,
+      "loss": 2.5879,
+      "step": 8151
+    },
+    {
+      "epoch": 0.6578968606246469,
+      "grad_norm": 0.6535033583641052,
+      "learning_rate": 0.00012951717943558735,
+      "loss": 2.5372,
+      "step": 8152
+    },
+    {
+      "epoch": 0.6579775643612299,
+      "grad_norm": 0.720730721950531,
+      "learning_rate": 0.0001295020955679916,
+      "loss": 2.5813,
+      "step": 8153
+    },
+    {
+      "epoch": 0.658058268097813,
+      "grad_norm": 0.7190384268760681,
+      "learning_rate": 0.00012948701096512655,
+      "loss": 2.5923,
+      "step": 8154
+    },
+    {
+      "epoch": 0.658138971834396,
+      "grad_norm": 0.6624464988708496,
+      "learning_rate": 0.0001294719256273681,
+      "loss": 2.5548,
+      "step": 8155
+    },
+    {
+      "epoch": 0.6582196755709789,
+      "grad_norm": 0.7839831709861755,
+      "learning_rate": 0.00012945683955509224,
+      "loss": 2.531,
+      "step": 8156
+    },
+    {
+      "epoch": 0.6583003793075619,
+      "grad_norm": 0.694970965385437,
+      "learning_rate": 0.00012944175274867497,
+      "loss": 2.4693,
+      "step": 8157
+    },
+    {
+      "epoch": 0.658381083044145,
+      "grad_norm": 0.7409366965293884,
+      "learning_rate": 0.0001294266652084922,
+      "loss": 2.5706,
+      "step": 8158
+    },
+    {
+      "epoch": 0.658461786780728,
+      "grad_norm": 0.7502163052558899,
+      "learning_rate": 0.00012941157693492002,
+      "loss": 2.6137,
+      "step": 8159
+    },
+    {
+      "epoch": 0.658542490517311,
+      "grad_norm": 0.6627129912376404,
+      "learning_rate": 0.00012939648792833447,
+      "loss": 2.5781,
+      "step": 8160
+    },
+    {
+      "epoch": 0.6586231942538939,
+      "grad_norm": 0.6775660514831543,
+      "learning_rate": 0.00012938139818911157,
+      "loss": 2.5441,
+      "step": 8161
+    },
+    {
+      "epoch": 0.6587038979904769,
+      "grad_norm": 0.7150553464889526,
+      "learning_rate": 0.00012936630771762748,
+      "loss": 2.5763,
+      "step": 8162
+    },
+    {
+      "epoch": 0.65878460172706,
+      "grad_norm": 0.7461466193199158,
+      "learning_rate": 0.0001293512165142582,
+      "loss": 2.54,
+      "step": 8163
+    },
+    {
+      "epoch": 0.658865305463643,
+      "grad_norm": 0.7635199427604675,
+      "learning_rate": 0.00012933612457937988,
+      "loss": 2.5763,
+      "step": 8164
+    },
+    {
+      "epoch": 0.658946009200226,
+      "grad_norm": 0.7360543608665466,
+      "learning_rate": 0.00012932103191336865,
+      "loss": 2.5968,
+      "step": 8165
+    },
+    {
+      "epoch": 0.6590267129368089,
+      "grad_norm": 0.6482167840003967,
+      "learning_rate": 0.0001293059385166007,
+      "loss": 2.5704,
+      "step": 8166
+    },
+    {
+      "epoch": 0.659107416673392,
+      "grad_norm": 0.7024737596511841,
+      "learning_rate": 0.00012929084438945208,
+      "loss": 2.6221,
+      "step": 8167
+    },
+    {
+      "epoch": 0.659188120409975,
+      "grad_norm": 0.7192068696022034,
+      "learning_rate": 0.0001292757495322991,
+      "loss": 2.5574,
+      "step": 8168
+    },
+    {
+      "epoch": 0.659268824146558,
+      "grad_norm": 0.6900508403778076,
+      "learning_rate": 0.0001292606539455179,
+      "loss": 2.5969,
+      "step": 8169
+    },
+    {
+      "epoch": 0.6593495278831409,
+      "grad_norm": 0.7522475719451904,
+      "learning_rate": 0.00012924555762948474,
+      "loss": 2.592,
+      "step": 8170
+    },
+    {
+      "epoch": 0.659430231619724,
+      "grad_norm": 0.6610947251319885,
+      "learning_rate": 0.00012923046058457583,
+      "loss": 2.5404,
+      "step": 8171
+    },
+    {
+      "epoch": 0.659510935356307,
+      "grad_norm": 0.667628288269043,
+      "learning_rate": 0.00012921536281116738,
+      "loss": 2.5551,
+      "step": 8172
+    },
+    {
+      "epoch": 0.65959163909289,
+      "grad_norm": 0.7119980454444885,
+      "learning_rate": 0.00012920026430963578,
+      "loss": 2.6002,
+      "step": 8173
+    },
+    {
+      "epoch": 0.659672342829473,
+      "grad_norm": 0.712166428565979,
+      "learning_rate": 0.00012918516508035724,
+      "loss": 2.626,
+      "step": 8174
+    },
+    {
+      "epoch": 0.659753046566056,
+      "grad_norm": 0.6993290185928345,
+      "learning_rate": 0.0001291700651237081,
+      "loss": 2.6311,
+      "step": 8175
+    },
+    {
+      "epoch": 0.659833750302639,
+      "grad_norm": 0.6889405250549316,
+      "learning_rate": 0.0001291549644400647,
+      "loss": 2.6483,
+      "step": 8176
+    },
+    {
+      "epoch": 0.659914454039222,
+      "grad_norm": 0.7120937705039978,
+      "learning_rate": 0.00012913986302980334,
+      "loss": 2.5489,
+      "step": 8177
+    },
+    {
+      "epoch": 0.659995157775805,
+      "grad_norm": 0.7112947106361389,
+      "learning_rate": 0.00012912476089330043,
+      "loss": 2.6393,
+      "step": 8178
+    },
+    {
+      "epoch": 0.6600758615123881,
+      "grad_norm": 0.710342526435852,
+      "learning_rate": 0.00012910965803093237,
+      "loss": 2.5897,
+      "step": 8179
+    },
+    {
+      "epoch": 0.660156565248971,
+      "grad_norm": 0.6506931185722351,
+      "learning_rate": 0.0001290945544430755,
+      "loss": 2.6429,
+      "step": 8180
+    },
+    {
+      "epoch": 0.660237268985554,
+      "grad_norm": 0.7147021293640137,
+      "learning_rate": 0.00012907945013010633,
+      "loss": 2.5521,
+      "step": 8181
+    },
+    {
+      "epoch": 0.660317972722137,
+      "grad_norm": 0.6802387833595276,
+      "learning_rate": 0.0001290643450924012,
+      "loss": 2.581,
+      "step": 8182
+    },
+    {
+      "epoch": 0.6603986764587201,
+      "grad_norm": 0.7599670886993408,
+      "learning_rate": 0.00012904923933033664,
+      "loss": 2.5532,
+      "step": 8183
+    },
+    {
+      "epoch": 0.6604793801953031,
+      "grad_norm": 0.7105657458305359,
+      "learning_rate": 0.0001290341328442891,
+      "loss": 2.5744,
+      "step": 8184
+    },
+    {
+      "epoch": 0.660560083931886,
+      "grad_norm": 0.6786425113677979,
+      "learning_rate": 0.00012901902563463506,
+      "loss": 2.5326,
+      "step": 8185
+    },
+    {
+      "epoch": 0.660640787668469,
+      "grad_norm": 0.7305583357810974,
+      "learning_rate": 0.00012900391770175106,
+      "loss": 2.6103,
+      "step": 8186
+    },
+    {
+      "epoch": 0.6607214914050521,
+      "grad_norm": 0.6578992605209351,
+      "learning_rate": 0.00012898880904601363,
+      "loss": 2.5833,
+      "step": 8187
+    },
+    {
+      "epoch": 0.6608021951416351,
+      "grad_norm": 0.6498856544494629,
+      "learning_rate": 0.00012897369966779926,
+      "loss": 2.6333,
+      "step": 8188
+    },
+    {
+      "epoch": 0.660882898878218,
+      "grad_norm": 0.7065569162368774,
+      "learning_rate": 0.00012895858956748458,
+      "loss": 2.5326,
+      "step": 8189
+    },
+    {
+      "epoch": 0.660963602614801,
+      "grad_norm": 0.7676446437835693,
+      "learning_rate": 0.00012894347874544613,
+      "loss": 2.6233,
+      "step": 8190
+    },
+    {
+      "epoch": 0.6610443063513841,
+      "grad_norm": 0.6794395446777344,
+      "learning_rate": 0.00012892836720206056,
+      "loss": 2.5426,
+      "step": 8191
+    },
+    {
+      "epoch": 0.6611250100879671,
+      "grad_norm": 0.7448986768722534,
+      "learning_rate": 0.00012891325493770444,
+      "loss": 2.5832,
+      "step": 8192
+    },
+    {
+      "epoch": 0.6612057138245501,
+      "grad_norm": 0.7789760231971741,
+      "learning_rate": 0.0001288981419527544,
+      "loss": 2.6393,
+      "step": 8193
+    },
+    {
+      "epoch": 0.661286417561133,
+      "grad_norm": 0.7425827980041504,
+      "learning_rate": 0.00012888302824758718,
+      "loss": 2.6159,
+      "step": 8194
+    },
+    {
+      "epoch": 0.6613671212977161,
+      "grad_norm": 0.6677481532096863,
+      "learning_rate": 0.00012886791382257936,
+      "loss": 2.5399,
+      "step": 8195
+    },
+    {
+      "epoch": 0.6614478250342991,
+      "grad_norm": 0.698397159576416,
+      "learning_rate": 0.0001288527986781077,
+      "loss": 2.5443,
+      "step": 8196
+    },
+    {
+      "epoch": 0.6615285287708821,
+      "grad_norm": 0.6862680315971375,
+      "learning_rate": 0.00012883768281454885,
+      "loss": 2.5843,
+      "step": 8197
+    },
+    {
+      "epoch": 0.6616092325074651,
+      "grad_norm": 0.7421948313713074,
+      "learning_rate": 0.00012882256623227955,
+      "loss": 2.5885,
+      "step": 8198
+    },
+    {
+      "epoch": 0.6616899362440481,
+      "grad_norm": 0.7453073859214783,
+      "learning_rate": 0.00012880744893167654,
+      "loss": 2.5821,
+      "step": 8199
+    },
+    {
+      "epoch": 0.6617706399806311,
+      "grad_norm": 0.668218195438385,
+      "learning_rate": 0.00012879233091311667,
+      "loss": 2.5941,
+      "step": 8200
+    },
+    {
+      "epoch": 0.6618513437172141,
+      "grad_norm": 0.6864587664604187,
+      "learning_rate": 0.00012877721217697657,
+      "loss": 2.5321,
+      "step": 8201
+    },
+    {
+      "epoch": 0.6619320474537971,
+      "grad_norm": 0.6521022319793701,
+      "learning_rate": 0.00012876209272363317,
+      "loss": 2.5945,
+      "step": 8202
+    },
+    {
+      "epoch": 0.6620127511903802,
+      "grad_norm": 0.7564631104469299,
+      "learning_rate": 0.00012874697255346325,
+      "loss": 2.5901,
+      "step": 8203
+    },
+    {
+      "epoch": 0.6620934549269631,
+      "grad_norm": 0.731991171836853,
+      "learning_rate": 0.00012873185166684356,
+      "loss": 2.649,
+      "step": 8204
+    },
+    {
+      "epoch": 0.6621741586635461,
+      "grad_norm": 0.6804815530776978,
+      "learning_rate": 0.00012871673006415108,
+      "loss": 2.5417,
+      "step": 8205
+    },
+    {
+      "epoch": 0.6622548624001291,
+      "grad_norm": 0.6862792372703552,
+      "learning_rate": 0.0001287016077457626,
+      "loss": 2.6118,
+      "step": 8206
+    },
+    {
+      "epoch": 0.6623355661367122,
+      "grad_norm": 0.7013735175132751,
+      "learning_rate": 0.00012868648471205503,
+      "loss": 2.6296,
+      "step": 8207
+    },
+    {
+      "epoch": 0.6624162698732952,
+      "grad_norm": 0.7284584045410156,
+      "learning_rate": 0.00012867136096340529,
+      "loss": 2.6547,
+      "step": 8208
+    },
+    {
+      "epoch": 0.6624969736098781,
+      "grad_norm": 0.714546799659729,
+      "learning_rate": 0.00012865623650019025,
+      "loss": 2.5955,
+      "step": 8209
+    },
+    {
+      "epoch": 0.6625776773464611,
+      "grad_norm": 0.7645453214645386,
+      "learning_rate": 0.0001286411113227869,
+      "loss": 2.6132,
+      "step": 8210
+    },
+    {
+      "epoch": 0.6626583810830441,
+      "grad_norm": 0.6615093946456909,
+      "learning_rate": 0.0001286259854315722,
+      "loss": 2.5701,
+      "step": 8211
+    },
+    {
+      "epoch": 0.6627390848196272,
+      "grad_norm": 0.6565523147583008,
+      "learning_rate": 0.0001286108588269231,
+      "loss": 2.57,
+      "step": 8212
+    },
+    {
+      "epoch": 0.6628197885562102,
+      "grad_norm": 0.7173478007316589,
+      "learning_rate": 0.00012859573150921666,
+      "loss": 2.589,
+      "step": 8213
+    },
+    {
+      "epoch": 0.6629004922927931,
+      "grad_norm": 0.7069580554962158,
+      "learning_rate": 0.00012858060347882975,
+      "loss": 2.6146,
+      "step": 8214
+    },
+    {
+      "epoch": 0.6629811960293761,
+      "grad_norm": 0.7004678249359131,
+      "learning_rate": 0.00012856547473613953,
+      "loss": 2.5735,
+      "step": 8215
+    },
+    {
+      "epoch": 0.6630618997659592,
+      "grad_norm": 0.6589130163192749,
+      "learning_rate": 0.00012855034528152305,
+      "loss": 2.5731,
+      "step": 8216
+    },
+    {
+      "epoch": 0.6631426035025422,
+      "grad_norm": 0.7223117351531982,
+      "learning_rate": 0.0001285352151153573,
+      "loss": 2.5262,
+      "step": 8217
+    },
+    {
+      "epoch": 0.6632233072391251,
+      "grad_norm": 0.7045131325721741,
+      "learning_rate": 0.0001285200842380194,
+      "loss": 2.5789,
+      "step": 8218
+    },
+    {
+      "epoch": 0.6633040109757081,
+      "grad_norm": 0.7002174854278564,
+      "learning_rate": 0.00012850495264988645,
+      "loss": 2.6386,
+      "step": 8219
+    },
+    {
+      "epoch": 0.6633847147122912,
+      "grad_norm": 0.6844584941864014,
+      "learning_rate": 0.00012848982035133555,
+      "loss": 2.5394,
+      "step": 8220
+    },
+    {
+      "epoch": 0.6634654184488742,
+      "grad_norm": 0.7154871821403503,
+      "learning_rate": 0.00012847468734274387,
+      "loss": 2.5927,
+      "step": 8221
+    },
+    {
+      "epoch": 0.6635461221854572,
+      "grad_norm": 0.6856776475906372,
+      "learning_rate": 0.00012845955362448855,
+      "loss": 2.5694,
+      "step": 8222
+    },
+    {
+      "epoch": 0.6636268259220401,
+      "grad_norm": 0.7069089412689209,
+      "learning_rate": 0.00012844441919694676,
+      "loss": 2.5856,
+      "step": 8223
+    },
+    {
+      "epoch": 0.6637075296586232,
+      "grad_norm": 0.7084143161773682,
+      "learning_rate": 0.00012842928406049567,
+      "loss": 2.6301,
+      "step": 8224
+    },
+    {
+      "epoch": 0.6637882333952062,
+      "grad_norm": 0.6790862679481506,
+      "learning_rate": 0.00012841414821551252,
+      "loss": 2.5586,
+      "step": 8225
+    },
+    {
+      "epoch": 0.6638689371317892,
+      "grad_norm": 0.6537249684333801,
+      "learning_rate": 0.00012839901166237453,
+      "loss": 2.5652,
+      "step": 8226
+    },
+    {
+      "epoch": 0.6639496408683722,
+      "grad_norm": 0.6670125126838684,
+      "learning_rate": 0.00012838387440145893,
+      "loss": 2.5438,
+      "step": 8227
+    },
+    {
+      "epoch": 0.6640303446049552,
+      "grad_norm": 0.7202955484390259,
+      "learning_rate": 0.00012836873643314297,
+      "loss": 2.5632,
+      "step": 8228
+    },
+    {
+      "epoch": 0.6641110483415382,
+      "grad_norm": 0.6844765543937683,
+      "learning_rate": 0.00012835359775780394,
+      "loss": 2.5595,
+      "step": 8229
+    },
+    {
+      "epoch": 0.6641917520781212,
+      "grad_norm": 0.6557698249816895,
+      "learning_rate": 0.00012833845837581916,
+      "loss": 2.5998,
+      "step": 8230
+    },
+    {
+      "epoch": 0.6642724558147042,
+      "grad_norm": 0.6741784811019897,
+      "learning_rate": 0.0001283233182875659,
+      "loss": 2.5591,
+      "step": 8231
+    },
+    {
+      "epoch": 0.6643531595512873,
+      "grad_norm": 0.6926484704017639,
+      "learning_rate": 0.00012830817749342154,
+      "loss": 2.5557,
+      "step": 8232
+    },
+    {
+      "epoch": 0.6644338632878702,
+      "grad_norm": 0.6866984367370605,
+      "learning_rate": 0.00012829303599376336,
+      "loss": 2.5646,
+      "step": 8233
+    },
+    {
+      "epoch": 0.6645145670244532,
+      "grad_norm": 0.6772707104682922,
+      "learning_rate": 0.0001282778937889688,
+      "loss": 2.6028,
+      "step": 8234
+    },
+    {
+      "epoch": 0.6645952707610362,
+      "grad_norm": 0.693236768245697,
+      "learning_rate": 0.00012826275087941518,
+      "loss": 2.611,
+      "step": 8235
+    },
+    {
+      "epoch": 0.6646759744976193,
+      "grad_norm": 0.7181996703147888,
+      "learning_rate": 0.00012824760726547993,
+      "loss": 2.6081,
+      "step": 8236
+    },
+    {
+      "epoch": 0.6647566782342023,
+      "grad_norm": 0.6845484375953674,
+      "learning_rate": 0.00012823246294754048,
+      "loss": 2.5544,
+      "step": 8237
+    },
+    {
+      "epoch": 0.6648373819707852,
+      "grad_norm": 0.7106444239616394,
+      "learning_rate": 0.00012821731792597425,
+      "loss": 2.552,
+      "step": 8238
+    },
+    {
+      "epoch": 0.6649180857073682,
+      "grad_norm": 0.6930601000785828,
+      "learning_rate": 0.0001282021722011587,
+      "loss": 2.5401,
+      "step": 8239
+    },
+    {
+      "epoch": 0.6649987894439513,
+      "grad_norm": 0.6658228039741516,
+      "learning_rate": 0.00012818702577347129,
+      "loss": 2.6287,
+      "step": 8240
+    },
+    {
+      "epoch": 0.6650794931805343,
+      "grad_norm": 0.6919803619384766,
+      "learning_rate": 0.0001281718786432895,
+      "loss": 2.6142,
+      "step": 8241
+    },
+    {
+      "epoch": 0.6651601969171173,
+      "grad_norm": 0.6675698757171631,
+      "learning_rate": 0.00012815673081099086,
+      "loss": 2.5325,
+      "step": 8242
+    },
+    {
+      "epoch": 0.6652409006537002,
+      "grad_norm": 0.6669798493385315,
+      "learning_rate": 0.0001281415822769529,
+      "loss": 2.5355,
+      "step": 8243
+    },
+    {
+      "epoch": 0.6653216043902833,
+      "grad_norm": 0.6449857950210571,
+      "learning_rate": 0.00012812643304155316,
+      "loss": 2.5968,
+      "step": 8244
+    },
+    {
+      "epoch": 0.6654023081268663,
+      "grad_norm": 0.6972789168357849,
+      "learning_rate": 0.00012811128310516914,
+      "loss": 2.6133,
+      "step": 8245
+    },
+    {
+      "epoch": 0.6654830118634493,
+      "grad_norm": 0.7179878354072571,
+      "learning_rate": 0.0001280961324681785,
+      "loss": 2.5793,
+      "step": 8246
+    },
+    {
+      "epoch": 0.6655637156000322,
+      "grad_norm": 0.6736378073692322,
+      "learning_rate": 0.0001280809811309588,
+      "loss": 2.5543,
+      "step": 8247
+    },
+    {
+      "epoch": 0.6656444193366153,
+      "grad_norm": 0.7376420497894287,
+      "learning_rate": 0.00012806582909388763,
+      "loss": 2.5501,
+      "step": 8248
+    },
+    {
+      "epoch": 0.6657251230731983,
+      "grad_norm": 0.7163094878196716,
+      "learning_rate": 0.00012805067635734263,
+      "loss": 2.5538,
+      "step": 8249
+    },
+    {
+      "epoch": 0.6658058268097813,
+      "grad_norm": 0.7699353694915771,
+      "learning_rate": 0.00012803552292170144,
+      "loss": 2.5925,
+      "step": 8250
+    },
+    {
+      "epoch": 0.6658865305463643,
+      "grad_norm": 0.6504995822906494,
+      "learning_rate": 0.00012802036878734177,
+      "loss": 2.5944,
+      "step": 8251
+    },
+    {
+      "epoch": 0.6659672342829474,
+      "grad_norm": 0.7150379419326782,
+      "learning_rate": 0.0001280052139546412,
+      "loss": 2.5959,
+      "step": 8252
+    },
+    {
+      "epoch": 0.6660479380195303,
+      "grad_norm": 0.7562555074691772,
+      "learning_rate": 0.00012799005842397757,
+      "loss": 2.6041,
+      "step": 8253
+    },
+    {
+      "epoch": 0.6661286417561133,
+      "grad_norm": 0.7242838740348816,
+      "learning_rate": 0.00012797490219572846,
+      "loss": 2.6152,
+      "step": 8254
+    },
+    {
+      "epoch": 0.6662093454926963,
+      "grad_norm": 0.7062848210334778,
+      "learning_rate": 0.00012795974527027168,
+      "loss": 2.596,
+      "step": 8255
+    },
+    {
+      "epoch": 0.6662900492292794,
+      "grad_norm": 0.8179726004600525,
+      "learning_rate": 0.00012794458764798497,
+      "loss": 2.5792,
+      "step": 8256
+    },
+    {
+      "epoch": 0.6663707529658623,
+      "grad_norm": 0.692166268825531,
+      "learning_rate": 0.00012792942932924608,
+      "loss": 2.6025,
+      "step": 8257
+    },
+    {
+      "epoch": 0.6664514567024453,
+      "grad_norm": 0.6540334224700928,
+      "learning_rate": 0.0001279142703144328,
+      "loss": 2.5119,
+      "step": 8258
+    },
+    {
+      "epoch": 0.6665321604390283,
+      "grad_norm": 0.7087461352348328,
+      "learning_rate": 0.00012789911060392294,
+      "loss": 2.5808,
+      "step": 8259
+    },
+    {
+      "epoch": 0.6666128641756114,
+      "grad_norm": 0.6897622346878052,
+      "learning_rate": 0.0001278839501980943,
+      "loss": 2.5811,
+      "step": 8260
+    },
+    {
+      "epoch": 0.6666935679121944,
+      "grad_norm": 0.6653634905815125,
+      "learning_rate": 0.00012786878909732473,
+      "loss": 2.5498,
+      "step": 8261
+    },
+    {
+      "epoch": 0.6667742716487773,
+      "grad_norm": 0.6541483402252197,
+      "learning_rate": 0.0001278536273019921,
+      "loss": 2.605,
+      "step": 8262
+    },
+    {
+      "epoch": 0.6668549753853603,
+      "grad_norm": 0.6748146414756775,
+      "learning_rate": 0.00012783846481247428,
+      "loss": 2.5571,
+      "step": 8263
+    },
+    {
+      "epoch": 0.6669356791219433,
+      "grad_norm": 0.7258282899856567,
+      "learning_rate": 0.00012782330162914915,
+      "loss": 2.5562,
+      "step": 8264
+    },
+    {
+      "epoch": 0.6670163828585264,
+      "grad_norm": 0.6963080167770386,
+      "learning_rate": 0.00012780813775239457,
+      "loss": 2.6467,
+      "step": 8265
+    },
+    {
+      "epoch": 0.6670970865951094,
+      "grad_norm": 0.6627718806266785,
+      "learning_rate": 0.00012779297318258855,
+      "loss": 2.5369,
+      "step": 8266
+    },
+    {
+      "epoch": 0.6671777903316923,
+      "grad_norm": 0.7026168704032898,
+      "learning_rate": 0.00012777780792010897,
+      "loss": 2.5639,
+      "step": 8267
+    },
+    {
+      "epoch": 0.6672584940682753,
+      "grad_norm": 0.6969077587127686,
+      "learning_rate": 0.0001277626419653338,
+      "loss": 2.517,
+      "step": 8268
+    },
+    {
+      "epoch": 0.6673391978048584,
+      "grad_norm": 0.6918485760688782,
+      "learning_rate": 0.00012774747531864102,
+      "loss": 2.6388,
+      "step": 8269
+    },
+    {
+      "epoch": 0.6674199015414414,
+      "grad_norm": 0.6661256551742554,
+      "learning_rate": 0.00012773230798040862,
+      "loss": 2.5477,
+      "step": 8270
+    },
+    {
+      "epoch": 0.6675006052780244,
+      "grad_norm": 0.6778402328491211,
+      "learning_rate": 0.0001277171399510146,
+      "loss": 2.6032,
+      "step": 8271
+    },
+    {
+      "epoch": 0.6675813090146073,
+      "grad_norm": 0.6464864611625671,
+      "learning_rate": 0.00012770197123083702,
+      "loss": 2.5396,
+      "step": 8272
+    },
+    {
+      "epoch": 0.6676620127511904,
+      "grad_norm": 0.7154508233070374,
+      "learning_rate": 0.0001276868018202539,
+      "loss": 2.6163,
+      "step": 8273
+    },
+    {
+      "epoch": 0.6677427164877734,
+      "grad_norm": 0.6849631071090698,
+      "learning_rate": 0.0001276716317196433,
+      "loss": 2.549,
+      "step": 8274
+    },
+    {
+      "epoch": 0.6678234202243564,
+      "grad_norm": 0.6696017980575562,
+      "learning_rate": 0.00012765646092938334,
+      "loss": 2.5046,
+      "step": 8275
+    },
+    {
+      "epoch": 0.6679041239609393,
+      "grad_norm": 0.668153703212738,
+      "learning_rate": 0.00012764128944985203,
+      "loss": 2.5422,
+      "step": 8276
+    },
+    {
+      "epoch": 0.6679848276975224,
+      "grad_norm": 0.6600282192230225,
+      "learning_rate": 0.00012762611728142756,
+      "loss": 2.6117,
+      "step": 8277
+    },
+    {
+      "epoch": 0.6680655314341054,
+      "grad_norm": 0.6691608428955078,
+      "learning_rate": 0.000127610944424488,
+      "loss": 2.5761,
+      "step": 8278
+    },
+    {
+      "epoch": 0.6681462351706884,
+      "grad_norm": 0.695142924785614,
+      "learning_rate": 0.00012759577087941156,
+      "loss": 2.6123,
+      "step": 8279
+    },
+    {
+      "epoch": 0.6682269389072714,
+      "grad_norm": 0.6846559643745422,
+      "learning_rate": 0.00012758059664657635,
+      "loss": 2.5882,
+      "step": 8280
+    },
+    {
+      "epoch": 0.6683076426438544,
+      "grad_norm": 0.7616459131240845,
+      "learning_rate": 0.0001275654217263606,
+      "loss": 2.5559,
+      "step": 8281
+    },
+    {
+      "epoch": 0.6683883463804374,
+      "grad_norm": 0.6995570063591003,
+      "learning_rate": 0.00012755024611914246,
+      "loss": 2.5336,
+      "step": 8282
+    },
+    {
+      "epoch": 0.6684690501170204,
+      "grad_norm": 0.7199691534042358,
+      "learning_rate": 0.0001275350698253002,
+      "loss": 2.6618,
+      "step": 8283
+    },
+    {
+      "epoch": 0.6685497538536034,
+      "grad_norm": 0.6938748955726624,
+      "learning_rate": 0.000127519892845212,
+      "loss": 2.574,
+      "step": 8284
+    },
+    {
+      "epoch": 0.6686304575901865,
+      "grad_norm": 0.6827714443206787,
+      "learning_rate": 0.00012750471517925614,
+      "loss": 2.5647,
+      "step": 8285
+    },
+    {
+      "epoch": 0.6687111613267694,
+      "grad_norm": 0.6684606671333313,
+      "learning_rate": 0.00012748953682781083,
+      "loss": 2.528,
+      "step": 8286
+    },
+    {
+      "epoch": 0.6687918650633524,
+      "grad_norm": 0.6842156052589417,
+      "learning_rate": 0.00012747435779125448,
+      "loss": 2.5521,
+      "step": 8287
+    },
+    {
+      "epoch": 0.6688725687999354,
+      "grad_norm": 0.7440506219863892,
+      "learning_rate": 0.0001274591780699653,
+      "loss": 2.5646,
+      "step": 8288
+    },
+    {
+      "epoch": 0.6689532725365185,
+      "grad_norm": 0.769922137260437,
+      "learning_rate": 0.0001274439976643216,
+      "loss": 2.6104,
+      "step": 8289
+    },
+    {
+      "epoch": 0.6690339762731015,
+      "grad_norm": 0.7793089747428894,
+      "learning_rate": 0.00012742881657470175,
+      "loss": 2.6348,
+      "step": 8290
+    },
+    {
+      "epoch": 0.6691146800096844,
+      "grad_norm": 0.695060133934021,
+      "learning_rate": 0.0001274136348014841,
+      "loss": 2.5797,
+      "step": 8291
+    },
+    {
+      "epoch": 0.6691953837462674,
+      "grad_norm": 0.7089917659759521,
+      "learning_rate": 0.00012739845234504697,
+      "loss": 2.5431,
+      "step": 8292
+    },
+    {
+      "epoch": 0.6692760874828505,
+      "grad_norm": 0.7542717456817627,
+      "learning_rate": 0.00012738326920576885,
+      "loss": 2.6172,
+      "step": 8293
+    },
+    {
+      "epoch": 0.6693567912194335,
+      "grad_norm": 0.6947969794273376,
+      "learning_rate": 0.00012736808538402802,
+      "loss": 2.6026,
+      "step": 8294
+    },
+    {
+      "epoch": 0.6694374949560165,
+      "grad_norm": 0.6696321368217468,
+      "learning_rate": 0.00012735290088020302,
+      "loss": 2.5592,
+      "step": 8295
+    },
+    {
+      "epoch": 0.6695181986925994,
+      "grad_norm": 0.7001518607139587,
+      "learning_rate": 0.0001273377156946722,
+      "loss": 2.5994,
+      "step": 8296
+    },
+    {
+      "epoch": 0.6695989024291825,
+      "grad_norm": 0.6708101630210876,
+      "learning_rate": 0.000127322529827814,
+      "loss": 2.6392,
+      "step": 8297
+    },
+    {
+      "epoch": 0.6696796061657655,
+      "grad_norm": 0.6282601952552795,
+      "learning_rate": 0.000127307343280007,
+      "loss": 2.5762,
+      "step": 8298
+    },
+    {
+      "epoch": 0.6697603099023485,
+      "grad_norm": 0.6879595518112183,
+      "learning_rate": 0.0001272921560516296,
+      "loss": 2.5507,
+      "step": 8299
+    },
+    {
+      "epoch": 0.6698410136389314,
+      "grad_norm": 0.6108266115188599,
+      "learning_rate": 0.00012727696814306033,
+      "loss": 2.5865,
+      "step": 8300
+    },
+    {
+      "epoch": 0.6699217173755145,
+      "grad_norm": 0.6763970851898193,
+      "learning_rate": 0.0001272617795546777,
+      "loss": 2.6439,
+      "step": 8301
+    },
+    {
+      "epoch": 0.6700024211120975,
+      "grad_norm": 0.6997560858726501,
+      "learning_rate": 0.00012724659028686027,
+      "loss": 2.5291,
+      "step": 8302
+    },
+    {
+      "epoch": 0.6700831248486805,
+      "grad_norm": 0.675714910030365,
+      "learning_rate": 0.0001272314003399866,
+      "loss": 2.5452,
+      "step": 8303
+    },
+    {
+      "epoch": 0.6701638285852635,
+      "grad_norm": 0.6847789883613586,
+      "learning_rate": 0.00012721620971443525,
+      "loss": 2.6111,
+      "step": 8304
+    },
+    {
+      "epoch": 0.6702445323218466,
+      "grad_norm": 0.7283920645713806,
+      "learning_rate": 0.0001272010184105848,
+      "loss": 2.6322,
+      "step": 8305
+    },
+    {
+      "epoch": 0.6703252360584295,
+      "grad_norm": 0.7551796436309814,
+      "learning_rate": 0.00012718582642881382,
+      "loss": 2.5728,
+      "step": 8306
+    },
+    {
+      "epoch": 0.6704059397950125,
+      "grad_norm": 0.694526195526123,
+      "learning_rate": 0.00012717063376950104,
+      "loss": 2.6241,
+      "step": 8307
+    },
+    {
+      "epoch": 0.6704866435315955,
+      "grad_norm": 0.6956443190574646,
+      "learning_rate": 0.00012715544043302504,
+      "loss": 2.5531,
+      "step": 8308
+    },
+    {
+      "epoch": 0.6705673472681786,
+      "grad_norm": 0.7649452686309814,
+      "learning_rate": 0.00012714024641976446,
+      "loss": 2.5462,
+      "step": 8309
+    },
+    {
+      "epoch": 0.6706480510047615,
+      "grad_norm": 0.7711065411567688,
+      "learning_rate": 0.00012712505173009797,
+      "loss": 2.5878,
+      "step": 8310
+    },
+    {
+      "epoch": 0.6707287547413445,
+      "grad_norm": 0.68077552318573,
+      "learning_rate": 0.00012710985636440434,
+      "loss": 2.5668,
+      "step": 8311
+    },
+    {
+      "epoch": 0.6708094584779275,
+      "grad_norm": 0.7181024551391602,
+      "learning_rate": 0.0001270946603230622,
+      "loss": 2.6104,
+      "step": 8312
+    },
+    {
+      "epoch": 0.6708901622145105,
+      "grad_norm": 0.7136553525924683,
+      "learning_rate": 0.0001270794636064503,
+      "loss": 2.5282,
+      "step": 8313
+    },
+    {
+      "epoch": 0.6709708659510936,
+      "grad_norm": 0.880094587802887,
+      "learning_rate": 0.00012706426621494736,
+      "loss": 2.5837,
+      "step": 8314
+    },
+    {
+      "epoch": 0.6710515696876765,
+      "grad_norm": 0.7438541054725647,
+      "learning_rate": 0.00012704906814893217,
+      "loss": 2.5577,
+      "step": 8315
+    },
+    {
+      "epoch": 0.6711322734242595,
+      "grad_norm": 0.8197470903396606,
+      "learning_rate": 0.00012703386940878352,
+      "loss": 2.569,
+      "step": 8316
+    },
+    {
+      "epoch": 0.6712129771608425,
+      "grad_norm": 0.7728317975997925,
+      "learning_rate": 0.00012701866999488014,
+      "loss": 2.6407,
+      "step": 8317
+    },
+    {
+      "epoch": 0.6712936808974256,
+      "grad_norm": 0.7594823837280273,
+      "learning_rate": 0.0001270034699076009,
+      "loss": 2.5789,
+      "step": 8318
+    },
+    {
+      "epoch": 0.6713743846340086,
+      "grad_norm": 0.7502284646034241,
+      "learning_rate": 0.0001269882691473246,
+      "loss": 2.6068,
+      "step": 8319
+    },
+    {
+      "epoch": 0.6714550883705915,
+      "grad_norm": 0.7355664372444153,
+      "learning_rate": 0.0001269730677144301,
+      "loss": 2.6055,
+      "step": 8320
+    },
+    {
+      "epoch": 0.6715357921071745,
+      "grad_norm": 0.7218407392501831,
+      "learning_rate": 0.0001269578656092962,
+      "loss": 2.5953,
+      "step": 8321
+    },
+    {
+      "epoch": 0.6716164958437576,
+      "grad_norm": 0.6932538747787476,
+      "learning_rate": 0.00012694266283230185,
+      "loss": 2.5795,
+      "step": 8322
+    },
+    {
+      "epoch": 0.6716971995803406,
+      "grad_norm": 0.7337260246276855,
+      "learning_rate": 0.00012692745938382591,
+      "loss": 2.5606,
+      "step": 8323
+    },
+    {
+      "epoch": 0.6717779033169236,
+      "grad_norm": 0.6959026455879211,
+      "learning_rate": 0.00012691225526424731,
+      "loss": 2.5688,
+      "step": 8324
+    },
+    {
+      "epoch": 0.6718586070535065,
+      "grad_norm": 0.7352995872497559,
+      "learning_rate": 0.00012689705047394493,
+      "loss": 2.6308,
+      "step": 8325
+    },
+    {
+      "epoch": 0.6719393107900896,
+      "grad_norm": 0.7023616433143616,
+      "learning_rate": 0.00012688184501329777,
+      "loss": 2.6462,
+      "step": 8326
+    },
+    {
+      "epoch": 0.6720200145266726,
+      "grad_norm": 0.6581354737281799,
+      "learning_rate": 0.00012686663888268474,
+      "loss": 2.5997,
+      "step": 8327
+    },
+    {
+      "epoch": 0.6721007182632556,
+      "grad_norm": 0.6332606077194214,
+      "learning_rate": 0.00012685143208248484,
+      "loss": 2.6348,
+      "step": 8328
+    },
+    {
+      "epoch": 0.6721814219998385,
+      "grad_norm": 0.6826457977294922,
+      "learning_rate": 0.00012683622461307707,
+      "loss": 2.5092,
+      "step": 8329
+    },
+    {
+      "epoch": 0.6722621257364216,
+      "grad_norm": 0.7641614079475403,
+      "learning_rate": 0.00012682101647484042,
+      "loss": 2.7098,
+      "step": 8330
+    },
+    {
+      "epoch": 0.6723428294730046,
+      "grad_norm": 0.7153630256652832,
+      "learning_rate": 0.00012680580766815394,
+      "loss": 2.5647,
+      "step": 8331
+    },
+    {
+      "epoch": 0.6724235332095876,
+      "grad_norm": 0.6746379137039185,
+      "learning_rate": 0.00012679059819339664,
+      "loss": 2.6187,
+      "step": 8332
+    },
+    {
+      "epoch": 0.6725042369461706,
+      "grad_norm": 0.6748883128166199,
+      "learning_rate": 0.00012677538805094764,
+      "loss": 2.6045,
+      "step": 8333
+    },
+    {
+      "epoch": 0.6725849406827537,
+      "grad_norm": 0.7366370558738708,
+      "learning_rate": 0.00012676017724118596,
+      "loss": 2.5789,
+      "step": 8334
+    },
+    {
+      "epoch": 0.6726656444193366,
+      "grad_norm": 0.7381749153137207,
+      "learning_rate": 0.00012674496576449074,
+      "loss": 2.5958,
+      "step": 8335
+    },
+    {
+      "epoch": 0.6727463481559196,
+      "grad_norm": 0.7109243869781494,
+      "learning_rate": 0.00012672975362124103,
+      "loss": 2.5874,
+      "step": 8336
+    },
+    {
+      "epoch": 0.6728270518925026,
+      "grad_norm": 0.6904270052909851,
+      "learning_rate": 0.00012671454081181595,
+      "loss": 2.5891,
+      "step": 8337
+    },
+    {
+      "epoch": 0.6729077556290857,
+      "grad_norm": 0.6809365749359131,
+      "learning_rate": 0.00012669932733659476,
+      "loss": 2.5904,
+      "step": 8338
+    },
+    {
+      "epoch": 0.6729884593656686,
+      "grad_norm": 0.7527552843093872,
+      "learning_rate": 0.00012668411319595647,
+      "loss": 2.5602,
+      "step": 8339
+    },
+    {
+      "epoch": 0.6730691631022516,
+      "grad_norm": 0.6746577620506287,
+      "learning_rate": 0.00012666889839028038,
+      "loss": 2.5468,
+      "step": 8340
+    },
+    {
+      "epoch": 0.6731498668388346,
+      "grad_norm": 0.6904895305633545,
+      "learning_rate": 0.00012665368291994562,
+      "loss": 2.623,
+      "step": 8341
+    },
+    {
+      "epoch": 0.6732305705754177,
+      "grad_norm": 0.6495908498764038,
+      "learning_rate": 0.00012663846678533135,
+      "loss": 2.5843,
+      "step": 8342
+    },
+    {
+      "epoch": 0.6733112743120007,
+      "grad_norm": 0.6782342195510864,
+      "learning_rate": 0.00012662324998681692,
+      "loss": 2.6141,
+      "step": 8343
+    },
+    {
+      "epoch": 0.6733919780485836,
+      "grad_norm": 0.7090504765510559,
+      "learning_rate": 0.0001266080325247815,
+      "loss": 2.6654,
+      "step": 8344
+    },
+    {
+      "epoch": 0.6734726817851666,
+      "grad_norm": 0.7085515856742859,
+      "learning_rate": 0.00012659281439960434,
+      "loss": 2.5394,
+      "step": 8345
+    },
+    {
+      "epoch": 0.6735533855217497,
+      "grad_norm": 0.6813806295394897,
+      "learning_rate": 0.00012657759561166473,
+      "loss": 2.6522,
+      "step": 8346
+    },
+    {
+      "epoch": 0.6736340892583327,
+      "grad_norm": 0.726378858089447,
+      "learning_rate": 0.00012656237616134197,
+      "loss": 2.5922,
+      "step": 8347
+    },
+    {
+      "epoch": 0.6737147929949157,
+      "grad_norm": 0.6323714256286621,
+      "learning_rate": 0.00012654715604901534,
+      "loss": 2.4938,
+      "step": 8348
+    },
+    {
+      "epoch": 0.6737954967314986,
+      "grad_norm": 0.6925889253616333,
+      "learning_rate": 0.0001265319352750642,
+      "loss": 2.635,
+      "step": 8349
+    },
+    {
+      "epoch": 0.6738762004680817,
+      "grad_norm": 0.6676003932952881,
+      "learning_rate": 0.00012651671383986788,
+      "loss": 2.558,
+      "step": 8350
+    },
+    {
+      "epoch": 0.6739569042046647,
+      "grad_norm": 0.7464616298675537,
+      "learning_rate": 0.00012650149174380575,
+      "loss": 2.5777,
+      "step": 8351
+    },
+    {
+      "epoch": 0.6740376079412477,
+      "grad_norm": 0.6611667275428772,
+      "learning_rate": 0.00012648626898725715,
+      "loss": 2.5779,
+      "step": 8352
+    },
+    {
+      "epoch": 0.6741183116778307,
+      "grad_norm": 0.7391866445541382,
+      "learning_rate": 0.00012647104557060148,
+      "loss": 2.5624,
+      "step": 8353
+    },
+    {
+      "epoch": 0.6741990154144137,
+      "grad_norm": 0.7107826471328735,
+      "learning_rate": 0.00012645582149421817,
+      "loss": 2.5744,
+      "step": 8354
+    },
+    {
+      "epoch": 0.6742797191509967,
+      "grad_norm": 0.7385339736938477,
+      "learning_rate": 0.00012644059675848666,
+      "loss": 2.5752,
+      "step": 8355
+    },
+    {
+      "epoch": 0.6743604228875797,
+      "grad_norm": 0.6887345314025879,
+      "learning_rate": 0.00012642537136378634,
+      "loss": 2.5794,
+      "step": 8356
+    },
+    {
+      "epoch": 0.6744411266241627,
+      "grad_norm": 0.6934933662414551,
+      "learning_rate": 0.00012641014531049666,
+      "loss": 2.5361,
+      "step": 8357
+    },
+    {
+      "epoch": 0.6745218303607458,
+      "grad_norm": 0.7437291741371155,
+      "learning_rate": 0.00012639491859899716,
+      "loss": 2.5741,
+      "step": 8358
+    },
+    {
+      "epoch": 0.6746025340973287,
+      "grad_norm": 0.7088494896888733,
+      "learning_rate": 0.00012637969122966729,
+      "loss": 2.6449,
+      "step": 8359
+    },
+    {
+      "epoch": 0.6746832378339117,
+      "grad_norm": 0.7496390342712402,
+      "learning_rate": 0.00012636446320288654,
+      "loss": 2.6109,
+      "step": 8360
+    },
+    {
+      "epoch": 0.6747639415704947,
+      "grad_norm": 0.6949843764305115,
+      "learning_rate": 0.00012634923451903447,
+      "loss": 2.5769,
+      "step": 8361
+    },
+    {
+      "epoch": 0.6748446453070778,
+      "grad_norm": 0.7192673087120056,
+      "learning_rate": 0.00012633400517849056,
+      "loss": 2.6053,
+      "step": 8362
+    },
+    {
+      "epoch": 0.6749253490436607,
+      "grad_norm": 0.7003379464149475,
+      "learning_rate": 0.00012631877518163442,
+      "loss": 2.5745,
+      "step": 8363
+    },
+    {
+      "epoch": 0.6750060527802437,
+      "grad_norm": 0.7499879002571106,
+      "learning_rate": 0.00012630354452884563,
+      "loss": 2.6077,
+      "step": 8364
+    },
+    {
+      "epoch": 0.6750867565168267,
+      "grad_norm": 0.7047405242919922,
+      "learning_rate": 0.00012628831322050377,
+      "loss": 2.5955,
+      "step": 8365
+    },
+    {
+      "epoch": 0.6751674602534097,
+      "grad_norm": 0.7463203072547913,
+      "learning_rate": 0.00012627308125698838,
+      "loss": 2.5421,
+      "step": 8366
+    },
+    {
+      "epoch": 0.6752481639899928,
+      "grad_norm": 0.7377086877822876,
+      "learning_rate": 0.00012625784863867914,
+      "loss": 2.5804,
+      "step": 8367
+    },
+    {
+      "epoch": 0.6753288677265757,
+      "grad_norm": 0.7136400938034058,
+      "learning_rate": 0.00012624261536595566,
+      "loss": 2.5673,
+      "step": 8368
+    },
+    {
+      "epoch": 0.6754095714631587,
+      "grad_norm": 0.6923615336418152,
+      "learning_rate": 0.0001262273814391976,
+      "loss": 2.5832,
+      "step": 8369
+    },
+    {
+      "epoch": 0.6754902751997417,
+      "grad_norm": 0.7495028972625732,
+      "learning_rate": 0.00012621214685878469,
+      "loss": 2.5943,
+      "step": 8370
+    },
+    {
+      "epoch": 0.6755709789363248,
+      "grad_norm": 0.6751434206962585,
+      "learning_rate": 0.0001261969116250965,
+      "loss": 2.5495,
+      "step": 8371
+    },
+    {
+      "epoch": 0.6756516826729078,
+      "grad_norm": 0.7055973410606384,
+      "learning_rate": 0.00012618167573851284,
+      "loss": 2.5651,
+      "step": 8372
+    },
+    {
+      "epoch": 0.6757323864094907,
+      "grad_norm": 0.7479640245437622,
+      "learning_rate": 0.00012616643919941337,
+      "loss": 2.653,
+      "step": 8373
+    },
+    {
+      "epoch": 0.6758130901460737,
+      "grad_norm": 0.7075015902519226,
+      "learning_rate": 0.00012615120200817778,
+      "loss": 2.5787,
+      "step": 8374
+    },
+    {
+      "epoch": 0.6758937938826568,
+      "grad_norm": 0.7513934969902039,
+      "learning_rate": 0.00012613596416518593,
+      "loss": 2.6099,
+      "step": 8375
+    },
+    {
+      "epoch": 0.6759744976192398,
+      "grad_norm": 0.6742326021194458,
+      "learning_rate": 0.00012612072567081754,
+      "loss": 2.5335,
+      "step": 8376
+    },
+    {
+      "epoch": 0.6760552013558228,
+      "grad_norm": 0.7271459698677063,
+      "learning_rate": 0.00012610548652545239,
+      "loss": 2.6082,
+      "step": 8377
+    },
+    {
+      "epoch": 0.6761359050924057,
+      "grad_norm": 0.7481515407562256,
+      "learning_rate": 0.00012609024672947022,
+      "loss": 2.5805,
+      "step": 8378
+    },
+    {
+      "epoch": 0.6762166088289888,
+      "grad_norm": 0.7484803199768066,
+      "learning_rate": 0.00012607500628325093,
+      "loss": 2.6099,
+      "step": 8379
+    },
+    {
+      "epoch": 0.6762973125655718,
+      "grad_norm": 0.7462390661239624,
+      "learning_rate": 0.00012605976518717435,
+      "loss": 2.6054,
+      "step": 8380
+    },
+    {
+      "epoch": 0.6763780163021548,
+      "grad_norm": 0.7014410495758057,
+      "learning_rate": 0.00012604452344162028,
+      "loss": 2.5614,
+      "step": 8381
+    },
+    {
+      "epoch": 0.6764587200387377,
+      "grad_norm": 0.6902963519096375,
+      "learning_rate": 0.0001260292810469686,
+      "loss": 2.5813,
+      "step": 8382
+    },
+    {
+      "epoch": 0.6765394237753208,
+      "grad_norm": 0.6646186113357544,
+      "learning_rate": 0.00012601403800359919,
+      "loss": 2.545,
+      "step": 8383
+    },
+    {
+      "epoch": 0.6766201275119038,
+      "grad_norm": 0.7067462801933289,
+      "learning_rate": 0.00012599879431189197,
+      "loss": 2.6195,
+      "step": 8384
+    },
+    {
+      "epoch": 0.6767008312484868,
+      "grad_norm": 0.7263965010643005,
+      "learning_rate": 0.0001259835499722268,
+      "loss": 2.5929,
+      "step": 8385
+    },
+    {
+      "epoch": 0.6767815349850698,
+      "grad_norm": 0.6672000885009766,
+      "learning_rate": 0.0001259683049849837,
+      "loss": 2.5561,
+      "step": 8386
+    },
+    {
+      "epoch": 0.6768622387216529,
+      "grad_norm": 0.6543236374855042,
+      "learning_rate": 0.0001259530593505425,
+      "loss": 2.6256,
+      "step": 8387
+    },
+    {
+      "epoch": 0.6769429424582358,
+      "grad_norm": 0.6532339453697205,
+      "learning_rate": 0.00012593781306928324,
+      "loss": 2.5074,
+      "step": 8388
+    },
+    {
+      "epoch": 0.6770236461948188,
+      "grad_norm": 0.7442833185195923,
+      "learning_rate": 0.00012592256614158591,
+      "loss": 2.6124,
+      "step": 8389
+    },
+    {
+      "epoch": 0.6771043499314018,
+      "grad_norm": 0.786685585975647,
+      "learning_rate": 0.00012590731856783043,
+      "loss": 2.6077,
+      "step": 8390
+    },
+    {
+      "epoch": 0.6771850536679849,
+      "grad_norm": 0.7952337265014648,
+      "learning_rate": 0.00012589207034839687,
+      "loss": 2.5894,
+      "step": 8391
+    },
+    {
+      "epoch": 0.6772657574045678,
+      "grad_norm": 0.7847954034805298,
+      "learning_rate": 0.00012587682148366524,
+      "loss": 2.4934,
+      "step": 8392
+    },
+    {
+      "epoch": 0.6773464611411508,
+      "grad_norm": 0.6769007444381714,
+      "learning_rate": 0.00012586157197401552,
+      "loss": 2.5695,
+      "step": 8393
+    },
+    {
+      "epoch": 0.6774271648777338,
+      "grad_norm": 0.6583757996559143,
+      "learning_rate": 0.00012584632181982788,
+      "loss": 2.5866,
+      "step": 8394
+    },
+    {
+      "epoch": 0.6775078686143169,
+      "grad_norm": 0.7375823855400085,
+      "learning_rate": 0.0001258310710214823,
+      "loss": 2.5141,
+      "step": 8395
+    },
+    {
+      "epoch": 0.6775885723508999,
+      "grad_norm": 0.6901078224182129,
+      "learning_rate": 0.00012581581957935896,
+      "loss": 2.5732,
+      "step": 8396
+    },
+    {
+      "epoch": 0.6776692760874828,
+      "grad_norm": 0.687152624130249,
+      "learning_rate": 0.0001258005674938379,
+      "loss": 2.5916,
+      "step": 8397
+    },
+    {
+      "epoch": 0.6777499798240658,
+      "grad_norm": 0.7198586463928223,
+      "learning_rate": 0.00012578531476529917,
+      "loss": 2.5626,
+      "step": 8398
+    },
+    {
+      "epoch": 0.6778306835606489,
+      "grad_norm": 0.7417474985122681,
+      "learning_rate": 0.00012577006139412309,
+      "loss": 2.5486,
+      "step": 8399
+    },
+    {
+      "epoch": 0.6779113872972319,
+      "grad_norm": 0.6588087677955627,
+      "learning_rate": 0.0001257548073806897,
+      "loss": 2.6123,
+      "step": 8400
+    },
+    {
+      "epoch": 0.6779920910338149,
+      "grad_norm": 0.7211382389068604,
+      "learning_rate": 0.00012573955272537915,
+      "loss": 2.6402,
+      "step": 8401
+    },
+    {
+      "epoch": 0.6780727947703978,
+      "grad_norm": 0.7196084856987,
+      "learning_rate": 0.00012572429742857167,
+      "loss": 2.51,
+      "step": 8402
+    },
+    {
+      "epoch": 0.6781534985069809,
+      "grad_norm": 0.6399394273757935,
+      "learning_rate": 0.00012570904149064748,
+      "loss": 2.5309,
+      "step": 8403
+    },
+    {
+      "epoch": 0.6782342022435639,
+      "grad_norm": 0.6969572305679321,
+      "learning_rate": 0.00012569378491198674,
+      "loss": 2.5829,
+      "step": 8404
+    },
+    {
+      "epoch": 0.6783149059801469,
+      "grad_norm": 0.8005492091178894,
+      "learning_rate": 0.00012567852769296975,
+      "loss": 2.6277,
+      "step": 8405
+    },
+    {
+      "epoch": 0.6783956097167299,
+      "grad_norm": 0.6786207556724548,
+      "learning_rate": 0.0001256632698339767,
+      "loss": 2.5839,
+      "step": 8406
+    },
+    {
+      "epoch": 0.6784763134533129,
+      "grad_norm": 0.7047130465507507,
+      "learning_rate": 0.0001256480113353879,
+      "loss": 2.533,
+      "step": 8407
+    },
+    {
+      "epoch": 0.6785570171898959,
+      "grad_norm": 0.7640479803085327,
+      "learning_rate": 0.0001256327521975836,
+      "loss": 2.5855,
+      "step": 8408
+    },
+    {
+      "epoch": 0.6786377209264789,
+      "grad_norm": 0.728111207485199,
+      "learning_rate": 0.00012561749242094412,
+      "loss": 2.6184,
+      "step": 8409
+    },
+    {
+      "epoch": 0.6787184246630619,
+      "grad_norm": 0.7842772603034973,
+      "learning_rate": 0.00012560223200584975,
+      "loss": 2.5915,
+      "step": 8410
+    },
+    {
+      "epoch": 0.678799128399645,
+      "grad_norm": 0.7129092812538147,
+      "learning_rate": 0.00012558697095268085,
+      "loss": 2.6526,
+      "step": 8411
+    },
+    {
+      "epoch": 0.6788798321362279,
+      "grad_norm": 0.751103401184082,
+      "learning_rate": 0.00012557170926181773,
+      "loss": 2.605,
+      "step": 8412
+    },
+    {
+      "epoch": 0.6789605358728109,
+      "grad_norm": 0.6850594878196716,
+      "learning_rate": 0.0001255564469336408,
+      "loss": 2.6047,
+      "step": 8413
+    },
+    {
+      "epoch": 0.6790412396093939,
+      "grad_norm": 0.703037679195404,
+      "learning_rate": 0.00012554118396853036,
+      "loss": 2.653,
+      "step": 8414
+    },
+    {
+      "epoch": 0.6791219433459769,
+      "grad_norm": 0.8097915053367615,
+      "learning_rate": 0.0001255259203668669,
+      "loss": 2.5937,
+      "step": 8415
+    },
+    {
+      "epoch": 0.67920264708256,
+      "grad_norm": 0.700351357460022,
+      "learning_rate": 0.00012551065612903076,
+      "loss": 2.6089,
+      "step": 8416
+    },
+    {
+      "epoch": 0.6792833508191429,
+      "grad_norm": 0.6760888695716858,
+      "learning_rate": 0.00012549539125540236,
+      "loss": 2.547,
+      "step": 8417
+    },
+    {
+      "epoch": 0.6793640545557259,
+      "grad_norm": 0.6751723289489746,
+      "learning_rate": 0.0001254801257463622,
+      "loss": 2.625,
+      "step": 8418
+    },
+    {
+      "epoch": 0.6794447582923089,
+      "grad_norm": 0.6928921937942505,
+      "learning_rate": 0.00012546485960229065,
+      "loss": 2.5671,
+      "step": 8419
+    },
+    {
+      "epoch": 0.679525462028892,
+      "grad_norm": 0.6541565656661987,
+      "learning_rate": 0.0001254495928235683,
+      "loss": 2.5837,
+      "step": 8420
+    },
+    {
+      "epoch": 0.679606165765475,
+      "grad_norm": 0.6228676438331604,
+      "learning_rate": 0.00012543432541057555,
+      "loss": 2.5798,
+      "step": 8421
+    },
+    {
+      "epoch": 0.6796868695020579,
+      "grad_norm": 0.7620853185653687,
+      "learning_rate": 0.0001254190573636929,
+      "loss": 2.5885,
+      "step": 8422
+    },
+    {
+      "epoch": 0.6797675732386409,
+      "grad_norm": 0.7425604462623596,
+      "learning_rate": 0.0001254037886833009,
+      "loss": 2.6124,
+      "step": 8423
+    },
+    {
+      "epoch": 0.679848276975224,
+      "grad_norm": 0.7150974273681641,
+      "learning_rate": 0.0001253885193697801,
+      "loss": 2.5423,
+      "step": 8424
+    },
+    {
+      "epoch": 0.679928980711807,
+      "grad_norm": 0.672649621963501,
+      "learning_rate": 0.000125373249423511,
+      "loss": 2.5563,
+      "step": 8425
+    },
+    {
+      "epoch": 0.6800096844483899,
+      "grad_norm": 0.6913620829582214,
+      "learning_rate": 0.00012535797884487425,
+      "loss": 2.5261,
+      "step": 8426
+    },
+    {
+      "epoch": 0.6800903881849729,
+      "grad_norm": 0.712123692035675,
+      "learning_rate": 0.00012534270763425034,
+      "loss": 2.5958,
+      "step": 8427
+    },
+    {
+      "epoch": 0.680171091921556,
+      "grad_norm": 0.7593061327934265,
+      "learning_rate": 0.00012532743579201993,
+      "loss": 2.6036,
+      "step": 8428
+    },
+    {
+      "epoch": 0.680251795658139,
+      "grad_norm": 0.7108714580535889,
+      "learning_rate": 0.0001253121633185636,
+      "loss": 2.6004,
+      "step": 8429
+    },
+    {
+      "epoch": 0.680332499394722,
+      "grad_norm": 0.7142449021339417,
+      "learning_rate": 0.00012529689021426198,
+      "loss": 2.588,
+      "step": 8430
+    },
+    {
+      "epoch": 0.6804132031313049,
+      "grad_norm": 0.7579841017723083,
+      "learning_rate": 0.00012528161647949574,
+      "loss": 2.5927,
+      "step": 8431
+    },
+    {
+      "epoch": 0.680493906867888,
+      "grad_norm": 0.6522083878517151,
+      "learning_rate": 0.00012526634211464555,
+      "loss": 2.5619,
+      "step": 8432
+    },
+    {
+      "epoch": 0.680574610604471,
+      "grad_norm": 0.7681782245635986,
+      "learning_rate": 0.00012525106712009203,
+      "loss": 2.6065,
+      "step": 8433
+    },
+    {
+      "epoch": 0.680655314341054,
+      "grad_norm": 0.6900169253349304,
+      "learning_rate": 0.00012523579149621594,
+      "loss": 2.5507,
+      "step": 8434
+    },
+    {
+      "epoch": 0.680736018077637,
+      "grad_norm": 0.6907666325569153,
+      "learning_rate": 0.00012522051524339794,
+      "loss": 2.5213,
+      "step": 8435
+    },
+    {
+      "epoch": 0.68081672181422,
+      "grad_norm": 0.7202023267745972,
+      "learning_rate": 0.0001252052383620188,
+      "loss": 2.6367,
+      "step": 8436
+    },
+    {
+      "epoch": 0.680897425550803,
+      "grad_norm": 0.7893621325492859,
+      "learning_rate": 0.00012518996085245925,
+      "loss": 2.6066,
+      "step": 8437
+    },
+    {
+      "epoch": 0.680978129287386,
+      "grad_norm": 0.7693532109260559,
+      "learning_rate": 0.00012517468271509998,
+      "loss": 2.5346,
+      "step": 8438
+    },
+    {
+      "epoch": 0.681058833023969,
+      "grad_norm": 0.7976840734481812,
+      "learning_rate": 0.0001251594039503218,
+      "loss": 2.5991,
+      "step": 8439
+    },
+    {
+      "epoch": 0.6811395367605521,
+      "grad_norm": 0.7671225666999817,
+      "learning_rate": 0.00012514412455850554,
+      "loss": 2.5959,
+      "step": 8440
+    },
+    {
+      "epoch": 0.681220240497135,
+      "grad_norm": 0.7143450975418091,
+      "learning_rate": 0.00012512884454003194,
+      "loss": 2.5828,
+      "step": 8441
+    },
+    {
+      "epoch": 0.681300944233718,
+      "grad_norm": 0.6821861863136292,
+      "learning_rate": 0.00012511356389528192,
+      "loss": 2.5908,
+      "step": 8442
+    },
+    {
+      "epoch": 0.681381647970301,
+      "grad_norm": 0.7279960513114929,
+      "learning_rate": 0.00012509828262463615,
+      "loss": 2.578,
+      "step": 8443
+    },
+    {
+      "epoch": 0.6814623517068841,
+      "grad_norm": 0.6503065824508667,
+      "learning_rate": 0.0001250830007284756,
+      "loss": 2.525,
+      "step": 8444
+    },
+    {
+      "epoch": 0.681543055443467,
+      "grad_norm": 0.7276029586791992,
+      "learning_rate": 0.00012506771820718112,
+      "loss": 2.584,
+      "step": 8445
+    },
+    {
+      "epoch": 0.68162375918005,
+      "grad_norm": 0.7635578513145447,
+      "learning_rate": 0.00012505243506113356,
+      "loss": 2.627,
+      "step": 8446
+    },
+    {
+      "epoch": 0.681704462916633,
+      "grad_norm": 0.7086981534957886,
+      "learning_rate": 0.00012503715129071386,
+      "loss": 2.6164,
+      "step": 8447
+    },
+    {
+      "epoch": 0.6817851666532161,
+      "grad_norm": 0.7144165635108948,
+      "learning_rate": 0.00012502186689630285,
+      "loss": 2.5642,
+      "step": 8448
+    },
+    {
+      "epoch": 0.6818658703897991,
+      "grad_norm": 0.8135093450546265,
+      "learning_rate": 0.00012500658187828155,
+      "loss": 2.6161,
+      "step": 8449
+    },
+    {
+      "epoch": 0.681946574126382,
+      "grad_norm": 0.7223377227783203,
+      "learning_rate": 0.00012499129623703086,
+      "loss": 2.6192,
+      "step": 8450
+    },
+    {
+      "epoch": 0.682027277862965,
+      "grad_norm": 0.7189127206802368,
+      "learning_rate": 0.00012497600997293172,
+      "loss": 2.6086,
+      "step": 8451
+    },
+    {
+      "epoch": 0.6821079815995481,
+      "grad_norm": 0.6742144823074341,
+      "learning_rate": 0.00012496072308636514,
+      "loss": 2.5747,
+      "step": 8452
+    },
+    {
+      "epoch": 0.6821886853361311,
+      "grad_norm": 0.7432419657707214,
+      "learning_rate": 0.0001249454355777121,
+      "loss": 2.5687,
+      "step": 8453
+    },
+    {
+      "epoch": 0.6822693890727141,
+      "grad_norm": 0.6140317320823669,
+      "learning_rate": 0.00012493014744735357,
+      "loss": 2.5371,
+      "step": 8454
+    },
+    {
+      "epoch": 0.682350092809297,
+      "grad_norm": 0.7215768098831177,
+      "learning_rate": 0.0001249148586956706,
+      "loss": 2.6806,
+      "step": 8455
+    },
+    {
+      "epoch": 0.6824307965458801,
+      "grad_norm": 0.7485790252685547,
+      "learning_rate": 0.0001248995693230442,
+      "loss": 2.575,
+      "step": 8456
+    },
+    {
+      "epoch": 0.6825115002824631,
+      "grad_norm": 0.744349479675293,
+      "learning_rate": 0.00012488427932985552,
+      "loss": 2.5961,
+      "step": 8457
+    },
+    {
+      "epoch": 0.6825922040190461,
+      "grad_norm": 0.6784959435462952,
+      "learning_rate": 0.0001248689887164855,
+      "loss": 2.5501,
+      "step": 8458
+    },
+    {
+      "epoch": 0.682672907755629,
+      "grad_norm": 0.6664010286331177,
+      "learning_rate": 0.0001248536974833153,
+      "loss": 2.5741,
+      "step": 8459
+    },
+    {
+      "epoch": 0.6827536114922121,
+      "grad_norm": 0.7185953259468079,
+      "learning_rate": 0.00012483840563072592,
+      "loss": 2.5875,
+      "step": 8460
+    },
+    {
+      "epoch": 0.6828343152287951,
+      "grad_norm": 0.6553035378456116,
+      "learning_rate": 0.00012482311315909864,
+      "loss": 2.5321,
+      "step": 8461
+    },
+    {
+      "epoch": 0.6829150189653781,
+      "grad_norm": 0.6713398694992065,
+      "learning_rate": 0.00012480782006881442,
+      "loss": 2.6207,
+      "step": 8462
+    },
+    {
+      "epoch": 0.6829957227019611,
+      "grad_norm": 0.6733734607696533,
+      "learning_rate": 0.00012479252636025452,
+      "loss": 2.5746,
+      "step": 8463
+    },
+    {
+      "epoch": 0.6830764264385442,
+      "grad_norm": 0.7257994413375854,
+      "learning_rate": 0.00012477723203380004,
+      "loss": 2.5837,
+      "step": 8464
+    },
+    {
+      "epoch": 0.6831571301751271,
+      "grad_norm": 0.716242253780365,
+      "learning_rate": 0.00012476193708983214,
+      "loss": 2.5611,
+      "step": 8465
+    },
+    {
+      "epoch": 0.6832378339117101,
+      "grad_norm": 0.6797829866409302,
+      "learning_rate": 0.0001247466415287321,
+      "loss": 2.5763,
+      "step": 8466
+    },
+    {
+      "epoch": 0.6833185376482931,
+      "grad_norm": 0.679931640625,
+      "learning_rate": 0.000124731345350881,
+      "loss": 2.606,
+      "step": 8467
+    },
+    {
+      "epoch": 0.6833992413848761,
+      "grad_norm": 0.6767866611480713,
+      "learning_rate": 0.00012471604855666016,
+      "loss": 2.5682,
+      "step": 8468
+    },
+    {
+      "epoch": 0.6834799451214592,
+      "grad_norm": 0.7297048568725586,
+      "learning_rate": 0.00012470075114645078,
+      "loss": 2.5527,
+      "step": 8469
+    },
+    {
+      "epoch": 0.6835606488580421,
+      "grad_norm": 0.6882644295692444,
+      "learning_rate": 0.0001246854531206341,
+      "loss": 2.5712,
+      "step": 8470
+    },
+    {
+      "epoch": 0.6836413525946251,
+      "grad_norm": 0.7129159569740295,
+      "learning_rate": 0.00012467015447959143,
+      "loss": 2.5627,
+      "step": 8471
+    },
+    {
+      "epoch": 0.6837220563312081,
+      "grad_norm": 0.6671481728553772,
+      "learning_rate": 0.000124654855223704,
+      "loss": 2.6226,
+      "step": 8472
+    },
+    {
+      "epoch": 0.6838027600677912,
+      "grad_norm": 0.7096946835517883,
+      "learning_rate": 0.00012463955535335313,
+      "loss": 2.5373,
+      "step": 8473
+    },
+    {
+      "epoch": 0.6838834638043741,
+      "grad_norm": 0.6781395077705383,
+      "learning_rate": 0.00012462425486892012,
+      "loss": 2.5607,
+      "step": 8474
+    },
+    {
+      "epoch": 0.6839641675409571,
+      "grad_norm": 0.6777891516685486,
+      "learning_rate": 0.00012460895377078632,
+      "loss": 2.5991,
+      "step": 8475
+    },
+    {
+      "epoch": 0.6840448712775401,
+      "grad_norm": 0.7175275087356567,
+      "learning_rate": 0.00012459365205933306,
+      "loss": 2.6006,
+      "step": 8476
+    },
+    {
+      "epoch": 0.6841255750141232,
+      "grad_norm": 0.6832807660102844,
+      "learning_rate": 0.00012457834973494174,
+      "loss": 2.5757,
+      "step": 8477
+    },
+    {
+      "epoch": 0.6842062787507062,
+      "grad_norm": 0.7002938985824585,
+      "learning_rate": 0.00012456304679799366,
+      "loss": 2.554,
+      "step": 8478
+    },
+    {
+      "epoch": 0.6842869824872891,
+      "grad_norm": 0.7236241698265076,
+      "learning_rate": 0.00012454774324887027,
+      "loss": 2.6054,
+      "step": 8479
+    },
+    {
+      "epoch": 0.6843676862238721,
+      "grad_norm": 0.7327216267585754,
+      "learning_rate": 0.00012453243908795288,
+      "loss": 2.6101,
+      "step": 8480
+    },
+    {
+      "epoch": 0.6844483899604552,
+      "grad_norm": 0.7414156794548035,
+      "learning_rate": 0.00012451713431562306,
+      "loss": 2.5505,
+      "step": 8481
+    },
+    {
+      "epoch": 0.6845290936970382,
+      "grad_norm": 0.697795569896698,
+      "learning_rate": 0.00012450182893226214,
+      "loss": 2.539,
+      "step": 8482
+    },
+    {
+      "epoch": 0.6846097974336212,
+      "grad_norm": 0.7053593397140503,
+      "learning_rate": 0.00012448652293825158,
+      "loss": 2.6045,
+      "step": 8483
+    },
+    {
+      "epoch": 0.6846905011702041,
+      "grad_norm": 0.6710856556892395,
+      "learning_rate": 0.00012447121633397287,
+      "loss": 2.554,
+      "step": 8484
+    },
+    {
+      "epoch": 0.6847712049067872,
+      "grad_norm": 0.754454493522644,
+      "learning_rate": 0.0001244559091198075,
+      "loss": 2.5523,
+      "step": 8485
+    },
+    {
+      "epoch": 0.6848519086433702,
+      "grad_norm": 0.6468656659126282,
+      "learning_rate": 0.0001244406012961369,
+      "loss": 2.5931,
+      "step": 8486
+    },
+    {
+      "epoch": 0.6849326123799532,
+      "grad_norm": 0.7169063091278076,
+      "learning_rate": 0.00012442529286334266,
+      "loss": 2.5743,
+      "step": 8487
+    },
+    {
+      "epoch": 0.6850133161165362,
+      "grad_norm": 0.6737040877342224,
+      "learning_rate": 0.00012440998382180627,
+      "loss": 2.5734,
+      "step": 8488
+    },
+    {
+      "epoch": 0.6850940198531192,
+      "grad_norm": 0.7026428580284119,
+      "learning_rate": 0.0001243946741719093,
+      "loss": 2.4994,
+      "step": 8489
+    },
+    {
+      "epoch": 0.6851747235897022,
+      "grad_norm": 0.7378512024879456,
+      "learning_rate": 0.00012437936391403322,
+      "loss": 2.5611,
+      "step": 8490
+    },
+    {
+      "epoch": 0.6852554273262852,
+      "grad_norm": 0.7379863262176514,
+      "learning_rate": 0.0001243640530485597,
+      "loss": 2.538,
+      "step": 8491
+    },
+    {
+      "epoch": 0.6853361310628682,
+      "grad_norm": 0.68398118019104,
+      "learning_rate": 0.00012434874157587027,
+      "loss": 2.5593,
+      "step": 8492
+    },
+    {
+      "epoch": 0.6854168347994513,
+      "grad_norm": 0.6780444383621216,
+      "learning_rate": 0.0001243334294963466,
+      "loss": 2.5068,
+      "step": 8493
+    },
+    {
+      "epoch": 0.6854975385360342,
+      "grad_norm": 0.7425427436828613,
+      "learning_rate": 0.0001243181168103702,
+      "loss": 2.6607,
+      "step": 8494
+    },
+    {
+      "epoch": 0.6855782422726172,
+      "grad_norm": 0.7563300132751465,
+      "learning_rate": 0.0001243028035183228,
+      "loss": 2.5915,
+      "step": 8495
+    },
+    {
+      "epoch": 0.6856589460092002,
+      "grad_norm": 0.6746618151664734,
+      "learning_rate": 0.000124287489620586,
+      "loss": 2.5399,
+      "step": 8496
+    },
+    {
+      "epoch": 0.6857396497457833,
+      "grad_norm": 0.7100487947463989,
+      "learning_rate": 0.00012427217511754146,
+      "loss": 2.5927,
+      "step": 8497
+    },
+    {
+      "epoch": 0.6858203534823663,
+      "grad_norm": 0.6487080454826355,
+      "learning_rate": 0.00012425686000957088,
+      "loss": 2.5582,
+      "step": 8498
+    },
+    {
+      "epoch": 0.6859010572189492,
+      "grad_norm": 0.6577199697494507,
+      "learning_rate": 0.00012424154429705592,
+      "loss": 2.5589,
+      "step": 8499
+    },
+    {
+      "epoch": 0.6859817609555322,
+      "grad_norm": 0.6748726963996887,
+      "learning_rate": 0.00012422622798037832,
+      "loss": 2.5651,
+      "step": 8500
+    },
+    {
+      "epoch": 0.6860624646921153,
+      "grad_norm": 0.7159377336502075,
+      "learning_rate": 0.0001242109110599198,
+      "loss": 2.569,
+      "step": 8501
+    },
+    {
+      "epoch": 0.6861431684286983,
+      "grad_norm": 0.6772934198379517,
+      "learning_rate": 0.00012419559353606208,
+      "loss": 2.5533,
+      "step": 8502
+    },
+    {
+      "epoch": 0.6862238721652812,
+      "grad_norm": 0.6776062846183777,
+      "learning_rate": 0.00012418027540918693,
+      "loss": 2.5704,
+      "step": 8503
+    },
+    {
+      "epoch": 0.6863045759018642,
+      "grad_norm": 0.7009913921356201,
+      "learning_rate": 0.00012416495667967608,
+      "loss": 2.5928,
+      "step": 8504
+    },
+    {
+      "epoch": 0.6863852796384473,
+      "grad_norm": 0.607571005821228,
+      "learning_rate": 0.00012414963734791137,
+      "loss": 2.5459,
+      "step": 8505
+    },
+    {
+      "epoch": 0.6864659833750303,
+      "grad_norm": 0.6798292398452759,
+      "learning_rate": 0.00012413431741427458,
+      "loss": 2.6585,
+      "step": 8506
+    },
+    {
+      "epoch": 0.6865466871116133,
+      "grad_norm": 0.7892771363258362,
+      "learning_rate": 0.00012411899687914747,
+      "loss": 2.5781,
+      "step": 8507
+    },
+    {
+      "epoch": 0.6866273908481962,
+      "grad_norm": 0.6683816909790039,
+      "learning_rate": 0.00012410367574291199,
+      "loss": 2.5598,
+      "step": 8508
+    },
+    {
+      "epoch": 0.6867080945847793,
+      "grad_norm": 0.7591805458068848,
+      "learning_rate": 0.00012408835400594983,
+      "loss": 2.6478,
+      "step": 8509
+    },
+    {
+      "epoch": 0.6867887983213623,
+      "grad_norm": 0.6896353960037231,
+      "learning_rate": 0.00012407303166864293,
+      "loss": 2.5418,
+      "step": 8510
+    },
+    {
+      "epoch": 0.6868695020579453,
+      "grad_norm": 0.6657233834266663,
+      "learning_rate": 0.00012405770873137316,
+      "loss": 2.5753,
+      "step": 8511
+    },
+    {
+      "epoch": 0.6869502057945283,
+      "grad_norm": 0.6775455474853516,
+      "learning_rate": 0.00012404238519452237,
+      "loss": 2.4902,
+      "step": 8512
+    },
+    {
+      "epoch": 0.6870309095311113,
+      "grad_norm": 0.6572847962379456,
+      "learning_rate": 0.00012402706105847254,
+      "loss": 2.6189,
+      "step": 8513
+    },
+    {
+      "epoch": 0.6871116132676943,
+      "grad_norm": 0.7159940004348755,
+      "learning_rate": 0.00012401173632360557,
+      "loss": 2.5928,
+      "step": 8514
+    },
+    {
+      "epoch": 0.6871923170042773,
+      "grad_norm": 0.7178850173950195,
+      "learning_rate": 0.0001239964109903033,
+      "loss": 2.5342,
+      "step": 8515
+    },
+    {
+      "epoch": 0.6872730207408603,
+      "grad_norm": 0.6761649250984192,
+      "learning_rate": 0.00012398108505894774,
+      "loss": 2.5716,
+      "step": 8516
+    },
+    {
+      "epoch": 0.6873537244774433,
+      "grad_norm": 0.6831200122833252,
+      "learning_rate": 0.0001239657585299209,
+      "loss": 2.5506,
+      "step": 8517
+    },
+    {
+      "epoch": 0.6874344282140263,
+      "grad_norm": 0.7064316868782043,
+      "learning_rate": 0.00012395043140360468,
+      "loss": 2.541,
+      "step": 8518
+    },
+    {
+      "epoch": 0.6875151319506093,
+      "grad_norm": 0.7269963026046753,
+      "learning_rate": 0.00012393510368038113,
+      "loss": 2.541,
+      "step": 8519
+    },
+    {
+      "epoch": 0.6875958356871923,
+      "grad_norm": 0.6651471257209778,
+      "learning_rate": 0.00012391977536063218,
+      "loss": 2.5476,
+      "step": 8520
+    },
+    {
+      "epoch": 0.6876765394237753,
+      "grad_norm": 0.7649257779121399,
+      "learning_rate": 0.00012390444644473994,
+      "loss": 2.601,
+      "step": 8521
+    },
+    {
+      "epoch": 0.6877572431603584,
+      "grad_norm": 0.6637376546859741,
+      "learning_rate": 0.0001238891169330864,
+      "loss": 2.5582,
+      "step": 8522
+    },
+    {
+      "epoch": 0.6878379468969413,
+      "grad_norm": 0.6609189510345459,
+      "learning_rate": 0.0001238737868260536,
+      "loss": 2.5795,
+      "step": 8523
+    },
+    {
+      "epoch": 0.6879186506335243,
+      "grad_norm": 0.657494843006134,
+      "learning_rate": 0.00012385845612402363,
+      "loss": 2.6005,
+      "step": 8524
+    },
+    {
+      "epoch": 0.6879993543701073,
+      "grad_norm": 0.6780641674995422,
+      "learning_rate": 0.00012384312482737858,
+      "loss": 2.514,
+      "step": 8525
+    },
+    {
+      "epoch": 0.6880800581066904,
+      "grad_norm": 0.7310795187950134,
+      "learning_rate": 0.00012382779293650052,
+      "loss": 2.5707,
+      "step": 8526
+    },
+    {
+      "epoch": 0.6881607618432733,
+      "grad_norm": 0.6722557544708252,
+      "learning_rate": 0.0001238124604517716,
+      "loss": 2.5897,
+      "step": 8527
+    },
+    {
+      "epoch": 0.6882414655798563,
+      "grad_norm": 0.6502346992492676,
+      "learning_rate": 0.0001237971273735739,
+      "loss": 2.5554,
+      "step": 8528
+    },
+    {
+      "epoch": 0.6883221693164393,
+      "grad_norm": 0.6993897557258606,
+      "learning_rate": 0.0001237817937022896,
+      "loss": 2.6328,
+      "step": 8529
+    },
+    {
+      "epoch": 0.6884028730530224,
+      "grad_norm": 0.7069644331932068,
+      "learning_rate": 0.00012376645943830083,
+      "loss": 2.5957,
+      "step": 8530
+    },
+    {
+      "epoch": 0.6884835767896054,
+      "grad_norm": 0.7193333506584167,
+      "learning_rate": 0.00012375112458198973,
+      "loss": 2.6505,
+      "step": 8531
+    },
+    {
+      "epoch": 0.6885642805261883,
+      "grad_norm": 0.6821088194847107,
+      "learning_rate": 0.00012373578913373853,
+      "loss": 2.6129,
+      "step": 8532
+    },
+    {
+      "epoch": 0.6886449842627713,
+      "grad_norm": 0.6499428749084473,
+      "learning_rate": 0.00012372045309392947,
+      "loss": 2.6053,
+      "step": 8533
+    },
+    {
+      "epoch": 0.6887256879993544,
+      "grad_norm": 0.7469449639320374,
+      "learning_rate": 0.00012370511646294464,
+      "loss": 2.6423,
+      "step": 8534
+    },
+    {
+      "epoch": 0.6888063917359374,
+      "grad_norm": 0.7326325178146362,
+      "learning_rate": 0.00012368977924116637,
+      "loss": 2.5708,
+      "step": 8535
+    },
+    {
+      "epoch": 0.6888870954725204,
+      "grad_norm": 0.7459580302238464,
+      "learning_rate": 0.00012367444142897686,
+      "loss": 2.544,
+      "step": 8536
+    },
+    {
+      "epoch": 0.6889677992091033,
+      "grad_norm": 0.7198929786682129,
+      "learning_rate": 0.00012365910302675843,
+      "loss": 2.6295,
+      "step": 8537
+    },
+    {
+      "epoch": 0.6890485029456864,
+      "grad_norm": 0.8139802813529968,
+      "learning_rate": 0.0001236437640348933,
+      "loss": 2.549,
+      "step": 8538
+    },
+    {
+      "epoch": 0.6891292066822694,
+      "grad_norm": 0.6497162580490112,
+      "learning_rate": 0.00012362842445376372,
+      "loss": 2.5849,
+      "step": 8539
+    },
+    {
+      "epoch": 0.6892099104188524,
+      "grad_norm": 0.7378165125846863,
+      "learning_rate": 0.00012361308428375208,
+      "loss": 2.606,
+      "step": 8540
+    },
+    {
+      "epoch": 0.6892906141554354,
+      "grad_norm": 0.6807567477226257,
+      "learning_rate": 0.00012359774352524062,
+      "loss": 2.5892,
+      "step": 8541
+    },
+    {
+      "epoch": 0.6893713178920184,
+      "grad_norm": 0.6639370918273926,
+      "learning_rate": 0.0001235824021786117,
+      "loss": 2.5249,
+      "step": 8542
+    },
+    {
+      "epoch": 0.6894520216286014,
+      "grad_norm": 0.7140880823135376,
+      "learning_rate": 0.00012356706024424773,
+      "loss": 2.5877,
+      "step": 8543
+    },
+    {
+      "epoch": 0.6895327253651844,
+      "grad_norm": 0.7079257965087891,
+      "learning_rate": 0.00012355171772253097,
+      "loss": 2.6011,
+      "step": 8544
+    },
+    {
+      "epoch": 0.6896134291017674,
+      "grad_norm": 0.7150856852531433,
+      "learning_rate": 0.00012353637461384387,
+      "loss": 2.549,
+      "step": 8545
+    },
+    {
+      "epoch": 0.6896941328383505,
+      "grad_norm": 0.6896397471427917,
+      "learning_rate": 0.00012352103091856876,
+      "loss": 2.5452,
+      "step": 8546
+    },
+    {
+      "epoch": 0.6897748365749334,
+      "grad_norm": 0.696964681148529,
+      "learning_rate": 0.00012350568663708808,
+      "loss": 2.5075,
+      "step": 8547
+    },
+    {
+      "epoch": 0.6898555403115164,
+      "grad_norm": 0.6926069855690002,
+      "learning_rate": 0.00012349034176978427,
+      "loss": 2.5905,
+      "step": 8548
+    },
+    {
+      "epoch": 0.6899362440480994,
+      "grad_norm": 0.6949423551559448,
+      "learning_rate": 0.00012347499631703968,
+      "loss": 2.5284,
+      "step": 8549
+    },
+    {
+      "epoch": 0.6900169477846825,
+      "grad_norm": 0.6480536460876465,
+      "learning_rate": 0.0001234596502792369,
+      "loss": 2.5713,
+      "step": 8550
+    },
+    {
+      "epoch": 0.6900976515212655,
+      "grad_norm": 0.6990019679069519,
+      "learning_rate": 0.00012344430365675825,
+      "loss": 2.5826,
+      "step": 8551
+    },
+    {
+      "epoch": 0.6901783552578484,
+      "grad_norm": 0.7063903212547302,
+      "learning_rate": 0.00012342895644998627,
+      "loss": 2.5271,
+      "step": 8552
+    },
+    {
+      "epoch": 0.6902590589944314,
+      "grad_norm": 0.7037132978439331,
+      "learning_rate": 0.0001234136086593035,
+      "loss": 2.5855,
+      "step": 8553
+    },
+    {
+      "epoch": 0.6903397627310145,
+      "grad_norm": 0.679701030254364,
+      "learning_rate": 0.00012339826028509235,
+      "loss": 2.5577,
+      "step": 8554
+    },
+    {
+      "epoch": 0.6904204664675975,
+      "grad_norm": 0.7088965773582458,
+      "learning_rate": 0.0001233829113277354,
+      "loss": 2.5767,
+      "step": 8555
+    },
+    {
+      "epoch": 0.6905011702041804,
+      "grad_norm": 0.7115551829338074,
+      "learning_rate": 0.00012336756178761517,
+      "loss": 2.5651,
+      "step": 8556
+    },
+    {
+      "epoch": 0.6905818739407634,
+      "grad_norm": 0.6778836250305176,
+      "learning_rate": 0.00012335221166511425,
+      "loss": 2.6388,
+      "step": 8557
+    },
+    {
+      "epoch": 0.6906625776773465,
+      "grad_norm": 0.6358879804611206,
+      "learning_rate": 0.00012333686096061515,
+      "loss": 2.5493,
+      "step": 8558
+    },
+    {
+      "epoch": 0.6907432814139295,
+      "grad_norm": 0.688197135925293,
+      "learning_rate": 0.00012332150967450046,
+      "loss": 2.5707,
+      "step": 8559
+    },
+    {
+      "epoch": 0.6908239851505125,
+      "grad_norm": 0.6931524872779846,
+      "learning_rate": 0.0001233061578071528,
+      "loss": 2.5561,
+      "step": 8560
+    },
+    {
+      "epoch": 0.6909046888870954,
+      "grad_norm": 0.6684975624084473,
+      "learning_rate": 0.00012329080535895478,
+      "loss": 2.6442,
+      "step": 8561
+    },
+    {
+      "epoch": 0.6909853926236785,
+      "grad_norm": 0.6865811347961426,
+      "learning_rate": 0.00012327545233028898,
+      "loss": 2.564,
+      "step": 8562
+    },
+    {
+      "epoch": 0.6910660963602615,
+      "grad_norm": 0.6999006867408752,
+      "learning_rate": 0.0001232600987215381,
+      "loss": 2.5607,
+      "step": 8563
+    },
+    {
+      "epoch": 0.6911468000968445,
+      "grad_norm": 0.6734526753425598,
+      "learning_rate": 0.0001232447445330847,
+      "loss": 2.5261,
+      "step": 8564
+    },
+    {
+      "epoch": 0.6912275038334275,
+      "grad_norm": 0.7447343468666077,
+      "learning_rate": 0.00012322938976531153,
+      "loss": 2.5359,
+      "step": 8565
+    },
+    {
+      "epoch": 0.6913082075700105,
+      "grad_norm": 0.6498517394065857,
+      "learning_rate": 0.00012321403441860126,
+      "loss": 2.5345,
+      "step": 8566
+    },
+    {
+      "epoch": 0.6913889113065935,
+      "grad_norm": 0.692933976650238,
+      "learning_rate": 0.00012319867849333658,
+      "loss": 2.6293,
+      "step": 8567
+    },
+    {
+      "epoch": 0.6914696150431765,
+      "grad_norm": 0.728430449962616,
+      "learning_rate": 0.00012318332198990015,
+      "loss": 2.618,
+      "step": 8568
+    },
+    {
+      "epoch": 0.6915503187797595,
+      "grad_norm": 0.7029061913490295,
+      "learning_rate": 0.00012316796490867478,
+      "loss": 2.6151,
+      "step": 8569
+    },
+    {
+      "epoch": 0.6916310225163425,
+      "grad_norm": 0.6692330241203308,
+      "learning_rate": 0.00012315260725004313,
+      "loss": 2.5511,
+      "step": 8570
+    },
+    {
+      "epoch": 0.6917117262529255,
+      "grad_norm": 0.6811983585357666,
+      "learning_rate": 0.000123137249014388,
+      "loss": 2.6337,
+      "step": 8571
+    },
+    {
+      "epoch": 0.6917924299895085,
+      "grad_norm": 0.7387441992759705,
+      "learning_rate": 0.00012312189020209212,
+      "loss": 2.5679,
+      "step": 8572
+    },
+    {
+      "epoch": 0.6918731337260915,
+      "grad_norm": 0.7180185914039612,
+      "learning_rate": 0.0001231065308135383,
+      "loss": 2.639,
+      "step": 8573
+    },
+    {
+      "epoch": 0.6919538374626745,
+      "grad_norm": 0.6997829079627991,
+      "learning_rate": 0.00012309117084910936,
+      "loss": 2.5392,
+      "step": 8574
+    },
+    {
+      "epoch": 0.6920345411992576,
+      "grad_norm": 0.7004552483558655,
+      "learning_rate": 0.00012307581030918807,
+      "loss": 2.6033,
+      "step": 8575
+    },
+    {
+      "epoch": 0.6921152449358405,
+      "grad_norm": 0.7183418273925781,
+      "learning_rate": 0.00012306044919415724,
+      "loss": 2.6302,
+      "step": 8576
+    },
+    {
+      "epoch": 0.6921959486724235,
+      "grad_norm": 0.6645712852478027,
+      "learning_rate": 0.00012304508750439976,
+      "loss": 2.5401,
+      "step": 8577
+    },
+    {
+      "epoch": 0.6922766524090065,
+      "grad_norm": 0.6455898284912109,
+      "learning_rate": 0.00012302972524029848,
+      "loss": 2.5084,
+      "step": 8578
+    },
+    {
+      "epoch": 0.6923573561455896,
+      "grad_norm": 0.6933849453926086,
+      "learning_rate": 0.00012301436240223622,
+      "loss": 2.5734,
+      "step": 8579
+    },
+    {
+      "epoch": 0.6924380598821726,
+      "grad_norm": 0.7967655658721924,
+      "learning_rate": 0.00012299899899059587,
+      "loss": 2.5721,
+      "step": 8580
+    },
+    {
+      "epoch": 0.6925187636187555,
+      "grad_norm": 0.706730306148529,
+      "learning_rate": 0.0001229836350057604,
+      "loss": 2.6216,
+      "step": 8581
+    },
+    {
+      "epoch": 0.6925994673553385,
+      "grad_norm": 0.7021105885505676,
+      "learning_rate": 0.0001229682704481126,
+      "loss": 2.4877,
+      "step": 8582
+    },
+    {
+      "epoch": 0.6926801710919216,
+      "grad_norm": 0.7197253108024597,
+      "learning_rate": 0.00012295290531803553,
+      "loss": 2.6124,
+      "step": 8583
+    },
+    {
+      "epoch": 0.6927608748285046,
+      "grad_norm": 0.7559605836868286,
+      "learning_rate": 0.00012293753961591198,
+      "loss": 2.6391,
+      "step": 8584
+    },
+    {
+      "epoch": 0.6928415785650875,
+      "grad_norm": 0.7074676752090454,
+      "learning_rate": 0.00012292217334212505,
+      "loss": 2.5949,
+      "step": 8585
+    },
+    {
+      "epoch": 0.6929222823016705,
+      "grad_norm": 0.6843528747558594,
+      "learning_rate": 0.00012290680649705763,
+      "loss": 2.4981,
+      "step": 8586
+    },
+    {
+      "epoch": 0.6930029860382536,
+      "grad_norm": 0.6853117942810059,
+      "learning_rate": 0.00012289143908109266,
+      "loss": 2.6352,
+      "step": 8587
+    },
+    {
+      "epoch": 0.6930836897748366,
+      "grad_norm": 0.6545630097389221,
+      "learning_rate": 0.00012287607109461325,
+      "loss": 2.5344,
+      "step": 8588
+    },
+    {
+      "epoch": 0.6931643935114196,
+      "grad_norm": 0.7377945184707642,
+      "learning_rate": 0.00012286070253800233,
+      "loss": 2.5895,
+      "step": 8589
+    },
+    {
+      "epoch": 0.6932450972480025,
+      "grad_norm": 0.6919971108436584,
+      "learning_rate": 0.00012284533341164295,
+      "loss": 2.5825,
+      "step": 8590
+    },
+    {
+      "epoch": 0.6933258009845856,
+      "grad_norm": 0.6911910176277161,
+      "learning_rate": 0.00012282996371591816,
+      "loss": 2.6008,
+      "step": 8591
+    },
+    {
+      "epoch": 0.6934065047211686,
+      "grad_norm": 0.7486373782157898,
+      "learning_rate": 0.00012281459345121095,
+      "loss": 2.6056,
+      "step": 8592
+    },
+    {
+      "epoch": 0.6934872084577516,
+      "grad_norm": 0.6829040050506592,
+      "learning_rate": 0.00012279922261790443,
+      "loss": 2.5161,
+      "step": 8593
+    },
+    {
+      "epoch": 0.6935679121943346,
+      "grad_norm": 0.7410104870796204,
+      "learning_rate": 0.00012278385121638173,
+      "loss": 2.6114,
+      "step": 8594
+    },
+    {
+      "epoch": 0.6936486159309176,
+      "grad_norm": 0.7355940937995911,
+      "learning_rate": 0.00012276847924702587,
+      "loss": 2.6371,
+      "step": 8595
+    },
+    {
+      "epoch": 0.6937293196675006,
+      "grad_norm": 0.650641679763794,
+      "learning_rate": 0.00012275310671022003,
+      "loss": 2.5568,
+      "step": 8596
+    },
+    {
+      "epoch": 0.6938100234040836,
+      "grad_norm": 0.661573052406311,
+      "learning_rate": 0.00012273773360634726,
+      "loss": 2.5828,
+      "step": 8597
+    },
+    {
+      "epoch": 0.6938907271406666,
+      "grad_norm": 0.6848435401916504,
+      "learning_rate": 0.00012272235993579072,
+      "loss": 2.5226,
+      "step": 8598
+    },
+    {
+      "epoch": 0.6939714308772497,
+      "grad_norm": 0.7015430927276611,
+      "learning_rate": 0.0001227069856989336,
+      "loss": 2.6156,
+      "step": 8599
+    },
+    {
+      "epoch": 0.6940521346138326,
+      "grad_norm": 0.7058628797531128,
+      "learning_rate": 0.000122691610896159,
+      "loss": 2.6007,
+      "step": 8600
+    },
+    {
+      "epoch": 0.6941328383504156,
+      "grad_norm": 0.6589432954788208,
+      "learning_rate": 0.0001226762355278502,
+      "loss": 2.5551,
+      "step": 8601
+    },
+    {
+      "epoch": 0.6942135420869986,
+      "grad_norm": 0.6875284910202026,
+      "learning_rate": 0.0001226608595943903,
+      "loss": 2.5537,
+      "step": 8602
+    },
+    {
+      "epoch": 0.6942942458235817,
+      "grad_norm": 0.7178356051445007,
+      "learning_rate": 0.00012264548309616252,
+      "loss": 2.655,
+      "step": 8603
+    },
+    {
+      "epoch": 0.6943749495601647,
+      "grad_norm": 0.7327077388763428,
+      "learning_rate": 0.00012263010603355017,
+      "loss": 2.5574,
+      "step": 8604
+    },
+    {
+      "epoch": 0.6944556532967476,
+      "grad_norm": 0.6318337917327881,
+      "learning_rate": 0.0001226147284069364,
+      "loss": 2.577,
+      "step": 8605
+    },
+    {
+      "epoch": 0.6945363570333306,
+      "grad_norm": 0.674872875213623,
+      "learning_rate": 0.00012259935021670444,
+      "loss": 2.6225,
+      "step": 8606
+    },
+    {
+      "epoch": 0.6946170607699137,
+      "grad_norm": 0.6554198861122131,
+      "learning_rate": 0.0001225839714632376,
+      "loss": 2.5951,
+      "step": 8607
+    },
+    {
+      "epoch": 0.6946977645064967,
+      "grad_norm": 0.7086453437805176,
+      "learning_rate": 0.00012256859214691918,
+      "loss": 2.622,
+      "step": 8608
+    },
+    {
+      "epoch": 0.6947784682430796,
+      "grad_norm": 0.6609488129615784,
+      "learning_rate": 0.00012255321226813245,
+      "loss": 2.5623,
+      "step": 8609
+    },
+    {
+      "epoch": 0.6948591719796626,
+      "grad_norm": 0.7504609823226929,
+      "learning_rate": 0.00012253783182726075,
+      "loss": 2.5264,
+      "step": 8610
+    },
+    {
+      "epoch": 0.6949398757162457,
+      "grad_norm": 0.6702934503555298,
+      "learning_rate": 0.00012252245082468733,
+      "loss": 2.5877,
+      "step": 8611
+    },
+    {
+      "epoch": 0.6950205794528287,
+      "grad_norm": 0.7116326689720154,
+      "learning_rate": 0.00012250706926079553,
+      "loss": 2.5629,
+      "step": 8612
+    },
+    {
+      "epoch": 0.6951012831894117,
+      "grad_norm": 0.7495368719100952,
+      "learning_rate": 0.00012249168713596875,
+      "loss": 2.5731,
+      "step": 8613
+    },
+    {
+      "epoch": 0.6951819869259946,
+      "grad_norm": 0.7434844970703125,
+      "learning_rate": 0.0001224763044505904,
+      "loss": 2.6008,
+      "step": 8614
+    },
+    {
+      "epoch": 0.6952626906625777,
+      "grad_norm": 0.719667375087738,
+      "learning_rate": 0.00012246092120504371,
+      "loss": 2.6051,
+      "step": 8615
+    },
+    {
+      "epoch": 0.6953433943991607,
+      "grad_norm": 0.7189086079597473,
+      "learning_rate": 0.00012244553739971216,
+      "loss": 2.5662,
+      "step": 8616
+    },
+    {
+      "epoch": 0.6954240981357437,
+      "grad_norm": 0.7222673892974854,
+      "learning_rate": 0.00012243015303497917,
+      "loss": 2.609,
+      "step": 8617
+    },
+    {
+      "epoch": 0.6955048018723267,
+      "grad_norm": 0.7323142290115356,
+      "learning_rate": 0.00012241476811122813,
+      "loss": 2.5458,
+      "step": 8618
+    },
+    {
+      "epoch": 0.6955855056089096,
+      "grad_norm": 0.7374032735824585,
+      "learning_rate": 0.00012239938262884246,
+      "loss": 2.6147,
+      "step": 8619
+    },
+    {
+      "epoch": 0.6956662093454927,
+      "grad_norm": 0.6707843542098999,
+      "learning_rate": 0.00012238399658820562,
+      "loss": 2.6462,
+      "step": 8620
+    },
+    {
+      "epoch": 0.6957469130820757,
+      "grad_norm": 0.7603243589401245,
+      "learning_rate": 0.0001223686099897011,
+      "loss": 2.6295,
+      "step": 8621
+    },
+    {
+      "epoch": 0.6958276168186587,
+      "grad_norm": 0.6966906785964966,
+      "learning_rate": 0.00012235322283371232,
+      "loss": 2.545,
+      "step": 8622
+    },
+    {
+      "epoch": 0.6959083205552417,
+      "grad_norm": 0.6757891774177551,
+      "learning_rate": 0.0001223378351206228,
+      "loss": 2.5548,
+      "step": 8623
+    },
+    {
+      "epoch": 0.6959890242918247,
+      "grad_norm": 0.6901456713676453,
+      "learning_rate": 0.00012232244685081605,
+      "loss": 2.5734,
+      "step": 8624
+    },
+    {
+      "epoch": 0.6960697280284077,
+      "grad_norm": 0.6942903995513916,
+      "learning_rate": 0.00012230705802467558,
+      "loss": 2.5495,
+      "step": 8625
+    },
+    {
+      "epoch": 0.6961504317649907,
+      "grad_norm": 0.6774815320968628,
+      "learning_rate": 0.0001222916686425849,
+      "loss": 2.5076,
+      "step": 8626
+    },
+    {
+      "epoch": 0.6962311355015737,
+      "grad_norm": 0.8037571310997009,
+      "learning_rate": 0.00012227627870492754,
+      "loss": 2.6737,
+      "step": 8627
+    },
+    {
+      "epoch": 0.6963118392381568,
+      "grad_norm": 0.7027560472488403,
+      "learning_rate": 0.0001222608882120871,
+      "loss": 2.5401,
+      "step": 8628
+    },
+    {
+      "epoch": 0.6963925429747397,
+      "grad_norm": 0.6651299595832825,
+      "learning_rate": 0.00012224549716444714,
+      "loss": 2.5835,
+      "step": 8629
+    },
+    {
+      "epoch": 0.6964732467113227,
+      "grad_norm": 0.7082433104515076,
+      "learning_rate": 0.00012223010556239124,
+      "loss": 2.5622,
+      "step": 8630
+    },
+    {
+      "epoch": 0.6965539504479057,
+      "grad_norm": 0.7993464469909668,
+      "learning_rate": 0.00012221471340630305,
+      "loss": 2.655,
+      "step": 8631
+    },
+    {
+      "epoch": 0.6966346541844888,
+      "grad_norm": 0.7375298142433167,
+      "learning_rate": 0.00012219932069656606,
+      "loss": 2.598,
+      "step": 8632
+    },
+    {
+      "epoch": 0.6967153579210718,
+      "grad_norm": 0.6915456652641296,
+      "learning_rate": 0.00012218392743356397,
+      "loss": 2.5649,
+      "step": 8633
+    },
+    {
+      "epoch": 0.6967960616576547,
+      "grad_norm": 0.679256021976471,
+      "learning_rate": 0.00012216853361768045,
+      "loss": 2.545,
+      "step": 8634
+    },
+    {
+      "epoch": 0.6968767653942377,
+      "grad_norm": 0.7234694361686707,
+      "learning_rate": 0.0001221531392492991,
+      "loss": 2.5863,
+      "step": 8635
+    },
+    {
+      "epoch": 0.6969574691308208,
+      "grad_norm": 0.7053319811820984,
+      "learning_rate": 0.00012213774432880364,
+      "loss": 2.5829,
+      "step": 8636
+    },
+    {
+      "epoch": 0.6970381728674038,
+      "grad_norm": 0.7584449648857117,
+      "learning_rate": 0.00012212234885657772,
+      "loss": 2.5855,
+      "step": 8637
+    },
+    {
+      "epoch": 0.6971188766039867,
+      "grad_norm": 0.7098579406738281,
+      "learning_rate": 0.00012210695283300501,
+      "loss": 2.6057,
+      "step": 8638
+    },
+    {
+      "epoch": 0.6971995803405697,
+      "grad_norm": 0.7350205779075623,
+      "learning_rate": 0.00012209155625846928,
+      "loss": 2.546,
+      "step": 8639
+    },
+    {
+      "epoch": 0.6972802840771528,
+      "grad_norm": 0.6842331290245056,
+      "learning_rate": 0.0001220761591333542,
+      "loss": 2.5602,
+      "step": 8640
+    },
+    {
+      "epoch": 0.6973609878137358,
+      "grad_norm": 0.6731252074241638,
+      "learning_rate": 0.00012206076145804354,
+      "loss": 2.4676,
+      "step": 8641
+    },
+    {
+      "epoch": 0.6974416915503188,
+      "grad_norm": 0.7271167635917664,
+      "learning_rate": 0.00012204536323292104,
+      "loss": 2.5605,
+      "step": 8642
+    },
+    {
+      "epoch": 0.6975223952869017,
+      "grad_norm": 0.6860780715942383,
+      "learning_rate": 0.00012202996445837043,
+      "loss": 2.5041,
+      "step": 8643
+    },
+    {
+      "epoch": 0.6976030990234848,
+      "grad_norm": 0.7134578824043274,
+      "learning_rate": 0.00012201456513477554,
+      "loss": 2.614,
+      "step": 8644
+    },
+    {
+      "epoch": 0.6976838027600678,
+      "grad_norm": 0.6995248198509216,
+      "learning_rate": 0.00012199916526252014,
+      "loss": 2.5087,
+      "step": 8645
+    },
+    {
+      "epoch": 0.6977645064966508,
+      "grad_norm": 0.7280197143554688,
+      "learning_rate": 0.00012198376484198803,
+      "loss": 2.5723,
+      "step": 8646
+    },
+    {
+      "epoch": 0.6978452102332338,
+      "grad_norm": 0.6898967623710632,
+      "learning_rate": 0.00012196836387356306,
+      "loss": 2.6073,
+      "step": 8647
+    },
+    {
+      "epoch": 0.6979259139698168,
+      "grad_norm": 0.6670758128166199,
+      "learning_rate": 0.00012195296235762901,
+      "loss": 2.5276,
+      "step": 8648
+    },
+    {
+      "epoch": 0.6980066177063998,
+      "grad_norm": 0.6862780451774597,
+      "learning_rate": 0.00012193756029456973,
+      "loss": 2.5363,
+      "step": 8649
+    },
+    {
+      "epoch": 0.6980873214429828,
+      "grad_norm": 0.6568876504898071,
+      "learning_rate": 0.00012192215768476916,
+      "loss": 2.5828,
+      "step": 8650
+    },
+    {
+      "epoch": 0.6981680251795658,
+      "grad_norm": 0.7237746119499207,
+      "learning_rate": 0.00012190675452861107,
+      "loss": 2.6076,
+      "step": 8651
+    },
+    {
+      "epoch": 0.6982487289161489,
+      "grad_norm": 0.6831536293029785,
+      "learning_rate": 0.00012189135082647943,
+      "loss": 2.5199,
+      "step": 8652
+    },
+    {
+      "epoch": 0.6983294326527318,
+      "grad_norm": 0.6767029166221619,
+      "learning_rate": 0.00012187594657875805,
+      "loss": 2.5859,
+      "step": 8653
+    },
+    {
+      "epoch": 0.6984101363893148,
+      "grad_norm": 0.6977167129516602,
+      "learning_rate": 0.00012186054178583092,
+      "loss": 2.5831,
+      "step": 8654
+    },
+    {
+      "epoch": 0.6984908401258978,
+      "grad_norm": 0.6369525194168091,
+      "learning_rate": 0.00012184513644808197,
+      "loss": 2.5839,
+      "step": 8655
+    },
+    {
+      "epoch": 0.6985715438624809,
+      "grad_norm": 0.6814634203910828,
+      "learning_rate": 0.00012182973056589508,
+      "loss": 2.5493,
+      "step": 8656
+    },
+    {
+      "epoch": 0.6986522475990639,
+      "grad_norm": 0.6895000338554382,
+      "learning_rate": 0.00012181432413965428,
+      "loss": 2.5616,
+      "step": 8657
+    },
+    {
+      "epoch": 0.6987329513356468,
+      "grad_norm": 0.6689717769622803,
+      "learning_rate": 0.00012179891716974345,
+      "loss": 2.5481,
+      "step": 8658
+    },
+    {
+      "epoch": 0.6988136550722298,
+      "grad_norm": 0.6945160031318665,
+      "learning_rate": 0.00012178350965654666,
+      "loss": 2.5781,
+      "step": 8659
+    },
+    {
+      "epoch": 0.6988943588088129,
+      "grad_norm": 0.7226110696792603,
+      "learning_rate": 0.00012176810160044785,
+      "loss": 2.5767,
+      "step": 8660
+    },
+    {
+      "epoch": 0.6989750625453959,
+      "grad_norm": 0.6810569167137146,
+      "learning_rate": 0.00012175269300183105,
+      "loss": 2.5184,
+      "step": 8661
+    },
+    {
+      "epoch": 0.6990557662819789,
+      "grad_norm": 0.727281928062439,
+      "learning_rate": 0.0001217372838610803,
+      "loss": 2.5972,
+      "step": 8662
+    },
+    {
+      "epoch": 0.6991364700185618,
+      "grad_norm": 0.7111573219299316,
+      "learning_rate": 0.00012172187417857959,
+      "loss": 2.6445,
+      "step": 8663
+    },
+    {
+      "epoch": 0.6992171737551449,
+      "grad_norm": 0.6808965802192688,
+      "learning_rate": 0.00012170646395471296,
+      "loss": 2.5191,
+      "step": 8664
+    },
+    {
+      "epoch": 0.6992978774917279,
+      "grad_norm": 0.7063688635826111,
+      "learning_rate": 0.00012169105318986455,
+      "loss": 2.6021,
+      "step": 8665
+    },
+    {
+      "epoch": 0.6993785812283109,
+      "grad_norm": 0.6522886753082275,
+      "learning_rate": 0.0001216756418844184,
+      "loss": 2.5697,
+      "step": 8666
+    },
+    {
+      "epoch": 0.6994592849648938,
+      "grad_norm": 0.6706095337867737,
+      "learning_rate": 0.00012166023003875859,
+      "loss": 2.5706,
+      "step": 8667
+    },
+    {
+      "epoch": 0.6995399887014769,
+      "grad_norm": 0.6744416356086731,
+      "learning_rate": 0.00012164481765326923,
+      "loss": 2.5713,
+      "step": 8668
+    },
+    {
+      "epoch": 0.6996206924380599,
+      "grad_norm": 0.7385411858558655,
+      "learning_rate": 0.0001216294047283344,
+      "loss": 2.5543,
+      "step": 8669
+    },
+    {
+      "epoch": 0.6997013961746429,
+      "grad_norm": 0.7286678552627563,
+      "learning_rate": 0.0001216139912643383,
+      "loss": 2.588,
+      "step": 8670
+    },
+    {
+      "epoch": 0.6997820999112259,
+      "grad_norm": 0.7065937519073486,
+      "learning_rate": 0.00012159857726166503,
+      "loss": 2.5475,
+      "step": 8671
+    },
+    {
+      "epoch": 0.6998628036478088,
+      "grad_norm": 0.6609788537025452,
+      "learning_rate": 0.00012158316272069874,
+      "loss": 2.5664,
+      "step": 8672
+    },
+    {
+      "epoch": 0.6999435073843919,
+      "grad_norm": 0.7360579371452332,
+      "learning_rate": 0.00012156774764182364,
+      "loss": 2.5822,
+      "step": 8673
+    },
+    {
+      "epoch": 0.7000242111209749,
+      "grad_norm": 0.6265058517456055,
+      "learning_rate": 0.00012155233202542384,
+      "loss": 2.5849,
+      "step": 8674
+    },
+    {
+      "epoch": 0.7001049148575579,
+      "grad_norm": 0.646976888179779,
+      "learning_rate": 0.00012153691587188363,
+      "loss": 2.5839,
+      "step": 8675
+    },
+    {
+      "epoch": 0.7001856185941409,
+      "grad_norm": 0.6634985208511353,
+      "learning_rate": 0.0001215214991815872,
+      "loss": 2.5434,
+      "step": 8676
+    },
+    {
+      "epoch": 0.700266322330724,
+      "grad_norm": 0.6757560968399048,
+      "learning_rate": 0.00012150608195491871,
+      "loss": 2.6186,
+      "step": 8677
+    },
+    {
+      "epoch": 0.7003470260673069,
+      "grad_norm": 0.7077112197875977,
+      "learning_rate": 0.00012149066419226247,
+      "loss": 2.5757,
+      "step": 8678
+    },
+    {
+      "epoch": 0.7004277298038899,
+      "grad_norm": 0.698226273059845,
+      "learning_rate": 0.00012147524589400268,
+      "loss": 2.5307,
+      "step": 8679
+    },
+    {
+      "epoch": 0.7005084335404729,
+      "grad_norm": 0.6782405376434326,
+      "learning_rate": 0.00012145982706052361,
+      "loss": 2.5582,
+      "step": 8680
+    },
+    {
+      "epoch": 0.700589137277056,
+      "grad_norm": 0.6832882165908813,
+      "learning_rate": 0.0001214444076922096,
+      "loss": 2.574,
+      "step": 8681
+    },
+    {
+      "epoch": 0.7006698410136389,
+      "grad_norm": 0.7182612419128418,
+      "learning_rate": 0.00012142898778944485,
+      "loss": 2.6457,
+      "step": 8682
+    },
+    {
+      "epoch": 0.7007505447502219,
+      "grad_norm": 0.7043644785881042,
+      "learning_rate": 0.00012141356735261373,
+      "loss": 2.5244,
+      "step": 8683
+    },
+    {
+      "epoch": 0.7008312484868049,
+      "grad_norm": 0.6942669749259949,
+      "learning_rate": 0.00012139814638210054,
+      "loss": 2.5507,
+      "step": 8684
+    },
+    {
+      "epoch": 0.700911952223388,
+      "grad_norm": 0.8412066102027893,
+      "learning_rate": 0.00012138272487828959,
+      "loss": 2.6025,
+      "step": 8685
+    },
+    {
+      "epoch": 0.700992655959971,
+      "grad_norm": 0.6906788945198059,
+      "learning_rate": 0.00012136730284156525,
+      "loss": 2.5259,
+      "step": 8686
+    },
+    {
+      "epoch": 0.7010733596965539,
+      "grad_norm": 0.7258631587028503,
+      "learning_rate": 0.00012135188027231188,
+      "loss": 2.6311,
+      "step": 8687
+    },
+    {
+      "epoch": 0.7011540634331369,
+      "grad_norm": 0.6294744610786438,
+      "learning_rate": 0.00012133645717091382,
+      "loss": 2.5969,
+      "step": 8688
+    },
+    {
+      "epoch": 0.70123476716972,
+      "grad_norm": 0.6994131207466125,
+      "learning_rate": 0.00012132103353775548,
+      "loss": 2.5954,
+      "step": 8689
+    },
+    {
+      "epoch": 0.701315470906303,
+      "grad_norm": 0.671441912651062,
+      "learning_rate": 0.00012130560937322124,
+      "loss": 2.5628,
+      "step": 8690
+    },
+    {
+      "epoch": 0.701396174642886,
+      "grad_norm": 0.6915482878684998,
+      "learning_rate": 0.00012129018467769555,
+      "loss": 2.5173,
+      "step": 8691
+    },
+    {
+      "epoch": 0.7014768783794689,
+      "grad_norm": 0.6810318231582642,
+      "learning_rate": 0.00012127475945156279,
+      "loss": 2.6186,
+      "step": 8692
+    },
+    {
+      "epoch": 0.701557582116052,
+      "grad_norm": 0.7931910157203674,
+      "learning_rate": 0.00012125933369520741,
+      "loss": 2.6243,
+      "step": 8693
+    },
+    {
+      "epoch": 0.701638285852635,
+      "grad_norm": 0.6843162178993225,
+      "learning_rate": 0.00012124390740901386,
+      "loss": 2.6072,
+      "step": 8694
+    },
+    {
+      "epoch": 0.701718989589218,
+      "grad_norm": 0.672115683555603,
+      "learning_rate": 0.0001212284805933666,
+      "loss": 2.6027,
+      "step": 8695
+    },
+    {
+      "epoch": 0.7017996933258009,
+      "grad_norm": 0.65242600440979,
+      "learning_rate": 0.00012121305324865014,
+      "loss": 2.5128,
+      "step": 8696
+    },
+    {
+      "epoch": 0.701880397062384,
+      "grad_norm": 0.7253173589706421,
+      "learning_rate": 0.00012119762537524893,
+      "loss": 2.5776,
+      "step": 8697
+    },
+    {
+      "epoch": 0.701961100798967,
+      "grad_norm": 0.6536431312561035,
+      "learning_rate": 0.00012118219697354745,
+      "loss": 2.5656,
+      "step": 8698
+    },
+    {
+      "epoch": 0.70204180453555,
+      "grad_norm": 0.7121500372886658,
+      "learning_rate": 0.00012116676804393028,
+      "loss": 2.5878,
+      "step": 8699
+    },
+    {
+      "epoch": 0.702122508272133,
+      "grad_norm": 0.676449716091156,
+      "learning_rate": 0.00012115133858678191,
+      "loss": 2.6624,
+      "step": 8700
+    },
+    {
+      "epoch": 0.702203212008716,
+      "grad_norm": 0.7230382561683655,
+      "learning_rate": 0.0001211359086024869,
+      "loss": 2.5461,
+      "step": 8701
+    },
+    {
+      "epoch": 0.702283915745299,
+      "grad_norm": 0.6679937839508057,
+      "learning_rate": 0.00012112047809142979,
+      "loss": 2.5568,
+      "step": 8702
+    },
+    {
+      "epoch": 0.702364619481882,
+      "grad_norm": 0.6627704501152039,
+      "learning_rate": 0.0001211050470539952,
+      "loss": 2.4819,
+      "step": 8703
+    },
+    {
+      "epoch": 0.702445323218465,
+      "grad_norm": 0.6680646538734436,
+      "learning_rate": 0.0001210896154905676,
+      "loss": 2.5722,
+      "step": 8704
+    },
+    {
+      "epoch": 0.7025260269550481,
+      "grad_norm": 0.7406336665153503,
+      "learning_rate": 0.00012107418340153167,
+      "loss": 2.5722,
+      "step": 8705
+    },
+    {
+      "epoch": 0.702606730691631,
+      "grad_norm": 0.6634557247161865,
+      "learning_rate": 0.00012105875078727203,
+      "loss": 2.5747,
+      "step": 8706
+    },
+    {
+      "epoch": 0.702687434428214,
+      "grad_norm": 0.6521568894386292,
+      "learning_rate": 0.00012104331764817325,
+      "loss": 2.555,
+      "step": 8707
+    },
+    {
+      "epoch": 0.702768138164797,
+      "grad_norm": 0.677606463432312,
+      "learning_rate": 0.00012102788398461999,
+      "loss": 2.5544,
+      "step": 8708
+    },
+    {
+      "epoch": 0.7028488419013801,
+      "grad_norm": 0.6593700051307678,
+      "learning_rate": 0.0001210124497969969,
+      "loss": 2.5252,
+      "step": 8709
+    },
+    {
+      "epoch": 0.7029295456379631,
+      "grad_norm": 0.686903715133667,
+      "learning_rate": 0.00012099701508568863,
+      "loss": 2.6513,
+      "step": 8710
+    },
+    {
+      "epoch": 0.703010249374546,
+      "grad_norm": 0.6395620107650757,
+      "learning_rate": 0.00012098157985107987,
+      "loss": 2.5169,
+      "step": 8711
+    },
+    {
+      "epoch": 0.703090953111129,
+      "grad_norm": 0.7387555837631226,
+      "learning_rate": 0.00012096614409355526,
+      "loss": 2.5741,
+      "step": 8712
+    },
+    {
+      "epoch": 0.7031716568477121,
+      "grad_norm": 0.665900707244873,
+      "learning_rate": 0.00012095070781349957,
+      "loss": 2.5068,
+      "step": 8713
+    },
+    {
+      "epoch": 0.7032523605842951,
+      "grad_norm": 0.6983458399772644,
+      "learning_rate": 0.00012093527101129745,
+      "loss": 2.5028,
+      "step": 8714
+    },
+    {
+      "epoch": 0.703333064320878,
+      "grad_norm": 0.6250826120376587,
+      "learning_rate": 0.00012091983368733366,
+      "loss": 2.5765,
+      "step": 8715
+    },
+    {
+      "epoch": 0.703413768057461,
+      "grad_norm": 0.7031501531600952,
+      "learning_rate": 0.00012090439584199294,
+      "loss": 2.5885,
+      "step": 8716
+    },
+    {
+      "epoch": 0.7034944717940441,
+      "grad_norm": 0.7140926122665405,
+      "learning_rate": 0.00012088895747566002,
+      "loss": 2.6278,
+      "step": 8717
+    },
+    {
+      "epoch": 0.7035751755306271,
+      "grad_norm": 0.6753602027893066,
+      "learning_rate": 0.00012087351858871969,
+      "loss": 2.5664,
+      "step": 8718
+    },
+    {
+      "epoch": 0.7036558792672101,
+      "grad_norm": 0.7150039076805115,
+      "learning_rate": 0.0001208580791815567,
+      "loss": 2.6739,
+      "step": 8719
+    },
+    {
+      "epoch": 0.703736583003793,
+      "grad_norm": 0.7120389342308044,
+      "learning_rate": 0.00012084263925455583,
+      "loss": 2.565,
+      "step": 8720
+    },
+    {
+      "epoch": 0.703817286740376,
+      "grad_norm": 0.7775784134864807,
+      "learning_rate": 0.00012082719880810194,
+      "loss": 2.5861,
+      "step": 8721
+    },
+    {
+      "epoch": 0.7038979904769591,
+      "grad_norm": 0.6704322695732117,
+      "learning_rate": 0.0001208117578425798,
+      "loss": 2.5957,
+      "step": 8722
+    },
+    {
+      "epoch": 0.7039786942135421,
+      "grad_norm": 0.6761276721954346,
+      "learning_rate": 0.00012079631635837426,
+      "loss": 2.5472,
+      "step": 8723
+    },
+    {
+      "epoch": 0.7040593979501251,
+      "grad_norm": 0.7639868855476379,
+      "learning_rate": 0.00012078087435587016,
+      "loss": 2.6053,
+      "step": 8724
+    },
+    {
+      "epoch": 0.704140101686708,
+      "grad_norm": 0.7490074038505554,
+      "learning_rate": 0.0001207654318354523,
+      "loss": 2.5517,
+      "step": 8725
+    },
+    {
+      "epoch": 0.7042208054232911,
+      "grad_norm": 0.7068852782249451,
+      "learning_rate": 0.00012074998879750566,
+      "loss": 2.5357,
+      "step": 8726
+    },
+    {
+      "epoch": 0.7043015091598741,
+      "grad_norm": 0.7273775935173035,
+      "learning_rate": 0.00012073454524241503,
+      "loss": 2.6028,
+      "step": 8727
+    },
+    {
+      "epoch": 0.7043822128964571,
+      "grad_norm": 0.7146363258361816,
+      "learning_rate": 0.00012071910117056533,
+      "loss": 2.5982,
+      "step": 8728
+    },
+    {
+      "epoch": 0.7044629166330401,
+      "grad_norm": 0.7631390690803528,
+      "learning_rate": 0.00012070365658234149,
+      "loss": 2.6021,
+      "step": 8729
+    },
+    {
+      "epoch": 0.7045436203696231,
+      "grad_norm": 0.7065283060073853,
+      "learning_rate": 0.00012068821147812839,
+      "loss": 2.5538,
+      "step": 8730
+    },
+    {
+      "epoch": 0.7046243241062061,
+      "grad_norm": 0.7914319634437561,
+      "learning_rate": 0.00012067276585831097,
+      "loss": 2.5617,
+      "step": 8731
+    },
+    {
+      "epoch": 0.7047050278427891,
+      "grad_norm": 0.7036565542221069,
+      "learning_rate": 0.0001206573197232742,
+      "loss": 2.5354,
+      "step": 8732
+    },
+    {
+      "epoch": 0.7047857315793721,
+      "grad_norm": 0.657116711139679,
+      "learning_rate": 0.00012064187307340303,
+      "loss": 2.5084,
+      "step": 8733
+    },
+    {
+      "epoch": 0.7048664353159552,
+      "grad_norm": 0.7246817946434021,
+      "learning_rate": 0.00012062642590908242,
+      "loss": 2.5737,
+      "step": 8734
+    },
+    {
+      "epoch": 0.7049471390525381,
+      "grad_norm": 0.6895857453346252,
+      "learning_rate": 0.00012061097823069736,
+      "loss": 2.5792,
+      "step": 8735
+    },
+    {
+      "epoch": 0.7050278427891211,
+      "grad_norm": 0.7654988169670105,
+      "learning_rate": 0.00012059553003863282,
+      "loss": 2.5302,
+      "step": 8736
+    },
+    {
+      "epoch": 0.7051085465257041,
+      "grad_norm": 0.7611668109893799,
+      "learning_rate": 0.00012058008133327387,
+      "loss": 2.6073,
+      "step": 8737
+    },
+    {
+      "epoch": 0.7051892502622872,
+      "grad_norm": 0.728729784488678,
+      "learning_rate": 0.00012056463211500546,
+      "loss": 2.5714,
+      "step": 8738
+    },
+    {
+      "epoch": 0.7052699539988702,
+      "grad_norm": 0.7251634001731873,
+      "learning_rate": 0.00012054918238421271,
+      "loss": 2.627,
+      "step": 8739
+    },
+    {
+      "epoch": 0.7053506577354531,
+      "grad_norm": 0.827745795249939,
+      "learning_rate": 0.00012053373214128056,
+      "loss": 2.6303,
+      "step": 8740
+    },
+    {
+      "epoch": 0.7054313614720361,
+      "grad_norm": 0.6837510466575623,
+      "learning_rate": 0.00012051828138659416,
+      "loss": 2.5837,
+      "step": 8741
+    },
+    {
+      "epoch": 0.7055120652086192,
+      "grad_norm": 0.6763553619384766,
+      "learning_rate": 0.00012050283012053856,
+      "loss": 2.575,
+      "step": 8742
+    },
+    {
+      "epoch": 0.7055927689452022,
+      "grad_norm": 0.6779605150222778,
+      "learning_rate": 0.00012048737834349886,
+      "loss": 2.588,
+      "step": 8743
+    },
+    {
+      "epoch": 0.7056734726817852,
+      "grad_norm": 0.7207251191139221,
+      "learning_rate": 0.00012047192605586008,
+      "loss": 2.6182,
+      "step": 8744
+    },
+    {
+      "epoch": 0.7057541764183681,
+      "grad_norm": 0.6681165099143982,
+      "learning_rate": 0.00012045647325800742,
+      "loss": 2.5595,
+      "step": 8745
+    },
+    {
+      "epoch": 0.7058348801549512,
+      "grad_norm": 0.7520970702171326,
+      "learning_rate": 0.00012044101995032594,
+      "loss": 2.6306,
+      "step": 8746
+    },
+    {
+      "epoch": 0.7059155838915342,
+      "grad_norm": 0.7148429155349731,
+      "learning_rate": 0.00012042556613320087,
+      "loss": 2.5749,
+      "step": 8747
+    },
+    {
+      "epoch": 0.7059962876281172,
+      "grad_norm": 0.619369626045227,
+      "learning_rate": 0.00012041011180701729,
+      "loss": 2.5382,
+      "step": 8748
+    },
+    {
+      "epoch": 0.7060769913647001,
+      "grad_norm": 0.7450816035270691,
+      "learning_rate": 0.00012039465697216032,
+      "loss": 2.5547,
+      "step": 8749
+    },
+    {
+      "epoch": 0.7061576951012832,
+      "grad_norm": 0.7324537634849548,
+      "learning_rate": 0.00012037920162901521,
+      "loss": 2.5756,
+      "step": 8750
+    },
+    {
+      "epoch": 0.7062383988378662,
+      "grad_norm": 0.7881754636764526,
+      "learning_rate": 0.00012036374577796715,
+      "loss": 2.6376,
+      "step": 8751
+    },
+    {
+      "epoch": 0.7063191025744492,
+      "grad_norm": 0.7095965147018433,
+      "learning_rate": 0.00012034828941940128,
+      "loss": 2.5454,
+      "step": 8752
+    },
+    {
+      "epoch": 0.7063998063110322,
+      "grad_norm": 0.7142949104309082,
+      "learning_rate": 0.00012033283255370287,
+      "loss": 2.5738,
+      "step": 8753
+    },
+    {
+      "epoch": 0.7064805100476153,
+      "grad_norm": 0.6592378616333008,
+      "learning_rate": 0.0001203173751812571,
+      "loss": 2.5473,
+      "step": 8754
+    },
+    {
+      "epoch": 0.7065612137841982,
+      "grad_norm": 0.6964332461357117,
+      "learning_rate": 0.00012030191730244926,
+      "loss": 2.5829,
+      "step": 8755
+    },
+    {
+      "epoch": 0.7066419175207812,
+      "grad_norm": 0.707539975643158,
+      "learning_rate": 0.00012028645891766455,
+      "loss": 2.5652,
+      "step": 8756
+    },
+    {
+      "epoch": 0.7067226212573642,
+      "grad_norm": 0.6991387009620667,
+      "learning_rate": 0.00012027100002728824,
+      "loss": 2.5874,
+      "step": 8757
+    },
+    {
+      "epoch": 0.7068033249939473,
+      "grad_norm": 0.665746808052063,
+      "learning_rate": 0.00012025554063170566,
+      "loss": 2.5163,
+      "step": 8758
+    },
+    {
+      "epoch": 0.7068840287305302,
+      "grad_norm": 0.696130096912384,
+      "learning_rate": 0.00012024008073130204,
+      "loss": 2.5748,
+      "step": 8759
+    },
+    {
+      "epoch": 0.7069647324671132,
+      "grad_norm": 0.698885440826416,
+      "learning_rate": 0.00012022462032646269,
+      "loss": 2.5561,
+      "step": 8760
+    },
+    {
+      "epoch": 0.7070454362036962,
+      "grad_norm": 0.7052211761474609,
+      "learning_rate": 0.00012020915941757292,
+      "loss": 2.5979,
+      "step": 8761
+    },
+    {
+      "epoch": 0.7071261399402793,
+      "grad_norm": 0.7370811104774475,
+      "learning_rate": 0.00012019369800501808,
+      "loss": 2.5623,
+      "step": 8762
+    },
+    {
+      "epoch": 0.7072068436768623,
+      "grad_norm": 0.6699148416519165,
+      "learning_rate": 0.00012017823608918352,
+      "loss": 2.5816,
+      "step": 8763
+    },
+    {
+      "epoch": 0.7072875474134452,
+      "grad_norm": 0.6712930798530579,
+      "learning_rate": 0.00012016277367045457,
+      "loss": 2.5495,
+      "step": 8764
+    },
+    {
+      "epoch": 0.7073682511500282,
+      "grad_norm": 0.7238204479217529,
+      "learning_rate": 0.00012014731074921659,
+      "loss": 2.5936,
+      "step": 8765
+    },
+    {
+      "epoch": 0.7074489548866113,
+      "grad_norm": 0.7303668856620789,
+      "learning_rate": 0.00012013184732585494,
+      "loss": 2.6366,
+      "step": 8766
+    },
+    {
+      "epoch": 0.7075296586231943,
+      "grad_norm": 0.6883132457733154,
+      "learning_rate": 0.00012011638340075505,
+      "loss": 2.534,
+      "step": 8767
+    },
+    {
+      "epoch": 0.7076103623597773,
+      "grad_norm": 0.7057133316993713,
+      "learning_rate": 0.00012010091897430229,
+      "loss": 2.6035,
+      "step": 8768
+    },
+    {
+      "epoch": 0.7076910660963602,
+      "grad_norm": 0.7069352269172668,
+      "learning_rate": 0.0001200854540468821,
+      "loss": 2.5047,
+      "step": 8769
+    },
+    {
+      "epoch": 0.7077717698329433,
+      "grad_norm": 0.7192478775978088,
+      "learning_rate": 0.00012006998861887985,
+      "loss": 2.5698,
+      "step": 8770
+    },
+    {
+      "epoch": 0.7078524735695263,
+      "grad_norm": 0.6992887854576111,
+      "learning_rate": 0.00012005452269068107,
+      "loss": 2.5631,
+      "step": 8771
+    },
+    {
+      "epoch": 0.7079331773061093,
+      "grad_norm": 0.676154613494873,
+      "learning_rate": 0.00012003905626267114,
+      "loss": 2.5255,
+      "step": 8772
+    },
+    {
+      "epoch": 0.7080138810426923,
+      "grad_norm": 0.672269880771637,
+      "learning_rate": 0.00012002358933523555,
+      "loss": 2.5766,
+      "step": 8773
+    },
+    {
+      "epoch": 0.7080945847792752,
+      "grad_norm": 0.7334566712379456,
+      "learning_rate": 0.00012000812190875976,
+      "loss": 2.6068,
+      "step": 8774
+    },
+    {
+      "epoch": 0.7081752885158583,
+      "grad_norm": 0.6599388122558594,
+      "learning_rate": 0.00011999265398362931,
+      "loss": 2.6032,
+      "step": 8775
+    },
+    {
+      "epoch": 0.7082559922524413,
+      "grad_norm": 0.7158498167991638,
+      "learning_rate": 0.00011997718556022958,
+      "loss": 2.599,
+      "step": 8776
+    },
+    {
+      "epoch": 0.7083366959890243,
+      "grad_norm": 0.7470360994338989,
+      "learning_rate": 0.00011996171663894624,
+      "loss": 2.58,
+      "step": 8777
+    },
+    {
+      "epoch": 0.7084173997256072,
+      "grad_norm": 0.6251266002655029,
+      "learning_rate": 0.00011994624722016472,
+      "loss": 2.5996,
+      "step": 8778
+    },
+    {
+      "epoch": 0.7084981034621903,
+      "grad_norm": 0.6649689078330994,
+      "learning_rate": 0.00011993077730427058,
+      "loss": 2.6025,
+      "step": 8779
+    },
+    {
+      "epoch": 0.7085788071987733,
+      "grad_norm": 0.7554693818092346,
+      "learning_rate": 0.00011991530689164939,
+      "loss": 2.6207,
+      "step": 8780
+    },
+    {
+      "epoch": 0.7086595109353563,
+      "grad_norm": 0.7941430807113647,
+      "learning_rate": 0.00011989983598268661,
+      "loss": 2.584,
+      "step": 8781
+    },
+    {
+      "epoch": 0.7087402146719393,
+      "grad_norm": 0.7257998585700989,
+      "learning_rate": 0.00011988436457776799,
+      "loss": 2.6152,
+      "step": 8782
+    },
+    {
+      "epoch": 0.7088209184085223,
+      "grad_norm": 0.716354489326477,
+      "learning_rate": 0.00011986889267727899,
+      "loss": 2.585,
+      "step": 8783
+    },
+    {
+      "epoch": 0.7089016221451053,
+      "grad_norm": 0.7094400525093079,
+      "learning_rate": 0.00011985342028160525,
+      "loss": 2.5759,
+      "step": 8784
+    },
+    {
+      "epoch": 0.7089823258816883,
+      "grad_norm": 0.7211421728134155,
+      "learning_rate": 0.0001198379473911324,
+      "loss": 2.5645,
+      "step": 8785
+    },
+    {
+      "epoch": 0.7090630296182713,
+      "grad_norm": 0.7166693806648254,
+      "learning_rate": 0.000119822474006246,
+      "loss": 2.5357,
+      "step": 8786
+    },
+    {
+      "epoch": 0.7091437333548544,
+      "grad_norm": 0.6702254414558411,
+      "learning_rate": 0.00011980700012733175,
+      "loss": 2.5353,
+      "step": 8787
+    },
+    {
+      "epoch": 0.7092244370914373,
+      "grad_norm": 0.6784049868583679,
+      "learning_rate": 0.0001197915257547753,
+      "loss": 2.4942,
+      "step": 8788
+    },
+    {
+      "epoch": 0.7093051408280203,
+      "grad_norm": 0.6914299726486206,
+      "learning_rate": 0.00011977605088896226,
+      "loss": 2.5682,
+      "step": 8789
+    },
+    {
+      "epoch": 0.7093858445646033,
+      "grad_norm": 0.7324358820915222,
+      "learning_rate": 0.00011976057553027837,
+      "loss": 2.564,
+      "step": 8790
+    },
+    {
+      "epoch": 0.7094665483011864,
+      "grad_norm": 0.6927928924560547,
+      "learning_rate": 0.00011974509967910927,
+      "loss": 2.5728,
+      "step": 8791
+    },
+    {
+      "epoch": 0.7095472520377694,
+      "grad_norm": 0.6795603036880493,
+      "learning_rate": 0.00011972962333584066,
+      "loss": 2.588,
+      "step": 8792
+    },
+    {
+      "epoch": 0.7096279557743523,
+      "grad_norm": 0.7132226228713989,
+      "learning_rate": 0.00011971414650085828,
+      "loss": 2.5759,
+      "step": 8793
+    },
+    {
+      "epoch": 0.7097086595109353,
+      "grad_norm": 0.737195611000061,
+      "learning_rate": 0.00011969866917454782,
+      "loss": 2.5721,
+      "step": 8794
+    },
+    {
+      "epoch": 0.7097893632475184,
+      "grad_norm": 0.6776021718978882,
+      "learning_rate": 0.00011968319135729507,
+      "loss": 2.5794,
+      "step": 8795
+    },
+    {
+      "epoch": 0.7098700669841014,
+      "grad_norm": 0.7113735675811768,
+      "learning_rate": 0.0001196677130494857,
+      "loss": 2.5595,
+      "step": 8796
+    },
+    {
+      "epoch": 0.7099507707206844,
+      "grad_norm": 0.6277747750282288,
+      "learning_rate": 0.0001196522342515055,
+      "loss": 2.5003,
+      "step": 8797
+    },
+    {
+      "epoch": 0.7100314744572673,
+      "grad_norm": 0.6982879042625427,
+      "learning_rate": 0.00011963675496374028,
+      "loss": 2.542,
+      "step": 8798
+    },
+    {
+      "epoch": 0.7101121781938504,
+      "grad_norm": 0.7019705176353455,
+      "learning_rate": 0.00011962127518657578,
+      "loss": 2.5723,
+      "step": 8799
+    },
+    {
+      "epoch": 0.7101928819304334,
+      "grad_norm": 0.6831088662147522,
+      "learning_rate": 0.00011960579492039783,
+      "loss": 2.5676,
+      "step": 8800
+    },
+    {
+      "epoch": 0.7102735856670164,
+      "grad_norm": 0.6744031310081482,
+      "learning_rate": 0.0001195903141655922,
+      "loss": 2.58,
+      "step": 8801
+    },
+    {
+      "epoch": 0.7103542894035993,
+      "grad_norm": 0.6873177289962769,
+      "learning_rate": 0.00011957483292254473,
+      "loss": 2.6289,
+      "step": 8802
+    },
+    {
+      "epoch": 0.7104349931401824,
+      "grad_norm": 0.6340685486793518,
+      "learning_rate": 0.00011955935119164125,
+      "loss": 2.5688,
+      "step": 8803
+    },
+    {
+      "epoch": 0.7105156968767654,
+      "grad_norm": 0.7147708535194397,
+      "learning_rate": 0.00011954386897326764,
+      "loss": 2.5471,
+      "step": 8804
+    },
+    {
+      "epoch": 0.7105964006133484,
+      "grad_norm": 0.699605405330658,
+      "learning_rate": 0.00011952838626780971,
+      "loss": 2.6122,
+      "step": 8805
+    },
+    {
+      "epoch": 0.7106771043499314,
+      "grad_norm": 0.6685385704040527,
+      "learning_rate": 0.00011951290307565335,
+      "loss": 2.5423,
+      "step": 8806
+    },
+    {
+      "epoch": 0.7107578080865145,
+      "grad_norm": 0.6884726881980896,
+      "learning_rate": 0.00011949741939718439,
+      "loss": 2.5243,
+      "step": 8807
+    },
+    {
+      "epoch": 0.7108385118230974,
+      "grad_norm": 0.6991142630577087,
+      "learning_rate": 0.00011948193523278884,
+      "loss": 2.6271,
+      "step": 8808
+    },
+    {
+      "epoch": 0.7109192155596804,
+      "grad_norm": 0.6964353919029236,
+      "learning_rate": 0.00011946645058285253,
+      "loss": 2.6296,
+      "step": 8809
+    },
+    {
+      "epoch": 0.7109999192962634,
+      "grad_norm": 0.7592040300369263,
+      "learning_rate": 0.00011945096544776136,
+      "loss": 2.6601,
+      "step": 8810
+    },
+    {
+      "epoch": 0.7110806230328465,
+      "grad_norm": 0.7146934866905212,
+      "learning_rate": 0.00011943547982790131,
+      "loss": 2.54,
+      "step": 8811
+    },
+    {
+      "epoch": 0.7111613267694294,
+      "grad_norm": 0.6991123557090759,
+      "learning_rate": 0.00011941999372365827,
+      "loss": 2.5978,
+      "step": 8812
+    },
+    {
+      "epoch": 0.7112420305060124,
+      "grad_norm": 0.6835920810699463,
+      "learning_rate": 0.00011940450713541822,
+      "loss": 2.6096,
+      "step": 8813
+    },
+    {
+      "epoch": 0.7113227342425954,
+      "grad_norm": 0.6913917660713196,
+      "learning_rate": 0.00011938902006356716,
+      "loss": 2.5624,
+      "step": 8814
+    },
+    {
+      "epoch": 0.7114034379791785,
+      "grad_norm": 0.6620622873306274,
+      "learning_rate": 0.00011937353250849102,
+      "loss": 2.6211,
+      "step": 8815
+    },
+    {
+      "epoch": 0.7114841417157615,
+      "grad_norm": 0.6738792061805725,
+      "learning_rate": 0.00011935804447057581,
+      "loss": 2.5889,
+      "step": 8816
+    },
+    {
+      "epoch": 0.7115648454523444,
+      "grad_norm": 0.7101936936378479,
+      "learning_rate": 0.00011934255595020751,
+      "loss": 2.5846,
+      "step": 8817
+    },
+    {
+      "epoch": 0.7116455491889274,
+      "grad_norm": 0.6843911409378052,
+      "learning_rate": 0.00011932706694777216,
+      "loss": 2.5757,
+      "step": 8818
+    },
+    {
+      "epoch": 0.7117262529255105,
+      "grad_norm": 0.7217971086502075,
+      "learning_rate": 0.0001193115774636558,
+      "loss": 2.6174,
+      "step": 8819
+    },
+    {
+      "epoch": 0.7118069566620935,
+      "grad_norm": 0.6706245541572571,
+      "learning_rate": 0.00011929608749824445,
+      "loss": 2.5893,
+      "step": 8820
+    },
+    {
+      "epoch": 0.7118876603986765,
+      "grad_norm": 0.7057672739028931,
+      "learning_rate": 0.00011928059705192413,
+      "loss": 2.5426,
+      "step": 8821
+    },
+    {
+      "epoch": 0.7119683641352594,
+      "grad_norm": 0.7354697585105896,
+      "learning_rate": 0.00011926510612508095,
+      "loss": 2.5741,
+      "step": 8822
+    },
+    {
+      "epoch": 0.7120490678718424,
+      "grad_norm": 0.6618186235427856,
+      "learning_rate": 0.00011924961471810096,
+      "loss": 2.6007,
+      "step": 8823
+    },
+    {
+      "epoch": 0.7121297716084255,
+      "grad_norm": 0.6733995676040649,
+      "learning_rate": 0.00011923412283137028,
+      "loss": 2.5739,
+      "step": 8824
+    },
+    {
+      "epoch": 0.7122104753450085,
+      "grad_norm": 0.7324833869934082,
+      "learning_rate": 0.00011921863046527497,
+      "loss": 2.5461,
+      "step": 8825
+    },
+    {
+      "epoch": 0.7122911790815915,
+      "grad_norm": 0.6753048896789551,
+      "learning_rate": 0.00011920313762020113,
+      "loss": 2.5066,
+      "step": 8826
+    },
+    {
+      "epoch": 0.7123718828181744,
+      "grad_norm": 0.7861250638961792,
+      "learning_rate": 0.00011918764429653489,
+      "loss": 2.5229,
+      "step": 8827
+    },
+    {
+      "epoch": 0.7124525865547575,
+      "grad_norm": 0.7037342190742493,
+      "learning_rate": 0.00011917215049466244,
+      "loss": 2.5443,
+      "step": 8828
+    },
+    {
+      "epoch": 0.7125332902913405,
+      "grad_norm": 0.7112773060798645,
+      "learning_rate": 0.00011915665621496985,
+      "loss": 2.5656,
+      "step": 8829
+    },
+    {
+      "epoch": 0.7126139940279235,
+      "grad_norm": 0.6384316682815552,
+      "learning_rate": 0.00011914116145784333,
+      "loss": 2.5526,
+      "step": 8830
+    },
+    {
+      "epoch": 0.7126946977645064,
+      "grad_norm": 0.6673600077629089,
+      "learning_rate": 0.000119125666223669,
+      "loss": 2.5868,
+      "step": 8831
+    },
+    {
+      "epoch": 0.7127754015010895,
+      "grad_norm": 0.6927722692489624,
+      "learning_rate": 0.0001191101705128331,
+      "loss": 2.6237,
+      "step": 8832
+    },
+    {
+      "epoch": 0.7128561052376725,
+      "grad_norm": 0.7410106658935547,
+      "learning_rate": 0.00011909467432572182,
+      "loss": 2.5652,
+      "step": 8833
+    },
+    {
+      "epoch": 0.7129368089742555,
+      "grad_norm": 0.6780139803886414,
+      "learning_rate": 0.0001190791776627213,
+      "loss": 2.5343,
+      "step": 8834
+    },
+    {
+      "epoch": 0.7130175127108385,
+      "grad_norm": 0.7147949934005737,
+      "learning_rate": 0.00011906368052421781,
+      "loss": 2.5368,
+      "step": 8835
+    },
+    {
+      "epoch": 0.7130982164474216,
+      "grad_norm": 0.7092324495315552,
+      "learning_rate": 0.00011904818291059759,
+      "loss": 2.538,
+      "step": 8836
+    },
+    {
+      "epoch": 0.7131789201840045,
+      "grad_norm": 0.761763870716095,
+      "learning_rate": 0.00011903268482224684,
+      "loss": 2.5984,
+      "step": 8837
+    },
+    {
+      "epoch": 0.7132596239205875,
+      "grad_norm": 0.7011365294456482,
+      "learning_rate": 0.00011901718625955182,
+      "loss": 2.5383,
+      "step": 8838
+    },
+    {
+      "epoch": 0.7133403276571705,
+      "grad_norm": 0.7982703447341919,
+      "learning_rate": 0.00011900168722289882,
+      "loss": 2.5714,
+      "step": 8839
+    },
+    {
+      "epoch": 0.7134210313937536,
+      "grad_norm": 0.6788253784179688,
+      "learning_rate": 0.00011898618771267412,
+      "loss": 2.5675,
+      "step": 8840
+    },
+    {
+      "epoch": 0.7135017351303365,
+      "grad_norm": 0.6245018243789673,
+      "learning_rate": 0.00011897068772926397,
+      "loss": 2.5497,
+      "step": 8841
+    },
+    {
+      "epoch": 0.7135824388669195,
+      "grad_norm": 0.732109785079956,
+      "learning_rate": 0.0001189551872730547,
+      "loss": 2.5043,
+      "step": 8842
+    },
+    {
+      "epoch": 0.7136631426035025,
+      "grad_norm": 0.7640885710716248,
+      "learning_rate": 0.0001189396863444326,
+      "loss": 2.5974,
+      "step": 8843
+    },
+    {
+      "epoch": 0.7137438463400856,
+      "grad_norm": 0.6806808710098267,
+      "learning_rate": 0.00011892418494378403,
+      "loss": 2.5911,
+      "step": 8844
+    },
+    {
+      "epoch": 0.7138245500766686,
+      "grad_norm": 0.6730000376701355,
+      "learning_rate": 0.00011890868307149528,
+      "loss": 2.5405,
+      "step": 8845
+    },
+    {
+      "epoch": 0.7139052538132515,
+      "grad_norm": 0.6881929636001587,
+      "learning_rate": 0.00011889318072795275,
+      "loss": 2.6083,
+      "step": 8846
+    },
+    {
+      "epoch": 0.7139859575498345,
+      "grad_norm": 0.7079598307609558,
+      "learning_rate": 0.00011887767791354275,
+      "loss": 2.5743,
+      "step": 8847
+    },
+    {
+      "epoch": 0.7140666612864176,
+      "grad_norm": 0.6760475635528564,
+      "learning_rate": 0.00011886217462865166,
+      "loss": 2.5925,
+      "step": 8848
+    },
+    {
+      "epoch": 0.7141473650230006,
+      "grad_norm": 0.6851043701171875,
+      "learning_rate": 0.00011884667087366587,
+      "loss": 2.5839,
+      "step": 8849
+    },
+    {
+      "epoch": 0.7142280687595836,
+      "grad_norm": 0.6805267930030823,
+      "learning_rate": 0.00011883116664897178,
+      "loss": 2.562,
+      "step": 8850
+    },
+    {
+      "epoch": 0.7143087724961665,
+      "grad_norm": 0.6720704436302185,
+      "learning_rate": 0.00011881566195495581,
+      "loss": 2.5381,
+      "step": 8851
+    },
+    {
+      "epoch": 0.7143894762327496,
+      "grad_norm": 0.718166172504425,
+      "learning_rate": 0.00011880015679200436,
+      "loss": 2.5912,
+      "step": 8852
+    },
+    {
+      "epoch": 0.7144701799693326,
+      "grad_norm": 0.6643497943878174,
+      "learning_rate": 0.00011878465116050383,
+      "loss": 2.5122,
+      "step": 8853
+    },
+    {
+      "epoch": 0.7145508837059156,
+      "grad_norm": 0.705186665058136,
+      "learning_rate": 0.00011876914506084074,
+      "loss": 2.617,
+      "step": 8854
+    },
+    {
+      "epoch": 0.7146315874424986,
+      "grad_norm": 0.6417848467826843,
+      "learning_rate": 0.00011875363849340144,
+      "loss": 2.5552,
+      "step": 8855
+    },
+    {
+      "epoch": 0.7147122911790816,
+      "grad_norm": 0.6861358880996704,
+      "learning_rate": 0.00011873813145857249,
+      "loss": 2.6324,
+      "step": 8856
+    },
+    {
+      "epoch": 0.7147929949156646,
+      "grad_norm": 0.7134111523628235,
+      "learning_rate": 0.00011872262395674027,
+      "loss": 2.5892,
+      "step": 8857
+    },
+    {
+      "epoch": 0.7148736986522476,
+      "grad_norm": 0.7177506685256958,
+      "learning_rate": 0.00011870711598829135,
+      "loss": 2.5677,
+      "step": 8858
+    },
+    {
+      "epoch": 0.7149544023888306,
+      "grad_norm": 0.6435763835906982,
+      "learning_rate": 0.00011869160755361219,
+      "loss": 2.5452,
+      "step": 8859
+    },
+    {
+      "epoch": 0.7150351061254137,
+      "grad_norm": 0.6443132758140564,
+      "learning_rate": 0.00011867609865308935,
+      "loss": 2.5566,
+      "step": 8860
+    },
+    {
+      "epoch": 0.7151158098619966,
+      "grad_norm": 0.7132347822189331,
+      "learning_rate": 0.00011866058928710925,
+      "loss": 2.565,
+      "step": 8861
+    },
+    {
+      "epoch": 0.7151965135985796,
+      "grad_norm": 0.7803207039833069,
+      "learning_rate": 0.00011864507945605854,
+      "loss": 2.556,
+      "step": 8862
+    },
+    {
+      "epoch": 0.7152772173351626,
+      "grad_norm": 0.7277950644493103,
+      "learning_rate": 0.00011862956916032367,
+      "loss": 2.5623,
+      "step": 8863
+    },
+    {
+      "epoch": 0.7153579210717457,
+      "grad_norm": 0.6812277436256409,
+      "learning_rate": 0.00011861405840029125,
+      "loss": 2.6146,
+      "step": 8864
+    },
+    {
+      "epoch": 0.7154386248083286,
+      "grad_norm": 0.7170509099960327,
+      "learning_rate": 0.00011859854717634786,
+      "loss": 2.52,
+      "step": 8865
+    },
+    {
+      "epoch": 0.7155193285449116,
+      "grad_norm": 0.7282906174659729,
+      "learning_rate": 0.00011858303548888004,
+      "loss": 2.5605,
+      "step": 8866
+    },
+    {
+      "epoch": 0.7156000322814946,
+      "grad_norm": 0.7290246486663818,
+      "learning_rate": 0.00011856752333827439,
+      "loss": 2.6292,
+      "step": 8867
+    },
+    {
+      "epoch": 0.7156807360180777,
+      "grad_norm": 0.6870024800300598,
+      "learning_rate": 0.00011855201072491752,
+      "loss": 2.6396,
+      "step": 8868
+    },
+    {
+      "epoch": 0.7157614397546607,
+      "grad_norm": 0.7336156964302063,
+      "learning_rate": 0.00011853649764919605,
+      "loss": 2.6356,
+      "step": 8869
+    },
+    {
+      "epoch": 0.7158421434912436,
+      "grad_norm": 0.7181294560432434,
+      "learning_rate": 0.00011852098411149661,
+      "loss": 2.5163,
+      "step": 8870
+    },
+    {
+      "epoch": 0.7159228472278266,
+      "grad_norm": 0.7355513572692871,
+      "learning_rate": 0.00011850547011220583,
+      "loss": 2.5485,
+      "step": 8871
+    },
+    {
+      "epoch": 0.7160035509644097,
+      "grad_norm": 0.7005351185798645,
+      "learning_rate": 0.00011848995565171038,
+      "loss": 2.5187,
+      "step": 8872
+    },
+    {
+      "epoch": 0.7160842547009927,
+      "grad_norm": 0.6550194025039673,
+      "learning_rate": 0.00011847444073039686,
+      "loss": 2.5174,
+      "step": 8873
+    },
+    {
+      "epoch": 0.7161649584375757,
+      "grad_norm": 0.6568251252174377,
+      "learning_rate": 0.00011845892534865202,
+      "loss": 2.5128,
+      "step": 8874
+    },
+    {
+      "epoch": 0.7162456621741586,
+      "grad_norm": 0.6359419226646423,
+      "learning_rate": 0.0001184434095068625,
+      "loss": 2.5967,
+      "step": 8875
+    },
+    {
+      "epoch": 0.7163263659107416,
+      "grad_norm": 0.6730023622512817,
+      "learning_rate": 0.00011842789320541504,
+      "loss": 2.5243,
+      "step": 8876
+    },
+    {
+      "epoch": 0.7164070696473247,
+      "grad_norm": 0.6750187277793884,
+      "learning_rate": 0.00011841237644469625,
+      "loss": 2.602,
+      "step": 8877
+    },
+    {
+      "epoch": 0.7164877733839077,
+      "grad_norm": 0.7039143443107605,
+      "learning_rate": 0.00011839685922509291,
+      "loss": 2.5345,
+      "step": 8878
+    },
+    {
+      "epoch": 0.7165684771204907,
+      "grad_norm": 0.6602306962013245,
+      "learning_rate": 0.00011838134154699177,
+      "loss": 2.5995,
+      "step": 8879
+    },
+    {
+      "epoch": 0.7166491808570736,
+      "grad_norm": 0.6744598150253296,
+      "learning_rate": 0.00011836582341077955,
+      "loss": 2.6005,
+      "step": 8880
+    },
+    {
+      "epoch": 0.7167298845936567,
+      "grad_norm": 0.7136051058769226,
+      "learning_rate": 0.00011835030481684302,
+      "loss": 2.5424,
+      "step": 8881
+    },
+    {
+      "epoch": 0.7168105883302397,
+      "grad_norm": 0.7085986137390137,
+      "learning_rate": 0.00011833478576556889,
+      "loss": 2.5912,
+      "step": 8882
+    },
+    {
+      "epoch": 0.7168912920668227,
+      "grad_norm": 0.7635689377784729,
+      "learning_rate": 0.00011831926625734398,
+      "loss": 2.5836,
+      "step": 8883
+    },
+    {
+      "epoch": 0.7169719958034056,
+      "grad_norm": 0.6543256640434265,
+      "learning_rate": 0.00011830374629255508,
+      "loss": 2.5442,
+      "step": 8884
+    },
+    {
+      "epoch": 0.7170526995399887,
+      "grad_norm": 0.663840115070343,
+      "learning_rate": 0.00011828822587158896,
+      "loss": 2.5529,
+      "step": 8885
+    },
+    {
+      "epoch": 0.7171334032765717,
+      "grad_norm": 0.6868027448654175,
+      "learning_rate": 0.00011827270499483247,
+      "loss": 2.6678,
+      "step": 8886
+    },
+    {
+      "epoch": 0.7172141070131547,
+      "grad_norm": 0.649172842502594,
+      "learning_rate": 0.00011825718366267238,
+      "loss": 2.57,
+      "step": 8887
+    },
+    {
+      "epoch": 0.7172948107497377,
+      "grad_norm": 0.6818440556526184,
+      "learning_rate": 0.00011824166187549554,
+      "loss": 2.5602,
+      "step": 8888
+    },
+    {
+      "epoch": 0.7173755144863208,
+      "grad_norm": 0.7222314476966858,
+      "learning_rate": 0.00011822613963368885,
+      "loss": 2.5526,
+      "step": 8889
+    },
+    {
+      "epoch": 0.7174562182229037,
+      "grad_norm": 0.7309598922729492,
+      "learning_rate": 0.00011821061693763909,
+      "loss": 2.5515,
+      "step": 8890
+    },
+    {
+      "epoch": 0.7175369219594867,
+      "grad_norm": 0.6935746669769287,
+      "learning_rate": 0.00011819509378773314,
+      "loss": 2.5506,
+      "step": 8891
+    },
+    {
+      "epoch": 0.7176176256960697,
+      "grad_norm": 0.6754423975944519,
+      "learning_rate": 0.00011817957018435792,
+      "loss": 2.5621,
+      "step": 8892
+    },
+    {
+      "epoch": 0.7176983294326528,
+      "grad_norm": 0.7087355852127075,
+      "learning_rate": 0.00011816404612790026,
+      "loss": 2.5708,
+      "step": 8893
+    },
+    {
+      "epoch": 0.7177790331692357,
+      "grad_norm": 0.726820707321167,
+      "learning_rate": 0.0001181485216187471,
+      "loss": 2.5741,
+      "step": 8894
+    },
+    {
+      "epoch": 0.7178597369058187,
+      "grad_norm": 0.6539922952651978,
+      "learning_rate": 0.00011813299665728532,
+      "loss": 2.613,
+      "step": 8895
+    },
+    {
+      "epoch": 0.7179404406424017,
+      "grad_norm": 0.7008066773414612,
+      "learning_rate": 0.00011811747124390189,
+      "loss": 2.6029,
+      "step": 8896
+    },
+    {
+      "epoch": 0.7180211443789848,
+      "grad_norm": 0.6900522708892822,
+      "learning_rate": 0.00011810194537898374,
+      "loss": 2.5716,
+      "step": 8897
+    },
+    {
+      "epoch": 0.7181018481155678,
+      "grad_norm": 0.675345242023468,
+      "learning_rate": 0.00011808641906291776,
+      "loss": 2.5742,
+      "step": 8898
+    },
+    {
+      "epoch": 0.7181825518521507,
+      "grad_norm": 0.6697559356689453,
+      "learning_rate": 0.00011807089229609092,
+      "loss": 2.5717,
+      "step": 8899
+    },
+    {
+      "epoch": 0.7182632555887337,
+      "grad_norm": 0.6874344944953918,
+      "learning_rate": 0.00011805536507889021,
+      "loss": 2.5394,
+      "step": 8900
+    },
+    {
+      "epoch": 0.7183439593253168,
+      "grad_norm": 0.6675494313240051,
+      "learning_rate": 0.00011803983741170263,
+      "loss": 2.5655,
+      "step": 8901
+    },
+    {
+      "epoch": 0.7184246630618998,
+      "grad_norm": 0.6937244534492493,
+      "learning_rate": 0.00011802430929491517,
+      "loss": 2.5676,
+      "step": 8902
+    },
+    {
+      "epoch": 0.7185053667984828,
+      "grad_norm": 0.7591496109962463,
+      "learning_rate": 0.00011800878072891474,
+      "loss": 2.5849,
+      "step": 8903
+    },
+    {
+      "epoch": 0.7185860705350657,
+      "grad_norm": 0.6503129005432129,
+      "learning_rate": 0.00011799325171408846,
+      "loss": 2.5416,
+      "step": 8904
+    },
+    {
+      "epoch": 0.7186667742716488,
+      "grad_norm": 0.6450222134590149,
+      "learning_rate": 0.00011797772225082333,
+      "loss": 2.5395,
+      "step": 8905
+    },
+    {
+      "epoch": 0.7187474780082318,
+      "grad_norm": 0.7317619919776917,
+      "learning_rate": 0.00011796219233950632,
+      "loss": 2.609,
+      "step": 8906
+    },
+    {
+      "epoch": 0.7188281817448148,
+      "grad_norm": 0.7585787773132324,
+      "learning_rate": 0.00011794666198052455,
+      "loss": 2.5556,
+      "step": 8907
+    },
+    {
+      "epoch": 0.7189088854813978,
+      "grad_norm": 0.6718214750289917,
+      "learning_rate": 0.00011793113117426505,
+      "loss": 2.5914,
+      "step": 8908
+    },
+    {
+      "epoch": 0.7189895892179808,
+      "grad_norm": 0.6459314823150635,
+      "learning_rate": 0.00011791559992111487,
+      "loss": 2.5956,
+      "step": 8909
+    },
+    {
+      "epoch": 0.7190702929545638,
+      "grad_norm": 0.6592775583267212,
+      "learning_rate": 0.00011790006822146113,
+      "loss": 2.5568,
+      "step": 8910
+    },
+    {
+      "epoch": 0.7191509966911468,
+      "grad_norm": 0.7277452349662781,
+      "learning_rate": 0.0001178845360756909,
+      "loss": 2.5989,
+      "step": 8911
+    },
+    {
+      "epoch": 0.7192317004277298,
+      "grad_norm": 0.7020131945610046,
+      "learning_rate": 0.00011786900348419128,
+      "loss": 2.645,
+      "step": 8912
+    },
+    {
+      "epoch": 0.7193124041643129,
+      "grad_norm": 0.6746636629104614,
+      "learning_rate": 0.00011785347044734938,
+      "loss": 2.5173,
+      "step": 8913
+    },
+    {
+      "epoch": 0.7193931079008958,
+      "grad_norm": 0.6782798171043396,
+      "learning_rate": 0.0001178379369655523,
+      "loss": 2.6007,
+      "step": 8914
+    },
+    {
+      "epoch": 0.7194738116374788,
+      "grad_norm": 0.705498218536377,
+      "learning_rate": 0.00011782240303918724,
+      "loss": 2.5408,
+      "step": 8915
+    },
+    {
+      "epoch": 0.7195545153740618,
+      "grad_norm": 0.675532341003418,
+      "learning_rate": 0.00011780686866864128,
+      "loss": 2.5188,
+      "step": 8916
+    },
+    {
+      "epoch": 0.7196352191106449,
+      "grad_norm": 0.6552390456199646,
+      "learning_rate": 0.00011779133385430161,
+      "loss": 2.5409,
+      "step": 8917
+    },
+    {
+      "epoch": 0.7197159228472279,
+      "grad_norm": 0.6589654088020325,
+      "learning_rate": 0.00011777579859655544,
+      "loss": 2.5447,
+      "step": 8918
+    },
+    {
+      "epoch": 0.7197966265838108,
+      "grad_norm": 0.7548382878303528,
+      "learning_rate": 0.00011776026289578985,
+      "loss": 2.5239,
+      "step": 8919
+    },
+    {
+      "epoch": 0.7198773303203938,
+      "grad_norm": 0.697325587272644,
+      "learning_rate": 0.00011774472675239207,
+      "loss": 2.5887,
+      "step": 8920
+    },
+    {
+      "epoch": 0.7199580340569769,
+      "grad_norm": 0.734462320804596,
+      "learning_rate": 0.00011772919016674934,
+      "loss": 2.5847,
+      "step": 8921
+    },
+    {
+      "epoch": 0.7200387377935599,
+      "grad_norm": 0.6736955642700195,
+      "learning_rate": 0.00011771365313924886,
+      "loss": 2.558,
+      "step": 8922
+    },
+    {
+      "epoch": 0.7201194415301428,
+      "grad_norm": 0.7157856822013855,
+      "learning_rate": 0.00011769811567027784,
+      "loss": 2.6199,
+      "step": 8923
+    },
+    {
+      "epoch": 0.7202001452667258,
+      "grad_norm": 0.7045830488204956,
+      "learning_rate": 0.0001176825777602235,
+      "loss": 2.576,
+      "step": 8924
+    },
+    {
+      "epoch": 0.7202808490033088,
+      "grad_norm": 0.6875419020652771,
+      "learning_rate": 0.00011766703940947308,
+      "loss": 2.6045,
+      "step": 8925
+    },
+    {
+      "epoch": 0.7203615527398919,
+      "grad_norm": 0.7313494086265564,
+      "learning_rate": 0.00011765150061841387,
+      "loss": 2.5388,
+      "step": 8926
+    },
+    {
+      "epoch": 0.7204422564764749,
+      "grad_norm": 0.7223608493804932,
+      "learning_rate": 0.00011763596138743313,
+      "loss": 2.5466,
+      "step": 8927
+    },
+    {
+      "epoch": 0.7205229602130578,
+      "grad_norm": 0.7289614081382751,
+      "learning_rate": 0.00011762042171691816,
+      "loss": 2.5862,
+      "step": 8928
+    },
+    {
+      "epoch": 0.7206036639496408,
+      "grad_norm": 0.7098878026008606,
+      "learning_rate": 0.00011760488160725617,
+      "loss": 2.5497,
+      "step": 8929
+    },
+    {
+      "epoch": 0.7206843676862239,
+      "grad_norm": 0.7096838355064392,
+      "learning_rate": 0.00011758934105883452,
+      "loss": 2.558,
+      "step": 8930
+    },
+    {
+      "epoch": 0.7207650714228069,
+      "grad_norm": 0.7334743738174438,
+      "learning_rate": 0.00011757380007204055,
+      "loss": 2.5966,
+      "step": 8931
+    },
+    {
+      "epoch": 0.7208457751593899,
+      "grad_norm": 0.7192476391792297,
+      "learning_rate": 0.00011755825864726149,
+      "loss": 2.5307,
+      "step": 8932
+    },
+    {
+      "epoch": 0.7209264788959728,
+      "grad_norm": 0.7329632043838501,
+      "learning_rate": 0.00011754271678488478,
+      "loss": 2.6453,
+      "step": 8933
+    },
+    {
+      "epoch": 0.7210071826325559,
+      "grad_norm": 0.6827974915504456,
+      "learning_rate": 0.00011752717448529766,
+      "loss": 2.5507,
+      "step": 8934
+    },
+    {
+      "epoch": 0.7210878863691389,
+      "grad_norm": 0.8292449116706848,
+      "learning_rate": 0.00011751163174888756,
+      "loss": 2.6178,
+      "step": 8935
+    },
+    {
+      "epoch": 0.7211685901057219,
+      "grad_norm": 0.6504058837890625,
+      "learning_rate": 0.00011749608857604183,
+      "loss": 2.574,
+      "step": 8936
+    },
+    {
+      "epoch": 0.7212492938423049,
+      "grad_norm": 0.6567742824554443,
+      "learning_rate": 0.00011748054496714785,
+      "loss": 2.45,
+      "step": 8937
+    },
+    {
+      "epoch": 0.7213299975788879,
+      "grad_norm": 0.6699101328849792,
+      "learning_rate": 0.00011746500092259296,
+      "loss": 2.5827,
+      "step": 8938
+    },
+    {
+      "epoch": 0.7214107013154709,
+      "grad_norm": 0.7664934992790222,
+      "learning_rate": 0.0001174494564427646,
+      "loss": 2.5246,
+      "step": 8939
+    },
+    {
+      "epoch": 0.7214914050520539,
+      "grad_norm": 0.7276309132575989,
+      "learning_rate": 0.00011743391152805017,
+      "loss": 2.6096,
+      "step": 8940
+    },
+    {
+      "epoch": 0.7215721087886369,
+      "grad_norm": 0.7248005867004395,
+      "learning_rate": 0.0001174183661788371,
+      "loss": 2.6362,
+      "step": 8941
+    },
+    {
+      "epoch": 0.72165281252522,
+      "grad_norm": 0.7773801684379578,
+      "learning_rate": 0.00011740282039551282,
+      "loss": 2.547,
+      "step": 8942
+    },
+    {
+      "epoch": 0.7217335162618029,
+      "grad_norm": 0.7346466779708862,
+      "learning_rate": 0.00011738727417846476,
+      "loss": 2.5635,
+      "step": 8943
+    },
+    {
+      "epoch": 0.7218142199983859,
+      "grad_norm": 0.7042707800865173,
+      "learning_rate": 0.0001173717275280804,
+      "loss": 2.5593,
+      "step": 8944
+    },
+    {
+      "epoch": 0.7218949237349689,
+      "grad_norm": 0.6894899010658264,
+      "learning_rate": 0.00011735618044474712,
+      "loss": 2.5272,
+      "step": 8945
+    },
+    {
+      "epoch": 0.721975627471552,
+      "grad_norm": 0.6643744111061096,
+      "learning_rate": 0.00011734063292885249,
+      "loss": 2.6001,
+      "step": 8946
+    },
+    {
+      "epoch": 0.722056331208135,
+      "grad_norm": 0.7543076276779175,
+      "learning_rate": 0.00011732508498078396,
+      "loss": 2.558,
+      "step": 8947
+    },
+    {
+      "epoch": 0.7221370349447179,
+      "grad_norm": 0.7065596580505371,
+      "learning_rate": 0.00011730953660092903,
+      "loss": 2.6255,
+      "step": 8948
+    },
+    {
+      "epoch": 0.7222177386813009,
+      "grad_norm": 0.6968158483505249,
+      "learning_rate": 0.0001172939877896752,
+      "loss": 2.5277,
+      "step": 8949
+    },
+    {
+      "epoch": 0.722298442417884,
+      "grad_norm": 0.6918557286262512,
+      "learning_rate": 0.00011727843854740996,
+      "loss": 2.5456,
+      "step": 8950
+    },
+    {
+      "epoch": 0.722379146154467,
+      "grad_norm": 0.7262142300605774,
+      "learning_rate": 0.00011726288887452088,
+      "loss": 2.5345,
+      "step": 8951
+    },
+    {
+      "epoch": 0.7224598498910499,
+      "grad_norm": 0.7423329949378967,
+      "learning_rate": 0.00011724733877139548,
+      "loss": 2.6335,
+      "step": 8952
+    },
+    {
+      "epoch": 0.7225405536276329,
+      "grad_norm": 0.7734495997428894,
+      "learning_rate": 0.00011723178823842136,
+      "loss": 2.5951,
+      "step": 8953
+    },
+    {
+      "epoch": 0.722621257364216,
+      "grad_norm": 0.6792804598808289,
+      "learning_rate": 0.00011721623727598597,
+      "loss": 2.5927,
+      "step": 8954
+    },
+    {
+      "epoch": 0.722701961100799,
+      "grad_norm": 0.7971853017807007,
+      "learning_rate": 0.00011720068588447697,
+      "loss": 2.5451,
+      "step": 8955
+    },
+    {
+      "epoch": 0.722782664837382,
+      "grad_norm": 0.7264395356178284,
+      "learning_rate": 0.00011718513406428189,
+      "loss": 2.5769,
+      "step": 8956
+    },
+    {
+      "epoch": 0.7228633685739649,
+      "grad_norm": 0.6536725759506226,
+      "learning_rate": 0.0001171695818157884,
+      "loss": 2.6285,
+      "step": 8957
+    },
+    {
+      "epoch": 0.722944072310548,
+      "grad_norm": 0.6676235198974609,
+      "learning_rate": 0.000117154029139384,
+      "loss": 2.5896,
+      "step": 8958
+    },
+    {
+      "epoch": 0.723024776047131,
+      "grad_norm": 0.7104088664054871,
+      "learning_rate": 0.00011713847603545636,
+      "loss": 2.5606,
+      "step": 8959
+    },
+    {
+      "epoch": 0.723105479783714,
+      "grad_norm": 0.6646785140037537,
+      "learning_rate": 0.0001171229225043931,
+      "loss": 2.5617,
+      "step": 8960
+    },
+    {
+      "epoch": 0.723186183520297,
+      "grad_norm": 0.7148672342300415,
+      "learning_rate": 0.00011710736854658186,
+      "loss": 2.5855,
+      "step": 8961
+    },
+    {
+      "epoch": 0.72326688725688,
+      "grad_norm": 0.6864955425262451,
+      "learning_rate": 0.00011709181416241028,
+      "loss": 2.6098,
+      "step": 8962
+    },
+    {
+      "epoch": 0.723347590993463,
+      "grad_norm": 0.7049087285995483,
+      "learning_rate": 0.00011707625935226602,
+      "loss": 2.506,
+      "step": 8963
+    },
+    {
+      "epoch": 0.723428294730046,
+      "grad_norm": 0.6419759392738342,
+      "learning_rate": 0.00011706070411653672,
+      "loss": 2.5485,
+      "step": 8964
+    },
+    {
+      "epoch": 0.723508998466629,
+      "grad_norm": 0.6879174709320068,
+      "learning_rate": 0.00011704514845561007,
+      "loss": 2.5373,
+      "step": 8965
+    },
+    {
+      "epoch": 0.7235897022032121,
+      "grad_norm": 0.6473780274391174,
+      "learning_rate": 0.00011702959236987378,
+      "loss": 2.5479,
+      "step": 8966
+    },
+    {
+      "epoch": 0.723670405939795,
+      "grad_norm": 0.6924241185188293,
+      "learning_rate": 0.00011701403585971553,
+      "loss": 2.5679,
+      "step": 8967
+    },
+    {
+      "epoch": 0.723751109676378,
+      "grad_norm": 0.7452483773231506,
+      "learning_rate": 0.00011699847892552305,
+      "loss": 2.5043,
+      "step": 8968
+    },
+    {
+      "epoch": 0.723831813412961,
+      "grad_norm": 0.7517218589782715,
+      "learning_rate": 0.00011698292156768402,
+      "loss": 2.5554,
+      "step": 8969
+    },
+    {
+      "epoch": 0.7239125171495441,
+      "grad_norm": 0.6492432355880737,
+      "learning_rate": 0.00011696736378658618,
+      "loss": 2.6091,
+      "step": 8970
+    },
+    {
+      "epoch": 0.723993220886127,
+      "grad_norm": 0.740093469619751,
+      "learning_rate": 0.0001169518055826173,
+      "loss": 2.5629,
+      "step": 8971
+    },
+    {
+      "epoch": 0.72407392462271,
+      "grad_norm": 0.7186923027038574,
+      "learning_rate": 0.00011693624695616509,
+      "loss": 2.5537,
+      "step": 8972
+    },
+    {
+      "epoch": 0.724154628359293,
+      "grad_norm": 0.7066059112548828,
+      "learning_rate": 0.00011692068790761737,
+      "loss": 2.5115,
+      "step": 8973
+    },
+    {
+      "epoch": 0.7242353320958761,
+      "grad_norm": 0.7031805515289307,
+      "learning_rate": 0.00011690512843736185,
+      "loss": 2.596,
+      "step": 8974
+    },
+    {
+      "epoch": 0.7243160358324591,
+      "grad_norm": 0.7308956384658813,
+      "learning_rate": 0.00011688956854578635,
+      "loss": 2.6311,
+      "step": 8975
+    },
+    {
+      "epoch": 0.724396739569042,
+      "grad_norm": 0.6926052570343018,
+      "learning_rate": 0.00011687400823327863,
+      "loss": 2.5659,
+      "step": 8976
+    },
+    {
+      "epoch": 0.724477443305625,
+      "grad_norm": 0.69638991355896,
+      "learning_rate": 0.00011685844750022654,
+      "loss": 2.4792,
+      "step": 8977
+    },
+    {
+      "epoch": 0.724558147042208,
+      "grad_norm": 0.6858355402946472,
+      "learning_rate": 0.00011684288634701785,
+      "loss": 2.5707,
+      "step": 8978
+    },
+    {
+      "epoch": 0.7246388507787911,
+      "grad_norm": 0.6673639416694641,
+      "learning_rate": 0.00011682732477404044,
+      "loss": 2.5627,
+      "step": 8979
+    },
+    {
+      "epoch": 0.7247195545153741,
+      "grad_norm": 0.7174322605133057,
+      "learning_rate": 0.00011681176278168206,
+      "loss": 2.5801,
+      "step": 8980
+    },
+    {
+      "epoch": 0.724800258251957,
+      "grad_norm": 0.6840930581092834,
+      "learning_rate": 0.00011679620037033064,
+      "loss": 2.4994,
+      "step": 8981
+    },
+    {
+      "epoch": 0.72488096198854,
+      "grad_norm": 0.7179884910583496,
+      "learning_rate": 0.00011678063754037399,
+      "loss": 2.6408,
+      "step": 8982
+    },
+    {
+      "epoch": 0.7249616657251231,
+      "grad_norm": 0.6564825773239136,
+      "learning_rate": 0.00011676507429219998,
+      "loss": 2.5412,
+      "step": 8983
+    },
+    {
+      "epoch": 0.7250423694617061,
+      "grad_norm": 0.7020624876022339,
+      "learning_rate": 0.00011674951062619652,
+      "loss": 2.5778,
+      "step": 8984
+    },
+    {
+      "epoch": 0.7251230731982891,
+      "grad_norm": 0.8061255812644958,
+      "learning_rate": 0.00011673394654275145,
+      "loss": 2.5581,
+      "step": 8985
+    },
+    {
+      "epoch": 0.725203776934872,
+      "grad_norm": 0.7653982043266296,
+      "learning_rate": 0.00011671838204225267,
+      "loss": 2.5324,
+      "step": 8986
+    },
+    {
+      "epoch": 0.7252844806714551,
+      "grad_norm": 0.7168377041816711,
+      "learning_rate": 0.00011670281712508816,
+      "loss": 2.6357,
+      "step": 8987
+    },
+    {
+      "epoch": 0.7253651844080381,
+      "grad_norm": 0.6860470771789551,
+      "learning_rate": 0.00011668725179164575,
+      "loss": 2.5367,
+      "step": 8988
+    },
+    {
+      "epoch": 0.7254458881446211,
+      "grad_norm": 0.7175878286361694,
+      "learning_rate": 0.00011667168604231342,
+      "loss": 2.549,
+      "step": 8989
+    },
+    {
+      "epoch": 0.725526591881204,
+      "grad_norm": 0.7124783992767334,
+      "learning_rate": 0.00011665611987747907,
+      "loss": 2.5566,
+      "step": 8990
+    },
+    {
+      "epoch": 0.7256072956177871,
+      "grad_norm": 0.6575417518615723,
+      "learning_rate": 0.00011664055329753067,
+      "loss": 2.5455,
+      "step": 8991
+    },
+    {
+      "epoch": 0.7256879993543701,
+      "grad_norm": 0.6576877236366272,
+      "learning_rate": 0.00011662498630285623,
+      "loss": 2.5596,
+      "step": 8992
+    },
+    {
+      "epoch": 0.7257687030909531,
+      "grad_norm": 0.7235110402107239,
+      "learning_rate": 0.00011660941889384365,
+      "loss": 2.6199,
+      "step": 8993
+    },
+    {
+      "epoch": 0.7258494068275361,
+      "grad_norm": 0.6623982787132263,
+      "learning_rate": 0.00011659385107088092,
+      "loss": 2.5642,
+      "step": 8994
+    },
+    {
+      "epoch": 0.7259301105641192,
+      "grad_norm": 0.7113857865333557,
+      "learning_rate": 0.00011657828283435605,
+      "loss": 2.5631,
+      "step": 8995
+    },
+    {
+      "epoch": 0.7260108143007021,
+      "grad_norm": 0.7076124548912048,
+      "learning_rate": 0.00011656271418465702,
+      "loss": 2.5141,
+      "step": 8996
+    },
+    {
+      "epoch": 0.7260915180372851,
+      "grad_norm": 0.7534562349319458,
+      "learning_rate": 0.00011654714512217188,
+      "loss": 2.5896,
+      "step": 8997
+    },
+    {
+      "epoch": 0.7261722217738681,
+      "grad_norm": 0.7393170595169067,
+      "learning_rate": 0.00011653157564728865,
+      "loss": 2.5848,
+      "step": 8998
+    },
+    {
+      "epoch": 0.7262529255104512,
+      "grad_norm": 0.6829591989517212,
+      "learning_rate": 0.0001165160057603953,
+      "loss": 2.5439,
+      "step": 8999
+    },
+    {
+      "epoch": 0.7263336292470342,
+      "grad_norm": 0.6527189016342163,
+      "learning_rate": 0.00011650043546187995,
+      "loss": 2.5655,
+      "step": 9000
+    },
+    {
+      "epoch": 0.7263336292470342,
+      "eval_loss": 2.487652063369751,
+      "eval_runtime": 845.9129,
+      "eval_samples_per_second": 3.097,
+      "eval_steps_per_second": 0.517,
+      "step": 9000
+    },
+    {
+      "epoch": 0.7264143329836171,
+      "grad_norm": 0.6545615196228027,
+      "learning_rate": 0.00011648486475213058,
+      "loss": 2.5366,
+      "step": 9001
+    },
+    {
+      "epoch": 0.7264950367202001,
+      "grad_norm": 0.6854971647262573,
+      "learning_rate": 0.00011646929363153529,
+      "loss": 2.5832,
+      "step": 9002
+    },
+    {
+      "epoch": 0.7265757404567832,
+      "grad_norm": 0.7745552062988281,
+      "learning_rate": 0.00011645372210048218,
+      "loss": 2.5854,
+      "step": 9003
+    },
+    {
+      "epoch": 0.7266564441933662,
+      "grad_norm": 0.7159156203269958,
+      "learning_rate": 0.00011643815015935928,
+      "loss": 2.614,
+      "step": 9004
+    },
+    {
+      "epoch": 0.7267371479299491,
+      "grad_norm": 0.700074315071106,
+      "learning_rate": 0.00011642257780855475,
+      "loss": 2.6124,
+      "step": 9005
+    },
+    {
+      "epoch": 0.7268178516665321,
+      "grad_norm": 0.7367869019508362,
+      "learning_rate": 0.0001164070050484566,
+      "loss": 2.5512,
+      "step": 9006
+    },
+    {
+      "epoch": 0.7268985554031152,
+      "grad_norm": 0.6623905897140503,
+      "learning_rate": 0.00011639143187945301,
+      "loss": 2.5724,
+      "step": 9007
+    },
+    {
+      "epoch": 0.7269792591396982,
+      "grad_norm": 0.7111610770225525,
+      "learning_rate": 0.0001163758583019321,
+      "loss": 2.547,
+      "step": 9008
+    },
+    {
+      "epoch": 0.7270599628762812,
+      "grad_norm": 0.6860959529876709,
+      "learning_rate": 0.00011636028431628199,
+      "loss": 2.532,
+      "step": 9009
+    },
+    {
+      "epoch": 0.7271406666128641,
+      "grad_norm": 0.7606309056282043,
+      "learning_rate": 0.00011634470992289084,
+      "loss": 2.5214,
+      "step": 9010
+    },
+    {
+      "epoch": 0.7272213703494472,
+      "grad_norm": 0.6440508365631104,
+      "learning_rate": 0.00011632913512214677,
+      "loss": 2.5554,
+      "step": 9011
+    },
+    {
+      "epoch": 0.7273020740860302,
+      "grad_norm": 0.6770462393760681,
+      "learning_rate": 0.00011631355991443796,
+      "loss": 2.5877,
+      "step": 9012
+    },
+    {
+      "epoch": 0.7273827778226132,
+      "grad_norm": 0.6419155597686768,
+      "learning_rate": 0.00011629798430015262,
+      "loss": 2.5337,
+      "step": 9013
+    },
+    {
+      "epoch": 0.7274634815591962,
+      "grad_norm": 0.6782121658325195,
+      "learning_rate": 0.00011628240827967891,
+      "loss": 2.5152,
+      "step": 9014
+    },
+    {
+      "epoch": 0.7275441852957792,
+      "grad_norm": 0.6972285509109497,
+      "learning_rate": 0.00011626683185340501,
+      "loss": 2.5628,
+      "step": 9015
+    },
+    {
+      "epoch": 0.7276248890323622,
+      "grad_norm": 0.6823342442512512,
+      "learning_rate": 0.00011625125502171914,
+      "loss": 2.5977,
+      "step": 9016
+    },
+    {
+      "epoch": 0.7277055927689452,
+      "grad_norm": 0.723311722278595,
+      "learning_rate": 0.0001162356777850095,
+      "loss": 2.5772,
+      "step": 9017
+    },
+    {
+      "epoch": 0.7277862965055282,
+      "grad_norm": 0.7395427227020264,
+      "learning_rate": 0.00011622010014366435,
+      "loss": 2.6068,
+      "step": 9018
+    },
+    {
+      "epoch": 0.7278670002421113,
+      "grad_norm": 0.6970974206924438,
+      "learning_rate": 0.00011620452209807192,
+      "loss": 2.5577,
+      "step": 9019
+    },
+    {
+      "epoch": 0.7279477039786942,
+      "grad_norm": 0.6921418309211731,
+      "learning_rate": 0.0001161889436486204,
+      "loss": 2.5476,
+      "step": 9020
+    },
+    {
+      "epoch": 0.7280284077152772,
+      "grad_norm": 0.7243841886520386,
+      "learning_rate": 0.0001161733647956981,
+      "loss": 2.579,
+      "step": 9021
+    },
+    {
+      "epoch": 0.7281091114518602,
+      "grad_norm": 0.7240262627601624,
+      "learning_rate": 0.0001161577855396933,
+      "loss": 2.5959,
+      "step": 9022
+    },
+    {
+      "epoch": 0.7281898151884433,
+      "grad_norm": 0.7215476632118225,
+      "learning_rate": 0.0001161422058809942,
+      "loss": 2.5979,
+      "step": 9023
+    },
+    {
+      "epoch": 0.7282705189250263,
+      "grad_norm": 0.7109708786010742,
+      "learning_rate": 0.00011612662581998917,
+      "loss": 2.5912,
+      "step": 9024
+    },
+    {
+      "epoch": 0.7283512226616092,
+      "grad_norm": 0.6814073920249939,
+      "learning_rate": 0.00011611104535706645,
+      "loss": 2.5742,
+      "step": 9025
+    },
+    {
+      "epoch": 0.7284319263981922,
+      "grad_norm": 0.6788144707679749,
+      "learning_rate": 0.0001160954644926144,
+      "loss": 2.5656,
+      "step": 9026
+    },
+    {
+      "epoch": 0.7285126301347752,
+      "grad_norm": 0.7312989830970764,
+      "learning_rate": 0.00011607988322702126,
+      "loss": 2.5877,
+      "step": 9027
+    },
+    {
+      "epoch": 0.7285933338713583,
+      "grad_norm": 0.6725338697433472,
+      "learning_rate": 0.0001160643015606754,
+      "loss": 2.5261,
+      "step": 9028
+    },
+    {
+      "epoch": 0.7286740376079412,
+      "grad_norm": 0.7439326047897339,
+      "learning_rate": 0.00011604871949396516,
+      "loss": 2.603,
+      "step": 9029
+    },
+    {
+      "epoch": 0.7287547413445242,
+      "grad_norm": 0.7091783285140991,
+      "learning_rate": 0.00011603313702727889,
+      "loss": 2.5227,
+      "step": 9030
+    },
+    {
+      "epoch": 0.7288354450811072,
+      "grad_norm": 0.7474398016929626,
+      "learning_rate": 0.00011601755416100492,
+      "loss": 2.616,
+      "step": 9031
+    },
+    {
+      "epoch": 0.7289161488176903,
+      "grad_norm": 0.6904098987579346,
+      "learning_rate": 0.00011600197089553162,
+      "loss": 2.556,
+      "step": 9032
+    },
+    {
+      "epoch": 0.7289968525542733,
+      "grad_norm": 0.7305783033370972,
+      "learning_rate": 0.00011598638723124739,
+      "loss": 2.5633,
+      "step": 9033
+    },
+    {
+      "epoch": 0.7290775562908562,
+      "grad_norm": 0.6626651883125305,
+      "learning_rate": 0.00011597080316854062,
+      "loss": 2.5862,
+      "step": 9034
+    },
+    {
+      "epoch": 0.7291582600274392,
+      "grad_norm": 0.683102548122406,
+      "learning_rate": 0.00011595521870779968,
+      "loss": 2.5629,
+      "step": 9035
+    },
+    {
+      "epoch": 0.7292389637640223,
+      "grad_norm": 0.7486757636070251,
+      "learning_rate": 0.00011593963384941295,
+      "loss": 2.5831,
+      "step": 9036
+    },
+    {
+      "epoch": 0.7293196675006053,
+      "grad_norm": 0.8059591054916382,
+      "learning_rate": 0.00011592404859376888,
+      "loss": 2.6414,
+      "step": 9037
+    },
+    {
+      "epoch": 0.7294003712371883,
+      "grad_norm": 0.8371721506118774,
+      "learning_rate": 0.00011590846294125594,
+      "loss": 2.643,
+      "step": 9038
+    },
+    {
+      "epoch": 0.7294810749737712,
+      "grad_norm": 0.7216931581497192,
+      "learning_rate": 0.00011589287689226246,
+      "loss": 2.6,
+      "step": 9039
+    },
+    {
+      "epoch": 0.7295617787103543,
+      "grad_norm": 0.6940354704856873,
+      "learning_rate": 0.00011587729044717701,
+      "loss": 2.546,
+      "step": 9040
+    },
+    {
+      "epoch": 0.7296424824469373,
+      "grad_norm": 0.6888829469680786,
+      "learning_rate": 0.00011586170360638792,
+      "loss": 2.5878,
+      "step": 9041
+    },
+    {
+      "epoch": 0.7297231861835203,
+      "grad_norm": 0.6863886117935181,
+      "learning_rate": 0.00011584611637028373,
+      "loss": 2.5389,
+      "step": 9042
+    },
+    {
+      "epoch": 0.7298038899201033,
+      "grad_norm": 0.6670756936073303,
+      "learning_rate": 0.00011583052873925294,
+      "loss": 2.5465,
+      "step": 9043
+    },
+    {
+      "epoch": 0.7298845936566863,
+      "grad_norm": 0.7441220879554749,
+      "learning_rate": 0.00011581494071368392,
+      "loss": 2.5679,
+      "step": 9044
+    },
+    {
+      "epoch": 0.7299652973932693,
+      "grad_norm": 0.7135717272758484,
+      "learning_rate": 0.0001157993522939653,
+      "loss": 2.5341,
+      "step": 9045
+    },
+    {
+      "epoch": 0.7300460011298523,
+      "grad_norm": 0.6837992072105408,
+      "learning_rate": 0.00011578376348048547,
+      "loss": 2.5233,
+      "step": 9046
+    },
+    {
+      "epoch": 0.7301267048664353,
+      "grad_norm": 0.706666886806488,
+      "learning_rate": 0.00011576817427363302,
+      "loss": 2.6109,
+      "step": 9047
+    },
+    {
+      "epoch": 0.7302074086030184,
+      "grad_norm": 0.6856269240379333,
+      "learning_rate": 0.00011575258467379646,
+      "loss": 2.5651,
+      "step": 9048
+    },
+    {
+      "epoch": 0.7302881123396013,
+      "grad_norm": 0.6931480169296265,
+      "learning_rate": 0.00011573699468136427,
+      "loss": 2.6031,
+      "step": 9049
+    },
+    {
+      "epoch": 0.7303688160761843,
+      "grad_norm": 0.6558480858802795,
+      "learning_rate": 0.00011572140429672508,
+      "loss": 2.5661,
+      "step": 9050
+    },
+    {
+      "epoch": 0.7304495198127673,
+      "grad_norm": 0.6468425393104553,
+      "learning_rate": 0.00011570581352026742,
+      "loss": 2.5171,
+      "step": 9051
+    },
+    {
+      "epoch": 0.7305302235493504,
+      "grad_norm": 0.7204702496528625,
+      "learning_rate": 0.00011569022235237974,
+      "loss": 2.5861,
+      "step": 9052
+    },
+    {
+      "epoch": 0.7306109272859334,
+      "grad_norm": 0.7536416053771973,
+      "learning_rate": 0.00011567463079345078,
+      "loss": 2.633,
+      "step": 9053
+    },
+    {
+      "epoch": 0.7306916310225163,
+      "grad_norm": 0.6597960591316223,
+      "learning_rate": 0.00011565903884386904,
+      "loss": 2.5327,
+      "step": 9054
+    },
+    {
+      "epoch": 0.7307723347590993,
+      "grad_norm": 0.689153254032135,
+      "learning_rate": 0.0001156434465040231,
+      "loss": 2.5397,
+      "step": 9055
+    },
+    {
+      "epoch": 0.7308530384956824,
+      "grad_norm": 0.7664844393730164,
+      "learning_rate": 0.00011562785377430159,
+      "loss": 2.4852,
+      "step": 9056
+    },
+    {
+      "epoch": 0.7309337422322654,
+      "grad_norm": 0.7122881412506104,
+      "learning_rate": 0.0001156122606550931,
+      "loss": 2.5401,
+      "step": 9057
+    },
+    {
+      "epoch": 0.7310144459688483,
+      "grad_norm": 0.6937551498413086,
+      "learning_rate": 0.00011559666714678627,
+      "loss": 2.5705,
+      "step": 9058
+    },
+    {
+      "epoch": 0.7310951497054313,
+      "grad_norm": 0.6504047513008118,
+      "learning_rate": 0.00011558107324976974,
+      "loss": 2.5638,
+      "step": 9059
+    },
+    {
+      "epoch": 0.7311758534420144,
+      "grad_norm": 0.7759538888931274,
+      "learning_rate": 0.0001155654789644321,
+      "loss": 2.5864,
+      "step": 9060
+    },
+    {
+      "epoch": 0.7312565571785974,
+      "grad_norm": 0.719859778881073,
+      "learning_rate": 0.00011554988429116207,
+      "loss": 2.519,
+      "step": 9061
+    },
+    {
+      "epoch": 0.7313372609151804,
+      "grad_norm": 0.7159178853034973,
+      "learning_rate": 0.00011553428923034826,
+      "loss": 2.5301,
+      "step": 9062
+    },
+    {
+      "epoch": 0.7314179646517633,
+      "grad_norm": 0.6584001183509827,
+      "learning_rate": 0.00011551869378237934,
+      "loss": 2.4716,
+      "step": 9063
+    },
+    {
+      "epoch": 0.7314986683883464,
+      "grad_norm": 0.6548463702201843,
+      "learning_rate": 0.00011550309794764405,
+      "loss": 2.5637,
+      "step": 9064
+    },
+    {
+      "epoch": 0.7315793721249294,
+      "grad_norm": 0.73887699842453,
+      "learning_rate": 0.000115487501726531,
+      "loss": 2.5813,
+      "step": 9065
+    },
+    {
+      "epoch": 0.7316600758615124,
+      "grad_norm": 0.7856181859970093,
+      "learning_rate": 0.00011547190511942893,
+      "loss": 2.592,
+      "step": 9066
+    },
+    {
+      "epoch": 0.7317407795980954,
+      "grad_norm": 0.7040740847587585,
+      "learning_rate": 0.00011545630812672654,
+      "loss": 2.5324,
+      "step": 9067
+    },
+    {
+      "epoch": 0.7318214833346784,
+      "grad_norm": 0.7316064238548279,
+      "learning_rate": 0.00011544071074881253,
+      "loss": 2.5487,
+      "step": 9068
+    },
+    {
+      "epoch": 0.7319021870712614,
+      "grad_norm": 0.7020413279533386,
+      "learning_rate": 0.00011542511298607568,
+      "loss": 2.5179,
+      "step": 9069
+    },
+    {
+      "epoch": 0.7319828908078444,
+      "grad_norm": 0.672605574131012,
+      "learning_rate": 0.00011540951483890468,
+      "loss": 2.5367,
+      "step": 9070
+    },
+    {
+      "epoch": 0.7320635945444274,
+      "grad_norm": 0.7668856382369995,
+      "learning_rate": 0.00011539391630768828,
+      "loss": 2.6089,
+      "step": 9071
+    },
+    {
+      "epoch": 0.7321442982810105,
+      "grad_norm": 0.6641809940338135,
+      "learning_rate": 0.00011537831739281524,
+      "loss": 2.5411,
+      "step": 9072
+    },
+    {
+      "epoch": 0.7322250020175934,
+      "grad_norm": 0.7142000198364258,
+      "learning_rate": 0.00011536271809467434,
+      "loss": 2.5469,
+      "step": 9073
+    },
+    {
+      "epoch": 0.7323057057541764,
+      "grad_norm": 0.7266140580177307,
+      "learning_rate": 0.00011534711841365435,
+      "loss": 2.5565,
+      "step": 9074
+    },
+    {
+      "epoch": 0.7323864094907594,
+      "grad_norm": 0.6763899326324463,
+      "learning_rate": 0.00011533151835014407,
+      "loss": 2.551,
+      "step": 9075
+    },
+    {
+      "epoch": 0.7324671132273425,
+      "grad_norm": 0.6517418026924133,
+      "learning_rate": 0.00011531591790453224,
+      "loss": 2.5415,
+      "step": 9076
+    },
+    {
+      "epoch": 0.7325478169639255,
+      "grad_norm": 0.6602214574813843,
+      "learning_rate": 0.00011530031707720772,
+      "loss": 2.593,
+      "step": 9077
+    },
+    {
+      "epoch": 0.7326285207005084,
+      "grad_norm": 0.7448844313621521,
+      "learning_rate": 0.00011528471586855931,
+      "loss": 2.5598,
+      "step": 9078
+    },
+    {
+      "epoch": 0.7327092244370914,
+      "grad_norm": 0.7197073698043823,
+      "learning_rate": 0.00011526911427897579,
+      "loss": 2.5128,
+      "step": 9079
+    },
+    {
+      "epoch": 0.7327899281736744,
+      "grad_norm": 0.7245968580245972,
+      "learning_rate": 0.00011525351230884606,
+      "loss": 2.5016,
+      "step": 9080
+    },
+    {
+      "epoch": 0.7328706319102575,
+      "grad_norm": 0.6715837717056274,
+      "learning_rate": 0.00011523790995855892,
+      "loss": 2.5469,
+      "step": 9081
+    },
+    {
+      "epoch": 0.7329513356468405,
+      "grad_norm": 0.7143638730049133,
+      "learning_rate": 0.00011522230722850325,
+      "loss": 2.5164,
+      "step": 9082
+    },
+    {
+      "epoch": 0.7330320393834234,
+      "grad_norm": 0.6809647083282471,
+      "learning_rate": 0.00011520670411906787,
+      "loss": 2.6071,
+      "step": 9083
+    },
+    {
+      "epoch": 0.7331127431200064,
+      "grad_norm": 0.7160956859588623,
+      "learning_rate": 0.00011519110063064167,
+      "loss": 2.5346,
+      "step": 9084
+    },
+    {
+      "epoch": 0.7331934468565895,
+      "grad_norm": 0.6814724802970886,
+      "learning_rate": 0.00011517549676361357,
+      "loss": 2.5499,
+      "step": 9085
+    },
+    {
+      "epoch": 0.7332741505931725,
+      "grad_norm": 0.6914821267127991,
+      "learning_rate": 0.00011515989251837239,
+      "loss": 2.5386,
+      "step": 9086
+    },
+    {
+      "epoch": 0.7333548543297554,
+      "grad_norm": 0.7292554378509521,
+      "learning_rate": 0.00011514428789530705,
+      "loss": 2.5642,
+      "step": 9087
+    },
+    {
+      "epoch": 0.7334355580663384,
+      "grad_norm": 0.6894826292991638,
+      "learning_rate": 0.00011512868289480647,
+      "loss": 2.6131,
+      "step": 9088
+    },
+    {
+      "epoch": 0.7335162618029215,
+      "grad_norm": 0.658770740032196,
+      "learning_rate": 0.00011511307751725957,
+      "loss": 2.5594,
+      "step": 9089
+    },
+    {
+      "epoch": 0.7335969655395045,
+      "grad_norm": 0.7508681416511536,
+      "learning_rate": 0.0001150974717630553,
+      "loss": 2.595,
+      "step": 9090
+    },
+    {
+      "epoch": 0.7336776692760875,
+      "grad_norm": 0.69661545753479,
+      "learning_rate": 0.00011508186563258256,
+      "loss": 2.5803,
+      "step": 9091
+    },
+    {
+      "epoch": 0.7337583730126704,
+      "grad_norm": 0.7277412414550781,
+      "learning_rate": 0.00011506625912623028,
+      "loss": 2.5456,
+      "step": 9092
+    },
+    {
+      "epoch": 0.7338390767492535,
+      "grad_norm": 0.658329963684082,
+      "learning_rate": 0.00011505065224438745,
+      "loss": 2.5177,
+      "step": 9093
+    },
+    {
+      "epoch": 0.7339197804858365,
+      "grad_norm": 0.7277211546897888,
+      "learning_rate": 0.00011503504498744302,
+      "loss": 2.553,
+      "step": 9094
+    },
+    {
+      "epoch": 0.7340004842224195,
+      "grad_norm": 0.7240201830863953,
+      "learning_rate": 0.00011501943735578598,
+      "loss": 2.5851,
+      "step": 9095
+    },
+    {
+      "epoch": 0.7340811879590025,
+      "grad_norm": 0.6565662026405334,
+      "learning_rate": 0.00011500382934980529,
+      "loss": 2.5865,
+      "step": 9096
+    },
+    {
+      "epoch": 0.7341618916955855,
+      "grad_norm": 0.658268392086029,
+      "learning_rate": 0.00011498822096988995,
+      "loss": 2.5402,
+      "step": 9097
+    },
+    {
+      "epoch": 0.7342425954321685,
+      "grad_norm": 0.7305087447166443,
+      "learning_rate": 0.00011497261221642894,
+      "loss": 2.5483,
+      "step": 9098
+    },
+    {
+      "epoch": 0.7343232991687515,
+      "grad_norm": 0.7271504402160645,
+      "learning_rate": 0.00011495700308981134,
+      "loss": 2.5303,
+      "step": 9099
+    },
+    {
+      "epoch": 0.7344040029053345,
+      "grad_norm": 0.70429527759552,
+      "learning_rate": 0.0001149413935904261,
+      "loss": 2.5878,
+      "step": 9100
+    },
+    {
+      "epoch": 0.7344847066419176,
+      "grad_norm": 0.7168769836425781,
+      "learning_rate": 0.00011492578371866229,
+      "loss": 2.6017,
+      "step": 9101
+    },
+    {
+      "epoch": 0.7345654103785005,
+      "grad_norm": 0.7131996154785156,
+      "learning_rate": 0.00011491017347490891,
+      "loss": 2.5439,
+      "step": 9102
+    },
+    {
+      "epoch": 0.7346461141150835,
+      "grad_norm": 0.660321056842804,
+      "learning_rate": 0.00011489456285955504,
+      "loss": 2.5236,
+      "step": 9103
+    },
+    {
+      "epoch": 0.7347268178516665,
+      "grad_norm": 0.6742995977401733,
+      "learning_rate": 0.00011487895187298977,
+      "loss": 2.5375,
+      "step": 9104
+    },
+    {
+      "epoch": 0.7348075215882496,
+      "grad_norm": 0.6380610466003418,
+      "learning_rate": 0.00011486334051560206,
+      "loss": 2.5173,
+      "step": 9105
+    },
+    {
+      "epoch": 0.7348882253248326,
+      "grad_norm": 0.6948198080062866,
+      "learning_rate": 0.0001148477287877811,
+      "loss": 2.5247,
+      "step": 9106
+    },
+    {
+      "epoch": 0.7349689290614155,
+      "grad_norm": 0.7088696360588074,
+      "learning_rate": 0.00011483211668991591,
+      "loss": 2.587,
+      "step": 9107
+    },
+    {
+      "epoch": 0.7350496327979985,
+      "grad_norm": 0.6278921961784363,
+      "learning_rate": 0.00011481650422239556,
+      "loss": 2.5652,
+      "step": 9108
+    },
+    {
+      "epoch": 0.7351303365345816,
+      "grad_norm": 0.6901956796646118,
+      "learning_rate": 0.00011480089138560926,
+      "loss": 2.5964,
+      "step": 9109
+    },
+    {
+      "epoch": 0.7352110402711646,
+      "grad_norm": 0.7264819145202637,
+      "learning_rate": 0.00011478527817994604,
+      "loss": 2.5437,
+      "step": 9110
+    },
+    {
+      "epoch": 0.7352917440077475,
+      "grad_norm": 0.6940708756446838,
+      "learning_rate": 0.00011476966460579501,
+      "loss": 2.5761,
+      "step": 9111
+    },
+    {
+      "epoch": 0.7353724477443305,
+      "grad_norm": 0.689588189125061,
+      "learning_rate": 0.00011475405066354536,
+      "loss": 2.5457,
+      "step": 9112
+    },
+    {
+      "epoch": 0.7354531514809136,
+      "grad_norm": 0.6938436031341553,
+      "learning_rate": 0.00011473843635358618,
+      "loss": 2.6026,
+      "step": 9113
+    },
+    {
+      "epoch": 0.7355338552174966,
+      "grad_norm": 0.7122177481651306,
+      "learning_rate": 0.00011472282167630663,
+      "loss": 2.5701,
+      "step": 9114
+    },
+    {
+      "epoch": 0.7356145589540796,
+      "grad_norm": 0.6667213439941406,
+      "learning_rate": 0.00011470720663209591,
+      "loss": 2.5944,
+      "step": 9115
+    },
+    {
+      "epoch": 0.7356952626906625,
+      "grad_norm": 0.705910861492157,
+      "learning_rate": 0.00011469159122134314,
+      "loss": 2.6183,
+      "step": 9116
+    },
+    {
+      "epoch": 0.7357759664272456,
+      "grad_norm": 0.709937572479248,
+      "learning_rate": 0.00011467597544443751,
+      "loss": 2.5153,
+      "step": 9117
+    },
+    {
+      "epoch": 0.7358566701638286,
+      "grad_norm": 0.6870958805084229,
+      "learning_rate": 0.00011466035930176822,
+      "loss": 2.5334,
+      "step": 9118
+    },
+    {
+      "epoch": 0.7359373739004116,
+      "grad_norm": 0.7274392247200012,
+      "learning_rate": 0.00011464474279372443,
+      "loss": 2.5336,
+      "step": 9119
+    },
+    {
+      "epoch": 0.7360180776369946,
+      "grad_norm": 0.6360952258110046,
+      "learning_rate": 0.0001146291259206954,
+      "loss": 2.5604,
+      "step": 9120
+    },
+    {
+      "epoch": 0.7360987813735776,
+      "grad_norm": 0.7990559935569763,
+      "learning_rate": 0.00011461350868307028,
+      "loss": 2.624,
+      "step": 9121
+    },
+    {
+      "epoch": 0.7361794851101606,
+      "grad_norm": 0.6670079827308655,
+      "learning_rate": 0.00011459789108123835,
+      "loss": 2.5761,
+      "step": 9122
+    },
+    {
+      "epoch": 0.7362601888467436,
+      "grad_norm": 0.6994437575340271,
+      "learning_rate": 0.00011458227311558877,
+      "loss": 2.5679,
+      "step": 9123
+    },
+    {
+      "epoch": 0.7363408925833266,
+      "grad_norm": 0.7428358197212219,
+      "learning_rate": 0.00011456665478651087,
+      "loss": 2.5874,
+      "step": 9124
+    },
+    {
+      "epoch": 0.7364215963199097,
+      "grad_norm": 0.7079486846923828,
+      "learning_rate": 0.00011455103609439387,
+      "loss": 2.5999,
+      "step": 9125
+    },
+    {
+      "epoch": 0.7365023000564926,
+      "grad_norm": 0.646244466304779,
+      "learning_rate": 0.00011453541703962695,
+      "loss": 2.5053,
+      "step": 9126
+    },
+    {
+      "epoch": 0.7365830037930756,
+      "grad_norm": 0.6671318411827087,
+      "learning_rate": 0.0001145197976225995,
+      "loss": 2.5277,
+      "step": 9127
+    },
+    {
+      "epoch": 0.7366637075296586,
+      "grad_norm": 0.7060399055480957,
+      "learning_rate": 0.00011450417784370072,
+      "loss": 2.6092,
+      "step": 9128
+    },
+    {
+      "epoch": 0.7367444112662416,
+      "grad_norm": 0.741547703742981,
+      "learning_rate": 0.00011448855770331989,
+      "loss": 2.6121,
+      "step": 9129
+    },
+    {
+      "epoch": 0.7368251150028247,
+      "grad_norm": 0.710267961025238,
+      "learning_rate": 0.00011447293720184636,
+      "loss": 2.5141,
+      "step": 9130
+    },
+    {
+      "epoch": 0.7369058187394076,
+      "grad_norm": 0.6914308071136475,
+      "learning_rate": 0.0001144573163396694,
+      "loss": 2.5489,
+      "step": 9131
+    },
+    {
+      "epoch": 0.7369865224759906,
+      "grad_norm": 0.7051414847373962,
+      "learning_rate": 0.0001144416951171783,
+      "loss": 2.5925,
+      "step": 9132
+    },
+    {
+      "epoch": 0.7370672262125736,
+      "grad_norm": 0.6765387058258057,
+      "learning_rate": 0.00011442607353476245,
+      "loss": 2.5864,
+      "step": 9133
+    },
+    {
+      "epoch": 0.7371479299491567,
+      "grad_norm": 0.706672191619873,
+      "learning_rate": 0.00011441045159281108,
+      "loss": 2.4823,
+      "step": 9134
+    },
+    {
+      "epoch": 0.7372286336857397,
+      "grad_norm": 0.7534066438674927,
+      "learning_rate": 0.00011439482929171362,
+      "loss": 2.5728,
+      "step": 9135
+    },
+    {
+      "epoch": 0.7373093374223226,
+      "grad_norm": 0.6628777384757996,
+      "learning_rate": 0.00011437920663185939,
+      "loss": 2.5538,
+      "step": 9136
+    },
+    {
+      "epoch": 0.7373900411589056,
+      "grad_norm": 0.6575733423233032,
+      "learning_rate": 0.00011436358361363773,
+      "loss": 2.4802,
+      "step": 9137
+    },
+    {
+      "epoch": 0.7374707448954887,
+      "grad_norm": 0.7629329562187195,
+      "learning_rate": 0.00011434796023743803,
+      "loss": 2.6169,
+      "step": 9138
+    },
+    {
+      "epoch": 0.7375514486320717,
+      "grad_norm": 0.7148225903511047,
+      "learning_rate": 0.00011433233650364965,
+      "loss": 2.6335,
+      "step": 9139
+    },
+    {
+      "epoch": 0.7376321523686546,
+      "grad_norm": 0.705210268497467,
+      "learning_rate": 0.00011431671241266198,
+      "loss": 2.6261,
+      "step": 9140
+    },
+    {
+      "epoch": 0.7377128561052376,
+      "grad_norm": 0.7137441635131836,
+      "learning_rate": 0.00011430108796486441,
+      "loss": 2.5021,
+      "step": 9141
+    },
+    {
+      "epoch": 0.7377935598418207,
+      "grad_norm": 0.6979854702949524,
+      "learning_rate": 0.00011428546316064635,
+      "loss": 2.5436,
+      "step": 9142
+    },
+    {
+      "epoch": 0.7378742635784037,
+      "grad_norm": 0.6568784713745117,
+      "learning_rate": 0.00011426983800039721,
+      "loss": 2.5882,
+      "step": 9143
+    },
+    {
+      "epoch": 0.7379549673149867,
+      "grad_norm": 0.666606605052948,
+      "learning_rate": 0.00011425421248450638,
+      "loss": 2.5472,
+      "step": 9144
+    },
+    {
+      "epoch": 0.7380356710515696,
+      "grad_norm": 0.7240840792655945,
+      "learning_rate": 0.00011423858661336333,
+      "loss": 2.6057,
+      "step": 9145
+    },
+    {
+      "epoch": 0.7381163747881527,
+      "grad_norm": 0.7342149615287781,
+      "learning_rate": 0.0001142229603873575,
+      "loss": 2.508,
+      "step": 9146
+    },
+    {
+      "epoch": 0.7381970785247357,
+      "grad_norm": 0.7089941501617432,
+      "learning_rate": 0.0001142073338068783,
+      "loss": 2.6115,
+      "step": 9147
+    },
+    {
+      "epoch": 0.7382777822613187,
+      "grad_norm": 0.6883555054664612,
+      "learning_rate": 0.00011419170687231519,
+      "loss": 2.5254,
+      "step": 9148
+    },
+    {
+      "epoch": 0.7383584859979017,
+      "grad_norm": 0.6819528937339783,
+      "learning_rate": 0.00011417607958405765,
+      "loss": 2.5498,
+      "step": 9149
+    },
+    {
+      "epoch": 0.7384391897344847,
+      "grad_norm": 0.7348979711532593,
+      "learning_rate": 0.00011416045194249516,
+      "loss": 2.5547,
+      "step": 9150
+    },
+    {
+      "epoch": 0.7385198934710677,
+      "grad_norm": 0.6733320355415344,
+      "learning_rate": 0.00011414482394801719,
+      "loss": 2.5985,
+      "step": 9151
+    },
+    {
+      "epoch": 0.7386005972076507,
+      "grad_norm": 0.714771032333374,
+      "learning_rate": 0.00011412919560101327,
+      "loss": 2.571,
+      "step": 9152
+    },
+    {
+      "epoch": 0.7386813009442337,
+      "grad_norm": 0.7010024189949036,
+      "learning_rate": 0.0001141135669018728,
+      "loss": 2.5755,
+      "step": 9153
+    },
+    {
+      "epoch": 0.7387620046808168,
+      "grad_norm": 0.7014826536178589,
+      "learning_rate": 0.00011409793785098536,
+      "loss": 2.6033,
+      "step": 9154
+    },
+    {
+      "epoch": 0.7388427084173997,
+      "grad_norm": 0.7286051511764526,
+      "learning_rate": 0.0001140823084487405,
+      "loss": 2.515,
+      "step": 9155
+    },
+    {
+      "epoch": 0.7389234121539827,
+      "grad_norm": 0.669365406036377,
+      "learning_rate": 0.00011406667869552768,
+      "loss": 2.506,
+      "step": 9156
+    },
+    {
+      "epoch": 0.7390041158905657,
+      "grad_norm": 0.6886852979660034,
+      "learning_rate": 0.00011405104859173645,
+      "loss": 2.6123,
+      "step": 9157
+    },
+    {
+      "epoch": 0.7390848196271488,
+      "grad_norm": 0.6344162225723267,
+      "learning_rate": 0.00011403541813775635,
+      "loss": 2.5483,
+      "step": 9158
+    },
+    {
+      "epoch": 0.7391655233637318,
+      "grad_norm": 0.7043579816818237,
+      "learning_rate": 0.00011401978733397694,
+      "loss": 2.5545,
+      "step": 9159
+    },
+    {
+      "epoch": 0.7392462271003147,
+      "grad_norm": 0.7960262298583984,
+      "learning_rate": 0.00011400415618078781,
+      "loss": 2.5666,
+      "step": 9160
+    },
+    {
+      "epoch": 0.7393269308368977,
+      "grad_norm": 0.6771546006202698,
+      "learning_rate": 0.00011398852467857848,
+      "loss": 2.6016,
+      "step": 9161
+    },
+    {
+      "epoch": 0.7394076345734808,
+      "grad_norm": 0.6522069573402405,
+      "learning_rate": 0.00011397289282773855,
+      "loss": 2.5493,
+      "step": 9162
+    },
+    {
+      "epoch": 0.7394883383100638,
+      "grad_norm": 0.6804657578468323,
+      "learning_rate": 0.00011395726062865762,
+      "loss": 2.5856,
+      "step": 9163
+    },
+    {
+      "epoch": 0.7395690420466468,
+      "grad_norm": 0.7562841176986694,
+      "learning_rate": 0.00011394162808172526,
+      "loss": 2.557,
+      "step": 9164
+    },
+    {
+      "epoch": 0.7396497457832297,
+      "grad_norm": 0.6464113593101501,
+      "learning_rate": 0.00011392599518733107,
+      "loss": 2.5292,
+      "step": 9165
+    },
+    {
+      "epoch": 0.7397304495198128,
+      "grad_norm": 0.7469549775123596,
+      "learning_rate": 0.00011391036194586466,
+      "loss": 2.6168,
+      "step": 9166
+    },
+    {
+      "epoch": 0.7398111532563958,
+      "grad_norm": 0.7095946669578552,
+      "learning_rate": 0.00011389472835771572,
+      "loss": 2.5468,
+      "step": 9167
+    },
+    {
+      "epoch": 0.7398918569929788,
+      "grad_norm": 0.7376375794410706,
+      "learning_rate": 0.00011387909442327382,
+      "loss": 2.5576,
+      "step": 9168
+    },
+    {
+      "epoch": 0.7399725607295617,
+      "grad_norm": 0.736727774143219,
+      "learning_rate": 0.00011386346014292859,
+      "loss": 2.6034,
+      "step": 9169
+    },
+    {
+      "epoch": 0.7400532644661448,
+      "grad_norm": 0.7026904821395874,
+      "learning_rate": 0.00011384782551706967,
+      "loss": 2.5848,
+      "step": 9170
+    },
+    {
+      "epoch": 0.7401339682027278,
+      "grad_norm": 0.6894888877868652,
+      "learning_rate": 0.00011383219054608678,
+      "loss": 2.5475,
+      "step": 9171
+    },
+    {
+      "epoch": 0.7402146719393108,
+      "grad_norm": 0.6754137277603149,
+      "learning_rate": 0.00011381655523036954,
+      "loss": 2.5124,
+      "step": 9172
+    },
+    {
+      "epoch": 0.7402953756758938,
+      "grad_norm": 0.7935643196105957,
+      "learning_rate": 0.00011380091957030762,
+      "loss": 2.5898,
+      "step": 9173
+    },
+    {
+      "epoch": 0.7403760794124769,
+      "grad_norm": 0.7017118334770203,
+      "learning_rate": 0.0001137852835662907,
+      "loss": 2.6139,
+      "step": 9174
+    },
+    {
+      "epoch": 0.7404567831490598,
+      "grad_norm": 0.7246189117431641,
+      "learning_rate": 0.00011376964721870847,
+      "loss": 2.4627,
+      "step": 9175
+    },
+    {
+      "epoch": 0.7405374868856428,
+      "grad_norm": 0.6835598349571228,
+      "learning_rate": 0.00011375401052795064,
+      "loss": 2.5707,
+      "step": 9176
+    },
+    {
+      "epoch": 0.7406181906222258,
+      "grad_norm": 0.6439787745475769,
+      "learning_rate": 0.00011373837349440693,
+      "loss": 2.5161,
+      "step": 9177
+    },
+    {
+      "epoch": 0.7406988943588089,
+      "grad_norm": 0.7249091267585754,
+      "learning_rate": 0.00011372273611846704,
+      "loss": 2.5054,
+      "step": 9178
+    },
+    {
+      "epoch": 0.7407795980953918,
+      "grad_norm": 0.7653267979621887,
+      "learning_rate": 0.0001137070984005207,
+      "loss": 2.6016,
+      "step": 9179
+    },
+    {
+      "epoch": 0.7408603018319748,
+      "grad_norm": 0.7195165157318115,
+      "learning_rate": 0.0001136914603409576,
+      "loss": 2.5931,
+      "step": 9180
+    },
+    {
+      "epoch": 0.7409410055685578,
+      "grad_norm": 0.7093746662139893,
+      "learning_rate": 0.00011367582194016756,
+      "loss": 2.5567,
+      "step": 9181
+    },
+    {
+      "epoch": 0.7410217093051408,
+      "grad_norm": 0.6868107318878174,
+      "learning_rate": 0.00011366018319854026,
+      "loss": 2.5769,
+      "step": 9182
+    },
+    {
+      "epoch": 0.7411024130417239,
+      "grad_norm": 0.6870261430740356,
+      "learning_rate": 0.00011364454411646552,
+      "loss": 2.5418,
+      "step": 9183
+    },
+    {
+      "epoch": 0.7411831167783068,
+      "grad_norm": 0.7034662365913391,
+      "learning_rate": 0.00011362890469433306,
+      "loss": 2.5798,
+      "step": 9184
+    },
+    {
+      "epoch": 0.7412638205148898,
+      "grad_norm": 0.7200794816017151,
+      "learning_rate": 0.00011361326493253264,
+      "loss": 2.5523,
+      "step": 9185
+    },
+    {
+      "epoch": 0.7413445242514728,
+      "grad_norm": 0.7034540772438049,
+      "learning_rate": 0.0001135976248314541,
+      "loss": 2.5107,
+      "step": 9186
+    },
+    {
+      "epoch": 0.7414252279880559,
+      "grad_norm": 0.7155053019523621,
+      "learning_rate": 0.00011358198439148721,
+      "loss": 2.5804,
+      "step": 9187
+    },
+    {
+      "epoch": 0.7415059317246389,
+      "grad_norm": 0.6965398788452148,
+      "learning_rate": 0.00011356634361302175,
+      "loss": 2.5532,
+      "step": 9188
+    },
+    {
+      "epoch": 0.7415866354612218,
+      "grad_norm": 0.65416419506073,
+      "learning_rate": 0.00011355070249644755,
+      "loss": 2.5411,
+      "step": 9189
+    },
+    {
+      "epoch": 0.7416673391978048,
+      "grad_norm": 0.6798486709594727,
+      "learning_rate": 0.0001135350610421544,
+      "loss": 2.4957,
+      "step": 9190
+    },
+    {
+      "epoch": 0.7417480429343879,
+      "grad_norm": 0.6839874386787415,
+      "learning_rate": 0.00011351941925053218,
+      "loss": 2.5745,
+      "step": 9191
+    },
+    {
+      "epoch": 0.7418287466709709,
+      "grad_norm": 0.7374398708343506,
+      "learning_rate": 0.00011350377712197068,
+      "loss": 2.4923,
+      "step": 9192
+    },
+    {
+      "epoch": 0.7419094504075538,
+      "grad_norm": 0.7517396807670593,
+      "learning_rate": 0.00011348813465685974,
+      "loss": 2.538,
+      "step": 9193
+    },
+    {
+      "epoch": 0.7419901541441368,
+      "grad_norm": 0.6670863628387451,
+      "learning_rate": 0.00011347249185558926,
+      "loss": 2.5442,
+      "step": 9194
+    },
+    {
+      "epoch": 0.7420708578807199,
+      "grad_norm": 0.6508080363273621,
+      "learning_rate": 0.00011345684871854905,
+      "loss": 2.6665,
+      "step": 9195
+    },
+    {
+      "epoch": 0.7421515616173029,
+      "grad_norm": 0.6935258507728577,
+      "learning_rate": 0.00011344120524612898,
+      "loss": 2.5388,
+      "step": 9196
+    },
+    {
+      "epoch": 0.7422322653538859,
+      "grad_norm": 0.696067750453949,
+      "learning_rate": 0.00011342556143871897,
+      "loss": 2.574,
+      "step": 9197
+    },
+    {
+      "epoch": 0.7423129690904688,
+      "grad_norm": 0.7486966252326965,
+      "learning_rate": 0.00011340991729670882,
+      "loss": 2.5924,
+      "step": 9198
+    },
+    {
+      "epoch": 0.7423936728270519,
+      "grad_norm": 0.676407516002655,
+      "learning_rate": 0.00011339427282048854,
+      "loss": 2.5907,
+      "step": 9199
+    },
+    {
+      "epoch": 0.7424743765636349,
+      "grad_norm": 0.7241318225860596,
+      "learning_rate": 0.00011337862801044792,
+      "loss": 2.5685,
+      "step": 9200
+    },
+    {
+      "epoch": 0.7425550803002179,
+      "grad_norm": 0.7012883424758911,
+      "learning_rate": 0.00011336298286697692,
+      "loss": 2.56,
+      "step": 9201
+    },
+    {
+      "epoch": 0.7426357840368009,
+      "grad_norm": 0.7313060164451599,
+      "learning_rate": 0.0001133473373904655,
+      "loss": 2.632,
+      "step": 9202
+    },
+    {
+      "epoch": 0.742716487773384,
+      "grad_norm": 0.6829206943511963,
+      "learning_rate": 0.00011333169158130353,
+      "loss": 2.5006,
+      "step": 9203
+    },
+    {
+      "epoch": 0.7427971915099669,
+      "grad_norm": 0.7324578166007996,
+      "learning_rate": 0.00011331604543988093,
+      "loss": 2.5004,
+      "step": 9204
+    },
+    {
+      "epoch": 0.7428778952465499,
+      "grad_norm": 0.6761097311973572,
+      "learning_rate": 0.00011330039896658766,
+      "loss": 2.5516,
+      "step": 9205
+    },
+    {
+      "epoch": 0.7429585989831329,
+      "grad_norm": 0.6909754276275635,
+      "learning_rate": 0.00011328475216181369,
+      "loss": 2.5273,
+      "step": 9206
+    },
+    {
+      "epoch": 0.743039302719716,
+      "grad_norm": 0.6420674324035645,
+      "learning_rate": 0.00011326910502594899,
+      "loss": 2.5507,
+      "step": 9207
+    },
+    {
+      "epoch": 0.7431200064562989,
+      "grad_norm": 0.6442455053329468,
+      "learning_rate": 0.0001132534575593835,
+      "loss": 2.542,
+      "step": 9208
+    },
+    {
+      "epoch": 0.7432007101928819,
+      "grad_norm": 0.7053101658821106,
+      "learning_rate": 0.0001132378097625072,
+      "loss": 2.5116,
+      "step": 9209
+    },
+    {
+      "epoch": 0.7432814139294649,
+      "grad_norm": 0.7570765614509583,
+      "learning_rate": 0.00011322216163571007,
+      "loss": 2.5576,
+      "step": 9210
+    },
+    {
+      "epoch": 0.743362117666048,
+      "grad_norm": 0.6937675476074219,
+      "learning_rate": 0.00011320651317938214,
+      "loss": 2.6212,
+      "step": 9211
+    },
+    {
+      "epoch": 0.743442821402631,
+      "grad_norm": 0.6741313934326172,
+      "learning_rate": 0.00011319086439391333,
+      "loss": 2.5723,
+      "step": 9212
+    },
+    {
+      "epoch": 0.7435235251392139,
+      "grad_norm": 0.711358904838562,
+      "learning_rate": 0.00011317521527969374,
+      "loss": 2.5713,
+      "step": 9213
+    },
+    {
+      "epoch": 0.7436042288757969,
+      "grad_norm": 0.7443268895149231,
+      "learning_rate": 0.00011315956583711331,
+      "loss": 2.5301,
+      "step": 9214
+    },
+    {
+      "epoch": 0.74368493261238,
+      "grad_norm": 0.7001742720603943,
+      "learning_rate": 0.00011314391606656212,
+      "loss": 2.5545,
+      "step": 9215
+    },
+    {
+      "epoch": 0.743765636348963,
+      "grad_norm": 0.7294990420341492,
+      "learning_rate": 0.00011312826596843019,
+      "loss": 2.5897,
+      "step": 9216
+    },
+    {
+      "epoch": 0.743846340085546,
+      "grad_norm": 0.706924319267273,
+      "learning_rate": 0.00011311261554310753,
+      "loss": 2.6477,
+      "step": 9217
+    },
+    {
+      "epoch": 0.7439270438221289,
+      "grad_norm": 0.7065039277076721,
+      "learning_rate": 0.00011309696479098423,
+      "loss": 2.5326,
+      "step": 9218
+    },
+    {
+      "epoch": 0.744007747558712,
+      "grad_norm": 0.6502599716186523,
+      "learning_rate": 0.00011308131371245037,
+      "loss": 2.5833,
+      "step": 9219
+    },
+    {
+      "epoch": 0.744088451295295,
+      "grad_norm": 0.7135158181190491,
+      "learning_rate": 0.00011306566230789592,
+      "loss": 2.5686,
+      "step": 9220
+    },
+    {
+      "epoch": 0.744169155031878,
+      "grad_norm": 0.7239195108413696,
+      "learning_rate": 0.00011305001057771101,
+      "loss": 2.6303,
+      "step": 9221
+    },
+    {
+      "epoch": 0.744249858768461,
+      "grad_norm": 0.6442604660987854,
+      "learning_rate": 0.00011303435852228574,
+      "loss": 2.5495,
+      "step": 9222
+    },
+    {
+      "epoch": 0.744330562505044,
+      "grad_norm": 0.6700316071510315,
+      "learning_rate": 0.0001130187061420102,
+      "loss": 2.5575,
+      "step": 9223
+    },
+    {
+      "epoch": 0.744411266241627,
+      "grad_norm": 0.7532816529273987,
+      "learning_rate": 0.00011300305343727446,
+      "loss": 2.5174,
+      "step": 9224
+    },
+    {
+      "epoch": 0.74449196997821,
+      "grad_norm": 0.7614738941192627,
+      "learning_rate": 0.00011298740040846862,
+      "loss": 2.5995,
+      "step": 9225
+    },
+    {
+      "epoch": 0.744572673714793,
+      "grad_norm": 0.6781208515167236,
+      "learning_rate": 0.00011297174705598283,
+      "loss": 2.5225,
+      "step": 9226
+    },
+    {
+      "epoch": 0.744653377451376,
+      "grad_norm": 0.680525541305542,
+      "learning_rate": 0.0001129560933802072,
+      "loss": 2.5844,
+      "step": 9227
+    },
+    {
+      "epoch": 0.744734081187959,
+      "grad_norm": 0.7196657657623291,
+      "learning_rate": 0.00011294043938153185,
+      "loss": 2.564,
+      "step": 9228
+    },
+    {
+      "epoch": 0.744814784924542,
+      "grad_norm": 0.6997412443161011,
+      "learning_rate": 0.00011292478506034694,
+      "loss": 2.6486,
+      "step": 9229
+    },
+    {
+      "epoch": 0.744895488661125,
+      "grad_norm": 0.7438939809799194,
+      "learning_rate": 0.00011290913041704256,
+      "loss": 2.5667,
+      "step": 9230
+    },
+    {
+      "epoch": 0.744976192397708,
+      "grad_norm": 0.7391374707221985,
+      "learning_rate": 0.00011289347545200892,
+      "loss": 2.5974,
+      "step": 9231
+    },
+    {
+      "epoch": 0.745056896134291,
+      "grad_norm": 0.7845481634140015,
+      "learning_rate": 0.0001128778201656362,
+      "loss": 2.5168,
+      "step": 9232
+    },
+    {
+      "epoch": 0.745137599870874,
+      "grad_norm": 0.728712797164917,
+      "learning_rate": 0.00011286216455831449,
+      "loss": 2.5241,
+      "step": 9233
+    },
+    {
+      "epoch": 0.745218303607457,
+      "grad_norm": 0.7310191988945007,
+      "learning_rate": 0.00011284650863043407,
+      "loss": 2.5777,
+      "step": 9234
+    },
+    {
+      "epoch": 0.74529900734404,
+      "grad_norm": 0.6661474704742432,
+      "learning_rate": 0.00011283085238238503,
+      "loss": 2.5471,
+      "step": 9235
+    },
+    {
+      "epoch": 0.7453797110806231,
+      "grad_norm": 0.7697983384132385,
+      "learning_rate": 0.00011281519581455761,
+      "loss": 2.587,
+      "step": 9236
+    },
+    {
+      "epoch": 0.745460414817206,
+      "grad_norm": 0.7336567640304565,
+      "learning_rate": 0.00011279953892734203,
+      "loss": 2.5756,
+      "step": 9237
+    },
+    {
+      "epoch": 0.745541118553789,
+      "grad_norm": 0.6192059516906738,
+      "learning_rate": 0.00011278388172112848,
+      "loss": 2.5038,
+      "step": 9238
+    },
+    {
+      "epoch": 0.745621822290372,
+      "grad_norm": 0.7180300354957581,
+      "learning_rate": 0.00011276822419630719,
+      "loss": 2.5469,
+      "step": 9239
+    },
+    {
+      "epoch": 0.7457025260269551,
+      "grad_norm": 0.7583367824554443,
+      "learning_rate": 0.00011275256635326837,
+      "loss": 2.6274,
+      "step": 9240
+    },
+    {
+      "epoch": 0.7457832297635381,
+      "grad_norm": 0.6848096251487732,
+      "learning_rate": 0.00011273690819240221,
+      "loss": 2.5117,
+      "step": 9241
+    },
+    {
+      "epoch": 0.745863933500121,
+      "grad_norm": 0.6830503344535828,
+      "learning_rate": 0.00011272124971409907,
+      "loss": 2.5114,
+      "step": 9242
+    },
+    {
+      "epoch": 0.745944637236704,
+      "grad_norm": 0.780240535736084,
+      "learning_rate": 0.0001127055909187491,
+      "loss": 2.6432,
+      "step": 9243
+    },
+    {
+      "epoch": 0.7460253409732871,
+      "grad_norm": 0.7421274185180664,
+      "learning_rate": 0.00011268993180674261,
+      "loss": 2.5723,
+      "step": 9244
+    },
+    {
+      "epoch": 0.7461060447098701,
+      "grad_norm": 0.6695685386657715,
+      "learning_rate": 0.00011267427237846986,
+      "loss": 2.5335,
+      "step": 9245
+    },
+    {
+      "epoch": 0.746186748446453,
+      "grad_norm": 0.8390316963195801,
+      "learning_rate": 0.00011265861263432104,
+      "loss": 2.5125,
+      "step": 9246
+    },
+    {
+      "epoch": 0.746267452183036,
+      "grad_norm": 0.7030535936355591,
+      "learning_rate": 0.00011264295257468658,
+      "loss": 2.5986,
+      "step": 9247
+    },
+    {
+      "epoch": 0.7463481559196191,
+      "grad_norm": 0.6754253506660461,
+      "learning_rate": 0.00011262729219995669,
+      "loss": 2.5067,
+      "step": 9248
+    },
+    {
+      "epoch": 0.7464288596562021,
+      "grad_norm": 0.6809592843055725,
+      "learning_rate": 0.00011261163151052163,
+      "loss": 2.5359,
+      "step": 9249
+    },
+    {
+      "epoch": 0.7465095633927851,
+      "grad_norm": 0.6546878218650818,
+      "learning_rate": 0.00011259597050677178,
+      "loss": 2.5357,
+      "step": 9250
+    },
+    {
+      "epoch": 0.746590267129368,
+      "grad_norm": 0.6514731645584106,
+      "learning_rate": 0.00011258030918909739,
+      "loss": 2.5591,
+      "step": 9251
+    },
+    {
+      "epoch": 0.7466709708659511,
+      "grad_norm": 0.6981258392333984,
+      "learning_rate": 0.0001125646475578888,
+      "loss": 2.6171,
+      "step": 9252
+    },
+    {
+      "epoch": 0.7467516746025341,
+      "grad_norm": 0.6763784885406494,
+      "learning_rate": 0.00011254898561353639,
+      "loss": 2.5455,
+      "step": 9253
+    },
+    {
+      "epoch": 0.7468323783391171,
+      "grad_norm": 0.6241726279258728,
+      "learning_rate": 0.00011253332335643043,
+      "loss": 2.6073,
+      "step": 9254
+    },
+    {
+      "epoch": 0.7469130820757001,
+      "grad_norm": 0.6810312271118164,
+      "learning_rate": 0.00011251766078696132,
+      "loss": 2.5285,
+      "step": 9255
+    },
+    {
+      "epoch": 0.7469937858122832,
+      "grad_norm": 0.6603971123695374,
+      "learning_rate": 0.00011250199790551934,
+      "loss": 2.5985,
+      "step": 9256
+    },
+    {
+      "epoch": 0.7470744895488661,
+      "grad_norm": 0.69618159532547,
+      "learning_rate": 0.0001124863347124949,
+      "loss": 2.5728,
+      "step": 9257
+    },
+    {
+      "epoch": 0.7471551932854491,
+      "grad_norm": 0.6878889203071594,
+      "learning_rate": 0.00011247067120827837,
+      "loss": 2.5459,
+      "step": 9258
+    },
+    {
+      "epoch": 0.7472358970220321,
+      "grad_norm": 0.6613149046897888,
+      "learning_rate": 0.00011245500739326011,
+      "loss": 2.6559,
+      "step": 9259
+    },
+    {
+      "epoch": 0.7473166007586152,
+      "grad_norm": 0.6397448778152466,
+      "learning_rate": 0.00011243934326783053,
+      "loss": 2.5712,
+      "step": 9260
+    },
+    {
+      "epoch": 0.7473973044951981,
+      "grad_norm": 0.6804259419441223,
+      "learning_rate": 0.00011242367883237996,
+      "loss": 2.6143,
+      "step": 9261
+    },
+    {
+      "epoch": 0.7474780082317811,
+      "grad_norm": 0.8029066324234009,
+      "learning_rate": 0.00011240801408729884,
+      "loss": 2.5702,
+      "step": 9262
+    },
+    {
+      "epoch": 0.7475587119683641,
+      "grad_norm": 0.7086285948753357,
+      "learning_rate": 0.00011239234903297761,
+      "loss": 2.6113,
+      "step": 9263
+    },
+    {
+      "epoch": 0.7476394157049472,
+      "grad_norm": 0.6980452537536621,
+      "learning_rate": 0.00011237668366980665,
+      "loss": 2.6355,
+      "step": 9264
+    },
+    {
+      "epoch": 0.7477201194415302,
+      "grad_norm": 0.6906906962394714,
+      "learning_rate": 0.00011236101799817636,
+      "loss": 2.5605,
+      "step": 9265
+    },
+    {
+      "epoch": 0.7478008231781131,
+      "grad_norm": 0.7412894368171692,
+      "learning_rate": 0.00011234535201847716,
+      "loss": 2.6073,
+      "step": 9266
+    },
+    {
+      "epoch": 0.7478815269146961,
+      "grad_norm": 0.6949330568313599,
+      "learning_rate": 0.00011232968573109955,
+      "loss": 2.5623,
+      "step": 9267
+    },
+    {
+      "epoch": 0.7479622306512792,
+      "grad_norm": 0.6916515827178955,
+      "learning_rate": 0.00011231401913643393,
+      "loss": 2.5348,
+      "step": 9268
+    },
+    {
+      "epoch": 0.7480429343878622,
+      "grad_norm": 0.7576180696487427,
+      "learning_rate": 0.0001122983522348708,
+      "loss": 2.5968,
+      "step": 9269
+    },
+    {
+      "epoch": 0.7481236381244452,
+      "grad_norm": 0.6734197735786438,
+      "learning_rate": 0.00011228268502680052,
+      "loss": 2.5185,
+      "step": 9270
+    },
+    {
+      "epoch": 0.7482043418610281,
+      "grad_norm": 0.6952544450759888,
+      "learning_rate": 0.00011226701751261367,
+      "loss": 2.57,
+      "step": 9271
+    },
+    {
+      "epoch": 0.7482850455976112,
+      "grad_norm": 0.6504654884338379,
+      "learning_rate": 0.00011225134969270068,
+      "loss": 2.5677,
+      "step": 9272
+    },
+    {
+      "epoch": 0.7483657493341942,
+      "grad_norm": 0.6843643188476562,
+      "learning_rate": 0.00011223568156745198,
+      "loss": 2.5686,
+      "step": 9273
+    },
+    {
+      "epoch": 0.7484464530707772,
+      "grad_norm": 0.6786371469497681,
+      "learning_rate": 0.00011222001313725816,
+      "loss": 2.5024,
+      "step": 9274
+    },
+    {
+      "epoch": 0.7485271568073602,
+      "grad_norm": 0.6431117057800293,
+      "learning_rate": 0.00011220434440250967,
+      "loss": 2.5206,
+      "step": 9275
+    },
+    {
+      "epoch": 0.7486078605439432,
+      "grad_norm": 0.699547290802002,
+      "learning_rate": 0.000112188675363597,
+      "loss": 2.5974,
+      "step": 9276
+    },
+    {
+      "epoch": 0.7486885642805262,
+      "grad_norm": 0.6870436072349548,
+      "learning_rate": 0.00011217300602091067,
+      "loss": 2.5303,
+      "step": 9277
+    },
+    {
+      "epoch": 0.7487692680171092,
+      "grad_norm": 0.7032173871994019,
+      "learning_rate": 0.0001121573363748412,
+      "loss": 2.5045,
+      "step": 9278
+    },
+    {
+      "epoch": 0.7488499717536922,
+      "grad_norm": 0.6890417337417603,
+      "learning_rate": 0.00011214166642577917,
+      "loss": 2.5945,
+      "step": 9279
+    },
+    {
+      "epoch": 0.7489306754902753,
+      "grad_norm": 0.7257806062698364,
+      "learning_rate": 0.00011212599617411506,
+      "loss": 2.6013,
+      "step": 9280
+    },
+    {
+      "epoch": 0.7490113792268582,
+      "grad_norm": 0.722561240196228,
+      "learning_rate": 0.0001121103256202394,
+      "loss": 2.5809,
+      "step": 9281
+    },
+    {
+      "epoch": 0.7490920829634412,
+      "grad_norm": 0.7360994219779968,
+      "learning_rate": 0.00011209465476454277,
+      "loss": 2.5036,
+      "step": 9282
+    },
+    {
+      "epoch": 0.7491727867000242,
+      "grad_norm": 0.6561676263809204,
+      "learning_rate": 0.00011207898360741574,
+      "loss": 2.5302,
+      "step": 9283
+    },
+    {
+      "epoch": 0.7492534904366072,
+      "grad_norm": 0.7454147338867188,
+      "learning_rate": 0.00011206331214924887,
+      "loss": 2.5511,
+      "step": 9284
+    },
+    {
+      "epoch": 0.7493341941731902,
+      "grad_norm": 0.7085482478141785,
+      "learning_rate": 0.00011204764039043275,
+      "loss": 2.5743,
+      "step": 9285
+    },
+    {
+      "epoch": 0.7494148979097732,
+      "grad_norm": 0.691872775554657,
+      "learning_rate": 0.0001120319683313579,
+      "loss": 2.5414,
+      "step": 9286
+    },
+    {
+      "epoch": 0.7494956016463562,
+      "grad_norm": 0.6661050915718079,
+      "learning_rate": 0.00011201629597241496,
+      "loss": 2.5418,
+      "step": 9287
+    },
+    {
+      "epoch": 0.7495763053829392,
+      "grad_norm": 0.7440990805625916,
+      "learning_rate": 0.00011200062331399452,
+      "loss": 2.5543,
+      "step": 9288
+    },
+    {
+      "epoch": 0.7496570091195223,
+      "grad_norm": 0.6655303835868835,
+      "learning_rate": 0.00011198495035648715,
+      "loss": 2.5629,
+      "step": 9289
+    },
+    {
+      "epoch": 0.7497377128561052,
+      "grad_norm": 0.7550996541976929,
+      "learning_rate": 0.00011196927710028353,
+      "loss": 2.5376,
+      "step": 9290
+    },
+    {
+      "epoch": 0.7498184165926882,
+      "grad_norm": 0.692915678024292,
+      "learning_rate": 0.00011195360354577422,
+      "loss": 2.4661,
+      "step": 9291
+    },
+    {
+      "epoch": 0.7498991203292712,
+      "grad_norm": 0.7572253346443176,
+      "learning_rate": 0.00011193792969334985,
+      "loss": 2.5641,
+      "step": 9292
+    },
+    {
+      "epoch": 0.7499798240658543,
+      "grad_norm": 0.6550531387329102,
+      "learning_rate": 0.00011192225554340107,
+      "loss": 2.5591,
+      "step": 9293
+    },
+    {
+      "epoch": 0.7500605278024373,
+      "grad_norm": 0.677130401134491,
+      "learning_rate": 0.0001119065810963185,
+      "loss": 2.5859,
+      "step": 9294
+    },
+    {
+      "epoch": 0.7501412315390202,
+      "grad_norm": 0.680673360824585,
+      "learning_rate": 0.00011189090635249287,
+      "loss": 2.5343,
+      "step": 9295
+    },
+    {
+      "epoch": 0.7502219352756032,
+      "grad_norm": 0.7574957609176636,
+      "learning_rate": 0.00011187523131231472,
+      "loss": 2.5966,
+      "step": 9296
+    },
+    {
+      "epoch": 0.7503026390121863,
+      "grad_norm": 0.7099971175193787,
+      "learning_rate": 0.00011185955597617474,
+      "loss": 2.5547,
+      "step": 9297
+    },
+    {
+      "epoch": 0.7503833427487693,
+      "grad_norm": 0.7153162956237793,
+      "learning_rate": 0.00011184388034446367,
+      "loss": 2.5986,
+      "step": 9298
+    },
+    {
+      "epoch": 0.7504640464853523,
+      "grad_norm": 0.7154852747917175,
+      "learning_rate": 0.00011182820441757212,
+      "loss": 2.5214,
+      "step": 9299
+    },
+    {
+      "epoch": 0.7505447502219352,
+      "grad_norm": 0.6899208426475525,
+      "learning_rate": 0.00011181252819589081,
+      "loss": 2.5026,
+      "step": 9300
+    },
+    {
+      "epoch": 0.7506254539585183,
+      "grad_norm": 0.6719048023223877,
+      "learning_rate": 0.00011179685167981041,
+      "loss": 2.5915,
+      "step": 9301
+    },
+    {
+      "epoch": 0.7507061576951013,
+      "grad_norm": 0.6664413213729858,
+      "learning_rate": 0.00011178117486972164,
+      "loss": 2.5479,
+      "step": 9302
+    },
+    {
+      "epoch": 0.7507868614316843,
+      "grad_norm": 0.7433286905288696,
+      "learning_rate": 0.00011176549776601517,
+      "loss": 2.5941,
+      "step": 9303
+    },
+    {
+      "epoch": 0.7508675651682672,
+      "grad_norm": 0.7868518233299255,
+      "learning_rate": 0.00011174982036908177,
+      "loss": 2.5537,
+      "step": 9304
+    },
+    {
+      "epoch": 0.7509482689048503,
+      "grad_norm": 0.7037336826324463,
+      "learning_rate": 0.0001117341426793121,
+      "loss": 2.568,
+      "step": 9305
+    },
+    {
+      "epoch": 0.7510289726414333,
+      "grad_norm": 0.6630405783653259,
+      "learning_rate": 0.00011171846469709697,
+      "loss": 2.4906,
+      "step": 9306
+    },
+    {
+      "epoch": 0.7511096763780163,
+      "grad_norm": 0.7398669719696045,
+      "learning_rate": 0.00011170278642282701,
+      "loss": 2.574,
+      "step": 9307
+    },
+    {
+      "epoch": 0.7511903801145993,
+      "grad_norm": 0.7557641267776489,
+      "learning_rate": 0.00011168710785689304,
+      "loss": 2.5237,
+      "step": 9308
+    },
+    {
+      "epoch": 0.7512710838511824,
+      "grad_norm": 0.6883708238601685,
+      "learning_rate": 0.00011167142899968581,
+      "loss": 2.5643,
+      "step": 9309
+    },
+    {
+      "epoch": 0.7513517875877653,
+      "grad_norm": 0.6623669862747192,
+      "learning_rate": 0.00011165574985159606,
+      "loss": 2.5319,
+      "step": 9310
+    },
+    {
+      "epoch": 0.7514324913243483,
+      "grad_norm": 0.6938778758049011,
+      "learning_rate": 0.00011164007041301454,
+      "loss": 2.5083,
+      "step": 9311
+    },
+    {
+      "epoch": 0.7515131950609313,
+      "grad_norm": 0.718534529209137,
+      "learning_rate": 0.00011162439068433204,
+      "loss": 2.4791,
+      "step": 9312
+    },
+    {
+      "epoch": 0.7515938987975144,
+      "grad_norm": 0.672113299369812,
+      "learning_rate": 0.00011160871066593934,
+      "loss": 2.5264,
+      "step": 9313
+    },
+    {
+      "epoch": 0.7516746025340973,
+      "grad_norm": 0.6854343414306641,
+      "learning_rate": 0.00011159303035822723,
+      "loss": 2.5734,
+      "step": 9314
+    },
+    {
+      "epoch": 0.7517553062706803,
+      "grad_norm": 0.6494589447975159,
+      "learning_rate": 0.0001115773497615865,
+      "loss": 2.5564,
+      "step": 9315
+    },
+    {
+      "epoch": 0.7518360100072633,
+      "grad_norm": 0.7219608426094055,
+      "learning_rate": 0.00011156166887640793,
+      "loss": 2.6049,
+      "step": 9316
+    },
+    {
+      "epoch": 0.7519167137438464,
+      "grad_norm": 0.6892502903938293,
+      "learning_rate": 0.00011154598770308236,
+      "loss": 2.5333,
+      "step": 9317
+    },
+    {
+      "epoch": 0.7519974174804294,
+      "grad_norm": 0.6670175790786743,
+      "learning_rate": 0.0001115303062420006,
+      "loss": 2.5882,
+      "step": 9318
+    },
+    {
+      "epoch": 0.7520781212170123,
+      "grad_norm": 0.7367776036262512,
+      "learning_rate": 0.00011151462449355347,
+      "loss": 2.5634,
+      "step": 9319
+    },
+    {
+      "epoch": 0.7521588249535953,
+      "grad_norm": 0.6971952319145203,
+      "learning_rate": 0.00011149894245813182,
+      "loss": 2.5323,
+      "step": 9320
+    },
+    {
+      "epoch": 0.7522395286901784,
+      "grad_norm": 0.6555755734443665,
+      "learning_rate": 0.00011148326013612642,
+      "loss": 2.5597,
+      "step": 9321
+    },
+    {
+      "epoch": 0.7523202324267614,
+      "grad_norm": 0.7004384994506836,
+      "learning_rate": 0.00011146757752792819,
+      "loss": 2.4761,
+      "step": 9322
+    },
+    {
+      "epoch": 0.7524009361633444,
+      "grad_norm": 0.7151978015899658,
+      "learning_rate": 0.00011145189463392791,
+      "loss": 2.5825,
+      "step": 9323
+    },
+    {
+      "epoch": 0.7524816398999273,
+      "grad_norm": 0.7176918387413025,
+      "learning_rate": 0.00011143621145451653,
+      "loss": 2.6112,
+      "step": 9324
+    },
+    {
+      "epoch": 0.7525623436365104,
+      "grad_norm": 0.7156146168708801,
+      "learning_rate": 0.00011142052799008487,
+      "loss": 2.5293,
+      "step": 9325
+    },
+    {
+      "epoch": 0.7526430473730934,
+      "grad_norm": 0.7360113263130188,
+      "learning_rate": 0.00011140484424102375,
+      "loss": 2.5703,
+      "step": 9326
+    },
+    {
+      "epoch": 0.7527237511096764,
+      "grad_norm": 0.65630042552948,
+      "learning_rate": 0.00011138916020772414,
+      "loss": 2.5224,
+      "step": 9327
+    },
+    {
+      "epoch": 0.7528044548462594,
+      "grad_norm": 0.7088161110877991,
+      "learning_rate": 0.00011137347589057687,
+      "loss": 2.6673,
+      "step": 9328
+    },
+    {
+      "epoch": 0.7528851585828424,
+      "grad_norm": 0.7335243821144104,
+      "learning_rate": 0.00011135779128997283,
+      "loss": 2.5693,
+      "step": 9329
+    },
+    {
+      "epoch": 0.7529658623194254,
+      "grad_norm": 0.7166211605072021,
+      "learning_rate": 0.00011134210640630298,
+      "loss": 2.5612,
+      "step": 9330
+    },
+    {
+      "epoch": 0.7530465660560084,
+      "grad_norm": 0.7324960231781006,
+      "learning_rate": 0.00011132642123995816,
+      "loss": 2.5682,
+      "step": 9331
+    },
+    {
+      "epoch": 0.7531272697925914,
+      "grad_norm": 0.7133917808532715,
+      "learning_rate": 0.00011131073579132936,
+      "loss": 2.6131,
+      "step": 9332
+    },
+    {
+      "epoch": 0.7532079735291743,
+      "grad_norm": 0.678741455078125,
+      "learning_rate": 0.0001112950500608074,
+      "loss": 2.6109,
+      "step": 9333
+    },
+    {
+      "epoch": 0.7532886772657574,
+      "grad_norm": 0.7000784277915955,
+      "learning_rate": 0.0001112793640487833,
+      "loss": 2.5087,
+      "step": 9334
+    },
+    {
+      "epoch": 0.7533693810023404,
+      "grad_norm": 0.719976544380188,
+      "learning_rate": 0.00011126367775564795,
+      "loss": 2.4665,
+      "step": 9335
+    },
+    {
+      "epoch": 0.7534500847389234,
+      "grad_norm": 0.7127155065536499,
+      "learning_rate": 0.00011124799118179232,
+      "loss": 2.5254,
+      "step": 9336
+    },
+    {
+      "epoch": 0.7535307884755064,
+      "grad_norm": 0.6306474804878235,
+      "learning_rate": 0.00011123230432760734,
+      "loss": 2.5487,
+      "step": 9337
+    },
+    {
+      "epoch": 0.7536114922120895,
+      "grad_norm": 0.667019784450531,
+      "learning_rate": 0.00011121661719348397,
+      "loss": 2.5576,
+      "step": 9338
+    },
+    {
+      "epoch": 0.7536921959486724,
+      "grad_norm": 0.6869673132896423,
+      "learning_rate": 0.00011120092977981318,
+      "loss": 2.544,
+      "step": 9339
+    },
+    {
+      "epoch": 0.7537728996852554,
+      "grad_norm": 0.6688670516014099,
+      "learning_rate": 0.00011118524208698596,
+      "loss": 2.6017,
+      "step": 9340
+    },
+    {
+      "epoch": 0.7538536034218384,
+      "grad_norm": 0.6717860102653503,
+      "learning_rate": 0.00011116955411539325,
+      "loss": 2.5571,
+      "step": 9341
+    },
+    {
+      "epoch": 0.7539343071584215,
+      "grad_norm": 0.7113999724388123,
+      "learning_rate": 0.00011115386586542604,
+      "loss": 2.5684,
+      "step": 9342
+    },
+    {
+      "epoch": 0.7540150108950044,
+      "grad_norm": 0.6687907576560974,
+      "learning_rate": 0.00011113817733747536,
+      "loss": 2.548,
+      "step": 9343
+    },
+    {
+      "epoch": 0.7540957146315874,
+      "grad_norm": 0.6828920841217041,
+      "learning_rate": 0.00011112248853193219,
+      "loss": 2.5544,
+      "step": 9344
+    },
+    {
+      "epoch": 0.7541764183681704,
+      "grad_norm": 0.6793262362480164,
+      "learning_rate": 0.00011110679944918749,
+      "loss": 2.4655,
+      "step": 9345
+    },
+    {
+      "epoch": 0.7542571221047535,
+      "grad_norm": 0.6812230348587036,
+      "learning_rate": 0.00011109111008963235,
+      "loss": 2.5473,
+      "step": 9346
+    },
+    {
+      "epoch": 0.7543378258413365,
+      "grad_norm": 0.6838300824165344,
+      "learning_rate": 0.00011107542045365775,
+      "loss": 2.5248,
+      "step": 9347
+    },
+    {
+      "epoch": 0.7544185295779194,
+      "grad_norm": 0.7101932764053345,
+      "learning_rate": 0.0001110597305416547,
+      "loss": 2.5235,
+      "step": 9348
+    },
+    {
+      "epoch": 0.7544992333145024,
+      "grad_norm": 0.7136144042015076,
+      "learning_rate": 0.0001110440403540143,
+      "loss": 2.5592,
+      "step": 9349
+    },
+    {
+      "epoch": 0.7545799370510855,
+      "grad_norm": 0.6673154234886169,
+      "learning_rate": 0.00011102834989112751,
+      "loss": 2.4962,
+      "step": 9350
+    },
+    {
+      "epoch": 0.7546606407876685,
+      "grad_norm": 0.6849049925804138,
+      "learning_rate": 0.00011101265915338544,
+      "loss": 2.5793,
+      "step": 9351
+    },
+    {
+      "epoch": 0.7547413445242515,
+      "grad_norm": 0.7239733338356018,
+      "learning_rate": 0.0001109969681411791,
+      "loss": 2.5556,
+      "step": 9352
+    },
+    {
+      "epoch": 0.7548220482608344,
+      "grad_norm": 0.6738215684890747,
+      "learning_rate": 0.00011098127685489955,
+      "loss": 2.6181,
+      "step": 9353
+    },
+    {
+      "epoch": 0.7549027519974175,
+      "grad_norm": 0.6212114095687866,
+      "learning_rate": 0.00011096558529493787,
+      "loss": 2.5509,
+      "step": 9354
+    },
+    {
+      "epoch": 0.7549834557340005,
+      "grad_norm": 0.6801952123641968,
+      "learning_rate": 0.00011094989346168517,
+      "loss": 2.6454,
+      "step": 9355
+    },
+    {
+      "epoch": 0.7550641594705835,
+      "grad_norm": 0.6605944037437439,
+      "learning_rate": 0.0001109342013555325,
+      "loss": 2.5218,
+      "step": 9356
+    },
+    {
+      "epoch": 0.7551448632071665,
+      "grad_norm": 0.6486438512802124,
+      "learning_rate": 0.00011091850897687096,
+      "loss": 2.5431,
+      "step": 9357
+    },
+    {
+      "epoch": 0.7552255669437495,
+      "grad_norm": 0.6701794266700745,
+      "learning_rate": 0.0001109028163260916,
+      "loss": 2.563,
+      "step": 9358
+    },
+    {
+      "epoch": 0.7553062706803325,
+      "grad_norm": 0.6486446261405945,
+      "learning_rate": 0.00011088712340358555,
+      "loss": 2.5147,
+      "step": 9359
+    },
+    {
+      "epoch": 0.7553869744169155,
+      "grad_norm": 0.695197582244873,
+      "learning_rate": 0.00011087143020974396,
+      "loss": 2.5707,
+      "step": 9360
+    },
+    {
+      "epoch": 0.7554676781534985,
+      "grad_norm": 0.6910821199417114,
+      "learning_rate": 0.00011085573674495791,
+      "loss": 2.5797,
+      "step": 9361
+    },
+    {
+      "epoch": 0.7555483818900816,
+      "grad_norm": 0.7084208726882935,
+      "learning_rate": 0.00011084004300961852,
+      "loss": 2.5362,
+      "step": 9362
+    },
+    {
+      "epoch": 0.7556290856266645,
+      "grad_norm": 0.6750916242599487,
+      "learning_rate": 0.00011082434900411691,
+      "loss": 2.5554,
+      "step": 9363
+    },
+    {
+      "epoch": 0.7557097893632475,
+      "grad_norm": 0.6711466908454895,
+      "learning_rate": 0.0001108086547288442,
+      "loss": 2.5577,
+      "step": 9364
+    },
+    {
+      "epoch": 0.7557904930998305,
+      "grad_norm": 0.7267118096351624,
+      "learning_rate": 0.00011079296018419163,
+      "loss": 2.5422,
+      "step": 9365
+    },
+    {
+      "epoch": 0.7558711968364136,
+      "grad_norm": 0.692730188369751,
+      "learning_rate": 0.00011077726537055021,
+      "loss": 2.5281,
+      "step": 9366
+    },
+    {
+      "epoch": 0.7559519005729965,
+      "grad_norm": 0.7071926593780518,
+      "learning_rate": 0.00011076157028831122,
+      "loss": 2.5273,
+      "step": 9367
+    },
+    {
+      "epoch": 0.7560326043095795,
+      "grad_norm": 0.7662521600723267,
+      "learning_rate": 0.00011074587493786574,
+      "loss": 2.5433,
+      "step": 9368
+    },
+    {
+      "epoch": 0.7561133080461625,
+      "grad_norm": 0.7173436880111694,
+      "learning_rate": 0.00011073017931960496,
+      "loss": 2.579,
+      "step": 9369
+    },
+    {
+      "epoch": 0.7561940117827456,
+      "grad_norm": 0.6401154398918152,
+      "learning_rate": 0.00011071448343392008,
+      "loss": 2.5189,
+      "step": 9370
+    },
+    {
+      "epoch": 0.7562747155193286,
+      "grad_norm": 0.6510714292526245,
+      "learning_rate": 0.00011069878728120224,
+      "loss": 2.5682,
+      "step": 9371
+    },
+    {
+      "epoch": 0.7563554192559115,
+      "grad_norm": 0.7189988493919373,
+      "learning_rate": 0.00011068309086184269,
+      "loss": 2.5247,
+      "step": 9372
+    },
+    {
+      "epoch": 0.7564361229924945,
+      "grad_norm": 0.678753137588501,
+      "learning_rate": 0.00011066739417623258,
+      "loss": 2.5083,
+      "step": 9373
+    },
+    {
+      "epoch": 0.7565168267290776,
+      "grad_norm": 0.6903115510940552,
+      "learning_rate": 0.0001106516972247631,
+      "loss": 2.5658,
+      "step": 9374
+    },
+    {
+      "epoch": 0.7565975304656606,
+      "grad_norm": 0.6772382855415344,
+      "learning_rate": 0.0001106360000078255,
+      "loss": 2.5445,
+      "step": 9375
+    },
+    {
+      "epoch": 0.7566782342022436,
+      "grad_norm": 0.6655055284500122,
+      "learning_rate": 0.00011062030252581097,
+      "loss": 2.5186,
+      "step": 9376
+    },
+    {
+      "epoch": 0.7567589379388265,
+      "grad_norm": 0.7173851728439331,
+      "learning_rate": 0.00011060460477911074,
+      "loss": 2.5297,
+      "step": 9377
+    },
+    {
+      "epoch": 0.7568396416754096,
+      "grad_norm": 0.6891282200813293,
+      "learning_rate": 0.00011058890676811606,
+      "loss": 2.5706,
+      "step": 9378
+    },
+    {
+      "epoch": 0.7569203454119926,
+      "grad_norm": 0.7053082585334778,
+      "learning_rate": 0.0001105732084932181,
+      "loss": 2.5475,
+      "step": 9379
+    },
+    {
+      "epoch": 0.7570010491485756,
+      "grad_norm": 0.7503373622894287,
+      "learning_rate": 0.00011055750995480818,
+      "loss": 2.6438,
+      "step": 9380
+    },
+    {
+      "epoch": 0.7570817528851586,
+      "grad_norm": 0.6703453660011292,
+      "learning_rate": 0.0001105418111532775,
+      "loss": 2.5485,
+      "step": 9381
+    },
+    {
+      "epoch": 0.7571624566217416,
+      "grad_norm": 0.6651757955551147,
+      "learning_rate": 0.00011052611208901733,
+      "loss": 2.6079,
+      "step": 9382
+    },
+    {
+      "epoch": 0.7572431603583246,
+      "grad_norm": 0.6738902926445007,
+      "learning_rate": 0.00011051041276241895,
+      "loss": 2.5279,
+      "step": 9383
+    },
+    {
+      "epoch": 0.7573238640949076,
+      "grad_norm": 0.6803816556930542,
+      "learning_rate": 0.00011049471317387357,
+      "loss": 2.5972,
+      "step": 9384
+    },
+    {
+      "epoch": 0.7574045678314906,
+      "grad_norm": 0.7127584218978882,
+      "learning_rate": 0.00011047901332377253,
+      "loss": 2.5275,
+      "step": 9385
+    },
+    {
+      "epoch": 0.7574852715680735,
+      "grad_norm": 0.7655676007270813,
+      "learning_rate": 0.00011046331321250711,
+      "loss": 2.6491,
+      "step": 9386
+    },
+    {
+      "epoch": 0.7575659753046566,
+      "grad_norm": 0.7005762457847595,
+      "learning_rate": 0.00011044761284046854,
+      "loss": 2.5266,
+      "step": 9387
+    },
+    {
+      "epoch": 0.7576466790412396,
+      "grad_norm": 0.701931357383728,
+      "learning_rate": 0.00011043191220804817,
+      "loss": 2.5556,
+      "step": 9388
+    },
+    {
+      "epoch": 0.7577273827778226,
+      "grad_norm": 0.6888757944107056,
+      "learning_rate": 0.00011041621131563724,
+      "loss": 2.5654,
+      "step": 9389
+    },
+    {
+      "epoch": 0.7578080865144056,
+      "grad_norm": 0.7119149565696716,
+      "learning_rate": 0.00011040051016362711,
+      "loss": 2.5925,
+      "step": 9390
+    },
+    {
+      "epoch": 0.7578887902509887,
+      "grad_norm": 0.7378301024436951,
+      "learning_rate": 0.00011038480875240911,
+      "loss": 2.5604,
+      "step": 9391
+    },
+    {
+      "epoch": 0.7579694939875716,
+      "grad_norm": 0.7221272587776184,
+      "learning_rate": 0.00011036910708237449,
+      "loss": 2.5293,
+      "step": 9392
+    },
+    {
+      "epoch": 0.7580501977241546,
+      "grad_norm": 0.6895891427993774,
+      "learning_rate": 0.00011035340515391465,
+      "loss": 2.5177,
+      "step": 9393
+    },
+    {
+      "epoch": 0.7581309014607376,
+      "grad_norm": 0.6812298893928528,
+      "learning_rate": 0.00011033770296742086,
+      "loss": 2.6345,
+      "step": 9394
+    },
+    {
+      "epoch": 0.7582116051973207,
+      "grad_norm": 0.6733750700950623,
+      "learning_rate": 0.00011032200052328449,
+      "loss": 2.5548,
+      "step": 9395
+    },
+    {
+      "epoch": 0.7582923089339036,
+      "grad_norm": 0.7667728066444397,
+      "learning_rate": 0.00011030629782189692,
+      "loss": 2.5858,
+      "step": 9396
+    },
+    {
+      "epoch": 0.7583730126704866,
+      "grad_norm": 0.6809018850326538,
+      "learning_rate": 0.00011029059486364946,
+      "loss": 2.6028,
+      "step": 9397
+    },
+    {
+      "epoch": 0.7584537164070696,
+      "grad_norm": 0.6817305684089661,
+      "learning_rate": 0.00011027489164893345,
+      "loss": 2.5594,
+      "step": 9398
+    },
+    {
+      "epoch": 0.7585344201436527,
+      "grad_norm": 0.6936343908309937,
+      "learning_rate": 0.00011025918817814027,
+      "loss": 2.4997,
+      "step": 9399
+    },
+    {
+      "epoch": 0.7586151238802357,
+      "grad_norm": 0.7046801447868347,
+      "learning_rate": 0.00011024348445166133,
+      "loss": 2.5199,
+      "step": 9400
+    },
+    {
+      "epoch": 0.7586958276168186,
+      "grad_norm": 0.7247316241264343,
+      "learning_rate": 0.00011022778046988798,
+      "loss": 2.5233,
+      "step": 9401
+    },
+    {
+      "epoch": 0.7587765313534016,
+      "grad_norm": 0.675652265548706,
+      "learning_rate": 0.00011021207623321162,
+      "loss": 2.5213,
+      "step": 9402
+    },
+    {
+      "epoch": 0.7588572350899847,
+      "grad_norm": 0.6866120100021362,
+      "learning_rate": 0.0001101963717420236,
+      "loss": 2.6026,
+      "step": 9403
+    },
+    {
+      "epoch": 0.7589379388265677,
+      "grad_norm": 0.7168806791305542,
+      "learning_rate": 0.00011018066699671534,
+      "loss": 2.5707,
+      "step": 9404
+    },
+    {
+      "epoch": 0.7590186425631507,
+      "grad_norm": 0.6858265995979309,
+      "learning_rate": 0.00011016496199767825,
+      "loss": 2.5313,
+      "step": 9405
+    },
+    {
+      "epoch": 0.7590993462997336,
+      "grad_norm": 0.7064315676689148,
+      "learning_rate": 0.00011014925674530375,
+      "loss": 2.5362,
+      "step": 9406
+    },
+    {
+      "epoch": 0.7591800500363167,
+      "grad_norm": 0.658385694026947,
+      "learning_rate": 0.00011013355123998324,
+      "loss": 2.5773,
+      "step": 9407
+    },
+    {
+      "epoch": 0.7592607537728997,
+      "grad_norm": 0.7112493515014648,
+      "learning_rate": 0.00011011784548210813,
+      "loss": 2.589,
+      "step": 9408
+    },
+    {
+      "epoch": 0.7593414575094827,
+      "grad_norm": 0.6835871934890747,
+      "learning_rate": 0.00011010213947206986,
+      "loss": 2.5952,
+      "step": 9409
+    },
+    {
+      "epoch": 0.7594221612460657,
+      "grad_norm": 0.6920506358146667,
+      "learning_rate": 0.00011008643321025989,
+      "loss": 2.5433,
+      "step": 9410
+    },
+    {
+      "epoch": 0.7595028649826487,
+      "grad_norm": 0.7239150404930115,
+      "learning_rate": 0.00011007072669706962,
+      "loss": 2.5291,
+      "step": 9411
+    },
+    {
+      "epoch": 0.7595835687192317,
+      "grad_norm": 0.644568145275116,
+      "learning_rate": 0.00011005501993289052,
+      "loss": 2.5324,
+      "step": 9412
+    },
+    {
+      "epoch": 0.7596642724558147,
+      "grad_norm": 0.6604863405227661,
+      "learning_rate": 0.00011003931291811405,
+      "loss": 2.561,
+      "step": 9413
+    },
+    {
+      "epoch": 0.7597449761923977,
+      "grad_norm": 0.7056753635406494,
+      "learning_rate": 0.00011002360565313164,
+      "loss": 2.6537,
+      "step": 9414
+    },
+    {
+      "epoch": 0.7598256799289808,
+      "grad_norm": 0.6712720394134521,
+      "learning_rate": 0.00011000789813833476,
+      "loss": 2.5222,
+      "step": 9415
+    },
+    {
+      "epoch": 0.7599063836655637,
+      "grad_norm": 0.6829253435134888,
+      "learning_rate": 0.00010999219037411492,
+      "loss": 2.5156,
+      "step": 9416
+    },
+    {
+      "epoch": 0.7599870874021467,
+      "grad_norm": 0.7386518120765686,
+      "learning_rate": 0.00010997648236086359,
+      "loss": 2.5378,
+      "step": 9417
+    },
+    {
+      "epoch": 0.7600677911387297,
+      "grad_norm": 0.6711105108261108,
+      "learning_rate": 0.00010996077409897223,
+      "loss": 2.4985,
+      "step": 9418
+    },
+    {
+      "epoch": 0.7601484948753128,
+      "grad_norm": 0.6936883926391602,
+      "learning_rate": 0.00010994506558883233,
+      "loss": 2.4912,
+      "step": 9419
+    },
+    {
+      "epoch": 0.7602291986118958,
+      "grad_norm": 0.6927978992462158,
+      "learning_rate": 0.00010992935683083541,
+      "loss": 2.5526,
+      "step": 9420
+    },
+    {
+      "epoch": 0.7603099023484787,
+      "grad_norm": 0.7661495804786682,
+      "learning_rate": 0.00010991364782537297,
+      "loss": 2.5778,
+      "step": 9421
+    },
+    {
+      "epoch": 0.7603906060850617,
+      "grad_norm": 0.7092108726501465,
+      "learning_rate": 0.0001098979385728365,
+      "loss": 2.6557,
+      "step": 9422
+    },
+    {
+      "epoch": 0.7604713098216448,
+      "grad_norm": 0.696666419506073,
+      "learning_rate": 0.00010988222907361754,
+      "loss": 2.4897,
+      "step": 9423
+    },
+    {
+      "epoch": 0.7605520135582278,
+      "grad_norm": 0.6836280822753906,
+      "learning_rate": 0.00010986651932810756,
+      "loss": 2.5146,
+      "step": 9424
+    },
+    {
+      "epoch": 0.7606327172948107,
+      "grad_norm": 0.7269579768180847,
+      "learning_rate": 0.00010985080933669815,
+      "loss": 2.5314,
+      "step": 9425
+    },
+    {
+      "epoch": 0.7607134210313937,
+      "grad_norm": 0.6862092018127441,
+      "learning_rate": 0.00010983509909978085,
+      "loss": 2.5415,
+      "step": 9426
+    },
+    {
+      "epoch": 0.7607941247679768,
+      "grad_norm": 0.7068747878074646,
+      "learning_rate": 0.00010981938861774713,
+      "loss": 2.5919,
+      "step": 9427
+    },
+    {
+      "epoch": 0.7608748285045598,
+      "grad_norm": 0.699999213218689,
+      "learning_rate": 0.0001098036778909886,
+      "loss": 2.5175,
+      "step": 9428
+    },
+    {
+      "epoch": 0.7609555322411428,
+      "grad_norm": 0.6642772555351257,
+      "learning_rate": 0.0001097879669198968,
+      "loss": 2.5721,
+      "step": 9429
+    },
+    {
+      "epoch": 0.7610362359777257,
+      "grad_norm": 0.7100533843040466,
+      "learning_rate": 0.00010977225570486323,
+      "loss": 2.5189,
+      "step": 9430
+    },
+    {
+      "epoch": 0.7611169397143088,
+      "grad_norm": 0.7289063930511475,
+      "learning_rate": 0.00010975654424627955,
+      "loss": 2.6139,
+      "step": 9431
+    },
+    {
+      "epoch": 0.7611976434508918,
+      "grad_norm": 0.7289659380912781,
+      "learning_rate": 0.00010974083254453726,
+      "loss": 2.5201,
+      "step": 9432
+    },
+    {
+      "epoch": 0.7612783471874748,
+      "grad_norm": 0.7389557957649231,
+      "learning_rate": 0.000109725120600028,
+      "loss": 2.559,
+      "step": 9433
+    },
+    {
+      "epoch": 0.7613590509240578,
+      "grad_norm": 0.7021538615226746,
+      "learning_rate": 0.00010970940841314327,
+      "loss": 2.6353,
+      "step": 9434
+    },
+    {
+      "epoch": 0.7614397546606407,
+      "grad_norm": 0.6614113450050354,
+      "learning_rate": 0.0001096936959842747,
+      "loss": 2.54,
+      "step": 9435
+    },
+    {
+      "epoch": 0.7615204583972238,
+      "grad_norm": 0.6905426979064941,
+      "learning_rate": 0.00010967798331381392,
+      "loss": 2.5845,
+      "step": 9436
+    },
+    {
+      "epoch": 0.7616011621338068,
+      "grad_norm": 0.8183904886245728,
+      "learning_rate": 0.00010966227040215247,
+      "loss": 2.5255,
+      "step": 9437
+    },
+    {
+      "epoch": 0.7616818658703898,
+      "grad_norm": 0.7404630780220032,
+      "learning_rate": 0.00010964655724968199,
+      "loss": 2.5726,
+      "step": 9438
+    },
+    {
+      "epoch": 0.7617625696069728,
+      "grad_norm": 0.657127320766449,
+      "learning_rate": 0.0001096308438567941,
+      "loss": 2.6233,
+      "step": 9439
+    },
+    {
+      "epoch": 0.7618432733435558,
+      "grad_norm": 0.7417906522750854,
+      "learning_rate": 0.00010961513022388039,
+      "loss": 2.6361,
+      "step": 9440
+    },
+    {
+      "epoch": 0.7619239770801388,
+      "grad_norm": 0.6930029988288879,
+      "learning_rate": 0.00010959941635133249,
+      "loss": 2.5164,
+      "step": 9441
+    },
+    {
+      "epoch": 0.7620046808167218,
+      "grad_norm": 0.6897261738777161,
+      "learning_rate": 0.00010958370223954207,
+      "loss": 2.5626,
+      "step": 9442
+    },
+    {
+      "epoch": 0.7620853845533048,
+      "grad_norm": 0.6737398505210876,
+      "learning_rate": 0.00010956798788890072,
+      "loss": 2.5342,
+      "step": 9443
+    },
+    {
+      "epoch": 0.7621660882898879,
+      "grad_norm": 0.6550001502037048,
+      "learning_rate": 0.0001095522732998001,
+      "loss": 2.5604,
+      "step": 9444
+    },
+    {
+      "epoch": 0.7622467920264708,
+      "grad_norm": 0.7184637784957886,
+      "learning_rate": 0.00010953655847263187,
+      "loss": 2.6006,
+      "step": 9445
+    },
+    {
+      "epoch": 0.7623274957630538,
+      "grad_norm": 0.6188609600067139,
+      "learning_rate": 0.00010952084340778766,
+      "loss": 2.4875,
+      "step": 9446
+    },
+    {
+      "epoch": 0.7624081994996368,
+      "grad_norm": 0.6550862789154053,
+      "learning_rate": 0.00010950512810565917,
+      "loss": 2.5794,
+      "step": 9447
+    },
+    {
+      "epoch": 0.7624889032362199,
+      "grad_norm": 0.6659231781959534,
+      "learning_rate": 0.000109489412566638,
+      "loss": 2.5137,
+      "step": 9448
+    },
+    {
+      "epoch": 0.7625696069728028,
+      "grad_norm": 0.749376118183136,
+      "learning_rate": 0.00010947369679111592,
+      "loss": 2.5923,
+      "step": 9449
+    },
+    {
+      "epoch": 0.7626503107093858,
+      "grad_norm": 0.6597894430160522,
+      "learning_rate": 0.0001094579807794845,
+      "loss": 2.5677,
+      "step": 9450
+    },
+    {
+      "epoch": 0.7627310144459688,
+      "grad_norm": 0.7194519639015198,
+      "learning_rate": 0.00010944226453213548,
+      "loss": 2.5754,
+      "step": 9451
+    },
+    {
+      "epoch": 0.7628117181825519,
+      "grad_norm": 0.6734583377838135,
+      "learning_rate": 0.00010942654804946057,
+      "loss": 2.535,
+      "step": 9452
+    },
+    {
+      "epoch": 0.7628924219191349,
+      "grad_norm": 0.7171904444694519,
+      "learning_rate": 0.00010941083133185146,
+      "loss": 2.5431,
+      "step": 9453
+    },
+    {
+      "epoch": 0.7629731256557178,
+      "grad_norm": 0.6760339736938477,
+      "learning_rate": 0.00010939511437969978,
+      "loss": 2.5163,
+      "step": 9454
+    },
+    {
+      "epoch": 0.7630538293923008,
+      "grad_norm": 0.6720966696739197,
+      "learning_rate": 0.00010937939719339731,
+      "loss": 2.5621,
+      "step": 9455
+    },
+    {
+      "epoch": 0.7631345331288839,
+      "grad_norm": 0.6374503970146179,
+      "learning_rate": 0.00010936367977333574,
+      "loss": 2.5007,
+      "step": 9456
+    },
+    {
+      "epoch": 0.7632152368654669,
+      "grad_norm": 0.6407146453857422,
+      "learning_rate": 0.00010934796211990684,
+      "loss": 2.5724,
+      "step": 9457
+    },
+    {
+      "epoch": 0.7632959406020499,
+      "grad_norm": 0.6685383319854736,
+      "learning_rate": 0.00010933224423350225,
+      "loss": 2.501,
+      "step": 9458
+    },
+    {
+      "epoch": 0.7633766443386328,
+      "grad_norm": 0.664806604385376,
+      "learning_rate": 0.00010931652611451373,
+      "loss": 2.6174,
+      "step": 9459
+    },
+    {
+      "epoch": 0.7634573480752159,
+      "grad_norm": 0.6383369565010071,
+      "learning_rate": 0.00010930080776333303,
+      "loss": 2.557,
+      "step": 9460
+    },
+    {
+      "epoch": 0.7635380518117989,
+      "grad_norm": 0.6747864484786987,
+      "learning_rate": 0.0001092850891803519,
+      "loss": 2.5406,
+      "step": 9461
+    },
+    {
+      "epoch": 0.7636187555483819,
+      "grad_norm": 0.7312811613082886,
+      "learning_rate": 0.00010926937036596205,
+      "loss": 2.5903,
+      "step": 9462
+    },
+    {
+      "epoch": 0.7636994592849649,
+      "grad_norm": 0.645847737789154,
+      "learning_rate": 0.00010925365132055529,
+      "loss": 2.5254,
+      "step": 9463
+    },
+    {
+      "epoch": 0.7637801630215479,
+      "grad_norm": 0.6466063857078552,
+      "learning_rate": 0.00010923793204452335,
+      "loss": 2.5322,
+      "step": 9464
+    },
+    {
+      "epoch": 0.7638608667581309,
+      "grad_norm": 0.6450574994087219,
+      "learning_rate": 0.000109222212538258,
+      "loss": 2.522,
+      "step": 9465
+    },
+    {
+      "epoch": 0.7639415704947139,
+      "grad_norm": 0.6491848826408386,
+      "learning_rate": 0.00010920649280215096,
+      "loss": 2.5545,
+      "step": 9466
+    },
+    {
+      "epoch": 0.7640222742312969,
+      "grad_norm": 0.6888336539268494,
+      "learning_rate": 0.0001091907728365941,
+      "loss": 2.5217,
+      "step": 9467
+    },
+    {
+      "epoch": 0.76410297796788,
+      "grad_norm": 0.702557384967804,
+      "learning_rate": 0.00010917505264197914,
+      "loss": 2.5351,
+      "step": 9468
+    },
+    {
+      "epoch": 0.7641836817044629,
+      "grad_norm": 0.6552408933639526,
+      "learning_rate": 0.0001091593322186979,
+      "loss": 2.5115,
+      "step": 9469
+    },
+    {
+      "epoch": 0.7642643854410459,
+      "grad_norm": 0.7514002919197083,
+      "learning_rate": 0.00010914361156714212,
+      "loss": 2.5196,
+      "step": 9470
+    },
+    {
+      "epoch": 0.7643450891776289,
+      "grad_norm": 0.6692500710487366,
+      "learning_rate": 0.00010912789068770366,
+      "loss": 2.5639,
+      "step": 9471
+    },
+    {
+      "epoch": 0.764425792914212,
+      "grad_norm": 0.6567397117614746,
+      "learning_rate": 0.0001091121695807743,
+      "loss": 2.5027,
+      "step": 9472
+    },
+    {
+      "epoch": 0.764506496650795,
+      "grad_norm": 0.6876057982444763,
+      "learning_rate": 0.00010909644824674587,
+      "loss": 2.519,
+      "step": 9473
+    },
+    {
+      "epoch": 0.7645872003873779,
+      "grad_norm": 0.747949481010437,
+      "learning_rate": 0.00010908072668601017,
+      "loss": 2.5604,
+      "step": 9474
+    },
+    {
+      "epoch": 0.7646679041239609,
+      "grad_norm": 0.6371368169784546,
+      "learning_rate": 0.000109065004898959,
+      "loss": 2.5853,
+      "step": 9475
+    },
+    {
+      "epoch": 0.764748607860544,
+      "grad_norm": 0.6472185254096985,
+      "learning_rate": 0.00010904928288598422,
+      "loss": 2.5662,
+      "step": 9476
+    },
+    {
+      "epoch": 0.764829311597127,
+      "grad_norm": 0.7009313702583313,
+      "learning_rate": 0.00010903356064747765,
+      "loss": 2.5244,
+      "step": 9477
+    },
+    {
+      "epoch": 0.76491001533371,
+      "grad_norm": 0.7405661940574646,
+      "learning_rate": 0.00010901783818383116,
+      "loss": 2.4963,
+      "step": 9478
+    },
+    {
+      "epoch": 0.7649907190702929,
+      "grad_norm": 0.7693421840667725,
+      "learning_rate": 0.00010900211549543658,
+      "loss": 2.6018,
+      "step": 9479
+    },
+    {
+      "epoch": 0.765071422806876,
+      "grad_norm": 0.6965410709381104,
+      "learning_rate": 0.00010898639258268571,
+      "loss": 2.627,
+      "step": 9480
+    },
+    {
+      "epoch": 0.765152126543459,
+      "grad_norm": 0.7167130708694458,
+      "learning_rate": 0.00010897066944597046,
+      "loss": 2.5298,
+      "step": 9481
+    },
+    {
+      "epoch": 0.765232830280042,
+      "grad_norm": 0.7159689664840698,
+      "learning_rate": 0.00010895494608568268,
+      "loss": 2.5179,
+      "step": 9482
+    },
+    {
+      "epoch": 0.7653135340166249,
+      "grad_norm": 0.7329332232475281,
+      "learning_rate": 0.00010893922250221423,
+      "loss": 2.6498,
+      "step": 9483
+    },
+    {
+      "epoch": 0.765394237753208,
+      "grad_norm": 0.6912567019462585,
+      "learning_rate": 0.000108923498695957,
+      "loss": 2.5679,
+      "step": 9484
+    },
+    {
+      "epoch": 0.765474941489791,
+      "grad_norm": 0.7030324935913086,
+      "learning_rate": 0.00010890777466730285,
+      "loss": 2.5678,
+      "step": 9485
+    },
+    {
+      "epoch": 0.765555645226374,
+      "grad_norm": 0.7238864898681641,
+      "learning_rate": 0.00010889205041664365,
+      "loss": 2.5525,
+      "step": 9486
+    },
+    {
+      "epoch": 0.765636348962957,
+      "grad_norm": 0.6623672842979431,
+      "learning_rate": 0.00010887632594437134,
+      "loss": 2.4857,
+      "step": 9487
+    },
+    {
+      "epoch": 0.7657170526995399,
+      "grad_norm": 0.726645827293396,
+      "learning_rate": 0.00010886060125087776,
+      "loss": 2.5405,
+      "step": 9488
+    },
+    {
+      "epoch": 0.765797756436123,
+      "grad_norm": 0.6624459624290466,
+      "learning_rate": 0.00010884487633655487,
+      "loss": 2.5538,
+      "step": 9489
+    },
+    {
+      "epoch": 0.765878460172706,
+      "grad_norm": 0.7198002934455872,
+      "learning_rate": 0.00010882915120179453,
+      "loss": 2.5808,
+      "step": 9490
+    },
+    {
+      "epoch": 0.765959163909289,
+      "grad_norm": 0.7545582056045532,
+      "learning_rate": 0.00010881342584698862,
+      "loss": 2.6059,
+      "step": 9491
+    },
+    {
+      "epoch": 0.766039867645872,
+      "grad_norm": 0.6748257279396057,
+      "learning_rate": 0.00010879770027252915,
+      "loss": 2.5203,
+      "step": 9492
+    },
+    {
+      "epoch": 0.766120571382455,
+      "grad_norm": 0.7376208901405334,
+      "learning_rate": 0.00010878197447880796,
+      "loss": 2.5255,
+      "step": 9493
+    },
+    {
+      "epoch": 0.766201275119038,
+      "grad_norm": 0.7589401006698608,
+      "learning_rate": 0.00010876624846621704,
+      "loss": 2.6304,
+      "step": 9494
+    },
+    {
+      "epoch": 0.766281978855621,
+      "grad_norm": 0.6963146924972534,
+      "learning_rate": 0.00010875052223514827,
+      "loss": 2.5547,
+      "step": 9495
+    },
+    {
+      "epoch": 0.766362682592204,
+      "grad_norm": 0.6660788059234619,
+      "learning_rate": 0.00010873479578599361,
+      "loss": 2.5922,
+      "step": 9496
+    },
+    {
+      "epoch": 0.7664433863287871,
+      "grad_norm": 0.7506482005119324,
+      "learning_rate": 0.00010871906911914502,
+      "loss": 2.5383,
+      "step": 9497
+    },
+    {
+      "epoch": 0.76652409006537,
+      "grad_norm": 0.7514285445213318,
+      "learning_rate": 0.00010870334223499443,
+      "loss": 2.5551,
+      "step": 9498
+    },
+    {
+      "epoch": 0.766604793801953,
+      "grad_norm": 0.6461809873580933,
+      "learning_rate": 0.00010868761513393379,
+      "loss": 2.5367,
+      "step": 9499
+    },
+    {
+      "epoch": 0.766685497538536,
+      "grad_norm": 0.6328238844871521,
+      "learning_rate": 0.00010867188781635512,
+      "loss": 2.5505,
+      "step": 9500
+    },
+    {
+      "epoch": 0.7667662012751191,
+      "grad_norm": 0.7090224027633667,
+      "learning_rate": 0.00010865616028265027,
+      "loss": 2.5921,
+      "step": 9501
+    },
+    {
+      "epoch": 0.766846905011702,
+      "grad_norm": 0.6404605507850647,
+      "learning_rate": 0.0001086404325332113,
+      "loss": 2.5357,
+      "step": 9502
+    },
+    {
+      "epoch": 0.766927608748285,
+      "grad_norm": 0.652477502822876,
+      "learning_rate": 0.00010862470456843016,
+      "loss": 2.5277,
+      "step": 9503
+    },
+    {
+      "epoch": 0.767008312484868,
+      "grad_norm": 0.7045448422431946,
+      "learning_rate": 0.00010860897638869887,
+      "loss": 2.5712,
+      "step": 9504
+    },
+    {
+      "epoch": 0.7670890162214511,
+      "grad_norm": 0.7024295926094055,
+      "learning_rate": 0.00010859324799440936,
+      "loss": 2.5976,
+      "step": 9505
+    },
+    {
+      "epoch": 0.7671697199580341,
+      "grad_norm": 0.7165585160255432,
+      "learning_rate": 0.00010857751938595364,
+      "loss": 2.5378,
+      "step": 9506
+    },
+    {
+      "epoch": 0.767250423694617,
+      "grad_norm": 0.7037522196769714,
+      "learning_rate": 0.0001085617905637237,
+      "loss": 2.554,
+      "step": 9507
+    },
+    {
+      "epoch": 0.7673311274312,
+      "grad_norm": 0.738210916519165,
+      "learning_rate": 0.00010854606152811163,
+      "loss": 2.5102,
+      "step": 9508
+    },
+    {
+      "epoch": 0.7674118311677831,
+      "grad_norm": 0.7500020861625671,
+      "learning_rate": 0.0001085303322795093,
+      "loss": 2.5908,
+      "step": 9509
+    },
+    {
+      "epoch": 0.7674925349043661,
+      "grad_norm": 0.7669610977172852,
+      "learning_rate": 0.00010851460281830883,
+      "loss": 2.5119,
+      "step": 9510
+    },
+    {
+      "epoch": 0.7675732386409491,
+      "grad_norm": 0.6619212031364441,
+      "learning_rate": 0.00010849887314490217,
+      "loss": 2.5622,
+      "step": 9511
+    },
+    {
+      "epoch": 0.767653942377532,
+      "grad_norm": 0.7142546772956848,
+      "learning_rate": 0.00010848314325968136,
+      "loss": 2.596,
+      "step": 9512
+    },
+    {
+      "epoch": 0.7677346461141151,
+      "grad_norm": 0.7365403175354004,
+      "learning_rate": 0.0001084674131630385,
+      "loss": 2.5695,
+      "step": 9513
+    },
+    {
+      "epoch": 0.7678153498506981,
+      "grad_norm": 0.7843711972236633,
+      "learning_rate": 0.00010845168285536555,
+      "loss": 2.5707,
+      "step": 9514
+    },
+    {
+      "epoch": 0.7678960535872811,
+      "grad_norm": 0.6391385197639465,
+      "learning_rate": 0.00010843595233705454,
+      "loss": 2.5523,
+      "step": 9515
+    },
+    {
+      "epoch": 0.7679767573238641,
+      "grad_norm": 0.6955631971359253,
+      "learning_rate": 0.00010842022160849758,
+      "loss": 2.5072,
+      "step": 9516
+    },
+    {
+      "epoch": 0.7680574610604471,
+      "grad_norm": 0.7291388511657715,
+      "learning_rate": 0.00010840449067008665,
+      "loss": 2.5786,
+      "step": 9517
+    },
+    {
+      "epoch": 0.7681381647970301,
+      "grad_norm": 0.7988889813423157,
+      "learning_rate": 0.00010838875952221387,
+      "loss": 2.5622,
+      "step": 9518
+    },
+    {
+      "epoch": 0.7682188685336131,
+      "grad_norm": 0.726271390914917,
+      "learning_rate": 0.00010837302816527129,
+      "loss": 2.5479,
+      "step": 9519
+    },
+    {
+      "epoch": 0.7682995722701961,
+      "grad_norm": 0.7305205464363098,
+      "learning_rate": 0.00010835729659965095,
+      "loss": 2.5946,
+      "step": 9520
+    },
+    {
+      "epoch": 0.7683802760067792,
+      "grad_norm": 0.7843366265296936,
+      "learning_rate": 0.00010834156482574493,
+      "loss": 2.5212,
+      "step": 9521
+    },
+    {
+      "epoch": 0.7684609797433621,
+      "grad_norm": 0.6988845467567444,
+      "learning_rate": 0.00010832583284394529,
+      "loss": 2.5174,
+      "step": 9522
+    },
+    {
+      "epoch": 0.7685416834799451,
+      "grad_norm": 0.7088077068328857,
+      "learning_rate": 0.00010831010065464414,
+      "loss": 2.5253,
+      "step": 9523
+    },
+    {
+      "epoch": 0.7686223872165281,
+      "grad_norm": 0.7447031140327454,
+      "learning_rate": 0.00010829436825823358,
+      "loss": 2.6045,
+      "step": 9524
+    },
+    {
+      "epoch": 0.7687030909531112,
+      "grad_norm": 0.6865237951278687,
+      "learning_rate": 0.00010827863565510566,
+      "loss": 2.558,
+      "step": 9525
+    },
+    {
+      "epoch": 0.7687837946896942,
+      "grad_norm": 0.7748900651931763,
+      "learning_rate": 0.0001082629028456525,
+      "loss": 2.5694,
+      "step": 9526
+    },
+    {
+      "epoch": 0.7688644984262771,
+      "grad_norm": 0.7031759023666382,
+      "learning_rate": 0.00010824716983026622,
+      "loss": 2.5171,
+      "step": 9527
+    },
+    {
+      "epoch": 0.7689452021628601,
+      "grad_norm": 0.7627702355384827,
+      "learning_rate": 0.00010823143660933888,
+      "loss": 2.5715,
+      "step": 9528
+    },
+    {
+      "epoch": 0.7690259058994432,
+      "grad_norm": 0.707815945148468,
+      "learning_rate": 0.00010821570318326264,
+      "loss": 2.5281,
+      "step": 9529
+    },
+    {
+      "epoch": 0.7691066096360262,
+      "grad_norm": 0.6833841800689697,
+      "learning_rate": 0.00010819996955242962,
+      "loss": 2.5702,
+      "step": 9530
+    },
+    {
+      "epoch": 0.7691873133726091,
+      "grad_norm": 0.7029415369033813,
+      "learning_rate": 0.00010818423571723189,
+      "loss": 2.5331,
+      "step": 9531
+    },
+    {
+      "epoch": 0.7692680171091921,
+      "grad_norm": 0.6442921161651611,
+      "learning_rate": 0.00010816850167806161,
+      "loss": 2.5423,
+      "step": 9532
+    },
+    {
+      "epoch": 0.7693487208457752,
+      "grad_norm": 0.7259004712104797,
+      "learning_rate": 0.00010815276743531093,
+      "loss": 2.6014,
+      "step": 9533
+    },
+    {
+      "epoch": 0.7694294245823582,
+      "grad_norm": 0.6483473777770996,
+      "learning_rate": 0.00010813703298937199,
+      "loss": 2.5268,
+      "step": 9534
+    },
+    {
+      "epoch": 0.7695101283189412,
+      "grad_norm": 0.6805520057678223,
+      "learning_rate": 0.00010812129834063691,
+      "loss": 2.5536,
+      "step": 9535
+    },
+    {
+      "epoch": 0.7695908320555241,
+      "grad_norm": 0.7120587825775146,
+      "learning_rate": 0.00010810556348949783,
+      "loss": 2.518,
+      "step": 9536
+    },
+    {
+      "epoch": 0.7696715357921071,
+      "grad_norm": 0.7280872464179993,
+      "learning_rate": 0.00010808982843634692,
+      "loss": 2.5525,
+      "step": 9537
+    },
+    {
+      "epoch": 0.7697522395286902,
+      "grad_norm": 0.68332439661026,
+      "learning_rate": 0.00010807409318157636,
+      "loss": 2.6318,
+      "step": 9538
+    },
+    {
+      "epoch": 0.7698329432652732,
+      "grad_norm": 0.655352771282196,
+      "learning_rate": 0.00010805835772557826,
+      "loss": 2.5781,
+      "step": 9539
+    },
+    {
+      "epoch": 0.7699136470018562,
+      "grad_norm": 0.7675400972366333,
+      "learning_rate": 0.00010804262206874484,
+      "loss": 2.5542,
+      "step": 9540
+    },
+    {
+      "epoch": 0.7699943507384391,
+      "grad_norm": 0.6676837205886841,
+      "learning_rate": 0.00010802688621146826,
+      "loss": 2.5411,
+      "step": 9541
+    },
+    {
+      "epoch": 0.7700750544750222,
+      "grad_norm": 0.7378436326980591,
+      "learning_rate": 0.00010801115015414067,
+      "loss": 2.5416,
+      "step": 9542
+    },
+    {
+      "epoch": 0.7701557582116052,
+      "grad_norm": 0.7330371141433716,
+      "learning_rate": 0.0001079954138971543,
+      "loss": 2.5154,
+      "step": 9543
+    },
+    {
+      "epoch": 0.7702364619481882,
+      "grad_norm": 0.6792974472045898,
+      "learning_rate": 0.00010797967744090131,
+      "loss": 2.5328,
+      "step": 9544
+    },
+    {
+      "epoch": 0.7703171656847712,
+      "grad_norm": 0.7129618525505066,
+      "learning_rate": 0.00010796394078577392,
+      "loss": 2.5688,
+      "step": 9545
+    },
+    {
+      "epoch": 0.7703978694213542,
+      "grad_norm": 0.6900608539581299,
+      "learning_rate": 0.00010794820393216429,
+      "loss": 2.5659,
+      "step": 9546
+    },
+    {
+      "epoch": 0.7704785731579372,
+      "grad_norm": 0.6798564195632935,
+      "learning_rate": 0.00010793246688046464,
+      "loss": 2.5746,
+      "step": 9547
+    },
+    {
+      "epoch": 0.7705592768945202,
+      "grad_norm": 0.7132395505905151,
+      "learning_rate": 0.00010791672963106715,
+      "loss": 2.6277,
+      "step": 9548
+    },
+    {
+      "epoch": 0.7706399806311032,
+      "grad_norm": 0.6762476563453674,
+      "learning_rate": 0.0001079009921843641,
+      "loss": 2.5265,
+      "step": 9549
+    },
+    {
+      "epoch": 0.7707206843676863,
+      "grad_norm": 0.7223351001739502,
+      "learning_rate": 0.00010788525454074765,
+      "loss": 2.6255,
+      "step": 9550
+    },
+    {
+      "epoch": 0.7708013881042692,
+      "grad_norm": 0.7383624315261841,
+      "learning_rate": 0.00010786951670061008,
+      "loss": 2.5744,
+      "step": 9551
+    },
+    {
+      "epoch": 0.7708820918408522,
+      "grad_norm": 0.6677328944206238,
+      "learning_rate": 0.00010785377866434355,
+      "loss": 2.5594,
+      "step": 9552
+    },
+    {
+      "epoch": 0.7709627955774352,
+      "grad_norm": 0.6572195887565613,
+      "learning_rate": 0.00010783804043234032,
+      "loss": 2.5582,
+      "step": 9553
+    },
+    {
+      "epoch": 0.7710434993140183,
+      "grad_norm": 0.6837800741195679,
+      "learning_rate": 0.00010782230200499265,
+      "loss": 2.5311,
+      "step": 9554
+    },
+    {
+      "epoch": 0.7711242030506013,
+      "grad_norm": 0.7232153415679932,
+      "learning_rate": 0.00010780656338269277,
+      "loss": 2.5074,
+      "step": 9555
+    },
+    {
+      "epoch": 0.7712049067871842,
+      "grad_norm": 0.6722296476364136,
+      "learning_rate": 0.00010779082456583291,
+      "loss": 2.551,
+      "step": 9556
+    },
+    {
+      "epoch": 0.7712856105237672,
+      "grad_norm": 0.6461100578308105,
+      "learning_rate": 0.00010777508555480535,
+      "loss": 2.5723,
+      "step": 9557
+    },
+    {
+      "epoch": 0.7713663142603503,
+      "grad_norm": 0.6573290824890137,
+      "learning_rate": 0.0001077593463500023,
+      "loss": 2.4967,
+      "step": 9558
+    },
+    {
+      "epoch": 0.7714470179969333,
+      "grad_norm": 0.7184738516807556,
+      "learning_rate": 0.0001077436069518161,
+      "loss": 2.6703,
+      "step": 9559
+    },
+    {
+      "epoch": 0.7715277217335162,
+      "grad_norm": 0.7226557731628418,
+      "learning_rate": 0.00010772786736063895,
+      "loss": 2.6118,
+      "step": 9560
+    },
+    {
+      "epoch": 0.7716084254700992,
+      "grad_norm": 0.6800956130027771,
+      "learning_rate": 0.00010771212757686318,
+      "loss": 2.578,
+      "step": 9561
+    },
+    {
+      "epoch": 0.7716891292066823,
+      "grad_norm": 0.6657535433769226,
+      "learning_rate": 0.00010769638760088099,
+      "loss": 2.5291,
+      "step": 9562
+    },
+    {
+      "epoch": 0.7717698329432653,
+      "grad_norm": 0.620527982711792,
+      "learning_rate": 0.00010768064743308471,
+      "loss": 2.5518,
+      "step": 9563
+    },
+    {
+      "epoch": 0.7718505366798483,
+      "grad_norm": 0.693760097026825,
+      "learning_rate": 0.00010766490707386663,
+      "loss": 2.52,
+      "step": 9564
+    },
+    {
+      "epoch": 0.7719312404164312,
+      "grad_norm": 0.6674148440361023,
+      "learning_rate": 0.000107649166523619,
+      "loss": 2.5197,
+      "step": 9565
+    },
+    {
+      "epoch": 0.7720119441530143,
+      "grad_norm": 0.6844033598899841,
+      "learning_rate": 0.00010763342578273419,
+      "loss": 2.5842,
+      "step": 9566
+    },
+    {
+      "epoch": 0.7720926478895973,
+      "grad_norm": 0.6891880035400391,
+      "learning_rate": 0.00010761768485160442,
+      "loss": 2.5349,
+      "step": 9567
+    },
+    {
+      "epoch": 0.7721733516261803,
+      "grad_norm": 0.7157394289970398,
+      "learning_rate": 0.00010760194373062204,
+      "loss": 2.5762,
+      "step": 9568
+    },
+    {
+      "epoch": 0.7722540553627633,
+      "grad_norm": 0.7522526383399963,
+      "learning_rate": 0.00010758620242017936,
+      "loss": 2.5348,
+      "step": 9569
+    },
+    {
+      "epoch": 0.7723347590993463,
+      "grad_norm": 0.6817746162414551,
+      "learning_rate": 0.00010757046092066869,
+      "loss": 2.5836,
+      "step": 9570
+    },
+    {
+      "epoch": 0.7724154628359293,
+      "grad_norm": 0.7274518013000488,
+      "learning_rate": 0.00010755471923248232,
+      "loss": 2.5276,
+      "step": 9571
+    },
+    {
+      "epoch": 0.7724961665725123,
+      "grad_norm": 0.6735557913780212,
+      "learning_rate": 0.00010753897735601264,
+      "loss": 2.6116,
+      "step": 9572
+    },
+    {
+      "epoch": 0.7725768703090953,
+      "grad_norm": 0.6626406908035278,
+      "learning_rate": 0.00010752323529165186,
+      "loss": 2.5778,
+      "step": 9573
+    },
+    {
+      "epoch": 0.7726575740456784,
+      "grad_norm": 0.6627367734909058,
+      "learning_rate": 0.00010750749303979246,
+      "loss": 2.5839,
+      "step": 9574
+    },
+    {
+      "epoch": 0.7727382777822613,
+      "grad_norm": 0.6658251881599426,
+      "learning_rate": 0.0001074917506008267,
+      "loss": 2.5233,
+      "step": 9575
+    },
+    {
+      "epoch": 0.7728189815188443,
+      "grad_norm": 0.6969848871231079,
+      "learning_rate": 0.00010747600797514692,
+      "loss": 2.5169,
+      "step": 9576
+    },
+    {
+      "epoch": 0.7728996852554273,
+      "grad_norm": 0.7313554883003235,
+      "learning_rate": 0.00010746026516314549,
+      "loss": 2.5528,
+      "step": 9577
+    },
+    {
+      "epoch": 0.7729803889920104,
+      "grad_norm": 0.6467077136039734,
+      "learning_rate": 0.00010744452216521472,
+      "loss": 2.5158,
+      "step": 9578
+    },
+    {
+      "epoch": 0.7730610927285934,
+      "grad_norm": 0.6808056235313416,
+      "learning_rate": 0.00010742877898174702,
+      "loss": 2.5346,
+      "step": 9579
+    },
+    {
+      "epoch": 0.7731417964651763,
+      "grad_norm": 0.7537400722503662,
+      "learning_rate": 0.00010741303561313474,
+      "loss": 2.5621,
+      "step": 9580
+    },
+    {
+      "epoch": 0.7732225002017593,
+      "grad_norm": 0.6715610027313232,
+      "learning_rate": 0.00010739729205977021,
+      "loss": 2.5384,
+      "step": 9581
+    },
+    {
+      "epoch": 0.7733032039383424,
+      "grad_norm": 0.7129234075546265,
+      "learning_rate": 0.00010738154832204586,
+      "loss": 2.5639,
+      "step": 9582
+    },
+    {
+      "epoch": 0.7733839076749254,
+      "grad_norm": 0.7156025171279907,
+      "learning_rate": 0.00010736580440035397,
+      "loss": 2.5427,
+      "step": 9583
+    },
+    {
+      "epoch": 0.7734646114115084,
+      "grad_norm": 0.7394191026687622,
+      "learning_rate": 0.00010735006029508703,
+      "loss": 2.5809,
+      "step": 9584
+    },
+    {
+      "epoch": 0.7735453151480913,
+      "grad_norm": 0.7117684483528137,
+      "learning_rate": 0.00010733431600663737,
+      "loss": 2.5807,
+      "step": 9585
+    },
+    {
+      "epoch": 0.7736260188846744,
+      "grad_norm": 0.6622862219810486,
+      "learning_rate": 0.00010731857153539737,
+      "loss": 2.5277,
+      "step": 9586
+    },
+    {
+      "epoch": 0.7737067226212574,
+      "grad_norm": 0.7744547128677368,
+      "learning_rate": 0.00010730282688175943,
+      "loss": 2.6119,
+      "step": 9587
+    },
+    {
+      "epoch": 0.7737874263578404,
+      "grad_norm": 0.6804926991462708,
+      "learning_rate": 0.00010728708204611597,
+      "loss": 2.534,
+      "step": 9588
+    },
+    {
+      "epoch": 0.7738681300944233,
+      "grad_norm": 0.7115367650985718,
+      "learning_rate": 0.00010727133702885937,
+      "loss": 2.542,
+      "step": 9589
+    },
+    {
+      "epoch": 0.7739488338310063,
+      "grad_norm": 0.7623847723007202,
+      "learning_rate": 0.00010725559183038205,
+      "loss": 2.587,
+      "step": 9590
+    },
+    {
+      "epoch": 0.7740295375675894,
+      "grad_norm": 0.6612982153892517,
+      "learning_rate": 0.00010723984645107641,
+      "loss": 2.5257,
+      "step": 9591
+    },
+    {
+      "epoch": 0.7741102413041724,
+      "grad_norm": 0.7553900480270386,
+      "learning_rate": 0.00010722410089133488,
+      "loss": 2.6311,
+      "step": 9592
+    },
+    {
+      "epoch": 0.7741909450407554,
+      "grad_norm": 0.7541414499282837,
+      "learning_rate": 0.00010720835515154983,
+      "loss": 2.5978,
+      "step": 9593
+    },
+    {
+      "epoch": 0.7742716487773383,
+      "grad_norm": 0.6690947413444519,
+      "learning_rate": 0.00010719260923211376,
+      "loss": 2.568,
+      "step": 9594
+    },
+    {
+      "epoch": 0.7743523525139214,
+      "grad_norm": 0.7282151579856873,
+      "learning_rate": 0.00010717686313341909,
+      "loss": 2.5375,
+      "step": 9595
+    },
+    {
+      "epoch": 0.7744330562505044,
+      "grad_norm": 0.6862902045249939,
+      "learning_rate": 0.00010716111685585821,
+      "loss": 2.5503,
+      "step": 9596
+    },
+    {
+      "epoch": 0.7745137599870874,
+      "grad_norm": 0.7076265811920166,
+      "learning_rate": 0.00010714537039982357,
+      "loss": 2.4766,
+      "step": 9597
+    },
+    {
+      "epoch": 0.7745944637236704,
+      "grad_norm": 0.7063891887664795,
+      "learning_rate": 0.00010712962376570761,
+      "loss": 2.5822,
+      "step": 9598
+    },
+    {
+      "epoch": 0.7746751674602534,
+      "grad_norm": 0.6975609064102173,
+      "learning_rate": 0.00010711387695390282,
+      "loss": 2.597,
+      "step": 9599
+    },
+    {
+      "epoch": 0.7747558711968364,
+      "grad_norm": 0.6790002584457397,
+      "learning_rate": 0.0001070981299648016,
+      "loss": 2.5705,
+      "step": 9600
+    },
+    {
+      "epoch": 0.7748365749334194,
+      "grad_norm": 0.6493679881095886,
+      "learning_rate": 0.00010708238279879643,
+      "loss": 2.49,
+      "step": 9601
+    },
+    {
+      "epoch": 0.7749172786700024,
+      "grad_norm": 0.6741142868995667,
+      "learning_rate": 0.00010706663545627977,
+      "loss": 2.6008,
+      "step": 9602
+    },
+    {
+      "epoch": 0.7749979824065855,
+      "grad_norm": 0.6753309965133667,
+      "learning_rate": 0.00010705088793764408,
+      "loss": 2.536,
+      "step": 9603
+    },
+    {
+      "epoch": 0.7750786861431684,
+      "grad_norm": 0.6879377365112305,
+      "learning_rate": 0.00010703514024328183,
+      "loss": 2.5884,
+      "step": 9604
+    },
+    {
+      "epoch": 0.7751593898797514,
+      "grad_norm": 0.6535949110984802,
+      "learning_rate": 0.00010701939237358549,
+      "loss": 2.5489,
+      "step": 9605
+    },
+    {
+      "epoch": 0.7752400936163344,
+      "grad_norm": 0.7308230400085449,
+      "learning_rate": 0.00010700364432894756,
+      "loss": 2.5679,
+      "step": 9606
+    },
+    {
+      "epoch": 0.7753207973529175,
+      "grad_norm": 0.7016584277153015,
+      "learning_rate": 0.00010698789610976052,
+      "loss": 2.5678,
+      "step": 9607
+    },
+    {
+      "epoch": 0.7754015010895005,
+      "grad_norm": 0.7181541323661804,
+      "learning_rate": 0.00010697214771641682,
+      "loss": 2.5004,
+      "step": 9608
+    },
+    {
+      "epoch": 0.7754822048260834,
+      "grad_norm": 0.6414844989776611,
+      "learning_rate": 0.00010695639914930895,
+      "loss": 2.4896,
+      "step": 9609
+    },
+    {
+      "epoch": 0.7755629085626664,
+      "grad_norm": 0.7288017868995667,
+      "learning_rate": 0.00010694065040882943,
+      "loss": 2.5945,
+      "step": 9610
+    },
+    {
+      "epoch": 0.7756436122992495,
+      "grad_norm": 0.6808066368103027,
+      "learning_rate": 0.00010692490149537079,
+      "loss": 2.5973,
+      "step": 9611
+    },
+    {
+      "epoch": 0.7757243160358325,
+      "grad_norm": 0.7924454212188721,
+      "learning_rate": 0.00010690915240932553,
+      "loss": 2.5448,
+      "step": 9612
+    },
+    {
+      "epoch": 0.7758050197724154,
+      "grad_norm": 0.6466094851493835,
+      "learning_rate": 0.00010689340315108606,
+      "loss": 2.5065,
+      "step": 9613
+    },
+    {
+      "epoch": 0.7758857235089984,
+      "grad_norm": 0.6775460243225098,
+      "learning_rate": 0.00010687765372104502,
+      "loss": 2.5238,
+      "step": 9614
+    },
+    {
+      "epoch": 0.7759664272455815,
+      "grad_norm": 0.6901230812072754,
+      "learning_rate": 0.00010686190411959484,
+      "loss": 2.5109,
+      "step": 9615
+    },
+    {
+      "epoch": 0.7760471309821645,
+      "grad_norm": 0.7032039165496826,
+      "learning_rate": 0.00010684615434712808,
+      "loss": 2.6094,
+      "step": 9616
+    },
+    {
+      "epoch": 0.7761278347187475,
+      "grad_norm": 0.7008969187736511,
+      "learning_rate": 0.00010683040440403727,
+      "loss": 2.5758,
+      "step": 9617
+    },
+    {
+      "epoch": 0.7762085384553304,
+      "grad_norm": 0.6909677386283875,
+      "learning_rate": 0.00010681465429071491,
+      "loss": 2.5373,
+      "step": 9618
+    },
+    {
+      "epoch": 0.7762892421919135,
+      "grad_norm": 0.699030339717865,
+      "learning_rate": 0.00010679890400755355,
+      "loss": 2.577,
+      "step": 9619
+    },
+    {
+      "epoch": 0.7763699459284965,
+      "grad_norm": 0.7012344598770142,
+      "learning_rate": 0.00010678315355494575,
+      "loss": 2.5205,
+      "step": 9620
+    },
+    {
+      "epoch": 0.7764506496650795,
+      "grad_norm": 0.7693915367126465,
+      "learning_rate": 0.000106767402933284,
+      "loss": 2.5947,
+      "step": 9621
+    },
+    {
+      "epoch": 0.7765313534016625,
+      "grad_norm": 0.7635772228240967,
+      "learning_rate": 0.00010675165214296093,
+      "loss": 2.6221,
+      "step": 9622
+    },
+    {
+      "epoch": 0.7766120571382455,
+      "grad_norm": 0.701411783695221,
+      "learning_rate": 0.000106735901184369,
+      "loss": 2.5236,
+      "step": 9623
+    },
+    {
+      "epoch": 0.7766927608748285,
+      "grad_norm": 0.7283998727798462,
+      "learning_rate": 0.00010672015005790079,
+      "loss": 2.5581,
+      "step": 9624
+    },
+    {
+      "epoch": 0.7767734646114115,
+      "grad_norm": 0.7069897055625916,
+      "learning_rate": 0.0001067043987639489,
+      "loss": 2.5541,
+      "step": 9625
+    },
+    {
+      "epoch": 0.7768541683479945,
+      "grad_norm": 0.7419753074645996,
+      "learning_rate": 0.00010668864730290586,
+      "loss": 2.5992,
+      "step": 9626
+    },
+    {
+      "epoch": 0.7769348720845776,
+      "grad_norm": 0.6651501059532166,
+      "learning_rate": 0.00010667289567516426,
+      "loss": 2.546,
+      "step": 9627
+    },
+    {
+      "epoch": 0.7770155758211605,
+      "grad_norm": 0.7265670895576477,
+      "learning_rate": 0.00010665714388111665,
+      "loss": 2.611,
+      "step": 9628
+    },
+    {
+      "epoch": 0.7770962795577435,
+      "grad_norm": 0.6520028114318848,
+      "learning_rate": 0.00010664139192115559,
+      "loss": 2.5433,
+      "step": 9629
+    },
+    {
+      "epoch": 0.7771769832943265,
+      "grad_norm": 0.6990057826042175,
+      "learning_rate": 0.0001066256397956737,
+      "loss": 2.5325,
+      "step": 9630
+    },
+    {
+      "epoch": 0.7772576870309096,
+      "grad_norm": 0.7353312373161316,
+      "learning_rate": 0.00010660988750506355,
+      "loss": 2.4707,
+      "step": 9631
+    },
+    {
+      "epoch": 0.7773383907674926,
+      "grad_norm": 0.6810272932052612,
+      "learning_rate": 0.00010659413504971774,
+      "loss": 2.5618,
+      "step": 9632
+    },
+    {
+      "epoch": 0.7774190945040755,
+      "grad_norm": 0.6480081081390381,
+      "learning_rate": 0.00010657838243002883,
+      "loss": 2.4543,
+      "step": 9633
+    },
+    {
+      "epoch": 0.7774997982406585,
+      "grad_norm": 0.6617380976676941,
+      "learning_rate": 0.00010656262964638942,
+      "loss": 2.5628,
+      "step": 9634
+    },
+    {
+      "epoch": 0.7775805019772416,
+      "grad_norm": 0.6761382222175598,
+      "learning_rate": 0.00010654687669919212,
+      "loss": 2.5433,
+      "step": 9635
+    },
+    {
+      "epoch": 0.7776612057138246,
+      "grad_norm": 0.6733867526054382,
+      "learning_rate": 0.00010653112358882957,
+      "loss": 2.5282,
+      "step": 9636
+    },
+    {
+      "epoch": 0.7777419094504076,
+      "grad_norm": 0.6854631304740906,
+      "learning_rate": 0.00010651537031569433,
+      "loss": 2.5997,
+      "step": 9637
+    },
+    {
+      "epoch": 0.7778226131869905,
+      "grad_norm": 0.7451226115226746,
+      "learning_rate": 0.00010649961688017904,
+      "loss": 2.5058,
+      "step": 9638
+    },
+    {
+      "epoch": 0.7779033169235735,
+      "grad_norm": 0.6744229197502136,
+      "learning_rate": 0.0001064838632826763,
+      "loss": 2.5962,
+      "step": 9639
+    },
+    {
+      "epoch": 0.7779840206601566,
+      "grad_norm": 0.7568119764328003,
+      "learning_rate": 0.00010646810952357873,
+      "loss": 2.5896,
+      "step": 9640
+    },
+    {
+      "epoch": 0.7780647243967396,
+      "grad_norm": 0.6860085725784302,
+      "learning_rate": 0.00010645235560327899,
+      "loss": 2.5675,
+      "step": 9641
+    },
+    {
+      "epoch": 0.7781454281333225,
+      "grad_norm": 0.6491742134094238,
+      "learning_rate": 0.00010643660152216965,
+      "loss": 2.5374,
+      "step": 9642
+    },
+    {
+      "epoch": 0.7782261318699055,
+      "grad_norm": 0.6664023399353027,
+      "learning_rate": 0.0001064208472806434,
+      "loss": 2.4679,
+      "step": 9643
+    },
+    {
+      "epoch": 0.7783068356064886,
+      "grad_norm": 0.6595140099525452,
+      "learning_rate": 0.00010640509287909284,
+      "loss": 2.5045,
+      "step": 9644
+    },
+    {
+      "epoch": 0.7783875393430716,
+      "grad_norm": 0.6788576245307922,
+      "learning_rate": 0.0001063893383179106,
+      "loss": 2.5706,
+      "step": 9645
+    },
+    {
+      "epoch": 0.7784682430796546,
+      "grad_norm": 0.6741334199905396,
+      "learning_rate": 0.00010637358359748939,
+      "loss": 2.5763,
+      "step": 9646
+    },
+    {
+      "epoch": 0.7785489468162375,
+      "grad_norm": 0.6837517023086548,
+      "learning_rate": 0.0001063578287182218,
+      "loss": 2.5484,
+      "step": 9647
+    },
+    {
+      "epoch": 0.7786296505528206,
+      "grad_norm": 0.6604229211807251,
+      "learning_rate": 0.00010634207368050048,
+      "loss": 2.5465,
+      "step": 9648
+    },
+    {
+      "epoch": 0.7787103542894036,
+      "grad_norm": 0.6528951525688171,
+      "learning_rate": 0.00010632631848471813,
+      "loss": 2.5409,
+      "step": 9649
+    },
+    {
+      "epoch": 0.7787910580259866,
+      "grad_norm": 0.6615377068519592,
+      "learning_rate": 0.00010631056313126734,
+      "loss": 2.5545,
+      "step": 9650
+    },
+    {
+      "epoch": 0.7788717617625696,
+      "grad_norm": 0.666033923625946,
+      "learning_rate": 0.00010629480762054089,
+      "loss": 2.5341,
+      "step": 9651
+    },
+    {
+      "epoch": 0.7789524654991526,
+      "grad_norm": 0.7022622227668762,
+      "learning_rate": 0.00010627905195293135,
+      "loss": 2.5206,
+      "step": 9652
+    },
+    {
+      "epoch": 0.7790331692357356,
+      "grad_norm": 0.7175850868225098,
+      "learning_rate": 0.00010626329612883141,
+      "loss": 2.5912,
+      "step": 9653
+    },
+    {
+      "epoch": 0.7791138729723186,
+      "grad_norm": 0.6592069268226624,
+      "learning_rate": 0.00010624754014863379,
+      "loss": 2.5076,
+      "step": 9654
+    },
+    {
+      "epoch": 0.7791945767089016,
+      "grad_norm": 0.645893931388855,
+      "learning_rate": 0.0001062317840127311,
+      "loss": 2.5124,
+      "step": 9655
+    },
+    {
+      "epoch": 0.7792752804454847,
+      "grad_norm": 0.6638232469558716,
+      "learning_rate": 0.00010621602772151607,
+      "loss": 2.5182,
+      "step": 9656
+    },
+    {
+      "epoch": 0.7793559841820676,
+      "grad_norm": 0.6718387603759766,
+      "learning_rate": 0.0001062002712753814,
+      "loss": 2.4773,
+      "step": 9657
+    },
+    {
+      "epoch": 0.7794366879186506,
+      "grad_norm": 0.6402876377105713,
+      "learning_rate": 0.00010618451467471972,
+      "loss": 2.5557,
+      "step": 9658
+    },
+    {
+      "epoch": 0.7795173916552336,
+      "grad_norm": 0.6898398399353027,
+      "learning_rate": 0.00010616875791992382,
+      "loss": 2.5557,
+      "step": 9659
+    },
+    {
+      "epoch": 0.7795980953918167,
+      "grad_norm": 0.6718475222587585,
+      "learning_rate": 0.00010615300101138633,
+      "loss": 2.5335,
+      "step": 9660
+    },
+    {
+      "epoch": 0.7796787991283997,
+      "grad_norm": 0.6436911225318909,
+      "learning_rate": 0.00010613724394949995,
+      "loss": 2.5214,
+      "step": 9661
+    },
+    {
+      "epoch": 0.7797595028649826,
+      "grad_norm": 0.7554156184196472,
+      "learning_rate": 0.00010612148673465743,
+      "loss": 2.5526,
+      "step": 9662
+    },
+    {
+      "epoch": 0.7798402066015656,
+      "grad_norm": 0.6728504300117493,
+      "learning_rate": 0.00010610572936725147,
+      "loss": 2.5935,
+      "step": 9663
+    },
+    {
+      "epoch": 0.7799209103381487,
+      "grad_norm": 0.6793323159217834,
+      "learning_rate": 0.00010608997184767476,
+      "loss": 2.5515,
+      "step": 9664
+    },
+    {
+      "epoch": 0.7800016140747317,
+      "grad_norm": 0.7242898941040039,
+      "learning_rate": 0.00010607421417631999,
+      "loss": 2.5332,
+      "step": 9665
+    },
+    {
+      "epoch": 0.7800823178113147,
+      "grad_norm": 0.6719244718551636,
+      "learning_rate": 0.00010605845635357996,
+      "loss": 2.5191,
+      "step": 9666
+    },
+    {
+      "epoch": 0.7801630215478976,
+      "grad_norm": 0.6836631894111633,
+      "learning_rate": 0.00010604269837984737,
+      "loss": 2.6489,
+      "step": 9667
+    },
+    {
+      "epoch": 0.7802437252844807,
+      "grad_norm": 0.6833824515342712,
+      "learning_rate": 0.00010602694025551496,
+      "loss": 2.4906,
+      "step": 9668
+    },
+    {
+      "epoch": 0.7803244290210637,
+      "grad_norm": 0.7449159026145935,
+      "learning_rate": 0.0001060111819809754,
+      "loss": 2.5301,
+      "step": 9669
+    },
+    {
+      "epoch": 0.7804051327576467,
+      "grad_norm": 0.7149158120155334,
+      "learning_rate": 0.00010599542355662149,
+      "loss": 2.5097,
+      "step": 9670
+    },
+    {
+      "epoch": 0.7804858364942296,
+      "grad_norm": 0.6616973876953125,
+      "learning_rate": 0.00010597966498284595,
+      "loss": 2.5928,
+      "step": 9671
+    },
+    {
+      "epoch": 0.7805665402308127,
+      "grad_norm": 0.6556531190872192,
+      "learning_rate": 0.00010596390626004154,
+      "loss": 2.5543,
+      "step": 9672
+    },
+    {
+      "epoch": 0.7806472439673957,
+      "grad_norm": 0.6585283875465393,
+      "learning_rate": 0.000105948147388601,
+      "loss": 2.5244,
+      "step": 9673
+    },
+    {
+      "epoch": 0.7807279477039787,
+      "grad_norm": 0.6484133005142212,
+      "learning_rate": 0.00010593238836891704,
+      "loss": 2.4996,
+      "step": 9674
+    },
+    {
+      "epoch": 0.7808086514405617,
+      "grad_norm": 0.6681119799613953,
+      "learning_rate": 0.00010591662920138248,
+      "loss": 2.5322,
+      "step": 9675
+    },
+    {
+      "epoch": 0.7808893551771448,
+      "grad_norm": 0.709403395652771,
+      "learning_rate": 0.00010590086988639005,
+      "loss": 2.5554,
+      "step": 9676
+    },
+    {
+      "epoch": 0.7809700589137277,
+      "grad_norm": 0.6734669804573059,
+      "learning_rate": 0.00010588511042433251,
+      "loss": 2.5452,
+      "step": 9677
+    },
+    {
+      "epoch": 0.7810507626503107,
+      "grad_norm": 0.6800141930580139,
+      "learning_rate": 0.00010586935081560268,
+      "loss": 2.5154,
+      "step": 9678
+    },
+    {
+      "epoch": 0.7811314663868937,
+      "grad_norm": 0.7757244110107422,
+      "learning_rate": 0.00010585359106059326,
+      "loss": 2.5935,
+      "step": 9679
+    },
+    {
+      "epoch": 0.7812121701234768,
+      "grad_norm": 0.7288491725921631,
+      "learning_rate": 0.00010583783115969699,
+      "loss": 2.5276,
+      "step": 9680
+    },
+    {
+      "epoch": 0.7812928738600597,
+      "grad_norm": 0.6785164475440979,
+      "learning_rate": 0.00010582207111330678,
+      "loss": 2.5907,
+      "step": 9681
+    },
+    {
+      "epoch": 0.7813735775966427,
+      "grad_norm": 0.6651367545127869,
+      "learning_rate": 0.0001058063109218153,
+      "loss": 2.545,
+      "step": 9682
+    },
+    {
+      "epoch": 0.7814542813332257,
+      "grad_norm": 0.6657043695449829,
+      "learning_rate": 0.0001057905505856154,
+      "loss": 2.5548,
+      "step": 9683
+    },
+    {
+      "epoch": 0.7815349850698088,
+      "grad_norm": 0.6486692428588867,
+      "learning_rate": 0.00010577479010509986,
+      "loss": 2.5589,
+      "step": 9684
+    },
+    {
+      "epoch": 0.7816156888063918,
+      "grad_norm": 0.700749397277832,
+      "learning_rate": 0.0001057590294806614,
+      "loss": 2.6008,
+      "step": 9685
+    },
+    {
+      "epoch": 0.7816963925429747,
+      "grad_norm": 0.647051215171814,
+      "learning_rate": 0.00010574326871269289,
+      "loss": 2.4894,
+      "step": 9686
+    },
+    {
+      "epoch": 0.7817770962795577,
+      "grad_norm": 0.6932066679000854,
+      "learning_rate": 0.00010572750780158713,
+      "loss": 2.5256,
+      "step": 9687
+    },
+    {
+      "epoch": 0.7818578000161408,
+      "grad_norm": 0.6330733895301819,
+      "learning_rate": 0.00010571174674773689,
+      "loss": 2.5242,
+      "step": 9688
+    },
+    {
+      "epoch": 0.7819385037527238,
+      "grad_norm": 0.6476379036903381,
+      "learning_rate": 0.00010569598555153499,
+      "loss": 2.552,
+      "step": 9689
+    },
+    {
+      "epoch": 0.7820192074893068,
+      "grad_norm": 0.661204993724823,
+      "learning_rate": 0.00010568022421337424,
+      "loss": 2.4869,
+      "step": 9690
+    },
+    {
+      "epoch": 0.7820999112258897,
+      "grad_norm": 0.6663263440132141,
+      "learning_rate": 0.00010566446273364746,
+      "loss": 2.5134,
+      "step": 9691
+    },
+    {
+      "epoch": 0.7821806149624727,
+      "grad_norm": 0.6982834339141846,
+      "learning_rate": 0.00010564870111274748,
+      "loss": 2.5755,
+      "step": 9692
+    },
+    {
+      "epoch": 0.7822613186990558,
+      "grad_norm": 0.6266167759895325,
+      "learning_rate": 0.00010563293935106706,
+      "loss": 2.5413,
+      "step": 9693
+    },
+    {
+      "epoch": 0.7823420224356388,
+      "grad_norm": 0.6484279632568359,
+      "learning_rate": 0.0001056171774489991,
+      "loss": 2.5579,
+      "step": 9694
+    },
+    {
+      "epoch": 0.7824227261722217,
+      "grad_norm": 0.674933910369873,
+      "learning_rate": 0.00010560141540693638,
+      "loss": 2.5364,
+      "step": 9695
+    },
+    {
+      "epoch": 0.7825034299088047,
+      "grad_norm": 0.7961840033531189,
+      "learning_rate": 0.00010558565322527174,
+      "loss": 2.5143,
+      "step": 9696
+    },
+    {
+      "epoch": 0.7825841336453878,
+      "grad_norm": 0.697158694267273,
+      "learning_rate": 0.00010556989090439804,
+      "loss": 2.5341,
+      "step": 9697
+    },
+    {
+      "epoch": 0.7826648373819708,
+      "grad_norm": 0.6912708282470703,
+      "learning_rate": 0.00010555412844470806,
+      "loss": 2.5331,
+      "step": 9698
+    },
+    {
+      "epoch": 0.7827455411185538,
+      "grad_norm": 0.7078350186347961,
+      "learning_rate": 0.00010553836584659474,
+      "loss": 2.5752,
+      "step": 9699
+    },
+    {
+      "epoch": 0.7828262448551367,
+      "grad_norm": 0.6421065926551819,
+      "learning_rate": 0.00010552260311045082,
+      "loss": 2.5393,
+      "step": 9700
+    },
+    {
+      "epoch": 0.7829069485917198,
+      "grad_norm": 0.644120454788208,
+      "learning_rate": 0.00010550684023666918,
+      "loss": 2.5062,
+      "step": 9701
+    },
+    {
+      "epoch": 0.7829876523283028,
+      "grad_norm": 0.7038589715957642,
+      "learning_rate": 0.00010549107722564275,
+      "loss": 2.6074,
+      "step": 9702
+    },
+    {
+      "epoch": 0.7830683560648858,
+      "grad_norm": 0.6692953109741211,
+      "learning_rate": 0.00010547531407776427,
+      "loss": 2.5801,
+      "step": 9703
+    },
+    {
+      "epoch": 0.7831490598014688,
+      "grad_norm": 0.7059200406074524,
+      "learning_rate": 0.00010545955079342669,
+      "loss": 2.5579,
+      "step": 9704
+    },
+    {
+      "epoch": 0.7832297635380518,
+      "grad_norm": 0.7126718759536743,
+      "learning_rate": 0.0001054437873730228,
+      "loss": 2.5764,
+      "step": 9705
+    },
+    {
+      "epoch": 0.7833104672746348,
+      "grad_norm": 0.696784257888794,
+      "learning_rate": 0.0001054280238169455,
+      "loss": 2.5256,
+      "step": 9706
+    },
+    {
+      "epoch": 0.7833911710112178,
+      "grad_norm": 0.7473082542419434,
+      "learning_rate": 0.00010541226012558767,
+      "loss": 2.5983,
+      "step": 9707
+    },
+    {
+      "epoch": 0.7834718747478008,
+      "grad_norm": 0.6598967909812927,
+      "learning_rate": 0.00010539649629934219,
+      "loss": 2.5267,
+      "step": 9708
+    },
+    {
+      "epoch": 0.7835525784843839,
+      "grad_norm": 0.7168934345245361,
+      "learning_rate": 0.00010538073233860188,
+      "loss": 2.5278,
+      "step": 9709
+    },
+    {
+      "epoch": 0.7836332822209668,
+      "grad_norm": 0.6848951578140259,
+      "learning_rate": 0.00010536496824375968,
+      "loss": 2.5267,
+      "step": 9710
+    },
+    {
+      "epoch": 0.7837139859575498,
+      "grad_norm": 0.7276272773742676,
+      "learning_rate": 0.0001053492040152084,
+      "loss": 2.5706,
+      "step": 9711
+    },
+    {
+      "epoch": 0.7837946896941328,
+      "grad_norm": 0.6929399371147156,
+      "learning_rate": 0.00010533343965334101,
+      "loss": 2.5184,
+      "step": 9712
+    },
+    {
+      "epoch": 0.7838753934307159,
+      "grad_norm": 0.7497181296348572,
+      "learning_rate": 0.00010531767515855037,
+      "loss": 2.5626,
+      "step": 9713
+    },
+    {
+      "epoch": 0.7839560971672989,
+      "grad_norm": 0.6536200046539307,
+      "learning_rate": 0.00010530191053122935,
+      "loss": 2.5909,
+      "step": 9714
+    },
+    {
+      "epoch": 0.7840368009038818,
+      "grad_norm": 0.6750395894050598,
+      "learning_rate": 0.00010528614577177087,
+      "loss": 2.5119,
+      "step": 9715
+    },
+    {
+      "epoch": 0.7841175046404648,
+      "grad_norm": 0.6284878849983215,
+      "learning_rate": 0.00010527038088056782,
+      "loss": 2.5417,
+      "step": 9716
+    },
+    {
+      "epoch": 0.7841982083770479,
+      "grad_norm": 0.6529444456100464,
+      "learning_rate": 0.00010525461585801308,
+      "loss": 2.5865,
+      "step": 9717
+    },
+    {
+      "epoch": 0.7842789121136309,
+      "grad_norm": 0.7332968711853027,
+      "learning_rate": 0.00010523885070449959,
+      "loss": 2.561,
+      "step": 9718
+    },
+    {
+      "epoch": 0.7843596158502139,
+      "grad_norm": 0.7054178714752197,
+      "learning_rate": 0.00010522308542042025,
+      "loss": 2.623,
+      "step": 9719
+    },
+    {
+      "epoch": 0.7844403195867968,
+      "grad_norm": 0.6837820410728455,
+      "learning_rate": 0.00010520732000616798,
+      "loss": 2.5586,
+      "step": 9720
+    },
+    {
+      "epoch": 0.7845210233233799,
+      "grad_norm": 0.7339439392089844,
+      "learning_rate": 0.00010519155446213565,
+      "loss": 2.5374,
+      "step": 9721
+    },
+    {
+      "epoch": 0.7846017270599629,
+      "grad_norm": 0.7625028491020203,
+      "learning_rate": 0.00010517578878871624,
+      "loss": 2.5663,
+      "step": 9722
+    },
+    {
+      "epoch": 0.7846824307965459,
+      "grad_norm": 0.6749752759933472,
+      "learning_rate": 0.00010516002298630263,
+      "loss": 2.5744,
+      "step": 9723
+    },
+    {
+      "epoch": 0.7847631345331288,
+      "grad_norm": 0.6702882647514343,
+      "learning_rate": 0.00010514425705528776,
+      "loss": 2.6247,
+      "step": 9724
+    },
+    {
+      "epoch": 0.7848438382697119,
+      "grad_norm": 0.6641737222671509,
+      "learning_rate": 0.00010512849099606457,
+      "loss": 2.5792,
+      "step": 9725
+    },
+    {
+      "epoch": 0.7849245420062949,
+      "grad_norm": 0.7522993683815002,
+      "learning_rate": 0.00010511272480902597,
+      "loss": 2.5941,
+      "step": 9726
+    },
+    {
+      "epoch": 0.7850052457428779,
+      "grad_norm": 0.7507709860801697,
+      "learning_rate": 0.00010509695849456487,
+      "loss": 2.5312,
+      "step": 9727
+    },
+    {
+      "epoch": 0.7850859494794609,
+      "grad_norm": 0.7101978063583374,
+      "learning_rate": 0.0001050811920530743,
+      "loss": 2.5833,
+      "step": 9728
+    },
+    {
+      "epoch": 0.785166653216044,
+      "grad_norm": 0.6814672946929932,
+      "learning_rate": 0.0001050654254849471,
+      "loss": 2.5466,
+      "step": 9729
+    },
+    {
+      "epoch": 0.7852473569526269,
+      "grad_norm": 0.7250106930732727,
+      "learning_rate": 0.0001050496587905763,
+      "loss": 2.5144,
+      "step": 9730
+    },
+    {
+      "epoch": 0.7853280606892099,
+      "grad_norm": 0.7125658392906189,
+      "learning_rate": 0.00010503389197035474,
+      "loss": 2.5384,
+      "step": 9731
+    },
+    {
+      "epoch": 0.7854087644257929,
+      "grad_norm": 0.7076827883720398,
+      "learning_rate": 0.00010501812502467547,
+      "loss": 2.4879,
+      "step": 9732
+    },
+    {
+      "epoch": 0.785489468162376,
+      "grad_norm": 0.632216215133667,
+      "learning_rate": 0.00010500235795393141,
+      "loss": 2.5678,
+      "step": 9733
+    },
+    {
+      "epoch": 0.785570171898959,
+      "grad_norm": 0.7376949191093445,
+      "learning_rate": 0.00010498659075851551,
+      "loss": 2.5024,
+      "step": 9734
+    },
+    {
+      "epoch": 0.7856508756355419,
+      "grad_norm": 0.6730546951293945,
+      "learning_rate": 0.00010497082343882072,
+      "loss": 2.5001,
+      "step": 9735
+    },
+    {
+      "epoch": 0.7857315793721249,
+      "grad_norm": 0.6958187818527222,
+      "learning_rate": 0.00010495505599524002,
+      "loss": 2.538,
+      "step": 9736
+    },
+    {
+      "epoch": 0.785812283108708,
+      "grad_norm": 0.6882508397102356,
+      "learning_rate": 0.00010493928842816638,
+      "loss": 2.5247,
+      "step": 9737
+    },
+    {
+      "epoch": 0.785892986845291,
+      "grad_norm": 0.711086630821228,
+      "learning_rate": 0.00010492352073799276,
+      "loss": 2.5721,
+      "step": 9738
+    },
+    {
+      "epoch": 0.7859736905818739,
+      "grad_norm": 0.7217094898223877,
+      "learning_rate": 0.00010490775292511214,
+      "loss": 2.5827,
+      "step": 9739
+    },
+    {
+      "epoch": 0.7860543943184569,
+      "grad_norm": 0.6812087893486023,
+      "learning_rate": 0.0001048919849899175,
+      "loss": 2.532,
+      "step": 9740
+    },
+    {
+      "epoch": 0.7861350980550399,
+      "grad_norm": 0.7449110150337219,
+      "learning_rate": 0.00010487621693280176,
+      "loss": 2.5611,
+      "step": 9741
+    },
+    {
+      "epoch": 0.786215801791623,
+      "grad_norm": 0.7297104001045227,
+      "learning_rate": 0.00010486044875415797,
+      "loss": 2.5173,
+      "step": 9742
+    },
+    {
+      "epoch": 0.786296505528206,
+      "grad_norm": 0.6741474270820618,
+      "learning_rate": 0.0001048446804543791,
+      "loss": 2.5451,
+      "step": 9743
+    },
+    {
+      "epoch": 0.7863772092647889,
+      "grad_norm": 0.6450859308242798,
+      "learning_rate": 0.00010482891203385812,
+      "loss": 2.551,
+      "step": 9744
+    },
+    {
+      "epoch": 0.7864579130013719,
+      "grad_norm": 0.6867123246192932,
+      "learning_rate": 0.00010481314349298805,
+      "loss": 2.4875,
+      "step": 9745
+    },
+    {
+      "epoch": 0.786538616737955,
+      "grad_norm": 0.6951552629470825,
+      "learning_rate": 0.00010479737483216183,
+      "loss": 2.6253,
+      "step": 9746
+    },
+    {
+      "epoch": 0.786619320474538,
+      "grad_norm": 0.6786869764328003,
+      "learning_rate": 0.0001047816060517725,
+      "loss": 2.5551,
+      "step": 9747
+    },
+    {
+      "epoch": 0.786700024211121,
+      "grad_norm": 0.698957622051239,
+      "learning_rate": 0.00010476583715221306,
+      "loss": 2.5554,
+      "step": 9748
+    },
+    {
+      "epoch": 0.7867807279477039,
+      "grad_norm": 0.6407502889633179,
+      "learning_rate": 0.00010475006813387648,
+      "loss": 2.5112,
+      "step": 9749
+    },
+    {
+      "epoch": 0.786861431684287,
+      "grad_norm": 0.660418689250946,
+      "learning_rate": 0.00010473429899715581,
+      "loss": 2.5557,
+      "step": 9750
+    },
+    {
+      "epoch": 0.78694213542087,
+      "grad_norm": 0.71445631980896,
+      "learning_rate": 0.00010471852974244403,
+      "loss": 2.5169,
+      "step": 9751
+    },
+    {
+      "epoch": 0.787022839157453,
+      "grad_norm": 0.6620494723320007,
+      "learning_rate": 0.00010470276037013414,
+      "loss": 2.5517,
+      "step": 9752
+    },
+    {
+      "epoch": 0.787103542894036,
+      "grad_norm": 0.6921235918998718,
+      "learning_rate": 0.00010468699088061917,
+      "loss": 2.5246,
+      "step": 9753
+    },
+    {
+      "epoch": 0.787184246630619,
+      "grad_norm": 0.6617140769958496,
+      "learning_rate": 0.00010467122127429214,
+      "loss": 2.4941,
+      "step": 9754
+    },
+    {
+      "epoch": 0.787264950367202,
+      "grad_norm": 0.6549816727638245,
+      "learning_rate": 0.00010465545155154608,
+      "loss": 2.5189,
+      "step": 9755
+    },
+    {
+      "epoch": 0.787345654103785,
+      "grad_norm": 0.7030060887336731,
+      "learning_rate": 0.00010463968171277396,
+      "loss": 2.5058,
+      "step": 9756
+    },
+    {
+      "epoch": 0.787426357840368,
+      "grad_norm": 0.7294049859046936,
+      "learning_rate": 0.00010462391175836886,
+      "loss": 2.5166,
+      "step": 9757
+    },
+    {
+      "epoch": 0.787507061576951,
+      "grad_norm": 0.6407562494277954,
+      "learning_rate": 0.00010460814168872382,
+      "loss": 2.5391,
+      "step": 9758
+    },
+    {
+      "epoch": 0.787587765313534,
+      "grad_norm": 0.8024646639823914,
+      "learning_rate": 0.0001045923715042318,
+      "loss": 2.7034,
+      "step": 9759
+    },
+    {
+      "epoch": 0.787668469050117,
+      "grad_norm": 0.7160943150520325,
+      "learning_rate": 0.00010457660120528592,
+      "loss": 2.6016,
+      "step": 9760
+    },
+    {
+      "epoch": 0.7877491727867,
+      "grad_norm": 0.6987707018852234,
+      "learning_rate": 0.00010456083079227916,
+      "loss": 2.5428,
+      "step": 9761
+    },
+    {
+      "epoch": 0.7878298765232831,
+      "grad_norm": 0.7235369086265564,
+      "learning_rate": 0.00010454506026560453,
+      "loss": 2.517,
+      "step": 9762
+    },
+    {
+      "epoch": 0.787910580259866,
+      "grad_norm": 0.6827502846717834,
+      "learning_rate": 0.00010452928962565518,
+      "loss": 2.5777,
+      "step": 9763
+    },
+    {
+      "epoch": 0.787991283996449,
+      "grad_norm": 0.71755450963974,
+      "learning_rate": 0.00010451351887282408,
+      "loss": 2.6004,
+      "step": 9764
+    },
+    {
+      "epoch": 0.788071987733032,
+      "grad_norm": 0.6988046765327454,
+      "learning_rate": 0.00010449774800750427,
+      "loss": 2.6116,
+      "step": 9765
+    },
+    {
+      "epoch": 0.7881526914696151,
+      "grad_norm": 0.6959548592567444,
+      "learning_rate": 0.00010448197703008884,
+      "loss": 2.5856,
+      "step": 9766
+    },
+    {
+      "epoch": 0.7882333952061981,
+      "grad_norm": 0.687042772769928,
+      "learning_rate": 0.00010446620594097079,
+      "loss": 2.5167,
+      "step": 9767
+    },
+    {
+      "epoch": 0.788314098942781,
+      "grad_norm": 0.6950173377990723,
+      "learning_rate": 0.00010445043474054325,
+      "loss": 2.5157,
+      "step": 9768
+    },
+    {
+      "epoch": 0.788394802679364,
+      "grad_norm": 0.680768609046936,
+      "learning_rate": 0.00010443466342919926,
+      "loss": 2.6177,
+      "step": 9769
+    },
+    {
+      "epoch": 0.7884755064159471,
+      "grad_norm": 0.7790142893791199,
+      "learning_rate": 0.00010441889200733181,
+      "loss": 2.5761,
+      "step": 9770
+    },
+    {
+      "epoch": 0.7885562101525301,
+      "grad_norm": 0.6207798719406128,
+      "learning_rate": 0.00010440312047533406,
+      "loss": 2.5305,
+      "step": 9771
+    },
+    {
+      "epoch": 0.7886369138891131,
+      "grad_norm": 0.7143635749816895,
+      "learning_rate": 0.00010438734883359903,
+      "loss": 2.5922,
+      "step": 9772
+    },
+    {
+      "epoch": 0.788717617625696,
+      "grad_norm": 0.7234248518943787,
+      "learning_rate": 0.00010437157708251977,
+      "loss": 2.6051,
+      "step": 9773
+    },
+    {
+      "epoch": 0.7887983213622791,
+      "grad_norm": 0.6602753400802612,
+      "learning_rate": 0.00010435580522248942,
+      "loss": 2.6002,
+      "step": 9774
+    },
+    {
+      "epoch": 0.7888790250988621,
+      "grad_norm": 0.6929246783256531,
+      "learning_rate": 0.00010434003325390101,
+      "loss": 2.5798,
+      "step": 9775
+    },
+    {
+      "epoch": 0.7889597288354451,
+      "grad_norm": 0.7355811595916748,
+      "learning_rate": 0.00010432426117714762,
+      "loss": 2.5859,
+      "step": 9776
+    },
+    {
+      "epoch": 0.789040432572028,
+      "grad_norm": 0.7009611129760742,
+      "learning_rate": 0.00010430848899262233,
+      "loss": 2.5535,
+      "step": 9777
+    },
+    {
+      "epoch": 0.7891211363086111,
+      "grad_norm": 0.6699070930480957,
+      "learning_rate": 0.00010429271670071823,
+      "loss": 2.5687,
+      "step": 9778
+    },
+    {
+      "epoch": 0.7892018400451941,
+      "grad_norm": 0.6632630228996277,
+      "learning_rate": 0.00010427694430182844,
+      "loss": 2.5359,
+      "step": 9779
+    },
+    {
+      "epoch": 0.7892825437817771,
+      "grad_norm": 0.7256911993026733,
+      "learning_rate": 0.000104261171796346,
+      "loss": 2.5432,
+      "step": 9780
+    },
+    {
+      "epoch": 0.7893632475183601,
+      "grad_norm": 0.6654312610626221,
+      "learning_rate": 0.000104245399184664,
+      "loss": 2.5432,
+      "step": 9781
+    },
+    {
+      "epoch": 0.7894439512549432,
+      "grad_norm": 0.6808900237083435,
+      "learning_rate": 0.00010422962646717557,
+      "loss": 2.4951,
+      "step": 9782
+    },
+    {
+      "epoch": 0.7895246549915261,
+      "grad_norm": 0.6655945181846619,
+      "learning_rate": 0.00010421385364427378,
+      "loss": 2.5152,
+      "step": 9783
+    },
+    {
+      "epoch": 0.7896053587281091,
+      "grad_norm": 0.8399274349212646,
+      "learning_rate": 0.00010419808071635178,
+      "loss": 2.5688,
+      "step": 9784
+    },
+    {
+      "epoch": 0.7896860624646921,
+      "grad_norm": 0.6412226557731628,
+      "learning_rate": 0.00010418230768380262,
+      "loss": 2.5527,
+      "step": 9785
+    },
+    {
+      "epoch": 0.7897667662012752,
+      "grad_norm": 0.6505058407783508,
+      "learning_rate": 0.0001041665345470194,
+      "loss": 2.5768,
+      "step": 9786
+    },
+    {
+      "epoch": 0.7898474699378581,
+      "grad_norm": 0.6297653317451477,
+      "learning_rate": 0.00010415076130639526,
+      "loss": 2.5372,
+      "step": 9787
+    },
+    {
+      "epoch": 0.7899281736744411,
+      "grad_norm": 0.6524460315704346,
+      "learning_rate": 0.00010413498796232331,
+      "loss": 2.5047,
+      "step": 9788
+    },
+    {
+      "epoch": 0.7900088774110241,
+      "grad_norm": 0.6637924313545227,
+      "learning_rate": 0.00010411921451519662,
+      "loss": 2.508,
+      "step": 9789
+    },
+    {
+      "epoch": 0.7900895811476072,
+      "grad_norm": 0.6423435211181641,
+      "learning_rate": 0.00010410344096540836,
+      "loss": 2.4597,
+      "step": 9790
+    },
+    {
+      "epoch": 0.7901702848841902,
+      "grad_norm": 0.6361977458000183,
+      "learning_rate": 0.00010408766731335163,
+      "loss": 2.5921,
+      "step": 9791
+    },
+    {
+      "epoch": 0.7902509886207731,
+      "grad_norm": 0.6792182922363281,
+      "learning_rate": 0.00010407189355941953,
+      "loss": 2.5543,
+      "step": 9792
+    },
+    {
+      "epoch": 0.7903316923573561,
+      "grad_norm": 0.6998419761657715,
+      "learning_rate": 0.00010405611970400519,
+      "loss": 2.5333,
+      "step": 9793
+    },
+    {
+      "epoch": 0.7904123960939391,
+      "grad_norm": 0.6730015873908997,
+      "learning_rate": 0.00010404034574750174,
+      "loss": 2.596,
+      "step": 9794
+    },
+    {
+      "epoch": 0.7904930998305222,
+      "grad_norm": 0.7120258808135986,
+      "learning_rate": 0.00010402457169030235,
+      "loss": 2.5314,
+      "step": 9795
+    },
+    {
+      "epoch": 0.7905738035671052,
+      "grad_norm": 0.6553651690483093,
+      "learning_rate": 0.0001040087975328001,
+      "loss": 2.4973,
+      "step": 9796
+    },
+    {
+      "epoch": 0.7906545073036881,
+      "grad_norm": 0.6506681442260742,
+      "learning_rate": 0.00010399302327538812,
+      "loss": 2.588,
+      "step": 9797
+    },
+    {
+      "epoch": 0.7907352110402711,
+      "grad_norm": 0.6737257242202759,
+      "learning_rate": 0.00010397724891845957,
+      "loss": 2.5454,
+      "step": 9798
+    },
+    {
+      "epoch": 0.7908159147768542,
+      "grad_norm": 0.670120894908905,
+      "learning_rate": 0.00010396147446240756,
+      "loss": 2.4926,
+      "step": 9799
+    },
+    {
+      "epoch": 0.7908966185134372,
+      "grad_norm": 0.7028468251228333,
+      "learning_rate": 0.00010394569990762529,
+      "loss": 2.5727,
+      "step": 9800
+    },
+    {
+      "epoch": 0.7909773222500202,
+      "grad_norm": 0.7084455490112305,
+      "learning_rate": 0.00010392992525450584,
+      "loss": 2.547,
+      "step": 9801
+    },
+    {
+      "epoch": 0.7910580259866031,
+      "grad_norm": 0.732694685459137,
+      "learning_rate": 0.0001039141505034424,
+      "loss": 2.5871,
+      "step": 9802
+    },
+    {
+      "epoch": 0.7911387297231862,
+      "grad_norm": 0.7214515209197998,
+      "learning_rate": 0.00010389837565482807,
+      "loss": 2.5672,
+      "step": 9803
+    },
+    {
+      "epoch": 0.7912194334597692,
+      "grad_norm": 0.6495330333709717,
+      "learning_rate": 0.00010388260070905604,
+      "loss": 2.5266,
+      "step": 9804
+    },
+    {
+      "epoch": 0.7913001371963522,
+      "grad_norm": 0.6930941343307495,
+      "learning_rate": 0.00010386682566651945,
+      "loss": 2.5734,
+      "step": 9805
+    },
+    {
+      "epoch": 0.7913808409329351,
+      "grad_norm": 0.714214563369751,
+      "learning_rate": 0.00010385105052761148,
+      "loss": 2.4987,
+      "step": 9806
+    },
+    {
+      "epoch": 0.7914615446695182,
+      "grad_norm": 0.7525388598442078,
+      "learning_rate": 0.00010383527529272523,
+      "loss": 2.5427,
+      "step": 9807
+    },
+    {
+      "epoch": 0.7915422484061012,
+      "grad_norm": 0.6088642477989197,
+      "learning_rate": 0.00010381949996225389,
+      "loss": 2.5018,
+      "step": 9808
+    },
+    {
+      "epoch": 0.7916229521426842,
+      "grad_norm": 0.6797540187835693,
+      "learning_rate": 0.00010380372453659066,
+      "loss": 2.5235,
+      "step": 9809
+    },
+    {
+      "epoch": 0.7917036558792672,
+      "grad_norm": 0.6754054427146912,
+      "learning_rate": 0.00010378794901612865,
+      "loss": 2.5343,
+      "step": 9810
+    },
+    {
+      "epoch": 0.7917843596158503,
+      "grad_norm": 0.7375015020370483,
+      "learning_rate": 0.00010377217340126106,
+      "loss": 2.6101,
+      "step": 9811
+    },
+    {
+      "epoch": 0.7918650633524332,
+      "grad_norm": 0.6487904191017151,
+      "learning_rate": 0.00010375639769238103,
+      "loss": 2.5408,
+      "step": 9812
+    },
+    {
+      "epoch": 0.7919457670890162,
+      "grad_norm": 0.7280275821685791,
+      "learning_rate": 0.00010374062188988176,
+      "loss": 2.5503,
+      "step": 9813
+    },
+    {
+      "epoch": 0.7920264708255992,
+      "grad_norm": 0.6944922208786011,
+      "learning_rate": 0.00010372484599415644,
+      "loss": 2.5815,
+      "step": 9814
+    },
+    {
+      "epoch": 0.7921071745621823,
+      "grad_norm": 0.6970139741897583,
+      "learning_rate": 0.00010370907000559818,
+      "loss": 2.546,
+      "step": 9815
+    },
+    {
+      "epoch": 0.7921878782987652,
+      "grad_norm": 0.7338151335716248,
+      "learning_rate": 0.00010369329392460023,
+      "loss": 2.5449,
+      "step": 9816
+    },
+    {
+      "epoch": 0.7922685820353482,
+      "grad_norm": 0.7763465642929077,
+      "learning_rate": 0.00010367751775155574,
+      "loss": 2.5331,
+      "step": 9817
+    },
+    {
+      "epoch": 0.7923492857719312,
+      "grad_norm": 0.6892645955085754,
+      "learning_rate": 0.00010366174148685786,
+      "loss": 2.5617,
+      "step": 9818
+    },
+    {
+      "epoch": 0.7924299895085143,
+      "grad_norm": 0.7388250231742859,
+      "learning_rate": 0.00010364596513089984,
+      "loss": 2.5236,
+      "step": 9819
+    },
+    {
+      "epoch": 0.7925106932450973,
+      "grad_norm": 0.7035132646560669,
+      "learning_rate": 0.00010363018868407482,
+      "loss": 2.5711,
+      "step": 9820
+    },
+    {
+      "epoch": 0.7925913969816802,
+      "grad_norm": 0.7087043523788452,
+      "learning_rate": 0.00010361441214677603,
+      "loss": 2.5416,
+      "step": 9821
+    },
+    {
+      "epoch": 0.7926721007182632,
+      "grad_norm": 0.7173168063163757,
+      "learning_rate": 0.00010359863551939664,
+      "loss": 2.529,
+      "step": 9822
+    },
+    {
+      "epoch": 0.7927528044548463,
+      "grad_norm": 0.7007408738136292,
+      "learning_rate": 0.00010358285880232983,
+      "loss": 2.5287,
+      "step": 9823
+    },
+    {
+      "epoch": 0.7928335081914293,
+      "grad_norm": 0.7731965780258179,
+      "learning_rate": 0.0001035670819959688,
+      "loss": 2.5913,
+      "step": 9824
+    },
+    {
+      "epoch": 0.7929142119280123,
+      "grad_norm": 0.6625120639801025,
+      "learning_rate": 0.00010355130510070681,
+      "loss": 2.5815,
+      "step": 9825
+    },
+    {
+      "epoch": 0.7929949156645952,
+      "grad_norm": 0.6628395318984985,
+      "learning_rate": 0.00010353552811693699,
+      "loss": 2.512,
+      "step": 9826
+    },
+    {
+      "epoch": 0.7930756194011783,
+      "grad_norm": 0.6565915942192078,
+      "learning_rate": 0.00010351975104505256,
+      "loss": 2.54,
+      "step": 9827
+    },
+    {
+      "epoch": 0.7931563231377613,
+      "grad_norm": 0.6581636667251587,
+      "learning_rate": 0.00010350397388544672,
+      "loss": 2.5462,
+      "step": 9828
+    },
+    {
+      "epoch": 0.7932370268743443,
+      "grad_norm": 0.705668568611145,
+      "learning_rate": 0.0001034881966385127,
+      "loss": 2.5241,
+      "step": 9829
+    },
+    {
+      "epoch": 0.7933177306109273,
+      "grad_norm": 0.7047126293182373,
+      "learning_rate": 0.00010347241930464373,
+      "loss": 2.5275,
+      "step": 9830
+    },
+    {
+      "epoch": 0.7933984343475103,
+      "grad_norm": 0.6285849213600159,
+      "learning_rate": 0.00010345664188423296,
+      "loss": 2.518,
+      "step": 9831
+    },
+    {
+      "epoch": 0.7934791380840933,
+      "grad_norm": 0.697542130947113,
+      "learning_rate": 0.00010344086437767366,
+      "loss": 2.5219,
+      "step": 9832
+    },
+    {
+      "epoch": 0.7935598418206763,
+      "grad_norm": 0.6349283456802368,
+      "learning_rate": 0.00010342508678535903,
+      "loss": 2.5277,
+      "step": 9833
+    },
+    {
+      "epoch": 0.7936405455572593,
+      "grad_norm": 0.7084335088729858,
+      "learning_rate": 0.00010340930910768225,
+      "loss": 2.476,
+      "step": 9834
+    },
+    {
+      "epoch": 0.7937212492938424,
+      "grad_norm": 0.6714156866073608,
+      "learning_rate": 0.00010339353134503662,
+      "loss": 2.556,
+      "step": 9835
+    },
+    {
+      "epoch": 0.7938019530304253,
+      "grad_norm": 0.6687895059585571,
+      "learning_rate": 0.00010337775349781527,
+      "loss": 2.5756,
+      "step": 9836
+    },
+    {
+      "epoch": 0.7938826567670083,
+      "grad_norm": 0.669784665107727,
+      "learning_rate": 0.00010336197556641152,
+      "loss": 2.5545,
+      "step": 9837
+    },
+    {
+      "epoch": 0.7939633605035913,
+      "grad_norm": 0.6738600134849548,
+      "learning_rate": 0.0001033461975512185,
+      "loss": 2.5807,
+      "step": 9838
+    },
+    {
+      "epoch": 0.7940440642401744,
+      "grad_norm": 0.691443681716919,
+      "learning_rate": 0.00010333041945262953,
+      "loss": 2.5279,
+      "step": 9839
+    },
+    {
+      "epoch": 0.7941247679767574,
+      "grad_norm": 0.6283861398696899,
+      "learning_rate": 0.0001033146412710378,
+      "loss": 2.5355,
+      "step": 9840
+    },
+    {
+      "epoch": 0.7942054717133403,
+      "grad_norm": 0.6491204500198364,
+      "learning_rate": 0.00010329886300683655,
+      "loss": 2.5431,
+      "step": 9841
+    },
+    {
+      "epoch": 0.7942861754499233,
+      "grad_norm": 0.6673988103866577,
+      "learning_rate": 0.00010328308466041898,
+      "loss": 2.5845,
+      "step": 9842
+    },
+    {
+      "epoch": 0.7943668791865063,
+      "grad_norm": 0.6669130325317383,
+      "learning_rate": 0.00010326730623217837,
+      "loss": 2.5348,
+      "step": 9843
+    },
+    {
+      "epoch": 0.7944475829230894,
+      "grad_norm": 0.7003189921379089,
+      "learning_rate": 0.00010325152772250795,
+      "loss": 2.5779,
+      "step": 9844
+    },
+    {
+      "epoch": 0.7945282866596723,
+      "grad_norm": 0.6602177619934082,
+      "learning_rate": 0.00010323574913180097,
+      "loss": 2.5527,
+      "step": 9845
+    },
+    {
+      "epoch": 0.7946089903962553,
+      "grad_norm": 0.7053726315498352,
+      "learning_rate": 0.00010321997046045066,
+      "loss": 2.566,
+      "step": 9846
+    },
+    {
+      "epoch": 0.7946896941328383,
+      "grad_norm": 0.7428076863288879,
+      "learning_rate": 0.00010320419170885025,
+      "loss": 2.5348,
+      "step": 9847
+    },
+    {
+      "epoch": 0.7947703978694214,
+      "grad_norm": 0.7029163837432861,
+      "learning_rate": 0.00010318841287739303,
+      "loss": 2.5387,
+      "step": 9848
+    },
+    {
+      "epoch": 0.7948511016060044,
+      "grad_norm": 0.6159133911132812,
+      "learning_rate": 0.00010317263396647221,
+      "loss": 2.5408,
+      "step": 9849
+    },
+    {
+      "epoch": 0.7949318053425873,
+      "grad_norm": 0.6748857498168945,
+      "learning_rate": 0.00010315685497648106,
+      "loss": 2.5299,
+      "step": 9850
+    },
+    {
+      "epoch": 0.7950125090791703,
+      "grad_norm": 0.6281898021697998,
+      "learning_rate": 0.00010314107590781284,
+      "loss": 2.5202,
+      "step": 9851
+    },
+    {
+      "epoch": 0.7950932128157534,
+      "grad_norm": 0.6602163910865784,
+      "learning_rate": 0.00010312529676086078,
+      "loss": 2.5119,
+      "step": 9852
+    },
+    {
+      "epoch": 0.7951739165523364,
+      "grad_norm": 0.6665403246879578,
+      "learning_rate": 0.00010310951753601818,
+      "loss": 2.5913,
+      "step": 9853
+    },
+    {
+      "epoch": 0.7952546202889194,
+      "grad_norm": 0.6705873012542725,
+      "learning_rate": 0.00010309373823367827,
+      "loss": 2.6039,
+      "step": 9854
+    },
+    {
+      "epoch": 0.7953353240255023,
+      "grad_norm": 0.6571313738822937,
+      "learning_rate": 0.0001030779588542343,
+      "loss": 2.5629,
+      "step": 9855
+    },
+    {
+      "epoch": 0.7954160277620854,
+      "grad_norm": 0.6597230434417725,
+      "learning_rate": 0.00010306217939807956,
+      "loss": 2.5569,
+      "step": 9856
+    },
+    {
+      "epoch": 0.7954967314986684,
+      "grad_norm": 0.7098817229270935,
+      "learning_rate": 0.00010304639986560733,
+      "loss": 2.4736,
+      "step": 9857
+    },
+    {
+      "epoch": 0.7955774352352514,
+      "grad_norm": 0.628663957118988,
+      "learning_rate": 0.00010303062025721082,
+      "loss": 2.5241,
+      "step": 9858
+    },
+    {
+      "epoch": 0.7956581389718343,
+      "grad_norm": 0.630843460559845,
+      "learning_rate": 0.00010301484057328333,
+      "loss": 2.5604,
+      "step": 9859
+    },
+    {
+      "epoch": 0.7957388427084174,
+      "grad_norm": 0.7457596659660339,
+      "learning_rate": 0.00010299906081421813,
+      "loss": 2.5675,
+      "step": 9860
+    },
+    {
+      "epoch": 0.7958195464450004,
+      "grad_norm": 0.6566091775894165,
+      "learning_rate": 0.00010298328098040851,
+      "loss": 2.4918,
+      "step": 9861
+    },
+    {
+      "epoch": 0.7959002501815834,
+      "grad_norm": 0.657357931137085,
+      "learning_rate": 0.00010296750107224773,
+      "loss": 2.5268,
+      "step": 9862
+    },
+    {
+      "epoch": 0.7959809539181664,
+      "grad_norm": 0.7021927833557129,
+      "learning_rate": 0.00010295172109012905,
+      "loss": 2.528,
+      "step": 9863
+    },
+    {
+      "epoch": 0.7960616576547495,
+      "grad_norm": 0.662053108215332,
+      "learning_rate": 0.00010293594103444578,
+      "loss": 2.5483,
+      "step": 9864
+    },
+    {
+      "epoch": 0.7961423613913324,
+      "grad_norm": 0.776407778263092,
+      "learning_rate": 0.00010292016090559118,
+      "loss": 2.6089,
+      "step": 9865
+    },
+    {
+      "epoch": 0.7962230651279154,
+      "grad_norm": 0.6499512791633606,
+      "learning_rate": 0.00010290438070395854,
+      "loss": 2.5609,
+      "step": 9866
+    },
+    {
+      "epoch": 0.7963037688644984,
+      "grad_norm": 0.6802246570587158,
+      "learning_rate": 0.00010288860042994113,
+      "loss": 2.5217,
+      "step": 9867
+    },
+    {
+      "epoch": 0.7963844726010815,
+      "grad_norm": 0.6371235847473145,
+      "learning_rate": 0.00010287282008393224,
+      "loss": 2.4783,
+      "step": 9868
+    },
+    {
+      "epoch": 0.7964651763376644,
+      "grad_norm": 0.7070169448852539,
+      "learning_rate": 0.00010285703966632518,
+      "loss": 2.5006,
+      "step": 9869
+    },
+    {
+      "epoch": 0.7965458800742474,
+      "grad_norm": 0.657738208770752,
+      "learning_rate": 0.00010284125917751323,
+      "loss": 2.551,
+      "step": 9870
+    },
+    {
+      "epoch": 0.7966265838108304,
+      "grad_norm": 0.7936853170394897,
+      "learning_rate": 0.00010282547861788964,
+      "loss": 2.574,
+      "step": 9871
+    },
+    {
+      "epoch": 0.7967072875474135,
+      "grad_norm": 0.675715982913971,
+      "learning_rate": 0.00010280969798784779,
+      "loss": 2.5288,
+      "step": 9872
+    },
+    {
+      "epoch": 0.7967879912839965,
+      "grad_norm": 0.6980394124984741,
+      "learning_rate": 0.00010279391728778092,
+      "loss": 2.5437,
+      "step": 9873
+    },
+    {
+      "epoch": 0.7968686950205794,
+      "grad_norm": 0.6580469608306885,
+      "learning_rate": 0.00010277813651808226,
+      "loss": 2.5574,
+      "step": 9874
+    },
+    {
+      "epoch": 0.7969493987571624,
+      "grad_norm": 0.6960238218307495,
+      "learning_rate": 0.00010276235567914522,
+      "loss": 2.5477,
+      "step": 9875
+    },
+    {
+      "epoch": 0.7970301024937455,
+      "grad_norm": 0.704140841960907,
+      "learning_rate": 0.00010274657477136304,
+      "loss": 2.5099,
+      "step": 9876
+    },
+    {
+      "epoch": 0.7971108062303285,
+      "grad_norm": 0.7238990068435669,
+      "learning_rate": 0.00010273079379512906,
+      "loss": 2.6182,
+      "step": 9877
+    },
+    {
+      "epoch": 0.7971915099669115,
+      "grad_norm": 0.6527700424194336,
+      "learning_rate": 0.00010271501275083657,
+      "loss": 2.5148,
+      "step": 9878
+    },
+    {
+      "epoch": 0.7972722137034944,
+      "grad_norm": 0.6665365695953369,
+      "learning_rate": 0.00010269923163887884,
+      "loss": 2.5624,
+      "step": 9879
+    },
+    {
+      "epoch": 0.7973529174400775,
+      "grad_norm": 0.7304019927978516,
+      "learning_rate": 0.0001026834504596492,
+      "loss": 2.5537,
+      "step": 9880
+    },
+    {
+      "epoch": 0.7974336211766605,
+      "grad_norm": 0.6645877957344055,
+      "learning_rate": 0.00010266766921354099,
+      "loss": 2.5381,
+      "step": 9881
+    },
+    {
+      "epoch": 0.7975143249132435,
+      "grad_norm": 0.6817314624786377,
+      "learning_rate": 0.00010265188790094744,
+      "loss": 2.5399,
+      "step": 9882
+    },
+    {
+      "epoch": 0.7975950286498265,
+      "grad_norm": 0.7477232217788696,
+      "learning_rate": 0.00010263610652226194,
+      "loss": 2.6461,
+      "step": 9883
+    },
+    {
+      "epoch": 0.7976757323864095,
+      "grad_norm": 0.7087170481681824,
+      "learning_rate": 0.00010262032507787777,
+      "loss": 2.5469,
+      "step": 9884
+    },
+    {
+      "epoch": 0.7977564361229925,
+      "grad_norm": 0.7093435525894165,
+      "learning_rate": 0.00010260454356818825,
+      "loss": 2.5606,
+      "step": 9885
+    },
+    {
+      "epoch": 0.7978371398595755,
+      "grad_norm": 0.6662636399269104,
+      "learning_rate": 0.00010258876199358672,
+      "loss": 2.5415,
+      "step": 9886
+    },
+    {
+      "epoch": 0.7979178435961585,
+      "grad_norm": 0.6829736232757568,
+      "learning_rate": 0.00010257298035446644,
+      "loss": 2.5618,
+      "step": 9887
+    },
+    {
+      "epoch": 0.7979985473327416,
+      "grad_norm": 0.6872264742851257,
+      "learning_rate": 0.00010255719865122077,
+      "loss": 2.5629,
+      "step": 9888
+    },
+    {
+      "epoch": 0.7980792510693245,
+      "grad_norm": 0.6988633871078491,
+      "learning_rate": 0.00010254141688424303,
+      "loss": 2.5191,
+      "step": 9889
+    },
+    {
+      "epoch": 0.7981599548059075,
+      "grad_norm": 0.6787285804748535,
+      "learning_rate": 0.00010252563505392654,
+      "loss": 2.5003,
+      "step": 9890
+    },
+    {
+      "epoch": 0.7982406585424905,
+      "grad_norm": 0.6703466773033142,
+      "learning_rate": 0.00010250985316066461,
+      "loss": 2.5442,
+      "step": 9891
+    },
+    {
+      "epoch": 0.7983213622790736,
+      "grad_norm": 0.6463642120361328,
+      "learning_rate": 0.0001024940712048506,
+      "loss": 2.5236,
+      "step": 9892
+    },
+    {
+      "epoch": 0.7984020660156566,
+      "grad_norm": 0.6835207939147949,
+      "learning_rate": 0.0001024782891868778,
+      "loss": 2.5094,
+      "step": 9893
+    },
+    {
+      "epoch": 0.7984827697522395,
+      "grad_norm": 0.6621001958847046,
+      "learning_rate": 0.00010246250710713956,
+      "loss": 2.5456,
+      "step": 9894
+    },
+    {
+      "epoch": 0.7985634734888225,
+      "grad_norm": 0.6675469875335693,
+      "learning_rate": 0.0001024467249660292,
+      "loss": 2.5312,
+      "step": 9895
+    },
+    {
+      "epoch": 0.7986441772254055,
+      "grad_norm": 0.7357796430587769,
+      "learning_rate": 0.00010243094276394007,
+      "loss": 2.5374,
+      "step": 9896
+    },
+    {
+      "epoch": 0.7987248809619886,
+      "grad_norm": 0.7005879878997803,
+      "learning_rate": 0.00010241516050126549,
+      "loss": 2.5667,
+      "step": 9897
+    },
+    {
+      "epoch": 0.7988055846985715,
+      "grad_norm": 0.669870913028717,
+      "learning_rate": 0.0001023993781783988,
+      "loss": 2.533,
+      "step": 9898
+    },
+    {
+      "epoch": 0.7988862884351545,
+      "grad_norm": 0.7584091424942017,
+      "learning_rate": 0.00010238359579573333,
+      "loss": 2.5995,
+      "step": 9899
+    },
+    {
+      "epoch": 0.7989669921717375,
+      "grad_norm": 0.6931570172309875,
+      "learning_rate": 0.00010236781335366239,
+      "loss": 2.5506,
+      "step": 9900
+    },
+    {
+      "epoch": 0.7990476959083206,
+      "grad_norm": 0.6810948848724365,
+      "learning_rate": 0.0001023520308525794,
+      "loss": 2.5048,
+      "step": 9901
+    },
+    {
+      "epoch": 0.7991283996449036,
+      "grad_norm": 0.6857194900512695,
+      "learning_rate": 0.00010233624829287765,
+      "loss": 2.5559,
+      "step": 9902
+    },
+    {
+      "epoch": 0.7992091033814865,
+      "grad_norm": 0.6685707569122314,
+      "learning_rate": 0.00010232046567495046,
+      "loss": 2.5661,
+      "step": 9903
+    },
+    {
+      "epoch": 0.7992898071180695,
+      "grad_norm": 0.6626694202423096,
+      "learning_rate": 0.00010230468299919121,
+      "loss": 2.6293,
+      "step": 9904
+    },
+    {
+      "epoch": 0.7993705108546526,
+      "grad_norm": 0.6407302021980286,
+      "learning_rate": 0.00010228890026599323,
+      "loss": 2.5552,
+      "step": 9905
+    },
+    {
+      "epoch": 0.7994512145912356,
+      "grad_norm": 0.762235701084137,
+      "learning_rate": 0.00010227311747574986,
+      "loss": 2.4904,
+      "step": 9906
+    },
+    {
+      "epoch": 0.7995319183278186,
+      "grad_norm": 0.703507661819458,
+      "learning_rate": 0.0001022573346288545,
+      "loss": 2.5684,
+      "step": 9907
+    },
+    {
+      "epoch": 0.7996126220644015,
+      "grad_norm": 0.82541823387146,
+      "learning_rate": 0.00010224155172570043,
+      "loss": 2.521,
+      "step": 9908
+    },
+    {
+      "epoch": 0.7996933258009846,
+      "grad_norm": 0.6836804747581482,
+      "learning_rate": 0.00010222576876668104,
+      "loss": 2.5364,
+      "step": 9909
+    },
+    {
+      "epoch": 0.7997740295375676,
+      "grad_norm": 0.7388977408409119,
+      "learning_rate": 0.00010220998575218966,
+      "loss": 2.5724,
+      "step": 9910
+    },
+    {
+      "epoch": 0.7998547332741506,
+      "grad_norm": 0.7380896806716919,
+      "learning_rate": 0.00010219420268261966,
+      "loss": 2.5918,
+      "step": 9911
+    },
+    {
+      "epoch": 0.7999354370107336,
+      "grad_norm": 0.7303522825241089,
+      "learning_rate": 0.00010217841955836442,
+      "loss": 2.5432,
+      "step": 9912
+    },
+    {
+      "epoch": 0.8000161407473166,
+      "grad_norm": 0.6859301924705505,
+      "learning_rate": 0.00010216263637981727,
+      "loss": 2.5734,
+      "step": 9913
+    },
+    {
+      "epoch": 0.8000968444838996,
+      "grad_norm": 0.731910228729248,
+      "learning_rate": 0.00010214685314737154,
+      "loss": 2.5227,
+      "step": 9914
+    },
+    {
+      "epoch": 0.8001775482204826,
+      "grad_norm": 0.7105006575584412,
+      "learning_rate": 0.00010213106986142062,
+      "loss": 2.5335,
+      "step": 9915
+    },
+    {
+      "epoch": 0.8002582519570656,
+      "grad_norm": 0.7337056994438171,
+      "learning_rate": 0.00010211528652235786,
+      "loss": 2.6204,
+      "step": 9916
+    },
+    {
+      "epoch": 0.8003389556936487,
+      "grad_norm": 0.7350614666938782,
+      "learning_rate": 0.00010209950313057668,
+      "loss": 2.5264,
+      "step": 9917
+    },
+    {
+      "epoch": 0.8004196594302316,
+      "grad_norm": 0.6411921977996826,
+      "learning_rate": 0.00010208371968647036,
+      "loss": 2.4642,
+      "step": 9918
+    },
+    {
+      "epoch": 0.8005003631668146,
+      "grad_norm": 0.7601611018180847,
+      "learning_rate": 0.00010206793619043229,
+      "loss": 2.6249,
+      "step": 9919
+    },
+    {
+      "epoch": 0.8005810669033976,
+      "grad_norm": 0.7086012363433838,
+      "learning_rate": 0.00010205215264285585,
+      "loss": 2.5508,
+      "step": 9920
+    },
+    {
+      "epoch": 0.8006617706399807,
+      "grad_norm": 0.7267128825187683,
+      "learning_rate": 0.00010203636904413443,
+      "loss": 2.5109,
+      "step": 9921
+    },
+    {
+      "epoch": 0.8007424743765637,
+      "grad_norm": 0.7606067657470703,
+      "learning_rate": 0.00010202058539466132,
+      "loss": 2.5172,
+      "step": 9922
+    },
+    {
+      "epoch": 0.8008231781131466,
+      "grad_norm": 0.7610498666763306,
+      "learning_rate": 0.00010200480169483,
+      "loss": 2.5085,
+      "step": 9923
+    },
+    {
+      "epoch": 0.8009038818497296,
+      "grad_norm": 0.7604225873947144,
+      "learning_rate": 0.00010198901794503373,
+      "loss": 2.5615,
+      "step": 9924
+    },
+    {
+      "epoch": 0.8009845855863127,
+      "grad_norm": 0.739532470703125,
+      "learning_rate": 0.00010197323414566596,
+      "loss": 2.5574,
+      "step": 9925
+    },
+    {
+      "epoch": 0.8010652893228957,
+      "grad_norm": 0.6913303136825562,
+      "learning_rate": 0.00010195745029712003,
+      "loss": 2.5403,
+      "step": 9926
+    },
+    {
+      "epoch": 0.8011459930594786,
+      "grad_norm": 0.6963592767715454,
+      "learning_rate": 0.0001019416663997893,
+      "loss": 2.5615,
+      "step": 9927
+    },
+    {
+      "epoch": 0.8012266967960616,
+      "grad_norm": 0.681481122970581,
+      "learning_rate": 0.0001019258824540672,
+      "loss": 2.5125,
+      "step": 9928
+    },
+    {
+      "epoch": 0.8013074005326447,
+      "grad_norm": 0.7192744016647339,
+      "learning_rate": 0.00010191009846034709,
+      "loss": 2.5952,
+      "step": 9929
+    },
+    {
+      "epoch": 0.8013881042692277,
+      "grad_norm": 0.7030046582221985,
+      "learning_rate": 0.00010189431441902228,
+      "loss": 2.5445,
+      "step": 9930
+    },
+    {
+      "epoch": 0.8014688080058107,
+      "grad_norm": 0.6180598139762878,
+      "learning_rate": 0.00010187853033048622,
+      "loss": 2.4902,
+      "step": 9931
+    },
+    {
+      "epoch": 0.8015495117423936,
+      "grad_norm": 0.7479971051216125,
+      "learning_rate": 0.0001018627461951323,
+      "loss": 2.5703,
+      "step": 9932
+    },
+    {
+      "epoch": 0.8016302154789767,
+      "grad_norm": 0.7339857220649719,
+      "learning_rate": 0.00010184696201335387,
+      "loss": 2.5744,
+      "step": 9933
+    },
+    {
+      "epoch": 0.8017109192155597,
+      "grad_norm": 0.6741397380828857,
+      "learning_rate": 0.00010183117778554432,
+      "loss": 2.5777,
+      "step": 9934
+    },
+    {
+      "epoch": 0.8017916229521427,
+      "grad_norm": 0.6731706857681274,
+      "learning_rate": 0.00010181539351209699,
+      "loss": 2.5438,
+      "step": 9935
+    },
+    {
+      "epoch": 0.8018723266887257,
+      "grad_norm": 0.6929418444633484,
+      "learning_rate": 0.00010179960919340535,
+      "loss": 2.5308,
+      "step": 9936
+    },
+    {
+      "epoch": 0.8019530304253087,
+      "grad_norm": 0.7383175492286682,
+      "learning_rate": 0.00010178382482986271,
+      "loss": 2.5623,
+      "step": 9937
+    },
+    {
+      "epoch": 0.8020337341618917,
+      "grad_norm": 0.6872193217277527,
+      "learning_rate": 0.00010176804042186252,
+      "loss": 2.5271,
+      "step": 9938
+    },
+    {
+      "epoch": 0.8021144378984747,
+      "grad_norm": 0.7354295253753662,
+      "learning_rate": 0.00010175225596979816,
+      "loss": 2.5122,
+      "step": 9939
+    },
+    {
+      "epoch": 0.8021951416350577,
+      "grad_norm": 0.7589237689971924,
+      "learning_rate": 0.00010173647147406297,
+      "loss": 2.5529,
+      "step": 9940
+    },
+    {
+      "epoch": 0.8022758453716408,
+      "grad_norm": 0.6998353004455566,
+      "learning_rate": 0.00010172068693505037,
+      "loss": 2.4683,
+      "step": 9941
+    },
+    {
+      "epoch": 0.8023565491082237,
+      "grad_norm": 0.6816055178642273,
+      "learning_rate": 0.00010170490235315377,
+      "loss": 2.567,
+      "step": 9942
+    },
+    {
+      "epoch": 0.8024372528448067,
+      "grad_norm": 0.7188318371772766,
+      "learning_rate": 0.00010168911772876652,
+      "loss": 2.5631,
+      "step": 9943
+    },
+    {
+      "epoch": 0.8025179565813897,
+      "grad_norm": 0.6925922632217407,
+      "learning_rate": 0.00010167333306228209,
+      "loss": 2.4872,
+      "step": 9944
+    },
+    {
+      "epoch": 0.8025986603179727,
+      "grad_norm": 0.7081493735313416,
+      "learning_rate": 0.00010165754835409377,
+      "loss": 2.5482,
+      "step": 9945
+    },
+    {
+      "epoch": 0.8026793640545558,
+      "grad_norm": 0.6838935613632202,
+      "learning_rate": 0.00010164176360459505,
+      "loss": 2.541,
+      "step": 9946
+    },
+    {
+      "epoch": 0.8027600677911387,
+      "grad_norm": 0.6959214210510254,
+      "learning_rate": 0.00010162597881417928,
+      "loss": 2.4574,
+      "step": 9947
+    },
+    {
+      "epoch": 0.8028407715277217,
+      "grad_norm": 0.693004310131073,
+      "learning_rate": 0.00010161019398323986,
+      "loss": 2.5553,
+      "step": 9948
+    },
+    {
+      "epoch": 0.8029214752643047,
+      "grad_norm": 0.6683690547943115,
+      "learning_rate": 0.00010159440911217022,
+      "loss": 2.5501,
+      "step": 9949
+    },
+    {
+      "epoch": 0.8030021790008878,
+      "grad_norm": 0.6797001361846924,
+      "learning_rate": 0.0001015786242013637,
+      "loss": 2.5731,
+      "step": 9950
+    },
+    {
+      "epoch": 0.8030828827374707,
+      "grad_norm": 0.6621012091636658,
+      "learning_rate": 0.00010156283925121375,
+      "loss": 2.5278,
+      "step": 9951
+    },
+    {
+      "epoch": 0.8031635864740537,
+      "grad_norm": 0.7024650573730469,
+      "learning_rate": 0.00010154705426211377,
+      "loss": 2.5939,
+      "step": 9952
+    },
+    {
+      "epoch": 0.8032442902106367,
+      "grad_norm": 0.6756548285484314,
+      "learning_rate": 0.00010153126923445714,
+      "loss": 2.5797,
+      "step": 9953
+    },
+    {
+      "epoch": 0.8033249939472198,
+      "grad_norm": 0.6560662984848022,
+      "learning_rate": 0.00010151548416863732,
+      "loss": 2.5358,
+      "step": 9954
+    },
+    {
+      "epoch": 0.8034056976838028,
+      "grad_norm": 0.7172456979751587,
+      "learning_rate": 0.00010149969906504766,
+      "loss": 2.5054,
+      "step": 9955
+    },
+    {
+      "epoch": 0.8034864014203857,
+      "grad_norm": 0.6379461288452148,
+      "learning_rate": 0.00010148391392408152,
+      "loss": 2.5341,
+      "step": 9956
+    },
+    {
+      "epoch": 0.8035671051569687,
+      "grad_norm": 0.6553892493247986,
+      "learning_rate": 0.00010146812874613243,
+      "loss": 2.5618,
+      "step": 9957
+    },
+    {
+      "epoch": 0.8036478088935518,
+      "grad_norm": 0.6940072178840637,
+      "learning_rate": 0.00010145234353159372,
+      "loss": 2.5686,
+      "step": 9958
+    },
+    {
+      "epoch": 0.8037285126301348,
+      "grad_norm": 0.6641896963119507,
+      "learning_rate": 0.00010143655828085878,
+      "loss": 2.5188,
+      "step": 9959
+    },
+    {
+      "epoch": 0.8038092163667178,
+      "grad_norm": 0.6622887253761292,
+      "learning_rate": 0.00010142077299432111,
+      "loss": 2.54,
+      "step": 9960
+    },
+    {
+      "epoch": 0.8038899201033007,
+      "grad_norm": 0.7216808795928955,
+      "learning_rate": 0.000101404987672374,
+      "loss": 2.5775,
+      "step": 9961
+    },
+    {
+      "epoch": 0.8039706238398838,
+      "grad_norm": 0.6544952988624573,
+      "learning_rate": 0.00010138920231541095,
+      "loss": 2.6066,
+      "step": 9962
+    },
+    {
+      "epoch": 0.8040513275764668,
+      "grad_norm": 0.6869354248046875,
+      "learning_rate": 0.00010137341692382539,
+      "loss": 2.5157,
+      "step": 9963
+    },
+    {
+      "epoch": 0.8041320313130498,
+      "grad_norm": 0.6731898784637451,
+      "learning_rate": 0.00010135763149801063,
+      "loss": 2.4369,
+      "step": 9964
+    },
+    {
+      "epoch": 0.8042127350496328,
+      "grad_norm": 0.6943373084068298,
+      "learning_rate": 0.00010134184603836017,
+      "loss": 2.5529,
+      "step": 9965
+    },
+    {
+      "epoch": 0.8042934387862158,
+      "grad_norm": 0.729928195476532,
+      "learning_rate": 0.00010132606054526739,
+      "loss": 2.5814,
+      "step": 9966
+    },
+    {
+      "epoch": 0.8043741425227988,
+      "grad_norm": 0.6491130590438843,
+      "learning_rate": 0.00010131027501912571,
+      "loss": 2.5246,
+      "step": 9967
+    },
+    {
+      "epoch": 0.8044548462593818,
+      "grad_norm": 0.747756838798523,
+      "learning_rate": 0.00010129448946032857,
+      "loss": 2.513,
+      "step": 9968
+    },
+    {
+      "epoch": 0.8045355499959648,
+      "grad_norm": 0.6449645757675171,
+      "learning_rate": 0.00010127870386926935,
+      "loss": 2.5232,
+      "step": 9969
+    },
+    {
+      "epoch": 0.8046162537325479,
+      "grad_norm": 0.6425037980079651,
+      "learning_rate": 0.0001012629182463415,
+      "loss": 2.5065,
+      "step": 9970
+    },
+    {
+      "epoch": 0.8046969574691308,
+      "grad_norm": 0.7340624332427979,
+      "learning_rate": 0.00010124713259193843,
+      "loss": 2.5325,
+      "step": 9971
+    },
+    {
+      "epoch": 0.8047776612057138,
+      "grad_norm": 0.7308940291404724,
+      "learning_rate": 0.00010123134690645352,
+      "loss": 2.5717,
+      "step": 9972
+    },
+    {
+      "epoch": 0.8048583649422968,
+      "grad_norm": 0.7128338813781738,
+      "learning_rate": 0.00010121556119028028,
+      "loss": 2.5548,
+      "step": 9973
+    },
+    {
+      "epoch": 0.8049390686788799,
+      "grad_norm": 0.7027677893638611,
+      "learning_rate": 0.00010119977544381207,
+      "loss": 2.5311,
+      "step": 9974
+    },
+    {
+      "epoch": 0.8050197724154629,
+      "grad_norm": 0.7022054195404053,
+      "learning_rate": 0.00010118398966744229,
+      "loss": 2.5177,
+      "step": 9975
+    },
+    {
+      "epoch": 0.8051004761520458,
+      "grad_norm": 0.7382696270942688,
+      "learning_rate": 0.00010116820386156441,
+      "loss": 2.532,
+      "step": 9976
+    },
+    {
+      "epoch": 0.8051811798886288,
+      "grad_norm": 0.6968613862991333,
+      "learning_rate": 0.00010115241802657181,
+      "loss": 2.536,
+      "step": 9977
+    },
+    {
+      "epoch": 0.8052618836252119,
+      "grad_norm": 0.8277899026870728,
+      "learning_rate": 0.00010113663216285798,
+      "loss": 2.5963,
+      "step": 9978
+    },
+    {
+      "epoch": 0.8053425873617949,
+      "grad_norm": 0.677707314491272,
+      "learning_rate": 0.00010112084627081629,
+      "loss": 2.5041,
+      "step": 9979
+    },
+    {
+      "epoch": 0.8054232910983778,
+      "grad_norm": 0.6943314075469971,
+      "learning_rate": 0.00010110506035084017,
+      "loss": 2.4776,
+      "step": 9980
+    },
+    {
+      "epoch": 0.8055039948349608,
+      "grad_norm": 0.6948177218437195,
+      "learning_rate": 0.00010108927440332306,
+      "loss": 2.5306,
+      "step": 9981
+    },
+    {
+      "epoch": 0.8055846985715439,
+      "grad_norm": 0.6873918771743774,
+      "learning_rate": 0.0001010734884286584,
+      "loss": 2.5783,
+      "step": 9982
+    },
+    {
+      "epoch": 0.8056654023081269,
+      "grad_norm": 0.6370649933815002,
+      "learning_rate": 0.00010105770242723958,
+      "loss": 2.5584,
+      "step": 9983
+    },
+    {
+      "epoch": 0.8057461060447099,
+      "grad_norm": 0.7594422698020935,
+      "learning_rate": 0.00010104191639946008,
+      "loss": 2.543,
+      "step": 9984
+    },
+    {
+      "epoch": 0.8058268097812928,
+      "grad_norm": 0.697380542755127,
+      "learning_rate": 0.00010102613034571327,
+      "loss": 2.5295,
+      "step": 9985
+    },
+    {
+      "epoch": 0.8059075135178759,
+      "grad_norm": 0.6597251892089844,
+      "learning_rate": 0.00010101034426639264,
+      "loss": 2.5917,
+      "step": 9986
+    },
+    {
+      "epoch": 0.8059882172544589,
+      "grad_norm": 0.6583479046821594,
+      "learning_rate": 0.00010099455816189156,
+      "loss": 2.6206,
+      "step": 9987
+    },
+    {
+      "epoch": 0.8060689209910419,
+      "grad_norm": 0.6603943705558777,
+      "learning_rate": 0.00010097877203260349,
+      "loss": 2.5223,
+      "step": 9988
+    },
+    {
+      "epoch": 0.8061496247276249,
+      "grad_norm": 0.716454267501831,
+      "learning_rate": 0.00010096298587892188,
+      "loss": 2.5572,
+      "step": 9989
+    },
+    {
+      "epoch": 0.806230328464208,
+      "grad_norm": 0.6511488556861877,
+      "learning_rate": 0.00010094719970124016,
+      "loss": 2.5815,
+      "step": 9990
+    },
+    {
+      "epoch": 0.8063110322007909,
+      "grad_norm": 0.6969261169433594,
+      "learning_rate": 0.00010093141349995173,
+      "loss": 2.5902,
+      "step": 9991
+    },
+    {
+      "epoch": 0.8063917359373739,
+      "grad_norm": 0.7012695074081421,
+      "learning_rate": 0.00010091562727545001,
+      "loss": 2.5134,
+      "step": 9992
+    },
+    {
+      "epoch": 0.8064724396739569,
+      "grad_norm": 0.6368406414985657,
+      "learning_rate": 0.00010089984102812848,
+      "loss": 2.568,
+      "step": 9993
+    },
+    {
+      "epoch": 0.80655314341054,
+      "grad_norm": 0.6552153825759888,
+      "learning_rate": 0.00010088405475838059,
+      "loss": 2.5101,
+      "step": 9994
+    },
+    {
+      "epoch": 0.8066338471471229,
+      "grad_norm": 0.6949633359909058,
+      "learning_rate": 0.00010086826846659974,
+      "loss": 2.5427,
+      "step": 9995
+    },
+    {
+      "epoch": 0.8067145508837059,
+      "grad_norm": 0.6593093872070312,
+      "learning_rate": 0.00010085248215317935,
+      "loss": 2.5551,
+      "step": 9996
+    },
+    {
+      "epoch": 0.8067952546202889,
+      "grad_norm": 0.6963745355606079,
+      "learning_rate": 0.00010083669581851287,
+      "loss": 2.4956,
+      "step": 9997
+    },
+    {
+      "epoch": 0.8068759583568719,
+      "grad_norm": 0.7093523144721985,
+      "learning_rate": 0.00010082090946299377,
+      "loss": 2.5876,
+      "step": 9998
+    },
+    {
+      "epoch": 0.806956662093455,
+      "grad_norm": 0.6796671152114868,
+      "learning_rate": 0.00010080512308701544,
+      "loss": 2.5302,
+      "step": 9999
+    },
+    {
+      "epoch": 0.8070373658300379,
+      "grad_norm": 0.7170542478561401,
+      "learning_rate": 0.00010078933669097135,
+      "loss": 2.5886,
+      "step": 10000
+    },
+    {
+      "epoch": 0.8070373658300379,
+      "eval_loss": 2.4734926223754883,
+      "eval_runtime": 788.2594,
+      "eval_samples_per_second": 3.324,
+      "eval_steps_per_second": 0.554,
+      "step": 10000
+    },
+    {
+      "epoch": 0.8071180695666209,
+      "grad_norm": 0.6566126346588135,
+      "learning_rate": 0.0001007735502752549,
+      "loss": 2.4441,
+      "step": 10001
+    },
+    {
+      "epoch": 0.8071987733032039,
+      "grad_norm": 0.6739515662193298,
+      "learning_rate": 0.00010075776384025957,
+      "loss": 2.5767,
+      "step": 10002
+    },
+    {
+      "epoch": 0.807279477039787,
+      "grad_norm": 0.6334208846092224,
+      "learning_rate": 0.00010074197738637881,
+      "loss": 2.5321,
+      "step": 10003
+    },
+    {
+      "epoch": 0.80736018077637,
+      "grad_norm": 0.6764520406723022,
+      "learning_rate": 0.000100726190914006,
+      "loss": 2.5144,
+      "step": 10004
+    },
+    {
+      "epoch": 0.8074408845129529,
+      "grad_norm": 0.7090082764625549,
+      "learning_rate": 0.00010071040442353464,
+      "loss": 2.5626,
+      "step": 10005
+    },
+    {
+      "epoch": 0.8075215882495359,
+      "grad_norm": 0.6915304064750671,
+      "learning_rate": 0.00010069461791535814,
+      "loss": 2.5261,
+      "step": 10006
+    },
+    {
+      "epoch": 0.807602291986119,
+      "grad_norm": 0.6685747504234314,
+      "learning_rate": 0.00010067883138986991,
+      "loss": 2.492,
+      "step": 10007
+    },
+    {
+      "epoch": 0.807682995722702,
+      "grad_norm": 0.7179074883460999,
+      "learning_rate": 0.00010066304484746347,
+      "loss": 2.4601,
+      "step": 10008
+    },
+    {
+      "epoch": 0.807763699459285,
+      "grad_norm": 0.7032761573791504,
+      "learning_rate": 0.00010064725828853219,
+      "loss": 2.578,
+      "step": 10009
+    },
+    {
+      "epoch": 0.8078444031958679,
+      "grad_norm": 0.710322916507721,
+      "learning_rate": 0.00010063147171346959,
+      "loss": 2.5514,
+      "step": 10010
+    },
+    {
+      "epoch": 0.807925106932451,
+      "grad_norm": 0.6552841067314148,
+      "learning_rate": 0.00010061568512266903,
+      "loss": 2.5474,
+      "step": 10011
+    },
+    {
+      "epoch": 0.808005810669034,
+      "grad_norm": 0.6862452626228333,
+      "learning_rate": 0.00010059989851652398,
+      "loss": 2.5772,
+      "step": 10012
+    },
+    {
+      "epoch": 0.808086514405617,
+      "grad_norm": 0.7123851180076599,
+      "learning_rate": 0.00010058411189542788,
+      "loss": 2.4936,
+      "step": 10013
+    },
+    {
+      "epoch": 0.8081672181421999,
+      "grad_norm": 0.6889944672584534,
+      "learning_rate": 0.00010056832525977422,
+      "loss": 2.5041,
+      "step": 10014
+    },
+    {
+      "epoch": 0.808247921878783,
+      "grad_norm": 0.6986924409866333,
+      "learning_rate": 0.0001005525386099564,
+      "loss": 2.5591,
+      "step": 10015
+    },
+    {
+      "epoch": 0.808328625615366,
+      "grad_norm": 0.6935306787490845,
+      "learning_rate": 0.00010053675194636787,
+      "loss": 2.5423,
+      "step": 10016
+    },
+    {
+      "epoch": 0.808409329351949,
+      "grad_norm": 0.6751969456672668,
+      "learning_rate": 0.00010052096526940207,
+      "loss": 2.5666,
+      "step": 10017
+    },
+    {
+      "epoch": 0.808490033088532,
+      "grad_norm": 0.676909327507019,
+      "learning_rate": 0.00010050517857945243,
+      "loss": 2.5394,
+      "step": 10018
+    },
+    {
+      "epoch": 0.808570736825115,
+      "grad_norm": 0.7439377307891846,
+      "learning_rate": 0.00010048939187691246,
+      "loss": 2.5011,
+      "step": 10019
+    },
+    {
+      "epoch": 0.808651440561698,
+      "grad_norm": 0.6594791412353516,
+      "learning_rate": 0.00010047360516217554,
+      "loss": 2.5159,
+      "step": 10020
+    },
+    {
+      "epoch": 0.808732144298281,
+      "grad_norm": 0.7013304233551025,
+      "learning_rate": 0.00010045781843563517,
+      "loss": 2.5439,
+      "step": 10021
+    },
+    {
+      "epoch": 0.808812848034864,
+      "grad_norm": 0.7537491917610168,
+      "learning_rate": 0.00010044203169768476,
+      "loss": 2.5837,
+      "step": 10022
+    },
+    {
+      "epoch": 0.8088935517714471,
+      "grad_norm": 0.7273866534233093,
+      "learning_rate": 0.00010042624494871773,
+      "loss": 2.5546,
+      "step": 10023
+    },
+    {
+      "epoch": 0.80897425550803,
+      "grad_norm": 0.6716369986534119,
+      "learning_rate": 0.0001004104581891276,
+      "loss": 2.5264,
+      "step": 10024
+    },
+    {
+      "epoch": 0.809054959244613,
+      "grad_norm": 0.7544769644737244,
+      "learning_rate": 0.00010039467141930777,
+      "loss": 2.5502,
+      "step": 10025
+    },
+    {
+      "epoch": 0.809135662981196,
+      "grad_norm": 0.8713179230690002,
+      "learning_rate": 0.0001003788846396517,
+      "loss": 2.5178,
+      "step": 10026
+    },
+    {
+      "epoch": 0.8092163667177791,
+      "grad_norm": 0.6704887747764587,
+      "learning_rate": 0.00010036309785055283,
+      "loss": 2.5136,
+      "step": 10027
+    },
+    {
+      "epoch": 0.809297070454362,
+      "grad_norm": 0.7308552861213684,
+      "learning_rate": 0.00010034731105240458,
+      "loss": 2.4781,
+      "step": 10028
+    },
+    {
+      "epoch": 0.809377774190945,
+      "grad_norm": 0.7214144468307495,
+      "learning_rate": 0.00010033152424560049,
+      "loss": 2.5946,
+      "step": 10029
+    },
+    {
+      "epoch": 0.809458477927528,
+      "grad_norm": 0.6946821808815002,
+      "learning_rate": 0.00010031573743053393,
+      "loss": 2.4937,
+      "step": 10030
+    },
+    {
+      "epoch": 0.8095391816641111,
+      "grad_norm": 0.7348416447639465,
+      "learning_rate": 0.00010029995060759833,
+      "loss": 2.5959,
+      "step": 10031
+    },
+    {
+      "epoch": 0.8096198854006941,
+      "grad_norm": 0.7482579350471497,
+      "learning_rate": 0.00010028416377718721,
+      "loss": 2.6,
+      "step": 10032
+    },
+    {
+      "epoch": 0.809700589137277,
+      "grad_norm": 0.7114939093589783,
+      "learning_rate": 0.00010026837693969397,
+      "loss": 2.5376,
+      "step": 10033
+    },
+    {
+      "epoch": 0.80978129287386,
+      "grad_norm": 0.6559228897094727,
+      "learning_rate": 0.00010025259009551209,
+      "loss": 2.4961,
+      "step": 10034
+    },
+    {
+      "epoch": 0.8098619966104431,
+      "grad_norm": 0.7494906187057495,
+      "learning_rate": 0.00010023680324503501,
+      "loss": 2.5723,
+      "step": 10035
+    },
+    {
+      "epoch": 0.8099427003470261,
+      "grad_norm": 0.7207093834877014,
+      "learning_rate": 0.00010022101638865618,
+      "loss": 2.5523,
+      "step": 10036
+    },
+    {
+      "epoch": 0.8100234040836091,
+      "grad_norm": 0.6730504035949707,
+      "learning_rate": 0.00010020522952676903,
+      "loss": 2.5135,
+      "step": 10037
+    },
+    {
+      "epoch": 0.810104107820192,
+      "grad_norm": 0.6805168390274048,
+      "learning_rate": 0.000100189442659767,
+      "loss": 2.5598,
+      "step": 10038
+    },
+    {
+      "epoch": 0.8101848115567751,
+      "grad_norm": 0.6639137268066406,
+      "learning_rate": 0.00010017365578804358,
+      "loss": 2.5152,
+      "step": 10039
+    },
+    {
+      "epoch": 0.8102655152933581,
+      "grad_norm": 0.6604194641113281,
+      "learning_rate": 0.00010015786891199221,
+      "loss": 2.5302,
+      "step": 10040
+    },
+    {
+      "epoch": 0.8103462190299411,
+      "grad_norm": 0.7664934992790222,
+      "learning_rate": 0.00010014208203200634,
+      "loss": 2.5437,
+      "step": 10041
+    },
+    {
+      "epoch": 0.8104269227665241,
+      "grad_norm": 0.7404079437255859,
+      "learning_rate": 0.00010012629514847942,
+      "loss": 2.6559,
+      "step": 10042
+    },
+    {
+      "epoch": 0.8105076265031071,
+      "grad_norm": 0.694006085395813,
+      "learning_rate": 0.00010011050826180488,
+      "loss": 2.5571,
+      "step": 10043
+    },
+    {
+      "epoch": 0.8105883302396901,
+      "grad_norm": 0.7007058262825012,
+      "learning_rate": 0.00010009472137237616,
+      "loss": 2.5639,
+      "step": 10044
+    },
+    {
+      "epoch": 0.8106690339762731,
+      "grad_norm": 0.7331913113594055,
+      "learning_rate": 0.00010007893448058678,
+      "loss": 2.5499,
+      "step": 10045
+    },
+    {
+      "epoch": 0.8107497377128561,
+      "grad_norm": 0.7636487483978271,
+      "learning_rate": 0.00010006314758683015,
+      "loss": 2.6068,
+      "step": 10046
+    },
+    {
+      "epoch": 0.810830441449439,
+      "grad_norm": 0.6505223512649536,
+      "learning_rate": 0.0001000473606914997,
+      "loss": 2.5313,
+      "step": 10047
+    },
+    {
+      "epoch": 0.8109111451860221,
+      "grad_norm": 0.6425966620445251,
+      "learning_rate": 0.00010003157379498886,
+      "loss": 2.5998,
+      "step": 10048
+    },
+    {
+      "epoch": 0.8109918489226051,
+      "grad_norm": 0.7163281440734863,
+      "learning_rate": 0.00010001578689769116,
+      "loss": 2.5493,
+      "step": 10049
+    },
+    {
+      "epoch": 0.8110725526591881,
+      "grad_norm": 0.7345306873321533,
+      "learning_rate": 0.0001,
+      "loss": 2.5609,
+      "step": 10050
+    },
+    {
+      "epoch": 0.8111532563957711,
+      "grad_norm": 0.6808427572250366,
+      "learning_rate": 9.998421310230884e-05,
+      "loss": 2.4823,
+      "step": 10051
+    },
+    {
+      "epoch": 0.8112339601323542,
+      "grad_norm": 0.7456082105636597,
+      "learning_rate": 9.996842620501115e-05,
+      "loss": 2.4782,
+      "step": 10052
+    },
+    {
+      "epoch": 0.8113146638689371,
+      "grad_norm": 0.7061728239059448,
+      "learning_rate": 9.995263930850034e-05,
+      "loss": 2.4906,
+      "step": 10053
+    },
+    {
+      "epoch": 0.8113953676055201,
+      "grad_norm": 0.691663920879364,
+      "learning_rate": 9.993685241316986e-05,
+      "loss": 2.5842,
+      "step": 10054
+    },
+    {
+      "epoch": 0.8114760713421031,
+      "grad_norm": 0.6899400353431702,
+      "learning_rate": 9.992106551941325e-05,
+      "loss": 2.5628,
+      "step": 10055
+    },
+    {
+      "epoch": 0.8115567750786862,
+      "grad_norm": 0.6909289360046387,
+      "learning_rate": 9.990527862762385e-05,
+      "loss": 2.5173,
+      "step": 10056
+    },
+    {
+      "epoch": 0.8116374788152692,
+      "grad_norm": 0.6507968306541443,
+      "learning_rate": 9.988949173819514e-05,
+      "loss": 2.5763,
+      "step": 10057
+    },
+    {
+      "epoch": 0.8117181825518521,
+      "grad_norm": 0.6972371339797974,
+      "learning_rate": 9.98737048515206e-05,
+      "loss": 2.604,
+      "step": 10058
+    },
+    {
+      "epoch": 0.8117988862884351,
+      "grad_norm": 0.6500107049942017,
+      "learning_rate": 9.985791796799368e-05,
+      "loss": 2.509,
+      "step": 10059
+    },
+    {
+      "epoch": 0.8118795900250182,
+      "grad_norm": 0.704501211643219,
+      "learning_rate": 9.98421310880078e-05,
+      "loss": 2.5773,
+      "step": 10060
+    },
+    {
+      "epoch": 0.8119602937616012,
+      "grad_norm": 0.7037203311920166,
+      "learning_rate": 9.982634421195641e-05,
+      "loss": 2.5968,
+      "step": 10061
+    },
+    {
+      "epoch": 0.8120409974981841,
+      "grad_norm": 0.7161232829093933,
+      "learning_rate": 9.981055734023304e-05,
+      "loss": 2.5373,
+      "step": 10062
+    },
+    {
+      "epoch": 0.8121217012347671,
+      "grad_norm": 0.6602928638458252,
+      "learning_rate": 9.979477047323099e-05,
+      "loss": 2.5851,
+      "step": 10063
+    },
+    {
+      "epoch": 0.8122024049713502,
+      "grad_norm": 0.6685947775840759,
+      "learning_rate": 9.977898361134383e-05,
+      "loss": 2.5543,
+      "step": 10064
+    },
+    {
+      "epoch": 0.8122831087079332,
+      "grad_norm": 0.6772760152816772,
+      "learning_rate": 9.976319675496502e-05,
+      "loss": 2.5355,
+      "step": 10065
+    },
+    {
+      "epoch": 0.8123638124445162,
+      "grad_norm": 0.6140885949134827,
+      "learning_rate": 9.974740990448792e-05,
+      "loss": 2.489,
+      "step": 10066
+    },
+    {
+      "epoch": 0.8124445161810991,
+      "grad_norm": 0.6597142219543457,
+      "learning_rate": 9.973162306030604e-05,
+      "loss": 2.5619,
+      "step": 10067
+    },
+    {
+      "epoch": 0.8125252199176822,
+      "grad_norm": 0.6768592000007629,
+      "learning_rate": 9.971583622281281e-05,
+      "loss": 2.5107,
+      "step": 10068
+    },
+    {
+      "epoch": 0.8126059236542652,
+      "grad_norm": 0.682296633720398,
+      "learning_rate": 9.970004939240168e-05,
+      "loss": 2.5003,
+      "step": 10069
+    },
+    {
+      "epoch": 0.8126866273908482,
+      "grad_norm": 0.7356325387954712,
+      "learning_rate": 9.96842625694661e-05,
+      "loss": 2.5864,
+      "step": 10070
+    },
+    {
+      "epoch": 0.8127673311274312,
+      "grad_norm": 0.6818091869354248,
+      "learning_rate": 9.966847575439956e-05,
+      "loss": 2.5375,
+      "step": 10071
+    },
+    {
+      "epoch": 0.8128480348640142,
+      "grad_norm": 0.6954368352890015,
+      "learning_rate": 9.965268894759543e-05,
+      "loss": 2.5314,
+      "step": 10072
+    },
+    {
+      "epoch": 0.8129287386005972,
+      "grad_norm": 0.6759306192398071,
+      "learning_rate": 9.963690214944721e-05,
+      "loss": 2.5881,
+      "step": 10073
+    },
+    {
+      "epoch": 0.8130094423371802,
+      "grad_norm": 0.6546545624732971,
+      "learning_rate": 9.962111536034832e-05,
+      "loss": 2.5264,
+      "step": 10074
+    },
+    {
+      "epoch": 0.8130901460737632,
+      "grad_norm": 0.6709586977958679,
+      "learning_rate": 9.960532858069226e-05,
+      "loss": 2.5906,
+      "step": 10075
+    },
+    {
+      "epoch": 0.8131708498103463,
+      "grad_norm": 0.7310851812362671,
+      "learning_rate": 9.958954181087241e-05,
+      "loss": 2.5134,
+      "step": 10076
+    },
+    {
+      "epoch": 0.8132515535469292,
+      "grad_norm": 0.6793027520179749,
+      "learning_rate": 9.957375505128227e-05,
+      "loss": 2.5387,
+      "step": 10077
+    },
+    {
+      "epoch": 0.8133322572835122,
+      "grad_norm": 0.6965875029563904,
+      "learning_rate": 9.955796830231528e-05,
+      "loss": 2.5649,
+      "step": 10078
+    },
+    {
+      "epoch": 0.8134129610200952,
+      "grad_norm": 0.6597574353218079,
+      "learning_rate": 9.954218156436485e-05,
+      "loss": 2.5281,
+      "step": 10079
+    },
+    {
+      "epoch": 0.8134936647566783,
+      "grad_norm": 0.7911555171012878,
+      "learning_rate": 9.952639483782445e-05,
+      "loss": 2.535,
+      "step": 10080
+    },
+    {
+      "epoch": 0.8135743684932613,
+      "grad_norm": 0.7405688762664795,
+      "learning_rate": 9.951060812308757e-05,
+      "loss": 2.5303,
+      "step": 10081
+    },
+    {
+      "epoch": 0.8136550722298442,
+      "grad_norm": 0.6961480379104614,
+      "learning_rate": 9.949482142054758e-05,
+      "loss": 2.4959,
+      "step": 10082
+    },
+    {
+      "epoch": 0.8137357759664272,
+      "grad_norm": 0.6761718392372131,
+      "learning_rate": 9.947903473059797e-05,
+      "loss": 2.5591,
+      "step": 10083
+    },
+    {
+      "epoch": 0.8138164797030103,
+      "grad_norm": 0.7383104562759399,
+      "learning_rate": 9.946324805363218e-05,
+      "loss": 2.5848,
+      "step": 10084
+    },
+    {
+      "epoch": 0.8138971834395933,
+      "grad_norm": 0.6495873928070068,
+      "learning_rate": 9.944746139004364e-05,
+      "loss": 2.4972,
+      "step": 10085
+    },
+    {
+      "epoch": 0.8139778871761763,
+      "grad_norm": 0.7247152328491211,
+      "learning_rate": 9.94316747402258e-05,
+      "loss": 2.5361,
+      "step": 10086
+    },
+    {
+      "epoch": 0.8140585909127592,
+      "grad_norm": 0.6965751051902771,
+      "learning_rate": 9.941588810457215e-05,
+      "loss": 2.4997,
+      "step": 10087
+    },
+    {
+      "epoch": 0.8141392946493423,
+      "grad_norm": 0.7138223648071289,
+      "learning_rate": 9.940010148347603e-05,
+      "loss": 2.5226,
+      "step": 10088
+    },
+    {
+      "epoch": 0.8142199983859253,
+      "grad_norm": 0.6571210622787476,
+      "learning_rate": 9.938431487733099e-05,
+      "loss": 2.5388,
+      "step": 10089
+    },
+    {
+      "epoch": 0.8143007021225083,
+      "grad_norm": 0.6721277832984924,
+      "learning_rate": 9.936852828653042e-05,
+      "loss": 2.5219,
+      "step": 10090
+    },
+    {
+      "epoch": 0.8143814058590912,
+      "grad_norm": 0.647520124912262,
+      "learning_rate": 9.935274171146782e-05,
+      "loss": 2.6199,
+      "step": 10091
+    },
+    {
+      "epoch": 0.8144621095956743,
+      "grad_norm": 0.6892204284667969,
+      "learning_rate": 9.933695515253654e-05,
+      "loss": 2.5132,
+      "step": 10092
+    },
+    {
+      "epoch": 0.8145428133322573,
+      "grad_norm": 0.6979050636291504,
+      "learning_rate": 9.932116861013008e-05,
+      "loss": 2.5148,
+      "step": 10093
+    },
+    {
+      "epoch": 0.8146235170688403,
+      "grad_norm": 0.6682664752006531,
+      "learning_rate": 9.930538208464189e-05,
+      "loss": 2.5795,
+      "step": 10094
+    },
+    {
+      "epoch": 0.8147042208054233,
+      "grad_norm": 0.734121561050415,
+      "learning_rate": 9.928959557646537e-05,
+      "loss": 2.5469,
+      "step": 10095
+    },
+    {
+      "epoch": 0.8147849245420064,
+      "grad_norm": 0.6669620275497437,
+      "learning_rate": 9.9273809085994e-05,
+      "loss": 2.5277,
+      "step": 10096
+    },
+    {
+      "epoch": 0.8148656282785893,
+      "grad_norm": 0.6750600934028625,
+      "learning_rate": 9.925802261362124e-05,
+      "loss": 2.5869,
+      "step": 10097
+    },
+    {
+      "epoch": 0.8149463320151723,
+      "grad_norm": 0.6813061237335205,
+      "learning_rate": 9.924223615974044e-05,
+      "loss": 2.585,
+      "step": 10098
+    },
+    {
+      "epoch": 0.8150270357517553,
+      "grad_norm": 0.6775497794151306,
+      "learning_rate": 9.92264497247451e-05,
+      "loss": 2.5353,
+      "step": 10099
+    },
+    {
+      "epoch": 0.8151077394883383,
+      "grad_norm": 0.6877530813217163,
+      "learning_rate": 9.92106633090287e-05,
+      "loss": 2.5349,
+      "step": 10100
+    },
+    {
+      "epoch": 0.8151884432249213,
+      "grad_norm": 0.6984169483184814,
+      "learning_rate": 9.91948769129846e-05,
+      "loss": 2.5986,
+      "step": 10101
+    },
+    {
+      "epoch": 0.8152691469615043,
+      "grad_norm": 0.7144806981086731,
+      "learning_rate": 9.917909053700626e-05,
+      "loss": 2.5797,
+      "step": 10102
+    },
+    {
+      "epoch": 0.8153498506980873,
+      "grad_norm": 0.6494203209877014,
+      "learning_rate": 9.916330418148715e-05,
+      "loss": 2.5035,
+      "step": 10103
+    },
+    {
+      "epoch": 0.8154305544346703,
+      "grad_norm": 0.6669752597808838,
+      "learning_rate": 9.914751784682069e-05,
+      "loss": 2.5489,
+      "step": 10104
+    },
+    {
+      "epoch": 0.8155112581712534,
+      "grad_norm": 0.6557981371879578,
+      "learning_rate": 9.913173153340029e-05,
+      "loss": 2.5266,
+      "step": 10105
+    },
+    {
+      "epoch": 0.8155919619078363,
+      "grad_norm": 0.6633948087692261,
+      "learning_rate": 9.911594524161941e-05,
+      "loss": 2.5263,
+      "step": 10106
+    },
+    {
+      "epoch": 0.8156726656444193,
+      "grad_norm": 0.7191522717475891,
+      "learning_rate": 9.910015897187154e-05,
+      "loss": 2.5625,
+      "step": 10107
+    },
+    {
+      "epoch": 0.8157533693810023,
+      "grad_norm": 0.7089062929153442,
+      "learning_rate": 9.908437272455001e-05,
+      "loss": 2.5644,
+      "step": 10108
+    },
+    {
+      "epoch": 0.8158340731175854,
+      "grad_norm": 0.7662761211395264,
+      "learning_rate": 9.906858650004831e-05,
+      "loss": 2.5875,
+      "step": 10109
+    },
+    {
+      "epoch": 0.8159147768541684,
+      "grad_norm": 0.6658861041069031,
+      "learning_rate": 9.905280029875988e-05,
+      "loss": 2.5818,
+      "step": 10110
+    },
+    {
+      "epoch": 0.8159954805907513,
+      "grad_norm": 0.7229514718055725,
+      "learning_rate": 9.903701412107815e-05,
+      "loss": 2.5421,
+      "step": 10111
+    },
+    {
+      "epoch": 0.8160761843273343,
+      "grad_norm": 0.7295149564743042,
+      "learning_rate": 9.902122796739652e-05,
+      "loss": 2.5298,
+      "step": 10112
+    },
+    {
+      "epoch": 0.8161568880639174,
+      "grad_norm": 0.6805420517921448,
+      "learning_rate": 9.900544183810849e-05,
+      "loss": 2.6693,
+      "step": 10113
+    },
+    {
+      "epoch": 0.8162375918005004,
+      "grad_norm": 0.6560602188110352,
+      "learning_rate": 9.898965573360738e-05,
+      "loss": 2.5445,
+      "step": 10114
+    },
+    {
+      "epoch": 0.8163182955370833,
+      "grad_norm": 0.690396785736084,
+      "learning_rate": 9.897386965428674e-05,
+      "loss": 2.5281,
+      "step": 10115
+    },
+    {
+      "epoch": 0.8163989992736663,
+      "grad_norm": 0.6905054450035095,
+      "learning_rate": 9.895808360053998e-05,
+      "loss": 2.5406,
+      "step": 10116
+    },
+    {
+      "epoch": 0.8164797030102494,
+      "grad_norm": 0.6905301213264465,
+      "learning_rate": 9.894229757276045e-05,
+      "loss": 2.5458,
+      "step": 10117
+    },
+    {
+      "epoch": 0.8165604067468324,
+      "grad_norm": 0.6827620267868042,
+      "learning_rate": 9.892651157134162e-05,
+      "loss": 2.4403,
+      "step": 10118
+    },
+    {
+      "epoch": 0.8166411104834154,
+      "grad_norm": 0.7614343166351318,
+      "learning_rate": 9.891072559667697e-05,
+      "loss": 2.6369,
+      "step": 10119
+    },
+    {
+      "epoch": 0.8167218142199983,
+      "grad_norm": 0.6913704872131348,
+      "learning_rate": 9.889493964915985e-05,
+      "loss": 2.5914,
+      "step": 10120
+    },
+    {
+      "epoch": 0.8168025179565814,
+      "grad_norm": 0.7026088237762451,
+      "learning_rate": 9.887915372918372e-05,
+      "loss": 2.5139,
+      "step": 10121
+    },
+    {
+      "epoch": 0.8168832216931644,
+      "grad_norm": 0.7064465284347534,
+      "learning_rate": 9.886336783714203e-05,
+      "loss": 2.549,
+      "step": 10122
+    },
+    {
+      "epoch": 0.8169639254297474,
+      "grad_norm": 0.7345553040504456,
+      "learning_rate": 9.884758197342821e-05,
+      "loss": 2.5887,
+      "step": 10123
+    },
+    {
+      "epoch": 0.8170446291663304,
+      "grad_norm": 0.6916251182556152,
+      "learning_rate": 9.883179613843563e-05,
+      "loss": 2.5659,
+      "step": 10124
+    },
+    {
+      "epoch": 0.8171253329029134,
+      "grad_norm": 0.6428200602531433,
+      "learning_rate": 9.881601033255771e-05,
+      "loss": 2.5379,
+      "step": 10125
+    },
+    {
+      "epoch": 0.8172060366394964,
+      "grad_norm": 0.7433571815490723,
+      "learning_rate": 9.880022455618796e-05,
+      "loss": 2.5751,
+      "step": 10126
+    },
+    {
+      "epoch": 0.8172867403760794,
+      "grad_norm": 0.733256995677948,
+      "learning_rate": 9.878443880971974e-05,
+      "loss": 2.4971,
+      "step": 10127
+    },
+    {
+      "epoch": 0.8173674441126624,
+      "grad_norm": 0.708289384841919,
+      "learning_rate": 9.876865309354646e-05,
+      "loss": 2.635,
+      "step": 10128
+    },
+    {
+      "epoch": 0.8174481478492455,
+      "grad_norm": 0.6877188682556152,
+      "learning_rate": 9.87528674080616e-05,
+      "loss": 2.5827,
+      "step": 10129
+    },
+    {
+      "epoch": 0.8175288515858284,
+      "grad_norm": 0.7108712792396545,
+      "learning_rate": 9.873708175365852e-05,
+      "loss": 2.5643,
+      "step": 10130
+    },
+    {
+      "epoch": 0.8176095553224114,
+      "grad_norm": 0.7435629367828369,
+      "learning_rate": 9.872129613073065e-05,
+      "loss": 2.5267,
+      "step": 10131
+    },
+    {
+      "epoch": 0.8176902590589944,
+      "grad_norm": 0.669913113117218,
+      "learning_rate": 9.870551053967148e-05,
+      "loss": 2.5684,
+      "step": 10132
+    },
+    {
+      "epoch": 0.8177709627955775,
+      "grad_norm": 0.6981424689292908,
+      "learning_rate": 9.868972498087431e-05,
+      "loss": 2.592,
+      "step": 10133
+    },
+    {
+      "epoch": 0.8178516665321605,
+      "grad_norm": 0.6661834716796875,
+      "learning_rate": 9.867393945473263e-05,
+      "loss": 2.5082,
+      "step": 10134
+    },
+    {
+      "epoch": 0.8179323702687434,
+      "grad_norm": 0.6611261367797852,
+      "learning_rate": 9.865815396163987e-05,
+      "loss": 2.556,
+      "step": 10135
+    },
+    {
+      "epoch": 0.8180130740053264,
+      "grad_norm": 0.6732283234596252,
+      "learning_rate": 9.86423685019894e-05,
+      "loss": 2.5668,
+      "step": 10136
+    },
+    {
+      "epoch": 0.8180937777419095,
+      "grad_norm": 0.6768637299537659,
+      "learning_rate": 9.862658307617465e-05,
+      "loss": 2.5467,
+      "step": 10137
+    },
+    {
+      "epoch": 0.8181744814784925,
+      "grad_norm": 0.6943596601486206,
+      "learning_rate": 9.861079768458904e-05,
+      "loss": 2.5989,
+      "step": 10138
+    },
+    {
+      "epoch": 0.8182551852150755,
+      "grad_norm": 0.7369638681411743,
+      "learning_rate": 9.859501232762601e-05,
+      "loss": 2.5189,
+      "step": 10139
+    },
+    {
+      "epoch": 0.8183358889516584,
+      "grad_norm": 0.7443112730979919,
+      "learning_rate": 9.857922700567892e-05,
+      "loss": 2.5979,
+      "step": 10140
+    },
+    {
+      "epoch": 0.8184165926882415,
+      "grad_norm": 0.6726163029670715,
+      "learning_rate": 9.85634417191412e-05,
+      "loss": 2.5451,
+      "step": 10141
+    },
+    {
+      "epoch": 0.8184972964248245,
+      "grad_norm": 0.720492422580719,
+      "learning_rate": 9.854765646840632e-05,
+      "loss": 2.6116,
+      "step": 10142
+    },
+    {
+      "epoch": 0.8185780001614075,
+      "grad_norm": 0.6998233795166016,
+      "learning_rate": 9.85318712538676e-05,
+      "loss": 2.556,
+      "step": 10143
+    },
+    {
+      "epoch": 0.8186587038979904,
+      "grad_norm": 0.7580110430717468,
+      "learning_rate": 9.851608607591848e-05,
+      "loss": 2.5222,
+      "step": 10144
+    },
+    {
+      "epoch": 0.8187394076345735,
+      "grad_norm": 0.6893007755279541,
+      "learning_rate": 9.85003009349524e-05,
+      "loss": 2.4639,
+      "step": 10145
+    },
+    {
+      "epoch": 0.8188201113711565,
+      "grad_norm": 0.6448441743850708,
+      "learning_rate": 9.84845158313627e-05,
+      "loss": 2.5249,
+      "step": 10146
+    },
+    {
+      "epoch": 0.8189008151077395,
+      "grad_norm": 0.7591872215270996,
+      "learning_rate": 9.846873076554285e-05,
+      "loss": 2.5173,
+      "step": 10147
+    },
+    {
+      "epoch": 0.8189815188443225,
+      "grad_norm": 0.6994685530662537,
+      "learning_rate": 9.845294573788626e-05,
+      "loss": 2.5181,
+      "step": 10148
+    },
+    {
+      "epoch": 0.8190622225809054,
+      "grad_norm": 0.6822378635406494,
+      "learning_rate": 9.843716074878628e-05,
+      "loss": 2.5109,
+      "step": 10149
+    },
+    {
+      "epoch": 0.8191429263174885,
+      "grad_norm": 0.6730359792709351,
+      "learning_rate": 9.842137579863632e-05,
+      "loss": 2.5402,
+      "step": 10150
+    },
+    {
+      "epoch": 0.8192236300540715,
+      "grad_norm": 0.6280627846717834,
+      "learning_rate": 9.840559088782984e-05,
+      "loss": 2.4806,
+      "step": 10151
+    },
+    {
+      "epoch": 0.8193043337906545,
+      "grad_norm": 0.6887876391410828,
+      "learning_rate": 9.838980601676017e-05,
+      "loss": 2.5498,
+      "step": 10152
+    },
+    {
+      "epoch": 0.8193850375272375,
+      "grad_norm": 0.7823790907859802,
+      "learning_rate": 9.837402118582075e-05,
+      "loss": 2.467,
+      "step": 10153
+    },
+    {
+      "epoch": 0.8194657412638205,
+      "grad_norm": 0.8109384179115295,
+      "learning_rate": 9.835823639540496e-05,
+      "loss": 2.5898,
+      "step": 10154
+    },
+    {
+      "epoch": 0.8195464450004035,
+      "grad_norm": 0.6883066892623901,
+      "learning_rate": 9.834245164590624e-05,
+      "loss": 2.5589,
+      "step": 10155
+    },
+    {
+      "epoch": 0.8196271487369865,
+      "grad_norm": 0.7291175723075867,
+      "learning_rate": 9.832666693771794e-05,
+      "loss": 2.5317,
+      "step": 10156
+    },
+    {
+      "epoch": 0.8197078524735695,
+      "grad_norm": 0.6819449663162231,
+      "learning_rate": 9.831088227123346e-05,
+      "loss": 2.5513,
+      "step": 10157
+    },
+    {
+      "epoch": 0.8197885562101526,
+      "grad_norm": 0.7038870453834534,
+      "learning_rate": 9.829509764684626e-05,
+      "loss": 2.5301,
+      "step": 10158
+    },
+    {
+      "epoch": 0.8198692599467355,
+      "grad_norm": 0.7483033537864685,
+      "learning_rate": 9.827931306494965e-05,
+      "loss": 2.5273,
+      "step": 10159
+    },
+    {
+      "epoch": 0.8199499636833185,
+      "grad_norm": 0.6998303532600403,
+      "learning_rate": 9.826352852593705e-05,
+      "loss": 2.5083,
+      "step": 10160
+    },
+    {
+      "epoch": 0.8200306674199015,
+      "grad_norm": 0.6865512728691101,
+      "learning_rate": 9.824774403020188e-05,
+      "loss": 2.5693,
+      "step": 10161
+    },
+    {
+      "epoch": 0.8201113711564846,
+      "grad_norm": 0.8144257068634033,
+      "learning_rate": 9.823195957813749e-05,
+      "loss": 2.6052,
+      "step": 10162
+    },
+    {
+      "epoch": 0.8201920748930676,
+      "grad_norm": 0.6920810341835022,
+      "learning_rate": 9.821617517013729e-05,
+      "loss": 2.5467,
+      "step": 10163
+    },
+    {
+      "epoch": 0.8202727786296505,
+      "grad_norm": 0.7538061141967773,
+      "learning_rate": 9.820039080659469e-05,
+      "loss": 2.5933,
+      "step": 10164
+    },
+    {
+      "epoch": 0.8203534823662335,
+      "grad_norm": 0.6744310259819031,
+      "learning_rate": 9.818460648790302e-05,
+      "loss": 2.5633,
+      "step": 10165
+    },
+    {
+      "epoch": 0.8204341861028166,
+      "grad_norm": 0.6943854689598083,
+      "learning_rate": 9.816882221445571e-05,
+      "loss": 2.5868,
+      "step": 10166
+    },
+    {
+      "epoch": 0.8205148898393996,
+      "grad_norm": 0.6486902832984924,
+      "learning_rate": 9.815303798664614e-05,
+      "loss": 2.4983,
+      "step": 10167
+    },
+    {
+      "epoch": 0.8205955935759826,
+      "grad_norm": 0.6699065566062927,
+      "learning_rate": 9.813725380486773e-05,
+      "loss": 2.563,
+      "step": 10168
+    },
+    {
+      "epoch": 0.8206762973125655,
+      "grad_norm": 0.6547110080718994,
+      "learning_rate": 9.812146966951379e-05,
+      "loss": 2.5404,
+      "step": 10169
+    },
+    {
+      "epoch": 0.8207570010491486,
+      "grad_norm": 0.692592203617096,
+      "learning_rate": 9.810568558097774e-05,
+      "loss": 2.5625,
+      "step": 10170
+    },
+    {
+      "epoch": 0.8208377047857316,
+      "grad_norm": 0.6696702837944031,
+      "learning_rate": 9.808990153965296e-05,
+      "loss": 2.5866,
+      "step": 10171
+    },
+    {
+      "epoch": 0.8209184085223146,
+      "grad_norm": 0.6425998210906982,
+      "learning_rate": 9.807411754593282e-05,
+      "loss": 2.5487,
+      "step": 10172
+    },
+    {
+      "epoch": 0.8209991122588975,
+      "grad_norm": 0.6849769949913025,
+      "learning_rate": 9.805833360021069e-05,
+      "loss": 2.5772,
+      "step": 10173
+    },
+    {
+      "epoch": 0.8210798159954806,
+      "grad_norm": 0.7451414465904236,
+      "learning_rate": 9.804254970288001e-05,
+      "loss": 2.5089,
+      "step": 10174
+    },
+    {
+      "epoch": 0.8211605197320636,
+      "grad_norm": 0.7134390473365784,
+      "learning_rate": 9.802676585433408e-05,
+      "loss": 2.541,
+      "step": 10175
+    },
+    {
+      "epoch": 0.8212412234686466,
+      "grad_norm": 0.7490564584732056,
+      "learning_rate": 9.801098205496627e-05,
+      "loss": 2.5299,
+      "step": 10176
+    },
+    {
+      "epoch": 0.8213219272052296,
+      "grad_norm": 0.6614408493041992,
+      "learning_rate": 9.799519830517005e-05,
+      "loss": 2.5252,
+      "step": 10177
+    },
+    {
+      "epoch": 0.8214026309418127,
+      "grad_norm": 0.761049211025238,
+      "learning_rate": 9.797941460533869e-05,
+      "loss": 2.5153,
+      "step": 10178
+    },
+    {
+      "epoch": 0.8214833346783956,
+      "grad_norm": 0.6352702379226685,
+      "learning_rate": 9.796363095586561e-05,
+      "loss": 2.5407,
+      "step": 10179
+    },
+    {
+      "epoch": 0.8215640384149786,
+      "grad_norm": 0.684212863445282,
+      "learning_rate": 9.794784735714417e-05,
+      "loss": 2.5425,
+      "step": 10180
+    },
+    {
+      "epoch": 0.8216447421515616,
+      "grad_norm": 0.652987539768219,
+      "learning_rate": 9.793206380956772e-05,
+      "loss": 2.5542,
+      "step": 10181
+    },
+    {
+      "epoch": 0.8217254458881447,
+      "grad_norm": 0.6912897229194641,
+      "learning_rate": 9.791628031352966e-05,
+      "loss": 2.5041,
+      "step": 10182
+    },
+    {
+      "epoch": 0.8218061496247276,
+      "grad_norm": 0.7025408744812012,
+      "learning_rate": 9.790049686942333e-05,
+      "loss": 2.5296,
+      "step": 10183
+    },
+    {
+      "epoch": 0.8218868533613106,
+      "grad_norm": 0.7580777406692505,
+      "learning_rate": 9.788471347764215e-05,
+      "loss": 2.578,
+      "step": 10184
+    },
+    {
+      "epoch": 0.8219675570978936,
+      "grad_norm": 0.7044378519058228,
+      "learning_rate": 9.78689301385794e-05,
+      "loss": 2.5093,
+      "step": 10185
+    },
+    {
+      "epoch": 0.8220482608344767,
+      "grad_norm": 0.7339754700660706,
+      "learning_rate": 9.785314685262849e-05,
+      "loss": 2.5202,
+      "step": 10186
+    },
+    {
+      "epoch": 0.8221289645710597,
+      "grad_norm": 0.6872244477272034,
+      "learning_rate": 9.783736362018277e-05,
+      "loss": 2.541,
+      "step": 10187
+    },
+    {
+      "epoch": 0.8222096683076426,
+      "grad_norm": 0.7052434682846069,
+      "learning_rate": 9.78215804416356e-05,
+      "loss": 2.4968,
+      "step": 10188
+    },
+    {
+      "epoch": 0.8222903720442256,
+      "grad_norm": 0.6739610433578491,
+      "learning_rate": 9.780579731738033e-05,
+      "loss": 2.5137,
+      "step": 10189
+    },
+    {
+      "epoch": 0.8223710757808087,
+      "grad_norm": 0.6842939853668213,
+      "learning_rate": 9.779001424781035e-05,
+      "loss": 2.5329,
+      "step": 10190
+    },
+    {
+      "epoch": 0.8224517795173917,
+      "grad_norm": 0.7057977914810181,
+      "learning_rate": 9.777423123331898e-05,
+      "loss": 2.5657,
+      "step": 10191
+    },
+    {
+      "epoch": 0.8225324832539747,
+      "grad_norm": 0.6748424172401428,
+      "learning_rate": 9.775844827429958e-05,
+      "loss": 2.6104,
+      "step": 10192
+    },
+    {
+      "epoch": 0.8226131869905576,
+      "grad_norm": 0.6492514610290527,
+      "learning_rate": 9.774266537114555e-05,
+      "loss": 2.58,
+      "step": 10193
+    },
+    {
+      "epoch": 0.8226938907271407,
+      "grad_norm": 0.6987641453742981,
+      "learning_rate": 9.772688252425016e-05,
+      "loss": 2.5301,
+      "step": 10194
+    },
+    {
+      "epoch": 0.8227745944637237,
+      "grad_norm": 0.710921585559845,
+      "learning_rate": 9.771109973400679e-05,
+      "loss": 2.6245,
+      "step": 10195
+    },
+    {
+      "epoch": 0.8228552982003067,
+      "grad_norm": 0.6673738360404968,
+      "learning_rate": 9.769531700080883e-05,
+      "loss": 2.5205,
+      "step": 10196
+    },
+    {
+      "epoch": 0.8229360019368896,
+      "grad_norm": 0.6705252528190613,
+      "learning_rate": 9.767953432504958e-05,
+      "loss": 2.4932,
+      "step": 10197
+    },
+    {
+      "epoch": 0.8230167056734727,
+      "grad_norm": 0.6587076783180237,
+      "learning_rate": 9.766375170712237e-05,
+      "loss": 2.5085,
+      "step": 10198
+    },
+    {
+      "epoch": 0.8230974094100557,
+      "grad_norm": 0.7285338640213013,
+      "learning_rate": 9.764796914742061e-05,
+      "loss": 2.5481,
+      "step": 10199
+    },
+    {
+      "epoch": 0.8231781131466387,
+      "grad_norm": 0.6971831321716309,
+      "learning_rate": 9.763218664633763e-05,
+      "loss": 2.6092,
+      "step": 10200
+    },
+    {
+      "epoch": 0.8232588168832217,
+      "grad_norm": 0.6940265893936157,
+      "learning_rate": 9.761640420426669e-05,
+      "loss": 2.5325,
+      "step": 10201
+    },
+    {
+      "epoch": 0.8233395206198046,
+      "grad_norm": 0.6612978577613831,
+      "learning_rate": 9.76006218216012e-05,
+      "loss": 2.5532,
+      "step": 10202
+    },
+    {
+      "epoch": 0.8234202243563877,
+      "grad_norm": 0.6707638502120972,
+      "learning_rate": 9.758483949873453e-05,
+      "loss": 2.512,
+      "step": 10203
+    },
+    {
+      "epoch": 0.8235009280929707,
+      "grad_norm": 0.6636764407157898,
+      "learning_rate": 9.756905723605994e-05,
+      "loss": 2.5446,
+      "step": 10204
+    },
+    {
+      "epoch": 0.8235816318295537,
+      "grad_norm": 0.6996643543243408,
+      "learning_rate": 9.755327503397081e-05,
+      "loss": 2.5504,
+      "step": 10205
+    },
+    {
+      "epoch": 0.8236623355661367,
+      "grad_norm": 0.604487955570221,
+      "learning_rate": 9.753749289286046e-05,
+      "loss": 2.4767,
+      "step": 10206
+    },
+    {
+      "epoch": 0.8237430393027197,
+      "grad_norm": 0.6484553217887878,
+      "learning_rate": 9.752171081312222e-05,
+      "loss": 2.5522,
+      "step": 10207
+    },
+    {
+      "epoch": 0.8238237430393027,
+      "grad_norm": 0.6890987753868103,
+      "learning_rate": 9.75059287951494e-05,
+      "loss": 2.5545,
+      "step": 10208
+    },
+    {
+      "epoch": 0.8239044467758857,
+      "grad_norm": 0.6786034107208252,
+      "learning_rate": 9.749014683933541e-05,
+      "loss": 2.591,
+      "step": 10209
+    },
+    {
+      "epoch": 0.8239851505124687,
+      "grad_norm": 0.751192033290863,
+      "learning_rate": 9.747436494607349e-05,
+      "loss": 2.5335,
+      "step": 10210
+    },
+    {
+      "epoch": 0.8240658542490518,
+      "grad_norm": 0.6611589789390564,
+      "learning_rate": 9.7458583115757e-05,
+      "loss": 2.5104,
+      "step": 10211
+    },
+    {
+      "epoch": 0.8241465579856347,
+      "grad_norm": 0.6602892875671387,
+      "learning_rate": 9.744280134877926e-05,
+      "loss": 2.5319,
+      "step": 10212
+    },
+    {
+      "epoch": 0.8242272617222177,
+      "grad_norm": 0.6856467127799988,
+      "learning_rate": 9.742701964553359e-05,
+      "loss": 2.5418,
+      "step": 10213
+    },
+    {
+      "epoch": 0.8243079654588007,
+      "grad_norm": 0.6810153126716614,
+      "learning_rate": 9.741123800641332e-05,
+      "loss": 2.5691,
+      "step": 10214
+    },
+    {
+      "epoch": 0.8243886691953838,
+      "grad_norm": 0.7044229507446289,
+      "learning_rate": 9.739545643181175e-05,
+      "loss": 2.5911,
+      "step": 10215
+    },
+    {
+      "epoch": 0.8244693729319668,
+      "grad_norm": 0.6689271330833435,
+      "learning_rate": 9.737967492212225e-05,
+      "loss": 2.5374,
+      "step": 10216
+    },
+    {
+      "epoch": 0.8245500766685497,
+      "grad_norm": 0.6558904051780701,
+      "learning_rate": 9.736389347773807e-05,
+      "loss": 2.5118,
+      "step": 10217
+    },
+    {
+      "epoch": 0.8246307804051327,
+      "grad_norm": 0.6900291442871094,
+      "learning_rate": 9.734811209905255e-05,
+      "loss": 2.515,
+      "step": 10218
+    },
+    {
+      "epoch": 0.8247114841417158,
+      "grad_norm": 0.7129492163658142,
+      "learning_rate": 9.733233078645907e-05,
+      "loss": 2.5191,
+      "step": 10219
+    },
+    {
+      "epoch": 0.8247921878782988,
+      "grad_norm": 0.7031866908073425,
+      "learning_rate": 9.731654954035082e-05,
+      "loss": 2.5616,
+      "step": 10220
+    },
+    {
+      "epoch": 0.8248728916148818,
+      "grad_norm": 0.6418820023536682,
+      "learning_rate": 9.730076836112118e-05,
+      "loss": 2.537,
+      "step": 10221
+    },
+    {
+      "epoch": 0.8249535953514647,
+      "grad_norm": 0.6731035113334656,
+      "learning_rate": 9.728498724916347e-05,
+      "loss": 2.5483,
+      "step": 10222
+    },
+    {
+      "epoch": 0.8250342990880478,
+      "grad_norm": 0.6941342353820801,
+      "learning_rate": 9.726920620487096e-05,
+      "loss": 2.5314,
+      "step": 10223
+    },
+    {
+      "epoch": 0.8251150028246308,
+      "grad_norm": 0.6808927059173584,
+      "learning_rate": 9.725342522863696e-05,
+      "loss": 2.5521,
+      "step": 10224
+    },
+    {
+      "epoch": 0.8251957065612138,
+      "grad_norm": 0.6873155832290649,
+      "learning_rate": 9.723764432085481e-05,
+      "loss": 2.5205,
+      "step": 10225
+    },
+    {
+      "epoch": 0.8252764102977967,
+      "grad_norm": 0.8590287566184998,
+      "learning_rate": 9.722186348191776e-05,
+      "loss": 2.5378,
+      "step": 10226
+    },
+    {
+      "epoch": 0.8253571140343798,
+      "grad_norm": 0.691523015499115,
+      "learning_rate": 9.720608271221912e-05,
+      "loss": 2.5062,
+      "step": 10227
+    },
+    {
+      "epoch": 0.8254378177709628,
+      "grad_norm": 0.6695523262023926,
+      "learning_rate": 9.719030201215226e-05,
+      "loss": 2.5164,
+      "step": 10228
+    },
+    {
+      "epoch": 0.8255185215075458,
+      "grad_norm": 0.745516300201416,
+      "learning_rate": 9.717452138211037e-05,
+      "loss": 2.5207,
+      "step": 10229
+    },
+    {
+      "epoch": 0.8255992252441288,
+      "grad_norm": 0.6628115773200989,
+      "learning_rate": 9.715874082248679e-05,
+      "loss": 2.5293,
+      "step": 10230
+    },
+    {
+      "epoch": 0.8256799289807119,
+      "grad_norm": 0.6531884074211121,
+      "learning_rate": 9.714296033367482e-05,
+      "loss": 2.4812,
+      "step": 10231
+    },
+    {
+      "epoch": 0.8257606327172948,
+      "grad_norm": 0.7444833517074585,
+      "learning_rate": 9.712717991606777e-05,
+      "loss": 2.5422,
+      "step": 10232
+    },
+    {
+      "epoch": 0.8258413364538778,
+      "grad_norm": 0.7013139128684998,
+      "learning_rate": 9.711139957005888e-05,
+      "loss": 2.5117,
+      "step": 10233
+    },
+    {
+      "epoch": 0.8259220401904608,
+      "grad_norm": 0.6588132977485657,
+      "learning_rate": 9.709561929604147e-05,
+      "loss": 2.5257,
+      "step": 10234
+    },
+    {
+      "epoch": 0.8260027439270439,
+      "grad_norm": 0.7538537383079529,
+      "learning_rate": 9.707983909440886e-05,
+      "loss": 2.5225,
+      "step": 10235
+    },
+    {
+      "epoch": 0.8260834476636268,
+      "grad_norm": Infinity,
+      "learning_rate": 9.707983909440886e-05,
+      "loss": 2.5532,
+      "step": 10236
+    },
+    {
+      "epoch": 0.8261641514002098,
+      "grad_norm": 0.7414929270744324,
+      "learning_rate": 9.706405896555425e-05,
+      "loss": 2.5653,
+      "step": 10237
+    },
+    {
+      "epoch": 0.8262448551367928,
+      "grad_norm": 0.757057785987854,
+      "learning_rate": 9.704827890987097e-05,
+      "loss": 2.5732,
+      "step": 10238
+    },
+    {
+      "epoch": 0.8263255588733759,
+      "grad_norm": 0.730721652507782,
+      "learning_rate": 9.703249892775232e-05,
+      "loss": 2.5317,
+      "step": 10239
+    },
+    {
+      "epoch": 0.8264062626099589,
+      "grad_norm": 0.6943208575248718,
+      "learning_rate": 9.701671901959151e-05,
+      "loss": 2.5849,
+      "step": 10240
+    },
+    {
+      "epoch": 0.8264869663465418,
+      "grad_norm": 0.7111102938652039,
+      "learning_rate": 9.700093918578188e-05,
+      "loss": 2.5007,
+      "step": 10241
+    },
+    {
+      "epoch": 0.8265676700831248,
+      "grad_norm": 0.7240251302719116,
+      "learning_rate": 9.69851594267167e-05,
+      "loss": 2.5002,
+      "step": 10242
+    },
+    {
+      "epoch": 0.8266483738197079,
+      "grad_norm": 0.6624411344528198,
+      "learning_rate": 9.696937974278922e-05,
+      "loss": 2.5175,
+      "step": 10243
+    },
+    {
+      "epoch": 0.8267290775562909,
+      "grad_norm": 0.6972576975822449,
+      "learning_rate": 9.695360013439269e-05,
+      "loss": 2.5285,
+      "step": 10244
+    },
+    {
+      "epoch": 0.8268097812928739,
+      "grad_norm": 0.684446394443512,
+      "learning_rate": 9.693782060192046e-05,
+      "loss": 2.57,
+      "step": 10245
+    },
+    {
+      "epoch": 0.8268904850294568,
+      "grad_norm": 0.6920011639595032,
+      "learning_rate": 9.692204114576573e-05,
+      "loss": 2.5042,
+      "step": 10246
+    },
+    {
+      "epoch": 0.8269711887660399,
+      "grad_norm": 0.7526013851165771,
+      "learning_rate": 9.690626176632176e-05,
+      "loss": 2.5878,
+      "step": 10247
+    },
+    {
+      "epoch": 0.8270518925026229,
+      "grad_norm": 0.6936177611351013,
+      "learning_rate": 9.689048246398184e-05,
+      "loss": 2.5572,
+      "step": 10248
+    },
+    {
+      "epoch": 0.8271325962392059,
+      "grad_norm": 0.672168493270874,
+      "learning_rate": 9.687470323913922e-05,
+      "loss": 2.5127,
+      "step": 10249
+    },
+    {
+      "epoch": 0.8272132999757889,
+      "grad_norm": 0.6847899556159973,
+      "learning_rate": 9.685892409218717e-05,
+      "loss": 2.5443,
+      "step": 10250
+    },
+    {
+      "epoch": 0.8272940037123718,
+      "grad_norm": 0.6877103447914124,
+      "learning_rate": 9.684314502351894e-05,
+      "loss": 2.4924,
+      "step": 10251
+    },
+    {
+      "epoch": 0.8273747074489549,
+      "grad_norm": 0.6894243359565735,
+      "learning_rate": 9.682736603352783e-05,
+      "loss": 2.5107,
+      "step": 10252
+    },
+    {
+      "epoch": 0.8274554111855379,
+      "grad_norm": 0.7318278551101685,
+      "learning_rate": 9.681158712260698e-05,
+      "loss": 2.5276,
+      "step": 10253
+    },
+    {
+      "epoch": 0.8275361149221209,
+      "grad_norm": 0.6949039101600647,
+      "learning_rate": 9.679580829114975e-05,
+      "loss": 2.5128,
+      "step": 10254
+    },
+    {
+      "epoch": 0.8276168186587038,
+      "grad_norm": 0.6523800492286682,
+      "learning_rate": 9.678002953954939e-05,
+      "loss": 2.5584,
+      "step": 10255
+    },
+    {
+      "epoch": 0.8276975223952869,
+      "grad_norm": 0.6914480328559875,
+      "learning_rate": 9.676425086819905e-05,
+      "loss": 2.5597,
+      "step": 10256
+    },
+    {
+      "epoch": 0.8277782261318699,
+      "grad_norm": 0.7107869982719421,
+      "learning_rate": 9.674847227749206e-05,
+      "loss": 2.5009,
+      "step": 10257
+    },
+    {
+      "epoch": 0.8278589298684529,
+      "grad_norm": 0.7066758275032043,
+      "learning_rate": 9.673269376782166e-05,
+      "loss": 2.4599,
+      "step": 10258
+    },
+    {
+      "epoch": 0.8279396336050359,
+      "grad_norm": 0.7147037982940674,
+      "learning_rate": 9.671691533958104e-05,
+      "loss": 2.4478,
+      "step": 10259
+    },
+    {
+      "epoch": 0.828020337341619,
+      "grad_norm": 0.666265606880188,
+      "learning_rate": 9.670113699316347e-05,
+      "loss": 2.5652,
+      "step": 10260
+    },
+    {
+      "epoch": 0.8281010410782019,
+      "grad_norm": 0.7026315927505493,
+      "learning_rate": 9.668535872896225e-05,
+      "loss": 2.5397,
+      "step": 10261
+    },
+    {
+      "epoch": 0.8281817448147849,
+      "grad_norm": 0.6611438393592834,
+      "learning_rate": 9.66695805473705e-05,
+      "loss": 2.5628,
+      "step": 10262
+    },
+    {
+      "epoch": 0.8282624485513679,
+      "grad_norm": 0.7211201190948486,
+      "learning_rate": 9.66538024487815e-05,
+      "loss": 2.5551,
+      "step": 10263
+    },
+    {
+      "epoch": 0.828343152287951,
+      "grad_norm": 0.7224553227424622,
+      "learning_rate": 9.663802443358849e-05,
+      "loss": 2.5329,
+      "step": 10264
+    },
+    {
+      "epoch": 0.8284238560245339,
+      "grad_norm": 0.6805843710899353,
+      "learning_rate": 9.662224650218474e-05,
+      "loss": 2.5744,
+      "step": 10265
+    },
+    {
+      "epoch": 0.8285045597611169,
+      "grad_norm": 0.7101335525512695,
+      "learning_rate": 9.66064686549634e-05,
+      "loss": 2.5281,
+      "step": 10266
+    },
+    {
+      "epoch": 0.8285852634976999,
+      "grad_norm": 0.7208443284034729,
+      "learning_rate": 9.659069089231774e-05,
+      "loss": 2.5326,
+      "step": 10267
+    },
+    {
+      "epoch": 0.828665967234283,
+      "grad_norm": 0.747894287109375,
+      "learning_rate": 9.6574913214641e-05,
+      "loss": 2.4909,
+      "step": 10268
+    },
+    {
+      "epoch": 0.828746670970866,
+      "grad_norm": 0.6618027091026306,
+      "learning_rate": 9.655913562232635e-05,
+      "loss": 2.6091,
+      "step": 10269
+    },
+    {
+      "epoch": 0.8288273747074489,
+      "grad_norm": 0.7101535201072693,
+      "learning_rate": 9.654335811576704e-05,
+      "loss": 2.5194,
+      "step": 10270
+    },
+    {
+      "epoch": 0.8289080784440319,
+      "grad_norm": 0.727763831615448,
+      "learning_rate": 9.652758069535631e-05,
+      "loss": 2.5767,
+      "step": 10271
+    },
+    {
+      "epoch": 0.828988782180615,
+      "grad_norm": 0.6936737895011902,
+      "learning_rate": 9.65118033614873e-05,
+      "loss": 2.498,
+      "step": 10272
+    },
+    {
+      "epoch": 0.829069485917198,
+      "grad_norm": 0.699462354183197,
+      "learning_rate": 9.64960261145533e-05,
+      "loss": 2.5033,
+      "step": 10273
+    },
+    {
+      "epoch": 0.829150189653781,
+      "grad_norm": 0.7024868726730347,
+      "learning_rate": 9.648024895494749e-05,
+      "loss": 2.5937,
+      "step": 10274
+    },
+    {
+      "epoch": 0.8292308933903639,
+      "grad_norm": 0.7028421759605408,
+      "learning_rate": 9.646447188306305e-05,
+      "loss": 2.5528,
+      "step": 10275
+    },
+    {
+      "epoch": 0.829311597126947,
+      "grad_norm": 0.7216476202011108,
+      "learning_rate": 9.644869489929321e-05,
+      "loss": 2.5298,
+      "step": 10276
+    },
+    {
+      "epoch": 0.82939230086353,
+      "grad_norm": 0.6815251111984253,
+      "learning_rate": 9.643291800403123e-05,
+      "loss": 2.5138,
+      "step": 10277
+    },
+    {
+      "epoch": 0.829473004600113,
+      "grad_norm": 0.6961970925331116,
+      "learning_rate": 9.64171411976702e-05,
+      "loss": 2.5441,
+      "step": 10278
+    },
+    {
+      "epoch": 0.829553708336696,
+      "grad_norm": 0.7317311763763428,
+      "learning_rate": 9.640136448060337e-05,
+      "loss": 2.5885,
+      "step": 10279
+    },
+    {
+      "epoch": 0.829634412073279,
+      "grad_norm": 0.729086697101593,
+      "learning_rate": 9.638558785322396e-05,
+      "loss": 2.475,
+      "step": 10280
+    },
+    {
+      "epoch": 0.829715115809862,
+      "grad_norm": 0.7790165543556213,
+      "learning_rate": 9.636981131592521e-05,
+      "loss": 2.5538,
+      "step": 10281
+    },
+    {
+      "epoch": 0.829795819546445,
+      "grad_norm": 0.7066864967346191,
+      "learning_rate": 9.635403486910018e-05,
+      "loss": 2.5916,
+      "step": 10282
+    },
+    {
+      "epoch": 0.829876523283028,
+      "grad_norm": 0.7070252299308777,
+      "learning_rate": 9.633825851314215e-05,
+      "loss": 2.5879,
+      "step": 10283
+    },
+    {
+      "epoch": 0.829957227019611,
+      "grad_norm": 0.7604004740715027,
+      "learning_rate": 9.63224822484443e-05,
+      "loss": 2.5298,
+      "step": 10284
+    },
+    {
+      "epoch": 0.830037930756194,
+      "grad_norm": 0.7548386454582214,
+      "learning_rate": 9.63067060753998e-05,
+      "loss": 2.5313,
+      "step": 10285
+    },
+    {
+      "epoch": 0.830118634492777,
+      "grad_norm": 0.7241540551185608,
+      "learning_rate": 9.629092999440183e-05,
+      "loss": 2.5498,
+      "step": 10286
+    },
+    {
+      "epoch": 0.83019933822936,
+      "grad_norm": 0.6748291850090027,
+      "learning_rate": 9.627515400584361e-05,
+      "loss": 2.523,
+      "step": 10287
+    },
+    {
+      "epoch": 0.8302800419659431,
+      "grad_norm": 0.6624683141708374,
+      "learning_rate": 9.625937811011826e-05,
+      "loss": 2.568,
+      "step": 10288
+    },
+    {
+      "epoch": 0.830360745702526,
+      "grad_norm": 0.6681114435195923,
+      "learning_rate": 9.624360230761899e-05,
+      "loss": 2.5255,
+      "step": 10289
+    },
+    {
+      "epoch": 0.830441449439109,
+      "grad_norm": 0.6895325183868408,
+      "learning_rate": 9.622782659873899e-05,
+      "loss": 2.5275,
+      "step": 10290
+    },
+    {
+      "epoch": 0.830522153175692,
+      "grad_norm": 0.7257826924324036,
+      "learning_rate": 9.621205098387137e-05,
+      "loss": 2.5102,
+      "step": 10291
+    },
+    {
+      "epoch": 0.8306028569122751,
+      "grad_norm": 0.6567066311836243,
+      "learning_rate": 9.619627546340935e-05,
+      "loss": 2.5721,
+      "step": 10292
+    },
+    {
+      "epoch": 0.8306835606488581,
+      "grad_norm": 0.6571428179740906,
+      "learning_rate": 9.61805000377461e-05,
+      "loss": 2.5014,
+      "step": 10293
+    },
+    {
+      "epoch": 0.830764264385441,
+      "grad_norm": 0.7807042598724365,
+      "learning_rate": 9.61647247072748e-05,
+      "loss": 2.632,
+      "step": 10294
+    },
+    {
+      "epoch": 0.830844968122024,
+      "grad_norm": 0.6688913702964783,
+      "learning_rate": 9.614894947238854e-05,
+      "loss": 2.5457,
+      "step": 10295
+    },
+    {
+      "epoch": 0.8309256718586071,
+      "grad_norm": 0.7769338488578796,
+      "learning_rate": 9.613317433348055e-05,
+      "loss": 2.4775,
+      "step": 10296
+    },
+    {
+      "epoch": 0.8310063755951901,
+      "grad_norm": 0.7089162468910217,
+      "learning_rate": 9.611739929094399e-05,
+      "loss": 2.4887,
+      "step": 10297
+    },
+    {
+      "epoch": 0.8310870793317731,
+      "grad_norm": 0.6901174783706665,
+      "learning_rate": 9.610162434517196e-05,
+      "loss": 2.6127,
+      "step": 10298
+    },
+    {
+      "epoch": 0.831167783068356,
+      "grad_norm": 0.6862173676490784,
+      "learning_rate": 9.608584949655764e-05,
+      "loss": 2.5432,
+      "step": 10299
+    },
+    {
+      "epoch": 0.8312484868049391,
+      "grad_norm": 0.6789367198944092,
+      "learning_rate": 9.607007474549418e-05,
+      "loss": 2.5135,
+      "step": 10300
+    },
+    {
+      "epoch": 0.8313291905415221,
+      "grad_norm": 0.6548805832862854,
+      "learning_rate": 9.605430009237474e-05,
+      "loss": 2.5466,
+      "step": 10301
+    },
+    {
+      "epoch": 0.8314098942781051,
+      "grad_norm": 0.6873800158500671,
+      "learning_rate": 9.603852553759244e-05,
+      "loss": 2.4954,
+      "step": 10302
+    },
+    {
+      "epoch": 0.831490598014688,
+      "grad_norm": 0.6816138029098511,
+      "learning_rate": 9.602275108154046e-05,
+      "loss": 2.5556,
+      "step": 10303
+    },
+    {
+      "epoch": 0.831571301751271,
+      "grad_norm": 0.6890314221382141,
+      "learning_rate": 9.600697672461189e-05,
+      "loss": 2.5253,
+      "step": 10304
+    },
+    {
+      "epoch": 0.8316520054878541,
+      "grad_norm": 0.6217427849769592,
+      "learning_rate": 9.599120246719992e-05,
+      "loss": 2.53,
+      "step": 10305
+    },
+    {
+      "epoch": 0.8317327092244371,
+      "grad_norm": 0.6638299226760864,
+      "learning_rate": 9.59754283096977e-05,
+      "loss": 2.5323,
+      "step": 10306
+    },
+    {
+      "epoch": 0.8318134129610201,
+      "grad_norm": 0.6834245920181274,
+      "learning_rate": 9.595965425249828e-05,
+      "loss": 2.5339,
+      "step": 10307
+    },
+    {
+      "epoch": 0.831894116697603,
+      "grad_norm": 0.8013476729393005,
+      "learning_rate": 9.594388029599484e-05,
+      "loss": 2.4925,
+      "step": 10308
+    },
+    {
+      "epoch": 0.8319748204341861,
+      "grad_norm": 0.7677187323570251,
+      "learning_rate": 9.592810644058049e-05,
+      "loss": 2.5717,
+      "step": 10309
+    },
+    {
+      "epoch": 0.8320555241707691,
+      "grad_norm": 0.6558046340942383,
+      "learning_rate": 9.591233268664841e-05,
+      "loss": 2.5631,
+      "step": 10310
+    },
+    {
+      "epoch": 0.8321362279073521,
+      "grad_norm": 0.6648481488227844,
+      "learning_rate": 9.589655903459165e-05,
+      "loss": 2.5232,
+      "step": 10311
+    },
+    {
+      "epoch": 0.8322169316439351,
+      "grad_norm": 0.6907756328582764,
+      "learning_rate": 9.588078548480338e-05,
+      "loss": 2.4804,
+      "step": 10312
+    },
+    {
+      "epoch": 0.8322976353805182,
+      "grad_norm": 0.6924928426742554,
+      "learning_rate": 9.586501203767675e-05,
+      "loss": 2.4648,
+      "step": 10313
+    },
+    {
+      "epoch": 0.8323783391171011,
+      "grad_norm": 0.7654799222946167,
+      "learning_rate": 9.584923869360477e-05,
+      "loss": 2.6184,
+      "step": 10314
+    },
+    {
+      "epoch": 0.8324590428536841,
+      "grad_norm": 0.7056179046630859,
+      "learning_rate": 9.58334654529806e-05,
+      "loss": 2.5862,
+      "step": 10315
+    },
+    {
+      "epoch": 0.8325397465902671,
+      "grad_norm": 0.7245064973831177,
+      "learning_rate": 9.581769231619743e-05,
+      "loss": 2.4866,
+      "step": 10316
+    },
+    {
+      "epoch": 0.8326204503268502,
+      "grad_norm": 0.6782355308532715,
+      "learning_rate": 9.580191928364824e-05,
+      "loss": 2.5519,
+      "step": 10317
+    },
+    {
+      "epoch": 0.8327011540634331,
+      "grad_norm": 0.6910805106163025,
+      "learning_rate": 9.578614635572621e-05,
+      "loss": 2.542,
+      "step": 10318
+    },
+    {
+      "epoch": 0.8327818578000161,
+      "grad_norm": 0.6858026385307312,
+      "learning_rate": 9.577037353282444e-05,
+      "loss": 2.5601,
+      "step": 10319
+    },
+    {
+      "epoch": 0.8328625615365991,
+      "grad_norm": 0.6886423230171204,
+      "learning_rate": 9.5754600815336e-05,
+      "loss": 2.5817,
+      "step": 10320
+    },
+    {
+      "epoch": 0.8329432652731822,
+      "grad_norm": 0.7585750818252563,
+      "learning_rate": 9.573882820365402e-05,
+      "loss": 2.5153,
+      "step": 10321
+    },
+    {
+      "epoch": 0.8330239690097652,
+      "grad_norm": 0.7004472613334656,
+      "learning_rate": 9.57230556981716e-05,
+      "loss": 2.5456,
+      "step": 10322
+    },
+    {
+      "epoch": 0.8331046727463481,
+      "grad_norm": 0.6530508399009705,
+      "learning_rate": 9.570728329928179e-05,
+      "loss": 2.5453,
+      "step": 10323
+    },
+    {
+      "epoch": 0.8331853764829311,
+      "grad_norm": 0.6767956614494324,
+      "learning_rate": 9.569151100737769e-05,
+      "loss": 2.5311,
+      "step": 10324
+    },
+    {
+      "epoch": 0.8332660802195142,
+      "grad_norm": 0.6835905909538269,
+      "learning_rate": 9.56757388228524e-05,
+      "loss": 2.5417,
+      "step": 10325
+    },
+    {
+      "epoch": 0.8333467839560972,
+      "grad_norm": 0.6582748889923096,
+      "learning_rate": 9.565996674609901e-05,
+      "loss": 2.5144,
+      "step": 10326
+    },
+    {
+      "epoch": 0.8334274876926802,
+      "grad_norm": 0.6815205216407776,
+      "learning_rate": 9.56441947775106e-05,
+      "loss": 2.5272,
+      "step": 10327
+    },
+    {
+      "epoch": 0.8335081914292631,
+      "grad_norm": 0.6810150146484375,
+      "learning_rate": 9.562842291748022e-05,
+      "loss": 2.5475,
+      "step": 10328
+    },
+    {
+      "epoch": 0.8335888951658462,
+      "grad_norm": 0.7220990657806396,
+      "learning_rate": 9.5612651166401e-05,
+      "loss": 2.54,
+      "step": 10329
+    },
+    {
+      "epoch": 0.8336695989024292,
+      "grad_norm": 0.6840164065361023,
+      "learning_rate": 9.559687952466596e-05,
+      "loss": 2.5987,
+      "step": 10330
+    },
+    {
+      "epoch": 0.8337503026390122,
+      "grad_norm": 0.7085031867027283,
+      "learning_rate": 9.558110799266819e-05,
+      "loss": 2.5674,
+      "step": 10331
+    },
+    {
+      "epoch": 0.8338310063755952,
+      "grad_norm": 0.6658117175102234,
+      "learning_rate": 9.55653365708008e-05,
+      "loss": 2.5793,
+      "step": 10332
+    },
+    {
+      "epoch": 0.8339117101121782,
+      "grad_norm": 0.782648503780365,
+      "learning_rate": 9.554956525945677e-05,
+      "loss": 2.5463,
+      "step": 10333
+    },
+    {
+      "epoch": 0.8339924138487612,
+      "grad_norm": 0.6999937891960144,
+      "learning_rate": 9.553379405902922e-05,
+      "loss": 2.5961,
+      "step": 10334
+    },
+    {
+      "epoch": 0.8340731175853442,
+      "grad_norm": 0.6681220531463623,
+      "learning_rate": 9.55180229699112e-05,
+      "loss": 2.6055,
+      "step": 10335
+    },
+    {
+      "epoch": 0.8341538213219272,
+      "grad_norm": 0.7127133011817932,
+      "learning_rate": 9.550225199249577e-05,
+      "loss": 2.5571,
+      "step": 10336
+    },
+    {
+      "epoch": 0.8342345250585103,
+      "grad_norm": 0.6939001679420471,
+      "learning_rate": 9.548648112717596e-05,
+      "loss": 2.5653,
+      "step": 10337
+    },
+    {
+      "epoch": 0.8343152287950932,
+      "grad_norm": 0.7483924031257629,
+      "learning_rate": 9.547071037434487e-05,
+      "loss": 2.5316,
+      "step": 10338
+    },
+    {
+      "epoch": 0.8343959325316762,
+      "grad_norm": 0.7975850105285645,
+      "learning_rate": 9.545493973439548e-05,
+      "loss": 2.6039,
+      "step": 10339
+    },
+    {
+      "epoch": 0.8344766362682592,
+      "grad_norm": 0.6893026232719421,
+      "learning_rate": 9.543916920772087e-05,
+      "loss": 2.5797,
+      "step": 10340
+    },
+    {
+      "epoch": 0.8345573400048423,
+      "grad_norm": 0.752869188785553,
+      "learning_rate": 9.542339879471409e-05,
+      "loss": 2.5677,
+      "step": 10341
+    },
+    {
+      "epoch": 0.8346380437414253,
+      "grad_norm": 0.7336339354515076,
+      "learning_rate": 9.540762849576822e-05,
+      "loss": 2.5212,
+      "step": 10342
+    },
+    {
+      "epoch": 0.8347187474780082,
+      "grad_norm": 0.7742713689804077,
+      "learning_rate": 9.539185831127621e-05,
+      "loss": 2.5599,
+      "step": 10343
+    },
+    {
+      "epoch": 0.8347994512145912,
+      "grad_norm": 0.7205352783203125,
+      "learning_rate": 9.537608824163114e-05,
+      "loss": 2.5591,
+      "step": 10344
+    },
+    {
+      "epoch": 0.8348801549511743,
+      "grad_norm": 0.7794787287712097,
+      "learning_rate": 9.536031828722605e-05,
+      "loss": 2.5858,
+      "step": 10345
+    },
+    {
+      "epoch": 0.8349608586877573,
+      "grad_norm": 0.7129528522491455,
+      "learning_rate": 9.534454844845396e-05,
+      "loss": 2.5591,
+      "step": 10346
+    },
+    {
+      "epoch": 0.8350415624243402,
+      "grad_norm": 0.731038510799408,
+      "learning_rate": 9.532877872570787e-05,
+      "loss": 2.5774,
+      "step": 10347
+    },
+    {
+      "epoch": 0.8351222661609232,
+      "grad_norm": 0.7706510424613953,
+      "learning_rate": 9.531300911938087e-05,
+      "loss": 2.6102,
+      "step": 10348
+    },
+    {
+      "epoch": 0.8352029698975063,
+      "grad_norm": 0.6890363097190857,
+      "learning_rate": 9.52972396298659e-05,
+      "loss": 2.5393,
+      "step": 10349
+    },
+    {
+      "epoch": 0.8352836736340893,
+      "grad_norm": 0.6792402863502502,
+      "learning_rate": 9.528147025755601e-05,
+      "loss": 2.5607,
+      "step": 10350
+    },
+    {
+      "epoch": 0.8353643773706723,
+      "grad_norm": 0.7097377777099609,
+      "learning_rate": 9.526570100284422e-05,
+      "loss": 2.5681,
+      "step": 10351
+    },
+    {
+      "epoch": 0.8354450811072552,
+      "grad_norm": 0.7530940771102905,
+      "learning_rate": 9.524993186612353e-05,
+      "loss": 2.5405,
+      "step": 10352
+    },
+    {
+      "epoch": 0.8355257848438382,
+      "grad_norm": 0.714080810546875,
+      "learning_rate": 9.523416284778696e-05,
+      "loss": 2.5365,
+      "step": 10353
+    },
+    {
+      "epoch": 0.8356064885804213,
+      "grad_norm": 0.6745832562446594,
+      "learning_rate": 9.521839394822752e-05,
+      "loss": 2.5553,
+      "step": 10354
+    },
+    {
+      "epoch": 0.8356871923170043,
+      "grad_norm": 0.7163450121879578,
+      "learning_rate": 9.52026251678382e-05,
+      "loss": 2.5074,
+      "step": 10355
+    },
+    {
+      "epoch": 0.8357678960535873,
+      "grad_norm": 0.6876534223556519,
+      "learning_rate": 9.518685650701197e-05,
+      "loss": 2.5652,
+      "step": 10356
+    },
+    {
+      "epoch": 0.8358485997901702,
+      "grad_norm": 0.6424533128738403,
+      "learning_rate": 9.517108796614187e-05,
+      "loss": 2.4823,
+      "step": 10357
+    },
+    {
+      "epoch": 0.8359293035267533,
+      "grad_norm": 0.646802544593811,
+      "learning_rate": 9.515531954562094e-05,
+      "loss": 2.5602,
+      "step": 10358
+    },
+    {
+      "epoch": 0.8360100072633363,
+      "grad_norm": 0.7266993522644043,
+      "learning_rate": 9.513955124584205e-05,
+      "loss": 2.5384,
+      "step": 10359
+    },
+    {
+      "epoch": 0.8360907109999193,
+      "grad_norm": 0.7358742356300354,
+      "learning_rate": 9.512378306719826e-05,
+      "loss": 2.5798,
+      "step": 10360
+    },
+    {
+      "epoch": 0.8361714147365022,
+      "grad_norm": 0.7191498279571533,
+      "learning_rate": 9.510801501008256e-05,
+      "loss": 2.5229,
+      "step": 10361
+    },
+    {
+      "epoch": 0.8362521184730853,
+      "grad_norm": 0.7058876156806946,
+      "learning_rate": 9.509224707488788e-05,
+      "loss": 2.5146,
+      "step": 10362
+    },
+    {
+      "epoch": 0.8363328222096683,
+      "grad_norm": 0.7348346710205078,
+      "learning_rate": 9.507647926200725e-05,
+      "loss": 2.5878,
+      "step": 10363
+    },
+    {
+      "epoch": 0.8364135259462513,
+      "grad_norm": 0.7464115619659424,
+      "learning_rate": 9.506071157183366e-05,
+      "loss": 2.6056,
+      "step": 10364
+    },
+    {
+      "epoch": 0.8364942296828343,
+      "grad_norm": 0.7077332139015198,
+      "learning_rate": 9.504494400476e-05,
+      "loss": 2.5161,
+      "step": 10365
+    },
+    {
+      "epoch": 0.8365749334194174,
+      "grad_norm": 0.7381827235221863,
+      "learning_rate": 9.502917656117928e-05,
+      "loss": 2.519,
+      "step": 10366
+    },
+    {
+      "epoch": 0.8366556371560003,
+      "grad_norm": 0.743180513381958,
+      "learning_rate": 9.501340924148452e-05,
+      "loss": 2.6149,
+      "step": 10367
+    },
+    {
+      "epoch": 0.8367363408925833,
+      "grad_norm": 0.6496078372001648,
+      "learning_rate": 9.499764204606863e-05,
+      "loss": 2.4969,
+      "step": 10368
+    },
+    {
+      "epoch": 0.8368170446291663,
+      "grad_norm": 0.6796541810035706,
+      "learning_rate": 9.498187497532454e-05,
+      "loss": 2.5304,
+      "step": 10369
+    },
+    {
+      "epoch": 0.8368977483657494,
+      "grad_norm": 0.6555948853492737,
+      "learning_rate": 9.496610802964529e-05,
+      "loss": 2.6029,
+      "step": 10370
+    },
+    {
+      "epoch": 0.8369784521023323,
+      "grad_norm": 0.6990405321121216,
+      "learning_rate": 9.495034120942374e-05,
+      "loss": 2.5286,
+      "step": 10371
+    },
+    {
+      "epoch": 0.8370591558389153,
+      "grad_norm": 0.7417613863945007,
+      "learning_rate": 9.49345745150529e-05,
+      "loss": 2.5301,
+      "step": 10372
+    },
+    {
+      "epoch": 0.8371398595754983,
+      "grad_norm": 0.6809872388839722,
+      "learning_rate": 9.49188079469257e-05,
+      "loss": 2.5075,
+      "step": 10373
+    },
+    {
+      "epoch": 0.8372205633120814,
+      "grad_norm": 0.6537099480628967,
+      "learning_rate": 9.490304150543514e-05,
+      "loss": 2.5515,
+      "step": 10374
+    },
+    {
+      "epoch": 0.8373012670486644,
+      "grad_norm": 0.6660431027412415,
+      "learning_rate": 9.488727519097407e-05,
+      "loss": 2.549,
+      "step": 10375
+    },
+    {
+      "epoch": 0.8373819707852473,
+      "grad_norm": 0.7257838249206543,
+      "learning_rate": 9.487150900393546e-05,
+      "loss": 2.546,
+      "step": 10376
+    },
+    {
+      "epoch": 0.8374626745218303,
+      "grad_norm": 0.742085874080658,
+      "learning_rate": 9.485574294471226e-05,
+      "loss": 2.5302,
+      "step": 10377
+    },
+    {
+      "epoch": 0.8375433782584134,
+      "grad_norm": 0.659934401512146,
+      "learning_rate": 9.48399770136974e-05,
+      "loss": 2.5553,
+      "step": 10378
+    },
+    {
+      "epoch": 0.8376240819949964,
+      "grad_norm": 0.7219613790512085,
+      "learning_rate": 9.482421121128377e-05,
+      "loss": 2.6186,
+      "step": 10379
+    },
+    {
+      "epoch": 0.8377047857315794,
+      "grad_norm": 0.706444263458252,
+      "learning_rate": 9.480844553786436e-05,
+      "loss": 2.5082,
+      "step": 10380
+    },
+    {
+      "epoch": 0.8377854894681623,
+      "grad_norm": 0.7527014017105103,
+      "learning_rate": 9.479267999383204e-05,
+      "loss": 2.5625,
+      "step": 10381
+    },
+    {
+      "epoch": 0.8378661932047454,
+      "grad_norm": 0.7488746643066406,
+      "learning_rate": 9.477691457957976e-05,
+      "loss": 2.528,
+      "step": 10382
+    },
+    {
+      "epoch": 0.8379468969413284,
+      "grad_norm": 0.7394229173660278,
+      "learning_rate": 9.476114929550045e-05,
+      "loss": 2.5387,
+      "step": 10383
+    },
+    {
+      "epoch": 0.8380276006779114,
+      "grad_norm": 0.7490981817245483,
+      "learning_rate": 9.474538414198695e-05,
+      "loss": 2.548,
+      "step": 10384
+    },
+    {
+      "epoch": 0.8381083044144944,
+      "grad_norm": 0.7203173041343689,
+      "learning_rate": 9.472961911943222e-05,
+      "loss": 2.5547,
+      "step": 10385
+    },
+    {
+      "epoch": 0.8381890081510774,
+      "grad_norm": 0.6929850578308105,
+      "learning_rate": 9.471385422822917e-05,
+      "loss": 2.4831,
+      "step": 10386
+    },
+    {
+      "epoch": 0.8382697118876604,
+      "grad_norm": 0.6303263902664185,
+      "learning_rate": 9.469808946877067e-05,
+      "loss": 2.4569,
+      "step": 10387
+    },
+    {
+      "epoch": 0.8383504156242434,
+      "grad_norm": 0.6986981630325317,
+      "learning_rate": 9.468232484144964e-05,
+      "loss": 2.5278,
+      "step": 10388
+    },
+    {
+      "epoch": 0.8384311193608264,
+      "grad_norm": 0.6910964846611023,
+      "learning_rate": 9.466656034665898e-05,
+      "loss": 2.5657,
+      "step": 10389
+    },
+    {
+      "epoch": 0.8385118230974095,
+      "grad_norm": 0.6571134924888611,
+      "learning_rate": 9.465079598479163e-05,
+      "loss": 2.6017,
+      "step": 10390
+    },
+    {
+      "epoch": 0.8385925268339924,
+      "grad_norm": 0.7117733359336853,
+      "learning_rate": 9.463503175624034e-05,
+      "loss": 2.56,
+      "step": 10391
+    },
+    {
+      "epoch": 0.8386732305705754,
+      "grad_norm": 0.7052998542785645,
+      "learning_rate": 9.461926766139813e-05,
+      "loss": 2.4998,
+      "step": 10392
+    },
+    {
+      "epoch": 0.8387539343071584,
+      "grad_norm": 0.7306597232818604,
+      "learning_rate": 9.460350370065786e-05,
+      "loss": 2.5292,
+      "step": 10393
+    },
+    {
+      "epoch": 0.8388346380437415,
+      "grad_norm": 0.681069552898407,
+      "learning_rate": 9.458773987441235e-05,
+      "loss": 2.5469,
+      "step": 10394
+    },
+    {
+      "epoch": 0.8389153417803245,
+      "grad_norm": 0.6681767702102661,
+      "learning_rate": 9.45719761830545e-05,
+      "loss": 2.5476,
+      "step": 10395
+    },
+    {
+      "epoch": 0.8389960455169074,
+      "grad_norm": 0.6759339570999146,
+      "learning_rate": 9.455621262697723e-05,
+      "loss": 2.4806,
+      "step": 10396
+    },
+    {
+      "epoch": 0.8390767492534904,
+      "grad_norm": 0.695829451084137,
+      "learning_rate": 9.454044920657333e-05,
+      "loss": 2.5255,
+      "step": 10397
+    },
+    {
+      "epoch": 0.8391574529900735,
+      "grad_norm": 0.686568558216095,
+      "learning_rate": 9.452468592223572e-05,
+      "loss": 2.5655,
+      "step": 10398
+    },
+    {
+      "epoch": 0.8392381567266565,
+      "grad_norm": 0.6529035568237305,
+      "learning_rate": 9.45089227743573e-05,
+      "loss": 2.5026,
+      "step": 10399
+    },
+    {
+      "epoch": 0.8393188604632394,
+      "grad_norm": 0.6809061765670776,
+      "learning_rate": 9.449315976333082e-05,
+      "loss": 2.5549,
+      "step": 10400
+    },
+    {
+      "epoch": 0.8393995641998224,
+      "grad_norm": 0.6920269727706909,
+      "learning_rate": 9.447739688954919e-05,
+      "loss": 2.517,
+      "step": 10401
+    },
+    {
+      "epoch": 0.8394802679364055,
+      "grad_norm": 0.6626712083816528,
+      "learning_rate": 9.446163415340526e-05,
+      "loss": 2.605,
+      "step": 10402
+    },
+    {
+      "epoch": 0.8395609716729885,
+      "grad_norm": 0.6912916898727417,
+      "learning_rate": 9.444587155529195e-05,
+      "loss": 2.588,
+      "step": 10403
+    },
+    {
+      "epoch": 0.8396416754095715,
+      "grad_norm": 0.6771352291107178,
+      "learning_rate": 9.443010909560198e-05,
+      "loss": 2.5148,
+      "step": 10404
+    },
+    {
+      "epoch": 0.8397223791461544,
+      "grad_norm": 0.7015509009361267,
+      "learning_rate": 9.441434677472827e-05,
+      "loss": 2.5425,
+      "step": 10405
+    },
+    {
+      "epoch": 0.8398030828827374,
+      "grad_norm": 0.6789976358413696,
+      "learning_rate": 9.439858459306364e-05,
+      "loss": 2.598,
+      "step": 10406
+    },
+    {
+      "epoch": 0.8398837866193205,
+      "grad_norm": 0.674391508102417,
+      "learning_rate": 9.438282255100091e-05,
+      "loss": 2.5581,
+      "step": 10407
+    },
+    {
+      "epoch": 0.8399644903559035,
+      "grad_norm": 0.6944772005081177,
+      "learning_rate": 9.436706064893294e-05,
+      "loss": 2.5591,
+      "step": 10408
+    },
+    {
+      "epoch": 0.8400451940924865,
+      "grad_norm": 0.6750832200050354,
+      "learning_rate": 9.435129888725259e-05,
+      "loss": 2.533,
+      "step": 10409
+    },
+    {
+      "epoch": 0.8401258978290694,
+      "grad_norm": 0.6927465200424194,
+      "learning_rate": 9.433553726635257e-05,
+      "loss": 2.536,
+      "step": 10410
+    },
+    {
+      "epoch": 0.8402066015656525,
+      "grad_norm": 0.6399651765823364,
+      "learning_rate": 9.431977578662578e-05,
+      "loss": 2.5123,
+      "step": 10411
+    },
+    {
+      "epoch": 0.8402873053022355,
+      "grad_norm": 0.7588143944740295,
+      "learning_rate": 9.430401444846505e-05,
+      "loss": 2.6133,
+      "step": 10412
+    },
+    {
+      "epoch": 0.8403680090388185,
+      "grad_norm": 0.8010972738265991,
+      "learning_rate": 9.428825325226313e-05,
+      "loss": 2.5407,
+      "step": 10413
+    },
+    {
+      "epoch": 0.8404487127754015,
+      "grad_norm": 0.6847307085990906,
+      "learning_rate": 9.427249219841288e-05,
+      "loss": 2.5912,
+      "step": 10414
+    },
+    {
+      "epoch": 0.8405294165119845,
+      "grad_norm": 0.7005963325500488,
+      "learning_rate": 9.425673128730716e-05,
+      "loss": 2.5059,
+      "step": 10415
+    },
+    {
+      "epoch": 0.8406101202485675,
+      "grad_norm": 0.7383962273597717,
+      "learning_rate": 9.424097051933862e-05,
+      "loss": 2.5157,
+      "step": 10416
+    },
+    {
+      "epoch": 0.8406908239851505,
+      "grad_norm": 0.7078843712806702,
+      "learning_rate": 9.422520989490018e-05,
+      "loss": 2.6093,
+      "step": 10417
+    },
+    {
+      "epoch": 0.8407715277217335,
+      "grad_norm": 0.7449501752853394,
+      "learning_rate": 9.42094494143846e-05,
+      "loss": 2.594,
+      "step": 10418
+    },
+    {
+      "epoch": 0.8408522314583166,
+      "grad_norm": 0.6823872923851013,
+      "learning_rate": 9.419368907818473e-05,
+      "loss": 2.5653,
+      "step": 10419
+    },
+    {
+      "epoch": 0.8409329351948995,
+      "grad_norm": 0.7403056025505066,
+      "learning_rate": 9.417792888669325e-05,
+      "loss": 2.5296,
+      "step": 10420
+    },
+    {
+      "epoch": 0.8410136389314825,
+      "grad_norm": 0.6858980655670166,
+      "learning_rate": 9.4162168840303e-05,
+      "loss": 2.5401,
+      "step": 10421
+    },
+    {
+      "epoch": 0.8410943426680655,
+      "grad_norm": 0.692348837852478,
+      "learning_rate": 9.41464089394068e-05,
+      "loss": 2.4797,
+      "step": 10422
+    },
+    {
+      "epoch": 0.8411750464046486,
+      "grad_norm": 0.6939836144447327,
+      "learning_rate": 9.413064918439736e-05,
+      "loss": 2.505,
+      "step": 10423
+    },
+    {
+      "epoch": 0.8412557501412316,
+      "grad_norm": 0.7334314584732056,
+      "learning_rate": 9.411488957566748e-05,
+      "loss": 2.5792,
+      "step": 10424
+    },
+    {
+      "epoch": 0.8413364538778145,
+      "grad_norm": 0.6977920532226562,
+      "learning_rate": 9.409913011360999e-05,
+      "loss": 2.5204,
+      "step": 10425
+    },
+    {
+      "epoch": 0.8414171576143975,
+      "grad_norm": 0.7121822834014893,
+      "learning_rate": 9.408337079861756e-05,
+      "loss": 2.571,
+      "step": 10426
+    },
+    {
+      "epoch": 0.8414978613509806,
+      "grad_norm": 0.761476993560791,
+      "learning_rate": 9.406761163108297e-05,
+      "loss": 2.5845,
+      "step": 10427
+    },
+    {
+      "epoch": 0.8415785650875636,
+      "grad_norm": 0.7160221934318542,
+      "learning_rate": 9.405185261139906e-05,
+      "loss": 2.5331,
+      "step": 10428
+    },
+    {
+      "epoch": 0.8416592688241465,
+      "grad_norm": 0.6828827857971191,
+      "learning_rate": 9.40360937399585e-05,
+      "loss": 2.5596,
+      "step": 10429
+    },
+    {
+      "epoch": 0.8417399725607295,
+      "grad_norm": 0.756473183631897,
+      "learning_rate": 9.402033501715406e-05,
+      "loss": 2.6107,
+      "step": 10430
+    },
+    {
+      "epoch": 0.8418206762973126,
+      "grad_norm": 0.7486895322799683,
+      "learning_rate": 9.400457644337853e-05,
+      "loss": 2.5388,
+      "step": 10431
+    },
+    {
+      "epoch": 0.8419013800338956,
+      "grad_norm": 0.7759146690368652,
+      "learning_rate": 9.398881801902461e-05,
+      "loss": 2.5559,
+      "step": 10432
+    },
+    {
+      "epoch": 0.8419820837704786,
+      "grad_norm": 0.71756911277771,
+      "learning_rate": 9.397305974448506e-05,
+      "loss": 2.6109,
+      "step": 10433
+    },
+    {
+      "epoch": 0.8420627875070615,
+      "grad_norm": 0.7741644382476807,
+      "learning_rate": 9.395730162015261e-05,
+      "loss": 2.5664,
+      "step": 10434
+    },
+    {
+      "epoch": 0.8421434912436446,
+      "grad_norm": 0.7155938744544983,
+      "learning_rate": 9.394154364642006e-05,
+      "loss": 2.5693,
+      "step": 10435
+    },
+    {
+      "epoch": 0.8422241949802276,
+      "grad_norm": 0.6862725019454956,
+      "learning_rate": 9.392578582368002e-05,
+      "loss": 2.4942,
+      "step": 10436
+    },
+    {
+      "epoch": 0.8423048987168106,
+      "grad_norm": 0.6698417067527771,
+      "learning_rate": 9.391002815232528e-05,
+      "loss": 2.5258,
+      "step": 10437
+    },
+    {
+      "epoch": 0.8423856024533936,
+      "grad_norm": 0.7756468057632446,
+      "learning_rate": 9.389427063274858e-05,
+      "loss": 2.5008,
+      "step": 10438
+    },
+    {
+      "epoch": 0.8424663061899766,
+      "grad_norm": 0.6579857468605042,
+      "learning_rate": 9.387851326534259e-05,
+      "loss": 2.5335,
+      "step": 10439
+    },
+    {
+      "epoch": 0.8425470099265596,
+      "grad_norm": 0.7673436403274536,
+      "learning_rate": 9.386275605050006e-05,
+      "loss": 2.5646,
+      "step": 10440
+    },
+    {
+      "epoch": 0.8426277136631426,
+      "grad_norm": 0.7377188205718994,
+      "learning_rate": 9.384699898861372e-05,
+      "loss": 2.568,
+      "step": 10441
+    },
+    {
+      "epoch": 0.8427084173997256,
+      "grad_norm": 0.6502123475074768,
+      "learning_rate": 9.38312420800762e-05,
+      "loss": 2.6091,
+      "step": 10442
+    },
+    {
+      "epoch": 0.8427891211363087,
+      "grad_norm": 0.729852020740509,
+      "learning_rate": 9.381548532528026e-05,
+      "loss": 2.4873,
+      "step": 10443
+    },
+    {
+      "epoch": 0.8428698248728916,
+      "grad_norm": 0.7419102191925049,
+      "learning_rate": 9.379972872461865e-05,
+      "loss": 2.4966,
+      "step": 10444
+    },
+    {
+      "epoch": 0.8429505286094746,
+      "grad_norm": 0.6921093463897705,
+      "learning_rate": 9.378397227848395e-05,
+      "loss": 2.4895,
+      "step": 10445
+    },
+    {
+      "epoch": 0.8430312323460576,
+      "grad_norm": 0.7697325944900513,
+      "learning_rate": 9.376821598726892e-05,
+      "loss": 2.5779,
+      "step": 10446
+    },
+    {
+      "epoch": 0.8431119360826407,
+      "grad_norm": 0.6441029906272888,
+      "learning_rate": 9.375245985136626e-05,
+      "loss": 2.4909,
+      "step": 10447
+    },
+    {
+      "epoch": 0.8431926398192237,
+      "grad_norm": 0.6962057948112488,
+      "learning_rate": 9.373670387116861e-05,
+      "loss": 2.5602,
+      "step": 10448
+    },
+    {
+      "epoch": 0.8432733435558066,
+      "grad_norm": 0.7030641436576843,
+      "learning_rate": 9.372094804706867e-05,
+      "loss": 2.5641,
+      "step": 10449
+    },
+    {
+      "epoch": 0.8433540472923896,
+      "grad_norm": 0.6969063878059387,
+      "learning_rate": 9.370519237945912e-05,
+      "loss": 2.5555,
+      "step": 10450
+    },
+    {
+      "epoch": 0.8434347510289727,
+      "grad_norm": 0.7169879674911499,
+      "learning_rate": 9.368943686873267e-05,
+      "loss": 2.5258,
+      "step": 10451
+    },
+    {
+      "epoch": 0.8435154547655557,
+      "grad_norm": 0.7198735475540161,
+      "learning_rate": 9.36736815152819e-05,
+      "loss": 2.5192,
+      "step": 10452
+    },
+    {
+      "epoch": 0.8435961585021386,
+      "grad_norm": 0.6613535284996033,
+      "learning_rate": 9.365792631949951e-05,
+      "loss": 2.5596,
+      "step": 10453
+    },
+    {
+      "epoch": 0.8436768622387216,
+      "grad_norm": 0.6377065777778625,
+      "learning_rate": 9.364217128177824e-05,
+      "loss": 2.5518,
+      "step": 10454
+    },
+    {
+      "epoch": 0.8437575659753046,
+      "grad_norm": 0.6670635938644409,
+      "learning_rate": 9.362641640251063e-05,
+      "loss": 2.4793,
+      "step": 10455
+    },
+    {
+      "epoch": 0.8438382697118877,
+      "grad_norm": 0.6556122899055481,
+      "learning_rate": 9.361066168208939e-05,
+      "loss": 2.5492,
+      "step": 10456
+    },
+    {
+      "epoch": 0.8439189734484707,
+      "grad_norm": 0.7262280583381653,
+      "learning_rate": 9.35949071209072e-05,
+      "loss": 2.6059,
+      "step": 10457
+    },
+    {
+      "epoch": 0.8439996771850536,
+      "grad_norm": 0.702953040599823,
+      "learning_rate": 9.357915271935662e-05,
+      "loss": 2.5445,
+      "step": 10458
+    },
+    {
+      "epoch": 0.8440803809216366,
+      "grad_norm": 0.6619930267333984,
+      "learning_rate": 9.356339847783036e-05,
+      "loss": 2.5688,
+      "step": 10459
+    },
+    {
+      "epoch": 0.8441610846582197,
+      "grad_norm": 0.7038032412528992,
+      "learning_rate": 9.354764439672106e-05,
+      "loss": 2.5195,
+      "step": 10460
+    },
+    {
+      "epoch": 0.8442417883948027,
+      "grad_norm": 0.6615132689476013,
+      "learning_rate": 9.353189047642129e-05,
+      "loss": 2.5176,
+      "step": 10461
+    },
+    {
+      "epoch": 0.8443224921313857,
+      "grad_norm": 0.6524826288223267,
+      "learning_rate": 9.351613671732372e-05,
+      "loss": 2.4294,
+      "step": 10462
+    },
+    {
+      "epoch": 0.8444031958679686,
+      "grad_norm": 0.6526279449462891,
+      "learning_rate": 9.350038311982099e-05,
+      "loss": 2.595,
+      "step": 10463
+    },
+    {
+      "epoch": 0.8444838996045517,
+      "grad_norm": 0.6610859632492065,
+      "learning_rate": 9.348462968430569e-05,
+      "loss": 2.5311,
+      "step": 10464
+    },
+    {
+      "epoch": 0.8445646033411347,
+      "grad_norm": 0.6835470795631409,
+      "learning_rate": 9.346887641117045e-05,
+      "loss": 2.5694,
+      "step": 10465
+    },
+    {
+      "epoch": 0.8446453070777177,
+      "grad_norm": 0.6768551468849182,
+      "learning_rate": 9.345312330080787e-05,
+      "loss": 2.6082,
+      "step": 10466
+    },
+    {
+      "epoch": 0.8447260108143007,
+      "grad_norm": 0.6368672847747803,
+      "learning_rate": 9.343737035361059e-05,
+      "loss": 2.5221,
+      "step": 10467
+    },
+    {
+      "epoch": 0.8448067145508837,
+      "grad_norm": 0.6952844858169556,
+      "learning_rate": 9.34216175699712e-05,
+      "loss": 2.5003,
+      "step": 10468
+    },
+    {
+      "epoch": 0.8448874182874667,
+      "grad_norm": 0.6663931012153625,
+      "learning_rate": 9.340586495028227e-05,
+      "loss": 2.5469,
+      "step": 10469
+    },
+    {
+      "epoch": 0.8449681220240497,
+      "grad_norm": 0.6840688586235046,
+      "learning_rate": 9.339011249493647e-05,
+      "loss": 2.5499,
+      "step": 10470
+    },
+    {
+      "epoch": 0.8450488257606327,
+      "grad_norm": 0.6832869052886963,
+      "learning_rate": 9.337436020432632e-05,
+      "loss": 2.5492,
+      "step": 10471
+    },
+    {
+      "epoch": 0.8451295294972158,
+      "grad_norm": 0.7444044947624207,
+      "learning_rate": 9.335860807884442e-05,
+      "loss": 2.5791,
+      "step": 10472
+    },
+    {
+      "epoch": 0.8452102332337987,
+      "grad_norm": 0.6821839809417725,
+      "learning_rate": 9.334285611888339e-05,
+      "loss": 2.4772,
+      "step": 10473
+    },
+    {
+      "epoch": 0.8452909369703817,
+      "grad_norm": 0.6209141612052917,
+      "learning_rate": 9.332710432483577e-05,
+      "loss": 2.5656,
+      "step": 10474
+    },
+    {
+      "epoch": 0.8453716407069647,
+      "grad_norm": 0.6531212329864502,
+      "learning_rate": 9.331135269709415e-05,
+      "loss": 2.5285,
+      "step": 10475
+    },
+    {
+      "epoch": 0.8454523444435478,
+      "grad_norm": 0.6418079137802124,
+      "learning_rate": 9.329560123605115e-05,
+      "loss": 2.5503,
+      "step": 10476
+    },
+    {
+      "epoch": 0.8455330481801308,
+      "grad_norm": 0.6636360287666321,
+      "learning_rate": 9.327984994209924e-05,
+      "loss": 2.528,
+      "step": 10477
+    },
+    {
+      "epoch": 0.8456137519167137,
+      "grad_norm": 0.6196488738059998,
+      "learning_rate": 9.326409881563102e-05,
+      "loss": 2.4907,
+      "step": 10478
+    },
+    {
+      "epoch": 0.8456944556532967,
+      "grad_norm": 0.6339137554168701,
+      "learning_rate": 9.324834785703913e-05,
+      "loss": 2.4672,
+      "step": 10479
+    },
+    {
+      "epoch": 0.8457751593898798,
+      "grad_norm": 0.6803932189941406,
+      "learning_rate": 9.323259706671602e-05,
+      "loss": 2.5538,
+      "step": 10480
+    },
+    {
+      "epoch": 0.8458558631264628,
+      "grad_norm": 0.6815275549888611,
+      "learning_rate": 9.321684644505429e-05,
+      "loss": 2.5291,
+      "step": 10481
+    },
+    {
+      "epoch": 0.8459365668630457,
+      "grad_norm": 0.6497374773025513,
+      "learning_rate": 9.320109599244646e-05,
+      "loss": 2.5499,
+      "step": 10482
+    },
+    {
+      "epoch": 0.8460172705996287,
+      "grad_norm": 0.7966926097869873,
+      "learning_rate": 9.318534570928512e-05,
+      "loss": 2.523,
+      "step": 10483
+    },
+    {
+      "epoch": 0.8460979743362118,
+      "grad_norm": 0.6532156467437744,
+      "learning_rate": 9.316959559596276e-05,
+      "loss": 2.5138,
+      "step": 10484
+    },
+    {
+      "epoch": 0.8461786780727948,
+      "grad_norm": 0.7292522192001343,
+      "learning_rate": 9.315384565287193e-05,
+      "loss": 2.5413,
+      "step": 10485
+    },
+    {
+      "epoch": 0.8462593818093778,
+      "grad_norm": 0.7610795497894287,
+      "learning_rate": 9.313809588040519e-05,
+      "loss": 2.5071,
+      "step": 10486
+    },
+    {
+      "epoch": 0.8463400855459607,
+      "grad_norm": 0.7038258910179138,
+      "learning_rate": 9.312234627895502e-05,
+      "loss": 2.5568,
+      "step": 10487
+    },
+    {
+      "epoch": 0.8464207892825438,
+      "grad_norm": 0.7136046290397644,
+      "learning_rate": 9.310659684891395e-05,
+      "loss": 2.5372,
+      "step": 10488
+    },
+    {
+      "epoch": 0.8465014930191268,
+      "grad_norm": 0.7512896060943604,
+      "learning_rate": 9.309084759067452e-05,
+      "loss": 2.5821,
+      "step": 10489
+    },
+    {
+      "epoch": 0.8465821967557098,
+      "grad_norm": 0.7436400651931763,
+      "learning_rate": 9.307509850462922e-05,
+      "loss": 2.5489,
+      "step": 10490
+    },
+    {
+      "epoch": 0.8466629004922928,
+      "grad_norm": 0.6858603954315186,
+      "learning_rate": 9.305934959117056e-05,
+      "loss": 2.5622,
+      "step": 10491
+    },
+    {
+      "epoch": 0.8467436042288758,
+      "grad_norm": 0.707185685634613,
+      "learning_rate": 9.304360085069107e-05,
+      "loss": 2.5275,
+      "step": 10492
+    },
+    {
+      "epoch": 0.8468243079654588,
+      "grad_norm": 0.7207933068275452,
+      "learning_rate": 9.302785228358322e-05,
+      "loss": 2.5877,
+      "step": 10493
+    },
+    {
+      "epoch": 0.8469050117020418,
+      "grad_norm": 0.6470080614089966,
+      "learning_rate": 9.30121038902395e-05,
+      "loss": 2.5117,
+      "step": 10494
+    },
+    {
+      "epoch": 0.8469857154386248,
+      "grad_norm": 0.75248783826828,
+      "learning_rate": 9.299635567105247e-05,
+      "loss": 2.5259,
+      "step": 10495
+    },
+    {
+      "epoch": 0.8470664191752079,
+      "grad_norm": 0.7150708436965942,
+      "learning_rate": 9.298060762641452e-05,
+      "loss": 2.551,
+      "step": 10496
+    },
+    {
+      "epoch": 0.8471471229117908,
+      "grad_norm": 0.6865069270133972,
+      "learning_rate": 9.296485975671818e-05,
+      "loss": 2.5184,
+      "step": 10497
+    },
+    {
+      "epoch": 0.8472278266483738,
+      "grad_norm": 0.7188237309455872,
+      "learning_rate": 9.294911206235593e-05,
+      "loss": 2.5207,
+      "step": 10498
+    },
+    {
+      "epoch": 0.8473085303849568,
+      "grad_norm": 0.6907880902290344,
+      "learning_rate": 9.293336454372026e-05,
+      "loss": 2.5544,
+      "step": 10499
+    },
+    {
+      "epoch": 0.8473892341215399,
+      "grad_norm": 0.7626079320907593,
+      "learning_rate": 9.291761720120358e-05,
+      "loss": 2.5741,
+      "step": 10500
+    },
+    {
+      "epoch": 0.8474699378581229,
+      "grad_norm": 0.6731963753700256,
+      "learning_rate": 9.29018700351984e-05,
+      "loss": 2.5433,
+      "step": 10501
+    },
+    {
+      "epoch": 0.8475506415947058,
+      "grad_norm": 0.7256288528442383,
+      "learning_rate": 9.288612304609723e-05,
+      "loss": 2.5131,
+      "step": 10502
+    },
+    {
+      "epoch": 0.8476313453312888,
+      "grad_norm": 0.7129119634628296,
+      "learning_rate": 9.287037623429242e-05,
+      "loss": 2.5054,
+      "step": 10503
+    },
+    {
+      "epoch": 0.8477120490678719,
+      "grad_norm": 0.6711156964302063,
+      "learning_rate": 9.285462960017644e-05,
+      "loss": 2.5671,
+      "step": 10504
+    },
+    {
+      "epoch": 0.8477927528044549,
+      "grad_norm": 0.7268081903457642,
+      "learning_rate": 9.283888314414184e-05,
+      "loss": 2.5627,
+      "step": 10505
+    },
+    {
+      "epoch": 0.8478734565410379,
+      "grad_norm": 0.8635050058364868,
+      "learning_rate": 9.282313686658094e-05,
+      "loss": 2.517,
+      "step": 10506
+    },
+    {
+      "epoch": 0.8479541602776208,
+      "grad_norm": 0.7077138423919678,
+      "learning_rate": 9.280739076788624e-05,
+      "loss": 2.5551,
+      "step": 10507
+    },
+    {
+      "epoch": 0.8480348640142038,
+      "grad_norm": 0.6312204599380493,
+      "learning_rate": 9.279164484845018e-05,
+      "loss": 2.5329,
+      "step": 10508
+    },
+    {
+      "epoch": 0.8481155677507869,
+      "grad_norm": 0.6749829649925232,
+      "learning_rate": 9.277589910866516e-05,
+      "loss": 2.5092,
+      "step": 10509
+    },
+    {
+      "epoch": 0.8481962714873699,
+      "grad_norm": 0.753391683101654,
+      "learning_rate": 9.27601535489236e-05,
+      "loss": 2.6244,
+      "step": 10510
+    },
+    {
+      "epoch": 0.8482769752239528,
+      "grad_norm": 0.7230119109153748,
+      "learning_rate": 9.2744408169618e-05,
+      "loss": 2.5021,
+      "step": 10511
+    },
+    {
+      "epoch": 0.8483576789605358,
+      "grad_norm": 0.6759157776832581,
+      "learning_rate": 9.272866297114067e-05,
+      "loss": 2.5399,
+      "step": 10512
+    },
+    {
+      "epoch": 0.8484383826971189,
+      "grad_norm": 0.7049473524093628,
+      "learning_rate": 9.271291795388406e-05,
+      "loss": 2.5024,
+      "step": 10513
+    },
+    {
+      "epoch": 0.8485190864337019,
+      "grad_norm": 0.6579850912094116,
+      "learning_rate": 9.269717311824058e-05,
+      "loss": 2.5019,
+      "step": 10514
+    },
+    {
+      "epoch": 0.8485997901702849,
+      "grad_norm": 0.7091391086578369,
+      "learning_rate": 9.268142846460265e-05,
+      "loss": 2.5785,
+      "step": 10515
+    },
+    {
+      "epoch": 0.8486804939068678,
+      "grad_norm": 0.6612898707389832,
+      "learning_rate": 9.266568399336266e-05,
+      "loss": 2.5046,
+      "step": 10516
+    },
+    {
+      "epoch": 0.8487611976434509,
+      "grad_norm": 0.6348623633384705,
+      "learning_rate": 9.264993970491298e-05,
+      "loss": 2.543,
+      "step": 10517
+    },
+    {
+      "epoch": 0.8488419013800339,
+      "grad_norm": 0.688360869884491,
+      "learning_rate": 9.263419559964604e-05,
+      "loss": 2.5294,
+      "step": 10518
+    },
+    {
+      "epoch": 0.8489226051166169,
+      "grad_norm": 0.6483190059661865,
+      "learning_rate": 9.261845167795418e-05,
+      "loss": 2.5623,
+      "step": 10519
+    },
+    {
+      "epoch": 0.8490033088531999,
+      "grad_norm": 0.689379096031189,
+      "learning_rate": 9.26027079402298e-05,
+      "loss": 2.4871,
+      "step": 10520
+    },
+    {
+      "epoch": 0.8490840125897829,
+      "grad_norm": 0.6627655625343323,
+      "learning_rate": 9.25869643868653e-05,
+      "loss": 2.5353,
+      "step": 10521
+    },
+    {
+      "epoch": 0.8491647163263659,
+      "grad_norm": 0.6701192259788513,
+      "learning_rate": 9.2571221018253e-05,
+      "loss": 2.5003,
+      "step": 10522
+    },
+    {
+      "epoch": 0.8492454200629489,
+      "grad_norm": 0.7413944005966187,
+      "learning_rate": 9.255547783478529e-05,
+      "loss": 2.5473,
+      "step": 10523
+    },
+    {
+      "epoch": 0.8493261237995319,
+      "grad_norm": 0.6490365266799927,
+      "learning_rate": 9.253973483685455e-05,
+      "loss": 2.5168,
+      "step": 10524
+    },
+    {
+      "epoch": 0.849406827536115,
+      "grad_norm": 0.7303688526153564,
+      "learning_rate": 9.25239920248531e-05,
+      "loss": 2.5953,
+      "step": 10525
+    },
+    {
+      "epoch": 0.8494875312726979,
+      "grad_norm": 0.7132991552352905,
+      "learning_rate": 9.250824939917331e-05,
+      "loss": 2.475,
+      "step": 10526
+    },
+    {
+      "epoch": 0.8495682350092809,
+      "grad_norm": 0.6935676336288452,
+      "learning_rate": 9.249250696020753e-05,
+      "loss": 2.5212,
+      "step": 10527
+    },
+    {
+      "epoch": 0.8496489387458639,
+      "grad_norm": 0.732961118221283,
+      "learning_rate": 9.247676470834814e-05,
+      "loss": 2.5848,
+      "step": 10528
+    },
+    {
+      "epoch": 0.849729642482447,
+      "grad_norm": 0.6899160146713257,
+      "learning_rate": 9.246102264398739e-05,
+      "loss": 2.4551,
+      "step": 10529
+    },
+    {
+      "epoch": 0.84981034621903,
+      "grad_norm": 0.6941123604774475,
+      "learning_rate": 9.244528076751766e-05,
+      "loss": 2.5441,
+      "step": 10530
+    },
+    {
+      "epoch": 0.8498910499556129,
+      "grad_norm": 0.7351016998291016,
+      "learning_rate": 9.242953907933134e-05,
+      "loss": 2.6519,
+      "step": 10531
+    },
+    {
+      "epoch": 0.8499717536921959,
+      "grad_norm": 0.7156691551208496,
+      "learning_rate": 9.241379757982065e-05,
+      "loss": 2.573,
+      "step": 10532
+    },
+    {
+      "epoch": 0.850052457428779,
+      "grad_norm": 0.7137688994407654,
+      "learning_rate": 9.239805626937797e-05,
+      "loss": 2.5688,
+      "step": 10533
+    },
+    {
+      "epoch": 0.850133161165362,
+      "grad_norm": 0.7018687129020691,
+      "learning_rate": 9.238231514839559e-05,
+      "loss": 2.5725,
+      "step": 10534
+    },
+    {
+      "epoch": 0.850213864901945,
+      "grad_norm": 0.6723659634590149,
+      "learning_rate": 9.236657421726583e-05,
+      "loss": 2.5661,
+      "step": 10535
+    },
+    {
+      "epoch": 0.8502945686385279,
+      "grad_norm": 0.7105850577354431,
+      "learning_rate": 9.235083347638098e-05,
+      "loss": 2.5676,
+      "step": 10536
+    },
+    {
+      "epoch": 0.850375272375111,
+      "grad_norm": 0.682601809501648,
+      "learning_rate": 9.233509292613341e-05,
+      "loss": 2.5489,
+      "step": 10537
+    },
+    {
+      "epoch": 0.850455976111694,
+      "grad_norm": 0.6703988313674927,
+      "learning_rate": 9.231935256691531e-05,
+      "loss": 2.5349,
+      "step": 10538
+    },
+    {
+      "epoch": 0.850536679848277,
+      "grad_norm": 0.6430882215499878,
+      "learning_rate": 9.230361239911903e-05,
+      "loss": 2.4959,
+      "step": 10539
+    },
+    {
+      "epoch": 0.8506173835848599,
+      "grad_norm": 0.7164519429206848,
+      "learning_rate": 9.228787242313687e-05,
+      "loss": 2.4999,
+      "step": 10540
+    },
+    {
+      "epoch": 0.850698087321443,
+      "grad_norm": 0.7463028430938721,
+      "learning_rate": 9.227213263936107e-05,
+      "loss": 2.545,
+      "step": 10541
+    },
+    {
+      "epoch": 0.850778791058026,
+      "grad_norm": 0.650577187538147,
+      "learning_rate": 9.22563930481839e-05,
+      "loss": 2.5707,
+      "step": 10542
+    },
+    {
+      "epoch": 0.850859494794609,
+      "grad_norm": 0.6808211207389832,
+      "learning_rate": 9.224065364999768e-05,
+      "loss": 2.5236,
+      "step": 10543
+    },
+    {
+      "epoch": 0.850940198531192,
+      "grad_norm": 0.6947758793830872,
+      "learning_rate": 9.222491444519467e-05,
+      "loss": 2.555,
+      "step": 10544
+    },
+    {
+      "epoch": 0.851020902267775,
+      "grad_norm": 0.6805624961853027,
+      "learning_rate": 9.22091754341671e-05,
+      "loss": 2.517,
+      "step": 10545
+    },
+    {
+      "epoch": 0.851101606004358,
+      "grad_norm": 0.6645655035972595,
+      "learning_rate": 9.219343661730724e-05,
+      "loss": 2.5237,
+      "step": 10546
+    },
+    {
+      "epoch": 0.851182309740941,
+      "grad_norm": 0.6912586092948914,
+      "learning_rate": 9.217769799500738e-05,
+      "loss": 2.5345,
+      "step": 10547
+    },
+    {
+      "epoch": 0.851263013477524,
+      "grad_norm": 0.6713781356811523,
+      "learning_rate": 9.21619595676597e-05,
+      "loss": 2.56,
+      "step": 10548
+    },
+    {
+      "epoch": 0.8513437172141071,
+      "grad_norm": 0.7031502723693848,
+      "learning_rate": 9.214622133565648e-05,
+      "loss": 2.4885,
+      "step": 10549
+    },
+    {
+      "epoch": 0.85142442095069,
+      "grad_norm": 0.6616455316543579,
+      "learning_rate": 9.213048329938997e-05,
+      "loss": 2.5101,
+      "step": 10550
+    },
+    {
+      "epoch": 0.851505124687273,
+      "grad_norm": 0.711077094078064,
+      "learning_rate": 9.211474545925236e-05,
+      "loss": 2.6264,
+      "step": 10551
+    },
+    {
+      "epoch": 0.851585828423856,
+      "grad_norm": 0.7534502744674683,
+      "learning_rate": 9.209900781563592e-05,
+      "loss": 2.5417,
+      "step": 10552
+    },
+    {
+      "epoch": 0.8516665321604391,
+      "grad_norm": 0.7405222058296204,
+      "learning_rate": 9.208327036893288e-05,
+      "loss": 2.546,
+      "step": 10553
+    },
+    {
+      "epoch": 0.8517472358970221,
+      "grad_norm": 0.7014057040214539,
+      "learning_rate": 9.20675331195354e-05,
+      "loss": 2.5211,
+      "step": 10554
+    },
+    {
+      "epoch": 0.851827939633605,
+      "grad_norm": 0.6984074115753174,
+      "learning_rate": 9.205179606783573e-05,
+      "loss": 2.5181,
+      "step": 10555
+    },
+    {
+      "epoch": 0.851908643370188,
+      "grad_norm": 0.7312670350074768,
+      "learning_rate": 9.203605921422613e-05,
+      "loss": 2.5345,
+      "step": 10556
+    },
+    {
+      "epoch": 0.851989347106771,
+      "grad_norm": 0.6861104369163513,
+      "learning_rate": 9.202032255909871e-05,
+      "loss": 2.5426,
+      "step": 10557
+    },
+    {
+      "epoch": 0.8520700508433541,
+      "grad_norm": 0.6989030838012695,
+      "learning_rate": 9.200458610284571e-05,
+      "loss": 2.5221,
+      "step": 10558
+    },
+    {
+      "epoch": 0.852150754579937,
+      "grad_norm": 0.6645115613937378,
+      "learning_rate": 9.198884984585932e-05,
+      "loss": 2.4755,
+      "step": 10559
+    },
+    {
+      "epoch": 0.85223145831652,
+      "grad_norm": 0.6577785015106201,
+      "learning_rate": 9.197311378853176e-05,
+      "loss": 2.5491,
+      "step": 10560
+    },
+    {
+      "epoch": 0.852312162053103,
+      "grad_norm": 0.7311568856239319,
+      "learning_rate": 9.195737793125517e-05,
+      "loss": 2.5653,
+      "step": 10561
+    },
+    {
+      "epoch": 0.8523928657896861,
+      "grad_norm": 0.6469970345497131,
+      "learning_rate": 9.194164227442174e-05,
+      "loss": 2.5384,
+      "step": 10562
+    },
+    {
+      "epoch": 0.8524735695262691,
+      "grad_norm": 0.6562933325767517,
+      "learning_rate": 9.19259068184237e-05,
+      "loss": 2.5644,
+      "step": 10563
+    },
+    {
+      "epoch": 0.852554273262852,
+      "grad_norm": 0.7740273475646973,
+      "learning_rate": 9.19101715636531e-05,
+      "loss": 2.5868,
+      "step": 10564
+    },
+    {
+      "epoch": 0.852634976999435,
+      "grad_norm": 0.6461195349693298,
+      "learning_rate": 9.18944365105022e-05,
+      "loss": 2.4862,
+      "step": 10565
+    },
+    {
+      "epoch": 0.8527156807360181,
+      "grad_norm": 0.7230537533760071,
+      "learning_rate": 9.187870165936313e-05,
+      "loss": 2.5125,
+      "step": 10566
+    },
+    {
+      "epoch": 0.8527963844726011,
+      "grad_norm": 0.6858233213424683,
+      "learning_rate": 9.186296701062805e-05,
+      "loss": 2.5463,
+      "step": 10567
+    },
+    {
+      "epoch": 0.8528770882091841,
+      "grad_norm": 0.717407763004303,
+      "learning_rate": 9.184723256468908e-05,
+      "loss": 2.5399,
+      "step": 10568
+    },
+    {
+      "epoch": 0.852957791945767,
+      "grad_norm": 0.7537745237350464,
+      "learning_rate": 9.18314983219384e-05,
+      "loss": 2.5164,
+      "step": 10569
+    },
+    {
+      "epoch": 0.8530384956823501,
+      "grad_norm": 0.7068665027618408,
+      "learning_rate": 9.181576428276814e-05,
+      "loss": 2.5747,
+      "step": 10570
+    },
+    {
+      "epoch": 0.8531191994189331,
+      "grad_norm": 0.8013456463813782,
+      "learning_rate": 9.18000304475704e-05,
+      "loss": 2.5401,
+      "step": 10571
+    },
+    {
+      "epoch": 0.8531999031555161,
+      "grad_norm": 0.6458969712257385,
+      "learning_rate": 9.178429681673741e-05,
+      "loss": 2.4781,
+      "step": 10572
+    },
+    {
+      "epoch": 0.8532806068920991,
+      "grad_norm": 0.7235112190246582,
+      "learning_rate": 9.176856339066114e-05,
+      "loss": 2.5753,
+      "step": 10573
+    },
+    {
+      "epoch": 0.8533613106286821,
+      "grad_norm": 0.6815706491470337,
+      "learning_rate": 9.175283016973382e-05,
+      "loss": 2.5526,
+      "step": 10574
+    },
+    {
+      "epoch": 0.8534420143652651,
+      "grad_norm": 0.739747166633606,
+      "learning_rate": 9.173709715434751e-05,
+      "loss": 2.5631,
+      "step": 10575
+    },
+    {
+      "epoch": 0.8535227181018481,
+      "grad_norm": 0.7325060963630676,
+      "learning_rate": 9.172136434489437e-05,
+      "loss": 2.4925,
+      "step": 10576
+    },
+    {
+      "epoch": 0.8536034218384311,
+      "grad_norm": 0.6505454182624817,
+      "learning_rate": 9.170563174176645e-05,
+      "loss": 2.5423,
+      "step": 10577
+    },
+    {
+      "epoch": 0.8536841255750142,
+      "grad_norm": 0.7267098426818848,
+      "learning_rate": 9.168989934535586e-05,
+      "loss": 2.5687,
+      "step": 10578
+    },
+    {
+      "epoch": 0.8537648293115971,
+      "grad_norm": 0.7264497876167297,
+      "learning_rate": 9.167416715605476e-05,
+      "loss": 2.5165,
+      "step": 10579
+    },
+    {
+      "epoch": 0.8538455330481801,
+      "grad_norm": 0.7473852634429932,
+      "learning_rate": 9.165843517425509e-05,
+      "loss": 2.5837,
+      "step": 10580
+    },
+    {
+      "epoch": 0.8539262367847631,
+      "grad_norm": 0.7249133586883545,
+      "learning_rate": 9.164270340034906e-05,
+      "loss": 2.5805,
+      "step": 10581
+    },
+    {
+      "epoch": 0.8540069405213462,
+      "grad_norm": 0.7463760375976562,
+      "learning_rate": 9.162697183472875e-05,
+      "loss": 2.5067,
+      "step": 10582
+    },
+    {
+      "epoch": 0.8540876442579292,
+      "grad_norm": 0.7125511169433594,
+      "learning_rate": 9.161124047778614e-05,
+      "loss": 2.5093,
+      "step": 10583
+    },
+    {
+      "epoch": 0.8541683479945121,
+      "grad_norm": 0.7247455716133118,
+      "learning_rate": 9.159550932991335e-05,
+      "loss": 2.5356,
+      "step": 10584
+    },
+    {
+      "epoch": 0.8542490517310951,
+      "grad_norm": 0.7593860030174255,
+      "learning_rate": 9.157977839150246e-05,
+      "loss": 2.5477,
+      "step": 10585
+    },
+    {
+      "epoch": 0.8543297554676782,
+      "grad_norm": 0.6758295297622681,
+      "learning_rate": 9.156404766294547e-05,
+      "loss": 2.4748,
+      "step": 10586
+    },
+    {
+      "epoch": 0.8544104592042612,
+      "grad_norm": 0.7114073634147644,
+      "learning_rate": 9.154831714463447e-05,
+      "loss": 2.5479,
+      "step": 10587
+    },
+    {
+      "epoch": 0.8544911629408442,
+      "grad_norm": 0.6881263256072998,
+      "learning_rate": 9.153258683696156e-05,
+      "loss": 2.5471,
+      "step": 10588
+    },
+    {
+      "epoch": 0.8545718666774271,
+      "grad_norm": 0.6509317755699158,
+      "learning_rate": 9.151685674031866e-05,
+      "loss": 2.5239,
+      "step": 10589
+    },
+    {
+      "epoch": 0.8546525704140102,
+      "grad_norm": 0.7754644751548767,
+      "learning_rate": 9.150112685509787e-05,
+      "loss": 2.5572,
+      "step": 10590
+    },
+    {
+      "epoch": 0.8547332741505932,
+      "grad_norm": 0.707080602645874,
+      "learning_rate": 9.148539718169118e-05,
+      "loss": 2.5572,
+      "step": 10591
+    },
+    {
+      "epoch": 0.8548139778871762,
+      "grad_norm": 0.6996685266494751,
+      "learning_rate": 9.146966772049073e-05,
+      "loss": 2.4968,
+      "step": 10592
+    },
+    {
+      "epoch": 0.8548946816237591,
+      "grad_norm": 0.6830589771270752,
+      "learning_rate": 9.145393847188841e-05,
+      "loss": 2.5795,
+      "step": 10593
+    },
+    {
+      "epoch": 0.8549753853603422,
+      "grad_norm": 0.7507784366607666,
+      "learning_rate": 9.143820943627628e-05,
+      "loss": 2.6135,
+      "step": 10594
+    },
+    {
+      "epoch": 0.8550560890969252,
+      "grad_norm": 0.673218309879303,
+      "learning_rate": 9.142248061404638e-05,
+      "loss": 2.5875,
+      "step": 10595
+    },
+    {
+      "epoch": 0.8551367928335082,
+      "grad_norm": 0.6861804723739624,
+      "learning_rate": 9.140675200559065e-05,
+      "loss": 2.5892,
+      "step": 10596
+    },
+    {
+      "epoch": 0.8552174965700912,
+      "grad_norm": 0.6928709149360657,
+      "learning_rate": 9.139102361130114e-05,
+      "loss": 2.5303,
+      "step": 10597
+    },
+    {
+      "epoch": 0.8552982003066743,
+      "grad_norm": 0.6958343386650085,
+      "learning_rate": 9.137529543156986e-05,
+      "loss": 2.5567,
+      "step": 10598
+    },
+    {
+      "epoch": 0.8553789040432572,
+      "grad_norm": 0.703845739364624,
+      "learning_rate": 9.135956746678873e-05,
+      "loss": 2.5215,
+      "step": 10599
+    },
+    {
+      "epoch": 0.8554596077798402,
+      "grad_norm": 0.7108649015426636,
+      "learning_rate": 9.134383971734975e-05,
+      "loss": 2.5687,
+      "step": 10600
+    },
+    {
+      "epoch": 0.8555403115164232,
+      "grad_norm": 0.7249850034713745,
+      "learning_rate": 9.132811218364495e-05,
+      "loss": 2.565,
+      "step": 10601
+    },
+    {
+      "epoch": 0.8556210152530063,
+      "grad_norm": 0.7060014009475708,
+      "learning_rate": 9.131238486606623e-05,
+      "loss": 2.5366,
+      "step": 10602
+    },
+    {
+      "epoch": 0.8557017189895892,
+      "grad_norm": 0.6915088891983032,
+      "learning_rate": 9.129665776500559e-05,
+      "loss": 2.527,
+      "step": 10603
+    },
+    {
+      "epoch": 0.8557824227261722,
+      "grad_norm": 0.7226938605308533,
+      "learning_rate": 9.128093088085503e-05,
+      "loss": 2.5999,
+      "step": 10604
+    },
+    {
+      "epoch": 0.8558631264627552,
+      "grad_norm": 0.6802428364753723,
+      "learning_rate": 9.126520421400641e-05,
+      "loss": 2.4788,
+      "step": 10605
+    },
+    {
+      "epoch": 0.8559438301993383,
+      "grad_norm": 0.7855350375175476,
+      "learning_rate": 9.124947776485175e-05,
+      "loss": 2.5349,
+      "step": 10606
+    },
+    {
+      "epoch": 0.8560245339359213,
+      "grad_norm": 0.6758337020874023,
+      "learning_rate": 9.123375153378296e-05,
+      "loss": 2.5874,
+      "step": 10607
+    },
+    {
+      "epoch": 0.8561052376725042,
+      "grad_norm": 0.675061821937561,
+      "learning_rate": 9.121802552119206e-05,
+      "loss": 2.5343,
+      "step": 10608
+    },
+    {
+      "epoch": 0.8561859414090872,
+      "grad_norm": 0.7044726014137268,
+      "learning_rate": 9.120229972747087e-05,
+      "loss": 2.5361,
+      "step": 10609
+    },
+    {
+      "epoch": 0.8562666451456702,
+      "grad_norm": 0.6324402689933777,
+      "learning_rate": 9.118657415301137e-05,
+      "loss": 2.5039,
+      "step": 10610
+    },
+    {
+      "epoch": 0.8563473488822533,
+      "grad_norm": 0.6621509790420532,
+      "learning_rate": 9.11708487982055e-05,
+      "loss": 2.5346,
+      "step": 10611
+    },
+    {
+      "epoch": 0.8564280526188363,
+      "grad_norm": 0.6709887981414795,
+      "learning_rate": 9.115512366344516e-05,
+      "loss": 2.5409,
+      "step": 10612
+    },
+    {
+      "epoch": 0.8565087563554192,
+      "grad_norm": 0.7237712740898132,
+      "learning_rate": 9.113939874912223e-05,
+      "loss": 2.5051,
+      "step": 10613
+    },
+    {
+      "epoch": 0.8565894600920022,
+      "grad_norm": 0.6646109223365784,
+      "learning_rate": 9.11236740556287e-05,
+      "loss": 2.5866,
+      "step": 10614
+    },
+    {
+      "epoch": 0.8566701638285853,
+      "grad_norm": 0.7131930589675903,
+      "learning_rate": 9.110794958335637e-05,
+      "loss": 2.5472,
+      "step": 10615
+    },
+    {
+      "epoch": 0.8567508675651683,
+      "grad_norm": 0.6662428975105286,
+      "learning_rate": 9.109222533269715e-05,
+      "loss": 2.4863,
+      "step": 10616
+    },
+    {
+      "epoch": 0.8568315713017512,
+      "grad_norm": 0.6527226567268372,
+      "learning_rate": 9.107650130404304e-05,
+      "loss": 2.5594,
+      "step": 10617
+    },
+    {
+      "epoch": 0.8569122750383342,
+      "grad_norm": 0.6639060378074646,
+      "learning_rate": 9.106077749778578e-05,
+      "loss": 2.5519,
+      "step": 10618
+    },
+    {
+      "epoch": 0.8569929787749173,
+      "grad_norm": 0.7088096737861633,
+      "learning_rate": 9.104505391431734e-05,
+      "loss": 2.5404,
+      "step": 10619
+    },
+    {
+      "epoch": 0.8570736825115003,
+      "grad_norm": 0.7155873775482178,
+      "learning_rate": 9.102933055402957e-05,
+      "loss": 2.5636,
+      "step": 10620
+    },
+    {
+      "epoch": 0.8571543862480833,
+      "grad_norm": 0.6522316932678223,
+      "learning_rate": 9.101360741731431e-05,
+      "loss": 2.5216,
+      "step": 10621
+    },
+    {
+      "epoch": 0.8572350899846662,
+      "grad_norm": 0.6515649557113647,
+      "learning_rate": 9.099788450456345e-05,
+      "loss": 2.5804,
+      "step": 10622
+    },
+    {
+      "epoch": 0.8573157937212493,
+      "grad_norm": 0.6791853904724121,
+      "learning_rate": 9.098216181616883e-05,
+      "loss": 2.5353,
+      "step": 10623
+    },
+    {
+      "epoch": 0.8573964974578323,
+      "grad_norm": 0.6946877241134644,
+      "learning_rate": 9.096643935252236e-05,
+      "loss": 2.5492,
+      "step": 10624
+    },
+    {
+      "epoch": 0.8574772011944153,
+      "grad_norm": 0.7235898375511169,
+      "learning_rate": 9.095071711401581e-05,
+      "loss": 2.5178,
+      "step": 10625
+    },
+    {
+      "epoch": 0.8575579049309983,
+      "grad_norm": 0.6740610003471375,
+      "learning_rate": 9.093499510104102e-05,
+      "loss": 2.5699,
+      "step": 10626
+    },
+    {
+      "epoch": 0.8576386086675813,
+      "grad_norm": 0.7441792488098145,
+      "learning_rate": 9.091927331398988e-05,
+      "loss": 2.579,
+      "step": 10627
+    },
+    {
+      "epoch": 0.8577193124041643,
+      "grad_norm": 0.6986937522888184,
+      "learning_rate": 9.090355175325416e-05,
+      "loss": 2.5556,
+      "step": 10628
+    },
+    {
+      "epoch": 0.8578000161407473,
+      "grad_norm": 0.6960151791572571,
+      "learning_rate": 9.08878304192257e-05,
+      "loss": 2.5448,
+      "step": 10629
+    },
+    {
+      "epoch": 0.8578807198773303,
+      "grad_norm": 0.6376819014549255,
+      "learning_rate": 9.087210931229636e-05,
+      "loss": 2.4636,
+      "step": 10630
+    },
+    {
+      "epoch": 0.8579614236139134,
+      "grad_norm": 0.752473771572113,
+      "learning_rate": 9.08563884328579e-05,
+      "loss": 2.5451,
+      "step": 10631
+    },
+    {
+      "epoch": 0.8580421273504963,
+      "grad_norm": 0.6879361867904663,
+      "learning_rate": 9.084066778130213e-05,
+      "loss": 2.5365,
+      "step": 10632
+    },
+    {
+      "epoch": 0.8581228310870793,
+      "grad_norm": 0.6630483865737915,
+      "learning_rate": 9.082494735802091e-05,
+      "loss": 2.5085,
+      "step": 10633
+    },
+    {
+      "epoch": 0.8582035348236623,
+      "grad_norm": 0.689602792263031,
+      "learning_rate": 9.080922716340594e-05,
+      "loss": 2.5087,
+      "step": 10634
+    },
+    {
+      "epoch": 0.8582842385602454,
+      "grad_norm": 0.7333599925041199,
+      "learning_rate": 9.079350719784905e-05,
+      "loss": 2.5476,
+      "step": 10635
+    },
+    {
+      "epoch": 0.8583649422968284,
+      "grad_norm": 0.6895802021026611,
+      "learning_rate": 9.077778746174204e-05,
+      "loss": 2.5099,
+      "step": 10636
+    },
+    {
+      "epoch": 0.8584456460334113,
+      "grad_norm": 0.7202162146568298,
+      "learning_rate": 9.076206795547668e-05,
+      "loss": 2.5197,
+      "step": 10637
+    },
+    {
+      "epoch": 0.8585263497699943,
+      "grad_norm": 0.6454200148582458,
+      "learning_rate": 9.074634867944472e-05,
+      "loss": 2.5303,
+      "step": 10638
+    },
+    {
+      "epoch": 0.8586070535065774,
+      "grad_norm": 0.6842506527900696,
+      "learning_rate": 9.073062963403795e-05,
+      "loss": 2.5051,
+      "step": 10639
+    },
+    {
+      "epoch": 0.8586877572431604,
+      "grad_norm": 0.6979129314422607,
+      "learning_rate": 9.071491081964815e-05,
+      "loss": 2.5209,
+      "step": 10640
+    },
+    {
+      "epoch": 0.8587684609797434,
+      "grad_norm": 0.6851540803909302,
+      "learning_rate": 9.0699192236667e-05,
+      "loss": 2.5003,
+      "step": 10641
+    },
+    {
+      "epoch": 0.8588491647163263,
+      "grad_norm": 0.7528585195541382,
+      "learning_rate": 9.068347388548627e-05,
+      "loss": 2.5524,
+      "step": 10642
+    },
+    {
+      "epoch": 0.8589298684529094,
+      "grad_norm": 0.6297397613525391,
+      "learning_rate": 9.06677557664978e-05,
+      "loss": 2.5412,
+      "step": 10643
+    },
+    {
+      "epoch": 0.8590105721894924,
+      "grad_norm": 0.7034026980400085,
+      "learning_rate": 9.06520378800932e-05,
+      "loss": 2.4958,
+      "step": 10644
+    },
+    {
+      "epoch": 0.8590912759260754,
+      "grad_norm": 0.690258800983429,
+      "learning_rate": 9.063632022666425e-05,
+      "loss": 2.4894,
+      "step": 10645
+    },
+    {
+      "epoch": 0.8591719796626583,
+      "grad_norm": 0.6449949145317078,
+      "learning_rate": 9.06206028066027e-05,
+      "loss": 2.507,
+      "step": 10646
+    },
+    {
+      "epoch": 0.8592526833992414,
+      "grad_norm": 0.6328588724136353,
+      "learning_rate": 9.060488562030023e-05,
+      "loss": 2.5503,
+      "step": 10647
+    },
+    {
+      "epoch": 0.8593333871358244,
+      "grad_norm": 0.6570547819137573,
+      "learning_rate": 9.058916866814858e-05,
+      "loss": 2.4993,
+      "step": 10648
+    },
+    {
+      "epoch": 0.8594140908724074,
+      "grad_norm": 0.7689602375030518,
+      "learning_rate": 9.057345195053945e-05,
+      "loss": 2.5498,
+      "step": 10649
+    },
+    {
+      "epoch": 0.8594947946089904,
+      "grad_norm": 0.6727081537246704,
+      "learning_rate": 9.055773546786454e-05,
+      "loss": 2.5172,
+      "step": 10650
+    },
+    {
+      "epoch": 0.8595754983455735,
+      "grad_norm": 0.694722056388855,
+      "learning_rate": 9.054201922051552e-05,
+      "loss": 2.5485,
+      "step": 10651
+    },
+    {
+      "epoch": 0.8596562020821564,
+      "grad_norm": 0.6638815999031067,
+      "learning_rate": 9.052630320888411e-05,
+      "loss": 2.5134,
+      "step": 10652
+    },
+    {
+      "epoch": 0.8597369058187394,
+      "grad_norm": 0.6600833535194397,
+      "learning_rate": 9.0510587433362e-05,
+      "loss": 2.5206,
+      "step": 10653
+    },
+    {
+      "epoch": 0.8598176095553224,
+      "grad_norm": 0.7193894386291504,
+      "learning_rate": 9.049487189434084e-05,
+      "loss": 2.5485,
+      "step": 10654
+    },
+    {
+      "epoch": 0.8598983132919055,
+      "grad_norm": 0.6651753187179565,
+      "learning_rate": 9.047915659221233e-05,
+      "loss": 2.5703,
+      "step": 10655
+    },
+    {
+      "epoch": 0.8599790170284884,
+      "grad_norm": 0.7346364855766296,
+      "learning_rate": 9.046344152736815e-05,
+      "loss": 2.5301,
+      "step": 10656
+    },
+    {
+      "epoch": 0.8600597207650714,
+      "grad_norm": 0.6681811809539795,
+      "learning_rate": 9.04477267001999e-05,
+      "loss": 2.5124,
+      "step": 10657
+    },
+    {
+      "epoch": 0.8601404245016544,
+      "grad_norm": 0.6928461790084839,
+      "learning_rate": 9.043201211109929e-05,
+      "loss": 2.5153,
+      "step": 10658
+    },
+    {
+      "epoch": 0.8602211282382374,
+      "grad_norm": 0.6957700252532959,
+      "learning_rate": 9.041629776045797e-05,
+      "loss": 2.4697,
+      "step": 10659
+    },
+    {
+      "epoch": 0.8603018319748205,
+      "grad_norm": 0.6361939311027527,
+      "learning_rate": 9.040058364866752e-05,
+      "loss": 2.5162,
+      "step": 10660
+    },
+    {
+      "epoch": 0.8603825357114034,
+      "grad_norm": 0.6827390193939209,
+      "learning_rate": 9.038486977611964e-05,
+      "loss": 2.4856,
+      "step": 10661
+    },
+    {
+      "epoch": 0.8604632394479864,
+      "grad_norm": 0.6638801097869873,
+      "learning_rate": 9.036915614320595e-05,
+      "loss": 2.5224,
+      "step": 10662
+    },
+    {
+      "epoch": 0.8605439431845694,
+      "grad_norm": 0.7249652743339539,
+      "learning_rate": 9.035344275031802e-05,
+      "loss": 2.5461,
+      "step": 10663
+    },
+    {
+      "epoch": 0.8606246469211525,
+      "grad_norm": 0.6693316102027893,
+      "learning_rate": 9.033772959784754e-05,
+      "loss": 2.5676,
+      "step": 10664
+    },
+    {
+      "epoch": 0.8607053506577355,
+      "grad_norm": 0.6787340641021729,
+      "learning_rate": 9.032201668618614e-05,
+      "loss": 2.5374,
+      "step": 10665
+    },
+    {
+      "epoch": 0.8607860543943184,
+      "grad_norm": 0.6581670641899109,
+      "learning_rate": 9.030630401572533e-05,
+      "loss": 2.5052,
+      "step": 10666
+    },
+    {
+      "epoch": 0.8608667581309014,
+      "grad_norm": 0.6975873112678528,
+      "learning_rate": 9.029059158685675e-05,
+      "loss": 2.4823,
+      "step": 10667
+    },
+    {
+      "epoch": 0.8609474618674845,
+      "grad_norm": 0.6632521748542786,
+      "learning_rate": 9.027487939997201e-05,
+      "loss": 2.5992,
+      "step": 10668
+    },
+    {
+      "epoch": 0.8610281656040675,
+      "grad_norm": 0.6793977618217468,
+      "learning_rate": 9.025916745546276e-05,
+      "loss": 2.5308,
+      "step": 10669
+    },
+    {
+      "epoch": 0.8611088693406505,
+      "grad_norm": 0.6499481797218323,
+      "learning_rate": 9.024345575372046e-05,
+      "loss": 2.4964,
+      "step": 10670
+    },
+    {
+      "epoch": 0.8611895730772334,
+      "grad_norm": 0.6858868598937988,
+      "learning_rate": 9.022774429513677e-05,
+      "loss": 2.5388,
+      "step": 10671
+    },
+    {
+      "epoch": 0.8612702768138165,
+      "grad_norm": 0.7586160898208618,
+      "learning_rate": 9.021203308010324e-05,
+      "loss": 2.5166,
+      "step": 10672
+    },
+    {
+      "epoch": 0.8613509805503995,
+      "grad_norm": 0.7179701328277588,
+      "learning_rate": 9.019632210901141e-05,
+      "loss": 2.5501,
+      "step": 10673
+    },
+    {
+      "epoch": 0.8614316842869825,
+      "grad_norm": 0.6830369830131531,
+      "learning_rate": 9.018061138225287e-05,
+      "loss": 2.4956,
+      "step": 10674
+    },
+    {
+      "epoch": 0.8615123880235654,
+      "grad_norm": 0.6710512042045593,
+      "learning_rate": 9.01649009002192e-05,
+      "loss": 2.5722,
+      "step": 10675
+    },
+    {
+      "epoch": 0.8615930917601485,
+      "grad_norm": 0.640011727809906,
+      "learning_rate": 9.014919066330186e-05,
+      "loss": 2.5197,
+      "step": 10676
+    },
+    {
+      "epoch": 0.8616737954967315,
+      "grad_norm": 0.6803860664367676,
+      "learning_rate": 9.013348067189245e-05,
+      "loss": 2.4794,
+      "step": 10677
+    },
+    {
+      "epoch": 0.8617544992333145,
+      "grad_norm": 0.6734865307807922,
+      "learning_rate": 9.011777092638251e-05,
+      "loss": 2.5831,
+      "step": 10678
+    },
+    {
+      "epoch": 0.8618352029698975,
+      "grad_norm": 0.6525718569755554,
+      "learning_rate": 9.010206142716353e-05,
+      "loss": 2.4925,
+      "step": 10679
+    },
+    {
+      "epoch": 0.8619159067064806,
+      "grad_norm": 0.6886672377586365,
+      "learning_rate": 9.008635217462706e-05,
+      "loss": 2.491,
+      "step": 10680
+    },
+    {
+      "epoch": 0.8619966104430635,
+      "grad_norm": 0.6397131085395813,
+      "learning_rate": 9.007064316916461e-05,
+      "loss": 2.4684,
+      "step": 10681
+    },
+    {
+      "epoch": 0.8620773141796465,
+      "grad_norm": 0.6308462023735046,
+      "learning_rate": 9.005493441116768e-05,
+      "loss": 2.504,
+      "step": 10682
+    },
+    {
+      "epoch": 0.8621580179162295,
+      "grad_norm": 0.7223808169364929,
+      "learning_rate": 9.003922590102778e-05,
+      "loss": 2.5342,
+      "step": 10683
+    },
+    {
+      "epoch": 0.8622387216528126,
+      "grad_norm": 0.687515914440155,
+      "learning_rate": 9.002351763913642e-05,
+      "loss": 2.4822,
+      "step": 10684
+    },
+    {
+      "epoch": 0.8623194253893955,
+      "grad_norm": 0.6888468265533447,
+      "learning_rate": 9.00078096258851e-05,
+      "loss": 2.5497,
+      "step": 10685
+    },
+    {
+      "epoch": 0.8624001291259785,
+      "grad_norm": 0.7429301738739014,
+      "learning_rate": 8.999210186166525e-05,
+      "loss": 2.624,
+      "step": 10686
+    },
+    {
+      "epoch": 0.8624808328625615,
+      "grad_norm": 0.6901945471763611,
+      "learning_rate": 8.997639434686839e-05,
+      "loss": 2.5268,
+      "step": 10687
+    },
+    {
+      "epoch": 0.8625615365991446,
+      "grad_norm": 0.7396681308746338,
+      "learning_rate": 8.9960687081886e-05,
+      "loss": 2.5427,
+      "step": 10688
+    },
+    {
+      "epoch": 0.8626422403357276,
+      "grad_norm": 0.6825531125068665,
+      "learning_rate": 8.99449800671095e-05,
+      "loss": 2.5722,
+      "step": 10689
+    },
+    {
+      "epoch": 0.8627229440723105,
+      "grad_norm": 0.6719860434532166,
+      "learning_rate": 8.992927330293039e-05,
+      "loss": 2.4939,
+      "step": 10690
+    },
+    {
+      "epoch": 0.8628036478088935,
+      "grad_norm": 0.644567608833313,
+      "learning_rate": 8.991356678974017e-05,
+      "loss": 2.5495,
+      "step": 10691
+    },
+    {
+      "epoch": 0.8628843515454766,
+      "grad_norm": 0.7066643834114075,
+      "learning_rate": 8.989786052793015e-05,
+      "loss": 2.5508,
+      "step": 10692
+    },
+    {
+      "epoch": 0.8629650552820596,
+      "grad_norm": 0.6697196364402771,
+      "learning_rate": 8.988215451789187e-05,
+      "loss": 2.5231,
+      "step": 10693
+    },
+    {
+      "epoch": 0.8630457590186426,
+      "grad_norm": 0.7143658399581909,
+      "learning_rate": 8.986644876001681e-05,
+      "loss": 2.5368,
+      "step": 10694
+    },
+    {
+      "epoch": 0.8631264627552255,
+      "grad_norm": 0.7597684264183044,
+      "learning_rate": 8.985074325469628e-05,
+      "loss": 2.5983,
+      "step": 10695
+    },
+    {
+      "epoch": 0.8632071664918086,
+      "grad_norm": 0.7418014407157898,
+      "learning_rate": 8.983503800232176e-05,
+      "loss": 2.5736,
+      "step": 10696
+    },
+    {
+      "epoch": 0.8632878702283916,
+      "grad_norm": 0.654435932636261,
+      "learning_rate": 8.981933300328468e-05,
+      "loss": 2.5389,
+      "step": 10697
+    },
+    {
+      "epoch": 0.8633685739649746,
+      "grad_norm": 0.658203661441803,
+      "learning_rate": 8.980362825797643e-05,
+      "loss": 2.5204,
+      "step": 10698
+    },
+    {
+      "epoch": 0.8634492777015575,
+      "grad_norm": 0.7132784724235535,
+      "learning_rate": 8.97879237667884e-05,
+      "loss": 2.4982,
+      "step": 10699
+    },
+    {
+      "epoch": 0.8635299814381406,
+      "grad_norm": 0.6901868581771851,
+      "learning_rate": 8.9772219530112e-05,
+      "loss": 2.5599,
+      "step": 10700
+    },
+    {
+      "epoch": 0.8636106851747236,
+      "grad_norm": 0.6241179704666138,
+      "learning_rate": 8.975651554833869e-05,
+      "loss": 2.5185,
+      "step": 10701
+    },
+    {
+      "epoch": 0.8636913889113066,
+      "grad_norm": 0.693692147731781,
+      "learning_rate": 8.974081182185974e-05,
+      "loss": 2.506,
+      "step": 10702
+    },
+    {
+      "epoch": 0.8637720926478896,
+      "grad_norm": 0.6699246168136597,
+      "learning_rate": 8.972510835106658e-05,
+      "loss": 2.557,
+      "step": 10703
+    },
+    {
+      "epoch": 0.8638527963844727,
+      "grad_norm": 0.7339062094688416,
+      "learning_rate": 8.970940513635059e-05,
+      "loss": 2.5614,
+      "step": 10704
+    },
+    {
+      "epoch": 0.8639335001210556,
+      "grad_norm": 0.7558815479278564,
+      "learning_rate": 8.969370217810311e-05,
+      "loss": 2.5949,
+      "step": 10705
+    },
+    {
+      "epoch": 0.8640142038576386,
+      "grad_norm": 0.6992602348327637,
+      "learning_rate": 8.96779994767155e-05,
+      "loss": 2.4755,
+      "step": 10706
+    },
+    {
+      "epoch": 0.8640949075942216,
+      "grad_norm": 0.6836397647857666,
+      "learning_rate": 8.966229703257915e-05,
+      "loss": 2.5172,
+      "step": 10707
+    },
+    {
+      "epoch": 0.8641756113308047,
+      "grad_norm": 0.7054563760757446,
+      "learning_rate": 8.964659484608537e-05,
+      "loss": 2.5186,
+      "step": 10708
+    },
+    {
+      "epoch": 0.8642563150673876,
+      "grad_norm": 0.7096611261367798,
+      "learning_rate": 8.963089291762551e-05,
+      "loss": 2.5157,
+      "step": 10709
+    },
+    {
+      "epoch": 0.8643370188039706,
+      "grad_norm": 0.657465934753418,
+      "learning_rate": 8.961519124759094e-05,
+      "loss": 2.5332,
+      "step": 10710
+    },
+    {
+      "epoch": 0.8644177225405536,
+      "grad_norm": 0.7490121126174927,
+      "learning_rate": 8.959948983637291e-05,
+      "loss": 2.512,
+      "step": 10711
+    },
+    {
+      "epoch": 0.8644984262771366,
+      "grad_norm": 0.7074166536331177,
+      "learning_rate": 8.958378868436279e-05,
+      "loss": 2.4745,
+      "step": 10712
+    },
+    {
+      "epoch": 0.8645791300137197,
+      "grad_norm": 0.7496227025985718,
+      "learning_rate": 8.956808779195188e-05,
+      "loss": 2.5533,
+      "step": 10713
+    },
+    {
+      "epoch": 0.8646598337503026,
+      "grad_norm": 0.6624657511711121,
+      "learning_rate": 8.95523871595315e-05,
+      "loss": 2.5346,
+      "step": 10714
+    },
+    {
+      "epoch": 0.8647405374868856,
+      "grad_norm": 0.6829125881195068,
+      "learning_rate": 8.953668678749292e-05,
+      "loss": 2.558,
+      "step": 10715
+    },
+    {
+      "epoch": 0.8648212412234686,
+      "grad_norm": 0.6954498887062073,
+      "learning_rate": 8.952098667622745e-05,
+      "loss": 2.5617,
+      "step": 10716
+    },
+    {
+      "epoch": 0.8649019449600517,
+      "grad_norm": 0.6722636818885803,
+      "learning_rate": 8.950528682612645e-05,
+      "loss": 2.5565,
+      "step": 10717
+    },
+    {
+      "epoch": 0.8649826486966347,
+      "grad_norm": 0.6793767213821411,
+      "learning_rate": 8.948958723758107e-05,
+      "loss": 2.5803,
+      "step": 10718
+    },
+    {
+      "epoch": 0.8650633524332176,
+      "grad_norm": 0.7159373760223389,
+      "learning_rate": 8.947388791098266e-05,
+      "loss": 2.5465,
+      "step": 10719
+    },
+    {
+      "epoch": 0.8651440561698006,
+      "grad_norm": 0.6823835372924805,
+      "learning_rate": 8.945818884672253e-05,
+      "loss": 2.5079,
+      "step": 10720
+    },
+    {
+      "epoch": 0.8652247599063837,
+      "grad_norm": 0.7521452903747559,
+      "learning_rate": 8.944249004519185e-05,
+      "loss": 2.5628,
+      "step": 10721
+    },
+    {
+      "epoch": 0.8653054636429667,
+      "grad_norm": 0.6774886846542358,
+      "learning_rate": 8.94267915067819e-05,
+      "loss": 2.6042,
+      "step": 10722
+    },
+    {
+      "epoch": 0.8653861673795497,
+      "grad_norm": 0.6915935277938843,
+      "learning_rate": 8.941109323188398e-05,
+      "loss": 2.5563,
+      "step": 10723
+    },
+    {
+      "epoch": 0.8654668711161326,
+      "grad_norm": 0.6609061360359192,
+      "learning_rate": 8.939539522088927e-05,
+      "loss": 2.5083,
+      "step": 10724
+    },
+    {
+      "epoch": 0.8655475748527157,
+      "grad_norm": 0.6457223892211914,
+      "learning_rate": 8.937969747418903e-05,
+      "loss": 2.573,
+      "step": 10725
+    },
+    {
+      "epoch": 0.8656282785892987,
+      "grad_norm": 0.6960360407829285,
+      "learning_rate": 8.936399999217455e-05,
+      "loss": 2.516,
+      "step": 10726
+    },
+    {
+      "epoch": 0.8657089823258817,
+      "grad_norm": 0.7269721627235413,
+      "learning_rate": 8.934830277523693e-05,
+      "loss": 2.5932,
+      "step": 10727
+    },
+    {
+      "epoch": 0.8657896860624646,
+      "grad_norm": 0.7057532668113708,
+      "learning_rate": 8.933260582376745e-05,
+      "loss": 2.5022,
+      "step": 10728
+    },
+    {
+      "epoch": 0.8658703897990477,
+      "grad_norm": 0.6698749661445618,
+      "learning_rate": 8.931690913815735e-05,
+      "loss": 2.5357,
+      "step": 10729
+    },
+    {
+      "epoch": 0.8659510935356307,
+      "grad_norm": 0.6616599559783936,
+      "learning_rate": 8.930121271879777e-05,
+      "loss": 2.4776,
+      "step": 10730
+    },
+    {
+      "epoch": 0.8660317972722137,
+      "grad_norm": 0.7457093000411987,
+      "learning_rate": 8.928551656607993e-05,
+      "loss": 2.5799,
+      "step": 10731
+    },
+    {
+      "epoch": 0.8661125010087967,
+      "grad_norm": 0.7199469804763794,
+      "learning_rate": 8.926982068039505e-05,
+      "loss": 2.5278,
+      "step": 10732
+    },
+    {
+      "epoch": 0.8661932047453798,
+      "grad_norm": 0.7579182386398315,
+      "learning_rate": 8.925412506213428e-05,
+      "loss": 2.5227,
+      "step": 10733
+    },
+    {
+      "epoch": 0.8662739084819627,
+      "grad_norm": 0.687455952167511,
+      "learning_rate": 8.92384297116888e-05,
+      "loss": 2.5099,
+      "step": 10734
+    },
+    {
+      "epoch": 0.8663546122185457,
+      "grad_norm": 0.7616521120071411,
+      "learning_rate": 8.922273462944978e-05,
+      "loss": 2.598,
+      "step": 10735
+    },
+    {
+      "epoch": 0.8664353159551287,
+      "grad_norm": 0.6730697751045227,
+      "learning_rate": 8.920703981580842e-05,
+      "loss": 2.5517,
+      "step": 10736
+    },
+    {
+      "epoch": 0.8665160196917118,
+      "grad_norm": 0.6769895553588867,
+      "learning_rate": 8.91913452711558e-05,
+      "loss": 2.5535,
+      "step": 10737
+    },
+    {
+      "epoch": 0.8665967234282947,
+      "grad_norm": 0.6284549832344055,
+      "learning_rate": 8.917565099588312e-05,
+      "loss": 2.4597,
+      "step": 10738
+    },
+    {
+      "epoch": 0.8666774271648777,
+      "grad_norm": 0.6900805830955505,
+      "learning_rate": 8.915995699038152e-05,
+      "loss": 2.5236,
+      "step": 10739
+    },
+    {
+      "epoch": 0.8667581309014607,
+      "grad_norm": 0.6842896938323975,
+      "learning_rate": 8.914426325504211e-05,
+      "loss": 2.5199,
+      "step": 10740
+    },
+    {
+      "epoch": 0.8668388346380438,
+      "grad_norm": 0.6637243628501892,
+      "learning_rate": 8.912856979025604e-05,
+      "loss": 2.5368,
+      "step": 10741
+    },
+    {
+      "epoch": 0.8669195383746268,
+      "grad_norm": 0.7474464178085327,
+      "learning_rate": 8.911287659641449e-05,
+      "loss": 2.4902,
+      "step": 10742
+    },
+    {
+      "epoch": 0.8670002421112097,
+      "grad_norm": 0.6977849006652832,
+      "learning_rate": 8.909718367390843e-05,
+      "loss": 2.5034,
+      "step": 10743
+    },
+    {
+      "epoch": 0.8670809458477927,
+      "grad_norm": 0.6968807578086853,
+      "learning_rate": 8.908149102312907e-05,
+      "loss": 2.5396,
+      "step": 10744
+    },
+    {
+      "epoch": 0.8671616495843758,
+      "grad_norm": 0.6656209230422974,
+      "learning_rate": 8.906579864446755e-05,
+      "loss": 2.5702,
+      "step": 10745
+    },
+    {
+      "epoch": 0.8672423533209588,
+      "grad_norm": 0.7079079151153564,
+      "learning_rate": 8.905010653831486e-05,
+      "loss": 2.5344,
+      "step": 10746
+    },
+    {
+      "epoch": 0.8673230570575418,
+      "grad_norm": 0.7423387765884399,
+      "learning_rate": 8.903441470506214e-05,
+      "loss": 2.5635,
+      "step": 10747
+    },
+    {
+      "epoch": 0.8674037607941247,
+      "grad_norm": 0.6607224941253662,
+      "learning_rate": 8.901872314510046e-05,
+      "loss": 2.54,
+      "step": 10748
+    },
+    {
+      "epoch": 0.8674844645307078,
+      "grad_norm": 0.6646947860717773,
+      "learning_rate": 8.900303185882095e-05,
+      "loss": 2.4661,
+      "step": 10749
+    },
+    {
+      "epoch": 0.8675651682672908,
+      "grad_norm": 0.6943496465682983,
+      "learning_rate": 8.89873408466146e-05,
+      "loss": 2.5213,
+      "step": 10750
+    },
+    {
+      "epoch": 0.8676458720038738,
+      "grad_norm": 0.7048123478889465,
+      "learning_rate": 8.89716501088725e-05,
+      "loss": 2.5529,
+      "step": 10751
+    },
+    {
+      "epoch": 0.8677265757404568,
+      "grad_norm": 0.654617428779602,
+      "learning_rate": 8.895595964598574e-05,
+      "loss": 2.5535,
+      "step": 10752
+    },
+    {
+      "epoch": 0.8678072794770398,
+      "grad_norm": 0.672063410282135,
+      "learning_rate": 8.894026945834531e-05,
+      "loss": 2.5279,
+      "step": 10753
+    },
+    {
+      "epoch": 0.8678879832136228,
+      "grad_norm": 0.7134148478507996,
+      "learning_rate": 8.892457954634225e-05,
+      "loss": 2.5403,
+      "step": 10754
+    },
+    {
+      "epoch": 0.8679686869502058,
+      "grad_norm": 0.6457598805427551,
+      "learning_rate": 8.890888991036768e-05,
+      "loss": 2.515,
+      "step": 10755
+    },
+    {
+      "epoch": 0.8680493906867888,
+      "grad_norm": 0.6725220084190369,
+      "learning_rate": 8.889320055081252e-05,
+      "loss": 2.4829,
+      "step": 10756
+    },
+    {
+      "epoch": 0.8681300944233719,
+      "grad_norm": 0.6425862312316895,
+      "learning_rate": 8.887751146806785e-05,
+      "loss": 2.4965,
+      "step": 10757
+    },
+    {
+      "epoch": 0.8682107981599548,
+      "grad_norm": 0.6654682755470276,
+      "learning_rate": 8.886182266252468e-05,
+      "loss": 2.48,
+      "step": 10758
+    },
+    {
+      "epoch": 0.8682915018965378,
+      "grad_norm": 0.7102493643760681,
+      "learning_rate": 8.884613413457398e-05,
+      "loss": 2.5415,
+      "step": 10759
+    },
+    {
+      "epoch": 0.8683722056331208,
+      "grad_norm": 0.6996567249298096,
+      "learning_rate": 8.883044588460677e-05,
+      "loss": 2.542,
+      "step": 10760
+    },
+    {
+      "epoch": 0.8684529093697038,
+      "grad_norm": 0.7011905312538147,
+      "learning_rate": 8.881475791301405e-05,
+      "loss": 2.5391,
+      "step": 10761
+    },
+    {
+      "epoch": 0.8685336131062869,
+      "grad_norm": 0.6508356928825378,
+      "learning_rate": 8.879907022018686e-05,
+      "loss": 2.4892,
+      "step": 10762
+    },
+    {
+      "epoch": 0.8686143168428698,
+      "grad_norm": 0.7104009985923767,
+      "learning_rate": 8.878338280651605e-05,
+      "loss": 2.5152,
+      "step": 10763
+    },
+    {
+      "epoch": 0.8686950205794528,
+      "grad_norm": 0.6501138210296631,
+      "learning_rate": 8.876769567239268e-05,
+      "loss": 2.5767,
+      "step": 10764
+    },
+    {
+      "epoch": 0.8687757243160358,
+      "grad_norm": 0.6463173031806946,
+      "learning_rate": 8.875200881820771e-05,
+      "loss": 2.4758,
+      "step": 10765
+    },
+    {
+      "epoch": 0.8688564280526189,
+      "grad_norm": 0.6494991779327393,
+      "learning_rate": 8.873632224435206e-05,
+      "loss": 2.5364,
+      "step": 10766
+    },
+    {
+      "epoch": 0.8689371317892018,
+      "grad_norm": 0.6926043033599854,
+      "learning_rate": 8.872063595121671e-05,
+      "loss": 2.5288,
+      "step": 10767
+    },
+    {
+      "epoch": 0.8690178355257848,
+      "grad_norm": 0.7076035737991333,
+      "learning_rate": 8.870494993919261e-05,
+      "loss": 2.5118,
+      "step": 10768
+    },
+    {
+      "epoch": 0.8690985392623678,
+      "grad_norm": 0.6456892490386963,
+      "learning_rate": 8.868926420867068e-05,
+      "loss": 2.4957,
+      "step": 10769
+    },
+    {
+      "epoch": 0.8691792429989509,
+      "grad_norm": 0.6585200428962708,
+      "learning_rate": 8.867357876004183e-05,
+      "loss": 2.5049,
+      "step": 10770
+    },
+    {
+      "epoch": 0.8692599467355339,
+      "grad_norm": 0.6893252730369568,
+      "learning_rate": 8.865789359369706e-05,
+      "loss": 2.4808,
+      "step": 10771
+    },
+    {
+      "epoch": 0.8693406504721168,
+      "grad_norm": 0.6700639724731445,
+      "learning_rate": 8.864220871002719e-05,
+      "loss": 2.5475,
+      "step": 10772
+    },
+    {
+      "epoch": 0.8694213542086998,
+      "grad_norm": 0.6551913619041443,
+      "learning_rate": 8.862652410942315e-05,
+      "loss": 2.5063,
+      "step": 10773
+    },
+    {
+      "epoch": 0.8695020579452829,
+      "grad_norm": 0.6870427131652832,
+      "learning_rate": 8.86108397922759e-05,
+      "loss": 2.5785,
+      "step": 10774
+    },
+    {
+      "epoch": 0.8695827616818659,
+      "grad_norm": 0.6489934325218201,
+      "learning_rate": 8.859515575897626e-05,
+      "loss": 2.5584,
+      "step": 10775
+    },
+    {
+      "epoch": 0.8696634654184489,
+      "grad_norm": 0.6726663112640381,
+      "learning_rate": 8.857947200991517e-05,
+      "loss": 2.5707,
+      "step": 10776
+    },
+    {
+      "epoch": 0.8697441691550318,
+      "grad_norm": 0.7696183323860168,
+      "learning_rate": 8.856378854548347e-05,
+      "loss": 2.501,
+      "step": 10777
+    },
+    {
+      "epoch": 0.8698248728916149,
+      "grad_norm": 0.7002642154693604,
+      "learning_rate": 8.854810536607212e-05,
+      "loss": 2.5792,
+      "step": 10778
+    },
+    {
+      "epoch": 0.8699055766281979,
+      "grad_norm": 0.6429435610771179,
+      "learning_rate": 8.853242247207185e-05,
+      "loss": 2.5463,
+      "step": 10779
+    },
+    {
+      "epoch": 0.8699862803647809,
+      "grad_norm": 0.7006216645240784,
+      "learning_rate": 8.851673986387358e-05,
+      "loss": 2.5698,
+      "step": 10780
+    },
+    {
+      "epoch": 0.8700669841013638,
+      "grad_norm": 0.7053292989730835,
+      "learning_rate": 8.850105754186824e-05,
+      "loss": 2.5468,
+      "step": 10781
+    },
+    {
+      "epoch": 0.8701476878379469,
+      "grad_norm": 0.6592122912406921,
+      "learning_rate": 8.848537550644654e-05,
+      "loss": 2.5271,
+      "step": 10782
+    },
+    {
+      "epoch": 0.8702283915745299,
+      "grad_norm": 0.679132342338562,
+      "learning_rate": 8.846969375799941e-05,
+      "loss": 2.5281,
+      "step": 10783
+    },
+    {
+      "epoch": 0.8703090953111129,
+      "grad_norm": 0.6868568062782288,
+      "learning_rate": 8.845401229691765e-05,
+      "loss": 2.5415,
+      "step": 10784
+    },
+    {
+      "epoch": 0.8703897990476959,
+      "grad_norm": 0.7060674428939819,
+      "learning_rate": 8.843833112359208e-05,
+      "loss": 2.5649,
+      "step": 10785
+    },
+    {
+      "epoch": 0.870470502784279,
+      "grad_norm": 0.6663981676101685,
+      "learning_rate": 8.842265023841352e-05,
+      "loss": 2.5055,
+      "step": 10786
+    },
+    {
+      "epoch": 0.8705512065208619,
+      "grad_norm": 0.7095218896865845,
+      "learning_rate": 8.840696964177282e-05,
+      "loss": 2.5442,
+      "step": 10787
+    },
+    {
+      "epoch": 0.8706319102574449,
+      "grad_norm": 0.6884104013442993,
+      "learning_rate": 8.839128933406069e-05,
+      "loss": 2.5285,
+      "step": 10788
+    },
+    {
+      "epoch": 0.8707126139940279,
+      "grad_norm": 0.6427462697029114,
+      "learning_rate": 8.837560931566798e-05,
+      "loss": 2.5197,
+      "step": 10789
+    },
+    {
+      "epoch": 0.870793317730611,
+      "grad_norm": 0.6870493292808533,
+      "learning_rate": 8.835992958698548e-05,
+      "loss": 2.4937,
+      "step": 10790
+    },
+    {
+      "epoch": 0.870874021467194,
+      "grad_norm": 0.7006319761276245,
+      "learning_rate": 8.834425014840398e-05,
+      "loss": 2.5148,
+      "step": 10791
+    },
+    {
+      "epoch": 0.8709547252037769,
+      "grad_norm": 0.690601646900177,
+      "learning_rate": 8.83285710003142e-05,
+      "loss": 2.5454,
+      "step": 10792
+    },
+    {
+      "epoch": 0.8710354289403599,
+      "grad_norm": 0.7205955982208252,
+      "learning_rate": 8.831289214310695e-05,
+      "loss": 2.5221,
+      "step": 10793
+    },
+    {
+      "epoch": 0.871116132676943,
+      "grad_norm": 0.7134295105934143,
+      "learning_rate": 8.8297213577173e-05,
+      "loss": 2.5626,
+      "step": 10794
+    },
+    {
+      "epoch": 0.871196836413526,
+      "grad_norm": 0.6560496091842651,
+      "learning_rate": 8.828153530290307e-05,
+      "loss": 2.5408,
+      "step": 10795
+    },
+    {
+      "epoch": 0.8712775401501089,
+      "grad_norm": 0.7055882215499878,
+      "learning_rate": 8.82658573206879e-05,
+      "loss": 2.5173,
+      "step": 10796
+    },
+    {
+      "epoch": 0.8713582438866919,
+      "grad_norm": 0.6751883029937744,
+      "learning_rate": 8.825017963091827e-05,
+      "loss": 2.5378,
+      "step": 10797
+    },
+    {
+      "epoch": 0.871438947623275,
+      "grad_norm": 0.6794824600219727,
+      "learning_rate": 8.823450223398485e-05,
+      "loss": 2.592,
+      "step": 10798
+    },
+    {
+      "epoch": 0.871519651359858,
+      "grad_norm": 0.675729513168335,
+      "learning_rate": 8.821882513027838e-05,
+      "loss": 2.5253,
+      "step": 10799
+    },
+    {
+      "epoch": 0.871600355096441,
+      "grad_norm": 0.7185894250869751,
+      "learning_rate": 8.820314832018962e-05,
+      "loss": 2.5073,
+      "step": 10800
+    },
+    {
+      "epoch": 0.8716810588330239,
+      "grad_norm": 0.6605187654495239,
+      "learning_rate": 8.818747180410921e-05,
+      "loss": 2.5141,
+      "step": 10801
+    },
+    {
+      "epoch": 0.871761762569607,
+      "grad_norm": 0.6955205798149109,
+      "learning_rate": 8.817179558242788e-05,
+      "loss": 2.5313,
+      "step": 10802
+    },
+    {
+      "epoch": 0.87184246630619,
+      "grad_norm": 0.6307928562164307,
+      "learning_rate": 8.815611965553638e-05,
+      "loss": 2.4975,
+      "step": 10803
+    },
+    {
+      "epoch": 0.871923170042773,
+      "grad_norm": 0.7283728122711182,
+      "learning_rate": 8.814044402382527e-05,
+      "loss": 2.4623,
+      "step": 10804
+    },
+    {
+      "epoch": 0.872003873779356,
+      "grad_norm": 0.7019702792167664,
+      "learning_rate": 8.81247686876853e-05,
+      "loss": 2.4755,
+      "step": 10805
+    },
+    {
+      "epoch": 0.872084577515939,
+      "grad_norm": 0.6769137382507324,
+      "learning_rate": 8.81090936475072e-05,
+      "loss": 2.59,
+      "step": 10806
+    },
+    {
+      "epoch": 0.872165281252522,
+      "grad_norm": 0.6185588836669922,
+      "learning_rate": 8.80934189036815e-05,
+      "loss": 2.5308,
+      "step": 10807
+    },
+    {
+      "epoch": 0.872245984989105,
+      "grad_norm": 0.7127000689506531,
+      "learning_rate": 8.807774445659894e-05,
+      "loss": 2.5301,
+      "step": 10808
+    },
+    {
+      "epoch": 0.872326688725688,
+      "grad_norm": 0.7039114236831665,
+      "learning_rate": 8.806207030665016e-05,
+      "loss": 2.5176,
+      "step": 10809
+    },
+    {
+      "epoch": 0.8724073924622711,
+      "grad_norm": 0.6763370633125305,
+      "learning_rate": 8.804639645422582e-05,
+      "loss": 2.5324,
+      "step": 10810
+    },
+    {
+      "epoch": 0.872488096198854,
+      "grad_norm": 0.7546409368515015,
+      "learning_rate": 8.803072289971648e-05,
+      "loss": 2.5446,
+      "step": 10811
+    },
+    {
+      "epoch": 0.872568799935437,
+      "grad_norm": 0.6916004419326782,
+      "learning_rate": 8.801504964351284e-05,
+      "loss": 2.5056,
+      "step": 10812
+    },
+    {
+      "epoch": 0.87264950367202,
+      "grad_norm": 0.7108416557312012,
+      "learning_rate": 8.799937668600552e-05,
+      "loss": 2.5966,
+      "step": 10813
+    },
+    {
+      "epoch": 0.872730207408603,
+      "grad_norm": 0.7146576046943665,
+      "learning_rate": 8.798370402758506e-05,
+      "loss": 2.5152,
+      "step": 10814
+    },
+    {
+      "epoch": 0.872810911145186,
+      "grad_norm": 0.6708142757415771,
+      "learning_rate": 8.796803166864211e-05,
+      "loss": 2.5248,
+      "step": 10815
+    },
+    {
+      "epoch": 0.872891614881769,
+      "grad_norm": 0.6687600612640381,
+      "learning_rate": 8.795235960956729e-05,
+      "loss": 2.4451,
+      "step": 10816
+    },
+    {
+      "epoch": 0.872972318618352,
+      "grad_norm": 0.724012553691864,
+      "learning_rate": 8.793668785075114e-05,
+      "loss": 2.4816,
+      "step": 10817
+    },
+    {
+      "epoch": 0.873053022354935,
+      "grad_norm": 0.6938769221305847,
+      "learning_rate": 8.792101639258426e-05,
+      "loss": 2.5435,
+      "step": 10818
+    },
+    {
+      "epoch": 0.8731337260915181,
+      "grad_norm": 0.7066235542297363,
+      "learning_rate": 8.790534523545724e-05,
+      "loss": 2.5167,
+      "step": 10819
+    },
+    {
+      "epoch": 0.873214429828101,
+      "grad_norm": 0.7129037380218506,
+      "learning_rate": 8.788967437976062e-05,
+      "loss": 2.5079,
+      "step": 10820
+    },
+    {
+      "epoch": 0.873295133564684,
+      "grad_norm": 0.6949728727340698,
+      "learning_rate": 8.787400382588497e-05,
+      "loss": 2.5564,
+      "step": 10821
+    },
+    {
+      "epoch": 0.873375837301267,
+      "grad_norm": 0.7924233675003052,
+      "learning_rate": 8.785833357422088e-05,
+      "loss": 2.5748,
+      "step": 10822
+    },
+    {
+      "epoch": 0.8734565410378501,
+      "grad_norm": 0.7486331462860107,
+      "learning_rate": 8.784266362515882e-05,
+      "loss": 2.565,
+      "step": 10823
+    },
+    {
+      "epoch": 0.8735372447744331,
+      "grad_norm": 0.7036460638046265,
+      "learning_rate": 8.782699397908935e-05,
+      "loss": 2.5101,
+      "step": 10824
+    },
+    {
+      "epoch": 0.873617948511016,
+      "grad_norm": 0.6691471338272095,
+      "learning_rate": 8.781132463640302e-05,
+      "loss": 2.5262,
+      "step": 10825
+    },
+    {
+      "epoch": 0.873698652247599,
+      "grad_norm": 0.6836682558059692,
+      "learning_rate": 8.779565559749037e-05,
+      "loss": 2.5651,
+      "step": 10826
+    },
+    {
+      "epoch": 0.8737793559841821,
+      "grad_norm": 0.6634507775306702,
+      "learning_rate": 8.777998686274185e-05,
+      "loss": 2.5383,
+      "step": 10827
+    },
+    {
+      "epoch": 0.8738600597207651,
+      "grad_norm": 0.6903105974197388,
+      "learning_rate": 8.7764318432548e-05,
+      "loss": 2.5659,
+      "step": 10828
+    },
+    {
+      "epoch": 0.8739407634573481,
+      "grad_norm": 0.737859308719635,
+      "learning_rate": 8.774865030729937e-05,
+      "loss": 2.5859,
+      "step": 10829
+    },
+    {
+      "epoch": 0.874021467193931,
+      "grad_norm": 0.696843683719635,
+      "learning_rate": 8.773298248738633e-05,
+      "loss": 2.5244,
+      "step": 10830
+    },
+    {
+      "epoch": 0.8741021709305141,
+      "grad_norm": 0.7342235445976257,
+      "learning_rate": 8.771731497319946e-05,
+      "loss": 2.5073,
+      "step": 10831
+    },
+    {
+      "epoch": 0.8741828746670971,
+      "grad_norm": 0.6676939725875854,
+      "learning_rate": 8.770164776512926e-05,
+      "loss": 2.5408,
+      "step": 10832
+    },
+    {
+      "epoch": 0.8742635784036801,
+      "grad_norm": 0.6957886219024658,
+      "learning_rate": 8.768598086356608e-05,
+      "loss": 2.5566,
+      "step": 10833
+    },
+    {
+      "epoch": 0.874344282140263,
+      "grad_norm": 0.6938990950584412,
+      "learning_rate": 8.767031426890046e-05,
+      "loss": 2.517,
+      "step": 10834
+    },
+    {
+      "epoch": 0.8744249858768461,
+      "grad_norm": 0.8387169241905212,
+      "learning_rate": 8.765464798152286e-05,
+      "loss": 2.5507,
+      "step": 10835
+    },
+    {
+      "epoch": 0.8745056896134291,
+      "grad_norm": 0.6396276354789734,
+      "learning_rate": 8.763898200182368e-05,
+      "loss": 2.5063,
+      "step": 10836
+    },
+    {
+      "epoch": 0.8745863933500121,
+      "grad_norm": 0.7122719883918762,
+      "learning_rate": 8.762331633019339e-05,
+      "loss": 2.5816,
+      "step": 10837
+    },
+    {
+      "epoch": 0.8746670970865951,
+      "grad_norm": 0.6807141304016113,
+      "learning_rate": 8.760765096702244e-05,
+      "loss": 2.6004,
+      "step": 10838
+    },
+    {
+      "epoch": 0.8747478008231782,
+      "grad_norm": 0.6764848232269287,
+      "learning_rate": 8.759198591270117e-05,
+      "loss": 2.5303,
+      "step": 10839
+    },
+    {
+      "epoch": 0.8748285045597611,
+      "grad_norm": 0.718515932559967,
+      "learning_rate": 8.757632116762006e-05,
+      "loss": 2.5088,
+      "step": 10840
+    },
+    {
+      "epoch": 0.8749092082963441,
+      "grad_norm": 0.7084362506866455,
+      "learning_rate": 8.75606567321695e-05,
+      "loss": 2.5496,
+      "step": 10841
+    },
+    {
+      "epoch": 0.8749899120329271,
+      "grad_norm": 0.7191734910011292,
+      "learning_rate": 8.754499260673991e-05,
+      "loss": 2.5525,
+      "step": 10842
+    },
+    {
+      "epoch": 0.8750706157695102,
+      "grad_norm": 0.7167977094650269,
+      "learning_rate": 8.752932879172164e-05,
+      "loss": 2.5479,
+      "step": 10843
+    },
+    {
+      "epoch": 0.8751513195060932,
+      "grad_norm": 0.6994979381561279,
+      "learning_rate": 8.751366528750511e-05,
+      "loss": 2.4942,
+      "step": 10844
+    },
+    {
+      "epoch": 0.8752320232426761,
+      "grad_norm": 0.7192725539207458,
+      "learning_rate": 8.749800209448068e-05,
+      "loss": 2.5233,
+      "step": 10845
+    },
+    {
+      "epoch": 0.8753127269792591,
+      "grad_norm": 0.7728807330131531,
+      "learning_rate": 8.748233921303871e-05,
+      "loss": 2.5698,
+      "step": 10846
+    },
+    {
+      "epoch": 0.8753934307158422,
+      "grad_norm": 0.7305434942245483,
+      "learning_rate": 8.746667664356956e-05,
+      "loss": 2.5096,
+      "step": 10847
+    },
+    {
+      "epoch": 0.8754741344524252,
+      "grad_norm": 0.7117629051208496,
+      "learning_rate": 8.745101438646365e-05,
+      "loss": 2.5272,
+      "step": 10848
+    },
+    {
+      "epoch": 0.8755548381890081,
+      "grad_norm": 0.7180361151695251,
+      "learning_rate": 8.743535244211121e-05,
+      "loss": 2.4718,
+      "step": 10849
+    },
+    {
+      "epoch": 0.8756355419255911,
+      "grad_norm": 0.6419457793235779,
+      "learning_rate": 8.741969081090263e-05,
+      "loss": 2.5407,
+      "step": 10850
+    },
+    {
+      "epoch": 0.8757162456621742,
+      "grad_norm": 0.7928328514099121,
+      "learning_rate": 8.740402949322827e-05,
+      "loss": 2.488,
+      "step": 10851
+    },
+    {
+      "epoch": 0.8757969493987572,
+      "grad_norm": 0.7449139952659607,
+      "learning_rate": 8.738836848947839e-05,
+      "loss": 2.5943,
+      "step": 10852
+    },
+    {
+      "epoch": 0.8758776531353402,
+      "grad_norm": 0.7919576168060303,
+      "learning_rate": 8.737270780004334e-05,
+      "loss": 2.5556,
+      "step": 10853
+    },
+    {
+      "epoch": 0.8759583568719231,
+      "grad_norm": 0.6867526769638062,
+      "learning_rate": 8.735704742531346e-05,
+      "loss": 2.5395,
+      "step": 10854
+    },
+    {
+      "epoch": 0.8760390606085062,
+      "grad_norm": 0.7195394039154053,
+      "learning_rate": 8.734138736567896e-05,
+      "loss": 2.4404,
+      "step": 10855
+    },
+    {
+      "epoch": 0.8761197643450892,
+      "grad_norm": 0.68385910987854,
+      "learning_rate": 8.732572762153016e-05,
+      "loss": 2.502,
+      "step": 10856
+    },
+    {
+      "epoch": 0.8762004680816722,
+      "grad_norm": 0.6957393884658813,
+      "learning_rate": 8.731006819325739e-05,
+      "loss": 2.5788,
+      "step": 10857
+    },
+    {
+      "epoch": 0.8762811718182552,
+      "grad_norm": 0.6973037123680115,
+      "learning_rate": 8.729440908125092e-05,
+      "loss": 2.4927,
+      "step": 10858
+    },
+    {
+      "epoch": 0.8763618755548382,
+      "grad_norm": 0.6535985469818115,
+      "learning_rate": 8.727875028590095e-05,
+      "loss": 2.596,
+      "step": 10859
+    },
+    {
+      "epoch": 0.8764425792914212,
+      "grad_norm": 0.7447848320007324,
+      "learning_rate": 8.726309180759777e-05,
+      "loss": 2.5825,
+      "step": 10860
+    },
+    {
+      "epoch": 0.8765232830280042,
+      "grad_norm": 0.7155942320823669,
+      "learning_rate": 8.724743364673168e-05,
+      "loss": 2.5105,
+      "step": 10861
+    },
+    {
+      "epoch": 0.8766039867645872,
+      "grad_norm": 0.6664694547653198,
+      "learning_rate": 8.723177580369285e-05,
+      "loss": 2.5244,
+      "step": 10862
+    },
+    {
+      "epoch": 0.8766846905011701,
+      "grad_norm": 0.7437852025032043,
+      "learning_rate": 8.721611827887153e-05,
+      "loss": 2.534,
+      "step": 10863
+    },
+    {
+      "epoch": 0.8767653942377532,
+      "grad_norm": 0.6752577424049377,
+      "learning_rate": 8.7200461072658e-05,
+      "loss": 2.5025,
+      "step": 10864
+    },
+    {
+      "epoch": 0.8768460979743362,
+      "grad_norm": 0.7420764565467834,
+      "learning_rate": 8.718480418544241e-05,
+      "loss": 2.5261,
+      "step": 10865
+    },
+    {
+      "epoch": 0.8769268017109192,
+      "grad_norm": 0.669384777545929,
+      "learning_rate": 8.7169147617615e-05,
+      "loss": 2.5258,
+      "step": 10866
+    },
+    {
+      "epoch": 0.8770075054475022,
+      "grad_norm": 0.6649587750434875,
+      "learning_rate": 8.715349136956599e-05,
+      "loss": 2.5308,
+      "step": 10867
+    },
+    {
+      "epoch": 0.8770882091840853,
+      "grad_norm": 0.728922426700592,
+      "learning_rate": 8.713783544168552e-05,
+      "loss": 2.5251,
+      "step": 10868
+    },
+    {
+      "epoch": 0.8771689129206682,
+      "grad_norm": 0.6957671642303467,
+      "learning_rate": 8.712217983436384e-05,
+      "loss": 2.5818,
+      "step": 10869
+    },
+    {
+      "epoch": 0.8772496166572512,
+      "grad_norm": 0.6796830892562866,
+      "learning_rate": 8.710652454799108e-05,
+      "loss": 2.5122,
+      "step": 10870
+    },
+    {
+      "epoch": 0.8773303203938342,
+      "grad_norm": 0.7230980396270752,
+      "learning_rate": 8.709086958295746e-05,
+      "loss": 2.5836,
+      "step": 10871
+    },
+    {
+      "epoch": 0.8774110241304173,
+      "grad_norm": 0.6992264986038208,
+      "learning_rate": 8.707521493965309e-05,
+      "loss": 2.5907,
+      "step": 10872
+    },
+    {
+      "epoch": 0.8774917278670002,
+      "grad_norm": 0.7066535353660583,
+      "learning_rate": 8.705956061846816e-05,
+      "loss": 2.5508,
+      "step": 10873
+    },
+    {
+      "epoch": 0.8775724316035832,
+      "grad_norm": 0.6559327244758606,
+      "learning_rate": 8.704390661979283e-05,
+      "loss": 2.611,
+      "step": 10874
+    },
+    {
+      "epoch": 0.8776531353401662,
+      "grad_norm": 0.6673287749290466,
+      "learning_rate": 8.70282529440172e-05,
+      "loss": 2.5778,
+      "step": 10875
+    },
+    {
+      "epoch": 0.8777338390767493,
+      "grad_norm": 0.6715971231460571,
+      "learning_rate": 8.701259959153139e-05,
+      "loss": 2.5342,
+      "step": 10876
+    },
+    {
+      "epoch": 0.8778145428133323,
+      "grad_norm": 0.7456488609313965,
+      "learning_rate": 8.699694656272557e-05,
+      "loss": 2.5365,
+      "step": 10877
+    },
+    {
+      "epoch": 0.8778952465499152,
+      "grad_norm": 0.6658159494400024,
+      "learning_rate": 8.698129385798983e-05,
+      "loss": 2.4387,
+      "step": 10878
+    },
+    {
+      "epoch": 0.8779759502864982,
+      "grad_norm": 0.6653816103935242,
+      "learning_rate": 8.696564147771427e-05,
+      "loss": 2.5791,
+      "step": 10879
+    },
+    {
+      "epoch": 0.8780566540230813,
+      "grad_norm": 0.6763200163841248,
+      "learning_rate": 8.694998942228902e-05,
+      "loss": 2.5356,
+      "step": 10880
+    },
+    {
+      "epoch": 0.8781373577596643,
+      "grad_norm": 0.6534504890441895,
+      "learning_rate": 8.69343376921041e-05,
+      "loss": 2.5358,
+      "step": 10881
+    },
+    {
+      "epoch": 0.8782180614962473,
+      "grad_norm": 0.6341667771339417,
+      "learning_rate": 8.691868628754967e-05,
+      "loss": 2.4927,
+      "step": 10882
+    },
+    {
+      "epoch": 0.8782987652328302,
+      "grad_norm": 0.6215559244155884,
+      "learning_rate": 8.690303520901579e-05,
+      "loss": 2.4312,
+      "step": 10883
+    },
+    {
+      "epoch": 0.8783794689694133,
+      "grad_norm": 0.6705841422080994,
+      "learning_rate": 8.688738445689248e-05,
+      "loss": 2.4778,
+      "step": 10884
+    },
+    {
+      "epoch": 0.8784601727059963,
+      "grad_norm": 0.680275559425354,
+      "learning_rate": 8.687173403156982e-05,
+      "loss": 2.5577,
+      "step": 10885
+    },
+    {
+      "epoch": 0.8785408764425793,
+      "grad_norm": 0.6918728351593018,
+      "learning_rate": 8.685608393343789e-05,
+      "loss": 2.5212,
+      "step": 10886
+    },
+    {
+      "epoch": 0.8786215801791623,
+      "grad_norm": 0.623636782169342,
+      "learning_rate": 8.68404341628867e-05,
+      "loss": 2.5131,
+      "step": 10887
+    },
+    {
+      "epoch": 0.8787022839157453,
+      "grad_norm": 0.7200562357902527,
+      "learning_rate": 8.682478472030628e-05,
+      "loss": 2.5517,
+      "step": 10888
+    },
+    {
+      "epoch": 0.8787829876523283,
+      "grad_norm": 0.6902644634246826,
+      "learning_rate": 8.680913560608666e-05,
+      "loss": 2.511,
+      "step": 10889
+    },
+    {
+      "epoch": 0.8788636913889113,
+      "grad_norm": 0.6855802536010742,
+      "learning_rate": 8.679348682061792e-05,
+      "loss": 2.5169,
+      "step": 10890
+    },
+    {
+      "epoch": 0.8789443951254943,
+      "grad_norm": 0.7229284048080444,
+      "learning_rate": 8.677783836428995e-05,
+      "loss": 2.5634,
+      "step": 10891
+    },
+    {
+      "epoch": 0.8790250988620774,
+      "grad_norm": 0.6350376605987549,
+      "learning_rate": 8.676219023749281e-05,
+      "loss": 2.443,
+      "step": 10892
+    },
+    {
+      "epoch": 0.8791058025986603,
+      "grad_norm": 0.6884307265281677,
+      "learning_rate": 8.674654244061653e-05,
+      "loss": 2.524,
+      "step": 10893
+    },
+    {
+      "epoch": 0.8791865063352433,
+      "grad_norm": 0.6571067571640015,
+      "learning_rate": 8.673089497405102e-05,
+      "loss": 2.5322,
+      "step": 10894
+    },
+    {
+      "epoch": 0.8792672100718263,
+      "grad_norm": 0.7078021764755249,
+      "learning_rate": 8.67152478381863e-05,
+      "loss": 2.5317,
+      "step": 10895
+    },
+    {
+      "epoch": 0.8793479138084094,
+      "grad_norm": 0.6809059381484985,
+      "learning_rate": 8.669960103341236e-05,
+      "loss": 2.5767,
+      "step": 10896
+    },
+    {
+      "epoch": 0.8794286175449924,
+      "grad_norm": 0.7399441003799438,
+      "learning_rate": 8.66839545601191e-05,
+      "loss": 2.5194,
+      "step": 10897
+    },
+    {
+      "epoch": 0.8795093212815753,
+      "grad_norm": 0.6762270927429199,
+      "learning_rate": 8.66683084186965e-05,
+      "loss": 2.5306,
+      "step": 10898
+    },
+    {
+      "epoch": 0.8795900250181583,
+      "grad_norm": 0.7394620776176453,
+      "learning_rate": 8.665266260953455e-05,
+      "loss": 2.4516,
+      "step": 10899
+    },
+    {
+      "epoch": 0.8796707287547414,
+      "grad_norm": 0.6775416135787964,
+      "learning_rate": 8.663701713302309e-05,
+      "loss": 2.5574,
+      "step": 10900
+    },
+    {
+      "epoch": 0.8797514324913244,
+      "grad_norm": 0.7630520462989807,
+      "learning_rate": 8.66213719895521e-05,
+      "loss": 2.5516,
+      "step": 10901
+    },
+    {
+      "epoch": 0.8798321362279073,
+      "grad_norm": 0.6555768847465515,
+      "learning_rate": 8.660572717951149e-05,
+      "loss": 2.5267,
+      "step": 10902
+    },
+    {
+      "epoch": 0.8799128399644903,
+      "grad_norm": 0.6899500489234924,
+      "learning_rate": 8.659008270329119e-05,
+      "loss": 2.4938,
+      "step": 10903
+    },
+    {
+      "epoch": 0.8799935437010734,
+      "grad_norm": 0.6939221024513245,
+      "learning_rate": 8.657443856128107e-05,
+      "loss": 2.5358,
+      "step": 10904
+    },
+    {
+      "epoch": 0.8800742474376564,
+      "grad_norm": 0.6454630494117737,
+      "learning_rate": 8.655879475387102e-05,
+      "loss": 2.5528,
+      "step": 10905
+    },
+    {
+      "epoch": 0.8801549511742394,
+      "grad_norm": 0.7142425775527954,
+      "learning_rate": 8.654315128145099e-05,
+      "loss": 2.5668,
+      "step": 10906
+    },
+    {
+      "epoch": 0.8802356549108223,
+      "grad_norm": 0.7512764930725098,
+      "learning_rate": 8.652750814441075e-05,
+      "loss": 2.5224,
+      "step": 10907
+    },
+    {
+      "epoch": 0.8803163586474054,
+      "grad_norm": 0.6599575877189636,
+      "learning_rate": 8.651186534314026e-05,
+      "loss": 2.5363,
+      "step": 10908
+    },
+    {
+      "epoch": 0.8803970623839884,
+      "grad_norm": 0.6787410974502563,
+      "learning_rate": 8.649622287802935e-05,
+      "loss": 2.4587,
+      "step": 10909
+    },
+    {
+      "epoch": 0.8804777661205714,
+      "grad_norm": 0.7124783396720886,
+      "learning_rate": 8.648058074946786e-05,
+      "loss": 2.5842,
+      "step": 10910
+    },
+    {
+      "epoch": 0.8805584698571544,
+      "grad_norm": 0.6698839664459229,
+      "learning_rate": 8.646493895784562e-05,
+      "loss": 2.513,
+      "step": 10911
+    },
+    {
+      "epoch": 0.8806391735937374,
+      "grad_norm": 0.6660044193267822,
+      "learning_rate": 8.644929750355249e-05,
+      "loss": 2.4996,
+      "step": 10912
+    },
+    {
+      "epoch": 0.8807198773303204,
+      "grad_norm": 0.7060455083847046,
+      "learning_rate": 8.643365638697828e-05,
+      "loss": 2.5497,
+      "step": 10913
+    },
+    {
+      "epoch": 0.8808005810669034,
+      "grad_norm": 0.6835277676582336,
+      "learning_rate": 8.641801560851281e-05,
+      "loss": 2.5198,
+      "step": 10914
+    },
+    {
+      "epoch": 0.8808812848034864,
+      "grad_norm": 0.6994042992591858,
+      "learning_rate": 8.640237516854595e-05,
+      "loss": 2.5692,
+      "step": 10915
+    },
+    {
+      "epoch": 0.8809619885400694,
+      "grad_norm": 0.6583377718925476,
+      "learning_rate": 8.63867350674674e-05,
+      "loss": 2.5025,
+      "step": 10916
+    },
+    {
+      "epoch": 0.8810426922766524,
+      "grad_norm": 0.6882332563400269,
+      "learning_rate": 8.637109530566698e-05,
+      "loss": 2.5343,
+      "step": 10917
+    },
+    {
+      "epoch": 0.8811233960132354,
+      "grad_norm": 0.6329876184463501,
+      "learning_rate": 8.635545588353449e-05,
+      "loss": 2.5335,
+      "step": 10918
+    },
+    {
+      "epoch": 0.8812040997498184,
+      "grad_norm": 0.713196337223053,
+      "learning_rate": 8.633981680145975e-05,
+      "loss": 2.4814,
+      "step": 10919
+    },
+    {
+      "epoch": 0.8812848034864014,
+      "grad_norm": 0.7388820648193359,
+      "learning_rate": 8.632417805983246e-05,
+      "loss": 2.4927,
+      "step": 10920
+    },
+    {
+      "epoch": 0.8813655072229845,
+      "grad_norm": 0.7316160798072815,
+      "learning_rate": 8.63085396590424e-05,
+      "loss": 2.508,
+      "step": 10921
+    },
+    {
+      "epoch": 0.8814462109595674,
+      "grad_norm": 0.6690139174461365,
+      "learning_rate": 8.629290159947934e-05,
+      "loss": 2.5719,
+      "step": 10922
+    },
+    {
+      "epoch": 0.8815269146961504,
+      "grad_norm": 0.6369553208351135,
+      "learning_rate": 8.627726388153297e-05,
+      "loss": 2.5277,
+      "step": 10923
+    },
+    {
+      "epoch": 0.8816076184327334,
+      "grad_norm": 0.6870365738868713,
+      "learning_rate": 8.626162650559306e-05,
+      "loss": 2.4731,
+      "step": 10924
+    },
+    {
+      "epoch": 0.8816883221693165,
+      "grad_norm": 0.6890872716903687,
+      "learning_rate": 8.624598947204938e-05,
+      "loss": 2.5417,
+      "step": 10925
+    },
+    {
+      "epoch": 0.8817690259058995,
+      "grad_norm": 0.6548230051994324,
+      "learning_rate": 8.623035278129156e-05,
+      "loss": 2.4888,
+      "step": 10926
+    },
+    {
+      "epoch": 0.8818497296424824,
+      "grad_norm": 0.6835262775421143,
+      "learning_rate": 8.621471643370933e-05,
+      "loss": 2.531,
+      "step": 10927
+    },
+    {
+      "epoch": 0.8819304333790654,
+      "grad_norm": 0.6910626292228699,
+      "learning_rate": 8.619908042969243e-05,
+      "loss": 2.4864,
+      "step": 10928
+    },
+    {
+      "epoch": 0.8820111371156485,
+      "grad_norm": 0.6727725267410278,
+      "learning_rate": 8.618344476963049e-05,
+      "loss": 2.5063,
+      "step": 10929
+    },
+    {
+      "epoch": 0.8820918408522315,
+      "grad_norm": 0.7285245656967163,
+      "learning_rate": 8.616780945391323e-05,
+      "loss": 2.5036,
+      "step": 10930
+    },
+    {
+      "epoch": 0.8821725445888144,
+      "grad_norm": 0.6561840176582336,
+      "learning_rate": 8.615217448293035e-05,
+      "loss": 2.5152,
+      "step": 10931
+    },
+    {
+      "epoch": 0.8822532483253974,
+      "grad_norm": 0.6524627208709717,
+      "learning_rate": 8.613653985707144e-05,
+      "loss": 2.4827,
+      "step": 10932
+    },
+    {
+      "epoch": 0.8823339520619805,
+      "grad_norm": 0.6815671920776367,
+      "learning_rate": 8.612090557672619e-05,
+      "loss": 2.5385,
+      "step": 10933
+    },
+    {
+      "epoch": 0.8824146557985635,
+      "grad_norm": 0.7479865550994873,
+      "learning_rate": 8.610527164228429e-05,
+      "loss": 2.5311,
+      "step": 10934
+    },
+    {
+      "epoch": 0.8824953595351465,
+      "grad_norm": 0.699504554271698,
+      "learning_rate": 8.608963805413535e-05,
+      "loss": 2.5332,
+      "step": 10935
+    },
+    {
+      "epoch": 0.8825760632717294,
+      "grad_norm": 0.7081198692321777,
+      "learning_rate": 8.607400481266896e-05,
+      "loss": 2.5636,
+      "step": 10936
+    },
+    {
+      "epoch": 0.8826567670083125,
+      "grad_norm": 0.7020730972290039,
+      "learning_rate": 8.605837191827478e-05,
+      "loss": 2.498,
+      "step": 10937
+    },
+    {
+      "epoch": 0.8827374707448955,
+      "grad_norm": 0.8004096150398254,
+      "learning_rate": 8.604273937134242e-05,
+      "loss": 2.5352,
+      "step": 10938
+    },
+    {
+      "epoch": 0.8828181744814785,
+      "grad_norm": 0.6399645209312439,
+      "learning_rate": 8.602710717226147e-05,
+      "loss": 2.5673,
+      "step": 10939
+    },
+    {
+      "epoch": 0.8828988782180615,
+      "grad_norm": 0.683195173740387,
+      "learning_rate": 8.601147532142153e-05,
+      "loss": 2.4812,
+      "step": 10940
+    },
+    {
+      "epoch": 0.8829795819546445,
+      "grad_norm": 0.7783642411231995,
+      "learning_rate": 8.599584381921224e-05,
+      "loss": 2.4812,
+      "step": 10941
+    },
+    {
+      "epoch": 0.8830602856912275,
+      "grad_norm": 0.7107423543930054,
+      "learning_rate": 8.598021266602308e-05,
+      "loss": 2.5527,
+      "step": 10942
+    },
+    {
+      "epoch": 0.8831409894278105,
+      "grad_norm": 0.6419345140457153,
+      "learning_rate": 8.596458186224365e-05,
+      "loss": 2.5642,
+      "step": 10943
+    },
+    {
+      "epoch": 0.8832216931643935,
+      "grad_norm": 0.6897309422492981,
+      "learning_rate": 8.59489514082636e-05,
+      "loss": 2.5743,
+      "step": 10944
+    },
+    {
+      "epoch": 0.8833023969009766,
+      "grad_norm": 0.6901495456695557,
+      "learning_rate": 8.593332130447236e-05,
+      "loss": 2.5139,
+      "step": 10945
+    },
+    {
+      "epoch": 0.8833831006375595,
+      "grad_norm": 0.6865388751029968,
+      "learning_rate": 8.591769155125953e-05,
+      "loss": 2.5281,
+      "step": 10946
+    },
+    {
+      "epoch": 0.8834638043741425,
+      "grad_norm": 0.7070403099060059,
+      "learning_rate": 8.590206214901465e-05,
+      "loss": 2.4648,
+      "step": 10947
+    },
+    {
+      "epoch": 0.8835445081107255,
+      "grad_norm": 0.6846395134925842,
+      "learning_rate": 8.588643309812721e-05,
+      "loss": 2.4792,
+      "step": 10948
+    },
+    {
+      "epoch": 0.8836252118473086,
+      "grad_norm": 0.6875495314598083,
+      "learning_rate": 8.587080439898675e-05,
+      "loss": 2.5126,
+      "step": 10949
+    },
+    {
+      "epoch": 0.8837059155838916,
+      "grad_norm": 0.670098066329956,
+      "learning_rate": 8.58551760519828e-05,
+      "loss": 2.4922,
+      "step": 10950
+    },
+    {
+      "epoch": 0.8837866193204745,
+      "grad_norm": 0.6675527691841125,
+      "learning_rate": 8.583954805750487e-05,
+      "loss": 2.499,
+      "step": 10951
+    },
+    {
+      "epoch": 0.8838673230570575,
+      "grad_norm": 0.6694127321243286,
+      "learning_rate": 8.582392041594236e-05,
+      "loss": 2.5286,
+      "step": 10952
+    },
+    {
+      "epoch": 0.8839480267936406,
+      "grad_norm": 0.7291092872619629,
+      "learning_rate": 8.580829312768482e-05,
+      "loss": 2.5705,
+      "step": 10953
+    },
+    {
+      "epoch": 0.8840287305302236,
+      "grad_norm": 0.709904670715332,
+      "learning_rate": 8.579266619312174e-05,
+      "loss": 2.5238,
+      "step": 10954
+    },
+    {
+      "epoch": 0.8841094342668065,
+      "grad_norm": 0.7037622332572937,
+      "learning_rate": 8.577703961264254e-05,
+      "loss": 2.5491,
+      "step": 10955
+    },
+    {
+      "epoch": 0.8841901380033895,
+      "grad_norm": 0.7553049325942993,
+      "learning_rate": 8.576141338663668e-05,
+      "loss": 2.5643,
+      "step": 10956
+    },
+    {
+      "epoch": 0.8842708417399726,
+      "grad_norm": 0.7177377343177795,
+      "learning_rate": 8.574578751549364e-05,
+      "loss": 2.49,
+      "step": 10957
+    },
+    {
+      "epoch": 0.8843515454765556,
+      "grad_norm": 0.682668149471283,
+      "learning_rate": 8.573016199960283e-05,
+      "loss": 2.5221,
+      "step": 10958
+    },
+    {
+      "epoch": 0.8844322492131386,
+      "grad_norm": 0.7508956789970398,
+      "learning_rate": 8.571453683935366e-05,
+      "loss": 2.5766,
+      "step": 10959
+    },
+    {
+      "epoch": 0.8845129529497215,
+      "grad_norm": 0.6495946645736694,
+      "learning_rate": 8.569891203513562e-05,
+      "loss": 2.534,
+      "step": 10960
+    },
+    {
+      "epoch": 0.8845936566863046,
+      "grad_norm": 0.7362824082374573,
+      "learning_rate": 8.568328758733806e-05,
+      "loss": 2.4614,
+      "step": 10961
+    },
+    {
+      "epoch": 0.8846743604228876,
+      "grad_norm": 0.6571496725082397,
+      "learning_rate": 8.566766349635037e-05,
+      "loss": 2.4393,
+      "step": 10962
+    },
+    {
+      "epoch": 0.8847550641594706,
+      "grad_norm": 0.7088329195976257,
+      "learning_rate": 8.5652039762562e-05,
+      "loss": 2.5476,
+      "step": 10963
+    },
+    {
+      "epoch": 0.8848357678960536,
+      "grad_norm": 0.6414440274238586,
+      "learning_rate": 8.56364163863623e-05,
+      "loss": 2.4668,
+      "step": 10964
+    },
+    {
+      "epoch": 0.8849164716326365,
+      "grad_norm": 0.7333478331565857,
+      "learning_rate": 8.562079336814063e-05,
+      "loss": 2.5151,
+      "step": 10965
+    },
+    {
+      "epoch": 0.8849971753692196,
+      "grad_norm": 0.638038694858551,
+      "learning_rate": 8.560517070828638e-05,
+      "loss": 2.5063,
+      "step": 10966
+    },
+    {
+      "epoch": 0.8850778791058026,
+      "grad_norm": 0.638921320438385,
+      "learning_rate": 8.558954840718896e-05,
+      "loss": 2.4769,
+      "step": 10967
+    },
+    {
+      "epoch": 0.8851585828423856,
+      "grad_norm": 0.6923465728759766,
+      "learning_rate": 8.557392646523759e-05,
+      "loss": 2.5388,
+      "step": 10968
+    },
+    {
+      "epoch": 0.8852392865789686,
+      "grad_norm": 0.7095212936401367,
+      "learning_rate": 8.555830488282169e-05,
+      "loss": 2.4955,
+      "step": 10969
+    },
+    {
+      "epoch": 0.8853199903155516,
+      "grad_norm": 0.689908504486084,
+      "learning_rate": 8.554268366033065e-05,
+      "loss": 2.4998,
+      "step": 10970
+    },
+    {
+      "epoch": 0.8854006940521346,
+      "grad_norm": 0.6551975011825562,
+      "learning_rate": 8.552706279815366e-05,
+      "loss": 2.4965,
+      "step": 10971
+    },
+    {
+      "epoch": 0.8854813977887176,
+      "grad_norm": 0.7239118218421936,
+      "learning_rate": 8.551144229668012e-05,
+      "loss": 2.5785,
+      "step": 10972
+    },
+    {
+      "epoch": 0.8855621015253006,
+      "grad_norm": 0.6743230819702148,
+      "learning_rate": 8.549582215629932e-05,
+      "loss": 2.5146,
+      "step": 10973
+    },
+    {
+      "epoch": 0.8856428052618837,
+      "grad_norm": 0.6991584300994873,
+      "learning_rate": 8.548020237740052e-05,
+      "loss": 2.5524,
+      "step": 10974
+    },
+    {
+      "epoch": 0.8857235089984666,
+      "grad_norm": 0.6605305075645447,
+      "learning_rate": 8.546458296037304e-05,
+      "loss": 2.5505,
+      "step": 10975
+    },
+    {
+      "epoch": 0.8858042127350496,
+      "grad_norm": 0.7011568546295166,
+      "learning_rate": 8.54489639056062e-05,
+      "loss": 2.4381,
+      "step": 10976
+    },
+    {
+      "epoch": 0.8858849164716326,
+      "grad_norm": 0.7015339136123657,
+      "learning_rate": 8.543334521348916e-05,
+      "loss": 2.5432,
+      "step": 10977
+    },
+    {
+      "epoch": 0.8859656202082157,
+      "grad_norm": 0.6892278790473938,
+      "learning_rate": 8.541772688441124e-05,
+      "loss": 2.5286,
+      "step": 10978
+    },
+    {
+      "epoch": 0.8860463239447987,
+      "grad_norm": 0.6680187582969666,
+      "learning_rate": 8.540210891876168e-05,
+      "loss": 2.439,
+      "step": 10979
+    },
+    {
+      "epoch": 0.8861270276813816,
+      "grad_norm": 0.7043240666389465,
+      "learning_rate": 8.538649131692975e-05,
+      "loss": 2.5558,
+      "step": 10980
+    },
+    {
+      "epoch": 0.8862077314179646,
+      "grad_norm": 0.6940229535102844,
+      "learning_rate": 8.537087407930463e-05,
+      "loss": 2.5219,
+      "step": 10981
+    },
+    {
+      "epoch": 0.8862884351545477,
+      "grad_norm": 0.6571553945541382,
+      "learning_rate": 8.535525720627558e-05,
+      "loss": 2.5054,
+      "step": 10982
+    },
+    {
+      "epoch": 0.8863691388911307,
+      "grad_norm": 0.6846656203269958,
+      "learning_rate": 8.533964069823182e-05,
+      "loss": 2.497,
+      "step": 10983
+    },
+    {
+      "epoch": 0.8864498426277136,
+      "grad_norm": 0.6838627457618713,
+      "learning_rate": 8.53240245555625e-05,
+      "loss": 2.5495,
+      "step": 10984
+    },
+    {
+      "epoch": 0.8865305463642966,
+      "grad_norm": 0.6825091242790222,
+      "learning_rate": 8.530840877865687e-05,
+      "loss": 2.5656,
+      "step": 10985
+    },
+    {
+      "epoch": 0.8866112501008797,
+      "grad_norm": 0.7368674278259277,
+      "learning_rate": 8.529279336790414e-05,
+      "loss": 2.5378,
+      "step": 10986
+    },
+    {
+      "epoch": 0.8866919538374627,
+      "grad_norm": 0.7333693504333496,
+      "learning_rate": 8.527717832369338e-05,
+      "loss": 2.506,
+      "step": 10987
+    },
+    {
+      "epoch": 0.8867726575740457,
+      "grad_norm": 0.6623306274414062,
+      "learning_rate": 8.526156364641384e-05,
+      "loss": 2.4824,
+      "step": 10988
+    },
+    {
+      "epoch": 0.8868533613106286,
+      "grad_norm": 0.6863973140716553,
+      "learning_rate": 8.524594933645468e-05,
+      "loss": 2.536,
+      "step": 10989
+    },
+    {
+      "epoch": 0.8869340650472117,
+      "grad_norm": 0.6805100440979004,
+      "learning_rate": 8.523033539420501e-05,
+      "loss": 2.4954,
+      "step": 10990
+    },
+    {
+      "epoch": 0.8870147687837947,
+      "grad_norm": 0.6672216653823853,
+      "learning_rate": 8.521472182005399e-05,
+      "loss": 2.4893,
+      "step": 10991
+    },
+    {
+      "epoch": 0.8870954725203777,
+      "grad_norm": 0.7310158610343933,
+      "learning_rate": 8.519910861439079e-05,
+      "loss": 2.5317,
+      "step": 10992
+    },
+    {
+      "epoch": 0.8871761762569607,
+      "grad_norm": 0.6820743083953857,
+      "learning_rate": 8.518349577760445e-05,
+      "loss": 2.4482,
+      "step": 10993
+    },
+    {
+      "epoch": 0.8872568799935437,
+      "grad_norm": 0.6660269498825073,
+      "learning_rate": 8.516788331008411e-05,
+      "loss": 2.5353,
+      "step": 10994
+    },
+    {
+      "epoch": 0.8873375837301267,
+      "grad_norm": 0.676243007183075,
+      "learning_rate": 8.51522712122189e-05,
+      "loss": 2.531,
+      "step": 10995
+    },
+    {
+      "epoch": 0.8874182874667097,
+      "grad_norm": 0.6677152514457703,
+      "learning_rate": 8.513665948439796e-05,
+      "loss": 2.4732,
+      "step": 10996
+    },
+    {
+      "epoch": 0.8874989912032927,
+      "grad_norm": 0.7341045141220093,
+      "learning_rate": 8.512104812701027e-05,
+      "loss": 2.5668,
+      "step": 10997
+    },
+    {
+      "epoch": 0.8875796949398758,
+      "grad_norm": 0.6475326418876648,
+      "learning_rate": 8.510543714044496e-05,
+      "loss": 2.5026,
+      "step": 10998
+    },
+    {
+      "epoch": 0.8876603986764587,
+      "grad_norm": 0.7335529923439026,
+      "learning_rate": 8.50898265250911e-05,
+      "loss": 2.4946,
+      "step": 10999
+    },
+    {
+      "epoch": 0.8877411024130417,
+      "grad_norm": 0.760108232498169,
+      "learning_rate": 8.507421628133772e-05,
+      "loss": 2.5697,
+      "step": 11000
+    },
+    {
+      "epoch": 0.8877411024130417,
+      "eval_loss": 2.450413465499878,
+      "eval_runtime": 975.281,
+      "eval_samples_per_second": 2.686,
+      "eval_steps_per_second": 0.448,
+      "step": 11000
+    },
+    {
+      "epoch": 0.8878218061496247,
+      "grad_norm": 0.6420160531997681,
+      "learning_rate": 8.505860640957391e-05,
+      "loss": 2.5842,
+      "step": 11001
+    },
+    {
+      "epoch": 0.8879025098862078,
+      "grad_norm": 0.6625204086303711,
+      "learning_rate": 8.50429969101887e-05,
+      "loss": 2.4771,
+      "step": 11002
+    },
+    {
+      "epoch": 0.8879832136227908,
+      "grad_norm": 0.7430149912834167,
+      "learning_rate": 8.502738778357107e-05,
+      "loss": 2.5509,
+      "step": 11003
+    },
+    {
+      "epoch": 0.8880639173593737,
+      "grad_norm": 0.663624107837677,
+      "learning_rate": 8.501177903011008e-05,
+      "loss": 2.504,
+      "step": 11004
+    },
+    {
+      "epoch": 0.8881446210959567,
+      "grad_norm": 0.6638087630271912,
+      "learning_rate": 8.499617065019476e-05,
+      "loss": 2.492,
+      "step": 11005
+    },
+    {
+      "epoch": 0.8882253248325398,
+      "grad_norm": 0.7321780323982239,
+      "learning_rate": 8.498056264421406e-05,
+      "loss": 2.5808,
+      "step": 11006
+    },
+    {
+      "epoch": 0.8883060285691228,
+      "grad_norm": 0.7108619809150696,
+      "learning_rate": 8.4964955012557e-05,
+      "loss": 2.6185,
+      "step": 11007
+    },
+    {
+      "epoch": 0.8883867323057058,
+      "grad_norm": 0.6745856404304504,
+      "learning_rate": 8.494934775561258e-05,
+      "loss": 2.576,
+      "step": 11008
+    },
+    {
+      "epoch": 0.8884674360422887,
+      "grad_norm": 0.8002225756645203,
+      "learning_rate": 8.493374087376976e-05,
+      "loss": 2.5598,
+      "step": 11009
+    },
+    {
+      "epoch": 0.8885481397788718,
+      "grad_norm": 0.6848840713500977,
+      "learning_rate": 8.491813436741746e-05,
+      "loss": 2.5218,
+      "step": 11010
+    },
+    {
+      "epoch": 0.8886288435154548,
+      "grad_norm": 0.6464105248451233,
+      "learning_rate": 8.490252823694471e-05,
+      "loss": 2.5503,
+      "step": 11011
+    },
+    {
+      "epoch": 0.8887095472520378,
+      "grad_norm": 0.7165790796279907,
+      "learning_rate": 8.488692248274045e-05,
+      "loss": 2.5104,
+      "step": 11012
+    },
+    {
+      "epoch": 0.8887902509886207,
+      "grad_norm": 0.6832898259162903,
+      "learning_rate": 8.487131710519355e-05,
+      "loss": 2.5379,
+      "step": 11013
+    },
+    {
+      "epoch": 0.8888709547252038,
+      "grad_norm": 0.6992432475090027,
+      "learning_rate": 8.485571210469296e-05,
+      "loss": 2.5388,
+      "step": 11014
+    },
+    {
+      "epoch": 0.8889516584617868,
+      "grad_norm": 0.6410119533538818,
+      "learning_rate": 8.484010748162765e-05,
+      "loss": 2.5237,
+      "step": 11015
+    },
+    {
+      "epoch": 0.8890323621983698,
+      "grad_norm": 0.716248095035553,
+      "learning_rate": 8.482450323638647e-05,
+      "loss": 2.4977,
+      "step": 11016
+    },
+    {
+      "epoch": 0.8891130659349528,
+      "grad_norm": 0.6620567440986633,
+      "learning_rate": 8.480889936935833e-05,
+      "loss": 2.5088,
+      "step": 11017
+    },
+    {
+      "epoch": 0.8891937696715357,
+      "grad_norm": 0.7311015129089355,
+      "learning_rate": 8.479329588093217e-05,
+      "loss": 2.5547,
+      "step": 11018
+    },
+    {
+      "epoch": 0.8892744734081188,
+      "grad_norm": 0.757203996181488,
+      "learning_rate": 8.477769277149676e-05,
+      "loss": 2.5681,
+      "step": 11019
+    },
+    {
+      "epoch": 0.8893551771447018,
+      "grad_norm": 0.6941282153129578,
+      "learning_rate": 8.476209004144107e-05,
+      "loss": 2.5078,
+      "step": 11020
+    },
+    {
+      "epoch": 0.8894358808812848,
+      "grad_norm": 0.6381667256355286,
+      "learning_rate": 8.474648769115396e-05,
+      "loss": 2.5371,
+      "step": 11021
+    },
+    {
+      "epoch": 0.8895165846178678,
+      "grad_norm": 0.7978621125221252,
+      "learning_rate": 8.473088572102422e-05,
+      "loss": 2.5384,
+      "step": 11022
+    },
+    {
+      "epoch": 0.8895972883544508,
+      "grad_norm": 0.7229189872741699,
+      "learning_rate": 8.471528413144072e-05,
+      "loss": 2.5469,
+      "step": 11023
+    },
+    {
+      "epoch": 0.8896779920910338,
+      "grad_norm": 0.705545961856842,
+      "learning_rate": 8.469968292279231e-05,
+      "loss": 2.5281,
+      "step": 11024
+    },
+    {
+      "epoch": 0.8897586958276168,
+      "grad_norm": 0.7259972095489502,
+      "learning_rate": 8.468408209546777e-05,
+      "loss": 2.5485,
+      "step": 11025
+    },
+    {
+      "epoch": 0.8898393995641998,
+      "grad_norm": 0.6859608888626099,
+      "learning_rate": 8.466848164985594e-05,
+      "loss": 2.5548,
+      "step": 11026
+    },
+    {
+      "epoch": 0.8899201033007829,
+      "grad_norm": 0.7036644816398621,
+      "learning_rate": 8.465288158634565e-05,
+      "loss": 2.5159,
+      "step": 11027
+    },
+    {
+      "epoch": 0.8900008070373658,
+      "grad_norm": 0.6899380087852478,
+      "learning_rate": 8.463728190532569e-05,
+      "loss": 2.5037,
+      "step": 11028
+    },
+    {
+      "epoch": 0.8900815107739488,
+      "grad_norm": 0.7428410649299622,
+      "learning_rate": 8.462168260718477e-05,
+      "loss": 2.5074,
+      "step": 11029
+    },
+    {
+      "epoch": 0.8901622145105318,
+      "grad_norm": 0.6724158525466919,
+      "learning_rate": 8.460608369231173e-05,
+      "loss": 2.5544,
+      "step": 11030
+    },
+    {
+      "epoch": 0.8902429182471149,
+      "grad_norm": 0.6516450643539429,
+      "learning_rate": 8.459048516109535e-05,
+      "loss": 2.5152,
+      "step": 11031
+    },
+    {
+      "epoch": 0.8903236219836979,
+      "grad_norm": 0.7013405561447144,
+      "learning_rate": 8.457488701392434e-05,
+      "loss": 2.5116,
+      "step": 11032
+    },
+    {
+      "epoch": 0.8904043257202808,
+      "grad_norm": 0.7207479476928711,
+      "learning_rate": 8.455928925118747e-05,
+      "loss": 2.6041,
+      "step": 11033
+    },
+    {
+      "epoch": 0.8904850294568638,
+      "grad_norm": 0.69600510597229,
+      "learning_rate": 8.454369187327348e-05,
+      "loss": 2.5794,
+      "step": 11034
+    },
+    {
+      "epoch": 0.8905657331934469,
+      "grad_norm": 0.6831288933753967,
+      "learning_rate": 8.452809488057108e-05,
+      "loss": 2.4682,
+      "step": 11035
+    },
+    {
+      "epoch": 0.8906464369300299,
+      "grad_norm": 0.6978991627693176,
+      "learning_rate": 8.451249827346901e-05,
+      "loss": 2.4862,
+      "step": 11036
+    },
+    {
+      "epoch": 0.8907271406666128,
+      "grad_norm": 0.6772337555885315,
+      "learning_rate": 8.4496902052356e-05,
+      "loss": 2.5357,
+      "step": 11037
+    },
+    {
+      "epoch": 0.8908078444031958,
+      "grad_norm": 0.6735778450965881,
+      "learning_rate": 8.448130621762067e-05,
+      "loss": 2.5115,
+      "step": 11038
+    },
+    {
+      "epoch": 0.8908885481397789,
+      "grad_norm": 0.6695345044136047,
+      "learning_rate": 8.446571076965177e-05,
+      "loss": 2.5083,
+      "step": 11039
+    },
+    {
+      "epoch": 0.8909692518763619,
+      "grad_norm": 0.685343325138092,
+      "learning_rate": 8.445011570883796e-05,
+      "loss": 2.5221,
+      "step": 11040
+    },
+    {
+      "epoch": 0.8910499556129449,
+      "grad_norm": 0.7030319571495056,
+      "learning_rate": 8.443452103556792e-05,
+      "loss": 2.5708,
+      "step": 11041
+    },
+    {
+      "epoch": 0.8911306593495278,
+      "grad_norm": 0.6910343766212463,
+      "learning_rate": 8.441892675023029e-05,
+      "loss": 2.5373,
+      "step": 11042
+    },
+    {
+      "epoch": 0.8912113630861109,
+      "grad_norm": 0.7207868099212646,
+      "learning_rate": 8.440333285321374e-05,
+      "loss": 2.5862,
+      "step": 11043
+    },
+    {
+      "epoch": 0.8912920668226939,
+      "grad_norm": 0.6780788898468018,
+      "learning_rate": 8.438773934490692e-05,
+      "loss": 2.562,
+      "step": 11044
+    },
+    {
+      "epoch": 0.8913727705592769,
+      "grad_norm": 0.7010074257850647,
+      "learning_rate": 8.437214622569842e-05,
+      "loss": 2.4556,
+      "step": 11045
+    },
+    {
+      "epoch": 0.8914534742958599,
+      "grad_norm": 0.6763667464256287,
+      "learning_rate": 8.435655349597689e-05,
+      "loss": 2.5402,
+      "step": 11046
+    },
+    {
+      "epoch": 0.891534178032443,
+      "grad_norm": 0.6870944499969482,
+      "learning_rate": 8.4340961156131e-05,
+      "loss": 2.5307,
+      "step": 11047
+    },
+    {
+      "epoch": 0.8916148817690259,
+      "grad_norm": 0.7835623025894165,
+      "learning_rate": 8.432536920654923e-05,
+      "loss": 2.4974,
+      "step": 11048
+    },
+    {
+      "epoch": 0.8916955855056089,
+      "grad_norm": 0.7551318407058716,
+      "learning_rate": 8.430977764762024e-05,
+      "loss": 2.5206,
+      "step": 11049
+    },
+    {
+      "epoch": 0.8917762892421919,
+      "grad_norm": 0.6486842632293701,
+      "learning_rate": 8.429418647973265e-05,
+      "loss": 2.4909,
+      "step": 11050
+    },
+    {
+      "epoch": 0.891856992978775,
+      "grad_norm": 0.6894064545631409,
+      "learning_rate": 8.427859570327494e-05,
+      "loss": 2.5846,
+      "step": 11051
+    },
+    {
+      "epoch": 0.8919376967153579,
+      "grad_norm": 0.7597395181655884,
+      "learning_rate": 8.426300531863571e-05,
+      "loss": 2.5259,
+      "step": 11052
+    },
+    {
+      "epoch": 0.8920184004519409,
+      "grad_norm": 0.6784652471542358,
+      "learning_rate": 8.42474153262036e-05,
+      "loss": 2.5048,
+      "step": 11053
+    },
+    {
+      "epoch": 0.8920991041885239,
+      "grad_norm": 0.7703847885131836,
+      "learning_rate": 8.4231825726367e-05,
+      "loss": 2.4962,
+      "step": 11054
+    },
+    {
+      "epoch": 0.892179807925107,
+      "grad_norm": 0.6646561026573181,
+      "learning_rate": 8.421623651951454e-05,
+      "loss": 2.491,
+      "step": 11055
+    },
+    {
+      "epoch": 0.89226051166169,
+      "grad_norm": 0.6901054978370667,
+      "learning_rate": 8.420064770603475e-05,
+      "loss": 2.515,
+      "step": 11056
+    },
+    {
+      "epoch": 0.8923412153982729,
+      "grad_norm": 0.6789328455924988,
+      "learning_rate": 8.41850592863161e-05,
+      "loss": 2.5481,
+      "step": 11057
+    },
+    {
+      "epoch": 0.8924219191348559,
+      "grad_norm": 0.6211017370223999,
+      "learning_rate": 8.41694712607471e-05,
+      "loss": 2.51,
+      "step": 11058
+    },
+    {
+      "epoch": 0.892502622871439,
+      "grad_norm": 0.6482260823249817,
+      "learning_rate": 8.415388362971626e-05,
+      "loss": 2.5418,
+      "step": 11059
+    },
+    {
+      "epoch": 0.892583326608022,
+      "grad_norm": 0.7627651691436768,
+      "learning_rate": 8.413829639361209e-05,
+      "loss": 2.5033,
+      "step": 11060
+    },
+    {
+      "epoch": 0.892664030344605,
+      "grad_norm": 0.6560852527618408,
+      "learning_rate": 8.412270955282302e-05,
+      "loss": 2.5442,
+      "step": 11061
+    },
+    {
+      "epoch": 0.8927447340811879,
+      "grad_norm": 0.7479087114334106,
+      "learning_rate": 8.410712310773752e-05,
+      "loss": 2.5189,
+      "step": 11062
+    },
+    {
+      "epoch": 0.892825437817771,
+      "grad_norm": 0.6970879435539246,
+      "learning_rate": 8.409153705874411e-05,
+      "loss": 2.5418,
+      "step": 11063
+    },
+    {
+      "epoch": 0.892906141554354,
+      "grad_norm": 0.6514548659324646,
+      "learning_rate": 8.407595140623113e-05,
+      "loss": 2.5277,
+      "step": 11064
+    },
+    {
+      "epoch": 0.892986845290937,
+      "grad_norm": 0.6745554804801941,
+      "learning_rate": 8.406036615058707e-05,
+      "loss": 2.5085,
+      "step": 11065
+    },
+    {
+      "epoch": 0.89306754902752,
+      "grad_norm": 0.7510363459587097,
+      "learning_rate": 8.404478129220037e-05,
+      "loss": 2.4941,
+      "step": 11066
+    },
+    {
+      "epoch": 0.8931482527641029,
+      "grad_norm": 0.6531470417976379,
+      "learning_rate": 8.402919683145941e-05,
+      "loss": 2.5363,
+      "step": 11067
+    },
+    {
+      "epoch": 0.893228956500686,
+      "grad_norm": 0.6861493587493896,
+      "learning_rate": 8.401361276875262e-05,
+      "loss": 2.6369,
+      "step": 11068
+    },
+    {
+      "epoch": 0.893309660237269,
+      "grad_norm": 0.6029497981071472,
+      "learning_rate": 8.39980291044684e-05,
+      "loss": 2.4953,
+      "step": 11069
+    },
+    {
+      "epoch": 0.893390363973852,
+      "grad_norm": 0.6831715106964111,
+      "learning_rate": 8.39824458389951e-05,
+      "loss": 2.5074,
+      "step": 11070
+    },
+    {
+      "epoch": 0.8934710677104349,
+      "grad_norm": 0.7076299786567688,
+      "learning_rate": 8.396686297272112e-05,
+      "loss": 2.5934,
+      "step": 11071
+    },
+    {
+      "epoch": 0.893551771447018,
+      "grad_norm": 0.6941438913345337,
+      "learning_rate": 8.395128050603487e-05,
+      "loss": 2.5338,
+      "step": 11072
+    },
+    {
+      "epoch": 0.893632475183601,
+      "grad_norm": 0.6867249011993408,
+      "learning_rate": 8.393569843932463e-05,
+      "loss": 2.5311,
+      "step": 11073
+    },
+    {
+      "epoch": 0.893713178920184,
+      "grad_norm": 0.623991847038269,
+      "learning_rate": 8.392011677297877e-05,
+      "loss": 2.5133,
+      "step": 11074
+    },
+    {
+      "epoch": 0.893793882656767,
+      "grad_norm": 0.6808422803878784,
+      "learning_rate": 8.390453550738564e-05,
+      "loss": 2.5398,
+      "step": 11075
+    },
+    {
+      "epoch": 0.89387458639335,
+      "grad_norm": 0.7136701345443726,
+      "learning_rate": 8.388895464293357e-05,
+      "loss": 2.5415,
+      "step": 11076
+    },
+    {
+      "epoch": 0.893955290129933,
+      "grad_norm": 0.6814287304878235,
+      "learning_rate": 8.387337418001084e-05,
+      "loss": 2.4782,
+      "step": 11077
+    },
+    {
+      "epoch": 0.894035993866516,
+      "grad_norm": 0.8101940155029297,
+      "learning_rate": 8.385779411900579e-05,
+      "loss": 2.5292,
+      "step": 11078
+    },
+    {
+      "epoch": 0.894116697603099,
+      "grad_norm": 0.7106796503067017,
+      "learning_rate": 8.384221446030676e-05,
+      "loss": 2.5819,
+      "step": 11079
+    },
+    {
+      "epoch": 0.8941974013396821,
+      "grad_norm": 0.7840015292167664,
+      "learning_rate": 8.382663520430191e-05,
+      "loss": 2.5243,
+      "step": 11080
+    },
+    {
+      "epoch": 0.894278105076265,
+      "grad_norm": 0.7037288546562195,
+      "learning_rate": 8.381105635137959e-05,
+      "loss": 2.5606,
+      "step": 11081
+    },
+    {
+      "epoch": 0.894358808812848,
+      "grad_norm": 0.671558678150177,
+      "learning_rate": 8.379547790192812e-05,
+      "loss": 2.4923,
+      "step": 11082
+    },
+    {
+      "epoch": 0.894439512549431,
+      "grad_norm": 0.6789675951004028,
+      "learning_rate": 8.377989985633567e-05,
+      "loss": 2.5281,
+      "step": 11083
+    },
+    {
+      "epoch": 0.8945202162860141,
+      "grad_norm": 0.6777840852737427,
+      "learning_rate": 8.37643222149905e-05,
+      "loss": 2.5159,
+      "step": 11084
+    },
+    {
+      "epoch": 0.8946009200225971,
+      "grad_norm": 0.6920693516731262,
+      "learning_rate": 8.374874497828089e-05,
+      "loss": 2.4952,
+      "step": 11085
+    },
+    {
+      "epoch": 0.89468162375918,
+      "grad_norm": 0.7394022941589355,
+      "learning_rate": 8.373316814659502e-05,
+      "loss": 2.5035,
+      "step": 11086
+    },
+    {
+      "epoch": 0.894762327495763,
+      "grad_norm": 0.625960648059845,
+      "learning_rate": 8.37175917203211e-05,
+      "loss": 2.5324,
+      "step": 11087
+    },
+    {
+      "epoch": 0.8948430312323461,
+      "grad_norm": 0.6848758459091187,
+      "learning_rate": 8.370201569984742e-05,
+      "loss": 2.5312,
+      "step": 11088
+    },
+    {
+      "epoch": 0.8949237349689291,
+      "grad_norm": 0.7207037210464478,
+      "learning_rate": 8.368644008556205e-05,
+      "loss": 2.5807,
+      "step": 11089
+    },
+    {
+      "epoch": 0.895004438705512,
+      "grad_norm": 0.7582261562347412,
+      "learning_rate": 8.367086487785326e-05,
+      "loss": 2.532,
+      "step": 11090
+    },
+    {
+      "epoch": 0.895085142442095,
+      "grad_norm": 0.6916806101799011,
+      "learning_rate": 8.36552900771092e-05,
+      "loss": 2.4772,
+      "step": 11091
+    },
+    {
+      "epoch": 0.8951658461786781,
+      "grad_norm": 0.6457386016845703,
+      "learning_rate": 8.363971568371805e-05,
+      "loss": 2.4952,
+      "step": 11092
+    },
+    {
+      "epoch": 0.8952465499152611,
+      "grad_norm": 0.7006754279136658,
+      "learning_rate": 8.362414169806792e-05,
+      "loss": 2.5818,
+      "step": 11093
+    },
+    {
+      "epoch": 0.8953272536518441,
+      "grad_norm": 0.6939932703971863,
+      "learning_rate": 8.3608568120547e-05,
+      "loss": 2.5411,
+      "step": 11094
+    },
+    {
+      "epoch": 0.895407957388427,
+      "grad_norm": 0.6314546465873718,
+      "learning_rate": 8.359299495154343e-05,
+      "loss": 2.5408,
+      "step": 11095
+    },
+    {
+      "epoch": 0.8954886611250101,
+      "grad_norm": 0.7202826738357544,
+      "learning_rate": 8.357742219144529e-05,
+      "loss": 2.4925,
+      "step": 11096
+    },
+    {
+      "epoch": 0.8955693648615931,
+      "grad_norm": 0.6475295424461365,
+      "learning_rate": 8.356184984064071e-05,
+      "loss": 2.5023,
+      "step": 11097
+    },
+    {
+      "epoch": 0.8956500685981761,
+      "grad_norm": 0.6161238551139832,
+      "learning_rate": 8.354627789951785e-05,
+      "loss": 2.5053,
+      "step": 11098
+    },
+    {
+      "epoch": 0.8957307723347591,
+      "grad_norm": 0.6919825077056885,
+      "learning_rate": 8.353070636846472e-05,
+      "loss": 2.5387,
+      "step": 11099
+    },
+    {
+      "epoch": 0.8958114760713421,
+      "grad_norm": 0.6374878883361816,
+      "learning_rate": 8.351513524786944e-05,
+      "loss": 2.5526,
+      "step": 11100
+    },
+    {
+      "epoch": 0.8958921798079251,
+      "grad_norm": 0.7041093707084656,
+      "learning_rate": 8.349956453812009e-05,
+      "loss": 2.5282,
+      "step": 11101
+    },
+    {
+      "epoch": 0.8959728835445081,
+      "grad_norm": 0.7252324819564819,
+      "learning_rate": 8.348399423960471e-05,
+      "loss": 2.5723,
+      "step": 11102
+    },
+    {
+      "epoch": 0.8960535872810911,
+      "grad_norm": 0.681682825088501,
+      "learning_rate": 8.346842435271137e-05,
+      "loss": 2.5284,
+      "step": 11103
+    },
+    {
+      "epoch": 0.8961342910176742,
+      "grad_norm": 0.7293850183486938,
+      "learning_rate": 8.34528548778281e-05,
+      "loss": 2.5014,
+      "step": 11104
+    },
+    {
+      "epoch": 0.8962149947542571,
+      "grad_norm": 0.7057846188545227,
+      "learning_rate": 8.343728581534299e-05,
+      "loss": 2.5502,
+      "step": 11105
+    },
+    {
+      "epoch": 0.8962956984908401,
+      "grad_norm": 0.6740830540657043,
+      "learning_rate": 8.342171716564398e-05,
+      "loss": 2.5205,
+      "step": 11106
+    },
+    {
+      "epoch": 0.8963764022274231,
+      "grad_norm": 0.6917470097541809,
+      "learning_rate": 8.340614892911907e-05,
+      "loss": 2.5216,
+      "step": 11107
+    },
+    {
+      "epoch": 0.8964571059640062,
+      "grad_norm": 0.7495635151863098,
+      "learning_rate": 8.339058110615638e-05,
+      "loss": 2.5509,
+      "step": 11108
+    },
+    {
+      "epoch": 0.8965378097005892,
+      "grad_norm": 0.6687765717506409,
+      "learning_rate": 8.33750136971438e-05,
+      "loss": 2.5286,
+      "step": 11109
+    },
+    {
+      "epoch": 0.8966185134371721,
+      "grad_norm": 0.6901381015777588,
+      "learning_rate": 8.335944670246931e-05,
+      "loss": 2.5545,
+      "step": 11110
+    },
+    {
+      "epoch": 0.8966992171737551,
+      "grad_norm": 0.6645506024360657,
+      "learning_rate": 8.334388012252094e-05,
+      "loss": 2.4883,
+      "step": 11111
+    },
+    {
+      "epoch": 0.8967799209103382,
+      "grad_norm": 0.6427997350692749,
+      "learning_rate": 8.332831395768662e-05,
+      "loss": 2.5103,
+      "step": 11112
+    },
+    {
+      "epoch": 0.8968606246469212,
+      "grad_norm": 0.7224035263061523,
+      "learning_rate": 8.331274820835425e-05,
+      "loss": 2.5086,
+      "step": 11113
+    },
+    {
+      "epoch": 0.8969413283835042,
+      "grad_norm": 0.6918233036994934,
+      "learning_rate": 8.329718287491188e-05,
+      "loss": 2.5222,
+      "step": 11114
+    },
+    {
+      "epoch": 0.8970220321200871,
+      "grad_norm": 0.735583484172821,
+      "learning_rate": 8.328161795774734e-05,
+      "loss": 2.5277,
+      "step": 11115
+    },
+    {
+      "epoch": 0.8971027358566702,
+      "grad_norm": 0.6624864339828491,
+      "learning_rate": 8.326605345724857e-05,
+      "loss": 2.532,
+      "step": 11116
+    },
+    {
+      "epoch": 0.8971834395932532,
+      "grad_norm": 0.6227770447731018,
+      "learning_rate": 8.325048937380352e-05,
+      "loss": 2.5386,
+      "step": 11117
+    },
+    {
+      "epoch": 0.8972641433298362,
+      "grad_norm": 0.6483022570610046,
+      "learning_rate": 8.323492570780004e-05,
+      "loss": 2.4958,
+      "step": 11118
+    },
+    {
+      "epoch": 0.8973448470664191,
+      "grad_norm": 0.7072618007659912,
+      "learning_rate": 8.321936245962602e-05,
+      "loss": 2.4931,
+      "step": 11119
+    },
+    {
+      "epoch": 0.8974255508030021,
+      "grad_norm": 0.6848764419555664,
+      "learning_rate": 8.320379962966937e-05,
+      "loss": 2.4549,
+      "step": 11120
+    },
+    {
+      "epoch": 0.8975062545395852,
+      "grad_norm": 0.6819620132446289,
+      "learning_rate": 8.318823721831795e-05,
+      "loss": 2.5156,
+      "step": 11121
+    },
+    {
+      "epoch": 0.8975869582761682,
+      "grad_norm": 0.6834476590156555,
+      "learning_rate": 8.31726752259596e-05,
+      "loss": 2.507,
+      "step": 11122
+    },
+    {
+      "epoch": 0.8976676620127512,
+      "grad_norm": 0.6785772442817688,
+      "learning_rate": 8.315711365298214e-05,
+      "loss": 2.5086,
+      "step": 11123
+    },
+    {
+      "epoch": 0.8977483657493341,
+      "grad_norm": 0.6303566098213196,
+      "learning_rate": 8.314155249977351e-05,
+      "loss": 2.5087,
+      "step": 11124
+    },
+    {
+      "epoch": 0.8978290694859172,
+      "grad_norm": 0.6544361710548401,
+      "learning_rate": 8.31259917667214e-05,
+      "loss": 2.505,
+      "step": 11125
+    },
+    {
+      "epoch": 0.8979097732225002,
+      "grad_norm": 0.8135818243026733,
+      "learning_rate": 8.311043145421369e-05,
+      "loss": 2.5139,
+      "step": 11126
+    },
+    {
+      "epoch": 0.8979904769590832,
+      "grad_norm": 0.6744341254234314,
+      "learning_rate": 8.309487156263818e-05,
+      "loss": 2.4797,
+      "step": 11127
+    },
+    {
+      "epoch": 0.8980711806956662,
+      "grad_norm": 0.6138790845870972,
+      "learning_rate": 8.307931209238267e-05,
+      "loss": 2.5334,
+      "step": 11128
+    },
+    {
+      "epoch": 0.8981518844322492,
+      "grad_norm": 0.702434241771698,
+      "learning_rate": 8.306375304383492e-05,
+      "loss": 2.5343,
+      "step": 11129
+    },
+    {
+      "epoch": 0.8982325881688322,
+      "grad_norm": 0.6787155270576477,
+      "learning_rate": 8.304819441738275e-05,
+      "loss": 2.507,
+      "step": 11130
+    },
+    {
+      "epoch": 0.8983132919054152,
+      "grad_norm": 0.6963719129562378,
+      "learning_rate": 8.303263621341386e-05,
+      "loss": 2.5238,
+      "step": 11131
+    },
+    {
+      "epoch": 0.8983939956419982,
+      "grad_norm": 0.6623271107673645,
+      "learning_rate": 8.3017078432316e-05,
+      "loss": 2.5206,
+      "step": 11132
+    },
+    {
+      "epoch": 0.8984746993785813,
+      "grad_norm": 0.777222752571106,
+      "learning_rate": 8.300152107447701e-05,
+      "loss": 2.5004,
+      "step": 11133
+    },
+    {
+      "epoch": 0.8985554031151642,
+      "grad_norm": 0.6788455247879028,
+      "learning_rate": 8.29859641402845e-05,
+      "loss": 2.5735,
+      "step": 11134
+    },
+    {
+      "epoch": 0.8986361068517472,
+      "grad_norm": 0.6595063209533691,
+      "learning_rate": 8.297040763012624e-05,
+      "loss": 2.4988,
+      "step": 11135
+    },
+    {
+      "epoch": 0.8987168105883302,
+      "grad_norm": 0.7105697989463806,
+      "learning_rate": 8.295485154438994e-05,
+      "loss": 2.5531,
+      "step": 11136
+    },
+    {
+      "epoch": 0.8987975143249133,
+      "grad_norm": 0.6884949803352356,
+      "learning_rate": 8.29392958834633e-05,
+      "loss": 2.5158,
+      "step": 11137
+    },
+    {
+      "epoch": 0.8988782180614963,
+      "grad_norm": 0.7178345322608948,
+      "learning_rate": 8.2923740647734e-05,
+      "loss": 2.5836,
+      "step": 11138
+    },
+    {
+      "epoch": 0.8989589217980792,
+      "grad_norm": 0.7000541687011719,
+      "learning_rate": 8.290818583758973e-05,
+      "loss": 2.5345,
+      "step": 11139
+    },
+    {
+      "epoch": 0.8990396255346622,
+      "grad_norm": 0.6808128952980042,
+      "learning_rate": 8.289263145341816e-05,
+      "loss": 2.5227,
+      "step": 11140
+    },
+    {
+      "epoch": 0.8991203292712453,
+      "grad_norm": 0.7047473788261414,
+      "learning_rate": 8.287707749560691e-05,
+      "loss": 2.477,
+      "step": 11141
+    },
+    {
+      "epoch": 0.8992010330078283,
+      "grad_norm": 0.6654812693595886,
+      "learning_rate": 8.286152396454365e-05,
+      "loss": 2.4575,
+      "step": 11142
+    },
+    {
+      "epoch": 0.8992817367444113,
+      "grad_norm": 0.6690360307693481,
+      "learning_rate": 8.284597086061603e-05,
+      "loss": 2.4755,
+      "step": 11143
+    },
+    {
+      "epoch": 0.8993624404809942,
+      "grad_norm": 0.7270147204399109,
+      "learning_rate": 8.283041818421164e-05,
+      "loss": 2.5893,
+      "step": 11144
+    },
+    {
+      "epoch": 0.8994431442175773,
+      "grad_norm": 0.5977498888969421,
+      "learning_rate": 8.28148659357181e-05,
+      "loss": 2.5108,
+      "step": 11145
+    },
+    {
+      "epoch": 0.8995238479541603,
+      "grad_norm": 0.694593071937561,
+      "learning_rate": 8.279931411552307e-05,
+      "loss": 2.5036,
+      "step": 11146
+    },
+    {
+      "epoch": 0.8996045516907433,
+      "grad_norm": 0.7395440936088562,
+      "learning_rate": 8.278376272401404e-05,
+      "loss": 2.5244,
+      "step": 11147
+    },
+    {
+      "epoch": 0.8996852554273262,
+      "grad_norm": 0.6483517289161682,
+      "learning_rate": 8.276821176157867e-05,
+      "loss": 2.5619,
+      "step": 11148
+    },
+    {
+      "epoch": 0.8997659591639093,
+      "grad_norm": 0.6996768116950989,
+      "learning_rate": 8.275266122860454e-05,
+      "loss": 2.5275,
+      "step": 11149
+    },
+    {
+      "epoch": 0.8998466629004923,
+      "grad_norm": 0.661122739315033,
+      "learning_rate": 8.273711112547914e-05,
+      "loss": 2.5053,
+      "step": 11150
+    },
+    {
+      "epoch": 0.8999273666370753,
+      "grad_norm": 0.6919111609458923,
+      "learning_rate": 8.272156145259006e-05,
+      "loss": 2.578,
+      "step": 11151
+    },
+    {
+      "epoch": 0.9000080703736583,
+      "grad_norm": 0.6680958867073059,
+      "learning_rate": 8.270601221032482e-05,
+      "loss": 2.4942,
+      "step": 11152
+    },
+    {
+      "epoch": 0.9000887741102414,
+      "grad_norm": 0.6782989501953125,
+      "learning_rate": 8.269046339907101e-05,
+      "loss": 2.5461,
+      "step": 11153
+    },
+    {
+      "epoch": 0.9001694778468243,
+      "grad_norm": 0.743468165397644,
+      "learning_rate": 8.267491501921605e-05,
+      "loss": 2.629,
+      "step": 11154
+    },
+    {
+      "epoch": 0.9002501815834073,
+      "grad_norm": 0.709562361240387,
+      "learning_rate": 8.265936707114751e-05,
+      "loss": 2.566,
+      "step": 11155
+    },
+    {
+      "epoch": 0.9003308853199903,
+      "grad_norm": 0.7075676918029785,
+      "learning_rate": 8.264381955525291e-05,
+      "loss": 2.5409,
+      "step": 11156
+    },
+    {
+      "epoch": 0.9004115890565734,
+      "grad_norm": 0.7021335959434509,
+      "learning_rate": 8.262827247191963e-05,
+      "loss": 2.5606,
+      "step": 11157
+    },
+    {
+      "epoch": 0.9004922927931563,
+      "grad_norm": 0.6507331132888794,
+      "learning_rate": 8.261272582153524e-05,
+      "loss": 2.5557,
+      "step": 11158
+    },
+    {
+      "epoch": 0.9005729965297393,
+      "grad_norm": 0.7182760238647461,
+      "learning_rate": 8.25971796044872e-05,
+      "loss": 2.5567,
+      "step": 11159
+    },
+    {
+      "epoch": 0.9006537002663223,
+      "grad_norm": 0.6632338762283325,
+      "learning_rate": 8.258163382116291e-05,
+      "loss": 2.5081,
+      "step": 11160
+    },
+    {
+      "epoch": 0.9007344040029054,
+      "grad_norm": 0.6889928579330444,
+      "learning_rate": 8.256608847194983e-05,
+      "loss": 2.5034,
+      "step": 11161
+    },
+    {
+      "epoch": 0.9008151077394884,
+      "grad_norm": 0.6374824047088623,
+      "learning_rate": 8.255054355723542e-05,
+      "loss": 2.4826,
+      "step": 11162
+    },
+    {
+      "epoch": 0.9008958114760713,
+      "grad_norm": 0.7100771069526672,
+      "learning_rate": 8.253499907740706e-05,
+      "loss": 2.4666,
+      "step": 11163
+    },
+    {
+      "epoch": 0.9009765152126543,
+      "grad_norm": 0.8141123652458191,
+      "learning_rate": 8.251945503285218e-05,
+      "loss": 2.5339,
+      "step": 11164
+    },
+    {
+      "epoch": 0.9010572189492374,
+      "grad_norm": 0.6621670722961426,
+      "learning_rate": 8.250391142395822e-05,
+      "loss": 2.4805,
+      "step": 11165
+    },
+    {
+      "epoch": 0.9011379226858204,
+      "grad_norm": 0.6624772548675537,
+      "learning_rate": 8.248836825111245e-05,
+      "loss": 2.5148,
+      "step": 11166
+    },
+    {
+      "epoch": 0.9012186264224034,
+      "grad_norm": 0.6783565282821655,
+      "learning_rate": 8.247282551470235e-05,
+      "loss": 2.4481,
+      "step": 11167
+    },
+    {
+      "epoch": 0.9012993301589863,
+      "grad_norm": 0.700089156627655,
+      "learning_rate": 8.245728321511525e-05,
+      "loss": 2.5649,
+      "step": 11168
+    },
+    {
+      "epoch": 0.9013800338955693,
+      "grad_norm": 0.6765339970588684,
+      "learning_rate": 8.244174135273852e-05,
+      "loss": 2.5221,
+      "step": 11169
+    },
+    {
+      "epoch": 0.9014607376321524,
+      "grad_norm": 0.6896056532859802,
+      "learning_rate": 8.242619992795948e-05,
+      "loss": 2.4742,
+      "step": 11170
+    },
+    {
+      "epoch": 0.9015414413687354,
+      "grad_norm": 0.7134374976158142,
+      "learning_rate": 8.241065894116547e-05,
+      "loss": 2.5231,
+      "step": 11171
+    },
+    {
+      "epoch": 0.9016221451053184,
+      "grad_norm": 0.6939442753791809,
+      "learning_rate": 8.239511839274385e-05,
+      "loss": 2.5159,
+      "step": 11172
+    },
+    {
+      "epoch": 0.9017028488419013,
+      "grad_norm": 0.6780345439910889,
+      "learning_rate": 8.237957828308187e-05,
+      "loss": 2.5474,
+      "step": 11173
+    },
+    {
+      "epoch": 0.9017835525784844,
+      "grad_norm": 0.6532382965087891,
+      "learning_rate": 8.236403861256687e-05,
+      "loss": 2.4982,
+      "step": 11174
+    },
+    {
+      "epoch": 0.9018642563150674,
+      "grad_norm": 0.6918137073516846,
+      "learning_rate": 8.234849938158615e-05,
+      "loss": 2.4657,
+      "step": 11175
+    },
+    {
+      "epoch": 0.9019449600516504,
+      "grad_norm": 0.6838762164115906,
+      "learning_rate": 8.233296059052695e-05,
+      "loss": 2.5405,
+      "step": 11176
+    },
+    {
+      "epoch": 0.9020256637882333,
+      "grad_norm": 0.7560290098190308,
+      "learning_rate": 8.231742223977653e-05,
+      "loss": 2.5379,
+      "step": 11177
+    },
+    {
+      "epoch": 0.9021063675248164,
+      "grad_norm": 0.6673319339752197,
+      "learning_rate": 8.230188432972221e-05,
+      "loss": 2.4669,
+      "step": 11178
+    },
+    {
+      "epoch": 0.9021870712613994,
+      "grad_norm": 0.7486294507980347,
+      "learning_rate": 8.228634686075116e-05,
+      "loss": 2.526,
+      "step": 11179
+    },
+    {
+      "epoch": 0.9022677749979824,
+      "grad_norm": 0.7012811303138733,
+      "learning_rate": 8.227080983325067e-05,
+      "loss": 2.5544,
+      "step": 11180
+    },
+    {
+      "epoch": 0.9023484787345654,
+      "grad_norm": 0.6807447075843811,
+      "learning_rate": 8.225527324760796e-05,
+      "loss": 2.5139,
+      "step": 11181
+    },
+    {
+      "epoch": 0.9024291824711484,
+      "grad_norm": 0.7594932317733765,
+      "learning_rate": 8.223973710421018e-05,
+      "loss": 2.539,
+      "step": 11182
+    },
+    {
+      "epoch": 0.9025098862077314,
+      "grad_norm": 0.6764204502105713,
+      "learning_rate": 8.22242014034446e-05,
+      "loss": 2.6128,
+      "step": 11183
+    },
+    {
+      "epoch": 0.9025905899443144,
+      "grad_norm": 0.6499967575073242,
+      "learning_rate": 8.220866614569837e-05,
+      "loss": 2.5459,
+      "step": 11184
+    },
+    {
+      "epoch": 0.9026712936808974,
+      "grad_norm": 0.673076331615448,
+      "learning_rate": 8.219313133135876e-05,
+      "loss": 2.5852,
+      "step": 11185
+    },
+    {
+      "epoch": 0.9027519974174805,
+      "grad_norm": 0.784854531288147,
+      "learning_rate": 8.21775969608128e-05,
+      "loss": 2.5586,
+      "step": 11186
+    },
+    {
+      "epoch": 0.9028327011540634,
+      "grad_norm": 0.658963680267334,
+      "learning_rate": 8.216206303444771e-05,
+      "loss": 2.4376,
+      "step": 11187
+    },
+    {
+      "epoch": 0.9029134048906464,
+      "grad_norm": 0.6456249356269836,
+      "learning_rate": 8.214652955265067e-05,
+      "loss": 2.5166,
+      "step": 11188
+    },
+    {
+      "epoch": 0.9029941086272294,
+      "grad_norm": 0.6940007209777832,
+      "learning_rate": 8.213099651580874e-05,
+      "loss": 2.4992,
+      "step": 11189
+    },
+    {
+      "epoch": 0.9030748123638125,
+      "grad_norm": 0.6661425828933716,
+      "learning_rate": 8.211546392430911e-05,
+      "loss": 2.5177,
+      "step": 11190
+    },
+    {
+      "epoch": 0.9031555161003955,
+      "grad_norm": 0.647834300994873,
+      "learning_rate": 8.20999317785389e-05,
+      "loss": 2.4666,
+      "step": 11191
+    },
+    {
+      "epoch": 0.9032362198369784,
+      "grad_norm": 0.7673383355140686,
+      "learning_rate": 8.208440007888515e-05,
+      "loss": 2.4852,
+      "step": 11192
+    },
+    {
+      "epoch": 0.9033169235735614,
+      "grad_norm": 0.7033390998840332,
+      "learning_rate": 8.206886882573498e-05,
+      "loss": 2.5549,
+      "step": 11193
+    },
+    {
+      "epoch": 0.9033976273101445,
+      "grad_norm": 0.6871141195297241,
+      "learning_rate": 8.205333801947548e-05,
+      "loss": 2.4585,
+      "step": 11194
+    },
+    {
+      "epoch": 0.9034783310467275,
+      "grad_norm": 0.7201984524726868,
+      "learning_rate": 8.20378076604937e-05,
+      "loss": 2.5271,
+      "step": 11195
+    },
+    {
+      "epoch": 0.9035590347833105,
+      "grad_norm": 0.704060971736908,
+      "learning_rate": 8.202227774917671e-05,
+      "loss": 2.4915,
+      "step": 11196
+    },
+    {
+      "epoch": 0.9036397385198934,
+      "grad_norm": 0.6833879947662354,
+      "learning_rate": 8.200674828591156e-05,
+      "loss": 2.4496,
+      "step": 11197
+    },
+    {
+      "epoch": 0.9037204422564765,
+      "grad_norm": 0.6564866304397583,
+      "learning_rate": 8.199121927108527e-05,
+      "loss": 2.4818,
+      "step": 11198
+    },
+    {
+      "epoch": 0.9038011459930595,
+      "grad_norm": 0.6970151662826538,
+      "learning_rate": 8.197569070508486e-05,
+      "loss": 2.5812,
+      "step": 11199
+    },
+    {
+      "epoch": 0.9038818497296425,
+      "grad_norm": 0.7147194743156433,
+      "learning_rate": 8.196016258829737e-05,
+      "loss": 2.5543,
+      "step": 11200
+    },
+    {
+      "epoch": 0.9039625534662254,
+      "grad_norm": 0.6357648968696594,
+      "learning_rate": 8.194463492110981e-05,
+      "loss": 2.5254,
+      "step": 11201
+    },
+    {
+      "epoch": 0.9040432572028085,
+      "grad_norm": 0.7113756537437439,
+      "learning_rate": 8.19291077039091e-05,
+      "loss": 2.5179,
+      "step": 11202
+    },
+    {
+      "epoch": 0.9041239609393915,
+      "grad_norm": 0.7252987623214722,
+      "learning_rate": 8.191358093708228e-05,
+      "loss": 2.5658,
+      "step": 11203
+    },
+    {
+      "epoch": 0.9042046646759745,
+      "grad_norm": 0.7095803618431091,
+      "learning_rate": 8.189805462101631e-05,
+      "loss": 2.583,
+      "step": 11204
+    },
+    {
+      "epoch": 0.9042853684125575,
+      "grad_norm": 0.7447760105133057,
+      "learning_rate": 8.188252875609812e-05,
+      "loss": 2.5608,
+      "step": 11205
+    },
+    {
+      "epoch": 0.9043660721491406,
+      "grad_norm": 0.6578439474105835,
+      "learning_rate": 8.186700334271468e-05,
+      "loss": 2.508,
+      "step": 11206
+    },
+    {
+      "epoch": 0.9044467758857235,
+      "grad_norm": 0.6776832938194275,
+      "learning_rate": 8.185147838125296e-05,
+      "loss": 2.6188,
+      "step": 11207
+    },
+    {
+      "epoch": 0.9045274796223065,
+      "grad_norm": 0.6559253931045532,
+      "learning_rate": 8.183595387209976e-05,
+      "loss": 2.5307,
+      "step": 11208
+    },
+    {
+      "epoch": 0.9046081833588895,
+      "grad_norm": 0.7078405022621155,
+      "learning_rate": 8.18204298156421e-05,
+      "loss": 2.5545,
+      "step": 11209
+    },
+    {
+      "epoch": 0.9046888870954726,
+      "grad_norm": 0.6790273189544678,
+      "learning_rate": 8.18049062122669e-05,
+      "loss": 2.4963,
+      "step": 11210
+    },
+    {
+      "epoch": 0.9047695908320555,
+      "grad_norm": 0.6888250708580017,
+      "learning_rate": 8.178938306236095e-05,
+      "loss": 2.5108,
+      "step": 11211
+    },
+    {
+      "epoch": 0.9048502945686385,
+      "grad_norm": 0.6438474059104919,
+      "learning_rate": 8.177386036631119e-05,
+      "loss": 2.4976,
+      "step": 11212
+    },
+    {
+      "epoch": 0.9049309983052215,
+      "grad_norm": 0.6786646842956543,
+      "learning_rate": 8.175833812450445e-05,
+      "loss": 2.4584,
+      "step": 11213
+    },
+    {
+      "epoch": 0.9050117020418046,
+      "grad_norm": 0.6480324268341064,
+      "learning_rate": 8.174281633732764e-05,
+      "loss": 2.5021,
+      "step": 11214
+    },
+    {
+      "epoch": 0.9050924057783876,
+      "grad_norm": 0.7232171893119812,
+      "learning_rate": 8.172729500516756e-05,
+      "loss": 2.4742,
+      "step": 11215
+    },
+    {
+      "epoch": 0.9051731095149705,
+      "grad_norm": 0.7048845291137695,
+      "learning_rate": 8.171177412841105e-05,
+      "loss": 2.518,
+      "step": 11216
+    },
+    {
+      "epoch": 0.9052538132515535,
+      "grad_norm": 0.6363180875778198,
+      "learning_rate": 8.169625370744496e-05,
+      "loss": 2.5154,
+      "step": 11217
+    },
+    {
+      "epoch": 0.9053345169881366,
+      "grad_norm": 0.7176045179367065,
+      "learning_rate": 8.168073374265605e-05,
+      "loss": 2.5182,
+      "step": 11218
+    },
+    {
+      "epoch": 0.9054152207247196,
+      "grad_norm": 0.7011643052101135,
+      "learning_rate": 8.166521423443112e-05,
+      "loss": 2.5615,
+      "step": 11219
+    },
+    {
+      "epoch": 0.9054959244613026,
+      "grad_norm": 0.6853327751159668,
+      "learning_rate": 8.164969518315704e-05,
+      "loss": 2.5057,
+      "step": 11220
+    },
+    {
+      "epoch": 0.9055766281978855,
+      "grad_norm": 0.6972528696060181,
+      "learning_rate": 8.163417658922049e-05,
+      "loss": 2.4949,
+      "step": 11221
+    },
+    {
+      "epoch": 0.9056573319344685,
+      "grad_norm": 0.6780978441238403,
+      "learning_rate": 8.161865845300824e-05,
+      "loss": 2.5601,
+      "step": 11222
+    },
+    {
+      "epoch": 0.9057380356710516,
+      "grad_norm": 0.6454098224639893,
+      "learning_rate": 8.160314077490711e-05,
+      "loss": 2.4203,
+      "step": 11223
+    },
+    {
+      "epoch": 0.9058187394076346,
+      "grad_norm": 0.7300907969474792,
+      "learning_rate": 8.158762355530378e-05,
+      "loss": 2.4818,
+      "step": 11224
+    },
+    {
+      "epoch": 0.9058994431442176,
+      "grad_norm": 0.682475745677948,
+      "learning_rate": 8.1572106794585e-05,
+      "loss": 2.4852,
+      "step": 11225
+    },
+    {
+      "epoch": 0.9059801468808005,
+      "grad_norm": 0.6666192412376404,
+      "learning_rate": 8.155659049313754e-05,
+      "loss": 2.5642,
+      "step": 11226
+    },
+    {
+      "epoch": 0.9060608506173836,
+      "grad_norm": 0.6873177886009216,
+      "learning_rate": 8.154107465134801e-05,
+      "loss": 2.5163,
+      "step": 11227
+    },
+    {
+      "epoch": 0.9061415543539666,
+      "grad_norm": 0.6704845428466797,
+      "learning_rate": 8.152555926960315e-05,
+      "loss": 2.5481,
+      "step": 11228
+    },
+    {
+      "epoch": 0.9062222580905496,
+      "grad_norm": 0.6340618133544922,
+      "learning_rate": 8.151004434828963e-05,
+      "loss": 2.4701,
+      "step": 11229
+    },
+    {
+      "epoch": 0.9063029618271325,
+      "grad_norm": 0.7886226177215576,
+      "learning_rate": 8.14945298877942e-05,
+      "loss": 2.5322,
+      "step": 11230
+    },
+    {
+      "epoch": 0.9063836655637156,
+      "grad_norm": 0.7086018919944763,
+      "learning_rate": 8.14790158885034e-05,
+      "loss": 2.4909,
+      "step": 11231
+    },
+    {
+      "epoch": 0.9064643693002986,
+      "grad_norm": 0.6791329979896545,
+      "learning_rate": 8.146350235080396e-05,
+      "loss": 2.4438,
+      "step": 11232
+    },
+    {
+      "epoch": 0.9065450730368816,
+      "grad_norm": 0.7070720791816711,
+      "learning_rate": 8.14479892750825e-05,
+      "loss": 2.528,
+      "step": 11233
+    },
+    {
+      "epoch": 0.9066257767734646,
+      "grad_norm": 0.6551348567008972,
+      "learning_rate": 8.143247666172564e-05,
+      "loss": 2.4747,
+      "step": 11234
+    },
+    {
+      "epoch": 0.9067064805100477,
+      "grad_norm": 0.6691645979881287,
+      "learning_rate": 8.141696451111997e-05,
+      "loss": 2.5038,
+      "step": 11235
+    },
+    {
+      "epoch": 0.9067871842466306,
+      "grad_norm": 0.6814864277839661,
+      "learning_rate": 8.14014528236522e-05,
+      "loss": 2.5737,
+      "step": 11236
+    },
+    {
+      "epoch": 0.9068678879832136,
+      "grad_norm": 0.7442377209663391,
+      "learning_rate": 8.138594159970877e-05,
+      "loss": 2.5839,
+      "step": 11237
+    },
+    {
+      "epoch": 0.9069485917197966,
+      "grad_norm": 0.6861338019371033,
+      "learning_rate": 8.137043083967634e-05,
+      "loss": 2.567,
+      "step": 11238
+    },
+    {
+      "epoch": 0.9070292954563797,
+      "grad_norm": 0.7056479454040527,
+      "learning_rate": 8.135492054394151e-05,
+      "loss": 2.5297,
+      "step": 11239
+    },
+    {
+      "epoch": 0.9071099991929626,
+      "grad_norm": 0.7166962623596191,
+      "learning_rate": 8.133941071289076e-05,
+      "loss": 2.4834,
+      "step": 11240
+    },
+    {
+      "epoch": 0.9071907029295456,
+      "grad_norm": 0.6285616159439087,
+      "learning_rate": 8.132390134691068e-05,
+      "loss": 2.5066,
+      "step": 11241
+    },
+    {
+      "epoch": 0.9072714066661286,
+      "grad_norm": 0.681915283203125,
+      "learning_rate": 8.130839244638783e-05,
+      "loss": 2.5387,
+      "step": 11242
+    },
+    {
+      "epoch": 0.9073521104027117,
+      "grad_norm": 0.6876898407936096,
+      "learning_rate": 8.129288401170866e-05,
+      "loss": 2.4465,
+      "step": 11243
+    },
+    {
+      "epoch": 0.9074328141392947,
+      "grad_norm": 0.657132625579834,
+      "learning_rate": 8.127737604325975e-05,
+      "loss": 2.499,
+      "step": 11244
+    },
+    {
+      "epoch": 0.9075135178758776,
+      "grad_norm": 0.6678825616836548,
+      "learning_rate": 8.126186854142752e-05,
+      "loss": 2.4872,
+      "step": 11245
+    },
+    {
+      "epoch": 0.9075942216124606,
+      "grad_norm": 0.7296879291534424,
+      "learning_rate": 8.124636150659858e-05,
+      "loss": 2.4783,
+      "step": 11246
+    },
+    {
+      "epoch": 0.9076749253490437,
+      "grad_norm": 0.7087056040763855,
+      "learning_rate": 8.12308549391593e-05,
+      "loss": 2.507,
+      "step": 11247
+    },
+    {
+      "epoch": 0.9077556290856267,
+      "grad_norm": 0.7099738121032715,
+      "learning_rate": 8.121534883949616e-05,
+      "loss": 2.5317,
+      "step": 11248
+    },
+    {
+      "epoch": 0.9078363328222097,
+      "grad_norm": 0.6421170830726624,
+      "learning_rate": 8.119984320799566e-05,
+      "loss": 2.5291,
+      "step": 11249
+    },
+    {
+      "epoch": 0.9079170365587926,
+      "grad_norm": 0.6835018396377563,
+      "learning_rate": 8.11843380450442e-05,
+      "loss": 2.5523,
+      "step": 11250
+    },
+    {
+      "epoch": 0.9079977402953757,
+      "grad_norm": 0.6638229489326477,
+      "learning_rate": 8.11688333510282e-05,
+      "loss": 2.5128,
+      "step": 11251
+    },
+    {
+      "epoch": 0.9080784440319587,
+      "grad_norm": 0.6783459186553955,
+      "learning_rate": 8.115332912633415e-05,
+      "loss": 2.5485,
+      "step": 11252
+    },
+    {
+      "epoch": 0.9081591477685417,
+      "grad_norm": 0.65911865234375,
+      "learning_rate": 8.113782537134838e-05,
+      "loss": 2.5408,
+      "step": 11253
+    },
+    {
+      "epoch": 0.9082398515051247,
+      "grad_norm": 0.6844244003295898,
+      "learning_rate": 8.112232208645729e-05,
+      "loss": 2.6067,
+      "step": 11254
+    },
+    {
+      "epoch": 0.9083205552417077,
+      "grad_norm": 0.6896870136260986,
+      "learning_rate": 8.110681927204729e-05,
+      "loss": 2.5444,
+      "step": 11255
+    },
+    {
+      "epoch": 0.9084012589782907,
+      "grad_norm": 0.6693820953369141,
+      "learning_rate": 8.109131692850473e-05,
+      "loss": 2.5118,
+      "step": 11256
+    },
+    {
+      "epoch": 0.9084819627148737,
+      "grad_norm": 0.6401854753494263,
+      "learning_rate": 8.107581505621599e-05,
+      "loss": 2.4811,
+      "step": 11257
+    },
+    {
+      "epoch": 0.9085626664514567,
+      "grad_norm": 0.6861663460731506,
+      "learning_rate": 8.106031365556743e-05,
+      "loss": 2.4633,
+      "step": 11258
+    },
+    {
+      "epoch": 0.9086433701880398,
+      "grad_norm": 0.6631655097007751,
+      "learning_rate": 8.104481272694533e-05,
+      "loss": 2.5748,
+      "step": 11259
+    },
+    {
+      "epoch": 0.9087240739246227,
+      "grad_norm": 0.6499454975128174,
+      "learning_rate": 8.102931227073604e-05,
+      "loss": 2.5573,
+      "step": 11260
+    },
+    {
+      "epoch": 0.9088047776612057,
+      "grad_norm": 0.7214524149894714,
+      "learning_rate": 8.10138122873259e-05,
+      "loss": 2.4905,
+      "step": 11261
+    },
+    {
+      "epoch": 0.9088854813977887,
+      "grad_norm": 0.6481152176856995,
+      "learning_rate": 8.099831277710122e-05,
+      "loss": 2.5073,
+      "step": 11262
+    },
+    {
+      "epoch": 0.9089661851343718,
+      "grad_norm": 0.6666486859321594,
+      "learning_rate": 8.09828137404482e-05,
+      "loss": 2.5379,
+      "step": 11263
+    },
+    {
+      "epoch": 0.9090468888709548,
+      "grad_norm": 0.7186474800109863,
+      "learning_rate": 8.096731517775319e-05,
+      "loss": 2.5164,
+      "step": 11264
+    },
+    {
+      "epoch": 0.9091275926075377,
+      "grad_norm": 0.6838653087615967,
+      "learning_rate": 8.095181708940245e-05,
+      "loss": 2.49,
+      "step": 11265
+    },
+    {
+      "epoch": 0.9092082963441207,
+      "grad_norm": 0.7740866541862488,
+      "learning_rate": 8.093631947578221e-05,
+      "loss": 2.5487,
+      "step": 11266
+    },
+    {
+      "epoch": 0.9092890000807038,
+      "grad_norm": 0.7198607325553894,
+      "learning_rate": 8.092082233727871e-05,
+      "loss": 2.4477,
+      "step": 11267
+    },
+    {
+      "epoch": 0.9093697038172868,
+      "grad_norm": 0.6454673409461975,
+      "learning_rate": 8.090532567427825e-05,
+      "loss": 2.523,
+      "step": 11268
+    },
+    {
+      "epoch": 0.9094504075538697,
+      "grad_norm": 0.6169581413269043,
+      "learning_rate": 8.088982948716692e-05,
+      "loss": 2.4924,
+      "step": 11269
+    },
+    {
+      "epoch": 0.9095311112904527,
+      "grad_norm": 0.7034861445426941,
+      "learning_rate": 8.0874333776331e-05,
+      "loss": 2.4756,
+      "step": 11270
+    },
+    {
+      "epoch": 0.9096118150270357,
+      "grad_norm": 0.7231355309486389,
+      "learning_rate": 8.085883854215671e-05,
+      "loss": 2.4963,
+      "step": 11271
+    },
+    {
+      "epoch": 0.9096925187636188,
+      "grad_norm": 0.6597892045974731,
+      "learning_rate": 8.084334378503017e-05,
+      "loss": 2.5617,
+      "step": 11272
+    },
+    {
+      "epoch": 0.9097732225002018,
+      "grad_norm": 0.7257365584373474,
+      "learning_rate": 8.082784950533759e-05,
+      "loss": 2.5293,
+      "step": 11273
+    },
+    {
+      "epoch": 0.9098539262367847,
+      "grad_norm": 0.7305313944816589,
+      "learning_rate": 8.081235570346512e-05,
+      "loss": 2.5355,
+      "step": 11274
+    },
+    {
+      "epoch": 0.9099346299733677,
+      "grad_norm": 0.6814435720443726,
+      "learning_rate": 8.07968623797989e-05,
+      "loss": 2.4842,
+      "step": 11275
+    },
+    {
+      "epoch": 0.9100153337099508,
+      "grad_norm": 0.7342902421951294,
+      "learning_rate": 8.078136953472506e-05,
+      "loss": 2.4817,
+      "step": 11276
+    },
+    {
+      "epoch": 0.9100960374465338,
+      "grad_norm": 0.6456516981124878,
+      "learning_rate": 8.076587716862973e-05,
+      "loss": 2.5119,
+      "step": 11277
+    },
+    {
+      "epoch": 0.9101767411831168,
+      "grad_norm": 0.7268881797790527,
+      "learning_rate": 8.075038528189906e-05,
+      "loss": 2.4614,
+      "step": 11278
+    },
+    {
+      "epoch": 0.9102574449196997,
+      "grad_norm": 0.6901549696922302,
+      "learning_rate": 8.073489387491906e-05,
+      "loss": 2.5411,
+      "step": 11279
+    },
+    {
+      "epoch": 0.9103381486562828,
+      "grad_norm": 0.6850160956382751,
+      "learning_rate": 8.071940294807588e-05,
+      "loss": 2.5078,
+      "step": 11280
+    },
+    {
+      "epoch": 0.9104188523928658,
+      "grad_norm": 0.6550731658935547,
+      "learning_rate": 8.070391250175558e-05,
+      "loss": 2.5502,
+      "step": 11281
+    },
+    {
+      "epoch": 0.9104995561294488,
+      "grad_norm": 0.7524412274360657,
+      "learning_rate": 8.068842253634421e-05,
+      "loss": 2.4699,
+      "step": 11282
+    },
+    {
+      "epoch": 0.9105802598660317,
+      "grad_norm": 0.6659243702888489,
+      "learning_rate": 8.067293305222784e-05,
+      "loss": 2.557,
+      "step": 11283
+    },
+    {
+      "epoch": 0.9106609636026148,
+      "grad_norm": 0.67015540599823,
+      "learning_rate": 8.065744404979251e-05,
+      "loss": 2.5929,
+      "step": 11284
+    },
+    {
+      "epoch": 0.9107416673391978,
+      "grad_norm": 0.7139000296592712,
+      "learning_rate": 8.064195552942422e-05,
+      "loss": 2.5262,
+      "step": 11285
+    },
+    {
+      "epoch": 0.9108223710757808,
+      "grad_norm": 0.6918016672134399,
+      "learning_rate": 8.062646749150899e-05,
+      "loss": 2.5161,
+      "step": 11286
+    },
+    {
+      "epoch": 0.9109030748123638,
+      "grad_norm": 0.7395541667938232,
+      "learning_rate": 8.061097993643289e-05,
+      "loss": 2.5351,
+      "step": 11287
+    },
+    {
+      "epoch": 0.9109837785489469,
+      "grad_norm": 0.6794499158859253,
+      "learning_rate": 8.05954928645818e-05,
+      "loss": 2.4617,
+      "step": 11288
+    },
+    {
+      "epoch": 0.9110644822855298,
+      "grad_norm": 0.6906577348709106,
+      "learning_rate": 8.058000627634176e-05,
+      "loss": 2.5701,
+      "step": 11289
+    },
+    {
+      "epoch": 0.9111451860221128,
+      "grad_norm": 0.6954079866409302,
+      "learning_rate": 8.056452017209874e-05,
+      "loss": 2.5137,
+      "step": 11290
+    },
+    {
+      "epoch": 0.9112258897586958,
+      "grad_norm": 0.7381381988525391,
+      "learning_rate": 8.054903455223866e-05,
+      "loss": 2.6666,
+      "step": 11291
+    },
+    {
+      "epoch": 0.9113065934952789,
+      "grad_norm": 0.6731518507003784,
+      "learning_rate": 8.053354941714749e-05,
+      "loss": 2.5173,
+      "step": 11292
+    },
+    {
+      "epoch": 0.9113872972318618,
+      "grad_norm": 0.6976885795593262,
+      "learning_rate": 8.051806476721116e-05,
+      "loss": 2.5089,
+      "step": 11293
+    },
+    {
+      "epoch": 0.9114680009684448,
+      "grad_norm": 0.6401965618133545,
+      "learning_rate": 8.050258060281562e-05,
+      "loss": 2.5295,
+      "step": 11294
+    },
+    {
+      "epoch": 0.9115487047050278,
+      "grad_norm": 0.7409671545028687,
+      "learning_rate": 8.048709692434667e-05,
+      "loss": 2.5074,
+      "step": 11295
+    },
+    {
+      "epoch": 0.9116294084416109,
+      "grad_norm": 0.6028234958648682,
+      "learning_rate": 8.04716137321903e-05,
+      "loss": 2.5437,
+      "step": 11296
+    },
+    {
+      "epoch": 0.9117101121781939,
+      "grad_norm": 0.727643609046936,
+      "learning_rate": 8.04561310267324e-05,
+      "loss": 2.5272,
+      "step": 11297
+    },
+    {
+      "epoch": 0.9117908159147768,
+      "grad_norm": 0.6912926435470581,
+      "learning_rate": 8.044064880835876e-05,
+      "loss": 2.5166,
+      "step": 11298
+    },
+    {
+      "epoch": 0.9118715196513598,
+      "grad_norm": 0.6971367001533508,
+      "learning_rate": 8.042516707745528e-05,
+      "loss": 2.5421,
+      "step": 11299
+    },
+    {
+      "epoch": 0.9119522233879429,
+      "grad_norm": 0.6722451448440552,
+      "learning_rate": 8.040968583440783e-05,
+      "loss": 2.5088,
+      "step": 11300
+    },
+    {
+      "epoch": 0.9120329271245259,
+      "grad_norm": 0.6469144225120544,
+      "learning_rate": 8.03942050796022e-05,
+      "loss": 2.4921,
+      "step": 11301
+    },
+    {
+      "epoch": 0.9121136308611089,
+      "grad_norm": 0.6709008812904358,
+      "learning_rate": 8.037872481342423e-05,
+      "loss": 2.4553,
+      "step": 11302
+    },
+    {
+      "epoch": 0.9121943345976918,
+      "grad_norm": 0.6540920734405518,
+      "learning_rate": 8.036324503625977e-05,
+      "loss": 2.489,
+      "step": 11303
+    },
+    {
+      "epoch": 0.9122750383342749,
+      "grad_norm": 0.6589755415916443,
+      "learning_rate": 8.034776574849453e-05,
+      "loss": 2.5195,
+      "step": 11304
+    },
+    {
+      "epoch": 0.9123557420708579,
+      "grad_norm": 0.676943838596344,
+      "learning_rate": 8.033228695051434e-05,
+      "loss": 2.4877,
+      "step": 11305
+    },
+    {
+      "epoch": 0.9124364458074409,
+      "grad_norm": 0.6509177088737488,
+      "learning_rate": 8.031680864270498e-05,
+      "loss": 2.5229,
+      "step": 11306
+    },
+    {
+      "epoch": 0.9125171495440239,
+      "grad_norm": 0.7480820417404175,
+      "learning_rate": 8.030133082545219e-05,
+      "loss": 2.5016,
+      "step": 11307
+    },
+    {
+      "epoch": 0.9125978532806069,
+      "grad_norm": 0.7130550742149353,
+      "learning_rate": 8.028585349914174e-05,
+      "loss": 2.5251,
+      "step": 11308
+    },
+    {
+      "epoch": 0.9126785570171899,
+      "grad_norm": 0.6959688067436218,
+      "learning_rate": 8.027037666415934e-05,
+      "loss": 2.4776,
+      "step": 11309
+    },
+    {
+      "epoch": 0.9127592607537729,
+      "grad_norm": 0.7540854215621948,
+      "learning_rate": 8.025490032089076e-05,
+      "loss": 2.5097,
+      "step": 11310
+    },
+    {
+      "epoch": 0.9128399644903559,
+      "grad_norm": 0.6921199560165405,
+      "learning_rate": 8.023942446972165e-05,
+      "loss": 2.5354,
+      "step": 11311
+    },
+    {
+      "epoch": 0.912920668226939,
+      "grad_norm": 0.649824857711792,
+      "learning_rate": 8.022394911103774e-05,
+      "loss": 2.5398,
+      "step": 11312
+    },
+    {
+      "epoch": 0.9130013719635219,
+      "grad_norm": 0.6951068639755249,
+      "learning_rate": 8.020847424522474e-05,
+      "loss": 2.5302,
+      "step": 11313
+    },
+    {
+      "epoch": 0.9130820757001049,
+      "grad_norm": 0.6906851530075073,
+      "learning_rate": 8.019299987266827e-05,
+      "loss": 2.581,
+      "step": 11314
+    },
+    {
+      "epoch": 0.9131627794366879,
+      "grad_norm": 0.6758459210395813,
+      "learning_rate": 8.0177525993754e-05,
+      "loss": 2.5208,
+      "step": 11315
+    },
+    {
+      "epoch": 0.913243483173271,
+      "grad_norm": 0.6915175318717957,
+      "learning_rate": 8.016205260886766e-05,
+      "loss": 2.5386,
+      "step": 11316
+    },
+    {
+      "epoch": 0.913324186909854,
+      "grad_norm": 0.7083550691604614,
+      "learning_rate": 8.014657971839476e-05,
+      "loss": 2.4895,
+      "step": 11317
+    },
+    {
+      "epoch": 0.9134048906464369,
+      "grad_norm": 0.7052562832832336,
+      "learning_rate": 8.013110732272102e-05,
+      "loss": 2.4896,
+      "step": 11318
+    },
+    {
+      "epoch": 0.9134855943830199,
+      "grad_norm": 0.7811834216117859,
+      "learning_rate": 8.011563542223206e-05,
+      "loss": 2.5082,
+      "step": 11319
+    },
+    {
+      "epoch": 0.913566298119603,
+      "grad_norm": 0.6207153797149658,
+      "learning_rate": 8.01001640173134e-05,
+      "loss": 2.4967,
+      "step": 11320
+    },
+    {
+      "epoch": 0.913647001856186,
+      "grad_norm": 0.7637950778007507,
+      "learning_rate": 8.008469310835065e-05,
+      "loss": 2.4907,
+      "step": 11321
+    },
+    {
+      "epoch": 0.913727705592769,
+      "grad_norm": 0.7263950705528259,
+      "learning_rate": 8.006922269572947e-05,
+      "loss": 2.5259,
+      "step": 11322
+    },
+    {
+      "epoch": 0.9138084093293519,
+      "grad_norm": 0.6965721845626831,
+      "learning_rate": 8.005375277983531e-05,
+      "loss": 2.5648,
+      "step": 11323
+    },
+    {
+      "epoch": 0.9138891130659349,
+      "grad_norm": 0.7146127223968506,
+      "learning_rate": 8.003828336105377e-05,
+      "loss": 2.53,
+      "step": 11324
+    },
+    {
+      "epoch": 0.913969816802518,
+      "grad_norm": 0.7083697319030762,
+      "learning_rate": 8.00228144397704e-05,
+      "loss": 2.4923,
+      "step": 11325
+    },
+    {
+      "epoch": 0.914050520539101,
+      "grad_norm": 0.7259312868118286,
+      "learning_rate": 8.000734601637074e-05,
+      "loss": 2.5303,
+      "step": 11326
+    },
+    {
+      "epoch": 0.9141312242756839,
+      "grad_norm": 0.7072086930274963,
+      "learning_rate": 7.999187809124025e-05,
+      "loss": 2.4662,
+      "step": 11327
+    },
+    {
+      "epoch": 0.9142119280122669,
+      "grad_norm": 0.7216035723686218,
+      "learning_rate": 7.997641066476445e-05,
+      "loss": 2.5069,
+      "step": 11328
+    },
+    {
+      "epoch": 0.91429263174885,
+      "grad_norm": 0.6925712823867798,
+      "learning_rate": 7.99609437373289e-05,
+      "loss": 2.5107,
+      "step": 11329
+    },
+    {
+      "epoch": 0.914373335485433,
+      "grad_norm": 0.6672701835632324,
+      "learning_rate": 7.994547730931896e-05,
+      "loss": 2.5248,
+      "step": 11330
+    },
+    {
+      "epoch": 0.914454039222016,
+      "grad_norm": 0.8058515787124634,
+      "learning_rate": 7.993001138112016e-05,
+      "loss": 2.4427,
+      "step": 11331
+    },
+    {
+      "epoch": 0.9145347429585989,
+      "grad_norm": 0.6942592859268188,
+      "learning_rate": 7.991454595311795e-05,
+      "loss": 2.6163,
+      "step": 11332
+    },
+    {
+      "epoch": 0.914615446695182,
+      "grad_norm": 0.7051894068717957,
+      "learning_rate": 7.989908102569774e-05,
+      "loss": 2.5327,
+      "step": 11333
+    },
+    {
+      "epoch": 0.914696150431765,
+      "grad_norm": 0.6824771761894226,
+      "learning_rate": 7.988361659924496e-05,
+      "loss": 2.4843,
+      "step": 11334
+    },
+    {
+      "epoch": 0.914776854168348,
+      "grad_norm": 0.6756488084793091,
+      "learning_rate": 7.98681526741451e-05,
+      "loss": 2.5215,
+      "step": 11335
+    },
+    {
+      "epoch": 0.914857557904931,
+      "grad_norm": 0.6988239288330078,
+      "learning_rate": 7.985268925078344e-05,
+      "loss": 2.5153,
+      "step": 11336
+    },
+    {
+      "epoch": 0.914938261641514,
+      "grad_norm": 0.6446006298065186,
+      "learning_rate": 7.983722632954544e-05,
+      "loss": 2.5081,
+      "step": 11337
+    },
+    {
+      "epoch": 0.915018965378097,
+      "grad_norm": 0.6828100681304932,
+      "learning_rate": 7.982176391081649e-05,
+      "loss": 2.5607,
+      "step": 11338
+    },
+    {
+      "epoch": 0.91509966911468,
+      "grad_norm": 0.659721851348877,
+      "learning_rate": 7.980630199498193e-05,
+      "loss": 2.531,
+      "step": 11339
+    },
+    {
+      "epoch": 0.915180372851263,
+      "grad_norm": 0.6298564076423645,
+      "learning_rate": 7.979084058242709e-05,
+      "loss": 2.513,
+      "step": 11340
+    },
+    {
+      "epoch": 0.9152610765878461,
+      "grad_norm": 0.664299726486206,
+      "learning_rate": 7.977537967353735e-05,
+      "loss": 2.5533,
+      "step": 11341
+    },
+    {
+      "epoch": 0.915341780324429,
+      "grad_norm": 0.7035108804702759,
+      "learning_rate": 7.975991926869801e-05,
+      "loss": 2.4868,
+      "step": 11342
+    },
+    {
+      "epoch": 0.915422484061012,
+      "grad_norm": 0.7428407073020935,
+      "learning_rate": 7.974445936829438e-05,
+      "loss": 2.5694,
+      "step": 11343
+    },
+    {
+      "epoch": 0.915503187797595,
+      "grad_norm": 0.6845505237579346,
+      "learning_rate": 7.972899997271176e-05,
+      "loss": 2.5092,
+      "step": 11344
+    },
+    {
+      "epoch": 0.9155838915341781,
+      "grad_norm": 0.7135340571403503,
+      "learning_rate": 7.971354108233551e-05,
+      "loss": 2.5157,
+      "step": 11345
+    },
+    {
+      "epoch": 0.915664595270761,
+      "grad_norm": 0.7032433152198792,
+      "learning_rate": 7.969808269755077e-05,
+      "loss": 2.5292,
+      "step": 11346
+    },
+    {
+      "epoch": 0.915745299007344,
+      "grad_norm": 0.6874690651893616,
+      "learning_rate": 7.96826248187429e-05,
+      "loss": 2.5312,
+      "step": 11347
+    },
+    {
+      "epoch": 0.915826002743927,
+      "grad_norm": 0.6497030258178711,
+      "learning_rate": 7.966716744629718e-05,
+      "loss": 2.505,
+      "step": 11348
+    },
+    {
+      "epoch": 0.9159067064805101,
+      "grad_norm": 0.6618520021438599,
+      "learning_rate": 7.965171058059874e-05,
+      "loss": 2.5287,
+      "step": 11349
+    },
+    {
+      "epoch": 0.9159874102170931,
+      "grad_norm": 0.6737041473388672,
+      "learning_rate": 7.963625422203288e-05,
+      "loss": 2.5494,
+      "step": 11350
+    },
+    {
+      "epoch": 0.916068113953676,
+      "grad_norm": 0.705646276473999,
+      "learning_rate": 7.96207983709848e-05,
+      "loss": 2.5402,
+      "step": 11351
+    },
+    {
+      "epoch": 0.916148817690259,
+      "grad_norm": 0.6852068901062012,
+      "learning_rate": 7.96053430278397e-05,
+      "loss": 2.51,
+      "step": 11352
+    },
+    {
+      "epoch": 0.9162295214268421,
+      "grad_norm": 0.7166822552680969,
+      "learning_rate": 7.958988819298274e-05,
+      "loss": 2.576,
+      "step": 11353
+    },
+    {
+      "epoch": 0.9163102251634251,
+      "grad_norm": 0.6349207162857056,
+      "learning_rate": 7.957443386679913e-05,
+      "loss": 2.5219,
+      "step": 11354
+    },
+    {
+      "epoch": 0.9163909289000081,
+      "grad_norm": 0.6504647135734558,
+      "learning_rate": 7.955898004967406e-05,
+      "loss": 2.4593,
+      "step": 11355
+    },
+    {
+      "epoch": 0.916471632636591,
+      "grad_norm": 0.7313871383666992,
+      "learning_rate": 7.95435267419926e-05,
+      "loss": 2.5616,
+      "step": 11356
+    },
+    {
+      "epoch": 0.9165523363731741,
+      "grad_norm": 0.6948587894439697,
+      "learning_rate": 7.95280739441399e-05,
+      "loss": 2.4608,
+      "step": 11357
+    },
+    {
+      "epoch": 0.9166330401097571,
+      "grad_norm": 0.6130328178405762,
+      "learning_rate": 7.95126216565012e-05,
+      "loss": 2.5563,
+      "step": 11358
+    },
+    {
+      "epoch": 0.9167137438463401,
+      "grad_norm": 0.7149228453636169,
+      "learning_rate": 7.949716987946145e-05,
+      "loss": 2.5664,
+      "step": 11359
+    },
+    {
+      "epoch": 0.916794447582923,
+      "grad_norm": 0.7452285289764404,
+      "learning_rate": 7.948171861340584e-05,
+      "loss": 2.525,
+      "step": 11360
+    },
+    {
+      "epoch": 0.9168751513195061,
+      "grad_norm": 0.6840611100196838,
+      "learning_rate": 7.946626785871945e-05,
+      "loss": 2.537,
+      "step": 11361
+    },
+    {
+      "epoch": 0.9169558550560891,
+      "grad_norm": 0.7269708514213562,
+      "learning_rate": 7.945081761578732e-05,
+      "loss": 2.5227,
+      "step": 11362
+    },
+    {
+      "epoch": 0.9170365587926721,
+      "grad_norm": 0.6521697044372559,
+      "learning_rate": 7.943536788499452e-05,
+      "loss": 2.54,
+      "step": 11363
+    },
+    {
+      "epoch": 0.9171172625292551,
+      "grad_norm": 0.6516863107681274,
+      "learning_rate": 7.941991866672618e-05,
+      "loss": 2.4788,
+      "step": 11364
+    },
+    {
+      "epoch": 0.9171979662658382,
+      "grad_norm": 0.7673580050468445,
+      "learning_rate": 7.94044699613672e-05,
+      "loss": 2.4678,
+      "step": 11365
+    },
+    {
+      "epoch": 0.9172786700024211,
+      "grad_norm": 0.6666994690895081,
+      "learning_rate": 7.938902176930268e-05,
+      "loss": 2.5251,
+      "step": 11366
+    },
+    {
+      "epoch": 0.9173593737390041,
+      "grad_norm": 0.7261863946914673,
+      "learning_rate": 7.937357409091761e-05,
+      "loss": 2.4977,
+      "step": 11367
+    },
+    {
+      "epoch": 0.9174400774755871,
+      "grad_norm": 0.6920679807662964,
+      "learning_rate": 7.9358126926597e-05,
+      "loss": 2.5367,
+      "step": 11368
+    },
+    {
+      "epoch": 0.9175207812121702,
+      "grad_norm": 0.6715712547302246,
+      "learning_rate": 7.93426802767258e-05,
+      "loss": 2.4898,
+      "step": 11369
+    },
+    {
+      "epoch": 0.9176014849487532,
+      "grad_norm": 0.7014333605766296,
+      "learning_rate": 7.932723414168904e-05,
+      "loss": 2.4507,
+      "step": 11370
+    },
+    {
+      "epoch": 0.9176821886853361,
+      "grad_norm": 0.6755761504173279,
+      "learning_rate": 7.931178852187163e-05,
+      "loss": 2.5895,
+      "step": 11371
+    },
+    {
+      "epoch": 0.9177628924219191,
+      "grad_norm": 0.6846731305122375,
+      "learning_rate": 7.929634341765852e-05,
+      "loss": 2.5002,
+      "step": 11372
+    },
+    {
+      "epoch": 0.9178435961585021,
+      "grad_norm": 0.6422831416130066,
+      "learning_rate": 7.928089882943466e-05,
+      "loss": 2.5326,
+      "step": 11373
+    },
+    {
+      "epoch": 0.9179242998950852,
+      "grad_norm": 0.7256442308425903,
+      "learning_rate": 7.9265454757585e-05,
+      "loss": 2.5706,
+      "step": 11374
+    },
+    {
+      "epoch": 0.9180050036316681,
+      "grad_norm": 0.6514387130737305,
+      "learning_rate": 7.925001120249436e-05,
+      "loss": 2.5349,
+      "step": 11375
+    },
+    {
+      "epoch": 0.9180857073682511,
+      "grad_norm": 0.7596457600593567,
+      "learning_rate": 7.923456816454768e-05,
+      "loss": 2.4767,
+      "step": 11376
+    },
+    {
+      "epoch": 0.9181664111048341,
+      "grad_norm": 0.673283040523529,
+      "learning_rate": 7.921912564412988e-05,
+      "loss": 2.5156,
+      "step": 11377
+    },
+    {
+      "epoch": 0.9182471148414172,
+      "grad_norm": 0.6964103579521179,
+      "learning_rate": 7.920368364162575e-05,
+      "loss": 2.5293,
+      "step": 11378
+    },
+    {
+      "epoch": 0.9183278185780002,
+      "grad_norm": 0.6765062212944031,
+      "learning_rate": 7.91882421574202e-05,
+      "loss": 2.5757,
+      "step": 11379
+    },
+    {
+      "epoch": 0.9184085223145831,
+      "grad_norm": 0.7039035558700562,
+      "learning_rate": 7.917280119189811e-05,
+      "loss": 2.513,
+      "step": 11380
+    },
+    {
+      "epoch": 0.9184892260511661,
+      "grad_norm": 0.6523976922035217,
+      "learning_rate": 7.915736074544419e-05,
+      "loss": 2.4712,
+      "step": 11381
+    },
+    {
+      "epoch": 0.9185699297877492,
+      "grad_norm": 0.7159552574157715,
+      "learning_rate": 7.914192081844334e-05,
+      "loss": 2.4713,
+      "step": 11382
+    },
+    {
+      "epoch": 0.9186506335243322,
+      "grad_norm": 0.7071694731712341,
+      "learning_rate": 7.912648141128036e-05,
+      "loss": 2.5367,
+      "step": 11383
+    },
+    {
+      "epoch": 0.9187313372609152,
+      "grad_norm": 0.6675183773040771,
+      "learning_rate": 7.911104252434e-05,
+      "loss": 2.5372,
+      "step": 11384
+    },
+    {
+      "epoch": 0.9188120409974981,
+      "grad_norm": 0.7293995022773743,
+      "learning_rate": 7.909560415800707e-05,
+      "loss": 2.5469,
+      "step": 11385
+    },
+    {
+      "epoch": 0.9188927447340812,
+      "grad_norm": 0.6774035096168518,
+      "learning_rate": 7.908016631266635e-05,
+      "loss": 2.5655,
+      "step": 11386
+    },
+    {
+      "epoch": 0.9189734484706642,
+      "grad_norm": 0.7068144083023071,
+      "learning_rate": 7.906472898870256e-05,
+      "loss": 2.5265,
+      "step": 11387
+    },
+    {
+      "epoch": 0.9190541522072472,
+      "grad_norm": 0.6756324172019958,
+      "learning_rate": 7.904929218650044e-05,
+      "loss": 2.4966,
+      "step": 11388
+    },
+    {
+      "epoch": 0.9191348559438302,
+      "grad_norm": 0.6964625120162964,
+      "learning_rate": 7.903385590644473e-05,
+      "loss": 2.5646,
+      "step": 11389
+    },
+    {
+      "epoch": 0.9192155596804132,
+      "grad_norm": 0.6760976314544678,
+      "learning_rate": 7.901842014892018e-05,
+      "loss": 2.5159,
+      "step": 11390
+    },
+    {
+      "epoch": 0.9192962634169962,
+      "grad_norm": 0.6648714542388916,
+      "learning_rate": 7.900298491431139e-05,
+      "loss": 2.5715,
+      "step": 11391
+    },
+    {
+      "epoch": 0.9193769671535792,
+      "grad_norm": 0.7492914199829102,
+      "learning_rate": 7.898755020300312e-05,
+      "loss": 2.5226,
+      "step": 11392
+    },
+    {
+      "epoch": 0.9194576708901622,
+      "grad_norm": 0.7041164040565491,
+      "learning_rate": 7.897211601538004e-05,
+      "loss": 2.5809,
+      "step": 11393
+    },
+    {
+      "epoch": 0.9195383746267453,
+      "grad_norm": 0.6746383309364319,
+      "learning_rate": 7.895668235182677e-05,
+      "loss": 2.5369,
+      "step": 11394
+    },
+    {
+      "epoch": 0.9196190783633282,
+      "grad_norm": 0.6486156582832336,
+      "learning_rate": 7.894124921272798e-05,
+      "loss": 2.5406,
+      "step": 11395
+    },
+    {
+      "epoch": 0.9196997820999112,
+      "grad_norm": 0.6828807592391968,
+      "learning_rate": 7.892581659846834e-05,
+      "loss": 2.5241,
+      "step": 11396
+    },
+    {
+      "epoch": 0.9197804858364942,
+      "grad_norm": 0.694970428943634,
+      "learning_rate": 7.891038450943242e-05,
+      "loss": 2.4402,
+      "step": 11397
+    },
+    {
+      "epoch": 0.9198611895730773,
+      "grad_norm": 0.7187039852142334,
+      "learning_rate": 7.889495294600484e-05,
+      "loss": 2.5052,
+      "step": 11398
+    },
+    {
+      "epoch": 0.9199418933096603,
+      "grad_norm": 0.6919832825660706,
+      "learning_rate": 7.887952190857024e-05,
+      "loss": 2.5078,
+      "step": 11399
+    },
+    {
+      "epoch": 0.9200225970462432,
+      "grad_norm": 0.7129504084587097,
+      "learning_rate": 7.886409139751313e-05,
+      "loss": 2.5047,
+      "step": 11400
+    },
+    {
+      "epoch": 0.9201033007828262,
+      "grad_norm": 0.6755272746086121,
+      "learning_rate": 7.88486614132181e-05,
+      "loss": 2.4821,
+      "step": 11401
+    },
+    {
+      "epoch": 0.9201840045194093,
+      "grad_norm": 0.7253937125205994,
+      "learning_rate": 7.883323195606973e-05,
+      "loss": 2.5062,
+      "step": 11402
+    },
+    {
+      "epoch": 0.9202647082559923,
+      "grad_norm": 0.7057155966758728,
+      "learning_rate": 7.881780302645257e-05,
+      "loss": 2.5475,
+      "step": 11403
+    },
+    {
+      "epoch": 0.9203454119925752,
+      "grad_norm": 0.713869571685791,
+      "learning_rate": 7.880237462475111e-05,
+      "loss": 2.5335,
+      "step": 11404
+    },
+    {
+      "epoch": 0.9204261157291582,
+      "grad_norm": 0.769648551940918,
+      "learning_rate": 7.878694675134987e-05,
+      "loss": 2.4944,
+      "step": 11405
+    },
+    {
+      "epoch": 0.9205068194657413,
+      "grad_norm": 0.6444964408874512,
+      "learning_rate": 7.877151940663343e-05,
+      "loss": 2.5755,
+      "step": 11406
+    },
+    {
+      "epoch": 0.9205875232023243,
+      "grad_norm": 0.6811819672584534,
+      "learning_rate": 7.875609259098618e-05,
+      "loss": 2.5475,
+      "step": 11407
+    },
+    {
+      "epoch": 0.9206682269389073,
+      "grad_norm": 0.6959417462348938,
+      "learning_rate": 7.874066630479259e-05,
+      "loss": 2.5095,
+      "step": 11408
+    },
+    {
+      "epoch": 0.9207489306754902,
+      "grad_norm": 0.6721363067626953,
+      "learning_rate": 7.872524054843724e-05,
+      "loss": 2.5166,
+      "step": 11409
+    },
+    {
+      "epoch": 0.9208296344120733,
+      "grad_norm": 0.713122546672821,
+      "learning_rate": 7.870981532230447e-05,
+      "loss": 2.5084,
+      "step": 11410
+    },
+    {
+      "epoch": 0.9209103381486563,
+      "grad_norm": 0.7059469819068909,
+      "learning_rate": 7.869439062677876e-05,
+      "loss": 2.437,
+      "step": 11411
+    },
+    {
+      "epoch": 0.9209910418852393,
+      "grad_norm": 0.6808314323425293,
+      "learning_rate": 7.867896646224454e-05,
+      "loss": 2.5658,
+      "step": 11412
+    },
+    {
+      "epoch": 0.9210717456218223,
+      "grad_norm": 0.7060894966125488,
+      "learning_rate": 7.86635428290862e-05,
+      "loss": 2.515,
+      "step": 11413
+    },
+    {
+      "epoch": 0.9211524493584053,
+      "grad_norm": 0.7538465857505798,
+      "learning_rate": 7.864811972768813e-05,
+      "loss": 2.4448,
+      "step": 11414
+    },
+    {
+      "epoch": 0.9212331530949883,
+      "grad_norm": 0.6824522018432617,
+      "learning_rate": 7.863269715843478e-05,
+      "loss": 2.503,
+      "step": 11415
+    },
+    {
+      "epoch": 0.9213138568315713,
+      "grad_norm": 0.7068174481391907,
+      "learning_rate": 7.861727512171044e-05,
+      "loss": 2.5198,
+      "step": 11416
+    },
+    {
+      "epoch": 0.9213945605681543,
+      "grad_norm": 0.6742961406707764,
+      "learning_rate": 7.860185361789948e-05,
+      "loss": 2.5167,
+      "step": 11417
+    },
+    {
+      "epoch": 0.9214752643047374,
+      "grad_norm": 0.7643383741378784,
+      "learning_rate": 7.858643264738628e-05,
+      "loss": 2.5508,
+      "step": 11418
+    },
+    {
+      "epoch": 0.9215559680413203,
+      "grad_norm": 0.6737802028656006,
+      "learning_rate": 7.857101221055518e-05,
+      "loss": 2.589,
+      "step": 11419
+    },
+    {
+      "epoch": 0.9216366717779033,
+      "grad_norm": 0.668214738368988,
+      "learning_rate": 7.855559230779043e-05,
+      "loss": 2.4747,
+      "step": 11420
+    },
+    {
+      "epoch": 0.9217173755144863,
+      "grad_norm": 0.6933084726333618,
+      "learning_rate": 7.854017293947638e-05,
+      "loss": 2.5171,
+      "step": 11421
+    },
+    {
+      "epoch": 0.9217980792510694,
+      "grad_norm": 0.6320228576660156,
+      "learning_rate": 7.852475410599736e-05,
+      "loss": 2.5213,
+      "step": 11422
+    },
+    {
+      "epoch": 0.9218787829876524,
+      "grad_norm": 0.6578245759010315,
+      "learning_rate": 7.850933580773756e-05,
+      "loss": 2.5085,
+      "step": 11423
+    },
+    {
+      "epoch": 0.9219594867242353,
+      "grad_norm": 0.6741796135902405,
+      "learning_rate": 7.849391804508129e-05,
+      "loss": 2.5294,
+      "step": 11424
+    },
+    {
+      "epoch": 0.9220401904608183,
+      "grad_norm": 0.6875781416893005,
+      "learning_rate": 7.847850081841285e-05,
+      "loss": 2.5034,
+      "step": 11425
+    },
+    {
+      "epoch": 0.9221208941974013,
+      "grad_norm": 0.6515244245529175,
+      "learning_rate": 7.846308412811638e-05,
+      "loss": 2.4707,
+      "step": 11426
+    },
+    {
+      "epoch": 0.9222015979339844,
+      "grad_norm": 0.7326812148094177,
+      "learning_rate": 7.844766797457615e-05,
+      "loss": 2.5049,
+      "step": 11427
+    },
+    {
+      "epoch": 0.9222823016705674,
+      "grad_norm": 0.7539918422698975,
+      "learning_rate": 7.84322523581764e-05,
+      "loss": 2.4726,
+      "step": 11428
+    },
+    {
+      "epoch": 0.9223630054071503,
+      "grad_norm": 0.745468020439148,
+      "learning_rate": 7.841683727930129e-05,
+      "loss": 2.5003,
+      "step": 11429
+    },
+    {
+      "epoch": 0.9224437091437333,
+      "grad_norm": 0.726362943649292,
+      "learning_rate": 7.840142273833499e-05,
+      "loss": 2.5056,
+      "step": 11430
+    },
+    {
+      "epoch": 0.9225244128803164,
+      "grad_norm": 0.7275403738021851,
+      "learning_rate": 7.838600873566175e-05,
+      "loss": 2.5188,
+      "step": 11431
+    },
+    {
+      "epoch": 0.9226051166168994,
+      "grad_norm": 0.6908789873123169,
+      "learning_rate": 7.837059527166563e-05,
+      "loss": 2.5349,
+      "step": 11432
+    },
+    {
+      "epoch": 0.9226858203534823,
+      "grad_norm": 0.7220396399497986,
+      "learning_rate": 7.835518234673079e-05,
+      "loss": 2.4863,
+      "step": 11433
+    },
+    {
+      "epoch": 0.9227665240900653,
+      "grad_norm": 0.6516178846359253,
+      "learning_rate": 7.833976996124142e-05,
+      "loss": 2.556,
+      "step": 11434
+    },
+    {
+      "epoch": 0.9228472278266484,
+      "grad_norm": 0.6958726644515991,
+      "learning_rate": 7.832435811558163e-05,
+      "loss": 2.5286,
+      "step": 11435
+    },
+    {
+      "epoch": 0.9229279315632314,
+      "grad_norm": 0.7734121680259705,
+      "learning_rate": 7.830894681013546e-05,
+      "loss": 2.5087,
+      "step": 11436
+    },
+    {
+      "epoch": 0.9230086352998144,
+      "grad_norm": 0.709064245223999,
+      "learning_rate": 7.829353604528703e-05,
+      "loss": 2.4817,
+      "step": 11437
+    },
+    {
+      "epoch": 0.9230893390363973,
+      "grad_norm": 0.7224971652030945,
+      "learning_rate": 7.827812582142045e-05,
+      "loss": 2.5179,
+      "step": 11438
+    },
+    {
+      "epoch": 0.9231700427729804,
+      "grad_norm": 0.7139936685562134,
+      "learning_rate": 7.826271613891973e-05,
+      "loss": 2.537,
+      "step": 11439
+    },
+    {
+      "epoch": 0.9232507465095634,
+      "grad_norm": 0.671138346195221,
+      "learning_rate": 7.824730699816896e-05,
+      "loss": 2.4865,
+      "step": 11440
+    },
+    {
+      "epoch": 0.9233314502461464,
+      "grad_norm": 0.6547425389289856,
+      "learning_rate": 7.823189839955218e-05,
+      "loss": 2.509,
+      "step": 11441
+    },
+    {
+      "epoch": 0.9234121539827294,
+      "grad_norm": 0.719765305519104,
+      "learning_rate": 7.821649034345338e-05,
+      "loss": 2.591,
+      "step": 11442
+    },
+    {
+      "epoch": 0.9234928577193124,
+      "grad_norm": 0.7128504514694214,
+      "learning_rate": 7.820108283025656e-05,
+      "loss": 2.541,
+      "step": 11443
+    },
+    {
+      "epoch": 0.9235735614558954,
+      "grad_norm": 0.7711538672447205,
+      "learning_rate": 7.818567586034577e-05,
+      "loss": 2.5388,
+      "step": 11444
+    },
+    {
+      "epoch": 0.9236542651924784,
+      "grad_norm": 0.7151121497154236,
+      "learning_rate": 7.817026943410494e-05,
+      "loss": 2.5539,
+      "step": 11445
+    },
+    {
+      "epoch": 0.9237349689290614,
+      "grad_norm": 0.7009569406509399,
+      "learning_rate": 7.815486355191805e-05,
+      "loss": 2.4793,
+      "step": 11446
+    },
+    {
+      "epoch": 0.9238156726656445,
+      "grad_norm": 0.7251109480857849,
+      "learning_rate": 7.813945821416909e-05,
+      "loss": 2.5406,
+      "step": 11447
+    },
+    {
+      "epoch": 0.9238963764022274,
+      "grad_norm": 0.6907934546470642,
+      "learning_rate": 7.812405342124196e-05,
+      "loss": 2.5069,
+      "step": 11448
+    },
+    {
+      "epoch": 0.9239770801388104,
+      "grad_norm": 0.699207067489624,
+      "learning_rate": 7.810864917352061e-05,
+      "loss": 2.4844,
+      "step": 11449
+    },
+    {
+      "epoch": 0.9240577838753934,
+      "grad_norm": 0.718386173248291,
+      "learning_rate": 7.809324547138893e-05,
+      "loss": 2.5666,
+      "step": 11450
+    },
+    {
+      "epoch": 0.9241384876119765,
+      "grad_norm": 0.6420444846153259,
+      "learning_rate": 7.807784231523089e-05,
+      "loss": 2.506,
+      "step": 11451
+    },
+    {
+      "epoch": 0.9242191913485595,
+      "grad_norm": 0.6777252554893494,
+      "learning_rate": 7.806243970543028e-05,
+      "loss": 2.487,
+      "step": 11452
+    },
+    {
+      "epoch": 0.9242998950851424,
+      "grad_norm": 0.6907702684402466,
+      "learning_rate": 7.804703764237102e-05,
+      "loss": 2.5284,
+      "step": 11453
+    },
+    {
+      "epoch": 0.9243805988217254,
+      "grad_norm": 0.6383422613143921,
+      "learning_rate": 7.803163612643698e-05,
+      "loss": 2.4704,
+      "step": 11454
+    },
+    {
+      "epoch": 0.9244613025583085,
+      "grad_norm": 0.6879577040672302,
+      "learning_rate": 7.801623515801198e-05,
+      "loss": 2.5103,
+      "step": 11455
+    },
+    {
+      "epoch": 0.9245420062948915,
+      "grad_norm": 0.6856719851493835,
+      "learning_rate": 7.800083473747986e-05,
+      "loss": 2.5086,
+      "step": 11456
+    },
+    {
+      "epoch": 0.9246227100314744,
+      "grad_norm": 0.7463707327842712,
+      "learning_rate": 7.79854348652245e-05,
+      "loss": 2.5456,
+      "step": 11457
+    },
+    {
+      "epoch": 0.9247034137680574,
+      "grad_norm": 0.7352643013000488,
+      "learning_rate": 7.79700355416296e-05,
+      "loss": 2.5335,
+      "step": 11458
+    },
+    {
+      "epoch": 0.9247841175046405,
+      "grad_norm": 0.7525908350944519,
+      "learning_rate": 7.795463676707897e-05,
+      "loss": 2.5855,
+      "step": 11459
+    },
+    {
+      "epoch": 0.9248648212412235,
+      "grad_norm": 0.7323870658874512,
+      "learning_rate": 7.79392385419565e-05,
+      "loss": 2.5471,
+      "step": 11460
+    },
+    {
+      "epoch": 0.9249455249778065,
+      "grad_norm": 0.7443860769271851,
+      "learning_rate": 7.792384086664582e-05,
+      "loss": 2.5449,
+      "step": 11461
+    },
+    {
+      "epoch": 0.9250262287143894,
+      "grad_norm": 0.6928641200065613,
+      "learning_rate": 7.790844374153073e-05,
+      "loss": 2.505,
+      "step": 11462
+    },
+    {
+      "epoch": 0.9251069324509725,
+      "grad_norm": 0.6491222381591797,
+      "learning_rate": 7.789304716699498e-05,
+      "loss": 2.5447,
+      "step": 11463
+    },
+    {
+      "epoch": 0.9251876361875555,
+      "grad_norm": 0.7351166009902954,
+      "learning_rate": 7.78776511434223e-05,
+      "loss": 2.524,
+      "step": 11464
+    },
+    {
+      "epoch": 0.9252683399241385,
+      "grad_norm": 0.6680036783218384,
+      "learning_rate": 7.786225567119637e-05,
+      "loss": 2.5019,
+      "step": 11465
+    },
+    {
+      "epoch": 0.9253490436607215,
+      "grad_norm": 0.7070801258087158,
+      "learning_rate": 7.784686075070089e-05,
+      "loss": 2.5052,
+      "step": 11466
+    },
+    {
+      "epoch": 0.9254297473973045,
+      "grad_norm": 0.7095211148262024,
+      "learning_rate": 7.783146638231957e-05,
+      "loss": 2.4998,
+      "step": 11467
+    },
+    {
+      "epoch": 0.9255104511338875,
+      "grad_norm": 0.6725812554359436,
+      "learning_rate": 7.781607256643604e-05,
+      "loss": 2.4909,
+      "step": 11468
+    },
+    {
+      "epoch": 0.9255911548704705,
+      "grad_norm": 0.684177577495575,
+      "learning_rate": 7.780067930343396e-05,
+      "loss": 2.5636,
+      "step": 11469
+    },
+    {
+      "epoch": 0.9256718586070535,
+      "grad_norm": 0.703419029712677,
+      "learning_rate": 7.778528659369702e-05,
+      "loss": 2.4295,
+      "step": 11470
+    },
+    {
+      "epoch": 0.9257525623436366,
+      "grad_norm": 0.6850195527076721,
+      "learning_rate": 7.776989443760877e-05,
+      "loss": 2.5143,
+      "step": 11471
+    },
+    {
+      "epoch": 0.9258332660802195,
+      "grad_norm": 0.7322348952293396,
+      "learning_rate": 7.775450283555286e-05,
+      "loss": 2.5616,
+      "step": 11472
+    },
+    {
+      "epoch": 0.9259139698168025,
+      "grad_norm": 0.6924510598182678,
+      "learning_rate": 7.77391117879129e-05,
+      "loss": 2.4796,
+      "step": 11473
+    },
+    {
+      "epoch": 0.9259946735533855,
+      "grad_norm": 0.7006441354751587,
+      "learning_rate": 7.772372129507249e-05,
+      "loss": 2.5142,
+      "step": 11474
+    },
+    {
+      "epoch": 0.9260753772899685,
+      "grad_norm": 0.6379218697547913,
+      "learning_rate": 7.770833135741513e-05,
+      "loss": 2.5366,
+      "step": 11475
+    },
+    {
+      "epoch": 0.9261560810265516,
+      "grad_norm": 0.676163375377655,
+      "learning_rate": 7.769294197532448e-05,
+      "loss": 2.4936,
+      "step": 11476
+    },
+    {
+      "epoch": 0.9262367847631345,
+      "grad_norm": 0.6964210271835327,
+      "learning_rate": 7.767755314918399e-05,
+      "loss": 2.429,
+      "step": 11477
+    },
+    {
+      "epoch": 0.9263174884997175,
+      "grad_norm": 0.7017048597335815,
+      "learning_rate": 7.766216487937722e-05,
+      "loss": 2.5488,
+      "step": 11478
+    },
+    {
+      "epoch": 0.9263981922363005,
+      "grad_norm": 0.6742509603500366,
+      "learning_rate": 7.76467771662877e-05,
+      "loss": 2.5121,
+      "step": 11479
+    },
+    {
+      "epoch": 0.9264788959728836,
+      "grad_norm": 0.6751403212547302,
+      "learning_rate": 7.763139001029893e-05,
+      "loss": 2.5897,
+      "step": 11480
+    },
+    {
+      "epoch": 0.9265595997094666,
+      "grad_norm": 0.6639657616615295,
+      "learning_rate": 7.761600341179439e-05,
+      "loss": 2.5015,
+      "step": 11481
+    },
+    {
+      "epoch": 0.9266403034460495,
+      "grad_norm": 0.6332827210426331,
+      "learning_rate": 7.760061737115756e-05,
+      "loss": 2.5518,
+      "step": 11482
+    },
+    {
+      "epoch": 0.9267210071826325,
+      "grad_norm": 0.6751062870025635,
+      "learning_rate": 7.758523188877192e-05,
+      "loss": 2.4252,
+      "step": 11483
+    },
+    {
+      "epoch": 0.9268017109192156,
+      "grad_norm": 0.6763231754302979,
+      "learning_rate": 7.756984696502084e-05,
+      "loss": 2.5683,
+      "step": 11484
+    },
+    {
+      "epoch": 0.9268824146557986,
+      "grad_norm": 0.6480380296707153,
+      "learning_rate": 7.755446260028784e-05,
+      "loss": 2.558,
+      "step": 11485
+    },
+    {
+      "epoch": 0.9269631183923815,
+      "grad_norm": 0.6925072073936462,
+      "learning_rate": 7.753907879495634e-05,
+      "loss": 2.5374,
+      "step": 11486
+    },
+    {
+      "epoch": 0.9270438221289645,
+      "grad_norm": 0.6771834492683411,
+      "learning_rate": 7.752369554940966e-05,
+      "loss": 2.5652,
+      "step": 11487
+    },
+    {
+      "epoch": 0.9271245258655476,
+      "grad_norm": 0.6747026443481445,
+      "learning_rate": 7.750831286403124e-05,
+      "loss": 2.5076,
+      "step": 11488
+    },
+    {
+      "epoch": 0.9272052296021306,
+      "grad_norm": 0.6727211475372314,
+      "learning_rate": 7.749293073920448e-05,
+      "loss": 2.4774,
+      "step": 11489
+    },
+    {
+      "epoch": 0.9272859333387136,
+      "grad_norm": 0.6334055066108704,
+      "learning_rate": 7.747754917531272e-05,
+      "loss": 2.5245,
+      "step": 11490
+    },
+    {
+      "epoch": 0.9273666370752965,
+      "grad_norm": 0.740700900554657,
+      "learning_rate": 7.746216817273928e-05,
+      "loss": 2.5485,
+      "step": 11491
+    },
+    {
+      "epoch": 0.9274473408118796,
+      "grad_norm": 0.6500691771507263,
+      "learning_rate": 7.744678773186757e-05,
+      "loss": 2.5277,
+      "step": 11492
+    },
+    {
+      "epoch": 0.9275280445484626,
+      "grad_norm": 0.6592985987663269,
+      "learning_rate": 7.743140785308084e-05,
+      "loss": 2.5304,
+      "step": 11493
+    },
+    {
+      "epoch": 0.9276087482850456,
+      "grad_norm": 0.6980452537536621,
+      "learning_rate": 7.741602853676241e-05,
+      "loss": 2.544,
+      "step": 11494
+    },
+    {
+      "epoch": 0.9276894520216286,
+      "grad_norm": 0.643190860748291,
+      "learning_rate": 7.740064978329555e-05,
+      "loss": 2.5167,
+      "step": 11495
+    },
+    {
+      "epoch": 0.9277701557582116,
+      "grad_norm": 0.6789804100990295,
+      "learning_rate": 7.738527159306366e-05,
+      "loss": 2.5117,
+      "step": 11496
+    },
+    {
+      "epoch": 0.9278508594947946,
+      "grad_norm": 0.7109663486480713,
+      "learning_rate": 7.736989396644987e-05,
+      "loss": 2.5294,
+      "step": 11497
+    },
+    {
+      "epoch": 0.9279315632313776,
+      "grad_norm": 0.6752706170082092,
+      "learning_rate": 7.735451690383746e-05,
+      "loss": 2.4851,
+      "step": 11498
+    },
+    {
+      "epoch": 0.9280122669679606,
+      "grad_norm": 0.6947829723358154,
+      "learning_rate": 7.733914040560972e-05,
+      "loss": 2.5792,
+      "step": 11499
+    },
+    {
+      "epoch": 0.9280929707045437,
+      "grad_norm": 0.6701157689094543,
+      "learning_rate": 7.732376447214981e-05,
+      "loss": 2.4884,
+      "step": 11500
+    },
+    {
+      "epoch": 0.9281736744411266,
+      "grad_norm": 0.64533531665802,
+      "learning_rate": 7.730838910384097e-05,
+      "loss": 2.4644,
+      "step": 11501
+    },
+    {
+      "epoch": 0.9282543781777096,
+      "grad_norm": 0.6664395332336426,
+      "learning_rate": 7.729301430106644e-05,
+      "loss": 2.5286,
+      "step": 11502
+    },
+    {
+      "epoch": 0.9283350819142926,
+      "grad_norm": 0.6982395648956299,
+      "learning_rate": 7.72776400642093e-05,
+      "loss": 2.5092,
+      "step": 11503
+    },
+    {
+      "epoch": 0.9284157856508757,
+      "grad_norm": 0.6656171679496765,
+      "learning_rate": 7.726226639365278e-05,
+      "loss": 2.4945,
+      "step": 11504
+    },
+    {
+      "epoch": 0.9284964893874587,
+      "grad_norm": 0.6213308572769165,
+      "learning_rate": 7.724689328978001e-05,
+      "loss": 2.5042,
+      "step": 11505
+    },
+    {
+      "epoch": 0.9285771931240416,
+      "grad_norm": 0.6855599880218506,
+      "learning_rate": 7.723152075297414e-05,
+      "loss": 2.5207,
+      "step": 11506
+    },
+    {
+      "epoch": 0.9286578968606246,
+      "grad_norm": 0.7724171280860901,
+      "learning_rate": 7.721614878361828e-05,
+      "loss": 2.4842,
+      "step": 11507
+    },
+    {
+      "epoch": 0.9287386005972077,
+      "grad_norm": 0.708634614944458,
+      "learning_rate": 7.720077738209559e-05,
+      "loss": 2.58,
+      "step": 11508
+    },
+    {
+      "epoch": 0.9288193043337907,
+      "grad_norm": 0.6766082644462585,
+      "learning_rate": 7.718540654878907e-05,
+      "loss": 2.492,
+      "step": 11509
+    },
+    {
+      "epoch": 0.9289000080703737,
+      "grad_norm": 0.6856982707977295,
+      "learning_rate": 7.717003628408187e-05,
+      "loss": 2.5186,
+      "step": 11510
+    },
+    {
+      "epoch": 0.9289807118069566,
+      "grad_norm": 0.680647611618042,
+      "learning_rate": 7.715466658835705e-05,
+      "loss": 2.5305,
+      "step": 11511
+    },
+    {
+      "epoch": 0.9290614155435397,
+      "grad_norm": 0.7174721360206604,
+      "learning_rate": 7.713929746199771e-05,
+      "loss": 2.4498,
+      "step": 11512
+    },
+    {
+      "epoch": 0.9291421192801227,
+      "grad_norm": 0.6507031321525574,
+      "learning_rate": 7.712392890538676e-05,
+      "loss": 2.5334,
+      "step": 11513
+    },
+    {
+      "epoch": 0.9292228230167057,
+      "grad_norm": 0.7545748353004456,
+      "learning_rate": 7.710856091890732e-05,
+      "loss": 2.505,
+      "step": 11514
+    },
+    {
+      "epoch": 0.9293035267532886,
+      "grad_norm": 0.6978560090065002,
+      "learning_rate": 7.709319350294242e-05,
+      "loss": 2.5243,
+      "step": 11515
+    },
+    {
+      "epoch": 0.9293842304898717,
+      "grad_norm": 0.6620199084281921,
+      "learning_rate": 7.707782665787497e-05,
+      "loss": 2.5114,
+      "step": 11516
+    },
+    {
+      "epoch": 0.9294649342264547,
+      "grad_norm": 0.7160476446151733,
+      "learning_rate": 7.7062460384088e-05,
+      "loss": 2.5322,
+      "step": 11517
+    },
+    {
+      "epoch": 0.9295456379630377,
+      "grad_norm": 0.6637005805969238,
+      "learning_rate": 7.704709468196454e-05,
+      "loss": 2.456,
+      "step": 11518
+    },
+    {
+      "epoch": 0.9296263416996207,
+      "grad_norm": 0.6668851375579834,
+      "learning_rate": 7.703172955188742e-05,
+      "loss": 2.5251,
+      "step": 11519
+    },
+    {
+      "epoch": 0.9297070454362037,
+      "grad_norm": 0.6840329170227051,
+      "learning_rate": 7.701636499423965e-05,
+      "loss": 2.5068,
+      "step": 11520
+    },
+    {
+      "epoch": 0.9297877491727867,
+      "grad_norm": 0.695122241973877,
+      "learning_rate": 7.700100100940415e-05,
+      "loss": 2.4822,
+      "step": 11521
+    },
+    {
+      "epoch": 0.9298684529093697,
+      "grad_norm": 0.6784923672676086,
+      "learning_rate": 7.698563759776382e-05,
+      "loss": 2.4978,
+      "step": 11522
+    },
+    {
+      "epoch": 0.9299491566459527,
+      "grad_norm": 0.6949357986450195,
+      "learning_rate": 7.697027475970154e-05,
+      "loss": 2.5392,
+      "step": 11523
+    },
+    {
+      "epoch": 0.9300298603825358,
+      "grad_norm": 0.7128093242645264,
+      "learning_rate": 7.695491249560025e-05,
+      "loss": 2.455,
+      "step": 11524
+    },
+    {
+      "epoch": 0.9301105641191187,
+      "grad_norm": 0.6534962058067322,
+      "learning_rate": 7.693955080584277e-05,
+      "loss": 2.5272,
+      "step": 11525
+    },
+    {
+      "epoch": 0.9301912678557017,
+      "grad_norm": 0.6893511414527893,
+      "learning_rate": 7.692418969081194e-05,
+      "loss": 2.5366,
+      "step": 11526
+    },
+    {
+      "epoch": 0.9302719715922847,
+      "grad_norm": 0.6335335373878479,
+      "learning_rate": 7.690882915089064e-05,
+      "loss": 2.5781,
+      "step": 11527
+    },
+    {
+      "epoch": 0.9303526753288677,
+      "grad_norm": 0.7264769077301025,
+      "learning_rate": 7.689346918646172e-05,
+      "loss": 2.5322,
+      "step": 11528
+    },
+    {
+      "epoch": 0.9304333790654508,
+      "grad_norm": 0.7156329154968262,
+      "learning_rate": 7.68781097979079e-05,
+      "loss": 2.5558,
+      "step": 11529
+    },
+    {
+      "epoch": 0.9305140828020337,
+      "grad_norm": 0.6914563775062561,
+      "learning_rate": 7.686275098561203e-05,
+      "loss": 2.5058,
+      "step": 11530
+    },
+    {
+      "epoch": 0.9305947865386167,
+      "grad_norm": 0.6939939260482788,
+      "learning_rate": 7.684739274995691e-05,
+      "loss": 2.4764,
+      "step": 11531
+    },
+    {
+      "epoch": 0.9306754902751997,
+      "grad_norm": 0.7103014588356018,
+      "learning_rate": 7.683203509132526e-05,
+      "loss": 2.5062,
+      "step": 11532
+    },
+    {
+      "epoch": 0.9307561940117828,
+      "grad_norm": 0.6558870077133179,
+      "learning_rate": 7.681667801009985e-05,
+      "loss": 2.4869,
+      "step": 11533
+    },
+    {
+      "epoch": 0.9308368977483658,
+      "grad_norm": 0.7280104160308838,
+      "learning_rate": 7.680132150666348e-05,
+      "loss": 2.566,
+      "step": 11534
+    },
+    {
+      "epoch": 0.9309176014849487,
+      "grad_norm": 0.6814180612564087,
+      "learning_rate": 7.678596558139875e-05,
+      "loss": 2.4926,
+      "step": 11535
+    },
+    {
+      "epoch": 0.9309983052215317,
+      "grad_norm": 0.6916589736938477,
+      "learning_rate": 7.677061023468846e-05,
+      "loss": 2.5189,
+      "step": 11536
+    },
+    {
+      "epoch": 0.9310790089581148,
+      "grad_norm": 0.6527554988861084,
+      "learning_rate": 7.675525546691533e-05,
+      "loss": 2.4969,
+      "step": 11537
+    },
+    {
+      "epoch": 0.9311597126946978,
+      "grad_norm": 0.6458954811096191,
+      "learning_rate": 7.673990127846196e-05,
+      "loss": 2.5159,
+      "step": 11538
+    },
+    {
+      "epoch": 0.9312404164312807,
+      "grad_norm": 0.6704902052879333,
+      "learning_rate": 7.672454766971105e-05,
+      "loss": 2.49,
+      "step": 11539
+    },
+    {
+      "epoch": 0.9313211201678637,
+      "grad_norm": 0.6599698066711426,
+      "learning_rate": 7.670919464104527e-05,
+      "loss": 2.4872,
+      "step": 11540
+    },
+    {
+      "epoch": 0.9314018239044468,
+      "grad_norm": 0.7638888955116272,
+      "learning_rate": 7.669384219284722e-05,
+      "loss": 2.5228,
+      "step": 11541
+    },
+    {
+      "epoch": 0.9314825276410298,
+      "grad_norm": 0.6911981105804443,
+      "learning_rate": 7.667849032549954e-05,
+      "loss": 2.4675,
+      "step": 11542
+    },
+    {
+      "epoch": 0.9315632313776128,
+      "grad_norm": 0.6414669156074524,
+      "learning_rate": 7.666313903938486e-05,
+      "loss": 2.5137,
+      "step": 11543
+    },
+    {
+      "epoch": 0.9316439351141957,
+      "grad_norm": 0.7552139759063721,
+      "learning_rate": 7.66477883348858e-05,
+      "loss": 2.5778,
+      "step": 11544
+    },
+    {
+      "epoch": 0.9317246388507788,
+      "grad_norm": 0.6738760471343994,
+      "learning_rate": 7.663243821238484e-05,
+      "loss": 2.5326,
+      "step": 11545
+    },
+    {
+      "epoch": 0.9318053425873618,
+      "grad_norm": 0.7406899333000183,
+      "learning_rate": 7.661708867226459e-05,
+      "loss": 2.4608,
+      "step": 11546
+    },
+    {
+      "epoch": 0.9318860463239448,
+      "grad_norm": 0.7261415719985962,
+      "learning_rate": 7.660173971490769e-05,
+      "loss": 2.5684,
+      "step": 11547
+    },
+    {
+      "epoch": 0.9319667500605278,
+      "grad_norm": 0.636542797088623,
+      "learning_rate": 7.658639134069654e-05,
+      "loss": 2.5159,
+      "step": 11548
+    },
+    {
+      "epoch": 0.9320474537971108,
+      "grad_norm": 0.7730209231376648,
+      "learning_rate": 7.657104355001373e-05,
+      "loss": 2.487,
+      "step": 11549
+    },
+    {
+      "epoch": 0.9321281575336938,
+      "grad_norm": 0.6553641557693481,
+      "learning_rate": 7.655569634324178e-05,
+      "loss": 2.5105,
+      "step": 11550
+    },
+    {
+      "epoch": 0.9322088612702768,
+      "grad_norm": 0.7008326649665833,
+      "learning_rate": 7.654034972076314e-05,
+      "loss": 2.492,
+      "step": 11551
+    },
+    {
+      "epoch": 0.9322895650068598,
+      "grad_norm": 0.7074279189109802,
+      "learning_rate": 7.65250036829603e-05,
+      "loss": 2.5221,
+      "step": 11552
+    },
+    {
+      "epoch": 0.9323702687434429,
+      "grad_norm": 0.7235530018806458,
+      "learning_rate": 7.650965823021578e-05,
+      "loss": 2.5285,
+      "step": 11553
+    },
+    {
+      "epoch": 0.9324509724800258,
+      "grad_norm": 0.7601436376571655,
+      "learning_rate": 7.649431336291194e-05,
+      "loss": 2.5071,
+      "step": 11554
+    },
+    {
+      "epoch": 0.9325316762166088,
+      "grad_norm": 0.6446424126625061,
+      "learning_rate": 7.647896908143127e-05,
+      "loss": 2.5032,
+      "step": 11555
+    },
+    {
+      "epoch": 0.9326123799531918,
+      "grad_norm": 0.7032139897346497,
+      "learning_rate": 7.646362538615614e-05,
+      "loss": 2.6096,
+      "step": 11556
+    },
+    {
+      "epoch": 0.9326930836897749,
+      "grad_norm": 0.6727899312973022,
+      "learning_rate": 7.644828227746904e-05,
+      "loss": 2.5041,
+      "step": 11557
+    },
+    {
+      "epoch": 0.9327737874263579,
+      "grad_norm": 0.6817529201507568,
+      "learning_rate": 7.643293975575229e-05,
+      "loss": 2.4474,
+      "step": 11558
+    },
+    {
+      "epoch": 0.9328544911629408,
+      "grad_norm": 0.6374444365501404,
+      "learning_rate": 7.641759782138827e-05,
+      "loss": 2.5204,
+      "step": 11559
+    },
+    {
+      "epoch": 0.9329351948995238,
+      "grad_norm": 0.6889457702636719,
+      "learning_rate": 7.640225647475939e-05,
+      "loss": 2.6344,
+      "step": 11560
+    },
+    {
+      "epoch": 0.9330158986361069,
+      "grad_norm": 0.6657958626747131,
+      "learning_rate": 7.638691571624794e-05,
+      "loss": 2.4672,
+      "step": 11561
+    },
+    {
+      "epoch": 0.9330966023726899,
+      "grad_norm": 0.6425464749336243,
+      "learning_rate": 7.637157554623627e-05,
+      "loss": 2.4756,
+      "step": 11562
+    },
+    {
+      "epoch": 0.9331773061092729,
+      "grad_norm": 0.7193450927734375,
+      "learning_rate": 7.635623596510675e-05,
+      "loss": 2.4969,
+      "step": 11563
+    },
+    {
+      "epoch": 0.9332580098458558,
+      "grad_norm": 0.6595252156257629,
+      "learning_rate": 7.634089697324159e-05,
+      "loss": 2.4647,
+      "step": 11564
+    },
+    {
+      "epoch": 0.9333387135824389,
+      "grad_norm": 0.6505268812179565,
+      "learning_rate": 7.632555857102312e-05,
+      "loss": 2.5059,
+      "step": 11565
+    },
+    {
+      "epoch": 0.9334194173190219,
+      "grad_norm": 0.6877838969230652,
+      "learning_rate": 7.631022075883365e-05,
+      "loss": 2.4855,
+      "step": 11566
+    },
+    {
+      "epoch": 0.9335001210556049,
+      "grad_norm": 0.6376198530197144,
+      "learning_rate": 7.629488353705538e-05,
+      "loss": 2.5024,
+      "step": 11567
+    },
+    {
+      "epoch": 0.9335808247921878,
+      "grad_norm": 0.6807642579078674,
+      "learning_rate": 7.627954690607058e-05,
+      "loss": 2.4954,
+      "step": 11568
+    },
+    {
+      "epoch": 0.9336615285287709,
+      "grad_norm": 0.6785219311714172,
+      "learning_rate": 7.62642108662615e-05,
+      "loss": 2.4854,
+      "step": 11569
+    },
+    {
+      "epoch": 0.9337422322653539,
+      "grad_norm": 0.8159591555595398,
+      "learning_rate": 7.624887541801032e-05,
+      "loss": 2.524,
+      "step": 11570
+    },
+    {
+      "epoch": 0.9338229360019369,
+      "grad_norm": 0.6912592053413391,
+      "learning_rate": 7.62335405616992e-05,
+      "loss": 2.5111,
+      "step": 11571
+    },
+    {
+      "epoch": 0.9339036397385199,
+      "grad_norm": 0.6772454977035522,
+      "learning_rate": 7.621820629771041e-05,
+      "loss": 2.5603,
+      "step": 11572
+    },
+    {
+      "epoch": 0.933984343475103,
+      "grad_norm": 0.6720221638679504,
+      "learning_rate": 7.620287262642613e-05,
+      "loss": 2.5016,
+      "step": 11573
+    },
+    {
+      "epoch": 0.9340650472116859,
+      "grad_norm": 0.651935338973999,
+      "learning_rate": 7.618753954822841e-05,
+      "loss": 2.445,
+      "step": 11574
+    },
+    {
+      "epoch": 0.9341457509482689,
+      "grad_norm": 0.6731166839599609,
+      "learning_rate": 7.617220706349947e-05,
+      "loss": 2.4703,
+      "step": 11575
+    },
+    {
+      "epoch": 0.9342264546848519,
+      "grad_norm": 0.6283879280090332,
+      "learning_rate": 7.615687517262143e-05,
+      "loss": 2.5232,
+      "step": 11576
+    },
+    {
+      "epoch": 0.9343071584214349,
+      "grad_norm": 0.7193455696105957,
+      "learning_rate": 7.614154387597638e-05,
+      "loss": 2.5268,
+      "step": 11577
+    },
+    {
+      "epoch": 0.934387862158018,
+      "grad_norm": 0.6992828845977783,
+      "learning_rate": 7.61262131739464e-05,
+      "loss": 2.5834,
+      "step": 11578
+    },
+    {
+      "epoch": 0.9344685658946009,
+      "grad_norm": 0.6501220464706421,
+      "learning_rate": 7.611088306691365e-05,
+      "loss": 2.5146,
+      "step": 11579
+    },
+    {
+      "epoch": 0.9345492696311839,
+      "grad_norm": 0.7246220111846924,
+      "learning_rate": 7.60955535552601e-05,
+      "loss": 2.5665,
+      "step": 11580
+    },
+    {
+      "epoch": 0.9346299733677669,
+      "grad_norm": 0.7190428376197815,
+      "learning_rate": 7.608022463936783e-05,
+      "loss": 2.5061,
+      "step": 11581
+    },
+    {
+      "epoch": 0.93471067710435,
+      "grad_norm": 0.7144324779510498,
+      "learning_rate": 7.606489631961893e-05,
+      "loss": 2.4982,
+      "step": 11582
+    },
+    {
+      "epoch": 0.9347913808409329,
+      "grad_norm": 0.7144657373428345,
+      "learning_rate": 7.604956859639535e-05,
+      "loss": 2.5506,
+      "step": 11583
+    },
+    {
+      "epoch": 0.9348720845775159,
+      "grad_norm": 0.6596626043319702,
+      "learning_rate": 7.603424147007913e-05,
+      "loss": 2.4911,
+      "step": 11584
+    },
+    {
+      "epoch": 0.9349527883140989,
+      "grad_norm": 0.7090883851051331,
+      "learning_rate": 7.601891494105227e-05,
+      "loss": 2.5087,
+      "step": 11585
+    },
+    {
+      "epoch": 0.935033492050682,
+      "grad_norm": 0.6679760217666626,
+      "learning_rate": 7.600358900969671e-05,
+      "loss": 2.497,
+      "step": 11586
+    },
+    {
+      "epoch": 0.935114195787265,
+      "grad_norm": 0.6795344948768616,
+      "learning_rate": 7.598826367639447e-05,
+      "loss": 2.4839,
+      "step": 11587
+    },
+    {
+      "epoch": 0.9351948995238479,
+      "grad_norm": 0.6378790736198425,
+      "learning_rate": 7.597293894152744e-05,
+      "loss": 2.4656,
+      "step": 11588
+    },
+    {
+      "epoch": 0.9352756032604309,
+      "grad_norm": 0.6646658182144165,
+      "learning_rate": 7.595761480547762e-05,
+      "loss": 2.4739,
+      "step": 11589
+    },
+    {
+      "epoch": 0.935356306997014,
+      "grad_norm": 0.6662073731422424,
+      "learning_rate": 7.594229126862687e-05,
+      "loss": 2.4872,
+      "step": 11590
+    },
+    {
+      "epoch": 0.935437010733597,
+      "grad_norm": 0.6698113679885864,
+      "learning_rate": 7.592696833135708e-05,
+      "loss": 2.4964,
+      "step": 11591
+    },
+    {
+      "epoch": 0.93551771447018,
+      "grad_norm": 0.6520004272460938,
+      "learning_rate": 7.59116459940502e-05,
+      "loss": 2.5616,
+      "step": 11592
+    },
+    {
+      "epoch": 0.9355984182067629,
+      "grad_norm": 0.6675869226455688,
+      "learning_rate": 7.589632425708806e-05,
+      "loss": 2.4854,
+      "step": 11593
+    },
+    {
+      "epoch": 0.935679121943346,
+      "grad_norm": 0.6914103031158447,
+      "learning_rate": 7.588100312085251e-05,
+      "loss": 2.5252,
+      "step": 11594
+    },
+    {
+      "epoch": 0.935759825679929,
+      "grad_norm": 0.7283286452293396,
+      "learning_rate": 7.586568258572546e-05,
+      "loss": 2.543,
+      "step": 11595
+    },
+    {
+      "epoch": 0.935840529416512,
+      "grad_norm": 0.6881958246231079,
+      "learning_rate": 7.585036265208864e-05,
+      "loss": 2.4499,
+      "step": 11596
+    },
+    {
+      "epoch": 0.935921233153095,
+      "grad_norm": 0.7733677625656128,
+      "learning_rate": 7.58350433203239e-05,
+      "loss": 2.5595,
+      "step": 11597
+    },
+    {
+      "epoch": 0.936001936889678,
+      "grad_norm": 0.672711968421936,
+      "learning_rate": 7.58197245908131e-05,
+      "loss": 2.4757,
+      "step": 11598
+    },
+    {
+      "epoch": 0.936082640626261,
+      "grad_norm": 0.691780686378479,
+      "learning_rate": 7.580440646393794e-05,
+      "loss": 2.5134,
+      "step": 11599
+    },
+    {
+      "epoch": 0.936163344362844,
+      "grad_norm": 0.6935102343559265,
+      "learning_rate": 7.578908894008021e-05,
+      "loss": 2.5128,
+      "step": 11600
+    },
+    {
+      "epoch": 0.936244048099427,
+      "grad_norm": 0.7005696892738342,
+      "learning_rate": 7.57737720196217e-05,
+      "loss": 2.5338,
+      "step": 11601
+    },
+    {
+      "epoch": 0.93632475183601,
+      "grad_norm": 0.6729815602302551,
+      "learning_rate": 7.575845570294409e-05,
+      "loss": 2.5373,
+      "step": 11602
+    },
+    {
+      "epoch": 0.936405455572593,
+      "grad_norm": 0.6694760918617249,
+      "learning_rate": 7.574313999042913e-05,
+      "loss": 2.5165,
+      "step": 11603
+    },
+    {
+      "epoch": 0.936486159309176,
+      "grad_norm": 0.6425337791442871,
+      "learning_rate": 7.572782488245854e-05,
+      "loss": 2.5102,
+      "step": 11604
+    },
+    {
+      "epoch": 0.936566863045759,
+      "grad_norm": 0.6613046526908875,
+      "learning_rate": 7.571251037941405e-05,
+      "loss": 2.5108,
+      "step": 11605
+    },
+    {
+      "epoch": 0.9366475667823421,
+      "grad_norm": 0.7396309971809387,
+      "learning_rate": 7.569719648167723e-05,
+      "loss": 2.5261,
+      "step": 11606
+    },
+    {
+      "epoch": 0.936728270518925,
+      "grad_norm": 0.6783239245414734,
+      "learning_rate": 7.568188318962981e-05,
+      "loss": 2.5725,
+      "step": 11607
+    },
+    {
+      "epoch": 0.936808974255508,
+      "grad_norm": 0.7591684460639954,
+      "learning_rate": 7.566657050365345e-05,
+      "loss": 2.5085,
+      "step": 11608
+    },
+    {
+      "epoch": 0.936889677992091,
+      "grad_norm": 0.6805615425109863,
+      "learning_rate": 7.565125842412974e-05,
+      "loss": 2.5598,
+      "step": 11609
+    },
+    {
+      "epoch": 0.9369703817286741,
+      "grad_norm": 0.680203378200531,
+      "learning_rate": 7.563594695144032e-05,
+      "loss": 2.5072,
+      "step": 11610
+    },
+    {
+      "epoch": 0.9370510854652571,
+      "grad_norm": 0.7035777568817139,
+      "learning_rate": 7.56206360859668e-05,
+      "loss": 2.4882,
+      "step": 11611
+    },
+    {
+      "epoch": 0.93713178920184,
+      "grad_norm": 0.7457048892974854,
+      "learning_rate": 7.560532582809075e-05,
+      "loss": 2.4975,
+      "step": 11612
+    },
+    {
+      "epoch": 0.937212492938423,
+      "grad_norm": 0.702055037021637,
+      "learning_rate": 7.559001617819374e-05,
+      "loss": 2.5522,
+      "step": 11613
+    },
+    {
+      "epoch": 0.9372931966750061,
+      "grad_norm": 0.7618527412414551,
+      "learning_rate": 7.557470713665738e-05,
+      "loss": 2.5503,
+      "step": 11614
+    },
+    {
+      "epoch": 0.9373739004115891,
+      "grad_norm": 0.8611559867858887,
+      "learning_rate": 7.555939870386312e-05,
+      "loss": 2.4866,
+      "step": 11615
+    },
+    {
+      "epoch": 0.937454604148172,
+      "grad_norm": 0.7285227179527283,
+      "learning_rate": 7.554409088019254e-05,
+      "loss": 2.4855,
+      "step": 11616
+    },
+    {
+      "epoch": 0.937535307884755,
+      "grad_norm": 0.7512121796607971,
+      "learning_rate": 7.552878366602716e-05,
+      "loss": 2.5496,
+      "step": 11617
+    },
+    {
+      "epoch": 0.9376160116213381,
+      "grad_norm": 0.7353625297546387,
+      "learning_rate": 7.551347706174844e-05,
+      "loss": 2.5754,
+      "step": 11618
+    },
+    {
+      "epoch": 0.9376967153579211,
+      "grad_norm": 0.7131205797195435,
+      "learning_rate": 7.549817106773788e-05,
+      "loss": 2.4927,
+      "step": 11619
+    },
+    {
+      "epoch": 0.9377774190945041,
+      "grad_norm": 0.6562477946281433,
+      "learning_rate": 7.548286568437695e-05,
+      "loss": 2.5247,
+      "step": 11620
+    },
+    {
+      "epoch": 0.937858122831087,
+      "grad_norm": 0.7094948887825012,
+      "learning_rate": 7.546756091204713e-05,
+      "loss": 2.5084,
+      "step": 11621
+    },
+    {
+      "epoch": 0.9379388265676701,
+      "grad_norm": 0.6890475153923035,
+      "learning_rate": 7.545225675112977e-05,
+      "loss": 2.5178,
+      "step": 11622
+    },
+    {
+      "epoch": 0.9380195303042531,
+      "grad_norm": 0.6801474094390869,
+      "learning_rate": 7.543695320200634e-05,
+      "loss": 2.5457,
+      "step": 11623
+    },
+    {
+      "epoch": 0.9381002340408361,
+      "grad_norm": 0.7093712687492371,
+      "learning_rate": 7.54216502650583e-05,
+      "loss": 2.6122,
+      "step": 11624
+    },
+    {
+      "epoch": 0.9381809377774191,
+      "grad_norm": 0.7246927618980408,
+      "learning_rate": 7.540634794066695e-05,
+      "loss": 2.5251,
+      "step": 11625
+    },
+    {
+      "epoch": 0.9382616415140022,
+      "grad_norm": 0.7358111143112183,
+      "learning_rate": 7.539104622921368e-05,
+      "loss": 2.5444,
+      "step": 11626
+    },
+    {
+      "epoch": 0.9383423452505851,
+      "grad_norm": 0.6915993690490723,
+      "learning_rate": 7.53757451310799e-05,
+      "loss": 2.448,
+      "step": 11627
+    },
+    {
+      "epoch": 0.9384230489871681,
+      "grad_norm": 0.6864039301872253,
+      "learning_rate": 7.536044464664689e-05,
+      "loss": 2.5267,
+      "step": 11628
+    },
+    {
+      "epoch": 0.9385037527237511,
+      "grad_norm": 0.664799690246582,
+      "learning_rate": 7.534514477629602e-05,
+      "loss": 2.5602,
+      "step": 11629
+    },
+    {
+      "epoch": 0.9385844564603341,
+      "grad_norm": 0.6770062446594238,
+      "learning_rate": 7.532984552040862e-05,
+      "loss": 2.5034,
+      "step": 11630
+    },
+    {
+      "epoch": 0.9386651601969171,
+      "grad_norm": 0.6961095929145813,
+      "learning_rate": 7.531454687936592e-05,
+      "loss": 2.4523,
+      "step": 11631
+    },
+    {
+      "epoch": 0.9387458639335001,
+      "grad_norm": 0.6776804327964783,
+      "learning_rate": 7.529924885354924e-05,
+      "loss": 2.5526,
+      "step": 11632
+    },
+    {
+      "epoch": 0.9388265676700831,
+      "grad_norm": 0.785796582698822,
+      "learning_rate": 7.528395144333988e-05,
+      "loss": 2.5256,
+      "step": 11633
+    },
+    {
+      "epoch": 0.9389072714066661,
+      "grad_norm": 0.7016655206680298,
+      "learning_rate": 7.526865464911902e-05,
+      "loss": 2.4781,
+      "step": 11634
+    },
+    {
+      "epoch": 0.9389879751432492,
+      "grad_norm": 0.7027767300605774,
+      "learning_rate": 7.525335847126795e-05,
+      "loss": 2.5287,
+      "step": 11635
+    },
+    {
+      "epoch": 0.9390686788798321,
+      "grad_norm": 0.710624098777771,
+      "learning_rate": 7.523806291016787e-05,
+      "loss": 2.5486,
+      "step": 11636
+    },
+    {
+      "epoch": 0.9391493826164151,
+      "grad_norm": 0.7029656767845154,
+      "learning_rate": 7.52227679662e-05,
+      "loss": 2.5244,
+      "step": 11637
+    },
+    {
+      "epoch": 0.9392300863529981,
+      "grad_norm": 0.7417333722114563,
+      "learning_rate": 7.520747363974551e-05,
+      "loss": 2.5561,
+      "step": 11638
+    },
+    {
+      "epoch": 0.9393107900895812,
+      "grad_norm": 0.6595067381858826,
+      "learning_rate": 7.519217993118559e-05,
+      "loss": 2.617,
+      "step": 11639
+    },
+    {
+      "epoch": 0.9393914938261642,
+      "grad_norm": 0.6808187365531921,
+      "learning_rate": 7.517688684090141e-05,
+      "loss": 2.5279,
+      "step": 11640
+    },
+    {
+      "epoch": 0.9394721975627471,
+      "grad_norm": 0.6618706583976746,
+      "learning_rate": 7.516159436927408e-05,
+      "loss": 2.4976,
+      "step": 11641
+    },
+    {
+      "epoch": 0.9395529012993301,
+      "grad_norm": 0.6979385018348694,
+      "learning_rate": 7.514630251668475e-05,
+      "loss": 2.4542,
+      "step": 11642
+    },
+    {
+      "epoch": 0.9396336050359132,
+      "grad_norm": 0.6380844116210938,
+      "learning_rate": 7.513101128351454e-05,
+      "loss": 2.48,
+      "step": 11643
+    },
+    {
+      "epoch": 0.9397143087724962,
+      "grad_norm": 0.6390014290809631,
+      "learning_rate": 7.511572067014452e-05,
+      "loss": 2.5111,
+      "step": 11644
+    },
+    {
+      "epoch": 0.9397950125090792,
+      "grad_norm": 0.7592498064041138,
+      "learning_rate": 7.510043067695578e-05,
+      "loss": 2.5161,
+      "step": 11645
+    },
+    {
+      "epoch": 0.9398757162456621,
+      "grad_norm": 0.6269322037696838,
+      "learning_rate": 7.508514130432945e-05,
+      "loss": 2.491,
+      "step": 11646
+    },
+    {
+      "epoch": 0.9399564199822452,
+      "grad_norm": 0.6372053623199463,
+      "learning_rate": 7.506985255264646e-05,
+      "loss": 2.4826,
+      "step": 11647
+    },
+    {
+      "epoch": 0.9400371237188282,
+      "grad_norm": 0.6962460875511169,
+      "learning_rate": 7.505456442228794e-05,
+      "loss": 2.5605,
+      "step": 11648
+    },
+    {
+      "epoch": 0.9401178274554112,
+      "grad_norm": 0.7931656241416931,
+      "learning_rate": 7.503927691363491e-05,
+      "loss": 2.4909,
+      "step": 11649
+    },
+    {
+      "epoch": 0.9401985311919941,
+      "grad_norm": 0.688792884349823,
+      "learning_rate": 7.502399002706832e-05,
+      "loss": 2.4888,
+      "step": 11650
+    },
+    {
+      "epoch": 0.9402792349285772,
+      "grad_norm": 0.6683691143989563,
+      "learning_rate": 7.500870376296918e-05,
+      "loss": 2.5233,
+      "step": 11651
+    },
+    {
+      "epoch": 0.9403599386651602,
+      "grad_norm": 0.6537527441978455,
+      "learning_rate": 7.499341812171846e-05,
+      "loss": 2.5061,
+      "step": 11652
+    },
+    {
+      "epoch": 0.9404406424017432,
+      "grad_norm": 0.6657658219337463,
+      "learning_rate": 7.497813310369717e-05,
+      "loss": 2.4844,
+      "step": 11653
+    },
+    {
+      "epoch": 0.9405213461383262,
+      "grad_norm": 0.6865110993385315,
+      "learning_rate": 7.496284870928618e-05,
+      "loss": 2.4986,
+      "step": 11654
+    },
+    {
+      "epoch": 0.9406020498749093,
+      "grad_norm": 0.6724923849105835,
+      "learning_rate": 7.494756493886644e-05,
+      "loss": 2.4818,
+      "step": 11655
+    },
+    {
+      "epoch": 0.9406827536114922,
+      "grad_norm": 0.6478626728057861,
+      "learning_rate": 7.493228179281892e-05,
+      "loss": 2.5321,
+      "step": 11656
+    },
+    {
+      "epoch": 0.9407634573480752,
+      "grad_norm": 0.6474425792694092,
+      "learning_rate": 7.491699927152443e-05,
+      "loss": 2.5276,
+      "step": 11657
+    },
+    {
+      "epoch": 0.9408441610846582,
+      "grad_norm": 0.6736220717430115,
+      "learning_rate": 7.490171737536387e-05,
+      "loss": 2.4734,
+      "step": 11658
+    },
+    {
+      "epoch": 0.9409248648212413,
+      "grad_norm": 0.6714746952056885,
+      "learning_rate": 7.488643610471815e-05,
+      "loss": 2.5754,
+      "step": 11659
+    },
+    {
+      "epoch": 0.9410055685578242,
+      "grad_norm": 0.6714532375335693,
+      "learning_rate": 7.487115545996805e-05,
+      "loss": 2.4855,
+      "step": 11660
+    },
+    {
+      "epoch": 0.9410862722944072,
+      "grad_norm": 0.7601683139801025,
+      "learning_rate": 7.485587544149447e-05,
+      "loss": 2.4887,
+      "step": 11661
+    },
+    {
+      "epoch": 0.9411669760309902,
+      "grad_norm": 0.7655646204948425,
+      "learning_rate": 7.484059604967821e-05,
+      "loss": 2.4904,
+      "step": 11662
+    },
+    {
+      "epoch": 0.9412476797675733,
+      "grad_norm": 0.6841822862625122,
+      "learning_rate": 7.482531728490006e-05,
+      "loss": 2.5272,
+      "step": 11663
+    },
+    {
+      "epoch": 0.9413283835041563,
+      "grad_norm": 0.7683621048927307,
+      "learning_rate": 7.481003914754078e-05,
+      "loss": 2.5218,
+      "step": 11664
+    },
+    {
+      "epoch": 0.9414090872407392,
+      "grad_norm": 0.6597647070884705,
+      "learning_rate": 7.479476163798124e-05,
+      "loss": 2.4925,
+      "step": 11665
+    },
+    {
+      "epoch": 0.9414897909773222,
+      "grad_norm": 0.6573941111564636,
+      "learning_rate": 7.477948475660208e-05,
+      "loss": 2.4854,
+      "step": 11666
+    },
+    {
+      "epoch": 0.9415704947139053,
+      "grad_norm": 0.6639125943183899,
+      "learning_rate": 7.476420850378407e-05,
+      "loss": 2.5207,
+      "step": 11667
+    },
+    {
+      "epoch": 0.9416511984504883,
+      "grad_norm": 0.6770366430282593,
+      "learning_rate": 7.474893287990796e-05,
+      "loss": 2.5167,
+      "step": 11668
+    },
+    {
+      "epoch": 0.9417319021870713,
+      "grad_norm": 0.6908389925956726,
+      "learning_rate": 7.473365788535447e-05,
+      "loss": 2.4606,
+      "step": 11669
+    },
+    {
+      "epoch": 0.9418126059236542,
+      "grad_norm": 0.6625069975852966,
+      "learning_rate": 7.471838352050427e-05,
+      "loss": 2.5344,
+      "step": 11670
+    },
+    {
+      "epoch": 0.9418933096602373,
+      "grad_norm": 0.6690869331359863,
+      "learning_rate": 7.470310978573803e-05,
+      "loss": 2.4507,
+      "step": 11671
+    },
+    {
+      "epoch": 0.9419740133968203,
+      "grad_norm": 0.6741886734962463,
+      "learning_rate": 7.468783668143645e-05,
+      "loss": 2.5755,
+      "step": 11672
+    },
+    {
+      "epoch": 0.9420547171334033,
+      "grad_norm": 0.6876424551010132,
+      "learning_rate": 7.467256420798009e-05,
+      "loss": 2.483,
+      "step": 11673
+    },
+    {
+      "epoch": 0.9421354208699863,
+      "grad_norm": 0.7044318318367004,
+      "learning_rate": 7.465729236574965e-05,
+      "loss": 2.5025,
+      "step": 11674
+    },
+    {
+      "epoch": 0.9422161246065693,
+      "grad_norm": 0.6608660817146301,
+      "learning_rate": 7.46420211551258e-05,
+      "loss": 2.5253,
+      "step": 11675
+    },
+    {
+      "epoch": 0.9422968283431523,
+      "grad_norm": 0.6944260001182556,
+      "learning_rate": 7.4626750576489e-05,
+      "loss": 2.5002,
+      "step": 11676
+    },
+    {
+      "epoch": 0.9423775320797353,
+      "grad_norm": 0.7304964065551758,
+      "learning_rate": 7.46114806302199e-05,
+      "loss": 2.5501,
+      "step": 11677
+    },
+    {
+      "epoch": 0.9424582358163183,
+      "grad_norm": 0.688525378704071,
+      "learning_rate": 7.459621131669911e-05,
+      "loss": 2.5291,
+      "step": 11678
+    },
+    {
+      "epoch": 0.9425389395529012,
+      "grad_norm": 0.7388432025909424,
+      "learning_rate": 7.45809426363071e-05,
+      "loss": 2.5391,
+      "step": 11679
+    },
+    {
+      "epoch": 0.9426196432894843,
+      "grad_norm": 0.6777819991111755,
+      "learning_rate": 7.456567458942447e-05,
+      "loss": 2.5425,
+      "step": 11680
+    },
+    {
+      "epoch": 0.9427003470260673,
+      "grad_norm": 0.7208845615386963,
+      "learning_rate": 7.455040717643169e-05,
+      "loss": 2.5306,
+      "step": 11681
+    },
+    {
+      "epoch": 0.9427810507626503,
+      "grad_norm": 0.745384693145752,
+      "learning_rate": 7.453514039770934e-05,
+      "loss": 2.4695,
+      "step": 11682
+    },
+    {
+      "epoch": 0.9428617544992333,
+      "grad_norm": 0.7088115215301514,
+      "learning_rate": 7.451987425363782e-05,
+      "loss": 2.5413,
+      "step": 11683
+    },
+    {
+      "epoch": 0.9429424582358163,
+      "grad_norm": 0.7287998795509338,
+      "learning_rate": 7.450460874459762e-05,
+      "loss": 2.5773,
+      "step": 11684
+    },
+    {
+      "epoch": 0.9430231619723993,
+      "grad_norm": 0.6897092461585999,
+      "learning_rate": 7.448934387096928e-05,
+      "loss": 2.5255,
+      "step": 11685
+    },
+    {
+      "epoch": 0.9431038657089823,
+      "grad_norm": 0.6227227449417114,
+      "learning_rate": 7.447407963313313e-05,
+      "loss": 2.5027,
+      "step": 11686
+    },
+    {
+      "epoch": 0.9431845694455653,
+      "grad_norm": 0.6954305768013,
+      "learning_rate": 7.445881603146964e-05,
+      "loss": 2.5477,
+      "step": 11687
+    },
+    {
+      "epoch": 0.9432652731821484,
+      "grad_norm": 0.7860052585601807,
+      "learning_rate": 7.444355306635924e-05,
+      "loss": 2.469,
+      "step": 11688
+    },
+    {
+      "epoch": 0.9433459769187313,
+      "grad_norm": 0.6851965188980103,
+      "learning_rate": 7.442829073818227e-05,
+      "loss": 2.4997,
+      "step": 11689
+    },
+    {
+      "epoch": 0.9434266806553143,
+      "grad_norm": 0.7011744379997253,
+      "learning_rate": 7.441302904731916e-05,
+      "loss": 2.5399,
+      "step": 11690
+    },
+    {
+      "epoch": 0.9435073843918973,
+      "grad_norm": 0.703167200088501,
+      "learning_rate": 7.439776799415028e-05,
+      "loss": 2.5323,
+      "step": 11691
+    },
+    {
+      "epoch": 0.9435880881284804,
+      "grad_norm": 0.6747310161590576,
+      "learning_rate": 7.438250757905591e-05,
+      "loss": 2.5406,
+      "step": 11692
+    },
+    {
+      "epoch": 0.9436687918650634,
+      "grad_norm": 0.8631153106689453,
+      "learning_rate": 7.436724780241642e-05,
+      "loss": 2.5215,
+      "step": 11693
+    },
+    {
+      "epoch": 0.9437494956016463,
+      "grad_norm": 0.6919798254966736,
+      "learning_rate": 7.435198866461214e-05,
+      "loss": 2.4654,
+      "step": 11694
+    },
+    {
+      "epoch": 0.9438301993382293,
+      "grad_norm": 0.6747070550918579,
+      "learning_rate": 7.433673016602332e-05,
+      "loss": 2.5186,
+      "step": 11695
+    },
+    {
+      "epoch": 0.9439109030748124,
+      "grad_norm": 0.7368776798248291,
+      "learning_rate": 7.432147230703026e-05,
+      "loss": 2.5365,
+      "step": 11696
+    },
+    {
+      "epoch": 0.9439916068113954,
+      "grad_norm": 0.7443639636039734,
+      "learning_rate": 7.430621508801325e-05,
+      "loss": 2.4966,
+      "step": 11697
+    },
+    {
+      "epoch": 0.9440723105479784,
+      "grad_norm": 0.7371395230293274,
+      "learning_rate": 7.429095850935255e-05,
+      "loss": 2.4638,
+      "step": 11698
+    },
+    {
+      "epoch": 0.9441530142845613,
+      "grad_norm": 0.6917321681976318,
+      "learning_rate": 7.427570257142832e-05,
+      "loss": 2.5341,
+      "step": 11699
+    },
+    {
+      "epoch": 0.9442337180211444,
+      "grad_norm": 0.7704101800918579,
+      "learning_rate": 7.426044727462085e-05,
+      "loss": 2.5144,
+      "step": 11700
+    },
+    {
+      "epoch": 0.9443144217577274,
+      "grad_norm": 0.692197859287262,
+      "learning_rate": 7.424519261931036e-05,
+      "loss": 2.5293,
+      "step": 11701
+    },
+    {
+      "epoch": 0.9443951254943104,
+      "grad_norm": 0.7305885553359985,
+      "learning_rate": 7.422993860587695e-05,
+      "loss": 2.5236,
+      "step": 11702
+    },
+    {
+      "epoch": 0.9444758292308933,
+      "grad_norm": 0.6955052018165588,
+      "learning_rate": 7.421468523470081e-05,
+      "loss": 2.4765,
+      "step": 11703
+    },
+    {
+      "epoch": 0.9445565329674764,
+      "grad_norm": 0.7394432425498962,
+      "learning_rate": 7.419943250616216e-05,
+      "loss": 2.5053,
+      "step": 11704
+    },
+    {
+      "epoch": 0.9446372367040594,
+      "grad_norm": 0.679044246673584,
+      "learning_rate": 7.418418042064108e-05,
+      "loss": 2.5413,
+      "step": 11705
+    },
+    {
+      "epoch": 0.9447179404406424,
+      "grad_norm": 0.7153440117835999,
+      "learning_rate": 7.41689289785177e-05,
+      "loss": 2.4938,
+      "step": 11706
+    },
+    {
+      "epoch": 0.9447986441772254,
+      "grad_norm": 0.697068452835083,
+      "learning_rate": 7.415367818017217e-05,
+      "loss": 2.5157,
+      "step": 11707
+    },
+    {
+      "epoch": 0.9448793479138085,
+      "grad_norm": 0.664616048336029,
+      "learning_rate": 7.41384280259845e-05,
+      "loss": 2.4859,
+      "step": 11708
+    },
+    {
+      "epoch": 0.9449600516503914,
+      "grad_norm": 0.7275365591049194,
+      "learning_rate": 7.412317851633479e-05,
+      "loss": 2.523,
+      "step": 11709
+    },
+    {
+      "epoch": 0.9450407553869744,
+      "grad_norm": 0.7408944368362427,
+      "learning_rate": 7.410792965160318e-05,
+      "loss": 2.4994,
+      "step": 11710
+    },
+    {
+      "epoch": 0.9451214591235574,
+      "grad_norm": 0.7222678065299988,
+      "learning_rate": 7.40926814321696e-05,
+      "loss": 2.5084,
+      "step": 11711
+    },
+    {
+      "epoch": 0.9452021628601405,
+      "grad_norm": 0.7242292761802673,
+      "learning_rate": 7.407743385841412e-05,
+      "loss": 2.5165,
+      "step": 11712
+    },
+    {
+      "epoch": 0.9452828665967234,
+      "grad_norm": 0.6634014844894409,
+      "learning_rate": 7.406218693071677e-05,
+      "loss": 2.4947,
+      "step": 11713
+    },
+    {
+      "epoch": 0.9453635703333064,
+      "grad_norm": 0.8126605153083801,
+      "learning_rate": 7.404694064945751e-05,
+      "loss": 2.5553,
+      "step": 11714
+    },
+    {
+      "epoch": 0.9454442740698894,
+      "grad_norm": 0.679344654083252,
+      "learning_rate": 7.403169501501632e-05,
+      "loss": 2.5475,
+      "step": 11715
+    },
+    {
+      "epoch": 0.9455249778064725,
+      "grad_norm": 0.7584314346313477,
+      "learning_rate": 7.401645002777318e-05,
+      "loss": 2.5498,
+      "step": 11716
+    },
+    {
+      "epoch": 0.9456056815430555,
+      "grad_norm": 0.7191590666770935,
+      "learning_rate": 7.400120568810806e-05,
+      "loss": 2.5161,
+      "step": 11717
+    },
+    {
+      "epoch": 0.9456863852796384,
+      "grad_norm": 0.6738762855529785,
+      "learning_rate": 7.398596199640084e-05,
+      "loss": 2.4819,
+      "step": 11718
+    },
+    {
+      "epoch": 0.9457670890162214,
+      "grad_norm": 0.7305885553359985,
+      "learning_rate": 7.397071895303143e-05,
+      "loss": 2.4842,
+      "step": 11719
+    },
+    {
+      "epoch": 0.9458477927528045,
+      "grad_norm": 0.6885291337966919,
+      "learning_rate": 7.395547655837976e-05,
+      "loss": 2.5016,
+      "step": 11720
+    },
+    {
+      "epoch": 0.9459284964893875,
+      "grad_norm": 0.6807307600975037,
+      "learning_rate": 7.394023481282568e-05,
+      "loss": 2.4949,
+      "step": 11721
+    },
+    {
+      "epoch": 0.9460092002259705,
+      "grad_norm": 0.6683849096298218,
+      "learning_rate": 7.392499371674907e-05,
+      "loss": 2.4974,
+      "step": 11722
+    },
+    {
+      "epoch": 0.9460899039625534,
+      "grad_norm": 0.6615697741508484,
+      "learning_rate": 7.39097532705298e-05,
+      "loss": 2.4744,
+      "step": 11723
+    },
+    {
+      "epoch": 0.9461706076991365,
+      "grad_norm": 0.6463690996170044,
+      "learning_rate": 7.389451347454765e-05,
+      "loss": 2.478,
+      "step": 11724
+    },
+    {
+      "epoch": 0.9462513114357195,
+      "grad_norm": 0.6848269701004028,
+      "learning_rate": 7.387927432918247e-05,
+      "loss": 2.5491,
+      "step": 11725
+    },
+    {
+      "epoch": 0.9463320151723025,
+      "grad_norm": 0.7251551747322083,
+      "learning_rate": 7.386403583481409e-05,
+      "loss": 2.4936,
+      "step": 11726
+    },
+    {
+      "epoch": 0.9464127189088855,
+      "grad_norm": 0.6562095284461975,
+      "learning_rate": 7.384879799182223e-05,
+      "loss": 2.4895,
+      "step": 11727
+    },
+    {
+      "epoch": 0.9464934226454685,
+      "grad_norm": 0.6891352534294128,
+      "learning_rate": 7.383356080058668e-05,
+      "loss": 2.508,
+      "step": 11728
+    },
+    {
+      "epoch": 0.9465741263820515,
+      "grad_norm": 0.7220255136489868,
+      "learning_rate": 7.381832426148719e-05,
+      "loss": 2.5181,
+      "step": 11729
+    },
+    {
+      "epoch": 0.9466548301186345,
+      "grad_norm": 0.7213689088821411,
+      "learning_rate": 7.38030883749035e-05,
+      "loss": 2.5136,
+      "step": 11730
+    },
+    {
+      "epoch": 0.9467355338552175,
+      "grad_norm": 0.6711129546165466,
+      "learning_rate": 7.378785314121535e-05,
+      "loss": 2.5463,
+      "step": 11731
+    },
+    {
+      "epoch": 0.9468162375918004,
+      "grad_norm": 0.6380139589309692,
+      "learning_rate": 7.377261856080239e-05,
+      "loss": 2.5092,
+      "step": 11732
+    },
+    {
+      "epoch": 0.9468969413283835,
+      "grad_norm": 0.66046142578125,
+      "learning_rate": 7.375738463404437e-05,
+      "loss": 2.5561,
+      "step": 11733
+    },
+    {
+      "epoch": 0.9469776450649665,
+      "grad_norm": 0.6857354044914246,
+      "learning_rate": 7.37421513613209e-05,
+      "loss": 2.5774,
+      "step": 11734
+    },
+    {
+      "epoch": 0.9470583488015495,
+      "grad_norm": 0.6811589598655701,
+      "learning_rate": 7.372691874301163e-05,
+      "loss": 2.4918,
+      "step": 11735
+    },
+    {
+      "epoch": 0.9471390525381325,
+      "grad_norm": 0.6401017308235168,
+      "learning_rate": 7.37116867794963e-05,
+      "loss": 2.4994,
+      "step": 11736
+    },
+    {
+      "epoch": 0.9472197562747156,
+      "grad_norm": 0.6967078447341919,
+      "learning_rate": 7.369645547115438e-05,
+      "loss": 2.5809,
+      "step": 11737
+    },
+    {
+      "epoch": 0.9473004600112985,
+      "grad_norm": 0.6695219278335571,
+      "learning_rate": 7.368122481836557e-05,
+      "loss": 2.4735,
+      "step": 11738
+    },
+    {
+      "epoch": 0.9473811637478815,
+      "grad_norm": 0.6540528535842896,
+      "learning_rate": 7.366599482150944e-05,
+      "loss": 2.4998,
+      "step": 11739
+    },
+    {
+      "epoch": 0.9474618674844645,
+      "grad_norm": 0.700683057308197,
+      "learning_rate": 7.365076548096556e-05,
+      "loss": 2.5258,
+      "step": 11740
+    },
+    {
+      "epoch": 0.9475425712210476,
+      "grad_norm": 0.7125419974327087,
+      "learning_rate": 7.363553679711347e-05,
+      "loss": 2.4653,
+      "step": 11741
+    },
+    {
+      "epoch": 0.9476232749576305,
+      "grad_norm": 0.7285346984863281,
+      "learning_rate": 7.362030877033275e-05,
+      "loss": 2.5523,
+      "step": 11742
+    },
+    {
+      "epoch": 0.9477039786942135,
+      "grad_norm": 0.7310814261436462,
+      "learning_rate": 7.360508140100288e-05,
+      "loss": 2.5027,
+      "step": 11743
+    },
+    {
+      "epoch": 0.9477846824307965,
+      "grad_norm": 0.746961772441864,
+      "learning_rate": 7.358985468950335e-05,
+      "loss": 2.5485,
+      "step": 11744
+    },
+    {
+      "epoch": 0.9478653861673796,
+      "grad_norm": 0.6880186796188354,
+      "learning_rate": 7.357462863621369e-05,
+      "loss": 2.5243,
+      "step": 11745
+    },
+    {
+      "epoch": 0.9479460899039626,
+      "grad_norm": 0.6406471133232117,
+      "learning_rate": 7.355940324151339e-05,
+      "loss": 2.512,
+      "step": 11746
+    },
+    {
+      "epoch": 0.9480267936405455,
+      "grad_norm": 0.6503005027770996,
+      "learning_rate": 7.354417850578184e-05,
+      "loss": 2.5318,
+      "step": 11747
+    },
+    {
+      "epoch": 0.9481074973771285,
+      "grad_norm": 0.6458879113197327,
+      "learning_rate": 7.352895442939852e-05,
+      "loss": 2.5451,
+      "step": 11748
+    },
+    {
+      "epoch": 0.9481882011137116,
+      "grad_norm": 0.7382936477661133,
+      "learning_rate": 7.351373101274288e-05,
+      "loss": 2.5393,
+      "step": 11749
+    },
+    {
+      "epoch": 0.9482689048502946,
+      "grad_norm": 0.7366087436676025,
+      "learning_rate": 7.349850825619429e-05,
+      "loss": 2.5591,
+      "step": 11750
+    },
+    {
+      "epoch": 0.9483496085868776,
+      "grad_norm": 0.6652588248252869,
+      "learning_rate": 7.348328616013213e-05,
+      "loss": 2.5348,
+      "step": 11751
+    },
+    {
+      "epoch": 0.9484303123234605,
+      "grad_norm": 0.7515435814857483,
+      "learning_rate": 7.346806472493584e-05,
+      "loss": 2.5208,
+      "step": 11752
+    },
+    {
+      "epoch": 0.9485110160600436,
+      "grad_norm": 0.7161263227462769,
+      "learning_rate": 7.345284395098469e-05,
+      "loss": 2.5518,
+      "step": 11753
+    },
+    {
+      "epoch": 0.9485917197966266,
+      "grad_norm": 0.7433953285217285,
+      "learning_rate": 7.343762383865807e-05,
+      "loss": 2.5914,
+      "step": 11754
+    },
+    {
+      "epoch": 0.9486724235332096,
+      "grad_norm": 0.674991250038147,
+      "learning_rate": 7.342240438833532e-05,
+      "loss": 2.5566,
+      "step": 11755
+    },
+    {
+      "epoch": 0.9487531272697926,
+      "grad_norm": 0.7511670589447021,
+      "learning_rate": 7.34071856003957e-05,
+      "loss": 2.5253,
+      "step": 11756
+    },
+    {
+      "epoch": 0.9488338310063756,
+      "grad_norm": 0.6672492623329163,
+      "learning_rate": 7.339196747521853e-05,
+      "loss": 2.4887,
+      "step": 11757
+    },
+    {
+      "epoch": 0.9489145347429586,
+      "grad_norm": 0.6826158761978149,
+      "learning_rate": 7.337675001318312e-05,
+      "loss": 2.5072,
+      "step": 11758
+    },
+    {
+      "epoch": 0.9489952384795416,
+      "grad_norm": 0.7189450860023499,
+      "learning_rate": 7.336153321466867e-05,
+      "loss": 2.5583,
+      "step": 11759
+    },
+    {
+      "epoch": 0.9490759422161246,
+      "grad_norm": 0.6923015117645264,
+      "learning_rate": 7.33463170800544e-05,
+      "loss": 2.5416,
+      "step": 11760
+    },
+    {
+      "epoch": 0.9491566459527077,
+      "grad_norm": 0.690060555934906,
+      "learning_rate": 7.333110160971963e-05,
+      "loss": 2.4931,
+      "step": 11761
+    },
+    {
+      "epoch": 0.9492373496892906,
+      "grad_norm": 0.6887977719306946,
+      "learning_rate": 7.331588680404354e-05,
+      "loss": 2.4676,
+      "step": 11762
+    },
+    {
+      "epoch": 0.9493180534258736,
+      "grad_norm": 0.8573753237724304,
+      "learning_rate": 7.330067266340528e-05,
+      "loss": 2.5074,
+      "step": 11763
+    },
+    {
+      "epoch": 0.9493987571624566,
+      "grad_norm": 0.6760974526405334,
+      "learning_rate": 7.328545918818403e-05,
+      "loss": 2.5395,
+      "step": 11764
+    },
+    {
+      "epoch": 0.9494794608990397,
+      "grad_norm": 0.6946160197257996,
+      "learning_rate": 7.327024637875901e-05,
+      "loss": 2.535,
+      "step": 11765
+    },
+    {
+      "epoch": 0.9495601646356226,
+      "grad_norm": 0.6851378679275513,
+      "learning_rate": 7.32550342355093e-05,
+      "loss": 2.487,
+      "step": 11766
+    },
+    {
+      "epoch": 0.9496408683722056,
+      "grad_norm": 0.6480168104171753,
+      "learning_rate": 7.323982275881404e-05,
+      "loss": 2.513,
+      "step": 11767
+    },
+    {
+      "epoch": 0.9497215721087886,
+      "grad_norm": 0.6492218971252441,
+      "learning_rate": 7.322461194905239e-05,
+      "loss": 2.4532,
+      "step": 11768
+    },
+    {
+      "epoch": 0.9498022758453717,
+      "grad_norm": 0.6670051217079163,
+      "learning_rate": 7.320940180660337e-05,
+      "loss": 2.5258,
+      "step": 11769
+    },
+    {
+      "epoch": 0.9498829795819547,
+      "grad_norm": 0.6678066253662109,
+      "learning_rate": 7.319419233184608e-05,
+      "loss": 2.5388,
+      "step": 11770
+    },
+    {
+      "epoch": 0.9499636833185376,
+      "grad_norm": 0.693545937538147,
+      "learning_rate": 7.31789835251596e-05,
+      "loss": 2.5304,
+      "step": 11771
+    },
+    {
+      "epoch": 0.9500443870551206,
+      "grad_norm": 0.680486798286438,
+      "learning_rate": 7.316377538692297e-05,
+      "loss": 2.5024,
+      "step": 11772
+    },
+    {
+      "epoch": 0.9501250907917037,
+      "grad_norm": 0.7271847128868103,
+      "learning_rate": 7.314856791751518e-05,
+      "loss": 2.5947,
+      "step": 11773
+    },
+    {
+      "epoch": 0.9502057945282867,
+      "grad_norm": 0.6889839172363281,
+      "learning_rate": 7.31333611173153e-05,
+      "loss": 2.5135,
+      "step": 11774
+    },
+    {
+      "epoch": 0.9502864982648697,
+      "grad_norm": 0.7431777119636536,
+      "learning_rate": 7.311815498670226e-05,
+      "loss": 2.5856,
+      "step": 11775
+    },
+    {
+      "epoch": 0.9503672020014526,
+      "grad_norm": 0.7168101072311401,
+      "learning_rate": 7.310294952605508e-05,
+      "loss": 2.4383,
+      "step": 11776
+    },
+    {
+      "epoch": 0.9504479057380357,
+      "grad_norm": 0.654803454875946,
+      "learning_rate": 7.308774473575271e-05,
+      "loss": 2.4908,
+      "step": 11777
+    },
+    {
+      "epoch": 0.9505286094746187,
+      "grad_norm": 0.6810718774795532,
+      "learning_rate": 7.307254061617412e-05,
+      "loss": 2.5073,
+      "step": 11778
+    },
+    {
+      "epoch": 0.9506093132112017,
+      "grad_norm": 0.637980043888092,
+      "learning_rate": 7.305733716769817e-05,
+      "loss": 2.5686,
+      "step": 11779
+    },
+    {
+      "epoch": 0.9506900169477847,
+      "grad_norm": 0.6549471020698547,
+      "learning_rate": 7.30421343907038e-05,
+      "loss": 2.5502,
+      "step": 11780
+    },
+    {
+      "epoch": 0.9507707206843676,
+      "grad_norm": 0.7087163329124451,
+      "learning_rate": 7.302693228556994e-05,
+      "loss": 2.4773,
+      "step": 11781
+    },
+    {
+      "epoch": 0.9508514244209507,
+      "grad_norm": 0.6230717897415161,
+      "learning_rate": 7.301173085267541e-05,
+      "loss": 2.4806,
+      "step": 11782
+    },
+    {
+      "epoch": 0.9509321281575337,
+      "grad_norm": 0.7145688533782959,
+      "learning_rate": 7.299653009239911e-05,
+      "loss": 2.5259,
+      "step": 11783
+    },
+    {
+      "epoch": 0.9510128318941167,
+      "grad_norm": 0.679100513458252,
+      "learning_rate": 7.298133000511988e-05,
+      "loss": 2.5012,
+      "step": 11784
+    },
+    {
+      "epoch": 0.9510935356306996,
+      "grad_norm": 0.7057691216468811,
+      "learning_rate": 7.29661305912165e-05,
+      "loss": 2.4826,
+      "step": 11785
+    },
+    {
+      "epoch": 0.9511742393672827,
+      "grad_norm": 0.65343177318573,
+      "learning_rate": 7.295093185106782e-05,
+      "loss": 2.4553,
+      "step": 11786
+    },
+    {
+      "epoch": 0.9512549431038657,
+      "grad_norm": 0.7948461174964905,
+      "learning_rate": 7.293573378505268e-05,
+      "loss": 2.478,
+      "step": 11787
+    },
+    {
+      "epoch": 0.9513356468404487,
+      "grad_norm": 0.6511468887329102,
+      "learning_rate": 7.292053639354975e-05,
+      "loss": 2.4862,
+      "step": 11788
+    },
+    {
+      "epoch": 0.9514163505770317,
+      "grad_norm": 0.7293919324874878,
+      "learning_rate": 7.290533967693782e-05,
+      "loss": 2.5956,
+      "step": 11789
+    },
+    {
+      "epoch": 0.9514970543136148,
+      "grad_norm": 0.6691277623176575,
+      "learning_rate": 7.289014363559567e-05,
+      "loss": 2.5659,
+      "step": 11790
+    },
+    {
+      "epoch": 0.9515777580501977,
+      "grad_norm": 0.7054625749588013,
+      "learning_rate": 7.287494826990203e-05,
+      "loss": 2.5875,
+      "step": 11791
+    },
+    {
+      "epoch": 0.9516584617867807,
+      "grad_norm": 0.6597220301628113,
+      "learning_rate": 7.285975358023555e-05,
+      "loss": 2.5215,
+      "step": 11792
+    },
+    {
+      "epoch": 0.9517391655233637,
+      "grad_norm": 0.6719489097595215,
+      "learning_rate": 7.284455956697497e-05,
+      "loss": 2.4752,
+      "step": 11793
+    },
+    {
+      "epoch": 0.9518198692599468,
+      "grad_norm": 0.7325637340545654,
+      "learning_rate": 7.2829366230499e-05,
+      "loss": 2.5504,
+      "step": 11794
+    },
+    {
+      "epoch": 0.9519005729965297,
+      "grad_norm": 0.637668788433075,
+      "learning_rate": 7.281417357118619e-05,
+      "loss": 2.5105,
+      "step": 11795
+    },
+    {
+      "epoch": 0.9519812767331127,
+      "grad_norm": 0.7815340161323547,
+      "learning_rate": 7.279898158941525e-05,
+      "loss": 2.4998,
+      "step": 11796
+    },
+    {
+      "epoch": 0.9520619804696957,
+      "grad_norm": 0.6555821299552917,
+      "learning_rate": 7.278379028556481e-05,
+      "loss": 2.4326,
+      "step": 11797
+    },
+    {
+      "epoch": 0.9521426842062788,
+      "grad_norm": 0.7298933863639832,
+      "learning_rate": 7.276859966001344e-05,
+      "loss": 2.4779,
+      "step": 11798
+    },
+    {
+      "epoch": 0.9522233879428618,
+      "grad_norm": 0.683455765247345,
+      "learning_rate": 7.275340971313974e-05,
+      "loss": 2.4416,
+      "step": 11799
+    },
+    {
+      "epoch": 0.9523040916794447,
+      "grad_norm": 0.6353151798248291,
+      "learning_rate": 7.273822044532232e-05,
+      "loss": 2.4777,
+      "step": 11800
+    },
+    {
+      "epoch": 0.9523847954160277,
+      "grad_norm": 0.6898894309997559,
+      "learning_rate": 7.27230318569397e-05,
+      "loss": 2.5351,
+      "step": 11801
+    },
+    {
+      "epoch": 0.9524654991526108,
+      "grad_norm": 0.6528690457344055,
+      "learning_rate": 7.270784394837041e-05,
+      "loss": 2.5145,
+      "step": 11802
+    },
+    {
+      "epoch": 0.9525462028891938,
+      "grad_norm": 0.6432619094848633,
+      "learning_rate": 7.269265671999304e-05,
+      "loss": 2.5002,
+      "step": 11803
+    },
+    {
+      "epoch": 0.9526269066257768,
+      "grad_norm": 0.7317861318588257,
+      "learning_rate": 7.267747017218601e-05,
+      "loss": 2.5318,
+      "step": 11804
+    },
+    {
+      "epoch": 0.9527076103623597,
+      "grad_norm": 0.7581185698509216,
+      "learning_rate": 7.266228430532785e-05,
+      "loss": 2.5313,
+      "step": 11805
+    },
+    {
+      "epoch": 0.9527883140989428,
+      "grad_norm": 0.7316486239433289,
+      "learning_rate": 7.264709911979702e-05,
+      "loss": 2.5147,
+      "step": 11806
+    },
+    {
+      "epoch": 0.9528690178355258,
+      "grad_norm": 0.7378978729248047,
+      "learning_rate": 7.263191461597199e-05,
+      "loss": 2.5149,
+      "step": 11807
+    },
+    {
+      "epoch": 0.9529497215721088,
+      "grad_norm": 0.6603738069534302,
+      "learning_rate": 7.26167307942312e-05,
+      "loss": 2.4684,
+      "step": 11808
+    },
+    {
+      "epoch": 0.9530304253086918,
+      "grad_norm": 0.7566502690315247,
+      "learning_rate": 7.260154765495302e-05,
+      "loss": 2.5535,
+      "step": 11809
+    },
+    {
+      "epoch": 0.9531111290452748,
+      "grad_norm": 0.693067729473114,
+      "learning_rate": 7.258636519851596e-05,
+      "loss": 2.5103,
+      "step": 11810
+    },
+    {
+      "epoch": 0.9531918327818578,
+      "grad_norm": 0.7049208283424377,
+      "learning_rate": 7.257118342529826e-05,
+      "loss": 2.5482,
+      "step": 11811
+    },
+    {
+      "epoch": 0.9532725365184408,
+      "grad_norm": 0.6986998319625854,
+      "learning_rate": 7.25560023356784e-05,
+      "loss": 2.4921,
+      "step": 11812
+    },
+    {
+      "epoch": 0.9533532402550238,
+      "grad_norm": 0.7079482674598694,
+      "learning_rate": 7.254082193003476e-05,
+      "loss": 2.5339,
+      "step": 11813
+    },
+    {
+      "epoch": 0.9534339439916069,
+      "grad_norm": 0.7283922433853149,
+      "learning_rate": 7.252564220874553e-05,
+      "loss": 2.5056,
+      "step": 11814
+    },
+    {
+      "epoch": 0.9535146477281898,
+      "grad_norm": 0.6965533494949341,
+      "learning_rate": 7.251046317218914e-05,
+      "loss": 2.5512,
+      "step": 11815
+    },
+    {
+      "epoch": 0.9535953514647728,
+      "grad_norm": 0.7367159128189087,
+      "learning_rate": 7.24952848207439e-05,
+      "loss": 2.5015,
+      "step": 11816
+    },
+    {
+      "epoch": 0.9536760552013558,
+      "grad_norm": 0.6959818601608276,
+      "learning_rate": 7.248010715478802e-05,
+      "loss": 2.4969,
+      "step": 11817
+    },
+    {
+      "epoch": 0.9537567589379389,
+      "grad_norm": 0.69304358959198,
+      "learning_rate": 7.246493017469981e-05,
+      "loss": 2.5098,
+      "step": 11818
+    },
+    {
+      "epoch": 0.9538374626745219,
+      "grad_norm": 0.6830596327781677,
+      "learning_rate": 7.244975388085757e-05,
+      "loss": 2.5206,
+      "step": 11819
+    },
+    {
+      "epoch": 0.9539181664111048,
+      "grad_norm": 0.7354303598403931,
+      "learning_rate": 7.243457827363944e-05,
+      "loss": 2.5223,
+      "step": 11820
+    },
+    {
+      "epoch": 0.9539988701476878,
+      "grad_norm": 0.7046182751655579,
+      "learning_rate": 7.241940335342366e-05,
+      "loss": 2.4931,
+      "step": 11821
+    },
+    {
+      "epoch": 0.9540795738842709,
+      "grad_norm": 0.6990540623664856,
+      "learning_rate": 7.240422912058843e-05,
+      "loss": 2.4302,
+      "step": 11822
+    },
+    {
+      "epoch": 0.9541602776208539,
+      "grad_norm": 0.7562115788459778,
+      "learning_rate": 7.238905557551202e-05,
+      "loss": 2.5118,
+      "step": 11823
+    },
+    {
+      "epoch": 0.9542409813574368,
+      "grad_norm": 0.8212862014770508,
+      "learning_rate": 7.237388271857248e-05,
+      "loss": 2.5476,
+      "step": 11824
+    },
+    {
+      "epoch": 0.9543216850940198,
+      "grad_norm": 0.7095397710800171,
+      "learning_rate": 7.235871055014798e-05,
+      "loss": 2.5073,
+      "step": 11825
+    },
+    {
+      "epoch": 0.9544023888306029,
+      "grad_norm": 0.7174660563468933,
+      "learning_rate": 7.23435390706167e-05,
+      "loss": 2.4553,
+      "step": 11826
+    },
+    {
+      "epoch": 0.9544830925671859,
+      "grad_norm": 0.7121314406394958,
+      "learning_rate": 7.23283682803567e-05,
+      "loss": 2.5164,
+      "step": 11827
+    },
+    {
+      "epoch": 0.9545637963037689,
+      "grad_norm": 0.7354126572608948,
+      "learning_rate": 7.231319817974609e-05,
+      "loss": 2.5413,
+      "step": 11828
+    },
+    {
+      "epoch": 0.9546445000403518,
+      "grad_norm": 0.7770543694496155,
+      "learning_rate": 7.2298028769163e-05,
+      "loss": 2.5244,
+      "step": 11829
+    },
+    {
+      "epoch": 0.9547252037769349,
+      "grad_norm": 0.6770393252372742,
+      "learning_rate": 7.228286004898541e-05,
+      "loss": 2.4707,
+      "step": 11830
+    },
+    {
+      "epoch": 0.9548059075135179,
+      "grad_norm": 0.6916880011558533,
+      "learning_rate": 7.22676920195914e-05,
+      "loss": 2.506,
+      "step": 11831
+    },
+    {
+      "epoch": 0.9548866112501009,
+      "grad_norm": 0.6299161314964294,
+      "learning_rate": 7.225252468135901e-05,
+      "loss": 2.5042,
+      "step": 11832
+    },
+    {
+      "epoch": 0.9549673149866839,
+      "grad_norm": 0.7081227898597717,
+      "learning_rate": 7.223735803466623e-05,
+      "loss": 2.5537,
+      "step": 11833
+    },
+    {
+      "epoch": 0.9550480187232668,
+      "grad_norm": 0.6600900888442993,
+      "learning_rate": 7.222219207989104e-05,
+      "loss": 2.5329,
+      "step": 11834
+    },
+    {
+      "epoch": 0.9551287224598499,
+      "grad_norm": 0.6715366244316101,
+      "learning_rate": 7.22070268174115e-05,
+      "loss": 2.5273,
+      "step": 11835
+    },
+    {
+      "epoch": 0.9552094261964329,
+      "grad_norm": 0.6655930280685425,
+      "learning_rate": 7.219186224760543e-05,
+      "loss": 2.4254,
+      "step": 11836
+    },
+    {
+      "epoch": 0.9552901299330159,
+      "grad_norm": 0.6925715208053589,
+      "learning_rate": 7.217669837085088e-05,
+      "loss": 2.5104,
+      "step": 11837
+    },
+    {
+      "epoch": 0.9553708336695989,
+      "grad_norm": 0.7132978439331055,
+      "learning_rate": 7.216153518752571e-05,
+      "loss": 2.5238,
+      "step": 11838
+    },
+    {
+      "epoch": 0.9554515374061819,
+      "grad_norm": 0.661651611328125,
+      "learning_rate": 7.214637269800791e-05,
+      "loss": 2.445,
+      "step": 11839
+    },
+    {
+      "epoch": 0.9555322411427649,
+      "grad_norm": 0.6635430455207825,
+      "learning_rate": 7.213121090267528e-05,
+      "loss": 2.4707,
+      "step": 11840
+    },
+    {
+      "epoch": 0.9556129448793479,
+      "grad_norm": 0.6303616166114807,
+      "learning_rate": 7.211604980190571e-05,
+      "loss": 2.4923,
+      "step": 11841
+    },
+    {
+      "epoch": 0.9556936486159309,
+      "grad_norm": 0.7027459144592285,
+      "learning_rate": 7.210088939607708e-05,
+      "loss": 2.5592,
+      "step": 11842
+    },
+    {
+      "epoch": 0.955774352352514,
+      "grad_norm": 0.6539996862411499,
+      "learning_rate": 7.208572968556722e-05,
+      "loss": 2.5256,
+      "step": 11843
+    },
+    {
+      "epoch": 0.9558550560890969,
+      "grad_norm": 0.7019872069358826,
+      "learning_rate": 7.207057067075393e-05,
+      "loss": 2.488,
+      "step": 11844
+    },
+    {
+      "epoch": 0.9559357598256799,
+      "grad_norm": 0.6848211288452148,
+      "learning_rate": 7.205541235201507e-05,
+      "loss": 2.4883,
+      "step": 11845
+    },
+    {
+      "epoch": 0.9560164635622629,
+      "grad_norm": 0.7806351184844971,
+      "learning_rate": 7.204025472972834e-05,
+      "loss": 2.5563,
+      "step": 11846
+    },
+    {
+      "epoch": 0.956097167298846,
+      "grad_norm": 0.7327724695205688,
+      "learning_rate": 7.202509780427156e-05,
+      "loss": 2.5275,
+      "step": 11847
+    },
+    {
+      "epoch": 0.956177871035429,
+      "grad_norm": 0.6805681586265564,
+      "learning_rate": 7.200994157602248e-05,
+      "loss": 2.4723,
+      "step": 11848
+    },
+    {
+      "epoch": 0.9562585747720119,
+      "grad_norm": 0.7053409814834595,
+      "learning_rate": 7.19947860453588e-05,
+      "loss": 2.4471,
+      "step": 11849
+    },
+    {
+      "epoch": 0.9563392785085949,
+      "grad_norm": 0.6783127188682556,
+      "learning_rate": 7.197963121265826e-05,
+      "loss": 2.4586,
+      "step": 11850
+    },
+    {
+      "epoch": 0.956419982245178,
+      "grad_norm": 0.6639916300773621,
+      "learning_rate": 7.196447707829857e-05,
+      "loss": 2.4966,
+      "step": 11851
+    },
+    {
+      "epoch": 0.956500685981761,
+      "grad_norm": 0.684066891670227,
+      "learning_rate": 7.194932364265739e-05,
+      "loss": 2.5676,
+      "step": 11852
+    },
+    {
+      "epoch": 0.9565813897183439,
+      "grad_norm": 0.7872990965843201,
+      "learning_rate": 7.193417090611239e-05,
+      "loss": 2.5101,
+      "step": 11853
+    },
+    {
+      "epoch": 0.9566620934549269,
+      "grad_norm": 0.7543401122093201,
+      "learning_rate": 7.19190188690412e-05,
+      "loss": 2.5503,
+      "step": 11854
+    },
+    {
+      "epoch": 0.95674279719151,
+      "grad_norm": 0.6514382362365723,
+      "learning_rate": 7.190386753182152e-05,
+      "loss": 2.4902,
+      "step": 11855
+    },
+    {
+      "epoch": 0.956823500928093,
+      "grad_norm": 0.6867108345031738,
+      "learning_rate": 7.188871689483087e-05,
+      "loss": 2.5054,
+      "step": 11856
+    },
+    {
+      "epoch": 0.956904204664676,
+      "grad_norm": 0.6536040306091309,
+      "learning_rate": 7.187356695844687e-05,
+      "loss": 2.5462,
+      "step": 11857
+    },
+    {
+      "epoch": 0.9569849084012589,
+      "grad_norm": 0.690237820148468,
+      "learning_rate": 7.185841772304711e-05,
+      "loss": 2.5673,
+      "step": 11858
+    },
+    {
+      "epoch": 0.957065612137842,
+      "grad_norm": 0.6699091196060181,
+      "learning_rate": 7.184326918900915e-05,
+      "loss": 2.4733,
+      "step": 11859
+    },
+    {
+      "epoch": 0.957146315874425,
+      "grad_norm": 0.6482241153717041,
+      "learning_rate": 7.18281213567105e-05,
+      "loss": 2.4897,
+      "step": 11860
+    },
+    {
+      "epoch": 0.957227019611008,
+      "grad_norm": 0.686130166053772,
+      "learning_rate": 7.181297422652874e-05,
+      "loss": 2.4596,
+      "step": 11861
+    },
+    {
+      "epoch": 0.957307723347591,
+      "grad_norm": 0.6507205367088318,
+      "learning_rate": 7.179782779884132e-05,
+      "loss": 2.5527,
+      "step": 11862
+    },
+    {
+      "epoch": 0.957388427084174,
+      "grad_norm": 0.6578813195228577,
+      "learning_rate": 7.178268207402577e-05,
+      "loss": 2.4975,
+      "step": 11863
+    },
+    {
+      "epoch": 0.957469130820757,
+      "grad_norm": 0.6931977272033691,
+      "learning_rate": 7.176753705245956e-05,
+      "loss": 2.5533,
+      "step": 11864
+    },
+    {
+      "epoch": 0.95754983455734,
+      "grad_norm": 0.7306256890296936,
+      "learning_rate": 7.17523927345201e-05,
+      "loss": 2.534,
+      "step": 11865
+    },
+    {
+      "epoch": 0.957630538293923,
+      "grad_norm": 0.6337448358535767,
+      "learning_rate": 7.173724912058483e-05,
+      "loss": 2.5015,
+      "step": 11866
+    },
+    {
+      "epoch": 0.9577112420305061,
+      "grad_norm": 0.6561456322669983,
+      "learning_rate": 7.172210621103124e-05,
+      "loss": 2.4946,
+      "step": 11867
+    },
+    {
+      "epoch": 0.957791945767089,
+      "grad_norm": 0.6341130137443542,
+      "learning_rate": 7.170696400623666e-05,
+      "loss": 2.5611,
+      "step": 11868
+    },
+    {
+      "epoch": 0.957872649503672,
+      "grad_norm": 0.7202804088592529,
+      "learning_rate": 7.169182250657849e-05,
+      "loss": 2.5209,
+      "step": 11869
+    },
+    {
+      "epoch": 0.957953353240255,
+      "grad_norm": 0.6620556712150574,
+      "learning_rate": 7.167668171243408e-05,
+      "loss": 2.4895,
+      "step": 11870
+    },
+    {
+      "epoch": 0.9580340569768381,
+      "grad_norm": 0.6842508912086487,
+      "learning_rate": 7.166154162418087e-05,
+      "loss": 2.4417,
+      "step": 11871
+    },
+    {
+      "epoch": 0.958114760713421,
+      "grad_norm": 0.7539907693862915,
+      "learning_rate": 7.164640224219608e-05,
+      "loss": 2.5153,
+      "step": 11872
+    },
+    {
+      "epoch": 0.958195464450004,
+      "grad_norm": 0.6524286270141602,
+      "learning_rate": 7.163126356685703e-05,
+      "loss": 2.509,
+      "step": 11873
+    },
+    {
+      "epoch": 0.958276168186587,
+      "grad_norm": 0.7022691965103149,
+      "learning_rate": 7.16161255985411e-05,
+      "loss": 2.5223,
+      "step": 11874
+    },
+    {
+      "epoch": 0.9583568719231701,
+      "grad_norm": 0.6659076809883118,
+      "learning_rate": 7.160098833762549e-05,
+      "loss": 2.5231,
+      "step": 11875
+    },
+    {
+      "epoch": 0.9584375756597531,
+      "grad_norm": 0.6756494641304016,
+      "learning_rate": 7.15858517844875e-05,
+      "loss": 2.5017,
+      "step": 11876
+    },
+    {
+      "epoch": 0.958518279396336,
+      "grad_norm": 0.729850709438324,
+      "learning_rate": 7.157071593950436e-05,
+      "loss": 2.4583,
+      "step": 11877
+    },
+    {
+      "epoch": 0.958598983132919,
+      "grad_norm": 0.7155230641365051,
+      "learning_rate": 7.155558080305326e-05,
+      "loss": 2.4753,
+      "step": 11878
+    },
+    {
+      "epoch": 0.9586796868695021,
+      "grad_norm": 0.6553284525871277,
+      "learning_rate": 7.154044637551147e-05,
+      "loss": 2.5093,
+      "step": 11879
+    },
+    {
+      "epoch": 0.9587603906060851,
+      "grad_norm": 0.6516379117965698,
+      "learning_rate": 7.152531265725617e-05,
+      "loss": 2.4996,
+      "step": 11880
+    },
+    {
+      "epoch": 0.9588410943426681,
+      "grad_norm": 0.6871184706687927,
+      "learning_rate": 7.151017964866449e-05,
+      "loss": 2.5322,
+      "step": 11881
+    },
+    {
+      "epoch": 0.958921798079251,
+      "grad_norm": 0.6998933553695679,
+      "learning_rate": 7.149504735011358e-05,
+      "loss": 2.5328,
+      "step": 11882
+    },
+    {
+      "epoch": 0.959002501815834,
+      "grad_norm": 0.7065120935440063,
+      "learning_rate": 7.147991576198065e-05,
+      "loss": 2.5251,
+      "step": 11883
+    },
+    {
+      "epoch": 0.9590832055524171,
+      "grad_norm": 0.6718337535858154,
+      "learning_rate": 7.146478488464275e-05,
+      "loss": 2.5596,
+      "step": 11884
+    },
+    {
+      "epoch": 0.9591639092890001,
+      "grad_norm": 0.6394883990287781,
+      "learning_rate": 7.144965471847698e-05,
+      "loss": 2.5022,
+      "step": 11885
+    },
+    {
+      "epoch": 0.9592446130255831,
+      "grad_norm": 0.6867207288742065,
+      "learning_rate": 7.143452526386045e-05,
+      "loss": 2.4927,
+      "step": 11886
+    },
+    {
+      "epoch": 0.959325316762166,
+      "grad_norm": 0.6710157990455627,
+      "learning_rate": 7.141939652117026e-05,
+      "loss": 2.5127,
+      "step": 11887
+    },
+    {
+      "epoch": 0.9594060204987491,
+      "grad_norm": 0.6286540627479553,
+      "learning_rate": 7.14042684907834e-05,
+      "loss": 2.4966,
+      "step": 11888
+    },
+    {
+      "epoch": 0.9594867242353321,
+      "grad_norm": 0.7295787334442139,
+      "learning_rate": 7.13891411730769e-05,
+      "loss": 2.5127,
+      "step": 11889
+    },
+    {
+      "epoch": 0.9595674279719151,
+      "grad_norm": 0.646084189414978,
+      "learning_rate": 7.137401456842784e-05,
+      "loss": 2.5575,
+      "step": 11890
+    },
+    {
+      "epoch": 0.959648131708498,
+      "grad_norm": 0.7884495258331299,
+      "learning_rate": 7.135888867721312e-05,
+      "loss": 2.4807,
+      "step": 11891
+    },
+    {
+      "epoch": 0.9597288354450811,
+      "grad_norm": 0.638469934463501,
+      "learning_rate": 7.134376349980977e-05,
+      "loss": 2.4989,
+      "step": 11892
+    },
+    {
+      "epoch": 0.9598095391816641,
+      "grad_norm": 0.6802849769592285,
+      "learning_rate": 7.132863903659476e-05,
+      "loss": 2.5139,
+      "step": 11893
+    },
+    {
+      "epoch": 0.9598902429182471,
+      "grad_norm": 0.6657521724700928,
+      "learning_rate": 7.131351528794499e-05,
+      "loss": 2.4488,
+      "step": 11894
+    },
+    {
+      "epoch": 0.9599709466548301,
+      "grad_norm": 0.6537562012672424,
+      "learning_rate": 7.129839225423741e-05,
+      "loss": 2.4664,
+      "step": 11895
+    },
+    {
+      "epoch": 0.9600516503914132,
+      "grad_norm": 0.689637303352356,
+      "learning_rate": 7.128326993584897e-05,
+      "loss": 2.582,
+      "step": 11896
+    },
+    {
+      "epoch": 0.9601323541279961,
+      "grad_norm": 0.6701640486717224,
+      "learning_rate": 7.126814833315646e-05,
+      "loss": 2.4963,
+      "step": 11897
+    },
+    {
+      "epoch": 0.9602130578645791,
+      "grad_norm": 0.7466658353805542,
+      "learning_rate": 7.125302744653677e-05,
+      "loss": 2.5015,
+      "step": 11898
+    },
+    {
+      "epoch": 0.9602937616011621,
+      "grad_norm": 0.6487225294113159,
+      "learning_rate": 7.123790727636685e-05,
+      "loss": 2.5393,
+      "step": 11899
+    },
+    {
+      "epoch": 0.9603744653377452,
+      "grad_norm": 0.7204654216766357,
+      "learning_rate": 7.122278782302343e-05,
+      "loss": 2.4668,
+      "step": 11900
+    },
+    {
+      "epoch": 0.9604551690743282,
+      "grad_norm": 0.6852861046791077,
+      "learning_rate": 7.120766908688336e-05,
+      "loss": 2.5893,
+      "step": 11901
+    },
+    {
+      "epoch": 0.9605358728109111,
+      "grad_norm": 0.6483901739120483,
+      "learning_rate": 7.119255106832344e-05,
+      "loss": 2.48,
+      "step": 11902
+    },
+    {
+      "epoch": 0.9606165765474941,
+      "grad_norm": 0.6670375466346741,
+      "learning_rate": 7.117743376772049e-05,
+      "loss": 2.5225,
+      "step": 11903
+    },
+    {
+      "epoch": 0.9606972802840772,
+      "grad_norm": 0.6805974841117859,
+      "learning_rate": 7.116231718545118e-05,
+      "loss": 2.4652,
+      "step": 11904
+    },
+    {
+      "epoch": 0.9607779840206602,
+      "grad_norm": 0.6700397729873657,
+      "learning_rate": 7.114720132189232e-05,
+      "loss": 2.5115,
+      "step": 11905
+    },
+    {
+      "epoch": 0.9608586877572431,
+      "grad_norm": 0.7167409062385559,
+      "learning_rate": 7.113208617742066e-05,
+      "loss": 2.5062,
+      "step": 11906
+    },
+    {
+      "epoch": 0.9609393914938261,
+      "grad_norm": 0.7337077856063843,
+      "learning_rate": 7.111697175241286e-05,
+      "loss": 2.5768,
+      "step": 11907
+    },
+    {
+      "epoch": 0.9610200952304092,
+      "grad_norm": 0.6681819558143616,
+      "learning_rate": 7.110185804724558e-05,
+      "loss": 2.5058,
+      "step": 11908
+    },
+    {
+      "epoch": 0.9611007989669922,
+      "grad_norm": 0.7235603332519531,
+      "learning_rate": 7.10867450622956e-05,
+      "loss": 2.4606,
+      "step": 11909
+    },
+    {
+      "epoch": 0.9611815027035752,
+      "grad_norm": 0.6931360363960266,
+      "learning_rate": 7.107163279793947e-05,
+      "loss": 2.5129,
+      "step": 11910
+    },
+    {
+      "epoch": 0.9612622064401581,
+      "grad_norm": 0.7331648468971252,
+      "learning_rate": 7.105652125455388e-05,
+      "loss": 2.4916,
+      "step": 11911
+    },
+    {
+      "epoch": 0.9613429101767412,
+      "grad_norm": 0.6538143754005432,
+      "learning_rate": 7.104141043251545e-05,
+      "loss": 2.5184,
+      "step": 11912
+    },
+    {
+      "epoch": 0.9614236139133242,
+      "grad_norm": 0.7018921375274658,
+      "learning_rate": 7.102630033220077e-05,
+      "loss": 2.5446,
+      "step": 11913
+    },
+    {
+      "epoch": 0.9615043176499072,
+      "grad_norm": 0.7528507709503174,
+      "learning_rate": 7.10111909539864e-05,
+      "loss": 2.4404,
+      "step": 11914
+    },
+    {
+      "epoch": 0.9615850213864902,
+      "grad_norm": 0.7258831858634949,
+      "learning_rate": 7.099608229824894e-05,
+      "loss": 2.4758,
+      "step": 11915
+    },
+    {
+      "epoch": 0.9616657251230732,
+      "grad_norm": 0.6954349875450134,
+      "learning_rate": 7.098097436536498e-05,
+      "loss": 2.4894,
+      "step": 11916
+    },
+    {
+      "epoch": 0.9617464288596562,
+      "grad_norm": 0.691584050655365,
+      "learning_rate": 7.096586715571092e-05,
+      "loss": 2.544,
+      "step": 11917
+    },
+    {
+      "epoch": 0.9618271325962392,
+      "grad_norm": 0.7107009291648865,
+      "learning_rate": 7.095076066966337e-05,
+      "loss": 2.4994,
+      "step": 11918
+    },
+    {
+      "epoch": 0.9619078363328222,
+      "grad_norm": 0.6492058634757996,
+      "learning_rate": 7.093565490759881e-05,
+      "loss": 2.5751,
+      "step": 11919
+    },
+    {
+      "epoch": 0.9619885400694053,
+      "grad_norm": 0.6817753314971924,
+      "learning_rate": 7.092054986989371e-05,
+      "loss": 2.5129,
+      "step": 11920
+    },
+    {
+      "epoch": 0.9620692438059882,
+      "grad_norm": 0.6991822123527527,
+      "learning_rate": 7.090544555692448e-05,
+      "loss": 2.5728,
+      "step": 11921
+    },
+    {
+      "epoch": 0.9621499475425712,
+      "grad_norm": 0.6627625226974487,
+      "learning_rate": 7.089034196906768e-05,
+      "loss": 2.4479,
+      "step": 11922
+    },
+    {
+      "epoch": 0.9622306512791542,
+      "grad_norm": 0.6889652013778687,
+      "learning_rate": 7.087523910669957e-05,
+      "loss": 2.5323,
+      "step": 11923
+    },
+    {
+      "epoch": 0.9623113550157373,
+      "grad_norm": 0.7863786816596985,
+      "learning_rate": 7.086013697019667e-05,
+      "loss": 2.5146,
+      "step": 11924
+    },
+    {
+      "epoch": 0.9623920587523203,
+      "grad_norm": 0.6885324716567993,
+      "learning_rate": 7.084503555993536e-05,
+      "loss": 2.5072,
+      "step": 11925
+    },
+    {
+      "epoch": 0.9624727624889032,
+      "grad_norm": 0.619239091873169,
+      "learning_rate": 7.082993487629192e-05,
+      "loss": 2.4622,
+      "step": 11926
+    },
+    {
+      "epoch": 0.9625534662254862,
+      "grad_norm": 0.6762447953224182,
+      "learning_rate": 7.081483491964278e-05,
+      "loss": 2.5155,
+      "step": 11927
+    },
+    {
+      "epoch": 0.9626341699620693,
+      "grad_norm": 0.6559715867042542,
+      "learning_rate": 7.079973569036424e-05,
+      "loss": 2.4729,
+      "step": 11928
+    },
+    {
+      "epoch": 0.9627148736986523,
+      "grad_norm": 0.633280873298645,
+      "learning_rate": 7.078463718883261e-05,
+      "loss": 2.4715,
+      "step": 11929
+    },
+    {
+      "epoch": 0.9627955774352353,
+      "grad_norm": 0.7740094065666199,
+      "learning_rate": 7.07695394154242e-05,
+      "loss": 2.4871,
+      "step": 11930
+    },
+    {
+      "epoch": 0.9628762811718182,
+      "grad_norm": 0.7103284597396851,
+      "learning_rate": 7.075444237051527e-05,
+      "loss": 2.5299,
+      "step": 11931
+    },
+    {
+      "epoch": 0.9629569849084013,
+      "grad_norm": 0.6800934076309204,
+      "learning_rate": 7.073934605448212e-05,
+      "loss": 2.5919,
+      "step": 11932
+    },
+    {
+      "epoch": 0.9630376886449843,
+      "grad_norm": 0.6680917143821716,
+      "learning_rate": 7.072425046770092e-05,
+      "loss": 2.4942,
+      "step": 11933
+    },
+    {
+      "epoch": 0.9631183923815673,
+      "grad_norm": 0.7248062491416931,
+      "learning_rate": 7.070915561054792e-05,
+      "loss": 2.4956,
+      "step": 11934
+    },
+    {
+      "epoch": 0.9631990961181502,
+      "grad_norm": 0.6635782122612,
+      "learning_rate": 7.069406148339936e-05,
+      "loss": 2.4658,
+      "step": 11935
+    },
+    {
+      "epoch": 0.9632797998547332,
+      "grad_norm": 0.6751061081886292,
+      "learning_rate": 7.067896808663137e-05,
+      "loss": 2.4912,
+      "step": 11936
+    },
+    {
+      "epoch": 0.9633605035913163,
+      "grad_norm": 0.7476027607917786,
+      "learning_rate": 7.066387542062013e-05,
+      "loss": 2.4858,
+      "step": 11937
+    },
+    {
+      "epoch": 0.9634412073278993,
+      "grad_norm": 0.6770931482315063,
+      "learning_rate": 7.064878348574183e-05,
+      "loss": 2.4574,
+      "step": 11938
+    },
+    {
+      "epoch": 0.9635219110644823,
+      "grad_norm": 0.7105392813682556,
+      "learning_rate": 7.063369228237255e-05,
+      "loss": 2.5523,
+      "step": 11939
+    },
+    {
+      "epoch": 0.9636026148010652,
+      "grad_norm": 0.6806207299232483,
+      "learning_rate": 7.061860181088842e-05,
+      "loss": 2.4992,
+      "step": 11940
+    },
+    {
+      "epoch": 0.9636833185376483,
+      "grad_norm": 0.7059600353240967,
+      "learning_rate": 7.060351207166558e-05,
+      "loss": 2.5778,
+      "step": 11941
+    },
+    {
+      "epoch": 0.9637640222742313,
+      "grad_norm": 0.6306884288787842,
+      "learning_rate": 7.058842306508002e-05,
+      "loss": 2.5389,
+      "step": 11942
+    },
+    {
+      "epoch": 0.9638447260108143,
+      "grad_norm": 0.6997150778770447,
+      "learning_rate": 7.057333479150783e-05,
+      "loss": 2.5077,
+      "step": 11943
+    },
+    {
+      "epoch": 0.9639254297473973,
+      "grad_norm": 0.7073743343353271,
+      "learning_rate": 7.05582472513251e-05,
+      "loss": 2.5274,
+      "step": 11944
+    },
+    {
+      "epoch": 0.9640061334839803,
+      "grad_norm": 0.6768803596496582,
+      "learning_rate": 7.054316044490777e-05,
+      "loss": 2.5155,
+      "step": 11945
+    },
+    {
+      "epoch": 0.9640868372205633,
+      "grad_norm": 0.6792057752609253,
+      "learning_rate": 7.052807437263189e-05,
+      "loss": 2.5509,
+      "step": 11946
+    },
+    {
+      "epoch": 0.9641675409571463,
+      "grad_norm": 0.6883981823921204,
+      "learning_rate": 7.051298903487344e-05,
+      "loss": 2.5176,
+      "step": 11947
+    },
+    {
+      "epoch": 0.9642482446937293,
+      "grad_norm": 0.6934401392936707,
+      "learning_rate": 7.049790443200844e-05,
+      "loss": 2.502,
+      "step": 11948
+    },
+    {
+      "epoch": 0.9643289484303124,
+      "grad_norm": 0.6882597804069519,
+      "learning_rate": 7.048282056441269e-05,
+      "loss": 2.487,
+      "step": 11949
+    },
+    {
+      "epoch": 0.9644096521668953,
+      "grad_norm": 0.6972896456718445,
+      "learning_rate": 7.046773743246225e-05,
+      "loss": 2.5304,
+      "step": 11950
+    },
+    {
+      "epoch": 0.9644903559034783,
+      "grad_norm": 0.6591988205909729,
+      "learning_rate": 7.045265503653303e-05,
+      "loss": 2.4734,
+      "step": 11951
+    },
+    {
+      "epoch": 0.9645710596400613,
+      "grad_norm": 0.6890063285827637,
+      "learning_rate": 7.043757337700082e-05,
+      "loss": 2.5289,
+      "step": 11952
+    },
+    {
+      "epoch": 0.9646517633766444,
+      "grad_norm": 0.6931065917015076,
+      "learning_rate": 7.042249245424157e-05,
+      "loss": 2.484,
+      "step": 11953
+    },
+    {
+      "epoch": 0.9647324671132274,
+      "grad_norm": 0.6943762898445129,
+      "learning_rate": 7.040741226863117e-05,
+      "loss": 2.501,
+      "step": 11954
+    },
+    {
+      "epoch": 0.9648131708498103,
+      "grad_norm": 0.677154004573822,
+      "learning_rate": 7.039233282054536e-05,
+      "loss": 2.4976,
+      "step": 11955
+    },
+    {
+      "epoch": 0.9648938745863933,
+      "grad_norm": 0.6662883758544922,
+      "learning_rate": 7.037725411036003e-05,
+      "loss": 2.4928,
+      "step": 11956
+    },
+    {
+      "epoch": 0.9649745783229764,
+      "grad_norm": 0.6854663491249084,
+      "learning_rate": 7.0362176138451e-05,
+      "loss": 2.4657,
+      "step": 11957
+    },
+    {
+      "epoch": 0.9650552820595594,
+      "grad_norm": 0.6703238487243652,
+      "learning_rate": 7.034709890519397e-05,
+      "loss": 2.4879,
+      "step": 11958
+    },
+    {
+      "epoch": 0.9651359857961423,
+      "grad_norm": 0.7023652791976929,
+      "learning_rate": 7.033202241096474e-05,
+      "loss": 2.4619,
+      "step": 11959
+    },
+    {
+      "epoch": 0.9652166895327253,
+      "grad_norm": 0.6950454711914062,
+      "learning_rate": 7.031694665613911e-05,
+      "loss": 2.5125,
+      "step": 11960
+    },
+    {
+      "epoch": 0.9652973932693084,
+      "grad_norm": 0.6740411520004272,
+      "learning_rate": 7.030187164109272e-05,
+      "loss": 2.436,
+      "step": 11961
+    },
+    {
+      "epoch": 0.9653780970058914,
+      "grad_norm": 0.6697152256965637,
+      "learning_rate": 7.028679736620132e-05,
+      "loss": 2.5513,
+      "step": 11962
+    },
+    {
+      "epoch": 0.9654588007424744,
+      "grad_norm": 0.6920599937438965,
+      "learning_rate": 7.027172383184061e-05,
+      "loss": 2.5264,
+      "step": 11963
+    },
+    {
+      "epoch": 0.9655395044790573,
+      "grad_norm": 0.6493465304374695,
+      "learning_rate": 7.025665103838627e-05,
+      "loss": 2.4834,
+      "step": 11964
+    },
+    {
+      "epoch": 0.9656202082156404,
+      "grad_norm": 0.684092104434967,
+      "learning_rate": 7.02415789862139e-05,
+      "loss": 2.4662,
+      "step": 11965
+    },
+    {
+      "epoch": 0.9657009119522234,
+      "grad_norm": 0.7161515355110168,
+      "learning_rate": 7.022650767569921e-05,
+      "loss": 2.4648,
+      "step": 11966
+    },
+    {
+      "epoch": 0.9657816156888064,
+      "grad_norm": 0.6994524002075195,
+      "learning_rate": 7.021143710721778e-05,
+      "loss": 2.5186,
+      "step": 11967
+    },
+    {
+      "epoch": 0.9658623194253894,
+      "grad_norm": 0.7105295062065125,
+      "learning_rate": 7.019636728114518e-05,
+      "loss": 2.5132,
+      "step": 11968
+    },
+    {
+      "epoch": 0.9659430231619724,
+      "grad_norm": 0.7182292938232422,
+      "learning_rate": 7.018129819785702e-05,
+      "loss": 2.5469,
+      "step": 11969
+    },
+    {
+      "epoch": 0.9660237268985554,
+      "grad_norm": 0.7021759152412415,
+      "learning_rate": 7.016622985772887e-05,
+      "loss": 2.5477,
+      "step": 11970
+    },
+    {
+      "epoch": 0.9661044306351384,
+      "grad_norm": 0.6751413941383362,
+      "learning_rate": 7.015116226113624e-05,
+      "loss": 2.5174,
+      "step": 11971
+    },
+    {
+      "epoch": 0.9661851343717214,
+      "grad_norm": 0.6341918110847473,
+      "learning_rate": 7.013609540845468e-05,
+      "loss": 2.4778,
+      "step": 11972
+    },
+    {
+      "epoch": 0.9662658381083045,
+      "grad_norm": 0.7080956697463989,
+      "learning_rate": 7.012102930005971e-05,
+      "loss": 2.5304,
+      "step": 11973
+    },
+    {
+      "epoch": 0.9663465418448874,
+      "grad_norm": 0.6367003321647644,
+      "learning_rate": 7.010596393632674e-05,
+      "loss": 2.4857,
+      "step": 11974
+    },
+    {
+      "epoch": 0.9664272455814704,
+      "grad_norm": 0.6841328740119934,
+      "learning_rate": 7.009089931763131e-05,
+      "loss": 2.5365,
+      "step": 11975
+    },
+    {
+      "epoch": 0.9665079493180534,
+      "grad_norm": 0.6568236351013184,
+      "learning_rate": 7.00758354443489e-05,
+      "loss": 2.5286,
+      "step": 11976
+    },
+    {
+      "epoch": 0.9665886530546365,
+      "grad_norm": 0.7071812152862549,
+      "learning_rate": 7.006077231685485e-05,
+      "loss": 2.458,
+      "step": 11977
+    },
+    {
+      "epoch": 0.9666693567912195,
+      "grad_norm": 0.6997712850570679,
+      "learning_rate": 7.004570993552462e-05,
+      "loss": 2.4571,
+      "step": 11978
+    },
+    {
+      "epoch": 0.9667500605278024,
+      "grad_norm": 0.6920793056488037,
+      "learning_rate": 7.003064830073359e-05,
+      "loss": 2.4172,
+      "step": 11979
+    },
+    {
+      "epoch": 0.9668307642643854,
+      "grad_norm": 0.6823387742042542,
+      "learning_rate": 7.001558741285718e-05,
+      "loss": 2.4895,
+      "step": 11980
+    },
+    {
+      "epoch": 0.9669114680009685,
+      "grad_norm": 0.7309569716453552,
+      "learning_rate": 7.000052727227068e-05,
+      "loss": 2.502,
+      "step": 11981
+    },
+    {
+      "epoch": 0.9669921717375515,
+      "grad_norm": 0.734708845615387,
+      "learning_rate": 6.998546787934946e-05,
+      "loss": 2.4918,
+      "step": 11982
+    },
+    {
+      "epoch": 0.9670728754741345,
+      "grad_norm": 0.690406084060669,
+      "learning_rate": 6.997040923446889e-05,
+      "loss": 2.4994,
+      "step": 11983
+    },
+    {
+      "epoch": 0.9671535792107174,
+      "grad_norm": 0.7126687169075012,
+      "learning_rate": 6.995535133800416e-05,
+      "loss": 2.4824,
+      "step": 11984
+    },
+    {
+      "epoch": 0.9672342829473004,
+      "grad_norm": 0.7020599246025085,
+      "learning_rate": 6.994029419033062e-05,
+      "loss": 2.4889,
+      "step": 11985
+    },
+    {
+      "epoch": 0.9673149866838835,
+      "grad_norm": 0.7690796852111816,
+      "learning_rate": 6.992523779182356e-05,
+      "loss": 2.4997,
+      "step": 11986
+    },
+    {
+      "epoch": 0.9673956904204665,
+      "grad_norm": 0.6635778546333313,
+      "learning_rate": 6.991018214285816e-05,
+      "loss": 2.4989,
+      "step": 11987
+    },
+    {
+      "epoch": 0.9674763941570494,
+      "grad_norm": 0.7088577747344971,
+      "learning_rate": 6.989512724380967e-05,
+      "loss": 2.549,
+      "step": 11988
+    },
+    {
+      "epoch": 0.9675570978936324,
+      "grad_norm": 0.6420924663543701,
+      "learning_rate": 6.988007309505333e-05,
+      "loss": 2.4585,
+      "step": 11989
+    },
+    {
+      "epoch": 0.9676378016302155,
+      "grad_norm": 0.7902400493621826,
+      "learning_rate": 6.986501969696428e-05,
+      "loss": 2.5009,
+      "step": 11990
+    },
+    {
+      "epoch": 0.9677185053667985,
+      "grad_norm": 0.700907289981842,
+      "learning_rate": 6.984996704991773e-05,
+      "loss": 2.4778,
+      "step": 11991
+    },
+    {
+      "epoch": 0.9677992091033815,
+      "grad_norm": 0.664378821849823,
+      "learning_rate": 6.983491515428883e-05,
+      "loss": 2.5116,
+      "step": 11992
+    },
+    {
+      "epoch": 0.9678799128399644,
+      "grad_norm": 0.6314663887023926,
+      "learning_rate": 6.981986401045266e-05,
+      "loss": 2.4588,
+      "step": 11993
+    },
+    {
+      "epoch": 0.9679606165765475,
+      "grad_norm": 0.6521078944206238,
+      "learning_rate": 6.980481361878438e-05,
+      "loss": 2.5224,
+      "step": 11994
+    },
+    {
+      "epoch": 0.9680413203131305,
+      "grad_norm": 0.6336014270782471,
+      "learning_rate": 6.978976397965907e-05,
+      "loss": 2.4297,
+      "step": 11995
+    },
+    {
+      "epoch": 0.9681220240497135,
+      "grad_norm": 0.7321500778198242,
+      "learning_rate": 6.977471509345183e-05,
+      "loss": 2.5252,
+      "step": 11996
+    },
+    {
+      "epoch": 0.9682027277862965,
+      "grad_norm": 0.686950147151947,
+      "learning_rate": 6.97596669605377e-05,
+      "loss": 2.5188,
+      "step": 11997
+    },
+    {
+      "epoch": 0.9682834315228795,
+      "grad_norm": 0.729343056678772,
+      "learning_rate": 6.97446195812917e-05,
+      "loss": 2.5157,
+      "step": 11998
+    },
+    {
+      "epoch": 0.9683641352594625,
+      "grad_norm": 0.6447068452835083,
+      "learning_rate": 6.972957295608889e-05,
+      "loss": 2.5041,
+      "step": 11999
+    },
+    {
+      "epoch": 0.9684448389960455,
+      "grad_norm": 0.6847280859947205,
+      "learning_rate": 6.971452708530423e-05,
+      "loss": 2.443,
+      "step": 12000
+    },
+    {
+      "epoch": 0.9684448389960455,
+      "eval_loss": 2.431878089904785,
+      "eval_runtime": 758.167,
+      "eval_samples_per_second": 3.456,
+      "eval_steps_per_second": 0.576,
+      "step": 12000
+    },
+    {
+      "epoch": 0.9685255427326285,
+      "grad_norm": 0.6440466046333313,
+      "learning_rate": 6.969948196931272e-05,
+      "loss": 2.5091,
+      "step": 12001
+    },
+    {
+      "epoch": 0.9686062464692116,
+      "grad_norm": 0.6570029258728027,
+      "learning_rate": 6.968443760848937e-05,
+      "loss": 2.491,
+      "step": 12002
+    },
+    {
+      "epoch": 0.9686869502057945,
+      "grad_norm": 0.7610877752304077,
+      "learning_rate": 6.966939400320905e-05,
+      "loss": 2.4713,
+      "step": 12003
+    },
+    {
+      "epoch": 0.9687676539423775,
+      "grad_norm": 0.7187781929969788,
+      "learning_rate": 6.965435115384669e-05,
+      "loss": 2.4303,
+      "step": 12004
+    },
+    {
+      "epoch": 0.9688483576789605,
+      "grad_norm": 0.7668420672416687,
+      "learning_rate": 6.963930906077727e-05,
+      "loss": 2.5513,
+      "step": 12005
+    },
+    {
+      "epoch": 0.9689290614155436,
+      "grad_norm": 0.7025619745254517,
+      "learning_rate": 6.96242677243756e-05,
+      "loss": 2.4349,
+      "step": 12006
+    },
+    {
+      "epoch": 0.9690097651521266,
+      "grad_norm": 0.7066935896873474,
+      "learning_rate": 6.960922714501657e-05,
+      "loss": 2.5465,
+      "step": 12007
+    },
+    {
+      "epoch": 0.9690904688887095,
+      "grad_norm": 0.6758970618247986,
+      "learning_rate": 6.95941873230751e-05,
+      "loss": 2.4827,
+      "step": 12008
+    },
+    {
+      "epoch": 0.9691711726252925,
+      "grad_norm": 0.7108862996101379,
+      "learning_rate": 6.957914825892591e-05,
+      "loss": 2.5412,
+      "step": 12009
+    },
+    {
+      "epoch": 0.9692518763618756,
+      "grad_norm": 0.660784125328064,
+      "learning_rate": 6.956410995294389e-05,
+      "loss": 2.5173,
+      "step": 12010
+    },
+    {
+      "epoch": 0.9693325800984586,
+      "grad_norm": 0.6966561079025269,
+      "learning_rate": 6.954907240550377e-05,
+      "loss": 2.5196,
+      "step": 12011
+    },
+    {
+      "epoch": 0.9694132838350416,
+      "grad_norm": 0.6889416575431824,
+      "learning_rate": 6.953403561698042e-05,
+      "loss": 2.5351,
+      "step": 12012
+    },
+    {
+      "epoch": 0.9694939875716245,
+      "grad_norm": 0.7578341960906982,
+      "learning_rate": 6.951899958774852e-05,
+      "loss": 2.5184,
+      "step": 12013
+    },
+    {
+      "epoch": 0.9695746913082076,
+      "grad_norm": 0.6735317707061768,
+      "learning_rate": 6.950396431818282e-05,
+      "loss": 2.4592,
+      "step": 12014
+    },
+    {
+      "epoch": 0.9696553950447906,
+      "grad_norm": 0.6903232932090759,
+      "learning_rate": 6.948892980865806e-05,
+      "loss": 2.5212,
+      "step": 12015
+    },
+    {
+      "epoch": 0.9697360987813736,
+      "grad_norm": 0.6477165818214417,
+      "learning_rate": 6.94738960595489e-05,
+      "loss": 2.4423,
+      "step": 12016
+    },
+    {
+      "epoch": 0.9698168025179565,
+      "grad_norm": 0.6778751015663147,
+      "learning_rate": 6.945886307123007e-05,
+      "loss": 2.547,
+      "step": 12017
+    },
+    {
+      "epoch": 0.9698975062545396,
+      "grad_norm": 0.690558135509491,
+      "learning_rate": 6.944383084407623e-05,
+      "loss": 2.5081,
+      "step": 12018
+    },
+    {
+      "epoch": 0.9699782099911226,
+      "grad_norm": 0.7210639119148254,
+      "learning_rate": 6.942879937846196e-05,
+      "loss": 2.496,
+      "step": 12019
+    },
+    {
+      "epoch": 0.9700589137277056,
+      "grad_norm": 0.7182444930076599,
+      "learning_rate": 6.941376867476194e-05,
+      "loss": 2.6138,
+      "step": 12020
+    },
+    {
+      "epoch": 0.9701396174642886,
+      "grad_norm": 0.6929295063018799,
+      "learning_rate": 6.939873873335077e-05,
+      "loss": 2.4828,
+      "step": 12021
+    },
+    {
+      "epoch": 0.9702203212008716,
+      "grad_norm": 0.6919693350791931,
+      "learning_rate": 6.938370955460298e-05,
+      "loss": 2.5123,
+      "step": 12022
+    },
+    {
+      "epoch": 0.9703010249374546,
+      "grad_norm": 0.6475244164466858,
+      "learning_rate": 6.93686811388932e-05,
+      "loss": 2.4992,
+      "step": 12023
+    },
+    {
+      "epoch": 0.9703817286740376,
+      "grad_norm": 0.6728265881538391,
+      "learning_rate": 6.935365348659597e-05,
+      "loss": 2.4486,
+      "step": 12024
+    },
+    {
+      "epoch": 0.9704624324106206,
+      "grad_norm": 0.6791470646858215,
+      "learning_rate": 6.933862659808582e-05,
+      "loss": 2.4657,
+      "step": 12025
+    },
+    {
+      "epoch": 0.9705431361472037,
+      "grad_norm": 0.7611662745475769,
+      "learning_rate": 6.932360047373721e-05,
+      "loss": 2.5243,
+      "step": 12026
+    },
+    {
+      "epoch": 0.9706238398837866,
+      "grad_norm": 0.6642355918884277,
+      "learning_rate": 6.930857511392467e-05,
+      "loss": 2.5308,
+      "step": 12027
+    },
+    {
+      "epoch": 0.9707045436203696,
+      "grad_norm": 0.7270805239677429,
+      "learning_rate": 6.92935505190227e-05,
+      "loss": 2.4708,
+      "step": 12028
+    },
+    {
+      "epoch": 0.9707852473569526,
+      "grad_norm": 0.6706295013427734,
+      "learning_rate": 6.927852668940568e-05,
+      "loss": 2.5136,
+      "step": 12029
+    },
+    {
+      "epoch": 0.9708659510935357,
+      "grad_norm": 0.6923376321792603,
+      "learning_rate": 6.92635036254481e-05,
+      "loss": 2.5238,
+      "step": 12030
+    },
+    {
+      "epoch": 0.9709466548301187,
+      "grad_norm": 0.7154483199119568,
+      "learning_rate": 6.924848132752436e-05,
+      "loss": 2.488,
+      "step": 12031
+    },
+    {
+      "epoch": 0.9710273585667016,
+      "grad_norm": 0.6675701141357422,
+      "learning_rate": 6.923345979600884e-05,
+      "loss": 2.5066,
+      "step": 12032
+    },
+    {
+      "epoch": 0.9711080623032846,
+      "grad_norm": 0.7282043695449829,
+      "learning_rate": 6.921843903127592e-05,
+      "loss": 2.5096,
+      "step": 12033
+    },
+    {
+      "epoch": 0.9711887660398677,
+      "grad_norm": 0.663526177406311,
+      "learning_rate": 6.92034190337e-05,
+      "loss": 2.5276,
+      "step": 12034
+    },
+    {
+      "epoch": 0.9712694697764507,
+      "grad_norm": 0.7491087913513184,
+      "learning_rate": 6.918839980365534e-05,
+      "loss": 2.5044,
+      "step": 12035
+    },
+    {
+      "epoch": 0.9713501735130337,
+      "grad_norm": 0.6977766156196594,
+      "learning_rate": 6.917338134151629e-05,
+      "loss": 2.6102,
+      "step": 12036
+    },
+    {
+      "epoch": 0.9714308772496166,
+      "grad_norm": 0.6447446346282959,
+      "learning_rate": 6.915836364765722e-05,
+      "loss": 2.5137,
+      "step": 12037
+    },
+    {
+      "epoch": 0.9715115809861996,
+      "grad_norm": 0.6801442503929138,
+      "learning_rate": 6.91433467224523e-05,
+      "loss": 2.5145,
+      "step": 12038
+    },
+    {
+      "epoch": 0.9715922847227827,
+      "grad_norm": 0.6843627691268921,
+      "learning_rate": 6.912833056627583e-05,
+      "loss": 2.6099,
+      "step": 12039
+    },
+    {
+      "epoch": 0.9716729884593657,
+      "grad_norm": 0.6862856149673462,
+      "learning_rate": 6.911331517950209e-05,
+      "loss": 2.5358,
+      "step": 12040
+    },
+    {
+      "epoch": 0.9717536921959486,
+      "grad_norm": 0.6835047602653503,
+      "learning_rate": 6.909830056250527e-05,
+      "loss": 2.5257,
+      "step": 12041
+    },
+    {
+      "epoch": 0.9718343959325316,
+      "grad_norm": 0.6958080530166626,
+      "learning_rate": 6.908328671565956e-05,
+      "loss": 2.5008,
+      "step": 12042
+    },
+    {
+      "epoch": 0.9719150996691147,
+      "grad_norm": 0.7556219100952148,
+      "learning_rate": 6.906827363933917e-05,
+      "loss": 2.5283,
+      "step": 12043
+    },
+    {
+      "epoch": 0.9719958034056977,
+      "grad_norm": 0.7074917554855347,
+      "learning_rate": 6.90532613339183e-05,
+      "loss": 2.4898,
+      "step": 12044
+    },
+    {
+      "epoch": 0.9720765071422807,
+      "grad_norm": 0.6456350684165955,
+      "learning_rate": 6.903824979977101e-05,
+      "loss": 2.4989,
+      "step": 12045
+    },
+    {
+      "epoch": 0.9721572108788636,
+      "grad_norm": 0.6609941720962524,
+      "learning_rate": 6.902323903727146e-05,
+      "loss": 2.4883,
+      "step": 12046
+    },
+    {
+      "epoch": 0.9722379146154467,
+      "grad_norm": 0.7132936716079712,
+      "learning_rate": 6.90082290467938e-05,
+      "loss": 2.4983,
+      "step": 12047
+    },
+    {
+      "epoch": 0.9723186183520297,
+      "grad_norm": 0.6686434745788574,
+      "learning_rate": 6.899321982871206e-05,
+      "loss": 2.4862,
+      "step": 12048
+    },
+    {
+      "epoch": 0.9723993220886127,
+      "grad_norm": 0.6792194247245789,
+      "learning_rate": 6.897821138340033e-05,
+      "loss": 2.5368,
+      "step": 12049
+    },
+    {
+      "epoch": 0.9724800258251957,
+      "grad_norm": 0.6829379796981812,
+      "learning_rate": 6.896320371123268e-05,
+      "loss": 2.4842,
+      "step": 12050
+    },
+    {
+      "epoch": 0.9725607295617787,
+      "grad_norm": 0.7459573745727539,
+      "learning_rate": 6.894819681258312e-05,
+      "loss": 2.5023,
+      "step": 12051
+    },
+    {
+      "epoch": 0.9726414332983617,
+      "grad_norm": 0.6700068712234497,
+      "learning_rate": 6.893319068782566e-05,
+      "loss": 2.552,
+      "step": 12052
+    },
+    {
+      "epoch": 0.9727221370349447,
+      "grad_norm": 0.7093638777732849,
+      "learning_rate": 6.891818533733434e-05,
+      "loss": 2.445,
+      "step": 12053
+    },
+    {
+      "epoch": 0.9728028407715277,
+      "grad_norm": 0.703599214553833,
+      "learning_rate": 6.890318076148304e-05,
+      "loss": 2.5536,
+      "step": 12054
+    },
+    {
+      "epoch": 0.9728835445081108,
+      "grad_norm": 0.6214482188224792,
+      "learning_rate": 6.888817696064578e-05,
+      "loss": 2.5188,
+      "step": 12055
+    },
+    {
+      "epoch": 0.9729642482446937,
+      "grad_norm": 0.6893547773361206,
+      "learning_rate": 6.887317393519645e-05,
+      "loss": 2.5596,
+      "step": 12056
+    },
+    {
+      "epoch": 0.9730449519812767,
+      "grad_norm": 0.6282656788825989,
+      "learning_rate": 6.885817168550903e-05,
+      "loss": 2.4873,
+      "step": 12057
+    },
+    {
+      "epoch": 0.9731256557178597,
+      "grad_norm": 0.6979188323020935,
+      "learning_rate": 6.884317021195737e-05,
+      "loss": 2.5358,
+      "step": 12058
+    },
+    {
+      "epoch": 0.9732063594544428,
+      "grad_norm": 0.7925785183906555,
+      "learning_rate": 6.882816951491533e-05,
+      "loss": 2.5358,
+      "step": 12059
+    },
+    {
+      "epoch": 0.9732870631910258,
+      "grad_norm": 0.6449821591377258,
+      "learning_rate": 6.881316959475684e-05,
+      "loss": 2.4784,
+      "step": 12060
+    },
+    {
+      "epoch": 0.9733677669276087,
+      "grad_norm": 0.7013393044471741,
+      "learning_rate": 6.879817045185565e-05,
+      "loss": 2.4804,
+      "step": 12061
+    },
+    {
+      "epoch": 0.9734484706641917,
+      "grad_norm": 0.8338057398796082,
+      "learning_rate": 6.878317208658559e-05,
+      "loss": 2.512,
+      "step": 12062
+    },
+    {
+      "epoch": 0.9735291744007748,
+      "grad_norm": 0.6815133094787598,
+      "learning_rate": 6.876817449932054e-05,
+      "loss": 2.467,
+      "step": 12063
+    },
+    {
+      "epoch": 0.9736098781373578,
+      "grad_norm": 0.659156858921051,
+      "learning_rate": 6.87531776904342e-05,
+      "loss": 2.503,
+      "step": 12064
+    },
+    {
+      "epoch": 0.9736905818739408,
+      "grad_norm": 0.7149603962898254,
+      "learning_rate": 6.873818166030033e-05,
+      "loss": 2.5135,
+      "step": 12065
+    },
+    {
+      "epoch": 0.9737712856105237,
+      "grad_norm": 0.7010510563850403,
+      "learning_rate": 6.872318640929272e-05,
+      "loss": 2.5133,
+      "step": 12066
+    },
+    {
+      "epoch": 0.9738519893471068,
+      "grad_norm": 0.6247616410255432,
+      "learning_rate": 6.870819193778504e-05,
+      "loss": 2.5189,
+      "step": 12067
+    },
+    {
+      "epoch": 0.9739326930836898,
+      "grad_norm": 0.6938940286636353,
+      "learning_rate": 6.869319824615101e-05,
+      "loss": 2.5053,
+      "step": 12068
+    },
+    {
+      "epoch": 0.9740133968202728,
+      "grad_norm": 0.7636895179748535,
+      "learning_rate": 6.867820533476436e-05,
+      "loss": 2.4989,
+      "step": 12069
+    },
+    {
+      "epoch": 0.9740941005568557,
+      "grad_norm": 0.6489234566688538,
+      "learning_rate": 6.866321320399869e-05,
+      "loss": 2.4935,
+      "step": 12070
+    },
+    {
+      "epoch": 0.9741748042934388,
+      "grad_norm": 0.6752095818519592,
+      "learning_rate": 6.864822185422764e-05,
+      "loss": 2.4835,
+      "step": 12071
+    },
+    {
+      "epoch": 0.9742555080300218,
+      "grad_norm": 0.6947118639945984,
+      "learning_rate": 6.863323128582486e-05,
+      "loss": 2.504,
+      "step": 12072
+    },
+    {
+      "epoch": 0.9743362117666048,
+      "grad_norm": 0.6815536618232727,
+      "learning_rate": 6.861824149916398e-05,
+      "loss": 2.5369,
+      "step": 12073
+    },
+    {
+      "epoch": 0.9744169155031878,
+      "grad_norm": 0.6550236344337463,
+      "learning_rate": 6.860325249461852e-05,
+      "loss": 2.4753,
+      "step": 12074
+    },
+    {
+      "epoch": 0.9744976192397709,
+      "grad_norm": 0.6833250522613525,
+      "learning_rate": 6.858826427256209e-05,
+      "loss": 2.4687,
+      "step": 12075
+    },
+    {
+      "epoch": 0.9745783229763538,
+      "grad_norm": 0.6925075650215149,
+      "learning_rate": 6.857327683336824e-05,
+      "loss": 2.5363,
+      "step": 12076
+    },
+    {
+      "epoch": 0.9746590267129368,
+      "grad_norm": 0.6754821538925171,
+      "learning_rate": 6.855829017741046e-05,
+      "loss": 2.4696,
+      "step": 12077
+    },
+    {
+      "epoch": 0.9747397304495198,
+      "grad_norm": 0.7360671162605286,
+      "learning_rate": 6.854330430506228e-05,
+      "loss": 2.5144,
+      "step": 12078
+    },
+    {
+      "epoch": 0.9748204341861029,
+      "grad_norm": 0.6814733743667603,
+      "learning_rate": 6.852831921669723e-05,
+      "loss": 2.5059,
+      "step": 12079
+    },
+    {
+      "epoch": 0.9749011379226858,
+      "grad_norm": 0.7106744647026062,
+      "learning_rate": 6.851333491268869e-05,
+      "loss": 2.453,
+      "step": 12080
+    },
+    {
+      "epoch": 0.9749818416592688,
+      "grad_norm": 0.6623831987380981,
+      "learning_rate": 6.849835139341015e-05,
+      "loss": 2.5244,
+      "step": 12081
+    },
+    {
+      "epoch": 0.9750625453958518,
+      "grad_norm": 0.6723372936248779,
+      "learning_rate": 6.848336865923506e-05,
+      "loss": 2.5159,
+      "step": 12082
+    },
+    {
+      "epoch": 0.9751432491324349,
+      "grad_norm": 0.7256618142127991,
+      "learning_rate": 6.84683867105368e-05,
+      "loss": 2.494,
+      "step": 12083
+    },
+    {
+      "epoch": 0.9752239528690179,
+      "grad_norm": 0.6881731152534485,
+      "learning_rate": 6.845340554768874e-05,
+      "loss": 2.4374,
+      "step": 12084
+    },
+    {
+      "epoch": 0.9753046566056008,
+      "grad_norm": 0.6759666204452515,
+      "learning_rate": 6.843842517106434e-05,
+      "loss": 2.5082,
+      "step": 12085
+    },
+    {
+      "epoch": 0.9753853603421838,
+      "grad_norm": 0.6983315348625183,
+      "learning_rate": 6.842344558103684e-05,
+      "loss": 2.5191,
+      "step": 12086
+    },
+    {
+      "epoch": 0.9754660640787668,
+      "grad_norm": 0.6805596351623535,
+      "learning_rate": 6.840846677797959e-05,
+      "loss": 2.5289,
+      "step": 12087
+    },
+    {
+      "epoch": 0.9755467678153499,
+      "grad_norm": 0.712942361831665,
+      "learning_rate": 6.839348876226595e-05,
+      "loss": 2.5544,
+      "step": 12088
+    },
+    {
+      "epoch": 0.9756274715519329,
+      "grad_norm": 0.6931124329566956,
+      "learning_rate": 6.837851153426924e-05,
+      "loss": 2.5407,
+      "step": 12089
+    },
+    {
+      "epoch": 0.9757081752885158,
+      "grad_norm": 0.6939486265182495,
+      "learning_rate": 6.836353509436264e-05,
+      "loss": 2.5236,
+      "step": 12090
+    },
+    {
+      "epoch": 0.9757888790250988,
+      "grad_norm": 0.7434083223342896,
+      "learning_rate": 6.834855944291944e-05,
+      "loss": 2.4903,
+      "step": 12091
+    },
+    {
+      "epoch": 0.9758695827616819,
+      "grad_norm": 0.672177255153656,
+      "learning_rate": 6.833358458031292e-05,
+      "loss": 2.4995,
+      "step": 12092
+    },
+    {
+      "epoch": 0.9759502864982649,
+      "grad_norm": 0.6631280779838562,
+      "learning_rate": 6.831861050691619e-05,
+      "loss": 2.4689,
+      "step": 12093
+    },
+    {
+      "epoch": 0.9760309902348479,
+      "grad_norm": 0.7485793232917786,
+      "learning_rate": 6.830363722310253e-05,
+      "loss": 2.5526,
+      "step": 12094
+    },
+    {
+      "epoch": 0.9761116939714308,
+      "grad_norm": 0.6592193245887756,
+      "learning_rate": 6.828866472924511e-05,
+      "loss": 2.4425,
+      "step": 12095
+    },
+    {
+      "epoch": 0.9761923977080139,
+      "grad_norm": 0.6479860544204712,
+      "learning_rate": 6.827369302571703e-05,
+      "loss": 2.4637,
+      "step": 12096
+    },
+    {
+      "epoch": 0.9762731014445969,
+      "grad_norm": 0.6694966554641724,
+      "learning_rate": 6.825872211289146e-05,
+      "loss": 2.5256,
+      "step": 12097
+    },
+    {
+      "epoch": 0.9763538051811799,
+      "grad_norm": 0.675751805305481,
+      "learning_rate": 6.82437519911415e-05,
+      "loss": 2.5021,
+      "step": 12098
+    },
+    {
+      "epoch": 0.9764345089177628,
+      "grad_norm": 0.7255450487136841,
+      "learning_rate": 6.822878266084026e-05,
+      "loss": 2.5275,
+      "step": 12099
+    },
+    {
+      "epoch": 0.9765152126543459,
+      "grad_norm": 0.7034213542938232,
+      "learning_rate": 6.821381412236079e-05,
+      "loss": 2.5432,
+      "step": 12100
+    },
+    {
+      "epoch": 0.9765959163909289,
+      "grad_norm": 0.6808038949966431,
+      "learning_rate": 6.819884637607619e-05,
+      "loss": 2.5044,
+      "step": 12101
+    },
+    {
+      "epoch": 0.9766766201275119,
+      "grad_norm": 0.6601580381393433,
+      "learning_rate": 6.818387942235945e-05,
+      "loss": 2.4602,
+      "step": 12102
+    },
+    {
+      "epoch": 0.9767573238640949,
+      "grad_norm": 0.7163928151130676,
+      "learning_rate": 6.816891326158359e-05,
+      "loss": 2.4785,
+      "step": 12103
+    },
+    {
+      "epoch": 0.976838027600678,
+      "grad_norm": 0.6616904735565186,
+      "learning_rate": 6.815394789412164e-05,
+      "loss": 2.5081,
+      "step": 12104
+    },
+    {
+      "epoch": 0.9769187313372609,
+      "grad_norm": 0.6476422548294067,
+      "learning_rate": 6.813898332034657e-05,
+      "loss": 2.4624,
+      "step": 12105
+    },
+    {
+      "epoch": 0.9769994350738439,
+      "grad_norm": 0.6468440890312195,
+      "learning_rate": 6.812401954063131e-05,
+      "loss": 2.4948,
+      "step": 12106
+    },
+    {
+      "epoch": 0.9770801388104269,
+      "grad_norm": 0.6988391876220703,
+      "learning_rate": 6.810905655534878e-05,
+      "loss": 2.4958,
+      "step": 12107
+    },
+    {
+      "epoch": 0.97716084254701,
+      "grad_norm": 0.6777953505516052,
+      "learning_rate": 6.809409436487196e-05,
+      "loss": 2.5304,
+      "step": 12108
+    },
+    {
+      "epoch": 0.9772415462835929,
+      "grad_norm": 0.7115550637245178,
+      "learning_rate": 6.807913296957368e-05,
+      "loss": 2.5321,
+      "step": 12109
+    },
+    {
+      "epoch": 0.9773222500201759,
+      "grad_norm": 0.737823486328125,
+      "learning_rate": 6.806417236982684e-05,
+      "loss": 2.5121,
+      "step": 12110
+    },
+    {
+      "epoch": 0.9774029537567589,
+      "grad_norm": 0.6797437071800232,
+      "learning_rate": 6.804921256600439e-05,
+      "loss": 2.4783,
+      "step": 12111
+    },
+    {
+      "epoch": 0.977483657493342,
+      "grad_norm": 0.7240802645683289,
+      "learning_rate": 6.803425355847897e-05,
+      "loss": 2.4949,
+      "step": 12112
+    },
+    {
+      "epoch": 0.977564361229925,
+      "grad_norm": 0.6433781981468201,
+      "learning_rate": 6.801929534762357e-05,
+      "loss": 2.4937,
+      "step": 12113
+    },
+    {
+      "epoch": 0.9776450649665079,
+      "grad_norm": 0.6935293078422546,
+      "learning_rate": 6.800433793381095e-05,
+      "loss": 2.5025,
+      "step": 12114
+    },
+    {
+      "epoch": 0.9777257687030909,
+      "grad_norm": 0.699780285358429,
+      "learning_rate": 6.798938131741383e-05,
+      "loss": 2.5231,
+      "step": 12115
+    },
+    {
+      "epoch": 0.977806472439674,
+      "grad_norm": 0.6414729952812195,
+      "learning_rate": 6.7974425498805e-05,
+      "loss": 2.4422,
+      "step": 12116
+    },
+    {
+      "epoch": 0.977887176176257,
+      "grad_norm": 0.6733608841896057,
+      "learning_rate": 6.795947047835722e-05,
+      "loss": 2.4873,
+      "step": 12117
+    },
+    {
+      "epoch": 0.97796787991284,
+      "grad_norm": 0.6985765099525452,
+      "learning_rate": 6.794451625644318e-05,
+      "loss": 2.4994,
+      "step": 12118
+    },
+    {
+      "epoch": 0.9780485836494229,
+      "grad_norm": 0.6429893374443054,
+      "learning_rate": 6.792956283343559e-05,
+      "loss": 2.4968,
+      "step": 12119
+    },
+    {
+      "epoch": 0.978129287386006,
+      "grad_norm": 0.7129024267196655,
+      "learning_rate": 6.79146102097071e-05,
+      "loss": 2.5457,
+      "step": 12120
+    },
+    {
+      "epoch": 0.978209991122589,
+      "grad_norm": 0.6811943650245667,
+      "learning_rate": 6.789965838563047e-05,
+      "loss": 2.5012,
+      "step": 12121
+    },
+    {
+      "epoch": 0.978290694859172,
+      "grad_norm": 0.7269948720932007,
+      "learning_rate": 6.788470736157821e-05,
+      "loss": 2.5124,
+      "step": 12122
+    },
+    {
+      "epoch": 0.978371398595755,
+      "grad_norm": 0.7396084666252136,
+      "learning_rate": 6.786975713792299e-05,
+      "loss": 2.5631,
+      "step": 12123
+    },
+    {
+      "epoch": 0.978452102332338,
+      "grad_norm": 0.6880094408988953,
+      "learning_rate": 6.785480771503745e-05,
+      "loss": 2.5103,
+      "step": 12124
+    },
+    {
+      "epoch": 0.978532806068921,
+      "grad_norm": 0.737095057964325,
+      "learning_rate": 6.783985909329409e-05,
+      "loss": 2.5062,
+      "step": 12125
+    },
+    {
+      "epoch": 0.978613509805504,
+      "grad_norm": 0.6540948152542114,
+      "learning_rate": 6.782491127306552e-05,
+      "loss": 2.5568,
+      "step": 12126
+    },
+    {
+      "epoch": 0.978694213542087,
+      "grad_norm": 0.669706404209137,
+      "learning_rate": 6.780996425472427e-05,
+      "loss": 2.5156,
+      "step": 12127
+    },
+    {
+      "epoch": 0.97877491727867,
+      "grad_norm": 0.6722843647003174,
+      "learning_rate": 6.779501803864286e-05,
+      "loss": 2.4784,
+      "step": 12128
+    },
+    {
+      "epoch": 0.978855621015253,
+      "grad_norm": 0.6545475125312805,
+      "learning_rate": 6.778007262519377e-05,
+      "loss": 2.5159,
+      "step": 12129
+    },
+    {
+      "epoch": 0.978936324751836,
+      "grad_norm": 0.7010136246681213,
+      "learning_rate": 6.776512801474953e-05,
+      "loss": 2.5244,
+      "step": 12130
+    },
+    {
+      "epoch": 0.979017028488419,
+      "grad_norm": 0.6912714242935181,
+      "learning_rate": 6.775018420768253e-05,
+      "loss": 2.5223,
+      "step": 12131
+    },
+    {
+      "epoch": 0.9790977322250021,
+      "grad_norm": 0.6864827275276184,
+      "learning_rate": 6.773524120436525e-05,
+      "loss": 2.5027,
+      "step": 12132
+    },
+    {
+      "epoch": 0.979178435961585,
+      "grad_norm": 0.7586981058120728,
+      "learning_rate": 6.77202990051701e-05,
+      "loss": 2.4554,
+      "step": 12133
+    },
+    {
+      "epoch": 0.979259139698168,
+      "grad_norm": 0.6487839818000793,
+      "learning_rate": 6.770535761046948e-05,
+      "loss": 2.5035,
+      "step": 12134
+    },
+    {
+      "epoch": 0.979339843434751,
+      "grad_norm": 0.7193071246147156,
+      "learning_rate": 6.769041702063575e-05,
+      "loss": 2.4669,
+      "step": 12135
+    },
+    {
+      "epoch": 0.9794205471713341,
+      "grad_norm": 0.7118960618972778,
+      "learning_rate": 6.76754772360413e-05,
+      "loss": 2.493,
+      "step": 12136
+    },
+    {
+      "epoch": 0.9795012509079171,
+      "grad_norm": 0.6617394685745239,
+      "learning_rate": 6.766053825705847e-05,
+      "loss": 2.4771,
+      "step": 12137
+    },
+    {
+      "epoch": 0.9795819546445,
+      "grad_norm": 0.7664859294891357,
+      "learning_rate": 6.764560008405953e-05,
+      "loss": 2.5191,
+      "step": 12138
+    },
+    {
+      "epoch": 0.979662658381083,
+      "grad_norm": 0.708063542842865,
+      "learning_rate": 6.763066271741682e-05,
+      "loss": 2.5521,
+      "step": 12139
+    },
+    {
+      "epoch": 0.979743362117666,
+      "grad_norm": 0.6951049566268921,
+      "learning_rate": 6.761572615750267e-05,
+      "loss": 2.4708,
+      "step": 12140
+    },
+    {
+      "epoch": 0.9798240658542491,
+      "grad_norm": 0.6914932727813721,
+      "learning_rate": 6.760079040468921e-05,
+      "loss": 2.5101,
+      "step": 12141
+    },
+    {
+      "epoch": 0.9799047695908321,
+      "grad_norm": 0.6843075752258301,
+      "learning_rate": 6.758585545934876e-05,
+      "loss": 2.4932,
+      "step": 12142
+    },
+    {
+      "epoch": 0.979985473327415,
+      "grad_norm": 0.6567733883857727,
+      "learning_rate": 6.757092132185354e-05,
+      "loss": 2.4577,
+      "step": 12143
+    },
+    {
+      "epoch": 0.980066177063998,
+      "grad_norm": 0.6874415874481201,
+      "learning_rate": 6.75559879925757e-05,
+      "loss": 2.4818,
+      "step": 12144
+    },
+    {
+      "epoch": 0.9801468808005811,
+      "grad_norm": 0.7274627685546875,
+      "learning_rate": 6.754105547188746e-05,
+      "loss": 2.523,
+      "step": 12145
+    },
+    {
+      "epoch": 0.9802275845371641,
+      "grad_norm": 0.6991173028945923,
+      "learning_rate": 6.7526123760161e-05,
+      "loss": 2.4864,
+      "step": 12146
+    },
+    {
+      "epoch": 0.980308288273747,
+      "grad_norm": 0.670078456401825,
+      "learning_rate": 6.75111928577684e-05,
+      "loss": 2.4889,
+      "step": 12147
+    },
+    {
+      "epoch": 0.98038899201033,
+      "grad_norm": 0.6653482913970947,
+      "learning_rate": 6.749626276508178e-05,
+      "loss": 2.4652,
+      "step": 12148
+    },
+    {
+      "epoch": 0.9804696957469131,
+      "grad_norm": 0.7329251766204834,
+      "learning_rate": 6.748133348247326e-05,
+      "loss": 2.518,
+      "step": 12149
+    },
+    {
+      "epoch": 0.9805503994834961,
+      "grad_norm": 0.7792871594429016,
+      "learning_rate": 6.746640501031495e-05,
+      "loss": 2.5018,
+      "step": 12150
+    },
+    {
+      "epoch": 0.9806311032200791,
+      "grad_norm": 0.6962797045707703,
+      "learning_rate": 6.745147734897883e-05,
+      "loss": 2.4388,
+      "step": 12151
+    },
+    {
+      "epoch": 0.980711806956662,
+      "grad_norm": 0.6981272101402283,
+      "learning_rate": 6.7436550498837e-05,
+      "loss": 2.4886,
+      "step": 12152
+    },
+    {
+      "epoch": 0.9807925106932451,
+      "grad_norm": 0.6696565747261047,
+      "learning_rate": 6.742162446026146e-05,
+      "loss": 2.5258,
+      "step": 12153
+    },
+    {
+      "epoch": 0.9808732144298281,
+      "grad_norm": 0.6922139525413513,
+      "learning_rate": 6.740669923362417e-05,
+      "loss": 2.493,
+      "step": 12154
+    },
+    {
+      "epoch": 0.9809539181664111,
+      "grad_norm": 0.6745694875717163,
+      "learning_rate": 6.739177481929715e-05,
+      "loss": 2.5209,
+      "step": 12155
+    },
+    {
+      "epoch": 0.9810346219029941,
+      "grad_norm": 0.7023215889930725,
+      "learning_rate": 6.737685121765238e-05,
+      "loss": 2.4987,
+      "step": 12156
+    },
+    {
+      "epoch": 0.9811153256395772,
+      "grad_norm": 0.6337805390357971,
+      "learning_rate": 6.73619284290617e-05,
+      "loss": 2.4838,
+      "step": 12157
+    },
+    {
+      "epoch": 0.9811960293761601,
+      "grad_norm": 0.6747817397117615,
+      "learning_rate": 6.73470064538971e-05,
+      "loss": 2.4834,
+      "step": 12158
+    },
+    {
+      "epoch": 0.9812767331127431,
+      "grad_norm": 0.6714580655097961,
+      "learning_rate": 6.733208529253047e-05,
+      "loss": 2.4724,
+      "step": 12159
+    },
+    {
+      "epoch": 0.9813574368493261,
+      "grad_norm": 0.6927861571311951,
+      "learning_rate": 6.731716494533364e-05,
+      "loss": 2.495,
+      "step": 12160
+    },
+    {
+      "epoch": 0.9814381405859092,
+      "grad_norm": 0.6576036214828491,
+      "learning_rate": 6.73022454126785e-05,
+      "loss": 2.5415,
+      "step": 12161
+    },
+    {
+      "epoch": 0.9815188443224921,
+      "grad_norm": 0.6495294570922852,
+      "learning_rate": 6.728732669493691e-05,
+      "loss": 2.4889,
+      "step": 12162
+    },
+    {
+      "epoch": 0.9815995480590751,
+      "grad_norm": 0.6680364012718201,
+      "learning_rate": 6.72724087924806e-05,
+      "loss": 2.4733,
+      "step": 12163
+    },
+    {
+      "epoch": 0.9816802517956581,
+      "grad_norm": 0.6816582083702087,
+      "learning_rate": 6.725749170568143e-05,
+      "loss": 2.4688,
+      "step": 12164
+    },
+    {
+      "epoch": 0.9817609555322412,
+      "grad_norm": 0.6995956897735596,
+      "learning_rate": 6.724257543491116e-05,
+      "loss": 2.4962,
+      "step": 12165
+    },
+    {
+      "epoch": 0.9818416592688242,
+      "grad_norm": 0.6728340983390808,
+      "learning_rate": 6.722765998054157e-05,
+      "loss": 2.5218,
+      "step": 12166
+    },
+    {
+      "epoch": 0.9819223630054071,
+      "grad_norm": 0.6835319995880127,
+      "learning_rate": 6.721274534294433e-05,
+      "loss": 2.4845,
+      "step": 12167
+    },
+    {
+      "epoch": 0.9820030667419901,
+      "grad_norm": 0.6969910264015198,
+      "learning_rate": 6.719783152249119e-05,
+      "loss": 2.4983,
+      "step": 12168
+    },
+    {
+      "epoch": 0.9820837704785732,
+      "grad_norm": 0.7327036261558533,
+      "learning_rate": 6.718291851955383e-05,
+      "loss": 2.5893,
+      "step": 12169
+    },
+    {
+      "epoch": 0.9821644742151562,
+      "grad_norm": 0.7092839479446411,
+      "learning_rate": 6.716800633450393e-05,
+      "loss": 2.5104,
+      "step": 12170
+    },
+    {
+      "epoch": 0.9822451779517392,
+      "grad_norm": 0.7384308576583862,
+      "learning_rate": 6.715309496771311e-05,
+      "loss": 2.5066,
+      "step": 12171
+    },
+    {
+      "epoch": 0.9823258816883221,
+      "grad_norm": 0.6744845509529114,
+      "learning_rate": 6.713818441955308e-05,
+      "loss": 2.469,
+      "step": 12172
+    },
+    {
+      "epoch": 0.9824065854249052,
+      "grad_norm": 0.6497980952262878,
+      "learning_rate": 6.712327469039536e-05,
+      "loss": 2.4943,
+      "step": 12173
+    },
+    {
+      "epoch": 0.9824872891614882,
+      "grad_norm": 0.6550357937812805,
+      "learning_rate": 6.710836578061156e-05,
+      "loss": 2.5019,
+      "step": 12174
+    },
+    {
+      "epoch": 0.9825679928980712,
+      "grad_norm": 0.6813549995422363,
+      "learning_rate": 6.709345769057331e-05,
+      "loss": 2.4314,
+      "step": 12175
+    },
+    {
+      "epoch": 0.9826486966346542,
+      "grad_norm": 0.6636531352996826,
+      "learning_rate": 6.707855042065209e-05,
+      "loss": 2.5202,
+      "step": 12176
+    },
+    {
+      "epoch": 0.9827294003712372,
+      "grad_norm": 0.6684894561767578,
+      "learning_rate": 6.706364397121944e-05,
+      "loss": 2.4353,
+      "step": 12177
+    },
+    {
+      "epoch": 0.9828101041078202,
+      "grad_norm": 0.6813677549362183,
+      "learning_rate": 6.704873834264688e-05,
+      "loss": 2.4254,
+      "step": 12178
+    },
+    {
+      "epoch": 0.9828908078444032,
+      "grad_norm": 0.6584975719451904,
+      "learning_rate": 6.70338335353059e-05,
+      "loss": 2.5647,
+      "step": 12179
+    },
+    {
+      "epoch": 0.9829715115809862,
+      "grad_norm": 0.6959114074707031,
+      "learning_rate": 6.701892954956796e-05,
+      "loss": 2.5203,
+      "step": 12180
+    },
+    {
+      "epoch": 0.9830522153175693,
+      "grad_norm": 0.6399044990539551,
+      "learning_rate": 6.700402638580452e-05,
+      "loss": 2.4697,
+      "step": 12181
+    },
+    {
+      "epoch": 0.9831329190541522,
+      "grad_norm": 0.6838750839233398,
+      "learning_rate": 6.698912404438702e-05,
+      "loss": 2.5261,
+      "step": 12182
+    },
+    {
+      "epoch": 0.9832136227907352,
+      "grad_norm": 0.6286367177963257,
+      "learning_rate": 6.697422252568679e-05,
+      "loss": 2.4264,
+      "step": 12183
+    },
+    {
+      "epoch": 0.9832943265273182,
+      "grad_norm": 0.901637852191925,
+      "learning_rate": 6.695932183007528e-05,
+      "loss": 2.4908,
+      "step": 12184
+    },
+    {
+      "epoch": 0.9833750302639013,
+      "grad_norm": 0.8361458778381348,
+      "learning_rate": 6.694442195792386e-05,
+      "loss": 2.5183,
+      "step": 12185
+    },
+    {
+      "epoch": 0.9834557340004842,
+      "grad_norm": 0.7033401727676392,
+      "learning_rate": 6.692952290960384e-05,
+      "loss": 2.5702,
+      "step": 12186
+    },
+    {
+      "epoch": 0.9835364377370672,
+      "grad_norm": 0.669486939907074,
+      "learning_rate": 6.691462468548653e-05,
+      "loss": 2.5143,
+      "step": 12187
+    },
+    {
+      "epoch": 0.9836171414736502,
+      "grad_norm": 0.7043797969818115,
+      "learning_rate": 6.689972728594329e-05,
+      "loss": 2.5638,
+      "step": 12188
+    },
+    {
+      "epoch": 0.9836978452102332,
+      "grad_norm": 0.6532511115074158,
+      "learning_rate": 6.688483071134537e-05,
+      "loss": 2.5227,
+      "step": 12189
+    },
+    {
+      "epoch": 0.9837785489468163,
+      "grad_norm": 0.7363922595977783,
+      "learning_rate": 6.6869934962064e-05,
+      "loss": 2.4953,
+      "step": 12190
+    },
+    {
+      "epoch": 0.9838592526833992,
+      "grad_norm": 0.6746651530265808,
+      "learning_rate": 6.685504003847051e-05,
+      "loss": 2.5021,
+      "step": 12191
+    },
+    {
+      "epoch": 0.9839399564199822,
+      "grad_norm": 0.665459930896759,
+      "learning_rate": 6.684014594093604e-05,
+      "loss": 2.5126,
+      "step": 12192
+    },
+    {
+      "epoch": 0.9840206601565652,
+      "grad_norm": 0.6618975400924683,
+      "learning_rate": 6.682525266983179e-05,
+      "loss": 2.5046,
+      "step": 12193
+    },
+    {
+      "epoch": 0.9841013638931483,
+      "grad_norm": 0.6536173224449158,
+      "learning_rate": 6.6810360225529e-05,
+      "loss": 2.4222,
+      "step": 12194
+    },
+    {
+      "epoch": 0.9841820676297313,
+      "grad_norm": 0.6882187724113464,
+      "learning_rate": 6.679546860839876e-05,
+      "loss": 2.475,
+      "step": 12195
+    },
+    {
+      "epoch": 0.9842627713663142,
+      "grad_norm": 0.6941187977790833,
+      "learning_rate": 6.678057781881224e-05,
+      "loss": 2.5642,
+      "step": 12196
+    },
+    {
+      "epoch": 0.9843434751028972,
+      "grad_norm": 0.7057064175605774,
+      "learning_rate": 6.676568785714057e-05,
+      "loss": 2.4817,
+      "step": 12197
+    },
+    {
+      "epoch": 0.9844241788394803,
+      "grad_norm": 0.6455948352813721,
+      "learning_rate": 6.675079872375487e-05,
+      "loss": 2.5206,
+      "step": 12198
+    },
+    {
+      "epoch": 0.9845048825760633,
+      "grad_norm": 0.6559014320373535,
+      "learning_rate": 6.673591041902613e-05,
+      "loss": 2.4082,
+      "step": 12199
+    },
+    {
+      "epoch": 0.9845855863126463,
+      "grad_norm": 0.6732046008110046,
+      "learning_rate": 6.672102294332542e-05,
+      "loss": 2.5472,
+      "step": 12200
+    },
+    {
+      "epoch": 0.9846662900492292,
+      "grad_norm": 0.7074914574623108,
+      "learning_rate": 6.670613629702391e-05,
+      "loss": 2.5243,
+      "step": 12201
+    },
+    {
+      "epoch": 0.9847469937858123,
+      "grad_norm": 0.6780694127082825,
+      "learning_rate": 6.669125048049246e-05,
+      "loss": 2.494,
+      "step": 12202
+    },
+    {
+      "epoch": 0.9848276975223953,
+      "grad_norm": 0.6361132264137268,
+      "learning_rate": 6.66763654941021e-05,
+      "loss": 2.4764,
+      "step": 12203
+    },
+    {
+      "epoch": 0.9849084012589783,
+      "grad_norm": 0.752727210521698,
+      "learning_rate": 6.666148133822387e-05,
+      "loss": 2.4942,
+      "step": 12204
+    },
+    {
+      "epoch": 0.9849891049955612,
+      "grad_norm": 0.7282724976539612,
+      "learning_rate": 6.664659801322863e-05,
+      "loss": 2.471,
+      "step": 12205
+    },
+    {
+      "epoch": 0.9850698087321443,
+      "grad_norm": 0.6977601051330566,
+      "learning_rate": 6.663171551948736e-05,
+      "loss": 2.4695,
+      "step": 12206
+    },
+    {
+      "epoch": 0.9851505124687273,
+      "grad_norm": 0.6957824230194092,
+      "learning_rate": 6.661683385737101e-05,
+      "loss": 2.5096,
+      "step": 12207
+    },
+    {
+      "epoch": 0.9852312162053103,
+      "grad_norm": 0.6197221279144287,
+      "learning_rate": 6.660195302725037e-05,
+      "loss": 2.4199,
+      "step": 12208
+    },
+    {
+      "epoch": 0.9853119199418933,
+      "grad_norm": 0.747558057308197,
+      "learning_rate": 6.658707302949638e-05,
+      "loss": 2.5988,
+      "step": 12209
+    },
+    {
+      "epoch": 0.9853926236784764,
+      "grad_norm": 0.6593184471130371,
+      "learning_rate": 6.657219386447989e-05,
+      "loss": 2.4837,
+      "step": 12210
+    },
+    {
+      "epoch": 0.9854733274150593,
+      "grad_norm": 0.6795992255210876,
+      "learning_rate": 6.655731553257169e-05,
+      "loss": 2.498,
+      "step": 12211
+    },
+    {
+      "epoch": 0.9855540311516423,
+      "grad_norm": 0.7588422298431396,
+      "learning_rate": 6.65424380341426e-05,
+      "loss": 2.444,
+      "step": 12212
+    },
+    {
+      "epoch": 0.9856347348882253,
+      "grad_norm": 0.7791433930397034,
+      "learning_rate": 6.652756136956342e-05,
+      "loss": 2.4893,
+      "step": 12213
+    },
+    {
+      "epoch": 0.9857154386248084,
+      "grad_norm": 0.6320767998695374,
+      "learning_rate": 6.651268553920493e-05,
+      "loss": 2.4831,
+      "step": 12214
+    },
+    {
+      "epoch": 0.9857961423613913,
+      "grad_norm": 0.6818140745162964,
+      "learning_rate": 6.649781054343783e-05,
+      "loss": 2.4316,
+      "step": 12215
+    },
+    {
+      "epoch": 0.9858768460979743,
+      "grad_norm": 0.7460113763809204,
+      "learning_rate": 6.648293638263285e-05,
+      "loss": 2.5335,
+      "step": 12216
+    },
+    {
+      "epoch": 0.9859575498345573,
+      "grad_norm": 0.714074432849884,
+      "learning_rate": 6.646806305716079e-05,
+      "loss": 2.4573,
+      "step": 12217
+    },
+    {
+      "epoch": 0.9860382535711404,
+      "grad_norm": 0.6815951466560364,
+      "learning_rate": 6.645319056739217e-05,
+      "loss": 2.4758,
+      "step": 12218
+    },
+    {
+      "epoch": 0.9861189573077234,
+      "grad_norm": 0.6842799782752991,
+      "learning_rate": 6.643831891369775e-05,
+      "loss": 2.4998,
+      "step": 12219
+    },
+    {
+      "epoch": 0.9861996610443063,
+      "grad_norm": 0.6725212335586548,
+      "learning_rate": 6.642344809644818e-05,
+      "loss": 2.5179,
+      "step": 12220
+    },
+    {
+      "epoch": 0.9862803647808893,
+      "grad_norm": 0.7859417796134949,
+      "learning_rate": 6.640857811601402e-05,
+      "loss": 2.5801,
+      "step": 12221
+    },
+    {
+      "epoch": 0.9863610685174724,
+      "grad_norm": 0.6438577771186829,
+      "learning_rate": 6.639370897276591e-05,
+      "loss": 2.4659,
+      "step": 12222
+    },
+    {
+      "epoch": 0.9864417722540554,
+      "grad_norm": 0.7036609053611755,
+      "learning_rate": 6.637884066707447e-05,
+      "loss": 2.5637,
+      "step": 12223
+    },
+    {
+      "epoch": 0.9865224759906384,
+      "grad_norm": 0.6756969094276428,
+      "learning_rate": 6.636397319931016e-05,
+      "loss": 2.5381,
+      "step": 12224
+    },
+    {
+      "epoch": 0.9866031797272213,
+      "grad_norm": 0.6907589435577393,
+      "learning_rate": 6.634910656984354e-05,
+      "loss": 2.4927,
+      "step": 12225
+    },
+    {
+      "epoch": 0.9866838834638044,
+      "grad_norm": 0.7347010374069214,
+      "learning_rate": 6.63342407790452e-05,
+      "loss": 2.5131,
+      "step": 12226
+    },
+    {
+      "epoch": 0.9867645872003874,
+      "grad_norm": 0.6835876107215881,
+      "learning_rate": 6.631937582728555e-05,
+      "loss": 2.4611,
+      "step": 12227
+    },
+    {
+      "epoch": 0.9868452909369704,
+      "grad_norm": 0.8199172616004944,
+      "learning_rate": 6.630451171493511e-05,
+      "loss": 2.5341,
+      "step": 12228
+    },
+    {
+      "epoch": 0.9869259946735534,
+      "grad_norm": 0.7537188529968262,
+      "learning_rate": 6.62896484423643e-05,
+      "loss": 2.5218,
+      "step": 12229
+    },
+    {
+      "epoch": 0.9870066984101364,
+      "grad_norm": 0.7254310250282288,
+      "learning_rate": 6.62747860099436e-05,
+      "loss": 2.4766,
+      "step": 12230
+    },
+    {
+      "epoch": 0.9870874021467194,
+      "grad_norm": 0.6852995157241821,
+      "learning_rate": 6.625992441804338e-05,
+      "loss": 2.548,
+      "step": 12231
+    },
+    {
+      "epoch": 0.9871681058833024,
+      "grad_norm": 0.7089388966560364,
+      "learning_rate": 6.624506366703402e-05,
+      "loss": 2.5125,
+      "step": 12232
+    },
+    {
+      "epoch": 0.9872488096198854,
+      "grad_norm": 0.7114216685295105,
+      "learning_rate": 6.623020375728597e-05,
+      "loss": 2.5408,
+      "step": 12233
+    },
+    {
+      "epoch": 0.9873295133564685,
+      "grad_norm": 0.7891978025436401,
+      "learning_rate": 6.621534468916946e-05,
+      "loss": 2.5946,
+      "step": 12234
+    },
+    {
+      "epoch": 0.9874102170930514,
+      "grad_norm": 0.671399712562561,
+      "learning_rate": 6.620048646305488e-05,
+      "loss": 2.4732,
+      "step": 12235
+    },
+    {
+      "epoch": 0.9874909208296344,
+      "grad_norm": 0.6712855696678162,
+      "learning_rate": 6.618562907931256e-05,
+      "loss": 2.4376,
+      "step": 12236
+    },
+    {
+      "epoch": 0.9875716245662174,
+      "grad_norm": 0.7183727025985718,
+      "learning_rate": 6.617077253831272e-05,
+      "loss": 2.5406,
+      "step": 12237
+    },
+    {
+      "epoch": 0.9876523283028005,
+      "grad_norm": 0.6857761144638062,
+      "learning_rate": 6.615591684042568e-05,
+      "loss": 2.5279,
+      "step": 12238
+    },
+    {
+      "epoch": 0.9877330320393835,
+      "grad_norm": 0.7268103957176208,
+      "learning_rate": 6.614106198602165e-05,
+      "loss": 2.5283,
+      "step": 12239
+    },
+    {
+      "epoch": 0.9878137357759664,
+      "grad_norm": 0.6703717708587646,
+      "learning_rate": 6.612620797547087e-05,
+      "loss": 2.4254,
+      "step": 12240
+    },
+    {
+      "epoch": 0.9878944395125494,
+      "grad_norm": 0.7110719680786133,
+      "learning_rate": 6.611135480914352e-05,
+      "loss": 2.496,
+      "step": 12241
+    },
+    {
+      "epoch": 0.9879751432491324,
+      "grad_norm": 0.7268263697624207,
+      "learning_rate": 6.609650248740983e-05,
+      "loss": 2.5489,
+      "step": 12242
+    },
+    {
+      "epoch": 0.9880558469857155,
+      "grad_norm": 0.7413432598114014,
+      "learning_rate": 6.60816510106399e-05,
+      "loss": 2.4998,
+      "step": 12243
+    },
+    {
+      "epoch": 0.9881365507222984,
+      "grad_norm": 0.7443360090255737,
+      "learning_rate": 6.606680037920389e-05,
+      "loss": 2.5282,
+      "step": 12244
+    },
+    {
+      "epoch": 0.9882172544588814,
+      "grad_norm": 0.7787832021713257,
+      "learning_rate": 6.605195059347191e-05,
+      "loss": 2.5221,
+      "step": 12245
+    },
+    {
+      "epoch": 0.9882979581954644,
+      "grad_norm": 0.6921473741531372,
+      "learning_rate": 6.603710165381409e-05,
+      "loss": 2.5434,
+      "step": 12246
+    },
+    {
+      "epoch": 0.9883786619320475,
+      "grad_norm": 0.737328827381134,
+      "learning_rate": 6.602225356060044e-05,
+      "loss": 2.5222,
+      "step": 12247
+    },
+    {
+      "epoch": 0.9884593656686305,
+      "grad_norm": 0.698823094367981,
+      "learning_rate": 6.600740631420106e-05,
+      "loss": 2.528,
+      "step": 12248
+    },
+    {
+      "epoch": 0.9885400694052134,
+      "grad_norm": 0.6735067963600159,
+      "learning_rate": 6.599255991498601e-05,
+      "loss": 2.4942,
+      "step": 12249
+    },
+    {
+      "epoch": 0.9886207731417964,
+      "grad_norm": 0.659622311592102,
+      "learning_rate": 6.59777143633252e-05,
+      "loss": 2.4822,
+      "step": 12250
+    },
+    {
+      "epoch": 0.9887014768783795,
+      "grad_norm": 0.6973726153373718,
+      "learning_rate": 6.596286965958872e-05,
+      "loss": 2.5499,
+      "step": 12251
+    },
+    {
+      "epoch": 0.9887821806149625,
+      "grad_norm": 0.6771909594535828,
+      "learning_rate": 6.594802580414651e-05,
+      "loss": 2.4968,
+      "step": 12252
+    },
+    {
+      "epoch": 0.9888628843515455,
+      "grad_norm": 0.68080073595047,
+      "learning_rate": 6.593318279736849e-05,
+      "loss": 2.5142,
+      "step": 12253
+    },
+    {
+      "epoch": 0.9889435880881284,
+      "grad_norm": NaN,
+      "learning_rate": 6.593318279736849e-05,
+      "loss": 2.466,
+      "step": 12254
+    },
+    {
+      "epoch": 0.9890242918247115,
+      "grad_norm": 0.6865221858024597,
+      "learning_rate": 6.591834063962461e-05,
+      "loss": 2.4894,
+      "step": 12255
+    },
+    {
+      "epoch": 0.9891049955612945,
+      "grad_norm": 0.7050445079803467,
+      "learning_rate": 6.590349933128478e-05,
+      "loss": 2.5733,
+      "step": 12256
+    },
+    {
+      "epoch": 0.9891856992978775,
+      "grad_norm": 0.6971526741981506,
+      "learning_rate": 6.588865887271887e-05,
+      "loss": 2.4997,
+      "step": 12257
+    },
+    {
+      "epoch": 0.9892664030344605,
+      "grad_norm": 0.6465088725090027,
+      "learning_rate": 6.587381926429674e-05,
+      "loss": 2.5155,
+      "step": 12258
+    },
+    {
+      "epoch": 0.9893471067710435,
+      "grad_norm": 0.6521422266960144,
+      "learning_rate": 6.585898050638823e-05,
+      "loss": 2.4803,
+      "step": 12259
+    },
+    {
+      "epoch": 0.9894278105076265,
+      "grad_norm": 0.6798849105834961,
+      "learning_rate": 6.584414259936324e-05,
+      "loss": 2.5301,
+      "step": 12260
+    },
+    {
+      "epoch": 0.9895085142442095,
+      "grad_norm": 0.6903446912765503,
+      "learning_rate": 6.582930554359144e-05,
+      "loss": 2.4662,
+      "step": 12261
+    },
+    {
+      "epoch": 0.9895892179807925,
+      "grad_norm": 0.7183516621589661,
+      "learning_rate": 6.581446933944267e-05,
+      "loss": 2.4711,
+      "step": 12262
+    },
+    {
+      "epoch": 0.9896699217173756,
+      "grad_norm": 0.702738344669342,
+      "learning_rate": 6.579963398728671e-05,
+      "loss": 2.531,
+      "step": 12263
+    },
+    {
+      "epoch": 0.9897506254539585,
+      "grad_norm": 0.7187048196792603,
+      "learning_rate": 6.578479948749325e-05,
+      "loss": 2.4933,
+      "step": 12264
+    },
+    {
+      "epoch": 0.9898313291905415,
+      "grad_norm": 0.6988784670829773,
+      "learning_rate": 6.576996584043202e-05,
+      "loss": 2.5179,
+      "step": 12265
+    },
+    {
+      "epoch": 0.9899120329271245,
+      "grad_norm": 0.7434641122817993,
+      "learning_rate": 6.575513304647276e-05,
+      "loss": 2.5157,
+      "step": 12266
+    },
+    {
+      "epoch": 0.9899927366637076,
+      "grad_norm": 0.667881429195404,
+      "learning_rate": 6.574030110598505e-05,
+      "loss": 2.5152,
+      "step": 12267
+    },
+    {
+      "epoch": 0.9900734404002905,
+      "grad_norm": 0.6766676902770996,
+      "learning_rate": 6.572547001933862e-05,
+      "loss": 2.5041,
+      "step": 12268
+    },
+    {
+      "epoch": 0.9901541441368735,
+      "grad_norm": 0.6531797051429749,
+      "learning_rate": 6.571063978690311e-05,
+      "loss": 2.5457,
+      "step": 12269
+    },
+    {
+      "epoch": 0.9902348478734565,
+      "grad_norm": 0.6557255983352661,
+      "learning_rate": 6.569581040904804e-05,
+      "loss": 2.5253,
+      "step": 12270
+    },
+    {
+      "epoch": 0.9903155516100396,
+      "grad_norm": 0.6818893551826477,
+      "learning_rate": 6.568098188614304e-05,
+      "loss": 2.5031,
+      "step": 12271
+    },
+    {
+      "epoch": 0.9903962553466226,
+      "grad_norm": 0.6644853949546814,
+      "learning_rate": 6.56661542185577e-05,
+      "loss": 2.5285,
+      "step": 12272
+    },
+    {
+      "epoch": 0.9904769590832055,
+      "grad_norm": 0.6035603284835815,
+      "learning_rate": 6.565132740666155e-05,
+      "loss": 2.46,
+      "step": 12273
+    },
+    {
+      "epoch": 0.9905576628197885,
+      "grad_norm": 0.7061343193054199,
+      "learning_rate": 6.56365014508241e-05,
+      "loss": 2.4731,
+      "step": 12274
+    },
+    {
+      "epoch": 0.9906383665563716,
+      "grad_norm": 0.6981248259544373,
+      "learning_rate": 6.562167635141486e-05,
+      "loss": 2.4518,
+      "step": 12275
+    },
+    {
+      "epoch": 0.9907190702929546,
+      "grad_norm": 0.6718073487281799,
+      "learning_rate": 6.560685210880334e-05,
+      "loss": 2.4919,
+      "step": 12276
+    },
+    {
+      "epoch": 0.9907997740295376,
+      "grad_norm": 0.7095392942428589,
+      "learning_rate": 6.559202872335893e-05,
+      "loss": 2.5284,
+      "step": 12277
+    },
+    {
+      "epoch": 0.9908804777661205,
+      "grad_norm": 0.7052092552185059,
+      "learning_rate": 6.557720619545111e-05,
+      "loss": 2.4781,
+      "step": 12278
+    },
+    {
+      "epoch": 0.9909611815027036,
+      "grad_norm": 0.653570830821991,
+      "learning_rate": 6.556238452544934e-05,
+      "loss": 2.5293,
+      "step": 12279
+    },
+    {
+      "epoch": 0.9910418852392866,
+      "grad_norm": 0.6705330610275269,
+      "learning_rate": 6.554756371372293e-05,
+      "loss": 2.4437,
+      "step": 12280
+    },
+    {
+      "epoch": 0.9911225889758696,
+      "grad_norm": 0.6494189500808716,
+      "learning_rate": 6.553274376064127e-05,
+      "loss": 2.4833,
+      "step": 12281
+    },
+    {
+      "epoch": 0.9912032927124526,
+      "grad_norm": 0.6497724652290344,
+      "learning_rate": 6.551792466657378e-05,
+      "loss": 2.4803,
+      "step": 12282
+    },
+    {
+      "epoch": 0.9912839964490356,
+      "grad_norm": 0.7740494608879089,
+      "learning_rate": 6.550310643188972e-05,
+      "loss": 2.4907,
+      "step": 12283
+    },
+    {
+      "epoch": 0.9913647001856186,
+      "grad_norm": 0.699562668800354,
+      "learning_rate": 6.548828905695843e-05,
+      "loss": 2.4576,
+      "step": 12284
+    },
+    {
+      "epoch": 0.9914454039222016,
+      "grad_norm": 0.8123162984848022,
+      "learning_rate": 6.547347254214921e-05,
+      "loss": 2.5118,
+      "step": 12285
+    },
+    {
+      "epoch": 0.9915261076587846,
+      "grad_norm": 0.7227715253829956,
+      "learning_rate": 6.545865688783129e-05,
+      "loss": 2.4688,
+      "step": 12286
+    },
+    {
+      "epoch": 0.9916068113953677,
+      "grad_norm": 0.6498493552207947,
+      "learning_rate": 6.544384209437392e-05,
+      "loss": 2.477,
+      "step": 12287
+    },
+    {
+      "epoch": 0.9916875151319506,
+      "grad_norm": 0.6427823901176453,
+      "learning_rate": 6.542902816214636e-05,
+      "loss": 2.4388,
+      "step": 12288
+    },
+    {
+      "epoch": 0.9917682188685336,
+      "grad_norm": 0.6803679466247559,
+      "learning_rate": 6.541421509151778e-05,
+      "loss": 2.5095,
+      "step": 12289
+    },
+    {
+      "epoch": 0.9918489226051166,
+      "grad_norm": 0.7025790810585022,
+      "learning_rate": 6.539940288285734e-05,
+      "loss": 2.4881,
+      "step": 12290
+    },
+    {
+      "epoch": 0.9919296263416996,
+      "grad_norm": 0.6899270415306091,
+      "learning_rate": 6.538459153653424e-05,
+      "loss": 2.486,
+      "step": 12291
+    },
+    {
+      "epoch": 0.9920103300782827,
+      "grad_norm": 0.7379609942436218,
+      "learning_rate": 6.536978105291762e-05,
+      "loss": 2.5368,
+      "step": 12292
+    },
+    {
+      "epoch": 0.9920910338148656,
+      "grad_norm": 0.7279202342033386,
+      "learning_rate": 6.535497143237657e-05,
+      "loss": 2.5275,
+      "step": 12293
+    },
+    {
+      "epoch": 0.9921717375514486,
+      "grad_norm": 0.6810527443885803,
+      "learning_rate": 6.53401626752802e-05,
+      "loss": 2.5053,
+      "step": 12294
+    },
+    {
+      "epoch": 0.9922524412880316,
+      "grad_norm": 0.6578424572944641,
+      "learning_rate": 6.532535478199759e-05,
+      "loss": 2.5334,
+      "step": 12295
+    },
+    {
+      "epoch": 0.9923331450246147,
+      "grad_norm": 0.6819284558296204,
+      "learning_rate": 6.531054775289778e-05,
+      "loss": 2.4879,
+      "step": 12296
+    },
+    {
+      "epoch": 0.9924138487611976,
+      "grad_norm": 0.6524500846862793,
+      "learning_rate": 6.529574158834977e-05,
+      "loss": 2.5349,
+      "step": 12297
+    },
+    {
+      "epoch": 0.9924945524977806,
+      "grad_norm": 0.6853352785110474,
+      "learning_rate": 6.528093628872263e-05,
+      "loss": 2.4217,
+      "step": 12298
+    },
+    {
+      "epoch": 0.9925752562343636,
+      "grad_norm": 0.6731893420219421,
+      "learning_rate": 6.526613185438529e-05,
+      "loss": 2.4739,
+      "step": 12299
+    },
+    {
+      "epoch": 0.9926559599709467,
+      "grad_norm": 0.6515606641769409,
+      "learning_rate": 6.525132828570673e-05,
+      "loss": 2.5348,
+      "step": 12300
+    },
+    {
+      "epoch": 0.9927366637075297,
+      "grad_norm": 0.6819963455200195,
+      "learning_rate": 6.523652558305596e-05,
+      "loss": 2.5052,
+      "step": 12301
+    },
+    {
+      "epoch": 0.9928173674441126,
+      "grad_norm": 0.6521475911140442,
+      "learning_rate": 6.522172374680177e-05,
+      "loss": 2.5283,
+      "step": 12302
+    },
+    {
+      "epoch": 0.9928980711806956,
+      "grad_norm": 0.6488186717033386,
+      "learning_rate": 6.520692277731315e-05,
+      "loss": 2.4779,
+      "step": 12303
+    },
+    {
+      "epoch": 0.9929787749172787,
+      "grad_norm": 0.6509760022163391,
+      "learning_rate": 6.519212267495903e-05,
+      "loss": 2.5426,
+      "step": 12304
+    },
+    {
+      "epoch": 0.9930594786538617,
+      "grad_norm": 0.621366560459137,
+      "learning_rate": 6.517732344010814e-05,
+      "loss": 2.4804,
+      "step": 12305
+    },
+    {
+      "epoch": 0.9931401823904447,
+      "grad_norm": 0.6907268166542053,
+      "learning_rate": 6.516252507312938e-05,
+      "loss": 2.4883,
+      "step": 12306
+    },
+    {
+      "epoch": 0.9932208861270276,
+      "grad_norm": 0.7739343643188477,
+      "learning_rate": 6.514772757439157e-05,
+      "loss": 2.481,
+      "step": 12307
+    },
+    {
+      "epoch": 0.9933015898636107,
+      "grad_norm": 0.6794601082801819,
+      "learning_rate": 6.513293094426352e-05,
+      "loss": 2.5244,
+      "step": 12308
+    },
+    {
+      "epoch": 0.9933822936001937,
+      "grad_norm": 0.7189902663230896,
+      "learning_rate": 6.511813518311394e-05,
+      "loss": 2.5221,
+      "step": 12309
+    },
+    {
+      "epoch": 0.9934629973367767,
+      "grad_norm": 0.733318030834198,
+      "learning_rate": 6.510334029131163e-05,
+      "loss": 2.521,
+      "step": 12310
+    },
+    {
+      "epoch": 0.9935437010733597,
+      "grad_norm": 0.7584299445152283,
+      "learning_rate": 6.508854626922531e-05,
+      "loss": 2.4962,
+      "step": 12311
+    },
+    {
+      "epoch": 0.9936244048099427,
+      "grad_norm": 0.6442410349845886,
+      "learning_rate": 6.507375311722366e-05,
+      "loss": 2.4775,
+      "step": 12312
+    },
+    {
+      "epoch": 0.9937051085465257,
+      "grad_norm": 0.6609243154525757,
+      "learning_rate": 6.505896083567536e-05,
+      "loss": 2.4706,
+      "step": 12313
+    },
+    {
+      "epoch": 0.9937858122831087,
+      "grad_norm": 0.6527631878852844,
+      "learning_rate": 6.504416942494914e-05,
+      "loss": 2.4612,
+      "step": 12314
+    },
+    {
+      "epoch": 0.9938665160196917,
+      "grad_norm": 0.6798218488693237,
+      "learning_rate": 6.502937888541357e-05,
+      "loss": 2.5502,
+      "step": 12315
+    },
+    {
+      "epoch": 0.9939472197562748,
+      "grad_norm": 0.6573790907859802,
+      "learning_rate": 6.501458921743728e-05,
+      "loss": 2.5598,
+      "step": 12316
+    },
+    {
+      "epoch": 0.9940279234928577,
+      "grad_norm": 0.6945913434028625,
+      "learning_rate": 6.49998004213889e-05,
+      "loss": 2.5323,
+      "step": 12317
+    },
+    {
+      "epoch": 0.9941086272294407,
+      "grad_norm": 0.7609078288078308,
+      "learning_rate": 6.498501249763697e-05,
+      "loss": 2.5211,
+      "step": 12318
+    },
+    {
+      "epoch": 0.9941893309660237,
+      "grad_norm": 0.6878666281700134,
+      "learning_rate": 6.497022544655006e-05,
+      "loss": 2.5366,
+      "step": 12319
+    },
+    {
+      "epoch": 0.9942700347026068,
+      "grad_norm": 0.6675810813903809,
+      "learning_rate": 6.495543926849674e-05,
+      "loss": 2.512,
+      "step": 12320
+    },
+    {
+      "epoch": 0.9943507384391898,
+      "grad_norm": 0.7285950779914856,
+      "learning_rate": 6.494065396384544e-05,
+      "loss": 2.4741,
+      "step": 12321
+    },
+    {
+      "epoch": 0.9944314421757727,
+      "grad_norm": 0.6287158131599426,
+      "learning_rate": 6.49258695329647e-05,
+      "loss": 2.4824,
+      "step": 12322
+    },
+    {
+      "epoch": 0.9945121459123557,
+      "grad_norm": 0.6506727337837219,
+      "learning_rate": 6.491108597622296e-05,
+      "loss": 2.5126,
+      "step": 12323
+    },
+    {
+      "epoch": 0.9945928496489388,
+      "grad_norm": 0.7679052352905273,
+      "learning_rate": 6.489630329398869e-05,
+      "loss": 2.5503,
+      "step": 12324
+    },
+    {
+      "epoch": 0.9946735533855218,
+      "grad_norm": 0.637184202671051,
+      "learning_rate": 6.488152148663029e-05,
+      "loss": 2.5098,
+      "step": 12325
+    },
+    {
+      "epoch": 0.9947542571221047,
+      "grad_norm": 0.6747186779975891,
+      "learning_rate": 6.486674055451619e-05,
+      "loss": 2.5154,
+      "step": 12326
+    },
+    {
+      "epoch": 0.9948349608586877,
+      "grad_norm": 0.7288245558738708,
+      "learning_rate": 6.485196049801476e-05,
+      "loss": 2.5077,
+      "step": 12327
+    },
+    {
+      "epoch": 0.9949156645952708,
+      "grad_norm": 0.6914251446723938,
+      "learning_rate": 6.483718131749435e-05,
+      "loss": 2.4877,
+      "step": 12328
+    },
+    {
+      "epoch": 0.9949963683318538,
+      "grad_norm": 0.7224392294883728,
+      "learning_rate": 6.48224030133233e-05,
+      "loss": 2.4862,
+      "step": 12329
+    },
+    {
+      "epoch": 0.9950770720684368,
+      "grad_norm": 0.7365561723709106,
+      "learning_rate": 6.480762558586995e-05,
+      "loss": 2.477,
+      "step": 12330
+    },
+    {
+      "epoch": 0.9951577758050197,
+      "grad_norm": 0.7673236131668091,
+      "learning_rate": 6.47928490355025e-05,
+      "loss": 2.5423,
+      "step": 12331
+    },
+    {
+      "epoch": 0.9952384795416028,
+      "grad_norm": 0.6638002395629883,
+      "learning_rate": 6.477807336258931e-05,
+      "loss": 2.5007,
+      "step": 12332
+    },
+    {
+      "epoch": 0.9953191832781858,
+      "grad_norm": 0.6415974497795105,
+      "learning_rate": 6.476329856749864e-05,
+      "loss": 2.4924,
+      "step": 12333
+    },
+    {
+      "epoch": 0.9953998870147688,
+      "grad_norm": 0.7129398584365845,
+      "learning_rate": 6.474852465059864e-05,
+      "loss": 2.5313,
+      "step": 12334
+    },
+    {
+      "epoch": 0.9954805907513518,
+      "grad_norm": 0.6896344423294067,
+      "learning_rate": 6.473375161225756e-05,
+      "loss": 2.5073,
+      "step": 12335
+    },
+    {
+      "epoch": 0.9955612944879348,
+      "grad_norm": 0.7009317874908447,
+      "learning_rate": 6.47189794528436e-05,
+      "loss": 2.574,
+      "step": 12336
+    },
+    {
+      "epoch": 0.9956419982245178,
+      "grad_norm": 0.6555172801017761,
+      "learning_rate": 6.470420817272488e-05,
+      "loss": 2.4769,
+      "step": 12337
+    },
+    {
+      "epoch": 0.9957227019611008,
+      "grad_norm": 0.7569532990455627,
+      "learning_rate": 6.468943777226954e-05,
+      "loss": 2.4691,
+      "step": 12338
+    },
+    {
+      "epoch": 0.9958034056976838,
+      "grad_norm": 0.68092280626297,
+      "learning_rate": 6.467466825184569e-05,
+      "loss": 2.4793,
+      "step": 12339
+    },
+    {
+      "epoch": 0.9958841094342669,
+      "grad_norm": 0.6977378726005554,
+      "learning_rate": 6.465989961182152e-05,
+      "loss": 2.4678,
+      "step": 12340
+    },
+    {
+      "epoch": 0.9959648131708498,
+      "grad_norm": 0.6702281832695007,
+      "learning_rate": 6.4645131852565e-05,
+      "loss": 2.5398,
+      "step": 12341
+    },
+    {
+      "epoch": 0.9960455169074328,
+      "grad_norm": 0.7584038972854614,
+      "learning_rate": 6.46303649744442e-05,
+      "loss": 2.5355,
+      "step": 12342
+    },
+    {
+      "epoch": 0.9961262206440158,
+      "grad_norm": 0.6779505610466003,
+      "learning_rate": 6.461559897782718e-05,
+      "loss": 2.4828,
+      "step": 12343
+    },
+    {
+      "epoch": 0.9962069243805988,
+      "grad_norm": 0.6968233585357666,
+      "learning_rate": 6.460083386308192e-05,
+      "loss": 2.5108,
+      "step": 12344
+    },
+    {
+      "epoch": 0.9962876281171819,
+      "grad_norm": 0.7114594578742981,
+      "learning_rate": 6.45860696305764e-05,
+      "loss": 2.5236,
+      "step": 12345
+    },
+    {
+      "epoch": 0.9963683318537648,
+      "grad_norm": 0.6850530505180359,
+      "learning_rate": 6.457130628067865e-05,
+      "loss": 2.458,
+      "step": 12346
+    },
+    {
+      "epoch": 0.9964490355903478,
+      "grad_norm": 0.7135400772094727,
+      "learning_rate": 6.455654381375651e-05,
+      "loss": 2.539,
+      "step": 12347
+    },
+    {
+      "epoch": 0.9965297393269308,
+      "grad_norm": 0.6736366748809814,
+      "learning_rate": 6.454178223017797e-05,
+      "loss": 2.4721,
+      "step": 12348
+    },
+    {
+      "epoch": 0.9966104430635139,
+      "grad_norm": 0.6806206107139587,
+      "learning_rate": 6.45270215303109e-05,
+      "loss": 2.5035,
+      "step": 12349
+    },
+    {
+      "epoch": 0.9966911468000968,
+      "grad_norm": 0.7120711803436279,
+      "learning_rate": 6.451226171452318e-05,
+      "loss": 2.5344,
+      "step": 12350
+    },
+    {
+      "epoch": 0.9967718505366798,
+      "grad_norm": 0.6865986585617065,
+      "learning_rate": 6.449750278318264e-05,
+      "loss": 2.4807,
+      "step": 12351
+    },
+    {
+      "epoch": 0.9968525542732628,
+      "grad_norm": 0.6461294889450073,
+      "learning_rate": 6.448274473665717e-05,
+      "loss": 2.4878,
+      "step": 12352
+    },
+    {
+      "epoch": 0.9969332580098459,
+      "grad_norm": 0.7090638279914856,
+      "learning_rate": 6.446798757531454e-05,
+      "loss": 2.4599,
+      "step": 12353
+    },
+    {
+      "epoch": 0.9970139617464289,
+      "grad_norm": 0.6933324337005615,
+      "learning_rate": 6.445323129952252e-05,
+      "loss": 2.5398,
+      "step": 12354
+    },
+    {
+      "epoch": 0.9970946654830118,
+      "grad_norm": 0.7018197774887085,
+      "learning_rate": 6.443847590964888e-05,
+      "loss": 2.5159,
+      "step": 12355
+    },
+    {
+      "epoch": 0.9971753692195948,
+      "grad_norm": 0.7292604446411133,
+      "learning_rate": 6.442372140606145e-05,
+      "loss": 2.4934,
+      "step": 12356
+    },
+    {
+      "epoch": 0.9972560729561779,
+      "grad_norm": 0.6686378121376038,
+      "learning_rate": 6.440896778912783e-05,
+      "loss": 2.5076,
+      "step": 12357
+    },
+    {
+      "epoch": 0.9973367766927609,
+      "grad_norm": 0.7194764018058777,
+      "learning_rate": 6.439421505921576e-05,
+      "loss": 2.4958,
+      "step": 12358
+    },
+    {
+      "epoch": 0.9974174804293439,
+      "grad_norm": 0.662467360496521,
+      "learning_rate": 6.437946321669296e-05,
+      "loss": 2.5202,
+      "step": 12359
+    },
+    {
+      "epoch": 0.9974981841659268,
+      "grad_norm": 0.7222515940666199,
+      "learning_rate": 6.436471226192703e-05,
+      "loss": 2.5058,
+      "step": 12360
+    },
+    {
+      "epoch": 0.9975788879025099,
+      "grad_norm": 0.6354855895042419,
+      "learning_rate": 6.434996219528562e-05,
+      "loss": 2.4849,
+      "step": 12361
+    },
+    {
+      "epoch": 0.9976595916390929,
+      "grad_norm": 0.7689539790153503,
+      "learning_rate": 6.433521301713636e-05,
+      "loss": 2.4959,
+      "step": 12362
+    },
+    {
+      "epoch": 0.9977402953756759,
+      "grad_norm": 0.6894338130950928,
+      "learning_rate": 6.43204647278468e-05,
+      "loss": 2.5098,
+      "step": 12363
+    },
+    {
+      "epoch": 0.9978209991122589,
+      "grad_norm": 0.7694165110588074,
+      "learning_rate": 6.430571732778451e-05,
+      "loss": 2.513,
+      "step": 12364
+    },
+    {
+      "epoch": 0.9979017028488419,
+      "grad_norm": 0.6512044668197632,
+      "learning_rate": 6.42909708173171e-05,
+      "loss": 2.4785,
+      "step": 12365
+    },
+    {
+      "epoch": 0.9979824065854249,
+      "grad_norm": 0.6605672836303711,
+      "learning_rate": 6.427622519681201e-05,
+      "loss": 2.4804,
+      "step": 12366
+    },
+    {
+      "epoch": 0.9980631103220079,
+      "grad_norm": 0.7123624086380005,
+      "learning_rate": 6.426148046663677e-05,
+      "loss": 2.4854,
+      "step": 12367
+    },
+    {
+      "epoch": 0.9981438140585909,
+      "grad_norm": 0.662645697593689,
+      "learning_rate": 6.424673662715886e-05,
+      "loss": 2.5314,
+      "step": 12368
+    },
+    {
+      "epoch": 0.998224517795174,
+      "grad_norm": 0.6482149362564087,
+      "learning_rate": 6.423199367874573e-05,
+      "loss": 2.4492,
+      "step": 12369
+    },
+    {
+      "epoch": 0.9983052215317569,
+      "grad_norm": 0.6545752286911011,
+      "learning_rate": 6.421725162176482e-05,
+      "loss": 2.5042,
+      "step": 12370
+    },
+    {
+      "epoch": 0.9983859252683399,
+      "grad_norm": 0.6698874235153198,
+      "learning_rate": 6.420251045658353e-05,
+      "loss": 2.4523,
+      "step": 12371
+    },
+    {
+      "epoch": 0.9984666290049229,
+      "grad_norm": 0.6961477398872375,
+      "learning_rate": 6.418777018356929e-05,
+      "loss": 2.556,
+      "step": 12372
+    },
+    {
+      "epoch": 0.998547332741506,
+      "grad_norm": 0.67090904712677,
+      "learning_rate": 6.41730308030894e-05,
+      "loss": 2.5237,
+      "step": 12373
+    },
+    {
+      "epoch": 0.998628036478089,
+      "grad_norm": 0.6828685402870178,
+      "learning_rate": 6.415829231551124e-05,
+      "loss": 2.453,
+      "step": 12374
+    },
+    {
+      "epoch": 0.9987087402146719,
+      "grad_norm": 0.6699565649032593,
+      "learning_rate": 6.414355472120213e-05,
+      "loss": 2.4632,
+      "step": 12375
+    },
+    {
+      "epoch": 0.9987894439512549,
+      "grad_norm": 0.6918730735778809,
+      "learning_rate": 6.412881802052936e-05,
+      "loss": 2.4532,
+      "step": 12376
+    },
+    {
+      "epoch": 0.998870147687838,
+      "grad_norm": 0.7222442030906677,
+      "learning_rate": 6.411408221386021e-05,
+      "loss": 2.5113,
+      "step": 12377
+    },
+    {
+      "epoch": 0.998950851424421,
+      "grad_norm": 0.7479627132415771,
+      "learning_rate": 6.409934730156195e-05,
+      "loss": 2.4857,
+      "step": 12378
+    },
+    {
+      "epoch": 0.999031555161004,
+      "grad_norm": 0.6552882194519043,
+      "learning_rate": 6.40846132840018e-05,
+      "loss": 2.4816,
+      "step": 12379
+    },
+    {
+      "epoch": 0.9991122588975869,
+      "grad_norm": 0.5990073084831238,
+      "learning_rate": 6.406988016154694e-05,
+      "loss": 2.4753,
+      "step": 12380
+    },
+    {
+      "epoch": 0.99919296263417,
+      "grad_norm": 0.6671901941299438,
+      "learning_rate": 6.405514793456465e-05,
+      "loss": 2.5298,
+      "step": 12381
+    },
+    {
+      "epoch": 0.999273666370753,
+      "grad_norm": 0.6630427241325378,
+      "learning_rate": 6.4040416603422e-05,
+      "loss": 2.485,
+      "step": 12382
+    },
+    {
+      "epoch": 0.999354370107336,
+      "grad_norm": 0.6873636841773987,
+      "learning_rate": 6.402568616848614e-05,
+      "loss": 2.4902,
+      "step": 12383
+    },
+    {
+      "epoch": 0.9994350738439189,
+      "grad_norm": 0.6912413239479065,
+      "learning_rate": 6.401095663012424e-05,
+      "loss": 2.5339,
+      "step": 12384
+    },
+    {
+      "epoch": 0.999515777580502,
+      "grad_norm": 0.6491912603378296,
+      "learning_rate": 6.39962279887034e-05,
+      "loss": 2.5367,
+      "step": 12385
+    },
+    {
+      "epoch": 0.999596481317085,
+      "grad_norm": 0.6668288111686707,
+      "learning_rate": 6.398150024459065e-05,
+      "loss": 2.5294,
+      "step": 12386
+    },
+    {
+      "epoch": 0.999677185053668,
+      "grad_norm": 0.6603856086730957,
+      "learning_rate": 6.396677339815306e-05,
+      "loss": 2.4378,
+      "step": 12387
+    },
+    {
+      "epoch": 0.999757888790251,
+      "grad_norm": 0.6461218595504761,
+      "learning_rate": 6.395204744975772e-05,
+      "loss": 2.4835,
+      "step": 12388
+    },
+    {
+      "epoch": 0.999838592526834,
+      "grad_norm": 0.6621688604354858,
+      "learning_rate": 6.39373223997715e-05,
+      "loss": 2.4834,
+      "step": 12389
+    },
+    {
+      "epoch": 0.999919296263417,
+      "grad_norm": 0.6758724451065063,
+      "learning_rate": 6.392259824856153e-05,
+      "loss": 2.4549,
+      "step": 12390
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1304112672805786,
+      "learning_rate": 6.390787499649473e-05,
+      "loss": 2.5547,
+      "step": 12391
+    },
+    {
+      "epoch": 1.000080703736583,
+      "grad_norm": 0.6919478178024292,
+      "learning_rate": 6.389315264393801e-05,
+      "loss": 2.47,
+      "step": 12392
+    },
+    {
+      "epoch": 1.000161407473166,
+      "grad_norm": 0.6916815638542175,
+      "learning_rate": 6.38784311912583e-05,
+      "loss": 2.4636,
+      "step": 12393
+    },
+    {
+      "epoch": 1.000242111209749,
+      "grad_norm": 0.6627040505409241,
+      "learning_rate": 6.386371063882252e-05,
+      "loss": 2.5094,
+      "step": 12394
+    },
+    {
+      "epoch": 1.000322814946332,
+      "grad_norm": 0.6408648490905762,
+      "learning_rate": 6.384899098699754e-05,
+      "loss": 2.426,
+      "step": 12395
+    },
+    {
+      "epoch": 1.000403518682915,
+      "grad_norm": 0.70432448387146,
+      "learning_rate": 6.38342722361502e-05,
+      "loss": 2.4861,
+      "step": 12396
+    },
+    {
+      "epoch": 1.000484222419498,
+      "grad_norm": 0.7115964889526367,
+      "learning_rate": 6.381955438664735e-05,
+      "loss": 2.4824,
+      "step": 12397
+    },
+    {
+      "epoch": 1.000564926156081,
+      "grad_norm": 0.6547040939331055,
+      "learning_rate": 6.380483743885574e-05,
+      "loss": 2.488,
+      "step": 12398
+    },
+    {
+      "epoch": 1.000645629892664,
+      "grad_norm": 0.6916625499725342,
+      "learning_rate": 6.379012139314223e-05,
+      "loss": 2.4864,
+      "step": 12399
+    },
+    {
+      "epoch": 1.0007263336292471,
+      "grad_norm": 0.6311133503913879,
+      "learning_rate": 6.377540624987352e-05,
+      "loss": 2.4672,
+      "step": 12400
+    },
+    {
+      "epoch": 1.00080703736583,
+      "grad_norm": 0.7115580439567566,
+      "learning_rate": 6.376069200941642e-05,
+      "loss": 2.4359,
+      "step": 12401
+    },
+    {
+      "epoch": 1.000887741102413,
+      "grad_norm": 0.6734051704406738,
+      "learning_rate": 6.374597867213756e-05,
+      "loss": 2.4896,
+      "step": 12402
+    },
+    {
+      "epoch": 1.000968444838996,
+      "grad_norm": 0.6910715699195862,
+      "learning_rate": 6.373126623840368e-05,
+      "loss": 2.4502,
+      "step": 12403
+    },
+    {
+      "epoch": 1.001049148575579,
+      "grad_norm": 0.6807514429092407,
+      "learning_rate": 6.37165547085815e-05,
+      "loss": 2.4791,
+      "step": 12404
+    },
+    {
+      "epoch": 1.0011298523121621,
+      "grad_norm": 0.679350733757019,
+      "learning_rate": 6.370184408303759e-05,
+      "loss": 2.4758,
+      "step": 12405
+    },
+    {
+      "epoch": 1.001210556048745,
+      "grad_norm": 0.6516300439834595,
+      "learning_rate": 6.36871343621386e-05,
+      "loss": 2.4338,
+      "step": 12406
+    },
+    {
+      "epoch": 1.001291259785328,
+      "grad_norm": 0.7033620476722717,
+      "learning_rate": 6.367242554625119e-05,
+      "loss": 2.429,
+      "step": 12407
+    },
+    {
+      "epoch": 1.0013719635219112,
+      "grad_norm": 0.6750274896621704,
+      "learning_rate": 6.365771763574186e-05,
+      "loss": 2.4283,
+      "step": 12408
+    },
+    {
+      "epoch": 1.001452667258494,
+      "grad_norm": 0.7188721895217896,
+      "learning_rate": 6.364301063097722e-05,
+      "loss": 2.4509,
+      "step": 12409
+    },
+    {
+      "epoch": 1.001533370995077,
+      "grad_norm": 0.6936308741569519,
+      "learning_rate": 6.362830453232379e-05,
+      "loss": 2.4469,
+      "step": 12410
+    },
+    {
+      "epoch": 1.00161407473166,
+      "grad_norm": 0.673060953617096,
+      "learning_rate": 6.361359934014808e-05,
+      "loss": 2.4444,
+      "step": 12411
+    },
+    {
+      "epoch": 1.001694778468243,
+      "grad_norm": 0.7465113997459412,
+      "learning_rate": 6.359889505481658e-05,
+      "loss": 2.4376,
+      "step": 12412
+    },
+    {
+      "epoch": 1.0017754822048262,
+      "grad_norm": 0.7180366516113281,
+      "learning_rate": 6.358419167669582e-05,
+      "loss": 2.4223,
+      "step": 12413
+    },
+    {
+      "epoch": 1.001856185941409,
+      "grad_norm": 0.6582302451133728,
+      "learning_rate": 6.356948920615214e-05,
+      "loss": 2.4723,
+      "step": 12414
+    },
+    {
+      "epoch": 1.001936889677992,
+      "grad_norm": 0.6452654600143433,
+      "learning_rate": 6.3554787643552e-05,
+      "loss": 2.4609,
+      "step": 12415
+    },
+    {
+      "epoch": 1.0020175934145752,
+      "grad_norm": 0.7170321345329285,
+      "learning_rate": 6.354008698926185e-05,
+      "loss": 2.5377,
+      "step": 12416
+    },
+    {
+      "epoch": 1.002098297151158,
+      "grad_norm": 0.6483680605888367,
+      "learning_rate": 6.352538724364809e-05,
+      "loss": 2.4349,
+      "step": 12417
+    },
+    {
+      "epoch": 1.0021790008877411,
+      "grad_norm": 0.6567494869232178,
+      "learning_rate": 6.351068840707697e-05,
+      "loss": 2.4421,
+      "step": 12418
+    },
+    {
+      "epoch": 1.002259704624324,
+      "grad_norm": 0.7498565912246704,
+      "learning_rate": 6.349599047991488e-05,
+      "loss": 2.4212,
+      "step": 12419
+    },
+    {
+      "epoch": 1.002340408360907,
+      "grad_norm": 0.6894906759262085,
+      "learning_rate": 6.348129346252816e-05,
+      "loss": 2.4356,
+      "step": 12420
+    },
+    {
+      "epoch": 1.0024211120974902,
+      "grad_norm": 0.657361626625061,
+      "learning_rate": 6.346659735528304e-05,
+      "loss": 2.4164,
+      "step": 12421
+    },
+    {
+      "epoch": 1.002501815834073,
+      "grad_norm": 0.6369211673736572,
+      "learning_rate": 6.345190215854581e-05,
+      "loss": 2.4229,
+      "step": 12422
+    },
+    {
+      "epoch": 1.0025825195706561,
+      "grad_norm": 0.7033721208572388,
+      "learning_rate": 6.343720787268277e-05,
+      "loss": 2.5052,
+      "step": 12423
+    },
+    {
+      "epoch": 1.0026632233072392,
+      "grad_norm": 0.7125518918037415,
+      "learning_rate": 6.342251449806003e-05,
+      "loss": 2.514,
+      "step": 12424
+    },
+    {
+      "epoch": 1.002743927043822,
+      "grad_norm": 0.7355595827102661,
+      "learning_rate": 6.340782203504385e-05,
+      "loss": 2.4459,
+      "step": 12425
+    },
+    {
+      "epoch": 1.0028246307804052,
+      "grad_norm": 0.7244594693183899,
+      "learning_rate": 6.339313048400042e-05,
+      "loss": 2.452,
+      "step": 12426
+    },
+    {
+      "epoch": 1.002905334516988,
+      "grad_norm": 0.7112728357315063,
+      "learning_rate": 6.337843984529585e-05,
+      "loss": 2.4951,
+      "step": 12427
+    },
+    {
+      "epoch": 1.0029860382535711,
+      "grad_norm": 0.7235615849494934,
+      "learning_rate": 6.336375011929628e-05,
+      "loss": 2.4697,
+      "step": 12428
+    },
+    {
+      "epoch": 1.0030667419901542,
+      "grad_norm": 0.653865396976471,
+      "learning_rate": 6.334906130636784e-05,
+      "loss": 2.4804,
+      "step": 12429
+    },
+    {
+      "epoch": 1.003147445726737,
+      "grad_norm": 0.7845149636268616,
+      "learning_rate": 6.33343734068766e-05,
+      "loss": 2.5415,
+      "step": 12430
+    },
+    {
+      "epoch": 1.0032281494633202,
+      "grad_norm": 0.7356342077255249,
+      "learning_rate": 6.33196864211886e-05,
+      "loss": 2.5321,
+      "step": 12431
+    },
+    {
+      "epoch": 1.0033088531999033,
+      "grad_norm": 0.6828265190124512,
+      "learning_rate": 6.330500034966991e-05,
+      "loss": 2.3849,
+      "step": 12432
+    },
+    {
+      "epoch": 1.0033895569364861,
+      "grad_norm": 0.7226579189300537,
+      "learning_rate": 6.329031519268658e-05,
+      "loss": 2.512,
+      "step": 12433
+    },
+    {
+      "epoch": 1.0034702606730692,
+      "grad_norm": 0.6490235924720764,
+      "learning_rate": 6.327563095060449e-05,
+      "loss": 2.487,
+      "step": 12434
+    },
+    {
+      "epoch": 1.003550964409652,
+      "grad_norm": 0.6889309883117676,
+      "learning_rate": 6.326094762378969e-05,
+      "loss": 2.4677,
+      "step": 12435
+    },
+    {
+      "epoch": 1.0036316681462352,
+      "grad_norm": 0.695854127407074,
+      "learning_rate": 6.324626521260815e-05,
+      "loss": 2.4362,
+      "step": 12436
+    },
+    {
+      "epoch": 1.0037123718828183,
+      "grad_norm": 0.7045256495475769,
+      "learning_rate": 6.32315837174257e-05,
+      "loss": 2.4307,
+      "step": 12437
+    },
+    {
+      "epoch": 1.0037930756194011,
+      "grad_norm": 0.662604570388794,
+      "learning_rate": 6.321690313860833e-05,
+      "loss": 2.4271,
+      "step": 12438
+    },
+    {
+      "epoch": 1.0038737793559842,
+      "grad_norm": 0.7682240009307861,
+      "learning_rate": 6.320222347652191e-05,
+      "loss": 2.4617,
+      "step": 12439
+    },
+    {
+      "epoch": 1.0039544830925673,
+      "grad_norm": 0.6599584817886353,
+      "learning_rate": 6.318754473153221e-05,
+      "loss": 2.405,
+      "step": 12440
+    },
+    {
+      "epoch": 1.0040351868291502,
+      "grad_norm": 0.7423116564750671,
+      "learning_rate": 6.317286690400515e-05,
+      "loss": 2.5496,
+      "step": 12441
+    },
+    {
+      "epoch": 1.0041158905657332,
+      "grad_norm": 0.6928953528404236,
+      "learning_rate": 6.315818999430654e-05,
+      "loss": 2.4265,
+      "step": 12442
+    },
+    {
+      "epoch": 1.0041965943023161,
+      "grad_norm": 0.699990451335907,
+      "learning_rate": 6.314351400280211e-05,
+      "loss": 2.4747,
+      "step": 12443
+    },
+    {
+      "epoch": 1.0042772980388992,
+      "grad_norm": 0.673384964466095,
+      "learning_rate": 6.312883892985765e-05,
+      "loss": 2.4891,
+      "step": 12444
+    },
+    {
+      "epoch": 1.0043580017754823,
+      "grad_norm": 0.6668596863746643,
+      "learning_rate": 6.311416477583893e-05,
+      "loss": 2.4312,
+      "step": 12445
+    },
+    {
+      "epoch": 1.0044387055120652,
+      "grad_norm": 0.6931218504905701,
+      "learning_rate": 6.309949154111163e-05,
+      "loss": 2.4907,
+      "step": 12446
+    },
+    {
+      "epoch": 1.0045194092486482,
+      "grad_norm": 0.687683641910553,
+      "learning_rate": 6.308481922604146e-05,
+      "loss": 2.4302,
+      "step": 12447
+    },
+    {
+      "epoch": 1.004600112985231,
+      "grad_norm": 0.6887302398681641,
+      "learning_rate": 6.30701478309941e-05,
+      "loss": 2.4749,
+      "step": 12448
+    },
+    {
+      "epoch": 1.0046808167218142,
+      "grad_norm": 0.6713404655456543,
+      "learning_rate": 6.305547735633522e-05,
+      "loss": 2.5046,
+      "step": 12449
+    },
+    {
+      "epoch": 1.0047615204583973,
+      "grad_norm": 0.7147336006164551,
+      "learning_rate": 6.304080780243038e-05,
+      "loss": 2.4578,
+      "step": 12450
+    },
+    {
+      "epoch": 1.0048422241949801,
+      "grad_norm": 0.87425297498703,
+      "learning_rate": 6.30261391696452e-05,
+      "loss": 2.4487,
+      "step": 12451
+    },
+    {
+      "epoch": 1.0049229279315632,
+      "grad_norm": 0.6641440987586975,
+      "learning_rate": 6.301147145834534e-05,
+      "loss": 2.4657,
+      "step": 12452
+    },
+    {
+      "epoch": 1.0050036316681463,
+      "grad_norm": 0.7311998009681702,
+      "learning_rate": 6.299680466889626e-05,
+      "loss": 2.4784,
+      "step": 12453
+    },
+    {
+      "epoch": 1.0050843354047292,
+      "grad_norm": 0.6722697615623474,
+      "learning_rate": 6.298213880166354e-05,
+      "loss": 2.4653,
+      "step": 12454
+    },
+    {
+      "epoch": 1.0051650391413123,
+      "grad_norm": 0.6886328458786011,
+      "learning_rate": 6.29674738570127e-05,
+      "loss": 2.3949,
+      "step": 12455
+    },
+    {
+      "epoch": 1.0052457428778951,
+      "grad_norm": 0.684688925743103,
+      "learning_rate": 6.295280983530921e-05,
+      "loss": 2.4334,
+      "step": 12456
+    },
+    {
+      "epoch": 1.0053264466144782,
+      "grad_norm": 0.7436798214912415,
+      "learning_rate": 6.293814673691853e-05,
+      "loss": 2.5316,
+      "step": 12457
+    },
+    {
+      "epoch": 1.0054071503510613,
+      "grad_norm": 0.7401304244995117,
+      "learning_rate": 6.292348456220615e-05,
+      "loss": 2.4556,
+      "step": 12458
+    },
+    {
+      "epoch": 1.0054878540876442,
+      "grad_norm": 0.7330329418182373,
+      "learning_rate": 6.290882331153742e-05,
+      "loss": 2.4321,
+      "step": 12459
+    },
+    {
+      "epoch": 1.0055685578242273,
+      "grad_norm": 0.8005052208900452,
+      "learning_rate": 6.289416298527776e-05,
+      "loss": 2.415,
+      "step": 12460
+    },
+    {
+      "epoch": 1.0056492615608104,
+      "grad_norm": 0.8047310709953308,
+      "learning_rate": 6.28795035837926e-05,
+      "loss": 2.4144,
+      "step": 12461
+    },
+    {
+      "epoch": 1.0057299652973932,
+      "grad_norm": 0.7384032011032104,
+      "learning_rate": 6.28648451074472e-05,
+      "loss": 2.5237,
+      "step": 12462
+    },
+    {
+      "epoch": 1.0058106690339763,
+      "grad_norm": 0.7240314483642578,
+      "learning_rate": 6.285018755660695e-05,
+      "loss": 2.4894,
+      "step": 12463
+    },
+    {
+      "epoch": 1.0058913727705592,
+      "grad_norm": 0.6901080012321472,
+      "learning_rate": 6.283553093163712e-05,
+      "loss": 2.4244,
+      "step": 12464
+    },
+    {
+      "epoch": 1.0059720765071423,
+      "grad_norm": 0.6572268605232239,
+      "learning_rate": 6.282087523290304e-05,
+      "loss": 2.456,
+      "step": 12465
+    },
+    {
+      "epoch": 1.0060527802437254,
+      "grad_norm": 0.7207481861114502,
+      "learning_rate": 6.28062204607699e-05,
+      "loss": 2.4153,
+      "step": 12466
+    },
+    {
+      "epoch": 1.0061334839803082,
+      "grad_norm": 0.6901980042457581,
+      "learning_rate": 6.279156661560299e-05,
+      "loss": 2.4776,
+      "step": 12467
+    },
+    {
+      "epoch": 1.0062141877168913,
+      "grad_norm": 0.7003545761108398,
+      "learning_rate": 6.277691369776752e-05,
+      "loss": 2.4206,
+      "step": 12468
+    },
+    {
+      "epoch": 1.0062948914534744,
+      "grad_norm": 0.6978366374969482,
+      "learning_rate": 6.276226170762865e-05,
+      "loss": 2.3866,
+      "step": 12469
+    },
+    {
+      "epoch": 1.0063755951900573,
+      "grad_norm": 0.6763097643852234,
+      "learning_rate": 6.274761064555154e-05,
+      "loss": 2.5439,
+      "step": 12470
+    },
+    {
+      "epoch": 1.0064562989266403,
+      "grad_norm": 0.7146836519241333,
+      "learning_rate": 6.273296051190139e-05,
+      "loss": 2.5486,
+      "step": 12471
+    },
+    {
+      "epoch": 1.0065370026632232,
+      "grad_norm": 0.7448136806488037,
+      "learning_rate": 6.271831130704326e-05,
+      "loss": 2.4539,
+      "step": 12472
+    },
+    {
+      "epoch": 1.0066177063998063,
+      "grad_norm": 0.6918472051620483,
+      "learning_rate": 6.270366303134226e-05,
+      "loss": 2.4756,
+      "step": 12473
+    },
+    {
+      "epoch": 1.0066984101363894,
+      "grad_norm": 0.7067514657974243,
+      "learning_rate": 6.26890156851635e-05,
+      "loss": 2.4925,
+      "step": 12474
+    },
+    {
+      "epoch": 1.0067791138729723,
+      "grad_norm": 0.6517517566680908,
+      "learning_rate": 6.267436926887197e-05,
+      "loss": 2.4339,
+      "step": 12475
+    },
+    {
+      "epoch": 1.0068598176095553,
+      "grad_norm": 0.673367977142334,
+      "learning_rate": 6.265972378283274e-05,
+      "loss": 2.416,
+      "step": 12476
+    },
+    {
+      "epoch": 1.0069405213461384,
+      "grad_norm": 0.7190212607383728,
+      "learning_rate": 6.26450792274108e-05,
+      "loss": 2.4822,
+      "step": 12477
+    },
+    {
+      "epoch": 1.0070212250827213,
+      "grad_norm": 0.7568029165267944,
+      "learning_rate": 6.263043560297112e-05,
+      "loss": 2.4607,
+      "step": 12478
+    },
+    {
+      "epoch": 1.0071019288193044,
+      "grad_norm": 0.6860609650611877,
+      "learning_rate": 6.261579290987866e-05,
+      "loss": 2.4429,
+      "step": 12479
+    },
+    {
+      "epoch": 1.0071826325558872,
+      "grad_norm": 0.7066059112548828,
+      "learning_rate": 6.260115114849839e-05,
+      "loss": 2.5504,
+      "step": 12480
+    },
+    {
+      "epoch": 1.0072633362924703,
+      "grad_norm": 0.6857946515083313,
+      "learning_rate": 6.25865103191952e-05,
+      "loss": 2.4776,
+      "step": 12481
+    },
+    {
+      "epoch": 1.0073440400290534,
+      "grad_norm": 0.6879859566688538,
+      "learning_rate": 6.257187042233396e-05,
+      "loss": 2.3651,
+      "step": 12482
+    },
+    {
+      "epoch": 1.0074247437656363,
+      "grad_norm": 0.6900867223739624,
+      "learning_rate": 6.255723145827954e-05,
+      "loss": 2.4644,
+      "step": 12483
+    },
+    {
+      "epoch": 1.0075054475022194,
+      "grad_norm": 0.7144716382026672,
+      "learning_rate": 6.254259342739683e-05,
+      "loss": 2.4219,
+      "step": 12484
+    },
+    {
+      "epoch": 1.0075861512388025,
+      "grad_norm": 0.674619197845459,
+      "learning_rate": 6.252795633005056e-05,
+      "loss": 2.5038,
+      "step": 12485
+    },
+    {
+      "epoch": 1.0076668549753853,
+      "grad_norm": 0.7036965489387512,
+      "learning_rate": 6.251332016660558e-05,
+      "loss": 2.4784,
+      "step": 12486
+    },
+    {
+      "epoch": 1.0077475587119684,
+      "grad_norm": 0.7046369910240173,
+      "learning_rate": 6.249868493742668e-05,
+      "loss": 2.514,
+      "step": 12487
+    },
+    {
+      "epoch": 1.0078282624485513,
+      "grad_norm": 0.6933087110519409,
+      "learning_rate": 6.248405064287854e-05,
+      "loss": 2.4855,
+      "step": 12488
+    },
+    {
+      "epoch": 1.0079089661851344,
+      "grad_norm": 0.7210546731948853,
+      "learning_rate": 6.246941728332594e-05,
+      "loss": 2.5101,
+      "step": 12489
+    },
+    {
+      "epoch": 1.0079896699217175,
+      "grad_norm": 0.6738288402557373,
+      "learning_rate": 6.245478485913361e-05,
+      "loss": 2.4891,
+      "step": 12490
+    },
+    {
+      "epoch": 1.0080703736583003,
+      "grad_norm": 0.7023273706436157,
+      "learning_rate": 6.244015337066611e-05,
+      "loss": 2.4977,
+      "step": 12491
+    },
+    {
+      "epoch": 1.0081510773948834,
+      "grad_norm": 0.6761355996131897,
+      "learning_rate": 6.24255228182882e-05,
+      "loss": 2.4948,
+      "step": 12492
+    },
+    {
+      "epoch": 1.0082317811314665,
+      "grad_norm": 0.6427976489067078,
+      "learning_rate": 6.241089320236448e-05,
+      "loss": 2.466,
+      "step": 12493
+    },
+    {
+      "epoch": 1.0083124848680494,
+      "grad_norm": 0.6907719969749451,
+      "learning_rate": 6.23962645232596e-05,
+      "loss": 2.437,
+      "step": 12494
+    },
+    {
+      "epoch": 1.0083931886046325,
+      "grad_norm": 0.709032416343689,
+      "learning_rate": 6.238163678133807e-05,
+      "loss": 2.4298,
+      "step": 12495
+    },
+    {
+      "epoch": 1.0084738923412153,
+      "grad_norm": 0.7395734786987305,
+      "learning_rate": 6.236700997696448e-05,
+      "loss": 2.4502,
+      "step": 12496
+    },
+    {
+      "epoch": 1.0085545960777984,
+      "grad_norm": 0.6535435914993286,
+      "learning_rate": 6.23523841105034e-05,
+      "loss": 2.4494,
+      "step": 12497
+    },
+    {
+      "epoch": 1.0086352998143815,
+      "grad_norm": 0.6597761511802673,
+      "learning_rate": 6.23377591823193e-05,
+      "loss": 2.4377,
+      "step": 12498
+    },
+    {
+      "epoch": 1.0087160035509644,
+      "grad_norm": 0.6610515713691711,
+      "learning_rate": 6.232313519277668e-05,
+      "loss": 2.4328,
+      "step": 12499
+    },
+    {
+      "epoch": 1.0087967072875474,
+      "grad_norm": 0.6785424947738647,
+      "learning_rate": 6.230851214224009e-05,
+      "loss": 2.457,
+      "step": 12500
+    },
+    {
+      "epoch": 1.0088774110241303,
+      "grad_norm": 0.6939748525619507,
+      "learning_rate": 6.229389003107383e-05,
+      "loss": 2.383,
+      "step": 12501
+    },
+    {
+      "epoch": 1.0089581147607134,
+      "grad_norm": 0.7592256665229797,
+      "learning_rate": 6.22792688596424e-05,
+      "loss": 2.4665,
+      "step": 12502
+    },
+    {
+      "epoch": 1.0090388184972965,
+      "grad_norm": 0.6751298308372498,
+      "learning_rate": 6.226464862831023e-05,
+      "loss": 2.491,
+      "step": 12503
+    },
+    {
+      "epoch": 1.0091195222338794,
+      "grad_norm": 0.682771623134613,
+      "learning_rate": 6.225002933744164e-05,
+      "loss": 2.4275,
+      "step": 12504
+    },
+    {
+      "epoch": 1.0092002259704624,
+      "grad_norm": 0.7314651608467102,
+      "learning_rate": 6.223541098740098e-05,
+      "loss": 2.4489,
+      "step": 12505
+    },
+    {
+      "epoch": 1.0092809297070455,
+      "grad_norm": 0.7132120132446289,
+      "learning_rate": 6.222079357855261e-05,
+      "loss": 2.4819,
+      "step": 12506
+    },
+    {
+      "epoch": 1.0093616334436284,
+      "grad_norm": 0.6571424007415771,
+      "learning_rate": 6.220617711126082e-05,
+      "loss": 2.455,
+      "step": 12507
+    },
+    {
+      "epoch": 1.0094423371802115,
+      "grad_norm": 0.7675301432609558,
+      "learning_rate": 6.21915615858899e-05,
+      "loss": 2.5282,
+      "step": 12508
+    },
+    {
+      "epoch": 1.0095230409167943,
+      "grad_norm": 0.6907868385314941,
+      "learning_rate": 6.217694700280408e-05,
+      "loss": 2.4639,
+      "step": 12509
+    },
+    {
+      "epoch": 1.0096037446533774,
+      "grad_norm": 0.7223815321922302,
+      "learning_rate": 6.216233336236764e-05,
+      "loss": 2.4682,
+      "step": 12510
+    },
+    {
+      "epoch": 1.0096844483899605,
+      "grad_norm": 0.7325109839439392,
+      "learning_rate": 6.214772066494474e-05,
+      "loss": 2.4591,
+      "step": 12511
+    },
+    {
+      "epoch": 1.0097651521265434,
+      "grad_norm": 0.6589400768280029,
+      "learning_rate": 6.213310891089957e-05,
+      "loss": 2.4883,
+      "step": 12512
+    },
+    {
+      "epoch": 1.0098458558631265,
+      "grad_norm": 0.6692262291908264,
+      "learning_rate": 6.211849810059635e-05,
+      "loss": 2.4635,
+      "step": 12513
+    },
+    {
+      "epoch": 1.0099265595997096,
+      "grad_norm": 0.7352520823478699,
+      "learning_rate": 6.210388823439914e-05,
+      "loss": 2.4743,
+      "step": 12514
+    },
+    {
+      "epoch": 1.0100072633362924,
+      "grad_norm": 0.6631996035575867,
+      "learning_rate": 6.208927931267212e-05,
+      "loss": 2.4848,
+      "step": 12515
+    },
+    {
+      "epoch": 1.0100879670728755,
+      "grad_norm": 0.6985767483711243,
+      "learning_rate": 6.207467133577937e-05,
+      "loss": 2.5044,
+      "step": 12516
+    },
+    {
+      "epoch": 1.0101686708094584,
+      "grad_norm": 0.665635347366333,
+      "learning_rate": 6.206006430408494e-05,
+      "loss": 2.4718,
+      "step": 12517
+    },
+    {
+      "epoch": 1.0102493745460415,
+      "grad_norm": 0.6859133243560791,
+      "learning_rate": 6.204545821795286e-05,
+      "loss": 2.4702,
+      "step": 12518
+    },
+    {
+      "epoch": 1.0103300782826246,
+      "grad_norm": 0.6578841805458069,
+      "learning_rate": 6.203085307774722e-05,
+      "loss": 2.4614,
+      "step": 12519
+    },
+    {
+      "epoch": 1.0104107820192074,
+      "grad_norm": 0.717523455619812,
+      "learning_rate": 6.201624888383194e-05,
+      "loss": 2.4412,
+      "step": 12520
+    },
+    {
+      "epoch": 1.0104914857557905,
+      "grad_norm": 0.7333831787109375,
+      "learning_rate": 6.200164563657103e-05,
+      "loss": 2.4157,
+      "step": 12521
+    },
+    {
+      "epoch": 1.0105721894923736,
+      "grad_norm": 0.6968720555305481,
+      "learning_rate": 6.198704333632845e-05,
+      "loss": 2.4556,
+      "step": 12522
+    },
+    {
+      "epoch": 1.0106528932289565,
+      "grad_norm": 0.6533070802688599,
+      "learning_rate": 6.19724419834681e-05,
+      "loss": 2.43,
+      "step": 12523
+    },
+    {
+      "epoch": 1.0107335969655395,
+      "grad_norm": 0.7341824769973755,
+      "learning_rate": 6.195784157835391e-05,
+      "loss": 2.5326,
+      "step": 12524
+    },
+    {
+      "epoch": 1.0108143007021224,
+      "grad_norm": 0.752912163734436,
+      "learning_rate": 6.194324212134974e-05,
+      "loss": 2.4282,
+      "step": 12525
+    },
+    {
+      "epoch": 1.0108950044387055,
+      "grad_norm": 0.6538611650466919,
+      "learning_rate": 6.192864361281951e-05,
+      "loss": 2.4135,
+      "step": 12526
+    },
+    {
+      "epoch": 1.0109757081752886,
+      "grad_norm": 0.6931454539299011,
+      "learning_rate": 6.191404605312695e-05,
+      "loss": 2.5097,
+      "step": 12527
+    },
+    {
+      "epoch": 1.0110564119118715,
+      "grad_norm": 0.6317688822746277,
+      "learning_rate": 6.18994494426359e-05,
+      "loss": 2.4977,
+      "step": 12528
+    },
+    {
+      "epoch": 1.0111371156484545,
+      "grad_norm": 0.6793715953826904,
+      "learning_rate": 6.188485378171024e-05,
+      "loss": 2.4619,
+      "step": 12529
+    },
+    {
+      "epoch": 1.0112178193850376,
+      "grad_norm": 0.6696654558181763,
+      "learning_rate": 6.187025907071361e-05,
+      "loss": 2.4658,
+      "step": 12530
+    },
+    {
+      "epoch": 1.0112985231216205,
+      "grad_norm": 0.6788807511329651,
+      "learning_rate": 6.185566531000979e-05,
+      "loss": 2.4793,
+      "step": 12531
+    },
+    {
+      "epoch": 1.0113792268582036,
+      "grad_norm": 0.6933971643447876,
+      "learning_rate": 6.184107249996253e-05,
+      "loss": 2.4772,
+      "step": 12532
+    },
+    {
+      "epoch": 1.0114599305947864,
+      "grad_norm": 0.6866000294685364,
+      "learning_rate": 6.182648064093546e-05,
+      "loss": 2.428,
+      "step": 12533
+    },
+    {
+      "epoch": 1.0115406343313695,
+      "grad_norm": 0.7013841271400452,
+      "learning_rate": 6.181188973329229e-05,
+      "loss": 2.5273,
+      "step": 12534
+    },
+    {
+      "epoch": 1.0116213380679526,
+      "grad_norm": 0.6569108963012695,
+      "learning_rate": 6.179729977739669e-05,
+      "loss": 2.4125,
+      "step": 12535
+    },
+    {
+      "epoch": 1.0117020418045355,
+      "grad_norm": 0.7503486275672913,
+      "learning_rate": 6.17827107736122e-05,
+      "loss": 2.4385,
+      "step": 12536
+    },
+    {
+      "epoch": 1.0117827455411186,
+      "grad_norm": 0.6757314205169678,
+      "learning_rate": 6.176812272230246e-05,
+      "loss": 2.4364,
+      "step": 12537
+    },
+    {
+      "epoch": 1.0118634492777017,
+      "grad_norm": 0.6567254662513733,
+      "learning_rate": 6.175353562383106e-05,
+      "loss": 2.4992,
+      "step": 12538
+    },
+    {
+      "epoch": 1.0119441530142845,
+      "grad_norm": 0.7564988732337952,
+      "learning_rate": 6.17389494785615e-05,
+      "loss": 2.4777,
+      "step": 12539
+    },
+    {
+      "epoch": 1.0120248567508676,
+      "grad_norm": 0.6972391605377197,
+      "learning_rate": 6.172436428685735e-05,
+      "loss": 2.5041,
+      "step": 12540
+    },
+    {
+      "epoch": 1.0121055604874505,
+      "grad_norm": 0.6861580610275269,
+      "learning_rate": 6.170978004908209e-05,
+      "loss": 2.4684,
+      "step": 12541
+    },
+    {
+      "epoch": 1.0121862642240336,
+      "grad_norm": 0.6621903777122498,
+      "learning_rate": 6.169519676559921e-05,
+      "loss": 2.4614,
+      "step": 12542
+    },
+    {
+      "epoch": 1.0122669679606167,
+      "grad_norm": 0.6879795789718628,
+      "learning_rate": 6.168061443677215e-05,
+      "loss": 2.4765,
+      "step": 12543
+    },
+    {
+      "epoch": 1.0123476716971995,
+      "grad_norm": 0.6361081004142761,
+      "learning_rate": 6.166603306296434e-05,
+      "loss": 2.4792,
+      "step": 12544
+    },
+    {
+      "epoch": 1.0124283754337826,
+      "grad_norm": 0.6660729050636292,
+      "learning_rate": 6.165145264453924e-05,
+      "loss": 2.489,
+      "step": 12545
+    },
+    {
+      "epoch": 1.0125090791703655,
+      "grad_norm": 0.6900594234466553,
+      "learning_rate": 6.163687318186015e-05,
+      "loss": 2.4543,
+      "step": 12546
+    },
+    {
+      "epoch": 1.0125897829069486,
+      "grad_norm": 0.7195869088172913,
+      "learning_rate": 6.162229467529046e-05,
+      "loss": 2.4137,
+      "step": 12547
+    },
+    {
+      "epoch": 1.0126704866435317,
+      "grad_norm": 0.7030326128005981,
+      "learning_rate": 6.16077171251935e-05,
+      "loss": 2.4657,
+      "step": 12548
+    },
+    {
+      "epoch": 1.0127511903801145,
+      "grad_norm": 0.6712052822113037,
+      "learning_rate": 6.15931405319326e-05,
+      "loss": 2.4718,
+      "step": 12549
+    },
+    {
+      "epoch": 1.0128318941166976,
+      "grad_norm": 0.7471029162406921,
+      "learning_rate": 6.157856489587102e-05,
+      "loss": 2.4705,
+      "step": 12550
+    },
+    {
+      "epoch": 1.0129125978532807,
+      "grad_norm": 0.6813762187957764,
+      "learning_rate": 6.15639902173721e-05,
+      "loss": 2.4479,
+      "step": 12551
+    },
+    {
+      "epoch": 1.0129933015898636,
+      "grad_norm": 0.6657249927520752,
+      "learning_rate": 6.154941649679894e-05,
+      "loss": 2.4911,
+      "step": 12552
+    },
+    {
+      "epoch": 1.0130740053264466,
+      "grad_norm": 0.6700132489204407,
+      "learning_rate": 6.153484373451483e-05,
+      "loss": 2.4962,
+      "step": 12553
+    },
+    {
+      "epoch": 1.0131547090630295,
+      "grad_norm": 0.7058695554733276,
+      "learning_rate": 6.152027193088302e-05,
+      "loss": 2.3935,
+      "step": 12554
+    },
+    {
+      "epoch": 1.0132354127996126,
+      "grad_norm": 0.7390396595001221,
+      "learning_rate": 6.150570108626658e-05,
+      "loss": 2.4454,
+      "step": 12555
+    },
+    {
+      "epoch": 1.0133161165361957,
+      "grad_norm": 0.7251414060592651,
+      "learning_rate": 6.149113120102869e-05,
+      "loss": 2.4146,
+      "step": 12556
+    },
+    {
+      "epoch": 1.0133968202727786,
+      "grad_norm": 0.8262537717819214,
+      "learning_rate": 6.14765622755325e-05,
+      "loss": 2.4638,
+      "step": 12557
+    },
+    {
+      "epoch": 1.0134775240093616,
+      "grad_norm": 0.7184064984321594,
+      "learning_rate": 6.146199431014106e-05,
+      "loss": 2.3958,
+      "step": 12558
+    },
+    {
+      "epoch": 1.0135582277459447,
+      "grad_norm": 0.7544865012168884,
+      "learning_rate": 6.144742730521746e-05,
+      "loss": 2.4662,
+      "step": 12559
+    },
+    {
+      "epoch": 1.0136389314825276,
+      "grad_norm": 0.6866207718849182,
+      "learning_rate": 6.143286126112475e-05,
+      "loss": 2.4951,
+      "step": 12560
+    },
+    {
+      "epoch": 1.0137196352191107,
+      "grad_norm": 0.6566087603569031,
+      "learning_rate": 6.1418296178226e-05,
+      "loss": 2.4002,
+      "step": 12561
+    },
+    {
+      "epoch": 1.0138003389556935,
+      "grad_norm": 0.6999008059501648,
+      "learning_rate": 6.140373205688411e-05,
+      "loss": 2.5306,
+      "step": 12562
+    },
+    {
+      "epoch": 1.0138810426922766,
+      "grad_norm": 0.6682353615760803,
+      "learning_rate": 6.138916889746212e-05,
+      "loss": 2.5565,
+      "step": 12563
+    },
+    {
+      "epoch": 1.0139617464288597,
+      "grad_norm": 0.7443362474441528,
+      "learning_rate": 6.137460670032298e-05,
+      "loss": 2.3958,
+      "step": 12564
+    },
+    {
+      "epoch": 1.0140424501654426,
+      "grad_norm": 0.6542403697967529,
+      "learning_rate": 6.136004546582958e-05,
+      "loss": 2.4394,
+      "step": 12565
+    },
+    {
+      "epoch": 1.0141231539020257,
+      "grad_norm": 0.6524317264556885,
+      "learning_rate": 6.134548519434488e-05,
+      "loss": 2.4979,
+      "step": 12566
+    },
+    {
+      "epoch": 1.0142038576386088,
+      "grad_norm": 0.6605600118637085,
+      "learning_rate": 6.133092588623174e-05,
+      "loss": 2.4827,
+      "step": 12567
+    },
+    {
+      "epoch": 1.0142845613751916,
+      "grad_norm": 0.7114397883415222,
+      "learning_rate": 6.1316367541853e-05,
+      "loss": 2.4799,
+      "step": 12568
+    },
+    {
+      "epoch": 1.0143652651117747,
+      "grad_norm": 0.6607296466827393,
+      "learning_rate": 6.130181016157148e-05,
+      "loss": 2.4991,
+      "step": 12569
+    },
+    {
+      "epoch": 1.0144459688483576,
+      "grad_norm": 0.6750844717025757,
+      "learning_rate": 6.128725374575005e-05,
+      "loss": 2.4451,
+      "step": 12570
+    },
+    {
+      "epoch": 1.0145266725849407,
+      "grad_norm": 0.6978901624679565,
+      "learning_rate": 6.127269829475141e-05,
+      "loss": 2.4608,
+      "step": 12571
+    },
+    {
+      "epoch": 1.0146073763215238,
+      "grad_norm": 0.676343560218811,
+      "learning_rate": 6.125814380893838e-05,
+      "loss": 2.4536,
+      "step": 12572
+    },
+    {
+      "epoch": 1.0146880800581066,
+      "grad_norm": 0.7082604765892029,
+      "learning_rate": 6.124359028867368e-05,
+      "loss": 2.45,
+      "step": 12573
+    },
+    {
+      "epoch": 1.0147687837946897,
+      "grad_norm": 0.7049853205680847,
+      "learning_rate": 6.122903773432003e-05,
+      "loss": 2.4378,
+      "step": 12574
+    },
+    {
+      "epoch": 1.0148494875312728,
+      "grad_norm": 0.6329593062400818,
+      "learning_rate": 6.121448614624009e-05,
+      "loss": 2.4386,
+      "step": 12575
+    },
+    {
+      "epoch": 1.0149301912678557,
+      "grad_norm": 0.7249468564987183,
+      "learning_rate": 6.119993552479655e-05,
+      "loss": 2.5191,
+      "step": 12576
+    },
+    {
+      "epoch": 1.0150108950044388,
+      "grad_norm": 0.7028193473815918,
+      "learning_rate": 6.118538587035206e-05,
+      "loss": 2.4376,
+      "step": 12577
+    },
+    {
+      "epoch": 1.0150915987410216,
+      "grad_norm": 0.697382926940918,
+      "learning_rate": 6.117083718326917e-05,
+      "loss": 2.4797,
+      "step": 12578
+    },
+    {
+      "epoch": 1.0151723024776047,
+      "grad_norm": 0.7386965155601501,
+      "learning_rate": 6.115628946391055e-05,
+      "loss": 2.4512,
+      "step": 12579
+    },
+    {
+      "epoch": 1.0152530062141878,
+      "grad_norm": 0.6614577174186707,
+      "learning_rate": 6.114174271263875e-05,
+      "loss": 2.4404,
+      "step": 12580
+    },
+    {
+      "epoch": 1.0153337099507707,
+      "grad_norm": 0.6927464604377747,
+      "learning_rate": 6.112719692981627e-05,
+      "loss": 2.47,
+      "step": 12581
+    },
+    {
+      "epoch": 1.0154144136873537,
+      "grad_norm": 0.7004262208938599,
+      "learning_rate": 6.111265211580566e-05,
+      "loss": 2.4212,
+      "step": 12582
+    },
+    {
+      "epoch": 1.0154951174239368,
+      "grad_norm": 0.71146559715271,
+      "learning_rate": 6.109810827096942e-05,
+      "loss": 2.4431,
+      "step": 12583
+    },
+    {
+      "epoch": 1.0155758211605197,
+      "grad_norm": 0.6857032775878906,
+      "learning_rate": 6.108356539567e-05,
+      "loss": 2.453,
+      "step": 12584
+    },
+    {
+      "epoch": 1.0156565248971028,
+      "grad_norm": 0.6976168155670166,
+      "learning_rate": 6.106902349026986e-05,
+      "loss": 2.4718,
+      "step": 12585
+    },
+    {
+      "epoch": 1.0157372286336857,
+      "grad_norm": 0.7158414125442505,
+      "learning_rate": 6.105448255513146e-05,
+      "loss": 2.425,
+      "step": 12586
+    },
+    {
+      "epoch": 1.0158179323702687,
+      "grad_norm": 0.6611737608909607,
+      "learning_rate": 6.103994259061714e-05,
+      "loss": 2.4563,
+      "step": 12587
+    },
+    {
+      "epoch": 1.0158986361068518,
+      "grad_norm": 0.7262980937957764,
+      "learning_rate": 6.102540359708926e-05,
+      "loss": 2.4538,
+      "step": 12588
+    },
+    {
+      "epoch": 1.0159793398434347,
+      "grad_norm": 0.7123451828956604,
+      "learning_rate": 6.10108655749102e-05,
+      "loss": 2.4677,
+      "step": 12589
+    },
+    {
+      "epoch": 1.0160600435800178,
+      "grad_norm": 0.7135589122772217,
+      "learning_rate": 6.099632852444235e-05,
+      "loss": 2.4312,
+      "step": 12590
+    },
+    {
+      "epoch": 1.0161407473166009,
+      "grad_norm": 0.6509461998939514,
+      "learning_rate": 6.09817924460479e-05,
+      "loss": 2.4716,
+      "step": 12591
+    },
+    {
+      "epoch": 1.0162214510531837,
+      "grad_norm": 0.8835915923118591,
+      "learning_rate": 6.096725734008919e-05,
+      "loss": 2.4817,
+      "step": 12592
+    },
+    {
+      "epoch": 1.0163021547897668,
+      "grad_norm": 0.7084136605262756,
+      "learning_rate": 6.095272320692846e-05,
+      "loss": 2.483,
+      "step": 12593
+    },
+    {
+      "epoch": 1.0163828585263497,
+      "grad_norm": 0.6866818070411682,
+      "learning_rate": 6.0938190046927934e-05,
+      "loss": 2.4838,
+      "step": 12594
+    },
+    {
+      "epoch": 1.0164635622629328,
+      "grad_norm": 0.7297510504722595,
+      "learning_rate": 6.0923657860449824e-05,
+      "loss": 2.4675,
+      "step": 12595
+    },
+    {
+      "epoch": 1.0165442659995159,
+      "grad_norm": 0.6735619306564331,
+      "learning_rate": 6.090912664785633e-05,
+      "loss": 2.444,
+      "step": 12596
+    },
+    {
+      "epoch": 1.0166249697360987,
+      "grad_norm": 0.7046451568603516,
+      "learning_rate": 6.0894596409509565e-05,
+      "loss": 2.4757,
+      "step": 12597
+    },
+    {
+      "epoch": 1.0167056734726818,
+      "grad_norm": 0.6646085977554321,
+      "learning_rate": 6.0880067145771656e-05,
+      "loss": 2.4772,
+      "step": 12598
+    },
+    {
+      "epoch": 1.0167863772092647,
+      "grad_norm": 0.7217094302177429,
+      "learning_rate": 6.086553885700478e-05,
+      "loss": 2.4589,
+      "step": 12599
+    },
+    {
+      "epoch": 1.0168670809458478,
+      "grad_norm": 0.647378146648407,
+      "learning_rate": 6.085101154357093e-05,
+      "loss": 2.4327,
+      "step": 12600
+    },
+    {
+      "epoch": 1.0169477846824309,
+      "grad_norm": 0.6907125115394592,
+      "learning_rate": 6.083648520583223e-05,
+      "loss": 2.467,
+      "step": 12601
+    },
+    {
+      "epoch": 1.0170284884190137,
+      "grad_norm": 0.690433919429779,
+      "learning_rate": 6.0821959844150687e-05,
+      "loss": 2.488,
+      "step": 12602
+    },
+    {
+      "epoch": 1.0171091921555968,
+      "grad_norm": 0.6528738737106323,
+      "learning_rate": 6.080743545888833e-05,
+      "loss": 2.5028,
+      "step": 12603
+    },
+    {
+      "epoch": 1.01718989589218,
+      "grad_norm": 0.6962323784828186,
+      "learning_rate": 6.079291205040711e-05,
+      "loss": 2.5381,
+      "step": 12604
+    },
+    {
+      "epoch": 1.0172705996287628,
+      "grad_norm": 0.7386075854301453,
+      "learning_rate": 6.077838961906902e-05,
+      "loss": 2.4445,
+      "step": 12605
+    },
+    {
+      "epoch": 1.0173513033653458,
+      "grad_norm": 0.7382189631462097,
+      "learning_rate": 6.0763868165236025e-05,
+      "loss": 2.4926,
+      "step": 12606
+    },
+    {
+      "epoch": 1.0174320071019287,
+      "grad_norm": 0.7291865944862366,
+      "learning_rate": 6.074934768926995e-05,
+      "loss": 2.4624,
+      "step": 12607
+    },
+    {
+      "epoch": 1.0175127108385118,
+      "grad_norm": 0.754843533039093,
+      "learning_rate": 6.073482819153275e-05,
+      "loss": 2.4291,
+      "step": 12608
+    },
+    {
+      "epoch": 1.017593414575095,
+      "grad_norm": 0.6827771663665771,
+      "learning_rate": 6.072030967238628e-05,
+      "loss": 2.453,
+      "step": 12609
+    },
+    {
+      "epoch": 1.0176741183116778,
+      "grad_norm": 0.7138541340827942,
+      "learning_rate": 6.0705792132192355e-05,
+      "loss": 2.5172,
+      "step": 12610
+    },
+    {
+      "epoch": 1.0177548220482608,
+      "grad_norm": 0.6539924740791321,
+      "learning_rate": 6.06912755713128e-05,
+      "loss": 2.4393,
+      "step": 12611
+    },
+    {
+      "epoch": 1.017835525784844,
+      "grad_norm": 0.7021273970603943,
+      "learning_rate": 6.067675999010945e-05,
+      "loss": 2.4519,
+      "step": 12612
+    },
+    {
+      "epoch": 1.0179162295214268,
+      "grad_norm": 0.7124225497245789,
+      "learning_rate": 6.0662245388944004e-05,
+      "loss": 2.4417,
+      "step": 12613
+    },
+    {
+      "epoch": 1.0179969332580099,
+      "grad_norm": 0.7214948534965515,
+      "learning_rate": 6.064773176817823e-05,
+      "loss": 2.4708,
+      "step": 12614
+    },
+    {
+      "epoch": 1.0180776369945927,
+      "grad_norm": 0.6738584041595459,
+      "learning_rate": 6.063321912817386e-05,
+      "loss": 2.4574,
+      "step": 12615
+    },
+    {
+      "epoch": 1.0181583407311758,
+      "grad_norm": 0.7215890884399414,
+      "learning_rate": 6.061870746929257e-05,
+      "loss": 2.4903,
+      "step": 12616
+    },
+    {
+      "epoch": 1.018239044467759,
+      "grad_norm": 0.6720155477523804,
+      "learning_rate": 6.0604196791896016e-05,
+      "loss": 2.4251,
+      "step": 12617
+    },
+    {
+      "epoch": 1.0183197482043418,
+      "grad_norm": 0.7046420574188232,
+      "learning_rate": 6.058968709634587e-05,
+      "loss": 2.446,
+      "step": 12618
+    },
+    {
+      "epoch": 1.0184004519409249,
+      "grad_norm": 0.6419540047645569,
+      "learning_rate": 6.0575178383003764e-05,
+      "loss": 2.4052,
+      "step": 12619
+    },
+    {
+      "epoch": 1.018481155677508,
+      "grad_norm": 0.6948695182800293,
+      "learning_rate": 6.0560670652231235e-05,
+      "loss": 2.5068,
+      "step": 12620
+    },
+    {
+      "epoch": 1.0185618594140908,
+      "grad_norm": 0.7274870276451111,
+      "learning_rate": 6.05461639043899e-05,
+      "loss": 2.4705,
+      "step": 12621
+    },
+    {
+      "epoch": 1.018642563150674,
+      "grad_norm": 0.6809766292572021,
+      "learning_rate": 6.053165813984134e-05,
+      "loss": 2.3767,
+      "step": 12622
+    },
+    {
+      "epoch": 1.0187232668872568,
+      "grad_norm": 0.6197625994682312,
+      "learning_rate": 6.0517153358946985e-05,
+      "loss": 2.4639,
+      "step": 12623
+    },
+    {
+      "epoch": 1.0188039706238399,
+      "grad_norm": 0.6613010764122009,
+      "learning_rate": 6.050264956206837e-05,
+      "loss": 2.5155,
+      "step": 12624
+    },
+    {
+      "epoch": 1.018884674360423,
+      "grad_norm": 0.7335553765296936,
+      "learning_rate": 6.0488146749567e-05,
+      "loss": 2.5344,
+      "step": 12625
+    },
+    {
+      "epoch": 1.0189653780970058,
+      "grad_norm": 0.7175146341323853,
+      "learning_rate": 6.047364492180428e-05,
+      "loss": 2.4972,
+      "step": 12626
+    },
+    {
+      "epoch": 1.019046081833589,
+      "grad_norm": 0.6825357675552368,
+      "learning_rate": 6.045914407914166e-05,
+      "loss": 2.4356,
+      "step": 12627
+    },
+    {
+      "epoch": 1.019126785570172,
+      "grad_norm": 0.6369633078575134,
+      "learning_rate": 6.044464422194056e-05,
+      "loss": 2.4692,
+      "step": 12628
+    },
+    {
+      "epoch": 1.0192074893067549,
+      "grad_norm": 0.7407073378562927,
+      "learning_rate": 6.0430145350562264e-05,
+      "loss": 2.4565,
+      "step": 12629
+    },
+    {
+      "epoch": 1.019288193043338,
+      "grad_norm": 0.6836552619934082,
+      "learning_rate": 6.041564746536821e-05,
+      "loss": 2.4357,
+      "step": 12630
+    },
+    {
+      "epoch": 1.0193688967799208,
+      "grad_norm": 0.6778741478919983,
+      "learning_rate": 6.040115056671972e-05,
+      "loss": 2.424,
+      "step": 12631
+    },
+    {
+      "epoch": 1.019449600516504,
+      "grad_norm": 0.6440724730491638,
+      "learning_rate": 6.0386654654978035e-05,
+      "loss": 2.4455,
+      "step": 12632
+    },
+    {
+      "epoch": 1.019530304253087,
+      "grad_norm": 0.681376039981842,
+      "learning_rate": 6.0372159730504476e-05,
+      "loss": 2.4562,
+      "step": 12633
+    },
+    {
+      "epoch": 1.0196110079896699,
+      "grad_norm": 0.657462477684021,
+      "learning_rate": 6.035766579366029e-05,
+      "loss": 2.4315,
+      "step": 12634
+    },
+    {
+      "epoch": 1.019691711726253,
+      "grad_norm": 0.6540380716323853,
+      "learning_rate": 6.0343172844806706e-05,
+      "loss": 2.4789,
+      "step": 12635
+    },
+    {
+      "epoch": 1.019772415462836,
+      "grad_norm": 0.711883008480072,
+      "learning_rate": 6.03286808843049e-05,
+      "loss": 2.4178,
+      "step": 12636
+    },
+    {
+      "epoch": 1.019853119199419,
+      "grad_norm": 0.6746736168861389,
+      "learning_rate": 6.031418991251607e-05,
+      "loss": 2.4351,
+      "step": 12637
+    },
+    {
+      "epoch": 1.019933822936002,
+      "grad_norm": 0.677237331867218,
+      "learning_rate": 6.02996999298014e-05,
+      "loss": 2.4335,
+      "step": 12638
+    },
+    {
+      "epoch": 1.0200145266725849,
+      "grad_norm": 0.6950497627258301,
+      "learning_rate": 6.0285210936521955e-05,
+      "loss": 2.5178,
+      "step": 12639
+    },
+    {
+      "epoch": 1.020095230409168,
+      "grad_norm": 0.6349243521690369,
+      "learning_rate": 6.027072293303885e-05,
+      "loss": 2.4405,
+      "step": 12640
+    },
+    {
+      "epoch": 1.020175934145751,
+      "grad_norm": 0.744276762008667,
+      "learning_rate": 6.0256235919713236e-05,
+      "loss": 2.5156,
+      "step": 12641
+    },
+    {
+      "epoch": 1.020256637882334,
+      "grad_norm": 0.7697997689247131,
+      "learning_rate": 6.0241749896906075e-05,
+      "loss": 2.4393,
+      "step": 12642
+    },
+    {
+      "epoch": 1.020337341618917,
+      "grad_norm": 0.7784204483032227,
+      "learning_rate": 6.022726486497844e-05,
+      "loss": 2.4565,
+      "step": 12643
+    },
+    {
+      "epoch": 1.0204180453555,
+      "grad_norm": 0.7434312701225281,
+      "learning_rate": 6.021278082429136e-05,
+      "loss": 2.4637,
+      "step": 12644
+    },
+    {
+      "epoch": 1.020498749092083,
+      "grad_norm": 0.7770118117332458,
+      "learning_rate": 6.019829777520575e-05,
+      "loss": 2.4998,
+      "step": 12645
+    },
+    {
+      "epoch": 1.020579452828666,
+      "grad_norm": 0.7021752595901489,
+      "learning_rate": 6.01838157180826e-05,
+      "loss": 2.4661,
+      "step": 12646
+    },
+    {
+      "epoch": 1.0206601565652489,
+      "grad_norm": 0.6812437176704407,
+      "learning_rate": 6.0169334653282895e-05,
+      "loss": 2.4611,
+      "step": 12647
+    },
+    {
+      "epoch": 1.020740860301832,
+      "grad_norm": 0.757724940776825,
+      "learning_rate": 6.0154854581167455e-05,
+      "loss": 2.4427,
+      "step": 12648
+    },
+    {
+      "epoch": 1.020821564038415,
+      "grad_norm": 0.7386252880096436,
+      "learning_rate": 6.014037550209718e-05,
+      "loss": 2.424,
+      "step": 12649
+    },
+    {
+      "epoch": 1.020902267774998,
+      "grad_norm": 0.7138059735298157,
+      "learning_rate": 6.012589741643295e-05,
+      "loss": 2.4951,
+      "step": 12650
+    },
+    {
+      "epoch": 1.020982971511581,
+      "grad_norm": 0.714022159576416,
+      "learning_rate": 6.011142032453561e-05,
+      "loss": 2.4398,
+      "step": 12651
+    },
+    {
+      "epoch": 1.0210636752481639,
+      "grad_norm": 0.6961550712585449,
+      "learning_rate": 6.00969442267659e-05,
+      "loss": 2.4495,
+      "step": 12652
+    },
+    {
+      "epoch": 1.021144378984747,
+      "grad_norm": 0.7196643948554993,
+      "learning_rate": 6.008246912348467e-05,
+      "loss": 2.4449,
+      "step": 12653
+    },
+    {
+      "epoch": 1.02122508272133,
+      "grad_norm": 0.6163341999053955,
+      "learning_rate": 6.006799501505268e-05,
+      "loss": 2.4108,
+      "step": 12654
+    },
+    {
+      "epoch": 1.021305786457913,
+      "grad_norm": 0.6657030582427979,
+      "learning_rate": 6.005352190183061e-05,
+      "loss": 2.4328,
+      "step": 12655
+    },
+    {
+      "epoch": 1.021386490194496,
+      "grad_norm": 0.7183353900909424,
+      "learning_rate": 6.00390497841792e-05,
+      "loss": 2.4912,
+      "step": 12656
+    },
+    {
+      "epoch": 1.021467193931079,
+      "grad_norm": 0.6912575364112854,
+      "learning_rate": 6.002457866245916e-05,
+      "loss": 2.4597,
+      "step": 12657
+    },
+    {
+      "epoch": 1.021547897667662,
+      "grad_norm": 0.7395210266113281,
+      "learning_rate": 6.0010108537031084e-05,
+      "loss": 2.4823,
+      "step": 12658
+    },
+    {
+      "epoch": 1.021628601404245,
+      "grad_norm": 0.722618043422699,
+      "learning_rate": 5.9995639408255636e-05,
+      "loss": 2.4924,
+      "step": 12659
+    },
+    {
+      "epoch": 1.021709305140828,
+      "grad_norm": 0.739009439945221,
+      "learning_rate": 5.998117127649344e-05,
+      "loss": 2.4454,
+      "step": 12660
+    },
+    {
+      "epoch": 1.021790008877411,
+      "grad_norm": 0.7017633318901062,
+      "learning_rate": 5.996670414210506e-05,
+      "loss": 2.5058,
+      "step": 12661
+    },
+    {
+      "epoch": 1.021870712613994,
+      "grad_norm": 0.742664635181427,
+      "learning_rate": 5.9952238005451046e-05,
+      "loss": 2.436,
+      "step": 12662
+    },
+    {
+      "epoch": 1.021951416350577,
+      "grad_norm": 0.6865660548210144,
+      "learning_rate": 5.9937772866892e-05,
+      "loss": 2.4364,
+      "step": 12663
+    },
+    {
+      "epoch": 1.02203212008716,
+      "grad_norm": 0.7376219034194946,
+      "learning_rate": 5.992330872678833e-05,
+      "loss": 2.4975,
+      "step": 12664
+    },
+    {
+      "epoch": 1.0221128238237431,
+      "grad_norm": 0.6496078372001648,
+      "learning_rate": 5.990884558550054e-05,
+      "loss": 2.4651,
+      "step": 12665
+    },
+    {
+      "epoch": 1.022193527560326,
+      "grad_norm": 0.7178322076797485,
+      "learning_rate": 5.989438344338915e-05,
+      "loss": 2.5015,
+      "step": 12666
+    },
+    {
+      "epoch": 1.022274231296909,
+      "grad_norm": 0.7084102034568787,
+      "learning_rate": 5.987992230081459e-05,
+      "loss": 2.4741,
+      "step": 12667
+    },
+    {
+      "epoch": 1.022354935033492,
+      "grad_norm": 0.6634935736656189,
+      "learning_rate": 5.986546215813722e-05,
+      "loss": 2.4255,
+      "step": 12668
+    },
+    {
+      "epoch": 1.022435638770075,
+      "grad_norm": 0.6897543668746948,
+      "learning_rate": 5.985100301571742e-05,
+      "loss": 2.4682,
+      "step": 12669
+    },
+    {
+      "epoch": 1.0225163425066581,
+      "grad_norm": 0.6643948554992676,
+      "learning_rate": 5.9836544873915614e-05,
+      "loss": 2.4009,
+      "step": 12670
+    },
+    {
+      "epoch": 1.022597046243241,
+      "grad_norm": 0.681252658367157,
+      "learning_rate": 5.982208773309208e-05,
+      "loss": 2.4542,
+      "step": 12671
+    },
+    {
+      "epoch": 1.022677749979824,
+      "grad_norm": 0.7608681917190552,
+      "learning_rate": 5.980763159360714e-05,
+      "loss": 2.5614,
+      "step": 12672
+    },
+    {
+      "epoch": 1.0227584537164072,
+      "grad_norm": 0.6855095028877258,
+      "learning_rate": 5.979317645582112e-05,
+      "loss": 2.4505,
+      "step": 12673
+    },
+    {
+      "epoch": 1.02283915745299,
+      "grad_norm": 0.6846089363098145,
+      "learning_rate": 5.97787223200942e-05,
+      "loss": 2.4438,
+      "step": 12674
+    },
+    {
+      "epoch": 1.0229198611895731,
+      "grad_norm": 0.7198090553283691,
+      "learning_rate": 5.9764269186786684e-05,
+      "loss": 2.4469,
+      "step": 12675
+    },
+    {
+      "epoch": 1.023000564926156,
+      "grad_norm": 0.7120245099067688,
+      "learning_rate": 5.9749817056258764e-05,
+      "loss": 2.4626,
+      "step": 12676
+    },
+    {
+      "epoch": 1.023081268662739,
+      "grad_norm": 0.6839897036552429,
+      "learning_rate": 5.973536592887059e-05,
+      "loss": 2.4384,
+      "step": 12677
+    },
+    {
+      "epoch": 1.0231619723993222,
+      "grad_norm": 0.7053773999214172,
+      "learning_rate": 5.9720915804982356e-05,
+      "loss": 2.4554,
+      "step": 12678
+    },
+    {
+      "epoch": 1.023242676135905,
+      "grad_norm": 0.7114294767379761,
+      "learning_rate": 5.970646668495421e-05,
+      "loss": 2.3964,
+      "step": 12679
+    },
+    {
+      "epoch": 1.0233233798724881,
+      "grad_norm": 0.7001516819000244,
+      "learning_rate": 5.9692018569146224e-05,
+      "loss": 2.5216,
+      "step": 12680
+    },
+    {
+      "epoch": 1.0234040836090712,
+      "grad_norm": 0.6715773940086365,
+      "learning_rate": 5.96775714579185e-05,
+      "loss": 2.4595,
+      "step": 12681
+    },
+    {
+      "epoch": 1.023484787345654,
+      "grad_norm": 0.6856278777122498,
+      "learning_rate": 5.96631253516311e-05,
+      "loss": 2.4637,
+      "step": 12682
+    },
+    {
+      "epoch": 1.0235654910822372,
+      "grad_norm": 0.6785625219345093,
+      "learning_rate": 5.96486802506441e-05,
+      "loss": 2.4615,
+      "step": 12683
+    },
+    {
+      "epoch": 1.02364619481882,
+      "grad_norm": 0.6834213137626648,
+      "learning_rate": 5.963423615531743e-05,
+      "loss": 2.4729,
+      "step": 12684
+    },
+    {
+      "epoch": 1.023726898555403,
+      "grad_norm": 0.6729516386985779,
+      "learning_rate": 5.961979306601109e-05,
+      "loss": 2.4013,
+      "step": 12685
+    },
+    {
+      "epoch": 1.0238076022919862,
+      "grad_norm": 0.6785775423049927,
+      "learning_rate": 5.960535098308511e-05,
+      "loss": 2.4825,
+      "step": 12686
+    },
+    {
+      "epoch": 1.023888306028569,
+      "grad_norm": 0.67277991771698,
+      "learning_rate": 5.959090990689934e-05,
+      "loss": 2.4606,
+      "step": 12687
+    },
+    {
+      "epoch": 1.0239690097651521,
+      "grad_norm": 0.7679588198661804,
+      "learning_rate": 5.957646983781373e-05,
+      "loss": 2.5234,
+      "step": 12688
+    },
+    {
+      "epoch": 1.0240497135017352,
+      "grad_norm": 0.6597407460212708,
+      "learning_rate": 5.956203077618821e-05,
+      "loss": 2.4699,
+      "step": 12689
+    },
+    {
+      "epoch": 1.024130417238318,
+      "grad_norm": 0.6743008494377136,
+      "learning_rate": 5.9547592722382525e-05,
+      "loss": 2.4266,
+      "step": 12690
+    },
+    {
+      "epoch": 1.0242111209749012,
+      "grad_norm": 0.7223396897315979,
+      "learning_rate": 5.953315567675657e-05,
+      "loss": 2.5117,
+      "step": 12691
+    },
+    {
+      "epoch": 1.024291824711484,
+      "grad_norm": 0.6729528307914734,
+      "learning_rate": 5.951871963967022e-05,
+      "loss": 2.4586,
+      "step": 12692
+    },
+    {
+      "epoch": 1.0243725284480671,
+      "grad_norm": 0.6523739695549011,
+      "learning_rate": 5.950428461148314e-05,
+      "loss": 2.4408,
+      "step": 12693
+    },
+    {
+      "epoch": 1.0244532321846502,
+      "grad_norm": 0.6830984950065613,
+      "learning_rate": 5.9489850592555164e-05,
+      "loss": 2.4094,
+      "step": 12694
+    },
+    {
+      "epoch": 1.024533935921233,
+      "grad_norm": 0.6223493814468384,
+      "learning_rate": 5.9475417583246006e-05,
+      "loss": 2.4105,
+      "step": 12695
+    },
+    {
+      "epoch": 1.0246146396578162,
+      "grad_norm": 0.6506635546684265,
+      "learning_rate": 5.9460985583915374e-05,
+      "loss": 2.4451,
+      "step": 12696
+    },
+    {
+      "epoch": 1.024695343394399,
+      "grad_norm": 0.7626760005950928,
+      "learning_rate": 5.944655459492293e-05,
+      "loss": 2.4643,
+      "step": 12697
+    },
+    {
+      "epoch": 1.0247760471309821,
+      "grad_norm": 0.7074631452560425,
+      "learning_rate": 5.943212461662837e-05,
+      "loss": 2.4662,
+      "step": 12698
+    },
+    {
+      "epoch": 1.0248567508675652,
+      "grad_norm": 0.718083918094635,
+      "learning_rate": 5.9417695649391346e-05,
+      "loss": 2.4686,
+      "step": 12699
+    },
+    {
+      "epoch": 1.024937454604148,
+      "grad_norm": 0.6850628852844238,
+      "learning_rate": 5.9403267693571384e-05,
+      "loss": 2.4542,
+      "step": 12700
+    },
+    {
+      "epoch": 1.0250181583407312,
+      "grad_norm": 0.6662585735321045,
+      "learning_rate": 5.938884074952812e-05,
+      "loss": 2.4676,
+      "step": 12701
+    },
+    {
+      "epoch": 1.0250988620773143,
+      "grad_norm": 0.6806240677833557,
+      "learning_rate": 5.9374414817621114e-05,
+      "loss": 2.4243,
+      "step": 12702
+    },
+    {
+      "epoch": 1.0251795658138971,
+      "grad_norm": 0.6763548851013184,
+      "learning_rate": 5.9359989898209876e-05,
+      "loss": 2.4389,
+      "step": 12703
+    },
+    {
+      "epoch": 1.0252602695504802,
+      "grad_norm": 0.7390143275260925,
+      "learning_rate": 5.934556599165393e-05,
+      "loss": 2.4667,
+      "step": 12704
+    },
+    {
+      "epoch": 1.025340973287063,
+      "grad_norm": 0.6159299612045288,
+      "learning_rate": 5.933114309831276e-05,
+      "loss": 2.3832,
+      "step": 12705
+    },
+    {
+      "epoch": 1.0254216770236462,
+      "grad_norm": 0.6779586672782898,
+      "learning_rate": 5.931672121854579e-05,
+      "loss": 2.4615,
+      "step": 12706
+    },
+    {
+      "epoch": 1.0255023807602293,
+      "grad_norm": 0.643800675868988,
+      "learning_rate": 5.930230035271247e-05,
+      "loss": 2.4725,
+      "step": 12707
+    },
+    {
+      "epoch": 1.0255830844968121,
+      "grad_norm": 0.6605903506278992,
+      "learning_rate": 5.928788050117227e-05,
+      "loss": 2.4332,
+      "step": 12708
+    },
+    {
+      "epoch": 1.0256637882333952,
+      "grad_norm": 0.7046334743499756,
+      "learning_rate": 5.927346166428446e-05,
+      "loss": 2.4445,
+      "step": 12709
+    },
+    {
+      "epoch": 1.0257444919699783,
+      "grad_norm": 0.6536325216293335,
+      "learning_rate": 5.925904384240843e-05,
+      "loss": 2.4168,
+      "step": 12710
+    },
+    {
+      "epoch": 1.0258251957065612,
+      "grad_norm": 0.6861097812652588,
+      "learning_rate": 5.9244627035903564e-05,
+      "loss": 2.512,
+      "step": 12711
+    },
+    {
+      "epoch": 1.0259058994431443,
+      "grad_norm": 0.6782278418540955,
+      "learning_rate": 5.923021124512911e-05,
+      "loss": 2.4667,
+      "step": 12712
+    },
+    {
+      "epoch": 1.0259866031797271,
+      "grad_norm": 0.724435031414032,
+      "learning_rate": 5.921579647044436e-05,
+      "loss": 2.4828,
+      "step": 12713
+    },
+    {
+      "epoch": 1.0260673069163102,
+      "grad_norm": 0.6690630316734314,
+      "learning_rate": 5.9201382712208575e-05,
+      "loss": 2.4832,
+      "step": 12714
+    },
+    {
+      "epoch": 1.0261480106528933,
+      "grad_norm": 0.7045348286628723,
+      "learning_rate": 5.9186969970781015e-05,
+      "loss": 2.4576,
+      "step": 12715
+    },
+    {
+      "epoch": 1.0262287143894762,
+      "grad_norm": 0.673321008682251,
+      "learning_rate": 5.9172558246520796e-05,
+      "loss": 2.3986,
+      "step": 12716
+    },
+    {
+      "epoch": 1.0263094181260592,
+      "grad_norm": 0.7184785008430481,
+      "learning_rate": 5.915814753978717e-05,
+      "loss": 2.4008,
+      "step": 12717
+    },
+    {
+      "epoch": 1.0263901218626423,
+      "grad_norm": 0.6971293091773987,
+      "learning_rate": 5.914373785093931e-05,
+      "loss": 2.4559,
+      "step": 12718
+    },
+    {
+      "epoch": 1.0264708255992252,
+      "grad_norm": 0.6941563487052917,
+      "learning_rate": 5.912932918033626e-05,
+      "loss": 2.4787,
+      "step": 12719
+    },
+    {
+      "epoch": 1.0265515293358083,
+      "grad_norm": 0.6276142001152039,
+      "learning_rate": 5.911492152833715e-05,
+      "loss": 2.4275,
+      "step": 12720
+    },
+    {
+      "epoch": 1.0266322330723912,
+      "grad_norm": 0.715928316116333,
+      "learning_rate": 5.9100514895301106e-05,
+      "loss": 2.4127,
+      "step": 12721
+    },
+    {
+      "epoch": 1.0267129368089742,
+      "grad_norm": 0.7004076838493347,
+      "learning_rate": 5.908610928158713e-05,
+      "loss": 2.4651,
+      "step": 12722
+    },
+    {
+      "epoch": 1.0267936405455573,
+      "grad_norm": 0.6761921048164368,
+      "learning_rate": 5.907170468755425e-05,
+      "loss": 2.4245,
+      "step": 12723
+    },
+    {
+      "epoch": 1.0268743442821402,
+      "grad_norm": 0.7246574759483337,
+      "learning_rate": 5.9057301113561515e-05,
+      "loss": 2.4489,
+      "step": 12724
+    },
+    {
+      "epoch": 1.0269550480187233,
+      "grad_norm": 0.7196606397628784,
+      "learning_rate": 5.904289855996783e-05,
+      "loss": 2.4357,
+      "step": 12725
+    },
+    {
+      "epoch": 1.0270357517553064,
+      "grad_norm": 0.7142692804336548,
+      "learning_rate": 5.902849702713216e-05,
+      "loss": 2.4821,
+      "step": 12726
+    },
+    {
+      "epoch": 1.0271164554918892,
+      "grad_norm": 0.7207832336425781,
+      "learning_rate": 5.9014096515413454e-05,
+      "loss": 2.4337,
+      "step": 12727
+    },
+    {
+      "epoch": 1.0271971592284723,
+      "grad_norm": 0.6865695714950562,
+      "learning_rate": 5.899969702517063e-05,
+      "loss": 2.4549,
+      "step": 12728
+    },
+    {
+      "epoch": 1.0272778629650552,
+      "grad_norm": 0.7136662006378174,
+      "learning_rate": 5.898529855676249e-05,
+      "loss": 2.4606,
+      "step": 12729
+    },
+    {
+      "epoch": 1.0273585667016383,
+      "grad_norm": 0.701885998249054,
+      "learning_rate": 5.897090111054795e-05,
+      "loss": 2.4913,
+      "step": 12730
+    },
+    {
+      "epoch": 1.0274392704382214,
+      "grad_norm": 0.6671354174613953,
+      "learning_rate": 5.8956504686885805e-05,
+      "loss": 2.4064,
+      "step": 12731
+    },
+    {
+      "epoch": 1.0275199741748042,
+      "grad_norm": 0.6720621585845947,
+      "learning_rate": 5.894210928613484e-05,
+      "loss": 2.4908,
+      "step": 12732
+    },
+    {
+      "epoch": 1.0276006779113873,
+      "grad_norm": 0.7530980706214905,
+      "learning_rate": 5.892771490865383e-05,
+      "loss": 2.4486,
+      "step": 12733
+    },
+    {
+      "epoch": 1.0276813816479704,
+      "grad_norm": 0.6771122813224792,
+      "learning_rate": 5.891332155480158e-05,
+      "loss": 2.3954,
+      "step": 12734
+    },
+    {
+      "epoch": 1.0277620853845533,
+      "grad_norm": 0.6779236793518066,
+      "learning_rate": 5.889892922493671e-05,
+      "loss": 2.4404,
+      "step": 12735
+    },
+    {
+      "epoch": 1.0278427891211364,
+      "grad_norm": 0.7593358755111694,
+      "learning_rate": 5.8884537919417974e-05,
+      "loss": 2.4997,
+      "step": 12736
+    },
+    {
+      "epoch": 1.0279234928577192,
+      "grad_norm": 0.672686755657196,
+      "learning_rate": 5.8870147638604044e-05,
+      "loss": 2.5394,
+      "step": 12737
+    },
+    {
+      "epoch": 1.0280041965943023,
+      "grad_norm": 0.6727546453475952,
+      "learning_rate": 5.885575838285353e-05,
+      "loss": 2.4554,
+      "step": 12738
+    },
+    {
+      "epoch": 1.0280849003308854,
+      "grad_norm": 0.7092764377593994,
+      "learning_rate": 5.884137015252507e-05,
+      "loss": 2.4568,
+      "step": 12739
+    },
+    {
+      "epoch": 1.0281656040674683,
+      "grad_norm": 0.6988070011138916,
+      "learning_rate": 5.882698294797728e-05,
+      "loss": 2.4453,
+      "step": 12740
+    },
+    {
+      "epoch": 1.0282463078040514,
+      "grad_norm": 0.7578697204589844,
+      "learning_rate": 5.8812596769568676e-05,
+      "loss": 2.5648,
+      "step": 12741
+    },
+    {
+      "epoch": 1.0283270115406344,
+      "grad_norm": 0.6523683667182922,
+      "learning_rate": 5.879821161765782e-05,
+      "loss": 2.4088,
+      "step": 12742
+    },
+    {
+      "epoch": 1.0284077152772173,
+      "grad_norm": 0.6797270178794861,
+      "learning_rate": 5.878382749260323e-05,
+      "loss": 2.4465,
+      "step": 12743
+    },
+    {
+      "epoch": 1.0284884190138004,
+      "grad_norm": 0.6823786497116089,
+      "learning_rate": 5.876944439476345e-05,
+      "loss": 2.5053,
+      "step": 12744
+    },
+    {
+      "epoch": 1.0285691227503833,
+      "grad_norm": 0.6840088367462158,
+      "learning_rate": 5.875506232449686e-05,
+      "loss": 2.3771,
+      "step": 12745
+    },
+    {
+      "epoch": 1.0286498264869663,
+      "grad_norm": 0.6985318064689636,
+      "learning_rate": 5.8740681282161914e-05,
+      "loss": 2.4456,
+      "step": 12746
+    },
+    {
+      "epoch": 1.0287305302235494,
+      "grad_norm": 0.7102388739585876,
+      "learning_rate": 5.872630126811707e-05,
+      "loss": 2.4802,
+      "step": 12747
+    },
+    {
+      "epoch": 1.0288112339601323,
+      "grad_norm": 0.7917937636375427,
+      "learning_rate": 5.871192228272067e-05,
+      "loss": 2.4606,
+      "step": 12748
+    },
+    {
+      "epoch": 1.0288919376967154,
+      "grad_norm": 0.683397114276886,
+      "learning_rate": 5.86975443263311e-05,
+      "loss": 2.5011,
+      "step": 12749
+    },
+    {
+      "epoch": 1.0289726414332985,
+      "grad_norm": 0.7543408870697021,
+      "learning_rate": 5.8683167399306724e-05,
+      "loss": 2.4705,
+      "step": 12750
+    },
+    {
+      "epoch": 1.0290533451698813,
+      "grad_norm": 0.6946283578872681,
+      "learning_rate": 5.866879150200579e-05,
+      "loss": 2.4986,
+      "step": 12751
+    },
+    {
+      "epoch": 1.0291340489064644,
+      "grad_norm": 0.6535125374794006,
+      "learning_rate": 5.8654416634786605e-05,
+      "loss": 2.4203,
+      "step": 12752
+    },
+    {
+      "epoch": 1.0292147526430473,
+      "grad_norm": 0.7470195889472961,
+      "learning_rate": 5.8640042798007455e-05,
+      "loss": 2.5103,
+      "step": 12753
+    },
+    {
+      "epoch": 1.0292954563796304,
+      "grad_norm": 0.6782363653182983,
+      "learning_rate": 5.8625669992026535e-05,
+      "loss": 2.4087,
+      "step": 12754
+    },
+    {
+      "epoch": 1.0293761601162135,
+      "grad_norm": 0.7601497173309326,
+      "learning_rate": 5.861129821720207e-05,
+      "loss": 2.4752,
+      "step": 12755
+    },
+    {
+      "epoch": 1.0294568638527963,
+      "grad_norm": 0.6875388026237488,
+      "learning_rate": 5.859692747389227e-05,
+      "loss": 2.448,
+      "step": 12756
+    },
+    {
+      "epoch": 1.0295375675893794,
+      "grad_norm": 0.7153629064559937,
+      "learning_rate": 5.858255776245525e-05,
+      "loss": 2.4641,
+      "step": 12757
+    },
+    {
+      "epoch": 1.0296182713259623,
+      "grad_norm": 0.682954728603363,
+      "learning_rate": 5.8568189083249145e-05,
+      "loss": 2.441,
+      "step": 12758
+    },
+    {
+      "epoch": 1.0296989750625454,
+      "grad_norm": 0.6959100961685181,
+      "learning_rate": 5.855382143663209e-05,
+      "loss": 2.4316,
+      "step": 12759
+    },
+    {
+      "epoch": 1.0297796787991285,
+      "grad_norm": 0.7062023878097534,
+      "learning_rate": 5.8539454822962167e-05,
+      "loss": 2.4287,
+      "step": 12760
+    },
+    {
+      "epoch": 1.0298603825357113,
+      "grad_norm": 0.706523597240448,
+      "learning_rate": 5.852508924259736e-05,
+      "loss": 2.4596,
+      "step": 12761
+    },
+    {
+      "epoch": 1.0299410862722944,
+      "grad_norm": 0.6908385753631592,
+      "learning_rate": 5.851072469589578e-05,
+      "loss": 2.4428,
+      "step": 12762
+    },
+    {
+      "epoch": 1.0300217900088775,
+      "grad_norm": 0.6810726523399353,
+      "learning_rate": 5.8496361183215386e-05,
+      "loss": 2.4902,
+      "step": 12763
+    },
+    {
+      "epoch": 1.0301024937454604,
+      "grad_norm": 0.661613941192627,
+      "learning_rate": 5.8481998704914156e-05,
+      "loss": 2.4256,
+      "step": 12764
+    },
+    {
+      "epoch": 1.0301831974820435,
+      "grad_norm": 0.6633132100105286,
+      "learning_rate": 5.846763726135005e-05,
+      "loss": 2.4512,
+      "step": 12765
+    },
+    {
+      "epoch": 1.0302639012186263,
+      "grad_norm": 0.6991820335388184,
+      "learning_rate": 5.8453276852881025e-05,
+      "loss": 2.3747,
+      "step": 12766
+    },
+    {
+      "epoch": 1.0303446049552094,
+      "grad_norm": 0.7392076253890991,
+      "learning_rate": 5.843891747986487e-05,
+      "loss": 2.438,
+      "step": 12767
+    },
+    {
+      "epoch": 1.0304253086917925,
+      "grad_norm": 0.6371724605560303,
+      "learning_rate": 5.842455914265958e-05,
+      "loss": 2.4627,
+      "step": 12768
+    },
+    {
+      "epoch": 1.0305060124283754,
+      "grad_norm": 0.6475048661231995,
+      "learning_rate": 5.841020184162298e-05,
+      "loss": 2.4883,
+      "step": 12769
+    },
+    {
+      "epoch": 1.0305867161649584,
+      "grad_norm": 0.6848995685577393,
+      "learning_rate": 5.839584557711283e-05,
+      "loss": 2.4452,
+      "step": 12770
+    },
+    {
+      "epoch": 1.0306674199015415,
+      "grad_norm": 0.7345505952835083,
+      "learning_rate": 5.838149034948697e-05,
+      "loss": 2.5121,
+      "step": 12771
+    },
+    {
+      "epoch": 1.0307481236381244,
+      "grad_norm": 0.715373158454895,
+      "learning_rate": 5.836713615910318e-05,
+      "loss": 2.4549,
+      "step": 12772
+    },
+    {
+      "epoch": 1.0308288273747075,
+      "grad_norm": 0.7371035814285278,
+      "learning_rate": 5.8352783006319166e-05,
+      "loss": 2.4633,
+      "step": 12773
+    },
+    {
+      "epoch": 1.0309095311112904,
+      "grad_norm": 0.6843077540397644,
+      "learning_rate": 5.833843089149267e-05,
+      "loss": 2.4067,
+      "step": 12774
+    },
+    {
+      "epoch": 1.0309902348478734,
+      "grad_norm": 0.7398965954780579,
+      "learning_rate": 5.832407981498136e-05,
+      "loss": 2.5199,
+      "step": 12775
+    },
+    {
+      "epoch": 1.0310709385844565,
+      "grad_norm": 0.6860283017158508,
+      "learning_rate": 5.830972977714294e-05,
+      "loss": 2.4564,
+      "step": 12776
+    },
+    {
+      "epoch": 1.0311516423210394,
+      "grad_norm": 0.683893084526062,
+      "learning_rate": 5.829538077833503e-05,
+      "loss": 2.4635,
+      "step": 12777
+    },
+    {
+      "epoch": 1.0312323460576225,
+      "grad_norm": 0.6412089467048645,
+      "learning_rate": 5.828103281891525e-05,
+      "loss": 2.4806,
+      "step": 12778
+    },
+    {
+      "epoch": 1.0313130497942056,
+      "grad_norm": 0.646393895149231,
+      "learning_rate": 5.826668589924123e-05,
+      "loss": 2.4674,
+      "step": 12779
+    },
+    {
+      "epoch": 1.0313937535307884,
+      "grad_norm": 0.6805605292320251,
+      "learning_rate": 5.825234001967044e-05,
+      "loss": 2.5145,
+      "step": 12780
+    },
+    {
+      "epoch": 1.0314744572673715,
+      "grad_norm": 0.681532084941864,
+      "learning_rate": 5.8237995180560455e-05,
+      "loss": 2.5041,
+      "step": 12781
+    },
+    {
+      "epoch": 1.0315551610039544,
+      "grad_norm": 0.6971312165260315,
+      "learning_rate": 5.8223651382268865e-05,
+      "loss": 2.5324,
+      "step": 12782
+    },
+    {
+      "epoch": 1.0316358647405375,
+      "grad_norm": 0.6634463667869568,
+      "learning_rate": 5.8209308625153026e-05,
+      "loss": 2.5086,
+      "step": 12783
+    },
+    {
+      "epoch": 1.0317165684771206,
+      "grad_norm": 0.6752117276191711,
+      "learning_rate": 5.819496690957047e-05,
+      "loss": 2.4805,
+      "step": 12784
+    },
+    {
+      "epoch": 1.0317972722137034,
+      "grad_norm": 0.7242109775543213,
+      "learning_rate": 5.818062623587861e-05,
+      "loss": 2.4205,
+      "step": 12785
+    },
+    {
+      "epoch": 1.0318779759502865,
+      "grad_norm": 0.7338563203811646,
+      "learning_rate": 5.816628660443486e-05,
+      "loss": 2.4277,
+      "step": 12786
+    },
+    {
+      "epoch": 1.0319586796868696,
+      "grad_norm": 0.6764293313026428,
+      "learning_rate": 5.81519480155966e-05,
+      "loss": 2.5096,
+      "step": 12787
+    },
+    {
+      "epoch": 1.0320393834234525,
+      "grad_norm": 0.6757099032402039,
+      "learning_rate": 5.813761046972124e-05,
+      "loss": 2.468,
+      "step": 12788
+    },
+    {
+      "epoch": 1.0321200871600356,
+      "grad_norm": 0.7072502374649048,
+      "learning_rate": 5.8123273967166017e-05,
+      "loss": 2.4642,
+      "step": 12789
+    },
+    {
+      "epoch": 1.0322007908966184,
+      "grad_norm": 0.6470256447792053,
+      "learning_rate": 5.810893850828827e-05,
+      "loss": 2.4146,
+      "step": 12790
+    },
+    {
+      "epoch": 1.0322814946332015,
+      "grad_norm": 0.7403351068496704,
+      "learning_rate": 5.809460409344527e-05,
+      "loss": 2.512,
+      "step": 12791
+    },
+    {
+      "epoch": 1.0323621983697846,
+      "grad_norm": 0.6711490154266357,
+      "learning_rate": 5.808027072299432e-05,
+      "loss": 2.4602,
+      "step": 12792
+    },
+    {
+      "epoch": 1.0324429021063675,
+      "grad_norm": 0.7920248508453369,
+      "learning_rate": 5.806593839729258e-05,
+      "loss": 2.4512,
+      "step": 12793
+    },
+    {
+      "epoch": 1.0325236058429506,
+      "grad_norm": 0.6442045569419861,
+      "learning_rate": 5.805160711669725e-05,
+      "loss": 2.4165,
+      "step": 12794
+    },
+    {
+      "epoch": 1.0326043095795336,
+      "grad_norm": 0.6681340932846069,
+      "learning_rate": 5.803727688156553e-05,
+      "loss": 2.4296,
+      "step": 12795
+    },
+    {
+      "epoch": 1.0326850133161165,
+      "grad_norm": 0.6653337478637695,
+      "learning_rate": 5.802294769225457e-05,
+      "loss": 2.5165,
+      "step": 12796
+    },
+    {
+      "epoch": 1.0327657170526996,
+      "grad_norm": 0.6444782018661499,
+      "learning_rate": 5.8008619549121476e-05,
+      "loss": 2.4266,
+      "step": 12797
+    },
+    {
+      "epoch": 1.0328464207892825,
+      "grad_norm": 0.6741451621055603,
+      "learning_rate": 5.7994292452523394e-05,
+      "loss": 2.4837,
+      "step": 12798
+    },
+    {
+      "epoch": 1.0329271245258655,
+      "grad_norm": 0.6629341840744019,
+      "learning_rate": 5.797996640281731e-05,
+      "loss": 2.4368,
+      "step": 12799
+    },
+    {
+      "epoch": 1.0330078282624486,
+      "grad_norm": 0.6755850315093994,
+      "learning_rate": 5.796564140036029e-05,
+      "loss": 2.4834,
+      "step": 12800
+    },
+    {
+      "epoch": 1.0330885319990315,
+      "grad_norm": 0.7271782755851746,
+      "learning_rate": 5.795131744550942e-05,
+      "loss": 2.5025,
+      "step": 12801
+    },
+    {
+      "epoch": 1.0331692357356146,
+      "grad_norm": 0.6870545744895935,
+      "learning_rate": 5.7936994538621605e-05,
+      "loss": 2.4443,
+      "step": 12802
+    },
+    {
+      "epoch": 1.0332499394721975,
+      "grad_norm": 0.7231935858726501,
+      "learning_rate": 5.792267268005382e-05,
+      "loss": 2.4917,
+      "step": 12803
+    },
+    {
+      "epoch": 1.0333306432087805,
+      "grad_norm": 0.6905832290649414,
+      "learning_rate": 5.790835187016307e-05,
+      "loss": 2.4902,
+      "step": 12804
+    },
+    {
+      "epoch": 1.0334113469453636,
+      "grad_norm": 0.711814284324646,
+      "learning_rate": 5.789403210930613e-05,
+      "loss": 2.4579,
+      "step": 12805
+    },
+    {
+      "epoch": 1.0334920506819465,
+      "grad_norm": 0.6982280015945435,
+      "learning_rate": 5.787971339784004e-05,
+      "loss": 2.5275,
+      "step": 12806
+    },
+    {
+      "epoch": 1.0335727544185296,
+      "grad_norm": 0.6871493458747864,
+      "learning_rate": 5.7865395736121575e-05,
+      "loss": 2.4401,
+      "step": 12807
+    },
+    {
+      "epoch": 1.0336534581551127,
+      "grad_norm": 0.6898353099822998,
+      "learning_rate": 5.785107912450763e-05,
+      "loss": 2.4005,
+      "step": 12808
+    },
+    {
+      "epoch": 1.0337341618916955,
+      "grad_norm": 0.6264411807060242,
+      "learning_rate": 5.7836763563354946e-05,
+      "loss": 2.4497,
+      "step": 12809
+    },
+    {
+      "epoch": 1.0338148656282786,
+      "grad_norm": 0.6997092962265015,
+      "learning_rate": 5.782244905302032e-05,
+      "loss": 2.4388,
+      "step": 12810
+    },
+    {
+      "epoch": 1.0338955693648615,
+      "grad_norm": 0.6834601759910583,
+      "learning_rate": 5.7808135593860555e-05,
+      "loss": 2.4298,
+      "step": 12811
+    },
+    {
+      "epoch": 1.0339762731014446,
+      "grad_norm": 0.664315402507782,
+      "learning_rate": 5.77938231862323e-05,
+      "loss": 2.4289,
+      "step": 12812
+    },
+    {
+      "epoch": 1.0340569768380277,
+      "grad_norm": 0.6660603284835815,
+      "learning_rate": 5.7779511830492306e-05,
+      "loss": 2.4772,
+      "step": 12813
+    },
+    {
+      "epoch": 1.0341376805746105,
+      "grad_norm": 0.6457028388977051,
+      "learning_rate": 5.776520152699728e-05,
+      "loss": 2.4408,
+      "step": 12814
+    },
+    {
+      "epoch": 1.0342183843111936,
+      "grad_norm": 0.7132207155227661,
+      "learning_rate": 5.7750892276103794e-05,
+      "loss": 2.4953,
+      "step": 12815
+    },
+    {
+      "epoch": 1.0342990880477767,
+      "grad_norm": 0.7397382259368896,
+      "learning_rate": 5.773658407816848e-05,
+      "loss": 2.4396,
+      "step": 12816
+    },
+    {
+      "epoch": 1.0343797917843596,
+      "grad_norm": 0.6951746344566345,
+      "learning_rate": 5.7722276933548034e-05,
+      "loss": 2.5021,
+      "step": 12817
+    },
+    {
+      "epoch": 1.0344604955209427,
+      "grad_norm": 0.6789736151695251,
+      "learning_rate": 5.7707970842598935e-05,
+      "loss": 2.4883,
+      "step": 12818
+    },
+    {
+      "epoch": 1.0345411992575255,
+      "grad_norm": 0.7231541872024536,
+      "learning_rate": 5.7693665805677747e-05,
+      "loss": 2.4761,
+      "step": 12819
+    },
+    {
+      "epoch": 1.0346219029941086,
+      "grad_norm": 0.685943603515625,
+      "learning_rate": 5.767936182314104e-05,
+      "loss": 2.4489,
+      "step": 12820
+    },
+    {
+      "epoch": 1.0347026067306917,
+      "grad_norm": 0.7081817984580994,
+      "learning_rate": 5.7665058895345236e-05,
+      "loss": 2.4329,
+      "step": 12821
+    },
+    {
+      "epoch": 1.0347833104672746,
+      "grad_norm": 0.6700818538665771,
+      "learning_rate": 5.7650757022646804e-05,
+      "loss": 2.4252,
+      "step": 12822
+    },
+    {
+      "epoch": 1.0348640142038577,
+      "grad_norm": 0.6712214946746826,
+      "learning_rate": 5.763645620540223e-05,
+      "loss": 2.419,
+      "step": 12823
+    },
+    {
+      "epoch": 1.0349447179404407,
+      "grad_norm": 0.6732817888259888,
+      "learning_rate": 5.762215644396793e-05,
+      "loss": 2.3928,
+      "step": 12824
+    },
+    {
+      "epoch": 1.0350254216770236,
+      "grad_norm": 0.6689301133155823,
+      "learning_rate": 5.760785773870024e-05,
+      "loss": 2.3981,
+      "step": 12825
+    },
+    {
+      "epoch": 1.0351061254136067,
+      "grad_norm": 0.6822957992553711,
+      "learning_rate": 5.759356008995556e-05,
+      "loss": 2.5265,
+      "step": 12826
+    },
+    {
+      "epoch": 1.0351868291501896,
+      "grad_norm": 0.7316287755966187,
+      "learning_rate": 5.7579263498090194e-05,
+      "loss": 2.4132,
+      "step": 12827
+    },
+    {
+      "epoch": 1.0352675328867726,
+      "grad_norm": 0.6688703894615173,
+      "learning_rate": 5.756496796346047e-05,
+      "loss": 2.4195,
+      "step": 12828
+    },
+    {
+      "epoch": 1.0353482366233557,
+      "grad_norm": 0.6894570589065552,
+      "learning_rate": 5.755067348642268e-05,
+      "loss": 2.4897,
+      "step": 12829
+    },
+    {
+      "epoch": 1.0354289403599386,
+      "grad_norm": 0.7635753750801086,
+      "learning_rate": 5.753638006733311e-05,
+      "loss": 2.4643,
+      "step": 12830
+    },
+    {
+      "epoch": 1.0355096440965217,
+      "grad_norm": 0.6353672742843628,
+      "learning_rate": 5.75220877065479e-05,
+      "loss": 2.4533,
+      "step": 12831
+    },
+    {
+      "epoch": 1.0355903478331048,
+      "grad_norm": 0.6725208759307861,
+      "learning_rate": 5.750779640442332e-05,
+      "loss": 2.4958,
+      "step": 12832
+    },
+    {
+      "epoch": 1.0356710515696876,
+      "grad_norm": 0.7350767254829407,
+      "learning_rate": 5.749350616131556e-05,
+      "loss": 2.4192,
+      "step": 12833
+    },
+    {
+      "epoch": 1.0357517553062707,
+      "grad_norm": 0.7322222590446472,
+      "learning_rate": 5.7479216977580695e-05,
+      "loss": 2.4719,
+      "step": 12834
+    },
+    {
+      "epoch": 1.0358324590428536,
+      "grad_norm": 0.7233425974845886,
+      "learning_rate": 5.7464928853574904e-05,
+      "loss": 2.4707,
+      "step": 12835
+    },
+    {
+      "epoch": 1.0359131627794367,
+      "grad_norm": 0.7117420434951782,
+      "learning_rate": 5.745064178965427e-05,
+      "loss": 2.4463,
+      "step": 12836
+    },
+    {
+      "epoch": 1.0359938665160198,
+      "grad_norm": 0.7615050077438354,
+      "learning_rate": 5.743635578617486e-05,
+      "loss": 2.4256,
+      "step": 12837
+    },
+    {
+      "epoch": 1.0360745702526026,
+      "grad_norm": 0.7056093215942383,
+      "learning_rate": 5.7422070843492734e-05,
+      "loss": 2.4628,
+      "step": 12838
+    },
+    {
+      "epoch": 1.0361552739891857,
+      "grad_norm": 0.685989499092102,
+      "learning_rate": 5.740778696196389e-05,
+      "loss": 2.4271,
+      "step": 12839
+    },
+    {
+      "epoch": 1.0362359777257688,
+      "grad_norm": 0.7286686301231384,
+      "learning_rate": 5.739350414194439e-05,
+      "loss": 2.4984,
+      "step": 12840
+    },
+    {
+      "epoch": 1.0363166814623517,
+      "grad_norm": 0.6939802765846252,
+      "learning_rate": 5.737922238379009e-05,
+      "loss": 2.4601,
+      "step": 12841
+    },
+    {
+      "epoch": 1.0363973851989348,
+      "grad_norm": 0.7077060341835022,
+      "learning_rate": 5.736494168785698e-05,
+      "loss": 2.4264,
+      "step": 12842
+    },
+    {
+      "epoch": 1.0364780889355176,
+      "grad_norm": 0.667086124420166,
+      "learning_rate": 5.7350662054501016e-05,
+      "loss": 2.4733,
+      "step": 12843
+    },
+    {
+      "epoch": 1.0365587926721007,
+      "grad_norm": 0.6531338691711426,
+      "learning_rate": 5.7336383484078004e-05,
+      "loss": 2.4709,
+      "step": 12844
+    },
+    {
+      "epoch": 1.0366394964086838,
+      "grad_norm": 0.7141630053520203,
+      "learning_rate": 5.732210597694383e-05,
+      "loss": 2.4747,
+      "step": 12845
+    },
+    {
+      "epoch": 1.0367202001452667,
+      "grad_norm": 0.7186396718025208,
+      "learning_rate": 5.730782953345435e-05,
+      "loss": 2.4401,
+      "step": 12846
+    },
+    {
+      "epoch": 1.0368009038818498,
+      "grad_norm": 0.6709686517715454,
+      "learning_rate": 5.7293554153965345e-05,
+      "loss": 2.456,
+      "step": 12847
+    },
+    {
+      "epoch": 1.0368816076184326,
+      "grad_norm": 0.6867267489433289,
+      "learning_rate": 5.727927983883261e-05,
+      "loss": 2.4522,
+      "step": 12848
+    },
+    {
+      "epoch": 1.0369623113550157,
+      "grad_norm": 0.7016724348068237,
+      "learning_rate": 5.7265006588411926e-05,
+      "loss": 2.4348,
+      "step": 12849
+    },
+    {
+      "epoch": 1.0370430150915988,
+      "grad_norm": 0.6764764785766602,
+      "learning_rate": 5.725073440305896e-05,
+      "loss": 2.4241,
+      "step": 12850
+    },
+    {
+      "epoch": 1.0371237188281817,
+      "grad_norm": 0.6965062618255615,
+      "learning_rate": 5.7236463283129435e-05,
+      "loss": 2.4559,
+      "step": 12851
+    },
+    {
+      "epoch": 1.0372044225647647,
+      "grad_norm": 0.6878135800361633,
+      "learning_rate": 5.7222193228979037e-05,
+      "loss": 2.4874,
+      "step": 12852
+    },
+    {
+      "epoch": 1.0372851263013478,
+      "grad_norm": 0.6576557755470276,
+      "learning_rate": 5.720792424096344e-05,
+      "loss": 2.4273,
+      "step": 12853
+    },
+    {
+      "epoch": 1.0373658300379307,
+      "grad_norm": 0.7463123798370361,
+      "learning_rate": 5.719365631943818e-05,
+      "loss": 2.4933,
+      "step": 12854
+    },
+    {
+      "epoch": 1.0374465337745138,
+      "grad_norm": 0.6920896768569946,
+      "learning_rate": 5.7179389464758914e-05,
+      "loss": 2.4799,
+      "step": 12855
+    },
+    {
+      "epoch": 1.0375272375110969,
+      "grad_norm": 0.7330591082572937,
+      "learning_rate": 5.71651236772812e-05,
+      "loss": 2.469,
+      "step": 12856
+    },
+    {
+      "epoch": 1.0376079412476797,
+      "grad_norm": 0.6766076683998108,
+      "learning_rate": 5.715085895736057e-05,
+      "loss": 2.4787,
+      "step": 12857
+    },
+    {
+      "epoch": 1.0376886449842628,
+      "grad_norm": 0.724278450012207,
+      "learning_rate": 5.713659530535255e-05,
+      "loss": 2.4524,
+      "step": 12858
+    },
+    {
+      "epoch": 1.0377693487208457,
+      "grad_norm": 0.6816281676292419,
+      "learning_rate": 5.712233272161265e-05,
+      "loss": 2.4993,
+      "step": 12859
+    },
+    {
+      "epoch": 1.0378500524574288,
+      "grad_norm": 0.7186439633369446,
+      "learning_rate": 5.710807120649626e-05,
+      "loss": 2.4108,
+      "step": 12860
+    },
+    {
+      "epoch": 1.0379307561940119,
+      "grad_norm": 0.6616777181625366,
+      "learning_rate": 5.709381076035887e-05,
+      "loss": 2.4797,
+      "step": 12861
+    },
+    {
+      "epoch": 1.0380114599305947,
+      "grad_norm": 0.6956895589828491,
+      "learning_rate": 5.7079551383555906e-05,
+      "loss": 2.4017,
+      "step": 12862
+    },
+    {
+      "epoch": 1.0380921636671778,
+      "grad_norm": 0.6650584936141968,
+      "learning_rate": 5.706529307644268e-05,
+      "loss": 2.4808,
+      "step": 12863
+    },
+    {
+      "epoch": 1.0381728674037607,
+      "grad_norm": 0.6362698674201965,
+      "learning_rate": 5.705103583937458e-05,
+      "loss": 2.4077,
+      "step": 12864
+    },
+    {
+      "epoch": 1.0382535711403438,
+      "grad_norm": 0.6962565183639526,
+      "learning_rate": 5.703677967270697e-05,
+      "loss": 2.4715,
+      "step": 12865
+    },
+    {
+      "epoch": 1.0383342748769269,
+      "grad_norm": 0.6927294135093689,
+      "learning_rate": 5.702252457679509e-05,
+      "loss": 2.4983,
+      "step": 12866
+    },
+    {
+      "epoch": 1.0384149786135097,
+      "grad_norm": 0.7107497453689575,
+      "learning_rate": 5.70082705519942e-05,
+      "loss": 2.4198,
+      "step": 12867
+    },
+    {
+      "epoch": 1.0384956823500928,
+      "grad_norm": 0.6459221243858337,
+      "learning_rate": 5.6994017598659634e-05,
+      "loss": 2.4423,
+      "step": 12868
+    },
+    {
+      "epoch": 1.038576386086676,
+      "grad_norm": 0.705563485622406,
+      "learning_rate": 5.697976571714658e-05,
+      "loss": 2.5346,
+      "step": 12869
+    },
+    {
+      "epoch": 1.0386570898232588,
+      "grad_norm": 0.7424784898757935,
+      "learning_rate": 5.696551490781021e-05,
+      "loss": 2.4824,
+      "step": 12870
+    },
+    {
+      "epoch": 1.0387377935598419,
+      "grad_norm": 0.6820988059043884,
+      "learning_rate": 5.695126517100569e-05,
+      "loss": 2.4965,
+      "step": 12871
+    },
+    {
+      "epoch": 1.0388184972964247,
+      "grad_norm": 0.8209595680236816,
+      "learning_rate": 5.6937016507088225e-05,
+      "loss": 2.475,
+      "step": 12872
+    },
+    {
+      "epoch": 1.0388992010330078,
+      "grad_norm": 0.7407695055007935,
+      "learning_rate": 5.6922768916412815e-05,
+      "loss": 2.4683,
+      "step": 12873
+    },
+    {
+      "epoch": 1.038979904769591,
+      "grad_norm": 0.7335677742958069,
+      "learning_rate": 5.690852239933462e-05,
+      "loss": 2.4621,
+      "step": 12874
+    },
+    {
+      "epoch": 1.0390606085061738,
+      "grad_norm": 0.6731325387954712,
+      "learning_rate": 5.689427695620873e-05,
+      "loss": 2.4882,
+      "step": 12875
+    },
+    {
+      "epoch": 1.0391413122427569,
+      "grad_norm": 0.7256175875663757,
+      "learning_rate": 5.68800325873901e-05,
+      "loss": 2.4827,
+      "step": 12876
+    },
+    {
+      "epoch": 1.03922201597934,
+      "grad_norm": 0.711928129196167,
+      "learning_rate": 5.686578929323377e-05,
+      "loss": 2.4447,
+      "step": 12877
+    },
+    {
+      "epoch": 1.0393027197159228,
+      "grad_norm": 0.6445996165275574,
+      "learning_rate": 5.685154707409473e-05,
+      "loss": 2.453,
+      "step": 12878
+    },
+    {
+      "epoch": 1.039383423452506,
+      "grad_norm": 0.6656066179275513,
+      "learning_rate": 5.6837305930327923e-05,
+      "loss": 2.4863,
+      "step": 12879
+    },
+    {
+      "epoch": 1.0394641271890888,
+      "grad_norm": 0.6844663619995117,
+      "learning_rate": 5.682306586228828e-05,
+      "loss": 2.4524,
+      "step": 12880
+    },
+    {
+      "epoch": 1.0395448309256718,
+      "grad_norm": 0.6436383724212646,
+      "learning_rate": 5.6808826870330746e-05,
+      "loss": 2.4137,
+      "step": 12881
+    },
+    {
+      "epoch": 1.039625534662255,
+      "grad_norm": 0.6731196641921997,
+      "learning_rate": 5.6794588954810104e-05,
+      "loss": 2.4176,
+      "step": 12882
+    },
+    {
+      "epoch": 1.0397062383988378,
+      "grad_norm": 0.6994587779045105,
+      "learning_rate": 5.678035211608125e-05,
+      "loss": 2.4651,
+      "step": 12883
+    },
+    {
+      "epoch": 1.0397869421354209,
+      "grad_norm": 0.6912599205970764,
+      "learning_rate": 5.6766116354499e-05,
+      "loss": 2.3918,
+      "step": 12884
+    },
+    {
+      "epoch": 1.039867645872004,
+      "grad_norm": 0.7627033591270447,
+      "learning_rate": 5.6751881670418185e-05,
+      "loss": 2.4278,
+      "step": 12885
+    },
+    {
+      "epoch": 1.0399483496085868,
+      "grad_norm": 0.7107213139533997,
+      "learning_rate": 5.6737648064193485e-05,
+      "loss": 2.5249,
+      "step": 12886
+    },
+    {
+      "epoch": 1.04002905334517,
+      "grad_norm": 0.7254211902618408,
+      "learning_rate": 5.672341553617968e-05,
+      "loss": 2.4454,
+      "step": 12887
+    },
+    {
+      "epoch": 1.0401097570817528,
+      "grad_norm": 0.6776205897331238,
+      "learning_rate": 5.670918408673149e-05,
+      "loss": 2.4333,
+      "step": 12888
+    },
+    {
+      "epoch": 1.0401904608183359,
+      "grad_norm": 0.6824465394020081,
+      "learning_rate": 5.669495371620359e-05,
+      "loss": 2.427,
+      "step": 12889
+    },
+    {
+      "epoch": 1.040271164554919,
+      "grad_norm": 0.6633001565933228,
+      "learning_rate": 5.668072442495066e-05,
+      "loss": 2.4874,
+      "step": 12890
+    },
+    {
+      "epoch": 1.0403518682915018,
+      "grad_norm": 0.6655289530754089,
+      "learning_rate": 5.666649621332735e-05,
+      "loss": 2.5023,
+      "step": 12891
+    },
+    {
+      "epoch": 1.040432572028085,
+      "grad_norm": 0.6892853379249573,
+      "learning_rate": 5.665226908168818e-05,
+      "loss": 2.4505,
+      "step": 12892
+    },
+    {
+      "epoch": 1.040513275764668,
+      "grad_norm": 0.7154649496078491,
+      "learning_rate": 5.6638043030387774e-05,
+      "loss": 2.4916,
+      "step": 12893
+    },
+    {
+      "epoch": 1.0405939795012509,
+      "grad_norm": 0.6780592799186707,
+      "learning_rate": 5.662381805978074e-05,
+      "loss": 2.4116,
+      "step": 12894
+    },
+    {
+      "epoch": 1.040674683237834,
+      "grad_norm": 0.6737352013587952,
+      "learning_rate": 5.66095941702215e-05,
+      "loss": 2.3903,
+      "step": 12895
+    },
+    {
+      "epoch": 1.0407553869744168,
+      "grad_norm": 0.7623820304870605,
+      "learning_rate": 5.659537136206461e-05,
+      "loss": 2.4334,
+      "step": 12896
+    },
+    {
+      "epoch": 1.040836090711,
+      "grad_norm": 0.7043081521987915,
+      "learning_rate": 5.65811496356645e-05,
+      "loss": 2.4403,
+      "step": 12897
+    },
+    {
+      "epoch": 1.040916794447583,
+      "grad_norm": 0.6704873442649841,
+      "learning_rate": 5.6566928991375654e-05,
+      "loss": 2.4416,
+      "step": 12898
+    },
+    {
+      "epoch": 1.0409974981841659,
+      "grad_norm": 0.6556837558746338,
+      "learning_rate": 5.6552709429552474e-05,
+      "loss": 2.4904,
+      "step": 12899
+    },
+    {
+      "epoch": 1.041078201920749,
+      "grad_norm": 0.6926451325416565,
+      "learning_rate": 5.653849095054935e-05,
+      "loss": 2.4889,
+      "step": 12900
+    },
+    {
+      "epoch": 1.041158905657332,
+      "grad_norm": 0.6407613158226013,
+      "learning_rate": 5.6524273554720674e-05,
+      "loss": 2.3951,
+      "step": 12901
+    },
+    {
+      "epoch": 1.041239609393915,
+      "grad_norm": 0.7812615633010864,
+      "learning_rate": 5.651005724242071e-05,
+      "loss": 2.4535,
+      "step": 12902
+    },
+    {
+      "epoch": 1.041320313130498,
+      "grad_norm": 0.6868990659713745,
+      "learning_rate": 5.6495842014003796e-05,
+      "loss": 2.4373,
+      "step": 12903
+    },
+    {
+      "epoch": 1.0414010168670809,
+      "grad_norm": 0.6467776894569397,
+      "learning_rate": 5.648162786982427e-05,
+      "loss": 2.4929,
+      "step": 12904
+    },
+    {
+      "epoch": 1.041481720603664,
+      "grad_norm": 0.6588063836097717,
+      "learning_rate": 5.64674148102363e-05,
+      "loss": 2.4445,
+      "step": 12905
+    },
+    {
+      "epoch": 1.041562424340247,
+      "grad_norm": 0.6880654096603394,
+      "learning_rate": 5.6453202835594136e-05,
+      "loss": 2.4298,
+      "step": 12906
+    },
+    {
+      "epoch": 1.04164312807683,
+      "grad_norm": 0.7471407055854797,
+      "learning_rate": 5.6438991946251996e-05,
+      "loss": 2.4669,
+      "step": 12907
+    },
+    {
+      "epoch": 1.041723831813413,
+      "grad_norm": 0.7069533467292786,
+      "learning_rate": 5.6424782142564034e-05,
+      "loss": 2.4498,
+      "step": 12908
+    },
+    {
+      "epoch": 1.0418045355499959,
+      "grad_norm": 0.7013602256774902,
+      "learning_rate": 5.641057342488443e-05,
+      "loss": 2.4993,
+      "step": 12909
+    },
+    {
+      "epoch": 1.041885239286579,
+      "grad_norm": 0.6870697736740112,
+      "learning_rate": 5.6396365793567305e-05,
+      "loss": 2.5338,
+      "step": 12910
+    },
+    {
+      "epoch": 1.041965943023162,
+      "grad_norm": 0.6569130420684814,
+      "learning_rate": 5.638215924896669e-05,
+      "loss": 2.4538,
+      "step": 12911
+    },
+    {
+      "epoch": 1.042046646759745,
+      "grad_norm": 0.6900331377983093,
+      "learning_rate": 5.636795379143669e-05,
+      "loss": 2.4013,
+      "step": 12912
+    },
+    {
+      "epoch": 1.042127350496328,
+      "grad_norm": 0.6800071001052856,
+      "learning_rate": 5.635374942133136e-05,
+      "loss": 2.4733,
+      "step": 12913
+    },
+    {
+      "epoch": 1.042208054232911,
+      "grad_norm": 0.703601598739624,
+      "learning_rate": 5.6339546139004663e-05,
+      "loss": 2.432,
+      "step": 12914
+    },
+    {
+      "epoch": 1.042288757969494,
+      "grad_norm": 0.6781988739967346,
+      "learning_rate": 5.6325343944810594e-05,
+      "loss": 2.4418,
+      "step": 12915
+    },
+    {
+      "epoch": 1.042369461706077,
+      "grad_norm": 0.7247167825698853,
+      "learning_rate": 5.6311142839103125e-05,
+      "loss": 2.5133,
+      "step": 12916
+    },
+    {
+      "epoch": 1.04245016544266,
+      "grad_norm": 0.7738155126571655,
+      "learning_rate": 5.629694282223619e-05,
+      "loss": 2.5137,
+      "step": 12917
+    },
+    {
+      "epoch": 1.042530869179243,
+      "grad_norm": 0.74723219871521,
+      "learning_rate": 5.628274389456367e-05,
+      "loss": 2.3996,
+      "step": 12918
+    },
+    {
+      "epoch": 1.042611572915826,
+      "grad_norm": 0.7245466709136963,
+      "learning_rate": 5.6268546056439456e-05,
+      "loss": 2.4213,
+      "step": 12919
+    },
+    {
+      "epoch": 1.042692276652409,
+      "grad_norm": 0.6307608485221863,
+      "learning_rate": 5.625434930821742e-05,
+      "loss": 2.4195,
+      "step": 12920
+    },
+    {
+      "epoch": 1.042772980388992,
+      "grad_norm": 0.7138007879257202,
+      "learning_rate": 5.6240153650251326e-05,
+      "loss": 2.463,
+      "step": 12921
+    },
+    {
+      "epoch": 1.042853684125575,
+      "grad_norm": 0.779659628868103,
+      "learning_rate": 5.622595908289498e-05,
+      "loss": 2.4898,
+      "step": 12922
+    },
+    {
+      "epoch": 1.042934387862158,
+      "grad_norm": 0.7144278287887573,
+      "learning_rate": 5.621176560650221e-05,
+      "loss": 2.4083,
+      "step": 12923
+    },
+    {
+      "epoch": 1.043015091598741,
+      "grad_norm": 0.7724754214286804,
+      "learning_rate": 5.619757322142667e-05,
+      "loss": 2.3917,
+      "step": 12924
+    },
+    {
+      "epoch": 1.043095795335324,
+      "grad_norm": 0.7667245268821716,
+      "learning_rate": 5.618338192802208e-05,
+      "loss": 2.4943,
+      "step": 12925
+    },
+    {
+      "epoch": 1.043176499071907,
+      "grad_norm": 0.6528030037879944,
+      "learning_rate": 5.616919172664221e-05,
+      "loss": 2.4323,
+      "step": 12926
+    },
+    {
+      "epoch": 1.04325720280849,
+      "grad_norm": 0.6790263652801514,
+      "learning_rate": 5.6155002617640615e-05,
+      "loss": 2.4304,
+      "step": 12927
+    },
+    {
+      "epoch": 1.043337906545073,
+      "grad_norm": 0.7554369568824768,
+      "learning_rate": 5.614081460137097e-05,
+      "loss": 2.4637,
+      "step": 12928
+    },
+    {
+      "epoch": 1.043418610281656,
+      "grad_norm": 0.7126293182373047,
+      "learning_rate": 5.612662767818686e-05,
+      "loss": 2.4765,
+      "step": 12929
+    },
+    {
+      "epoch": 1.0434993140182391,
+      "grad_norm": 0.6705749034881592,
+      "learning_rate": 5.611244184844189e-05,
+      "loss": 2.4746,
+      "step": 12930
+    },
+    {
+      "epoch": 1.043580017754822,
+      "grad_norm": 0.6595145463943481,
+      "learning_rate": 5.609825711248958e-05,
+      "loss": 2.463,
+      "step": 12931
+    },
+    {
+      "epoch": 1.043660721491405,
+      "grad_norm": 0.6942049860954285,
+      "learning_rate": 5.6084073470683476e-05,
+      "loss": 2.5101,
+      "step": 12932
+    },
+    {
+      "epoch": 1.043741425227988,
+      "grad_norm": 0.7285810708999634,
+      "learning_rate": 5.6069890923377087e-05,
+      "loss": 2.467,
+      "step": 12933
+    },
+    {
+      "epoch": 1.043822128964571,
+      "grad_norm": 0.7702928185462952,
+      "learning_rate": 5.605570947092382e-05,
+      "loss": 2.4998,
+      "step": 12934
+    },
+    {
+      "epoch": 1.0439028327011541,
+      "grad_norm": 0.6631895899772644,
+      "learning_rate": 5.604152911367713e-05,
+      "loss": 2.4277,
+      "step": 12935
+    },
+    {
+      "epoch": 1.043983536437737,
+      "grad_norm": 0.6447882652282715,
+      "learning_rate": 5.6027349851990494e-05,
+      "loss": 2.4868,
+      "step": 12936
+    },
+    {
+      "epoch": 1.04406424017432,
+      "grad_norm": 0.695160448551178,
+      "learning_rate": 5.6013171686217205e-05,
+      "loss": 2.3917,
+      "step": 12937
+    },
+    {
+      "epoch": 1.0441449439109032,
+      "grad_norm": 0.6579271554946899,
+      "learning_rate": 5.5998994616710656e-05,
+      "loss": 2.4245,
+      "step": 12938
+    },
+    {
+      "epoch": 1.044225647647486,
+      "grad_norm": 0.7053574323654175,
+      "learning_rate": 5.598481864382419e-05,
+      "loss": 2.4809,
+      "step": 12939
+    },
+    {
+      "epoch": 1.0443063513840691,
+      "grad_norm": 0.7008736729621887,
+      "learning_rate": 5.5970643767911105e-05,
+      "loss": 2.4481,
+      "step": 12940
+    },
+    {
+      "epoch": 1.044387055120652,
+      "grad_norm": 0.6577918529510498,
+      "learning_rate": 5.5956469989324644e-05,
+      "loss": 2.4211,
+      "step": 12941
+    },
+    {
+      "epoch": 1.044467758857235,
+      "grad_norm": 0.6662739515304565,
+      "learning_rate": 5.594229730841815e-05,
+      "loss": 2.4607,
+      "step": 12942
+    },
+    {
+      "epoch": 1.0445484625938182,
+      "grad_norm": 0.6637060046195984,
+      "learning_rate": 5.592812572554471e-05,
+      "loss": 2.4388,
+      "step": 12943
+    },
+    {
+      "epoch": 1.044629166330401,
+      "grad_norm": 0.7282097935676575,
+      "learning_rate": 5.5913955241057605e-05,
+      "loss": 2.4536,
+      "step": 12944
+    },
+    {
+      "epoch": 1.0447098700669841,
+      "grad_norm": 0.6470810174942017,
+      "learning_rate": 5.589978585530997e-05,
+      "loss": 2.4032,
+      "step": 12945
+    },
+    {
+      "epoch": 1.0447905738035672,
+      "grad_norm": 0.6958881616592407,
+      "learning_rate": 5.588561756865498e-05,
+      "loss": 2.4577,
+      "step": 12946
+    },
+    {
+      "epoch": 1.04487127754015,
+      "grad_norm": 0.6999812722206116,
+      "learning_rate": 5.587145038144569e-05,
+      "loss": 2.454,
+      "step": 12947
+    },
+    {
+      "epoch": 1.0449519812767332,
+      "grad_norm": 0.6919988989830017,
+      "learning_rate": 5.58572842940352e-05,
+      "loss": 2.4505,
+      "step": 12948
+    },
+    {
+      "epoch": 1.045032685013316,
+      "grad_norm": 0.6813084483146667,
+      "learning_rate": 5.584311930677659e-05,
+      "loss": 2.4873,
+      "step": 12949
+    },
+    {
+      "epoch": 1.0451133887498991,
+      "grad_norm": 0.6587427854537964,
+      "learning_rate": 5.582895542002286e-05,
+      "loss": 2.4658,
+      "step": 12950
+    },
+    {
+      "epoch": 1.0451940924864822,
+      "grad_norm": 0.6942041516304016,
+      "learning_rate": 5.581479263412703e-05,
+      "loss": 2.47,
+      "step": 12951
+    },
+    {
+      "epoch": 1.045274796223065,
+      "grad_norm": 0.7330117225646973,
+      "learning_rate": 5.58006309494421e-05,
+      "loss": 2.4826,
+      "step": 12952
+    },
+    {
+      "epoch": 1.0453554999596482,
+      "grad_norm": 0.7197144031524658,
+      "learning_rate": 5.578647036632096e-05,
+      "loss": 2.4425,
+      "step": 12953
+    },
+    {
+      "epoch": 1.045436203696231,
+      "grad_norm": 0.7442573308944702,
+      "learning_rate": 5.577231088511654e-05,
+      "loss": 2.4946,
+      "step": 12954
+    },
+    {
+      "epoch": 1.0455169074328141,
+      "grad_norm": 0.7039753198623657,
+      "learning_rate": 5.575815250618179e-05,
+      "loss": 2.4188,
+      "step": 12955
+    },
+    {
+      "epoch": 1.0455976111693972,
+      "grad_norm": 0.7374606728553772,
+      "learning_rate": 5.574399522986951e-05,
+      "loss": 2.3916,
+      "step": 12956
+    },
+    {
+      "epoch": 1.04567831490598,
+      "grad_norm": 0.6358140707015991,
+      "learning_rate": 5.572983905653253e-05,
+      "loss": 2.4502,
+      "step": 12957
+    },
+    {
+      "epoch": 1.0457590186425632,
+      "grad_norm": 0.712858259677887,
+      "learning_rate": 5.5715683986523694e-05,
+      "loss": 2.4746,
+      "step": 12958
+    },
+    {
+      "epoch": 1.0458397223791462,
+      "grad_norm": 0.6757933497428894,
+      "learning_rate": 5.5701530020195756e-05,
+      "loss": 2.4836,
+      "step": 12959
+    },
+    {
+      "epoch": 1.045920426115729,
+      "grad_norm": 0.7509831786155701,
+      "learning_rate": 5.568737715790151e-05,
+      "loss": 2.4061,
+      "step": 12960
+    },
+    {
+      "epoch": 1.0460011298523122,
+      "grad_norm": 0.7120335102081299,
+      "learning_rate": 5.5673225399993646e-05,
+      "loss": 2.4772,
+      "step": 12961
+    },
+    {
+      "epoch": 1.046081833588895,
+      "grad_norm": 0.7213751673698425,
+      "learning_rate": 5.5659074746824924e-05,
+      "loss": 2.4637,
+      "step": 12962
+    },
+    {
+      "epoch": 1.0461625373254781,
+      "grad_norm": 0.7161290645599365,
+      "learning_rate": 5.5644925198747934e-05,
+      "loss": 2.4552,
+      "step": 12963
+    },
+    {
+      "epoch": 1.0462432410620612,
+      "grad_norm": 0.7303922772407532,
+      "learning_rate": 5.563077675611534e-05,
+      "loss": 2.5091,
+      "step": 12964
+    },
+    {
+      "epoch": 1.046323944798644,
+      "grad_norm": 0.7051636576652527,
+      "learning_rate": 5.561662941927981e-05,
+      "loss": 2.3717,
+      "step": 12965
+    },
+    {
+      "epoch": 1.0464046485352272,
+      "grad_norm": 0.6880733370780945,
+      "learning_rate": 5.5602483188593866e-05,
+      "loss": 2.4205,
+      "step": 12966
+    },
+    {
+      "epoch": 1.0464853522718103,
+      "grad_norm": 0.6942360401153564,
+      "learning_rate": 5.558833806441008e-05,
+      "loss": 2.4601,
+      "step": 12967
+    },
+    {
+      "epoch": 1.0465660560083931,
+      "grad_norm": 0.7264992594718933,
+      "learning_rate": 5.5574194047081016e-05,
+      "loss": 2.4612,
+      "step": 12968
+    },
+    {
+      "epoch": 1.0466467597449762,
+      "grad_norm": 0.7502472996711731,
+      "learning_rate": 5.5560051136959166e-05,
+      "loss": 2.4099,
+      "step": 12969
+    },
+    {
+      "epoch": 1.046727463481559,
+      "grad_norm": 0.691694438457489,
+      "learning_rate": 5.5545909334397004e-05,
+      "loss": 2.5071,
+      "step": 12970
+    },
+    {
+      "epoch": 1.0468081672181422,
+      "grad_norm": 0.7120653986930847,
+      "learning_rate": 5.5531768639747026e-05,
+      "loss": 2.4066,
+      "step": 12971
+    },
+    {
+      "epoch": 1.0468888709547253,
+      "grad_norm": 0.6501363515853882,
+      "learning_rate": 5.551762905336159e-05,
+      "loss": 2.4186,
+      "step": 12972
+    },
+    {
+      "epoch": 1.0469695746913081,
+      "grad_norm": 0.6924965977668762,
+      "learning_rate": 5.5503490575593095e-05,
+      "loss": 2.4864,
+      "step": 12973
+    },
+    {
+      "epoch": 1.0470502784278912,
+      "grad_norm": 0.6772900819778442,
+      "learning_rate": 5.548935320679398e-05,
+      "loss": 2.4101,
+      "step": 12974
+    },
+    {
+      "epoch": 1.0471309821644743,
+      "grad_norm": 0.6950967311859131,
+      "learning_rate": 5.54752169473165e-05,
+      "loss": 2.4893,
+      "step": 12975
+    },
+    {
+      "epoch": 1.0472116859010572,
+      "grad_norm": 0.6663516163825989,
+      "learning_rate": 5.5461081797512994e-05,
+      "loss": 2.4136,
+      "step": 12976
+    },
+    {
+      "epoch": 1.0472923896376403,
+      "grad_norm": 0.7337449789047241,
+      "learning_rate": 5.5446947757735754e-05,
+      "loss": 2.473,
+      "step": 12977
+    },
+    {
+      "epoch": 1.0473730933742231,
+      "grad_norm": 0.6808840036392212,
+      "learning_rate": 5.543281482833709e-05,
+      "loss": 2.4473,
+      "step": 12978
+    },
+    {
+      "epoch": 1.0474537971108062,
+      "grad_norm": 0.6472508907318115,
+      "learning_rate": 5.5418683009669124e-05,
+      "loss": 2.4077,
+      "step": 12979
+    },
+    {
+      "epoch": 1.0475345008473893,
+      "grad_norm": 0.6904192566871643,
+      "learning_rate": 5.540455230208409e-05,
+      "loss": 2.482,
+      "step": 12980
+    },
+    {
+      "epoch": 1.0476152045839722,
+      "grad_norm": 0.6781610250473022,
+      "learning_rate": 5.5390422705934264e-05,
+      "loss": 2.4458,
+      "step": 12981
+    },
+    {
+      "epoch": 1.0476959083205553,
+      "grad_norm": 0.7130050659179688,
+      "learning_rate": 5.5376294221571666e-05,
+      "loss": 2.5136,
+      "step": 12982
+    },
+    {
+      "epoch": 1.0477766120571383,
+      "grad_norm": 0.7727184891700745,
+      "learning_rate": 5.536216684934846e-05,
+      "loss": 2.5346,
+      "step": 12983
+    },
+    {
+      "epoch": 1.0478573157937212,
+      "grad_norm": 0.7177208662033081,
+      "learning_rate": 5.534804058961679e-05,
+      "loss": 2.4153,
+      "step": 12984
+    },
+    {
+      "epoch": 1.0479380195303043,
+      "grad_norm": 0.7333023548126221,
+      "learning_rate": 5.5333915442728634e-05,
+      "loss": 2.4171,
+      "step": 12985
+    },
+    {
+      "epoch": 1.0480187232668872,
+      "grad_norm": 0.658423125743866,
+      "learning_rate": 5.5319791409036046e-05,
+      "loss": 2.446,
+      "step": 12986
+    },
+    {
+      "epoch": 1.0480994270034703,
+      "grad_norm": 0.8305184841156006,
+      "learning_rate": 5.5305668488891114e-05,
+      "loss": 2.5026,
+      "step": 12987
+    },
+    {
+      "epoch": 1.0481801307400533,
+      "grad_norm": 0.7083305716514587,
+      "learning_rate": 5.52915466826457e-05,
+      "loss": 2.5366,
+      "step": 12988
+    },
+    {
+      "epoch": 1.0482608344766362,
+      "grad_norm": 0.7924454212188721,
+      "learning_rate": 5.5277425990651824e-05,
+      "loss": 2.528,
+      "step": 12989
+    },
+    {
+      "epoch": 1.0483415382132193,
+      "grad_norm": 0.633376955986023,
+      "learning_rate": 5.5263306413261384e-05,
+      "loss": 2.4442,
+      "step": 12990
+    },
+    {
+      "epoch": 1.0484222419498024,
+      "grad_norm": 0.7387240529060364,
+      "learning_rate": 5.5249187950826295e-05,
+      "loss": 2.4761,
+      "step": 12991
+    },
+    {
+      "epoch": 1.0485029456863852,
+      "grad_norm": 0.6796224117279053,
+      "learning_rate": 5.523507060369843e-05,
+      "loss": 2.4828,
+      "step": 12992
+    },
+    {
+      "epoch": 1.0485836494229683,
+      "grad_norm": 0.6925581097602844,
+      "learning_rate": 5.5220954372229604e-05,
+      "loss": 2.4861,
+      "step": 12993
+    },
+    {
+      "epoch": 1.0486643531595512,
+      "grad_norm": 0.6854318380355835,
+      "learning_rate": 5.5206839256771704e-05,
+      "loss": 2.473,
+      "step": 12994
+    },
+    {
+      "epoch": 1.0487450568961343,
+      "grad_norm": 0.706375241279602,
+      "learning_rate": 5.519272525767643e-05,
+      "loss": 2.4284,
+      "step": 12995
+    },
+    {
+      "epoch": 1.0488257606327174,
+      "grad_norm": 0.6917428374290466,
+      "learning_rate": 5.517861237529556e-05,
+      "loss": 2.4702,
+      "step": 12996
+    },
+    {
+      "epoch": 1.0489064643693002,
+      "grad_norm": 0.6903818845748901,
+      "learning_rate": 5.516450060998086e-05,
+      "loss": 2.4679,
+      "step": 12997
+    },
+    {
+      "epoch": 1.0489871681058833,
+      "grad_norm": 0.6403356194496155,
+      "learning_rate": 5.515038996208398e-05,
+      "loss": 2.396,
+      "step": 12998
+    },
+    {
+      "epoch": 1.0490678718424662,
+      "grad_norm": 0.6491792798042297,
+      "learning_rate": 5.513628043195662e-05,
+      "loss": 2.4543,
+      "step": 12999
+    },
+    {
+      "epoch": 1.0491485755790493,
+      "grad_norm": 0.687303900718689,
+      "learning_rate": 5.512217201995043e-05,
+      "loss": 2.4716,
+      "step": 13000
+    },
+    {
+      "epoch": 1.0491485755790493,
+      "eval_loss": 2.4177169799804688,
+      "eval_runtime": 763.9215,
+      "eval_samples_per_second": 3.43,
+      "eval_steps_per_second": 0.572,
+      "step": 13000
+    },
+    {
+      "epoch": 1.0492292793156324,
+      "grad_norm": 0.7020761370658875,
+      "learning_rate": 5.510806472641701e-05,
+      "loss": 2.3591,
+      "step": 13001
+    },
+    {
+      "epoch": 1.0493099830522152,
+      "grad_norm": 0.6978075504302979,
+      "learning_rate": 5.509395855170798e-05,
+      "loss": 2.4585,
+      "step": 13002
+    },
+    {
+      "epoch": 1.0493906867887983,
+      "grad_norm": 0.7327752113342285,
+      "learning_rate": 5.5079853496174925e-05,
+      "loss": 2.5265,
+      "step": 13003
+    },
+    {
+      "epoch": 1.0494713905253814,
+      "grad_norm": 0.7552505135536194,
+      "learning_rate": 5.50657495601693e-05,
+      "loss": 2.4821,
+      "step": 13004
+    },
+    {
+      "epoch": 1.0495520942619643,
+      "grad_norm": 0.7100770473480225,
+      "learning_rate": 5.5051646744042664e-05,
+      "loss": 2.4566,
+      "step": 13005
+    },
+    {
+      "epoch": 1.0496327979985474,
+      "grad_norm": 0.7008209824562073,
+      "learning_rate": 5.503754504814651e-05,
+      "loss": 2.4476,
+      "step": 13006
+    },
+    {
+      "epoch": 1.0497135017351304,
+      "grad_norm": 0.640724241733551,
+      "learning_rate": 5.502344447283223e-05,
+      "loss": 2.437,
+      "step": 13007
+    },
+    {
+      "epoch": 1.0497942054717133,
+      "grad_norm": 0.7064981460571289,
+      "learning_rate": 5.5009345018451297e-05,
+      "loss": 2.5129,
+      "step": 13008
+    },
+    {
+      "epoch": 1.0498749092082964,
+      "grad_norm": 0.6729782223701477,
+      "learning_rate": 5.49952466853551e-05,
+      "loss": 2.4867,
+      "step": 13009
+    },
+    {
+      "epoch": 1.0499556129448793,
+      "grad_norm": 0.7245302200317383,
+      "learning_rate": 5.4981149473894966e-05,
+      "loss": 2.4485,
+      "step": 13010
+    },
+    {
+      "epoch": 1.0500363166814624,
+      "grad_norm": 0.6686248779296875,
+      "learning_rate": 5.4967053384422294e-05,
+      "loss": 2.4314,
+      "step": 13011
+    },
+    {
+      "epoch": 1.0501170204180454,
+      "grad_norm": 0.6790863871574402,
+      "learning_rate": 5.495295841728836e-05,
+      "loss": 2.4847,
+      "step": 13012
+    },
+    {
+      "epoch": 1.0501977241546283,
+      "grad_norm": 0.6516931653022766,
+      "learning_rate": 5.49388645728445e-05,
+      "loss": 2.4306,
+      "step": 13013
+    },
+    {
+      "epoch": 1.0502784278912114,
+      "grad_norm": 0.6967600584030151,
+      "learning_rate": 5.492477185144189e-05,
+      "loss": 2.4942,
+      "step": 13014
+    },
+    {
+      "epoch": 1.0503591316277943,
+      "grad_norm": 0.696246325969696,
+      "learning_rate": 5.491068025343178e-05,
+      "loss": 2.4647,
+      "step": 13015
+    },
+    {
+      "epoch": 1.0504398353643774,
+      "grad_norm": 0.6962751150131226,
+      "learning_rate": 5.489658977916543e-05,
+      "loss": 2.5095,
+      "step": 13016
+    },
+    {
+      "epoch": 1.0505205391009604,
+      "grad_norm": 0.6982631087303162,
+      "learning_rate": 5.488250042899392e-05,
+      "loss": 2.4327,
+      "step": 13017
+    },
+    {
+      "epoch": 1.0506012428375433,
+      "grad_norm": 0.6932644844055176,
+      "learning_rate": 5.486841220326845e-05,
+      "loss": 2.4777,
+      "step": 13018
+    },
+    {
+      "epoch": 1.0506819465741264,
+      "grad_norm": 0.6923339366912842,
+      "learning_rate": 5.485432510234012e-05,
+      "loss": 2.4321,
+      "step": 13019
+    },
+    {
+      "epoch": 1.0507626503107095,
+      "grad_norm": 0.7445859313011169,
+      "learning_rate": 5.4840239126560015e-05,
+      "loss": 2.4425,
+      "step": 13020
+    },
+    {
+      "epoch": 1.0508433540472923,
+      "grad_norm": 0.7122324705123901,
+      "learning_rate": 5.48261542762792e-05,
+      "loss": 2.4545,
+      "step": 13021
+    },
+    {
+      "epoch": 1.0509240577838754,
+      "grad_norm": 0.734779417514801,
+      "learning_rate": 5.4812070551848736e-05,
+      "loss": 2.4764,
+      "step": 13022
+    },
+    {
+      "epoch": 1.0510047615204583,
+      "grad_norm": 0.6544109582901001,
+      "learning_rate": 5.4797987953619566e-05,
+      "loss": 2.4492,
+      "step": 13023
+    },
+    {
+      "epoch": 1.0510854652570414,
+      "grad_norm": 0.6366097331047058,
+      "learning_rate": 5.4783906481942704e-05,
+      "loss": 2.4695,
+      "step": 13024
+    },
+    {
+      "epoch": 1.0511661689936245,
+      "grad_norm": 0.6966270804405212,
+      "learning_rate": 5.476982613716908e-05,
+      "loss": 2.4505,
+      "step": 13025
+    },
+    {
+      "epoch": 1.0512468727302073,
+      "grad_norm": 0.7010120153427124,
+      "learning_rate": 5.4755746919649665e-05,
+      "loss": 2.4545,
+      "step": 13026
+    },
+    {
+      "epoch": 1.0513275764667904,
+      "grad_norm": 0.6704719662666321,
+      "learning_rate": 5.474166882973526e-05,
+      "loss": 2.3899,
+      "step": 13027
+    },
+    {
+      "epoch": 1.0514082802033735,
+      "grad_norm": 0.757152259349823,
+      "learning_rate": 5.472759186777679e-05,
+      "loss": 2.5112,
+      "step": 13028
+    },
+    {
+      "epoch": 1.0514889839399564,
+      "grad_norm": 0.6668868660926819,
+      "learning_rate": 5.471351603412509e-05,
+      "loss": 2.4797,
+      "step": 13029
+    },
+    {
+      "epoch": 1.0515696876765395,
+      "grad_norm": 0.7919496893882751,
+      "learning_rate": 5.4699441329130887e-05,
+      "loss": 2.4874,
+      "step": 13030
+    },
+    {
+      "epoch": 1.0516503914131223,
+      "grad_norm": 0.7595484852790833,
+      "learning_rate": 5.468536775314506e-05,
+      "loss": 2.4621,
+      "step": 13031
+    },
+    {
+      "epoch": 1.0517310951497054,
+      "grad_norm": 0.6575995683670044,
+      "learning_rate": 5.467129530651835e-05,
+      "loss": 2.4474,
+      "step": 13032
+    },
+    {
+      "epoch": 1.0518117988862885,
+      "grad_norm": 0.6817733645439148,
+      "learning_rate": 5.4657223989601425e-05,
+      "loss": 2.4329,
+      "step": 13033
+    },
+    {
+      "epoch": 1.0518925026228714,
+      "grad_norm": 0.722882091999054,
+      "learning_rate": 5.464315380274501e-05,
+      "loss": 2.4544,
+      "step": 13034
+    },
+    {
+      "epoch": 1.0519732063594545,
+      "grad_norm": 0.6957377791404724,
+      "learning_rate": 5.4629084746299796e-05,
+      "loss": 2.5669,
+      "step": 13035
+    },
+    {
+      "epoch": 1.0520539100960375,
+      "grad_norm": 0.6749420166015625,
+      "learning_rate": 5.461501682061636e-05,
+      "loss": 2.5053,
+      "step": 13036
+    },
+    {
+      "epoch": 1.0521346138326204,
+      "grad_norm": 0.8158369064331055,
+      "learning_rate": 5.4600950026045326e-05,
+      "loss": 2.429,
+      "step": 13037
+    },
+    {
+      "epoch": 1.0522153175692035,
+      "grad_norm": 0.6960736513137817,
+      "learning_rate": 5.458688436293735e-05,
+      "loss": 2.4731,
+      "step": 13038
+    },
+    {
+      "epoch": 1.0522960213057864,
+      "grad_norm": 0.6686301231384277,
+      "learning_rate": 5.457281983164287e-05,
+      "loss": 2.4495,
+      "step": 13039
+    },
+    {
+      "epoch": 1.0523767250423695,
+      "grad_norm": 0.6691476106643677,
+      "learning_rate": 5.455875643251248e-05,
+      "loss": 2.4329,
+      "step": 13040
+    },
+    {
+      "epoch": 1.0524574287789525,
+      "grad_norm": 0.7737297415733337,
+      "learning_rate": 5.454469416589666e-05,
+      "loss": 2.4664,
+      "step": 13041
+    },
+    {
+      "epoch": 1.0525381325155354,
+      "grad_norm": 0.7848188281059265,
+      "learning_rate": 5.453063303214588e-05,
+      "loss": 2.4799,
+      "step": 13042
+    },
+    {
+      "epoch": 1.0526188362521185,
+      "grad_norm": 0.7831119894981384,
+      "learning_rate": 5.45165730316106e-05,
+      "loss": 2.5076,
+      "step": 13043
+    },
+    {
+      "epoch": 1.0526995399887016,
+      "grad_norm": 0.691635012626648,
+      "learning_rate": 5.4502514164641196e-05,
+      "loss": 2.4866,
+      "step": 13044
+    },
+    {
+      "epoch": 1.0527802437252844,
+      "grad_norm": 0.6667110919952393,
+      "learning_rate": 5.4488456431588106e-05,
+      "loss": 2.4162,
+      "step": 13045
+    },
+    {
+      "epoch": 1.0528609474618675,
+      "grad_norm": 0.7201905846595764,
+      "learning_rate": 5.447439983280163e-05,
+      "loss": 2.498,
+      "step": 13046
+    },
+    {
+      "epoch": 1.0529416511984504,
+      "grad_norm": 0.8538106083869934,
+      "learning_rate": 5.44603443686321e-05,
+      "loss": 2.4477,
+      "step": 13047
+    },
+    {
+      "epoch": 1.0530223549350335,
+      "grad_norm": 0.6661962270736694,
+      "learning_rate": 5.444629003942987e-05,
+      "loss": 2.5253,
+      "step": 13048
+    },
+    {
+      "epoch": 1.0531030586716166,
+      "grad_norm": 0.7239834666252136,
+      "learning_rate": 5.4432236845545146e-05,
+      "loss": 2.4786,
+      "step": 13049
+    },
+    {
+      "epoch": 1.0531837624081994,
+      "grad_norm": 0.7328412532806396,
+      "learning_rate": 5.4418184787328186e-05,
+      "loss": 2.4841,
+      "step": 13050
+    },
+    {
+      "epoch": 1.0532644661447825,
+      "grad_norm": 0.6395559310913086,
+      "learning_rate": 5.440413386512922e-05,
+      "loss": 2.3544,
+      "step": 13051
+    },
+    {
+      "epoch": 1.0533451698813656,
+      "grad_norm": 0.6632471084594727,
+      "learning_rate": 5.43900840792984e-05,
+      "loss": 2.4753,
+      "step": 13052
+    },
+    {
+      "epoch": 1.0534258736179485,
+      "grad_norm": 0.7262828350067139,
+      "learning_rate": 5.4376035430185935e-05,
+      "loss": 2.4162,
+      "step": 13053
+    },
+    {
+      "epoch": 1.0535065773545316,
+      "grad_norm": 0.7897952198982239,
+      "learning_rate": 5.436198791814196e-05,
+      "loss": 2.4571,
+      "step": 13054
+    },
+    {
+      "epoch": 1.0535872810911144,
+      "grad_norm": 0.7281489372253418,
+      "learning_rate": 5.434794154351651e-05,
+      "loss": 2.4531,
+      "step": 13055
+    },
+    {
+      "epoch": 1.0536679848276975,
+      "grad_norm": 0.7322356700897217,
+      "learning_rate": 5.4333896306659694e-05,
+      "loss": 2.4102,
+      "step": 13056
+    },
+    {
+      "epoch": 1.0537486885642806,
+      "grad_norm": 0.7657945156097412,
+      "learning_rate": 5.4319852207921554e-05,
+      "loss": 2.4526,
+      "step": 13057
+    },
+    {
+      "epoch": 1.0538293923008635,
+      "grad_norm": 0.6732973456382751,
+      "learning_rate": 5.430580924765214e-05,
+      "loss": 2.4516,
+      "step": 13058
+    },
+    {
+      "epoch": 1.0539100960374466,
+      "grad_norm": 0.663398027420044,
+      "learning_rate": 5.429176742620137e-05,
+      "loss": 2.4437,
+      "step": 13059
+    },
+    {
+      "epoch": 1.0539907997740294,
+      "grad_norm": 0.6363258957862854,
+      "learning_rate": 5.4277726743919244e-05,
+      "loss": 2.414,
+      "step": 13060
+    },
+    {
+      "epoch": 1.0540715035106125,
+      "grad_norm": 0.6600647568702698,
+      "learning_rate": 5.426368720115568e-05,
+      "loss": 2.4319,
+      "step": 13061
+    },
+    {
+      "epoch": 1.0541522072471956,
+      "grad_norm": 0.6941983699798584,
+      "learning_rate": 5.4249648798260574e-05,
+      "loss": 2.5247,
+      "step": 13062
+    },
+    {
+      "epoch": 1.0542329109837785,
+      "grad_norm": 0.7419719099998474,
+      "learning_rate": 5.423561153558383e-05,
+      "loss": 2.5088,
+      "step": 13063
+    },
+    {
+      "epoch": 1.0543136147203616,
+      "grad_norm": 0.708073079586029,
+      "learning_rate": 5.4221575413475326e-05,
+      "loss": 2.4037,
+      "step": 13064
+    },
+    {
+      "epoch": 1.0543943184569446,
+      "grad_norm": 0.7081628441810608,
+      "learning_rate": 5.4207540432284764e-05,
+      "loss": 2.4556,
+      "step": 13065
+    },
+    {
+      "epoch": 1.0544750221935275,
+      "grad_norm": 0.7058689594268799,
+      "learning_rate": 5.419350659236201e-05,
+      "loss": 2.4244,
+      "step": 13066
+    },
+    {
+      "epoch": 1.0545557259301106,
+      "grad_norm": 0.6858707070350647,
+      "learning_rate": 5.417947389405684e-05,
+      "loss": 2.4431,
+      "step": 13067
+    },
+    {
+      "epoch": 1.0546364296666935,
+      "grad_norm": 0.6769983768463135,
+      "learning_rate": 5.416544233771893e-05,
+      "loss": 2.4257,
+      "step": 13068
+    },
+    {
+      "epoch": 1.0547171334032766,
+      "grad_norm": 0.7128089070320129,
+      "learning_rate": 5.4151411923698e-05,
+      "loss": 2.4558,
+      "step": 13069
+    },
+    {
+      "epoch": 1.0547978371398596,
+      "grad_norm": 0.6419198513031006,
+      "learning_rate": 5.413738265234374e-05,
+      "loss": 2.4421,
+      "step": 13070
+    },
+    {
+      "epoch": 1.0548785408764425,
+      "grad_norm": 0.760848879814148,
+      "learning_rate": 5.4123354524005784e-05,
+      "loss": 2.4427,
+      "step": 13071
+    },
+    {
+      "epoch": 1.0549592446130256,
+      "grad_norm": 0.6749173998832703,
+      "learning_rate": 5.410932753903377e-05,
+      "loss": 2.4902,
+      "step": 13072
+    },
+    {
+      "epoch": 1.0550399483496087,
+      "grad_norm": 0.6908800601959229,
+      "learning_rate": 5.4095301697777265e-05,
+      "loss": 2.4219,
+      "step": 13073
+    },
+    {
+      "epoch": 1.0551206520861915,
+      "grad_norm": 0.6779965758323669,
+      "learning_rate": 5.408127700058587e-05,
+      "loss": 2.4533,
+      "step": 13074
+    },
+    {
+      "epoch": 1.0552013558227746,
+      "grad_norm": 0.6832355260848999,
+      "learning_rate": 5.406725344780906e-05,
+      "loss": 2.418,
+      "step": 13075
+    },
+    {
+      "epoch": 1.0552820595593575,
+      "grad_norm": 0.6766698956489563,
+      "learning_rate": 5.4053231039796357e-05,
+      "loss": 2.4493,
+      "step": 13076
+    },
+    {
+      "epoch": 1.0553627632959406,
+      "grad_norm": 0.7256276607513428,
+      "learning_rate": 5.4039209776897285e-05,
+      "loss": 2.4126,
+      "step": 13077
+    },
+    {
+      "epoch": 1.0554434670325237,
+      "grad_norm": 0.6687275171279907,
+      "learning_rate": 5.4025189659461196e-05,
+      "loss": 2.435,
+      "step": 13078
+    },
+    {
+      "epoch": 1.0555241707691065,
+      "grad_norm": 0.6800444722175598,
+      "learning_rate": 5.401117068783758e-05,
+      "loss": 2.4608,
+      "step": 13079
+    },
+    {
+      "epoch": 1.0556048745056896,
+      "grad_norm": 0.6947116851806641,
+      "learning_rate": 5.399715286237583e-05,
+      "loss": 2.4908,
+      "step": 13080
+    },
+    {
+      "epoch": 1.0556855782422727,
+      "grad_norm": 0.6907915472984314,
+      "learning_rate": 5.398313618342521e-05,
+      "loss": 2.4805,
+      "step": 13081
+    },
+    {
+      "epoch": 1.0557662819788556,
+      "grad_norm": 0.7429100275039673,
+      "learning_rate": 5.396912065133516e-05,
+      "loss": 2.458,
+      "step": 13082
+    },
+    {
+      "epoch": 1.0558469857154387,
+      "grad_norm": 0.7186924815177917,
+      "learning_rate": 5.3955106266454994e-05,
+      "loss": 2.4924,
+      "step": 13083
+    },
+    {
+      "epoch": 1.0559276894520215,
+      "grad_norm": 0.7017999887466431,
+      "learning_rate": 5.394109302913391e-05,
+      "loss": 2.4103,
+      "step": 13084
+    },
+    {
+      "epoch": 1.0560083931886046,
+      "grad_norm": 0.7318955659866333,
+      "learning_rate": 5.392708093972117e-05,
+      "loss": 2.4424,
+      "step": 13085
+    },
+    {
+      "epoch": 1.0560890969251877,
+      "grad_norm": 0.6278600692749023,
+      "learning_rate": 5.391306999856602e-05,
+      "loss": 2.4433,
+      "step": 13086
+    },
+    {
+      "epoch": 1.0561698006617706,
+      "grad_norm": 0.6895800232887268,
+      "learning_rate": 5.389906020601767e-05,
+      "loss": 2.4275,
+      "step": 13087
+    },
+    {
+      "epoch": 1.0562505043983537,
+      "grad_norm": 0.7197345495223999,
+      "learning_rate": 5.388505156242522e-05,
+      "loss": 2.4309,
+      "step": 13088
+    },
+    {
+      "epoch": 1.0563312081349367,
+      "grad_norm": 0.636433482170105,
+      "learning_rate": 5.3871044068137824e-05,
+      "loss": 2.4258,
+      "step": 13089
+    },
+    {
+      "epoch": 1.0564119118715196,
+      "grad_norm": 0.6884748339653015,
+      "learning_rate": 5.3857037723504634e-05,
+      "loss": 2.4543,
+      "step": 13090
+    },
+    {
+      "epoch": 1.0564926156081027,
+      "grad_norm": 0.7277036309242249,
+      "learning_rate": 5.384303252887464e-05,
+      "loss": 2.4911,
+      "step": 13091
+    },
+    {
+      "epoch": 1.0565733193446856,
+      "grad_norm": 0.6940809488296509,
+      "learning_rate": 5.38290284845969e-05,
+      "loss": 2.4112,
+      "step": 13092
+    },
+    {
+      "epoch": 1.0566540230812687,
+      "grad_norm": 0.6729177236557007,
+      "learning_rate": 5.3815025591020526e-05,
+      "loss": 2.4394,
+      "step": 13093
+    },
+    {
+      "epoch": 1.0567347268178517,
+      "grad_norm": 0.6941854357719421,
+      "learning_rate": 5.3801023848494416e-05,
+      "loss": 2.4263,
+      "step": 13094
+    },
+    {
+      "epoch": 1.0568154305544346,
+      "grad_norm": 0.7046812772750854,
+      "learning_rate": 5.3787023257367554e-05,
+      "loss": 2.5196,
+      "step": 13095
+    },
+    {
+      "epoch": 1.0568961342910177,
+      "grad_norm": 0.6896177530288696,
+      "learning_rate": 5.377302381798891e-05,
+      "loss": 2.4178,
+      "step": 13096
+    },
+    {
+      "epoch": 1.0569768380276008,
+      "grad_norm": 0.6693699955940247,
+      "learning_rate": 5.375902553070731e-05,
+      "loss": 2.4908,
+      "step": 13097
+    },
+    {
+      "epoch": 1.0570575417641837,
+      "grad_norm": 0.6751677989959717,
+      "learning_rate": 5.3745028395871674e-05,
+      "loss": 2.4222,
+      "step": 13098
+    },
+    {
+      "epoch": 1.0571382455007667,
+      "grad_norm": 0.7666265368461609,
+      "learning_rate": 5.373103241383088e-05,
+      "loss": 2.4965,
+      "step": 13099
+    },
+    {
+      "epoch": 1.0572189492373496,
+      "grad_norm": 0.8069329857826233,
+      "learning_rate": 5.3717037584933674e-05,
+      "loss": 2.4988,
+      "step": 13100
+    },
+    {
+      "epoch": 1.0572996529739327,
+      "grad_norm": 0.7160749435424805,
+      "learning_rate": 5.370304390952887e-05,
+      "loss": 2.4311,
+      "step": 13101
+    },
+    {
+      "epoch": 1.0573803567105158,
+      "grad_norm": 0.6936448812484741,
+      "learning_rate": 5.368905138796523e-05,
+      "loss": 2.4877,
+      "step": 13102
+    },
+    {
+      "epoch": 1.0574610604470986,
+      "grad_norm": 0.7202793955802917,
+      "learning_rate": 5.3675060020591494e-05,
+      "loss": 2.4841,
+      "step": 13103
+    },
+    {
+      "epoch": 1.0575417641836817,
+      "grad_norm": 0.7750168442726135,
+      "learning_rate": 5.366106980775636e-05,
+      "loss": 2.4828,
+      "step": 13104
+    },
+    {
+      "epoch": 1.0576224679202646,
+      "grad_norm": 0.7079972624778748,
+      "learning_rate": 5.364708074980849e-05,
+      "loss": 2.4912,
+      "step": 13105
+    },
+    {
+      "epoch": 1.0577031716568477,
+      "grad_norm": 0.704066276550293,
+      "learning_rate": 5.363309284709657e-05,
+      "loss": 2.4731,
+      "step": 13106
+    },
+    {
+      "epoch": 1.0577838753934308,
+      "grad_norm": 0.7040490508079529,
+      "learning_rate": 5.361910609996915e-05,
+      "loss": 2.3811,
+      "step": 13107
+    },
+    {
+      "epoch": 1.0578645791300136,
+      "grad_norm": 0.6669453978538513,
+      "learning_rate": 5.360512050877484e-05,
+      "loss": 2.5372,
+      "step": 13108
+    },
+    {
+      "epoch": 1.0579452828665967,
+      "grad_norm": 0.7197996973991394,
+      "learning_rate": 5.359113607386226e-05,
+      "loss": 2.4612,
+      "step": 13109
+    },
+    {
+      "epoch": 1.0580259866031798,
+      "grad_norm": 0.7192320823669434,
+      "learning_rate": 5.3577152795579824e-05,
+      "loss": 2.4636,
+      "step": 13110
+    },
+    {
+      "epoch": 1.0581066903397627,
+      "grad_norm": 0.6907937526702881,
+      "learning_rate": 5.35631706742761e-05,
+      "loss": 2.4791,
+      "step": 13111
+    },
+    {
+      "epoch": 1.0581873940763458,
+      "grad_norm": 0.687035083770752,
+      "learning_rate": 5.354918971029954e-05,
+      "loss": 2.4706,
+      "step": 13112
+    },
+    {
+      "epoch": 1.0582680978129286,
+      "grad_norm": 0.6666533350944519,
+      "learning_rate": 5.353520990399861e-05,
+      "loss": 2.4789,
+      "step": 13113
+    },
+    {
+      "epoch": 1.0583488015495117,
+      "grad_norm": 0.6261809468269348,
+      "learning_rate": 5.35212312557217e-05,
+      "loss": 2.4485,
+      "step": 13114
+    },
+    {
+      "epoch": 1.0584295052860948,
+      "grad_norm": 0.6740814447402954,
+      "learning_rate": 5.350725376581725e-05,
+      "loss": 2.47,
+      "step": 13115
+    },
+    {
+      "epoch": 1.0585102090226777,
+      "grad_norm": 0.7634154558181763,
+      "learning_rate": 5.3493277434633526e-05,
+      "loss": 2.4685,
+      "step": 13116
+    },
+    {
+      "epoch": 1.0585909127592608,
+      "grad_norm": 0.6674611568450928,
+      "learning_rate": 5.34793022625189e-05,
+      "loss": 2.4362,
+      "step": 13117
+    },
+    {
+      "epoch": 1.0586716164958438,
+      "grad_norm": 0.7584757804870605,
+      "learning_rate": 5.346532824982167e-05,
+      "loss": 2.499,
+      "step": 13118
+    },
+    {
+      "epoch": 1.0587523202324267,
+      "grad_norm": 0.6453456282615662,
+      "learning_rate": 5.345135539689015e-05,
+      "loss": 2.4341,
+      "step": 13119
+    },
+    {
+      "epoch": 1.0588330239690098,
+      "grad_norm": 0.70013427734375,
+      "learning_rate": 5.343738370407247e-05,
+      "loss": 2.3448,
+      "step": 13120
+    },
+    {
+      "epoch": 1.0589137277055927,
+      "grad_norm": 0.6763362884521484,
+      "learning_rate": 5.342341317171693e-05,
+      "loss": 2.4234,
+      "step": 13121
+    },
+    {
+      "epoch": 1.0589944314421758,
+      "grad_norm": 0.6896576881408691,
+      "learning_rate": 5.3409443800171664e-05,
+      "loss": 2.4753,
+      "step": 13122
+    },
+    {
+      "epoch": 1.0590751351787588,
+      "grad_norm": 0.6984997987747192,
+      "learning_rate": 5.339547558978486e-05,
+      "loss": 2.4581,
+      "step": 13123
+    },
+    {
+      "epoch": 1.0591558389153417,
+      "grad_norm": 0.7276118993759155,
+      "learning_rate": 5.338150854090462e-05,
+      "loss": 2.4765,
+      "step": 13124
+    },
+    {
+      "epoch": 1.0592365426519248,
+      "grad_norm": 0.6943252086639404,
+      "learning_rate": 5.336754265387911e-05,
+      "loss": 2.4514,
+      "step": 13125
+    },
+    {
+      "epoch": 1.0593172463885079,
+      "grad_norm": 0.7070014476776123,
+      "learning_rate": 5.335357792905628e-05,
+      "loss": 2.4365,
+      "step": 13126
+    },
+    {
+      "epoch": 1.0593979501250907,
+      "grad_norm": 0.6887189149856567,
+      "learning_rate": 5.333961436678422e-05,
+      "loss": 2.4834,
+      "step": 13127
+    },
+    {
+      "epoch": 1.0594786538616738,
+      "grad_norm": 0.8150162696838379,
+      "learning_rate": 5.332565196741098e-05,
+      "loss": 2.4474,
+      "step": 13128
+    },
+    {
+      "epoch": 1.0595593575982567,
+      "grad_norm": 0.6681316494941711,
+      "learning_rate": 5.331169073128447e-05,
+      "loss": 2.4888,
+      "step": 13129
+    },
+    {
+      "epoch": 1.0596400613348398,
+      "grad_norm": 0.6696690320968628,
+      "learning_rate": 5.329773065875267e-05,
+      "loss": 2.3874,
+      "step": 13130
+    },
+    {
+      "epoch": 1.0597207650714229,
+      "grad_norm": 0.729807436466217,
+      "learning_rate": 5.32837717501635e-05,
+      "loss": 2.4442,
+      "step": 13131
+    },
+    {
+      "epoch": 1.0598014688080057,
+      "grad_norm": 0.6959047913551331,
+      "learning_rate": 5.326981400586486e-05,
+      "loss": 2.4697,
+      "step": 13132
+    },
+    {
+      "epoch": 1.0598821725445888,
+      "grad_norm": 0.667294442653656,
+      "learning_rate": 5.3255857426204606e-05,
+      "loss": 2.3986,
+      "step": 13133
+    },
+    {
+      "epoch": 1.059962876281172,
+      "grad_norm": 0.6953842639923096,
+      "learning_rate": 5.3241902011530566e-05,
+      "loss": 2.396,
+      "step": 13134
+    },
+    {
+      "epoch": 1.0600435800177548,
+      "grad_norm": 0.6544597148895264,
+      "learning_rate": 5.32279477621906e-05,
+      "loss": 2.426,
+      "step": 13135
+    },
+    {
+      "epoch": 1.0601242837543379,
+      "grad_norm": 0.708017885684967,
+      "learning_rate": 5.321399467853241e-05,
+      "loss": 2.4931,
+      "step": 13136
+    },
+    {
+      "epoch": 1.0602049874909207,
+      "grad_norm": 0.6669809818267822,
+      "learning_rate": 5.3200042760903764e-05,
+      "loss": 2.4354,
+      "step": 13137
+    },
+    {
+      "epoch": 1.0602856912275038,
+      "grad_norm": 1.0144098997116089,
+      "learning_rate": 5.3186092009652435e-05,
+      "loss": 2.4803,
+      "step": 13138
+    },
+    {
+      "epoch": 1.060366394964087,
+      "grad_norm": 0.7213768362998962,
+      "learning_rate": 5.317214242512601e-05,
+      "loss": 2.4318,
+      "step": 13139
+    },
+    {
+      "epoch": 1.0604470987006698,
+      "grad_norm": 0.6429069638252258,
+      "learning_rate": 5.315819400767223e-05,
+      "loss": 2.458,
+      "step": 13140
+    },
+    {
+      "epoch": 1.0605278024372529,
+      "grad_norm": 0.6480485796928406,
+      "learning_rate": 5.3144246757638714e-05,
+      "loss": 2.4586,
+      "step": 13141
+    },
+    {
+      "epoch": 1.060608506173836,
+      "grad_norm": 0.7037697434425354,
+      "learning_rate": 5.3130300675373035e-05,
+      "loss": 2.4698,
+      "step": 13142
+    },
+    {
+      "epoch": 1.0606892099104188,
+      "grad_norm": 0.7307559251785278,
+      "learning_rate": 5.3116355761222725e-05,
+      "loss": 2.4027,
+      "step": 13143
+    },
+    {
+      "epoch": 1.060769913647002,
+      "grad_norm": 0.6684615612030029,
+      "learning_rate": 5.310241201553547e-05,
+      "loss": 2.478,
+      "step": 13144
+    },
+    {
+      "epoch": 1.0608506173835848,
+      "grad_norm": 0.7018016576766968,
+      "learning_rate": 5.308846943865866e-05,
+      "loss": 2.4229,
+      "step": 13145
+    },
+    {
+      "epoch": 1.0609313211201679,
+      "grad_norm": 0.7538621425628662,
+      "learning_rate": 5.307452803093982e-05,
+      "loss": 2.5201,
+      "step": 13146
+    },
+    {
+      "epoch": 1.061012024856751,
+      "grad_norm": 0.6957963109016418,
+      "learning_rate": 5.306058779272645e-05,
+      "loss": 2.4233,
+      "step": 13147
+    },
+    {
+      "epoch": 1.0610927285933338,
+      "grad_norm": 0.6280590295791626,
+      "learning_rate": 5.304664872436588e-05,
+      "loss": 2.5117,
+      "step": 13148
+    },
+    {
+      "epoch": 1.061173432329917,
+      "grad_norm": 0.6937280297279358,
+      "learning_rate": 5.3032710826205564e-05,
+      "loss": 2.4889,
+      "step": 13149
+    },
+    {
+      "epoch": 1.0612541360664998,
+      "grad_norm": 0.6750391125679016,
+      "learning_rate": 5.3018774098592884e-05,
+      "loss": 2.4472,
+      "step": 13150
+    },
+    {
+      "epoch": 1.0613348398030829,
+      "grad_norm": 0.6931902766227722,
+      "learning_rate": 5.300483854187519e-05,
+      "loss": 2.3883,
+      "step": 13151
+    },
+    {
+      "epoch": 1.061415543539666,
+      "grad_norm": 0.6982774138450623,
+      "learning_rate": 5.2990904156399726e-05,
+      "loss": 2.4688,
+      "step": 13152
+    },
+    {
+      "epoch": 1.0614962472762488,
+      "grad_norm": 0.6873522996902466,
+      "learning_rate": 5.297697094251382e-05,
+      "loss": 2.4818,
+      "step": 13153
+    },
+    {
+      "epoch": 1.061576951012832,
+      "grad_norm": 0.635377049446106,
+      "learning_rate": 5.296303890056471e-05,
+      "loss": 2.3906,
+      "step": 13154
+    },
+    {
+      "epoch": 1.061657654749415,
+      "grad_norm": 0.6368159651756287,
+      "learning_rate": 5.294910803089963e-05,
+      "loss": 2.4714,
+      "step": 13155
+    },
+    {
+      "epoch": 1.0617383584859978,
+      "grad_norm": 0.7147238254547119,
+      "learning_rate": 5.293517833386576e-05,
+      "loss": 2.4746,
+      "step": 13156
+    },
+    {
+      "epoch": 1.061819062222581,
+      "grad_norm": 0.742189884185791,
+      "learning_rate": 5.2921249809810326e-05,
+      "loss": 2.3913,
+      "step": 13157
+    },
+    {
+      "epoch": 1.061899765959164,
+      "grad_norm": 0.6665734648704529,
+      "learning_rate": 5.290732245908038e-05,
+      "loss": 2.4263,
+      "step": 13158
+    },
+    {
+      "epoch": 1.0619804696957469,
+      "grad_norm": 0.6894757747650146,
+      "learning_rate": 5.2893396282023055e-05,
+      "loss": 2.4204,
+      "step": 13159
+    },
+    {
+      "epoch": 1.06206117343233,
+      "grad_norm": 0.6394561529159546,
+      "learning_rate": 5.287947127898546e-05,
+      "loss": 2.4183,
+      "step": 13160
+    },
+    {
+      "epoch": 1.0621418771689128,
+      "grad_norm": 0.7422548532485962,
+      "learning_rate": 5.2865547450314576e-05,
+      "loss": 2.4454,
+      "step": 13161
+    },
+    {
+      "epoch": 1.062222580905496,
+      "grad_norm": 0.7486133575439453,
+      "learning_rate": 5.285162479635748e-05,
+      "loss": 2.4856,
+      "step": 13162
+    },
+    {
+      "epoch": 1.062303284642079,
+      "grad_norm": 0.6743031144142151,
+      "learning_rate": 5.283770331746112e-05,
+      "loss": 2.4318,
+      "step": 13163
+    },
+    {
+      "epoch": 1.0623839883786619,
+      "grad_norm": 0.6461686491966248,
+      "learning_rate": 5.282378301397248e-05,
+      "loss": 2.4133,
+      "step": 13164
+    },
+    {
+      "epoch": 1.062464692115245,
+      "grad_norm": 0.6745431423187256,
+      "learning_rate": 5.28098638862385e-05,
+      "loss": 2.4463,
+      "step": 13165
+    },
+    {
+      "epoch": 1.0625453958518278,
+      "grad_norm": 0.6646310687065125,
+      "learning_rate": 5.279594593460606e-05,
+      "loss": 2.4211,
+      "step": 13166
+    },
+    {
+      "epoch": 1.062626099588411,
+      "grad_norm": 0.6789249777793884,
+      "learning_rate": 5.278202915942207e-05,
+      "loss": 2.4832,
+      "step": 13167
+    },
+    {
+      "epoch": 1.062706803324994,
+      "grad_norm": 0.7082679867744446,
+      "learning_rate": 5.2768113561033326e-05,
+      "loss": 2.4303,
+      "step": 13168
+    },
+    {
+      "epoch": 1.0627875070615769,
+      "grad_norm": 0.6875587701797485,
+      "learning_rate": 5.275419913978664e-05,
+      "loss": 2.4601,
+      "step": 13169
+    },
+    {
+      "epoch": 1.06286821079816,
+      "grad_norm": 0.6556203961372375,
+      "learning_rate": 5.274028589602886e-05,
+      "loss": 2.4359,
+      "step": 13170
+    },
+    {
+      "epoch": 1.062948914534743,
+      "grad_norm": 0.7280015349388123,
+      "learning_rate": 5.272637383010666e-05,
+      "loss": 2.4999,
+      "step": 13171
+    },
+    {
+      "epoch": 1.063029618271326,
+      "grad_norm": 0.664654016494751,
+      "learning_rate": 5.271246294236678e-05,
+      "loss": 2.3951,
+      "step": 13172
+    },
+    {
+      "epoch": 1.063110322007909,
+      "grad_norm": 0.6941719055175781,
+      "learning_rate": 5.2698553233155945e-05,
+      "loss": 2.45,
+      "step": 13173
+    },
+    {
+      "epoch": 1.0631910257444919,
+      "grad_norm": 0.7212931513786316,
+      "learning_rate": 5.268464470282082e-05,
+      "loss": 2.4615,
+      "step": 13174
+    },
+    {
+      "epoch": 1.063271729481075,
+      "grad_norm": 0.6877106428146362,
+      "learning_rate": 5.2670737351708014e-05,
+      "loss": 2.4495,
+      "step": 13175
+    },
+    {
+      "epoch": 1.063352433217658,
+      "grad_norm": 0.737718939781189,
+      "learning_rate": 5.26568311801642e-05,
+      "loss": 2.4971,
+      "step": 13176
+    },
+    {
+      "epoch": 1.063433136954241,
+      "grad_norm": 0.6909129619598389,
+      "learning_rate": 5.264292618853587e-05,
+      "loss": 2.4889,
+      "step": 13177
+    },
+    {
+      "epoch": 1.063513840690824,
+      "grad_norm": 0.6750304102897644,
+      "learning_rate": 5.262902237716961e-05,
+      "loss": 2.4779,
+      "step": 13178
+    },
+    {
+      "epoch": 1.063594544427407,
+      "grad_norm": 0.7256019115447998,
+      "learning_rate": 5.2615119746411954e-05,
+      "loss": 2.4904,
+      "step": 13179
+    },
+    {
+      "epoch": 1.06367524816399,
+      "grad_norm": 0.7335983514785767,
+      "learning_rate": 5.26012182966094e-05,
+      "loss": 2.4357,
+      "step": 13180
+    },
+    {
+      "epoch": 1.063755951900573,
+      "grad_norm": 0.6534200310707092,
+      "learning_rate": 5.258731802810837e-05,
+      "loss": 2.4213,
+      "step": 13181
+    },
+    {
+      "epoch": 1.063836655637156,
+      "grad_norm": 0.6899768114089966,
+      "learning_rate": 5.257341894125529e-05,
+      "loss": 2.4963,
+      "step": 13182
+    },
+    {
+      "epoch": 1.063917359373739,
+      "grad_norm": 0.7016159892082214,
+      "learning_rate": 5.25595210363966e-05,
+      "loss": 2.4583,
+      "step": 13183
+    },
+    {
+      "epoch": 1.063998063110322,
+      "grad_norm": 0.6868152022361755,
+      "learning_rate": 5.2545624313878636e-05,
+      "loss": 2.4523,
+      "step": 13184
+    },
+    {
+      "epoch": 1.064078766846905,
+      "grad_norm": 0.7442622184753418,
+      "learning_rate": 5.2531728774047785e-05,
+      "loss": 2.425,
+      "step": 13185
+    },
+    {
+      "epoch": 1.064159470583488,
+      "grad_norm": 0.6900869011878967,
+      "learning_rate": 5.251783441725037e-05,
+      "loss": 2.459,
+      "step": 13186
+    },
+    {
+      "epoch": 1.0642401743200711,
+      "grad_norm": 0.6910288333892822,
+      "learning_rate": 5.25039412438326e-05,
+      "loss": 2.4882,
+      "step": 13187
+    },
+    {
+      "epoch": 1.064320878056654,
+      "grad_norm": 0.7644359469413757,
+      "learning_rate": 5.249004925414076e-05,
+      "loss": 2.4663,
+      "step": 13188
+    },
+    {
+      "epoch": 1.064401581793237,
+      "grad_norm": 0.6703082919120789,
+      "learning_rate": 5.247615844852114e-05,
+      "loss": 2.4309,
+      "step": 13189
+    },
+    {
+      "epoch": 1.06448228552982,
+      "grad_norm": 0.6449835896492004,
+      "learning_rate": 5.246226882731983e-05,
+      "loss": 2.4307,
+      "step": 13190
+    },
+    {
+      "epoch": 1.064562989266403,
+      "grad_norm": 0.7332713603973389,
+      "learning_rate": 5.244838039088305e-05,
+      "loss": 2.3763,
+      "step": 13191
+    },
+    {
+      "epoch": 1.0646436930029861,
+      "grad_norm": 0.7626641988754272,
+      "learning_rate": 5.2434493139556974e-05,
+      "loss": 2.4167,
+      "step": 13192
+    },
+    {
+      "epoch": 1.064724396739569,
+      "grad_norm": 0.6924002170562744,
+      "learning_rate": 5.2420607073687614e-05,
+      "loss": 2.4751,
+      "step": 13193
+    },
+    {
+      "epoch": 1.064805100476152,
+      "grad_norm": 0.6815003156661987,
+      "learning_rate": 5.2406722193621074e-05,
+      "loss": 2.4731,
+      "step": 13194
+    },
+    {
+      "epoch": 1.064885804212735,
+      "grad_norm": 0.7632609009742737,
+      "learning_rate": 5.239283849970347e-05,
+      "loss": 2.4562,
+      "step": 13195
+    },
+    {
+      "epoch": 1.064966507949318,
+      "grad_norm": 0.7157592177391052,
+      "learning_rate": 5.23789559922808e-05,
+      "loss": 2.4507,
+      "step": 13196
+    },
+    {
+      "epoch": 1.065047211685901,
+      "grad_norm": 0.7035543918609619,
+      "learning_rate": 5.2365074671699e-05,
+      "loss": 2.4616,
+      "step": 13197
+    },
+    {
+      "epoch": 1.065127915422484,
+      "grad_norm": 0.7566644549369812,
+      "learning_rate": 5.235119453830406e-05,
+      "loss": 2.4751,
+      "step": 13198
+    },
+    {
+      "epoch": 1.065208619159067,
+      "grad_norm": 0.7030916213989258,
+      "learning_rate": 5.233731559244194e-05,
+      "loss": 2.381,
+      "step": 13199
+    },
+    {
+      "epoch": 1.0652893228956501,
+      "grad_norm": 0.7663755416870117,
+      "learning_rate": 5.232343783445847e-05,
+      "loss": 2.4822,
+      "step": 13200
+    },
+    {
+      "epoch": 1.065370026632233,
+      "grad_norm": 0.717767596244812,
+      "learning_rate": 5.230956126469955e-05,
+      "loss": 2.4807,
+      "step": 13201
+    },
+    {
+      "epoch": 1.065450730368816,
+      "grad_norm": 0.6920818090438843,
+      "learning_rate": 5.229568588351108e-05,
+      "loss": 2.4643,
+      "step": 13202
+    },
+    {
+      "epoch": 1.0655314341053992,
+      "grad_norm": 0.6812553405761719,
+      "learning_rate": 5.228181169123877e-05,
+      "loss": 2.4443,
+      "step": 13203
+    },
+    {
+      "epoch": 1.065612137841982,
+      "grad_norm": 0.7241889834403992,
+      "learning_rate": 5.226793868822846e-05,
+      "loss": 2.4581,
+      "step": 13204
+    },
+    {
+      "epoch": 1.0656928415785651,
+      "grad_norm": 0.7254642248153687,
+      "learning_rate": 5.225406687482588e-05,
+      "loss": 2.4999,
+      "step": 13205
+    },
+    {
+      "epoch": 1.065773545315148,
+      "grad_norm": 0.7316950559616089,
+      "learning_rate": 5.2240196251376764e-05,
+      "loss": 2.4493,
+      "step": 13206
+    },
+    {
+      "epoch": 1.065854249051731,
+      "grad_norm": 0.7208307385444641,
+      "learning_rate": 5.22263268182268e-05,
+      "loss": 2.5083,
+      "step": 13207
+    },
+    {
+      "epoch": 1.0659349527883142,
+      "grad_norm": 0.6552214622497559,
+      "learning_rate": 5.22124585757217e-05,
+      "loss": 2.4662,
+      "step": 13208
+    },
+    {
+      "epoch": 1.066015656524897,
+      "grad_norm": 0.7949681878089905,
+      "learning_rate": 5.219859152420701e-05,
+      "loss": 2.4584,
+      "step": 13209
+    },
+    {
+      "epoch": 1.0660963602614801,
+      "grad_norm": 0.7012154459953308,
+      "learning_rate": 5.2184725664028366e-05,
+      "loss": 2.4702,
+      "step": 13210
+    },
+    {
+      "epoch": 1.066177063998063,
+      "grad_norm": 0.7431927919387817,
+      "learning_rate": 5.217086099553136e-05,
+      "loss": 2.4422,
+      "step": 13211
+    },
+    {
+      "epoch": 1.066257767734646,
+      "grad_norm": 0.7235366702079773,
+      "learning_rate": 5.2156997519061554e-05,
+      "loss": 2.4173,
+      "step": 13212
+    },
+    {
+      "epoch": 1.0663384714712292,
+      "grad_norm": 0.7475029826164246,
+      "learning_rate": 5.214313523496439e-05,
+      "loss": 2.4924,
+      "step": 13213
+    },
+    {
+      "epoch": 1.066419175207812,
+      "grad_norm": 0.6326786875724792,
+      "learning_rate": 5.212927414358542e-05,
+      "loss": 2.4154,
+      "step": 13214
+    },
+    {
+      "epoch": 1.0664998789443951,
+      "grad_norm": 0.6755837798118591,
+      "learning_rate": 5.211541424527004e-05,
+      "loss": 2.4248,
+      "step": 13215
+    },
+    {
+      "epoch": 1.0665805826809782,
+      "grad_norm": 0.645395040512085,
+      "learning_rate": 5.210155554036373e-05,
+      "loss": 2.4078,
+      "step": 13216
+    },
+    {
+      "epoch": 1.066661286417561,
+      "grad_norm": 0.799913763999939,
+      "learning_rate": 5.208769802921185e-05,
+      "loss": 2.5067,
+      "step": 13217
+    },
+    {
+      "epoch": 1.0667419901541442,
+      "grad_norm": 0.7056344747543335,
+      "learning_rate": 5.207384171215983e-05,
+      "loss": 2.4817,
+      "step": 13218
+    },
+    {
+      "epoch": 1.0668226938907273,
+      "grad_norm": 0.7082187533378601,
+      "learning_rate": 5.205998658955291e-05,
+      "loss": 2.4495,
+      "step": 13219
+    },
+    {
+      "epoch": 1.0669033976273101,
+      "grad_norm": 0.6948464512825012,
+      "learning_rate": 5.204613266173646e-05,
+      "loss": 2.4584,
+      "step": 13220
+    },
+    {
+      "epoch": 1.0669841013638932,
+      "grad_norm": 0.7812542915344238,
+      "learning_rate": 5.203227992905575e-05,
+      "loss": 2.4803,
+      "step": 13221
+    },
+    {
+      "epoch": 1.067064805100476,
+      "grad_norm": 0.6892200708389282,
+      "learning_rate": 5.201842839185598e-05,
+      "loss": 2.4424,
+      "step": 13222
+    },
+    {
+      "epoch": 1.0671455088370592,
+      "grad_norm": 0.6982070803642273,
+      "learning_rate": 5.20045780504824e-05,
+      "loss": 2.4654,
+      "step": 13223
+    },
+    {
+      "epoch": 1.0672262125736423,
+      "grad_norm": 0.6799101233482361,
+      "learning_rate": 5.1990728905280205e-05,
+      "loss": 2.4748,
+      "step": 13224
+    },
+    {
+      "epoch": 1.0673069163102251,
+      "grad_norm": 0.6703687906265259,
+      "learning_rate": 5.1976880956594544e-05,
+      "loss": 2.4459,
+      "step": 13225
+    },
+    {
+      "epoch": 1.0673876200468082,
+      "grad_norm": 0.6821435689926147,
+      "learning_rate": 5.196303420477053e-05,
+      "loss": 2.4517,
+      "step": 13226
+    },
+    {
+      "epoch": 1.067468323783391,
+      "grad_norm": 0.6369695067405701,
+      "learning_rate": 5.194918865015328e-05,
+      "loss": 2.4388,
+      "step": 13227
+    },
+    {
+      "epoch": 1.0675490275199742,
+      "grad_norm": 0.6465736627578735,
+      "learning_rate": 5.1935344293087885e-05,
+      "loss": 2.3839,
+      "step": 13228
+    },
+    {
+      "epoch": 1.0676297312565572,
+      "grad_norm": 0.6745415329933167,
+      "learning_rate": 5.192150113391933e-05,
+      "loss": 2.4676,
+      "step": 13229
+    },
+    {
+      "epoch": 1.0677104349931401,
+      "grad_norm": 0.7605211138725281,
+      "learning_rate": 5.190765917299263e-05,
+      "loss": 2.4764,
+      "step": 13230
+    },
+    {
+      "epoch": 1.0677911387297232,
+      "grad_norm": 0.7040959596633911,
+      "learning_rate": 5.1893818410652825e-05,
+      "loss": 2.4727,
+      "step": 13231
+    },
+    {
+      "epoch": 1.0678718424663063,
+      "grad_norm": 0.6718928813934326,
+      "learning_rate": 5.1879978847244785e-05,
+      "loss": 2.4308,
+      "step": 13232
+    },
+    {
+      "epoch": 1.0679525462028892,
+      "grad_norm": 0.6788188219070435,
+      "learning_rate": 5.1866140483113445e-05,
+      "loss": 2.4278,
+      "step": 13233
+    },
+    {
+      "epoch": 1.0680332499394722,
+      "grad_norm": 0.7310218811035156,
+      "learning_rate": 5.185230331860371e-05,
+      "loss": 2.4585,
+      "step": 13234
+    },
+    {
+      "epoch": 1.068113953676055,
+      "grad_norm": 0.8092277646064758,
+      "learning_rate": 5.183846735406044e-05,
+      "loss": 2.4128,
+      "step": 13235
+    },
+    {
+      "epoch": 1.0681946574126382,
+      "grad_norm": 0.6469862461090088,
+      "learning_rate": 5.182463258982846e-05,
+      "loss": 2.4315,
+      "step": 13236
+    },
+    {
+      "epoch": 1.0682753611492213,
+      "grad_norm": 0.7948115468025208,
+      "learning_rate": 5.181079902625261e-05,
+      "loss": 2.5127,
+      "step": 13237
+    },
+    {
+      "epoch": 1.0683560648858041,
+      "grad_norm": 0.6988852620124817,
+      "learning_rate": 5.179696666367757e-05,
+      "loss": 2.432,
+      "step": 13238
+    },
+    {
+      "epoch": 1.0684367686223872,
+      "grad_norm": 0.6914555430412292,
+      "learning_rate": 5.1783135502448124e-05,
+      "loss": 2.4748,
+      "step": 13239
+    },
+    {
+      "epoch": 1.0685174723589703,
+      "grad_norm": 0.7586313486099243,
+      "learning_rate": 5.176930554290902e-05,
+      "loss": 2.4522,
+      "step": 13240
+    },
+    {
+      "epoch": 1.0685981760955532,
+      "grad_norm": 0.6763948798179626,
+      "learning_rate": 5.175547678540487e-05,
+      "loss": 2.4477,
+      "step": 13241
+    },
+    {
+      "epoch": 1.0686788798321363,
+      "grad_norm": 0.7625983357429504,
+      "learning_rate": 5.1741649230280334e-05,
+      "loss": 2.4725,
+      "step": 13242
+    },
+    {
+      "epoch": 1.0687595835687191,
+      "grad_norm": 0.6574710011482239,
+      "learning_rate": 5.172782287788005e-05,
+      "loss": 2.4212,
+      "step": 13243
+    },
+    {
+      "epoch": 1.0688402873053022,
+      "grad_norm": 0.770062267780304,
+      "learning_rate": 5.1713997728548615e-05,
+      "loss": 2.5065,
+      "step": 13244
+    },
+    {
+      "epoch": 1.0689209910418853,
+      "grad_norm": 0.7719037532806396,
+      "learning_rate": 5.170017378263057e-05,
+      "loss": 2.5082,
+      "step": 13245
+    },
+    {
+      "epoch": 1.0690016947784682,
+      "grad_norm": 0.7106119394302368,
+      "learning_rate": 5.168635104047046e-05,
+      "loss": 2.4922,
+      "step": 13246
+    },
+    {
+      "epoch": 1.0690823985150513,
+      "grad_norm": 0.711815595626831,
+      "learning_rate": 5.167252950241281e-05,
+      "loss": 2.498,
+      "step": 13247
+    },
+    {
+      "epoch": 1.0691631022516344,
+      "grad_norm": 0.6926038265228271,
+      "learning_rate": 5.165870916880201e-05,
+      "loss": 2.4464,
+      "step": 13248
+    },
+    {
+      "epoch": 1.0692438059882172,
+      "grad_norm": 0.6959360241889954,
+      "learning_rate": 5.164489003998254e-05,
+      "loss": 2.4668,
+      "step": 13249
+    },
+    {
+      "epoch": 1.0693245097248003,
+      "grad_norm": 0.7165184617042542,
+      "learning_rate": 5.1631072116298875e-05,
+      "loss": 2.4198,
+      "step": 13250
+    },
+    {
+      "epoch": 1.0694052134613832,
+      "grad_norm": 0.7133236527442932,
+      "learning_rate": 5.161725539809527e-05,
+      "loss": 2.4691,
+      "step": 13251
+    },
+    {
+      "epoch": 1.0694859171979663,
+      "grad_norm": 0.7057758569717407,
+      "learning_rate": 5.160343988571613e-05,
+      "loss": 2.466,
+      "step": 13252
+    },
+    {
+      "epoch": 1.0695666209345494,
+      "grad_norm": 0.6808326244354248,
+      "learning_rate": 5.158962557950583e-05,
+      "loss": 2.4248,
+      "step": 13253
+    },
+    {
+      "epoch": 1.0696473246711322,
+      "grad_norm": 0.7166025638580322,
+      "learning_rate": 5.1575812479808563e-05,
+      "loss": 2.4753,
+      "step": 13254
+    },
+    {
+      "epoch": 1.0697280284077153,
+      "grad_norm": 0.7395358085632324,
+      "learning_rate": 5.156200058696863e-05,
+      "loss": 2.485,
+      "step": 13255
+    },
+    {
+      "epoch": 1.0698087321442982,
+      "grad_norm": 0.681106686592102,
+      "learning_rate": 5.154818990133026e-05,
+      "loss": 2.5077,
+      "step": 13256
+    },
+    {
+      "epoch": 1.0698894358808813,
+      "grad_norm": 0.7517002820968628,
+      "learning_rate": 5.153438042323766e-05,
+      "loss": 2.5093,
+      "step": 13257
+    },
+    {
+      "epoch": 1.0699701396174643,
+      "grad_norm": 0.6516926288604736,
+      "learning_rate": 5.152057215303499e-05,
+      "loss": 2.4416,
+      "step": 13258
+    },
+    {
+      "epoch": 1.0700508433540472,
+      "grad_norm": 0.6930893063545227,
+      "learning_rate": 5.150676509106638e-05,
+      "loss": 2.506,
+      "step": 13259
+    },
+    {
+      "epoch": 1.0701315470906303,
+      "grad_norm": 0.7737041115760803,
+      "learning_rate": 5.1492959237675986e-05,
+      "loss": 2.4355,
+      "step": 13260
+    },
+    {
+      "epoch": 1.0702122508272134,
+      "grad_norm": 0.7274872660636902,
+      "learning_rate": 5.14791545932078e-05,
+      "loss": 2.5552,
+      "step": 13261
+    },
+    {
+      "epoch": 1.0702929545637963,
+      "grad_norm": 0.7112408876419067,
+      "learning_rate": 5.146535115800593e-05,
+      "loss": 2.4041,
+      "step": 13262
+    },
+    {
+      "epoch": 1.0703736583003793,
+      "grad_norm": 0.6822024583816528,
+      "learning_rate": 5.1451548932414415e-05,
+      "loss": 2.4346,
+      "step": 13263
+    },
+    {
+      "epoch": 1.0704543620369624,
+      "grad_norm": 0.6590598225593567,
+      "learning_rate": 5.1437747916777165e-05,
+      "loss": 2.3946,
+      "step": 13264
+    },
+    {
+      "epoch": 1.0705350657735453,
+      "grad_norm": 0.643014132976532,
+      "learning_rate": 5.142394811143818e-05,
+      "loss": 2.4455,
+      "step": 13265
+    },
+    {
+      "epoch": 1.0706157695101284,
+      "grad_norm": 0.6480194926261902,
+      "learning_rate": 5.141014951674139e-05,
+      "loss": 2.4304,
+      "step": 13266
+    },
+    {
+      "epoch": 1.0706964732467112,
+      "grad_norm": 0.6933526992797852,
+      "learning_rate": 5.139635213303069e-05,
+      "loss": 2.4627,
+      "step": 13267
+    },
+    {
+      "epoch": 1.0707771769832943,
+      "grad_norm": 0.6832638382911682,
+      "learning_rate": 5.138255596064995e-05,
+      "loss": 2.4645,
+      "step": 13268
+    },
+    {
+      "epoch": 1.0708578807198774,
+      "grad_norm": 0.6579757928848267,
+      "learning_rate": 5.1368760999943034e-05,
+      "loss": 2.3928,
+      "step": 13269
+    },
+    {
+      "epoch": 1.0709385844564603,
+      "grad_norm": 0.6658132672309875,
+      "learning_rate": 5.1354967251253684e-05,
+      "loss": 2.4732,
+      "step": 13270
+    },
+    {
+      "epoch": 1.0710192881930434,
+      "grad_norm": 0.7610828876495361,
+      "learning_rate": 5.13411747149257e-05,
+      "loss": 2.4781,
+      "step": 13271
+    },
+    {
+      "epoch": 1.0710999919296262,
+      "grad_norm": 0.682858943939209,
+      "learning_rate": 5.1327383391302895e-05,
+      "loss": 2.4545,
+      "step": 13272
+    },
+    {
+      "epoch": 1.0711806956662093,
+      "grad_norm": 0.7461360692977905,
+      "learning_rate": 5.131359328072887e-05,
+      "loss": 2.4647,
+      "step": 13273
+    },
+    {
+      "epoch": 1.0712613994027924,
+      "grad_norm": 0.6767961382865906,
+      "learning_rate": 5.129980438354738e-05,
+      "loss": 2.4562,
+      "step": 13274
+    },
+    {
+      "epoch": 1.0713421031393753,
+      "grad_norm": 0.6768184304237366,
+      "learning_rate": 5.1286016700102066e-05,
+      "loss": 2.4662,
+      "step": 13275
+    },
+    {
+      "epoch": 1.0714228068759584,
+      "grad_norm": 0.7022743225097656,
+      "learning_rate": 5.1272230230736554e-05,
+      "loss": 2.4321,
+      "step": 13276
+    },
+    {
+      "epoch": 1.0715035106125415,
+      "grad_norm": 0.725488007068634,
+      "learning_rate": 5.125844497579444e-05,
+      "loss": 2.457,
+      "step": 13277
+    },
+    {
+      "epoch": 1.0715842143491243,
+      "grad_norm": 0.7542931437492371,
+      "learning_rate": 5.124466093561928e-05,
+      "loss": 2.4302,
+      "step": 13278
+    },
+    {
+      "epoch": 1.0716649180857074,
+      "grad_norm": 0.6598316431045532,
+      "learning_rate": 5.123087811055467e-05,
+      "loss": 2.4552,
+      "step": 13279
+    },
+    {
+      "epoch": 1.0717456218222903,
+      "grad_norm": 0.7533490061759949,
+      "learning_rate": 5.1217096500944017e-05,
+      "loss": 2.4778,
+      "step": 13280
+    },
+    {
+      "epoch": 1.0718263255588734,
+      "grad_norm": 0.6890795826911926,
+      "learning_rate": 5.1203316107130825e-05,
+      "loss": 2.4349,
+      "step": 13281
+    },
+    {
+      "epoch": 1.0719070292954564,
+      "grad_norm": 0.7004082202911377,
+      "learning_rate": 5.118953692945862e-05,
+      "loss": 2.4645,
+      "step": 13282
+    },
+    {
+      "epoch": 1.0719877330320393,
+      "grad_norm": 0.7409259676933289,
+      "learning_rate": 5.117575896827068e-05,
+      "loss": 2.4734,
+      "step": 13283
+    },
+    {
+      "epoch": 1.0720684367686224,
+      "grad_norm": 0.7035481929779053,
+      "learning_rate": 5.116198222391046e-05,
+      "loss": 2.5027,
+      "step": 13284
+    },
+    {
+      "epoch": 1.0721491405052055,
+      "grad_norm": 0.7146698236465454,
+      "learning_rate": 5.114820669672132e-05,
+      "loss": 2.4623,
+      "step": 13285
+    },
+    {
+      "epoch": 1.0722298442417884,
+      "grad_norm": 0.7813882231712341,
+      "learning_rate": 5.113443238704656e-05,
+      "loss": 2.4644,
+      "step": 13286
+    },
+    {
+      "epoch": 1.0723105479783714,
+      "grad_norm": 0.6592430472373962,
+      "learning_rate": 5.1120659295229486e-05,
+      "loss": 2.4682,
+      "step": 13287
+    },
+    {
+      "epoch": 1.0723912517149543,
+      "grad_norm": 0.7047967910766602,
+      "learning_rate": 5.1106887421613395e-05,
+      "loss": 2.4368,
+      "step": 13288
+    },
+    {
+      "epoch": 1.0724719554515374,
+      "grad_norm": 0.700977087020874,
+      "learning_rate": 5.109311676654143e-05,
+      "loss": 2.4471,
+      "step": 13289
+    },
+    {
+      "epoch": 1.0725526591881205,
+      "grad_norm": 0.6821093559265137,
+      "learning_rate": 5.107934733035684e-05,
+      "loss": 2.433,
+      "step": 13290
+    },
+    {
+      "epoch": 1.0726333629247033,
+      "grad_norm": 0.6579930186271667,
+      "learning_rate": 5.1065579113402794e-05,
+      "loss": 2.4527,
+      "step": 13291
+    },
+    {
+      "epoch": 1.0727140666612864,
+      "grad_norm": 0.658514678478241,
+      "learning_rate": 5.105181211602248e-05,
+      "loss": 2.4443,
+      "step": 13292
+    },
+    {
+      "epoch": 1.0727947703978695,
+      "grad_norm": 0.6963977217674255,
+      "learning_rate": 5.103804633855891e-05,
+      "loss": 2.4699,
+      "step": 13293
+    },
+    {
+      "epoch": 1.0728754741344524,
+      "grad_norm": 0.6670787334442139,
+      "learning_rate": 5.102428178135522e-05,
+      "loss": 2.4672,
+      "step": 13294
+    },
+    {
+      "epoch": 1.0729561778710355,
+      "grad_norm": 0.6959822773933411,
+      "learning_rate": 5.1010518444754454e-05,
+      "loss": 2.4338,
+      "step": 13295
+    },
+    {
+      "epoch": 1.0730368816076183,
+      "grad_norm": 0.6534817218780518,
+      "learning_rate": 5.0996756329099614e-05,
+      "loss": 2.4491,
+      "step": 13296
+    },
+    {
+      "epoch": 1.0731175853442014,
+      "grad_norm": 0.7265146970748901,
+      "learning_rate": 5.098299543473371e-05,
+      "loss": 2.4718,
+      "step": 13297
+    },
+    {
+      "epoch": 1.0731982890807845,
+      "grad_norm": 0.6554745435714722,
+      "learning_rate": 5.0969235761999746e-05,
+      "loss": 2.4286,
+      "step": 13298
+    },
+    {
+      "epoch": 1.0732789928173674,
+      "grad_norm": 0.7003172039985657,
+      "learning_rate": 5.095547731124053e-05,
+      "loss": 2.4182,
+      "step": 13299
+    },
+    {
+      "epoch": 1.0733596965539505,
+      "grad_norm": 0.6700341105461121,
+      "learning_rate": 5.094172008279904e-05,
+      "loss": 2.428,
+      "step": 13300
+    },
+    {
+      "epoch": 1.0734404002905333,
+      "grad_norm": 0.7290289402008057,
+      "learning_rate": 5.0927964077018164e-05,
+      "loss": 2.4324,
+      "step": 13301
+    },
+    {
+      "epoch": 1.0735211040271164,
+      "grad_norm": 0.6999204158782959,
+      "learning_rate": 5.0914209294240644e-05,
+      "loss": 2.5386,
+      "step": 13302
+    },
+    {
+      "epoch": 1.0736018077636995,
+      "grad_norm": 0.7008000612258911,
+      "learning_rate": 5.090045573480935e-05,
+      "loss": 2.5295,
+      "step": 13303
+    },
+    {
+      "epoch": 1.0736825115002824,
+      "grad_norm": 0.7023071646690369,
+      "learning_rate": 5.088670339906705e-05,
+      "loss": 2.4418,
+      "step": 13304
+    },
+    {
+      "epoch": 1.0737632152368655,
+      "grad_norm": 0.627174437046051,
+      "learning_rate": 5.0872952287356525e-05,
+      "loss": 2.3782,
+      "step": 13305
+    },
+    {
+      "epoch": 1.0738439189734486,
+      "grad_norm": 0.6992766857147217,
+      "learning_rate": 5.0859202400020364e-05,
+      "loss": 2.4698,
+      "step": 13306
+    },
+    {
+      "epoch": 1.0739246227100314,
+      "grad_norm": 0.7189817428588867,
+      "learning_rate": 5.084545373740138e-05,
+      "loss": 2.5248,
+      "step": 13307
+    },
+    {
+      "epoch": 1.0740053264466145,
+      "grad_norm": 0.6849164962768555,
+      "learning_rate": 5.0831706299842216e-05,
+      "loss": 2.4084,
+      "step": 13308
+    },
+    {
+      "epoch": 1.0740860301831976,
+      "grad_norm": 0.6985825300216675,
+      "learning_rate": 5.0817960087685424e-05,
+      "loss": 2.4893,
+      "step": 13309
+    },
+    {
+      "epoch": 1.0741667339197805,
+      "grad_norm": 0.6519783139228821,
+      "learning_rate": 5.080421510127362e-05,
+      "loss": 2.5144,
+      "step": 13310
+    },
+    {
+      "epoch": 1.0742474376563635,
+      "grad_norm": 0.6605731248855591,
+      "learning_rate": 5.079047134094941e-05,
+      "loss": 2.4487,
+      "step": 13311
+    },
+    {
+      "epoch": 1.0743281413929464,
+      "grad_norm": 0.7236705422401428,
+      "learning_rate": 5.077672880705526e-05,
+      "loss": 2.4578,
+      "step": 13312
+    },
+    {
+      "epoch": 1.0744088451295295,
+      "grad_norm": 0.7126381397247314,
+      "learning_rate": 5.07629874999337e-05,
+      "loss": 2.4528,
+      "step": 13313
+    },
+    {
+      "epoch": 1.0744895488661126,
+      "grad_norm": 0.7247878313064575,
+      "learning_rate": 5.0749247419927236e-05,
+      "loss": 2.563,
+      "step": 13314
+    },
+    {
+      "epoch": 1.0745702526026955,
+      "grad_norm": 0.728349506855011,
+      "learning_rate": 5.0735508567378234e-05,
+      "loss": 2.4229,
+      "step": 13315
+    },
+    {
+      "epoch": 1.0746509563392785,
+      "grad_norm": 0.6593719124794006,
+      "learning_rate": 5.072177094262913e-05,
+      "loss": 2.4853,
+      "step": 13316
+    },
+    {
+      "epoch": 1.0747316600758614,
+      "grad_norm": 0.6519735455513,
+      "learning_rate": 5.070803454602231e-05,
+      "loss": 2.4507,
+      "step": 13317
+    },
+    {
+      "epoch": 1.0748123638124445,
+      "grad_norm": 0.6660017371177673,
+      "learning_rate": 5.0694299377900115e-05,
+      "loss": 2.4286,
+      "step": 13318
+    },
+    {
+      "epoch": 1.0748930675490276,
+      "grad_norm": 0.7506695985794067,
+      "learning_rate": 5.0680565438604876e-05,
+      "loss": 2.4841,
+      "step": 13319
+    },
+    {
+      "epoch": 1.0749737712856104,
+      "grad_norm": 0.6855955719947815,
+      "learning_rate": 5.0666832728478863e-05,
+      "loss": 2.3817,
+      "step": 13320
+    },
+    {
+      "epoch": 1.0750544750221935,
+      "grad_norm": 0.7151634693145752,
+      "learning_rate": 5.065310124786438e-05,
+      "loss": 2.3984,
+      "step": 13321
+    },
+    {
+      "epoch": 1.0751351787587766,
+      "grad_norm": 0.6551649570465088,
+      "learning_rate": 5.063937099710356e-05,
+      "loss": 2.4574,
+      "step": 13322
+    },
+    {
+      "epoch": 1.0752158824953595,
+      "grad_norm": 0.7443479895591736,
+      "learning_rate": 5.062564197653865e-05,
+      "loss": 2.52,
+      "step": 13323
+    },
+    {
+      "epoch": 1.0752965862319426,
+      "grad_norm": 0.7554972767829895,
+      "learning_rate": 5.061191418651186e-05,
+      "loss": 2.483,
+      "step": 13324
+    },
+    {
+      "epoch": 1.0753772899685254,
+      "grad_norm": 0.7661007642745972,
+      "learning_rate": 5.059818762736521e-05,
+      "loss": 2.566,
+      "step": 13325
+    },
+    {
+      "epoch": 1.0754579937051085,
+      "grad_norm": 0.7416480183601379,
+      "learning_rate": 5.058446229944087e-05,
+      "loss": 2.465,
+      "step": 13326
+    },
+    {
+      "epoch": 1.0755386974416916,
+      "grad_norm": 0.6997848749160767,
+      "learning_rate": 5.057073820308089e-05,
+      "loss": 2.4936,
+      "step": 13327
+    },
+    {
+      "epoch": 1.0756194011782745,
+      "grad_norm": 0.7570235133171082,
+      "learning_rate": 5.0557015338627345e-05,
+      "loss": 2.519,
+      "step": 13328
+    },
+    {
+      "epoch": 1.0757001049148576,
+      "grad_norm": 0.7910803556442261,
+      "learning_rate": 5.0543293706422214e-05,
+      "loss": 2.4932,
+      "step": 13329
+    },
+    {
+      "epoch": 1.0757808086514407,
+      "grad_norm": 0.7068312168121338,
+      "learning_rate": 5.052957330680752e-05,
+      "loss": 2.4489,
+      "step": 13330
+    },
+    {
+      "epoch": 1.0758615123880235,
+      "grad_norm": 0.7818215489387512,
+      "learning_rate": 5.051585414012514e-05,
+      "loss": 2.4467,
+      "step": 13331
+    },
+    {
+      "epoch": 1.0759422161246066,
+      "grad_norm": 0.7359446287155151,
+      "learning_rate": 5.0502136206717046e-05,
+      "loss": 2.4348,
+      "step": 13332
+    },
+    {
+      "epoch": 1.0760229198611895,
+      "grad_norm": 0.694726824760437,
+      "learning_rate": 5.0488419506925124e-05,
+      "loss": 2.4554,
+      "step": 13333
+    },
+    {
+      "epoch": 1.0761036235977726,
+      "grad_norm": 0.6776530742645264,
+      "learning_rate": 5.047470404109118e-05,
+      "loss": 2.4206,
+      "step": 13334
+    },
+    {
+      "epoch": 1.0761843273343557,
+      "grad_norm": 0.6977556943893433,
+      "learning_rate": 5.0460989809557066e-05,
+      "loss": 2.4748,
+      "step": 13335
+    },
+    {
+      "epoch": 1.0762650310709385,
+      "grad_norm": 0.6888061761856079,
+      "learning_rate": 5.044727681266459e-05,
+      "loss": 2.4129,
+      "step": 13336
+    },
+    {
+      "epoch": 1.0763457348075216,
+      "grad_norm": 0.744110643863678,
+      "learning_rate": 5.043356505075549e-05,
+      "loss": 2.4815,
+      "step": 13337
+    },
+    {
+      "epoch": 1.0764264385441047,
+      "grad_norm": 0.6726455688476562,
+      "learning_rate": 5.041985452417154e-05,
+      "loss": 2.4299,
+      "step": 13338
+    },
+    {
+      "epoch": 1.0765071422806876,
+      "grad_norm": 0.6755545735359192,
+      "learning_rate": 5.040614523325441e-05,
+      "loss": 2.4188,
+      "step": 13339
+    },
+    {
+      "epoch": 1.0765878460172706,
+      "grad_norm": 0.7152739763259888,
+      "learning_rate": 5.039243717834582e-05,
+      "loss": 2.4366,
+      "step": 13340
+    },
+    {
+      "epoch": 1.0766685497538535,
+      "grad_norm": 0.7253085374832153,
+      "learning_rate": 5.037873035978733e-05,
+      "loss": 2.4681,
+      "step": 13341
+    },
+    {
+      "epoch": 1.0767492534904366,
+      "grad_norm": 0.6780266165733337,
+      "learning_rate": 5.03650247779206e-05,
+      "loss": 2.5163,
+      "step": 13342
+    },
+    {
+      "epoch": 1.0768299572270197,
+      "grad_norm": 0.7440996170043945,
+      "learning_rate": 5.035132043308722e-05,
+      "loss": 2.4831,
+      "step": 13343
+    },
+    {
+      "epoch": 1.0769106609636026,
+      "grad_norm": 0.6619833111763,
+      "learning_rate": 5.0337617325628695e-05,
+      "loss": 2.433,
+      "step": 13344
+    },
+    {
+      "epoch": 1.0769913647001856,
+      "grad_norm": 0.7518059015274048,
+      "learning_rate": 5.032391545588656e-05,
+      "loss": 2.4241,
+      "step": 13345
+    },
+    {
+      "epoch": 1.0770720684367687,
+      "grad_norm": 0.6592784523963928,
+      "learning_rate": 5.031021482420231e-05,
+      "loss": 2.4902,
+      "step": 13346
+    },
+    {
+      "epoch": 1.0771527721733516,
+      "grad_norm": 0.7192299365997314,
+      "learning_rate": 5.029651543091739e-05,
+      "loss": 2.4445,
+      "step": 13347
+    },
+    {
+      "epoch": 1.0772334759099347,
+      "grad_norm": 0.7376793622970581,
+      "learning_rate": 5.028281727637323e-05,
+      "loss": 2.4532,
+      "step": 13348
+    },
+    {
+      "epoch": 1.0773141796465175,
+      "grad_norm": 0.7344524264335632,
+      "learning_rate": 5.026912036091127e-05,
+      "loss": 2.4193,
+      "step": 13349
+    },
+    {
+      "epoch": 1.0773948833831006,
+      "grad_norm": 0.7343986630439758,
+      "learning_rate": 5.0255424684872785e-05,
+      "loss": 2.4912,
+      "step": 13350
+    },
+    {
+      "epoch": 1.0774755871196837,
+      "grad_norm": 0.7103631496429443,
+      "learning_rate": 5.024173024859916e-05,
+      "loss": 2.4611,
+      "step": 13351
+    },
+    {
+      "epoch": 1.0775562908562666,
+      "grad_norm": 0.7554094791412354,
+      "learning_rate": 5.022803705243169e-05,
+      "loss": 2.4875,
+      "step": 13352
+    },
+    {
+      "epoch": 1.0776369945928497,
+      "grad_norm": 0.6754978895187378,
+      "learning_rate": 5.0214345096711655e-05,
+      "loss": 2.4585,
+      "step": 13353
+    },
+    {
+      "epoch": 1.0777176983294328,
+      "grad_norm": 0.690747857093811,
+      "learning_rate": 5.020065438178026e-05,
+      "loss": 2.4751,
+      "step": 13354
+    },
+    {
+      "epoch": 1.0777984020660156,
+      "grad_norm": 0.7012028694152832,
+      "learning_rate": 5.018696490797874e-05,
+      "loss": 2.4443,
+      "step": 13355
+    },
+    {
+      "epoch": 1.0778791058025987,
+      "grad_norm": 0.6788459420204163,
+      "learning_rate": 5.017327667564831e-05,
+      "loss": 2.4135,
+      "step": 13356
+    },
+    {
+      "epoch": 1.0779598095391816,
+      "grad_norm": 0.6662794351577759,
+      "learning_rate": 5.015958968512997e-05,
+      "loss": 2.3801,
+      "step": 13357
+    },
+    {
+      "epoch": 1.0780405132757647,
+      "grad_norm": 0.7873939275741577,
+      "learning_rate": 5.0145903936764994e-05,
+      "loss": 2.4629,
+      "step": 13358
+    },
+    {
+      "epoch": 1.0781212170123478,
+      "grad_norm": 0.7484980225563049,
+      "learning_rate": 5.0132219430894455e-05,
+      "loss": 2.4307,
+      "step": 13359
+    },
+    {
+      "epoch": 1.0782019207489306,
+      "grad_norm": 0.7559076547622681,
+      "learning_rate": 5.011853616785932e-05,
+      "loss": 2.4846,
+      "step": 13360
+    },
+    {
+      "epoch": 1.0782826244855137,
+      "grad_norm": 0.6822710633277893,
+      "learning_rate": 5.010485414800066e-05,
+      "loss": 2.4448,
+      "step": 13361
+    },
+    {
+      "epoch": 1.0783633282220966,
+      "grad_norm": 0.6665955185890198,
+      "learning_rate": 5.0091173371659496e-05,
+      "loss": 2.4562,
+      "step": 13362
+    },
+    {
+      "epoch": 1.0784440319586797,
+      "grad_norm": 0.6645659804344177,
+      "learning_rate": 5.0077493839176714e-05,
+      "loss": 2.4545,
+      "step": 13363
+    },
+    {
+      "epoch": 1.0785247356952627,
+      "grad_norm": 0.6648181080818176,
+      "learning_rate": 5.0063815550893276e-05,
+      "loss": 2.4565,
+      "step": 13364
+    },
+    {
+      "epoch": 1.0786054394318456,
+      "grad_norm": 0.6679299473762512,
+      "learning_rate": 5.005013850715014e-05,
+      "loss": 2.4301,
+      "step": 13365
+    },
+    {
+      "epoch": 1.0786861431684287,
+      "grad_norm": 0.7116484642028809,
+      "learning_rate": 5.003646270828808e-05,
+      "loss": 2.4174,
+      "step": 13366
+    },
+    {
+      "epoch": 1.0787668469050118,
+      "grad_norm": 0.6850735545158386,
+      "learning_rate": 5.002278815464798e-05,
+      "loss": 2.4386,
+      "step": 13367
+    },
+    {
+      "epoch": 1.0788475506415947,
+      "grad_norm": 0.6613513827323914,
+      "learning_rate": 5.00091148465706e-05,
+      "loss": 2.4038,
+      "step": 13368
+    },
+    {
+      "epoch": 1.0789282543781777,
+      "grad_norm": 0.659635603427887,
+      "learning_rate": 4.9995442784396827e-05,
+      "loss": 2.4346,
+      "step": 13369
+    },
+    {
+      "epoch": 1.0790089581147608,
+      "grad_norm": 0.6775132417678833,
+      "learning_rate": 4.998177196846731e-05,
+      "loss": 2.4853,
+      "step": 13370
+    },
+    {
+      "epoch": 1.0790896618513437,
+      "grad_norm": 0.719860851764679,
+      "learning_rate": 4.996810239912277e-05,
+      "loss": 2.4018,
+      "step": 13371
+    },
+    {
+      "epoch": 1.0791703655879268,
+      "grad_norm": 0.7316389083862305,
+      "learning_rate": 4.9954434076703946e-05,
+      "loss": 2.424,
+      "step": 13372
+    },
+    {
+      "epoch": 1.0792510693245096,
+      "grad_norm": 0.6779622435569763,
+      "learning_rate": 4.99407670015514e-05,
+      "loss": 2.4743,
+      "step": 13373
+    },
+    {
+      "epoch": 1.0793317730610927,
+      "grad_norm": 0.7357139587402344,
+      "learning_rate": 4.992710117400581e-05,
+      "loss": 2.4385,
+      "step": 13374
+    },
+    {
+      "epoch": 1.0794124767976758,
+      "grad_norm": 0.671441912651062,
+      "learning_rate": 4.9913436594407784e-05,
+      "loss": 2.3988,
+      "step": 13375
+    },
+    {
+      "epoch": 1.0794931805342587,
+      "grad_norm": 0.7205149531364441,
+      "learning_rate": 4.9899773263097804e-05,
+      "loss": 2.4594,
+      "step": 13376
+    },
+    {
+      "epoch": 1.0795738842708418,
+      "grad_norm": 0.702910840511322,
+      "learning_rate": 4.988611118041644e-05,
+      "loss": 2.4831,
+      "step": 13377
+    },
+    {
+      "epoch": 1.0796545880074246,
+      "grad_norm": 0.6977962255477905,
+      "learning_rate": 4.987245034670418e-05,
+      "loss": 2.422,
+      "step": 13378
+    },
+    {
+      "epoch": 1.0797352917440077,
+      "grad_norm": 0.7106757760047913,
+      "learning_rate": 4.985879076230149e-05,
+      "loss": 2.4073,
+      "step": 13379
+    },
+    {
+      "epoch": 1.0798159954805908,
+      "grad_norm": 0.7046806812286377,
+      "learning_rate": 4.9845132427548814e-05,
+      "loss": 2.4065,
+      "step": 13380
+    },
+    {
+      "epoch": 1.0798966992171737,
+      "grad_norm": 0.7476605772972107,
+      "learning_rate": 4.9831475342786574e-05,
+      "loss": 2.4886,
+      "step": 13381
+    },
+    {
+      "epoch": 1.0799774029537568,
+      "grad_norm": 0.696977972984314,
+      "learning_rate": 4.981781950835508e-05,
+      "loss": 2.4732,
+      "step": 13382
+    },
+    {
+      "epoch": 1.0800581066903399,
+      "grad_norm": 0.6596804857254028,
+      "learning_rate": 4.98041649245947e-05,
+      "loss": 2.4497,
+      "step": 13383
+    },
+    {
+      "epoch": 1.0801388104269227,
+      "grad_norm": 0.7216050028800964,
+      "learning_rate": 4.979051159184573e-05,
+      "loss": 2.4745,
+      "step": 13384
+    },
+    {
+      "epoch": 1.0802195141635058,
+      "grad_norm": 0.6636630296707153,
+      "learning_rate": 4.977685951044852e-05,
+      "loss": 2.4904,
+      "step": 13385
+    },
+    {
+      "epoch": 1.0803002179000887,
+      "grad_norm": 0.7030208110809326,
+      "learning_rate": 4.97632086807432e-05,
+      "loss": 2.4302,
+      "step": 13386
+    },
+    {
+      "epoch": 1.0803809216366718,
+      "grad_norm": 0.7158327102661133,
+      "learning_rate": 4.974955910307004e-05,
+      "loss": 2.4735,
+      "step": 13387
+    },
+    {
+      "epoch": 1.0804616253732549,
+      "grad_norm": 0.6736464500427246,
+      "learning_rate": 4.9735910777769234e-05,
+      "loss": 2.4334,
+      "step": 13388
+    },
+    {
+      "epoch": 1.0805423291098377,
+      "grad_norm": 0.6913403272628784,
+      "learning_rate": 4.972226370518092e-05,
+      "loss": 2.468,
+      "step": 13389
+    },
+    {
+      "epoch": 1.0806230328464208,
+      "grad_norm": 0.7006524205207825,
+      "learning_rate": 4.970861788564522e-05,
+      "loss": 2.4598,
+      "step": 13390
+    },
+    {
+      "epoch": 1.080703736583004,
+      "grad_norm": 0.6892947554588318,
+      "learning_rate": 4.969497331950227e-05,
+      "loss": 2.4297,
+      "step": 13391
+    },
+    {
+      "epoch": 1.0807844403195868,
+      "grad_norm": 0.7270283699035645,
+      "learning_rate": 4.968133000709203e-05,
+      "loss": 2.5344,
+      "step": 13392
+    },
+    {
+      "epoch": 1.0808651440561698,
+      "grad_norm": 0.735342264175415,
+      "learning_rate": 4.9667687948754594e-05,
+      "loss": 2.4431,
+      "step": 13393
+    },
+    {
+      "epoch": 1.0809458477927527,
+      "grad_norm": 0.6869279146194458,
+      "learning_rate": 4.9654047144829974e-05,
+      "loss": 2.5581,
+      "step": 13394
+    },
+    {
+      "epoch": 1.0810265515293358,
+      "grad_norm": 0.6975715160369873,
+      "learning_rate": 4.964040759565808e-05,
+      "loss": 2.4328,
+      "step": 13395
+    },
+    {
+      "epoch": 1.0811072552659189,
+      "grad_norm": 0.7312532067298889,
+      "learning_rate": 4.9626769301578856e-05,
+      "loss": 2.4686,
+      "step": 13396
+    },
+    {
+      "epoch": 1.0811879590025018,
+      "grad_norm": 0.7824496626853943,
+      "learning_rate": 4.9613132262932215e-05,
+      "loss": 2.4564,
+      "step": 13397
+    },
+    {
+      "epoch": 1.0812686627390848,
+      "grad_norm": 0.7337941527366638,
+      "learning_rate": 4.959949648005805e-05,
+      "loss": 2.4752,
+      "step": 13398
+    },
+    {
+      "epoch": 1.081349366475668,
+      "grad_norm": 0.7450836300849915,
+      "learning_rate": 4.958586195329617e-05,
+      "loss": 2.4457,
+      "step": 13399
+    },
+    {
+      "epoch": 1.0814300702122508,
+      "grad_norm": 0.6990504860877991,
+      "learning_rate": 4.9572228682986385e-05,
+      "loss": 2.4172,
+      "step": 13400
+    },
+    {
+      "epoch": 1.0815107739488339,
+      "grad_norm": 0.7293999791145325,
+      "learning_rate": 4.955859666946853e-05,
+      "loss": 2.5295,
+      "step": 13401
+    },
+    {
+      "epoch": 1.0815914776854167,
+      "grad_norm": 0.6872537136077881,
+      "learning_rate": 4.9544965913082264e-05,
+      "loss": 2.5029,
+      "step": 13402
+    },
+    {
+      "epoch": 1.0816721814219998,
+      "grad_norm": 0.6821706891059875,
+      "learning_rate": 4.953133641416733e-05,
+      "loss": 2.4738,
+      "step": 13403
+    },
+    {
+      "epoch": 1.081752885158583,
+      "grad_norm": 0.6811527609825134,
+      "learning_rate": 4.951770817306346e-05,
+      "loss": 2.4323,
+      "step": 13404
+    },
+    {
+      "epoch": 1.0818335888951658,
+      "grad_norm": 0.7138943076133728,
+      "learning_rate": 4.950408119011023e-05,
+      "loss": 2.5155,
+      "step": 13405
+    },
+    {
+      "epoch": 1.0819142926317489,
+      "grad_norm": 0.6777952909469604,
+      "learning_rate": 4.949045546564729e-05,
+      "loss": 2.4414,
+      "step": 13406
+    },
+    {
+      "epoch": 1.0819949963683317,
+      "grad_norm": 0.7065548896789551,
+      "learning_rate": 4.9476831000014276e-05,
+      "loss": 2.4913,
+      "step": 13407
+    },
+    {
+      "epoch": 1.0820757001049148,
+      "grad_norm": 0.7286355495452881,
+      "learning_rate": 4.9463207793550626e-05,
+      "loss": 2.4171,
+      "step": 13408
+    },
+    {
+      "epoch": 1.082156403841498,
+      "grad_norm": 0.6703049540519714,
+      "learning_rate": 4.944958584659597e-05,
+      "loss": 2.4387,
+      "step": 13409
+    },
+    {
+      "epoch": 1.0822371075780808,
+      "grad_norm": 0.6572019457817078,
+      "learning_rate": 4.943596515948983e-05,
+      "loss": 2.4324,
+      "step": 13410
+    },
+    {
+      "epoch": 1.0823178113146639,
+      "grad_norm": 0.6722360849380493,
+      "learning_rate": 4.942234573257156e-05,
+      "loss": 2.4802,
+      "step": 13411
+    },
+    {
+      "epoch": 1.082398515051247,
+      "grad_norm": 0.7122535109519958,
+      "learning_rate": 4.9408727566180655e-05,
+      "loss": 2.4531,
+      "step": 13412
+    },
+    {
+      "epoch": 1.0824792187878298,
+      "grad_norm": 0.6769903898239136,
+      "learning_rate": 4.9395110660656505e-05,
+      "loss": 2.4549,
+      "step": 13413
+    },
+    {
+      "epoch": 1.082559922524413,
+      "grad_norm": 0.766251266002655,
+      "learning_rate": 4.938149501633852e-05,
+      "loss": 2.4416,
+      "step": 13414
+    },
+    {
+      "epoch": 1.082640626260996,
+      "grad_norm": 0.6677987575531006,
+      "learning_rate": 4.936788063356596e-05,
+      "loss": 2.4578,
+      "step": 13415
+    },
+    {
+      "epoch": 1.0827213299975789,
+      "grad_norm": 0.7461380362510681,
+      "learning_rate": 4.9354267512678156e-05,
+      "loss": 2.4776,
+      "step": 13416
+    },
+    {
+      "epoch": 1.082802033734162,
+      "grad_norm": 0.6681976914405823,
+      "learning_rate": 4.934065565401443e-05,
+      "loss": 2.5044,
+      "step": 13417
+    },
+    {
+      "epoch": 1.0828827374707448,
+      "grad_norm": 0.6809324622154236,
+      "learning_rate": 4.932704505791397e-05,
+      "loss": 2.4651,
+      "step": 13418
+    },
+    {
+      "epoch": 1.082963441207328,
+      "grad_norm": 0.6926563382148743,
+      "learning_rate": 4.931343572471596e-05,
+      "loss": 2.4633,
+      "step": 13419
+    },
+    {
+      "epoch": 1.083044144943911,
+      "grad_norm": 0.6451820135116577,
+      "learning_rate": 4.929982765475971e-05,
+      "loss": 2.474,
+      "step": 13420
+    },
+    {
+      "epoch": 1.0831248486804939,
+      "grad_norm": 0.7088493704795837,
+      "learning_rate": 4.9286220848384247e-05,
+      "loss": 2.462,
+      "step": 13421
+    },
+    {
+      "epoch": 1.083205552417077,
+      "grad_norm": 0.7819172739982605,
+      "learning_rate": 4.9272615305928725e-05,
+      "loss": 2.4534,
+      "step": 13422
+    },
+    {
+      "epoch": 1.0832862561536598,
+      "grad_norm": 0.6579666137695312,
+      "learning_rate": 4.925901102773227e-05,
+      "loss": 2.4101,
+      "step": 13423
+    },
+    {
+      "epoch": 1.083366959890243,
+      "grad_norm": 0.6999555230140686,
+      "learning_rate": 4.924540801413385e-05,
+      "loss": 2.4534,
+      "step": 13424
+    },
+    {
+      "epoch": 1.083447663626826,
+      "grad_norm": 0.7034400105476379,
+      "learning_rate": 4.9231806265472555e-05,
+      "loss": 2.4741,
+      "step": 13425
+    },
+    {
+      "epoch": 1.0835283673634089,
+      "grad_norm": 0.6595034599304199,
+      "learning_rate": 4.921820578208739e-05,
+      "loss": 2.4011,
+      "step": 13426
+    },
+    {
+      "epoch": 1.083609071099992,
+      "grad_norm": 0.666419267654419,
+      "learning_rate": 4.920460656431723e-05,
+      "loss": 2.4399,
+      "step": 13427
+    },
+    {
+      "epoch": 1.083689774836575,
+      "grad_norm": 0.7058294415473938,
+      "learning_rate": 4.919100861250108e-05,
+      "loss": 2.434,
+      "step": 13428
+    },
+    {
+      "epoch": 1.083770478573158,
+      "grad_norm": 0.7045806050300598,
+      "learning_rate": 4.917741192697779e-05,
+      "loss": 2.4616,
+      "step": 13429
+    },
+    {
+      "epoch": 1.083851182309741,
+      "grad_norm": 0.6565639972686768,
+      "learning_rate": 4.916381650808626e-05,
+      "loss": 2.3864,
+      "step": 13430
+    },
+    {
+      "epoch": 1.0839318860463238,
+      "grad_norm": 0.6939674615859985,
+      "learning_rate": 4.9150222356165295e-05,
+      "loss": 2.4217,
+      "step": 13431
+    },
+    {
+      "epoch": 1.084012589782907,
+      "grad_norm": 0.7240599989891052,
+      "learning_rate": 4.913662947155373e-05,
+      "loss": 2.447,
+      "step": 13432
+    },
+    {
+      "epoch": 1.08409329351949,
+      "grad_norm": 0.7369012832641602,
+      "learning_rate": 4.9123037854590336e-05,
+      "loss": 2.4588,
+      "step": 13433
+    },
+    {
+      "epoch": 1.0841739972560729,
+      "grad_norm": 0.714269757270813,
+      "learning_rate": 4.9109447505613803e-05,
+      "loss": 2.4921,
+      "step": 13434
+    },
+    {
+      "epoch": 1.084254700992656,
+      "grad_norm": 0.7541659474372864,
+      "learning_rate": 4.909585842496287e-05,
+      "loss": 2.4191,
+      "step": 13435
+    },
+    {
+      "epoch": 1.084335404729239,
+      "grad_norm": 0.7245596051216125,
+      "learning_rate": 4.9082270612976243e-05,
+      "loss": 2.4904,
+      "step": 13436
+    },
+    {
+      "epoch": 1.084416108465822,
+      "grad_norm": 0.7301090359687805,
+      "learning_rate": 4.90686840699925e-05,
+      "loss": 2.4461,
+      "step": 13437
+    },
+    {
+      "epoch": 1.084496812202405,
+      "grad_norm": 0.7404102683067322,
+      "learning_rate": 4.905509879635028e-05,
+      "loss": 2.4826,
+      "step": 13438
+    },
+    {
+      "epoch": 1.0845775159389879,
+      "grad_norm": 0.7053710222244263,
+      "learning_rate": 4.9041514792388175e-05,
+      "loss": 2.4231,
+      "step": 13439
+    },
+    {
+      "epoch": 1.084658219675571,
+      "grad_norm": 0.6171362400054932,
+      "learning_rate": 4.9027932058444724e-05,
+      "loss": 2.4472,
+      "step": 13440
+    },
+    {
+      "epoch": 1.084738923412154,
+      "grad_norm": 0.7367038130760193,
+      "learning_rate": 4.901435059485845e-05,
+      "loss": 2.4847,
+      "step": 13441
+    },
+    {
+      "epoch": 1.084819627148737,
+      "grad_norm": 0.754828691482544,
+      "learning_rate": 4.900077040196788e-05,
+      "loss": 2.4731,
+      "step": 13442
+    },
+    {
+      "epoch": 1.08490033088532,
+      "grad_norm": 0.7380684018135071,
+      "learning_rate": 4.8987191480111386e-05,
+      "loss": 2.4227,
+      "step": 13443
+    },
+    {
+      "epoch": 1.084981034621903,
+      "grad_norm": 0.6711444854736328,
+      "learning_rate": 4.897361382962742e-05,
+      "loss": 2.4744,
+      "step": 13444
+    },
+    {
+      "epoch": 1.085061738358486,
+      "grad_norm": 0.7709227204322815,
+      "learning_rate": 4.896003745085438e-05,
+      "loss": 2.5422,
+      "step": 13445
+    },
+    {
+      "epoch": 1.085142442095069,
+      "grad_norm": 0.6778519153594971,
+      "learning_rate": 4.8946462344130675e-05,
+      "loss": 2.4757,
+      "step": 13446
+    },
+    {
+      "epoch": 1.085223145831652,
+      "grad_norm": 0.7390698194503784,
+      "learning_rate": 4.893288850979454e-05,
+      "loss": 2.4214,
+      "step": 13447
+    },
+    {
+      "epoch": 1.085303849568235,
+      "grad_norm": 0.6632684469223022,
+      "learning_rate": 4.891931594818432e-05,
+      "loss": 2.4689,
+      "step": 13448
+    },
+    {
+      "epoch": 1.085384553304818,
+      "grad_norm": 0.68693608045578,
+      "learning_rate": 4.890574465963827e-05,
+      "loss": 2.4788,
+      "step": 13449
+    },
+    {
+      "epoch": 1.085465257041401,
+      "grad_norm": 0.6910344362258911,
+      "learning_rate": 4.8892174644494625e-05,
+      "loss": 2.4611,
+      "step": 13450
+    },
+    {
+      "epoch": 1.085545960777984,
+      "grad_norm": 0.6935380101203918,
+      "learning_rate": 4.887860590309158e-05,
+      "loss": 2.4481,
+      "step": 13451
+    },
+    {
+      "epoch": 1.085626664514567,
+      "grad_norm": 0.7086954712867737,
+      "learning_rate": 4.886503843576735e-05,
+      "loss": 2.4583,
+      "step": 13452
+    },
+    {
+      "epoch": 1.08570736825115,
+      "grad_norm": 0.7447777986526489,
+      "learning_rate": 4.8851472242859994e-05,
+      "loss": 2.5035,
+      "step": 13453
+    },
+    {
+      "epoch": 1.085788071987733,
+      "grad_norm": 0.6896036267280579,
+      "learning_rate": 4.8837907324707656e-05,
+      "loss": 2.4622,
+      "step": 13454
+    },
+    {
+      "epoch": 1.085868775724316,
+      "grad_norm": 0.7261155247688293,
+      "learning_rate": 4.882434368164843e-05,
+      "loss": 2.4958,
+      "step": 13455
+    },
+    {
+      "epoch": 1.085949479460899,
+      "grad_norm": 0.6868197321891785,
+      "learning_rate": 4.881078131402031e-05,
+      "loss": 2.4952,
+      "step": 13456
+    },
+    {
+      "epoch": 1.0860301831974821,
+      "grad_norm": 0.6338867545127869,
+      "learning_rate": 4.879722022216132e-05,
+      "loss": 2.4553,
+      "step": 13457
+    },
+    {
+      "epoch": 1.086110886934065,
+      "grad_norm": 0.7214454412460327,
+      "learning_rate": 4.878366040640946e-05,
+      "loss": 2.4433,
+      "step": 13458
+    },
+    {
+      "epoch": 1.086191590670648,
+      "grad_norm": 0.6871301531791687,
+      "learning_rate": 4.877010186710266e-05,
+      "loss": 2.4118,
+      "step": 13459
+    },
+    {
+      "epoch": 1.0862722944072312,
+      "grad_norm": 0.6845650672912598,
+      "learning_rate": 4.875654460457883e-05,
+      "loss": 2.4684,
+      "step": 13460
+    },
+    {
+      "epoch": 1.086352998143814,
+      "grad_norm": 0.7027513980865479,
+      "learning_rate": 4.8742988619175865e-05,
+      "loss": 2.4569,
+      "step": 13461
+    },
+    {
+      "epoch": 1.0864337018803971,
+      "grad_norm": 0.6428621411323547,
+      "learning_rate": 4.8729433911231646e-05,
+      "loss": 2.4211,
+      "step": 13462
+    },
+    {
+      "epoch": 1.08651440561698,
+      "grad_norm": 0.6921488046646118,
+      "learning_rate": 4.8715880481083934e-05,
+      "loss": 2.4668,
+      "step": 13463
+    },
+    {
+      "epoch": 1.086595109353563,
+      "grad_norm": 0.7001025676727295,
+      "learning_rate": 4.870232832907051e-05,
+      "loss": 2.4685,
+      "step": 13464
+    },
+    {
+      "epoch": 1.0866758130901462,
+      "grad_norm": 0.7460644245147705,
+      "learning_rate": 4.868877745552922e-05,
+      "loss": 2.3922,
+      "step": 13465
+    },
+    {
+      "epoch": 1.086756516826729,
+      "grad_norm": 0.7418891191482544,
+      "learning_rate": 4.867522786079768e-05,
+      "loss": 2.3777,
+      "step": 13466
+    },
+    {
+      "epoch": 1.0868372205633121,
+      "grad_norm": 0.6430083513259888,
+      "learning_rate": 4.8661679545213625e-05,
+      "loss": 2.4385,
+      "step": 13467
+    },
+    {
+      "epoch": 1.086917924299895,
+      "grad_norm": 0.6963593363761902,
+      "learning_rate": 4.864813250911475e-05,
+      "loss": 2.4083,
+      "step": 13468
+    },
+    {
+      "epoch": 1.086998628036478,
+      "grad_norm": 0.6796097159385681,
+      "learning_rate": 4.8634586752838606e-05,
+      "loss": 2.4984,
+      "step": 13469
+    },
+    {
+      "epoch": 1.0870793317730612,
+      "grad_norm": 0.6845307946205139,
+      "learning_rate": 4.862104227672281e-05,
+      "loss": 2.4168,
+      "step": 13470
+    },
+    {
+      "epoch": 1.087160035509644,
+      "grad_norm": 0.705348014831543,
+      "learning_rate": 4.8607499081105e-05,
+      "loss": 2.4216,
+      "step": 13471
+    },
+    {
+      "epoch": 1.087240739246227,
+      "grad_norm": 0.6906474828720093,
+      "learning_rate": 4.8593957166322636e-05,
+      "loss": 2.4955,
+      "step": 13472
+    },
+    {
+      "epoch": 1.0873214429828102,
+      "grad_norm": 0.696489691734314,
+      "learning_rate": 4.858041653271323e-05,
+      "loss": 2.4186,
+      "step": 13473
+    },
+    {
+      "epoch": 1.087402146719393,
+      "grad_norm": 0.6997761726379395,
+      "learning_rate": 4.856687718061429e-05,
+      "loss": 2.441,
+      "step": 13474
+    },
+    {
+      "epoch": 1.0874828504559761,
+      "grad_norm": 0.6515649557113647,
+      "learning_rate": 4.8553339110363184e-05,
+      "loss": 2.3997,
+      "step": 13475
+    },
+    {
+      "epoch": 1.087563554192559,
+      "grad_norm": 0.6902725696563721,
+      "learning_rate": 4.853980232229734e-05,
+      "loss": 2.4765,
+      "step": 13476
+    },
+    {
+      "epoch": 1.087644257929142,
+      "grad_norm": 0.6832055449485779,
+      "learning_rate": 4.852626681675415e-05,
+      "loss": 2.411,
+      "step": 13477
+    },
+    {
+      "epoch": 1.0877249616657252,
+      "grad_norm": 0.668520987033844,
+      "learning_rate": 4.8512732594070984e-05,
+      "loss": 2.4742,
+      "step": 13478
+    },
+    {
+      "epoch": 1.087805665402308,
+      "grad_norm": 0.7019832134246826,
+      "learning_rate": 4.849919965458507e-05,
+      "loss": 2.4638,
+      "step": 13479
+    },
+    {
+      "epoch": 1.0878863691388911,
+      "grad_norm": 0.6986027359962463,
+      "learning_rate": 4.8485667998633724e-05,
+      "loss": 2.4866,
+      "step": 13480
+    },
+    {
+      "epoch": 1.0879670728754742,
+      "grad_norm": 0.659037709236145,
+      "learning_rate": 4.8472137626554195e-05,
+      "loss": 2.4821,
+      "step": 13481
+    },
+    {
+      "epoch": 1.088047776612057,
+      "grad_norm": 0.6506801247596741,
+      "learning_rate": 4.8458608538683694e-05,
+      "loss": 2.4686,
+      "step": 13482
+    },
+    {
+      "epoch": 1.0881284803486402,
+      "grad_norm": 0.7136878967285156,
+      "learning_rate": 4.844508073535939e-05,
+      "loss": 2.4523,
+      "step": 13483
+    },
+    {
+      "epoch": 1.088209184085223,
+      "grad_norm": 0.6663414239883423,
+      "learning_rate": 4.843155421691848e-05,
+      "loss": 2.4287,
+      "step": 13484
+    },
+    {
+      "epoch": 1.0882898878218061,
+      "grad_norm": 0.7192783355712891,
+      "learning_rate": 4.8418028983698006e-05,
+      "loss": 2.4433,
+      "step": 13485
+    },
+    {
+      "epoch": 1.0883705915583892,
+      "grad_norm": 0.6620980501174927,
+      "learning_rate": 4.8404505036035086e-05,
+      "loss": 2.4823,
+      "step": 13486
+    },
+    {
+      "epoch": 1.088451295294972,
+      "grad_norm": 0.6282123327255249,
+      "learning_rate": 4.83909823742668e-05,
+      "loss": 2.4641,
+      "step": 13487
+    },
+    {
+      "epoch": 1.0885319990315552,
+      "grad_norm": 0.6384354829788208,
+      "learning_rate": 4.837746099873012e-05,
+      "loss": 2.4234,
+      "step": 13488
+    },
+    {
+      "epoch": 1.0886127027681383,
+      "grad_norm": 0.6550076603889465,
+      "learning_rate": 4.836394090976204e-05,
+      "loss": 2.4743,
+      "step": 13489
+    },
+    {
+      "epoch": 1.0886934065047211,
+      "grad_norm": 0.6987888216972351,
+      "learning_rate": 4.8350422107699545e-05,
+      "loss": 2.4263,
+      "step": 13490
+    },
+    {
+      "epoch": 1.0887741102413042,
+      "grad_norm": 0.7012613415718079,
+      "learning_rate": 4.833690459287953e-05,
+      "loss": 2.4801,
+      "step": 13491
+    },
+    {
+      "epoch": 1.088854813977887,
+      "grad_norm": 0.6986923217773438,
+      "learning_rate": 4.832338836563891e-05,
+      "loss": 2.426,
+      "step": 13492
+    },
+    {
+      "epoch": 1.0889355177144702,
+      "grad_norm": 0.6936241984367371,
+      "learning_rate": 4.830987342631453e-05,
+      "loss": 2.4361,
+      "step": 13493
+    },
+    {
+      "epoch": 1.0890162214510533,
+      "grad_norm": 0.6612359881401062,
+      "learning_rate": 4.8296359775243275e-05,
+      "loss": 2.4385,
+      "step": 13494
+    },
+    {
+      "epoch": 1.0890969251876361,
+      "grad_norm": 0.6927692294120789,
+      "learning_rate": 4.828284741276183e-05,
+      "loss": 2.4692,
+      "step": 13495
+    },
+    {
+      "epoch": 1.0891776289242192,
+      "grad_norm": 0.6710225343704224,
+      "learning_rate": 4.8269336339207036e-05,
+      "loss": 2.4078,
+      "step": 13496
+    },
+    {
+      "epoch": 1.0892583326608023,
+      "grad_norm": 0.639076828956604,
+      "learning_rate": 4.825582655491564e-05,
+      "loss": 2.4368,
+      "step": 13497
+    },
+    {
+      "epoch": 1.0893390363973852,
+      "grad_norm": 0.7050483226776123,
+      "learning_rate": 4.824231806022426e-05,
+      "loss": 2.4308,
+      "step": 13498
+    },
+    {
+      "epoch": 1.0894197401339683,
+      "grad_norm": 0.7097769975662231,
+      "learning_rate": 4.822881085546962e-05,
+      "loss": 2.4378,
+      "step": 13499
+    },
+    {
+      "epoch": 1.0895004438705511,
+      "grad_norm": 0.6939458847045898,
+      "learning_rate": 4.821530494098834e-05,
+      "loss": 2.4678,
+      "step": 13500
+    },
+    {
+      "epoch": 1.0895811476071342,
+      "grad_norm": 0.6797441840171814,
+      "learning_rate": 4.8201800317117016e-05,
+      "loss": 2.4837,
+      "step": 13501
+    },
+    {
+      "epoch": 1.0896618513437173,
+      "grad_norm": 0.7451521158218384,
+      "learning_rate": 4.818829698419225e-05,
+      "loss": 2.4651,
+      "step": 13502
+    },
+    {
+      "epoch": 1.0897425550803002,
+      "grad_norm": 0.6749109625816345,
+      "learning_rate": 4.8174794942550585e-05,
+      "loss": 2.4569,
+      "step": 13503
+    },
+    {
+      "epoch": 1.0898232588168832,
+      "grad_norm": 0.6321636438369751,
+      "learning_rate": 4.8161294192528474e-05,
+      "loss": 2.4049,
+      "step": 13504
+    },
+    {
+      "epoch": 1.0899039625534663,
+      "grad_norm": 0.7002367377281189,
+      "learning_rate": 4.8147794734462415e-05,
+      "loss": 2.4489,
+      "step": 13505
+    },
+    {
+      "epoch": 1.0899846662900492,
+      "grad_norm": 0.758057713508606,
+      "learning_rate": 4.813429656868889e-05,
+      "loss": 2.436,
+      "step": 13506
+    },
+    {
+      "epoch": 1.0900653700266323,
+      "grad_norm": 0.6665529012680054,
+      "learning_rate": 4.812079969554424e-05,
+      "loss": 2.3805,
+      "step": 13507
+    },
+    {
+      "epoch": 1.0901460737632152,
+      "grad_norm": 0.6962547898292542,
+      "learning_rate": 4.810730411536487e-05,
+      "loss": 2.4203,
+      "step": 13508
+    },
+    {
+      "epoch": 1.0902267774997982,
+      "grad_norm": 0.6860647201538086,
+      "learning_rate": 4.809380982848712e-05,
+      "loss": 2.4482,
+      "step": 13509
+    },
+    {
+      "epoch": 1.0903074812363813,
+      "grad_norm": 0.7045090198516846,
+      "learning_rate": 4.808031683524733e-05,
+      "loss": 2.4155,
+      "step": 13510
+    },
+    {
+      "epoch": 1.0903881849729642,
+      "grad_norm": 0.6609304547309875,
+      "learning_rate": 4.806682513598176e-05,
+      "loss": 2.4295,
+      "step": 13511
+    },
+    {
+      "epoch": 1.0904688887095473,
+      "grad_norm": 0.7647323608398438,
+      "learning_rate": 4.8053334731026665e-05,
+      "loss": 2.4704,
+      "step": 13512
+    },
+    {
+      "epoch": 1.0905495924461301,
+      "grad_norm": 0.677449643611908,
+      "learning_rate": 4.803984562071829e-05,
+      "loss": 2.4501,
+      "step": 13513
+    },
+    {
+      "epoch": 1.0906302961827132,
+      "grad_norm": 0.645866334438324,
+      "learning_rate": 4.8026357805392754e-05,
+      "loss": 2.427,
+      "step": 13514
+    },
+    {
+      "epoch": 1.0907109999192963,
+      "grad_norm": 0.6968488097190857,
+      "learning_rate": 4.801287128538624e-05,
+      "loss": 2.3933,
+      "step": 13515
+    },
+    {
+      "epoch": 1.0907917036558792,
+      "grad_norm": 0.7137444615364075,
+      "learning_rate": 4.799938606103491e-05,
+      "loss": 2.4611,
+      "step": 13516
+    },
+    {
+      "epoch": 1.0908724073924623,
+      "grad_norm": 0.6860007047653198,
+      "learning_rate": 4.7985902132674765e-05,
+      "loss": 2.4252,
+      "step": 13517
+    },
+    {
+      "epoch": 1.0909531111290454,
+      "grad_norm": 0.726290762424469,
+      "learning_rate": 4.797241950064192e-05,
+      "loss": 2.44,
+      "step": 13518
+    },
+    {
+      "epoch": 1.0910338148656282,
+      "grad_norm": 0.6833362579345703,
+      "learning_rate": 4.795893816527241e-05,
+      "loss": 2.4199,
+      "step": 13519
+    },
+    {
+      "epoch": 1.0911145186022113,
+      "grad_norm": 0.7412242293357849,
+      "learning_rate": 4.794545812690212e-05,
+      "loss": 2.5412,
+      "step": 13520
+    },
+    {
+      "epoch": 1.0911952223387944,
+      "grad_norm": 0.6882274150848389,
+      "learning_rate": 4.793197938586712e-05,
+      "loss": 2.473,
+      "step": 13521
+    },
+    {
+      "epoch": 1.0912759260753773,
+      "grad_norm": 0.7334007024765015,
+      "learning_rate": 4.791850194250335e-05,
+      "loss": 2.4357,
+      "step": 13522
+    },
+    {
+      "epoch": 1.0913566298119604,
+      "grad_norm": 0.6564081311225891,
+      "learning_rate": 4.790502579714661e-05,
+      "loss": 2.4425,
+      "step": 13523
+    },
+    {
+      "epoch": 1.0914373335485432,
+      "grad_norm": 0.7045762538909912,
+      "learning_rate": 4.78915509501328e-05,
+      "loss": 2.4929,
+      "step": 13524
+    },
+    {
+      "epoch": 1.0915180372851263,
+      "grad_norm": 0.7512505650520325,
+      "learning_rate": 4.787807740179776e-05,
+      "loss": 2.4187,
+      "step": 13525
+    },
+    {
+      "epoch": 1.0915987410217094,
+      "grad_norm": 0.6592997908592224,
+      "learning_rate": 4.786460515247732e-05,
+      "loss": 2.4344,
+      "step": 13526
+    },
+    {
+      "epoch": 1.0916794447582923,
+      "grad_norm": 0.6721770763397217,
+      "learning_rate": 4.785113420250715e-05,
+      "loss": 2.4415,
+      "step": 13527
+    },
+    {
+      "epoch": 1.0917601484948753,
+      "grad_norm": 0.7544431686401367,
+      "learning_rate": 4.783766455222305e-05,
+      "loss": 2.4831,
+      "step": 13528
+    },
+    {
+      "epoch": 1.0918408522314582,
+      "grad_norm": 0.7226355671882629,
+      "learning_rate": 4.782419620196073e-05,
+      "loss": 2.4807,
+      "step": 13529
+    },
+    {
+      "epoch": 1.0919215559680413,
+      "grad_norm": 0.6386340260505676,
+      "learning_rate": 4.78107291520558e-05,
+      "loss": 2.4062,
+      "step": 13530
+    },
+    {
+      "epoch": 1.0920022597046244,
+      "grad_norm": 0.6670595407485962,
+      "learning_rate": 4.7797263402843926e-05,
+      "loss": 2.4009,
+      "step": 13531
+    },
+    {
+      "epoch": 1.0920829634412073,
+      "grad_norm": 0.6600756049156189,
+      "learning_rate": 4.778379895466071e-05,
+      "loss": 2.4321,
+      "step": 13532
+    },
+    {
+      "epoch": 1.0921636671777903,
+      "grad_norm": 0.7190701961517334,
+      "learning_rate": 4.77703358078417e-05,
+      "loss": 2.4229,
+      "step": 13533
+    },
+    {
+      "epoch": 1.0922443709143734,
+      "grad_norm": 0.6554828882217407,
+      "learning_rate": 4.775687396272247e-05,
+      "loss": 2.442,
+      "step": 13534
+    },
+    {
+      "epoch": 1.0923250746509563,
+      "grad_norm": 0.6720205545425415,
+      "learning_rate": 4.774341341963853e-05,
+      "loss": 2.4994,
+      "step": 13535
+    },
+    {
+      "epoch": 1.0924057783875394,
+      "grad_norm": 0.7161003947257996,
+      "learning_rate": 4.7729954178925295e-05,
+      "loss": 2.4666,
+      "step": 13536
+    },
+    {
+      "epoch": 1.0924864821241222,
+      "grad_norm": 0.6817156672477722,
+      "learning_rate": 4.771649624091824e-05,
+      "loss": 2.4203,
+      "step": 13537
+    },
+    {
+      "epoch": 1.0925671858607053,
+      "grad_norm": 0.7167035937309265,
+      "learning_rate": 4.770303960595277e-05,
+      "loss": 2.4214,
+      "step": 13538
+    },
+    {
+      "epoch": 1.0926478895972884,
+      "grad_norm": 0.6373945474624634,
+      "learning_rate": 4.768958427436429e-05,
+      "loss": 2.485,
+      "step": 13539
+    },
+    {
+      "epoch": 1.0927285933338713,
+      "grad_norm": 0.7361387014389038,
+      "learning_rate": 4.767613024648808e-05,
+      "loss": 2.5192,
+      "step": 13540
+    },
+    {
+      "epoch": 1.0928092970704544,
+      "grad_norm": 0.7034375667572021,
+      "learning_rate": 4.766267752265947e-05,
+      "loss": 2.4324,
+      "step": 13541
+    },
+    {
+      "epoch": 1.0928900008070375,
+      "grad_norm": 0.7355689406394958,
+      "learning_rate": 4.7649226103213765e-05,
+      "loss": 2.5048,
+      "step": 13542
+    },
+    {
+      "epoch": 1.0929707045436203,
+      "grad_norm": 0.7120445966720581,
+      "learning_rate": 4.7635775988486176e-05,
+      "loss": 2.449,
+      "step": 13543
+    },
+    {
+      "epoch": 1.0930514082802034,
+      "grad_norm": 0.695888876914978,
+      "learning_rate": 4.7622327178811935e-05,
+      "loss": 2.4974,
+      "step": 13544
+    },
+    {
+      "epoch": 1.0931321120167863,
+      "grad_norm": 0.6953639984130859,
+      "learning_rate": 4.760887967452625e-05,
+      "loss": 2.3927,
+      "step": 13545
+    },
+    {
+      "epoch": 1.0932128157533694,
+      "grad_norm": 0.6457183957099915,
+      "learning_rate": 4.759543347596421e-05,
+      "loss": 2.4501,
+      "step": 13546
+    },
+    {
+      "epoch": 1.0932935194899525,
+      "grad_norm": 0.7259296774864197,
+      "learning_rate": 4.7581988583460946e-05,
+      "loss": 2.4896,
+      "step": 13547
+    },
+    {
+      "epoch": 1.0933742232265353,
+      "grad_norm": 0.6897724270820618,
+      "learning_rate": 4.7568544997351586e-05,
+      "loss": 2.4181,
+      "step": 13548
+    },
+    {
+      "epoch": 1.0934549269631184,
+      "grad_norm": 0.6723688840866089,
+      "learning_rate": 4.755510271797111e-05,
+      "loss": 2.5097,
+      "step": 13549
+    },
+    {
+      "epoch": 1.0935356306997015,
+      "grad_norm": 0.7353307604789734,
+      "learning_rate": 4.754166174565456e-05,
+      "loss": 2.4548,
+      "step": 13550
+    },
+    {
+      "epoch": 1.0936163344362844,
+      "grad_norm": 0.7334069013595581,
+      "learning_rate": 4.752822208073693e-05,
+      "loss": 2.5113,
+      "step": 13551
+    },
+    {
+      "epoch": 1.0936970381728675,
+      "grad_norm": 0.6581420302391052,
+      "learning_rate": 4.751478372355317e-05,
+      "loss": 2.4546,
+      "step": 13552
+    },
+    {
+      "epoch": 1.0937777419094503,
+      "grad_norm": 0.7890802621841431,
+      "learning_rate": 4.75013466744382e-05,
+      "loss": 2.4092,
+      "step": 13553
+    },
+    {
+      "epoch": 1.0938584456460334,
+      "grad_norm": 0.7226595282554626,
+      "learning_rate": 4.7487910933726895e-05,
+      "loss": 2.457,
+      "step": 13554
+    },
+    {
+      "epoch": 1.0939391493826165,
+      "grad_norm": 0.7108014225959778,
+      "learning_rate": 4.7474476501754165e-05,
+      "loss": 2.471,
+      "step": 13555
+    },
+    {
+      "epoch": 1.0940198531191994,
+      "grad_norm": 0.6864863038063049,
+      "learning_rate": 4.746104337885473e-05,
+      "loss": 2.4778,
+      "step": 13556
+    },
+    {
+      "epoch": 1.0941005568557824,
+      "grad_norm": 0.6890624165534973,
+      "learning_rate": 4.744761156536345e-05,
+      "loss": 2.456,
+      "step": 13557
+    },
+    {
+      "epoch": 1.0941812605923653,
+      "grad_norm": 0.7052781581878662,
+      "learning_rate": 4.743418106161509e-05,
+      "loss": 2.4796,
+      "step": 13558
+    },
+    {
+      "epoch": 1.0942619643289484,
+      "grad_norm": 0.6569164991378784,
+      "learning_rate": 4.742075186794431e-05,
+      "loss": 2.469,
+      "step": 13559
+    },
+    {
+      "epoch": 1.0943426680655315,
+      "grad_norm": 0.7302874326705933,
+      "learning_rate": 4.7407323984685836e-05,
+      "loss": 2.4543,
+      "step": 13560
+    },
+    {
+      "epoch": 1.0944233718021144,
+      "grad_norm": 0.6499345898628235,
+      "learning_rate": 4.7393897412174335e-05,
+      "loss": 2.4037,
+      "step": 13561
+    },
+    {
+      "epoch": 1.0945040755386974,
+      "grad_norm": 0.6643944382667542,
+      "learning_rate": 4.7380472150744416e-05,
+      "loss": 2.4067,
+      "step": 13562
+    },
+    {
+      "epoch": 1.0945847792752805,
+      "grad_norm": 0.7491872906684875,
+      "learning_rate": 4.736704820073069e-05,
+      "loss": 2.4277,
+      "step": 13563
+    },
+    {
+      "epoch": 1.0946654830118634,
+      "grad_norm": 0.7319512367248535,
+      "learning_rate": 4.735362556246773e-05,
+      "loss": 2.4588,
+      "step": 13564
+    },
+    {
+      "epoch": 1.0947461867484465,
+      "grad_norm": 0.7404350638389587,
+      "learning_rate": 4.734020423629001e-05,
+      "loss": 2.432,
+      "step": 13565
+    },
+    {
+      "epoch": 1.0948268904850296,
+      "grad_norm": 0.6462193727493286,
+      "learning_rate": 4.732678422253206e-05,
+      "loss": 2.4417,
+      "step": 13566
+    },
+    {
+      "epoch": 1.0949075942216124,
+      "grad_norm": 0.6711323857307434,
+      "learning_rate": 4.731336552152836e-05,
+      "loss": 2.4023,
+      "step": 13567
+    },
+    {
+      "epoch": 1.0949882979581955,
+      "grad_norm": 0.658261239528656,
+      "learning_rate": 4.729994813361329e-05,
+      "loss": 2.4132,
+      "step": 13568
+    },
+    {
+      "epoch": 1.0950690016947784,
+      "grad_norm": 0.8081904053688049,
+      "learning_rate": 4.728653205912127e-05,
+      "loss": 2.4412,
+      "step": 13569
+    },
+    {
+      "epoch": 1.0951497054313615,
+      "grad_norm": 0.6620786786079407,
+      "learning_rate": 4.727311729838666e-05,
+      "loss": 2.4357,
+      "step": 13570
+    },
+    {
+      "epoch": 1.0952304091679446,
+      "grad_norm": 0.7026848793029785,
+      "learning_rate": 4.725970385174381e-05,
+      "loss": 2.4159,
+      "step": 13571
+    },
+    {
+      "epoch": 1.0953111129045274,
+      "grad_norm": 0.7017392516136169,
+      "learning_rate": 4.7246291719526995e-05,
+      "loss": 2.4253,
+      "step": 13572
+    },
+    {
+      "epoch": 1.0953918166411105,
+      "grad_norm": 0.710172712802887,
+      "learning_rate": 4.7232880902070483e-05,
+      "loss": 2.4057,
+      "step": 13573
+    },
+    {
+      "epoch": 1.0954725203776934,
+      "grad_norm": 0.7208876013755798,
+      "learning_rate": 4.721947139970856e-05,
+      "loss": 2.4803,
+      "step": 13574
+    },
+    {
+      "epoch": 1.0955532241142765,
+      "grad_norm": 0.693219006061554,
+      "learning_rate": 4.720606321277534e-05,
+      "loss": 2.3611,
+      "step": 13575
+    },
+    {
+      "epoch": 1.0956339278508596,
+      "grad_norm": 0.737206757068634,
+      "learning_rate": 4.7192656341605026e-05,
+      "loss": 2.3873,
+      "step": 13576
+    },
+    {
+      "epoch": 1.0957146315874424,
+      "grad_norm": 0.6605268120765686,
+      "learning_rate": 4.717925078653179e-05,
+      "loss": 2.4155,
+      "step": 13577
+    },
+    {
+      "epoch": 1.0957953353240255,
+      "grad_norm": 0.7143047451972961,
+      "learning_rate": 4.716584654788967e-05,
+      "loss": 2.4526,
+      "step": 13578
+    },
+    {
+      "epoch": 1.0958760390606086,
+      "grad_norm": 0.6980953216552734,
+      "learning_rate": 4.715244362601277e-05,
+      "loss": 2.4422,
+      "step": 13579
+    },
+    {
+      "epoch": 1.0959567427971915,
+      "grad_norm": 0.6852009892463684,
+      "learning_rate": 4.713904202123515e-05,
+      "loss": 2.4599,
+      "step": 13580
+    },
+    {
+      "epoch": 1.0960374465337746,
+      "grad_norm": 0.7436656355857849,
+      "learning_rate": 4.712564173389074e-05,
+      "loss": 2.4441,
+      "step": 13581
+    },
+    {
+      "epoch": 1.0961181502703574,
+      "grad_norm": 0.7090624570846558,
+      "learning_rate": 4.711224276431352e-05,
+      "loss": 2.4741,
+      "step": 13582
+    },
+    {
+      "epoch": 1.0961988540069405,
+      "grad_norm": 0.6611043810844421,
+      "learning_rate": 4.709884511283753e-05,
+      "loss": 2.4589,
+      "step": 13583
+    },
+    {
+      "epoch": 1.0962795577435236,
+      "grad_norm": 0.6932426691055298,
+      "learning_rate": 4.708544877979658e-05,
+      "loss": 2.4199,
+      "step": 13584
+    },
+    {
+      "epoch": 1.0963602614801065,
+      "grad_norm": 0.7629422545433044,
+      "learning_rate": 4.707205376552456e-05,
+      "loss": 2.4588,
+      "step": 13585
+    },
+    {
+      "epoch": 1.0964409652166895,
+      "grad_norm": 0.8116739392280579,
+      "learning_rate": 4.705866007035531e-05,
+      "loss": 2.472,
+      "step": 13586
+    },
+    {
+      "epoch": 1.0965216689532726,
+      "grad_norm": 0.6711297631263733,
+      "learning_rate": 4.704526769462269e-05,
+      "loss": 2.4086,
+      "step": 13587
+    },
+    {
+      "epoch": 1.0966023726898555,
+      "grad_norm": 0.716015636920929,
+      "learning_rate": 4.703187663866037e-05,
+      "loss": 2.4411,
+      "step": 13588
+    },
+    {
+      "epoch": 1.0966830764264386,
+      "grad_norm": 0.6982430219650269,
+      "learning_rate": 4.701848690280215e-05,
+      "loss": 2.4438,
+      "step": 13589
+    },
+    {
+      "epoch": 1.0967637801630215,
+      "grad_norm": 0.7183159589767456,
+      "learning_rate": 4.7005098487381785e-05,
+      "loss": 2.4464,
+      "step": 13590
+    },
+    {
+      "epoch": 1.0968444838996045,
+      "grad_norm": 0.6983399391174316,
+      "learning_rate": 4.699171139273284e-05,
+      "loss": 2.4354,
+      "step": 13591
+    },
+    {
+      "epoch": 1.0969251876361876,
+      "grad_norm": 0.7157938480377197,
+      "learning_rate": 4.697832561918901e-05,
+      "loss": 2.4393,
+      "step": 13592
+    },
+    {
+      "epoch": 1.0970058913727705,
+      "grad_norm": 0.6991363763809204,
+      "learning_rate": 4.696494116708392e-05,
+      "loss": 2.4723,
+      "step": 13593
+    },
+    {
+      "epoch": 1.0970865951093536,
+      "grad_norm": 0.6722309589385986,
+      "learning_rate": 4.695155803675112e-05,
+      "loss": 2.447,
+      "step": 13594
+    },
+    {
+      "epoch": 1.0971672988459367,
+      "grad_norm": 0.6492688655853271,
+      "learning_rate": 4.6938176228524175e-05,
+      "loss": 2.4213,
+      "step": 13595
+    },
+    {
+      "epoch": 1.0972480025825195,
+      "grad_norm": 0.6941642165184021,
+      "learning_rate": 4.6924795742736616e-05,
+      "loss": 2.4714,
+      "step": 13596
+    },
+    {
+      "epoch": 1.0973287063191026,
+      "grad_norm": 0.7506042122840881,
+      "learning_rate": 4.691141657972185e-05,
+      "loss": 2.4563,
+      "step": 13597
+    },
+    {
+      "epoch": 1.0974094100556855,
+      "grad_norm": 0.7032836675643921,
+      "learning_rate": 4.6898038739813356e-05,
+      "loss": 2.4824,
+      "step": 13598
+    },
+    {
+      "epoch": 1.0974901137922686,
+      "grad_norm": 0.6908734440803528,
+      "learning_rate": 4.6884662223344575e-05,
+      "loss": 2.4486,
+      "step": 13599
+    },
+    {
+      "epoch": 1.0975708175288517,
+      "grad_norm": 0.714971661567688,
+      "learning_rate": 4.687128703064883e-05,
+      "loss": 2.4372,
+      "step": 13600
+    },
+    {
+      "epoch": 1.0976515212654345,
+      "grad_norm": 0.6989198327064514,
+      "learning_rate": 4.6857913162059486e-05,
+      "loss": 2.395,
+      "step": 13601
+    },
+    {
+      "epoch": 1.0977322250020176,
+      "grad_norm": 0.7163406014442444,
+      "learning_rate": 4.684454061790987e-05,
+      "loss": 2.4868,
+      "step": 13602
+    },
+    {
+      "epoch": 1.0978129287386005,
+      "grad_norm": 0.6600626707077026,
+      "learning_rate": 4.6831169398533245e-05,
+      "loss": 2.5134,
+      "step": 13603
+    },
+    {
+      "epoch": 1.0978936324751836,
+      "grad_norm": 0.6657080054283142,
+      "learning_rate": 4.681779950426286e-05,
+      "loss": 2.4701,
+      "step": 13604
+    },
+    {
+      "epoch": 1.0979743362117667,
+      "grad_norm": 0.665860116481781,
+      "learning_rate": 4.680443093543194e-05,
+      "loss": 2.4593,
+      "step": 13605
+    },
+    {
+      "epoch": 1.0980550399483495,
+      "grad_norm": 0.7000327110290527,
+      "learning_rate": 4.679106369237368e-05,
+      "loss": 2.4523,
+      "step": 13606
+    },
+    {
+      "epoch": 1.0981357436849326,
+      "grad_norm": 0.6969157457351685,
+      "learning_rate": 4.677769777542118e-05,
+      "loss": 2.4935,
+      "step": 13607
+    },
+    {
+      "epoch": 1.0982164474215157,
+      "grad_norm": 0.6864836812019348,
+      "learning_rate": 4.676433318490757e-05,
+      "loss": 2.457,
+      "step": 13608
+    },
+    {
+      "epoch": 1.0982971511580986,
+      "grad_norm": 0.7331364750862122,
+      "learning_rate": 4.675096992116598e-05,
+      "loss": 2.4253,
+      "step": 13609
+    },
+    {
+      "epoch": 1.0983778548946816,
+      "grad_norm": 0.75,
+      "learning_rate": 4.673760798452936e-05,
+      "loss": 2.4147,
+      "step": 13610
+    },
+    {
+      "epoch": 1.0984585586312647,
+      "grad_norm": 0.6589440703392029,
+      "learning_rate": 4.6724247375330786e-05,
+      "loss": 2.4718,
+      "step": 13611
+    },
+    {
+      "epoch": 1.0985392623678476,
+      "grad_norm": 0.7032667994499207,
+      "learning_rate": 4.671088809390324e-05,
+      "loss": 2.4724,
+      "step": 13612
+    },
+    {
+      "epoch": 1.0986199661044307,
+      "grad_norm": 0.7544135451316833,
+      "learning_rate": 4.6697530140579646e-05,
+      "loss": 2.4804,
+      "step": 13613
+    },
+    {
+      "epoch": 1.0987006698410136,
+      "grad_norm": 0.6503081917762756,
+      "learning_rate": 4.668417351569295e-05,
+      "loss": 2.3829,
+      "step": 13614
+    },
+    {
+      "epoch": 1.0987813735775966,
+      "grad_norm": 0.6928786039352417,
+      "learning_rate": 4.667081821957605e-05,
+      "loss": 2.5678,
+      "step": 13615
+    },
+    {
+      "epoch": 1.0988620773141797,
+      "grad_norm": 0.6652864217758179,
+      "learning_rate": 4.665746425256173e-05,
+      "loss": 2.4585,
+      "step": 13616
+    },
+    {
+      "epoch": 1.0989427810507626,
+      "grad_norm": 0.700265109539032,
+      "learning_rate": 4.664411161498283e-05,
+      "loss": 2.4785,
+      "step": 13617
+    },
+    {
+      "epoch": 1.0990234847873457,
+      "grad_norm": 0.7443608045578003,
+      "learning_rate": 4.663076030717216e-05,
+      "loss": 2.4869,
+      "step": 13618
+    },
+    {
+      "epoch": 1.0991041885239285,
+      "grad_norm": 0.7037705779075623,
+      "learning_rate": 4.6617410329462477e-05,
+      "loss": 2.4518,
+      "step": 13619
+    },
+    {
+      "epoch": 1.0991848922605116,
+      "grad_norm": 0.7528365850448608,
+      "learning_rate": 4.660406168218643e-05,
+      "loss": 2.4616,
+      "step": 13620
+    },
+    {
+      "epoch": 1.0992655959970947,
+      "grad_norm": 0.7149221301078796,
+      "learning_rate": 4.659071436567676e-05,
+      "loss": 2.4661,
+      "step": 13621
+    },
+    {
+      "epoch": 1.0993462997336776,
+      "grad_norm": 0.7212862968444824,
+      "learning_rate": 4.657736838026608e-05,
+      "loss": 2.4424,
+      "step": 13622
+    },
+    {
+      "epoch": 1.0994270034702607,
+      "grad_norm": 0.6934216022491455,
+      "learning_rate": 4.6564023726287045e-05,
+      "loss": 2.4633,
+      "step": 13623
+    },
+    {
+      "epoch": 1.0995077072068438,
+      "grad_norm": 0.7244036793708801,
+      "learning_rate": 4.655068040407221e-05,
+      "loss": 2.409,
+      "step": 13624
+    },
+    {
+      "epoch": 1.0995884109434266,
+      "grad_norm": 0.6911318898200989,
+      "learning_rate": 4.653733841395419e-05,
+      "loss": 2.5117,
+      "step": 13625
+    },
+    {
+      "epoch": 1.0996691146800097,
+      "grad_norm": 0.7579816579818726,
+      "learning_rate": 4.65239977562654e-05,
+      "loss": 2.4927,
+      "step": 13626
+    },
+    {
+      "epoch": 1.0997498184165928,
+      "grad_norm": 0.7699651122093201,
+      "learning_rate": 4.651065843133837e-05,
+      "loss": 2.4083,
+      "step": 13627
+    },
+    {
+      "epoch": 1.0998305221531757,
+      "grad_norm": 0.6669431328773499,
+      "learning_rate": 4.649732043950561e-05,
+      "loss": 2.4402,
+      "step": 13628
+    },
+    {
+      "epoch": 1.0999112258897588,
+      "grad_norm": 0.7134940028190613,
+      "learning_rate": 4.6483983781099426e-05,
+      "loss": 2.4275,
+      "step": 13629
+    },
+    {
+      "epoch": 1.0999919296263416,
+      "grad_norm": 0.7107651233673096,
+      "learning_rate": 4.647064845645227e-05,
+      "loss": 2.4654,
+      "step": 13630
+    },
+    {
+      "epoch": 1.1000726333629247,
+      "grad_norm": 0.7101391553878784,
+      "learning_rate": 4.645731446589652e-05,
+      "loss": 2.4357,
+      "step": 13631
+    },
+    {
+      "epoch": 1.1001533370995078,
+      "grad_norm": 0.7511606216430664,
+      "learning_rate": 4.6443981809764405e-05,
+      "loss": 2.5016,
+      "step": 13632
+    },
+    {
+      "epoch": 1.1002340408360907,
+      "grad_norm": 0.7315953373908997,
+      "learning_rate": 4.6430650488388226e-05,
+      "loss": 2.4541,
+      "step": 13633
+    },
+    {
+      "epoch": 1.1003147445726738,
+      "grad_norm": 0.6701769232749939,
+      "learning_rate": 4.6417320502100316e-05,
+      "loss": 2.4071,
+      "step": 13634
+    },
+    {
+      "epoch": 1.1003954483092566,
+      "grad_norm": 0.7164294123649597,
+      "learning_rate": 4.6403991851232876e-05,
+      "loss": 2.478,
+      "step": 13635
+    },
+    {
+      "epoch": 1.1004761520458397,
+      "grad_norm": 0.7003894448280334,
+      "learning_rate": 4.639066453611802e-05,
+      "loss": 2.4686,
+      "step": 13636
+    },
+    {
+      "epoch": 1.1005568557824228,
+      "grad_norm": 0.6855250000953674,
+      "learning_rate": 4.6377338557087957e-05,
+      "loss": 2.4531,
+      "step": 13637
+    },
+    {
+      "epoch": 1.1006375595190057,
+      "grad_norm": 0.6581299901008606,
+      "learning_rate": 4.6364013914474816e-05,
+      "loss": 2.4511,
+      "step": 13638
+    },
+    {
+      "epoch": 1.1007182632555887,
+      "grad_norm": 0.7599080204963684,
+      "learning_rate": 4.6350690608610604e-05,
+      "loss": 2.5143,
+      "step": 13639
+    },
+    {
+      "epoch": 1.1007989669921718,
+      "grad_norm": 0.7029981017112732,
+      "learning_rate": 4.633736863982744e-05,
+      "loss": 2.4541,
+      "step": 13640
+    },
+    {
+      "epoch": 1.1008796707287547,
+      "grad_norm": 0.7378708720207214,
+      "learning_rate": 4.6324048008457357e-05,
+      "loss": 2.4319,
+      "step": 13641
+    },
+    {
+      "epoch": 1.1009603744653378,
+      "grad_norm": 0.7087826728820801,
+      "learning_rate": 4.631072871483226e-05,
+      "loss": 2.4148,
+      "step": 13642
+    },
+    {
+      "epoch": 1.1010410782019207,
+      "grad_norm": 0.7000819444656372,
+      "learning_rate": 4.629741075928415e-05,
+      "loss": 2.4692,
+      "step": 13643
+    },
+    {
+      "epoch": 1.1011217819385037,
+      "grad_norm": 0.7363965511322021,
+      "learning_rate": 4.628409414214496e-05,
+      "loss": 2.4584,
+      "step": 13644
+    },
+    {
+      "epoch": 1.1012024856750868,
+      "grad_norm": 0.6691753268241882,
+      "learning_rate": 4.627077886374656e-05,
+      "loss": 2.4356,
+      "step": 13645
+    },
+    {
+      "epoch": 1.1012831894116697,
+      "grad_norm": 0.6864185929298401,
+      "learning_rate": 4.625746492442078e-05,
+      "loss": 2.4713,
+      "step": 13646
+    },
+    {
+      "epoch": 1.1013638931482528,
+      "grad_norm": 0.714318573474884,
+      "learning_rate": 4.624415232449947e-05,
+      "loss": 2.4482,
+      "step": 13647
+    },
+    {
+      "epoch": 1.1014445968848359,
+      "grad_norm": 0.6383495330810547,
+      "learning_rate": 4.623084106431444e-05,
+      "loss": 2.4248,
+      "step": 13648
+    },
+    {
+      "epoch": 1.1015253006214187,
+      "grad_norm": 0.7014495730400085,
+      "learning_rate": 4.6217531144197365e-05,
+      "loss": 2.4393,
+      "step": 13649
+    },
+    {
+      "epoch": 1.1016060043580018,
+      "grad_norm": 0.8128634095191956,
+      "learning_rate": 4.620422256448e-05,
+      "loss": 2.4741,
+      "step": 13650
+    },
+    {
+      "epoch": 1.1016867080945847,
+      "grad_norm": 0.7333208322525024,
+      "learning_rate": 4.619091532549408e-05,
+      "loss": 2.4288,
+      "step": 13651
+    },
+    {
+      "epoch": 1.1017674118311678,
+      "grad_norm": 0.7023218274116516,
+      "learning_rate": 4.617760942757117e-05,
+      "loss": 2.5025,
+      "step": 13652
+    },
+    {
+      "epoch": 1.1018481155677509,
+      "grad_norm": 0.6420873403549194,
+      "learning_rate": 4.616430487104292e-05,
+      "loss": 2.4165,
+      "step": 13653
+    },
+    {
+      "epoch": 1.1019288193043337,
+      "grad_norm": 0.6767684817314148,
+      "learning_rate": 4.615100165624092e-05,
+      "loss": 2.4642,
+      "step": 13654
+    },
+    {
+      "epoch": 1.1020095230409168,
+      "grad_norm": 0.7361159920692444,
+      "learning_rate": 4.613769978349672e-05,
+      "loss": 2.5343,
+      "step": 13655
+    },
+    {
+      "epoch": 1.1020902267775,
+      "grad_norm": 0.6642624735832214,
+      "learning_rate": 4.6124399253141846e-05,
+      "loss": 2.3769,
+      "step": 13656
+    },
+    {
+      "epoch": 1.1021709305140828,
+      "grad_norm": 0.6912256479263306,
+      "learning_rate": 4.611110006550781e-05,
+      "loss": 2.455,
+      "step": 13657
+    },
+    {
+      "epoch": 1.1022516342506659,
+      "grad_norm": 0.7419310212135315,
+      "learning_rate": 4.609780222092599e-05,
+      "loss": 2.4171,
+      "step": 13658
+    },
+    {
+      "epoch": 1.1023323379872487,
+      "grad_norm": 0.718953549861908,
+      "learning_rate": 4.6084505719727835e-05,
+      "loss": 2.4791,
+      "step": 13659
+    },
+    {
+      "epoch": 1.1024130417238318,
+      "grad_norm": 0.7904248237609863,
+      "learning_rate": 4.607121056224477e-05,
+      "loss": 2.4429,
+      "step": 13660
+    },
+    {
+      "epoch": 1.102493745460415,
+      "grad_norm": 0.6743534803390503,
+      "learning_rate": 4.605791674880808e-05,
+      "loss": 2.4481,
+      "step": 13661
+    },
+    {
+      "epoch": 1.1025744491969978,
+      "grad_norm": 0.6829143166542053,
+      "learning_rate": 4.6044624279749106e-05,
+      "loss": 2.4078,
+      "step": 13662
+    },
+    {
+      "epoch": 1.1026551529335809,
+      "grad_norm": 0.6803167462348938,
+      "learning_rate": 4.6031333155399136e-05,
+      "loss": 2.4509,
+      "step": 13663
+    },
+    {
+      "epoch": 1.1027358566701637,
+      "grad_norm": 0.7474592328071594,
+      "learning_rate": 4.601804337608943e-05,
+      "loss": 2.4563,
+      "step": 13664
+    },
+    {
+      "epoch": 1.1028165604067468,
+      "grad_norm": 0.6753630042076111,
+      "learning_rate": 4.6004754942151174e-05,
+      "loss": 2.4285,
+      "step": 13665
+    },
+    {
+      "epoch": 1.10289726414333,
+      "grad_norm": 0.7990161180496216,
+      "learning_rate": 4.599146785391558e-05,
+      "loss": 2.4907,
+      "step": 13666
+    },
+    {
+      "epoch": 1.1029779678799128,
+      "grad_norm": 0.8161290287971497,
+      "learning_rate": 4.597818211171383e-05,
+      "loss": 2.4599,
+      "step": 13667
+    },
+    {
+      "epoch": 1.1030586716164958,
+      "grad_norm": 0.6813610792160034,
+      "learning_rate": 4.596489771587695e-05,
+      "loss": 2.4484,
+      "step": 13668
+    },
+    {
+      "epoch": 1.103139375353079,
+      "grad_norm": 0.6598966121673584,
+      "learning_rate": 4.5951614666736076e-05,
+      "loss": 2.4326,
+      "step": 13669
+    },
+    {
+      "epoch": 1.1032200790896618,
+      "grad_norm": 0.7084827423095703,
+      "learning_rate": 4.593833296462228e-05,
+      "loss": 2.4188,
+      "step": 13670
+    },
+    {
+      "epoch": 1.1033007828262449,
+      "grad_norm": 0.6876685619354248,
+      "learning_rate": 4.59250526098665e-05,
+      "loss": 2.4482,
+      "step": 13671
+    },
+    {
+      "epoch": 1.103381486562828,
+      "grad_norm": 0.7292699813842773,
+      "learning_rate": 4.591177360279978e-05,
+      "loss": 2.4452,
+      "step": 13672
+    },
+    {
+      "epoch": 1.1034621902994108,
+      "grad_norm": 0.7057675123214722,
+      "learning_rate": 4.589849594375304e-05,
+      "loss": 2.4336,
+      "step": 13673
+    },
+    {
+      "epoch": 1.103542894035994,
+      "grad_norm": 0.7684180736541748,
+      "learning_rate": 4.5885219633057196e-05,
+      "loss": 2.4453,
+      "step": 13674
+    },
+    {
+      "epoch": 1.1036235977725768,
+      "grad_norm": 0.7107112407684326,
+      "learning_rate": 4.5871944671043154e-05,
+      "loss": 2.4116,
+      "step": 13675
+    },
+    {
+      "epoch": 1.1037043015091599,
+      "grad_norm": 0.659501314163208,
+      "learning_rate": 4.585867105804177e-05,
+      "loss": 2.4907,
+      "step": 13676
+    },
+    {
+      "epoch": 1.103785005245743,
+      "grad_norm": 0.7553967833518982,
+      "learning_rate": 4.5845398794383786e-05,
+      "loss": 2.3982,
+      "step": 13677
+    },
+    {
+      "epoch": 1.1038657089823258,
+      "grad_norm": 0.6861104965209961,
+      "learning_rate": 4.583212788040003e-05,
+      "loss": 2.416,
+      "step": 13678
+    },
+    {
+      "epoch": 1.103946412718909,
+      "grad_norm": 0.6546811461448669,
+      "learning_rate": 4.5818858316421254e-05,
+      "loss": 2.4506,
+      "step": 13679
+    },
+    {
+      "epoch": 1.1040271164554918,
+      "grad_norm": 0.7012909650802612,
+      "learning_rate": 4.58055901027782e-05,
+      "loss": 2.439,
+      "step": 13680
+    },
+    {
+      "epoch": 1.1041078201920749,
+      "grad_norm": 0.7594780325889587,
+      "learning_rate": 4.5792323239801446e-05,
+      "loss": 2.4437,
+      "step": 13681
+    },
+    {
+      "epoch": 1.104188523928658,
+      "grad_norm": 0.6576492190361023,
+      "learning_rate": 4.577905772782172e-05,
+      "loss": 2.443,
+      "step": 13682
+    },
+    {
+      "epoch": 1.1042692276652408,
+      "grad_norm": 0.6751925349235535,
+      "learning_rate": 4.576579356716963e-05,
+      "loss": 2.507,
+      "step": 13683
+    },
+    {
+      "epoch": 1.104349931401824,
+      "grad_norm": 0.7206710577011108,
+      "learning_rate": 4.575253075817567e-05,
+      "loss": 2.4236,
+      "step": 13684
+    },
+    {
+      "epoch": 1.104430635138407,
+      "grad_norm": 0.7736170291900635,
+      "learning_rate": 4.5739269301170485e-05,
+      "loss": 2.4095,
+      "step": 13685
+    },
+    {
+      "epoch": 1.1045113388749899,
+      "grad_norm": 0.6901736855506897,
+      "learning_rate": 4.572600919648457e-05,
+      "loss": 2.4519,
+      "step": 13686
+    },
+    {
+      "epoch": 1.104592042611573,
+      "grad_norm": 0.7762539982795715,
+      "learning_rate": 4.571275044444836e-05,
+      "loss": 2.5018,
+      "step": 13687
+    },
+    {
+      "epoch": 1.1046727463481558,
+      "grad_norm": 0.7231423854827881,
+      "learning_rate": 4.569949304539232e-05,
+      "loss": 2.4553,
+      "step": 13688
+    },
+    {
+      "epoch": 1.104753450084739,
+      "grad_norm": 0.7713531255722046,
+      "learning_rate": 4.568623699964688e-05,
+      "loss": 2.49,
+      "step": 13689
+    },
+    {
+      "epoch": 1.104834153821322,
+      "grad_norm": 0.7355079650878906,
+      "learning_rate": 4.5672982307542354e-05,
+      "loss": 2.5191,
+      "step": 13690
+    },
+    {
+      "epoch": 1.1049148575579049,
+      "grad_norm": 0.6916452050209045,
+      "learning_rate": 4.565972896940913e-05,
+      "loss": 2.3867,
+      "step": 13691
+    },
+    {
+      "epoch": 1.104995561294488,
+      "grad_norm": 0.6622549295425415,
+      "learning_rate": 4.5646476985577544e-05,
+      "loss": 2.4364,
+      "step": 13692
+    },
+    {
+      "epoch": 1.105076265031071,
+      "grad_norm": 0.6683297157287598,
+      "learning_rate": 4.563322635637779e-05,
+      "loss": 2.43,
+      "step": 13693
+    },
+    {
+      "epoch": 1.105156968767654,
+      "grad_norm": 0.6857880353927612,
+      "learning_rate": 4.561997708214015e-05,
+      "loss": 2.4515,
+      "step": 13694
+    },
+    {
+      "epoch": 1.105237672504237,
+      "grad_norm": 0.7473817467689514,
+      "learning_rate": 4.5606729163194807e-05,
+      "loss": 2.442,
+      "step": 13695
+    },
+    {
+      "epoch": 1.1053183762408199,
+      "grad_norm": 0.6988846063613892,
+      "learning_rate": 4.559348259987203e-05,
+      "loss": 2.3886,
+      "step": 13696
+    },
+    {
+      "epoch": 1.105399079977403,
+      "grad_norm": 0.6450650691986084,
+      "learning_rate": 4.5580237392501836e-05,
+      "loss": 2.4647,
+      "step": 13697
+    },
+    {
+      "epoch": 1.105479783713986,
+      "grad_norm": 0.7669623494148254,
+      "learning_rate": 4.556699354141439e-05,
+      "loss": 2.4362,
+      "step": 13698
+    },
+    {
+      "epoch": 1.105560487450569,
+      "grad_norm": 0.7019730806350708,
+      "learning_rate": 4.55537510469398e-05,
+      "loss": 2.49,
+      "step": 13699
+    },
+    {
+      "epoch": 1.105641191187152,
+      "grad_norm": 0.6736636757850647,
+      "learning_rate": 4.5540509909408e-05,
+      "loss": 2.43,
+      "step": 13700
+    },
+    {
+      "epoch": 1.105721894923735,
+      "grad_norm": 0.6872034668922424,
+      "learning_rate": 4.552727012914907e-05,
+      "loss": 2.4507,
+      "step": 13701
+    },
+    {
+      "epoch": 1.105802598660318,
+      "grad_norm": 0.6726621985435486,
+      "learning_rate": 4.5514031706492986e-05,
+      "loss": 2.4193,
+      "step": 13702
+    },
+    {
+      "epoch": 1.105883302396901,
+      "grad_norm": 0.7345453500747681,
+      "learning_rate": 4.550079464176963e-05,
+      "loss": 2.4257,
+      "step": 13703
+    },
+    {
+      "epoch": 1.105964006133484,
+      "grad_norm": 0.6764804124832153,
+      "learning_rate": 4.548755893530894e-05,
+      "loss": 2.4656,
+      "step": 13704
+    },
+    {
+      "epoch": 1.106044709870067,
+      "grad_norm": 0.6915058493614197,
+      "learning_rate": 4.5474324587440766e-05,
+      "loss": 2.4148,
+      "step": 13705
+    },
+    {
+      "epoch": 1.10612541360665,
+      "grad_norm": 0.7960236668586731,
+      "learning_rate": 4.5461091598494954e-05,
+      "loss": 2.4148,
+      "step": 13706
+    },
+    {
+      "epoch": 1.106206117343233,
+      "grad_norm": 0.7058970928192139,
+      "learning_rate": 4.544785996880131e-05,
+      "loss": 2.4795,
+      "step": 13707
+    },
+    {
+      "epoch": 1.106286821079816,
+      "grad_norm": 0.6979549527168274,
+      "learning_rate": 4.5434629698689634e-05,
+      "loss": 2.4329,
+      "step": 13708
+    },
+    {
+      "epoch": 1.1063675248163989,
+      "grad_norm": 0.6805241107940674,
+      "learning_rate": 4.5421400788489586e-05,
+      "loss": 2.4303,
+      "step": 13709
+    },
+    {
+      "epoch": 1.106448228552982,
+      "grad_norm": 0.7566354274749756,
+      "learning_rate": 4.5408173238530905e-05,
+      "loss": 2.4769,
+      "step": 13710
+    },
+    {
+      "epoch": 1.106528932289565,
+      "grad_norm": 0.647773802280426,
+      "learning_rate": 4.539494704914324e-05,
+      "loss": 2.4037,
+      "step": 13711
+    },
+    {
+      "epoch": 1.106609636026148,
+      "grad_norm": 0.7248135209083557,
+      "learning_rate": 4.538172222065628e-05,
+      "loss": 2.4366,
+      "step": 13712
+    },
+    {
+      "epoch": 1.106690339762731,
+      "grad_norm": 0.6861057281494141,
+      "learning_rate": 4.536849875339953e-05,
+      "loss": 2.456,
+      "step": 13713
+    },
+    {
+      "epoch": 1.106771043499314,
+      "grad_norm": 0.7386166453361511,
+      "learning_rate": 4.5355276647702605e-05,
+      "loss": 2.4806,
+      "step": 13714
+    },
+    {
+      "epoch": 1.106851747235897,
+      "grad_norm": 0.664402961730957,
+      "learning_rate": 4.534205590389503e-05,
+      "loss": 2.4846,
+      "step": 13715
+    },
+    {
+      "epoch": 1.10693245097248,
+      "grad_norm": 0.8123969435691833,
+      "learning_rate": 4.5328836522306296e-05,
+      "loss": 2.4945,
+      "step": 13716
+    },
+    {
+      "epoch": 1.1070131547090631,
+      "grad_norm": 0.7375624775886536,
+      "learning_rate": 4.5315618503265865e-05,
+      "loss": 2.4533,
+      "step": 13717
+    },
+    {
+      "epoch": 1.107093858445646,
+      "grad_norm": 0.70960932970047,
+      "learning_rate": 4.53024018471032e-05,
+      "loss": 2.4351,
+      "step": 13718
+    },
+    {
+      "epoch": 1.107174562182229,
+      "grad_norm": 0.7170885801315308,
+      "learning_rate": 4.5289186554147645e-05,
+      "loss": 2.4654,
+      "step": 13719
+    },
+    {
+      "epoch": 1.107255265918812,
+      "grad_norm": 0.6986895203590393,
+      "learning_rate": 4.5275972624728556e-05,
+      "loss": 2.4079,
+      "step": 13720
+    },
+    {
+      "epoch": 1.107335969655395,
+      "grad_norm": 0.6948813796043396,
+      "learning_rate": 4.526276005917532e-05,
+      "loss": 2.4981,
+      "step": 13721
+    },
+    {
+      "epoch": 1.1074166733919781,
+      "grad_norm": 0.7719457149505615,
+      "learning_rate": 4.524954885781717e-05,
+      "loss": 2.4853,
+      "step": 13722
+    },
+    {
+      "epoch": 1.107497377128561,
+      "grad_norm": 0.652686357498169,
+      "learning_rate": 4.5236339020983363e-05,
+      "loss": 2.3672,
+      "step": 13723
+    },
+    {
+      "epoch": 1.107578080865144,
+      "grad_norm": 0.7517427802085876,
+      "learning_rate": 4.5223130549003144e-05,
+      "loss": 2.3947,
+      "step": 13724
+    },
+    {
+      "epoch": 1.107658784601727,
+      "grad_norm": 0.6755498647689819,
+      "learning_rate": 4.5209923442205705e-05,
+      "loss": 2.4173,
+      "step": 13725
+    },
+    {
+      "epoch": 1.10773948833831,
+      "grad_norm": 0.6801806688308716,
+      "learning_rate": 4.519671770092019e-05,
+      "loss": 2.4366,
+      "step": 13726
+    },
+    {
+      "epoch": 1.1078201920748931,
+      "grad_norm": 0.6665045619010925,
+      "learning_rate": 4.5183513325475724e-05,
+      "loss": 2.4797,
+      "step": 13727
+    },
+    {
+      "epoch": 1.107900895811476,
+      "grad_norm": 0.7303451299667358,
+      "learning_rate": 4.517031031620145e-05,
+      "loss": 2.4487,
+      "step": 13728
+    },
+    {
+      "epoch": 1.107981599548059,
+      "grad_norm": 0.7241206765174866,
+      "learning_rate": 4.515710867342632e-05,
+      "loss": 2.4632,
+      "step": 13729
+    },
+    {
+      "epoch": 1.1080623032846422,
+      "grad_norm": 0.738835334777832,
+      "learning_rate": 4.514390839747941e-05,
+      "loss": 2.3937,
+      "step": 13730
+    },
+    {
+      "epoch": 1.108143007021225,
+      "grad_norm": 0.7062843441963196,
+      "learning_rate": 4.5130709488689726e-05,
+      "loss": 2.4576,
+      "step": 13731
+    },
+    {
+      "epoch": 1.1082237107578081,
+      "grad_norm": 0.7074100971221924,
+      "learning_rate": 4.511751194738616e-05,
+      "loss": 2.4843,
+      "step": 13732
+    },
+    {
+      "epoch": 1.108304414494391,
+      "grad_norm": 0.751742959022522,
+      "learning_rate": 4.510431577389765e-05,
+      "loss": 2.4607,
+      "step": 13733
+    },
+    {
+      "epoch": 1.108385118230974,
+      "grad_norm": 0.7370054125785828,
+      "learning_rate": 4.50911209685531e-05,
+      "loss": 2.4877,
+      "step": 13734
+    },
+    {
+      "epoch": 1.1084658219675572,
+      "grad_norm": 0.6410251259803772,
+      "learning_rate": 4.507792753168135e-05,
+      "loss": 2.4254,
+      "step": 13735
+    },
+    {
+      "epoch": 1.10854652570414,
+      "grad_norm": 0.7141317129135132,
+      "learning_rate": 4.506473546361121e-05,
+      "loss": 2.4962,
+      "step": 13736
+    },
+    {
+      "epoch": 1.1086272294407231,
+      "grad_norm": 0.6903412342071533,
+      "learning_rate": 4.50515447646715e-05,
+      "loss": 2.4315,
+      "step": 13737
+    },
+    {
+      "epoch": 1.1087079331773062,
+      "grad_norm": 0.7068564891815186,
+      "learning_rate": 4.50383554351909e-05,
+      "loss": 2.5795,
+      "step": 13738
+    },
+    {
+      "epoch": 1.108788636913889,
+      "grad_norm": 0.6880627274513245,
+      "learning_rate": 4.5025167475498154e-05,
+      "loss": 2.4399,
+      "step": 13739
+    },
+    {
+      "epoch": 1.1088693406504722,
+      "grad_norm": 0.6721192598342896,
+      "learning_rate": 4.5011980885921965e-05,
+      "loss": 2.4651,
+      "step": 13740
+    },
+    {
+      "epoch": 1.108950044387055,
+      "grad_norm": 0.7084259986877441,
+      "learning_rate": 4.499879566679093e-05,
+      "loss": 2.4121,
+      "step": 13741
+    },
+    {
+      "epoch": 1.109030748123638,
+      "grad_norm": 0.6809335947036743,
+      "learning_rate": 4.498561181843368e-05,
+      "loss": 2.4714,
+      "step": 13742
+    },
+    {
+      "epoch": 1.1091114518602212,
+      "grad_norm": 0.690416693687439,
+      "learning_rate": 4.497242934117879e-05,
+      "loss": 2.4744,
+      "step": 13743
+    },
+    {
+      "epoch": 1.109192155596804,
+      "grad_norm": 0.728522002696991,
+      "learning_rate": 4.495924823535483e-05,
+      "loss": 2.4374,
+      "step": 13744
+    },
+    {
+      "epoch": 1.1092728593333872,
+      "grad_norm": 0.7000796794891357,
+      "learning_rate": 4.494606850129026e-05,
+      "loss": 2.4635,
+      "step": 13745
+    },
+    {
+      "epoch": 1.1093535630699702,
+      "grad_norm": 0.824645459651947,
+      "learning_rate": 4.493289013931353e-05,
+      "loss": 2.3724,
+      "step": 13746
+    },
+    {
+      "epoch": 1.109434266806553,
+      "grad_norm": 0.6561198830604553,
+      "learning_rate": 4.491971314975321e-05,
+      "loss": 2.3726,
+      "step": 13747
+    },
+    {
+      "epoch": 1.1095149705431362,
+      "grad_norm": 0.7067599892616272,
+      "learning_rate": 4.490653753293757e-05,
+      "loss": 2.4285,
+      "step": 13748
+    },
+    {
+      "epoch": 1.109595674279719,
+      "grad_norm": 0.6954898834228516,
+      "learning_rate": 4.489336328919503e-05,
+      "loss": 2.4252,
+      "step": 13749
+    },
+    {
+      "epoch": 1.1096763780163021,
+      "grad_norm": 0.6683667302131653,
+      "learning_rate": 4.4880190418853974e-05,
+      "loss": 2.4815,
+      "step": 13750
+    },
+    {
+      "epoch": 1.1097570817528852,
+      "grad_norm": 0.7554971575737,
+      "learning_rate": 4.486701892224261e-05,
+      "loss": 2.5036,
+      "step": 13751
+    },
+    {
+      "epoch": 1.109837785489468,
+      "grad_norm": 0.7043242454528809,
+      "learning_rate": 4.485384879968926e-05,
+      "loss": 2.3757,
+      "step": 13752
+    },
+    {
+      "epoch": 1.1099184892260512,
+      "grad_norm": 0.8016893863677979,
+      "learning_rate": 4.4840680051522186e-05,
+      "loss": 2.4655,
+      "step": 13753
+    },
+    {
+      "epoch": 1.1099991929626343,
+      "grad_norm": 0.7022131085395813,
+      "learning_rate": 4.4827512678069515e-05,
+      "loss": 2.475,
+      "step": 13754
+    },
+    {
+      "epoch": 1.1100798966992171,
+      "grad_norm": 0.6963247656822205,
+      "learning_rate": 4.4814346679659455e-05,
+      "loss": 2.4866,
+      "step": 13755
+    },
+    {
+      "epoch": 1.1101606004358002,
+      "grad_norm": 0.6980907917022705,
+      "learning_rate": 4.4801182056620125e-05,
+      "loss": 2.4322,
+      "step": 13756
+    },
+    {
+      "epoch": 1.110241304172383,
+      "grad_norm": 0.68063884973526,
+      "learning_rate": 4.478801880927964e-05,
+      "loss": 2.426,
+      "step": 13757
+    },
+    {
+      "epoch": 1.1103220079089662,
+      "grad_norm": 0.7454195618629456,
+      "learning_rate": 4.477485693796605e-05,
+      "loss": 2.5042,
+      "step": 13758
+    },
+    {
+      "epoch": 1.1104027116455493,
+      "grad_norm": 0.685975193977356,
+      "learning_rate": 4.476169644300737e-05,
+      "loss": 2.4874,
+      "step": 13759
+    },
+    {
+      "epoch": 1.1104834153821321,
+      "grad_norm": 0.7060961723327637,
+      "learning_rate": 4.4748537324731664e-05,
+      "loss": 2.4126,
+      "step": 13760
+    },
+    {
+      "epoch": 1.1105641191187152,
+      "grad_norm": 0.6794416904449463,
+      "learning_rate": 4.4735379583466795e-05,
+      "loss": 2.4112,
+      "step": 13761
+    },
+    {
+      "epoch": 1.1106448228552983,
+      "grad_norm": 0.6854961514472961,
+      "learning_rate": 4.472222321954073e-05,
+      "loss": 2.4909,
+      "step": 13762
+    },
+    {
+      "epoch": 1.1107255265918812,
+      "grad_norm": 0.7660776972770691,
+      "learning_rate": 4.470906823328139e-05,
+      "loss": 2.5021,
+      "step": 13763
+    },
+    {
+      "epoch": 1.1108062303284643,
+      "grad_norm": 0.7027743458747864,
+      "learning_rate": 4.4695914625016564e-05,
+      "loss": 2.4375,
+      "step": 13764
+    },
+    {
+      "epoch": 1.1108869340650471,
+      "grad_norm": 0.6896719336509705,
+      "learning_rate": 4.468276239507413e-05,
+      "loss": 2.4574,
+      "step": 13765
+    },
+    {
+      "epoch": 1.1109676378016302,
+      "grad_norm": 0.685141384601593,
+      "learning_rate": 4.4669611543781844e-05,
+      "loss": 2.4311,
+      "step": 13766
+    },
+    {
+      "epoch": 1.1110483415382133,
+      "grad_norm": 0.7108263373374939,
+      "learning_rate": 4.465646207146746e-05,
+      "loss": 2.4565,
+      "step": 13767
+    },
+    {
+      "epoch": 1.1111290452747962,
+      "grad_norm": 0.63578861951828,
+      "learning_rate": 4.464331397845873e-05,
+      "loss": 2.449,
+      "step": 13768
+    },
+    {
+      "epoch": 1.1112097490113793,
+      "grad_norm": 0.6917306780815125,
+      "learning_rate": 4.463016726508335e-05,
+      "loss": 2.4681,
+      "step": 13769
+    },
+    {
+      "epoch": 1.1112904527479621,
+      "grad_norm": 0.7328054308891296,
+      "learning_rate": 4.4617021931668914e-05,
+      "loss": 2.404,
+      "step": 13770
+    },
+    {
+      "epoch": 1.1113711564845452,
+      "grad_norm": 0.6501660943031311,
+      "learning_rate": 4.460387797854305e-05,
+      "loss": 2.4228,
+      "step": 13771
+    },
+    {
+      "epoch": 1.1114518602211283,
+      "grad_norm": 0.6656771302223206,
+      "learning_rate": 4.459073540603336e-05,
+      "loss": 2.4814,
+      "step": 13772
+    },
+    {
+      "epoch": 1.1115325639577112,
+      "grad_norm": 0.671017587184906,
+      "learning_rate": 4.457759421446742e-05,
+      "loss": 2.4605,
+      "step": 13773
+    },
+    {
+      "epoch": 1.1116132676942942,
+      "grad_norm": 0.6715343594551086,
+      "learning_rate": 4.456445440417267e-05,
+      "loss": 2.424,
+      "step": 13774
+    },
+    {
+      "epoch": 1.1116939714308773,
+      "grad_norm": 0.7051515579223633,
+      "learning_rate": 4.4551315975476626e-05,
+      "loss": 2.4358,
+      "step": 13775
+    },
+    {
+      "epoch": 1.1117746751674602,
+      "grad_norm": 0.7810437679290771,
+      "learning_rate": 4.453817892870673e-05,
+      "loss": 2.4718,
+      "step": 13776
+    },
+    {
+      "epoch": 1.1118553789040433,
+      "grad_norm": 0.7072561383247375,
+      "learning_rate": 4.4525043264190405e-05,
+      "loss": 2.4429,
+      "step": 13777
+    },
+    {
+      "epoch": 1.1119360826406264,
+      "grad_norm": 0.7949702143669128,
+      "learning_rate": 4.4511908982255e-05,
+      "loss": 2.4413,
+      "step": 13778
+    },
+    {
+      "epoch": 1.1120167863772092,
+      "grad_norm": 0.6716235876083374,
+      "learning_rate": 4.449877608322792e-05,
+      "loss": 2.427,
+      "step": 13779
+    },
+    {
+      "epoch": 1.1120974901137923,
+      "grad_norm": 0.7332563996315002,
+      "learning_rate": 4.448564456743638e-05,
+      "loss": 2.4567,
+      "step": 13780
+    },
+    {
+      "epoch": 1.1121781938503752,
+      "grad_norm": 0.7264607548713684,
+      "learning_rate": 4.447251443520769e-05,
+      "loss": 2.4844,
+      "step": 13781
+    },
+    {
+      "epoch": 1.1122588975869583,
+      "grad_norm": 0.7819967865943909,
+      "learning_rate": 4.4459385686869136e-05,
+      "loss": 2.5129,
+      "step": 13782
+    },
+    {
+      "epoch": 1.1123396013235414,
+      "grad_norm": 0.7587651610374451,
+      "learning_rate": 4.4446258322747824e-05,
+      "loss": 2.4714,
+      "step": 13783
+    },
+    {
+      "epoch": 1.1124203050601242,
+      "grad_norm": 0.6392871141433716,
+      "learning_rate": 4.443313234317099e-05,
+      "loss": 2.462,
+      "step": 13784
+    },
+    {
+      "epoch": 1.1125010087967073,
+      "grad_norm": 0.6609585881233215,
+      "learning_rate": 4.442000774846574e-05,
+      "loss": 2.4566,
+      "step": 13785
+    },
+    {
+      "epoch": 1.1125817125332902,
+      "grad_norm": 0.762924075126648,
+      "learning_rate": 4.440688453895919e-05,
+      "loss": 2.4613,
+      "step": 13786
+    },
+    {
+      "epoch": 1.1126624162698733,
+      "grad_norm": 0.7096089124679565,
+      "learning_rate": 4.4393762714978394e-05,
+      "loss": 2.4195,
+      "step": 13787
+    },
+    {
+      "epoch": 1.1127431200064564,
+      "grad_norm": 0.6663284301757812,
+      "learning_rate": 4.438064227685039e-05,
+      "loss": 2.422,
+      "step": 13788
+    },
+    {
+      "epoch": 1.1128238237430392,
+      "grad_norm": 0.6653628945350647,
+      "learning_rate": 4.436752322490221e-05,
+      "loss": 2.4477,
+      "step": 13789
+    },
+    {
+      "epoch": 1.1129045274796223,
+      "grad_norm": 0.6527605056762695,
+      "learning_rate": 4.435440555946073e-05,
+      "loss": 2.3874,
+      "step": 13790
+    },
+    {
+      "epoch": 1.1129852312162054,
+      "grad_norm": 0.6801275014877319,
+      "learning_rate": 4.4341289280852935e-05,
+      "loss": 2.4474,
+      "step": 13791
+    },
+    {
+      "epoch": 1.1130659349527883,
+      "grad_norm": 0.729905366897583,
+      "learning_rate": 4.432817438940574e-05,
+      "loss": 2.4711,
+      "step": 13792
+    },
+    {
+      "epoch": 1.1131466386893714,
+      "grad_norm": 0.7074751853942871,
+      "learning_rate": 4.431506088544593e-05,
+      "loss": 2.451,
+      "step": 13793
+    },
+    {
+      "epoch": 1.1132273424259542,
+      "grad_norm": 0.7241154313087463,
+      "learning_rate": 4.430194876930035e-05,
+      "loss": 2.4883,
+      "step": 13794
+    },
+    {
+      "epoch": 1.1133080461625373,
+      "grad_norm": 0.6549142003059387,
+      "learning_rate": 4.428883804129586e-05,
+      "loss": 2.4243,
+      "step": 13795
+    },
+    {
+      "epoch": 1.1133887498991204,
+      "grad_norm": 0.7046780586242676,
+      "learning_rate": 4.427572870175907e-05,
+      "loss": 2.4143,
+      "step": 13796
+    },
+    {
+      "epoch": 1.1134694536357033,
+      "grad_norm": 0.6563952565193176,
+      "learning_rate": 4.426262075101682e-05,
+      "loss": 2.416,
+      "step": 13797
+    },
+    {
+      "epoch": 1.1135501573722864,
+      "grad_norm": 0.7002081871032715,
+      "learning_rate": 4.4249514189395803e-05,
+      "loss": 2.3673,
+      "step": 13798
+    },
+    {
+      "epoch": 1.1136308611088694,
+      "grad_norm": 0.6766571998596191,
+      "learning_rate": 4.423640901722259e-05,
+      "loss": 2.4941,
+      "step": 13799
+    },
+    {
+      "epoch": 1.1137115648454523,
+      "grad_norm": 0.7404381632804871,
+      "learning_rate": 4.422330523482383e-05,
+      "loss": 2.4794,
+      "step": 13800
+    },
+    {
+      "epoch": 1.1137922685820354,
+      "grad_norm": 0.6670998930931091,
+      "learning_rate": 4.421020284252614e-05,
+      "loss": 2.5131,
+      "step": 13801
+    },
+    {
+      "epoch": 1.1138729723186183,
+      "grad_norm": 0.803720235824585,
+      "learning_rate": 4.4197101840655995e-05,
+      "loss": 2.4751,
+      "step": 13802
+    },
+    {
+      "epoch": 1.1139536760552013,
+      "grad_norm": 0.6532074809074402,
+      "learning_rate": 4.4184002229539947e-05,
+      "loss": 2.4147,
+      "step": 13803
+    },
+    {
+      "epoch": 1.1140343797917844,
+      "grad_norm": 0.6548035144805908,
+      "learning_rate": 4.417090400950447e-05,
+      "loss": 2.4601,
+      "step": 13804
+    },
+    {
+      "epoch": 1.1141150835283673,
+      "grad_norm": 0.6971763968467712,
+      "learning_rate": 4.415780718087603e-05,
+      "loss": 2.4752,
+      "step": 13805
+    },
+    {
+      "epoch": 1.1141957872649504,
+      "grad_norm": 0.6624024510383606,
+      "learning_rate": 4.414471174398098e-05,
+      "loss": 2.4183,
+      "step": 13806
+    },
+    {
+      "epoch": 1.1142764910015335,
+      "grad_norm": 0.6571507453918457,
+      "learning_rate": 4.4131617699145714e-05,
+      "loss": 2.4747,
+      "step": 13807
+    },
+    {
+      "epoch": 1.1143571947381163,
+      "grad_norm": 0.7165808081626892,
+      "learning_rate": 4.411852504669658e-05,
+      "loss": 2.453,
+      "step": 13808
+    },
+    {
+      "epoch": 1.1144378984746994,
+      "grad_norm": 0.6708057522773743,
+      "learning_rate": 4.410543378695988e-05,
+      "loss": 2.4858,
+      "step": 13809
+    },
+    {
+      "epoch": 1.1145186022112823,
+      "grad_norm": 0.889302134513855,
+      "learning_rate": 4.409234392026187e-05,
+      "loss": 2.4333,
+      "step": 13810
+    },
+    {
+      "epoch": 1.1145993059478654,
+      "grad_norm": 0.7440677881240845,
+      "learning_rate": 4.407925544692884e-05,
+      "loss": 2.49,
+      "step": 13811
+    },
+    {
+      "epoch": 1.1146800096844485,
+      "grad_norm": 0.6688372492790222,
+      "learning_rate": 4.406616836728691e-05,
+      "loss": 2.4663,
+      "step": 13812
+    },
+    {
+      "epoch": 1.1147607134210313,
+      "grad_norm": 0.7108204364776611,
+      "learning_rate": 4.4053082681662264e-05,
+      "loss": 2.4843,
+      "step": 13813
+    },
+    {
+      "epoch": 1.1148414171576144,
+      "grad_norm": 0.7270475029945374,
+      "learning_rate": 4.4039998390381087e-05,
+      "loss": 2.4158,
+      "step": 13814
+    },
+    {
+      "epoch": 1.1149221208941973,
+      "grad_norm": 0.7243396639823914,
+      "learning_rate": 4.402691549376939e-05,
+      "loss": 2.3969,
+      "step": 13815
+    },
+    {
+      "epoch": 1.1150028246307804,
+      "grad_norm": 0.6687803268432617,
+      "learning_rate": 4.4013833992153285e-05,
+      "loss": 2.42,
+      "step": 13816
+    },
+    {
+      "epoch": 1.1150835283673635,
+      "grad_norm": 0.6892626285552979,
+      "learning_rate": 4.400075388585877e-05,
+      "loss": 2.4086,
+      "step": 13817
+    },
+    {
+      "epoch": 1.1151642321039463,
+      "grad_norm": 0.7556231021881104,
+      "learning_rate": 4.398767517521186e-05,
+      "loss": 2.4201,
+      "step": 13818
+    },
+    {
+      "epoch": 1.1152449358405294,
+      "grad_norm": 0.6872838735580444,
+      "learning_rate": 4.397459786053851e-05,
+      "loss": 2.4143,
+      "step": 13819
+    },
+    {
+      "epoch": 1.1153256395771125,
+      "grad_norm": 0.6681817770004272,
+      "learning_rate": 4.396152194216463e-05,
+      "loss": 2.4404,
+      "step": 13820
+    },
+    {
+      "epoch": 1.1154063433136954,
+      "grad_norm": 0.7107201218605042,
+      "learning_rate": 4.394844742041614e-05,
+      "loss": 2.4503,
+      "step": 13821
+    },
+    {
+      "epoch": 1.1154870470502785,
+      "grad_norm": 0.706541121006012,
+      "learning_rate": 4.3935374295618824e-05,
+      "loss": 2.5106,
+      "step": 13822
+    },
+    {
+      "epoch": 1.1155677507868615,
+      "grad_norm": 0.6659905910491943,
+      "learning_rate": 4.392230256809854e-05,
+      "loss": 2.3839,
+      "step": 13823
+    },
+    {
+      "epoch": 1.1156484545234444,
+      "grad_norm": 0.7125810980796814,
+      "learning_rate": 4.3909232238181095e-05,
+      "loss": 2.4463,
+      "step": 13824
+    },
+    {
+      "epoch": 1.1157291582600275,
+      "grad_norm": 0.6581901907920837,
+      "learning_rate": 4.389616330619217e-05,
+      "loss": 2.4004,
+      "step": 13825
+    },
+    {
+      "epoch": 1.1158098619966104,
+      "grad_norm": 0.7660872340202332,
+      "learning_rate": 4.388309577245752e-05,
+      "loss": 2.4685,
+      "step": 13826
+    },
+    {
+      "epoch": 1.1158905657331935,
+      "grad_norm": 0.699526846408844,
+      "learning_rate": 4.387002963730281e-05,
+      "loss": 2.4131,
+      "step": 13827
+    },
+    {
+      "epoch": 1.1159712694697765,
+      "grad_norm": 0.7031015753746033,
+      "learning_rate": 4.3856964901053685e-05,
+      "loss": 2.4476,
+      "step": 13828
+    },
+    {
+      "epoch": 1.1160519732063594,
+      "grad_norm": 0.6876828074455261,
+      "learning_rate": 4.384390156403575e-05,
+      "loss": 2.4402,
+      "step": 13829
+    },
+    {
+      "epoch": 1.1161326769429425,
+      "grad_norm": 0.7188935279846191,
+      "learning_rate": 4.3830839626574626e-05,
+      "loss": 2.4473,
+      "step": 13830
+    },
+    {
+      "epoch": 1.1162133806795254,
+      "grad_norm": 0.6825287938117981,
+      "learning_rate": 4.381777908899577e-05,
+      "loss": 2.4757,
+      "step": 13831
+    },
+    {
+      "epoch": 1.1162940844161084,
+      "grad_norm": 0.718267560005188,
+      "learning_rate": 4.380471995162472e-05,
+      "loss": 2.483,
+      "step": 13832
+    },
+    {
+      "epoch": 1.1163747881526915,
+      "grad_norm": 0.6526767611503601,
+      "learning_rate": 4.379166221478697e-05,
+      "loss": 2.4161,
+      "step": 13833
+    },
+    {
+      "epoch": 1.1164554918892744,
+      "grad_norm": 0.7541480660438538,
+      "learning_rate": 4.37786058788079e-05,
+      "loss": 2.4876,
+      "step": 13834
+    },
+    {
+      "epoch": 1.1165361956258575,
+      "grad_norm": 0.7144232988357544,
+      "learning_rate": 4.376555094401294e-05,
+      "loss": 2.4153,
+      "step": 13835
+    },
+    {
+      "epoch": 1.1166168993624406,
+      "grad_norm": 0.7544882297515869,
+      "learning_rate": 4.3752497410727445e-05,
+      "loss": 2.4634,
+      "step": 13836
+    },
+    {
+      "epoch": 1.1166976030990234,
+      "grad_norm": 0.7263267040252686,
+      "learning_rate": 4.373944527927674e-05,
+      "loss": 2.5189,
+      "step": 13837
+    },
+    {
+      "epoch": 1.1167783068356065,
+      "grad_norm": 0.7709252834320068,
+      "learning_rate": 4.3726394549986135e-05,
+      "loss": 2.5036,
+      "step": 13838
+    },
+    {
+      "epoch": 1.1168590105721894,
+      "grad_norm": 0.6849128007888794,
+      "learning_rate": 4.3713345223180866e-05,
+      "loss": 2.414,
+      "step": 13839
+    },
+    {
+      "epoch": 1.1169397143087725,
+      "grad_norm": 0.6807512044906616,
+      "learning_rate": 4.3700297299186224e-05,
+      "loss": 2.4924,
+      "step": 13840
+    },
+    {
+      "epoch": 1.1170204180453556,
+      "grad_norm": 0.6894977688789368,
+      "learning_rate": 4.3687250778327294e-05,
+      "loss": 2.4183,
+      "step": 13841
+    },
+    {
+      "epoch": 1.1171011217819384,
+      "grad_norm": 0.6657617092132568,
+      "learning_rate": 4.367420566092928e-05,
+      "loss": 2.448,
+      "step": 13842
+    },
+    {
+      "epoch": 1.1171818255185215,
+      "grad_norm": 0.7104446291923523,
+      "learning_rate": 4.366116194731733e-05,
+      "loss": 2.4862,
+      "step": 13843
+    },
+    {
+      "epoch": 1.1172625292551046,
+      "grad_norm": 0.7485257387161255,
+      "learning_rate": 4.3648119637816465e-05,
+      "loss": 2.4253,
+      "step": 13844
+    },
+    {
+      "epoch": 1.1173432329916875,
+      "grad_norm": 0.7079899907112122,
+      "learning_rate": 4.363507873275177e-05,
+      "loss": 2.4235,
+      "step": 13845
+    },
+    {
+      "epoch": 1.1174239367282706,
+      "grad_norm": 0.6891573667526245,
+      "learning_rate": 4.3622039232448274e-05,
+      "loss": 2.4382,
+      "step": 13846
+    },
+    {
+      "epoch": 1.1175046404648534,
+      "grad_norm": 0.6886103749275208,
+      "learning_rate": 4.360900113723086e-05,
+      "loss": 2.5115,
+      "step": 13847
+    },
+    {
+      "epoch": 1.1175853442014365,
+      "grad_norm": 0.7511457800865173,
+      "learning_rate": 4.35959644474246e-05,
+      "loss": 2.4071,
+      "step": 13848
+    },
+    {
+      "epoch": 1.1176660479380196,
+      "grad_norm": 0.6526182293891907,
+      "learning_rate": 4.358292916335437e-05,
+      "loss": 2.4242,
+      "step": 13849
+    },
+    {
+      "epoch": 1.1177467516746025,
+      "grad_norm": 0.7385138273239136,
+      "learning_rate": 4.356989528534499e-05,
+      "loss": 2.4459,
+      "step": 13850
+    },
+    {
+      "epoch": 1.1178274554111856,
+      "grad_norm": 0.6668610572814941,
+      "learning_rate": 4.355686281372132e-05,
+      "loss": 2.4188,
+      "step": 13851
+    },
+    {
+      "epoch": 1.1179081591477686,
+      "grad_norm": 0.6950691342353821,
+      "learning_rate": 4.354383174880818e-05,
+      "loss": 2.4339,
+      "step": 13852
+    },
+    {
+      "epoch": 1.1179888628843515,
+      "grad_norm": 0.7017496824264526,
+      "learning_rate": 4.3530802090930375e-05,
+      "loss": 2.4733,
+      "step": 13853
+    },
+    {
+      "epoch": 1.1180695666209346,
+      "grad_norm": 0.8118221759796143,
+      "learning_rate": 4.351777384041254e-05,
+      "loss": 2.4826,
+      "step": 13854
+    },
+    {
+      "epoch": 1.1181502703575175,
+      "grad_norm": 0.7233164310455322,
+      "learning_rate": 4.350474699757945e-05,
+      "loss": 2.4637,
+      "step": 13855
+    },
+    {
+      "epoch": 1.1182309740941005,
+      "grad_norm": 0.6354575157165527,
+      "learning_rate": 4.349172156275576e-05,
+      "loss": 2.4487,
+      "step": 13856
+    },
+    {
+      "epoch": 1.1183116778306836,
+      "grad_norm": 0.6776937246322632,
+      "learning_rate": 4.347869753626606e-05,
+      "loss": 2.4292,
+      "step": 13857
+    },
+    {
+      "epoch": 1.1183923815672665,
+      "grad_norm": 0.6656864881515503,
+      "learning_rate": 4.3465674918434953e-05,
+      "loss": 2.484,
+      "step": 13858
+    },
+    {
+      "epoch": 1.1184730853038496,
+      "grad_norm": 0.7659650444984436,
+      "learning_rate": 4.345265370958702e-05,
+      "loss": 2.4181,
+      "step": 13859
+    },
+    {
+      "epoch": 1.1185537890404325,
+      "grad_norm": 0.6546063423156738,
+      "learning_rate": 4.3439633910046764e-05,
+      "loss": 2.4657,
+      "step": 13860
+    },
+    {
+      "epoch": 1.1186344927770155,
+      "grad_norm": 0.6869762539863586,
+      "learning_rate": 4.342661552013869e-05,
+      "loss": 2.513,
+      "step": 13861
+    },
+    {
+      "epoch": 1.1187151965135986,
+      "grad_norm": 0.6633490324020386,
+      "learning_rate": 4.3413598540187275e-05,
+      "loss": 2.4716,
+      "step": 13862
+    },
+    {
+      "epoch": 1.1187959002501815,
+      "grad_norm": 0.7238267660140991,
+      "learning_rate": 4.340058297051687e-05,
+      "loss": 2.4353,
+      "step": 13863
+    },
+    {
+      "epoch": 1.1188766039867646,
+      "grad_norm": 0.67429119348526,
+      "learning_rate": 4.3387568811451875e-05,
+      "loss": 2.4808,
+      "step": 13864
+    },
+    {
+      "epoch": 1.1189573077233477,
+      "grad_norm": 0.6901153326034546,
+      "learning_rate": 4.33745560633167e-05,
+      "loss": 2.4785,
+      "step": 13865
+    },
+    {
+      "epoch": 1.1190380114599305,
+      "grad_norm": 0.7227689027786255,
+      "learning_rate": 4.336154472643556e-05,
+      "loss": 2.4414,
+      "step": 13866
+    },
+    {
+      "epoch": 1.1191187151965136,
+      "grad_norm": 0.713793933391571,
+      "learning_rate": 4.33485348011328e-05,
+      "loss": 2.5136,
+      "step": 13867
+    },
+    {
+      "epoch": 1.1191994189330967,
+      "grad_norm": 0.6495655179023743,
+      "learning_rate": 4.333552628773263e-05,
+      "loss": 2.4267,
+      "step": 13868
+    },
+    {
+      "epoch": 1.1192801226696796,
+      "grad_norm": 0.7265790104866028,
+      "learning_rate": 4.3322519186559274e-05,
+      "loss": 2.4406,
+      "step": 13869
+    },
+    {
+      "epoch": 1.1193608264062627,
+      "grad_norm": 0.6700571179389954,
+      "learning_rate": 4.330951349793688e-05,
+      "loss": 2.4457,
+      "step": 13870
+    },
+    {
+      "epoch": 1.1194415301428455,
+      "grad_norm": 0.7112334966659546,
+      "learning_rate": 4.3296509222189616e-05,
+      "loss": 2.4788,
+      "step": 13871
+    },
+    {
+      "epoch": 1.1195222338794286,
+      "grad_norm": 0.7056662440299988,
+      "learning_rate": 4.32835063596416e-05,
+      "loss": 2.5195,
+      "step": 13872
+    },
+    {
+      "epoch": 1.1196029376160117,
+      "grad_norm": 0.7198836207389832,
+      "learning_rate": 4.327050491061683e-05,
+      "loss": 2.4827,
+      "step": 13873
+    },
+    {
+      "epoch": 1.1196836413525946,
+      "grad_norm": 0.7384079694747925,
+      "learning_rate": 4.325750487543936e-05,
+      "loss": 2.4556,
+      "step": 13874
+    },
+    {
+      "epoch": 1.1197643450891777,
+      "grad_norm": 0.7315430641174316,
+      "learning_rate": 4.324450625443324e-05,
+      "loss": 2.4302,
+      "step": 13875
+    },
+    {
+      "epoch": 1.1198450488257605,
+      "grad_norm": 0.6692587733268738,
+      "learning_rate": 4.323150904792234e-05,
+      "loss": 2.5283,
+      "step": 13876
+    },
+    {
+      "epoch": 1.1199257525623436,
+      "grad_norm": 0.7407168745994568,
+      "learning_rate": 4.321851325623063e-05,
+      "loss": 2.4757,
+      "step": 13877
+    },
+    {
+      "epoch": 1.1200064562989267,
+      "grad_norm": 0.7387246489524841,
+      "learning_rate": 4.3205518879682e-05,
+      "loss": 2.5025,
+      "step": 13878
+    },
+    {
+      "epoch": 1.1200871600355096,
+      "grad_norm": 0.8058405518531799,
+      "learning_rate": 4.319252591860031e-05,
+      "loss": 2.4951,
+      "step": 13879
+    },
+    {
+      "epoch": 1.1201678637720927,
+      "grad_norm": 0.6964818835258484,
+      "learning_rate": 4.317953437330936e-05,
+      "loss": 2.4462,
+      "step": 13880
+    },
+    {
+      "epoch": 1.1202485675086757,
+      "grad_norm": 0.6904557347297668,
+      "learning_rate": 4.316654424413294e-05,
+      "loss": 2.3981,
+      "step": 13881
+    },
+    {
+      "epoch": 1.1203292712452586,
+      "grad_norm": 0.6555196046829224,
+      "learning_rate": 4.315355553139485e-05,
+      "loss": 2.418,
+      "step": 13882
+    },
+    {
+      "epoch": 1.1204099749818417,
+      "grad_norm": 0.7745094299316406,
+      "learning_rate": 4.3140568235418724e-05,
+      "loss": 2.4635,
+      "step": 13883
+    },
+    {
+      "epoch": 1.1204906787184246,
+      "grad_norm": 0.686676025390625,
+      "learning_rate": 4.312758235652825e-05,
+      "loss": 2.4847,
+      "step": 13884
+    },
+    {
+      "epoch": 1.1205713824550076,
+      "grad_norm": 0.6937002539634705,
+      "learning_rate": 4.311459789504714e-05,
+      "loss": 2.4632,
+      "step": 13885
+    },
+    {
+      "epoch": 1.1206520861915907,
+      "grad_norm": 0.7024590373039246,
+      "learning_rate": 4.310161485129891e-05,
+      "loss": 2.4268,
+      "step": 13886
+    },
+    {
+      "epoch": 1.1207327899281736,
+      "grad_norm": 0.6848484873771667,
+      "learning_rate": 4.308863322560717e-05,
+      "loss": 2.4895,
+      "step": 13887
+    },
+    {
+      "epoch": 1.1208134936647567,
+      "grad_norm": 0.7071602940559387,
+      "learning_rate": 4.307565301829546e-05,
+      "loss": 2.4348,
+      "step": 13888
+    },
+    {
+      "epoch": 1.1208941974013398,
+      "grad_norm": 0.6868199706077576,
+      "learning_rate": 4.3062674229687274e-05,
+      "loss": 2.4613,
+      "step": 13889
+    },
+    {
+      "epoch": 1.1209749011379226,
+      "grad_norm": 0.7283496260643005,
+      "learning_rate": 4.304969686010608e-05,
+      "loss": 2.478,
+      "step": 13890
+    },
+    {
+      "epoch": 1.1210556048745057,
+      "grad_norm": 0.6907255053520203,
+      "learning_rate": 4.303672090987535e-05,
+      "loss": 2.4431,
+      "step": 13891
+    },
+    {
+      "epoch": 1.1211363086110886,
+      "grad_norm": 0.675089418888092,
+      "learning_rate": 4.302374637931841e-05,
+      "loss": 2.4398,
+      "step": 13892
+    },
+    {
+      "epoch": 1.1212170123476717,
+      "grad_norm": 0.6929863095283508,
+      "learning_rate": 4.301077326875863e-05,
+      "loss": 2.3909,
+      "step": 13893
+    },
+    {
+      "epoch": 1.1212977160842548,
+      "grad_norm": 0.6746132969856262,
+      "learning_rate": 4.29978015785194e-05,
+      "loss": 2.4726,
+      "step": 13894
+    },
+    {
+      "epoch": 1.1213784198208376,
+      "grad_norm": 0.720781147480011,
+      "learning_rate": 4.298483130892392e-05,
+      "loss": 2.4445,
+      "step": 13895
+    },
+    {
+      "epoch": 1.1214591235574207,
+      "grad_norm": 0.6624416708946228,
+      "learning_rate": 4.297186246029549e-05,
+      "loss": 2.3868,
+      "step": 13896
+    },
+    {
+      "epoch": 1.1215398272940038,
+      "grad_norm": 0.7849127054214478,
+      "learning_rate": 4.295889503295731e-05,
+      "loss": 2.4479,
+      "step": 13897
+    },
+    {
+      "epoch": 1.1216205310305867,
+      "grad_norm": 0.6655337810516357,
+      "learning_rate": 4.294592902723259e-05,
+      "loss": 2.5093,
+      "step": 13898
+    },
+    {
+      "epoch": 1.1217012347671698,
+      "grad_norm": 0.7055402398109436,
+      "learning_rate": 4.293296444344445e-05,
+      "loss": 2.4385,
+      "step": 13899
+    },
+    {
+      "epoch": 1.1217819385037526,
+      "grad_norm": 0.7388767600059509,
+      "learning_rate": 4.2920001281916e-05,
+      "loss": 2.4863,
+      "step": 13900
+    },
+    {
+      "epoch": 1.1218626422403357,
+      "grad_norm": 0.6915223002433777,
+      "learning_rate": 4.2907039542970373e-05,
+      "loss": 2.4218,
+      "step": 13901
+    },
+    {
+      "epoch": 1.1219433459769188,
+      "grad_norm": 0.7124893665313721,
+      "learning_rate": 4.289407922693053e-05,
+      "loss": 2.4514,
+      "step": 13902
+    },
+    {
+      "epoch": 1.1220240497135017,
+      "grad_norm": 0.6552406549453735,
+      "learning_rate": 4.28811203341195e-05,
+      "loss": 2.4558,
+      "step": 13903
+    },
+    {
+      "epoch": 1.1221047534500848,
+      "grad_norm": 0.6641791462898254,
+      "learning_rate": 4.286816286486031e-05,
+      "loss": 2.4277,
+      "step": 13904
+    },
+    {
+      "epoch": 1.1221854571866678,
+      "grad_norm": 0.677733838558197,
+      "learning_rate": 4.285520681947579e-05,
+      "loss": 2.4861,
+      "step": 13905
+    },
+    {
+      "epoch": 1.1222661609232507,
+      "grad_norm": 0.6572888493537903,
+      "learning_rate": 4.284225219828891e-05,
+      "loss": 2.4657,
+      "step": 13906
+    },
+    {
+      "epoch": 1.1223468646598338,
+      "grad_norm": 0.6923860907554626,
+      "learning_rate": 4.2829299001622546e-05,
+      "loss": 2.4857,
+      "step": 13907
+    },
+    {
+      "epoch": 1.1224275683964167,
+      "grad_norm": 0.6971977949142456,
+      "learning_rate": 4.281634722979947e-05,
+      "loss": 2.4434,
+      "step": 13908
+    },
+    {
+      "epoch": 1.1225082721329998,
+      "grad_norm": 0.6828060746192932,
+      "learning_rate": 4.2803396883142456e-05,
+      "loss": 2.4342,
+      "step": 13909
+    },
+    {
+      "epoch": 1.1225889758695828,
+      "grad_norm": 0.7001270651817322,
+      "learning_rate": 4.279044796197438e-05,
+      "loss": 2.5222,
+      "step": 13910
+    },
+    {
+      "epoch": 1.1226696796061657,
+      "grad_norm": 0.6425578594207764,
+      "learning_rate": 4.277750046661785e-05,
+      "loss": 2.42,
+      "step": 13911
+    },
+    {
+      "epoch": 1.1227503833427488,
+      "grad_norm": 0.6498209834098816,
+      "learning_rate": 4.2764554397395585e-05,
+      "loss": 2.4448,
+      "step": 13912
+    },
+    {
+      "epoch": 1.1228310870793319,
+      "grad_norm": 0.6894031763076782,
+      "learning_rate": 4.275160975463025e-05,
+      "loss": 2.4508,
+      "step": 13913
+    },
+    {
+      "epoch": 1.1229117908159147,
+      "grad_norm": 0.7286608219146729,
+      "learning_rate": 4.273866653864448e-05,
+      "loss": 2.4557,
+      "step": 13914
+    },
+    {
+      "epoch": 1.1229924945524978,
+      "grad_norm": 0.753826379776001,
+      "learning_rate": 4.272572474976079e-05,
+      "loss": 2.4635,
+      "step": 13915
+    },
+    {
+      "epoch": 1.1230731982890807,
+      "grad_norm": 0.6715937256813049,
+      "learning_rate": 4.271278438830174e-05,
+      "loss": 2.5107,
+      "step": 13916
+    },
+    {
+      "epoch": 1.1231539020256638,
+      "grad_norm": 0.6833200454711914,
+      "learning_rate": 4.26998454545899e-05,
+      "loss": 2.4883,
+      "step": 13917
+    },
+    {
+      "epoch": 1.1232346057622469,
+      "grad_norm": 0.6763597130775452,
+      "learning_rate": 4.2686907948947666e-05,
+      "loss": 2.4178,
+      "step": 13918
+    },
+    {
+      "epoch": 1.1233153094988297,
+      "grad_norm": 0.7336227297782898,
+      "learning_rate": 4.26739718716975e-05,
+      "loss": 2.4542,
+      "step": 13919
+    },
+    {
+      "epoch": 1.1233960132354128,
+      "grad_norm": 0.6583260297775269,
+      "learning_rate": 4.2661037223161806e-05,
+      "loss": 2.3998,
+      "step": 13920
+    },
+    {
+      "epoch": 1.1234767169719957,
+      "grad_norm": 0.6444356441497803,
+      "learning_rate": 4.264810400366295e-05,
+      "loss": 2.4354,
+      "step": 13921
+    },
+    {
+      "epoch": 1.1235574207085788,
+      "grad_norm": 0.6786002516746521,
+      "learning_rate": 4.2635172213523255e-05,
+      "loss": 2.3989,
+      "step": 13922
+    },
+    {
+      "epoch": 1.1236381244451619,
+      "grad_norm": 0.6838372349739075,
+      "learning_rate": 4.262224185306507e-05,
+      "loss": 2.4431,
+      "step": 13923
+    },
+    {
+      "epoch": 1.1237188281817447,
+      "grad_norm": 0.7516793012619019,
+      "learning_rate": 4.260931292261056e-05,
+      "loss": 2.4373,
+      "step": 13924
+    },
+    {
+      "epoch": 1.1237995319183278,
+      "grad_norm": 0.6860260367393494,
+      "learning_rate": 4.2596385422481985e-05,
+      "loss": 2.4457,
+      "step": 13925
+    },
+    {
+      "epoch": 1.123880235654911,
+      "grad_norm": 0.6556448936462402,
+      "learning_rate": 4.2583459353001595e-05,
+      "loss": 2.4165,
+      "step": 13926
+    },
+    {
+      "epoch": 1.1239609393914938,
+      "grad_norm": 0.729131281375885,
+      "learning_rate": 4.257053471449144e-05,
+      "loss": 2.4124,
+      "step": 13927
+    },
+    {
+      "epoch": 1.1240416431280769,
+      "grad_norm": 0.6941910982131958,
+      "learning_rate": 4.2557611507273684e-05,
+      "loss": 2.4095,
+      "step": 13928
+    },
+    {
+      "epoch": 1.12412234686466,
+      "grad_norm": 0.6390536427497864,
+      "learning_rate": 4.25446897316704e-05,
+      "loss": 2.4221,
+      "step": 13929
+    },
+    {
+      "epoch": 1.1242030506012428,
+      "grad_norm": 0.7034881114959717,
+      "learning_rate": 4.253176938800365e-05,
+      "loss": 2.4685,
+      "step": 13930
+    },
+    {
+      "epoch": 1.124283754337826,
+      "grad_norm": 0.6975526809692383,
+      "learning_rate": 4.251885047659542e-05,
+      "loss": 2.4771,
+      "step": 13931
+    },
+    {
+      "epoch": 1.1243644580744088,
+      "grad_norm": 0.7020023465156555,
+      "learning_rate": 4.2505932997767695e-05,
+      "loss": 2.4746,
+      "step": 13932
+    },
+    {
+      "epoch": 1.1244451618109919,
+      "grad_norm": 0.7207093238830566,
+      "learning_rate": 4.2493016951842444e-05,
+      "loss": 2.4707,
+      "step": 13933
+    },
+    {
+      "epoch": 1.124525865547575,
+      "grad_norm": 0.7711251974105835,
+      "learning_rate": 4.24801023391415e-05,
+      "loss": 2.5104,
+      "step": 13934
+    },
+    {
+      "epoch": 1.1246065692841578,
+      "grad_norm": 0.7324040532112122,
+      "learning_rate": 4.246718915998677e-05,
+      "loss": 2.4257,
+      "step": 13935
+    },
+    {
+      "epoch": 1.124687273020741,
+      "grad_norm": 0.6532757878303528,
+      "learning_rate": 4.2454277414700116e-05,
+      "loss": 2.3708,
+      "step": 13936
+    },
+    {
+      "epoch": 1.1247679767573238,
+      "grad_norm": 0.6933012008666992,
+      "learning_rate": 4.244136710360325e-05,
+      "loss": 2.4985,
+      "step": 13937
+    },
+    {
+      "epoch": 1.1248486804939068,
+      "grad_norm": 0.6787589192390442,
+      "learning_rate": 4.242845822701798e-05,
+      "loss": 2.402,
+      "step": 13938
+    },
+    {
+      "epoch": 1.12492938423049,
+      "grad_norm": 0.6567786931991577,
+      "learning_rate": 4.241555078526602e-05,
+      "loss": 2.4295,
+      "step": 13939
+    },
+    {
+      "epoch": 1.1250100879670728,
+      "grad_norm": 0.6962547302246094,
+      "learning_rate": 4.2402644778669074e-05,
+      "loss": 2.4006,
+      "step": 13940
+    },
+    {
+      "epoch": 1.125090791703656,
+      "grad_norm": 0.7152721285820007,
+      "learning_rate": 4.238974020754877e-05,
+      "loss": 2.4757,
+      "step": 13941
+    },
+    {
+      "epoch": 1.125171495440239,
+      "grad_norm": 0.6869861483573914,
+      "learning_rate": 4.237683707222677e-05,
+      "loss": 2.3877,
+      "step": 13942
+    },
+    {
+      "epoch": 1.1252521991768218,
+      "grad_norm": 0.6951470971107483,
+      "learning_rate": 4.236393537302459e-05,
+      "loss": 2.3755,
+      "step": 13943
+    },
+    {
+      "epoch": 1.125332902913405,
+      "grad_norm": 0.6997567415237427,
+      "learning_rate": 4.2351035110263805e-05,
+      "loss": 2.4731,
+      "step": 13944
+    },
+    {
+      "epoch": 1.125413606649988,
+      "grad_norm": 0.6765854358673096,
+      "learning_rate": 4.23381362842659e-05,
+      "loss": 2.4004,
+      "step": 13945
+    },
+    {
+      "epoch": 1.1254943103865709,
+      "grad_norm": 0.7046722173690796,
+      "learning_rate": 4.2325238895352426e-05,
+      "loss": 2.4379,
+      "step": 13946
+    },
+    {
+      "epoch": 1.125575014123154,
+      "grad_norm": 0.6862985491752625,
+      "learning_rate": 4.231234294384472e-05,
+      "loss": 2.4614,
+      "step": 13947
+    },
+    {
+      "epoch": 1.1256557178597368,
+      "grad_norm": 0.6637778282165527,
+      "learning_rate": 4.229944843006422e-05,
+      "loss": 2.4412,
+      "step": 13948
+    },
+    {
+      "epoch": 1.12573642159632,
+      "grad_norm": 0.7042228579521179,
+      "learning_rate": 4.228655535433231e-05,
+      "loss": 2.4296,
+      "step": 13949
+    },
+    {
+      "epoch": 1.1258171253329028,
+      "grad_norm": 0.6767764687538147,
+      "learning_rate": 4.227366371697029e-05,
+      "loss": 2.409,
+      "step": 13950
+    },
+    {
+      "epoch": 1.1258978290694859,
+      "grad_norm": 0.6886798143386841,
+      "learning_rate": 4.226077351829948e-05,
+      "loss": 2.4786,
+      "step": 13951
+    },
+    {
+      "epoch": 1.125978532806069,
+      "grad_norm": 0.7723653316497803,
+      "learning_rate": 4.224788475864115e-05,
+      "loss": 2.4111,
+      "step": 13952
+    },
+    {
+      "epoch": 1.1260592365426518,
+      "grad_norm": 0.7614055275917053,
+      "learning_rate": 4.2234997438316473e-05,
+      "loss": 2.5055,
+      "step": 13953
+    },
+    {
+      "epoch": 1.126139940279235,
+      "grad_norm": 0.7195241451263428,
+      "learning_rate": 4.222211155764665e-05,
+      "loss": 2.411,
+      "step": 13954
+    },
+    {
+      "epoch": 1.126220644015818,
+      "grad_norm": 0.7130021452903748,
+      "learning_rate": 4.220922711695288e-05,
+      "loss": 2.4819,
+      "step": 13955
+    },
+    {
+      "epoch": 1.1263013477524009,
+      "grad_norm": 0.6972241401672363,
+      "learning_rate": 4.2196344116556194e-05,
+      "loss": 2.4611,
+      "step": 13956
+    },
+    {
+      "epoch": 1.126382051488984,
+      "grad_norm": 0.7023231387138367,
+      "learning_rate": 4.218346255677772e-05,
+      "loss": 2.4509,
+      "step": 13957
+    },
+    {
+      "epoch": 1.126462755225567,
+      "grad_norm": 0.6959301829338074,
+      "learning_rate": 4.2170582437938534e-05,
+      "loss": 2.4441,
+      "step": 13958
+    },
+    {
+      "epoch": 1.12654345896215,
+      "grad_norm": 0.7423149347305298,
+      "learning_rate": 4.2157703760359555e-05,
+      "loss": 2.4452,
+      "step": 13959
+    },
+    {
+      "epoch": 1.126624162698733,
+      "grad_norm": 0.6587820053100586,
+      "learning_rate": 4.214482652436177e-05,
+      "loss": 2.3936,
+      "step": 13960
+    },
+    {
+      "epoch": 1.1267048664353159,
+      "grad_norm": 0.6601768136024475,
+      "learning_rate": 4.213195073026618e-05,
+      "loss": 2.453,
+      "step": 13961
+    },
+    {
+      "epoch": 1.126785570171899,
+      "grad_norm": 0.6986891031265259,
+      "learning_rate": 4.2119076378393676e-05,
+      "loss": 2.452,
+      "step": 13962
+    },
+    {
+      "epoch": 1.126866273908482,
+      "grad_norm": 0.7207025289535522,
+      "learning_rate": 4.2106203469065055e-05,
+      "loss": 2.4048,
+      "step": 13963
+    },
+    {
+      "epoch": 1.126946977645065,
+      "grad_norm": 0.6731177568435669,
+      "learning_rate": 4.2093332002601184e-05,
+      "loss": 2.4573,
+      "step": 13964
+    },
+    {
+      "epoch": 1.127027681381648,
+      "grad_norm": 0.7330070734024048,
+      "learning_rate": 4.208046197932288e-05,
+      "loss": 2.4274,
+      "step": 13965
+    },
+    {
+      "epoch": 1.1271083851182309,
+      "grad_norm": 0.7008770704269409,
+      "learning_rate": 4.206759339955084e-05,
+      "loss": 2.4933,
+      "step": 13966
+    },
+    {
+      "epoch": 1.127189088854814,
+      "grad_norm": 0.8309584259986877,
+      "learning_rate": 4.20547262636058e-05,
+      "loss": 2.3857,
+      "step": 13967
+    },
+    {
+      "epoch": 1.127269792591397,
+      "grad_norm": 0.6705843210220337,
+      "learning_rate": 4.204186057180849e-05,
+      "loss": 2.4303,
+      "step": 13968
+    },
+    {
+      "epoch": 1.12735049632798,
+      "grad_norm": 0.7526851296424866,
+      "learning_rate": 4.202899632447949e-05,
+      "loss": 2.455,
+      "step": 13969
+    },
+    {
+      "epoch": 1.127431200064563,
+      "grad_norm": 0.6690995097160339,
+      "learning_rate": 4.201613352193943e-05,
+      "loss": 2.4398,
+      "step": 13970
+    },
+    {
+      "epoch": 1.127511903801146,
+      "grad_norm": 0.6946840286254883,
+      "learning_rate": 4.20032721645089e-05,
+      "loss": 2.4032,
+      "step": 13971
+    },
+    {
+      "epoch": 1.127592607537729,
+      "grad_norm": 0.7438863515853882,
+      "learning_rate": 4.1990412252508426e-05,
+      "loss": 2.4644,
+      "step": 13972
+    },
+    {
+      "epoch": 1.127673311274312,
+      "grad_norm": 0.6975359916687012,
+      "learning_rate": 4.197755378625852e-05,
+      "loss": 2.3991,
+      "step": 13973
+    },
+    {
+      "epoch": 1.1277540150108951,
+      "grad_norm": 0.6799279451370239,
+      "learning_rate": 4.196469676607968e-05,
+      "loss": 2.4328,
+      "step": 13974
+    },
+    {
+      "epoch": 1.127834718747478,
+      "grad_norm": 0.7014481425285339,
+      "learning_rate": 4.1951841192292274e-05,
+      "loss": 2.5045,
+      "step": 13975
+    },
+    {
+      "epoch": 1.127915422484061,
+      "grad_norm": 0.7074011564254761,
+      "learning_rate": 4.1938987065216716e-05,
+      "loss": 2.4583,
+      "step": 13976
+    },
+    {
+      "epoch": 1.127996126220644,
+      "grad_norm": 0.7246339917182922,
+      "learning_rate": 4.192613438517338e-05,
+      "loss": 2.447,
+      "step": 13977
+    },
+    {
+      "epoch": 1.128076829957227,
+      "grad_norm": 0.6757462620735168,
+      "learning_rate": 4.191328315248262e-05,
+      "loss": 2.4181,
+      "step": 13978
+    },
+    {
+      "epoch": 1.12815753369381,
+      "grad_norm": 0.6758493185043335,
+      "learning_rate": 4.1900433367464644e-05,
+      "loss": 2.4837,
+      "step": 13979
+    },
+    {
+      "epoch": 1.128238237430393,
+      "grad_norm": 0.6782165765762329,
+      "learning_rate": 4.1887585030439736e-05,
+      "loss": 2.3946,
+      "step": 13980
+    },
+    {
+      "epoch": 1.128318941166976,
+      "grad_norm": 0.7176415324211121,
+      "learning_rate": 4.187473814172812e-05,
+      "loss": 2.4538,
+      "step": 13981
+    },
+    {
+      "epoch": 1.128399644903559,
+      "grad_norm": 0.6636224985122681,
+      "learning_rate": 4.186189270164997e-05,
+      "loss": 2.4493,
+      "step": 13982
+    },
+    {
+      "epoch": 1.128480348640142,
+      "grad_norm": 0.6613143086433411,
+      "learning_rate": 4.184904871052544e-05,
+      "loss": 2.4994,
+      "step": 13983
+    },
+    {
+      "epoch": 1.128561052376725,
+      "grad_norm": 0.7148364186286926,
+      "learning_rate": 4.183620616867465e-05,
+      "loss": 2.4673,
+      "step": 13984
+    },
+    {
+      "epoch": 1.128641756113308,
+      "grad_norm": 0.6657952070236206,
+      "learning_rate": 4.1823365076417606e-05,
+      "loss": 2.3915,
+      "step": 13985
+    },
+    {
+      "epoch": 1.128722459849891,
+      "grad_norm": 0.7135687470436096,
+      "learning_rate": 4.181052543407439e-05,
+      "loss": 2.4961,
+      "step": 13986
+    },
+    {
+      "epoch": 1.1288031635864741,
+      "grad_norm": 0.7245377898216248,
+      "learning_rate": 4.179768724196501e-05,
+      "loss": 2.4519,
+      "step": 13987
+    },
+    {
+      "epoch": 1.128883867323057,
+      "grad_norm": 0.6832938194274902,
+      "learning_rate": 4.1784850500409376e-05,
+      "loss": 2.4471,
+      "step": 13988
+    },
+    {
+      "epoch": 1.12896457105964,
+      "grad_norm": 0.7303032279014587,
+      "learning_rate": 4.177201520972746e-05,
+      "loss": 2.3906,
+      "step": 13989
+    },
+    {
+      "epoch": 1.1290452747962232,
+      "grad_norm": 0.698581874370575,
+      "learning_rate": 4.175918137023911e-05,
+      "loss": 2.4667,
+      "step": 13990
+    },
+    {
+      "epoch": 1.129125978532806,
+      "grad_norm": 0.69133061170578,
+      "learning_rate": 4.174634898226422e-05,
+      "loss": 2.4285,
+      "step": 13991
+    },
+    {
+      "epoch": 1.1292066822693891,
+      "grad_norm": 0.7029501795768738,
+      "learning_rate": 4.1733518046122576e-05,
+      "loss": 2.4839,
+      "step": 13992
+    },
+    {
+      "epoch": 1.129287386005972,
+      "grad_norm": 0.7566521167755127,
+      "learning_rate": 4.172068856213398e-05,
+      "loss": 2.5019,
+      "step": 13993
+    },
+    {
+      "epoch": 1.129368089742555,
+      "grad_norm": 0.697998046875,
+      "learning_rate": 4.1707860530618204e-05,
+      "loss": 2.4305,
+      "step": 13994
+    },
+    {
+      "epoch": 1.1294487934791382,
+      "grad_norm": 0.674194872379303,
+      "learning_rate": 4.169503395189489e-05,
+      "loss": 2.4361,
+      "step": 13995
+    },
+    {
+      "epoch": 1.129529497215721,
+      "grad_norm": 0.6936436891555786,
+      "learning_rate": 4.168220882628373e-05,
+      "loss": 2.518,
+      "step": 13996
+    },
+    {
+      "epoch": 1.1296102009523041,
+      "grad_norm": 0.6831670999526978,
+      "learning_rate": 4.166938515410442e-05,
+      "loss": 2.4197,
+      "step": 13997
+    },
+    {
+      "epoch": 1.129690904688887,
+      "grad_norm": 0.7323662638664246,
+      "learning_rate": 4.165656293567647e-05,
+      "loss": 2.4555,
+      "step": 13998
+    },
+    {
+      "epoch": 1.12977160842547,
+      "grad_norm": 0.7699782848358154,
+      "learning_rate": 4.164374217131948e-05,
+      "loss": 2.4456,
+      "step": 13999
+    },
+    {
+      "epoch": 1.1298523121620532,
+      "grad_norm": 0.7009051442146301,
+      "learning_rate": 4.163092286135297e-05,
+      "loss": 2.4429,
+      "step": 14000
+    },
+    {
+      "epoch": 1.1298523121620532,
+      "eval_loss": 2.4034411907196045,
+      "eval_runtime": 771.1158,
+      "eval_samples_per_second": 3.398,
+      "eval_steps_per_second": 0.567,
+      "step": 14000
+    },
+    {
+      "epoch": 1.129933015898636,
+      "grad_norm": 0.674665093421936,
+      "learning_rate": 4.1618105006096456e-05,
+      "loss": 2.4127,
+      "step": 14001
+    },
+    {
+      "epoch": 1.1300137196352191,
+      "grad_norm": 0.7332403659820557,
+      "learning_rate": 4.1605288605869365e-05,
+      "loss": 2.4854,
+      "step": 14002
+    },
+    {
+      "epoch": 1.1300944233718022,
+      "grad_norm": 0.70233553647995,
+      "learning_rate": 4.159247366099117e-05,
+      "loss": 2.4433,
+      "step": 14003
+    },
+    {
+      "epoch": 1.130175127108385,
+      "grad_norm": 0.6259445548057556,
+      "learning_rate": 4.157966017178118e-05,
+      "loss": 2.3605,
+      "step": 14004
+    },
+    {
+      "epoch": 1.1302558308449682,
+      "grad_norm": 0.717408299446106,
+      "learning_rate": 4.1566848138558755e-05,
+      "loss": 2.4378,
+      "step": 14005
+    },
+    {
+      "epoch": 1.130336534581551,
+      "grad_norm": 0.6973297595977783,
+      "learning_rate": 4.155403756164323e-05,
+      "loss": 2.4363,
+      "step": 14006
+    },
+    {
+      "epoch": 1.1304172383181341,
+      "grad_norm": 0.7204940915107727,
+      "learning_rate": 4.154122844135391e-05,
+      "loss": 2.4814,
+      "step": 14007
+    },
+    {
+      "epoch": 1.1304979420547172,
+      "grad_norm": 0.8976696133613586,
+      "learning_rate": 4.1528420778009935e-05,
+      "loss": 2.4654,
+      "step": 14008
+    },
+    {
+      "epoch": 1.1305786457913,
+      "grad_norm": 0.7270354628562927,
+      "learning_rate": 4.151561457193057e-05,
+      "loss": 2.4088,
+      "step": 14009
+    },
+    {
+      "epoch": 1.1306593495278832,
+      "grad_norm": 0.7200367450714111,
+      "learning_rate": 4.1502809823434985e-05,
+      "loss": 2.4412,
+      "step": 14010
+    },
+    {
+      "epoch": 1.130740053264466,
+      "grad_norm": 0.7593986392021179,
+      "learning_rate": 4.149000653284227e-05,
+      "loss": 2.5058,
+      "step": 14011
+    },
+    {
+      "epoch": 1.1308207570010491,
+      "grad_norm": 0.7322795987129211,
+      "learning_rate": 4.147720470047155e-05,
+      "loss": 2.4899,
+      "step": 14012
+    },
+    {
+      "epoch": 1.1309014607376322,
+      "grad_norm": 0.6649030447006226,
+      "learning_rate": 4.1464404326641905e-05,
+      "loss": 2.4358,
+      "step": 14013
+    },
+    {
+      "epoch": 1.130982164474215,
+      "grad_norm": 0.7258814573287964,
+      "learning_rate": 4.145160541167228e-05,
+      "loss": 2.4732,
+      "step": 14014
+    },
+    {
+      "epoch": 1.1310628682107982,
+      "grad_norm": 0.7414976358413696,
+      "learning_rate": 4.1438807955881695e-05,
+      "loss": 2.4157,
+      "step": 14015
+    },
+    {
+      "epoch": 1.1311435719473812,
+      "grad_norm": 0.6813236474990845,
+      "learning_rate": 4.142601195958914e-05,
+      "loss": 2.3966,
+      "step": 14016
+    },
+    {
+      "epoch": 1.131224275683964,
+      "grad_norm": 0.6715923547744751,
+      "learning_rate": 4.141321742311344e-05,
+      "loss": 2.4358,
+      "step": 14017
+    },
+    {
+      "epoch": 1.1313049794205472,
+      "grad_norm": 0.7174912691116333,
+      "learning_rate": 4.14004243467735e-05,
+      "loss": 2.4838,
+      "step": 14018
+    },
+    {
+      "epoch": 1.1313856831571303,
+      "grad_norm": 0.6945109963417053,
+      "learning_rate": 4.138763273088821e-05,
+      "loss": 2.4674,
+      "step": 14019
+    },
+    {
+      "epoch": 1.1314663868937131,
+      "grad_norm": 0.6759494543075562,
+      "learning_rate": 4.137484257577629e-05,
+      "loss": 2.4659,
+      "step": 14020
+    },
+    {
+      "epoch": 1.1315470906302962,
+      "grad_norm": 0.7077876925468445,
+      "learning_rate": 4.1362053881756534e-05,
+      "loss": 2.4731,
+      "step": 14021
+    },
+    {
+      "epoch": 1.131627794366879,
+      "grad_norm": 0.6769500970840454,
+      "learning_rate": 4.1349266649147654e-05,
+      "loss": 2.3606,
+      "step": 14022
+    },
+    {
+      "epoch": 1.1317084981034622,
+      "grad_norm": 0.7104208469390869,
+      "learning_rate": 4.1336480878268424e-05,
+      "loss": 2.4626,
+      "step": 14023
+    },
+    {
+      "epoch": 1.1317892018400453,
+      "grad_norm": 0.7102686762809753,
+      "learning_rate": 4.132369656943741e-05,
+      "loss": 2.4545,
+      "step": 14024
+    },
+    {
+      "epoch": 1.1318699055766281,
+      "grad_norm": 0.7773897647857666,
+      "learning_rate": 4.1310913722973256e-05,
+      "loss": 2.5107,
+      "step": 14025
+    },
+    {
+      "epoch": 1.1319506093132112,
+      "grad_norm": 0.6427130103111267,
+      "learning_rate": 4.1298132339194585e-05,
+      "loss": 2.4349,
+      "step": 14026
+    },
+    {
+      "epoch": 1.132031313049794,
+      "grad_norm": 0.6725162863731384,
+      "learning_rate": 4.128535241841987e-05,
+      "loss": 2.4566,
+      "step": 14027
+    },
+    {
+      "epoch": 1.1321120167863772,
+      "grad_norm": 0.7182251214981079,
+      "learning_rate": 4.127257396096764e-05,
+      "loss": 2.4472,
+      "step": 14028
+    },
+    {
+      "epoch": 1.1321927205229603,
+      "grad_norm": 0.6712302565574646,
+      "learning_rate": 4.1259796967156426e-05,
+      "loss": 2.4326,
+      "step": 14029
+    },
+    {
+      "epoch": 1.1322734242595431,
+      "grad_norm": 0.7726041078567505,
+      "learning_rate": 4.124702143730459e-05,
+      "loss": 2.4994,
+      "step": 14030
+    },
+    {
+      "epoch": 1.1323541279961262,
+      "grad_norm": 0.651899516582489,
+      "learning_rate": 4.123424737173056e-05,
+      "loss": 2.4244,
+      "step": 14031
+    },
+    {
+      "epoch": 1.1324348317327093,
+      "grad_norm": 0.6646261215209961,
+      "learning_rate": 4.12214747707527e-05,
+      "loss": 2.5027,
+      "step": 14032
+    },
+    {
+      "epoch": 1.1325155354692922,
+      "grad_norm": 0.729098916053772,
+      "learning_rate": 4.120870363468933e-05,
+      "loss": 2.5117,
+      "step": 14033
+    },
+    {
+      "epoch": 1.1325962392058753,
+      "grad_norm": 0.7056638598442078,
+      "learning_rate": 4.119593396385876e-05,
+      "loss": 2.4279,
+      "step": 14034
+    },
+    {
+      "epoch": 1.1326769429424584,
+      "grad_norm": 0.7051844000816345,
+      "learning_rate": 4.1183165758579255e-05,
+      "loss": 2.3844,
+      "step": 14035
+    },
+    {
+      "epoch": 1.1327576466790412,
+      "grad_norm": 0.6954311728477478,
+      "learning_rate": 4.1170399019168984e-05,
+      "loss": 2.4041,
+      "step": 14036
+    },
+    {
+      "epoch": 1.1328383504156243,
+      "grad_norm": 0.650044858455658,
+      "learning_rate": 4.1157633745946135e-05,
+      "loss": 2.4397,
+      "step": 14037
+    },
+    {
+      "epoch": 1.1329190541522072,
+      "grad_norm": 0.6974380016326904,
+      "learning_rate": 4.114486993922888e-05,
+      "loss": 2.4391,
+      "step": 14038
+    },
+    {
+      "epoch": 1.1329997578887903,
+      "grad_norm": 0.7252807021141052,
+      "learning_rate": 4.113210759933536e-05,
+      "loss": 2.4471,
+      "step": 14039
+    },
+    {
+      "epoch": 1.1330804616253733,
+      "grad_norm": 0.7001414895057678,
+      "learning_rate": 4.111934672658354e-05,
+      "loss": 2.402,
+      "step": 14040
+    },
+    {
+      "epoch": 1.1331611653619562,
+      "grad_norm": 0.7420533895492554,
+      "learning_rate": 4.110658732129153e-05,
+      "loss": 2.4987,
+      "step": 14041
+    },
+    {
+      "epoch": 1.1332418690985393,
+      "grad_norm": 0.6850644946098328,
+      "learning_rate": 4.1093829383777315e-05,
+      "loss": 2.4355,
+      "step": 14042
+    },
+    {
+      "epoch": 1.1333225728351222,
+      "grad_norm": 0.6905977725982666,
+      "learning_rate": 4.108107291435885e-05,
+      "loss": 2.4818,
+      "step": 14043
+    },
+    {
+      "epoch": 1.1334032765717053,
+      "grad_norm": 0.6555112600326538,
+      "learning_rate": 4.106831791335407e-05,
+      "loss": 2.425,
+      "step": 14044
+    },
+    {
+      "epoch": 1.1334839803082883,
+      "grad_norm": 0.6570355892181396,
+      "learning_rate": 4.105556438108089e-05,
+      "loss": 2.4232,
+      "step": 14045
+    },
+    {
+      "epoch": 1.1335646840448712,
+      "grad_norm": 0.7910747528076172,
+      "learning_rate": 4.104281231785708e-05,
+      "loss": 2.484,
+      "step": 14046
+    },
+    {
+      "epoch": 1.1336453877814543,
+      "grad_norm": 0.6581952571868896,
+      "learning_rate": 4.103006172400052e-05,
+      "loss": 2.4102,
+      "step": 14047
+    },
+    {
+      "epoch": 1.1337260915180374,
+      "grad_norm": 0.6834773421287537,
+      "learning_rate": 4.1017312599828994e-05,
+      "loss": 2.4602,
+      "step": 14048
+    },
+    {
+      "epoch": 1.1338067952546202,
+      "grad_norm": 0.7588350772857666,
+      "learning_rate": 4.1004564945660195e-05,
+      "loss": 2.5059,
+      "step": 14049
+    },
+    {
+      "epoch": 1.1338874989912033,
+      "grad_norm": 0.6604699492454529,
+      "learning_rate": 4.099181876181185e-05,
+      "loss": 2.4403,
+      "step": 14050
+    },
+    {
+      "epoch": 1.1339682027277862,
+      "grad_norm": 0.6957669258117676,
+      "learning_rate": 4.097907404860163e-05,
+      "loss": 2.4218,
+      "step": 14051
+    },
+    {
+      "epoch": 1.1340489064643693,
+      "grad_norm": 0.7091849446296692,
+      "learning_rate": 4.0966330806347166e-05,
+      "loss": 2.4396,
+      "step": 14052
+    },
+    {
+      "epoch": 1.1341296102009524,
+      "grad_norm": 0.6637482047080994,
+      "learning_rate": 4.095358903536605e-05,
+      "loss": 2.4514,
+      "step": 14053
+    },
+    {
+      "epoch": 1.1342103139375352,
+      "grad_norm": 0.7485960125923157,
+      "learning_rate": 4.0940848735975846e-05,
+      "loss": 2.4401,
+      "step": 14054
+    },
+    {
+      "epoch": 1.1342910176741183,
+      "grad_norm": 0.6509774327278137,
+      "learning_rate": 4.092810990849411e-05,
+      "loss": 2.4575,
+      "step": 14055
+    },
+    {
+      "epoch": 1.1343717214107012,
+      "grad_norm": 0.7151626348495483,
+      "learning_rate": 4.091537255323825e-05,
+      "loss": 2.45,
+      "step": 14056
+    },
+    {
+      "epoch": 1.1344524251472843,
+      "grad_norm": 0.7536267042160034,
+      "learning_rate": 4.0902636670525764e-05,
+      "loss": 2.497,
+      "step": 14057
+    },
+    {
+      "epoch": 1.1345331288838674,
+      "grad_norm": 0.7779545783996582,
+      "learning_rate": 4.0889902260674086e-05,
+      "loss": 2.412,
+      "step": 14058
+    },
+    {
+      "epoch": 1.1346138326204502,
+      "grad_norm": 0.7211748957633972,
+      "learning_rate": 4.087716932400052e-05,
+      "loss": 2.4727,
+      "step": 14059
+    },
+    {
+      "epoch": 1.1346945363570333,
+      "grad_norm": 0.6710701584815979,
+      "learning_rate": 4.086443786082245e-05,
+      "loss": 2.4318,
+      "step": 14060
+    },
+    {
+      "epoch": 1.1347752400936164,
+      "grad_norm": 0.7072857022285461,
+      "learning_rate": 4.085170787145717e-05,
+      "loss": 2.4672,
+      "step": 14061
+    },
+    {
+      "epoch": 1.1348559438301993,
+      "grad_norm": 0.6475152969360352,
+      "learning_rate": 4.083897935622194e-05,
+      "loss": 2.4104,
+      "step": 14062
+    },
+    {
+      "epoch": 1.1349366475667824,
+      "grad_norm": 0.7408067584037781,
+      "learning_rate": 4.0826252315433986e-05,
+      "loss": 2.4129,
+      "step": 14063
+    },
+    {
+      "epoch": 1.1350173513033655,
+      "grad_norm": 0.732540488243103,
+      "learning_rate": 4.081352674941056e-05,
+      "loss": 2.4209,
+      "step": 14064
+    },
+    {
+      "epoch": 1.1350980550399483,
+      "grad_norm": 0.6933332681655884,
+      "learning_rate": 4.080080265846872e-05,
+      "loss": 2.3797,
+      "step": 14065
+    },
+    {
+      "epoch": 1.1351787587765314,
+      "grad_norm": 0.6507896780967712,
+      "learning_rate": 4.078808004292561e-05,
+      "loss": 2.4372,
+      "step": 14066
+    },
+    {
+      "epoch": 1.1352594625131143,
+      "grad_norm": 0.729292094707489,
+      "learning_rate": 4.0775358903098384e-05,
+      "loss": 2.5513,
+      "step": 14067
+    },
+    {
+      "epoch": 1.1353401662496974,
+      "grad_norm": 0.692757248878479,
+      "learning_rate": 4.076263923930398e-05,
+      "loss": 2.4228,
+      "step": 14068
+    },
+    {
+      "epoch": 1.1354208699862804,
+      "grad_norm": 0.7028260231018066,
+      "learning_rate": 4.074992105185946e-05,
+      "loss": 2.4478,
+      "step": 14069
+    },
+    {
+      "epoch": 1.1355015737228633,
+      "grad_norm": 0.65067058801651,
+      "learning_rate": 4.073720434108179e-05,
+      "loss": 2.3729,
+      "step": 14070
+    },
+    {
+      "epoch": 1.1355822774594464,
+      "grad_norm": 0.6884061098098755,
+      "learning_rate": 4.0724489107287933e-05,
+      "loss": 2.3693,
+      "step": 14071
+    },
+    {
+      "epoch": 1.1356629811960293,
+      "grad_norm": 0.70686936378479,
+      "learning_rate": 4.071177535079472e-05,
+      "loss": 2.4989,
+      "step": 14072
+    },
+    {
+      "epoch": 1.1357436849326124,
+      "grad_norm": 0.6792482733726501,
+      "learning_rate": 4.0699063071919016e-05,
+      "loss": 2.393,
+      "step": 14073
+    },
+    {
+      "epoch": 1.1358243886691954,
+      "grad_norm": 0.7231085896492004,
+      "learning_rate": 4.0686352270977745e-05,
+      "loss": 2.4597,
+      "step": 14074
+    },
+    {
+      "epoch": 1.1359050924057783,
+      "grad_norm": 0.8024532198905945,
+      "learning_rate": 4.067364294828758e-05,
+      "loss": 2.4409,
+      "step": 14075
+    },
+    {
+      "epoch": 1.1359857961423614,
+      "grad_norm": 0.6761424541473389,
+      "learning_rate": 4.066093510416532e-05,
+      "loss": 2.4598,
+      "step": 14076
+    },
+    {
+      "epoch": 1.1360664998789445,
+      "grad_norm": 0.7075559496879578,
+      "learning_rate": 4.064822873892771e-05,
+      "loss": 2.4649,
+      "step": 14077
+    },
+    {
+      "epoch": 1.1361472036155273,
+      "grad_norm": 0.6292272806167603,
+      "learning_rate": 4.063552385289134e-05,
+      "loss": 2.445,
+      "step": 14078
+    },
+    {
+      "epoch": 1.1362279073521104,
+      "grad_norm": 0.6435273885726929,
+      "learning_rate": 4.06228204463729e-05,
+      "loss": 2.4105,
+      "step": 14079
+    },
+    {
+      "epoch": 1.1363086110886935,
+      "grad_norm": 0.7135637402534485,
+      "learning_rate": 4.061011851968903e-05,
+      "loss": 2.3907,
+      "step": 14080
+    },
+    {
+      "epoch": 1.1363893148252764,
+      "grad_norm": 0.7424013614654541,
+      "learning_rate": 4.059741807315621e-05,
+      "loss": 2.4405,
+      "step": 14081
+    },
+    {
+      "epoch": 1.1364700185618595,
+      "grad_norm": 0.6649916768074036,
+      "learning_rate": 4.0584719107091016e-05,
+      "loss": 2.4314,
+      "step": 14082
+    },
+    {
+      "epoch": 1.1365507222984423,
+      "grad_norm": 0.6700563430786133,
+      "learning_rate": 4.0572021621809944e-05,
+      "loss": 2.4093,
+      "step": 14083
+    },
+    {
+      "epoch": 1.1366314260350254,
+      "grad_norm": 0.6740709543228149,
+      "learning_rate": 4.055932561762942e-05,
+      "loss": 2.4301,
+      "step": 14084
+    },
+    {
+      "epoch": 1.1367121297716085,
+      "grad_norm": 0.7039555907249451,
+      "learning_rate": 4.0546631094865895e-05,
+      "loss": 2.4427,
+      "step": 14085
+    },
+    {
+      "epoch": 1.1367928335081914,
+      "grad_norm": 0.7461164593696594,
+      "learning_rate": 4.053393805383573e-05,
+      "loss": 2.3865,
+      "step": 14086
+    },
+    {
+      "epoch": 1.1368735372447745,
+      "grad_norm": 0.6808290481567383,
+      "learning_rate": 4.0521246494855316e-05,
+      "loss": 2.3738,
+      "step": 14087
+    },
+    {
+      "epoch": 1.1369542409813573,
+      "grad_norm": 0.6942760944366455,
+      "learning_rate": 4.0508556418240875e-05,
+      "loss": 2.4351,
+      "step": 14088
+    },
+    {
+      "epoch": 1.1370349447179404,
+      "grad_norm": 0.7615510821342468,
+      "learning_rate": 4.049586782430872e-05,
+      "loss": 2.3968,
+      "step": 14089
+    },
+    {
+      "epoch": 1.1371156484545235,
+      "grad_norm": 0.7240662574768066,
+      "learning_rate": 4.048318071337512e-05,
+      "loss": 2.4046,
+      "step": 14090
+    },
+    {
+      "epoch": 1.1371963521911064,
+      "grad_norm": 0.7286471128463745,
+      "learning_rate": 4.047049508575621e-05,
+      "loss": 2.4039,
+      "step": 14091
+    },
+    {
+      "epoch": 1.1372770559276895,
+      "grad_norm": 0.7031459212303162,
+      "learning_rate": 4.045781094176816e-05,
+      "loss": 2.4494,
+      "step": 14092
+    },
+    {
+      "epoch": 1.1373577596642725,
+      "grad_norm": 0.7116301655769348,
+      "learning_rate": 4.0445128281727116e-05,
+      "loss": 2.3991,
+      "step": 14093
+    },
+    {
+      "epoch": 1.1374384634008554,
+      "grad_norm": 0.6719788312911987,
+      "learning_rate": 4.043244710594914e-05,
+      "loss": 2.4823,
+      "step": 14094
+    },
+    {
+      "epoch": 1.1375191671374385,
+      "grad_norm": 0.6770508885383606,
+      "learning_rate": 4.041976741475031e-05,
+      "loss": 2.4362,
+      "step": 14095
+    },
+    {
+      "epoch": 1.1375998708740216,
+      "grad_norm": 0.6808609962463379,
+      "learning_rate": 4.040708920844666e-05,
+      "loss": 2.435,
+      "step": 14096
+    },
+    {
+      "epoch": 1.1376805746106045,
+      "grad_norm": 0.7445514798164368,
+      "learning_rate": 4.0394412487354074e-05,
+      "loss": 2.4749,
+      "step": 14097
+    },
+    {
+      "epoch": 1.1377612783471875,
+      "grad_norm": 0.7024775743484497,
+      "learning_rate": 4.038173725178854e-05,
+      "loss": 2.4354,
+      "step": 14098
+    },
+    {
+      "epoch": 1.1378419820837704,
+      "grad_norm": 0.6925685405731201,
+      "learning_rate": 4.0369063502066e-05,
+      "loss": 2.4462,
+      "step": 14099
+    },
+    {
+      "epoch": 1.1379226858203535,
+      "grad_norm": 0.6970539689064026,
+      "learning_rate": 4.035639123850223e-05,
+      "loss": 2.3842,
+      "step": 14100
+    },
+    {
+      "epoch": 1.1380033895569364,
+      "grad_norm": 0.6571836471557617,
+      "learning_rate": 4.0343720461413107e-05,
+      "loss": 2.4213,
+      "step": 14101
+    },
+    {
+      "epoch": 1.1380840932935194,
+      "grad_norm": 0.7264918684959412,
+      "learning_rate": 4.033105117111441e-05,
+      "loss": 2.4697,
+      "step": 14102
+    },
+    {
+      "epoch": 1.1381647970301025,
+      "grad_norm": 0.6929560899734497,
+      "learning_rate": 4.03183833679219e-05,
+      "loss": 2.461,
+      "step": 14103
+    },
+    {
+      "epoch": 1.1382455007666854,
+      "grad_norm": 0.6533559560775757,
+      "learning_rate": 4.030571705215128e-05,
+      "loss": 2.4336,
+      "step": 14104
+    },
+    {
+      "epoch": 1.1383262045032685,
+      "grad_norm": 0.7372364401817322,
+      "learning_rate": 4.0293052224118234e-05,
+      "loss": 2.4396,
+      "step": 14105
+    },
+    {
+      "epoch": 1.1384069082398516,
+      "grad_norm": 0.6736310720443726,
+      "learning_rate": 4.028038888413844e-05,
+      "loss": 2.4123,
+      "step": 14106
+    },
+    {
+      "epoch": 1.1384876119764344,
+      "grad_norm": 0.6898338794708252,
+      "learning_rate": 4.026772703252742e-05,
+      "loss": 2.431,
+      "step": 14107
+    },
+    {
+      "epoch": 1.1385683157130175,
+      "grad_norm": 0.7933369278907776,
+      "learning_rate": 4.02550666696008e-05,
+      "loss": 2.4669,
+      "step": 14108
+    },
+    {
+      "epoch": 1.1386490194496006,
+      "grad_norm": 0.7218122482299805,
+      "learning_rate": 4.024240779567412e-05,
+      "loss": 2.3761,
+      "step": 14109
+    },
+    {
+      "epoch": 1.1387297231861835,
+      "grad_norm": 0.7018248438835144,
+      "learning_rate": 4.022975041106281e-05,
+      "loss": 2.4011,
+      "step": 14110
+    },
+    {
+      "epoch": 1.1388104269227666,
+      "grad_norm": 0.6709668040275574,
+      "learning_rate": 4.0217094516082364e-05,
+      "loss": 2.426,
+      "step": 14111
+    },
+    {
+      "epoch": 1.1388911306593494,
+      "grad_norm": 0.7241504192352295,
+      "learning_rate": 4.0204440111048195e-05,
+      "loss": 2.4085,
+      "step": 14112
+    },
+    {
+      "epoch": 1.1389718343959325,
+      "grad_norm": 0.731347382068634,
+      "learning_rate": 4.0191787196275675e-05,
+      "loss": 2.502,
+      "step": 14113
+    },
+    {
+      "epoch": 1.1390525381325156,
+      "grad_norm": 0.6630167365074158,
+      "learning_rate": 4.0179135772080166e-05,
+      "loss": 2.3999,
+      "step": 14114
+    },
+    {
+      "epoch": 1.1391332418690985,
+      "grad_norm": 0.7094748616218567,
+      "learning_rate": 4.016648583877698e-05,
+      "loss": 2.4666,
+      "step": 14115
+    },
+    {
+      "epoch": 1.1392139456056816,
+      "grad_norm": 0.7262436151504517,
+      "learning_rate": 4.0153837396681395e-05,
+      "loss": 2.4369,
+      "step": 14116
+    },
+    {
+      "epoch": 1.1392946493422644,
+      "grad_norm": 0.6796039938926697,
+      "learning_rate": 4.014119044610859e-05,
+      "loss": 2.4607,
+      "step": 14117
+    },
+    {
+      "epoch": 1.1393753530788475,
+      "grad_norm": 0.6690036058425903,
+      "learning_rate": 4.0128544987373785e-05,
+      "loss": 2.4145,
+      "step": 14118
+    },
+    {
+      "epoch": 1.1394560568154306,
+      "grad_norm": 0.6987181305885315,
+      "learning_rate": 4.011590102079219e-05,
+      "loss": 2.4294,
+      "step": 14119
+    },
+    {
+      "epoch": 1.1395367605520135,
+      "grad_norm": 0.6756789684295654,
+      "learning_rate": 4.0103258546678836e-05,
+      "loss": 2.396,
+      "step": 14120
+    },
+    {
+      "epoch": 1.1396174642885966,
+      "grad_norm": 0.7027772068977356,
+      "learning_rate": 4.009061756534885e-05,
+      "loss": 2.3971,
+      "step": 14121
+    },
+    {
+      "epoch": 1.1396981680251796,
+      "grad_norm": 0.6872174143791199,
+      "learning_rate": 4.007797807711732e-05,
+      "loss": 2.4297,
+      "step": 14122
+    },
+    {
+      "epoch": 1.1397788717617625,
+      "grad_norm": 0.7213007211685181,
+      "learning_rate": 4.006534008229914e-05,
+      "loss": 2.4792,
+      "step": 14123
+    },
+    {
+      "epoch": 1.1398595754983456,
+      "grad_norm": 0.6771649122238159,
+      "learning_rate": 4.0052703581209395e-05,
+      "loss": 2.4397,
+      "step": 14124
+    },
+    {
+      "epoch": 1.1399402792349287,
+      "grad_norm": 0.6577184796333313,
+      "learning_rate": 4.0040068574163013e-05,
+      "loss": 2.4113,
+      "step": 14125
+    },
+    {
+      "epoch": 1.1400209829715116,
+      "grad_norm": 0.7493160367012024,
+      "learning_rate": 4.002743506147483e-05,
+      "loss": 2.4454,
+      "step": 14126
+    },
+    {
+      "epoch": 1.1401016867080946,
+      "grad_norm": 0.6820357441902161,
+      "learning_rate": 4.0014803043459726e-05,
+      "loss": 2.4126,
+      "step": 14127
+    },
+    {
+      "epoch": 1.1401823904446775,
+      "grad_norm": 0.7177188992500305,
+      "learning_rate": 4.000217252043258e-05,
+      "loss": 2.4355,
+      "step": 14128
+    },
+    {
+      "epoch": 1.1402630941812606,
+      "grad_norm": 0.654371440410614,
+      "learning_rate": 3.998954349270808e-05,
+      "loss": 2.4932,
+      "step": 14129
+    },
+    {
+      "epoch": 1.1403437979178437,
+      "grad_norm": 0.7029837965965271,
+      "learning_rate": 3.997691596060104e-05,
+      "loss": 2.4341,
+      "step": 14130
+    },
+    {
+      "epoch": 1.1404245016544265,
+      "grad_norm": 0.7971171140670776,
+      "learning_rate": 3.996428992442615e-05,
+      "loss": 2.4466,
+      "step": 14131
+    },
+    {
+      "epoch": 1.1405052053910096,
+      "grad_norm": 0.6941849589347839,
+      "learning_rate": 3.9951665384498114e-05,
+      "loss": 2.4861,
+      "step": 14132
+    },
+    {
+      "epoch": 1.1405859091275925,
+      "grad_norm": 0.6657733917236328,
+      "learning_rate": 3.993904234113153e-05,
+      "loss": 2.4266,
+      "step": 14133
+    },
+    {
+      "epoch": 1.1406666128641756,
+      "grad_norm": 0.6780329346656799,
+      "learning_rate": 3.9926420794641e-05,
+      "loss": 2.458,
+      "step": 14134
+    },
+    {
+      "epoch": 1.1407473166007587,
+      "grad_norm": 0.7070702910423279,
+      "learning_rate": 3.991380074534109e-05,
+      "loss": 2.368,
+      "step": 14135
+    },
+    {
+      "epoch": 1.1408280203373415,
+      "grad_norm": 0.7186575531959534,
+      "learning_rate": 3.990118219354635e-05,
+      "loss": 2.4611,
+      "step": 14136
+    },
+    {
+      "epoch": 1.1409087240739246,
+      "grad_norm": 0.7171763777732849,
+      "learning_rate": 3.988856513957123e-05,
+      "loss": 2.4315,
+      "step": 14137
+    },
+    {
+      "epoch": 1.1409894278105077,
+      "grad_norm": 0.7090228796005249,
+      "learning_rate": 3.987594958373025e-05,
+      "loss": 2.4668,
+      "step": 14138
+    },
+    {
+      "epoch": 1.1410701315470906,
+      "grad_norm": 0.6523951888084412,
+      "learning_rate": 3.986333552633773e-05,
+      "loss": 2.4392,
+      "step": 14139
+    },
+    {
+      "epoch": 1.1411508352836737,
+      "grad_norm": 0.706000804901123,
+      "learning_rate": 3.98507229677081e-05,
+      "loss": 2.4382,
+      "step": 14140
+    },
+    {
+      "epoch": 1.1412315390202568,
+      "grad_norm": 0.6537537574768066,
+      "learning_rate": 3.983811190815571e-05,
+      "loss": 2.456,
+      "step": 14141
+    },
+    {
+      "epoch": 1.1413122427568396,
+      "grad_norm": 0.7509549856185913,
+      "learning_rate": 3.982550234799479e-05,
+      "loss": 2.4744,
+      "step": 14142
+    },
+    {
+      "epoch": 1.1413929464934227,
+      "grad_norm": 0.7188650965690613,
+      "learning_rate": 3.981289428753967e-05,
+      "loss": 2.4632,
+      "step": 14143
+    },
+    {
+      "epoch": 1.1414736502300056,
+      "grad_norm": 0.7563674449920654,
+      "learning_rate": 3.9800287727104544e-05,
+      "loss": 2.5063,
+      "step": 14144
+    },
+    {
+      "epoch": 1.1415543539665887,
+      "grad_norm": 0.8374128341674805,
+      "learning_rate": 3.978768266700361e-05,
+      "loss": 2.4942,
+      "step": 14145
+    },
+    {
+      "epoch": 1.1416350577031718,
+      "grad_norm": 0.7020177841186523,
+      "learning_rate": 3.9775079107551027e-05,
+      "loss": 2.4404,
+      "step": 14146
+    },
+    {
+      "epoch": 1.1417157614397546,
+      "grad_norm": 0.7326170802116394,
+      "learning_rate": 3.9762477049060895e-05,
+      "loss": 2.4127,
+      "step": 14147
+    },
+    {
+      "epoch": 1.1417964651763377,
+      "grad_norm": 0.6661173105239868,
+      "learning_rate": 3.974987649184734e-05,
+      "loss": 2.4649,
+      "step": 14148
+    },
+    {
+      "epoch": 1.1418771689129206,
+      "grad_norm": 0.7186033129692078,
+      "learning_rate": 3.973727743622432e-05,
+      "loss": 2.4275,
+      "step": 14149
+    },
+    {
+      "epoch": 1.1419578726495037,
+      "grad_norm": 0.7193881869316101,
+      "learning_rate": 3.972467988250588e-05,
+      "loss": 2.4997,
+      "step": 14150
+    },
+    {
+      "epoch": 1.1420385763860867,
+      "grad_norm": 0.7139542102813721,
+      "learning_rate": 3.971208383100601e-05,
+      "loss": 2.4211,
+      "step": 14151
+    },
+    {
+      "epoch": 1.1421192801226696,
+      "grad_norm": 0.6840166449546814,
+      "learning_rate": 3.969948928203856e-05,
+      "loss": 2.4504,
+      "step": 14152
+    },
+    {
+      "epoch": 1.1421999838592527,
+      "grad_norm": 0.8261072039604187,
+      "learning_rate": 3.968689623591747e-05,
+      "loss": 2.4901,
+      "step": 14153
+    },
+    {
+      "epoch": 1.1422806875958358,
+      "grad_norm": 0.7636086940765381,
+      "learning_rate": 3.96743046929566e-05,
+      "loss": 2.4202,
+      "step": 14154
+    },
+    {
+      "epoch": 1.1423613913324187,
+      "grad_norm": 0.7477976679801941,
+      "learning_rate": 3.966171465346973e-05,
+      "loss": 2.492,
+      "step": 14155
+    },
+    {
+      "epoch": 1.1424420950690017,
+      "grad_norm": 0.7516389489173889,
+      "learning_rate": 3.9649126117770665e-05,
+      "loss": 2.4512,
+      "step": 14156
+    },
+    {
+      "epoch": 1.1425227988055846,
+      "grad_norm": 0.6987521648406982,
+      "learning_rate": 3.9636539086173174e-05,
+      "loss": 2.4005,
+      "step": 14157
+    },
+    {
+      "epoch": 1.1426035025421677,
+      "grad_norm": 0.7242532968521118,
+      "learning_rate": 3.962395355899088e-05,
+      "loss": 2.4414,
+      "step": 14158
+    },
+    {
+      "epoch": 1.1426842062787508,
+      "grad_norm": 0.6616180539131165,
+      "learning_rate": 3.961136953653749e-05,
+      "loss": 2.4442,
+      "step": 14159
+    },
+    {
+      "epoch": 1.1427649100153336,
+      "grad_norm": 0.7165415287017822,
+      "learning_rate": 3.959878701912667e-05,
+      "loss": 2.4658,
+      "step": 14160
+    },
+    {
+      "epoch": 1.1428456137519167,
+      "grad_norm": 0.6619318127632141,
+      "learning_rate": 3.9586206007071926e-05,
+      "loss": 2.3803,
+      "step": 14161
+    },
+    {
+      "epoch": 1.1429263174884996,
+      "grad_norm": 0.6654838919639587,
+      "learning_rate": 3.957362650068684e-05,
+      "loss": 2.4584,
+      "step": 14162
+    },
+    {
+      "epoch": 1.1430070212250827,
+      "grad_norm": 0.6947140097618103,
+      "learning_rate": 3.956104850028496e-05,
+      "loss": 2.4236,
+      "step": 14163
+    },
+    {
+      "epoch": 1.1430877249616658,
+      "grad_norm": 0.6510412096977234,
+      "learning_rate": 3.954847200617973e-05,
+      "loss": 2.3589,
+      "step": 14164
+    },
+    {
+      "epoch": 1.1431684286982486,
+      "grad_norm": 0.7550667524337769,
+      "learning_rate": 3.95358970186846e-05,
+      "loss": 2.419,
+      "step": 14165
+    },
+    {
+      "epoch": 1.1432491324348317,
+      "grad_norm": 0.7898361682891846,
+      "learning_rate": 3.9523323538112975e-05,
+      "loss": 2.4549,
+      "step": 14166
+    },
+    {
+      "epoch": 1.1433298361714148,
+      "grad_norm": 0.7162390947341919,
+      "learning_rate": 3.9510751564778246e-05,
+      "loss": 2.4493,
+      "step": 14167
+    },
+    {
+      "epoch": 1.1434105399079977,
+      "grad_norm": 0.8251990079879761,
+      "learning_rate": 3.949818109899367e-05,
+      "loss": 2.4474,
+      "step": 14168
+    },
+    {
+      "epoch": 1.1434912436445808,
+      "grad_norm": 0.6739209890365601,
+      "learning_rate": 3.948561214107258e-05,
+      "loss": 2.4564,
+      "step": 14169
+    },
+    {
+      "epoch": 1.1435719473811639,
+      "grad_norm": 0.6606340408325195,
+      "learning_rate": 3.9473044691328254e-05,
+      "loss": 2.3838,
+      "step": 14170
+    },
+    {
+      "epoch": 1.1436526511177467,
+      "grad_norm": 0.7297452092170715,
+      "learning_rate": 3.946047875007384e-05,
+      "loss": 2.4673,
+      "step": 14171
+    },
+    {
+      "epoch": 1.1437333548543298,
+      "grad_norm": 0.7382420301437378,
+      "learning_rate": 3.9447914317622546e-05,
+      "loss": 2.4279,
+      "step": 14172
+    },
+    {
+      "epoch": 1.1438140585909127,
+      "grad_norm": 0.6947354674339294,
+      "learning_rate": 3.9435351394287546e-05,
+      "loss": 2.4553,
+      "step": 14173
+    },
+    {
+      "epoch": 1.1438947623274958,
+      "grad_norm": 0.670369565486908,
+      "learning_rate": 3.942278998038183e-05,
+      "loss": 2.4285,
+      "step": 14174
+    },
+    {
+      "epoch": 1.1439754660640788,
+      "grad_norm": 0.7097954154014587,
+      "learning_rate": 3.941023007621859e-05,
+      "loss": 2.477,
+      "step": 14175
+    },
+    {
+      "epoch": 1.1440561698006617,
+      "grad_norm": 0.6490213871002197,
+      "learning_rate": 3.9397671682110826e-05,
+      "loss": 2.3943,
+      "step": 14176
+    },
+    {
+      "epoch": 1.1441368735372448,
+      "grad_norm": 0.6505936980247498,
+      "learning_rate": 3.938511479837147e-05,
+      "loss": 2.4188,
+      "step": 14177
+    },
+    {
+      "epoch": 1.1442175772738277,
+      "grad_norm": 0.6696773767471313,
+      "learning_rate": 3.9372559425313496e-05,
+      "loss": 2.4377,
+      "step": 14178
+    },
+    {
+      "epoch": 1.1442982810104108,
+      "grad_norm": 0.6747034192085266,
+      "learning_rate": 3.936000556324982e-05,
+      "loss": 2.4111,
+      "step": 14179
+    },
+    {
+      "epoch": 1.1443789847469938,
+      "grad_norm": 0.7766546607017517,
+      "learning_rate": 3.934745321249336e-05,
+      "loss": 2.3873,
+      "step": 14180
+    },
+    {
+      "epoch": 1.1444596884835767,
+      "grad_norm": 0.7608100175857544,
+      "learning_rate": 3.933490237335688e-05,
+      "loss": 2.4567,
+      "step": 14181
+    },
+    {
+      "epoch": 1.1445403922201598,
+      "grad_norm": 0.7724356055259705,
+      "learning_rate": 3.9322353046153205e-05,
+      "loss": 2.4729,
+      "step": 14182
+    },
+    {
+      "epoch": 1.1446210959567429,
+      "grad_norm": 0.6908414363861084,
+      "learning_rate": 3.930980523119515e-05,
+      "loss": 2.41,
+      "step": 14183
+    },
+    {
+      "epoch": 1.1447017996933257,
+      "grad_norm": 0.7209733128547668,
+      "learning_rate": 3.9297258928795356e-05,
+      "loss": 2.4629,
+      "step": 14184
+    },
+    {
+      "epoch": 1.1447825034299088,
+      "grad_norm": 0.7116519212722778,
+      "learning_rate": 3.928471413926651e-05,
+      "loss": 2.5081,
+      "step": 14185
+    },
+    {
+      "epoch": 1.144863207166492,
+      "grad_norm": 0.6704578995704651,
+      "learning_rate": 3.9272170862921365e-05,
+      "loss": 2.494,
+      "step": 14186
+    },
+    {
+      "epoch": 1.1449439109030748,
+      "grad_norm": 0.6914607882499695,
+      "learning_rate": 3.9259629100072435e-05,
+      "loss": 2.3979,
+      "step": 14187
+    },
+    {
+      "epoch": 1.1450246146396579,
+      "grad_norm": 0.7413245439529419,
+      "learning_rate": 3.924708885103233e-05,
+      "loss": 2.4534,
+      "step": 14188
+    },
+    {
+      "epoch": 1.1451053183762407,
+      "grad_norm": 0.7411661744117737,
+      "learning_rate": 3.923455011611362e-05,
+      "loss": 2.4191,
+      "step": 14189
+    },
+    {
+      "epoch": 1.1451860221128238,
+      "grad_norm": 0.6581972241401672,
+      "learning_rate": 3.9222012895628716e-05,
+      "loss": 2.4494,
+      "step": 14190
+    },
+    {
+      "epoch": 1.145266725849407,
+      "grad_norm": 0.6628647446632385,
+      "learning_rate": 3.920947718989013e-05,
+      "loss": 2.4483,
+      "step": 14191
+    },
+    {
+      "epoch": 1.1453474295859898,
+      "grad_norm": 0.7068151831626892,
+      "learning_rate": 3.9196942999210316e-05,
+      "loss": 2.4549,
+      "step": 14192
+    },
+    {
+      "epoch": 1.1454281333225729,
+      "grad_norm": 0.6727713942527771,
+      "learning_rate": 3.918441032390159e-05,
+      "loss": 2.4261,
+      "step": 14193
+    },
+    {
+      "epoch": 1.1455088370591557,
+      "grad_norm": 0.6680718660354614,
+      "learning_rate": 3.9171879164276334e-05,
+      "loss": 2.4705,
+      "step": 14194
+    },
+    {
+      "epoch": 1.1455895407957388,
+      "grad_norm": 0.710096538066864,
+      "learning_rate": 3.915934952064685e-05,
+      "loss": 2.474,
+      "step": 14195
+    },
+    {
+      "epoch": 1.145670244532322,
+      "grad_norm": 0.6927496790885925,
+      "learning_rate": 3.9146821393325414e-05,
+      "loss": 2.3979,
+      "step": 14196
+    },
+    {
+      "epoch": 1.1457509482689048,
+      "grad_norm": 0.6887550354003906,
+      "learning_rate": 3.913429478262427e-05,
+      "loss": 2.4588,
+      "step": 14197
+    },
+    {
+      "epoch": 1.1458316520054879,
+      "grad_norm": 0.6847062706947327,
+      "learning_rate": 3.912176968885559e-05,
+      "loss": 2.4602,
+      "step": 14198
+    },
+    {
+      "epoch": 1.145912355742071,
+      "grad_norm": 0.6832349300384521,
+      "learning_rate": 3.91092461123316e-05,
+      "loss": 2.4672,
+      "step": 14199
+    },
+    {
+      "epoch": 1.1459930594786538,
+      "grad_norm": 0.6789066791534424,
+      "learning_rate": 3.909672405336432e-05,
+      "loss": 2.5029,
+      "step": 14200
+    },
+    {
+      "epoch": 1.146073763215237,
+      "grad_norm": 0.6953951120376587,
+      "learning_rate": 3.9084203512265885e-05,
+      "loss": 2.4223,
+      "step": 14201
+    },
+    {
+      "epoch": 1.1461544669518198,
+      "grad_norm": 0.6629688739776611,
+      "learning_rate": 3.907168448934836e-05,
+      "loss": 2.4028,
+      "step": 14202
+    },
+    {
+      "epoch": 1.1462351706884029,
+      "grad_norm": 0.6661216020584106,
+      "learning_rate": 3.90591669849237e-05,
+      "loss": 2.4668,
+      "step": 14203
+    },
+    {
+      "epoch": 1.146315874424986,
+      "grad_norm": 0.6814442276954651,
+      "learning_rate": 3.9046650999303894e-05,
+      "loss": 2.4273,
+      "step": 14204
+    },
+    {
+      "epoch": 1.1463965781615688,
+      "grad_norm": 0.6678626537322998,
+      "learning_rate": 3.903413653280088e-05,
+      "loss": 2.444,
+      "step": 14205
+    },
+    {
+      "epoch": 1.146477281898152,
+      "grad_norm": 0.6703703999519348,
+      "learning_rate": 3.902162358572655e-05,
+      "loss": 2.4273,
+      "step": 14206
+    },
+    {
+      "epoch": 1.1465579856347348,
+      "grad_norm": 0.7052578926086426,
+      "learning_rate": 3.900911215839276e-05,
+      "loss": 2.4397,
+      "step": 14207
+    },
+    {
+      "epoch": 1.1466386893713179,
+      "grad_norm": 0.6792036294937134,
+      "learning_rate": 3.899660225111136e-05,
+      "loss": 2.439,
+      "step": 14208
+    },
+    {
+      "epoch": 1.146719393107901,
+      "grad_norm": 0.6995401978492737,
+      "learning_rate": 3.898409386419407e-05,
+      "loss": 2.5002,
+      "step": 14209
+    },
+    {
+      "epoch": 1.1468000968444838,
+      "grad_norm": 0.6527338027954102,
+      "learning_rate": 3.897158699795265e-05,
+      "loss": 2.4523,
+      "step": 14210
+    },
+    {
+      "epoch": 1.146880800581067,
+      "grad_norm": 0.7509400248527527,
+      "learning_rate": 3.8959081652698814e-05,
+      "loss": 2.4193,
+      "step": 14211
+    },
+    {
+      "epoch": 1.14696150431765,
+      "grad_norm": 0.6985350251197815,
+      "learning_rate": 3.894657782874426e-05,
+      "loss": 2.4251,
+      "step": 14212
+    },
+    {
+      "epoch": 1.1470422080542328,
+      "grad_norm": 0.6831483840942383,
+      "learning_rate": 3.893407552640055e-05,
+      "loss": 2.4172,
+      "step": 14213
+    },
+    {
+      "epoch": 1.147122911790816,
+      "grad_norm": 0.7281469702720642,
+      "learning_rate": 3.892157474597929e-05,
+      "loss": 2.4451,
+      "step": 14214
+    },
+    {
+      "epoch": 1.147203615527399,
+      "grad_norm": 0.7326027750968933,
+      "learning_rate": 3.8909075487792066e-05,
+      "loss": 2.3926,
+      "step": 14215
+    },
+    {
+      "epoch": 1.1472843192639819,
+      "grad_norm": 0.7030496597290039,
+      "learning_rate": 3.889657775215036e-05,
+      "loss": 2.435,
+      "step": 14216
+    },
+    {
+      "epoch": 1.147365023000565,
+      "grad_norm": 0.6915596127510071,
+      "learning_rate": 3.888408153936568e-05,
+      "loss": 2.4622,
+      "step": 14217
+    },
+    {
+      "epoch": 1.1474457267371478,
+      "grad_norm": 0.678600013256073,
+      "learning_rate": 3.8871586849749474e-05,
+      "loss": 2.4264,
+      "step": 14218
+    },
+    {
+      "epoch": 1.147526430473731,
+      "grad_norm": 0.7487786412239075,
+      "learning_rate": 3.885909368361308e-05,
+      "loss": 2.4038,
+      "step": 14219
+    },
+    {
+      "epoch": 1.147607134210314,
+      "grad_norm": 0.6658064723014832,
+      "learning_rate": 3.8846602041267886e-05,
+      "loss": 2.4079,
+      "step": 14220
+    },
+    {
+      "epoch": 1.1476878379468969,
+      "grad_norm": 0.6985111832618713,
+      "learning_rate": 3.883411192302527e-05,
+      "loss": 2.481,
+      "step": 14221
+    },
+    {
+      "epoch": 1.14776854168348,
+      "grad_norm": 0.7056208848953247,
+      "learning_rate": 3.8821623329196445e-05,
+      "loss": 2.4409,
+      "step": 14222
+    },
+    {
+      "epoch": 1.1478492454200628,
+      "grad_norm": 0.7107830047607422,
+      "learning_rate": 3.880913626009268e-05,
+      "loss": 2.4578,
+      "step": 14223
+    },
+    {
+      "epoch": 1.147929949156646,
+      "grad_norm": 0.6678555607795715,
+      "learning_rate": 3.87966507160252e-05,
+      "loss": 2.4548,
+      "step": 14224
+    },
+    {
+      "epoch": 1.148010652893229,
+      "grad_norm": 0.6699830293655396,
+      "learning_rate": 3.8784166697305157e-05,
+      "loss": 2.3763,
+      "step": 14225
+    },
+    {
+      "epoch": 1.1480913566298119,
+      "grad_norm": 0.7695464491844177,
+      "learning_rate": 3.8771684204243716e-05,
+      "loss": 2.4774,
+      "step": 14226
+    },
+    {
+      "epoch": 1.148172060366395,
+      "grad_norm": 0.7801330089569092,
+      "learning_rate": 3.8759203237151954e-05,
+      "loss": 2.4598,
+      "step": 14227
+    },
+    {
+      "epoch": 1.148252764102978,
+      "grad_norm": 0.7029622793197632,
+      "learning_rate": 3.8746723796340955e-05,
+      "loss": 2.3901,
+      "step": 14228
+    },
+    {
+      "epoch": 1.148333467839561,
+      "grad_norm": 0.7472359538078308,
+      "learning_rate": 3.873424588212169e-05,
+      "loss": 2.4724,
+      "step": 14229
+    },
+    {
+      "epoch": 1.148414171576144,
+      "grad_norm": 0.6621725559234619,
+      "learning_rate": 3.872176949480517e-05,
+      "loss": 2.4523,
+      "step": 14230
+    },
+    {
+      "epoch": 1.148494875312727,
+      "grad_norm": 0.722658634185791,
+      "learning_rate": 3.8709294634702376e-05,
+      "loss": 2.4032,
+      "step": 14231
+    },
+    {
+      "epoch": 1.14857557904931,
+      "grad_norm": 0.7743202447891235,
+      "learning_rate": 3.869682130212413e-05,
+      "loss": 2.4373,
+      "step": 14232
+    },
+    {
+      "epoch": 1.148656282785893,
+      "grad_norm": 0.6906178593635559,
+      "learning_rate": 3.868434949738136e-05,
+      "loss": 2.4765,
+      "step": 14233
+    },
+    {
+      "epoch": 1.148736986522476,
+      "grad_norm": 0.6708275675773621,
+      "learning_rate": 3.86718792207849e-05,
+      "loss": 2.4263,
+      "step": 14234
+    },
+    {
+      "epoch": 1.148817690259059,
+      "grad_norm": 0.6992776989936829,
+      "learning_rate": 3.8659410472645494e-05,
+      "loss": 2.378,
+      "step": 14235
+    },
+    {
+      "epoch": 1.148898393995642,
+      "grad_norm": 0.7229011058807373,
+      "learning_rate": 3.864694325327389e-05,
+      "loss": 2.4075,
+      "step": 14236
+    },
+    {
+      "epoch": 1.148979097732225,
+      "grad_norm": 0.6622509956359863,
+      "learning_rate": 3.863447756298091e-05,
+      "loss": 2.3954,
+      "step": 14237
+    },
+    {
+      "epoch": 1.149059801468808,
+      "grad_norm": 0.7233534455299377,
+      "learning_rate": 3.862201340207712e-05,
+      "loss": 2.4506,
+      "step": 14238
+    },
+    {
+      "epoch": 1.149140505205391,
+      "grad_norm": 0.716869056224823,
+      "learning_rate": 3.860955077087321e-05,
+      "loss": 2.4304,
+      "step": 14239
+    },
+    {
+      "epoch": 1.149221208941974,
+      "grad_norm": 0.6550257205963135,
+      "learning_rate": 3.8597089669679766e-05,
+      "loss": 2.4261,
+      "step": 14240
+    },
+    {
+      "epoch": 1.149301912678557,
+      "grad_norm": 0.6981741786003113,
+      "learning_rate": 3.858463009880738e-05,
+      "loss": 2.4115,
+      "step": 14241
+    },
+    {
+      "epoch": 1.14938261641514,
+      "grad_norm": 0.6792196035385132,
+      "learning_rate": 3.8572172058566534e-05,
+      "loss": 2.4195,
+      "step": 14242
+    },
+    {
+      "epoch": 1.149463320151723,
+      "grad_norm": 0.7278807163238525,
+      "learning_rate": 3.855971554926773e-05,
+      "loss": 2.418,
+      "step": 14243
+    },
+    {
+      "epoch": 1.1495440238883061,
+      "grad_norm": 0.6451076865196228,
+      "learning_rate": 3.8547260571221456e-05,
+      "loss": 2.4591,
+      "step": 14244
+    },
+    {
+      "epoch": 1.149624727624889,
+      "grad_norm": 0.7052451968193054,
+      "learning_rate": 3.853480712473805e-05,
+      "loss": 2.4023,
+      "step": 14245
+    },
+    {
+      "epoch": 1.149705431361472,
+      "grad_norm": 0.7016182541847229,
+      "learning_rate": 3.852235521012793e-05,
+      "loss": 2.4959,
+      "step": 14246
+    },
+    {
+      "epoch": 1.1497861350980552,
+      "grad_norm": 0.7287492156028748,
+      "learning_rate": 3.850990482770141e-05,
+      "loss": 2.3884,
+      "step": 14247
+    },
+    {
+      "epoch": 1.149866838834638,
+      "grad_norm": 0.6648508310317993,
+      "learning_rate": 3.84974559777688e-05,
+      "loss": 2.4632,
+      "step": 14248
+    },
+    {
+      "epoch": 1.1499475425712211,
+      "grad_norm": 0.7387828230857849,
+      "learning_rate": 3.848500866064036e-05,
+      "loss": 2.4053,
+      "step": 14249
+    },
+    {
+      "epoch": 1.150028246307804,
+      "grad_norm": 0.7230356931686401,
+      "learning_rate": 3.847256287662635e-05,
+      "loss": 2.5128,
+      "step": 14250
+    },
+    {
+      "epoch": 1.150108950044387,
+      "grad_norm": 0.7209547162055969,
+      "learning_rate": 3.846011862603686e-05,
+      "loss": 2.4626,
+      "step": 14251
+    },
+    {
+      "epoch": 1.1501896537809702,
+      "grad_norm": 0.7177916765213013,
+      "learning_rate": 3.844767590918209e-05,
+      "loss": 2.4469,
+      "step": 14252
+    },
+    {
+      "epoch": 1.150270357517553,
+      "grad_norm": 0.7850151658058167,
+      "learning_rate": 3.843523472637216e-05,
+      "loss": 2.4731,
+      "step": 14253
+    },
+    {
+      "epoch": 1.150351061254136,
+      "grad_norm": 0.7051519155502319,
+      "learning_rate": 3.8422795077917084e-05,
+      "loss": 2.3696,
+      "step": 14254
+    },
+    {
+      "epoch": 1.150431764990719,
+      "grad_norm": 0.7434025406837463,
+      "learning_rate": 3.841035696412692e-05,
+      "loss": 2.444,
+      "step": 14255
+    },
+    {
+      "epoch": 1.150512468727302,
+      "grad_norm": 0.7404719591140747,
+      "learning_rate": 3.839792038531166e-05,
+      "loss": 2.4415,
+      "step": 14256
+    },
+    {
+      "epoch": 1.1505931724638851,
+      "grad_norm": 0.6883764266967773,
+      "learning_rate": 3.838548534178125e-05,
+      "loss": 2.4887,
+      "step": 14257
+    },
+    {
+      "epoch": 1.150673876200468,
+      "grad_norm": 0.6697155237197876,
+      "learning_rate": 3.83730518338456e-05,
+      "loss": 2.3721,
+      "step": 14258
+    },
+    {
+      "epoch": 1.150754579937051,
+      "grad_norm": 0.68825763463974,
+      "learning_rate": 3.836061986181459e-05,
+      "loss": 2.4712,
+      "step": 14259
+    },
+    {
+      "epoch": 1.1508352836736342,
+      "grad_norm": 0.6810611486434937,
+      "learning_rate": 3.8348189425998114e-05,
+      "loss": 2.3995,
+      "step": 14260
+    },
+    {
+      "epoch": 1.150915987410217,
+      "grad_norm": 0.6718329787254333,
+      "learning_rate": 3.8335760526705866e-05,
+      "loss": 2.4068,
+      "step": 14261
+    },
+    {
+      "epoch": 1.1509966911468001,
+      "grad_norm": 0.694618284702301,
+      "learning_rate": 3.832333316424767e-05,
+      "loss": 2.458,
+      "step": 14262
+    },
+    {
+      "epoch": 1.151077394883383,
+      "grad_norm": 0.6824250817298889,
+      "learning_rate": 3.8310907338933266e-05,
+      "loss": 2.4623,
+      "step": 14263
+    },
+    {
+      "epoch": 1.151158098619966,
+      "grad_norm": 0.6875178217887878,
+      "learning_rate": 3.8298483051072264e-05,
+      "loss": 2.4827,
+      "step": 14264
+    },
+    {
+      "epoch": 1.1512388023565492,
+      "grad_norm": 0.7868281602859497,
+      "learning_rate": 3.828606030097437e-05,
+      "loss": 2.4638,
+      "step": 14265
+    },
+    {
+      "epoch": 1.151319506093132,
+      "grad_norm": 0.7003639936447144,
+      "learning_rate": 3.8273639088949165e-05,
+      "loss": 2.4885,
+      "step": 14266
+    },
+    {
+      "epoch": 1.1514002098297151,
+      "grad_norm": 0.6965197920799255,
+      "learning_rate": 3.826121941530623e-05,
+      "loss": 2.3983,
+      "step": 14267
+    },
+    {
+      "epoch": 1.151480913566298,
+      "grad_norm": 0.7241101264953613,
+      "learning_rate": 3.824880128035509e-05,
+      "loss": 2.4598,
+      "step": 14268
+    },
+    {
+      "epoch": 1.151561617302881,
+      "grad_norm": 0.700764536857605,
+      "learning_rate": 3.823638468440528e-05,
+      "loss": 2.3627,
+      "step": 14269
+    },
+    {
+      "epoch": 1.1516423210394642,
+      "grad_norm": 0.6889846324920654,
+      "learning_rate": 3.822396962776619e-05,
+      "loss": 2.4442,
+      "step": 14270
+    },
+    {
+      "epoch": 1.151723024776047,
+      "grad_norm": 0.6660009026527405,
+      "learning_rate": 3.8211556110747245e-05,
+      "loss": 2.403,
+      "step": 14271
+    },
+    {
+      "epoch": 1.1518037285126301,
+      "grad_norm": 0.6537240743637085,
+      "learning_rate": 3.819914413365785e-05,
+      "loss": 2.4358,
+      "step": 14272
+    },
+    {
+      "epoch": 1.1518844322492132,
+      "grad_norm": 0.6852741837501526,
+      "learning_rate": 3.818673369680735e-05,
+      "loss": 2.4272,
+      "step": 14273
+    },
+    {
+      "epoch": 1.151965135985796,
+      "grad_norm": 0.701874852180481,
+      "learning_rate": 3.817432480050501e-05,
+      "loss": 2.4419,
+      "step": 14274
+    },
+    {
+      "epoch": 1.1520458397223792,
+      "grad_norm": 0.7089500427246094,
+      "learning_rate": 3.816191744506011e-05,
+      "loss": 2.4537,
+      "step": 14275
+    },
+    {
+      "epoch": 1.1521265434589623,
+      "grad_norm": 0.698564887046814,
+      "learning_rate": 3.8149511630781866e-05,
+      "loss": 2.3991,
+      "step": 14276
+    },
+    {
+      "epoch": 1.1522072471955451,
+      "grad_norm": 0.6940335035324097,
+      "learning_rate": 3.813710735797947e-05,
+      "loss": 2.5022,
+      "step": 14277
+    },
+    {
+      "epoch": 1.1522879509321282,
+      "grad_norm": 0.6916826367378235,
+      "learning_rate": 3.812470462696208e-05,
+      "loss": 2.4449,
+      "step": 14278
+    },
+    {
+      "epoch": 1.152368654668711,
+      "grad_norm": 0.7115256190299988,
+      "learning_rate": 3.811230343803882e-05,
+      "loss": 2.4371,
+      "step": 14279
+    },
+    {
+      "epoch": 1.1524493584052942,
+      "grad_norm": 0.6857369542121887,
+      "learning_rate": 3.80999037915187e-05,
+      "loss": 2.4426,
+      "step": 14280
+    },
+    {
+      "epoch": 1.1525300621418773,
+      "grad_norm": 0.7605363130569458,
+      "learning_rate": 3.808750568771079e-05,
+      "loss": 2.4999,
+      "step": 14281
+    },
+    {
+      "epoch": 1.1526107658784601,
+      "grad_norm": 0.6604358553886414,
+      "learning_rate": 3.8075109126924115e-05,
+      "loss": 2.419,
+      "step": 14282
+    },
+    {
+      "epoch": 1.1526914696150432,
+      "grad_norm": 0.6945412755012512,
+      "learning_rate": 3.806271410946756e-05,
+      "loss": 2.4555,
+      "step": 14283
+    },
+    {
+      "epoch": 1.152772173351626,
+      "grad_norm": 0.7205908894538879,
+      "learning_rate": 3.805032063565007e-05,
+      "loss": 2.4745,
+      "step": 14284
+    },
+    {
+      "epoch": 1.1528528770882092,
+      "grad_norm": 0.7198025584220886,
+      "learning_rate": 3.8037928705780554e-05,
+      "loss": 2.4358,
+      "step": 14285
+    },
+    {
+      "epoch": 1.1529335808247922,
+      "grad_norm": 0.7231044769287109,
+      "learning_rate": 3.802553832016781e-05,
+      "loss": 2.4713,
+      "step": 14286
+    },
+    {
+      "epoch": 1.1530142845613751,
+      "grad_norm": 0.6878815293312073,
+      "learning_rate": 3.80131494791206e-05,
+      "loss": 2.4479,
+      "step": 14287
+    },
+    {
+      "epoch": 1.1530949882979582,
+      "grad_norm": 0.6930533647537231,
+      "learning_rate": 3.800076218294779e-05,
+      "loss": 2.3912,
+      "step": 14288
+    },
+    {
+      "epoch": 1.1531756920345413,
+      "grad_norm": 0.703521192073822,
+      "learning_rate": 3.798837643195808e-05,
+      "loss": 2.451,
+      "step": 14289
+    },
+    {
+      "epoch": 1.1532563957711242,
+      "grad_norm": 0.7099746465682983,
+      "learning_rate": 3.79759922264601e-05,
+      "loss": 2.4957,
+      "step": 14290
+    },
+    {
+      "epoch": 1.1533370995077072,
+      "grad_norm": 0.7268218398094177,
+      "learning_rate": 3.7963609566762527e-05,
+      "loss": 2.4242,
+      "step": 14291
+    },
+    {
+      "epoch": 1.1534178032442903,
+      "grad_norm": 0.7465239763259888,
+      "learning_rate": 3.7951228453174004e-05,
+      "loss": 2.3867,
+      "step": 14292
+    },
+    {
+      "epoch": 1.1534985069808732,
+      "grad_norm": 0.704584002494812,
+      "learning_rate": 3.793884888600302e-05,
+      "loss": 2.5009,
+      "step": 14293
+    },
+    {
+      "epoch": 1.1535792107174563,
+      "grad_norm": 0.7057262063026428,
+      "learning_rate": 3.792647086555816e-05,
+      "loss": 2.4381,
+      "step": 14294
+    },
+    {
+      "epoch": 1.1536599144540391,
+      "grad_norm": 0.7045955061912537,
+      "learning_rate": 3.791409439214794e-05,
+      "loss": 2.4456,
+      "step": 14295
+    },
+    {
+      "epoch": 1.1537406181906222,
+      "grad_norm": 0.705476701259613,
+      "learning_rate": 3.790171946608074e-05,
+      "loss": 2.466,
+      "step": 14296
+    },
+    {
+      "epoch": 1.1538213219272053,
+      "grad_norm": 0.7128286957740784,
+      "learning_rate": 3.788934608766503e-05,
+      "loss": 2.4891,
+      "step": 14297
+    },
+    {
+      "epoch": 1.1539020256637882,
+      "grad_norm": 0.678144633769989,
+      "learning_rate": 3.787697425720918e-05,
+      "loss": 2.4453,
+      "step": 14298
+    },
+    {
+      "epoch": 1.1539827294003713,
+      "grad_norm": 0.754216730594635,
+      "learning_rate": 3.786460397502151e-05,
+      "loss": 2.4331,
+      "step": 14299
+    },
+    {
+      "epoch": 1.1540634331369541,
+      "grad_norm": 0.6881092190742493,
+      "learning_rate": 3.7852235241410325e-05,
+      "loss": 2.3692,
+      "step": 14300
+    },
+    {
+      "epoch": 1.1541441368735372,
+      "grad_norm": 0.7498507499694824,
+      "learning_rate": 3.783986805668395e-05,
+      "loss": 2.4556,
+      "step": 14301
+    },
+    {
+      "epoch": 1.1542248406101203,
+      "grad_norm": 0.6312216520309448,
+      "learning_rate": 3.7827502421150496e-05,
+      "loss": 2.4727,
+      "step": 14302
+    },
+    {
+      "epoch": 1.1543055443467032,
+      "grad_norm": 0.7156404256820679,
+      "learning_rate": 3.781513833511822e-05,
+      "loss": 2.4003,
+      "step": 14303
+    },
+    {
+      "epoch": 1.1543862480832863,
+      "grad_norm": 0.6589376926422119,
+      "learning_rate": 3.7802775798895226e-05,
+      "loss": 2.4461,
+      "step": 14304
+    },
+    {
+      "epoch": 1.1544669518198694,
+      "grad_norm": 0.7259865999221802,
+      "learning_rate": 3.77904148127897e-05,
+      "loss": 2.4021,
+      "step": 14305
+    },
+    {
+      "epoch": 1.1545476555564522,
+      "grad_norm": 0.7248456478118896,
+      "learning_rate": 3.777805537710961e-05,
+      "loss": 2.4784,
+      "step": 14306
+    },
+    {
+      "epoch": 1.1546283592930353,
+      "grad_norm": 0.7085593342781067,
+      "learning_rate": 3.7765697492163034e-05,
+      "loss": 2.4394,
+      "step": 14307
+    },
+    {
+      "epoch": 1.1547090630296182,
+      "grad_norm": 0.7394313216209412,
+      "learning_rate": 3.775334115825796e-05,
+      "loss": 2.5055,
+      "step": 14308
+    },
+    {
+      "epoch": 1.1547897667662013,
+      "grad_norm": 0.7231999039649963,
+      "learning_rate": 3.7740986375702336e-05,
+      "loss": 2.4551,
+      "step": 14309
+    },
+    {
+      "epoch": 1.1548704705027844,
+      "grad_norm": 0.6875953078269958,
+      "learning_rate": 3.7728633144804084e-05,
+      "loss": 2.4641,
+      "step": 14310
+    },
+    {
+      "epoch": 1.1549511742393672,
+      "grad_norm": 0.7477203607559204,
+      "learning_rate": 3.7716281465871094e-05,
+      "loss": 2.4929,
+      "step": 14311
+    },
+    {
+      "epoch": 1.1550318779759503,
+      "grad_norm": 0.6653971076011658,
+      "learning_rate": 3.770393133921115e-05,
+      "loss": 2.4819,
+      "step": 14312
+    },
+    {
+      "epoch": 1.1551125817125332,
+      "grad_norm": 0.7267318964004517,
+      "learning_rate": 3.769158276513209e-05,
+      "loss": 2.4568,
+      "step": 14313
+    },
+    {
+      "epoch": 1.1551932854491163,
+      "grad_norm": 0.6675654053688049,
+      "learning_rate": 3.76792357439417e-05,
+      "loss": 2.4789,
+      "step": 14314
+    },
+    {
+      "epoch": 1.1552739891856993,
+      "grad_norm": 0.6847487688064575,
+      "learning_rate": 3.7666890275947616e-05,
+      "loss": 2.4034,
+      "step": 14315
+    },
+    {
+      "epoch": 1.1553546929222822,
+      "grad_norm": 0.811553418636322,
+      "learning_rate": 3.765454636145758e-05,
+      "loss": 2.5051,
+      "step": 14316
+    },
+    {
+      "epoch": 1.1554353966588653,
+      "grad_norm": 0.690026581287384,
+      "learning_rate": 3.7642204000779204e-05,
+      "loss": 2.4477,
+      "step": 14317
+    },
+    {
+      "epoch": 1.1555161003954484,
+      "grad_norm": 0.695810079574585,
+      "learning_rate": 3.762986319422013e-05,
+      "loss": 2.4516,
+      "step": 14318
+    },
+    {
+      "epoch": 1.1555968041320313,
+      "grad_norm": 0.6869217753410339,
+      "learning_rate": 3.7617523942087886e-05,
+      "loss": 2.3802,
+      "step": 14319
+    },
+    {
+      "epoch": 1.1556775078686143,
+      "grad_norm": 0.7109078764915466,
+      "learning_rate": 3.7605186244690016e-05,
+      "loss": 2.4306,
+      "step": 14320
+    },
+    {
+      "epoch": 1.1557582116051974,
+      "grad_norm": 0.7385044693946838,
+      "learning_rate": 3.759285010233404e-05,
+      "loss": 2.4288,
+      "step": 14321
+    },
+    {
+      "epoch": 1.1558389153417803,
+      "grad_norm": 0.6775605082511902,
+      "learning_rate": 3.7580515515327355e-05,
+      "loss": 2.4155,
+      "step": 14322
+    },
+    {
+      "epoch": 1.1559196190783634,
+      "grad_norm": 0.7325694561004639,
+      "learning_rate": 3.7568182483977375e-05,
+      "loss": 2.5035,
+      "step": 14323
+    },
+    {
+      "epoch": 1.1560003228149462,
+      "grad_norm": 0.6896799206733704,
+      "learning_rate": 3.7555851008591526e-05,
+      "loss": 2.4739,
+      "step": 14324
+    },
+    {
+      "epoch": 1.1560810265515293,
+      "grad_norm": 0.7086506485939026,
+      "learning_rate": 3.7543521089477065e-05,
+      "loss": 2.4815,
+      "step": 14325
+    },
+    {
+      "epoch": 1.1561617302881124,
+      "grad_norm": 0.6886687874794006,
+      "learning_rate": 3.753119272694132e-05,
+      "loss": 2.4261,
+      "step": 14326
+    },
+    {
+      "epoch": 1.1562424340246953,
+      "grad_norm": 0.675136148929596,
+      "learning_rate": 3.751886592129155e-05,
+      "loss": 2.3946,
+      "step": 14327
+    },
+    {
+      "epoch": 1.1563231377612784,
+      "grad_norm": 0.706729531288147,
+      "learning_rate": 3.7506540672834964e-05,
+      "loss": 2.4199,
+      "step": 14328
+    },
+    {
+      "epoch": 1.1564038414978612,
+      "grad_norm": 0.6790904998779297,
+      "learning_rate": 3.749421698187875e-05,
+      "loss": 2.4419,
+      "step": 14329
+    },
+    {
+      "epoch": 1.1564845452344443,
+      "grad_norm": 0.6688171029090881,
+      "learning_rate": 3.748189484873007e-05,
+      "loss": 2.4516,
+      "step": 14330
+    },
+    {
+      "epoch": 1.1565652489710274,
+      "grad_norm": 0.6782420873641968,
+      "learning_rate": 3.746957427369596e-05,
+      "loss": 2.4586,
+      "step": 14331
+    },
+    {
+      "epoch": 1.1566459527076103,
+      "grad_norm": 0.7633399367332458,
+      "learning_rate": 3.7457255257083514e-05,
+      "loss": 2.3776,
+      "step": 14332
+    },
+    {
+      "epoch": 1.1567266564441934,
+      "grad_norm": 0.680000364780426,
+      "learning_rate": 3.744493779919976e-05,
+      "loss": 2.4978,
+      "step": 14333
+    },
+    {
+      "epoch": 1.1568073601807765,
+      "grad_norm": 0.6993350386619568,
+      "learning_rate": 3.743262190035171e-05,
+      "loss": 2.3974,
+      "step": 14334
+    },
+    {
+      "epoch": 1.1568880639173593,
+      "grad_norm": 0.7316375374794006,
+      "learning_rate": 3.7420307560846234e-05,
+      "loss": 2.4423,
+      "step": 14335
+    },
+    {
+      "epoch": 1.1569687676539424,
+      "grad_norm": 0.7384842038154602,
+      "learning_rate": 3.7407994780990285e-05,
+      "loss": 2.4604,
+      "step": 14336
+    },
+    {
+      "epoch": 1.1570494713905255,
+      "grad_norm": 0.6980708837509155,
+      "learning_rate": 3.739568356109072e-05,
+      "loss": 2.4408,
+      "step": 14337
+    },
+    {
+      "epoch": 1.1571301751271084,
+      "grad_norm": 0.6510182619094849,
+      "learning_rate": 3.738337390145438e-05,
+      "loss": 2.4076,
+      "step": 14338
+    },
+    {
+      "epoch": 1.1572108788636915,
+      "grad_norm": 0.7458614706993103,
+      "learning_rate": 3.737106580238804e-05,
+      "loss": 2.4976,
+      "step": 14339
+    },
+    {
+      "epoch": 1.1572915826002743,
+      "grad_norm": 0.6663469672203064,
+      "learning_rate": 3.735875926419849e-05,
+      "loss": 2.4414,
+      "step": 14340
+    },
+    {
+      "epoch": 1.1573722863368574,
+      "grad_norm": 0.6611858606338501,
+      "learning_rate": 3.7346454287192355e-05,
+      "loss": 2.3783,
+      "step": 14341
+    },
+    {
+      "epoch": 1.1574529900734405,
+      "grad_norm": 0.6605291366577148,
+      "learning_rate": 3.7334150871676364e-05,
+      "loss": 2.4291,
+      "step": 14342
+    },
+    {
+      "epoch": 1.1575336938100234,
+      "grad_norm": 0.6879985928535461,
+      "learning_rate": 3.7321849017957186e-05,
+      "loss": 2.4229,
+      "step": 14343
+    },
+    {
+      "epoch": 1.1576143975466064,
+      "grad_norm": 0.7466493844985962,
+      "learning_rate": 3.7309548726341334e-05,
+      "loss": 2.4278,
+      "step": 14344
+    },
+    {
+      "epoch": 1.1576951012831893,
+      "grad_norm": 0.7476457357406616,
+      "learning_rate": 3.72972499971354e-05,
+      "loss": 2.4944,
+      "step": 14345
+    },
+    {
+      "epoch": 1.1577758050197724,
+      "grad_norm": 0.6339364647865295,
+      "learning_rate": 3.728495283064594e-05,
+      "loss": 2.3753,
+      "step": 14346
+    },
+    {
+      "epoch": 1.1578565087563555,
+      "grad_norm": 0.6885230541229248,
+      "learning_rate": 3.7272657227179355e-05,
+      "loss": 2.4519,
+      "step": 14347
+    },
+    {
+      "epoch": 1.1579372124929384,
+      "grad_norm": 0.7561741471290588,
+      "learning_rate": 3.7260363187042126e-05,
+      "loss": 2.4808,
+      "step": 14348
+    },
+    {
+      "epoch": 1.1580179162295214,
+      "grad_norm": 0.8007705211639404,
+      "learning_rate": 3.724807071054062e-05,
+      "loss": 2.4649,
+      "step": 14349
+    },
+    {
+      "epoch": 1.1580986199661045,
+      "grad_norm": 0.6920937895774841,
+      "learning_rate": 3.72357797979813e-05,
+      "loss": 2.4145,
+      "step": 14350
+    },
+    {
+      "epoch": 1.1581793237026874,
+      "grad_norm": 0.7310675978660583,
+      "learning_rate": 3.7223490449670364e-05,
+      "loss": 2.4475,
+      "step": 14351
+    },
+    {
+      "epoch": 1.1582600274392705,
+      "grad_norm": 0.6600463390350342,
+      "learning_rate": 3.7211202665914155e-05,
+      "loss": 2.3938,
+      "step": 14352
+    },
+    {
+      "epoch": 1.1583407311758536,
+      "grad_norm": 0.690258800983429,
+      "learning_rate": 3.719891644701894e-05,
+      "loss": 2.3944,
+      "step": 14353
+    },
+    {
+      "epoch": 1.1584214349124364,
+      "grad_norm": 0.7075135111808777,
+      "learning_rate": 3.718663179329085e-05,
+      "loss": 2.3931,
+      "step": 14354
+    },
+    {
+      "epoch": 1.1585021386490195,
+      "grad_norm": 0.7416332960128784,
+      "learning_rate": 3.71743487050361e-05,
+      "loss": 2.4566,
+      "step": 14355
+    },
+    {
+      "epoch": 1.1585828423856024,
+      "grad_norm": 0.7459710836410522,
+      "learning_rate": 3.7162067182560846e-05,
+      "loss": 2.4232,
+      "step": 14356
+    },
+    {
+      "epoch": 1.1586635461221855,
+      "grad_norm": 0.7265400886535645,
+      "learning_rate": 3.71497872261711e-05,
+      "loss": 2.4798,
+      "step": 14357
+    },
+    {
+      "epoch": 1.1587442498587683,
+      "grad_norm": 0.7142636775970459,
+      "learning_rate": 3.713750883617294e-05,
+      "loss": 2.4576,
+      "step": 14358
+    },
+    {
+      "epoch": 1.1588249535953514,
+      "grad_norm": 0.7279871702194214,
+      "learning_rate": 3.712523201287239e-05,
+      "loss": 2.439,
+      "step": 14359
+    },
+    {
+      "epoch": 1.1589056573319345,
+      "grad_norm": 0.7151274681091309,
+      "learning_rate": 3.7112956756575414e-05,
+      "loss": 2.4684,
+      "step": 14360
+    },
+    {
+      "epoch": 1.1589863610685174,
+      "grad_norm": 0.7142657041549683,
+      "learning_rate": 3.7100683067587946e-05,
+      "loss": 2.4582,
+      "step": 14361
+    },
+    {
+      "epoch": 1.1590670648051005,
+      "grad_norm": 0.7716035842895508,
+      "learning_rate": 3.7088410946215914e-05,
+      "loss": 2.5038,
+      "step": 14362
+    },
+    {
+      "epoch": 1.1591477685416836,
+      "grad_norm": 0.7232338190078735,
+      "learning_rate": 3.707614039276509e-05,
+      "loss": 2.4558,
+      "step": 14363
+    },
+    {
+      "epoch": 1.1592284722782664,
+      "grad_norm": 0.7388719916343689,
+      "learning_rate": 3.706387140754134e-05,
+      "loss": 2.4535,
+      "step": 14364
+    },
+    {
+      "epoch": 1.1593091760148495,
+      "grad_norm": 0.7022652626037598,
+      "learning_rate": 3.7051603990850425e-05,
+      "loss": 2.4479,
+      "step": 14365
+    },
+    {
+      "epoch": 1.1593898797514326,
+      "grad_norm": 0.7861798405647278,
+      "learning_rate": 3.703933814299813e-05,
+      "loss": 2.4219,
+      "step": 14366
+    },
+    {
+      "epoch": 1.1594705834880155,
+      "grad_norm": 0.6928723454475403,
+      "learning_rate": 3.7027073864290074e-05,
+      "loss": 2.4401,
+      "step": 14367
+    },
+    {
+      "epoch": 1.1595512872245985,
+      "grad_norm": 0.6312821507453918,
+      "learning_rate": 3.701481115503194e-05,
+      "loss": 2.3975,
+      "step": 14368
+    },
+    {
+      "epoch": 1.1596319909611814,
+      "grad_norm": 0.7008257508277893,
+      "learning_rate": 3.700255001552937e-05,
+      "loss": 2.4988,
+      "step": 14369
+    },
+    {
+      "epoch": 1.1597126946977645,
+      "grad_norm": 0.6664693355560303,
+      "learning_rate": 3.699029044608792e-05,
+      "loss": 2.4123,
+      "step": 14370
+    },
+    {
+      "epoch": 1.1597933984343476,
+      "grad_norm": 0.6613842844963074,
+      "learning_rate": 3.6978032447013145e-05,
+      "loss": 2.4802,
+      "step": 14371
+    },
+    {
+      "epoch": 1.1598741021709305,
+      "grad_norm": 0.707788348197937,
+      "learning_rate": 3.696577601861057e-05,
+      "loss": 2.4432,
+      "step": 14372
+    },
+    {
+      "epoch": 1.1599548059075135,
+      "grad_norm": 0.6547604203224182,
+      "learning_rate": 3.695352116118561e-05,
+      "loss": 2.412,
+      "step": 14373
+    },
+    {
+      "epoch": 1.1600355096440964,
+      "grad_norm": 0.7238109707832336,
+      "learning_rate": 3.69412678750437e-05,
+      "loss": 2.4858,
+      "step": 14374
+    },
+    {
+      "epoch": 1.1601162133806795,
+      "grad_norm": 0.8156580328941345,
+      "learning_rate": 3.692901616049026e-05,
+      "loss": 2.4063,
+      "step": 14375
+    },
+    {
+      "epoch": 1.1601969171172626,
+      "grad_norm": 0.7035481333732605,
+      "learning_rate": 3.6916766017830585e-05,
+      "loss": 2.4586,
+      "step": 14376
+    },
+    {
+      "epoch": 1.1602776208538454,
+      "grad_norm": 0.7523401379585266,
+      "learning_rate": 3.690451744736999e-05,
+      "loss": 2.4262,
+      "step": 14377
+    },
+    {
+      "epoch": 1.1603583245904285,
+      "grad_norm": 0.6740732192993164,
+      "learning_rate": 3.689227044941376e-05,
+      "loss": 2.5215,
+      "step": 14378
+    },
+    {
+      "epoch": 1.1604390283270116,
+      "grad_norm": 0.6502695083618164,
+      "learning_rate": 3.6880025024267115e-05,
+      "loss": 2.4292,
+      "step": 14379
+    },
+    {
+      "epoch": 1.1605197320635945,
+      "grad_norm": 0.7000409364700317,
+      "learning_rate": 3.686778117223524e-05,
+      "loss": 2.4323,
+      "step": 14380
+    },
+    {
+      "epoch": 1.1606004358001776,
+      "grad_norm": 0.7415478229522705,
+      "learning_rate": 3.68555388936233e-05,
+      "loss": 2.4515,
+      "step": 14381
+    },
+    {
+      "epoch": 1.1606811395367607,
+      "grad_norm": 0.6890547871589661,
+      "learning_rate": 3.684329818873641e-05,
+      "loss": 2.4115,
+      "step": 14382
+    },
+    {
+      "epoch": 1.1607618432733435,
+      "grad_norm": 0.8238685727119446,
+      "learning_rate": 3.68310590578796e-05,
+      "loss": 2.4666,
+      "step": 14383
+    },
+    {
+      "epoch": 1.1608425470099266,
+      "grad_norm": 0.8098889589309692,
+      "learning_rate": 3.681882150135791e-05,
+      "loss": 2.4667,
+      "step": 14384
+    },
+    {
+      "epoch": 1.1609232507465095,
+      "grad_norm": 0.6932713985443115,
+      "learning_rate": 3.680658551947639e-05,
+      "loss": 2.4574,
+      "step": 14385
+    },
+    {
+      "epoch": 1.1610039544830926,
+      "grad_norm": 0.7062943577766418,
+      "learning_rate": 3.6794351112539915e-05,
+      "loss": 2.4408,
+      "step": 14386
+    },
+    {
+      "epoch": 1.1610846582196757,
+      "grad_norm": 0.7859255075454712,
+      "learning_rate": 3.678211828085343e-05,
+      "loss": 2.3946,
+      "step": 14387
+    },
+    {
+      "epoch": 1.1611653619562585,
+      "grad_norm": 0.674609899520874,
+      "learning_rate": 3.676988702472181e-05,
+      "loss": 2.4456,
+      "step": 14388
+    },
+    {
+      "epoch": 1.1612460656928416,
+      "grad_norm": 0.7068402171134949,
+      "learning_rate": 3.675765734444989e-05,
+      "loss": 2.4393,
+      "step": 14389
+    },
+    {
+      "epoch": 1.1613267694294245,
+      "grad_norm": 0.7276526689529419,
+      "learning_rate": 3.674542924034246e-05,
+      "loss": 2.456,
+      "step": 14390
+    },
+    {
+      "epoch": 1.1614074731660076,
+      "grad_norm": 0.7670585513114929,
+      "learning_rate": 3.673320271270433e-05,
+      "loss": 2.3774,
+      "step": 14391
+    },
+    {
+      "epoch": 1.1614881769025907,
+      "grad_norm": 0.702173113822937,
+      "learning_rate": 3.672097776184013e-05,
+      "loss": 2.3974,
+      "step": 14392
+    },
+    {
+      "epoch": 1.1615688806391735,
+      "grad_norm": 0.6922066807746887,
+      "learning_rate": 3.670875438805457e-05,
+      "loss": 2.4035,
+      "step": 14393
+    },
+    {
+      "epoch": 1.1616495843757566,
+      "grad_norm": 0.6675707697868347,
+      "learning_rate": 3.6696532591652335e-05,
+      "loss": 2.4369,
+      "step": 14394
+    },
+    {
+      "epoch": 1.1617302881123397,
+      "grad_norm": 0.6939712762832642,
+      "learning_rate": 3.668431237293796e-05,
+      "loss": 2.4265,
+      "step": 14395
+    },
+    {
+      "epoch": 1.1618109918489226,
+      "grad_norm": 0.719510018825531,
+      "learning_rate": 3.667209373221602e-05,
+      "loss": 2.4686,
+      "step": 14396
+    },
+    {
+      "epoch": 1.1618916955855056,
+      "grad_norm": 0.7167489528656006,
+      "learning_rate": 3.665987666979104e-05,
+      "loss": 2.5077,
+      "step": 14397
+    },
+    {
+      "epoch": 1.1619723993220887,
+      "grad_norm": 0.6539514064788818,
+      "learning_rate": 3.664766118596754e-05,
+      "loss": 2.4476,
+      "step": 14398
+    },
+    {
+      "epoch": 1.1620531030586716,
+      "grad_norm": 0.6926440596580505,
+      "learning_rate": 3.6635447281049876e-05,
+      "loss": 2.4336,
+      "step": 14399
+    },
+    {
+      "epoch": 1.1621338067952547,
+      "grad_norm": 0.7124993205070496,
+      "learning_rate": 3.662323495534252e-05,
+      "loss": 2.3938,
+      "step": 14400
+    },
+    {
+      "epoch": 1.1622145105318376,
+      "grad_norm": 0.7073954939842224,
+      "learning_rate": 3.661102420914986e-05,
+      "loss": 2.4232,
+      "step": 14401
+    },
+    {
+      "epoch": 1.1622952142684206,
+      "grad_norm": 0.7491076588630676,
+      "learning_rate": 3.659881504277613e-05,
+      "loss": 2.5047,
+      "step": 14402
+    },
+    {
+      "epoch": 1.1623759180050037,
+      "grad_norm": 0.6698675155639648,
+      "learning_rate": 3.658660745652568e-05,
+      "loss": 2.4164,
+      "step": 14403
+    },
+    {
+      "epoch": 1.1624566217415866,
+      "grad_norm": 0.6576815843582153,
+      "learning_rate": 3.657440145070276e-05,
+      "loss": 2.4368,
+      "step": 14404
+    },
+    {
+      "epoch": 1.1625373254781697,
+      "grad_norm": 0.8236953020095825,
+      "learning_rate": 3.6562197025611524e-05,
+      "loss": 2.5041,
+      "step": 14405
+    },
+    {
+      "epoch": 1.1626180292147525,
+      "grad_norm": 0.7391532063484192,
+      "learning_rate": 3.6549994181556157e-05,
+      "loss": 2.4556,
+      "step": 14406
+    },
+    {
+      "epoch": 1.1626987329513356,
+      "grad_norm": 0.6529936790466309,
+      "learning_rate": 3.653779291884084e-05,
+      "loss": 2.4559,
+      "step": 14407
+    },
+    {
+      "epoch": 1.1627794366879187,
+      "grad_norm": 0.7101796269416809,
+      "learning_rate": 3.652559323776957e-05,
+      "loss": 2.3937,
+      "step": 14408
+    },
+    {
+      "epoch": 1.1628601404245016,
+      "grad_norm": 0.6890308260917664,
+      "learning_rate": 3.651339513864645e-05,
+      "loss": 2.4694,
+      "step": 14409
+    },
+    {
+      "epoch": 1.1629408441610847,
+      "grad_norm": 0.6919918060302734,
+      "learning_rate": 3.650119862177548e-05,
+      "loss": 2.4793,
+      "step": 14410
+    },
+    {
+      "epoch": 1.1630215478976678,
+      "grad_norm": 0.6553575992584229,
+      "learning_rate": 3.6489003687460624e-05,
+      "loss": 2.454,
+      "step": 14411
+    },
+    {
+      "epoch": 1.1631022516342506,
+      "grad_norm": 0.7095460891723633,
+      "learning_rate": 3.6476810336005804e-05,
+      "loss": 2.4672,
+      "step": 14412
+    },
+    {
+      "epoch": 1.1631829553708337,
+      "grad_norm": 0.738301694393158,
+      "learning_rate": 3.6464618567714935e-05,
+      "loss": 2.4369,
+      "step": 14413
+    },
+    {
+      "epoch": 1.1632636591074166,
+      "grad_norm": 0.7574542760848999,
+      "learning_rate": 3.645242838289189e-05,
+      "loss": 2.4981,
+      "step": 14414
+    },
+    {
+      "epoch": 1.1633443628439997,
+      "grad_norm": 0.6780585646629333,
+      "learning_rate": 3.64402397818404e-05,
+      "loss": 2.4811,
+      "step": 14415
+    },
+    {
+      "epoch": 1.1634250665805828,
+      "grad_norm": 0.7050060629844666,
+      "learning_rate": 3.6428052764864287e-05,
+      "loss": 2.4607,
+      "step": 14416
+    },
+    {
+      "epoch": 1.1635057703171656,
+      "grad_norm": 0.6946923136711121,
+      "learning_rate": 3.6415867332267316e-05,
+      "loss": 2.4482,
+      "step": 14417
+    },
+    {
+      "epoch": 1.1635864740537487,
+      "grad_norm": 0.7202015519142151,
+      "learning_rate": 3.64036834843531e-05,
+      "loss": 2.4764,
+      "step": 14418
+    },
+    {
+      "epoch": 1.1636671777903316,
+      "grad_norm": 0.7845996618270874,
+      "learning_rate": 3.639150122142534e-05,
+      "loss": 2.4926,
+      "step": 14419
+    },
+    {
+      "epoch": 1.1637478815269147,
+      "grad_norm": 0.6924630403518677,
+      "learning_rate": 3.6379320543787645e-05,
+      "loss": 2.4664,
+      "step": 14420
+    },
+    {
+      "epoch": 1.1638285852634978,
+      "grad_norm": 0.7225920557975769,
+      "learning_rate": 3.636714145174358e-05,
+      "loss": 2.4638,
+      "step": 14421
+    },
+    {
+      "epoch": 1.1639092890000806,
+      "grad_norm": 0.6587103605270386,
+      "learning_rate": 3.63549639455967e-05,
+      "loss": 2.3629,
+      "step": 14422
+    },
+    {
+      "epoch": 1.1639899927366637,
+      "grad_norm": 0.7537658214569092,
+      "learning_rate": 3.634278802565051e-05,
+      "loss": 2.4971,
+      "step": 14423
+    },
+    {
+      "epoch": 1.1640706964732468,
+      "grad_norm": 0.6881381273269653,
+      "learning_rate": 3.633061369220841e-05,
+      "loss": 2.3737,
+      "step": 14424
+    },
+    {
+      "epoch": 1.1641514002098297,
+      "grad_norm": 0.693779468536377,
+      "learning_rate": 3.6318440945573864e-05,
+      "loss": 2.4346,
+      "step": 14425
+    },
+    {
+      "epoch": 1.1642321039464127,
+      "grad_norm": 0.777563750743866,
+      "learning_rate": 3.6306269786050265e-05,
+      "loss": 2.4288,
+      "step": 14426
+    },
+    {
+      "epoch": 1.1643128076829958,
+      "grad_norm": 0.6786738634109497,
+      "learning_rate": 3.629410021394087e-05,
+      "loss": 2.4094,
+      "step": 14427
+    },
+    {
+      "epoch": 1.1643935114195787,
+      "grad_norm": 0.7478442788124084,
+      "learning_rate": 3.628193222954904e-05,
+      "loss": 2.4163,
+      "step": 14428
+    },
+    {
+      "epoch": 1.1644742151561618,
+      "grad_norm": 0.6530766487121582,
+      "learning_rate": 3.626976583317803e-05,
+      "loss": 2.4328,
+      "step": 14429
+    },
+    {
+      "epoch": 1.1645549188927447,
+      "grad_norm": 0.6665371060371399,
+      "learning_rate": 3.6257601025131026e-05,
+      "loss": 2.4006,
+      "step": 14430
+    },
+    {
+      "epoch": 1.1646356226293277,
+      "grad_norm": 0.7184741497039795,
+      "learning_rate": 3.624543780571125e-05,
+      "loss": 2.462,
+      "step": 14431
+    },
+    {
+      "epoch": 1.1647163263659108,
+      "grad_norm": 0.7039462327957153,
+      "learning_rate": 3.6233276175221794e-05,
+      "loss": 2.4321,
+      "step": 14432
+    },
+    {
+      "epoch": 1.1647970301024937,
+      "grad_norm": 0.7039144039154053,
+      "learning_rate": 3.622111613396584e-05,
+      "loss": 2.4399,
+      "step": 14433
+    },
+    {
+      "epoch": 1.1648777338390768,
+      "grad_norm": 0.6690253615379333,
+      "learning_rate": 3.620895768224635e-05,
+      "loss": 2.3976,
+      "step": 14434
+    },
+    {
+      "epoch": 1.1649584375756596,
+      "grad_norm": 0.7048032879829407,
+      "learning_rate": 3.6196800820366384e-05,
+      "loss": 2.4848,
+      "step": 14435
+    },
+    {
+      "epoch": 1.1650391413122427,
+      "grad_norm": 0.668971836566925,
+      "learning_rate": 3.618464554862896e-05,
+      "loss": 2.4614,
+      "step": 14436
+    },
+    {
+      "epoch": 1.1651198450488258,
+      "grad_norm": 0.704858660697937,
+      "learning_rate": 3.617249186733695e-05,
+      "loss": 2.3962,
+      "step": 14437
+    },
+    {
+      "epoch": 1.1652005487854087,
+      "grad_norm": 0.692435085773468,
+      "learning_rate": 3.6160339776793296e-05,
+      "loss": 2.4059,
+      "step": 14438
+    },
+    {
+      "epoch": 1.1652812525219918,
+      "grad_norm": 0.6774182319641113,
+      "learning_rate": 3.614818927730085e-05,
+      "loss": 2.4975,
+      "step": 14439
+    },
+    {
+      "epoch": 1.1653619562585749,
+      "grad_norm": 0.6507411003112793,
+      "learning_rate": 3.613604036916243e-05,
+      "loss": 2.5029,
+      "step": 14440
+    },
+    {
+      "epoch": 1.1654426599951577,
+      "grad_norm": 0.7223206162452698,
+      "learning_rate": 3.612389305268084e-05,
+      "loss": 2.4599,
+      "step": 14441
+    },
+    {
+      "epoch": 1.1655233637317408,
+      "grad_norm": 0.6523364186286926,
+      "learning_rate": 3.611174732815883e-05,
+      "loss": 2.4521,
+      "step": 14442
+    },
+    {
+      "epoch": 1.165604067468324,
+      "grad_norm": 0.6668452024459839,
+      "learning_rate": 3.6099603195899046e-05,
+      "loss": 2.4082,
+      "step": 14443
+    },
+    {
+      "epoch": 1.1656847712049068,
+      "grad_norm": 0.6878299117088318,
+      "learning_rate": 3.60874606562042e-05,
+      "loss": 2.4144,
+      "step": 14444
+    },
+    {
+      "epoch": 1.1657654749414899,
+      "grad_norm": 0.6662277579307556,
+      "learning_rate": 3.6075319709376895e-05,
+      "loss": 2.438,
+      "step": 14445
+    },
+    {
+      "epoch": 1.1658461786780727,
+      "grad_norm": 0.721422553062439,
+      "learning_rate": 3.606318035571976e-05,
+      "loss": 2.4414,
+      "step": 14446
+    },
+    {
+      "epoch": 1.1659268824146558,
+      "grad_norm": 0.6739782691001892,
+      "learning_rate": 3.6051042595535264e-05,
+      "loss": 2.4093,
+      "step": 14447
+    },
+    {
+      "epoch": 1.166007586151239,
+      "grad_norm": 0.6890884637832642,
+      "learning_rate": 3.603890642912596e-05,
+      "loss": 2.4385,
+      "step": 14448
+    },
+    {
+      "epoch": 1.1660882898878218,
+      "grad_norm": 0.6503998637199402,
+      "learning_rate": 3.602677185679433e-05,
+      "loss": 2.4498,
+      "step": 14449
+    },
+    {
+      "epoch": 1.1661689936244048,
+      "grad_norm": 0.6748046875,
+      "learning_rate": 3.601463887884271e-05,
+      "loss": 2.3739,
+      "step": 14450
+    },
+    {
+      "epoch": 1.1662496973609877,
+      "grad_norm": 0.6843422651290894,
+      "learning_rate": 3.600250749557358e-05,
+      "loss": 2.4323,
+      "step": 14451
+    },
+    {
+      "epoch": 1.1663304010975708,
+      "grad_norm": 0.7061208486557007,
+      "learning_rate": 3.599037770728929e-05,
+      "loss": 2.4611,
+      "step": 14452
+    },
+    {
+      "epoch": 1.166411104834154,
+      "grad_norm": 0.6614537239074707,
+      "learning_rate": 3.597824951429208e-05,
+      "loss": 2.4656,
+      "step": 14453
+    },
+    {
+      "epoch": 1.1664918085707368,
+      "grad_norm": 0.6620328426361084,
+      "learning_rate": 3.596612291688424e-05,
+      "loss": 2.415,
+      "step": 14454
+    },
+    {
+      "epoch": 1.1665725123073198,
+      "grad_norm": 0.6936565041542053,
+      "learning_rate": 3.595399791536804e-05,
+      "loss": 2.4655,
+      "step": 14455
+    },
+    {
+      "epoch": 1.166653216043903,
+      "grad_norm": 0.6766063570976257,
+      "learning_rate": 3.594187451004559e-05,
+      "loss": 2.4628,
+      "step": 14456
+    },
+    {
+      "epoch": 1.1667339197804858,
+      "grad_norm": 0.6588734984397888,
+      "learning_rate": 3.592975270121909e-05,
+      "loss": 2.4503,
+      "step": 14457
+    },
+    {
+      "epoch": 1.1668146235170689,
+      "grad_norm": 0.7290894985198975,
+      "learning_rate": 3.591763248919062e-05,
+      "loss": 2.5075,
+      "step": 14458
+    },
+    {
+      "epoch": 1.1668953272536517,
+      "grad_norm": 0.6952784657478333,
+      "learning_rate": 3.590551387426231e-05,
+      "loss": 2.4258,
+      "step": 14459
+    },
+    {
+      "epoch": 1.1669760309902348,
+      "grad_norm": 0.6737042665481567,
+      "learning_rate": 3.5893396856736096e-05,
+      "loss": 2.4459,
+      "step": 14460
+    },
+    {
+      "epoch": 1.167056734726818,
+      "grad_norm": 0.6616976857185364,
+      "learning_rate": 3.588128143691397e-05,
+      "loss": 2.4726,
+      "step": 14461
+    },
+    {
+      "epoch": 1.1671374384634008,
+      "grad_norm": 0.7017171382904053,
+      "learning_rate": 3.5869167615098e-05,
+      "loss": 2.375,
+      "step": 14462
+    },
+    {
+      "epoch": 1.1672181421999839,
+      "grad_norm": 0.7153809666633606,
+      "learning_rate": 3.585705539158997e-05,
+      "loss": 2.4271,
+      "step": 14463
+    },
+    {
+      "epoch": 1.1672988459365667,
+      "grad_norm": 0.749196469783783,
+      "learning_rate": 3.584494476669179e-05,
+      "loss": 2.4713,
+      "step": 14464
+    },
+    {
+      "epoch": 1.1673795496731498,
+      "grad_norm": 0.6593676209449768,
+      "learning_rate": 3.583283574070533e-05,
+      "loss": 2.4276,
+      "step": 14465
+    },
+    {
+      "epoch": 1.167460253409733,
+      "grad_norm": 0.6949084401130676,
+      "learning_rate": 3.5820728313932295e-05,
+      "loss": 2.4128,
+      "step": 14466
+    },
+    {
+      "epoch": 1.1675409571463158,
+      "grad_norm": 0.6795482039451599,
+      "learning_rate": 3.5808622486674484e-05,
+      "loss": 2.485,
+      "step": 14467
+    },
+    {
+      "epoch": 1.1676216608828989,
+      "grad_norm": 0.6763483881950378,
+      "learning_rate": 3.5796518259233625e-05,
+      "loss": 2.4063,
+      "step": 14468
+    },
+    {
+      "epoch": 1.167702364619482,
+      "grad_norm": 0.665687620639801,
+      "learning_rate": 3.578441563191133e-05,
+      "loss": 2.437,
+      "step": 14469
+    },
+    {
+      "epoch": 1.1677830683560648,
+      "grad_norm": 0.6338435411453247,
+      "learning_rate": 3.577231460500926e-05,
+      "loss": 2.3747,
+      "step": 14470
+    },
+    {
+      "epoch": 1.167863772092648,
+      "grad_norm": 0.7031865119934082,
+      "learning_rate": 3.5760215178829e-05,
+      "loss": 2.3952,
+      "step": 14471
+    },
+    {
+      "epoch": 1.167944475829231,
+      "grad_norm": 0.7544599771499634,
+      "learning_rate": 3.5748117353672106e-05,
+      "loss": 2.3941,
+      "step": 14472
+    },
+    {
+      "epoch": 1.1680251795658139,
+      "grad_norm": 0.7271532416343689,
+      "learning_rate": 3.5736021129840083e-05,
+      "loss": 2.4371,
+      "step": 14473
+    },
+    {
+      "epoch": 1.168105883302397,
+      "grad_norm": 0.709048867225647,
+      "learning_rate": 3.572392650763441e-05,
+      "loss": 2.482,
+      "step": 14474
+    },
+    {
+      "epoch": 1.1681865870389798,
+      "grad_norm": 0.6894589066505432,
+      "learning_rate": 3.571183348735653e-05,
+      "loss": 2.4347,
+      "step": 14475
+    },
+    {
+      "epoch": 1.168267290775563,
+      "grad_norm": 0.6680620908737183,
+      "learning_rate": 3.5699742069307774e-05,
+      "loss": 2.3995,
+      "step": 14476
+    },
+    {
+      "epoch": 1.168347994512146,
+      "grad_norm": 0.701669454574585,
+      "learning_rate": 3.568765225378954e-05,
+      "loss": 2.4045,
+      "step": 14477
+    },
+    {
+      "epoch": 1.1684286982487289,
+      "grad_norm": 0.7102392911911011,
+      "learning_rate": 3.567556404110315e-05,
+      "loss": 2.4695,
+      "step": 14478
+    },
+    {
+      "epoch": 1.168509401985312,
+      "grad_norm": 0.6820430755615234,
+      "learning_rate": 3.566347743154982e-05,
+      "loss": 2.4155,
+      "step": 14479
+    },
+    {
+      "epoch": 1.1685901057218948,
+      "grad_norm": 0.6611022353172302,
+      "learning_rate": 3.565139242543081e-05,
+      "loss": 2.3992,
+      "step": 14480
+    },
+    {
+      "epoch": 1.168670809458478,
+      "grad_norm": 0.6844382882118225,
+      "learning_rate": 3.5639309023047306e-05,
+      "loss": 2.4345,
+      "step": 14481
+    },
+    {
+      "epoch": 1.168751513195061,
+      "grad_norm": 0.7557988166809082,
+      "learning_rate": 3.5627227224700464e-05,
+      "loss": 2.4454,
+      "step": 14482
+    },
+    {
+      "epoch": 1.1688322169316439,
+      "grad_norm": 0.6652555465698242,
+      "learning_rate": 3.5615147030691384e-05,
+      "loss": 2.3749,
+      "step": 14483
+    },
+    {
+      "epoch": 1.168912920668227,
+      "grad_norm": 0.6912989020347595,
+      "learning_rate": 3.56030684413212e-05,
+      "loss": 2.4737,
+      "step": 14484
+    },
+    {
+      "epoch": 1.16899362440481,
+      "grad_norm": 0.735103964805603,
+      "learning_rate": 3.559099145689083e-05,
+      "loss": 2.4098,
+      "step": 14485
+    },
+    {
+      "epoch": 1.169074328141393,
+      "grad_norm": 0.6873028874397278,
+      "learning_rate": 3.557891607770133e-05,
+      "loss": 2.4247,
+      "step": 14486
+    },
+    {
+      "epoch": 1.169155031877976,
+      "grad_norm": 0.7364680171012878,
+      "learning_rate": 3.556684230405367e-05,
+      "loss": 2.4314,
+      "step": 14487
+    },
+    {
+      "epoch": 1.169235735614559,
+      "grad_norm": 0.679122269153595,
+      "learning_rate": 3.55547701362487e-05,
+      "loss": 2.4196,
+      "step": 14488
+    },
+    {
+      "epoch": 1.169316439351142,
+      "grad_norm": 0.6783872246742249,
+      "learning_rate": 3.554269957458731e-05,
+      "loss": 2.4212,
+      "step": 14489
+    },
+    {
+      "epoch": 1.169397143087725,
+      "grad_norm": 0.7434942126274109,
+      "learning_rate": 3.553063061937034e-05,
+      "loss": 2.4139,
+      "step": 14490
+    },
+    {
+      "epoch": 1.1694778468243079,
+      "grad_norm": 0.6799852252006531,
+      "learning_rate": 3.55185632708986e-05,
+      "loss": 2.4252,
+      "step": 14491
+    },
+    {
+      "epoch": 1.169558550560891,
+      "grad_norm": 0.7040107250213623,
+      "learning_rate": 3.5506497529472795e-05,
+      "loss": 2.3937,
+      "step": 14492
+    },
+    {
+      "epoch": 1.169639254297474,
+      "grad_norm": 0.7350315451622009,
+      "learning_rate": 3.549443339539368e-05,
+      "loss": 2.4063,
+      "step": 14493
+    },
+    {
+      "epoch": 1.169719958034057,
+      "grad_norm": 0.694521963596344,
+      "learning_rate": 3.548237086896192e-05,
+      "loss": 2.4715,
+      "step": 14494
+    },
+    {
+      "epoch": 1.16980066177064,
+      "grad_norm": 0.6648221015930176,
+      "learning_rate": 3.5470309950478096e-05,
+      "loss": 2.4365,
+      "step": 14495
+    },
+    {
+      "epoch": 1.1698813655072229,
+      "grad_norm": 0.688024640083313,
+      "learning_rate": 3.545825064024284e-05,
+      "loss": 2.449,
+      "step": 14496
+    },
+    {
+      "epoch": 1.169962069243806,
+      "grad_norm": 0.6743311882019043,
+      "learning_rate": 3.544619293855672e-05,
+      "loss": 2.4283,
+      "step": 14497
+    },
+    {
+      "epoch": 1.170042772980389,
+      "grad_norm": 0.669119119644165,
+      "learning_rate": 3.543413684572019e-05,
+      "loss": 2.4363,
+      "step": 14498
+    },
+    {
+      "epoch": 1.170123476716972,
+      "grad_norm": 0.6998667120933533,
+      "learning_rate": 3.5422082362033745e-05,
+      "loss": 2.425,
+      "step": 14499
+    },
+    {
+      "epoch": 1.170204180453555,
+      "grad_norm": 0.7681630253791809,
+      "learning_rate": 3.5410029487797845e-05,
+      "loss": 2.4382,
+      "step": 14500
+    },
+    {
+      "epoch": 1.170284884190138,
+      "grad_norm": 0.6925049424171448,
+      "learning_rate": 3.539797822331279e-05,
+      "loss": 2.4261,
+      "step": 14501
+    },
+    {
+      "epoch": 1.170365587926721,
+      "grad_norm": 0.7145542502403259,
+      "learning_rate": 3.538592856887901e-05,
+      "loss": 2.4681,
+      "step": 14502
+    },
+    {
+      "epoch": 1.170446291663304,
+      "grad_norm": 0.6441611647605896,
+      "learning_rate": 3.537388052479684e-05,
+      "loss": 2.4187,
+      "step": 14503
+    },
+    {
+      "epoch": 1.1705269953998871,
+      "grad_norm": 0.6622560620307922,
+      "learning_rate": 3.5361834091366466e-05,
+      "loss": 2.4615,
+      "step": 14504
+    },
+    {
+      "epoch": 1.17060769913647,
+      "grad_norm": 0.6987677812576294,
+      "learning_rate": 3.5349789268888144e-05,
+      "loss": 2.413,
+      "step": 14505
+    },
+    {
+      "epoch": 1.170688402873053,
+      "grad_norm": 0.668358325958252,
+      "learning_rate": 3.533774605766207e-05,
+      "loss": 2.5146,
+      "step": 14506
+    },
+    {
+      "epoch": 1.170769106609636,
+      "grad_norm": 0.7514958381652832,
+      "learning_rate": 3.532570445798844e-05,
+      "loss": 2.4474,
+      "step": 14507
+    },
+    {
+      "epoch": 1.170849810346219,
+      "grad_norm": 0.6454465389251709,
+      "learning_rate": 3.5313664470167276e-05,
+      "loss": 2.3911,
+      "step": 14508
+    },
+    {
+      "epoch": 1.170930514082802,
+      "grad_norm": 0.6653602719306946,
+      "learning_rate": 3.5301626094498674e-05,
+      "loss": 2.4223,
+      "step": 14509
+    },
+    {
+      "epoch": 1.171011217819385,
+      "grad_norm": 0.6782815456390381,
+      "learning_rate": 3.5289589331282715e-05,
+      "loss": 2.457,
+      "step": 14510
+    },
+    {
+      "epoch": 1.171091921555968,
+      "grad_norm": 0.720973014831543,
+      "learning_rate": 3.527755418081932e-05,
+      "loss": 2.4541,
+      "step": 14511
+    },
+    {
+      "epoch": 1.171172625292551,
+      "grad_norm": 0.6300156712532043,
+      "learning_rate": 3.526552064340841e-05,
+      "loss": 2.4451,
+      "step": 14512
+    },
+    {
+      "epoch": 1.171253329029134,
+      "grad_norm": 0.7660964727401733,
+      "learning_rate": 3.5253488719350026e-05,
+      "loss": 2.5031,
+      "step": 14513
+    },
+    {
+      "epoch": 1.1713340327657171,
+      "grad_norm": 0.6931602358818054,
+      "learning_rate": 3.5241458408943905e-05,
+      "loss": 2.4249,
+      "step": 14514
+    },
+    {
+      "epoch": 1.1714147365023,
+      "grad_norm": 0.6863045692443848,
+      "learning_rate": 3.522942971248993e-05,
+      "loss": 2.4429,
+      "step": 14515
+    },
+    {
+      "epoch": 1.171495440238883,
+      "grad_norm": 0.6993531584739685,
+      "learning_rate": 3.521740263028791e-05,
+      "loss": 2.3864,
+      "step": 14516
+    },
+    {
+      "epoch": 1.1715761439754662,
+      "grad_norm": 0.807991087436676,
+      "learning_rate": 3.520537716263753e-05,
+      "loss": 2.459,
+      "step": 14517
+    },
+    {
+      "epoch": 1.171656847712049,
+      "grad_norm": 0.6722908020019531,
+      "learning_rate": 3.519335330983852e-05,
+      "loss": 2.4426,
+      "step": 14518
+    },
+    {
+      "epoch": 1.1717375514486321,
+      "grad_norm": 0.6934377551078796,
+      "learning_rate": 3.5181331072190585e-05,
+      "loss": 2.4326,
+      "step": 14519
+    },
+    {
+      "epoch": 1.171818255185215,
+      "grad_norm": 0.6532938480377197,
+      "learning_rate": 3.516931044999329e-05,
+      "loss": 2.3778,
+      "step": 14520
+    },
+    {
+      "epoch": 1.171898958921798,
+      "grad_norm": 0.6779183745384216,
+      "learning_rate": 3.5157291443546247e-05,
+      "loss": 2.4089,
+      "step": 14521
+    },
+    {
+      "epoch": 1.1719796626583812,
+      "grad_norm": 0.687005877494812,
+      "learning_rate": 3.514527405314899e-05,
+      "loss": 2.4669,
+      "step": 14522
+    },
+    {
+      "epoch": 1.172060366394964,
+      "grad_norm": 0.6804830431938171,
+      "learning_rate": 3.5133258279101045e-05,
+      "loss": 2.4789,
+      "step": 14523
+    },
+    {
+      "epoch": 1.1721410701315471,
+      "grad_norm": 0.8345538973808289,
+      "learning_rate": 3.512124412170187e-05,
+      "loss": 2.4506,
+      "step": 14524
+    },
+    {
+      "epoch": 1.17222177386813,
+      "grad_norm": 0.6571901440620422,
+      "learning_rate": 3.510923158125088e-05,
+      "loss": 2.4911,
+      "step": 14525
+    },
+    {
+      "epoch": 1.172302477604713,
+      "grad_norm": 0.6607047915458679,
+      "learning_rate": 3.5097220658047504e-05,
+      "loss": 2.4882,
+      "step": 14526
+    },
+    {
+      "epoch": 1.1723831813412962,
+      "grad_norm": 0.6883669495582581,
+      "learning_rate": 3.508521135239101e-05,
+      "loss": 2.4083,
+      "step": 14527
+    },
+    {
+      "epoch": 1.172463885077879,
+      "grad_norm": 0.6792941689491272,
+      "learning_rate": 3.5073203664580746e-05,
+      "loss": 2.368,
+      "step": 14528
+    },
+    {
+      "epoch": 1.172544588814462,
+      "grad_norm": 0.6675198674201965,
+      "learning_rate": 3.506119759491598e-05,
+      "loss": 2.4193,
+      "step": 14529
+    },
+    {
+      "epoch": 1.1726252925510452,
+      "grad_norm": 0.7267464399337769,
+      "learning_rate": 3.504919314369591e-05,
+      "loss": 2.3906,
+      "step": 14530
+    },
+    {
+      "epoch": 1.172705996287628,
+      "grad_norm": 0.6927710175514221,
+      "learning_rate": 3.503719031121973e-05,
+      "loss": 2.4082,
+      "step": 14531
+    },
+    {
+      "epoch": 1.1727867000242111,
+      "grad_norm": 0.7231000065803528,
+      "learning_rate": 3.502518909778656e-05,
+      "loss": 2.4845,
+      "step": 14532
+    },
+    {
+      "epoch": 1.1728674037607942,
+      "grad_norm": 0.7087520360946655,
+      "learning_rate": 3.5013189503695544e-05,
+      "loss": 2.4622,
+      "step": 14533
+    },
+    {
+      "epoch": 1.172948107497377,
+      "grad_norm": 0.6669846177101135,
+      "learning_rate": 3.5001191529245716e-05,
+      "loss": 2.4151,
+      "step": 14534
+    },
+    {
+      "epoch": 1.1730288112339602,
+      "grad_norm": 0.7338447570800781,
+      "learning_rate": 3.4989195174736134e-05,
+      "loss": 2.4274,
+      "step": 14535
+    },
+    {
+      "epoch": 1.173109514970543,
+      "grad_norm": 0.7032054662704468,
+      "learning_rate": 3.497720044046572e-05,
+      "loss": 2.4066,
+      "step": 14536
+    },
+    {
+      "epoch": 1.1731902187071261,
+      "grad_norm": 0.6571083068847656,
+      "learning_rate": 3.496520732673344e-05,
+      "loss": 2.4581,
+      "step": 14537
+    },
+    {
+      "epoch": 1.1732709224437092,
+      "grad_norm": 0.6618444919586182,
+      "learning_rate": 3.495321583383819e-05,
+      "loss": 2.3675,
+      "step": 14538
+    },
+    {
+      "epoch": 1.173351626180292,
+      "grad_norm": 0.6597652435302734,
+      "learning_rate": 3.4941225962078885e-05,
+      "loss": 2.416,
+      "step": 14539
+    },
+    {
+      "epoch": 1.1734323299168752,
+      "grad_norm": 0.682634711265564,
+      "learning_rate": 3.492923771175425e-05,
+      "loss": 2.5081,
+      "step": 14540
+    },
+    {
+      "epoch": 1.173513033653458,
+      "grad_norm": 0.7046132683753967,
+      "learning_rate": 3.49172510831631e-05,
+      "loss": 2.4439,
+      "step": 14541
+    },
+    {
+      "epoch": 1.1735937373900411,
+      "grad_norm": 0.6734833717346191,
+      "learning_rate": 3.4905266076604196e-05,
+      "loss": 2.4348,
+      "step": 14542
+    },
+    {
+      "epoch": 1.1736744411266242,
+      "grad_norm": 0.6624744534492493,
+      "learning_rate": 3.4893282692376214e-05,
+      "loss": 2.4364,
+      "step": 14543
+    },
+    {
+      "epoch": 1.173755144863207,
+      "grad_norm": 0.8425754308700562,
+      "learning_rate": 3.4881300930777815e-05,
+      "loss": 2.4803,
+      "step": 14544
+    },
+    {
+      "epoch": 1.1738358485997902,
+      "grad_norm": 0.6438888311386108,
+      "learning_rate": 3.486932079210766e-05,
+      "loss": 2.3973,
+      "step": 14545
+    },
+    {
+      "epoch": 1.1739165523363733,
+      "grad_norm": 0.650399923324585,
+      "learning_rate": 3.485734227666424e-05,
+      "loss": 2.4183,
+      "step": 14546
+    },
+    {
+      "epoch": 1.1739972560729561,
+      "grad_norm": 0.6857002973556519,
+      "learning_rate": 3.4845365384746144e-05,
+      "loss": 2.4061,
+      "step": 14547
+    },
+    {
+      "epoch": 1.1740779598095392,
+      "grad_norm": 0.6680994629859924,
+      "learning_rate": 3.483339011665189e-05,
+      "loss": 2.421,
+      "step": 14548
+    },
+    {
+      "epoch": 1.1741586635461223,
+      "grad_norm": 0.6440950632095337,
+      "learning_rate": 3.482141647267987e-05,
+      "loss": 2.3914,
+      "step": 14549
+    },
+    {
+      "epoch": 1.1742393672827052,
+      "grad_norm": 0.7329740524291992,
+      "learning_rate": 3.480944445312853e-05,
+      "loss": 2.4805,
+      "step": 14550
+    },
+    {
+      "epoch": 1.1743200710192883,
+      "grad_norm": 0.6848189234733582,
+      "learning_rate": 3.4797474058296245e-05,
+      "loss": 2.3611,
+      "step": 14551
+    },
+    {
+      "epoch": 1.1744007747558711,
+      "grad_norm": 0.6994072794914246,
+      "learning_rate": 3.478550528848134e-05,
+      "loss": 2.5106,
+      "step": 14552
+    },
+    {
+      "epoch": 1.1744814784924542,
+      "grad_norm": 0.6826444268226624,
+      "learning_rate": 3.477353814398212e-05,
+      "loss": 2.467,
+      "step": 14553
+    },
+    {
+      "epoch": 1.1745621822290373,
+      "grad_norm": 0.6658408045768738,
+      "learning_rate": 3.476157262509683e-05,
+      "loss": 2.423,
+      "step": 14554
+    },
+    {
+      "epoch": 1.1746428859656202,
+      "grad_norm": 0.6963697075843811,
+      "learning_rate": 3.474960873212372e-05,
+      "loss": 2.457,
+      "step": 14555
+    },
+    {
+      "epoch": 1.1747235897022033,
+      "grad_norm": 0.7574479579925537,
+      "learning_rate": 3.4737646465360894e-05,
+      "loss": 2.4292,
+      "step": 14556
+    },
+    {
+      "epoch": 1.1748042934387861,
+      "grad_norm": 0.7494931817054749,
+      "learning_rate": 3.472568582510652e-05,
+      "loss": 2.4395,
+      "step": 14557
+    },
+    {
+      "epoch": 1.1748849971753692,
+      "grad_norm": 0.7062687873840332,
+      "learning_rate": 3.471372681165872e-05,
+      "loss": 2.4561,
+      "step": 14558
+    },
+    {
+      "epoch": 1.1749657009119523,
+      "grad_norm": 0.6875349879264832,
+      "learning_rate": 3.4701769425315465e-05,
+      "loss": 2.4728,
+      "step": 14559
+    },
+    {
+      "epoch": 1.1750464046485352,
+      "grad_norm": 0.7009960412979126,
+      "learning_rate": 3.46898136663748e-05,
+      "loss": 2.5364,
+      "step": 14560
+    },
+    {
+      "epoch": 1.1751271083851182,
+      "grad_norm": 0.673791766166687,
+      "learning_rate": 3.467785953513475e-05,
+      "loss": 2.4611,
+      "step": 14561
+    },
+    {
+      "epoch": 1.1752078121217013,
+      "grad_norm": 0.7166882753372192,
+      "learning_rate": 3.4665907031893164e-05,
+      "loss": 2.4451,
+      "step": 14562
+    },
+    {
+      "epoch": 1.1752885158582842,
+      "grad_norm": 0.6868429780006409,
+      "learning_rate": 3.465395615694791e-05,
+      "loss": 2.4282,
+      "step": 14563
+    },
+    {
+      "epoch": 1.1753692195948673,
+      "grad_norm": 0.7212893962860107,
+      "learning_rate": 3.464200691059697e-05,
+      "loss": 2.4239,
+      "step": 14564
+    },
+    {
+      "epoch": 1.1754499233314502,
+      "grad_norm": 0.7213432192802429,
+      "learning_rate": 3.463005929313802e-05,
+      "loss": 2.4872,
+      "step": 14565
+    },
+    {
+      "epoch": 1.1755306270680332,
+      "grad_norm": 0.6805179119110107,
+      "learning_rate": 3.461811330486887e-05,
+      "loss": 2.4192,
+      "step": 14566
+    },
+    {
+      "epoch": 1.1756113308046163,
+      "grad_norm": 0.6746333241462708,
+      "learning_rate": 3.460616894608725e-05,
+      "loss": 2.3911,
+      "step": 14567
+    },
+    {
+      "epoch": 1.1756920345411992,
+      "grad_norm": 0.7388630509376526,
+      "learning_rate": 3.459422621709088e-05,
+      "loss": 2.4758,
+      "step": 14568
+    },
+    {
+      "epoch": 1.1757727382777823,
+      "grad_norm": 0.7730274200439453,
+      "learning_rate": 3.458228511817731e-05,
+      "loss": 2.4159,
+      "step": 14569
+    },
+    {
+      "epoch": 1.1758534420143651,
+      "grad_norm": 0.721075177192688,
+      "learning_rate": 3.457034564964422e-05,
+      "loss": 2.4673,
+      "step": 14570
+    },
+    {
+      "epoch": 1.1759341457509482,
+      "grad_norm": 0.6647645235061646,
+      "learning_rate": 3.4558407811789184e-05,
+      "loss": 2.395,
+      "step": 14571
+    },
+    {
+      "epoch": 1.1760148494875313,
+      "grad_norm": 0.7155466675758362,
+      "learning_rate": 3.454647160490965e-05,
+      "loss": 2.503,
+      "step": 14572
+    },
+    {
+      "epoch": 1.1760955532241142,
+      "grad_norm": 0.6789268851280212,
+      "learning_rate": 3.453453702930314e-05,
+      "loss": 2.401,
+      "step": 14573
+    },
+    {
+      "epoch": 1.1761762569606973,
+      "grad_norm": 0.7488093376159668,
+      "learning_rate": 3.4522604085267105e-05,
+      "loss": 2.4434,
+      "step": 14574
+    },
+    {
+      "epoch": 1.1762569606972804,
+      "grad_norm": 0.7954889535903931,
+      "learning_rate": 3.451067277309893e-05,
+      "loss": 2.5302,
+      "step": 14575
+    },
+    {
+      "epoch": 1.1763376644338632,
+      "grad_norm": 0.7008484601974487,
+      "learning_rate": 3.4498743093095975e-05,
+      "loss": 2.3935,
+      "step": 14576
+    },
+    {
+      "epoch": 1.1764183681704463,
+      "grad_norm": 0.6725437641143799,
+      "learning_rate": 3.448681504555561e-05,
+      "loss": 2.399,
+      "step": 14577
+    },
+    {
+      "epoch": 1.1764990719070294,
+      "grad_norm": 0.6778931617736816,
+      "learning_rate": 3.4474888630775026e-05,
+      "loss": 2.4178,
+      "step": 14578
+    },
+    {
+      "epoch": 1.1765797756436123,
+      "grad_norm": 0.7043762803077698,
+      "learning_rate": 3.44629638490515e-05,
+      "loss": 2.5581,
+      "step": 14579
+    },
+    {
+      "epoch": 1.1766604793801954,
+      "grad_norm": 0.6848085522651672,
+      "learning_rate": 3.445104070068227e-05,
+      "loss": 2.436,
+      "step": 14580
+    },
+    {
+      "epoch": 1.1767411831167782,
+      "grad_norm": 0.7504082322120667,
+      "learning_rate": 3.443911918596441e-05,
+      "loss": 2.4138,
+      "step": 14581
+    },
+    {
+      "epoch": 1.1768218868533613,
+      "grad_norm": 0.7441161870956421,
+      "learning_rate": 3.442719930519508e-05,
+      "loss": 2.4333,
+      "step": 14582
+    },
+    {
+      "epoch": 1.1769025905899444,
+      "grad_norm": 0.663894772529602,
+      "learning_rate": 3.4415281058671354e-05,
+      "loss": 2.4672,
+      "step": 14583
+    },
+    {
+      "epoch": 1.1769832943265273,
+      "grad_norm": 0.6814345121383667,
+      "learning_rate": 3.440336444669027e-05,
+      "loss": 2.4196,
+      "step": 14584
+    },
+    {
+      "epoch": 1.1770639980631104,
+      "grad_norm": 0.7566598057746887,
+      "learning_rate": 3.439144946954881e-05,
+      "loss": 2.4586,
+      "step": 14585
+    },
+    {
+      "epoch": 1.1771447017996932,
+      "grad_norm": 0.7324996590614319,
+      "learning_rate": 3.4379536127543934e-05,
+      "loss": 2.4286,
+      "step": 14586
+    },
+    {
+      "epoch": 1.1772254055362763,
+      "grad_norm": 0.6632608771324158,
+      "learning_rate": 3.436762442097259e-05,
+      "loss": 2.4713,
+      "step": 14587
+    },
+    {
+      "epoch": 1.1773061092728594,
+      "grad_norm": 0.7246156930923462,
+      "learning_rate": 3.4355714350131564e-05,
+      "loss": 2.4374,
+      "step": 14588
+    },
+    {
+      "epoch": 1.1773868130094423,
+      "grad_norm": 0.7096351981163025,
+      "learning_rate": 3.4343805915317737e-05,
+      "loss": 2.4649,
+      "step": 14589
+    },
+    {
+      "epoch": 1.1774675167460253,
+      "grad_norm": 0.7090620398521423,
+      "learning_rate": 3.433189911682793e-05,
+      "loss": 2.396,
+      "step": 14590
+    },
+    {
+      "epoch": 1.1775482204826084,
+      "grad_norm": 0.7782440185546875,
+      "learning_rate": 3.431999395495882e-05,
+      "loss": 2.4506,
+      "step": 14591
+    },
+    {
+      "epoch": 1.1776289242191913,
+      "grad_norm": 0.6933457851409912,
+      "learning_rate": 3.4308090430007155e-05,
+      "loss": 2.3985,
+      "step": 14592
+    },
+    {
+      "epoch": 1.1777096279557744,
+      "grad_norm": 0.6935414671897888,
+      "learning_rate": 3.429618854226959e-05,
+      "loss": 2.4372,
+      "step": 14593
+    },
+    {
+      "epoch": 1.1777903316923575,
+      "grad_norm": 0.6971156597137451,
+      "learning_rate": 3.428428829204276e-05,
+      "loss": 2.4837,
+      "step": 14594
+    },
+    {
+      "epoch": 1.1778710354289403,
+      "grad_norm": 0.6460022926330566,
+      "learning_rate": 3.427238967962325e-05,
+      "loss": 2.3742,
+      "step": 14595
+    },
+    {
+      "epoch": 1.1779517391655234,
+      "grad_norm": 0.6941941976547241,
+      "learning_rate": 3.426049270530763e-05,
+      "loss": 2.4706,
+      "step": 14596
+    },
+    {
+      "epoch": 1.1780324429021063,
+      "grad_norm": 0.7062166333198547,
+      "learning_rate": 3.424859736939236e-05,
+      "loss": 2.3893,
+      "step": 14597
+    },
+    {
+      "epoch": 1.1781131466386894,
+      "grad_norm": 0.6586433053016663,
+      "learning_rate": 3.42367036721739e-05,
+      "loss": 2.4385,
+      "step": 14598
+    },
+    {
+      "epoch": 1.1781938503752725,
+      "grad_norm": 0.6781242489814758,
+      "learning_rate": 3.422481161394869e-05,
+      "loss": 2.3876,
+      "step": 14599
+    },
+    {
+      "epoch": 1.1782745541118553,
+      "grad_norm": 0.710127592086792,
+      "learning_rate": 3.421292119501316e-05,
+      "loss": 2.4067,
+      "step": 14600
+    },
+    {
+      "epoch": 1.1783552578484384,
+      "grad_norm": 0.6856096982955933,
+      "learning_rate": 3.420103241566357e-05,
+      "loss": 2.4855,
+      "step": 14601
+    },
+    {
+      "epoch": 1.1784359615850213,
+      "grad_norm": 0.7173380851745605,
+      "learning_rate": 3.4189145276196245e-05,
+      "loss": 2.4871,
+      "step": 14602
+    },
+    {
+      "epoch": 1.1785166653216044,
+      "grad_norm": 0.6895382404327393,
+      "learning_rate": 3.417725977690745e-05,
+      "loss": 2.4066,
+      "step": 14603
+    },
+    {
+      "epoch": 1.1785973690581875,
+      "grad_norm": 0.7417690753936768,
+      "learning_rate": 3.416537591809341e-05,
+      "loss": 2.3779,
+      "step": 14604
+    },
+    {
+      "epoch": 1.1786780727947703,
+      "grad_norm": 0.7258411049842834,
+      "learning_rate": 3.4153493700050286e-05,
+      "loss": 2.4334,
+      "step": 14605
+    },
+    {
+      "epoch": 1.1787587765313534,
+      "grad_norm": 0.65704345703125,
+      "learning_rate": 3.414161312307427e-05,
+      "loss": 2.4531,
+      "step": 14606
+    },
+    {
+      "epoch": 1.1788394802679365,
+      "grad_norm": 0.6937118172645569,
+      "learning_rate": 3.4129734187461374e-05,
+      "loss": 2.4562,
+      "step": 14607
+    },
+    {
+      "epoch": 1.1789201840045194,
+      "grad_norm": 0.7331998348236084,
+      "learning_rate": 3.411785689350768e-05,
+      "loss": 2.4418,
+      "step": 14608
+    },
+    {
+      "epoch": 1.1790008877411025,
+      "grad_norm": 0.666582465171814,
+      "learning_rate": 3.410598124150924e-05,
+      "loss": 2.4154,
+      "step": 14609
+    },
+    {
+      "epoch": 1.1790815914776853,
+      "grad_norm": 0.6684321165084839,
+      "learning_rate": 3.409410723176197e-05,
+      "loss": 2.4155,
+      "step": 14610
+    },
+    {
+      "epoch": 1.1791622952142684,
+      "grad_norm": 0.6413382291793823,
+      "learning_rate": 3.408223486456184e-05,
+      "loss": 2.3924,
+      "step": 14611
+    },
+    {
+      "epoch": 1.1792429989508515,
+      "grad_norm": 0.7081305384635925,
+      "learning_rate": 3.407036414020475e-05,
+      "loss": 2.3811,
+      "step": 14612
+    },
+    {
+      "epoch": 1.1793237026874344,
+      "grad_norm": 0.7550063133239746,
+      "learning_rate": 3.405849505898645e-05,
+      "loss": 2.4425,
+      "step": 14613
+    },
+    {
+      "epoch": 1.1794044064240174,
+      "grad_norm": 0.677200198173523,
+      "learning_rate": 3.404662762120288e-05,
+      "loss": 2.5182,
+      "step": 14614
+    },
+    {
+      "epoch": 1.1794851101606003,
+      "grad_norm": 0.6829770803451538,
+      "learning_rate": 3.4034761827149745e-05,
+      "loss": 2.5068,
+      "step": 14615
+    },
+    {
+      "epoch": 1.1795658138971834,
+      "grad_norm": 0.7069409489631653,
+      "learning_rate": 3.4022897677122815e-05,
+      "loss": 2.4449,
+      "step": 14616
+    },
+    {
+      "epoch": 1.1796465176337665,
+      "grad_norm": 0.6604448556900024,
+      "learning_rate": 3.4011035171417696e-05,
+      "loss": 2.3996,
+      "step": 14617
+    },
+    {
+      "epoch": 1.1797272213703494,
+      "grad_norm": 0.6577324271202087,
+      "learning_rate": 3.3999174310330084e-05,
+      "loss": 2.4723,
+      "step": 14618
+    },
+    {
+      "epoch": 1.1798079251069324,
+      "grad_norm": 0.8159187436103821,
+      "learning_rate": 3.398731509415561e-05,
+      "loss": 2.4655,
+      "step": 14619
+    },
+    {
+      "epoch": 1.1798886288435155,
+      "grad_norm": 0.7170652747154236,
+      "learning_rate": 3.397545752318977e-05,
+      "loss": 2.5095,
+      "step": 14620
+    },
+    {
+      "epoch": 1.1799693325800984,
+      "grad_norm": 0.6865009665489197,
+      "learning_rate": 3.396360159772812e-05,
+      "loss": 2.4358,
+      "step": 14621
+    },
+    {
+      "epoch": 1.1800500363166815,
+      "grad_norm": 0.6485020518302917,
+      "learning_rate": 3.3951747318066175e-05,
+      "loss": 2.4576,
+      "step": 14622
+    },
+    {
+      "epoch": 1.1801307400532646,
+      "grad_norm": 0.6626582145690918,
+      "learning_rate": 3.39398946844993e-05,
+      "loss": 2.4824,
+      "step": 14623
+    },
+    {
+      "epoch": 1.1802114437898474,
+      "grad_norm": 0.718588650226593,
+      "learning_rate": 3.392804369732293e-05,
+      "loss": 2.4211,
+      "step": 14624
+    },
+    {
+      "epoch": 1.1802921475264305,
+      "grad_norm": 0.7449582815170288,
+      "learning_rate": 3.391619435683243e-05,
+      "loss": 2.444,
+      "step": 14625
+    },
+    {
+      "epoch": 1.1803728512630134,
+      "grad_norm": 0.6988492012023926,
+      "learning_rate": 3.3904346663323115e-05,
+      "loss": 2.4262,
+      "step": 14626
+    },
+    {
+      "epoch": 1.1804535549995965,
+      "grad_norm": 0.6779490113258362,
+      "learning_rate": 3.389250061709025e-05,
+      "loss": 2.4751,
+      "step": 14627
+    },
+    {
+      "epoch": 1.1805342587361796,
+      "grad_norm": 0.6883673667907715,
+      "learning_rate": 3.388065621842912e-05,
+      "loss": 2.4995,
+      "step": 14628
+    },
+    {
+      "epoch": 1.1806149624727624,
+      "grad_norm": 0.7112017273902893,
+      "learning_rate": 3.386881346763483e-05,
+      "loss": 2.4181,
+      "step": 14629
+    },
+    {
+      "epoch": 1.1806956662093455,
+      "grad_norm": 0.6960459351539612,
+      "learning_rate": 3.385697236500258e-05,
+      "loss": 2.4888,
+      "step": 14630
+    },
+    {
+      "epoch": 1.1807763699459284,
+      "grad_norm": 0.6874156594276428,
+      "learning_rate": 3.3845132910827484e-05,
+      "loss": 2.4175,
+      "step": 14631
+    },
+    {
+      "epoch": 1.1808570736825115,
+      "grad_norm": 0.7075642347335815,
+      "learning_rate": 3.383329510540463e-05,
+      "loss": 2.4315,
+      "step": 14632
+    },
+    {
+      "epoch": 1.1809377774190946,
+      "grad_norm": 0.674907386302948,
+      "learning_rate": 3.3821458949028995e-05,
+      "loss": 2.4216,
+      "step": 14633
+    },
+    {
+      "epoch": 1.1810184811556774,
+      "grad_norm": 0.7008463740348816,
+      "learning_rate": 3.380962444199559e-05,
+      "loss": 2.4114,
+      "step": 14634
+    },
+    {
+      "epoch": 1.1810991848922605,
+      "grad_norm": 0.6784217953681946,
+      "learning_rate": 3.379779158459937e-05,
+      "loss": 2.3663,
+      "step": 14635
+    },
+    {
+      "epoch": 1.1811798886288436,
+      "grad_norm": 0.7174829244613647,
+      "learning_rate": 3.378596037713525e-05,
+      "loss": 2.4582,
+      "step": 14636
+    },
+    {
+      "epoch": 1.1812605923654265,
+      "grad_norm": 0.7106035947799683,
+      "learning_rate": 3.3774130819898065e-05,
+      "loss": 2.5095,
+      "step": 14637
+    },
+    {
+      "epoch": 1.1813412961020096,
+      "grad_norm": 0.809107780456543,
+      "learning_rate": 3.3762302913182696e-05,
+      "loss": 2.4942,
+      "step": 14638
+    },
+    {
+      "epoch": 1.1814219998385926,
+      "grad_norm": 0.7150272727012634,
+      "learning_rate": 3.375047665728386e-05,
+      "loss": 2.378,
+      "step": 14639
+    },
+    {
+      "epoch": 1.1815027035751755,
+      "grad_norm": 0.7016271352767944,
+      "learning_rate": 3.373865205249632e-05,
+      "loss": 2.4393,
+      "step": 14640
+    },
+    {
+      "epoch": 1.1815834073117586,
+      "grad_norm": 0.6387282013893127,
+      "learning_rate": 3.372682909911481e-05,
+      "loss": 2.4399,
+      "step": 14641
+    },
+    {
+      "epoch": 1.1816641110483415,
+      "grad_norm": 0.834181010723114,
+      "learning_rate": 3.371500779743393e-05,
+      "loss": 2.4312,
+      "step": 14642
+    },
+    {
+      "epoch": 1.1817448147849245,
+      "grad_norm": 0.6690472960472107,
+      "learning_rate": 3.370318814774832e-05,
+      "loss": 2.407,
+      "step": 14643
+    },
+    {
+      "epoch": 1.1818255185215076,
+      "grad_norm": 0.6594302654266357,
+      "learning_rate": 3.369137015035256e-05,
+      "loss": 2.4275,
+      "step": 14644
+    },
+    {
+      "epoch": 1.1819062222580905,
+      "grad_norm": 0.7284699082374573,
+      "learning_rate": 3.3679553805541194e-05,
+      "loss": 2.3981,
+      "step": 14645
+    },
+    {
+      "epoch": 1.1819869259946736,
+      "grad_norm": 0.7109572291374207,
+      "learning_rate": 3.366773911360871e-05,
+      "loss": 2.4345,
+      "step": 14646
+    },
+    {
+      "epoch": 1.1820676297312565,
+      "grad_norm": 0.6874241828918457,
+      "learning_rate": 3.3655926074849566e-05,
+      "loss": 2.4488,
+      "step": 14647
+    },
+    {
+      "epoch": 1.1821483334678395,
+      "grad_norm": 0.6698973178863525,
+      "learning_rate": 3.364411468955819e-05,
+      "loss": 2.42,
+      "step": 14648
+    },
+    {
+      "epoch": 1.1822290372044226,
+      "grad_norm": 0.7816089391708374,
+      "learning_rate": 3.3632304958028915e-05,
+      "loss": 2.4638,
+      "step": 14649
+    },
+    {
+      "epoch": 1.1823097409410055,
+      "grad_norm": 0.6718220710754395,
+      "learning_rate": 3.3620496880556075e-05,
+      "loss": 2.413,
+      "step": 14650
+    },
+    {
+      "epoch": 1.1823904446775886,
+      "grad_norm": 0.753463089466095,
+      "learning_rate": 3.360869045743401e-05,
+      "loss": 2.3772,
+      "step": 14651
+    },
+    {
+      "epoch": 1.1824711484141717,
+      "grad_norm": 0.7031456828117371,
+      "learning_rate": 3.359688568895689e-05,
+      "loss": 2.4198,
+      "step": 14652
+    },
+    {
+      "epoch": 1.1825518521507545,
+      "grad_norm": 0.7857323288917542,
+      "learning_rate": 3.358508257541897e-05,
+      "loss": 2.4223,
+      "step": 14653
+    },
+    {
+      "epoch": 1.1826325558873376,
+      "grad_norm": 0.7779297828674316,
+      "learning_rate": 3.357328111711439e-05,
+      "loss": 2.5266,
+      "step": 14654
+    },
+    {
+      "epoch": 1.1827132596239207,
+      "grad_norm": 0.7382386326789856,
+      "learning_rate": 3.356148131433728e-05,
+      "loss": 2.4673,
+      "step": 14655
+    },
+    {
+      "epoch": 1.1827939633605036,
+      "grad_norm": 0.7868054509162903,
+      "learning_rate": 3.354968316738174e-05,
+      "loss": 2.4285,
+      "step": 14656
+    },
+    {
+      "epoch": 1.1828746670970867,
+      "grad_norm": 0.7007591724395752,
+      "learning_rate": 3.353788667654183e-05,
+      "loss": 2.4054,
+      "step": 14657
+    },
+    {
+      "epoch": 1.1829553708336695,
+      "grad_norm": 0.6627741456031799,
+      "learning_rate": 3.352609184211148e-05,
+      "loss": 2.4224,
+      "step": 14658
+    },
+    {
+      "epoch": 1.1830360745702526,
+      "grad_norm": 0.6865360736846924,
+      "learning_rate": 3.351429866438469e-05,
+      "loss": 2.4084,
+      "step": 14659
+    },
+    {
+      "epoch": 1.1831167783068357,
+      "grad_norm": 0.7572095990180969,
+      "learning_rate": 3.3502507143655404e-05,
+      "loss": 2.4339,
+      "step": 14660
+    },
+    {
+      "epoch": 1.1831974820434186,
+      "grad_norm": 0.6907969117164612,
+      "learning_rate": 3.349071728021743e-05,
+      "loss": 2.4578,
+      "step": 14661
+    },
+    {
+      "epoch": 1.1832781857800017,
+      "grad_norm": 0.6618743538856506,
+      "learning_rate": 3.347892907436465e-05,
+      "loss": 2.4131,
+      "step": 14662
+    },
+    {
+      "epoch": 1.1833588895165845,
+      "grad_norm": 0.777159571647644,
+      "learning_rate": 3.346714252639084e-05,
+      "loss": 2.419,
+      "step": 14663
+    },
+    {
+      "epoch": 1.1834395932531676,
+      "grad_norm": 0.666344165802002,
+      "learning_rate": 3.345535763658975e-05,
+      "loss": 2.4155,
+      "step": 14664
+    },
+    {
+      "epoch": 1.1835202969897507,
+      "grad_norm": 0.708848774433136,
+      "learning_rate": 3.3443574405255095e-05,
+      "loss": 2.4794,
+      "step": 14665
+    },
+    {
+      "epoch": 1.1836010007263336,
+      "grad_norm": 0.7247438430786133,
+      "learning_rate": 3.3431792832680555e-05,
+      "loss": 2.4445,
+      "step": 14666
+    },
+    {
+      "epoch": 1.1836817044629167,
+      "grad_norm": 0.6870034337043762,
+      "learning_rate": 3.342001291915978e-05,
+      "loss": 2.4309,
+      "step": 14667
+    },
+    {
+      "epoch": 1.1837624081994997,
+      "grad_norm": 0.7088049650192261,
+      "learning_rate": 3.340823466498629e-05,
+      "loss": 2.4456,
+      "step": 14668
+    },
+    {
+      "epoch": 1.1838431119360826,
+      "grad_norm": 0.695148229598999,
+      "learning_rate": 3.3396458070453676e-05,
+      "loss": 2.4018,
+      "step": 14669
+    },
+    {
+      "epoch": 1.1839238156726657,
+      "grad_norm": 0.7947117686271667,
+      "learning_rate": 3.3384683135855444e-05,
+      "loss": 2.4099,
+      "step": 14670
+    },
+    {
+      "epoch": 1.1840045194092486,
+      "grad_norm": 0.7268195748329163,
+      "learning_rate": 3.337290986148502e-05,
+      "loss": 2.3955,
+      "step": 14671
+    },
+    {
+      "epoch": 1.1840852231458316,
+      "grad_norm": 0.6932024955749512,
+      "learning_rate": 3.336113824763585e-05,
+      "loss": 2.4046,
+      "step": 14672
+    },
+    {
+      "epoch": 1.1841659268824147,
+      "grad_norm": 0.7408114671707153,
+      "learning_rate": 3.3349368294601334e-05,
+      "loss": 2.4186,
+      "step": 14673
+    },
+    {
+      "epoch": 1.1842466306189976,
+      "grad_norm": 0.6678428053855896,
+      "learning_rate": 3.3337600002674765e-05,
+      "loss": 2.4324,
+      "step": 14674
+    },
+    {
+      "epoch": 1.1843273343555807,
+      "grad_norm": 0.7221381664276123,
+      "learning_rate": 3.3325833372149416e-05,
+      "loss": 2.4474,
+      "step": 14675
+    },
+    {
+      "epoch": 1.1844080380921636,
+      "grad_norm": 0.6971224546432495,
+      "learning_rate": 3.3314068403318654e-05,
+      "loss": 2.4197,
+      "step": 14676
+    },
+    {
+      "epoch": 1.1844887418287466,
+      "grad_norm": 0.65053391456604,
+      "learning_rate": 3.3302305096475604e-05,
+      "loss": 2.4169,
+      "step": 14677
+    },
+    {
+      "epoch": 1.1845694455653297,
+      "grad_norm": 0.7231155633926392,
+      "learning_rate": 3.3290543451913457e-05,
+      "loss": 2.4222,
+      "step": 14678
+    },
+    {
+      "epoch": 1.1846501493019126,
+      "grad_norm": 0.6458824872970581,
+      "learning_rate": 3.3278783469925345e-05,
+      "loss": 2.422,
+      "step": 14679
+    },
+    {
+      "epoch": 1.1847308530384957,
+      "grad_norm": 0.6783488392829895,
+      "learning_rate": 3.32670251508044e-05,
+      "loss": 2.4231,
+      "step": 14680
+    },
+    {
+      "epoch": 1.1848115567750788,
+      "grad_norm": 0.6742293238639832,
+      "learning_rate": 3.3255268494843586e-05,
+      "loss": 2.409,
+      "step": 14681
+    },
+    {
+      "epoch": 1.1848922605116616,
+      "grad_norm": 0.7455186247825623,
+      "learning_rate": 3.3243513502335956e-05,
+      "loss": 2.4121,
+      "step": 14682
+    },
+    {
+      "epoch": 1.1849729642482447,
+      "grad_norm": 0.7042234539985657,
+      "learning_rate": 3.323176017357451e-05,
+      "loss": 2.4574,
+      "step": 14683
+    },
+    {
+      "epoch": 1.1850536679848278,
+      "grad_norm": 0.7897992134094238,
+      "learning_rate": 3.3220008508852094e-05,
+      "loss": 2.4796,
+      "step": 14684
+    },
+    {
+      "epoch": 1.1851343717214107,
+      "grad_norm": 0.6894058585166931,
+      "learning_rate": 3.3208258508461644e-05,
+      "loss": 2.4125,
+      "step": 14685
+    },
+    {
+      "epoch": 1.1852150754579938,
+      "grad_norm": 0.7574072480201721,
+      "learning_rate": 3.319651017269597e-05,
+      "loss": 2.4714,
+      "step": 14686
+    },
+    {
+      "epoch": 1.1852957791945766,
+      "grad_norm": 0.7457531094551086,
+      "learning_rate": 3.3184763501847905e-05,
+      "loss": 2.4793,
+      "step": 14687
+    },
+    {
+      "epoch": 1.1853764829311597,
+      "grad_norm": 0.6819709539413452,
+      "learning_rate": 3.317301849621018e-05,
+      "loss": 2.4563,
+      "step": 14688
+    },
+    {
+      "epoch": 1.1854571866677428,
+      "grad_norm": 0.6998026371002197,
+      "learning_rate": 3.316127515607555e-05,
+      "loss": 2.4548,
+      "step": 14689
+    },
+    {
+      "epoch": 1.1855378904043257,
+      "grad_norm": 0.7148768305778503,
+      "learning_rate": 3.314953348173664e-05,
+      "loss": 2.4897,
+      "step": 14690
+    },
+    {
+      "epoch": 1.1856185941409088,
+      "grad_norm": 0.6581987738609314,
+      "learning_rate": 3.31377934734861e-05,
+      "loss": 2.4683,
+      "step": 14691
+    },
+    {
+      "epoch": 1.1856992978774916,
+      "grad_norm": 0.7493093609809875,
+      "learning_rate": 3.312605513161653e-05,
+      "loss": 2.4564,
+      "step": 14692
+    },
+    {
+      "epoch": 1.1857800016140747,
+      "grad_norm": 0.7095562219619751,
+      "learning_rate": 3.311431845642051e-05,
+      "loss": 2.4595,
+      "step": 14693
+    },
+    {
+      "epoch": 1.1858607053506578,
+      "grad_norm": 0.8045323491096497,
+      "learning_rate": 3.310258344819047e-05,
+      "loss": 2.5044,
+      "step": 14694
+    },
+    {
+      "epoch": 1.1859414090872407,
+      "grad_norm": 0.7381219267845154,
+      "learning_rate": 3.3090850107218943e-05,
+      "loss": 2.415,
+      "step": 14695
+    },
+    {
+      "epoch": 1.1860221128238237,
+      "grad_norm": 0.6859883069992065,
+      "learning_rate": 3.307911843379832e-05,
+      "loss": 2.4314,
+      "step": 14696
+    },
+    {
+      "epoch": 1.1861028165604068,
+      "grad_norm": 0.7084196209907532,
+      "learning_rate": 3.306738842822099e-05,
+      "loss": 2.4404,
+      "step": 14697
+    },
+    {
+      "epoch": 1.1861835202969897,
+      "grad_norm": 0.6964806318283081,
+      "learning_rate": 3.305566009077932e-05,
+      "loss": 2.4391,
+      "step": 14698
+    },
+    {
+      "epoch": 1.1862642240335728,
+      "grad_norm": 0.7272049188613892,
+      "learning_rate": 3.304393342176562e-05,
+      "loss": 2.4395,
+      "step": 14699
+    },
+    {
+      "epoch": 1.1863449277701559,
+      "grad_norm": 0.6651458144187927,
+      "learning_rate": 3.303220842147209e-05,
+      "loss": 2.4059,
+      "step": 14700
+    },
+    {
+      "epoch": 1.1864256315067387,
+      "grad_norm": 0.7599130868911743,
+      "learning_rate": 3.302048509019099e-05,
+      "loss": 2.5044,
+      "step": 14701
+    },
+    {
+      "epoch": 1.1865063352433218,
+      "grad_norm": 0.6694391965866089,
+      "learning_rate": 3.3008763428214505e-05,
+      "loss": 2.4817,
+      "step": 14702
+    },
+    {
+      "epoch": 1.1865870389799047,
+      "grad_norm": 0.7176856398582458,
+      "learning_rate": 3.299704343583473e-05,
+      "loss": 2.4702,
+      "step": 14703
+    },
+    {
+      "epoch": 1.1866677427164878,
+      "grad_norm": 0.7133145332336426,
+      "learning_rate": 3.298532511334378e-05,
+      "loss": 2.4685,
+      "step": 14704
+    },
+    {
+      "epoch": 1.1867484464530709,
+      "grad_norm": 0.7170277833938599,
+      "learning_rate": 3.297360846103371e-05,
+      "loss": 2.4203,
+      "step": 14705
+    },
+    {
+      "epoch": 1.1868291501896537,
+      "grad_norm": 0.6853376626968384,
+      "learning_rate": 3.296189347919652e-05,
+      "loss": 2.4067,
+      "step": 14706
+    },
+    {
+      "epoch": 1.1869098539262368,
+      "grad_norm": 0.7269156575202942,
+      "learning_rate": 3.2950180168124175e-05,
+      "loss": 2.4211,
+      "step": 14707
+    },
+    {
+      "epoch": 1.1869905576628197,
+      "grad_norm": 0.8649005889892578,
+      "learning_rate": 3.2938468528108626e-05,
+      "loss": 2.4611,
+      "step": 14708
+    },
+    {
+      "epoch": 1.1870712613994028,
+      "grad_norm": 0.7256221771240234,
+      "learning_rate": 3.292675855944177e-05,
+      "loss": 2.4618,
+      "step": 14709
+    },
+    {
+      "epoch": 1.1871519651359859,
+      "grad_norm": 0.6854279637336731,
+      "learning_rate": 3.291505026241539e-05,
+      "loss": 2.4466,
+      "step": 14710
+    },
+    {
+      "epoch": 1.1872326688725687,
+      "grad_norm": 0.7182712554931641,
+      "learning_rate": 3.2903343637321316e-05,
+      "loss": 2.4847,
+      "step": 14711
+    },
+    {
+      "epoch": 1.1873133726091518,
+      "grad_norm": 0.6795300841331482,
+      "learning_rate": 3.289163868445134e-05,
+      "loss": 2.4407,
+      "step": 14712
+    },
+    {
+      "epoch": 1.187394076345735,
+      "grad_norm": 0.685146689414978,
+      "learning_rate": 3.287993540409713e-05,
+      "loss": 2.4537,
+      "step": 14713
+    },
+    {
+      "epoch": 1.1874747800823178,
+      "grad_norm": 0.7891005873680115,
+      "learning_rate": 3.2868233796550375e-05,
+      "loss": 2.4085,
+      "step": 14714
+    },
+    {
+      "epoch": 1.1875554838189009,
+      "grad_norm": 0.6521769762039185,
+      "learning_rate": 3.2856533862102724e-05,
+      "loss": 2.4174,
+      "step": 14715
+    },
+    {
+      "epoch": 1.1876361875554837,
+      "grad_norm": 0.7486612200737,
+      "learning_rate": 3.284483560104575e-05,
+      "loss": 2.4072,
+      "step": 14716
+    },
+    {
+      "epoch": 1.1877168912920668,
+      "grad_norm": 0.6895913481712341,
+      "learning_rate": 3.283313901367103e-05,
+      "loss": 2.4398,
+      "step": 14717
+    },
+    {
+      "epoch": 1.18779759502865,
+      "grad_norm": 0.6595678329467773,
+      "learning_rate": 3.282144410027009e-05,
+      "loss": 2.4407,
+      "step": 14718
+    },
+    {
+      "epoch": 1.1878782987652328,
+      "grad_norm": 0.7724249958992004,
+      "learning_rate": 3.280975086113435e-05,
+      "loss": 2.464,
+      "step": 14719
+    },
+    {
+      "epoch": 1.1879590025018159,
+      "grad_norm": 0.659472644329071,
+      "learning_rate": 3.279805929655524e-05,
+      "loss": 2.4774,
+      "step": 14720
+    },
+    {
+      "epoch": 1.1880397062383987,
+      "grad_norm": 0.7187919020652771,
+      "learning_rate": 3.27863694068242e-05,
+      "loss": 2.4767,
+      "step": 14721
+    },
+    {
+      "epoch": 1.1881204099749818,
+      "grad_norm": 0.7740198373794556,
+      "learning_rate": 3.2774681192232506e-05,
+      "loss": 2.4762,
+      "step": 14722
+    },
+    {
+      "epoch": 1.188201113711565,
+      "grad_norm": 0.700591504573822,
+      "learning_rate": 3.2762994653071464e-05,
+      "loss": 2.448,
+      "step": 14723
+    },
+    {
+      "epoch": 1.1882818174481478,
+      "grad_norm": 0.7168558239936829,
+      "learning_rate": 3.275130978963237e-05,
+      "loss": 2.4084,
+      "step": 14724
+    },
+    {
+      "epoch": 1.1883625211847308,
+      "grad_norm": 0.8039551973342896,
+      "learning_rate": 3.273962660220646e-05,
+      "loss": 2.3849,
+      "step": 14725
+    },
+    {
+      "epoch": 1.188443224921314,
+      "grad_norm": 0.6453016400337219,
+      "learning_rate": 3.27279450910848e-05,
+      "loss": 2.3856,
+      "step": 14726
+    },
+    {
+      "epoch": 1.1885239286578968,
+      "grad_norm": 0.7194651365280151,
+      "learning_rate": 3.2716265256558644e-05,
+      "loss": 2.4337,
+      "step": 14727
+    },
+    {
+      "epoch": 1.1886046323944799,
+      "grad_norm": 0.7298597097396851,
+      "learning_rate": 3.270458709891906e-05,
+      "loss": 2.4491,
+      "step": 14728
+    },
+    {
+      "epoch": 1.188685336131063,
+      "grad_norm": 0.7127524614334106,
+      "learning_rate": 3.269291061845705e-05,
+      "loss": 2.4319,
+      "step": 14729
+    },
+    {
+      "epoch": 1.1887660398676458,
+      "grad_norm": 0.6782705783843994,
+      "learning_rate": 3.2681235815463654e-05,
+      "loss": 2.4375,
+      "step": 14730
+    },
+    {
+      "epoch": 1.188846743604229,
+      "grad_norm": 0.7418326735496521,
+      "learning_rate": 3.266956269022987e-05,
+      "loss": 2.4149,
+      "step": 14731
+    },
+    {
+      "epoch": 1.1889274473408118,
+      "grad_norm": 0.7442455291748047,
+      "learning_rate": 3.265789124304654e-05,
+      "loss": 2.3935,
+      "step": 14732
+    },
+    {
+      "epoch": 1.1890081510773949,
+      "grad_norm": 0.7238253951072693,
+      "learning_rate": 3.264622147420461e-05,
+      "loss": 2.4592,
+      "step": 14733
+    },
+    {
+      "epoch": 1.189088854813978,
+      "grad_norm": 0.6488127708435059,
+      "learning_rate": 3.2634553383994925e-05,
+      "loss": 2.3468,
+      "step": 14734
+    },
+    {
+      "epoch": 1.1891695585505608,
+      "grad_norm": 0.7182446718215942,
+      "learning_rate": 3.2622886972708246e-05,
+      "loss": 2.4457,
+      "step": 14735
+    },
+    {
+      "epoch": 1.189250262287144,
+      "grad_norm": 0.6885523796081543,
+      "learning_rate": 3.261122224063534e-05,
+      "loss": 2.3943,
+      "step": 14736
+    },
+    {
+      "epoch": 1.1893309660237268,
+      "grad_norm": 0.653367817401886,
+      "learning_rate": 3.259955918806693e-05,
+      "loss": 2.4188,
+      "step": 14737
+    },
+    {
+      "epoch": 1.1894116697603099,
+      "grad_norm": 0.6968675851821899,
+      "learning_rate": 3.2587897815293686e-05,
+      "loss": 2.4276,
+      "step": 14738
+    },
+    {
+      "epoch": 1.189492373496893,
+      "grad_norm": 0.6827409267425537,
+      "learning_rate": 3.257623812260626e-05,
+      "loss": 2.4417,
+      "step": 14739
+    },
+    {
+      "epoch": 1.1895730772334758,
+      "grad_norm": 0.6807438731193542,
+      "learning_rate": 3.256458011029523e-05,
+      "loss": 2.4495,
+      "step": 14740
+    },
+    {
+      "epoch": 1.189653780970059,
+      "grad_norm": 0.6692882180213928,
+      "learning_rate": 3.255292377865116e-05,
+      "loss": 2.3789,
+      "step": 14741
+    },
+    {
+      "epoch": 1.189734484706642,
+      "grad_norm": 0.6581685543060303,
+      "learning_rate": 3.2541269127964515e-05,
+      "loss": 2.4073,
+      "step": 14742
+    },
+    {
+      "epoch": 1.1898151884432249,
+      "grad_norm": 0.6458544731140137,
+      "learning_rate": 3.252961615852578e-05,
+      "loss": 2.4657,
+      "step": 14743
+    },
+    {
+      "epoch": 1.189895892179808,
+      "grad_norm": 0.6971322298049927,
+      "learning_rate": 3.251796487062541e-05,
+      "loss": 2.4404,
+      "step": 14744
+    },
+    {
+      "epoch": 1.189976595916391,
+      "grad_norm": 0.6770374178886414,
+      "learning_rate": 3.2506315264553724e-05,
+      "loss": 2.4329,
+      "step": 14745
+    },
+    {
+      "epoch": 1.190057299652974,
+      "grad_norm": 0.7634715437889099,
+      "learning_rate": 3.2494667340601085e-05,
+      "loss": 2.4234,
+      "step": 14746
+    },
+    {
+      "epoch": 1.190138003389557,
+      "grad_norm": 0.7717967629432678,
+      "learning_rate": 3.24830210990578e-05,
+      "loss": 2.5009,
+      "step": 14747
+    },
+    {
+      "epoch": 1.1902187071261399,
+      "grad_norm": 0.7133559584617615,
+      "learning_rate": 3.2471376540214124e-05,
+      "loss": 2.4272,
+      "step": 14748
+    },
+    {
+      "epoch": 1.190299410862723,
+      "grad_norm": 0.7273291349411011,
+      "learning_rate": 3.245973366436027e-05,
+      "loss": 2.4174,
+      "step": 14749
+    },
+    {
+      "epoch": 1.190380114599306,
+      "grad_norm": 0.6955052614212036,
+      "learning_rate": 3.244809247178643e-05,
+      "loss": 2.3605,
+      "step": 14750
+    },
+    {
+      "epoch": 1.190460818335889,
+      "grad_norm": 0.7072615027427673,
+      "learning_rate": 3.2436452962782685e-05,
+      "loss": 2.4897,
+      "step": 14751
+    },
+    {
+      "epoch": 1.190541522072472,
+      "grad_norm": 0.7095344662666321,
+      "learning_rate": 3.242481513763913e-05,
+      "loss": 2.4172,
+      "step": 14752
+    },
+    {
+      "epoch": 1.1906222258090549,
+      "grad_norm": 0.7260944247245789,
+      "learning_rate": 3.2413178996645864e-05,
+      "loss": 2.4272,
+      "step": 14753
+    },
+    {
+      "epoch": 1.190702929545638,
+      "grad_norm": 0.6601141691207886,
+      "learning_rate": 3.2401544540092824e-05,
+      "loss": 2.4072,
+      "step": 14754
+    },
+    {
+      "epoch": 1.190783633282221,
+      "grad_norm": 0.6684936881065369,
+      "learning_rate": 3.238991176827e-05,
+      "loss": 2.3968,
+      "step": 14755
+    },
+    {
+      "epoch": 1.190864337018804,
+      "grad_norm": 0.7264483571052551,
+      "learning_rate": 3.23782806814673e-05,
+      "loss": 2.4263,
+      "step": 14756
+    },
+    {
+      "epoch": 1.190945040755387,
+      "grad_norm": 0.6927621960639954,
+      "learning_rate": 3.2366651279974614e-05,
+      "loss": 2.4495,
+      "step": 14757
+    },
+    {
+      "epoch": 1.19102574449197,
+      "grad_norm": 0.7007272243499756,
+      "learning_rate": 3.2355023564081775e-05,
+      "loss": 2.4373,
+      "step": 14758
+    },
+    {
+      "epoch": 1.191106448228553,
+      "grad_norm": 0.6756663918495178,
+      "learning_rate": 3.234339753407857e-05,
+      "loss": 2.4148,
+      "step": 14759
+    },
+    {
+      "epoch": 1.191187151965136,
+      "grad_norm": 0.6741094589233398,
+      "learning_rate": 3.233177319025479e-05,
+      "loss": 2.3976,
+      "step": 14760
+    },
+    {
+      "epoch": 1.1912678557017191,
+      "grad_norm": 0.7098578810691833,
+      "learning_rate": 3.2320150532900085e-05,
+      "loss": 2.4326,
+      "step": 14761
+    },
+    {
+      "epoch": 1.191348559438302,
+      "grad_norm": 0.750271737575531,
+      "learning_rate": 3.230852956230413e-05,
+      "loss": 2.4766,
+      "step": 14762
+    },
+    {
+      "epoch": 1.191429263174885,
+      "grad_norm": 0.68764728307724,
+      "learning_rate": 3.229691027875661e-05,
+      "loss": 2.4128,
+      "step": 14763
+    },
+    {
+      "epoch": 1.191509966911468,
+      "grad_norm": 0.656295657157898,
+      "learning_rate": 3.228529268254702e-05,
+      "loss": 2.3928,
+      "step": 14764
+    },
+    {
+      "epoch": 1.191590670648051,
+      "grad_norm": 0.6690353155136108,
+      "learning_rate": 3.2273676773964955e-05,
+      "loss": 2.408,
+      "step": 14765
+    },
+    {
+      "epoch": 1.1916713743846339,
+      "grad_norm": 0.8111640214920044,
+      "learning_rate": 3.22620625532999e-05,
+      "loss": 2.4644,
+      "step": 14766
+    },
+    {
+      "epoch": 1.191752078121217,
+      "grad_norm": 0.7329768538475037,
+      "learning_rate": 3.2250450020841316e-05,
+      "loss": 2.4235,
+      "step": 14767
+    },
+    {
+      "epoch": 1.1918327818578,
+      "grad_norm": 0.6902688145637512,
+      "learning_rate": 3.223883917687861e-05,
+      "loss": 2.3883,
+      "step": 14768
+    },
+    {
+      "epoch": 1.191913485594383,
+      "grad_norm": 0.797249972820282,
+      "learning_rate": 3.2227230021701205e-05,
+      "loss": 2.523,
+      "step": 14769
+    },
+    {
+      "epoch": 1.191994189330966,
+      "grad_norm": 0.6294408440589905,
+      "learning_rate": 3.221562255559834e-05,
+      "loss": 2.4156,
+      "step": 14770
+    },
+    {
+      "epoch": 1.192074893067549,
+      "grad_norm": 0.7326164245605469,
+      "learning_rate": 3.220401677885936e-05,
+      "loss": 2.3828,
+      "step": 14771
+    },
+    {
+      "epoch": 1.192155596804132,
+      "grad_norm": 0.783747673034668,
+      "learning_rate": 3.219241269177351e-05,
+      "loss": 2.4321,
+      "step": 14772
+    },
+    {
+      "epoch": 1.192236300540715,
+      "grad_norm": 0.7415335178375244,
+      "learning_rate": 3.2180810294630005e-05,
+      "loss": 2.4446,
+      "step": 14773
+    },
+    {
+      "epoch": 1.1923170042772981,
+      "grad_norm": 0.7125591039657593,
+      "learning_rate": 3.2169209587717966e-05,
+      "loss": 2.3914,
+      "step": 14774
+    },
+    {
+      "epoch": 1.192397708013881,
+      "grad_norm": 0.6714075207710266,
+      "learning_rate": 3.215761057132652e-05,
+      "loss": 2.3918,
+      "step": 14775
+    },
+    {
+      "epoch": 1.192478411750464,
+      "grad_norm": 0.7147830724716187,
+      "learning_rate": 3.214601324574481e-05,
+      "loss": 2.4389,
+      "step": 14776
+    },
+    {
+      "epoch": 1.192559115487047,
+      "grad_norm": 0.6780480146408081,
+      "learning_rate": 3.2134417611261755e-05,
+      "loss": 2.4119,
+      "step": 14777
+    },
+    {
+      "epoch": 1.19263981922363,
+      "grad_norm": 0.7473881840705872,
+      "learning_rate": 3.212282366816645e-05,
+      "loss": 2.4547,
+      "step": 14778
+    },
+    {
+      "epoch": 1.1927205229602131,
+      "grad_norm": 0.7418377995491028,
+      "learning_rate": 3.211123141674784e-05,
+      "loss": 2.4156,
+      "step": 14779
+    },
+    {
+      "epoch": 1.192801226696796,
+      "grad_norm": 0.687524139881134,
+      "learning_rate": 3.209964085729477e-05,
+      "loss": 2.4309,
+      "step": 14780
+    },
+    {
+      "epoch": 1.192881930433379,
+      "grad_norm": 0.6965883374214172,
+      "learning_rate": 3.208805199009615e-05,
+      "loss": 2.4028,
+      "step": 14781
+    },
+    {
+      "epoch": 1.192962634169962,
+      "grad_norm": 0.7024682760238647,
+      "learning_rate": 3.207646481544082e-05,
+      "loss": 2.4482,
+      "step": 14782
+    },
+    {
+      "epoch": 1.193043337906545,
+      "grad_norm": 0.6835834383964539,
+      "learning_rate": 3.2064879333617514e-05,
+      "loss": 2.3898,
+      "step": 14783
+    },
+    {
+      "epoch": 1.1931240416431281,
+      "grad_norm": 0.7002003788948059,
+      "learning_rate": 3.2053295544915e-05,
+      "loss": 2.487,
+      "step": 14784
+    },
+    {
+      "epoch": 1.193204745379711,
+      "grad_norm": 0.7128168940544128,
+      "learning_rate": 3.2041713449622e-05,
+      "loss": 2.4591,
+      "step": 14785
+    },
+    {
+      "epoch": 1.193285449116294,
+      "grad_norm": 0.6897242665290833,
+      "learning_rate": 3.203013304802712e-05,
+      "loss": 2.4458,
+      "step": 14786
+    },
+    {
+      "epoch": 1.1933661528528772,
+      "grad_norm": 0.7281817197799683,
+      "learning_rate": 3.2018554340419004e-05,
+      "loss": 2.3772,
+      "step": 14787
+    },
+    {
+      "epoch": 1.19344685658946,
+      "grad_norm": 0.6956086754798889,
+      "learning_rate": 3.200697732708619e-05,
+      "loss": 2.4316,
+      "step": 14788
+    },
+    {
+      "epoch": 1.1935275603260431,
+      "grad_norm": 0.7679805159568787,
+      "learning_rate": 3.199540200831729e-05,
+      "loss": 2.4464,
+      "step": 14789
+    },
+    {
+      "epoch": 1.1936082640626262,
+      "grad_norm": 0.6993041634559631,
+      "learning_rate": 3.19838283844007e-05,
+      "loss": 2.3881,
+      "step": 14790
+    },
+    {
+      "epoch": 1.193688967799209,
+      "grad_norm": 0.689618706703186,
+      "learning_rate": 3.197225645562493e-05,
+      "loss": 2.4184,
+      "step": 14791
+    },
+    {
+      "epoch": 1.1937696715357922,
+      "grad_norm": 0.6896520853042603,
+      "learning_rate": 3.1960686222278354e-05,
+      "loss": 2.4484,
+      "step": 14792
+    },
+    {
+      "epoch": 1.193850375272375,
+      "grad_norm": 0.6743811368942261,
+      "learning_rate": 3.1949117684649334e-05,
+      "loss": 2.4636,
+      "step": 14793
+    },
+    {
+      "epoch": 1.1939310790089581,
+      "grad_norm": 0.7028046250343323,
+      "learning_rate": 3.1937550843026163e-05,
+      "loss": 2.4576,
+      "step": 14794
+    },
+    {
+      "epoch": 1.1940117827455412,
+      "grad_norm": 0.7219679951667786,
+      "learning_rate": 3.192598569769718e-05,
+      "loss": 2.4495,
+      "step": 14795
+    },
+    {
+      "epoch": 1.194092486482124,
+      "grad_norm": 0.731438159942627,
+      "learning_rate": 3.191442224895056e-05,
+      "loss": 2.4699,
+      "step": 14796
+    },
+    {
+      "epoch": 1.1941731902187072,
+      "grad_norm": 0.6731431484222412,
+      "learning_rate": 3.19028604970745e-05,
+      "loss": 2.4292,
+      "step": 14797
+    },
+    {
+      "epoch": 1.19425389395529,
+      "grad_norm": 0.6720147728919983,
+      "learning_rate": 3.1891300442357174e-05,
+      "loss": 2.4482,
+      "step": 14798
+    },
+    {
+      "epoch": 1.1943345976918731,
+      "grad_norm": 0.7504273653030396,
+      "learning_rate": 3.187974208508667e-05,
+      "loss": 2.4233,
+      "step": 14799
+    },
+    {
+      "epoch": 1.1944153014284562,
+      "grad_norm": 0.6882641315460205,
+      "learning_rate": 3.186818542555108e-05,
+      "loss": 2.4633,
+      "step": 14800
+    },
+    {
+      "epoch": 1.194496005165039,
+      "grad_norm": 0.7337899208068848,
+      "learning_rate": 3.1856630464038385e-05,
+      "loss": 2.4257,
+      "step": 14801
+    },
+    {
+      "epoch": 1.1945767089016222,
+      "grad_norm": 0.7026493549346924,
+      "learning_rate": 3.1845077200836636e-05,
+      "loss": 2.482,
+      "step": 14802
+    },
+    {
+      "epoch": 1.1946574126382052,
+      "grad_norm": 0.763351321220398,
+      "learning_rate": 3.1833525636233675e-05,
+      "loss": 2.4428,
+      "step": 14803
+    },
+    {
+      "epoch": 1.194738116374788,
+      "grad_norm": 0.6568076610565186,
+      "learning_rate": 3.182197577051745e-05,
+      "loss": 2.4373,
+      "step": 14804
+    },
+    {
+      "epoch": 1.1948188201113712,
+      "grad_norm": 0.6954717040061951,
+      "learning_rate": 3.1810427603975844e-05,
+      "loss": 2.4582,
+      "step": 14805
+    },
+    {
+      "epoch": 1.1948995238479543,
+      "grad_norm": 0.7130215167999268,
+      "learning_rate": 3.179888113689661e-05,
+      "loss": 2.443,
+      "step": 14806
+    },
+    {
+      "epoch": 1.1949802275845371,
+      "grad_norm": 0.6789865493774414,
+      "learning_rate": 3.178733636956752e-05,
+      "loss": 2.4138,
+      "step": 14807
+    },
+    {
+      "epoch": 1.1950609313211202,
+      "grad_norm": 0.7725361585617065,
+      "learning_rate": 3.177579330227633e-05,
+      "loss": 2.4783,
+      "step": 14808
+    },
+    {
+      "epoch": 1.195141635057703,
+      "grad_norm": 0.6952371001243591,
+      "learning_rate": 3.17642519353107e-05,
+      "loss": 2.4571,
+      "step": 14809
+    },
+    {
+      "epoch": 1.1952223387942862,
+      "grad_norm": 0.7541885375976562,
+      "learning_rate": 3.1752712268958275e-05,
+      "loss": 2.4075,
+      "step": 14810
+    },
+    {
+      "epoch": 1.1953030425308693,
+      "grad_norm": 0.6974624395370483,
+      "learning_rate": 3.174117430350671e-05,
+      "loss": 2.4525,
+      "step": 14811
+    },
+    {
+      "epoch": 1.1953837462674521,
+      "grad_norm": 0.7293709516525269,
+      "learning_rate": 3.172963803924347e-05,
+      "loss": 2.4646,
+      "step": 14812
+    },
+    {
+      "epoch": 1.1954644500040352,
+      "grad_norm": 0.6944144368171692,
+      "learning_rate": 3.1718103476456106e-05,
+      "loss": 2.462,
+      "step": 14813
+    },
+    {
+      "epoch": 1.195545153740618,
+      "grad_norm": 0.6415363550186157,
+      "learning_rate": 3.170657061543214e-05,
+      "loss": 2.4086,
+      "step": 14814
+    },
+    {
+      "epoch": 1.1956258574772012,
+      "grad_norm": 0.6511349081993103,
+      "learning_rate": 3.169503945645892e-05,
+      "loss": 2.4376,
+      "step": 14815
+    },
+    {
+      "epoch": 1.1957065612137843,
+      "grad_norm": 0.7420210242271423,
+      "learning_rate": 3.1683509999823854e-05,
+      "loss": 2.4317,
+      "step": 14816
+    },
+    {
+      "epoch": 1.1957872649503671,
+      "grad_norm": 0.7291967272758484,
+      "learning_rate": 3.1671982245814316e-05,
+      "loss": 2.4369,
+      "step": 14817
+    },
+    {
+      "epoch": 1.1958679686869502,
+      "grad_norm": 0.685743510723114,
+      "learning_rate": 3.166045619471758e-05,
+      "loss": 2.465,
+      "step": 14818
+    },
+    {
+      "epoch": 1.1959486724235333,
+      "grad_norm": 0.7130060195922852,
+      "learning_rate": 3.164893184682093e-05,
+      "loss": 2.4305,
+      "step": 14819
+    },
+    {
+      "epoch": 1.1960293761601162,
+      "grad_norm": 0.694508969783783,
+      "learning_rate": 3.163740920241156e-05,
+      "loss": 2.4278,
+      "step": 14820
+    },
+    {
+      "epoch": 1.1961100798966993,
+      "grad_norm": 0.6478514075279236,
+      "learning_rate": 3.162588826177669e-05,
+      "loss": 2.4721,
+      "step": 14821
+    },
+    {
+      "epoch": 1.1961907836332821,
+      "grad_norm": 0.6586465835571289,
+      "learning_rate": 3.1614369025203386e-05,
+      "loss": 2.4716,
+      "step": 14822
+    },
+    {
+      "epoch": 1.1962714873698652,
+      "grad_norm": 0.7558106184005737,
+      "learning_rate": 3.160285149297876e-05,
+      "loss": 2.4656,
+      "step": 14823
+    },
+    {
+      "epoch": 1.1963521911064483,
+      "grad_norm": 0.7208340764045715,
+      "learning_rate": 3.1591335665389896e-05,
+      "loss": 2.4374,
+      "step": 14824
+    },
+    {
+      "epoch": 1.1964328948430312,
+      "grad_norm": 0.70301353931427,
+      "learning_rate": 3.157982154272375e-05,
+      "loss": 2.397,
+      "step": 14825
+    },
+    {
+      "epoch": 1.1965135985796143,
+      "grad_norm": 0.6857609152793884,
+      "learning_rate": 3.15683091252673e-05,
+      "loss": 2.4258,
+      "step": 14826
+    },
+    {
+      "epoch": 1.1965943023161971,
+      "grad_norm": 0.6954602003097534,
+      "learning_rate": 3.155679841330747e-05,
+      "loss": 2.4566,
+      "step": 14827
+    },
+    {
+      "epoch": 1.1966750060527802,
+      "grad_norm": 0.6923913955688477,
+      "learning_rate": 3.154528940713113e-05,
+      "loss": 2.4,
+      "step": 14828
+    },
+    {
+      "epoch": 1.1967557097893633,
+      "grad_norm": 0.6641134023666382,
+      "learning_rate": 3.1533782107025124e-05,
+      "loss": 2.4721,
+      "step": 14829
+    },
+    {
+      "epoch": 1.1968364135259462,
+      "grad_norm": 0.7470134496688843,
+      "learning_rate": 3.152227651327627e-05,
+      "loss": 2.4253,
+      "step": 14830
+    },
+    {
+      "epoch": 1.1969171172625293,
+      "grad_norm": 0.7234545350074768,
+      "learning_rate": 3.151077262617126e-05,
+      "loss": 2.4109,
+      "step": 14831
+    },
+    {
+      "epoch": 1.1969978209991123,
+      "grad_norm": 0.7814013957977295,
+      "learning_rate": 3.149927044599682e-05,
+      "loss": 2.4522,
+      "step": 14832
+    },
+    {
+      "epoch": 1.1970785247356952,
+      "grad_norm": 0.6825435161590576,
+      "learning_rate": 3.1487769973039624e-05,
+      "loss": 2.4728,
+      "step": 14833
+    },
+    {
+      "epoch": 1.1971592284722783,
+      "grad_norm": 0.7091361880302429,
+      "learning_rate": 3.147627120758634e-05,
+      "loss": 2.4615,
+      "step": 14834
+    },
+    {
+      "epoch": 1.1972399322088614,
+      "grad_norm": 0.7271433472633362,
+      "learning_rate": 3.146477414992346e-05,
+      "loss": 2.4154,
+      "step": 14835
+    },
+    {
+      "epoch": 1.1973206359454442,
+      "grad_norm": 0.6557306051254272,
+      "learning_rate": 3.145327880033756e-05,
+      "loss": 2.4348,
+      "step": 14836
+    },
+    {
+      "epoch": 1.1974013396820273,
+      "grad_norm": 0.6667891144752502,
+      "learning_rate": 3.1441785159115166e-05,
+      "loss": 2.4123,
+      "step": 14837
+    },
+    {
+      "epoch": 1.1974820434186102,
+      "grad_norm": 0.6755266189575195,
+      "learning_rate": 3.143029322654266e-05,
+      "loss": 2.4287,
+      "step": 14838
+    },
+    {
+      "epoch": 1.1975627471551933,
+      "grad_norm": 0.7647396922111511,
+      "learning_rate": 3.1418803002906475e-05,
+      "loss": 2.4343,
+      "step": 14839
+    },
+    {
+      "epoch": 1.1976434508917764,
+      "grad_norm": 0.7288243174552917,
+      "learning_rate": 3.140731448849305e-05,
+      "loss": 2.4536,
+      "step": 14840
+    },
+    {
+      "epoch": 1.1977241546283592,
+      "grad_norm": 0.6126244068145752,
+      "learning_rate": 3.1395827683588605e-05,
+      "loss": 2.4187,
+      "step": 14841
+    },
+    {
+      "epoch": 1.1978048583649423,
+      "grad_norm": 0.6773896217346191,
+      "learning_rate": 3.138434258847948e-05,
+      "loss": 2.3916,
+      "step": 14842
+    },
+    {
+      "epoch": 1.1978855621015252,
+      "grad_norm": 0.724413275718689,
+      "learning_rate": 3.1372859203451934e-05,
+      "loss": 2.4614,
+      "step": 14843
+    },
+    {
+      "epoch": 1.1979662658381083,
+      "grad_norm": 0.7043039798736572,
+      "learning_rate": 3.136137752879209e-05,
+      "loss": 2.4343,
+      "step": 14844
+    },
+    {
+      "epoch": 1.1980469695746914,
+      "grad_norm": 0.7543383240699768,
+      "learning_rate": 3.134989756478615e-05,
+      "loss": 2.4345,
+      "step": 14845
+    },
+    {
+      "epoch": 1.1981276733112742,
+      "grad_norm": 0.7193408608436584,
+      "learning_rate": 3.1338419311720244e-05,
+      "loss": 2.4728,
+      "step": 14846
+    },
+    {
+      "epoch": 1.1982083770478573,
+      "grad_norm": 0.8090186715126038,
+      "learning_rate": 3.132694276988038e-05,
+      "loss": 2.4246,
+      "step": 14847
+    },
+    {
+      "epoch": 1.1982890807844404,
+      "grad_norm": 0.7154600620269775,
+      "learning_rate": 3.131546793955261e-05,
+      "loss": 2.4061,
+      "step": 14848
+    },
+    {
+      "epoch": 1.1983697845210233,
+      "grad_norm": 0.6987032890319824,
+      "learning_rate": 3.130399482102293e-05,
+      "loss": 2.4525,
+      "step": 14849
+    },
+    {
+      "epoch": 1.1984504882576064,
+      "grad_norm": 0.7123507261276245,
+      "learning_rate": 3.129252341457727e-05,
+      "loss": 2.4017,
+      "step": 14850
+    },
+    {
+      "epoch": 1.1985311919941894,
+      "grad_norm": 0.6475987434387207,
+      "learning_rate": 3.128105372050153e-05,
+      "loss": 2.4617,
+      "step": 14851
+    },
+    {
+      "epoch": 1.1986118957307723,
+      "grad_norm": 0.6799046993255615,
+      "learning_rate": 3.126958573908156e-05,
+      "loss": 2.4337,
+      "step": 14852
+    },
+    {
+      "epoch": 1.1986925994673554,
+      "grad_norm": 0.6910607218742371,
+      "learning_rate": 3.125811947060322e-05,
+      "loss": 2.415,
+      "step": 14853
+    },
+    {
+      "epoch": 1.1987733032039383,
+      "grad_norm": 0.6879963278770447,
+      "learning_rate": 3.124665491535219e-05,
+      "loss": 2.4912,
+      "step": 14854
+    },
+    {
+      "epoch": 1.1988540069405214,
+      "grad_norm": 0.7038810849189758,
+      "learning_rate": 3.123519207361425e-05,
+      "loss": 2.4528,
+      "step": 14855
+    },
+    {
+      "epoch": 1.1989347106771044,
+      "grad_norm": 0.6771957278251648,
+      "learning_rate": 3.1223730945675104e-05,
+      "loss": 2.4524,
+      "step": 14856
+    },
+    {
+      "epoch": 1.1990154144136873,
+      "grad_norm": 0.7529320120811462,
+      "learning_rate": 3.1212271531820336e-05,
+      "loss": 2.4667,
+      "step": 14857
+    },
+    {
+      "epoch": 1.1990961181502704,
+      "grad_norm": 0.6498474478721619,
+      "learning_rate": 3.1200813832335574e-05,
+      "loss": 2.3863,
+      "step": 14858
+    },
+    {
+      "epoch": 1.1991768218868533,
+      "grad_norm": 0.7587705850601196,
+      "learning_rate": 3.1189357847506383e-05,
+      "loss": 2.4962,
+      "step": 14859
+    },
+    {
+      "epoch": 1.1992575256234363,
+      "grad_norm": 0.674013078212738,
+      "learning_rate": 3.117790357761825e-05,
+      "loss": 2.3939,
+      "step": 14860
+    },
+    {
+      "epoch": 1.1993382293600194,
+      "grad_norm": 0.6546844840049744,
+      "learning_rate": 3.116645102295668e-05,
+      "loss": 2.4775,
+      "step": 14861
+    },
+    {
+      "epoch": 1.1994189330966023,
+      "grad_norm": 0.7558320760726929,
+      "learning_rate": 3.11550001838071e-05,
+      "loss": 2.3918,
+      "step": 14862
+    },
+    {
+      "epoch": 1.1994996368331854,
+      "grad_norm": 0.7074883580207825,
+      "learning_rate": 3.114355106045486e-05,
+      "loss": 2.3969,
+      "step": 14863
+    },
+    {
+      "epoch": 1.1995803405697685,
+      "grad_norm": 0.706078290939331,
+      "learning_rate": 3.1132103653185305e-05,
+      "loss": 2.5028,
+      "step": 14864
+    },
+    {
+      "epoch": 1.1996610443063513,
+      "grad_norm": 0.6883544921875,
+      "learning_rate": 3.1120657962283764e-05,
+      "loss": 2.4407,
+      "step": 14865
+    },
+    {
+      "epoch": 1.1997417480429344,
+      "grad_norm": 0.6905466914176941,
+      "learning_rate": 3.110921398803551e-05,
+      "loss": 2.3893,
+      "step": 14866
+    },
+    {
+      "epoch": 1.1998224517795173,
+      "grad_norm": 0.6584910154342651,
+      "learning_rate": 3.109777173072569e-05,
+      "loss": 2.4515,
+      "step": 14867
+    },
+    {
+      "epoch": 1.1999031555161004,
+      "grad_norm": 0.6957471370697021,
+      "learning_rate": 3.108633119063951e-05,
+      "loss": 2.4483,
+      "step": 14868
+    },
+    {
+      "epoch": 1.1999838592526835,
+      "grad_norm": 0.6716276407241821,
+      "learning_rate": 3.1074892368062095e-05,
+      "loss": 2.4298,
+      "step": 14869
+    },
+    {
+      "epoch": 1.2000645629892663,
+      "grad_norm": 0.7350820302963257,
+      "learning_rate": 3.1063455263278543e-05,
+      "loss": 2.4088,
+      "step": 14870
+    },
+    {
+      "epoch": 1.2001452667258494,
+      "grad_norm": 0.7409771680831909,
+      "learning_rate": 3.105201987657388e-05,
+      "loss": 2.4089,
+      "step": 14871
+    },
+    {
+      "epoch": 1.2002259704624323,
+      "grad_norm": 0.7273266911506653,
+      "learning_rate": 3.104058620823315e-05,
+      "loss": 2.5149,
+      "step": 14872
+    },
+    {
+      "epoch": 1.2003066741990154,
+      "grad_norm": 0.6793962717056274,
+      "learning_rate": 3.102915425854124e-05,
+      "loss": 2.4422,
+      "step": 14873
+    },
+    {
+      "epoch": 1.2003873779355985,
+      "grad_norm": 0.72386234998703,
+      "learning_rate": 3.101772402778309e-05,
+      "loss": 2.4756,
+      "step": 14874
+    },
+    {
+      "epoch": 1.2004680816721813,
+      "grad_norm": 0.6530055999755859,
+      "learning_rate": 3.1006295516243625e-05,
+      "loss": 2.4145,
+      "step": 14875
+    },
+    {
+      "epoch": 1.2005487854087644,
+      "grad_norm": 0.7288365960121155,
+      "learning_rate": 3.099486872420758e-05,
+      "loss": 2.4565,
+      "step": 14876
+    },
+    {
+      "epoch": 1.2006294891453475,
+      "grad_norm": 0.6982102394104004,
+      "learning_rate": 3.09834436519598e-05,
+      "loss": 2.4788,
+      "step": 14877
+    },
+    {
+      "epoch": 1.2007101928819304,
+      "grad_norm": 0.7208256125450134,
+      "learning_rate": 3.0972020299785007e-05,
+      "loss": 2.4186,
+      "step": 14878
+    },
+    {
+      "epoch": 1.2007908966185135,
+      "grad_norm": 0.6928278803825378,
+      "learning_rate": 3.096059866796791e-05,
+      "loss": 2.4177,
+      "step": 14879
+    },
+    {
+      "epoch": 1.2008716003550965,
+      "grad_norm": 0.7145438194274902,
+      "learning_rate": 3.094917875679317e-05,
+      "loss": 2.4796,
+      "step": 14880
+    },
+    {
+      "epoch": 1.2009523040916794,
+      "grad_norm": 0.7126322388648987,
+      "learning_rate": 3.093776056654539e-05,
+      "loss": 2.4926,
+      "step": 14881
+    },
+    {
+      "epoch": 1.2010330078282625,
+      "grad_norm": 0.7775046825408936,
+      "learning_rate": 3.092634409750919e-05,
+      "loss": 2.4386,
+      "step": 14882
+    },
+    {
+      "epoch": 1.2011137115648454,
+      "grad_norm": 0.6387330889701843,
+      "learning_rate": 3.091492934996901e-05,
+      "loss": 2.4302,
+      "step": 14883
+    },
+    {
+      "epoch": 1.2011944153014285,
+      "grad_norm": 0.6883525252342224,
+      "learning_rate": 3.090351632420939e-05,
+      "loss": 2.4644,
+      "step": 14884
+    },
+    {
+      "epoch": 1.2012751190380115,
+      "grad_norm": 0.6698900461196899,
+      "learning_rate": 3.0892105020514795e-05,
+      "loss": 2.414,
+      "step": 14885
+    },
+    {
+      "epoch": 1.2013558227745944,
+      "grad_norm": 0.7124409079551697,
+      "learning_rate": 3.088069543916956e-05,
+      "loss": 2.4275,
+      "step": 14886
+    },
+    {
+      "epoch": 1.2014365265111775,
+      "grad_norm": 0.6996601223945618,
+      "learning_rate": 3.0869287580458076e-05,
+      "loss": 2.4725,
+      "step": 14887
+    },
+    {
+      "epoch": 1.2015172302477604,
+      "grad_norm": 0.653087317943573,
+      "learning_rate": 3.085788144466468e-05,
+      "loss": 2.383,
+      "step": 14888
+    },
+    {
+      "epoch": 1.2015979339843434,
+      "grad_norm": 0.7426899671554565,
+      "learning_rate": 3.0846477032073554e-05,
+      "loss": 2.4064,
+      "step": 14889
+    },
+    {
+      "epoch": 1.2016786377209265,
+      "grad_norm": 0.6417646408081055,
+      "learning_rate": 3.083507434296903e-05,
+      "loss": 2.3964,
+      "step": 14890
+    },
+    {
+      "epoch": 1.2017593414575094,
+      "grad_norm": 0.6301923394203186,
+      "learning_rate": 3.0823673377635274e-05,
+      "loss": 2.4285,
+      "step": 14891
+    },
+    {
+      "epoch": 1.2018400451940925,
+      "grad_norm": 0.7621259093284607,
+      "learning_rate": 3.081227413635638e-05,
+      "loss": 2.4731,
+      "step": 14892
+    },
+    {
+      "epoch": 1.2019207489306756,
+      "grad_norm": 0.6637598872184753,
+      "learning_rate": 3.080087661941648e-05,
+      "loss": 2.4126,
+      "step": 14893
+    },
+    {
+      "epoch": 1.2020014526672584,
+      "grad_norm": 0.6820287108421326,
+      "learning_rate": 3.078948082709964e-05,
+      "loss": 2.4108,
+      "step": 14894
+    },
+    {
+      "epoch": 1.2020821564038415,
+      "grad_norm": 0.7090989351272583,
+      "learning_rate": 3.077808675968983e-05,
+      "loss": 2.4678,
+      "step": 14895
+    },
+    {
+      "epoch": 1.2021628601404246,
+      "grad_norm": 0.7242181897163391,
+      "learning_rate": 3.076669441747105e-05,
+      "loss": 2.5346,
+      "step": 14896
+    },
+    {
+      "epoch": 1.2022435638770075,
+      "grad_norm": 0.7790088653564453,
+      "learning_rate": 3.075530380072722e-05,
+      "loss": 2.4436,
+      "step": 14897
+    },
+    {
+      "epoch": 1.2023242676135906,
+      "grad_norm": 0.6828821301460266,
+      "learning_rate": 3.074391490974225e-05,
+      "loss": 2.3767,
+      "step": 14898
+    },
+    {
+      "epoch": 1.2024049713501734,
+      "grad_norm": 0.709815502166748,
+      "learning_rate": 3.0732527744799945e-05,
+      "loss": 2.4139,
+      "step": 14899
+    },
+    {
+      "epoch": 1.2024856750867565,
+      "grad_norm": 0.6561180353164673,
+      "learning_rate": 3.07211423061841e-05,
+      "loss": 2.399,
+      "step": 14900
+    },
+    {
+      "epoch": 1.2025663788233396,
+      "grad_norm": 0.7122004628181458,
+      "learning_rate": 3.0709758594178495e-05,
+      "loss": 2.4314,
+      "step": 14901
+    },
+    {
+      "epoch": 1.2026470825599225,
+      "grad_norm": 0.6817516684532166,
+      "learning_rate": 3.0698376609066825e-05,
+      "loss": 2.4241,
+      "step": 14902
+    },
+    {
+      "epoch": 1.2027277862965056,
+      "grad_norm": 0.6848475337028503,
+      "learning_rate": 3.068699635113277e-05,
+      "loss": 2.4583,
+      "step": 14903
+    },
+    {
+      "epoch": 1.2028084900330884,
+      "grad_norm": 0.6567823886871338,
+      "learning_rate": 3.067561782065999e-05,
+      "loss": 2.3818,
+      "step": 14904
+    },
+    {
+      "epoch": 1.2028891937696715,
+      "grad_norm": 0.7373961806297302,
+      "learning_rate": 3.066424101793198e-05,
+      "loss": 2.4075,
+      "step": 14905
+    },
+    {
+      "epoch": 1.2029698975062546,
+      "grad_norm": 0.6968079209327698,
+      "learning_rate": 3.0652865943232346e-05,
+      "loss": 2.4701,
+      "step": 14906
+    },
+    {
+      "epoch": 1.2030506012428375,
+      "grad_norm": 0.7356292009353638,
+      "learning_rate": 3.064149259684459e-05,
+      "loss": 2.4188,
+      "step": 14907
+    },
+    {
+      "epoch": 1.2031313049794206,
+      "grad_norm": 0.7144857048988342,
+      "learning_rate": 3.063012097905211e-05,
+      "loss": 2.4411,
+      "step": 14908
+    },
+    {
+      "epoch": 1.2032120087160036,
+      "grad_norm": 0.734531044960022,
+      "learning_rate": 3.0618751090138365e-05,
+      "loss": 2.4595,
+      "step": 14909
+    },
+    {
+      "epoch": 1.2032927124525865,
+      "grad_norm": 0.6658234000205994,
+      "learning_rate": 3.060738293038669e-05,
+      "loss": 2.4206,
+      "step": 14910
+    },
+    {
+      "epoch": 1.2033734161891696,
+      "grad_norm": 0.678424596786499,
+      "learning_rate": 3.059601650008044e-05,
+      "loss": 2.4704,
+      "step": 14911
+    },
+    {
+      "epoch": 1.2034541199257527,
+      "grad_norm": 0.6852440237998962,
+      "learning_rate": 3.058465179950287e-05,
+      "loss": 2.46,
+      "step": 14912
+    },
+    {
+      "epoch": 1.2035348236623356,
+      "grad_norm": 0.702881395816803,
+      "learning_rate": 3.057328882893724e-05,
+      "loss": 2.4372,
+      "step": 14913
+    },
+    {
+      "epoch": 1.2036155273989186,
+      "grad_norm": 0.6978999972343445,
+      "learning_rate": 3.056192758866676e-05,
+      "loss": 2.401,
+      "step": 14914
+    },
+    {
+      "epoch": 1.2036962311355015,
+      "grad_norm": 0.7070993185043335,
+      "learning_rate": 3.055056807897454e-05,
+      "loss": 2.3967,
+      "step": 14915
+    },
+    {
+      "epoch": 1.2037769348720846,
+      "grad_norm": 0.7159305810928345,
+      "learning_rate": 3.0539210300143693e-05,
+      "loss": 2.4388,
+      "step": 14916
+    },
+    {
+      "epoch": 1.2038576386086675,
+      "grad_norm": 0.6920869946479797,
+      "learning_rate": 3.0527854252457333e-05,
+      "loss": 2.441,
+      "step": 14917
+    },
+    {
+      "epoch": 1.2039383423452505,
+      "grad_norm": 0.7014884352684021,
+      "learning_rate": 3.0516499936198417e-05,
+      "loss": 2.4115,
+      "step": 14918
+    },
+    {
+      "epoch": 1.2040190460818336,
+      "grad_norm": 0.6754150986671448,
+      "learning_rate": 3.0505147351649955e-05,
+      "loss": 2.3722,
+      "step": 14919
+    },
+    {
+      "epoch": 1.2040997498184165,
+      "grad_norm": 0.7681791186332703,
+      "learning_rate": 3.0493796499094874e-05,
+      "loss": 2.4331,
+      "step": 14920
+    },
+    {
+      "epoch": 1.2041804535549996,
+      "grad_norm": 0.7265221476554871,
+      "learning_rate": 3.0482447378816082e-05,
+      "loss": 2.4806,
+      "step": 14921
+    },
+    {
+      "epoch": 1.2042611572915827,
+      "grad_norm": 0.6841520667076111,
+      "learning_rate": 3.047109999109642e-05,
+      "loss": 2.3896,
+      "step": 14922
+    },
+    {
+      "epoch": 1.2043418610281655,
+      "grad_norm": 0.746347963809967,
+      "learning_rate": 3.0459754336218737e-05,
+      "loss": 2.4081,
+      "step": 14923
+    },
+    {
+      "epoch": 1.2044225647647486,
+      "grad_norm": 0.6679818034172058,
+      "learning_rate": 3.0448410414465712e-05,
+      "loss": 2.4206,
+      "step": 14924
+    },
+    {
+      "epoch": 1.2045032685013317,
+      "grad_norm": 0.7122265100479126,
+      "learning_rate": 3.0437068226120114e-05,
+      "loss": 2.4217,
+      "step": 14925
+    },
+    {
+      "epoch": 1.2045839722379146,
+      "grad_norm": 0.7023499011993408,
+      "learning_rate": 3.0425727771464618e-05,
+      "loss": 2.4597,
+      "step": 14926
+    },
+    {
+      "epoch": 1.2046646759744977,
+      "grad_norm": 0.7304259538650513,
+      "learning_rate": 3.0414389050781876e-05,
+      "loss": 2.4915,
+      "step": 14927
+    },
+    {
+      "epoch": 1.2047453797110805,
+      "grad_norm": 0.7209908962249756,
+      "learning_rate": 3.0403052064354442e-05,
+      "loss": 2.4163,
+      "step": 14928
+    },
+    {
+      "epoch": 1.2048260834476636,
+      "grad_norm": 0.7367275953292847,
+      "learning_rate": 3.0391716812464865e-05,
+      "loss": 2.4192,
+      "step": 14929
+    },
+    {
+      "epoch": 1.2049067871842467,
+      "grad_norm": 0.6576591730117798,
+      "learning_rate": 3.0380383295395674e-05,
+      "loss": 2.4606,
+      "step": 14930
+    },
+    {
+      "epoch": 1.2049874909208296,
+      "grad_norm": 0.7082500457763672,
+      "learning_rate": 3.0369051513429315e-05,
+      "loss": 2.4079,
+      "step": 14931
+    },
+    {
+      "epoch": 1.2050681946574127,
+      "grad_norm": 0.6770346760749817,
+      "learning_rate": 3.03577214668482e-05,
+      "loss": 2.45,
+      "step": 14932
+    },
+    {
+      "epoch": 1.2051488983939955,
+      "grad_norm": 0.6979790925979614,
+      "learning_rate": 3.034639315593476e-05,
+      "loss": 2.3966,
+      "step": 14933
+    },
+    {
+      "epoch": 1.2052296021305786,
+      "grad_norm": 0.6863394975662231,
+      "learning_rate": 3.033506658097124e-05,
+      "loss": 2.4637,
+      "step": 14934
+    },
+    {
+      "epoch": 1.2053103058671617,
+      "grad_norm": 0.7522799372673035,
+      "learning_rate": 3.0323741742239963e-05,
+      "loss": 2.4585,
+      "step": 14935
+    },
+    {
+      "epoch": 1.2053910096037446,
+      "grad_norm": 0.7119878530502319,
+      "learning_rate": 3.031241864002321e-05,
+      "loss": 2.4473,
+      "step": 14936
+    },
+    {
+      "epoch": 1.2054717133403277,
+      "grad_norm": 0.690861701965332,
+      "learning_rate": 3.030109727460312e-05,
+      "loss": 2.4564,
+      "step": 14937
+    },
+    {
+      "epoch": 1.2055524170769107,
+      "grad_norm": 0.6825447082519531,
+      "learning_rate": 3.0289777646261886e-05,
+      "loss": 2.4511,
+      "step": 14938
+    },
+    {
+      "epoch": 1.2056331208134936,
+      "grad_norm": 0.7404600977897644,
+      "learning_rate": 3.027845975528164e-05,
+      "loss": 2.4461,
+      "step": 14939
+    },
+    {
+      "epoch": 1.2057138245500767,
+      "grad_norm": 0.6871766448020935,
+      "learning_rate": 3.026714360194437e-05,
+      "loss": 2.4486,
+      "step": 14940
+    },
+    {
+      "epoch": 1.2057945282866598,
+      "grad_norm": 0.6646476984024048,
+      "learning_rate": 3.02558291865322e-05,
+      "loss": 2.378,
+      "step": 14941
+    },
+    {
+      "epoch": 1.2058752320232426,
+      "grad_norm": 0.6998385787010193,
+      "learning_rate": 3.024451650932707e-05,
+      "loss": 2.4646,
+      "step": 14942
+    },
+    {
+      "epoch": 1.2059559357598257,
+      "grad_norm": 0.6763097047805786,
+      "learning_rate": 3.023320557061098e-05,
+      "loss": 2.3971,
+      "step": 14943
+    },
+    {
+      "epoch": 1.2060366394964086,
+      "grad_norm": 0.7409633994102478,
+      "learning_rate": 3.0221896370665736e-05,
+      "loss": 2.4405,
+      "step": 14944
+    },
+    {
+      "epoch": 1.2061173432329917,
+      "grad_norm": 0.6972076892852783,
+      "learning_rate": 3.0210588909773242e-05,
+      "loss": 2.3935,
+      "step": 14945
+    },
+    {
+      "epoch": 1.2061980469695748,
+      "grad_norm": 0.6898512840270996,
+      "learning_rate": 3.0199283188215333e-05,
+      "loss": 2.4173,
+      "step": 14946
+    },
+    {
+      "epoch": 1.2062787507061576,
+      "grad_norm": 0.6878097057342529,
+      "learning_rate": 3.0187979206273707e-05,
+      "loss": 2.44,
+      "step": 14947
+    },
+    {
+      "epoch": 1.2063594544427407,
+      "grad_norm": 0.6629695296287537,
+      "learning_rate": 3.0176676964230143e-05,
+      "loss": 2.3836,
+      "step": 14948
+    },
+    {
+      "epoch": 1.2064401581793236,
+      "grad_norm": 0.717654824256897,
+      "learning_rate": 3.0165376462366336e-05,
+      "loss": 2.415,
+      "step": 14949
+    },
+    {
+      "epoch": 1.2065208619159067,
+      "grad_norm": 0.7526129484176636,
+      "learning_rate": 3.0154077700963867e-05,
+      "loss": 2.4985,
+      "step": 14950
+    },
+    {
+      "epoch": 1.2066015656524898,
+      "grad_norm": 0.6867300271987915,
+      "learning_rate": 3.014278068030435e-05,
+      "loss": 2.395,
+      "step": 14951
+    },
+    {
+      "epoch": 1.2066822693890726,
+      "grad_norm": 0.7321466207504272,
+      "learning_rate": 3.0131485400669356e-05,
+      "loss": 2.4503,
+      "step": 14952
+    },
+    {
+      "epoch": 1.2067629731256557,
+      "grad_norm": 0.6915534734725952,
+      "learning_rate": 3.0120191862340387e-05,
+      "loss": 2.398,
+      "step": 14953
+    },
+    {
+      "epoch": 1.2068436768622388,
+      "grad_norm": 0.7017377018928528,
+      "learning_rate": 3.01089000655989e-05,
+      "loss": 2.4367,
+      "step": 14954
+    },
+    {
+      "epoch": 1.2069243805988217,
+      "grad_norm": 0.7032245397567749,
+      "learning_rate": 3.0097610010726353e-05,
+      "loss": 2.4078,
+      "step": 14955
+    },
+    {
+      "epoch": 1.2070050843354048,
+      "grad_norm": 0.6795478463172913,
+      "learning_rate": 3.008632169800406e-05,
+      "loss": 2.3508,
+      "step": 14956
+    },
+    {
+      "epoch": 1.2070857880719879,
+      "grad_norm": 0.7149559855461121,
+      "learning_rate": 3.007503512771339e-05,
+      "loss": 2.4023,
+      "step": 14957
+    },
+    {
+      "epoch": 1.2071664918085707,
+      "grad_norm": 0.724756121635437,
+      "learning_rate": 3.006375030013563e-05,
+      "loss": 2.4439,
+      "step": 14958
+    },
+    {
+      "epoch": 1.2072471955451538,
+      "grad_norm": 0.7233348488807678,
+      "learning_rate": 3.005246721555205e-05,
+      "loss": 2.3819,
+      "step": 14959
+    },
+    {
+      "epoch": 1.2073278992817367,
+      "grad_norm": 0.700322151184082,
+      "learning_rate": 3.0041185874243815e-05,
+      "loss": 2.4222,
+      "step": 14960
+    },
+    {
+      "epoch": 1.2074086030183198,
+      "grad_norm": 0.7268145680427551,
+      "learning_rate": 3.002990627649209e-05,
+      "loss": 2.4698,
+      "step": 14961
+    },
+    {
+      "epoch": 1.2074893067549028,
+      "grad_norm": 0.6885111331939697,
+      "learning_rate": 3.001862842257801e-05,
+      "loss": 2.4505,
+      "step": 14962
+    },
+    {
+      "epoch": 1.2075700104914857,
+      "grad_norm": 0.7237974405288696,
+      "learning_rate": 3.0007352312782632e-05,
+      "loss": 2.422,
+      "step": 14963
+    },
+    {
+      "epoch": 1.2076507142280688,
+      "grad_norm": 0.7214741110801697,
+      "learning_rate": 2.9996077947387015e-05,
+      "loss": 2.4428,
+      "step": 14964
+    },
+    {
+      "epoch": 1.2077314179646517,
+      "grad_norm": 0.7264460921287537,
+      "learning_rate": 2.998480532667215e-05,
+      "loss": 2.4669,
+      "step": 14965
+    },
+    {
+      "epoch": 1.2078121217012348,
+      "grad_norm": 0.7055517435073853,
+      "learning_rate": 2.9973534450918928e-05,
+      "loss": 2.5082,
+      "step": 14966
+    },
+    {
+      "epoch": 1.2078928254378178,
+      "grad_norm": 0.6886781454086304,
+      "learning_rate": 2.9962265320408268e-05,
+      "loss": 2.4697,
+      "step": 14967
+    },
+    {
+      "epoch": 1.2079735291744007,
+      "grad_norm": 0.6875878572463989,
+      "learning_rate": 2.9950997935421076e-05,
+      "loss": 2.4384,
+      "step": 14968
+    },
+    {
+      "epoch": 1.2080542329109838,
+      "grad_norm": 0.7586886882781982,
+      "learning_rate": 2.99397322962381e-05,
+      "loss": 2.4088,
+      "step": 14969
+    },
+    {
+      "epoch": 1.2081349366475669,
+      "grad_norm": 0.6744365096092224,
+      "learning_rate": 2.992846840314013e-05,
+      "loss": 2.4109,
+      "step": 14970
+    },
+    {
+      "epoch": 1.2082156403841497,
+      "grad_norm": 0.6589661240577698,
+      "learning_rate": 2.9917206256407893e-05,
+      "loss": 2.4386,
+      "step": 14971
+    },
+    {
+      "epoch": 1.2082963441207328,
+      "grad_norm": 0.6787264943122864,
+      "learning_rate": 2.990594585632208e-05,
+      "loss": 2.401,
+      "step": 14972
+    },
+    {
+      "epoch": 1.2083770478573157,
+      "grad_norm": 0.710517406463623,
+      "learning_rate": 2.9894687203163317e-05,
+      "loss": 2.4813,
+      "step": 14973
+    },
+    {
+      "epoch": 1.2084577515938988,
+      "grad_norm": 0.676110029220581,
+      "learning_rate": 2.988343029721221e-05,
+      "loss": 2.4654,
+      "step": 14974
+    },
+    {
+      "epoch": 1.2085384553304819,
+      "grad_norm": 0.6940518617630005,
+      "learning_rate": 2.9872175138749336e-05,
+      "loss": 2.4188,
+      "step": 14975
+    },
+    {
+      "epoch": 1.2086191590670647,
+      "grad_norm": 0.6849910020828247,
+      "learning_rate": 2.9860921728055147e-05,
+      "loss": 2.384,
+      "step": 14976
+    },
+    {
+      "epoch": 1.2086998628036478,
+      "grad_norm": 0.6902467608451843,
+      "learning_rate": 2.9849670065410128e-05,
+      "loss": 2.4364,
+      "step": 14977
+    },
+    {
+      "epoch": 1.2087805665402307,
+      "grad_norm": 0.6742224097251892,
+      "learning_rate": 2.9838420151094747e-05,
+      "loss": 2.5085,
+      "step": 14978
+    },
+    {
+      "epoch": 1.2088612702768138,
+      "grad_norm": 0.6635094285011292,
+      "learning_rate": 2.9827171985389303e-05,
+      "loss": 2.3635,
+      "step": 14979
+    },
+    {
+      "epoch": 1.2089419740133969,
+      "grad_norm": 0.7189158201217651,
+      "learning_rate": 2.9815925568574165e-05,
+      "loss": 2.458,
+      "step": 14980
+    },
+    {
+      "epoch": 1.2090226777499797,
+      "grad_norm": 0.7370143532752991,
+      "learning_rate": 2.9804680900929628e-05,
+      "loss": 2.4543,
+      "step": 14981
+    },
+    {
+      "epoch": 1.2091033814865628,
+      "grad_norm": 0.7410217523574829,
+      "learning_rate": 2.979343798273593e-05,
+      "loss": 2.4537,
+      "step": 14982
+    },
+    {
+      "epoch": 1.209184085223146,
+      "grad_norm": 0.7525770664215088,
+      "learning_rate": 2.9782196814273277e-05,
+      "loss": 2.5147,
+      "step": 14983
+    },
+    {
+      "epoch": 1.2092647889597288,
+      "grad_norm": 0.7302291393280029,
+      "learning_rate": 2.9770957395821863e-05,
+      "loss": 2.4711,
+      "step": 14984
+    },
+    {
+      "epoch": 1.2093454926963119,
+      "grad_norm": 0.7154920101165771,
+      "learning_rate": 2.975971972766175e-05,
+      "loss": 2.5224,
+      "step": 14985
+    },
+    {
+      "epoch": 1.209426196432895,
+      "grad_norm": 0.6827684640884399,
+      "learning_rate": 2.9748483810073025e-05,
+      "loss": 2.4477,
+      "step": 14986
+    },
+    {
+      "epoch": 1.2095069001694778,
+      "grad_norm": 0.7753484845161438,
+      "learning_rate": 2.973724964333575e-05,
+      "loss": 2.4257,
+      "step": 14987
+    },
+    {
+      "epoch": 1.209587603906061,
+      "grad_norm": 0.7146809101104736,
+      "learning_rate": 2.9726017227729862e-05,
+      "loss": 2.3953,
+      "step": 14988
+    },
+    {
+      "epoch": 1.2096683076426438,
+      "grad_norm": 0.7360730767250061,
+      "learning_rate": 2.9714786563535313e-05,
+      "loss": 2.3774,
+      "step": 14989
+    },
+    {
+      "epoch": 1.2097490113792269,
+      "grad_norm": 0.7159923911094666,
+      "learning_rate": 2.970355765103201e-05,
+      "loss": 2.4068,
+      "step": 14990
+    },
+    {
+      "epoch": 1.20982971511581,
+      "grad_norm": 0.6732171773910522,
+      "learning_rate": 2.969233049049982e-05,
+      "loss": 2.4215,
+      "step": 14991
+    },
+    {
+      "epoch": 1.2099104188523928,
+      "grad_norm": 0.749812126159668,
+      "learning_rate": 2.968110508221853e-05,
+      "loss": 2.4415,
+      "step": 14992
+    },
+    {
+      "epoch": 1.209991122588976,
+      "grad_norm": 0.7185530662536621,
+      "learning_rate": 2.9669881426467916e-05,
+      "loss": 2.4536,
+      "step": 14993
+    },
+    {
+      "epoch": 1.2100718263255588,
+      "grad_norm": 0.6757143139839172,
+      "learning_rate": 2.9658659523527733e-05,
+      "loss": 2.3892,
+      "step": 14994
+    },
+    {
+      "epoch": 1.2101525300621419,
+      "grad_norm": 0.7187495231628418,
+      "learning_rate": 2.96474393736776e-05,
+      "loss": 2.434,
+      "step": 14995
+    },
+    {
+      "epoch": 1.210233233798725,
+      "grad_norm": 0.7016372680664062,
+      "learning_rate": 2.9636220977197182e-05,
+      "loss": 2.4903,
+      "step": 14996
+    },
+    {
+      "epoch": 1.2103139375353078,
+      "grad_norm": 0.7528983950614929,
+      "learning_rate": 2.9625004334366103e-05,
+      "loss": 2.3829,
+      "step": 14997
+    },
+    {
+      "epoch": 1.210394641271891,
+      "grad_norm": 0.6735692024230957,
+      "learning_rate": 2.9613789445463837e-05,
+      "loss": 2.3844,
+      "step": 14998
+    },
+    {
+      "epoch": 1.210475345008474,
+      "grad_norm": 0.6825322508811951,
+      "learning_rate": 2.9602576310769935e-05,
+      "loss": 2.4691,
+      "step": 14999
+    },
+    {
+      "epoch": 1.2105560487450568,
+      "grad_norm": 0.7507675290107727,
+      "learning_rate": 2.959136493056389e-05,
+      "loss": 2.4605,
+      "step": 15000
+    },
+    {
+      "epoch": 1.2105560487450568,
+      "eval_loss": 2.3882925510406494,
+      "eval_runtime": 1014.0781,
+      "eval_samples_per_second": 2.584,
+      "eval_steps_per_second": 0.431,
+      "step": 15000
+    },
+    {
+      "epoch": 1.21063675248164,
+      "grad_norm": 0.6937146782875061,
+      "learning_rate": 2.9580155305125044e-05,
+      "loss": 2.4444,
+      "step": 15001
+    },
+    {
+      "epoch": 1.210717456218223,
+      "grad_norm": 0.6572179794311523,
+      "learning_rate": 2.9568947434732775e-05,
+      "loss": 2.4373,
+      "step": 15002
+    },
+    {
+      "epoch": 1.2107981599548059,
+      "grad_norm": 0.7420738935470581,
+      "learning_rate": 2.955774131966651e-05,
+      "loss": 2.4046,
+      "step": 15003
+    },
+    {
+      "epoch": 1.210878863691389,
+      "grad_norm": 0.7952237129211426,
+      "learning_rate": 2.954653696020543e-05,
+      "loss": 2.4082,
+      "step": 15004
+    },
+    {
+      "epoch": 1.2109595674279718,
+      "grad_norm": 0.6640750765800476,
+      "learning_rate": 2.9535334356628817e-05,
+      "loss": 2.4109,
+      "step": 15005
+    },
+    {
+      "epoch": 1.211040271164555,
+      "grad_norm": 0.6968019008636475,
+      "learning_rate": 2.952413350921588e-05,
+      "loss": 2.3991,
+      "step": 15006
+    },
+    {
+      "epoch": 1.211120974901138,
+      "grad_norm": 0.7174221277236938,
+      "learning_rate": 2.9512934418245787e-05,
+      "loss": 2.3909,
+      "step": 15007
+    },
+    {
+      "epoch": 1.2112016786377209,
+      "grad_norm": 0.6854268908500671,
+      "learning_rate": 2.9501737083997595e-05,
+      "loss": 2.4321,
+      "step": 15008
+    },
+    {
+      "epoch": 1.211282382374304,
+      "grad_norm": 0.6705672740936279,
+      "learning_rate": 2.949054150675039e-05,
+      "loss": 2.4749,
+      "step": 15009
+    },
+    {
+      "epoch": 1.2113630861108868,
+      "grad_norm": 0.7871068716049194,
+      "learning_rate": 2.9479347686783244e-05,
+      "loss": 2.424,
+      "step": 15010
+    },
+    {
+      "epoch": 1.21144378984747,
+      "grad_norm": 0.8194620609283447,
+      "learning_rate": 2.946815562437506e-05,
+      "loss": 2.461,
+      "step": 15011
+    },
+    {
+      "epoch": 1.211524493584053,
+      "grad_norm": 0.673367977142334,
+      "learning_rate": 2.9456965319804818e-05,
+      "loss": 2.4212,
+      "step": 15012
+    },
+    {
+      "epoch": 1.2116051973206359,
+      "grad_norm": 0.6630001068115234,
+      "learning_rate": 2.9445776773351397e-05,
+      "loss": 2.4393,
+      "step": 15013
+    },
+    {
+      "epoch": 1.211685901057219,
+      "grad_norm": 0.676170825958252,
+      "learning_rate": 2.943458998529365e-05,
+      "loss": 2.3889,
+      "step": 15014
+    },
+    {
+      "epoch": 1.211766604793802,
+      "grad_norm": 0.6951417326927185,
+      "learning_rate": 2.942340495591037e-05,
+      "loss": 2.4088,
+      "step": 15015
+    },
+    {
+      "epoch": 1.211847308530385,
+      "grad_norm": 0.6909857988357544,
+      "learning_rate": 2.941222168548037e-05,
+      "loss": 2.4282,
+      "step": 15016
+    },
+    {
+      "epoch": 1.211928012266968,
+      "grad_norm": 0.653264045715332,
+      "learning_rate": 2.9401040174282292e-05,
+      "loss": 2.4369,
+      "step": 15017
+    },
+    {
+      "epoch": 1.2120087160035509,
+      "grad_norm": 0.6994543075561523,
+      "learning_rate": 2.938986042259484e-05,
+      "loss": 2.419,
+      "step": 15018
+    },
+    {
+      "epoch": 1.212089419740134,
+      "grad_norm": 0.709015965461731,
+      "learning_rate": 2.9378682430696668e-05,
+      "loss": 2.4747,
+      "step": 15019
+    },
+    {
+      "epoch": 1.212170123476717,
+      "grad_norm": 0.6899579167366028,
+      "learning_rate": 2.9367506198866313e-05,
+      "loss": 2.4134,
+      "step": 15020
+    },
+    {
+      "epoch": 1.2122508272133,
+      "grad_norm": 0.6811912059783936,
+      "learning_rate": 2.9356331727382337e-05,
+      "loss": 2.449,
+      "step": 15021
+    },
+    {
+      "epoch": 1.212331530949883,
+      "grad_norm": 0.8119748830795288,
+      "learning_rate": 2.9345159016523237e-05,
+      "loss": 2.4463,
+      "step": 15022
+    },
+    {
+      "epoch": 1.2124122346864659,
+      "grad_norm": 0.7323578000068665,
+      "learning_rate": 2.9333988066567463e-05,
+      "loss": 2.4305,
+      "step": 15023
+    },
+    {
+      "epoch": 1.212492938423049,
+      "grad_norm": 0.6639837622642517,
+      "learning_rate": 2.9322818877793436e-05,
+      "loss": 2.4237,
+      "step": 15024
+    },
+    {
+      "epoch": 1.212573642159632,
+      "grad_norm": 0.669623076915741,
+      "learning_rate": 2.9311651450479516e-05,
+      "loss": 2.4436,
+      "step": 15025
+    },
+    {
+      "epoch": 1.212654345896215,
+      "grad_norm": 0.7200437784194946,
+      "learning_rate": 2.9300485784904054e-05,
+      "loss": 2.4399,
+      "step": 15026
+    },
+    {
+      "epoch": 1.212735049632798,
+      "grad_norm": 0.7015525102615356,
+      "learning_rate": 2.9289321881345254e-05,
+      "loss": 2.4696,
+      "step": 15027
+    },
+    {
+      "epoch": 1.212815753369381,
+      "grad_norm": 0.74539715051651,
+      "learning_rate": 2.9278159740081402e-05,
+      "loss": 2.4204,
+      "step": 15028
+    },
+    {
+      "epoch": 1.212896457105964,
+      "grad_norm": 0.6373662352561951,
+      "learning_rate": 2.9266999361390713e-05,
+      "loss": 2.4273,
+      "step": 15029
+    },
+    {
+      "epoch": 1.212977160842547,
+      "grad_norm": 0.8213370442390442,
+      "learning_rate": 2.9255840745551256e-05,
+      "loss": 2.4166,
+      "step": 15030
+    },
+    {
+      "epoch": 1.2130578645791301,
+      "grad_norm": 0.7386181354522705,
+      "learning_rate": 2.9244683892841185e-05,
+      "loss": 2.3973,
+      "step": 15031
+    },
+    {
+      "epoch": 1.213138568315713,
+      "grad_norm": 0.7939273118972778,
+      "learning_rate": 2.9233528803538534e-05,
+      "loss": 2.5593,
+      "step": 15032
+    },
+    {
+      "epoch": 1.213219272052296,
+      "grad_norm": 0.7580689191818237,
+      "learning_rate": 2.9222375477921347e-05,
+      "loss": 2.4255,
+      "step": 15033
+    },
+    {
+      "epoch": 1.213299975788879,
+      "grad_norm": 0.7680409550666809,
+      "learning_rate": 2.9211223916267573e-05,
+      "loss": 2.4447,
+      "step": 15034
+    },
+    {
+      "epoch": 1.213380679525462,
+      "grad_norm": 0.6998565196990967,
+      "learning_rate": 2.9200074118855135e-05,
+      "loss": 2.4061,
+      "step": 15035
+    },
+    {
+      "epoch": 1.2134613832620451,
+      "grad_norm": 0.6673001050949097,
+      "learning_rate": 2.9188926085961954e-05,
+      "loss": 2.3989,
+      "step": 15036
+    },
+    {
+      "epoch": 1.213542086998628,
+      "grad_norm": 0.683215320110321,
+      "learning_rate": 2.9177779817865815e-05,
+      "loss": 2.4078,
+      "step": 15037
+    },
+    {
+      "epoch": 1.213622790735211,
+      "grad_norm": 0.696967363357544,
+      "learning_rate": 2.9166635314844527e-05,
+      "loss": 2.4224,
+      "step": 15038
+    },
+    {
+      "epoch": 1.213703494471794,
+      "grad_norm": 0.6930364370346069,
+      "learning_rate": 2.915549257717588e-05,
+      "loss": 2.4112,
+      "step": 15039
+    },
+    {
+      "epoch": 1.213784198208377,
+      "grad_norm": 0.7387405633926392,
+      "learning_rate": 2.914435160513752e-05,
+      "loss": 2.4458,
+      "step": 15040
+    },
+    {
+      "epoch": 1.21386490194496,
+      "grad_norm": 0.6615941524505615,
+      "learning_rate": 2.913321239900714e-05,
+      "loss": 2.4406,
+      "step": 15041
+    },
+    {
+      "epoch": 1.213945605681543,
+      "grad_norm": 0.7520569562911987,
+      "learning_rate": 2.912207495906235e-05,
+      "loss": 2.3991,
+      "step": 15042
+    },
+    {
+      "epoch": 1.214026309418126,
+      "grad_norm": 0.6952454447746277,
+      "learning_rate": 2.911093928558072e-05,
+      "loss": 2.4404,
+      "step": 15043
+    },
+    {
+      "epoch": 1.2141070131547091,
+      "grad_norm": 0.7595344185829163,
+      "learning_rate": 2.9099805378839794e-05,
+      "loss": 2.551,
+      "step": 15044
+    },
+    {
+      "epoch": 1.214187716891292,
+      "grad_norm": 0.6645220518112183,
+      "learning_rate": 2.9088673239117094e-05,
+      "loss": 2.4167,
+      "step": 15045
+    },
+    {
+      "epoch": 1.214268420627875,
+      "grad_norm": 0.6433377861976624,
+      "learning_rate": 2.907754286668998e-05,
+      "loss": 2.3873,
+      "step": 15046
+    },
+    {
+      "epoch": 1.2143491243644582,
+      "grad_norm": 0.6806936860084534,
+      "learning_rate": 2.9066414261835894e-05,
+      "loss": 2.3868,
+      "step": 15047
+    },
+    {
+      "epoch": 1.214429828101041,
+      "grad_norm": 0.7261343598365784,
+      "learning_rate": 2.905528742483222e-05,
+      "loss": 2.4785,
+      "step": 15048
+    },
+    {
+      "epoch": 1.2145105318376241,
+      "grad_norm": 0.6495440602302551,
+      "learning_rate": 2.9044162355956196e-05,
+      "loss": 2.4167,
+      "step": 15049
+    },
+    {
+      "epoch": 1.214591235574207,
+      "grad_norm": 0.6816607117652893,
+      "learning_rate": 2.9033039055485135e-05,
+      "loss": 2.459,
+      "step": 15050
+    },
+    {
+      "epoch": 1.21467193931079,
+      "grad_norm": 0.6624214053153992,
+      "learning_rate": 2.902191752369624e-05,
+      "loss": 2.4498,
+      "step": 15051
+    },
+    {
+      "epoch": 1.2147526430473732,
+      "grad_norm": 0.6800024509429932,
+      "learning_rate": 2.9010797760866737e-05,
+      "loss": 2.4442,
+      "step": 15052
+    },
+    {
+      "epoch": 1.214833346783956,
+      "grad_norm": 0.711705207824707,
+      "learning_rate": 2.8999679767273667e-05,
+      "loss": 2.422,
+      "step": 15053
+    },
+    {
+      "epoch": 1.2149140505205391,
+      "grad_norm": 0.6854784488677979,
+      "learning_rate": 2.898856354319419e-05,
+      "loss": 2.4567,
+      "step": 15054
+    },
+    {
+      "epoch": 1.214994754257122,
+      "grad_norm": 0.6676114797592163,
+      "learning_rate": 2.8977449088905373e-05,
+      "loss": 2.3913,
+      "step": 15055
+    },
+    {
+      "epoch": 1.215075457993705,
+      "grad_norm": 0.6893348693847656,
+      "learning_rate": 2.8966336404684145e-05,
+      "loss": 2.4407,
+      "step": 15056
+    },
+    {
+      "epoch": 1.2151561617302882,
+      "grad_norm": 0.6749289035797119,
+      "learning_rate": 2.8955225490807514e-05,
+      "loss": 2.409,
+      "step": 15057
+    },
+    {
+      "epoch": 1.215236865466871,
+      "grad_norm": 0.6998956203460693,
+      "learning_rate": 2.8944116347552387e-05,
+      "loss": 2.4297,
+      "step": 15058
+    },
+    {
+      "epoch": 1.2153175692034541,
+      "grad_norm": 0.7040024399757385,
+      "learning_rate": 2.8933008975195596e-05,
+      "loss": 2.4262,
+      "step": 15059
+    },
+    {
+      "epoch": 1.2153982729400372,
+      "grad_norm": 0.6638362407684326,
+      "learning_rate": 2.8921903374014005e-05,
+      "loss": 2.4355,
+      "step": 15060
+    },
+    {
+      "epoch": 1.21547897667662,
+      "grad_norm": 0.6864547729492188,
+      "learning_rate": 2.8910799544284407e-05,
+      "loss": 2.4493,
+      "step": 15061
+    },
+    {
+      "epoch": 1.2155596804132032,
+      "grad_norm": 0.707383394241333,
+      "learning_rate": 2.8899697486283474e-05,
+      "loss": 2.4604,
+      "step": 15062
+    },
+    {
+      "epoch": 1.2156403841497863,
+      "grad_norm": 0.7121397852897644,
+      "learning_rate": 2.888859720028795e-05,
+      "loss": 2.4272,
+      "step": 15063
+    },
+    {
+      "epoch": 1.2157210878863691,
+      "grad_norm": 0.7600439786911011,
+      "learning_rate": 2.8877498686574455e-05,
+      "loss": 2.4499,
+      "step": 15064
+    },
+    {
+      "epoch": 1.2158017916229522,
+      "grad_norm": 0.6654962301254272,
+      "learning_rate": 2.886640194541962e-05,
+      "loss": 2.4632,
+      "step": 15065
+    },
+    {
+      "epoch": 1.215882495359535,
+      "grad_norm": 0.7138063311576843,
+      "learning_rate": 2.8855306977099994e-05,
+      "loss": 2.4321,
+      "step": 15066
+    },
+    {
+      "epoch": 1.2159631990961182,
+      "grad_norm": 0.672604501247406,
+      "learning_rate": 2.884421378189208e-05,
+      "loss": 2.4026,
+      "step": 15067
+    },
+    {
+      "epoch": 1.2160439028327013,
+      "grad_norm": 0.6894693970680237,
+      "learning_rate": 2.8833122360072405e-05,
+      "loss": 2.4213,
+      "step": 15068
+    },
+    {
+      "epoch": 1.2161246065692841,
+      "grad_norm": 0.6784985065460205,
+      "learning_rate": 2.8822032711917325e-05,
+      "loss": 2.4207,
+      "step": 15069
+    },
+    {
+      "epoch": 1.2162053103058672,
+      "grad_norm": 0.6569294929504395,
+      "learning_rate": 2.8810944837703248e-05,
+      "loss": 2.4142,
+      "step": 15070
+    },
+    {
+      "epoch": 1.21628601404245,
+      "grad_norm": 0.7240702509880066,
+      "learning_rate": 2.879985873770654e-05,
+      "loss": 2.4173,
+      "step": 15071
+    },
+    {
+      "epoch": 1.2163667177790332,
+      "grad_norm": 0.6935575604438782,
+      "learning_rate": 2.8788774412203444e-05,
+      "loss": 2.4487,
+      "step": 15072
+    },
+    {
+      "epoch": 1.2164474215156162,
+      "grad_norm": 0.6903246641159058,
+      "learning_rate": 2.8777691861470234e-05,
+      "loss": 2.4193,
+      "step": 15073
+    },
+    {
+      "epoch": 1.216528125252199,
+      "grad_norm": 0.7982182502746582,
+      "learning_rate": 2.8766611085783123e-05,
+      "loss": 2.492,
+      "step": 15074
+    },
+    {
+      "epoch": 1.2166088289887822,
+      "grad_norm": 0.6958058476448059,
+      "learning_rate": 2.875553208541827e-05,
+      "loss": 2.4198,
+      "step": 15075
+    },
+    {
+      "epoch": 1.2166895327253653,
+      "grad_norm": 0.6869969964027405,
+      "learning_rate": 2.8744454860651794e-05,
+      "loss": 2.3768,
+      "step": 15076
+    },
+    {
+      "epoch": 1.2167702364619482,
+      "grad_norm": 0.7263007760047913,
+      "learning_rate": 2.8733379411759796e-05,
+      "loss": 2.386,
+      "step": 15077
+    },
+    {
+      "epoch": 1.2168509401985312,
+      "grad_norm": 0.7010302543640137,
+      "learning_rate": 2.872230573901825e-05,
+      "loss": 2.4417,
+      "step": 15078
+    },
+    {
+      "epoch": 1.216931643935114,
+      "grad_norm": 0.818980872631073,
+      "learning_rate": 2.8711233842703156e-05,
+      "loss": 2.433,
+      "step": 15079
+    },
+    {
+      "epoch": 1.2170123476716972,
+      "grad_norm": 0.6937929391860962,
+      "learning_rate": 2.87001637230905e-05,
+      "loss": 2.379,
+      "step": 15080
+    },
+    {
+      "epoch": 1.2170930514082803,
+      "grad_norm": 0.6954175233840942,
+      "learning_rate": 2.868909538045612e-05,
+      "loss": 2.4296,
+      "step": 15081
+    },
+    {
+      "epoch": 1.2171737551448631,
+      "grad_norm": 0.7177354097366333,
+      "learning_rate": 2.8678028815075887e-05,
+      "loss": 2.3978,
+      "step": 15082
+    },
+    {
+      "epoch": 1.2172544588814462,
+      "grad_norm": 0.7100846171379089,
+      "learning_rate": 2.8666964027225607e-05,
+      "loss": 2.4566,
+      "step": 15083
+    },
+    {
+      "epoch": 1.217335162618029,
+      "grad_norm": 0.6909635066986084,
+      "learning_rate": 2.8655901017181064e-05,
+      "loss": 2.4772,
+      "step": 15084
+    },
+    {
+      "epoch": 1.2174158663546122,
+      "grad_norm": 0.7319501638412476,
+      "learning_rate": 2.8644839785217947e-05,
+      "loss": 2.4402,
+      "step": 15085
+    },
+    {
+      "epoch": 1.2174965700911953,
+      "grad_norm": 0.6691421270370483,
+      "learning_rate": 2.8633780331611958e-05,
+      "loss": 2.4465,
+      "step": 15086
+    },
+    {
+      "epoch": 1.2175772738277781,
+      "grad_norm": 0.7028824687004089,
+      "learning_rate": 2.8622722656638745e-05,
+      "loss": 2.4765,
+      "step": 15087
+    },
+    {
+      "epoch": 1.2176579775643612,
+      "grad_norm": 0.7428398728370667,
+      "learning_rate": 2.861166676057383e-05,
+      "loss": 2.441,
+      "step": 15088
+    },
+    {
+      "epoch": 1.2177386813009443,
+      "grad_norm": 0.6715269684791565,
+      "learning_rate": 2.8600612643692803e-05,
+      "loss": 2.4621,
+      "step": 15089
+    },
+    {
+      "epoch": 1.2178193850375272,
+      "grad_norm": 0.6768512725830078,
+      "learning_rate": 2.8589560306271168e-05,
+      "loss": 2.4257,
+      "step": 15090
+    },
+    {
+      "epoch": 1.2179000887741103,
+      "grad_norm": 0.7442535758018494,
+      "learning_rate": 2.8578509748584326e-05,
+      "loss": 2.424,
+      "step": 15091
+    },
+    {
+      "epoch": 1.2179807925106934,
+      "grad_norm": 0.7275974154472351,
+      "learning_rate": 2.8567460970907722e-05,
+      "loss": 2.4698,
+      "step": 15092
+    },
+    {
+      "epoch": 1.2180614962472762,
+      "grad_norm": 0.7050346732139587,
+      "learning_rate": 2.8556413973516727e-05,
+      "loss": 2.4734,
+      "step": 15093
+    },
+    {
+      "epoch": 1.2181421999838593,
+      "grad_norm": 0.7325939536094666,
+      "learning_rate": 2.854536875668664e-05,
+      "loss": 2.4166,
+      "step": 15094
+    },
+    {
+      "epoch": 1.2182229037204422,
+      "grad_norm": 0.6764184236526489,
+      "learning_rate": 2.8534325320692746e-05,
+      "loss": 2.4742,
+      "step": 15095
+    },
+    {
+      "epoch": 1.2183036074570253,
+      "grad_norm": 0.7405500411987305,
+      "learning_rate": 2.8523283665810318e-05,
+      "loss": 2.3959,
+      "step": 15096
+    },
+    {
+      "epoch": 1.2183843111936083,
+      "grad_norm": 0.6714199185371399,
+      "learning_rate": 2.8512243792314465e-05,
+      "loss": 2.4571,
+      "step": 15097
+    },
+    {
+      "epoch": 1.2184650149301912,
+      "grad_norm": 0.6779391169548035,
+      "learning_rate": 2.8501205700480372e-05,
+      "loss": 2.3745,
+      "step": 15098
+    },
+    {
+      "epoch": 1.2185457186667743,
+      "grad_norm": 0.6876079440116882,
+      "learning_rate": 2.8490169390583134e-05,
+      "loss": 2.4432,
+      "step": 15099
+    },
+    {
+      "epoch": 1.2186264224033572,
+      "grad_norm": 0.7092362642288208,
+      "learning_rate": 2.8479134862897826e-05,
+      "loss": 2.4716,
+      "step": 15100
+    },
+    {
+      "epoch": 1.2187071261399403,
+      "grad_norm": 0.6901989579200745,
+      "learning_rate": 2.8468102117699414e-05,
+      "loss": 2.417,
+      "step": 15101
+    },
+    {
+      "epoch": 1.2187878298765233,
+      "grad_norm": 0.7011592984199524,
+      "learning_rate": 2.8457071155262884e-05,
+      "loss": 2.4439,
+      "step": 15102
+    },
+    {
+      "epoch": 1.2188685336131062,
+      "grad_norm": 0.6923472285270691,
+      "learning_rate": 2.8446041975863146e-05,
+      "loss": 2.4247,
+      "step": 15103
+    },
+    {
+      "epoch": 1.2189492373496893,
+      "grad_norm": 0.6948748230934143,
+      "learning_rate": 2.843501457977509e-05,
+      "loss": 2.3902,
+      "step": 15104
+    },
+    {
+      "epoch": 1.2190299410862724,
+      "grad_norm": 0.7034386396408081,
+      "learning_rate": 2.842398896727354e-05,
+      "loss": 2.4277,
+      "step": 15105
+    },
+    {
+      "epoch": 1.2191106448228552,
+      "grad_norm": 0.7965617775917053,
+      "learning_rate": 2.8412965138633318e-05,
+      "loss": 2.435,
+      "step": 15106
+    },
+    {
+      "epoch": 1.2191913485594383,
+      "grad_norm": 0.7371121644973755,
+      "learning_rate": 2.8401943094129112e-05,
+      "loss": 2.3928,
+      "step": 15107
+    },
+    {
+      "epoch": 1.2192720522960214,
+      "grad_norm": 0.7079561352729797,
+      "learning_rate": 2.839092283403564e-05,
+      "loss": 2.4706,
+      "step": 15108
+    },
+    {
+      "epoch": 1.2193527560326043,
+      "grad_norm": 0.6711337566375732,
+      "learning_rate": 2.8379904358627584e-05,
+      "loss": 2.4272,
+      "step": 15109
+    },
+    {
+      "epoch": 1.2194334597691874,
+      "grad_norm": 0.6840410828590393,
+      "learning_rate": 2.836888766817951e-05,
+      "loss": 2.4174,
+      "step": 15110
+    },
+    {
+      "epoch": 1.2195141635057702,
+      "grad_norm": 0.700366199016571,
+      "learning_rate": 2.8357872762965986e-05,
+      "loss": 2.4667,
+      "step": 15111
+    },
+    {
+      "epoch": 1.2195948672423533,
+      "grad_norm": 0.7090682983398438,
+      "learning_rate": 2.8346859643261593e-05,
+      "loss": 2.3748,
+      "step": 15112
+    },
+    {
+      "epoch": 1.2196755709789364,
+      "grad_norm": 0.7965148687362671,
+      "learning_rate": 2.8335848309340717e-05,
+      "loss": 2.5138,
+      "step": 15113
+    },
+    {
+      "epoch": 1.2197562747155193,
+      "grad_norm": 0.7845773696899414,
+      "learning_rate": 2.8324838761477833e-05,
+      "loss": 2.4274,
+      "step": 15114
+    },
+    {
+      "epoch": 1.2198369784521024,
+      "grad_norm": 0.6545087099075317,
+      "learning_rate": 2.831383099994731e-05,
+      "loss": 2.4311,
+      "step": 15115
+    },
+    {
+      "epoch": 1.2199176821886852,
+      "grad_norm": 0.6846331357955933,
+      "learning_rate": 2.830282502502356e-05,
+      "loss": 2.4239,
+      "step": 15116
+    },
+    {
+      "epoch": 1.2199983859252683,
+      "grad_norm": 0.7062236070632935,
+      "learning_rate": 2.8291820836980798e-05,
+      "loss": 2.4429,
+      "step": 15117
+    },
+    {
+      "epoch": 1.2200790896618514,
+      "grad_norm": 0.7526285648345947,
+      "learning_rate": 2.8280818436093315e-05,
+      "loss": 2.4882,
+      "step": 15118
+    },
+    {
+      "epoch": 1.2201597933984343,
+      "grad_norm": 0.6853364109992981,
+      "learning_rate": 2.8269817822635337e-05,
+      "loss": 2.3803,
+      "step": 15119
+    },
+    {
+      "epoch": 1.2202404971350174,
+      "grad_norm": 0.7796143293380737,
+      "learning_rate": 2.8258818996880964e-05,
+      "loss": 2.4157,
+      "step": 15120
+    },
+    {
+      "epoch": 1.2203212008716005,
+      "grad_norm": 0.7202157378196716,
+      "learning_rate": 2.824782195910437e-05,
+      "loss": 2.5101,
+      "step": 15121
+    },
+    {
+      "epoch": 1.2204019046081833,
+      "grad_norm": 0.6730707287788391,
+      "learning_rate": 2.8236826709579644e-05,
+      "loss": 2.4397,
+      "step": 15122
+    },
+    {
+      "epoch": 1.2204826083447664,
+      "grad_norm": 0.7840865850448608,
+      "learning_rate": 2.8225833248580745e-05,
+      "loss": 2.4452,
+      "step": 15123
+    },
+    {
+      "epoch": 1.2205633120813493,
+      "grad_norm": 0.8323497772216797,
+      "learning_rate": 2.821484157638171e-05,
+      "loss": 2.4775,
+      "step": 15124
+    },
+    {
+      "epoch": 1.2206440158179324,
+      "grad_norm": 0.6699438691139221,
+      "learning_rate": 2.8203851693256466e-05,
+      "loss": 2.3958,
+      "step": 15125
+    },
+    {
+      "epoch": 1.2207247195545154,
+      "grad_norm": 0.6711557507514954,
+      "learning_rate": 2.8192863599478923e-05,
+      "loss": 2.477,
+      "step": 15126
+    },
+    {
+      "epoch": 1.2208054232910983,
+      "grad_norm": 0.6255797743797302,
+      "learning_rate": 2.8181877295322922e-05,
+      "loss": 2.4222,
+      "step": 15127
+    },
+    {
+      "epoch": 1.2208861270276814,
+      "grad_norm": 0.7313731908798218,
+      "learning_rate": 2.8170892781062297e-05,
+      "loss": 2.4343,
+      "step": 15128
+    },
+    {
+      "epoch": 1.2209668307642643,
+      "grad_norm": 0.6611476540565491,
+      "learning_rate": 2.815991005697076e-05,
+      "loss": 2.3844,
+      "step": 15129
+    },
+    {
+      "epoch": 1.2210475345008474,
+      "grad_norm": 0.7293661236763,
+      "learning_rate": 2.8148929123322065e-05,
+      "loss": 2.3912,
+      "step": 15130
+    },
+    {
+      "epoch": 1.2211282382374304,
+      "grad_norm": 0.7150777578353882,
+      "learning_rate": 2.8137949980389866e-05,
+      "loss": 2.4227,
+      "step": 15131
+    },
+    {
+      "epoch": 1.2212089419740133,
+      "grad_norm": 0.7001000642776489,
+      "learning_rate": 2.8126972628447845e-05,
+      "loss": 2.4751,
+      "step": 15132
+    },
+    {
+      "epoch": 1.2212896457105964,
+      "grad_norm": 0.7106043100357056,
+      "learning_rate": 2.8115997067769505e-05,
+      "loss": 2.4127,
+      "step": 15133
+    },
+    {
+      "epoch": 1.2213703494471795,
+      "grad_norm": 0.6969115138053894,
+      "learning_rate": 2.810502329862842e-05,
+      "loss": 2.4073,
+      "step": 15134
+    },
+    {
+      "epoch": 1.2214510531837623,
+      "grad_norm": 0.7493317127227783,
+      "learning_rate": 2.8094051321298098e-05,
+      "loss": 2.4541,
+      "step": 15135
+    },
+    {
+      "epoch": 1.2215317569203454,
+      "grad_norm": 0.6499322652816772,
+      "learning_rate": 2.808308113605198e-05,
+      "loss": 2.4057,
+      "step": 15136
+    },
+    {
+      "epoch": 1.2216124606569285,
+      "grad_norm": 0.6716788411140442,
+      "learning_rate": 2.807211274316347e-05,
+      "loss": 2.3856,
+      "step": 15137
+    },
+    {
+      "epoch": 1.2216931643935114,
+      "grad_norm": 0.7724741101264954,
+      "learning_rate": 2.8061146142905958e-05,
+      "loss": 2.4652,
+      "step": 15138
+    },
+    {
+      "epoch": 1.2217738681300945,
+      "grad_norm": 0.7014325261116028,
+      "learning_rate": 2.8050181335552718e-05,
+      "loss": 2.4506,
+      "step": 15139
+    },
+    {
+      "epoch": 1.2218545718666773,
+      "grad_norm": 0.6705317497253418,
+      "learning_rate": 2.8039218321377026e-05,
+      "loss": 2.4581,
+      "step": 15140
+    },
+    {
+      "epoch": 1.2219352756032604,
+      "grad_norm": 0.709973931312561,
+      "learning_rate": 2.8028257100652156e-05,
+      "loss": 2.427,
+      "step": 15141
+    },
+    {
+      "epoch": 1.2220159793398435,
+      "grad_norm": 0.7021297812461853,
+      "learning_rate": 2.801729767365122e-05,
+      "loss": 2.3784,
+      "step": 15142
+    },
+    {
+      "epoch": 1.2220966830764264,
+      "grad_norm": 0.7431899905204773,
+      "learning_rate": 2.8006340040647393e-05,
+      "loss": 2.4135,
+      "step": 15143
+    },
+    {
+      "epoch": 1.2221773868130095,
+      "grad_norm": 0.6724472045898438,
+      "learning_rate": 2.7995384201913765e-05,
+      "loss": 2.3966,
+      "step": 15144
+    },
+    {
+      "epoch": 1.2222580905495923,
+      "grad_norm": 0.7381375432014465,
+      "learning_rate": 2.7984430157723384e-05,
+      "loss": 2.4853,
+      "step": 15145
+    },
+    {
+      "epoch": 1.2223387942861754,
+      "grad_norm": 0.6809988617897034,
+      "learning_rate": 2.7973477908349255e-05,
+      "loss": 2.408,
+      "step": 15146
+    },
+    {
+      "epoch": 1.2224194980227585,
+      "grad_norm": 0.7042898535728455,
+      "learning_rate": 2.7962527454064337e-05,
+      "loss": 2.3981,
+      "step": 15147
+    },
+    {
+      "epoch": 1.2225002017593414,
+      "grad_norm": 0.7096118330955505,
+      "learning_rate": 2.7951578795141576e-05,
+      "loss": 2.4175,
+      "step": 15148
+    },
+    {
+      "epoch": 1.2225809054959245,
+      "grad_norm": 0.7271720767021179,
+      "learning_rate": 2.794063193185378e-05,
+      "loss": 2.4193,
+      "step": 15149
+    },
+    {
+      "epoch": 1.2226616092325076,
+      "grad_norm": 0.7000352740287781,
+      "learning_rate": 2.7929686864473792e-05,
+      "loss": 2.422,
+      "step": 15150
+    },
+    {
+      "epoch": 1.2227423129690904,
+      "grad_norm": 0.6983076333999634,
+      "learning_rate": 2.791874359327443e-05,
+      "loss": 2.4613,
+      "step": 15151
+    },
+    {
+      "epoch": 1.2228230167056735,
+      "grad_norm": 0.7520100474357605,
+      "learning_rate": 2.7907802118528383e-05,
+      "loss": 2.4147,
+      "step": 15152
+    },
+    {
+      "epoch": 1.2229037204422566,
+      "grad_norm": 0.7056650519371033,
+      "learning_rate": 2.789686244050834e-05,
+      "loss": 2.4568,
+      "step": 15153
+    },
+    {
+      "epoch": 1.2229844241788395,
+      "grad_norm": 0.7092614769935608,
+      "learning_rate": 2.7885924559486975e-05,
+      "loss": 2.4758,
+      "step": 15154
+    },
+    {
+      "epoch": 1.2230651279154225,
+      "grad_norm": 0.702521562576294,
+      "learning_rate": 2.7874988475736885e-05,
+      "loss": 2.4893,
+      "step": 15155
+    },
+    {
+      "epoch": 1.2231458316520054,
+      "grad_norm": 0.7454921007156372,
+      "learning_rate": 2.786405418953061e-05,
+      "loss": 2.4277,
+      "step": 15156
+    },
+    {
+      "epoch": 1.2232265353885885,
+      "grad_norm": 0.659503161907196,
+      "learning_rate": 2.7853121701140694e-05,
+      "loss": 2.4664,
+      "step": 15157
+    },
+    {
+      "epoch": 1.2233072391251716,
+      "grad_norm": 0.6368914842605591,
+      "learning_rate": 2.7842191010839556e-05,
+      "loss": 2.3728,
+      "step": 15158
+    },
+    {
+      "epoch": 1.2233879428617545,
+      "grad_norm": 0.7076737880706787,
+      "learning_rate": 2.783126211889965e-05,
+      "loss": 2.4204,
+      "step": 15159
+    },
+    {
+      "epoch": 1.2234686465983375,
+      "grad_norm": 0.718100905418396,
+      "learning_rate": 2.7820335025593325e-05,
+      "loss": 2.478,
+      "step": 15160
+    },
+    {
+      "epoch": 1.2235493503349204,
+      "grad_norm": 0.6804678440093994,
+      "learning_rate": 2.7809409731192972e-05,
+      "loss": 2.3755,
+      "step": 15161
+    },
+    {
+      "epoch": 1.2236300540715035,
+      "grad_norm": 0.7068643569946289,
+      "learning_rate": 2.77984862359708e-05,
+      "loss": 2.3713,
+      "step": 15162
+    },
+    {
+      "epoch": 1.2237107578080866,
+      "grad_norm": 0.7047072052955627,
+      "learning_rate": 2.7787564540199097e-05,
+      "loss": 2.4264,
+      "step": 15163
+    },
+    {
+      "epoch": 1.2237914615446694,
+      "grad_norm": 0.6985021829605103,
+      "learning_rate": 2.7776644644150076e-05,
+      "loss": 2.4101,
+      "step": 15164
+    },
+    {
+      "epoch": 1.2238721652812525,
+      "grad_norm": 0.7543687224388123,
+      "learning_rate": 2.776572654809583e-05,
+      "loss": 2.3722,
+      "step": 15165
+    },
+    {
+      "epoch": 1.2239528690178356,
+      "grad_norm": 0.7199926972389221,
+      "learning_rate": 2.7754810252308473e-05,
+      "loss": 2.3819,
+      "step": 15166
+    },
+    {
+      "epoch": 1.2240335727544185,
+      "grad_norm": 0.696756899356842,
+      "learning_rate": 2.7743895757060156e-05,
+      "loss": 2.4245,
+      "step": 15167
+    },
+    {
+      "epoch": 1.2241142764910016,
+      "grad_norm": 0.7848933339118958,
+      "learning_rate": 2.773298306262281e-05,
+      "loss": 2.4725,
+      "step": 15168
+    },
+    {
+      "epoch": 1.2241949802275847,
+      "grad_norm": 0.6819389462471008,
+      "learning_rate": 2.7722072169268432e-05,
+      "loss": 2.4338,
+      "step": 15169
+    },
+    {
+      "epoch": 1.2242756839641675,
+      "grad_norm": 0.7185801267623901,
+      "learning_rate": 2.7711163077268977e-05,
+      "loss": 2.4745,
+      "step": 15170
+    },
+    {
+      "epoch": 1.2243563877007506,
+      "grad_norm": 0.7645030617713928,
+      "learning_rate": 2.7700255786896278e-05,
+      "loss": 2.4677,
+      "step": 15171
+    },
+    {
+      "epoch": 1.2244370914373335,
+      "grad_norm": 0.6559275388717651,
+      "learning_rate": 2.7689350298422202e-05,
+      "loss": 2.386,
+      "step": 15172
+    },
+    {
+      "epoch": 1.2245177951739166,
+      "grad_norm": 0.6965066194534302,
+      "learning_rate": 2.767844661211856e-05,
+      "loss": 2.4022,
+      "step": 15173
+    },
+    {
+      "epoch": 1.2245984989104994,
+      "grad_norm": 0.6618858575820923,
+      "learning_rate": 2.7667544728257057e-05,
+      "loss": 2.3541,
+      "step": 15174
+    },
+    {
+      "epoch": 1.2246792026470825,
+      "grad_norm": 0.6635501980781555,
+      "learning_rate": 2.765664464710941e-05,
+      "loss": 2.3984,
+      "step": 15175
+    },
+    {
+      "epoch": 1.2247599063836656,
+      "grad_norm": 0.6987191438674927,
+      "learning_rate": 2.764574636894729e-05,
+      "loss": 2.4637,
+      "step": 15176
+    },
+    {
+      "epoch": 1.2248406101202485,
+      "grad_norm": 0.7289232611656189,
+      "learning_rate": 2.7634849894042303e-05,
+      "loss": 2.4033,
+      "step": 15177
+    },
+    {
+      "epoch": 1.2249213138568316,
+      "grad_norm": 0.7245565056800842,
+      "learning_rate": 2.762395522266602e-05,
+      "loss": 2.4281,
+      "step": 15178
+    },
+    {
+      "epoch": 1.2250020175934146,
+      "grad_norm": 0.6946065425872803,
+      "learning_rate": 2.761306235508997e-05,
+      "loss": 2.3869,
+      "step": 15179
+    },
+    {
+      "epoch": 1.2250827213299975,
+      "grad_norm": 0.6381784677505493,
+      "learning_rate": 2.7602171291585666e-05,
+      "loss": 2.404,
+      "step": 15180
+    },
+    {
+      "epoch": 1.2251634250665806,
+      "grad_norm": 0.6893685460090637,
+      "learning_rate": 2.759128203242446e-05,
+      "loss": 2.4807,
+      "step": 15181
+    },
+    {
+      "epoch": 1.2252441288031637,
+      "grad_norm": 0.6640260815620422,
+      "learning_rate": 2.7580394577877787e-05,
+      "loss": 2.4036,
+      "step": 15182
+    },
+    {
+      "epoch": 1.2253248325397466,
+      "grad_norm": 0.7125177979469299,
+      "learning_rate": 2.7569508928217026e-05,
+      "loss": 2.3869,
+      "step": 15183
+    },
+    {
+      "epoch": 1.2254055362763296,
+      "grad_norm": 0.657865583896637,
+      "learning_rate": 2.7558625083713397e-05,
+      "loss": 2.3869,
+      "step": 15184
+    },
+    {
+      "epoch": 1.2254862400129125,
+      "grad_norm": 0.6776065230369568,
+      "learning_rate": 2.7547743044638197e-05,
+      "loss": 2.4128,
+      "step": 15185
+    },
+    {
+      "epoch": 1.2255669437494956,
+      "grad_norm": 0.7126299738883972,
+      "learning_rate": 2.753686281126263e-05,
+      "loss": 2.4465,
+      "step": 15186
+    },
+    {
+      "epoch": 1.2256476474860787,
+      "grad_norm": 0.6918273568153381,
+      "learning_rate": 2.7525984383857873e-05,
+      "loss": 2.428,
+      "step": 15187
+    },
+    {
+      "epoch": 1.2257283512226615,
+      "grad_norm": 0.7742759585380554,
+      "learning_rate": 2.7515107762695025e-05,
+      "loss": 2.4299,
+      "step": 15188
+    },
+    {
+      "epoch": 1.2258090549592446,
+      "grad_norm": 0.7194607853889465,
+      "learning_rate": 2.7504232948045205e-05,
+      "loss": 2.4315,
+      "step": 15189
+    },
+    {
+      "epoch": 1.2258897586958275,
+      "grad_norm": 0.6962646245956421,
+      "learning_rate": 2.7493359940179363e-05,
+      "loss": 2.4494,
+      "step": 15190
+    },
+    {
+      "epoch": 1.2259704624324106,
+      "grad_norm": 0.6681686639785767,
+      "learning_rate": 2.7482488739368538e-05,
+      "loss": 2.427,
+      "step": 15191
+    },
+    {
+      "epoch": 1.2260511661689937,
+      "grad_norm": 0.6589877009391785,
+      "learning_rate": 2.747161934588366e-05,
+      "loss": 2.4333,
+      "step": 15192
+    },
+    {
+      "epoch": 1.2261318699055765,
+      "grad_norm": 0.7415218949317932,
+      "learning_rate": 2.746075175999564e-05,
+      "loss": 2.4203,
+      "step": 15193
+    },
+    {
+      "epoch": 1.2262125736421596,
+      "grad_norm": 0.7371910214424133,
+      "learning_rate": 2.7449885981975276e-05,
+      "loss": 2.4684,
+      "step": 15194
+    },
+    {
+      "epoch": 1.2262932773787427,
+      "grad_norm": 0.7010802626609802,
+      "learning_rate": 2.7439022012093407e-05,
+      "loss": 2.4625,
+      "step": 15195
+    },
+    {
+      "epoch": 1.2263739811153256,
+      "grad_norm": 0.7125125527381897,
+      "learning_rate": 2.7428159850620773e-05,
+      "loss": 2.4075,
+      "step": 15196
+    },
+    {
+      "epoch": 1.2264546848519087,
+      "grad_norm": 0.701133668422699,
+      "learning_rate": 2.7417299497828107e-05,
+      "loss": 2.4525,
+      "step": 15197
+    },
+    {
+      "epoch": 1.2265353885884918,
+      "grad_norm": 0.7543410658836365,
+      "learning_rate": 2.7406440953986078e-05,
+      "loss": 2.474,
+      "step": 15198
+    },
+    {
+      "epoch": 1.2266160923250746,
+      "grad_norm": 0.69012051820755,
+      "learning_rate": 2.7395584219365323e-05,
+      "loss": 2.4853,
+      "step": 15199
+    },
+    {
+      "epoch": 1.2266967960616577,
+      "grad_norm": 0.6559048295021057,
+      "learning_rate": 2.7384729294236378e-05,
+      "loss": 2.4252,
+      "step": 15200
+    },
+    {
+      "epoch": 1.2267774997982406,
+      "grad_norm": 0.6603518128395081,
+      "learning_rate": 2.7373876178869794e-05,
+      "loss": 2.4047,
+      "step": 15201
+    },
+    {
+      "epoch": 1.2268582035348237,
+      "grad_norm": 0.7159265279769897,
+      "learning_rate": 2.736302487353609e-05,
+      "loss": 2.4352,
+      "step": 15202
+    },
+    {
+      "epoch": 1.2269389072714068,
+      "grad_norm": 0.6784560084342957,
+      "learning_rate": 2.735217537850565e-05,
+      "loss": 2.3933,
+      "step": 15203
+    },
+    {
+      "epoch": 1.2270196110079896,
+      "grad_norm": 0.7341950535774231,
+      "learning_rate": 2.7341327694048903e-05,
+      "loss": 2.4514,
+      "step": 15204
+    },
+    {
+      "epoch": 1.2271003147445727,
+      "grad_norm": 0.726046621799469,
+      "learning_rate": 2.7330481820436204e-05,
+      "loss": 2.4427,
+      "step": 15205
+    },
+    {
+      "epoch": 1.2271810184811556,
+      "grad_norm": 0.6897192001342773,
+      "learning_rate": 2.7319637757937854e-05,
+      "loss": 2.4587,
+      "step": 15206
+    },
+    {
+      "epoch": 1.2272617222177387,
+      "grad_norm": 0.6981058716773987,
+      "learning_rate": 2.7308795506824124e-05,
+      "loss": 2.4297,
+      "step": 15207
+    },
+    {
+      "epoch": 1.2273424259543217,
+      "grad_norm": 0.694583535194397,
+      "learning_rate": 2.729795506736522e-05,
+      "loss": 2.3608,
+      "step": 15208
+    },
+    {
+      "epoch": 1.2274231296909046,
+      "grad_norm": 0.710192084312439,
+      "learning_rate": 2.728711643983136e-05,
+      "loss": 2.3733,
+      "step": 15209
+    },
+    {
+      "epoch": 1.2275038334274877,
+      "grad_norm": 0.7203633785247803,
+      "learning_rate": 2.7276279624492595e-05,
+      "loss": 2.389,
+      "step": 15210
+    },
+    {
+      "epoch": 1.2275845371640708,
+      "grad_norm": 0.7298668622970581,
+      "learning_rate": 2.726544462161905e-05,
+      "loss": 2.3981,
+      "step": 15211
+    },
+    {
+      "epoch": 1.2276652409006537,
+      "grad_norm": 0.6640039682388306,
+      "learning_rate": 2.725461143148078e-05,
+      "loss": 2.4073,
+      "step": 15212
+    },
+    {
+      "epoch": 1.2277459446372367,
+      "grad_norm": 0.7203015685081482,
+      "learning_rate": 2.724378005434772e-05,
+      "loss": 2.4901,
+      "step": 15213
+    },
+    {
+      "epoch": 1.2278266483738198,
+      "grad_norm": 0.6668895483016968,
+      "learning_rate": 2.723295049048985e-05,
+      "loss": 2.4482,
+      "step": 15214
+    },
+    {
+      "epoch": 1.2279073521104027,
+      "grad_norm": 0.7551584839820862,
+      "learning_rate": 2.7222122740177103e-05,
+      "loss": 2.4877,
+      "step": 15215
+    },
+    {
+      "epoch": 1.2279880558469858,
+      "grad_norm": 0.707202672958374,
+      "learning_rate": 2.721129680367923e-05,
+      "loss": 2.4577,
+      "step": 15216
+    },
+    {
+      "epoch": 1.2280687595835686,
+      "grad_norm": 0.685153603553772,
+      "learning_rate": 2.7200472681266155e-05,
+      "loss": 2.476,
+      "step": 15217
+    },
+    {
+      "epoch": 1.2281494633201517,
+      "grad_norm": 0.6843041181564331,
+      "learning_rate": 2.718965037320762e-05,
+      "loss": 2.4164,
+      "step": 15218
+    },
+    {
+      "epoch": 1.2282301670567348,
+      "grad_norm": 0.6548978686332703,
+      "learning_rate": 2.7178829879773306e-05,
+      "loss": 2.4187,
+      "step": 15219
+    },
+    {
+      "epoch": 1.2283108707933177,
+      "grad_norm": 0.7037245035171509,
+      "learning_rate": 2.7168011201232902e-05,
+      "loss": 2.3621,
+      "step": 15220
+    },
+    {
+      "epoch": 1.2283915745299008,
+      "grad_norm": 0.6540676951408386,
+      "learning_rate": 2.7157194337856074e-05,
+      "loss": 2.4542,
+      "step": 15221
+    },
+    {
+      "epoch": 1.2284722782664836,
+      "grad_norm": 0.7699899673461914,
+      "learning_rate": 2.7146379289912338e-05,
+      "loss": 2.4639,
+      "step": 15222
+    },
+    {
+      "epoch": 1.2285529820030667,
+      "grad_norm": 0.7178743481636047,
+      "learning_rate": 2.713556605767128e-05,
+      "loss": 2.4222,
+      "step": 15223
+    },
+    {
+      "epoch": 1.2286336857396498,
+      "grad_norm": 0.6749793887138367,
+      "learning_rate": 2.7124754641402383e-05,
+      "loss": 2.4323,
+      "step": 15224
+    },
+    {
+      "epoch": 1.2287143894762327,
+      "grad_norm": 0.7035594582557678,
+      "learning_rate": 2.711394504137513e-05,
+      "loss": 2.4466,
+      "step": 15225
+    },
+    {
+      "epoch": 1.2287950932128158,
+      "grad_norm": 0.6518487930297852,
+      "learning_rate": 2.7103137257858868e-05,
+      "loss": 2.4969,
+      "step": 15226
+    },
+    {
+      "epoch": 1.2288757969493989,
+      "grad_norm": 0.6739057898521423,
+      "learning_rate": 2.7092331291122974e-05,
+      "loss": 2.406,
+      "step": 15227
+    },
+    {
+      "epoch": 1.2289565006859817,
+      "grad_norm": 0.6584770083427429,
+      "learning_rate": 2.7081527141436767e-05,
+      "loss": 2.4304,
+      "step": 15228
+    },
+    {
+      "epoch": 1.2290372044225648,
+      "grad_norm": 0.6846301555633545,
+      "learning_rate": 2.7070724809069514e-05,
+      "loss": 2.3995,
+      "step": 15229
+    },
+    {
+      "epoch": 1.2291179081591477,
+      "grad_norm": 0.6778364777565002,
+      "learning_rate": 2.705992429429044e-05,
+      "loss": 2.38,
+      "step": 15230
+    },
+    {
+      "epoch": 1.2291986118957308,
+      "grad_norm": 0.6957302689552307,
+      "learning_rate": 2.7049125597368753e-05,
+      "loss": 2.3973,
+      "step": 15231
+    },
+    {
+      "epoch": 1.2292793156323139,
+      "grad_norm": 0.730269193649292,
+      "learning_rate": 2.7038328718573514e-05,
+      "loss": 2.4829,
+      "step": 15232
+    },
+    {
+      "epoch": 1.2293600193688967,
+      "grad_norm": 0.7114049196243286,
+      "learning_rate": 2.702753365817384e-05,
+      "loss": 2.3902,
+      "step": 15233
+    },
+    {
+      "epoch": 1.2294407231054798,
+      "grad_norm": 0.7137531638145447,
+      "learning_rate": 2.7016740416438823e-05,
+      "loss": 2.3957,
+      "step": 15234
+    },
+    {
+      "epoch": 1.2295214268420627,
+      "grad_norm": 0.7178330421447754,
+      "learning_rate": 2.7005948993637386e-05,
+      "loss": 2.4429,
+      "step": 15235
+    },
+    {
+      "epoch": 1.2296021305786458,
+      "grad_norm": 0.6767767071723938,
+      "learning_rate": 2.6995159390038506e-05,
+      "loss": 2.4009,
+      "step": 15236
+    },
+    {
+      "epoch": 1.2296828343152288,
+      "grad_norm": 0.7713541984558105,
+      "learning_rate": 2.6984371605911086e-05,
+      "loss": 2.4326,
+      "step": 15237
+    },
+    {
+      "epoch": 1.2297635380518117,
+      "grad_norm": 0.7218228578567505,
+      "learning_rate": 2.6973585641523992e-05,
+      "loss": 2.4358,
+      "step": 15238
+    },
+    {
+      "epoch": 1.2298442417883948,
+      "grad_norm": 0.6782575249671936,
+      "learning_rate": 2.696280149714604e-05,
+      "loss": 2.3844,
+      "step": 15239
+    },
+    {
+      "epoch": 1.2299249455249779,
+      "grad_norm": 0.6825734972953796,
+      "learning_rate": 2.6952019173045982e-05,
+      "loss": 2.4621,
+      "step": 15240
+    },
+    {
+      "epoch": 1.2300056492615608,
+      "grad_norm": 0.6587522625923157,
+      "learning_rate": 2.6941238669492608e-05,
+      "loss": 2.4465,
+      "step": 15241
+    },
+    {
+      "epoch": 1.2300863529981438,
+      "grad_norm": 0.6898796558380127,
+      "learning_rate": 2.6930459986754498e-05,
+      "loss": 2.4469,
+      "step": 15242
+    },
+    {
+      "epoch": 1.230167056734727,
+      "grad_norm": 0.6764062643051147,
+      "learning_rate": 2.6919683125100338e-05,
+      "loss": 2.4476,
+      "step": 15243
+    },
+    {
+      "epoch": 1.2302477604713098,
+      "grad_norm": 0.6647047400474548,
+      "learning_rate": 2.6908908084798733e-05,
+      "loss": 2.3677,
+      "step": 15244
+    },
+    {
+      "epoch": 1.2303284642078929,
+      "grad_norm": 0.7091608047485352,
+      "learning_rate": 2.6898134866118174e-05,
+      "loss": 2.4605,
+      "step": 15245
+    },
+    {
+      "epoch": 1.2304091679444757,
+      "grad_norm": 0.691007137298584,
+      "learning_rate": 2.6887363469327188e-05,
+      "loss": 2.4397,
+      "step": 15246
+    },
+    {
+      "epoch": 1.2304898716810588,
+      "grad_norm": 0.6685532927513123,
+      "learning_rate": 2.6876593894694214e-05,
+      "loss": 2.4279,
+      "step": 15247
+    },
+    {
+      "epoch": 1.230570575417642,
+      "grad_norm": 0.684474766254425,
+      "learning_rate": 2.686582614248767e-05,
+      "loss": 2.4162,
+      "step": 15248
+    },
+    {
+      "epoch": 1.2306512791542248,
+      "grad_norm": 0.657293975353241,
+      "learning_rate": 2.6855060212975915e-05,
+      "loss": 2.4337,
+      "step": 15249
+    },
+    {
+      "epoch": 1.2307319828908079,
+      "grad_norm": 0.7136504650115967,
+      "learning_rate": 2.684429610642729e-05,
+      "loss": 2.4156,
+      "step": 15250
+    },
+    {
+      "epoch": 1.2308126866273907,
+      "grad_norm": 0.6564410924911499,
+      "learning_rate": 2.6833533823110013e-05,
+      "loss": 2.5101,
+      "step": 15251
+    },
+    {
+      "epoch": 1.2308933903639738,
+      "grad_norm": 0.6628747582435608,
+      "learning_rate": 2.682277336329233e-05,
+      "loss": 2.3933,
+      "step": 15252
+    },
+    {
+      "epoch": 1.230974094100557,
+      "grad_norm": 0.7362595796585083,
+      "learning_rate": 2.681201472724244e-05,
+      "loss": 2.4541,
+      "step": 15253
+    },
+    {
+      "epoch": 1.2310547978371398,
+      "grad_norm": 0.7604697346687317,
+      "learning_rate": 2.680125791522844e-05,
+      "loss": 2.4383,
+      "step": 15254
+    },
+    {
+      "epoch": 1.2311355015737229,
+      "grad_norm": 0.7128429412841797,
+      "learning_rate": 2.6790502927518434e-05,
+      "loss": 2.4492,
+      "step": 15255
+    },
+    {
+      "epoch": 1.231216205310306,
+      "grad_norm": 0.6761955618858337,
+      "learning_rate": 2.677974976438047e-05,
+      "loss": 2.4355,
+      "step": 15256
+    },
+    {
+      "epoch": 1.2312969090468888,
+      "grad_norm": 0.6687077879905701,
+      "learning_rate": 2.6768998426082538e-05,
+      "loss": 2.4317,
+      "step": 15257
+    },
+    {
+      "epoch": 1.231377612783472,
+      "grad_norm": 0.7423825860023499,
+      "learning_rate": 2.675824891289259e-05,
+      "loss": 2.4216,
+      "step": 15258
+    },
+    {
+      "epoch": 1.231458316520055,
+      "grad_norm": 0.671130359172821,
+      "learning_rate": 2.6747501225078542e-05,
+      "loss": 2.4775,
+      "step": 15259
+    },
+    {
+      "epoch": 1.2315390202566379,
+      "grad_norm": 0.7421461939811707,
+      "learning_rate": 2.6736755362908273e-05,
+      "loss": 2.4042,
+      "step": 15260
+    },
+    {
+      "epoch": 1.231619723993221,
+      "grad_norm": 0.7084131240844727,
+      "learning_rate": 2.6726011326649547e-05,
+      "loss": 2.4506,
+      "step": 15261
+    },
+    {
+      "epoch": 1.2317004277298038,
+      "grad_norm": 0.641852855682373,
+      "learning_rate": 2.671526911657015e-05,
+      "loss": 2.4261,
+      "step": 15262
+    },
+    {
+      "epoch": 1.231781131466387,
+      "grad_norm": 0.7627724409103394,
+      "learning_rate": 2.670452873293785e-05,
+      "loss": 2.4647,
+      "step": 15263
+    },
+    {
+      "epoch": 1.23186183520297,
+      "grad_norm": 0.6638163924217224,
+      "learning_rate": 2.669379017602026e-05,
+      "loss": 2.4208,
+      "step": 15264
+    },
+    {
+      "epoch": 1.2319425389395529,
+      "grad_norm": 0.6815361380577087,
+      "learning_rate": 2.668305344608505e-05,
+      "loss": 2.4404,
+      "step": 15265
+    },
+    {
+      "epoch": 1.232023242676136,
+      "grad_norm": 0.6466485857963562,
+      "learning_rate": 2.6672318543399823e-05,
+      "loss": 2.4327,
+      "step": 15266
+    },
+    {
+      "epoch": 1.2321039464127188,
+      "grad_norm": 0.7119305729866028,
+      "learning_rate": 2.6661585468232042e-05,
+      "loss": 2.4266,
+      "step": 15267
+    },
+    {
+      "epoch": 1.232184650149302,
+      "grad_norm": 0.7245718836784363,
+      "learning_rate": 2.6650854220849286e-05,
+      "loss": 2.4484,
+      "step": 15268
+    },
+    {
+      "epoch": 1.232265353885885,
+      "grad_norm": 0.7050287127494812,
+      "learning_rate": 2.6640124801518972e-05,
+      "loss": 2.4441,
+      "step": 15269
+    },
+    {
+      "epoch": 1.2323460576224678,
+      "grad_norm": 0.6906494498252869,
+      "learning_rate": 2.6629397210508556e-05,
+      "loss": 2.4297,
+      "step": 15270
+    },
+    {
+      "epoch": 1.232426761359051,
+      "grad_norm": 0.7224171757698059,
+      "learning_rate": 2.661867144808532e-05,
+      "loss": 2.4279,
+      "step": 15271
+    },
+    {
+      "epoch": 1.232507465095634,
+      "grad_norm": 0.688804030418396,
+      "learning_rate": 2.6607947514516606e-05,
+      "loss": 2.4741,
+      "step": 15272
+    },
+    {
+      "epoch": 1.232588168832217,
+      "grad_norm": 0.6462350487709045,
+      "learning_rate": 2.6597225410069726e-05,
+      "loss": 2.4499,
+      "step": 15273
+    },
+    {
+      "epoch": 1.2326688725688,
+      "grad_norm": 0.6860110759735107,
+      "learning_rate": 2.658650513501184e-05,
+      "loss": 2.4488,
+      "step": 15274
+    },
+    {
+      "epoch": 1.2327495763053828,
+      "grad_norm": 0.7158305644989014,
+      "learning_rate": 2.6575786689610138e-05,
+      "loss": 2.4318,
+      "step": 15275
+    },
+    {
+      "epoch": 1.232830280041966,
+      "grad_norm": 0.7740959525108337,
+      "learning_rate": 2.6565070074131804e-05,
+      "loss": 2.4824,
+      "step": 15276
+    },
+    {
+      "epoch": 1.232910983778549,
+      "grad_norm": 0.7573856711387634,
+      "learning_rate": 2.6554355288843847e-05,
+      "loss": 2.4034,
+      "step": 15277
+    },
+    {
+      "epoch": 1.2329916875151319,
+      "grad_norm": 0.6809369921684265,
+      "learning_rate": 2.654364233401332e-05,
+      "loss": 2.5085,
+      "step": 15278
+    },
+    {
+      "epoch": 1.233072391251715,
+      "grad_norm": 0.6695643067359924,
+      "learning_rate": 2.6532931209907307e-05,
+      "loss": 2.4697,
+      "step": 15279
+    },
+    {
+      "epoch": 1.2331530949882978,
+      "grad_norm": 0.7218750715255737,
+      "learning_rate": 2.6522221916792655e-05,
+      "loss": 2.4753,
+      "step": 15280
+    },
+    {
+      "epoch": 1.233233798724881,
+      "grad_norm": 0.8171822428703308,
+      "learning_rate": 2.6511514454936314e-05,
+      "loss": 2.45,
+      "step": 15281
+    },
+    {
+      "epoch": 1.233314502461464,
+      "grad_norm": 0.7234573364257812,
+      "learning_rate": 2.6500808824605162e-05,
+      "loss": 2.3963,
+      "step": 15282
+    },
+    {
+      "epoch": 1.2333952061980469,
+      "grad_norm": 0.6993409395217896,
+      "learning_rate": 2.6490105026065948e-05,
+      "loss": 2.4449,
+      "step": 15283
+    },
+    {
+      "epoch": 1.23347590993463,
+      "grad_norm": 0.7984449863433838,
+      "learning_rate": 2.6479403059585472e-05,
+      "loss": 2.4322,
+      "step": 15284
+    },
+    {
+      "epoch": 1.233556613671213,
+      "grad_norm": 0.683971107006073,
+      "learning_rate": 2.6468702925430466e-05,
+      "loss": 2.4125,
+      "step": 15285
+    },
+    {
+      "epoch": 1.233637317407796,
+      "grad_norm": 0.6739822626113892,
+      "learning_rate": 2.6458004623867617e-05,
+      "loss": 2.4487,
+      "step": 15286
+    },
+    {
+      "epoch": 1.233718021144379,
+      "grad_norm": 0.7003912925720215,
+      "learning_rate": 2.644730815516351e-05,
+      "loss": 2.4437,
+      "step": 15287
+    },
+    {
+      "epoch": 1.233798724880962,
+      "grad_norm": 0.7011744379997253,
+      "learning_rate": 2.643661351958474e-05,
+      "loss": 2.4798,
+      "step": 15288
+    },
+    {
+      "epoch": 1.233879428617545,
+      "grad_norm": 0.7003397941589355,
+      "learning_rate": 2.6425920717397867e-05,
+      "loss": 2.4554,
+      "step": 15289
+    },
+    {
+      "epoch": 1.233960132354128,
+      "grad_norm": 0.6682165265083313,
+      "learning_rate": 2.6415229748869374e-05,
+      "loss": 2.4252,
+      "step": 15290
+    },
+    {
+      "epoch": 1.234040836090711,
+      "grad_norm": 0.6712457537651062,
+      "learning_rate": 2.6404540614265715e-05,
+      "loss": 2.4225,
+      "step": 15291
+    },
+    {
+      "epoch": 1.234121539827294,
+      "grad_norm": 0.654464602470398,
+      "learning_rate": 2.63938533138533e-05,
+      "loss": 2.4462,
+      "step": 15292
+    },
+    {
+      "epoch": 1.234202243563877,
+      "grad_norm": 0.7311797738075256,
+      "learning_rate": 2.638316784789845e-05,
+      "loss": 2.502,
+      "step": 15293
+    },
+    {
+      "epoch": 1.23428294730046,
+      "grad_norm": 0.6836559176445007,
+      "learning_rate": 2.6372484216667492e-05,
+      "loss": 2.5134,
+      "step": 15294
+    },
+    {
+      "epoch": 1.234363651037043,
+      "grad_norm": 0.6961826086044312,
+      "learning_rate": 2.636180242042672e-05,
+      "loss": 2.4479,
+      "step": 15295
+    },
+    {
+      "epoch": 1.234444354773626,
+      "grad_norm": 0.6824259161949158,
+      "learning_rate": 2.635112245944229e-05,
+      "loss": 2.4299,
+      "step": 15296
+    },
+    {
+      "epoch": 1.234525058510209,
+      "grad_norm": 0.7594609260559082,
+      "learning_rate": 2.634044433398042e-05,
+      "loss": 2.4469,
+      "step": 15297
+    },
+    {
+      "epoch": 1.234605762246792,
+      "grad_norm": 0.7044653296470642,
+      "learning_rate": 2.632976804430721e-05,
+      "loss": 2.447,
+      "step": 15298
+    },
+    {
+      "epoch": 1.234686465983375,
+      "grad_norm": 0.6986916065216064,
+      "learning_rate": 2.631909359068876e-05,
+      "loss": 2.4705,
+      "step": 15299
+    },
+    {
+      "epoch": 1.234767169719958,
+      "grad_norm": 0.7025431990623474,
+      "learning_rate": 2.630842097339111e-05,
+      "loss": 2.3951,
+      "step": 15300
+    },
+    {
+      "epoch": 1.2348478734565411,
+      "grad_norm": 0.6533786058425903,
+      "learning_rate": 2.6297750192680237e-05,
+      "loss": 2.3769,
+      "step": 15301
+    },
+    {
+      "epoch": 1.234928577193124,
+      "grad_norm": 0.6575472354888916,
+      "learning_rate": 2.628708124882212e-05,
+      "loss": 2.4293,
+      "step": 15302
+    },
+    {
+      "epoch": 1.235009280929707,
+      "grad_norm": 0.6712046265602112,
+      "learning_rate": 2.6276414142082584e-05,
+      "loss": 2.4819,
+      "step": 15303
+    },
+    {
+      "epoch": 1.2350899846662902,
+      "grad_norm": 0.6947652101516724,
+      "learning_rate": 2.6265748872727535e-05,
+      "loss": 2.449,
+      "step": 15304
+    },
+    {
+      "epoch": 1.235170688402873,
+      "grad_norm": 0.6881443858146667,
+      "learning_rate": 2.62550854410228e-05,
+      "loss": 2.3991,
+      "step": 15305
+    },
+    {
+      "epoch": 1.2352513921394561,
+      "grad_norm": 0.6681519746780396,
+      "learning_rate": 2.624442384723407e-05,
+      "loss": 2.4005,
+      "step": 15306
+    },
+    {
+      "epoch": 1.235332095876039,
+      "grad_norm": 0.6728120446205139,
+      "learning_rate": 2.62337640916271e-05,
+      "loss": 2.4242,
+      "step": 15307
+    },
+    {
+      "epoch": 1.235412799612622,
+      "grad_norm": 0.707360029220581,
+      "learning_rate": 2.622310617446755e-05,
+      "loss": 2.4385,
+      "step": 15308
+    },
+    {
+      "epoch": 1.2354935033492052,
+      "grad_norm": 0.6890079975128174,
+      "learning_rate": 2.6212450096021058e-05,
+      "loss": 2.443,
+      "step": 15309
+    },
+    {
+      "epoch": 1.235574207085788,
+      "grad_norm": 0.7022379636764526,
+      "learning_rate": 2.620179585655318e-05,
+      "loss": 2.3982,
+      "step": 15310
+    },
+    {
+      "epoch": 1.235654910822371,
+      "grad_norm": 0.7283182740211487,
+      "learning_rate": 2.61911434563295e-05,
+      "loss": 2.4197,
+      "step": 15311
+    },
+    {
+      "epoch": 1.235735614558954,
+      "grad_norm": 0.6721852421760559,
+      "learning_rate": 2.6180492895615426e-05,
+      "loss": 2.4356,
+      "step": 15312
+    },
+    {
+      "epoch": 1.235816318295537,
+      "grad_norm": 0.6817916631698608,
+      "learning_rate": 2.616984417467645e-05,
+      "loss": 2.4325,
+      "step": 15313
+    },
+    {
+      "epoch": 1.2358970220321202,
+      "grad_norm": 0.6826596260070801,
+      "learning_rate": 2.6159197293777972e-05,
+      "loss": 2.4043,
+      "step": 15314
+    },
+    {
+      "epoch": 1.235977725768703,
+      "grad_norm": 0.7135530114173889,
+      "learning_rate": 2.6148552253185288e-05,
+      "loss": 2.4269,
+      "step": 15315
+    },
+    {
+      "epoch": 1.236058429505286,
+      "grad_norm": 0.7027753591537476,
+      "learning_rate": 2.6137909053163722e-05,
+      "loss": 2.4266,
+      "step": 15316
+    },
+    {
+      "epoch": 1.2361391332418692,
+      "grad_norm": 0.6597041487693787,
+      "learning_rate": 2.6127267693978552e-05,
+      "loss": 2.4073,
+      "step": 15317
+    },
+    {
+      "epoch": 1.236219836978452,
+      "grad_norm": 0.6450026631355286,
+      "learning_rate": 2.6116628175894974e-05,
+      "loss": 2.4299,
+      "step": 15318
+    },
+    {
+      "epoch": 1.2363005407150351,
+      "grad_norm": 0.7740476727485657,
+      "learning_rate": 2.6105990499178156e-05,
+      "loss": 2.4088,
+      "step": 15319
+    },
+    {
+      "epoch": 1.2363812444516182,
+      "grad_norm": 0.6460183262825012,
+      "learning_rate": 2.609535466409322e-05,
+      "loss": 2.4311,
+      "step": 15320
+    },
+    {
+      "epoch": 1.236461948188201,
+      "grad_norm": 0.6514838337898254,
+      "learning_rate": 2.608472067090525e-05,
+      "loss": 2.4069,
+      "step": 15321
+    },
+    {
+      "epoch": 1.2365426519247842,
+      "grad_norm": 0.7281234860420227,
+      "learning_rate": 2.6074088519879237e-05,
+      "loss": 2.4245,
+      "step": 15322
+    },
+    {
+      "epoch": 1.236623355661367,
+      "grad_norm": 0.752983570098877,
+      "learning_rate": 2.606345821128018e-05,
+      "loss": 2.4149,
+      "step": 15323
+    },
+    {
+      "epoch": 1.2367040593979501,
+      "grad_norm": 0.6912856101989746,
+      "learning_rate": 2.6052829745373054e-05,
+      "loss": 2.4489,
+      "step": 15324
+    },
+    {
+      "epoch": 1.236784763134533,
+      "grad_norm": 0.6719293594360352,
+      "learning_rate": 2.604220312242267e-05,
+      "loss": 2.457,
+      "step": 15325
+    },
+    {
+      "epoch": 1.236865466871116,
+      "grad_norm": 0.7440586090087891,
+      "learning_rate": 2.6031578342693918e-05,
+      "loss": 2.4657,
+      "step": 15326
+    },
+    {
+      "epoch": 1.2369461706076992,
+      "grad_norm": 0.694442629814148,
+      "learning_rate": 2.602095540645162e-05,
+      "loss": 2.4422,
+      "step": 15327
+    },
+    {
+      "epoch": 1.237026874344282,
+      "grad_norm": 0.7186843752861023,
+      "learning_rate": 2.601033431396046e-05,
+      "loss": 2.4229,
+      "step": 15328
+    },
+    {
+      "epoch": 1.2371075780808651,
+      "grad_norm": 0.7401825785636902,
+      "learning_rate": 2.5999715065485153e-05,
+      "loss": 2.45,
+      "step": 15329
+    },
+    {
+      "epoch": 1.2371882818174482,
+      "grad_norm": 0.6710138916969299,
+      "learning_rate": 2.598909766129045e-05,
+      "loss": 2.4074,
+      "step": 15330
+    },
+    {
+      "epoch": 1.237268985554031,
+      "grad_norm": 0.7867769598960876,
+      "learning_rate": 2.5978482101640867e-05,
+      "loss": 2.4709,
+      "step": 15331
+    },
+    {
+      "epoch": 1.2373496892906142,
+      "grad_norm": 0.7076219916343689,
+      "learning_rate": 2.5967868386801e-05,
+      "loss": 2.4887,
+      "step": 15332
+    },
+    {
+      "epoch": 1.2374303930271973,
+      "grad_norm": 0.7277626991271973,
+      "learning_rate": 2.5957256517035378e-05,
+      "loss": 2.4295,
+      "step": 15333
+    },
+    {
+      "epoch": 1.2375110967637801,
+      "grad_norm": 0.7339804768562317,
+      "learning_rate": 2.5946646492608506e-05,
+      "loss": 2.4624,
+      "step": 15334
+    },
+    {
+      "epoch": 1.2375918005003632,
+      "grad_norm": 0.6707656383514404,
+      "learning_rate": 2.593603831378475e-05,
+      "loss": 2.4159,
+      "step": 15335
+    },
+    {
+      "epoch": 1.237672504236946,
+      "grad_norm": 0.7118813991546631,
+      "learning_rate": 2.592543198082852e-05,
+      "loss": 2.4496,
+      "step": 15336
+    },
+    {
+      "epoch": 1.2377532079735292,
+      "grad_norm": 0.675167977809906,
+      "learning_rate": 2.591482749400419e-05,
+      "loss": 2.4519,
+      "step": 15337
+    },
+    {
+      "epoch": 1.2378339117101123,
+      "grad_norm": 0.8245306611061096,
+      "learning_rate": 2.5904224853575986e-05,
+      "loss": 2.4732,
+      "step": 15338
+    },
+    {
+      "epoch": 1.2379146154466951,
+      "grad_norm": 0.7411863207817078,
+      "learning_rate": 2.5893624059808184e-05,
+      "loss": 2.4458,
+      "step": 15339
+    },
+    {
+      "epoch": 1.2379953191832782,
+      "grad_norm": 0.6864522695541382,
+      "learning_rate": 2.5883025112964997e-05,
+      "loss": 2.4264,
+      "step": 15340
+    },
+    {
+      "epoch": 1.238076022919861,
+      "grad_norm": 0.6585919260978699,
+      "learning_rate": 2.5872428013310567e-05,
+      "loss": 2.3904,
+      "step": 15341
+    },
+    {
+      "epoch": 1.2381567266564442,
+      "grad_norm": 0.6605508327484131,
+      "learning_rate": 2.5861832761108995e-05,
+      "loss": 2.4828,
+      "step": 15342
+    },
+    {
+      "epoch": 1.2382374303930272,
+      "grad_norm": 0.7353223562240601,
+      "learning_rate": 2.5851239356624392e-05,
+      "loss": 2.4335,
+      "step": 15343
+    },
+    {
+      "epoch": 1.2383181341296101,
+      "grad_norm": 0.6907783150672913,
+      "learning_rate": 2.5840647800120688e-05,
+      "loss": 2.4394,
+      "step": 15344
+    },
+    {
+      "epoch": 1.2383988378661932,
+      "grad_norm": 0.7239590287208557,
+      "learning_rate": 2.5830058091861896e-05,
+      "loss": 2.4221,
+      "step": 15345
+    },
+    {
+      "epoch": 1.2384795416027763,
+      "grad_norm": 0.7001412510871887,
+      "learning_rate": 2.5819470232111975e-05,
+      "loss": 2.4521,
+      "step": 15346
+    },
+    {
+      "epoch": 1.2385602453393592,
+      "grad_norm": 0.6983658671379089,
+      "learning_rate": 2.580888422113473e-05,
+      "loss": 2.4839,
+      "step": 15347
+    },
+    {
+      "epoch": 1.2386409490759422,
+      "grad_norm": 0.7829005718231201,
+      "learning_rate": 2.5798300059194037e-05,
+      "loss": 2.4546,
+      "step": 15348
+    },
+    {
+      "epoch": 1.2387216528125253,
+      "grad_norm": 0.7248061299324036,
+      "learning_rate": 2.5787717746553664e-05,
+      "loss": 2.4341,
+      "step": 15349
+    },
+    {
+      "epoch": 1.2388023565491082,
+      "grad_norm": 0.7921163439750671,
+      "learning_rate": 2.577713728347736e-05,
+      "loss": 2.475,
+      "step": 15350
+    },
+    {
+      "epoch": 1.2388830602856913,
+      "grad_norm": 0.6571238040924072,
+      "learning_rate": 2.5766558670228813e-05,
+      "loss": 2.4636,
+      "step": 15351
+    },
+    {
+      "epoch": 1.2389637640222741,
+      "grad_norm": 0.7436683177947998,
+      "learning_rate": 2.575598190707168e-05,
+      "loss": 2.4868,
+      "step": 15352
+    },
+    {
+      "epoch": 1.2390444677588572,
+      "grad_norm": 0.6471900939941406,
+      "learning_rate": 2.5745406994269573e-05,
+      "loss": 2.4349,
+      "step": 15353
+    },
+    {
+      "epoch": 1.2391251714954403,
+      "grad_norm": 0.6612011194229126,
+      "learning_rate": 2.5734833932086012e-05,
+      "loss": 2.4088,
+      "step": 15354
+    },
+    {
+      "epoch": 1.2392058752320232,
+      "grad_norm": 0.6882977485656738,
+      "learning_rate": 2.572426272078451e-05,
+      "loss": 2.4344,
+      "step": 15355
+    },
+    {
+      "epoch": 1.2392865789686063,
+      "grad_norm": 0.6836830973625183,
+      "learning_rate": 2.5713693360628565e-05,
+      "loss": 2.4325,
+      "step": 15356
+    },
+    {
+      "epoch": 1.2393672827051891,
+      "grad_norm": 0.712127149105072,
+      "learning_rate": 2.5703125851881536e-05,
+      "loss": 2.4505,
+      "step": 15357
+    },
+    {
+      "epoch": 1.2394479864417722,
+      "grad_norm": 0.7162468433380127,
+      "learning_rate": 2.5692560194806837e-05,
+      "loss": 2.4167,
+      "step": 15358
+    },
+    {
+      "epoch": 1.2395286901783553,
+      "grad_norm": 0.7770177125930786,
+      "learning_rate": 2.568199638966777e-05,
+      "loss": 2.4072,
+      "step": 15359
+    },
+    {
+      "epoch": 1.2396093939149382,
+      "grad_norm": 0.7049651741981506,
+      "learning_rate": 2.5671434436727636e-05,
+      "loss": 2.434,
+      "step": 15360
+    },
+    {
+      "epoch": 1.2396900976515213,
+      "grad_norm": 0.7793349027633667,
+      "learning_rate": 2.566087433624964e-05,
+      "loss": 2.4762,
+      "step": 15361
+    },
+    {
+      "epoch": 1.2397708013881044,
+      "grad_norm": 0.6776690483093262,
+      "learning_rate": 2.5650316088497018e-05,
+      "loss": 2.402,
+      "step": 15362
+    },
+    {
+      "epoch": 1.2398515051246872,
+      "grad_norm": 0.7207701802253723,
+      "learning_rate": 2.5639759693732834e-05,
+      "loss": 2.4398,
+      "step": 15363
+    },
+    {
+      "epoch": 1.2399322088612703,
+      "grad_norm": 0.759787917137146,
+      "learning_rate": 2.5629205152220215e-05,
+      "loss": 2.4268,
+      "step": 15364
+    },
+    {
+      "epoch": 1.2400129125978534,
+      "grad_norm": 0.6906142830848694,
+      "learning_rate": 2.5618652464222215e-05,
+      "loss": 2.4075,
+      "step": 15365
+    },
+    {
+      "epoch": 1.2400936163344363,
+      "grad_norm": 0.7002954483032227,
+      "learning_rate": 2.560810163000187e-05,
+      "loss": 2.4516,
+      "step": 15366
+    },
+    {
+      "epoch": 1.2401743200710194,
+      "grad_norm": 0.7287559509277344,
+      "learning_rate": 2.5597552649822053e-05,
+      "loss": 2.4975,
+      "step": 15367
+    },
+    {
+      "epoch": 1.2402550238076022,
+      "grad_norm": 0.6523926854133606,
+      "learning_rate": 2.558700552394572e-05,
+      "loss": 2.4085,
+      "step": 15368
+    },
+    {
+      "epoch": 1.2403357275441853,
+      "grad_norm": 0.7289387583732605,
+      "learning_rate": 2.5576460252635727e-05,
+      "loss": 2.4789,
+      "step": 15369
+    },
+    {
+      "epoch": 1.2404164312807684,
+      "grad_norm": 0.6613432765007019,
+      "learning_rate": 2.5565916836154878e-05,
+      "loss": 2.4263,
+      "step": 15370
+    },
+    {
+      "epoch": 1.2404971350173513,
+      "grad_norm": 0.7275245785713196,
+      "learning_rate": 2.555537527476597e-05,
+      "loss": 2.4652,
+      "step": 15371
+    },
+    {
+      "epoch": 1.2405778387539343,
+      "grad_norm": 0.6726976037025452,
+      "learning_rate": 2.554483556873173e-05,
+      "loss": 2.4092,
+      "step": 15372
+    },
+    {
+      "epoch": 1.2406585424905172,
+      "grad_norm": 0.6908233761787415,
+      "learning_rate": 2.5534297718314794e-05,
+      "loss": 2.3678,
+      "step": 15373
+    },
+    {
+      "epoch": 1.2407392462271003,
+      "grad_norm": 0.6893147826194763,
+      "learning_rate": 2.5523761723777806e-05,
+      "loss": 2.4625,
+      "step": 15374
+    },
+    {
+      "epoch": 1.2408199499636834,
+      "grad_norm": 0.7640267014503479,
+      "learning_rate": 2.551322758538339e-05,
+      "loss": 2.446,
+      "step": 15375
+    },
+    {
+      "epoch": 1.2409006537002663,
+      "grad_norm": 0.7187458276748657,
+      "learning_rate": 2.550269530339402e-05,
+      "loss": 2.4215,
+      "step": 15376
+    },
+    {
+      "epoch": 1.2409813574368493,
+      "grad_norm": 0.8041789531707764,
+      "learning_rate": 2.5492164878072234e-05,
+      "loss": 2.5085,
+      "step": 15377
+    },
+    {
+      "epoch": 1.2410620611734324,
+      "grad_norm": 0.6582188010215759,
+      "learning_rate": 2.5481636309680445e-05,
+      "loss": 2.467,
+      "step": 15378
+    },
+    {
+      "epoch": 1.2411427649100153,
+      "grad_norm": 0.705731213092804,
+      "learning_rate": 2.5471109598481112e-05,
+      "loss": 2.3764,
+      "step": 15379
+    },
+    {
+      "epoch": 1.2412234686465984,
+      "grad_norm": 0.6918940544128418,
+      "learning_rate": 2.5460584744736495e-05,
+      "loss": 2.4513,
+      "step": 15380
+    },
+    {
+      "epoch": 1.2413041723831812,
+      "grad_norm": 0.7402673959732056,
+      "learning_rate": 2.5450061748708975e-05,
+      "loss": 2.5133,
+      "step": 15381
+    },
+    {
+      "epoch": 1.2413848761197643,
+      "grad_norm": 0.6740667223930359,
+      "learning_rate": 2.543954061066083e-05,
+      "loss": 2.4649,
+      "step": 15382
+    },
+    {
+      "epoch": 1.2414655798563474,
+      "grad_norm": 0.6665407419204712,
+      "learning_rate": 2.5429021330854197e-05,
+      "loss": 2.4321,
+      "step": 15383
+    },
+    {
+      "epoch": 1.2415462835929303,
+      "grad_norm": 0.7324530482292175,
+      "learning_rate": 2.5418503909551296e-05,
+      "loss": 2.3574,
+      "step": 15384
+    },
+    {
+      "epoch": 1.2416269873295134,
+      "grad_norm": 0.7117868661880493,
+      "learning_rate": 2.5407988347014255e-05,
+      "loss": 2.4552,
+      "step": 15385
+    },
+    {
+      "epoch": 1.2417076910660962,
+      "grad_norm": 0.7162930965423584,
+      "learning_rate": 2.5397474643505103e-05,
+      "loss": 2.4135,
+      "step": 15386
+    },
+    {
+      "epoch": 1.2417883948026793,
+      "grad_norm": 0.7301257848739624,
+      "learning_rate": 2.5386962799285895e-05,
+      "loss": 2.4277,
+      "step": 15387
+    },
+    {
+      "epoch": 1.2418690985392624,
+      "grad_norm": 0.7404977679252625,
+      "learning_rate": 2.5376452814618645e-05,
+      "loss": 2.478,
+      "step": 15388
+    },
+    {
+      "epoch": 1.2419498022758453,
+      "grad_norm": 0.6546272039413452,
+      "learning_rate": 2.536594468976522e-05,
+      "loss": 2.4879,
+      "step": 15389
+    },
+    {
+      "epoch": 1.2420305060124284,
+      "grad_norm": 0.6501599550247192,
+      "learning_rate": 2.5355438424987565e-05,
+      "loss": 2.3964,
+      "step": 15390
+    },
+    {
+      "epoch": 1.2421112097490115,
+      "grad_norm": 0.6711748242378235,
+      "learning_rate": 2.5344934020547496e-05,
+      "loss": 2.4123,
+      "step": 15391
+    },
+    {
+      "epoch": 1.2421919134855943,
+      "grad_norm": 0.6803534030914307,
+      "learning_rate": 2.5334431476706823e-05,
+      "loss": 2.4271,
+      "step": 15392
+    },
+    {
+      "epoch": 1.2422726172221774,
+      "grad_norm": 0.7407296299934387,
+      "learning_rate": 2.5323930793727302e-05,
+      "loss": 2.49,
+      "step": 15393
+    },
+    {
+      "epoch": 1.2423533209587605,
+      "grad_norm": 0.701870858669281,
+      "learning_rate": 2.5313431971870617e-05,
+      "loss": 2.4534,
+      "step": 15394
+    },
+    {
+      "epoch": 1.2424340246953434,
+      "grad_norm": 0.6658090353012085,
+      "learning_rate": 2.5302935011398475e-05,
+      "loss": 2.4581,
+      "step": 15395
+    },
+    {
+      "epoch": 1.2425147284319265,
+      "grad_norm": 0.6616473197937012,
+      "learning_rate": 2.529243991257243e-05,
+      "loss": 2.4169,
+      "step": 15396
+    },
+    {
+      "epoch": 1.2425954321685093,
+      "grad_norm": 0.6714773178100586,
+      "learning_rate": 2.5281946675654067e-05,
+      "loss": 2.4159,
+      "step": 15397
+    },
+    {
+      "epoch": 1.2426761359050924,
+      "grad_norm": 0.6789337396621704,
+      "learning_rate": 2.5271455300904935e-05,
+      "loss": 2.4211,
+      "step": 15398
+    },
+    {
+      "epoch": 1.2427568396416755,
+      "grad_norm": 0.6793739795684814,
+      "learning_rate": 2.5260965788586456e-05,
+      "loss": 2.4337,
+      "step": 15399
+    },
+    {
+      "epoch": 1.2428375433782584,
+      "grad_norm": 0.6432294249534607,
+      "learning_rate": 2.5250478138960076e-05,
+      "loss": 2.4268,
+      "step": 15400
+    },
+    {
+      "epoch": 1.2429182471148414,
+      "grad_norm": 0.6960669159889221,
+      "learning_rate": 2.523999235228718e-05,
+      "loss": 2.3535,
+      "step": 15401
+    },
+    {
+      "epoch": 1.2429989508514243,
+      "grad_norm": 0.6724488735198975,
+      "learning_rate": 2.5229508428829096e-05,
+      "loss": 2.4294,
+      "step": 15402
+    },
+    {
+      "epoch": 1.2430796545880074,
+      "grad_norm": 0.636105477809906,
+      "learning_rate": 2.521902636884711e-05,
+      "loss": 2.4438,
+      "step": 15403
+    },
+    {
+      "epoch": 1.2431603583245905,
+      "grad_norm": 0.6865580677986145,
+      "learning_rate": 2.52085461726025e-05,
+      "loss": 2.4473,
+      "step": 15404
+    },
+    {
+      "epoch": 1.2432410620611734,
+      "grad_norm": 0.6740261316299438,
+      "learning_rate": 2.5198067840356398e-05,
+      "loss": 2.4642,
+      "step": 15405
+    },
+    {
+      "epoch": 1.2433217657977564,
+      "grad_norm": 0.7241789698600769,
+      "learning_rate": 2.518759137236998e-05,
+      "loss": 2.4294,
+      "step": 15406
+    },
+    {
+      "epoch": 1.2434024695343395,
+      "grad_norm": 0.6839794516563416,
+      "learning_rate": 2.5177116768904373e-05,
+      "loss": 2.4697,
+      "step": 15407
+    },
+    {
+      "epoch": 1.2434831732709224,
+      "grad_norm": 0.677390992641449,
+      "learning_rate": 2.5166644030220578e-05,
+      "loss": 2.4411,
+      "step": 15408
+    },
+    {
+      "epoch": 1.2435638770075055,
+      "grad_norm": 0.709065854549408,
+      "learning_rate": 2.515617315657962e-05,
+      "loss": 2.4392,
+      "step": 15409
+    },
+    {
+      "epoch": 1.2436445807440886,
+      "grad_norm": 0.6735498905181885,
+      "learning_rate": 2.514570414824249e-05,
+      "loss": 2.3924,
+      "step": 15410
+    },
+    {
+      "epoch": 1.2437252844806714,
+      "grad_norm": 0.6729374527931213,
+      "learning_rate": 2.513523700547007e-05,
+      "loss": 2.4464,
+      "step": 15411
+    },
+    {
+      "epoch": 1.2438059882172545,
+      "grad_norm": 0.7232720851898193,
+      "learning_rate": 2.5124771728523244e-05,
+      "loss": 2.3975,
+      "step": 15412
+    },
+    {
+      "epoch": 1.2438866919538374,
+      "grad_norm": 0.7467584609985352,
+      "learning_rate": 2.5114308317662837e-05,
+      "loss": 2.4191,
+      "step": 15413
+    },
+    {
+      "epoch": 1.2439673956904205,
+      "grad_norm": 0.6951141953468323,
+      "learning_rate": 2.5103846773149642e-05,
+      "loss": 2.4207,
+      "step": 15414
+    },
+    {
+      "epoch": 1.2440480994270036,
+      "grad_norm": 0.6427489519119263,
+      "learning_rate": 2.5093387095244336e-05,
+      "loss": 2.3539,
+      "step": 15415
+    },
+    {
+      "epoch": 1.2441288031635864,
+      "grad_norm": 0.729580283164978,
+      "learning_rate": 2.5082929284207644e-05,
+      "loss": 2.4464,
+      "step": 15416
+    },
+    {
+      "epoch": 1.2442095069001695,
+      "grad_norm": 0.7247009873390198,
+      "learning_rate": 2.5072473340300207e-05,
+      "loss": 2.4294,
+      "step": 15417
+    },
+    {
+      "epoch": 1.2442902106367524,
+      "grad_norm": 0.7037674784660339,
+      "learning_rate": 2.5062019263782577e-05,
+      "loss": 2.4294,
+      "step": 15418
+    },
+    {
+      "epoch": 1.2443709143733355,
+      "grad_norm": 0.6997841596603394,
+      "learning_rate": 2.5051567054915303e-05,
+      "loss": 2.4976,
+      "step": 15419
+    },
+    {
+      "epoch": 1.2444516181099186,
+      "grad_norm": 0.7001172304153442,
+      "learning_rate": 2.504111671395891e-05,
+      "loss": 2.371,
+      "step": 15420
+    },
+    {
+      "epoch": 1.2445323218465014,
+      "grad_norm": 0.6781473159790039,
+      "learning_rate": 2.5030668241173827e-05,
+      "loss": 2.4124,
+      "step": 15421
+    },
+    {
+      "epoch": 1.2446130255830845,
+      "grad_norm": 0.7053182125091553,
+      "learning_rate": 2.5020221636820463e-05,
+      "loss": 2.4109,
+      "step": 15422
+    },
+    {
+      "epoch": 1.2446937293196676,
+      "grad_norm": 0.68635493516922,
+      "learning_rate": 2.50097769011592e-05,
+      "loss": 2.4548,
+      "step": 15423
+    },
+    {
+      "epoch": 1.2447744330562505,
+      "grad_norm": 0.7015564441680908,
+      "learning_rate": 2.4999334034450293e-05,
+      "loss": 2.4537,
+      "step": 15424
+    },
+    {
+      "epoch": 1.2448551367928335,
+      "grad_norm": 0.694054901599884,
+      "learning_rate": 2.4988893036954043e-05,
+      "loss": 2.4396,
+      "step": 15425
+    },
+    {
+      "epoch": 1.2449358405294164,
+      "grad_norm": 0.702518880367279,
+      "learning_rate": 2.4978453908930665e-05,
+      "loss": 2.4015,
+      "step": 15426
+    },
+    {
+      "epoch": 1.2450165442659995,
+      "grad_norm": 0.7237387895584106,
+      "learning_rate": 2.4968016650640348e-05,
+      "loss": 2.4257,
+      "step": 15427
+    },
+    {
+      "epoch": 1.2450972480025826,
+      "grad_norm": 0.7133163809776306,
+      "learning_rate": 2.4957581262343154e-05,
+      "loss": 2.4532,
+      "step": 15428
+    },
+    {
+      "epoch": 1.2451779517391655,
+      "grad_norm": 0.8339287042617798,
+      "learning_rate": 2.4947147744299203e-05,
+      "loss": 2.4621,
+      "step": 15429
+    },
+    {
+      "epoch": 1.2452586554757485,
+      "grad_norm": 0.7620034217834473,
+      "learning_rate": 2.493671609676852e-05,
+      "loss": 2.365,
+      "step": 15430
+    },
+    {
+      "epoch": 1.2453393592123314,
+      "grad_norm": 0.7445465922355652,
+      "learning_rate": 2.4926286320011094e-05,
+      "loss": 2.4764,
+      "step": 15431
+    },
+    {
+      "epoch": 1.2454200629489145,
+      "grad_norm": 0.7366160154342651,
+      "learning_rate": 2.4915858414286852e-05,
+      "loss": 2.4597,
+      "step": 15432
+    },
+    {
+      "epoch": 1.2455007666854976,
+      "grad_norm": 0.7098437547683716,
+      "learning_rate": 2.490543237985572e-05,
+      "loss": 2.4202,
+      "step": 15433
+    },
+    {
+      "epoch": 1.2455814704220805,
+      "grad_norm": 0.6483333706855774,
+      "learning_rate": 2.4895008216977478e-05,
+      "loss": 2.4108,
+      "step": 15434
+    },
+    {
+      "epoch": 1.2456621741586635,
+      "grad_norm": 0.6797904968261719,
+      "learning_rate": 2.4884585925911963e-05,
+      "loss": 2.4414,
+      "step": 15435
+    },
+    {
+      "epoch": 1.2457428778952466,
+      "grad_norm": 0.6853424310684204,
+      "learning_rate": 2.4874165506918957e-05,
+      "loss": 2.4226,
+      "step": 15436
+    },
+    {
+      "epoch": 1.2458235816318295,
+      "grad_norm": 0.6861590147018433,
+      "learning_rate": 2.4863746960258094e-05,
+      "loss": 2.3748,
+      "step": 15437
+    },
+    {
+      "epoch": 1.2459042853684126,
+      "grad_norm": 0.7360263466835022,
+      "learning_rate": 2.4853330286189058e-05,
+      "loss": 2.4441,
+      "step": 15438
+    },
+    {
+      "epoch": 1.2459849891049957,
+      "grad_norm": 0.6894183158874512,
+      "learning_rate": 2.4842915484971496e-05,
+      "loss": 2.3495,
+      "step": 15439
+    },
+    {
+      "epoch": 1.2460656928415785,
+      "grad_norm": 0.7570669651031494,
+      "learning_rate": 2.4832502556864923e-05,
+      "loss": 2.4622,
+      "step": 15440
+    },
+    {
+      "epoch": 1.2461463965781616,
+      "grad_norm": 0.6986069083213806,
+      "learning_rate": 2.4822091502128876e-05,
+      "loss": 2.3647,
+      "step": 15441
+    },
+    {
+      "epoch": 1.2462271003147445,
+      "grad_norm": 0.681450366973877,
+      "learning_rate": 2.481168232102279e-05,
+      "loss": 2.3872,
+      "step": 15442
+    },
+    {
+      "epoch": 1.2463078040513276,
+      "grad_norm": 0.7241837978363037,
+      "learning_rate": 2.480127501380618e-05,
+      "loss": 2.4692,
+      "step": 15443
+    },
+    {
+      "epoch": 1.2463885077879107,
+      "grad_norm": 0.6575295329093933,
+      "learning_rate": 2.479086958073834e-05,
+      "loss": 2.5057,
+      "step": 15444
+    },
+    {
+      "epoch": 1.2464692115244935,
+      "grad_norm": 0.7289770841598511,
+      "learning_rate": 2.478046602207864e-05,
+      "loss": 2.4164,
+      "step": 15445
+    },
+    {
+      "epoch": 1.2465499152610766,
+      "grad_norm": 0.6682024598121643,
+      "learning_rate": 2.4770064338086374e-05,
+      "loss": 2.4466,
+      "step": 15446
+    },
+    {
+      "epoch": 1.2466306189976595,
+      "grad_norm": 0.7238918542861938,
+      "learning_rate": 2.475966452902072e-05,
+      "loss": 2.4367,
+      "step": 15447
+    },
+    {
+      "epoch": 1.2467113227342426,
+      "grad_norm": 0.6825705170631409,
+      "learning_rate": 2.4749266595140918e-05,
+      "loss": 2.4337,
+      "step": 15448
+    },
+    {
+      "epoch": 1.2467920264708257,
+      "grad_norm": 0.7352269887924194,
+      "learning_rate": 2.4738870536706126e-05,
+      "loss": 2.4103,
+      "step": 15449
+    },
+    {
+      "epoch": 1.2468727302074085,
+      "grad_norm": 0.658930778503418,
+      "learning_rate": 2.4728476353975394e-05,
+      "loss": 2.4281,
+      "step": 15450
+    },
+    {
+      "epoch": 1.2469534339439916,
+      "grad_norm": 0.6933601498603821,
+      "learning_rate": 2.4718084047207778e-05,
+      "loss": 2.4502,
+      "step": 15451
+    },
+    {
+      "epoch": 1.2470341376805747,
+      "grad_norm": 0.6901879906654358,
+      "learning_rate": 2.4707693616662308e-05,
+      "loss": 2.4057,
+      "step": 15452
+    },
+    {
+      "epoch": 1.2471148414171576,
+      "grad_norm": 0.7648913860321045,
+      "learning_rate": 2.469730506259792e-05,
+      "loss": 2.4163,
+      "step": 15453
+    },
+    {
+      "epoch": 1.2471955451537406,
+      "grad_norm": 0.6496175527572632,
+      "learning_rate": 2.4686918385273537e-05,
+      "loss": 2.4373,
+      "step": 15454
+    },
+    {
+      "epoch": 1.2472762488903237,
+      "grad_norm": 0.6949105858802795,
+      "learning_rate": 2.4676533584948048e-05,
+      "loss": 2.4108,
+      "step": 15455
+    },
+    {
+      "epoch": 1.2473569526269066,
+      "grad_norm": 0.7018688321113586,
+      "learning_rate": 2.4666150661880206e-05,
+      "loss": 2.4589,
+      "step": 15456
+    },
+    {
+      "epoch": 1.2474376563634897,
+      "grad_norm": 0.7141219973564148,
+      "learning_rate": 2.4655769616328827e-05,
+      "loss": 2.4022,
+      "step": 15457
+    },
+    {
+      "epoch": 1.2475183601000726,
+      "grad_norm": 0.7276743054389954,
+      "learning_rate": 2.4645390448552608e-05,
+      "loss": 2.4443,
+      "step": 15458
+    },
+    {
+      "epoch": 1.2475990638366556,
+      "grad_norm": 0.6861153244972229,
+      "learning_rate": 2.463501315881027e-05,
+      "loss": 2.4478,
+      "step": 15459
+    },
+    {
+      "epoch": 1.2476797675732387,
+      "grad_norm": 0.7252256274223328,
+      "learning_rate": 2.462463774736038e-05,
+      "loss": 2.446,
+      "step": 15460
+    },
+    {
+      "epoch": 1.2477604713098216,
+      "grad_norm": 0.6914857625961304,
+      "learning_rate": 2.4614264214461557e-05,
+      "loss": 2.4294,
+      "step": 15461
+    },
+    {
+      "epoch": 1.2478411750464047,
+      "grad_norm": 0.6815036535263062,
+      "learning_rate": 2.460389256037232e-05,
+      "loss": 2.4389,
+      "step": 15462
+    },
+    {
+      "epoch": 1.2479218787829875,
+      "grad_norm": 0.7420194745063782,
+      "learning_rate": 2.4593522785351176e-05,
+      "loss": 2.4932,
+      "step": 15463
+    },
+    {
+      "epoch": 1.2480025825195706,
+      "grad_norm": 0.6622182130813599,
+      "learning_rate": 2.4583154889656556e-05,
+      "loss": 2.4327,
+      "step": 15464
+    },
+    {
+      "epoch": 1.2480832862561537,
+      "grad_norm": 0.6527934074401855,
+      "learning_rate": 2.457278887354689e-05,
+      "loss": 2.3857,
+      "step": 15465
+    },
+    {
+      "epoch": 1.2481639899927366,
+      "grad_norm": 0.6942344903945923,
+      "learning_rate": 2.4562424737280465e-05,
+      "loss": 2.4181,
+      "step": 15466
+    },
+    {
+      "epoch": 1.2482446937293197,
+      "grad_norm": 0.7449823021888733,
+      "learning_rate": 2.45520624811156e-05,
+      "loss": 2.4575,
+      "step": 15467
+    },
+    {
+      "epoch": 1.2483253974659028,
+      "grad_norm": 0.6905208826065063,
+      "learning_rate": 2.4541702105310605e-05,
+      "loss": 2.3858,
+      "step": 15468
+    },
+    {
+      "epoch": 1.2484061012024856,
+      "grad_norm": 0.6928502917289734,
+      "learning_rate": 2.4531343610123603e-05,
+      "loss": 2.4212,
+      "step": 15469
+    },
+    {
+      "epoch": 1.2484868049390687,
+      "grad_norm": 0.7182145118713379,
+      "learning_rate": 2.45209869958128e-05,
+      "loss": 2.4063,
+      "step": 15470
+    },
+    {
+      "epoch": 1.2485675086756518,
+      "grad_norm": 0.7379452586174011,
+      "learning_rate": 2.4510632262636314e-05,
+      "loss": 2.4612,
+      "step": 15471
+    },
+    {
+      "epoch": 1.2486482124122347,
+      "grad_norm": 0.6663349270820618,
+      "learning_rate": 2.450027941085219e-05,
+      "loss": 2.4583,
+      "step": 15472
+    },
+    {
+      "epoch": 1.2487289161488178,
+      "grad_norm": 0.7266560792922974,
+      "learning_rate": 2.4489928440718467e-05,
+      "loss": 2.4483,
+      "step": 15473
+    },
+    {
+      "epoch": 1.2488096198854006,
+      "grad_norm": 0.7046550512313843,
+      "learning_rate": 2.447957935249311e-05,
+      "loss": 2.4087,
+      "step": 15474
+    },
+    {
+      "epoch": 1.2488903236219837,
+      "grad_norm": 0.684248685836792,
+      "learning_rate": 2.4469232146434084e-05,
+      "loss": 2.4352,
+      "step": 15475
+    },
+    {
+      "epoch": 1.2489710273585668,
+      "grad_norm": 0.6864973902702332,
+      "learning_rate": 2.4458886822799198e-05,
+      "loss": 2.3872,
+      "step": 15476
+    },
+    {
+      "epoch": 1.2490517310951497,
+      "grad_norm": 0.6964752674102783,
+      "learning_rate": 2.444854338184631e-05,
+      "loss": 2.437,
+      "step": 15477
+    },
+    {
+      "epoch": 1.2491324348317328,
+      "grad_norm": 0.6755973100662231,
+      "learning_rate": 2.4438201823833252e-05,
+      "loss": 2.4302,
+      "step": 15478
+    },
+    {
+      "epoch": 1.2492131385683156,
+      "grad_norm": 0.6434857249259949,
+      "learning_rate": 2.44278621490177e-05,
+      "loss": 2.406,
+      "step": 15479
+    },
+    {
+      "epoch": 1.2492938423048987,
+      "grad_norm": 0.7342328429222107,
+      "learning_rate": 2.441752435765736e-05,
+      "loss": 2.451,
+      "step": 15480
+    },
+    {
+      "epoch": 1.2493745460414818,
+      "grad_norm": 0.7486860752105713,
+      "learning_rate": 2.44071884500099e-05,
+      "loss": 2.4536,
+      "step": 15481
+    },
+    {
+      "epoch": 1.2494552497780647,
+      "grad_norm": 0.7274537086486816,
+      "learning_rate": 2.4396854426332903e-05,
+      "loss": 2.4599,
+      "step": 15482
+    },
+    {
+      "epoch": 1.2495359535146477,
+      "grad_norm": 0.7580124735832214,
+      "learning_rate": 2.4386522286883918e-05,
+      "loss": 2.4038,
+      "step": 15483
+    },
+    {
+      "epoch": 1.2496166572512308,
+      "grad_norm": 0.6776975393295288,
+      "learning_rate": 2.4376192031920488e-05,
+      "loss": 2.4246,
+      "step": 15484
+    },
+    {
+      "epoch": 1.2496973609878137,
+      "grad_norm": 0.6899511814117432,
+      "learning_rate": 2.4365863661699996e-05,
+      "loss": 2.3922,
+      "step": 15485
+    },
+    {
+      "epoch": 1.2497780647243968,
+      "grad_norm": 0.7487930059432983,
+      "learning_rate": 2.4355537176479903e-05,
+      "loss": 2.4573,
+      "step": 15486
+    },
+    {
+      "epoch": 1.2498587684609797,
+      "grad_norm": 0.7306599617004395,
+      "learning_rate": 2.4345212576517575e-05,
+      "loss": 2.4745,
+      "step": 15487
+    },
+    {
+      "epoch": 1.2499394721975627,
+      "grad_norm": 0.7152543067932129,
+      "learning_rate": 2.43348898620703e-05,
+      "loss": 2.4768,
+      "step": 15488
+    },
+    {
+      "epoch": 1.2500201759341458,
+      "grad_norm": 0.6576277017593384,
+      "learning_rate": 2.432456903339535e-05,
+      "loss": 2.4289,
+      "step": 15489
+    },
+    {
+      "epoch": 1.2501008796707287,
+      "grad_norm": 0.6974572539329529,
+      "learning_rate": 2.4314250090749956e-05,
+      "loss": 2.4218,
+      "step": 15490
+    },
+    {
+      "epoch": 1.2501815834073118,
+      "grad_norm": 0.7869577407836914,
+      "learning_rate": 2.4303933034391323e-05,
+      "loss": 2.3899,
+      "step": 15491
+    },
+    {
+      "epoch": 1.2502622871438946,
+      "grad_norm": 0.6723129749298096,
+      "learning_rate": 2.42936178645765e-05,
+      "loss": 2.4238,
+      "step": 15492
+    },
+    {
+      "epoch": 1.2503429908804777,
+      "grad_norm": 0.6839526891708374,
+      "learning_rate": 2.428330458156265e-05,
+      "loss": 2.4037,
+      "step": 15493
+    },
+    {
+      "epoch": 1.2504236946170608,
+      "grad_norm": 0.6866093277931213,
+      "learning_rate": 2.4272993185606796e-05,
+      "loss": 2.4228,
+      "step": 15494
+    },
+    {
+      "epoch": 1.2505043983536437,
+      "grad_norm": 0.6992947459220886,
+      "learning_rate": 2.426268367696588e-05,
+      "loss": 2.4248,
+      "step": 15495
+    },
+    {
+      "epoch": 1.2505851020902268,
+      "grad_norm": 0.6836698651313782,
+      "learning_rate": 2.4252376055896862e-05,
+      "loss": 2.5387,
+      "step": 15496
+    },
+    {
+      "epoch": 1.2506658058268099,
+      "grad_norm": 0.6990752816200256,
+      "learning_rate": 2.4242070322656663e-05,
+      "loss": 2.4438,
+      "step": 15497
+    },
+    {
+      "epoch": 1.2507465095633927,
+      "grad_norm": 0.7143029570579529,
+      "learning_rate": 2.4231766477502082e-05,
+      "loss": 2.4,
+      "step": 15498
+    },
+    {
+      "epoch": 1.2508272132999758,
+      "grad_norm": 0.6585043668746948,
+      "learning_rate": 2.422146452068994e-05,
+      "loss": 2.4256,
+      "step": 15499
+    },
+    {
+      "epoch": 1.250907917036559,
+      "grad_norm": 0.739107072353363,
+      "learning_rate": 2.421116445247702e-05,
+      "loss": 2.428,
+      "step": 15500
+    },
+    {
+      "epoch": 1.2509886207731418,
+      "grad_norm": 0.6675287485122681,
+      "learning_rate": 2.420086627311997e-05,
+      "loss": 2.5095,
+      "step": 15501
+    },
+    {
+      "epoch": 1.2510693245097249,
+      "grad_norm": 0.7133405804634094,
+      "learning_rate": 2.4190569982875467e-05,
+      "loss": 2.4719,
+      "step": 15502
+    },
+    {
+      "epoch": 1.2511500282463077,
+      "grad_norm": 0.710904061794281,
+      "learning_rate": 2.4180275582000134e-05,
+      "loss": 2.4449,
+      "step": 15503
+    },
+    {
+      "epoch": 1.2512307319828908,
+      "grad_norm": 0.7088729739189148,
+      "learning_rate": 2.4169983070750525e-05,
+      "loss": 2.4059,
+      "step": 15504
+    },
+    {
+      "epoch": 1.2513114357194737,
+      "grad_norm": 0.7187358736991882,
+      "learning_rate": 2.4159692449383152e-05,
+      "loss": 2.4577,
+      "step": 15505
+    },
+    {
+      "epoch": 1.2513921394560568,
+      "grad_norm": 0.7531955242156982,
+      "learning_rate": 2.4149403718154497e-05,
+      "loss": 2.4101,
+      "step": 15506
+    },
+    {
+      "epoch": 1.2514728431926398,
+      "grad_norm": 0.7565199136734009,
+      "learning_rate": 2.413911687732101e-05,
+      "loss": 2.4805,
+      "step": 15507
+    },
+    {
+      "epoch": 1.2515535469292227,
+      "grad_norm": 0.706471860408783,
+      "learning_rate": 2.4128831927139008e-05,
+      "loss": 2.4494,
+      "step": 15508
+    },
+    {
+      "epoch": 1.2516342506658058,
+      "grad_norm": 0.7022314667701721,
+      "learning_rate": 2.4118548867864832e-05,
+      "loss": 2.4442,
+      "step": 15509
+    },
+    {
+      "epoch": 1.251714954402389,
+      "grad_norm": 0.6885591745376587,
+      "learning_rate": 2.4108267699754806e-05,
+      "loss": 2.4186,
+      "step": 15510
+    },
+    {
+      "epoch": 1.2517956581389718,
+      "grad_norm": 0.6963610649108887,
+      "learning_rate": 2.409798842306511e-05,
+      "loss": 2.4209,
+      "step": 15511
+    },
+    {
+      "epoch": 1.2518763618755548,
+      "grad_norm": 0.7117185592651367,
+      "learning_rate": 2.4087711038051942e-05,
+      "loss": 2.4106,
+      "step": 15512
+    },
+    {
+      "epoch": 1.251957065612138,
+      "grad_norm": 0.6944519281387329,
+      "learning_rate": 2.407743554497146e-05,
+      "loss": 2.4493,
+      "step": 15513
+    },
+    {
+      "epoch": 1.2520377693487208,
+      "grad_norm": 0.689818263053894,
+      "learning_rate": 2.406716194407974e-05,
+      "loss": 2.4358,
+      "step": 15514
+    },
+    {
+      "epoch": 1.2521184730853039,
+      "grad_norm": 0.8132768273353577,
+      "learning_rate": 2.4056890235632846e-05,
+      "loss": 2.4574,
+      "step": 15515
+    },
+    {
+      "epoch": 1.252199176821887,
+      "grad_norm": 0.6855002045631409,
+      "learning_rate": 2.4046620419886777e-05,
+      "loss": 2.4118,
+      "step": 15516
+    },
+    {
+      "epoch": 1.2522798805584698,
+      "grad_norm": 0.6616373658180237,
+      "learning_rate": 2.4036352497097458e-05,
+      "loss": 2.4332,
+      "step": 15517
+    },
+    {
+      "epoch": 1.252360584295053,
+      "grad_norm": 0.6657225489616394,
+      "learning_rate": 2.4026086467520803e-05,
+      "loss": 2.3989,
+      "step": 15518
+    },
+    {
+      "epoch": 1.2524412880316358,
+      "grad_norm": 0.6796447038650513,
+      "learning_rate": 2.4015822331412664e-05,
+      "loss": 2.4269,
+      "step": 15519
+    },
+    {
+      "epoch": 1.2525219917682189,
+      "grad_norm": 0.7168079614639282,
+      "learning_rate": 2.400556008902889e-05,
+      "loss": 2.4263,
+      "step": 15520
+    },
+    {
+      "epoch": 1.2526026955048017,
+      "grad_norm": 0.6985058188438416,
+      "learning_rate": 2.3995299740625186e-05,
+      "loss": 2.437,
+      "step": 15521
+    },
+    {
+      "epoch": 1.2526833992413848,
+      "grad_norm": 0.7078086137771606,
+      "learning_rate": 2.3985041286457287e-05,
+      "loss": 2.3996,
+      "step": 15522
+    },
+    {
+      "epoch": 1.252764102977968,
+      "grad_norm": 0.6989054083824158,
+      "learning_rate": 2.3974784726780865e-05,
+      "loss": 2.4717,
+      "step": 15523
+    },
+    {
+      "epoch": 1.2528448067145508,
+      "grad_norm": 0.747606098651886,
+      "learning_rate": 2.396453006185153e-05,
+      "loss": 2.4228,
+      "step": 15524
+    },
+    {
+      "epoch": 1.2529255104511339,
+      "grad_norm": 0.7500887513160706,
+      "learning_rate": 2.3954277291924876e-05,
+      "loss": 2.4636,
+      "step": 15525
+    },
+    {
+      "epoch": 1.253006214187717,
+      "grad_norm": 0.7710712552070618,
+      "learning_rate": 2.3944026417256437e-05,
+      "loss": 2.4405,
+      "step": 15526
+    },
+    {
+      "epoch": 1.2530869179242998,
+      "grad_norm": 0.7278285622596741,
+      "learning_rate": 2.3933777438101657e-05,
+      "loss": 2.4279,
+      "step": 15527
+    },
+    {
+      "epoch": 1.253167621660883,
+      "grad_norm": 0.6979010701179504,
+      "learning_rate": 2.3923530354715973e-05,
+      "loss": 2.4272,
+      "step": 15528
+    },
+    {
+      "epoch": 1.253248325397466,
+      "grad_norm": 0.7330336570739746,
+      "learning_rate": 2.3913285167354804e-05,
+      "loss": 2.3861,
+      "step": 15529
+    },
+    {
+      "epoch": 1.2533290291340489,
+      "grad_norm": 0.675499677658081,
+      "learning_rate": 2.3903041876273436e-05,
+      "loss": 2.3987,
+      "step": 15530
+    },
+    {
+      "epoch": 1.253409732870632,
+      "grad_norm": 0.6854682564735413,
+      "learning_rate": 2.3892800481727186e-05,
+      "loss": 2.4085,
+      "step": 15531
+    },
+    {
+      "epoch": 1.253490436607215,
+      "grad_norm": 0.713810384273529,
+      "learning_rate": 2.388256098397129e-05,
+      "loss": 2.3897,
+      "step": 15532
+    },
+    {
+      "epoch": 1.253571140343798,
+      "grad_norm": 0.683214545249939,
+      "learning_rate": 2.3872323383260953e-05,
+      "loss": 2.4526,
+      "step": 15533
+    },
+    {
+      "epoch": 1.253651844080381,
+      "grad_norm": 0.6718357801437378,
+      "learning_rate": 2.3862087679851318e-05,
+      "loss": 2.4612,
+      "step": 15534
+    },
+    {
+      "epoch": 1.2537325478169639,
+      "grad_norm": 0.722283124923706,
+      "learning_rate": 2.3851853873997488e-05,
+      "loss": 2.4163,
+      "step": 15535
+    },
+    {
+      "epoch": 1.253813251553547,
+      "grad_norm": 0.689393162727356,
+      "learning_rate": 2.384162196595453e-05,
+      "loss": 2.3984,
+      "step": 15536
+    },
+    {
+      "epoch": 1.2538939552901298,
+      "grad_norm": 0.7146410346031189,
+      "learning_rate": 2.3831391955977412e-05,
+      "loss": 2.4442,
+      "step": 15537
+    },
+    {
+      "epoch": 1.253974659026713,
+      "grad_norm": 0.6651021838188171,
+      "learning_rate": 2.3821163844321104e-05,
+      "loss": 2.4064,
+      "step": 15538
+    },
+    {
+      "epoch": 1.254055362763296,
+      "grad_norm": 0.7088985443115234,
+      "learning_rate": 2.381093763124056e-05,
+      "loss": 2.4831,
+      "step": 15539
+    },
+    {
+      "epoch": 1.2541360664998789,
+      "grad_norm": 0.661375105381012,
+      "learning_rate": 2.3800713316990588e-05,
+      "loss": 2.3657,
+      "step": 15540
+    },
+    {
+      "epoch": 1.254216770236462,
+      "grad_norm": 0.6870979070663452,
+      "learning_rate": 2.3790490901826012e-05,
+      "loss": 2.4208,
+      "step": 15541
+    },
+    {
+      "epoch": 1.254297473973045,
+      "grad_norm": 0.6256219148635864,
+      "learning_rate": 2.3780270386001657e-05,
+      "loss": 2.4182,
+      "step": 15542
+    },
+    {
+      "epoch": 1.254378177709628,
+      "grad_norm": 0.7070638537406921,
+      "learning_rate": 2.377005176977215e-05,
+      "loss": 2.3758,
+      "step": 15543
+    },
+    {
+      "epoch": 1.254458881446211,
+      "grad_norm": 0.6571370363235474,
+      "learning_rate": 2.3759835053392242e-05,
+      "loss": 2.3927,
+      "step": 15544
+    },
+    {
+      "epoch": 1.254539585182794,
+      "grad_norm": 0.644263744354248,
+      "learning_rate": 2.3749620237116565e-05,
+      "loss": 2.3992,
+      "step": 15545
+    },
+    {
+      "epoch": 1.254620288919377,
+      "grad_norm": 0.7127394676208496,
+      "learning_rate": 2.3739407321199648e-05,
+      "loss": 2.3942,
+      "step": 15546
+    },
+    {
+      "epoch": 1.25470099265596,
+      "grad_norm": 0.7274866104125977,
+      "learning_rate": 2.372919630589605e-05,
+      "loss": 2.5232,
+      "step": 15547
+    },
+    {
+      "epoch": 1.2547816963925431,
+      "grad_norm": 0.690138041973114,
+      "learning_rate": 2.3718987191460274e-05,
+      "loss": 2.4371,
+      "step": 15548
+    },
+    {
+      "epoch": 1.254862400129126,
+      "grad_norm": 0.6990681886672974,
+      "learning_rate": 2.3708779978146724e-05,
+      "loss": 2.4568,
+      "step": 15549
+    },
+    {
+      "epoch": 1.254943103865709,
+      "grad_norm": 0.7430790662765503,
+      "learning_rate": 2.3698574666209793e-05,
+      "loss": 2.423,
+      "step": 15550
+    },
+    {
+      "epoch": 1.255023807602292,
+      "grad_norm": 0.6991416215896606,
+      "learning_rate": 2.3688371255903828e-05,
+      "loss": 2.4529,
+      "step": 15551
+    },
+    {
+      "epoch": 1.255104511338875,
+      "grad_norm": 0.6733322739601135,
+      "learning_rate": 2.367816974748317e-05,
+      "loss": 2.4531,
+      "step": 15552
+    },
+    {
+      "epoch": 1.2551852150754579,
+      "grad_norm": 0.7460463047027588,
+      "learning_rate": 2.3667970141202e-05,
+      "loss": 2.4267,
+      "step": 15553
+    },
+    {
+      "epoch": 1.255265918812041,
+      "grad_norm": 0.6784021854400635,
+      "learning_rate": 2.3657772437314517e-05,
+      "loss": 2.4996,
+      "step": 15554
+    },
+    {
+      "epoch": 1.255346622548624,
+      "grad_norm": 0.7499529719352722,
+      "learning_rate": 2.3647576636074975e-05,
+      "loss": 2.4749,
+      "step": 15555
+    },
+    {
+      "epoch": 1.255427326285207,
+      "grad_norm": 0.6698335409164429,
+      "learning_rate": 2.3637382737737368e-05,
+      "loss": 2.4499,
+      "step": 15556
+    },
+    {
+      "epoch": 1.25550803002179,
+      "grad_norm": 0.6644846200942993,
+      "learning_rate": 2.3627190742555806e-05,
+      "loss": 2.397,
+      "step": 15557
+    },
+    {
+      "epoch": 1.255588733758373,
+      "grad_norm": 0.7041488289833069,
+      "learning_rate": 2.3617000650784315e-05,
+      "loss": 2.4012,
+      "step": 15558
+    },
+    {
+      "epoch": 1.255669437494956,
+      "grad_norm": 0.72523033618927,
+      "learning_rate": 2.3606812462676798e-05,
+      "loss": 2.4151,
+      "step": 15559
+    },
+    {
+      "epoch": 1.255750141231539,
+      "grad_norm": 0.77669757604599,
+      "learning_rate": 2.3596626178487225e-05,
+      "loss": 2.4478,
+      "step": 15560
+    },
+    {
+      "epoch": 1.2558308449681221,
+      "grad_norm": 0.6919559836387634,
+      "learning_rate": 2.3586441798469462e-05,
+      "loss": 2.4548,
+      "step": 15561
+    },
+    {
+      "epoch": 1.255911548704705,
+      "grad_norm": 0.7613349556922913,
+      "learning_rate": 2.3576259322877292e-05,
+      "loss": 2.4475,
+      "step": 15562
+    },
+    {
+      "epoch": 1.255992252441288,
+      "grad_norm": 0.6738333106040955,
+      "learning_rate": 2.3566078751964515e-05,
+      "loss": 2.4242,
+      "step": 15563
+    },
+    {
+      "epoch": 1.256072956177871,
+      "grad_norm": 0.7242118716239929,
+      "learning_rate": 2.355590008598486e-05,
+      "loss": 2.4047,
+      "step": 15564
+    },
+    {
+      "epoch": 1.256153659914454,
+      "grad_norm": 0.7117685675621033,
+      "learning_rate": 2.354572332519199e-05,
+      "loss": 2.4473,
+      "step": 15565
+    },
+    {
+      "epoch": 1.256234363651037,
+      "grad_norm": 0.7466531991958618,
+      "learning_rate": 2.3535548469839564e-05,
+      "loss": 2.453,
+      "step": 15566
+    },
+    {
+      "epoch": 1.25631506738762,
+      "grad_norm": 0.6750668883323669,
+      "learning_rate": 2.3525375520181136e-05,
+      "loss": 2.4367,
+      "step": 15567
+    },
+    {
+      "epoch": 1.256395771124203,
+      "grad_norm": 0.7640851736068726,
+      "learning_rate": 2.35152044764703e-05,
+      "loss": 2.5014,
+      "step": 15568
+    },
+    {
+      "epoch": 1.256476474860786,
+      "grad_norm": 0.7198928594589233,
+      "learning_rate": 2.3505035338960456e-05,
+      "loss": 2.5138,
+      "step": 15569
+    },
+    {
+      "epoch": 1.256557178597369,
+      "grad_norm": 0.7079946398735046,
+      "learning_rate": 2.349486810790511e-05,
+      "loss": 2.4172,
+      "step": 15570
+    },
+    {
+      "epoch": 1.2566378823339521,
+      "grad_norm": 0.7477186918258667,
+      "learning_rate": 2.3484702783557655e-05,
+      "loss": 2.4224,
+      "step": 15571
+    },
+    {
+      "epoch": 1.256718586070535,
+      "grad_norm": 0.6875394582748413,
+      "learning_rate": 2.3474539366171388e-05,
+      "loss": 2.4621,
+      "step": 15572
+    },
+    {
+      "epoch": 1.256799289807118,
+      "grad_norm": 0.7164824604988098,
+      "learning_rate": 2.346437785599964e-05,
+      "loss": 2.4416,
+      "step": 15573
+    },
+    {
+      "epoch": 1.2568799935437012,
+      "grad_norm": 0.7031935453414917,
+      "learning_rate": 2.3454218253295668e-05,
+      "loss": 2.3943,
+      "step": 15574
+    },
+    {
+      "epoch": 1.256960697280284,
+      "grad_norm": 0.6739614009857178,
+      "learning_rate": 2.3444060558312665e-05,
+      "loss": 2.4114,
+      "step": 15575
+    },
+    {
+      "epoch": 1.2570414010168671,
+      "grad_norm": 0.6710866689682007,
+      "learning_rate": 2.3433904771303794e-05,
+      "loss": 2.4077,
+      "step": 15576
+    },
+    {
+      "epoch": 1.2571221047534502,
+      "grad_norm": 0.6589750051498413,
+      "learning_rate": 2.342375089252219e-05,
+      "loss": 2.3494,
+      "step": 15577
+    },
+    {
+      "epoch": 1.257202808490033,
+      "grad_norm": 0.7018333077430725,
+      "learning_rate": 2.3413598922220857e-05,
+      "loss": 2.459,
+      "step": 15578
+    },
+    {
+      "epoch": 1.2572835122266162,
+      "grad_norm": 0.7735301852226257,
+      "learning_rate": 2.3403448860652842e-05,
+      "loss": 2.4524,
+      "step": 15579
+    },
+    {
+      "epoch": 1.257364215963199,
+      "grad_norm": 0.7009726762771606,
+      "learning_rate": 2.339330070807113e-05,
+      "loss": 2.4244,
+      "step": 15580
+    },
+    {
+      "epoch": 1.2574449196997821,
+      "grad_norm": 0.671521008014679,
+      "learning_rate": 2.3383154464728595e-05,
+      "loss": 2.3808,
+      "step": 15581
+    },
+    {
+      "epoch": 1.257525623436365,
+      "grad_norm": 0.7736711502075195,
+      "learning_rate": 2.3373010130878126e-05,
+      "loss": 2.4936,
+      "step": 15582
+    },
+    {
+      "epoch": 1.257606327172948,
+      "grad_norm": 0.6987056136131287,
+      "learning_rate": 2.336286770677255e-05,
+      "loss": 2.4484,
+      "step": 15583
+    },
+    {
+      "epoch": 1.2576870309095312,
+      "grad_norm": 0.6337067484855652,
+      "learning_rate": 2.3352727192664635e-05,
+      "loss": 2.4196,
+      "step": 15584
+    },
+    {
+      "epoch": 1.257767734646114,
+      "grad_norm": 0.6832795143127441,
+      "learning_rate": 2.3342588588807123e-05,
+      "loss": 2.3681,
+      "step": 15585
+    },
+    {
+      "epoch": 1.257848438382697,
+      "grad_norm": 0.7208079695701599,
+      "learning_rate": 2.3332451895452688e-05,
+      "loss": 2.4436,
+      "step": 15586
+    },
+    {
+      "epoch": 1.2579291421192802,
+      "grad_norm": 0.6607621312141418,
+      "learning_rate": 2.3322317112853986e-05,
+      "loss": 2.4088,
+      "step": 15587
+    },
+    {
+      "epoch": 1.258009845855863,
+      "grad_norm": 0.7261247038841248,
+      "learning_rate": 2.331218424126356e-05,
+      "loss": 2.4389,
+      "step": 15588
+    },
+    {
+      "epoch": 1.2580905495924462,
+      "grad_norm": 0.6187729239463806,
+      "learning_rate": 2.3302053280933954e-05,
+      "loss": 2.3568,
+      "step": 15589
+    },
+    {
+      "epoch": 1.2581712533290292,
+      "grad_norm": 0.6196430921554565,
+      "learning_rate": 2.3291924232117713e-05,
+      "loss": 2.4285,
+      "step": 15590
+    },
+    {
+      "epoch": 1.258251957065612,
+      "grad_norm": 0.7271853685379028,
+      "learning_rate": 2.3281797095067193e-05,
+      "loss": 2.4058,
+      "step": 15591
+    },
+    {
+      "epoch": 1.2583326608021952,
+      "grad_norm": 0.7141130566596985,
+      "learning_rate": 2.327167187003484e-05,
+      "loss": 2.3971,
+      "step": 15592
+    },
+    {
+      "epoch": 1.2584133645387783,
+      "grad_norm": 0.680743932723999,
+      "learning_rate": 2.3261548557273027e-05,
+      "loss": 2.4387,
+      "step": 15593
+    },
+    {
+      "epoch": 1.2584940682753611,
+      "grad_norm": 0.718173086643219,
+      "learning_rate": 2.3251427157033955e-05,
+      "loss": 2.43,
+      "step": 15594
+    },
+    {
+      "epoch": 1.2585747720119442,
+      "grad_norm": 0.7600045800209045,
+      "learning_rate": 2.324130766956998e-05,
+      "loss": 2.4584,
+      "step": 15595
+    },
+    {
+      "epoch": 1.258655475748527,
+      "grad_norm": 0.7432500123977661,
+      "learning_rate": 2.3231190095133294e-05,
+      "loss": 2.4717,
+      "step": 15596
+    },
+    {
+      "epoch": 1.2587361794851102,
+      "grad_norm": 0.6603000164031982,
+      "learning_rate": 2.3221074433975988e-05,
+      "loss": 2.3952,
+      "step": 15597
+    },
+    {
+      "epoch": 1.258816883221693,
+      "grad_norm": 0.7020140290260315,
+      "learning_rate": 2.3210960686350213e-05,
+      "loss": 2.4064,
+      "step": 15598
+    },
+    {
+      "epoch": 1.2588975869582761,
+      "grad_norm": 0.7434887290000916,
+      "learning_rate": 2.320084885250804e-05,
+      "loss": 2.4708,
+      "step": 15599
+    },
+    {
+      "epoch": 1.2589782906948592,
+      "grad_norm": 0.6626797318458557,
+      "learning_rate": 2.3190738932701482e-05,
+      "loss": 2.4503,
+      "step": 15600
+    },
+    {
+      "epoch": 1.259058994431442,
+      "grad_norm": 0.7880598902702332,
+      "learning_rate": 2.3180630927182466e-05,
+      "loss": 2.384,
+      "step": 15601
+    },
+    {
+      "epoch": 1.2591396981680252,
+      "grad_norm": 0.7766147255897522,
+      "learning_rate": 2.3170524836202933e-05,
+      "loss": 2.4019,
+      "step": 15602
+    },
+    {
+      "epoch": 1.2592204019046083,
+      "grad_norm": 0.7817980051040649,
+      "learning_rate": 2.3160420660014792e-05,
+      "loss": 2.4729,
+      "step": 15603
+    },
+    {
+      "epoch": 1.2593011056411911,
+      "grad_norm": 0.6915614604949951,
+      "learning_rate": 2.3150318398869787e-05,
+      "loss": 2.4028,
+      "step": 15604
+    },
+    {
+      "epoch": 1.2593818093777742,
+      "grad_norm": 0.690882682800293,
+      "learning_rate": 2.3140218053019714e-05,
+      "loss": 2.4386,
+      "step": 15605
+    },
+    {
+      "epoch": 1.2594625131143573,
+      "grad_norm": 0.6670350432395935,
+      "learning_rate": 2.3130119622716382e-05,
+      "loss": 2.4224,
+      "step": 15606
+    },
+    {
+      "epoch": 1.2595432168509402,
+      "grad_norm": 0.6680006980895996,
+      "learning_rate": 2.3120023108211375e-05,
+      "loss": 2.3475,
+      "step": 15607
+    },
+    {
+      "epoch": 1.2596239205875233,
+      "grad_norm": 0.7003577947616577,
+      "learning_rate": 2.310992850975636e-05,
+      "loss": 2.4198,
+      "step": 15608
+    },
+    {
+      "epoch": 1.2597046243241061,
+      "grad_norm": 0.7444167733192444,
+      "learning_rate": 2.3099835827602944e-05,
+      "loss": 2.3756,
+      "step": 15609
+    },
+    {
+      "epoch": 1.2597853280606892,
+      "grad_norm": 0.6757989525794983,
+      "learning_rate": 2.3089745062002612e-05,
+      "loss": 2.3955,
+      "step": 15610
+    },
+    {
+      "epoch": 1.259866031797272,
+      "grad_norm": 0.6955820322036743,
+      "learning_rate": 2.3079656213206878e-05,
+      "loss": 2.4031,
+      "step": 15611
+    },
+    {
+      "epoch": 1.2599467355338552,
+      "grad_norm": 0.6646408438682556,
+      "learning_rate": 2.3069569281467184e-05,
+      "loss": 2.4246,
+      "step": 15612
+    },
+    {
+      "epoch": 1.2600274392704383,
+      "grad_norm": 0.6922882199287415,
+      "learning_rate": 2.3059484267034958e-05,
+      "loss": 2.4157,
+      "step": 15613
+    },
+    {
+      "epoch": 1.2601081430070211,
+      "grad_norm": 0.8092310428619385,
+      "learning_rate": 2.3049401170161468e-05,
+      "loss": 2.4137,
+      "step": 15614
+    },
+    {
+      "epoch": 1.2601888467436042,
+      "grad_norm": 0.7024559378623962,
+      "learning_rate": 2.3039319991098063e-05,
+      "loss": 2.4497,
+      "step": 15615
+    },
+    {
+      "epoch": 1.2602695504801873,
+      "grad_norm": 0.7096099853515625,
+      "learning_rate": 2.302924073009597e-05,
+      "loss": 2.4045,
+      "step": 15616
+    },
+    {
+      "epoch": 1.2603502542167702,
+      "grad_norm": 0.6777564287185669,
+      "learning_rate": 2.3019163387406406e-05,
+      "loss": 2.4607,
+      "step": 15617
+    },
+    {
+      "epoch": 1.2604309579533532,
+      "grad_norm": 0.7564159035682678,
+      "learning_rate": 2.300908796328052e-05,
+      "loss": 2.4985,
+      "step": 15618
+    },
+    {
+      "epoch": 1.2605116616899363,
+      "grad_norm": 0.7432986497879028,
+      "learning_rate": 2.2999014457969447e-05,
+      "loss": 2.4326,
+      "step": 15619
+    },
+    {
+      "epoch": 1.2605923654265192,
+      "grad_norm": 0.7178141474723816,
+      "learning_rate": 2.2988942871724182e-05,
+      "loss": 2.4118,
+      "step": 15620
+    },
+    {
+      "epoch": 1.2606730691631023,
+      "grad_norm": 0.7074497938156128,
+      "learning_rate": 2.2978873204795782e-05,
+      "loss": 2.4163,
+      "step": 15621
+    },
+    {
+      "epoch": 1.2607537728996854,
+      "grad_norm": 0.670200765132904,
+      "learning_rate": 2.2968805457435217e-05,
+      "loss": 2.4081,
+      "step": 15622
+    },
+    {
+      "epoch": 1.2608344766362682,
+      "grad_norm": 0.7258187532424927,
+      "learning_rate": 2.2958739629893355e-05,
+      "loss": 2.4889,
+      "step": 15623
+    },
+    {
+      "epoch": 1.2609151803728513,
+      "grad_norm": 0.6999781727790833,
+      "learning_rate": 2.2948675722421086e-05,
+      "loss": 2.3945,
+      "step": 15624
+    },
+    {
+      "epoch": 1.2609958841094342,
+      "grad_norm": 0.7030084133148193,
+      "learning_rate": 2.2938613735269243e-05,
+      "loss": 2.4509,
+      "step": 15625
+    },
+    {
+      "epoch": 1.2610765878460173,
+      "grad_norm": 0.6875420212745667,
+      "learning_rate": 2.292855366868858e-05,
+      "loss": 2.3658,
+      "step": 15626
+    },
+    {
+      "epoch": 1.2611572915826001,
+      "grad_norm": 0.7375235557556152,
+      "learning_rate": 2.2918495522929817e-05,
+      "loss": 2.4308,
+      "step": 15627
+    },
+    {
+      "epoch": 1.2612379953191832,
+      "grad_norm": 0.7021106481552124,
+      "learning_rate": 2.2908439298243644e-05,
+      "loss": 2.4046,
+      "step": 15628
+    },
+    {
+      "epoch": 1.2613186990557663,
+      "grad_norm": 0.76661616563797,
+      "learning_rate": 2.2898384994880716e-05,
+      "loss": 2.5156,
+      "step": 15629
+    },
+    {
+      "epoch": 1.2613994027923492,
+      "grad_norm": 0.6684869527816772,
+      "learning_rate": 2.2888332613091558e-05,
+      "loss": 2.4342,
+      "step": 15630
+    },
+    {
+      "epoch": 1.2614801065289323,
+      "grad_norm": 0.6878669261932373,
+      "learning_rate": 2.2878282153126706e-05,
+      "loss": 2.4544,
+      "step": 15631
+    },
+    {
+      "epoch": 1.2615608102655154,
+      "grad_norm": 0.6659132838249207,
+      "learning_rate": 2.2868233615236702e-05,
+      "loss": 2.4341,
+      "step": 15632
+    },
+    {
+      "epoch": 1.2616415140020982,
+      "grad_norm": 0.657474160194397,
+      "learning_rate": 2.2858186999671905e-05,
+      "loss": 2.3515,
+      "step": 15633
+    },
+    {
+      "epoch": 1.2617222177386813,
+      "grad_norm": 0.7245650291442871,
+      "learning_rate": 2.284814230668274e-05,
+      "loss": 2.3983,
+      "step": 15634
+    },
+    {
+      "epoch": 1.2618029214752644,
+      "grad_norm": 0.6400195360183716,
+      "learning_rate": 2.2838099536519554e-05,
+      "loss": 2.3535,
+      "step": 15635
+    },
+    {
+      "epoch": 1.2618836252118473,
+      "grad_norm": 0.6719450950622559,
+      "learning_rate": 2.282805868943262e-05,
+      "loss": 2.3906,
+      "step": 15636
+    },
+    {
+      "epoch": 1.2619643289484304,
+      "grad_norm": 0.682746946811676,
+      "learning_rate": 2.2818019765672207e-05,
+      "loss": 2.4045,
+      "step": 15637
+    },
+    {
+      "epoch": 1.2620450326850134,
+      "grad_norm": 0.6631760597229004,
+      "learning_rate": 2.2807982765488513e-05,
+      "loss": 2.4896,
+      "step": 15638
+    },
+    {
+      "epoch": 1.2621257364215963,
+      "grad_norm": 0.782202422618866,
+      "learning_rate": 2.279794768913164e-05,
+      "loss": 2.4628,
+      "step": 15639
+    },
+    {
+      "epoch": 1.2622064401581794,
+      "grad_norm": 0.7579823732376099,
+      "learning_rate": 2.278791453685173e-05,
+      "loss": 2.4635,
+      "step": 15640
+    },
+    {
+      "epoch": 1.2622871438947623,
+      "grad_norm": 0.665096640586853,
+      "learning_rate": 2.277788330889884e-05,
+      "loss": 2.4899,
+      "step": 15641
+    },
+    {
+      "epoch": 1.2623678476313454,
+      "grad_norm": 0.7635685205459595,
+      "learning_rate": 2.2767854005522936e-05,
+      "loss": 2.4146,
+      "step": 15642
+    },
+    {
+      "epoch": 1.2624485513679282,
+      "grad_norm": 0.7579118609428406,
+      "learning_rate": 2.2757826626974e-05,
+      "loss": 2.3692,
+      "step": 15643
+    },
+    {
+      "epoch": 1.2625292551045113,
+      "grad_norm": 0.6772074699401855,
+      "learning_rate": 2.2747801173501938e-05,
+      "loss": 2.3954,
+      "step": 15644
+    },
+    {
+      "epoch": 1.2626099588410944,
+      "grad_norm": 0.7028382420539856,
+      "learning_rate": 2.2737777645356606e-05,
+      "loss": 2.4799,
+      "step": 15645
+    },
+    {
+      "epoch": 1.2626906625776773,
+      "grad_norm": 0.7152617573738098,
+      "learning_rate": 2.2727756042787818e-05,
+      "loss": 2.4095,
+      "step": 15646
+    },
+    {
+      "epoch": 1.2627713663142603,
+      "grad_norm": 0.7286608219146729,
+      "learning_rate": 2.271773636604535e-05,
+      "loss": 2.4496,
+      "step": 15647
+    },
+    {
+      "epoch": 1.2628520700508434,
+      "grad_norm": 0.7006896734237671,
+      "learning_rate": 2.2707718615378935e-05,
+      "loss": 2.4128,
+      "step": 15648
+    },
+    {
+      "epoch": 1.2629327737874263,
+      "grad_norm": 0.6856697797775269,
+      "learning_rate": 2.2697702791038177e-05,
+      "loss": 2.4169,
+      "step": 15649
+    },
+    {
+      "epoch": 1.2630134775240094,
+      "grad_norm": 0.7582918405532837,
+      "learning_rate": 2.268768889327275e-05,
+      "loss": 2.4007,
+      "step": 15650
+    },
+    {
+      "epoch": 1.2630941812605925,
+      "grad_norm": 0.664633572101593,
+      "learning_rate": 2.2677676922332237e-05,
+      "loss": 2.3876,
+      "step": 15651
+    },
+    {
+      "epoch": 1.2631748849971753,
+      "grad_norm": 0.7283070087432861,
+      "learning_rate": 2.266766687846611e-05,
+      "loss": 2.4175,
+      "step": 15652
+    },
+    {
+      "epoch": 1.2632555887337584,
+      "grad_norm": 0.7309537529945374,
+      "learning_rate": 2.2657658761923863e-05,
+      "loss": 2.3998,
+      "step": 15653
+    },
+    {
+      "epoch": 1.2633362924703415,
+      "grad_norm": 0.6386510133743286,
+      "learning_rate": 2.2647652572954968e-05,
+      "loss": 2.3723,
+      "step": 15654
+    },
+    {
+      "epoch": 1.2634169962069244,
+      "grad_norm": 0.6805689930915833,
+      "learning_rate": 2.263764831180876e-05,
+      "loss": 2.3989,
+      "step": 15655
+    },
+    {
+      "epoch": 1.2634976999435072,
+      "grad_norm": 0.7147208452224731,
+      "learning_rate": 2.2627645978734536e-05,
+      "loss": 2.4748,
+      "step": 15656
+    },
+    {
+      "epoch": 1.2635784036800903,
+      "grad_norm": 0.6835155487060547,
+      "learning_rate": 2.2617645573981683e-05,
+      "loss": 2.4266,
+      "step": 15657
+    },
+    {
+      "epoch": 1.2636591074166734,
+      "grad_norm": 0.7631552219390869,
+      "learning_rate": 2.2607647097799368e-05,
+      "loss": 2.4152,
+      "step": 15658
+    },
+    {
+      "epoch": 1.2637398111532563,
+      "grad_norm": 0.6793624758720398,
+      "learning_rate": 2.2597650550436777e-05,
+      "loss": 2.3491,
+      "step": 15659
+    },
+    {
+      "epoch": 1.2638205148898394,
+      "grad_norm": 0.6465637683868408,
+      "learning_rate": 2.2587655932143083e-05,
+      "loss": 2.3774,
+      "step": 15660
+    },
+    {
+      "epoch": 1.2639012186264225,
+      "grad_norm": 0.6920284628868103,
+      "learning_rate": 2.2577663243167368e-05,
+      "loss": 2.4321,
+      "step": 15661
+    },
+    {
+      "epoch": 1.2639819223630053,
+      "grad_norm": 0.6922522783279419,
+      "learning_rate": 2.256767248375866e-05,
+      "loss": 2.4242,
+      "step": 15662
+    },
+    {
+      "epoch": 1.2640626260995884,
+      "grad_norm": 0.6811214089393616,
+      "learning_rate": 2.255768365416595e-05,
+      "loss": 2.4101,
+      "step": 15663
+    },
+    {
+      "epoch": 1.2641433298361715,
+      "grad_norm": 0.6704947352409363,
+      "learning_rate": 2.2547696754638238e-05,
+      "loss": 2.4792,
+      "step": 15664
+    },
+    {
+      "epoch": 1.2642240335727544,
+      "grad_norm": 0.6814701557159424,
+      "learning_rate": 2.2537711785424354e-05,
+      "loss": 2.4429,
+      "step": 15665
+    },
+    {
+      "epoch": 1.2643047373093375,
+      "grad_norm": 0.6778244972229004,
+      "learning_rate": 2.252772874677318e-05,
+      "loss": 2.3882,
+      "step": 15666
+    },
+    {
+      "epoch": 1.2643854410459205,
+      "grad_norm": 0.6570093035697937,
+      "learning_rate": 2.2517747638933518e-05,
+      "loss": 2.4162,
+      "step": 15667
+    },
+    {
+      "epoch": 1.2644661447825034,
+      "grad_norm": 0.6973466873168945,
+      "learning_rate": 2.2507768462154133e-05,
+      "loss": 2.3646,
+      "step": 15668
+    },
+    {
+      "epoch": 1.2645468485190865,
+      "grad_norm": 0.7258623242378235,
+      "learning_rate": 2.2497791216683715e-05,
+      "loss": 2.404,
+      "step": 15669
+    },
+    {
+      "epoch": 1.2646275522556694,
+      "grad_norm": 0.7462170124053955,
+      "learning_rate": 2.248781590277097e-05,
+      "loss": 2.5076,
+      "step": 15670
+    },
+    {
+      "epoch": 1.2647082559922525,
+      "grad_norm": 0.7070441246032715,
+      "learning_rate": 2.247784252066444e-05,
+      "loss": 2.3817,
+      "step": 15671
+    },
+    {
+      "epoch": 1.2647889597288353,
+      "grad_norm": 0.7150183916091919,
+      "learning_rate": 2.246787107061272e-05,
+      "loss": 2.461,
+      "step": 15672
+    },
+    {
+      "epoch": 1.2648696634654184,
+      "grad_norm": 0.668436586856842,
+      "learning_rate": 2.2457901552864347e-05,
+      "loss": 2.466,
+      "step": 15673
+    },
+    {
+      "epoch": 1.2649503672020015,
+      "grad_norm": 0.7011097073554993,
+      "learning_rate": 2.2447933967667745e-05,
+      "loss": 2.4582,
+      "step": 15674
+    },
+    {
+      "epoch": 1.2650310709385844,
+      "grad_norm": 0.7149096727371216,
+      "learning_rate": 2.243796831527134e-05,
+      "loss": 2.4461,
+      "step": 15675
+    },
+    {
+      "epoch": 1.2651117746751674,
+      "grad_norm": 0.6810914278030396,
+      "learning_rate": 2.2428004595923525e-05,
+      "loss": 2.4043,
+      "step": 15676
+    },
+    {
+      "epoch": 1.2651924784117505,
+      "grad_norm": 0.7700765132904053,
+      "learning_rate": 2.241804280987261e-05,
+      "loss": 2.4197,
+      "step": 15677
+    },
+    {
+      "epoch": 1.2652731821483334,
+      "grad_norm": 0.6897448897361755,
+      "learning_rate": 2.240808295736686e-05,
+      "loss": 2.4052,
+      "step": 15678
+    },
+    {
+      "epoch": 1.2653538858849165,
+      "grad_norm": 0.7092932462692261,
+      "learning_rate": 2.2398125038654515e-05,
+      "loss": 2.4088,
+      "step": 15679
+    },
+    {
+      "epoch": 1.2654345896214996,
+      "grad_norm": 0.6930294632911682,
+      "learning_rate": 2.2388169053983777e-05,
+      "loss": 2.4504,
+      "step": 15680
+    },
+    {
+      "epoch": 1.2655152933580824,
+      "grad_norm": 0.7056782245635986,
+      "learning_rate": 2.237821500360271e-05,
+      "loss": 2.3975,
+      "step": 15681
+    },
+    {
+      "epoch": 1.2655959970946655,
+      "grad_norm": 0.651772141456604,
+      "learning_rate": 2.236826288775944e-05,
+      "loss": 2.3941,
+      "step": 15682
+    },
+    {
+      "epoch": 1.2656767008312486,
+      "grad_norm": 0.7254980206489563,
+      "learning_rate": 2.2358312706702012e-05,
+      "loss": 2.4149,
+      "step": 15683
+    },
+    {
+      "epoch": 1.2657574045678315,
+      "grad_norm": 0.6553635597229004,
+      "learning_rate": 2.2348364460678373e-05,
+      "loss": 2.4099,
+      "step": 15684
+    },
+    {
+      "epoch": 1.2658381083044146,
+      "grad_norm": 0.6952616572380066,
+      "learning_rate": 2.233841814993646e-05,
+      "loss": 2.384,
+      "step": 15685
+    },
+    {
+      "epoch": 1.2659188120409974,
+      "grad_norm": 0.72947096824646,
+      "learning_rate": 2.2328473774724178e-05,
+      "loss": 2.5033,
+      "step": 15686
+    },
+    {
+      "epoch": 1.2659995157775805,
+      "grad_norm": 0.7419683933258057,
+      "learning_rate": 2.231853133528937e-05,
+      "loss": 2.4881,
+      "step": 15687
+    },
+    {
+      "epoch": 1.2660802195141634,
+      "grad_norm": 0.7125211358070374,
+      "learning_rate": 2.2308590831879827e-05,
+      "loss": 2.4334,
+      "step": 15688
+    },
+    {
+      "epoch": 1.2661609232507465,
+      "grad_norm": 0.6668617129325867,
+      "learning_rate": 2.2298652264743315e-05,
+      "loss": 2.4144,
+      "step": 15689
+    },
+    {
+      "epoch": 1.2662416269873296,
+      "grad_norm": 0.8075512051582336,
+      "learning_rate": 2.2288715634127465e-05,
+      "loss": 2.421,
+      "step": 15690
+    },
+    {
+      "epoch": 1.2663223307239124,
+      "grad_norm": 0.6894629001617432,
+      "learning_rate": 2.2278780940279965e-05,
+      "loss": 2.4142,
+      "step": 15691
+    },
+    {
+      "epoch": 1.2664030344604955,
+      "grad_norm": 0.7418074011802673,
+      "learning_rate": 2.226884818344841e-05,
+      "loss": 2.4214,
+      "step": 15692
+    },
+    {
+      "epoch": 1.2664837381970786,
+      "grad_norm": 0.6724219918251038,
+      "learning_rate": 2.225891736388037e-05,
+      "loss": 2.4455,
+      "step": 15693
+    },
+    {
+      "epoch": 1.2665644419336615,
+      "grad_norm": 0.7202882766723633,
+      "learning_rate": 2.224898848182331e-05,
+      "loss": 2.4017,
+      "step": 15694
+    },
+    {
+      "epoch": 1.2666451456702446,
+      "grad_norm": 0.7671259641647339,
+      "learning_rate": 2.2239061537524698e-05,
+      "loss": 2.4386,
+      "step": 15695
+    },
+    {
+      "epoch": 1.2667258494068276,
+      "grad_norm": 0.7154317498207092,
+      "learning_rate": 2.222913653123194e-05,
+      "loss": 2.3754,
+      "step": 15696
+    },
+    {
+      "epoch": 1.2668065531434105,
+      "grad_norm": 0.7203264236450195,
+      "learning_rate": 2.221921346319239e-05,
+      "loss": 2.3926,
+      "step": 15697
+    },
+    {
+      "epoch": 1.2668872568799936,
+      "grad_norm": 0.7104187607765198,
+      "learning_rate": 2.2209292333653365e-05,
+      "loss": 2.4528,
+      "step": 15698
+    },
+    {
+      "epoch": 1.2669679606165767,
+      "grad_norm": 0.7650138139724731,
+      "learning_rate": 2.2199373142862158e-05,
+      "loss": 2.4372,
+      "step": 15699
+    },
+    {
+      "epoch": 1.2670486643531595,
+      "grad_norm": 0.6796044111251831,
+      "learning_rate": 2.2189455891065903e-05,
+      "loss": 2.415,
+      "step": 15700
+    },
+    {
+      "epoch": 1.2671293680897426,
+      "grad_norm": 0.6749297380447388,
+      "learning_rate": 2.2179540578511813e-05,
+      "loss": 2.4337,
+      "step": 15701
+    },
+    {
+      "epoch": 1.2672100718263255,
+      "grad_norm": 0.7330272793769836,
+      "learning_rate": 2.216962720544703e-05,
+      "loss": 2.4322,
+      "step": 15702
+    },
+    {
+      "epoch": 1.2672907755629086,
+      "grad_norm": 0.6793510913848877,
+      "learning_rate": 2.215971577211855e-05,
+      "loss": 2.4473,
+      "step": 15703
+    },
+    {
+      "epoch": 1.2673714792994915,
+      "grad_norm": 0.7477267384529114,
+      "learning_rate": 2.2149806278773433e-05,
+      "loss": 2.4699,
+      "step": 15704
+    },
+    {
+      "epoch": 1.2674521830360745,
+      "grad_norm": 0.7048643827438354,
+      "learning_rate": 2.213989872565867e-05,
+      "loss": 2.4341,
+      "step": 15705
+    },
+    {
+      "epoch": 1.2675328867726576,
+      "grad_norm": 0.647433340549469,
+      "learning_rate": 2.2129993113021108e-05,
+      "loss": 2.423,
+      "step": 15706
+    },
+    {
+      "epoch": 1.2676135905092405,
+      "grad_norm": 0.6886507272720337,
+      "learning_rate": 2.2120089441107706e-05,
+      "loss": 2.4185,
+      "step": 15707
+    },
+    {
+      "epoch": 1.2676942942458236,
+      "grad_norm": 0.6720516085624695,
+      "learning_rate": 2.2110187710165242e-05,
+      "loss": 2.4587,
+      "step": 15708
+    },
+    {
+      "epoch": 1.2677749979824067,
+      "grad_norm": 0.676665723323822,
+      "learning_rate": 2.2100287920440543e-05,
+      "loss": 2.4241,
+      "step": 15709
+    },
+    {
+      "epoch": 1.2678557017189895,
+      "grad_norm": 0.6939559578895569,
+      "learning_rate": 2.209039007218028e-05,
+      "loss": 2.3974,
+      "step": 15710
+    },
+    {
+      "epoch": 1.2679364054555726,
+      "grad_norm": 0.6485786437988281,
+      "learning_rate": 2.2080494165631137e-05,
+      "loss": 2.4041,
+      "step": 15711
+    },
+    {
+      "epoch": 1.2680171091921557,
+      "grad_norm": 0.668319582939148,
+      "learning_rate": 2.2070600201039802e-05,
+      "loss": 2.4705,
+      "step": 15712
+    },
+    {
+      "epoch": 1.2680978129287386,
+      "grad_norm": 0.6837478280067444,
+      "learning_rate": 2.206070817865279e-05,
+      "loss": 2.4474,
+      "step": 15713
+    },
+    {
+      "epoch": 1.2681785166653217,
+      "grad_norm": 0.7000131011009216,
+      "learning_rate": 2.2050818098716664e-05,
+      "loss": 2.4463,
+      "step": 15714
+    },
+    {
+      "epoch": 1.2682592204019045,
+      "grad_norm": 0.7063068151473999,
+      "learning_rate": 2.204092996147794e-05,
+      "loss": 2.4226,
+      "step": 15715
+    },
+    {
+      "epoch": 1.2683399241384876,
+      "grad_norm": 0.6497172117233276,
+      "learning_rate": 2.2031043767183003e-05,
+      "loss": 2.3678,
+      "step": 15716
+    },
+    {
+      "epoch": 1.2684206278750705,
+      "grad_norm": 0.6558645963668823,
+      "learning_rate": 2.2021159516078262e-05,
+      "loss": 2.4021,
+      "step": 15717
+    },
+    {
+      "epoch": 1.2685013316116536,
+      "grad_norm": 0.7411713600158691,
+      "learning_rate": 2.2011277208410062e-05,
+      "loss": 2.4346,
+      "step": 15718
+    },
+    {
+      "epoch": 1.2685820353482367,
+      "grad_norm": 0.7275578379631042,
+      "learning_rate": 2.2001396844424714e-05,
+      "loss": 2.4262,
+      "step": 15719
+    },
+    {
+      "epoch": 1.2686627390848195,
+      "grad_norm": 0.7010936141014099,
+      "learning_rate": 2.199151842436844e-05,
+      "loss": 2.4774,
+      "step": 15720
+    },
+    {
+      "epoch": 1.2687434428214026,
+      "grad_norm": 0.7551137208938599,
+      "learning_rate": 2.1981641948487462e-05,
+      "loss": 2.5286,
+      "step": 15721
+    },
+    {
+      "epoch": 1.2688241465579857,
+      "grad_norm": 0.6510799527168274,
+      "learning_rate": 2.1971767417027888e-05,
+      "loss": 2.3813,
+      "step": 15722
+    },
+    {
+      "epoch": 1.2689048502945686,
+      "grad_norm": 0.636050283908844,
+      "learning_rate": 2.196189483023584e-05,
+      "loss": 2.4226,
+      "step": 15723
+    },
+    {
+      "epoch": 1.2689855540311517,
+      "grad_norm": 0.6939265131950378,
+      "learning_rate": 2.1952024188357368e-05,
+      "loss": 2.4516,
+      "step": 15724
+    },
+    {
+      "epoch": 1.2690662577677347,
+      "grad_norm": 0.6715239882469177,
+      "learning_rate": 2.1942155491638494e-05,
+      "loss": 2.4358,
+      "step": 15725
+    },
+    {
+      "epoch": 1.2691469615043176,
+      "grad_norm": 0.740680456161499,
+      "learning_rate": 2.1932288740325123e-05,
+      "loss": 2.4135,
+      "step": 15726
+    },
+    {
+      "epoch": 1.2692276652409007,
+      "grad_norm": 0.6969335079193115,
+      "learning_rate": 2.1922423934663193e-05,
+      "loss": 2.43,
+      "step": 15727
+    },
+    {
+      "epoch": 1.2693083689774838,
+      "grad_norm": 0.6390758156776428,
+      "learning_rate": 2.1912561074898554e-05,
+      "loss": 2.4492,
+      "step": 15728
+    },
+    {
+      "epoch": 1.2693890727140666,
+      "grad_norm": 0.7129701375961304,
+      "learning_rate": 2.190270016127701e-05,
+      "loss": 2.3799,
+      "step": 15729
+    },
+    {
+      "epoch": 1.2694697764506497,
+      "grad_norm": 0.7309553027153015,
+      "learning_rate": 2.1892841194044332e-05,
+      "loss": 2.4955,
+      "step": 15730
+    },
+    {
+      "epoch": 1.2695504801872326,
+      "grad_norm": 0.7257225513458252,
+      "learning_rate": 2.1882984173446252e-05,
+      "loss": 2.4184,
+      "step": 15731
+    },
+    {
+      "epoch": 1.2696311839238157,
+      "grad_norm": 0.7434510588645935,
+      "learning_rate": 2.1873129099728384e-05,
+      "loss": 2.453,
+      "step": 15732
+    },
+    {
+      "epoch": 1.2697118876603986,
+      "grad_norm": 0.6643160581588745,
+      "learning_rate": 2.1863275973136356e-05,
+      "loss": 2.3619,
+      "step": 15733
+    },
+    {
+      "epoch": 1.2697925913969816,
+      "grad_norm": 0.6677344441413879,
+      "learning_rate": 2.1853424793915778e-05,
+      "loss": 2.406,
+      "step": 15734
+    },
+    {
+      "epoch": 1.2698732951335647,
+      "grad_norm": 0.760028064250946,
+      "learning_rate": 2.1843575562312092e-05,
+      "loss": 2.5479,
+      "step": 15735
+    },
+    {
+      "epoch": 1.2699539988701476,
+      "grad_norm": 0.6668389439582825,
+      "learning_rate": 2.183372827857082e-05,
+      "loss": 2.4104,
+      "step": 15736
+    },
+    {
+      "epoch": 1.2700347026067307,
+      "grad_norm": 0.651155412197113,
+      "learning_rate": 2.182388294293736e-05,
+      "loss": 2.3738,
+      "step": 15737
+    },
+    {
+      "epoch": 1.2701154063433138,
+      "grad_norm": 0.736907958984375,
+      "learning_rate": 2.1814039555657084e-05,
+      "loss": 2.4179,
+      "step": 15738
+    },
+    {
+      "epoch": 1.2701961100798966,
+      "grad_norm": 0.7068225741386414,
+      "learning_rate": 2.180419811697534e-05,
+      "loss": 2.3911,
+      "step": 15739
+    },
+    {
+      "epoch": 1.2702768138164797,
+      "grad_norm": 0.6959261894226074,
+      "learning_rate": 2.1794358627137368e-05,
+      "loss": 2.452,
+      "step": 15740
+    },
+    {
+      "epoch": 1.2703575175530628,
+      "grad_norm": 0.6886181235313416,
+      "learning_rate": 2.1784521086388442e-05,
+      "loss": 2.4166,
+      "step": 15741
+    },
+    {
+      "epoch": 1.2704382212896457,
+      "grad_norm": 0.6494541168212891,
+      "learning_rate": 2.177468549497369e-05,
+      "loss": 2.3589,
+      "step": 15742
+    },
+    {
+      "epoch": 1.2705189250262288,
+      "grad_norm": 0.7008326649665833,
+      "learning_rate": 2.1764851853138247e-05,
+      "loss": 2.3697,
+      "step": 15743
+    },
+    {
+      "epoch": 1.2705996287628119,
+      "grad_norm": 0.6800456643104553,
+      "learning_rate": 2.1755020161127238e-05,
+      "loss": 2.4162,
+      "step": 15744
+    },
+    {
+      "epoch": 1.2706803324993947,
+      "grad_norm": 0.6836018562316895,
+      "learning_rate": 2.1745190419185634e-05,
+      "loss": 2.3977,
+      "step": 15745
+    },
+    {
+      "epoch": 1.2707610362359778,
+      "grad_norm": 0.6489691138267517,
+      "learning_rate": 2.173536262755844e-05,
+      "loss": 2.464,
+      "step": 15746
+    },
+    {
+      "epoch": 1.2708417399725607,
+      "grad_norm": 0.7309786677360535,
+      "learning_rate": 2.172553678649061e-05,
+      "loss": 2.4065,
+      "step": 15747
+    },
+    {
+      "epoch": 1.2709224437091438,
+      "grad_norm": 0.6752686500549316,
+      "learning_rate": 2.1715712896227004e-05,
+      "loss": 2.3935,
+      "step": 15748
+    },
+    {
+      "epoch": 1.2710031474457266,
+      "grad_norm": 0.7039850354194641,
+      "learning_rate": 2.1705890957012465e-05,
+      "loss": 2.4605,
+      "step": 15749
+    },
+    {
+      "epoch": 1.2710838511823097,
+      "grad_norm": 0.6904652714729309,
+      "learning_rate": 2.169607096909182e-05,
+      "loss": 2.4264,
+      "step": 15750
+    },
+    {
+      "epoch": 1.2711645549188928,
+      "grad_norm": 0.7104331254959106,
+      "learning_rate": 2.168625293270974e-05,
+      "loss": 2.378,
+      "step": 15751
+    },
+    {
+      "epoch": 1.2712452586554757,
+      "grad_norm": 0.6732800602912903,
+      "learning_rate": 2.167643684811096e-05,
+      "loss": 2.4216,
+      "step": 15752
+    },
+    {
+      "epoch": 1.2713259623920588,
+      "grad_norm": 0.7207335829734802,
+      "learning_rate": 2.166662271554011e-05,
+      "loss": 2.3861,
+      "step": 15753
+    },
+    {
+      "epoch": 1.2714066661286418,
+      "grad_norm": 0.7561055421829224,
+      "learning_rate": 2.1656810535241813e-05,
+      "loss": 2.4753,
+      "step": 15754
+    },
+    {
+      "epoch": 1.2714873698652247,
+      "grad_norm": 0.7018210887908936,
+      "learning_rate": 2.1647000307460564e-05,
+      "loss": 2.401,
+      "step": 15755
+    },
+    {
+      "epoch": 1.2715680736018078,
+      "grad_norm": 0.6908013224601746,
+      "learning_rate": 2.163719203244089e-05,
+      "loss": 2.4451,
+      "step": 15756
+    },
+    {
+      "epoch": 1.2716487773383909,
+      "grad_norm": 0.734909176826477,
+      "learning_rate": 2.162738571042723e-05,
+      "loss": 2.4221,
+      "step": 15757
+    },
+    {
+      "epoch": 1.2717294810749737,
+      "grad_norm": 0.7047279477119446,
+      "learning_rate": 2.1617581341663973e-05,
+      "loss": 2.4149,
+      "step": 15758
+    },
+    {
+      "epoch": 1.2718101848115568,
+      "grad_norm": 0.6875640749931335,
+      "learning_rate": 2.1607778926395496e-05,
+      "loss": 2.3874,
+      "step": 15759
+    },
+    {
+      "epoch": 1.2718908885481397,
+      "grad_norm": 0.7300851345062256,
+      "learning_rate": 2.159797846486611e-05,
+      "loss": 2.4706,
+      "step": 15760
+    },
+    {
+      "epoch": 1.2719715922847228,
+      "grad_norm": 0.733775794506073,
+      "learning_rate": 2.1588179957320022e-05,
+      "loss": 2.4208,
+      "step": 15761
+    },
+    {
+      "epoch": 1.2720522960213057,
+      "grad_norm": 0.8375213742256165,
+      "learning_rate": 2.1578383404001458e-05,
+      "loss": 2.4672,
+      "step": 15762
+    },
+    {
+      "epoch": 1.2721329997578887,
+      "grad_norm": 0.7276780009269714,
+      "learning_rate": 2.15685888051546e-05,
+      "loss": 2.4536,
+      "step": 15763
+    },
+    {
+      "epoch": 1.2722137034944718,
+      "grad_norm": 0.7765224575996399,
+      "learning_rate": 2.1558796161023508e-05,
+      "loss": 2.3671,
+      "step": 15764
+    },
+    {
+      "epoch": 1.2722944072310547,
+      "grad_norm": 0.7225642204284668,
+      "learning_rate": 2.1549005471852256e-05,
+      "loss": 2.4316,
+      "step": 15765
+    },
+    {
+      "epoch": 1.2723751109676378,
+      "grad_norm": 0.6959484219551086,
+      "learning_rate": 2.1539216737884904e-05,
+      "loss": 2.4581,
+      "step": 15766
+    },
+    {
+      "epoch": 1.2724558147042209,
+      "grad_norm": 0.6943621039390564,
+      "learning_rate": 2.1529429959365332e-05,
+      "loss": 2.4372,
+      "step": 15767
+    },
+    {
+      "epoch": 1.2725365184408037,
+      "grad_norm": 0.7067148089408875,
+      "learning_rate": 2.151964513653746e-05,
+      "loss": 2.431,
+      "step": 15768
+    },
+    {
+      "epoch": 1.2726172221773868,
+      "grad_norm": 0.8317076563835144,
+      "learning_rate": 2.150986226964521e-05,
+      "loss": 2.4177,
+      "step": 15769
+    },
+    {
+      "epoch": 1.27269792591397,
+      "grad_norm": 0.7390087246894836,
+      "learning_rate": 2.150008135893239e-05,
+      "loss": 2.4711,
+      "step": 15770
+    },
+    {
+      "epoch": 1.2727786296505528,
+      "grad_norm": 0.6829150915145874,
+      "learning_rate": 2.1490302404642725e-05,
+      "loss": 2.4477,
+      "step": 15771
+    },
+    {
+      "epoch": 1.2728593333871359,
+      "grad_norm": 0.7355613708496094,
+      "learning_rate": 2.148052540701995e-05,
+      "loss": 2.493,
+      "step": 15772
+    },
+    {
+      "epoch": 1.272940037123719,
+      "grad_norm": 0.6872289776802063,
+      "learning_rate": 2.1470750366307747e-05,
+      "loss": 2.4363,
+      "step": 15773
+    },
+    {
+      "epoch": 1.2730207408603018,
+      "grad_norm": 0.7753220796585083,
+      "learning_rate": 2.1460977282749705e-05,
+      "loss": 2.4376,
+      "step": 15774
+    },
+    {
+      "epoch": 1.273101444596885,
+      "grad_norm": 0.6717056632041931,
+      "learning_rate": 2.145120615658942e-05,
+      "loss": 2.4383,
+      "step": 15775
+    },
+    {
+      "epoch": 1.2731821483334678,
+      "grad_norm": 0.7441569566726685,
+      "learning_rate": 2.1441436988070428e-05,
+      "loss": 2.462,
+      "step": 15776
+    },
+    {
+      "epoch": 1.2732628520700509,
+      "grad_norm": 0.6824371814727783,
+      "learning_rate": 2.143166977743615e-05,
+      "loss": 2.4173,
+      "step": 15777
+    },
+    {
+      "epoch": 1.2733435558066337,
+      "grad_norm": 0.7310225963592529,
+      "learning_rate": 2.1421904524930038e-05,
+      "loss": 2.4222,
+      "step": 15778
+    },
+    {
+      "epoch": 1.2734242595432168,
+      "grad_norm": 0.7198066115379333,
+      "learning_rate": 2.141214123079548e-05,
+      "loss": 2.4262,
+      "step": 15779
+    },
+    {
+      "epoch": 1.2735049632798,
+      "grad_norm": 0.7081776857376099,
+      "learning_rate": 2.1402379895275783e-05,
+      "loss": 2.4473,
+      "step": 15780
+    },
+    {
+      "epoch": 1.2735856670163828,
+      "grad_norm": 0.6909368634223938,
+      "learning_rate": 2.1392620518614235e-05,
+      "loss": 2.4528,
+      "step": 15781
+    },
+    {
+      "epoch": 1.2736663707529658,
+      "grad_norm": 0.7170675992965698,
+      "learning_rate": 2.1382863101054107e-05,
+      "loss": 2.4214,
+      "step": 15782
+    },
+    {
+      "epoch": 1.273747074489549,
+      "grad_norm": 0.6992846727371216,
+      "learning_rate": 2.1373107642838497e-05,
+      "loss": 2.4397,
+      "step": 15783
+    },
+    {
+      "epoch": 1.2738277782261318,
+      "grad_norm": 0.7245237231254578,
+      "learning_rate": 2.1363354144210578e-05,
+      "loss": 2.373,
+      "step": 15784
+    },
+    {
+      "epoch": 1.273908481962715,
+      "grad_norm": 0.6929232478141785,
+      "learning_rate": 2.1353602605413435e-05,
+      "loss": 2.4297,
+      "step": 15785
+    },
+    {
+      "epoch": 1.273989185699298,
+      "grad_norm": 0.7243950366973877,
+      "learning_rate": 2.134385302669013e-05,
+      "loss": 2.3856,
+      "step": 15786
+    },
+    {
+      "epoch": 1.2740698894358808,
+      "grad_norm": 0.6712679266929626,
+      "learning_rate": 2.133410540828359e-05,
+      "loss": 2.3818,
+      "step": 15787
+    },
+    {
+      "epoch": 1.274150593172464,
+      "grad_norm": 0.7433474063873291,
+      "learning_rate": 2.1324359750436774e-05,
+      "loss": 2.4148,
+      "step": 15788
+    },
+    {
+      "epoch": 1.274231296909047,
+      "grad_norm": 0.7225894927978516,
+      "learning_rate": 2.1314616053392577e-05,
+      "loss": 2.395,
+      "step": 15789
+    },
+    {
+      "epoch": 1.2743120006456299,
+      "grad_norm": 0.7026889324188232,
+      "learning_rate": 2.130487431739383e-05,
+      "loss": 2.4693,
+      "step": 15790
+    },
+    {
+      "epoch": 1.274392704382213,
+      "grad_norm": 0.6898565292358398,
+      "learning_rate": 2.1295134542683325e-05,
+      "loss": 2.3643,
+      "step": 15791
+    },
+    {
+      "epoch": 1.2744734081187958,
+      "grad_norm": 0.7212820649147034,
+      "learning_rate": 2.1285396729503826e-05,
+      "loss": 2.4178,
+      "step": 15792
+    },
+    {
+      "epoch": 1.274554111855379,
+      "grad_norm": 0.7149149179458618,
+      "learning_rate": 2.127566087809798e-05,
+      "loss": 2.4023,
+      "step": 15793
+    },
+    {
+      "epoch": 1.2746348155919618,
+      "grad_norm": 0.7039671540260315,
+      "learning_rate": 2.126592698870846e-05,
+      "loss": 2.4667,
+      "step": 15794
+    },
+    {
+      "epoch": 1.2747155193285449,
+      "grad_norm": 0.806849479675293,
+      "learning_rate": 2.1256195061577877e-05,
+      "loss": 2.4741,
+      "step": 15795
+    },
+    {
+      "epoch": 1.274796223065128,
+      "grad_norm": 0.7544776797294617,
+      "learning_rate": 2.124646509694872e-05,
+      "loss": 2.4258,
+      "step": 15796
+    },
+    {
+      "epoch": 1.2748769268017108,
+      "grad_norm": 0.6946810483932495,
+      "learning_rate": 2.1236737095063518e-05,
+      "loss": 2.4088,
+      "step": 15797
+    },
+    {
+      "epoch": 1.274957630538294,
+      "grad_norm": 0.7714219093322754,
+      "learning_rate": 2.1227011056164714e-05,
+      "loss": 2.4705,
+      "step": 15798
+    },
+    {
+      "epoch": 1.275038334274877,
+      "grad_norm": 0.6789658665657043,
+      "learning_rate": 2.121728698049471e-05,
+      "loss": 2.4692,
+      "step": 15799
+    },
+    {
+      "epoch": 1.2751190380114599,
+      "grad_norm": 0.7003477215766907,
+      "learning_rate": 2.120756486829586e-05,
+      "loss": 2.4437,
+      "step": 15800
+    },
+    {
+      "epoch": 1.275199741748043,
+      "grad_norm": 0.6802948117256165,
+      "learning_rate": 2.1197844719810455e-05,
+      "loss": 2.4002,
+      "step": 15801
+    },
+    {
+      "epoch": 1.275280445484626,
+      "grad_norm": 0.67823326587677,
+      "learning_rate": 2.1188126535280773e-05,
+      "loss": 2.5119,
+      "step": 15802
+    },
+    {
+      "epoch": 1.275361149221209,
+      "grad_norm": 0.6580843925476074,
+      "learning_rate": 2.1178410314948972e-05,
+      "loss": 2.3814,
+      "step": 15803
+    },
+    {
+      "epoch": 1.275441852957792,
+      "grad_norm": 0.681642472743988,
+      "learning_rate": 2.1168696059057226e-05,
+      "loss": 2.4206,
+      "step": 15804
+    },
+    {
+      "epoch": 1.275522556694375,
+      "grad_norm": 0.7483543753623962,
+      "learning_rate": 2.1158983767847674e-05,
+      "loss": 2.4633,
+      "step": 15805
+    },
+    {
+      "epoch": 1.275603260430958,
+      "grad_norm": 0.6565235257148743,
+      "learning_rate": 2.11492734415623e-05,
+      "loss": 2.4145,
+      "step": 15806
+    },
+    {
+      "epoch": 1.275683964167541,
+      "grad_norm": 0.6606764793395996,
+      "learning_rate": 2.1139565080443157e-05,
+      "loss": 2.3935,
+      "step": 15807
+    },
+    {
+      "epoch": 1.275764667904124,
+      "grad_norm": 0.7915800213813782,
+      "learning_rate": 2.1129858684732206e-05,
+      "loss": 2.4288,
+      "step": 15808
+    },
+    {
+      "epoch": 1.275845371640707,
+      "grad_norm": 0.6763594746589661,
+      "learning_rate": 2.112015425467133e-05,
+      "loss": 2.4147,
+      "step": 15809
+    },
+    {
+      "epoch": 1.2759260753772899,
+      "grad_norm": 0.6886053085327148,
+      "learning_rate": 2.1110451790502405e-05,
+      "loss": 2.3798,
+      "step": 15810
+    },
+    {
+      "epoch": 1.276006779113873,
+      "grad_norm": 0.686122715473175,
+      "learning_rate": 2.110075129246728e-05,
+      "loss": 2.3896,
+      "step": 15811
+    },
+    {
+      "epoch": 1.276087482850456,
+      "grad_norm": 0.6989614367485046,
+      "learning_rate": 2.109105276080764e-05,
+      "loss": 2.4533,
+      "step": 15812
+    },
+    {
+      "epoch": 1.276168186587039,
+      "grad_norm": 0.6818450689315796,
+      "learning_rate": 2.1081356195765232e-05,
+      "loss": 2.4012,
+      "step": 15813
+    },
+    {
+      "epoch": 1.276248890323622,
+      "grad_norm": 0.7492663860321045,
+      "learning_rate": 2.107166159758176e-05,
+      "loss": 2.4269,
+      "step": 15814
+    },
+    {
+      "epoch": 1.276329594060205,
+      "grad_norm": 0.6752359867095947,
+      "learning_rate": 2.1061968966498767e-05,
+      "loss": 2.4478,
+      "step": 15815
+    },
+    {
+      "epoch": 1.276410297796788,
+      "grad_norm": 0.6784162521362305,
+      "learning_rate": 2.1052278302757854e-05,
+      "loss": 2.4853,
+      "step": 15816
+    },
+    {
+      "epoch": 1.276491001533371,
+      "grad_norm": 0.7273215651512146,
+      "learning_rate": 2.104258960660055e-05,
+      "loss": 2.4365,
+      "step": 15817
+    },
+    {
+      "epoch": 1.2765717052699541,
+      "grad_norm": 0.7021621465682983,
+      "learning_rate": 2.1032902878268323e-05,
+      "loss": 2.4665,
+      "step": 15818
+    },
+    {
+      "epoch": 1.276652409006537,
+      "grad_norm": 0.666828989982605,
+      "learning_rate": 2.102321811800253e-05,
+      "loss": 2.3922,
+      "step": 15819
+    },
+    {
+      "epoch": 1.27673311274312,
+      "grad_norm": 0.6780487298965454,
+      "learning_rate": 2.1013535326044608e-05,
+      "loss": 2.4072,
+      "step": 15820
+    },
+    {
+      "epoch": 1.276813816479703,
+      "grad_norm": 0.6474688053131104,
+      "learning_rate": 2.1003854502635888e-05,
+      "loss": 2.4145,
+      "step": 15821
+    },
+    {
+      "epoch": 1.276894520216286,
+      "grad_norm": 0.6712753772735596,
+      "learning_rate": 2.0994175648017587e-05,
+      "loss": 2.4349,
+      "step": 15822
+    },
+    {
+      "epoch": 1.2769752239528689,
+      "grad_norm": 0.6705189943313599,
+      "learning_rate": 2.098449876243096e-05,
+      "loss": 2.4376,
+      "step": 15823
+    },
+    {
+      "epoch": 1.277055927689452,
+      "grad_norm": 0.6794685125350952,
+      "learning_rate": 2.0974823846117197e-05,
+      "loss": 2.3717,
+      "step": 15824
+    },
+    {
+      "epoch": 1.277136631426035,
+      "grad_norm": 0.7145677804946899,
+      "learning_rate": 2.0965150899317364e-05,
+      "loss": 2.3829,
+      "step": 15825
+    },
+    {
+      "epoch": 1.277217335162618,
+      "grad_norm": 0.7043245434761047,
+      "learning_rate": 2.095547992227257e-05,
+      "loss": 2.405,
+      "step": 15826
+    },
+    {
+      "epoch": 1.277298038899201,
+      "grad_norm": 0.7969205379486084,
+      "learning_rate": 2.0945810915223873e-05,
+      "loss": 2.4115,
+      "step": 15827
+    },
+    {
+      "epoch": 1.277378742635784,
+      "grad_norm": 0.657482385635376,
+      "learning_rate": 2.0936143878412186e-05,
+      "loss": 2.372,
+      "step": 15828
+    },
+    {
+      "epoch": 1.277459446372367,
+      "grad_norm": 0.7315167784690857,
+      "learning_rate": 2.0926478812078466e-05,
+      "loss": 2.4372,
+      "step": 15829
+    },
+    {
+      "epoch": 1.27754015010895,
+      "grad_norm": 0.6985061764717102,
+      "learning_rate": 2.09168157164636e-05,
+      "loss": 2.3901,
+      "step": 15830
+    },
+    {
+      "epoch": 1.2776208538455331,
+      "grad_norm": 0.6906184554100037,
+      "learning_rate": 2.0907154591808408e-05,
+      "loss": 2.4562,
+      "step": 15831
+    },
+    {
+      "epoch": 1.277701557582116,
+      "grad_norm": 0.655094563961029,
+      "learning_rate": 2.0897495438353676e-05,
+      "loss": 2.451,
+      "step": 15832
+    },
+    {
+      "epoch": 1.277782261318699,
+      "grad_norm": 0.7663134932518005,
+      "learning_rate": 2.0887838256340143e-05,
+      "loss": 2.4634,
+      "step": 15833
+    },
+    {
+      "epoch": 1.2778629650552822,
+      "grad_norm": 0.7164491415023804,
+      "learning_rate": 2.087818304600849e-05,
+      "loss": 2.4624,
+      "step": 15834
+    },
+    {
+      "epoch": 1.277943668791865,
+      "grad_norm": 0.6962822079658508,
+      "learning_rate": 2.0868529807599336e-05,
+      "loss": 2.4325,
+      "step": 15835
+    },
+    {
+      "epoch": 1.2780243725284481,
+      "grad_norm": 0.702985405921936,
+      "learning_rate": 2.0858878541353255e-05,
+      "loss": 2.4219,
+      "step": 15836
+    },
+    {
+      "epoch": 1.278105076265031,
+      "grad_norm": 0.7605595588684082,
+      "learning_rate": 2.0849229247510826e-05,
+      "loss": 2.4201,
+      "step": 15837
+    },
+    {
+      "epoch": 1.278185780001614,
+      "grad_norm": 0.8479344248771667,
+      "learning_rate": 2.083958192631249e-05,
+      "loss": 2.4689,
+      "step": 15838
+    },
+    {
+      "epoch": 1.278266483738197,
+      "grad_norm": 0.7241235375404358,
+      "learning_rate": 2.082993657799869e-05,
+      "loss": 2.4861,
+      "step": 15839
+    },
+    {
+      "epoch": 1.27834718747478,
+      "grad_norm": 0.7069835066795349,
+      "learning_rate": 2.0820293202809827e-05,
+      "loss": 2.3759,
+      "step": 15840
+    },
+    {
+      "epoch": 1.2784278912113631,
+      "grad_norm": 0.6606370210647583,
+      "learning_rate": 2.0810651800986237e-05,
+      "loss": 2.4444,
+      "step": 15841
+    },
+    {
+      "epoch": 1.278508594947946,
+      "grad_norm": 0.6608174443244934,
+      "learning_rate": 2.08010123727682e-05,
+      "loss": 2.4339,
+      "step": 15842
+    },
+    {
+      "epoch": 1.278589298684529,
+      "grad_norm": 0.751000702381134,
+      "learning_rate": 2.0791374918396e-05,
+      "loss": 2.4327,
+      "step": 15843
+    },
+    {
+      "epoch": 1.2786700024211122,
+      "grad_norm": 0.7223808765411377,
+      "learning_rate": 2.0781739438109748e-05,
+      "loss": 2.3573,
+      "step": 15844
+    },
+    {
+      "epoch": 1.278750706157695,
+      "grad_norm": 0.6872109770774841,
+      "learning_rate": 2.0772105932149642e-05,
+      "loss": 2.3973,
+      "step": 15845
+    },
+    {
+      "epoch": 1.2788314098942781,
+      "grad_norm": 0.6967385411262512,
+      "learning_rate": 2.0762474400755762e-05,
+      "loss": 2.4622,
+      "step": 15846
+    },
+    {
+      "epoch": 1.2789121136308612,
+      "grad_norm": 0.7289159893989563,
+      "learning_rate": 2.0752844844168163e-05,
+      "loss": 2.4507,
+      "step": 15847
+    },
+    {
+      "epoch": 1.278992817367444,
+      "grad_norm": 0.7735978364944458,
+      "learning_rate": 2.0743217262626802e-05,
+      "loss": 2.4341,
+      "step": 15848
+    },
+    {
+      "epoch": 1.2790735211040272,
+      "grad_norm": 0.7209177017211914,
+      "learning_rate": 2.0733591656371655e-05,
+      "loss": 2.4024,
+      "step": 15849
+    },
+    {
+      "epoch": 1.2791542248406103,
+      "grad_norm": 0.6789259314537048,
+      "learning_rate": 2.0723968025642604e-05,
+      "loss": 2.3809,
+      "step": 15850
+    },
+    {
+      "epoch": 1.2792349285771931,
+      "grad_norm": 0.6972812414169312,
+      "learning_rate": 2.0714346370679495e-05,
+      "loss": 2.3986,
+      "step": 15851
+    },
+    {
+      "epoch": 1.2793156323137762,
+      "grad_norm": 0.7144166827201843,
+      "learning_rate": 2.070472669172213e-05,
+      "loss": 2.4241,
+      "step": 15852
+    },
+    {
+      "epoch": 1.279396336050359,
+      "grad_norm": 0.7325223088264465,
+      "learning_rate": 2.0695108989010282e-05,
+      "loss": 2.452,
+      "step": 15853
+    },
+    {
+      "epoch": 1.2794770397869422,
+      "grad_norm": 0.6900116205215454,
+      "learning_rate": 2.0685493262783608e-05,
+      "loss": 2.4091,
+      "step": 15854
+    },
+    {
+      "epoch": 1.279557743523525,
+      "grad_norm": 0.6846197843551636,
+      "learning_rate": 2.0675879513281758e-05,
+      "loss": 2.4337,
+      "step": 15855
+    },
+    {
+      "epoch": 1.2796384472601081,
+      "grad_norm": 0.6901541352272034,
+      "learning_rate": 2.0666267740744372e-05,
+      "loss": 2.4586,
+      "step": 15856
+    },
+    {
+      "epoch": 1.2797191509966912,
+      "grad_norm": 0.6842665672302246,
+      "learning_rate": 2.0656657945410953e-05,
+      "loss": 2.4383,
+      "step": 15857
+    },
+    {
+      "epoch": 1.279799854733274,
+      "grad_norm": 0.7450493574142456,
+      "learning_rate": 2.0647050127521028e-05,
+      "loss": 2.4308,
+      "step": 15858
+    },
+    {
+      "epoch": 1.2798805584698572,
+      "grad_norm": 0.6928436160087585,
+      "learning_rate": 2.0637444287314033e-05,
+      "loss": 2.4726,
+      "step": 15859
+    },
+    {
+      "epoch": 1.2799612622064402,
+      "grad_norm": 0.6539968252182007,
+      "learning_rate": 2.06278404250294e-05,
+      "loss": 2.3983,
+      "step": 15860
+    },
+    {
+      "epoch": 1.280041965943023,
+      "grad_norm": 0.7183163166046143,
+      "learning_rate": 2.0618238540906444e-05,
+      "loss": 2.4172,
+      "step": 15861
+    },
+    {
+      "epoch": 1.2801226696796062,
+      "grad_norm": 0.7070814371109009,
+      "learning_rate": 2.0608638635184507e-05,
+      "loss": 2.4018,
+      "step": 15862
+    },
+    {
+      "epoch": 1.2802033734161893,
+      "grad_norm": 0.7589142918586731,
+      "learning_rate": 2.0599040708102847e-05,
+      "loss": 2.4175,
+      "step": 15863
+    },
+    {
+      "epoch": 1.2802840771527721,
+      "grad_norm": 0.6945414543151855,
+      "learning_rate": 2.0589444759900613e-05,
+      "loss": 2.4093,
+      "step": 15864
+    },
+    {
+      "epoch": 1.2803647808893552,
+      "grad_norm": 0.685482919216156,
+      "learning_rate": 2.0579850790817003e-05,
+      "loss": 2.4388,
+      "step": 15865
+    },
+    {
+      "epoch": 1.280445484625938,
+      "grad_norm": 0.7089706063270569,
+      "learning_rate": 2.0570258801091148e-05,
+      "loss": 2.3779,
+      "step": 15866
+    },
+    {
+      "epoch": 1.2805261883625212,
+      "grad_norm": 0.6994217038154602,
+      "learning_rate": 2.0560668790962046e-05,
+      "loss": 2.3757,
+      "step": 15867
+    },
+    {
+      "epoch": 1.280606892099104,
+      "grad_norm": 0.7170232534408569,
+      "learning_rate": 2.055108076066874e-05,
+      "loss": 2.4087,
+      "step": 15868
+    },
+    {
+      "epoch": 1.2806875958356871,
+      "grad_norm": 0.7008751034736633,
+      "learning_rate": 2.0541494710450206e-05,
+      "loss": 2.4384,
+      "step": 15869
+    },
+    {
+      "epoch": 1.2807682995722702,
+      "grad_norm": 0.6795800924301147,
+      "learning_rate": 2.053191064054527e-05,
+      "loss": 2.415,
+      "step": 15870
+    },
+    {
+      "epoch": 1.280849003308853,
+      "grad_norm": 0.6650210022926331,
+      "learning_rate": 2.0522328551192882e-05,
+      "loss": 2.4421,
+      "step": 15871
+    },
+    {
+      "epoch": 1.2809297070454362,
+      "grad_norm": 0.7045374512672424,
+      "learning_rate": 2.0512748442631858e-05,
+      "loss": 2.4285,
+      "step": 15872
+    },
+    {
+      "epoch": 1.2810104107820193,
+      "grad_norm": 0.6585350632667542,
+      "learning_rate": 2.0503170315100883e-05,
+      "loss": 2.3806,
+      "step": 15873
+    },
+    {
+      "epoch": 1.2810911145186021,
+      "grad_norm": 0.7833496332168579,
+      "learning_rate": 2.0493594168838725e-05,
+      "loss": 2.4557,
+      "step": 15874
+    },
+    {
+      "epoch": 1.2811718182551852,
+      "grad_norm": 0.7237457036972046,
+      "learning_rate": 2.0484020004084048e-05,
+      "loss": 2.3966,
+      "step": 15875
+    },
+    {
+      "epoch": 1.2812525219917683,
+      "grad_norm": 0.7416609525680542,
+      "learning_rate": 2.0474447821075426e-05,
+      "loss": 2.3729,
+      "step": 15876
+    },
+    {
+      "epoch": 1.2813332257283512,
+      "grad_norm": 0.7148095369338989,
+      "learning_rate": 2.046487762005146e-05,
+      "loss": 2.4163,
+      "step": 15877
+    },
+    {
+      "epoch": 1.2814139294649343,
+      "grad_norm": 0.670281171798706,
+      "learning_rate": 2.0455309401250632e-05,
+      "loss": 2.383,
+      "step": 15878
+    },
+    {
+      "epoch": 1.2814946332015174,
+      "grad_norm": 0.6968950629234314,
+      "learning_rate": 2.0445743164911457e-05,
+      "loss": 2.3967,
+      "step": 15879
+    },
+    {
+      "epoch": 1.2815753369381002,
+      "grad_norm": 0.783441960811615,
+      "learning_rate": 2.0436178911272298e-05,
+      "loss": 2.455,
+      "step": 15880
+    },
+    {
+      "epoch": 1.2816560406746833,
+      "grad_norm": 0.709032416343689,
+      "learning_rate": 2.0426616640571518e-05,
+      "loss": 2.4207,
+      "step": 15881
+    },
+    {
+      "epoch": 1.2817367444112662,
+      "grad_norm": 0.6727990508079529,
+      "learning_rate": 2.0417056353047504e-05,
+      "loss": 2.4115,
+      "step": 15882
+    },
+    {
+      "epoch": 1.2818174481478493,
+      "grad_norm": 0.7336034774780273,
+      "learning_rate": 2.0407498048938445e-05,
+      "loss": 2.43,
+      "step": 15883
+    },
+    {
+      "epoch": 1.2818981518844321,
+      "grad_norm": 0.7649042010307312,
+      "learning_rate": 2.0397941728482604e-05,
+      "loss": 2.4655,
+      "step": 15884
+    },
+    {
+      "epoch": 1.2819788556210152,
+      "grad_norm": 0.7218052744865417,
+      "learning_rate": 2.038838739191816e-05,
+      "loss": 2.4872,
+      "step": 15885
+    },
+    {
+      "epoch": 1.2820595593575983,
+      "grad_norm": 0.7192350625991821,
+      "learning_rate": 2.0378835039483178e-05,
+      "loss": 2.4751,
+      "step": 15886
+    },
+    {
+      "epoch": 1.2821402630941812,
+      "grad_norm": 0.7059212923049927,
+      "learning_rate": 2.0369284671415768e-05,
+      "loss": 2.43,
+      "step": 15887
+    },
+    {
+      "epoch": 1.2822209668307643,
+      "grad_norm": 0.7387098073959351,
+      "learning_rate": 2.0359736287953956e-05,
+      "loss": 2.4281,
+      "step": 15888
+    },
+    {
+      "epoch": 1.2823016705673473,
+      "grad_norm": 0.7454321980476379,
+      "learning_rate": 2.035018988933568e-05,
+      "loss": 2.4372,
+      "step": 15889
+    },
+    {
+      "epoch": 1.2823823743039302,
+      "grad_norm": 0.6822765469551086,
+      "learning_rate": 2.034064547579888e-05,
+      "loss": 2.3728,
+      "step": 15890
+    },
+    {
+      "epoch": 1.2824630780405133,
+      "grad_norm": 0.6917527914047241,
+      "learning_rate": 2.0331103047581412e-05,
+      "loss": 2.3997,
+      "step": 15891
+    },
+    {
+      "epoch": 1.2825437817770964,
+      "grad_norm": 0.6734376549720764,
+      "learning_rate": 2.032156260492113e-05,
+      "loss": 2.4495,
+      "step": 15892
+    },
+    {
+      "epoch": 1.2826244855136792,
+      "grad_norm": 0.7222443222999573,
+      "learning_rate": 2.0312024148055776e-05,
+      "loss": 2.3466,
+      "step": 15893
+    },
+    {
+      "epoch": 1.2827051892502623,
+      "grad_norm": 0.703714907169342,
+      "learning_rate": 2.030248767722309e-05,
+      "loss": 2.4599,
+      "step": 15894
+    },
+    {
+      "epoch": 1.2827858929868454,
+      "grad_norm": 0.655161440372467,
+      "learning_rate": 2.029295319266078e-05,
+      "loss": 2.3896,
+      "step": 15895
+    },
+    {
+      "epoch": 1.2828665967234283,
+      "grad_norm": 0.6449242234230042,
+      "learning_rate": 2.028342069460639e-05,
+      "loss": 2.3511,
+      "step": 15896
+    },
+    {
+      "epoch": 1.2829473004600114,
+      "grad_norm": 0.6578382849693298,
+      "learning_rate": 2.027389018329755e-05,
+      "loss": 2.3678,
+      "step": 15897
+    },
+    {
+      "epoch": 1.2830280041965942,
+      "grad_norm": 0.7047572731971741,
+      "learning_rate": 2.0264361658971797e-05,
+      "loss": 2.4522,
+      "step": 15898
+    },
+    {
+      "epoch": 1.2831087079331773,
+      "grad_norm": 0.7310267090797424,
+      "learning_rate": 2.0254835121866554e-05,
+      "loss": 2.4117,
+      "step": 15899
+    },
+    {
+      "epoch": 1.2831894116697602,
+      "grad_norm": 0.7020776867866516,
+      "learning_rate": 2.024531057221927e-05,
+      "loss": 2.4033,
+      "step": 15900
+    },
+    {
+      "epoch": 1.2832701154063433,
+      "grad_norm": 0.6967746615409851,
+      "learning_rate": 2.023578801026733e-05,
+      "loss": 2.3491,
+      "step": 15901
+    },
+    {
+      "epoch": 1.2833508191429264,
+      "grad_norm": 0.7062339782714844,
+      "learning_rate": 2.022626743624807e-05,
+      "loss": 2.4598,
+      "step": 15902
+    },
+    {
+      "epoch": 1.2834315228795092,
+      "grad_norm": 0.730625331401825,
+      "learning_rate": 2.0216748850398748e-05,
+      "loss": 2.4995,
+      "step": 15903
+    },
+    {
+      "epoch": 1.2835122266160923,
+      "grad_norm": 0.6634403467178345,
+      "learning_rate": 2.020723225295662e-05,
+      "loss": 2.3843,
+      "step": 15904
+    },
+    {
+      "epoch": 1.2835929303526754,
+      "grad_norm": 0.6924816966056824,
+      "learning_rate": 2.019771764415883e-05,
+      "loss": 2.4258,
+      "step": 15905
+    },
+    {
+      "epoch": 1.2836736340892583,
+      "grad_norm": 0.7127227187156677,
+      "learning_rate": 2.018820502424251e-05,
+      "loss": 2.4038,
+      "step": 15906
+    },
+    {
+      "epoch": 1.2837543378258414,
+      "grad_norm": 0.7108431458473206,
+      "learning_rate": 2.0178694393444785e-05,
+      "loss": 2.4571,
+      "step": 15907
+    },
+    {
+      "epoch": 1.2838350415624245,
+      "grad_norm": 0.7478229999542236,
+      "learning_rate": 2.016918575200262e-05,
+      "loss": 2.4526,
+      "step": 15908
+    },
+    {
+      "epoch": 1.2839157452990073,
+      "grad_norm": 0.65651935338974,
+      "learning_rate": 2.015967910015303e-05,
+      "loss": 2.434,
+      "step": 15909
+    },
+    {
+      "epoch": 1.2839964490355904,
+      "grad_norm": 0.7285312414169312,
+      "learning_rate": 2.015017443813294e-05,
+      "loss": 2.3857,
+      "step": 15910
+    },
+    {
+      "epoch": 1.2840771527721733,
+      "grad_norm": 0.6947231292724609,
+      "learning_rate": 2.014067176617923e-05,
+      "loss": 2.4294,
+      "step": 15911
+    },
+    {
+      "epoch": 1.2841578565087564,
+      "grad_norm": 0.6965867877006531,
+      "learning_rate": 2.0131171084528744e-05,
+      "loss": 2.4514,
+      "step": 15912
+    },
+    {
+      "epoch": 1.2842385602453392,
+      "grad_norm": 0.6962311863899231,
+      "learning_rate": 2.0121672393418246e-05,
+      "loss": 2.4391,
+      "step": 15913
+    },
+    {
+      "epoch": 1.2843192639819223,
+      "grad_norm": 0.6687992215156555,
+      "learning_rate": 2.01121756930845e-05,
+      "loss": 2.4266,
+      "step": 15914
+    },
+    {
+      "epoch": 1.2843999677185054,
+      "grad_norm": 0.7118954658508301,
+      "learning_rate": 2.0102680983764145e-05,
+      "loss": 2.3436,
+      "step": 15915
+    },
+    {
+      "epoch": 1.2844806714550883,
+      "grad_norm": 0.6866199970245361,
+      "learning_rate": 2.009318826569382e-05,
+      "loss": 2.3719,
+      "step": 15916
+    },
+    {
+      "epoch": 1.2845613751916714,
+      "grad_norm": 0.6701404452323914,
+      "learning_rate": 2.008369753911016e-05,
+      "loss": 2.4875,
+      "step": 15917
+    },
+    {
+      "epoch": 1.2846420789282544,
+      "grad_norm": 0.7020917534828186,
+      "learning_rate": 2.007420880424963e-05,
+      "loss": 2.3871,
+      "step": 15918
+    },
+    {
+      "epoch": 1.2847227826648373,
+      "grad_norm": 0.6865704655647278,
+      "learning_rate": 2.006472206134875e-05,
+      "loss": 2.3815,
+      "step": 15919
+    },
+    {
+      "epoch": 1.2848034864014204,
+      "grad_norm": 0.7106871008872986,
+      "learning_rate": 2.0055237310643948e-05,
+      "loss": 2.4276,
+      "step": 15920
+    },
+    {
+      "epoch": 1.2848841901380035,
+      "grad_norm": 0.6891976594924927,
+      "learning_rate": 2.004575455237161e-05,
+      "loss": 2.3641,
+      "step": 15921
+    },
+    {
+      "epoch": 1.2849648938745863,
+      "grad_norm": 0.6385056972503662,
+      "learning_rate": 2.0036273786768067e-05,
+      "loss": 2.3898,
+      "step": 15922
+    },
+    {
+      "epoch": 1.2850455976111694,
+      "grad_norm": 0.7038321495056152,
+      "learning_rate": 2.0026795014069633e-05,
+      "loss": 2.4688,
+      "step": 15923
+    },
+    {
+      "epoch": 1.2851263013477525,
+      "grad_norm": 0.6310208439826965,
+      "learning_rate": 2.0017318234512494e-05,
+      "loss": 2.3821,
+      "step": 15924
+    },
+    {
+      "epoch": 1.2852070050843354,
+      "grad_norm": 0.6989426016807556,
+      "learning_rate": 2.0007843448332865e-05,
+      "loss": 2.434,
+      "step": 15925
+    },
+    {
+      "epoch": 1.2852877088209185,
+      "grad_norm": 0.6666426658630371,
+      "learning_rate": 1.9998370655766886e-05,
+      "loss": 2.4687,
+      "step": 15926
+    },
+    {
+      "epoch": 1.2853684125575013,
+      "grad_norm": 0.6421633958816528,
+      "learning_rate": 1.9988899857050648e-05,
+      "loss": 2.4269,
+      "step": 15927
+    },
+    {
+      "epoch": 1.2854491162940844,
+      "grad_norm": 0.7229343056678772,
+      "learning_rate": 1.997943105242016e-05,
+      "loss": 2.4139,
+      "step": 15928
+    },
+    {
+      "epoch": 1.2855298200306673,
+      "grad_norm": 0.7168964743614197,
+      "learning_rate": 1.9969964242111427e-05,
+      "loss": 2.405,
+      "step": 15929
+    },
+    {
+      "epoch": 1.2856105237672504,
+      "grad_norm": 0.6824480891227722,
+      "learning_rate": 1.99604994263604e-05,
+      "loss": 2.3955,
+      "step": 15930
+    },
+    {
+      "epoch": 1.2856912275038335,
+      "grad_norm": 0.670956552028656,
+      "learning_rate": 1.995103660540294e-05,
+      "loss": 2.3743,
+      "step": 15931
+    },
+    {
+      "epoch": 1.2857719312404163,
+      "grad_norm": 0.7057971954345703,
+      "learning_rate": 1.9941575779474864e-05,
+      "loss": 2.4496,
+      "step": 15932
+    },
+    {
+      "epoch": 1.2858526349769994,
+      "grad_norm": 0.7802264094352722,
+      "learning_rate": 1.9932116948812052e-05,
+      "loss": 2.4231,
+      "step": 15933
+    },
+    {
+      "epoch": 1.2859333387135825,
+      "grad_norm": 0.7151160836219788,
+      "learning_rate": 1.992266011365016e-05,
+      "loss": 2.4319,
+      "step": 15934
+    },
+    {
+      "epoch": 1.2860140424501654,
+      "grad_norm": 0.7078769207000732,
+      "learning_rate": 1.991320527422489e-05,
+      "loss": 2.4037,
+      "step": 15935
+    },
+    {
+      "epoch": 1.2860947461867485,
+      "grad_norm": 0.7483938336372375,
+      "learning_rate": 1.9903752430771927e-05,
+      "loss": 2.4946,
+      "step": 15936
+    },
+    {
+      "epoch": 1.2861754499233315,
+      "grad_norm": 0.7774620056152344,
+      "learning_rate": 1.9894301583526808e-05,
+      "loss": 2.4536,
+      "step": 15937
+    },
+    {
+      "epoch": 1.2862561536599144,
+      "grad_norm": 0.7311348915100098,
+      "learning_rate": 1.988485273272509e-05,
+      "loss": 2.4178,
+      "step": 15938
+    },
+    {
+      "epoch": 1.2863368573964975,
+      "grad_norm": 0.6821309328079224,
+      "learning_rate": 1.9875405878602282e-05,
+      "loss": 2.4851,
+      "step": 15939
+    },
+    {
+      "epoch": 1.2864175611330806,
+      "grad_norm": 0.7081651091575623,
+      "learning_rate": 1.9865961021393785e-05,
+      "loss": 2.4377,
+      "step": 15940
+    },
+    {
+      "epoch": 1.2864982648696635,
+      "grad_norm": 0.8093439340591431,
+      "learning_rate": 1.9856518161335014e-05,
+      "loss": 2.4681,
+      "step": 15941
+    },
+    {
+      "epoch": 1.2865789686062465,
+      "grad_norm": 0.6769521832466125,
+      "learning_rate": 1.984707729866131e-05,
+      "loss": 2.4231,
+      "step": 15942
+    },
+    {
+      "epoch": 1.2866596723428294,
+      "grad_norm": 0.6973356604576111,
+      "learning_rate": 1.983763843360795e-05,
+      "loss": 2.4144,
+      "step": 15943
+    },
+    {
+      "epoch": 1.2867403760794125,
+      "grad_norm": 0.7814682722091675,
+      "learning_rate": 1.9828201566410197e-05,
+      "loss": 2.3935,
+      "step": 15944
+    },
+    {
+      "epoch": 1.2868210798159954,
+      "grad_norm": 0.7545498609542847,
+      "learning_rate": 1.9818766697303236e-05,
+      "loss": 2.4136,
+      "step": 15945
+    },
+    {
+      "epoch": 1.2869017835525784,
+      "grad_norm": 0.7165581583976746,
+      "learning_rate": 1.9809333826522225e-05,
+      "loss": 2.3757,
+      "step": 15946
+    },
+    {
+      "epoch": 1.2869824872891615,
+      "grad_norm": 0.6812456846237183,
+      "learning_rate": 1.9799902954302208e-05,
+      "loss": 2.4143,
+      "step": 15947
+    },
+    {
+      "epoch": 1.2870631910257444,
+      "grad_norm": 0.7231366634368896,
+      "learning_rate": 1.9790474080878262e-05,
+      "loss": 2.4837,
+      "step": 15948
+    },
+    {
+      "epoch": 1.2871438947623275,
+      "grad_norm": 0.690916121006012,
+      "learning_rate": 1.9781047206485393e-05,
+      "loss": 2.4513,
+      "step": 15949
+    },
+    {
+      "epoch": 1.2872245984989106,
+      "grad_norm": 0.6608129143714905,
+      "learning_rate": 1.9771622331358485e-05,
+      "loss": 2.3908,
+      "step": 15950
+    },
+    {
+      "epoch": 1.2873053022354934,
+      "grad_norm": 0.7194501161575317,
+      "learning_rate": 1.976219945573249e-05,
+      "loss": 2.38,
+      "step": 15951
+    },
+    {
+      "epoch": 1.2873860059720765,
+      "grad_norm": 0.7315083146095276,
+      "learning_rate": 1.9752778579842213e-05,
+      "loss": 2.4351,
+      "step": 15952
+    },
+    {
+      "epoch": 1.2874667097086596,
+      "grad_norm": 0.7313492298126221,
+      "learning_rate": 1.974335970392246e-05,
+      "loss": 2.3531,
+      "step": 15953
+    },
+    {
+      "epoch": 1.2875474134452425,
+      "grad_norm": 0.6982418894767761,
+      "learning_rate": 1.9733942828207985e-05,
+      "loss": 2.4319,
+      "step": 15954
+    },
+    {
+      "epoch": 1.2876281171818256,
+      "grad_norm": 0.6664792895317078,
+      "learning_rate": 1.972452795293347e-05,
+      "loss": 2.3981,
+      "step": 15955
+    },
+    {
+      "epoch": 1.2877088209184087,
+      "grad_norm": 0.6849696040153503,
+      "learning_rate": 1.9715115078333578e-05,
+      "loss": 2.3952,
+      "step": 15956
+    },
+    {
+      "epoch": 1.2877895246549915,
+      "grad_norm": 0.7355225086212158,
+      "learning_rate": 1.9705704204642873e-05,
+      "loss": 2.4556,
+      "step": 15957
+    },
+    {
+      "epoch": 1.2878702283915746,
+      "grad_norm": 0.6850876808166504,
+      "learning_rate": 1.9696295332095906e-05,
+      "loss": 2.3873,
+      "step": 15958
+    },
+    {
+      "epoch": 1.2879509321281575,
+      "grad_norm": 0.6449069976806641,
+      "learning_rate": 1.9686888460927198e-05,
+      "loss": 2.4226,
+      "step": 15959
+    },
+    {
+      "epoch": 1.2880316358647406,
+      "grad_norm": 0.7517794966697693,
+      "learning_rate": 1.967748359137114e-05,
+      "loss": 2.377,
+      "step": 15960
+    },
+    {
+      "epoch": 1.2881123396013234,
+      "grad_norm": 0.6861303448677063,
+      "learning_rate": 1.9668080723662162e-05,
+      "loss": 2.4451,
+      "step": 15961
+    },
+    {
+      "epoch": 1.2881930433379065,
+      "grad_norm": 0.7025154829025269,
+      "learning_rate": 1.9658679858034602e-05,
+      "loss": 2.3856,
+      "step": 15962
+    },
+    {
+      "epoch": 1.2882737470744896,
+      "grad_norm": 0.6775577068328857,
+      "learning_rate": 1.964928099472275e-05,
+      "loss": 2.4383,
+      "step": 15963
+    },
+    {
+      "epoch": 1.2883544508110725,
+      "grad_norm": 0.6889605522155762,
+      "learning_rate": 1.963988413396086e-05,
+      "loss": 2.3766,
+      "step": 15964
+    },
+    {
+      "epoch": 1.2884351545476556,
+      "grad_norm": 0.6697166562080383,
+      "learning_rate": 1.9630489275983156e-05,
+      "loss": 2.44,
+      "step": 15965
+    },
+    {
+      "epoch": 1.2885158582842386,
+      "grad_norm": 0.6895437836647034,
+      "learning_rate": 1.96210964210237e-05,
+      "loss": 2.4242,
+      "step": 15966
+    },
+    {
+      "epoch": 1.2885965620208215,
+      "grad_norm": 0.6955164670944214,
+      "learning_rate": 1.9611705569316652e-05,
+      "loss": 2.3915,
+      "step": 15967
+    },
+    {
+      "epoch": 1.2886772657574046,
+      "grad_norm": 0.7133461236953735,
+      "learning_rate": 1.960231672109605e-05,
+      "loss": 2.4307,
+      "step": 15968
+    },
+    {
+      "epoch": 1.2887579694939877,
+      "grad_norm": 0.6874761581420898,
+      "learning_rate": 1.9592929876595857e-05,
+      "loss": 2.4371,
+      "step": 15969
+    },
+    {
+      "epoch": 1.2888386732305706,
+      "grad_norm": 0.7168406248092651,
+      "learning_rate": 1.9583545036050044e-05,
+      "loss": 2.4681,
+      "step": 15970
+    },
+    {
+      "epoch": 1.2889193769671536,
+      "grad_norm": 0.701874852180481,
+      "learning_rate": 1.9574162199692492e-05,
+      "loss": 2.4746,
+      "step": 15971
+    },
+    {
+      "epoch": 1.2890000807037365,
+      "grad_norm": 0.7118390202522278,
+      "learning_rate": 1.9564781367757058e-05,
+      "loss": 2.4139,
+      "step": 15972
+    },
+    {
+      "epoch": 1.2890807844403196,
+      "grad_norm": 0.6597239971160889,
+      "learning_rate": 1.955540254047753e-05,
+      "loss": 2.4346,
+      "step": 15973
+    },
+    {
+      "epoch": 1.2891614881769025,
+      "grad_norm": 0.7461068630218506,
+      "learning_rate": 1.9546025718087645e-05,
+      "loss": 2.4331,
+      "step": 15974
+    },
+    {
+      "epoch": 1.2892421919134855,
+      "grad_norm": 0.6992977857589722,
+      "learning_rate": 1.953665090082115e-05,
+      "loss": 2.424,
+      "step": 15975
+    },
+    {
+      "epoch": 1.2893228956500686,
+      "grad_norm": 0.6674031615257263,
+      "learning_rate": 1.9527278088911617e-05,
+      "loss": 2.4545,
+      "step": 15976
+    },
+    {
+      "epoch": 1.2894035993866515,
+      "grad_norm": 0.7377402782440186,
+      "learning_rate": 1.9517907282592662e-05,
+      "loss": 2.4625,
+      "step": 15977
+    },
+    {
+      "epoch": 1.2894843031232346,
+      "grad_norm": 0.720579206943512,
+      "learning_rate": 1.950853848209788e-05,
+      "loss": 2.4073,
+      "step": 15978
+    },
+    {
+      "epoch": 1.2895650068598177,
+      "grad_norm": 0.7221893668174744,
+      "learning_rate": 1.9499171687660688e-05,
+      "loss": 2.4056,
+      "step": 15979
+    },
+    {
+      "epoch": 1.2896457105964005,
+      "grad_norm": 0.7409725189208984,
+      "learning_rate": 1.9489806899514574e-05,
+      "loss": 2.3899,
+      "step": 15980
+    },
+    {
+      "epoch": 1.2897264143329836,
+      "grad_norm": 0.6946583986282349,
+      "learning_rate": 1.948044411789296e-05,
+      "loss": 2.4832,
+      "step": 15981
+    },
+    {
+      "epoch": 1.2898071180695667,
+      "grad_norm": 0.7031306028366089,
+      "learning_rate": 1.9471083343029096e-05,
+      "loss": 2.4265,
+      "step": 15982
+    },
+    {
+      "epoch": 1.2898878218061496,
+      "grad_norm": 0.660093367099762,
+      "learning_rate": 1.946172457515637e-05,
+      "loss": 2.4883,
+      "step": 15983
+    },
+    {
+      "epoch": 1.2899685255427327,
+      "grad_norm": 0.700641930103302,
+      "learning_rate": 1.945236781450802e-05,
+      "loss": 2.4096,
+      "step": 15984
+    },
+    {
+      "epoch": 1.2900492292793158,
+      "grad_norm": 0.7350760698318481,
+      "learning_rate": 1.9443013061317205e-05,
+      "loss": 2.4161,
+      "step": 15985
+    },
+    {
+      "epoch": 1.2901299330158986,
+      "grad_norm": 0.7567386031150818,
+      "learning_rate": 1.9433660315817072e-05,
+      "loss": 2.3978,
+      "step": 15986
+    },
+    {
+      "epoch": 1.2902106367524817,
+      "grad_norm": 0.7471369504928589,
+      "learning_rate": 1.9424309578240717e-05,
+      "loss": 2.4079,
+      "step": 15987
+    },
+    {
+      "epoch": 1.2902913404890646,
+      "grad_norm": 0.6630815267562866,
+      "learning_rate": 1.941496084882124e-05,
+      "loss": 2.4223,
+      "step": 15988
+    },
+    {
+      "epoch": 1.2903720442256477,
+      "grad_norm": 0.687224268913269,
+      "learning_rate": 1.940561412779155e-05,
+      "loss": 2.4413,
+      "step": 15989
+    },
+    {
+      "epoch": 1.2904527479622305,
+      "grad_norm": 0.6989685297012329,
+      "learning_rate": 1.9396269415384637e-05,
+      "loss": 2.3651,
+      "step": 15990
+    },
+    {
+      "epoch": 1.2905334516988136,
+      "grad_norm": 0.7256720066070557,
+      "learning_rate": 1.938692671183342e-05,
+      "loss": 2.4526,
+      "step": 15991
+    },
+    {
+      "epoch": 1.2906141554353967,
+      "grad_norm": 0.692032516002655,
+      "learning_rate": 1.9377586017370685e-05,
+      "loss": 2.3936,
+      "step": 15992
+    },
+    {
+      "epoch": 1.2906948591719796,
+      "grad_norm": 0.6733511686325073,
+      "learning_rate": 1.936824733222925e-05,
+      "loss": 2.4691,
+      "step": 15993
+    },
+    {
+      "epoch": 1.2907755629085627,
+      "grad_norm": 0.6698563098907471,
+      "learning_rate": 1.935891065664187e-05,
+      "loss": 2.3904,
+      "step": 15994
+    },
+    {
+      "epoch": 1.2908562666451457,
+      "grad_norm": 0.660521388053894,
+      "learning_rate": 1.934957599084123e-05,
+      "loss": 2.4647,
+      "step": 15995
+    },
+    {
+      "epoch": 1.2909369703817286,
+      "grad_norm": 0.6714615821838379,
+      "learning_rate": 1.9340243335059982e-05,
+      "loss": 2.403,
+      "step": 15996
+    },
+    {
+      "epoch": 1.2910176741183117,
+      "grad_norm": 0.726099967956543,
+      "learning_rate": 1.9330912689530746e-05,
+      "loss": 2.4101,
+      "step": 15997
+    },
+    {
+      "epoch": 1.2910983778548948,
+      "grad_norm": 0.6585896015167236,
+      "learning_rate": 1.932158405448601e-05,
+      "loss": 2.3813,
+      "step": 15998
+    },
+    {
+      "epoch": 1.2911790815914777,
+      "grad_norm": 0.7967908382415771,
+      "learning_rate": 1.9312257430158286e-05,
+      "loss": 2.4188,
+      "step": 15999
+    },
+    {
+      "epoch": 1.2912597853280607,
+      "grad_norm": 0.7340367436408997,
+      "learning_rate": 1.9302932816780063e-05,
+      "loss": 2.4642,
+      "step": 16000
+    },
+    {
+      "epoch": 1.2912597853280607,
+      "eval_loss": 2.3791537284851074,
+      "eval_runtime": 780.6124,
+      "eval_samples_per_second": 3.356,
+      "eval_steps_per_second": 0.56,
+      "step": 16000
+    },
+    {
+      "epoch": 1.2913404890646438,
+      "grad_norm": 0.6778663992881775,
+      "learning_rate": 1.929361021458367e-05,
+      "loss": 2.4057,
+      "step": 16001
+    },
+    {
+      "epoch": 1.2914211928012267,
+      "grad_norm": 0.6982381343841553,
+      "learning_rate": 1.9284289623801477e-05,
+      "loss": 2.4376,
+      "step": 16002
+    },
+    {
+      "epoch": 1.2915018965378098,
+      "grad_norm": 0.6956612467765808,
+      "learning_rate": 1.927497104466578e-05,
+      "loss": 2.4485,
+      "step": 16003
+    },
+    {
+      "epoch": 1.2915826002743926,
+      "grad_norm": 0.6780211925506592,
+      "learning_rate": 1.9265654477408825e-05,
+      "loss": 2.4233,
+      "step": 16004
+    },
+    {
+      "epoch": 1.2916633040109757,
+      "grad_norm": 0.6869028806686401,
+      "learning_rate": 1.92563399222628e-05,
+      "loss": 2.4156,
+      "step": 16005
+    },
+    {
+      "epoch": 1.2917440077475586,
+      "grad_norm": 0.6402696967124939,
+      "learning_rate": 1.9247027379459848e-05,
+      "loss": 2.4208,
+      "step": 16006
+    },
+    {
+      "epoch": 1.2918247114841417,
+      "grad_norm": 0.6868177652359009,
+      "learning_rate": 1.92377168492321e-05,
+      "loss": 2.4067,
+      "step": 16007
+    },
+    {
+      "epoch": 1.2919054152207248,
+      "grad_norm": 0.7152438759803772,
+      "learning_rate": 1.922840833181152e-05,
+      "loss": 2.3944,
+      "step": 16008
+    },
+    {
+      "epoch": 1.2919861189573076,
+      "grad_norm": 0.6467335820198059,
+      "learning_rate": 1.921910182743015e-05,
+      "loss": 2.4064,
+      "step": 16009
+    },
+    {
+      "epoch": 1.2920668226938907,
+      "grad_norm": 0.6918551325798035,
+      "learning_rate": 1.9209797336319956e-05,
+      "loss": 2.4457,
+      "step": 16010
+    },
+    {
+      "epoch": 1.2921475264304738,
+      "grad_norm": 0.7308588027954102,
+      "learning_rate": 1.920049485871278e-05,
+      "loss": 2.3785,
+      "step": 16011
+    },
+    {
+      "epoch": 1.2922282301670567,
+      "grad_norm": 0.6918718814849854,
+      "learning_rate": 1.9191194394840472e-05,
+      "loss": 2.4645,
+      "step": 16012
+    },
+    {
+      "epoch": 1.2923089339036398,
+      "grad_norm": 0.7048078775405884,
+      "learning_rate": 1.9181895944934848e-05,
+      "loss": 2.4082,
+      "step": 16013
+    },
+    {
+      "epoch": 1.2923896376402229,
+      "grad_norm": 0.7175794839859009,
+      "learning_rate": 1.917259950922763e-05,
+      "loss": 2.4521,
+      "step": 16014
+    },
+    {
+      "epoch": 1.2924703413768057,
+      "grad_norm": 0.6895543932914734,
+      "learning_rate": 1.916330508795051e-05,
+      "loss": 2.4058,
+      "step": 16015
+    },
+    {
+      "epoch": 1.2925510451133888,
+      "grad_norm": 0.6951895952224731,
+      "learning_rate": 1.9154012681335176e-05,
+      "loss": 2.4274,
+      "step": 16016
+    },
+    {
+      "epoch": 1.2926317488499717,
+      "grad_norm": 0.6807428598403931,
+      "learning_rate": 1.9144722289613148e-05,
+      "loss": 2.4008,
+      "step": 16017
+    },
+    {
+      "epoch": 1.2927124525865548,
+      "grad_norm": 0.6643410325050354,
+      "learning_rate": 1.9135433913015997e-05,
+      "loss": 2.4036,
+      "step": 16018
+    },
+    {
+      "epoch": 1.2927931563231376,
+      "grad_norm": 0.7283294796943665,
+      "learning_rate": 1.912614755177522e-05,
+      "loss": 2.4118,
+      "step": 16019
+    },
+    {
+      "epoch": 1.2928738600597207,
+      "grad_norm": 0.7516021132469177,
+      "learning_rate": 1.911686320612227e-05,
+      "loss": 2.3983,
+      "step": 16020
+    },
+    {
+      "epoch": 1.2929545637963038,
+      "grad_norm": 0.7314203381538391,
+      "learning_rate": 1.91075808762885e-05,
+      "loss": 2.4352,
+      "step": 16021
+    },
+    {
+      "epoch": 1.2930352675328867,
+      "grad_norm": 0.6904106736183167,
+      "learning_rate": 1.9098300562505266e-05,
+      "loss": 2.3734,
+      "step": 16022
+    },
+    {
+      "epoch": 1.2931159712694698,
+      "grad_norm": 0.6936709880828857,
+      "learning_rate": 1.9089022265003863e-05,
+      "loss": 2.4356,
+      "step": 16023
+    },
+    {
+      "epoch": 1.2931966750060528,
+      "grad_norm": 0.6753442883491516,
+      "learning_rate": 1.9079745984015528e-05,
+      "loss": 2.4713,
+      "step": 16024
+    },
+    {
+      "epoch": 1.2932773787426357,
+      "grad_norm": 0.7185340523719788,
+      "learning_rate": 1.9070471719771445e-05,
+      "loss": 2.4021,
+      "step": 16025
+    },
+    {
+      "epoch": 1.2933580824792188,
+      "grad_norm": 0.7486871480941772,
+      "learning_rate": 1.9061199472502798e-05,
+      "loss": 2.4144,
+      "step": 16026
+    },
+    {
+      "epoch": 1.2934387862158019,
+      "grad_norm": 0.6790735721588135,
+      "learning_rate": 1.90519292424406e-05,
+      "loss": 2.413,
+      "step": 16027
+    },
+    {
+      "epoch": 1.2935194899523847,
+      "grad_norm": 0.7104402780532837,
+      "learning_rate": 1.9042661029815922e-05,
+      "loss": 2.452,
+      "step": 16028
+    },
+    {
+      "epoch": 1.2936001936889678,
+      "grad_norm": 0.6975364685058594,
+      "learning_rate": 1.9033394834859796e-05,
+      "loss": 2.4169,
+      "step": 16029
+    },
+    {
+      "epoch": 1.293680897425551,
+      "grad_norm": 0.7619667649269104,
+      "learning_rate": 1.9024130657803085e-05,
+      "loss": 2.4106,
+      "step": 16030
+    },
+    {
+      "epoch": 1.2937616011621338,
+      "grad_norm": 0.6600254774093628,
+      "learning_rate": 1.9014868498876716e-05,
+      "loss": 2.3955,
+      "step": 16031
+    },
+    {
+      "epoch": 1.2938423048987169,
+      "grad_norm": 0.6790784597396851,
+      "learning_rate": 1.9005608358311533e-05,
+      "loss": 2.437,
+      "step": 16032
+    },
+    {
+      "epoch": 1.2939230086352997,
+      "grad_norm": 0.7085568308830261,
+      "learning_rate": 1.899635023633828e-05,
+      "loss": 2.4729,
+      "step": 16033
+    },
+    {
+      "epoch": 1.2940037123718828,
+      "grad_norm": 0.6940603256225586,
+      "learning_rate": 1.8987094133187732e-05,
+      "loss": 2.4099,
+      "step": 16034
+    },
+    {
+      "epoch": 1.2940844161084657,
+      "grad_norm": 0.7387171387672424,
+      "learning_rate": 1.897784004909058e-05,
+      "loss": 2.4509,
+      "step": 16035
+    },
+    {
+      "epoch": 1.2941651198450488,
+      "grad_norm": 0.8263981938362122,
+      "learning_rate": 1.8968587984277463e-05,
+      "loss": 2.4208,
+      "step": 16036
+    },
+    {
+      "epoch": 1.2942458235816319,
+      "grad_norm": 0.7393552660942078,
+      "learning_rate": 1.8959337938978937e-05,
+      "loss": 2.4458,
+      "step": 16037
+    },
+    {
+      "epoch": 1.2943265273182147,
+      "grad_norm": 0.652787983417511,
+      "learning_rate": 1.895008991342555e-05,
+      "loss": 2.3593,
+      "step": 16038
+    },
+    {
+      "epoch": 1.2944072310547978,
+      "grad_norm": 0.6533015370368958,
+      "learning_rate": 1.8940843907847817e-05,
+      "loss": 2.4538,
+      "step": 16039
+    },
+    {
+      "epoch": 1.294487934791381,
+      "grad_norm": 0.6723785400390625,
+      "learning_rate": 1.8931599922476106e-05,
+      "loss": 2.4528,
+      "step": 16040
+    },
+    {
+      "epoch": 1.2945686385279638,
+      "grad_norm": 0.693242073059082,
+      "learning_rate": 1.892235795754085e-05,
+      "loss": 2.4006,
+      "step": 16041
+    },
+    {
+      "epoch": 1.2946493422645469,
+      "grad_norm": 0.6849604845046997,
+      "learning_rate": 1.8913118013272403e-05,
+      "loss": 2.3758,
+      "step": 16042
+    },
+    {
+      "epoch": 1.29473004600113,
+      "grad_norm": 0.7252739667892456,
+      "learning_rate": 1.8903880089900983e-05,
+      "loss": 2.4101,
+      "step": 16043
+    },
+    {
+      "epoch": 1.2948107497377128,
+      "grad_norm": 0.720431923866272,
+      "learning_rate": 1.8894644187656864e-05,
+      "loss": 2.4241,
+      "step": 16044
+    },
+    {
+      "epoch": 1.294891453474296,
+      "grad_norm": 0.6936169862747192,
+      "learning_rate": 1.8885410306770225e-05,
+      "loss": 2.4225,
+      "step": 16045
+    },
+    {
+      "epoch": 1.294972157210879,
+      "grad_norm": 0.7698646187782288,
+      "learning_rate": 1.8876178447471193e-05,
+      "loss": 2.4031,
+      "step": 16046
+    },
+    {
+      "epoch": 1.2950528609474619,
+      "grad_norm": 0.6800495982170105,
+      "learning_rate": 1.8866948609989854e-05,
+      "loss": 2.3679,
+      "step": 16047
+    },
+    {
+      "epoch": 1.295133564684045,
+      "grad_norm": 0.7348111867904663,
+      "learning_rate": 1.8857720794556267e-05,
+      "loss": 2.4263,
+      "step": 16048
+    },
+    {
+      "epoch": 1.2952142684206278,
+      "grad_norm": 0.6614782214164734,
+      "learning_rate": 1.8848495001400356e-05,
+      "loss": 2.4396,
+      "step": 16049
+    },
+    {
+      "epoch": 1.295294972157211,
+      "grad_norm": 0.6683650612831116,
+      "learning_rate": 1.8839271230752075e-05,
+      "loss": 2.4189,
+      "step": 16050
+    },
+    {
+      "epoch": 1.2953756758937938,
+      "grad_norm": 0.711040198802948,
+      "learning_rate": 1.8830049482841328e-05,
+      "loss": 2.3974,
+      "step": 16051
+    },
+    {
+      "epoch": 1.2954563796303769,
+      "grad_norm": 0.6663193702697754,
+      "learning_rate": 1.882082975789795e-05,
+      "loss": 2.4196,
+      "step": 16052
+    },
+    {
+      "epoch": 1.29553708336696,
+      "grad_norm": 0.6551210284233093,
+      "learning_rate": 1.881161205615166e-05,
+      "loss": 2.3793,
+      "step": 16053
+    },
+    {
+      "epoch": 1.2956177871035428,
+      "grad_norm": 0.6849039793014526,
+      "learning_rate": 1.8802396377832243e-05,
+      "loss": 2.3941,
+      "step": 16054
+    },
+    {
+      "epoch": 1.295698490840126,
+      "grad_norm": 0.7642949223518372,
+      "learning_rate": 1.8793182723169357e-05,
+      "loss": 2.4296,
+      "step": 16055
+    },
+    {
+      "epoch": 1.295779194576709,
+      "grad_norm": 0.7104716897010803,
+      "learning_rate": 1.878397109239263e-05,
+      "loss": 2.4124,
+      "step": 16056
+    },
+    {
+      "epoch": 1.2958598983132918,
+      "grad_norm": 0.6822344064712524,
+      "learning_rate": 1.877476148573164e-05,
+      "loss": 2.4072,
+      "step": 16057
+    },
+    {
+      "epoch": 1.295940602049875,
+      "grad_norm": 0.6824066042900085,
+      "learning_rate": 1.8765553903415956e-05,
+      "loss": 2.4137,
+      "step": 16058
+    },
+    {
+      "epoch": 1.296021305786458,
+      "grad_norm": 0.7083307504653931,
+      "learning_rate": 1.875634834567498e-05,
+      "loss": 2.4423,
+      "step": 16059
+    },
+    {
+      "epoch": 1.2961020095230409,
+      "grad_norm": 0.7301077246665955,
+      "learning_rate": 1.874714481273818e-05,
+      "loss": 2.3926,
+      "step": 16060
+    },
+    {
+      "epoch": 1.296182713259624,
+      "grad_norm": 0.685656726360321,
+      "learning_rate": 1.873794330483496e-05,
+      "loss": 2.4409,
+      "step": 16061
+    },
+    {
+      "epoch": 1.296263416996207,
+      "grad_norm": 0.6916719675064087,
+      "learning_rate": 1.8728743822194584e-05,
+      "loss": 2.4141,
+      "step": 16062
+    },
+    {
+      "epoch": 1.29634412073279,
+      "grad_norm": 0.7188845276832581,
+      "learning_rate": 1.871954636504636e-05,
+      "loss": 2.4186,
+      "step": 16063
+    },
+    {
+      "epoch": 1.2964248244693728,
+      "grad_norm": 0.6637440919876099,
+      "learning_rate": 1.8710350933619504e-05,
+      "loss": 2.4526,
+      "step": 16064
+    },
+    {
+      "epoch": 1.2965055282059559,
+      "grad_norm": 0.7000349760055542,
+      "learning_rate": 1.87011575281432e-05,
+      "loss": 2.4096,
+      "step": 16065
+    },
+    {
+      "epoch": 1.296586231942539,
+      "grad_norm": 0.693513810634613,
+      "learning_rate": 1.8691966148846573e-05,
+      "loss": 2.3931,
+      "step": 16066
+    },
+    {
+      "epoch": 1.2966669356791218,
+      "grad_norm": 0.6928985118865967,
+      "learning_rate": 1.8682776795958678e-05,
+      "loss": 2.4384,
+      "step": 16067
+    },
+    {
+      "epoch": 1.296747639415705,
+      "grad_norm": 0.6474096179008484,
+      "learning_rate": 1.8673589469708585e-05,
+      "loss": 2.3985,
+      "step": 16068
+    },
+    {
+      "epoch": 1.296828343152288,
+      "grad_norm": 0.6827313899993896,
+      "learning_rate": 1.866440417032521e-05,
+      "loss": 2.4607,
+      "step": 16069
+    },
+    {
+      "epoch": 1.2969090468888709,
+      "grad_norm": 0.7183445692062378,
+      "learning_rate": 1.8655220898037485e-05,
+      "loss": 2.4396,
+      "step": 16070
+    },
+    {
+      "epoch": 1.296989750625454,
+      "grad_norm": 0.6997376680374146,
+      "learning_rate": 1.8646039653074333e-05,
+      "loss": 2.4627,
+      "step": 16071
+    },
+    {
+      "epoch": 1.297070454362037,
+      "grad_norm": 0.7358444333076477,
+      "learning_rate": 1.8636860435664493e-05,
+      "loss": 2.4165,
+      "step": 16072
+    },
+    {
+      "epoch": 1.29715115809862,
+      "grad_norm": 0.8126270771026611,
+      "learning_rate": 1.8627683246036787e-05,
+      "loss": 2.4681,
+      "step": 16073
+    },
+    {
+      "epoch": 1.297231861835203,
+      "grad_norm": 0.7364177107810974,
+      "learning_rate": 1.8618508084419918e-05,
+      "loss": 2.44,
+      "step": 16074
+    },
+    {
+      "epoch": 1.297312565571786,
+      "grad_norm": 0.7480010390281677,
+      "learning_rate": 1.8609334951042567e-05,
+      "loss": 2.4759,
+      "step": 16075
+    },
+    {
+      "epoch": 1.297393269308369,
+      "grad_norm": 0.6563693284988403,
+      "learning_rate": 1.8600163846133335e-05,
+      "loss": 2.3865,
+      "step": 16076
+    },
+    {
+      "epoch": 1.297473973044952,
+      "grad_norm": 0.6961230039596558,
+      "learning_rate": 1.8590994769920832e-05,
+      "loss": 2.3851,
+      "step": 16077
+    },
+    {
+      "epoch": 1.297554676781535,
+      "grad_norm": 0.7137415409088135,
+      "learning_rate": 1.8581827722633527e-05,
+      "loss": 2.4115,
+      "step": 16078
+    },
+    {
+      "epoch": 1.297635380518118,
+      "grad_norm": 0.6579335331916809,
+      "learning_rate": 1.85726627044999e-05,
+      "loss": 2.4464,
+      "step": 16079
+    },
+    {
+      "epoch": 1.2977160842547009,
+      "grad_norm": 0.7069905400276184,
+      "learning_rate": 1.8563499715748366e-05,
+      "loss": 2.4057,
+      "step": 16080
+    },
+    {
+      "epoch": 1.297796787991284,
+      "grad_norm": 0.771925687789917,
+      "learning_rate": 1.8554338756607325e-05,
+      "loss": 2.4696,
+      "step": 16081
+    },
+    {
+      "epoch": 1.297877491727867,
+      "grad_norm": 0.7268456816673279,
+      "learning_rate": 1.8545179827305048e-05,
+      "loss": 2.3949,
+      "step": 16082
+    },
+    {
+      "epoch": 1.29795819546445,
+      "grad_norm": 0.7049130797386169,
+      "learning_rate": 1.8536022928069796e-05,
+      "loss": 2.4448,
+      "step": 16083
+    },
+    {
+      "epoch": 1.298038899201033,
+      "grad_norm": 0.6716888546943665,
+      "learning_rate": 1.852686805912982e-05,
+      "loss": 2.3356,
+      "step": 16084
+    },
+    {
+      "epoch": 1.298119602937616,
+      "grad_norm": 0.666386604309082,
+      "learning_rate": 1.851771522071325e-05,
+      "loss": 2.4226,
+      "step": 16085
+    },
+    {
+      "epoch": 1.298200306674199,
+      "grad_norm": 0.7084901332855225,
+      "learning_rate": 1.8508564413048223e-05,
+      "loss": 2.4452,
+      "step": 16086
+    },
+    {
+      "epoch": 1.298281010410782,
+      "grad_norm": 0.6615412831306458,
+      "learning_rate": 1.8499415636362815e-05,
+      "loss": 2.4193,
+      "step": 16087
+    },
+    {
+      "epoch": 1.2983617141473651,
+      "grad_norm": 0.7143606543540955,
+      "learning_rate": 1.849026889088499e-05,
+      "loss": 2.4513,
+      "step": 16088
+    },
+    {
+      "epoch": 1.298442417883948,
+      "grad_norm": 0.7241482734680176,
+      "learning_rate": 1.8481124176842723e-05,
+      "loss": 2.458,
+      "step": 16089
+    },
+    {
+      "epoch": 1.298523121620531,
+      "grad_norm": 0.6762149930000305,
+      "learning_rate": 1.8471981494463963e-05,
+      "loss": 2.4386,
+      "step": 16090
+    },
+    {
+      "epoch": 1.2986038253571142,
+      "grad_norm": 0.6672768592834473,
+      "learning_rate": 1.8462840843976525e-05,
+      "loss": 2.375,
+      "step": 16091
+    },
+    {
+      "epoch": 1.298684529093697,
+      "grad_norm": 0.6871693134307861,
+      "learning_rate": 1.8453702225608226e-05,
+      "loss": 2.4342,
+      "step": 16092
+    },
+    {
+      "epoch": 1.2987652328302801,
+      "grad_norm": 0.6771275401115417,
+      "learning_rate": 1.8444565639586864e-05,
+      "loss": 2.402,
+      "step": 16093
+    },
+    {
+      "epoch": 1.298845936566863,
+      "grad_norm": 0.6627403497695923,
+      "learning_rate": 1.8435431086140077e-05,
+      "loss": 2.4667,
+      "step": 16094
+    },
+    {
+      "epoch": 1.298926640303446,
+      "grad_norm": 0.7001610398292542,
+      "learning_rate": 1.8426298565495538e-05,
+      "loss": 2.4396,
+      "step": 16095
+    },
+    {
+      "epoch": 1.299007344040029,
+      "grad_norm": 0.7574489712715149,
+      "learning_rate": 1.8417168077880908e-05,
+      "loss": 2.4601,
+      "step": 16096
+    },
+    {
+      "epoch": 1.299088047776612,
+      "grad_norm": 0.7771055698394775,
+      "learning_rate": 1.840803962352372e-05,
+      "loss": 2.4371,
+      "step": 16097
+    },
+    {
+      "epoch": 1.299168751513195,
+      "grad_norm": 0.6738649606704712,
+      "learning_rate": 1.8398913202651457e-05,
+      "loss": 2.3921,
+      "step": 16098
+    },
+    {
+      "epoch": 1.299249455249778,
+      "grad_norm": 0.7014862895011902,
+      "learning_rate": 1.8389788815491583e-05,
+      "loss": 2.451,
+      "step": 16099
+    },
+    {
+      "epoch": 1.299330158986361,
+      "grad_norm": 0.7026070952415466,
+      "learning_rate": 1.8380666462271523e-05,
+      "loss": 2.4583,
+      "step": 16100
+    },
+    {
+      "epoch": 1.2994108627229441,
+      "grad_norm": 0.6904535293579102,
+      "learning_rate": 1.8371546143218588e-05,
+      "loss": 2.4453,
+      "step": 16101
+    },
+    {
+      "epoch": 1.299491566459527,
+      "grad_norm": 0.6974804997444153,
+      "learning_rate": 1.8362427858560093e-05,
+      "loss": 2.4291,
+      "step": 16102
+    },
+    {
+      "epoch": 1.29957227019611,
+      "grad_norm": 0.6826989650726318,
+      "learning_rate": 1.8353311608523326e-05,
+      "loss": 2.4183,
+      "step": 16103
+    },
+    {
+      "epoch": 1.2996529739326932,
+      "grad_norm": 0.6804787516593933,
+      "learning_rate": 1.8344197393335448e-05,
+      "loss": 2.434,
+      "step": 16104
+    },
+    {
+      "epoch": 1.299733677669276,
+      "grad_norm": 0.7144587635993958,
+      "learning_rate": 1.8335085213223613e-05,
+      "loss": 2.4296,
+      "step": 16105
+    },
+    {
+      "epoch": 1.2998143814058591,
+      "grad_norm": 0.7228755354881287,
+      "learning_rate": 1.8325975068414924e-05,
+      "loss": 2.3987,
+      "step": 16106
+    },
+    {
+      "epoch": 1.2998950851424422,
+      "grad_norm": 0.7417716383934021,
+      "learning_rate": 1.8316866959136438e-05,
+      "loss": 2.4076,
+      "step": 16107
+    },
+    {
+      "epoch": 1.299975788879025,
+      "grad_norm": 0.6737387776374817,
+      "learning_rate": 1.8307760885615154e-05,
+      "loss": 2.4175,
+      "step": 16108
+    },
+    {
+      "epoch": 1.3000564926156082,
+      "grad_norm": 0.7294918298721313,
+      "learning_rate": 1.8298656848078035e-05,
+      "loss": 2.4022,
+      "step": 16109
+    },
+    {
+      "epoch": 1.300137196352191,
+      "grad_norm": 0.7200861573219299,
+      "learning_rate": 1.828955484675193e-05,
+      "loss": 2.4018,
+      "step": 16110
+    },
+    {
+      "epoch": 1.3002179000887741,
+      "grad_norm": 0.7704176306724548,
+      "learning_rate": 1.8280454881863718e-05,
+      "loss": 2.4539,
+      "step": 16111
+    },
+    {
+      "epoch": 1.300298603825357,
+      "grad_norm": 0.6790730953216553,
+      "learning_rate": 1.8271356953640184e-05,
+      "loss": 2.4196,
+      "step": 16112
+    },
+    {
+      "epoch": 1.30037930756194,
+      "grad_norm": 0.7165740132331848,
+      "learning_rate": 1.8262261062308096e-05,
+      "loss": 2.4234,
+      "step": 16113
+    },
+    {
+      "epoch": 1.3004600112985232,
+      "grad_norm": 0.7716830372810364,
+      "learning_rate": 1.82531672080941e-05,
+      "loss": 2.4255,
+      "step": 16114
+    },
+    {
+      "epoch": 1.300540715035106,
+      "grad_norm": 0.6525317430496216,
+      "learning_rate": 1.824407539122488e-05,
+      "loss": 2.4482,
+      "step": 16115
+    },
+    {
+      "epoch": 1.3006214187716891,
+      "grad_norm": 0.7397769093513489,
+      "learning_rate": 1.8234985611927003e-05,
+      "loss": 2.33,
+      "step": 16116
+    },
+    {
+      "epoch": 1.3007021225082722,
+      "grad_norm": 0.7106032967567444,
+      "learning_rate": 1.822589787042702e-05,
+      "loss": 2.485,
+      "step": 16117
+    },
+    {
+      "epoch": 1.300782826244855,
+      "grad_norm": 0.7030045390129089,
+      "learning_rate": 1.8216812166951425e-05,
+      "loss": 2.454,
+      "step": 16118
+    },
+    {
+      "epoch": 1.3008635299814382,
+      "grad_norm": 0.7075662612915039,
+      "learning_rate": 1.8207728501726683e-05,
+      "loss": 2.4589,
+      "step": 16119
+    },
+    {
+      "epoch": 1.3009442337180213,
+      "grad_norm": 0.6700533032417297,
+      "learning_rate": 1.819864687497912e-05,
+      "loss": 2.4398,
+      "step": 16120
+    },
+    {
+      "epoch": 1.3010249374546041,
+      "grad_norm": 0.6951712369918823,
+      "learning_rate": 1.8189567286935117e-05,
+      "loss": 2.3998,
+      "step": 16121
+    },
+    {
+      "epoch": 1.3011056411911872,
+      "grad_norm": 0.708344578742981,
+      "learning_rate": 1.818048973782097e-05,
+      "loss": 2.4142,
+      "step": 16122
+    },
+    {
+      "epoch": 1.30118634492777,
+      "grad_norm": 0.7078592777252197,
+      "learning_rate": 1.817141422786287e-05,
+      "loss": 2.451,
+      "step": 16123
+    },
+    {
+      "epoch": 1.3012670486643532,
+      "grad_norm": 0.7111849784851074,
+      "learning_rate": 1.816234075728703e-05,
+      "loss": 2.4762,
+      "step": 16124
+    },
+    {
+      "epoch": 1.301347752400936,
+      "grad_norm": 0.6716348528862,
+      "learning_rate": 1.8153269326319588e-05,
+      "loss": 2.4373,
+      "step": 16125
+    },
+    {
+      "epoch": 1.3014284561375191,
+      "grad_norm": 0.6592512130737305,
+      "learning_rate": 1.8144199935186623e-05,
+      "loss": 2.412,
+      "step": 16126
+    },
+    {
+      "epoch": 1.3015091598741022,
+      "grad_norm": 0.6958334445953369,
+      "learning_rate": 1.8135132584114167e-05,
+      "loss": 2.4077,
+      "step": 16127
+    },
+    {
+      "epoch": 1.301589863610685,
+      "grad_norm": 0.6911341547966003,
+      "learning_rate": 1.8126067273328207e-05,
+      "loss": 2.409,
+      "step": 16128
+    },
+    {
+      "epoch": 1.3016705673472682,
+      "grad_norm": 0.676114022731781,
+      "learning_rate": 1.8117004003054693e-05,
+      "loss": 2.4463,
+      "step": 16129
+    },
+    {
+      "epoch": 1.3017512710838512,
+      "grad_norm": 0.6493322849273682,
+      "learning_rate": 1.810794277351947e-05,
+      "loss": 2.4377,
+      "step": 16130
+    },
+    {
+      "epoch": 1.3018319748204341,
+      "grad_norm": 0.6938454508781433,
+      "learning_rate": 1.8098883584948367e-05,
+      "loss": 2.4298,
+      "step": 16131
+    },
+    {
+      "epoch": 1.3019126785570172,
+      "grad_norm": 0.69407719373703,
+      "learning_rate": 1.8089826437567214e-05,
+      "loss": 2.4107,
+      "step": 16132
+    },
+    {
+      "epoch": 1.3019933822936003,
+      "grad_norm": 0.6898862719535828,
+      "learning_rate": 1.8080771331601664e-05,
+      "loss": 2.4182,
+      "step": 16133
+    },
+    {
+      "epoch": 1.3020740860301832,
+      "grad_norm": 0.7377758026123047,
+      "learning_rate": 1.807171826727744e-05,
+      "loss": 2.4112,
+      "step": 16134
+    },
+    {
+      "epoch": 1.3021547897667662,
+      "grad_norm": 0.674057126045227,
+      "learning_rate": 1.8062667244820154e-05,
+      "loss": 2.4276,
+      "step": 16135
+    },
+    {
+      "epoch": 1.3022354935033493,
+      "grad_norm": 0.7087522745132446,
+      "learning_rate": 1.8053618264455384e-05,
+      "loss": 2.4338,
+      "step": 16136
+    },
+    {
+      "epoch": 1.3023161972399322,
+      "grad_norm": 0.70958411693573,
+      "learning_rate": 1.8044571326408667e-05,
+      "loss": 2.4369,
+      "step": 16137
+    },
+    {
+      "epoch": 1.3023969009765153,
+      "grad_norm": 0.7023837566375732,
+      "learning_rate": 1.803552643090548e-05,
+      "loss": 2.4185,
+      "step": 16138
+    },
+    {
+      "epoch": 1.3024776047130981,
+      "grad_norm": 0.708543598651886,
+      "learning_rate": 1.8026483578171216e-05,
+      "loss": 2.4053,
+      "step": 16139
+    },
+    {
+      "epoch": 1.3025583084496812,
+      "grad_norm": 0.748601496219635,
+      "learning_rate": 1.8017442768431257e-05,
+      "loss": 2.3948,
+      "step": 16140
+    },
+    {
+      "epoch": 1.302639012186264,
+      "grad_norm": 0.6626949310302734,
+      "learning_rate": 1.800840400191096e-05,
+      "loss": 2.4636,
+      "step": 16141
+    },
+    {
+      "epoch": 1.3027197159228472,
+      "grad_norm": 0.7079617977142334,
+      "learning_rate": 1.7999367278835534e-05,
+      "loss": 2.4091,
+      "step": 16142
+    },
+    {
+      "epoch": 1.3028004196594303,
+      "grad_norm": 0.7025624513626099,
+      "learning_rate": 1.7990332599430225e-05,
+      "loss": 2.3732,
+      "step": 16143
+    },
+    {
+      "epoch": 1.3028811233960131,
+      "grad_norm": 0.7365758419036865,
+      "learning_rate": 1.7981299963920205e-05,
+      "loss": 2.4725,
+      "step": 16144
+    },
+    {
+      "epoch": 1.3029618271325962,
+      "grad_norm": 0.7511963248252869,
+      "learning_rate": 1.7972269372530615e-05,
+      "loss": 2.4304,
+      "step": 16145
+    },
+    {
+      "epoch": 1.3030425308691793,
+      "grad_norm": 0.7055985331535339,
+      "learning_rate": 1.796324082548644e-05,
+      "loss": 2.4259,
+      "step": 16146
+    },
+    {
+      "epoch": 1.3031232346057622,
+      "grad_norm": 0.691162645816803,
+      "learning_rate": 1.7954214323012775e-05,
+      "loss": 2.4262,
+      "step": 16147
+    },
+    {
+      "epoch": 1.3032039383423453,
+      "grad_norm": 0.7179710268974304,
+      "learning_rate": 1.7945189865334587e-05,
+      "loss": 2.4301,
+      "step": 16148
+    },
+    {
+      "epoch": 1.3032846420789284,
+      "grad_norm": 0.7391623258590698,
+      "learning_rate": 1.7936167452676744e-05,
+      "loss": 2.4302,
+      "step": 16149
+    },
+    {
+      "epoch": 1.3033653458155112,
+      "grad_norm": 0.7297981381416321,
+      "learning_rate": 1.7927147085264117e-05,
+      "loss": 2.3911,
+      "step": 16150
+    },
+    {
+      "epoch": 1.3034460495520943,
+      "grad_norm": 0.7571932673454285,
+      "learning_rate": 1.7918128763321552e-05,
+      "loss": 2.4348,
+      "step": 16151
+    },
+    {
+      "epoch": 1.3035267532886774,
+      "grad_norm": 0.7074765563011169,
+      "learning_rate": 1.7909112487073754e-05,
+      "loss": 2.4164,
+      "step": 16152
+    },
+    {
+      "epoch": 1.3036074570252603,
+      "grad_norm": 0.7534131407737732,
+      "learning_rate": 1.7900098256745467e-05,
+      "loss": 2.3784,
+      "step": 16153
+    },
+    {
+      "epoch": 1.3036881607618434,
+      "grad_norm": 0.675398588180542,
+      "learning_rate": 1.789108607256136e-05,
+      "loss": 2.4305,
+      "step": 16154
+    },
+    {
+      "epoch": 1.3037688644984262,
+      "grad_norm": 0.7099249362945557,
+      "learning_rate": 1.7882075934746002e-05,
+      "loss": 2.4053,
+      "step": 16155
+    },
+    {
+      "epoch": 1.3038495682350093,
+      "grad_norm": 0.6914681196212769,
+      "learning_rate": 1.787306784352397e-05,
+      "loss": 2.3902,
+      "step": 16156
+    },
+    {
+      "epoch": 1.3039302719715922,
+      "grad_norm": 0.6956958770751953,
+      "learning_rate": 1.786406179911977e-05,
+      "loss": 2.4026,
+      "step": 16157
+    },
+    {
+      "epoch": 1.3040109757081753,
+      "grad_norm": 0.6873000860214233,
+      "learning_rate": 1.7855057801757857e-05,
+      "loss": 2.4082,
+      "step": 16158
+    },
+    {
+      "epoch": 1.3040916794447583,
+      "grad_norm": 0.7340587377548218,
+      "learning_rate": 1.7846055851662625e-05,
+      "loss": 2.4894,
+      "step": 16159
+    },
+    {
+      "epoch": 1.3041723831813412,
+      "grad_norm": 0.6956963539123535,
+      "learning_rate": 1.7837055949058444e-05,
+      "loss": 2.3976,
+      "step": 16160
+    },
+    {
+      "epoch": 1.3042530869179243,
+      "grad_norm": 0.7654300332069397,
+      "learning_rate": 1.782805809416962e-05,
+      "loss": 2.4272,
+      "step": 16161
+    },
+    {
+      "epoch": 1.3043337906545074,
+      "grad_norm": 0.7735971212387085,
+      "learning_rate": 1.7819062287220368e-05,
+      "loss": 2.4513,
+      "step": 16162
+    },
+    {
+      "epoch": 1.3044144943910903,
+      "grad_norm": 0.6897203326225281,
+      "learning_rate": 1.7810068528434908e-05,
+      "loss": 2.3974,
+      "step": 16163
+    },
+    {
+      "epoch": 1.3044951981276733,
+      "grad_norm": 0.7328432202339172,
+      "learning_rate": 1.780107681803741e-05,
+      "loss": 2.4455,
+      "step": 16164
+    },
+    {
+      "epoch": 1.3045759018642564,
+      "grad_norm": 0.7098489999771118,
+      "learning_rate": 1.7792087156251924e-05,
+      "loss": 2.4173,
+      "step": 16165
+    },
+    {
+      "epoch": 1.3046566056008393,
+      "grad_norm": 0.6593194007873535,
+      "learning_rate": 1.7783099543302518e-05,
+      "loss": 2.4102,
+      "step": 16166
+    },
+    {
+      "epoch": 1.3047373093374224,
+      "grad_norm": 0.7329291105270386,
+      "learning_rate": 1.7774113979413188e-05,
+      "loss": 2.4856,
+      "step": 16167
+    },
+    {
+      "epoch": 1.3048180130740052,
+      "grad_norm": 0.7033355236053467,
+      "learning_rate": 1.776513046480788e-05,
+      "loss": 2.4503,
+      "step": 16168
+    },
+    {
+      "epoch": 1.3048987168105883,
+      "grad_norm": 0.7063608765602112,
+      "learning_rate": 1.7756148999710486e-05,
+      "loss": 2.4523,
+      "step": 16169
+    },
+    {
+      "epoch": 1.3049794205471712,
+      "grad_norm": 0.6905883550643921,
+      "learning_rate": 1.774716958434487e-05,
+      "loss": 2.4149,
+      "step": 16170
+    },
+    {
+      "epoch": 1.3050601242837543,
+      "grad_norm": 0.694551408290863,
+      "learning_rate": 1.7738192218934778e-05,
+      "loss": 2.437,
+      "step": 16171
+    },
+    {
+      "epoch": 1.3051408280203374,
+      "grad_norm": 0.7173176407814026,
+      "learning_rate": 1.772921690370396e-05,
+      "loss": 2.4817,
+      "step": 16172
+    },
+    {
+      "epoch": 1.3052215317569202,
+      "grad_norm": 0.7197130918502808,
+      "learning_rate": 1.7720243638876153e-05,
+      "loss": 2.4481,
+      "step": 16173
+    },
+    {
+      "epoch": 1.3053022354935033,
+      "grad_norm": 0.710811197757721,
+      "learning_rate": 1.771127242467493e-05,
+      "loss": 2.397,
+      "step": 16174
+    },
+    {
+      "epoch": 1.3053829392300864,
+      "grad_norm": 0.9194550514221191,
+      "learning_rate": 1.7702303261323894e-05,
+      "loss": 2.5206,
+      "step": 16175
+    },
+    {
+      "epoch": 1.3054636429666693,
+      "grad_norm": 0.7003832459449768,
+      "learning_rate": 1.769333614904659e-05,
+      "loss": 2.4175,
+      "step": 16176
+    },
+    {
+      "epoch": 1.3055443467032524,
+      "grad_norm": 0.7161554098129272,
+      "learning_rate": 1.768437108806651e-05,
+      "loss": 2.3892,
+      "step": 16177
+    },
+    {
+      "epoch": 1.3056250504398355,
+      "grad_norm": 0.6516181826591492,
+      "learning_rate": 1.767540807860707e-05,
+      "loss": 2.4361,
+      "step": 16178
+    },
+    {
+      "epoch": 1.3057057541764183,
+      "grad_norm": 0.7518061399459839,
+      "learning_rate": 1.7666447120891662e-05,
+      "loss": 2.4572,
+      "step": 16179
+    },
+    {
+      "epoch": 1.3057864579130014,
+      "grad_norm": 0.735388994216919,
+      "learning_rate": 1.7657488215143637e-05,
+      "loss": 2.3965,
+      "step": 16180
+    },
+    {
+      "epoch": 1.3058671616495845,
+      "grad_norm": 0.6994282007217407,
+      "learning_rate": 1.764853136158622e-05,
+      "loss": 2.4052,
+      "step": 16181
+    },
+    {
+      "epoch": 1.3059478653861674,
+      "grad_norm": 0.7095311880111694,
+      "learning_rate": 1.7639576560442684e-05,
+      "loss": 2.4818,
+      "step": 16182
+    },
+    {
+      "epoch": 1.3060285691227504,
+      "grad_norm": 0.6527207493782043,
+      "learning_rate": 1.7630623811936208e-05,
+      "loss": 2.3962,
+      "step": 16183
+    },
+    {
+      "epoch": 1.3061092728593333,
+      "grad_norm": 0.6668451428413391,
+      "learning_rate": 1.7621673116289882e-05,
+      "loss": 2.4514,
+      "step": 16184
+    },
+    {
+      "epoch": 1.3061899765959164,
+      "grad_norm": 0.7119911909103394,
+      "learning_rate": 1.7612724473726795e-05,
+      "loss": 2.4313,
+      "step": 16185
+    },
+    {
+      "epoch": 1.3062706803324993,
+      "grad_norm": 0.706249475479126,
+      "learning_rate": 1.7603777884469984e-05,
+      "loss": 2.4131,
+      "step": 16186
+    },
+    {
+      "epoch": 1.3063513840690824,
+      "grad_norm": 0.6634086966514587,
+      "learning_rate": 1.759483334874241e-05,
+      "loss": 2.3532,
+      "step": 16187
+    },
+    {
+      "epoch": 1.3064320878056654,
+      "grad_norm": 0.8096393942832947,
+      "learning_rate": 1.7585890866766995e-05,
+      "loss": 2.4485,
+      "step": 16188
+    },
+    {
+      "epoch": 1.3065127915422483,
+      "grad_norm": 0.675308883190155,
+      "learning_rate": 1.7576950438766615e-05,
+      "loss": 2.388,
+      "step": 16189
+    },
+    {
+      "epoch": 1.3065934952788314,
+      "grad_norm": 0.738275408744812,
+      "learning_rate": 1.756801206496411e-05,
+      "loss": 2.4485,
+      "step": 16190
+    },
+    {
+      "epoch": 1.3066741990154145,
+      "grad_norm": 0.7045620083808899,
+      "learning_rate": 1.755907574558221e-05,
+      "loss": 2.3985,
+      "step": 16191
+    },
+    {
+      "epoch": 1.3067549027519973,
+      "grad_norm": 0.6499879360198975,
+      "learning_rate": 1.755014148084363e-05,
+      "loss": 2.3992,
+      "step": 16192
+    },
+    {
+      "epoch": 1.3068356064885804,
+      "grad_norm": 0.7101179361343384,
+      "learning_rate": 1.7541209270971083e-05,
+      "loss": 2.4217,
+      "step": 16193
+    },
+    {
+      "epoch": 1.3069163102251635,
+      "grad_norm": 0.6865181922912598,
+      "learning_rate": 1.7532279116187124e-05,
+      "loss": 2.4805,
+      "step": 16194
+    },
+    {
+      "epoch": 1.3069970139617464,
+      "grad_norm": 0.7710141539573669,
+      "learning_rate": 1.752335101671434e-05,
+      "loss": 2.3654,
+      "step": 16195
+    },
+    {
+      "epoch": 1.3070777176983295,
+      "grad_norm": 0.695936381816864,
+      "learning_rate": 1.7514424972775244e-05,
+      "loss": 2.4315,
+      "step": 16196
+    },
+    {
+      "epoch": 1.3071584214349126,
+      "grad_norm": 0.6781535148620605,
+      "learning_rate": 1.7505500984592304e-05,
+      "loss": 2.4238,
+      "step": 16197
+    },
+    {
+      "epoch": 1.3072391251714954,
+      "grad_norm": 0.6549252271652222,
+      "learning_rate": 1.7496579052387918e-05,
+      "loss": 2.3766,
+      "step": 16198
+    },
+    {
+      "epoch": 1.3073198289080785,
+      "grad_norm": 0.6599059700965881,
+      "learning_rate": 1.7487659176384474e-05,
+      "loss": 2.4613,
+      "step": 16199
+    },
+    {
+      "epoch": 1.3074005326446614,
+      "grad_norm": 0.6742514967918396,
+      "learning_rate": 1.7478741356804228e-05,
+      "loss": 2.3917,
+      "step": 16200
+    },
+    {
+      "epoch": 1.3074812363812445,
+      "grad_norm": 0.6542397141456604,
+      "learning_rate": 1.746982559386946e-05,
+      "loss": 2.44,
+      "step": 16201
+    },
+    {
+      "epoch": 1.3075619401178273,
+      "grad_norm": 0.7200478315353394,
+      "learning_rate": 1.74609118878024e-05,
+      "loss": 2.4324,
+      "step": 16202
+    },
+    {
+      "epoch": 1.3076426438544104,
+      "grad_norm": 0.717628002166748,
+      "learning_rate": 1.745200023882515e-05,
+      "loss": 2.3996,
+      "step": 16203
+    },
+    {
+      "epoch": 1.3077233475909935,
+      "grad_norm": 0.7350025177001953,
+      "learning_rate": 1.744309064715983e-05,
+      "loss": 2.4812,
+      "step": 16204
+    },
+    {
+      "epoch": 1.3078040513275764,
+      "grad_norm": 0.7253599762916565,
+      "learning_rate": 1.74341831130285e-05,
+      "loss": 2.4454,
+      "step": 16205
+    },
+    {
+      "epoch": 1.3078847550641595,
+      "grad_norm": 0.7537909746170044,
+      "learning_rate": 1.7425277636653193e-05,
+      "loss": 2.4247,
+      "step": 16206
+    },
+    {
+      "epoch": 1.3079654588007426,
+      "grad_norm": 0.7563284039497375,
+      "learning_rate": 1.7416374218255783e-05,
+      "loss": 2.3893,
+      "step": 16207
+    },
+    {
+      "epoch": 1.3080461625373254,
+      "grad_norm": 0.7118926048278809,
+      "learning_rate": 1.740747285805818e-05,
+      "loss": 2.4146,
+      "step": 16208
+    },
+    {
+      "epoch": 1.3081268662739085,
+      "grad_norm": 0.7805569171905518,
+      "learning_rate": 1.7398573556282304e-05,
+      "loss": 2.396,
+      "step": 16209
+    },
+    {
+      "epoch": 1.3082075700104916,
+      "grad_norm": 0.7357630133628845,
+      "learning_rate": 1.738967631314987e-05,
+      "loss": 2.5405,
+      "step": 16210
+    },
+    {
+      "epoch": 1.3082882737470745,
+      "grad_norm": 0.6670438647270203,
+      "learning_rate": 1.7380781128882652e-05,
+      "loss": 2.4452,
+      "step": 16211
+    },
+    {
+      "epoch": 1.3083689774836575,
+      "grad_norm": 0.7374427318572998,
+      "learning_rate": 1.7371888003702353e-05,
+      "loss": 2.5143,
+      "step": 16212
+    },
+    {
+      "epoch": 1.3084496812202406,
+      "grad_norm": 0.672207236289978,
+      "learning_rate": 1.736299693783058e-05,
+      "loss": 2.4178,
+      "step": 16213
+    },
+    {
+      "epoch": 1.3085303849568235,
+      "grad_norm": 0.6926576495170593,
+      "learning_rate": 1.735410793148894e-05,
+      "loss": 2.3466,
+      "step": 16214
+    },
+    {
+      "epoch": 1.3086110886934066,
+      "grad_norm": 0.6928917169570923,
+      "learning_rate": 1.734522098489899e-05,
+      "loss": 2.4654,
+      "step": 16215
+    },
+    {
+      "epoch": 1.3086917924299895,
+      "grad_norm": 0.6536242961883545,
+      "learning_rate": 1.733633609828217e-05,
+      "loss": 2.3761,
+      "step": 16216
+    },
+    {
+      "epoch": 1.3087724961665725,
+      "grad_norm": 0.6993953585624695,
+      "learning_rate": 1.732745327185994e-05,
+      "loss": 2.3963,
+      "step": 16217
+    },
+    {
+      "epoch": 1.3088531999031554,
+      "grad_norm": 0.6851957440376282,
+      "learning_rate": 1.731857250585368e-05,
+      "loss": 2.4253,
+      "step": 16218
+    },
+    {
+      "epoch": 1.3089339036397385,
+      "grad_norm": 0.6620005965232849,
+      "learning_rate": 1.7309693800484728e-05,
+      "loss": 2.4302,
+      "step": 16219
+    },
+    {
+      "epoch": 1.3090146073763216,
+      "grad_norm": 0.6704410314559937,
+      "learning_rate": 1.7300817155974356e-05,
+      "loss": 2.4065,
+      "step": 16220
+    },
+    {
+      "epoch": 1.3090953111129044,
+      "grad_norm": 0.6882327198982239,
+      "learning_rate": 1.7291942572543807e-05,
+      "loss": 2.4526,
+      "step": 16221
+    },
+    {
+      "epoch": 1.3091760148494875,
+      "grad_norm": 0.6971533298492432,
+      "learning_rate": 1.7283070050414275e-05,
+      "loss": 2.4076,
+      "step": 16222
+    },
+    {
+      "epoch": 1.3092567185860706,
+      "grad_norm": 0.6662544012069702,
+      "learning_rate": 1.7274199589806827e-05,
+      "loss": 2.3678,
+      "step": 16223
+    },
+    {
+      "epoch": 1.3093374223226535,
+      "grad_norm": 0.6342894434928894,
+      "learning_rate": 1.726533119094258e-05,
+      "loss": 2.3424,
+      "step": 16224
+    },
+    {
+      "epoch": 1.3094181260592366,
+      "grad_norm": 0.6808488965034485,
+      "learning_rate": 1.7256464854042577e-05,
+      "loss": 2.4286,
+      "step": 16225
+    },
+    {
+      "epoch": 1.3094988297958197,
+      "grad_norm": 0.6417922973632812,
+      "learning_rate": 1.7247600579327738e-05,
+      "loss": 2.3677,
+      "step": 16226
+    },
+    {
+      "epoch": 1.3095795335324025,
+      "grad_norm": 0.7267102599143982,
+      "learning_rate": 1.7238738367019002e-05,
+      "loss": 2.3974,
+      "step": 16227
+    },
+    {
+      "epoch": 1.3096602372689856,
+      "grad_norm": 0.6915002465248108,
+      "learning_rate": 1.722987821733725e-05,
+      "loss": 2.4429,
+      "step": 16228
+    },
+    {
+      "epoch": 1.3097409410055685,
+      "grad_norm": 0.6930112242698669,
+      "learning_rate": 1.7221020130503296e-05,
+      "loss": 2.4272,
+      "step": 16229
+    },
+    {
+      "epoch": 1.3098216447421516,
+      "grad_norm": 0.7049465179443359,
+      "learning_rate": 1.7212164106737904e-05,
+      "loss": 2.4089,
+      "step": 16230
+    },
+    {
+      "epoch": 1.3099023484787344,
+      "grad_norm": 0.7230044603347778,
+      "learning_rate": 1.720331014626182e-05,
+      "loss": 2.4313,
+      "step": 16231
+    },
+    {
+      "epoch": 1.3099830522153175,
+      "grad_norm": 0.6513530015945435,
+      "learning_rate": 1.7194458249295665e-05,
+      "loss": 2.3293,
+      "step": 16232
+    },
+    {
+      "epoch": 1.3100637559519006,
+      "grad_norm": 0.6880534291267395,
+      "learning_rate": 1.718560841606005e-05,
+      "loss": 2.4556,
+      "step": 16233
+    },
+    {
+      "epoch": 1.3101444596884835,
+      "grad_norm": 0.7075292468070984,
+      "learning_rate": 1.717676064677559e-05,
+      "loss": 2.4747,
+      "step": 16234
+    },
+    {
+      "epoch": 1.3102251634250666,
+      "grad_norm": 0.7713594436645508,
+      "learning_rate": 1.7167914941662723e-05,
+      "loss": 2.4135,
+      "step": 16235
+    },
+    {
+      "epoch": 1.3103058671616497,
+      "grad_norm": 0.7883979082107544,
+      "learning_rate": 1.7159071300941943e-05,
+      "loss": 2.418,
+      "step": 16236
+    },
+    {
+      "epoch": 1.3103865708982325,
+      "grad_norm": 0.6588975787162781,
+      "learning_rate": 1.7150229724833655e-05,
+      "loss": 2.3295,
+      "step": 16237
+    },
+    {
+      "epoch": 1.3104672746348156,
+      "grad_norm": 0.679086446762085,
+      "learning_rate": 1.7141390213558217e-05,
+      "loss": 2.413,
+      "step": 16238
+    },
+    {
+      "epoch": 1.3105479783713987,
+      "grad_norm": 0.6803067326545715,
+      "learning_rate": 1.713255276733592e-05,
+      "loss": 2.4338,
+      "step": 16239
+    },
+    {
+      "epoch": 1.3106286821079816,
+      "grad_norm": 0.7041650414466858,
+      "learning_rate": 1.712371738638704e-05,
+      "loss": 2.469,
+      "step": 16240
+    },
+    {
+      "epoch": 1.3107093858445646,
+      "grad_norm": 0.6560962796211243,
+      "learning_rate": 1.711488407093178e-05,
+      "loss": 2.4353,
+      "step": 16241
+    },
+    {
+      "epoch": 1.3107900895811477,
+      "grad_norm": 0.6637921333312988,
+      "learning_rate": 1.7106052821190244e-05,
+      "loss": 2.3996,
+      "step": 16242
+    },
+    {
+      "epoch": 1.3108707933177306,
+      "grad_norm": 0.8131709098815918,
+      "learning_rate": 1.7097223637382565e-05,
+      "loss": 2.466,
+      "step": 16243
+    },
+    {
+      "epoch": 1.3109514970543137,
+      "grad_norm": 0.6637253165245056,
+      "learning_rate": 1.708839651972881e-05,
+      "loss": 2.3811,
+      "step": 16244
+    },
+    {
+      "epoch": 1.3110322007908966,
+      "grad_norm": 0.71912682056427,
+      "learning_rate": 1.7079571468448917e-05,
+      "loss": 2.4175,
+      "step": 16245
+    },
+    {
+      "epoch": 1.3111129045274796,
+      "grad_norm": 0.7028010487556458,
+      "learning_rate": 1.7070748483762854e-05,
+      "loss": 2.41,
+      "step": 16246
+    },
+    {
+      "epoch": 1.3111936082640625,
+      "grad_norm": 0.7241945862770081,
+      "learning_rate": 1.7061927565890522e-05,
+      "loss": 2.4171,
+      "step": 16247
+    },
+    {
+      "epoch": 1.3112743120006456,
+      "grad_norm": 0.7039221525192261,
+      "learning_rate": 1.705310871505177e-05,
+      "loss": 2.4154,
+      "step": 16248
+    },
+    {
+      "epoch": 1.3113550157372287,
+      "grad_norm": 0.672444760799408,
+      "learning_rate": 1.704429193146636e-05,
+      "loss": 2.4025,
+      "step": 16249
+    },
+    {
+      "epoch": 1.3114357194738115,
+      "grad_norm": 0.7240859866142273,
+      "learning_rate": 1.7035477215354068e-05,
+      "loss": 2.3864,
+      "step": 16250
+    },
+    {
+      "epoch": 1.3115164232103946,
+      "grad_norm": 0.7379294633865356,
+      "learning_rate": 1.7026664566934536e-05,
+      "loss": 2.4663,
+      "step": 16251
+    },
+    {
+      "epoch": 1.3115971269469777,
+      "grad_norm": 0.6928708553314209,
+      "learning_rate": 1.7017853986427425e-05,
+      "loss": 2.4407,
+      "step": 16252
+    },
+    {
+      "epoch": 1.3116778306835606,
+      "grad_norm": 0.6304093599319458,
+      "learning_rate": 1.7009045474052298e-05,
+      "loss": 2.4755,
+      "step": 16253
+    },
+    {
+      "epoch": 1.3117585344201437,
+      "grad_norm": 0.6945829391479492,
+      "learning_rate": 1.700023903002872e-05,
+      "loss": 2.3817,
+      "step": 16254
+    },
+    {
+      "epoch": 1.3118392381567268,
+      "grad_norm": 0.6899009346961975,
+      "learning_rate": 1.6991434654576133e-05,
+      "loss": 2.3989,
+      "step": 16255
+    },
+    {
+      "epoch": 1.3119199418933096,
+      "grad_norm": 0.7359157204627991,
+      "learning_rate": 1.6982632347913985e-05,
+      "loss": 2.3788,
+      "step": 16256
+    },
+    {
+      "epoch": 1.3120006456298927,
+      "grad_norm": 0.6562486886978149,
+      "learning_rate": 1.6973832110261658e-05,
+      "loss": 2.3955,
+      "step": 16257
+    },
+    {
+      "epoch": 1.3120813493664758,
+      "grad_norm": 0.6772989630699158,
+      "learning_rate": 1.696503394183846e-05,
+      "loss": 2.4788,
+      "step": 16258
+    },
+    {
+      "epoch": 1.3121620531030587,
+      "grad_norm": 0.7214391231536865,
+      "learning_rate": 1.695623784286363e-05,
+      "loss": 2.3836,
+      "step": 16259
+    },
+    {
+      "epoch": 1.3122427568396418,
+      "grad_norm": 0.7041679620742798,
+      "learning_rate": 1.6947443813556495e-05,
+      "loss": 2.4547,
+      "step": 16260
+    },
+    {
+      "epoch": 1.3123234605762246,
+      "grad_norm": 0.6819555163383484,
+      "learning_rate": 1.6938651854136135e-05,
+      "loss": 2.468,
+      "step": 16261
+    },
+    {
+      "epoch": 1.3124041643128077,
+      "grad_norm": 0.6466858983039856,
+      "learning_rate": 1.6929861964821693e-05,
+      "loss": 2.4572,
+      "step": 16262
+    },
+    {
+      "epoch": 1.3124848680493906,
+      "grad_norm": 0.688709557056427,
+      "learning_rate": 1.6921074145832248e-05,
+      "loss": 2.3891,
+      "step": 16263
+    },
+    {
+      "epoch": 1.3125655717859737,
+      "grad_norm": 0.6896470785140991,
+      "learning_rate": 1.69122883973868e-05,
+      "loss": 2.3825,
+      "step": 16264
+    },
+    {
+      "epoch": 1.3126462755225567,
+      "grad_norm": 0.8242524266242981,
+      "learning_rate": 1.690350471970431e-05,
+      "loss": 2.4804,
+      "step": 16265
+    },
+    {
+      "epoch": 1.3127269792591396,
+      "grad_norm": 0.7506044507026672,
+      "learning_rate": 1.689472311300373e-05,
+      "loss": 2.4671,
+      "step": 16266
+    },
+    {
+      "epoch": 1.3128076829957227,
+      "grad_norm": 0.6776263117790222,
+      "learning_rate": 1.688594357750386e-05,
+      "loss": 2.4646,
+      "step": 16267
+    },
+    {
+      "epoch": 1.3128883867323058,
+      "grad_norm": 0.6843759417533875,
+      "learning_rate": 1.6877166113423548e-05,
+      "loss": 2.4147,
+      "step": 16268
+    },
+    {
+      "epoch": 1.3129690904688887,
+      "grad_norm": 0.6650474667549133,
+      "learning_rate": 1.686839072098153e-05,
+      "loss": 2.4379,
+      "step": 16269
+    },
+    {
+      "epoch": 1.3130497942054717,
+      "grad_norm": 0.6636466383934021,
+      "learning_rate": 1.6859617400396533e-05,
+      "loss": 2.4334,
+      "step": 16270
+    },
+    {
+      "epoch": 1.3131304979420548,
+      "grad_norm": 0.649217963218689,
+      "learning_rate": 1.685084615188719e-05,
+      "loss": 2.319,
+      "step": 16271
+    },
+    {
+      "epoch": 1.3132112016786377,
+      "grad_norm": 0.7343039512634277,
+      "learning_rate": 1.6842076975672126e-05,
+      "loss": 2.3844,
+      "step": 16272
+    },
+    {
+      "epoch": 1.3132919054152208,
+      "grad_norm": 0.6916847825050354,
+      "learning_rate": 1.6833309871969894e-05,
+      "loss": 2.4544,
+      "step": 16273
+    },
+    {
+      "epoch": 1.3133726091518036,
+      "grad_norm": 0.6762102842330933,
+      "learning_rate": 1.6824544840998967e-05,
+      "loss": 2.3912,
+      "step": 16274
+    },
+    {
+      "epoch": 1.3134533128883867,
+      "grad_norm": 0.7327221035957336,
+      "learning_rate": 1.68157818829778e-05,
+      "loss": 2.4403,
+      "step": 16275
+    },
+    {
+      "epoch": 1.3135340166249696,
+      "grad_norm": 0.7362363338470459,
+      "learning_rate": 1.6807020998124812e-05,
+      "loss": 2.5169,
+      "step": 16276
+    },
+    {
+      "epoch": 1.3136147203615527,
+      "grad_norm": 0.6882300972938538,
+      "learning_rate": 1.679826218665832e-05,
+      "loss": 2.4139,
+      "step": 16277
+    },
+    {
+      "epoch": 1.3136954240981358,
+      "grad_norm": 0.7146984934806824,
+      "learning_rate": 1.6789505448796615e-05,
+      "loss": 2.4738,
+      "step": 16278
+    },
+    {
+      "epoch": 1.3137761278347186,
+      "grad_norm": 0.6581223607063293,
+      "learning_rate": 1.6780750784757947e-05,
+      "loss": 2.4617,
+      "step": 16279
+    },
+    {
+      "epoch": 1.3138568315713017,
+      "grad_norm": 0.7729318141937256,
+      "learning_rate": 1.6771998194760518e-05,
+      "loss": 2.4541,
+      "step": 16280
+    },
+    {
+      "epoch": 1.3139375353078848,
+      "grad_norm": 0.7617159485816956,
+      "learning_rate": 1.6763247679022442e-05,
+      "loss": 2.4727,
+      "step": 16281
+    },
+    {
+      "epoch": 1.3140182390444677,
+      "grad_norm": 0.6640555262565613,
+      "learning_rate": 1.6754499237761844e-05,
+      "loss": 2.4717,
+      "step": 16282
+    },
+    {
+      "epoch": 1.3140989427810508,
+      "grad_norm": 0.7289882898330688,
+      "learning_rate": 1.6745752871196707e-05,
+      "loss": 2.4515,
+      "step": 16283
+    },
+    {
+      "epoch": 1.3141796465176339,
+      "grad_norm": 0.7075887322425842,
+      "learning_rate": 1.6737008579545043e-05,
+      "loss": 2.4586,
+      "step": 16284
+    },
+    {
+      "epoch": 1.3142603502542167,
+      "grad_norm": 0.7152252197265625,
+      "learning_rate": 1.672826636302477e-05,
+      "loss": 2.512,
+      "step": 16285
+    },
+    {
+      "epoch": 1.3143410539907998,
+      "grad_norm": 0.6875295639038086,
+      "learning_rate": 1.6719526221853808e-05,
+      "loss": 2.4049,
+      "step": 16286
+    },
+    {
+      "epoch": 1.314421757727383,
+      "grad_norm": 0.6812484860420227,
+      "learning_rate": 1.671078815624991e-05,
+      "loss": 2.3705,
+      "step": 16287
+    },
+    {
+      "epoch": 1.3145024614639658,
+      "grad_norm": 0.664282500743866,
+      "learning_rate": 1.6702052166430904e-05,
+      "loss": 2.3776,
+      "step": 16288
+    },
+    {
+      "epoch": 1.3145831652005489,
+      "grad_norm": 0.7460842728614807,
+      "learning_rate": 1.66933182526145e-05,
+      "loss": 2.4525,
+      "step": 16289
+    },
+    {
+      "epoch": 1.3146638689371317,
+      "grad_norm": 0.6555477380752563,
+      "learning_rate": 1.6684586415018366e-05,
+      "loss": 2.3902,
+      "step": 16290
+    },
+    {
+      "epoch": 1.3147445726737148,
+      "grad_norm": 0.7191921472549438,
+      "learning_rate": 1.6675856653860135e-05,
+      "loss": 2.4957,
+      "step": 16291
+    },
+    {
+      "epoch": 1.3148252764102977,
+      "grad_norm": 0.738667368888855,
+      "learning_rate": 1.666712896935738e-05,
+      "loss": 2.4182,
+      "step": 16292
+    },
+    {
+      "epoch": 1.3149059801468808,
+      "grad_norm": 0.6764421463012695,
+      "learning_rate": 1.6658403361727593e-05,
+      "loss": 2.4179,
+      "step": 16293
+    },
+    {
+      "epoch": 1.3149866838834638,
+      "grad_norm": 0.6981594562530518,
+      "learning_rate": 1.6649679831188247e-05,
+      "loss": 2.4288,
+      "step": 16294
+    },
+    {
+      "epoch": 1.3150673876200467,
+      "grad_norm": 0.6657801866531372,
+      "learning_rate": 1.6640958377956784e-05,
+      "loss": 2.3716,
+      "step": 16295
+    },
+    {
+      "epoch": 1.3151480913566298,
+      "grad_norm": 0.7238973379135132,
+      "learning_rate": 1.6632239002250505e-05,
+      "loss": 2.438,
+      "step": 16296
+    },
+    {
+      "epoch": 1.3152287950932129,
+      "grad_norm": 0.6727766990661621,
+      "learning_rate": 1.6623521704286772e-05,
+      "loss": 2.4406,
+      "step": 16297
+    },
+    {
+      "epoch": 1.3153094988297958,
+      "grad_norm": 0.6741603016853333,
+      "learning_rate": 1.661480648428282e-05,
+      "loss": 2.4379,
+      "step": 16298
+    },
+    {
+      "epoch": 1.3153902025663788,
+      "grad_norm": 0.7174610495567322,
+      "learning_rate": 1.6606093342455865e-05,
+      "loss": 2.4368,
+      "step": 16299
+    },
+    {
+      "epoch": 1.315470906302962,
+      "grad_norm": 0.6604920029640198,
+      "learning_rate": 1.6597382279023057e-05,
+      "loss": 2.4431,
+      "step": 16300
+    },
+    {
+      "epoch": 1.3155516100395448,
+      "grad_norm": 0.6930821537971497,
+      "learning_rate": 1.6588673294201494e-05,
+      "loss": 2.4064,
+      "step": 16301
+    },
+    {
+      "epoch": 1.3156323137761279,
+      "grad_norm": 0.6489799618721008,
+      "learning_rate": 1.657996638820826e-05,
+      "loss": 2.4256,
+      "step": 16302
+    },
+    {
+      "epoch": 1.315713017512711,
+      "grad_norm": 0.6781083345413208,
+      "learning_rate": 1.65712615612603e-05,
+      "loss": 2.4731,
+      "step": 16303
+    },
+    {
+      "epoch": 1.3157937212492938,
+      "grad_norm": 0.6710748076438904,
+      "learning_rate": 1.656255881357458e-05,
+      "loss": 2.4065,
+      "step": 16304
+    },
+    {
+      "epoch": 1.315874424985877,
+      "grad_norm": 0.7099822163581848,
+      "learning_rate": 1.655385814536804e-05,
+      "loss": 2.3978,
+      "step": 16305
+    },
+    {
+      "epoch": 1.3159551287224598,
+      "grad_norm": 0.7215133905410767,
+      "learning_rate": 1.6545159556857447e-05,
+      "loss": 2.4655,
+      "step": 16306
+    },
+    {
+      "epoch": 1.3160358324590429,
+      "grad_norm": 0.7705253958702087,
+      "learning_rate": 1.6536463048259643e-05,
+      "loss": 2.4576,
+      "step": 16307
+    },
+    {
+      "epoch": 1.3161165361956257,
+      "grad_norm": 0.6232311725616455,
+      "learning_rate": 1.6527768619791372e-05,
+      "loss": 2.3923,
+      "step": 16308
+    },
+    {
+      "epoch": 1.3161972399322088,
+      "grad_norm": 0.6599528789520264,
+      "learning_rate": 1.6519076271669264e-05,
+      "loss": 2.4236,
+      "step": 16309
+    },
+    {
+      "epoch": 1.316277943668792,
+      "grad_norm": 0.6598034501075745,
+      "learning_rate": 1.6510386004110023e-05,
+      "loss": 2.368,
+      "step": 16310
+    },
+    {
+      "epoch": 1.3163586474053748,
+      "grad_norm": 0.6949655413627625,
+      "learning_rate": 1.650169781733022e-05,
+      "loss": 2.4277,
+      "step": 16311
+    },
+    {
+      "epoch": 1.3164393511419579,
+      "grad_norm": 0.6838186383247375,
+      "learning_rate": 1.6493011711546358e-05,
+      "loss": 2.4413,
+      "step": 16312
+    },
+    {
+      "epoch": 1.316520054878541,
+      "grad_norm": 0.7026765942573547,
+      "learning_rate": 1.6484327686974933e-05,
+      "loss": 2.4628,
+      "step": 16313
+    },
+    {
+      "epoch": 1.3166007586151238,
+      "grad_norm": 0.745360791683197,
+      "learning_rate": 1.647564574383237e-05,
+      "loss": 2.4358,
+      "step": 16314
+    },
+    {
+      "epoch": 1.316681462351707,
+      "grad_norm": 0.676225483417511,
+      "learning_rate": 1.6466965882335083e-05,
+      "loss": 2.4119,
+      "step": 16315
+    },
+    {
+      "epoch": 1.31676216608829,
+      "grad_norm": 0.6767755150794983,
+      "learning_rate": 1.6458288102699325e-05,
+      "loss": 2.4322,
+      "step": 16316
+    },
+    {
+      "epoch": 1.3168428698248729,
+      "grad_norm": 0.6957309246063232,
+      "learning_rate": 1.6449612405141424e-05,
+      "loss": 2.4327,
+      "step": 16317
+    },
+    {
+      "epoch": 1.316923573561456,
+      "grad_norm": 0.6773050427436829,
+      "learning_rate": 1.64409387898776e-05,
+      "loss": 2.4207,
+      "step": 16318
+    },
+    {
+      "epoch": 1.3170042772980388,
+      "grad_norm": 0.7319278717041016,
+      "learning_rate": 1.6432267257123978e-05,
+      "loss": 2.445,
+      "step": 16319
+    },
+    {
+      "epoch": 1.317084981034622,
+      "grad_norm": 0.7531326413154602,
+      "learning_rate": 1.6423597807096714e-05,
+      "loss": 2.3948,
+      "step": 16320
+    },
+    {
+      "epoch": 1.3171656847712048,
+      "grad_norm": 0.6741669178009033,
+      "learning_rate": 1.6414930440011854e-05,
+      "loss": 2.4177,
+      "step": 16321
+    },
+    {
+      "epoch": 1.3172463885077879,
+      "grad_norm": 0.6814963221549988,
+      "learning_rate": 1.640626515608543e-05,
+      "loss": 2.4419,
+      "step": 16322
+    },
+    {
+      "epoch": 1.317327092244371,
+      "grad_norm": 0.6740893721580505,
+      "learning_rate": 1.6397601955533392e-05,
+      "loss": 2.3516,
+      "step": 16323
+    },
+    {
+      "epoch": 1.3174077959809538,
+      "grad_norm": 0.7172163724899292,
+      "learning_rate": 1.6388940838571675e-05,
+      "loss": 2.4665,
+      "step": 16324
+    },
+    {
+      "epoch": 1.317488499717537,
+      "grad_norm": 0.6690489053726196,
+      "learning_rate": 1.6380281805416085e-05,
+      "loss": 2.3957,
+      "step": 16325
+    },
+    {
+      "epoch": 1.31756920345412,
+      "grad_norm": 0.7182994484901428,
+      "learning_rate": 1.6371624856282462e-05,
+      "loss": 2.4456,
+      "step": 16326
+    },
+    {
+      "epoch": 1.3176499071907029,
+      "grad_norm": 0.6324366927146912,
+      "learning_rate": 1.636296999138659e-05,
+      "loss": 2.4111,
+      "step": 16327
+    },
+    {
+      "epoch": 1.317730610927286,
+      "grad_norm": 0.6740162372589111,
+      "learning_rate": 1.6354317210944093e-05,
+      "loss": 2.451,
+      "step": 16328
+    },
+    {
+      "epoch": 1.317811314663869,
+      "grad_norm": 0.6964122653007507,
+      "learning_rate": 1.6345666515170665e-05,
+      "loss": 2.4269,
+      "step": 16329
+    },
+    {
+      "epoch": 1.317892018400452,
+      "grad_norm": 0.7093058824539185,
+      "learning_rate": 1.6337017904281915e-05,
+      "loss": 2.4686,
+      "step": 16330
+    },
+    {
+      "epoch": 1.317972722137035,
+      "grad_norm": 0.693233072757721,
+      "learning_rate": 1.6328371378493367e-05,
+      "loss": 2.4149,
+      "step": 16331
+    },
+    {
+      "epoch": 1.318053425873618,
+      "grad_norm": 0.6418019533157349,
+      "learning_rate": 1.631972693802052e-05,
+      "loss": 2.4268,
+      "step": 16332
+    },
+    {
+      "epoch": 1.318134129610201,
+      "grad_norm": 0.6815310120582581,
+      "learning_rate": 1.631108458307883e-05,
+      "loss": 2.4274,
+      "step": 16333
+    },
+    {
+      "epoch": 1.318214833346784,
+      "grad_norm": 0.6774280071258545,
+      "learning_rate": 1.630244431388369e-05,
+      "loss": 2.3927,
+      "step": 16334
+    },
+    {
+      "epoch": 1.3182955370833669,
+      "grad_norm": 0.688090443611145,
+      "learning_rate": 1.6293806130650413e-05,
+      "loss": 2.4013,
+      "step": 16335
+    },
+    {
+      "epoch": 1.31837624081995,
+      "grad_norm": 0.7300553321838379,
+      "learning_rate": 1.6285170033594288e-05,
+      "loss": 2.4716,
+      "step": 16336
+    },
+    {
+      "epoch": 1.3184569445565328,
+      "grad_norm": 0.6798286437988281,
+      "learning_rate": 1.627653602293059e-05,
+      "loss": 2.3893,
+      "step": 16337
+    },
+    {
+      "epoch": 1.318537648293116,
+      "grad_norm": 0.6699275970458984,
+      "learning_rate": 1.6267904098874442e-05,
+      "loss": 2.4446,
+      "step": 16338
+    },
+    {
+      "epoch": 1.318618352029699,
+      "grad_norm": 0.7632322311401367,
+      "learning_rate": 1.6259274261641e-05,
+      "loss": 2.4434,
+      "step": 16339
+    },
+    {
+      "epoch": 1.3186990557662819,
+      "grad_norm": 0.7156099677085876,
+      "learning_rate": 1.6250646511445343e-05,
+      "loss": 2.4142,
+      "step": 16340
+    },
+    {
+      "epoch": 1.318779759502865,
+      "grad_norm": 0.7525599598884583,
+      "learning_rate": 1.6242020848502505e-05,
+      "loss": 2.3543,
+      "step": 16341
+    },
+    {
+      "epoch": 1.318860463239448,
+      "grad_norm": 0.7063113451004028,
+      "learning_rate": 1.623339727302745e-05,
+      "loss": 2.4754,
+      "step": 16342
+    },
+    {
+      "epoch": 1.318941166976031,
+      "grad_norm": 0.7138137221336365,
+      "learning_rate": 1.6224775785235123e-05,
+      "loss": 2.4223,
+      "step": 16343
+    },
+    {
+      "epoch": 1.319021870712614,
+      "grad_norm": 0.6976706981658936,
+      "learning_rate": 1.6216156385340352e-05,
+      "loss": 2.4878,
+      "step": 16344
+    },
+    {
+      "epoch": 1.319102574449197,
+      "grad_norm": 0.6931003332138062,
+      "learning_rate": 1.6207539073557974e-05,
+      "loss": 2.39,
+      "step": 16345
+    },
+    {
+      "epoch": 1.31918327818578,
+      "grad_norm": 0.6919357180595398,
+      "learning_rate": 1.6198923850102765e-05,
+      "loss": 2.4197,
+      "step": 16346
+    },
+    {
+      "epoch": 1.319263981922363,
+      "grad_norm": 0.7453805804252625,
+      "learning_rate": 1.619031071518945e-05,
+      "loss": 2.4226,
+      "step": 16347
+    },
+    {
+      "epoch": 1.3193446856589461,
+      "grad_norm": 0.6990562677383423,
+      "learning_rate": 1.6181699669032658e-05,
+      "loss": 2.3925,
+      "step": 16348
+    },
+    {
+      "epoch": 1.319425389395529,
+      "grad_norm": 0.6974303126335144,
+      "learning_rate": 1.6173090711847006e-05,
+      "loss": 2.445,
+      "step": 16349
+    },
+    {
+      "epoch": 1.319506093132112,
+      "grad_norm": 0.7278286814689636,
+      "learning_rate": 1.6164483843847057e-05,
+      "loss": 2.3869,
+      "step": 16350
+    },
+    {
+      "epoch": 1.319586796868695,
+      "grad_norm": 0.7282646298408508,
+      "learning_rate": 1.6155879065247326e-05,
+      "loss": 2.3694,
+      "step": 16351
+    },
+    {
+      "epoch": 1.319667500605278,
+      "grad_norm": 0.7329844832420349,
+      "learning_rate": 1.6147276376262255e-05,
+      "loss": 2.4369,
+      "step": 16352
+    },
+    {
+      "epoch": 1.319748204341861,
+      "grad_norm": 0.6499385833740234,
+      "learning_rate": 1.613867577710627e-05,
+      "loss": 2.441,
+      "step": 16353
+    },
+    {
+      "epoch": 1.319828908078444,
+      "grad_norm": 0.7026061415672302,
+      "learning_rate": 1.6130077267993683e-05,
+      "loss": 2.4117,
+      "step": 16354
+    },
+    {
+      "epoch": 1.319909611815027,
+      "grad_norm": 0.7007814049720764,
+      "learning_rate": 1.6121480849138803e-05,
+      "loss": 2.4287,
+      "step": 16355
+    },
+    {
+      "epoch": 1.31999031555161,
+      "grad_norm": 0.6525697708129883,
+      "learning_rate": 1.611288652075591e-05,
+      "loss": 2.3969,
+      "step": 16356
+    },
+    {
+      "epoch": 1.320071019288193,
+      "grad_norm": 0.7268216609954834,
+      "learning_rate": 1.610429428305914e-05,
+      "loss": 2.4227,
+      "step": 16357
+    },
+    {
+      "epoch": 1.3201517230247761,
+      "grad_norm": 0.6665107011795044,
+      "learning_rate": 1.6095704136262668e-05,
+      "loss": 2.3694,
+      "step": 16358
+    },
+    {
+      "epoch": 1.320232426761359,
+      "grad_norm": 0.6832399368286133,
+      "learning_rate": 1.60871160805806e-05,
+      "loss": 2.4001,
+      "step": 16359
+    },
+    {
+      "epoch": 1.320313130497942,
+      "grad_norm": 0.6788592338562012,
+      "learning_rate": 1.6078530116226897e-05,
+      "loss": 2.4294,
+      "step": 16360
+    },
+    {
+      "epoch": 1.3203938342345252,
+      "grad_norm": 0.7147449254989624,
+      "learning_rate": 1.6069946243415625e-05,
+      "loss": 2.3904,
+      "step": 16361
+    },
+    {
+      "epoch": 1.320474537971108,
+      "grad_norm": 0.7014418840408325,
+      "learning_rate": 1.6061364462360683e-05,
+      "loss": 2.4026,
+      "step": 16362
+    },
+    {
+      "epoch": 1.3205552417076911,
+      "grad_norm": 0.6867612600326538,
+      "learning_rate": 1.6052784773275987e-05,
+      "loss": 2.4092,
+      "step": 16363
+    },
+    {
+      "epoch": 1.3206359454442742,
+      "grad_norm": 0.6588961482048035,
+      "learning_rate": 1.6044207176375303e-05,
+      "loss": 2.4588,
+      "step": 16364
+    },
+    {
+      "epoch": 1.320716649180857,
+      "grad_norm": 0.688671350479126,
+      "learning_rate": 1.6035631671872444e-05,
+      "loss": 2.3957,
+      "step": 16365
+    },
+    {
+      "epoch": 1.3207973529174402,
+      "grad_norm": 0.7548064589500427,
+      "learning_rate": 1.6027058259981154e-05,
+      "loss": 2.4168,
+      "step": 16366
+    },
+    {
+      "epoch": 1.320878056654023,
+      "grad_norm": 0.7251972556114197,
+      "learning_rate": 1.6018486940915044e-05,
+      "loss": 2.4704,
+      "step": 16367
+    },
+    {
+      "epoch": 1.3209587603906061,
+      "grad_norm": 0.73149174451828,
+      "learning_rate": 1.6009917714887778e-05,
+      "loss": 2.4597,
+      "step": 16368
+    },
+    {
+      "epoch": 1.321039464127189,
+      "grad_norm": 0.6741003394126892,
+      "learning_rate": 1.600135058211294e-05,
+      "loss": 2.3876,
+      "step": 16369
+    },
+    {
+      "epoch": 1.321120167863772,
+      "grad_norm": 0.6891310214996338,
+      "learning_rate": 1.5992785542804e-05,
+      "loss": 2.4229,
+      "step": 16370
+    },
+    {
+      "epoch": 1.3212008716003552,
+      "grad_norm": 0.7529458403587341,
+      "learning_rate": 1.5984222597174415e-05,
+      "loss": 2.45,
+      "step": 16371
+    },
+    {
+      "epoch": 1.321281575336938,
+      "grad_norm": 0.708134651184082,
+      "learning_rate": 1.5975661745437664e-05,
+      "loss": 2.454,
+      "step": 16372
+    },
+    {
+      "epoch": 1.321362279073521,
+      "grad_norm": 0.7511130571365356,
+      "learning_rate": 1.596710298780705e-05,
+      "loss": 2.4201,
+      "step": 16373
+    },
+    {
+      "epoch": 1.3214429828101042,
+      "grad_norm": 0.6599537134170532,
+      "learning_rate": 1.595854632449588e-05,
+      "loss": 2.3982,
+      "step": 16374
+    },
+    {
+      "epoch": 1.321523686546687,
+      "grad_norm": 0.6821228861808777,
+      "learning_rate": 1.5949991755717453e-05,
+      "loss": 2.4525,
+      "step": 16375
+    },
+    {
+      "epoch": 1.3216043902832701,
+      "grad_norm": 0.6872302293777466,
+      "learning_rate": 1.5941439281684923e-05,
+      "loss": 2.3631,
+      "step": 16376
+    },
+    {
+      "epoch": 1.3216850940198532,
+      "grad_norm": 0.6650066375732422,
+      "learning_rate": 1.5932888902611453e-05,
+      "loss": 2.3718,
+      "step": 16377
+    },
+    {
+      "epoch": 1.321765797756436,
+      "grad_norm": 0.6620016694068909,
+      "learning_rate": 1.5924340618710143e-05,
+      "loss": 2.4076,
+      "step": 16378
+    },
+    {
+      "epoch": 1.3218465014930192,
+      "grad_norm": 0.694807231426239,
+      "learning_rate": 1.5915794430194066e-05,
+      "loss": 2.4369,
+      "step": 16379
+    },
+    {
+      "epoch": 1.321927205229602,
+      "grad_norm": 0.6810131669044495,
+      "learning_rate": 1.590725033727616e-05,
+      "loss": 2.4151,
+      "step": 16380
+    },
+    {
+      "epoch": 1.3220079089661851,
+      "grad_norm": 0.768846333026886,
+      "learning_rate": 1.58987083401694e-05,
+      "loss": 2.4991,
+      "step": 16381
+    },
+    {
+      "epoch": 1.322088612702768,
+      "grad_norm": 0.6581698656082153,
+      "learning_rate": 1.5890168439086672e-05,
+      "loss": 2.4263,
+      "step": 16382
+    },
+    {
+      "epoch": 1.322169316439351,
+      "grad_norm": 0.7267034649848938,
+      "learning_rate": 1.5881630634240818e-05,
+      "loss": 2.4219,
+      "step": 16383
+    },
+    {
+      "epoch": 1.3222500201759342,
+      "grad_norm": 0.7391555905342102,
+      "learning_rate": 1.5873094925844612e-05,
+      "loss": 2.427,
+      "step": 16384
+    },
+    {
+      "epoch": 1.322330723912517,
+      "grad_norm": 0.6612021923065186,
+      "learning_rate": 1.5864561314110815e-05,
+      "loss": 2.4108,
+      "step": 16385
+    },
+    {
+      "epoch": 1.3224114276491001,
+      "grad_norm": 0.7118437886238098,
+      "learning_rate": 1.585602979925206e-05,
+      "loss": 2.3839,
+      "step": 16386
+    },
+    {
+      "epoch": 1.3224921313856832,
+      "grad_norm": 0.6663616299629211,
+      "learning_rate": 1.5847500381480997e-05,
+      "loss": 2.4302,
+      "step": 16387
+    },
+    {
+      "epoch": 1.322572835122266,
+      "grad_norm": 0.6848715543746948,
+      "learning_rate": 1.583897306101022e-05,
+      "loss": 2.4228,
+      "step": 16388
+    },
+    {
+      "epoch": 1.3226535388588492,
+      "grad_norm": 0.680895209312439,
+      "learning_rate": 1.5830447838052208e-05,
+      "loss": 2.4457,
+      "step": 16389
+    },
+    {
+      "epoch": 1.3227342425954323,
+      "grad_norm": 0.683276891708374,
+      "learning_rate": 1.582192471281946e-05,
+      "loss": 2.4412,
+      "step": 16390
+    },
+    {
+      "epoch": 1.3228149463320151,
+      "grad_norm": 0.7311880588531494,
+      "learning_rate": 1.5813403685524396e-05,
+      "loss": 2.4604,
+      "step": 16391
+    },
+    {
+      "epoch": 1.3228956500685982,
+      "grad_norm": 0.6769095659255981,
+      "learning_rate": 1.580488475637937e-05,
+      "loss": 2.4311,
+      "step": 16392
+    },
+    {
+      "epoch": 1.3229763538051813,
+      "grad_norm": 0.6683096289634705,
+      "learning_rate": 1.579636792559671e-05,
+      "loss": 2.445,
+      "step": 16393
+    },
+    {
+      "epoch": 1.3230570575417642,
+      "grad_norm": 0.7268782258033752,
+      "learning_rate": 1.5787853193388667e-05,
+      "loss": 2.4176,
+      "step": 16394
+    },
+    {
+      "epoch": 1.3231377612783473,
+      "grad_norm": 0.6878541707992554,
+      "learning_rate": 1.5779340559967494e-05,
+      "loss": 2.4615,
+      "step": 16395
+    },
+    {
+      "epoch": 1.3232184650149301,
+      "grad_norm": 0.7031291127204895,
+      "learning_rate": 1.577083002554527e-05,
+      "loss": 2.3726,
+      "step": 16396
+    },
+    {
+      "epoch": 1.3232991687515132,
+      "grad_norm": 0.7738708853721619,
+      "learning_rate": 1.5762321590334138e-05,
+      "loss": 2.5046,
+      "step": 16397
+    },
+    {
+      "epoch": 1.323379872488096,
+      "grad_norm": 0.6660913228988647,
+      "learning_rate": 1.575381525454619e-05,
+      "loss": 2.3759,
+      "step": 16398
+    },
+    {
+      "epoch": 1.3234605762246792,
+      "grad_norm": 0.6534021496772766,
+      "learning_rate": 1.574531101839335e-05,
+      "loss": 2.3983,
+      "step": 16399
+    },
+    {
+      "epoch": 1.3235412799612623,
+      "grad_norm": 0.6645511388778687,
+      "learning_rate": 1.5736808882087606e-05,
+      "loss": 2.3958,
+      "step": 16400
+    },
+    {
+      "epoch": 1.3236219836978451,
+      "grad_norm": 0.6723225712776184,
+      "learning_rate": 1.5728308845840855e-05,
+      "loss": 2.4248,
+      "step": 16401
+    },
+    {
+      "epoch": 1.3237026874344282,
+      "grad_norm": 0.6609976887702942,
+      "learning_rate": 1.5719810909864942e-05,
+      "loss": 2.3888,
+      "step": 16402
+    },
+    {
+      "epoch": 1.3237833911710113,
+      "grad_norm": 0.6713845729827881,
+      "learning_rate": 1.5711315074371635e-05,
+      "loss": 2.4474,
+      "step": 16403
+    },
+    {
+      "epoch": 1.3238640949075942,
+      "grad_norm": 0.701438307762146,
+      "learning_rate": 1.5702821339572726e-05,
+      "loss": 2.4673,
+      "step": 16404
+    },
+    {
+      "epoch": 1.3239447986441772,
+      "grad_norm": 0.7235428094863892,
+      "learning_rate": 1.5694329705679834e-05,
+      "loss": 2.3825,
+      "step": 16405
+    },
+    {
+      "epoch": 1.3240255023807603,
+      "grad_norm": 0.6785053610801697,
+      "learning_rate": 1.568584017290462e-05,
+      "loss": 2.4668,
+      "step": 16406
+    },
+    {
+      "epoch": 1.3241062061173432,
+      "grad_norm": 0.6918929815292358,
+      "learning_rate": 1.5677352741458705e-05,
+      "loss": 2.4329,
+      "step": 16407
+    },
+    {
+      "epoch": 1.3241869098539263,
+      "grad_norm": 0.7194826006889343,
+      "learning_rate": 1.5668867411553544e-05,
+      "loss": 2.3717,
+      "step": 16408
+    },
+    {
+      "epoch": 1.3242676135905094,
+      "grad_norm": 0.7299134731292725,
+      "learning_rate": 1.5660384183400658e-05,
+      "loss": 2.4695,
+      "step": 16409
+    },
+    {
+      "epoch": 1.3243483173270922,
+      "grad_norm": 0.7047600746154785,
+      "learning_rate": 1.565190305721147e-05,
+      "loss": 2.4525,
+      "step": 16410
+    },
+    {
+      "epoch": 1.3244290210636753,
+      "grad_norm": 0.685001015663147,
+      "learning_rate": 1.5643424033197328e-05,
+      "loss": 2.322,
+      "step": 16411
+    },
+    {
+      "epoch": 1.3245097248002582,
+      "grad_norm": 0.7696635127067566,
+      "learning_rate": 1.5634947111569588e-05,
+      "loss": 2.4464,
+      "step": 16412
+    },
+    {
+      "epoch": 1.3245904285368413,
+      "grad_norm": 0.7066066265106201,
+      "learning_rate": 1.5626472292539485e-05,
+      "loss": 2.4315,
+      "step": 16413
+    },
+    {
+      "epoch": 1.3246711322734241,
+      "grad_norm": 0.6553033590316772,
+      "learning_rate": 1.5617999576318276e-05,
+      "loss": 2.4296,
+      "step": 16414
+    },
+    {
+      "epoch": 1.3247518360100072,
+      "grad_norm": 0.7031354308128357,
+      "learning_rate": 1.560952896311707e-05,
+      "loss": 2.4565,
+      "step": 16415
+    },
+    {
+      "epoch": 1.3248325397465903,
+      "grad_norm": 0.7826353311538696,
+      "learning_rate": 1.560106045314701e-05,
+      "loss": 2.4275,
+      "step": 16416
+    },
+    {
+      "epoch": 1.3249132434831732,
+      "grad_norm": 0.6408981084823608,
+      "learning_rate": 1.559259404661916e-05,
+      "loss": 2.3869,
+      "step": 16417
+    },
+    {
+      "epoch": 1.3249939472197563,
+      "grad_norm": 0.7487547993659973,
+      "learning_rate": 1.558412974374448e-05,
+      "loss": 2.3678,
+      "step": 16418
+    },
+    {
+      "epoch": 1.3250746509563394,
+      "grad_norm": 0.7163991332054138,
+      "learning_rate": 1.5575667544733963e-05,
+      "loss": 2.397,
+      "step": 16419
+    },
+    {
+      "epoch": 1.3251553546929222,
+      "grad_norm": 0.6933553814888,
+      "learning_rate": 1.5567207449798515e-05,
+      "loss": 2.424,
+      "step": 16420
+    },
+    {
+      "epoch": 1.3252360584295053,
+      "grad_norm": 0.687406063079834,
+      "learning_rate": 1.5558749459148945e-05,
+      "loss": 2.4346,
+      "step": 16421
+    },
+    {
+      "epoch": 1.3253167621660884,
+      "grad_norm": 0.6781243681907654,
+      "learning_rate": 1.5550293572996054e-05,
+      "loss": 2.4526,
+      "step": 16422
+    },
+    {
+      "epoch": 1.3253974659026713,
+      "grad_norm": 0.6632506847381592,
+      "learning_rate": 1.5541839791550616e-05,
+      "loss": 2.4559,
+      "step": 16423
+    },
+    {
+      "epoch": 1.3254781696392544,
+      "grad_norm": 0.668396532535553,
+      "learning_rate": 1.5533388115023327e-05,
+      "loss": 2.4463,
+      "step": 16424
+    },
+    {
+      "epoch": 1.3255588733758372,
+      "grad_norm": 0.6853309869766235,
+      "learning_rate": 1.552493854362479e-05,
+      "loss": 2.429,
+      "step": 16425
+    },
+    {
+      "epoch": 1.3256395771124203,
+      "grad_norm": 0.7443413138389587,
+      "learning_rate": 1.5516491077565597e-05,
+      "loss": 2.4091,
+      "step": 16426
+    },
+    {
+      "epoch": 1.3257202808490032,
+      "grad_norm": 0.690170168876648,
+      "learning_rate": 1.550804571705632e-05,
+      "loss": 2.3942,
+      "step": 16427
+    },
+    {
+      "epoch": 1.3258009845855863,
+      "grad_norm": NaN,
+      "learning_rate": 1.550804571705632e-05,
+      "loss": 2.3788,
+      "step": 16428
+    },
+    {
+      "epoch": 1.3258816883221693,
+      "grad_norm": 0.6901132464408875,
+      "learning_rate": 1.5499602462307373e-05,
+      "loss": 2.3859,
+      "step": 16429
+    },
+    {
+      "epoch": 1.3259623920587522,
+      "grad_norm": 0.6639334559440613,
+      "learning_rate": 1.5491161313529223e-05,
+      "loss": 2.4271,
+      "step": 16430
+    },
+    {
+      "epoch": 1.3260430957953353,
+      "grad_norm": 0.7121936678886414,
+      "learning_rate": 1.548272227093227e-05,
+      "loss": 2.3818,
+      "step": 16431
+    },
+    {
+      "epoch": 1.3261237995319184,
+      "grad_norm": 0.6863218545913696,
+      "learning_rate": 1.5474285334726778e-05,
+      "loss": 2.3744,
+      "step": 16432
+    },
+    {
+      "epoch": 1.3262045032685013,
+      "grad_norm": 0.6697081327438354,
+      "learning_rate": 1.5465850505123057e-05,
+      "loss": 2.4001,
+      "step": 16433
+    },
+    {
+      "epoch": 1.3262852070050843,
+      "grad_norm": 0.7258912324905396,
+      "learning_rate": 1.5457417782331308e-05,
+      "loss": 2.4556,
+      "step": 16434
+    },
+    {
+      "epoch": 1.3263659107416674,
+      "grad_norm": 0.6930057406425476,
+      "learning_rate": 1.5448987166561712e-05,
+      "loss": 2.4979,
+      "step": 16435
+    },
+    {
+      "epoch": 1.3264466144782503,
+      "grad_norm": 0.6475574970245361,
+      "learning_rate": 1.5440558658024363e-05,
+      "loss": 2.3821,
+      "step": 16436
+    },
+    {
+      "epoch": 1.3265273182148334,
+      "grad_norm": 0.7489237785339355,
+      "learning_rate": 1.5432132256929367e-05,
+      "loss": 2.465,
+      "step": 16437
+    },
+    {
+      "epoch": 1.3266080219514165,
+      "grad_norm": 0.704391360282898,
+      "learning_rate": 1.5423707963486667e-05,
+      "loss": 2.433,
+      "step": 16438
+    },
+    {
+      "epoch": 1.3266887256879993,
+      "grad_norm": 0.669452965259552,
+      "learning_rate": 1.5415285777906253e-05,
+      "loss": 2.3981,
+      "step": 16439
+    },
+    {
+      "epoch": 1.3267694294245824,
+      "grad_norm": 0.6961604356765747,
+      "learning_rate": 1.540686570039802e-05,
+      "loss": 2.4684,
+      "step": 16440
+    },
+    {
+      "epoch": 1.3268501331611653,
+      "grad_norm": 0.6613924503326416,
+      "learning_rate": 1.539844773117185e-05,
+      "loss": 2.3711,
+      "step": 16441
+    },
+    {
+      "epoch": 1.3269308368977484,
+      "grad_norm": 0.7019763588905334,
+      "learning_rate": 1.5390031870437492e-05,
+      "loss": 2.3716,
+      "step": 16442
+    },
+    {
+      "epoch": 1.3270115406343312,
+      "grad_norm": 0.700176477432251,
+      "learning_rate": 1.5381618118404707e-05,
+      "loss": 2.4305,
+      "step": 16443
+    },
+    {
+      "epoch": 1.3270922443709143,
+      "grad_norm": 0.6716598272323608,
+      "learning_rate": 1.5373206475283197e-05,
+      "loss": 2.3835,
+      "step": 16444
+    },
+    {
+      "epoch": 1.3271729481074974,
+      "grad_norm": 0.6449697017669678,
+      "learning_rate": 1.53647969412826e-05,
+      "loss": 2.3707,
+      "step": 16445
+    },
+    {
+      "epoch": 1.3272536518440803,
+      "grad_norm": 0.7276685237884521,
+      "learning_rate": 1.535638951661249e-05,
+      "loss": 2.4313,
+      "step": 16446
+    },
+    {
+      "epoch": 1.3273343555806634,
+      "grad_norm": 0.7144705057144165,
+      "learning_rate": 1.5347984201482456e-05,
+      "loss": 2.4122,
+      "step": 16447
+    },
+    {
+      "epoch": 1.3274150593172465,
+      "grad_norm": 0.660225510597229,
+      "learning_rate": 1.53395809961019e-05,
+      "loss": 2.4282,
+      "step": 16448
+    },
+    {
+      "epoch": 1.3274957630538293,
+      "grad_norm": 0.7431676983833313,
+      "learning_rate": 1.5331179900680293e-05,
+      "loss": 2.3863,
+      "step": 16449
+    },
+    {
+      "epoch": 1.3275764667904124,
+      "grad_norm": 0.6670290231704712,
+      "learning_rate": 1.5322780915427036e-05,
+      "loss": 2.4266,
+      "step": 16450
+    },
+    {
+      "epoch": 1.3276571705269955,
+      "grad_norm": 0.711098313331604,
+      "learning_rate": 1.531438404055141e-05,
+      "loss": 2.4431,
+      "step": 16451
+    },
+    {
+      "epoch": 1.3277378742635784,
+      "grad_norm": 0.6908091902732849,
+      "learning_rate": 1.5305989276262688e-05,
+      "loss": 2.4153,
+      "step": 16452
+    },
+    {
+      "epoch": 1.3278185780001615,
+      "grad_norm": 0.7458107471466064,
+      "learning_rate": 1.5297596622770115e-05,
+      "loss": 2.4076,
+      "step": 16453
+    },
+    {
+      "epoch": 1.3278992817367445,
+      "grad_norm": 0.7406951189041138,
+      "learning_rate": 1.528920608028285e-05,
+      "loss": 2.3585,
+      "step": 16454
+    },
+    {
+      "epoch": 1.3279799854733274,
+      "grad_norm": 0.718824565410614,
+      "learning_rate": 1.5280817649010005e-05,
+      "loss": 2.4092,
+      "step": 16455
+    },
+    {
+      "epoch": 1.3280606892099105,
+      "grad_norm": 0.7163959741592407,
+      "learning_rate": 1.527243132916064e-05,
+      "loss": 2.4344,
+      "step": 16456
+    },
+    {
+      "epoch": 1.3281413929464934,
+      "grad_norm": 0.6695916652679443,
+      "learning_rate": 1.5264047120943793e-05,
+      "loss": 2.4144,
+      "step": 16457
+    },
+    {
+      "epoch": 1.3282220966830764,
+      "grad_norm": 0.6858509182929993,
+      "learning_rate": 1.5255665024568366e-05,
+      "loss": 2.4345,
+      "step": 16458
+    },
+    {
+      "epoch": 1.3283028004196593,
+      "grad_norm": 0.7277235388755798,
+      "learning_rate": 1.5247285040243297e-05,
+      "loss": 2.4219,
+      "step": 16459
+    },
+    {
+      "epoch": 1.3283835041562424,
+      "grad_norm": 0.6481949090957642,
+      "learning_rate": 1.5238907168177441e-05,
+      "loss": 2.4483,
+      "step": 16460
+    },
+    {
+      "epoch": 1.3284642078928255,
+      "grad_norm": 0.6956833600997925,
+      "learning_rate": 1.5230531408579574e-05,
+      "loss": 2.4241,
+      "step": 16461
+    },
+    {
+      "epoch": 1.3285449116294084,
+      "grad_norm": 0.7266185879707336,
+      "learning_rate": 1.522215776165845e-05,
+      "loss": 2.4577,
+      "step": 16462
+    },
+    {
+      "epoch": 1.3286256153659914,
+      "grad_norm": 0.725574254989624,
+      "learning_rate": 1.5213786227622773e-05,
+      "loss": 2.4451,
+      "step": 16463
+    },
+    {
+      "epoch": 1.3287063191025745,
+      "grad_norm": 0.7550850510597229,
+      "learning_rate": 1.5205416806681172e-05,
+      "loss": 2.4262,
+      "step": 16464
+    },
+    {
+      "epoch": 1.3287870228391574,
+      "grad_norm": 0.6391028761863708,
+      "learning_rate": 1.5197049499042237e-05,
+      "loss": 2.4116,
+      "step": 16465
+    },
+    {
+      "epoch": 1.3288677265757405,
+      "grad_norm": 0.6899027824401855,
+      "learning_rate": 1.5188684304914524e-05,
+      "loss": 2.3754,
+      "step": 16466
+    },
+    {
+      "epoch": 1.3289484303123236,
+      "grad_norm": 0.696681022644043,
+      "learning_rate": 1.518032122450649e-05,
+      "loss": 2.471,
+      "step": 16467
+    },
+    {
+      "epoch": 1.3290291340489064,
+      "grad_norm": 0.7090939283370972,
+      "learning_rate": 1.5171960258026551e-05,
+      "loss": 2.4153,
+      "step": 16468
+    },
+    {
+      "epoch": 1.3291098377854895,
+      "grad_norm": 0.7125746607780457,
+      "learning_rate": 1.5163601405683148e-05,
+      "loss": 2.4102,
+      "step": 16469
+    },
+    {
+      "epoch": 1.3291905415220726,
+      "grad_norm": 0.7407518029212952,
+      "learning_rate": 1.5155244667684531e-05,
+      "loss": 2.429,
+      "step": 16470
+    },
+    {
+      "epoch": 1.3292712452586555,
+      "grad_norm": 0.7401885390281677,
+      "learning_rate": 1.5146890044239004e-05,
+      "loss": 2.4577,
+      "step": 16471
+    },
+    {
+      "epoch": 1.3293519489952383,
+      "grad_norm": 0.7625757455825806,
+      "learning_rate": 1.5138537535554786e-05,
+      "loss": 2.3813,
+      "step": 16472
+    },
+    {
+      "epoch": 1.3294326527318214,
+      "grad_norm": 0.7423396706581116,
+      "learning_rate": 1.5130187141840057e-05,
+      "loss": 2.3797,
+      "step": 16473
+    },
+    {
+      "epoch": 1.3295133564684045,
+      "grad_norm": 0.7029228806495667,
+      "learning_rate": 1.5121838863302884e-05,
+      "loss": 2.4203,
+      "step": 16474
+    },
+    {
+      "epoch": 1.3295940602049874,
+      "grad_norm": 0.8062863349914551,
+      "learning_rate": 1.5113492700151378e-05,
+      "loss": 2.3743,
+      "step": 16475
+    },
+    {
+      "epoch": 1.3296747639415705,
+      "grad_norm": 0.7113343477249146,
+      "learning_rate": 1.5105148652593548e-05,
+      "loss": 2.3837,
+      "step": 16476
+    },
+    {
+      "epoch": 1.3297554676781536,
+      "grad_norm": 0.6733126044273376,
+      "learning_rate": 1.5096806720837309e-05,
+      "loss": 2.4677,
+      "step": 16477
+    },
+    {
+      "epoch": 1.3298361714147364,
+      "grad_norm": 0.6936657428741455,
+      "learning_rate": 1.5088466905090593e-05,
+      "loss": 2.3677,
+      "step": 16478
+    },
+    {
+      "epoch": 1.3299168751513195,
+      "grad_norm": 0.746746301651001,
+      "learning_rate": 1.5080129205561255e-05,
+      "loss": 2.423,
+      "step": 16479
+    },
+    {
+      "epoch": 1.3299975788879026,
+      "grad_norm": 0.6879116296768188,
+      "learning_rate": 1.5071793622457065e-05,
+      "loss": 2.4867,
+      "step": 16480
+    },
+    {
+      "epoch": 1.3300782826244855,
+      "grad_norm": 0.6841214299201965,
+      "learning_rate": 1.5063460155985776e-05,
+      "loss": 2.5015,
+      "step": 16481
+    },
+    {
+      "epoch": 1.3301589863610686,
+      "grad_norm": 0.6955111622810364,
+      "learning_rate": 1.5055128806355123e-05,
+      "loss": 2.3975,
+      "step": 16482
+    },
+    {
+      "epoch": 1.3302396900976516,
+      "grad_norm": 0.7084987163543701,
+      "learning_rate": 1.5046799573772673e-05,
+      "loss": 2.4511,
+      "step": 16483
+    },
+    {
+      "epoch": 1.3303203938342345,
+      "grad_norm": 0.6905840039253235,
+      "learning_rate": 1.5038472458446051e-05,
+      "loss": 2.3542,
+      "step": 16484
+    },
+    {
+      "epoch": 1.3304010975708176,
+      "grad_norm": 0.7182672023773193,
+      "learning_rate": 1.5030147460582788e-05,
+      "loss": 2.3673,
+      "step": 16485
+    },
+    {
+      "epoch": 1.3304818013074005,
+      "grad_norm": 0.6805183291435242,
+      "learning_rate": 1.5021824580390353e-05,
+      "loss": 2.3751,
+      "step": 16486
+    },
+    {
+      "epoch": 1.3305625050439835,
+      "grad_norm": 0.6278836727142334,
+      "learning_rate": 1.5013503818076202e-05,
+      "loss": 2.3508,
+      "step": 16487
+    },
+    {
+      "epoch": 1.3306432087805664,
+      "grad_norm": 0.664000391960144,
+      "learning_rate": 1.500518517384768e-05,
+      "loss": 2.4039,
+      "step": 16488
+    },
+    {
+      "epoch": 1.3307239125171495,
+      "grad_norm": 0.6906681060791016,
+      "learning_rate": 1.4996868647912155e-05,
+      "loss": 2.4068,
+      "step": 16489
+    },
+    {
+      "epoch": 1.3308046162537326,
+      "grad_norm": 0.6756102442741394,
+      "learning_rate": 1.4988554240476826e-05,
+      "loss": 2.4423,
+      "step": 16490
+    },
+    {
+      "epoch": 1.3308853199903155,
+      "grad_norm": 0.7013095021247864,
+      "learning_rate": 1.4980241951748964e-05,
+      "loss": 2.3536,
+      "step": 16491
+    },
+    {
+      "epoch": 1.3309660237268985,
+      "grad_norm": 0.6689851880073547,
+      "learning_rate": 1.4971931781935732e-05,
+      "loss": 2.4192,
+      "step": 16492
+    },
+    {
+      "epoch": 1.3310467274634816,
+      "grad_norm": 0.6411572694778442,
+      "learning_rate": 1.4963623731244202e-05,
+      "loss": 2.4012,
+      "step": 16493
+    },
+    {
+      "epoch": 1.3311274312000645,
+      "grad_norm": 0.7209812998771667,
+      "learning_rate": 1.4955317799881453e-05,
+      "loss": 2.378,
+      "step": 16494
+    },
+    {
+      "epoch": 1.3312081349366476,
+      "grad_norm": 0.7041119933128357,
+      "learning_rate": 1.4947013988054504e-05,
+      "loss": 2.4047,
+      "step": 16495
+    },
+    {
+      "epoch": 1.3312888386732307,
+      "grad_norm": 0.6928852796554565,
+      "learning_rate": 1.4938712295970292e-05,
+      "loss": 2.4489,
+      "step": 16496
+    },
+    {
+      "epoch": 1.3313695424098135,
+      "grad_norm": 0.6923524141311646,
+      "learning_rate": 1.4930412723835718e-05,
+      "loss": 2.3752,
+      "step": 16497
+    },
+    {
+      "epoch": 1.3314502461463966,
+      "grad_norm": 0.7034686803817749,
+      "learning_rate": 1.4922115271857662e-05,
+      "loss": 2.3898,
+      "step": 16498
+    },
+    {
+      "epoch": 1.3315309498829797,
+      "grad_norm": 0.6717320084571838,
+      "learning_rate": 1.4913819940242856e-05,
+      "loss": 2.3629,
+      "step": 16499
+    },
+    {
+      "epoch": 1.3316116536195626,
+      "grad_norm": 0.6885079741477966,
+      "learning_rate": 1.4905526729198083e-05,
+      "loss": 2.4321,
+      "step": 16500
+    },
+    {
+      "epoch": 1.3316923573561457,
+      "grad_norm": 0.662452757358551,
+      "learning_rate": 1.489723563893004e-05,
+      "loss": 2.4532,
+      "step": 16501
+    },
+    {
+      "epoch": 1.3317730610927285,
+      "grad_norm": 0.6650903224945068,
+      "learning_rate": 1.4888946669645332e-05,
+      "loss": 2.4347,
+      "step": 16502
+    },
+    {
+      "epoch": 1.3318537648293116,
+      "grad_norm": 0.7217590808868408,
+      "learning_rate": 1.4880659821550546e-05,
+      "loss": 2.4641,
+      "step": 16503
+    },
+    {
+      "epoch": 1.3319344685658945,
+      "grad_norm": 0.7063763737678528,
+      "learning_rate": 1.4872375094852232e-05,
+      "loss": 2.4365,
+      "step": 16504
+    },
+    {
+      "epoch": 1.3320151723024776,
+      "grad_norm": 0.7366454005241394,
+      "learning_rate": 1.4864092489756853e-05,
+      "loss": 2.4223,
+      "step": 16505
+    },
+    {
+      "epoch": 1.3320958760390607,
+      "grad_norm": 0.7132206559181213,
+      "learning_rate": 1.4855812006470838e-05,
+      "loss": 2.4404,
+      "step": 16506
+    },
+    {
+      "epoch": 1.3321765797756435,
+      "grad_norm": 0.665553867816925,
+      "learning_rate": 1.484753364520055e-05,
+      "loss": 2.3818,
+      "step": 16507
+    },
+    {
+      "epoch": 1.3322572835122266,
+      "grad_norm": 0.7854028344154358,
+      "learning_rate": 1.483925740615234e-05,
+      "loss": 2.4111,
+      "step": 16508
+    },
+    {
+      "epoch": 1.3323379872488097,
+      "grad_norm": 0.7331317663192749,
+      "learning_rate": 1.4830983289532418e-05,
+      "loss": 2.4446,
+      "step": 16509
+    },
+    {
+      "epoch": 1.3324186909853926,
+      "grad_norm": 0.670315146446228,
+      "learning_rate": 1.4822711295547042e-05,
+      "loss": 2.4017,
+      "step": 16510
+    },
+    {
+      "epoch": 1.3324993947219756,
+      "grad_norm": 0.7242144346237183,
+      "learning_rate": 1.481444142440237e-05,
+      "loss": 2.4281,
+      "step": 16511
+    },
+    {
+      "epoch": 1.3325800984585587,
+      "grad_norm": 0.7108538746833801,
+      "learning_rate": 1.4806173676304468e-05,
+      "loss": 2.4331,
+      "step": 16512
+    },
+    {
+      "epoch": 1.3326608021951416,
+      "grad_norm": 0.658989667892456,
+      "learning_rate": 1.479790805145943e-05,
+      "loss": 2.4321,
+      "step": 16513
+    },
+    {
+      "epoch": 1.3327415059317247,
+      "grad_norm": 0.6596404314041138,
+      "learning_rate": 1.4789644550073233e-05,
+      "loss": 2.3817,
+      "step": 16514
+    },
+    {
+      "epoch": 1.3328222096683078,
+      "grad_norm": 0.6922028064727783,
+      "learning_rate": 1.4781383172351837e-05,
+      "loss": 2.399,
+      "step": 16515
+    },
+    {
+      "epoch": 1.3329029134048906,
+      "grad_norm": 0.750747799873352,
+      "learning_rate": 1.4773123918501141e-05,
+      "loss": 2.4502,
+      "step": 16516
+    },
+    {
+      "epoch": 1.3329836171414737,
+      "grad_norm": 0.6887632608413696,
+      "learning_rate": 1.4764866788727006e-05,
+      "loss": 2.3636,
+      "step": 16517
+    },
+    {
+      "epoch": 1.3330643208780566,
+      "grad_norm": 0.6751166582107544,
+      "learning_rate": 1.4756611783235163e-05,
+      "loss": 2.3956,
+      "step": 16518
+    },
+    {
+      "epoch": 1.3331450246146397,
+      "grad_norm": 0.679040253162384,
+      "learning_rate": 1.4748358902231395e-05,
+      "loss": 2.4044,
+      "step": 16519
+    },
+    {
+      "epoch": 1.3332257283512225,
+      "grad_norm": 0.6396780610084534,
+      "learning_rate": 1.4740108145921373e-05,
+      "loss": 2.4114,
+      "step": 16520
+    },
+    {
+      "epoch": 1.3333064320878056,
+      "grad_norm": 0.6686230301856995,
+      "learning_rate": 1.4731859514510738e-05,
+      "loss": 2.4535,
+      "step": 16521
+    },
+    {
+      "epoch": 1.3333871358243887,
+      "grad_norm": 0.6693681478500366,
+      "learning_rate": 1.472361300820505e-05,
+      "loss": 2.3885,
+      "step": 16522
+    },
+    {
+      "epoch": 1.3334678395609716,
+      "grad_norm": 0.7700718641281128,
+      "learning_rate": 1.4715368627209836e-05,
+      "loss": 2.3939,
+      "step": 16523
+    },
+    {
+      "epoch": 1.3335485432975547,
+      "grad_norm": 0.7203121781349182,
+      "learning_rate": 1.4707126371730561e-05,
+      "loss": 2.4644,
+      "step": 16524
+    },
+    {
+      "epoch": 1.3336292470341378,
+      "grad_norm": 0.7798308730125427,
+      "learning_rate": 1.4698886241972665e-05,
+      "loss": 2.4293,
+      "step": 16525
+    },
+    {
+      "epoch": 1.3337099507707206,
+      "grad_norm": 0.7017160654067993,
+      "learning_rate": 1.4690648238141503e-05,
+      "loss": 2.4327,
+      "step": 16526
+    },
+    {
+      "epoch": 1.3337906545073037,
+      "grad_norm": 0.6522603631019592,
+      "learning_rate": 1.468241236044241e-05,
+      "loss": 2.3955,
+      "step": 16527
+    },
+    {
+      "epoch": 1.3338713582438868,
+      "grad_norm": 0.766222357749939,
+      "learning_rate": 1.4674178609080602e-05,
+      "loss": 2.4652,
+      "step": 16528
+    },
+    {
+      "epoch": 1.3339520619804697,
+      "grad_norm": 0.7351565361022949,
+      "learning_rate": 1.4665946984261303e-05,
+      "loss": 2.4607,
+      "step": 16529
+    },
+    {
+      "epoch": 1.3340327657170528,
+      "grad_norm": 0.6817728281021118,
+      "learning_rate": 1.4657717486189693e-05,
+      "loss": 2.3687,
+      "step": 16530
+    },
+    {
+      "epoch": 1.3341134694536356,
+      "grad_norm": 0.7401643395423889,
+      "learning_rate": 1.464949011507083e-05,
+      "loss": 2.4179,
+      "step": 16531
+    },
+    {
+      "epoch": 1.3341941731902187,
+      "grad_norm": 0.7783530354499817,
+      "learning_rate": 1.4641264871109784e-05,
+      "loss": 2.4088,
+      "step": 16532
+    },
+    {
+      "epoch": 1.3342748769268016,
+      "grad_norm": 0.6761943697929382,
+      "learning_rate": 1.4633041754511534e-05,
+      "loss": 2.4141,
+      "step": 16533
+    },
+    {
+      "epoch": 1.3343555806633847,
+      "grad_norm": 0.6842260360717773,
+      "learning_rate": 1.4624820765481073e-05,
+      "loss": 2.4918,
+      "step": 16534
+    },
+    {
+      "epoch": 1.3344362843999678,
+      "grad_norm": 0.6906094551086426,
+      "learning_rate": 1.4616601904223225e-05,
+      "loss": 2.4576,
+      "step": 16535
+    },
+    {
+      "epoch": 1.3345169881365506,
+      "grad_norm": 0.6549125909805298,
+      "learning_rate": 1.4608385170942829e-05,
+      "loss": 2.3748,
+      "step": 16536
+    },
+    {
+      "epoch": 1.3345976918731337,
+      "grad_norm": 0.6603896617889404,
+      "learning_rate": 1.4600170565844728e-05,
+      "loss": 2.3739,
+      "step": 16537
+    },
+    {
+      "epoch": 1.3346783956097168,
+      "grad_norm": 0.6413096189498901,
+      "learning_rate": 1.4591958089133606e-05,
+      "loss": 2.3979,
+      "step": 16538
+    },
+    {
+      "epoch": 1.3347590993462997,
+      "grad_norm": 0.7085204720497131,
+      "learning_rate": 1.4583747741014142e-05,
+      "loss": 2.4185,
+      "step": 16539
+    },
+    {
+      "epoch": 1.3348398030828827,
+      "grad_norm": 0.6517937183380127,
+      "learning_rate": 1.4575539521690983e-05,
+      "loss": 2.3938,
+      "step": 16540
+    },
+    {
+      "epoch": 1.3349205068194658,
+      "grad_norm": 0.6326449513435364,
+      "learning_rate": 1.4567333431368658e-05,
+      "loss": 2.4613,
+      "step": 16541
+    },
+    {
+      "epoch": 1.3350012105560487,
+      "grad_norm": 0.8046317100524902,
+      "learning_rate": 1.4559129470251708e-05,
+      "loss": 2.4547,
+      "step": 16542
+    },
+    {
+      "epoch": 1.3350819142926318,
+      "grad_norm": 0.6661570072174072,
+      "learning_rate": 1.455092763854462e-05,
+      "loss": 2.3636,
+      "step": 16543
+    },
+    {
+      "epoch": 1.3351626180292149,
+      "grad_norm": 0.6806541085243225,
+      "learning_rate": 1.454272793645176e-05,
+      "loss": 2.4309,
+      "step": 16544
+    },
+    {
+      "epoch": 1.3352433217657977,
+      "grad_norm": 0.651836097240448,
+      "learning_rate": 1.45345303641775e-05,
+      "loss": 2.3862,
+      "step": 16545
+    },
+    {
+      "epoch": 1.3353240255023808,
+      "grad_norm": 0.7448983192443848,
+      "learning_rate": 1.4526334921926165e-05,
+      "loss": 2.4654,
+      "step": 16546
+    },
+    {
+      "epoch": 1.3354047292389637,
+      "grad_norm": 0.6885285973548889,
+      "learning_rate": 1.4518141609901992e-05,
+      "loss": 2.3943,
+      "step": 16547
+    },
+    {
+      "epoch": 1.3354854329755468,
+      "grad_norm": 0.7204004526138306,
+      "learning_rate": 1.450995042830917e-05,
+      "loss": 2.4117,
+      "step": 16548
+    },
+    {
+      "epoch": 1.3355661367121296,
+      "grad_norm": 0.6551961898803711,
+      "learning_rate": 1.4501761377351864e-05,
+      "loss": 2.4269,
+      "step": 16549
+    },
+    {
+      "epoch": 1.3356468404487127,
+      "grad_norm": 0.7191253304481506,
+      "learning_rate": 1.4493574457234182e-05,
+      "loss": 2.3472,
+      "step": 16550
+    },
+    {
+      "epoch": 1.3357275441852958,
+      "grad_norm": 0.6793580651283264,
+      "learning_rate": 1.4485389668160121e-05,
+      "loss": 2.4264,
+      "step": 16551
+    },
+    {
+      "epoch": 1.3358082479218787,
+      "grad_norm": 0.704250693321228,
+      "learning_rate": 1.4477207010333682e-05,
+      "loss": 2.5236,
+      "step": 16552
+    },
+    {
+      "epoch": 1.3358889516584618,
+      "grad_norm": 0.6826470494270325,
+      "learning_rate": 1.4469026483958837e-05,
+      "loss": 2.4473,
+      "step": 16553
+    },
+    {
+      "epoch": 1.3359696553950449,
+      "grad_norm": 0.6646167039871216,
+      "learning_rate": 1.4460848089239399e-05,
+      "loss": 2.4232,
+      "step": 16554
+    },
+    {
+      "epoch": 1.3360503591316277,
+      "grad_norm": 0.7604451179504395,
+      "learning_rate": 1.4452671826379227e-05,
+      "loss": 2.4208,
+      "step": 16555
+    },
+    {
+      "epoch": 1.3361310628682108,
+      "grad_norm": 0.7129300236701965,
+      "learning_rate": 1.4444497695582093e-05,
+      "loss": 2.4304,
+      "step": 16556
+    },
+    {
+      "epoch": 1.336211766604794,
+      "grad_norm": 0.6769927740097046,
+      "learning_rate": 1.4436325697051733e-05,
+      "loss": 2.3467,
+      "step": 16557
+    },
+    {
+      "epoch": 1.3362924703413768,
+      "grad_norm": 0.6568608283996582,
+      "learning_rate": 1.4428155830991797e-05,
+      "loss": 2.4285,
+      "step": 16558
+    },
+    {
+      "epoch": 1.3363731740779599,
+      "grad_norm": 0.7687276005744934,
+      "learning_rate": 1.4419988097605919e-05,
+      "loss": 2.4815,
+      "step": 16559
+    },
+    {
+      "epoch": 1.336453877814543,
+      "grad_norm": 0.7001463770866394,
+      "learning_rate": 1.4411822497097638e-05,
+      "loss": 2.4629,
+      "step": 16560
+    },
+    {
+      "epoch": 1.3365345815511258,
+      "grad_norm": 0.7211995720863342,
+      "learning_rate": 1.4403659029670458e-05,
+      "loss": 2.4323,
+      "step": 16561
+    },
+    {
+      "epoch": 1.336615285287709,
+      "grad_norm": 0.7371769547462463,
+      "learning_rate": 1.439549769552787e-05,
+      "loss": 2.3962,
+      "step": 16562
+    },
+    {
+      "epoch": 1.3366959890242918,
+      "grad_norm": 0.7475463151931763,
+      "learning_rate": 1.4387338494873237e-05,
+      "loss": 2.3593,
+      "step": 16563
+    },
+    {
+      "epoch": 1.3367766927608749,
+      "grad_norm": 0.7215834856033325,
+      "learning_rate": 1.4379181427909916e-05,
+      "loss": 2.3687,
+      "step": 16564
+    },
+    {
+      "epoch": 1.3368573964974577,
+      "grad_norm": 0.7160200476646423,
+      "learning_rate": 1.4371026494841211e-05,
+      "loss": 2.3652,
+      "step": 16565
+    },
+    {
+      "epoch": 1.3369381002340408,
+      "grad_norm": 0.6636231541633606,
+      "learning_rate": 1.436287369587036e-05,
+      "loss": 2.4628,
+      "step": 16566
+    },
+    {
+      "epoch": 1.337018803970624,
+      "grad_norm": 0.657774806022644,
+      "learning_rate": 1.4354723031200556e-05,
+      "loss": 2.4082,
+      "step": 16567
+    },
+    {
+      "epoch": 1.3370995077072068,
+      "grad_norm": 0.7020300626754761,
+      "learning_rate": 1.4346574501034936e-05,
+      "loss": 2.3821,
+      "step": 16568
+    },
+    {
+      "epoch": 1.3371802114437898,
+      "grad_norm": 0.6800786256790161,
+      "learning_rate": 1.4338428105576595e-05,
+      "loss": 2.3839,
+      "step": 16569
+    },
+    {
+      "epoch": 1.337260915180373,
+      "grad_norm": 0.7176932692527771,
+      "learning_rate": 1.4330283845028536e-05,
+      "loss": 2.4614,
+      "step": 16570
+    },
+    {
+      "epoch": 1.3373416189169558,
+      "grad_norm": 0.7233355641365051,
+      "learning_rate": 1.432214171959374e-05,
+      "loss": 2.4048,
+      "step": 16571
+    },
+    {
+      "epoch": 1.3374223226535389,
+      "grad_norm": 0.7721874117851257,
+      "learning_rate": 1.4314001729475157e-05,
+      "loss": 2.4169,
+      "step": 16572
+    },
+    {
+      "epoch": 1.337503026390122,
+      "grad_norm": 0.7123380303382874,
+      "learning_rate": 1.4305863874875613e-05,
+      "loss": 2.3799,
+      "step": 16573
+    },
+    {
+      "epoch": 1.3375837301267048,
+      "grad_norm": 0.7297765016555786,
+      "learning_rate": 1.4297728155997958e-05,
+      "loss": 2.4655,
+      "step": 16574
+    },
+    {
+      "epoch": 1.337664433863288,
+      "grad_norm": 0.6806401610374451,
+      "learning_rate": 1.428959457304493e-05,
+      "loss": 2.4102,
+      "step": 16575
+    },
+    {
+      "epoch": 1.3377451375998708,
+      "grad_norm": 0.6811275482177734,
+      "learning_rate": 1.4281463126219264e-05,
+      "loss": 2.4298,
+      "step": 16576
+    },
+    {
+      "epoch": 1.3378258413364539,
+      "grad_norm": 0.6900678277015686,
+      "learning_rate": 1.427333381572361e-05,
+      "loss": 2.4745,
+      "step": 16577
+    },
+    {
+      "epoch": 1.3379065450730367,
+      "grad_norm": 0.7815307974815369,
+      "learning_rate": 1.4265206641760587e-05,
+      "loss": 2.3624,
+      "step": 16578
+    },
+    {
+      "epoch": 1.3379872488096198,
+      "grad_norm": 0.6948800683021545,
+      "learning_rate": 1.4257081604532708e-05,
+      "loss": 2.4142,
+      "step": 16579
+    },
+    {
+      "epoch": 1.338067952546203,
+      "grad_norm": 0.7387657165527344,
+      "learning_rate": 1.4248958704242488e-05,
+      "loss": 2.4241,
+      "step": 16580
+    },
+    {
+      "epoch": 1.3381486562827858,
+      "grad_norm": 0.7158597111701965,
+      "learning_rate": 1.4240837941092367e-05,
+      "loss": 2.4473,
+      "step": 16581
+    },
+    {
+      "epoch": 1.3382293600193689,
+      "grad_norm": 0.758674144744873,
+      "learning_rate": 1.423271931528477e-05,
+      "loss": 2.4504,
+      "step": 16582
+    },
+    {
+      "epoch": 1.338310063755952,
+      "grad_norm": 0.6904417872428894,
+      "learning_rate": 1.4224602827021982e-05,
+      "loss": 2.4288,
+      "step": 16583
+    },
+    {
+      "epoch": 1.3383907674925348,
+      "grad_norm": 0.6988760828971863,
+      "learning_rate": 1.4216488476506307e-05,
+      "loss": 2.3874,
+      "step": 16584
+    },
+    {
+      "epoch": 1.338471471229118,
+      "grad_norm": 0.6969872117042542,
+      "learning_rate": 1.4208376263940003e-05,
+      "loss": 2.3388,
+      "step": 16585
+    },
+    {
+      "epoch": 1.338552174965701,
+      "grad_norm": 0.687179684638977,
+      "learning_rate": 1.420026618952518e-05,
+      "loss": 2.431,
+      "step": 16586
+    },
+    {
+      "epoch": 1.3386328787022839,
+      "grad_norm": 0.6319810152053833,
+      "learning_rate": 1.4192158253464038e-05,
+      "loss": 2.4415,
+      "step": 16587
+    },
+    {
+      "epoch": 1.338713582438867,
+      "grad_norm": 0.7554977536201477,
+      "learning_rate": 1.4184052455958629e-05,
+      "loss": 2.3863,
+      "step": 16588
+    },
+    {
+      "epoch": 1.33879428617545,
+      "grad_norm": 0.7025974988937378,
+      "learning_rate": 1.4175948797210936e-05,
+      "loss": 2.3957,
+      "step": 16589
+    },
+    {
+      "epoch": 1.338874989912033,
+      "grad_norm": 0.7270370721817017,
+      "learning_rate": 1.4167847277422952e-05,
+      "loss": 2.4309,
+      "step": 16590
+    },
+    {
+      "epoch": 1.338955693648616,
+      "grad_norm": 0.7017608284950256,
+      "learning_rate": 1.4159747896796593e-05,
+      "loss": 2.4142,
+      "step": 16591
+    },
+    {
+      "epoch": 1.3390363973851989,
+      "grad_norm": 0.7114055156707764,
+      "learning_rate": 1.4151650655533687e-05,
+      "loss": 2.473,
+      "step": 16592
+    },
+    {
+      "epoch": 1.339117101121782,
+      "grad_norm": 0.6420357823371887,
+      "learning_rate": 1.4143555553836063e-05,
+      "loss": 2.3671,
+      "step": 16593
+    },
+    {
+      "epoch": 1.3391978048583648,
+      "grad_norm": 0.7067350745201111,
+      "learning_rate": 1.413546259190548e-05,
+      "loss": 2.4422,
+      "step": 16594
+    },
+    {
+      "epoch": 1.339278508594948,
+      "grad_norm": 0.7376763224601746,
+      "learning_rate": 1.4127371769943598e-05,
+      "loss": 2.4443,
+      "step": 16595
+    },
+    {
+      "epoch": 1.339359212331531,
+      "grad_norm": 0.646515965461731,
+      "learning_rate": 1.4119283088152092e-05,
+      "loss": 2.3949,
+      "step": 16596
+    },
+    {
+      "epoch": 1.3394399160681139,
+      "grad_norm": 0.6896061301231384,
+      "learning_rate": 1.411119654673254e-05,
+      "loss": 2.4535,
+      "step": 16597
+    },
+    {
+      "epoch": 1.339520619804697,
+      "grad_norm": 0.6992611289024353,
+      "learning_rate": 1.4103112145886489e-05,
+      "loss": 2.3983,
+      "step": 16598
+    },
+    {
+      "epoch": 1.33960132354128,
+      "grad_norm": 0.7176348567008972,
+      "learning_rate": 1.4095029885815426e-05,
+      "loss": 2.4671,
+      "step": 16599
+    },
+    {
+      "epoch": 1.339682027277863,
+      "grad_norm": 0.6635856628417969,
+      "learning_rate": 1.4086949766720759e-05,
+      "loss": 2.4235,
+      "step": 16600
+    },
+    {
+      "epoch": 1.339762731014446,
+      "grad_norm": 0.673332154750824,
+      "learning_rate": 1.4078871788803915e-05,
+      "loss": 2.4328,
+      "step": 16601
+    },
+    {
+      "epoch": 1.339843434751029,
+      "grad_norm": 0.6738821864128113,
+      "learning_rate": 1.407079595226617e-05,
+      "loss": 2.4786,
+      "step": 16602
+    },
+    {
+      "epoch": 1.339924138487612,
+      "grad_norm": 0.690605103969574,
+      "learning_rate": 1.4062722257308803e-05,
+      "loss": 2.4025,
+      "step": 16603
+    },
+    {
+      "epoch": 1.340004842224195,
+      "grad_norm": 0.7186758518218994,
+      "learning_rate": 1.4054650704133066e-05,
+      "loss": 2.4793,
+      "step": 16604
+    },
+    {
+      "epoch": 1.3400855459607781,
+      "grad_norm": 0.6484951376914978,
+      "learning_rate": 1.4046581292940075e-05,
+      "loss": 2.3855,
+      "step": 16605
+    },
+    {
+      "epoch": 1.340166249697361,
+      "grad_norm": 0.6993771195411682,
+      "learning_rate": 1.403851402393096e-05,
+      "loss": 2.3872,
+      "step": 16606
+    },
+    {
+      "epoch": 1.340246953433944,
+      "grad_norm": 0.7446531653404236,
+      "learning_rate": 1.403044889730678e-05,
+      "loss": 2.4253,
+      "step": 16607
+    },
+    {
+      "epoch": 1.340327657170527,
+      "grad_norm": 0.6873160004615784,
+      "learning_rate": 1.4022385913268542e-05,
+      "loss": 2.464,
+      "step": 16608
+    },
+    {
+      "epoch": 1.34040836090711,
+      "grad_norm": 0.6570948362350464,
+      "learning_rate": 1.4014325072017198e-05,
+      "loss": 2.4063,
+      "step": 16609
+    },
+    {
+      "epoch": 1.3404890646436929,
+      "grad_norm": 0.7209224104881287,
+      "learning_rate": 1.4006266373753651e-05,
+      "loss": 2.4827,
+      "step": 16610
+    },
+    {
+      "epoch": 1.340569768380276,
+      "grad_norm": 0.7283413410186768,
+      "learning_rate": 1.3998209818678732e-05,
+      "loss": 2.4009,
+      "step": 16611
+    },
+    {
+      "epoch": 1.340650472116859,
+      "grad_norm": 0.6650960445404053,
+      "learning_rate": 1.3990155406993221e-05,
+      "loss": 2.3576,
+      "step": 16612
+    },
+    {
+      "epoch": 1.340731175853442,
+      "grad_norm": 0.6857860088348389,
+      "learning_rate": 1.3982103138897873e-05,
+      "loss": 2.4686,
+      "step": 16613
+    },
+    {
+      "epoch": 1.340811879590025,
+      "grad_norm": 0.7065873146057129,
+      "learning_rate": 1.3974053014593402e-05,
+      "loss": 2.3999,
+      "step": 16614
+    },
+    {
+      "epoch": 1.340892583326608,
+      "grad_norm": 0.8093010783195496,
+      "learning_rate": 1.3966005034280372e-05,
+      "loss": 2.4273,
+      "step": 16615
+    },
+    {
+      "epoch": 1.340973287063191,
+      "grad_norm": 0.649132251739502,
+      "learning_rate": 1.3957959198159387e-05,
+      "loss": 2.3418,
+      "step": 16616
+    },
+    {
+      "epoch": 1.341053990799774,
+      "grad_norm": 0.7114978432655334,
+      "learning_rate": 1.3949915506430976e-05,
+      "loss": 2.4393,
+      "step": 16617
+    },
+    {
+      "epoch": 1.3411346945363571,
+      "grad_norm": 0.7989282608032227,
+      "learning_rate": 1.3941873959295615e-05,
+      "loss": 2.4044,
+      "step": 16618
+    },
+    {
+      "epoch": 1.34121539827294,
+      "grad_norm": 0.7373676896095276,
+      "learning_rate": 1.3933834556953707e-05,
+      "loss": 2.4758,
+      "step": 16619
+    },
+    {
+      "epoch": 1.341296102009523,
+      "grad_norm": 0.7076435089111328,
+      "learning_rate": 1.3925797299605647e-05,
+      "loss": 2.4429,
+      "step": 16620
+    },
+    {
+      "epoch": 1.3413768057461062,
+      "grad_norm": 0.6739028692245483,
+      "learning_rate": 1.39177621874517e-05,
+      "loss": 2.4275,
+      "step": 16621
+    },
+    {
+      "epoch": 1.341457509482689,
+      "grad_norm": 0.7134198546409607,
+      "learning_rate": 1.3909729220692125e-05,
+      "loss": 2.4541,
+      "step": 16622
+    },
+    {
+      "epoch": 1.3415382132192721,
+      "grad_norm": 0.6770301461219788,
+      "learning_rate": 1.3901698399527175e-05,
+      "loss": 2.4143,
+      "step": 16623
+    },
+    {
+      "epoch": 1.341618916955855,
+      "grad_norm": 0.7146373987197876,
+      "learning_rate": 1.3893669724156943e-05,
+      "loss": 2.4886,
+      "step": 16624
+    },
+    {
+      "epoch": 1.341699620692438,
+      "grad_norm": 0.6801536083221436,
+      "learning_rate": 1.3885643194781539e-05,
+      "loss": 2.4154,
+      "step": 16625
+    },
+    {
+      "epoch": 1.341780324429021,
+      "grad_norm": 0.7350363731384277,
+      "learning_rate": 1.3877618811601024e-05,
+      "loss": 2.3918,
+      "step": 16626
+    },
+    {
+      "epoch": 1.341861028165604,
+      "grad_norm": 0.7088882327079773,
+      "learning_rate": 1.3869596574815358e-05,
+      "loss": 2.412,
+      "step": 16627
+    },
+    {
+      "epoch": 1.3419417319021871,
+      "grad_norm": 0.7199791669845581,
+      "learning_rate": 1.3861576484624506e-05,
+      "loss": 2.3912,
+      "step": 16628
+    },
+    {
+      "epoch": 1.34202243563877,
+      "grad_norm": 0.692971408367157,
+      "learning_rate": 1.3853558541228328e-05,
+      "loss": 2.3826,
+      "step": 16629
+    },
+    {
+      "epoch": 1.342103139375353,
+      "grad_norm": 0.7524722814559937,
+      "learning_rate": 1.3845542744826679e-05,
+      "loss": 2.4227,
+      "step": 16630
+    },
+    {
+      "epoch": 1.3421838431119362,
+      "grad_norm": 0.6624585390090942,
+      "learning_rate": 1.3837529095619307e-05,
+      "loss": 2.3649,
+      "step": 16631
+    },
+    {
+      "epoch": 1.342264546848519,
+      "grad_norm": 0.6884489059448242,
+      "learning_rate": 1.3829517593805929e-05,
+      "loss": 2.3687,
+      "step": 16632
+    },
+    {
+      "epoch": 1.3423452505851021,
+      "grad_norm": 0.6766197085380554,
+      "learning_rate": 1.3821508239586246e-05,
+      "loss": 2.4191,
+      "step": 16633
+    },
+    {
+      "epoch": 1.3424259543216852,
+      "grad_norm": 0.6744453310966492,
+      "learning_rate": 1.3813501033159837e-05,
+      "loss": 2.4254,
+      "step": 16634
+    },
+    {
+      "epoch": 1.342506658058268,
+      "grad_norm": 0.6906216144561768,
+      "learning_rate": 1.3805495974726267e-05,
+      "loss": 2.4763,
+      "step": 16635
+    },
+    {
+      "epoch": 1.3425873617948512,
+      "grad_norm": 0.7052608132362366,
+      "learning_rate": 1.3797493064485078e-05,
+      "loss": 2.4307,
+      "step": 16636
+    },
+    {
+      "epoch": 1.342668065531434,
+      "grad_norm": 0.6701127290725708,
+      "learning_rate": 1.3789492302635653e-05,
+      "loss": 2.4529,
+      "step": 16637
+    },
+    {
+      "epoch": 1.3427487692680171,
+      "grad_norm": 0.7440397143363953,
+      "learning_rate": 1.3781493689377455e-05,
+      "loss": 2.4471,
+      "step": 16638
+    },
+    {
+      "epoch": 1.3428294730046,
+      "grad_norm": 0.7340207695960999,
+      "learning_rate": 1.3773497224909848e-05,
+      "loss": 2.4434,
+      "step": 16639
+    },
+    {
+      "epoch": 1.342910176741183,
+      "grad_norm": 0.6836793422698975,
+      "learning_rate": 1.376550290943205e-05,
+      "loss": 2.4072,
+      "step": 16640
+    },
+    {
+      "epoch": 1.3429908804777662,
+      "grad_norm": 0.6820472478866577,
+      "learning_rate": 1.3757510743143342e-05,
+      "loss": 2.4078,
+      "step": 16641
+    },
+    {
+      "epoch": 1.343071584214349,
+      "grad_norm": 0.6608061194419861,
+      "learning_rate": 1.3749520726242938e-05,
+      "loss": 2.3995,
+      "step": 16642
+    },
+    {
+      "epoch": 1.3431522879509321,
+      "grad_norm": 0.6582421064376831,
+      "learning_rate": 1.3741532858929906e-05,
+      "loss": 2.3768,
+      "step": 16643
+    },
+    {
+      "epoch": 1.3432329916875152,
+      "grad_norm": 0.7032744288444519,
+      "learning_rate": 1.3733547141403358e-05,
+      "loss": 2.4367,
+      "step": 16644
+    },
+    {
+      "epoch": 1.343313695424098,
+      "grad_norm": 0.7149307727813721,
+      "learning_rate": 1.3725563573862321e-05,
+      "loss": 2.4425,
+      "step": 16645
+    },
+    {
+      "epoch": 1.3433943991606812,
+      "grad_norm": 0.7375392913818359,
+      "learning_rate": 1.3717582156505793e-05,
+      "loss": 2.409,
+      "step": 16646
+    },
+    {
+      "epoch": 1.3434751028972642,
+      "grad_norm": 0.8422170877456665,
+      "learning_rate": 1.3709602889532624e-05,
+      "loss": 2.4758,
+      "step": 16647
+    },
+    {
+      "epoch": 1.343555806633847,
+      "grad_norm": 0.6542177796363831,
+      "learning_rate": 1.3701625773141712e-05,
+      "loss": 2.4199,
+      "step": 16648
+    },
+    {
+      "epoch": 1.3436365103704302,
+      "grad_norm": 0.6639342904090881,
+      "learning_rate": 1.3693650807531898e-05,
+      "loss": 2.4366,
+      "step": 16649
+    },
+    {
+      "epoch": 1.3437172141070133,
+      "grad_norm": 0.7270925045013428,
+      "learning_rate": 1.3685677992901901e-05,
+      "loss": 2.3745,
+      "step": 16650
+    },
+    {
+      "epoch": 1.3437979178435961,
+      "grad_norm": 0.7325547337532043,
+      "learning_rate": 1.367770732945044e-05,
+      "loss": 2.5053,
+      "step": 16651
+    },
+    {
+      "epoch": 1.3438786215801792,
+      "grad_norm": 0.7752320766448975,
+      "learning_rate": 1.3669738817376177e-05,
+      "loss": 2.4505,
+      "step": 16652
+    },
+    {
+      "epoch": 1.343959325316762,
+      "grad_norm": 0.6538182497024536,
+      "learning_rate": 1.3661772456877675e-05,
+      "loss": 2.4164,
+      "step": 16653
+    },
+    {
+      "epoch": 1.3440400290533452,
+      "grad_norm": 0.6886051297187805,
+      "learning_rate": 1.3653808248153487e-05,
+      "loss": 2.4156,
+      "step": 16654
+    },
+    {
+      "epoch": 1.344120732789928,
+      "grad_norm": 0.6990679502487183,
+      "learning_rate": 1.3645846191402134e-05,
+      "loss": 2.418,
+      "step": 16655
+    },
+    {
+      "epoch": 1.3442014365265111,
+      "grad_norm": 0.7006608247756958,
+      "learning_rate": 1.3637886286821999e-05,
+      "loss": 2.3987,
+      "step": 16656
+    },
+    {
+      "epoch": 1.3442821402630942,
+      "grad_norm": 0.6858758926391602,
+      "learning_rate": 1.3629928534611502e-05,
+      "loss": 2.3571,
+      "step": 16657
+    },
+    {
+      "epoch": 1.344362843999677,
+      "grad_norm": 0.7273774147033691,
+      "learning_rate": 1.3621972934968951e-05,
+      "loss": 2.4141,
+      "step": 16658
+    },
+    {
+      "epoch": 1.3444435477362602,
+      "grad_norm": 0.6770352721214294,
+      "learning_rate": 1.3614019488092633e-05,
+      "loss": 2.4602,
+      "step": 16659
+    },
+    {
+      "epoch": 1.3445242514728433,
+      "grad_norm": 0.7473095655441284,
+      "learning_rate": 1.3606068194180766e-05,
+      "loss": 2.3884,
+      "step": 16660
+    },
+    {
+      "epoch": 1.3446049552094261,
+      "grad_norm": 0.7271387577056885,
+      "learning_rate": 1.3598119053431512e-05,
+      "loss": 2.4705,
+      "step": 16661
+    },
+    {
+      "epoch": 1.3446856589460092,
+      "grad_norm": 0.658349335193634,
+      "learning_rate": 1.3590172066043006e-05,
+      "loss": 2.4271,
+      "step": 16662
+    },
+    {
+      "epoch": 1.3447663626825923,
+      "grad_norm": 0.6479319930076599,
+      "learning_rate": 1.3582227232213273e-05,
+      "loss": 2.3428,
+      "step": 16663
+    },
+    {
+      "epoch": 1.3448470664191752,
+      "grad_norm": 0.700951874256134,
+      "learning_rate": 1.3574284552140337e-05,
+      "loss": 2.4926,
+      "step": 16664
+    },
+    {
+      "epoch": 1.3449277701557583,
+      "grad_norm": 0.6699960231781006,
+      "learning_rate": 1.3566344026022171e-05,
+      "loss": 2.4372,
+      "step": 16665
+    },
+    {
+      "epoch": 1.3450084738923413,
+      "grad_norm": 0.6743033528327942,
+      "learning_rate": 1.3558405654056617e-05,
+      "loss": 2.4142,
+      "step": 16666
+    },
+    {
+      "epoch": 1.3450891776289242,
+      "grad_norm": 0.6619464755058289,
+      "learning_rate": 1.355046943644157e-05,
+      "loss": 2.4099,
+      "step": 16667
+    },
+    {
+      "epoch": 1.3451698813655073,
+      "grad_norm": 0.668084442615509,
+      "learning_rate": 1.3542535373374798e-05,
+      "loss": 2.3895,
+      "step": 16668
+    },
+    {
+      "epoch": 1.3452505851020902,
+      "grad_norm": 0.7954626679420471,
+      "learning_rate": 1.3534603465054052e-05,
+      "loss": 2.479,
+      "step": 16669
+    },
+    {
+      "epoch": 1.3453312888386733,
+      "grad_norm": 0.6742919683456421,
+      "learning_rate": 1.3526673711677008e-05,
+      "loss": 2.4289,
+      "step": 16670
+    },
+    {
+      "epoch": 1.3454119925752561,
+      "grad_norm": 0.6564723253250122,
+      "learning_rate": 1.3518746113441316e-05,
+      "loss": 2.404,
+      "step": 16671
+    },
+    {
+      "epoch": 1.3454926963118392,
+      "grad_norm": 0.6955705881118774,
+      "learning_rate": 1.3510820670544521e-05,
+      "loss": 2.4274,
+      "step": 16672
+    },
+    {
+      "epoch": 1.3455734000484223,
+      "grad_norm": 0.6687749028205872,
+      "learning_rate": 1.3502897383184154e-05,
+      "loss": 2.4564,
+      "step": 16673
+    },
+    {
+      "epoch": 1.3456541037850052,
+      "grad_norm": 0.7984250783920288,
+      "learning_rate": 1.34949762515577e-05,
+      "loss": 2.3426,
+      "step": 16674
+    },
+    {
+      "epoch": 1.3457348075215882,
+      "grad_norm": 0.7334223389625549,
+      "learning_rate": 1.348705727586258e-05,
+      "loss": 2.4712,
+      "step": 16675
+    },
+    {
+      "epoch": 1.3458155112581713,
+      "grad_norm": 0.6732765436172485,
+      "learning_rate": 1.3479140456296114e-05,
+      "loss": 2.424,
+      "step": 16676
+    },
+    {
+      "epoch": 1.3458962149947542,
+      "grad_norm": 0.7944334149360657,
+      "learning_rate": 1.3471225793055641e-05,
+      "loss": 2.3951,
+      "step": 16677
+    },
+    {
+      "epoch": 1.3459769187313373,
+      "grad_norm": 0.6829007863998413,
+      "learning_rate": 1.3463313286338408e-05,
+      "loss": 2.4158,
+      "step": 16678
+    },
+    {
+      "epoch": 1.3460576224679204,
+      "grad_norm": 0.7019640207290649,
+      "learning_rate": 1.345540293634161e-05,
+      "loss": 2.4093,
+      "step": 16679
+    },
+    {
+      "epoch": 1.3461383262045032,
+      "grad_norm": 0.6839374303817749,
+      "learning_rate": 1.3447494743262412e-05,
+      "loss": 2.3959,
+      "step": 16680
+    },
+    {
+      "epoch": 1.3462190299410863,
+      "grad_norm": 0.7211155295372009,
+      "learning_rate": 1.3439588707297911e-05,
+      "loss": 2.4052,
+      "step": 16681
+    },
+    {
+      "epoch": 1.3462997336776692,
+      "grad_norm": 0.73811274766922,
+      "learning_rate": 1.3431684828645109e-05,
+      "loss": 2.4179,
+      "step": 16682
+    },
+    {
+      "epoch": 1.3463804374142523,
+      "grad_norm": 0.6634721159934998,
+      "learning_rate": 1.3423783107501009e-05,
+      "loss": 2.379,
+      "step": 16683
+    },
+    {
+      "epoch": 1.3464611411508352,
+      "grad_norm": 0.6884057521820068,
+      "learning_rate": 1.3415883544062579e-05,
+      "loss": 2.4144,
+      "step": 16684
+    },
+    {
+      "epoch": 1.3465418448874182,
+      "grad_norm": 0.7239587306976318,
+      "learning_rate": 1.340798613852664e-05,
+      "loss": 2.3856,
+      "step": 16685
+    },
+    {
+      "epoch": 1.3466225486240013,
+      "grad_norm": 0.7201077342033386,
+      "learning_rate": 1.3400090891090033e-05,
+      "loss": 2.4552,
+      "step": 16686
+    },
+    {
+      "epoch": 1.3467032523605842,
+      "grad_norm": 0.7049584984779358,
+      "learning_rate": 1.3392197801949558e-05,
+      "loss": 2.4424,
+      "step": 16687
+    },
+    {
+      "epoch": 1.3467839560971673,
+      "grad_norm": 0.7240790128707886,
+      "learning_rate": 1.3384306871301877e-05,
+      "loss": 2.4156,
+      "step": 16688
+    },
+    {
+      "epoch": 1.3468646598337504,
+      "grad_norm": 0.7276458740234375,
+      "learning_rate": 1.337641809934369e-05,
+      "loss": 2.3882,
+      "step": 16689
+    },
+    {
+      "epoch": 1.3469453635703332,
+      "grad_norm": 0.6650896072387695,
+      "learning_rate": 1.3368531486271607e-05,
+      "loss": 2.396,
+      "step": 16690
+    },
+    {
+      "epoch": 1.3470260673069163,
+      "grad_norm": 0.6946447491645813,
+      "learning_rate": 1.3360647032282203e-05,
+      "loss": 2.3779,
+      "step": 16691
+    },
+    {
+      "epoch": 1.3471067710434994,
+      "grad_norm": 0.7507699728012085,
+      "learning_rate": 1.3352764737571932e-05,
+      "loss": 2.4378,
+      "step": 16692
+    },
+    {
+      "epoch": 1.3471874747800823,
+      "grad_norm": 0.6548876762390137,
+      "learning_rate": 1.334488460233725e-05,
+      "loss": 2.4181,
+      "step": 16693
+    },
+    {
+      "epoch": 1.3472681785166654,
+      "grad_norm": 0.7000874280929565,
+      "learning_rate": 1.3337006626774595e-05,
+      "loss": 2.4463,
+      "step": 16694
+    },
+    {
+      "epoch": 1.3473488822532484,
+      "grad_norm": 0.6487517356872559,
+      "learning_rate": 1.3329130811080249e-05,
+      "loss": 2.3703,
+      "step": 16695
+    },
+    {
+      "epoch": 1.3474295859898313,
+      "grad_norm": 0.6447827219963074,
+      "learning_rate": 1.3321257155450517e-05,
+      "loss": 2.3779,
+      "step": 16696
+    },
+    {
+      "epoch": 1.3475102897264144,
+      "grad_norm": 0.6309572458267212,
+      "learning_rate": 1.3313385660081667e-05,
+      "loss": 2.4443,
+      "step": 16697
+    },
+    {
+      "epoch": 1.3475909934629973,
+      "grad_norm": 0.6366227865219116,
+      "learning_rate": 1.330551632516982e-05,
+      "loss": 2.3418,
+      "step": 16698
+    },
+    {
+      "epoch": 1.3476716971995804,
+      "grad_norm": 0.6864019632339478,
+      "learning_rate": 1.3297649150911117e-05,
+      "loss": 2.4416,
+      "step": 16699
+    },
+    {
+      "epoch": 1.3477524009361632,
+      "grad_norm": 0.6807940006256104,
+      "learning_rate": 1.3289784137501671e-05,
+      "loss": 2.4465,
+      "step": 16700
+    },
+    {
+      "epoch": 1.3478331046727463,
+      "grad_norm": 0.6991185545921326,
+      "learning_rate": 1.3281921285137455e-05,
+      "loss": 2.3929,
+      "step": 16701
+    },
+    {
+      "epoch": 1.3479138084093294,
+      "grad_norm": 0.691908061504364,
+      "learning_rate": 1.3274060594014437e-05,
+      "loss": 2.4237,
+      "step": 16702
+    },
+    {
+      "epoch": 1.3479945121459123,
+      "grad_norm": 0.6909685730934143,
+      "learning_rate": 1.3266202064328548e-05,
+      "loss": 2.3695,
+      "step": 16703
+    },
+    {
+      "epoch": 1.3480752158824953,
+      "grad_norm": 0.6473715901374817,
+      "learning_rate": 1.325834569627562e-05,
+      "loss": 2.384,
+      "step": 16704
+    },
+    {
+      "epoch": 1.3481559196190784,
+      "grad_norm": 0.7433453798294067,
+      "learning_rate": 1.3250491490051454e-05,
+      "loss": 2.4546,
+      "step": 16705
+    },
+    {
+      "epoch": 1.3482366233556613,
+      "grad_norm": 0.7432501316070557,
+      "learning_rate": 1.3242639445851812e-05,
+      "loss": 2.4204,
+      "step": 16706
+    },
+    {
+      "epoch": 1.3483173270922444,
+      "grad_norm": 0.6661228537559509,
+      "learning_rate": 1.3234789563872397e-05,
+      "loss": 2.4454,
+      "step": 16707
+    },
+    {
+      "epoch": 1.3483980308288275,
+      "grad_norm": 0.7481260895729065,
+      "learning_rate": 1.3226941844308816e-05,
+      "loss": 2.4348,
+      "step": 16708
+    },
+    {
+      "epoch": 1.3484787345654103,
+      "grad_norm": 0.6986531019210815,
+      "learning_rate": 1.3219096287356669e-05,
+      "loss": 2.3622,
+      "step": 16709
+    },
+    {
+      "epoch": 1.3485594383019934,
+      "grad_norm": 0.7457645535469055,
+      "learning_rate": 1.321125289321149e-05,
+      "loss": 2.4399,
+      "step": 16710
+    },
+    {
+      "epoch": 1.3486401420385765,
+      "grad_norm": 0.6710307598114014,
+      "learning_rate": 1.3203411662068754e-05,
+      "loss": 2.3857,
+      "step": 16711
+    },
+    {
+      "epoch": 1.3487208457751594,
+      "grad_norm": 0.767304539680481,
+      "learning_rate": 1.3195572594123884e-05,
+      "loss": 2.4666,
+      "step": 16712
+    },
+    {
+      "epoch": 1.3488015495117425,
+      "grad_norm": 0.6720963716506958,
+      "learning_rate": 1.3187735689572289e-05,
+      "loss": 2.3952,
+      "step": 16713
+    },
+    {
+      "epoch": 1.3488822532483253,
+      "grad_norm": 0.6381734609603882,
+      "learning_rate": 1.3179900948609213e-05,
+      "loss": 2.3632,
+      "step": 16714
+    },
+    {
+      "epoch": 1.3489629569849084,
+      "grad_norm": 0.6697315573692322,
+      "learning_rate": 1.317206837142997e-05,
+      "loss": 2.4117,
+      "step": 16715
+    },
+    {
+      "epoch": 1.3490436607214913,
+      "grad_norm": 0.723676323890686,
+      "learning_rate": 1.3164237958229764e-05,
+      "loss": 2.3772,
+      "step": 16716
+    },
+    {
+      "epoch": 1.3491243644580744,
+      "grad_norm": 0.7021055817604065,
+      "learning_rate": 1.3156409709203732e-05,
+      "loss": 2.3808,
+      "step": 16717
+    },
+    {
+      "epoch": 1.3492050681946575,
+      "grad_norm": 0.7128920555114746,
+      "learning_rate": 1.3148583624546962e-05,
+      "loss": 2.3854,
+      "step": 16718
+    },
+    {
+      "epoch": 1.3492857719312403,
+      "grad_norm": 0.6684797406196594,
+      "learning_rate": 1.314075970445453e-05,
+      "loss": 2.3722,
+      "step": 16719
+    },
+    {
+      "epoch": 1.3493664756678234,
+      "grad_norm": 0.6710386276245117,
+      "learning_rate": 1.3132937949121426e-05,
+      "loss": 2.412,
+      "step": 16720
+    },
+    {
+      "epoch": 1.3494471794044065,
+      "grad_norm": 0.7207252979278564,
+      "learning_rate": 1.3125118358742572e-05,
+      "loss": 2.4506,
+      "step": 16721
+    },
+    {
+      "epoch": 1.3495278831409894,
+      "grad_norm": 0.685516893863678,
+      "learning_rate": 1.3117300933512865e-05,
+      "loss": 2.435,
+      "step": 16722
+    },
+    {
+      "epoch": 1.3496085868775725,
+      "grad_norm": 0.71708744764328,
+      "learning_rate": 1.3109485673627154e-05,
+      "loss": 2.4735,
+      "step": 16723
+    },
+    {
+      "epoch": 1.3496892906141555,
+      "grad_norm": 0.7293861508369446,
+      "learning_rate": 1.3101672579280166e-05,
+      "loss": 2.4545,
+      "step": 16724
+    },
+    {
+      "epoch": 1.3497699943507384,
+      "grad_norm": 0.6448976993560791,
+      "learning_rate": 1.3093861650666661e-05,
+      "loss": 2.386,
+      "step": 16725
+    },
+    {
+      "epoch": 1.3498506980873215,
+      "grad_norm": 0.8111226558685303,
+      "learning_rate": 1.3086052887981315e-05,
+      "loss": 2.4733,
+      "step": 16726
+    },
+    {
+      "epoch": 1.3499314018239044,
+      "grad_norm": 0.7673875093460083,
+      "learning_rate": 1.3078246291418706e-05,
+      "loss": 2.4119,
+      "step": 16727
+    },
+    {
+      "epoch": 1.3500121055604875,
+      "grad_norm": 0.7296731472015381,
+      "learning_rate": 1.307044186117341e-05,
+      "loss": 2.3724,
+      "step": 16728
+    },
+    {
+      "epoch": 1.3500928092970703,
+      "grad_norm": 0.6947155594825745,
+      "learning_rate": 1.306263959743994e-05,
+      "loss": 2.3989,
+      "step": 16729
+    },
+    {
+      "epoch": 1.3501735130336534,
+      "grad_norm": 0.6781659722328186,
+      "learning_rate": 1.3054839500412753e-05,
+      "loss": 2.429,
+      "step": 16730
+    },
+    {
+      "epoch": 1.3502542167702365,
+      "grad_norm": 0.7498819231987,
+      "learning_rate": 1.3047041570286244e-05,
+      "loss": 2.459,
+      "step": 16731
+    },
+    {
+      "epoch": 1.3503349205068194,
+      "grad_norm": 0.6651057004928589,
+      "learning_rate": 1.3039245807254774e-05,
+      "loss": 2.4049,
+      "step": 16732
+    },
+    {
+      "epoch": 1.3504156242434024,
+      "grad_norm": 0.6998507380485535,
+      "learning_rate": 1.3031452211512596e-05,
+      "loss": 2.4083,
+      "step": 16733
+    },
+    {
+      "epoch": 1.3504963279799855,
+      "grad_norm": 0.6522402167320251,
+      "learning_rate": 1.3023660783253966e-05,
+      "loss": 2.3987,
+      "step": 16734
+    },
+    {
+      "epoch": 1.3505770317165684,
+      "grad_norm": 0.6618130207061768,
+      "learning_rate": 1.3015871522673096e-05,
+      "loss": 2.4514,
+      "step": 16735
+    },
+    {
+      "epoch": 1.3506577354531515,
+      "grad_norm": 0.7139489650726318,
+      "learning_rate": 1.300808442996405e-05,
+      "loss": 2.484,
+      "step": 16736
+    },
+    {
+      "epoch": 1.3507384391897346,
+      "grad_norm": 0.6582522988319397,
+      "learning_rate": 1.3000299505320956e-05,
+      "loss": 2.4463,
+      "step": 16737
+    },
+    {
+      "epoch": 1.3508191429263174,
+      "grad_norm": 0.7115446329116821,
+      "learning_rate": 1.2992516748937811e-05,
+      "loss": 2.4795,
+      "step": 16738
+    },
+    {
+      "epoch": 1.3508998466629005,
+      "grad_norm": 0.7243752479553223,
+      "learning_rate": 1.2984736161008581e-05,
+      "loss": 2.4151,
+      "step": 16739
+    },
+    {
+      "epoch": 1.3509805503994836,
+      "grad_norm": 0.758084774017334,
+      "learning_rate": 1.297695774172719e-05,
+      "loss": 2.4028,
+      "step": 16740
+    },
+    {
+      "epoch": 1.3510612541360665,
+      "grad_norm": 0.6555618643760681,
+      "learning_rate": 1.2969181491287496e-05,
+      "loss": 2.4184,
+      "step": 16741
+    },
+    {
+      "epoch": 1.3511419578726496,
+      "grad_norm": 0.6657842993736267,
+      "learning_rate": 1.2961407409883331e-05,
+      "loss": 2.375,
+      "step": 16742
+    },
+    {
+      "epoch": 1.3512226616092324,
+      "grad_norm": 0.6355723142623901,
+      "learning_rate": 1.2953635497708382e-05,
+      "loss": 2.4202,
+      "step": 16743
+    },
+    {
+      "epoch": 1.3513033653458155,
+      "grad_norm": 0.7384408116340637,
+      "learning_rate": 1.2945865754956377e-05,
+      "loss": 2.4298,
+      "step": 16744
+    },
+    {
+      "epoch": 1.3513840690823984,
+      "grad_norm": 0.7300455570220947,
+      "learning_rate": 1.2938098181820979e-05,
+      "loss": 2.3842,
+      "step": 16745
+    },
+    {
+      "epoch": 1.3514647728189815,
+      "grad_norm": 0.7378895282745361,
+      "learning_rate": 1.2930332778495735e-05,
+      "loss": 2.4025,
+      "step": 16746
+    },
+    {
+      "epoch": 1.3515454765555646,
+      "grad_norm": 0.6542565822601318,
+      "learning_rate": 1.2922569545174212e-05,
+      "loss": 2.3995,
+      "step": 16747
+    },
+    {
+      "epoch": 1.3516261802921474,
+      "grad_norm": 0.669829249382019,
+      "learning_rate": 1.291480848204989e-05,
+      "loss": 2.3843,
+      "step": 16748
+    },
+    {
+      "epoch": 1.3517068840287305,
+      "grad_norm": 0.6747604608535767,
+      "learning_rate": 1.2907049589316167e-05,
+      "loss": 2.4108,
+      "step": 16749
+    },
+    {
+      "epoch": 1.3517875877653136,
+      "grad_norm": 0.7003559470176697,
+      "learning_rate": 1.2899292867166402e-05,
+      "loss": 2.4233,
+      "step": 16750
+    },
+    {
+      "epoch": 1.3518682915018965,
+      "grad_norm": 0.7365099191665649,
+      "learning_rate": 1.2891538315793994e-05,
+      "loss": 2.3592,
+      "step": 16751
+    },
+    {
+      "epoch": 1.3519489952384796,
+      "grad_norm": 0.6849377751350403,
+      "learning_rate": 1.2883785935392123e-05,
+      "loss": 2.3943,
+      "step": 16752
+    },
+    {
+      "epoch": 1.3520296989750626,
+      "grad_norm": 0.7263002395629883,
+      "learning_rate": 1.2876035726154045e-05,
+      "loss": 2.4078,
+      "step": 16753
+    },
+    {
+      "epoch": 1.3521104027116455,
+      "grad_norm": 0.7341182827949524,
+      "learning_rate": 1.2868287688272884e-05,
+      "loss": 2.3568,
+      "step": 16754
+    },
+    {
+      "epoch": 1.3521911064482286,
+      "grad_norm": 0.7281078100204468,
+      "learning_rate": 1.2860541821941796e-05,
+      "loss": 2.4073,
+      "step": 16755
+    },
+    {
+      "epoch": 1.3522718101848117,
+      "grad_norm": 0.6302868127822876,
+      "learning_rate": 1.285279812735376e-05,
+      "loss": 2.3946,
+      "step": 16756
+    },
+    {
+      "epoch": 1.3523525139213946,
+      "grad_norm": 0.7333062887191772,
+      "learning_rate": 1.28450566047018e-05,
+      "loss": 2.3892,
+      "step": 16757
+    },
+    {
+      "epoch": 1.3524332176579776,
+      "grad_norm": 0.74838787317276,
+      "learning_rate": 1.2837317254178882e-05,
+      "loss": 2.4844,
+      "step": 16758
+    },
+    {
+      "epoch": 1.3525139213945605,
+      "grad_norm": 0.7085757255554199,
+      "learning_rate": 1.2829580075977843e-05,
+      "loss": 2.3583,
+      "step": 16759
+    },
+    {
+      "epoch": 1.3525946251311436,
+      "grad_norm": 0.7182579040527344,
+      "learning_rate": 1.2821845070291527e-05,
+      "loss": 2.4326,
+      "step": 16760
+    },
+    {
+      "epoch": 1.3526753288677265,
+      "grad_norm": 0.6857885718345642,
+      "learning_rate": 1.2814112237312714e-05,
+      "loss": 2.4406,
+      "step": 16761
+    },
+    {
+      "epoch": 1.3527560326043095,
+      "grad_norm": 0.7629652619361877,
+      "learning_rate": 1.2806381577234139e-05,
+      "loss": 2.4839,
+      "step": 16762
+    },
+    {
+      "epoch": 1.3528367363408926,
+      "grad_norm": 0.6940319538116455,
+      "learning_rate": 1.2798653090248458e-05,
+      "loss": 2.3918,
+      "step": 16763
+    },
+    {
+      "epoch": 1.3529174400774755,
+      "grad_norm": 0.6825633645057678,
+      "learning_rate": 1.2790926776548318e-05,
+      "loss": 2.3828,
+      "step": 16764
+    },
+    {
+      "epoch": 1.3529981438140586,
+      "grad_norm": 0.6830280423164368,
+      "learning_rate": 1.278320263632622e-05,
+      "loss": 2.3727,
+      "step": 16765
+    },
+    {
+      "epoch": 1.3530788475506417,
+      "grad_norm": 0.6782984733581543,
+      "learning_rate": 1.2775480669774698e-05,
+      "loss": 2.3984,
+      "step": 16766
+    },
+    {
+      "epoch": 1.3531595512872245,
+      "grad_norm": 0.6939808130264282,
+      "learning_rate": 1.276776087708621e-05,
+      "loss": 2.3724,
+      "step": 16767
+    },
+    {
+      "epoch": 1.3532402550238076,
+      "grad_norm": 0.7562546133995056,
+      "learning_rate": 1.276004325845317e-05,
+      "loss": 2.4178,
+      "step": 16768
+    },
+    {
+      "epoch": 1.3533209587603907,
+      "grad_norm": 0.6692922115325928,
+      "learning_rate": 1.2752327814067877e-05,
+      "loss": 2.4072,
+      "step": 16769
+    },
+    {
+      "epoch": 1.3534016624969736,
+      "grad_norm": 0.6783415079116821,
+      "learning_rate": 1.2744614544122635e-05,
+      "loss": 2.3993,
+      "step": 16770
+    },
+    {
+      "epoch": 1.3534823662335567,
+      "grad_norm": 0.6608997583389282,
+      "learning_rate": 1.27369034488097e-05,
+      "loss": 2.3883,
+      "step": 16771
+    },
+    {
+      "epoch": 1.3535630699701398,
+      "grad_norm": 0.6849228739738464,
+      "learning_rate": 1.2729194528321231e-05,
+      "loss": 2.4009,
+      "step": 16772
+    },
+    {
+      "epoch": 1.3536437737067226,
+      "grad_norm": 0.7059305906295776,
+      "learning_rate": 1.2721487782849362e-05,
+      "loss": 2.508,
+      "step": 16773
+    },
+    {
+      "epoch": 1.3537244774433057,
+      "grad_norm": 0.6471492052078247,
+      "learning_rate": 1.2713783212586183e-05,
+      "loss": 2.3813,
+      "step": 16774
+    },
+    {
+      "epoch": 1.3538051811798886,
+      "grad_norm": 0.7108949422836304,
+      "learning_rate": 1.2706080817723687e-05,
+      "loss": 2.4189,
+      "step": 16775
+    },
+    {
+      "epoch": 1.3538858849164717,
+      "grad_norm": 0.6623945236206055,
+      "learning_rate": 1.269838059845383e-05,
+      "loss": 2.4128,
+      "step": 16776
+    },
+    {
+      "epoch": 1.3539665886530545,
+      "grad_norm": 0.6595518589019775,
+      "learning_rate": 1.269068255496857e-05,
+      "loss": 2.3984,
+      "step": 16777
+    },
+    {
+      "epoch": 1.3540472923896376,
+      "grad_norm": 0.6932248473167419,
+      "learning_rate": 1.2682986687459708e-05,
+      "loss": 2.3951,
+      "step": 16778
+    },
+    {
+      "epoch": 1.3541279961262207,
+      "grad_norm": 0.6914867162704468,
+      "learning_rate": 1.2675292996119059e-05,
+      "loss": 2.4602,
+      "step": 16779
+    },
+    {
+      "epoch": 1.3542086998628036,
+      "grad_norm": 0.6633034348487854,
+      "learning_rate": 1.266760148113838e-05,
+      "loss": 2.43,
+      "step": 16780
+    },
+    {
+      "epoch": 1.3542894035993867,
+      "grad_norm": 0.6987594366073608,
+      "learning_rate": 1.2659912142709363e-05,
+      "loss": 2.3962,
+      "step": 16781
+    },
+    {
+      "epoch": 1.3543701073359697,
+      "grad_norm": 0.7429597973823547,
+      "learning_rate": 1.2652224981023652e-05,
+      "loss": 2.4838,
+      "step": 16782
+    },
+    {
+      "epoch": 1.3544508110725526,
+      "grad_norm": 0.6402504444122314,
+      "learning_rate": 1.2644539996272808e-05,
+      "loss": 2.43,
+      "step": 16783
+    },
+    {
+      "epoch": 1.3545315148091357,
+      "grad_norm": 0.6763156652450562,
+      "learning_rate": 1.263685718864841e-05,
+      "loss": 2.4911,
+      "step": 16784
+    },
+    {
+      "epoch": 1.3546122185457188,
+      "grad_norm": 0.8133900165557861,
+      "learning_rate": 1.2629176558341881e-05,
+      "loss": 2.45,
+      "step": 16785
+    },
+    {
+      "epoch": 1.3546929222823016,
+      "grad_norm": 0.6946277022361755,
+      "learning_rate": 1.262149810554465e-05,
+      "loss": 2.43,
+      "step": 16786
+    },
+    {
+      "epoch": 1.3547736260188847,
+      "grad_norm": 0.7667170166969299,
+      "learning_rate": 1.2613821830448125e-05,
+      "loss": 2.4464,
+      "step": 16787
+    },
+    {
+      "epoch": 1.3548543297554676,
+      "grad_norm": 0.672662615776062,
+      "learning_rate": 1.2606147733243567e-05,
+      "loss": 2.3653,
+      "step": 16788
+    },
+    {
+      "epoch": 1.3549350334920507,
+      "grad_norm": 0.6856412291526794,
+      "learning_rate": 1.2598475814122258e-05,
+      "loss": 2.3924,
+      "step": 16789
+    },
+    {
+      "epoch": 1.3550157372286336,
+      "grad_norm": 0.6966650485992432,
+      "learning_rate": 1.2590806073275407e-05,
+      "loss": 2.4039,
+      "step": 16790
+    },
+    {
+      "epoch": 1.3550964409652166,
+      "grad_norm": 0.7397874593734741,
+      "learning_rate": 1.2583138510894143e-05,
+      "loss": 2.4769,
+      "step": 16791
+    },
+    {
+      "epoch": 1.3551771447017997,
+      "grad_norm": 0.6960996985435486,
+      "learning_rate": 1.2575473127169591e-05,
+      "loss": 2.4342,
+      "step": 16792
+    },
+    {
+      "epoch": 1.3552578484383826,
+      "grad_norm": 0.7324376702308655,
+      "learning_rate": 1.2567809922292795e-05,
+      "loss": 2.4779,
+      "step": 16793
+    },
+    {
+      "epoch": 1.3553385521749657,
+      "grad_norm": 0.6891930103302002,
+      "learning_rate": 1.2560148896454704e-05,
+      "loss": 2.4228,
+      "step": 16794
+    },
+    {
+      "epoch": 1.3554192559115488,
+      "grad_norm": 0.6919474601745605,
+      "learning_rate": 1.2552490049846278e-05,
+      "loss": 2.4178,
+      "step": 16795
+    },
+    {
+      "epoch": 1.3554999596481316,
+      "grad_norm": 0.7067604660987854,
+      "learning_rate": 1.2544833382658405e-05,
+      "loss": 2.457,
+      "step": 16796
+    },
+    {
+      "epoch": 1.3555806633847147,
+      "grad_norm": 0.7667992115020752,
+      "learning_rate": 1.253717889508188e-05,
+      "loss": 2.3951,
+      "step": 16797
+    },
+    {
+      "epoch": 1.3556613671212978,
+      "grad_norm": 0.6337998509407043,
+      "learning_rate": 1.2529526587307482e-05,
+      "loss": 2.3788,
+      "step": 16798
+    },
+    {
+      "epoch": 1.3557420708578807,
+      "grad_norm": 0.6591900587081909,
+      "learning_rate": 1.2521876459525927e-05,
+      "loss": 2.4101,
+      "step": 16799
+    },
+    {
+      "epoch": 1.3558227745944638,
+      "grad_norm": 0.7115298509597778,
+      "learning_rate": 1.2514228511927895e-05,
+      "loss": 2.4417,
+      "step": 16800
+    },
+    {
+      "epoch": 1.3559034783310469,
+      "grad_norm": 0.6851321458816528,
+      "learning_rate": 1.2506582744703965e-05,
+      "loss": 2.4081,
+      "step": 16801
+    },
+    {
+      "epoch": 1.3559841820676297,
+      "grad_norm": 0.7469603419303894,
+      "learning_rate": 1.249893915804471e-05,
+      "loss": 2.3703,
+      "step": 16802
+    },
+    {
+      "epoch": 1.3560648858042128,
+      "grad_norm": 0.6972614526748657,
+      "learning_rate": 1.2491297752140641e-05,
+      "loss": 2.3549,
+      "step": 16803
+    },
+    {
+      "epoch": 1.3561455895407957,
+      "grad_norm": 0.6669485569000244,
+      "learning_rate": 1.2483658527182151e-05,
+      "loss": 2.4261,
+      "step": 16804
+    },
+    {
+      "epoch": 1.3562262932773788,
+      "grad_norm": 0.7516919374465942,
+      "learning_rate": 1.247602148335968e-05,
+      "loss": 2.4323,
+      "step": 16805
+    },
+    {
+      "epoch": 1.3563069970139616,
+      "grad_norm": 0.7191836833953857,
+      "learning_rate": 1.2468386620863548e-05,
+      "loss": 2.4242,
+      "step": 16806
+    },
+    {
+      "epoch": 1.3563877007505447,
+      "grad_norm": 0.660237729549408,
+      "learning_rate": 1.2460753939884017e-05,
+      "loss": 2.4154,
+      "step": 16807
+    },
+    {
+      "epoch": 1.3564684044871278,
+      "grad_norm": 0.749531626701355,
+      "learning_rate": 1.2453123440611325e-05,
+      "loss": 2.4138,
+      "step": 16808
+    },
+    {
+      "epoch": 1.3565491082237107,
+      "grad_norm": 0.6808986067771912,
+      "learning_rate": 1.2445495123235673e-05,
+      "loss": 2.3918,
+      "step": 16809
+    },
+    {
+      "epoch": 1.3566298119602938,
+      "grad_norm": 0.686183750629425,
+      "learning_rate": 1.2437868987947133e-05,
+      "loss": 2.4172,
+      "step": 16810
+    },
+    {
+      "epoch": 1.3567105156968768,
+      "grad_norm": 0.6487868428230286,
+      "learning_rate": 1.2430245034935784e-05,
+      "loss": 2.4199,
+      "step": 16811
+    },
+    {
+      "epoch": 1.3567912194334597,
+      "grad_norm": 0.7352244257926941,
+      "learning_rate": 1.242262326439163e-05,
+      "loss": 2.3779,
+      "step": 16812
+    },
+    {
+      "epoch": 1.3568719231700428,
+      "grad_norm": 0.7250565886497498,
+      "learning_rate": 1.2415003676504644e-05,
+      "loss": 2.4106,
+      "step": 16813
+    },
+    {
+      "epoch": 1.3569526269066259,
+      "grad_norm": 0.6843926906585693,
+      "learning_rate": 1.2407386271464716e-05,
+      "loss": 2.3725,
+      "step": 16814
+    },
+    {
+      "epoch": 1.3570333306432087,
+      "grad_norm": 0.686326801776886,
+      "learning_rate": 1.2399771049461684e-05,
+      "loss": 2.3709,
+      "step": 16815
+    },
+    {
+      "epoch": 1.3571140343797918,
+      "grad_norm": 0.6796969771385193,
+      "learning_rate": 1.2392158010685373e-05,
+      "loss": 2.4545,
+      "step": 16816
+    },
+    {
+      "epoch": 1.357194738116375,
+      "grad_norm": 0.6469466090202332,
+      "learning_rate": 1.2384547155325466e-05,
+      "loss": 2.4263,
+      "step": 16817
+    },
+    {
+      "epoch": 1.3572754418529578,
+      "grad_norm": 0.7089909911155701,
+      "learning_rate": 1.2376938483571688e-05,
+      "loss": 2.378,
+      "step": 16818
+    },
+    {
+      "epoch": 1.3573561455895409,
+      "grad_norm": 0.7313235402107239,
+      "learning_rate": 1.2369331995613665e-05,
+      "loss": 2.46,
+      "step": 16819
+    },
+    {
+      "epoch": 1.3574368493261237,
+      "grad_norm": 0.7555651664733887,
+      "learning_rate": 1.2361727691640934e-05,
+      "loss": 2.531,
+      "step": 16820
+    },
+    {
+      "epoch": 1.3575175530627068,
+      "grad_norm": 0.7563485503196716,
+      "learning_rate": 1.2354125571843033e-05,
+      "loss": 2.4205,
+      "step": 16821
+    },
+    {
+      "epoch": 1.3575982567992897,
+      "grad_norm": 0.7996519804000854,
+      "learning_rate": 1.2346525636409434e-05,
+      "loss": 2.4223,
+      "step": 16822
+    },
+    {
+      "epoch": 1.3576789605358728,
+      "grad_norm": 0.7141731977462769,
+      "learning_rate": 1.233892788552955e-05,
+      "loss": 2.4554,
+      "step": 16823
+    },
+    {
+      "epoch": 1.3577596642724559,
+      "grad_norm": 0.6715070605278015,
+      "learning_rate": 1.233133231939273e-05,
+      "loss": 2.4386,
+      "step": 16824
+    },
+    {
+      "epoch": 1.3578403680090387,
+      "grad_norm": 0.6893020272254944,
+      "learning_rate": 1.2323738938188301e-05,
+      "loss": 2.4065,
+      "step": 16825
+    },
+    {
+      "epoch": 1.3579210717456218,
+      "grad_norm": 0.7542821764945984,
+      "learning_rate": 1.2316147742105454e-05,
+      "loss": 2.3974,
+      "step": 16826
+    },
+    {
+      "epoch": 1.358001775482205,
+      "grad_norm": 0.7177664041519165,
+      "learning_rate": 1.230855873133343e-05,
+      "loss": 2.4306,
+      "step": 16827
+    },
+    {
+      "epoch": 1.3580824792187878,
+      "grad_norm": 0.7056576013565063,
+      "learning_rate": 1.2300971906061354e-05,
+      "loss": 2.4238,
+      "step": 16828
+    },
+    {
+      "epoch": 1.3581631829553709,
+      "grad_norm": 0.686903715133667,
+      "learning_rate": 1.2293387266478296e-05,
+      "loss": 2.3902,
+      "step": 16829
+    },
+    {
+      "epoch": 1.358243886691954,
+      "grad_norm": 0.7377725839614868,
+      "learning_rate": 1.2285804812773293e-05,
+      "loss": 2.4294,
+      "step": 16830
+    },
+    {
+      "epoch": 1.3583245904285368,
+      "grad_norm": 0.6537891030311584,
+      "learning_rate": 1.227822454513532e-05,
+      "loss": 2.374,
+      "step": 16831
+    },
+    {
+      "epoch": 1.35840529416512,
+      "grad_norm": 0.684699296951294,
+      "learning_rate": 1.2270646463753288e-05,
+      "loss": 2.4105,
+      "step": 16832
+    },
+    {
+      "epoch": 1.3584859979017028,
+      "grad_norm": 0.7042316794395447,
+      "learning_rate": 1.2263070568816081e-05,
+      "loss": 2.4246,
+      "step": 16833
+    },
+    {
+      "epoch": 1.3585667016382859,
+      "grad_norm": 0.7610476613044739,
+      "learning_rate": 1.2255496860512505e-05,
+      "loss": 2.4581,
+      "step": 16834
+    },
+    {
+      "epoch": 1.3586474053748687,
+      "grad_norm": 0.6620839834213257,
+      "learning_rate": 1.224792533903134e-05,
+      "loss": 2.4138,
+      "step": 16835
+    },
+    {
+      "epoch": 1.3587281091114518,
+      "grad_norm": 0.6861035823822021,
+      "learning_rate": 1.2240356004561227e-05,
+      "loss": 2.4195,
+      "step": 16836
+    },
+    {
+      "epoch": 1.358808812848035,
+      "grad_norm": 0.7186882495880127,
+      "learning_rate": 1.2232788857290855e-05,
+      "loss": 2.404,
+      "step": 16837
+    },
+    {
+      "epoch": 1.3588895165846178,
+      "grad_norm": 0.7219386696815491,
+      "learning_rate": 1.2225223897408833e-05,
+      "loss": 2.3778,
+      "step": 16838
+    },
+    {
+      "epoch": 1.3589702203212009,
+      "grad_norm": 0.6935911774635315,
+      "learning_rate": 1.2217661125103663e-05,
+      "loss": 2.4617,
+      "step": 16839
+    },
+    {
+      "epoch": 1.359050924057784,
+      "grad_norm": 0.7885910272598267,
+      "learning_rate": 1.2210100540563828e-05,
+      "loss": 2.4467,
+      "step": 16840
+    },
+    {
+      "epoch": 1.3591316277943668,
+      "grad_norm": 0.6690255403518677,
+      "learning_rate": 1.220254214397778e-05,
+      "loss": 2.381,
+      "step": 16841
+    },
+    {
+      "epoch": 1.35921233153095,
+      "grad_norm": 0.7592741847038269,
+      "learning_rate": 1.2194985935533887e-05,
+      "loss": 2.4459,
+      "step": 16842
+    },
+    {
+      "epoch": 1.359293035267533,
+      "grad_norm": 0.827460527420044,
+      "learning_rate": 1.2187431915420466e-05,
+      "loss": 2.3842,
+      "step": 16843
+    },
+    {
+      "epoch": 1.3593737390041158,
+      "grad_norm": 0.7313764691352844,
+      "learning_rate": 1.2179880083825811e-05,
+      "loss": 2.3938,
+      "step": 16844
+    },
+    {
+      "epoch": 1.359454442740699,
+      "grad_norm": 0.7093486189842224,
+      "learning_rate": 1.2172330440938084e-05,
+      "loss": 2.4316,
+      "step": 16845
+    },
+    {
+      "epoch": 1.359535146477282,
+      "grad_norm": 0.6805742383003235,
+      "learning_rate": 1.2164782986945467e-05,
+      "loss": 2.4372,
+      "step": 16846
+    },
+    {
+      "epoch": 1.3596158502138649,
+      "grad_norm": 0.7525961399078369,
+      "learning_rate": 1.2157237722036064e-05,
+      "loss": 2.3867,
+      "step": 16847
+    },
+    {
+      "epoch": 1.359696553950448,
+      "grad_norm": 0.723896861076355,
+      "learning_rate": 1.2149694646397947e-05,
+      "loss": 2.4685,
+      "step": 16848
+    },
+    {
+      "epoch": 1.3597772576870308,
+      "grad_norm": 0.704448938369751,
+      "learning_rate": 1.2142153760219055e-05,
+      "loss": 2.4463,
+      "step": 16849
+    },
+    {
+      "epoch": 1.359857961423614,
+      "grad_norm": 0.7207927703857422,
+      "learning_rate": 1.2134615063687349e-05,
+      "loss": 2.3549,
+      "step": 16850
+    },
+    {
+      "epoch": 1.3599386651601968,
+      "grad_norm": 0.7106234431266785,
+      "learning_rate": 1.2127078556990724e-05,
+      "loss": 2.4145,
+      "step": 16851
+    },
+    {
+      "epoch": 1.3600193688967799,
+      "grad_norm": 0.7740694284439087,
+      "learning_rate": 1.2119544240316993e-05,
+      "loss": 2.3999,
+      "step": 16852
+    },
+    {
+      "epoch": 1.360100072633363,
+      "grad_norm": 0.6696181297302246,
+      "learning_rate": 1.2112012113853954e-05,
+      "loss": 2.4046,
+      "step": 16853
+    },
+    {
+      "epoch": 1.3601807763699458,
+      "grad_norm": 0.6758043169975281,
+      "learning_rate": 1.2104482177789334e-05,
+      "loss": 2.4021,
+      "step": 16854
+    },
+    {
+      "epoch": 1.360261480106529,
+      "grad_norm": 0.6659380793571472,
+      "learning_rate": 1.2096954432310758e-05,
+      "loss": 2.4145,
+      "step": 16855
+    },
+    {
+      "epoch": 1.360342183843112,
+      "grad_norm": 0.6889290809631348,
+      "learning_rate": 1.2089428877605858e-05,
+      "loss": 2.3486,
+      "step": 16856
+    },
+    {
+      "epoch": 1.3604228875796949,
+      "grad_norm": 0.6755563020706177,
+      "learning_rate": 1.2081905513862201e-05,
+      "loss": 2.4294,
+      "step": 16857
+    },
+    {
+      "epoch": 1.360503591316278,
+      "grad_norm": 0.7662243843078613,
+      "learning_rate": 1.2074384341267276e-05,
+      "loss": 2.414,
+      "step": 16858
+    },
+    {
+      "epoch": 1.360584295052861,
+      "grad_norm": 0.7432721853256226,
+      "learning_rate": 1.2066865360008517e-05,
+      "loss": 2.4314,
+      "step": 16859
+    },
+    {
+      "epoch": 1.360664998789444,
+      "grad_norm": 0.6465074419975281,
+      "learning_rate": 1.2059348570273366e-05,
+      "loss": 2.3349,
+      "step": 16860
+    },
+    {
+      "epoch": 1.360745702526027,
+      "grad_norm": 0.6940968632698059,
+      "learning_rate": 1.2051833972249105e-05,
+      "loss": 2.4539,
+      "step": 16861
+    },
+    {
+      "epoch": 1.36082640626261,
+      "grad_norm": 0.7211138010025024,
+      "learning_rate": 1.2044321566123019e-05,
+      "loss": 2.4041,
+      "step": 16862
+    },
+    {
+      "epoch": 1.360907109999193,
+      "grad_norm": 0.6746649146080017,
+      "learning_rate": 1.2036811352082367e-05,
+      "loss": 2.4329,
+      "step": 16863
+    },
+    {
+      "epoch": 1.360987813735776,
+      "grad_norm": 0.7502184510231018,
+      "learning_rate": 1.2029303330314345e-05,
+      "loss": 2.407,
+      "step": 16864
+    },
+    {
+      "epoch": 1.361068517472359,
+      "grad_norm": 0.7192596793174744,
+      "learning_rate": 1.2021797501006027e-05,
+      "loss": 2.3907,
+      "step": 16865
+    },
+    {
+      "epoch": 1.361149221208942,
+      "grad_norm": 0.6682254672050476,
+      "learning_rate": 1.2014293864344483e-05,
+      "loss": 2.391,
+      "step": 16866
+    },
+    {
+      "epoch": 1.3612299249455249,
+      "grad_norm": 0.680969774723053,
+      "learning_rate": 1.2006792420516755e-05,
+      "loss": 2.3479,
+      "step": 16867
+    },
+    {
+      "epoch": 1.361310628682108,
+      "grad_norm": 0.682671308517456,
+      "learning_rate": 1.1999293169709757e-05,
+      "loss": 2.4097,
+      "step": 16868
+    },
+    {
+      "epoch": 1.361391332418691,
+      "grad_norm": 0.7030573487281799,
+      "learning_rate": 1.199179611211041e-05,
+      "loss": 2.4514,
+      "step": 16869
+    },
+    {
+      "epoch": 1.361472036155274,
+      "grad_norm": 0.670630693435669,
+      "learning_rate": 1.1984301247905582e-05,
+      "loss": 2.3982,
+      "step": 16870
+    },
+    {
+      "epoch": 1.361552739891857,
+      "grad_norm": 0.6993644833564758,
+      "learning_rate": 1.1976808577282017e-05,
+      "loss": 2.4297,
+      "step": 16871
+    },
+    {
+      "epoch": 1.36163344362844,
+      "grad_norm": 0.7448122501373291,
+      "learning_rate": 1.1969318100426486e-05,
+      "loss": 2.3612,
+      "step": 16872
+    },
+    {
+      "epoch": 1.361714147365023,
+      "grad_norm": 0.7014498114585876,
+      "learning_rate": 1.1961829817525649e-05,
+      "loss": 2.3451,
+      "step": 16873
+    },
+    {
+      "epoch": 1.361794851101606,
+      "grad_norm": 0.7140750885009766,
+      "learning_rate": 1.195434372876616e-05,
+      "loss": 2.4231,
+      "step": 16874
+    },
+    {
+      "epoch": 1.3618755548381891,
+      "grad_norm": 0.7377427816390991,
+      "learning_rate": 1.1946859834334567e-05,
+      "loss": 2.4055,
+      "step": 16875
+    },
+    {
+      "epoch": 1.361956258574772,
+      "grad_norm": 0.7969191670417786,
+      "learning_rate": 1.1939378134417433e-05,
+      "loss": 2.3503,
+      "step": 16876
+    },
+    {
+      "epoch": 1.362036962311355,
+      "grad_norm": 0.6821554899215698,
+      "learning_rate": 1.1931898629201155e-05,
+      "loss": 2.4259,
+      "step": 16877
+    },
+    {
+      "epoch": 1.3621176660479382,
+      "grad_norm": 0.6598221659660339,
+      "learning_rate": 1.1924421318872182e-05,
+      "loss": 2.3833,
+      "step": 16878
+    },
+    {
+      "epoch": 1.362198369784521,
+      "grad_norm": 0.8031432628631592,
+      "learning_rate": 1.1916946203616863e-05,
+      "loss": 2.5077,
+      "step": 16879
+    },
+    {
+      "epoch": 1.362279073521104,
+      "grad_norm": 0.7247405648231506,
+      "learning_rate": 1.190947328362152e-05,
+      "loss": 2.426,
+      "step": 16880
+    },
+    {
+      "epoch": 1.362359777257687,
+      "grad_norm": 0.7256691455841064,
+      "learning_rate": 1.1902002559072344e-05,
+      "loss": 2.474,
+      "step": 16881
+    },
+    {
+      "epoch": 1.36244048099427,
+      "grad_norm": 0.7382180094718933,
+      "learning_rate": 1.1894534030155558e-05,
+      "loss": 2.4487,
+      "step": 16882
+    },
+    {
+      "epoch": 1.362521184730853,
+      "grad_norm": 0.700179398059845,
+      "learning_rate": 1.1887067697057297e-05,
+      "loss": 2.3836,
+      "step": 16883
+    },
+    {
+      "epoch": 1.362601888467436,
+      "grad_norm": 0.706106424331665,
+      "learning_rate": 1.1879603559963638e-05,
+      "loss": 2.4304,
+      "step": 16884
+    },
+    {
+      "epoch": 1.362682592204019,
+      "grad_norm": 0.7514815926551819,
+      "learning_rate": 1.1872141619060606e-05,
+      "loss": 2.4895,
+      "step": 16885
+    },
+    {
+      "epoch": 1.362763295940602,
+      "grad_norm": 0.6605612635612488,
+      "learning_rate": 1.1864681874534201e-05,
+      "loss": 2.3569,
+      "step": 16886
+    },
+    {
+      "epoch": 1.362843999677185,
+      "grad_norm": 0.6366496682167053,
+      "learning_rate": 1.1857224326570283e-05,
+      "loss": 2.3919,
+      "step": 16887
+    },
+    {
+      "epoch": 1.3629247034137681,
+      "grad_norm": 0.8100820183753967,
+      "learning_rate": 1.1849768975354736e-05,
+      "loss": 2.5063,
+      "step": 16888
+    },
+    {
+      "epoch": 1.363005407150351,
+      "grad_norm": 0.685127854347229,
+      "learning_rate": 1.1842315821073403e-05,
+      "loss": 2.4647,
+      "step": 16889
+    },
+    {
+      "epoch": 1.363086110886934,
+      "grad_norm": 0.696172833442688,
+      "learning_rate": 1.1834864863911987e-05,
+      "loss": 2.4224,
+      "step": 16890
+    },
+    {
+      "epoch": 1.3631668146235172,
+      "grad_norm": 0.6558032035827637,
+      "learning_rate": 1.1827416104056199e-05,
+      "loss": 2.3619,
+      "step": 16891
+    },
+    {
+      "epoch": 1.3632475183601,
+      "grad_norm": 0.744687020778656,
+      "learning_rate": 1.1819969541691689e-05,
+      "loss": 2.4669,
+      "step": 16892
+    },
+    {
+      "epoch": 1.3633282220966831,
+      "grad_norm": 0.6925212740898132,
+      "learning_rate": 1.1812525177004052e-05,
+      "loss": 2.3967,
+      "step": 16893
+    },
+    {
+      "epoch": 1.363408925833266,
+      "grad_norm": 0.6861244440078735,
+      "learning_rate": 1.1805083010178797e-05,
+      "loss": 2.3979,
+      "step": 16894
+    },
+    {
+      "epoch": 1.363489629569849,
+      "grad_norm": 0.6987108588218689,
+      "learning_rate": 1.179764304140143e-05,
+      "loss": 2.4263,
+      "step": 16895
+    },
+    {
+      "epoch": 1.363570333306432,
+      "grad_norm": 0.6940091848373413,
+      "learning_rate": 1.179020527085738e-05,
+      "loss": 2.4328,
+      "step": 16896
+    },
+    {
+      "epoch": 1.363651037043015,
+      "grad_norm": 0.6831968426704407,
+      "learning_rate": 1.1782769698731966e-05,
+      "loss": 2.427,
+      "step": 16897
+    },
+    {
+      "epoch": 1.3637317407795981,
+      "grad_norm": 0.7370985746383667,
+      "learning_rate": 1.177533632521054e-05,
+      "loss": 2.3711,
+      "step": 16898
+    },
+    {
+      "epoch": 1.363812444516181,
+      "grad_norm": 0.8176774978637695,
+      "learning_rate": 1.1767905150478376e-05,
+      "loss": 2.4337,
+      "step": 16899
+    },
+    {
+      "epoch": 1.363893148252764,
+      "grad_norm": 0.786318302154541,
+      "learning_rate": 1.1760476174720637e-05,
+      "loss": 2.5099,
+      "step": 16900
+    },
+    {
+      "epoch": 1.3639738519893472,
+      "grad_norm": 0.7309854626655579,
+      "learning_rate": 1.1753049398122495e-05,
+      "loss": 2.46,
+      "step": 16901
+    },
+    {
+      "epoch": 1.36405455572593,
+      "grad_norm": 0.7410863637924194,
+      "learning_rate": 1.1745624820869039e-05,
+      "loss": 2.4249,
+      "step": 16902
+    },
+    {
+      "epoch": 1.3641352594625131,
+      "grad_norm": 0.7059988379478455,
+      "learning_rate": 1.1738202443145308e-05,
+      "loss": 2.4964,
+      "step": 16903
+    },
+    {
+      "epoch": 1.3642159631990962,
+      "grad_norm": 0.7351845502853394,
+      "learning_rate": 1.1730782265136287e-05,
+      "loss": 2.4694,
+      "step": 16904
+    },
+    {
+      "epoch": 1.364296666935679,
+      "grad_norm": 0.6928153038024902,
+      "learning_rate": 1.1723364287026938e-05,
+      "loss": 2.426,
+      "step": 16905
+    },
+    {
+      "epoch": 1.3643773706722622,
+      "grad_norm": 0.759920060634613,
+      "learning_rate": 1.1715948509002083e-05,
+      "loss": 2.4359,
+      "step": 16906
+    },
+    {
+      "epoch": 1.3644580744088453,
+      "grad_norm": 0.6655696630477905,
+      "learning_rate": 1.1708534931246573e-05,
+      "loss": 2.4118,
+      "step": 16907
+    },
+    {
+      "epoch": 1.3645387781454281,
+      "grad_norm": 0.6912528872489929,
+      "learning_rate": 1.170112355394517e-05,
+      "loss": 2.4257,
+      "step": 16908
+    },
+    {
+      "epoch": 1.3646194818820112,
+      "grad_norm": 0.6612871289253235,
+      "learning_rate": 1.1693714377282604e-05,
+      "loss": 2.4192,
+      "step": 16909
+    },
+    {
+      "epoch": 1.364700185618594,
+      "grad_norm": 0.6548018455505371,
+      "learning_rate": 1.1686307401443486e-05,
+      "loss": 2.4054,
+      "step": 16910
+    },
+    {
+      "epoch": 1.3647808893551772,
+      "grad_norm": 0.7749961018562317,
+      "learning_rate": 1.1678902626612443e-05,
+      "loss": 2.44,
+      "step": 16911
+    },
+    {
+      "epoch": 1.36486159309176,
+      "grad_norm": 0.7187496423721313,
+      "learning_rate": 1.1671500052974039e-05,
+      "loss": 2.4033,
+      "step": 16912
+    },
+    {
+      "epoch": 1.3649422968283431,
+      "grad_norm": 0.7002814412117004,
+      "learning_rate": 1.1664099680712715e-05,
+      "loss": 2.4442,
+      "step": 16913
+    },
+    {
+      "epoch": 1.3650230005649262,
+      "grad_norm": 0.6852529644966125,
+      "learning_rate": 1.1656701510012946e-05,
+      "loss": 2.4253,
+      "step": 16914
+    },
+    {
+      "epoch": 1.365103704301509,
+      "grad_norm": 0.6922035813331604,
+      "learning_rate": 1.1649305541059142e-05,
+      "loss": 2.4406,
+      "step": 16915
+    },
+    {
+      "epoch": 1.3651844080380922,
+      "grad_norm": 0.6883397698402405,
+      "learning_rate": 1.1641911774035563e-05,
+      "loss": 2.4064,
+      "step": 16916
+    },
+    {
+      "epoch": 1.3652651117746752,
+      "grad_norm": 0.7101531624794006,
+      "learning_rate": 1.163452020912652e-05,
+      "loss": 2.4068,
+      "step": 16917
+    },
+    {
+      "epoch": 1.365345815511258,
+      "grad_norm": 0.728369951248169,
+      "learning_rate": 1.1627130846516231e-05,
+      "loss": 2.4319,
+      "step": 16918
+    },
+    {
+      "epoch": 1.3654265192478412,
+      "grad_norm": 0.6765053272247314,
+      "learning_rate": 1.161974368638884e-05,
+      "loss": 2.3922,
+      "step": 16919
+    },
+    {
+      "epoch": 1.3655072229844243,
+      "grad_norm": 0.6909242868423462,
+      "learning_rate": 1.1612358728928475e-05,
+      "loss": 2.4124,
+      "step": 16920
+    },
+    {
+      "epoch": 1.3655879267210072,
+      "grad_norm": 0.735650897026062,
+      "learning_rate": 1.1604975974319177e-05,
+      "loss": 2.5137,
+      "step": 16921
+    },
+    {
+      "epoch": 1.3656686304575902,
+      "grad_norm": 0.6587653756141663,
+      "learning_rate": 1.1597595422744934e-05,
+      "loss": 2.4163,
+      "step": 16922
+    },
+    {
+      "epoch": 1.3657493341941733,
+      "grad_norm": 0.700282096862793,
+      "learning_rate": 1.159021707438971e-05,
+      "loss": 2.4272,
+      "step": 16923
+    },
+    {
+      "epoch": 1.3658300379307562,
+      "grad_norm": 0.7175682783126831,
+      "learning_rate": 1.1582840929437365e-05,
+      "loss": 2.4598,
+      "step": 16924
+    },
+    {
+      "epoch": 1.3659107416673393,
+      "grad_norm": 0.6725881695747375,
+      "learning_rate": 1.157546698807176e-05,
+      "loss": 2.4064,
+      "step": 16925
+    },
+    {
+      "epoch": 1.3659914454039221,
+      "grad_norm": 0.7130467295646667,
+      "learning_rate": 1.1568095250476651e-05,
+      "loss": 2.3851,
+      "step": 16926
+    },
+    {
+      "epoch": 1.3660721491405052,
+      "grad_norm": 0.6859269142150879,
+      "learning_rate": 1.1560725716835785e-05,
+      "loss": 2.3577,
+      "step": 16927
+    },
+    {
+      "epoch": 1.366152852877088,
+      "grad_norm": 0.7037541270256042,
+      "learning_rate": 1.1553358387332824e-05,
+      "loss": 2.4402,
+      "step": 16928
+    },
+    {
+      "epoch": 1.3662335566136712,
+      "grad_norm": 0.7094031572341919,
+      "learning_rate": 1.1545993262151366e-05,
+      "loss": 2.4036,
+      "step": 16929
+    },
+    {
+      "epoch": 1.3663142603502543,
+      "grad_norm": 0.6953302025794983,
+      "learning_rate": 1.1538630341474965e-05,
+      "loss": 2.4192,
+      "step": 16930
+    },
+    {
+      "epoch": 1.3663949640868371,
+      "grad_norm": 0.7012252807617188,
+      "learning_rate": 1.1531269625487163e-05,
+      "loss": 2.4207,
+      "step": 16931
+    },
+    {
+      "epoch": 1.3664756678234202,
+      "grad_norm": 0.6616495847702026,
+      "learning_rate": 1.1523911114371366e-05,
+      "loss": 2.4187,
+      "step": 16932
+    },
+    {
+      "epoch": 1.3665563715600033,
+      "grad_norm": 0.6819868087768555,
+      "learning_rate": 1.1516554808310975e-05,
+      "loss": 2.448,
+      "step": 16933
+    },
+    {
+      "epoch": 1.3666370752965862,
+      "grad_norm": 0.6869969964027405,
+      "learning_rate": 1.1509200707489343e-05,
+      "loss": 2.4134,
+      "step": 16934
+    },
+    {
+      "epoch": 1.3667177790331693,
+      "grad_norm": 0.6600778698921204,
+      "learning_rate": 1.1501848812089733e-05,
+      "loss": 2.4159,
+      "step": 16935
+    },
+    {
+      "epoch": 1.3667984827697524,
+      "grad_norm": 0.668712317943573,
+      "learning_rate": 1.1494499122295398e-05,
+      "loss": 2.41,
+      "step": 16936
+    },
+    {
+      "epoch": 1.3668791865063352,
+      "grad_norm": 0.767365574836731,
+      "learning_rate": 1.1487151638289518e-05,
+      "loss": 2.3856,
+      "step": 16937
+    },
+    {
+      "epoch": 1.3669598902429183,
+      "grad_norm": 0.721546471118927,
+      "learning_rate": 1.1479806360255174e-05,
+      "loss": 2.4038,
+      "step": 16938
+    },
+    {
+      "epoch": 1.3670405939795012,
+      "grad_norm": 0.6796963810920715,
+      "learning_rate": 1.1472463288375456e-05,
+      "loss": 2.3698,
+      "step": 16939
+    },
+    {
+      "epoch": 1.3671212977160843,
+      "grad_norm": 0.7340671420097351,
+      "learning_rate": 1.1465122422833363e-05,
+      "loss": 2.4296,
+      "step": 16940
+    },
+    {
+      "epoch": 1.3672020014526671,
+      "grad_norm": 0.7173369526863098,
+      "learning_rate": 1.145778376381187e-05,
+      "loss": 2.3923,
+      "step": 16941
+    },
+    {
+      "epoch": 1.3672827051892502,
+      "grad_norm": 0.6683956980705261,
+      "learning_rate": 1.1450447311493839e-05,
+      "loss": 2.4092,
+      "step": 16942
+    },
+    {
+      "epoch": 1.3673634089258333,
+      "grad_norm": 0.6457851529121399,
+      "learning_rate": 1.1443113066062129e-05,
+      "loss": 2.3467,
+      "step": 16943
+    },
+    {
+      "epoch": 1.3674441126624162,
+      "grad_norm": 0.6870608925819397,
+      "learning_rate": 1.1435781027699532e-05,
+      "loss": 2.3766,
+      "step": 16944
+    },
+    {
+      "epoch": 1.3675248163989993,
+      "grad_norm": 0.6496049165725708,
+      "learning_rate": 1.1428451196588775e-05,
+      "loss": 2.4464,
+      "step": 16945
+    },
+    {
+      "epoch": 1.3676055201355823,
+      "grad_norm": 0.7554739117622375,
+      "learning_rate": 1.1421123572912551e-05,
+      "loss": 2.4243,
+      "step": 16946
+    },
+    {
+      "epoch": 1.3676862238721652,
+      "grad_norm": 0.7208122611045837,
+      "learning_rate": 1.1413798156853495e-05,
+      "loss": 2.3699,
+      "step": 16947
+    },
+    {
+      "epoch": 1.3677669276087483,
+      "grad_norm": 0.7072176337242126,
+      "learning_rate": 1.1406474948594126e-05,
+      "loss": 2.4011,
+      "step": 16948
+    },
+    {
+      "epoch": 1.3678476313453314,
+      "grad_norm": 0.7316476106643677,
+      "learning_rate": 1.1399153948316999e-05,
+      "loss": 2.4508,
+      "step": 16949
+    },
+    {
+      "epoch": 1.3679283350819142,
+      "grad_norm": 0.8518069386482239,
+      "learning_rate": 1.1391835156204577e-05,
+      "loss": 2.4197,
+      "step": 16950
+    },
+    {
+      "epoch": 1.3680090388184973,
+      "grad_norm": 0.6700364947319031,
+      "learning_rate": 1.1384518572439228e-05,
+      "loss": 2.4272,
+      "step": 16951
+    },
+    {
+      "epoch": 1.3680897425550804,
+      "grad_norm": 0.7007749676704407,
+      "learning_rate": 1.1377204197203317e-05,
+      "loss": 2.3777,
+      "step": 16952
+    },
+    {
+      "epoch": 1.3681704462916633,
+      "grad_norm": 0.6792053580284119,
+      "learning_rate": 1.1369892030679141e-05,
+      "loss": 2.4487,
+      "step": 16953
+    },
+    {
+      "epoch": 1.3682511500282464,
+      "grad_norm": 0.6913022398948669,
+      "learning_rate": 1.1362582073048932e-05,
+      "loss": 2.3757,
+      "step": 16954
+    },
+    {
+      "epoch": 1.3683318537648292,
+      "grad_norm": 0.648248016834259,
+      "learning_rate": 1.135527432449488e-05,
+      "loss": 2.3482,
+      "step": 16955
+    },
+    {
+      "epoch": 1.3684125575014123,
+      "grad_norm": 0.6711798906326294,
+      "learning_rate": 1.1347968785199115e-05,
+      "loss": 2.4096,
+      "step": 16956
+    },
+    {
+      "epoch": 1.3684932612379952,
+      "grad_norm": 0.6932381987571716,
+      "learning_rate": 1.1340665455343724e-05,
+      "loss": 2.3834,
+      "step": 16957
+    },
+    {
+      "epoch": 1.3685739649745783,
+      "grad_norm": 0.6890178918838501,
+      "learning_rate": 1.1333364335110697e-05,
+      "loss": 2.4182,
+      "step": 16958
+    },
+    {
+      "epoch": 1.3686546687111614,
+      "grad_norm": 0.6612519025802612,
+      "learning_rate": 1.1326065424681997e-05,
+      "loss": 2.3691,
+      "step": 16959
+    },
+    {
+      "epoch": 1.3687353724477442,
+      "grad_norm": 0.7123190760612488,
+      "learning_rate": 1.131876872423957e-05,
+      "loss": 2.3919,
+      "step": 16960
+    },
+    {
+      "epoch": 1.3688160761843273,
+      "grad_norm": 0.6615463495254517,
+      "learning_rate": 1.1311474233965214e-05,
+      "loss": 2.4266,
+      "step": 16961
+    },
+    {
+      "epoch": 1.3688967799209104,
+      "grad_norm": 0.7320190668106079,
+      "learning_rate": 1.130418195404076e-05,
+      "loss": 2.4268,
+      "step": 16962
+    },
+    {
+      "epoch": 1.3689774836574933,
+      "grad_norm": 0.6845116019248962,
+      "learning_rate": 1.1296891884647965e-05,
+      "loss": 2.3972,
+      "step": 16963
+    },
+    {
+      "epoch": 1.3690581873940764,
+      "grad_norm": 0.70455002784729,
+      "learning_rate": 1.1289604025968448e-05,
+      "loss": 2.4183,
+      "step": 16964
+    },
+    {
+      "epoch": 1.3691388911306595,
+      "grad_norm": 0.6952407956123352,
+      "learning_rate": 1.128231837818392e-05,
+      "loss": 2.4276,
+      "step": 16965
+    },
+    {
+      "epoch": 1.3692195948672423,
+      "grad_norm": 0.7939464449882507,
+      "learning_rate": 1.1275034941475938e-05,
+      "loss": 2.4072,
+      "step": 16966
+    },
+    {
+      "epoch": 1.3693002986038254,
+      "grad_norm": 0.6974930763244629,
+      "learning_rate": 1.1267753716026007e-05,
+      "loss": 2.4133,
+      "step": 16967
+    },
+    {
+      "epoch": 1.3693810023404085,
+      "grad_norm": 0.7187508344650269,
+      "learning_rate": 1.126047470201559e-05,
+      "loss": 2.3588,
+      "step": 16968
+    },
+    {
+      "epoch": 1.3694617060769914,
+      "grad_norm": 0.6887609958648682,
+      "learning_rate": 1.1253197899626134e-05,
+      "loss": 2.4322,
+      "step": 16969
+    },
+    {
+      "epoch": 1.3695424098135744,
+      "grad_norm": 0.679957389831543,
+      "learning_rate": 1.1245923309038964e-05,
+      "loss": 2.3907,
+      "step": 16970
+    },
+    {
+      "epoch": 1.3696231135501573,
+      "grad_norm": 0.7540870308876038,
+      "learning_rate": 1.1238650930435378e-05,
+      "loss": 2.4752,
+      "step": 16971
+    },
+    {
+      "epoch": 1.3697038172867404,
+      "grad_norm": 0.7697634100914001,
+      "learning_rate": 1.1231380763996635e-05,
+      "loss": 2.4366,
+      "step": 16972
+    },
+    {
+      "epoch": 1.3697845210233233,
+      "grad_norm": 0.6836850643157959,
+      "learning_rate": 1.1224112809903954e-05,
+      "loss": 2.3511,
+      "step": 16973
+    },
+    {
+      "epoch": 1.3698652247599064,
+      "grad_norm": 0.6904506683349609,
+      "learning_rate": 1.1216847068338421e-05,
+      "loss": 2.4109,
+      "step": 16974
+    },
+    {
+      "epoch": 1.3699459284964894,
+      "grad_norm": 0.6579318046569824,
+      "learning_rate": 1.1209583539481127e-05,
+      "loss": 2.4391,
+      "step": 16975
+    },
+    {
+      "epoch": 1.3700266322330723,
+      "grad_norm": 0.7107192277908325,
+      "learning_rate": 1.120232222351314e-05,
+      "loss": 2.399,
+      "step": 16976
+    },
+    {
+      "epoch": 1.3701073359696554,
+      "grad_norm": 0.7581583261489868,
+      "learning_rate": 1.119506312061539e-05,
+      "loss": 2.4817,
+      "step": 16977
+    },
+    {
+      "epoch": 1.3701880397062385,
+      "grad_norm": 0.6836642622947693,
+      "learning_rate": 1.11878062309688e-05,
+      "loss": 2.4415,
+      "step": 16978
+    },
+    {
+      "epoch": 1.3702687434428213,
+      "grad_norm": 0.6842699646949768,
+      "learning_rate": 1.118055155475426e-05,
+      "loss": 2.4045,
+      "step": 16979
+    },
+    {
+      "epoch": 1.3703494471794044,
+      "grad_norm": 0.7630519270896912,
+      "learning_rate": 1.1173299092152534e-05,
+      "loss": 2.4314,
+      "step": 16980
+    },
+    {
+      "epoch": 1.3704301509159875,
+      "grad_norm": 0.7334303259849548,
+      "learning_rate": 1.116604884334439e-05,
+      "loss": 2.3564,
+      "step": 16981
+    },
+    {
+      "epoch": 1.3705108546525704,
+      "grad_norm": 0.6929439306259155,
+      "learning_rate": 1.1158800808510538e-05,
+      "loss": 2.4258,
+      "step": 16982
+    },
+    {
+      "epoch": 1.3705915583891535,
+      "grad_norm": 0.6387187838554382,
+      "learning_rate": 1.1151554987831591e-05,
+      "loss": 2.3263,
+      "step": 16983
+    },
+    {
+      "epoch": 1.3706722621257363,
+      "grad_norm": 0.7279032468795776,
+      "learning_rate": 1.1144311381488136e-05,
+      "loss": 2.4074,
+      "step": 16984
+    },
+    {
+      "epoch": 1.3707529658623194,
+      "grad_norm": 0.7066916227340698,
+      "learning_rate": 1.113706998966072e-05,
+      "loss": 2.4358,
+      "step": 16985
+    },
+    {
+      "epoch": 1.3708336695989023,
+      "grad_norm": 0.6753098964691162,
+      "learning_rate": 1.1129830812529807e-05,
+      "loss": 2.4195,
+      "step": 16986
+    },
+    {
+      "epoch": 1.3709143733354854,
+      "grad_norm": 0.6728894114494324,
+      "learning_rate": 1.112259385027582e-05,
+      "loss": 2.3712,
+      "step": 16987
+    },
+    {
+      "epoch": 1.3709950770720685,
+      "grad_norm": 0.7251775860786438,
+      "learning_rate": 1.1115359103079115e-05,
+      "loss": 2.4063,
+      "step": 16988
+    },
+    {
+      "epoch": 1.3710757808086513,
+      "grad_norm": 0.6797254085540771,
+      "learning_rate": 1.1108126571120036e-05,
+      "loss": 2.395,
+      "step": 16989
+    },
+    {
+      "epoch": 1.3711564845452344,
+      "grad_norm": 0.7505605220794678,
+      "learning_rate": 1.1100896254578786e-05,
+      "loss": 2.4044,
+      "step": 16990
+    },
+    {
+      "epoch": 1.3712371882818175,
+      "grad_norm": 0.7126416563987732,
+      "learning_rate": 1.1093668153635594e-05,
+      "loss": 2.4043,
+      "step": 16991
+    },
+    {
+      "epoch": 1.3713178920184004,
+      "grad_norm": 0.6550771594047546,
+      "learning_rate": 1.1086442268470609e-05,
+      "loss": 2.3515,
+      "step": 16992
+    },
+    {
+      "epoch": 1.3713985957549835,
+      "grad_norm": 0.7253621816635132,
+      "learning_rate": 1.1079218599263874e-05,
+      "loss": 2.4109,
+      "step": 16993
+    },
+    {
+      "epoch": 1.3714792994915666,
+      "grad_norm": 0.7272186875343323,
+      "learning_rate": 1.1071997146195468e-05,
+      "loss": 2.3531,
+      "step": 16994
+    },
+    {
+      "epoch": 1.3715600032281494,
+      "grad_norm": 0.6841129660606384,
+      "learning_rate": 1.1064777909445345e-05,
+      "loss": 2.4031,
+      "step": 16995
+    },
+    {
+      "epoch": 1.3716407069647325,
+      "grad_norm": 0.692945659160614,
+      "learning_rate": 1.1057560889193441e-05,
+      "loss": 2.3858,
+      "step": 16996
+    },
+    {
+      "epoch": 1.3717214107013156,
+      "grad_norm": 0.721182644367218,
+      "learning_rate": 1.1050346085619612e-05,
+      "loss": 2.3871,
+      "step": 16997
+    },
+    {
+      "epoch": 1.3718021144378985,
+      "grad_norm": 0.722960889339447,
+      "learning_rate": 1.1043133498903702e-05,
+      "loss": 2.3452,
+      "step": 16998
+    },
+    {
+      "epoch": 1.3718828181744815,
+      "grad_norm": 0.7148451805114746,
+      "learning_rate": 1.1035923129225412e-05,
+      "loss": 2.3905,
+      "step": 16999
+    },
+    {
+      "epoch": 1.3719635219110644,
+      "grad_norm": 0.7118532061576843,
+      "learning_rate": 1.1028714976764486e-05,
+      "loss": 2.3894,
+      "step": 17000
+    },
+    {
+      "epoch": 1.3719635219110644,
+      "eval_loss": 2.3730249404907227,
+      "eval_runtime": 769.4165,
+      "eval_samples_per_second": 3.405,
+      "eval_steps_per_second": 0.568,
+      "step": 17000
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 20000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.95407905485312e+17,
+  "train_batch_size": 6,
+  "trial_name": null,
+  "trial_params": null
+}