diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,54714 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.999936,
+  "eval_steps": 500,
+  "global_step": 7812,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.000128,
+      "grad_norm": 129.30274963378906,
+      "learning_rate": 4.255319148936171e-06,
+      "loss": 10.5394,
+      "step": 1
+    },
+    {
+      "epoch": 0.000256,
+      "grad_norm": 116.76910400390625,
+      "learning_rate": 8.510638297872341e-06,
+      "loss": 10.4539,
+      "step": 2
+    },
+    {
+      "epoch": 0.000384,
+      "grad_norm": 111.25746154785156,
+      "learning_rate": 1.2765957446808511e-05,
+      "loss": 10.3586,
+      "step": 3
+    },
+    {
+      "epoch": 0.000512,
+      "grad_norm": 100.38201904296875,
+      "learning_rate": 1.7021276595744682e-05,
+      "loss": 9.5197,
+      "step": 4
+    },
+    {
+      "epoch": 0.00064,
+      "grad_norm": 41.0691032409668,
+      "learning_rate": 2.1276595744680852e-05,
+      "loss": 8.1775,
+      "step": 5
+    },
+    {
+      "epoch": 0.000768,
+      "grad_norm": 27.273649215698242,
+      "learning_rate": 2.5531914893617022e-05,
+      "loss": 7.3553,
+      "step": 6
+    },
+    {
+      "epoch": 0.000896,
+      "grad_norm": 22.083667755126953,
+      "learning_rate": 2.9787234042553195e-05,
+      "loss": 6.7702,
+      "step": 7
+    },
+    {
+      "epoch": 0.001024,
+      "grad_norm": 19.924449920654297,
+      "learning_rate": 3.4042553191489365e-05,
+      "loss": 6.7761,
+      "step": 8
+    },
+    {
+      "epoch": 0.001152,
+      "grad_norm": 16.61648178100586,
+      "learning_rate": 3.829787234042553e-05,
+      "loss": 6.2348,
+      "step": 9
+    },
+    {
+      "epoch": 0.00128,
+      "grad_norm": 15.290637016296387,
+      "learning_rate": 4.2553191489361704e-05,
+      "loss": 5.913,
+      "step": 10
+    },
+    {
+      "epoch": 0.001408,
+      "grad_norm": 12.637874603271484,
+      "learning_rate": 4.680851063829787e-05,
+      "loss": 5.7432,
+      "step": 11
+    },
+    {
+      "epoch": 0.001536,
+      "grad_norm": 8.566034317016602,
+      "learning_rate": 5.1063829787234044e-05,
+      "loss": 5.6379,
+      "step": 12
+    },
+    {
+      "epoch": 0.001664,
+      "grad_norm": 6.990787982940674,
+      "learning_rate": 5.531914893617021e-05,
+      "loss": 5.3746,
+      "step": 13
+    },
+    {
+      "epoch": 0.001792,
+      "grad_norm": 5.690971374511719,
+      "learning_rate": 5.957446808510639e-05,
+      "loss": 5.2501,
+      "step": 14
+    },
+    {
+      "epoch": 0.00192,
+      "grad_norm": 6.8315606117248535,
+      "learning_rate": 6.382978723404255e-05,
+      "loss": 5.1217,
+      "step": 15
+    },
+    {
+      "epoch": 0.002048,
+      "grad_norm": 6.799686908721924,
+      "learning_rate": 6.808510638297873e-05,
+      "loss": 4.996,
+      "step": 16
+    },
+    {
+      "epoch": 0.002176,
+      "grad_norm": 8.46674633026123,
+      "learning_rate": 7.234042553191488e-05,
+      "loss": 4.9009,
+      "step": 17
+    },
+    {
+      "epoch": 0.002304,
+      "grad_norm": 6.133491516113281,
+      "learning_rate": 7.659574468085106e-05,
+      "loss": 4.7067,
+      "step": 18
+    },
+    {
+      "epoch": 0.002432,
+      "grad_norm": 4.384441375732422,
+      "learning_rate": 8.085106382978724e-05,
+      "loss": 4.5724,
+      "step": 19
+    },
+    {
+      "epoch": 0.00256,
+      "grad_norm": 3.051910161972046,
+      "learning_rate": 8.510638297872341e-05,
+      "loss": 4.5241,
+      "step": 20
+    },
+    {
+      "epoch": 0.002688,
+      "grad_norm": 2.1846039295196533,
+      "learning_rate": 8.936170212765958e-05,
+      "loss": 4.4502,
+      "step": 21
+    },
+    {
+      "epoch": 0.002816,
+      "grad_norm": 1.9885467290878296,
+      "learning_rate": 9.361702127659574e-05,
+      "loss": 4.4328,
+      "step": 22
+    },
+    {
+      "epoch": 0.002944,
+      "grad_norm": 1.640302300453186,
+      "learning_rate": 9.787234042553192e-05,
+      "loss": 4.3293,
+      "step": 23
+    },
+    {
+      "epoch": 0.003072,
+      "grad_norm": 1.5435653924942017,
+      "learning_rate": 0.00010212765957446809,
+      "loss": 4.1607,
+      "step": 24
+    },
+    {
+      "epoch": 0.0032,
+      "grad_norm": 1.6526771783828735,
+      "learning_rate": 0.00010638297872340425,
+      "loss": 4.47,
+      "step": 25
+    },
+    {
+      "epoch": 0.003328,
+      "grad_norm": 1.5372318029403687,
+      "learning_rate": 0.00011063829787234042,
+      "loss": 4.5156,
+      "step": 26
+    },
+    {
+      "epoch": 0.003456,
+      "grad_norm": 1.3895479440689087,
+      "learning_rate": 0.0001148936170212766,
+      "loss": 4.1921,
+      "step": 27
+    },
+    {
+      "epoch": 0.003584,
+      "grad_norm": 1.311293125152588,
+      "learning_rate": 0.00011914893617021278,
+      "loss": 4.2487,
+      "step": 28
+    },
+    {
+      "epoch": 0.003712,
+      "grad_norm": 1.331913948059082,
+      "learning_rate": 0.00012340425531914893,
+      "loss": 4.1842,
+      "step": 29
+    },
+    {
+      "epoch": 0.00384,
+      "grad_norm": 1.1350797414779663,
+      "learning_rate": 0.0001276595744680851,
+      "loss": 4.2377,
+      "step": 30
+    },
+    {
+      "epoch": 0.003968,
+      "grad_norm": 1.260619878768921,
+      "learning_rate": 0.00013191489361702127,
+      "loss": 4.3074,
+      "step": 31
+    },
+    {
+      "epoch": 0.004096,
+      "grad_norm": 1.081208348274231,
+      "learning_rate": 0.00013617021276595746,
+      "loss": 4.3815,
+      "step": 32
+    },
+    {
+      "epoch": 0.004224,
+      "grad_norm": 1.059191346168518,
+      "learning_rate": 0.00014042553191489363,
+      "loss": 4.2211,
+      "step": 33
+    },
+    {
+      "epoch": 0.004352,
+      "grad_norm": 1.1371310949325562,
+      "learning_rate": 0.00014468085106382977,
+      "loss": 4.1937,
+      "step": 34
+    },
+    {
+      "epoch": 0.00448,
+      "grad_norm": 0.930898904800415,
+      "learning_rate": 0.00014893617021276596,
+      "loss": 4.1093,
+      "step": 35
+    },
+    {
+      "epoch": 0.004608,
+      "grad_norm": 0.9123896956443787,
+      "learning_rate": 0.00015319148936170213,
+      "loss": 4.1043,
+      "step": 36
+    },
+    {
+      "epoch": 0.004736,
+      "grad_norm": 0.843893826007843,
+      "learning_rate": 0.00015744680851063832,
+      "loss": 4.0225,
+      "step": 37
+    },
+    {
+      "epoch": 0.004864,
+      "grad_norm": 0.8537824749946594,
+      "learning_rate": 0.00016170212765957449,
+      "loss": 4.1961,
+      "step": 38
+    },
+    {
+      "epoch": 0.004992,
+      "grad_norm": 0.7867328524589539,
+      "learning_rate": 0.00016595744680851062,
+      "loss": 4.1816,
+      "step": 39
+    },
+    {
+      "epoch": 0.00512,
+      "grad_norm": 0.7540923953056335,
+      "learning_rate": 0.00017021276595744682,
+      "loss": 4.174,
+      "step": 40
+    },
+    {
+      "epoch": 0.005248,
+      "grad_norm": 0.7507907152175903,
+      "learning_rate": 0.00017446808510638298,
+      "loss": 3.9306,
+      "step": 41
+    },
+    {
+      "epoch": 0.005376,
+      "grad_norm": 0.7091206908226013,
+      "learning_rate": 0.00017872340425531915,
+      "loss": 4.2432,
+      "step": 42
+    },
+    {
+      "epoch": 0.005504,
+      "grad_norm": 0.6981675624847412,
+      "learning_rate": 0.00018297872340425532,
+      "loss": 4.1774,
+      "step": 43
+    },
+    {
+      "epoch": 0.005632,
+      "grad_norm": 0.6921675801277161,
+      "learning_rate": 0.00018723404255319148,
+      "loss": 4.157,
+      "step": 44
+    },
+    {
+      "epoch": 0.00576,
+      "grad_norm": 0.7346277236938477,
+      "learning_rate": 0.00019148936170212765,
+      "loss": 4.146,
+      "step": 45
+    },
+    {
+      "epoch": 0.005888,
+      "grad_norm": 0.6094359755516052,
+      "learning_rate": 0.00019574468085106384,
+      "loss": 4.239,
+      "step": 46
+    },
+    {
+      "epoch": 0.006016,
+      "grad_norm": 0.7275527119636536,
+      "learning_rate": 0.0002,
+      "loss": 4.1157,
+      "step": 47
+    },
+    {
+      "epoch": 0.006144,
+      "grad_norm": 0.5808286666870117,
+      "learning_rate": 0.00020425531914893618,
+      "loss": 4.2271,
+      "step": 48
+    },
+    {
+      "epoch": 0.006272,
+      "grad_norm": 0.5874955654144287,
+      "learning_rate": 0.00020851063829787234,
+      "loss": 4.061,
+      "step": 49
+    },
+    {
+      "epoch": 0.0064,
+      "grad_norm": 0.5500697493553162,
+      "learning_rate": 0.0002127659574468085,
+      "loss": 3.8623,
+      "step": 50
+    },
+    {
+      "epoch": 0.006528,
+      "grad_norm": 0.6067680716514587,
+      "learning_rate": 0.0002170212765957447,
+      "loss": 4.2082,
+      "step": 51
+    },
+    {
+      "epoch": 0.006656,
+      "grad_norm": 0.5269648432731628,
+      "learning_rate": 0.00022127659574468084,
+      "loss": 4.038,
+      "step": 52
+    },
+    {
+      "epoch": 0.006784,
+      "grad_norm": 0.5667698979377747,
+      "learning_rate": 0.000225531914893617,
+      "loss": 4.1465,
+      "step": 53
+    },
+    {
+      "epoch": 0.006912,
+      "grad_norm": 0.6308124661445618,
+      "learning_rate": 0.0002297872340425532,
+      "loss": 4.2249,
+      "step": 54
+    },
+    {
+      "epoch": 0.00704,
+      "grad_norm": 0.49904540181159973,
+      "learning_rate": 0.00023404255319148937,
+      "loss": 4.0785,
+      "step": 55
+    },
+    {
+      "epoch": 0.007168,
+      "grad_norm": 0.5441160202026367,
+      "learning_rate": 0.00023829787234042556,
+      "loss": 4.1333,
+      "step": 56
+    },
+    {
+      "epoch": 0.007296,
+      "grad_norm": 0.5134733319282532,
+      "learning_rate": 0.0002425531914893617,
+      "loss": 4.0535,
+      "step": 57
+    },
+    {
+      "epoch": 0.007424,
+      "grad_norm": 0.498975932598114,
+      "learning_rate": 0.00024680851063829787,
+      "loss": 4.1288,
+      "step": 58
+    },
+    {
+      "epoch": 0.007552,
+      "grad_norm": 0.44294339418411255,
+      "learning_rate": 0.00025106382978723403,
+      "loss": 3.9465,
+      "step": 59
+    },
+    {
+      "epoch": 0.00768,
+      "grad_norm": 0.8531392216682434,
+      "learning_rate": 0.0002553191489361702,
+      "loss": 3.9497,
+      "step": 60
+    },
+    {
+      "epoch": 0.007808,
+      "grad_norm": 0.4884328544139862,
+      "learning_rate": 0.00025957446808510637,
+      "loss": 4.091,
+      "step": 61
+    },
+    {
+      "epoch": 0.007936,
+      "grad_norm": 0.5359905362129211,
+      "learning_rate": 0.00026382978723404253,
+      "loss": 4.1551,
+      "step": 62
+    },
+    {
+      "epoch": 0.008064,
+      "grad_norm": 0.44776439666748047,
+      "learning_rate": 0.00026808510638297875,
+      "loss": 4.21,
+      "step": 63
+    },
+    {
+      "epoch": 0.008192,
+      "grad_norm": 0.4573459327220917,
+      "learning_rate": 0.0002723404255319149,
+      "loss": 4.1015,
+      "step": 64
+    },
+    {
+      "epoch": 0.00832,
+      "grad_norm": 0.4639029800891876,
+      "learning_rate": 0.0002765957446808511,
+      "loss": 4.1207,
+      "step": 65
+    },
+    {
+      "epoch": 0.008448,
+      "grad_norm": 0.451500803232193,
+      "learning_rate": 0.00028085106382978725,
+      "loss": 3.9107,
+      "step": 66
+    },
+    {
+      "epoch": 0.008576,
+      "grad_norm": 0.5121942758560181,
+      "learning_rate": 0.0002851063829787234,
+      "loss": 4.1452,
+      "step": 67
+    },
+    {
+      "epoch": 0.008704,
+      "grad_norm": 0.46146151423454285,
+      "learning_rate": 0.00028936170212765953,
+      "loss": 4.011,
+      "step": 68
+    },
+    {
+      "epoch": 0.008832,
+      "grad_norm": 0.48989012837409973,
+      "learning_rate": 0.00029361702127659575,
+      "loss": 4.0371,
+      "step": 69
+    },
+    {
+      "epoch": 0.00896,
+      "grad_norm": 0.4731649160385132,
+      "learning_rate": 0.0002978723404255319,
+      "loss": 4.0406,
+      "step": 70
+    },
+    {
+      "epoch": 0.009088,
+      "grad_norm": 0.4796810746192932,
+      "learning_rate": 0.0003021276595744681,
+      "loss": 4.0839,
+      "step": 71
+    },
+    {
+      "epoch": 0.009216,
+      "grad_norm": 0.4134310483932495,
+      "learning_rate": 0.00030638297872340425,
+      "loss": 3.8444,
+      "step": 72
+    },
+    {
+      "epoch": 0.009344,
+      "grad_norm": 0.456488698720932,
+      "learning_rate": 0.0003106382978723404,
+      "loss": 4.1335,
+      "step": 73
+    },
+    {
+      "epoch": 0.009472,
+      "grad_norm": 0.4094163775444031,
+      "learning_rate": 0.00031489361702127664,
+      "loss": 4.0029,
+      "step": 74
+    },
+    {
+      "epoch": 0.0096,
+      "grad_norm": 0.5195801258087158,
+      "learning_rate": 0.0003191489361702128,
+      "loss": 4.0127,
+      "step": 75
+    },
+    {
+      "epoch": 0.009728,
+      "grad_norm": 0.5037373304367065,
+      "learning_rate": 0.00032340425531914897,
+      "loss": 4.1532,
+      "step": 76
+    },
+    {
+      "epoch": 0.009856,
+      "grad_norm": 0.42778733372688293,
+      "learning_rate": 0.0003276595744680851,
+      "loss": 4.046,
+      "step": 77
+    },
+    {
+      "epoch": 0.009984,
+      "grad_norm": 0.4622833728790283,
+      "learning_rate": 0.00033191489361702125,
+      "loss": 3.9358,
+      "step": 78
+    },
+    {
+      "epoch": 0.010112,
+      "grad_norm": 0.44430774450302124,
+      "learning_rate": 0.0003361702127659574,
+      "loss": 4.0158,
+      "step": 79
+    },
+    {
+      "epoch": 0.01024,
+      "grad_norm": 0.4277341663837433,
+      "learning_rate": 0.00034042553191489364,
+      "loss": 4.0791,
+      "step": 80
+    },
+    {
+      "epoch": 0.010368,
+      "grad_norm": 0.5889245271682739,
+      "learning_rate": 0.0003446808510638298,
+      "loss": 3.9415,
+      "step": 81
+    },
+    {
+      "epoch": 0.010496,
+      "grad_norm": 0.4817613959312439,
+      "learning_rate": 0.00034893617021276597,
+      "loss": 4.0788,
+      "step": 82
+    },
+    {
+      "epoch": 0.010624,
+      "grad_norm": 0.44639843702316284,
+      "learning_rate": 0.00035319148936170213,
+      "loss": 4.0657,
+      "step": 83
+    },
+    {
+      "epoch": 0.010752,
+      "grad_norm": 0.4467894732952118,
+      "learning_rate": 0.0003574468085106383,
+      "loss": 3.9895,
+      "step": 84
+    },
+    {
+      "epoch": 0.01088,
+      "grad_norm": 0.5414591431617737,
+      "learning_rate": 0.0003617021276595745,
+      "loss": 4.1003,
+      "step": 85
+    },
+    {
+      "epoch": 0.011008,
+      "grad_norm": 0.48186618089675903,
+      "learning_rate": 0.00036595744680851063,
+      "loss": 3.9528,
+      "step": 86
+    },
+    {
+      "epoch": 0.011136,
+      "grad_norm": 0.48061543703079224,
+      "learning_rate": 0.0003702127659574468,
+      "loss": 4.0812,
+      "step": 87
+    },
+    {
+      "epoch": 0.011264,
+      "grad_norm": 0.49682289361953735,
+      "learning_rate": 0.00037446808510638297,
+      "loss": 4.011,
+      "step": 88
+    },
+    {
+      "epoch": 0.011392,
+      "grad_norm": 0.45431989431381226,
+      "learning_rate": 0.00037872340425531913,
+      "loss": 4.2069,
+      "step": 89
+    },
+    {
+      "epoch": 0.01152,
+      "grad_norm": 0.501205563545227,
+      "learning_rate": 0.0003829787234042553,
+      "loss": 4.1266,
+      "step": 90
+    },
+    {
+      "epoch": 0.011648,
+      "grad_norm": 0.48493492603302,
+      "learning_rate": 0.0003872340425531915,
+      "loss": 4.1509,
+      "step": 91
+    },
+    {
+      "epoch": 0.011776,
+      "grad_norm": 0.49257519841194153,
+      "learning_rate": 0.0003914893617021277,
+      "loss": 3.9219,
+      "step": 92
+    },
+    {
+      "epoch": 0.011904,
+      "grad_norm": 0.48081332445144653,
+      "learning_rate": 0.00039574468085106385,
+      "loss": 3.9078,
+      "step": 93
+    },
+    {
+      "epoch": 0.012032,
+      "grad_norm": 0.41292718052864075,
+      "learning_rate": 0.0004,
+      "loss": 3.9341,
+      "step": 94
+    },
+    {
+      "epoch": 0.01216,
+      "grad_norm": 0.4186377227306366,
+      "learning_rate": 0.00040425531914893613,
+      "loss": 3.9907,
+      "step": 95
+    },
+    {
+      "epoch": 0.012288,
+      "grad_norm": 0.4486132860183716,
+      "learning_rate": 0.00040851063829787235,
+      "loss": 3.9789,
+      "step": 96
+    },
+    {
+      "epoch": 0.012416,
+      "grad_norm": 0.4189164638519287,
+      "learning_rate": 0.0004127659574468085,
+      "loss": 3.8904,
+      "step": 97
+    },
+    {
+      "epoch": 0.012544,
+      "grad_norm": 0.4362741708755493,
+      "learning_rate": 0.0004170212765957447,
+      "loss": 4.1077,
+      "step": 98
+    },
+    {
+      "epoch": 0.012672,
+      "grad_norm": 0.39442557096481323,
+      "learning_rate": 0.00042127659574468085,
+      "loss": 3.9914,
+      "step": 99
+    },
+    {
+      "epoch": 0.0128,
+      "grad_norm": 0.3945353031158447,
+      "learning_rate": 0.000425531914893617,
+      "loss": 3.9016,
+      "step": 100
+    },
+    {
+      "epoch": 0.012928,
+      "grad_norm": 0.3992224633693695,
+      "learning_rate": 0.0004297872340425532,
+      "loss": 4.0718,
+      "step": 101
+    },
+    {
+      "epoch": 0.013056,
+      "grad_norm": 0.404905766248703,
+      "learning_rate": 0.0004340425531914894,
+      "loss": 3.8471,
+      "step": 102
+    },
+    {
+      "epoch": 0.013184,
+      "grad_norm": 0.458406537771225,
+      "learning_rate": 0.00043829787234042557,
+      "loss": 3.9559,
+      "step": 103
+    },
+    {
+      "epoch": 0.013312,
+      "grad_norm": 0.3796349763870239,
+      "learning_rate": 0.0004425531914893617,
+      "loss": 3.796,
+      "step": 104
+    },
+    {
+      "epoch": 0.01344,
+      "grad_norm": 0.47080284357070923,
+      "learning_rate": 0.00044680851063829785,
+      "loss": 3.9118,
+      "step": 105
+    },
+    {
+      "epoch": 0.013568,
+      "grad_norm": 0.41898369789123535,
+      "learning_rate": 0.000451063829787234,
+      "loss": 4.0004,
+      "step": 106
+    },
+    {
+      "epoch": 0.013696,
+      "grad_norm": 0.4580632448196411,
+      "learning_rate": 0.00045531914893617024,
+      "loss": 3.8332,
+      "step": 107
+    },
+    {
+      "epoch": 0.013824,
+      "grad_norm": 0.4071379005908966,
+      "learning_rate": 0.0004595744680851064,
+      "loss": 3.9318,
+      "step": 108
+    },
+    {
+      "epoch": 0.013952,
+      "grad_norm": 0.4497881829738617,
+      "learning_rate": 0.00046382978723404257,
+      "loss": 3.9175,
+      "step": 109
+    },
+    {
+      "epoch": 0.01408,
+      "grad_norm": 0.49049776792526245,
+      "learning_rate": 0.00046808510638297874,
+      "loss": 4.0421,
+      "step": 110
+    },
+    {
+      "epoch": 0.014208,
+      "grad_norm": 0.4861873984336853,
+      "learning_rate": 0.0004723404255319149,
+      "loss": 3.8317,
+      "step": 111
+    },
+    {
+      "epoch": 0.014336,
+      "grad_norm": 0.3996484875679016,
+      "learning_rate": 0.0004765957446808511,
+      "loss": 3.8542,
+      "step": 112
+    },
+    {
+      "epoch": 0.014464,
+      "grad_norm": 0.4877206087112427,
+      "learning_rate": 0.00048085106382978723,
+      "loss": 4.0346,
+      "step": 113
+    },
+    {
+      "epoch": 0.014592,
+      "grad_norm": 0.44443801045417786,
+      "learning_rate": 0.0004851063829787234,
+      "loss": 4.1039,
+      "step": 114
+    },
+    {
+      "epoch": 0.01472,
+      "grad_norm": 0.4804691672325134,
+      "learning_rate": 0.0004893617021276596,
+      "loss": 4.0837,
+      "step": 115
+    },
+    {
+      "epoch": 0.014848,
+      "grad_norm": 0.5219253897666931,
+      "learning_rate": 0.0004936170212765957,
+      "loss": 3.9347,
+      "step": 116
+    },
+    {
+      "epoch": 0.014976,
+      "grad_norm": 0.5126889944076538,
+      "learning_rate": 0.000497872340425532,
+      "loss": 4.038,
+      "step": 117
+    },
+    {
+      "epoch": 0.015104,
+      "grad_norm": 0.45488491654396057,
+      "learning_rate": 0.0005021276595744681,
+      "loss": 4.0688,
+      "step": 118
+    },
+    {
+      "epoch": 0.015232,
+      "grad_norm": 0.4320695400238037,
+      "learning_rate": 0.0005063829787234042,
+      "loss": 3.9479,
+      "step": 119
+    },
+    {
+      "epoch": 0.01536,
+      "grad_norm": 0.5098544955253601,
+      "learning_rate": 0.0005106382978723404,
+      "loss": 3.8067,
+      "step": 120
+    },
+    {
+      "epoch": 0.015488,
+      "grad_norm": 0.5286118388175964,
+      "learning_rate": 0.0005148936170212766,
+      "loss": 3.9413,
+      "step": 121
+    },
+    {
+      "epoch": 0.015616,
+      "grad_norm": 0.45877233147621155,
+      "learning_rate": 0.0005191489361702127,
+      "loss": 3.8891,
+      "step": 122
+    },
+    {
+      "epoch": 0.015744,
+      "grad_norm": 0.4693465232849121,
+      "learning_rate": 0.000523404255319149,
+      "loss": 3.952,
+      "step": 123
+    },
+    {
+      "epoch": 0.015872,
+      "grad_norm": 0.48408186435699463,
+      "learning_rate": 0.0005276595744680851,
+      "loss": 4.0822,
+      "step": 124
+    },
+    {
+      "epoch": 0.016,
+      "grad_norm": 0.45663052797317505,
+      "learning_rate": 0.0005319148936170213,
+      "loss": 3.848,
+      "step": 125
+    },
+    {
+      "epoch": 0.016128,
+      "grad_norm": 0.604242205619812,
+      "learning_rate": 0.0005361702127659575,
+      "loss": 3.9407,
+      "step": 126
+    },
+    {
+      "epoch": 0.016256,
+      "grad_norm": 0.46198374032974243,
+      "learning_rate": 0.0005404255319148936,
+      "loss": 3.9954,
+      "step": 127
+    },
+    {
+      "epoch": 0.016384,
+      "grad_norm": 0.5603082776069641,
+      "learning_rate": 0.0005446808510638298,
+      "loss": 4.1803,
+      "step": 128
+    },
+    {
+      "epoch": 0.016512,
+      "grad_norm": 0.6773263812065125,
+      "learning_rate": 0.000548936170212766,
+      "loss": 3.8956,
+      "step": 129
+    },
+    {
+      "epoch": 0.01664,
+      "grad_norm": 0.7231651544570923,
+      "learning_rate": 0.0005531914893617022,
+      "loss": 3.9902,
+      "step": 130
+    },
+    {
+      "epoch": 0.016768,
+      "grad_norm": 0.538757860660553,
+      "learning_rate": 0.0005574468085106383,
+      "loss": 3.9527,
+      "step": 131
+    },
+    {
+      "epoch": 0.016896,
+      "grad_norm": 0.603611171245575,
+      "learning_rate": 0.0005617021276595745,
+      "loss": 4.0051,
+      "step": 132
+    },
+    {
+      "epoch": 0.017024,
+      "grad_norm": 0.6359286904335022,
+      "learning_rate": 0.0005659574468085107,
+      "loss": 4.0369,
+      "step": 133
+    },
+    {
+      "epoch": 0.017152,
+      "grad_norm": 0.5824873447418213,
+      "learning_rate": 0.0005702127659574468,
+      "loss": 3.9938,
+      "step": 134
+    },
+    {
+      "epoch": 0.01728,
+      "grad_norm": 0.7265945076942444,
+      "learning_rate": 0.0005744680851063831,
+      "loss": 3.8844,
+      "step": 135
+    },
+    {
+      "epoch": 0.017408,
+      "grad_norm": 0.660447895526886,
+      "learning_rate": 0.0005787234042553191,
+      "loss": 3.8938,
+      "step": 136
+    },
+    {
+      "epoch": 0.017536,
+      "grad_norm": 0.819727897644043,
+      "learning_rate": 0.0005829787234042553,
+      "loss": 3.8421,
+      "step": 137
+    },
+    {
+      "epoch": 0.017664,
+      "grad_norm": 0.7334808707237244,
+      "learning_rate": 0.0005872340425531915,
+      "loss": 3.8305,
+      "step": 138
+    },
+    {
+      "epoch": 0.017792,
+      "grad_norm": 0.9528576135635376,
+      "learning_rate": 0.0005914893617021276,
+      "loss": 3.9298,
+      "step": 139
+    },
+    {
+      "epoch": 0.01792,
+      "grad_norm": 0.9138336777687073,
+      "learning_rate": 0.0005957446808510638,
+      "loss": 3.8899,
+      "step": 140
+    },
+    {
+      "epoch": 0.018048,
+      "grad_norm": 0.851485550403595,
+      "learning_rate": 0.0006,
+      "loss": 3.9338,
+      "step": 141
+    },
+    {
+      "epoch": 0.018176,
+      "grad_norm": 0.930862545967102,
+      "learning_rate": 0.0006042553191489362,
+      "loss": 3.9647,
+      "step": 142
+    },
+    {
+      "epoch": 0.018304,
+      "grad_norm": 1.4732348918914795,
+      "learning_rate": 0.0006085106382978724,
+      "loss": 3.9113,
+      "step": 143
+    },
+    {
+      "epoch": 0.018432,
+      "grad_norm": 0.8589973449707031,
+      "learning_rate": 0.0006127659574468085,
+      "loss": 3.877,
+      "step": 144
+    },
+    {
+      "epoch": 0.01856,
+      "grad_norm": 0.9976956248283386,
+      "learning_rate": 0.0006170212765957447,
+      "loss": 3.7145,
+      "step": 145
+    },
+    {
+      "epoch": 0.018688,
+      "grad_norm": 1.3879908323287964,
+      "learning_rate": 0.0006212765957446808,
+      "loss": 3.8989,
+      "step": 146
+    },
+    {
+      "epoch": 0.018816,
+      "grad_norm": 0.8569075465202332,
+      "learning_rate": 0.000625531914893617,
+      "loss": 3.6446,
+      "step": 147
+    },
+    {
+      "epoch": 0.018944,
+      "grad_norm": 1.044952154159546,
+      "learning_rate": 0.0006297872340425533,
+      "loss": 3.8333,
+      "step": 148
+    },
+    {
+      "epoch": 0.019072,
+      "grad_norm": 0.9681318998336792,
+      "learning_rate": 0.0006340425531914894,
+      "loss": 3.7408,
+      "step": 149
+    },
+    {
+      "epoch": 0.0192,
+      "grad_norm": 0.8293789029121399,
+      "learning_rate": 0.0006382978723404256,
+      "loss": 3.7941,
+      "step": 150
+    },
+    {
+      "epoch": 0.019328,
+      "grad_norm": 0.9052213430404663,
+      "learning_rate": 0.0006425531914893617,
+      "loss": 3.7057,
+      "step": 151
+    },
+    {
+      "epoch": 0.019456,
+      "grad_norm": 0.9938436150550842,
+      "learning_rate": 0.0006468085106382979,
+      "loss": 3.7271,
+      "step": 152
+    },
+    {
+      "epoch": 0.019584,
+      "grad_norm": 1.0418956279754639,
+      "learning_rate": 0.0006510638297872342,
+      "loss": 3.7244,
+      "step": 153
+    },
+    {
+      "epoch": 0.019712,
+      "grad_norm": 1.041827917098999,
+      "learning_rate": 0.0006553191489361702,
+      "loss": 3.7973,
+      "step": 154
+    },
+    {
+      "epoch": 0.01984,
+      "grad_norm": 1.2070434093475342,
+      "learning_rate": 0.0006595744680851064,
+      "loss": 3.763,
+      "step": 155
+    },
+    {
+      "epoch": 0.019968,
+      "grad_norm": 1.2098212242126465,
+      "learning_rate": 0.0006638297872340425,
+      "loss": 3.8179,
+      "step": 156
+    },
+    {
+      "epoch": 0.020096,
+      "grad_norm": 1.1066244840621948,
+      "learning_rate": 0.0006680851063829787,
+      "loss": 3.5696,
+      "step": 157
+    },
+    {
+      "epoch": 0.020224,
+      "grad_norm": 0.9764774441719055,
+      "learning_rate": 0.0006723404255319148,
+      "loss": 3.661,
+      "step": 158
+    },
+    {
+      "epoch": 0.020352,
+      "grad_norm": 1.2751606702804565,
+      "learning_rate": 0.000676595744680851,
+      "loss": 3.8932,
+      "step": 159
+    },
+    {
+      "epoch": 0.02048,
+      "grad_norm": 0.882090151309967,
+      "learning_rate": 0.0006808510638297873,
+      "loss": 3.7595,
+      "step": 160
+    },
+    {
+      "epoch": 0.020608,
+      "grad_norm": 0.9387248158454895,
+      "learning_rate": 0.0006851063829787234,
+      "loss": 3.6247,
+      "step": 161
+    },
+    {
+      "epoch": 0.020736,
+      "grad_norm": 0.966367781162262,
+      "learning_rate": 0.0006893617021276596,
+      "loss": 3.7138,
+      "step": 162
+    },
+    {
+      "epoch": 0.020864,
+      "grad_norm": 0.9478686451911926,
+      "learning_rate": 0.0006936170212765957,
+      "loss": 3.7331,
+      "step": 163
+    },
+    {
+      "epoch": 0.020992,
+      "grad_norm": 0.8556765913963318,
+      "learning_rate": 0.0006978723404255319,
+      "loss": 3.5807,
+      "step": 164
+    },
+    {
+      "epoch": 0.02112,
+      "grad_norm": 1.007807970046997,
+      "learning_rate": 0.0007021276595744682,
+      "loss": 3.7369,
+      "step": 165
+    },
+    {
+      "epoch": 0.021248,
+      "grad_norm": 0.9777787923812866,
+      "learning_rate": 0.0007063829787234043,
+      "loss": 3.691,
+      "step": 166
+    },
+    {
+      "epoch": 0.021376,
+      "grad_norm": 1.027316927909851,
+      "learning_rate": 0.0007106382978723405,
+      "loss": 3.7057,
+      "step": 167
+    },
+    {
+      "epoch": 0.021504,
+      "grad_norm": 0.9888322353363037,
+      "learning_rate": 0.0007148936170212766,
+      "loss": 3.5538,
+      "step": 168
+    },
+    {
+      "epoch": 0.021632,
+      "grad_norm": 0.938539445400238,
+      "learning_rate": 0.0007191489361702128,
+      "loss": 3.7273,
+      "step": 169
+    },
+    {
+      "epoch": 0.02176,
+      "grad_norm": 1.0349684953689575,
+      "learning_rate": 0.000723404255319149,
+      "loss": 3.5971,
+      "step": 170
+    },
+    {
+      "epoch": 0.021888,
+      "grad_norm": 0.8391501903533936,
+      "learning_rate": 0.0007276595744680852,
+      "loss": 3.6428,
+      "step": 171
+    },
+    {
+      "epoch": 0.022016,
+      "grad_norm": 0.9031051397323608,
+      "learning_rate": 0.0007319148936170213,
+      "loss": 3.5116,
+      "step": 172
+    },
+    {
+      "epoch": 0.022144,
+      "grad_norm": 0.9160420894622803,
+      "learning_rate": 0.0007361702127659574,
+      "loss": 3.5312,
+      "step": 173
+    },
+    {
+      "epoch": 0.022272,
+      "grad_norm": 0.956156849861145,
+      "learning_rate": 0.0007404255319148936,
+      "loss": 3.6581,
+      "step": 174
+    },
+    {
+      "epoch": 0.0224,
+      "grad_norm": 0.9484707713127136,
+      "learning_rate": 0.0007446808510638298,
+      "loss": 3.6847,
+      "step": 175
+    },
+    {
+      "epoch": 0.022528,
+      "grad_norm": 1.0067147016525269,
+      "learning_rate": 0.0007489361702127659,
+      "loss": 3.4275,
+      "step": 176
+    },
+    {
+      "epoch": 0.022656,
+      "grad_norm": 1.030314326286316,
+      "learning_rate": 0.0007531914893617022,
+      "loss": 3.7725,
+      "step": 177
+    },
+    {
+      "epoch": 0.022784,
+      "grad_norm": 0.979386031627655,
+      "learning_rate": 0.0007574468085106383,
+      "loss": 3.5085,
+      "step": 178
+    },
+    {
+      "epoch": 0.022912,
+      "grad_norm": 0.9709092974662781,
+      "learning_rate": 0.0007617021276595745,
+      "loss": 3.5108,
+      "step": 179
+    },
+    {
+      "epoch": 0.02304,
+      "grad_norm": 0.9393243789672852,
+      "learning_rate": 0.0007659574468085106,
+      "loss": 3.5364,
+      "step": 180
+    },
+    {
+      "epoch": 0.023168,
+      "grad_norm": 0.9973805546760559,
+      "learning_rate": 0.0007702127659574468,
+      "loss": 3.6058,
+      "step": 181
+    },
+    {
+      "epoch": 0.023296,
+      "grad_norm": 0.842243492603302,
+      "learning_rate": 0.000774468085106383,
+      "loss": 3.5425,
+      "step": 182
+    },
+    {
+      "epoch": 0.023424,
+      "grad_norm": 0.9282711148262024,
+      "learning_rate": 0.0007787234042553192,
+      "loss": 3.5091,
+      "step": 183
+    },
+    {
+      "epoch": 0.023552,
+      "grad_norm": 0.924363374710083,
+      "learning_rate": 0.0007829787234042554,
+      "loss": 3.4311,
+      "step": 184
+    },
+    {
+      "epoch": 0.02368,
+      "grad_norm": 0.9427180886268616,
+      "learning_rate": 0.0007872340425531915,
+      "loss": 3.6139,
+      "step": 185
+    },
+    {
+      "epoch": 0.023808,
+      "grad_norm": 0.8794881701469421,
+      "learning_rate": 0.0007914893617021277,
+      "loss": 3.6135,
+      "step": 186
+    },
+    {
+      "epoch": 0.023936,
+      "grad_norm": 0.8330989480018616,
+      "learning_rate": 0.0007957446808510639,
+      "loss": 3.4336,
+      "step": 187
+    },
+    {
+      "epoch": 0.024064,
+      "grad_norm": 0.9163820743560791,
+      "learning_rate": 0.0008,
+      "loss": 3.4939,
+      "step": 188
+    },
+    {
+      "epoch": 0.024192,
+      "grad_norm": 0.8766486048698425,
+      "learning_rate": 0.0008042553191489363,
+      "loss": 3.5041,
+      "step": 189
+    },
+    {
+      "epoch": 0.02432,
+      "grad_norm": 0.8489805459976196,
+      "learning_rate": 0.0008085106382978723,
+      "loss": 3.5156,
+      "step": 190
+    },
+    {
+      "epoch": 0.024448,
+      "grad_norm": 0.9160630106925964,
+      "learning_rate": 0.0008127659574468085,
+      "loss": 3.3641,
+      "step": 191
+    },
+    {
+      "epoch": 0.024576,
+      "grad_norm": 0.9346572160720825,
+      "learning_rate": 0.0008170212765957447,
+      "loss": 3.4519,
+      "step": 192
+    },
+    {
+      "epoch": 0.024704,
+      "grad_norm": 0.9234569072723389,
+      "learning_rate": 0.0008212765957446808,
+      "loss": 3.6078,
+      "step": 193
+    },
+    {
+      "epoch": 0.024832,
+      "grad_norm": 0.969478189945221,
+      "learning_rate": 0.000825531914893617,
+      "loss": 3.4634,
+      "step": 194
+    },
+    {
+      "epoch": 0.02496,
+      "grad_norm": 0.9091804027557373,
+      "learning_rate": 0.0008297872340425531,
+      "loss": 3.6636,
+      "step": 195
+    },
+    {
+      "epoch": 0.025088,
+      "grad_norm": 0.9273473024368286,
+      "learning_rate": 0.0008340425531914894,
+      "loss": 3.5663,
+      "step": 196
+    },
+    {
+      "epoch": 0.025216,
+      "grad_norm": 1.0131908655166626,
+      "learning_rate": 0.0008382978723404256,
+      "loss": 3.4593,
+      "step": 197
+    },
+    {
+      "epoch": 0.025344,
+      "grad_norm": 0.851618230342865,
+      "learning_rate": 0.0008425531914893617,
+      "loss": 3.5082,
+      "step": 198
+    },
+    {
+      "epoch": 0.025472,
+      "grad_norm": 0.9014102220535278,
+      "learning_rate": 0.0008468085106382979,
+      "loss": 3.5728,
+      "step": 199
+    },
+    {
+      "epoch": 0.0256,
+      "grad_norm": 0.9466091394424438,
+      "learning_rate": 0.000851063829787234,
+      "loss": 3.5582,
+      "step": 200
+    },
+    {
+      "epoch": 0.025728,
+      "grad_norm": 0.9483460187911987,
+      "learning_rate": 0.0008553191489361703,
+      "loss": 3.3827,
+      "step": 201
+    },
+    {
+      "epoch": 0.025856,
+      "grad_norm": 0.9222003817558289,
+      "learning_rate": 0.0008595744680851064,
+      "loss": 3.3668,
+      "step": 202
+    },
+    {
+      "epoch": 0.025984,
+      "grad_norm": 0.9708554744720459,
+      "learning_rate": 0.0008638297872340426,
+      "loss": 3.4732,
+      "step": 203
+    },
+    {
+      "epoch": 0.026112,
+      "grad_norm": 0.949671745300293,
+      "learning_rate": 0.0008680851063829788,
+      "loss": 3.479,
+      "step": 204
+    },
+    {
+      "epoch": 0.02624,
+      "grad_norm": 0.8843660950660706,
+      "learning_rate": 0.0008723404255319149,
+      "loss": 3.4997,
+      "step": 205
+    },
+    {
+      "epoch": 0.026368,
+      "grad_norm": 0.841330349445343,
+      "learning_rate": 0.0008765957446808511,
+      "loss": 3.4025,
+      "step": 206
+    },
+    {
+      "epoch": 0.026496,
+      "grad_norm": 0.9349532723426819,
+      "learning_rate": 0.0008808510638297873,
+      "loss": 3.336,
+      "step": 207
+    },
+    {
+      "epoch": 0.026624,
+      "grad_norm": 0.8519659638404846,
+      "learning_rate": 0.0008851063829787234,
+      "loss": 3.3747,
+      "step": 208
+    },
+    {
+      "epoch": 0.026752,
+      "grad_norm": 0.934549868106842,
+      "learning_rate": 0.0008893617021276596,
+      "loss": 3.404,
+      "step": 209
+    },
+    {
+      "epoch": 0.02688,
+      "grad_norm": 0.8963429927825928,
+      "learning_rate": 0.0008936170212765957,
+      "loss": 3.4165,
+      "step": 210
+    },
+    {
+      "epoch": 0.027008,
+      "grad_norm": 0.9446308016777039,
+      "learning_rate": 0.0008978723404255319,
+      "loss": 3.3677,
+      "step": 211
+    },
+    {
+      "epoch": 0.027136,
+      "grad_norm": 0.9304212927818298,
+      "learning_rate": 0.000902127659574468,
+      "loss": 3.391,
+      "step": 212
+    },
+    {
+      "epoch": 0.027264,
+      "grad_norm": 0.8521987795829773,
+      "learning_rate": 0.0009063829787234043,
+      "loss": 3.3743,
+      "step": 213
+    },
+    {
+      "epoch": 0.027392,
+      "grad_norm": 0.9352124333381653,
+      "learning_rate": 0.0009106382978723405,
+      "loss": 3.2886,
+      "step": 214
+    },
+    {
+      "epoch": 0.02752,
+      "grad_norm": 0.8904086947441101,
+      "learning_rate": 0.0009148936170212766,
+      "loss": 3.3385,
+      "step": 215
+    },
+    {
+      "epoch": 0.027648,
+      "grad_norm": 0.9283705353736877,
+      "learning_rate": 0.0009191489361702128,
+      "loss": 3.3386,
+      "step": 216
+    },
+    {
+      "epoch": 0.027776,
+      "grad_norm": 0.9547103643417358,
+      "learning_rate": 0.0009234042553191489,
+      "loss": 3.3793,
+      "step": 217
+    },
+    {
+      "epoch": 0.027904,
+      "grad_norm": 0.8626068830490112,
+      "learning_rate": 0.0009276595744680851,
+      "loss": 3.3696,
+      "step": 218
+    },
+    {
+      "epoch": 0.028032,
+      "grad_norm": 0.8564240336418152,
+      "learning_rate": 0.0009319148936170214,
+      "loss": 3.2413,
+      "step": 219
+    },
+    {
+      "epoch": 0.02816,
+      "grad_norm": 0.8179991841316223,
+      "learning_rate": 0.0009361702127659575,
+      "loss": 3.4977,
+      "step": 220
+    },
+    {
+      "epoch": 0.028288,
+      "grad_norm": 0.8745185136795044,
+      "learning_rate": 0.0009404255319148937,
+      "loss": 3.5078,
+      "step": 221
+    },
+    {
+      "epoch": 0.028416,
+      "grad_norm": 0.852754533290863,
+      "learning_rate": 0.0009446808510638298,
+      "loss": 3.4175,
+      "step": 222
+    },
+    {
+      "epoch": 0.028544,
+      "grad_norm": 0.7308538556098938,
+      "learning_rate": 0.000948936170212766,
+      "loss": 3.2847,
+      "step": 223
+    },
+    {
+      "epoch": 0.028672,
+      "grad_norm": 0.988552451133728,
+      "learning_rate": 0.0009531914893617022,
+      "loss": 3.342,
+      "step": 224
+    },
+    {
+      "epoch": 0.0288,
+      "grad_norm": 0.8555554747581482,
+      "learning_rate": 0.0009574468085106384,
+      "loss": 3.3043,
+      "step": 225
+    },
+    {
+      "epoch": 0.028928,
+      "grad_norm": 0.9265850186347961,
+      "learning_rate": 0.0009617021276595745,
+      "loss": 3.3641,
+      "step": 226
+    },
+    {
+      "epoch": 0.029056,
+      "grad_norm": 0.8759183287620544,
+      "learning_rate": 0.0009659574468085106,
+      "loss": 3.3156,
+      "step": 227
+    },
+    {
+      "epoch": 0.029184,
+      "grad_norm": 0.8514947295188904,
+      "learning_rate": 0.0009702127659574468,
+      "loss": 3.4576,
+      "step": 228
+    },
+    {
+      "epoch": 0.029312,
+      "grad_norm": 0.8790687322616577,
+      "learning_rate": 0.0009744680851063829,
+      "loss": 3.4077,
+      "step": 229
+    },
+    {
+      "epoch": 0.02944,
+      "grad_norm": 0.807448148727417,
+      "learning_rate": 0.0009787234042553192,
+      "loss": 3.1981,
+      "step": 230
+    },
+    {
+      "epoch": 0.029568,
+      "grad_norm": 0.8897572755813599,
+      "learning_rate": 0.0009829787234042554,
+      "loss": 3.3701,
+      "step": 231
+    },
+    {
+      "epoch": 0.029696,
+      "grad_norm": 0.9162553548812866,
+      "learning_rate": 0.0009872340425531915,
+      "loss": 3.3981,
+      "step": 232
+    },
+    {
+      "epoch": 0.029824,
+      "grad_norm": 0.8530787825584412,
+      "learning_rate": 0.0009914893617021276,
+      "loss": 3.4104,
+      "step": 233
+    },
+    {
+      "epoch": 0.029952,
+      "grad_norm": 0.8520109057426453,
+      "learning_rate": 0.000995744680851064,
+      "loss": 3.1925,
+      "step": 234
+    },
+    {
+      "epoch": 0.03008,
+      "grad_norm": 0.8010127544403076,
+      "learning_rate": 0.001,
+      "loss": 3.3716,
+      "step": 235
+    },
+    {
+      "epoch": 0.030208,
+      "grad_norm": 0.8690422177314758,
+      "learning_rate": 0.0009999999570221019,
+      "loss": 3.2535,
+      "step": 236
+    },
+    {
+      "epoch": 0.030336,
+      "grad_norm": 0.8539397120475769,
+      "learning_rate": 0.0009999998280884144,
+      "loss": 3.2887,
+      "step": 237
+    },
+    {
+      "epoch": 0.030464,
+      "grad_norm": 0.8677384257316589,
+      "learning_rate": 0.0009999996131989603,
+      "loss": 3.3122,
+      "step": 238
+    },
+    {
+      "epoch": 0.030592,
+      "grad_norm": 0.9448609352111816,
+      "learning_rate": 0.0009999993123537758,
+      "loss": 3.421,
+      "step": 239
+    },
+    {
+      "epoch": 0.03072,
+      "grad_norm": 0.853166401386261,
+      "learning_rate": 0.0009999989255529132,
+      "loss": 3.3236,
+      "step": 240
+    },
+    {
+      "epoch": 0.030848,
+      "grad_norm": 0.88984215259552,
+      "learning_rate": 0.0009999984527964388,
+      "loss": 3.3779,
+      "step": 241
+    },
+    {
+      "epoch": 0.030976,
+      "grad_norm": 0.8330724239349365,
+      "learning_rate": 0.000999997894084434,
+      "loss": 3.1445,
+      "step": 242
+    },
+    {
+      "epoch": 0.031104,
+      "grad_norm": 0.8758183717727661,
+      "learning_rate": 0.0009999972494169947,
+      "loss": 3.3848,
+      "step": 243
+    },
+    {
+      "epoch": 0.031232,
+      "grad_norm": 0.76685631275177,
+      "learning_rate": 0.0009999965187942315,
+      "loss": 3.4149,
+      "step": 244
+    },
+    {
+      "epoch": 0.03136,
+      "grad_norm": 0.8216967582702637,
+      "learning_rate": 0.0009999957022162706,
+      "loss": 3.3322,
+      "step": 245
+    },
+    {
+      "epoch": 0.031488,
+      "grad_norm": 0.7599855065345764,
+      "learning_rate": 0.0009999947996832518,
+      "loss": 3.2843,
+      "step": 246
+    },
+    {
+      "epoch": 0.031616,
+      "grad_norm": 0.7854336500167847,
+      "learning_rate": 0.0009999938111953305,
+      "loss": 3.3705,
+      "step": 247
+    },
+    {
+      "epoch": 0.031744,
+      "grad_norm": 0.8333333730697632,
+      "learning_rate": 0.0009999927367526768,
+      "loss": 3.2403,
+      "step": 248
+    },
+    {
+      "epoch": 0.031872,
+      "grad_norm": 0.7525842189788818,
+      "learning_rate": 0.0009999915763554753,
+      "loss": 3.2198,
+      "step": 249
+    },
+    {
+      "epoch": 0.032,
+      "grad_norm": 0.9552444815635681,
+      "learning_rate": 0.0009999903300039253,
+      "loss": 3.4971,
+      "step": 250
+    },
+    {
+      "epoch": 0.032128,
+      "grad_norm": 0.8945220112800598,
+      "learning_rate": 0.0009999889976982413,
+      "loss": 3.2391,
+      "step": 251
+    },
+    {
+      "epoch": 0.032256,
+      "grad_norm": 0.8347753286361694,
+      "learning_rate": 0.000999987579438652,
+      "loss": 3.4935,
+      "step": 252
+    },
+    {
+      "epoch": 0.032384,
+      "grad_norm": 0.8176755309104919,
+      "learning_rate": 0.0009999860752254014,
+      "loss": 3.3071,
+      "step": 253
+    },
+    {
+      "epoch": 0.032512,
+      "grad_norm": 0.8427179455757141,
+      "learning_rate": 0.0009999844850587486,
+      "loss": 3.231,
+      "step": 254
+    },
+    {
+      "epoch": 0.03264,
+      "grad_norm": 0.8210577368736267,
+      "learning_rate": 0.000999982808938966,
+      "loss": 3.4602,
+      "step": 255
+    },
+    {
+      "epoch": 0.032768,
+      "grad_norm": 0.9852181077003479,
+      "learning_rate": 0.0009999810468663425,
+      "loss": 3.3215,
+      "step": 256
+    },
+    {
+      "epoch": 0.032896,
+      "grad_norm": 0.8749747276306152,
+      "learning_rate": 0.0009999791988411805,
+      "loss": 3.3495,
+      "step": 257
+    },
+    {
+      "epoch": 0.033024,
+      "grad_norm": 0.8574191331863403,
+      "learning_rate": 0.0009999772648637984,
+      "loss": 3.2003,
+      "step": 258
+    },
+    {
+      "epoch": 0.033152,
+      "grad_norm": 0.7964820861816406,
+      "learning_rate": 0.000999975244934528,
+      "loss": 3.3638,
+      "step": 259
+    },
+    {
+      "epoch": 0.03328,
+      "grad_norm": 0.9202578067779541,
+      "learning_rate": 0.0009999731390537168,
+      "loss": 3.2962,
+      "step": 260
+    },
+    {
+      "epoch": 0.033408,
+      "grad_norm": 0.8890935778617859,
+      "learning_rate": 0.0009999709472217266,
+      "loss": 3.2612,
+      "step": 261
+    },
+    {
+      "epoch": 0.033536,
+      "grad_norm": 0.8193520903587341,
+      "learning_rate": 0.0009999686694389348,
+      "loss": 3.3331,
+      "step": 262
+    },
+    {
+      "epoch": 0.033664,
+      "grad_norm": 0.827944815158844,
+      "learning_rate": 0.0009999663057057323,
+      "loss": 3.2829,
+      "step": 263
+    },
+    {
+      "epoch": 0.033792,
+      "grad_norm": 0.8362346291542053,
+      "learning_rate": 0.0009999638560225258,
+      "loss": 3.3151,
+      "step": 264
+    },
+    {
+      "epoch": 0.03392,
+      "grad_norm": 0.8072309494018555,
+      "learning_rate": 0.0009999613203897364,
+      "loss": 3.2792,
+      "step": 265
+    },
+    {
+      "epoch": 0.034048,
+      "grad_norm": 0.8949719667434692,
+      "learning_rate": 0.0009999586988078,
+      "loss": 3.342,
+      "step": 266
+    },
+    {
+      "epoch": 0.034176,
+      "grad_norm": 0.7946468591690063,
+      "learning_rate": 0.0009999559912771672,
+      "loss": 3.2979,
+      "step": 267
+    },
+    {
+      "epoch": 0.034304,
+      "grad_norm": 0.823536217212677,
+      "learning_rate": 0.0009999531977983036,
+      "loss": 3.2273,
+      "step": 268
+    },
+    {
+      "epoch": 0.034432,
+      "grad_norm": 0.8271183967590332,
+      "learning_rate": 0.0009999503183716894,
+      "loss": 3.1719,
+      "step": 269
+    },
+    {
+      "epoch": 0.03456,
+      "grad_norm": 0.7533804178237915,
+      "learning_rate": 0.0009999473529978193,
+      "loss": 3.1213,
+      "step": 270
+    },
+    {
+      "epoch": 0.034688,
+      "grad_norm": 0.8764476776123047,
+      "learning_rate": 0.0009999443016772036,
+      "loss": 3.2809,
+      "step": 271
+    },
+    {
+      "epoch": 0.034816,
+      "grad_norm": 0.838826060295105,
+      "learning_rate": 0.0009999411644103665,
+      "loss": 3.3288,
+      "step": 272
+    },
+    {
+      "epoch": 0.034944,
+      "grad_norm": 0.8703794479370117,
+      "learning_rate": 0.0009999379411978473,
+      "loss": 3.303,
+      "step": 273
+    },
+    {
+      "epoch": 0.035072,
+      "grad_norm": 0.8487218022346497,
+      "learning_rate": 0.0009999346320402002,
+      "loss": 3.2203,
+      "step": 274
+    },
+    {
+      "epoch": 0.0352,
+      "grad_norm": 0.8788632750511169,
+      "learning_rate": 0.0009999312369379943,
+      "loss": 3.3296,
+      "step": 275
+    },
+    {
+      "epoch": 0.035328,
+      "grad_norm": 0.7742990851402283,
+      "learning_rate": 0.000999927755891813,
+      "loss": 3.3421,
+      "step": 276
+    },
+    {
+      "epoch": 0.035456,
+      "grad_norm": 0.7678495645523071,
+      "learning_rate": 0.0009999241889022549,
+      "loss": 3.2814,
+      "step": 277
+    },
+    {
+      "epoch": 0.035584,
+      "grad_norm": 0.8291451930999756,
+      "learning_rate": 0.000999920535969933,
+      "loss": 3.2306,
+      "step": 278
+    },
+    {
+      "epoch": 0.035712,
+      "grad_norm": 0.8035301566123962,
+      "learning_rate": 0.0009999167970954753,
+      "loss": 3.2435,
+      "step": 279
+    },
+    {
+      "epoch": 0.03584,
+      "grad_norm": 0.8229001760482788,
+      "learning_rate": 0.0009999129722795248,
+      "loss": 3.2304,
+      "step": 280
+    },
+    {
+      "epoch": 0.035968,
+      "grad_norm": 0.9090624451637268,
+      "learning_rate": 0.000999909061522739,
+      "loss": 3.4283,
+      "step": 281
+    },
+    {
+      "epoch": 0.036096,
+      "grad_norm": 0.8406767845153809,
+      "learning_rate": 0.0009999050648257899,
+      "loss": 3.1534,
+      "step": 282
+    },
+    {
+      "epoch": 0.036224,
+      "grad_norm": 0.906377375125885,
+      "learning_rate": 0.0009999009821893647,
+      "loss": 3.3032,
+      "step": 283
+    },
+    {
+      "epoch": 0.036352,
+      "grad_norm": 0.8107097148895264,
+      "learning_rate": 0.0009998968136141653,
+      "loss": 3.245,
+      "step": 284
+    },
+    {
+      "epoch": 0.03648,
+      "grad_norm": 0.7890430092811584,
+      "learning_rate": 0.0009998925591009084,
+      "loss": 3.231,
+      "step": 285
+    },
+    {
+      "epoch": 0.036608,
+      "grad_norm": 0.8039774298667908,
+      "learning_rate": 0.0009998882186503255,
+      "loss": 3.3317,
+      "step": 286
+    },
+    {
+      "epoch": 0.036736,
+      "grad_norm": 0.8253852725028992,
+      "learning_rate": 0.0009998837922631624,
+      "loss": 3.3291,
+      "step": 287
+    },
+    {
+      "epoch": 0.036864,
+      "grad_norm": 0.8264845609664917,
+      "learning_rate": 0.0009998792799401804,
+      "loss": 3.3314,
+      "step": 288
+    },
+    {
+      "epoch": 0.036992,
+      "grad_norm": 0.8145113587379456,
+      "learning_rate": 0.000999874681682155,
+      "loss": 3.2851,
+      "step": 289
+    },
+    {
+      "epoch": 0.03712,
+      "grad_norm": 0.8404815196990967,
+      "learning_rate": 0.0009998699974898767,
+      "loss": 3.306,
+      "step": 290
+    },
+    {
+      "epoch": 0.037248,
+      "grad_norm": 0.8717318773269653,
+      "learning_rate": 0.000999865227364151,
+      "loss": 3.3333,
+      "step": 291
+    },
+    {
+      "epoch": 0.037376,
+      "grad_norm": 0.8295497894287109,
+      "learning_rate": 0.0009998603713057977,
+      "loss": 3.2947,
+      "step": 292
+    },
+    {
+      "epoch": 0.037504,
+      "grad_norm": 0.8610597848892212,
+      "learning_rate": 0.0009998554293156516,
+      "loss": 3.239,
+      "step": 293
+    },
+    {
+      "epoch": 0.037632,
+      "grad_norm": 0.7773419618606567,
+      "learning_rate": 0.0009998504013945626,
+      "loss": 3.2211,
+      "step": 294
+    },
+    {
+      "epoch": 0.03776,
+      "grad_norm": 0.7946914434432983,
+      "learning_rate": 0.0009998452875433948,
+      "loss": 3.2235,
+      "step": 295
+    },
+    {
+      "epoch": 0.037888,
+      "grad_norm": 0.9075652956962585,
+      "learning_rate": 0.0009998400877630272,
+      "loss": 3.1875,
+      "step": 296
+    },
+    {
+      "epoch": 0.038016,
+      "grad_norm": 0.8102372884750366,
+      "learning_rate": 0.000999834802054354,
+      "loss": 3.0966,
+      "step": 297
+    },
+    {
+      "epoch": 0.038144,
+      "grad_norm": 0.8464699387550354,
+      "learning_rate": 0.0009998294304182837,
+      "loss": 3.2814,
+      "step": 298
+    },
+    {
+      "epoch": 0.038272,
+      "grad_norm": 0.8302551507949829,
+      "learning_rate": 0.0009998239728557397,
+      "loss": 3.1046,
+      "step": 299
+    },
+    {
+      "epoch": 0.0384,
+      "grad_norm": 0.7190431952476501,
+      "learning_rate": 0.0009998184293676603,
+      "loss": 3.2283,
+      "step": 300
+    },
+    {
+      "epoch": 0.038528,
+      "grad_norm": 0.8664412498474121,
+      "learning_rate": 0.0009998127999549988,
+      "loss": 3.3911,
+      "step": 301
+    },
+    {
+      "epoch": 0.038656,
+      "grad_norm": 0.8277519941329956,
+      "learning_rate": 0.0009998070846187223,
+      "loss": 3.2139,
+      "step": 302
+    },
+    {
+      "epoch": 0.038784,
+      "grad_norm": 0.8536484837532043,
+      "learning_rate": 0.000999801283359814,
+      "loss": 3.2374,
+      "step": 303
+    },
+    {
+      "epoch": 0.038912,
+      "grad_norm": 0.7816300988197327,
+      "learning_rate": 0.0009997953961792708,
+      "loss": 3.3369,
+      "step": 304
+    },
+    {
+      "epoch": 0.03904,
+      "grad_norm": 0.8025678396224976,
+      "learning_rate": 0.0009997894230781047,
+      "loss": 3.3676,
+      "step": 305
+    },
+    {
+      "epoch": 0.039168,
+      "grad_norm": 0.8144189715385437,
+      "learning_rate": 0.0009997833640573428,
+      "loss": 3.2657,
+      "step": 306
+    },
+    {
+      "epoch": 0.039296,
+      "grad_norm": 0.7556259036064148,
+      "learning_rate": 0.0009997772191180268,
+      "loss": 3.3331,
+      "step": 307
+    },
+    {
+      "epoch": 0.039424,
+      "grad_norm": 0.7968985438346863,
+      "learning_rate": 0.0009997709882612128,
+      "loss": 3.2455,
+      "step": 308
+    },
+    {
+      "epoch": 0.039552,
+      "grad_norm": 0.7812645435333252,
+      "learning_rate": 0.000999764671487972,
+      "loss": 3.2633,
+      "step": 309
+    },
+    {
+      "epoch": 0.03968,
+      "grad_norm": 0.8175634741783142,
+      "learning_rate": 0.0009997582687993904,
+      "loss": 3.2343,
+      "step": 310
+    },
+    {
+      "epoch": 0.039808,
+      "grad_norm": 0.7840164303779602,
+      "learning_rate": 0.000999751780196569,
+      "loss": 3.1622,
+      "step": 311
+    },
+    {
+      "epoch": 0.039936,
+      "grad_norm": 0.7789239883422852,
+      "learning_rate": 0.0009997452056806227,
+      "loss": 3.2136,
+      "step": 312
+    },
+    {
+      "epoch": 0.040064,
+      "grad_norm": 0.8194958567619324,
+      "learning_rate": 0.000999738545252682,
+      "loss": 3.2398,
+      "step": 313
+    },
+    {
+      "epoch": 0.040192,
+      "grad_norm": 0.7968689799308777,
+      "learning_rate": 0.0009997317989138919,
+      "loss": 3.3128,
+      "step": 314
+    },
+    {
+      "epoch": 0.04032,
+      "grad_norm": 0.8123764991760254,
+      "learning_rate": 0.0009997249666654122,
+      "loss": 3.223,
+      "step": 315
+    },
+    {
+      "epoch": 0.040448,
+      "grad_norm": 0.7803280353546143,
+      "learning_rate": 0.0009997180485084175,
+      "loss": 3.2955,
+      "step": 316
+    },
+    {
+      "epoch": 0.040576,
+      "grad_norm": 0.8365983366966248,
+      "learning_rate": 0.000999711044444097,
+      "loss": 3.3168,
+      "step": 317
+    },
+    {
+      "epoch": 0.040704,
+      "grad_norm": 0.7476377487182617,
+      "learning_rate": 0.0009997039544736546,
+      "loss": 3.2182,
+      "step": 318
+    },
+    {
+      "epoch": 0.040832,
+      "grad_norm": 0.7676250338554382,
+      "learning_rate": 0.0009996967785983097,
+      "loss": 3.1224,
+      "step": 319
+    },
+    {
+      "epoch": 0.04096,
+      "grad_norm": 0.7897032499313354,
+      "learning_rate": 0.0009996895168192955,
+      "loss": 3.2704,
+      "step": 320
+    },
+    {
+      "epoch": 0.041088,
+      "grad_norm": 0.7680679559707642,
+      "learning_rate": 0.0009996821691378602,
+      "loss": 3.2088,
+      "step": 321
+    },
+    {
+      "epoch": 0.041216,
+      "grad_norm": 0.7613722085952759,
+      "learning_rate": 0.0009996747355552674,
+      "loss": 3.2116,
+      "step": 322
+    },
+    {
+      "epoch": 0.041344,
+      "grad_norm": 0.797653317451477,
+      "learning_rate": 0.000999667216072795,
+      "loss": 3.1527,
+      "step": 323
+    },
+    {
+      "epoch": 0.041472,
+      "grad_norm": 0.8015227317810059,
+      "learning_rate": 0.0009996596106917352,
+      "loss": 3.2658,
+      "step": 324
+    },
+    {
+      "epoch": 0.0416,
+      "grad_norm": 0.760847806930542,
+      "learning_rate": 0.000999651919413396,
+      "loss": 3.1354,
+      "step": 325
+    },
+    {
+      "epoch": 0.041728,
+      "grad_norm": 0.8084747195243835,
+      "learning_rate": 0.0009996441422390992,
+      "loss": 3.3118,
+      "step": 326
+    },
+    {
+      "epoch": 0.041856,
+      "grad_norm": 0.8435327410697937,
+      "learning_rate": 0.0009996362791701821,
+      "loss": 3.1355,
+      "step": 327
+    },
+    {
+      "epoch": 0.041984,
+      "grad_norm": 0.8104134798049927,
+      "learning_rate": 0.0009996283302079964,
+      "loss": 3.1854,
+      "step": 328
+    },
+    {
+      "epoch": 0.042112,
+      "grad_norm": 0.8140692710876465,
+      "learning_rate": 0.0009996202953539083,
+      "loss": 3.2624,
+      "step": 329
+    },
+    {
+      "epoch": 0.04224,
+      "grad_norm": 0.9315882921218872,
+      "learning_rate": 0.0009996121746092997,
+      "loss": 3.3476,
+      "step": 330
+    },
+    {
+      "epoch": 0.042368,
+      "grad_norm": 0.8101044297218323,
+      "learning_rate": 0.000999603967975566,
+      "loss": 3.203,
+      "step": 331
+    },
+    {
+      "epoch": 0.042496,
+      "grad_norm": 0.96517014503479,
+      "learning_rate": 0.0009995956754541182,
+      "loss": 3.1896,
+      "step": 332
+    },
+    {
+      "epoch": 0.042624,
+      "grad_norm": 0.8159698247909546,
+      "learning_rate": 0.0009995872970463822,
+      "loss": 3.294,
+      "step": 333
+    },
+    {
+      "epoch": 0.042752,
+      "grad_norm": 0.879071831703186,
+      "learning_rate": 0.0009995788327537983,
+      "loss": 3.1761,
+      "step": 334
+    },
+    {
+      "epoch": 0.04288,
+      "grad_norm": 0.8579456806182861,
+      "learning_rate": 0.0009995702825778211,
+      "loss": 3.1901,
+      "step": 335
+    },
+    {
+      "epoch": 0.043008,
+      "grad_norm": 0.8130214810371399,
+      "learning_rate": 0.0009995616465199209,
+      "loss": 3.1209,
+      "step": 336
+    },
+    {
+      "epoch": 0.043136,
+      "grad_norm": 0.7347559928894043,
+      "learning_rate": 0.0009995529245815822,
+      "loss": 3.1517,
+      "step": 337
+    },
+    {
+      "epoch": 0.043264,
+      "grad_norm": 0.9961786270141602,
+      "learning_rate": 0.0009995441167643046,
+      "loss": 3.1623,
+      "step": 338
+    },
+    {
+      "epoch": 0.043392,
+      "grad_norm": 0.7655556201934814,
+      "learning_rate": 0.000999535223069602,
+      "loss": 3.2121,
+      "step": 339
+    },
+    {
+      "epoch": 0.04352,
+      "grad_norm": 0.8258215188980103,
+      "learning_rate": 0.0009995262434990037,
+      "loss": 3.245,
+      "step": 340
+    },
+    {
+      "epoch": 0.043648,
+      "grad_norm": 0.7866475582122803,
+      "learning_rate": 0.000999517178054053,
+      "loss": 3.2608,
+      "step": 341
+    },
+    {
+      "epoch": 0.043776,
+      "grad_norm": 0.797358512878418,
+      "learning_rate": 0.0009995080267363081,
+      "loss": 3.1894,
+      "step": 342
+    },
+    {
+      "epoch": 0.043904,
+      "grad_norm": 0.8028795123100281,
+      "learning_rate": 0.000999498789547343,
+      "loss": 3.1504,
+      "step": 343
+    },
+    {
+      "epoch": 0.044032,
+      "grad_norm": 0.7332743406295776,
+      "learning_rate": 0.0009994894664887453,
+      "loss": 3.2158,
+      "step": 344
+    },
+    {
+      "epoch": 0.04416,
+      "grad_norm": 0.8556720614433289,
+      "learning_rate": 0.0009994800575621175,
+      "loss": 3.2778,
+      "step": 345
+    },
+    {
+      "epoch": 0.044288,
+      "grad_norm": 0.8669919371604919,
+      "learning_rate": 0.0009994705627690776,
+      "loss": 3.1652,
+      "step": 346
+    },
+    {
+      "epoch": 0.044416,
+      "grad_norm": 0.7044753432273865,
+      "learning_rate": 0.0009994609821112575,
+      "loss": 3.1733,
+      "step": 347
+    },
+    {
+      "epoch": 0.044544,
+      "grad_norm": 0.7604653835296631,
+      "learning_rate": 0.0009994513155903043,
+      "loss": 3.2257,
+      "step": 348
+    },
+    {
+      "epoch": 0.044672,
+      "grad_norm": 0.8257603049278259,
+      "learning_rate": 0.0009994415632078796,
+      "loss": 3.1105,
+      "step": 349
+    },
+    {
+      "epoch": 0.0448,
+      "grad_norm": 0.7756049633026123,
+      "learning_rate": 0.0009994317249656607,
+      "loss": 3.0513,
+      "step": 350
+    },
+    {
+      "epoch": 0.044928,
+      "grad_norm": 0.7612481117248535,
+      "learning_rate": 0.000999421800865338,
+      "loss": 3.2163,
+      "step": 351
+    },
+    {
+      "epoch": 0.045056,
+      "grad_norm": 0.9298253655433655,
+      "learning_rate": 0.0009994117909086179,
+      "loss": 3.2725,
+      "step": 352
+    },
+    {
+      "epoch": 0.045184,
+      "grad_norm": 0.717602550983429,
+      "learning_rate": 0.0009994016950972213,
+      "loss": 3.1094,
+      "step": 353
+    },
+    {
+      "epoch": 0.045312,
+      "grad_norm": 0.7483286261558533,
+      "learning_rate": 0.000999391513432884,
+      "loss": 3.1475,
+      "step": 354
+    },
+    {
+      "epoch": 0.04544,
+      "grad_norm": 0.7910239696502686,
+      "learning_rate": 0.0009993812459173556,
+      "loss": 3.1971,
+      "step": 355
+    },
+    {
+      "epoch": 0.045568,
+      "grad_norm": 0.8064398169517517,
+      "learning_rate": 0.000999370892552402,
+      "loss": 3.1926,
+      "step": 356
+    },
+    {
+      "epoch": 0.045696,
+      "grad_norm": 0.769553005695343,
+      "learning_rate": 0.000999360453339803,
+      "loss": 3.1029,
+      "step": 357
+    },
+    {
+      "epoch": 0.045824,
+      "grad_norm": 0.818428099155426,
+      "learning_rate": 0.0009993499282813527,
+      "loss": 3.1881,
+      "step": 358
+    },
+    {
+      "epoch": 0.045952,
+      "grad_norm": 0.8190000653266907,
+      "learning_rate": 0.0009993393173788607,
+      "loss": 3.2707,
+      "step": 359
+    },
+    {
+      "epoch": 0.04608,
+      "grad_norm": 0.7646111249923706,
+      "learning_rate": 0.0009993286206341513,
+      "loss": 3.2351,
+      "step": 360
+    },
+    {
+      "epoch": 0.046208,
+      "grad_norm": 0.7863335013389587,
+      "learning_rate": 0.0009993178380490635,
+      "loss": 3.3552,
+      "step": 361
+    },
+    {
+      "epoch": 0.046336,
+      "grad_norm": 0.7234004735946655,
+      "learning_rate": 0.0009993069696254506,
+      "loss": 3.1511,
+      "step": 362
+    },
+    {
+      "epoch": 0.046464,
+      "grad_norm": 0.8437128663063049,
+      "learning_rate": 0.0009992960153651811,
+      "loss": 3.2757,
+      "step": 363
+    },
+    {
+      "epoch": 0.046592,
+      "grad_norm": 0.7486714124679565,
+      "learning_rate": 0.0009992849752701383,
+      "loss": 3.221,
+      "step": 364
+    },
+    {
+      "epoch": 0.04672,
+      "grad_norm": 0.7163797616958618,
+      "learning_rate": 0.00099927384934222,
+      "loss": 3.1493,
+      "step": 365
+    },
+    {
+      "epoch": 0.046848,
+      "grad_norm": 0.6881948113441467,
+      "learning_rate": 0.0009992626375833389,
+      "loss": 3.2611,
+      "step": 366
+    },
+    {
+      "epoch": 0.046976,
+      "grad_norm": 0.7824955582618713,
+      "learning_rate": 0.0009992513399954225,
+      "loss": 3.1717,
+      "step": 367
+    },
+    {
+      "epoch": 0.047104,
+      "grad_norm": 0.757205069065094,
+      "learning_rate": 0.0009992399565804128,
+      "loss": 3.024,
+      "step": 368
+    },
+    {
+      "epoch": 0.047232,
+      "grad_norm": 0.820587158203125,
+      "learning_rate": 0.000999228487340267,
+      "loss": 3.0869,
+      "step": 369
+    },
+    {
+      "epoch": 0.04736,
+      "grad_norm": 0.9545608758926392,
+      "learning_rate": 0.0009992169322769566,
+      "loss": 3.2943,
+      "step": 370
+    },
+    {
+      "epoch": 0.047488,
+      "grad_norm": 0.7984517812728882,
+      "learning_rate": 0.0009992052913924683,
+      "loss": 3.2228,
+      "step": 371
+    },
+    {
+      "epoch": 0.047616,
+      "grad_norm": 0.870311439037323,
+      "learning_rate": 0.000999193564688803,
+      "loss": 3.1716,
+      "step": 372
+    },
+    {
+      "epoch": 0.047744,
+      "grad_norm": 0.8146920800209045,
+      "learning_rate": 0.0009991817521679768,
+      "loss": 3.2535,
+      "step": 373
+    },
+    {
+      "epoch": 0.047872,
+      "grad_norm": 0.8337340354919434,
+      "learning_rate": 0.0009991698538320203,
+      "loss": 3.258,
+      "step": 374
+    },
+    {
+      "epoch": 0.048,
+      "grad_norm": 0.7986704707145691,
+      "learning_rate": 0.0009991578696829794,
+      "loss": 3.1198,
+      "step": 375
+    },
+    {
+      "epoch": 0.048128,
+      "grad_norm": 0.7395791411399841,
+      "learning_rate": 0.0009991457997229134,
+      "loss": 3.2348,
+      "step": 376
+    },
+    {
+      "epoch": 0.048256,
+      "grad_norm": 0.7468963861465454,
+      "learning_rate": 0.0009991336439538984,
+      "loss": 3.2296,
+      "step": 377
+    },
+    {
+      "epoch": 0.048384,
+      "grad_norm": 0.7218276858329773,
+      "learning_rate": 0.000999121402378023,
+      "loss": 3.109,
+      "step": 378
+    },
+    {
+      "epoch": 0.048512,
+      "grad_norm": 0.7256556153297424,
+      "learning_rate": 0.0009991090749973926,
+      "loss": 3.1101,
+      "step": 379
+    },
+    {
+      "epoch": 0.04864,
+      "grad_norm": 0.8670189380645752,
+      "learning_rate": 0.000999096661814126,
+      "loss": 3.3117,
+      "step": 380
+    },
+    {
+      "epoch": 0.048768,
+      "grad_norm": 0.8963426351547241,
+      "learning_rate": 0.0009990841628303572,
+      "loss": 3.1284,
+      "step": 381
+    },
+    {
+      "epoch": 0.048896,
+      "grad_norm": 0.685258150100708,
+      "learning_rate": 0.0009990715780482347,
+      "loss": 3.1156,
+      "step": 382
+    },
+    {
+      "epoch": 0.049024,
+      "grad_norm": 0.7792803049087524,
+      "learning_rate": 0.0009990589074699225,
+      "loss": 3.1621,
+      "step": 383
+    },
+    {
+      "epoch": 0.049152,
+      "grad_norm": 0.7119778990745544,
+      "learning_rate": 0.0009990461510975984,
+      "loss": 3.2378,
+      "step": 384
+    },
+    {
+      "epoch": 0.04928,
+      "grad_norm": 0.7149829268455505,
+      "learning_rate": 0.0009990333089334555,
+      "loss": 3.1203,
+      "step": 385
+    },
+    {
+      "epoch": 0.049408,
+      "grad_norm": 0.7387886047363281,
+      "learning_rate": 0.0009990203809797013,
+      "loss": 3.1301,
+      "step": 386
+    },
+    {
+      "epoch": 0.049536,
+      "grad_norm": 0.7041022777557373,
+      "learning_rate": 0.0009990073672385587,
+      "loss": 3.1086,
+      "step": 387
+    },
+    {
+      "epoch": 0.049664,
+      "grad_norm": 0.7611181139945984,
+      "learning_rate": 0.0009989942677122647,
+      "loss": 3.1104,
+      "step": 388
+    },
+    {
+      "epoch": 0.049792,
+      "grad_norm": 0.812367856502533,
+      "learning_rate": 0.0009989810824030713,
+      "loss": 3.1251,
+      "step": 389
+    },
+    {
+      "epoch": 0.04992,
+      "grad_norm": 0.7678342461585999,
+      "learning_rate": 0.000998967811313245,
+      "loss": 3.2409,
+      "step": 390
+    },
+    {
+      "epoch": 0.050048,
+      "grad_norm": 0.7483301162719727,
+      "learning_rate": 0.0009989544544450673,
+      "loss": 3.1614,
+      "step": 391
+    },
+    {
+      "epoch": 0.050176,
+      "grad_norm": 0.7870932817459106,
+      "learning_rate": 0.0009989410118008348,
+      "loss": 3.0902,
+      "step": 392
+    },
+    {
+      "epoch": 0.050304,
+      "grad_norm": 0.82729572057724,
+      "learning_rate": 0.000998927483382858,
+      "loss": 3.1859,
+      "step": 393
+    },
+    {
+      "epoch": 0.050432,
+      "grad_norm": 0.7566567659378052,
+      "learning_rate": 0.0009989138691934628,
+      "loss": 3.1406,
+      "step": 394
+    },
+    {
+      "epoch": 0.05056,
+      "grad_norm": 0.7829864621162415,
+      "learning_rate": 0.0009989001692349894,
+      "loss": 3.2631,
+      "step": 395
+    },
+    {
+      "epoch": 0.050688,
+      "grad_norm": 0.7357336282730103,
+      "learning_rate": 0.0009988863835097932,
+      "loss": 3.0317,
+      "step": 396
+    },
+    {
+      "epoch": 0.050816,
+      "grad_norm": 0.7818576693534851,
+      "learning_rate": 0.0009988725120202442,
+      "loss": 3.1777,
+      "step": 397
+    },
+    {
+      "epoch": 0.050944,
+      "grad_norm": 0.8287133574485779,
+      "learning_rate": 0.000998858554768727,
+      "loss": 3.0795,
+      "step": 398
+    },
+    {
+      "epoch": 0.051072,
+      "grad_norm": 0.78425532579422,
+      "learning_rate": 0.0009988445117576409,
+      "loss": 3.2403,
+      "step": 399
+    },
+    {
+      "epoch": 0.0512,
+      "grad_norm": 0.78499436378479,
+      "learning_rate": 0.0009988303829894002,
+      "loss": 3.1031,
+      "step": 400
+    },
+    {
+      "epoch": 0.051328,
+      "grad_norm": 0.8074891567230225,
+      "learning_rate": 0.0009988161684664335,
+      "loss": 3.151,
+      "step": 401
+    },
+    {
+      "epoch": 0.051456,
+      "grad_norm": 0.7951787710189819,
+      "learning_rate": 0.0009988018681911849,
+      "loss": 3.2585,
+      "step": 402
+    },
+    {
+      "epoch": 0.051584,
+      "grad_norm": 0.8176019191741943,
+      "learning_rate": 0.0009987874821661123,
+      "loss": 3.1498,
+      "step": 403
+    },
+    {
+      "epoch": 0.051712,
+      "grad_norm": 0.8019437789916992,
+      "learning_rate": 0.0009987730103936895,
+      "loss": 3.238,
+      "step": 404
+    },
+    {
+      "epoch": 0.05184,
+      "grad_norm": 0.7950243949890137,
+      "learning_rate": 0.0009987584528764035,
+      "loss": 3.1363,
+      "step": 405
+    },
+    {
+      "epoch": 0.051968,
+      "grad_norm": 0.8351478576660156,
+      "learning_rate": 0.0009987438096167576,
+      "loss": 3.1335,
+      "step": 406
+    },
+    {
+      "epoch": 0.052096,
+      "grad_norm": 0.8377797603607178,
+      "learning_rate": 0.0009987290806172688,
+      "loss": 3.1464,
+      "step": 407
+    },
+    {
+      "epoch": 0.052224,
+      "grad_norm": 0.8063503503799438,
+      "learning_rate": 0.0009987142658804693,
+      "loss": 3.2062,
+      "step": 408
+    },
+    {
+      "epoch": 0.052352,
+      "grad_norm": 0.7243967056274414,
+      "learning_rate": 0.0009986993654089059,
+      "loss": 3.2012,
+      "step": 409
+    },
+    {
+      "epoch": 0.05248,
+      "grad_norm": 0.7669918537139893,
+      "learning_rate": 0.00099868437920514,
+      "loss": 3.2996,
+      "step": 410
+    },
+    {
+      "epoch": 0.052608,
+      "grad_norm": 0.7197207808494568,
+      "learning_rate": 0.0009986693072717484,
+      "loss": 3.1462,
+      "step": 411
+    },
+    {
+      "epoch": 0.052736,
+      "grad_norm": 0.7624968886375427,
+      "learning_rate": 0.0009986541496113215,
+      "loss": 3.296,
+      "step": 412
+    },
+    {
+      "epoch": 0.052864,
+      "grad_norm": 0.774023175239563,
+      "learning_rate": 0.0009986389062264657,
+      "loss": 3.2147,
+      "step": 413
+    },
+    {
+      "epoch": 0.052992,
+      "grad_norm": 0.7212259769439697,
+      "learning_rate": 0.0009986235771198009,
+      "loss": 3.1001,
+      "step": 414
+    },
+    {
+      "epoch": 0.05312,
+      "grad_norm": 0.8833066821098328,
+      "learning_rate": 0.000998608162293963,
+      "loss": 3.0889,
+      "step": 415
+    },
+    {
+      "epoch": 0.053248,
+      "grad_norm": 0.7235613465309143,
+      "learning_rate": 0.0009985926617516012,
+      "loss": 3.1434,
+      "step": 416
+    },
+    {
+      "epoch": 0.053376,
+      "grad_norm": 0.783703088760376,
+      "learning_rate": 0.0009985770754953808,
+      "loss": 3.1736,
+      "step": 417
+    },
+    {
+      "epoch": 0.053504,
+      "grad_norm": 0.7828892469406128,
+      "learning_rate": 0.0009985614035279812,
+      "loss": 3.0042,
+      "step": 418
+    },
+    {
+      "epoch": 0.053632,
+      "grad_norm": 0.7796140909194946,
+      "learning_rate": 0.0009985456458520968,
+      "loss": 3.2019,
+      "step": 419
+    },
+    {
+      "epoch": 0.05376,
+      "grad_norm": 0.7701830863952637,
+      "learning_rate": 0.000998529802470436,
+      "loss": 3.1548,
+      "step": 420
+    },
+    {
+      "epoch": 0.053888,
+      "grad_norm": 0.7383383512496948,
+      "learning_rate": 0.0009985138733857224,
+      "loss": 3.0449,
+      "step": 421
+    },
+    {
+      "epoch": 0.054016,
+      "grad_norm": 0.7496457099914551,
+      "learning_rate": 0.000998497858600695,
+      "loss": 3.1257,
+      "step": 422
+    },
+    {
+      "epoch": 0.054144,
+      "grad_norm": 0.7348814010620117,
+      "learning_rate": 0.0009984817581181068,
+      "loss": 3.1759,
+      "step": 423
+    },
+    {
+      "epoch": 0.054272,
+      "grad_norm": 0.7162420749664307,
+      "learning_rate": 0.0009984655719407253,
+      "loss": 3.1606,
+      "step": 424
+    },
+    {
+      "epoch": 0.0544,
+      "grad_norm": 0.7207120656967163,
+      "learning_rate": 0.000998449300071333,
+      "loss": 3.0778,
+      "step": 425
+    },
+    {
+      "epoch": 0.054528,
+      "grad_norm": 0.7698966860771179,
+      "learning_rate": 0.0009984329425127279,
+      "loss": 3.1562,
+      "step": 426
+    },
+    {
+      "epoch": 0.054656,
+      "grad_norm": 0.8114136457443237,
+      "learning_rate": 0.0009984164992677214,
+      "loss": 3.1876,
+      "step": 427
+    },
+    {
+      "epoch": 0.054784,
+      "grad_norm": 0.7571989893913269,
+      "learning_rate": 0.0009983999703391407,
+      "loss": 3.1106,
+      "step": 428
+    },
+    {
+      "epoch": 0.054912,
+      "grad_norm": 0.7484201788902283,
+      "learning_rate": 0.000998383355729827,
+      "loss": 3.2265,
+      "step": 429
+    },
+    {
+      "epoch": 0.05504,
+      "grad_norm": 0.7098832130432129,
+      "learning_rate": 0.0009983666554426367,
+      "loss": 3.222,
+      "step": 430
+    },
+    {
+      "epoch": 0.055168,
+      "grad_norm": 0.7395052909851074,
+      "learning_rate": 0.000998349869480441,
+      "loss": 3.3194,
+      "step": 431
+    },
+    {
+      "epoch": 0.055296,
+      "grad_norm": 0.8089489340782166,
+      "learning_rate": 0.0009983329978461252,
+      "loss": 3.2356,
+      "step": 432
+    },
+    {
+      "epoch": 0.055424,
+      "grad_norm": 0.745945394039154,
+      "learning_rate": 0.0009983160405425899,
+      "loss": 3.061,
+      "step": 433
+    },
+    {
+      "epoch": 0.055552,
+      "grad_norm": 1.0568695068359375,
+      "learning_rate": 0.0009982989975727502,
+      "loss": 3.0885,
+      "step": 434
+    },
+    {
+      "epoch": 0.05568,
+      "grad_norm": 0.8035607933998108,
+      "learning_rate": 0.0009982818689395362,
+      "loss": 3.2153,
+      "step": 435
+    },
+    {
+      "epoch": 0.055808,
+      "grad_norm": 0.7507925033569336,
+      "learning_rate": 0.0009982646546458922,
+      "loss": 3.0335,
+      "step": 436
+    },
+    {
+      "epoch": 0.055936,
+      "grad_norm": 0.7124462127685547,
+      "learning_rate": 0.0009982473546947777,
+      "loss": 3.162,
+      "step": 437
+    },
+    {
+      "epoch": 0.056064,
+      "grad_norm": 0.7384085655212402,
+      "learning_rate": 0.0009982299690891668,
+      "loss": 3.0469,
+      "step": 438
+    },
+    {
+      "epoch": 0.056192,
+      "grad_norm": 0.7980152368545532,
+      "learning_rate": 0.0009982124978320482,
+      "loss": 3.1539,
+      "step": 439
+    },
+    {
+      "epoch": 0.05632,
+      "grad_norm": 0.6914937496185303,
+      "learning_rate": 0.0009981949409264255,
+      "loss": 3.1616,
+      "step": 440
+    },
+    {
+      "epoch": 0.056448,
+      "grad_norm": 0.8160804510116577,
+      "learning_rate": 0.0009981772983753167,
+      "loss": 3.1733,
+      "step": 441
+    },
+    {
+      "epoch": 0.056576,
+      "grad_norm": 0.7223599553108215,
+      "learning_rate": 0.000998159570181755,
+      "loss": 3.1922,
+      "step": 442
+    },
+    {
+      "epoch": 0.056704,
+      "grad_norm": 0.7005056142807007,
+      "learning_rate": 0.0009981417563487884,
+      "loss": 3.0893,
+      "step": 443
+    },
+    {
+      "epoch": 0.056832,
+      "grad_norm": 0.7256864905357361,
+      "learning_rate": 0.0009981238568794787,
+      "loss": 3.2238,
+      "step": 444
+    },
+    {
+      "epoch": 0.05696,
+      "grad_norm": 0.7256974577903748,
+      "learning_rate": 0.000998105871776903,
+      "loss": 3.0824,
+      "step": 445
+    },
+    {
+      "epoch": 0.057088,
+      "grad_norm": 0.7843765020370483,
+      "learning_rate": 0.0009980878010441537,
+      "loss": 3.2199,
+      "step": 446
+    },
+    {
+      "epoch": 0.057216,
+      "grad_norm": 0.6976195573806763,
+      "learning_rate": 0.000998069644684337,
+      "loss": 3.1079,
+      "step": 447
+    },
+    {
+      "epoch": 0.057344,
+      "grad_norm": 0.7711122632026672,
+      "learning_rate": 0.0009980514027005743,
+      "loss": 3.1403,
+      "step": 448
+    },
+    {
+      "epoch": 0.057472,
+      "grad_norm": 0.6713622212409973,
+      "learning_rate": 0.0009980330750960014,
+      "loss": 3.1314,
+      "step": 449
+    },
+    {
+      "epoch": 0.0576,
+      "grad_norm": 0.7053857445716858,
+      "learning_rate": 0.0009980146618737693,
+      "loss": 3.1581,
+      "step": 450
+    },
+    {
+      "epoch": 0.057728,
+      "grad_norm": 0.6859737634658813,
+      "learning_rate": 0.0009979961630370434,
+      "loss": 3.1192,
+      "step": 451
+    },
+    {
+      "epoch": 0.057856,
+      "grad_norm": 0.7403327822685242,
+      "learning_rate": 0.0009979775785890037,
+      "loss": 3.058,
+      "step": 452
+    },
+    {
+      "epoch": 0.057984,
+      "grad_norm": 0.7334280014038086,
+      "learning_rate": 0.0009979589085328453,
+      "loss": 3.1993,
+      "step": 453
+    },
+    {
+      "epoch": 0.058112,
+      "grad_norm": 0.652929425239563,
+      "learning_rate": 0.0009979401528717776,
+      "loss": 3.1178,
+      "step": 454
+    },
+    {
+      "epoch": 0.05824,
+      "grad_norm": 0.6972871422767639,
+      "learning_rate": 0.0009979213116090252,
+      "loss": 3.0119,
+      "step": 455
+    },
+    {
+      "epoch": 0.058368,
+      "grad_norm": 0.7019355893135071,
+      "learning_rate": 0.0009979023847478266,
+      "loss": 3.0544,
+      "step": 456
+    },
+    {
+      "epoch": 0.058496,
+      "grad_norm": 0.7604960203170776,
+      "learning_rate": 0.0009978833722914362,
+      "loss": 3.2123,
+      "step": 457
+    },
+    {
+      "epoch": 0.058624,
+      "grad_norm": 0.7128496766090393,
+      "learning_rate": 0.000997864274243122,
+      "loss": 3.1748,
+      "step": 458
+    },
+    {
+      "epoch": 0.058752,
+      "grad_norm": 0.7328530550003052,
+      "learning_rate": 0.0009978450906061673,
+      "loss": 3.091,
+      "step": 459
+    },
+    {
+      "epoch": 0.05888,
+      "grad_norm": 0.7276799082756042,
+      "learning_rate": 0.00099782582138387,
+      "loss": 3.0994,
+      "step": 460
+    },
+    {
+      "epoch": 0.059008,
+      "grad_norm": 0.856726348400116,
+      "learning_rate": 0.0009978064665795428,
+      "loss": 3.1889,
+      "step": 461
+    },
+    {
+      "epoch": 0.059136,
+      "grad_norm": 0.7617965936660767,
+      "learning_rate": 0.0009977870261965128,
+      "loss": 3.1114,
+      "step": 462
+    },
+    {
+      "epoch": 0.059264,
+      "grad_norm": 0.8421193957328796,
+      "learning_rate": 0.000997767500238122,
+      "loss": 3.1292,
+      "step": 463
+    },
+    {
+      "epoch": 0.059392,
+      "grad_norm": 0.760215699672699,
+      "learning_rate": 0.0009977478887077276,
+      "loss": 3.1342,
+      "step": 464
+    },
+    {
+      "epoch": 0.05952,
+      "grad_norm": 0.8205323219299316,
+      "learning_rate": 0.0009977281916087007,
+      "loss": 3.151,
+      "step": 465
+    },
+    {
+      "epoch": 0.059648,
+      "grad_norm": 0.7818612456321716,
+      "learning_rate": 0.0009977084089444275,
+      "loss": 3.1147,
+      "step": 466
+    },
+    {
+      "epoch": 0.059776,
+      "grad_norm": 0.791232168674469,
+      "learning_rate": 0.000997688540718309,
+      "loss": 3.1505,
+      "step": 467
+    },
+    {
+      "epoch": 0.059904,
+      "grad_norm": 0.779671311378479,
+      "learning_rate": 0.0009976685869337602,
+      "loss": 3.1545,
+      "step": 468
+    },
+    {
+      "epoch": 0.060032,
+      "grad_norm": 0.7566160559654236,
+      "learning_rate": 0.0009976485475942122,
+      "loss": 3.1821,
+      "step": 469
+    },
+    {
+      "epoch": 0.06016,
+      "grad_norm": 0.698099672794342,
+      "learning_rate": 0.0009976284227031097,
+      "loss": 3.1346,
+      "step": 470
+    },
+    {
+      "epoch": 0.060288,
+      "grad_norm": 0.7147216796875,
+      "learning_rate": 0.000997608212263912,
+      "loss": 3.0567,
+      "step": 471
+    },
+    {
+      "epoch": 0.060416,
+      "grad_norm": 0.7513561844825745,
+      "learning_rate": 0.000997587916280094,
+      "loss": 3.187,
+      "step": 472
+    },
+    {
+      "epoch": 0.060544,
+      "grad_norm": 0.6938210725784302,
+      "learning_rate": 0.0009975675347551447,
+      "loss": 3.101,
+      "step": 473
+    },
+    {
+      "epoch": 0.060672,
+      "grad_norm": 0.7488069534301758,
+      "learning_rate": 0.000997547067692568,
+      "loss": 3.1273,
+      "step": 474
+    },
+    {
+      "epoch": 0.0608,
+      "grad_norm": 0.796640157699585,
+      "learning_rate": 0.000997526515095882,
+      "loss": 3.222,
+      "step": 475
+    },
+    {
+      "epoch": 0.060928,
+      "grad_norm": 0.7998777627944946,
+      "learning_rate": 0.0009975058769686204,
+      "loss": 2.9964,
+      "step": 476
+    },
+    {
+      "epoch": 0.061056,
+      "grad_norm": 0.7299967408180237,
+      "learning_rate": 0.000997485153314331,
+      "loss": 3.102,
+      "step": 477
+    },
+    {
+      "epoch": 0.061184,
+      "grad_norm": 0.7408947348594666,
+      "learning_rate": 0.0009974643441365764,
+      "loss": 3.0728,
+      "step": 478
+    },
+    {
+      "epoch": 0.061312,
+      "grad_norm": 0.824144184589386,
+      "learning_rate": 0.000997443449438934,
+      "loss": 3.1638,
+      "step": 479
+    },
+    {
+      "epoch": 0.06144,
+      "grad_norm": 0.8477172255516052,
+      "learning_rate": 0.0009974224692249956,
+      "loss": 3.0425,
+      "step": 480
+    },
+    {
+      "epoch": 0.061568,
+      "grad_norm": 0.7110774517059326,
+      "learning_rate": 0.0009974014034983684,
+      "loss": 3.0711,
+      "step": 481
+    },
+    {
+      "epoch": 0.061696,
+      "grad_norm": 0.7557125091552734,
+      "learning_rate": 0.0009973802522626732,
+      "loss": 3.1675,
+      "step": 482
+    },
+    {
+      "epoch": 0.061824,
+      "grad_norm": 0.7310000658035278,
+      "learning_rate": 0.0009973590155215467,
+      "loss": 3.1546,
+      "step": 483
+    },
+    {
+      "epoch": 0.061952,
+      "grad_norm": 0.6954556703567505,
+      "learning_rate": 0.0009973376932786396,
+      "loss": 3.0417,
+      "step": 484
+    },
+    {
+      "epoch": 0.06208,
+      "grad_norm": 0.787783682346344,
+      "learning_rate": 0.0009973162855376172,
+      "loss": 3.1751,
+      "step": 485
+    },
+    {
+      "epoch": 0.062208,
+      "grad_norm": 0.7236363291740417,
+      "learning_rate": 0.0009972947923021602,
+      "loss": 3.1429,
+      "step": 486
+    },
+    {
+      "epoch": 0.062336,
+      "grad_norm": 0.7593803405761719,
+      "learning_rate": 0.000997273213575963,
+      "loss": 3.1875,
+      "step": 487
+    },
+    {
+      "epoch": 0.062464,
+      "grad_norm": 0.7542287111282349,
+      "learning_rate": 0.0009972515493627356,
+      "loss": 3.1439,
+      "step": 488
+    },
+    {
+      "epoch": 0.062592,
+      "grad_norm": 0.6806060075759888,
+      "learning_rate": 0.0009972297996662023,
+      "loss": 3.0512,
+      "step": 489
+    },
+    {
+      "epoch": 0.06272,
+      "grad_norm": 0.6804995536804199,
+      "learning_rate": 0.0009972079644901019,
+      "loss": 3.1022,
+      "step": 490
+    },
+    {
+      "epoch": 0.062848,
+      "grad_norm": 0.7489669322967529,
+      "learning_rate": 0.0009971860438381885,
+      "loss": 3.1388,
+      "step": 491
+    },
+    {
+      "epoch": 0.062976,
+      "grad_norm": 0.6942378878593445,
+      "learning_rate": 0.0009971640377142302,
+      "loss": 3.0603,
+      "step": 492
+    },
+    {
+      "epoch": 0.063104,
+      "grad_norm": 0.6896810531616211,
+      "learning_rate": 0.0009971419461220102,
+      "loss": 3.1763,
+      "step": 493
+    },
+    {
+      "epoch": 0.063232,
+      "grad_norm": 0.7135551571846008,
+      "learning_rate": 0.0009971197690653263,
+      "loss": 3.1012,
+      "step": 494
+    },
+    {
+      "epoch": 0.06336,
+      "grad_norm": 0.8038024306297302,
+      "learning_rate": 0.000997097506547991,
+      "loss": 3.1167,
+      "step": 495
+    },
+    {
+      "epoch": 0.063488,
+      "grad_norm": 0.7583320140838623,
+      "learning_rate": 0.0009970751585738314,
+      "loss": 3.053,
+      "step": 496
+    },
+    {
+      "epoch": 0.063616,
+      "grad_norm": 0.7410792708396912,
+      "learning_rate": 0.0009970527251466894,
+      "loss": 3.0533,
+      "step": 497
+    },
+    {
+      "epoch": 0.063744,
+      "grad_norm": 0.7397985458374023,
+      "learning_rate": 0.0009970302062704216,
+      "loss": 3.2184,
+      "step": 498
+    },
+    {
+      "epoch": 0.063872,
+      "grad_norm": 0.6921324133872986,
+      "learning_rate": 0.0009970076019488996,
+      "loss": 3.184,
+      "step": 499
+    },
+    {
+      "epoch": 0.064,
+      "grad_norm": 0.811834990978241,
+      "learning_rate": 0.0009969849121860088,
+      "loss": 3.0971,
+      "step": 500
+    },
+    {
+      "epoch": 0.064128,
+      "grad_norm": 0.7718203067779541,
+      "learning_rate": 0.0009969621369856501,
+      "loss": 3.2074,
+      "step": 501
+    },
+    {
+      "epoch": 0.064256,
+      "grad_norm": 0.7491363286972046,
+      "learning_rate": 0.0009969392763517387,
+      "loss": 3.1739,
+      "step": 502
+    },
+    {
+      "epoch": 0.064384,
+      "grad_norm": 0.6919291019439697,
+      "learning_rate": 0.0009969163302882048,
+      "loss": 3.1163,
+      "step": 503
+    },
+    {
+      "epoch": 0.064512,
+      "grad_norm": 0.7202510833740234,
+      "learning_rate": 0.0009968932987989927,
+      "loss": 3.1114,
+      "step": 504
+    },
+    {
+      "epoch": 0.06464,
+      "grad_norm": 0.696713924407959,
+      "learning_rate": 0.0009968701818880623,
+      "loss": 3.0081,
+      "step": 505
+    },
+    {
+      "epoch": 0.064768,
+      "grad_norm": 0.7257859110832214,
+      "learning_rate": 0.0009968469795593874,
+      "loss": 3.2376,
+      "step": 506
+    },
+    {
+      "epoch": 0.064896,
+      "grad_norm": 0.6806179285049438,
+      "learning_rate": 0.0009968236918169566,
+      "loss": 3.1489,
+      "step": 507
+    },
+    {
+      "epoch": 0.065024,
+      "grad_norm": 0.7255749702453613,
+      "learning_rate": 0.0009968003186647735,
+      "loss": 3.0656,
+      "step": 508
+    },
+    {
+      "epoch": 0.065152,
+      "grad_norm": 0.8080621957778931,
+      "learning_rate": 0.0009967768601068566,
+      "loss": 3.1158,
+      "step": 509
+    },
+    {
+      "epoch": 0.06528,
+      "grad_norm": 0.7103700637817383,
+      "learning_rate": 0.000996753316147238,
+      "loss": 2.966,
+      "step": 510
+    },
+    {
+      "epoch": 0.065408,
+      "grad_norm": 0.7424911260604858,
+      "learning_rate": 0.0009967296867899656,
+      "loss": 3.0878,
+      "step": 511
+    },
+    {
+      "epoch": 0.065536,
+      "grad_norm": 0.7657431960105896,
+      "learning_rate": 0.0009967059720391014,
+      "loss": 3.1063,
+      "step": 512
+    },
+    {
+      "epoch": 0.065664,
+      "grad_norm": 0.686854898929596,
+      "learning_rate": 0.0009966821718987223,
+      "loss": 3.081,
+      "step": 513
+    },
+    {
+      "epoch": 0.065792,
+      "grad_norm": 0.7803924679756165,
+      "learning_rate": 0.0009966582863729198,
+      "loss": 3.0711,
+      "step": 514
+    },
+    {
+      "epoch": 0.06592,
+      "grad_norm": 0.8141863346099854,
+      "learning_rate": 0.0009966343154658001,
+      "loss": 3.0621,
+      "step": 515
+    },
+    {
+      "epoch": 0.066048,
+      "grad_norm": 0.7635067105293274,
+      "learning_rate": 0.0009966102591814842,
+      "loss": 3.2288,
+      "step": 516
+    },
+    {
+      "epoch": 0.066176,
+      "grad_norm": 0.7859484553337097,
+      "learning_rate": 0.0009965861175241075,
+      "loss": 3.0525,
+      "step": 517
+    },
+    {
+      "epoch": 0.066304,
+      "grad_norm": 0.7338966131210327,
+      "learning_rate": 0.0009965618904978202,
+      "loss": 3.1054,
+      "step": 518
+    },
+    {
+      "epoch": 0.066432,
+      "grad_norm": 0.6967792510986328,
+      "learning_rate": 0.0009965375781067873,
+      "loss": 2.9824,
+      "step": 519
+    },
+    {
+      "epoch": 0.06656,
+      "grad_norm": 0.762470543384552,
+      "learning_rate": 0.0009965131803551886,
+      "loss": 3.2783,
+      "step": 520
+    },
+    {
+      "epoch": 0.066688,
+      "grad_norm": 0.9084627032279968,
+      "learning_rate": 0.0009964886972472179,
+      "loss": 3.144,
+      "step": 521
+    },
+    {
+      "epoch": 0.066816,
+      "grad_norm": 0.7518755197525024,
+      "learning_rate": 0.0009964641287870842,
+      "loss": 3.2221,
+      "step": 522
+    },
+    {
+      "epoch": 0.066944,
+      "grad_norm": 0.9937359094619751,
+      "learning_rate": 0.0009964394749790116,
+      "loss": 3.1817,
+      "step": 523
+    },
+    {
+      "epoch": 0.067072,
+      "grad_norm": 0.8334982991218567,
+      "learning_rate": 0.0009964147358272378,
+      "loss": 3.1431,
+      "step": 524
+    },
+    {
+      "epoch": 0.0672,
+      "grad_norm": 0.7354768514633179,
+      "learning_rate": 0.000996389911336016,
+      "loss": 3.0596,
+      "step": 525
+    },
+    {
+      "epoch": 0.067328,
+      "grad_norm": 0.7657933831214905,
+      "learning_rate": 0.000996365001509614,
+      "loss": 2.9875,
+      "step": 526
+    },
+    {
+      "epoch": 0.067456,
+      "grad_norm": 0.6428136229515076,
+      "learning_rate": 0.0009963400063523135,
+      "loss": 3.0384,
+      "step": 527
+    },
+    {
+      "epoch": 0.067584,
+      "grad_norm": 0.8008496761322021,
+      "learning_rate": 0.0009963149258684122,
+      "loss": 3.1785,
+      "step": 528
+    },
+    {
+      "epoch": 0.067712,
+      "grad_norm": 0.8338497281074524,
+      "learning_rate": 0.000996289760062221,
+      "loss": 3.1367,
+      "step": 529
+    },
+    {
+      "epoch": 0.06784,
+      "grad_norm": 0.7511110305786133,
+      "learning_rate": 0.000996264508938067,
+      "loss": 3.1204,
+      "step": 530
+    },
+    {
+      "epoch": 0.067968,
+      "grad_norm": 0.9002041220664978,
+      "learning_rate": 0.0009962391725002904,
+      "loss": 3.0578,
+      "step": 531
+    },
+    {
+      "epoch": 0.068096,
+      "grad_norm": 0.6773514747619629,
+      "learning_rate": 0.0009962137507532475,
+      "loss": 3.1398,
+      "step": 532
+    },
+    {
+      "epoch": 0.068224,
+      "grad_norm": 0.7734187841415405,
+      "learning_rate": 0.000996188243701308,
+      "loss": 3.0571,
+      "step": 533
+    },
+    {
+      "epoch": 0.068352,
+      "grad_norm": 0.792180061340332,
+      "learning_rate": 0.000996162651348857,
+      "loss": 3.0121,
+      "step": 534
+    },
+    {
+      "epoch": 0.06848,
+      "grad_norm": 0.762553870677948,
+      "learning_rate": 0.0009961369737002943,
+      "loss": 3.088,
+      "step": 535
+    },
+    {
+      "epoch": 0.068608,
+      "grad_norm": 0.6956556439399719,
+      "learning_rate": 0.0009961112107600343,
+      "loss": 3.1688,
+      "step": 536
+    },
+    {
+      "epoch": 0.068736,
+      "grad_norm": 0.6945220828056335,
+      "learning_rate": 0.0009960853625325054,
+      "loss": 2.9048,
+      "step": 537
+    },
+    {
+      "epoch": 0.068864,
+      "grad_norm": 0.786500871181488,
+      "learning_rate": 0.0009960594290221519,
+      "loss": 3.0332,
+      "step": 538
+    },
+    {
+      "epoch": 0.068992,
+      "grad_norm": 0.7891021966934204,
+      "learning_rate": 0.0009960334102334315,
+      "loss": 3.2088,
+      "step": 539
+    },
+    {
+      "epoch": 0.06912,
+      "grad_norm": 0.7082324028015137,
+      "learning_rate": 0.0009960073061708176,
+      "loss": 3.0282,
+      "step": 540
+    },
+    {
+      "epoch": 0.069248,
+      "grad_norm": 0.714262843132019,
+      "learning_rate": 0.0009959811168387973,
+      "loss": 3.1239,
+      "step": 541
+    },
+    {
+      "epoch": 0.069376,
+      "grad_norm": 0.6973041892051697,
+      "learning_rate": 0.0009959548422418733,
+      "loss": 3.0098,
+      "step": 542
+    },
+    {
+      "epoch": 0.069504,
+      "grad_norm": 0.6749856472015381,
+      "learning_rate": 0.0009959284823845623,
+      "loss": 3.053,
+      "step": 543
+    },
+    {
+      "epoch": 0.069632,
+      "grad_norm": 0.740902304649353,
+      "learning_rate": 0.0009959020372713957,
+      "loss": 3.0721,
+      "step": 544
+    },
+    {
+      "epoch": 0.06976,
+      "grad_norm": 0.7815244197845459,
+      "learning_rate": 0.00099587550690692,
+      "loss": 3.1627,
+      "step": 545
+    },
+    {
+      "epoch": 0.069888,
+      "grad_norm": 0.770220160484314,
+      "learning_rate": 0.000995848891295696,
+      "loss": 3.137,
+      "step": 546
+    },
+    {
+      "epoch": 0.070016,
+      "grad_norm": 0.8079702258110046,
+      "learning_rate": 0.0009958221904422992,
+      "loss": 3.1068,
+      "step": 547
+    },
+    {
+      "epoch": 0.070144,
+      "grad_norm": 0.7014492750167847,
+      "learning_rate": 0.00099579540435132,
+      "loss": 2.965,
+      "step": 548
+    },
+    {
+      "epoch": 0.070272,
+      "grad_norm": 0.7675507664680481,
+      "learning_rate": 0.0009957685330273629,
+      "loss": 3.0835,
+      "step": 549
+    },
+    {
+      "epoch": 0.0704,
+      "grad_norm": 0.7224851846694946,
+      "learning_rate": 0.0009957415764750475,
+      "loss": 3.1062,
+      "step": 550
+    },
+    {
+      "epoch": 0.070528,
+      "grad_norm": 0.7742204666137695,
+      "learning_rate": 0.000995714534699008,
+      "loss": 3.1677,
+      "step": 551
+    },
+    {
+      "epoch": 0.070656,
+      "grad_norm": 0.6700819134712219,
+      "learning_rate": 0.000995687407703893,
+      "loss": 3.0646,
+      "step": 552
+    },
+    {
+      "epoch": 0.070784,
+      "grad_norm": 0.6780673265457153,
+      "learning_rate": 0.0009956601954943663,
+      "loss": 3.2956,
+      "step": 553
+    },
+    {
+      "epoch": 0.070912,
+      "grad_norm": 0.7249671816825867,
+      "learning_rate": 0.0009956328980751058,
+      "loss": 3.1007,
+      "step": 554
+    },
+    {
+      "epoch": 0.07104,
+      "grad_norm": 0.6841483116149902,
+      "learning_rate": 0.000995605515450804,
+      "loss": 3.0822,
+      "step": 555
+    },
+    {
+      "epoch": 0.071168,
+      "grad_norm": 0.7107796669006348,
+      "learning_rate": 0.0009955780476261689,
+      "loss": 3.1321,
+      "step": 556
+    },
+    {
+      "epoch": 0.071296,
+      "grad_norm": 0.7167090773582458,
+      "learning_rate": 0.000995550494605922,
+      "loss": 3.0805,
+      "step": 557
+    },
+    {
+      "epoch": 0.071424,
+      "grad_norm": 0.7434768080711365,
+      "learning_rate": 0.0009955228563948003,
+      "loss": 3.1339,
+      "step": 558
+    },
+    {
+      "epoch": 0.071552,
+      "grad_norm": 0.691872775554657,
+      "learning_rate": 0.0009954951329975547,
+      "loss": 3.1107,
+      "step": 559
+    },
+    {
+      "epoch": 0.07168,
+      "grad_norm": 0.6434438228607178,
+      "learning_rate": 0.0009954673244189519,
+      "loss": 3.0173,
+      "step": 560
+    },
+    {
+      "epoch": 0.071808,
+      "grad_norm": 0.7157766222953796,
+      "learning_rate": 0.0009954394306637717,
+      "loss": 3.0273,
+      "step": 561
+    },
+    {
+      "epoch": 0.071936,
+      "grad_norm": 0.7790548205375671,
+      "learning_rate": 0.00099541145173681,
+      "loss": 3.1296,
+      "step": 562
+    },
+    {
+      "epoch": 0.072064,
+      "grad_norm": 0.7481710910797119,
+      "learning_rate": 0.0009953833876428762,
+      "loss": 3.1571,
+      "step": 563
+    },
+    {
+      "epoch": 0.072192,
+      "grad_norm": 0.7195072174072266,
+      "learning_rate": 0.0009953552383867952,
+      "loss": 3.1079,
+      "step": 564
+    },
+    {
+      "epoch": 0.07232,
+      "grad_norm": 0.7821452021598816,
+      "learning_rate": 0.000995327003973406,
+      "loss": 3.1179,
+      "step": 565
+    },
+    {
+      "epoch": 0.072448,
+      "grad_norm": 0.7327917218208313,
+      "learning_rate": 0.0009952986844075627,
+      "loss": 3.1254,
+      "step": 566
+    },
+    {
+      "epoch": 0.072576,
+      "grad_norm": 0.831008791923523,
+      "learning_rate": 0.0009952702796941334,
+      "loss": 3.1657,
+      "step": 567
+    },
+    {
+      "epoch": 0.072704,
+      "grad_norm": 0.7353545427322388,
+      "learning_rate": 0.0009952417898380014,
+      "loss": 3.0557,
+      "step": 568
+    },
+    {
+      "epoch": 0.072832,
+      "grad_norm": 0.7818930149078369,
+      "learning_rate": 0.0009952132148440644,
+      "loss": 3.2577,
+      "step": 569
+    },
+    {
+      "epoch": 0.07296,
+      "grad_norm": 0.7695721983909607,
+      "learning_rate": 0.0009951845547172347,
+      "loss": 3.0915,
+      "step": 570
+    },
+    {
+      "epoch": 0.073088,
+      "grad_norm": 0.7305608987808228,
+      "learning_rate": 0.0009951558094624395,
+      "loss": 3.0327,
+      "step": 571
+    },
+    {
+      "epoch": 0.073216,
+      "grad_norm": 0.7304173111915588,
+      "learning_rate": 0.0009951269790846203,
+      "loss": 3.1601,
+      "step": 572
+    },
+    {
+      "epoch": 0.073344,
+      "grad_norm": 0.7234534621238708,
+      "learning_rate": 0.0009950980635887333,
+      "loss": 3.1201,
+      "step": 573
+    },
+    {
+      "epoch": 0.073472,
+      "grad_norm": 0.7461772561073303,
+      "learning_rate": 0.0009950690629797494,
+      "loss": 3.0298,
+      "step": 574
+    },
+    {
+      "epoch": 0.0736,
+      "grad_norm": 0.7911774516105652,
+      "learning_rate": 0.0009950399772626543,
+      "loss": 3.1049,
+      "step": 575
+    },
+    {
+      "epoch": 0.073728,
+      "grad_norm": 0.7779978513717651,
+      "learning_rate": 0.0009950108064424481,
+      "loss": 3.0699,
+      "step": 576
+    },
+    {
+      "epoch": 0.073856,
+      "grad_norm": 0.7385353446006775,
+      "learning_rate": 0.0009949815505241458,
+      "loss": 3.1056,
+      "step": 577
+    },
+    {
+      "epoch": 0.073984,
+      "grad_norm": 0.700513482093811,
+      "learning_rate": 0.0009949522095127766,
+      "loss": 2.9028,
+      "step": 578
+    },
+    {
+      "epoch": 0.074112,
+      "grad_norm": 0.6979973912239075,
+      "learning_rate": 0.0009949227834133846,
+      "loss": 3.107,
+      "step": 579
+    },
+    {
+      "epoch": 0.07424,
+      "grad_norm": 0.704292356967926,
+      "learning_rate": 0.0009948932722310283,
+      "loss": 3.0599,
+      "step": 580
+    },
+    {
+      "epoch": 0.074368,
+      "grad_norm": 0.6718951463699341,
+      "learning_rate": 0.0009948636759707814,
+      "loss": 3.1816,
+      "step": 581
+    },
+    {
+      "epoch": 0.074496,
+      "grad_norm": 0.7304395437240601,
+      "learning_rate": 0.0009948339946377316,
+      "loss": 3.0809,
+      "step": 582
+    },
+    {
+      "epoch": 0.074624,
+      "grad_norm": 0.7129250168800354,
+      "learning_rate": 0.0009948042282369816,
+      "loss": 3.1042,
+      "step": 583
+    },
+    {
+      "epoch": 0.074752,
+      "grad_norm": 0.7862797975540161,
+      "learning_rate": 0.0009947743767736484,
+      "loss": 3.0656,
+      "step": 584
+    },
+    {
+      "epoch": 0.07488,
+      "grad_norm": 0.7385842800140381,
+      "learning_rate": 0.0009947444402528638,
+      "loss": 3.0688,
+      "step": 585
+    },
+    {
+      "epoch": 0.075008,
+      "grad_norm": 0.7365714311599731,
+      "learning_rate": 0.0009947144186797745,
+      "loss": 3.0959,
+      "step": 586
+    },
+    {
+      "epoch": 0.075136,
+      "grad_norm": 0.7907959222793579,
+      "learning_rate": 0.0009946843120595414,
+      "loss": 3.0567,
+      "step": 587
+    },
+    {
+      "epoch": 0.075264,
+      "grad_norm": 0.7481981515884399,
+      "learning_rate": 0.00099465412039734,
+      "loss": 3.0258,
+      "step": 588
+    },
+    {
+      "epoch": 0.075392,
+      "grad_norm": 0.7034714818000793,
+      "learning_rate": 0.000994623843698361,
+      "loss": 3.0622,
+      "step": 589
+    },
+    {
+      "epoch": 0.07552,
+      "grad_norm": 0.7911312580108643,
+      "learning_rate": 0.000994593481967809,
+      "loss": 3.1191,
+      "step": 590
+    },
+    {
+      "epoch": 0.075648,
+      "grad_norm": 0.7168077230453491,
+      "learning_rate": 0.0009945630352109038,
+      "loss": 3.0072,
+      "step": 591
+    },
+    {
+      "epoch": 0.075776,
+      "grad_norm": 0.7810152173042297,
+      "learning_rate": 0.0009945325034328792,
+      "loss": 2.9569,
+      "step": 592
+    },
+    {
+      "epoch": 0.075904,
+      "grad_norm": 0.7254882454872131,
+      "learning_rate": 0.0009945018866389843,
+      "loss": 3.1448,
+      "step": 593
+    },
+    {
+      "epoch": 0.076032,
+      "grad_norm": 0.6917237639427185,
+      "learning_rate": 0.0009944711848344822,
+      "loss": 3.0596,
+      "step": 594
+    },
+    {
+      "epoch": 0.07616,
+      "grad_norm": 0.9037342071533203,
+      "learning_rate": 0.000994440398024651,
+      "loss": 3.0208,
+      "step": 595
+    },
+    {
+      "epoch": 0.076288,
+      "grad_norm": 0.7262395620346069,
+      "learning_rate": 0.0009944095262147835,
+      "loss": 3.1581,
+      "step": 596
+    },
+    {
+      "epoch": 0.076416,
+      "grad_norm": 0.868431568145752,
+      "learning_rate": 0.0009943785694101865,
+      "loss": 3.1486,
+      "step": 597
+    },
+    {
+      "epoch": 0.076544,
+      "grad_norm": 0.6920374631881714,
+      "learning_rate": 0.0009943475276161823,
+      "loss": 3.0052,
+      "step": 598
+    },
+    {
+      "epoch": 0.076672,
+      "grad_norm": 0.6977366805076599,
+      "learning_rate": 0.0009943164008381071,
+      "loss": 3.0292,
+      "step": 599
+    },
+    {
+      "epoch": 0.0768,
+      "grad_norm": 0.7593752145767212,
+      "learning_rate": 0.000994285189081312,
+      "loss": 3.1525,
+      "step": 600
+    },
+    {
+      "epoch": 0.076928,
+      "grad_norm": 0.7664857506752014,
+      "learning_rate": 0.0009942538923511626,
+      "loss": 3.0471,
+      "step": 601
+    },
+    {
+      "epoch": 0.077056,
+      "grad_norm": 0.6776736378669739,
+      "learning_rate": 0.000994222510653039,
+      "loss": 3.0566,
+      "step": 602
+    },
+    {
+      "epoch": 0.077184,
+      "grad_norm": 0.7835485339164734,
+      "learning_rate": 0.0009941910439923366,
+      "loss": 3.2421,
+      "step": 603
+    },
+    {
+      "epoch": 0.077312,
+      "grad_norm": 0.7698847055435181,
+      "learning_rate": 0.0009941594923744647,
+      "loss": 3.2288,
+      "step": 604
+    },
+    {
+      "epoch": 0.07744,
+      "grad_norm": 0.7154125571250916,
+      "learning_rate": 0.000994127855804847,
+      "loss": 3.1387,
+      "step": 605
+    },
+    {
+      "epoch": 0.077568,
+      "grad_norm": 0.6782771348953247,
+      "learning_rate": 0.0009940961342889225,
+      "loss": 3.1883,
+      "step": 606
+    },
+    {
+      "epoch": 0.077696,
+      "grad_norm": 0.904245913028717,
+      "learning_rate": 0.0009940643278321447,
+      "loss": 3.084,
+      "step": 607
+    },
+    {
+      "epoch": 0.077824,
+      "grad_norm": 0.7120671272277832,
+      "learning_rate": 0.0009940324364399809,
+      "loss": 2.9734,
+      "step": 608
+    },
+    {
+      "epoch": 0.077952,
+      "grad_norm": 0.7763145565986633,
+      "learning_rate": 0.0009940004601179141,
+      "loss": 3.1292,
+      "step": 609
+    },
+    {
+      "epoch": 0.07808,
+      "grad_norm": 0.807473361492157,
+      "learning_rate": 0.0009939683988714413,
+      "loss": 3.0418,
+      "step": 610
+    },
+    {
+      "epoch": 0.078208,
+      "grad_norm": 0.6777582168579102,
+      "learning_rate": 0.0009939362527060743,
+      "loss": 3.0167,
+      "step": 611
+    },
+    {
+      "epoch": 0.078336,
+      "grad_norm": 1.498221755027771,
+      "learning_rate": 0.0009939040216273391,
+      "loss": 3.1167,
+      "step": 612
+    },
+    {
+      "epoch": 0.078464,
+      "grad_norm": 0.7088596820831299,
+      "learning_rate": 0.0009938717056407768,
+      "loss": 3.0094,
+      "step": 613
+    },
+    {
+      "epoch": 0.078592,
+      "grad_norm": 0.7785171866416931,
+      "learning_rate": 0.0009938393047519427,
+      "loss": 3.051,
+      "step": 614
+    },
+    {
+      "epoch": 0.07872,
+      "grad_norm": 0.7279220223426819,
+      "learning_rate": 0.0009938068189664071,
+      "loss": 3.092,
+      "step": 615
+    },
+    {
+      "epoch": 0.078848,
+      "grad_norm": 0.7990803718566895,
+      "learning_rate": 0.0009937742482897549,
+      "loss": 3.1432,
+      "step": 616
+    },
+    {
+      "epoch": 0.078976,
+      "grad_norm": 0.6258861422538757,
+      "learning_rate": 0.0009937415927275847,
+      "loss": 2.989,
+      "step": 617
+    },
+    {
+      "epoch": 0.079104,
+      "grad_norm": 0.6657810211181641,
+      "learning_rate": 0.000993708852285511,
+      "loss": 3.1236,
+      "step": 618
+    },
+    {
+      "epoch": 0.079232,
+      "grad_norm": 0.7562732696533203,
+      "learning_rate": 0.0009936760269691621,
+      "loss": 3.1827,
+      "step": 619
+    },
+    {
+      "epoch": 0.07936,
+      "grad_norm": 0.6699399948120117,
+      "learning_rate": 0.0009936431167841806,
+      "loss": 3.0384,
+      "step": 620
+    },
+    {
+      "epoch": 0.079488,
+      "grad_norm": 0.7054303884506226,
+      "learning_rate": 0.000993610121736225,
+      "loss": 3.0586,
+      "step": 621
+    },
+    {
+      "epoch": 0.079616,
+      "grad_norm": 0.7201679944992065,
+      "learning_rate": 0.0009935770418309667,
+      "loss": 3.0479,
+      "step": 622
+    },
+    {
+      "epoch": 0.079744,
+      "grad_norm": 0.7237951159477234,
+      "learning_rate": 0.000993543877074093,
+      "loss": 3.0638,
+      "step": 623
+    },
+    {
+      "epoch": 0.079872,
+      "grad_norm": 0.7305510640144348,
+      "learning_rate": 0.0009935106274713053,
+      "loss": 3.1077,
+      "step": 624
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7390991449356079,
+      "learning_rate": 0.0009934772930283193,
+      "loss": 3.0308,
+      "step": 625
+    },
+    {
+      "epoch": 0.080128,
+      "grad_norm": 0.6952308416366577,
+      "learning_rate": 0.0009934438737508658,
+      "loss": 3.0412,
+      "step": 626
+    },
+    {
+      "epoch": 0.080256,
+      "grad_norm": 0.6755799055099487,
+      "learning_rate": 0.00099341036964469,
+      "loss": 3.0061,
+      "step": 627
+    },
+    {
+      "epoch": 0.080384,
+      "grad_norm": 0.7121164202690125,
+      "learning_rate": 0.0009933767807155515,
+      "loss": 2.9626,
+      "step": 628
+    },
+    {
+      "epoch": 0.080512,
+      "grad_norm": 0.6506229639053345,
+      "learning_rate": 0.0009933431069692245,
+      "loss": 3.0989,
+      "step": 629
+    },
+    {
+      "epoch": 0.08064,
+      "grad_norm": 0.6785044074058533,
+      "learning_rate": 0.0009933093484114982,
+      "loss": 3.1104,
+      "step": 630
+    },
+    {
+      "epoch": 0.080768,
+      "grad_norm": 0.6575117111206055,
+      "learning_rate": 0.000993275505048176,
+      "loss": 3.1135,
+      "step": 631
+    },
+    {
+      "epoch": 0.080896,
+      "grad_norm": 0.7005763649940491,
+      "learning_rate": 0.000993241576885076,
+      "loss": 2.9179,
+      "step": 632
+    },
+    {
+      "epoch": 0.081024,
+      "grad_norm": 0.7527945637702942,
+      "learning_rate": 0.0009932075639280306,
+      "loss": 3.0541,
+      "step": 633
+    },
+    {
+      "epoch": 0.081152,
+      "grad_norm": 0.7555251717567444,
+      "learning_rate": 0.0009931734661828874,
+      "loss": 3.1269,
+      "step": 634
+    },
+    {
+      "epoch": 0.08128,
+      "grad_norm": 0.8016023635864258,
+      "learning_rate": 0.000993139283655508,
+      "loss": 3.0795,
+      "step": 635
+    },
+    {
+      "epoch": 0.081408,
+      "grad_norm": 0.7260803580284119,
+      "learning_rate": 0.0009931050163517687,
+      "loss": 3.1911,
+      "step": 636
+    },
+    {
+      "epoch": 0.081536,
+      "grad_norm": 0.7294242978096008,
+      "learning_rate": 0.0009930706642775607,
+      "loss": 2.9863,
+      "step": 637
+    },
+    {
+      "epoch": 0.081664,
+      "grad_norm": 0.7383239269256592,
+      "learning_rate": 0.0009930362274387892,
+      "loss": 3.0592,
+      "step": 638
+    },
+    {
+      "epoch": 0.081792,
+      "grad_norm": 0.6834824681282043,
+      "learning_rate": 0.0009930017058413744,
+      "loss": 2.896,
+      "step": 639
+    },
+    {
+      "epoch": 0.08192,
+      "grad_norm": 0.6990280151367188,
+      "learning_rate": 0.000992967099491251,
+      "loss": 3.1014,
+      "step": 640
+    },
+    {
+      "epoch": 0.082048,
+      "grad_norm": 0.6729111075401306,
+      "learning_rate": 0.0009929324083943682,
+      "loss": 3.006,
+      "step": 641
+    },
+    {
+      "epoch": 0.082176,
+      "grad_norm": 0.6575155258178711,
+      "learning_rate": 0.00099289763255669,
+      "loss": 2.9695,
+      "step": 642
+    },
+    {
+      "epoch": 0.082304,
+      "grad_norm": 0.6534914970397949,
+      "learning_rate": 0.0009928627719841944,
+      "loss": 3.1624,
+      "step": 643
+    },
+    {
+      "epoch": 0.082432,
+      "grad_norm": 0.7516581416130066,
+      "learning_rate": 0.0009928278266828747,
+      "loss": 3.0719,
+      "step": 644
+    },
+    {
+      "epoch": 0.08256,
+      "grad_norm": 0.7106656432151794,
+      "learning_rate": 0.0009927927966587383,
+      "loss": 3.1371,
+      "step": 645
+    },
+    {
+      "epoch": 0.082688,
+      "grad_norm": 0.6625559329986572,
+      "learning_rate": 0.000992757681917807,
+      "loss": 3.0921,
+      "step": 646
+    },
+    {
+      "epoch": 0.082816,
+      "grad_norm": 0.7367047071456909,
+      "learning_rate": 0.0009927224824661178,
+      "loss": 3.0653,
+      "step": 647
+    },
+    {
+      "epoch": 0.082944,
+      "grad_norm": 0.7739534974098206,
+      "learning_rate": 0.0009926871983097216,
+      "loss": 3.1004,
+      "step": 648
+    },
+    {
+      "epoch": 0.083072,
+      "grad_norm": 0.7578924298286438,
+      "learning_rate": 0.0009926518294546844,
+      "loss": 3.1871,
+      "step": 649
+    },
+    {
+      "epoch": 0.0832,
+      "grad_norm": 0.685286819934845,
+      "learning_rate": 0.0009926163759070863,
+      "loss": 3.029,
+      "step": 650
+    },
+    {
+      "epoch": 0.083328,
+      "grad_norm": 0.6690321564674377,
+      "learning_rate": 0.0009925808376730224,
+      "loss": 3.1493,
+      "step": 651
+    },
+    {
+      "epoch": 0.083456,
+      "grad_norm": 0.7086144685745239,
+      "learning_rate": 0.000992545214758602,
+      "loss": 2.9849,
+      "step": 652
+    },
+    {
+      "epoch": 0.083584,
+      "grad_norm": 0.747015118598938,
+      "learning_rate": 0.000992509507169949,
+      "loss": 3.09,
+      "step": 653
+    },
+    {
+      "epoch": 0.083712,
+      "grad_norm": 0.667730450630188,
+      "learning_rate": 0.0009924737149132022,
+      "loss": 3.0727,
+      "step": 654
+    },
+    {
+      "epoch": 0.08384,
+      "grad_norm": 0.6947174072265625,
+      "learning_rate": 0.0009924378379945146,
+      "loss": 3.0651,
+      "step": 655
+    },
+    {
+      "epoch": 0.083968,
+      "grad_norm": 0.7511754631996155,
+      "learning_rate": 0.0009924018764200537,
+      "loss": 3.1098,
+      "step": 656
+    },
+    {
+      "epoch": 0.084096,
+      "grad_norm": 0.7187621593475342,
+      "learning_rate": 0.0009923658301960018,
+      "loss": 3.1445,
+      "step": 657
+    },
+    {
+      "epoch": 0.084224,
+      "grad_norm": 0.6992619037628174,
+      "learning_rate": 0.0009923296993285557,
+      "loss": 3.0277,
+      "step": 658
+    },
+    {
+      "epoch": 0.084352,
+      "grad_norm": 0.7203529477119446,
+      "learning_rate": 0.0009922934838239269,
+      "loss": 3.0331,
+      "step": 659
+    },
+    {
+      "epoch": 0.08448,
+      "grad_norm": 0.698060929775238,
+      "learning_rate": 0.0009922571836883409,
+      "loss": 3.1225,
+      "step": 660
+    },
+    {
+      "epoch": 0.084608,
+      "grad_norm": 0.6899997591972351,
+      "learning_rate": 0.0009922207989280383,
+      "loss": 3.0495,
+      "step": 661
+    },
+    {
+      "epoch": 0.084736,
+      "grad_norm": 0.7230764627456665,
+      "learning_rate": 0.000992184329549274,
+      "loss": 3.0063,
+      "step": 662
+    },
+    {
+      "epoch": 0.084864,
+      "grad_norm": 0.7283497452735901,
+      "learning_rate": 0.0009921477755583177,
+      "loss": 2.9951,
+      "step": 663
+    },
+    {
+      "epoch": 0.084992,
+      "grad_norm": 0.762915313243866,
+      "learning_rate": 0.0009921111369614532,
+      "loss": 3.0217,
+      "step": 664
+    },
+    {
+      "epoch": 0.08512,
+      "grad_norm": 0.7147346138954163,
+      "learning_rate": 0.0009920744137649792,
+      "loss": 3.0288,
+      "step": 665
+    },
+    {
+      "epoch": 0.085248,
+      "grad_norm": 0.7030810117721558,
+      "learning_rate": 0.000992037605975209,
+      "loss": 3.09,
+      "step": 666
+    },
+    {
+      "epoch": 0.085376,
+      "grad_norm": 0.6879552006721497,
+      "learning_rate": 0.00099200071359847,
+      "loss": 3.0663,
+      "step": 667
+    },
+    {
+      "epoch": 0.085504,
+      "grad_norm": 0.6789368987083435,
+      "learning_rate": 0.0009919637366411047,
+      "loss": 3.1581,
+      "step": 668
+    },
+    {
+      "epoch": 0.085632,
+      "grad_norm": 0.6937747001647949,
+      "learning_rate": 0.0009919266751094696,
+      "loss": 3.174,
+      "step": 669
+    },
+    {
+      "epoch": 0.08576,
+      "grad_norm": 0.6882796883583069,
+      "learning_rate": 0.0009918895290099363,
+      "loss": 3.0706,
+      "step": 670
+    },
+    {
+      "epoch": 0.085888,
+      "grad_norm": 0.7130433917045593,
+      "learning_rate": 0.0009918522983488904,
+      "loss": 3.0475,
+      "step": 671
+    },
+    {
+      "epoch": 0.086016,
+      "grad_norm": 0.7150790691375732,
+      "learning_rate": 0.0009918149831327323,
+      "loss": 2.9482,
+      "step": 672
+    },
+    {
+      "epoch": 0.086144,
+      "grad_norm": 0.7804268598556519,
+      "learning_rate": 0.0009917775833678771,
+      "loss": 3.1034,
+      "step": 673
+    },
+    {
+      "epoch": 0.086272,
+      "grad_norm": 0.6963019967079163,
+      "learning_rate": 0.000991740099060754,
+      "loss": 3.2092,
+      "step": 674
+    },
+    {
+      "epoch": 0.0864,
+      "grad_norm": 0.7491478323936462,
+      "learning_rate": 0.0009917025302178073,
+      "loss": 3.0717,
+      "step": 675
+    },
+    {
+      "epoch": 0.086528,
+      "grad_norm": 0.7937915325164795,
+      "learning_rate": 0.0009916648768454955,
+      "loss": 3.0125,
+      "step": 676
+    },
+    {
+      "epoch": 0.086656,
+      "grad_norm": 0.7034295797348022,
+      "learning_rate": 0.000991627138950291,
+      "loss": 2.9976,
+      "step": 677
+    },
+    {
+      "epoch": 0.086784,
+      "grad_norm": 0.6956719160079956,
+      "learning_rate": 0.0009915893165386824,
+      "loss": 3.1221,
+      "step": 678
+    },
+    {
+      "epoch": 0.086912,
+      "grad_norm": 0.7081146836280823,
+      "learning_rate": 0.0009915514096171711,
+      "loss": 3.1433,
+      "step": 679
+    },
+    {
+      "epoch": 0.08704,
+      "grad_norm": 0.6890046000480652,
+      "learning_rate": 0.0009915134181922739,
+      "loss": 3.1336,
+      "step": 680
+    },
+    {
+      "epoch": 0.087168,
+      "grad_norm": 0.7390891313552856,
+      "learning_rate": 0.000991475342270522,
+      "loss": 3.1501,
+      "step": 681
+    },
+    {
+      "epoch": 0.087296,
+      "grad_norm": 0.7421786785125732,
+      "learning_rate": 0.0009914371818584612,
+      "loss": 3.0679,
+      "step": 682
+    },
+    {
+      "epoch": 0.087424,
+      "grad_norm": 0.7227280735969543,
+      "learning_rate": 0.0009913989369626515,
+      "loss": 3.0433,
+      "step": 683
+    },
+    {
+      "epoch": 0.087552,
+      "grad_norm": 0.8287850618362427,
+      "learning_rate": 0.0009913606075896675,
+      "loss": 3.0964,
+      "step": 684
+    },
+    {
+      "epoch": 0.08768,
+      "grad_norm": 0.7628472447395325,
+      "learning_rate": 0.000991322193746099,
+      "loss": 3.1709,
+      "step": 685
+    },
+    {
+      "epoch": 0.087808,
+      "grad_norm": 0.6621315479278564,
+      "learning_rate": 0.0009912836954385496,
+      "loss": 3.0462,
+      "step": 686
+    },
+    {
+      "epoch": 0.087936,
+      "grad_norm": 0.8023290038108826,
+      "learning_rate": 0.0009912451126736373,
+      "loss": 3.0671,
+      "step": 687
+    },
+    {
+      "epoch": 0.088064,
+      "grad_norm": 0.7438244223594666,
+      "learning_rate": 0.0009912064454579953,
+      "loss": 3.0825,
+      "step": 688
+    },
+    {
+      "epoch": 0.088192,
+      "grad_norm": 0.8270880579948425,
+      "learning_rate": 0.0009911676937982706,
+      "loss": 3.0452,
+      "step": 689
+    },
+    {
+      "epoch": 0.08832,
+      "grad_norm": 0.9641112089157104,
+      "learning_rate": 0.0009911288577011254,
+      "loss": 3.0886,
+      "step": 690
+    },
+    {
+      "epoch": 0.088448,
+      "grad_norm": 0.7076590061187744,
+      "learning_rate": 0.0009910899371732358,
+      "loss": 3.0732,
+      "step": 691
+    },
+    {
+      "epoch": 0.088576,
+      "grad_norm": 0.700249433517456,
+      "learning_rate": 0.0009910509322212928,
+      "loss": 3.0012,
+      "step": 692
+    },
+    {
+      "epoch": 0.088704,
+      "grad_norm": 0.6963532567024231,
+      "learning_rate": 0.000991011842852002,
+      "loss": 2.9255,
+      "step": 693
+    },
+    {
+      "epoch": 0.088832,
+      "grad_norm": 0.6746599674224854,
+      "learning_rate": 0.0009909726690720828,
+      "loss": 2.9708,
+      "step": 694
+    },
+    {
+      "epoch": 0.08896,
+      "grad_norm": 0.690865695476532,
+      "learning_rate": 0.00099093341088827,
+      "loss": 2.9454,
+      "step": 695
+    },
+    {
+      "epoch": 0.089088,
+      "grad_norm": 0.7075537443161011,
+      "learning_rate": 0.0009908940683073127,
+      "loss": 3.0666,
+      "step": 696
+    },
+    {
+      "epoch": 0.089216,
+      "grad_norm": 0.7539401054382324,
+      "learning_rate": 0.000990854641335974,
+      "loss": 3.0535,
+      "step": 697
+    },
+    {
+      "epoch": 0.089344,
+      "grad_norm": 0.6586560010910034,
+      "learning_rate": 0.000990815129981032,
+      "loss": 3.0417,
+      "step": 698
+    },
+    {
+      "epoch": 0.089472,
+      "grad_norm": 0.7366626262664795,
+      "learning_rate": 0.000990775534249279,
+      "loss": 3.0018,
+      "step": 699
+    },
+    {
+      "epoch": 0.0896,
+      "grad_norm": 0.7050108909606934,
+      "learning_rate": 0.0009907358541475222,
+      "loss": 3.2299,
+      "step": 700
+    },
+    {
+      "epoch": 0.089728,
+      "grad_norm": 0.7515972256660461,
+      "learning_rate": 0.0009906960896825829,
+      "loss": 2.9768,
+      "step": 701
+    },
+    {
+      "epoch": 0.089856,
+      "grad_norm": 0.645246684551239,
+      "learning_rate": 0.0009906562408612974,
+      "loss": 3.0237,
+      "step": 702
+    },
+    {
+      "epoch": 0.089984,
+      "grad_norm": 0.638218879699707,
+      "learning_rate": 0.0009906163076905158,
+      "loss": 2.8425,
+      "step": 703
+    },
+    {
+      "epoch": 0.090112,
+      "grad_norm": 0.7679662704467773,
+      "learning_rate": 0.000990576290177103,
+      "loss": 3.1436,
+      "step": 704
+    },
+    {
+      "epoch": 0.09024,
+      "grad_norm": 0.7061696648597717,
+      "learning_rate": 0.0009905361883279387,
+      "loss": 3.0997,
+      "step": 705
+    },
+    {
+      "epoch": 0.090368,
+      "grad_norm": 0.7256093621253967,
+      "learning_rate": 0.000990496002149917,
+      "loss": 3.0648,
+      "step": 706
+    },
+    {
+      "epoch": 0.090496,
+      "grad_norm": 0.7010204792022705,
+      "learning_rate": 0.0009904557316499463,
+      "loss": 3.014,
+      "step": 707
+    },
+    {
+      "epoch": 0.090624,
+      "grad_norm": 0.7349609732627869,
+      "learning_rate": 0.0009904153768349492,
+      "loss": 3.0275,
+      "step": 708
+    },
+    {
+      "epoch": 0.090752,
+      "grad_norm": 0.6804260611534119,
+      "learning_rate": 0.0009903749377118637,
+      "loss": 3.058,
+      "step": 709
+    },
+    {
+      "epoch": 0.09088,
+      "grad_norm": 0.7278128862380981,
+      "learning_rate": 0.0009903344142876414,
+      "loss": 3.0964,
+      "step": 710
+    },
+    {
+      "epoch": 0.091008,
+      "grad_norm": 0.705495297908783,
+      "learning_rate": 0.0009902938065692486,
+      "loss": 3.1314,
+      "step": 711
+    },
+    {
+      "epoch": 0.091136,
+      "grad_norm": 0.7630688548088074,
+      "learning_rate": 0.000990253114563667,
+      "loss": 3.0378,
+      "step": 712
+    },
+    {
+      "epoch": 0.091264,
+      "grad_norm": 0.674720048904419,
+      "learning_rate": 0.000990212338277891,
+      "loss": 2.9626,
+      "step": 713
+    },
+    {
+      "epoch": 0.091392,
+      "grad_norm": 0.6776097416877747,
+      "learning_rate": 0.0009901714777189312,
+      "loss": 3.0826,
+      "step": 714
+    },
+    {
+      "epoch": 0.09152,
+      "grad_norm": 0.7493191361427307,
+      "learning_rate": 0.0009901305328938118,
+      "loss": 3.0247,
+      "step": 715
+    },
+    {
+      "epoch": 0.091648,
+      "grad_norm": 0.8133392930030823,
+      "learning_rate": 0.000990089503809572,
+      "loss": 3.0566,
+      "step": 716
+    },
+    {
+      "epoch": 0.091776,
+      "grad_norm": 0.7083328366279602,
+      "learning_rate": 0.0009900483904732645,
+      "loss": 3.1245,
+      "step": 717
+    },
+    {
+      "epoch": 0.091904,
+      "grad_norm": 0.6564400792121887,
+      "learning_rate": 0.0009900071928919578,
+      "loss": 2.9578,
+      "step": 718
+    },
+    {
+      "epoch": 0.092032,
+      "grad_norm": 0.7109162211418152,
+      "learning_rate": 0.0009899659110727337,
+      "loss": 3.076,
+      "step": 719
+    },
+    {
+      "epoch": 0.09216,
+      "grad_norm": 0.6364444494247437,
+      "learning_rate": 0.0009899245450226896,
+      "loss": 2.9729,
+      "step": 720
+    },
+    {
+      "epoch": 0.092288,
+      "grad_norm": 0.6459901928901672,
+      "learning_rate": 0.0009898830947489366,
+      "loss": 3.0473,
+      "step": 721
+    },
+    {
+      "epoch": 0.092416,
+      "grad_norm": 0.7545374035835266,
+      "learning_rate": 0.0009898415602586001,
+      "loss": 3.0341,
+      "step": 722
+    },
+    {
+      "epoch": 0.092544,
+      "grad_norm": 0.6912007927894592,
+      "learning_rate": 0.000989799941558821,
+      "loss": 3.0208,
+      "step": 723
+    },
+    {
+      "epoch": 0.092672,
+      "grad_norm": 0.7314330339431763,
+      "learning_rate": 0.000989758238656753,
+      "loss": 3.0562,
+      "step": 724
+    },
+    {
+      "epoch": 0.0928,
+      "grad_norm": 0.7523088455200195,
+      "learning_rate": 0.0009897164515595667,
+      "loss": 3.0532,
+      "step": 725
+    },
+    {
+      "epoch": 0.092928,
+      "grad_norm": 0.6732398271560669,
+      "learning_rate": 0.0009896745802744449,
+      "loss": 2.8975,
+      "step": 726
+    },
+    {
+      "epoch": 0.093056,
+      "grad_norm": 0.6710033416748047,
+      "learning_rate": 0.000989632624808586,
+      "loss": 3.1168,
+      "step": 727
+    },
+    {
+      "epoch": 0.093184,
+      "grad_norm": 0.7314496636390686,
+      "learning_rate": 0.0009895905851692026,
+      "loss": 2.9838,
+      "step": 728
+    },
+    {
+      "epoch": 0.093312,
+      "grad_norm": 0.6288522481918335,
+      "learning_rate": 0.0009895484613635216,
+      "loss": 3.0154,
+      "step": 729
+    },
+    {
+      "epoch": 0.09344,
+      "grad_norm": 0.6682552099227905,
+      "learning_rate": 0.0009895062533987848,
+      "loss": 2.9608,
+      "step": 730
+    },
+    {
+      "epoch": 0.093568,
+      "grad_norm": 0.6916200518608093,
+      "learning_rate": 0.0009894639612822486,
+      "loss": 3.0555,
+      "step": 731
+    },
+    {
+      "epoch": 0.093696,
+      "grad_norm": 0.6921561360359192,
+      "learning_rate": 0.0009894215850211824,
+      "loss": 2.9746,
+      "step": 732
+    },
+    {
+      "epoch": 0.093824,
+      "grad_norm": 0.6782969832420349,
+      "learning_rate": 0.0009893791246228724,
+      "loss": 2.9646,
+      "step": 733
+    },
+    {
+      "epoch": 0.093952,
+      "grad_norm": 0.68351811170578,
+      "learning_rate": 0.0009893365800946175,
+      "loss": 3.0646,
+      "step": 734
+    },
+    {
+      "epoch": 0.09408,
+      "grad_norm": 0.7342271208763123,
+      "learning_rate": 0.0009892939514437313,
+      "loss": 3.1165,
+      "step": 735
+    },
+    {
+      "epoch": 0.094208,
+      "grad_norm": 0.7102157473564148,
+      "learning_rate": 0.0009892512386775428,
+      "loss": 3.0997,
+      "step": 736
+    },
+    {
+      "epoch": 0.094336,
+      "grad_norm": 0.6602867245674133,
+      "learning_rate": 0.000989208441803394,
+      "loss": 3.0211,
+      "step": 737
+    },
+    {
+      "epoch": 0.094464,
+      "grad_norm": 0.7778097987174988,
+      "learning_rate": 0.0009891655608286433,
+      "loss": 3.1166,
+      "step": 738
+    },
+    {
+      "epoch": 0.094592,
+      "grad_norm": 0.6952223181724548,
+      "learning_rate": 0.0009891225957606613,
+      "loss": 2.9628,
+      "step": 739
+    },
+    {
+      "epoch": 0.09472,
+      "grad_norm": 0.7545922994613647,
+      "learning_rate": 0.000989079546606835,
+      "loss": 2.9878,
+      "step": 740
+    },
+    {
+      "epoch": 0.094848,
+      "grad_norm": 0.6703514456748962,
+      "learning_rate": 0.0009890364133745646,
+      "loss": 2.9932,
+      "step": 741
+    },
+    {
+      "epoch": 0.094976,
+      "grad_norm": 0.688248336315155,
+      "learning_rate": 0.0009889931960712652,
+      "loss": 3.0606,
+      "step": 742
+    },
+    {
+      "epoch": 0.095104,
+      "grad_norm": 0.689236044883728,
+      "learning_rate": 0.0009889498947043668,
+      "loss": 3.1017,
+      "step": 743
+    },
+    {
+      "epoch": 0.095232,
+      "grad_norm": 0.669778048992157,
+      "learning_rate": 0.000988906509281313,
+      "loss": 3.0348,
+      "step": 744
+    },
+    {
+      "epoch": 0.09536,
+      "grad_norm": 0.7055254578590393,
+      "learning_rate": 0.0009888630398095623,
+      "loss": 3.05,
+      "step": 745
+    },
+    {
+      "epoch": 0.095488,
+      "grad_norm": 0.6773854494094849,
+      "learning_rate": 0.0009888194862965878,
+      "loss": 3.1228,
+      "step": 746
+    },
+    {
+      "epoch": 0.095616,
+      "grad_norm": 0.6839637160301208,
+      "learning_rate": 0.0009887758487498765,
+      "loss": 3.1132,
+      "step": 747
+    },
+    {
+      "epoch": 0.095744,
+      "grad_norm": 0.7668929696083069,
+      "learning_rate": 0.0009887321271769304,
+      "loss": 2.9545,
+      "step": 748
+    },
+    {
+      "epoch": 0.095872,
+      "grad_norm": 0.7026340365409851,
+      "learning_rate": 0.000988688321585266,
+      "loss": 3.0571,
+      "step": 749
+    },
+    {
+      "epoch": 0.096,
+      "grad_norm": 0.6941034197807312,
+      "learning_rate": 0.0009886444319824137,
+      "loss": 3.0099,
+      "step": 750
+    },
+    {
+      "epoch": 0.096128,
+      "grad_norm": 0.7253655195236206,
+      "learning_rate": 0.0009886004583759187,
+      "loss": 3.0347,
+      "step": 751
+    },
+    {
+      "epoch": 0.096256,
+      "grad_norm": 0.6975129246711731,
+      "learning_rate": 0.0009885564007733405,
+      "loss": 3.0595,
+      "step": 752
+    },
+    {
+      "epoch": 0.096384,
+      "grad_norm": 0.7365577220916748,
+      "learning_rate": 0.0009885122591822533,
+      "loss": 3.0155,
+      "step": 753
+    },
+    {
+      "epoch": 0.096512,
+      "grad_norm": 0.7185386419296265,
+      "learning_rate": 0.0009884680336102452,
+      "loss": 3.1196,
+      "step": 754
+    },
+    {
+      "epoch": 0.09664,
+      "grad_norm": 0.7287877798080444,
+      "learning_rate": 0.0009884237240649193,
+      "loss": 3.0546,
+      "step": 755
+    },
+    {
+      "epoch": 0.096768,
+      "grad_norm": 0.781643271446228,
+      "learning_rate": 0.0009883793305538932,
+      "loss": 3.0948,
+      "step": 756
+    },
+    {
+      "epoch": 0.096896,
+      "grad_norm": 0.7226229906082153,
+      "learning_rate": 0.0009883348530847983,
+      "loss": 2.9673,
+      "step": 757
+    },
+    {
+      "epoch": 0.097024,
+      "grad_norm": 0.6505287885665894,
+      "learning_rate": 0.0009882902916652809,
+      "loss": 3.0931,
+      "step": 758
+    },
+    {
+      "epoch": 0.097152,
+      "grad_norm": 0.6687285900115967,
+      "learning_rate": 0.0009882456463030016,
+      "loss": 2.8933,
+      "step": 759
+    },
+    {
+      "epoch": 0.09728,
+      "grad_norm": 0.6824723482131958,
+      "learning_rate": 0.0009882009170056353,
+      "loss": 3.0383,
+      "step": 760
+    },
+    {
+      "epoch": 0.097408,
+      "grad_norm": 0.730706512928009,
+      "learning_rate": 0.000988156103780872,
+      "loss": 2.9651,
+      "step": 761
+    },
+    {
+      "epoch": 0.097536,
+      "grad_norm": 0.6420467495918274,
+      "learning_rate": 0.000988111206636415,
+      "loss": 2.9905,
+      "step": 762
+    },
+    {
+      "epoch": 0.097664,
+      "grad_norm": 0.719586193561554,
+      "learning_rate": 0.000988066225579983,
+      "loss": 3.1945,
+      "step": 763
+    },
+    {
+      "epoch": 0.097792,
+      "grad_norm": 0.6795921921730042,
+      "learning_rate": 0.0009880211606193088,
+      "loss": 2.8984,
+      "step": 764
+    },
+    {
+      "epoch": 0.09792,
+      "grad_norm": 0.7220088839530945,
+      "learning_rate": 0.0009879760117621392,
+      "loss": 2.9765,
+      "step": 765
+    },
+    {
+      "epoch": 0.098048,
+      "grad_norm": 0.7150560021400452,
+      "learning_rate": 0.0009879307790162362,
+      "loss": 2.8898,
+      "step": 766
+    },
+    {
+      "epoch": 0.098176,
+      "grad_norm": 0.7640425562858582,
+      "learning_rate": 0.000987885462389376,
+      "loss": 2.955,
+      "step": 767
+    },
+    {
+      "epoch": 0.098304,
+      "grad_norm": 0.6977807283401489,
+      "learning_rate": 0.0009878400618893483,
+      "loss": 3.0531,
+      "step": 768
+    },
+    {
+      "epoch": 0.098432,
+      "grad_norm": 0.6772504448890686,
+      "learning_rate": 0.0009877945775239585,
+      "loss": 3.0334,
+      "step": 769
+    },
+    {
+      "epoch": 0.09856,
+      "grad_norm": 0.6757044196128845,
+      "learning_rate": 0.0009877490093010259,
+      "loss": 3.0737,
+      "step": 770
+    },
+    {
+      "epoch": 0.098688,
+      "grad_norm": 0.6773708462715149,
+      "learning_rate": 0.0009877033572283842,
+      "loss": 3.1026,
+      "step": 771
+    },
+    {
+      "epoch": 0.098816,
+      "grad_norm": 0.6826951503753662,
+      "learning_rate": 0.0009876576213138815,
+      "loss": 3.1571,
+      "step": 772
+    },
+    {
+      "epoch": 0.098944,
+      "grad_norm": 0.6875942349433899,
+      "learning_rate": 0.0009876118015653803,
+      "loss": 2.9538,
+      "step": 773
+    },
+    {
+      "epoch": 0.099072,
+      "grad_norm": 0.668457567691803,
+      "learning_rate": 0.0009875658979907574,
+      "loss": 3.1074,
+      "step": 774
+    },
+    {
+      "epoch": 0.0992,
+      "grad_norm": 0.660249650478363,
+      "learning_rate": 0.0009875199105979042,
+      "loss": 3.0199,
+      "step": 775
+    },
+    {
+      "epoch": 0.099328,
+      "grad_norm": 0.6975038051605225,
+      "learning_rate": 0.0009874738393947267,
+      "loss": 3.0205,
+      "step": 776
+    },
+    {
+      "epoch": 0.099456,
+      "grad_norm": 0.6172084808349609,
+      "learning_rate": 0.0009874276843891449,
+      "loss": 2.9902,
+      "step": 777
+    },
+    {
+      "epoch": 0.099584,
+      "grad_norm": 0.6787610650062561,
+      "learning_rate": 0.0009873814455890935,
+      "loss": 3.046,
+      "step": 778
+    },
+    {
+      "epoch": 0.099712,
+      "grad_norm": 0.6908243298530579,
+      "learning_rate": 0.0009873351230025211,
+      "loss": 3.0434,
+      "step": 779
+    },
+    {
+      "epoch": 0.09984,
+      "grad_norm": 0.7081867456436157,
+      "learning_rate": 0.0009872887166373916,
+      "loss": 3.0345,
+      "step": 780
+    },
+    {
+      "epoch": 0.099968,
+      "grad_norm": 0.6856307983398438,
+      "learning_rate": 0.0009872422265016826,
+      "loss": 3.0559,
+      "step": 781
+    },
+    {
+      "epoch": 0.100096,
+      "grad_norm": 0.72980135679245,
+      "learning_rate": 0.0009871956526033862,
+      "loss": 2.9158,
+      "step": 782
+    },
+    {
+      "epoch": 0.100224,
+      "grad_norm": 0.6752789616584778,
+      "learning_rate": 0.0009871489949505091,
+      "loss": 3.0251,
+      "step": 783
+    },
+    {
+      "epoch": 0.100352,
+      "grad_norm": 0.7256962656974792,
+      "learning_rate": 0.0009871022535510722,
+      "loss": 2.9644,
+      "step": 784
+    },
+    {
+      "epoch": 0.10048,
+      "grad_norm": 0.6545035243034363,
+      "learning_rate": 0.0009870554284131111,
+      "loss": 2.9613,
+      "step": 785
+    },
+    {
+      "epoch": 0.100608,
+      "grad_norm": 0.7104776501655579,
+      "learning_rate": 0.0009870085195446752,
+      "loss": 3.0804,
+      "step": 786
+    },
+    {
+      "epoch": 0.100736,
+      "grad_norm": 0.7600831389427185,
+      "learning_rate": 0.000986961526953829,
+      "loss": 3.0801,
+      "step": 787
+    },
+    {
+      "epoch": 0.100864,
+      "grad_norm": 0.7726160883903503,
+      "learning_rate": 0.000986914450648651,
+      "loss": 2.942,
+      "step": 788
+    },
+    {
+      "epoch": 0.100992,
+      "grad_norm": 0.7255537509918213,
+      "learning_rate": 0.0009868672906372341,
+      "loss": 3.0949,
+      "step": 789
+    },
+    {
+      "epoch": 0.10112,
+      "grad_norm": 0.6741006374359131,
+      "learning_rate": 0.0009868200469276857,
+      "loss": 3.1419,
+      "step": 790
+    },
+    {
+      "epoch": 0.101248,
+      "grad_norm": 0.6867875456809998,
+      "learning_rate": 0.0009867727195281275,
+      "loss": 3.0712,
+      "step": 791
+    },
+    {
+      "epoch": 0.101376,
+      "grad_norm": 0.68589186668396,
+      "learning_rate": 0.0009867253084466956,
+      "loss": 3.1549,
+      "step": 792
+    },
+    {
+      "epoch": 0.101504,
+      "grad_norm": 0.6061535477638245,
+      "learning_rate": 0.0009866778136915408,
+      "loss": 2.9811,
+      "step": 793
+    },
+    {
+      "epoch": 0.101632,
+      "grad_norm": 0.6912597417831421,
+      "learning_rate": 0.0009866302352708276,
+      "loss": 3.043,
+      "step": 794
+    },
+    {
+      "epoch": 0.10176,
+      "grad_norm": 0.6708470582962036,
+      "learning_rate": 0.0009865825731927354,
+      "loss": 3.0798,
+      "step": 795
+    },
+    {
+      "epoch": 0.101888,
+      "grad_norm": 0.694206714630127,
+      "learning_rate": 0.000986534827465458,
+      "loss": 3.0608,
+      "step": 796
+    },
+    {
+      "epoch": 0.102016,
+      "grad_norm": 0.6785531640052795,
+      "learning_rate": 0.0009864869980972034,
+      "loss": 2.9613,
+      "step": 797
+    },
+    {
+      "epoch": 0.102144,
+      "grad_norm": 0.6426734924316406,
+      "learning_rate": 0.000986439085096194,
+      "loss": 2.9017,
+      "step": 798
+    },
+    {
+      "epoch": 0.102272,
+      "grad_norm": 0.7030400037765503,
+      "learning_rate": 0.0009863910884706666,
+      "loss": 3.1597,
+      "step": 799
+    },
+    {
+      "epoch": 0.1024,
+      "grad_norm": 0.689471423625946,
+      "learning_rate": 0.0009863430082288722,
+      "loss": 3.1361,
+      "step": 800
+    },
+    {
+      "epoch": 0.102528,
+      "grad_norm": 0.7326629757881165,
+      "learning_rate": 0.0009862948443790766,
+      "loss": 2.8999,
+      "step": 801
+    },
+    {
+      "epoch": 0.102656,
+      "grad_norm": 0.6239827871322632,
+      "learning_rate": 0.0009862465969295597,
+      "loss": 3.0167,
+      "step": 802
+    },
+    {
+      "epoch": 0.102784,
+      "grad_norm": 0.7145435810089111,
+      "learning_rate": 0.0009861982658886158,
+      "loss": 3.0475,
+      "step": 803
+    },
+    {
+      "epoch": 0.102912,
+      "grad_norm": 0.7064559459686279,
+      "learning_rate": 0.0009861498512645532,
+      "loss": 3.1276,
+      "step": 804
+    },
+    {
+      "epoch": 0.10304,
+      "grad_norm": 0.7117116451263428,
+      "learning_rate": 0.0009861013530656955,
+      "loss": 3.0297,
+      "step": 805
+    },
+    {
+      "epoch": 0.103168,
+      "grad_norm": 0.6756002306938171,
+      "learning_rate": 0.0009860527713003798,
+      "loss": 2.8862,
+      "step": 806
+    },
+    {
+      "epoch": 0.103296,
+      "grad_norm": 0.7354007959365845,
+      "learning_rate": 0.0009860041059769579,
+      "loss": 3.0906,
+      "step": 807
+    },
+    {
+      "epoch": 0.103424,
+      "grad_norm": 0.7439422607421875,
+      "learning_rate": 0.000985955357103796,
+      "loss": 3.1357,
+      "step": 808
+    },
+    {
+      "epoch": 0.103552,
+      "grad_norm": 0.7092074751853943,
+      "learning_rate": 0.0009859065246892744,
+      "loss": 2.9167,
+      "step": 809
+    },
+    {
+      "epoch": 0.10368,
+      "grad_norm": 0.5974611043930054,
+      "learning_rate": 0.000985857608741788,
+      "loss": 2.9146,
+      "step": 810
+    },
+    {
+      "epoch": 0.103808,
+      "grad_norm": 0.6998751163482666,
+      "learning_rate": 0.0009858086092697463,
+      "loss": 3.0178,
+      "step": 811
+    },
+    {
+      "epoch": 0.103936,
+      "grad_norm": 0.6987960934638977,
+      "learning_rate": 0.0009857595262815727,
+      "loss": 3.0328,
+      "step": 812
+    },
+    {
+      "epoch": 0.104064,
+      "grad_norm": 0.6748073697090149,
+      "learning_rate": 0.000985710359785705,
+      "loss": 3.0561,
+      "step": 813
+    },
+    {
+      "epoch": 0.104192,
+      "grad_norm": 0.6286702156066895,
+      "learning_rate": 0.0009856611097905955,
+      "loss": 2.9767,
+      "step": 814
+    },
+    {
+      "epoch": 0.10432,
+      "grad_norm": 0.6694437861442566,
+      "learning_rate": 0.0009856117763047114,
+      "loss": 3.1646,
+      "step": 815
+    },
+    {
+      "epoch": 0.104448,
+      "grad_norm": 0.6423718929290771,
+      "learning_rate": 0.0009855623593365331,
+      "loss": 3.088,
+      "step": 816
+    },
+    {
+      "epoch": 0.104576,
+      "grad_norm": 0.6717129349708557,
+      "learning_rate": 0.000985512858894556,
+      "loss": 2.8879,
+      "step": 817
+    },
+    {
+      "epoch": 0.104704,
+      "grad_norm": 0.7187593579292297,
+      "learning_rate": 0.0009854632749872899,
+      "loss": 3.0344,
+      "step": 818
+    },
+    {
+      "epoch": 0.104832,
+      "grad_norm": 0.6191635131835938,
+      "learning_rate": 0.0009854136076232587,
+      "loss": 3.0587,
+      "step": 819
+    },
+    {
+      "epoch": 0.10496,
+      "grad_norm": 0.798617959022522,
+      "learning_rate": 0.0009853638568110012,
+      "loss": 3.0587,
+      "step": 820
+    },
+    {
+      "epoch": 0.105088,
+      "grad_norm": 0.800321638584137,
+      "learning_rate": 0.0009853140225590698,
+      "loss": 3.1221,
+      "step": 821
+    },
+    {
+      "epoch": 0.105216,
+      "grad_norm": 0.7488496899604797,
+      "learning_rate": 0.0009852641048760317,
+      "loss": 3.0398,
+      "step": 822
+    },
+    {
+      "epoch": 0.105344,
+      "grad_norm": 0.7354324460029602,
+      "learning_rate": 0.0009852141037704682,
+      "loss": 2.9601,
+      "step": 823
+    },
+    {
+      "epoch": 0.105472,
+      "grad_norm": 0.6556381583213806,
+      "learning_rate": 0.0009851640192509753,
+      "loss": 2.853,
+      "step": 824
+    },
+    {
+      "epoch": 0.1056,
+      "grad_norm": 0.698973536491394,
+      "learning_rate": 0.0009851138513261631,
+      "loss": 3.1229,
+      "step": 825
+    },
+    {
+      "epoch": 0.105728,
+      "grad_norm": 0.6646164655685425,
+      "learning_rate": 0.0009850636000046558,
+      "loss": 2.9714,
+      "step": 826
+    },
+    {
+      "epoch": 0.105856,
+      "grad_norm": 0.6660075187683105,
+      "learning_rate": 0.0009850132652950922,
+      "loss": 2.9742,
+      "step": 827
+    },
+    {
+      "epoch": 0.105984,
+      "grad_norm": 0.7060567140579224,
+      "learning_rate": 0.0009849628472061258,
+      "loss": 2.9942,
+      "step": 828
+    },
+    {
+      "epoch": 0.106112,
+      "grad_norm": 0.7005155682563782,
+      "learning_rate": 0.0009849123457464236,
+      "loss": 2.9736,
+      "step": 829
+    },
+    {
+      "epoch": 0.10624,
+      "grad_norm": 0.6527014970779419,
+      "learning_rate": 0.0009848617609246677,
+      "loss": 2.9855,
+      "step": 830
+    },
+    {
+      "epoch": 0.106368,
+      "grad_norm": 0.7236108183860779,
+      "learning_rate": 0.0009848110927495538,
+      "loss": 2.9739,
+      "step": 831
+    },
+    {
+      "epoch": 0.106496,
+      "grad_norm": 0.7211366295814514,
+      "learning_rate": 0.000984760341229793,
+      "loss": 2.8637,
+      "step": 832
+    },
+    {
+      "epoch": 0.106624,
+      "grad_norm": 0.6863318085670471,
+      "learning_rate": 0.0009847095063741096,
+      "loss": 3.135,
+      "step": 833
+    },
+    {
+      "epoch": 0.106752,
+      "grad_norm": 0.7190727591514587,
+      "learning_rate": 0.000984658588191243,
+      "loss": 3.0029,
+      "step": 834
+    },
+    {
+      "epoch": 0.10688,
+      "grad_norm": 0.7159355282783508,
+      "learning_rate": 0.0009846075866899462,
+      "loss": 3.0768,
+      "step": 835
+    },
+    {
+      "epoch": 0.107008,
+      "grad_norm": 0.6249812245368958,
+      "learning_rate": 0.0009845565018789873,
+      "loss": 3.0405,
+      "step": 836
+    },
+    {
+      "epoch": 0.107136,
+      "grad_norm": 0.6748888492584229,
+      "learning_rate": 0.0009845053337671481,
+      "loss": 3.0023,
+      "step": 837
+    },
+    {
+      "epoch": 0.107264,
+      "grad_norm": 0.7015107274055481,
+      "learning_rate": 0.0009844540823632252,
+      "loss": 3.0302,
+      "step": 838
+    },
+    {
+      "epoch": 0.107392,
+      "grad_norm": 0.7007777690887451,
+      "learning_rate": 0.0009844027476760295,
+      "loss": 3.0355,
+      "step": 839
+    },
+    {
+      "epoch": 0.10752,
+      "grad_norm": 0.701545000076294,
+      "learning_rate": 0.0009843513297143856,
+      "loss": 2.9461,
+      "step": 840
+    },
+    {
+      "epoch": 0.107648,
+      "grad_norm": 0.6545088887214661,
+      "learning_rate": 0.0009842998284871332,
+      "loss": 2.9514,
+      "step": 841
+    },
+    {
+      "epoch": 0.107776,
+      "grad_norm": 0.6611624360084534,
+      "learning_rate": 0.0009842482440031255,
+      "loss": 3.0429,
+      "step": 842
+    },
+    {
+      "epoch": 0.107904,
+      "grad_norm": 0.7076961994171143,
+      "learning_rate": 0.0009841965762712312,
+      "loss": 2.9869,
+      "step": 843
+    },
+    {
+      "epoch": 0.108032,
+      "grad_norm": 0.6754920482635498,
+      "learning_rate": 0.0009841448253003317,
+      "loss": 3.0597,
+      "step": 844
+    },
+    {
+      "epoch": 0.10816,
+      "grad_norm": 0.679995059967041,
+      "learning_rate": 0.0009840929910993244,
+      "loss": 3.105,
+      "step": 845
+    },
+    {
+      "epoch": 0.108288,
+      "grad_norm": 0.6516468524932861,
+      "learning_rate": 0.0009840410736771197,
+      "loss": 2.8662,
+      "step": 846
+    },
+    {
+      "epoch": 0.108416,
+      "grad_norm": 0.6538189053535461,
+      "learning_rate": 0.000983989073042643,
+      "loss": 3.0215,
+      "step": 847
+    },
+    {
+      "epoch": 0.108544,
+      "grad_norm": 0.7106131911277771,
+      "learning_rate": 0.0009839369892048336,
+      "loss": 3.1225,
+      "step": 848
+    },
+    {
+      "epoch": 0.108672,
+      "grad_norm": 0.6414303779602051,
+      "learning_rate": 0.0009838848221726455,
+      "loss": 2.9672,
+      "step": 849
+    },
+    {
+      "epoch": 0.1088,
+      "grad_norm": 0.6354730129241943,
+      "learning_rate": 0.000983832571955047,
+      "loss": 3.1875,
+      "step": 850
+    },
+    {
+      "epoch": 0.108928,
+      "grad_norm": 0.703874409198761,
+      "learning_rate": 0.00098378023856102,
+      "loss": 2.9933,
+      "step": 851
+    },
+    {
+      "epoch": 0.109056,
+      "grad_norm": 0.8571790456771851,
+      "learning_rate": 0.0009837278219995617,
+      "loss": 3.1022,
+      "step": 852
+    },
+    {
+      "epoch": 0.109184,
+      "grad_norm": 0.6648741364479065,
+      "learning_rate": 0.000983675322279683,
+      "loss": 3.0231,
+      "step": 853
+    },
+    {
+      "epoch": 0.109312,
+      "grad_norm": 0.7008213996887207,
+      "learning_rate": 0.0009836227394104092,
+      "loss": 3.0569,
+      "step": 854
+    },
+    {
+      "epoch": 0.10944,
+      "grad_norm": 0.6504847407341003,
+      "learning_rate": 0.0009835700734007796,
+      "loss": 3.0028,
+      "step": 855
+    },
+    {
+      "epoch": 0.109568,
+      "grad_norm": 0.6611882448196411,
+      "learning_rate": 0.0009835173242598485,
+      "loss": 3.0537,
+      "step": 856
+    },
+    {
+      "epoch": 0.109696,
+      "grad_norm": 0.6329249143600464,
+      "learning_rate": 0.0009834644919966841,
+      "loss": 2.9653,
+      "step": 857
+    },
+    {
+      "epoch": 0.109824,
+      "grad_norm": 0.6773561835289001,
+      "learning_rate": 0.0009834115766203686,
+      "loss": 2.9243,
+      "step": 858
+    },
+    {
+      "epoch": 0.109952,
+      "grad_norm": 0.6781663298606873,
+      "learning_rate": 0.000983358578139999,
+      "loss": 2.97,
+      "step": 859
+    },
+    {
+      "epoch": 0.11008,
+      "grad_norm": 0.6622476577758789,
+      "learning_rate": 0.000983305496564686,
+      "loss": 3.1151,
+      "step": 860
+    },
+    {
+      "epoch": 0.110208,
+      "grad_norm": 0.6058017611503601,
+      "learning_rate": 0.0009832523319035554,
+      "loss": 3.0524,
+      "step": 861
+    },
+    {
+      "epoch": 0.110336,
+      "grad_norm": 0.6185407042503357,
+      "learning_rate": 0.0009831990841657463,
+      "loss": 2.9205,
+      "step": 862
+    },
+    {
+      "epoch": 0.110464,
+      "grad_norm": 0.670077383518219,
+      "learning_rate": 0.0009831457533604133,
+      "loss": 3.0496,
+      "step": 863
+    },
+    {
+      "epoch": 0.110592,
+      "grad_norm": 0.7041655778884888,
+      "learning_rate": 0.000983092339496724,
+      "loss": 2.9705,
+      "step": 864
+    },
+    {
+      "epoch": 0.11072,
+      "grad_norm": 0.7642086148262024,
+      "learning_rate": 0.000983038842583861,
+      "loss": 3.1038,
+      "step": 865
+    },
+    {
+      "epoch": 0.110848,
+      "grad_norm": 0.7351483702659607,
+      "learning_rate": 0.0009829852626310211,
+      "loss": 2.8273,
+      "step": 866
+    },
+    {
+      "epoch": 0.110976,
+      "grad_norm": 0.7130418419837952,
+      "learning_rate": 0.0009829315996474155,
+      "loss": 2.9804,
+      "step": 867
+    },
+    {
+      "epoch": 0.111104,
+      "grad_norm": 0.6716287136077881,
+      "learning_rate": 0.0009828778536422691,
+      "loss": 3.0872,
+      "step": 868
+    },
+    {
+      "epoch": 0.111232,
+      "grad_norm": 0.6280295848846436,
+      "learning_rate": 0.0009828240246248217,
+      "loss": 3.1011,
+      "step": 869
+    },
+    {
+      "epoch": 0.11136,
+      "grad_norm": 0.7491655349731445,
+      "learning_rate": 0.0009827701126043271,
+      "loss": 3.1416,
+      "step": 870
+    },
+    {
+      "epoch": 0.111488,
+      "grad_norm": 0.6838396787643433,
+      "learning_rate": 0.0009827161175900534,
+      "loss": 2.9674,
+      "step": 871
+    },
+    {
+      "epoch": 0.111616,
+      "grad_norm": 0.6280179619789124,
+      "learning_rate": 0.0009826620395912832,
+      "loss": 3.0312,
+      "step": 872
+    },
+    {
+      "epoch": 0.111744,
+      "grad_norm": 0.6673280596733093,
+      "learning_rate": 0.0009826078786173125,
+      "loss": 3.0064,
+      "step": 873
+    },
+    {
+      "epoch": 0.111872,
+      "grad_norm": 0.6568505764007568,
+      "learning_rate": 0.000982553634677453,
+      "loss": 2.9871,
+      "step": 874
+    },
+    {
+      "epoch": 0.112,
+      "grad_norm": 0.6817496418952942,
+      "learning_rate": 0.000982499307781029,
+      "loss": 3.0198,
+      "step": 875
+    },
+    {
+      "epoch": 0.112128,
+      "grad_norm": 0.7104738354682922,
+      "learning_rate": 0.0009824448979373806,
+      "loss": 2.9603,
+      "step": 876
+    },
+    {
+      "epoch": 0.112256,
+      "grad_norm": 0.7548314929008484,
+      "learning_rate": 0.0009823904051558612,
+      "loss": 3.128,
+      "step": 877
+    },
+    {
+      "epoch": 0.112384,
+      "grad_norm": 0.6923811435699463,
+      "learning_rate": 0.0009823358294458388,
+      "loss": 3.1969,
+      "step": 878
+    },
+    {
+      "epoch": 0.112512,
+      "grad_norm": 0.6278347373008728,
+      "learning_rate": 0.0009822811708166959,
+      "loss": 3.0333,
+      "step": 879
+    },
+    {
+      "epoch": 0.11264,
+      "grad_norm": 0.6198648810386658,
+      "learning_rate": 0.0009822264292778282,
+      "loss": 3.0507,
+      "step": 880
+    },
+    {
+      "epoch": 0.112768,
+      "grad_norm": 0.7954167723655701,
+      "learning_rate": 0.0009821716048386473,
+      "loss": 2.9207,
+      "step": 881
+    },
+    {
+      "epoch": 0.112896,
+      "grad_norm": 0.685231626033783,
+      "learning_rate": 0.0009821166975085773,
+      "loss": 3.0447,
+      "step": 882
+    },
+    {
+      "epoch": 0.113024,
+      "grad_norm": 0.659488320350647,
+      "learning_rate": 0.0009820617072970582,
+      "loss": 2.9316,
+      "step": 883
+    },
+    {
+      "epoch": 0.113152,
+      "grad_norm": 0.6866512298583984,
+      "learning_rate": 0.0009820066342135427,
+      "loss": 3.0442,
+      "step": 884
+    },
+    {
+      "epoch": 0.11328,
+      "grad_norm": 0.6297601461410522,
+      "learning_rate": 0.0009819514782674992,
+      "loss": 2.9734,
+      "step": 885
+    },
+    {
+      "epoch": 0.113408,
+      "grad_norm": 0.6448697447776794,
+      "learning_rate": 0.0009818962394684092,
+      "loss": 3.0624,
+      "step": 886
+    },
+    {
+      "epoch": 0.113536,
+      "grad_norm": 0.6619642376899719,
+      "learning_rate": 0.000981840917825769,
+      "loss": 3.0462,
+      "step": 887
+    },
+    {
+      "epoch": 0.113664,
+      "grad_norm": 0.7215335965156555,
+      "learning_rate": 0.0009817855133490891,
+      "loss": 2.9871,
+      "step": 888
+    },
+    {
+      "epoch": 0.113792,
+      "grad_norm": 0.7220499515533447,
+      "learning_rate": 0.000981730026047894,
+      "loss": 2.9392,
+      "step": 889
+    },
+    {
+      "epoch": 0.11392,
+      "grad_norm": 0.6859092712402344,
+      "learning_rate": 0.0009816744559317228,
+      "loss": 2.9221,
+      "step": 890
+    },
+    {
+      "epoch": 0.114048,
+      "grad_norm": 0.6807576417922974,
+      "learning_rate": 0.0009816188030101285,
+      "loss": 3.0057,
+      "step": 891
+    },
+    {
+      "epoch": 0.114176,
+      "grad_norm": 0.6299679279327393,
+      "learning_rate": 0.0009815630672926788,
+      "loss": 3.0794,
+      "step": 892
+    },
+    {
+      "epoch": 0.114304,
+      "grad_norm": 0.6836426258087158,
+      "learning_rate": 0.0009815072487889549,
+      "loss": 3.1903,
+      "step": 893
+    },
+    {
+      "epoch": 0.114432,
+      "grad_norm": 0.6786308884620667,
+      "learning_rate": 0.000981451347508553,
+      "loss": 2.9425,
+      "step": 894
+    },
+    {
+      "epoch": 0.11456,
+      "grad_norm": 0.6136744022369385,
+      "learning_rate": 0.0009813953634610826,
+      "loss": 3.0139,
+      "step": 895
+    },
+    {
+      "epoch": 0.114688,
+      "grad_norm": 0.706629753112793,
+      "learning_rate": 0.0009813392966561687,
+      "loss": 2.9903,
+      "step": 896
+    },
+    {
+      "epoch": 0.114816,
+      "grad_norm": 0.6718335747718811,
+      "learning_rate": 0.0009812831471034495,
+      "loss": 3.089,
+      "step": 897
+    },
+    {
+      "epoch": 0.114944,
+      "grad_norm": 0.6958935260772705,
+      "learning_rate": 0.0009812269148125777,
+      "loss": 2.9675,
+      "step": 898
+    },
+    {
+      "epoch": 0.115072,
+      "grad_norm": 0.701296865940094,
+      "learning_rate": 0.0009811705997932205,
+      "loss": 2.9206,
+      "step": 899
+    },
+    {
+      "epoch": 0.1152,
+      "grad_norm": 0.677127480506897,
+      "learning_rate": 0.000981114202055059,
+      "loss": 2.8816,
+      "step": 900
+    },
+    {
+      "epoch": 0.115328,
+      "grad_norm": 0.7315416932106018,
+      "learning_rate": 0.0009810577216077885,
+      "loss": 3.0763,
+      "step": 901
+    },
+    {
+      "epoch": 0.115456,
+      "grad_norm": 0.7221956849098206,
+      "learning_rate": 0.000981001158461119,
+      "loss": 3.0761,
+      "step": 902
+    },
+    {
+      "epoch": 0.115584,
+      "grad_norm": 0.6711305975914001,
+      "learning_rate": 0.0009809445126247739,
+      "loss": 2.9588,
+      "step": 903
+    },
+    {
+      "epoch": 0.115712,
+      "grad_norm": 0.6964670419692993,
+      "learning_rate": 0.0009808877841084914,
+      "loss": 3.0097,
+      "step": 904
+    },
+    {
+      "epoch": 0.11584,
+      "grad_norm": 0.6524249911308289,
+      "learning_rate": 0.000980830972922024,
+      "loss": 3.0315,
+      "step": 905
+    },
+    {
+      "epoch": 0.115968,
+      "grad_norm": 0.633556067943573,
+      "learning_rate": 0.0009807740790751382,
+      "loss": 2.9139,
+      "step": 906
+    },
+    {
+      "epoch": 0.116096,
+      "grad_norm": 0.6520939469337463,
+      "learning_rate": 0.0009807171025776145,
+      "loss": 3.0239,
+      "step": 907
+    },
+    {
+      "epoch": 0.116224,
+      "grad_norm": 0.6795200705528259,
+      "learning_rate": 0.0009806600434392478,
+      "loss": 3.0616,
+      "step": 908
+    },
+    {
+      "epoch": 0.116352,
+      "grad_norm": 0.645137369632721,
+      "learning_rate": 0.0009806029016698474,
+      "loss": 3.0465,
+      "step": 909
+    },
+    {
+      "epoch": 0.11648,
+      "grad_norm": 0.6600620746612549,
+      "learning_rate": 0.0009805456772792368,
+      "loss": 3.0705,
+      "step": 910
+    },
+    {
+      "epoch": 0.116608,
+      "grad_norm": 0.7105903029441833,
+      "learning_rate": 0.0009804883702772531,
+      "loss": 3.0079,
+      "step": 911
+    },
+    {
+      "epoch": 0.116736,
+      "grad_norm": 0.6878994107246399,
+      "learning_rate": 0.0009804309806737482,
+      "loss": 2.9543,
+      "step": 912
+    },
+    {
+      "epoch": 0.116864,
+      "grad_norm": 0.6337210536003113,
+      "learning_rate": 0.0009803735084785883,
+      "loss": 2.9613,
+      "step": 913
+    },
+    {
+      "epoch": 0.116992,
+      "grad_norm": 0.6976180076599121,
+      "learning_rate": 0.000980315953701653,
+      "loss": 3.0568,
+      "step": 914
+    },
+    {
+      "epoch": 0.11712,
+      "grad_norm": 0.7079159021377563,
+      "learning_rate": 0.0009802583163528373,
+      "loss": 3.0544,
+      "step": 915
+    },
+    {
+      "epoch": 0.117248,
+      "grad_norm": 0.6428760290145874,
+      "learning_rate": 0.0009802005964420493,
+      "loss": 3.0969,
+      "step": 916
+    },
+    {
+      "epoch": 0.117376,
+      "grad_norm": 0.6490296125411987,
+      "learning_rate": 0.0009801427939792118,
+      "loss": 3.0194,
+      "step": 917
+    },
+    {
+      "epoch": 0.117504,
+      "grad_norm": 0.7238758206367493,
+      "learning_rate": 0.0009800849089742615,
+      "loss": 3.0541,
+      "step": 918
+    },
+    {
+      "epoch": 0.117632,
+      "grad_norm": 0.6917224526405334,
+      "learning_rate": 0.0009800269414371502,
+      "loss": 3.0178,
+      "step": 919
+    },
+    {
+      "epoch": 0.11776,
+      "grad_norm": 0.6229843497276306,
+      "learning_rate": 0.000979968891377842,
+      "loss": 3.1517,
+      "step": 920
+    },
+    {
+      "epoch": 0.117888,
+      "grad_norm": 0.6766431331634521,
+      "learning_rate": 0.0009799107588063178,
+      "loss": 2.9059,
+      "step": 921
+    },
+    {
+      "epoch": 0.118016,
+      "grad_norm": 0.6575608849525452,
+      "learning_rate": 0.0009798525437325703,
+      "loss": 3.193,
+      "step": 922
+    },
+    {
+      "epoch": 0.118144,
+      "grad_norm": 0.6358314752578735,
+      "learning_rate": 0.0009797942461666075,
+      "loss": 3.0516,
+      "step": 923
+    },
+    {
+      "epoch": 0.118272,
+      "grad_norm": 0.7094784379005432,
+      "learning_rate": 0.0009797358661184517,
+      "loss": 2.9891,
+      "step": 924
+    },
+    {
+      "epoch": 0.1184,
+      "grad_norm": 0.6845430731773376,
+      "learning_rate": 0.0009796774035981388,
+      "loss": 2.9461,
+      "step": 925
+    },
+    {
+      "epoch": 0.118528,
+      "grad_norm": 0.7136852741241455,
+      "learning_rate": 0.0009796188586157193,
+      "loss": 3.0323,
+      "step": 926
+    },
+    {
+      "epoch": 0.118656,
+      "grad_norm": 0.6409615278244019,
+      "learning_rate": 0.000979560231181258,
+      "loss": 2.8108,
+      "step": 927
+    },
+    {
+      "epoch": 0.118784,
+      "grad_norm": 0.6721232533454895,
+      "learning_rate": 0.0009795015213048333,
+      "loss": 3.076,
+      "step": 928
+    },
+    {
+      "epoch": 0.118912,
+      "grad_norm": 0.6529289484024048,
+      "learning_rate": 0.0009794427289965381,
+      "loss": 2.9431,
+      "step": 929
+    },
+    {
+      "epoch": 0.11904,
+      "grad_norm": 0.6708246469497681,
+      "learning_rate": 0.0009793838542664797,
+      "loss": 2.8882,
+      "step": 930
+    },
+    {
+      "epoch": 0.119168,
+      "grad_norm": 0.6168515682220459,
+      "learning_rate": 0.0009793248971247791,
+      "loss": 2.9579,
+      "step": 931
+    },
+    {
+      "epoch": 0.119296,
+      "grad_norm": 0.7099841237068176,
+      "learning_rate": 0.0009792658575815724,
+      "loss": 2.8773,
+      "step": 932
+    },
+    {
+      "epoch": 0.119424,
+      "grad_norm": 0.669234037399292,
+      "learning_rate": 0.0009792067356470082,
+      "loss": 3.0627,
+      "step": 933
+    },
+    {
+      "epoch": 0.119552,
+      "grad_norm": 0.6550662517547607,
+      "learning_rate": 0.0009791475313312508,
+      "loss": 3.0336,
+      "step": 934
+    },
+    {
+      "epoch": 0.11968,
+      "grad_norm": 0.6444387435913086,
+      "learning_rate": 0.000979088244644478,
+      "loss": 3.017,
+      "step": 935
+    },
+    {
+      "epoch": 0.119808,
+      "grad_norm": 0.6627427339553833,
+      "learning_rate": 0.000979028875596882,
+      "loss": 3.0479,
+      "step": 936
+    },
+    {
+      "epoch": 0.119936,
+      "grad_norm": 0.6445711255073547,
+      "learning_rate": 0.0009789694241986687,
+      "loss": 3.0049,
+      "step": 937
+    },
+    {
+      "epoch": 0.120064,
+      "grad_norm": 0.7065802812576294,
+      "learning_rate": 0.0009789098904600588,
+      "loss": 3.0053,
+      "step": 938
+    },
+    {
+      "epoch": 0.120192,
+      "grad_norm": 0.6788486242294312,
+      "learning_rate": 0.0009788502743912867,
+      "loss": 2.985,
+      "step": 939
+    },
+    {
+      "epoch": 0.12032,
+      "grad_norm": 0.7369581460952759,
+      "learning_rate": 0.000978790576002601,
+      "loss": 2.9527,
+      "step": 940
+    },
+    {
+      "epoch": 0.120448,
+      "grad_norm": 0.6423206925392151,
+      "learning_rate": 0.000978730795304265,
+      "loss": 2.9244,
+      "step": 941
+    },
+    {
+      "epoch": 0.120576,
+      "grad_norm": 0.6415398120880127,
+      "learning_rate": 0.0009786709323065552,
+      "loss": 3.0303,
+      "step": 942
+    },
+    {
+      "epoch": 0.120704,
+      "grad_norm": 0.604226291179657,
+      "learning_rate": 0.0009786109870197628,
+      "loss": 3.0354,
+      "step": 943
+    },
+    {
+      "epoch": 0.120832,
+      "grad_norm": 0.6808715462684631,
+      "learning_rate": 0.0009785509594541933,
+      "loss": 2.9731,
+      "step": 944
+    },
+    {
+      "epoch": 0.12096,
+      "grad_norm": 0.6742005348205566,
+      "learning_rate": 0.000978490849620166,
+      "loss": 3.1239,
+      "step": 945
+    },
+    {
+      "epoch": 0.121088,
+      "grad_norm": 0.6544292569160461,
+      "learning_rate": 0.0009784306575280144,
+      "loss": 3.0094,
+      "step": 946
+    },
+    {
+      "epoch": 0.121216,
+      "grad_norm": 0.6111042499542236,
+      "learning_rate": 0.0009783703831880866,
+      "loss": 2.9054,
+      "step": 947
+    },
+    {
+      "epoch": 0.121344,
+      "grad_norm": 0.7061547636985779,
+      "learning_rate": 0.000978310026610744,
+      "loss": 2.9045,
+      "step": 948
+    },
+    {
+      "epoch": 0.121472,
+      "grad_norm": 0.6494823694229126,
+      "learning_rate": 0.0009782495878063629,
+      "loss": 2.9482,
+      "step": 949
+    },
+    {
+      "epoch": 0.1216,
+      "grad_norm": 0.656141459941864,
+      "learning_rate": 0.000978189066785333,
+      "loss": 3.0161,
+      "step": 950
+    },
+    {
+      "epoch": 0.121728,
+      "grad_norm": 0.6820494532585144,
+      "learning_rate": 0.0009781284635580591,
+      "loss": 2.9663,
+      "step": 951
+    },
+    {
+      "epoch": 0.121856,
+      "grad_norm": 0.7006909251213074,
+      "learning_rate": 0.0009780677781349595,
+      "loss": 3.0943,
+      "step": 952
+    },
+    {
+      "epoch": 0.121984,
+      "grad_norm": 0.6687823534011841,
+      "learning_rate": 0.0009780070105264666,
+      "loss": 3.0386,
+      "step": 953
+    },
+    {
+      "epoch": 0.122112,
+      "grad_norm": 0.6484476923942566,
+      "learning_rate": 0.0009779461607430267,
+      "loss": 3.0651,
+      "step": 954
+    },
+    {
+      "epoch": 0.12224,
+      "grad_norm": 0.7069973349571228,
+      "learning_rate": 0.0009778852287951013,
+      "loss": 3.0661,
+      "step": 955
+    },
+    {
+      "epoch": 0.122368,
+      "grad_norm": 0.6811779141426086,
+      "learning_rate": 0.000977824214693165,
+      "loss": 3.017,
+      "step": 956
+    },
+    {
+      "epoch": 0.122496,
+      "grad_norm": 0.6682418584823608,
+      "learning_rate": 0.0009777631184477065,
+      "loss": 2.9809,
+      "step": 957
+    },
+    {
+      "epoch": 0.122624,
+      "grad_norm": 0.6194751262664795,
+      "learning_rate": 0.0009777019400692291,
+      "loss": 2.9837,
+      "step": 958
+    },
+    {
+      "epoch": 0.122752,
+      "grad_norm": 0.6731294393539429,
+      "learning_rate": 0.0009776406795682505,
+      "loss": 3.0092,
+      "step": 959
+    },
+    {
+      "epoch": 0.12288,
+      "grad_norm": 0.601652979850769,
+      "learning_rate": 0.0009775793369553017,
+      "loss": 2.9253,
+      "step": 960
+    },
+    {
+      "epoch": 0.123008,
+      "grad_norm": 0.6316988468170166,
+      "learning_rate": 0.0009775179122409284,
+      "loss": 2.8814,
+      "step": 961
+    },
+    {
+      "epoch": 0.123136,
+      "grad_norm": 0.6373597383499146,
+      "learning_rate": 0.00097745640543569,
+      "loss": 2.9355,
+      "step": 962
+    },
+    {
+      "epoch": 0.123264,
+      "grad_norm": 0.6404913067817688,
+      "learning_rate": 0.0009773948165501605,
+      "loss": 2.9559,
+      "step": 963
+    },
+    {
+      "epoch": 0.123392,
+      "grad_norm": 0.7208508849143982,
+      "learning_rate": 0.0009773331455949275,
+      "loss": 2.9452,
+      "step": 964
+    },
+    {
+      "epoch": 0.12352,
+      "grad_norm": 0.6533061265945435,
+      "learning_rate": 0.000977271392580593,
+      "loss": 3.0641,
+      "step": 965
+    },
+    {
+      "epoch": 0.123648,
+      "grad_norm": 0.652422308921814,
+      "learning_rate": 0.0009772095575177732,
+      "loss": 3.0209,
+      "step": 966
+    },
+    {
+      "epoch": 0.123776,
+      "grad_norm": 0.6725317239761353,
+      "learning_rate": 0.0009771476404170983,
+      "loss": 3.095,
+      "step": 967
+    },
+    {
+      "epoch": 0.123904,
+      "grad_norm": 0.6512139439582825,
+      "learning_rate": 0.0009770856412892122,
+      "loss": 3.0029,
+      "step": 968
+    },
+    {
+      "epoch": 0.124032,
+      "grad_norm": 0.647534966468811,
+      "learning_rate": 0.0009770235601447737,
+      "loss": 3.1243,
+      "step": 969
+    },
+    {
+      "epoch": 0.12416,
+      "grad_norm": 0.6864913105964661,
+      "learning_rate": 0.0009769613969944548,
+      "loss": 2.9787,
+      "step": 970
+    },
+    {
+      "epoch": 0.124288,
+      "grad_norm": 0.6798150539398193,
+      "learning_rate": 0.0009768991518489425,
+      "loss": 2.9837,
+      "step": 971
+    },
+    {
+      "epoch": 0.124416,
+      "grad_norm": 0.6908317804336548,
+      "learning_rate": 0.0009768368247189375,
+      "loss": 3.0107,
+      "step": 972
+    },
+    {
+      "epoch": 0.124544,
+      "grad_norm": 0.6326662302017212,
+      "learning_rate": 0.0009767744156151542,
+      "loss": 2.909,
+      "step": 973
+    },
+    {
+      "epoch": 0.124672,
+      "grad_norm": 0.661371111869812,
+      "learning_rate": 0.0009767119245483215,
+      "loss": 3.0347,
+      "step": 974
+    },
+    {
+      "epoch": 0.1248,
+      "grad_norm": 0.6334081888198853,
+      "learning_rate": 0.0009766493515291827,
+      "loss": 3.0695,
+      "step": 975
+    },
+    {
+      "epoch": 0.124928,
+      "grad_norm": 0.6248986721038818,
+      "learning_rate": 0.0009765866965684945,
+      "loss": 2.9017,
+      "step": 976
+    },
+    {
+      "epoch": 0.125056,
+      "grad_norm": 0.5554507970809937,
+      "learning_rate": 0.0009765239596770279,
+      "loss": 2.9821,
+      "step": 977
+    },
+    {
+      "epoch": 0.125184,
+      "grad_norm": 0.6525313258171082,
+      "learning_rate": 0.0009764611408655685,
+      "loss": 3.0707,
+      "step": 978
+    },
+    {
+      "epoch": 0.125312,
+      "grad_norm": 0.6555005311965942,
+      "learning_rate": 0.0009763982401449153,
+      "loss": 3.0267,
+      "step": 979
+    },
+    {
+      "epoch": 0.12544,
+      "grad_norm": 0.6786180138587952,
+      "learning_rate": 0.0009763352575258817,
+      "loss": 3.0255,
+      "step": 980
+    },
+    {
+      "epoch": 0.125568,
+      "grad_norm": 0.6427340507507324,
+      "learning_rate": 0.0009762721930192952,
+      "loss": 2.9915,
+      "step": 981
+    },
+    {
+      "epoch": 0.125696,
+      "grad_norm": 0.6113302111625671,
+      "learning_rate": 0.0009762090466359974,
+      "loss": 2.8979,
+      "step": 982
+    },
+    {
+      "epoch": 0.125824,
+      "grad_norm": 0.6821973323822021,
+      "learning_rate": 0.0009761458183868437,
+      "loss": 3.0654,
+      "step": 983
+    },
+    {
+      "epoch": 0.125952,
+      "grad_norm": 0.6569185853004456,
+      "learning_rate": 0.0009760825082827039,
+      "loss": 3.0301,
+      "step": 984
+    },
+    {
+      "epoch": 0.12608,
+      "grad_norm": 0.6979755163192749,
+      "learning_rate": 0.0009760191163344617,
+      "loss": 2.8523,
+      "step": 985
+    },
+    {
+      "epoch": 0.126208,
+      "grad_norm": 0.6744502186775208,
+      "learning_rate": 0.000975955642553015,
+      "loss": 2.8998,
+      "step": 986
+    },
+    {
+      "epoch": 0.126336,
+      "grad_norm": 0.6875883340835571,
+      "learning_rate": 0.0009758920869492754,
+      "loss": 2.8748,
+      "step": 987
+    },
+    {
+      "epoch": 0.126464,
+      "grad_norm": 0.6680119633674622,
+      "learning_rate": 0.0009758284495341692,
+      "loss": 2.9972,
+      "step": 988
+    },
+    {
+      "epoch": 0.126592,
+      "grad_norm": 0.6757093667984009,
+      "learning_rate": 0.0009757647303186361,
+      "loss": 2.975,
+      "step": 989
+    },
+    {
+      "epoch": 0.12672,
+      "grad_norm": 0.6539942026138306,
+      "learning_rate": 0.0009757009293136303,
+      "loss": 2.9738,
+      "step": 990
+    },
+    {
+      "epoch": 0.126848,
+      "grad_norm": 0.6346407532691956,
+      "learning_rate": 0.00097563704653012,
+      "loss": 2.9314,
+      "step": 991
+    },
+    {
+      "epoch": 0.126976,
+      "grad_norm": 0.7313452959060669,
+      "learning_rate": 0.0009755730819790876,
+      "loss": 3.1981,
+      "step": 992
+    },
+    {
+      "epoch": 0.127104,
+      "grad_norm": 0.6172248125076294,
+      "learning_rate": 0.0009755090356715288,
+      "loss": 2.9656,
+      "step": 993
+    },
+    {
+      "epoch": 0.127232,
+      "grad_norm": 0.6944584846496582,
+      "learning_rate": 0.0009754449076184543,
+      "loss": 2.9663,
+      "step": 994
+    },
+    {
+      "epoch": 0.12736,
+      "grad_norm": 0.6243419647216797,
+      "learning_rate": 0.0009753806978308883,
+      "loss": 3.0869,
+      "step": 995
+    },
+    {
+      "epoch": 0.127488,
+      "grad_norm": 0.6634061336517334,
+      "learning_rate": 0.0009753164063198693,
+      "loss": 3.0,
+      "step": 996
+    },
+    {
+      "epoch": 0.127616,
+      "grad_norm": 0.6448676586151123,
+      "learning_rate": 0.0009752520330964496,
+      "loss": 3.0231,
+      "step": 997
+    },
+    {
+      "epoch": 0.127744,
+      "grad_norm": 0.6205151081085205,
+      "learning_rate": 0.000975187578171696,
+      "loss": 2.8803,
+      "step": 998
+    },
+    {
+      "epoch": 0.127872,
+      "grad_norm": 0.659408688545227,
+      "learning_rate": 0.0009751230415566887,
+      "loss": 3.1365,
+      "step": 999
+    },
+    {
+      "epoch": 0.128,
+      "grad_norm": 0.6099840402603149,
+      "learning_rate": 0.0009750584232625226,
+      "loss": 3.1019,
+      "step": 1000
+    },
+    {
+      "epoch": 0.128128,
+      "grad_norm": 0.6046210527420044,
+      "learning_rate": 0.0009749937233003061,
+      "loss": 2.8066,
+      "step": 1001
+    },
+    {
+      "epoch": 0.128256,
+      "grad_norm": 0.6078466176986694,
+      "learning_rate": 0.0009749289416811621,
+      "loss": 3.0049,
+      "step": 1002
+    },
+    {
+      "epoch": 0.128384,
+      "grad_norm": 0.6424407362937927,
+      "learning_rate": 0.000974864078416227,
+      "loss": 3.0378,
+      "step": 1003
+    },
+    {
+      "epoch": 0.128512,
+      "grad_norm": 0.6950150728225708,
+      "learning_rate": 0.0009747991335166519,
+      "loss": 3.0013,
+      "step": 1004
+    },
+    {
+      "epoch": 0.12864,
+      "grad_norm": 0.7067024111747742,
+      "learning_rate": 0.0009747341069936013,
+      "loss": 2.9354,
+      "step": 1005
+    },
+    {
+      "epoch": 0.128768,
+      "grad_norm": 0.6437323689460754,
+      "learning_rate": 0.000974668998858254,
+      "loss": 2.9491,
+      "step": 1006
+    },
+    {
+      "epoch": 0.128896,
+      "grad_norm": 0.716408371925354,
+      "learning_rate": 0.0009746038091218031,
+      "loss": 2.973,
+      "step": 1007
+    },
+    {
+      "epoch": 0.129024,
+      "grad_norm": 0.6274816393852234,
+      "learning_rate": 0.0009745385377954555,
+      "loss": 2.9348,
+      "step": 1008
+    },
+    {
+      "epoch": 0.129152,
+      "grad_norm": 0.621276319026947,
+      "learning_rate": 0.0009744731848904315,
+      "loss": 2.9777,
+      "step": 1009
+    },
+    {
+      "epoch": 0.12928,
+      "grad_norm": 0.622287392616272,
+      "learning_rate": 0.0009744077504179669,
+      "loss": 2.9105,
+      "step": 1010
+    },
+    {
+      "epoch": 0.129408,
+      "grad_norm": 0.6601535677909851,
+      "learning_rate": 0.0009743422343893098,
+      "loss": 2.9588,
+      "step": 1011
+    },
+    {
+      "epoch": 0.129536,
+      "grad_norm": 0.6426204442977905,
+      "learning_rate": 0.0009742766368157239,
+      "loss": 2.9451,
+      "step": 1012
+    },
+    {
+      "epoch": 0.129664,
+      "grad_norm": 0.637910783290863,
+      "learning_rate": 0.0009742109577084856,
+      "loss": 3.0266,
+      "step": 1013
+    },
+    {
+      "epoch": 0.129792,
+      "grad_norm": 0.9439998865127563,
+      "learning_rate": 0.0009741451970788862,
+      "loss": 2.8455,
+      "step": 1014
+    },
+    {
+      "epoch": 0.12992,
+      "grad_norm": 0.6799437403678894,
+      "learning_rate": 0.0009740793549382307,
+      "loss": 3.175,
+      "step": 1015
+    },
+    {
+      "epoch": 0.130048,
+      "grad_norm": 0.6601764559745789,
+      "learning_rate": 0.0009740134312978382,
+      "loss": 3.0329,
+      "step": 1016
+    },
+    {
+      "epoch": 0.130176,
+      "grad_norm": 0.6045657396316528,
+      "learning_rate": 0.0009739474261690415,
+      "loss": 3.0164,
+      "step": 1017
+    },
+    {
+      "epoch": 0.130304,
+      "grad_norm": 0.7210801243782043,
+      "learning_rate": 0.0009738813395631878,
+      "loss": 3.0328,
+      "step": 1018
+    },
+    {
+      "epoch": 0.130432,
+      "grad_norm": 0.6395325660705566,
+      "learning_rate": 0.0009738151714916381,
+      "loss": 2.808,
+      "step": 1019
+    },
+    {
+      "epoch": 0.13056,
+      "grad_norm": 0.6457556486129761,
+      "learning_rate": 0.0009737489219657675,
+      "loss": 2.9747,
+      "step": 1020
+    },
+    {
+      "epoch": 0.130688,
+      "grad_norm": 0.6614893674850464,
+      "learning_rate": 0.000973682590996965,
+      "loss": 3.0019,
+      "step": 1021
+    },
+    {
+      "epoch": 0.130816,
+      "grad_norm": 0.6543712019920349,
+      "learning_rate": 0.0009736161785966337,
+      "loss": 3.1628,
+      "step": 1022
+    },
+    {
+      "epoch": 0.130944,
+      "grad_norm": 0.6071307063102722,
+      "learning_rate": 0.0009735496847761908,
+      "loss": 2.9978,
+      "step": 1023
+    },
+    {
+      "epoch": 0.131072,
+      "grad_norm": 0.6605586409568787,
+      "learning_rate": 0.0009734831095470671,
+      "loss": 2.95,
+      "step": 1024
+    },
+    {
+      "epoch": 0.1312,
+      "grad_norm": 0.6144652366638184,
+      "learning_rate": 0.0009734164529207079,
+      "loss": 2.8919,
+      "step": 1025
+    },
+    {
+      "epoch": 0.131328,
+      "grad_norm": 0.6489017605781555,
+      "learning_rate": 0.0009733497149085719,
+      "loss": 2.9439,
+      "step": 1026
+    },
+    {
+      "epoch": 0.131456,
+      "grad_norm": 0.6536964774131775,
+      "learning_rate": 0.0009732828955221327,
+      "loss": 2.9848,
+      "step": 1027
+    },
+    {
+      "epoch": 0.131584,
+      "grad_norm": 0.6011839509010315,
+      "learning_rate": 0.0009732159947728767,
+      "loss": 2.9769,
+      "step": 1028
+    },
+    {
+      "epoch": 0.131712,
+      "grad_norm": 0.7403272390365601,
+      "learning_rate": 0.0009731490126723052,
+      "loss": 2.9247,
+      "step": 1029
+    },
+    {
+      "epoch": 0.13184,
+      "grad_norm": 0.6234657764434814,
+      "learning_rate": 0.0009730819492319335,
+      "loss": 3.0691,
+      "step": 1030
+    },
+    {
+      "epoch": 0.131968,
+      "grad_norm": 0.5914653539657593,
+      "learning_rate": 0.0009730148044632901,
+      "loss": 2.9486,
+      "step": 1031
+    },
+    {
+      "epoch": 0.132096,
+      "grad_norm": 0.6170773506164551,
+      "learning_rate": 0.0009729475783779181,
+      "loss": 3.0597,
+      "step": 1032
+    },
+    {
+      "epoch": 0.132224,
+      "grad_norm": 0.6669968366622925,
+      "learning_rate": 0.0009728802709873747,
+      "loss": 2.8668,
+      "step": 1033
+    },
+    {
+      "epoch": 0.132352,
+      "grad_norm": 0.6765456795692444,
+      "learning_rate": 0.0009728128823032305,
+      "loss": 3.0749,
+      "step": 1034
+    },
+    {
+      "epoch": 0.13248,
+      "grad_norm": 0.601654589176178,
+      "learning_rate": 0.0009727454123370705,
+      "loss": 2.8826,
+      "step": 1035
+    },
+    {
+      "epoch": 0.132608,
+      "grad_norm": 0.6667168140411377,
+      "learning_rate": 0.0009726778611004935,
+      "loss": 3.0536,
+      "step": 1036
+    },
+    {
+      "epoch": 0.132736,
+      "grad_norm": 0.6341603398323059,
+      "learning_rate": 0.0009726102286051127,
+      "loss": 3.0627,
+      "step": 1037
+    },
+    {
+      "epoch": 0.132864,
+      "grad_norm": 0.5939716696739197,
+      "learning_rate": 0.0009725425148625542,
+      "loss": 2.9936,
+      "step": 1038
+    },
+    {
+      "epoch": 0.132992,
+      "grad_norm": 0.6556738615036011,
+      "learning_rate": 0.0009724747198844595,
+      "loss": 3.0709,
+      "step": 1039
+    },
+    {
+      "epoch": 0.13312,
+      "grad_norm": 0.6814432144165039,
+      "learning_rate": 0.0009724068436824831,
+      "loss": 3.0927,
+      "step": 1040
+    },
+    {
+      "epoch": 0.133248,
+      "grad_norm": 0.6294246912002563,
+      "learning_rate": 0.0009723388862682936,
+      "loss": 2.8581,
+      "step": 1041
+    },
+    {
+      "epoch": 0.133376,
+      "grad_norm": 0.6464622616767883,
+      "learning_rate": 0.0009722708476535736,
+      "loss": 3.0752,
+      "step": 1042
+    },
+    {
+      "epoch": 0.133504,
+      "grad_norm": 0.6203954815864563,
+      "learning_rate": 0.0009722027278500201,
+      "loss": 3.0461,
+      "step": 1043
+    },
+    {
+      "epoch": 0.133632,
+      "grad_norm": 0.6666882634162903,
+      "learning_rate": 0.0009721345268693432,
+      "loss": 3.0712,
+      "step": 1044
+    },
+    {
+      "epoch": 0.13376,
+      "grad_norm": 0.6522104144096375,
+      "learning_rate": 0.0009720662447232677,
+      "loss": 3.0612,
+      "step": 1045
+    },
+    {
+      "epoch": 0.133888,
+      "grad_norm": 0.6397411227226257,
+      "learning_rate": 0.0009719978814235322,
+      "loss": 3.0613,
+      "step": 1046
+    },
+    {
+      "epoch": 0.134016,
+      "grad_norm": 0.6949122548103333,
+      "learning_rate": 0.000971929436981889,
+      "loss": 3.061,
+      "step": 1047
+    },
+    {
+      "epoch": 0.134144,
+      "grad_norm": 0.6485607624053955,
+      "learning_rate": 0.0009718609114101045,
+      "loss": 2.9839,
+      "step": 1048
+    },
+    {
+      "epoch": 0.134272,
+      "grad_norm": 0.6722410917282104,
+      "learning_rate": 0.0009717923047199591,
+      "loss": 2.8982,
+      "step": 1049
+    },
+    {
+      "epoch": 0.1344,
+      "grad_norm": 0.6328402161598206,
+      "learning_rate": 0.0009717236169232468,
+      "loss": 3.0648,
+      "step": 1050
+    },
+    {
+      "epoch": 0.134528,
+      "grad_norm": 0.6418887376785278,
+      "learning_rate": 0.0009716548480317763,
+      "loss": 2.8986,
+      "step": 1051
+    },
+    {
+      "epoch": 0.134656,
+      "grad_norm": 0.7011194825172424,
+      "learning_rate": 0.0009715859980573694,
+      "loss": 3.0393,
+      "step": 1052
+    },
+    {
+      "epoch": 0.134784,
+      "grad_norm": 0.6392652988433838,
+      "learning_rate": 0.0009715170670118624,
+      "loss": 2.8914,
+      "step": 1053
+    },
+    {
+      "epoch": 0.134912,
+      "grad_norm": 0.6300926208496094,
+      "learning_rate": 0.0009714480549071052,
+      "loss": 3.0182,
+      "step": 1054
+    },
+    {
+      "epoch": 0.13504,
+      "grad_norm": 0.6420726180076599,
+      "learning_rate": 0.000971378961754962,
+      "loss": 2.989,
+      "step": 1055
+    },
+    {
+      "epoch": 0.135168,
+      "grad_norm": 0.634863555431366,
+      "learning_rate": 0.0009713097875673105,
+      "loss": 2.9334,
+      "step": 1056
+    },
+    {
+      "epoch": 0.135296,
+      "grad_norm": 0.6257927417755127,
+      "learning_rate": 0.0009712405323560427,
+      "loss": 2.9791,
+      "step": 1057
+    },
+    {
+      "epoch": 0.135424,
+      "grad_norm": 0.6221749782562256,
+      "learning_rate": 0.0009711711961330643,
+      "loss": 2.972,
+      "step": 1058
+    },
+    {
+      "epoch": 0.135552,
+      "grad_norm": 0.6771379709243774,
+      "learning_rate": 0.0009711017789102948,
+      "loss": 3.1148,
+      "step": 1059
+    },
+    {
+      "epoch": 0.13568,
+      "grad_norm": 0.6548649668693542,
+      "learning_rate": 0.0009710322806996682,
+      "loss": 3.1112,
+      "step": 1060
+    },
+    {
+      "epoch": 0.135808,
+      "grad_norm": 0.692046582698822,
+      "learning_rate": 0.0009709627015131319,
+      "loss": 3.0789,
+      "step": 1061
+    },
+    {
+      "epoch": 0.135936,
+      "grad_norm": 0.7310773134231567,
+      "learning_rate": 0.0009708930413626473,
+      "loss": 3.0114,
+      "step": 1062
+    },
+    {
+      "epoch": 0.136064,
+      "grad_norm": 0.633556604385376,
+      "learning_rate": 0.0009708233002601898,
+      "loss": 2.8835,
+      "step": 1063
+    },
+    {
+      "epoch": 0.136192,
+      "grad_norm": 0.6581055521965027,
+      "learning_rate": 0.0009707534782177487,
+      "loss": 3.0116,
+      "step": 1064
+    },
+    {
+      "epoch": 0.13632,
+      "grad_norm": 0.6508828997612,
+      "learning_rate": 0.0009706835752473273,
+      "loss": 3.1069,
+      "step": 1065
+    },
+    {
+      "epoch": 0.136448,
+      "grad_norm": 0.6214575171470642,
+      "learning_rate": 0.0009706135913609426,
+      "loss": 2.9621,
+      "step": 1066
+    },
+    {
+      "epoch": 0.136576,
+      "grad_norm": 0.5992044806480408,
+      "learning_rate": 0.0009705435265706258,
+      "loss": 2.9071,
+      "step": 1067
+    },
+    {
+      "epoch": 0.136704,
+      "grad_norm": 0.6063728928565979,
+      "learning_rate": 0.0009704733808884218,
+      "loss": 2.9886,
+      "step": 1068
+    },
+    {
+      "epoch": 0.136832,
+      "grad_norm": 0.6093555092811584,
+      "learning_rate": 0.0009704031543263893,
+      "loss": 2.908,
+      "step": 1069
+    },
+    {
+      "epoch": 0.13696,
+      "grad_norm": 0.5946210026741028,
+      "learning_rate": 0.0009703328468966015,
+      "loss": 3.047,
+      "step": 1070
+    },
+    {
+      "epoch": 0.137088,
+      "grad_norm": 0.6072921752929688,
+      "learning_rate": 0.0009702624586111445,
+      "loss": 2.9833,
+      "step": 1071
+    },
+    {
+      "epoch": 0.137216,
+      "grad_norm": 0.7103402614593506,
+      "learning_rate": 0.0009701919894821192,
+      "loss": 3.0713,
+      "step": 1072
+    },
+    {
+      "epoch": 0.137344,
+      "grad_norm": 0.6476485729217529,
+      "learning_rate": 0.0009701214395216399,
+      "loss": 3.0404,
+      "step": 1073
+    },
+    {
+      "epoch": 0.137472,
+      "grad_norm": 0.6150424480438232,
+      "learning_rate": 0.0009700508087418351,
+      "loss": 3.0179,
+      "step": 1074
+    },
+    {
+      "epoch": 0.1376,
+      "grad_norm": 0.6502296924591064,
+      "learning_rate": 0.0009699800971548469,
+      "loss": 3.0137,
+      "step": 1075
+    },
+    {
+      "epoch": 0.137728,
+      "grad_norm": 0.6373313069343567,
+      "learning_rate": 0.0009699093047728316,
+      "loss": 3.06,
+      "step": 1076
+    },
+    {
+      "epoch": 0.137856,
+      "grad_norm": 0.6414070129394531,
+      "learning_rate": 0.0009698384316079591,
+      "loss": 3.0184,
+      "step": 1077
+    },
+    {
+      "epoch": 0.137984,
+      "grad_norm": 0.5951650142669678,
+      "learning_rate": 0.0009697674776724135,
+      "loss": 3.0334,
+      "step": 1078
+    },
+    {
+      "epoch": 0.138112,
+      "grad_norm": 0.6243633031845093,
+      "learning_rate": 0.0009696964429783923,
+      "loss": 2.9272,
+      "step": 1079
+    },
+    {
+      "epoch": 0.13824,
+      "grad_norm": 0.6515641808509827,
+      "learning_rate": 0.0009696253275381074,
+      "loss": 2.9502,
+      "step": 1080
+    },
+    {
+      "epoch": 0.138368,
+      "grad_norm": 0.6037840247154236,
+      "learning_rate": 0.0009695541313637843,
+      "loss": 3.0282,
+      "step": 1081
+    },
+    {
+      "epoch": 0.138496,
+      "grad_norm": 0.6015642881393433,
+      "learning_rate": 0.0009694828544676626,
+      "loss": 3.0485,
+      "step": 1082
+    },
+    {
+      "epoch": 0.138624,
+      "grad_norm": 0.6054241061210632,
+      "learning_rate": 0.0009694114968619954,
+      "loss": 2.9373,
+      "step": 1083
+    },
+    {
+      "epoch": 0.138752,
+      "grad_norm": 0.6412133574485779,
+      "learning_rate": 0.0009693400585590501,
+      "loss": 2.8795,
+      "step": 1084
+    },
+    {
+      "epoch": 0.13888,
+      "grad_norm": 0.5880333185195923,
+      "learning_rate": 0.0009692685395711076,
+      "loss": 2.8473,
+      "step": 1085
+    },
+    {
+      "epoch": 0.139008,
+      "grad_norm": 0.6330735683441162,
+      "learning_rate": 0.000969196939910463,
+      "loss": 3.0045,
+      "step": 1086
+    },
+    {
+      "epoch": 0.139136,
+      "grad_norm": 0.6261857748031616,
+      "learning_rate": 0.000969125259589425,
+      "loss": 2.757,
+      "step": 1087
+    },
+    {
+      "epoch": 0.139264,
+      "grad_norm": 0.6684296727180481,
+      "learning_rate": 0.0009690534986203164,
+      "loss": 2.947,
+      "step": 1088
+    },
+    {
+      "epoch": 0.139392,
+      "grad_norm": 0.6041046380996704,
+      "learning_rate": 0.0009689816570154734,
+      "loss": 2.9651,
+      "step": 1089
+    },
+    {
+      "epoch": 0.13952,
+      "grad_norm": 0.5682059526443481,
+      "learning_rate": 0.0009689097347872469,
+      "loss": 2.8784,
+      "step": 1090
+    },
+    {
+      "epoch": 0.139648,
+      "grad_norm": 0.639103353023529,
+      "learning_rate": 0.0009688377319480009,
+      "loss": 2.9208,
+      "step": 1091
+    },
+    {
+      "epoch": 0.139776,
+      "grad_norm": 0.6057471036911011,
+      "learning_rate": 0.0009687656485101132,
+      "loss": 2.9835,
+      "step": 1092
+    },
+    {
+      "epoch": 0.139904,
+      "grad_norm": 0.6550599336624146,
+      "learning_rate": 0.0009686934844859765,
+      "loss": 3.0523,
+      "step": 1093
+    },
+    {
+      "epoch": 0.140032,
+      "grad_norm": 0.6509894728660583,
+      "learning_rate": 0.0009686212398879962,
+      "loss": 3.1677,
+      "step": 1094
+    },
+    {
+      "epoch": 0.14016,
+      "grad_norm": 0.6307494044303894,
+      "learning_rate": 0.000968548914728592,
+      "loss": 3.0702,
+      "step": 1095
+    },
+    {
+      "epoch": 0.140288,
+      "grad_norm": 0.618074357509613,
+      "learning_rate": 0.0009684765090201974,
+      "loss": 2.9448,
+      "step": 1096
+    },
+    {
+      "epoch": 0.140416,
+      "grad_norm": 0.6310997605323792,
+      "learning_rate": 0.0009684040227752599,
+      "loss": 3.0442,
+      "step": 1097
+    },
+    {
+      "epoch": 0.140544,
+      "grad_norm": 0.588813841342926,
+      "learning_rate": 0.0009683314560062408,
+      "loss": 2.7897,
+      "step": 1098
+    },
+    {
+      "epoch": 0.140672,
+      "grad_norm": 0.5723446607589722,
+      "learning_rate": 0.0009682588087256148,
+      "loss": 3.0228,
+      "step": 1099
+    },
+    {
+      "epoch": 0.1408,
+      "grad_norm": 0.6476001143455505,
+      "learning_rate": 0.0009681860809458713,
+      "loss": 2.9343,
+      "step": 1100
+    },
+    {
+      "epoch": 0.140928,
+      "grad_norm": 0.6207176446914673,
+      "learning_rate": 0.0009681132726795128,
+      "loss": 3.077,
+      "step": 1101
+    },
+    {
+      "epoch": 0.141056,
+      "grad_norm": 0.7014726996421814,
+      "learning_rate": 0.0009680403839390557,
+      "loss": 2.9082,
+      "step": 1102
+    },
+    {
+      "epoch": 0.141184,
+      "grad_norm": 0.6927246451377869,
+      "learning_rate": 0.0009679674147370308,
+      "loss": 3.0071,
+      "step": 1103
+    },
+    {
+      "epoch": 0.141312,
+      "grad_norm": 0.6302134394645691,
+      "learning_rate": 0.0009678943650859819,
+      "loss": 2.9382,
+      "step": 1104
+    },
+    {
+      "epoch": 0.14144,
+      "grad_norm": 0.5867714881896973,
+      "learning_rate": 0.0009678212349984676,
+      "loss": 2.8748,
+      "step": 1105
+    },
+    {
+      "epoch": 0.141568,
+      "grad_norm": 0.6058904528617859,
+      "learning_rate": 0.0009677480244870596,
+      "loss": 2.9764,
+      "step": 1106
+    },
+    {
+      "epoch": 0.141696,
+      "grad_norm": 0.6179157495498657,
+      "learning_rate": 0.0009676747335643435,
+      "loss": 3.0077,
+      "step": 1107
+    },
+    {
+      "epoch": 0.141824,
+      "grad_norm": 0.6128743290901184,
+      "learning_rate": 0.0009676013622429187,
+      "loss": 2.9789,
+      "step": 1108
+    },
+    {
+      "epoch": 0.141952,
+      "grad_norm": 0.6552680730819702,
+      "learning_rate": 0.0009675279105353992,
+      "loss": 2.9663,
+      "step": 1109
+    },
+    {
+      "epoch": 0.14208,
+      "grad_norm": 0.5670120716094971,
+      "learning_rate": 0.0009674543784544117,
+      "loss": 2.9957,
+      "step": 1110
+    },
+    {
+      "epoch": 0.142208,
+      "grad_norm": 0.6377847194671631,
+      "learning_rate": 0.0009673807660125973,
+      "loss": 2.9523,
+      "step": 1111
+    },
+    {
+      "epoch": 0.142336,
+      "grad_norm": 0.6481389403343201,
+      "learning_rate": 0.0009673070732226108,
+      "loss": 2.8959,
+      "step": 1112
+    },
+    {
+      "epoch": 0.142464,
+      "grad_norm": 0.5892378091812134,
+      "learning_rate": 0.000967233300097121,
+      "loss": 2.9076,
+      "step": 1113
+    },
+    {
+      "epoch": 0.142592,
+      "grad_norm": 0.6507986783981323,
+      "learning_rate": 0.0009671594466488104,
+      "loss": 2.8901,
+      "step": 1114
+    },
+    {
+      "epoch": 0.14272,
+      "grad_norm": 0.6125216484069824,
+      "learning_rate": 0.000967085512890375,
+      "loss": 3.0185,
+      "step": 1115
+    },
+    {
+      "epoch": 0.142848,
+      "grad_norm": 0.5896008014678955,
+      "learning_rate": 0.000967011498834525,
+      "loss": 2.8372,
+      "step": 1116
+    },
+    {
+      "epoch": 0.142976,
+      "grad_norm": 0.6252095103263855,
+      "learning_rate": 0.0009669374044939845,
+      "loss": 2.9438,
+      "step": 1117
+    },
+    {
+      "epoch": 0.143104,
+      "grad_norm": 0.5920199155807495,
+      "learning_rate": 0.0009668632298814908,
+      "loss": 2.8432,
+      "step": 1118
+    },
+    {
+      "epoch": 0.143232,
+      "grad_norm": 0.6739546060562134,
+      "learning_rate": 0.0009667889750097956,
+      "loss": 2.9676,
+      "step": 1119
+    },
+    {
+      "epoch": 0.14336,
+      "grad_norm": 0.6336777210235596,
+      "learning_rate": 0.000966714639891664,
+      "loss": 3.053,
+      "step": 1120
+    },
+    {
+      "epoch": 0.143488,
+      "grad_norm": 0.6989814639091492,
+      "learning_rate": 0.0009666402245398754,
+      "loss": 3.0392,
+      "step": 1121
+    },
+    {
+      "epoch": 0.143616,
+      "grad_norm": 0.6583889722824097,
+      "learning_rate": 0.0009665657289672222,
+      "loss": 2.8697,
+      "step": 1122
+    },
+    {
+      "epoch": 0.143744,
+      "grad_norm": 0.5975735187530518,
+      "learning_rate": 0.0009664911531865114,
+      "loss": 3.0478,
+      "step": 1123
+    },
+    {
+      "epoch": 0.143872,
+      "grad_norm": 0.5998668670654297,
+      "learning_rate": 0.0009664164972105633,
+      "loss": 2.8733,
+      "step": 1124
+    },
+    {
+      "epoch": 0.144,
+      "grad_norm": 0.6030800938606262,
+      "learning_rate": 0.0009663417610522123,
+      "loss": 3.0715,
+      "step": 1125
+    },
+    {
+      "epoch": 0.144128,
+      "grad_norm": 0.5905584692955017,
+      "learning_rate": 0.0009662669447243061,
+      "loss": 2.963,
+      "step": 1126
+    },
+    {
+      "epoch": 0.144256,
+      "grad_norm": 0.6239269971847534,
+      "learning_rate": 0.0009661920482397068,
+      "loss": 2.9331,
+      "step": 1127
+    },
+    {
+      "epoch": 0.144384,
+      "grad_norm": 0.6499000191688538,
+      "learning_rate": 0.0009661170716112897,
+      "loss": 3.0175,
+      "step": 1128
+    },
+    {
+      "epoch": 0.144512,
+      "grad_norm": 0.5724586248397827,
+      "learning_rate": 0.0009660420148519444,
+      "loss": 3.0364,
+      "step": 1129
+    },
+    {
+      "epoch": 0.14464,
+      "grad_norm": 0.6048414707183838,
+      "learning_rate": 0.000965966877974574,
+      "loss": 2.9122,
+      "step": 1130
+    },
+    {
+      "epoch": 0.144768,
+      "grad_norm": 0.5595748424530029,
+      "learning_rate": 0.0009658916609920951,
+      "loss": 2.9795,
+      "step": 1131
+    },
+    {
+      "epoch": 0.144896,
+      "grad_norm": 0.5689665675163269,
+      "learning_rate": 0.0009658163639174389,
+      "loss": 2.999,
+      "step": 1132
+    },
+    {
+      "epoch": 0.145024,
+      "grad_norm": 0.5877953767776489,
+      "learning_rate": 0.0009657409867635492,
+      "loss": 2.8969,
+      "step": 1133
+    },
+    {
+      "epoch": 0.145152,
+      "grad_norm": 0.613582968711853,
+      "learning_rate": 0.0009656655295433848,
+      "loss": 3.066,
+      "step": 1134
+    },
+    {
+      "epoch": 0.14528,
+      "grad_norm": 0.670566976070404,
+      "learning_rate": 0.0009655899922699174,
+      "loss": 3.0737,
+      "step": 1135
+    },
+    {
+      "epoch": 0.145408,
+      "grad_norm": 0.625819742679596,
+      "learning_rate": 0.0009655143749561325,
+      "loss": 2.8998,
+      "step": 1136
+    },
+    {
+      "epoch": 0.145536,
+      "grad_norm": 0.6011262536048889,
+      "learning_rate": 0.0009654386776150302,
+      "loss": 2.9305,
+      "step": 1137
+    },
+    {
+      "epoch": 0.145664,
+      "grad_norm": 0.6002070903778076,
+      "learning_rate": 0.0009653629002596231,
+      "loss": 2.9111,
+      "step": 1138
+    },
+    {
+      "epoch": 0.145792,
+      "grad_norm": 0.6683277487754822,
+      "learning_rate": 0.0009652870429029386,
+      "loss": 2.958,
+      "step": 1139
+    },
+    {
+      "epoch": 0.14592,
+      "grad_norm": 0.6033244132995605,
+      "learning_rate": 0.0009652111055580174,
+      "loss": 3.0078,
+      "step": 1140
+    },
+    {
+      "epoch": 0.146048,
+      "grad_norm": 0.6155250072479248,
+      "learning_rate": 0.0009651350882379139,
+      "loss": 2.9026,
+      "step": 1141
+    },
+    {
+      "epoch": 0.146176,
+      "grad_norm": 0.6328603625297546,
+      "learning_rate": 0.0009650589909556964,
+      "loss": 3.1307,
+      "step": 1142
+    },
+    {
+      "epoch": 0.146304,
+      "grad_norm": 0.6020199656486511,
+      "learning_rate": 0.000964982813724447,
+      "loss": 2.9337,
+      "step": 1143
+    },
+    {
+      "epoch": 0.146432,
+      "grad_norm": 0.6337321400642395,
+      "learning_rate": 0.0009649065565572614,
+      "loss": 3.0488,
+      "step": 1144
+    },
+    {
+      "epoch": 0.14656,
+      "grad_norm": 0.6233793497085571,
+      "learning_rate": 0.0009648302194672489,
+      "loss": 2.9996,
+      "step": 1145
+    },
+    {
+      "epoch": 0.146688,
+      "grad_norm": 0.6500163078308105,
+      "learning_rate": 0.0009647538024675331,
+      "loss": 3.0071,
+      "step": 1146
+    },
+    {
+      "epoch": 0.146816,
+      "grad_norm": 0.5729029774665833,
+      "learning_rate": 0.0009646773055712506,
+      "loss": 2.9112,
+      "step": 1147
+    },
+    {
+      "epoch": 0.146944,
+      "grad_norm": 0.5561197996139526,
+      "learning_rate": 0.0009646007287915523,
+      "loss": 2.8922,
+      "step": 1148
+    },
+    {
+      "epoch": 0.147072,
+      "grad_norm": 0.6311776041984558,
+      "learning_rate": 0.0009645240721416028,
+      "loss": 2.9418,
+      "step": 1149
+    },
+    {
+      "epoch": 0.1472,
+      "grad_norm": 0.621874988079071,
+      "learning_rate": 0.00096444733563458,
+      "loss": 3.1875,
+      "step": 1150
+    },
+    {
+      "epoch": 0.147328,
+      "grad_norm": 0.6198594570159912,
+      "learning_rate": 0.0009643705192836758,
+      "loss": 3.0331,
+      "step": 1151
+    },
+    {
+      "epoch": 0.147456,
+      "grad_norm": 0.5904159545898438,
+      "learning_rate": 0.0009642936231020958,
+      "loss": 3.0822,
+      "step": 1152
+    },
+    {
+      "epoch": 0.147584,
+      "grad_norm": 0.5682401061058044,
+      "learning_rate": 0.0009642166471030596,
+      "loss": 2.9793,
+      "step": 1153
+    },
+    {
+      "epoch": 0.147712,
+      "grad_norm": 0.7354381680488586,
+      "learning_rate": 0.0009641395912998,
+      "loss": 3.0004,
+      "step": 1154
+    },
+    {
+      "epoch": 0.14784,
+      "grad_norm": 0.5955519676208496,
+      "learning_rate": 0.0009640624557055639,
+      "loss": 3.0244,
+      "step": 1155
+    },
+    {
+      "epoch": 0.147968,
+      "grad_norm": 0.5871164798736572,
+      "learning_rate": 0.0009639852403336118,
+      "loss": 3.0655,
+      "step": 1156
+    },
+    {
+      "epoch": 0.148096,
+      "grad_norm": 0.6155728101730347,
+      "learning_rate": 0.0009639079451972179,
+      "loss": 2.9833,
+      "step": 1157
+    },
+    {
+      "epoch": 0.148224,
+      "grad_norm": 0.6171334981918335,
+      "learning_rate": 0.0009638305703096701,
+      "loss": 3.043,
+      "step": 1158
+    },
+    {
+      "epoch": 0.148352,
+      "grad_norm": 0.6144945025444031,
+      "learning_rate": 0.0009637531156842702,
+      "loss": 2.9244,
+      "step": 1159
+    },
+    {
+      "epoch": 0.14848,
+      "grad_norm": 0.593903124332428,
+      "learning_rate": 0.0009636755813343334,
+      "loss": 3.0688,
+      "step": 1160
+    },
+    {
+      "epoch": 0.148608,
+      "grad_norm": 0.6018420457839966,
+      "learning_rate": 0.0009635979672731887,
+      "loss": 2.958,
+      "step": 1161
+    },
+    {
+      "epoch": 0.148736,
+      "grad_norm": 0.5881418585777283,
+      "learning_rate": 0.000963520273514179,
+      "loss": 2.9571,
+      "step": 1162
+    },
+    {
+      "epoch": 0.148864,
+      "grad_norm": 0.6241642832756042,
+      "learning_rate": 0.0009634425000706607,
+      "loss": 3.1093,
+      "step": 1163
+    },
+    {
+      "epoch": 0.148992,
+      "grad_norm": 0.612195611000061,
+      "learning_rate": 0.0009633646469560039,
+      "loss": 2.9336,
+      "step": 1164
+    },
+    {
+      "epoch": 0.14912,
+      "grad_norm": 0.6228492856025696,
+      "learning_rate": 0.0009632867141835926,
+      "loss": 2.9276,
+      "step": 1165
+    },
+    {
+      "epoch": 0.149248,
+      "grad_norm": 0.617311954498291,
+      "learning_rate": 0.000963208701766824,
+      "loss": 2.9726,
+      "step": 1166
+    },
+    {
+      "epoch": 0.149376,
+      "grad_norm": 0.5992480516433716,
+      "learning_rate": 0.0009631306097191099,
+      "loss": 2.8967,
+      "step": 1167
+    },
+    {
+      "epoch": 0.149504,
+      "grad_norm": 0.6234979629516602,
+      "learning_rate": 0.0009630524380538748,
+      "loss": 2.899,
+      "step": 1168
+    },
+    {
+      "epoch": 0.149632,
+      "grad_norm": 0.615425169467926,
+      "learning_rate": 0.0009629741867845573,
+      "loss": 3.0112,
+      "step": 1169
+    },
+    {
+      "epoch": 0.14976,
+      "grad_norm": 0.6082454919815063,
+      "learning_rate": 0.00096289585592461,
+      "loss": 2.9507,
+      "step": 1170
+    },
+    {
+      "epoch": 0.149888,
+      "grad_norm": 0.6156983375549316,
+      "learning_rate": 0.0009628174454874987,
+      "loss": 2.9857,
+      "step": 1171
+    },
+    {
+      "epoch": 0.150016,
+      "grad_norm": 0.6309418678283691,
+      "learning_rate": 0.0009627389554867031,
+      "loss": 3.0001,
+      "step": 1172
+    },
+    {
+      "epoch": 0.150144,
+      "grad_norm": 0.6456624269485474,
+      "learning_rate": 0.0009626603859357165,
+      "loss": 3.087,
+      "step": 1173
+    },
+    {
+      "epoch": 0.150272,
+      "grad_norm": 0.6040751338005066,
+      "learning_rate": 0.0009625817368480458,
+      "loss": 2.8509,
+      "step": 1174
+    },
+    {
+      "epoch": 0.1504,
+      "grad_norm": 0.6404918432235718,
+      "learning_rate": 0.000962503008237212,
+      "loss": 3.0181,
+      "step": 1175
+    },
+    {
+      "epoch": 0.150528,
+      "grad_norm": 0.6155198216438293,
+      "learning_rate": 0.0009624242001167493,
+      "loss": 3.0501,
+      "step": 1176
+    },
+    {
+      "epoch": 0.150656,
+      "grad_norm": 0.6024815440177917,
+      "learning_rate": 0.0009623453125002056,
+      "loss": 2.8866,
+      "step": 1177
+    },
+    {
+      "epoch": 0.150784,
+      "grad_norm": 0.6305459141731262,
+      "learning_rate": 0.0009622663454011427,
+      "loss": 3.0546,
+      "step": 1178
+    },
+    {
+      "epoch": 0.150912,
+      "grad_norm": 0.6155794262886047,
+      "learning_rate": 0.0009621872988331362,
+      "loss": 3.0372,
+      "step": 1179
+    },
+    {
+      "epoch": 0.15104,
+      "grad_norm": 0.6763031482696533,
+      "learning_rate": 0.0009621081728097747,
+      "loss": 2.8139,
+      "step": 1180
+    },
+    {
+      "epoch": 0.151168,
+      "grad_norm": 0.6106659173965454,
+      "learning_rate": 0.000962028967344661,
+      "loss": 2.9453,
+      "step": 1181
+    },
+    {
+      "epoch": 0.151296,
+      "grad_norm": 0.5846068263053894,
+      "learning_rate": 0.0009619496824514117,
+      "loss": 2.7877,
+      "step": 1182
+    },
+    {
+      "epoch": 0.151424,
+      "grad_norm": 0.5686424970626831,
+      "learning_rate": 0.0009618703181436566,
+      "loss": 3.0332,
+      "step": 1183
+    },
+    {
+      "epoch": 0.151552,
+      "grad_norm": 0.5943578481674194,
+      "learning_rate": 0.0009617908744350393,
+      "loss": 2.9441,
+      "step": 1184
+    },
+    {
+      "epoch": 0.15168,
+      "grad_norm": 0.657890796661377,
+      "learning_rate": 0.000961711351339217,
+      "loss": 3.0889,
+      "step": 1185
+    },
+    {
+      "epoch": 0.151808,
+      "grad_norm": 0.6085147857666016,
+      "learning_rate": 0.0009616317488698609,
+      "loss": 2.9484,
+      "step": 1186
+    },
+    {
+      "epoch": 0.151936,
+      "grad_norm": 0.606269121170044,
+      "learning_rate": 0.0009615520670406554,
+      "loss": 2.8945,
+      "step": 1187
+    },
+    {
+      "epoch": 0.152064,
+      "grad_norm": 0.5822179317474365,
+      "learning_rate": 0.000961472305865299,
+      "loss": 2.8954,
+      "step": 1188
+    },
+    {
+      "epoch": 0.152192,
+      "grad_norm": 0.6167764663696289,
+      "learning_rate": 0.0009613924653575032,
+      "loss": 2.9296,
+      "step": 1189
+    },
+    {
+      "epoch": 0.15232,
+      "grad_norm": 0.5957857370376587,
+      "learning_rate": 0.0009613125455309939,
+      "loss": 2.9719,
+      "step": 1190
+    },
+    {
+      "epoch": 0.152448,
+      "grad_norm": 0.6384730339050293,
+      "learning_rate": 0.0009612325463995099,
+      "loss": 3.025,
+      "step": 1191
+    },
+    {
+      "epoch": 0.152576,
+      "grad_norm": 0.6139668226242065,
+      "learning_rate": 0.000961152467976804,
+      "loss": 2.9532,
+      "step": 1192
+    },
+    {
+      "epoch": 0.152704,
+      "grad_norm": 0.666888952255249,
+      "learning_rate": 0.0009610723102766428,
+      "loss": 2.94,
+      "step": 1193
+    },
+    {
+      "epoch": 0.152832,
+      "grad_norm": 0.650660514831543,
+      "learning_rate": 0.0009609920733128064,
+      "loss": 2.9215,
+      "step": 1194
+    },
+    {
+      "epoch": 0.15296,
+      "grad_norm": 0.6266156435012817,
+      "learning_rate": 0.000960911757099088,
+      "loss": 2.8924,
+      "step": 1195
+    },
+    {
+      "epoch": 0.153088,
+      "grad_norm": 0.6209748387336731,
+      "learning_rate": 0.0009608313616492955,
+      "loss": 2.9075,
+      "step": 1196
+    },
+    {
+      "epoch": 0.153216,
+      "grad_norm": 0.5719965100288391,
+      "learning_rate": 0.0009607508869772495,
+      "loss": 2.8316,
+      "step": 1197
+    },
+    {
+      "epoch": 0.153344,
+      "grad_norm": 0.5614538192749023,
+      "learning_rate": 0.0009606703330967843,
+      "loss": 3.0841,
+      "step": 1198
+    },
+    {
+      "epoch": 0.153472,
+      "grad_norm": 0.5548755526542664,
+      "learning_rate": 0.0009605897000217485,
+      "loss": 2.883,
+      "step": 1199
+    },
+    {
+      "epoch": 0.1536,
+      "grad_norm": 0.5697600841522217,
+      "learning_rate": 0.0009605089877660036,
+      "loss": 3.1269,
+      "step": 1200
+    },
+    {
+      "epoch": 0.153728,
+      "grad_norm": 0.5912089943885803,
+      "learning_rate": 0.000960428196343425,
+      "loss": 3.0878,
+      "step": 1201
+    },
+    {
+      "epoch": 0.153856,
+      "grad_norm": 0.6009547710418701,
+      "learning_rate": 0.0009603473257679018,
+      "loss": 2.9242,
+      "step": 1202
+    },
+    {
+      "epoch": 0.153984,
+      "grad_norm": 0.6203259825706482,
+      "learning_rate": 0.0009602663760533363,
+      "loss": 2.9652,
+      "step": 1203
+    },
+    {
+      "epoch": 0.154112,
+      "grad_norm": 0.6111510396003723,
+      "learning_rate": 0.000960185347213645,
+      "loss": 3.0452,
+      "step": 1204
+    },
+    {
+      "epoch": 0.15424,
+      "grad_norm": 0.6303266286849976,
+      "learning_rate": 0.0009601042392627576,
+      "loss": 2.8333,
+      "step": 1205
+    },
+    {
+      "epoch": 0.154368,
+      "grad_norm": 0.6153172254562378,
+      "learning_rate": 0.0009600230522146176,
+      "loss": 2.8794,
+      "step": 1206
+    },
+    {
+      "epoch": 0.154496,
+      "grad_norm": 0.6007001399993896,
+      "learning_rate": 0.0009599417860831817,
+      "loss": 2.9474,
+      "step": 1207
+    },
+    {
+      "epoch": 0.154624,
+      "grad_norm": 0.6471025347709656,
+      "learning_rate": 0.0009598604408824207,
+      "loss": 2.962,
+      "step": 1208
+    },
+    {
+      "epoch": 0.154752,
+      "grad_norm": 0.6017709970474243,
+      "learning_rate": 0.0009597790166263189,
+      "loss": 3.0177,
+      "step": 1209
+    },
+    {
+      "epoch": 0.15488,
+      "grad_norm": 0.5858443379402161,
+      "learning_rate": 0.0009596975133288738,
+      "loss": 3.0199,
+      "step": 1210
+    },
+    {
+      "epoch": 0.155008,
+      "grad_norm": 0.5620368123054504,
+      "learning_rate": 0.0009596159310040969,
+      "loss": 2.9122,
+      "step": 1211
+    },
+    {
+      "epoch": 0.155136,
+      "grad_norm": 0.6088341474533081,
+      "learning_rate": 0.0009595342696660132,
+      "loss": 2.9836,
+      "step": 1212
+    },
+    {
+      "epoch": 0.155264,
+      "grad_norm": 0.6076828241348267,
+      "learning_rate": 0.000959452529328661,
+      "loss": 2.9879,
+      "step": 1213
+    },
+    {
+      "epoch": 0.155392,
+      "grad_norm": 0.5557029843330383,
+      "learning_rate": 0.0009593707100060926,
+      "loss": 2.8445,
+      "step": 1214
+    },
+    {
+      "epoch": 0.15552,
+      "grad_norm": 0.6402243375778198,
+      "learning_rate": 0.000959288811712374,
+      "loss": 2.9956,
+      "step": 1215
+    },
+    {
+      "epoch": 0.155648,
+      "grad_norm": 0.594367504119873,
+      "learning_rate": 0.0009592068344615837,
+      "loss": 2.9477,
+      "step": 1216
+    },
+    {
+      "epoch": 0.155776,
+      "grad_norm": 0.6257947683334351,
+      "learning_rate": 0.0009591247782678152,
+      "loss": 2.9394,
+      "step": 1217
+    },
+    {
+      "epoch": 0.155904,
+      "grad_norm": 0.6219844222068787,
+      "learning_rate": 0.0009590426431451746,
+      "loss": 3.0219,
+      "step": 1218
+    },
+    {
+      "epoch": 0.156032,
+      "grad_norm": 0.640549898147583,
+      "learning_rate": 0.000958960429107782,
+      "loss": 3.0277,
+      "step": 1219
+    },
+    {
+      "epoch": 0.15616,
+      "grad_norm": 0.623816728591919,
+      "learning_rate": 0.0009588781361697709,
+      "loss": 2.8579,
+      "step": 1220
+    },
+    {
+      "epoch": 0.156288,
+      "grad_norm": 0.6141059994697571,
+      "learning_rate": 0.0009587957643452885,
+      "loss": 2.9432,
+      "step": 1221
+    },
+    {
+      "epoch": 0.156416,
+      "grad_norm": 0.5777742266654968,
+      "learning_rate": 0.0009587133136484953,
+      "loss": 3.0006,
+      "step": 1222
+    },
+    {
+      "epoch": 0.156544,
+      "grad_norm": 0.6087720990180969,
+      "learning_rate": 0.0009586307840935657,
+      "loss": 2.9477,
+      "step": 1223
+    },
+    {
+      "epoch": 0.156672,
+      "grad_norm": 0.6071696877479553,
+      "learning_rate": 0.0009585481756946874,
+      "loss": 3.0299,
+      "step": 1224
+    },
+    {
+      "epoch": 0.1568,
+      "grad_norm": 0.5469344854354858,
+      "learning_rate": 0.0009584654884660618,
+      "loss": 2.8705,
+      "step": 1225
+    },
+    {
+      "epoch": 0.156928,
+      "grad_norm": 0.6058060526847839,
+      "learning_rate": 0.0009583827224219035,
+      "loss": 2.9487,
+      "step": 1226
+    },
+    {
+      "epoch": 0.157056,
+      "grad_norm": 0.5557302832603455,
+      "learning_rate": 0.0009582998775764413,
+      "loss": 3.0694,
+      "step": 1227
+    },
+    {
+      "epoch": 0.157184,
+      "grad_norm": 0.5437577962875366,
+      "learning_rate": 0.0009582169539439171,
+      "loss": 2.9614,
+      "step": 1228
+    },
+    {
+      "epoch": 0.157312,
+      "grad_norm": 0.583004355430603,
+      "learning_rate": 0.0009581339515385865,
+      "loss": 3.0463,
+      "step": 1229
+    },
+    {
+      "epoch": 0.15744,
+      "grad_norm": 0.573049008846283,
+      "learning_rate": 0.0009580508703747184,
+      "loss": 2.9581,
+      "step": 1230
+    },
+    {
+      "epoch": 0.157568,
+      "grad_norm": 0.5868643522262573,
+      "learning_rate": 0.0009579677104665956,
+      "loss": 3.0051,
+      "step": 1231
+    },
+    {
+      "epoch": 0.157696,
+      "grad_norm": 0.6016791462898254,
+      "learning_rate": 0.000957884471828514,
+      "loss": 2.9293,
+      "step": 1232
+    },
+    {
+      "epoch": 0.157824,
+      "grad_norm": 0.6156187057495117,
+      "learning_rate": 0.0009578011544747835,
+      "loss": 3.049,
+      "step": 1233
+    },
+    {
+      "epoch": 0.157952,
+      "grad_norm": 0.5872225761413574,
+      "learning_rate": 0.0009577177584197274,
+      "loss": 2.9969,
+      "step": 1234
+    },
+    {
+      "epoch": 0.15808,
+      "grad_norm": 0.614916205406189,
+      "learning_rate": 0.0009576342836776821,
+      "loss": 3.0432,
+      "step": 1235
+    },
+    {
+      "epoch": 0.158208,
+      "grad_norm": 0.6493079662322998,
+      "learning_rate": 0.0009575507302629982,
+      "loss": 3.0896,
+      "step": 1236
+    },
+    {
+      "epoch": 0.158336,
+      "grad_norm": 0.5872235298156738,
+      "learning_rate": 0.0009574670981900393,
+      "loss": 2.9212,
+      "step": 1237
+    },
+    {
+      "epoch": 0.158464,
+      "grad_norm": 0.605522632598877,
+      "learning_rate": 0.0009573833874731829,
+      "loss": 2.8771,
+      "step": 1238
+    },
+    {
+      "epoch": 0.158592,
+      "grad_norm": 0.5819229483604431,
+      "learning_rate": 0.0009572995981268198,
+      "loss": 3.067,
+      "step": 1239
+    },
+    {
+      "epoch": 0.15872,
+      "grad_norm": 0.6610497832298279,
+      "learning_rate": 0.0009572157301653542,
+      "loss": 3.03,
+      "step": 1240
+    },
+    {
+      "epoch": 0.158848,
+      "grad_norm": 0.5964094996452332,
+      "learning_rate": 0.0009571317836032041,
+      "loss": 2.9212,
+      "step": 1241
+    },
+    {
+      "epoch": 0.158976,
+      "grad_norm": 0.5858182907104492,
+      "learning_rate": 0.0009570477584548007,
+      "loss": 2.8688,
+      "step": 1242
+    },
+    {
+      "epoch": 0.159104,
+      "grad_norm": 0.6101973652839661,
+      "learning_rate": 0.0009569636547345894,
+      "loss": 2.9212,
+      "step": 1243
+    },
+    {
+      "epoch": 0.159232,
+      "grad_norm": 0.5965656042098999,
+      "learning_rate": 0.0009568794724570281,
+      "loss": 2.9467,
+      "step": 1244
+    },
+    {
+      "epoch": 0.15936,
+      "grad_norm": 0.6532838940620422,
+      "learning_rate": 0.0009567952116365889,
+      "loss": 2.9304,
+      "step": 1245
+    },
+    {
+      "epoch": 0.159488,
+      "grad_norm": 0.6184207797050476,
+      "learning_rate": 0.0009567108722877572,
+      "loss": 3.0394,
+      "step": 1246
+    },
+    {
+      "epoch": 0.159616,
+      "grad_norm": 0.6537927985191345,
+      "learning_rate": 0.0009566264544250318,
+      "loss": 3.0339,
+      "step": 1247
+    },
+    {
+      "epoch": 0.159744,
+      "grad_norm": 0.5770497918128967,
+      "learning_rate": 0.0009565419580629253,
+      "loss": 2.9034,
+      "step": 1248
+    },
+    {
+      "epoch": 0.159872,
+      "grad_norm": 0.606183648109436,
+      "learning_rate": 0.0009564573832159637,
+      "loss": 2.9835,
+      "step": 1249
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5639417171478271,
+      "learning_rate": 0.0009563727298986859,
+      "loss": 3.0488,
+      "step": 1250
+    },
+    {
+      "epoch": 0.160128,
+      "grad_norm": 0.5924633741378784,
+      "learning_rate": 0.0009562879981256454,
+      "loss": 3.0215,
+      "step": 1251
+    },
+    {
+      "epoch": 0.160256,
+      "grad_norm": 0.5760909914970398,
+      "learning_rate": 0.0009562031879114081,
+      "loss": 3.0286,
+      "step": 1252
+    },
+    {
+      "epoch": 0.160384,
+      "grad_norm": 0.5676677823066711,
+      "learning_rate": 0.0009561182992705541,
+      "loss": 2.9334,
+      "step": 1253
+    },
+    {
+      "epoch": 0.160512,
+      "grad_norm": 0.5835487842559814,
+      "learning_rate": 0.0009560333322176767,
+      "loss": 2.8197,
+      "step": 1254
+    },
+    {
+      "epoch": 0.16064,
+      "grad_norm": 0.6029694080352783,
+      "learning_rate": 0.0009559482867673825,
+      "loss": 3.0028,
+      "step": 1255
+    },
+    {
+      "epoch": 0.160768,
+      "grad_norm": 0.6231985092163086,
+      "learning_rate": 0.0009558631629342922,
+      "loss": 3.0731,
+      "step": 1256
+    },
+    {
+      "epoch": 0.160896,
+      "grad_norm": 0.6275836229324341,
+      "learning_rate": 0.0009557779607330393,
+      "loss": 3.0768,
+      "step": 1257
+    },
+    {
+      "epoch": 0.161024,
+      "grad_norm": 0.5926213264465332,
+      "learning_rate": 0.0009556926801782713,
+      "loss": 3.0374,
+      "step": 1258
+    },
+    {
+      "epoch": 0.161152,
+      "grad_norm": 0.591346263885498,
+      "learning_rate": 0.0009556073212846485,
+      "loss": 2.8727,
+      "step": 1259
+    },
+    {
+      "epoch": 0.16128,
+      "grad_norm": 0.6016525626182556,
+      "learning_rate": 0.0009555218840668454,
+      "loss": 2.887,
+      "step": 1260
+    },
+    {
+      "epoch": 0.161408,
+      "grad_norm": 0.5752759575843811,
+      "learning_rate": 0.0009554363685395495,
+      "loss": 2.9285,
+      "step": 1261
+    },
+    {
+      "epoch": 0.161536,
+      "grad_norm": 0.5657200217247009,
+      "learning_rate": 0.0009553507747174622,
+      "loss": 3.0227,
+      "step": 1262
+    },
+    {
+      "epoch": 0.161664,
+      "grad_norm": 0.581520676612854,
+      "learning_rate": 0.0009552651026152976,
+      "loss": 2.905,
+      "step": 1263
+    },
+    {
+      "epoch": 0.161792,
+      "grad_norm": 0.5485462546348572,
+      "learning_rate": 0.0009551793522477841,
+      "loss": 2.9429,
+      "step": 1264
+    },
+    {
+      "epoch": 0.16192,
+      "grad_norm": 0.561663806438446,
+      "learning_rate": 0.0009550935236296629,
+      "loss": 3.1076,
+      "step": 1265
+    },
+    {
+      "epoch": 0.162048,
+      "grad_norm": 0.5918262600898743,
+      "learning_rate": 0.0009550076167756892,
+      "loss": 2.954,
+      "step": 1266
+    },
+    {
+      "epoch": 0.162176,
+      "grad_norm": 0.5685549974441528,
+      "learning_rate": 0.0009549216317006313,
+      "loss": 2.9121,
+      "step": 1267
+    },
+    {
+      "epoch": 0.162304,
+      "grad_norm": 0.5591074824333191,
+      "learning_rate": 0.0009548355684192711,
+      "loss": 2.9539,
+      "step": 1268
+    },
+    {
+      "epoch": 0.162432,
+      "grad_norm": 0.5609291791915894,
+      "learning_rate": 0.0009547494269464036,
+      "loss": 3.1175,
+      "step": 1269
+    },
+    {
+      "epoch": 0.16256,
+      "grad_norm": 0.6031808257102966,
+      "learning_rate": 0.0009546632072968378,
+      "loss": 3.0615,
+      "step": 1270
+    },
+    {
+      "epoch": 0.162688,
+      "grad_norm": 0.6263890862464905,
+      "learning_rate": 0.0009545769094853958,
+      "loss": 3.0068,
+      "step": 1271
+    },
+    {
+      "epoch": 0.162816,
+      "grad_norm": 0.5580376982688904,
+      "learning_rate": 0.000954490533526913,
+      "loss": 2.9426,
+      "step": 1272
+    },
+    {
+      "epoch": 0.162944,
+      "grad_norm": 0.623832106590271,
+      "learning_rate": 0.0009544040794362388,
+      "loss": 3.0524,
+      "step": 1273
+    },
+    {
+      "epoch": 0.163072,
+      "grad_norm": 0.5840739011764526,
+      "learning_rate": 0.0009543175472282352,
+      "loss": 2.9695,
+      "step": 1274
+    },
+    {
+      "epoch": 0.1632,
+      "grad_norm": 0.6291914582252502,
+      "learning_rate": 0.0009542309369177784,
+      "loss": 3.0398,
+      "step": 1275
+    },
+    {
+      "epoch": 0.163328,
+      "grad_norm": 0.5482161641120911,
+      "learning_rate": 0.0009541442485197577,
+      "loss": 2.9473,
+      "step": 1276
+    },
+    {
+      "epoch": 0.163456,
+      "grad_norm": 0.5956851243972778,
+      "learning_rate": 0.0009540574820490757,
+      "loss": 2.7672,
+      "step": 1277
+    },
+    {
+      "epoch": 0.163584,
+      "grad_norm": 0.6388895511627197,
+      "learning_rate": 0.0009539706375206487,
+      "loss": 2.9471,
+      "step": 1278
+    },
+    {
+      "epoch": 0.163712,
+      "grad_norm": 0.6522027254104614,
+      "learning_rate": 0.0009538837149494064,
+      "loss": 2.9905,
+      "step": 1279
+    },
+    {
+      "epoch": 0.16384,
+      "grad_norm": 0.6101669073104858,
+      "learning_rate": 0.0009537967143502915,
+      "loss": 2.9371,
+      "step": 1280
+    },
+    {
+      "epoch": 0.163968,
+      "grad_norm": 0.6271051168441772,
+      "learning_rate": 0.0009537096357382604,
+      "loss": 3.046,
+      "step": 1281
+    },
+    {
+      "epoch": 0.164096,
+      "grad_norm": 0.7442458868026733,
+      "learning_rate": 0.0009536224791282833,
+      "loss": 2.968,
+      "step": 1282
+    },
+    {
+      "epoch": 0.164224,
+      "grad_norm": 0.6459609270095825,
+      "learning_rate": 0.000953535244535343,
+      "loss": 3.0394,
+      "step": 1283
+    },
+    {
+      "epoch": 0.164352,
+      "grad_norm": 0.5712761282920837,
+      "learning_rate": 0.0009534479319744365,
+      "loss": 3.0356,
+      "step": 1284
+    },
+    {
+      "epoch": 0.16448,
+      "grad_norm": 0.5749444365501404,
+      "learning_rate": 0.0009533605414605735,
+      "loss": 2.9715,
+      "step": 1285
+    },
+    {
+      "epoch": 0.164608,
+      "grad_norm": 0.5793547034263611,
+      "learning_rate": 0.0009532730730087778,
+      "loss": 2.9554,
+      "step": 1286
+    },
+    {
+      "epoch": 0.164736,
+      "grad_norm": 0.5837928056716919,
+      "learning_rate": 0.0009531855266340859,
+      "loss": 2.7503,
+      "step": 1287
+    },
+    {
+      "epoch": 0.164864,
+      "grad_norm": 0.6083358526229858,
+      "learning_rate": 0.0009530979023515483,
+      "loss": 2.9607,
+      "step": 1288
+    },
+    {
+      "epoch": 0.164992,
+      "grad_norm": 0.6265530586242676,
+      "learning_rate": 0.0009530102001762284,
+      "loss": 3.1082,
+      "step": 1289
+    },
+    {
+      "epoch": 0.16512,
+      "grad_norm": 0.5550141334533691,
+      "learning_rate": 0.0009529224201232034,
+      "loss": 3.0671,
+      "step": 1290
+    },
+    {
+      "epoch": 0.165248,
+      "grad_norm": 0.6140491962432861,
+      "learning_rate": 0.0009528345622075635,
+      "loss": 3.0207,
+      "step": 1291
+    },
+    {
+      "epoch": 0.165376,
+      "grad_norm": 0.6177128553390503,
+      "learning_rate": 0.000952746626444413,
+      "loss": 2.7937,
+      "step": 1292
+    },
+    {
+      "epoch": 0.165504,
+      "grad_norm": 0.5868253111839294,
+      "learning_rate": 0.0009526586128488685,
+      "loss": 3.1033,
+      "step": 1293
+    },
+    {
+      "epoch": 0.165632,
+      "grad_norm": 0.6476940512657166,
+      "learning_rate": 0.0009525705214360609,
+      "loss": 2.9978,
+      "step": 1294
+    },
+    {
+      "epoch": 0.16576,
+      "grad_norm": 0.6326400637626648,
+      "learning_rate": 0.0009524823522211339,
+      "loss": 2.9217,
+      "step": 1295
+    },
+    {
+      "epoch": 0.165888,
+      "grad_norm": 0.5758768320083618,
+      "learning_rate": 0.000952394105219245,
+      "loss": 3.0545,
+      "step": 1296
+    },
+    {
+      "epoch": 0.166016,
+      "grad_norm": 0.5573798418045044,
+      "learning_rate": 0.0009523057804455648,
+      "loss": 3.0304,
+      "step": 1297
+    },
+    {
+      "epoch": 0.166144,
+      "grad_norm": 0.5978743433952332,
+      "learning_rate": 0.0009522173779152773,
+      "loss": 2.9905,
+      "step": 1298
+    },
+    {
+      "epoch": 0.166272,
+      "grad_norm": 0.6142465472221375,
+      "learning_rate": 0.00095212889764358,
+      "loss": 3.1329,
+      "step": 1299
+    },
+    {
+      "epoch": 0.1664,
+      "grad_norm": 0.5844963192939758,
+      "learning_rate": 0.0009520403396456838,
+      "loss": 2.8291,
+      "step": 1300
+    },
+    {
+      "epoch": 0.166528,
+      "grad_norm": 0.6500221490859985,
+      "learning_rate": 0.0009519517039368126,
+      "loss": 3.0632,
+      "step": 1301
+    },
+    {
+      "epoch": 0.166656,
+      "grad_norm": 0.6658205389976501,
+      "learning_rate": 0.000951862990532204,
+      "loss": 2.9763,
+      "step": 1302
+    },
+    {
+      "epoch": 0.166784,
+      "grad_norm": 0.6000604629516602,
+      "learning_rate": 0.000951774199447109,
+      "loss": 2.9742,
+      "step": 1303
+    },
+    {
+      "epoch": 0.166912,
+      "grad_norm": 0.6095767617225647,
+      "learning_rate": 0.0009516853306967915,
+      "loss": 3.0094,
+      "step": 1304
+    },
+    {
+      "epoch": 0.16704,
+      "grad_norm": 0.5994210839271545,
+      "learning_rate": 0.0009515963842965294,
+      "loss": 2.9982,
+      "step": 1305
+    },
+    {
+      "epoch": 0.167168,
+      "grad_norm": 0.6205915212631226,
+      "learning_rate": 0.0009515073602616135,
+      "loss": 2.8625,
+      "step": 1306
+    },
+    {
+      "epoch": 0.167296,
+      "grad_norm": 0.5760748386383057,
+      "learning_rate": 0.000951418258607348,
+      "loss": 3.0991,
+      "step": 1307
+    },
+    {
+      "epoch": 0.167424,
+      "grad_norm": 0.5779820084571838,
+      "learning_rate": 0.0009513290793490505,
+      "loss": 2.9587,
+      "step": 1308
+    },
+    {
+      "epoch": 0.167552,
+      "grad_norm": 0.6242061853408813,
+      "learning_rate": 0.0009512398225020522,
+      "loss": 2.9349,
+      "step": 1309
+    },
+    {
+      "epoch": 0.16768,
+      "grad_norm": 0.6009894013404846,
+      "learning_rate": 0.000951150488081697,
+      "loss": 2.9698,
+      "step": 1310
+    },
+    {
+      "epoch": 0.167808,
+      "grad_norm": 0.5750643014907837,
+      "learning_rate": 0.0009510610761033427,
+      "loss": 2.9432,
+      "step": 1311
+    },
+    {
+      "epoch": 0.167936,
+      "grad_norm": 0.5858772397041321,
+      "learning_rate": 0.0009509715865823604,
+      "loss": 2.9146,
+      "step": 1312
+    },
+    {
+      "epoch": 0.168064,
+      "grad_norm": 0.5911533832550049,
+      "learning_rate": 0.0009508820195341342,
+      "loss": 2.88,
+      "step": 1313
+    },
+    {
+      "epoch": 0.168192,
+      "grad_norm": 0.5741074085235596,
+      "learning_rate": 0.0009507923749740618,
+      "loss": 2.8935,
+      "step": 1314
+    },
+    {
+      "epoch": 0.16832,
+      "grad_norm": 0.5498982071876526,
+      "learning_rate": 0.000950702652917554,
+      "loss": 2.8911,
+      "step": 1315
+    },
+    {
+      "epoch": 0.168448,
+      "grad_norm": 0.5669998526573181,
+      "learning_rate": 0.0009506128533800354,
+      "loss": 2.9608,
+      "step": 1316
+    },
+    {
+      "epoch": 0.168576,
+      "grad_norm": 0.5820481777191162,
+      "learning_rate": 0.0009505229763769431,
+      "loss": 2.889,
+      "step": 1317
+    },
+    {
+      "epoch": 0.168704,
+      "grad_norm": 0.6051968336105347,
+      "learning_rate": 0.0009504330219237283,
+      "loss": 2.8852,
+      "step": 1318
+    },
+    {
+      "epoch": 0.168832,
+      "grad_norm": 0.5505192279815674,
+      "learning_rate": 0.0009503429900358553,
+      "loss": 3.0426,
+      "step": 1319
+    },
+    {
+      "epoch": 0.16896,
+      "grad_norm": 0.5932660102844238,
+      "learning_rate": 0.0009502528807288014,
+      "loss": 2.9382,
+      "step": 1320
+    },
+    {
+      "epoch": 0.169088,
+      "grad_norm": 0.5918727517127991,
+      "learning_rate": 0.0009501626940180575,
+      "loss": 3.0011,
+      "step": 1321
+    },
+    {
+      "epoch": 0.169216,
+      "grad_norm": 0.571088969707489,
+      "learning_rate": 0.0009500724299191278,
+      "loss": 2.8444,
+      "step": 1322
+    },
+    {
+      "epoch": 0.169344,
+      "grad_norm": 0.5727904438972473,
+      "learning_rate": 0.0009499820884475296,
+      "loss": 2.8296,
+      "step": 1323
+    },
+    {
+      "epoch": 0.169472,
+      "grad_norm": 0.6826344132423401,
+      "learning_rate": 0.000949891669618794,
+      "loss": 2.969,
+      "step": 1324
+    },
+    {
+      "epoch": 0.1696,
+      "grad_norm": 0.6024475693702698,
+      "learning_rate": 0.0009498011734484646,
+      "loss": 3.0081,
+      "step": 1325
+    },
+    {
+      "epoch": 0.169728,
+      "grad_norm": 0.5945368409156799,
+      "learning_rate": 0.0009497105999520989,
+      "loss": 2.9732,
+      "step": 1326
+    },
+    {
+      "epoch": 0.169856,
+      "grad_norm": 0.5822340846061707,
+      "learning_rate": 0.0009496199491452679,
+      "loss": 2.9444,
+      "step": 1327
+    },
+    {
+      "epoch": 0.169984,
+      "grad_norm": 0.5683667659759521,
+      "learning_rate": 0.0009495292210435548,
+      "loss": 2.9423,
+      "step": 1328
+    },
+    {
+      "epoch": 0.170112,
+      "grad_norm": 0.5851700901985168,
+      "learning_rate": 0.0009494384156625575,
+      "loss": 3.0065,
+      "step": 1329
+    },
+    {
+      "epoch": 0.17024,
+      "grad_norm": 0.661036491394043,
+      "learning_rate": 0.000949347533017886,
+      "loss": 3.0816,
+      "step": 1330
+    },
+    {
+      "epoch": 0.170368,
+      "grad_norm": 0.6698899269104004,
+      "learning_rate": 0.0009492565731251644,
+      "loss": 3.1839,
+      "step": 1331
+    },
+    {
+      "epoch": 0.170496,
+      "grad_norm": 0.5690553188323975,
+      "learning_rate": 0.0009491655360000297,
+      "loss": 2.9499,
+      "step": 1332
+    },
+    {
+      "epoch": 0.170624,
+      "grad_norm": 0.6314891576766968,
+      "learning_rate": 0.0009490744216581322,
+      "loss": 2.9474,
+      "step": 1333
+    },
+    {
+      "epoch": 0.170752,
+      "grad_norm": 0.6237809658050537,
+      "learning_rate": 0.0009489832301151353,
+      "loss": 2.9886,
+      "step": 1334
+    },
+    {
+      "epoch": 0.17088,
+      "grad_norm": 0.5221965909004211,
+      "learning_rate": 0.0009488919613867162,
+      "loss": 2.8715,
+      "step": 1335
+    },
+    {
+      "epoch": 0.171008,
+      "grad_norm": 0.5470190048217773,
+      "learning_rate": 0.000948800615488565,
+      "loss": 2.981,
+      "step": 1336
+    },
+    {
+      "epoch": 0.171136,
+      "grad_norm": 0.601141095161438,
+      "learning_rate": 0.0009487091924363851,
+      "loss": 3.0036,
+      "step": 1337
+    },
+    {
+      "epoch": 0.171264,
+      "grad_norm": 0.6507043242454529,
+      "learning_rate": 0.0009486176922458929,
+      "loss": 2.8625,
+      "step": 1338
+    },
+    {
+      "epoch": 0.171392,
+      "grad_norm": 0.6538265347480774,
+      "learning_rate": 0.0009485261149328188,
+      "loss": 2.9015,
+      "step": 1339
+    },
+    {
+      "epoch": 0.17152,
+      "grad_norm": 0.6120591163635254,
+      "learning_rate": 0.0009484344605129056,
+      "loss": 2.8916,
+      "step": 1340
+    },
+    {
+      "epoch": 0.171648,
+      "grad_norm": 0.5865566730499268,
+      "learning_rate": 0.0009483427290019101,
+      "loss": 3.0674,
+      "step": 1341
+    },
+    {
+      "epoch": 0.171776,
+      "grad_norm": 0.5813671946525574,
+      "learning_rate": 0.0009482509204156018,
+      "loss": 2.8776,
+      "step": 1342
+    },
+    {
+      "epoch": 0.171904,
+      "grad_norm": 0.5532426834106445,
+      "learning_rate": 0.0009481590347697637,
+      "loss": 3.0473,
+      "step": 1343
+    },
+    {
+      "epoch": 0.172032,
+      "grad_norm": 0.6300868391990662,
+      "learning_rate": 0.0009480670720801921,
+      "loss": 2.8025,
+      "step": 1344
+    },
+    {
+      "epoch": 0.17216,
+      "grad_norm": 0.6718496680259705,
+      "learning_rate": 0.0009479750323626962,
+      "loss": 3.0212,
+      "step": 1345
+    },
+    {
+      "epoch": 0.172288,
+      "grad_norm": 0.6408380270004272,
+      "learning_rate": 0.000947882915633099,
+      "loss": 2.9894,
+      "step": 1346
+    },
+    {
+      "epoch": 0.172416,
+      "grad_norm": 0.6614367961883545,
+      "learning_rate": 0.0009477907219072361,
+      "loss": 3.2025,
+      "step": 1347
+    },
+    {
+      "epoch": 0.172544,
+      "grad_norm": 0.6120317578315735,
+      "learning_rate": 0.0009476984512009571,
+      "loss": 2.9496,
+      "step": 1348
+    },
+    {
+      "epoch": 0.172672,
+      "grad_norm": 0.5792599320411682,
+      "learning_rate": 0.0009476061035301241,
+      "loss": 2.9684,
+      "step": 1349
+    },
+    {
+      "epoch": 0.1728,
+      "grad_norm": 0.6294567584991455,
+      "learning_rate": 0.0009475136789106128,
+      "loss": 3.0714,
+      "step": 1350
+    },
+    {
+      "epoch": 0.172928,
+      "grad_norm": 0.5535847544670105,
+      "learning_rate": 0.0009474211773583121,
+      "loss": 2.8369,
+      "step": 1351
+    },
+    {
+      "epoch": 0.173056,
+      "grad_norm": 0.568334698677063,
+      "learning_rate": 0.000947328598889124,
+      "loss": 3.0976,
+      "step": 1352
+    },
+    {
+      "epoch": 0.173184,
+      "grad_norm": 0.6524762511253357,
+      "learning_rate": 0.000947235943518964,
+      "loss": 2.991,
+      "step": 1353
+    },
+    {
+      "epoch": 0.173312,
+      "grad_norm": 0.5786418318748474,
+      "learning_rate": 0.0009471432112637604,
+      "loss": 2.9118,
+      "step": 1354
+    },
+    {
+      "epoch": 0.17344,
+      "grad_norm": 0.6123725771903992,
+      "learning_rate": 0.000947050402139455,
+      "loss": 2.9963,
+      "step": 1355
+    },
+    {
+      "epoch": 0.173568,
+      "grad_norm": 0.5556647777557373,
+      "learning_rate": 0.0009469575161620029,
+      "loss": 2.9979,
+      "step": 1356
+    },
+    {
+      "epoch": 0.173696,
+      "grad_norm": 0.5829709768295288,
+      "learning_rate": 0.0009468645533473721,
+      "loss": 3.0037,
+      "step": 1357
+    },
+    {
+      "epoch": 0.173824,
+      "grad_norm": 0.6372557282447815,
+      "learning_rate": 0.0009467715137115442,
+      "loss": 3.0152,
+      "step": 1358
+    },
+    {
+      "epoch": 0.173952,
+      "grad_norm": 0.6376338601112366,
+      "learning_rate": 0.0009466783972705136,
+      "loss": 2.9063,
+      "step": 1359
+    },
+    {
+      "epoch": 0.17408,
+      "grad_norm": 0.6565738320350647,
+      "learning_rate": 0.0009465852040402883,
+      "loss": 3.0314,
+      "step": 1360
+    },
+    {
+      "epoch": 0.174208,
+      "grad_norm": 0.5980199575424194,
+      "learning_rate": 0.000946491934036889,
+      "loss": 2.9868,
+      "step": 1361
+    },
+    {
+      "epoch": 0.174336,
+      "grad_norm": 0.5884917378425598,
+      "learning_rate": 0.00094639858727635,
+      "loss": 2.9392,
+      "step": 1362
+    },
+    {
+      "epoch": 0.174464,
+      "grad_norm": 0.604090690612793,
+      "learning_rate": 0.0009463051637747191,
+      "loss": 2.9411,
+      "step": 1363
+    },
+    {
+      "epoch": 0.174592,
+      "grad_norm": 0.5877172946929932,
+      "learning_rate": 0.0009462116635480562,
+      "loss": 2.7586,
+      "step": 1364
+    },
+    {
+      "epoch": 0.17472,
+      "grad_norm": 0.5268228054046631,
+      "learning_rate": 0.0009461180866124356,
+      "loss": 2.9147,
+      "step": 1365
+    },
+    {
+      "epoch": 0.174848,
+      "grad_norm": 0.5975393652915955,
+      "learning_rate": 0.000946024432983944,
+      "loss": 3.1189,
+      "step": 1366
+    },
+    {
+      "epoch": 0.174976,
+      "grad_norm": 0.5890174508094788,
+      "learning_rate": 0.0009459307026786816,
+      "loss": 2.9538,
+      "step": 1367
+    },
+    {
+      "epoch": 0.175104,
+      "grad_norm": 0.5746464729309082,
+      "learning_rate": 0.0009458368957127617,
+      "loss": 3.0591,
+      "step": 1368
+    },
+    {
+      "epoch": 0.175232,
+      "grad_norm": 0.5904214978218079,
+      "learning_rate": 0.000945743012102311,
+      "loss": 2.931,
+      "step": 1369
+    },
+    {
+      "epoch": 0.17536,
+      "grad_norm": 0.5789846777915955,
+      "learning_rate": 0.0009456490518634688,
+      "loss": 2.8761,
+      "step": 1370
+    },
+    {
+      "epoch": 0.175488,
+      "grad_norm": 0.5404038429260254,
+      "learning_rate": 0.0009455550150123883,
+      "loss": 2.9458,
+      "step": 1371
+    },
+    {
+      "epoch": 0.175616,
+      "grad_norm": 0.5811572074890137,
+      "learning_rate": 0.0009454609015652353,
+      "loss": 2.9651,
+      "step": 1372
+    },
+    {
+      "epoch": 0.175744,
+      "grad_norm": 0.7197185158729553,
+      "learning_rate": 0.0009453667115381892,
+      "loss": 2.9896,
+      "step": 1373
+    },
+    {
+      "epoch": 0.175872,
+      "grad_norm": 0.5985124111175537,
+      "learning_rate": 0.0009452724449474422,
+      "loss": 2.9919,
+      "step": 1374
+    },
+    {
+      "epoch": 0.176,
+      "grad_norm": 0.5494886040687561,
+      "learning_rate": 0.0009451781018091999,
+      "loss": 3.005,
+      "step": 1375
+    },
+    {
+      "epoch": 0.176128,
+      "grad_norm": 0.5366721749305725,
+      "learning_rate": 0.000945083682139681,
+      "loss": 2.9139,
+      "step": 1376
+    },
+    {
+      "epoch": 0.176256,
+      "grad_norm": 0.7650388479232788,
+      "learning_rate": 0.0009449891859551172,
+      "loss": 3.0429,
+      "step": 1377
+    },
+    {
+      "epoch": 0.176384,
+      "grad_norm": 0.6005625128746033,
+      "learning_rate": 0.0009448946132717536,
+      "loss": 2.9164,
+      "step": 1378
+    },
+    {
+      "epoch": 0.176512,
+      "grad_norm": 0.5769376754760742,
+      "learning_rate": 0.0009447999641058485,
+      "loss": 2.9579,
+      "step": 1379
+    },
+    {
+      "epoch": 0.17664,
+      "grad_norm": 0.5981265902519226,
+      "learning_rate": 0.0009447052384736728,
+      "loss": 2.8975,
+      "step": 1380
+    },
+    {
+      "epoch": 0.176768,
+      "grad_norm": 0.6351479887962341,
+      "learning_rate": 0.0009446104363915112,
+      "loss": 2.9857,
+      "step": 1381
+    },
+    {
+      "epoch": 0.176896,
+      "grad_norm": 0.5744738578796387,
+      "learning_rate": 0.0009445155578756611,
+      "loss": 2.9625,
+      "step": 1382
+    },
+    {
+      "epoch": 0.177024,
+      "grad_norm": 0.6122178435325623,
+      "learning_rate": 0.0009444206029424334,
+      "loss": 2.9202,
+      "step": 1383
+    },
+    {
+      "epoch": 0.177152,
+      "grad_norm": 0.5826151967048645,
+      "learning_rate": 0.0009443255716081521,
+      "loss": 2.939,
+      "step": 1384
+    },
+    {
+      "epoch": 0.17728,
+      "grad_norm": 0.5438986420631409,
+      "learning_rate": 0.0009442304638891537,
+      "loss": 2.9909,
+      "step": 1385
+    },
+    {
+      "epoch": 0.177408,
+      "grad_norm": 0.6030034422874451,
+      "learning_rate": 0.0009441352798017886,
+      "loss": 2.9615,
+      "step": 1386
+    },
+    {
+      "epoch": 0.177536,
+      "grad_norm": 0.5625852942466736,
+      "learning_rate": 0.0009440400193624201,
+      "loss": 2.9622,
+      "step": 1387
+    },
+    {
+      "epoch": 0.177664,
+      "grad_norm": 0.5554730296134949,
+      "learning_rate": 0.0009439446825874247,
+      "loss": 2.9961,
+      "step": 1388
+    },
+    {
+      "epoch": 0.177792,
+      "grad_norm": 0.5628294944763184,
+      "learning_rate": 0.0009438492694931915,
+      "loss": 2.9864,
+      "step": 1389
+    },
+    {
+      "epoch": 0.17792,
+      "grad_norm": 0.5831946134567261,
+      "learning_rate": 0.0009437537800961234,
+      "loss": 2.9591,
+      "step": 1390
+    },
+    {
+      "epoch": 0.178048,
+      "grad_norm": 0.5962980389595032,
+      "learning_rate": 0.0009436582144126361,
+      "loss": 3.0199,
+      "step": 1391
+    },
+    {
+      "epoch": 0.178176,
+      "grad_norm": 0.5960446000099182,
+      "learning_rate": 0.0009435625724591583,
+      "loss": 2.9681,
+      "step": 1392
+    },
+    {
+      "epoch": 0.178304,
+      "grad_norm": 0.5895053148269653,
+      "learning_rate": 0.0009434668542521323,
+      "loss": 3.0413,
+      "step": 1393
+    },
+    {
+      "epoch": 0.178432,
+      "grad_norm": 0.6043865084648132,
+      "learning_rate": 0.0009433710598080127,
+      "loss": 2.9394,
+      "step": 1394
+    },
+    {
+      "epoch": 0.17856,
+      "grad_norm": 0.5843538045883179,
+      "learning_rate": 0.000943275189143268,
+      "loss": 2.9388,
+      "step": 1395
+    },
+    {
+      "epoch": 0.178688,
+      "grad_norm": 0.5861196517944336,
+      "learning_rate": 0.0009431792422743794,
+      "loss": 2.8675,
+      "step": 1396
+    },
+    {
+      "epoch": 0.178816,
+      "grad_norm": 0.5593687295913696,
+      "learning_rate": 0.0009430832192178413,
+      "loss": 2.9362,
+      "step": 1397
+    },
+    {
+      "epoch": 0.178944,
+      "grad_norm": 0.6233148574829102,
+      "learning_rate": 0.0009429871199901613,
+      "loss": 2.9993,
+      "step": 1398
+    },
+    {
+      "epoch": 0.179072,
+      "grad_norm": 0.6054777503013611,
+      "learning_rate": 0.0009428909446078598,
+      "loss": 3.088,
+      "step": 1399
+    },
+    {
+      "epoch": 0.1792,
+      "grad_norm": 0.5517017245292664,
+      "learning_rate": 0.0009427946930874703,
+      "loss": 2.8929,
+      "step": 1400
+    },
+    {
+      "epoch": 0.179328,
+      "grad_norm": 0.6232598423957825,
+      "learning_rate": 0.0009426983654455398,
+      "loss": 2.9639,
+      "step": 1401
+    },
+    {
+      "epoch": 0.179456,
+      "grad_norm": 0.619859516620636,
+      "learning_rate": 0.0009426019616986281,
+      "loss": 3.0685,
+      "step": 1402
+    },
+    {
+      "epoch": 0.179584,
+      "grad_norm": 0.5578510165214539,
+      "learning_rate": 0.0009425054818633081,
+      "loss": 3.0126,
+      "step": 1403
+    },
+    {
+      "epoch": 0.179712,
+      "grad_norm": 0.5943408608436584,
+      "learning_rate": 0.0009424089259561657,
+      "loss": 2.8479,
+      "step": 1404
+    },
+    {
+      "epoch": 0.17984,
+      "grad_norm": 0.6176751852035522,
+      "learning_rate": 0.0009423122939938003,
+      "loss": 2.9841,
+      "step": 1405
+    },
+    {
+      "epoch": 0.179968,
+      "grad_norm": 0.6411997079849243,
+      "learning_rate": 0.0009422155859928236,
+      "loss": 3.0468,
+      "step": 1406
+    },
+    {
+      "epoch": 0.180096,
+      "grad_norm": 0.567071795463562,
+      "learning_rate": 0.0009421188019698612,
+      "loss": 3.0455,
+      "step": 1407
+    },
+    {
+      "epoch": 0.180224,
+      "grad_norm": 0.5943470001220703,
+      "learning_rate": 0.0009420219419415514,
+      "loss": 3.0331,
+      "step": 1408
+    },
+    {
+      "epoch": 0.180352,
+      "grad_norm": 0.5916371941566467,
+      "learning_rate": 0.0009419250059245452,
+      "loss": 2.8617,
+      "step": 1409
+    },
+    {
+      "epoch": 0.18048,
+      "grad_norm": 0.5153305530548096,
+      "learning_rate": 0.0009418279939355074,
+      "loss": 2.7354,
+      "step": 1410
+    },
+    {
+      "epoch": 0.180608,
+      "grad_norm": 0.5837420225143433,
+      "learning_rate": 0.0009417309059911152,
+      "loss": 2.9974,
+      "step": 1411
+    },
+    {
+      "epoch": 0.180736,
+      "grad_norm": 0.60513836145401,
+      "learning_rate": 0.0009416337421080593,
+      "loss": 2.926,
+      "step": 1412
+    },
+    {
+      "epoch": 0.180864,
+      "grad_norm": 0.5677710771560669,
+      "learning_rate": 0.0009415365023030433,
+      "loss": 2.9596,
+      "step": 1413
+    },
+    {
+      "epoch": 0.180992,
+      "grad_norm": 0.5925608277320862,
+      "learning_rate": 0.0009414391865927838,
+      "loss": 2.9195,
+      "step": 1414
+    },
+    {
+      "epoch": 0.18112,
+      "grad_norm": 0.5300722718238831,
+      "learning_rate": 0.0009413417949940107,
+      "loss": 2.8971,
+      "step": 1415
+    },
+    {
+      "epoch": 0.181248,
+      "grad_norm": 0.5379540920257568,
+      "learning_rate": 0.0009412443275234663,
+      "loss": 2.9979,
+      "step": 1416
+    },
+    {
+      "epoch": 0.181376,
+      "grad_norm": 0.6440203189849854,
+      "learning_rate": 0.0009411467841979068,
+      "loss": 2.9207,
+      "step": 1417
+    },
+    {
+      "epoch": 0.181504,
+      "grad_norm": 0.6228487491607666,
+      "learning_rate": 0.0009410491650341009,
+      "loss": 3.0481,
+      "step": 1418
+    },
+    {
+      "epoch": 0.181632,
+      "grad_norm": 0.5807985067367554,
+      "learning_rate": 0.0009409514700488304,
+      "loss": 3.035,
+      "step": 1419
+    },
+    {
+      "epoch": 0.18176,
+      "grad_norm": 0.6051082015037537,
+      "learning_rate": 0.0009408536992588902,
+      "loss": 2.9845,
+      "step": 1420
+    },
+    {
+      "epoch": 0.181888,
+      "grad_norm": 0.596237063407898,
+      "learning_rate": 0.0009407558526810883,
+      "loss": 3.0444,
+      "step": 1421
+    },
+    {
+      "epoch": 0.182016,
+      "grad_norm": 0.5752583742141724,
+      "learning_rate": 0.0009406579303322458,
+      "loss": 3.0352,
+      "step": 1422
+    },
+    {
+      "epoch": 0.182144,
+      "grad_norm": 0.5452525019645691,
+      "learning_rate": 0.0009405599322291962,
+      "loss": 2.8574,
+      "step": 1423
+    },
+    {
+      "epoch": 0.182272,
+      "grad_norm": 0.6223734021186829,
+      "learning_rate": 0.0009404618583887872,
+      "loss": 3.1187,
+      "step": 1424
+    },
+    {
+      "epoch": 0.1824,
+      "grad_norm": 0.534267008304596,
+      "learning_rate": 0.0009403637088278783,
+      "loss": 2.9156,
+      "step": 1425
+    },
+    {
+      "epoch": 0.182528,
+      "grad_norm": 0.5576037168502808,
+      "learning_rate": 0.0009402654835633427,
+      "loss": 3.0558,
+      "step": 1426
+    },
+    {
+      "epoch": 0.182656,
+      "grad_norm": 0.5968494415283203,
+      "learning_rate": 0.0009401671826120665,
+      "loss": 3.0355,
+      "step": 1427
+    },
+    {
+      "epoch": 0.182784,
+      "grad_norm": 0.5726848840713501,
+      "learning_rate": 0.0009400688059909487,
+      "loss": 3.1837,
+      "step": 1428
+    },
+    {
+      "epoch": 0.182912,
+      "grad_norm": 0.5722002387046814,
+      "learning_rate": 0.0009399703537169016,
+      "loss": 2.9971,
+      "step": 1429
+    },
+    {
+      "epoch": 0.18304,
+      "grad_norm": 0.5902926325798035,
+      "learning_rate": 0.0009398718258068502,
+      "loss": 2.8949,
+      "step": 1430
+    },
+    {
+      "epoch": 0.183168,
+      "grad_norm": 0.536436140537262,
+      "learning_rate": 0.0009397732222777322,
+      "loss": 2.9257,
+      "step": 1431
+    },
+    {
+      "epoch": 0.183296,
+      "grad_norm": 0.5620376467704773,
+      "learning_rate": 0.0009396745431464993,
+      "loss": 3.0436,
+      "step": 1432
+    },
+    {
+      "epoch": 0.183424,
+      "grad_norm": 0.5422829985618591,
+      "learning_rate": 0.000939575788430115,
+      "loss": 2.8843,
+      "step": 1433
+    },
+    {
+      "epoch": 0.183552,
+      "grad_norm": 0.6089709997177124,
+      "learning_rate": 0.0009394769581455569,
+      "loss": 2.8814,
+      "step": 1434
+    },
+    {
+      "epoch": 0.18368,
+      "grad_norm": 0.5851966738700867,
+      "learning_rate": 0.0009393780523098146,
+      "loss": 2.975,
+      "step": 1435
+    },
+    {
+      "epoch": 0.183808,
+      "grad_norm": 0.5569199323654175,
+      "learning_rate": 0.0009392790709398917,
+      "loss": 2.8404,
+      "step": 1436
+    },
+    {
+      "epoch": 0.183936,
+      "grad_norm": 0.6108258962631226,
+      "learning_rate": 0.0009391800140528037,
+      "loss": 3.0013,
+      "step": 1437
+    },
+    {
+      "epoch": 0.184064,
+      "grad_norm": 0.5767658352851868,
+      "learning_rate": 0.0009390808816655801,
+      "loss": 2.8028,
+      "step": 1438
+    },
+    {
+      "epoch": 0.184192,
+      "grad_norm": 0.5856661200523376,
+      "learning_rate": 0.0009389816737952623,
+      "loss": 3.0213,
+      "step": 1439
+    },
+    {
+      "epoch": 0.18432,
+      "grad_norm": 0.5704496502876282,
+      "learning_rate": 0.0009388823904589061,
+      "loss": 2.8974,
+      "step": 1440
+    },
+    {
+      "epoch": 0.184448,
+      "grad_norm": 0.5648722648620605,
+      "learning_rate": 0.0009387830316735788,
+      "loss": 2.9387,
+      "step": 1441
+    },
+    {
+      "epoch": 0.184576,
+      "grad_norm": 0.564189612865448,
+      "learning_rate": 0.0009386835974563615,
+      "loss": 3.0016,
+      "step": 1442
+    },
+    {
+      "epoch": 0.184704,
+      "grad_norm": 0.6131287813186646,
+      "learning_rate": 0.0009385840878243483,
+      "loss": 2.9366,
+      "step": 1443
+    },
+    {
+      "epoch": 0.184832,
+      "grad_norm": 0.6616742014884949,
+      "learning_rate": 0.0009384845027946458,
+      "loss": 2.9888,
+      "step": 1444
+    },
+    {
+      "epoch": 0.18496,
+      "grad_norm": 0.58185875415802,
+      "learning_rate": 0.000938384842384374,
+      "loss": 2.8288,
+      "step": 1445
+    },
+    {
+      "epoch": 0.185088,
+      "grad_norm": 0.5773332715034485,
+      "learning_rate": 0.0009382851066106655,
+      "loss": 3.0813,
+      "step": 1446
+    },
+    {
+      "epoch": 0.185216,
+      "grad_norm": 0.5878926515579224,
+      "learning_rate": 0.0009381852954906661,
+      "loss": 2.8924,
+      "step": 1447
+    },
+    {
+      "epoch": 0.185344,
+      "grad_norm": 0.6392579674720764,
+      "learning_rate": 0.0009380854090415347,
+      "loss": 2.8872,
+      "step": 1448
+    },
+    {
+      "epoch": 0.185472,
+      "grad_norm": 0.5709802508354187,
+      "learning_rate": 0.0009379854472804425,
+      "loss": 2.8537,
+      "step": 1449
+    },
+    {
+      "epoch": 0.1856,
+      "grad_norm": 0.5672786235809326,
+      "learning_rate": 0.0009378854102245746,
+      "loss": 2.981,
+      "step": 1450
+    },
+    {
+      "epoch": 0.185728,
+      "grad_norm": 0.5494188666343689,
+      "learning_rate": 0.0009377852978911281,
+      "loss": 2.8771,
+      "step": 1451
+    },
+    {
+      "epoch": 0.185856,
+      "grad_norm": 0.5745264291763306,
+      "learning_rate": 0.0009376851102973139,
+      "loss": 2.8728,
+      "step": 1452
+    },
+    {
+      "epoch": 0.185984,
+      "grad_norm": 0.6030881404876709,
+      "learning_rate": 0.0009375848474603549,
+      "loss": 2.9917,
+      "step": 1453
+    },
+    {
+      "epoch": 0.186112,
+      "grad_norm": 0.6081180572509766,
+      "learning_rate": 0.0009374845093974878,
+      "loss": 3.108,
+      "step": 1454
+    },
+    {
+      "epoch": 0.18624,
+      "grad_norm": 0.5667645335197449,
+      "learning_rate": 0.0009373840961259619,
+      "loss": 2.8547,
+      "step": 1455
+    },
+    {
+      "epoch": 0.186368,
+      "grad_norm": 0.5367847084999084,
+      "learning_rate": 0.0009372836076630392,
+      "loss": 2.9231,
+      "step": 1456
+    },
+    {
+      "epoch": 0.186496,
+      "grad_norm": 0.5197865962982178,
+      "learning_rate": 0.0009371830440259947,
+      "loss": 3.0139,
+      "step": 1457
+    },
+    {
+      "epoch": 0.186624,
+      "grad_norm": 0.5701393485069275,
+      "learning_rate": 0.0009370824052321168,
+      "loss": 2.9484,
+      "step": 1458
+    },
+    {
+      "epoch": 0.186752,
+      "grad_norm": 0.6064413189888,
+      "learning_rate": 0.0009369816912987065,
+      "loss": 2.9903,
+      "step": 1459
+    },
+    {
+      "epoch": 0.18688,
+      "grad_norm": 0.603203296661377,
+      "learning_rate": 0.0009368809022430773,
+      "loss": 2.9531,
+      "step": 1460
+    },
+    {
+      "epoch": 0.187008,
+      "grad_norm": 0.5580971837043762,
+      "learning_rate": 0.0009367800380825564,
+      "loss": 2.9199,
+      "step": 1461
+    },
+    {
+      "epoch": 0.187136,
+      "grad_norm": 0.575786828994751,
+      "learning_rate": 0.0009366790988344834,
+      "loss": 2.8899,
+      "step": 1462
+    },
+    {
+      "epoch": 0.187264,
+      "grad_norm": 0.585085391998291,
+      "learning_rate": 0.0009365780845162108,
+      "loss": 3.0238,
+      "step": 1463
+    },
+    {
+      "epoch": 0.187392,
+      "grad_norm": 0.5848841071128845,
+      "learning_rate": 0.0009364769951451044,
+      "loss": 2.9042,
+      "step": 1464
+    },
+    {
+      "epoch": 0.18752,
+      "grad_norm": 0.5672663450241089,
+      "learning_rate": 0.0009363758307385422,
+      "loss": 2.9395,
+      "step": 1465
+    },
+    {
+      "epoch": 0.187648,
+      "grad_norm": 0.5840806365013123,
+      "learning_rate": 0.0009362745913139159,
+      "loss": 3.1057,
+      "step": 1466
+    },
+    {
+      "epoch": 0.187776,
+      "grad_norm": 0.6160809993743896,
+      "learning_rate": 0.0009361732768886298,
+      "loss": 3.0609,
+      "step": 1467
+    },
+    {
+      "epoch": 0.187904,
+      "grad_norm": 0.5637272000312805,
+      "learning_rate": 0.0009360718874801006,
+      "loss": 2.8335,
+      "step": 1468
+    },
+    {
+      "epoch": 0.188032,
+      "grad_norm": 0.5598779320716858,
+      "learning_rate": 0.0009359704231057586,
+      "loss": 2.9177,
+      "step": 1469
+    },
+    {
+      "epoch": 0.18816,
+      "grad_norm": 0.5385131239891052,
+      "learning_rate": 0.0009358688837830468,
+      "loss": 2.9821,
+      "step": 1470
+    },
+    {
+      "epoch": 0.188288,
+      "grad_norm": 0.6065384149551392,
+      "learning_rate": 0.0009357672695294208,
+      "loss": 3.0129,
+      "step": 1471
+    },
+    {
+      "epoch": 0.188416,
+      "grad_norm": 0.6106814742088318,
+      "learning_rate": 0.0009356655803623492,
+      "loss": 3.0046,
+      "step": 1472
+    },
+    {
+      "epoch": 0.188544,
+      "grad_norm": 0.5522772669792175,
+      "learning_rate": 0.0009355638162993139,
+      "loss": 2.956,
+      "step": 1473
+    },
+    {
+      "epoch": 0.188672,
+      "grad_norm": 0.5487381219863892,
+      "learning_rate": 0.0009354619773578088,
+      "loss": 2.9636,
+      "step": 1474
+    },
+    {
+      "epoch": 0.1888,
+      "grad_norm": 0.5319175124168396,
+      "learning_rate": 0.0009353600635553416,
+      "loss": 3.0249,
+      "step": 1475
+    },
+    {
+      "epoch": 0.188928,
+      "grad_norm": 0.5981173515319824,
+      "learning_rate": 0.0009352580749094324,
+      "loss": 2.9006,
+      "step": 1476
+    },
+    {
+      "epoch": 0.189056,
+      "grad_norm": 0.6037370562553406,
+      "learning_rate": 0.0009351560114376142,
+      "loss": 2.955,
+      "step": 1477
+    },
+    {
+      "epoch": 0.189184,
+      "grad_norm": 0.6774312257766724,
+      "learning_rate": 0.0009350538731574329,
+      "loss": 2.9485,
+      "step": 1478
+    },
+    {
+      "epoch": 0.189312,
+      "grad_norm": 0.6284586191177368,
+      "learning_rate": 0.0009349516600864471,
+      "loss": 3.05,
+      "step": 1479
+    },
+    {
+      "epoch": 0.18944,
+      "grad_norm": 0.6268891096115112,
+      "learning_rate": 0.0009348493722422287,
+      "loss": 3.0413,
+      "step": 1480
+    },
+    {
+      "epoch": 0.189568,
+      "grad_norm": 0.6178195476531982,
+      "learning_rate": 0.0009347470096423618,
+      "loss": 3.0825,
+      "step": 1481
+    },
+    {
+      "epoch": 0.189696,
+      "grad_norm": 0.549744188785553,
+      "learning_rate": 0.0009346445723044441,
+      "loss": 2.9948,
+      "step": 1482
+    },
+    {
+      "epoch": 0.189824,
+      "grad_norm": 0.5896495580673218,
+      "learning_rate": 0.0009345420602460855,
+      "loss": 2.8968,
+      "step": 1483
+    },
+    {
+      "epoch": 0.189952,
+      "grad_norm": 0.5850705504417419,
+      "learning_rate": 0.0009344394734849091,
+      "loss": 2.9518,
+      "step": 1484
+    },
+    {
+      "epoch": 0.19008,
+      "grad_norm": 0.589493453502655,
+      "learning_rate": 0.0009343368120385507,
+      "loss": 2.9814,
+      "step": 1485
+    },
+    {
+      "epoch": 0.190208,
+      "grad_norm": 0.6199533939361572,
+      "learning_rate": 0.0009342340759246591,
+      "loss": 3.0046,
+      "step": 1486
+    },
+    {
+      "epoch": 0.190336,
+      "grad_norm": 0.49656233191490173,
+      "learning_rate": 0.0009341312651608958,
+      "loss": 2.8714,
+      "step": 1487
+    },
+    {
+      "epoch": 0.190464,
+      "grad_norm": 0.5583142042160034,
+      "learning_rate": 0.0009340283797649352,
+      "loss": 2.8819,
+      "step": 1488
+    },
+    {
+      "epoch": 0.190592,
+      "grad_norm": 0.5915510654449463,
+      "learning_rate": 0.0009339254197544642,
+      "loss": 3.0141,
+      "step": 1489
+    },
+    {
+      "epoch": 0.19072,
+      "grad_norm": 0.5923678874969482,
+      "learning_rate": 0.0009338223851471832,
+      "loss": 2.8886,
+      "step": 1490
+    },
+    {
+      "epoch": 0.190848,
+      "grad_norm": 0.5282290577888489,
+      "learning_rate": 0.0009337192759608049,
+      "loss": 2.8565,
+      "step": 1491
+    },
+    {
+      "epoch": 0.190976,
+      "grad_norm": 0.5592545866966248,
+      "learning_rate": 0.000933616092213055,
+      "loss": 3.0226,
+      "step": 1492
+    },
+    {
+      "epoch": 0.191104,
+      "grad_norm": 0.6165173053741455,
+      "learning_rate": 0.0009335128339216719,
+      "loss": 2.9217,
+      "step": 1493
+    },
+    {
+      "epoch": 0.191232,
+      "grad_norm": 0.6189430952072144,
+      "learning_rate": 0.0009334095011044069,
+      "loss": 2.8777,
+      "step": 1494
+    },
+    {
+      "epoch": 0.19136,
+      "grad_norm": 0.6431520581245422,
+      "learning_rate": 0.0009333060937790242,
+      "loss": 2.9045,
+      "step": 1495
+    },
+    {
+      "epoch": 0.191488,
+      "grad_norm": 0.5628201365470886,
+      "learning_rate": 0.0009332026119633006,
+      "loss": 2.9511,
+      "step": 1496
+    },
+    {
+      "epoch": 0.191616,
+      "grad_norm": 0.5679343938827515,
+      "learning_rate": 0.000933099055675026,
+      "loss": 2.9057,
+      "step": 1497
+    },
+    {
+      "epoch": 0.191744,
+      "grad_norm": 0.5732449889183044,
+      "learning_rate": 0.0009329954249320027,
+      "loss": 2.8735,
+      "step": 1498
+    },
+    {
+      "epoch": 0.191872,
+      "grad_norm": 0.543315589427948,
+      "learning_rate": 0.0009328917197520461,
+      "loss": 2.8537,
+      "step": 1499
+    },
+    {
+      "epoch": 0.192,
+      "grad_norm": 0.546286404132843,
+      "learning_rate": 0.0009327879401529846,
+      "loss": 2.9983,
+      "step": 1500
+    },
+    {
+      "epoch": 0.192128,
+      "grad_norm": 0.5538829565048218,
+      "learning_rate": 0.0009326840861526586,
+      "loss": 2.8655,
+      "step": 1501
+    },
+    {
+      "epoch": 0.192256,
+      "grad_norm": 0.5417448878288269,
+      "learning_rate": 0.0009325801577689223,
+      "loss": 2.8758,
+      "step": 1502
+    },
+    {
+      "epoch": 0.192384,
+      "grad_norm": 0.550911545753479,
+      "learning_rate": 0.0009324761550196419,
+      "loss": 2.893,
+      "step": 1503
+    },
+    {
+      "epoch": 0.192512,
+      "grad_norm": 0.60374516248703,
+      "learning_rate": 0.0009323720779226967,
+      "loss": 3.08,
+      "step": 1504
+    },
+    {
+      "epoch": 0.19264,
+      "grad_norm": 0.5500683188438416,
+      "learning_rate": 0.0009322679264959789,
+      "loss": 2.8562,
+      "step": 1505
+    },
+    {
+      "epoch": 0.192768,
+      "grad_norm": 0.6039919257164001,
+      "learning_rate": 0.000932163700757393,
+      "loss": 2.847,
+      "step": 1506
+    },
+    {
+      "epoch": 0.192896,
+      "grad_norm": 0.5904696583747864,
+      "learning_rate": 0.0009320594007248573,
+      "loss": 3.0359,
+      "step": 1507
+    },
+    {
+      "epoch": 0.193024,
+      "grad_norm": 0.5734100341796875,
+      "learning_rate": 0.0009319550264163015,
+      "loss": 2.8984,
+      "step": 1508
+    },
+    {
+      "epoch": 0.193152,
+      "grad_norm": 0.584735631942749,
+      "learning_rate": 0.0009318505778496691,
+      "loss": 2.9807,
+      "step": 1509
+    },
+    {
+      "epoch": 0.19328,
+      "grad_norm": 0.6232760548591614,
+      "learning_rate": 0.0009317460550429159,
+      "loss": 2.9303,
+      "step": 1510
+    },
+    {
+      "epoch": 0.193408,
+      "grad_norm": 0.5761833190917969,
+      "learning_rate": 0.0009316414580140105,
+      "loss": 2.9282,
+      "step": 1511
+    },
+    {
+      "epoch": 0.193536,
+      "grad_norm": 0.6299952864646912,
+      "learning_rate": 0.0009315367867809345,
+      "loss": 2.9894,
+      "step": 1512
+    },
+    {
+      "epoch": 0.193664,
+      "grad_norm": 0.5804581642150879,
+      "learning_rate": 0.0009314320413616821,
+      "loss": 2.8959,
+      "step": 1513
+    },
+    {
+      "epoch": 0.193792,
+      "grad_norm": 0.597433865070343,
+      "learning_rate": 0.0009313272217742603,
+      "loss": 2.9649,
+      "step": 1514
+    },
+    {
+      "epoch": 0.19392,
+      "grad_norm": 0.5818958878517151,
+      "learning_rate": 0.0009312223280366885,
+      "loss": 2.8919,
+      "step": 1515
+    },
+    {
+      "epoch": 0.194048,
+      "grad_norm": 0.5227886438369751,
+      "learning_rate": 0.0009311173601669996,
+      "loss": 2.8656,
+      "step": 1516
+    },
+    {
+      "epoch": 0.194176,
+      "grad_norm": 0.5447340607643127,
+      "learning_rate": 0.0009310123181832382,
+      "loss": 2.9098,
+      "step": 1517
+    },
+    {
+      "epoch": 0.194304,
+      "grad_norm": 0.5548421144485474,
+      "learning_rate": 0.0009309072021034628,
+      "loss": 3.0071,
+      "step": 1518
+    },
+    {
+      "epoch": 0.194432,
+      "grad_norm": 0.5115775465965271,
+      "learning_rate": 0.0009308020119457439,
+      "loss": 2.9097,
+      "step": 1519
+    },
+    {
+      "epoch": 0.19456,
+      "grad_norm": 0.49590587615966797,
+      "learning_rate": 0.0009306967477281647,
+      "loss": 2.813,
+      "step": 1520
+    },
+    {
+      "epoch": 0.194688,
+      "grad_norm": 0.5472809076309204,
+      "learning_rate": 0.0009305914094688216,
+      "loss": 2.8727,
+      "step": 1521
+    },
+    {
+      "epoch": 0.194816,
+      "grad_norm": 0.5982186794281006,
+      "learning_rate": 0.0009304859971858233,
+      "loss": 2.9076,
+      "step": 1522
+    },
+    {
+      "epoch": 0.194944,
+      "grad_norm": 0.5688288807868958,
+      "learning_rate": 0.0009303805108972916,
+      "loss": 2.8932,
+      "step": 1523
+    },
+    {
+      "epoch": 0.195072,
+      "grad_norm": 0.6057001948356628,
+      "learning_rate": 0.0009302749506213606,
+      "loss": 3.0112,
+      "step": 1524
+    },
+    {
+      "epoch": 0.1952,
+      "grad_norm": 0.575221061706543,
+      "learning_rate": 0.0009301693163761773,
+      "loss": 2.9896,
+      "step": 1525
+    },
+    {
+      "epoch": 0.195328,
+      "grad_norm": 0.6200656294822693,
+      "learning_rate": 0.0009300636081799017,
+      "loss": 3.0505,
+      "step": 1526
+    },
+    {
+      "epoch": 0.195456,
+      "grad_norm": 0.6064473390579224,
+      "learning_rate": 0.0009299578260507061,
+      "loss": 3.0567,
+      "step": 1527
+    },
+    {
+      "epoch": 0.195584,
+      "grad_norm": 0.5797837376594543,
+      "learning_rate": 0.0009298519700067757,
+      "loss": 2.9778,
+      "step": 1528
+    },
+    {
+      "epoch": 0.195712,
+      "grad_norm": 0.5538948774337769,
+      "learning_rate": 0.0009297460400663084,
+      "loss": 2.9643,
+      "step": 1529
+    },
+    {
+      "epoch": 0.19584,
+      "grad_norm": 0.5869251489639282,
+      "learning_rate": 0.0009296400362475148,
+      "loss": 2.772,
+      "step": 1530
+    },
+    {
+      "epoch": 0.195968,
+      "grad_norm": 0.6814356446266174,
+      "learning_rate": 0.000929533958568618,
+      "loss": 3.1002,
+      "step": 1531
+    },
+    {
+      "epoch": 0.196096,
+      "grad_norm": 0.5591349601745605,
+      "learning_rate": 0.0009294278070478544,
+      "loss": 3.0232,
+      "step": 1532
+    },
+    {
+      "epoch": 0.196224,
+      "grad_norm": 0.5213653445243835,
+      "learning_rate": 0.0009293215817034722,
+      "loss": 2.9444,
+      "step": 1533
+    },
+    {
+      "epoch": 0.196352,
+      "grad_norm": 0.5450800061225891,
+      "learning_rate": 0.000929215282553733,
+      "loss": 3.1093,
+      "step": 1534
+    },
+    {
+      "epoch": 0.19648,
+      "grad_norm": 0.5859785676002502,
+      "learning_rate": 0.0009291089096169109,
+      "loss": 2.9429,
+      "step": 1535
+    },
+    {
+      "epoch": 0.196608,
+      "grad_norm": 0.5450800657272339,
+      "learning_rate": 0.0009290024629112926,
+      "loss": 2.9593,
+      "step": 1536
+    },
+    {
+      "epoch": 0.196736,
+      "grad_norm": 0.5136170387268066,
+      "learning_rate": 0.0009288959424551774,
+      "loss": 2.9573,
+      "step": 1537
+    },
+    {
+      "epoch": 0.196864,
+      "grad_norm": 0.5486687421798706,
+      "learning_rate": 0.0009287893482668774,
+      "loss": 2.9347,
+      "step": 1538
+    },
+    {
+      "epoch": 0.196992,
+      "grad_norm": 0.6012721061706543,
+      "learning_rate": 0.0009286826803647177,
+      "loss": 3.0311,
+      "step": 1539
+    },
+    {
+      "epoch": 0.19712,
+      "grad_norm": 0.5951274633407593,
+      "learning_rate": 0.0009285759387670356,
+      "loss": 2.9877,
+      "step": 1540
+    },
+    {
+      "epoch": 0.197248,
+      "grad_norm": 0.5983397960662842,
+      "learning_rate": 0.0009284691234921809,
+      "loss": 2.9835,
+      "step": 1541
+    },
+    {
+      "epoch": 0.197376,
+      "grad_norm": 0.6208573579788208,
+      "learning_rate": 0.0009283622345585168,
+      "loss": 2.9039,
+      "step": 1542
+    },
+    {
+      "epoch": 0.197504,
+      "grad_norm": 0.5884132981300354,
+      "learning_rate": 0.0009282552719844185,
+      "loss": 3.0378,
+      "step": 1543
+    },
+    {
+      "epoch": 0.197632,
+      "grad_norm": 0.5949737429618835,
+      "learning_rate": 0.0009281482357882743,
+      "loss": 3.0773,
+      "step": 1544
+    },
+    {
+      "epoch": 0.19776,
+      "grad_norm": 0.6186710000038147,
+      "learning_rate": 0.0009280411259884846,
+      "loss": 2.9041,
+      "step": 1545
+    },
+    {
+      "epoch": 0.197888,
+      "grad_norm": 0.5800589323043823,
+      "learning_rate": 0.0009279339426034633,
+      "loss": 2.8832,
+      "step": 1546
+    },
+    {
+      "epoch": 0.198016,
+      "grad_norm": 0.6050989627838135,
+      "learning_rate": 0.0009278266856516362,
+      "loss": 2.9552,
+      "step": 1547
+    },
+    {
+      "epoch": 0.198144,
+      "grad_norm": 0.5646042227745056,
+      "learning_rate": 0.000927719355151442,
+      "loss": 2.8777,
+      "step": 1548
+    },
+    {
+      "epoch": 0.198272,
+      "grad_norm": 0.6017725467681885,
+      "learning_rate": 0.0009276119511213321,
+      "loss": 2.913,
+      "step": 1549
+    },
+    {
+      "epoch": 0.1984,
+      "grad_norm": 0.5790621042251587,
+      "learning_rate": 0.0009275044735797705,
+      "loss": 3.0106,
+      "step": 1550
+    },
+    {
+      "epoch": 0.198528,
+      "grad_norm": 0.5870450139045715,
+      "learning_rate": 0.0009273969225452339,
+      "loss": 3.0299,
+      "step": 1551
+    },
+    {
+      "epoch": 0.198656,
+      "grad_norm": 0.543690025806427,
+      "learning_rate": 0.0009272892980362114,
+      "loss": 2.9841,
+      "step": 1552
+    },
+    {
+      "epoch": 0.198784,
+      "grad_norm": 0.5220555067062378,
+      "learning_rate": 0.000927181600071205,
+      "loss": 3.0678,
+      "step": 1553
+    },
+    {
+      "epoch": 0.198912,
+      "grad_norm": 0.5806132555007935,
+      "learning_rate": 0.0009270738286687293,
+      "loss": 2.9628,
+      "step": 1554
+    },
+    {
+      "epoch": 0.19904,
+      "grad_norm": 0.5786036252975464,
+      "learning_rate": 0.0009269659838473113,
+      "loss": 3.0107,
+      "step": 1555
+    },
+    {
+      "epoch": 0.199168,
+      "grad_norm": 0.6475144624710083,
+      "learning_rate": 0.0009268580656254911,
+      "loss": 2.9075,
+      "step": 1556
+    },
+    {
+      "epoch": 0.199296,
+      "grad_norm": 0.5902179479598999,
+      "learning_rate": 0.0009267500740218205,
+      "loss": 2.9632,
+      "step": 1557
+    },
+    {
+      "epoch": 0.199424,
+      "grad_norm": 0.5175977945327759,
+      "learning_rate": 0.000926642009054865,
+      "loss": 3.0341,
+      "step": 1558
+    },
+    {
+      "epoch": 0.199552,
+      "grad_norm": 0.6151183843612671,
+      "learning_rate": 0.0009265338707432023,
+      "loss": 2.8635,
+      "step": 1559
+    },
+    {
+      "epoch": 0.19968,
+      "grad_norm": 0.5659031271934509,
+      "learning_rate": 0.0009264256591054222,
+      "loss": 3.0128,
+      "step": 1560
+    },
+    {
+      "epoch": 0.199808,
+      "grad_norm": 0.5296604633331299,
+      "learning_rate": 0.0009263173741601279,
+      "loss": 2.7911,
+      "step": 1561
+    },
+    {
+      "epoch": 0.199936,
+      "grad_norm": 0.5206165313720703,
+      "learning_rate": 0.0009262090159259344,
+      "loss": 2.9005,
+      "step": 1562
+    },
+    {
+      "epoch": 0.200064,
+      "grad_norm": 0.5643444657325745,
+      "learning_rate": 0.0009261005844214703,
+      "loss": 3.0296,
+      "step": 1563
+    },
+    {
+      "epoch": 0.200192,
+      "grad_norm": 0.5839098691940308,
+      "learning_rate": 0.0009259920796653758,
+      "loss": 3.0417,
+      "step": 1564
+    },
+    {
+      "epoch": 0.20032,
+      "grad_norm": 0.5470393300056458,
+      "learning_rate": 0.0009258835016763043,
+      "loss": 2.9505,
+      "step": 1565
+    },
+    {
+      "epoch": 0.200448,
+      "grad_norm": 0.5801128149032593,
+      "learning_rate": 0.0009257748504729218,
+      "loss": 2.996,
+      "step": 1566
+    },
+    {
+      "epoch": 0.200576,
+      "grad_norm": 0.567602813243866,
+      "learning_rate": 0.0009256661260739063,
+      "loss": 2.9805,
+      "step": 1567
+    },
+    {
+      "epoch": 0.200704,
+      "grad_norm": 0.5802134871482849,
+      "learning_rate": 0.000925557328497949,
+      "loss": 2.9942,
+      "step": 1568
+    },
+    {
+      "epoch": 0.200832,
+      "grad_norm": 0.5646215081214905,
+      "learning_rate": 0.0009254484577637534,
+      "loss": 2.8529,
+      "step": 1569
+    },
+    {
+      "epoch": 0.20096,
+      "grad_norm": 0.6057287454605103,
+      "learning_rate": 0.0009253395138900358,
+      "loss": 2.8494,
+      "step": 1570
+    },
+    {
+      "epoch": 0.201088,
+      "grad_norm": 0.5700502991676331,
+      "learning_rate": 0.0009252304968955247,
+      "loss": 2.921,
+      "step": 1571
+    },
+    {
+      "epoch": 0.201216,
+      "grad_norm": 0.617802619934082,
+      "learning_rate": 0.0009251214067989616,
+      "loss": 3.0914,
+      "step": 1572
+    },
+    {
+      "epoch": 0.201344,
+      "grad_norm": 0.5670329332351685,
+      "learning_rate": 0.0009250122436191002,
+      "loss": 2.886,
+      "step": 1573
+    },
+    {
+      "epoch": 0.201472,
+      "grad_norm": 0.5999025702476501,
+      "learning_rate": 0.0009249030073747069,
+      "loss": 2.951,
+      "step": 1574
+    },
+    {
+      "epoch": 0.2016,
+      "grad_norm": 0.5880758166313171,
+      "learning_rate": 0.0009247936980845608,
+      "loss": 3.0165,
+      "step": 1575
+    },
+    {
+      "epoch": 0.201728,
+      "grad_norm": 2.124382734298706,
+      "learning_rate": 0.0009246843157674534,
+      "loss": 2.9381,
+      "step": 1576
+    },
+    {
+      "epoch": 0.201856,
+      "grad_norm": 0.564065158367157,
+      "learning_rate": 0.0009245748604421886,
+      "loss": 3.0042,
+      "step": 1577
+    },
+    {
+      "epoch": 0.201984,
+      "grad_norm": 0.5859578847885132,
+      "learning_rate": 0.0009244653321275833,
+      "loss": 2.9425,
+      "step": 1578
+    },
+    {
+      "epoch": 0.202112,
+      "grad_norm": 0.5504744052886963,
+      "learning_rate": 0.0009243557308424666,
+      "loss": 3.0271,
+      "step": 1579
+    },
+    {
+      "epoch": 0.20224,
+      "grad_norm": 0.6082347631454468,
+      "learning_rate": 0.0009242460566056803,
+      "loss": 2.8803,
+      "step": 1580
+    },
+    {
+      "epoch": 0.202368,
+      "grad_norm": 0.5363054871559143,
+      "learning_rate": 0.0009241363094360784,
+      "loss": 2.9101,
+      "step": 1581
+    },
+    {
+      "epoch": 0.202496,
+      "grad_norm": 0.5918108224868774,
+      "learning_rate": 0.0009240264893525281,
+      "loss": 3.0755,
+      "step": 1582
+    },
+    {
+      "epoch": 0.202624,
+      "grad_norm": 0.5524587631225586,
+      "learning_rate": 0.0009239165963739084,
+      "loss": 3.0226,
+      "step": 1583
+    },
+    {
+      "epoch": 0.202752,
+      "grad_norm": 0.5251649022102356,
+      "learning_rate": 0.0009238066305191113,
+      "loss": 2.9721,
+      "step": 1584
+    },
+    {
+      "epoch": 0.20288,
+      "grad_norm": 0.6157435178756714,
+      "learning_rate": 0.0009236965918070412,
+      "loss": 2.9755,
+      "step": 1585
+    },
+    {
+      "epoch": 0.203008,
+      "grad_norm": 0.6009249091148376,
+      "learning_rate": 0.0009235864802566152,
+      "loss": 2.9539,
+      "step": 1586
+    },
+    {
+      "epoch": 0.203136,
+      "grad_norm": 0.5496651530265808,
+      "learning_rate": 0.0009234762958867626,
+      "loss": 2.8603,
+      "step": 1587
+    },
+    {
+      "epoch": 0.203264,
+      "grad_norm": 0.6282094717025757,
+      "learning_rate": 0.0009233660387164254,
+      "loss": 2.9022,
+      "step": 1588
+    },
+    {
+      "epoch": 0.203392,
+      "grad_norm": 0.5855196118354797,
+      "learning_rate": 0.0009232557087645579,
+      "loss": 2.8298,
+      "step": 1589
+    },
+    {
+      "epoch": 0.20352,
+      "grad_norm": 0.5574777722358704,
+      "learning_rate": 0.0009231453060501274,
+      "loss": 2.8854,
+      "step": 1590
+    },
+    {
+      "epoch": 0.203648,
+      "grad_norm": 0.5931380987167358,
+      "learning_rate": 0.0009230348305921131,
+      "loss": 2.9971,
+      "step": 1591
+    },
+    {
+      "epoch": 0.203776,
+      "grad_norm": 0.5480794906616211,
+      "learning_rate": 0.0009229242824095073,
+      "loss": 3.0626,
+      "step": 1592
+    },
+    {
+      "epoch": 0.203904,
+      "grad_norm": 0.5436170101165771,
+      "learning_rate": 0.0009228136615213144,
+      "loss": 2.9566,
+      "step": 1593
+    },
+    {
+      "epoch": 0.204032,
+      "grad_norm": 0.6178094148635864,
+      "learning_rate": 0.0009227029679465514,
+      "loss": 3.0176,
+      "step": 1594
+    },
+    {
+      "epoch": 0.20416,
+      "grad_norm": 0.5751908421516418,
+      "learning_rate": 0.0009225922017042479,
+      "loss": 2.8514,
+      "step": 1595
+    },
+    {
+      "epoch": 0.204288,
+      "grad_norm": 0.5603744387626648,
+      "learning_rate": 0.0009224813628134457,
+      "loss": 2.9522,
+      "step": 1596
+    },
+    {
+      "epoch": 0.204416,
+      "grad_norm": 0.564110279083252,
+      "learning_rate": 0.0009223704512931994,
+      "loss": 3.0472,
+      "step": 1597
+    },
+    {
+      "epoch": 0.204544,
+      "grad_norm": 0.5471618175506592,
+      "learning_rate": 0.0009222594671625759,
+      "loss": 2.9077,
+      "step": 1598
+    },
+    {
+      "epoch": 0.204672,
+      "grad_norm": 0.548700213432312,
+      "learning_rate": 0.000922148410440655,
+      "loss": 3.0663,
+      "step": 1599
+    },
+    {
+      "epoch": 0.2048,
+      "grad_norm": 0.5859079957008362,
+      "learning_rate": 0.0009220372811465282,
+      "loss": 2.9653,
+      "step": 1600
+    },
+    {
+      "epoch": 0.204928,
+      "grad_norm": 0.6208023428916931,
+      "learning_rate": 0.0009219260792993001,
+      "loss": 2.9315,
+      "step": 1601
+    },
+    {
+      "epoch": 0.205056,
+      "grad_norm": 0.5638324022293091,
+      "learning_rate": 0.0009218148049180877,
+      "loss": 2.9363,
+      "step": 1602
+    },
+    {
+      "epoch": 0.205184,
+      "grad_norm": 0.575494647026062,
+      "learning_rate": 0.0009217034580220201,
+      "loss": 2.8097,
+      "step": 1603
+    },
+    {
+      "epoch": 0.205312,
+      "grad_norm": 0.5974637866020203,
+      "learning_rate": 0.0009215920386302394,
+      "loss": 2.9654,
+      "step": 1604
+    },
+    {
+      "epoch": 0.20544,
+      "grad_norm": 0.6177108883857727,
+      "learning_rate": 0.0009214805467618997,
+      "loss": 2.9635,
+      "step": 1605
+    },
+    {
+      "epoch": 0.205568,
+      "grad_norm": 0.5513960123062134,
+      "learning_rate": 0.0009213689824361678,
+      "loss": 3.0709,
+      "step": 1606
+    },
+    {
+      "epoch": 0.205696,
+      "grad_norm": 0.5796489715576172,
+      "learning_rate": 0.0009212573456722226,
+      "loss": 3.0287,
+      "step": 1607
+    },
+    {
+      "epoch": 0.205824,
+      "grad_norm": 0.5631668567657471,
+      "learning_rate": 0.0009211456364892564,
+      "loss": 2.8517,
+      "step": 1608
+    },
+    {
+      "epoch": 0.205952,
+      "grad_norm": 0.5153703093528748,
+      "learning_rate": 0.0009210338549064728,
+      "loss": 2.8945,
+      "step": 1609
+    },
+    {
+      "epoch": 0.20608,
+      "grad_norm": 0.6049119830131531,
+      "learning_rate": 0.0009209220009430885,
+      "loss": 3.0113,
+      "step": 1610
+    },
+    {
+      "epoch": 0.206208,
+      "grad_norm": 0.5580521821975708,
+      "learning_rate": 0.0009208100746183326,
+      "loss": 2.8193,
+      "step": 1611
+    },
+    {
+      "epoch": 0.206336,
+      "grad_norm": 0.5967240929603577,
+      "learning_rate": 0.0009206980759514464,
+      "loss": 2.9697,
+      "step": 1612
+    },
+    {
+      "epoch": 0.206464,
+      "grad_norm": 0.5748556852340698,
+      "learning_rate": 0.0009205860049616837,
+      "loss": 2.8934,
+      "step": 1613
+    },
+    {
+      "epoch": 0.206592,
+      "grad_norm": 0.6059197783470154,
+      "learning_rate": 0.000920473861668311,
+      "loss": 2.8642,
+      "step": 1614
+    },
+    {
+      "epoch": 0.20672,
+      "grad_norm": 0.5981786847114563,
+      "learning_rate": 0.0009203616460906068,
+      "loss": 2.9611,
+      "step": 1615
+    },
+    {
+      "epoch": 0.206848,
+      "grad_norm": 0.5657117962837219,
+      "learning_rate": 0.0009202493582478625,
+      "loss": 2.867,
+      "step": 1616
+    },
+    {
+      "epoch": 0.206976,
+      "grad_norm": 0.5388901233673096,
+      "learning_rate": 0.0009201369981593814,
+      "loss": 2.9987,
+      "step": 1617
+    },
+    {
+      "epoch": 0.207104,
+      "grad_norm": 0.5569252967834473,
+      "learning_rate": 0.0009200245658444799,
+      "loss": 2.9897,
+      "step": 1618
+    },
+    {
+      "epoch": 0.207232,
+      "grad_norm": 0.5674579739570618,
+      "learning_rate": 0.0009199120613224861,
+      "loss": 2.8522,
+      "step": 1619
+    },
+    {
+      "epoch": 0.20736,
+      "grad_norm": 0.5259724259376526,
+      "learning_rate": 0.0009197994846127409,
+      "loss": 2.9165,
+      "step": 1620
+    },
+    {
+      "epoch": 0.207488,
+      "grad_norm": 0.5774664878845215,
+      "learning_rate": 0.0009196868357345976,
+      "loss": 3.1289,
+      "step": 1621
+    },
+    {
+      "epoch": 0.207616,
+      "grad_norm": 0.5677288174629211,
+      "learning_rate": 0.0009195741147074217,
+      "loss": 3.0903,
+      "step": 1622
+    },
+    {
+      "epoch": 0.207744,
+      "grad_norm": 0.5523772835731506,
+      "learning_rate": 0.0009194613215505915,
+      "loss": 2.8796,
+      "step": 1623
+    },
+    {
+      "epoch": 0.207872,
+      "grad_norm": 0.5352883338928223,
+      "learning_rate": 0.0009193484562834973,
+      "loss": 2.9181,
+      "step": 1624
+    },
+    {
+      "epoch": 0.208,
+      "grad_norm": 0.6119636297225952,
+      "learning_rate": 0.0009192355189255418,
+      "loss": 3.022,
+      "step": 1625
+    },
+    {
+      "epoch": 0.208128,
+      "grad_norm": 0.5815951228141785,
+      "learning_rate": 0.0009191225094961407,
+      "loss": 2.811,
+      "step": 1626
+    },
+    {
+      "epoch": 0.208256,
+      "grad_norm": 0.6034047603607178,
+      "learning_rate": 0.000919009428014721,
+      "loss": 2.9311,
+      "step": 1627
+    },
+    {
+      "epoch": 0.208384,
+      "grad_norm": 0.5794652104377747,
+      "learning_rate": 0.0009188962745007233,
+      "loss": 2.9794,
+      "step": 1628
+    },
+    {
+      "epoch": 0.208512,
+      "grad_norm": 0.5475305318832397,
+      "learning_rate": 0.0009187830489735995,
+      "loss": 3.0152,
+      "step": 1629
+    },
+    {
+      "epoch": 0.20864,
+      "grad_norm": 0.529373049736023,
+      "learning_rate": 0.0009186697514528148,
+      "loss": 3.0522,
+      "step": 1630
+    },
+    {
+      "epoch": 0.208768,
+      "grad_norm": 0.5764333009719849,
+      "learning_rate": 0.0009185563819578462,
+      "loss": 2.9802,
+      "step": 1631
+    },
+    {
+      "epoch": 0.208896,
+      "grad_norm": 0.5849448442459106,
+      "learning_rate": 0.0009184429405081832,
+      "loss": 2.9789,
+      "step": 1632
+    },
+    {
+      "epoch": 0.209024,
+      "grad_norm": 0.5767343640327454,
+      "learning_rate": 0.0009183294271233277,
+      "loss": 3.0077,
+      "step": 1633
+    },
+    {
+      "epoch": 0.209152,
+      "grad_norm": 0.5651653409004211,
+      "learning_rate": 0.000918215841822794,
+      "loss": 3.0016,
+      "step": 1634
+    },
+    {
+      "epoch": 0.20928,
+      "grad_norm": 0.5356111526489258,
+      "learning_rate": 0.0009181021846261087,
+      "loss": 2.7907,
+      "step": 1635
+    },
+    {
+      "epoch": 0.209408,
+      "grad_norm": 0.5639657974243164,
+      "learning_rate": 0.0009179884555528109,
+      "loss": 2.9593,
+      "step": 1636
+    },
+    {
+      "epoch": 0.209536,
+      "grad_norm": 0.5674639344215393,
+      "learning_rate": 0.0009178746546224517,
+      "loss": 2.9569,
+      "step": 1637
+    },
+    {
+      "epoch": 0.209664,
+      "grad_norm": 0.5897277593612671,
+      "learning_rate": 0.000917760781854595,
+      "loss": 2.9308,
+      "step": 1638
+    },
+    {
+      "epoch": 0.209792,
+      "grad_norm": 0.5392431020736694,
+      "learning_rate": 0.0009176468372688168,
+      "loss": 3.0061,
+      "step": 1639
+    },
+    {
+      "epoch": 0.20992,
+      "grad_norm": 0.5617191791534424,
+      "learning_rate": 0.0009175328208847056,
+      "loss": 2.9557,
+      "step": 1640
+    },
+    {
+      "epoch": 0.210048,
+      "grad_norm": 0.5266217589378357,
+      "learning_rate": 0.000917418732721862,
+      "loss": 3.0352,
+      "step": 1641
+    },
+    {
+      "epoch": 0.210176,
+      "grad_norm": 0.5281474590301514,
+      "learning_rate": 0.000917304572799899,
+      "loss": 3.1155,
+      "step": 1642
+    },
+    {
+      "epoch": 0.210304,
+      "grad_norm": 0.5831418633460999,
+      "learning_rate": 0.0009171903411384422,
+      "loss": 3.0107,
+      "step": 1643
+    },
+    {
+      "epoch": 0.210432,
+      "grad_norm": 0.5657524466514587,
+      "learning_rate": 0.0009170760377571292,
+      "loss": 2.8442,
+      "step": 1644
+    },
+    {
+      "epoch": 0.21056,
+      "grad_norm": 0.5452412366867065,
+      "learning_rate": 0.0009169616626756102,
+      "loss": 2.9309,
+      "step": 1645
+    },
+    {
+      "epoch": 0.210688,
+      "grad_norm": 0.5496721267700195,
+      "learning_rate": 0.0009168472159135476,
+      "loss": 3.0865,
+      "step": 1646
+    },
+    {
+      "epoch": 0.210816,
+      "grad_norm": 0.535876989364624,
+      "learning_rate": 0.000916732697490616,
+      "loss": 2.9873,
+      "step": 1647
+    },
+    {
+      "epoch": 0.210944,
+      "grad_norm": 0.595379650592804,
+      "learning_rate": 0.0009166181074265025,
+      "loss": 2.8894,
+      "step": 1648
+    },
+    {
+      "epoch": 0.211072,
+      "grad_norm": 0.5277078151702881,
+      "learning_rate": 0.0009165034457409066,
+      "loss": 2.8442,
+      "step": 1649
+    },
+    {
+      "epoch": 0.2112,
+      "grad_norm": 0.526021420955658,
+      "learning_rate": 0.0009163887124535397,
+      "loss": 3.0534,
+      "step": 1650
+    },
+    {
+      "epoch": 0.211328,
+      "grad_norm": 0.5715261101722717,
+      "learning_rate": 0.0009162739075841262,
+      "loss": 2.9888,
+      "step": 1651
+    },
+    {
+      "epoch": 0.211456,
+      "grad_norm": 0.5544307231903076,
+      "learning_rate": 0.000916159031152402,
+      "loss": 3.0075,
+      "step": 1652
+    },
+    {
+      "epoch": 0.211584,
+      "grad_norm": 0.5213819742202759,
+      "learning_rate": 0.0009160440831781157,
+      "loss": 2.8761,
+      "step": 1653
+    },
+    {
+      "epoch": 0.211712,
+      "grad_norm": 0.5173313617706299,
+      "learning_rate": 0.0009159290636810283,
+      "loss": 3.0533,
+      "step": 1654
+    },
+    {
+      "epoch": 0.21184,
+      "grad_norm": 0.554438054561615,
+      "learning_rate": 0.0009158139726809132,
+      "loss": 3.0541,
+      "step": 1655
+    },
+    {
+      "epoch": 0.211968,
+      "grad_norm": 0.5534675121307373,
+      "learning_rate": 0.0009156988101975554,
+      "loss": 2.952,
+      "step": 1656
+    },
+    {
+      "epoch": 0.212096,
+      "grad_norm": 0.5288779735565186,
+      "learning_rate": 0.0009155835762507531,
+      "loss": 3.0051,
+      "step": 1657
+    },
+    {
+      "epoch": 0.212224,
+      "grad_norm": 0.5666531324386597,
+      "learning_rate": 0.0009154682708603162,
+      "loss": 2.915,
+      "step": 1658
+    },
+    {
+      "epoch": 0.212352,
+      "grad_norm": 0.5739021897315979,
+      "learning_rate": 0.0009153528940460669,
+      "loss": 2.9623,
+      "step": 1659
+    },
+    {
+      "epoch": 0.21248,
+      "grad_norm": 0.5277852416038513,
+      "learning_rate": 0.0009152374458278401,
+      "loss": 2.8447,
+      "step": 1660
+    },
+    {
+      "epoch": 0.212608,
+      "grad_norm": 0.5538998246192932,
+      "learning_rate": 0.0009151219262254825,
+      "loss": 2.9077,
+      "step": 1661
+    },
+    {
+      "epoch": 0.212736,
+      "grad_norm": 0.5257222056388855,
+      "learning_rate": 0.0009150063352588531,
+      "loss": 2.8991,
+      "step": 1662
+    },
+    {
+      "epoch": 0.212864,
+      "grad_norm": 0.5435817837715149,
+      "learning_rate": 0.0009148906729478236,
+      "loss": 2.8106,
+      "step": 1663
+    },
+    {
+      "epoch": 0.212992,
+      "grad_norm": 0.571153461933136,
+      "learning_rate": 0.0009147749393122775,
+      "loss": 3.0363,
+      "step": 1664
+    },
+    {
+      "epoch": 0.21312,
+      "grad_norm": 0.6040127873420715,
+      "learning_rate": 0.000914659134372111,
+      "loss": 2.9681,
+      "step": 1665
+    },
+    {
+      "epoch": 0.213248,
+      "grad_norm": 0.5574022531509399,
+      "learning_rate": 0.0009145432581472322,
+      "loss": 2.9536,
+      "step": 1666
+    },
+    {
+      "epoch": 0.213376,
+      "grad_norm": 0.5823351740837097,
+      "learning_rate": 0.0009144273106575613,
+      "loss": 2.899,
+      "step": 1667
+    },
+    {
+      "epoch": 0.213504,
+      "grad_norm": 0.5689985752105713,
+      "learning_rate": 0.0009143112919230314,
+      "loss": 2.9285,
+      "step": 1668
+    },
+    {
+      "epoch": 0.213632,
+      "grad_norm": 0.5642018914222717,
+      "learning_rate": 0.0009141952019635873,
+      "loss": 3.0597,
+      "step": 1669
+    },
+    {
+      "epoch": 0.21376,
+      "grad_norm": 0.6142164468765259,
+      "learning_rate": 0.0009140790407991861,
+      "loss": 3.0505,
+      "step": 1670
+    },
+    {
+      "epoch": 0.213888,
+      "grad_norm": 0.5443561673164368,
+      "learning_rate": 0.0009139628084497975,
+      "loss": 2.9659,
+      "step": 1671
+    },
+    {
+      "epoch": 0.214016,
+      "grad_norm": 0.5128235816955566,
+      "learning_rate": 0.0009138465049354031,
+      "loss": 2.9647,
+      "step": 1672
+    },
+    {
+      "epoch": 0.214144,
+      "grad_norm": 0.5615318417549133,
+      "learning_rate": 0.0009137301302759968,
+      "loss": 2.9804,
+      "step": 1673
+    },
+    {
+      "epoch": 0.214272,
+      "grad_norm": 0.5283939242362976,
+      "learning_rate": 0.0009136136844915846,
+      "loss": 3.0091,
+      "step": 1674
+    },
+    {
+      "epoch": 0.2144,
+      "grad_norm": 0.551540732383728,
+      "learning_rate": 0.000913497167602185,
+      "loss": 3.0016,
+      "step": 1675
+    },
+    {
+      "epoch": 0.214528,
+      "grad_norm": 0.5556532144546509,
+      "learning_rate": 0.0009133805796278287,
+      "loss": 2.9292,
+      "step": 1676
+    },
+    {
+      "epoch": 0.214656,
+      "grad_norm": 0.6225051283836365,
+      "learning_rate": 0.0009132639205885585,
+      "loss": 2.8883,
+      "step": 1677
+    },
+    {
+      "epoch": 0.214784,
+      "grad_norm": 0.6277577877044678,
+      "learning_rate": 0.0009131471905044293,
+      "loss": 2.9148,
+      "step": 1678
+    },
+    {
+      "epoch": 0.214912,
+      "grad_norm": 0.5365939736366272,
+      "learning_rate": 0.0009130303893955084,
+      "loss": 2.9472,
+      "step": 1679
+    },
+    {
+      "epoch": 0.21504,
+      "grad_norm": 0.5846887826919556,
+      "learning_rate": 0.0009129135172818753,
+      "loss": 2.9407,
+      "step": 1680
+    },
+    {
+      "epoch": 0.215168,
+      "grad_norm": 0.5067850351333618,
+      "learning_rate": 0.0009127965741836218,
+      "loss": 2.9924,
+      "step": 1681
+    },
+    {
+      "epoch": 0.215296,
+      "grad_norm": 0.5378647446632385,
+      "learning_rate": 0.0009126795601208515,
+      "loss": 2.9268,
+      "step": 1682
+    },
+    {
+      "epoch": 0.215424,
+      "grad_norm": 0.5232348442077637,
+      "learning_rate": 0.0009125624751136807,
+      "loss": 2.9811,
+      "step": 1683
+    },
+    {
+      "epoch": 0.215552,
+      "grad_norm": 0.554328203201294,
+      "learning_rate": 0.0009124453191822376,
+      "loss": 3.0211,
+      "step": 1684
+    },
+    {
+      "epoch": 0.21568,
+      "grad_norm": 0.5367416739463806,
+      "learning_rate": 0.0009123280923466626,
+      "loss": 2.8811,
+      "step": 1685
+    },
+    {
+      "epoch": 0.215808,
+      "grad_norm": 0.5860336422920227,
+      "learning_rate": 0.0009122107946271086,
+      "loss": 2.9454,
+      "step": 1686
+    },
+    {
+      "epoch": 0.215936,
+      "grad_norm": 0.5598700046539307,
+      "learning_rate": 0.00091209342604374,
+      "loss": 3.0253,
+      "step": 1687
+    },
+    {
+      "epoch": 0.216064,
+      "grad_norm": 0.5433297157287598,
+      "learning_rate": 0.000911975986616734,
+      "loss": 2.905,
+      "step": 1688
+    },
+    {
+      "epoch": 0.216192,
+      "grad_norm": 0.5585367679595947,
+      "learning_rate": 0.0009118584763662803,
+      "loss": 2.893,
+      "step": 1689
+    },
+    {
+      "epoch": 0.21632,
+      "grad_norm": 0.601786196231842,
+      "learning_rate": 0.0009117408953125795,
+      "loss": 2.8591,
+      "step": 1690
+    },
+    {
+      "epoch": 0.216448,
+      "grad_norm": 0.5585361123085022,
+      "learning_rate": 0.0009116232434758456,
+      "loss": 2.9667,
+      "step": 1691
+    },
+    {
+      "epoch": 0.216576,
+      "grad_norm": 0.5605311393737793,
+      "learning_rate": 0.0009115055208763041,
+      "loss": 2.9708,
+      "step": 1692
+    },
+    {
+      "epoch": 0.216704,
+      "grad_norm": 0.5497050881385803,
+      "learning_rate": 0.0009113877275341931,
+      "loss": 2.9594,
+      "step": 1693
+    },
+    {
+      "epoch": 0.216832,
+      "grad_norm": 0.5313854813575745,
+      "learning_rate": 0.0009112698634697624,
+      "loss": 2.8728,
+      "step": 1694
+    },
+    {
+      "epoch": 0.21696,
+      "grad_norm": 0.5228835940361023,
+      "learning_rate": 0.0009111519287032743,
+      "loss": 3.0214,
+      "step": 1695
+    },
+    {
+      "epoch": 0.217088,
+      "grad_norm": 0.561218798160553,
+      "learning_rate": 0.0009110339232550034,
+      "loss": 2.915,
+      "step": 1696
+    },
+    {
+      "epoch": 0.217216,
+      "grad_norm": 0.6116324067115784,
+      "learning_rate": 0.0009109158471452357,
+      "loss": 3.0034,
+      "step": 1697
+    },
+    {
+      "epoch": 0.217344,
+      "grad_norm": 0.568450391292572,
+      "learning_rate": 0.0009107977003942702,
+      "loss": 2.939,
+      "step": 1698
+    },
+    {
+      "epoch": 0.217472,
+      "grad_norm": 0.5406863689422607,
+      "learning_rate": 0.0009106794830224179,
+      "loss": 2.8909,
+      "step": 1699
+    },
+    {
+      "epoch": 0.2176,
+      "grad_norm": 0.5565332174301147,
+      "learning_rate": 0.0009105611950500012,
+      "loss": 2.9938,
+      "step": 1700
+    },
+    {
+      "epoch": 0.217728,
+      "grad_norm": 0.559166669845581,
+      "learning_rate": 0.0009104428364973553,
+      "loss": 2.8378,
+      "step": 1701
+    },
+    {
+      "epoch": 0.217856,
+      "grad_norm": 0.5835161209106445,
+      "learning_rate": 0.0009103244073848278,
+      "loss": 3.1303,
+      "step": 1702
+    },
+    {
+      "epoch": 0.217984,
+      "grad_norm": 0.5451844334602356,
+      "learning_rate": 0.0009102059077327778,
+      "loss": 3.0575,
+      "step": 1703
+    },
+    {
+      "epoch": 0.218112,
+      "grad_norm": 0.5609920620918274,
+      "learning_rate": 0.0009100873375615765,
+      "loss": 2.8795,
+      "step": 1704
+    },
+    {
+      "epoch": 0.21824,
+      "grad_norm": 0.5540197491645813,
+      "learning_rate": 0.0009099686968916082,
+      "loss": 2.8878,
+      "step": 1705
+    },
+    {
+      "epoch": 0.218368,
+      "grad_norm": 0.48126086592674255,
+      "learning_rate": 0.0009098499857432676,
+      "loss": 2.9293,
+      "step": 1706
+    },
+    {
+      "epoch": 0.218496,
+      "grad_norm": 0.5308560132980347,
+      "learning_rate": 0.0009097312041369634,
+      "loss": 2.8366,
+      "step": 1707
+    },
+    {
+      "epoch": 0.218624,
+      "grad_norm": 0.5198960900306702,
+      "learning_rate": 0.000909612352093115,
+      "loss": 2.8209,
+      "step": 1708
+    },
+    {
+      "epoch": 0.218752,
+      "grad_norm": 0.5664052963256836,
+      "learning_rate": 0.0009094934296321549,
+      "loss": 2.9882,
+      "step": 1709
+    },
+    {
+      "epoch": 0.21888,
+      "grad_norm": 0.5311546921730042,
+      "learning_rate": 0.0009093744367745267,
+      "loss": 2.9374,
+      "step": 1710
+    },
+    {
+      "epoch": 0.219008,
+      "grad_norm": 0.5171306729316711,
+      "learning_rate": 0.0009092553735406871,
+      "loss": 3.0301,
+      "step": 1711
+    },
+    {
+      "epoch": 0.219136,
+      "grad_norm": 0.5814864039421082,
+      "learning_rate": 0.0009091362399511043,
+      "loss": 2.8933,
+      "step": 1712
+    },
+    {
+      "epoch": 0.219264,
+      "grad_norm": 0.5591967105865479,
+      "learning_rate": 0.0009090170360262586,
+      "loss": 2.9149,
+      "step": 1713
+    },
+    {
+      "epoch": 0.219392,
+      "grad_norm": 0.57537841796875,
+      "learning_rate": 0.0009088977617866428,
+      "loss": 2.9487,
+      "step": 1714
+    },
+    {
+      "epoch": 0.21952,
+      "grad_norm": 0.5295172333717346,
+      "learning_rate": 0.0009087784172527614,
+      "loss": 2.9306,
+      "step": 1715
+    },
+    {
+      "epoch": 0.219648,
+      "grad_norm": 0.5500990152359009,
+      "learning_rate": 0.0009086590024451311,
+      "loss": 2.8125,
+      "step": 1716
+    },
+    {
+      "epoch": 0.219776,
+      "grad_norm": 0.5544302463531494,
+      "learning_rate": 0.0009085395173842807,
+      "loss": 2.9181,
+      "step": 1717
+    },
+    {
+      "epoch": 0.219904,
+      "grad_norm": 0.575838565826416,
+      "learning_rate": 0.000908419962090751,
+      "loss": 3.0199,
+      "step": 1718
+    },
+    {
+      "epoch": 0.220032,
+      "grad_norm": 0.5310478806495667,
+      "learning_rate": 0.000908300336585095,
+      "loss": 2.8839,
+      "step": 1719
+    },
+    {
+      "epoch": 0.22016,
+      "grad_norm": 0.566171407699585,
+      "learning_rate": 0.0009081806408878778,
+      "loss": 3.011,
+      "step": 1720
+    },
+    {
+      "epoch": 0.220288,
+      "grad_norm": 0.5573983788490295,
+      "learning_rate": 0.0009080608750196764,
+      "loss": 2.975,
+      "step": 1721
+    },
+    {
+      "epoch": 0.220416,
+      "grad_norm": 0.5381975173950195,
+      "learning_rate": 0.0009079410390010798,
+      "loss": 2.9884,
+      "step": 1722
+    },
+    {
+      "epoch": 0.220544,
+      "grad_norm": 0.5417394638061523,
+      "learning_rate": 0.0009078211328526895,
+      "loss": 2.7408,
+      "step": 1723
+    },
+    {
+      "epoch": 0.220672,
+      "grad_norm": 0.5979185700416565,
+      "learning_rate": 0.0009077011565951185,
+      "loss": 2.8844,
+      "step": 1724
+    },
+    {
+      "epoch": 0.2208,
+      "grad_norm": 0.5393586754798889,
+      "learning_rate": 0.0009075811102489923,
+      "loss": 3.017,
+      "step": 1725
+    },
+    {
+      "epoch": 0.220928,
+      "grad_norm": 0.532637357711792,
+      "learning_rate": 0.0009074609938349481,
+      "loss": 3.0198,
+      "step": 1726
+    },
+    {
+      "epoch": 0.221056,
+      "grad_norm": 0.5806271433830261,
+      "learning_rate": 0.0009073408073736354,
+      "loss": 2.9079,
+      "step": 1727
+    },
+    {
+      "epoch": 0.221184,
+      "grad_norm": 0.5469481945037842,
+      "learning_rate": 0.0009072205508857154,
+      "loss": 3.0499,
+      "step": 1728
+    },
+    {
+      "epoch": 0.221312,
+      "grad_norm": 0.542588472366333,
+      "learning_rate": 0.0009071002243918621,
+      "loss": 2.9889,
+      "step": 1729
+    },
+    {
+      "epoch": 0.22144,
+      "grad_norm": 0.5996057391166687,
+      "learning_rate": 0.0009069798279127605,
+      "loss": 3.0468,
+      "step": 1730
+    },
+    {
+      "epoch": 0.221568,
+      "grad_norm": 0.5646742582321167,
+      "learning_rate": 0.0009068593614691085,
+      "loss": 3.0295,
+      "step": 1731
+    },
+    {
+      "epoch": 0.221696,
+      "grad_norm": 0.534864068031311,
+      "learning_rate": 0.0009067388250816156,
+      "loss": 2.9851,
+      "step": 1732
+    },
+    {
+      "epoch": 0.221824,
+      "grad_norm": 0.561005175113678,
+      "learning_rate": 0.0009066182187710031,
+      "loss": 3.0063,
+      "step": 1733
+    },
+    {
+      "epoch": 0.221952,
+      "grad_norm": 0.5051330924034119,
+      "learning_rate": 0.0009064975425580051,
+      "loss": 2.7608,
+      "step": 1734
+    },
+    {
+      "epoch": 0.22208,
+      "grad_norm": 0.5232276320457458,
+      "learning_rate": 0.0009063767964633669,
+      "loss": 2.9159,
+      "step": 1735
+    },
+    {
+      "epoch": 0.222208,
+      "grad_norm": 0.6101745963096619,
+      "learning_rate": 0.0009062559805078463,
+      "loss": 2.9477,
+      "step": 1736
+    },
+    {
+      "epoch": 0.222336,
+      "grad_norm": 0.5451213121414185,
+      "learning_rate": 0.0009061350947122129,
+      "loss": 2.9282,
+      "step": 1737
+    },
+    {
+      "epoch": 0.222464,
+      "grad_norm": 0.5313256978988647,
+      "learning_rate": 0.0009060141390972485,
+      "loss": 2.8278,
+      "step": 1738
+    },
+    {
+      "epoch": 0.222592,
+      "grad_norm": 0.551654577255249,
+      "learning_rate": 0.0009058931136837465,
+      "loss": 3.047,
+      "step": 1739
+    },
+    {
+      "epoch": 0.22272,
+      "grad_norm": 0.5677871704101562,
+      "learning_rate": 0.0009057720184925129,
+      "loss": 2.9713,
+      "step": 1740
+    },
+    {
+      "epoch": 0.222848,
+      "grad_norm": 0.5783289074897766,
+      "learning_rate": 0.0009056508535443651,
+      "loss": 2.8844,
+      "step": 1741
+    },
+    {
+      "epoch": 0.222976,
+      "grad_norm": 0.5367400050163269,
+      "learning_rate": 0.0009055296188601329,
+      "loss": 3.0019,
+      "step": 1742
+    },
+    {
+      "epoch": 0.223104,
+      "grad_norm": 0.5881387591362,
+      "learning_rate": 0.0009054083144606579,
+      "loss": 2.8059,
+      "step": 1743
+    },
+    {
+      "epoch": 0.223232,
+      "grad_norm": 0.581623375415802,
+      "learning_rate": 0.0009052869403667938,
+      "loss": 3.0192,
+      "step": 1744
+    },
+    {
+      "epoch": 0.22336,
+      "grad_norm": 0.5396224856376648,
+      "learning_rate": 0.000905165496599406,
+      "loss": 2.953,
+      "step": 1745
+    },
+    {
+      "epoch": 0.223488,
+      "grad_norm": 0.5176796913146973,
+      "learning_rate": 0.0009050439831793726,
+      "loss": 2.9984,
+      "step": 1746
+    },
+    {
+      "epoch": 0.223616,
+      "grad_norm": 0.5469228029251099,
+      "learning_rate": 0.0009049224001275825,
+      "loss": 2.981,
+      "step": 1747
+    },
+    {
+      "epoch": 0.223744,
+      "grad_norm": 0.5127195119857788,
+      "learning_rate": 0.0009048007474649377,
+      "loss": 2.9585,
+      "step": 1748
+    },
+    {
+      "epoch": 0.223872,
+      "grad_norm": 0.541542112827301,
+      "learning_rate": 0.0009046790252123514,
+      "loss": 2.922,
+      "step": 1749
+    },
+    {
+      "epoch": 0.224,
+      "grad_norm": 0.5639825463294983,
+      "learning_rate": 0.0009045572333907495,
+      "loss": 3.0384,
+      "step": 1750
+    },
+    {
+      "epoch": 0.224128,
+      "grad_norm": 0.547157347202301,
+      "learning_rate": 0.000904435372021069,
+      "loss": 2.9291,
+      "step": 1751
+    },
+    {
+      "epoch": 0.224256,
+      "grad_norm": 0.5881614089012146,
+      "learning_rate": 0.0009043134411242593,
+      "loss": 2.9044,
+      "step": 1752
+    },
+    {
+      "epoch": 0.224384,
+      "grad_norm": 0.5653748512268066,
+      "learning_rate": 0.0009041914407212821,
+      "loss": 3.0287,
+      "step": 1753
+    },
+    {
+      "epoch": 0.224512,
+      "grad_norm": 0.5357649922370911,
+      "learning_rate": 0.0009040693708331103,
+      "loss": 2.8602,
+      "step": 1754
+    },
+    {
+      "epoch": 0.22464,
+      "grad_norm": 0.5618146061897278,
+      "learning_rate": 0.0009039472314807293,
+      "loss": 2.8957,
+      "step": 1755
+    },
+    {
+      "epoch": 0.224768,
+      "grad_norm": 0.5773375630378723,
+      "learning_rate": 0.0009038250226851362,
+      "loss": 2.9793,
+      "step": 1756
+    },
+    {
+      "epoch": 0.224896,
+      "grad_norm": 0.47514912486076355,
+      "learning_rate": 0.0009037027444673402,
+      "loss": 2.913,
+      "step": 1757
+    },
+    {
+      "epoch": 0.225024,
+      "grad_norm": 0.5318259596824646,
+      "learning_rate": 0.0009035803968483625,
+      "loss": 3.0846,
+      "step": 1758
+    },
+    {
+      "epoch": 0.225152,
+      "grad_norm": 0.5315955877304077,
+      "learning_rate": 0.0009034579798492356,
+      "loss": 2.9231,
+      "step": 1759
+    },
+    {
+      "epoch": 0.22528,
+      "grad_norm": 0.556702196598053,
+      "learning_rate": 0.0009033354934910048,
+      "loss": 3.0244,
+      "step": 1760
+    },
+    {
+      "epoch": 0.225408,
+      "grad_norm": 0.5497405529022217,
+      "learning_rate": 0.0009032129377947267,
+      "loss": 2.9743,
+      "step": 1761
+    },
+    {
+      "epoch": 0.225536,
+      "grad_norm": 0.5288832187652588,
+      "learning_rate": 0.0009030903127814703,
+      "loss": 2.8476,
+      "step": 1762
+    },
+    {
+      "epoch": 0.225664,
+      "grad_norm": 0.5570463538169861,
+      "learning_rate": 0.0009029676184723161,
+      "loss": 2.9255,
+      "step": 1763
+    },
+    {
+      "epoch": 0.225792,
+      "grad_norm": 0.557093620300293,
+      "learning_rate": 0.0009028448548883566,
+      "loss": 2.9607,
+      "step": 1764
+    },
+    {
+      "epoch": 0.22592,
+      "grad_norm": 0.5550035834312439,
+      "learning_rate": 0.0009027220220506965,
+      "loss": 2.7967,
+      "step": 1765
+    },
+    {
+      "epoch": 0.226048,
+      "grad_norm": 0.5749984383583069,
+      "learning_rate": 0.0009025991199804517,
+      "loss": 2.9339,
+      "step": 1766
+    },
+    {
+      "epoch": 0.226176,
+      "grad_norm": 0.6294952034950256,
+      "learning_rate": 0.0009024761486987512,
+      "loss": 2.9732,
+      "step": 1767
+    },
+    {
+      "epoch": 0.226304,
+      "grad_norm": 0.5389283299446106,
+      "learning_rate": 0.0009023531082267347,
+      "loss": 2.8906,
+      "step": 1768
+    },
+    {
+      "epoch": 0.226432,
+      "grad_norm": 0.533857524394989,
+      "learning_rate": 0.0009022299985855545,
+      "loss": 2.9885,
+      "step": 1769
+    },
+    {
+      "epoch": 0.22656,
+      "grad_norm": 0.5407208800315857,
+      "learning_rate": 0.0009021068197963744,
+      "loss": 2.8719,
+      "step": 1770
+    },
+    {
+      "epoch": 0.226688,
+      "grad_norm": 0.5447088479995728,
+      "learning_rate": 0.0009019835718803704,
+      "loss": 2.8844,
+      "step": 1771
+    },
+    {
+      "epoch": 0.226816,
+      "grad_norm": 0.524085283279419,
+      "learning_rate": 0.0009018602548587302,
+      "loss": 2.9044,
+      "step": 1772
+    },
+    {
+      "epoch": 0.226944,
+      "grad_norm": 0.5438936352729797,
+      "learning_rate": 0.0009017368687526535,
+      "loss": 2.8655,
+      "step": 1773
+    },
+    {
+      "epoch": 0.227072,
+      "grad_norm": 0.5245216488838196,
+      "learning_rate": 0.0009016134135833516,
+      "loss": 2.8648,
+      "step": 1774
+    },
+    {
+      "epoch": 0.2272,
+      "grad_norm": 0.5462217926979065,
+      "learning_rate": 0.000901489889372048,
+      "loss": 3.0287,
+      "step": 1775
+    },
+    {
+      "epoch": 0.227328,
+      "grad_norm": 0.5281918048858643,
+      "learning_rate": 0.000901366296139978,
+      "loss": 2.9909,
+      "step": 1776
+    },
+    {
+      "epoch": 0.227456,
+      "grad_norm": 0.5556467771530151,
+      "learning_rate": 0.0009012426339083887,
+      "loss": 2.9795,
+      "step": 1777
+    },
+    {
+      "epoch": 0.227584,
+      "grad_norm": 0.568077802658081,
+      "learning_rate": 0.0009011189026985389,
+      "loss": 2.9336,
+      "step": 1778
+    },
+    {
+      "epoch": 0.227712,
+      "grad_norm": 0.6372084617614746,
+      "learning_rate": 0.0009009951025316997,
+      "loss": 2.8643,
+      "step": 1779
+    },
+    {
+      "epoch": 0.22784,
+      "grad_norm": 0.5367746949195862,
+      "learning_rate": 0.0009008712334291535,
+      "loss": 3.0281,
+      "step": 1780
+    },
+    {
+      "epoch": 0.227968,
+      "grad_norm": 0.5598137974739075,
+      "learning_rate": 0.0009007472954121952,
+      "loss": 2.9637,
+      "step": 1781
+    },
+    {
+      "epoch": 0.228096,
+      "grad_norm": 0.5241551399230957,
+      "learning_rate": 0.0009006232885021308,
+      "loss": 3.0091,
+      "step": 1782
+    },
+    {
+      "epoch": 0.228224,
+      "grad_norm": 0.5413448810577393,
+      "learning_rate": 0.0009004992127202787,
+      "loss": 2.9503,
+      "step": 1783
+    },
+    {
+      "epoch": 0.228352,
+      "grad_norm": 0.5281667113304138,
+      "learning_rate": 0.0009003750680879689,
+      "loss": 2.903,
+      "step": 1784
+    },
+    {
+      "epoch": 0.22848,
+      "grad_norm": 0.535030722618103,
+      "learning_rate": 0.0009002508546265434,
+      "loss": 2.9541,
+      "step": 1785
+    },
+    {
+      "epoch": 0.228608,
+      "grad_norm": 0.5041018128395081,
+      "learning_rate": 0.0009001265723573559,
+      "loss": 2.983,
+      "step": 1786
+    },
+    {
+      "epoch": 0.228736,
+      "grad_norm": 0.5259836912155151,
+      "learning_rate": 0.0009000022213017719,
+      "loss": 2.9208,
+      "step": 1787
+    },
+    {
+      "epoch": 0.228864,
+      "grad_norm": 0.5219681262969971,
+      "learning_rate": 0.0008998778014811689,
+      "loss": 3.0167,
+      "step": 1788
+    },
+    {
+      "epoch": 0.228992,
+      "grad_norm": 0.5274616479873657,
+      "learning_rate": 0.000899753312916936,
+      "loss": 2.9523,
+      "step": 1789
+    },
+    {
+      "epoch": 0.22912,
+      "grad_norm": 0.5145535469055176,
+      "learning_rate": 0.0008996287556304743,
+      "loss": 2.9546,
+      "step": 1790
+    },
+    {
+      "epoch": 0.229248,
+      "grad_norm": 0.5169640779495239,
+      "learning_rate": 0.0008995041296431965,
+      "loss": 2.8893,
+      "step": 1791
+    },
+    {
+      "epoch": 0.229376,
+      "grad_norm": 0.5476436018943787,
+      "learning_rate": 0.0008993794349765274,
+      "loss": 2.8966,
+      "step": 1792
+    },
+    {
+      "epoch": 0.229504,
+      "grad_norm": 0.5510210394859314,
+      "learning_rate": 0.0008992546716519034,
+      "loss": 2.9488,
+      "step": 1793
+    },
+    {
+      "epoch": 0.229632,
+      "grad_norm": 0.5487595200538635,
+      "learning_rate": 0.000899129839690773,
+      "loss": 2.9846,
+      "step": 1794
+    },
+    {
+      "epoch": 0.22976,
+      "grad_norm": 0.5857563018798828,
+      "learning_rate": 0.0008990049391145959,
+      "loss": 3.1314,
+      "step": 1795
+    },
+    {
+      "epoch": 0.229888,
+      "grad_norm": 0.5296247601509094,
+      "learning_rate": 0.0008988799699448441,
+      "loss": 2.9617,
+      "step": 1796
+    },
+    {
+      "epoch": 0.230016,
+      "grad_norm": 0.5183902382850647,
+      "learning_rate": 0.0008987549322030013,
+      "loss": 2.8384,
+      "step": 1797
+    },
+    {
+      "epoch": 0.230144,
+      "grad_norm": 0.5463996529579163,
+      "learning_rate": 0.0008986298259105629,
+      "loss": 2.9498,
+      "step": 1798
+    },
+    {
+      "epoch": 0.230272,
+      "grad_norm": 0.5053161382675171,
+      "learning_rate": 0.000898504651089036,
+      "loss": 2.8687,
+      "step": 1799
+    },
+    {
+      "epoch": 0.2304,
+      "grad_norm": 0.5879713892936707,
+      "learning_rate": 0.0008983794077599398,
+      "loss": 3.0212,
+      "step": 1800
+    },
+    {
+      "epoch": 0.230528,
+      "grad_norm": 0.5704714059829712,
+      "learning_rate": 0.0008982540959448049,
+      "loss": 2.8572,
+      "step": 1801
+    },
+    {
+      "epoch": 0.230656,
+      "grad_norm": 0.5492478609085083,
+      "learning_rate": 0.000898128715665174,
+      "loss": 2.9512,
+      "step": 1802
+    },
+    {
+      "epoch": 0.230784,
+      "grad_norm": 0.5554349422454834,
+      "learning_rate": 0.0008980032669426015,
+      "loss": 3.0218,
+      "step": 1803
+    },
+    {
+      "epoch": 0.230912,
+      "grad_norm": 0.5691182613372803,
+      "learning_rate": 0.0008978777497986532,
+      "loss": 2.8977,
+      "step": 1804
+    },
+    {
+      "epoch": 0.23104,
+      "grad_norm": 0.5251091718673706,
+      "learning_rate": 0.0008977521642549072,
+      "loss": 2.8623,
+      "step": 1805
+    },
+    {
+      "epoch": 0.231168,
+      "grad_norm": 0.5471358299255371,
+      "learning_rate": 0.0008976265103329531,
+      "loss": 3.0253,
+      "step": 1806
+    },
+    {
+      "epoch": 0.231296,
+      "grad_norm": 0.528927743434906,
+      "learning_rate": 0.0008975007880543921,
+      "loss": 2.9924,
+      "step": 1807
+    },
+    {
+      "epoch": 0.231424,
+      "grad_norm": 0.5360991358757019,
+      "learning_rate": 0.0008973749974408374,
+      "loss": 3.0168,
+      "step": 1808
+    },
+    {
+      "epoch": 0.231552,
+      "grad_norm": 0.646999716758728,
+      "learning_rate": 0.0008972491385139138,
+      "loss": 2.9111,
+      "step": 1809
+    },
+    {
+      "epoch": 0.23168,
+      "grad_norm": 0.5610525012016296,
+      "learning_rate": 0.0008971232112952581,
+      "loss": 2.9382,
+      "step": 1810
+    },
+    {
+      "epoch": 0.231808,
+      "grad_norm": 0.6139883399009705,
+      "learning_rate": 0.0008969972158065185,
+      "loss": 2.9849,
+      "step": 1811
+    },
+    {
+      "epoch": 0.231936,
+      "grad_norm": 0.5908984541893005,
+      "learning_rate": 0.0008968711520693551,
+      "loss": 2.7904,
+      "step": 1812
+    },
+    {
+      "epoch": 0.232064,
+      "grad_norm": 0.6074075698852539,
+      "learning_rate": 0.0008967450201054397,
+      "loss": 2.8463,
+      "step": 1813
+    },
+    {
+      "epoch": 0.232192,
+      "grad_norm": 0.5473753213882446,
+      "learning_rate": 0.000896618819936456,
+      "loss": 3.045,
+      "step": 1814
+    },
+    {
+      "epoch": 0.23232,
+      "grad_norm": 0.5228095650672913,
+      "learning_rate": 0.0008964925515840991,
+      "loss": 2.923,
+      "step": 1815
+    },
+    {
+      "epoch": 0.232448,
+      "grad_norm": 0.5399818420410156,
+      "learning_rate": 0.000896366215070076,
+      "loss": 2.8543,
+      "step": 1816
+    },
+    {
+      "epoch": 0.232576,
+      "grad_norm": 0.6007784605026245,
+      "learning_rate": 0.0008962398104161054,
+      "loss": 2.9237,
+      "step": 1817
+    },
+    {
+      "epoch": 0.232704,
+      "grad_norm": 0.5462977886199951,
+      "learning_rate": 0.0008961133376439178,
+      "loss": 2.9024,
+      "step": 1818
+    },
+    {
+      "epoch": 0.232832,
+      "grad_norm": 0.5432478785514832,
+      "learning_rate": 0.0008959867967752556,
+      "loss": 2.9592,
+      "step": 1819
+    },
+    {
+      "epoch": 0.23296,
+      "grad_norm": 0.500706672668457,
+      "learning_rate": 0.000895860187831872,
+      "loss": 2.9519,
+      "step": 1820
+    },
+    {
+      "epoch": 0.233088,
+      "grad_norm": 0.500938892364502,
+      "learning_rate": 0.0008957335108355331,
+      "loss": 2.8988,
+      "step": 1821
+    },
+    {
+      "epoch": 0.233216,
+      "grad_norm": 0.5368850231170654,
+      "learning_rate": 0.0008956067658080158,
+      "loss": 2.886,
+      "step": 1822
+    },
+    {
+      "epoch": 0.233344,
+      "grad_norm": 0.5447342991828918,
+      "learning_rate": 0.0008954799527711094,
+      "loss": 2.9087,
+      "step": 1823
+    },
+    {
+      "epoch": 0.233472,
+      "grad_norm": 0.551532506942749,
+      "learning_rate": 0.0008953530717466142,
+      "loss": 2.8604,
+      "step": 1824
+    },
+    {
+      "epoch": 0.2336,
+      "grad_norm": 0.5507795810699463,
+      "learning_rate": 0.0008952261227563426,
+      "loss": 2.8614,
+      "step": 1825
+    },
+    {
+      "epoch": 0.233728,
+      "grad_norm": 0.5282488465309143,
+      "learning_rate": 0.0008950991058221188,
+      "loss": 3.0387,
+      "step": 1826
+    },
+    {
+      "epoch": 0.233856,
+      "grad_norm": 0.5533960461616516,
+      "learning_rate": 0.0008949720209657781,
+      "loss": 2.9406,
+      "step": 1827
+    },
+    {
+      "epoch": 0.233984,
+      "grad_norm": 0.5354134440422058,
+      "learning_rate": 0.0008948448682091683,
+      "loss": 3.0007,
+      "step": 1828
+    },
+    {
+      "epoch": 0.234112,
+      "grad_norm": 0.5428692102432251,
+      "learning_rate": 0.0008947176475741481,
+      "loss": 2.9122,
+      "step": 1829
+    },
+    {
+      "epoch": 0.23424,
+      "grad_norm": 0.49443745613098145,
+      "learning_rate": 0.0008945903590825883,
+      "loss": 2.9129,
+      "step": 1830
+    },
+    {
+      "epoch": 0.234368,
+      "grad_norm": 0.5412538647651672,
+      "learning_rate": 0.0008944630027563714,
+      "loss": 3.0975,
+      "step": 1831
+    },
+    {
+      "epoch": 0.234496,
+      "grad_norm": 0.5399856567382812,
+      "learning_rate": 0.0008943355786173912,
+      "loss": 3.0398,
+      "step": 1832
+    },
+    {
+      "epoch": 0.234624,
+      "grad_norm": 0.5044770836830139,
+      "learning_rate": 0.0008942080866875537,
+      "loss": 2.858,
+      "step": 1833
+    },
+    {
+      "epoch": 0.234752,
+      "grad_norm": 0.510504961013794,
+      "learning_rate": 0.000894080526988776,
+      "loss": 3.0583,
+      "step": 1834
+    },
+    {
+      "epoch": 0.23488,
+      "grad_norm": 0.5240353345870972,
+      "learning_rate": 0.0008939528995429869,
+      "loss": 2.9771,
+      "step": 1835
+    },
+    {
+      "epoch": 0.235008,
+      "grad_norm": 0.5510261058807373,
+      "learning_rate": 0.0008938252043721274,
+      "loss": 2.849,
+      "step": 1836
+    },
+    {
+      "epoch": 0.235136,
+      "grad_norm": 0.507845401763916,
+      "learning_rate": 0.0008936974414981498,
+      "loss": 2.9869,
+      "step": 1837
+    },
+    {
+      "epoch": 0.235264,
+      "grad_norm": 0.522064208984375,
+      "learning_rate": 0.0008935696109430177,
+      "loss": 2.9064,
+      "step": 1838
+    },
+    {
+      "epoch": 0.235392,
+      "grad_norm": 0.5701350569725037,
+      "learning_rate": 0.0008934417127287069,
+      "loss": 3.0222,
+      "step": 1839
+    },
+    {
+      "epoch": 0.23552,
+      "grad_norm": 0.5841301083564758,
+      "learning_rate": 0.0008933137468772046,
+      "loss": 2.9281,
+      "step": 1840
+    },
+    {
+      "epoch": 0.235648,
+      "grad_norm": 0.5277387499809265,
+      "learning_rate": 0.0008931857134105093,
+      "loss": 2.9248,
+      "step": 1841
+    },
+    {
+      "epoch": 0.235776,
+      "grad_norm": 0.5638837814331055,
+      "learning_rate": 0.0008930576123506317,
+      "loss": 2.8546,
+      "step": 1842
+    },
+    {
+      "epoch": 0.235904,
+      "grad_norm": 0.5322207808494568,
+      "learning_rate": 0.000892929443719594,
+      "loss": 2.9986,
+      "step": 1843
+    },
+    {
+      "epoch": 0.236032,
+      "grad_norm": 0.5209859609603882,
+      "learning_rate": 0.0008928012075394296,
+      "loss": 2.7706,
+      "step": 1844
+    },
+    {
+      "epoch": 0.23616,
+      "grad_norm": 0.5709221959114075,
+      "learning_rate": 0.000892672903832184,
+      "loss": 2.8702,
+      "step": 1845
+    },
+    {
+      "epoch": 0.236288,
+      "grad_norm": 0.5095044374465942,
+      "learning_rate": 0.0008925445326199138,
+      "loss": 2.9225,
+      "step": 1846
+    },
+    {
+      "epoch": 0.236416,
+      "grad_norm": 0.5157075524330139,
+      "learning_rate": 0.0008924160939246877,
+      "loss": 2.8939,
+      "step": 1847
+    },
+    {
+      "epoch": 0.236544,
+      "grad_norm": 0.5400156378746033,
+      "learning_rate": 0.0008922875877685858,
+      "loss": 2.9944,
+      "step": 1848
+    },
+    {
+      "epoch": 0.236672,
+      "grad_norm": 0.5252795815467834,
+      "learning_rate": 0.0008921590141737,
+      "loss": 2.9079,
+      "step": 1849
+    },
+    {
+      "epoch": 0.2368,
+      "grad_norm": 0.5527716875076294,
+      "learning_rate": 0.0008920303731621331,
+      "loss": 2.8956,
+      "step": 1850
+    },
+    {
+      "epoch": 0.236928,
+      "grad_norm": 0.5234727263450623,
+      "learning_rate": 0.0008919016647560002,
+      "loss": 3.0567,
+      "step": 1851
+    },
+    {
+      "epoch": 0.237056,
+      "grad_norm": 0.5504688620567322,
+      "learning_rate": 0.0008917728889774279,
+      "loss": 2.9181,
+      "step": 1852
+    },
+    {
+      "epoch": 0.237184,
+      "grad_norm": 0.5643249154090881,
+      "learning_rate": 0.0008916440458485542,
+      "loss": 3.0306,
+      "step": 1853
+    },
+    {
+      "epoch": 0.237312,
+      "grad_norm": 0.5127208828926086,
+      "learning_rate": 0.0008915151353915288,
+      "loss": 2.9868,
+      "step": 1854
+    },
+    {
+      "epoch": 0.23744,
+      "grad_norm": 0.48380428552627563,
+      "learning_rate": 0.0008913861576285127,
+      "loss": 2.8613,
+      "step": 1855
+    },
+    {
+      "epoch": 0.237568,
+      "grad_norm": 0.4982568919658661,
+      "learning_rate": 0.0008912571125816786,
+      "loss": 2.9621,
+      "step": 1856
+    },
+    {
+      "epoch": 0.237696,
+      "grad_norm": 0.5077729821205139,
+      "learning_rate": 0.0008911280002732112,
+      "loss": 3.0516,
+      "step": 1857
+    },
+    {
+      "epoch": 0.237824,
+      "grad_norm": 0.4985038638114929,
+      "learning_rate": 0.0008909988207253061,
+      "loss": 3.0933,
+      "step": 1858
+    },
+    {
+      "epoch": 0.237952,
+      "grad_norm": 0.49355971813201904,
+      "learning_rate": 0.0008908695739601709,
+      "loss": 2.9216,
+      "step": 1859
+    },
+    {
+      "epoch": 0.23808,
+      "grad_norm": 0.5409252643585205,
+      "learning_rate": 0.0008907402600000248,
+      "loss": 2.9419,
+      "step": 1860
+    },
+    {
+      "epoch": 0.238208,
+      "grad_norm": 0.5241426229476929,
+      "learning_rate": 0.0008906108788670979,
+      "loss": 2.9487,
+      "step": 1861
+    },
+    {
+      "epoch": 0.238336,
+      "grad_norm": 0.5749977827072144,
+      "learning_rate": 0.0008904814305836326,
+      "loss": 2.8493,
+      "step": 1862
+    },
+    {
+      "epoch": 0.238464,
+      "grad_norm": 0.6001923084259033,
+      "learning_rate": 0.0008903519151718826,
+      "loss": 2.8958,
+      "step": 1863
+    },
+    {
+      "epoch": 0.238592,
+      "grad_norm": 0.5242201685905457,
+      "learning_rate": 0.000890222332654113,
+      "loss": 2.9161,
+      "step": 1864
+    },
+    {
+      "epoch": 0.23872,
+      "grad_norm": 0.5513917207717896,
+      "learning_rate": 0.0008900926830526005,
+      "loss": 3.0275,
+      "step": 1865
+    },
+    {
+      "epoch": 0.238848,
+      "grad_norm": 0.6036017537117004,
+      "learning_rate": 0.0008899629663896336,
+      "loss": 2.9521,
+      "step": 1866
+    },
+    {
+      "epoch": 0.238976,
+      "grad_norm": 0.5910593271255493,
+      "learning_rate": 0.0008898331826875117,
+      "loss": 3.1023,
+      "step": 1867
+    },
+    {
+      "epoch": 0.239104,
+      "grad_norm": 0.49502435326576233,
+      "learning_rate": 0.0008897033319685466,
+      "loss": 2.9171,
+      "step": 1868
+    },
+    {
+      "epoch": 0.239232,
+      "grad_norm": 0.5122207403182983,
+      "learning_rate": 0.0008895734142550609,
+      "loss": 2.7822,
+      "step": 1869
+    },
+    {
+      "epoch": 0.23936,
+      "grad_norm": 0.527217447757721,
+      "learning_rate": 0.0008894434295693888,
+      "loss": 2.8972,
+      "step": 1870
+    },
+    {
+      "epoch": 0.239488,
+      "grad_norm": 0.5834856629371643,
+      "learning_rate": 0.0008893133779338764,
+      "loss": 2.9882,
+      "step": 1871
+    },
+    {
+      "epoch": 0.239616,
+      "grad_norm": 0.5820335149765015,
+      "learning_rate": 0.0008891832593708811,
+      "loss": 2.9422,
+      "step": 1872
+    },
+    {
+      "epoch": 0.239744,
+      "grad_norm": 0.5474259853363037,
+      "learning_rate": 0.0008890530739027717,
+      "loss": 2.9707,
+      "step": 1873
+    },
+    {
+      "epoch": 0.239872,
+      "grad_norm": 0.5736411809921265,
+      "learning_rate": 0.0008889228215519285,
+      "loss": 3.0353,
+      "step": 1874
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5455046892166138,
+      "learning_rate": 0.0008887925023407437,
+      "loss": 2.8237,
+      "step": 1875
+    },
+    {
+      "epoch": 0.240128,
+      "grad_norm": 0.5274345278739929,
+      "learning_rate": 0.0008886621162916203,
+      "loss": 2.8819,
+      "step": 1876
+    },
+    {
+      "epoch": 0.240256,
+      "grad_norm": 0.5909514427185059,
+      "learning_rate": 0.0008885316634269733,
+      "loss": 2.9299,
+      "step": 1877
+    },
+    {
+      "epoch": 0.240384,
+      "grad_norm": 0.6402825713157654,
+      "learning_rate": 0.0008884011437692294,
+      "loss": 3.0493,
+      "step": 1878
+    },
+    {
+      "epoch": 0.240512,
+      "grad_norm": 0.5250142812728882,
+      "learning_rate": 0.0008882705573408259,
+      "loss": 2.9337,
+      "step": 1879
+    },
+    {
+      "epoch": 0.24064,
+      "grad_norm": 0.5297059416770935,
+      "learning_rate": 0.0008881399041642124,
+      "loss": 2.8821,
+      "step": 1880
+    },
+    {
+      "epoch": 0.240768,
+      "grad_norm": 0.5268275141716003,
+      "learning_rate": 0.0008880091842618497,
+      "loss": 2.8838,
+      "step": 1881
+    },
+    {
+      "epoch": 0.240896,
+      "grad_norm": 0.5360270142555237,
+      "learning_rate": 0.0008878783976562102,
+      "loss": 3.0412,
+      "step": 1882
+    },
+    {
+      "epoch": 0.241024,
+      "grad_norm": 0.51416015625,
+      "learning_rate": 0.0008877475443697773,
+      "loss": 2.923,
+      "step": 1883
+    },
+    {
+      "epoch": 0.241152,
+      "grad_norm": 0.5579512715339661,
+      "learning_rate": 0.0008876166244250463,
+      "loss": 2.8322,
+      "step": 1884
+    },
+    {
+      "epoch": 0.24128,
+      "grad_norm": 0.5464280843734741,
+      "learning_rate": 0.0008874856378445241,
+      "loss": 2.8788,
+      "step": 1885
+    },
+    {
+      "epoch": 0.241408,
+      "grad_norm": 0.5634199380874634,
+      "learning_rate": 0.0008873545846507285,
+      "loss": 3.1176,
+      "step": 1886
+    },
+    {
+      "epoch": 0.241536,
+      "grad_norm": 0.5769438147544861,
+      "learning_rate": 0.0008872234648661892,
+      "loss": 2.9057,
+      "step": 1887
+    },
+    {
+      "epoch": 0.241664,
+      "grad_norm": 0.5444268584251404,
+      "learning_rate": 0.0008870922785134472,
+      "loss": 3.1557,
+      "step": 1888
+    },
+    {
+      "epoch": 0.241792,
+      "grad_norm": 0.5009043216705322,
+      "learning_rate": 0.0008869610256150551,
+      "loss": 2.9085,
+      "step": 1889
+    },
+    {
+      "epoch": 0.24192,
+      "grad_norm": 0.5056285262107849,
+      "learning_rate": 0.0008868297061935766,
+      "loss": 2.7998,
+      "step": 1890
+    },
+    {
+      "epoch": 0.242048,
+      "grad_norm": 0.5120295286178589,
+      "learning_rate": 0.000886698320271587,
+      "loss": 2.914,
+      "step": 1891
+    },
+    {
+      "epoch": 0.242176,
+      "grad_norm": 0.47762757539749146,
+      "learning_rate": 0.0008865668678716734,
+      "loss": 2.8344,
+      "step": 1892
+    },
+    {
+      "epoch": 0.242304,
+      "grad_norm": 0.492083340883255,
+      "learning_rate": 0.0008864353490164335,
+      "loss": 3.0136,
+      "step": 1893
+    },
+    {
+      "epoch": 0.242432,
+      "grad_norm": 0.5220906734466553,
+      "learning_rate": 0.0008863037637284772,
+      "loss": 2.8524,
+      "step": 1894
+    },
+    {
+      "epoch": 0.24256,
+      "grad_norm": 0.5517160892486572,
+      "learning_rate": 0.0008861721120304256,
+      "loss": 2.9492,
+      "step": 1895
+    },
+    {
+      "epoch": 0.242688,
+      "grad_norm": 0.5325133800506592,
+      "learning_rate": 0.0008860403939449109,
+      "loss": 2.9565,
+      "step": 1896
+    },
+    {
+      "epoch": 0.242816,
+      "grad_norm": 0.5263796448707581,
+      "learning_rate": 0.0008859086094945772,
+      "loss": 2.979,
+      "step": 1897
+    },
+    {
+      "epoch": 0.242944,
+      "grad_norm": 0.4817315340042114,
+      "learning_rate": 0.0008857767587020798,
+      "loss": 2.8681,
+      "step": 1898
+    },
+    {
+      "epoch": 0.243072,
+      "grad_norm": 0.5058117508888245,
+      "learning_rate": 0.000885644841590085,
+      "loss": 2.9225,
+      "step": 1899
+    },
+    {
+      "epoch": 0.2432,
+      "grad_norm": 0.5333071351051331,
+      "learning_rate": 0.0008855128581812714,
+      "loss": 2.9241,
+      "step": 1900
+    },
+    {
+      "epoch": 0.243328,
+      "grad_norm": 0.487117201089859,
+      "learning_rate": 0.0008853808084983281,
+      "loss": 2.9851,
+      "step": 1901
+    },
+    {
+      "epoch": 0.243456,
+      "grad_norm": 0.5410962104797363,
+      "learning_rate": 0.0008852486925639561,
+      "loss": 3.0303,
+      "step": 1902
+    },
+    {
+      "epoch": 0.243584,
+      "grad_norm": 0.4850640296936035,
+      "learning_rate": 0.0008851165104008679,
+      "loss": 2.8341,
+      "step": 1903
+    },
+    {
+      "epoch": 0.243712,
+      "grad_norm": 0.5050856471061707,
+      "learning_rate": 0.0008849842620317865,
+      "loss": 2.7905,
+      "step": 1904
+    },
+    {
+      "epoch": 0.24384,
+      "grad_norm": 0.48446694016456604,
+      "learning_rate": 0.0008848519474794476,
+      "loss": 2.9594,
+      "step": 1905
+    },
+    {
+      "epoch": 0.243968,
+      "grad_norm": 0.5225907564163208,
+      "learning_rate": 0.0008847195667665973,
+      "loss": 2.9675,
+      "step": 1906
+    },
+    {
+      "epoch": 0.244096,
+      "grad_norm": 0.5080874562263489,
+      "learning_rate": 0.0008845871199159934,
+      "loss": 2.841,
+      "step": 1907
+    },
+    {
+      "epoch": 0.244224,
+      "grad_norm": 0.5296223163604736,
+      "learning_rate": 0.0008844546069504051,
+      "loss": 2.9658,
+      "step": 1908
+    },
+    {
+      "epoch": 0.244352,
+      "grad_norm": 0.5384196639060974,
+      "learning_rate": 0.0008843220278926128,
+      "loss": 2.7634,
+      "step": 1909
+    },
+    {
+      "epoch": 0.24448,
+      "grad_norm": 0.5314280986785889,
+      "learning_rate": 0.0008841893827654086,
+      "loss": 2.863,
+      "step": 1910
+    },
+    {
+      "epoch": 0.244608,
+      "grad_norm": 0.5369045734405518,
+      "learning_rate": 0.0008840566715915955,
+      "loss": 2.817,
+      "step": 1911
+    },
+    {
+      "epoch": 0.244736,
+      "grad_norm": 0.4814787805080414,
+      "learning_rate": 0.0008839238943939882,
+      "loss": 2.783,
+      "step": 1912
+    },
+    {
+      "epoch": 0.244864,
+      "grad_norm": 0.5458283424377441,
+      "learning_rate": 0.0008837910511954127,
+      "loss": 2.9767,
+      "step": 1913
+    },
+    {
+      "epoch": 0.244992,
+      "grad_norm": 0.5649486184120178,
+      "learning_rate": 0.0008836581420187062,
+      "loss": 2.9581,
+      "step": 1914
+    },
+    {
+      "epoch": 0.24512,
+      "grad_norm": 0.5456324815750122,
+      "learning_rate": 0.0008835251668867172,
+      "loss": 3.1142,
+      "step": 1915
+    },
+    {
+      "epoch": 0.245248,
+      "grad_norm": 0.49204909801483154,
+      "learning_rate": 0.0008833921258223059,
+      "loss": 2.8113,
+      "step": 1916
+    },
+    {
+      "epoch": 0.245376,
+      "grad_norm": 0.5287555456161499,
+      "learning_rate": 0.0008832590188483436,
+      "loss": 2.7806,
+      "step": 1917
+    },
+    {
+      "epoch": 0.245504,
+      "grad_norm": 0.5136240124702454,
+      "learning_rate": 0.0008831258459877127,
+      "loss": 2.9233,
+      "step": 1918
+    },
+    {
+      "epoch": 0.245632,
+      "grad_norm": 0.5206447243690491,
+      "learning_rate": 0.0008829926072633074,
+      "loss": 2.93,
+      "step": 1919
+    },
+    {
+      "epoch": 0.24576,
+      "grad_norm": 0.5652109980583191,
+      "learning_rate": 0.0008828593026980327,
+      "loss": 2.9705,
+      "step": 1920
+    },
+    {
+      "epoch": 0.245888,
+      "grad_norm": 1.3820905685424805,
+      "learning_rate": 0.0008827259323148055,
+      "loss": 2.7169,
+      "step": 1921
+    },
+    {
+      "epoch": 0.246016,
+      "grad_norm": 0.548683226108551,
+      "learning_rate": 0.0008825924961365537,
+      "loss": 2.9285,
+      "step": 1922
+    },
+    {
+      "epoch": 0.246144,
+      "grad_norm": 0.5451627373695374,
+      "learning_rate": 0.0008824589941862163,
+      "loss": 2.8767,
+      "step": 1923
+    },
+    {
+      "epoch": 0.246272,
+      "grad_norm": 0.5414646863937378,
+      "learning_rate": 0.0008823254264867439,
+      "loss": 2.8697,
+      "step": 1924
+    },
+    {
+      "epoch": 0.2464,
+      "grad_norm": 0.5076912641525269,
+      "learning_rate": 0.0008821917930610985,
+      "loss": 3.0572,
+      "step": 1925
+    },
+    {
+      "epoch": 0.246528,
+      "grad_norm": 0.5330734848976135,
+      "learning_rate": 0.0008820580939322532,
+      "loss": 2.9932,
+      "step": 1926
+    },
+    {
+      "epoch": 0.246656,
+      "grad_norm": 0.49790504574775696,
+      "learning_rate": 0.0008819243291231922,
+      "loss": 2.8516,
+      "step": 1927
+    },
+    {
+      "epoch": 0.246784,
+      "grad_norm": 0.5462942719459534,
+      "learning_rate": 0.0008817904986569114,
+      "loss": 3.0776,
+      "step": 1928
+    },
+    {
+      "epoch": 0.246912,
+      "grad_norm": 0.5138339996337891,
+      "learning_rate": 0.0008816566025564179,
+      "loss": 2.9985,
+      "step": 1929
+    },
+    {
+      "epoch": 0.24704,
+      "grad_norm": 0.5643506050109863,
+      "learning_rate": 0.0008815226408447296,
+      "loss": 2.9941,
+      "step": 1930
+    },
+    {
+      "epoch": 0.247168,
+      "grad_norm": 0.5343838334083557,
+      "learning_rate": 0.0008813886135448767,
+      "loss": 2.8538,
+      "step": 1931
+    },
+    {
+      "epoch": 0.247296,
+      "grad_norm": 0.5316978693008423,
+      "learning_rate": 0.0008812545206798994,
+      "loss": 2.9981,
+      "step": 1932
+    },
+    {
+      "epoch": 0.247424,
+      "grad_norm": 0.5169166326522827,
+      "learning_rate": 0.0008811203622728503,
+      "loss": 2.9638,
+      "step": 1933
+    },
+    {
+      "epoch": 0.247552,
+      "grad_norm": 0.4998607039451599,
+      "learning_rate": 0.0008809861383467925,
+      "loss": 2.9571,
+      "step": 1934
+    },
+    {
+      "epoch": 0.24768,
+      "grad_norm": 0.5215280055999756,
+      "learning_rate": 0.0008808518489248009,
+      "loss": 2.9376,
+      "step": 1935
+    },
+    {
+      "epoch": 0.247808,
+      "grad_norm": 0.5398836135864258,
+      "learning_rate": 0.0008807174940299612,
+      "loss": 2.9423,
+      "step": 1936
+    },
+    {
+      "epoch": 0.247936,
+      "grad_norm": 0.49361884593963623,
+      "learning_rate": 0.0008805830736853705,
+      "loss": 2.7989,
+      "step": 1937
+    },
+    {
+      "epoch": 0.248064,
+      "grad_norm": 0.4572981297969818,
+      "learning_rate": 0.0008804485879141374,
+      "loss": 2.7494,
+      "step": 1938
+    },
+    {
+      "epoch": 0.248192,
+      "grad_norm": 0.48900774121284485,
+      "learning_rate": 0.0008803140367393815,
+      "loss": 2.9724,
+      "step": 1939
+    },
+    {
+      "epoch": 0.24832,
+      "grad_norm": 0.5258393883705139,
+      "learning_rate": 0.0008801794201842337,
+      "loss": 2.957,
+      "step": 1940
+    },
+    {
+      "epoch": 0.248448,
+      "grad_norm": 0.5480587482452393,
+      "learning_rate": 0.0008800447382718362,
+      "loss": 2.9387,
+      "step": 1941
+    },
+    {
+      "epoch": 0.248576,
+      "grad_norm": 0.5881758332252502,
+      "learning_rate": 0.0008799099910253423,
+      "loss": 2.8815,
+      "step": 1942
+    },
+    {
+      "epoch": 0.248704,
+      "grad_norm": 0.5309457182884216,
+      "learning_rate": 0.0008797751784679166,
+      "loss": 3.0339,
+      "step": 1943
+    },
+    {
+      "epoch": 0.248832,
+      "grad_norm": 0.5167729258537292,
+      "learning_rate": 0.000879640300622735,
+      "loss": 2.953,
+      "step": 1944
+    },
+    {
+      "epoch": 0.24896,
+      "grad_norm": 0.49267807602882385,
+      "learning_rate": 0.0008795053575129846,
+      "loss": 3.0513,
+      "step": 1945
+    },
+    {
+      "epoch": 0.249088,
+      "grad_norm": 0.5285442471504211,
+      "learning_rate": 0.0008793703491618637,
+      "loss": 2.915,
+      "step": 1946
+    },
+    {
+      "epoch": 0.249216,
+      "grad_norm": 0.5443438291549683,
+      "learning_rate": 0.0008792352755925817,
+      "loss": 2.9688,
+      "step": 1947
+    },
+    {
+      "epoch": 0.249344,
+      "grad_norm": 0.4722561240196228,
+      "learning_rate": 0.0008791001368283594,
+      "loss": 2.8028,
+      "step": 1948
+    },
+    {
+      "epoch": 0.249472,
+      "grad_norm": 0.5246741771697998,
+      "learning_rate": 0.0008789649328924285,
+      "loss": 2.998,
+      "step": 1949
+    },
+    {
+      "epoch": 0.2496,
+      "grad_norm": 0.495604544878006,
+      "learning_rate": 0.0008788296638080324,
+      "loss": 2.9391,
+      "step": 1950
+    },
+    {
+      "epoch": 0.249728,
+      "grad_norm": 0.49381253123283386,
+      "learning_rate": 0.0008786943295984253,
+      "loss": 2.9209,
+      "step": 1951
+    },
+    {
+      "epoch": 0.249856,
+      "grad_norm": 0.5202034711837769,
+      "learning_rate": 0.0008785589302868727,
+      "loss": 2.897,
+      "step": 1952
+    },
+    {
+      "epoch": 0.249984,
+      "grad_norm": 0.5506734251976013,
+      "learning_rate": 0.0008784234658966515,
+      "loss": 2.8379,
+      "step": 1953
+    },
+    {
+      "epoch": 0.250112,
+      "grad_norm": 0.5358812212944031,
+      "learning_rate": 0.0008782879364510494,
+      "loss": 2.8275,
+      "step": 1954
+    },
+    {
+      "epoch": 0.25024,
+      "grad_norm": 0.7368479371070862,
+      "learning_rate": 0.0008781523419733654,
+      "loss": 2.9778,
+      "step": 1955
+    },
+    {
+      "epoch": 0.250368,
+      "grad_norm": 0.5023691654205322,
+      "learning_rate": 0.00087801668248691,
+      "loss": 2.9582,
+      "step": 1956
+    },
+    {
+      "epoch": 0.250496,
+      "grad_norm": 0.5233824849128723,
+      "learning_rate": 0.0008778809580150045,
+      "loss": 2.9107,
+      "step": 1957
+    },
+    {
+      "epoch": 0.250624,
+      "grad_norm": 0.49735280871391296,
+      "learning_rate": 0.0008777451685809816,
+      "loss": 2.9536,
+      "step": 1958
+    },
+    {
+      "epoch": 0.250752,
+      "grad_norm": 0.4946553111076355,
+      "learning_rate": 0.0008776093142081849,
+      "loss": 2.9322,
+      "step": 1959
+    },
+    {
+      "epoch": 0.25088,
+      "grad_norm": 0.5046801567077637,
+      "learning_rate": 0.0008774733949199696,
+      "loss": 2.8833,
+      "step": 1960
+    },
+    {
+      "epoch": 0.251008,
+      "grad_norm": 0.5753873586654663,
+      "learning_rate": 0.0008773374107397016,
+      "loss": 2.9337,
+      "step": 1961
+    },
+    {
+      "epoch": 0.251136,
+      "grad_norm": 0.5156358480453491,
+      "learning_rate": 0.0008772013616907584,
+      "loss": 2.9122,
+      "step": 1962
+    },
+    {
+      "epoch": 0.251264,
+      "grad_norm": 0.5381879210472107,
+      "learning_rate": 0.000877065247796528,
+      "loss": 2.9116,
+      "step": 1963
+    },
+    {
+      "epoch": 0.251392,
+      "grad_norm": 0.5189423561096191,
+      "learning_rate": 0.0008769290690804103,
+      "loss": 2.9225,
+      "step": 1964
+    },
+    {
+      "epoch": 0.25152,
+      "grad_norm": 0.5379394292831421,
+      "learning_rate": 0.0008767928255658157,
+      "loss": 3.0813,
+      "step": 1965
+    },
+    {
+      "epoch": 0.251648,
+      "grad_norm": 0.5315592288970947,
+      "learning_rate": 0.0008766565172761664,
+      "loss": 3.034,
+      "step": 1966
+    },
+    {
+      "epoch": 0.251776,
+      "grad_norm": 0.5891066789627075,
+      "learning_rate": 0.0008765201442348954,
+      "loss": 2.886,
+      "step": 1967
+    },
+    {
+      "epoch": 0.251904,
+      "grad_norm": 0.6017338037490845,
+      "learning_rate": 0.0008763837064654463,
+      "loss": 2.9238,
+      "step": 1968
+    },
+    {
+      "epoch": 0.252032,
+      "grad_norm": 0.5427408218383789,
+      "learning_rate": 0.0008762472039912747,
+      "loss": 2.8885,
+      "step": 1969
+    },
+    {
+      "epoch": 0.25216,
+      "grad_norm": 0.5011466145515442,
+      "learning_rate": 0.0008761106368358469,
+      "loss": 2.9177,
+      "step": 1970
+    },
+    {
+      "epoch": 0.252288,
+      "grad_norm": 0.5592702627182007,
+      "learning_rate": 0.0008759740050226406,
+      "loss": 2.9021,
+      "step": 1971
+    },
+    {
+      "epoch": 0.252416,
+      "grad_norm": 0.5122145414352417,
+      "learning_rate": 0.0008758373085751439,
+      "loss": 2.8893,
+      "step": 1972
+    },
+    {
+      "epoch": 0.252544,
+      "grad_norm": 0.5265568494796753,
+      "learning_rate": 0.0008757005475168569,
+      "loss": 2.8896,
+      "step": 1973
+    },
+    {
+      "epoch": 0.252672,
+      "grad_norm": 0.5205619931221008,
+      "learning_rate": 0.0008755637218712902,
+      "loss": 2.939,
+      "step": 1974
+    },
+    {
+      "epoch": 0.2528,
+      "grad_norm": 0.5008439421653748,
+      "learning_rate": 0.0008754268316619659,
+      "loss": 2.9413,
+      "step": 1975
+    },
+    {
+      "epoch": 0.252928,
+      "grad_norm": 0.5466175675392151,
+      "learning_rate": 0.0008752898769124169,
+      "loss": 3.0466,
+      "step": 1976
+    },
+    {
+      "epoch": 0.253056,
+      "grad_norm": 0.5089293718338013,
+      "learning_rate": 0.0008751528576461873,
+      "loss": 3.0001,
+      "step": 1977
+    },
+    {
+      "epoch": 0.253184,
+      "grad_norm": 0.5147605538368225,
+      "learning_rate": 0.0008750157738868323,
+      "loss": 2.8358,
+      "step": 1978
+    },
+    {
+      "epoch": 0.253312,
+      "grad_norm": 0.48467934131622314,
+      "learning_rate": 0.0008748786256579182,
+      "loss": 2.9458,
+      "step": 1979
+    },
+    {
+      "epoch": 0.25344,
+      "grad_norm": 0.5093725919723511,
+      "learning_rate": 0.0008747414129830224,
+      "loss": 2.8782,
+      "step": 1980
+    },
+    {
+      "epoch": 0.253568,
+      "grad_norm": 0.5186693072319031,
+      "learning_rate": 0.0008746041358857334,
+      "loss": 2.9754,
+      "step": 1981
+    },
+    {
+      "epoch": 0.253696,
+      "grad_norm": 0.5057083964347839,
+      "learning_rate": 0.0008744667943896506,
+      "loss": 3.0624,
+      "step": 1982
+    },
+    {
+      "epoch": 0.253824,
+      "grad_norm": 0.5122122168540955,
+      "learning_rate": 0.0008743293885183848,
+      "loss": 2.8134,
+      "step": 1983
+    },
+    {
+      "epoch": 0.253952,
+      "grad_norm": 0.5324753522872925,
+      "learning_rate": 0.0008741919182955573,
+      "loss": 2.7863,
+      "step": 1984
+    },
+    {
+      "epoch": 0.25408,
+      "grad_norm": 0.549465537071228,
+      "learning_rate": 0.0008740543837448012,
+      "loss": 3.0443,
+      "step": 1985
+    },
+    {
+      "epoch": 0.254208,
+      "grad_norm": 0.5372951626777649,
+      "learning_rate": 0.00087391678488976,
+      "loss": 2.841,
+      "step": 1986
+    },
+    {
+      "epoch": 0.254336,
+      "grad_norm": 0.5399136543273926,
+      "learning_rate": 0.0008737791217540887,
+      "loss": 2.9134,
+      "step": 1987
+    },
+    {
+      "epoch": 0.254464,
+      "grad_norm": 0.5628214478492737,
+      "learning_rate": 0.0008736413943614532,
+      "loss": 2.8972,
+      "step": 1988
+    },
+    {
+      "epoch": 0.254592,
+      "grad_norm": 0.5166627764701843,
+      "learning_rate": 0.0008735036027355303,
+      "loss": 2.8489,
+      "step": 1989
+    },
+    {
+      "epoch": 0.25472,
+      "grad_norm": 0.5244573950767517,
+      "learning_rate": 0.0008733657469000081,
+      "loss": 3.0115,
+      "step": 1990
+    },
+    {
+      "epoch": 0.254848,
+      "grad_norm": 0.5829212665557861,
+      "learning_rate": 0.0008732278268785856,
+      "loss": 2.8593,
+      "step": 1991
+    },
+    {
+      "epoch": 0.254976,
+      "grad_norm": 0.5187703371047974,
+      "learning_rate": 0.0008730898426949727,
+      "loss": 2.8714,
+      "step": 1992
+    },
+    {
+      "epoch": 0.255104,
+      "grad_norm": 1.055019736289978,
+      "learning_rate": 0.0008729517943728909,
+      "loss": 2.9639,
+      "step": 1993
+    },
+    {
+      "epoch": 0.255232,
+      "grad_norm": 0.5738641619682312,
+      "learning_rate": 0.0008728136819360717,
+      "loss": 2.7896,
+      "step": 1994
+    },
+    {
+      "epoch": 0.25536,
+      "grad_norm": 0.5085809826850891,
+      "learning_rate": 0.0008726755054082588,
+      "loss": 2.9493,
+      "step": 1995
+    },
+    {
+      "epoch": 0.255488,
+      "grad_norm": 0.5190675258636475,
+      "learning_rate": 0.0008725372648132059,
+      "loss": 3.0332,
+      "step": 1996
+    },
+    {
+      "epoch": 0.255616,
+      "grad_norm": 0.5140447616577148,
+      "learning_rate": 0.0008723989601746784,
+      "loss": 2.8298,
+      "step": 1997
+    },
+    {
+      "epoch": 0.255744,
+      "grad_norm": 0.5125830173492432,
+      "learning_rate": 0.0008722605915164525,
+      "loss": 2.9325,
+      "step": 1998
+    },
+    {
+      "epoch": 0.255872,
+      "grad_norm": 0.48862674832344055,
+      "learning_rate": 0.0008721221588623154,
+      "loss": 2.9654,
+      "step": 1999
+    },
+    {
+      "epoch": 0.256,
+      "grad_norm": 0.5227726101875305,
+      "learning_rate": 0.000871983662236065,
+      "loss": 3.0293,
+      "step": 2000
+    },
+    {
+      "epoch": 0.256128,
+      "grad_norm": 0.5234312415122986,
+      "learning_rate": 0.0008718451016615108,
+      "loss": 2.9527,
+      "step": 2001
+    },
+    {
+      "epoch": 0.256256,
+      "grad_norm": 0.5035701990127563,
+      "learning_rate": 0.0008717064771624727,
+      "loss": 3.0308,
+      "step": 2002
+    },
+    {
+      "epoch": 0.256384,
+      "grad_norm": 0.5329359769821167,
+      "learning_rate": 0.000871567788762782,
+      "loss": 2.9756,
+      "step": 2003
+    },
+    {
+      "epoch": 0.256512,
+      "grad_norm": 0.5298618674278259,
+      "learning_rate": 0.000871429036486281,
+      "loss": 2.9492,
+      "step": 2004
+    },
+    {
+      "epoch": 0.25664,
+      "grad_norm": 0.5442545413970947,
+      "learning_rate": 0.0008712902203568225,
+      "loss": 2.8499,
+      "step": 2005
+    },
+    {
+      "epoch": 0.256768,
+      "grad_norm": 0.5055107474327087,
+      "learning_rate": 0.0008711513403982707,
+      "loss": 3.0339,
+      "step": 2006
+    },
+    {
+      "epoch": 0.256896,
+      "grad_norm": 0.5208272933959961,
+      "learning_rate": 0.0008710123966345007,
+      "loss": 2.9231,
+      "step": 2007
+    },
+    {
+      "epoch": 0.257024,
+      "grad_norm": 0.4805013835430145,
+      "learning_rate": 0.0008708733890893987,
+      "loss": 2.8668,
+      "step": 2008
+    },
+    {
+      "epoch": 0.257152,
+      "grad_norm": 0.4827541708946228,
+      "learning_rate": 0.0008707343177868617,
+      "loss": 2.9237,
+      "step": 2009
+    },
+    {
+      "epoch": 0.25728,
+      "grad_norm": 0.5196141004562378,
+      "learning_rate": 0.0008705951827507974,
+      "loss": 2.9336,
+      "step": 2010
+    },
+    {
+      "epoch": 0.257408,
+      "grad_norm": 0.48739907145500183,
+      "learning_rate": 0.0008704559840051248,
+      "loss": 2.8714,
+      "step": 2011
+    },
+    {
+      "epoch": 0.257536,
+      "grad_norm": 0.5339064002037048,
+      "learning_rate": 0.0008703167215737739,
+      "loss": 2.7789,
+      "step": 2012
+    },
+    {
+      "epoch": 0.257664,
+      "grad_norm": 0.5569707751274109,
+      "learning_rate": 0.0008701773954806855,
+      "loss": 2.9514,
+      "step": 2013
+    },
+    {
+      "epoch": 0.257792,
+      "grad_norm": 0.5323115587234497,
+      "learning_rate": 0.0008700380057498114,
+      "loss": 2.9166,
+      "step": 2014
+    },
+    {
+      "epoch": 0.25792,
+      "grad_norm": 0.5200791358947754,
+      "learning_rate": 0.0008698985524051143,
+      "loss": 2.8797,
+      "step": 2015
+    },
+    {
+      "epoch": 0.258048,
+      "grad_norm": 0.5077555775642395,
+      "learning_rate": 0.0008697590354705678,
+      "loss": 2.9767,
+      "step": 2016
+    },
+    {
+      "epoch": 0.258176,
+      "grad_norm": 0.5556850433349609,
+      "learning_rate": 0.0008696194549701564,
+      "loss": 2.8845,
+      "step": 2017
+    },
+    {
+      "epoch": 0.258304,
+      "grad_norm": 0.5353317856788635,
+      "learning_rate": 0.0008694798109278757,
+      "loss": 3.0244,
+      "step": 2018
+    },
+    {
+      "epoch": 0.258432,
+      "grad_norm": 0.5033209919929504,
+      "learning_rate": 0.0008693401033677321,
+      "loss": 2.9431,
+      "step": 2019
+    },
+    {
+      "epoch": 0.25856,
+      "grad_norm": 0.48652470111846924,
+      "learning_rate": 0.000869200332313743,
+      "loss": 2.9457,
+      "step": 2020
+    },
+    {
+      "epoch": 0.258688,
+      "grad_norm": 0.48623284697532654,
+      "learning_rate": 0.0008690604977899368,
+      "loss": 2.847,
+      "step": 2021
+    },
+    {
+      "epoch": 0.258816,
+      "grad_norm": 0.49265360832214355,
+      "learning_rate": 0.0008689205998203522,
+      "loss": 3.0337,
+      "step": 2022
+    },
+    {
+      "epoch": 0.258944,
+      "grad_norm": 0.48109617829322815,
+      "learning_rate": 0.0008687806384290399,
+      "loss": 2.7936,
+      "step": 2023
+    },
+    {
+      "epoch": 0.259072,
+      "grad_norm": 0.5052147507667542,
+      "learning_rate": 0.0008686406136400604,
+      "loss": 2.8482,
+      "step": 2024
+    },
+    {
+      "epoch": 0.2592,
+      "grad_norm": 0.531146228313446,
+      "learning_rate": 0.0008685005254774858,
+      "loss": 2.9252,
+      "step": 2025
+    },
+    {
+      "epoch": 0.259328,
+      "grad_norm": 0.5143641829490662,
+      "learning_rate": 0.0008683603739653989,
+      "loss": 2.8822,
+      "step": 2026
+    },
+    {
+      "epoch": 0.259456,
+      "grad_norm": 0.46483293175697327,
+      "learning_rate": 0.0008682201591278933,
+      "loss": 2.8384,
+      "step": 2027
+    },
+    {
+      "epoch": 0.259584,
+      "grad_norm": 0.4861757159233093,
+      "learning_rate": 0.0008680798809890736,
+      "loss": 2.9489,
+      "step": 2028
+    },
+    {
+      "epoch": 0.259712,
+      "grad_norm": 0.5075407028198242,
+      "learning_rate": 0.0008679395395730552,
+      "loss": 2.8704,
+      "step": 2029
+    },
+    {
+      "epoch": 0.25984,
+      "grad_norm": 0.529055655002594,
+      "learning_rate": 0.0008677991349039643,
+      "loss": 3.0455,
+      "step": 2030
+    },
+    {
+      "epoch": 0.259968,
+      "grad_norm": 0.5465389490127563,
+      "learning_rate": 0.0008676586670059383,
+      "loss": 2.9175,
+      "step": 2031
+    },
+    {
+      "epoch": 0.260096,
+      "grad_norm": 0.5384780168533325,
+      "learning_rate": 0.0008675181359031252,
+      "loss": 2.868,
+      "step": 2032
+    },
+    {
+      "epoch": 0.260224,
+      "grad_norm": 0.5506133437156677,
+      "learning_rate": 0.0008673775416196839,
+      "loss": 2.9754,
+      "step": 2033
+    },
+    {
+      "epoch": 0.260352,
+      "grad_norm": 0.535566508769989,
+      "learning_rate": 0.0008672368841797841,
+      "loss": 2.9188,
+      "step": 2034
+    },
+    {
+      "epoch": 0.26048,
+      "grad_norm": 0.548651397228241,
+      "learning_rate": 0.0008670961636076067,
+      "loss": 2.9662,
+      "step": 2035
+    },
+    {
+      "epoch": 0.260608,
+      "grad_norm": 0.5245262384414673,
+      "learning_rate": 0.0008669553799273428,
+      "loss": 2.9679,
+      "step": 2036
+    },
+    {
+      "epoch": 0.260736,
+      "grad_norm": 0.5167974829673767,
+      "learning_rate": 0.0008668145331631953,
+      "loss": 2.8212,
+      "step": 2037
+    },
+    {
+      "epoch": 0.260864,
+      "grad_norm": 0.5037155151367188,
+      "learning_rate": 0.0008666736233393768,
+      "loss": 2.9459,
+      "step": 2038
+    },
+    {
+      "epoch": 0.260992,
+      "grad_norm": 0.4820737838745117,
+      "learning_rate": 0.0008665326504801117,
+      "loss": 2.9742,
+      "step": 2039
+    },
+    {
+      "epoch": 0.26112,
+      "grad_norm": 0.4637087881565094,
+      "learning_rate": 0.0008663916146096348,
+      "loss": 3.0183,
+      "step": 2040
+    },
+    {
+      "epoch": 0.261248,
+      "grad_norm": 0.48485812544822693,
+      "learning_rate": 0.0008662505157521918,
+      "loss": 3.0368,
+      "step": 2041
+    },
+    {
+      "epoch": 0.261376,
+      "grad_norm": 0.5528088808059692,
+      "learning_rate": 0.0008661093539320391,
+      "loss": 2.9484,
+      "step": 2042
+    },
+    {
+      "epoch": 0.261504,
+      "grad_norm": 0.5247149467468262,
+      "learning_rate": 0.0008659681291734441,
+      "loss": 2.9881,
+      "step": 2043
+    },
+    {
+      "epoch": 0.261632,
+      "grad_norm": 0.5311048030853271,
+      "learning_rate": 0.0008658268415006851,
+      "loss": 2.8171,
+      "step": 2044
+    },
+    {
+      "epoch": 0.26176,
+      "grad_norm": 0.5387604832649231,
+      "learning_rate": 0.0008656854909380511,
+      "loss": 3.003,
+      "step": 2045
+    },
+    {
+      "epoch": 0.261888,
+      "grad_norm": 0.5462945103645325,
+      "learning_rate": 0.0008655440775098417,
+      "loss": 2.8913,
+      "step": 2046
+    },
+    {
+      "epoch": 0.262016,
+      "grad_norm": 0.5693079829216003,
+      "learning_rate": 0.0008654026012403677,
+      "loss": 2.9629,
+      "step": 2047
+    },
+    {
+      "epoch": 0.262144,
+      "grad_norm": 0.49810266494750977,
+      "learning_rate": 0.0008652610621539505,
+      "loss": 3.0535,
+      "step": 2048
+    },
+    {
+      "epoch": 0.262272,
+      "grad_norm": 0.5120519399642944,
+      "learning_rate": 0.0008651194602749221,
+      "loss": 2.912,
+      "step": 2049
+    },
+    {
+      "epoch": 0.2624,
+      "grad_norm": 0.5165218710899353,
+      "learning_rate": 0.0008649777956276257,
+      "loss": 2.8929,
+      "step": 2050
+    },
+    {
+      "epoch": 0.262528,
+      "grad_norm": 0.5299156308174133,
+      "learning_rate": 0.0008648360682364151,
+      "loss": 2.7366,
+      "step": 2051
+    },
+    {
+      "epoch": 0.262656,
+      "grad_norm": 0.5429794192314148,
+      "learning_rate": 0.0008646942781256548,
+      "loss": 2.7872,
+      "step": 2052
+    },
+    {
+      "epoch": 0.262784,
+      "grad_norm": 0.5046778321266174,
+      "learning_rate": 0.0008645524253197201,
+      "loss": 2.9235,
+      "step": 2053
+    },
+    {
+      "epoch": 0.262912,
+      "grad_norm": 0.5390233993530273,
+      "learning_rate": 0.0008644105098429974,
+      "loss": 2.9491,
+      "step": 2054
+    },
+    {
+      "epoch": 0.26304,
+      "grad_norm": 0.5340114235877991,
+      "learning_rate": 0.0008642685317198833,
+      "loss": 2.9392,
+      "step": 2055
+    },
+    {
+      "epoch": 0.263168,
+      "grad_norm": 0.5169486999511719,
+      "learning_rate": 0.0008641264909747858,
+      "loss": 2.8016,
+      "step": 2056
+    },
+    {
+      "epoch": 0.263296,
+      "grad_norm": 0.5475737452507019,
+      "learning_rate": 0.0008639843876321231,
+      "loss": 2.9573,
+      "step": 2057
+    },
+    {
+      "epoch": 0.263424,
+      "grad_norm": 0.4866578280925751,
+      "learning_rate": 0.0008638422217163244,
+      "loss": 2.9155,
+      "step": 2058
+    },
+    {
+      "epoch": 0.263552,
+      "grad_norm": 0.5058185458183289,
+      "learning_rate": 0.00086369999325183,
+      "loss": 2.9228,
+      "step": 2059
+    },
+    {
+      "epoch": 0.26368,
+      "grad_norm": 0.507924497127533,
+      "learning_rate": 0.0008635577022630901,
+      "loss": 2.9174,
+      "step": 2060
+    },
+    {
+      "epoch": 0.263808,
+      "grad_norm": 0.5135180950164795,
+      "learning_rate": 0.0008634153487745666,
+      "loss": 2.8917,
+      "step": 2061
+    },
+    {
+      "epoch": 0.263936,
+      "grad_norm": 0.5074484944343567,
+      "learning_rate": 0.0008632729328107317,
+      "loss": 2.8797,
+      "step": 2062
+    },
+    {
+      "epoch": 0.264064,
+      "grad_norm": 0.540285587310791,
+      "learning_rate": 0.000863130454396068,
+      "loss": 2.944,
+      "step": 2063
+    },
+    {
+      "epoch": 0.264192,
+      "grad_norm": 0.49974554777145386,
+      "learning_rate": 0.0008629879135550694,
+      "loss": 3.0382,
+      "step": 2064
+    },
+    {
+      "epoch": 0.26432,
+      "grad_norm": 0.4749911427497864,
+      "learning_rate": 0.0008628453103122404,
+      "loss": 2.9851,
+      "step": 2065
+    },
+    {
+      "epoch": 0.264448,
+      "grad_norm": 0.49232587218284607,
+      "learning_rate": 0.0008627026446920963,
+      "loss": 2.9586,
+      "step": 2066
+    },
+    {
+      "epoch": 0.264576,
+      "grad_norm": 0.5131406784057617,
+      "learning_rate": 0.0008625599167191626,
+      "loss": 3.0048,
+      "step": 2067
+    },
+    {
+      "epoch": 0.264704,
+      "grad_norm": 0.5285881757736206,
+      "learning_rate": 0.0008624171264179761,
+      "loss": 2.8846,
+      "step": 2068
+    },
+    {
+      "epoch": 0.264832,
+      "grad_norm": 0.5315921306610107,
+      "learning_rate": 0.000862274273813084,
+      "loss": 2.9602,
+      "step": 2069
+    },
+    {
+      "epoch": 0.26496,
+      "grad_norm": 0.5786234736442566,
+      "learning_rate": 0.0008621313589290445,
+      "loss": 3.0839,
+      "step": 2070
+    },
+    {
+      "epoch": 0.265088,
+      "grad_norm": 0.523207426071167,
+      "learning_rate": 0.0008619883817904262,
+      "loss": 2.9252,
+      "step": 2071
+    },
+    {
+      "epoch": 0.265216,
+      "grad_norm": 0.540807843208313,
+      "learning_rate": 0.0008618453424218084,
+      "loss": 2.8772,
+      "step": 2072
+    },
+    {
+      "epoch": 0.265344,
+      "grad_norm": 0.5260081887245178,
+      "learning_rate": 0.0008617022408477815,
+      "loss": 3.0535,
+      "step": 2073
+    },
+    {
+      "epoch": 0.265472,
+      "grad_norm": 0.48770636320114136,
+      "learning_rate": 0.000861559077092946,
+      "loss": 3.0139,
+      "step": 2074
+    },
+    {
+      "epoch": 0.2656,
+      "grad_norm": 0.5326758027076721,
+      "learning_rate": 0.0008614158511819138,
+      "loss": 2.9466,
+      "step": 2075
+    },
+    {
+      "epoch": 0.265728,
+      "grad_norm": 0.48201248049736023,
+      "learning_rate": 0.0008612725631393068,
+      "loss": 3.0258,
+      "step": 2076
+    },
+    {
+      "epoch": 0.265856,
+      "grad_norm": 0.5121851563453674,
+      "learning_rate": 0.0008611292129897581,
+      "loss": 2.8906,
+      "step": 2077
+    },
+    {
+      "epoch": 0.265984,
+      "grad_norm": 0.4921637773513794,
+      "learning_rate": 0.000860985800757911,
+      "loss": 2.8828,
+      "step": 2078
+    },
+    {
+      "epoch": 0.266112,
+      "grad_norm": 0.499891996383667,
+      "learning_rate": 0.0008608423264684198,
+      "loss": 2.8621,
+      "step": 2079
+    },
+    {
+      "epoch": 0.26624,
+      "grad_norm": 0.5032579898834229,
+      "learning_rate": 0.0008606987901459496,
+      "loss": 2.8332,
+      "step": 2080
+    },
+    {
+      "epoch": 0.266368,
+      "grad_norm": 0.508699357509613,
+      "learning_rate": 0.0008605551918151755,
+      "loss": 2.8089,
+      "step": 2081
+    },
+    {
+      "epoch": 0.266496,
+      "grad_norm": 0.47923555970191956,
+      "learning_rate": 0.0008604115315007843,
+      "loss": 2.9697,
+      "step": 2082
+    },
+    {
+      "epoch": 0.266624,
+      "grad_norm": 0.5003945231437683,
+      "learning_rate": 0.0008602678092274725,
+      "loss": 2.8829,
+      "step": 2083
+    },
+    {
+      "epoch": 0.266752,
+      "grad_norm": 0.5228596329689026,
+      "learning_rate": 0.0008601240250199476,
+      "loss": 2.9865,
+      "step": 2084
+    },
+    {
+      "epoch": 0.26688,
+      "grad_norm": 0.5228360295295715,
+      "learning_rate": 0.000859980178902928,
+      "loss": 3.0115,
+      "step": 2085
+    },
+    {
+      "epoch": 0.267008,
+      "grad_norm": 0.49090850353240967,
+      "learning_rate": 0.0008598362709011424,
+      "loss": 2.9267,
+      "step": 2086
+    },
+    {
+      "epoch": 0.267136,
+      "grad_norm": 0.49136295914649963,
+      "learning_rate": 0.0008596923010393303,
+      "loss": 2.9597,
+      "step": 2087
+    },
+    {
+      "epoch": 0.267264,
+      "grad_norm": 0.5341232419013977,
+      "learning_rate": 0.0008595482693422417,
+      "loss": 3.101,
+      "step": 2088
+    },
+    {
+      "epoch": 0.267392,
+      "grad_norm": 0.5146908164024353,
+      "learning_rate": 0.0008594041758346371,
+      "loss": 2.797,
+      "step": 2089
+    },
+    {
+      "epoch": 0.26752,
+      "grad_norm": 0.49823591113090515,
+      "learning_rate": 0.0008592600205412883,
+      "loss": 2.9735,
+      "step": 2090
+    },
+    {
+      "epoch": 0.267648,
+      "grad_norm": 0.5240467190742493,
+      "learning_rate": 0.0008591158034869773,
+      "loss": 3.0317,
+      "step": 2091
+    },
+    {
+      "epoch": 0.267776,
+      "grad_norm": 0.5306592583656311,
+      "learning_rate": 0.0008589715246964963,
+      "loss": 2.9374,
+      "step": 2092
+    },
+    {
+      "epoch": 0.267904,
+      "grad_norm": 0.542686939239502,
+      "learning_rate": 0.0008588271841946484,
+      "loss": 2.9028,
+      "step": 2093
+    },
+    {
+      "epoch": 0.268032,
+      "grad_norm": 0.5377242565155029,
+      "learning_rate": 0.000858682782006248,
+      "loss": 2.9987,
+      "step": 2094
+    },
+    {
+      "epoch": 0.26816,
+      "grad_norm": 0.5923547744750977,
+      "learning_rate": 0.000858538318156119,
+      "loss": 2.9906,
+      "step": 2095
+    },
+    {
+      "epoch": 0.268288,
+      "grad_norm": 0.5137338638305664,
+      "learning_rate": 0.0008583937926690965,
+      "loss": 2.8925,
+      "step": 2096
+    },
+    {
+      "epoch": 0.268416,
+      "grad_norm": 0.552821934223175,
+      "learning_rate": 0.0008582492055700264,
+      "loss": 2.9909,
+      "step": 2097
+    },
+    {
+      "epoch": 0.268544,
+      "grad_norm": 0.4974272549152374,
+      "learning_rate": 0.0008581045568837645,
+      "loss": 2.8466,
+      "step": 2098
+    },
+    {
+      "epoch": 0.268672,
+      "grad_norm": 0.5621693134307861,
+      "learning_rate": 0.0008579598466351779,
+      "loss": 2.9919,
+      "step": 2099
+    },
+    {
+      "epoch": 0.2688,
+      "grad_norm": 0.5093082189559937,
+      "learning_rate": 0.0008578150748491438,
+      "loss": 2.8134,
+      "step": 2100
+    },
+    {
+      "epoch": 0.268928,
+      "grad_norm": 0.5065686106681824,
+      "learning_rate": 0.0008576702415505501,
+      "loss": 3.0115,
+      "step": 2101
+    },
+    {
+      "epoch": 0.269056,
+      "grad_norm": 0.5110481977462769,
+      "learning_rate": 0.0008575253467642954,
+      "loss": 2.9646,
+      "step": 2102
+    },
+    {
+      "epoch": 0.269184,
+      "grad_norm": 0.5351948142051697,
+      "learning_rate": 0.000857380390515289,
+      "loss": 2.9487,
+      "step": 2103
+    },
+    {
+      "epoch": 0.269312,
+      "grad_norm": 0.5070205926895142,
+      "learning_rate": 0.0008572353728284501,
+      "loss": 2.9225,
+      "step": 2104
+    },
+    {
+      "epoch": 0.26944,
+      "grad_norm": 0.5263282656669617,
+      "learning_rate": 0.0008570902937287092,
+      "loss": 3.0406,
+      "step": 2105
+    },
+    {
+      "epoch": 0.269568,
+      "grad_norm": 0.5243058204650879,
+      "learning_rate": 0.0008569451532410074,
+      "loss": 2.928,
+      "step": 2106
+    },
+    {
+      "epoch": 0.269696,
+      "grad_norm": 0.5621421933174133,
+      "learning_rate": 0.0008567999513902953,
+      "loss": 2.9921,
+      "step": 2107
+    },
+    {
+      "epoch": 0.269824,
+      "grad_norm": 0.5412895679473877,
+      "learning_rate": 0.0008566546882015354,
+      "loss": 2.6927,
+      "step": 2108
+    },
+    {
+      "epoch": 0.269952,
+      "grad_norm": 0.5112060904502869,
+      "learning_rate": 0.0008565093636996999,
+      "loss": 2.8477,
+      "step": 2109
+    },
+    {
+      "epoch": 0.27008,
+      "grad_norm": 0.5135743021965027,
+      "learning_rate": 0.0008563639779097716,
+      "loss": 2.9713,
+      "step": 2110
+    },
+    {
+      "epoch": 0.270208,
+      "grad_norm": 0.5890532732009888,
+      "learning_rate": 0.0008562185308567443,
+      "loss": 2.9255,
+      "step": 2111
+    },
+    {
+      "epoch": 0.270336,
+      "grad_norm": 0.5350236892700195,
+      "learning_rate": 0.000856073022565622,
+      "loss": 2.9259,
+      "step": 2112
+    },
+    {
+      "epoch": 0.270464,
+      "grad_norm": 0.5126326680183411,
+      "learning_rate": 0.0008559274530614189,
+      "loss": 2.9323,
+      "step": 2113
+    },
+    {
+      "epoch": 0.270592,
+      "grad_norm": 0.487209290266037,
+      "learning_rate": 0.0008557818223691607,
+      "loss": 2.8677,
+      "step": 2114
+    },
+    {
+      "epoch": 0.27072,
+      "grad_norm": 0.47234296798706055,
+      "learning_rate": 0.0008556361305138825,
+      "loss": 3.0006,
+      "step": 2115
+    },
+    {
+      "epoch": 0.270848,
+      "grad_norm": 0.5286137461662292,
+      "learning_rate": 0.0008554903775206304,
+      "loss": 2.9172,
+      "step": 2116
+    },
+    {
+      "epoch": 0.270976,
+      "grad_norm": 0.48669183254241943,
+      "learning_rate": 0.0008553445634144614,
+      "loss": 2.8729,
+      "step": 2117
+    },
+    {
+      "epoch": 0.271104,
+      "grad_norm": 0.509513795375824,
+      "learning_rate": 0.0008551986882204424,
+      "loss": 2.9413,
+      "step": 2118
+    },
+    {
+      "epoch": 0.271232,
+      "grad_norm": 0.5105999708175659,
+      "learning_rate": 0.0008550527519636511,
+      "loss": 2.9847,
+      "step": 2119
+    },
+    {
+      "epoch": 0.27136,
+      "grad_norm": 0.5052416324615479,
+      "learning_rate": 0.0008549067546691756,
+      "loss": 2.9243,
+      "step": 2120
+    },
+    {
+      "epoch": 0.271488,
+      "grad_norm": 0.4960881173610687,
+      "learning_rate": 0.0008547606963621145,
+      "loss": 2.9268,
+      "step": 2121
+    },
+    {
+      "epoch": 0.271616,
+      "grad_norm": 0.5650200843811035,
+      "learning_rate": 0.000854614577067577,
+      "loss": 3.0192,
+      "step": 2122
+    },
+    {
+      "epoch": 0.271744,
+      "grad_norm": 0.6402601003646851,
+      "learning_rate": 0.0008544683968106826,
+      "loss": 2.9502,
+      "step": 2123
+    },
+    {
+      "epoch": 0.271872,
+      "grad_norm": 0.5338484644889832,
+      "learning_rate": 0.0008543221556165614,
+      "loss": 2.9272,
+      "step": 2124
+    },
+    {
+      "epoch": 0.272,
+      "grad_norm": 0.5013724565505981,
+      "learning_rate": 0.0008541758535103541,
+      "loss": 3.0497,
+      "step": 2125
+    },
+    {
+      "epoch": 0.272128,
+      "grad_norm": 0.529005765914917,
+      "learning_rate": 0.0008540294905172117,
+      "loss": 2.9277,
+      "step": 2126
+    },
+    {
+      "epoch": 0.272256,
+      "grad_norm": 0.5064907073974609,
+      "learning_rate": 0.0008538830666622954,
+      "loss": 2.9121,
+      "step": 2127
+    },
+    {
+      "epoch": 0.272384,
+      "grad_norm": 0.5066137909889221,
+      "learning_rate": 0.0008537365819707776,
+      "loss": 2.8418,
+      "step": 2128
+    },
+    {
+      "epoch": 0.272512,
+      "grad_norm": 0.4868587851524353,
+      "learning_rate": 0.0008535900364678402,
+      "loss": 2.8772,
+      "step": 2129
+    },
+    {
+      "epoch": 0.27264,
+      "grad_norm": 0.5333967804908752,
+      "learning_rate": 0.0008534434301786766,
+      "loss": 2.8576,
+      "step": 2130
+    },
+    {
+      "epoch": 0.272768,
+      "grad_norm": 0.49044179916381836,
+      "learning_rate": 0.0008532967631284898,
+      "loss": 2.8791,
+      "step": 2131
+    },
+    {
+      "epoch": 0.272896,
+      "grad_norm": 0.4689520299434662,
+      "learning_rate": 0.0008531500353424937,
+      "loss": 2.9332,
+      "step": 2132
+    },
+    {
+      "epoch": 0.273024,
+      "grad_norm": 0.5359063744544983,
+      "learning_rate": 0.0008530032468459126,
+      "loss": 2.8856,
+      "step": 2133
+    },
+    {
+      "epoch": 0.273152,
+      "grad_norm": 0.5670987963676453,
+      "learning_rate": 0.0008528563976639807,
+      "loss": 2.9788,
+      "step": 2134
+    },
+    {
+      "epoch": 0.27328,
+      "grad_norm": 0.5089085698127747,
+      "learning_rate": 0.0008527094878219434,
+      "loss": 2.8778,
+      "step": 2135
+    },
+    {
+      "epoch": 0.273408,
+      "grad_norm": 0.5071350932121277,
+      "learning_rate": 0.0008525625173450564,
+      "loss": 3.0142,
+      "step": 2136
+    },
+    {
+      "epoch": 0.273536,
+      "grad_norm": 0.5504944324493408,
+      "learning_rate": 0.0008524154862585853,
+      "loss": 3.0096,
+      "step": 2137
+    },
+    {
+      "epoch": 0.273664,
+      "grad_norm": 0.5275861620903015,
+      "learning_rate": 0.0008522683945878066,
+      "loss": 2.8933,
+      "step": 2138
+    },
+    {
+      "epoch": 0.273792,
+      "grad_norm": 0.5219005346298218,
+      "learning_rate": 0.000852121242358007,
+      "loss": 2.8273,
+      "step": 2139
+    },
+    {
+      "epoch": 0.27392,
+      "grad_norm": 0.5288229584693909,
+      "learning_rate": 0.0008519740295944838,
+      "loss": 2.9303,
+      "step": 2140
+    },
+    {
+      "epoch": 0.274048,
+      "grad_norm": 0.5196417570114136,
+      "learning_rate": 0.0008518267563225444,
+      "loss": 3.0062,
+      "step": 2141
+    },
+    {
+      "epoch": 0.274176,
+      "grad_norm": 0.492921382188797,
+      "learning_rate": 0.0008516794225675069,
+      "loss": 3.0126,
+      "step": 2142
+    },
+    {
+      "epoch": 0.274304,
+      "grad_norm": 0.47151386737823486,
+      "learning_rate": 0.0008515320283546995,
+      "loss": 2.9658,
+      "step": 2143
+    },
+    {
+      "epoch": 0.274432,
+      "grad_norm": 0.5093326568603516,
+      "learning_rate": 0.0008513845737094612,
+      "loss": 2.8518,
+      "step": 2144
+    },
+    {
+      "epoch": 0.27456,
+      "grad_norm": 0.4952482581138611,
+      "learning_rate": 0.0008512370586571409,
+      "loss": 2.9698,
+      "step": 2145
+    },
+    {
+      "epoch": 0.274688,
+      "grad_norm": 0.4868939220905304,
+      "learning_rate": 0.0008510894832230987,
+      "loss": 2.9622,
+      "step": 2146
+    },
+    {
+      "epoch": 0.274816,
+      "grad_norm": 0.4903799593448639,
+      "learning_rate": 0.000850941847432704,
+      "loss": 2.9044,
+      "step": 2147
+    },
+    {
+      "epoch": 0.274944,
+      "grad_norm": 0.5940061807632446,
+      "learning_rate": 0.000850794151311337,
+      "loss": 2.9298,
+      "step": 2148
+    },
+    {
+      "epoch": 0.275072,
+      "grad_norm": 0.48996442556381226,
+      "learning_rate": 0.000850646394884389,
+      "loss": 2.8917,
+      "step": 2149
+    },
+    {
+      "epoch": 0.2752,
+      "grad_norm": 0.48524510860443115,
+      "learning_rate": 0.0008504985781772605,
+      "loss": 2.921,
+      "step": 2150
+    },
+    {
+      "epoch": 0.275328,
+      "grad_norm": 0.4882016181945801,
+      "learning_rate": 0.0008503507012153632,
+      "loss": 2.9625,
+      "step": 2151
+    },
+    {
+      "epoch": 0.275456,
+      "grad_norm": 0.4798775017261505,
+      "learning_rate": 0.0008502027640241187,
+      "loss": 3.0496,
+      "step": 2152
+    },
+    {
+      "epoch": 0.275584,
+      "grad_norm": 0.534584105014801,
+      "learning_rate": 0.000850054766628959,
+      "loss": 2.9701,
+      "step": 2153
+    },
+    {
+      "epoch": 0.275712,
+      "grad_norm": 0.5125247836112976,
+      "learning_rate": 0.0008499067090553269,
+      "loss": 3.0213,
+      "step": 2154
+    },
+    {
+      "epoch": 0.27584,
+      "grad_norm": 1.0365999937057495,
+      "learning_rate": 0.0008497585913286752,
+      "loss": 3.0709,
+      "step": 2155
+    },
+    {
+      "epoch": 0.275968,
+      "grad_norm": 0.5339630842208862,
+      "learning_rate": 0.0008496104134744667,
+      "loss": 2.9981,
+      "step": 2156
+    },
+    {
+      "epoch": 0.276096,
+      "grad_norm": 0.5207678079605103,
+      "learning_rate": 0.000849462175518175,
+      "loss": 2.8532,
+      "step": 2157
+    },
+    {
+      "epoch": 0.276224,
+      "grad_norm": 0.5268397927284241,
+      "learning_rate": 0.0008493138774852841,
+      "loss": 2.8861,
+      "step": 2158
+    },
+    {
+      "epoch": 0.276352,
+      "grad_norm": 0.5397483706474304,
+      "learning_rate": 0.0008491655194012881,
+      "loss": 2.8377,
+      "step": 2159
+    },
+    {
+      "epoch": 0.27648,
+      "grad_norm": 0.5007504224777222,
+      "learning_rate": 0.0008490171012916914,
+      "loss": 2.9713,
+      "step": 2160
+    },
+    {
+      "epoch": 0.276608,
+      "grad_norm": 0.48868128657341003,
+      "learning_rate": 0.0008488686231820088,
+      "loss": 2.9028,
+      "step": 2161
+    },
+    {
+      "epoch": 0.276736,
+      "grad_norm": 0.5563438534736633,
+      "learning_rate": 0.0008487200850977657,
+      "loss": 2.9377,
+      "step": 2162
+    },
+    {
+      "epoch": 0.276864,
+      "grad_norm": 0.49761515855789185,
+      "learning_rate": 0.0008485714870644969,
+      "loss": 2.8427,
+      "step": 2163
+    },
+    {
+      "epoch": 0.276992,
+      "grad_norm": 0.4809052050113678,
+      "learning_rate": 0.0008484228291077488,
+      "loss": 2.8484,
+      "step": 2164
+    },
+    {
+      "epoch": 0.27712,
+      "grad_norm": 0.5223020911216736,
+      "learning_rate": 0.0008482741112530769,
+      "loss": 2.8706,
+      "step": 2165
+    },
+    {
+      "epoch": 0.277248,
+      "grad_norm": 0.4774249196052551,
+      "learning_rate": 0.0008481253335260478,
+      "loss": 2.8498,
+      "step": 2166
+    },
+    {
+      "epoch": 0.277376,
+      "grad_norm": 0.5010533928871155,
+      "learning_rate": 0.000847976495952238,
+      "loss": 2.9161,
+      "step": 2167
+    },
+    {
+      "epoch": 0.277504,
+      "grad_norm": 0.5006215572357178,
+      "learning_rate": 0.0008478275985572346,
+      "loss": 2.8705,
+      "step": 2168
+    },
+    {
+      "epoch": 0.277632,
+      "grad_norm": 0.6531100273132324,
+      "learning_rate": 0.0008476786413666345,
+      "loss": 2.9097,
+      "step": 2169
+    },
+    {
+      "epoch": 0.27776,
+      "grad_norm": 0.5282934904098511,
+      "learning_rate": 0.0008475296244060454,
+      "loss": 2.7357,
+      "step": 2170
+    },
+    {
+      "epoch": 0.277888,
+      "grad_norm": 0.5993438363075256,
+      "learning_rate": 0.0008473805477010849,
+      "loss": 2.8518,
+      "step": 2171
+    },
+    {
+      "epoch": 0.278016,
+      "grad_norm": 0.5443519949913025,
+      "learning_rate": 0.0008472314112773811,
+      "loss": 3.0708,
+      "step": 2172
+    },
+    {
+      "epoch": 0.278144,
+      "grad_norm": 0.512251615524292,
+      "learning_rate": 0.0008470822151605723,
+      "loss": 2.8952,
+      "step": 2173
+    },
+    {
+      "epoch": 0.278272,
+      "grad_norm": 0.5098617672920227,
+      "learning_rate": 0.0008469329593763069,
+      "loss": 2.8142,
+      "step": 2174
+    },
+    {
+      "epoch": 0.2784,
+      "grad_norm": 0.537761390209198,
+      "learning_rate": 0.0008467836439502439,
+      "loss": 3.0313,
+      "step": 2175
+    },
+    {
+      "epoch": 0.278528,
+      "grad_norm": 0.5045440196990967,
+      "learning_rate": 0.000846634268908052,
+      "loss": 3.0119,
+      "step": 2176
+    },
+    {
+      "epoch": 0.278656,
+      "grad_norm": 0.5316609740257263,
+      "learning_rate": 0.000846484834275411,
+      "loss": 2.8094,
+      "step": 2177
+    },
+    {
+      "epoch": 0.278784,
+      "grad_norm": 0.4961012899875641,
+      "learning_rate": 0.0008463353400780101,
+      "loss": 3.0169,
+      "step": 2178
+    },
+    {
+      "epoch": 0.278912,
+      "grad_norm": 0.5272672176361084,
+      "learning_rate": 0.0008461857863415492,
+      "loss": 2.8874,
+      "step": 2179
+    },
+    {
+      "epoch": 0.27904,
+      "grad_norm": 0.5152226686477661,
+      "learning_rate": 0.0008460361730917383,
+      "loss": 2.9655,
+      "step": 2180
+    },
+    {
+      "epoch": 0.279168,
+      "grad_norm": 0.5269408226013184,
+      "learning_rate": 0.0008458865003542975,
+      "loss": 3.0829,
+      "step": 2181
+    },
+    {
+      "epoch": 0.279296,
+      "grad_norm": 0.5601498484611511,
+      "learning_rate": 0.0008457367681549575,
+      "loss": 3.0497,
+      "step": 2182
+    },
+    {
+      "epoch": 0.279424,
+      "grad_norm": 0.49007099866867065,
+      "learning_rate": 0.0008455869765194591,
+      "loss": 2.8836,
+      "step": 2183
+    },
+    {
+      "epoch": 0.279552,
+      "grad_norm": 0.530005156993866,
+      "learning_rate": 0.000845437125473553,
+      "loss": 2.8008,
+      "step": 2184
+    },
+    {
+      "epoch": 0.27968,
+      "grad_norm": 0.531674861907959,
+      "learning_rate": 0.0008452872150430002,
+      "loss": 2.9136,
+      "step": 2185
+    },
+    {
+      "epoch": 0.279808,
+      "grad_norm": 0.5207239985466003,
+      "learning_rate": 0.0008451372452535724,
+      "loss": 3.0001,
+      "step": 2186
+    },
+    {
+      "epoch": 0.279936,
+      "grad_norm": 0.4993184208869934,
+      "learning_rate": 0.000844987216131051,
+      "loss": 3.0023,
+      "step": 2187
+    },
+    {
+      "epoch": 0.280064,
+      "grad_norm": 0.4853539764881134,
+      "learning_rate": 0.0008448371277012276,
+      "loss": 2.9095,
+      "step": 2188
+    },
+    {
+      "epoch": 0.280192,
+      "grad_norm": 0.5109748840332031,
+      "learning_rate": 0.0008446869799899042,
+      "loss": 2.7832,
+      "step": 2189
+    },
+    {
+      "epoch": 0.28032,
+      "grad_norm": 0.48652562499046326,
+      "learning_rate": 0.000844536773022893,
+      "loss": 2.9806,
+      "step": 2190
+    },
+    {
+      "epoch": 0.280448,
+      "grad_norm": 0.48760443925857544,
+      "learning_rate": 0.0008443865068260164,
+      "loss": 2.9542,
+      "step": 2191
+    },
+    {
+      "epoch": 0.280576,
+      "grad_norm": 0.5412243604660034,
+      "learning_rate": 0.0008442361814251069,
+      "loss": 2.8832,
+      "step": 2192
+    },
+    {
+      "epoch": 0.280704,
+      "grad_norm": 0.49491411447525024,
+      "learning_rate": 0.0008440857968460069,
+      "loss": 2.9374,
+      "step": 2193
+    },
+    {
+      "epoch": 0.280832,
+      "grad_norm": 0.5309852361679077,
+      "learning_rate": 0.0008439353531145695,
+      "loss": 3.0851,
+      "step": 2194
+    },
+    {
+      "epoch": 0.28096,
+      "grad_norm": 0.4881364703178406,
+      "learning_rate": 0.0008437848502566575,
+      "loss": 2.9858,
+      "step": 2195
+    },
+    {
+      "epoch": 0.281088,
+      "grad_norm": 0.49775564670562744,
+      "learning_rate": 0.0008436342882981445,
+      "loss": 2.918,
+      "step": 2196
+    },
+    {
+      "epoch": 0.281216,
+      "grad_norm": 0.5031439065933228,
+      "learning_rate": 0.0008434836672649135,
+      "loss": 2.9576,
+      "step": 2197
+    },
+    {
+      "epoch": 0.281344,
+      "grad_norm": 0.518322765827179,
+      "learning_rate": 0.0008433329871828582,
+      "loss": 2.9748,
+      "step": 2198
+    },
+    {
+      "epoch": 0.281472,
+      "grad_norm": 0.5010822415351868,
+      "learning_rate": 0.0008431822480778819,
+      "loss": 2.8664,
+      "step": 2199
+    },
+    {
+      "epoch": 0.2816,
+      "grad_norm": 0.4954446256160736,
+      "learning_rate": 0.0008430314499758986,
+      "loss": 2.9035,
+      "step": 2200
+    },
+    {
+      "epoch": 0.281728,
+      "grad_norm": 0.49386176466941833,
+      "learning_rate": 0.0008428805929028326,
+      "loss": 2.8361,
+      "step": 2201
+    },
+    {
+      "epoch": 0.281856,
+      "grad_norm": 0.5233529806137085,
+      "learning_rate": 0.0008427296768846175,
+      "loss": 2.9609,
+      "step": 2202
+    },
+    {
+      "epoch": 0.281984,
+      "grad_norm": 0.5602116584777832,
+      "learning_rate": 0.0008425787019471978,
+      "loss": 2.8403,
+      "step": 2203
+    },
+    {
+      "epoch": 0.282112,
+      "grad_norm": 0.4752119183540344,
+      "learning_rate": 0.0008424276681165275,
+      "loss": 2.9722,
+      "step": 2204
+    },
+    {
+      "epoch": 0.28224,
+      "grad_norm": 0.49233490228652954,
+      "learning_rate": 0.0008422765754185715,
+      "loss": 2.8875,
+      "step": 2205
+    },
+    {
+      "epoch": 0.282368,
+      "grad_norm": 0.5380737781524658,
+      "learning_rate": 0.000842125423879304,
+      "loss": 2.8036,
+      "step": 2206
+    },
+    {
+      "epoch": 0.282496,
+      "grad_norm": 0.5068323612213135,
+      "learning_rate": 0.00084197421352471,
+      "loss": 2.9985,
+      "step": 2207
+    },
+    {
+      "epoch": 0.282624,
+      "grad_norm": 0.4625907242298126,
+      "learning_rate": 0.0008418229443807841,
+      "loss": 2.8312,
+      "step": 2208
+    },
+    {
+      "epoch": 0.282752,
+      "grad_norm": 0.5034356117248535,
+      "learning_rate": 0.0008416716164735314,
+      "loss": 2.9291,
+      "step": 2209
+    },
+    {
+      "epoch": 0.28288,
+      "grad_norm": 0.5381683707237244,
+      "learning_rate": 0.0008415202298289668,
+      "loss": 2.9678,
+      "step": 2210
+    },
+    {
+      "epoch": 0.283008,
+      "grad_norm": 0.5304538607597351,
+      "learning_rate": 0.0008413687844731155,
+      "loss": 2.9934,
+      "step": 2211
+    },
+    {
+      "epoch": 0.283136,
+      "grad_norm": 0.5080546736717224,
+      "learning_rate": 0.0008412172804320126,
+      "loss": 2.88,
+      "step": 2212
+    },
+    {
+      "epoch": 0.283264,
+      "grad_norm": 0.507810115814209,
+      "learning_rate": 0.0008410657177317034,
+      "loss": 2.911,
+      "step": 2213
+    },
+    {
+      "epoch": 0.283392,
+      "grad_norm": 0.4965450167655945,
+      "learning_rate": 0.0008409140963982435,
+      "loss": 2.8709,
+      "step": 2214
+    },
+    {
+      "epoch": 0.28352,
+      "grad_norm": 0.48740383982658386,
+      "learning_rate": 0.0008407624164576981,
+      "loss": 2.8645,
+      "step": 2215
+    },
+    {
+      "epoch": 0.283648,
+      "grad_norm": 0.49104613065719604,
+      "learning_rate": 0.0008406106779361429,
+      "loss": 2.9743,
+      "step": 2216
+    },
+    {
+      "epoch": 0.283776,
+      "grad_norm": 0.48062846064567566,
+      "learning_rate": 0.0008404588808596634,
+      "loss": 2.9506,
+      "step": 2217
+    },
+    {
+      "epoch": 0.283904,
+      "grad_norm": 0.5338069200515747,
+      "learning_rate": 0.0008403070252543555,
+      "loss": 2.9408,
+      "step": 2218
+    },
+    {
+      "epoch": 0.284032,
+      "grad_norm": 0.5060318112373352,
+      "learning_rate": 0.0008401551111463245,
+      "loss": 2.9288,
+      "step": 2219
+    },
+    {
+      "epoch": 0.28416,
+      "grad_norm": 0.5320688486099243,
+      "learning_rate": 0.0008400031385616867,
+      "loss": 2.9757,
+      "step": 2220
+    },
+    {
+      "epoch": 0.284288,
+      "grad_norm": 0.5152877569198608,
+      "learning_rate": 0.0008398511075265677,
+      "loss": 3.0439,
+      "step": 2221
+    },
+    {
+      "epoch": 0.284416,
+      "grad_norm": 0.4876433312892914,
+      "learning_rate": 0.0008396990180671033,
+      "loss": 2.909,
+      "step": 2222
+    },
+    {
+      "epoch": 0.284544,
+      "grad_norm": 0.583520770072937,
+      "learning_rate": 0.0008395468702094398,
+      "loss": 2.8185,
+      "step": 2223
+    },
+    {
+      "epoch": 0.284672,
+      "grad_norm": 0.5256274342536926,
+      "learning_rate": 0.0008393946639797328,
+      "loss": 2.931,
+      "step": 2224
+    },
+    {
+      "epoch": 0.2848,
+      "grad_norm": 0.506183385848999,
+      "learning_rate": 0.0008392423994041485,
+      "loss": 3.0201,
+      "step": 2225
+    },
+    {
+      "epoch": 0.284928,
+      "grad_norm": 0.5341640114784241,
+      "learning_rate": 0.000839090076508863,
+      "loss": 2.8662,
+      "step": 2226
+    },
+    {
+      "epoch": 0.285056,
+      "grad_norm": 0.5014682412147522,
+      "learning_rate": 0.000838937695320062,
+      "loss": 2.8871,
+      "step": 2227
+    },
+    {
+      "epoch": 0.285184,
+      "grad_norm": 0.49816277623176575,
+      "learning_rate": 0.0008387852558639421,
+      "loss": 2.8338,
+      "step": 2228
+    },
+    {
+      "epoch": 0.285312,
+      "grad_norm": 0.49649056792259216,
+      "learning_rate": 0.0008386327581667091,
+      "loss": 2.9389,
+      "step": 2229
+    },
+    {
+      "epoch": 0.28544,
+      "grad_norm": 0.4932112991809845,
+      "learning_rate": 0.0008384802022545793,
+      "loss": 2.7877,
+      "step": 2230
+    },
+    {
+      "epoch": 0.285568,
+      "grad_norm": 0.5017111301422119,
+      "learning_rate": 0.0008383275881537786,
+      "loss": 2.9394,
+      "step": 2231
+    },
+    {
+      "epoch": 0.285696,
+      "grad_norm": 0.5759559273719788,
+      "learning_rate": 0.0008381749158905432,
+      "loss": 2.8288,
+      "step": 2232
+    },
+    {
+      "epoch": 0.285824,
+      "grad_norm": 0.4756017029285431,
+      "learning_rate": 0.0008380221854911194,
+      "loss": 2.9053,
+      "step": 2233
+    },
+    {
+      "epoch": 0.285952,
+      "grad_norm": 0.5252525806427002,
+      "learning_rate": 0.0008378693969817632,
+      "loss": 2.7067,
+      "step": 2234
+    },
+    {
+      "epoch": 0.28608,
+      "grad_norm": 0.5745764970779419,
+      "learning_rate": 0.0008377165503887409,
+      "loss": 2.8719,
+      "step": 2235
+    },
+    {
+      "epoch": 0.286208,
+      "grad_norm": 0.5504339337348938,
+      "learning_rate": 0.0008375636457383282,
+      "loss": 3.008,
+      "step": 2236
+    },
+    {
+      "epoch": 0.286336,
+      "grad_norm": 0.47700682282447815,
+      "learning_rate": 0.0008374106830568117,
+      "loss": 2.8622,
+      "step": 2237
+    },
+    {
+      "epoch": 0.286464,
+      "grad_norm": 0.5017300248146057,
+      "learning_rate": 0.000837257662370487,
+      "loss": 2.8431,
+      "step": 2238
+    },
+    {
+      "epoch": 0.286592,
+      "grad_norm": 0.49018168449401855,
+      "learning_rate": 0.0008371045837056604,
+      "loss": 3.0024,
+      "step": 2239
+    },
+    {
+      "epoch": 0.28672,
+      "grad_norm": 0.5128681659698486,
+      "learning_rate": 0.0008369514470886476,
+      "loss": 2.8561,
+      "step": 2240
+    },
+    {
+      "epoch": 0.286848,
+      "grad_norm": 0.5200481414794922,
+      "learning_rate": 0.000836798252545775,
+      "loss": 2.8996,
+      "step": 2241
+    },
+    {
+      "epoch": 0.286976,
+      "grad_norm": 0.4662335216999054,
+      "learning_rate": 0.0008366450001033783,
+      "loss": 2.8406,
+      "step": 2242
+    },
+    {
+      "epoch": 0.287104,
+      "grad_norm": 0.5330715179443359,
+      "learning_rate": 0.0008364916897878032,
+      "loss": 2.9093,
+      "step": 2243
+    },
+    {
+      "epoch": 0.287232,
+      "grad_norm": 0.4797719419002533,
+      "learning_rate": 0.0008363383216254057,
+      "loss": 3.0354,
+      "step": 2244
+    },
+    {
+      "epoch": 0.28736,
+      "grad_norm": 0.4953981339931488,
+      "learning_rate": 0.0008361848956425515,
+      "loss": 2.9058,
+      "step": 2245
+    },
+    {
+      "epoch": 0.287488,
+      "grad_norm": 0.5141475796699524,
+      "learning_rate": 0.0008360314118656165,
+      "loss": 2.9145,
+      "step": 2246
+    },
+    {
+      "epoch": 0.287616,
+      "grad_norm": 0.49981844425201416,
+      "learning_rate": 0.0008358778703209861,
+      "loss": 3.008,
+      "step": 2247
+    },
+    {
+      "epoch": 0.287744,
+      "grad_norm": 0.4718305766582489,
+      "learning_rate": 0.0008357242710350561,
+      "loss": 2.8875,
+      "step": 2248
+    },
+    {
+      "epoch": 0.287872,
+      "grad_norm": 0.49809014797210693,
+      "learning_rate": 0.0008355706140342318,
+      "loss": 2.832,
+      "step": 2249
+    },
+    {
+      "epoch": 0.288,
+      "grad_norm": 0.5109177231788635,
+      "learning_rate": 0.0008354168993449285,
+      "loss": 2.8382,
+      "step": 2250
+    },
+    {
+      "epoch": 0.288128,
+      "grad_norm": 0.5107190608978271,
+      "learning_rate": 0.0008352631269935718,
+      "loss": 2.9902,
+      "step": 2251
+    },
+    {
+      "epoch": 0.288256,
+      "grad_norm": 0.48891526460647583,
+      "learning_rate": 0.0008351092970065969,
+      "loss": 3.0427,
+      "step": 2252
+    },
+    {
+      "epoch": 0.288384,
+      "grad_norm": 0.4806212782859802,
+      "learning_rate": 0.000834955409410449,
+      "loss": 2.9022,
+      "step": 2253
+    },
+    {
+      "epoch": 0.288512,
+      "grad_norm": 0.47188320755958557,
+      "learning_rate": 0.0008348014642315831,
+      "loss": 2.9835,
+      "step": 2254
+    },
+    {
+      "epoch": 0.28864,
+      "grad_norm": 0.47357991337776184,
+      "learning_rate": 0.0008346474614964641,
+      "loss": 3.0028,
+      "step": 2255
+    },
+    {
+      "epoch": 0.288768,
+      "grad_norm": 0.46828097105026245,
+      "learning_rate": 0.0008344934012315668,
+      "loss": 3.0324,
+      "step": 2256
+    },
+    {
+      "epoch": 0.288896,
+      "grad_norm": 0.49793025851249695,
+      "learning_rate": 0.0008343392834633761,
+      "loss": 2.7782,
+      "step": 2257
+    },
+    {
+      "epoch": 0.289024,
+      "grad_norm": 0.48624387383461,
+      "learning_rate": 0.0008341851082183867,
+      "loss": 2.9213,
+      "step": 2258
+    },
+    {
+      "epoch": 0.289152,
+      "grad_norm": 0.46720096468925476,
+      "learning_rate": 0.0008340308755231027,
+      "loss": 2.9653,
+      "step": 2259
+    },
+    {
+      "epoch": 0.28928,
+      "grad_norm": 0.4911794364452362,
+      "learning_rate": 0.0008338765854040392,
+      "loss": 2.779,
+      "step": 2260
+    },
+    {
+      "epoch": 0.289408,
+      "grad_norm": 0.5098861455917358,
+      "learning_rate": 0.0008337222378877196,
+      "loss": 2.9421,
+      "step": 2261
+    },
+    {
+      "epoch": 0.289536,
+      "grad_norm": 0.5335205793380737,
+      "learning_rate": 0.0008335678330006789,
+      "loss": 2.8495,
+      "step": 2262
+    },
+    {
+      "epoch": 0.289664,
+      "grad_norm": 0.5269430875778198,
+      "learning_rate": 0.0008334133707694604,
+      "loss": 2.9857,
+      "step": 2263
+    },
+    {
+      "epoch": 0.289792,
+      "grad_norm": 0.5004192590713501,
+      "learning_rate": 0.0008332588512206181,
+      "loss": 3.1438,
+      "step": 2264
+    },
+    {
+      "epoch": 0.28992,
+      "grad_norm": 0.503845751285553,
+      "learning_rate": 0.000833104274380716,
+      "loss": 2.9843,
+      "step": 2265
+    },
+    {
+      "epoch": 0.290048,
+      "grad_norm": 0.48945069313049316,
+      "learning_rate": 0.0008329496402763273,
+      "loss": 2.9767,
+      "step": 2266
+    },
+    {
+      "epoch": 0.290176,
+      "grad_norm": 0.47048747539520264,
+      "learning_rate": 0.0008327949489340358,
+      "loss": 2.8641,
+      "step": 2267
+    },
+    {
+      "epoch": 0.290304,
+      "grad_norm": 0.5081391334533691,
+      "learning_rate": 0.0008326402003804343,
+      "loss": 2.8832,
+      "step": 2268
+    },
+    {
+      "epoch": 0.290432,
+      "grad_norm": 0.49963289499282837,
+      "learning_rate": 0.000832485394642126,
+      "loss": 2.7583,
+      "step": 2269
+    },
+    {
+      "epoch": 0.29056,
+      "grad_norm": 0.5315260887145996,
+      "learning_rate": 0.000832330531745724,
+      "loss": 2.9179,
+      "step": 2270
+    },
+    {
+      "epoch": 0.290688,
+      "grad_norm": 0.5159884691238403,
+      "learning_rate": 0.000832175611717851,
+      "loss": 2.9714,
+      "step": 2271
+    },
+    {
+      "epoch": 0.290816,
+      "grad_norm": 0.49588194489479065,
+      "learning_rate": 0.0008320206345851393,
+      "loss": 2.7616,
+      "step": 2272
+    },
+    {
+      "epoch": 0.290944,
+      "grad_norm": 0.4957731068134308,
+      "learning_rate": 0.0008318656003742313,
+      "loss": 3.0012,
+      "step": 2273
+    },
+    {
+      "epoch": 0.291072,
+      "grad_norm": 0.4984471797943115,
+      "learning_rate": 0.0008317105091117794,
+      "loss": 2.9733,
+      "step": 2274
+    },
+    {
+      "epoch": 0.2912,
+      "grad_norm": 0.5100799202919006,
+      "learning_rate": 0.0008315553608244453,
+      "loss": 2.8554,
+      "step": 2275
+    },
+    {
+      "epoch": 0.291328,
+      "grad_norm": 0.4962720274925232,
+      "learning_rate": 0.0008314001555389012,
+      "loss": 2.7421,
+      "step": 2276
+    },
+    {
+      "epoch": 0.291456,
+      "grad_norm": 0.4858592450618744,
+      "learning_rate": 0.0008312448932818283,
+      "loss": 2.926,
+      "step": 2277
+    },
+    {
+      "epoch": 0.291584,
+      "grad_norm": 0.5084744691848755,
+      "learning_rate": 0.0008310895740799181,
+      "loss": 3.0818,
+      "step": 2278
+    },
+    {
+      "epoch": 0.291712,
+      "grad_norm": 0.5392442941665649,
+      "learning_rate": 0.0008309341979598719,
+      "loss": 2.8487,
+      "step": 2279
+    },
+    {
+      "epoch": 0.29184,
+      "grad_norm": 0.5312164425849915,
+      "learning_rate": 0.0008307787649484004,
+      "loss": 3.035,
+      "step": 2280
+    },
+    {
+      "epoch": 0.291968,
+      "grad_norm": 0.4878517687320709,
+      "learning_rate": 0.0008306232750722247,
+      "loss": 2.9428,
+      "step": 2281
+    },
+    {
+      "epoch": 0.292096,
+      "grad_norm": 0.4959985315799713,
+      "learning_rate": 0.0008304677283580749,
+      "loss": 2.9427,
+      "step": 2282
+    },
+    {
+      "epoch": 0.292224,
+      "grad_norm": 0.49351736903190613,
+      "learning_rate": 0.0008303121248326915,
+      "loss": 2.8595,
+      "step": 2283
+    },
+    {
+      "epoch": 0.292352,
+      "grad_norm": 0.4854571521282196,
+      "learning_rate": 0.0008301564645228248,
+      "loss": 2.8919,
+      "step": 2284
+    },
+    {
+      "epoch": 0.29248,
+      "grad_norm": 0.5084532499313354,
+      "learning_rate": 0.0008300007474552341,
+      "loss": 2.9215,
+      "step": 2285
+    },
+    {
+      "epoch": 0.292608,
+      "grad_norm": 0.5293973088264465,
+      "learning_rate": 0.0008298449736566893,
+      "loss": 2.817,
+      "step": 2286
+    },
+    {
+      "epoch": 0.292736,
+      "grad_norm": 0.5711443424224854,
+      "learning_rate": 0.0008296891431539695,
+      "loss": 3.0618,
+      "step": 2287
+    },
+    {
+      "epoch": 0.292864,
+      "grad_norm": 0.5013183355331421,
+      "learning_rate": 0.0008295332559738641,
+      "loss": 2.8719,
+      "step": 2288
+    },
+    {
+      "epoch": 0.292992,
+      "grad_norm": 0.5161945819854736,
+      "learning_rate": 0.0008293773121431716,
+      "loss": 2.9993,
+      "step": 2289
+    },
+    {
+      "epoch": 0.29312,
+      "grad_norm": 0.554534375667572,
+      "learning_rate": 0.0008292213116887008,
+      "loss": 2.9232,
+      "step": 2290
+    },
+    {
+      "epoch": 0.293248,
+      "grad_norm": 0.5291886925697327,
+      "learning_rate": 0.0008290652546372696,
+      "loss": 2.8758,
+      "step": 2291
+    },
+    {
+      "epoch": 0.293376,
+      "grad_norm": 0.4893510043621063,
+      "learning_rate": 0.0008289091410157066,
+      "loss": 3.0398,
+      "step": 2292
+    },
+    {
+      "epoch": 0.293504,
+      "grad_norm": 0.552004337310791,
+      "learning_rate": 0.000828752970850849,
+      "loss": 2.8005,
+      "step": 2293
+    },
+    {
+      "epoch": 0.293632,
+      "grad_norm": 0.5638794302940369,
+      "learning_rate": 0.0008285967441695446,
+      "loss": 2.8842,
+      "step": 2294
+    },
+    {
+      "epoch": 0.29376,
+      "grad_norm": 0.514528214931488,
+      "learning_rate": 0.0008284404609986504,
+      "loss": 2.9063,
+      "step": 2295
+    },
+    {
+      "epoch": 0.293888,
+      "grad_norm": 0.5137476921081543,
+      "learning_rate": 0.0008282841213650332,
+      "loss": 2.8635,
+      "step": 2296
+    },
+    {
+      "epoch": 0.294016,
+      "grad_norm": 0.49658823013305664,
+      "learning_rate": 0.0008281277252955699,
+      "loss": 2.8399,
+      "step": 2297
+    },
+    {
+      "epoch": 0.294144,
+      "grad_norm": 0.49617335200309753,
+      "learning_rate": 0.0008279712728171466,
+      "loss": 3.0232,
+      "step": 2298
+    },
+    {
+      "epoch": 0.294272,
+      "grad_norm": 0.47013232111930847,
+      "learning_rate": 0.0008278147639566595,
+      "loss": 3.0137,
+      "step": 2299
+    },
+    {
+      "epoch": 0.2944,
+      "grad_norm": 0.4929681420326233,
+      "learning_rate": 0.0008276581987410139,
+      "loss": 3.0045,
+      "step": 2300
+    },
+    {
+      "epoch": 0.294528,
+      "grad_norm": 0.5088479518890381,
+      "learning_rate": 0.0008275015771971255,
+      "loss": 2.8244,
+      "step": 2301
+    },
+    {
+      "epoch": 0.294656,
+      "grad_norm": 0.47106486558914185,
+      "learning_rate": 0.0008273448993519193,
+      "loss": 2.8511,
+      "step": 2302
+    },
+    {
+      "epoch": 0.294784,
+      "grad_norm": 0.5212969183921814,
+      "learning_rate": 0.0008271881652323298,
+      "loss": 2.8504,
+      "step": 2303
+    },
+    {
+      "epoch": 0.294912,
+      "grad_norm": 0.5146804451942444,
+      "learning_rate": 0.0008270313748653018,
+      "loss": 2.8765,
+      "step": 2304
+    },
+    {
+      "epoch": 0.29504,
+      "grad_norm": 0.4860549569129944,
+      "learning_rate": 0.0008268745282777891,
+      "loss": 2.9851,
+      "step": 2305
+    },
+    {
+      "epoch": 0.295168,
+      "grad_norm": 0.5100993514060974,
+      "learning_rate": 0.0008267176254967556,
+      "loss": 3.0006,
+      "step": 2306
+    },
+    {
+      "epoch": 0.295296,
+      "grad_norm": 0.56208735704422,
+      "learning_rate": 0.0008265606665491745,
+      "loss": 2.7917,
+      "step": 2307
+    },
+    {
+      "epoch": 0.295424,
+      "grad_norm": 0.47039687633514404,
+      "learning_rate": 0.0008264036514620293,
+      "loss": 2.7671,
+      "step": 2308
+    },
+    {
+      "epoch": 0.295552,
+      "grad_norm": 0.5244276523590088,
+      "learning_rate": 0.0008262465802623121,
+      "loss": 2.8708,
+      "step": 2309
+    },
+    {
+      "epoch": 0.29568,
+      "grad_norm": 0.4734605848789215,
+      "learning_rate": 0.0008260894529770257,
+      "loss": 2.8369,
+      "step": 2310
+    },
+    {
+      "epoch": 0.295808,
+      "grad_norm": 0.519145667552948,
+      "learning_rate": 0.0008259322696331819,
+      "loss": 2.8175,
+      "step": 2311
+    },
+    {
+      "epoch": 0.295936,
+      "grad_norm": 0.49185317754745483,
+      "learning_rate": 0.0008257750302578026,
+      "loss": 3.0409,
+      "step": 2312
+    },
+    {
+      "epoch": 0.296064,
+      "grad_norm": 0.4860426187515259,
+      "learning_rate": 0.0008256177348779188,
+      "loss": 2.9497,
+      "step": 2313
+    },
+    {
+      "epoch": 0.296192,
+      "grad_norm": 0.6175550818443298,
+      "learning_rate": 0.0008254603835205714,
+      "loss": 2.9357,
+      "step": 2314
+    },
+    {
+      "epoch": 0.29632,
+      "grad_norm": 0.4957858622074127,
+      "learning_rate": 0.000825302976212811,
+      "loss": 2.9237,
+      "step": 2315
+    },
+    {
+      "epoch": 0.296448,
+      "grad_norm": 0.47185614705085754,
+      "learning_rate": 0.0008251455129816979,
+      "loss": 2.8828,
+      "step": 2316
+    },
+    {
+      "epoch": 0.296576,
+      "grad_norm": 0.5223645567893982,
+      "learning_rate": 0.0008249879938543017,
+      "loss": 2.7825,
+      "step": 2317
+    },
+    {
+      "epoch": 0.296704,
+      "grad_norm": 0.5006940364837646,
+      "learning_rate": 0.0008248304188577017,
+      "loss": 2.9035,
+      "step": 2318
+    },
+    {
+      "epoch": 0.296832,
+      "grad_norm": 0.5160130262374878,
+      "learning_rate": 0.0008246727880189869,
+      "loss": 2.9664,
+      "step": 2319
+    },
+    {
+      "epoch": 0.29696,
+      "grad_norm": 0.5200162529945374,
+      "learning_rate": 0.0008245151013652561,
+      "loss": 3.0148,
+      "step": 2320
+    },
+    {
+      "epoch": 0.297088,
+      "grad_norm": 0.492624968290329,
+      "learning_rate": 0.0008243573589236171,
+      "loss": 2.9711,
+      "step": 2321
+    },
+    {
+      "epoch": 0.297216,
+      "grad_norm": 0.5800040364265442,
+      "learning_rate": 0.0008241995607211879,
+      "loss": 2.8451,
+      "step": 2322
+    },
+    {
+      "epoch": 0.297344,
+      "grad_norm": 0.5232494473457336,
+      "learning_rate": 0.0008240417067850957,
+      "loss": 2.978,
+      "step": 2323
+    },
+    {
+      "epoch": 0.297472,
+      "grad_norm": 0.49256467819213867,
+      "learning_rate": 0.0008238837971424775,
+      "loss": 2.9963,
+      "step": 2324
+    },
+    {
+      "epoch": 0.2976,
+      "grad_norm": 0.549264669418335,
+      "learning_rate": 0.0008237258318204799,
+      "loss": 2.9924,
+      "step": 2325
+    },
+    {
+      "epoch": 0.297728,
+      "grad_norm": 0.49674519896507263,
+      "learning_rate": 0.0008235678108462587,
+      "loss": 2.8339,
+      "step": 2326
+    },
+    {
+      "epoch": 0.297856,
+      "grad_norm": 0.5176485776901245,
+      "learning_rate": 0.0008234097342469798,
+      "loss": 2.884,
+      "step": 2327
+    },
+    {
+      "epoch": 0.297984,
+      "grad_norm": 0.5019426941871643,
+      "learning_rate": 0.0008232516020498183,
+      "loss": 2.7923,
+      "step": 2328
+    },
+    {
+      "epoch": 0.298112,
+      "grad_norm": 0.4909980595111847,
+      "learning_rate": 0.0008230934142819588,
+      "loss": 2.8466,
+      "step": 2329
+    },
+    {
+      "epoch": 0.29824,
+      "grad_norm": 0.47746631503105164,
+      "learning_rate": 0.000822935170970596,
+      "loss": 2.7809,
+      "step": 2330
+    },
+    {
+      "epoch": 0.298368,
+      "grad_norm": 0.5025290250778198,
+      "learning_rate": 0.0008227768721429334,
+      "loss": 2.8521,
+      "step": 2331
+    },
+    {
+      "epoch": 0.298496,
+      "grad_norm": 1.2027992010116577,
+      "learning_rate": 0.0008226185178261846,
+      "loss": 2.9256,
+      "step": 2332
+    },
+    {
+      "epoch": 0.298624,
+      "grad_norm": 0.4587000906467438,
+      "learning_rate": 0.0008224601080475723,
+      "loss": 2.904,
+      "step": 2333
+    },
+    {
+      "epoch": 0.298752,
+      "grad_norm": 0.4788459241390228,
+      "learning_rate": 0.0008223016428343293,
+      "loss": 2.9985,
+      "step": 2334
+    },
+    {
+      "epoch": 0.29888,
+      "grad_norm": 0.48574501276016235,
+      "learning_rate": 0.0008221431222136975,
+      "loss": 2.876,
+      "step": 2335
+    },
+    {
+      "epoch": 0.299008,
+      "grad_norm": 0.4867737293243408,
+      "learning_rate": 0.0008219845462129283,
+      "loss": 2.9781,
+      "step": 2336
+    },
+    {
+      "epoch": 0.299136,
+      "grad_norm": 0.47482481598854065,
+      "learning_rate": 0.0008218259148592828,
+      "loss": 2.7588,
+      "step": 2337
+    },
+    {
+      "epoch": 0.299264,
+      "grad_norm": 0.4822710454463959,
+      "learning_rate": 0.0008216672281800316,
+      "loss": 2.8707,
+      "step": 2338
+    },
+    {
+      "epoch": 0.299392,
+      "grad_norm": 0.5051573514938354,
+      "learning_rate": 0.0008215084862024549,
+      "loss": 2.9221,
+      "step": 2339
+    },
+    {
+      "epoch": 0.29952,
+      "grad_norm": 0.472768098115921,
+      "learning_rate": 0.0008213496889538421,
+      "loss": 2.9064,
+      "step": 2340
+    },
+    {
+      "epoch": 0.299648,
+      "grad_norm": 0.49367108941078186,
+      "learning_rate": 0.0008211908364614924,
+      "loss": 2.9728,
+      "step": 2341
+    },
+    {
+      "epoch": 0.299776,
+      "grad_norm": 0.5038679242134094,
+      "learning_rate": 0.0008210319287527143,
+      "loss": 2.8064,
+      "step": 2342
+    },
+    {
+      "epoch": 0.299904,
+      "grad_norm": 0.4789201021194458,
+      "learning_rate": 0.000820872965854826,
+      "loss": 2.9503,
+      "step": 2343
+    },
+    {
+      "epoch": 0.300032,
+      "grad_norm": 0.7603449821472168,
+      "learning_rate": 0.0008207139477951549,
+      "loss": 2.909,
+      "step": 2344
+    },
+    {
+      "epoch": 0.30016,
+      "grad_norm": 0.4836367964744568,
+      "learning_rate": 0.0008205548746010382,
+      "loss": 2.8508,
+      "step": 2345
+    },
+    {
+      "epoch": 0.300288,
+      "grad_norm": 0.5140978097915649,
+      "learning_rate": 0.0008203957462998224,
+      "loss": 2.9975,
+      "step": 2346
+    },
+    {
+      "epoch": 0.300416,
+      "grad_norm": 0.47527194023132324,
+      "learning_rate": 0.0008202365629188633,
+      "loss": 2.9685,
+      "step": 2347
+    },
+    {
+      "epoch": 0.300544,
+      "grad_norm": 0.47923389077186584,
+      "learning_rate": 0.0008200773244855267,
+      "loss": 2.9502,
+      "step": 2348
+    },
+    {
+      "epoch": 0.300672,
+      "grad_norm": 0.5184609889984131,
+      "learning_rate": 0.0008199180310271873,
+      "loss": 2.941,
+      "step": 2349
+    },
+    {
+      "epoch": 0.3008,
+      "grad_norm": 0.4521273672580719,
+      "learning_rate": 0.0008197586825712295,
+      "loss": 2.986,
+      "step": 2350
+    },
+    {
+      "epoch": 0.300928,
+      "grad_norm": 0.5023333430290222,
+      "learning_rate": 0.0008195992791450474,
+      "loss": 2.8959,
+      "step": 2351
+    },
+    {
+      "epoch": 0.301056,
+      "grad_norm": 0.4809386432170868,
+      "learning_rate": 0.0008194398207760439,
+      "loss": 2.866,
+      "step": 2352
+    },
+    {
+      "epoch": 0.301184,
+      "grad_norm": 0.513736367225647,
+      "learning_rate": 0.0008192803074916321,
+      "loss": 3.0785,
+      "step": 2353
+    },
+    {
+      "epoch": 0.301312,
+      "grad_norm": 0.4863099157810211,
+      "learning_rate": 0.0008191207393192339,
+      "loss": 2.9924,
+      "step": 2354
+    },
+    {
+      "epoch": 0.30144,
+      "grad_norm": 0.53351891040802,
+      "learning_rate": 0.0008189611162862812,
+      "loss": 3.0079,
+      "step": 2355
+    },
+    {
+      "epoch": 0.301568,
+      "grad_norm": 0.49441277980804443,
+      "learning_rate": 0.0008188014384202147,
+      "loss": 2.8675,
+      "step": 2356
+    },
+    {
+      "epoch": 0.301696,
+      "grad_norm": 0.5466651916503906,
+      "learning_rate": 0.0008186417057484852,
+      "loss": 2.8584,
+      "step": 2357
+    },
+    {
+      "epoch": 0.301824,
+      "grad_norm": 0.496905654668808,
+      "learning_rate": 0.0008184819182985524,
+      "loss": 2.8819,
+      "step": 2358
+    },
+    {
+      "epoch": 0.301952,
+      "grad_norm": 0.5154985785484314,
+      "learning_rate": 0.0008183220760978858,
+      "loss": 2.8035,
+      "step": 2359
+    },
+    {
+      "epoch": 0.30208,
+      "grad_norm": 0.47623082995414734,
+      "learning_rate": 0.000818162179173964,
+      "loss": 3.0158,
+      "step": 2360
+    },
+    {
+      "epoch": 0.302208,
+      "grad_norm": 0.5186389088630676,
+      "learning_rate": 0.0008180022275542749,
+      "loss": 3.0951,
+      "step": 2361
+    },
+    {
+      "epoch": 0.302336,
+      "grad_norm": 0.5449215173721313,
+      "learning_rate": 0.0008178422212663166,
+      "loss": 2.9131,
+      "step": 2362
+    },
+    {
+      "epoch": 0.302464,
+      "grad_norm": 0.4795946776866913,
+      "learning_rate": 0.0008176821603375954,
+      "loss": 2.9311,
+      "step": 2363
+    },
+    {
+      "epoch": 0.302592,
+      "grad_norm": 0.5176295638084412,
+      "learning_rate": 0.0008175220447956281,
+      "loss": 2.8921,
+      "step": 2364
+    },
+    {
+      "epoch": 0.30272,
+      "grad_norm": 0.4621509611606598,
+      "learning_rate": 0.0008173618746679406,
+      "loss": 2.9069,
+      "step": 2365
+    },
+    {
+      "epoch": 0.302848,
+      "grad_norm": 0.5111235976219177,
+      "learning_rate": 0.0008172016499820671,
+      "loss": 3.0183,
+      "step": 2366
+    },
+    {
+      "epoch": 0.302976,
+      "grad_norm": 0.5597186088562012,
+      "learning_rate": 0.000817041370765553,
+      "loss": 2.9735,
+      "step": 2367
+    },
+    {
+      "epoch": 0.303104,
+      "grad_norm": 0.48399439454078674,
+      "learning_rate": 0.0008168810370459518,
+      "loss": 2.8826,
+      "step": 2368
+    },
+    {
+      "epoch": 0.303232,
+      "grad_norm": 0.5399102568626404,
+      "learning_rate": 0.0008167206488508267,
+      "loss": 2.9327,
+      "step": 2369
+    },
+    {
+      "epoch": 0.30336,
+      "grad_norm": 0.4973888099193573,
+      "learning_rate": 0.0008165602062077502,
+      "loss": 2.937,
+      "step": 2370
+    },
+    {
+      "epoch": 0.303488,
+      "grad_norm": 0.48604968190193176,
+      "learning_rate": 0.0008163997091443045,
+      "loss": 2.8089,
+      "step": 2371
+    },
+    {
+      "epoch": 0.303616,
+      "grad_norm": 0.508804976940155,
+      "learning_rate": 0.0008162391576880808,
+      "loss": 2.9842,
+      "step": 2372
+    },
+    {
+      "epoch": 0.303744,
+      "grad_norm": 0.5130351781845093,
+      "learning_rate": 0.0008160785518666796,
+      "loss": 2.876,
+      "step": 2373
+    },
+    {
+      "epoch": 0.303872,
+      "grad_norm": 0.48926323652267456,
+      "learning_rate": 0.0008159178917077111,
+      "loss": 2.9872,
+      "step": 2374
+    },
+    {
+      "epoch": 0.304,
+      "grad_norm": 0.46356263756752014,
+      "learning_rate": 0.0008157571772387946,
+      "loss": 2.8154,
+      "step": 2375
+    },
+    {
+      "epoch": 0.304128,
+      "grad_norm": 0.49358996748924255,
+      "learning_rate": 0.0008155964084875587,
+      "loss": 2.9669,
+      "step": 2376
+    },
+    {
+      "epoch": 0.304256,
+      "grad_norm": 0.5273545980453491,
+      "learning_rate": 0.0008154355854816416,
+      "loss": 3.044,
+      "step": 2377
+    },
+    {
+      "epoch": 0.304384,
+      "grad_norm": 0.5308858156204224,
+      "learning_rate": 0.0008152747082486904,
+      "loss": 2.972,
+      "step": 2378
+    },
+    {
+      "epoch": 0.304512,
+      "grad_norm": 0.467502623796463,
+      "learning_rate": 0.0008151137768163619,
+      "loss": 2.9017,
+      "step": 2379
+    },
+    {
+      "epoch": 0.30464,
+      "grad_norm": 0.49239829182624817,
+      "learning_rate": 0.000814952791212322,
+      "loss": 2.9274,
+      "step": 2380
+    },
+    {
+      "epoch": 0.304768,
+      "grad_norm": 0.5434689521789551,
+      "learning_rate": 0.0008147917514642462,
+      "loss": 2.948,
+      "step": 2381
+    },
+    {
+      "epoch": 0.304896,
+      "grad_norm": 0.5068210959434509,
+      "learning_rate": 0.0008146306575998188,
+      "loss": 2.971,
+      "step": 2382
+    },
+    {
+      "epoch": 0.305024,
+      "grad_norm": 0.4755307734012604,
+      "learning_rate": 0.0008144695096467341,
+      "loss": 2.9878,
+      "step": 2383
+    },
+    {
+      "epoch": 0.305152,
+      "grad_norm": 0.4968573749065399,
+      "learning_rate": 0.0008143083076326947,
+      "loss": 2.9569,
+      "step": 2384
+    },
+    {
+      "epoch": 0.30528,
+      "grad_norm": 0.4792605936527252,
+      "learning_rate": 0.0008141470515854135,
+      "loss": 2.8245,
+      "step": 2385
+    },
+    {
+      "epoch": 0.305408,
+      "grad_norm": 0.4739339053630829,
+      "learning_rate": 0.0008139857415326124,
+      "loss": 2.9302,
+      "step": 2386
+    },
+    {
+      "epoch": 0.305536,
+      "grad_norm": 0.4726722836494446,
+      "learning_rate": 0.0008138243775020223,
+      "loss": 2.8736,
+      "step": 2387
+    },
+    {
+      "epoch": 0.305664,
+      "grad_norm": 0.49579352140426636,
+      "learning_rate": 0.0008136629595213834,
+      "loss": 2.8697,
+      "step": 2388
+    },
+    {
+      "epoch": 0.305792,
+      "grad_norm": 0.4968126714229584,
+      "learning_rate": 0.0008135014876184454,
+      "loss": 2.8891,
+      "step": 2389
+    },
+    {
+      "epoch": 0.30592,
+      "grad_norm": 0.49273163080215454,
+      "learning_rate": 0.0008133399618209674,
+      "loss": 2.9602,
+      "step": 2390
+    },
+    {
+      "epoch": 0.306048,
+      "grad_norm": 0.5092888474464417,
+      "learning_rate": 0.0008131783821567174,
+      "loss": 2.919,
+      "step": 2391
+    },
+    {
+      "epoch": 0.306176,
+      "grad_norm": 0.5169079303741455,
+      "learning_rate": 0.0008130167486534729,
+      "loss": 2.8856,
+      "step": 2392
+    },
+    {
+      "epoch": 0.306304,
+      "grad_norm": 0.4822143316268921,
+      "learning_rate": 0.0008128550613390205,
+      "loss": 2.9249,
+      "step": 2393
+    },
+    {
+      "epoch": 0.306432,
+      "grad_norm": 0.5248920917510986,
+      "learning_rate": 0.0008126933202411559,
+      "loss": 2.8843,
+      "step": 2394
+    },
+    {
+      "epoch": 0.30656,
+      "grad_norm": 0.49162906408309937,
+      "learning_rate": 0.0008125315253876849,
+      "loss": 2.8837,
+      "step": 2395
+    },
+    {
+      "epoch": 0.306688,
+      "grad_norm": 0.516188383102417,
+      "learning_rate": 0.0008123696768064212,
+      "loss": 2.8382,
+      "step": 2396
+    },
+    {
+      "epoch": 0.306816,
+      "grad_norm": 0.4689294993877411,
+      "learning_rate": 0.0008122077745251888,
+      "loss": 2.8518,
+      "step": 2397
+    },
+    {
+      "epoch": 0.306944,
+      "grad_norm": 0.4818526804447174,
+      "learning_rate": 0.0008120458185718205,
+      "loss": 2.9277,
+      "step": 2398
+    },
+    {
+      "epoch": 0.307072,
+      "grad_norm": 0.4924405813217163,
+      "learning_rate": 0.0008118838089741585,
+      "loss": 2.7618,
+      "step": 2399
+    },
+    {
+      "epoch": 0.3072,
+      "grad_norm": 0.49621692299842834,
+      "learning_rate": 0.0008117217457600541,
+      "loss": 2.802,
+      "step": 2400
+    },
+    {
+      "epoch": 0.307328,
+      "grad_norm": 0.4668716490268707,
+      "learning_rate": 0.0008115596289573678,
+      "loss": 2.9158,
+      "step": 2401
+    },
+    {
+      "epoch": 0.307456,
+      "grad_norm": 0.49292874336242676,
+      "learning_rate": 0.0008113974585939694,
+      "loss": 2.8966,
+      "step": 2402
+    },
+    {
+      "epoch": 0.307584,
+      "grad_norm": 0.49537140130996704,
+      "learning_rate": 0.0008112352346977377,
+      "loss": 2.946,
+      "step": 2403
+    },
+    {
+      "epoch": 0.307712,
+      "grad_norm": 0.46481630206108093,
+      "learning_rate": 0.0008110729572965612,
+      "loss": 2.8531,
+      "step": 2404
+    },
+    {
+      "epoch": 0.30784,
+      "grad_norm": 0.46709275245666504,
+      "learning_rate": 0.0008109106264183369,
+      "loss": 2.9547,
+      "step": 2405
+    },
+    {
+      "epoch": 0.307968,
+      "grad_norm": 0.46974191069602966,
+      "learning_rate": 0.0008107482420909718,
+      "loss": 2.8381,
+      "step": 2406
+    },
+    {
+      "epoch": 0.308096,
+      "grad_norm": 0.4874151349067688,
+      "learning_rate": 0.0008105858043423811,
+      "loss": 2.9185,
+      "step": 2407
+    },
+    {
+      "epoch": 0.308224,
+      "grad_norm": 0.5290341973304749,
+      "learning_rate": 0.0008104233132004901,
+      "loss": 3.0612,
+      "step": 2408
+    },
+    {
+      "epoch": 0.308352,
+      "grad_norm": 0.4832834303379059,
+      "learning_rate": 0.0008102607686932328,
+      "loss": 2.843,
+      "step": 2409
+    },
+    {
+      "epoch": 0.30848,
+      "grad_norm": 0.4897901117801666,
+      "learning_rate": 0.0008100981708485526,
+      "loss": 2.8084,
+      "step": 2410
+    },
+    {
+      "epoch": 0.308608,
+      "grad_norm": 0.5063967704772949,
+      "learning_rate": 0.0008099355196944018,
+      "loss": 2.8963,
+      "step": 2411
+    },
+    {
+      "epoch": 0.308736,
+      "grad_norm": 0.5579289793968201,
+      "learning_rate": 0.000809772815258742,
+      "loss": 2.9437,
+      "step": 2412
+    },
+    {
+      "epoch": 0.308864,
+      "grad_norm": 0.5106964111328125,
+      "learning_rate": 0.0008096100575695442,
+      "loss": 2.8636,
+      "step": 2413
+    },
+    {
+      "epoch": 0.308992,
+      "grad_norm": 0.5232910513877869,
+      "learning_rate": 0.0008094472466547882,
+      "loss": 3.0101,
+      "step": 2414
+    },
+    {
+      "epoch": 0.30912,
+      "grad_norm": 0.5063114762306213,
+      "learning_rate": 0.0008092843825424629,
+      "loss": 3.0431,
+      "step": 2415
+    },
+    {
+      "epoch": 0.309248,
+      "grad_norm": 0.4977165758609772,
+      "learning_rate": 0.0008091214652605669,
+      "loss": 3.0127,
+      "step": 2416
+    },
+    {
+      "epoch": 0.309376,
+      "grad_norm": 0.48981809616088867,
+      "learning_rate": 0.0008089584948371074,
+      "loss": 3.0093,
+      "step": 2417
+    },
+    {
+      "epoch": 0.309504,
+      "grad_norm": 0.45608076453208923,
+      "learning_rate": 0.0008087954713001007,
+      "loss": 2.7939,
+      "step": 2418
+    },
+    {
+      "epoch": 0.309632,
+      "grad_norm": 0.4704652726650238,
+      "learning_rate": 0.0008086323946775727,
+      "loss": 2.9254,
+      "step": 2419
+    },
+    {
+      "epoch": 0.30976,
+      "grad_norm": 0.5216735601425171,
+      "learning_rate": 0.0008084692649975582,
+      "loss": 2.7304,
+      "step": 2420
+    },
+    {
+      "epoch": 0.309888,
+      "grad_norm": 0.5145627856254578,
+      "learning_rate": 0.0008083060822881008,
+      "loss": 2.9664,
+      "step": 2421
+    },
+    {
+      "epoch": 0.310016,
+      "grad_norm": 0.5313540697097778,
+      "learning_rate": 0.0008081428465772538,
+      "loss": 2.935,
+      "step": 2422
+    },
+    {
+      "epoch": 0.310144,
+      "grad_norm": 0.5123348832130432,
+      "learning_rate": 0.0008079795578930791,
+      "loss": 2.868,
+      "step": 2423
+    },
+    {
+      "epoch": 0.310272,
+      "grad_norm": 0.49067458510398865,
+      "learning_rate": 0.000807816216263648,
+      "loss": 2.9656,
+      "step": 2424
+    },
+    {
+      "epoch": 0.3104,
+      "grad_norm": 0.5014142394065857,
+      "learning_rate": 0.0008076528217170408,
+      "loss": 2.9094,
+      "step": 2425
+    },
+    {
+      "epoch": 0.310528,
+      "grad_norm": 0.4593740999698639,
+      "learning_rate": 0.0008074893742813469,
+      "loss": 2.9674,
+      "step": 2426
+    },
+    {
+      "epoch": 0.310656,
+      "grad_norm": 0.524004340171814,
+      "learning_rate": 0.0008073258739846649,
+      "loss": 2.9353,
+      "step": 2427
+    },
+    {
+      "epoch": 0.310784,
+      "grad_norm": 0.49850088357925415,
+      "learning_rate": 0.0008071623208551023,
+      "loss": 2.8629,
+      "step": 2428
+    },
+    {
+      "epoch": 0.310912,
+      "grad_norm": 0.4967794120311737,
+      "learning_rate": 0.0008069987149207759,
+      "loss": 2.9594,
+      "step": 2429
+    },
+    {
+      "epoch": 0.31104,
+      "grad_norm": 0.4760918617248535,
+      "learning_rate": 0.0008068350562098112,
+      "loss": 2.8991,
+      "step": 2430
+    },
+    {
+      "epoch": 0.311168,
+      "grad_norm": 0.5188255310058594,
+      "learning_rate": 0.0008066713447503433,
+      "loss": 2.8986,
+      "step": 2431
+    },
+    {
+      "epoch": 0.311296,
+      "grad_norm": 0.47959768772125244,
+      "learning_rate": 0.000806507580570516,
+      "loss": 2.8933,
+      "step": 2432
+    },
+    {
+      "epoch": 0.311424,
+      "grad_norm": 0.48403486609458923,
+      "learning_rate": 0.0008063437636984823,
+      "loss": 2.8218,
+      "step": 2433
+    },
+    {
+      "epoch": 0.311552,
+      "grad_norm": 0.4924117624759674,
+      "learning_rate": 0.0008061798941624041,
+      "loss": 2.8877,
+      "step": 2434
+    },
+    {
+      "epoch": 0.31168,
+      "grad_norm": 0.5012858510017395,
+      "learning_rate": 0.0008060159719904525,
+      "loss": 2.7562,
+      "step": 2435
+    },
+    {
+      "epoch": 0.311808,
+      "grad_norm": 0.46351921558380127,
+      "learning_rate": 0.0008058519972108079,
+      "loss": 2.9356,
+      "step": 2436
+    },
+    {
+      "epoch": 0.311936,
+      "grad_norm": 0.5388748645782471,
+      "learning_rate": 0.0008056879698516591,
+      "loss": 2.9556,
+      "step": 2437
+    },
+    {
+      "epoch": 0.312064,
+      "grad_norm": 0.48392948508262634,
+      "learning_rate": 0.0008055238899412045,
+      "loss": 2.9313,
+      "step": 2438
+    },
+    {
+      "epoch": 0.312192,
+      "grad_norm": 0.47482502460479736,
+      "learning_rate": 0.0008053597575076512,
+      "loss": 2.8856,
+      "step": 2439
+    },
+    {
+      "epoch": 0.31232,
+      "grad_norm": 0.5010195970535278,
+      "learning_rate": 0.0008051955725792155,
+      "loss": 2.8968,
+      "step": 2440
+    },
+    {
+      "epoch": 0.312448,
+      "grad_norm": 0.4686266779899597,
+      "learning_rate": 0.0008050313351841228,
+      "loss": 2.9029,
+      "step": 2441
+    },
+    {
+      "epoch": 0.312576,
+      "grad_norm": 0.45737335085868835,
+      "learning_rate": 0.0008048670453506074,
+      "loss": 2.824,
+      "step": 2442
+    },
+    {
+      "epoch": 0.312704,
+      "grad_norm": 0.4832533299922943,
+      "learning_rate": 0.0008047027031069126,
+      "loss": 2.829,
+      "step": 2443
+    },
+    {
+      "epoch": 0.312832,
+      "grad_norm": 0.48089537024497986,
+      "learning_rate": 0.0008045383084812906,
+      "loss": 2.9534,
+      "step": 2444
+    },
+    {
+      "epoch": 0.31296,
+      "grad_norm": 0.4302358329296112,
+      "learning_rate": 0.0008043738615020029,
+      "loss": 2.8033,
+      "step": 2445
+    },
+    {
+      "epoch": 0.313088,
+      "grad_norm": 0.48767080903053284,
+      "learning_rate": 0.0008042093621973198,
+      "loss": 2.9239,
+      "step": 2446
+    },
+    {
+      "epoch": 0.313216,
+      "grad_norm": 0.44325512647628784,
+      "learning_rate": 0.0008040448105955208,
+      "loss": 2.9894,
+      "step": 2447
+    },
+    {
+      "epoch": 0.313344,
+      "grad_norm": 0.48578137159347534,
+      "learning_rate": 0.0008038802067248939,
+      "loss": 2.9394,
+      "step": 2448
+    },
+    {
+      "epoch": 0.313472,
+      "grad_norm": 0.4667922556400299,
+      "learning_rate": 0.0008037155506137366,
+      "loss": 2.7359,
+      "step": 2449
+    },
+    {
+      "epoch": 0.3136,
+      "grad_norm": 0.5178252458572388,
+      "learning_rate": 0.0008035508422903553,
+      "loss": 2.9129,
+      "step": 2450
+    },
+    {
+      "epoch": 0.313728,
+      "grad_norm": 0.4937523305416107,
+      "learning_rate": 0.000803386081783065,
+      "loss": 2.8897,
+      "step": 2451
+    },
+    {
+      "epoch": 0.313856,
+      "grad_norm": 0.45950478315353394,
+      "learning_rate": 0.0008032212691201901,
+      "loss": 2.8094,
+      "step": 2452
+    },
+    {
+      "epoch": 0.313984,
+      "grad_norm": 0.46415990591049194,
+      "learning_rate": 0.0008030564043300639,
+      "loss": 2.893,
+      "step": 2453
+    },
+    {
+      "epoch": 0.314112,
+      "grad_norm": 0.4707229435443878,
+      "learning_rate": 0.0008028914874410283,
+      "loss": 2.9179,
+      "step": 2454
+    },
+    {
+      "epoch": 0.31424,
+      "grad_norm": 0.46848633885383606,
+      "learning_rate": 0.0008027265184814348,
+      "loss": 2.8841,
+      "step": 2455
+    },
+    {
+      "epoch": 0.314368,
+      "grad_norm": 0.4920593202114105,
+      "learning_rate": 0.0008025614974796431,
+      "loss": 2.9225,
+      "step": 2456
+    },
+    {
+      "epoch": 0.314496,
+      "grad_norm": 0.44019538164138794,
+      "learning_rate": 0.0008023964244640224,
+      "loss": 2.7868,
+      "step": 2457
+    },
+    {
+      "epoch": 0.314624,
+      "grad_norm": 0.4505968987941742,
+      "learning_rate": 0.0008022312994629508,
+      "loss": 2.8118,
+      "step": 2458
+    },
+    {
+      "epoch": 0.314752,
+      "grad_norm": 0.49876928329467773,
+      "learning_rate": 0.0008020661225048149,
+      "loss": 3.0642,
+      "step": 2459
+    },
+    {
+      "epoch": 0.31488,
+      "grad_norm": 0.4765637218952179,
+      "learning_rate": 0.0008019008936180108,
+      "loss": 2.9723,
+      "step": 2460
+    },
+    {
+      "epoch": 0.315008,
+      "grad_norm": 0.5069573521614075,
+      "learning_rate": 0.0008017356128309431,
+      "loss": 2.8225,
+      "step": 2461
+    },
+    {
+      "epoch": 0.315136,
+      "grad_norm": 0.5046855807304382,
+      "learning_rate": 0.0008015702801720255,
+      "loss": 2.8858,
+      "step": 2462
+    },
+    {
+      "epoch": 0.315264,
+      "grad_norm": 0.5210908055305481,
+      "learning_rate": 0.0008014048956696807,
+      "loss": 2.8798,
+      "step": 2463
+    },
+    {
+      "epoch": 0.315392,
+      "grad_norm": 0.48583316802978516,
+      "learning_rate": 0.0008012394593523403,
+      "loss": 2.8664,
+      "step": 2464
+    },
+    {
+      "epoch": 0.31552,
+      "grad_norm": 0.4979773163795471,
+      "learning_rate": 0.0008010739712484443,
+      "loss": 2.9316,
+      "step": 2465
+    },
+    {
+      "epoch": 0.315648,
+      "grad_norm": 0.49733492732048035,
+      "learning_rate": 0.0008009084313864423,
+      "loss": 2.8084,
+      "step": 2466
+    },
+    {
+      "epoch": 0.315776,
+      "grad_norm": 0.473783940076828,
+      "learning_rate": 0.0008007428397947925,
+      "loss": 2.8423,
+      "step": 2467
+    },
+    {
+      "epoch": 0.315904,
+      "grad_norm": 0.5263046622276306,
+      "learning_rate": 0.0008005771965019622,
+      "loss": 2.7864,
+      "step": 2468
+    },
+    {
+      "epoch": 0.316032,
+      "grad_norm": 0.49627596139907837,
+      "learning_rate": 0.0008004115015364272,
+      "loss": 2.963,
+      "step": 2469
+    },
+    {
+      "epoch": 0.31616,
+      "grad_norm": 0.5166580080986023,
+      "learning_rate": 0.0008002457549266724,
+      "loss": 2.8578,
+      "step": 2470
+    },
+    {
+      "epoch": 0.316288,
+      "grad_norm": 0.5513797402381897,
+      "learning_rate": 0.0008000799567011914,
+      "loss": 2.8514,
+      "step": 2471
+    },
+    {
+      "epoch": 0.316416,
+      "grad_norm": 0.49006354808807373,
+      "learning_rate": 0.0007999141068884872,
+      "loss": 2.8871,
+      "step": 2472
+    },
+    {
+      "epoch": 0.316544,
+      "grad_norm": 0.5233575701713562,
+      "learning_rate": 0.0007997482055170712,
+      "loss": 2.9514,
+      "step": 2473
+    },
+    {
+      "epoch": 0.316672,
+      "grad_norm": 0.48910221457481384,
+      "learning_rate": 0.0007995822526154636,
+      "loss": 2.8618,
+      "step": 2474
+    },
+    {
+      "epoch": 0.3168,
+      "grad_norm": 0.5080544948577881,
+      "learning_rate": 0.0007994162482121936,
+      "loss": 3.0055,
+      "step": 2475
+    },
+    {
+      "epoch": 0.316928,
+      "grad_norm": 0.4739373028278351,
+      "learning_rate": 0.0007992501923357997,
+      "loss": 2.7848,
+      "step": 2476
+    },
+    {
+      "epoch": 0.317056,
+      "grad_norm": 0.49846142530441284,
+      "learning_rate": 0.0007990840850148282,
+      "loss": 2.8935,
+      "step": 2477
+    },
+    {
+      "epoch": 0.317184,
+      "grad_norm": 0.47560712695121765,
+      "learning_rate": 0.0007989179262778355,
+      "loss": 2.8116,
+      "step": 2478
+    },
+    {
+      "epoch": 0.317312,
+      "grad_norm": 0.4787268042564392,
+      "learning_rate": 0.0007987517161533857,
+      "loss": 2.777,
+      "step": 2479
+    },
+    {
+      "epoch": 0.31744,
+      "grad_norm": 0.5078412294387817,
+      "learning_rate": 0.0007985854546700525,
+      "loss": 2.9354,
+      "step": 2480
+    },
+    {
+      "epoch": 0.317568,
+      "grad_norm": 0.5310080051422119,
+      "learning_rate": 0.0007984191418564182,
+      "loss": 2.8267,
+      "step": 2481
+    },
+    {
+      "epoch": 0.317696,
+      "grad_norm": 0.49502506852149963,
+      "learning_rate": 0.0007982527777410737,
+      "loss": 2.9275,
+      "step": 2482
+    },
+    {
+      "epoch": 0.317824,
+      "grad_norm": 0.5073422193527222,
+      "learning_rate": 0.0007980863623526193,
+      "loss": 3.0924,
+      "step": 2483
+    },
+    {
+      "epoch": 0.317952,
+      "grad_norm": 0.5026443600654602,
+      "learning_rate": 0.0007979198957196634,
+      "loss": 2.8037,
+      "step": 2484
+    },
+    {
+      "epoch": 0.31808,
+      "grad_norm": 0.45057377219200134,
+      "learning_rate": 0.0007977533778708236,
+      "loss": 2.9443,
+      "step": 2485
+    },
+    {
+      "epoch": 0.318208,
+      "grad_norm": 0.4696612060070038,
+      "learning_rate": 0.0007975868088347263,
+      "loss": 2.8716,
+      "step": 2486
+    },
+    {
+      "epoch": 0.318336,
+      "grad_norm": 0.4621131122112274,
+      "learning_rate": 0.0007974201886400066,
+      "loss": 2.9602,
+      "step": 2487
+    },
+    {
+      "epoch": 0.318464,
+      "grad_norm": 0.5243457555770874,
+      "learning_rate": 0.0007972535173153086,
+      "loss": 2.8543,
+      "step": 2488
+    },
+    {
+      "epoch": 0.318592,
+      "grad_norm": 0.4850609600543976,
+      "learning_rate": 0.0007970867948892848,
+      "loss": 2.9606,
+      "step": 2489
+    },
+    {
+      "epoch": 0.31872,
+      "grad_norm": 0.49778926372528076,
+      "learning_rate": 0.000796920021390597,
+      "loss": 2.7507,
+      "step": 2490
+    },
+    {
+      "epoch": 0.318848,
+      "grad_norm": 0.5029168128967285,
+      "learning_rate": 0.0007967531968479151,
+      "loss": 3.0214,
+      "step": 2491
+    },
+    {
+      "epoch": 0.318976,
+      "grad_norm": 0.4638972878456116,
+      "learning_rate": 0.0007965863212899185,
+      "loss": 2.9995,
+      "step": 2492
+    },
+    {
+      "epoch": 0.319104,
+      "grad_norm": 0.5274518132209778,
+      "learning_rate": 0.0007964193947452949,
+      "loss": 2.8892,
+      "step": 2493
+    },
+    {
+      "epoch": 0.319232,
+      "grad_norm": 0.4662211835384369,
+      "learning_rate": 0.0007962524172427411,
+      "loss": 2.9489,
+      "step": 2494
+    },
+    {
+      "epoch": 0.31936,
+      "grad_norm": 0.40964534878730774,
+      "learning_rate": 0.0007960853888109622,
+      "loss": 2.8247,
+      "step": 2495
+    },
+    {
+      "epoch": 0.319488,
+      "grad_norm": 0.4678458869457245,
+      "learning_rate": 0.0007959183094786725,
+      "loss": 3.0526,
+      "step": 2496
+    },
+    {
+      "epoch": 0.319616,
+      "grad_norm": 0.4730255901813507,
+      "learning_rate": 0.0007957511792745949,
+      "loss": 2.7679,
+      "step": 2497
+    },
+    {
+      "epoch": 0.319744,
+      "grad_norm": 0.5288103222846985,
+      "learning_rate": 0.0007955839982274607,
+      "loss": 2.9244,
+      "step": 2498
+    },
+    {
+      "epoch": 0.319872,
+      "grad_norm": 0.5185258984565735,
+      "learning_rate": 0.0007954167663660108,
+      "loss": 2.8106,
+      "step": 2499
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5098527073860168,
+      "learning_rate": 0.0007952494837189937,
+      "loss": 2.946,
+      "step": 2500
+    },
+    {
+      "epoch": 0.320128,
+      "grad_norm": 0.49485132098197937,
+      "learning_rate": 0.0007950821503151678,
+      "loss": 3.04,
+      "step": 2501
+    },
+    {
+      "epoch": 0.320256,
+      "grad_norm": 0.4809635877609253,
+      "learning_rate": 0.0007949147661832992,
+      "loss": 3.0026,
+      "step": 2502
+    },
+    {
+      "epoch": 0.320384,
+      "grad_norm": 0.4584812819957733,
+      "learning_rate": 0.0007947473313521634,
+      "loss": 3.0586,
+      "step": 2503
+    },
+    {
+      "epoch": 0.320512,
+      "grad_norm": 0.48823538422584534,
+      "learning_rate": 0.0007945798458505442,
+      "loss": 2.8497,
+      "step": 2504
+    },
+    {
+      "epoch": 0.32064,
+      "grad_norm": 0.4633082449436188,
+      "learning_rate": 0.0007944123097072347,
+      "loss": 2.9931,
+      "step": 2505
+    },
+    {
+      "epoch": 0.320768,
+      "grad_norm": 0.46133753657341003,
+      "learning_rate": 0.0007942447229510359,
+      "loss": 2.8489,
+      "step": 2506
+    },
+    {
+      "epoch": 0.320896,
+      "grad_norm": 0.47339701652526855,
+      "learning_rate": 0.000794077085610758,
+      "loss": 2.9704,
+      "step": 2507
+    },
+    {
+      "epoch": 0.321024,
+      "grad_norm": 0.5004094839096069,
+      "learning_rate": 0.00079390939771522,
+      "loss": 2.8169,
+      "step": 2508
+    },
+    {
+      "epoch": 0.321152,
+      "grad_norm": 0.4719487428665161,
+      "learning_rate": 0.000793741659293249,
+      "loss": 2.8514,
+      "step": 2509
+    },
+    {
+      "epoch": 0.32128,
+      "grad_norm": 0.4801078736782074,
+      "learning_rate": 0.0007935738703736817,
+      "loss": 2.8623,
+      "step": 2510
+    },
+    {
+      "epoch": 0.321408,
+      "grad_norm": 0.4726249575614929,
+      "learning_rate": 0.0007934060309853626,
+      "loss": 2.9501,
+      "step": 2511
+    },
+    {
+      "epoch": 0.321536,
+      "grad_norm": 0.4488205015659332,
+      "learning_rate": 0.0007932381411571453,
+      "loss": 2.8661,
+      "step": 2512
+    },
+    {
+      "epoch": 0.321664,
+      "grad_norm": 0.5044283866882324,
+      "learning_rate": 0.0007930702009178918,
+      "loss": 2.8469,
+      "step": 2513
+    },
+    {
+      "epoch": 0.321792,
+      "grad_norm": 0.48469874262809753,
+      "learning_rate": 0.0007929022102964734,
+      "loss": 2.8627,
+      "step": 2514
+    },
+    {
+      "epoch": 0.32192,
+      "grad_norm": 0.47914919257164,
+      "learning_rate": 0.0007927341693217695,
+      "loss": 2.9556,
+      "step": 2515
+    },
+    {
+      "epoch": 0.322048,
+      "grad_norm": 0.4567905366420746,
+      "learning_rate": 0.0007925660780226683,
+      "loss": 2.9973,
+      "step": 2516
+    },
+    {
+      "epoch": 0.322176,
+      "grad_norm": 0.4677960276603699,
+      "learning_rate": 0.0007923979364280663,
+      "loss": 2.8779,
+      "step": 2517
+    },
+    {
+      "epoch": 0.322304,
+      "grad_norm": 0.46520453691482544,
+      "learning_rate": 0.0007922297445668694,
+      "loss": 2.8797,
+      "step": 2518
+    },
+    {
+      "epoch": 0.322432,
+      "grad_norm": 0.447179913520813,
+      "learning_rate": 0.0007920615024679916,
+      "loss": 2.8674,
+      "step": 2519
+    },
+    {
+      "epoch": 0.32256,
+      "grad_norm": 0.4928196966648102,
+      "learning_rate": 0.0007918932101603555,
+      "loss": 3.0065,
+      "step": 2520
+    },
+    {
+      "epoch": 0.322688,
+      "grad_norm": 0.5056754350662231,
+      "learning_rate": 0.0007917248676728929,
+      "loss": 2.9075,
+      "step": 2521
+    },
+    {
+      "epoch": 0.322816,
+      "grad_norm": 0.49443066120147705,
+      "learning_rate": 0.0007915564750345435,
+      "loss": 2.9489,
+      "step": 2522
+    },
+    {
+      "epoch": 0.322944,
+      "grad_norm": 0.47672945261001587,
+      "learning_rate": 0.0007913880322742559,
+      "loss": 2.9368,
+      "step": 2523
+    },
+    {
+      "epoch": 0.323072,
+      "grad_norm": 0.48887988924980164,
+      "learning_rate": 0.0007912195394209877,
+      "loss": 2.7947,
+      "step": 2524
+    },
+    {
+      "epoch": 0.3232,
+      "grad_norm": 0.4442354738712311,
+      "learning_rate": 0.0007910509965037045,
+      "loss": 2.8982,
+      "step": 2525
+    },
+    {
+      "epoch": 0.323328,
+      "grad_norm": 0.4569530189037323,
+      "learning_rate": 0.0007908824035513808,
+      "loss": 2.9779,
+      "step": 2526
+    },
+    {
+      "epoch": 0.323456,
+      "grad_norm": 0.4435310959815979,
+      "learning_rate": 0.0007907137605929997,
+      "loss": 2.838,
+      "step": 2527
+    },
+    {
+      "epoch": 0.323584,
+      "grad_norm": 0.4726199209690094,
+      "learning_rate": 0.000790545067657553,
+      "loss": 2.8741,
+      "step": 2528
+    },
+    {
+      "epoch": 0.323712,
+      "grad_norm": 0.46060502529144287,
+      "learning_rate": 0.0007903763247740407,
+      "loss": 2.9261,
+      "step": 2529
+    },
+    {
+      "epoch": 0.32384,
+      "grad_norm": 0.479848712682724,
+      "learning_rate": 0.0007902075319714721,
+      "loss": 2.8943,
+      "step": 2530
+    },
+    {
+      "epoch": 0.323968,
+      "grad_norm": 0.5796263813972473,
+      "learning_rate": 0.0007900386892788642,
+      "loss": 3.0156,
+      "step": 2531
+    },
+    {
+      "epoch": 0.324096,
+      "grad_norm": 0.4818500578403473,
+      "learning_rate": 0.0007898697967252433,
+      "loss": 2.8906,
+      "step": 2532
+    },
+    {
+      "epoch": 0.324224,
+      "grad_norm": 0.4410395622253418,
+      "learning_rate": 0.0007897008543396438,
+      "loss": 2.9169,
+      "step": 2533
+    },
+    {
+      "epoch": 0.324352,
+      "grad_norm": 0.48457151651382446,
+      "learning_rate": 0.000789531862151109,
+      "loss": 2.8941,
+      "step": 2534
+    },
+    {
+      "epoch": 0.32448,
+      "grad_norm": 0.4650956690311432,
+      "learning_rate": 0.0007893628201886907,
+      "loss": 2.9379,
+      "step": 2535
+    },
+    {
+      "epoch": 0.324608,
+      "grad_norm": 0.4716746211051941,
+      "learning_rate": 0.0007891937284814488,
+      "loss": 2.8854,
+      "step": 2536
+    },
+    {
+      "epoch": 0.324736,
+      "grad_norm": 0.430833637714386,
+      "learning_rate": 0.0007890245870584522,
+      "loss": 2.8788,
+      "step": 2537
+    },
+    {
+      "epoch": 0.324864,
+      "grad_norm": 0.4749464690685272,
+      "learning_rate": 0.0007888553959487787,
+      "loss": 2.9982,
+      "step": 2538
+    },
+    {
+      "epoch": 0.324992,
+      "grad_norm": 0.4525414705276489,
+      "learning_rate": 0.0007886861551815138,
+      "loss": 2.7695,
+      "step": 2539
+    },
+    {
+      "epoch": 0.32512,
+      "grad_norm": 0.4450726807117462,
+      "learning_rate": 0.0007885168647857524,
+      "loss": 2.8473,
+      "step": 2540
+    },
+    {
+      "epoch": 0.325248,
+      "grad_norm": 0.4866255819797516,
+      "learning_rate": 0.0007883475247905969,
+      "loss": 2.9393,
+      "step": 2541
+    },
+    {
+      "epoch": 0.325376,
+      "grad_norm": 0.5236878395080566,
+      "learning_rate": 0.0007881781352251591,
+      "loss": 2.8753,
+      "step": 2542
+    },
+    {
+      "epoch": 0.325504,
+      "grad_norm": 0.5095987915992737,
+      "learning_rate": 0.0007880086961185589,
+      "loss": 2.976,
+      "step": 2543
+    },
+    {
+      "epoch": 0.325632,
+      "grad_norm": 0.4403684735298157,
+      "learning_rate": 0.0007878392074999253,
+      "loss": 2.8752,
+      "step": 2544
+    },
+    {
+      "epoch": 0.32576,
+      "grad_norm": 0.4837266504764557,
+      "learning_rate": 0.0007876696693983946,
+      "loss": 2.9656,
+      "step": 2545
+    },
+    {
+      "epoch": 0.325888,
+      "grad_norm": 0.47327524423599243,
+      "learning_rate": 0.000787500081843113,
+      "loss": 2.8881,
+      "step": 2546
+    },
+    {
+      "epoch": 0.326016,
+      "grad_norm": 0.4423697888851166,
+      "learning_rate": 0.0007873304448632344,
+      "loss": 2.8181,
+      "step": 2547
+    },
+    {
+      "epoch": 0.326144,
+      "grad_norm": 0.46050408482551575,
+      "learning_rate": 0.0007871607584879211,
+      "loss": 2.9022,
+      "step": 2548
+    },
+    {
+      "epoch": 0.326272,
+      "grad_norm": 0.47034740447998047,
+      "learning_rate": 0.0007869910227463446,
+      "loss": 2.9141,
+      "step": 2549
+    },
+    {
+      "epoch": 0.3264,
+      "grad_norm": 0.47077763080596924,
+      "learning_rate": 0.000786821237667684,
+      "loss": 2.9543,
+      "step": 2550
+    },
+    {
+      "epoch": 0.326528,
+      "grad_norm": 0.4654685854911804,
+      "learning_rate": 0.0007866514032811276,
+      "loss": 2.7864,
+      "step": 2551
+    },
+    {
+      "epoch": 0.326656,
+      "grad_norm": 0.506328284740448,
+      "learning_rate": 0.0007864815196158718,
+      "loss": 2.8589,
+      "step": 2552
+    },
+    {
+      "epoch": 0.326784,
+      "grad_norm": 0.4537164568901062,
+      "learning_rate": 0.0007863115867011216,
+      "loss": 2.817,
+      "step": 2553
+    },
+    {
+      "epoch": 0.326912,
+      "grad_norm": 0.4540252685546875,
+      "learning_rate": 0.0007861416045660905,
+      "loss": 2.9276,
+      "step": 2554
+    },
+    {
+      "epoch": 0.32704,
+      "grad_norm": 0.49236607551574707,
+      "learning_rate": 0.0007859715732400004,
+      "loss": 2.8663,
+      "step": 2555
+    },
+    {
+      "epoch": 0.327168,
+      "grad_norm": 0.5092237591743469,
+      "learning_rate": 0.0007858014927520814,
+      "loss": 2.9746,
+      "step": 2556
+    },
+    {
+      "epoch": 0.327296,
+      "grad_norm": 0.5062278509140015,
+      "learning_rate": 0.0007856313631315726,
+      "loss": 2.9013,
+      "step": 2557
+    },
+    {
+      "epoch": 0.327424,
+      "grad_norm": 0.4860278367996216,
+      "learning_rate": 0.0007854611844077213,
+      "loss": 3.0595,
+      "step": 2558
+    },
+    {
+      "epoch": 0.327552,
+      "grad_norm": 0.48558688163757324,
+      "learning_rate": 0.0007852909566097828,
+      "loss": 2.9202,
+      "step": 2559
+    },
+    {
+      "epoch": 0.32768,
+      "grad_norm": 0.4853699505329132,
+      "learning_rate": 0.0007851206797670216,
+      "loss": 2.7824,
+      "step": 2560
+    },
+    {
+      "epoch": 0.327808,
+      "grad_norm": 0.4717356860637665,
+      "learning_rate": 0.0007849503539087101,
+      "loss": 2.7706,
+      "step": 2561
+    },
+    {
+      "epoch": 0.327936,
+      "grad_norm": 0.4459903836250305,
+      "learning_rate": 0.0007847799790641295,
+      "loss": 2.841,
+      "step": 2562
+    },
+    {
+      "epoch": 0.328064,
+      "grad_norm": 0.49329453706741333,
+      "learning_rate": 0.0007846095552625688,
+      "loss": 3.0215,
+      "step": 2563
+    },
+    {
+      "epoch": 0.328192,
+      "grad_norm": 0.499096155166626,
+      "learning_rate": 0.0007844390825333263,
+      "loss": 3.0154,
+      "step": 2564
+    },
+    {
+      "epoch": 0.32832,
+      "grad_norm": 0.4604116678237915,
+      "learning_rate": 0.0007842685609057077,
+      "loss": 2.8667,
+      "step": 2565
+    },
+    {
+      "epoch": 0.328448,
+      "grad_norm": 0.5083147287368774,
+      "learning_rate": 0.0007840979904090282,
+      "loss": 2.8709,
+      "step": 2566
+    },
+    {
+      "epoch": 0.328576,
+      "grad_norm": 0.43596771359443665,
+      "learning_rate": 0.0007839273710726107,
+      "loss": 2.8169,
+      "step": 2567
+    },
+    {
+      "epoch": 0.328704,
+      "grad_norm": 0.4484316408634186,
+      "learning_rate": 0.0007837567029257864,
+      "loss": 2.8461,
+      "step": 2568
+    },
+    {
+      "epoch": 0.328832,
+      "grad_norm": 0.4814988374710083,
+      "learning_rate": 0.0007835859859978951,
+      "loss": 2.8973,
+      "step": 2569
+    },
+    {
+      "epoch": 0.32896,
+      "grad_norm": 0.487408846616745,
+      "learning_rate": 0.0007834152203182854,
+      "loss": 3.0962,
+      "step": 2570
+    },
+    {
+      "epoch": 0.329088,
+      "grad_norm": 0.46855103969573975,
+      "learning_rate": 0.0007832444059163137,
+      "loss": 3.0038,
+      "step": 2571
+    },
+    {
+      "epoch": 0.329216,
+      "grad_norm": 0.4571864604949951,
+      "learning_rate": 0.0007830735428213451,
+      "loss": 2.9094,
+      "step": 2572
+    },
+    {
+      "epoch": 0.329344,
+      "grad_norm": 0.48822835087776184,
+      "learning_rate": 0.0007829026310627527,
+      "loss": 2.9844,
+      "step": 2573
+    },
+    {
+      "epoch": 0.329472,
+      "grad_norm": 0.46997174620628357,
+      "learning_rate": 0.0007827316706699181,
+      "loss": 2.8557,
+      "step": 2574
+    },
+    {
+      "epoch": 0.3296,
+      "grad_norm": 0.4517820179462433,
+      "learning_rate": 0.0007825606616722318,
+      "loss": 3.0907,
+      "step": 2575
+    },
+    {
+      "epoch": 0.329728,
+      "grad_norm": 0.4879182279109955,
+      "learning_rate": 0.0007823896040990921,
+      "loss": 2.9725,
+      "step": 2576
+    },
+    {
+      "epoch": 0.329856,
+      "grad_norm": 0.4874756336212158,
+      "learning_rate": 0.0007822184979799057,
+      "loss": 2.9435,
+      "step": 2577
+    },
+    {
+      "epoch": 0.329984,
+      "grad_norm": 0.5001745223999023,
+      "learning_rate": 0.0007820473433440876,
+      "loss": 2.9019,
+      "step": 2578
+    },
+    {
+      "epoch": 0.330112,
+      "grad_norm": 0.46457627415657043,
+      "learning_rate": 0.0007818761402210615,
+      "loss": 3.0009,
+      "step": 2579
+    },
+    {
+      "epoch": 0.33024,
+      "grad_norm": 0.5468002557754517,
+      "learning_rate": 0.0007817048886402592,
+      "loss": 2.8521,
+      "step": 2580
+    },
+    {
+      "epoch": 0.330368,
+      "grad_norm": 0.4407159090042114,
+      "learning_rate": 0.0007815335886311206,
+      "loss": 2.9363,
+      "step": 2581
+    },
+    {
+      "epoch": 0.330496,
+      "grad_norm": 0.4811733663082123,
+      "learning_rate": 0.0007813622402230943,
+      "loss": 2.9174,
+      "step": 2582
+    },
+    {
+      "epoch": 0.330624,
+      "grad_norm": 0.44949010014533997,
+      "learning_rate": 0.0007811908434456372,
+      "loss": 2.764,
+      "step": 2583
+    },
+    {
+      "epoch": 0.330752,
+      "grad_norm": 0.45394304394721985,
+      "learning_rate": 0.0007810193983282141,
+      "loss": 2.8772,
+      "step": 2584
+    },
+    {
+      "epoch": 0.33088,
+      "grad_norm": 0.47455713152885437,
+      "learning_rate": 0.0007808479049002987,
+      "loss": 2.9541,
+      "step": 2585
+    },
+    {
+      "epoch": 0.331008,
+      "grad_norm": 0.47033190727233887,
+      "learning_rate": 0.0007806763631913725,
+      "loss": 2.8764,
+      "step": 2586
+    },
+    {
+      "epoch": 0.331136,
+      "grad_norm": 0.5444204211235046,
+      "learning_rate": 0.0007805047732309256,
+      "loss": 2.7914,
+      "step": 2587
+    },
+    {
+      "epoch": 0.331264,
+      "grad_norm": 0.4693557918071747,
+      "learning_rate": 0.0007803331350484563,
+      "loss": 2.8605,
+      "step": 2588
+    },
+    {
+      "epoch": 0.331392,
+      "grad_norm": 0.4406556785106659,
+      "learning_rate": 0.0007801614486734711,
+      "loss": 2.8056,
+      "step": 2589
+    },
+    {
+      "epoch": 0.33152,
+      "grad_norm": 0.47865644097328186,
+      "learning_rate": 0.0007799897141354853,
+      "loss": 2.8692,
+      "step": 2590
+    },
+    {
+      "epoch": 0.331648,
+      "grad_norm": 0.5154767632484436,
+      "learning_rate": 0.0007798179314640214,
+      "loss": 2.9652,
+      "step": 2591
+    },
+    {
+      "epoch": 0.331776,
+      "grad_norm": 0.48273369669914246,
+      "learning_rate": 0.000779646100688611,
+      "loss": 2.915,
+      "step": 2592
+    },
+    {
+      "epoch": 0.331904,
+      "grad_norm": 0.4819542467594147,
+      "learning_rate": 0.0007794742218387943,
+      "loss": 2.95,
+      "step": 2593
+    },
+    {
+      "epoch": 0.332032,
+      "grad_norm": 0.4575668275356293,
+      "learning_rate": 0.0007793022949441189,
+      "loss": 2.8759,
+      "step": 2594
+    },
+    {
+      "epoch": 0.33216,
+      "grad_norm": 0.6341890096664429,
+      "learning_rate": 0.000779130320034141,
+      "loss": 3.0062,
+      "step": 2595
+    },
+    {
+      "epoch": 0.332288,
+      "grad_norm": 0.45544272661209106,
+      "learning_rate": 0.000778958297138425,
+      "loss": 2.9871,
+      "step": 2596
+    },
+    {
+      "epoch": 0.332416,
+      "grad_norm": 0.44684037566185,
+      "learning_rate": 0.0007787862262865438,
+      "loss": 2.8431,
+      "step": 2597
+    },
+    {
+      "epoch": 0.332544,
+      "grad_norm": 0.4882149398326874,
+      "learning_rate": 0.0007786141075080785,
+      "loss": 2.9647,
+      "step": 2598
+    },
+    {
+      "epoch": 0.332672,
+      "grad_norm": 0.4739358425140381,
+      "learning_rate": 0.000778441940832618,
+      "loss": 2.9583,
+      "step": 2599
+    },
+    {
+      "epoch": 0.3328,
+      "grad_norm": 0.4898494482040405,
+      "learning_rate": 0.00077826972628976,
+      "loss": 2.8594,
+      "step": 2600
+    },
+    {
+      "epoch": 0.332928,
+      "grad_norm": 0.4745839536190033,
+      "learning_rate": 0.00077809746390911,
+      "loss": 3.0179,
+      "step": 2601
+    },
+    {
+      "epoch": 0.333056,
+      "grad_norm": 0.44942760467529297,
+      "learning_rate": 0.0007779251537202818,
+      "loss": 2.8442,
+      "step": 2602
+    },
+    {
+      "epoch": 0.333184,
+      "grad_norm": 0.48985934257507324,
+      "learning_rate": 0.0007777527957528981,
+      "loss": 2.9839,
+      "step": 2603
+    },
+    {
+      "epoch": 0.333312,
+      "grad_norm": 0.5002795457839966,
+      "learning_rate": 0.0007775803900365885,
+      "loss": 2.7719,
+      "step": 2604
+    },
+    {
+      "epoch": 0.33344,
+      "grad_norm": 0.4596644639968872,
+      "learning_rate": 0.0007774079366009919,
+      "loss": 2.9777,
+      "step": 2605
+    },
+    {
+      "epoch": 0.333568,
+      "grad_norm": 0.5231437683105469,
+      "learning_rate": 0.0007772354354757549,
+      "loss": 2.9209,
+      "step": 2606
+    },
+    {
+      "epoch": 0.333696,
+      "grad_norm": 0.5029871463775635,
+      "learning_rate": 0.0007770628866905329,
+      "loss": 2.9421,
+      "step": 2607
+    },
+    {
+      "epoch": 0.333824,
+      "grad_norm": 0.4868071675300598,
+      "learning_rate": 0.0007768902902749884,
+      "loss": 2.9039,
+      "step": 2608
+    },
+    {
+      "epoch": 0.333952,
+      "grad_norm": 0.507334291934967,
+      "learning_rate": 0.0007767176462587931,
+      "loss": 2.9395,
+      "step": 2609
+    },
+    {
+      "epoch": 0.33408,
+      "grad_norm": 0.4206216633319855,
+      "learning_rate": 0.0007765449546716266,
+      "loss": 2.7179,
+      "step": 2610
+    },
+    {
+      "epoch": 0.334208,
+      "grad_norm": 0.5274062752723694,
+      "learning_rate": 0.0007763722155431761,
+      "loss": 2.832,
+      "step": 2611
+    },
+    {
+      "epoch": 0.334336,
+      "grad_norm": 0.4797396659851074,
+      "learning_rate": 0.000776199428903138,
+      "loss": 2.8569,
+      "step": 2612
+    },
+    {
+      "epoch": 0.334464,
+      "grad_norm": 0.4408458173274994,
+      "learning_rate": 0.000776026594781216,
+      "loss": 2.8942,
+      "step": 2613
+    },
+    {
+      "epoch": 0.334592,
+      "grad_norm": 0.46456828713417053,
+      "learning_rate": 0.0007758537132071224,
+      "loss": 2.8594,
+      "step": 2614
+    },
+    {
+      "epoch": 0.33472,
+      "grad_norm": 0.5235244631767273,
+      "learning_rate": 0.0007756807842105774,
+      "loss": 2.8201,
+      "step": 2615
+    },
+    {
+      "epoch": 0.334848,
+      "grad_norm": 0.48998400568962097,
+      "learning_rate": 0.0007755078078213098,
+      "loss": 2.8448,
+      "step": 2616
+    },
+    {
+      "epoch": 0.334976,
+      "grad_norm": 0.4846147298812866,
+      "learning_rate": 0.000775334784069056,
+      "loss": 2.9637,
+      "step": 2617
+    },
+    {
+      "epoch": 0.335104,
+      "grad_norm": 0.4855690002441406,
+      "learning_rate": 0.0007751617129835609,
+      "loss": 2.9661,
+      "step": 2618
+    },
+    {
+      "epoch": 0.335232,
+      "grad_norm": 0.44451600313186646,
+      "learning_rate": 0.0007749885945945774,
+      "loss": 2.8832,
+      "step": 2619
+    },
+    {
+      "epoch": 0.33536,
+      "grad_norm": 0.46829962730407715,
+      "learning_rate": 0.0007748154289318665,
+      "loss": 2.9912,
+      "step": 2620
+    },
+    {
+      "epoch": 0.335488,
+      "grad_norm": 0.47119155526161194,
+      "learning_rate": 0.0007746422160251975,
+      "loss": 2.9092,
+      "step": 2621
+    },
+    {
+      "epoch": 0.335616,
+      "grad_norm": 0.4769349694252014,
+      "learning_rate": 0.0007744689559043475,
+      "loss": 2.9506,
+      "step": 2622
+    },
+    {
+      "epoch": 0.335744,
+      "grad_norm": 0.4406980872154236,
+      "learning_rate": 0.0007742956485991022,
+      "loss": 2.9371,
+      "step": 2623
+    },
+    {
+      "epoch": 0.335872,
+      "grad_norm": 0.43092581629753113,
+      "learning_rate": 0.000774122294139255,
+      "loss": 2.7978,
+      "step": 2624
+    },
+    {
+      "epoch": 0.336,
+      "grad_norm": 0.5187844038009644,
+      "learning_rate": 0.0007739488925546075,
+      "loss": 2.9479,
+      "step": 2625
+    },
+    {
+      "epoch": 0.336128,
+      "grad_norm": 0.49655312299728394,
+      "learning_rate": 0.0007737754438749693,
+      "loss": 2.8565,
+      "step": 2626
+    },
+    {
+      "epoch": 0.336256,
+      "grad_norm": 0.46754682064056396,
+      "learning_rate": 0.0007736019481301588,
+      "loss": 2.9492,
+      "step": 2627
+    },
+    {
+      "epoch": 0.336384,
+      "grad_norm": 0.5002387166023254,
+      "learning_rate": 0.0007734284053500013,
+      "loss": 2.9341,
+      "step": 2628
+    },
+    {
+      "epoch": 0.336512,
+      "grad_norm": 0.4983181655406952,
+      "learning_rate": 0.000773254815564331,
+      "loss": 3.0273,
+      "step": 2629
+    },
+    {
+      "epoch": 0.33664,
+      "grad_norm": 0.5606125593185425,
+      "learning_rate": 0.0007730811788029902,
+      "loss": 2.8949,
+      "step": 2630
+    },
+    {
+      "epoch": 0.336768,
+      "grad_norm": 0.5023104548454285,
+      "learning_rate": 0.0007729074950958289,
+      "loss": 3.0129,
+      "step": 2631
+    },
+    {
+      "epoch": 0.336896,
+      "grad_norm": 0.4907623827457428,
+      "learning_rate": 0.0007727337644727055,
+      "loss": 3.0,
+      "step": 2632
+    },
+    {
+      "epoch": 0.337024,
+      "grad_norm": 0.47121483087539673,
+      "learning_rate": 0.000772559986963486,
+      "loss": 2.9257,
+      "step": 2633
+    },
+    {
+      "epoch": 0.337152,
+      "grad_norm": 0.487299382686615,
+      "learning_rate": 0.000772386162598045,
+      "loss": 2.9589,
+      "step": 2634
+    },
+    {
+      "epoch": 0.33728,
+      "grad_norm": 0.47607800364494324,
+      "learning_rate": 0.0007722122914062649,
+      "loss": 2.9179,
+      "step": 2635
+    },
+    {
+      "epoch": 0.337408,
+      "grad_norm": 0.4825735092163086,
+      "learning_rate": 0.0007720383734180361,
+      "loss": 2.914,
+      "step": 2636
+    },
+    {
+      "epoch": 0.337536,
+      "grad_norm": 0.5055685043334961,
+      "learning_rate": 0.0007718644086632573,
+      "loss": 2.9986,
+      "step": 2637
+    },
+    {
+      "epoch": 0.337664,
+      "grad_norm": 0.45996713638305664,
+      "learning_rate": 0.000771690397171835,
+      "loss": 2.9596,
+      "step": 2638
+    },
+    {
+      "epoch": 0.337792,
+      "grad_norm": 0.4450676143169403,
+      "learning_rate": 0.0007715163389736834,
+      "loss": 2.8049,
+      "step": 2639
+    },
+    {
+      "epoch": 0.33792,
+      "grad_norm": 0.46981316804885864,
+      "learning_rate": 0.0007713422340987257,
+      "loss": 2.801,
+      "step": 2640
+    },
+    {
+      "epoch": 0.338048,
+      "grad_norm": 0.446167916059494,
+      "learning_rate": 0.0007711680825768921,
+      "loss": 2.8747,
+      "step": 2641
+    },
+    {
+      "epoch": 0.338176,
+      "grad_norm": 0.45793554186820984,
+      "learning_rate": 0.0007709938844381217,
+      "loss": 2.9079,
+      "step": 2642
+    },
+    {
+      "epoch": 0.338304,
+      "grad_norm": 0.4841660261154175,
+      "learning_rate": 0.0007708196397123607,
+      "loss": 2.9336,
+      "step": 2643
+    },
+    {
+      "epoch": 0.338432,
+      "grad_norm": 0.841526985168457,
+      "learning_rate": 0.0007706453484295641,
+      "loss": 2.9698,
+      "step": 2644
+    },
+    {
+      "epoch": 0.33856,
+      "grad_norm": 0.4851662814617157,
+      "learning_rate": 0.0007704710106196946,
+      "loss": 2.821,
+      "step": 2645
+    },
+    {
+      "epoch": 0.338688,
+      "grad_norm": 0.4588889181613922,
+      "learning_rate": 0.0007702966263127227,
+      "loss": 2.9862,
+      "step": 2646
+    },
+    {
+      "epoch": 0.338816,
+      "grad_norm": 0.45303109288215637,
+      "learning_rate": 0.0007701221955386273,
+      "loss": 2.8499,
+      "step": 2647
+    },
+    {
+      "epoch": 0.338944,
+      "grad_norm": 0.46600204706192017,
+      "learning_rate": 0.0007699477183273948,
+      "loss": 2.9219,
+      "step": 2648
+    },
+    {
+      "epoch": 0.339072,
+      "grad_norm": 0.44709786772727966,
+      "learning_rate": 0.00076977319470902,
+      "loss": 2.9828,
+      "step": 2649
+    },
+    {
+      "epoch": 0.3392,
+      "grad_norm": 0.4619612991809845,
+      "learning_rate": 0.0007695986247135059,
+      "loss": 2.9637,
+      "step": 2650
+    },
+    {
+      "epoch": 0.339328,
+      "grad_norm": 0.4531804025173187,
+      "learning_rate": 0.0007694240083708624,
+      "loss": 2.983,
+      "step": 2651
+    },
+    {
+      "epoch": 0.339456,
+      "grad_norm": 0.4943966567516327,
+      "learning_rate": 0.0007692493457111085,
+      "loss": 2.9676,
+      "step": 2652
+    },
+    {
+      "epoch": 0.339584,
+      "grad_norm": 0.45841851830482483,
+      "learning_rate": 0.0007690746367642707,
+      "loss": 2.9488,
+      "step": 2653
+    },
+    {
+      "epoch": 0.339712,
+      "grad_norm": 0.46747100353240967,
+      "learning_rate": 0.0007688998815603836,
+      "loss": 2.85,
+      "step": 2654
+    },
+    {
+      "epoch": 0.33984,
+      "grad_norm": 0.4490174353122711,
+      "learning_rate": 0.0007687250801294892,
+      "loss": 3.1344,
+      "step": 2655
+    },
+    {
+      "epoch": 0.339968,
+      "grad_norm": 0.45749959349632263,
+      "learning_rate": 0.0007685502325016385,
+      "loss": 2.9723,
+      "step": 2656
+    },
+    {
+      "epoch": 0.340096,
+      "grad_norm": 0.48814666271209717,
+      "learning_rate": 0.0007683753387068893,
+      "loss": 3.0699,
+      "step": 2657
+    },
+    {
+      "epoch": 0.340224,
+      "grad_norm": 0.4789627194404602,
+      "learning_rate": 0.0007682003987753083,
+      "loss": 2.9304,
+      "step": 2658
+    },
+    {
+      "epoch": 0.340352,
+      "grad_norm": 0.4523337781429291,
+      "learning_rate": 0.0007680254127369695,
+      "loss": 2.9391,
+      "step": 2659
+    },
+    {
+      "epoch": 0.34048,
+      "grad_norm": 0.49218568205833435,
+      "learning_rate": 0.0007678503806219549,
+      "loss": 2.8195,
+      "step": 2660
+    },
+    {
+      "epoch": 0.340608,
+      "grad_norm": 0.44549286365509033,
+      "learning_rate": 0.0007676753024603546,
+      "loss": 2.8982,
+      "step": 2661
+    },
+    {
+      "epoch": 0.340736,
+      "grad_norm": 0.46618589758872986,
+      "learning_rate": 0.0007675001782822668,
+      "loss": 2.9803,
+      "step": 2662
+    },
+    {
+      "epoch": 0.340864,
+      "grad_norm": 0.48853960633277893,
+      "learning_rate": 0.0007673250081177972,
+      "loss": 2.9112,
+      "step": 2663
+    },
+    {
+      "epoch": 0.340992,
+      "grad_norm": 0.4656594395637512,
+      "learning_rate": 0.0007671497919970597,
+      "loss": 2.8873,
+      "step": 2664
+    },
+    {
+      "epoch": 0.34112,
+      "grad_norm": 0.46643224358558655,
+      "learning_rate": 0.0007669745299501758,
+      "loss": 2.7966,
+      "step": 2665
+    },
+    {
+      "epoch": 0.341248,
+      "grad_norm": 0.4557456374168396,
+      "learning_rate": 0.0007667992220072751,
+      "loss": 2.8442,
+      "step": 2666
+    },
+    {
+      "epoch": 0.341376,
+      "grad_norm": 0.46900469064712524,
+      "learning_rate": 0.0007666238681984952,
+      "loss": 2.9974,
+      "step": 2667
+    },
+    {
+      "epoch": 0.341504,
+      "grad_norm": 0.4714982807636261,
+      "learning_rate": 0.0007664484685539814,
+      "loss": 2.8885,
+      "step": 2668
+    },
+    {
+      "epoch": 0.341632,
+      "grad_norm": 0.5068759918212891,
+      "learning_rate": 0.0007662730231038869,
+      "loss": 2.8386,
+      "step": 2669
+    },
+    {
+      "epoch": 0.34176,
+      "grad_norm": 0.5082533955574036,
+      "learning_rate": 0.0007660975318783729,
+      "loss": 2.8546,
+      "step": 2670
+    },
+    {
+      "epoch": 0.341888,
+      "grad_norm": 0.4969204366207123,
+      "learning_rate": 0.0007659219949076081,
+      "loss": 2.9312,
+      "step": 2671
+    },
+    {
+      "epoch": 0.342016,
+      "grad_norm": 0.47328123450279236,
+      "learning_rate": 0.0007657464122217697,
+      "loss": 2.9124,
+      "step": 2672
+    },
+    {
+      "epoch": 0.342144,
+      "grad_norm": 0.4735209345817566,
+      "learning_rate": 0.0007655707838510423,
+      "loss": 2.869,
+      "step": 2673
+    },
+    {
+      "epoch": 0.342272,
+      "grad_norm": 0.46270933747291565,
+      "learning_rate": 0.0007653951098256183,
+      "loss": 2.8075,
+      "step": 2674
+    },
+    {
+      "epoch": 0.3424,
+      "grad_norm": 0.45459479093551636,
+      "learning_rate": 0.0007652193901756983,
+      "loss": 2.974,
+      "step": 2675
+    },
+    {
+      "epoch": 0.342528,
+      "grad_norm": 0.42626282572746277,
+      "learning_rate": 0.0007650436249314903,
+      "loss": 2.9142,
+      "step": 2676
+    },
+    {
+      "epoch": 0.342656,
+      "grad_norm": 0.46878039836883545,
+      "learning_rate": 0.0007648678141232106,
+      "loss": 2.9379,
+      "step": 2677
+    },
+    {
+      "epoch": 0.342784,
+      "grad_norm": 0.48528507351875305,
+      "learning_rate": 0.000764691957781083,
+      "loss": 3.0562,
+      "step": 2678
+    },
+    {
+      "epoch": 0.342912,
+      "grad_norm": 0.47151026129722595,
+      "learning_rate": 0.0007645160559353393,
+      "loss": 3.1207,
+      "step": 2679
+    },
+    {
+      "epoch": 0.34304,
+      "grad_norm": 0.4615163207054138,
+      "learning_rate": 0.000764340108616219,
+      "loss": 2.9344,
+      "step": 2680
+    },
+    {
+      "epoch": 0.343168,
+      "grad_norm": 0.48443862795829773,
+      "learning_rate": 0.0007641641158539697,
+      "loss": 2.8022,
+      "step": 2681
+    },
+    {
+      "epoch": 0.343296,
+      "grad_norm": 0.592052698135376,
+      "learning_rate": 0.0007639880776788464,
+      "loss": 2.9206,
+      "step": 2682
+    },
+    {
+      "epoch": 0.343424,
+      "grad_norm": 0.4360868036746979,
+      "learning_rate": 0.0007638119941211119,
+      "loss": 2.9824,
+      "step": 2683
+    },
+    {
+      "epoch": 0.343552,
+      "grad_norm": 0.4936276972293854,
+      "learning_rate": 0.0007636358652110375,
+      "loss": 2.877,
+      "step": 2684
+    },
+    {
+      "epoch": 0.34368,
+      "grad_norm": 0.4459110200405121,
+      "learning_rate": 0.0007634596909789013,
+      "loss": 2.8317,
+      "step": 2685
+    },
+    {
+      "epoch": 0.343808,
+      "grad_norm": 0.47784632444381714,
+      "learning_rate": 0.0007632834714549901,
+      "loss": 2.8107,
+      "step": 2686
+    },
+    {
+      "epoch": 0.343936,
+      "grad_norm": 0.47855278849601746,
+      "learning_rate": 0.0007631072066695978,
+      "loss": 3.0156,
+      "step": 2687
+    },
+    {
+      "epoch": 0.344064,
+      "grad_norm": 0.4738227128982544,
+      "learning_rate": 0.0007629308966530268,
+      "loss": 2.8494,
+      "step": 2688
+    },
+    {
+      "epoch": 0.344192,
+      "grad_norm": 0.4760821759700775,
+      "learning_rate": 0.0007627545414355861,
+      "loss": 3.0143,
+      "step": 2689
+    },
+    {
+      "epoch": 0.34432,
+      "grad_norm": 0.43768996000289917,
+      "learning_rate": 0.0007625781410475938,
+      "loss": 2.9708,
+      "step": 2690
+    },
+    {
+      "epoch": 0.344448,
+      "grad_norm": 0.4436054229736328,
+      "learning_rate": 0.000762401695519375,
+      "loss": 2.9073,
+      "step": 2691
+    },
+    {
+      "epoch": 0.344576,
+      "grad_norm": 0.4539101719856262,
+      "learning_rate": 0.0007622252048812628,
+      "loss": 2.7917,
+      "step": 2692
+    },
+    {
+      "epoch": 0.344704,
+      "grad_norm": 0.47362783551216125,
+      "learning_rate": 0.0007620486691635977,
+      "loss": 3.0079,
+      "step": 2693
+    },
+    {
+      "epoch": 0.344832,
+      "grad_norm": 0.47198593616485596,
+      "learning_rate": 0.0007618720883967285,
+      "loss": 2.8266,
+      "step": 2694
+    },
+    {
+      "epoch": 0.34496,
+      "grad_norm": 0.5452716946601868,
+      "learning_rate": 0.0007616954626110115,
+      "loss": 2.7698,
+      "step": 2695
+    },
+    {
+      "epoch": 0.345088,
+      "grad_norm": 0.46432846784591675,
+      "learning_rate": 0.0007615187918368106,
+      "loss": 3.035,
+      "step": 2696
+    },
+    {
+      "epoch": 0.345216,
+      "grad_norm": 0.48362624645233154,
+      "learning_rate": 0.0007613420761044976,
+      "loss": 2.8305,
+      "step": 2697
+    },
+    {
+      "epoch": 0.345344,
+      "grad_norm": 0.4661448001861572,
+      "learning_rate": 0.000761165315444452,
+      "loss": 2.9263,
+      "step": 2698
+    },
+    {
+      "epoch": 0.345472,
+      "grad_norm": 0.4571998119354248,
+      "learning_rate": 0.0007609885098870609,
+      "loss": 2.9651,
+      "step": 2699
+    },
+    {
+      "epoch": 0.3456,
+      "grad_norm": 0.4711521565914154,
+      "learning_rate": 0.0007608116594627195,
+      "loss": 2.9743,
+      "step": 2700
+    },
+    {
+      "epoch": 0.345728,
+      "grad_norm": 0.41744089126586914,
+      "learning_rate": 0.0007606347642018301,
+      "loss": 2.8319,
+      "step": 2701
+    },
+    {
+      "epoch": 0.345856,
+      "grad_norm": 0.4359124004840851,
+      "learning_rate": 0.0007604578241348032,
+      "loss": 2.8904,
+      "step": 2702
+    },
+    {
+      "epoch": 0.345984,
+      "grad_norm": 0.48975709080696106,
+      "learning_rate": 0.0007602808392920569,
+      "loss": 2.9656,
+      "step": 2703
+    },
+    {
+      "epoch": 0.346112,
+      "grad_norm": 0.4369654059410095,
+      "learning_rate": 0.0007601038097040169,
+      "loss": 3.0469,
+      "step": 2704
+    },
+    {
+      "epoch": 0.34624,
+      "grad_norm": 0.48666349053382874,
+      "learning_rate": 0.0007599267354011166,
+      "loss": 2.9709,
+      "step": 2705
+    },
+    {
+      "epoch": 0.346368,
+      "grad_norm": 0.4796736240386963,
+      "learning_rate": 0.0007597496164137972,
+      "loss": 2.9528,
+      "step": 2706
+    },
+    {
+      "epoch": 0.346496,
+      "grad_norm": 0.45433393120765686,
+      "learning_rate": 0.0007595724527725074,
+      "loss": 3.0391,
+      "step": 2707
+    },
+    {
+      "epoch": 0.346624,
+      "grad_norm": 0.4854256212711334,
+      "learning_rate": 0.0007593952445077035,
+      "loss": 3.0061,
+      "step": 2708
+    },
+    {
+      "epoch": 0.346752,
+      "grad_norm": 0.457088828086853,
+      "learning_rate": 0.0007592179916498503,
+      "loss": 2.8341,
+      "step": 2709
+    },
+    {
+      "epoch": 0.34688,
+      "grad_norm": 0.420558363199234,
+      "learning_rate": 0.0007590406942294191,
+      "loss": 2.9203,
+      "step": 2710
+    },
+    {
+      "epoch": 0.347008,
+      "grad_norm": 0.47258326411247253,
+      "learning_rate": 0.0007588633522768895,
+      "loss": 2.9583,
+      "step": 2711
+    },
+    {
+      "epoch": 0.347136,
+      "grad_norm": 0.45688822865486145,
+      "learning_rate": 0.0007586859658227486,
+      "loss": 2.8077,
+      "step": 2712
+    },
+    {
+      "epoch": 0.347264,
+      "grad_norm": 0.4900491535663605,
+      "learning_rate": 0.0007585085348974912,
+      "loss": 2.9605,
+      "step": 2713
+    },
+    {
+      "epoch": 0.347392,
+      "grad_norm": 0.46367472410202026,
+      "learning_rate": 0.0007583310595316199,
+      "loss": 2.8925,
+      "step": 2714
+    },
+    {
+      "epoch": 0.34752,
+      "grad_norm": 0.5012758374214172,
+      "learning_rate": 0.0007581535397556445,
+      "loss": 2.8205,
+      "step": 2715
+    },
+    {
+      "epoch": 0.347648,
+      "grad_norm": 0.4762772023677826,
+      "learning_rate": 0.000757975975600083,
+      "loss": 2.8449,
+      "step": 2716
+    },
+    {
+      "epoch": 0.347776,
+      "grad_norm": 0.5466546416282654,
+      "learning_rate": 0.0007577983670954603,
+      "loss": 2.7598,
+      "step": 2717
+    },
+    {
+      "epoch": 0.347904,
+      "grad_norm": 0.46906834840774536,
+      "learning_rate": 0.0007576207142723099,
+      "loss": 2.8503,
+      "step": 2718
+    },
+    {
+      "epoch": 0.348032,
+      "grad_norm": 0.4513189196586609,
+      "learning_rate": 0.000757443017161172,
+      "loss": 2.7157,
+      "step": 2719
+    },
+    {
+      "epoch": 0.34816,
+      "grad_norm": 0.47062331438064575,
+      "learning_rate": 0.0007572652757925951,
+      "loss": 2.9098,
+      "step": 2720
+    },
+    {
+      "epoch": 0.348288,
+      "grad_norm": 0.4711534380912781,
+      "learning_rate": 0.0007570874901971347,
+      "loss": 2.9067,
+      "step": 2721
+    },
+    {
+      "epoch": 0.348416,
+      "grad_norm": 0.4824124276638031,
+      "learning_rate": 0.0007569096604053543,
+      "loss": 2.973,
+      "step": 2722
+    },
+    {
+      "epoch": 0.348544,
+      "grad_norm": 0.4797900319099426,
+      "learning_rate": 0.0007567317864478247,
+      "loss": 2.6915,
+      "step": 2723
+    },
+    {
+      "epoch": 0.348672,
+      "grad_norm": 0.48488032817840576,
+      "learning_rate": 0.0007565538683551251,
+      "loss": 2.8719,
+      "step": 2724
+    },
+    {
+      "epoch": 0.3488,
+      "grad_norm": 0.4877001643180847,
+      "learning_rate": 0.0007563759061578411,
+      "loss": 2.9595,
+      "step": 2725
+    },
+    {
+      "epoch": 0.348928,
+      "grad_norm": 0.44492217898368835,
+      "learning_rate": 0.0007561978998865666,
+      "loss": 2.8833,
+      "step": 2726
+    },
+    {
+      "epoch": 0.349056,
+      "grad_norm": 0.4525908827781677,
+      "learning_rate": 0.0007560198495719032,
+      "loss": 2.9153,
+      "step": 2727
+    },
+    {
+      "epoch": 0.349184,
+      "grad_norm": 0.4565339684486389,
+      "learning_rate": 0.0007558417552444596,
+      "loss": 2.8062,
+      "step": 2728
+    },
+    {
+      "epoch": 0.349312,
+      "grad_norm": 0.42276471853256226,
+      "learning_rate": 0.0007556636169348523,
+      "loss": 2.8522,
+      "step": 2729
+    },
+    {
+      "epoch": 0.34944,
+      "grad_norm": 0.4775993227958679,
+      "learning_rate": 0.0007554854346737052,
+      "loss": 2.9168,
+      "step": 2730
+    },
+    {
+      "epoch": 0.349568,
+      "grad_norm": 0.47085049748420715,
+      "learning_rate": 0.00075530720849165,
+      "loss": 2.9665,
+      "step": 2731
+    },
+    {
+      "epoch": 0.349696,
+      "grad_norm": 0.4458191394805908,
+      "learning_rate": 0.0007551289384193261,
+      "loss": 2.832,
+      "step": 2732
+    },
+    {
+      "epoch": 0.349824,
+      "grad_norm": 0.5024060010910034,
+      "learning_rate": 0.0007549506244873799,
+      "loss": 2.9513,
+      "step": 2733
+    },
+    {
+      "epoch": 0.349952,
+      "grad_norm": 0.4633769989013672,
+      "learning_rate": 0.0007547722667264656,
+      "loss": 2.9687,
+      "step": 2734
+    },
+    {
+      "epoch": 0.35008,
+      "grad_norm": 0.4250565767288208,
+      "learning_rate": 0.0007545938651672453,
+      "loss": 2.7668,
+      "step": 2735
+    },
+    {
+      "epoch": 0.350208,
+      "grad_norm": 0.5241614580154419,
+      "learning_rate": 0.0007544154198403879,
+      "loss": 2.9475,
+      "step": 2736
+    },
+    {
+      "epoch": 0.350336,
+      "grad_norm": 0.4740312397480011,
+      "learning_rate": 0.0007542369307765704,
+      "loss": 2.9339,
+      "step": 2737
+    },
+    {
+      "epoch": 0.350464,
+      "grad_norm": 0.47248902916908264,
+      "learning_rate": 0.0007540583980064771,
+      "loss": 2.9657,
+      "step": 2738
+    },
+    {
+      "epoch": 0.350592,
+      "grad_norm": 0.49559473991394043,
+      "learning_rate": 0.0007538798215608,
+      "loss": 2.7828,
+      "step": 2739
+    },
+    {
+      "epoch": 0.35072,
+      "grad_norm": 0.4645918607711792,
+      "learning_rate": 0.0007537012014702382,
+      "loss": 2.8469,
+      "step": 2740
+    },
+    {
+      "epoch": 0.350848,
+      "grad_norm": 0.49176499247550964,
+      "learning_rate": 0.0007535225377654989,
+      "loss": 2.8949,
+      "step": 2741
+    },
+    {
+      "epoch": 0.350976,
+      "grad_norm": 0.5089744329452515,
+      "learning_rate": 0.0007533438304772962,
+      "loss": 2.8959,
+      "step": 2742
+    },
+    {
+      "epoch": 0.351104,
+      "grad_norm": 0.4917677640914917,
+      "learning_rate": 0.0007531650796363521,
+      "loss": 2.86,
+      "step": 2743
+    },
+    {
+      "epoch": 0.351232,
+      "grad_norm": 0.41693082451820374,
+      "learning_rate": 0.0007529862852733957,
+      "loss": 2.9011,
+      "step": 2744
+    },
+    {
+      "epoch": 0.35136,
+      "grad_norm": 0.44940268993377686,
+      "learning_rate": 0.000752807447419164,
+      "loss": 2.9258,
+      "step": 2745
+    },
+    {
+      "epoch": 0.351488,
+      "grad_norm": 0.45848071575164795,
+      "learning_rate": 0.0007526285661044014,
+      "loss": 2.8133,
+      "step": 2746
+    },
+    {
+      "epoch": 0.351616,
+      "grad_norm": 0.5148753523826599,
+      "learning_rate": 0.0007524496413598595,
+      "loss": 2.9402,
+      "step": 2747
+    },
+    {
+      "epoch": 0.351744,
+      "grad_norm": 0.4515235722064972,
+      "learning_rate": 0.0007522706732162976,
+      "loss": 3.0297,
+      "step": 2748
+    },
+    {
+      "epoch": 0.351872,
+      "grad_norm": 0.4520319104194641,
+      "learning_rate": 0.0007520916617044825,
+      "loss": 3.0588,
+      "step": 2749
+    },
+    {
+      "epoch": 0.352,
+      "grad_norm": 0.46438106894493103,
+      "learning_rate": 0.0007519126068551882,
+      "loss": 2.9482,
+      "step": 2750
+    },
+    {
+      "epoch": 0.352128,
+      "grad_norm": 0.4730311930179596,
+      "learning_rate": 0.0007517335086991964,
+      "loss": 3.0413,
+      "step": 2751
+    },
+    {
+      "epoch": 0.352256,
+      "grad_norm": 0.47054600715637207,
+      "learning_rate": 0.000751554367267296,
+      "loss": 3.0681,
+      "step": 2752
+    },
+    {
+      "epoch": 0.352384,
+      "grad_norm": 0.42875197529792786,
+      "learning_rate": 0.0007513751825902837,
+      "loss": 2.8611,
+      "step": 2753
+    },
+    {
+      "epoch": 0.352512,
+      "grad_norm": 0.4872433543205261,
+      "learning_rate": 0.0007511959546989632,
+      "loss": 2.8327,
+      "step": 2754
+    },
+    {
+      "epoch": 0.35264,
+      "grad_norm": 0.4465016722679138,
+      "learning_rate": 0.000751016683624146,
+      "loss": 2.9782,
+      "step": 2755
+    },
+    {
+      "epoch": 0.352768,
+      "grad_norm": 0.4368813633918762,
+      "learning_rate": 0.0007508373693966509,
+      "loss": 2.8482,
+      "step": 2756
+    },
+    {
+      "epoch": 0.352896,
+      "grad_norm": 0.4380091428756714,
+      "learning_rate": 0.0007506580120473041,
+      "loss": 2.872,
+      "step": 2757
+    },
+    {
+      "epoch": 0.353024,
+      "grad_norm": 0.5826338529586792,
+      "learning_rate": 0.0007504786116069391,
+      "loss": 2.9154,
+      "step": 2758
+    },
+    {
+      "epoch": 0.353152,
+      "grad_norm": 0.49614566564559937,
+      "learning_rate": 0.0007502991681063969,
+      "loss": 2.9605,
+      "step": 2759
+    },
+    {
+      "epoch": 0.35328,
+      "grad_norm": 0.4909805655479431,
+      "learning_rate": 0.0007501196815765261,
+      "loss": 2.908,
+      "step": 2760
+    },
+    {
+      "epoch": 0.353408,
+      "grad_norm": 0.4376322329044342,
+      "learning_rate": 0.0007499401520481824,
+      "loss": 2.8871,
+      "step": 2761
+    },
+    {
+      "epoch": 0.353536,
+      "grad_norm": 0.49457138776779175,
+      "learning_rate": 0.0007497605795522288,
+      "loss": 2.9141,
+      "step": 2762
+    },
+    {
+      "epoch": 0.353664,
+      "grad_norm": 0.48794761300086975,
+      "learning_rate": 0.0007495809641195364,
+      "loss": 2.8046,
+      "step": 2763
+    },
+    {
+      "epoch": 0.353792,
+      "grad_norm": 0.4625846743583679,
+      "learning_rate": 0.0007494013057809827,
+      "loss": 2.945,
+      "step": 2764
+    },
+    {
+      "epoch": 0.35392,
+      "grad_norm": 0.44658219814300537,
+      "learning_rate": 0.0007492216045674532,
+      "loss": 2.9412,
+      "step": 2765
+    },
+    {
+      "epoch": 0.354048,
+      "grad_norm": 0.4922688901424408,
+      "learning_rate": 0.0007490418605098408,
+      "loss": 2.8563,
+      "step": 2766
+    },
+    {
+      "epoch": 0.354176,
+      "grad_norm": 0.4991559684276581,
+      "learning_rate": 0.0007488620736390454,
+      "loss": 2.8508,
+      "step": 2767
+    },
+    {
+      "epoch": 0.354304,
+      "grad_norm": 0.4442526400089264,
+      "learning_rate": 0.0007486822439859744,
+      "loss": 2.833,
+      "step": 2768
+    },
+    {
+      "epoch": 0.354432,
+      "grad_norm": 0.4861162602901459,
+      "learning_rate": 0.0007485023715815426,
+      "loss": 2.8179,
+      "step": 2769
+    },
+    {
+      "epoch": 0.35456,
+      "grad_norm": 0.4867169260978699,
+      "learning_rate": 0.0007483224564566724,
+      "loss": 2.9389,
+      "step": 2770
+    },
+    {
+      "epoch": 0.354688,
+      "grad_norm": 0.48353302478790283,
+      "learning_rate": 0.0007481424986422932,
+      "loss": 2.8875,
+      "step": 2771
+    },
+    {
+      "epoch": 0.354816,
+      "grad_norm": 0.45972803235054016,
+      "learning_rate": 0.0007479624981693416,
+      "loss": 2.8434,
+      "step": 2772
+    },
+    {
+      "epoch": 0.354944,
+      "grad_norm": 0.4310835301876068,
+      "learning_rate": 0.000747782455068762,
+      "loss": 2.7979,
+      "step": 2773
+    },
+    {
+      "epoch": 0.355072,
+      "grad_norm": 0.4534873366355896,
+      "learning_rate": 0.0007476023693715058,
+      "loss": 2.856,
+      "step": 2774
+    },
+    {
+      "epoch": 0.3552,
+      "grad_norm": 0.5041418075561523,
+      "learning_rate": 0.000747422241108532,
+      "loss": 2.9877,
+      "step": 2775
+    },
+    {
+      "epoch": 0.355328,
+      "grad_norm": 0.4691498279571533,
+      "learning_rate": 0.0007472420703108064,
+      "loss": 2.9948,
+      "step": 2776
+    },
+    {
+      "epoch": 0.355456,
+      "grad_norm": 0.44899919629096985,
+      "learning_rate": 0.0007470618570093026,
+      "loss": 2.9984,
+      "step": 2777
+    },
+    {
+      "epoch": 0.355584,
+      "grad_norm": 0.4346155822277069,
+      "learning_rate": 0.0007468816012350016,
+      "loss": 2.8175,
+      "step": 2778
+    },
+    {
+      "epoch": 0.355712,
+      "grad_norm": 0.4378584027290344,
+      "learning_rate": 0.0007467013030188911,
+      "loss": 2.9116,
+      "step": 2779
+    },
+    {
+      "epoch": 0.35584,
+      "grad_norm": 0.45925962924957275,
+      "learning_rate": 0.0007465209623919668,
+      "loss": 2.9552,
+      "step": 2780
+    },
+    {
+      "epoch": 0.355968,
+      "grad_norm": 0.4410577118396759,
+      "learning_rate": 0.0007463405793852309,
+      "loss": 2.8712,
+      "step": 2781
+    },
+    {
+      "epoch": 0.356096,
+      "grad_norm": 0.4619900584220886,
+      "learning_rate": 0.0007461601540296937,
+      "loss": 2.8609,
+      "step": 2782
+    },
+    {
+      "epoch": 0.356224,
+      "grad_norm": 0.4966939389705658,
+      "learning_rate": 0.0007459796863563723,
+      "loss": 3.0609,
+      "step": 2783
+    },
+    {
+      "epoch": 0.356352,
+      "grad_norm": 0.46446093916893005,
+      "learning_rate": 0.0007457991763962911,
+      "loss": 2.9134,
+      "step": 2784
+    },
+    {
+      "epoch": 0.35648,
+      "grad_norm": 0.4754083454608917,
+      "learning_rate": 0.0007456186241804821,
+      "loss": 2.9866,
+      "step": 2785
+    },
+    {
+      "epoch": 0.356608,
+      "grad_norm": 0.4969669282436371,
+      "learning_rate": 0.000745438029739984,
+      "loss": 2.8082,
+      "step": 2786
+    },
+    {
+      "epoch": 0.356736,
+      "grad_norm": 0.4795287549495697,
+      "learning_rate": 0.0007452573931058432,
+      "loss": 2.82,
+      "step": 2787
+    },
+    {
+      "epoch": 0.356864,
+      "grad_norm": 0.4703952968120575,
+      "learning_rate": 0.0007450767143091133,
+      "loss": 2.9835,
+      "step": 2788
+    },
+    {
+      "epoch": 0.356992,
+      "grad_norm": 0.4619237780570984,
+      "learning_rate": 0.0007448959933808551,
+      "loss": 2.9111,
+      "step": 2789
+    },
+    {
+      "epoch": 0.35712,
+      "grad_norm": 0.5091763734817505,
+      "learning_rate": 0.0007447152303521365,
+      "loss": 2.8714,
+      "step": 2790
+    },
+    {
+      "epoch": 0.357248,
+      "grad_norm": 0.5213060975074768,
+      "learning_rate": 0.0007445344252540327,
+      "loss": 2.8492,
+      "step": 2791
+    },
+    {
+      "epoch": 0.357376,
+      "grad_norm": 0.49798455834388733,
+      "learning_rate": 0.0007443535781176265,
+      "loss": 2.9468,
+      "step": 2792
+    },
+    {
+      "epoch": 0.357504,
+      "grad_norm": 0.4522590935230255,
+      "learning_rate": 0.0007441726889740075,
+      "loss": 2.8089,
+      "step": 2793
+    },
+    {
+      "epoch": 0.357632,
+      "grad_norm": 0.4623050093650818,
+      "learning_rate": 0.0007439917578542723,
+      "loss": 2.8394,
+      "step": 2794
+    },
+    {
+      "epoch": 0.35776,
+      "grad_norm": 0.47724851965904236,
+      "learning_rate": 0.0007438107847895256,
+      "loss": 2.8504,
+      "step": 2795
+    },
+    {
+      "epoch": 0.357888,
+      "grad_norm": 0.46323734521865845,
+      "learning_rate": 0.0007436297698108782,
+      "loss": 2.8515,
+      "step": 2796
+    },
+    {
+      "epoch": 0.358016,
+      "grad_norm": 0.5190070271492004,
+      "learning_rate": 0.0007434487129494493,
+      "loss": 2.9184,
+      "step": 2797
+    },
+    {
+      "epoch": 0.358144,
+      "grad_norm": 0.4678052067756653,
+      "learning_rate": 0.0007432676142363642,
+      "loss": 3.0489,
+      "step": 2798
+    },
+    {
+      "epoch": 0.358272,
+      "grad_norm": 0.4264679253101349,
+      "learning_rate": 0.0007430864737027561,
+      "loss": 2.9739,
+      "step": 2799
+    },
+    {
+      "epoch": 0.3584,
+      "grad_norm": 0.4824303090572357,
+      "learning_rate": 0.0007429052913797648,
+      "loss": 2.9482,
+      "step": 2800
+    },
+    {
+      "epoch": 0.358528,
+      "grad_norm": 0.4766540825366974,
+      "learning_rate": 0.000742724067298538,
+      "loss": 2.8622,
+      "step": 2801
+    },
+    {
+      "epoch": 0.358656,
+      "grad_norm": 0.44895148277282715,
+      "learning_rate": 0.0007425428014902303,
+      "loss": 2.9134,
+      "step": 2802
+    },
+    {
+      "epoch": 0.358784,
+      "grad_norm": 0.492463082075119,
+      "learning_rate": 0.0007423614939860032,
+      "loss": 2.8379,
+      "step": 2803
+    },
+    {
+      "epoch": 0.358912,
+      "grad_norm": 0.44722744822502136,
+      "learning_rate": 0.0007421801448170256,
+      "loss": 2.8339,
+      "step": 2804
+    },
+    {
+      "epoch": 0.35904,
+      "grad_norm": 0.4558826684951782,
+      "learning_rate": 0.0007419987540144733,
+      "loss": 2.7451,
+      "step": 2805
+    },
+    {
+      "epoch": 0.359168,
+      "grad_norm": 0.48909372091293335,
+      "learning_rate": 0.0007418173216095299,
+      "loss": 2.8612,
+      "step": 2806
+    },
+    {
+      "epoch": 0.359296,
+      "grad_norm": 0.449277400970459,
+      "learning_rate": 0.0007416358476333855,
+      "loss": 2.8629,
+      "step": 2807
+    },
+    {
+      "epoch": 0.359424,
+      "grad_norm": 0.44733449816703796,
+      "learning_rate": 0.0007414543321172376,
+      "loss": 2.8625,
+      "step": 2808
+    },
+    {
+      "epoch": 0.359552,
+      "grad_norm": 0.43573474884033203,
+      "learning_rate": 0.0007412727750922908,
+      "loss": 2.9038,
+      "step": 2809
+    },
+    {
+      "epoch": 0.35968,
+      "grad_norm": 0.4705999195575714,
+      "learning_rate": 0.0007410911765897567,
+      "loss": 2.9216,
+      "step": 2810
+    },
+    {
+      "epoch": 0.359808,
+      "grad_norm": 0.4630375802516937,
+      "learning_rate": 0.0007409095366408547,
+      "loss": 2.7827,
+      "step": 2811
+    },
+    {
+      "epoch": 0.359936,
+      "grad_norm": 0.5195351839065552,
+      "learning_rate": 0.0007407278552768103,
+      "loss": 2.9183,
+      "step": 2812
+    },
+    {
+      "epoch": 0.360064,
+      "grad_norm": 0.4566921889781952,
+      "learning_rate": 0.0007405461325288569,
+      "loss": 2.8096,
+      "step": 2813
+    },
+    {
+      "epoch": 0.360192,
+      "grad_norm": 0.5006151795387268,
+      "learning_rate": 0.0007403643684282346,
+      "loss": 2.9268,
+      "step": 2814
+    },
+    {
+      "epoch": 0.36032,
+      "grad_norm": 0.46068084239959717,
+      "learning_rate": 0.000740182563006191,
+      "loss": 2.894,
+      "step": 2815
+    },
+    {
+      "epoch": 0.360448,
+      "grad_norm": 0.4505310356616974,
+      "learning_rate": 0.0007400007162939802,
+      "loss": 2.8872,
+      "step": 2816
+    },
+    {
+      "epoch": 0.360576,
+      "grad_norm": 0.4473731517791748,
+      "learning_rate": 0.000739818828322864,
+      "loss": 2.9606,
+      "step": 2817
+    },
+    {
+      "epoch": 0.360704,
+      "grad_norm": 0.5108419060707092,
+      "learning_rate": 0.000739636899124111,
+      "loss": 3.0024,
+      "step": 2818
+    },
+    {
+      "epoch": 0.360832,
+      "grad_norm": 0.45436760783195496,
+      "learning_rate": 0.0007394549287289969,
+      "loss": 2.9128,
+      "step": 2819
+    },
+    {
+      "epoch": 0.36096,
+      "grad_norm": 0.47473230957984924,
+      "learning_rate": 0.0007392729171688046,
+      "loss": 2.9829,
+      "step": 2820
+    },
+    {
+      "epoch": 0.361088,
+      "grad_norm": 0.4894045293331146,
+      "learning_rate": 0.000739090864474824,
+      "loss": 2.9708,
+      "step": 2821
+    },
+    {
+      "epoch": 0.361216,
+      "grad_norm": 0.4493611454963684,
+      "learning_rate": 0.0007389087706783518,
+      "loss": 2.7433,
+      "step": 2822
+    },
+    {
+      "epoch": 0.361344,
+      "grad_norm": 0.4856726825237274,
+      "learning_rate": 0.0007387266358106924,
+      "loss": 2.8639,
+      "step": 2823
+    },
+    {
+      "epoch": 0.361472,
+      "grad_norm": 0.49168726801872253,
+      "learning_rate": 0.0007385444599031567,
+      "loss": 2.9283,
+      "step": 2824
+    },
+    {
+      "epoch": 0.3616,
+      "grad_norm": 0.48293447494506836,
+      "learning_rate": 0.000738362242987063,
+      "loss": 2.8149,
+      "step": 2825
+    },
+    {
+      "epoch": 0.361728,
+      "grad_norm": 0.5469521880149841,
+      "learning_rate": 0.0007381799850937363,
+      "loss": 2.9908,
+      "step": 2826
+    },
+    {
+      "epoch": 0.361856,
+      "grad_norm": 0.528692901134491,
+      "learning_rate": 0.0007379976862545091,
+      "loss": 3.0055,
+      "step": 2827
+    },
+    {
+      "epoch": 0.361984,
+      "grad_norm": 0.4850149154663086,
+      "learning_rate": 0.0007378153465007202,
+      "loss": 3.0135,
+      "step": 2828
+    },
+    {
+      "epoch": 0.362112,
+      "grad_norm": 0.49843403697013855,
+      "learning_rate": 0.0007376329658637166,
+      "loss": 2.9256,
+      "step": 2829
+    },
+    {
+      "epoch": 0.36224,
+      "grad_norm": 0.4640624225139618,
+      "learning_rate": 0.0007374505443748511,
+      "loss": 2.7788,
+      "step": 2830
+    },
+    {
+      "epoch": 0.362368,
+      "grad_norm": 0.5423629879951477,
+      "learning_rate": 0.0007372680820654843,
+      "loss": 2.8486,
+      "step": 2831
+    },
+    {
+      "epoch": 0.362496,
+      "grad_norm": 0.44692981243133545,
+      "learning_rate": 0.0007370855789669837,
+      "loss": 2.8462,
+      "step": 2832
+    },
+    {
+      "epoch": 0.362624,
+      "grad_norm": 0.44344380497932434,
+      "learning_rate": 0.0007369030351107232,
+      "loss": 2.9703,
+      "step": 2833
+    },
+    {
+      "epoch": 0.362752,
+      "grad_norm": 0.48593655228614807,
+      "learning_rate": 0.0007367204505280847,
+      "loss": 2.8677,
+      "step": 2834
+    },
+    {
+      "epoch": 0.36288,
+      "grad_norm": 0.4586315155029297,
+      "learning_rate": 0.0007365378252504566,
+      "loss": 2.9148,
+      "step": 2835
+    },
+    {
+      "epoch": 0.363008,
+      "grad_norm": 0.49511969089508057,
+      "learning_rate": 0.0007363551593092342,
+      "loss": 2.9462,
+      "step": 2836
+    },
+    {
+      "epoch": 0.363136,
+      "grad_norm": 0.4542655050754547,
+      "learning_rate": 0.0007361724527358194,
+      "loss": 2.7899,
+      "step": 2837
+    },
+    {
+      "epoch": 0.363264,
+      "grad_norm": 0.43190744519233704,
+      "learning_rate": 0.0007359897055616225,
+      "loss": 3.0018,
+      "step": 2838
+    },
+    {
+      "epoch": 0.363392,
+      "grad_norm": 0.44152262806892395,
+      "learning_rate": 0.0007358069178180592,
+      "loss": 2.9112,
+      "step": 2839
+    },
+    {
+      "epoch": 0.36352,
+      "grad_norm": 0.40912720561027527,
+      "learning_rate": 0.000735624089536553,
+      "loss": 2.731,
+      "step": 2840
+    },
+    {
+      "epoch": 0.363648,
+      "grad_norm": 0.47236838936805725,
+      "learning_rate": 0.0007354412207485342,
+      "loss": 2.9968,
+      "step": 2841
+    },
+    {
+      "epoch": 0.363776,
+      "grad_norm": 0.4425683915615082,
+      "learning_rate": 0.0007352583114854402,
+      "loss": 2.8452,
+      "step": 2842
+    },
+    {
+      "epoch": 0.363904,
+      "grad_norm": 0.45909613370895386,
+      "learning_rate": 0.000735075361778715,
+      "loss": 2.9807,
+      "step": 2843
+    },
+    {
+      "epoch": 0.364032,
+      "grad_norm": 0.4459207355976105,
+      "learning_rate": 0.0007348923716598099,
+      "loss": 2.7998,
+      "step": 2844
+    },
+    {
+      "epoch": 0.36416,
+      "grad_norm": 0.4811474680900574,
+      "learning_rate": 0.0007347093411601831,
+      "loss": 2.9981,
+      "step": 2845
+    },
+    {
+      "epoch": 0.364288,
+      "grad_norm": 0.4658900797367096,
+      "learning_rate": 0.0007345262703112996,
+      "loss": 2.8808,
+      "step": 2846
+    },
+    {
+      "epoch": 0.364416,
+      "grad_norm": 0.49452322721481323,
+      "learning_rate": 0.0007343431591446312,
+      "loss": 2.9245,
+      "step": 2847
+    },
+    {
+      "epoch": 0.364544,
+      "grad_norm": 0.48551779985427856,
+      "learning_rate": 0.0007341600076916571,
+      "loss": 2.7377,
+      "step": 2848
+    },
+    {
+      "epoch": 0.364672,
+      "grad_norm": 0.48844122886657715,
+      "learning_rate": 0.0007339768159838631,
+      "loss": 3.0277,
+      "step": 2849
+    },
+    {
+      "epoch": 0.3648,
+      "grad_norm": 0.49005556106567383,
+      "learning_rate": 0.0007337935840527419,
+      "loss": 2.9689,
+      "step": 2850
+    },
+    {
+      "epoch": 0.364928,
+      "grad_norm": 0.4624228775501251,
+      "learning_rate": 0.000733610311929793,
+      "loss": 2.9326,
+      "step": 2851
+    },
+    {
+      "epoch": 0.365056,
+      "grad_norm": 0.47764500975608826,
+      "learning_rate": 0.0007334269996465235,
+      "loss": 2.7577,
+      "step": 2852
+    },
+    {
+      "epoch": 0.365184,
+      "grad_norm": 0.4675452411174774,
+      "learning_rate": 0.0007332436472344467,
+      "loss": 2.8429,
+      "step": 2853
+    },
+    {
+      "epoch": 0.365312,
+      "grad_norm": 0.43744608759880066,
+      "learning_rate": 0.0007330602547250827,
+      "loss": 2.9632,
+      "step": 2854
+    },
+    {
+      "epoch": 0.36544,
+      "grad_norm": 0.48239997029304504,
+      "learning_rate": 0.0007328768221499594,
+      "loss": 2.9471,
+      "step": 2855
+    },
+    {
+      "epoch": 0.365568,
+      "grad_norm": 0.4512200355529785,
+      "learning_rate": 0.0007326933495406103,
+      "loss": 2.8353,
+      "step": 2856
+    },
+    {
+      "epoch": 0.365696,
+      "grad_norm": 0.4707525372505188,
+      "learning_rate": 0.000732509836928577,
+      "loss": 2.8935,
+      "step": 2857
+    },
+    {
+      "epoch": 0.365824,
+      "grad_norm": 0.4261745512485504,
+      "learning_rate": 0.0007323262843454071,
+      "loss": 2.8908,
+      "step": 2858
+    },
+    {
+      "epoch": 0.365952,
+      "grad_norm": 0.4242784380912781,
+      "learning_rate": 0.0007321426918226557,
+      "loss": 2.8306,
+      "step": 2859
+    },
+    {
+      "epoch": 0.36608,
+      "grad_norm": 0.4405313730239868,
+      "learning_rate": 0.0007319590593918843,
+      "loss": 2.9672,
+      "step": 2860
+    },
+    {
+      "epoch": 0.366208,
+      "grad_norm": 0.4485013782978058,
+      "learning_rate": 0.0007317753870846614,
+      "loss": 2.9008,
+      "step": 2861
+    },
+    {
+      "epoch": 0.366336,
+      "grad_norm": 0.537553071975708,
+      "learning_rate": 0.0007315916749325625,
+      "loss": 2.942,
+      "step": 2862
+    },
+    {
+      "epoch": 0.366464,
+      "grad_norm": 0.44233742356300354,
+      "learning_rate": 0.0007314079229671699,
+      "loss": 2.8292,
+      "step": 2863
+    },
+    {
+      "epoch": 0.366592,
+      "grad_norm": 0.4790656864643097,
+      "learning_rate": 0.0007312241312200727,
+      "loss": 3.0025,
+      "step": 2864
+    },
+    {
+      "epoch": 0.36672,
+      "grad_norm": 0.4605633616447449,
+      "learning_rate": 0.0007310402997228667,
+      "loss": 2.8715,
+      "step": 2865
+    },
+    {
+      "epoch": 0.366848,
+      "grad_norm": 0.4453747868537903,
+      "learning_rate": 0.0007308564285071546,
+      "loss": 2.9692,
+      "step": 2866
+    },
+    {
+      "epoch": 0.366976,
+      "grad_norm": 0.4375332295894623,
+      "learning_rate": 0.0007306725176045464,
+      "loss": 2.8913,
+      "step": 2867
+    },
+    {
+      "epoch": 0.367104,
+      "grad_norm": 0.4315635561943054,
+      "learning_rate": 0.0007304885670466582,
+      "loss": 2.7754,
+      "step": 2868
+    },
+    {
+      "epoch": 0.367232,
+      "grad_norm": 0.4535144865512848,
+      "learning_rate": 0.000730304576865113,
+      "loss": 2.8079,
+      "step": 2869
+    },
+    {
+      "epoch": 0.36736,
+      "grad_norm": 0.4716770052909851,
+      "learning_rate": 0.0007301205470915412,
+      "loss": 2.8322,
+      "step": 2870
+    },
+    {
+      "epoch": 0.367488,
+      "grad_norm": 0.44651713967323303,
+      "learning_rate": 0.0007299364777575796,
+      "loss": 2.8443,
+      "step": 2871
+    },
+    {
+      "epoch": 0.367616,
+      "grad_norm": 0.4835733473300934,
+      "learning_rate": 0.000729752368894872,
+      "loss": 2.7885,
+      "step": 2872
+    },
+    {
+      "epoch": 0.367744,
+      "grad_norm": 0.8274149298667908,
+      "learning_rate": 0.0007295682205350684,
+      "loss": 2.7991,
+      "step": 2873
+    },
+    {
+      "epoch": 0.367872,
+      "grad_norm": 0.45919787883758545,
+      "learning_rate": 0.0007293840327098264,
+      "loss": 2.8486,
+      "step": 2874
+    },
+    {
+      "epoch": 0.368,
+      "grad_norm": 0.507565975189209,
+      "learning_rate": 0.0007291998054508101,
+      "loss": 2.9359,
+      "step": 2875
+    },
+    {
+      "epoch": 0.368128,
+      "grad_norm": 0.44228029251098633,
+      "learning_rate": 0.0007290155387896899,
+      "loss": 2.9361,
+      "step": 2876
+    },
+    {
+      "epoch": 0.368256,
+      "grad_norm": 0.435350626707077,
+      "learning_rate": 0.0007288312327581438,
+      "loss": 2.9721,
+      "step": 2877
+    },
+    {
+      "epoch": 0.368384,
+      "grad_norm": 0.4739178717136383,
+      "learning_rate": 0.0007286468873878559,
+      "loss": 2.8816,
+      "step": 2878
+    },
+    {
+      "epoch": 0.368512,
+      "grad_norm": 0.47682511806488037,
+      "learning_rate": 0.0007284625027105172,
+      "loss": 2.9194,
+      "step": 2879
+    },
+    {
+      "epoch": 0.36864,
+      "grad_norm": 0.4488813579082489,
+      "learning_rate": 0.0007282780787578258,
+      "loss": 2.8088,
+      "step": 2880
+    },
+    {
+      "epoch": 0.368768,
+      "grad_norm": 0.46638354659080505,
+      "learning_rate": 0.0007280936155614863,
+      "loss": 2.8803,
+      "step": 2881
+    },
+    {
+      "epoch": 0.368896,
+      "grad_norm": 0.4864390790462494,
+      "learning_rate": 0.0007279091131532101,
+      "loss": 2.912,
+      "step": 2882
+    },
+    {
+      "epoch": 0.369024,
+      "grad_norm": 0.5006654262542725,
+      "learning_rate": 0.000727724571564715,
+      "loss": 2.9722,
+      "step": 2883
+    },
+    {
+      "epoch": 0.369152,
+      "grad_norm": 0.46475720405578613,
+      "learning_rate": 0.000727539990827726,
+      "loss": 2.9887,
+      "step": 2884
+    },
+    {
+      "epoch": 0.36928,
+      "grad_norm": 0.501327633857727,
+      "learning_rate": 0.0007273553709739749,
+      "loss": 2.9763,
+      "step": 2885
+    },
+    {
+      "epoch": 0.369408,
+      "grad_norm": 0.4743506610393524,
+      "learning_rate": 0.0007271707120351996,
+      "loss": 2.9386,
+      "step": 2886
+    },
+    {
+      "epoch": 0.369536,
+      "grad_norm": 0.4601151943206787,
+      "learning_rate": 0.0007269860140431454,
+      "loss": 3.0203,
+      "step": 2887
+    },
+    {
+      "epoch": 0.369664,
+      "grad_norm": 0.4504337012767792,
+      "learning_rate": 0.000726801277029564,
+      "loss": 2.7948,
+      "step": 2888
+    },
+    {
+      "epoch": 0.369792,
+      "grad_norm": 0.5354257822036743,
+      "learning_rate": 0.0007266165010262138,
+      "loss": 2.9471,
+      "step": 2889
+    },
+    {
+      "epoch": 0.36992,
+      "grad_norm": 0.4520520269870758,
+      "learning_rate": 0.0007264316860648598,
+      "loss": 2.8507,
+      "step": 2890
+    },
+    {
+      "epoch": 0.370048,
+      "grad_norm": 0.4487895667552948,
+      "learning_rate": 0.000726246832177274,
+      "loss": 2.7872,
+      "step": 2891
+    },
+    {
+      "epoch": 0.370176,
+      "grad_norm": 0.48715221881866455,
+      "learning_rate": 0.000726061939395235,
+      "loss": 2.9862,
+      "step": 2892
+    },
+    {
+      "epoch": 0.370304,
+      "grad_norm": 0.46618106961250305,
+      "learning_rate": 0.0007258770077505275,
+      "loss": 2.9259,
+      "step": 2893
+    },
+    {
+      "epoch": 0.370432,
+      "grad_norm": 0.48139867186546326,
+      "learning_rate": 0.000725692037274944,
+      "loss": 2.833,
+      "step": 2894
+    },
+    {
+      "epoch": 0.37056,
+      "grad_norm": 0.4493372440338135,
+      "learning_rate": 0.0007255070280002829,
+      "loss": 2.7744,
+      "step": 2895
+    },
+    {
+      "epoch": 0.370688,
+      "grad_norm": 0.4387955069541931,
+      "learning_rate": 0.0007253219799583495,
+      "loss": 2.8385,
+      "step": 2896
+    },
+    {
+      "epoch": 0.370816,
+      "grad_norm": 0.416663259267807,
+      "learning_rate": 0.0007251368931809553,
+      "loss": 2.8885,
+      "step": 2897
+    },
+    {
+      "epoch": 0.370944,
+      "grad_norm": 0.45950615406036377,
+      "learning_rate": 0.0007249517676999192,
+      "loss": 2.9951,
+      "step": 2898
+    },
+    {
+      "epoch": 0.371072,
+      "grad_norm": 0.4518347382545471,
+      "learning_rate": 0.0007247666035470666,
+      "loss": 2.8524,
+      "step": 2899
+    },
+    {
+      "epoch": 0.3712,
+      "grad_norm": 0.4423445761203766,
+      "learning_rate": 0.0007245814007542289,
+      "loss": 2.7451,
+      "step": 2900
+    },
+    {
+      "epoch": 0.371328,
+      "grad_norm": 0.4935082495212555,
+      "learning_rate": 0.000724396159353245,
+      "loss": 2.9518,
+      "step": 2901
+    },
+    {
+      "epoch": 0.371456,
+      "grad_norm": 0.4129147529602051,
+      "learning_rate": 0.0007242108793759596,
+      "loss": 2.8187,
+      "step": 2902
+    },
+    {
+      "epoch": 0.371584,
+      "grad_norm": 0.45092612504959106,
+      "learning_rate": 0.0007240255608542251,
+      "loss": 2.933,
+      "step": 2903
+    },
+    {
+      "epoch": 0.371712,
+      "grad_norm": 0.43859758973121643,
+      "learning_rate": 0.0007238402038198994,
+      "loss": 2.9897,
+      "step": 2904
+    },
+    {
+      "epoch": 0.37184,
+      "grad_norm": 0.49950143694877625,
+      "learning_rate": 0.0007236548083048478,
+      "loss": 2.9178,
+      "step": 2905
+    },
+    {
+      "epoch": 0.371968,
+      "grad_norm": 0.46168580651283264,
+      "learning_rate": 0.0007234693743409417,
+      "loss": 2.9769,
+      "step": 2906
+    },
+    {
+      "epoch": 0.372096,
+      "grad_norm": 0.4487912654876709,
+      "learning_rate": 0.0007232839019600594,
+      "loss": 2.8931,
+      "step": 2907
+    },
+    {
+      "epoch": 0.372224,
+      "grad_norm": 0.43640950322151184,
+      "learning_rate": 0.0007230983911940861,
+      "loss": 2.7816,
+      "step": 2908
+    },
+    {
+      "epoch": 0.372352,
+      "grad_norm": 0.4629068970680237,
+      "learning_rate": 0.0007229128420749127,
+      "loss": 2.8728,
+      "step": 2909
+    },
+    {
+      "epoch": 0.37248,
+      "grad_norm": 0.4380590617656708,
+      "learning_rate": 0.0007227272546344377,
+      "loss": 2.8829,
+      "step": 2910
+    },
+    {
+      "epoch": 0.372608,
+      "grad_norm": 0.45174503326416016,
+      "learning_rate": 0.0007225416289045654,
+      "loss": 2.9965,
+      "step": 2911
+    },
+    {
+      "epoch": 0.372736,
+      "grad_norm": 0.48155173659324646,
+      "learning_rate": 0.0007223559649172073,
+      "loss": 2.7541,
+      "step": 2912
+    },
+    {
+      "epoch": 0.372864,
+      "grad_norm": 0.4813424348831177,
+      "learning_rate": 0.000722170262704281,
+      "loss": 2.9782,
+      "step": 2913
+    },
+    {
+      "epoch": 0.372992,
+      "grad_norm": 0.4519997537136078,
+      "learning_rate": 0.0007219845222977109,
+      "loss": 2.9284,
+      "step": 2914
+    },
+    {
+      "epoch": 0.37312,
+      "grad_norm": 0.46581873297691345,
+      "learning_rate": 0.000721798743729428,
+      "loss": 3.0328,
+      "step": 2915
+    },
+    {
+      "epoch": 0.373248,
+      "grad_norm": 0.42827707529067993,
+      "learning_rate": 0.0007216129270313697,
+      "loss": 2.7909,
+      "step": 2916
+    },
+    {
+      "epoch": 0.373376,
+      "grad_norm": 0.4339712858200073,
+      "learning_rate": 0.0007214270722354801,
+      "loss": 2.8937,
+      "step": 2917
+    },
+    {
+      "epoch": 0.373504,
+      "grad_norm": 0.5552771687507629,
+      "learning_rate": 0.00072124117937371,
+      "loss": 2.9589,
+      "step": 2918
+    },
+    {
+      "epoch": 0.373632,
+      "grad_norm": 0.4518507719039917,
+      "learning_rate": 0.000721055248478016,
+      "loss": 2.8937,
+      "step": 2919
+    },
+    {
+      "epoch": 0.37376,
+      "grad_norm": 0.46659064292907715,
+      "learning_rate": 0.0007208692795803622,
+      "loss": 2.9477,
+      "step": 2920
+    },
+    {
+      "epoch": 0.373888,
+      "grad_norm": 0.4712584912776947,
+      "learning_rate": 0.0007206832727127186,
+      "loss": 2.8732,
+      "step": 2921
+    },
+    {
+      "epoch": 0.374016,
+      "grad_norm": 0.4917088449001312,
+      "learning_rate": 0.0007204972279070622,
+      "loss": 2.9231,
+      "step": 2922
+    },
+    {
+      "epoch": 0.374144,
+      "grad_norm": 0.45806917548179626,
+      "learning_rate": 0.000720311145195376,
+      "loss": 2.855,
+      "step": 2923
+    },
+    {
+      "epoch": 0.374272,
+      "grad_norm": 0.4482679069042206,
+      "learning_rate": 0.00072012502460965,
+      "loss": 2.945,
+      "step": 2924
+    },
+    {
+      "epoch": 0.3744,
+      "grad_norm": 0.464298278093338,
+      "learning_rate": 0.0007199388661818802,
+      "loss": 2.9014,
+      "step": 2925
+    },
+    {
+      "epoch": 0.374528,
+      "grad_norm": 0.43307816982269287,
+      "learning_rate": 0.0007197526699440697,
+      "loss": 2.6801,
+      "step": 2926
+    },
+    {
+      "epoch": 0.374656,
+      "grad_norm": 0.4530610144138336,
+      "learning_rate": 0.0007195664359282275,
+      "loss": 2.8963,
+      "step": 2927
+    },
+    {
+      "epoch": 0.374784,
+      "grad_norm": 0.4543643295764923,
+      "learning_rate": 0.0007193801641663697,
+      "loss": 2.8356,
+      "step": 2928
+    },
+    {
+      "epoch": 0.374912,
+      "grad_norm": 0.4351753890514374,
+      "learning_rate": 0.0007191938546905183,
+      "loss": 2.7577,
+      "step": 2929
+    },
+    {
+      "epoch": 0.37504,
+      "grad_norm": 0.4449367821216583,
+      "learning_rate": 0.0007190075075327021,
+      "loss": 2.8872,
+      "step": 2930
+    },
+    {
+      "epoch": 0.375168,
+      "grad_norm": 0.48438531160354614,
+      "learning_rate": 0.0007188211227249565,
+      "loss": 2.8695,
+      "step": 2931
+    },
+    {
+      "epoch": 0.375296,
+      "grad_norm": 0.4982151985168457,
+      "learning_rate": 0.0007186347002993232,
+      "loss": 2.8958,
+      "step": 2932
+    },
+    {
+      "epoch": 0.375424,
+      "grad_norm": 0.47176551818847656,
+      "learning_rate": 0.0007184482402878501,
+      "loss": 2.9058,
+      "step": 2933
+    },
+    {
+      "epoch": 0.375552,
+      "grad_norm": 0.4563322067260742,
+      "learning_rate": 0.0007182617427225922,
+      "loss": 3.0017,
+      "step": 2934
+    },
+    {
+      "epoch": 0.37568,
+      "grad_norm": 0.47470685839653015,
+      "learning_rate": 0.0007180752076356102,
+      "loss": 2.8444,
+      "step": 2935
+    },
+    {
+      "epoch": 0.375808,
+      "grad_norm": 0.46827512979507446,
+      "learning_rate": 0.0007178886350589721,
+      "loss": 2.9242,
+      "step": 2936
+    },
+    {
+      "epoch": 0.375936,
+      "grad_norm": 0.46549075841903687,
+      "learning_rate": 0.0007177020250247514,
+      "loss": 2.9635,
+      "step": 2937
+    },
+    {
+      "epoch": 0.376064,
+      "grad_norm": 0.4309782385826111,
+      "learning_rate": 0.0007175153775650289,
+      "loss": 2.9906,
+      "step": 2938
+    },
+    {
+      "epoch": 0.376192,
+      "grad_norm": 0.5061914324760437,
+      "learning_rate": 0.0007173286927118914,
+      "loss": 3.0396,
+      "step": 2939
+    },
+    {
+      "epoch": 0.37632,
+      "grad_norm": 0.4561704397201538,
+      "learning_rate": 0.000717141970497432,
+      "loss": 2.986,
+      "step": 2940
+    },
+    {
+      "epoch": 0.376448,
+      "grad_norm": 0.4410429000854492,
+      "learning_rate": 0.0007169552109537507,
+      "loss": 2.8369,
+      "step": 2941
+    },
+    {
+      "epoch": 0.376576,
+      "grad_norm": 0.48430484533309937,
+      "learning_rate": 0.0007167684141129536,
+      "loss": 2.9231,
+      "step": 2942
+    },
+    {
+      "epoch": 0.376704,
+      "grad_norm": 0.43939387798309326,
+      "learning_rate": 0.0007165815800071529,
+      "loss": 2.9859,
+      "step": 2943
+    },
+    {
+      "epoch": 0.376832,
+      "grad_norm": 0.49426761269569397,
+      "learning_rate": 0.0007163947086684676,
+      "loss": 2.9047,
+      "step": 2944
+    },
+    {
+      "epoch": 0.37696,
+      "grad_norm": 0.4633600115776062,
+      "learning_rate": 0.0007162078001290235,
+      "loss": 2.8284,
+      "step": 2945
+    },
+    {
+      "epoch": 0.377088,
+      "grad_norm": 0.48416468501091003,
+      "learning_rate": 0.0007160208544209521,
+      "loss": 2.9308,
+      "step": 2946
+    },
+    {
+      "epoch": 0.377216,
+      "grad_norm": 0.4370930790901184,
+      "learning_rate": 0.0007158338715763913,
+      "loss": 2.7758,
+      "step": 2947
+    },
+    {
+      "epoch": 0.377344,
+      "grad_norm": 0.44038596749305725,
+      "learning_rate": 0.0007156468516274859,
+      "loss": 2.9447,
+      "step": 2948
+    },
+    {
+      "epoch": 0.377472,
+      "grad_norm": 0.4827476143836975,
+      "learning_rate": 0.0007154597946063867,
+      "loss": 2.7852,
+      "step": 2949
+    },
+    {
+      "epoch": 0.3776,
+      "grad_norm": 0.4602373540401459,
+      "learning_rate": 0.000715272700545251,
+      "loss": 2.951,
+      "step": 2950
+    },
+    {
+      "epoch": 0.377728,
+      "grad_norm": 0.4485637843608856,
+      "learning_rate": 0.0007150855694762424,
+      "loss": 2.9089,
+      "step": 2951
+    },
+    {
+      "epoch": 0.377856,
+      "grad_norm": 0.42191842198371887,
+      "learning_rate": 0.000714898401431531,
+      "loss": 2.9056,
+      "step": 2952
+    },
+    {
+      "epoch": 0.377984,
+      "grad_norm": 0.4374709725379944,
+      "learning_rate": 0.000714711196443293,
+      "loss": 2.9165,
+      "step": 2953
+    },
+    {
+      "epoch": 0.378112,
+      "grad_norm": 0.442513108253479,
+      "learning_rate": 0.0007145239545437112,
+      "loss": 3.0345,
+      "step": 2954
+    },
+    {
+      "epoch": 0.37824,
+      "grad_norm": 0.4086267650127411,
+      "learning_rate": 0.0007143366757649746,
+      "loss": 2.782,
+      "step": 2955
+    },
+    {
+      "epoch": 0.378368,
+      "grad_norm": 0.4141438603401184,
+      "learning_rate": 0.0007141493601392787,
+      "loss": 2.8468,
+      "step": 2956
+    },
+    {
+      "epoch": 0.378496,
+      "grad_norm": 0.4317202568054199,
+      "learning_rate": 0.0007139620076988251,
+      "loss": 2.7881,
+      "step": 2957
+    },
+    {
+      "epoch": 0.378624,
+      "grad_norm": 0.48754990100860596,
+      "learning_rate": 0.0007137746184758218,
+      "loss": 2.9689,
+      "step": 2958
+    },
+    {
+      "epoch": 0.378752,
+      "grad_norm": 0.45136964321136475,
+      "learning_rate": 0.0007135871925024835,
+      "loss": 2.9379,
+      "step": 2959
+    },
+    {
+      "epoch": 0.37888,
+      "grad_norm": 0.4639531970024109,
+      "learning_rate": 0.0007133997298110307,
+      "loss": 2.9383,
+      "step": 2960
+    },
+    {
+      "epoch": 0.379008,
+      "grad_norm": 0.5146617293357849,
+      "learning_rate": 0.0007132122304336903,
+      "loss": 3.0616,
+      "step": 2961
+    },
+    {
+      "epoch": 0.379136,
+      "grad_norm": 0.5123382806777954,
+      "learning_rate": 0.0007130246944026956,
+      "loss": 2.8144,
+      "step": 2962
+    },
+    {
+      "epoch": 0.379264,
+      "grad_norm": 0.4475536048412323,
+      "learning_rate": 0.0007128371217502867,
+      "loss": 2.8346,
+      "step": 2963
+    },
+    {
+      "epoch": 0.379392,
+      "grad_norm": 0.4387931823730469,
+      "learning_rate": 0.000712649512508709,
+      "loss": 2.8334,
+      "step": 2964
+    },
+    {
+      "epoch": 0.37952,
+      "grad_norm": 0.4352128207683563,
+      "learning_rate": 0.0007124618667102147,
+      "loss": 2.8082,
+      "step": 2965
+    },
+    {
+      "epoch": 0.379648,
+      "grad_norm": 0.44317954778671265,
+      "learning_rate": 0.0007122741843870626,
+      "loss": 2.829,
+      "step": 2966
+    },
+    {
+      "epoch": 0.379776,
+      "grad_norm": 0.4498663544654846,
+      "learning_rate": 0.0007120864655715171,
+      "loss": 2.8502,
+      "step": 2967
+    },
+    {
+      "epoch": 0.379904,
+      "grad_norm": 0.4274469017982483,
+      "learning_rate": 0.0007118987102958498,
+      "loss": 2.891,
+      "step": 2968
+    },
+    {
+      "epoch": 0.380032,
+      "grad_norm": 0.4761981964111328,
+      "learning_rate": 0.0007117109185923373,
+      "loss": 2.8595,
+      "step": 2969
+    },
+    {
+      "epoch": 0.38016,
+      "grad_norm": 0.4716610014438629,
+      "learning_rate": 0.0007115230904932639,
+      "loss": 3.0287,
+      "step": 2970
+    },
+    {
+      "epoch": 0.380288,
+      "grad_norm": 0.43786659836769104,
+      "learning_rate": 0.0007113352260309188,
+      "loss": 2.8423,
+      "step": 2971
+    },
+    {
+      "epoch": 0.380416,
+      "grad_norm": 0.46233025193214417,
+      "learning_rate": 0.0007111473252375983,
+      "loss": 2.8512,
+      "step": 2972
+    },
+    {
+      "epoch": 0.380544,
+      "grad_norm": 0.4262726604938507,
+      "learning_rate": 0.0007109593881456048,
+      "loss": 2.7826,
+      "step": 2973
+    },
+    {
+      "epoch": 0.380672,
+      "grad_norm": 0.4779355525970459,
+      "learning_rate": 0.000710771414787247,
+      "loss": 2.9846,
+      "step": 2974
+    },
+    {
+      "epoch": 0.3808,
+      "grad_norm": 0.45237651467323303,
+      "learning_rate": 0.0007105834051948395,
+      "loss": 2.8029,
+      "step": 2975
+    },
+    {
+      "epoch": 0.380928,
+      "grad_norm": 0.48962000012397766,
+      "learning_rate": 0.0007103953594007029,
+      "loss": 2.8431,
+      "step": 2976
+    },
+    {
+      "epoch": 0.381056,
+      "grad_norm": 0.5096243619918823,
+      "learning_rate": 0.0007102072774371654,
+      "loss": 2.8184,
+      "step": 2977
+    },
+    {
+      "epoch": 0.381184,
+      "grad_norm": 0.47960638999938965,
+      "learning_rate": 0.0007100191593365599,
+      "loss": 3.0134,
+      "step": 2978
+    },
+    {
+      "epoch": 0.381312,
+      "grad_norm": 0.4587400257587433,
+      "learning_rate": 0.000709831005131226,
+      "loss": 2.9736,
+      "step": 2979
+    },
+    {
+      "epoch": 0.38144,
+      "grad_norm": 0.48440834879875183,
+      "learning_rate": 0.00070964281485351,
+      "loss": 2.832,
+      "step": 2980
+    },
+    {
+      "epoch": 0.381568,
+      "grad_norm": 0.45781829953193665,
+      "learning_rate": 0.0007094545885357635,
+      "loss": 2.9425,
+      "step": 2981
+    },
+    {
+      "epoch": 0.381696,
+      "grad_norm": 0.4254377782344818,
+      "learning_rate": 0.0007092663262103451,
+      "loss": 2.7847,
+      "step": 2982
+    },
+    {
+      "epoch": 0.381824,
+      "grad_norm": 0.45893633365631104,
+      "learning_rate": 0.0007090780279096192,
+      "loss": 2.8358,
+      "step": 2983
+    },
+    {
+      "epoch": 0.381952,
+      "grad_norm": 0.45882534980773926,
+      "learning_rate": 0.0007088896936659566,
+      "loss": 2.941,
+      "step": 2984
+    },
+    {
+      "epoch": 0.38208,
+      "grad_norm": 0.46347323060035706,
+      "learning_rate": 0.0007087013235117338,
+      "loss": 2.9306,
+      "step": 2985
+    },
+    {
+      "epoch": 0.382208,
+      "grad_norm": 0.47694486379623413,
+      "learning_rate": 0.0007085129174793342,
+      "loss": 2.8339,
+      "step": 2986
+    },
+    {
+      "epoch": 0.382336,
+      "grad_norm": 0.4312053620815277,
+      "learning_rate": 0.0007083244756011465,
+      "loss": 2.8848,
+      "step": 2987
+    },
+    {
+      "epoch": 0.382464,
+      "grad_norm": 0.4300554692745209,
+      "learning_rate": 0.0007081359979095666,
+      "loss": 2.7562,
+      "step": 2988
+    },
+    {
+      "epoch": 0.382592,
+      "grad_norm": 0.47886788845062256,
+      "learning_rate": 0.0007079474844369957,
+      "loss": 3.0211,
+      "step": 2989
+    },
+    {
+      "epoch": 0.38272,
+      "grad_norm": 0.49065545201301575,
+      "learning_rate": 0.0007077589352158415,
+      "loss": 3.0203,
+      "step": 2990
+    },
+    {
+      "epoch": 0.382848,
+      "grad_norm": 0.45889657735824585,
+      "learning_rate": 0.0007075703502785177,
+      "loss": 2.8298,
+      "step": 2991
+    },
+    {
+      "epoch": 0.382976,
+      "grad_norm": 0.5127334594726562,
+      "learning_rate": 0.0007073817296574444,
+      "loss": 3.0046,
+      "step": 2992
+    },
+    {
+      "epoch": 0.383104,
+      "grad_norm": 0.4352664351463318,
+      "learning_rate": 0.0007071930733850476,
+      "loss": 2.9739,
+      "step": 2993
+    },
+    {
+      "epoch": 0.383232,
+      "grad_norm": 0.4355897605419159,
+      "learning_rate": 0.0007070043814937594,
+      "loss": 2.7149,
+      "step": 2994
+    },
+    {
+      "epoch": 0.38336,
+      "grad_norm": 0.4716731309890747,
+      "learning_rate": 0.0007068156540160182,
+      "loss": 2.8759,
+      "step": 2995
+    },
+    {
+      "epoch": 0.383488,
+      "grad_norm": 0.4631376266479492,
+      "learning_rate": 0.0007066268909842687,
+      "loss": 2.8542,
+      "step": 2996
+    },
+    {
+      "epoch": 0.383616,
+      "grad_norm": 0.4793536961078644,
+      "learning_rate": 0.000706438092430961,
+      "loss": 2.9202,
+      "step": 2997
+    },
+    {
+      "epoch": 0.383744,
+      "grad_norm": 0.4472249746322632,
+      "learning_rate": 0.000706249258388552,
+      "loss": 2.825,
+      "step": 2998
+    },
+    {
+      "epoch": 0.383872,
+      "grad_norm": 0.45877957344055176,
+      "learning_rate": 0.0007060603888895045,
+      "loss": 2.8722,
+      "step": 2999
+    },
+    {
+      "epoch": 0.384,
+      "grad_norm": 0.47359222173690796,
+      "learning_rate": 0.0007058714839662874,
+      "loss": 2.9235,
+      "step": 3000
+    },
+    {
+      "epoch": 0.384128,
+      "grad_norm": 0.45524320006370544,
+      "learning_rate": 0.0007056825436513753,
+      "loss": 2.8435,
+      "step": 3001
+    },
+    {
+      "epoch": 0.384256,
+      "grad_norm": 0.42198091745376587,
+      "learning_rate": 0.0007054935679772496,
+      "loss": 2.828,
+      "step": 3002
+    },
+    {
+      "epoch": 0.384384,
+      "grad_norm": 0.45144176483154297,
+      "learning_rate": 0.0007053045569763973,
+      "loss": 3.0157,
+      "step": 3003
+    },
+    {
+      "epoch": 0.384512,
+      "grad_norm": 0.4143078923225403,
+      "learning_rate": 0.0007051155106813114,
+      "loss": 2.9692,
+      "step": 3004
+    },
+    {
+      "epoch": 0.38464,
+      "grad_norm": 0.4335055947303772,
+      "learning_rate": 0.0007049264291244914,
+      "loss": 2.8655,
+      "step": 3005
+    },
+    {
+      "epoch": 0.384768,
+      "grad_norm": 0.4362388253211975,
+      "learning_rate": 0.0007047373123384426,
+      "loss": 2.8841,
+      "step": 3006
+    },
+    {
+      "epoch": 0.384896,
+      "grad_norm": 0.40873610973358154,
+      "learning_rate": 0.0007045481603556763,
+      "loss": 2.787,
+      "step": 3007
+    },
+    {
+      "epoch": 0.385024,
+      "grad_norm": 0.45844846963882446,
+      "learning_rate": 0.0007043589732087098,
+      "loss": 2.9147,
+      "step": 3008
+    },
+    {
+      "epoch": 0.385152,
+      "grad_norm": 0.4858155846595764,
+      "learning_rate": 0.0007041697509300667,
+      "loss": 2.9072,
+      "step": 3009
+    },
+    {
+      "epoch": 0.38528,
+      "grad_norm": 0.510715901851654,
+      "learning_rate": 0.0007039804935522765,
+      "loss": 3.018,
+      "step": 3010
+    },
+    {
+      "epoch": 0.385408,
+      "grad_norm": 0.4052485227584839,
+      "learning_rate": 0.0007037912011078749,
+      "loss": 2.8763,
+      "step": 3011
+    },
+    {
+      "epoch": 0.385536,
+      "grad_norm": 0.4930177628993988,
+      "learning_rate": 0.000703601873629403,
+      "loss": 2.9742,
+      "step": 3012
+    },
+    {
+      "epoch": 0.385664,
+      "grad_norm": 0.43468740582466125,
+      "learning_rate": 0.0007034125111494087,
+      "loss": 2.9096,
+      "step": 3013
+    },
+    {
+      "epoch": 0.385792,
+      "grad_norm": 0.4478032886981964,
+      "learning_rate": 0.0007032231137004457,
+      "loss": 2.8229,
+      "step": 3014
+    },
+    {
+      "epoch": 0.38592,
+      "grad_norm": 0.4665033519268036,
+      "learning_rate": 0.0007030336813150733,
+      "loss": 2.8813,
+      "step": 3015
+    },
+    {
+      "epoch": 0.386048,
+      "grad_norm": 0.5207582712173462,
+      "learning_rate": 0.0007028442140258575,
+      "loss": 3.0199,
+      "step": 3016
+    },
+    {
+      "epoch": 0.386176,
+      "grad_norm": 0.4530661404132843,
+      "learning_rate": 0.0007026547118653696,
+      "loss": 2.8128,
+      "step": 3017
+    },
+    {
+      "epoch": 0.386304,
+      "grad_norm": 0.4696804881095886,
+      "learning_rate": 0.0007024651748661875,
+      "loss": 2.7908,
+      "step": 3018
+    },
+    {
+      "epoch": 0.386432,
+      "grad_norm": 0.4973638355731964,
+      "learning_rate": 0.0007022756030608946,
+      "loss": 2.9861,
+      "step": 3019
+    },
+    {
+      "epoch": 0.38656,
+      "grad_norm": 0.4416906535625458,
+      "learning_rate": 0.0007020859964820805,
+      "loss": 2.9055,
+      "step": 3020
+    },
+    {
+      "epoch": 0.386688,
+      "grad_norm": 0.46195489168167114,
+      "learning_rate": 0.000701896355162341,
+      "loss": 2.9671,
+      "step": 3021
+    },
+    {
+      "epoch": 0.386816,
+      "grad_norm": 0.4480120539665222,
+      "learning_rate": 0.0007017066791342772,
+      "loss": 2.9453,
+      "step": 3022
+    },
+    {
+      "epoch": 0.386944,
+      "grad_norm": 0.43126121163368225,
+      "learning_rate": 0.0007015169684304969,
+      "loss": 2.8765,
+      "step": 3023
+    },
+    {
+      "epoch": 0.387072,
+      "grad_norm": 0.47261494398117065,
+      "learning_rate": 0.0007013272230836138,
+      "loss": 2.9186,
+      "step": 3024
+    },
+    {
+      "epoch": 0.3872,
+      "grad_norm": 0.48162180185317993,
+      "learning_rate": 0.000701137443126247,
+      "loss": 2.9497,
+      "step": 3025
+    },
+    {
+      "epoch": 0.387328,
+      "grad_norm": 0.47541943192481995,
+      "learning_rate": 0.0007009476285910218,
+      "loss": 2.8668,
+      "step": 3026
+    },
+    {
+      "epoch": 0.387456,
+      "grad_norm": 0.44699275493621826,
+      "learning_rate": 0.0007007577795105696,
+      "loss": 2.8733,
+      "step": 3027
+    },
+    {
+      "epoch": 0.387584,
+      "grad_norm": 0.4725266396999359,
+      "learning_rate": 0.0007005678959175279,
+      "loss": 3.0477,
+      "step": 3028
+    },
+    {
+      "epoch": 0.387712,
+      "grad_norm": 0.42495957016944885,
+      "learning_rate": 0.0007003779778445397,
+      "loss": 2.794,
+      "step": 3029
+    },
+    {
+      "epoch": 0.38784,
+      "grad_norm": 0.44257277250289917,
+      "learning_rate": 0.0007001880253242541,
+      "loss": 2.8636,
+      "step": 3030
+    },
+    {
+      "epoch": 0.387968,
+      "grad_norm": 0.42430779337882996,
+      "learning_rate": 0.0006999980383893261,
+      "loss": 2.797,
+      "step": 3031
+    },
+    {
+      "epoch": 0.388096,
+      "grad_norm": 0.43292656540870667,
+      "learning_rate": 0.0006998080170724167,
+      "loss": 2.9606,
+      "step": 3032
+    },
+    {
+      "epoch": 0.388224,
+      "grad_norm": 0.4437865912914276,
+      "learning_rate": 0.0006996179614061929,
+      "loss": 2.9004,
+      "step": 3033
+    },
+    {
+      "epoch": 0.388352,
+      "grad_norm": 0.44787824153900146,
+      "learning_rate": 0.0006994278714233273,
+      "loss": 2.8423,
+      "step": 3034
+    },
+    {
+      "epoch": 0.38848,
+      "grad_norm": 0.4528372585773468,
+      "learning_rate": 0.0006992377471564987,
+      "loss": 2.804,
+      "step": 3035
+    },
+    {
+      "epoch": 0.388608,
+      "grad_norm": 0.4523576498031616,
+      "learning_rate": 0.0006990475886383915,
+      "loss": 2.8497,
+      "step": 3036
+    },
+    {
+      "epoch": 0.388736,
+      "grad_norm": 0.4472360610961914,
+      "learning_rate": 0.0006988573959016963,
+      "loss": 2.8904,
+      "step": 3037
+    },
+    {
+      "epoch": 0.388864,
+      "grad_norm": 0.4421881139278412,
+      "learning_rate": 0.0006986671689791094,
+      "loss": 2.8144,
+      "step": 3038
+    },
+    {
+      "epoch": 0.388992,
+      "grad_norm": 0.48519840836524963,
+      "learning_rate": 0.0006984769079033331,
+      "loss": 3.0093,
+      "step": 3039
+    },
+    {
+      "epoch": 0.38912,
+      "grad_norm": 0.4381081163883209,
+      "learning_rate": 0.0006982866127070753,
+      "loss": 2.8804,
+      "step": 3040
+    },
+    {
+      "epoch": 0.389248,
+      "grad_norm": 0.4634467661380768,
+      "learning_rate": 0.0006980962834230499,
+      "loss": 2.8994,
+      "step": 3041
+    },
+    {
+      "epoch": 0.389376,
+      "grad_norm": 0.4717426896095276,
+      "learning_rate": 0.0006979059200839771,
+      "loss": 2.8262,
+      "step": 3042
+    },
+    {
+      "epoch": 0.389504,
+      "grad_norm": 0.4416886866092682,
+      "learning_rate": 0.0006977155227225822,
+      "loss": 2.9219,
+      "step": 3043
+    },
+    {
+      "epoch": 0.389632,
+      "grad_norm": 0.4685446619987488,
+      "learning_rate": 0.0006975250913715968,
+      "loss": 2.8066,
+      "step": 3044
+    },
+    {
+      "epoch": 0.38976,
+      "grad_norm": 0.4604648947715759,
+      "learning_rate": 0.0006973346260637583,
+      "loss": 2.9012,
+      "step": 3045
+    },
+    {
+      "epoch": 0.389888,
+      "grad_norm": 0.44598525762557983,
+      "learning_rate": 0.0006971441268318099,
+      "loss": 2.8057,
+      "step": 3046
+    },
+    {
+      "epoch": 0.390016,
+      "grad_norm": 0.4899541437625885,
+      "learning_rate": 0.0006969535937085005,
+      "loss": 2.9314,
+      "step": 3047
+    },
+    {
+      "epoch": 0.390144,
+      "grad_norm": 0.48988208174705505,
+      "learning_rate": 0.0006967630267265853,
+      "loss": 3.0131,
+      "step": 3048
+    },
+    {
+      "epoch": 0.390272,
+      "grad_norm": 0.4632672369480133,
+      "learning_rate": 0.0006965724259188246,
+      "loss": 2.8721,
+      "step": 3049
+    },
+    {
+      "epoch": 0.3904,
+      "grad_norm": 0.4209100306034088,
+      "learning_rate": 0.0006963817913179849,
+      "loss": 2.8141,
+      "step": 3050
+    },
+    {
+      "epoch": 0.390528,
+      "grad_norm": 0.4316747784614563,
+      "learning_rate": 0.0006961911229568387,
+      "loss": 3.0179,
+      "step": 3051
+    },
+    {
+      "epoch": 0.390656,
+      "grad_norm": 0.4249226152896881,
+      "learning_rate": 0.0006960004208681642,
+      "loss": 2.948,
+      "step": 3052
+    },
+    {
+      "epoch": 0.390784,
+      "grad_norm": 0.4469341039657593,
+      "learning_rate": 0.000695809685084745,
+      "loss": 2.8808,
+      "step": 3053
+    },
+    {
+      "epoch": 0.390912,
+      "grad_norm": 0.4261433780193329,
+      "learning_rate": 0.0006956189156393708,
+      "loss": 2.7561,
+      "step": 3054
+    },
+    {
+      "epoch": 0.39104,
+      "grad_norm": 0.4532381594181061,
+      "learning_rate": 0.0006954281125648373,
+      "loss": 2.9419,
+      "step": 3055
+    },
+    {
+      "epoch": 0.391168,
+      "grad_norm": 0.42453375458717346,
+      "learning_rate": 0.0006952372758939457,
+      "loss": 2.9029,
+      "step": 3056
+    },
+    {
+      "epoch": 0.391296,
+      "grad_norm": 0.4536336660385132,
+      "learning_rate": 0.0006950464056595029,
+      "loss": 2.8829,
+      "step": 3057
+    },
+    {
+      "epoch": 0.391424,
+      "grad_norm": 0.49446940422058105,
+      "learning_rate": 0.0006948555018943219,
+      "loss": 2.9518,
+      "step": 3058
+    },
+    {
+      "epoch": 0.391552,
+      "grad_norm": 0.4457966089248657,
+      "learning_rate": 0.0006946645646312211,
+      "loss": 2.7222,
+      "step": 3059
+    },
+    {
+      "epoch": 0.39168,
+      "grad_norm": 0.43920865654945374,
+      "learning_rate": 0.0006944735939030249,
+      "loss": 2.9819,
+      "step": 3060
+    },
+    {
+      "epoch": 0.391808,
+      "grad_norm": 0.4627797305583954,
+      "learning_rate": 0.0006942825897425633,
+      "loss": 2.9357,
+      "step": 3061
+    },
+    {
+      "epoch": 0.391936,
+      "grad_norm": 0.4886755049228668,
+      "learning_rate": 0.0006940915521826723,
+      "loss": 2.9256,
+      "step": 3062
+    },
+    {
+      "epoch": 0.392064,
+      "grad_norm": 0.46377137303352356,
+      "learning_rate": 0.0006939004812561933,
+      "loss": 2.9669,
+      "step": 3063
+    },
+    {
+      "epoch": 0.392192,
+      "grad_norm": 0.4547809064388275,
+      "learning_rate": 0.0006937093769959736,
+      "loss": 2.8429,
+      "step": 3064
+    },
+    {
+      "epoch": 0.39232,
+      "grad_norm": 0.42778074741363525,
+      "learning_rate": 0.0006935182394348664,
+      "loss": 2.7876,
+      "step": 3065
+    },
+    {
+      "epoch": 0.392448,
+      "grad_norm": 0.49064698815345764,
+      "learning_rate": 0.0006933270686057307,
+      "loss": 2.8154,
+      "step": 3066
+    },
+    {
+      "epoch": 0.392576,
+      "grad_norm": 0.46815866231918335,
+      "learning_rate": 0.0006931358645414304,
+      "loss": 3.0021,
+      "step": 3067
+    },
+    {
+      "epoch": 0.392704,
+      "grad_norm": 0.4475027620792389,
+      "learning_rate": 0.0006929446272748359,
+      "loss": 2.8174,
+      "step": 3068
+    },
+    {
+      "epoch": 0.392832,
+      "grad_norm": 0.4863116145133972,
+      "learning_rate": 0.0006927533568388231,
+      "loss": 3.0152,
+      "step": 3069
+    },
+    {
+      "epoch": 0.39296,
+      "grad_norm": 0.45754820108413696,
+      "learning_rate": 0.000692562053266274,
+      "loss": 2.8841,
+      "step": 3070
+    },
+    {
+      "epoch": 0.393088,
+      "grad_norm": 0.434680700302124,
+      "learning_rate": 0.0006923707165900754,
+      "loss": 2.7925,
+      "step": 3071
+    },
+    {
+      "epoch": 0.393216,
+      "grad_norm": 0.441206693649292,
+      "learning_rate": 0.0006921793468431203,
+      "loss": 2.8341,
+      "step": 3072
+    },
+    {
+      "epoch": 0.393344,
+      "grad_norm": 0.4100573658943176,
+      "learning_rate": 0.0006919879440583077,
+      "loss": 2.7867,
+      "step": 3073
+    },
+    {
+      "epoch": 0.393472,
+      "grad_norm": 0.43837717175483704,
+      "learning_rate": 0.0006917965082685418,
+      "loss": 2.8365,
+      "step": 3074
+    },
+    {
+      "epoch": 0.3936,
+      "grad_norm": 0.4473802149295807,
+      "learning_rate": 0.0006916050395067327,
+      "loss": 2.8859,
+      "step": 3075
+    },
+    {
+      "epoch": 0.393728,
+      "grad_norm": 0.41160866618156433,
+      "learning_rate": 0.0006914135378057959,
+      "loss": 2.8675,
+      "step": 3076
+    },
+    {
+      "epoch": 0.393856,
+      "grad_norm": 0.45181119441986084,
+      "learning_rate": 0.0006912220031986531,
+      "loss": 2.96,
+      "step": 3077
+    },
+    {
+      "epoch": 0.393984,
+      "grad_norm": 0.46188244223594666,
+      "learning_rate": 0.0006910304357182307,
+      "loss": 2.9294,
+      "step": 3078
+    },
+    {
+      "epoch": 0.394112,
+      "grad_norm": 0.45550060272216797,
+      "learning_rate": 0.0006908388353974622,
+      "loss": 2.8136,
+      "step": 3079
+    },
+    {
+      "epoch": 0.39424,
+      "grad_norm": 0.43088603019714355,
+      "learning_rate": 0.0006906472022692854,
+      "loss": 2.8241,
+      "step": 3080
+    },
+    {
+      "epoch": 0.394368,
+      "grad_norm": 0.44399771094322205,
+      "learning_rate": 0.0006904555363666442,
+      "loss": 2.8693,
+      "step": 3081
+    },
+    {
+      "epoch": 0.394496,
+      "grad_norm": 0.4570031762123108,
+      "learning_rate": 0.0006902638377224886,
+      "loss": 3.0306,
+      "step": 3082
+    },
+    {
+      "epoch": 0.394624,
+      "grad_norm": 0.4605392515659332,
+      "learning_rate": 0.0006900721063697732,
+      "loss": 2.9704,
+      "step": 3083
+    },
+    {
+      "epoch": 0.394752,
+      "grad_norm": 0.4443973898887634,
+      "learning_rate": 0.0006898803423414595,
+      "loss": 2.8336,
+      "step": 3084
+    },
+    {
+      "epoch": 0.39488,
+      "grad_norm": 0.4551134705543518,
+      "learning_rate": 0.0006896885456705137,
+      "loss": 2.8905,
+      "step": 3085
+    },
+    {
+      "epoch": 0.395008,
+      "grad_norm": 0.4606602191925049,
+      "learning_rate": 0.0006894967163899076,
+      "loss": 2.9383,
+      "step": 3086
+    },
+    {
+      "epoch": 0.395136,
+      "grad_norm": 0.46792832016944885,
+      "learning_rate": 0.0006893048545326192,
+      "loss": 2.9398,
+      "step": 3087
+    },
+    {
+      "epoch": 0.395264,
+      "grad_norm": 0.6507146954536438,
+      "learning_rate": 0.0006891129601316318,
+      "loss": 2.9688,
+      "step": 3088
+    },
+    {
+      "epoch": 0.395392,
+      "grad_norm": 0.49388590455055237,
+      "learning_rate": 0.0006889210332199343,
+      "loss": 3.0132,
+      "step": 3089
+    },
+    {
+      "epoch": 0.39552,
+      "grad_norm": 0.4477824866771698,
+      "learning_rate": 0.0006887290738305207,
+      "loss": 2.9668,
+      "step": 3090
+    },
+    {
+      "epoch": 0.395648,
+      "grad_norm": 0.46280917525291443,
+      "learning_rate": 0.0006885370819963917,
+      "loss": 2.8407,
+      "step": 3091
+    },
+    {
+      "epoch": 0.395776,
+      "grad_norm": 0.43270495533943176,
+      "learning_rate": 0.0006883450577505523,
+      "loss": 2.8503,
+      "step": 3092
+    },
+    {
+      "epoch": 0.395904,
+      "grad_norm": 0.4325033724308014,
+      "learning_rate": 0.0006881530011260142,
+      "loss": 2.9756,
+      "step": 3093
+    },
+    {
+      "epoch": 0.396032,
+      "grad_norm": 0.4527774453163147,
+      "learning_rate": 0.0006879609121557937,
+      "loss": 2.8278,
+      "step": 3094
+    },
+    {
+      "epoch": 0.39616,
+      "grad_norm": 0.4487600326538086,
+      "learning_rate": 0.0006877687908729137,
+      "loss": 2.9921,
+      "step": 3095
+    },
+    {
+      "epoch": 0.396288,
+      "grad_norm": 0.46167483925819397,
+      "learning_rate": 0.0006875766373104016,
+      "loss": 2.9343,
+      "step": 3096
+    },
+    {
+      "epoch": 0.396416,
+      "grad_norm": 0.41477397084236145,
+      "learning_rate": 0.0006873844515012908,
+      "loss": 2.8016,
+      "step": 3097
+    },
+    {
+      "epoch": 0.396544,
+      "grad_norm": 0.4322338104248047,
+      "learning_rate": 0.0006871922334786205,
+      "loss": 2.8385,
+      "step": 3098
+    },
+    {
+      "epoch": 0.396672,
+      "grad_norm": 0.43136197328567505,
+      "learning_rate": 0.0006869999832754355,
+      "loss": 2.9732,
+      "step": 3099
+    },
+    {
+      "epoch": 0.3968,
+      "grad_norm": 0.4648962914943695,
+      "learning_rate": 0.0006868077009247852,
+      "loss": 2.9195,
+      "step": 3100
+    },
+    {
+      "epoch": 0.396928,
+      "grad_norm": 0.4212260842323303,
+      "learning_rate": 0.0006866153864597253,
+      "loss": 2.9234,
+      "step": 3101
+    },
+    {
+      "epoch": 0.397056,
+      "grad_norm": 0.46229761838912964,
+      "learning_rate": 0.0006864230399133173,
+      "loss": 2.8914,
+      "step": 3102
+    },
+    {
+      "epoch": 0.397184,
+      "grad_norm": 0.4174051880836487,
+      "learning_rate": 0.0006862306613186275,
+      "loss": 2.967,
+      "step": 3103
+    },
+    {
+      "epoch": 0.397312,
+      "grad_norm": 0.4427741467952728,
+      "learning_rate": 0.0006860382507087279,
+      "loss": 2.936,
+      "step": 3104
+    },
+    {
+      "epoch": 0.39744,
+      "grad_norm": 0.5247324705123901,
+      "learning_rate": 0.0006858458081166964,
+      "loss": 2.9741,
+      "step": 3105
+    },
+    {
+      "epoch": 0.397568,
+      "grad_norm": 0.44159960746765137,
+      "learning_rate": 0.0006856533335756158,
+      "loss": 2.8597,
+      "step": 3106
+    },
+    {
+      "epoch": 0.397696,
+      "grad_norm": 0.5307297110557556,
+      "learning_rate": 0.0006854608271185752,
+      "loss": 2.8743,
+      "step": 3107
+    },
+    {
+      "epoch": 0.397824,
+      "grad_norm": 0.45344027876853943,
+      "learning_rate": 0.0006852682887786681,
+      "loss": 2.9495,
+      "step": 3108
+    },
+    {
+      "epoch": 0.397952,
+      "grad_norm": 0.48475703597068787,
+      "learning_rate": 0.0006850757185889945,
+      "loss": 3.0297,
+      "step": 3109
+    },
+    {
+      "epoch": 0.39808,
+      "grad_norm": 0.45889636874198914,
+      "learning_rate": 0.0006848831165826591,
+      "loss": 2.9952,
+      "step": 3110
+    },
+    {
+      "epoch": 0.398208,
+      "grad_norm": 0.42446252703666687,
+      "learning_rate": 0.0006846904827927728,
+      "loss": 2.8111,
+      "step": 3111
+    },
+    {
+      "epoch": 0.398336,
+      "grad_norm": 0.4360777735710144,
+      "learning_rate": 0.0006844978172524513,
+      "loss": 2.994,
+      "step": 3112
+    },
+    {
+      "epoch": 0.398464,
+      "grad_norm": 0.4618556499481201,
+      "learning_rate": 0.0006843051199948161,
+      "loss": 2.8947,
+      "step": 3113
+    },
+    {
+      "epoch": 0.398592,
+      "grad_norm": 0.45799195766448975,
+      "learning_rate": 0.0006841123910529943,
+      "loss": 3.0628,
+      "step": 3114
+    },
+    {
+      "epoch": 0.39872,
+      "grad_norm": 0.4510129988193512,
+      "learning_rate": 0.0006839196304601179,
+      "loss": 2.853,
+      "step": 3115
+    },
+    {
+      "epoch": 0.398848,
+      "grad_norm": 0.43226158618927,
+      "learning_rate": 0.0006837268382493248,
+      "loss": 2.9425,
+      "step": 3116
+    },
+    {
+      "epoch": 0.398976,
+      "grad_norm": 0.5076228976249695,
+      "learning_rate": 0.0006835340144537584,
+      "loss": 2.9615,
+      "step": 3117
+    },
+    {
+      "epoch": 0.399104,
+      "grad_norm": 0.43536415696144104,
+      "learning_rate": 0.0006833411591065673,
+      "loss": 3.0051,
+      "step": 3118
+    },
+    {
+      "epoch": 0.399232,
+      "grad_norm": 0.4361191391944885,
+      "learning_rate": 0.0006831482722409052,
+      "loss": 2.9061,
+      "step": 3119
+    },
+    {
+      "epoch": 0.39936,
+      "grad_norm": 0.4201210141181946,
+      "learning_rate": 0.0006829553538899321,
+      "loss": 2.9083,
+      "step": 3120
+    },
+    {
+      "epoch": 0.399488,
+      "grad_norm": 0.467176616191864,
+      "learning_rate": 0.0006827624040868125,
+      "loss": 2.9283,
+      "step": 3121
+    },
+    {
+      "epoch": 0.399616,
+      "grad_norm": 0.4064013659954071,
+      "learning_rate": 0.0006825694228647169,
+      "loss": 3.0528,
+      "step": 3122
+    },
+    {
+      "epoch": 0.399744,
+      "grad_norm": 0.42155101895332336,
+      "learning_rate": 0.0006823764102568211,
+      "loss": 2.9176,
+      "step": 3123
+    },
+    {
+      "epoch": 0.399872,
+      "grad_norm": 0.45396703481674194,
+      "learning_rate": 0.000682183366296306,
+      "loss": 2.924,
+      "step": 3124
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.47462981939315796,
+      "learning_rate": 0.0006819902910163581,
+      "loss": 2.8783,
+      "step": 3125
+    },
+    {
+      "epoch": 0.400128,
+      "grad_norm": 0.44676801562309265,
+      "learning_rate": 0.0006817971844501695,
+      "loss": 2.9712,
+      "step": 3126
+    },
+    {
+      "epoch": 0.400256,
+      "grad_norm": 0.41748175024986267,
+      "learning_rate": 0.0006816040466309372,
+      "loss": 2.835,
+      "step": 3127
+    },
+    {
+      "epoch": 0.400384,
+      "grad_norm": 0.45010074973106384,
+      "learning_rate": 0.0006814108775918641,
+      "loss": 2.7805,
+      "step": 3128
+    },
+    {
+      "epoch": 0.400512,
+      "grad_norm": 0.4543108642101288,
+      "learning_rate": 0.0006812176773661579,
+      "loss": 2.8162,
+      "step": 3129
+    },
+    {
+      "epoch": 0.40064,
+      "grad_norm": 0.44181913137435913,
+      "learning_rate": 0.0006810244459870322,
+      "loss": 2.8689,
+      "step": 3130
+    },
+    {
+      "epoch": 0.400768,
+      "grad_norm": 0.4721290171146393,
+      "learning_rate": 0.0006808311834877056,
+      "loss": 2.9394,
+      "step": 3131
+    },
+    {
+      "epoch": 0.400896,
+      "grad_norm": 0.3883224427700043,
+      "learning_rate": 0.0006806378899014023,
+      "loss": 2.8384,
+      "step": 3132
+    },
+    {
+      "epoch": 0.401024,
+      "grad_norm": 0.450553297996521,
+      "learning_rate": 0.0006804445652613514,
+      "loss": 2.9025,
+      "step": 3133
+    },
+    {
+      "epoch": 0.401152,
+      "grad_norm": 0.5066479444503784,
+      "learning_rate": 0.0006802512096007878,
+      "loss": 2.8603,
+      "step": 3134
+    },
+    {
+      "epoch": 0.40128,
+      "grad_norm": 0.44616761803627014,
+      "learning_rate": 0.0006800578229529518,
+      "loss": 2.9724,
+      "step": 3135
+    },
+    {
+      "epoch": 0.401408,
+      "grad_norm": 0.41665828227996826,
+      "learning_rate": 0.0006798644053510886,
+      "loss": 2.9569,
+      "step": 3136
+    },
+    {
+      "epoch": 0.401536,
+      "grad_norm": 0.4495328962802887,
+      "learning_rate": 0.0006796709568284488,
+      "loss": 2.9635,
+      "step": 3137
+    },
+    {
+      "epoch": 0.401664,
+      "grad_norm": 0.46058014035224915,
+      "learning_rate": 0.0006794774774182887,
+      "loss": 2.926,
+      "step": 3138
+    },
+    {
+      "epoch": 0.401792,
+      "grad_norm": 0.4509997069835663,
+      "learning_rate": 0.0006792839671538695,
+      "loss": 2.9636,
+      "step": 3139
+    },
+    {
+      "epoch": 0.40192,
+      "grad_norm": 0.43146735429763794,
+      "learning_rate": 0.0006790904260684581,
+      "loss": 2.8551,
+      "step": 3140
+    },
+    {
+      "epoch": 0.402048,
+      "grad_norm": 0.45595669746398926,
+      "learning_rate": 0.0006788968541953261,
+      "loss": 2.9463,
+      "step": 3141
+    },
+    {
+      "epoch": 0.402176,
+      "grad_norm": 0.458778440952301,
+      "learning_rate": 0.0006787032515677509,
+      "loss": 2.9169,
+      "step": 3142
+    },
+    {
+      "epoch": 0.402304,
+      "grad_norm": 0.4239840507507324,
+      "learning_rate": 0.0006785096182190151,
+      "loss": 2.9831,
+      "step": 3143
+    },
+    {
+      "epoch": 0.402432,
+      "grad_norm": 0.4316083490848541,
+      "learning_rate": 0.0006783159541824065,
+      "loss": 2.9958,
+      "step": 3144
+    },
+    {
+      "epoch": 0.40256,
+      "grad_norm": 0.5986521244049072,
+      "learning_rate": 0.0006781222594912181,
+      "loss": 2.8865,
+      "step": 3145
+    },
+    {
+      "epoch": 0.402688,
+      "grad_norm": 0.460986465215683,
+      "learning_rate": 0.0006779285341787484,
+      "loss": 2.8755,
+      "step": 3146
+    },
+    {
+      "epoch": 0.402816,
+      "grad_norm": 0.9214583039283752,
+      "learning_rate": 0.0006777347782783008,
+      "loss": 2.694,
+      "step": 3147
+    },
+    {
+      "epoch": 0.402944,
+      "grad_norm": 0.4762403070926666,
+      "learning_rate": 0.0006775409918231843,
+      "loss": 2.8908,
+      "step": 3148
+    },
+    {
+      "epoch": 0.403072,
+      "grad_norm": 0.6089789271354675,
+      "learning_rate": 0.0006773471748467133,
+      "loss": 2.8767,
+      "step": 3149
+    },
+    {
+      "epoch": 0.4032,
+      "grad_norm": 0.43882882595062256,
+      "learning_rate": 0.000677153327382207,
+      "loss": 2.8028,
+      "step": 3150
+    },
+    {
+      "epoch": 0.403328,
+      "grad_norm": 0.4701058864593506,
+      "learning_rate": 0.0006769594494629897,
+      "loss": 2.9016,
+      "step": 3151
+    },
+    {
+      "epoch": 0.403456,
+      "grad_norm": 0.4339085519313812,
+      "learning_rate": 0.0006767655411223916,
+      "loss": 2.7637,
+      "step": 3152
+    },
+    {
+      "epoch": 0.403584,
+      "grad_norm": 0.5074201226234436,
+      "learning_rate": 0.0006765716023937479,
+      "loss": 2.9184,
+      "step": 3153
+    },
+    {
+      "epoch": 0.403712,
+      "grad_norm": 0.4503526985645294,
+      "learning_rate": 0.0006763776333103988,
+      "loss": 2.9477,
+      "step": 3154
+    },
+    {
+      "epoch": 0.40384,
+      "grad_norm": 0.4295203387737274,
+      "learning_rate": 0.0006761836339056896,
+      "loss": 2.8424,
+      "step": 3155
+    },
+    {
+      "epoch": 0.403968,
+      "grad_norm": 0.43834933638572693,
+      "learning_rate": 0.0006759896042129713,
+      "loss": 2.8019,
+      "step": 3156
+    },
+    {
+      "epoch": 0.404096,
+      "grad_norm": 0.4396340847015381,
+      "learning_rate": 0.0006757955442655998,
+      "loss": 2.8724,
+      "step": 3157
+    },
+    {
+      "epoch": 0.404224,
+      "grad_norm": 0.4488854706287384,
+      "learning_rate": 0.0006756014540969362,
+      "loss": 3.0081,
+      "step": 3158
+    },
+    {
+      "epoch": 0.404352,
+      "grad_norm": 0.464618057012558,
+      "learning_rate": 0.0006754073337403469,
+      "loss": 2.9383,
+      "step": 3159
+    },
+    {
+      "epoch": 0.40448,
+      "grad_norm": 0.4500836730003357,
+      "learning_rate": 0.0006752131832292035,
+      "loss": 2.8971,
+      "step": 3160
+    },
+    {
+      "epoch": 0.404608,
+      "grad_norm": 0.48572155833244324,
+      "learning_rate": 0.0006750190025968827,
+      "loss": 2.8488,
+      "step": 3161
+    },
+    {
+      "epoch": 0.404736,
+      "grad_norm": 0.45179542899131775,
+      "learning_rate": 0.0006748247918767662,
+      "loss": 2.8628,
+      "step": 3162
+    },
+    {
+      "epoch": 0.404864,
+      "grad_norm": 0.4458301067352295,
+      "learning_rate": 0.0006746305511022414,
+      "loss": 2.971,
+      "step": 3163
+    },
+    {
+      "epoch": 0.404992,
+      "grad_norm": 0.4688001871109009,
+      "learning_rate": 0.0006744362803067002,
+      "loss": 2.8846,
+      "step": 3164
+    },
+    {
+      "epoch": 0.40512,
+      "grad_norm": 0.4576423764228821,
+      "learning_rate": 0.0006742419795235403,
+      "loss": 2.9799,
+      "step": 3165
+    },
+    {
+      "epoch": 0.405248,
+      "grad_norm": 0.49632343649864197,
+      "learning_rate": 0.000674047648786164,
+      "loss": 2.8586,
+      "step": 3166
+    },
+    {
+      "epoch": 0.405376,
+      "grad_norm": 0.44714346528053284,
+      "learning_rate": 0.0006738532881279793,
+      "loss": 2.9157,
+      "step": 3167
+    },
+    {
+      "epoch": 0.405504,
+      "grad_norm": 0.44664880633354187,
+      "learning_rate": 0.000673658897582399,
+      "loss": 2.8645,
+      "step": 3168
+    },
+    {
+      "epoch": 0.405632,
+      "grad_norm": 0.45871517062187195,
+      "learning_rate": 0.0006734644771828407,
+      "loss": 2.844,
+      "step": 3169
+    },
+    {
+      "epoch": 0.40576,
+      "grad_norm": 0.46869131922721863,
+      "learning_rate": 0.0006732700269627279,
+      "loss": 2.9226,
+      "step": 3170
+    },
+    {
+      "epoch": 0.405888,
+      "grad_norm": 0.4519689977169037,
+      "learning_rate": 0.0006730755469554888,
+      "loss": 2.8639,
+      "step": 3171
+    },
+    {
+      "epoch": 0.406016,
+      "grad_norm": 0.5095706582069397,
+      "learning_rate": 0.0006728810371945566,
+      "loss": 2.9763,
+      "step": 3172
+    },
+    {
+      "epoch": 0.406144,
+      "grad_norm": 0.4178374409675598,
+      "learning_rate": 0.00067268649771337,
+      "loss": 2.9025,
+      "step": 3173
+    },
+    {
+      "epoch": 0.406272,
+      "grad_norm": 0.4735436737537384,
+      "learning_rate": 0.0006724919285453725,
+      "loss": 2.929,
+      "step": 3174
+    },
+    {
+      "epoch": 0.4064,
+      "grad_norm": 0.4643951654434204,
+      "learning_rate": 0.0006722973297240127,
+      "loss": 2.8802,
+      "step": 3175
+    },
+    {
+      "epoch": 0.406528,
+      "grad_norm": 0.45067164301872253,
+      "learning_rate": 0.0006721027012827444,
+      "loss": 2.8043,
+      "step": 3176
+    },
+    {
+      "epoch": 0.406656,
+      "grad_norm": 0.4566052556037903,
+      "learning_rate": 0.0006719080432550269,
+      "loss": 2.8485,
+      "step": 3177
+    },
+    {
+      "epoch": 0.406784,
+      "grad_norm": 0.46093007922172546,
+      "learning_rate": 0.0006717133556743235,
+      "loss": 2.9712,
+      "step": 3178
+    },
+    {
+      "epoch": 0.406912,
+      "grad_norm": 0.4791119694709778,
+      "learning_rate": 0.0006715186385741037,
+      "loss": 2.9684,
+      "step": 3179
+    },
+    {
+      "epoch": 0.40704,
+      "grad_norm": 0.4475882351398468,
+      "learning_rate": 0.0006713238919878415,
+      "loss": 2.8393,
+      "step": 3180
+    },
+    {
+      "epoch": 0.407168,
+      "grad_norm": 0.4540812075138092,
+      "learning_rate": 0.0006711291159490161,
+      "loss": 2.936,
+      "step": 3181
+    },
+    {
+      "epoch": 0.407296,
+      "grad_norm": 0.4281275272369385,
+      "learning_rate": 0.0006709343104911119,
+      "loss": 2.7903,
+      "step": 3182
+    },
+    {
+      "epoch": 0.407424,
+      "grad_norm": 0.4153285324573517,
+      "learning_rate": 0.0006707394756476179,
+      "loss": 2.8059,
+      "step": 3183
+    },
+    {
+      "epoch": 0.407552,
+      "grad_norm": 0.4291643500328064,
+      "learning_rate": 0.0006705446114520288,
+      "loss": 2.9037,
+      "step": 3184
+    },
+    {
+      "epoch": 0.40768,
+      "grad_norm": 0.4371093213558197,
+      "learning_rate": 0.0006703497179378438,
+      "loss": 2.8227,
+      "step": 3185
+    },
+    {
+      "epoch": 0.407808,
+      "grad_norm": 0.41008368134498596,
+      "learning_rate": 0.0006701547951385675,
+      "loss": 2.8242,
+      "step": 3186
+    },
+    {
+      "epoch": 0.407936,
+      "grad_norm": 0.45269104838371277,
+      "learning_rate": 0.0006699598430877092,
+      "loss": 2.9421,
+      "step": 3187
+    },
+    {
+      "epoch": 0.408064,
+      "grad_norm": 0.45688512921333313,
+      "learning_rate": 0.0006697648618187836,
+      "loss": 2.8332,
+      "step": 3188
+    },
+    {
+      "epoch": 0.408192,
+      "grad_norm": 0.4551680386066437,
+      "learning_rate": 0.0006695698513653101,
+      "loss": 2.9544,
+      "step": 3189
+    },
+    {
+      "epoch": 0.40832,
+      "grad_norm": 0.5006901025772095,
+      "learning_rate": 0.0006693748117608133,
+      "loss": 2.9123,
+      "step": 3190
+    },
+    {
+      "epoch": 0.408448,
+      "grad_norm": 0.4649401605129242,
+      "learning_rate": 0.000669179743038823,
+      "loss": 2.8462,
+      "step": 3191
+    },
+    {
+      "epoch": 0.408576,
+      "grad_norm": 0.4609390199184418,
+      "learning_rate": 0.0006689846452328735,
+      "loss": 2.8033,
+      "step": 3192
+    },
+    {
+      "epoch": 0.408704,
+      "grad_norm": 0.4476522207260132,
+      "learning_rate": 0.0006687895183765043,
+      "loss": 2.768,
+      "step": 3193
+    },
+    {
+      "epoch": 0.408832,
+      "grad_norm": 0.43808048963546753,
+      "learning_rate": 0.0006685943625032603,
+      "loss": 2.9704,
+      "step": 3194
+    },
+    {
+      "epoch": 0.40896,
+      "grad_norm": 0.43268802762031555,
+      "learning_rate": 0.0006683991776466906,
+      "loss": 2.9108,
+      "step": 3195
+    },
+    {
+      "epoch": 0.409088,
+      "grad_norm": 0.4473373591899872,
+      "learning_rate": 0.0006682039638403503,
+      "loss": 2.8289,
+      "step": 3196
+    },
+    {
+      "epoch": 0.409216,
+      "grad_norm": 0.421802818775177,
+      "learning_rate": 0.0006680087211177985,
+      "loss": 2.8509,
+      "step": 3197
+    },
+    {
+      "epoch": 0.409344,
+      "grad_norm": 0.4535096287727356,
+      "learning_rate": 0.0006678134495125998,
+      "loss": 2.9703,
+      "step": 3198
+    },
+    {
+      "epoch": 0.409472,
+      "grad_norm": 0.422649085521698,
+      "learning_rate": 0.0006676181490583238,
+      "loss": 2.7602,
+      "step": 3199
+    },
+    {
+      "epoch": 0.4096,
+      "grad_norm": 0.45454171299934387,
+      "learning_rate": 0.0006674228197885446,
+      "loss": 2.8892,
+      "step": 3200
+    },
+    {
+      "epoch": 0.409728,
+      "grad_norm": 0.4481218755245209,
+      "learning_rate": 0.0006672274617368418,
+      "loss": 2.7488,
+      "step": 3201
+    },
+    {
+      "epoch": 0.409856,
+      "grad_norm": 0.48971325159072876,
+      "learning_rate": 0.0006670320749367997,
+      "loss": 2.8091,
+      "step": 3202
+    },
+    {
+      "epoch": 0.409984,
+      "grad_norm": 0.4822114408016205,
+      "learning_rate": 0.0006668366594220075,
+      "loss": 2.8929,
+      "step": 3203
+    },
+    {
+      "epoch": 0.410112,
+      "grad_norm": 0.4181433618068695,
+      "learning_rate": 0.0006666412152260595,
+      "loss": 2.9551,
+      "step": 3204
+    },
+    {
+      "epoch": 0.41024,
+      "grad_norm": 0.4663460850715637,
+      "learning_rate": 0.0006664457423825546,
+      "loss": 3.0142,
+      "step": 3205
+    },
+    {
+      "epoch": 0.410368,
+      "grad_norm": 0.437890887260437,
+      "learning_rate": 0.000666250240925097,
+      "loss": 2.8748,
+      "step": 3206
+    },
+    {
+      "epoch": 0.410496,
+      "grad_norm": 0.42608585953712463,
+      "learning_rate": 0.0006660547108872959,
+      "loss": 2.8459,
+      "step": 3207
+    },
+    {
+      "epoch": 0.410624,
+      "grad_norm": 0.46166375279426575,
+      "learning_rate": 0.0006658591523027643,
+      "loss": 2.9711,
+      "step": 3208
+    },
+    {
+      "epoch": 0.410752,
+      "grad_norm": 0.4234524071216583,
+      "learning_rate": 0.0006656635652051222,
+      "loss": 2.8403,
+      "step": 3209
+    },
+    {
+      "epoch": 0.41088,
+      "grad_norm": 0.44437092542648315,
+      "learning_rate": 0.0006654679496279926,
+      "loss": 2.9396,
+      "step": 3210
+    },
+    {
+      "epoch": 0.411008,
+      "grad_norm": 0.4451534152030945,
+      "learning_rate": 0.000665272305605004,
+      "loss": 2.8833,
+      "step": 3211
+    },
+    {
+      "epoch": 0.411136,
+      "grad_norm": 0.44552239775657654,
+      "learning_rate": 0.0006650766331697902,
+      "loss": 2.9453,
+      "step": 3212
+    },
+    {
+      "epoch": 0.411264,
+      "grad_norm": 0.44758284091949463,
+      "learning_rate": 0.0006648809323559895,
+      "loss": 2.7753,
+      "step": 3213
+    },
+    {
+      "epoch": 0.411392,
+      "grad_norm": 0.44451257586479187,
+      "learning_rate": 0.0006646852031972449,
+      "loss": 2.8592,
+      "step": 3214
+    },
+    {
+      "epoch": 0.41152,
+      "grad_norm": 0.4396018385887146,
+      "learning_rate": 0.0006644894457272047,
+      "loss": 2.8164,
+      "step": 3215
+    },
+    {
+      "epoch": 0.411648,
+      "grad_norm": 0.4517490565776825,
+      "learning_rate": 0.0006642936599795219,
+      "loss": 2.9472,
+      "step": 3216
+    },
+    {
+      "epoch": 0.411776,
+      "grad_norm": 0.4633621275424957,
+      "learning_rate": 0.0006640978459878543,
+      "loss": 2.8428,
+      "step": 3217
+    },
+    {
+      "epoch": 0.411904,
+      "grad_norm": 0.3884051740169525,
+      "learning_rate": 0.0006639020037858646,
+      "loss": 2.8109,
+      "step": 3218
+    },
+    {
+      "epoch": 0.412032,
+      "grad_norm": 0.4787423610687256,
+      "learning_rate": 0.0006637061334072203,
+      "loss": 2.8661,
+      "step": 3219
+    },
+    {
+      "epoch": 0.41216,
+      "grad_norm": 0.4761379659175873,
+      "learning_rate": 0.0006635102348855939,
+      "loss": 2.8672,
+      "step": 3220
+    },
+    {
+      "epoch": 0.412288,
+      "grad_norm": 0.48324206471443176,
+      "learning_rate": 0.0006633143082546624,
+      "loss": 2.8289,
+      "step": 3221
+    },
+    {
+      "epoch": 0.412416,
+      "grad_norm": 0.3948509395122528,
+      "learning_rate": 0.0006631183535481081,
+      "loss": 2.926,
+      "step": 3222
+    },
+    {
+      "epoch": 0.412544,
+      "grad_norm": 0.4443584978580475,
+      "learning_rate": 0.0006629223707996177,
+      "loss": 2.9169,
+      "step": 3223
+    },
+    {
+      "epoch": 0.412672,
+      "grad_norm": 0.41524747014045715,
+      "learning_rate": 0.0006627263600428832,
+      "loss": 2.9348,
+      "step": 3224
+    },
+    {
+      "epoch": 0.4128,
+      "grad_norm": 0.44109317660331726,
+      "learning_rate": 0.0006625303213116008,
+      "loss": 2.8985,
+      "step": 3225
+    },
+    {
+      "epoch": 0.412928,
+      "grad_norm": 0.4329018294811249,
+      "learning_rate": 0.0006623342546394719,
+      "loss": 2.8074,
+      "step": 3226
+    },
+    {
+      "epoch": 0.413056,
+      "grad_norm": 0.4529988467693329,
+      "learning_rate": 0.0006621381600602028,
+      "loss": 2.9466,
+      "step": 3227
+    },
+    {
+      "epoch": 0.413184,
+      "grad_norm": 0.4321211874485016,
+      "learning_rate": 0.0006619420376075042,
+      "loss": 2.7368,
+      "step": 3228
+    },
+    {
+      "epoch": 0.413312,
+      "grad_norm": 0.4653237760066986,
+      "learning_rate": 0.0006617458873150921,
+      "loss": 3.024,
+      "step": 3229
+    },
+    {
+      "epoch": 0.41344,
+      "grad_norm": 0.45334702730178833,
+      "learning_rate": 0.0006615497092166868,
+      "loss": 3.0949,
+      "step": 3230
+    },
+    {
+      "epoch": 0.413568,
+      "grad_norm": 0.5773935914039612,
+      "learning_rate": 0.0006613535033460135,
+      "loss": 2.8591,
+      "step": 3231
+    },
+    {
+      "epoch": 0.413696,
+      "grad_norm": 0.45192116498947144,
+      "learning_rate": 0.0006611572697368026,
+      "loss": 3.0104,
+      "step": 3232
+    },
+    {
+      "epoch": 0.413824,
+      "grad_norm": 0.437284380197525,
+      "learning_rate": 0.0006609610084227886,
+      "loss": 2.8629,
+      "step": 3233
+    },
+    {
+      "epoch": 0.413952,
+      "grad_norm": 0.42515096068382263,
+      "learning_rate": 0.0006607647194377113,
+      "loss": 2.9291,
+      "step": 3234
+    },
+    {
+      "epoch": 0.41408,
+      "grad_norm": 0.45288413763046265,
+      "learning_rate": 0.0006605684028153149,
+      "loss": 2.8808,
+      "step": 3235
+    },
+    {
+      "epoch": 0.414208,
+      "grad_norm": 0.4062768816947937,
+      "learning_rate": 0.0006603720585893486,
+      "loss": 2.8653,
+      "step": 3236
+    },
+    {
+      "epoch": 0.414336,
+      "grad_norm": 0.42415717244148254,
+      "learning_rate": 0.0006601756867935664,
+      "loss": 2.855,
+      "step": 3237
+    },
+    {
+      "epoch": 0.414464,
+      "grad_norm": 0.4227084219455719,
+      "learning_rate": 0.0006599792874617266,
+      "loss": 2.835,
+      "step": 3238
+    },
+    {
+      "epoch": 0.414592,
+      "grad_norm": 0.5027446746826172,
+      "learning_rate": 0.0006597828606275928,
+      "loss": 2.8703,
+      "step": 3239
+    },
+    {
+      "epoch": 0.41472,
+      "grad_norm": 0.4458296597003937,
+      "learning_rate": 0.0006595864063249326,
+      "loss": 2.8601,
+      "step": 3240
+    },
+    {
+      "epoch": 0.414848,
+      "grad_norm": 0.4928509593009949,
+      "learning_rate": 0.0006593899245875192,
+      "loss": 2.8583,
+      "step": 3241
+    },
+    {
+      "epoch": 0.414976,
+      "grad_norm": 0.4184534251689911,
+      "learning_rate": 0.00065919341544913,
+      "loss": 2.8851,
+      "step": 3242
+    },
+    {
+      "epoch": 0.415104,
+      "grad_norm": 0.44592544436454773,
+      "learning_rate": 0.0006589968789435472,
+      "loss": 2.9156,
+      "step": 3243
+    },
+    {
+      "epoch": 0.415232,
+      "grad_norm": 0.46028947830200195,
+      "learning_rate": 0.0006588003151045576,
+      "loss": 2.8321,
+      "step": 3244
+    },
+    {
+      "epoch": 0.41536,
+      "grad_norm": 0.5221034288406372,
+      "learning_rate": 0.0006586037239659528,
+      "loss": 2.9057,
+      "step": 3245
+    },
+    {
+      "epoch": 0.415488,
+      "grad_norm": 0.4340832531452179,
+      "learning_rate": 0.0006584071055615293,
+      "loss": 2.7941,
+      "step": 3246
+    },
+    {
+      "epoch": 0.415616,
+      "grad_norm": 0.4584273397922516,
+      "learning_rate": 0.0006582104599250878,
+      "loss": 3.0463,
+      "step": 3247
+    },
+    {
+      "epoch": 0.415744,
+      "grad_norm": 0.4859541654586792,
+      "learning_rate": 0.0006580137870904341,
+      "loss": 2.8399,
+      "step": 3248
+    },
+    {
+      "epoch": 0.415872,
+      "grad_norm": 0.43855202198028564,
+      "learning_rate": 0.0006578170870913787,
+      "loss": 2.867,
+      "step": 3249
+    },
+    {
+      "epoch": 0.416,
+      "grad_norm": 0.47458159923553467,
+      "learning_rate": 0.0006576203599617363,
+      "loss": 2.8899,
+      "step": 3250
+    },
+    {
+      "epoch": 0.416128,
+      "grad_norm": 0.511414647102356,
+      "learning_rate": 0.0006574236057353268,
+      "loss": 2.8971,
+      "step": 3251
+    },
+    {
+      "epoch": 0.416256,
+      "grad_norm": 0.4591068923473358,
+      "learning_rate": 0.0006572268244459745,
+      "loss": 2.7356,
+      "step": 3252
+    },
+    {
+      "epoch": 0.416384,
+      "grad_norm": 0.46457529067993164,
+      "learning_rate": 0.0006570300161275084,
+      "loss": 2.9058,
+      "step": 3253
+    },
+    {
+      "epoch": 0.416512,
+      "grad_norm": 0.48261523246765137,
+      "learning_rate": 0.0006568331808137619,
+      "loss": 2.8425,
+      "step": 3254
+    },
+    {
+      "epoch": 0.41664,
+      "grad_norm": 0.43980008363723755,
+      "learning_rate": 0.0006566363185385735,
+      "loss": 2.939,
+      "step": 3255
+    },
+    {
+      "epoch": 0.416768,
+      "grad_norm": 0.4363357722759247,
+      "learning_rate": 0.0006564394293357861,
+      "loss": 2.8251,
+      "step": 3256
+    },
+    {
+      "epoch": 0.416896,
+      "grad_norm": 0.4319329261779785,
+      "learning_rate": 0.0006562425132392472,
+      "loss": 3.062,
+      "step": 3257
+    },
+    {
+      "epoch": 0.417024,
+      "grad_norm": 0.42450985312461853,
+      "learning_rate": 0.0006560455702828089,
+      "loss": 2.9579,
+      "step": 3258
+    },
+    {
+      "epoch": 0.417152,
+      "grad_norm": 0.4666684865951538,
+      "learning_rate": 0.0006558486005003282,
+      "loss": 3.0056,
+      "step": 3259
+    },
+    {
+      "epoch": 0.41728,
+      "grad_norm": 0.4730471670627594,
+      "learning_rate": 0.0006556516039256661,
+      "loss": 2.881,
+      "step": 3260
+    },
+    {
+      "epoch": 0.417408,
+      "grad_norm": 0.4251174330711365,
+      "learning_rate": 0.000655454580592689,
+      "loss": 3.0115,
+      "step": 3261
+    },
+    {
+      "epoch": 0.417536,
+      "grad_norm": 0.42243823409080505,
+      "learning_rate": 0.0006552575305352671,
+      "loss": 2.9357,
+      "step": 3262
+    },
+    {
+      "epoch": 0.417664,
+      "grad_norm": 0.4705365002155304,
+      "learning_rate": 0.0006550604537872759,
+      "loss": 2.7708,
+      "step": 3263
+    },
+    {
+      "epoch": 0.417792,
+      "grad_norm": 0.4361517131328583,
+      "learning_rate": 0.0006548633503825952,
+      "loss": 2.9409,
+      "step": 3264
+    },
+    {
+      "epoch": 0.41792,
+      "grad_norm": 0.4532221853733063,
+      "learning_rate": 0.0006546662203551092,
+      "loss": 2.7335,
+      "step": 3265
+    },
+    {
+      "epoch": 0.418048,
+      "grad_norm": 0.4267335534095764,
+      "learning_rate": 0.0006544690637387068,
+      "loss": 2.9441,
+      "step": 3266
+    },
+    {
+      "epoch": 0.418176,
+      "grad_norm": 0.45892176032066345,
+      "learning_rate": 0.0006542718805672816,
+      "loss": 2.8603,
+      "step": 3267
+    },
+    {
+      "epoch": 0.418304,
+      "grad_norm": 0.4424954056739807,
+      "learning_rate": 0.0006540746708747317,
+      "loss": 2.9791,
+      "step": 3268
+    },
+    {
+      "epoch": 0.418432,
+      "grad_norm": 0.4663112163543701,
+      "learning_rate": 0.0006538774346949597,
+      "loss": 2.7488,
+      "step": 3269
+    },
+    {
+      "epoch": 0.41856,
+      "grad_norm": 0.4785551130771637,
+      "learning_rate": 0.0006536801720618727,
+      "loss": 2.8738,
+      "step": 3270
+    },
+    {
+      "epoch": 0.418688,
+      "grad_norm": 0.43908920884132385,
+      "learning_rate": 0.0006534828830093827,
+      "loss": 2.9527,
+      "step": 3271
+    },
+    {
+      "epoch": 0.418816,
+      "grad_norm": 0.4348186254501343,
+      "learning_rate": 0.0006532855675714054,
+      "loss": 2.7632,
+      "step": 3272
+    },
+    {
+      "epoch": 0.418944,
+      "grad_norm": 0.4681605398654938,
+      "learning_rate": 0.0006530882257818622,
+      "loss": 2.9611,
+      "step": 3273
+    },
+    {
+      "epoch": 0.419072,
+      "grad_norm": 0.44336262345314026,
+      "learning_rate": 0.0006528908576746784,
+      "loss": 2.9961,
+      "step": 3274
+    },
+    {
+      "epoch": 0.4192,
+      "grad_norm": 0.4510493874549866,
+      "learning_rate": 0.0006526934632837835,
+      "loss": 2.9547,
+      "step": 3275
+    },
+    {
+      "epoch": 0.419328,
+      "grad_norm": 0.454348623752594,
+      "learning_rate": 0.0006524960426431122,
+      "loss": 2.7688,
+      "step": 3276
+    },
+    {
+      "epoch": 0.419456,
+      "grad_norm": 0.45813876390457153,
+      "learning_rate": 0.0006522985957866032,
+      "loss": 2.9152,
+      "step": 3277
+    },
+    {
+      "epoch": 0.419584,
+      "grad_norm": 0.4582643210887909,
+      "learning_rate": 0.0006521011227482,
+      "loss": 2.9992,
+      "step": 3278
+    },
+    {
+      "epoch": 0.419712,
+      "grad_norm": 0.4724067747592926,
+      "learning_rate": 0.0006519036235618505,
+      "loss": 2.9706,
+      "step": 3279
+    },
+    {
+      "epoch": 0.41984,
+      "grad_norm": 0.4907130300998688,
+      "learning_rate": 0.0006517060982615071,
+      "loss": 3.0978,
+      "step": 3280
+    },
+    {
+      "epoch": 0.419968,
+      "grad_norm": 0.4406067132949829,
+      "learning_rate": 0.0006515085468811268,
+      "loss": 2.8678,
+      "step": 3281
+    },
+    {
+      "epoch": 0.420096,
+      "grad_norm": 0.40605098009109497,
+      "learning_rate": 0.0006513109694546706,
+      "loss": 2.8067,
+      "step": 3282
+    },
+    {
+      "epoch": 0.420224,
+      "grad_norm": 0.4438405930995941,
+      "learning_rate": 0.0006511133660161047,
+      "loss": 2.9831,
+      "step": 3283
+    },
+    {
+      "epoch": 0.420352,
+      "grad_norm": 0.44131365418434143,
+      "learning_rate": 0.0006509157365993994,
+      "loss": 2.7524,
+      "step": 3284
+    },
+    {
+      "epoch": 0.42048,
+      "grad_norm": 0.4518292248249054,
+      "learning_rate": 0.0006507180812385295,
+      "loss": 2.9054,
+      "step": 3285
+    },
+    {
+      "epoch": 0.420608,
+      "grad_norm": 0.4600540101528168,
+      "learning_rate": 0.000650520399967474,
+      "loss": 2.962,
+      "step": 3286
+    },
+    {
+      "epoch": 0.420736,
+      "grad_norm": 0.4239102900028229,
+      "learning_rate": 0.0006503226928202167,
+      "loss": 2.7564,
+      "step": 3287
+    },
+    {
+      "epoch": 0.420864,
+      "grad_norm": 0.48350197076797485,
+      "learning_rate": 0.000650124959830746,
+      "loss": 2.9025,
+      "step": 3288
+    },
+    {
+      "epoch": 0.420992,
+      "grad_norm": 0.44370517134666443,
+      "learning_rate": 0.0006499272010330543,
+      "loss": 2.9472,
+      "step": 3289
+    },
+    {
+      "epoch": 0.42112,
+      "grad_norm": 0.425523042678833,
+      "learning_rate": 0.0006497294164611385,
+      "loss": 2.8385,
+      "step": 3290
+    },
+    {
+      "epoch": 0.421248,
+      "grad_norm": 0.4094841480255127,
+      "learning_rate": 0.0006495316061490002,
+      "loss": 2.8249,
+      "step": 3291
+    },
+    {
+      "epoch": 0.421376,
+      "grad_norm": 0.45880308747291565,
+      "learning_rate": 0.0006493337701306454,
+      "loss": 3.0206,
+      "step": 3292
+    },
+    {
+      "epoch": 0.421504,
+      "grad_norm": 0.4019167721271515,
+      "learning_rate": 0.0006491359084400843,
+      "loss": 2.9582,
+      "step": 3293
+    },
+    {
+      "epoch": 0.421632,
+      "grad_norm": 0.4317949414253235,
+      "learning_rate": 0.0006489380211113315,
+      "loss": 2.9436,
+      "step": 3294
+    },
+    {
+      "epoch": 0.42176,
+      "grad_norm": 0.4239514470100403,
+      "learning_rate": 0.0006487401081784062,
+      "loss": 2.7979,
+      "step": 3295
+    },
+    {
+      "epoch": 0.421888,
+      "grad_norm": 0.425984263420105,
+      "learning_rate": 0.000648542169675332,
+      "loss": 2.8594,
+      "step": 3296
+    },
+    {
+      "epoch": 0.422016,
+      "grad_norm": 0.42507457733154297,
+      "learning_rate": 0.0006483442056361368,
+      "loss": 2.9026,
+      "step": 3297
+    },
+    {
+      "epoch": 0.422144,
+      "grad_norm": 0.4450777471065521,
+      "learning_rate": 0.000648146216094853,
+      "loss": 2.7602,
+      "step": 3298
+    },
+    {
+      "epoch": 0.422272,
+      "grad_norm": 0.4241431951522827,
+      "learning_rate": 0.0006479482010855172,
+      "loss": 2.8918,
+      "step": 3299
+    },
+    {
+      "epoch": 0.4224,
+      "grad_norm": 0.4700932502746582,
+      "learning_rate": 0.0006477501606421702,
+      "loss": 2.9503,
+      "step": 3300
+    },
+    {
+      "epoch": 0.422528,
+      "grad_norm": 0.46089789271354675,
+      "learning_rate": 0.000647552094798858,
+      "loss": 2.8685,
+      "step": 3301
+    },
+    {
+      "epoch": 0.422656,
+      "grad_norm": 0.42263108491897583,
+      "learning_rate": 0.00064735400358963,
+      "loss": 2.9551,
+      "step": 3302
+    },
+    {
+      "epoch": 0.422784,
+      "grad_norm": 0.4227675199508667,
+      "learning_rate": 0.0006471558870485407,
+      "loss": 2.6979,
+      "step": 3303
+    },
+    {
+      "epoch": 0.422912,
+      "grad_norm": 0.4684280455112457,
+      "learning_rate": 0.0006469577452096482,
+      "loss": 2.8185,
+      "step": 3304
+    },
+    {
+      "epoch": 0.42304,
+      "grad_norm": 0.44374898076057434,
+      "learning_rate": 0.0006467595781070157,
+      "loss": 2.8484,
+      "step": 3305
+    },
+    {
+      "epoch": 0.423168,
+      "grad_norm": 0.4017733633518219,
+      "learning_rate": 0.0006465613857747104,
+      "loss": 2.6637,
+      "step": 3306
+    },
+    {
+      "epoch": 0.423296,
+      "grad_norm": 0.42916741967201233,
+      "learning_rate": 0.0006463631682468037,
+      "loss": 2.8637,
+      "step": 3307
+    },
+    {
+      "epoch": 0.423424,
+      "grad_norm": 0.42470043897628784,
+      "learning_rate": 0.0006461649255573716,
+      "loss": 3.0343,
+      "step": 3308
+    },
+    {
+      "epoch": 0.423552,
+      "grad_norm": 0.4430844187736511,
+      "learning_rate": 0.0006459666577404943,
+      "loss": 2.9449,
+      "step": 3309
+    },
+    {
+      "epoch": 0.42368,
+      "grad_norm": 0.4011468291282654,
+      "learning_rate": 0.0006457683648302565,
+      "loss": 2.8196,
+      "step": 3310
+    },
+    {
+      "epoch": 0.423808,
+      "grad_norm": 0.4105871319770813,
+      "learning_rate": 0.0006455700468607469,
+      "loss": 2.8209,
+      "step": 3311
+    },
+    {
+      "epoch": 0.423936,
+      "grad_norm": 0.42936378717422485,
+      "learning_rate": 0.0006453717038660584,
+      "loss": 2.9539,
+      "step": 3312
+    },
+    {
+      "epoch": 0.424064,
+      "grad_norm": 0.4062698185443878,
+      "learning_rate": 0.0006451733358802888,
+      "loss": 2.8093,
+      "step": 3313
+    },
+    {
+      "epoch": 0.424192,
+      "grad_norm": 0.42204976081848145,
+      "learning_rate": 0.0006449749429375398,
+      "loss": 2.8165,
+      "step": 3314
+    },
+    {
+      "epoch": 0.42432,
+      "grad_norm": 0.44176122546195984,
+      "learning_rate": 0.0006447765250719174,
+      "loss": 2.7645,
+      "step": 3315
+    },
+    {
+      "epoch": 0.424448,
+      "grad_norm": 0.44014376401901245,
+      "learning_rate": 0.0006445780823175321,
+      "loss": 2.8936,
+      "step": 3316
+    },
+    {
+      "epoch": 0.424576,
+      "grad_norm": 0.44232070446014404,
+      "learning_rate": 0.0006443796147084981,
+      "loss": 2.9949,
+      "step": 3317
+    },
+    {
+      "epoch": 0.424704,
+      "grad_norm": 0.480686753988266,
+      "learning_rate": 0.0006441811222789347,
+      "loss": 3.0146,
+      "step": 3318
+    },
+    {
+      "epoch": 0.424832,
+      "grad_norm": 0.4178481101989746,
+      "learning_rate": 0.0006439826050629648,
+      "loss": 2.7405,
+      "step": 3319
+    },
+    {
+      "epoch": 0.42496,
+      "grad_norm": 0.4674568176269531,
+      "learning_rate": 0.0006437840630947161,
+      "loss": 2.816,
+      "step": 3320
+    },
+    {
+      "epoch": 0.425088,
+      "grad_norm": 0.4234991669654846,
+      "learning_rate": 0.0006435854964083198,
+      "loss": 2.7724,
+      "step": 3321
+    },
+    {
+      "epoch": 0.425216,
+      "grad_norm": 0.43948042392730713,
+      "learning_rate": 0.0006433869050379121,
+      "loss": 2.953,
+      "step": 3322
+    },
+    {
+      "epoch": 0.425344,
+      "grad_norm": 0.4700566530227661,
+      "learning_rate": 0.0006431882890176333,
+      "loss": 2.8991,
+      "step": 3323
+    },
+    {
+      "epoch": 0.425472,
+      "grad_norm": 0.41717013716697693,
+      "learning_rate": 0.0006429896483816276,
+      "loss": 3.0275,
+      "step": 3324
+    },
+    {
+      "epoch": 0.4256,
+      "grad_norm": 0.42725467681884766,
+      "learning_rate": 0.0006427909831640437,
+      "loss": 3.0189,
+      "step": 3325
+    },
+    {
+      "epoch": 0.425728,
+      "grad_norm": 0.46575942635536194,
+      "learning_rate": 0.0006425922933990343,
+      "loss": 2.9043,
+      "step": 3326
+    },
+    {
+      "epoch": 0.425856,
+      "grad_norm": 0.4470275342464447,
+      "learning_rate": 0.0006423935791207568,
+      "loss": 3.0643,
+      "step": 3327
+    },
+    {
+      "epoch": 0.425984,
+      "grad_norm": 0.422124981880188,
+      "learning_rate": 0.0006421948403633721,
+      "loss": 2.8299,
+      "step": 3328
+    },
+    {
+      "epoch": 0.426112,
+      "grad_norm": 0.4824879765510559,
+      "learning_rate": 0.000641996077161046,
+      "loss": 2.8648,
+      "step": 3329
+    },
+    {
+      "epoch": 0.42624,
+      "grad_norm": 0.42489567399024963,
+      "learning_rate": 0.000641797289547948,
+      "loss": 2.7664,
+      "step": 3330
+    },
+    {
+      "epoch": 0.426368,
+      "grad_norm": 0.47619807720184326,
+      "learning_rate": 0.0006415984775582521,
+      "loss": 2.923,
+      "step": 3331
+    },
+    {
+      "epoch": 0.426496,
+      "grad_norm": 0.4284994304180145,
+      "learning_rate": 0.0006413996412261363,
+      "loss": 2.95,
+      "step": 3332
+    },
+    {
+      "epoch": 0.426624,
+      "grad_norm": 0.3984729051589966,
+      "learning_rate": 0.0006412007805857828,
+      "loss": 2.7662,
+      "step": 3333
+    },
+    {
+      "epoch": 0.426752,
+      "grad_norm": 0.39650586247444153,
+      "learning_rate": 0.0006410018956713783,
+      "loss": 2.8021,
+      "step": 3334
+    },
+    {
+      "epoch": 0.42688,
+      "grad_norm": 0.45128771662712097,
+      "learning_rate": 0.0006408029865171134,
+      "loss": 2.8081,
+      "step": 3335
+    },
+    {
+      "epoch": 0.427008,
+      "grad_norm": 0.4389777183532715,
+      "learning_rate": 0.0006406040531571828,
+      "loss": 2.9107,
+      "step": 3336
+    },
+    {
+      "epoch": 0.427136,
+      "grad_norm": 0.44102853536605835,
+      "learning_rate": 0.0006404050956257852,
+      "loss": 2.8994,
+      "step": 3337
+    },
+    {
+      "epoch": 0.427264,
+      "grad_norm": 0.41355523467063904,
+      "learning_rate": 0.0006402061139571243,
+      "loss": 2.8671,
+      "step": 3338
+    },
+    {
+      "epoch": 0.427392,
+      "grad_norm": 0.4401109516620636,
+      "learning_rate": 0.0006400071081854068,
+      "loss": 3.0297,
+      "step": 3339
+    },
+    {
+      "epoch": 0.42752,
+      "grad_norm": 0.4350214898586273,
+      "learning_rate": 0.0006398080783448442,
+      "loss": 2.8474,
+      "step": 3340
+    },
+    {
+      "epoch": 0.427648,
+      "grad_norm": 0.4427169859409332,
+      "learning_rate": 0.0006396090244696523,
+      "loss": 2.8063,
+      "step": 3341
+    },
+    {
+      "epoch": 0.427776,
+      "grad_norm": 0.3963504731655121,
+      "learning_rate": 0.0006394099465940505,
+      "loss": 2.8476,
+      "step": 3342
+    },
+    {
+      "epoch": 0.427904,
+      "grad_norm": 0.4371221363544464,
+      "learning_rate": 0.0006392108447522629,
+      "loss": 2.9792,
+      "step": 3343
+    },
+    {
+      "epoch": 0.428032,
+      "grad_norm": 0.4593190550804138,
+      "learning_rate": 0.000639011718978517,
+      "loss": 2.8606,
+      "step": 3344
+    },
+    {
+      "epoch": 0.42816,
+      "grad_norm": 0.4269559979438782,
+      "learning_rate": 0.0006388125693070451,
+      "loss": 2.8692,
+      "step": 3345
+    },
+    {
+      "epoch": 0.428288,
+      "grad_norm": 0.46427562832832336,
+      "learning_rate": 0.0006386133957720832,
+      "loss": 2.8013,
+      "step": 3346
+    },
+    {
+      "epoch": 0.428416,
+      "grad_norm": 0.39870554208755493,
+      "learning_rate": 0.0006384141984078719,
+      "loss": 2.861,
+      "step": 3347
+    },
+    {
+      "epoch": 0.428544,
+      "grad_norm": 0.4236806035041809,
+      "learning_rate": 0.000638214977248655,
+      "loss": 2.8946,
+      "step": 3348
+    },
+    {
+      "epoch": 0.428672,
+      "grad_norm": 0.4223303496837616,
+      "learning_rate": 0.0006380157323286812,
+      "loss": 2.9024,
+      "step": 3349
+    },
+    {
+      "epoch": 0.4288,
+      "grad_norm": 0.4085384011268616,
+      "learning_rate": 0.0006378164636822032,
+      "loss": 2.9584,
+      "step": 3350
+    },
+    {
+      "epoch": 0.428928,
+      "grad_norm": 0.43384429812431335,
+      "learning_rate": 0.0006376171713434771,
+      "loss": 2.9468,
+      "step": 3351
+    },
+    {
+      "epoch": 0.429056,
+      "grad_norm": 0.4931505620479584,
+      "learning_rate": 0.0006374178553467639,
+      "loss": 2.981,
+      "step": 3352
+    },
+    {
+      "epoch": 0.429184,
+      "grad_norm": 0.4186239540576935,
+      "learning_rate": 0.0006372185157263286,
+      "loss": 2.8432,
+      "step": 3353
+    },
+    {
+      "epoch": 0.429312,
+      "grad_norm": 0.4137890040874481,
+      "learning_rate": 0.0006370191525164394,
+      "loss": 2.9925,
+      "step": 3354
+    },
+    {
+      "epoch": 0.42944,
+      "grad_norm": 0.4346694350242615,
+      "learning_rate": 0.0006368197657513695,
+      "loss": 2.8908,
+      "step": 3355
+    },
+    {
+      "epoch": 0.429568,
+      "grad_norm": 0.43224164843559265,
+      "learning_rate": 0.0006366203554653957,
+      "loss": 2.8957,
+      "step": 3356
+    },
+    {
+      "epoch": 0.429696,
+      "grad_norm": 0.44849488139152527,
+      "learning_rate": 0.000636420921692799,
+      "loss": 2.9315,
+      "step": 3357
+    },
+    {
+      "epoch": 0.429824,
+      "grad_norm": 0.45557424426078796,
+      "learning_rate": 0.0006362214644678641,
+      "loss": 2.7828,
+      "step": 3358
+    },
+    {
+      "epoch": 0.429952,
+      "grad_norm": 0.4464561939239502,
+      "learning_rate": 0.0006360219838248806,
+      "loss": 2.9556,
+      "step": 3359
+    },
+    {
+      "epoch": 0.43008,
+      "grad_norm": 0.3985830545425415,
+      "learning_rate": 0.0006358224797981409,
+      "loss": 2.7417,
+      "step": 3360
+    },
+    {
+      "epoch": 0.430208,
+      "grad_norm": 0.45411217212677,
+      "learning_rate": 0.0006356229524219425,
+      "loss": 2.9172,
+      "step": 3361
+    },
+    {
+      "epoch": 0.430336,
+      "grad_norm": 0.4201340675354004,
+      "learning_rate": 0.0006354234017305862,
+      "loss": 2.8717,
+      "step": 3362
+    },
+    {
+      "epoch": 0.430464,
+      "grad_norm": 0.4478873908519745,
+      "learning_rate": 0.0006352238277583773,
+      "loss": 2.8107,
+      "step": 3363
+    },
+    {
+      "epoch": 0.430592,
+      "grad_norm": 0.4239387512207031,
+      "learning_rate": 0.0006350242305396247,
+      "loss": 2.9616,
+      "step": 3364
+    },
+    {
+      "epoch": 0.43072,
+      "grad_norm": 0.4246761202812195,
+      "learning_rate": 0.0006348246101086413,
+      "loss": 2.9351,
+      "step": 3365
+    },
+    {
+      "epoch": 0.430848,
+      "grad_norm": 0.4505438208580017,
+      "learning_rate": 0.0006346249664997448,
+      "loss": 2.9246,
+      "step": 3366
+    },
+    {
+      "epoch": 0.430976,
+      "grad_norm": 0.4659230709075928,
+      "learning_rate": 0.0006344252997472556,
+      "loss": 2.9631,
+      "step": 3367
+    },
+    {
+      "epoch": 0.431104,
+      "grad_norm": 0.4261031150817871,
+      "learning_rate": 0.0006342256098854992,
+      "loss": 2.7514,
+      "step": 3368
+    },
+    {
+      "epoch": 0.431232,
+      "grad_norm": 0.45321929454803467,
+      "learning_rate": 0.0006340258969488039,
+      "loss": 2.8904,
+      "step": 3369
+    },
+    {
+      "epoch": 0.43136,
+      "grad_norm": 0.43182751536369324,
+      "learning_rate": 0.0006338261609715037,
+      "loss": 2.9266,
+      "step": 3370
+    },
+    {
+      "epoch": 0.431488,
+      "grad_norm": 0.4460148811340332,
+      "learning_rate": 0.0006336264019879348,
+      "loss": 2.8702,
+      "step": 3371
+    },
+    {
+      "epoch": 0.431616,
+      "grad_norm": 0.43146607279777527,
+      "learning_rate": 0.0006334266200324381,
+      "loss": 2.9328,
+      "step": 3372
+    },
+    {
+      "epoch": 0.431744,
+      "grad_norm": 0.4471499025821686,
+      "learning_rate": 0.0006332268151393589,
+      "loss": 2.9151,
+      "step": 3373
+    },
+    {
+      "epoch": 0.431872,
+      "grad_norm": 0.4386645257472992,
+      "learning_rate": 0.0006330269873430454,
+      "loss": 2.7988,
+      "step": 3374
+    },
+    {
+      "epoch": 0.432,
+      "grad_norm": 0.45034968852996826,
+      "learning_rate": 0.0006328271366778509,
+      "loss": 2.8921,
+      "step": 3375
+    },
+    {
+      "epoch": 0.432128,
+      "grad_norm": 0.4472000002861023,
+      "learning_rate": 0.0006326272631781314,
+      "loss": 2.8063,
+      "step": 3376
+    },
+    {
+      "epoch": 0.432256,
+      "grad_norm": 0.4467780590057373,
+      "learning_rate": 0.000632427366878248,
+      "loss": 2.8218,
+      "step": 3377
+    },
+    {
+      "epoch": 0.432384,
+      "grad_norm": 0.4042727053165436,
+      "learning_rate": 0.000632227447812565,
+      "loss": 2.8144,
+      "step": 3378
+    },
+    {
+      "epoch": 0.432512,
+      "grad_norm": 0.5885103940963745,
+      "learning_rate": 0.0006320275060154507,
+      "loss": 2.9467,
+      "step": 3379
+    },
+    {
+      "epoch": 0.43264,
+      "grad_norm": 0.42153969407081604,
+      "learning_rate": 0.0006318275415212776,
+      "loss": 2.8093,
+      "step": 3380
+    },
+    {
+      "epoch": 0.432768,
+      "grad_norm": 0.42812371253967285,
+      "learning_rate": 0.0006316275543644218,
+      "loss": 2.9454,
+      "step": 3381
+    },
+    {
+      "epoch": 0.432896,
+      "grad_norm": 0.4147493541240692,
+      "learning_rate": 0.0006314275445792637,
+      "loss": 2.884,
+      "step": 3382
+    },
+    {
+      "epoch": 0.433024,
+      "grad_norm": 0.44156306982040405,
+      "learning_rate": 0.0006312275122001866,
+      "loss": 2.9978,
+      "step": 3383
+    },
+    {
+      "epoch": 0.433152,
+      "grad_norm": 0.42659661173820496,
+      "learning_rate": 0.0006310274572615791,
+      "loss": 2.8205,
+      "step": 3384
+    },
+    {
+      "epoch": 0.43328,
+      "grad_norm": 0.44071662425994873,
+      "learning_rate": 0.0006308273797978328,
+      "loss": 2.9358,
+      "step": 3385
+    },
+    {
+      "epoch": 0.433408,
+      "grad_norm": 0.46236732602119446,
+      "learning_rate": 0.000630627279843343,
+      "loss": 2.7294,
+      "step": 3386
+    },
+    {
+      "epoch": 0.433536,
+      "grad_norm": 0.4365192949771881,
+      "learning_rate": 0.0006304271574325095,
+      "loss": 2.7709,
+      "step": 3387
+    },
+    {
+      "epoch": 0.433664,
+      "grad_norm": 0.44091352820396423,
+      "learning_rate": 0.0006302270125997356,
+      "loss": 2.8581,
+      "step": 3388
+    },
+    {
+      "epoch": 0.433792,
+      "grad_norm": 0.4519268274307251,
+      "learning_rate": 0.0006300268453794287,
+      "loss": 2.8489,
+      "step": 3389
+    },
+    {
+      "epoch": 0.43392,
+      "grad_norm": 0.45855823159217834,
+      "learning_rate": 0.0006298266558059995,
+      "loss": 2.8892,
+      "step": 3390
+    },
+    {
+      "epoch": 0.434048,
+      "grad_norm": 0.4882846474647522,
+      "learning_rate": 0.000629626443913863,
+      "loss": 2.9046,
+      "step": 3391
+    },
+    {
+      "epoch": 0.434176,
+      "grad_norm": 0.4353890120983124,
+      "learning_rate": 0.0006294262097374382,
+      "loss": 2.879,
+      "step": 3392
+    },
+    {
+      "epoch": 0.434304,
+      "grad_norm": 0.4150359034538269,
+      "learning_rate": 0.0006292259533111474,
+      "loss": 2.8829,
+      "step": 3393
+    },
+    {
+      "epoch": 0.434432,
+      "grad_norm": 0.4573715329170227,
+      "learning_rate": 0.0006290256746694172,
+      "loss": 2.9786,
+      "step": 3394
+    },
+    {
+      "epoch": 0.43456,
+      "grad_norm": 0.41422373056411743,
+      "learning_rate": 0.0006288253738466777,
+      "loss": 2.8391,
+      "step": 3395
+    },
+    {
+      "epoch": 0.434688,
+      "grad_norm": 0.46019986271858215,
+      "learning_rate": 0.000628625050877363,
+      "loss": 3.0274,
+      "step": 3396
+    },
+    {
+      "epoch": 0.434816,
+      "grad_norm": 0.4546186923980713,
+      "learning_rate": 0.0006284247057959107,
+      "loss": 3.1279,
+      "step": 3397
+    },
+    {
+      "epoch": 0.434944,
+      "grad_norm": 0.4323049783706665,
+      "learning_rate": 0.0006282243386367628,
+      "loss": 2.9005,
+      "step": 3398
+    },
+    {
+      "epoch": 0.435072,
+      "grad_norm": 0.498582661151886,
+      "learning_rate": 0.0006280239494343647,
+      "loss": 2.8372,
+      "step": 3399
+    },
+    {
+      "epoch": 0.4352,
+      "grad_norm": 0.4398665130138397,
+      "learning_rate": 0.0006278235382231654,
+      "loss": 2.8916,
+      "step": 3400
+    },
+    {
+      "epoch": 0.435328,
+      "grad_norm": 0.4278354048728943,
+      "learning_rate": 0.000627623105037618,
+      "loss": 2.9215,
+      "step": 3401
+    },
+    {
+      "epoch": 0.435456,
+      "grad_norm": 0.45607736706733704,
+      "learning_rate": 0.0006274226499121792,
+      "loss": 2.8419,
+      "step": 3402
+    },
+    {
+      "epoch": 0.435584,
+      "grad_norm": 0.40356582403182983,
+      "learning_rate": 0.0006272221728813099,
+      "loss": 2.857,
+      "step": 3403
+    },
+    {
+      "epoch": 0.435712,
+      "grad_norm": 0.439904123544693,
+      "learning_rate": 0.0006270216739794741,
+      "loss": 2.7099,
+      "step": 3404
+    },
+    {
+      "epoch": 0.43584,
+      "grad_norm": 0.42117756605148315,
+      "learning_rate": 0.00062682115324114,
+      "loss": 2.8535,
+      "step": 3405
+    },
+    {
+      "epoch": 0.435968,
+      "grad_norm": 0.4396045506000519,
+      "learning_rate": 0.0006266206107007793,
+      "loss": 2.9152,
+      "step": 3406
+    },
+    {
+      "epoch": 0.436096,
+      "grad_norm": 0.438849538564682,
+      "learning_rate": 0.0006264200463928679,
+      "loss": 2.7294,
+      "step": 3407
+    },
+    {
+      "epoch": 0.436224,
+      "grad_norm": 0.4543686509132385,
+      "learning_rate": 0.0006262194603518847,
+      "loss": 2.9966,
+      "step": 3408
+    },
+    {
+      "epoch": 0.436352,
+      "grad_norm": 0.4484862685203552,
+      "learning_rate": 0.0006260188526123133,
+      "loss": 2.8734,
+      "step": 3409
+    },
+    {
+      "epoch": 0.43648,
+      "grad_norm": 0.475974977016449,
+      "learning_rate": 0.00062581822320864,
+      "loss": 2.9422,
+      "step": 3410
+    },
+    {
+      "epoch": 0.436608,
+      "grad_norm": 0.442452609539032,
+      "learning_rate": 0.0006256175721753556,
+      "loss": 2.7412,
+      "step": 3411
+    },
+    {
+      "epoch": 0.436736,
+      "grad_norm": 0.4197764992713928,
+      "learning_rate": 0.0006254168995469541,
+      "loss": 2.9667,
+      "step": 3412
+    },
+    {
+      "epoch": 0.436864,
+      "grad_norm": 0.42917072772979736,
+      "learning_rate": 0.0006252162053579337,
+      "loss": 2.8069,
+      "step": 3413
+    },
+    {
+      "epoch": 0.436992,
+      "grad_norm": 0.4029163122177124,
+      "learning_rate": 0.0006250154896427962,
+      "loss": 2.9591,
+      "step": 3414
+    },
+    {
+      "epoch": 0.43712,
+      "grad_norm": 0.4202183783054352,
+      "learning_rate": 0.0006248147524360464,
+      "loss": 2.9594,
+      "step": 3415
+    },
+    {
+      "epoch": 0.437248,
+      "grad_norm": 0.4438765048980713,
+      "learning_rate": 0.0006246139937721939,
+      "loss": 2.7771,
+      "step": 3416
+    },
+    {
+      "epoch": 0.437376,
+      "grad_norm": 0.4552188515663147,
+      "learning_rate": 0.0006244132136857512,
+      "loss": 2.8797,
+      "step": 3417
+    },
+    {
+      "epoch": 0.437504,
+      "grad_norm": 0.42662230134010315,
+      "learning_rate": 0.0006242124122112347,
+      "loss": 2.9487,
+      "step": 3418
+    },
+    {
+      "epoch": 0.437632,
+      "grad_norm": 0.41676023602485657,
+      "learning_rate": 0.0006240115893831644,
+      "loss": 2.8894,
+      "step": 3419
+    },
+    {
+      "epoch": 0.43776,
+      "grad_norm": 0.4313552975654602,
+      "learning_rate": 0.0006238107452360643,
+      "loss": 2.9688,
+      "step": 3420
+    },
+    {
+      "epoch": 0.437888,
+      "grad_norm": 0.5192545652389526,
+      "learning_rate": 0.0006236098798044618,
+      "loss": 2.8641,
+      "step": 3421
+    },
+    {
+      "epoch": 0.438016,
+      "grad_norm": 0.42516785860061646,
+      "learning_rate": 0.000623408993122888,
+      "loss": 2.8259,
+      "step": 3422
+    },
+    {
+      "epoch": 0.438144,
+      "grad_norm": 0.426675945520401,
+      "learning_rate": 0.0006232080852258776,
+      "loss": 2.8875,
+      "step": 3423
+    },
+    {
+      "epoch": 0.438272,
+      "grad_norm": 0.43668028712272644,
+      "learning_rate": 0.000623007156147969,
+      "loss": 2.8296,
+      "step": 3424
+    },
+    {
+      "epoch": 0.4384,
+      "grad_norm": 0.4308415651321411,
+      "learning_rate": 0.000622806205923704,
+      "loss": 2.8386,
+      "step": 3425
+    },
+    {
+      "epoch": 0.438528,
+      "grad_norm": 0.43884772062301636,
+      "learning_rate": 0.0006226052345876288,
+      "loss": 2.9484,
+      "step": 3426
+    },
+    {
+      "epoch": 0.438656,
+      "grad_norm": 0.4351074695587158,
+      "learning_rate": 0.0006224042421742924,
+      "loss": 2.8231,
+      "step": 3427
+    },
+    {
+      "epoch": 0.438784,
+      "grad_norm": 0.4449838697910309,
+      "learning_rate": 0.0006222032287182477,
+      "loss": 2.949,
+      "step": 3428
+    },
+    {
+      "epoch": 0.438912,
+      "grad_norm": 0.43482545018196106,
+      "learning_rate": 0.0006220021942540512,
+      "loss": 2.8595,
+      "step": 3429
+    },
+    {
+      "epoch": 0.43904,
+      "grad_norm": 0.47595250606536865,
+      "learning_rate": 0.000621801138816263,
+      "loss": 2.7583,
+      "step": 3430
+    },
+    {
+      "epoch": 0.439168,
+      "grad_norm": 0.4483700692653656,
+      "learning_rate": 0.0006216000624394472,
+      "loss": 2.9751,
+      "step": 3431
+    },
+    {
+      "epoch": 0.439296,
+      "grad_norm": 0.4694201648235321,
+      "learning_rate": 0.000621398965158171,
+      "loss": 3.0384,
+      "step": 3432
+    },
+    {
+      "epoch": 0.439424,
+      "grad_norm": 0.4767659306526184,
+      "learning_rate": 0.0006211978470070052,
+      "loss": 2.8004,
+      "step": 3433
+    },
+    {
+      "epoch": 0.439552,
+      "grad_norm": 0.44125038385391235,
+      "learning_rate": 0.0006209967080205244,
+      "loss": 2.8909,
+      "step": 3434
+    },
+    {
+      "epoch": 0.43968,
+      "grad_norm": 0.4338669180870056,
+      "learning_rate": 0.000620795548233307,
+      "loss": 2.8369,
+      "step": 3435
+    },
+    {
+      "epoch": 0.439808,
+      "grad_norm": 0.4470394551753998,
+      "learning_rate": 0.0006205943676799344,
+      "loss": 2.9775,
+      "step": 3436
+    },
+    {
+      "epoch": 0.439936,
+      "grad_norm": 0.45222875475883484,
+      "learning_rate": 0.0006203931663949918,
+      "loss": 2.6674,
+      "step": 3437
+    },
+    {
+      "epoch": 0.440064,
+      "grad_norm": 0.48931387066841125,
+      "learning_rate": 0.0006201919444130683,
+      "loss": 2.855,
+      "step": 3438
+    },
+    {
+      "epoch": 0.440192,
+      "grad_norm": 0.43883705139160156,
+      "learning_rate": 0.0006199907017687561,
+      "loss": 2.9437,
+      "step": 3439
+    },
+    {
+      "epoch": 0.44032,
+      "grad_norm": 0.4332881271839142,
+      "learning_rate": 0.0006197894384966512,
+      "loss": 2.8007,
+      "step": 3440
+    },
+    {
+      "epoch": 0.440448,
+      "grad_norm": 0.43893197178840637,
+      "learning_rate": 0.0006195881546313532,
+      "loss": 2.9739,
+      "step": 3441
+    },
+    {
+      "epoch": 0.440576,
+      "grad_norm": 0.4322003126144409,
+      "learning_rate": 0.0006193868502074651,
+      "loss": 2.8834,
+      "step": 3442
+    },
+    {
+      "epoch": 0.440704,
+      "grad_norm": 0.43789538741111755,
+      "learning_rate": 0.0006191855252595933,
+      "loss": 2.9054,
+      "step": 3443
+    },
+    {
+      "epoch": 0.440832,
+      "grad_norm": 0.4043145477771759,
+      "learning_rate": 0.0006189841798223479,
+      "loss": 2.8074,
+      "step": 3444
+    },
+    {
+      "epoch": 0.44096,
+      "grad_norm": 0.43007132411003113,
+      "learning_rate": 0.0006187828139303427,
+      "loss": 2.8168,
+      "step": 3445
+    },
+    {
+      "epoch": 0.441088,
+      "grad_norm": 0.4306334853172302,
+      "learning_rate": 0.0006185814276181948,
+      "loss": 2.8919,
+      "step": 3446
+    },
+    {
+      "epoch": 0.441216,
+      "grad_norm": 0.4639306664466858,
+      "learning_rate": 0.0006183800209205246,
+      "loss": 2.8644,
+      "step": 3447
+    },
+    {
+      "epoch": 0.441344,
+      "grad_norm": 0.4298896789550781,
+      "learning_rate": 0.0006181785938719565,
+      "loss": 2.9044,
+      "step": 3448
+    },
+    {
+      "epoch": 0.441472,
+      "grad_norm": 0.43429863452911377,
+      "learning_rate": 0.0006179771465071181,
+      "loss": 2.9072,
+      "step": 3449
+    },
+    {
+      "epoch": 0.4416,
+      "grad_norm": 0.44122400879859924,
+      "learning_rate": 0.0006177756788606405,
+      "loss": 2.7498,
+      "step": 3450
+    },
+    {
+      "epoch": 0.441728,
+      "grad_norm": 0.4352279603481293,
+      "learning_rate": 0.0006175741909671584,
+      "loss": 2.9038,
+      "step": 3451
+    },
+    {
+      "epoch": 0.441856,
+      "grad_norm": 0.4514691233634949,
+      "learning_rate": 0.0006173726828613097,
+      "loss": 2.9284,
+      "step": 3452
+    },
+    {
+      "epoch": 0.441984,
+      "grad_norm": 0.45021921396255493,
+      "learning_rate": 0.0006171711545777363,
+      "loss": 2.8562,
+      "step": 3453
+    },
+    {
+      "epoch": 0.442112,
+      "grad_norm": 0.4387170970439911,
+      "learning_rate": 0.000616969606151083,
+      "loss": 2.8603,
+      "step": 3454
+    },
+    {
+      "epoch": 0.44224,
+      "grad_norm": 0.4325452744960785,
+      "learning_rate": 0.0006167680376159982,
+      "loss": 2.9251,
+      "step": 3455
+    },
+    {
+      "epoch": 0.442368,
+      "grad_norm": 0.4036429524421692,
+      "learning_rate": 0.0006165664490071342,
+      "loss": 2.9891,
+      "step": 3456
+    },
+    {
+      "epoch": 0.442496,
+      "grad_norm": 0.4479140639305115,
+      "learning_rate": 0.0006163648403591462,
+      "loss": 2.8291,
+      "step": 3457
+    },
+    {
+      "epoch": 0.442624,
+      "grad_norm": 0.40760162472724915,
+      "learning_rate": 0.0006161632117066931,
+      "loss": 2.9145,
+      "step": 3458
+    },
+    {
+      "epoch": 0.442752,
+      "grad_norm": 0.4348607063293457,
+      "learning_rate": 0.0006159615630844374,
+      "loss": 2.9044,
+      "step": 3459
+    },
+    {
+      "epoch": 0.44288,
+      "grad_norm": 0.42111843824386597,
+      "learning_rate": 0.0006157598945270446,
+      "loss": 2.8874,
+      "step": 3460
+    },
+    {
+      "epoch": 0.443008,
+      "grad_norm": 0.44388481974601746,
+      "learning_rate": 0.0006155582060691837,
+      "loss": 2.8707,
+      "step": 3461
+    },
+    {
+      "epoch": 0.443136,
+      "grad_norm": 0.46995678544044495,
+      "learning_rate": 0.0006153564977455277,
+      "loss": 2.7839,
+      "step": 3462
+    },
+    {
+      "epoch": 0.443264,
+      "grad_norm": 0.45619434118270874,
+      "learning_rate": 0.0006151547695907525,
+      "loss": 3.0496,
+      "step": 3463
+    },
+    {
+      "epoch": 0.443392,
+      "grad_norm": 0.45141348242759705,
+      "learning_rate": 0.0006149530216395374,
+      "loss": 2.866,
+      "step": 3464
+    },
+    {
+      "epoch": 0.44352,
+      "grad_norm": 0.44188785552978516,
+      "learning_rate": 0.0006147512539265651,
+      "loss": 2.9477,
+      "step": 3465
+    },
+    {
+      "epoch": 0.443648,
+      "grad_norm": 0.44253042340278625,
+      "learning_rate": 0.000614549466486522,
+      "loss": 2.9214,
+      "step": 3466
+    },
+    {
+      "epoch": 0.443776,
+      "grad_norm": 0.42663800716400146,
+      "learning_rate": 0.0006143476593540976,
+      "loss": 2.9936,
+      "step": 3467
+    },
+    {
+      "epoch": 0.443904,
+      "grad_norm": 0.44325360655784607,
+      "learning_rate": 0.000614145832563985,
+      "loss": 2.9079,
+      "step": 3468
+    },
+    {
+      "epoch": 0.444032,
+      "grad_norm": 0.41754987835884094,
+      "learning_rate": 0.0006139439861508803,
+      "loss": 2.8656,
+      "step": 3469
+    },
+    {
+      "epoch": 0.44416,
+      "grad_norm": 0.43650999665260315,
+      "learning_rate": 0.0006137421201494836,
+      "loss": 2.905,
+      "step": 3470
+    },
+    {
+      "epoch": 0.444288,
+      "grad_norm": 0.4266263246536255,
+      "learning_rate": 0.0006135402345944978,
+      "loss": 2.8844,
+      "step": 3471
+    },
+    {
+      "epoch": 0.444416,
+      "grad_norm": 0.42880958318710327,
+      "learning_rate": 0.0006133383295206294,
+      "loss": 2.8077,
+      "step": 3472
+    },
+    {
+      "epoch": 0.444544,
+      "grad_norm": 0.45412591099739075,
+      "learning_rate": 0.0006131364049625882,
+      "loss": 2.8257,
+      "step": 3473
+    },
+    {
+      "epoch": 0.444672,
+      "grad_norm": 0.4847337603569031,
+      "learning_rate": 0.0006129344609550876,
+      "loss": 2.8829,
+      "step": 3474
+    },
+    {
+      "epoch": 0.4448,
+      "grad_norm": 0.4172142744064331,
+      "learning_rate": 0.0006127324975328436,
+      "loss": 2.9596,
+      "step": 3475
+    },
+    {
+      "epoch": 0.444928,
+      "grad_norm": 0.3953095078468323,
+      "learning_rate": 0.0006125305147305764,
+      "loss": 2.8603,
+      "step": 3476
+    },
+    {
+      "epoch": 0.445056,
+      "grad_norm": 0.5183844566345215,
+      "learning_rate": 0.0006123285125830093,
+      "loss": 2.8121,
+      "step": 3477
+    },
+    {
+      "epoch": 0.445184,
+      "grad_norm": 0.4071420729160309,
+      "learning_rate": 0.0006121264911248687,
+      "loss": 2.8148,
+      "step": 3478
+    },
+    {
+      "epoch": 0.445312,
+      "grad_norm": 0.43917977809906006,
+      "learning_rate": 0.0006119244503908843,
+      "loss": 2.8755,
+      "step": 3479
+    },
+    {
+      "epoch": 0.44544,
+      "grad_norm": 0.41379663348197937,
+      "learning_rate": 0.0006117223904157893,
+      "loss": 2.8811,
+      "step": 3480
+    },
+    {
+      "epoch": 0.445568,
+      "grad_norm": 0.4009621739387512,
+      "learning_rate": 0.0006115203112343203,
+      "loss": 2.7719,
+      "step": 3481
+    },
+    {
+      "epoch": 0.445696,
+      "grad_norm": 0.4354688227176666,
+      "learning_rate": 0.0006113182128812169,
+      "loss": 2.9788,
+      "step": 3482
+    },
+    {
+      "epoch": 0.445824,
+      "grad_norm": 0.464785635471344,
+      "learning_rate": 0.0006111160953912221,
+      "loss": 2.8484,
+      "step": 3483
+    },
+    {
+      "epoch": 0.445952,
+      "grad_norm": 0.4272063672542572,
+      "learning_rate": 0.0006109139587990825,
+      "loss": 2.7967,
+      "step": 3484
+    },
+    {
+      "epoch": 0.44608,
+      "grad_norm": 0.4661426842212677,
+      "learning_rate": 0.0006107118031395475,
+      "loss": 2.8457,
+      "step": 3485
+    },
+    {
+      "epoch": 0.446208,
+      "grad_norm": 0.43879982829093933,
+      "learning_rate": 0.0006105096284473701,
+      "loss": 2.9562,
+      "step": 3486
+    },
+    {
+      "epoch": 0.446336,
+      "grad_norm": 0.514254093170166,
+      "learning_rate": 0.0006103074347573063,
+      "loss": 2.8419,
+      "step": 3487
+    },
+    {
+      "epoch": 0.446464,
+      "grad_norm": 0.39470353722572327,
+      "learning_rate": 0.0006101052221041157,
+      "loss": 3.0055,
+      "step": 3488
+    },
+    {
+      "epoch": 0.446592,
+      "grad_norm": 0.403464138507843,
+      "learning_rate": 0.0006099029905225611,
+      "loss": 2.8025,
+      "step": 3489
+    },
+    {
+      "epoch": 0.44672,
+      "grad_norm": 0.4277206361293793,
+      "learning_rate": 0.000609700740047408,
+      "loss": 2.7612,
+      "step": 3490
+    },
+    {
+      "epoch": 0.446848,
+      "grad_norm": 0.5877236127853394,
+      "learning_rate": 0.0006094984707134262,
+      "loss": 2.8942,
+      "step": 3491
+    },
+    {
+      "epoch": 0.446976,
+      "grad_norm": 0.41732120513916016,
+      "learning_rate": 0.000609296182555388,
+      "loss": 2.7178,
+      "step": 3492
+    },
+    {
+      "epoch": 0.447104,
+      "grad_norm": 0.43678340315818787,
+      "learning_rate": 0.0006090938756080688,
+      "loss": 2.8697,
+      "step": 3493
+    },
+    {
+      "epoch": 0.447232,
+      "grad_norm": 0.4685511589050293,
+      "learning_rate": 0.0006088915499062475,
+      "loss": 2.842,
+      "step": 3494
+    },
+    {
+      "epoch": 0.44736,
+      "grad_norm": 0.42039552330970764,
+      "learning_rate": 0.0006086892054847065,
+      "loss": 2.8437,
+      "step": 3495
+    },
+    {
+      "epoch": 0.447488,
+      "grad_norm": 0.48491987586021423,
+      "learning_rate": 0.0006084868423782312,
+      "loss": 2.8905,
+      "step": 3496
+    },
+    {
+      "epoch": 0.447616,
+      "grad_norm": 0.43184635043144226,
+      "learning_rate": 0.0006082844606216098,
+      "loss": 3.0786,
+      "step": 3497
+    },
+    {
+      "epoch": 0.447744,
+      "grad_norm": 0.44290366768836975,
+      "learning_rate": 0.0006080820602496344,
+      "loss": 2.918,
+      "step": 3498
+    },
+    {
+      "epoch": 0.447872,
+      "grad_norm": 0.3999873101711273,
+      "learning_rate": 0.0006078796412970997,
+      "loss": 2.7052,
+      "step": 3499
+    },
+    {
+      "epoch": 0.448,
+      "grad_norm": 0.4308745265007019,
+      "learning_rate": 0.0006076772037988042,
+      "loss": 2.9489,
+      "step": 3500
+    },
+    {
+      "epoch": 0.448128,
+      "grad_norm": 0.42468205094337463,
+      "learning_rate": 0.0006074747477895489,
+      "loss": 2.9471,
+      "step": 3501
+    },
+    {
+      "epoch": 0.448256,
+      "grad_norm": 0.43171653151512146,
+      "learning_rate": 0.0006072722733041387,
+      "loss": 2.944,
+      "step": 3502
+    },
+    {
+      "epoch": 0.448384,
+      "grad_norm": 0.44100114703178406,
+      "learning_rate": 0.000607069780377381,
+      "loss": 2.888,
+      "step": 3503
+    },
+    {
+      "epoch": 0.448512,
+      "grad_norm": 0.41490495204925537,
+      "learning_rate": 0.0006068672690440867,
+      "loss": 2.8298,
+      "step": 3504
+    },
+    {
+      "epoch": 0.44864,
+      "grad_norm": 0.4191237986087799,
+      "learning_rate": 0.00060666473933907,
+      "loss": 2.8759,
+      "step": 3505
+    },
+    {
+      "epoch": 0.448768,
+      "grad_norm": 0.4297257363796234,
+      "learning_rate": 0.0006064621912971481,
+      "loss": 2.9013,
+      "step": 3506
+    },
+    {
+      "epoch": 0.448896,
+      "grad_norm": 0.4351956844329834,
+      "learning_rate": 0.0006062596249531414,
+      "loss": 2.8846,
+      "step": 3507
+    },
+    {
+      "epoch": 0.449024,
+      "grad_norm": 0.443943589925766,
+      "learning_rate": 0.0006060570403418731,
+      "loss": 2.9276,
+      "step": 3508
+    },
+    {
+      "epoch": 0.449152,
+      "grad_norm": 0.4364476501941681,
+      "learning_rate": 0.00060585443749817,
+      "loss": 2.8727,
+      "step": 3509
+    },
+    {
+      "epoch": 0.44928,
+      "grad_norm": 0.41835662722587585,
+      "learning_rate": 0.0006056518164568622,
+      "loss": 2.8748,
+      "step": 3510
+    },
+    {
+      "epoch": 0.449408,
+      "grad_norm": 0.41202035546302795,
+      "learning_rate": 0.0006054491772527821,
+      "loss": 2.913,
+      "step": 3511
+    },
+    {
+      "epoch": 0.449536,
+      "grad_norm": 0.4821150302886963,
+      "learning_rate": 0.0006052465199207661,
+      "loss": 2.8027,
+      "step": 3512
+    },
+    {
+      "epoch": 0.449664,
+      "grad_norm": 0.45584115386009216,
+      "learning_rate": 0.0006050438444956531,
+      "loss": 2.7337,
+      "step": 3513
+    },
+    {
+      "epoch": 0.449792,
+      "grad_norm": 0.4090549945831299,
+      "learning_rate": 0.0006048411510122855,
+      "loss": 2.8927,
+      "step": 3514
+    },
+    {
+      "epoch": 0.44992,
+      "grad_norm": 0.4302290976047516,
+      "learning_rate": 0.0006046384395055086,
+      "loss": 2.842,
+      "step": 3515
+    },
+    {
+      "epoch": 0.450048,
+      "grad_norm": 0.4104582965373993,
+      "learning_rate": 0.000604435710010171,
+      "loss": 2.8155,
+      "step": 3516
+    },
+    {
+      "epoch": 0.450176,
+      "grad_norm": 0.44897329807281494,
+      "learning_rate": 0.0006042329625611239,
+      "loss": 2.9424,
+      "step": 3517
+    },
+    {
+      "epoch": 0.450304,
+      "grad_norm": 0.4657040536403656,
+      "learning_rate": 0.0006040301971932222,
+      "loss": 2.9192,
+      "step": 3518
+    },
+    {
+      "epoch": 0.450432,
+      "grad_norm": 0.4133733808994293,
+      "learning_rate": 0.0006038274139413237,
+      "loss": 2.9263,
+      "step": 3519
+    },
+    {
+      "epoch": 0.45056,
+      "grad_norm": 0.44951021671295166,
+      "learning_rate": 0.0006036246128402891,
+      "loss": 2.7831,
+      "step": 3520
+    },
+    {
+      "epoch": 0.450688,
+      "grad_norm": 0.42116814851760864,
+      "learning_rate": 0.0006034217939249822,
+      "loss": 2.9238,
+      "step": 3521
+    },
+    {
+      "epoch": 0.450816,
+      "grad_norm": 0.4405933916568756,
+      "learning_rate": 0.00060321895723027,
+      "loss": 2.8502,
+      "step": 3522
+    },
+    {
+      "epoch": 0.450944,
+      "grad_norm": 0.4307357966899872,
+      "learning_rate": 0.0006030161027910223,
+      "loss": 2.8124,
+      "step": 3523
+    },
+    {
+      "epoch": 0.451072,
+      "grad_norm": 0.44363030791282654,
+      "learning_rate": 0.0006028132306421124,
+      "loss": 2.8863,
+      "step": 3524
+    },
+    {
+      "epoch": 0.4512,
+      "grad_norm": 0.4119660258293152,
+      "learning_rate": 0.0006026103408184162,
+      "loss": 2.8791,
+      "step": 3525
+    },
+    {
+      "epoch": 0.451328,
+      "grad_norm": 0.46098077297210693,
+      "learning_rate": 0.0006024074333548127,
+      "loss": 2.9482,
+      "step": 3526
+    },
+    {
+      "epoch": 0.451456,
+      "grad_norm": 0.41478919982910156,
+      "learning_rate": 0.0006022045082861843,
+      "loss": 2.8871,
+      "step": 3527
+    },
+    {
+      "epoch": 0.451584,
+      "grad_norm": 0.47909852862358093,
+      "learning_rate": 0.0006020015656474161,
+      "loss": 2.8992,
+      "step": 3528
+    },
+    {
+      "epoch": 0.451712,
+      "grad_norm": 0.4834597110748291,
+      "learning_rate": 0.0006017986054733962,
+      "loss": 2.8162,
+      "step": 3529
+    },
+    {
+      "epoch": 0.45184,
+      "grad_norm": 0.44385093450546265,
+      "learning_rate": 0.000601595627799016,
+      "loss": 2.8396,
+      "step": 3530
+    },
+    {
+      "epoch": 0.451968,
+      "grad_norm": 0.40124887228012085,
+      "learning_rate": 0.0006013926326591695,
+      "loss": 2.8916,
+      "step": 3531
+    },
+    {
+      "epoch": 0.452096,
+      "grad_norm": 0.4316502511501312,
+      "learning_rate": 0.0006011896200887539,
+      "loss": 2.8618,
+      "step": 3532
+    },
+    {
+      "epoch": 0.452224,
+      "grad_norm": 0.4415668845176697,
+      "learning_rate": 0.0006009865901226697,
+      "loss": 2.7607,
+      "step": 3533
+    },
+    {
+      "epoch": 0.452352,
+      "grad_norm": 0.44322916865348816,
+      "learning_rate": 0.0006007835427958198,
+      "loss": 3.0652,
+      "step": 3534
+    },
+    {
+      "epoch": 0.45248,
+      "grad_norm": 0.42921629548072815,
+      "learning_rate": 0.0006005804781431106,
+      "loss": 2.8618,
+      "step": 3535
+    },
+    {
+      "epoch": 0.452608,
+      "grad_norm": 0.44059181213378906,
+      "learning_rate": 0.0006003773961994511,
+      "loss": 2.814,
+      "step": 3536
+    },
+    {
+      "epoch": 0.452736,
+      "grad_norm": 0.45349255204200745,
+      "learning_rate": 0.0006001742969997534,
+      "loss": 2.9205,
+      "step": 3537
+    },
+    {
+      "epoch": 0.452864,
+      "grad_norm": 0.5859175324440002,
+      "learning_rate": 0.0005999711805789329,
+      "loss": 2.8259,
+      "step": 3538
+    },
+    {
+      "epoch": 0.452992,
+      "grad_norm": 0.47362446784973145,
+      "learning_rate": 0.0005997680469719076,
+      "loss": 2.8313,
+      "step": 3539
+    },
+    {
+      "epoch": 0.45312,
+      "grad_norm": 0.45135435461997986,
+      "learning_rate": 0.0005995648962135983,
+      "loss": 2.9405,
+      "step": 3540
+    },
+    {
+      "epoch": 0.453248,
+      "grad_norm": 0.4314587116241455,
+      "learning_rate": 0.0005993617283389289,
+      "loss": 2.8005,
+      "step": 3541
+    },
+    {
+      "epoch": 0.453376,
+      "grad_norm": 0.44209444522857666,
+      "learning_rate": 0.0005991585433828267,
+      "loss": 2.7706,
+      "step": 3542
+    },
+    {
+      "epoch": 0.453504,
+      "grad_norm": 0.47088706493377686,
+      "learning_rate": 0.0005989553413802214,
+      "loss": 2.8372,
+      "step": 3543
+    },
+    {
+      "epoch": 0.453632,
+      "grad_norm": 0.4303573966026306,
+      "learning_rate": 0.0005987521223660456,
+      "loss": 2.8011,
+      "step": 3544
+    },
+    {
+      "epoch": 0.45376,
+      "grad_norm": 0.4325943887233734,
+      "learning_rate": 0.0005985488863752351,
+      "loss": 2.7739,
+      "step": 3545
+    },
+    {
+      "epoch": 0.453888,
+      "grad_norm": 0.5064706206321716,
+      "learning_rate": 0.0005983456334427286,
+      "loss": 2.7264,
+      "step": 3546
+    },
+    {
+      "epoch": 0.454016,
+      "grad_norm": 0.43346285820007324,
+      "learning_rate": 0.0005981423636034676,
+      "loss": 2.7886,
+      "step": 3547
+    },
+    {
+      "epoch": 0.454144,
+      "grad_norm": 0.48280712962150574,
+      "learning_rate": 0.0005979390768923967,
+      "loss": 3.0139,
+      "step": 3548
+    },
+    {
+      "epoch": 0.454272,
+      "grad_norm": 0.43050989508628845,
+      "learning_rate": 0.000597735773344463,
+      "loss": 2.8056,
+      "step": 3549
+    },
+    {
+      "epoch": 0.4544,
+      "grad_norm": 0.43140602111816406,
+      "learning_rate": 0.0005975324529946165,
+      "loss": 2.9003,
+      "step": 3550
+    },
+    {
+      "epoch": 0.454528,
+      "grad_norm": 0.4823165535926819,
+      "learning_rate": 0.0005973291158778109,
+      "loss": 2.854,
+      "step": 3551
+    },
+    {
+      "epoch": 0.454656,
+      "grad_norm": 0.4367239475250244,
+      "learning_rate": 0.000597125762029002,
+      "loss": 2.9342,
+      "step": 3552
+    },
+    {
+      "epoch": 0.454784,
+      "grad_norm": 0.4442397356033325,
+      "learning_rate": 0.0005969223914831486,
+      "loss": 2.8633,
+      "step": 3553
+    },
+    {
+      "epoch": 0.454912,
+      "grad_norm": 0.47130605578422546,
+      "learning_rate": 0.0005967190042752123,
+      "loss": 2.8251,
+      "step": 3554
+    },
+    {
+      "epoch": 0.45504,
+      "grad_norm": 0.436886191368103,
+      "learning_rate": 0.000596515600440158,
+      "loss": 2.8815,
+      "step": 3555
+    },
+    {
+      "epoch": 0.455168,
+      "grad_norm": 0.5095201134681702,
+      "learning_rate": 0.0005963121800129532,
+      "loss": 2.9501,
+      "step": 3556
+    },
+    {
+      "epoch": 0.455296,
+      "grad_norm": 0.4319216012954712,
+      "learning_rate": 0.0005961087430285681,
+      "loss": 2.8848,
+      "step": 3557
+    },
+    {
+      "epoch": 0.455424,
+      "grad_norm": 0.4383789300918579,
+      "learning_rate": 0.0005959052895219757,
+      "loss": 2.8003,
+      "step": 3558
+    },
+    {
+      "epoch": 0.455552,
+      "grad_norm": 0.4213782846927643,
+      "learning_rate": 0.0005957018195281523,
+      "loss": 2.8789,
+      "step": 3559
+    },
+    {
+      "epoch": 0.45568,
+      "grad_norm": 0.41278448700904846,
+      "learning_rate": 0.0005954983330820766,
+      "loss": 2.981,
+      "step": 3560
+    },
+    {
+      "epoch": 0.455808,
+      "grad_norm": 0.40470579266548157,
+      "learning_rate": 0.0005952948302187306,
+      "loss": 2.827,
+      "step": 3561
+    },
+    {
+      "epoch": 0.455936,
+      "grad_norm": 0.4147280156612396,
+      "learning_rate": 0.0005950913109730982,
+      "loss": 2.877,
+      "step": 3562
+    },
+    {
+      "epoch": 0.456064,
+      "grad_norm": 0.40479788184165955,
+      "learning_rate": 0.0005948877753801673,
+      "loss": 2.8852,
+      "step": 3563
+    },
+    {
+      "epoch": 0.456192,
+      "grad_norm": 0.3982674777507782,
+      "learning_rate": 0.0005946842234749274,
+      "loss": 2.8313,
+      "step": 3564
+    },
+    {
+      "epoch": 0.45632,
+      "grad_norm": 0.42275652289390564,
+      "learning_rate": 0.0005944806552923721,
+      "loss": 2.8976,
+      "step": 3565
+    },
+    {
+      "epoch": 0.456448,
+      "grad_norm": 0.42956021428108215,
+      "learning_rate": 0.0005942770708674969,
+      "loss": 2.9332,
+      "step": 3566
+    },
+    {
+      "epoch": 0.456576,
+      "grad_norm": 0.4444783627986908,
+      "learning_rate": 0.0005940734702353002,
+      "loss": 2.9483,
+      "step": 3567
+    },
+    {
+      "epoch": 0.456704,
+      "grad_norm": 0.41162919998168945,
+      "learning_rate": 0.0005938698534307833,
+      "loss": 2.878,
+      "step": 3568
+    },
+    {
+      "epoch": 0.456832,
+      "grad_norm": 0.3997342884540558,
+      "learning_rate": 0.0005936662204889504,
+      "loss": 2.8911,
+      "step": 3569
+    },
+    {
+      "epoch": 0.45696,
+      "grad_norm": 0.45338770747184753,
+      "learning_rate": 0.0005934625714448084,
+      "loss": 2.852,
+      "step": 3570
+    },
+    {
+      "epoch": 0.457088,
+      "grad_norm": 0.4258294403553009,
+      "learning_rate": 0.0005932589063333668,
+      "loss": 2.9648,
+      "step": 3571
+    },
+    {
+      "epoch": 0.457216,
+      "grad_norm": 0.43920907378196716,
+      "learning_rate": 0.000593055225189638,
+      "loss": 3.011,
+      "step": 3572
+    },
+    {
+      "epoch": 0.457344,
+      "grad_norm": 0.41496169567108154,
+      "learning_rate": 0.0005928515280486372,
+      "loss": 2.9115,
+      "step": 3573
+    },
+    {
+      "epoch": 0.457472,
+      "grad_norm": 0.43396422266960144,
+      "learning_rate": 0.0005926478149453823,
+      "loss": 2.8709,
+      "step": 3574
+    },
+    {
+      "epoch": 0.4576,
+      "grad_norm": 0.40058958530426025,
+      "learning_rate": 0.000592444085914894,
+      "loss": 2.7886,
+      "step": 3575
+    },
+    {
+      "epoch": 0.457728,
+      "grad_norm": 0.41390982270240784,
+      "learning_rate": 0.0005922403409921956,
+      "loss": 2.9512,
+      "step": 3576
+    },
+    {
+      "epoch": 0.457856,
+      "grad_norm": 0.4247812032699585,
+      "learning_rate": 0.0005920365802123133,
+      "loss": 3.0104,
+      "step": 3577
+    },
+    {
+      "epoch": 0.457984,
+      "grad_norm": 0.4226842522621155,
+      "learning_rate": 0.0005918328036102758,
+      "loss": 2.8404,
+      "step": 3578
+    },
+    {
+      "epoch": 0.458112,
+      "grad_norm": 0.439054012298584,
+      "learning_rate": 0.000591629011221115,
+      "loss": 2.9438,
+      "step": 3579
+    },
+    {
+      "epoch": 0.45824,
+      "grad_norm": 0.43916329741477966,
+      "learning_rate": 0.0005914252030798645,
+      "loss": 2.7879,
+      "step": 3580
+    },
+    {
+      "epoch": 0.458368,
+      "grad_norm": 0.4155625104904175,
+      "learning_rate": 0.000591221379221562,
+      "loss": 2.7656,
+      "step": 3581
+    },
+    {
+      "epoch": 0.458496,
+      "grad_norm": 0.4175342917442322,
+      "learning_rate": 0.0005910175396812467,
+      "loss": 2.8551,
+      "step": 3582
+    },
+    {
+      "epoch": 0.458624,
+      "grad_norm": 0.43005234003067017,
+      "learning_rate": 0.0005908136844939611,
+      "loss": 2.976,
+      "step": 3583
+    },
+    {
+      "epoch": 0.458752,
+      "grad_norm": 0.42026224732398987,
+      "learning_rate": 0.0005906098136947507,
+      "loss": 2.8026,
+      "step": 3584
+    },
+    {
+      "epoch": 0.45888,
+      "grad_norm": 0.41256922483444214,
+      "learning_rate": 0.0005904059273186626,
+      "loss": 2.7778,
+      "step": 3585
+    },
+    {
+      "epoch": 0.459008,
+      "grad_norm": 0.4062056541442871,
+      "learning_rate": 0.0005902020254007475,
+      "loss": 2.877,
+      "step": 3586
+    },
+    {
+      "epoch": 0.459136,
+      "grad_norm": 0.42384034395217896,
+      "learning_rate": 0.0005899981079760585,
+      "loss": 2.902,
+      "step": 3587
+    },
+    {
+      "epoch": 0.459264,
+      "grad_norm": 0.43836331367492676,
+      "learning_rate": 0.0005897941750796516,
+      "loss": 2.9352,
+      "step": 3588
+    },
+    {
+      "epoch": 0.459392,
+      "grad_norm": 0.3880522549152374,
+      "learning_rate": 0.000589590226746585,
+      "loss": 2.8475,
+      "step": 3589
+    },
+    {
+      "epoch": 0.45952,
+      "grad_norm": 0.40906837582588196,
+      "learning_rate": 0.0005893862630119197,
+      "loss": 2.9878,
+      "step": 3590
+    },
+    {
+      "epoch": 0.459648,
+      "grad_norm": 0.4325112998485565,
+      "learning_rate": 0.0005891822839107195,
+      "loss": 2.9069,
+      "step": 3591
+    },
+    {
+      "epoch": 0.459776,
+      "grad_norm": 0.39466843008995056,
+      "learning_rate": 0.0005889782894780509,
+      "loss": 2.8846,
+      "step": 3592
+    },
+    {
+      "epoch": 0.459904,
+      "grad_norm": 0.4080761671066284,
+      "learning_rate": 0.0005887742797489828,
+      "loss": 2.9858,
+      "step": 3593
+    },
+    {
+      "epoch": 0.460032,
+      "grad_norm": 0.43036243319511414,
+      "learning_rate": 0.0005885702547585867,
+      "loss": 2.8364,
+      "step": 3594
+    },
+    {
+      "epoch": 0.46016,
+      "grad_norm": 0.41352003812789917,
+      "learning_rate": 0.0005883662145419373,
+      "loss": 2.9015,
+      "step": 3595
+    },
+    {
+      "epoch": 0.460288,
+      "grad_norm": 0.4091233015060425,
+      "learning_rate": 0.0005881621591341109,
+      "loss": 2.7017,
+      "step": 3596
+    },
+    {
+      "epoch": 0.460416,
+      "grad_norm": 0.4272184669971466,
+      "learning_rate": 0.0005879580885701874,
+      "loss": 2.8321,
+      "step": 3597
+    },
+    {
+      "epoch": 0.460544,
+      "grad_norm": 0.41143929958343506,
+      "learning_rate": 0.0005877540028852488,
+      "loss": 2.8074,
+      "step": 3598
+    },
+    {
+      "epoch": 0.460672,
+      "grad_norm": 0.4156115651130676,
+      "learning_rate": 0.0005875499021143799,
+      "loss": 2.9989,
+      "step": 3599
+    },
+    {
+      "epoch": 0.4608,
+      "grad_norm": 0.4126070737838745,
+      "learning_rate": 0.0005873457862926677,
+      "loss": 2.8501,
+      "step": 3600
+    },
+    {
+      "epoch": 0.460928,
+      "grad_norm": 0.45612066984176636,
+      "learning_rate": 0.0005871416554552021,
+      "loss": 2.9394,
+      "step": 3601
+    },
+    {
+      "epoch": 0.461056,
+      "grad_norm": 0.45564714074134827,
+      "learning_rate": 0.000586937509637076,
+      "loss": 2.781,
+      "step": 3602
+    },
+    {
+      "epoch": 0.461184,
+      "grad_norm": 0.5772602558135986,
+      "learning_rate": 0.0005867333488733839,
+      "loss": 2.8532,
+      "step": 3603
+    },
+    {
+      "epoch": 0.461312,
+      "grad_norm": 0.42484596371650696,
+      "learning_rate": 0.0005865291731992238,
+      "loss": 2.8652,
+      "step": 3604
+    },
+    {
+      "epoch": 0.46144,
+      "grad_norm": 0.46116140484809875,
+      "learning_rate": 0.0005863249826496955,
+      "loss": 2.9177,
+      "step": 3605
+    },
+    {
+      "epoch": 0.461568,
+      "grad_norm": 0.4043636620044708,
+      "learning_rate": 0.000586120777259902,
+      "loss": 2.9002,
+      "step": 3606
+    },
+    {
+      "epoch": 0.461696,
+      "grad_norm": 0.4285929799079895,
+      "learning_rate": 0.0005859165570649486,
+      "loss": 2.8899,
+      "step": 3607
+    },
+    {
+      "epoch": 0.461824,
+      "grad_norm": 0.4531913995742798,
+      "learning_rate": 0.0005857123220999428,
+      "loss": 2.8633,
+      "step": 3608
+    },
+    {
+      "epoch": 0.461952,
+      "grad_norm": 0.40930646657943726,
+      "learning_rate": 0.0005855080723999954,
+      "loss": 2.8921,
+      "step": 3609
+    },
+    {
+      "epoch": 0.46208,
+      "grad_norm": 0.4223979711532593,
+      "learning_rate": 0.0005853038080002188,
+      "loss": 2.7173,
+      "step": 3610
+    },
+    {
+      "epoch": 0.462208,
+      "grad_norm": 0.39840781688690186,
+      "learning_rate": 0.0005850995289357289,
+      "loss": 2.8299,
+      "step": 3611
+    },
+    {
+      "epoch": 0.462336,
+      "grad_norm": 0.4141884446144104,
+      "learning_rate": 0.0005848952352416433,
+      "loss": 2.8577,
+      "step": 3612
+    },
+    {
+      "epoch": 0.462464,
+      "grad_norm": 0.4698151648044586,
+      "learning_rate": 0.0005846909269530827,
+      "loss": 2.8521,
+      "step": 3613
+    },
+    {
+      "epoch": 0.462592,
+      "grad_norm": 0.46162110567092896,
+      "learning_rate": 0.0005844866041051698,
+      "loss": 2.9916,
+      "step": 3614
+    },
+    {
+      "epoch": 0.46272,
+      "grad_norm": 0.44259095191955566,
+      "learning_rate": 0.0005842822667330302,
+      "loss": 2.944,
+      "step": 3615
+    },
+    {
+      "epoch": 0.462848,
+      "grad_norm": 0.386894166469574,
+      "learning_rate": 0.0005840779148717919,
+      "loss": 2.8298,
+      "step": 3616
+    },
+    {
+      "epoch": 0.462976,
+      "grad_norm": 0.3865201771259308,
+      "learning_rate": 0.0005838735485565855,
+      "loss": 2.8634,
+      "step": 3617
+    },
+    {
+      "epoch": 0.463104,
+      "grad_norm": 0.4433255195617676,
+      "learning_rate": 0.0005836691678225434,
+      "loss": 2.7977,
+      "step": 3618
+    },
+    {
+      "epoch": 0.463232,
+      "grad_norm": 0.4138701558113098,
+      "learning_rate": 0.0005834647727048016,
+      "loss": 2.859,
+      "step": 3619
+    },
+    {
+      "epoch": 0.46336,
+      "grad_norm": 0.3968327045440674,
+      "learning_rate": 0.0005832603632384977,
+      "loss": 2.8758,
+      "step": 3620
+    },
+    {
+      "epoch": 0.463488,
+      "grad_norm": 0.38919568061828613,
+      "learning_rate": 0.0005830559394587724,
+      "loss": 2.6641,
+      "step": 3621
+    },
+    {
+      "epoch": 0.463616,
+      "grad_norm": 0.4490815997123718,
+      "learning_rate": 0.0005828515014007678,
+      "loss": 2.8086,
+      "step": 3622
+    },
+    {
+      "epoch": 0.463744,
+      "grad_norm": 0.41860833764076233,
+      "learning_rate": 0.0005826470490996298,
+      "loss": 2.923,
+      "step": 3623
+    },
+    {
+      "epoch": 0.463872,
+      "grad_norm": 0.42719566822052,
+      "learning_rate": 0.000582442582590506,
+      "loss": 2.8719,
+      "step": 3624
+    },
+    {
+      "epoch": 0.464,
+      "grad_norm": 0.4582633972167969,
+      "learning_rate": 0.0005822381019085465,
+      "loss": 2.8212,
+      "step": 3625
+    },
+    {
+      "epoch": 0.464128,
+      "grad_norm": 0.4350619614124298,
+      "learning_rate": 0.0005820336070889037,
+      "loss": 2.8647,
+      "step": 3626
+    },
+    {
+      "epoch": 0.464256,
+      "grad_norm": 0.4270089566707611,
+      "learning_rate": 0.0005818290981667332,
+      "loss": 2.8365,
+      "step": 3627
+    },
+    {
+      "epoch": 0.464384,
+      "grad_norm": 0.4216808080673218,
+      "learning_rate": 0.0005816245751771917,
+      "loss": 2.8615,
+      "step": 3628
+    },
+    {
+      "epoch": 0.464512,
+      "grad_norm": 0.41411131620407104,
+      "learning_rate": 0.0005814200381554397,
+      "loss": 3.0399,
+      "step": 3629
+    },
+    {
+      "epoch": 0.46464,
+      "grad_norm": 0.4174272119998932,
+      "learning_rate": 0.0005812154871366389,
+      "loss": 2.8492,
+      "step": 3630
+    },
+    {
+      "epoch": 0.464768,
+      "grad_norm": 0.41788309812545776,
+      "learning_rate": 0.0005810109221559548,
+      "loss": 2.8582,
+      "step": 3631
+    },
+    {
+      "epoch": 0.464896,
+      "grad_norm": 0.4298934042453766,
+      "learning_rate": 0.0005808063432485537,
+      "loss": 2.8732,
+      "step": 3632
+    },
+    {
+      "epoch": 0.465024,
+      "grad_norm": 0.44074392318725586,
+      "learning_rate": 0.0005806017504496054,
+      "loss": 2.8859,
+      "step": 3633
+    },
+    {
+      "epoch": 0.465152,
+      "grad_norm": 0.40001073479652405,
+      "learning_rate": 0.0005803971437942819,
+      "loss": 2.816,
+      "step": 3634
+    },
+    {
+      "epoch": 0.46528,
+      "grad_norm": 0.4387224018573761,
+      "learning_rate": 0.0005801925233177574,
+      "loss": 2.821,
+      "step": 3635
+    },
+    {
+      "epoch": 0.465408,
+      "grad_norm": 0.4188748896121979,
+      "learning_rate": 0.0005799878890552083,
+      "loss": 2.8795,
+      "step": 3636
+    },
+    {
+      "epoch": 0.465536,
+      "grad_norm": 0.429212361574173,
+      "learning_rate": 0.000579783241041814,
+      "loss": 2.8053,
+      "step": 3637
+    },
+    {
+      "epoch": 0.465664,
+      "grad_norm": 0.41584858298301697,
+      "learning_rate": 0.0005795785793127554,
+      "loss": 2.9381,
+      "step": 3638
+    },
+    {
+      "epoch": 0.465792,
+      "grad_norm": 0.4274567663669586,
+      "learning_rate": 0.0005793739039032165,
+      "loss": 2.9599,
+      "step": 3639
+    },
+    {
+      "epoch": 0.46592,
+      "grad_norm": 0.414823979139328,
+      "learning_rate": 0.0005791692148483834,
+      "loss": 2.9,
+      "step": 3640
+    },
+    {
+      "epoch": 0.466048,
+      "grad_norm": 0.4165601134300232,
+      "learning_rate": 0.0005789645121834445,
+      "loss": 2.8059,
+      "step": 3641
+    },
+    {
+      "epoch": 0.466176,
+      "grad_norm": 0.44540393352508545,
+      "learning_rate": 0.0005787597959435904,
+      "loss": 2.9516,
+      "step": 3642
+    },
+    {
+      "epoch": 0.466304,
+      "grad_norm": 0.49474748969078064,
+      "learning_rate": 0.0005785550661640144,
+      "loss": 2.8624,
+      "step": 3643
+    },
+    {
+      "epoch": 0.466432,
+      "grad_norm": 0.4183148741722107,
+      "learning_rate": 0.0005783503228799119,
+      "loss": 2.8433,
+      "step": 3644
+    },
+    {
+      "epoch": 0.46656,
+      "grad_norm": 0.413156658411026,
+      "learning_rate": 0.0005781455661264805,
+      "loss": 2.9225,
+      "step": 3645
+    },
+    {
+      "epoch": 0.466688,
+      "grad_norm": 0.4380357563495636,
+      "learning_rate": 0.0005779407959389204,
+      "loss": 2.9118,
+      "step": 3646
+    },
+    {
+      "epoch": 0.466816,
+      "grad_norm": 0.4428248703479767,
+      "learning_rate": 0.0005777360123524338,
+      "loss": 2.8933,
+      "step": 3647
+    },
+    {
+      "epoch": 0.466944,
+      "grad_norm": 0.4294101893901825,
+      "learning_rate": 0.0005775312154022255,
+      "loss": 2.9198,
+      "step": 3648
+    },
+    {
+      "epoch": 0.467072,
+      "grad_norm": 0.46569275856018066,
+      "learning_rate": 0.0005773264051235025,
+      "loss": 2.9651,
+      "step": 3649
+    },
+    {
+      "epoch": 0.4672,
+      "grad_norm": 0.4269278347492218,
+      "learning_rate": 0.0005771215815514741,
+      "loss": 2.9207,
+      "step": 3650
+    },
+    {
+      "epoch": 0.467328,
+      "grad_norm": 0.41563764214515686,
+      "learning_rate": 0.0005769167447213517,
+      "loss": 2.7876,
+      "step": 3651
+    },
+    {
+      "epoch": 0.467456,
+      "grad_norm": 0.43918082118034363,
+      "learning_rate": 0.0005767118946683491,
+      "loss": 2.9453,
+      "step": 3652
+    },
+    {
+      "epoch": 0.467584,
+      "grad_norm": 0.438470721244812,
+      "learning_rate": 0.0005765070314276826,
+      "loss": 2.8767,
+      "step": 3653
+    },
+    {
+      "epoch": 0.467712,
+      "grad_norm": 0.4370444416999817,
+      "learning_rate": 0.0005763021550345703,
+      "loss": 2.9,
+      "step": 3654
+    },
+    {
+      "epoch": 0.46784,
+      "grad_norm": 0.40168601274490356,
+      "learning_rate": 0.0005760972655242332,
+      "loss": 2.9332,
+      "step": 3655
+    },
+    {
+      "epoch": 0.467968,
+      "grad_norm": 0.39074862003326416,
+      "learning_rate": 0.0005758923629318937,
+      "loss": 2.9311,
+      "step": 3656
+    },
+    {
+      "epoch": 0.468096,
+      "grad_norm": 0.4057040810585022,
+      "learning_rate": 0.0005756874472927774,
+      "loss": 2.7992,
+      "step": 3657
+    },
+    {
+      "epoch": 0.468224,
+      "grad_norm": 0.40989139676094055,
+      "learning_rate": 0.0005754825186421113,
+      "loss": 2.9529,
+      "step": 3658
+    },
+    {
+      "epoch": 0.468352,
+      "grad_norm": 0.4117618203163147,
+      "learning_rate": 0.0005752775770151254,
+      "loss": 2.7977,
+      "step": 3659
+    },
+    {
+      "epoch": 0.46848,
+      "grad_norm": 0.43160802125930786,
+      "learning_rate": 0.0005750726224470513,
+      "loss": 2.8989,
+      "step": 3660
+    },
+    {
+      "epoch": 0.468608,
+      "grad_norm": 0.4618944823741913,
+      "learning_rate": 0.0005748676549731229,
+      "loss": 2.9043,
+      "step": 3661
+    },
+    {
+      "epoch": 0.468736,
+      "grad_norm": 0.4192894697189331,
+      "learning_rate": 0.0005746626746285767,
+      "loss": 2.8465,
+      "step": 3662
+    },
+    {
+      "epoch": 0.468864,
+      "grad_norm": 0.43456822633743286,
+      "learning_rate": 0.0005744576814486512,
+      "loss": 2.9878,
+      "step": 3663
+    },
+    {
+      "epoch": 0.468992,
+      "grad_norm": 0.41404667496681213,
+      "learning_rate": 0.0005742526754685871,
+      "loss": 2.8632,
+      "step": 3664
+    },
+    {
+      "epoch": 0.46912,
+      "grad_norm": 0.40751415491104126,
+      "learning_rate": 0.0005740476567236272,
+      "loss": 2.9249,
+      "step": 3665
+    },
+    {
+      "epoch": 0.469248,
+      "grad_norm": 0.4041231870651245,
+      "learning_rate": 0.0005738426252490168,
+      "loss": 2.8503,
+      "step": 3666
+    },
+    {
+      "epoch": 0.469376,
+      "grad_norm": 0.39410725235939026,
+      "learning_rate": 0.000573637581080003,
+      "loss": 2.8132,
+      "step": 3667
+    },
+    {
+      "epoch": 0.469504,
+      "grad_norm": 0.44120046496391296,
+      "learning_rate": 0.0005734325242518354,
+      "loss": 2.8266,
+      "step": 3668
+    },
+    {
+      "epoch": 0.469632,
+      "grad_norm": 0.40708616375923157,
+      "learning_rate": 0.0005732274547997655,
+      "loss": 2.9864,
+      "step": 3669
+    },
+    {
+      "epoch": 0.46976,
+      "grad_norm": 0.40904590487480164,
+      "learning_rate": 0.0005730223727590471,
+      "loss": 2.869,
+      "step": 3670
+    },
+    {
+      "epoch": 0.469888,
+      "grad_norm": 0.4124044179916382,
+      "learning_rate": 0.0005728172781649365,
+      "loss": 2.8352,
+      "step": 3671
+    },
+    {
+      "epoch": 0.470016,
+      "grad_norm": 0.452443927526474,
+      "learning_rate": 0.0005726121710526913,
+      "loss": 2.9534,
+      "step": 3672
+    },
+    {
+      "epoch": 0.470144,
+      "grad_norm": 0.4239620268344879,
+      "learning_rate": 0.0005724070514575724,
+      "loss": 2.8462,
+      "step": 3673
+    },
+    {
+      "epoch": 0.470272,
+      "grad_norm": 0.4175872802734375,
+      "learning_rate": 0.0005722019194148419,
+      "loss": 2.9223,
+      "step": 3674
+    },
+    {
+      "epoch": 0.4704,
+      "grad_norm": 0.4487774968147278,
+      "learning_rate": 0.0005719967749597643,
+      "loss": 2.8518,
+      "step": 3675
+    },
+    {
+      "epoch": 0.470528,
+      "grad_norm": 0.44167691469192505,
+      "learning_rate": 0.0005717916181276065,
+      "loss": 2.8375,
+      "step": 3676
+    },
+    {
+      "epoch": 0.470656,
+      "grad_norm": 0.4504028558731079,
+      "learning_rate": 0.0005715864489536372,
+      "loss": 2.9814,
+      "step": 3677
+    },
+    {
+      "epoch": 0.470784,
+      "grad_norm": 0.4219740033149719,
+      "learning_rate": 0.0005713812674731276,
+      "loss": 2.9157,
+      "step": 3678
+    },
+    {
+      "epoch": 0.470912,
+      "grad_norm": 0.4269733428955078,
+      "learning_rate": 0.0005711760737213504,
+      "loss": 2.7748,
+      "step": 3679
+    },
+    {
+      "epoch": 0.47104,
+      "grad_norm": 0.782888650894165,
+      "learning_rate": 0.000570970867733581,
+      "loss": 2.9569,
+      "step": 3680
+    },
+    {
+      "epoch": 0.471168,
+      "grad_norm": 0.43998897075653076,
+      "learning_rate": 0.0005707656495450968,
+      "loss": 2.8815,
+      "step": 3681
+    },
+    {
+      "epoch": 0.471296,
+      "grad_norm": 0.4469761252403259,
+      "learning_rate": 0.000570560419191177,
+      "loss": 3.0642,
+      "step": 3682
+    },
+    {
+      "epoch": 0.471424,
+      "grad_norm": 0.4116871953010559,
+      "learning_rate": 0.000570355176707103,
+      "loss": 2.849,
+      "step": 3683
+    },
+    {
+      "epoch": 0.471552,
+      "grad_norm": 0.4185531735420227,
+      "learning_rate": 0.0005701499221281586,
+      "loss": 2.8738,
+      "step": 3684
+    },
+    {
+      "epoch": 0.47168,
+      "grad_norm": 0.44185057282447815,
+      "learning_rate": 0.0005699446554896293,
+      "loss": 3.0521,
+      "step": 3685
+    },
+    {
+      "epoch": 0.471808,
+      "grad_norm": 0.4274960458278656,
+      "learning_rate": 0.0005697393768268031,
+      "loss": 2.9439,
+      "step": 3686
+    },
+    {
+      "epoch": 0.471936,
+      "grad_norm": 0.4229310154914856,
+      "learning_rate": 0.0005695340861749693,
+      "loss": 2.7986,
+      "step": 3687
+    },
+    {
+      "epoch": 0.472064,
+      "grad_norm": 0.4640122950077057,
+      "learning_rate": 0.00056932878356942,
+      "loss": 2.7673,
+      "step": 3688
+    },
+    {
+      "epoch": 0.472192,
+      "grad_norm": 0.4461619555950165,
+      "learning_rate": 0.0005691234690454491,
+      "loss": 2.8881,
+      "step": 3689
+    },
+    {
+      "epoch": 0.47232,
+      "grad_norm": 0.4353177845478058,
+      "learning_rate": 0.0005689181426383524,
+      "loss": 2.873,
+      "step": 3690
+    },
+    {
+      "epoch": 0.472448,
+      "grad_norm": 0.40220609307289124,
+      "learning_rate": 0.0005687128043834282,
+      "loss": 2.7787,
+      "step": 3691
+    },
+    {
+      "epoch": 0.472576,
+      "grad_norm": 0.43609854578971863,
+      "learning_rate": 0.0005685074543159763,
+      "loss": 2.958,
+      "step": 3692
+    },
+    {
+      "epoch": 0.472704,
+      "grad_norm": 0.46478235721588135,
+      "learning_rate": 0.0005683020924712988,
+      "loss": 2.8851,
+      "step": 3693
+    },
+    {
+      "epoch": 0.472832,
+      "grad_norm": 0.4217210114002228,
+      "learning_rate": 0.0005680967188846996,
+      "loss": 2.9594,
+      "step": 3694
+    },
+    {
+      "epoch": 0.47296,
+      "grad_norm": 0.42570334672927856,
+      "learning_rate": 0.0005678913335914851,
+      "loss": 3.0121,
+      "step": 3695
+    },
+    {
+      "epoch": 0.473088,
+      "grad_norm": 0.3829468786716461,
+      "learning_rate": 0.0005676859366269634,
+      "loss": 2.8441,
+      "step": 3696
+    },
+    {
+      "epoch": 0.473216,
+      "grad_norm": 0.41923388838768005,
+      "learning_rate": 0.0005674805280264445,
+      "loss": 2.836,
+      "step": 3697
+    },
+    {
+      "epoch": 0.473344,
+      "grad_norm": 0.41490986943244934,
+      "learning_rate": 0.0005672751078252403,
+      "loss": 2.7941,
+      "step": 3698
+    },
+    {
+      "epoch": 0.473472,
+      "grad_norm": 0.45712101459503174,
+      "learning_rate": 0.0005670696760586653,
+      "loss": 2.9892,
+      "step": 3699
+    },
+    {
+      "epoch": 0.4736,
+      "grad_norm": 0.4021947383880615,
+      "learning_rate": 0.0005668642327620356,
+      "loss": 2.7512,
+      "step": 3700
+    },
+    {
+      "epoch": 0.473728,
+      "grad_norm": 0.4418056011199951,
+      "learning_rate": 0.000566658777970669,
+      "loss": 2.7876,
+      "step": 3701
+    },
+    {
+      "epoch": 0.473856,
+      "grad_norm": 0.44348230957984924,
+      "learning_rate": 0.0005664533117198855,
+      "loss": 2.8727,
+      "step": 3702
+    },
+    {
+      "epoch": 0.473984,
+      "grad_norm": 0.45528340339660645,
+      "learning_rate": 0.0005662478340450076,
+      "loss": 2.6582,
+      "step": 3703
+    },
+    {
+      "epoch": 0.474112,
+      "grad_norm": 0.42434951663017273,
+      "learning_rate": 0.0005660423449813588,
+      "loss": 2.9565,
+      "step": 3704
+    },
+    {
+      "epoch": 0.47424,
+      "grad_norm": 0.4066430330276489,
+      "learning_rate": 0.0005658368445642654,
+      "loss": 2.8852,
+      "step": 3705
+    },
+    {
+      "epoch": 0.474368,
+      "grad_norm": 0.40451741218566895,
+      "learning_rate": 0.0005656313328290553,
+      "loss": 2.8334,
+      "step": 3706
+    },
+    {
+      "epoch": 0.474496,
+      "grad_norm": 0.421687513589859,
+      "learning_rate": 0.000565425809811058,
+      "loss": 2.8693,
+      "step": 3707
+    },
+    {
+      "epoch": 0.474624,
+      "grad_norm": 0.417242169380188,
+      "learning_rate": 0.0005652202755456055,
+      "loss": 2.9742,
+      "step": 3708
+    },
+    {
+      "epoch": 0.474752,
+      "grad_norm": 0.40506261587142944,
+      "learning_rate": 0.0005650147300680317,
+      "loss": 2.8538,
+      "step": 3709
+    },
+    {
+      "epoch": 0.47488,
+      "grad_norm": 0.4125186502933502,
+      "learning_rate": 0.000564809173413672,
+      "loss": 2.6665,
+      "step": 3710
+    },
+    {
+      "epoch": 0.475008,
+      "grad_norm": 0.42508620023727417,
+      "learning_rate": 0.000564603605617864,
+      "loss": 2.9342,
+      "step": 3711
+    },
+    {
+      "epoch": 0.475136,
+      "grad_norm": 0.4112227261066437,
+      "learning_rate": 0.0005643980267159473,
+      "loss": 2.9288,
+      "step": 3712
+    },
+    {
+      "epoch": 0.475264,
+      "grad_norm": 0.42235544323921204,
+      "learning_rate": 0.0005641924367432633,
+      "loss": 3.0214,
+      "step": 3713
+    },
+    {
+      "epoch": 0.475392,
+      "grad_norm": 0.40585485100746155,
+      "learning_rate": 0.0005639868357351553,
+      "loss": 2.6927,
+      "step": 3714
+    },
+    {
+      "epoch": 0.47552,
+      "grad_norm": 0.4484511911869049,
+      "learning_rate": 0.0005637812237269682,
+      "loss": 2.9765,
+      "step": 3715
+    },
+    {
+      "epoch": 0.475648,
+      "grad_norm": 0.41545403003692627,
+      "learning_rate": 0.0005635756007540494,
+      "loss": 2.9616,
+      "step": 3716
+    },
+    {
+      "epoch": 0.475776,
+      "grad_norm": 0.4827587902545929,
+      "learning_rate": 0.000563369966851748,
+      "loss": 2.8276,
+      "step": 3717
+    },
+    {
+      "epoch": 0.475904,
+      "grad_norm": 0.39885735511779785,
+      "learning_rate": 0.0005631643220554146,
+      "loss": 2.8653,
+      "step": 3718
+    },
+    {
+      "epoch": 0.476032,
+      "grad_norm": 0.4135931432247162,
+      "learning_rate": 0.0005629586664004017,
+      "loss": 2.8406,
+      "step": 3719
+    },
+    {
+      "epoch": 0.47616,
+      "grad_norm": 0.4833976626396179,
+      "learning_rate": 0.0005627529999220644,
+      "loss": 2.8837,
+      "step": 3720
+    },
+    {
+      "epoch": 0.476288,
+      "grad_norm": 0.4005281925201416,
+      "learning_rate": 0.0005625473226557588,
+      "loss": 2.8284,
+      "step": 3721
+    },
+    {
+      "epoch": 0.476416,
+      "grad_norm": 0.4236091077327728,
+      "learning_rate": 0.0005623416346368434,
+      "loss": 2.873,
+      "step": 3722
+    },
+    {
+      "epoch": 0.476544,
+      "grad_norm": 0.41567978262901306,
+      "learning_rate": 0.0005621359359006783,
+      "loss": 2.8435,
+      "step": 3723
+    },
+    {
+      "epoch": 0.476672,
+      "grad_norm": 0.4554838538169861,
+      "learning_rate": 0.0005619302264826255,
+      "loss": 2.9096,
+      "step": 3724
+    },
+    {
+      "epoch": 0.4768,
+      "grad_norm": 0.41543591022491455,
+      "learning_rate": 0.0005617245064180488,
+      "loss": 2.9032,
+      "step": 3725
+    },
+    {
+      "epoch": 0.476928,
+      "grad_norm": 0.4102315306663513,
+      "learning_rate": 0.0005615187757423137,
+      "loss": 2.7759,
+      "step": 3726
+    },
+    {
+      "epoch": 0.477056,
+      "grad_norm": 0.4458465576171875,
+      "learning_rate": 0.0005613130344907881,
+      "loss": 2.7716,
+      "step": 3727
+    },
+    {
+      "epoch": 0.477184,
+      "grad_norm": 0.43334054946899414,
+      "learning_rate": 0.0005611072826988409,
+      "loss": 2.8441,
+      "step": 3728
+    },
+    {
+      "epoch": 0.477312,
+      "grad_norm": 0.42163076996803284,
+      "learning_rate": 0.0005609015204018435,
+      "loss": 2.7789,
+      "step": 3729
+    },
+    {
+      "epoch": 0.47744,
+      "grad_norm": 0.4216097891330719,
+      "learning_rate": 0.0005606957476351687,
+      "loss": 2.9061,
+      "step": 3730
+    },
+    {
+      "epoch": 0.477568,
+      "grad_norm": 0.4248703718185425,
+      "learning_rate": 0.0005604899644341911,
+      "loss": 2.8007,
+      "step": 3731
+    },
+    {
+      "epoch": 0.477696,
+      "grad_norm": 0.4515250027179718,
+      "learning_rate": 0.0005602841708342876,
+      "loss": 2.8827,
+      "step": 3732
+    },
+    {
+      "epoch": 0.477824,
+      "grad_norm": 0.4271150827407837,
+      "learning_rate": 0.000560078366870836,
+      "loss": 3.0033,
+      "step": 3733
+    },
+    {
+      "epoch": 0.477952,
+      "grad_norm": 0.4050045311450958,
+      "learning_rate": 0.0005598725525792168,
+      "loss": 2.9193,
+      "step": 3734
+    },
+    {
+      "epoch": 0.47808,
+      "grad_norm": 0.40167683362960815,
+      "learning_rate": 0.0005596667279948117,
+      "loss": 2.9472,
+      "step": 3735
+    },
+    {
+      "epoch": 0.478208,
+      "grad_norm": 0.40627536177635193,
+      "learning_rate": 0.0005594608931530043,
+      "loss": 2.8531,
+      "step": 3736
+    },
+    {
+      "epoch": 0.478336,
+      "grad_norm": 0.4213855564594269,
+      "learning_rate": 0.00055925504808918,
+      "loss": 2.9051,
+      "step": 3737
+    },
+    {
+      "epoch": 0.478464,
+      "grad_norm": 0.4286533296108246,
+      "learning_rate": 0.0005590491928387262,
+      "loss": 2.844,
+      "step": 3738
+    },
+    {
+      "epoch": 0.478592,
+      "grad_norm": 0.39365431666374207,
+      "learning_rate": 0.0005588433274370313,
+      "loss": 2.8619,
+      "step": 3739
+    },
+    {
+      "epoch": 0.47872,
+      "grad_norm": 0.39499586820602417,
+      "learning_rate": 0.0005586374519194864,
+      "loss": 2.9414,
+      "step": 3740
+    },
+    {
+      "epoch": 0.478848,
+      "grad_norm": 0.42694416642189026,
+      "learning_rate": 0.0005584315663214836,
+      "loss": 2.8472,
+      "step": 3741
+    },
+    {
+      "epoch": 0.478976,
+      "grad_norm": 0.44388580322265625,
+      "learning_rate": 0.0005582256706784173,
+      "loss": 2.9024,
+      "step": 3742
+    },
+    {
+      "epoch": 0.479104,
+      "grad_norm": 0.4290860593318939,
+      "learning_rate": 0.0005580197650256832,
+      "loss": 2.8635,
+      "step": 3743
+    },
+    {
+      "epoch": 0.479232,
+      "grad_norm": 0.4921223819255829,
+      "learning_rate": 0.0005578138493986788,
+      "loss": 2.8362,
+      "step": 3744
+    },
+    {
+      "epoch": 0.47936,
+      "grad_norm": 0.4198877215385437,
+      "learning_rate": 0.0005576079238328034,
+      "loss": 2.9563,
+      "step": 3745
+    },
+    {
+      "epoch": 0.479488,
+      "grad_norm": 0.42848506569862366,
+      "learning_rate": 0.0005574019883634583,
+      "loss": 3.0245,
+      "step": 3746
+    },
+    {
+      "epoch": 0.479616,
+      "grad_norm": 0.4441917836666107,
+      "learning_rate": 0.0005571960430260457,
+      "loss": 2.8953,
+      "step": 3747
+    },
+    {
+      "epoch": 0.479744,
+      "grad_norm": 0.42467615008354187,
+      "learning_rate": 0.0005569900878559704,
+      "loss": 3.0101,
+      "step": 3748
+    },
+    {
+      "epoch": 0.479872,
+      "grad_norm": 0.4447496831417084,
+      "learning_rate": 0.0005567841228886381,
+      "loss": 2.8798,
+      "step": 3749
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4200349450111389,
+      "learning_rate": 0.000556578148159457,
+      "loss": 2.9416,
+      "step": 3750
+    },
+    {
+      "epoch": 0.480128,
+      "grad_norm": 0.4109404683113098,
+      "learning_rate": 0.0005563721637038362,
+      "loss": 2.9749,
+      "step": 3751
+    },
+    {
+      "epoch": 0.480256,
+      "grad_norm": 0.4231906533241272,
+      "learning_rate": 0.000556166169557187,
+      "loss": 2.9712,
+      "step": 3752
+    },
+    {
+      "epoch": 0.480384,
+      "grad_norm": 0.5183878540992737,
+      "learning_rate": 0.0005559601657549219,
+      "loss": 2.8872,
+      "step": 3753
+    },
+    {
+      "epoch": 0.480512,
+      "grad_norm": 0.4149218499660492,
+      "learning_rate": 0.0005557541523324558,
+      "loss": 2.9084,
+      "step": 3754
+    },
+    {
+      "epoch": 0.48064,
+      "grad_norm": 0.40129321813583374,
+      "learning_rate": 0.0005555481293252044,
+      "loss": 2.8683,
+      "step": 3755
+    },
+    {
+      "epoch": 0.480768,
+      "grad_norm": 0.4166179299354553,
+      "learning_rate": 0.0005553420967685857,
+      "loss": 2.9641,
+      "step": 3756
+    },
+    {
+      "epoch": 0.480896,
+      "grad_norm": 0.41869688034057617,
+      "learning_rate": 0.0005551360546980191,
+      "loss": 2.922,
+      "step": 3757
+    },
+    {
+      "epoch": 0.481024,
+      "grad_norm": 0.4644629955291748,
+      "learning_rate": 0.0005549300031489252,
+      "loss": 2.8937,
+      "step": 3758
+    },
+    {
+      "epoch": 0.481152,
+      "grad_norm": 0.41291892528533936,
+      "learning_rate": 0.0005547239421567271,
+      "loss": 2.9428,
+      "step": 3759
+    },
+    {
+      "epoch": 0.48128,
+      "grad_norm": 0.4388875365257263,
+      "learning_rate": 0.0005545178717568491,
+      "loss": 2.901,
+      "step": 3760
+    },
+    {
+      "epoch": 0.481408,
+      "grad_norm": 0.43995437026023865,
+      "learning_rate": 0.0005543117919847167,
+      "loss": 2.8591,
+      "step": 3761
+    },
+    {
+      "epoch": 0.481536,
+      "grad_norm": 0.4281708300113678,
+      "learning_rate": 0.0005541057028757576,
+      "loss": 2.9896,
+      "step": 3762
+    },
+    {
+      "epoch": 0.481664,
+      "grad_norm": 0.44005072116851807,
+      "learning_rate": 0.0005538996044654011,
+      "loss": 2.902,
+      "step": 3763
+    },
+    {
+      "epoch": 0.481792,
+      "grad_norm": 0.45427653193473816,
+      "learning_rate": 0.0005536934967890777,
+      "loss": 2.9634,
+      "step": 3764
+    },
+    {
+      "epoch": 0.48192,
+      "grad_norm": 0.40349259972572327,
+      "learning_rate": 0.0005534873798822196,
+      "loss": 2.893,
+      "step": 3765
+    },
+    {
+      "epoch": 0.482048,
+      "grad_norm": 0.4567594826221466,
+      "learning_rate": 0.0005532812537802611,
+      "loss": 2.8903,
+      "step": 3766
+    },
+    {
+      "epoch": 0.482176,
+      "grad_norm": 0.4455542266368866,
+      "learning_rate": 0.0005530751185186372,
+      "loss": 2.874,
+      "step": 3767
+    },
+    {
+      "epoch": 0.482304,
+      "grad_norm": 0.4491727948188782,
+      "learning_rate": 0.000552868974132785,
+      "loss": 2.9239,
+      "step": 3768
+    },
+    {
+      "epoch": 0.482432,
+      "grad_norm": 0.4227888584136963,
+      "learning_rate": 0.0005526628206581433,
+      "loss": 2.9997,
+      "step": 3769
+    },
+    {
+      "epoch": 0.48256,
+      "grad_norm": 0.3753412365913391,
+      "learning_rate": 0.0005524566581301525,
+      "loss": 2.8596,
+      "step": 3770
+    },
+    {
+      "epoch": 0.482688,
+      "grad_norm": 0.4268723130226135,
+      "learning_rate": 0.0005522504865842537,
+      "loss": 2.8141,
+      "step": 3771
+    },
+    {
+      "epoch": 0.482816,
+      "grad_norm": 0.4235062003135681,
+      "learning_rate": 0.0005520443060558906,
+      "loss": 2.8668,
+      "step": 3772
+    },
+    {
+      "epoch": 0.482944,
+      "grad_norm": 0.41288235783576965,
+      "learning_rate": 0.000551838116580508,
+      "loss": 3.0702,
+      "step": 3773
+    },
+    {
+      "epoch": 0.483072,
+      "grad_norm": 0.43362438678741455,
+      "learning_rate": 0.0005516319181935521,
+      "loss": 2.8772,
+      "step": 3774
+    },
+    {
+      "epoch": 0.4832,
+      "grad_norm": 0.42318612337112427,
+      "learning_rate": 0.0005514257109304711,
+      "loss": 2.7699,
+      "step": 3775
+    },
+    {
+      "epoch": 0.483328,
+      "grad_norm": 0.4197331666946411,
+      "learning_rate": 0.0005512194948267139,
+      "loss": 2.819,
+      "step": 3776
+    },
+    {
+      "epoch": 0.483456,
+      "grad_norm": 0.4580160081386566,
+      "learning_rate": 0.000551013269917732,
+      "loss": 2.9133,
+      "step": 3777
+    },
+    {
+      "epoch": 0.483584,
+      "grad_norm": 0.4434983730316162,
+      "learning_rate": 0.0005508070362389776,
+      "loss": 2.9374,
+      "step": 3778
+    },
+    {
+      "epoch": 0.483712,
+      "grad_norm": 0.462864488363266,
+      "learning_rate": 0.0005506007938259045,
+      "loss": 2.8133,
+      "step": 3779
+    },
+    {
+      "epoch": 0.48384,
+      "grad_norm": 0.4012368321418762,
+      "learning_rate": 0.0005503945427139684,
+      "loss": 2.7765,
+      "step": 3780
+    },
+    {
+      "epoch": 0.483968,
+      "grad_norm": 0.41424453258514404,
+      "learning_rate": 0.0005501882829386263,
+      "loss": 2.8939,
+      "step": 3781
+    },
+    {
+      "epoch": 0.484096,
+      "grad_norm": 0.43147045373916626,
+      "learning_rate": 0.0005499820145353363,
+      "loss": 2.8254,
+      "step": 3782
+    },
+    {
+      "epoch": 0.484224,
+      "grad_norm": 0.4190748333930969,
+      "learning_rate": 0.0005497757375395588,
+      "loss": 2.8132,
+      "step": 3783
+    },
+    {
+      "epoch": 0.484352,
+      "grad_norm": 0.4067164361476898,
+      "learning_rate": 0.0005495694519867549,
+      "loss": 2.98,
+      "step": 3784
+    },
+    {
+      "epoch": 0.48448,
+      "grad_norm": 0.5691153407096863,
+      "learning_rate": 0.0005493631579123876,
+      "loss": 2.9976,
+      "step": 3785
+    },
+    {
+      "epoch": 0.484608,
+      "grad_norm": 0.4473784267902374,
+      "learning_rate": 0.000549156855351921,
+      "loss": 2.923,
+      "step": 3786
+    },
+    {
+      "epoch": 0.484736,
+      "grad_norm": 0.42379269003868103,
+      "learning_rate": 0.0005489505443408211,
+      "loss": 2.9549,
+      "step": 3787
+    },
+    {
+      "epoch": 0.484864,
+      "grad_norm": 0.4310116171836853,
+      "learning_rate": 0.0005487442249145553,
+      "loss": 2.9143,
+      "step": 3788
+    },
+    {
+      "epoch": 0.484992,
+      "grad_norm": 0.41097840666770935,
+      "learning_rate": 0.0005485378971085921,
+      "loss": 2.9096,
+      "step": 3789
+    },
+    {
+      "epoch": 0.48512,
+      "grad_norm": 0.441872775554657,
+      "learning_rate": 0.0005483315609584016,
+      "loss": 2.7048,
+      "step": 3790
+    },
+    {
+      "epoch": 0.485248,
+      "grad_norm": 0.40166202187538147,
+      "learning_rate": 0.0005481252164994554,
+      "loss": 2.8175,
+      "step": 3791
+    },
+    {
+      "epoch": 0.485376,
+      "grad_norm": 0.45694923400878906,
+      "learning_rate": 0.0005479188637672269,
+      "loss": 2.9425,
+      "step": 3792
+    },
+    {
+      "epoch": 0.485504,
+      "grad_norm": 0.444218248128891,
+      "learning_rate": 0.00054771250279719,
+      "loss": 2.9537,
+      "step": 3793
+    },
+    {
+      "epoch": 0.485632,
+      "grad_norm": 0.4161939024925232,
+      "learning_rate": 0.0005475061336248207,
+      "loss": 2.9197,
+      "step": 3794
+    },
+    {
+      "epoch": 0.48576,
+      "grad_norm": 0.3976356089115143,
+      "learning_rate": 0.0005472997562855962,
+      "loss": 2.7349,
+      "step": 3795
+    },
+    {
+      "epoch": 0.485888,
+      "grad_norm": 0.46349263191223145,
+      "learning_rate": 0.0005470933708149955,
+      "loss": 2.8086,
+      "step": 3796
+    },
+    {
+      "epoch": 0.486016,
+      "grad_norm": 0.4093191921710968,
+      "learning_rate": 0.0005468869772484982,
+      "loss": 2.8282,
+      "step": 3797
+    },
+    {
+      "epoch": 0.486144,
+      "grad_norm": 0.41616398096084595,
+      "learning_rate": 0.000546680575621586,
+      "loss": 2.8136,
+      "step": 3798
+    },
+    {
+      "epoch": 0.486272,
+      "grad_norm": 0.4176335036754608,
+      "learning_rate": 0.0005464741659697419,
+      "loss": 2.87,
+      "step": 3799
+    },
+    {
+      "epoch": 0.4864,
+      "grad_norm": 0.39962151646614075,
+      "learning_rate": 0.0005462677483284496,
+      "loss": 2.8556,
+      "step": 3800
+    },
+    {
+      "epoch": 0.486528,
+      "grad_norm": 0.4289792478084564,
+      "learning_rate": 0.000546061322733195,
+      "loss": 2.9174,
+      "step": 3801
+    },
+    {
+      "epoch": 0.486656,
+      "grad_norm": 0.42548510432243347,
+      "learning_rate": 0.0005458548892194651,
+      "loss": 2.8559,
+      "step": 3802
+    },
+    {
+      "epoch": 0.486784,
+      "grad_norm": 0.4158097207546234,
+      "learning_rate": 0.0005456484478227482,
+      "loss": 2.8408,
+      "step": 3803
+    },
+    {
+      "epoch": 0.486912,
+      "grad_norm": 0.4012371599674225,
+      "learning_rate": 0.0005454419985785339,
+      "loss": 2.8845,
+      "step": 3804
+    },
+    {
+      "epoch": 0.48704,
+      "grad_norm": 0.4181024134159088,
+      "learning_rate": 0.0005452355415223131,
+      "loss": 2.9019,
+      "step": 3805
+    },
+    {
+      "epoch": 0.487168,
+      "grad_norm": 0.42169177532196045,
+      "learning_rate": 0.0005450290766895786,
+      "loss": 2.9073,
+      "step": 3806
+    },
+    {
+      "epoch": 0.487296,
+      "grad_norm": 0.3945174515247345,
+      "learning_rate": 0.0005448226041158236,
+      "loss": 2.7512,
+      "step": 3807
+    },
+    {
+      "epoch": 0.487424,
+      "grad_norm": 0.4113897979259491,
+      "learning_rate": 0.0005446161238365433,
+      "loss": 2.8945,
+      "step": 3808
+    },
+    {
+      "epoch": 0.487552,
+      "grad_norm": 0.4397762417793274,
+      "learning_rate": 0.0005444096358872341,
+      "loss": 2.8712,
+      "step": 3809
+    },
+    {
+      "epoch": 0.48768,
+      "grad_norm": 0.42273834347724915,
+      "learning_rate": 0.0005442031403033937,
+      "loss": 2.7011,
+      "step": 3810
+    },
+    {
+      "epoch": 0.487808,
+      "grad_norm": 0.4095867872238159,
+      "learning_rate": 0.0005439966371205209,
+      "loss": 2.6786,
+      "step": 3811
+    },
+    {
+      "epoch": 0.487936,
+      "grad_norm": 0.4808022081851959,
+      "learning_rate": 0.0005437901263741163,
+      "loss": 2.8409,
+      "step": 3812
+    },
+    {
+      "epoch": 0.488064,
+      "grad_norm": 0.4302924871444702,
+      "learning_rate": 0.0005435836080996812,
+      "loss": 2.819,
+      "step": 3813
+    },
+    {
+      "epoch": 0.488192,
+      "grad_norm": 0.4311966300010681,
+      "learning_rate": 0.0005433770823327186,
+      "loss": 2.972,
+      "step": 3814
+    },
+    {
+      "epoch": 0.48832,
+      "grad_norm": 0.41902199387550354,
+      "learning_rate": 0.0005431705491087326,
+      "loss": 2.9028,
+      "step": 3815
+    },
+    {
+      "epoch": 0.488448,
+      "grad_norm": 0.42489707469940186,
+      "learning_rate": 0.000542964008463229,
+      "loss": 2.7051,
+      "step": 3816
+    },
+    {
+      "epoch": 0.488576,
+      "grad_norm": 0.4083309471607208,
+      "learning_rate": 0.0005427574604317142,
+      "loss": 2.814,
+      "step": 3817
+    },
+    {
+      "epoch": 0.488704,
+      "grad_norm": 0.42183083295822144,
+      "learning_rate": 0.000542550905049696,
+      "loss": 2.8536,
+      "step": 3818
+    },
+    {
+      "epoch": 0.488832,
+      "grad_norm": 0.4530366063117981,
+      "learning_rate": 0.0005423443423526841,
+      "loss": 2.9082,
+      "step": 3819
+    },
+    {
+      "epoch": 0.48896,
+      "grad_norm": 0.46060335636138916,
+      "learning_rate": 0.000542137772376189,
+      "loss": 2.9268,
+      "step": 3820
+    },
+    {
+      "epoch": 0.489088,
+      "grad_norm": 0.43720853328704834,
+      "learning_rate": 0.0005419311951557221,
+      "loss": 2.8976,
+      "step": 3821
+    },
+    {
+      "epoch": 0.489216,
+      "grad_norm": 0.42510613799095154,
+      "learning_rate": 0.0005417246107267968,
+      "loss": 2.8866,
+      "step": 3822
+    },
+    {
+      "epoch": 0.489344,
+      "grad_norm": 0.42179277539253235,
+      "learning_rate": 0.0005415180191249271,
+      "loss": 2.7794,
+      "step": 3823
+    },
+    {
+      "epoch": 0.489472,
+      "grad_norm": 0.42388421297073364,
+      "learning_rate": 0.0005413114203856286,
+      "loss": 2.7639,
+      "step": 3824
+    },
+    {
+      "epoch": 0.4896,
+      "grad_norm": 0.4214480221271515,
+      "learning_rate": 0.0005411048145444182,
+      "loss": 2.9253,
+      "step": 3825
+    },
+    {
+      "epoch": 0.489728,
+      "grad_norm": 0.4297669231891632,
+      "learning_rate": 0.0005408982016368133,
+      "loss": 2.8631,
+      "step": 3826
+    },
+    {
+      "epoch": 0.489856,
+      "grad_norm": 0.44233953952789307,
+      "learning_rate": 0.0005406915816983336,
+      "loss": 2.8788,
+      "step": 3827
+    },
+    {
+      "epoch": 0.489984,
+      "grad_norm": 0.4594293236732483,
+      "learning_rate": 0.0005404849547644994,
+      "loss": 2.8173,
+      "step": 3828
+    },
+    {
+      "epoch": 0.490112,
+      "grad_norm": 0.43543028831481934,
+      "learning_rate": 0.0005402783208708319,
+      "loss": 2.8279,
+      "step": 3829
+    },
+    {
+      "epoch": 0.49024,
+      "grad_norm": 0.43814200162887573,
+      "learning_rate": 0.0005400716800528542,
+      "loss": 2.9231,
+      "step": 3830
+    },
+    {
+      "epoch": 0.490368,
+      "grad_norm": 0.4609832465648651,
+      "learning_rate": 0.0005398650323460903,
+      "loss": 2.8322,
+      "step": 3831
+    },
+    {
+      "epoch": 0.490496,
+      "grad_norm": 0.4258003830909729,
+      "learning_rate": 0.000539658377786065,
+      "loss": 2.9229,
+      "step": 3832
+    },
+    {
+      "epoch": 0.490624,
+      "grad_norm": 0.46023279428482056,
+      "learning_rate": 0.0005394517164083047,
+      "loss": 2.8752,
+      "step": 3833
+    },
+    {
+      "epoch": 0.490752,
+      "grad_norm": 0.42779216170310974,
+      "learning_rate": 0.0005392450482483372,
+      "loss": 2.8825,
+      "step": 3834
+    },
+    {
+      "epoch": 0.49088,
+      "grad_norm": 0.4077283442020416,
+      "learning_rate": 0.0005390383733416909,
+      "loss": 2.7896,
+      "step": 3835
+    },
+    {
+      "epoch": 0.491008,
+      "grad_norm": 0.39681825041770935,
+      "learning_rate": 0.0005388316917238955,
+      "loss": 2.8892,
+      "step": 3836
+    },
+    {
+      "epoch": 0.491136,
+      "grad_norm": 0.41111624240875244,
+      "learning_rate": 0.0005386250034304823,
+      "loss": 2.9085,
+      "step": 3837
+    },
+    {
+      "epoch": 0.491264,
+      "grad_norm": 0.42380425333976746,
+      "learning_rate": 0.0005384183084969831,
+      "loss": 2.8864,
+      "step": 3838
+    },
+    {
+      "epoch": 0.491392,
+      "grad_norm": 0.4503888785839081,
+      "learning_rate": 0.0005382116069589315,
+      "loss": 2.9084,
+      "step": 3839
+    },
+    {
+      "epoch": 0.49152,
+      "grad_norm": 0.3998926877975464,
+      "learning_rate": 0.0005380048988518613,
+      "loss": 2.8117,
+      "step": 3840
+    },
+    {
+      "epoch": 0.491648,
+      "grad_norm": 0.46822643280029297,
+      "learning_rate": 0.0005377981842113087,
+      "loss": 2.9709,
+      "step": 3841
+    },
+    {
+      "epoch": 0.491776,
+      "grad_norm": 0.4293217957019806,
+      "learning_rate": 0.0005375914630728099,
+      "loss": 2.9281,
+      "step": 3842
+    },
+    {
+      "epoch": 0.491904,
+      "grad_norm": 0.3933890759944916,
+      "learning_rate": 0.0005373847354719029,
+      "loss": 2.7881,
+      "step": 3843
+    },
+    {
+      "epoch": 0.492032,
+      "grad_norm": 0.4320515990257263,
+      "learning_rate": 0.0005371780014441264,
+      "loss": 2.8108,
+      "step": 3844
+    },
+    {
+      "epoch": 0.49216,
+      "grad_norm": 0.42600205540657043,
+      "learning_rate": 0.0005369712610250206,
+      "loss": 2.7793,
+      "step": 3845
+    },
+    {
+      "epoch": 0.492288,
+      "grad_norm": 0.3822498917579651,
+      "learning_rate": 0.0005367645142501262,
+      "loss": 2.893,
+      "step": 3846
+    },
+    {
+      "epoch": 0.492416,
+      "grad_norm": 0.39799216389656067,
+      "learning_rate": 0.0005365577611549855,
+      "loss": 2.7871,
+      "step": 3847
+    },
+    {
+      "epoch": 0.492544,
+      "grad_norm": 0.3962434232234955,
+      "learning_rate": 0.0005363510017751422,
+      "loss": 2.8547,
+      "step": 3848
+    },
+    {
+      "epoch": 0.492672,
+      "grad_norm": 0.3992100954055786,
+      "learning_rate": 0.0005361442361461402,
+      "loss": 2.8411,
+      "step": 3849
+    },
+    {
+      "epoch": 0.4928,
+      "grad_norm": 0.4112558662891388,
+      "learning_rate": 0.0005359374643035248,
+      "loss": 2.9817,
+      "step": 3850
+    },
+    {
+      "epoch": 0.492928,
+      "grad_norm": 0.4156324863433838,
+      "learning_rate": 0.0005357306862828427,
+      "loss": 2.942,
+      "step": 3851
+    },
+    {
+      "epoch": 0.493056,
+      "grad_norm": 0.4151874780654907,
+      "learning_rate": 0.0005355239021196416,
+      "loss": 2.8677,
+      "step": 3852
+    },
+    {
+      "epoch": 0.493184,
+      "grad_norm": 0.3959118723869324,
+      "learning_rate": 0.0005353171118494698,
+      "loss": 2.7534,
+      "step": 3853
+    },
+    {
+      "epoch": 0.493312,
+      "grad_norm": 0.40935155749320984,
+      "learning_rate": 0.0005351103155078768,
+      "loss": 2.7549,
+      "step": 3854
+    },
+    {
+      "epoch": 0.49344,
+      "grad_norm": 0.42022189497947693,
+      "learning_rate": 0.0005349035131304138,
+      "loss": 2.6716,
+      "step": 3855
+    },
+    {
+      "epoch": 0.493568,
+      "grad_norm": 0.4125000536441803,
+      "learning_rate": 0.0005346967047526321,
+      "loss": 2.8785,
+      "step": 3856
+    },
+    {
+      "epoch": 0.493696,
+      "grad_norm": 0.42368537187576294,
+      "learning_rate": 0.0005344898904100848,
+      "loss": 2.9049,
+      "step": 3857
+    },
+    {
+      "epoch": 0.493824,
+      "grad_norm": 0.4014261066913605,
+      "learning_rate": 0.0005342830701383254,
+      "loss": 2.825,
+      "step": 3858
+    },
+    {
+      "epoch": 0.493952,
+      "grad_norm": 0.4286510944366455,
+      "learning_rate": 0.0005340762439729088,
+      "loss": 2.8448,
+      "step": 3859
+    },
+    {
+      "epoch": 0.49408,
+      "grad_norm": 0.40717047452926636,
+      "learning_rate": 0.0005338694119493908,
+      "loss": 2.8168,
+      "step": 3860
+    },
+    {
+      "epoch": 0.494208,
+      "grad_norm": 0.4186759889125824,
+      "learning_rate": 0.0005336625741033283,
+      "loss": 2.7776,
+      "step": 3861
+    },
+    {
+      "epoch": 0.494336,
+      "grad_norm": 0.42026078701019287,
+      "learning_rate": 0.0005334557304702789,
+      "loss": 2.7794,
+      "step": 3862
+    },
+    {
+      "epoch": 0.494464,
+      "grad_norm": 0.4477492570877075,
+      "learning_rate": 0.0005332488810858017,
+      "loss": 2.763,
+      "step": 3863
+    },
+    {
+      "epoch": 0.494592,
+      "grad_norm": 0.45416659116744995,
+      "learning_rate": 0.0005330420259854564,
+      "loss": 2.9364,
+      "step": 3864
+    },
+    {
+      "epoch": 0.49472,
+      "grad_norm": 0.3771000802516937,
+      "learning_rate": 0.0005328351652048037,
+      "loss": 3.0761,
+      "step": 3865
+    },
+    {
+      "epoch": 0.494848,
+      "grad_norm": 0.4699905514717102,
+      "learning_rate": 0.0005326282987794056,
+      "loss": 2.9256,
+      "step": 3866
+    },
+    {
+      "epoch": 0.494976,
+      "grad_norm": 0.4205271303653717,
+      "learning_rate": 0.0005324214267448245,
+      "loss": 2.8584,
+      "step": 3867
+    },
+    {
+      "epoch": 0.495104,
+      "grad_norm": 0.46251732110977173,
+      "learning_rate": 0.0005322145491366243,
+      "loss": 2.8486,
+      "step": 3868
+    },
+    {
+      "epoch": 0.495232,
+      "grad_norm": 0.41948574781417847,
+      "learning_rate": 0.0005320076659903698,
+      "loss": 2.8637,
+      "step": 3869
+    },
+    {
+      "epoch": 0.49536,
+      "grad_norm": 0.39813032746315,
+      "learning_rate": 0.0005318007773416263,
+      "loss": 2.8132,
+      "step": 3870
+    },
+    {
+      "epoch": 0.495488,
+      "grad_norm": 0.4130594730377197,
+      "learning_rate": 0.0005315938832259606,
+      "loss": 2.7437,
+      "step": 3871
+    },
+    {
+      "epoch": 0.495616,
+      "grad_norm": 0.418660968542099,
+      "learning_rate": 0.00053138698367894,
+      "loss": 2.8585,
+      "step": 3872
+    },
+    {
+      "epoch": 0.495744,
+      "grad_norm": 0.42727676033973694,
+      "learning_rate": 0.0005311800787361332,
+      "loss": 2.9428,
+      "step": 3873
+    },
+    {
+      "epoch": 0.495872,
+      "grad_norm": 0.41886046528816223,
+      "learning_rate": 0.0005309731684331093,
+      "loss": 2.9446,
+      "step": 3874
+    },
+    {
+      "epoch": 0.496,
+      "grad_norm": 0.4045485556125641,
+      "learning_rate": 0.0005307662528054387,
+      "loss": 2.9279,
+      "step": 3875
+    },
+    {
+      "epoch": 0.496128,
+      "grad_norm": 0.3947528600692749,
+      "learning_rate": 0.0005305593318886924,
+      "loss": 2.9433,
+      "step": 3876
+    },
+    {
+      "epoch": 0.496256,
+      "grad_norm": 0.3990950286388397,
+      "learning_rate": 0.0005303524057184429,
+      "loss": 2.8255,
+      "step": 3877
+    },
+    {
+      "epoch": 0.496384,
+      "grad_norm": 0.44006070494651794,
+      "learning_rate": 0.0005301454743302629,
+      "loss": 2.8799,
+      "step": 3878
+    },
+    {
+      "epoch": 0.496512,
+      "grad_norm": 0.42290785908699036,
+      "learning_rate": 0.0005299385377597265,
+      "loss": 2.8956,
+      "step": 3879
+    },
+    {
+      "epoch": 0.49664,
+      "grad_norm": 0.43569809198379517,
+      "learning_rate": 0.0005297315960424081,
+      "loss": 2.7961,
+      "step": 3880
+    },
+    {
+      "epoch": 0.496768,
+      "grad_norm": 0.4483851194381714,
+      "learning_rate": 0.0005295246492138839,
+      "loss": 2.7767,
+      "step": 3881
+    },
+    {
+      "epoch": 0.496896,
+      "grad_norm": 0.4277383089065552,
+      "learning_rate": 0.0005293176973097302,
+      "loss": 2.9147,
+      "step": 3882
+    },
+    {
+      "epoch": 0.497024,
+      "grad_norm": 0.4581163823604584,
+      "learning_rate": 0.0005291107403655242,
+      "loss": 2.8707,
+      "step": 3883
+    },
+    {
+      "epoch": 0.497152,
+      "grad_norm": 0.4349813461303711,
+      "learning_rate": 0.0005289037784168448,
+      "loss": 2.9283,
+      "step": 3884
+    },
+    {
+      "epoch": 0.49728,
+      "grad_norm": 0.4218544363975525,
+      "learning_rate": 0.0005286968114992706,
+      "loss": 2.8646,
+      "step": 3885
+    },
+    {
+      "epoch": 0.497408,
+      "grad_norm": 0.42760777473449707,
+      "learning_rate": 0.0005284898396483819,
+      "loss": 2.8843,
+      "step": 3886
+    },
+    {
+      "epoch": 0.497536,
+      "grad_norm": 0.43409737944602966,
+      "learning_rate": 0.0005282828628997595,
+      "loss": 2.9909,
+      "step": 3887
+    },
+    {
+      "epoch": 0.497664,
+      "grad_norm": 0.42124083638191223,
+      "learning_rate": 0.000528075881288985,
+      "loss": 2.8205,
+      "step": 3888
+    },
+    {
+      "epoch": 0.497792,
+      "grad_norm": 0.3931414783000946,
+      "learning_rate": 0.0005278688948516412,
+      "loss": 2.7433,
+      "step": 3889
+    },
+    {
+      "epoch": 0.49792,
+      "grad_norm": 0.45035046339035034,
+      "learning_rate": 0.0005276619036233111,
+      "loss": 2.8184,
+      "step": 3890
+    },
+    {
+      "epoch": 0.498048,
+      "grad_norm": 0.4250778555870056,
+      "learning_rate": 0.0005274549076395792,
+      "loss": 2.8666,
+      "step": 3891
+    },
+    {
+      "epoch": 0.498176,
+      "grad_norm": 0.4184761643409729,
+      "learning_rate": 0.0005272479069360302,
+      "loss": 2.8167,
+      "step": 3892
+    },
+    {
+      "epoch": 0.498304,
+      "grad_norm": 0.459006130695343,
+      "learning_rate": 0.0005270409015482503,
+      "loss": 2.9138,
+      "step": 3893
+    },
+    {
+      "epoch": 0.498432,
+      "grad_norm": 0.44776856899261475,
+      "learning_rate": 0.0005268338915118258,
+      "loss": 2.9446,
+      "step": 3894
+    },
+    {
+      "epoch": 0.49856,
+      "grad_norm": 0.46756690740585327,
+      "learning_rate": 0.0005266268768623444,
+      "loss": 2.8468,
+      "step": 3895
+    },
+    {
+      "epoch": 0.498688,
+      "grad_norm": 0.43127208948135376,
+      "learning_rate": 0.0005264198576353942,
+      "loss": 2.9128,
+      "step": 3896
+    },
+    {
+      "epoch": 0.498816,
+      "grad_norm": 0.43572360277175903,
+      "learning_rate": 0.0005262128338665641,
+      "loss": 2.9821,
+      "step": 3897
+    },
+    {
+      "epoch": 0.498944,
+      "grad_norm": 0.4405980706214905,
+      "learning_rate": 0.000526005805591444,
+      "loss": 2.7686,
+      "step": 3898
+    },
+    {
+      "epoch": 0.499072,
+      "grad_norm": 0.4509694576263428,
+      "learning_rate": 0.0005257987728456244,
+      "loss": 2.8911,
+      "step": 3899
+    },
+    {
+      "epoch": 0.4992,
+      "grad_norm": 0.42567768692970276,
+      "learning_rate": 0.0005255917356646968,
+      "loss": 2.7669,
+      "step": 3900
+    },
+    {
+      "epoch": 0.499328,
+      "grad_norm": 0.43367359042167664,
+      "learning_rate": 0.0005253846940842528,
+      "loss": 2.9961,
+      "step": 3901
+    },
+    {
+      "epoch": 0.499456,
+      "grad_norm": 0.44393420219421387,
+      "learning_rate": 0.000525177648139886,
+      "loss": 2.9547,
+      "step": 3902
+    },
+    {
+      "epoch": 0.499584,
+      "grad_norm": 0.4038757383823395,
+      "learning_rate": 0.0005249705978671895,
+      "loss": 2.8499,
+      "step": 3903
+    },
+    {
+      "epoch": 0.499712,
+      "grad_norm": 0.4123300313949585,
+      "learning_rate": 0.0005247635433017576,
+      "loss": 2.9496,
+      "step": 3904
+    },
+    {
+      "epoch": 0.49984,
+      "grad_norm": 0.3855282664299011,
+      "learning_rate": 0.0005245564844791857,
+      "loss": 2.878,
+      "step": 3905
+    },
+    {
+      "epoch": 0.499968,
+      "grad_norm": 0.40939533710479736,
+      "learning_rate": 0.0005243494214350693,
+      "loss": 2.7326,
+      "step": 3906
+    },
+    {
+      "epoch": 0.500096,
+      "grad_norm": 0.40486177802085876,
+      "learning_rate": 0.0005241423542050051,
+      "loss": 2.8973,
+      "step": 3907
+    },
+    {
+      "epoch": 0.500224,
+      "grad_norm": 0.4247356951236725,
+      "learning_rate": 0.0005239352828245904,
+      "loss": 2.9261,
+      "step": 3908
+    },
+    {
+      "epoch": 0.500352,
+      "grad_norm": 0.4040723443031311,
+      "learning_rate": 0.000523728207329423,
+      "loss": 2.8422,
+      "step": 3909
+    },
+    {
+      "epoch": 0.50048,
+      "grad_norm": 0.4181481599807739,
+      "learning_rate": 0.0005235211277551018,
+      "loss": 2.9,
+      "step": 3910
+    },
+    {
+      "epoch": 0.500608,
+      "grad_norm": 0.4112429916858673,
+      "learning_rate": 0.000523314044137226,
+      "loss": 2.8968,
+      "step": 3911
+    },
+    {
+      "epoch": 0.500736,
+      "grad_norm": 0.47169414162635803,
+      "learning_rate": 0.0005231069565113957,
+      "loss": 2.7549,
+      "step": 3912
+    },
+    {
+      "epoch": 0.500864,
+      "grad_norm": 0.4226801097393036,
+      "learning_rate": 0.0005228998649132118,
+      "loss": 2.8828,
+      "step": 3913
+    },
+    {
+      "epoch": 0.500992,
+      "grad_norm": 0.45490962266921997,
+      "learning_rate": 0.0005226927693782754,
+      "loss": 2.9362,
+      "step": 3914
+    },
+    {
+      "epoch": 0.50112,
+      "grad_norm": 0.42030540108680725,
+      "learning_rate": 0.0005224856699421889,
+      "loss": 3.0472,
+      "step": 3915
+    },
+    {
+      "epoch": 0.501248,
+      "grad_norm": 0.41244572401046753,
+      "learning_rate": 0.0005222785666405553,
+      "loss": 2.7805,
+      "step": 3916
+    },
+    {
+      "epoch": 0.501376,
+      "grad_norm": 0.39253363013267517,
+      "learning_rate": 0.0005220714595089777,
+      "loss": 2.7781,
+      "step": 3917
+    },
+    {
+      "epoch": 0.501504,
+      "grad_norm": 0.3996981978416443,
+      "learning_rate": 0.0005218643485830605,
+      "loss": 2.9265,
+      "step": 3918
+    },
+    {
+      "epoch": 0.501632,
+      "grad_norm": 0.43807852268218994,
+      "learning_rate": 0.000521657233898408,
+      "loss": 2.7687,
+      "step": 3919
+    },
+    {
+      "epoch": 0.50176,
+      "grad_norm": 0.3962256908416748,
+      "learning_rate": 0.0005214501154906261,
+      "loss": 2.833,
+      "step": 3920
+    },
+    {
+      "epoch": 0.501888,
+      "grad_norm": 0.3950268030166626,
+      "learning_rate": 0.0005212429933953206,
+      "loss": 2.9122,
+      "step": 3921
+    },
+    {
+      "epoch": 0.502016,
+      "grad_norm": 0.4119686186313629,
+      "learning_rate": 0.0005210358676480983,
+      "loss": 2.9136,
+      "step": 3922
+    },
+    {
+      "epoch": 0.502144,
+      "grad_norm": 0.4059731364250183,
+      "learning_rate": 0.0005208287382845666,
+      "loss": 2.782,
+      "step": 3923
+    },
+    {
+      "epoch": 0.502272,
+      "grad_norm": 0.4271673858165741,
+      "learning_rate": 0.0005206216053403333,
+      "loss": 2.8669,
+      "step": 3924
+    },
+    {
+      "epoch": 0.5024,
+      "grad_norm": 0.40853336453437805,
+      "learning_rate": 0.0005204144688510068,
+      "loss": 2.8673,
+      "step": 3925
+    },
+    {
+      "epoch": 0.502528,
+      "grad_norm": 0.44123250246047974,
+      "learning_rate": 0.0005202073288521964,
+      "loss": 2.8419,
+      "step": 3926
+    },
+    {
+      "epoch": 0.502656,
+      "grad_norm": 0.40439680218696594,
+      "learning_rate": 0.0005200001853795123,
+      "loss": 2.8935,
+      "step": 3927
+    },
+    {
+      "epoch": 0.502784,
+      "grad_norm": 0.39985835552215576,
+      "learning_rate": 0.0005197930384685642,
+      "loss": 2.7455,
+      "step": 3928
+    },
+    {
+      "epoch": 0.502912,
+      "grad_norm": 0.4175058603286743,
+      "learning_rate": 0.0005195858881549632,
+      "loss": 2.9927,
+      "step": 3929
+    },
+    {
+      "epoch": 0.50304,
+      "grad_norm": 0.4484770894050598,
+      "learning_rate": 0.000519378734474321,
+      "loss": 2.7929,
+      "step": 3930
+    },
+    {
+      "epoch": 0.503168,
+      "grad_norm": 0.41283735632896423,
+      "learning_rate": 0.0005191715774622499,
+      "loss": 3.0296,
+      "step": 3931
+    },
+    {
+      "epoch": 0.503296,
+      "grad_norm": 0.5257746577262878,
+      "learning_rate": 0.0005189644171543624,
+      "loss": 2.8234,
+      "step": 3932
+    },
+    {
+      "epoch": 0.503424,
+      "grad_norm": 0.37903767824172974,
+      "learning_rate": 0.0005187572535862715,
+      "loss": 2.8254,
+      "step": 3933
+    },
+    {
+      "epoch": 0.503552,
+      "grad_norm": 0.41114577651023865,
+      "learning_rate": 0.0005185500867935913,
+      "loss": 2.8454,
+      "step": 3934
+    },
+    {
+      "epoch": 0.50368,
+      "grad_norm": 0.3976965844631195,
+      "learning_rate": 0.0005183429168119362,
+      "loss": 2.8582,
+      "step": 3935
+    },
+    {
+      "epoch": 0.503808,
+      "grad_norm": 0.40707284212112427,
+      "learning_rate": 0.000518135743676921,
+      "loss": 2.8431,
+      "step": 3936
+    },
+    {
+      "epoch": 0.503936,
+      "grad_norm": 0.3910399377346039,
+      "learning_rate": 0.0005179285674241614,
+      "loss": 2.9364,
+      "step": 3937
+    },
+    {
+      "epoch": 0.504064,
+      "grad_norm": 0.4019538462162018,
+      "learning_rate": 0.0005177213880892729,
+      "loss": 2.9035,
+      "step": 3938
+    },
+    {
+      "epoch": 0.504192,
+      "grad_norm": 0.4430789053440094,
+      "learning_rate": 0.0005175142057078727,
+      "loss": 2.9283,
+      "step": 3939
+    },
+    {
+      "epoch": 0.50432,
+      "grad_norm": 0.4261739253997803,
+      "learning_rate": 0.0005173070203155772,
+      "loss": 2.8773,
+      "step": 3940
+    },
+    {
+      "epoch": 0.504448,
+      "grad_norm": 0.40120503306388855,
+      "learning_rate": 0.0005170998319480043,
+      "loss": 2.7612,
+      "step": 3941
+    },
+    {
+      "epoch": 0.504576,
+      "grad_norm": 0.41164299845695496,
+      "learning_rate": 0.0005168926406407722,
+      "loss": 2.7831,
+      "step": 3942
+    },
+    {
+      "epoch": 0.504704,
+      "grad_norm": 0.4330790042877197,
+      "learning_rate": 0.0005166854464294993,
+      "loss": 2.8153,
+      "step": 3943
+    },
+    {
+      "epoch": 0.504832,
+      "grad_norm": 0.4457439184188843,
+      "learning_rate": 0.0005164782493498046,
+      "loss": 2.9901,
+      "step": 3944
+    },
+    {
+      "epoch": 0.50496,
+      "grad_norm": 0.4085686504840851,
+      "learning_rate": 0.000516271049437308,
+      "loss": 2.8119,
+      "step": 3945
+    },
+    {
+      "epoch": 0.505088,
+      "grad_norm": 0.41812077164649963,
+      "learning_rate": 0.0005160638467276293,
+      "loss": 2.8048,
+      "step": 3946
+    },
+    {
+      "epoch": 0.505216,
+      "grad_norm": 0.44970986247062683,
+      "learning_rate": 0.000515856641256389,
+      "loss": 2.9514,
+      "step": 3947
+    },
+    {
+      "epoch": 0.505344,
+      "grad_norm": 0.41997087001800537,
+      "learning_rate": 0.0005156494330592083,
+      "loss": 2.8585,
+      "step": 3948
+    },
+    {
+      "epoch": 0.505472,
+      "grad_norm": 0.45252150297164917,
+      "learning_rate": 0.0005154422221717086,
+      "loss": 2.9875,
+      "step": 3949
+    },
+    {
+      "epoch": 0.5056,
+      "grad_norm": 0.39206501841545105,
+      "learning_rate": 0.0005152350086295121,
+      "loss": 2.6999,
+      "step": 3950
+    },
+    {
+      "epoch": 0.505728,
+      "grad_norm": 0.37131550908088684,
+      "learning_rate": 0.0005150277924682408,
+      "loss": 2.8901,
+      "step": 3951
+    },
+    {
+      "epoch": 0.505856,
+      "grad_norm": 0.371511310338974,
+      "learning_rate": 0.0005148205737235177,
+      "loss": 2.8476,
+      "step": 3952
+    },
+    {
+      "epoch": 0.505984,
+      "grad_norm": 0.3848649263381958,
+      "learning_rate": 0.0005146133524309664,
+      "loss": 2.8609,
+      "step": 3953
+    },
+    {
+      "epoch": 0.506112,
+      "grad_norm": 0.4530373811721802,
+      "learning_rate": 0.0005144061286262102,
+      "loss": 2.908,
+      "step": 3954
+    },
+    {
+      "epoch": 0.50624,
+      "grad_norm": 0.4150950610637665,
+      "learning_rate": 0.0005141989023448735,
+      "loss": 2.7659,
+      "step": 3955
+    },
+    {
+      "epoch": 0.506368,
+      "grad_norm": 0.40299248695373535,
+      "learning_rate": 0.000513991673622581,
+      "loss": 2.8682,
+      "step": 3956
+    },
+    {
+      "epoch": 0.506496,
+      "grad_norm": 0.42051875591278076,
+      "learning_rate": 0.0005137844424949575,
+      "loss": 2.8849,
+      "step": 3957
+    },
+    {
+      "epoch": 0.506624,
+      "grad_norm": 0.39584577083587646,
+      "learning_rate": 0.0005135772089976285,
+      "loss": 2.819,
+      "step": 3958
+    },
+    {
+      "epoch": 0.506752,
+      "grad_norm": 0.42329519987106323,
+      "learning_rate": 0.00051336997316622,
+      "loss": 2.8601,
+      "step": 3959
+    },
+    {
+      "epoch": 0.50688,
+      "grad_norm": 0.42248770594596863,
+      "learning_rate": 0.0005131627350363581,
+      "loss": 2.9042,
+      "step": 3960
+    },
+    {
+      "epoch": 0.507008,
+      "grad_norm": 0.41717761754989624,
+      "learning_rate": 0.0005129554946436694,
+      "loss": 2.7435,
+      "step": 3961
+    },
+    {
+      "epoch": 0.507136,
+      "grad_norm": 0.4598080813884735,
+      "learning_rate": 0.0005127482520237808,
+      "loss": 2.954,
+      "step": 3962
+    },
+    {
+      "epoch": 0.507264,
+      "grad_norm": 0.47410136461257935,
+      "learning_rate": 0.0005125410072123202,
+      "loss": 2.9637,
+      "step": 3963
+    },
+    {
+      "epoch": 0.507392,
+      "grad_norm": 0.5828011631965637,
+      "learning_rate": 0.000512333760244915,
+      "loss": 3.0553,
+      "step": 3964
+    },
+    {
+      "epoch": 0.50752,
+      "grad_norm": 0.40129178762435913,
+      "learning_rate": 0.0005121265111571933,
+      "loss": 2.8851,
+      "step": 3965
+    },
+    {
+      "epoch": 0.507648,
+      "grad_norm": 0.38992995023727417,
+      "learning_rate": 0.0005119192599847837,
+      "loss": 2.8346,
+      "step": 3966
+    },
+    {
+      "epoch": 0.507776,
+      "grad_norm": 0.4479053020477295,
+      "learning_rate": 0.0005117120067633153,
+      "loss": 3.1012,
+      "step": 3967
+    },
+    {
+      "epoch": 0.507904,
+      "grad_norm": 0.41870760917663574,
+      "learning_rate": 0.0005115047515284173,
+      "loss": 2.8986,
+      "step": 3968
+    },
+    {
+      "epoch": 0.508032,
+      "grad_norm": 0.40679100155830383,
+      "learning_rate": 0.0005112974943157189,
+      "loss": 2.766,
+      "step": 3969
+    },
+    {
+      "epoch": 0.50816,
+      "grad_norm": 0.42168670892715454,
+      "learning_rate": 0.0005110902351608503,
+      "loss": 2.6021,
+      "step": 3970
+    },
+    {
+      "epoch": 0.508288,
+      "grad_norm": 0.421496719121933,
+      "learning_rate": 0.0005108829740994417,
+      "loss": 2.8172,
+      "step": 3971
+    },
+    {
+      "epoch": 0.508416,
+      "grad_norm": 0.41575464606285095,
+      "learning_rate": 0.0005106757111671235,
+      "loss": 2.7569,
+      "step": 3972
+    },
+    {
+      "epoch": 0.508544,
+      "grad_norm": 0.4060327410697937,
+      "learning_rate": 0.0005104684463995271,
+      "loss": 2.84,
+      "step": 3973
+    },
+    {
+      "epoch": 0.508672,
+      "grad_norm": 0.4417901933193207,
+      "learning_rate": 0.0005102611798322833,
+      "loss": 3.0362,
+      "step": 3974
+    },
+    {
+      "epoch": 0.5088,
+      "grad_norm": 0.39811065793037415,
+      "learning_rate": 0.0005100539115010236,
+      "loss": 2.7926,
+      "step": 3975
+    },
+    {
+      "epoch": 0.508928,
+      "grad_norm": 0.42787837982177734,
+      "learning_rate": 0.00050984664144138,
+      "loss": 2.7894,
+      "step": 3976
+    },
+    {
+      "epoch": 0.509056,
+      "grad_norm": 0.43865612149238586,
+      "learning_rate": 0.0005096393696889847,
+      "loss": 2.8576,
+      "step": 3977
+    },
+    {
+      "epoch": 0.509184,
+      "grad_norm": 0.44969964027404785,
+      "learning_rate": 0.0005094320962794701,
+      "loss": 2.8762,
+      "step": 3978
+    },
+    {
+      "epoch": 0.509312,
+      "grad_norm": 0.4528118968009949,
+      "learning_rate": 0.0005092248212484686,
+      "loss": 2.9901,
+      "step": 3979
+    },
+    {
+      "epoch": 0.50944,
+      "grad_norm": 0.42348694801330566,
+      "learning_rate": 0.0005090175446316133,
+      "loss": 3.0029,
+      "step": 3980
+    },
+    {
+      "epoch": 0.509568,
+      "grad_norm": 0.44461366534233093,
+      "learning_rate": 0.0005088102664645378,
+      "loss": 2.8491,
+      "step": 3981
+    },
+    {
+      "epoch": 0.509696,
+      "grad_norm": 0.3892180919647217,
+      "learning_rate": 0.0005086029867828752,
+      "loss": 2.9079,
+      "step": 3982
+    },
+    {
+      "epoch": 0.509824,
+      "grad_norm": 0.4239261746406555,
+      "learning_rate": 0.0005083957056222594,
+      "loss": 2.9184,
+      "step": 3983
+    },
+    {
+      "epoch": 0.509952,
+      "grad_norm": 0.39174461364746094,
+      "learning_rate": 0.0005081884230183247,
+      "loss": 2.749,
+      "step": 3984
+    },
+    {
+      "epoch": 0.51008,
+      "grad_norm": 0.40927764773368835,
+      "learning_rate": 0.0005079811390067048,
+      "loss": 2.794,
+      "step": 3985
+    },
+    {
+      "epoch": 0.510208,
+      "grad_norm": 0.4874249994754791,
+      "learning_rate": 0.000507773853623035,
+      "loss": 2.7718,
+      "step": 3986
+    },
+    {
+      "epoch": 0.510336,
+      "grad_norm": 0.4349079728126526,
+      "learning_rate": 0.0005075665669029494,
+      "loss": 2.7756,
+      "step": 3987
+    },
+    {
+      "epoch": 0.510464,
+      "grad_norm": 0.4389895498752594,
+      "learning_rate": 0.0005073592788820832,
+      "loss": 2.8423,
+      "step": 3988
+    },
+    {
+      "epoch": 0.510592,
+      "grad_norm": 0.4287116825580597,
+      "learning_rate": 0.0005071519895960717,
+      "loss": 2.8928,
+      "step": 3989
+    },
+    {
+      "epoch": 0.51072,
+      "grad_norm": 0.424008309841156,
+      "learning_rate": 0.0005069446990805501,
+      "loss": 2.7937,
+      "step": 3990
+    },
+    {
+      "epoch": 0.510848,
+      "grad_norm": 0.379194438457489,
+      "learning_rate": 0.0005067374073711544,
+      "loss": 2.8856,
+      "step": 3991
+    },
+    {
+      "epoch": 0.510976,
+      "grad_norm": 0.3957364559173584,
+      "learning_rate": 0.0005065301145035204,
+      "loss": 2.836,
+      "step": 3992
+    },
+    {
+      "epoch": 0.511104,
+      "grad_norm": 0.43100589513778687,
+      "learning_rate": 0.0005063228205132839,
+      "loss": 2.7906,
+      "step": 3993
+    },
+    {
+      "epoch": 0.511232,
+      "grad_norm": 0.44487181305885315,
+      "learning_rate": 0.000506115525436081,
+      "loss": 2.9571,
+      "step": 3994
+    },
+    {
+      "epoch": 0.51136,
+      "grad_norm": 0.3986407220363617,
+      "learning_rate": 0.0005059082293075487,
+      "loss": 2.8393,
+      "step": 3995
+    },
+    {
+      "epoch": 0.511488,
+      "grad_norm": 0.4443216621875763,
+      "learning_rate": 0.0005057009321633233,
+      "loss": 2.8918,
+      "step": 3996
+    },
+    {
+      "epoch": 0.511616,
+      "grad_norm": 0.40398937463760376,
+      "learning_rate": 0.0005054936340390414,
+      "loss": 2.8543,
+      "step": 3997
+    },
+    {
+      "epoch": 0.511744,
+      "grad_norm": 0.4124487340450287,
+      "learning_rate": 0.0005052863349703402,
+      "loss": 2.8742,
+      "step": 3998
+    },
+    {
+      "epoch": 0.511872,
+      "grad_norm": 0.40908434987068176,
+      "learning_rate": 0.0005050790349928566,
+      "loss": 3.01,
+      "step": 3999
+    },
+    {
+      "epoch": 0.512,
+      "grad_norm": 0.41094303131103516,
+      "learning_rate": 0.0005048717341422282,
+      "loss": 3.0012,
+      "step": 4000
+    },
+    {
+      "epoch": 0.512128,
+      "grad_norm": 0.41971835494041443,
+      "learning_rate": 0.0005046644324540921,
+      "loss": 2.8573,
+      "step": 4001
+    },
+    {
+      "epoch": 0.512256,
+      "grad_norm": 0.40278467535972595,
+      "learning_rate": 0.000504457129964086,
+      "loss": 2.9391,
+      "step": 4002
+    },
+    {
+      "epoch": 0.512384,
+      "grad_norm": 0.4455163776874542,
+      "learning_rate": 0.0005042498267078476,
+      "loss": 2.8087,
+      "step": 4003
+    },
+    {
+      "epoch": 0.512512,
+      "grad_norm": 0.4030039608478546,
+      "learning_rate": 0.0005040425227210148,
+      "loss": 2.8013,
+      "step": 4004
+    },
+    {
+      "epoch": 0.51264,
+      "grad_norm": 0.3906235992908478,
+      "learning_rate": 0.0005038352180392254,
+      "loss": 2.8286,
+      "step": 4005
+    },
+    {
+      "epoch": 0.512768,
+      "grad_norm": 0.4324887692928314,
+      "learning_rate": 0.0005036279126981176,
+      "loss": 2.798,
+      "step": 4006
+    },
+    {
+      "epoch": 0.512896,
+      "grad_norm": 0.4179579019546509,
+      "learning_rate": 0.0005034206067333296,
+      "loss": 2.8233,
+      "step": 4007
+    },
+    {
+      "epoch": 0.513024,
+      "grad_norm": 0.3984704613685608,
+      "learning_rate": 0.0005032133001804993,
+      "loss": 2.6588,
+      "step": 4008
+    },
+    {
+      "epoch": 0.513152,
+      "grad_norm": 0.38438180088996887,
+      "learning_rate": 0.0005030059930752658,
+      "loss": 2.6912,
+      "step": 4009
+    },
+    {
+      "epoch": 0.51328,
+      "grad_norm": 0.4341740310192108,
+      "learning_rate": 0.0005027986854532672,
+      "loss": 3.0044,
+      "step": 4010
+    },
+    {
+      "epoch": 0.513408,
+      "grad_norm": 0.4153798520565033,
+      "learning_rate": 0.0005025913773501421,
+      "loss": 2.9144,
+      "step": 4011
+    },
+    {
+      "epoch": 0.513536,
+      "grad_norm": 0.42244938015937805,
+      "learning_rate": 0.0005023840688015291,
+      "loss": 2.8455,
+      "step": 4012
+    },
+    {
+      "epoch": 0.513664,
+      "grad_norm": 0.4192439913749695,
+      "learning_rate": 0.0005021767598430671,
+      "loss": 2.9425,
+      "step": 4013
+    },
+    {
+      "epoch": 0.513792,
+      "grad_norm": 0.4251191020011902,
+      "learning_rate": 0.0005019694505103948,
+      "loss": 2.9345,
+      "step": 4014
+    },
+    {
+      "epoch": 0.51392,
+      "grad_norm": 0.4381127953529358,
+      "learning_rate": 0.0005017621408391513,
+      "loss": 2.9377,
+      "step": 4015
+    },
+    {
+      "epoch": 0.514048,
+      "grad_norm": 0.39507895708084106,
+      "learning_rate": 0.0005015548308649752,
+      "loss": 2.8529,
+      "step": 4016
+    },
+    {
+      "epoch": 0.514176,
+      "grad_norm": 0.41748204827308655,
+      "learning_rate": 0.0005013475206235056,
+      "loss": 2.9449,
+      "step": 4017
+    },
+    {
+      "epoch": 0.514304,
+      "grad_norm": 0.42555704712867737,
+      "learning_rate": 0.0005011402101503818,
+      "loss": 2.7619,
+      "step": 4018
+    },
+    {
+      "epoch": 0.514432,
+      "grad_norm": 0.4237884283065796,
+      "learning_rate": 0.0005009328994812425,
+      "loss": 2.8551,
+      "step": 4019
+    },
+    {
+      "epoch": 0.51456,
+      "grad_norm": 0.42393386363983154,
+      "learning_rate": 0.000500725588651727,
+      "loss": 2.7831,
+      "step": 4020
+    },
+    {
+      "epoch": 0.514688,
+      "grad_norm": 0.4221552610397339,
+      "learning_rate": 0.0005005182776974745,
+      "loss": 2.8548,
+      "step": 4021
+    },
+    {
+      "epoch": 0.514816,
+      "grad_norm": 0.4380204677581787,
+      "learning_rate": 0.0005003109666541238,
+      "loss": 2.9378,
+      "step": 4022
+    },
+    {
+      "epoch": 0.514944,
+      "grad_norm": 0.4485597610473633,
+      "learning_rate": 0.0005001036555573145,
+      "loss": 2.8011,
+      "step": 4023
+    },
+    {
+      "epoch": 0.515072,
+      "grad_norm": 0.4335438013076782,
+      "learning_rate": 0.0004998963444426855,
+      "loss": 2.8066,
+      "step": 4024
+    },
+    {
+      "epoch": 0.5152,
+      "grad_norm": 0.43467962741851807,
+      "learning_rate": 0.0004996890333458762,
+      "loss": 2.7679,
+      "step": 4025
+    },
+    {
+      "epoch": 0.515328,
+      "grad_norm": 0.4134734869003296,
+      "learning_rate": 0.0004994817223025256,
+      "loss": 2.8937,
+      "step": 4026
+    },
+    {
+      "epoch": 0.515456,
+      "grad_norm": 0.4034424126148224,
+      "learning_rate": 0.000499274411348273,
+      "loss": 3.0256,
+      "step": 4027
+    },
+    {
+      "epoch": 0.515584,
+      "grad_norm": 0.3807096481323242,
+      "learning_rate": 0.0004990671005187576,
+      "loss": 2.8368,
+      "step": 4028
+    },
+    {
+      "epoch": 0.515712,
+      "grad_norm": 0.41547316312789917,
+      "learning_rate": 0.0004988597898496183,
+      "loss": 2.903,
+      "step": 4029
+    },
+    {
+      "epoch": 0.51584,
+      "grad_norm": 0.42907509207725525,
+      "learning_rate": 0.0004986524793764945,
+      "loss": 2.7621,
+      "step": 4030
+    },
+    {
+      "epoch": 0.515968,
+      "grad_norm": 0.4107275903224945,
+      "learning_rate": 0.0004984451691350249,
+      "loss": 2.8177,
+      "step": 4031
+    },
+    {
+      "epoch": 0.516096,
+      "grad_norm": 0.4543997049331665,
+      "learning_rate": 0.0004982378591608488,
+      "loss": 2.9048,
+      "step": 4032
+    },
+    {
+      "epoch": 0.516224,
+      "grad_norm": 0.44772613048553467,
+      "learning_rate": 0.0004980305494896052,
+      "loss": 2.8449,
+      "step": 4033
+    },
+    {
+      "epoch": 0.516352,
+      "grad_norm": 0.4583573341369629,
+      "learning_rate": 0.0004978232401569329,
+      "loss": 2.7944,
+      "step": 4034
+    },
+    {
+      "epoch": 0.51648,
+      "grad_norm": 0.43442896008491516,
+      "learning_rate": 0.000497615931198471,
+      "loss": 2.9644,
+      "step": 4035
+    },
+    {
+      "epoch": 0.516608,
+      "grad_norm": 0.43407952785491943,
+      "learning_rate": 0.0004974086226498581,
+      "loss": 2.8197,
+      "step": 4036
+    },
+    {
+      "epoch": 0.516736,
+      "grad_norm": 0.3900356888771057,
+      "learning_rate": 0.0004972013145467329,
+      "loss": 2.9642,
+      "step": 4037
+    },
+    {
+      "epoch": 0.516864,
+      "grad_norm": 0.41613754630088806,
+      "learning_rate": 0.0004969940069247343,
+      "loss": 2.8454,
+      "step": 4038
+    },
+    {
+      "epoch": 0.516992,
+      "grad_norm": 0.4340950548648834,
+      "learning_rate": 0.0004967866998195008,
+      "loss": 2.8849,
+      "step": 4039
+    },
+    {
+      "epoch": 0.51712,
+      "grad_norm": 0.42494475841522217,
+      "learning_rate": 0.0004965793932666707,
+      "loss": 2.8871,
+      "step": 4040
+    },
+    {
+      "epoch": 0.517248,
+      "grad_norm": 0.4166974723339081,
+      "learning_rate": 0.0004963720873018827,
+      "loss": 3.0212,
+      "step": 4041
+    },
+    {
+      "epoch": 0.517376,
+      "grad_norm": 0.45957180857658386,
+      "learning_rate": 0.0004961647819607749,
+      "loss": 2.9388,
+      "step": 4042
+    },
+    {
+      "epoch": 0.517504,
+      "grad_norm": 0.416073203086853,
+      "learning_rate": 0.0004959574772789853,
+      "loss": 2.8648,
+      "step": 4043
+    },
+    {
+      "epoch": 0.517632,
+      "grad_norm": 0.41458383202552795,
+      "learning_rate": 0.0004957501732921524,
+      "loss": 2.8654,
+      "step": 4044
+    },
+    {
+      "epoch": 0.51776,
+      "grad_norm": 0.404540091753006,
+      "learning_rate": 0.000495542870035914,
+      "loss": 2.9508,
+      "step": 4045
+    },
+    {
+      "epoch": 0.517888,
+      "grad_norm": 0.45716553926467896,
+      "learning_rate": 0.000495335567545908,
+      "loss": 2.8988,
+      "step": 4046
+    },
+    {
+      "epoch": 0.518016,
+      "grad_norm": 0.4163787364959717,
+      "learning_rate": 0.0004951282658577718,
+      "loss": 2.7841,
+      "step": 4047
+    },
+    {
+      "epoch": 0.518144,
+      "grad_norm": 0.41052183508872986,
+      "learning_rate": 0.0004949209650071435,
+      "loss": 2.8654,
+      "step": 4048
+    },
+    {
+      "epoch": 0.518272,
+      "grad_norm": 0.42903000116348267,
+      "learning_rate": 0.0004947136650296601,
+      "loss": 2.8879,
+      "step": 4049
+    },
+    {
+      "epoch": 0.5184,
+      "grad_norm": 0.436210960149765,
+      "learning_rate": 0.0004945063659609587,
+      "loss": 2.8224,
+      "step": 4050
+    },
+    {
+      "epoch": 0.518528,
+      "grad_norm": 0.3764611482620239,
+      "learning_rate": 0.0004942990678366769,
+      "loss": 2.8091,
+      "step": 4051
+    },
+    {
+      "epoch": 0.518656,
+      "grad_norm": 0.3941909968852997,
+      "learning_rate": 0.0004940917706924513,
+      "loss": 2.9121,
+      "step": 4052
+    },
+    {
+      "epoch": 0.518784,
+      "grad_norm": 0.47499871253967285,
+      "learning_rate": 0.000493884474563919,
+      "loss": 2.9941,
+      "step": 4053
+    },
+    {
+      "epoch": 0.518912,
+      "grad_norm": 0.4053983986377716,
+      "learning_rate": 0.0004936771794867163,
+      "loss": 2.826,
+      "step": 4054
+    },
+    {
+      "epoch": 0.51904,
+      "grad_norm": 0.3915752172470093,
+      "learning_rate": 0.0004934698854964797,
+      "loss": 2.9069,
+      "step": 4055
+    },
+    {
+      "epoch": 0.519168,
+      "grad_norm": 0.47227537631988525,
+      "learning_rate": 0.0004932625926288456,
+      "loss": 3.0509,
+      "step": 4056
+    },
+    {
+      "epoch": 0.519296,
+      "grad_norm": 0.4294140040874481,
+      "learning_rate": 0.0004930553009194499,
+      "loss": 2.9438,
+      "step": 4057
+    },
+    {
+      "epoch": 0.519424,
+      "grad_norm": 0.4197215139865875,
+      "learning_rate": 0.0004928480104039284,
+      "loss": 2.8857,
+      "step": 4058
+    },
+    {
+      "epoch": 0.519552,
+      "grad_norm": 0.4169391393661499,
+      "learning_rate": 0.000492640721117917,
+      "loss": 2.8528,
+      "step": 4059
+    },
+    {
+      "epoch": 0.51968,
+      "grad_norm": 0.3908802568912506,
+      "learning_rate": 0.0004924334330970509,
+      "loss": 2.7897,
+      "step": 4060
+    },
+    {
+      "epoch": 0.519808,
+      "grad_norm": 0.4110635221004486,
+      "learning_rate": 0.0004922261463769651,
+      "loss": 3.03,
+      "step": 4061
+    },
+    {
+      "epoch": 0.519936,
+      "grad_norm": 0.40013009309768677,
+      "learning_rate": 0.0004920188609932951,
+      "loss": 2.7316,
+      "step": 4062
+    },
+    {
+      "epoch": 0.520064,
+      "grad_norm": 0.4293777644634247,
+      "learning_rate": 0.0004918115769816754,
+      "loss": 2.7864,
+      "step": 4063
+    },
+    {
+      "epoch": 0.520192,
+      "grad_norm": 0.4275892674922943,
+      "learning_rate": 0.0004916042943777406,
+      "loss": 2.8582,
+      "step": 4064
+    },
+    {
+      "epoch": 0.52032,
+      "grad_norm": 0.3951684236526489,
+      "learning_rate": 0.0004913970132171248,
+      "loss": 2.8512,
+      "step": 4065
+    },
+    {
+      "epoch": 0.520448,
+      "grad_norm": 0.4252420961856842,
+      "learning_rate": 0.0004911897335354624,
+      "loss": 2.8927,
+      "step": 4066
+    },
+    {
+      "epoch": 0.520576,
+      "grad_norm": 0.4033498466014862,
+      "learning_rate": 0.0004909824553683867,
+      "loss": 2.8881,
+      "step": 4067
+    },
+    {
+      "epoch": 0.520704,
+      "grad_norm": 0.4142298102378845,
+      "learning_rate": 0.0004907751787515315,
+      "loss": 2.795,
+      "step": 4068
+    },
+    {
+      "epoch": 0.520832,
+      "grad_norm": 0.39789843559265137,
+      "learning_rate": 0.0004905679037205302,
+      "loss": 2.8505,
+      "step": 4069
+    },
+    {
+      "epoch": 0.52096,
+      "grad_norm": 0.43670454621315,
+      "learning_rate": 0.0004903606303110152,
+      "loss": 2.8711,
+      "step": 4070
+    },
+    {
+      "epoch": 0.521088,
+      "grad_norm": 0.4129983186721802,
+      "learning_rate": 0.0004901533585586199,
+      "loss": 2.7573,
+      "step": 4071
+    },
+    {
+      "epoch": 0.521216,
+      "grad_norm": 0.40850868821144104,
+      "learning_rate": 0.0004899460884989764,
+      "loss": 2.8255,
+      "step": 4072
+    },
+    {
+      "epoch": 0.521344,
+      "grad_norm": 0.40776461362838745,
+      "learning_rate": 0.0004897388201677168,
+      "loss": 2.8605,
+      "step": 4073
+    },
+    {
+      "epoch": 0.521472,
+      "grad_norm": 0.4344709813594818,
+      "learning_rate": 0.000489531553600473,
+      "loss": 2.845,
+      "step": 4074
+    },
+    {
+      "epoch": 0.5216,
+      "grad_norm": 0.3882288932800293,
+      "learning_rate": 0.0004893242888328765,
+      "loss": 2.8151,
+      "step": 4075
+    },
+    {
+      "epoch": 0.521728,
+      "grad_norm": 0.4378066956996918,
+      "learning_rate": 0.0004891170259005585,
+      "loss": 2.9387,
+      "step": 4076
+    },
+    {
+      "epoch": 0.521856,
+      "grad_norm": 0.4379422664642334,
+      "learning_rate": 0.00048890976483915,
+      "loss": 2.7864,
+      "step": 4077
+    },
+    {
+      "epoch": 0.521984,
+      "grad_norm": 0.4040283262729645,
+      "learning_rate": 0.0004887025056842814,
+      "loss": 2.9328,
+      "step": 4078
+    },
+    {
+      "epoch": 0.522112,
+      "grad_norm": 0.44158414006233215,
+      "learning_rate": 0.000488495248471583,
+      "loss": 2.8969,
+      "step": 4079
+    },
+    {
+      "epoch": 0.52224,
+      "grad_norm": 0.41905903816223145,
+      "learning_rate": 0.0004882879932366846,
+      "loss": 2.8873,
+      "step": 4080
+    },
+    {
+      "epoch": 0.522368,
+      "grad_norm": 0.4367314577102661,
+      "learning_rate": 0.0004880807400152162,
+      "loss": 2.941,
+      "step": 4081
+    },
+    {
+      "epoch": 0.522496,
+      "grad_norm": 0.4216005802154541,
+      "learning_rate": 0.0004878734888428068,
+      "loss": 2.9567,
+      "step": 4082
+    },
+    {
+      "epoch": 0.522624,
+      "grad_norm": 0.43903887271881104,
+      "learning_rate": 0.0004876662397550851,
+      "loss": 2.9206,
+      "step": 4083
+    },
+    {
+      "epoch": 0.522752,
+      "grad_norm": 0.44283315539360046,
+      "learning_rate": 0.00048745899278767983,
+      "loss": 2.977,
+      "step": 4084
+    },
+    {
+      "epoch": 0.52288,
+      "grad_norm": 0.424873411655426,
+      "learning_rate": 0.00048725174797621917,
+      "loss": 2.8545,
+      "step": 4085
+    },
+    {
+      "epoch": 0.523008,
+      "grad_norm": 0.4066339433193207,
+      "learning_rate": 0.00048704450535633074,
+      "loss": 2.763,
+      "step": 4086
+    },
+    {
+      "epoch": 0.523136,
+      "grad_norm": 0.39463233947753906,
+      "learning_rate": 0.00048683726496364207,
+      "loss": 2.8248,
+      "step": 4087
+    },
+    {
+      "epoch": 0.523264,
+      "grad_norm": 0.4205288887023926,
+      "learning_rate": 0.00048663002683378015,
+      "loss": 2.837,
+      "step": 4088
+    },
+    {
+      "epoch": 0.523392,
+      "grad_norm": 0.4102119505405426,
+      "learning_rate": 0.00048642279100237143,
+      "loss": 2.8392,
+      "step": 4089
+    },
+    {
+      "epoch": 0.52352,
+      "grad_norm": 0.3912118971347809,
+      "learning_rate": 0.00048621555750504253,
+      "loss": 2.7117,
+      "step": 4090
+    },
+    {
+      "epoch": 0.523648,
+      "grad_norm": 0.4242720603942871,
+      "learning_rate": 0.000486008326377419,
+      "loss": 2.7952,
+      "step": 4091
+    },
+    {
+      "epoch": 0.523776,
+      "grad_norm": 0.4401116669178009,
+      "learning_rate": 0.0004858010976551265,
+      "loss": 2.8401,
+      "step": 4092
+    },
+    {
+      "epoch": 0.523904,
+      "grad_norm": 0.43263188004493713,
+      "learning_rate": 0.0004855938713737899,
+      "loss": 2.9477,
+      "step": 4093
+    },
+    {
+      "epoch": 0.524032,
+      "grad_norm": 0.4307785630226135,
+      "learning_rate": 0.00048538664756903375,
+      "loss": 2.888,
+      "step": 4094
+    },
+    {
+      "epoch": 0.52416,
+      "grad_norm": 0.5081241130828857,
+      "learning_rate": 0.0004851794262764823,
+      "loss": 2.8549,
+      "step": 4095
+    },
+    {
+      "epoch": 0.524288,
+      "grad_norm": 0.43126291036605835,
+      "learning_rate": 0.00048497220753175943,
+      "loss": 2.8056,
+      "step": 4096
+    },
+    {
+      "epoch": 0.524416,
+      "grad_norm": 0.41626474261283875,
+      "learning_rate": 0.0004847649913704881,
+      "loss": 2.7943,
+      "step": 4097
+    },
+    {
+      "epoch": 0.524544,
+      "grad_norm": 0.4101981818675995,
+      "learning_rate": 0.00048455777782829124,
+      "loss": 2.7271,
+      "step": 4098
+    },
+    {
+      "epoch": 0.524672,
+      "grad_norm": 0.3986057639122009,
+      "learning_rate": 0.00048435056694079166,
+      "loss": 2.9595,
+      "step": 4099
+    },
+    {
+      "epoch": 0.5248,
+      "grad_norm": 0.4265193045139313,
+      "learning_rate": 0.000484143358743611,
+      "loss": 2.8354,
+      "step": 4100
+    },
+    {
+      "epoch": 0.524928,
+      "grad_norm": 0.44263792037963867,
+      "learning_rate": 0.0004839361532723708,
+      "loss": 2.8174,
+      "step": 4101
+    },
+    {
+      "epoch": 0.525056,
+      "grad_norm": 0.44561734795570374,
+      "learning_rate": 0.0004837289505626921,
+      "loss": 2.938,
+      "step": 4102
+    },
+    {
+      "epoch": 0.525184,
+      "grad_norm": 0.4417126774787903,
+      "learning_rate": 0.0004835217506501954,
+      "loss": 2.9193,
+      "step": 4103
+    },
+    {
+      "epoch": 0.525312,
+      "grad_norm": 0.42325541377067566,
+      "learning_rate": 0.0004833145535705008,
+      "loss": 2.9017,
+      "step": 4104
+    },
+    {
+      "epoch": 0.52544,
+      "grad_norm": 0.4226759672164917,
+      "learning_rate": 0.0004831073593592279,
+      "loss": 2.8398,
+      "step": 4105
+    },
+    {
+      "epoch": 0.525568,
+      "grad_norm": 0.4147658944129944,
+      "learning_rate": 0.0004829001680519958,
+      "loss": 2.7592,
+      "step": 4106
+    },
+    {
+      "epoch": 0.525696,
+      "grad_norm": 0.4165031909942627,
+      "learning_rate": 0.0004826929796844228,
+      "loss": 2.7718,
+      "step": 4107
+    },
+    {
+      "epoch": 0.525824,
+      "grad_norm": 0.4314729571342468,
+      "learning_rate": 0.00048248579429212735,
+      "loss": 2.8703,
+      "step": 4108
+    },
+    {
+      "epoch": 0.525952,
+      "grad_norm": 0.43153223395347595,
+      "learning_rate": 0.00048227861191072705,
+      "loss": 2.9683,
+      "step": 4109
+    },
+    {
+      "epoch": 0.52608,
+      "grad_norm": 0.44214507937431335,
+      "learning_rate": 0.00048207143257583874,
+      "loss": 2.8647,
+      "step": 4110
+    },
+    {
+      "epoch": 0.526208,
+      "grad_norm": 0.38847607374191284,
+      "learning_rate": 0.00048186425632307906,
+      "loss": 3.0193,
+      "step": 4111
+    },
+    {
+      "epoch": 0.526336,
+      "grad_norm": 0.4789884388446808,
+      "learning_rate": 0.00048165708318806383,
+      "loss": 2.982,
+      "step": 4112
+    },
+    {
+      "epoch": 0.526464,
+      "grad_norm": 0.4402121901512146,
+      "learning_rate": 0.00048144991320640877,
+      "loss": 2.8378,
+      "step": 4113
+    },
+    {
+      "epoch": 0.526592,
+      "grad_norm": 0.3972822427749634,
+      "learning_rate": 0.00048124274641372873,
+      "loss": 2.8411,
+      "step": 4114
+    },
+    {
+      "epoch": 0.52672,
+      "grad_norm": 0.4221082925796509,
+      "learning_rate": 0.00048103558284563785,
+      "loss": 2.8785,
+      "step": 4115
+    },
+    {
+      "epoch": 0.526848,
+      "grad_norm": 0.40344661474227905,
+      "learning_rate": 0.00048082842253775027,
+      "loss": 2.8082,
+      "step": 4116
+    },
+    {
+      "epoch": 0.526976,
+      "grad_norm": 0.5079003572463989,
+      "learning_rate": 0.0004806212655256789,
+      "loss": 2.7434,
+      "step": 4117
+    },
+    {
+      "epoch": 0.527104,
+      "grad_norm": 0.43321117758750916,
+      "learning_rate": 0.00048041411184503684,
+      "loss": 2.8603,
+      "step": 4118
+    },
+    {
+      "epoch": 0.527232,
+      "grad_norm": 0.43519389629364014,
+      "learning_rate": 0.00048020696153143596,
+      "loss": 2.8195,
+      "step": 4119
+    },
+    {
+      "epoch": 0.52736,
+      "grad_norm": 0.4327753782272339,
+      "learning_rate": 0.0004799998146204879,
+      "loss": 2.9207,
+      "step": 4120
+    },
+    {
+      "epoch": 0.527488,
+      "grad_norm": 0.5465975999832153,
+      "learning_rate": 0.0004797926711478036,
+      "loss": 2.7775,
+      "step": 4121
+    },
+    {
+      "epoch": 0.527616,
+      "grad_norm": 0.46385103464126587,
+      "learning_rate": 0.00047958553114899326,
+      "loss": 2.7791,
+      "step": 4122
+    },
+    {
+      "epoch": 0.527744,
+      "grad_norm": 0.4153241813182831,
+      "learning_rate": 0.0004793783946596669,
+      "loss": 2.8023,
+      "step": 4123
+    },
+    {
+      "epoch": 0.527872,
+      "grad_norm": 0.4289769232273102,
+      "learning_rate": 0.0004791712617154336,
+      "loss": 2.7661,
+      "step": 4124
+    },
+    {
+      "epoch": 0.528,
+      "grad_norm": 0.41764822602272034,
+      "learning_rate": 0.00047896413235190175,
+      "loss": 2.6169,
+      "step": 4125
+    },
+    {
+      "epoch": 0.528128,
+      "grad_norm": 0.4486224055290222,
+      "learning_rate": 0.0004787570066046793,
+      "loss": 2.9542,
+      "step": 4126
+    },
+    {
+      "epoch": 0.528256,
+      "grad_norm": 0.4346499741077423,
+      "learning_rate": 0.0004785498845093739,
+      "loss": 2.8651,
+      "step": 4127
+    },
+    {
+      "epoch": 0.528384,
+      "grad_norm": 0.4521925449371338,
+      "learning_rate": 0.000478342766101592,
+      "loss": 2.9479,
+      "step": 4128
+    },
+    {
+      "epoch": 0.528512,
+      "grad_norm": 0.41116029024124146,
+      "learning_rate": 0.0004781356514169396,
+      "loss": 2.7968,
+      "step": 4129
+    },
+    {
+      "epoch": 0.52864,
+      "grad_norm": 0.415486216545105,
+      "learning_rate": 0.0004779285404910223,
+      "loss": 2.9631,
+      "step": 4130
+    },
+    {
+      "epoch": 0.528768,
+      "grad_norm": 0.39223119616508484,
+      "learning_rate": 0.0004777214333594448,
+      "loss": 2.8769,
+      "step": 4131
+    },
+    {
+      "epoch": 0.528896,
+      "grad_norm": 0.4170284569263458,
+      "learning_rate": 0.0004775143300578111,
+      "loss": 2.8922,
+      "step": 4132
+    },
+    {
+      "epoch": 0.529024,
+      "grad_norm": 0.39090612530708313,
+      "learning_rate": 0.0004773072306217247,
+      "loss": 2.9152,
+      "step": 4133
+    },
+    {
+      "epoch": 0.529152,
+      "grad_norm": 0.4349527955055237,
+      "learning_rate": 0.0004771001350867885,
+      "loss": 2.8545,
+      "step": 4134
+    },
+    {
+      "epoch": 0.52928,
+      "grad_norm": 0.37919172644615173,
+      "learning_rate": 0.00047689304348860436,
+      "loss": 2.9677,
+      "step": 4135
+    },
+    {
+      "epoch": 0.529408,
+      "grad_norm": 0.41449713706970215,
+      "learning_rate": 0.00047668595586277415,
+      "loss": 2.8615,
+      "step": 4136
+    },
+    {
+      "epoch": 0.529536,
+      "grad_norm": 0.3776123821735382,
+      "learning_rate": 0.0004764788722448983,
+      "loss": 2.877,
+      "step": 4137
+    },
+    {
+      "epoch": 0.529664,
+      "grad_norm": 0.412131130695343,
+      "learning_rate": 0.0004762717926705771,
+      "loss": 2.841,
+      "step": 4138
+    },
+    {
+      "epoch": 0.529792,
+      "grad_norm": 0.43055853247642517,
+      "learning_rate": 0.00047606471717540977,
+      "loss": 2.8049,
+      "step": 4139
+    },
+    {
+      "epoch": 0.52992,
+      "grad_norm": 0.4111311733722687,
+      "learning_rate": 0.00047585764579499495,
+      "loss": 2.7499,
+      "step": 4140
+    },
+    {
+      "epoch": 0.530048,
+      "grad_norm": 0.4218684732913971,
+      "learning_rate": 0.0004756505785649308,
+      "loss": 2.8219,
+      "step": 4141
+    },
+    {
+      "epoch": 0.530176,
+      "grad_norm": 0.42546480894088745,
+      "learning_rate": 0.00047544351552081455,
+      "loss": 2.9466,
+      "step": 4142
+    },
+    {
+      "epoch": 0.530304,
+      "grad_norm": 0.38852623105049133,
+      "learning_rate": 0.0004752364566982425,
+      "loss": 2.7126,
+      "step": 4143
+    },
+    {
+      "epoch": 0.530432,
+      "grad_norm": 0.3946860432624817,
+      "learning_rate": 0.0004750294021328105,
+      "loss": 2.688,
+      "step": 4144
+    },
+    {
+      "epoch": 0.53056,
+      "grad_norm": 0.40264514088630676,
+      "learning_rate": 0.00047482235186011403,
+      "loss": 2.9768,
+      "step": 4145
+    },
+    {
+      "epoch": 0.530688,
+      "grad_norm": 0.4401693046092987,
+      "learning_rate": 0.0004746153059157471,
+      "loss": 2.7333,
+      "step": 4146
+    },
+    {
+      "epoch": 0.530816,
+      "grad_norm": 0.4133301377296448,
+      "learning_rate": 0.0004744082643353033,
+      "loss": 2.9699,
+      "step": 4147
+    },
+    {
+      "epoch": 0.530944,
+      "grad_norm": 0.3945316970348358,
+      "learning_rate": 0.00047420122715437565,
+      "loss": 2.8681,
+      "step": 4148
+    },
+    {
+      "epoch": 0.531072,
+      "grad_norm": 0.40803050994873047,
+      "learning_rate": 0.00047399419440855614,
+      "loss": 2.7607,
+      "step": 4149
+    },
+    {
+      "epoch": 0.5312,
+      "grad_norm": 0.42406195402145386,
+      "learning_rate": 0.000473787166133436,
+      "loss": 2.7753,
+      "step": 4150
+    },
+    {
+      "epoch": 0.531328,
+      "grad_norm": 0.4014380872249603,
+      "learning_rate": 0.000473580142364606,
+      "loss": 2.9433,
+      "step": 4151
+    },
+    {
+      "epoch": 0.531456,
+      "grad_norm": 0.4244258999824524,
+      "learning_rate": 0.00047337312313765575,
+      "loss": 2.9291,
+      "step": 4152
+    },
+    {
+      "epoch": 0.531584,
+      "grad_norm": 0.43073782324790955,
+      "learning_rate": 0.00047316610848817435,
+      "loss": 2.9224,
+      "step": 4153
+    },
+    {
+      "epoch": 0.531712,
+      "grad_norm": 0.4392019510269165,
+      "learning_rate": 0.0004729590984517498,
+      "loss": 2.8982,
+      "step": 4154
+    },
+    {
+      "epoch": 0.53184,
+      "grad_norm": 0.4326994717121124,
+      "learning_rate": 0.0004727520930639698,
+      "loss": 2.8741,
+      "step": 4155
+    },
+    {
+      "epoch": 0.531968,
+      "grad_norm": 0.48569077253341675,
+      "learning_rate": 0.00047254509236042096,
+      "loss": 2.8039,
+      "step": 4156
+    },
+    {
+      "epoch": 0.532096,
+      "grad_norm": 0.42556121945381165,
+      "learning_rate": 0.00047233809637668906,
+      "loss": 2.9389,
+      "step": 4157
+    },
+    {
+      "epoch": 0.532224,
+      "grad_norm": 0.4398675858974457,
+      "learning_rate": 0.000472131105148359,
+      "loss": 2.8574,
+      "step": 4158
+    },
+    {
+      "epoch": 0.532352,
+      "grad_norm": 0.46172580122947693,
+      "learning_rate": 0.0004719241187110151,
+      "loss": 2.8926,
+      "step": 4159
+    },
+    {
+      "epoch": 0.53248,
+      "grad_norm": 0.4075527489185333,
+      "learning_rate": 0.00047171713710024074,
+      "loss": 2.9678,
+      "step": 4160
+    },
+    {
+      "epoch": 0.532608,
+      "grad_norm": 0.42629051208496094,
+      "learning_rate": 0.0004715101603516182,
+      "loss": 2.8543,
+      "step": 4161
+    },
+    {
+      "epoch": 0.532736,
+      "grad_norm": 0.3909108340740204,
+      "learning_rate": 0.0004713031885007296,
+      "loss": 2.8317,
+      "step": 4162
+    },
+    {
+      "epoch": 0.532864,
+      "grad_norm": 0.41872966289520264,
+      "learning_rate": 0.00047109622158315526,
+      "loss": 2.9045,
+      "step": 4163
+    },
+    {
+      "epoch": 0.532992,
+      "grad_norm": 0.411870539188385,
+      "learning_rate": 0.00047088925963447575,
+      "loss": 2.9199,
+      "step": 4164
+    },
+    {
+      "epoch": 0.53312,
+      "grad_norm": 0.40755483508110046,
+      "learning_rate": 0.0004706823026902699,
+      "loss": 2.8242,
+      "step": 4165
+    },
+    {
+      "epoch": 0.533248,
+      "grad_norm": 0.44240453839302063,
+      "learning_rate": 0.0004704753507861162,
+      "loss": 2.8387,
+      "step": 4166
+    },
+    {
+      "epoch": 0.533376,
+      "grad_norm": 0.43574488162994385,
+      "learning_rate": 0.00047026840395759196,
+      "loss": 2.7356,
+      "step": 4167
+    },
+    {
+      "epoch": 0.533504,
+      "grad_norm": 0.42288506031036377,
+      "learning_rate": 0.0004700614622402737,
+      "loss": 2.8417,
+      "step": 4168
+    },
+    {
+      "epoch": 0.533632,
+      "grad_norm": 0.4940197765827179,
+      "learning_rate": 0.00046985452566973725,
+      "loss": 2.9641,
+      "step": 4169
+    },
+    {
+      "epoch": 0.53376,
+      "grad_norm": 0.3987281024456024,
+      "learning_rate": 0.0004696475942815573,
+      "loss": 2.9567,
+      "step": 4170
+    },
+    {
+      "epoch": 0.533888,
+      "grad_norm": 0.42762672901153564,
+      "learning_rate": 0.00046944066811130776,
+      "loss": 2.8225,
+      "step": 4171
+    },
+    {
+      "epoch": 0.534016,
+      "grad_norm": 0.44472092390060425,
+      "learning_rate": 0.00046923374719456143,
+      "loss": 2.8293,
+      "step": 4172
+    },
+    {
+      "epoch": 0.534144,
+      "grad_norm": 0.4105190336704254,
+      "learning_rate": 0.00046902683156689075,
+      "loss": 2.811,
+      "step": 4173
+    },
+    {
+      "epoch": 0.534272,
+      "grad_norm": 0.403595894575119,
+      "learning_rate": 0.0004688199212638669,
+      "loss": 2.8791,
+      "step": 4174
+    },
+    {
+      "epoch": 0.5344,
+      "grad_norm": 0.42785388231277466,
+      "learning_rate": 0.00046861301632106,
+      "loss": 2.8826,
+      "step": 4175
+    },
+    {
+      "epoch": 0.534528,
+      "grad_norm": 0.42342400550842285,
+      "learning_rate": 0.00046840611677403943,
+      "loss": 3.001,
+      "step": 4176
+    },
+    {
+      "epoch": 0.534656,
+      "grad_norm": 0.40331363677978516,
+      "learning_rate": 0.00046819922265837376,
+      "loss": 2.8438,
+      "step": 4177
+    },
+    {
+      "epoch": 0.534784,
+      "grad_norm": 0.4263276159763336,
+      "learning_rate": 0.00046799233400963037,
+      "loss": 2.9253,
+      "step": 4178
+    },
+    {
+      "epoch": 0.534912,
+      "grad_norm": 0.44182291626930237,
+      "learning_rate": 0.0004677854508633757,
+      "loss": 2.9471,
+      "step": 4179
+    },
+    {
+      "epoch": 0.53504,
+      "grad_norm": 0.4070250988006592,
+      "learning_rate": 0.0004675785732551756,
+      "loss": 2.8834,
+      "step": 4180
+    },
+    {
+      "epoch": 0.535168,
+      "grad_norm": 0.4037683606147766,
+      "learning_rate": 0.00046737170122059446,
+      "loss": 2.8948,
+      "step": 4181
+    },
+    {
+      "epoch": 0.535296,
+      "grad_norm": 0.48121172189712524,
+      "learning_rate": 0.0004671648347951963,
+      "loss": 2.7828,
+      "step": 4182
+    },
+    {
+      "epoch": 0.535424,
+      "grad_norm": 0.41399574279785156,
+      "learning_rate": 0.0004669579740145436,
+      "loss": 2.8557,
+      "step": 4183
+    },
+    {
+      "epoch": 0.535552,
+      "grad_norm": 0.4016878604888916,
+      "learning_rate": 0.0004667511189141983,
+      "loss": 2.7757,
+      "step": 4184
+    },
+    {
+      "epoch": 0.53568,
+      "grad_norm": 0.4502201974391937,
+      "learning_rate": 0.0004665442695297212,
+      "loss": 2.9113,
+      "step": 4185
+    },
+    {
+      "epoch": 0.535808,
+      "grad_norm": 0.4380075931549072,
+      "learning_rate": 0.0004663374258966718,
+      "loss": 2.7977,
+      "step": 4186
+    },
+    {
+      "epoch": 0.535936,
+      "grad_norm": 0.4172612726688385,
+      "learning_rate": 0.00046613058805060933,
+      "loss": 2.86,
+      "step": 4187
+    },
+    {
+      "epoch": 0.536064,
+      "grad_norm": 0.41308149695396423,
+      "learning_rate": 0.0004659237560270914,
+      "loss": 2.8923,
+      "step": 4188
+    },
+    {
+      "epoch": 0.536192,
+      "grad_norm": 0.4179181754589081,
+      "learning_rate": 0.0004657169298616747,
+      "loss": 2.8824,
+      "step": 4189
+    },
+    {
+      "epoch": 0.53632,
+      "grad_norm": 0.4401620030403137,
+      "learning_rate": 0.00046551010958991536,
+      "loss": 3.0333,
+      "step": 4190
+    },
+    {
+      "epoch": 0.536448,
+      "grad_norm": 0.41828975081443787,
+      "learning_rate": 0.0004653032952473678,
+      "loss": 2.769,
+      "step": 4191
+    },
+    {
+      "epoch": 0.536576,
+      "grad_norm": 0.41021856665611267,
+      "learning_rate": 0.0004650964868695863,
+      "loss": 2.8394,
+      "step": 4192
+    },
+    {
+      "epoch": 0.536704,
+      "grad_norm": 0.44685450196266174,
+      "learning_rate": 0.0004648896844921232,
+      "loss": 2.9306,
+      "step": 4193
+    },
+    {
+      "epoch": 0.536832,
+      "grad_norm": 0.4502381682395935,
+      "learning_rate": 0.0004646828881505304,
+      "loss": 2.8722,
+      "step": 4194
+    },
+    {
+      "epoch": 0.53696,
+      "grad_norm": 0.4522649943828583,
+      "learning_rate": 0.0004644760978803586,
+      "loss": 2.7638,
+      "step": 4195
+    },
+    {
+      "epoch": 0.537088,
+      "grad_norm": 0.4058089554309845,
+      "learning_rate": 0.0004642693137171574,
+      "loss": 2.795,
+      "step": 4196
+    },
+    {
+      "epoch": 0.537216,
+      "grad_norm": 0.41998353600502014,
+      "learning_rate": 0.0004640625356964753,
+      "loss": 2.7229,
+      "step": 4197
+    },
+    {
+      "epoch": 0.537344,
+      "grad_norm": 0.39574572443962097,
+      "learning_rate": 0.00046385576385386004,
+      "loss": 2.8818,
+      "step": 4198
+    },
+    {
+      "epoch": 0.537472,
+      "grad_norm": 0.4094662070274353,
+      "learning_rate": 0.000463648998224858,
+      "loss": 2.6937,
+      "step": 4199
+    },
+    {
+      "epoch": 0.5376,
+      "grad_norm": 0.43437787890434265,
+      "learning_rate": 0.00046344223884501435,
+      "loss": 2.7331,
+      "step": 4200
+    },
+    {
+      "epoch": 0.537728,
+      "grad_norm": 0.44544631242752075,
+      "learning_rate": 0.0004632354857498738,
+      "loss": 2.976,
+      "step": 4201
+    },
+    {
+      "epoch": 0.537856,
+      "grad_norm": 0.4459388554096222,
+      "learning_rate": 0.00046302873897497956,
+      "loss": 2.8304,
+      "step": 4202
+    },
+    {
+      "epoch": 0.537984,
+      "grad_norm": 0.43447548151016235,
+      "learning_rate": 0.00046282199855587365,
+      "loss": 2.9727,
+      "step": 4203
+    },
+    {
+      "epoch": 0.538112,
+      "grad_norm": 0.42178866267204285,
+      "learning_rate": 0.00046261526452809717,
+      "loss": 2.8747,
+      "step": 4204
+    },
+    {
+      "epoch": 0.53824,
+      "grad_norm": 0.41361042857170105,
+      "learning_rate": 0.00046240853692719014,
+      "loss": 2.8922,
+      "step": 4205
+    },
+    {
+      "epoch": 0.538368,
+      "grad_norm": 0.4066367745399475,
+      "learning_rate": 0.00046220181578869146,
+      "loss": 2.8032,
+      "step": 4206
+    },
+    {
+      "epoch": 0.538496,
+      "grad_norm": 0.4288911521434784,
+      "learning_rate": 0.0004619951011481387,
+      "loss": 2.8311,
+      "step": 4207
+    },
+    {
+      "epoch": 0.538624,
+      "grad_norm": 0.44755324721336365,
+      "learning_rate": 0.0004617883930410688,
+      "loss": 2.9813,
+      "step": 4208
+    },
+    {
+      "epoch": 0.538752,
+      "grad_norm": 0.3936568796634674,
+      "learning_rate": 0.0004615816915030169,
+      "loss": 2.7709,
+      "step": 4209
+    },
+    {
+      "epoch": 0.53888,
+      "grad_norm": 0.48191314935684204,
+      "learning_rate": 0.00046137499656951774,
+      "loss": 2.8021,
+      "step": 4210
+    },
+    {
+      "epoch": 0.539008,
+      "grad_norm": 0.4170166552066803,
+      "learning_rate": 0.00046116830827610454,
+      "loss": 2.9701,
+      "step": 4211
+    },
+    {
+      "epoch": 0.539136,
+      "grad_norm": 0.45361581444740295,
+      "learning_rate": 0.0004609616266583092,
+      "loss": 2.8897,
+      "step": 4212
+    },
+    {
+      "epoch": 0.539264,
+      "grad_norm": 0.4155038595199585,
+      "learning_rate": 0.0004607549517516629,
+      "loss": 2.9142,
+      "step": 4213
+    },
+    {
+      "epoch": 0.539392,
+      "grad_norm": 0.41609594225883484,
+      "learning_rate": 0.0004605482835916954,
+      "loss": 2.8687,
+      "step": 4214
+    },
+    {
+      "epoch": 0.53952,
+      "grad_norm": 0.4668302834033966,
+      "learning_rate": 0.0004603416222139352,
+      "loss": 2.9364,
+      "step": 4215
+    },
+    {
+      "epoch": 0.539648,
+      "grad_norm": 0.6165719628334045,
+      "learning_rate": 0.0004601349676539099,
+      "loss": 2.9238,
+      "step": 4216
+    },
+    {
+      "epoch": 0.539776,
+      "grad_norm": 0.4305163621902466,
+      "learning_rate": 0.00045992831994714594,
+      "loss": 2.7819,
+      "step": 4217
+    },
+    {
+      "epoch": 0.539904,
+      "grad_norm": 0.42612236738204956,
+      "learning_rate": 0.00045972167912916806,
+      "loss": 2.7686,
+      "step": 4218
+    },
+    {
+      "epoch": 0.540032,
+      "grad_norm": 0.3975229859352112,
+      "learning_rate": 0.0004595150452355006,
+      "loss": 2.862,
+      "step": 4219
+    },
+    {
+      "epoch": 0.54016,
+      "grad_norm": 0.460857093334198,
+      "learning_rate": 0.0004593084183016664,
+      "loss": 2.9003,
+      "step": 4220
+    },
+    {
+      "epoch": 0.540288,
+      "grad_norm": 0.44309309124946594,
+      "learning_rate": 0.00045910179836318666,
+      "loss": 2.8917,
+      "step": 4221
+    },
+    {
+      "epoch": 0.540416,
+      "grad_norm": 0.44965246319770813,
+      "learning_rate": 0.00045889518545558195,
+      "loss": 2.9175,
+      "step": 4222
+    },
+    {
+      "epoch": 0.540544,
+      "grad_norm": 0.38287556171417236,
+      "learning_rate": 0.00045868857961437143,
+      "loss": 2.7837,
+      "step": 4223
+    },
+    {
+      "epoch": 0.540672,
+      "grad_norm": 0.44146445393562317,
+      "learning_rate": 0.00045848198087507295,
+      "loss": 3.0348,
+      "step": 4224
+    },
+    {
+      "epoch": 0.5408,
+      "grad_norm": 0.40589576959609985,
+      "learning_rate": 0.00045827538927320326,
+      "loss": 2.9439,
+      "step": 4225
+    },
+    {
+      "epoch": 0.540928,
+      "grad_norm": 0.43163546919822693,
+      "learning_rate": 0.000458068804844278,
+      "loss": 2.8719,
+      "step": 4226
+    },
+    {
+      "epoch": 0.541056,
+      "grad_norm": 0.4069448709487915,
+      "learning_rate": 0.00045786222762381125,
+      "loss": 2.7531,
+      "step": 4227
+    },
+    {
+      "epoch": 0.541184,
+      "grad_norm": 0.4272003769874573,
+      "learning_rate": 0.0004576556576473158,
+      "loss": 2.7899,
+      "step": 4228
+    },
+    {
+      "epoch": 0.541312,
+      "grad_norm": 0.38415271043777466,
+      "learning_rate": 0.00045744909495030393,
+      "loss": 2.8124,
+      "step": 4229
+    },
+    {
+      "epoch": 0.54144,
+      "grad_norm": 0.3980996906757355,
+      "learning_rate": 0.0004572425395682859,
+      "loss": 2.9543,
+      "step": 4230
+    },
+    {
+      "epoch": 0.541568,
+      "grad_norm": 0.47377145290374756,
+      "learning_rate": 0.00045703599153677104,
+      "loss": 2.7977,
+      "step": 4231
+    },
+    {
+      "epoch": 0.541696,
+      "grad_norm": 0.39165645837783813,
+      "learning_rate": 0.00045682945089126737,
+      "loss": 2.7788,
+      "step": 4232
+    },
+    {
+      "epoch": 0.541824,
+      "grad_norm": 0.45433980226516724,
+      "learning_rate": 0.0004566229176672814,
+      "loss": 2.738,
+      "step": 4233
+    },
+    {
+      "epoch": 0.541952,
+      "grad_norm": 0.42983779311180115,
+      "learning_rate": 0.0004564163919003189,
+      "loss": 2.79,
+      "step": 4234
+    },
+    {
+      "epoch": 0.54208,
+      "grad_norm": 0.4651046693325043,
+      "learning_rate": 0.00045620987362588394,
+      "loss": 2.904,
+      "step": 4235
+    },
+    {
+      "epoch": 0.542208,
+      "grad_norm": 0.42173781991004944,
+      "learning_rate": 0.0004560033628794792,
+      "loss": 2.8822,
+      "step": 4236
+    },
+    {
+      "epoch": 0.542336,
+      "grad_norm": 0.3867246210575104,
+      "learning_rate": 0.00045579685969660634,
+      "loss": 2.8474,
+      "step": 4237
+    },
+    {
+      "epoch": 0.542464,
+      "grad_norm": 0.4233472943305969,
+      "learning_rate": 0.00045559036411276595,
+      "loss": 2.7648,
+      "step": 4238
+    },
+    {
+      "epoch": 0.542592,
+      "grad_norm": 0.4405779242515564,
+      "learning_rate": 0.00045538387616345685,
+      "loss": 2.9558,
+      "step": 4239
+    },
+    {
+      "epoch": 0.54272,
+      "grad_norm": 0.4030233919620514,
+      "learning_rate": 0.0004551773958841765,
+      "loss": 2.8079,
+      "step": 4240
+    },
+    {
+      "epoch": 0.542848,
+      "grad_norm": 0.4240240156650543,
+      "learning_rate": 0.00045497092331042157,
+      "loss": 2.9835,
+      "step": 4241
+    },
+    {
+      "epoch": 0.542976,
+      "grad_norm": 0.410280704498291,
+      "learning_rate": 0.00045476445847768694,
+      "loss": 2.7874,
+      "step": 4242
+    },
+    {
+      "epoch": 0.543104,
+      "grad_norm": 0.3986964821815491,
+      "learning_rate": 0.00045455800142146617,
+      "loss": 2.8833,
+      "step": 4243
+    },
+    {
+      "epoch": 0.543232,
+      "grad_norm": 0.39263004064559937,
+      "learning_rate": 0.0004543515521772519,
+      "loss": 2.9172,
+      "step": 4244
+    },
+    {
+      "epoch": 0.54336,
+      "grad_norm": 0.4333422780036926,
+      "learning_rate": 0.00045414511078053506,
+      "loss": 2.9554,
+      "step": 4245
+    },
+    {
+      "epoch": 0.543488,
+      "grad_norm": 0.4328392744064331,
+      "learning_rate": 0.000453938677266805,
+      "loss": 2.8746,
+      "step": 4246
+    },
+    {
+      "epoch": 0.543616,
+      "grad_norm": 0.38661953806877136,
+      "learning_rate": 0.0004537322516715505,
+      "loss": 2.8615,
+      "step": 4247
+    },
+    {
+      "epoch": 0.543744,
+      "grad_norm": 0.3912937045097351,
+      "learning_rate": 0.00045352583403025825,
+      "loss": 2.8554,
+      "step": 4248
+    },
+    {
+      "epoch": 0.543872,
+      "grad_norm": 0.4091685712337494,
+      "learning_rate": 0.000453319424378414,
+      "loss": 2.8491,
+      "step": 4249
+    },
+    {
+      "epoch": 0.544,
+      "grad_norm": 0.4149726331233978,
+      "learning_rate": 0.0004531130227515019,
+      "loss": 2.8309,
+      "step": 4250
+    },
+    {
+      "epoch": 0.544128,
+      "grad_norm": 0.44890812039375305,
+      "learning_rate": 0.00045290662918500465,
+      "loss": 2.9807,
+      "step": 4251
+    },
+    {
+      "epoch": 0.544256,
+      "grad_norm": 0.41850748658180237,
+      "learning_rate": 0.00045270024371440385,
+      "loss": 2.8062,
+      "step": 4252
+    },
+    {
+      "epoch": 0.544384,
+      "grad_norm": 0.4253298044204712,
+      "learning_rate": 0.00045249386637517954,
+      "loss": 3.0061,
+      "step": 4253
+    },
+    {
+      "epoch": 0.544512,
+      "grad_norm": 0.4408901333808899,
+      "learning_rate": 0.00045228749720281023,
+      "loss": 2.8835,
+      "step": 4254
+    },
+    {
+      "epoch": 0.54464,
+      "grad_norm": 0.38186416029930115,
+      "learning_rate": 0.0004520811362327731,
+      "loss": 2.84,
+      "step": 4255
+    },
+    {
+      "epoch": 0.544768,
+      "grad_norm": 0.4522089958190918,
+      "learning_rate": 0.00045187478350054445,
+      "loss": 2.9598,
+      "step": 4256
+    },
+    {
+      "epoch": 0.544896,
+      "grad_norm": 0.4425143599510193,
+      "learning_rate": 0.0004516684390415985,
+      "loss": 2.8858,
+      "step": 4257
+    },
+    {
+      "epoch": 0.545024,
+      "grad_norm": 0.3960406184196472,
+      "learning_rate": 0.000451462102891408,
+      "loss": 2.8023,
+      "step": 4258
+    },
+    {
+      "epoch": 0.545152,
+      "grad_norm": 0.42064914107322693,
+      "learning_rate": 0.0004512557750854448,
+      "loss": 2.9974,
+      "step": 4259
+    },
+    {
+      "epoch": 0.54528,
+      "grad_norm": 0.43830689787864685,
+      "learning_rate": 0.000451049455659179,
+      "loss": 2.9792,
+      "step": 4260
+    },
+    {
+      "epoch": 0.545408,
+      "grad_norm": 0.43678778409957886,
+      "learning_rate": 0.00045084314464807914,
+      "loss": 2.8517,
+      "step": 4261
+    },
+    {
+      "epoch": 0.545536,
+      "grad_norm": 0.4082828462123871,
+      "learning_rate": 0.00045063684208761267,
+      "loss": 2.8311,
+      "step": 4262
+    },
+    {
+      "epoch": 0.545664,
+      "grad_norm": 0.40980401635169983,
+      "learning_rate": 0.00045043054801324533,
+      "loss": 2.869,
+      "step": 4263
+    },
+    {
+      "epoch": 0.545792,
+      "grad_norm": 0.40464290976524353,
+      "learning_rate": 0.00045022426246044134,
+      "loss": 2.8103,
+      "step": 4264
+    },
+    {
+      "epoch": 0.54592,
+      "grad_norm": 0.3949543237686157,
+      "learning_rate": 0.0004500179854646637,
+      "loss": 2.9328,
+      "step": 4265
+    },
+    {
+      "epoch": 0.546048,
+      "grad_norm": 0.4652846157550812,
+      "learning_rate": 0.00044981171706137376,
+      "loss": 2.8277,
+      "step": 4266
+    },
+    {
+      "epoch": 0.546176,
+      "grad_norm": 0.40444493293762207,
+      "learning_rate": 0.00044960545728603163,
+      "loss": 2.8755,
+      "step": 4267
+    },
+    {
+      "epoch": 0.546304,
+      "grad_norm": 0.4225592613220215,
+      "learning_rate": 0.0004493992061740956,
+      "loss": 3.0067,
+      "step": 4268
+    },
+    {
+      "epoch": 0.546432,
+      "grad_norm": 0.41372835636138916,
+      "learning_rate": 0.00044919296376102254,
+      "loss": 2.8915,
+      "step": 4269
+    },
+    {
+      "epoch": 0.54656,
+      "grad_norm": 0.4107300639152527,
+      "learning_rate": 0.0004489867300822681,
+      "loss": 2.8589,
+      "step": 4270
+    },
+    {
+      "epoch": 0.546688,
+      "grad_norm": 0.39674103260040283,
+      "learning_rate": 0.0004487805051732862,
+      "loss": 2.8592,
+      "step": 4271
+    },
+    {
+      "epoch": 0.546816,
+      "grad_norm": 0.4177449345588684,
+      "learning_rate": 0.00044857428906952914,
+      "loss": 2.8534,
+      "step": 4272
+    },
+    {
+      "epoch": 0.546944,
+      "grad_norm": 0.41664525866508484,
+      "learning_rate": 0.00044836808180644805,
+      "loss": 2.8683,
+      "step": 4273
+    },
+    {
+      "epoch": 0.547072,
+      "grad_norm": 0.4293774366378784,
+      "learning_rate": 0.00044816188341949205,
+      "loss": 2.9413,
+      "step": 4274
+    },
+    {
+      "epoch": 0.5472,
+      "grad_norm": 0.4021281599998474,
+      "learning_rate": 0.00044795569394410947,
+      "loss": 2.9056,
+      "step": 4275
+    },
+    {
+      "epoch": 0.547328,
+      "grad_norm": 0.4051416516304016,
+      "learning_rate": 0.00044774951341574633,
+      "loss": 2.6547,
+      "step": 4276
+    },
+    {
+      "epoch": 0.547456,
+      "grad_norm": 0.4421294927597046,
+      "learning_rate": 0.0004475433418698477,
+      "loss": 2.9601,
+      "step": 4277
+    },
+    {
+      "epoch": 0.547584,
+      "grad_norm": 0.42934751510620117,
+      "learning_rate": 0.0004473371793418567,
+      "loss": 2.8817,
+      "step": 4278
+    },
+    {
+      "epoch": 0.547712,
+      "grad_norm": 0.4177100658416748,
+      "learning_rate": 0.000447131025867215,
+      "loss": 2.9212,
+      "step": 4279
+    },
+    {
+      "epoch": 0.54784,
+      "grad_norm": 0.4063859283924103,
+      "learning_rate": 0.000446924881481363,
+      "loss": 2.9025,
+      "step": 4280
+    },
+    {
+      "epoch": 0.547968,
+      "grad_norm": 0.40288037061691284,
+      "learning_rate": 0.0004467187462197392,
+      "loss": 2.749,
+      "step": 4281
+    },
+    {
+      "epoch": 0.548096,
+      "grad_norm": 0.3931383192539215,
+      "learning_rate": 0.00044651262011778046,
+      "loss": 2.9305,
+      "step": 4282
+    },
+    {
+      "epoch": 0.548224,
+      "grad_norm": 0.3885703682899475,
+      "learning_rate": 0.00044630650321092236,
+      "loss": 2.7892,
+      "step": 4283
+    },
+    {
+      "epoch": 0.548352,
+      "grad_norm": 0.42350324988365173,
+      "learning_rate": 0.00044610039553459884,
+      "loss": 2.7551,
+      "step": 4284
+    },
+    {
+      "epoch": 0.54848,
+      "grad_norm": 0.4485780596733093,
+      "learning_rate": 0.0004458942971242423,
+      "loss": 2.8037,
+      "step": 4285
+    },
+    {
+      "epoch": 0.548608,
+      "grad_norm": 0.43546611070632935,
+      "learning_rate": 0.0004456882080152834,
+      "loss": 2.8782,
+      "step": 4286
+    },
+    {
+      "epoch": 0.548736,
+      "grad_norm": 0.3939485549926758,
+      "learning_rate": 0.00044548212824315104,
+      "loss": 2.9014,
+      "step": 4287
+    },
+    {
+      "epoch": 0.548864,
+      "grad_norm": 0.4284423887729645,
+      "learning_rate": 0.0004452760578432729,
+      "loss": 2.8762,
+      "step": 4288
+    },
+    {
+      "epoch": 0.548992,
+      "grad_norm": 0.4255518615245819,
+      "learning_rate": 0.0004450699968510749,
+      "loss": 2.9397,
+      "step": 4289
+    },
+    {
+      "epoch": 0.54912,
+      "grad_norm": 0.40468671917915344,
+      "learning_rate": 0.00044486394530198116,
+      "loss": 2.8503,
+      "step": 4290
+    },
+    {
+      "epoch": 0.549248,
+      "grad_norm": 0.4255591928958893,
+      "learning_rate": 0.00044465790323141446,
+      "loss": 2.8996,
+      "step": 4291
+    },
+    {
+      "epoch": 0.549376,
+      "grad_norm": 0.3982556462287903,
+      "learning_rate": 0.0004444518706747956,
+      "loss": 2.8161,
+      "step": 4292
+    },
+    {
+      "epoch": 0.549504,
+      "grad_norm": 0.3896838426589966,
+      "learning_rate": 0.0004442458476675443,
+      "loss": 2.8022,
+      "step": 4293
+    },
+    {
+      "epoch": 0.549632,
+      "grad_norm": 0.38936519622802734,
+      "learning_rate": 0.0004440398342450781,
+      "loss": 2.7662,
+      "step": 4294
+    },
+    {
+      "epoch": 0.54976,
+      "grad_norm": 0.38889196515083313,
+      "learning_rate": 0.0004438338304428132,
+      "loss": 2.8344,
+      "step": 4295
+    },
+    {
+      "epoch": 0.549888,
+      "grad_norm": 0.4199090600013733,
+      "learning_rate": 0.000443627836296164,
+      "loss": 2.8684,
+      "step": 4296
+    },
+    {
+      "epoch": 0.550016,
+      "grad_norm": 0.40086179971694946,
+      "learning_rate": 0.00044342185184054314,
+      "loss": 2.8826,
+      "step": 4297
+    },
+    {
+      "epoch": 0.550144,
+      "grad_norm": 0.4081622064113617,
+      "learning_rate": 0.00044321587711136195,
+      "loss": 2.9762,
+      "step": 4298
+    },
+    {
+      "epoch": 0.550272,
+      "grad_norm": 0.3993370532989502,
+      "learning_rate": 0.00044300991214402986,
+      "loss": 2.8887,
+      "step": 4299
+    },
+    {
+      "epoch": 0.5504,
+      "grad_norm": 0.40677574276924133,
+      "learning_rate": 0.00044280395697395437,
+      "loss": 2.7416,
+      "step": 4300
+    },
+    {
+      "epoch": 0.550528,
+      "grad_norm": 0.4117218554019928,
+      "learning_rate": 0.0004425980116365419,
+      "loss": 2.8965,
+      "step": 4301
+    },
+    {
+      "epoch": 0.550656,
+      "grad_norm": 0.39430704712867737,
+      "learning_rate": 0.00044239207616719647,
+      "loss": 2.8643,
+      "step": 4302
+    },
+    {
+      "epoch": 0.550784,
+      "grad_norm": 0.3918898403644562,
+      "learning_rate": 0.00044218615060132125,
+      "loss": 2.9647,
+      "step": 4303
+    },
+    {
+      "epoch": 0.550912,
+      "grad_norm": 0.42238661646842957,
+      "learning_rate": 0.0004419802349743169,
+      "loss": 2.8548,
+      "step": 4304
+    },
+    {
+      "epoch": 0.55104,
+      "grad_norm": 0.42495256662368774,
+      "learning_rate": 0.0004417743293215827,
+      "loss": 2.9316,
+      "step": 4305
+    },
+    {
+      "epoch": 0.551168,
+      "grad_norm": 0.4118393659591675,
+      "learning_rate": 0.0004415684336785164,
+      "loss": 2.7817,
+      "step": 4306
+    },
+    {
+      "epoch": 0.551296,
+      "grad_norm": 0.4216100573539734,
+      "learning_rate": 0.0004413625480805138,
+      "loss": 2.8701,
+      "step": 4307
+    },
+    {
+      "epoch": 0.551424,
+      "grad_norm": 0.40731993317604065,
+      "learning_rate": 0.0004411566725629688,
+      "loss": 2.7867,
+      "step": 4308
+    },
+    {
+      "epoch": 0.551552,
+      "grad_norm": 0.44419246912002563,
+      "learning_rate": 0.00044095080716127414,
+      "loss": 2.8506,
+      "step": 4309
+    },
+    {
+      "epoch": 0.55168,
+      "grad_norm": 0.4162459373474121,
+      "learning_rate": 0.0004407449519108202,
+      "loss": 2.7225,
+      "step": 4310
+    },
+    {
+      "epoch": 0.551808,
+      "grad_norm": 0.4659572243690491,
+      "learning_rate": 0.00044053910684699576,
+      "loss": 2.8331,
+      "step": 4311
+    },
+    {
+      "epoch": 0.551936,
+      "grad_norm": 0.38695067167282104,
+      "learning_rate": 0.00044033327200518834,
+      "loss": 2.9398,
+      "step": 4312
+    },
+    {
+      "epoch": 0.552064,
+      "grad_norm": 0.40453386306762695,
+      "learning_rate": 0.00044012744742078327,
+      "loss": 2.8557,
+      "step": 4313
+    },
+    {
+      "epoch": 0.552192,
+      "grad_norm": 0.4148349463939667,
+      "learning_rate": 0.0004399216331291641,
+      "loss": 2.8377,
+      "step": 4314
+    },
+    {
+      "epoch": 0.55232,
+      "grad_norm": 0.4601198136806488,
+      "learning_rate": 0.0004397158291657125,
+      "loss": 2.9499,
+      "step": 4315
+    },
+    {
+      "epoch": 0.552448,
+      "grad_norm": 0.42851683497428894,
+      "learning_rate": 0.0004395100355658089,
+      "loss": 2.8561,
+      "step": 4316
+    },
+    {
+      "epoch": 0.552576,
+      "grad_norm": 0.42031586170196533,
+      "learning_rate": 0.0004393042523648315,
+      "loss": 2.8721,
+      "step": 4317
+    },
+    {
+      "epoch": 0.552704,
+      "grad_norm": 0.4035525619983673,
+      "learning_rate": 0.0004390984795981566,
+      "loss": 2.9389,
+      "step": 4318
+    },
+    {
+      "epoch": 0.552832,
+      "grad_norm": 0.411304235458374,
+      "learning_rate": 0.0004388927173011592,
+      "loss": 2.8205,
+      "step": 4319
+    },
+    {
+      "epoch": 0.55296,
+      "grad_norm": 0.39843615889549255,
+      "learning_rate": 0.00043868696550921195,
+      "loss": 2.7872,
+      "step": 4320
+    },
+    {
+      "epoch": 0.553088,
+      "grad_norm": 0.4113139510154724,
+      "learning_rate": 0.00043848122425768627,
+      "loss": 2.9162,
+      "step": 4321
+    },
+    {
+      "epoch": 0.553216,
+      "grad_norm": 0.3954131305217743,
+      "learning_rate": 0.00043827549358195137,
+      "loss": 2.9255,
+      "step": 4322
+    },
+    {
+      "epoch": 0.553344,
+      "grad_norm": 0.39863985776901245,
+      "learning_rate": 0.0004380697735173745,
+      "loss": 2.8001,
+      "step": 4323
+    },
+    {
+      "epoch": 0.553472,
+      "grad_norm": 0.43243086338043213,
+      "learning_rate": 0.00043786406409932176,
+      "loss": 2.9931,
+      "step": 4324
+    },
+    {
+      "epoch": 0.5536,
+      "grad_norm": 0.4135066866874695,
+      "learning_rate": 0.00043765836536315667,
+      "loss": 2.8798,
+      "step": 4325
+    },
+    {
+      "epoch": 0.553728,
+      "grad_norm": 0.41665515303611755,
+      "learning_rate": 0.0004374526773442412,
+      "loss": 2.8803,
+      "step": 4326
+    },
+    {
+      "epoch": 0.553856,
+      "grad_norm": 0.42562544345855713,
+      "learning_rate": 0.00043724700007793574,
+      "loss": 2.7872,
+      "step": 4327
+    },
+    {
+      "epoch": 0.553984,
+      "grad_norm": 0.3987142741680145,
+      "learning_rate": 0.00043704133359959846,
+      "loss": 2.9258,
+      "step": 4328
+    },
+    {
+      "epoch": 0.554112,
+      "grad_norm": 0.40211623907089233,
+      "learning_rate": 0.00043683567794458556,
+      "loss": 2.9147,
+      "step": 4329
+    },
+    {
+      "epoch": 0.55424,
+      "grad_norm": 0.4094667136669159,
+      "learning_rate": 0.000436630033148252,
+      "loss": 2.8135,
+      "step": 4330
+    },
+    {
+      "epoch": 0.554368,
+      "grad_norm": 0.40571630001068115,
+      "learning_rate": 0.00043642439924595057,
+      "loss": 2.7425,
+      "step": 4331
+    },
+    {
+      "epoch": 0.554496,
+      "grad_norm": 0.42464011907577515,
+      "learning_rate": 0.00043621877627303185,
+      "loss": 2.7674,
+      "step": 4332
+    },
+    {
+      "epoch": 0.554624,
+      "grad_norm": 0.39678800106048584,
+      "learning_rate": 0.00043601316426484485,
+      "loss": 2.8816,
+      "step": 4333
+    },
+    {
+      "epoch": 0.554752,
+      "grad_norm": 0.41540005803108215,
+      "learning_rate": 0.00043580756325673674,
+      "loss": 2.8697,
+      "step": 4334
+    },
+    {
+      "epoch": 0.55488,
+      "grad_norm": 0.4477888345718384,
+      "learning_rate": 0.0004356019732840528,
+      "loss": 2.8311,
+      "step": 4335
+    },
+    {
+      "epoch": 0.555008,
+      "grad_norm": 0.4223763644695282,
+      "learning_rate": 0.000435396394382136,
+      "loss": 2.9029,
+      "step": 4336
+    },
+    {
+      "epoch": 0.555136,
+      "grad_norm": 0.43374308943748474,
+      "learning_rate": 0.00043519082658632815,
+      "loss": 2.8764,
+      "step": 4337
+    },
+    {
+      "epoch": 0.555264,
+      "grad_norm": 0.38434481620788574,
+      "learning_rate": 0.0004349852699319685,
+      "loss": 2.9369,
+      "step": 4338
+    },
+    {
+      "epoch": 0.555392,
+      "grad_norm": 0.40477830171585083,
+      "learning_rate": 0.00043477972445439447,
+      "loss": 2.8522,
+      "step": 4339
+    },
+    {
+      "epoch": 0.55552,
+      "grad_norm": 0.4334544539451599,
+      "learning_rate": 0.00043457419018894203,
+      "loss": 2.8642,
+      "step": 4340
+    },
+    {
+      "epoch": 0.555648,
+      "grad_norm": 0.4217802584171295,
+      "learning_rate": 0.0004343686671709448,
+      "loss": 2.786,
+      "step": 4341
+    },
+    {
+      "epoch": 0.555776,
+      "grad_norm": 0.45196589827537537,
+      "learning_rate": 0.00043416315543573465,
+      "loss": 2.9184,
+      "step": 4342
+    },
+    {
+      "epoch": 0.555904,
+      "grad_norm": 0.43512845039367676,
+      "learning_rate": 0.0004339576550186413,
+      "loss": 2.9123,
+      "step": 4343
+    },
+    {
+      "epoch": 0.556032,
+      "grad_norm": 0.3875661790370941,
+      "learning_rate": 0.0004337521659549925,
+      "loss": 2.768,
+      "step": 4344
+    },
+    {
+      "epoch": 0.55616,
+      "grad_norm": 0.44325709342956543,
+      "learning_rate": 0.0004335466882801146,
+      "loss": 2.7481,
+      "step": 4345
+    },
+    {
+      "epoch": 0.556288,
+      "grad_norm": 0.41433271765708923,
+      "learning_rate": 0.0004333412220293313,
+      "loss": 2.9568,
+      "step": 4346
+    },
+    {
+      "epoch": 0.556416,
+      "grad_norm": 0.4153602123260498,
+      "learning_rate": 0.0004331357672379646,
+      "loss": 2.8044,
+      "step": 4347
+    },
+    {
+      "epoch": 0.556544,
+      "grad_norm": 0.4072357714176178,
+      "learning_rate": 0.0004329303239413346,
+      "loss": 2.9357,
+      "step": 4348
+    },
+    {
+      "epoch": 0.556672,
+      "grad_norm": 0.37896108627319336,
+      "learning_rate": 0.00043272489217475963,
+      "loss": 2.7953,
+      "step": 4349
+    },
+    {
+      "epoch": 0.5568,
+      "grad_norm": 0.3956839144229889,
+      "learning_rate": 0.0004325194719735557,
+      "loss": 2.7898,
+      "step": 4350
+    },
+    {
+      "epoch": 0.556928,
+      "grad_norm": 0.4068792760372162,
+      "learning_rate": 0.00043231406337303663,
+      "loss": 3.0233,
+      "step": 4351
+    },
+    {
+      "epoch": 0.557056,
+      "grad_norm": 0.42828747630119324,
+      "learning_rate": 0.00043210866640851485,
+      "loss": 2.9232,
+      "step": 4352
+    },
+    {
+      "epoch": 0.557184,
+      "grad_norm": 0.4117404818534851,
+      "learning_rate": 0.00043190328111530045,
+      "loss": 2.8059,
+      "step": 4353
+    },
+    {
+      "epoch": 0.557312,
+      "grad_norm": 0.37598729133605957,
+      "learning_rate": 0.00043169790752870137,
+      "loss": 2.767,
+      "step": 4354
+    },
+    {
+      "epoch": 0.55744,
+      "grad_norm": 0.3808259069919586,
+      "learning_rate": 0.00043149254568402385,
+      "loss": 2.801,
+      "step": 4355
+    },
+    {
+      "epoch": 0.557568,
+      "grad_norm": 0.3864506185054779,
+      "learning_rate": 0.00043128719561657196,
+      "loss": 2.7783,
+      "step": 4356
+    },
+    {
+      "epoch": 0.557696,
+      "grad_norm": 0.40706804394721985,
+      "learning_rate": 0.0004310818573616476,
+      "loss": 2.9237,
+      "step": 4357
+    },
+    {
+      "epoch": 0.557824,
+      "grad_norm": 0.5432155728340149,
+      "learning_rate": 0.0004308765309545509,
+      "loss": 2.7743,
+      "step": 4358
+    },
+    {
+      "epoch": 0.557952,
+      "grad_norm": 0.40157443284988403,
+      "learning_rate": 0.00043067121643058004,
+      "loss": 2.7845,
+      "step": 4359
+    },
+    {
+      "epoch": 0.55808,
+      "grad_norm": 0.40856727957725525,
+      "learning_rate": 0.0004304659138250309,
+      "loss": 2.8666,
+      "step": 4360
+    },
+    {
+      "epoch": 0.558208,
+      "grad_norm": 0.4088335633277893,
+      "learning_rate": 0.00043026062317319704,
+      "loss": 2.9217,
+      "step": 4361
+    },
+    {
+      "epoch": 0.558336,
+      "grad_norm": 0.4386853575706482,
+      "learning_rate": 0.00043005534451037063,
+      "loss": 2.8824,
+      "step": 4362
+    },
+    {
+      "epoch": 0.558464,
+      "grad_norm": 0.40718284249305725,
+      "learning_rate": 0.0004298500778718415,
+      "loss": 2.7892,
+      "step": 4363
+    },
+    {
+      "epoch": 0.558592,
+      "grad_norm": 0.42483049631118774,
+      "learning_rate": 0.0004296448232928971,
+      "loss": 2.8536,
+      "step": 4364
+    },
+    {
+      "epoch": 0.55872,
+      "grad_norm": 0.43760159611701965,
+      "learning_rate": 0.0004294395808088232,
+      "loss": 2.7891,
+      "step": 4365
+    },
+    {
+      "epoch": 0.558848,
+      "grad_norm": 0.44260650873184204,
+      "learning_rate": 0.00042923435045490323,
+      "loss": 2.982,
+      "step": 4366
+    },
+    {
+      "epoch": 0.558976,
+      "grad_norm": 0.4153197705745697,
+      "learning_rate": 0.000429029132266419,
+      "loss": 2.8382,
+      "step": 4367
+    },
+    {
+      "epoch": 0.559104,
+      "grad_norm": 0.4129685163497925,
+      "learning_rate": 0.0004288239262786497,
+      "loss": 2.8661,
+      "step": 4368
+    },
+    {
+      "epoch": 0.559232,
+      "grad_norm": 0.401699036359787,
+      "learning_rate": 0.0004286187325268725,
+      "loss": 2.7435,
+      "step": 4369
+    },
+    {
+      "epoch": 0.55936,
+      "grad_norm": 0.38656917214393616,
+      "learning_rate": 0.00042841355104636283,
+      "loss": 2.9546,
+      "step": 4370
+    },
+    {
+      "epoch": 0.559488,
+      "grad_norm": 0.38249069452285767,
+      "learning_rate": 0.00042820838187239364,
+      "loss": 2.7328,
+      "step": 4371
+    },
+    {
+      "epoch": 0.559616,
+      "grad_norm": 0.4622862935066223,
+      "learning_rate": 0.00042800322504023573,
+      "loss": 2.8397,
+      "step": 4372
+    },
+    {
+      "epoch": 0.559744,
+      "grad_norm": 0.4481446444988251,
+      "learning_rate": 0.00042779808058515824,
+      "loss": 2.8454,
+      "step": 4373
+    },
+    {
+      "epoch": 0.559872,
+      "grad_norm": 0.42972517013549805,
+      "learning_rate": 0.00042759294854242773,
+      "loss": 2.9113,
+      "step": 4374
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.42473670840263367,
+      "learning_rate": 0.00042738782894730874,
+      "loss": 2.874,
+      "step": 4375
+    },
+    {
+      "epoch": 0.560128,
+      "grad_norm": 0.3801279664039612,
+      "learning_rate": 0.00042718272183506353,
+      "loss": 2.7966,
+      "step": 4376
+    },
+    {
+      "epoch": 0.560256,
+      "grad_norm": 0.438021183013916,
+      "learning_rate": 0.00042697762724095285,
+      "loss": 2.7675,
+      "step": 4377
+    },
+    {
+      "epoch": 0.560384,
+      "grad_norm": 0.4914986789226532,
+      "learning_rate": 0.00042677254520023463,
+      "loss": 2.8305,
+      "step": 4378
+    },
+    {
+      "epoch": 0.560512,
+      "grad_norm": 0.4221052825450897,
+      "learning_rate": 0.0004265674757481647,
+      "loss": 3.0006,
+      "step": 4379
+    },
+    {
+      "epoch": 0.56064,
+      "grad_norm": 0.43798381090164185,
+      "learning_rate": 0.00042636241891999706,
+      "loss": 2.8384,
+      "step": 4380
+    },
+    {
+      "epoch": 0.560768,
+      "grad_norm": 0.4466171860694885,
+      "learning_rate": 0.0004261573747509833,
+      "loss": 2.7067,
+      "step": 4381
+    },
+    {
+      "epoch": 0.560896,
+      "grad_norm": 0.4049871861934662,
+      "learning_rate": 0.00042595234327637277,
+      "loss": 2.7607,
+      "step": 4382
+    },
+    {
+      "epoch": 0.561024,
+      "grad_norm": 0.4108105003833771,
+      "learning_rate": 0.000425747324531413,
+      "loss": 2.8061,
+      "step": 4383
+    },
+    {
+      "epoch": 0.561152,
+      "grad_norm": 0.37825560569763184,
+      "learning_rate": 0.0004255423185513489,
+      "loss": 2.7887,
+      "step": 4384
+    },
+    {
+      "epoch": 0.56128,
+      "grad_norm": 0.4615241587162018,
+      "learning_rate": 0.00042533732537142335,
+      "loss": 2.8847,
+      "step": 4385
+    },
+    {
+      "epoch": 0.561408,
+      "grad_norm": 0.39594587683677673,
+      "learning_rate": 0.00042513234502687723,
+      "loss": 2.8821,
+      "step": 4386
+    },
+    {
+      "epoch": 0.561536,
+      "grad_norm": 0.40556126832962036,
+      "learning_rate": 0.00042492737755294883,
+      "loss": 2.7823,
+      "step": 4387
+    },
+    {
+      "epoch": 0.561664,
+      "grad_norm": 0.43309223651885986,
+      "learning_rate": 0.00042472242298487466,
+      "loss": 2.8364,
+      "step": 4388
+    },
+    {
+      "epoch": 0.561792,
+      "grad_norm": 0.4147423505783081,
+      "learning_rate": 0.0004245174813578887,
+      "loss": 2.6466,
+      "step": 4389
+    },
+    {
+      "epoch": 0.56192,
+      "grad_norm": 0.4079302251338959,
+      "learning_rate": 0.0004243125527072227,
+      "loss": 2.933,
+      "step": 4390
+    },
+    {
+      "epoch": 0.562048,
+      "grad_norm": 0.41053035855293274,
+      "learning_rate": 0.00042410763706810634,
+      "loss": 2.9024,
+      "step": 4391
+    },
+    {
+      "epoch": 0.562176,
+      "grad_norm": 0.43919986486434937,
+      "learning_rate": 0.000423902734475767,
+      "loss": 2.8171,
+      "step": 4392
+    },
+    {
+      "epoch": 0.562304,
+      "grad_norm": 0.4238097369670868,
+      "learning_rate": 0.00042369784496542977,
+      "loss": 2.8103,
+      "step": 4393
+    },
+    {
+      "epoch": 0.562432,
+      "grad_norm": 0.49002358317375183,
+      "learning_rate": 0.0004234929685723174,
+      "loss": 3.0019,
+      "step": 4394
+    },
+    {
+      "epoch": 0.56256,
+      "grad_norm": 0.39745473861694336,
+      "learning_rate": 0.0004232881053316509,
+      "loss": 2.9869,
+      "step": 4395
+    },
+    {
+      "epoch": 0.562688,
+      "grad_norm": 0.38512545824050903,
+      "learning_rate": 0.0004230832552786484,
+      "loss": 2.7917,
+      "step": 4396
+    },
+    {
+      "epoch": 0.562816,
+      "grad_norm": 0.43248119950294495,
+      "learning_rate": 0.00042287841844852595,
+      "loss": 2.8637,
+      "step": 4397
+    },
+    {
+      "epoch": 0.562944,
+      "grad_norm": 0.47860389947891235,
+      "learning_rate": 0.00042267359487649755,
+      "loss": 3.0122,
+      "step": 4398
+    },
+    {
+      "epoch": 0.563072,
+      "grad_norm": 0.4183955788612366,
+      "learning_rate": 0.0004224687845977746,
+      "loss": 2.8684,
+      "step": 4399
+    },
+    {
+      "epoch": 0.5632,
+      "grad_norm": 0.41327691078186035,
+      "learning_rate": 0.00042226398764756625,
+      "loss": 2.9551,
+      "step": 4400
+    },
+    {
+      "epoch": 0.563328,
+      "grad_norm": 0.3902265429496765,
+      "learning_rate": 0.0004220592040610798,
+      "loss": 2.7834,
+      "step": 4401
+    },
+    {
+      "epoch": 0.563456,
+      "grad_norm": 0.3780233860015869,
+      "learning_rate": 0.0004218544338735197,
+      "loss": 2.7967,
+      "step": 4402
+    },
+    {
+      "epoch": 0.563584,
+      "grad_norm": 0.41957995295524597,
+      "learning_rate": 0.0004216496771200881,
+      "loss": 2.938,
+      "step": 4403
+    },
+    {
+      "epoch": 0.563712,
+      "grad_norm": 0.4244241416454315,
+      "learning_rate": 0.0004214449338359856,
+      "loss": 2.8425,
+      "step": 4404
+    },
+    {
+      "epoch": 0.56384,
+      "grad_norm": 0.41686490178108215,
+      "learning_rate": 0.00042124020405640955,
+      "loss": 2.7557,
+      "step": 4405
+    },
+    {
+      "epoch": 0.563968,
+      "grad_norm": 0.46934375166893005,
+      "learning_rate": 0.00042103548781655554,
+      "loss": 2.7978,
+      "step": 4406
+    },
+    {
+      "epoch": 0.564096,
+      "grad_norm": 0.4215030372142792,
+      "learning_rate": 0.00042083078515161665,
+      "loss": 2.8968,
+      "step": 4407
+    },
+    {
+      "epoch": 0.564224,
+      "grad_norm": 0.40649399161338806,
+      "learning_rate": 0.00042062609609678353,
+      "loss": 2.8306,
+      "step": 4408
+    },
+    {
+      "epoch": 0.564352,
+      "grad_norm": 0.3990454375743866,
+      "learning_rate": 0.0004204214206872448,
+      "loss": 2.8926,
+      "step": 4409
+    },
+    {
+      "epoch": 0.56448,
+      "grad_norm": 0.40391603112220764,
+      "learning_rate": 0.0004202167589581863,
+      "loss": 2.7231,
+      "step": 4410
+    },
+    {
+      "epoch": 0.564608,
+      "grad_norm": 0.38020139932632446,
+      "learning_rate": 0.0004200121109447918,
+      "loss": 2.8884,
+      "step": 4411
+    },
+    {
+      "epoch": 0.564736,
+      "grad_norm": 0.40926510095596313,
+      "learning_rate": 0.00041980747668224285,
+      "loss": 2.7831,
+      "step": 4412
+    },
+    {
+      "epoch": 0.564864,
+      "grad_norm": 0.4134894609451294,
+      "learning_rate": 0.00041960285620571807,
+      "loss": 2.8399,
+      "step": 4413
+    },
+    {
+      "epoch": 0.564992,
+      "grad_norm": 0.4222732186317444,
+      "learning_rate": 0.00041939824955039464,
+      "loss": 2.8668,
+      "step": 4414
+    },
+    {
+      "epoch": 0.56512,
+      "grad_norm": 0.45631200075149536,
+      "learning_rate": 0.00041919365675144635,
+      "loss": 2.9656,
+      "step": 4415
+    },
+    {
+      "epoch": 0.565248,
+      "grad_norm": 0.44022300839424133,
+      "learning_rate": 0.0004189890778440454,
+      "loss": 2.8455,
+      "step": 4416
+    },
+    {
+      "epoch": 0.565376,
+      "grad_norm": 0.41950666904449463,
+      "learning_rate": 0.0004187845128633611,
+      "loss": 2.7624,
+      "step": 4417
+    },
+    {
+      "epoch": 0.565504,
+      "grad_norm": 0.3838654160499573,
+      "learning_rate": 0.00041857996184456047,
+      "loss": 2.7788,
+      "step": 4418
+    },
+    {
+      "epoch": 0.565632,
+      "grad_norm": 0.40251022577285767,
+      "learning_rate": 0.00041837542482280846,
+      "loss": 2.8964,
+      "step": 4419
+    },
+    {
+      "epoch": 0.56576,
+      "grad_norm": 0.4189419746398926,
+      "learning_rate": 0.0004181709018332671,
+      "loss": 2.8662,
+      "step": 4420
+    },
+    {
+      "epoch": 0.565888,
+      "grad_norm": 0.40775245428085327,
+      "learning_rate": 0.0004179663929110963,
+      "loss": 2.7773,
+      "step": 4421
+    },
+    {
+      "epoch": 0.566016,
+      "grad_norm": 0.43468379974365234,
+      "learning_rate": 0.0004177618980914536,
+      "loss": 2.8361,
+      "step": 4422
+    },
+    {
+      "epoch": 0.566144,
+      "grad_norm": 0.46034836769104004,
+      "learning_rate": 0.000417557417409494,
+      "loss": 2.7398,
+      "step": 4423
+    },
+    {
+      "epoch": 0.566272,
+      "grad_norm": 0.4264008402824402,
+      "learning_rate": 0.0004173529509003702,
+      "loss": 2.9508,
+      "step": 4424
+    },
+    {
+      "epoch": 0.5664,
+      "grad_norm": 0.4454164206981659,
+      "learning_rate": 0.00041714849859923227,
+      "loss": 2.8857,
+      "step": 4425
+    },
+    {
+      "epoch": 0.566528,
+      "grad_norm": 0.40050017833709717,
+      "learning_rate": 0.0004169440605412278,
+      "loss": 2.867,
+      "step": 4426
+    },
+    {
+      "epoch": 0.566656,
+      "grad_norm": 0.42341601848602295,
+      "learning_rate": 0.00041673963676150226,
+      "loss": 3.0365,
+      "step": 4427
+    },
+    {
+      "epoch": 0.566784,
+      "grad_norm": 0.3853769898414612,
+      "learning_rate": 0.0004165352272951985,
+      "loss": 2.8392,
+      "step": 4428
+    },
+    {
+      "epoch": 0.566912,
+      "grad_norm": 0.4306138753890991,
+      "learning_rate": 0.0004163308321774566,
+      "loss": 2.8194,
+      "step": 4429
+    },
+    {
+      "epoch": 0.56704,
+      "grad_norm": 0.3872254192829132,
+      "learning_rate": 0.0004161264514434148,
+      "loss": 2.8485,
+      "step": 4430
+    },
+    {
+      "epoch": 0.567168,
+      "grad_norm": 0.42119914293289185,
+      "learning_rate": 0.000415922085128208,
+      "loss": 2.9724,
+      "step": 4431
+    },
+    {
+      "epoch": 0.567296,
+      "grad_norm": 0.4442872703075409,
+      "learning_rate": 0.00041571773326696984,
+      "loss": 2.9516,
+      "step": 4432
+    },
+    {
+      "epoch": 0.567424,
+      "grad_norm": 0.406496524810791,
+      "learning_rate": 0.00041551339589483016,
+      "loss": 2.8971,
+      "step": 4433
+    },
+    {
+      "epoch": 0.567552,
+      "grad_norm": 0.422544926404953,
+      "learning_rate": 0.0004153090730469174,
+      "loss": 2.7045,
+      "step": 4434
+    },
+    {
+      "epoch": 0.56768,
+      "grad_norm": 0.4253782033920288,
+      "learning_rate": 0.0004151047647583568,
+      "loss": 2.9318,
+      "step": 4435
+    },
+    {
+      "epoch": 0.567808,
+      "grad_norm": 0.4122875928878784,
+      "learning_rate": 0.00041490047106427114,
+      "loss": 2.7165,
+      "step": 4436
+    },
+    {
+      "epoch": 0.567936,
+      "grad_norm": 0.36282041668891907,
+      "learning_rate": 0.00041469619199978124,
+      "loss": 2.7666,
+      "step": 4437
+    },
+    {
+      "epoch": 0.568064,
+      "grad_norm": 0.40950697660446167,
+      "learning_rate": 0.0004144919276000048,
+      "loss": 2.893,
+      "step": 4438
+    },
+    {
+      "epoch": 0.568192,
+      "grad_norm": 0.4241769015789032,
+      "learning_rate": 0.00041428767790005727,
+      "loss": 2.9449,
+      "step": 4439
+    },
+    {
+      "epoch": 0.56832,
+      "grad_norm": 0.44882431626319885,
+      "learning_rate": 0.0004140834429350515,
+      "loss": 2.8503,
+      "step": 4440
+    },
+    {
+      "epoch": 0.568448,
+      "grad_norm": 0.4020390212535858,
+      "learning_rate": 0.000413879222740098,
+      "loss": 2.8231,
+      "step": 4441
+    },
+    {
+      "epoch": 0.568576,
+      "grad_norm": 0.4021652638912201,
+      "learning_rate": 0.00041367501735030453,
+      "loss": 2.8591,
+      "step": 4442
+    },
+    {
+      "epoch": 0.568704,
+      "grad_norm": 0.41025447845458984,
+      "learning_rate": 0.0004134708268007764,
+      "loss": 2.9897,
+      "step": 4443
+    },
+    {
+      "epoch": 0.568832,
+      "grad_norm": 0.3899625241756439,
+      "learning_rate": 0.00041326665112661614,
+      "loss": 2.7611,
+      "step": 4444
+    },
+    {
+      "epoch": 0.56896,
+      "grad_norm": 0.40143871307373047,
+      "learning_rate": 0.0004130624903629242,
+      "loss": 2.8888,
+      "step": 4445
+    },
+    {
+      "epoch": 0.569088,
+      "grad_norm": 0.3960128426551819,
+      "learning_rate": 0.00041285834454479796,
+      "loss": 2.7387,
+      "step": 4446
+    },
+    {
+      "epoch": 0.569216,
+      "grad_norm": 0.4086760878562927,
+      "learning_rate": 0.00041265421370733247,
+      "loss": 2.8603,
+      "step": 4447
+    },
+    {
+      "epoch": 0.569344,
+      "grad_norm": 0.39002272486686707,
+      "learning_rate": 0.00041245009788562037,
+      "loss": 2.9209,
+      "step": 4448
+    },
+    {
+      "epoch": 0.569472,
+      "grad_norm": 0.41168245673179626,
+      "learning_rate": 0.0004122459971147514,
+      "loss": 2.8808,
+      "step": 4449
+    },
+    {
+      "epoch": 0.5696,
+      "grad_norm": 0.39500290155410767,
+      "learning_rate": 0.0004120419114298126,
+      "loss": 2.863,
+      "step": 4450
+    },
+    {
+      "epoch": 0.569728,
+      "grad_norm": 0.41496843099594116,
+      "learning_rate": 0.00041183784086588904,
+      "loss": 2.8453,
+      "step": 4451
+    },
+    {
+      "epoch": 0.569856,
+      "grad_norm": 0.4305526614189148,
+      "learning_rate": 0.00041163378545806286,
+      "loss": 2.8926,
+      "step": 4452
+    },
+    {
+      "epoch": 0.569984,
+      "grad_norm": 0.4178623557090759,
+      "learning_rate": 0.0004114297452414133,
+      "loss": 2.796,
+      "step": 4453
+    },
+    {
+      "epoch": 0.570112,
+      "grad_norm": 0.4280755817890167,
+      "learning_rate": 0.0004112257202510173,
+      "loss": 2.8317,
+      "step": 4454
+    },
+    {
+      "epoch": 0.57024,
+      "grad_norm": 0.39707931876182556,
+      "learning_rate": 0.0004110217105219492,
+      "loss": 2.7964,
+      "step": 4455
+    },
+    {
+      "epoch": 0.570368,
+      "grad_norm": 0.41847217082977295,
+      "learning_rate": 0.00041081771608928065,
+      "loss": 2.8609,
+      "step": 4456
+    },
+    {
+      "epoch": 0.570496,
+      "grad_norm": 0.3795264959335327,
+      "learning_rate": 0.0004106137369880804,
+      "loss": 2.7642,
+      "step": 4457
+    },
+    {
+      "epoch": 0.570624,
+      "grad_norm": 0.4026149809360504,
+      "learning_rate": 0.0004104097732534152,
+      "loss": 2.8875,
+      "step": 4458
+    },
+    {
+      "epoch": 0.570752,
+      "grad_norm": 0.47419479489326477,
+      "learning_rate": 0.0004102058249203483,
+      "loss": 2.9156,
+      "step": 4459
+    },
+    {
+      "epoch": 0.57088,
+      "grad_norm": 0.40231040120124817,
+      "learning_rate": 0.0004100018920239414,
+      "loss": 2.8928,
+      "step": 4460
+    },
+    {
+      "epoch": 0.571008,
+      "grad_norm": 0.44006770849227905,
+      "learning_rate": 0.00040979797459925253,
+      "loss": 3.0016,
+      "step": 4461
+    },
+    {
+      "epoch": 0.571136,
+      "grad_norm": 0.39198437333106995,
+      "learning_rate": 0.00040959407268133745,
+      "loss": 2.8352,
+      "step": 4462
+    },
+    {
+      "epoch": 0.571264,
+      "grad_norm": 0.3831428289413452,
+      "learning_rate": 0.0004093901863052495,
+      "loss": 2.8635,
+      "step": 4463
+    },
+    {
+      "epoch": 0.571392,
+      "grad_norm": 0.39949148893356323,
+      "learning_rate": 0.0004091863155060389,
+      "loss": 2.9037,
+      "step": 4464
+    },
+    {
+      "epoch": 0.57152,
+      "grad_norm": 0.3962692618370056,
+      "learning_rate": 0.0004089824603187534,
+      "loss": 2.8731,
+      "step": 4465
+    },
+    {
+      "epoch": 0.571648,
+      "grad_norm": 0.4010111689567566,
+      "learning_rate": 0.00040877862077843825,
+      "loss": 2.8976,
+      "step": 4466
+    },
+    {
+      "epoch": 0.571776,
+      "grad_norm": 0.4108634889125824,
+      "learning_rate": 0.00040857479692013566,
+      "loss": 2.8871,
+      "step": 4467
+    },
+    {
+      "epoch": 0.571904,
+      "grad_norm": 0.39448678493499756,
+      "learning_rate": 0.00040837098877888523,
+      "loss": 2.8068,
+      "step": 4468
+    },
+    {
+      "epoch": 0.572032,
+      "grad_norm": 0.40498095750808716,
+      "learning_rate": 0.0004081671963897241,
+      "loss": 2.8881,
+      "step": 4469
+    },
+    {
+      "epoch": 0.57216,
+      "grad_norm": 0.4646829664707184,
+      "learning_rate": 0.0004079634197876867,
+      "loss": 2.7585,
+      "step": 4470
+    },
+    {
+      "epoch": 0.572288,
+      "grad_norm": 0.3794872760772705,
+      "learning_rate": 0.00040775965900780443,
+      "loss": 2.7984,
+      "step": 4471
+    },
+    {
+      "epoch": 0.572416,
+      "grad_norm": 0.39949992299079895,
+      "learning_rate": 0.000407555914085106,
+      "loss": 2.6914,
+      "step": 4472
+    },
+    {
+      "epoch": 0.572544,
+      "grad_norm": 0.4166777729988098,
+      "learning_rate": 0.0004073521850546177,
+      "loss": 3.0751,
+      "step": 4473
+    },
+    {
+      "epoch": 0.572672,
+      "grad_norm": 0.4094710052013397,
+      "learning_rate": 0.00040714847195136294,
+      "loss": 2.8732,
+      "step": 4474
+    },
+    {
+      "epoch": 0.5728,
+      "grad_norm": 0.44674310088157654,
+      "learning_rate": 0.00040694477481036215,
+      "loss": 2.798,
+      "step": 4475
+    },
+    {
+      "epoch": 0.572928,
+      "grad_norm": 0.3585146367549896,
+      "learning_rate": 0.00040674109366663344,
+      "loss": 2.7768,
+      "step": 4476
+    },
+    {
+      "epoch": 0.573056,
+      "grad_norm": 0.3905830383300781,
+      "learning_rate": 0.00040653742855519166,
+      "loss": 2.805,
+      "step": 4477
+    },
+    {
+      "epoch": 0.573184,
+      "grad_norm": 0.43881604075431824,
+      "learning_rate": 0.00040633377951104966,
+      "loss": 2.76,
+      "step": 4478
+    },
+    {
+      "epoch": 0.573312,
+      "grad_norm": 0.3737005889415741,
+      "learning_rate": 0.00040613014656921677,
+      "loss": 2.7888,
+      "step": 4479
+    },
+    {
+      "epoch": 0.57344,
+      "grad_norm": 0.3912295401096344,
+      "learning_rate": 0.0004059265297646999,
+      "loss": 2.7858,
+      "step": 4480
+    },
+    {
+      "epoch": 0.573568,
+      "grad_norm": 0.41289886832237244,
+      "learning_rate": 0.0004057229291325032,
+      "loss": 2.8626,
+      "step": 4481
+    },
+    {
+      "epoch": 0.573696,
+      "grad_norm": 0.42647409439086914,
+      "learning_rate": 0.00040551934470762797,
+      "loss": 2.8133,
+      "step": 4482
+    },
+    {
+      "epoch": 0.573824,
+      "grad_norm": 0.40230104327201843,
+      "learning_rate": 0.00040531577652507255,
+      "loss": 2.8721,
+      "step": 4483
+    },
+    {
+      "epoch": 0.573952,
+      "grad_norm": 0.407715380191803,
+      "learning_rate": 0.000405112224619833,
+      "loss": 2.8125,
+      "step": 4484
+    },
+    {
+      "epoch": 0.57408,
+      "grad_norm": 0.41592034697532654,
+      "learning_rate": 0.00040490868902690204,
+      "loss": 2.7878,
+      "step": 4485
+    },
+    {
+      "epoch": 0.574208,
+      "grad_norm": 0.3848268389701843,
+      "learning_rate": 0.0004047051697812697,
+      "loss": 2.9139,
+      "step": 4486
+    },
+    {
+      "epoch": 0.574336,
+      "grad_norm": 0.42882201075553894,
+      "learning_rate": 0.00040450166691792327,
+      "loss": 2.8967,
+      "step": 4487
+    },
+    {
+      "epoch": 0.574464,
+      "grad_norm": 0.39693236351013184,
+      "learning_rate": 0.00040429818047184774,
+      "loss": 2.8116,
+      "step": 4488
+    },
+    {
+      "epoch": 0.574592,
+      "grad_norm": 0.43982645869255066,
+      "learning_rate": 0.00040409471047802436,
+      "loss": 2.8998,
+      "step": 4489
+    },
+    {
+      "epoch": 0.57472,
+      "grad_norm": 0.383804053068161,
+      "learning_rate": 0.000403891256971432,
+      "loss": 2.8892,
+      "step": 4490
+    },
+    {
+      "epoch": 0.574848,
+      "grad_norm": 0.3864831030368805,
+      "learning_rate": 0.0004036878199870469,
+      "loss": 2.8468,
+      "step": 4491
+    },
+    {
+      "epoch": 0.574976,
+      "grad_norm": 0.38534483313560486,
+      "learning_rate": 0.00040348439955984204,
+      "loss": 2.8702,
+      "step": 4492
+    },
+    {
+      "epoch": 0.575104,
+      "grad_norm": 0.4317382872104645,
+      "learning_rate": 0.0004032809957247877,
+      "loss": 2.835,
+      "step": 4493
+    },
+    {
+      "epoch": 0.575232,
+      "grad_norm": 0.4055645763874054,
+      "learning_rate": 0.0004030776085168516,
+      "loss": 2.7915,
+      "step": 4494
+    },
+    {
+      "epoch": 0.57536,
+      "grad_norm": 0.4167625904083252,
+      "learning_rate": 0.0004028742379709982,
+      "loss": 2.8819,
+      "step": 4495
+    },
+    {
+      "epoch": 0.575488,
+      "grad_norm": 0.41262754797935486,
+      "learning_rate": 0.000402670884122189,
+      "loss": 2.7858,
+      "step": 4496
+    },
+    {
+      "epoch": 0.575616,
+      "grad_norm": 0.43196430802345276,
+      "learning_rate": 0.00040246754700538353,
+      "loss": 2.9227,
+      "step": 4497
+    },
+    {
+      "epoch": 0.575744,
+      "grad_norm": 0.43682289123535156,
+      "learning_rate": 0.0004022642266555372,
+      "loss": 2.9469,
+      "step": 4498
+    },
+    {
+      "epoch": 0.575872,
+      "grad_norm": 0.40762603282928467,
+      "learning_rate": 0.00040206092310760347,
+      "loss": 2.967,
+      "step": 4499
+    },
+    {
+      "epoch": 0.576,
+      "grad_norm": 0.3890848159790039,
+      "learning_rate": 0.00040185763639653246,
+      "loss": 2.7407,
+      "step": 4500
+    },
+    {
+      "epoch": 0.576128,
+      "grad_norm": 0.40443557500839233,
+      "learning_rate": 0.0004016543665572714,
+      "loss": 2.8494,
+      "step": 4501
+    },
+    {
+      "epoch": 0.576256,
+      "grad_norm": 0.3928864896297455,
+      "learning_rate": 0.000401451113624765,
+      "loss": 2.9299,
+      "step": 4502
+    },
+    {
+      "epoch": 0.576384,
+      "grad_norm": 0.39031878113746643,
+      "learning_rate": 0.00040124787763395463,
+      "loss": 2.8325,
+      "step": 4503
+    },
+    {
+      "epoch": 0.576512,
+      "grad_norm": 0.4063706696033478,
+      "learning_rate": 0.0004010446586197788,
+      "loss": 2.97,
+      "step": 4504
+    },
+    {
+      "epoch": 0.57664,
+      "grad_norm": 0.43063127994537354,
+      "learning_rate": 0.00040084145661717325,
+      "loss": 2.9058,
+      "step": 4505
+    },
+    {
+      "epoch": 0.576768,
+      "grad_norm": 0.36924973130226135,
+      "learning_rate": 0.000400638271661071,
+      "loss": 2.9881,
+      "step": 4506
+    },
+    {
+      "epoch": 0.576896,
+      "grad_norm": 0.4026382863521576,
+      "learning_rate": 0.00040043510378640193,
+      "loss": 2.886,
+      "step": 4507
+    },
+    {
+      "epoch": 0.577024,
+      "grad_norm": 0.4677257537841797,
+      "learning_rate": 0.0004002319530280925,
+      "loss": 2.8378,
+      "step": 4508
+    },
+    {
+      "epoch": 0.577152,
+      "grad_norm": 0.38592976331710815,
+      "learning_rate": 0.0004000288194210671,
+      "loss": 2.8074,
+      "step": 4509
+    },
+    {
+      "epoch": 0.57728,
+      "grad_norm": 0.3912583589553833,
+      "learning_rate": 0.00039982570300024664,
+      "loss": 2.8257,
+      "step": 4510
+    },
+    {
+      "epoch": 0.577408,
+      "grad_norm": 0.4150262176990509,
+      "learning_rate": 0.00039962260380054903,
+      "loss": 2.9405,
+      "step": 4511
+    },
+    {
+      "epoch": 0.577536,
+      "grad_norm": 0.41791996359825134,
+      "learning_rate": 0.0003994195218568896,
+      "loss": 2.748,
+      "step": 4512
+    },
+    {
+      "epoch": 0.577664,
+      "grad_norm": 0.4235340654850006,
+      "learning_rate": 0.0003992164572041804,
+      "loss": 2.7898,
+      "step": 4513
+    },
+    {
+      "epoch": 0.577792,
+      "grad_norm": 0.40534335374832153,
+      "learning_rate": 0.00039901340987733036,
+      "loss": 3.0532,
+      "step": 4514
+    },
+    {
+      "epoch": 0.57792,
+      "grad_norm": 0.3946755826473236,
+      "learning_rate": 0.0003988103799112461,
+      "loss": 2.9683,
+      "step": 4515
+    },
+    {
+      "epoch": 0.578048,
+      "grad_norm": 0.4047011137008667,
+      "learning_rate": 0.0003986073673408306,
+      "loss": 2.8983,
+      "step": 4516
+    },
+    {
+      "epoch": 0.578176,
+      "grad_norm": 0.3818943202495575,
+      "learning_rate": 0.0003984043722009841,
+      "loss": 2.8296,
+      "step": 4517
+    },
+    {
+      "epoch": 0.578304,
+      "grad_norm": 0.41568657755851746,
+      "learning_rate": 0.0003982013945266039,
+      "loss": 3.0529,
+      "step": 4518
+    },
+    {
+      "epoch": 0.578432,
+      "grad_norm": 0.4088493287563324,
+      "learning_rate": 0.000397998434352584,
+      "loss": 2.8402,
+      "step": 4519
+    },
+    {
+      "epoch": 0.57856,
+      "grad_norm": 0.3998301923274994,
+      "learning_rate": 0.0003977954917138159,
+      "loss": 2.7684,
+      "step": 4520
+    },
+    {
+      "epoch": 0.578688,
+      "grad_norm": 0.42059803009033203,
+      "learning_rate": 0.00039759256664518753,
+      "loss": 2.7689,
+      "step": 4521
+    },
+    {
+      "epoch": 0.578816,
+      "grad_norm": 0.4558786153793335,
+      "learning_rate": 0.0003973896591815841,
+      "loss": 2.861,
+      "step": 4522
+    },
+    {
+      "epoch": 0.578944,
+      "grad_norm": 0.4109323024749756,
+      "learning_rate": 0.0003971867693578879,
+      "loss": 2.8615,
+      "step": 4523
+    },
+    {
+      "epoch": 0.579072,
+      "grad_norm": 0.4049813449382782,
+      "learning_rate": 0.00039698389720897774,
+      "loss": 2.8562,
+      "step": 4524
+    },
+    {
+      "epoch": 0.5792,
+      "grad_norm": 0.4467053711414337,
+      "learning_rate": 0.00039678104276973016,
+      "loss": 2.806,
+      "step": 4525
+    },
+    {
+      "epoch": 0.579328,
+      "grad_norm": 0.42956939339637756,
+      "learning_rate": 0.0003965782060750178,
+      "loss": 2.7752,
+      "step": 4526
+    },
+    {
+      "epoch": 0.579456,
+      "grad_norm": 0.445102334022522,
+      "learning_rate": 0.00039637538715971095,
+      "loss": 2.7913,
+      "step": 4527
+    },
+    {
+      "epoch": 0.579584,
+      "grad_norm": 0.45111846923828125,
+      "learning_rate": 0.00039617258605867633,
+      "loss": 2.8844,
+      "step": 4528
+    },
+    {
+      "epoch": 0.579712,
+      "grad_norm": 0.41817861795425415,
+      "learning_rate": 0.00039596980280677774,
+      "loss": 2.8932,
+      "step": 4529
+    },
+    {
+      "epoch": 0.57984,
+      "grad_norm": 0.41321733593940735,
+      "learning_rate": 0.0003957670374388762,
+      "loss": 2.9473,
+      "step": 4530
+    },
+    {
+      "epoch": 0.579968,
+      "grad_norm": 0.38630786538124084,
+      "learning_rate": 0.00039556428998982924,
+      "loss": 2.8501,
+      "step": 4531
+    },
+    {
+      "epoch": 0.580096,
+      "grad_norm": 0.4072386920452118,
+      "learning_rate": 0.00039536156049449153,
+      "loss": 2.7953,
+      "step": 4532
+    },
+    {
+      "epoch": 0.580224,
+      "grad_norm": 0.3894507884979248,
+      "learning_rate": 0.0003951588489877145,
+      "loss": 2.9232,
+      "step": 4533
+    },
+    {
+      "epoch": 0.580352,
+      "grad_norm": 0.44960924983024597,
+      "learning_rate": 0.0003949561555043469,
+      "loss": 2.9398,
+      "step": 4534
+    },
+    {
+      "epoch": 0.58048,
+      "grad_norm": 0.3769601285457611,
+      "learning_rate": 0.000394753480079234,
+      "loss": 2.9167,
+      "step": 4535
+    },
+    {
+      "epoch": 0.580608,
+      "grad_norm": 0.38681137561798096,
+      "learning_rate": 0.000394550822747218,
+      "loss": 2.7788,
+      "step": 4536
+    },
+    {
+      "epoch": 0.580736,
+      "grad_norm": 0.42727670073509216,
+      "learning_rate": 0.00039434818354313787,
+      "loss": 2.7283,
+      "step": 4537
+    },
+    {
+      "epoch": 0.580864,
+      "grad_norm": 0.4127942621707916,
+      "learning_rate": 0.00039414556250183,
+      "loss": 2.7701,
+      "step": 4538
+    },
+    {
+      "epoch": 0.580992,
+      "grad_norm": 0.3976554274559021,
+      "learning_rate": 0.0003939429596581271,
+      "loss": 2.8655,
+      "step": 4539
+    },
+    {
+      "epoch": 0.58112,
+      "grad_norm": 0.44627487659454346,
+      "learning_rate": 0.00039374037504685876,
+      "loss": 2.6719,
+      "step": 4540
+    },
+    {
+      "epoch": 0.581248,
+      "grad_norm": 0.3952978551387787,
+      "learning_rate": 0.00039353780870285204,
+      "loss": 2.836,
+      "step": 4541
+    },
+    {
+      "epoch": 0.581376,
+      "grad_norm": 0.40294861793518066,
+      "learning_rate": 0.00039333526066092993,
+      "loss": 2.9004,
+      "step": 4542
+    },
+    {
+      "epoch": 0.581504,
+      "grad_norm": 0.42145127058029175,
+      "learning_rate": 0.0003931327309559133,
+      "loss": 2.8786,
+      "step": 4543
+    },
+    {
+      "epoch": 0.581632,
+      "grad_norm": 0.41093525290489197,
+      "learning_rate": 0.000392930219622619,
+      "loss": 2.8844,
+      "step": 4544
+    },
+    {
+      "epoch": 0.58176,
+      "grad_norm": 0.4119255542755127,
+      "learning_rate": 0.0003927277266958614,
+      "loss": 2.8558,
+      "step": 4545
+    },
+    {
+      "epoch": 0.581888,
+      "grad_norm": 0.4012066423892975,
+      "learning_rate": 0.00039252525221045116,
+      "loss": 2.7698,
+      "step": 4546
+    },
+    {
+      "epoch": 0.582016,
+      "grad_norm": 0.39214983582496643,
+      "learning_rate": 0.00039232279620119586,
+      "loss": 2.9181,
+      "step": 4547
+    },
+    {
+      "epoch": 0.582144,
+      "grad_norm": 0.4473380148410797,
+      "learning_rate": 0.00039212035870290037,
+      "loss": 2.7375,
+      "step": 4548
+    },
+    {
+      "epoch": 0.582272,
+      "grad_norm": 0.4141892194747925,
+      "learning_rate": 0.00039191793975036583,
+      "loss": 2.977,
+      "step": 4549
+    },
+    {
+      "epoch": 0.5824,
+      "grad_norm": 0.45548468828201294,
+      "learning_rate": 0.0003917155393783903,
+      "loss": 2.7103,
+      "step": 4550
+    },
+    {
+      "epoch": 0.582528,
+      "grad_norm": 0.4304201602935791,
+      "learning_rate": 0.0003915131576217688,
+      "loss": 2.7018,
+      "step": 4551
+    },
+    {
+      "epoch": 0.582656,
+      "grad_norm": 0.4145083725452423,
+      "learning_rate": 0.0003913107945152934,
+      "loss": 2.8452,
+      "step": 4552
+    },
+    {
+      "epoch": 0.582784,
+      "grad_norm": 0.41397881507873535,
+      "learning_rate": 0.0003911084500937525,
+      "loss": 2.823,
+      "step": 4553
+    },
+    {
+      "epoch": 0.582912,
+      "grad_norm": 0.4547826051712036,
+      "learning_rate": 0.00039090612439193127,
+      "loss": 2.8622,
+      "step": 4554
+    },
+    {
+      "epoch": 0.58304,
+      "grad_norm": 0.43215927481651306,
+      "learning_rate": 0.00039070381744461217,
+      "loss": 2.8898,
+      "step": 4555
+    },
+    {
+      "epoch": 0.583168,
+      "grad_norm": 0.4003465175628662,
+      "learning_rate": 0.0003905015292865738,
+      "loss": 2.9395,
+      "step": 4556
+    },
+    {
+      "epoch": 0.583296,
+      "grad_norm": 0.422613263130188,
+      "learning_rate": 0.00039029925995259204,
+      "loss": 2.7313,
+      "step": 4557
+    },
+    {
+      "epoch": 0.583424,
+      "grad_norm": 0.40812626481056213,
+      "learning_rate": 0.00039009700947743907,
+      "loss": 2.7886,
+      "step": 4558
+    },
+    {
+      "epoch": 0.583552,
+      "grad_norm": 0.39730995893478394,
+      "learning_rate": 0.00038989477789588443,
+      "loss": 2.8622,
+      "step": 4559
+    },
+    {
+      "epoch": 0.58368,
+      "grad_norm": 0.41947972774505615,
+      "learning_rate": 0.00038969256524269393,
+      "loss": 2.9456,
+      "step": 4560
+    },
+    {
+      "epoch": 0.583808,
+      "grad_norm": 0.398448646068573,
+      "learning_rate": 0.00038949037155263007,
+      "loss": 2.6942,
+      "step": 4561
+    },
+    {
+      "epoch": 0.583936,
+      "grad_norm": 0.4343293011188507,
+      "learning_rate": 0.00038928819686045253,
+      "loss": 2.9113,
+      "step": 4562
+    },
+    {
+      "epoch": 0.584064,
+      "grad_norm": 0.376241534948349,
+      "learning_rate": 0.00038908604120091754,
+      "loss": 2.8678,
+      "step": 4563
+    },
+    {
+      "epoch": 0.584192,
+      "grad_norm": 0.38652652502059937,
+      "learning_rate": 0.00038888390460877794,
+      "loss": 2.9027,
+      "step": 4564
+    },
+    {
+      "epoch": 0.58432,
+      "grad_norm": 0.40761226415634155,
+      "learning_rate": 0.00038868178711878317,
+      "loss": 2.8536,
+      "step": 4565
+    },
+    {
+      "epoch": 0.584448,
+      "grad_norm": 0.3914927542209625,
+      "learning_rate": 0.0003884796887656798,
+      "loss": 2.8494,
+      "step": 4566
+    },
+    {
+      "epoch": 0.584576,
+      "grad_norm": 0.4287664294242859,
+      "learning_rate": 0.00038827760958421077,
+      "loss": 2.8314,
+      "step": 4567
+    },
+    {
+      "epoch": 0.584704,
+      "grad_norm": 0.41207054257392883,
+      "learning_rate": 0.0003880755496091158,
+      "loss": 2.8285,
+      "step": 4568
+    },
+    {
+      "epoch": 0.584832,
+      "grad_norm": 0.44358184933662415,
+      "learning_rate": 0.00038787350887513144,
+      "loss": 2.8423,
+      "step": 4569
+    },
+    {
+      "epoch": 0.58496,
+      "grad_norm": 0.4250726103782654,
+      "learning_rate": 0.00038767148741699063,
+      "loss": 2.9351,
+      "step": 4570
+    },
+    {
+      "epoch": 0.585088,
+      "grad_norm": 0.411933571100235,
+      "learning_rate": 0.00038746948526942353,
+      "loss": 2.7923,
+      "step": 4571
+    },
+    {
+      "epoch": 0.585216,
+      "grad_norm": 0.40414702892303467,
+      "learning_rate": 0.0003872675024671564,
+      "loss": 2.7473,
+      "step": 4572
+    },
+    {
+      "epoch": 0.585344,
+      "grad_norm": 0.41698548197746277,
+      "learning_rate": 0.00038706553904491257,
+      "loss": 2.8774,
+      "step": 4573
+    },
+    {
+      "epoch": 0.585472,
+      "grad_norm": 0.4121815264225006,
+      "learning_rate": 0.00038686359503741185,
+      "loss": 2.8207,
+      "step": 4574
+    },
+    {
+      "epoch": 0.5856,
+      "grad_norm": 0.4305737018585205,
+      "learning_rate": 0.0003866616704793706,
+      "loss": 2.7557,
+      "step": 4575
+    },
+    {
+      "epoch": 0.585728,
+      "grad_norm": 0.38933390378952026,
+      "learning_rate": 0.0003864597654055023,
+      "loss": 2.733,
+      "step": 4576
+    },
+    {
+      "epoch": 0.585856,
+      "grad_norm": 0.4103395342826843,
+      "learning_rate": 0.00038625787985051654,
+      "loss": 2.926,
+      "step": 4577
+    },
+    {
+      "epoch": 0.585984,
+      "grad_norm": 0.42509713768959045,
+      "learning_rate": 0.0003860560138491199,
+      "loss": 2.8867,
+      "step": 4578
+    },
+    {
+      "epoch": 0.586112,
+      "grad_norm": 0.42323920130729675,
+      "learning_rate": 0.0003858541674360151,
+      "loss": 2.8277,
+      "step": 4579
+    },
+    {
+      "epoch": 0.58624,
+      "grad_norm": 0.4284464120864868,
+      "learning_rate": 0.00038565234064590244,
+      "loss": 2.758,
+      "step": 4580
+    },
+    {
+      "epoch": 0.586368,
+      "grad_norm": 0.4026067554950714,
+      "learning_rate": 0.00038545053351347815,
+      "loss": 2.828,
+      "step": 4581
+    },
+    {
+      "epoch": 0.586496,
+      "grad_norm": 0.4181816577911377,
+      "learning_rate": 0.00038524874607343506,
+      "loss": 2.8099,
+      "step": 4582
+    },
+    {
+      "epoch": 0.586624,
+      "grad_norm": 0.4448768198490143,
+      "learning_rate": 0.00038504697836046274,
+      "loss": 2.8023,
+      "step": 4583
+    },
+    {
+      "epoch": 0.586752,
+      "grad_norm": 0.39640218019485474,
+      "learning_rate": 0.00038484523040924765,
+      "loss": 2.7434,
+      "step": 4584
+    },
+    {
+      "epoch": 0.58688,
+      "grad_norm": 0.4067550599575043,
+      "learning_rate": 0.0003846435022544724,
+      "loss": 2.7778,
+      "step": 4585
+    },
+    {
+      "epoch": 0.587008,
+      "grad_norm": 0.4234640300273895,
+      "learning_rate": 0.0003844417939308163,
+      "loss": 2.8495,
+      "step": 4586
+    },
+    {
+      "epoch": 0.587136,
+      "grad_norm": 0.40629374980926514,
+      "learning_rate": 0.0003842401054729556,
+      "loss": 2.8971,
+      "step": 4587
+    },
+    {
+      "epoch": 0.587264,
+      "grad_norm": 0.42675626277923584,
+      "learning_rate": 0.0003840384369155626,
+      "loss": 2.7879,
+      "step": 4588
+    },
+    {
+      "epoch": 0.587392,
+      "grad_norm": 0.39377298951148987,
+      "learning_rate": 0.0003838367882933068,
+      "loss": 2.8545,
+      "step": 4589
+    },
+    {
+      "epoch": 0.58752,
+      "grad_norm": 0.4275971055030823,
+      "learning_rate": 0.0003836351596408537,
+      "loss": 2.7985,
+      "step": 4590
+    },
+    {
+      "epoch": 0.587648,
+      "grad_norm": 0.39448603987693787,
+      "learning_rate": 0.00038343355099286583,
+      "loss": 2.7746,
+      "step": 4591
+    },
+    {
+      "epoch": 0.587776,
+      "grad_norm": 0.4326706528663635,
+      "learning_rate": 0.0003832319623840018,
+      "loss": 2.809,
+      "step": 4592
+    },
+    {
+      "epoch": 0.587904,
+      "grad_norm": 0.4068787693977356,
+      "learning_rate": 0.00038303039384891715,
+      "loss": 2.8726,
+      "step": 4593
+    },
+    {
+      "epoch": 0.588032,
+      "grad_norm": 0.42376211285591125,
+      "learning_rate": 0.0003828288454222638,
+      "loss": 2.7643,
+      "step": 4594
+    },
+    {
+      "epoch": 0.58816,
+      "grad_norm": 0.43122583627700806,
+      "learning_rate": 0.0003826273171386904,
+      "loss": 2.7982,
+      "step": 4595
+    },
+    {
+      "epoch": 0.588288,
+      "grad_norm": 0.4176432192325592,
+      "learning_rate": 0.0003824258090328418,
+      "loss": 2.877,
+      "step": 4596
+    },
+    {
+      "epoch": 0.588416,
+      "grad_norm": 0.38502609729766846,
+      "learning_rate": 0.0003822243211393596,
+      "loss": 2.8051,
+      "step": 4597
+    },
+    {
+      "epoch": 0.588544,
+      "grad_norm": 0.43909305334091187,
+      "learning_rate": 0.0003820228534928818,
+      "loss": 2.9163,
+      "step": 4598
+    },
+    {
+      "epoch": 0.588672,
+      "grad_norm": 0.3904136121273041,
+      "learning_rate": 0.0003818214061280435,
+      "loss": 2.8358,
+      "step": 4599
+    },
+    {
+      "epoch": 0.5888,
+      "grad_norm": 0.4050203859806061,
+      "learning_rate": 0.0003816199790794754,
+      "loss": 2.8659,
+      "step": 4600
+    },
+    {
+      "epoch": 0.588928,
+      "grad_norm": 0.41303691267967224,
+      "learning_rate": 0.0003814185723818053,
+      "loss": 2.8527,
+      "step": 4601
+    },
+    {
+      "epoch": 0.589056,
+      "grad_norm": 0.40307193994522095,
+      "learning_rate": 0.0003812171860696574,
+      "loss": 2.8261,
+      "step": 4602
+    },
+    {
+      "epoch": 0.589184,
+      "grad_norm": 0.40418559312820435,
+      "learning_rate": 0.00038101582017765226,
+      "loss": 2.7396,
+      "step": 4603
+    },
+    {
+      "epoch": 0.589312,
+      "grad_norm": 0.41955479979515076,
+      "learning_rate": 0.00038081447474040686,
+      "loss": 2.7899,
+      "step": 4604
+    },
+    {
+      "epoch": 0.58944,
+      "grad_norm": 0.4430882930755615,
+      "learning_rate": 0.00038061314979253506,
+      "loss": 2.7485,
+      "step": 4605
+    },
+    {
+      "epoch": 0.589568,
+      "grad_norm": 0.39909547567367554,
+      "learning_rate": 0.0003804118453686469,
+      "loss": 2.7594,
+      "step": 4606
+    },
+    {
+      "epoch": 0.589696,
+      "grad_norm": 0.47588175535202026,
+      "learning_rate": 0.0003802105615033488,
+      "loss": 3.1141,
+      "step": 4607
+    },
+    {
+      "epoch": 0.589824,
+      "grad_norm": 0.4102558493614197,
+      "learning_rate": 0.00038000929823124385,
+      "loss": 2.7738,
+      "step": 4608
+    },
+    {
+      "epoch": 0.589952,
+      "grad_norm": 0.43094733357429504,
+      "learning_rate": 0.0003798080555869318,
+      "loss": 2.7435,
+      "step": 4609
+    },
+    {
+      "epoch": 0.59008,
+      "grad_norm": 0.38973256945610046,
+      "learning_rate": 0.00037960683360500826,
+      "loss": 2.8129,
+      "step": 4610
+    },
+    {
+      "epoch": 0.590208,
+      "grad_norm": 0.4394755959510803,
+      "learning_rate": 0.0003794056323200657,
+      "loss": 2.7688,
+      "step": 4611
+    },
+    {
+      "epoch": 0.590336,
+      "grad_norm": 0.444799542427063,
+      "learning_rate": 0.00037920445176669313,
+      "loss": 2.7692,
+      "step": 4612
+    },
+    {
+      "epoch": 0.590464,
+      "grad_norm": 0.4100411832332611,
+      "learning_rate": 0.00037900329197947563,
+      "loss": 2.7827,
+      "step": 4613
+    },
+    {
+      "epoch": 0.590592,
+      "grad_norm": 0.41132843494415283,
+      "learning_rate": 0.00037880215299299487,
+      "loss": 2.835,
+      "step": 4614
+    },
+    {
+      "epoch": 0.59072,
+      "grad_norm": 0.44901037216186523,
+      "learning_rate": 0.0003786010348418292,
+      "loss": 2.8223,
+      "step": 4615
+    },
+    {
+      "epoch": 0.590848,
+      "grad_norm": 0.37999314069747925,
+      "learning_rate": 0.0003783999375605527,
+      "loss": 2.8178,
+      "step": 4616
+    },
+    {
+      "epoch": 0.590976,
+      "grad_norm": 0.4282040596008301,
+      "learning_rate": 0.0003781988611837369,
+      "loss": 3.0105,
+      "step": 4617
+    },
+    {
+      "epoch": 0.591104,
+      "grad_norm": 0.38680872321128845,
+      "learning_rate": 0.000377997805745949,
+      "loss": 2.944,
+      "step": 4618
+    },
+    {
+      "epoch": 0.591232,
+      "grad_norm": 0.40538156032562256,
+      "learning_rate": 0.00037779677128175245,
+      "loss": 2.9212,
+      "step": 4619
+    },
+    {
+      "epoch": 0.59136,
+      "grad_norm": 0.40318819880485535,
+      "learning_rate": 0.0003775957578257078,
+      "loss": 2.7893,
+      "step": 4620
+    },
+    {
+      "epoch": 0.591488,
+      "grad_norm": 0.4068361520767212,
+      "learning_rate": 0.0003773947654123713,
+      "loss": 2.7512,
+      "step": 4621
+    },
+    {
+      "epoch": 0.591616,
+      "grad_norm": 0.44049420952796936,
+      "learning_rate": 0.00037719379407629593,
+      "loss": 2.8466,
+      "step": 4622
+    },
+    {
+      "epoch": 0.591744,
+      "grad_norm": 0.43102166056632996,
+      "learning_rate": 0.0003769928438520312,
+      "loss": 2.8489,
+      "step": 4623
+    },
+    {
+      "epoch": 0.591872,
+      "grad_norm": 0.42289960384368896,
+      "learning_rate": 0.0003767919147741226,
+      "loss": 2.9822,
+      "step": 4624
+    },
+    {
+      "epoch": 0.592,
+      "grad_norm": 0.40847864747047424,
+      "learning_rate": 0.000376591006877112,
+      "loss": 2.7682,
+      "step": 4625
+    },
+    {
+      "epoch": 0.592128,
+      "grad_norm": 0.4043353199958801,
+      "learning_rate": 0.0003763901201955381,
+      "loss": 2.8359,
+      "step": 4626
+    },
+    {
+      "epoch": 0.592256,
+      "grad_norm": 0.3913222849369049,
+      "learning_rate": 0.00037618925476393567,
+      "loss": 2.9356,
+      "step": 4627
+    },
+    {
+      "epoch": 0.592384,
+      "grad_norm": 0.3801572024822235,
+      "learning_rate": 0.0003759884106168356,
+      "loss": 2.8605,
+      "step": 4628
+    },
+    {
+      "epoch": 0.592512,
+      "grad_norm": 0.40004250407218933,
+      "learning_rate": 0.00037578758778876545,
+      "loss": 2.7765,
+      "step": 4629
+    },
+    {
+      "epoch": 0.59264,
+      "grad_norm": 0.4680653512477875,
+      "learning_rate": 0.000375586786314249,
+      "loss": 2.9076,
+      "step": 4630
+    },
+    {
+      "epoch": 0.592768,
+      "grad_norm": 0.4274962842464447,
+      "learning_rate": 0.00037538600622780626,
+      "loss": 2.8569,
+      "step": 4631
+    },
+    {
+      "epoch": 0.592896,
+      "grad_norm": 0.44849923253059387,
+      "learning_rate": 0.0003751852475639536,
+      "loss": 2.8682,
+      "step": 4632
+    },
+    {
+      "epoch": 0.593024,
+      "grad_norm": 0.3967364430427551,
+      "learning_rate": 0.000374984510357204,
+      "loss": 2.9322,
+      "step": 4633
+    },
+    {
+      "epoch": 0.593152,
+      "grad_norm": 0.40151575207710266,
+      "learning_rate": 0.00037478379464206635,
+      "loss": 2.8699,
+      "step": 4634
+    },
+    {
+      "epoch": 0.59328,
+      "grad_norm": 0.43061354756355286,
+      "learning_rate": 0.00037458310045304584,
+      "loss": 2.9264,
+      "step": 4635
+    },
+    {
+      "epoch": 0.593408,
+      "grad_norm": 0.38285356760025024,
+      "learning_rate": 0.0003743824278246445,
+      "loss": 2.8516,
+      "step": 4636
+    },
+    {
+      "epoch": 0.593536,
+      "grad_norm": 0.4346286952495575,
+      "learning_rate": 0.00037418177679136,
+      "loss": 2.7379,
+      "step": 4637
+    },
+    {
+      "epoch": 0.593664,
+      "grad_norm": 0.4005469083786011,
+      "learning_rate": 0.00037398114738768686,
+      "loss": 2.75,
+      "step": 4638
+    },
+    {
+      "epoch": 0.593792,
+      "grad_norm": 0.394107460975647,
+      "learning_rate": 0.0003737805396481153,
+      "loss": 2.7611,
+      "step": 4639
+    },
+    {
+      "epoch": 0.59392,
+      "grad_norm": 0.4438163936138153,
+      "learning_rate": 0.0003735799536071322,
+      "loss": 2.8269,
+      "step": 4640
+    },
+    {
+      "epoch": 0.594048,
+      "grad_norm": 0.425460547208786,
+      "learning_rate": 0.00037337938929922075,
+      "loss": 2.7481,
+      "step": 4641
+    },
+    {
+      "epoch": 0.594176,
+      "grad_norm": 0.3959190845489502,
+      "learning_rate": 0.00037317884675886024,
+      "loss": 2.7494,
+      "step": 4642
+    },
+    {
+      "epoch": 0.594304,
+      "grad_norm": 0.4410264790058136,
+      "learning_rate": 0.00037297832602052605,
+      "loss": 2.794,
+      "step": 4643
+    },
+    {
+      "epoch": 0.594432,
+      "grad_norm": 0.4001834988594055,
+      "learning_rate": 0.0003727778271186901,
+      "loss": 2.8332,
+      "step": 4644
+    },
+    {
+      "epoch": 0.59456,
+      "grad_norm": 0.5711485147476196,
+      "learning_rate": 0.00037257735008782073,
+      "loss": 2.8483,
+      "step": 4645
+    },
+    {
+      "epoch": 0.594688,
+      "grad_norm": 0.4100117087364197,
+      "learning_rate": 0.00037237689496238214,
+      "loss": 2.9241,
+      "step": 4646
+    },
+    {
+      "epoch": 0.594816,
+      "grad_norm": 0.4308624267578125,
+      "learning_rate": 0.00037217646177683465,
+      "loss": 2.8886,
+      "step": 4647
+    },
+    {
+      "epoch": 0.594944,
+      "grad_norm": 0.44200599193573,
+      "learning_rate": 0.00037197605056563544,
+      "loss": 3.0399,
+      "step": 4648
+    },
+    {
+      "epoch": 0.595072,
+      "grad_norm": 0.38219213485717773,
+      "learning_rate": 0.00037177566136323726,
+      "loss": 2.8379,
+      "step": 4649
+    },
+    {
+      "epoch": 0.5952,
+      "grad_norm": 0.37804341316223145,
+      "learning_rate": 0.00037157529420408933,
+      "loss": 2.9898,
+      "step": 4650
+    },
+    {
+      "epoch": 0.595328,
+      "grad_norm": 0.4100346267223358,
+      "learning_rate": 0.0003713749491226372,
+      "loss": 2.8485,
+      "step": 4651
+    },
+    {
+      "epoch": 0.595456,
+      "grad_norm": 0.41512325406074524,
+      "learning_rate": 0.00037117462615332254,
+      "loss": 2.8529,
+      "step": 4652
+    },
+    {
+      "epoch": 0.595584,
+      "grad_norm": 0.39969602227211,
+      "learning_rate": 0.00037097432533058283,
+      "loss": 2.6922,
+      "step": 4653
+    },
+    {
+      "epoch": 0.595712,
+      "grad_norm": 0.40442222356796265,
+      "learning_rate": 0.00037077404668885264,
+      "loss": 2.8133,
+      "step": 4654
+    },
+    {
+      "epoch": 0.59584,
+      "grad_norm": 0.42178693413734436,
+      "learning_rate": 0.00037057379026256184,
+      "loss": 2.8145,
+      "step": 4655
+    },
+    {
+      "epoch": 0.595968,
+      "grad_norm": 0.45349469780921936,
+      "learning_rate": 0.000370373556086137,
+      "loss": 2.8573,
+      "step": 4656
+    },
+    {
+      "epoch": 0.596096,
+      "grad_norm": 0.3801501989364624,
+      "learning_rate": 0.0003701733441940007,
+      "loss": 2.8272,
+      "step": 4657
+    },
+    {
+      "epoch": 0.596224,
+      "grad_norm": 0.3974490463733673,
+      "learning_rate": 0.00036997315462057145,
+      "loss": 2.8207,
+      "step": 4658
+    },
+    {
+      "epoch": 0.596352,
+      "grad_norm": 0.3667813837528229,
+      "learning_rate": 0.00036977298740026445,
+      "loss": 2.8665,
+      "step": 4659
+    },
+    {
+      "epoch": 0.59648,
+      "grad_norm": 0.41169121861457825,
+      "learning_rate": 0.0003695728425674906,
+      "loss": 2.6631,
+      "step": 4660
+    },
+    {
+      "epoch": 0.596608,
+      "grad_norm": 0.440548300743103,
+      "learning_rate": 0.00036937272015665705,
+      "loss": 2.7999,
+      "step": 4661
+    },
+    {
+      "epoch": 0.596736,
+      "grad_norm": 0.39659005403518677,
+      "learning_rate": 0.00036917262020216726,
+      "loss": 2.8401,
+      "step": 4662
+    },
+    {
+      "epoch": 0.596864,
+      "grad_norm": 0.45176559686660767,
+      "learning_rate": 0.00036897254273842085,
+      "loss": 2.8527,
+      "step": 4663
+    },
+    {
+      "epoch": 0.596992,
+      "grad_norm": 0.40221843123435974,
+      "learning_rate": 0.0003687724877998133,
+      "loss": 2.8682,
+      "step": 4664
+    },
+    {
+      "epoch": 0.59712,
+      "grad_norm": 0.4261503219604492,
+      "learning_rate": 0.00036857245542073645,
+      "loss": 2.8805,
+      "step": 4665
+    },
+    {
+      "epoch": 0.597248,
+      "grad_norm": 0.4124809503555298,
+      "learning_rate": 0.00036837244563557814,
+      "loss": 2.8575,
+      "step": 4666
+    },
+    {
+      "epoch": 0.597376,
+      "grad_norm": 0.39208143949508667,
+      "learning_rate": 0.0003681724584787225,
+      "loss": 2.7539,
+      "step": 4667
+    },
+    {
+      "epoch": 0.597504,
+      "grad_norm": 0.39954090118408203,
+      "learning_rate": 0.00036797249398454933,
+      "loss": 2.8706,
+      "step": 4668
+    },
+    {
+      "epoch": 0.597632,
+      "grad_norm": 0.43096694350242615,
+      "learning_rate": 0.00036777255218743514,
+      "loss": 2.9107,
+      "step": 4669
+    },
+    {
+      "epoch": 0.59776,
+      "grad_norm": 0.4183279871940613,
+      "learning_rate": 0.00036757263312175217,
+      "loss": 2.8808,
+      "step": 4670
+    },
+    {
+      "epoch": 0.597888,
+      "grad_norm": 0.42403644323349,
+      "learning_rate": 0.0003673727368218687,
+      "loss": 2.9406,
+      "step": 4671
+    },
+    {
+      "epoch": 0.598016,
+      "grad_norm": 0.41394689679145813,
+      "learning_rate": 0.00036717286332214926,
+      "loss": 2.8537,
+      "step": 4672
+    },
+    {
+      "epoch": 0.598144,
+      "grad_norm": 0.4059096574783325,
+      "learning_rate": 0.0003669730126569545,
+      "loss": 2.7404,
+      "step": 4673
+    },
+    {
+      "epoch": 0.598272,
+      "grad_norm": 0.39557304978370667,
+      "learning_rate": 0.00036677318486064126,
+      "loss": 2.8123,
+      "step": 4674
+    },
+    {
+      "epoch": 0.5984,
+      "grad_norm": 0.42253443598747253,
+      "learning_rate": 0.0003665733799675619,
+      "loss": 2.8352,
+      "step": 4675
+    },
+    {
+      "epoch": 0.598528,
+      "grad_norm": 0.42171937227249146,
+      "learning_rate": 0.0003663735980120653,
+      "loss": 3.0312,
+      "step": 4676
+    },
+    {
+      "epoch": 0.598656,
+      "grad_norm": 0.38881590962409973,
+      "learning_rate": 0.00036617383902849644,
+      "loss": 2.8471,
+      "step": 4677
+    },
+    {
+      "epoch": 0.598784,
+      "grad_norm": 0.4172724485397339,
+      "learning_rate": 0.000365974103051196,
+      "loss": 2.7703,
+      "step": 4678
+    },
+    {
+      "epoch": 0.598912,
+      "grad_norm": 0.4066183567047119,
+      "learning_rate": 0.0003657743901145011,
+      "loss": 2.8203,
+      "step": 4679
+    },
+    {
+      "epoch": 0.59904,
+      "grad_norm": 0.4171360731124878,
+      "learning_rate": 0.00036557470025274454,
+      "loss": 2.7992,
+      "step": 4680
+    },
+    {
+      "epoch": 0.599168,
+      "grad_norm": 0.3824343979358673,
+      "learning_rate": 0.0003653750335002552,
+      "loss": 3.0341,
+      "step": 4681
+    },
+    {
+      "epoch": 0.599296,
+      "grad_norm": 0.41188833117485046,
+      "learning_rate": 0.00036517538989135865,
+      "loss": 2.9099,
+      "step": 4682
+    },
+    {
+      "epoch": 0.599424,
+      "grad_norm": 0.4469556212425232,
+      "learning_rate": 0.0003649757694603754,
+      "loss": 2.9278,
+      "step": 4683
+    },
+    {
+      "epoch": 0.599552,
+      "grad_norm": 0.40267783403396606,
+      "learning_rate": 0.00036477617224162285,
+      "loss": 2.9203,
+      "step": 4684
+    },
+    {
+      "epoch": 0.59968,
+      "grad_norm": 0.42181846499443054,
+      "learning_rate": 0.0003645765982694139,
+      "loss": 2.8359,
+      "step": 4685
+    },
+    {
+      "epoch": 0.599808,
+      "grad_norm": 0.43285080790519714,
+      "learning_rate": 0.0003643770475780576,
+      "loss": 2.8064,
+      "step": 4686
+    },
+    {
+      "epoch": 0.599936,
+      "grad_norm": 0.3923158049583435,
+      "learning_rate": 0.00036417752020185916,
+      "loss": 2.8128,
+      "step": 4687
+    },
+    {
+      "epoch": 0.600064,
+      "grad_norm": 0.40862080454826355,
+      "learning_rate": 0.0003639780161751196,
+      "loss": 2.7981,
+      "step": 4688
+    },
+    {
+      "epoch": 0.600192,
+      "grad_norm": 0.4189339280128479,
+      "learning_rate": 0.00036377853553213593,
+      "loss": 2.8536,
+      "step": 4689
+    },
+    {
+      "epoch": 0.60032,
+      "grad_norm": 0.4198634922504425,
+      "learning_rate": 0.00036357907830720116,
+      "loss": 2.869,
+      "step": 4690
+    },
+    {
+      "epoch": 0.600448,
+      "grad_norm": 0.39705684781074524,
+      "learning_rate": 0.0003633796445346044,
+      "loss": 2.7584,
+      "step": 4691
+    },
+    {
+      "epoch": 0.600576,
+      "grad_norm": 0.44507789611816406,
+      "learning_rate": 0.00036318023424863053,
+      "loss": 2.9009,
+      "step": 4692
+    },
+    {
+      "epoch": 0.600704,
+      "grad_norm": 0.4001615643501282,
+      "learning_rate": 0.00036298084748356076,
+      "loss": 2.8429,
+      "step": 4693
+    },
+    {
+      "epoch": 0.600832,
+      "grad_norm": 0.42016562819480896,
+      "learning_rate": 0.0003627814842736715,
+      "loss": 2.754,
+      "step": 4694
+    },
+    {
+      "epoch": 0.60096,
+      "grad_norm": 0.43571996688842773,
+      "learning_rate": 0.000362582144653236,
+      "loss": 2.8089,
+      "step": 4695
+    },
+    {
+      "epoch": 0.601088,
+      "grad_norm": 0.3975415527820587,
+      "learning_rate": 0.000362382828656523,
+      "loss": 3.0531,
+      "step": 4696
+    },
+    {
+      "epoch": 0.601216,
+      "grad_norm": 0.40989887714385986,
+      "learning_rate": 0.000362183536317797,
+      "loss": 2.9395,
+      "step": 4697
+    },
+    {
+      "epoch": 0.601344,
+      "grad_norm": 0.4147094488143921,
+      "learning_rate": 0.0003619842676713189,
+      "loss": 2.9296,
+      "step": 4698
+    },
+    {
+      "epoch": 0.601472,
+      "grad_norm": 0.3667636215686798,
+      "learning_rate": 0.00036178502275134507,
+      "loss": 2.7919,
+      "step": 4699
+    },
+    {
+      "epoch": 0.6016,
+      "grad_norm": 0.42983478307724,
+      "learning_rate": 0.0003615858015921283,
+      "loss": 2.791,
+      "step": 4700
+    },
+    {
+      "epoch": 0.601728,
+      "grad_norm": 0.36199381947517395,
+      "learning_rate": 0.00036138660422791667,
+      "loss": 2.85,
+      "step": 4701
+    },
+    {
+      "epoch": 0.601856,
+      "grad_norm": 0.417850524187088,
+      "learning_rate": 0.000361187430692955,
+      "loss": 2.8323,
+      "step": 4702
+    },
+    {
+      "epoch": 0.601984,
+      "grad_norm": 0.38273197412490845,
+      "learning_rate": 0.0003609882810214832,
+      "loss": 2.8529,
+      "step": 4703
+    },
+    {
+      "epoch": 0.602112,
+      "grad_norm": 0.43153995275497437,
+      "learning_rate": 0.0003607891552477373,
+      "loss": 2.847,
+      "step": 4704
+    },
+    {
+      "epoch": 0.60224,
+      "grad_norm": 0.4200153648853302,
+      "learning_rate": 0.00036059005340594957,
+      "loss": 2.8992,
+      "step": 4705
+    },
+    {
+      "epoch": 0.602368,
+      "grad_norm": 0.393309623003006,
+      "learning_rate": 0.0003603909755303479,
+      "loss": 2.7472,
+      "step": 4706
+    },
+    {
+      "epoch": 0.602496,
+      "grad_norm": 0.3958575427532196,
+      "learning_rate": 0.0003601919216551559,
+      "loss": 2.827,
+      "step": 4707
+    },
+    {
+      "epoch": 0.602624,
+      "grad_norm": 0.4056946337223053,
+      "learning_rate": 0.0003599928918145934,
+      "loss": 2.792,
+      "step": 4708
+    },
+    {
+      "epoch": 0.602752,
+      "grad_norm": 0.4033781588077545,
+      "learning_rate": 0.00035979388604287575,
+      "loss": 2.898,
+      "step": 4709
+    },
+    {
+      "epoch": 0.60288,
+      "grad_norm": 0.4044368267059326,
+      "learning_rate": 0.0003595949043742147,
+      "loss": 2.7694,
+      "step": 4710
+    },
+    {
+      "epoch": 0.603008,
+      "grad_norm": 0.3989121913909912,
+      "learning_rate": 0.00035939594684281735,
+      "loss": 2.8155,
+      "step": 4711
+    },
+    {
+      "epoch": 0.603136,
+      "grad_norm": 0.42482617497444153,
+      "learning_rate": 0.0003591970134828866,
+      "loss": 2.8795,
+      "step": 4712
+    },
+    {
+      "epoch": 0.603264,
+      "grad_norm": 0.37808436155319214,
+      "learning_rate": 0.00035899810432862166,
+      "loss": 2.8465,
+      "step": 4713
+    },
+    {
+      "epoch": 0.603392,
+      "grad_norm": 0.43895891308784485,
+      "learning_rate": 0.0003587992194142173,
+      "loss": 2.9094,
+      "step": 4714
+    },
+    {
+      "epoch": 0.60352,
+      "grad_norm": 0.43961450457572937,
+      "learning_rate": 0.00035860035877386385,
+      "loss": 2.9246,
+      "step": 4715
+    },
+    {
+      "epoch": 0.603648,
+      "grad_norm": 0.43334969878196716,
+      "learning_rate": 0.0003584015224417482,
+      "loss": 2.9384,
+      "step": 4716
+    },
+    {
+      "epoch": 0.603776,
+      "grad_norm": 0.41877445578575134,
+      "learning_rate": 0.0003582027104520523,
+      "loss": 2.88,
+      "step": 4717
+    },
+    {
+      "epoch": 0.603904,
+      "grad_norm": 0.4185437262058258,
+      "learning_rate": 0.0003580039228389541,
+      "loss": 2.8486,
+      "step": 4718
+    },
+    {
+      "epoch": 0.604032,
+      "grad_norm": 0.4227961003780365,
+      "learning_rate": 0.00035780515963662786,
+      "loss": 2.7391,
+      "step": 4719
+    },
+    {
+      "epoch": 0.60416,
+      "grad_norm": 0.4232025146484375,
+      "learning_rate": 0.0003576064208792433,
+      "loss": 2.8899,
+      "step": 4720
+    },
+    {
+      "epoch": 0.604288,
+      "grad_norm": 0.39582961797714233,
+      "learning_rate": 0.0003574077066009657,
+      "loss": 2.8509,
+      "step": 4721
+    },
+    {
+      "epoch": 0.604416,
+      "grad_norm": 0.3863828778266907,
+      "learning_rate": 0.0003572090168359563,
+      "loss": 2.6942,
+      "step": 4722
+    },
+    {
+      "epoch": 0.604544,
+      "grad_norm": 0.404024213552475,
+      "learning_rate": 0.0003570103516183724,
+      "loss": 2.8573,
+      "step": 4723
+    },
+    {
+      "epoch": 0.604672,
+      "grad_norm": 0.4166001081466675,
+      "learning_rate": 0.00035681171098236674,
+      "loss": 2.8911,
+      "step": 4724
+    },
+    {
+      "epoch": 0.6048,
+      "grad_norm": 0.4076208472251892,
+      "learning_rate": 0.0003566130949620878,
+      "loss": 2.7446,
+      "step": 4725
+    },
+    {
+      "epoch": 0.604928,
+      "grad_norm": 0.4487946331501007,
+      "learning_rate": 0.0003564145035916803,
+      "loss": 2.9435,
+      "step": 4726
+    },
+    {
+      "epoch": 0.605056,
+      "grad_norm": 0.42144620418548584,
+      "learning_rate": 0.000356215936905284,
+      "loss": 2.9475,
+      "step": 4727
+    },
+    {
+      "epoch": 0.605184,
+      "grad_norm": 0.39264583587646484,
+      "learning_rate": 0.00035601739493703515,
+      "loss": 2.8508,
+      "step": 4728
+    },
+    {
+      "epoch": 0.605312,
+      "grad_norm": 0.40834522247314453,
+      "learning_rate": 0.00035581887772106536,
+      "loss": 2.8106,
+      "step": 4729
+    },
+    {
+      "epoch": 0.60544,
+      "grad_norm": 0.42250579595565796,
+      "learning_rate": 0.00035562038529150184,
+      "loss": 2.8532,
+      "step": 4730
+    },
+    {
+      "epoch": 0.605568,
+      "grad_norm": 0.4140843152999878,
+      "learning_rate": 0.00035542191768246807,
+      "loss": 2.892,
+      "step": 4731
+    },
+    {
+      "epoch": 0.605696,
+      "grad_norm": 0.39055386185646057,
+      "learning_rate": 0.0003552234749280827,
+      "loss": 2.6337,
+      "step": 4732
+    },
+    {
+      "epoch": 0.605824,
+      "grad_norm": 0.39308464527130127,
+      "learning_rate": 0.00035502505706246025,
+      "loss": 2.8166,
+      "step": 4733
+    },
+    {
+      "epoch": 0.605952,
+      "grad_norm": 0.40459850430488586,
+      "learning_rate": 0.0003548266641197113,
+      "loss": 2.7079,
+      "step": 4734
+    },
+    {
+      "epoch": 0.60608,
+      "grad_norm": 0.426887184381485,
+      "learning_rate": 0.0003546282961339418,
+      "loss": 2.7381,
+      "step": 4735
+    },
+    {
+      "epoch": 0.606208,
+      "grad_norm": 0.402901291847229,
+      "learning_rate": 0.0003544299531392533,
+      "loss": 2.8827,
+      "step": 4736
+    },
+    {
+      "epoch": 0.606336,
+      "grad_norm": 0.4120728373527527,
+      "learning_rate": 0.00035423163516974356,
+      "loss": 2.7721,
+      "step": 4737
+    },
+    {
+      "epoch": 0.606464,
+      "grad_norm": 0.4046798348426819,
+      "learning_rate": 0.0003540333422595057,
+      "loss": 2.8365,
+      "step": 4738
+    },
+    {
+      "epoch": 0.606592,
+      "grad_norm": 0.4221363961696625,
+      "learning_rate": 0.0003538350744426285,
+      "loss": 2.8709,
+      "step": 4739
+    },
+    {
+      "epoch": 0.60672,
+      "grad_norm": 0.41032153367996216,
+      "learning_rate": 0.0003536368317531963,
+      "loss": 2.8234,
+      "step": 4740
+    },
+    {
+      "epoch": 0.606848,
+      "grad_norm": 0.40053024888038635,
+      "learning_rate": 0.00035343861422528974,
+      "loss": 2.7983,
+      "step": 4741
+    },
+    {
+      "epoch": 0.606976,
+      "grad_norm": 0.3842395842075348,
+      "learning_rate": 0.00035324042189298447,
+      "loss": 2.8671,
+      "step": 4742
+    },
+    {
+      "epoch": 0.607104,
+      "grad_norm": 0.3843987286090851,
+      "learning_rate": 0.00035304225479035193,
+      "loss": 2.7831,
+      "step": 4743
+    },
+    {
+      "epoch": 0.607232,
+      "grad_norm": 0.40148448944091797,
+      "learning_rate": 0.0003528441129514595,
+      "loss": 2.8258,
+      "step": 4744
+    },
+    {
+      "epoch": 0.60736,
+      "grad_norm": 0.398846834897995,
+      "learning_rate": 0.0003526459964103701,
+      "loss": 2.8508,
+      "step": 4745
+    },
+    {
+      "epoch": 0.607488,
+      "grad_norm": 0.3936963379383087,
+      "learning_rate": 0.000352447905201142,
+      "loss": 2.783,
+      "step": 4746
+    },
+    {
+      "epoch": 0.607616,
+      "grad_norm": 0.42755788564682007,
+      "learning_rate": 0.0003522498393578297,
+      "loss": 2.9895,
+      "step": 4747
+    },
+    {
+      "epoch": 0.607744,
+      "grad_norm": 0.39429041743278503,
+      "learning_rate": 0.0003520517989144829,
+      "loss": 2.8621,
+      "step": 4748
+    },
+    {
+      "epoch": 0.607872,
+      "grad_norm": 0.39354458451271057,
+      "learning_rate": 0.0003518537839051471,
+      "loss": 2.784,
+      "step": 4749
+    },
+    {
+      "epoch": 0.608,
+      "grad_norm": 0.4308680295944214,
+      "learning_rate": 0.0003516557943638632,
+      "loss": 2.8284,
+      "step": 4750
+    },
+    {
+      "epoch": 0.608128,
+      "grad_norm": 0.4267406761646271,
+      "learning_rate": 0.00035145783032466797,
+      "loss": 2.7299,
+      "step": 4751
+    },
+    {
+      "epoch": 0.608256,
+      "grad_norm": 0.39339423179626465,
+      "learning_rate": 0.00035125989182159393,
+      "loss": 2.9832,
+      "step": 4752
+    },
+    {
+      "epoch": 0.608384,
+      "grad_norm": 0.4253215193748474,
+      "learning_rate": 0.0003510619788886687,
+      "loss": 2.7781,
+      "step": 4753
+    },
+    {
+      "epoch": 0.608512,
+      "grad_norm": 0.3966854214668274,
+      "learning_rate": 0.0003508640915599159,
+      "loss": 2.9386,
+      "step": 4754
+    },
+    {
+      "epoch": 0.60864,
+      "grad_norm": 0.45508506894111633,
+      "learning_rate": 0.0003506662298693546,
+      "loss": 2.7479,
+      "step": 4755
+    },
+    {
+      "epoch": 0.608768,
+      "grad_norm": 0.4109669029712677,
+      "learning_rate": 0.00035046839385099977,
+      "loss": 2.8211,
+      "step": 4756
+    },
+    {
+      "epoch": 0.608896,
+      "grad_norm": 0.4231646656990051,
+      "learning_rate": 0.0003502705835388616,
+      "loss": 2.8075,
+      "step": 4757
+    },
+    {
+      "epoch": 0.609024,
+      "grad_norm": 0.41770023107528687,
+      "learning_rate": 0.00035007279896694576,
+      "loss": 2.8306,
+      "step": 4758
+    },
+    {
+      "epoch": 0.609152,
+      "grad_norm": 0.43888309597969055,
+      "learning_rate": 0.0003498750401692541,
+      "loss": 2.948,
+      "step": 4759
+    },
+    {
+      "epoch": 0.60928,
+      "grad_norm": 0.44682812690734863,
+      "learning_rate": 0.00034967730717978335,
+      "loss": 2.8862,
+      "step": 4760
+    },
+    {
+      "epoch": 0.609408,
+      "grad_norm": 0.42988529801368713,
+      "learning_rate": 0.0003494796000325261,
+      "loss": 2.9044,
+      "step": 4761
+    },
+    {
+      "epoch": 0.609536,
+      "grad_norm": 0.40498408675193787,
+      "learning_rate": 0.0003492819187614707,
+      "loss": 2.7906,
+      "step": 4762
+    },
+    {
+      "epoch": 0.609664,
+      "grad_norm": 0.4247494041919708,
+      "learning_rate": 0.00034908426340060075,
+      "loss": 2.7808,
+      "step": 4763
+    },
+    {
+      "epoch": 0.609792,
+      "grad_norm": 0.4085414409637451,
+      "learning_rate": 0.00034888663398389524,
+      "loss": 2.8797,
+      "step": 4764
+    },
+    {
+      "epoch": 0.60992,
+      "grad_norm": 0.4360335171222687,
+      "learning_rate": 0.0003486890305453293,
+      "loss": 2.9229,
+      "step": 4765
+    },
+    {
+      "epoch": 0.610048,
+      "grad_norm": 0.4120417833328247,
+      "learning_rate": 0.0003484914531188733,
+      "loss": 2.8715,
+      "step": 4766
+    },
+    {
+      "epoch": 0.610176,
+      "grad_norm": 0.3757292330265045,
+      "learning_rate": 0.00034829390173849297,
+      "loss": 2.8515,
+      "step": 4767
+    },
+    {
+      "epoch": 0.610304,
+      "grad_norm": 0.4303332567214966,
+      "learning_rate": 0.00034809637643814964,
+      "loss": 2.7495,
+      "step": 4768
+    },
+    {
+      "epoch": 0.610432,
+      "grad_norm": 0.42687296867370605,
+      "learning_rate": 0.0003478988772518001,
+      "loss": 3.0171,
+      "step": 4769
+    },
+    {
+      "epoch": 0.61056,
+      "grad_norm": 0.3974483609199524,
+      "learning_rate": 0.000347701404213397,
+      "loss": 2.8394,
+      "step": 4770
+    },
+    {
+      "epoch": 0.610688,
+      "grad_norm": 0.41046562790870667,
+      "learning_rate": 0.00034750395735688803,
+      "loss": 2.8177,
+      "step": 4771
+    },
+    {
+      "epoch": 0.610816,
+      "grad_norm": 0.4460611939430237,
+      "learning_rate": 0.00034730653671621663,
+      "loss": 2.8081,
+      "step": 4772
+    },
+    {
+      "epoch": 0.610944,
+      "grad_norm": 0.37635326385498047,
+      "learning_rate": 0.0003471091423253216,
+      "loss": 2.8241,
+      "step": 4773
+    },
+    {
+      "epoch": 0.611072,
+      "grad_norm": 0.40865176916122437,
+      "learning_rate": 0.0003469117742181378,
+      "loss": 2.8373,
+      "step": 4774
+    },
+    {
+      "epoch": 0.6112,
+      "grad_norm": 0.41607847809791565,
+      "learning_rate": 0.00034671443242859464,
+      "loss": 2.91,
+      "step": 4775
+    },
+    {
+      "epoch": 0.611328,
+      "grad_norm": 0.4538770616054535,
+      "learning_rate": 0.0003465171169906175,
+      "loss": 2.7991,
+      "step": 4776
+    },
+    {
+      "epoch": 0.611456,
+      "grad_norm": 0.41892874240875244,
+      "learning_rate": 0.00034631982793812744,
+      "loss": 2.8626,
+      "step": 4777
+    },
+    {
+      "epoch": 0.611584,
+      "grad_norm": 0.40284988284111023,
+      "learning_rate": 0.0003461225653050405,
+      "loss": 2.8024,
+      "step": 4778
+    },
+    {
+      "epoch": 0.611712,
+      "grad_norm": 0.4285210072994232,
+      "learning_rate": 0.0003459253291252684,
+      "loss": 2.7573,
+      "step": 4779
+    },
+    {
+      "epoch": 0.61184,
+      "grad_norm": 0.39288488030433655,
+      "learning_rate": 0.00034572811943271855,
+      "loss": 2.8149,
+      "step": 4780
+    },
+    {
+      "epoch": 0.611968,
+      "grad_norm": 0.4032428562641144,
+      "learning_rate": 0.0003455309362612934,
+      "loss": 2.8079,
+      "step": 4781
+    },
+    {
+      "epoch": 0.612096,
+      "grad_norm": 0.38669681549072266,
+      "learning_rate": 0.000345333779644891,
+      "loss": 2.8281,
+      "step": 4782
+    },
+    {
+      "epoch": 0.612224,
+      "grad_norm": 0.3964182138442993,
+      "learning_rate": 0.00034513664961740475,
+      "loss": 2.876,
+      "step": 4783
+    },
+    {
+      "epoch": 0.612352,
+      "grad_norm": 0.42288240790367126,
+      "learning_rate": 0.00034493954621272403,
+      "loss": 2.8339,
+      "step": 4784
+    },
+    {
+      "epoch": 0.61248,
+      "grad_norm": 0.41158801317214966,
+      "learning_rate": 0.00034474246946473286,
+      "loss": 2.7977,
+      "step": 4785
+    },
+    {
+      "epoch": 0.612608,
+      "grad_norm": 0.4405413866043091,
+      "learning_rate": 0.0003445454194073111,
+      "loss": 2.9297,
+      "step": 4786
+    },
+    {
+      "epoch": 0.612736,
+      "grad_norm": 0.44219568371772766,
+      "learning_rate": 0.00034434839607433394,
+      "loss": 2.9303,
+      "step": 4787
+    },
+    {
+      "epoch": 0.612864,
+      "grad_norm": 0.39459237456321716,
+      "learning_rate": 0.000344151399499672,
+      "loss": 2.9672,
+      "step": 4788
+    },
+    {
+      "epoch": 0.612992,
+      "grad_norm": 0.5903765559196472,
+      "learning_rate": 0.00034395442971719124,
+      "loss": 2.9634,
+      "step": 4789
+    },
+    {
+      "epoch": 0.61312,
+      "grad_norm": 0.5256094932556152,
+      "learning_rate": 0.0003437574867607529,
+      "loss": 2.8447,
+      "step": 4790
+    },
+    {
+      "epoch": 0.613248,
+      "grad_norm": 0.41846731305122375,
+      "learning_rate": 0.0003435605706642141,
+      "loss": 2.8101,
+      "step": 4791
+    },
+    {
+      "epoch": 0.613376,
+      "grad_norm": 0.389496386051178,
+      "learning_rate": 0.00034336368146142647,
+      "loss": 2.8553,
+      "step": 4792
+    },
+    {
+      "epoch": 0.613504,
+      "grad_norm": 0.4603519141674042,
+      "learning_rate": 0.0003431668191862382,
+      "loss": 2.7723,
+      "step": 4793
+    },
+    {
+      "epoch": 0.613632,
+      "grad_norm": 0.38991275429725647,
+      "learning_rate": 0.0003429699838724917,
+      "loss": 2.7879,
+      "step": 4794
+    },
+    {
+      "epoch": 0.61376,
+      "grad_norm": 0.41206449270248413,
+      "learning_rate": 0.0003427731755540256,
+      "loss": 2.8517,
+      "step": 4795
+    },
+    {
+      "epoch": 0.613888,
+      "grad_norm": 0.4376109838485718,
+      "learning_rate": 0.0003425763942646733,
+      "loss": 2.7628,
+      "step": 4796
+    },
+    {
+      "epoch": 0.614016,
+      "grad_norm": 0.42930614948272705,
+      "learning_rate": 0.00034237964003826373,
+      "loss": 2.8581,
+      "step": 4797
+    },
+    {
+      "epoch": 0.614144,
+      "grad_norm": 0.40500232577323914,
+      "learning_rate": 0.0003421829129086215,
+      "loss": 2.8706,
+      "step": 4798
+    },
+    {
+      "epoch": 0.614272,
+      "grad_norm": 0.411285936832428,
+      "learning_rate": 0.000341986212909566,
+      "loss": 2.8826,
+      "step": 4799
+    },
+    {
+      "epoch": 0.6144,
+      "grad_norm": 0.4203937351703644,
+      "learning_rate": 0.00034178954007491237,
+      "loss": 2.9,
+      "step": 4800
+    },
+    {
+      "epoch": 0.614528,
+      "grad_norm": 0.41201654076576233,
+      "learning_rate": 0.00034159289443847076,
+      "loss": 2.8801,
+      "step": 4801
+    },
+    {
+      "epoch": 0.614656,
+      "grad_norm": 0.40868106484413147,
+      "learning_rate": 0.00034139627603404723,
+      "loss": 3.0127,
+      "step": 4802
+    },
+    {
+      "epoch": 0.614784,
+      "grad_norm": 0.41430312395095825,
+      "learning_rate": 0.0003411996848954425,
+      "loss": 2.8558,
+      "step": 4803
+    },
+    {
+      "epoch": 0.614912,
+      "grad_norm": 0.4074108302593231,
+      "learning_rate": 0.0003410031210564528,
+      "loss": 2.8458,
+      "step": 4804
+    },
+    {
+      "epoch": 0.61504,
+      "grad_norm": 0.4064929187297821,
+      "learning_rate": 0.00034080658455087,
+      "loss": 2.9113,
+      "step": 4805
+    },
+    {
+      "epoch": 0.615168,
+      "grad_norm": 0.4014107286930084,
+      "learning_rate": 0.0003406100754124809,
+      "loss": 2.7919,
+      "step": 4806
+    },
+    {
+      "epoch": 0.615296,
+      "grad_norm": 0.3879109025001526,
+      "learning_rate": 0.0003404135936750674,
+      "loss": 2.8162,
+      "step": 4807
+    },
+    {
+      "epoch": 0.615424,
+      "grad_norm": 0.4039013087749481,
+      "learning_rate": 0.00034021713937240747,
+      "loss": 2.9329,
+      "step": 4808
+    },
+    {
+      "epoch": 0.615552,
+      "grad_norm": 0.41827327013015747,
+      "learning_rate": 0.00034002071253827357,
+      "loss": 2.785,
+      "step": 4809
+    },
+    {
+      "epoch": 0.61568,
+      "grad_norm": 0.40418604016304016,
+      "learning_rate": 0.00033982431320643357,
+      "loss": 2.9511,
+      "step": 4810
+    },
+    {
+      "epoch": 0.615808,
+      "grad_norm": 0.4373067319393158,
+      "learning_rate": 0.0003396279414106513,
+      "loss": 2.9369,
+      "step": 4811
+    },
+    {
+      "epoch": 0.615936,
+      "grad_norm": 0.3660423457622528,
+      "learning_rate": 0.000339431597184685,
+      "loss": 2.7305,
+      "step": 4812
+    },
+    {
+      "epoch": 0.616064,
+      "grad_norm": 0.5329440236091614,
+      "learning_rate": 0.00033923528056228875,
+      "loss": 2.8823,
+      "step": 4813
+    },
+    {
+      "epoch": 0.616192,
+      "grad_norm": 0.39465364813804626,
+      "learning_rate": 0.0003390389915772115,
+      "loss": 2.7322,
+      "step": 4814
+    },
+    {
+      "epoch": 0.61632,
+      "grad_norm": 0.4127886891365051,
+      "learning_rate": 0.00033884273026319747,
+      "loss": 2.6633,
+      "step": 4815
+    },
+    {
+      "epoch": 0.616448,
+      "grad_norm": 0.3869602382183075,
+      "learning_rate": 0.0003386464966539865,
+      "loss": 2.9617,
+      "step": 4816
+    },
+    {
+      "epoch": 0.616576,
+      "grad_norm": 0.40756940841674805,
+      "learning_rate": 0.0003384502907833134,
+      "loss": 2.9013,
+      "step": 4817
+    },
+    {
+      "epoch": 0.616704,
+      "grad_norm": 0.4077656865119934,
+      "learning_rate": 0.000338254112684908,
+      "loss": 2.8073,
+      "step": 4818
+    },
+    {
+      "epoch": 0.616832,
+      "grad_norm": 0.45330336689949036,
+      "learning_rate": 0.0003380579623924959,
+      "loss": 2.7798,
+      "step": 4819
+    },
+    {
+      "epoch": 0.61696,
+      "grad_norm": 0.4053998291492462,
+      "learning_rate": 0.0003378618399397972,
+      "loss": 2.7963,
+      "step": 4820
+    },
+    {
+      "epoch": 0.617088,
+      "grad_norm": 0.4133601486682892,
+      "learning_rate": 0.0003376657453605281,
+      "loss": 2.8895,
+      "step": 4821
+    },
+    {
+      "epoch": 0.617216,
+      "grad_norm": 0.41690996289253235,
+      "learning_rate": 0.00033746967868839917,
+      "loss": 2.8991,
+      "step": 4822
+    },
+    {
+      "epoch": 0.617344,
+      "grad_norm": 0.3864802420139313,
+      "learning_rate": 0.0003372736399571169,
+      "loss": 2.7663,
+      "step": 4823
+    },
+    {
+      "epoch": 0.617472,
+      "grad_norm": 0.4080016314983368,
+      "learning_rate": 0.0003370776292003823,
+      "loss": 2.9215,
+      "step": 4824
+    },
+    {
+      "epoch": 0.6176,
+      "grad_norm": 0.4392680823802948,
+      "learning_rate": 0.00033688164645189194,
+      "loss": 2.8216,
+      "step": 4825
+    },
+    {
+      "epoch": 0.617728,
+      "grad_norm": 0.43851685523986816,
+      "learning_rate": 0.0003366856917453377,
+      "loss": 2.9605,
+      "step": 4826
+    },
+    {
+      "epoch": 0.617856,
+      "grad_norm": 0.42302873730659485,
+      "learning_rate": 0.00033648976511440633,
+      "loss": 2.8606,
+      "step": 4827
+    },
+    {
+      "epoch": 0.617984,
+      "grad_norm": 0.42401960492134094,
+      "learning_rate": 0.00033629386659277984,
+      "loss": 2.6965,
+      "step": 4828
+    },
+    {
+      "epoch": 0.618112,
+      "grad_norm": 0.42720678448677063,
+      "learning_rate": 0.0003360979962141355,
+      "loss": 2.8449,
+      "step": 4829
+    },
+    {
+      "epoch": 0.61824,
+      "grad_norm": 0.4081781506538391,
+      "learning_rate": 0.0003359021540121457,
+      "loss": 2.8612,
+      "step": 4830
+    },
+    {
+      "epoch": 0.618368,
+      "grad_norm": 0.4336670935153961,
+      "learning_rate": 0.0003357063400204781,
+      "loss": 2.7677,
+      "step": 4831
+    },
+    {
+      "epoch": 0.618496,
+      "grad_norm": 0.3987661302089691,
+      "learning_rate": 0.0003355105542727954,
+      "loss": 2.9225,
+      "step": 4832
+    },
+    {
+      "epoch": 0.618624,
+      "grad_norm": 0.4044264554977417,
+      "learning_rate": 0.00033531479680275515,
+      "loss": 2.7481,
+      "step": 4833
+    },
+    {
+      "epoch": 0.618752,
+      "grad_norm": 0.3959006667137146,
+      "learning_rate": 0.00033511906764401066,
+      "loss": 2.8649,
+      "step": 4834
+    },
+    {
+      "epoch": 0.61888,
+      "grad_norm": 0.38154780864715576,
+      "learning_rate": 0.00033492336683020997,
+      "loss": 2.9044,
+      "step": 4835
+    },
+    {
+      "epoch": 0.619008,
+      "grad_norm": 0.4013727009296417,
+      "learning_rate": 0.0003347276943949961,
+      "loss": 2.7752,
+      "step": 4836
+    },
+    {
+      "epoch": 0.619136,
+      "grad_norm": 0.3768063485622406,
+      "learning_rate": 0.00033453205037200763,
+      "loss": 2.7732,
+      "step": 4837
+    },
+    {
+      "epoch": 0.619264,
+      "grad_norm": 0.4136853516101837,
+      "learning_rate": 0.00033433643479487775,
+      "loss": 2.9868,
+      "step": 4838
+    },
+    {
+      "epoch": 0.619392,
+      "grad_norm": 0.41531598567962646,
+      "learning_rate": 0.0003341408476972355,
+      "loss": 2.7776,
+      "step": 4839
+    },
+    {
+      "epoch": 0.61952,
+      "grad_norm": 0.4108033776283264,
+      "learning_rate": 0.00033394528911270425,
+      "loss": 2.7248,
+      "step": 4840
+    },
+    {
+      "epoch": 0.619648,
+      "grad_norm": 0.3838825523853302,
+      "learning_rate": 0.000333749759074903,
+      "loss": 2.786,
+      "step": 4841
+    },
+    {
+      "epoch": 0.619776,
+      "grad_norm": 0.3838747441768646,
+      "learning_rate": 0.0003335542576174455,
+      "loss": 2.842,
+      "step": 4842
+    },
+    {
+      "epoch": 0.619904,
+      "grad_norm": 0.3877181112766266,
+      "learning_rate": 0.00033335878477394056,
+      "loss": 2.7937,
+      "step": 4843
+    },
+    {
+      "epoch": 0.620032,
+      "grad_norm": 0.36422616243362427,
+      "learning_rate": 0.0003331633405779926,
+      "loss": 2.7823,
+      "step": 4844
+    },
+    {
+      "epoch": 0.62016,
+      "grad_norm": 0.38175421953201294,
+      "learning_rate": 0.0003329679250632004,
+      "loss": 2.7268,
+      "step": 4845
+    },
+    {
+      "epoch": 0.620288,
+      "grad_norm": 0.42853593826293945,
+      "learning_rate": 0.0003327725382631582,
+      "loss": 2.7914,
+      "step": 4846
+    },
+    {
+      "epoch": 0.620416,
+      "grad_norm": 0.4110375940799713,
+      "learning_rate": 0.00033257718021145543,
+      "loss": 2.8841,
+      "step": 4847
+    },
+    {
+      "epoch": 0.620544,
+      "grad_norm": 0.42038053274154663,
+      "learning_rate": 0.00033238185094167625,
+      "loss": 2.9142,
+      "step": 4848
+    },
+    {
+      "epoch": 0.620672,
+      "grad_norm": 0.43172115087509155,
+      "learning_rate": 0.00033218655048740023,
+      "loss": 2.8227,
+      "step": 4849
+    },
+    {
+      "epoch": 0.6208,
+      "grad_norm": 0.493123859167099,
+      "learning_rate": 0.0003319912788822016,
+      "loss": 2.7933,
+      "step": 4850
+    },
+    {
+      "epoch": 0.620928,
+      "grad_norm": 0.405548095703125,
+      "learning_rate": 0.00033179603615964974,
+      "loss": 2.8733,
+      "step": 4851
+    },
+    {
+      "epoch": 0.621056,
+      "grad_norm": 0.4167015850543976,
+      "learning_rate": 0.00033160082235330935,
+      "loss": 2.9224,
+      "step": 4852
+    },
+    {
+      "epoch": 0.621184,
+      "grad_norm": 0.4269377291202545,
+      "learning_rate": 0.0003314056374967399,
+      "loss": 2.8468,
+      "step": 4853
+    },
+    {
+      "epoch": 0.621312,
+      "grad_norm": 0.43306922912597656,
+      "learning_rate": 0.0003312104816234958,
+      "loss": 2.8882,
+      "step": 4854
+    },
+    {
+      "epoch": 0.62144,
+      "grad_norm": 0.4303545355796814,
+      "learning_rate": 0.00033101535476712674,
+      "loss": 2.8196,
+      "step": 4855
+    },
+    {
+      "epoch": 0.621568,
+      "grad_norm": 0.40924471616744995,
+      "learning_rate": 0.0003308202569611772,
+      "loss": 2.6337,
+      "step": 4856
+    },
+    {
+      "epoch": 0.621696,
+      "grad_norm": 0.41062062978744507,
+      "learning_rate": 0.00033062518823918663,
+      "loss": 2.8261,
+      "step": 4857
+    },
+    {
+      "epoch": 0.621824,
+      "grad_norm": 0.4039859473705292,
+      "learning_rate": 0.00033043014863468983,
+      "loss": 2.8358,
+      "step": 4858
+    },
+    {
+      "epoch": 0.621952,
+      "grad_norm": 0.43354830145835876,
+      "learning_rate": 0.0003302351381812164,
+      "loss": 2.8283,
+      "step": 4859
+    },
+    {
+      "epoch": 0.62208,
+      "grad_norm": 0.40152275562286377,
+      "learning_rate": 0.0003300401569122908,
+      "loss": 2.7525,
+      "step": 4860
+    },
+    {
+      "epoch": 0.622208,
+      "grad_norm": 0.40010184049606323,
+      "learning_rate": 0.00032984520486143256,
+      "loss": 2.8369,
+      "step": 4861
+    },
+    {
+      "epoch": 0.622336,
+      "grad_norm": 0.4062066972255707,
+      "learning_rate": 0.00032965028206215625,
+      "loss": 2.7926,
+      "step": 4862
+    },
+    {
+      "epoch": 0.622464,
+      "grad_norm": 0.42496970295906067,
+      "learning_rate": 0.0003294553885479713,
+      "loss": 2.9845,
+      "step": 4863
+    },
+    {
+      "epoch": 0.622592,
+      "grad_norm": 0.39128148555755615,
+      "learning_rate": 0.0003292605243523821,
+      "loss": 2.7391,
+      "step": 4864
+    },
+    {
+      "epoch": 0.62272,
+      "grad_norm": 0.4271163046360016,
+      "learning_rate": 0.00032906568950888827,
+      "loss": 2.7699,
+      "step": 4865
+    },
+    {
+      "epoch": 0.622848,
+      "grad_norm": 0.3885034918785095,
+      "learning_rate": 0.00032887088405098377,
+      "loss": 2.8634,
+      "step": 4866
+    },
+    {
+      "epoch": 0.622976,
+      "grad_norm": 0.4100918471813202,
+      "learning_rate": 0.0003286761080121585,
+      "loss": 2.7543,
+      "step": 4867
+    },
+    {
+      "epoch": 0.623104,
+      "grad_norm": 0.43279752135276794,
+      "learning_rate": 0.00032848136142589636,
+      "loss": 2.9199,
+      "step": 4868
+    },
+    {
+      "epoch": 0.623232,
+      "grad_norm": 0.4271816313266754,
+      "learning_rate": 0.00032828664432567653,
+      "loss": 2.8424,
+      "step": 4869
+    },
+    {
+      "epoch": 0.62336,
+      "grad_norm": 0.41509148478507996,
+      "learning_rate": 0.00032809195674497326,
+      "loss": 2.9035,
+      "step": 4870
+    },
+    {
+      "epoch": 0.623488,
+      "grad_norm": 0.4209059476852417,
+      "learning_rate": 0.0003278972987172556,
+      "loss": 2.8569,
+      "step": 4871
+    },
+    {
+      "epoch": 0.623616,
+      "grad_norm": 0.38560518622398376,
+      "learning_rate": 0.00032770267027598736,
+      "loss": 2.8025,
+      "step": 4872
+    },
+    {
+      "epoch": 0.623744,
+      "grad_norm": 0.3923930525779724,
+      "learning_rate": 0.0003275080714546277,
+      "loss": 2.8534,
+      "step": 4873
+    },
+    {
+      "epoch": 0.623872,
+      "grad_norm": 0.4211357533931732,
+      "learning_rate": 0.00032731350228663023,
+      "loss": 2.9531,
+      "step": 4874
+    },
+    {
+      "epoch": 0.624,
+      "grad_norm": 0.4002450406551361,
+      "learning_rate": 0.0003271189628054434,
+      "loss": 2.8617,
+      "step": 4875
+    },
+    {
+      "epoch": 0.624128,
+      "grad_norm": 0.4205063283443451,
+      "learning_rate": 0.0003269244530445112,
+      "loss": 2.8338,
+      "step": 4876
+    },
+    {
+      "epoch": 0.624256,
+      "grad_norm": 0.40131354331970215,
+      "learning_rate": 0.00032672997303727217,
+      "loss": 2.9065,
+      "step": 4877
+    },
+    {
+      "epoch": 0.624384,
+      "grad_norm": 0.37917354702949524,
+      "learning_rate": 0.0003265355228171594,
+      "loss": 2.812,
+      "step": 4878
+    },
+    {
+      "epoch": 0.624512,
+      "grad_norm": 0.4086361229419708,
+      "learning_rate": 0.00032634110241760116,
+      "loss": 2.8554,
+      "step": 4879
+    },
+    {
+      "epoch": 0.62464,
+      "grad_norm": 0.4234074652194977,
+      "learning_rate": 0.0003261467118720207,
+      "loss": 2.929,
+      "step": 4880
+    },
+    {
+      "epoch": 0.624768,
+      "grad_norm": 0.38650214672088623,
+      "learning_rate": 0.000325952351213836,
+      "loss": 2.9531,
+      "step": 4881
+    },
+    {
+      "epoch": 0.624896,
+      "grad_norm": 0.3920231759548187,
+      "learning_rate": 0.00032575802047645977,
+      "loss": 2.8896,
+      "step": 4882
+    },
+    {
+      "epoch": 0.625024,
+      "grad_norm": 0.41024333238601685,
+      "learning_rate": 0.0003255637196932999,
+      "loss": 2.9094,
+      "step": 4883
+    },
+    {
+      "epoch": 0.625152,
+      "grad_norm": 0.4149235785007477,
+      "learning_rate": 0.00032536944889775864,
+      "loss": 2.8778,
+      "step": 4884
+    },
+    {
+      "epoch": 0.62528,
+      "grad_norm": 0.3781399130821228,
+      "learning_rate": 0.0003251752081232338,
+      "loss": 2.9929,
+      "step": 4885
+    },
+    {
+      "epoch": 0.625408,
+      "grad_norm": 0.3891620934009552,
+      "learning_rate": 0.0003249809974031175,
+      "loss": 2.8959,
+      "step": 4886
+    },
+    {
+      "epoch": 0.625536,
+      "grad_norm": 0.40788954496383667,
+      "learning_rate": 0.00032478681677079654,
+      "loss": 2.7477,
+      "step": 4887
+    },
+    {
+      "epoch": 0.625664,
+      "grad_norm": 0.3857533037662506,
+      "learning_rate": 0.0003245926662596531,
+      "loss": 2.8673,
+      "step": 4888
+    },
+    {
+      "epoch": 0.625792,
+      "grad_norm": 0.41186803579330444,
+      "learning_rate": 0.00032439854590306396,
+      "loss": 2.855,
+      "step": 4889
+    },
+    {
+      "epoch": 0.62592,
+      "grad_norm": 0.37686923146247864,
+      "learning_rate": 0.0003242044557344003,
+      "loss": 2.8232,
+      "step": 4890
+    },
+    {
+      "epoch": 0.626048,
+      "grad_norm": 0.42702004313468933,
+      "learning_rate": 0.00032401039578702887,
+      "loss": 2.8735,
+      "step": 4891
+    },
+    {
+      "epoch": 0.626176,
+      "grad_norm": 0.3748891055583954,
+      "learning_rate": 0.0003238163660943106,
+      "loss": 2.7914,
+      "step": 4892
+    },
+    {
+      "epoch": 0.626304,
+      "grad_norm": 0.40607234835624695,
+      "learning_rate": 0.00032362236668960145,
+      "loss": 2.8733,
+      "step": 4893
+    },
+    {
+      "epoch": 0.626432,
+      "grad_norm": 0.4263097941875458,
+      "learning_rate": 0.0003234283976062521,
+      "loss": 2.961,
+      "step": 4894
+    },
+    {
+      "epoch": 0.62656,
+      "grad_norm": 0.4095201790332794,
+      "learning_rate": 0.0003232344588776083,
+      "loss": 2.7962,
+      "step": 4895
+    },
+    {
+      "epoch": 0.626688,
+      "grad_norm": 0.3886289894580841,
+      "learning_rate": 0.00032304055053701037,
+      "loss": 2.7875,
+      "step": 4896
+    },
+    {
+      "epoch": 0.626816,
+      "grad_norm": 0.42098385095596313,
+      "learning_rate": 0.0003228466726177932,
+      "loss": 2.7754,
+      "step": 4897
+    },
+    {
+      "epoch": 0.626944,
+      "grad_norm": 0.40146517753601074,
+      "learning_rate": 0.0003226528251532868,
+      "loss": 2.7989,
+      "step": 4898
+    },
+    {
+      "epoch": 0.627072,
+      "grad_norm": 0.4000326991081238,
+      "learning_rate": 0.0003224590081768157,
+      "loss": 2.7853,
+      "step": 4899
+    },
+    {
+      "epoch": 0.6272,
+      "grad_norm": 0.41949349641799927,
+      "learning_rate": 0.00032226522172169923,
+      "loss": 2.972,
+      "step": 4900
+    },
+    {
+      "epoch": 0.627328,
+      "grad_norm": 0.40498483180999756,
+      "learning_rate": 0.0003220714658212518,
+      "loss": 2.936,
+      "step": 4901
+    },
+    {
+      "epoch": 0.627456,
+      "grad_norm": 0.4065920114517212,
+      "learning_rate": 0.0003218777405087821,
+      "loss": 2.8913,
+      "step": 4902
+    },
+    {
+      "epoch": 0.627584,
+      "grad_norm": 0.39954817295074463,
+      "learning_rate": 0.00032168404581759357,
+      "loss": 2.9459,
+      "step": 4903
+    },
+    {
+      "epoch": 0.627712,
+      "grad_norm": 0.4022829532623291,
+      "learning_rate": 0.000321490381780985,
+      "loss": 2.7359,
+      "step": 4904
+    },
+    {
+      "epoch": 0.62784,
+      "grad_norm": 0.44910478591918945,
+      "learning_rate": 0.0003212967484322491,
+      "loss": 2.7104,
+      "step": 4905
+    },
+    {
+      "epoch": 0.627968,
+      "grad_norm": 0.4468386769294739,
+      "learning_rate": 0.00032110314580467404,
+      "loss": 2.9278,
+      "step": 4906
+    },
+    {
+      "epoch": 0.628096,
+      "grad_norm": 0.419699490070343,
+      "learning_rate": 0.0003209095739315421,
+      "loss": 2.8453,
+      "step": 4907
+    },
+    {
+      "epoch": 0.628224,
+      "grad_norm": 0.4107443392276764,
+      "learning_rate": 0.0003207160328461305,
+      "loss": 2.8567,
+      "step": 4908
+    },
+    {
+      "epoch": 0.628352,
+      "grad_norm": 0.390717476606369,
+      "learning_rate": 0.00032052252258171136,
+      "loss": 2.8445,
+      "step": 4909
+    },
+    {
+      "epoch": 0.62848,
+      "grad_norm": 0.424385666847229,
+      "learning_rate": 0.0003203290431715513,
+      "loss": 2.8905,
+      "step": 4910
+    },
+    {
+      "epoch": 0.628608,
+      "grad_norm": 0.38418474793434143,
+      "learning_rate": 0.00032013559464891164,
+      "loss": 2.7626,
+      "step": 4911
+    },
+    {
+      "epoch": 0.628736,
+      "grad_norm": 0.41109928488731384,
+      "learning_rate": 0.0003199421770470482,
+      "loss": 2.9258,
+      "step": 4912
+    },
+    {
+      "epoch": 0.628864,
+      "grad_norm": 0.43430301547050476,
+      "learning_rate": 0.00031974879039921213,
+      "loss": 2.8016,
+      "step": 4913
+    },
+    {
+      "epoch": 0.628992,
+      "grad_norm": 0.38910770416259766,
+      "learning_rate": 0.0003195554347386487,
+      "loss": 2.7074,
+      "step": 4914
+    },
+    {
+      "epoch": 0.62912,
+      "grad_norm": 0.3917587697505951,
+      "learning_rate": 0.00031936211009859784,
+      "loss": 2.8255,
+      "step": 4915
+    },
+    {
+      "epoch": 0.629248,
+      "grad_norm": 0.4068356156349182,
+      "learning_rate": 0.00031916881651229445,
+      "loss": 2.7547,
+      "step": 4916
+    },
+    {
+      "epoch": 0.629376,
+      "grad_norm": 0.41015058755874634,
+      "learning_rate": 0.0003189755540129679,
+      "loss": 2.6982,
+      "step": 4917
+    },
+    {
+      "epoch": 0.629504,
+      "grad_norm": 0.4100431501865387,
+      "learning_rate": 0.0003187823226338421,
+      "loss": 2.6502,
+      "step": 4918
+    },
+    {
+      "epoch": 0.629632,
+      "grad_norm": 0.43049007654190063,
+      "learning_rate": 0.000318589122408136,
+      "loss": 2.8648,
+      "step": 4919
+    },
+    {
+      "epoch": 0.62976,
+      "grad_norm": 0.4306660592556,
+      "learning_rate": 0.0003183959533690629,
+      "loss": 2.8527,
+      "step": 4920
+    },
+    {
+      "epoch": 0.629888,
+      "grad_norm": 0.4012849032878876,
+      "learning_rate": 0.00031820281554983054,
+      "loss": 2.8636,
+      "step": 4921
+    },
+    {
+      "epoch": 0.630016,
+      "grad_norm": 0.4075110852718353,
+      "learning_rate": 0.0003180097089836419,
+      "loss": 2.8251,
+      "step": 4922
+    },
+    {
+      "epoch": 0.630144,
+      "grad_norm": 0.41685524582862854,
+      "learning_rate": 0.0003178166337036941,
+      "loss": 2.9571,
+      "step": 4923
+    },
+    {
+      "epoch": 0.630272,
+      "grad_norm": 0.3739596903324127,
+      "learning_rate": 0.00031762358974317905,
+      "loss": 2.6814,
+      "step": 4924
+    },
+    {
+      "epoch": 0.6304,
+      "grad_norm": 0.42140382528305054,
+      "learning_rate": 0.00031743057713528325,
+      "loss": 2.8879,
+      "step": 4925
+    },
+    {
+      "epoch": 0.630528,
+      "grad_norm": 0.4748397171497345,
+      "learning_rate": 0.00031723759591318755,
+      "loss": 2.7392,
+      "step": 4926
+    },
+    {
+      "epoch": 0.630656,
+      "grad_norm": 0.40635794401168823,
+      "learning_rate": 0.0003170446461100681,
+      "loss": 2.9084,
+      "step": 4927
+    },
+    {
+      "epoch": 0.630784,
+      "grad_norm": 0.4129738509654999,
+      "learning_rate": 0.000316851727759095,
+      "loss": 2.752,
+      "step": 4928
+    },
+    {
+      "epoch": 0.630912,
+      "grad_norm": 0.3735859990119934,
+      "learning_rate": 0.00031665884089343295,
+      "loss": 2.7671,
+      "step": 4929
+    },
+    {
+      "epoch": 0.63104,
+      "grad_norm": 0.3965545892715454,
+      "learning_rate": 0.0003164659855462417,
+      "loss": 2.7826,
+      "step": 4930
+    },
+    {
+      "epoch": 0.631168,
+      "grad_norm": 0.42073702812194824,
+      "learning_rate": 0.00031627316175067514,
+      "loss": 3.0024,
+      "step": 4931
+    },
+    {
+      "epoch": 0.631296,
+      "grad_norm": 0.3790392577648163,
+      "learning_rate": 0.0003160803695398822,
+      "loss": 2.8329,
+      "step": 4932
+    },
+    {
+      "epoch": 0.631424,
+      "grad_norm": 0.40170353651046753,
+      "learning_rate": 0.00031588760894700577,
+      "loss": 2.855,
+      "step": 4933
+    },
+    {
+      "epoch": 0.631552,
+      "grad_norm": 0.42236393690109253,
+      "learning_rate": 0.0003156948800051839,
+      "loss": 2.879,
+      "step": 4934
+    },
+    {
+      "epoch": 0.63168,
+      "grad_norm": 0.41063523292541504,
+      "learning_rate": 0.0003155021827475488,
+      "loss": 2.7417,
+      "step": 4935
+    },
+    {
+      "epoch": 0.631808,
+      "grad_norm": 0.39441779255867004,
+      "learning_rate": 0.0003153095172072273,
+      "loss": 2.8901,
+      "step": 4936
+    },
+    {
+      "epoch": 0.631936,
+      "grad_norm": 0.3967224955558777,
+      "learning_rate": 0.00031511688341734096,
+      "loss": 2.7563,
+      "step": 4937
+    },
+    {
+      "epoch": 0.632064,
+      "grad_norm": 0.3905535936355591,
+      "learning_rate": 0.0003149242814110057,
+      "loss": 2.9072,
+      "step": 4938
+    },
+    {
+      "epoch": 0.632192,
+      "grad_norm": 0.3849788308143616,
+      "learning_rate": 0.00031473171122133205,
+      "loss": 2.8122,
+      "step": 4939
+    },
+    {
+      "epoch": 0.63232,
+      "grad_norm": 0.44086986780166626,
+      "learning_rate": 0.00031453917288142496,
+      "loss": 2.85,
+      "step": 4940
+    },
+    {
+      "epoch": 0.632448,
+      "grad_norm": 0.39657652378082275,
+      "learning_rate": 0.0003143466664243841,
+      "loss": 2.9056,
+      "step": 4941
+    },
+    {
+      "epoch": 0.632576,
+      "grad_norm": 0.39416131377220154,
+      "learning_rate": 0.0003141541918833037,
+      "loss": 2.8727,
+      "step": 4942
+    },
+    {
+      "epoch": 0.632704,
+      "grad_norm": 0.38360363245010376,
+      "learning_rate": 0.0003139617492912722,
+      "loss": 2.7505,
+      "step": 4943
+    },
+    {
+      "epoch": 0.632832,
+      "grad_norm": 0.4155833125114441,
+      "learning_rate": 0.00031376933868137264,
+      "loss": 2.7091,
+      "step": 4944
+    },
+    {
+      "epoch": 0.63296,
+      "grad_norm": 0.42820224165916443,
+      "learning_rate": 0.0003135769600866828,
+      "loss": 2.7779,
+      "step": 4945
+    },
+    {
+      "epoch": 0.633088,
+      "grad_norm": 0.40411654114723206,
+      "learning_rate": 0.0003133846135402748,
+      "loss": 2.7358,
+      "step": 4946
+    },
+    {
+      "epoch": 0.633216,
+      "grad_norm": 0.4116949141025543,
+      "learning_rate": 0.000313192299075215,
+      "loss": 2.8534,
+      "step": 4947
+    },
+    {
+      "epoch": 0.633344,
+      "grad_norm": 0.41559672355651855,
+      "learning_rate": 0.0003130000167245648,
+      "loss": 2.7497,
+      "step": 4948
+    },
+    {
+      "epoch": 0.633472,
+      "grad_norm": 0.4499170482158661,
+      "learning_rate": 0.00031280776652137937,
+      "loss": 2.8128,
+      "step": 4949
+    },
+    {
+      "epoch": 0.6336,
+      "grad_norm": 0.43346279859542847,
+      "learning_rate": 0.0003126155484987092,
+      "loss": 2.9048,
+      "step": 4950
+    },
+    {
+      "epoch": 0.633728,
+      "grad_norm": 0.39782360196113586,
+      "learning_rate": 0.0003124233626895985,
+      "loss": 2.8509,
+      "step": 4951
+    },
+    {
+      "epoch": 0.633856,
+      "grad_norm": 0.43292859196662903,
+      "learning_rate": 0.00031223120912708647,
+      "loss": 2.7827,
+      "step": 4952
+    },
+    {
+      "epoch": 0.633984,
+      "grad_norm": 0.4136907458305359,
+      "learning_rate": 0.0003120390878442063,
+      "loss": 2.7727,
+      "step": 4953
+    },
+    {
+      "epoch": 0.634112,
+      "grad_norm": 0.4923643171787262,
+      "learning_rate": 0.0003118469988739859,
+      "loss": 2.7752,
+      "step": 4954
+    },
+    {
+      "epoch": 0.63424,
+      "grad_norm": 0.40492069721221924,
+      "learning_rate": 0.0003116549422494478,
+      "loss": 2.8279,
+      "step": 4955
+    },
+    {
+      "epoch": 0.634368,
+      "grad_norm": 0.4053919017314911,
+      "learning_rate": 0.0003114629180036086,
+      "loss": 2.7181,
+      "step": 4956
+    },
+    {
+      "epoch": 0.634496,
+      "grad_norm": 0.4197677969932556,
+      "learning_rate": 0.0003112709261694793,
+      "loss": 2.8648,
+      "step": 4957
+    },
+    {
+      "epoch": 0.634624,
+      "grad_norm": 0.3886752128601074,
+      "learning_rate": 0.00031107896678006573,
+      "loss": 2.8223,
+      "step": 4958
+    },
+    {
+      "epoch": 0.634752,
+      "grad_norm": 0.4148769676685333,
+      "learning_rate": 0.0003108870398683681,
+      "loss": 2.9137,
+      "step": 4959
+    },
+    {
+      "epoch": 0.63488,
+      "grad_norm": 0.38311588764190674,
+      "learning_rate": 0.00031069514546738076,
+      "loss": 2.7858,
+      "step": 4960
+    },
+    {
+      "epoch": 0.635008,
+      "grad_norm": 0.39475366473197937,
+      "learning_rate": 0.00031050328361009245,
+      "loss": 2.8523,
+      "step": 4961
+    },
+    {
+      "epoch": 0.635136,
+      "grad_norm": 0.4232536256313324,
+      "learning_rate": 0.00031031145432948643,
+      "loss": 2.8635,
+      "step": 4962
+    },
+    {
+      "epoch": 0.635264,
+      "grad_norm": 0.42048895359039307,
+      "learning_rate": 0.00031011965765854056,
+      "loss": 2.7158,
+      "step": 4963
+    },
+    {
+      "epoch": 0.635392,
+      "grad_norm": 0.4051856994628906,
+      "learning_rate": 0.0003099278936302269,
+      "loss": 2.6807,
+      "step": 4964
+    },
+    {
+      "epoch": 0.63552,
+      "grad_norm": 0.41608893871307373,
+      "learning_rate": 0.00030973616227751165,
+      "loss": 2.9051,
+      "step": 4965
+    },
+    {
+      "epoch": 0.635648,
+      "grad_norm": 0.4299136996269226,
+      "learning_rate": 0.0003095444636333559,
+      "loss": 2.7791,
+      "step": 4966
+    },
+    {
+      "epoch": 0.635776,
+      "grad_norm": 0.4220398962497711,
+      "learning_rate": 0.0003093527977307149,
+      "loss": 2.7436,
+      "step": 4967
+    },
+    {
+      "epoch": 0.635904,
+      "grad_norm": 0.394222617149353,
+      "learning_rate": 0.00030916116460253783,
+      "loss": 2.8986,
+      "step": 4968
+    },
+    {
+      "epoch": 0.636032,
+      "grad_norm": 0.38906562328338623,
+      "learning_rate": 0.0003089695642817691,
+      "loss": 2.9173,
+      "step": 4969
+    },
+    {
+      "epoch": 0.63616,
+      "grad_norm": 0.4115164875984192,
+      "learning_rate": 0.0003087779968013471,
+      "loss": 2.801,
+      "step": 4970
+    },
+    {
+      "epoch": 0.636288,
+      "grad_norm": 0.4150785505771637,
+      "learning_rate": 0.00030858646219420415,
+      "loss": 2.8031,
+      "step": 4971
+    },
+    {
+      "epoch": 0.636416,
+      "grad_norm": 0.3877466917037964,
+      "learning_rate": 0.0003083949604932674,
+      "loss": 2.8937,
+      "step": 4972
+    },
+    {
+      "epoch": 0.636544,
+      "grad_norm": 0.3898378014564514,
+      "learning_rate": 0.00030820349173145824,
+      "loss": 2.7325,
+      "step": 4973
+    },
+    {
+      "epoch": 0.636672,
+      "grad_norm": 0.39726293087005615,
+      "learning_rate": 0.0003080120559416924,
+      "loss": 2.8767,
+      "step": 4974
+    },
+    {
+      "epoch": 0.6368,
+      "grad_norm": 0.47274133563041687,
+      "learning_rate": 0.00030782065315687975,
+      "loss": 2.7902,
+      "step": 4975
+    },
+    {
+      "epoch": 0.636928,
+      "grad_norm": 0.4185158312320709,
+      "learning_rate": 0.0003076292834099248,
+      "loss": 2.8281,
+      "step": 4976
+    },
+    {
+      "epoch": 0.637056,
+      "grad_norm": 0.4106575548648834,
+      "learning_rate": 0.00030743794673372606,
+      "loss": 2.8503,
+      "step": 4977
+    },
+    {
+      "epoch": 0.637184,
+      "grad_norm": 0.3962419927120209,
+      "learning_rate": 0.0003072466431611768,
+      "loss": 2.801,
+      "step": 4978
+    },
+    {
+      "epoch": 0.637312,
+      "grad_norm": 0.4268336296081543,
+      "learning_rate": 0.000307055372725164,
+      "loss": 2.7935,
+      "step": 4979
+    },
+    {
+      "epoch": 0.63744,
+      "grad_norm": 0.374625563621521,
+      "learning_rate": 0.0003068641354585697,
+      "loss": 2.8872,
+      "step": 4980
+    },
+    {
+      "epoch": 0.637568,
+      "grad_norm": 0.42541036009788513,
+      "learning_rate": 0.00030667293139426946,
+      "loss": 2.7645,
+      "step": 4981
+    },
+    {
+      "epoch": 0.637696,
+      "grad_norm": 0.38999226689338684,
+      "learning_rate": 0.00030648176056513355,
+      "loss": 2.8391,
+      "step": 4982
+    },
+    {
+      "epoch": 0.637824,
+      "grad_norm": 0.39741280674934387,
+      "learning_rate": 0.0003062906230040263,
+      "loss": 2.7895,
+      "step": 4983
+    },
+    {
+      "epoch": 0.637952,
+      "grad_norm": 0.4544917643070221,
+      "learning_rate": 0.0003060995187438068,
+      "loss": 2.827,
+      "step": 4984
+    },
+    {
+      "epoch": 0.63808,
+      "grad_norm": 0.37860938906669617,
+      "learning_rate": 0.00030590844781732795,
+      "loss": 2.7342,
+      "step": 4985
+    },
+    {
+      "epoch": 0.638208,
+      "grad_norm": 0.39850014448165894,
+      "learning_rate": 0.0003057174102574367,
+      "loss": 2.8597,
+      "step": 4986
+    },
+    {
+      "epoch": 0.638336,
+      "grad_norm": 0.4172910451889038,
+      "learning_rate": 0.00030552640609697514,
+      "loss": 2.8094,
+      "step": 4987
+    },
+    {
+      "epoch": 0.638464,
+      "grad_norm": 0.40661436319351196,
+      "learning_rate": 0.00030533543536877895,
+      "loss": 2.9195,
+      "step": 4988
+    },
+    {
+      "epoch": 0.638592,
+      "grad_norm": 0.38885557651519775,
+      "learning_rate": 0.00030514449810567826,
+      "loss": 2.7668,
+      "step": 4989
+    },
+    {
+      "epoch": 0.63872,
+      "grad_norm": 0.38999488949775696,
+      "learning_rate": 0.00030495359434049707,
+      "loss": 2.8256,
+      "step": 4990
+    },
+    {
+      "epoch": 0.638848,
+      "grad_norm": 0.3979494571685791,
+      "learning_rate": 0.0003047627241060544,
+      "loss": 2.7525,
+      "step": 4991
+    },
+    {
+      "epoch": 0.638976,
+      "grad_norm": 0.3798055648803711,
+      "learning_rate": 0.00030457188743516274,
+      "loss": 2.7619,
+      "step": 4992
+    },
+    {
+      "epoch": 0.639104,
+      "grad_norm": 0.4070138931274414,
+      "learning_rate": 0.00030438108436062916,
+      "loss": 2.7552,
+      "step": 4993
+    },
+    {
+      "epoch": 0.639232,
+      "grad_norm": 0.4727679491043091,
+      "learning_rate": 0.0003041903149152552,
+      "loss": 2.7801,
+      "step": 4994
+    },
+    {
+      "epoch": 0.63936,
+      "grad_norm": 0.39623865485191345,
+      "learning_rate": 0.0003039995791318358,
+      "loss": 2.7335,
+      "step": 4995
+    },
+    {
+      "epoch": 0.639488,
+      "grad_norm": 0.36961114406585693,
+      "learning_rate": 0.00030380887704316116,
+      "loss": 2.7277,
+      "step": 4996
+    },
+    {
+      "epoch": 0.639616,
+      "grad_norm": 0.4361562430858612,
+      "learning_rate": 0.00030361820868201494,
+      "loss": 2.8213,
+      "step": 4997
+    },
+    {
+      "epoch": 0.639744,
+      "grad_norm": 0.37268850207328796,
+      "learning_rate": 0.00030342757408117544,
+      "loss": 2.7407,
+      "step": 4998
+    },
+    {
+      "epoch": 0.639872,
+      "grad_norm": 0.38566792011260986,
+      "learning_rate": 0.0003032369732734148,
+      "loss": 2.756,
+      "step": 4999
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.40195512771606445,
+      "learning_rate": 0.0003030464062914995,
+      "loss": 2.7935,
+      "step": 5000
+    },
+    {
+      "epoch": 0.640128,
+      "grad_norm": 0.41509875655174255,
+      "learning_rate": 0.0003028558731681902,
+      "loss": 2.7942,
+      "step": 5001
+    },
+    {
+      "epoch": 0.640256,
+      "grad_norm": 0.39035630226135254,
+      "learning_rate": 0.0003026653739362418,
+      "loss": 2.7567,
+      "step": 5002
+    },
+    {
+      "epoch": 0.640384,
+      "grad_norm": 0.39688581228256226,
+      "learning_rate": 0.00030247490862840337,
+      "loss": 2.8692,
+      "step": 5003
+    },
+    {
+      "epoch": 0.640512,
+      "grad_norm": 0.3999447822570801,
+      "learning_rate": 0.000302284477277418,
+      "loss": 2.8005,
+      "step": 5004
+    },
+    {
+      "epoch": 0.64064,
+      "grad_norm": 0.40972116589546204,
+      "learning_rate": 0.0003020940799160229,
+      "loss": 2.7936,
+      "step": 5005
+    },
+    {
+      "epoch": 0.640768,
+      "grad_norm": 0.41044777631759644,
+      "learning_rate": 0.00030190371657695,
+      "loss": 2.8337,
+      "step": 5006
+    },
+    {
+      "epoch": 0.640896,
+      "grad_norm": 0.3848056197166443,
+      "learning_rate": 0.00030171338729292484,
+      "loss": 2.7829,
+      "step": 5007
+    },
+    {
+      "epoch": 0.641024,
+      "grad_norm": 0.42341914772987366,
+      "learning_rate": 0.000301523092096667,
+      "loss": 2.7516,
+      "step": 5008
+    },
+    {
+      "epoch": 0.641152,
+      "grad_norm": 0.416532039642334,
+      "learning_rate": 0.00030133283102089063,
+      "loss": 2.7961,
+      "step": 5009
+    },
+    {
+      "epoch": 0.64128,
+      "grad_norm": 0.43532639741897583,
+      "learning_rate": 0.00030114260409830384,
+      "loss": 2.9167,
+      "step": 5010
+    },
+    {
+      "epoch": 0.641408,
+      "grad_norm": 0.4173070192337036,
+      "learning_rate": 0.00030095241136160857,
+      "loss": 2.7831,
+      "step": 5011
+    },
+    {
+      "epoch": 0.641536,
+      "grad_norm": 0.4620778560638428,
+      "learning_rate": 0.00030076225284350144,
+      "loss": 2.7595,
+      "step": 5012
+    },
+    {
+      "epoch": 0.641664,
+      "grad_norm": 0.40984925627708435,
+      "learning_rate": 0.0003005721285766729,
+      "loss": 2.8797,
+      "step": 5013
+    },
+    {
+      "epoch": 0.641792,
+      "grad_norm": 0.46339908242225647,
+      "learning_rate": 0.00030038203859380716,
+      "loss": 2.8978,
+      "step": 5014
+    },
+    {
+      "epoch": 0.64192,
+      "grad_norm": 0.4156503975391388,
+      "learning_rate": 0.00030019198292758324,
+      "loss": 2.9516,
+      "step": 5015
+    },
+    {
+      "epoch": 0.642048,
+      "grad_norm": 0.4474954903125763,
+      "learning_rate": 0.00030000196161067394,
+      "loss": 2.8443,
+      "step": 5016
+    },
+    {
+      "epoch": 0.642176,
+      "grad_norm": 0.4051572382450104,
+      "learning_rate": 0.00029981197467574607,
+      "loss": 2.8224,
+      "step": 5017
+    },
+    {
+      "epoch": 0.642304,
+      "grad_norm": 0.4219511151313782,
+      "learning_rate": 0.0002996220221554604,
+      "loss": 2.638,
+      "step": 5018
+    },
+    {
+      "epoch": 0.642432,
+      "grad_norm": 0.3921409249305725,
+      "learning_rate": 0.00029943210408247216,
+      "loss": 2.7788,
+      "step": 5019
+    },
+    {
+      "epoch": 0.64256,
+      "grad_norm": 0.41471537947654724,
+      "learning_rate": 0.00029924222048943044,
+      "loss": 2.9005,
+      "step": 5020
+    },
+    {
+      "epoch": 0.642688,
+      "grad_norm": 0.4083428680896759,
+      "learning_rate": 0.00029905237140897845,
+      "loss": 2.7885,
+      "step": 5021
+    },
+    {
+      "epoch": 0.642816,
+      "grad_norm": 0.39140385389328003,
+      "learning_rate": 0.00029886255687375316,
+      "loss": 2.8494,
+      "step": 5022
+    },
+    {
+      "epoch": 0.642944,
+      "grad_norm": 0.4499935805797577,
+      "learning_rate": 0.0002986727769163862,
+      "loss": 2.9614,
+      "step": 5023
+    },
+    {
+      "epoch": 0.643072,
+      "grad_norm": 0.4641708433628082,
+      "learning_rate": 0.000298483031569503,
+      "loss": 2.8087,
+      "step": 5024
+    },
+    {
+      "epoch": 0.6432,
+      "grad_norm": 0.3983391225337982,
+      "learning_rate": 0.0002982933208657228,
+      "loss": 2.7835,
+      "step": 5025
+    },
+    {
+      "epoch": 0.643328,
+      "grad_norm": 0.4198138415813446,
+      "learning_rate": 0.0002981036448376591,
+      "loss": 2.7943,
+      "step": 5026
+    },
+    {
+      "epoch": 0.643456,
+      "grad_norm": 0.394598126411438,
+      "learning_rate": 0.00029791400351791955,
+      "loss": 2.8563,
+      "step": 5027
+    },
+    {
+      "epoch": 0.643584,
+      "grad_norm": 0.38938891887664795,
+      "learning_rate": 0.0002977243969391055,
+      "loss": 2.7242,
+      "step": 5028
+    },
+    {
+      "epoch": 0.643712,
+      "grad_norm": 0.4108050763607025,
+      "learning_rate": 0.0002975348251338126,
+      "loss": 2.9116,
+      "step": 5029
+    },
+    {
+      "epoch": 0.64384,
+      "grad_norm": 0.42031964659690857,
+      "learning_rate": 0.0002973452881346305,
+      "loss": 2.9266,
+      "step": 5030
+    },
+    {
+      "epoch": 0.643968,
+      "grad_norm": 0.4298689067363739,
+      "learning_rate": 0.00029715578597414266,
+      "loss": 2.7663,
+      "step": 5031
+    },
+    {
+      "epoch": 0.644096,
+      "grad_norm": 0.3988637924194336,
+      "learning_rate": 0.0002969663186849266,
+      "loss": 2.5969,
+      "step": 5032
+    },
+    {
+      "epoch": 0.644224,
+      "grad_norm": 0.4136901795864105,
+      "learning_rate": 0.00029677688629955437,
+      "loss": 2.9773,
+      "step": 5033
+    },
+    {
+      "epoch": 0.644352,
+      "grad_norm": 0.393038272857666,
+      "learning_rate": 0.00029658748885059126,
+      "loss": 2.9124,
+      "step": 5034
+    },
+    {
+      "epoch": 0.64448,
+      "grad_norm": 0.4212276339530945,
+      "learning_rate": 0.00029639812637059716,
+      "loss": 2.881,
+      "step": 5035
+    },
+    {
+      "epoch": 0.644608,
+      "grad_norm": 0.43191421031951904,
+      "learning_rate": 0.00029620879889212526,
+      "loss": 2.7289,
+      "step": 5036
+    },
+    {
+      "epoch": 0.644736,
+      "grad_norm": 0.7612763047218323,
+      "learning_rate": 0.0002960195064477235,
+      "loss": 2.8111,
+      "step": 5037
+    },
+    {
+      "epoch": 0.644864,
+      "grad_norm": 0.3903345465660095,
+      "learning_rate": 0.00029583024906993337,
+      "loss": 2.8027,
+      "step": 5038
+    },
+    {
+      "epoch": 0.644992,
+      "grad_norm": 0.4188162684440613,
+      "learning_rate": 0.00029564102679129024,
+      "loss": 2.8642,
+      "step": 5039
+    },
+    {
+      "epoch": 0.64512,
+      "grad_norm": 0.39472463726997375,
+      "learning_rate": 0.00029545183964432386,
+      "loss": 2.8674,
+      "step": 5040
+    },
+    {
+      "epoch": 0.645248,
+      "grad_norm": 0.3966025710105896,
+      "learning_rate": 0.0002952626876615575,
+      "loss": 2.8565,
+      "step": 5041
+    },
+    {
+      "epoch": 0.645376,
+      "grad_norm": 0.40463197231292725,
+      "learning_rate": 0.00029507357087550857,
+      "loss": 2.8115,
+      "step": 5042
+    },
+    {
+      "epoch": 0.645504,
+      "grad_norm": 0.4119217097759247,
+      "learning_rate": 0.0002948844893186886,
+      "loss": 2.8868,
+      "step": 5043
+    },
+    {
+      "epoch": 0.645632,
+      "grad_norm": 0.44049933552742004,
+      "learning_rate": 0.0002946954430236028,
+      "loss": 2.8717,
+      "step": 5044
+    },
+    {
+      "epoch": 0.64576,
+      "grad_norm": 0.3972229063510895,
+      "learning_rate": 0.0002945064320227505,
+      "loss": 2.8556,
+      "step": 5045
+    },
+    {
+      "epoch": 0.645888,
+      "grad_norm": 0.44193559885025024,
+      "learning_rate": 0.00029431745634862484,
+      "loss": 2.7503,
+      "step": 5046
+    },
+    {
+      "epoch": 0.646016,
+      "grad_norm": 0.4195716679096222,
+      "learning_rate": 0.0002941285160337128,
+      "loss": 2.8882,
+      "step": 5047
+    },
+    {
+      "epoch": 0.646144,
+      "grad_norm": 0.3918161988258362,
+      "learning_rate": 0.0002939396111104956,
+      "loss": 2.8803,
+      "step": 5048
+    },
+    {
+      "epoch": 0.646272,
+      "grad_norm": 0.42977625131607056,
+      "learning_rate": 0.0002937507416114481,
+      "loss": 2.7551,
+      "step": 5049
+    },
+    {
+      "epoch": 0.6464,
+      "grad_norm": 0.4123656153678894,
+      "learning_rate": 0.0002935619075690391,
+      "loss": 2.9377,
+      "step": 5050
+    },
+    {
+      "epoch": 0.646528,
+      "grad_norm": 0.404049277305603,
+      "learning_rate": 0.0002933731090157313,
+      "loss": 2.7853,
+      "step": 5051
+    },
+    {
+      "epoch": 0.646656,
+      "grad_norm": 0.4142031967639923,
+      "learning_rate": 0.0002931843459839817,
+      "loss": 2.8641,
+      "step": 5052
+    },
+    {
+      "epoch": 0.646784,
+      "grad_norm": 0.39210808277130127,
+      "learning_rate": 0.0002929956185062407,
+      "loss": 2.8413,
+      "step": 5053
+    },
+    {
+      "epoch": 0.646912,
+      "grad_norm": 0.39944136142730713,
+      "learning_rate": 0.00029280692661495247,
+      "loss": 2.8822,
+      "step": 5054
+    },
+    {
+      "epoch": 0.64704,
+      "grad_norm": 0.4656812846660614,
+      "learning_rate": 0.0002926182703425557,
+      "loss": 2.8914,
+      "step": 5055
+    },
+    {
+      "epoch": 0.647168,
+      "grad_norm": 0.3689979910850525,
+      "learning_rate": 0.00029242964972148243,
+      "loss": 2.7041,
+      "step": 5056
+    },
+    {
+      "epoch": 0.647296,
+      "grad_norm": 0.4074077308177948,
+      "learning_rate": 0.00029224106478415856,
+      "loss": 2.7945,
+      "step": 5057
+    },
+    {
+      "epoch": 0.647424,
+      "grad_norm": 0.38099026679992676,
+      "learning_rate": 0.0002920525155630044,
+      "loss": 2.651,
+      "step": 5058
+    },
+    {
+      "epoch": 0.647552,
+      "grad_norm": 0.405265212059021,
+      "learning_rate": 0.0002918640020904334,
+      "loss": 2.7085,
+      "step": 5059
+    },
+    {
+      "epoch": 0.64768,
+      "grad_norm": 0.4242308437824249,
+      "learning_rate": 0.0002916755243988533,
+      "loss": 2.8054,
+      "step": 5060
+    },
+    {
+      "epoch": 0.647808,
+      "grad_norm": 0.4028867185115814,
+      "learning_rate": 0.00029148708252066586,
+      "loss": 2.7654,
+      "step": 5061
+    },
+    {
+      "epoch": 0.647936,
+      "grad_norm": 0.39050063490867615,
+      "learning_rate": 0.0002912986764882662,
+      "loss": 2.7373,
+      "step": 5062
+    },
+    {
+      "epoch": 0.648064,
+      "grad_norm": 0.4043120741844177,
+      "learning_rate": 0.0002911103063340435,
+      "loss": 2.7906,
+      "step": 5063
+    },
+    {
+      "epoch": 0.648192,
+      "grad_norm": 0.40123510360717773,
+      "learning_rate": 0.0002909219720903808,
+      "loss": 2.8738,
+      "step": 5064
+    },
+    {
+      "epoch": 0.64832,
+      "grad_norm": 0.3932061791419983,
+      "learning_rate": 0.000290733673789655,
+      "loss": 2.7448,
+      "step": 5065
+    },
+    {
+      "epoch": 0.648448,
+      "grad_norm": 0.43300145864486694,
+      "learning_rate": 0.00029054541146423645,
+      "loss": 2.8212,
+      "step": 5066
+    },
+    {
+      "epoch": 0.648576,
+      "grad_norm": 0.38119062781333923,
+      "learning_rate": 0.00029035718514649014,
+      "loss": 2.7503,
+      "step": 5067
+    },
+    {
+      "epoch": 0.648704,
+      "grad_norm": 0.41039469838142395,
+      "learning_rate": 0.000290168994868774,
+      "loss": 2.8198,
+      "step": 5068
+    },
+    {
+      "epoch": 0.648832,
+      "grad_norm": 0.42145782709121704,
+      "learning_rate": 0.00028998084066344,
+      "loss": 2.8049,
+      "step": 5069
+    },
+    {
+      "epoch": 0.64896,
+      "grad_norm": 0.4044198989868164,
+      "learning_rate": 0.0002897927225628345,
+      "loss": 2.814,
+      "step": 5070
+    },
+    {
+      "epoch": 0.649088,
+      "grad_norm": 0.4226263463497162,
+      "learning_rate": 0.00028960464059929697,
+      "loss": 2.8896,
+      "step": 5071
+    },
+    {
+      "epoch": 0.649216,
+      "grad_norm": 0.3848384916782379,
+      "learning_rate": 0.0002894165948051608,
+      "loss": 2.8233,
+      "step": 5072
+    },
+    {
+      "epoch": 0.649344,
+      "grad_norm": 0.4452480673789978,
+      "learning_rate": 0.0002892285852127532,
+      "loss": 2.7654,
+      "step": 5073
+    },
+    {
+      "epoch": 0.649472,
+      "grad_norm": 0.3888813555240631,
+      "learning_rate": 0.0002890406118543951,
+      "loss": 2.7438,
+      "step": 5074
+    },
+    {
+      "epoch": 0.6496,
+      "grad_norm": 0.4203411638736725,
+      "learning_rate": 0.00028885267476240175,
+      "loss": 2.8334,
+      "step": 5075
+    },
+    {
+      "epoch": 0.649728,
+      "grad_norm": 0.41188639402389526,
+      "learning_rate": 0.0002886647739690814,
+      "loss": 2.8201,
+      "step": 5076
+    },
+    {
+      "epoch": 0.649856,
+      "grad_norm": 0.416334867477417,
+      "learning_rate": 0.0002884769095067364,
+      "loss": 2.8409,
+      "step": 5077
+    },
+    {
+      "epoch": 0.649984,
+      "grad_norm": 0.4008370041847229,
+      "learning_rate": 0.00028828908140766284,
+      "loss": 2.7721,
+      "step": 5078
+    },
+    {
+      "epoch": 0.650112,
+      "grad_norm": 0.41549554467201233,
+      "learning_rate": 0.00028810128970415034,
+      "loss": 2.7314,
+      "step": 5079
+    },
+    {
+      "epoch": 0.65024,
+      "grad_norm": 0.3999291658401489,
+      "learning_rate": 0.0002879135344284829,
+      "loss": 2.9084,
+      "step": 5080
+    },
+    {
+      "epoch": 0.650368,
+      "grad_norm": 0.4064357578754425,
+      "learning_rate": 0.00028772581561293745,
+      "loss": 2.8586,
+      "step": 5081
+    },
+    {
+      "epoch": 0.650496,
+      "grad_norm": 0.4103940427303314,
+      "learning_rate": 0.0002875381332897853,
+      "loss": 2.934,
+      "step": 5082
+    },
+    {
+      "epoch": 0.650624,
+      "grad_norm": 0.39495715498924255,
+      "learning_rate": 0.0002873504874912911,
+      "loss": 2.7349,
+      "step": 5083
+    },
+    {
+      "epoch": 0.650752,
+      "grad_norm": 0.4136445224285126,
+      "learning_rate": 0.00028716287824971345,
+      "loss": 2.8568,
+      "step": 5084
+    },
+    {
+      "epoch": 0.65088,
+      "grad_norm": 0.41024643182754517,
+      "learning_rate": 0.0002869753055973044,
+      "loss": 2.8794,
+      "step": 5085
+    },
+    {
+      "epoch": 0.651008,
+      "grad_norm": 0.4004479646682739,
+      "learning_rate": 0.0002867877695663099,
+      "loss": 2.7928,
+      "step": 5086
+    },
+    {
+      "epoch": 0.651136,
+      "grad_norm": 0.40798628330230713,
+      "learning_rate": 0.00028660027018896937,
+      "loss": 2.9384,
+      "step": 5087
+    },
+    {
+      "epoch": 0.651264,
+      "grad_norm": 0.40025416016578674,
+      "learning_rate": 0.00028641280749751655,
+      "loss": 2.7291,
+      "step": 5088
+    },
+    {
+      "epoch": 0.651392,
+      "grad_norm": 0.4083660840988159,
+      "learning_rate": 0.00028622538152417806,
+      "loss": 2.9069,
+      "step": 5089
+    },
+    {
+      "epoch": 0.65152,
+      "grad_norm": 0.40079373121261597,
+      "learning_rate": 0.00028603799230117493,
+      "loss": 2.8039,
+      "step": 5090
+    },
+    {
+      "epoch": 0.651648,
+      "grad_norm": 0.403460294008255,
+      "learning_rate": 0.0002858506398607214,
+      "loss": 2.9265,
+      "step": 5091
+    },
+    {
+      "epoch": 0.651776,
+      "grad_norm": 0.4278118908405304,
+      "learning_rate": 0.0002856633242350255,
+      "loss": 2.7888,
+      "step": 5092
+    },
+    {
+      "epoch": 0.651904,
+      "grad_norm": 0.40026402473449707,
+      "learning_rate": 0.000285476045456289,
+      "loss": 2.6062,
+      "step": 5093
+    },
+    {
+      "epoch": 0.652032,
+      "grad_norm": 0.39578014612197876,
+      "learning_rate": 0.00028528880355670725,
+      "loss": 2.7871,
+      "step": 5094
+    },
+    {
+      "epoch": 0.65216,
+      "grad_norm": 0.42356228828430176,
+      "learning_rate": 0.0002851015985684691,
+      "loss": 2.766,
+      "step": 5095
+    },
+    {
+      "epoch": 0.652288,
+      "grad_norm": 0.4290383458137512,
+      "learning_rate": 0.00028491443052375764,
+      "loss": 2.935,
+      "step": 5096
+    },
+    {
+      "epoch": 0.652416,
+      "grad_norm": 0.39189961552619934,
+      "learning_rate": 0.00028472729945474895,
+      "loss": 2.7979,
+      "step": 5097
+    },
+    {
+      "epoch": 0.652544,
+      "grad_norm": 0.40896332263946533,
+      "learning_rate": 0.00028454020539361326,
+      "loss": 2.8146,
+      "step": 5098
+    },
+    {
+      "epoch": 0.652672,
+      "grad_norm": 0.44575682282447815,
+      "learning_rate": 0.00028435314837251414,
+      "loss": 2.8317,
+      "step": 5099
+    },
+    {
+      "epoch": 0.6528,
+      "grad_norm": 0.37148261070251465,
+      "learning_rate": 0.0002841661284236088,
+      "loss": 2.8312,
+      "step": 5100
+    },
+    {
+      "epoch": 0.652928,
+      "grad_norm": 0.3920239806175232,
+      "learning_rate": 0.0002839791455790481,
+      "loss": 2.8166,
+      "step": 5101
+    },
+    {
+      "epoch": 0.653056,
+      "grad_norm": 0.3917125165462494,
+      "learning_rate": 0.0002837921998709765,
+      "loss": 2.8464,
+      "step": 5102
+    },
+    {
+      "epoch": 0.653184,
+      "grad_norm": 0.42281821370124817,
+      "learning_rate": 0.00028360529133153234,
+      "loss": 2.8106,
+      "step": 5103
+    },
+    {
+      "epoch": 0.653312,
+      "grad_norm": 0.43187567591667175,
+      "learning_rate": 0.0002834184199928472,
+      "loss": 2.7906,
+      "step": 5104
+    },
+    {
+      "epoch": 0.65344,
+      "grad_norm": 0.44374051690101624,
+      "learning_rate": 0.00028323158588704667,
+      "loss": 2.8487,
+      "step": 5105
+    },
+    {
+      "epoch": 0.653568,
+      "grad_norm": 0.42398974299430847,
+      "learning_rate": 0.0002830447890462492,
+      "loss": 2.8861,
+      "step": 5106
+    },
+    {
+      "epoch": 0.653696,
+      "grad_norm": 0.4026283919811249,
+      "learning_rate": 0.0002828580295025679,
+      "loss": 2.7887,
+      "step": 5107
+    },
+    {
+      "epoch": 0.653824,
+      "grad_norm": 0.43004757165908813,
+      "learning_rate": 0.00028267130728810866,
+      "loss": 2.9163,
+      "step": 5108
+    },
+    {
+      "epoch": 0.653952,
+      "grad_norm": 0.39522430300712585,
+      "learning_rate": 0.0002824846224349712,
+      "loss": 2.8013,
+      "step": 5109
+    },
+    {
+      "epoch": 0.65408,
+      "grad_norm": 0.38908180594444275,
+      "learning_rate": 0.0002822979749752486,
+      "loss": 2.8353,
+      "step": 5110
+    },
+    {
+      "epoch": 0.654208,
+      "grad_norm": 0.42507097125053406,
+      "learning_rate": 0.00028211136494102806,
+      "loss": 2.9002,
+      "step": 5111
+    },
+    {
+      "epoch": 0.654336,
+      "grad_norm": 0.4426031708717346,
+      "learning_rate": 0.00028192479236438983,
+      "loss": 2.8805,
+      "step": 5112
+    },
+    {
+      "epoch": 0.654464,
+      "grad_norm": 0.4225306808948517,
+      "learning_rate": 0.00028173825727740806,
+      "loss": 2.703,
+      "step": 5113
+    },
+    {
+      "epoch": 0.654592,
+      "grad_norm": 0.39362242817878723,
+      "learning_rate": 0.00028155175971215007,
+      "loss": 2.8942,
+      "step": 5114
+    },
+    {
+      "epoch": 0.65472,
+      "grad_norm": 0.41316846013069153,
+      "learning_rate": 0.00028136529970067714,
+      "loss": 2.8993,
+      "step": 5115
+    },
+    {
+      "epoch": 0.654848,
+      "grad_norm": 0.435039758682251,
+      "learning_rate": 0.00028117887727504355,
+      "loss": 2.7084,
+      "step": 5116
+    },
+    {
+      "epoch": 0.654976,
+      "grad_norm": 0.4118439257144928,
+      "learning_rate": 0.0002809924924672978,
+      "loss": 2.8048,
+      "step": 5117
+    },
+    {
+      "epoch": 0.655104,
+      "grad_norm": 0.38843733072280884,
+      "learning_rate": 0.0002808061453094817,
+      "loss": 2.7258,
+      "step": 5118
+    },
+    {
+      "epoch": 0.655232,
+      "grad_norm": 0.4267595410346985,
+      "learning_rate": 0.00028061983583363035,
+      "loss": 2.9178,
+      "step": 5119
+    },
+    {
+      "epoch": 0.65536,
+      "grad_norm": 0.39107751846313477,
+      "learning_rate": 0.00028043356407177254,
+      "loss": 2.8197,
+      "step": 5120
+    },
+    {
+      "epoch": 0.655488,
+      "grad_norm": 0.40090492367744446,
+      "learning_rate": 0.0002802473300559305,
+      "loss": 2.7226,
+      "step": 5121
+    },
+    {
+      "epoch": 0.655616,
+      "grad_norm": 0.40670809149742126,
+      "learning_rate": 0.00028006113381812,
+      "loss": 2.7632,
+      "step": 5122
+    },
+    {
+      "epoch": 0.655744,
+      "grad_norm": 0.4134945571422577,
+      "learning_rate": 0.00027987497539035014,
+      "loss": 2.9909,
+      "step": 5123
+    },
+    {
+      "epoch": 0.655872,
+      "grad_norm": 0.39535409212112427,
+      "learning_rate": 0.00027968885480462405,
+      "loss": 2.763,
+      "step": 5124
+    },
+    {
+      "epoch": 0.656,
+      "grad_norm": 0.3769369423389435,
+      "learning_rate": 0.00027950277209293776,
+      "loss": 2.891,
+      "step": 5125
+    },
+    {
+      "epoch": 0.656128,
+      "grad_norm": 0.37516218423843384,
+      "learning_rate": 0.00027931672728728134,
+      "loss": 2.7188,
+      "step": 5126
+    },
+    {
+      "epoch": 0.656256,
+      "grad_norm": 0.43105170130729675,
+      "learning_rate": 0.00027913072041963795,
+      "loss": 2.7577,
+      "step": 5127
+    },
+    {
+      "epoch": 0.656384,
+      "grad_norm": 0.4197482466697693,
+      "learning_rate": 0.0002789447515219842,
+      "loss": 2.9064,
+      "step": 5128
+    },
+    {
+      "epoch": 0.656512,
+      "grad_norm": 0.4162271022796631,
+      "learning_rate": 0.0002787588206262903,
+      "loss": 2.8195,
+      "step": 5129
+    },
+    {
+      "epoch": 0.65664,
+      "grad_norm": 0.4237973988056183,
+      "learning_rate": 0.00027857292776452004,
+      "loss": 2.7933,
+      "step": 5130
+    },
+    {
+      "epoch": 0.656768,
+      "grad_norm": 0.40686294436454773,
+      "learning_rate": 0.00027838707296863033,
+      "loss": 2.7311,
+      "step": 5131
+    },
+    {
+      "epoch": 0.656896,
+      "grad_norm": 0.4112490713596344,
+      "learning_rate": 0.0002782012562705721,
+      "loss": 2.8634,
+      "step": 5132
+    },
+    {
+      "epoch": 0.657024,
+      "grad_norm": 0.39399996399879456,
+      "learning_rate": 0.0002780154777022892,
+      "loss": 2.911,
+      "step": 5133
+    },
+    {
+      "epoch": 0.657152,
+      "grad_norm": 0.40658602118492126,
+      "learning_rate": 0.000277829737295719,
+      "loss": 2.8366,
+      "step": 5134
+    },
+    {
+      "epoch": 0.65728,
+      "grad_norm": 0.4284951388835907,
+      "learning_rate": 0.00027764403508279275,
+      "loss": 2.8272,
+      "step": 5135
+    },
+    {
+      "epoch": 0.657408,
+      "grad_norm": 0.4125773310661316,
+      "learning_rate": 0.0002774583710954346,
+      "loss": 2.7953,
+      "step": 5136
+    },
+    {
+      "epoch": 0.657536,
+      "grad_norm": 0.3831304609775543,
+      "learning_rate": 0.00027727274536556247,
+      "loss": 2.7268,
+      "step": 5137
+    },
+    {
+      "epoch": 0.657664,
+      "grad_norm": 0.43771547079086304,
+      "learning_rate": 0.0002770871579250872,
+      "loss": 2.8047,
+      "step": 5138
+    },
+    {
+      "epoch": 0.657792,
+      "grad_norm": 0.38871103525161743,
+      "learning_rate": 0.00027690160880591396,
+      "loss": 2.843,
+      "step": 5139
+    },
+    {
+      "epoch": 0.65792,
+      "grad_norm": 0.4287137985229492,
+      "learning_rate": 0.00027671609803994057,
+      "loss": 2.9494,
+      "step": 5140
+    },
+    {
+      "epoch": 0.658048,
+      "grad_norm": 0.4265073835849762,
+      "learning_rate": 0.0002765306256590584,
+      "loss": 2.7437,
+      "step": 5141
+    },
+    {
+      "epoch": 0.658176,
+      "grad_norm": 0.39877668023109436,
+      "learning_rate": 0.00027634519169515237,
+      "loss": 2.9147,
+      "step": 5142
+    },
+    {
+      "epoch": 0.658304,
+      "grad_norm": 0.4045288860797882,
+      "learning_rate": 0.0002761597961801006,
+      "loss": 2.8506,
+      "step": 5143
+    },
+    {
+      "epoch": 0.658432,
+      "grad_norm": 0.3970333933830261,
+      "learning_rate": 0.00027597443914577493,
+      "loss": 2.9317,
+      "step": 5144
+    },
+    {
+      "epoch": 0.65856,
+      "grad_norm": 0.40585288405418396,
+      "learning_rate": 0.0002757891206240404,
+      "loss": 2.8173,
+      "step": 5145
+    },
+    {
+      "epoch": 0.658688,
+      "grad_norm": 0.446503609418869,
+      "learning_rate": 0.00027560384064675503,
+      "loss": 2.7255,
+      "step": 5146
+    },
+    {
+      "epoch": 0.658816,
+      "grad_norm": 0.3997128903865814,
+      "learning_rate": 0.00027541859924577104,
+      "loss": 2.908,
+      "step": 5147
+    },
+    {
+      "epoch": 0.658944,
+      "grad_norm": 0.3838978111743927,
+      "learning_rate": 0.0002752333964529335,
+      "loss": 2.7315,
+      "step": 5148
+    },
+    {
+      "epoch": 0.659072,
+      "grad_norm": 0.4070565104484558,
+      "learning_rate": 0.0002750482323000808,
+      "loss": 2.7442,
+      "step": 5149
+    },
+    {
+      "epoch": 0.6592,
+      "grad_norm": 0.418671578168869,
+      "learning_rate": 0.00027486310681904485,
+      "loss": 2.8261,
+      "step": 5150
+    },
+    {
+      "epoch": 0.659328,
+      "grad_norm": 0.3974626362323761,
+      "learning_rate": 0.0002746780200416509,
+      "loss": 2.7333,
+      "step": 5151
+    },
+    {
+      "epoch": 0.659456,
+      "grad_norm": 0.4084816575050354,
+      "learning_rate": 0.00027449297199971714,
+      "loss": 2.8536,
+      "step": 5152
+    },
+    {
+      "epoch": 0.659584,
+      "grad_norm": 0.40231752395629883,
+      "learning_rate": 0.00027430796272505587,
+      "loss": 2.9072,
+      "step": 5153
+    },
+    {
+      "epoch": 0.659712,
+      "grad_norm": 0.3947332501411438,
+      "learning_rate": 0.0002741229922494724,
+      "loss": 2.9173,
+      "step": 5154
+    },
+    {
+      "epoch": 0.65984,
+      "grad_norm": 0.4148947298526764,
+      "learning_rate": 0.00027393806060476523,
+      "loss": 2.6924,
+      "step": 5155
+    },
+    {
+      "epoch": 0.659968,
+      "grad_norm": 0.41443055868148804,
+      "learning_rate": 0.0002737531678227261,
+      "loss": 2.9804,
+      "step": 5156
+    },
+    {
+      "epoch": 0.660096,
+      "grad_norm": 0.41855111718177795,
+      "learning_rate": 0.00027356831393514033,
+      "loss": 2.7782,
+      "step": 5157
+    },
+    {
+      "epoch": 0.660224,
+      "grad_norm": 0.40582534670829773,
+      "learning_rate": 0.00027338349897378637,
+      "loss": 2.7021,
+      "step": 5158
+    },
+    {
+      "epoch": 0.660352,
+      "grad_norm": 0.3922986090183258,
+      "learning_rate": 0.000273198722970436,
+      "loss": 2.9003,
+      "step": 5159
+    },
+    {
+      "epoch": 0.66048,
+      "grad_norm": 0.4141084551811218,
+      "learning_rate": 0.0002730139859568546,
+      "loss": 2.8632,
+      "step": 5160
+    },
+    {
+      "epoch": 0.660608,
+      "grad_norm": 0.3981490135192871,
+      "learning_rate": 0.00027282928796480043,
+      "loss": 2.7774,
+      "step": 5161
+    },
+    {
+      "epoch": 0.660736,
+      "grad_norm": 0.4036293029785156,
+      "learning_rate": 0.0002726446290260251,
+      "loss": 2.8611,
+      "step": 5162
+    },
+    {
+      "epoch": 0.660864,
+      "grad_norm": 0.42157310247421265,
+      "learning_rate": 0.000272460009172274,
+      "loss": 2.9259,
+      "step": 5163
+    },
+    {
+      "epoch": 0.660992,
+      "grad_norm": 0.415991872549057,
+      "learning_rate": 0.00027227542843528507,
+      "loss": 2.8989,
+      "step": 5164
+    },
+    {
+      "epoch": 0.66112,
+      "grad_norm": 0.4019545316696167,
+      "learning_rate": 0.00027209088684679013,
+      "loss": 2.835,
+      "step": 5165
+    },
+    {
+      "epoch": 0.661248,
+      "grad_norm": 0.39125940203666687,
+      "learning_rate": 0.0002719063844385138,
+      "loss": 2.7646,
+      "step": 5166
+    },
+    {
+      "epoch": 0.661376,
+      "grad_norm": 0.3923029601573944,
+      "learning_rate": 0.0002717219212421742,
+      "loss": 2.8332,
+      "step": 5167
+    },
+    {
+      "epoch": 0.661504,
+      "grad_norm": 0.4579806327819824,
+      "learning_rate": 0.00027153749728948284,
+      "loss": 2.7996,
+      "step": 5168
+    },
+    {
+      "epoch": 0.661632,
+      "grad_norm": 0.41038262844085693,
+      "learning_rate": 0.0002713531126121442,
+      "loss": 2.8436,
+      "step": 5169
+    },
+    {
+      "epoch": 0.66176,
+      "grad_norm": 0.39874348044395447,
+      "learning_rate": 0.00027116876724185633,
+      "loss": 2.889,
+      "step": 5170
+    },
+    {
+      "epoch": 0.661888,
+      "grad_norm": 0.4131772816181183,
+      "learning_rate": 0.00027098446121031006,
+      "loss": 2.8497,
+      "step": 5171
+    },
+    {
+      "epoch": 0.662016,
+      "grad_norm": 0.41099780797958374,
+      "learning_rate": 0.00027080019454919,
+      "loss": 2.8061,
+      "step": 5172
+    },
+    {
+      "epoch": 0.662144,
+      "grad_norm": 0.39785540103912354,
+      "learning_rate": 0.00027061596729017356,
+      "loss": 3.0232,
+      "step": 5173
+    },
+    {
+      "epoch": 0.662272,
+      "grad_norm": 0.38122403621673584,
+      "learning_rate": 0.00027043177946493147,
+      "loss": 2.6538,
+      "step": 5174
+    },
+    {
+      "epoch": 0.6624,
+      "grad_norm": 0.4231666326522827,
+      "learning_rate": 0.00027024763110512805,
+      "loss": 2.7566,
+      "step": 5175
+    },
+    {
+      "epoch": 0.662528,
+      "grad_norm": 0.4434388279914856,
+      "learning_rate": 0.0002700635222424204,
+      "loss": 2.7952,
+      "step": 5176
+    },
+    {
+      "epoch": 0.662656,
+      "grad_norm": 0.4110172986984253,
+      "learning_rate": 0.00026987945290845883,
+      "loss": 2.7901,
+      "step": 5177
+    },
+    {
+      "epoch": 0.662784,
+      "grad_norm": 0.4315597116947174,
+      "learning_rate": 0.0002696954231348871,
+      "loss": 2.8881,
+      "step": 5178
+    },
+    {
+      "epoch": 0.662912,
+      "grad_norm": 0.4028015732765198,
+      "learning_rate": 0.0002695114329533422,
+      "loss": 2.8246,
+      "step": 5179
+    },
+    {
+      "epoch": 0.66304,
+      "grad_norm": 0.4087422788143158,
+      "learning_rate": 0.0002693274823954537,
+      "loss": 2.7304,
+      "step": 5180
+    },
+    {
+      "epoch": 0.663168,
+      "grad_norm": 0.39912843704223633,
+      "learning_rate": 0.0002691435714928452,
+      "loss": 2.8002,
+      "step": 5181
+    },
+    {
+      "epoch": 0.663296,
+      "grad_norm": 0.4198690354824066,
+      "learning_rate": 0.0002689597002771333,
+      "loss": 2.8686,
+      "step": 5182
+    },
+    {
+      "epoch": 0.663424,
+      "grad_norm": 0.39108729362487793,
+      "learning_rate": 0.0002687758687799273,
+      "loss": 2.7943,
+      "step": 5183
+    },
+    {
+      "epoch": 0.663552,
+      "grad_norm": 0.4175972640514374,
+      "learning_rate": 0.0002685920770328301,
+      "loss": 2.8292,
+      "step": 5184
+    },
+    {
+      "epoch": 0.66368,
+      "grad_norm": 0.4132092595100403,
+      "learning_rate": 0.0002684083250674376,
+      "loss": 2.6606,
+      "step": 5185
+    },
+    {
+      "epoch": 0.663808,
+      "grad_norm": 0.3862355947494507,
+      "learning_rate": 0.0002682246129153387,
+      "loss": 2.7305,
+      "step": 5186
+    },
+    {
+      "epoch": 0.663936,
+      "grad_norm": 0.4030037820339203,
+      "learning_rate": 0.000268040940608116,
+      "loss": 2.9334,
+      "step": 5187
+    },
+    {
+      "epoch": 0.664064,
+      "grad_norm": 0.42082053422927856,
+      "learning_rate": 0.00026785730817734446,
+      "loss": 2.8101,
+      "step": 5188
+    },
+    {
+      "epoch": 0.664192,
+      "grad_norm": 0.43850210309028625,
+      "learning_rate": 0.000267673715654593,
+      "loss": 2.8415,
+      "step": 5189
+    },
+    {
+      "epoch": 0.66432,
+      "grad_norm": 0.41507911682128906,
+      "learning_rate": 0.00026749016307142307,
+      "loss": 2.7443,
+      "step": 5190
+    },
+    {
+      "epoch": 0.664448,
+      "grad_norm": 0.4448722004890442,
+      "learning_rate": 0.0002673066504593897,
+      "loss": 2.8838,
+      "step": 5191
+    },
+    {
+      "epoch": 0.664576,
+      "grad_norm": 0.394825279712677,
+      "learning_rate": 0.00026712317785004076,
+      "loss": 2.8261,
+      "step": 5192
+    },
+    {
+      "epoch": 0.664704,
+      "grad_norm": 0.3971739113330841,
+      "learning_rate": 0.00026693974527491725,
+      "loss": 2.7315,
+      "step": 5193
+    },
+    {
+      "epoch": 0.664832,
+      "grad_norm": 0.39141228795051575,
+      "learning_rate": 0.00026675635276555347,
+      "loss": 2.8836,
+      "step": 5194
+    },
+    {
+      "epoch": 0.66496,
+      "grad_norm": 0.4086637794971466,
+      "learning_rate": 0.0002665730003534764,
+      "loss": 2.8709,
+      "step": 5195
+    },
+    {
+      "epoch": 0.665088,
+      "grad_norm": 0.4084164798259735,
+      "learning_rate": 0.0002663896880702069,
+      "loss": 2.8283,
+      "step": 5196
+    },
+    {
+      "epoch": 0.665216,
+      "grad_norm": 0.37064284086227417,
+      "learning_rate": 0.0002662064159472583,
+      "loss": 2.9097,
+      "step": 5197
+    },
+    {
+      "epoch": 0.665344,
+      "grad_norm": 0.42759406566619873,
+      "learning_rate": 0.0002660231840161371,
+      "loss": 2.8519,
+      "step": 5198
+    },
+    {
+      "epoch": 0.665472,
+      "grad_norm": 0.40597647428512573,
+      "learning_rate": 0.00026583999230834294,
+      "loss": 2.935,
+      "step": 5199
+    },
+    {
+      "epoch": 0.6656,
+      "grad_norm": 0.41209933161735535,
+      "learning_rate": 0.0002656568408553689,
+      "loss": 2.7646,
+      "step": 5200
+    },
+    {
+      "epoch": 0.665728,
+      "grad_norm": 0.40767616033554077,
+      "learning_rate": 0.0002654737296887006,
+      "loss": 2.8192,
+      "step": 5201
+    },
+    {
+      "epoch": 0.665856,
+      "grad_norm": 0.3902677297592163,
+      "learning_rate": 0.00026529065883981707,
+      "loss": 2.8368,
+      "step": 5202
+    },
+    {
+      "epoch": 0.665984,
+      "grad_norm": 0.4035986363887787,
+      "learning_rate": 0.00026510762834019004,
+      "loss": 2.7312,
+      "step": 5203
+    },
+    {
+      "epoch": 0.666112,
+      "grad_norm": 0.3825324773788452,
+      "learning_rate": 0.000264924638221285,
+      "loss": 2.7315,
+      "step": 5204
+    },
+    {
+      "epoch": 0.66624,
+      "grad_norm": 0.393693208694458,
+      "learning_rate": 0.0002647416885145599,
+      "loss": 2.8948,
+      "step": 5205
+    },
+    {
+      "epoch": 0.666368,
+      "grad_norm": 0.42439955472946167,
+      "learning_rate": 0.0002645587792514659,
+      "loss": 2.8525,
+      "step": 5206
+    },
+    {
+      "epoch": 0.666496,
+      "grad_norm": 0.3959715664386749,
+      "learning_rate": 0.0002643759104634472,
+      "loss": 2.7698,
+      "step": 5207
+    },
+    {
+      "epoch": 0.666624,
+      "grad_norm": 0.3831789493560791,
+      "learning_rate": 0.00026419308218194085,
+      "loss": 2.8405,
+      "step": 5208
+    },
+    {
+      "epoch": 0.666752,
+      "grad_norm": 0.3830810785293579,
+      "learning_rate": 0.00026401029443837767,
+      "loss": 2.7267,
+      "step": 5209
+    },
+    {
+      "epoch": 0.66688,
+      "grad_norm": 0.42153894901275635,
+      "learning_rate": 0.00026382754726418043,
+      "loss": 2.7435,
+      "step": 5210
+    },
+    {
+      "epoch": 0.667008,
+      "grad_norm": 0.39695286750793457,
+      "learning_rate": 0.000263644840690766,
+      "loss": 2.7895,
+      "step": 5211
+    },
+    {
+      "epoch": 0.667136,
+      "grad_norm": 0.38016706705093384,
+      "learning_rate": 0.00026346217474954346,
+      "loss": 2.8629,
+      "step": 5212
+    },
+    {
+      "epoch": 0.667264,
+      "grad_norm": 0.45564788579940796,
+      "learning_rate": 0.00026327954947191534,
+      "loss": 2.9182,
+      "step": 5213
+    },
+    {
+      "epoch": 0.667392,
+      "grad_norm": 0.3884672522544861,
+      "learning_rate": 0.0002630969648892769,
+      "loss": 2.8727,
+      "step": 5214
+    },
+    {
+      "epoch": 0.66752,
+      "grad_norm": 0.4304999113082886,
+      "learning_rate": 0.00026291442103301666,
+      "loss": 2.8091,
+      "step": 5215
+    },
+    {
+      "epoch": 0.667648,
+      "grad_norm": 0.40186142921447754,
+      "learning_rate": 0.0002627319179345157,
+      "loss": 2.8392,
+      "step": 5216
+    },
+    {
+      "epoch": 0.667776,
+      "grad_norm": 0.3796350657939911,
+      "learning_rate": 0.00026254945562514877,
+      "loss": 2.8471,
+      "step": 5217
+    },
+    {
+      "epoch": 0.667904,
+      "grad_norm": 0.410748153924942,
+      "learning_rate": 0.0002623670341362834,
+      "loss": 2.7983,
+      "step": 5218
+    },
+    {
+      "epoch": 0.668032,
+      "grad_norm": 0.4282422661781311,
+      "learning_rate": 0.00026218465349927964,
+      "loss": 2.8535,
+      "step": 5219
+    },
+    {
+      "epoch": 0.66816,
+      "grad_norm": 0.40590932965278625,
+      "learning_rate": 0.000262002313745491,
+      "loss": 2.7828,
+      "step": 5220
+    },
+    {
+      "epoch": 0.668288,
+      "grad_norm": 0.3994084596633911,
+      "learning_rate": 0.00026182001490626373,
+      "loss": 2.7633,
+      "step": 5221
+    },
+    {
+      "epoch": 0.668416,
+      "grad_norm": 0.44352683424949646,
+      "learning_rate": 0.00026163775701293715,
+      "loss": 2.776,
+      "step": 5222
+    },
+    {
+      "epoch": 0.668544,
+      "grad_norm": 0.4049454629421234,
+      "learning_rate": 0.0002614555400968434,
+      "loss": 2.7854,
+      "step": 5223
+    },
+    {
+      "epoch": 0.668672,
+      "grad_norm": 0.4784989356994629,
+      "learning_rate": 0.0002612733641893076,
+      "loss": 2.9164,
+      "step": 5224
+    },
+    {
+      "epoch": 0.6688,
+      "grad_norm": 0.38240721821784973,
+      "learning_rate": 0.00026109122932164823,
+      "loss": 2.7808,
+      "step": 5225
+    },
+    {
+      "epoch": 0.668928,
+      "grad_norm": 0.45617231726646423,
+      "learning_rate": 0.00026090913552517624,
+      "loss": 2.9493,
+      "step": 5226
+    },
+    {
+      "epoch": 0.669056,
+      "grad_norm": 0.4626556932926178,
+      "learning_rate": 0.0002607270828311954,
+      "loss": 2.8648,
+      "step": 5227
+    },
+    {
+      "epoch": 0.669184,
+      "grad_norm": 0.4196780025959015,
+      "learning_rate": 0.00026054507127100314,
+      "loss": 2.8684,
+      "step": 5228
+    },
+    {
+      "epoch": 0.669312,
+      "grad_norm": 0.42023757100105286,
+      "learning_rate": 0.0002603631008758891,
+      "loss": 2.6919,
+      "step": 5229
+    },
+    {
+      "epoch": 0.66944,
+      "grad_norm": 0.44041839241981506,
+      "learning_rate": 0.00026018117167713617,
+      "loss": 2.8703,
+      "step": 5230
+    },
+    {
+      "epoch": 0.669568,
+      "grad_norm": 0.39826688170433044,
+      "learning_rate": 0.0002599992837060198,
+      "loss": 2.8252,
+      "step": 5231
+    },
+    {
+      "epoch": 0.669696,
+      "grad_norm": 0.41246360540390015,
+      "learning_rate": 0.0002598174369938091,
+      "loss": 2.8538,
+      "step": 5232
+    },
+    {
+      "epoch": 0.669824,
+      "grad_norm": 0.4284837543964386,
+      "learning_rate": 0.0002596356315717654,
+      "loss": 2.7981,
+      "step": 5233
+    },
+    {
+      "epoch": 0.669952,
+      "grad_norm": 0.4151444733142853,
+      "learning_rate": 0.0002594538674711432,
+      "loss": 2.7463,
+      "step": 5234
+    },
+    {
+      "epoch": 0.67008,
+      "grad_norm": 0.3953562378883362,
+      "learning_rate": 0.0002592721447231898,
+      "loss": 2.8391,
+      "step": 5235
+    },
+    {
+      "epoch": 0.670208,
+      "grad_norm": 0.39041489362716675,
+      "learning_rate": 0.00025909046335914534,
+      "loss": 2.9226,
+      "step": 5236
+    },
+    {
+      "epoch": 0.670336,
+      "grad_norm": 0.39723214507102966,
+      "learning_rate": 0.0002589088234102432,
+      "loss": 2.8839,
+      "step": 5237
+    },
+    {
+      "epoch": 0.670464,
+      "grad_norm": 0.4172837436199188,
+      "learning_rate": 0.00025872722490770915,
+      "loss": 2.8148,
+      "step": 5238
+    },
+    {
+      "epoch": 0.670592,
+      "grad_norm": 0.43569284677505493,
+      "learning_rate": 0.00025854566788276244,
+      "loss": 2.8664,
+      "step": 5239
+    },
+    {
+      "epoch": 0.67072,
+      "grad_norm": 0.3893626630306244,
+      "learning_rate": 0.0002583641523666146,
+      "loss": 2.6734,
+      "step": 5240
+    },
+    {
+      "epoch": 0.670848,
+      "grad_norm": 0.3926182687282562,
+      "learning_rate": 0.0002581826783904702,
+      "loss": 2.793,
+      "step": 5241
+    },
+    {
+      "epoch": 0.670976,
+      "grad_norm": 0.43659254908561707,
+      "learning_rate": 0.00025800124598552676,
+      "loss": 2.9135,
+      "step": 5242
+    },
+    {
+      "epoch": 0.671104,
+      "grad_norm": 0.40386638045310974,
+      "learning_rate": 0.0002578198551829747,
+      "loss": 2.9254,
+      "step": 5243
+    },
+    {
+      "epoch": 0.671232,
+      "grad_norm": 0.4184795618057251,
+      "learning_rate": 0.00025763850601399706,
+      "loss": 2.8678,
+      "step": 5244
+    },
+    {
+      "epoch": 0.67136,
+      "grad_norm": 0.39328640699386597,
+      "learning_rate": 0.00025745719850976976,
+      "loss": 2.7751,
+      "step": 5245
+    },
+    {
+      "epoch": 0.671488,
+      "grad_norm": 0.3857335150241852,
+      "learning_rate": 0.0002572759327014618,
+      "loss": 2.7291,
+      "step": 5246
+    },
+    {
+      "epoch": 0.671616,
+      "grad_norm": 0.4467376470565796,
+      "learning_rate": 0.00025709470862023516,
+      "loss": 2.8332,
+      "step": 5247
+    },
+    {
+      "epoch": 0.671744,
+      "grad_norm": 0.42435774207115173,
+      "learning_rate": 0.00025691352629724406,
+      "loss": 2.9134,
+      "step": 5248
+    },
+    {
+      "epoch": 0.671872,
+      "grad_norm": 0.40861132740974426,
+      "learning_rate": 0.0002567323857636359,
+      "loss": 2.7668,
+      "step": 5249
+    },
+    {
+      "epoch": 0.672,
+      "grad_norm": 0.40449368953704834,
+      "learning_rate": 0.00025655128705055085,
+      "loss": 2.7898,
+      "step": 5250
+    },
+    {
+      "epoch": 0.672128,
+      "grad_norm": 0.4328538775444031,
+      "learning_rate": 0.00025637023018912185,
+      "loss": 2.8262,
+      "step": 5251
+    },
+    {
+      "epoch": 0.672256,
+      "grad_norm": 0.4153282940387726,
+      "learning_rate": 0.0002561892152104744,
+      "loss": 2.8343,
+      "step": 5252
+    },
+    {
+      "epoch": 0.672384,
+      "grad_norm": 0.3661895990371704,
+      "learning_rate": 0.0002560082421457277,
+      "loss": 2.7948,
+      "step": 5253
+    },
+    {
+      "epoch": 0.672512,
+      "grad_norm": 0.3835928738117218,
+      "learning_rate": 0.0002558273110259925,
+      "loss": 2.8325,
+      "step": 5254
+    },
+    {
+      "epoch": 0.67264,
+      "grad_norm": 0.4084993600845337,
+      "learning_rate": 0.00025564642188237344,
+      "loss": 2.8152,
+      "step": 5255
+    },
+    {
+      "epoch": 0.672768,
+      "grad_norm": 0.3874063491821289,
+      "learning_rate": 0.0002554655747459672,
+      "loss": 2.7518,
+      "step": 5256
+    },
+    {
+      "epoch": 0.672896,
+      "grad_norm": 0.40424731373786926,
+      "learning_rate": 0.0002552847696478636,
+      "loss": 2.8241,
+      "step": 5257
+    },
+    {
+      "epoch": 0.673024,
+      "grad_norm": 0.382347047328949,
+      "learning_rate": 0.000255104006619145,
+      "loss": 2.7866,
+      "step": 5258
+    },
+    {
+      "epoch": 0.673152,
+      "grad_norm": 0.4206586480140686,
+      "learning_rate": 0.00025492328569088664,
+      "loss": 2.7786,
+      "step": 5259
+    },
+    {
+      "epoch": 0.67328,
+      "grad_norm": 0.42617422342300415,
+      "learning_rate": 0.00025474260689415686,
+      "loss": 2.9404,
+      "step": 5260
+    },
+    {
+      "epoch": 0.673408,
+      "grad_norm": 0.4369490444660187,
+      "learning_rate": 0.0002545619702600161,
+      "loss": 2.6956,
+      "step": 5261
+    },
+    {
+      "epoch": 0.673536,
+      "grad_norm": 0.38806819915771484,
+      "learning_rate": 0.0002543813758195181,
+      "loss": 2.8651,
+      "step": 5262
+    },
+    {
+      "epoch": 0.673664,
+      "grad_norm": 0.4380597472190857,
+      "learning_rate": 0.000254200823603709,
+      "loss": 2.7972,
+      "step": 5263
+    },
+    {
+      "epoch": 0.673792,
+      "grad_norm": 0.4066748321056366,
+      "learning_rate": 0.0002540203136436277,
+      "loss": 2.8606,
+      "step": 5264
+    },
+    {
+      "epoch": 0.67392,
+      "grad_norm": 0.41663801670074463,
+      "learning_rate": 0.00025383984597030635,
+      "loss": 2.8685,
+      "step": 5265
+    },
+    {
+      "epoch": 0.674048,
+      "grad_norm": 0.39280590415000916,
+      "learning_rate": 0.0002536594206147692,
+      "loss": 2.8149,
+      "step": 5266
+    },
+    {
+      "epoch": 0.674176,
+      "grad_norm": 0.3841207027435303,
+      "learning_rate": 0.00025347903760803325,
+      "loss": 2.8936,
+      "step": 5267
+    },
+    {
+      "epoch": 0.674304,
+      "grad_norm": 0.3996492922306061,
+      "learning_rate": 0.00025329869698110884,
+      "loss": 2.8913,
+      "step": 5268
+    },
+    {
+      "epoch": 0.674432,
+      "grad_norm": 0.40421798825263977,
+      "learning_rate": 0.00025311839876499844,
+      "loss": 2.713,
+      "step": 5269
+    },
+    {
+      "epoch": 0.67456,
+      "grad_norm": 0.4060192108154297,
+      "learning_rate": 0.0002529381429906974,
+      "loss": 2.9113,
+      "step": 5270
+    },
+    {
+      "epoch": 0.674688,
+      "grad_norm": 0.3970266580581665,
+      "learning_rate": 0.00025275792968919373,
+      "loss": 2.8604,
+      "step": 5271
+    },
+    {
+      "epoch": 0.674816,
+      "grad_norm": 0.4179273247718811,
+      "learning_rate": 0.00025257775889146825,
+      "loss": 2.8525,
+      "step": 5272
+    },
+    {
+      "epoch": 0.674944,
+      "grad_norm": 0.4038170874118805,
+      "learning_rate": 0.00025239763062849427,
+      "loss": 2.783,
+      "step": 5273
+    },
+    {
+      "epoch": 0.675072,
+      "grad_norm": 0.4190945327281952,
+      "learning_rate": 0.00025221754493123794,
+      "loss": 2.8473,
+      "step": 5274
+    },
+    {
+      "epoch": 0.6752,
+      "grad_norm": 0.37771737575531006,
+      "learning_rate": 0.0002520375018306584,
+      "loss": 2.7651,
+      "step": 5275
+    },
+    {
+      "epoch": 0.675328,
+      "grad_norm": 0.41996967792510986,
+      "learning_rate": 0.0002518575013577069,
+      "loss": 2.7862,
+      "step": 5276
+    },
+    {
+      "epoch": 0.675456,
+      "grad_norm": 0.4064623713493347,
+      "learning_rate": 0.0002516775435433276,
+      "loss": 2.7692,
+      "step": 5277
+    },
+    {
+      "epoch": 0.675584,
+      "grad_norm": 0.3851461410522461,
+      "learning_rate": 0.00025149762841845746,
+      "loss": 2.7276,
+      "step": 5278
+    },
+    {
+      "epoch": 0.675712,
+      "grad_norm": 0.40905508399009705,
+      "learning_rate": 0.00025131775601402583,
+      "loss": 2.8124,
+      "step": 5279
+    },
+    {
+      "epoch": 0.67584,
+      "grad_norm": 0.383443146944046,
+      "learning_rate": 0.00025113792636095465,
+      "loss": 2.8571,
+      "step": 5280
+    },
+    {
+      "epoch": 0.675968,
+      "grad_norm": 0.3971133232116699,
+      "learning_rate": 0.00025095813949015935,
+      "loss": 2.8674,
+      "step": 5281
+    },
+    {
+      "epoch": 0.676096,
+      "grad_norm": 0.4214121103286743,
+      "learning_rate": 0.0002507783954325467,
+      "loss": 2.8444,
+      "step": 5282
+    },
+    {
+      "epoch": 0.676224,
+      "grad_norm": 0.4117683172225952,
+      "learning_rate": 0.0002505986942190173,
+      "loss": 2.9026,
+      "step": 5283
+    },
+    {
+      "epoch": 0.676352,
+      "grad_norm": 0.3981657922267914,
+      "learning_rate": 0.0002504190358804637,
+      "loss": 2.8041,
+      "step": 5284
+    },
+    {
+      "epoch": 0.67648,
+      "grad_norm": 0.43165698647499084,
+      "learning_rate": 0.00025023942044777114,
+      "loss": 2.807,
+      "step": 5285
+    },
+    {
+      "epoch": 0.676608,
+      "grad_norm": 0.43119126558303833,
+      "learning_rate": 0.0002500598479518178,
+      "loss": 2.9266,
+      "step": 5286
+    },
+    {
+      "epoch": 0.676736,
+      "grad_norm": 0.41144880652427673,
+      "learning_rate": 0.00024988031842347407,
+      "loss": 2.7663,
+      "step": 5287
+    },
+    {
+      "epoch": 0.676864,
+      "grad_norm": 0.37657395005226135,
+      "learning_rate": 0.0002497008318936031,
+      "loss": 2.8525,
+      "step": 5288
+    },
+    {
+      "epoch": 0.676992,
+      "grad_norm": 0.3895387351512909,
+      "learning_rate": 0.000249521388393061,
+      "loss": 2.9882,
+      "step": 5289
+    },
+    {
+      "epoch": 0.67712,
+      "grad_norm": 0.4189384877681732,
+      "learning_rate": 0.00024934198795269604,
+      "loss": 2.8479,
+      "step": 5290
+    },
+    {
+      "epoch": 0.677248,
+      "grad_norm": 0.39337247610092163,
+      "learning_rate": 0.00024916263060334904,
+      "loss": 2.8299,
+      "step": 5291
+    },
+    {
+      "epoch": 0.677376,
+      "grad_norm": 0.4067697823047638,
+      "learning_rate": 0.000248983316375854,
+      "loss": 2.8771,
+      "step": 5292
+    },
+    {
+      "epoch": 0.677504,
+      "grad_norm": 0.4042368233203888,
+      "learning_rate": 0.0002488040453010369,
+      "loss": 2.8483,
+      "step": 5293
+    },
+    {
+      "epoch": 0.677632,
+      "grad_norm": 0.41775545477867126,
+      "learning_rate": 0.0002486248174097165,
+      "loss": 2.7964,
+      "step": 5294
+    },
+    {
+      "epoch": 0.67776,
+      "grad_norm": 0.44083699584007263,
+      "learning_rate": 0.00024844563273270404,
+      "loss": 2.8187,
+      "step": 5295
+    },
+    {
+      "epoch": 0.677888,
+      "grad_norm": 0.4050145447254181,
+      "learning_rate": 0.0002482664913008037,
+      "loss": 2.8598,
+      "step": 5296
+    },
+    {
+      "epoch": 0.678016,
+      "grad_norm": 0.3937114477157593,
+      "learning_rate": 0.0002480873931448119,
+      "loss": 2.9595,
+      "step": 5297
+    },
+    {
+      "epoch": 0.678144,
+      "grad_norm": 0.3999718725681305,
+      "learning_rate": 0.0002479083382955176,
+      "loss": 2.8388,
+      "step": 5298
+    },
+    {
+      "epoch": 0.678272,
+      "grad_norm": 0.415421724319458,
+      "learning_rate": 0.0002477293267837025,
+      "loss": 3.0251,
+      "step": 5299
+    },
+    {
+      "epoch": 0.6784,
+      "grad_norm": 0.395977258682251,
+      "learning_rate": 0.00024755035864014067,
+      "loss": 2.9057,
+      "step": 5300
+    },
+    {
+      "epoch": 0.678528,
+      "grad_norm": 0.3681545853614807,
+      "learning_rate": 0.0002473714338955987,
+      "loss": 2.89,
+      "step": 5301
+    },
+    {
+      "epoch": 0.678656,
+      "grad_norm": 0.4023945927619934,
+      "learning_rate": 0.0002471925525808361,
+      "loss": 2.7829,
+      "step": 5302
+    },
+    {
+      "epoch": 0.678784,
+      "grad_norm": 0.38578805327415466,
+      "learning_rate": 0.0002470137147266043,
+      "loss": 2.8094,
+      "step": 5303
+    },
+    {
+      "epoch": 0.678912,
+      "grad_norm": 0.38273152709007263,
+      "learning_rate": 0.000246834920363648,
+      "loss": 2.7635,
+      "step": 5304
+    },
+    {
+      "epoch": 0.67904,
+      "grad_norm": 0.3989138603210449,
+      "learning_rate": 0.00024665616952270386,
+      "loss": 2.7752,
+      "step": 5305
+    },
+    {
+      "epoch": 0.679168,
+      "grad_norm": 0.40970954298973083,
+      "learning_rate": 0.00024647746223450126,
+      "loss": 2.8696,
+      "step": 5306
+    },
+    {
+      "epoch": 0.679296,
+      "grad_norm": 0.41300952434539795,
+      "learning_rate": 0.0002462987985297619,
+      "loss": 2.8504,
+      "step": 5307
+    },
+    {
+      "epoch": 0.679424,
+      "grad_norm": 0.387833833694458,
+      "learning_rate": 0.0002461201784392002,
+      "loss": 2.9143,
+      "step": 5308
+    },
+    {
+      "epoch": 0.679552,
+      "grad_norm": 0.40905311703681946,
+      "learning_rate": 0.0002459416019935229,
+      "loss": 2.878,
+      "step": 5309
+    },
+    {
+      "epoch": 0.67968,
+      "grad_norm": 0.38554954528808594,
+      "learning_rate": 0.0002457630692234296,
+      "loss": 2.7878,
+      "step": 5310
+    },
+    {
+      "epoch": 0.679808,
+      "grad_norm": 0.41959506273269653,
+      "learning_rate": 0.0002455845801596121,
+      "loss": 2.9602,
+      "step": 5311
+    },
+    {
+      "epoch": 0.679936,
+      "grad_norm": 0.4071325361728668,
+      "learning_rate": 0.0002454061348327548,
+      "loss": 2.7789,
+      "step": 5312
+    },
+    {
+      "epoch": 0.680064,
+      "grad_norm": 0.3757457137107849,
+      "learning_rate": 0.00024522773327353444,
+      "loss": 2.7758,
+      "step": 5313
+    },
+    {
+      "epoch": 0.680192,
+      "grad_norm": 0.3987123370170593,
+      "learning_rate": 0.00024504937551262024,
+      "loss": 2.7615,
+      "step": 5314
+    },
+    {
+      "epoch": 0.68032,
+      "grad_norm": 0.40453779697418213,
+      "learning_rate": 0.00024487106158067407,
+      "loss": 2.8408,
+      "step": 5315
+    },
+    {
+      "epoch": 0.680448,
+      "grad_norm": 0.4230065941810608,
+      "learning_rate": 0.00024469279150834993,
+      "loss": 2.9635,
+      "step": 5316
+    },
+    {
+      "epoch": 0.680576,
+      "grad_norm": 0.3924727439880371,
+      "learning_rate": 0.00024451456532629493,
+      "loss": 2.7945,
+      "step": 5317
+    },
+    {
+      "epoch": 0.680704,
+      "grad_norm": 0.426673024892807,
+      "learning_rate": 0.00024433638306514796,
+      "loss": 2.697,
+      "step": 5318
+    },
+    {
+      "epoch": 0.680832,
+      "grad_norm": 0.3944379985332489,
+      "learning_rate": 0.00024415824475554045,
+      "loss": 2.8941,
+      "step": 5319
+    },
+    {
+      "epoch": 0.68096,
+      "grad_norm": 0.39052432775497437,
+      "learning_rate": 0.0002439801504280968,
+      "loss": 2.7465,
+      "step": 5320
+    },
+    {
+      "epoch": 0.681088,
+      "grad_norm": 0.39921873807907104,
+      "learning_rate": 0.0002438021001134334,
+      "loss": 2.798,
+      "step": 5321
+    },
+    {
+      "epoch": 0.681216,
+      "grad_norm": 0.3891114592552185,
+      "learning_rate": 0.00024362409384215904,
+      "loss": 2.9013,
+      "step": 5322
+    },
+    {
+      "epoch": 0.681344,
+      "grad_norm": 0.3977868854999542,
+      "learning_rate": 0.00024344613164487512,
+      "loss": 2.8364,
+      "step": 5323
+    },
+    {
+      "epoch": 0.681472,
+      "grad_norm": 0.39534756541252136,
+      "learning_rate": 0.0002432682135521752,
+      "loss": 2.7568,
+      "step": 5324
+    },
+    {
+      "epoch": 0.6816,
+      "grad_norm": 0.3855177164077759,
+      "learning_rate": 0.0002430903395946459,
+      "loss": 2.7024,
+      "step": 5325
+    },
+    {
+      "epoch": 0.681728,
+      "grad_norm": 0.4123527407646179,
+      "learning_rate": 0.0002429125098028655,
+      "loss": 2.8069,
+      "step": 5326
+    },
+    {
+      "epoch": 0.681856,
+      "grad_norm": 0.3763120174407959,
+      "learning_rate": 0.0002427347242074051,
+      "loss": 2.7081,
+      "step": 5327
+    },
+    {
+      "epoch": 0.681984,
+      "grad_norm": 0.42332616448402405,
+      "learning_rate": 0.00024255698283882793,
+      "loss": 2.8507,
+      "step": 5328
+    },
+    {
+      "epoch": 0.682112,
+      "grad_norm": 0.4103545546531677,
+      "learning_rate": 0.00024237928572769013,
+      "loss": 2.8372,
+      "step": 5329
+    },
+    {
+      "epoch": 0.68224,
+      "grad_norm": 0.3831997215747833,
+      "learning_rate": 0.00024220163290453972,
+      "loss": 2.7725,
+      "step": 5330
+    },
+    {
+      "epoch": 0.682368,
+      "grad_norm": 0.3998417258262634,
+      "learning_rate": 0.00024202402439991706,
+      "loss": 2.685,
+      "step": 5331
+    },
+    {
+      "epoch": 0.682496,
+      "grad_norm": 0.44757354259490967,
+      "learning_rate": 0.00024184646024435547,
+      "loss": 2.8478,
+      "step": 5332
+    },
+    {
+      "epoch": 0.682624,
+      "grad_norm": 0.413164883852005,
+      "learning_rate": 0.00024166894046838018,
+      "loss": 2.839,
+      "step": 5333
+    },
+    {
+      "epoch": 0.682752,
+      "grad_norm": 0.4086025059223175,
+      "learning_rate": 0.00024149146510250884,
+      "loss": 2.7757,
+      "step": 5334
+    },
+    {
+      "epoch": 0.68288,
+      "grad_norm": 0.4035036563873291,
+      "learning_rate": 0.00024131403417725156,
+      "loss": 2.8577,
+      "step": 5335
+    },
+    {
+      "epoch": 0.683008,
+      "grad_norm": 0.4030012786388397,
+      "learning_rate": 0.00024113664772311072,
+      "loss": 2.8257,
+      "step": 5336
+    },
+    {
+      "epoch": 0.683136,
+      "grad_norm": 0.41388168931007385,
+      "learning_rate": 0.000240959305770581,
+      "loss": 2.7994,
+      "step": 5337
+    },
+    {
+      "epoch": 0.683264,
+      "grad_norm": 0.4173181653022766,
+      "learning_rate": 0.0002407820083501498,
+      "loss": 2.92,
+      "step": 5338
+    },
+    {
+      "epoch": 0.683392,
+      "grad_norm": 0.4069133400917053,
+      "learning_rate": 0.00024060475549229633,
+      "loss": 2.9664,
+      "step": 5339
+    },
+    {
+      "epoch": 0.68352,
+      "grad_norm": 0.40762859582901,
+      "learning_rate": 0.0002404275472274927,
+      "loss": 2.8427,
+      "step": 5340
+    },
+    {
+      "epoch": 0.683648,
+      "grad_norm": 0.40038585662841797,
+      "learning_rate": 0.000240250383586203,
+      "loss": 2.8775,
+      "step": 5341
+    },
+    {
+      "epoch": 0.683776,
+      "grad_norm": 0.40370211005210876,
+      "learning_rate": 0.00024007326459888357,
+      "loss": 2.8645,
+      "step": 5342
+    },
+    {
+      "epoch": 0.683904,
+      "grad_norm": 0.4215662479400635,
+      "learning_rate": 0.0002398961902959833,
+      "loss": 2.8099,
+      "step": 5343
+    },
+    {
+      "epoch": 0.684032,
+      "grad_norm": 0.3943985104560852,
+      "learning_rate": 0.00023971916070794337,
+      "loss": 2.7723,
+      "step": 5344
+    },
+    {
+      "epoch": 0.68416,
+      "grad_norm": 0.40808042883872986,
+      "learning_rate": 0.0002395421758651969,
+      "loss": 2.815,
+      "step": 5345
+    },
+    {
+      "epoch": 0.684288,
+      "grad_norm": 0.3845255970954895,
+      "learning_rate": 0.00023936523579817003,
+      "loss": 2.7284,
+      "step": 5346
+    },
+    {
+      "epoch": 0.684416,
+      "grad_norm": 0.42593592405319214,
+      "learning_rate": 0.0002391883405372805,
+      "loss": 2.8711,
+      "step": 5347
+    },
+    {
+      "epoch": 0.684544,
+      "grad_norm": 0.39849549531936646,
+      "learning_rate": 0.00023901149011293905,
+      "loss": 2.941,
+      "step": 5348
+    },
+    {
+      "epoch": 0.684672,
+      "grad_norm": 0.41917353868484497,
+      "learning_rate": 0.00023883468455554807,
+      "loss": 2.858,
+      "step": 5349
+    },
+    {
+      "epoch": 0.6848,
+      "grad_norm": 0.4313604533672333,
+      "learning_rate": 0.0002386579238955025,
+      "loss": 2.8132,
+      "step": 5350
+    },
+    {
+      "epoch": 0.684928,
+      "grad_norm": 0.42837008833885193,
+      "learning_rate": 0.00023848120816318954,
+      "loss": 3.017,
+      "step": 5351
+    },
+    {
+      "epoch": 0.685056,
+      "grad_norm": 0.39096754789352417,
+      "learning_rate": 0.0002383045373889885,
+      "loss": 2.7996,
+      "step": 5352
+    },
+    {
+      "epoch": 0.685184,
+      "grad_norm": 0.4117313623428345,
+      "learning_rate": 0.00023812791160327153,
+      "loss": 2.756,
+      "step": 5353
+    },
+    {
+      "epoch": 0.685312,
+      "grad_norm": 0.4257199168205261,
+      "learning_rate": 0.0002379513308364024,
+      "loss": 2.8925,
+      "step": 5354
+    },
+    {
+      "epoch": 0.68544,
+      "grad_norm": 0.4021505117416382,
+      "learning_rate": 0.00023777479511873746,
+      "loss": 2.8961,
+      "step": 5355
+    },
+    {
+      "epoch": 0.685568,
+      "grad_norm": 0.4237460196018219,
+      "learning_rate": 0.00023759830448062498,
+      "loss": 2.9365,
+      "step": 5356
+    },
+    {
+      "epoch": 0.685696,
+      "grad_norm": 0.3965706527233124,
+      "learning_rate": 0.0002374218589524062,
+      "loss": 2.7673,
+      "step": 5357
+    },
+    {
+      "epoch": 0.685824,
+      "grad_norm": 0.39627522230148315,
+      "learning_rate": 0.00023724545856441392,
+      "loss": 2.8087,
+      "step": 5358
+    },
+    {
+      "epoch": 0.685952,
+      "grad_norm": 0.3891827166080475,
+      "learning_rate": 0.00023706910334697347,
+      "loss": 2.7148,
+      "step": 5359
+    },
+    {
+      "epoch": 0.68608,
+      "grad_norm": 0.45396173000335693,
+      "learning_rate": 0.0002368927933304021,
+      "loss": 2.8378,
+      "step": 5360
+    },
+    {
+      "epoch": 0.686208,
+      "grad_norm": 0.39677974581718445,
+      "learning_rate": 0.0002367165285450099,
+      "loss": 2.8684,
+      "step": 5361
+    },
+    {
+      "epoch": 0.686336,
+      "grad_norm": 0.41682976484298706,
+      "learning_rate": 0.00023654030902109873,
+      "loss": 2.7847,
+      "step": 5362
+    },
+    {
+      "epoch": 0.686464,
+      "grad_norm": 0.40857040882110596,
+      "learning_rate": 0.00023636413478896274,
+      "loss": 2.8111,
+      "step": 5363
+    },
+    {
+      "epoch": 0.686592,
+      "grad_norm": 0.3963772654533386,
+      "learning_rate": 0.00023618800587888827,
+      "loss": 2.6868,
+      "step": 5364
+    },
+    {
+      "epoch": 0.68672,
+      "grad_norm": 0.38795197010040283,
+      "learning_rate": 0.00023601192232115375,
+      "loss": 2.7484,
+      "step": 5365
+    },
+    {
+      "epoch": 0.686848,
+      "grad_norm": 0.4065529704093933,
+      "learning_rate": 0.00023583588414603047,
+      "loss": 2.942,
+      "step": 5366
+    },
+    {
+      "epoch": 0.686976,
+      "grad_norm": 0.3855719268321991,
+      "learning_rate": 0.0002356598913837809,
+      "loss": 2.7824,
+      "step": 5367
+    },
+    {
+      "epoch": 0.687104,
+      "grad_norm": 0.41944602131843567,
+      "learning_rate": 0.00023548394406466074,
+      "loss": 2.8702,
+      "step": 5368
+    },
+    {
+      "epoch": 0.687232,
+      "grad_norm": 0.39604681730270386,
+      "learning_rate": 0.00023530804221891705,
+      "loss": 2.7798,
+      "step": 5369
+    },
+    {
+      "epoch": 0.68736,
+      "grad_norm": 0.3844750225543976,
+      "learning_rate": 0.00023513218587678953,
+      "loss": 2.7907,
+      "step": 5370
+    },
+    {
+      "epoch": 0.687488,
+      "grad_norm": 0.431153804063797,
+      "learning_rate": 0.00023495637506850987,
+      "loss": 2.7019,
+      "step": 5371
+    },
+    {
+      "epoch": 0.687616,
+      "grad_norm": 0.39983904361724854,
+      "learning_rate": 0.00023478060982430195,
+      "loss": 2.7844,
+      "step": 5372
+    },
+    {
+      "epoch": 0.687744,
+      "grad_norm": 0.4107910394668579,
+      "learning_rate": 0.00023460489017438174,
+      "loss": 2.8924,
+      "step": 5373
+    },
+    {
+      "epoch": 0.687872,
+      "grad_norm": 0.3868301212787628,
+      "learning_rate": 0.00023442921614895778,
+      "loss": 2.7506,
+      "step": 5374
+    },
+    {
+      "epoch": 0.688,
+      "grad_norm": 0.37864989042282104,
+      "learning_rate": 0.0002342535877782302,
+      "loss": 2.7427,
+      "step": 5375
+    },
+    {
+      "epoch": 0.688128,
+      "grad_norm": 0.3947281539440155,
+      "learning_rate": 0.0002340780050923918,
+      "loss": 2.8837,
+      "step": 5376
+    },
+    {
+      "epoch": 0.688256,
+      "grad_norm": 0.4087953567504883,
+      "learning_rate": 0.00023390246812162718,
+      "loss": 2.9565,
+      "step": 5377
+    },
+    {
+      "epoch": 0.688384,
+      "grad_norm": 0.41392627358436584,
+      "learning_rate": 0.00023372697689611317,
+      "loss": 2.8517,
+      "step": 5378
+    },
+    {
+      "epoch": 0.688512,
+      "grad_norm": 0.42951709032058716,
+      "learning_rate": 0.0002335515314460187,
+      "loss": 2.8862,
+      "step": 5379
+    },
+    {
+      "epoch": 0.68864,
+      "grad_norm": 0.41394132375717163,
+      "learning_rate": 0.00023337613180150497,
+      "loss": 2.8277,
+      "step": 5380
+    },
+    {
+      "epoch": 0.688768,
+      "grad_norm": 0.3862796127796173,
+      "learning_rate": 0.00023320077799272493,
+      "loss": 2.8532,
+      "step": 5381
+    },
+    {
+      "epoch": 0.688896,
+      "grad_norm": 0.4154782295227051,
+      "learning_rate": 0.00023302547004982433,
+      "loss": 2.8823,
+      "step": 5382
+    },
+    {
+      "epoch": 0.689024,
+      "grad_norm": 0.40478822588920593,
+      "learning_rate": 0.0002328502080029405,
+      "loss": 2.8173,
+      "step": 5383
+    },
+    {
+      "epoch": 0.689152,
+      "grad_norm": 0.3842703402042389,
+      "learning_rate": 0.00023267499188220277,
+      "loss": 2.7287,
+      "step": 5384
+    },
+    {
+      "epoch": 0.68928,
+      "grad_norm": 0.4328641891479492,
+      "learning_rate": 0.00023249982171773322,
+      "loss": 2.8166,
+      "step": 5385
+    },
+    {
+      "epoch": 0.689408,
+      "grad_norm": 0.42197728157043457,
+      "learning_rate": 0.0002323246975396454,
+      "loss": 2.824,
+      "step": 5386
+    },
+    {
+      "epoch": 0.689536,
+      "grad_norm": 0.41290637850761414,
+      "learning_rate": 0.0002321496193780453,
+      "loss": 2.8288,
+      "step": 5387
+    },
+    {
+      "epoch": 0.689664,
+      "grad_norm": 0.4545500576496124,
+      "learning_rate": 0.0002319745872630306,
+      "loss": 2.8319,
+      "step": 5388
+    },
+    {
+      "epoch": 0.689792,
+      "grad_norm": 0.3930759131908417,
+      "learning_rate": 0.00023179960122469173,
+      "loss": 2.6846,
+      "step": 5389
+    },
+    {
+      "epoch": 0.68992,
+      "grad_norm": 0.4010675847530365,
+      "learning_rate": 0.00023162466129311073,
+      "loss": 2.9255,
+      "step": 5390
+    },
+    {
+      "epoch": 0.690048,
+      "grad_norm": 0.3782845735549927,
+      "learning_rate": 0.00023144976749836166,
+      "loss": 2.8343,
+      "step": 5391
+    },
+    {
+      "epoch": 0.690176,
+      "grad_norm": 0.4334155321121216,
+      "learning_rate": 0.00023127491987051085,
+      "loss": 2.8799,
+      "step": 5392
+    },
+    {
+      "epoch": 0.690304,
+      "grad_norm": 0.3932889401912689,
+      "learning_rate": 0.0002311001184396165,
+      "loss": 2.8201,
+      "step": 5393
+    },
+    {
+      "epoch": 0.690432,
+      "grad_norm": 0.3985103964805603,
+      "learning_rate": 0.00023092536323572933,
+      "loss": 2.7249,
+      "step": 5394
+    },
+    {
+      "epoch": 0.69056,
+      "grad_norm": 0.3914976418018341,
+      "learning_rate": 0.0002307506542888916,
+      "loss": 2.8334,
+      "step": 5395
+    },
+    {
+      "epoch": 0.690688,
+      "grad_norm": 0.43340083956718445,
+      "learning_rate": 0.00023057599162913756,
+      "loss": 2.7894,
+      "step": 5396
+    },
+    {
+      "epoch": 0.690816,
+      "grad_norm": 0.3905002176761627,
+      "learning_rate": 0.00023040137528649424,
+      "loss": 2.8992,
+      "step": 5397
+    },
+    {
+      "epoch": 0.690944,
+      "grad_norm": 0.37721338868141174,
+      "learning_rate": 0.00023022680529097994,
+      "loss": 2.8025,
+      "step": 5398
+    },
+    {
+      "epoch": 0.691072,
+      "grad_norm": 0.4344537556171417,
+      "learning_rate": 0.00023005228167260528,
+      "loss": 3.0054,
+      "step": 5399
+    },
+    {
+      "epoch": 0.6912,
+      "grad_norm": 0.4027358889579773,
+      "learning_rate": 0.00022987780446137292,
+      "loss": 2.9803,
+      "step": 5400
+    },
+    {
+      "epoch": 0.691328,
+      "grad_norm": 0.4238596558570862,
+      "learning_rate": 0.0002297033736872775,
+      "loss": 2.8727,
+      "step": 5401
+    },
+    {
+      "epoch": 0.691456,
+      "grad_norm": 0.3977165222167969,
+      "learning_rate": 0.00022952898938030552,
+      "loss": 2.9301,
+      "step": 5402
+    },
+    {
+      "epoch": 0.691584,
+      "grad_norm": 0.405979722738266,
+      "learning_rate": 0.0002293546515704358,
+      "loss": 2.8161,
+      "step": 5403
+    },
+    {
+      "epoch": 0.691712,
+      "grad_norm": 0.4010053277015686,
+      "learning_rate": 0.00022918036028763927,
+      "loss": 2.8807,
+      "step": 5404
+    },
+    {
+      "epoch": 0.69184,
+      "grad_norm": 0.4341963827610016,
+      "learning_rate": 0.0002290061155618784,
+      "loss": 2.9028,
+      "step": 5405
+    },
+    {
+      "epoch": 0.691968,
+      "grad_norm": 0.4171775281429291,
+      "learning_rate": 0.00022883191742310795,
+      "loss": 2.8485,
+      "step": 5406
+    },
+    {
+      "epoch": 0.692096,
+      "grad_norm": 0.40984734892845154,
+      "learning_rate": 0.00022865776590127445,
+      "loss": 2.7622,
+      "step": 5407
+    },
+    {
+      "epoch": 0.692224,
+      "grad_norm": 0.42258235812187195,
+      "learning_rate": 0.00022848366102631675,
+      "loss": 2.7931,
+      "step": 5408
+    },
+    {
+      "epoch": 0.692352,
+      "grad_norm": 0.39211612939834595,
+      "learning_rate": 0.00022830960282816515,
+      "loss": 2.8412,
+      "step": 5409
+    },
+    {
+      "epoch": 0.69248,
+      "grad_norm": 0.430715948343277,
+      "learning_rate": 0.00022813559133674278,
+      "loss": 2.7913,
+      "step": 5410
+    },
+    {
+      "epoch": 0.692608,
+      "grad_norm": 0.40293997526168823,
+      "learning_rate": 0.00022796162658196396,
+      "loss": 2.9339,
+      "step": 5411
+    },
+    {
+      "epoch": 0.692736,
+      "grad_norm": 0.37875792384147644,
+      "learning_rate": 0.00022778770859373503,
+      "loss": 2.679,
+      "step": 5412
+    },
+    {
+      "epoch": 0.692864,
+      "grad_norm": 0.4213627278804779,
+      "learning_rate": 0.000227613837401955,
+      "loss": 2.8438,
+      "step": 5413
+    },
+    {
+      "epoch": 0.692992,
+      "grad_norm": 0.4310801327228546,
+      "learning_rate": 0.00022744001303651407,
+      "loss": 2.8854,
+      "step": 5414
+    },
+    {
+      "epoch": 0.69312,
+      "grad_norm": 0.42655590176582336,
+      "learning_rate": 0.0002272662355272947,
+      "loss": 2.8057,
+      "step": 5415
+    },
+    {
+      "epoch": 0.693248,
+      "grad_norm": 0.3859892785549164,
+      "learning_rate": 0.0002270925049041712,
+      "loss": 2.8495,
+      "step": 5416
+    },
+    {
+      "epoch": 0.693376,
+      "grad_norm": 0.39362165331840515,
+      "learning_rate": 0.0002269188211970098,
+      "loss": 2.797,
+      "step": 5417
+    },
+    {
+      "epoch": 0.693504,
+      "grad_norm": 0.4246830344200134,
+      "learning_rate": 0.00022674518443566906,
+      "loss": 2.8667,
+      "step": 5418
+    },
+    {
+      "epoch": 0.693632,
+      "grad_norm": 0.41544005274772644,
+      "learning_rate": 0.00022657159464999888,
+      "loss": 2.7817,
+      "step": 5419
+    },
+    {
+      "epoch": 0.69376,
+      "grad_norm": 0.396819144487381,
+      "learning_rate": 0.00022639805186984148,
+      "loss": 2.8228,
+      "step": 5420
+    },
+    {
+      "epoch": 0.693888,
+      "grad_norm": 0.39982596039772034,
+      "learning_rate": 0.00022622455612503062,
+      "loss": 2.7302,
+      "step": 5421
+    },
+    {
+      "epoch": 0.694016,
+      "grad_norm": 0.382228821516037,
+      "learning_rate": 0.00022605110744539265,
+      "loss": 2.8001,
+      "step": 5422
+    },
+    {
+      "epoch": 0.694144,
+      "grad_norm": 0.38304832577705383,
+      "learning_rate": 0.00022587770586074519,
+      "loss": 2.9125,
+      "step": 5423
+    },
+    {
+      "epoch": 0.694272,
+      "grad_norm": 0.38940298557281494,
+      "learning_rate": 0.0002257043514008978,
+      "loss": 2.6794,
+      "step": 5424
+    },
+    {
+      "epoch": 0.6944,
+      "grad_norm": 0.39676252007484436,
+      "learning_rate": 0.0002255310440956525,
+      "loss": 2.9315,
+      "step": 5425
+    },
+    {
+      "epoch": 0.694528,
+      "grad_norm": 0.3741582930088043,
+      "learning_rate": 0.00022535778397480262,
+      "loss": 2.7915,
+      "step": 5426
+    },
+    {
+      "epoch": 0.694656,
+      "grad_norm": 0.38373124599456787,
+      "learning_rate": 0.0002251845710681336,
+      "loss": 2.7711,
+      "step": 5427
+    },
+    {
+      "epoch": 0.694784,
+      "grad_norm": 0.3917064666748047,
+      "learning_rate": 0.0002250114054054228,
+      "loss": 2.5723,
+      "step": 5428
+    },
+    {
+      "epoch": 0.694912,
+      "grad_norm": 0.41761231422424316,
+      "learning_rate": 0.00022483828701643933,
+      "loss": 2.9133,
+      "step": 5429
+    },
+    {
+      "epoch": 0.69504,
+      "grad_norm": 0.4308297634124756,
+      "learning_rate": 0.00022466521593094403,
+      "loss": 2.8304,
+      "step": 5430
+    },
+    {
+      "epoch": 0.695168,
+      "grad_norm": 0.3990941643714905,
+      "learning_rate": 0.0002244921921786901,
+      "loss": 2.7902,
+      "step": 5431
+    },
+    {
+      "epoch": 0.695296,
+      "grad_norm": 0.3984021842479706,
+      "learning_rate": 0.0002243192157894225,
+      "loss": 2.7067,
+      "step": 5432
+    },
+    {
+      "epoch": 0.695424,
+      "grad_norm": 0.4558500647544861,
+      "learning_rate": 0.00022414628679287769,
+      "loss": 2.9559,
+      "step": 5433
+    },
+    {
+      "epoch": 0.695552,
+      "grad_norm": 0.3850252330303192,
+      "learning_rate": 0.00022397340521878412,
+      "loss": 2.7728,
+      "step": 5434
+    },
+    {
+      "epoch": 0.69568,
+      "grad_norm": 0.3642334043979645,
+      "learning_rate": 0.00022380057109686215,
+      "loss": 2.7805,
+      "step": 5435
+    },
+    {
+      "epoch": 0.695808,
+      "grad_norm": 0.41980767250061035,
+      "learning_rate": 0.00022362778445682402,
+      "loss": 2.8616,
+      "step": 5436
+    },
+    {
+      "epoch": 0.695936,
+      "grad_norm": 0.4095541536808014,
+      "learning_rate": 0.00022345504532837368,
+      "loss": 2.8579,
+      "step": 5437
+    },
+    {
+      "epoch": 0.696064,
+      "grad_norm": 0.41860923171043396,
+      "learning_rate": 0.00022328235374120692,
+      "loss": 2.8564,
+      "step": 5438
+    },
+    {
+      "epoch": 0.696192,
+      "grad_norm": 0.38635218143463135,
+      "learning_rate": 0.0002231097097250115,
+      "loss": 2.7155,
+      "step": 5439
+    },
+    {
+      "epoch": 0.69632,
+      "grad_norm": 0.39118245244026184,
+      "learning_rate": 0.0002229371133094671,
+      "loss": 2.8965,
+      "step": 5440
+    },
+    {
+      "epoch": 0.696448,
+      "grad_norm": 0.4207175672054291,
+      "learning_rate": 0.00022276456452424497,
+      "loss": 2.9096,
+      "step": 5441
+    },
+    {
+      "epoch": 0.696576,
+      "grad_norm": 0.3989034593105316,
+      "learning_rate": 0.00022259206339900818,
+      "loss": 2.9229,
+      "step": 5442
+    },
+    {
+      "epoch": 0.696704,
+      "grad_norm": 0.4392186999320984,
+      "learning_rate": 0.00022241960996341166,
+      "loss": 2.896,
+      "step": 5443
+    },
+    {
+      "epoch": 0.696832,
+      "grad_norm": 0.3947724401950836,
+      "learning_rate": 0.0002222472042471022,
+      "loss": 2.9009,
+      "step": 5444
+    },
+    {
+      "epoch": 0.69696,
+      "grad_norm": 0.39487138390541077,
+      "learning_rate": 0.00022207484627971813,
+      "loss": 2.8829,
+      "step": 5445
+    },
+    {
+      "epoch": 0.697088,
+      "grad_norm": 0.3766847848892212,
+      "learning_rate": 0.0002219025360908901,
+      "loss": 2.6857,
+      "step": 5446
+    },
+    {
+      "epoch": 0.697216,
+      "grad_norm": 0.39538708329200745,
+      "learning_rate": 0.0002217302737102402,
+      "loss": 3.0598,
+      "step": 5447
+    },
+    {
+      "epoch": 0.697344,
+      "grad_norm": 0.40650659799575806,
+      "learning_rate": 0.00022155805916738213,
+      "loss": 2.8042,
+      "step": 5448
+    },
+    {
+      "epoch": 0.697472,
+      "grad_norm": 0.4267335534095764,
+      "learning_rate": 0.00022138589249192154,
+      "loss": 2.7568,
+      "step": 5449
+    },
+    {
+      "epoch": 0.6976,
+      "grad_norm": 0.4069296717643738,
+      "learning_rate": 0.0002212137737134562,
+      "loss": 2.842,
+      "step": 5450
+    },
+    {
+      "epoch": 0.697728,
+      "grad_norm": 0.38173937797546387,
+      "learning_rate": 0.00022104170286157504,
+      "loss": 2.841,
+      "step": 5451
+    },
+    {
+      "epoch": 0.697856,
+      "grad_norm": 0.39820408821105957,
+      "learning_rate": 0.00022086967996585898,
+      "loss": 2.7613,
+      "step": 5452
+    },
+    {
+      "epoch": 0.697984,
+      "grad_norm": 0.4072699546813965,
+      "learning_rate": 0.00022069770505588111,
+      "loss": 2.7093,
+      "step": 5453
+    },
+    {
+      "epoch": 0.698112,
+      "grad_norm": 0.40188413858413696,
+      "learning_rate": 0.00022052577816120567,
+      "loss": 2.7,
+      "step": 5454
+    },
+    {
+      "epoch": 0.69824,
+      "grad_norm": 0.4097534120082855,
+      "learning_rate": 0.00022035389931138893,
+      "loss": 2.8278,
+      "step": 5455
+    },
+    {
+      "epoch": 0.698368,
+      "grad_norm": 0.39850521087646484,
+      "learning_rate": 0.00022018206853597879,
+      "loss": 2.8647,
+      "step": 5456
+    },
+    {
+      "epoch": 0.698496,
+      "grad_norm": 0.3930126428604126,
+      "learning_rate": 0.000220010285864515,
+      "loss": 2.8662,
+      "step": 5457
+    },
+    {
+      "epoch": 0.698624,
+      "grad_norm": 0.4128915071487427,
+      "learning_rate": 0.00021983855132652887,
+      "loss": 2.7453,
+      "step": 5458
+    },
+    {
+      "epoch": 0.698752,
+      "grad_norm": 0.42025744915008545,
+      "learning_rate": 0.00021966686495154375,
+      "loss": 2.7916,
+      "step": 5459
+    },
+    {
+      "epoch": 0.69888,
+      "grad_norm": 0.3972541093826294,
+      "learning_rate": 0.00021949522676907436,
+      "loss": 2.7142,
+      "step": 5460
+    },
+    {
+      "epoch": 0.699008,
+      "grad_norm": 0.39933469891548157,
+      "learning_rate": 0.0002193236368086275,
+      "loss": 2.8028,
+      "step": 5461
+    },
+    {
+      "epoch": 0.699136,
+      "grad_norm": 0.40930604934692383,
+      "learning_rate": 0.0002191520950997014,
+      "loss": 2.8338,
+      "step": 5462
+    },
+    {
+      "epoch": 0.699264,
+      "grad_norm": 0.40863236784935,
+      "learning_rate": 0.000218980601671786,
+      "loss": 2.8133,
+      "step": 5463
+    },
+    {
+      "epoch": 0.699392,
+      "grad_norm": 0.39579638838768005,
+      "learning_rate": 0.00021880915655436302,
+      "loss": 2.8358,
+      "step": 5464
+    },
+    {
+      "epoch": 0.69952,
+      "grad_norm": 0.39159858226776123,
+      "learning_rate": 0.00021863775977690587,
+      "loss": 2.8346,
+      "step": 5465
+    },
+    {
+      "epoch": 0.699648,
+      "grad_norm": 0.4052850604057312,
+      "learning_rate": 0.00021846641136887945,
+      "loss": 2.7736,
+      "step": 5466
+    },
+    {
+      "epoch": 0.699776,
+      "grad_norm": 0.39821210503578186,
+      "learning_rate": 0.0002182951113597408,
+      "loss": 2.8354,
+      "step": 5467
+    },
+    {
+      "epoch": 0.699904,
+      "grad_norm": 0.41811227798461914,
+      "learning_rate": 0.00021812385977893845,
+      "loss": 2.8826,
+      "step": 5468
+    },
+    {
+      "epoch": 0.700032,
+      "grad_norm": 0.4142155051231384,
+      "learning_rate": 0.00021795265665591236,
+      "loss": 2.7768,
+      "step": 5469
+    },
+    {
+      "epoch": 0.70016,
+      "grad_norm": 0.3872862756252289,
+      "learning_rate": 0.0002177815020200944,
+      "loss": 2.7988,
+      "step": 5470
+    },
+    {
+      "epoch": 0.700288,
+      "grad_norm": 0.3802974820137024,
+      "learning_rate": 0.00021761039590090798,
+      "loss": 2.6934,
+      "step": 5471
+    },
+    {
+      "epoch": 0.700416,
+      "grad_norm": 0.37630292773246765,
+      "learning_rate": 0.0002174393383277683,
+      "loss": 2.7851,
+      "step": 5472
+    },
+    {
+      "epoch": 0.700544,
+      "grad_norm": 0.3792800307273865,
+      "learning_rate": 0.00021726832933008185,
+      "loss": 2.8571,
+      "step": 5473
+    },
+    {
+      "epoch": 0.700672,
+      "grad_norm": 0.4158037602901459,
+      "learning_rate": 0.00021709736893724745,
+      "loss": 2.9326,
+      "step": 5474
+    },
+    {
+      "epoch": 0.7008,
+      "grad_norm": 0.39698871970176697,
+      "learning_rate": 0.0002169264571786551,
+      "loss": 2.75,
+      "step": 5475
+    },
+    {
+      "epoch": 0.700928,
+      "grad_norm": 0.39466559886932373,
+      "learning_rate": 0.00021675559408368622,
+      "loss": 2.8369,
+      "step": 5476
+    },
+    {
+      "epoch": 0.701056,
+      "grad_norm": 0.4118227958679199,
+      "learning_rate": 0.00021658477968171453,
+      "loss": 2.8361,
+      "step": 5477
+    },
+    {
+      "epoch": 0.701184,
+      "grad_norm": 0.4156016707420349,
+      "learning_rate": 0.00021641401400210487,
+      "loss": 2.8326,
+      "step": 5478
+    },
+    {
+      "epoch": 0.701312,
+      "grad_norm": 0.3871689736843109,
+      "learning_rate": 0.00021624329707421375,
+      "loss": 2.7773,
+      "step": 5479
+    },
+    {
+      "epoch": 0.70144,
+      "grad_norm": 0.4321022629737854,
+      "learning_rate": 0.00021607262892738954,
+      "loss": 2.8336,
+      "step": 5480
+    },
+    {
+      "epoch": 0.701568,
+      "grad_norm": 0.3735867440700531,
+      "learning_rate": 0.00021590200959097173,
+      "loss": 2.7926,
+      "step": 5481
+    },
+    {
+      "epoch": 0.701696,
+      "grad_norm": 0.3938910961151123,
+      "learning_rate": 0.00021573143909429227,
+      "loss": 2.7759,
+      "step": 5482
+    },
+    {
+      "epoch": 0.701824,
+      "grad_norm": 0.38795652985572815,
+      "learning_rate": 0.0002155609174666739,
+      "loss": 2.8167,
+      "step": 5483
+    },
+    {
+      "epoch": 0.701952,
+      "grad_norm": 0.4103357493877411,
+      "learning_rate": 0.00021539044473743135,
+      "loss": 2.7174,
+      "step": 5484
+    },
+    {
+      "epoch": 0.70208,
+      "grad_norm": 0.4209286570549011,
+      "learning_rate": 0.00021522002093587078,
+      "loss": 2.9491,
+      "step": 5485
+    },
+    {
+      "epoch": 0.702208,
+      "grad_norm": 0.4094254672527313,
+      "learning_rate": 0.0002150496460912899,
+      "loss": 2.7778,
+      "step": 5486
+    },
+    {
+      "epoch": 0.702336,
+      "grad_norm": 0.38650429248809814,
+      "learning_rate": 0.00021487932023297852,
+      "loss": 2.7262,
+      "step": 5487
+    },
+    {
+      "epoch": 0.702464,
+      "grad_norm": 0.4114895761013031,
+      "learning_rate": 0.0002147090433902172,
+      "loss": 2.8379,
+      "step": 5488
+    },
+    {
+      "epoch": 0.702592,
+      "grad_norm": 0.3874468207359314,
+      "learning_rate": 0.00021453881559227884,
+      "loss": 2.7332,
+      "step": 5489
+    },
+    {
+      "epoch": 0.70272,
+      "grad_norm": 0.40873461961746216,
+      "learning_rate": 0.0002143686368684274,
+      "loss": 2.8085,
+      "step": 5490
+    },
+    {
+      "epoch": 0.702848,
+      "grad_norm": 0.389877051115036,
+      "learning_rate": 0.00021419850724791866,
+      "loss": 2.7057,
+      "step": 5491
+    },
+    {
+      "epoch": 0.702976,
+      "grad_norm": 0.4019429087638855,
+      "learning_rate": 0.0002140284267599998,
+      "loss": 2.8314,
+      "step": 5492
+    },
+    {
+      "epoch": 0.703104,
+      "grad_norm": 0.42326703667640686,
+      "learning_rate": 0.00021385839543390967,
+      "loss": 2.7641,
+      "step": 5493
+    },
+    {
+      "epoch": 0.703232,
+      "grad_norm": 0.44661328196525574,
+      "learning_rate": 0.0002136884132988784,
+      "loss": 2.8294,
+      "step": 5494
+    },
+    {
+      "epoch": 0.70336,
+      "grad_norm": 0.4235101640224457,
+      "learning_rate": 0.0002135184803841283,
+      "loss": 2.8252,
+      "step": 5495
+    },
+    {
+      "epoch": 0.703488,
+      "grad_norm": 0.39270707964897156,
+      "learning_rate": 0.00021334859671887234,
+      "loss": 2.9113,
+      "step": 5496
+    },
+    {
+      "epoch": 0.703616,
+      "grad_norm": 0.4228587746620178,
+      "learning_rate": 0.00021317876233231605,
+      "loss": 2.8017,
+      "step": 5497
+    },
+    {
+      "epoch": 0.703744,
+      "grad_norm": 0.3750213086605072,
+      "learning_rate": 0.00021300897725365554,
+      "loss": 2.8876,
+      "step": 5498
+    },
+    {
+      "epoch": 0.703872,
+      "grad_norm": 0.40063706040382385,
+      "learning_rate": 0.00021283924151207896,
+      "loss": 2.7337,
+      "step": 5499
+    },
+    {
+      "epoch": 0.704,
+      "grad_norm": 0.42321836948394775,
+      "learning_rate": 0.00021266955513676584,
+      "loss": 2.7331,
+      "step": 5500
+    },
+    {
+      "epoch": 0.704128,
+      "grad_norm": 0.41188952326774597,
+      "learning_rate": 0.0002124999181568872,
+      "loss": 2.8988,
+      "step": 5501
+    },
+    {
+      "epoch": 0.704256,
+      "grad_norm": 0.38546013832092285,
+      "learning_rate": 0.00021233033060160539,
+      "loss": 2.8281,
+      "step": 5502
+    },
+    {
+      "epoch": 0.704384,
+      "grad_norm": 0.38854676485061646,
+      "learning_rate": 0.00021216079250007492,
+      "loss": 2.8129,
+      "step": 5503
+    },
+    {
+      "epoch": 0.704512,
+      "grad_norm": 0.409949392080307,
+      "learning_rate": 0.00021199130388144095,
+      "loss": 2.934,
+      "step": 5504
+    },
+    {
+      "epoch": 0.70464,
+      "grad_norm": 0.4060867130756378,
+      "learning_rate": 0.00021182186477484093,
+      "loss": 2.7516,
+      "step": 5505
+    },
+    {
+      "epoch": 0.704768,
+      "grad_norm": 0.39702391624450684,
+      "learning_rate": 0.00021165247520940319,
+      "loss": 2.8894,
+      "step": 5506
+    },
+    {
+      "epoch": 0.704896,
+      "grad_norm": 0.39972731471061707,
+      "learning_rate": 0.00021148313521424778,
+      "loss": 2.7568,
+      "step": 5507
+    },
+    {
+      "epoch": 0.705024,
+      "grad_norm": 0.40539753437042236,
+      "learning_rate": 0.00021131384481848614,
+      "loss": 2.8509,
+      "step": 5508
+    },
+    {
+      "epoch": 0.705152,
+      "grad_norm": 0.4030393362045288,
+      "learning_rate": 0.0002111446040512212,
+      "loss": 2.7623,
+      "step": 5509
+    },
+    {
+      "epoch": 0.70528,
+      "grad_norm": 0.3873591125011444,
+      "learning_rate": 0.0002109754129415477,
+      "loss": 2.7489,
+      "step": 5510
+    },
+    {
+      "epoch": 0.705408,
+      "grad_norm": 0.420719712972641,
+      "learning_rate": 0.00021080627151855136,
+      "loss": 2.8454,
+      "step": 5511
+    },
+    {
+      "epoch": 0.705536,
+      "grad_norm": 0.3756382465362549,
+      "learning_rate": 0.00021063717981130954,
+      "loss": 2.7718,
+      "step": 5512
+    },
+    {
+      "epoch": 0.705664,
+      "grad_norm": 0.43856924772262573,
+      "learning_rate": 0.00021046813784889097,
+      "loss": 2.8238,
+      "step": 5513
+    },
+    {
+      "epoch": 0.705792,
+      "grad_norm": 0.38232898712158203,
+      "learning_rate": 0.00021029914566035617,
+      "loss": 2.6576,
+      "step": 5514
+    },
+    {
+      "epoch": 0.70592,
+      "grad_norm": 0.39326778054237366,
+      "learning_rate": 0.0002101302032747568,
+      "loss": 2.8309,
+      "step": 5515
+    },
+    {
+      "epoch": 0.706048,
+      "grad_norm": 0.3917228579521179,
+      "learning_rate": 0.00020996131072113588,
+      "loss": 2.9297,
+      "step": 5516
+    },
+    {
+      "epoch": 0.706176,
+      "grad_norm": 0.37779736518859863,
+      "learning_rate": 0.00020979246802852792,
+      "loss": 2.864,
+      "step": 5517
+    },
+    {
+      "epoch": 0.706304,
+      "grad_norm": 0.3978080451488495,
+      "learning_rate": 0.00020962367522595922,
+      "loss": 2.8321,
+      "step": 5518
+    },
+    {
+      "epoch": 0.706432,
+      "grad_norm": 0.4211970269680023,
+      "learning_rate": 0.00020945493234244712,
+      "loss": 2.7965,
+      "step": 5519
+    },
+    {
+      "epoch": 0.70656,
+      "grad_norm": 0.39318329095840454,
+      "learning_rate": 0.00020928623940700042,
+      "loss": 2.5964,
+      "step": 5520
+    },
+    {
+      "epoch": 0.706688,
+      "grad_norm": 0.4214593768119812,
+      "learning_rate": 0.00020911759644861937,
+      "loss": 2.875,
+      "step": 5521
+    },
+    {
+      "epoch": 0.706816,
+      "grad_norm": 0.4137212038040161,
+      "learning_rate": 0.00020894900349629575,
+      "loss": 2.8422,
+      "step": 5522
+    },
+    {
+      "epoch": 0.706944,
+      "grad_norm": 0.38349729776382446,
+      "learning_rate": 0.0002087804605790124,
+      "loss": 2.7436,
+      "step": 5523
+    },
+    {
+      "epoch": 0.707072,
+      "grad_norm": 0.39749884605407715,
+      "learning_rate": 0.00020861196772574398,
+      "loss": 2.6989,
+      "step": 5524
+    },
+    {
+      "epoch": 0.7072,
+      "grad_norm": 0.44051942229270935,
+      "learning_rate": 0.0002084435249654565,
+      "loss": 2.7696,
+      "step": 5525
+    },
+    {
+      "epoch": 0.707328,
+      "grad_norm": 0.41579487919807434,
+      "learning_rate": 0.00020827513232710715,
+      "loss": 2.8868,
+      "step": 5526
+    },
+    {
+      "epoch": 0.707456,
+      "grad_norm": 0.4090573191642761,
+      "learning_rate": 0.00020810678983964448,
+      "loss": 2.7782,
+      "step": 5527
+    },
+    {
+      "epoch": 0.707584,
+      "grad_norm": 0.3744186758995056,
+      "learning_rate": 0.00020793849753200855,
+      "loss": 2.9346,
+      "step": 5528
+    },
+    {
+      "epoch": 0.707712,
+      "grad_norm": 0.4099600613117218,
+      "learning_rate": 0.00020777025543313078,
+      "loss": 2.8081,
+      "step": 5529
+    },
+    {
+      "epoch": 0.70784,
+      "grad_norm": 0.4382067322731018,
+      "learning_rate": 0.00020760206357193372,
+      "loss": 2.873,
+      "step": 5530
+    },
+    {
+      "epoch": 0.707968,
+      "grad_norm": 0.4651094973087311,
+      "learning_rate": 0.00020743392197733191,
+      "loss": 2.8902,
+      "step": 5531
+    },
+    {
+      "epoch": 0.708096,
+      "grad_norm": 0.4139474928379059,
+      "learning_rate": 0.00020726583067823045,
+      "loss": 2.8359,
+      "step": 5532
+    },
+    {
+      "epoch": 0.708224,
+      "grad_norm": 0.39393213391304016,
+      "learning_rate": 0.00020709778970352656,
+      "loss": 2.8209,
+      "step": 5533
+    },
+    {
+      "epoch": 0.708352,
+      "grad_norm": 0.4301656186580658,
+      "learning_rate": 0.0002069297990821082,
+      "loss": 2.7742,
+      "step": 5534
+    },
+    {
+      "epoch": 0.70848,
+      "grad_norm": 0.42110753059387207,
+      "learning_rate": 0.00020676185884285493,
+      "loss": 2.918,
+      "step": 5535
+    },
+    {
+      "epoch": 0.708608,
+      "grad_norm": 0.39460355043411255,
+      "learning_rate": 0.0002065939690146376,
+      "loss": 2.8027,
+      "step": 5536
+    },
+    {
+      "epoch": 0.708736,
+      "grad_norm": 0.415504515171051,
+      "learning_rate": 0.00020642612962631853,
+      "loss": 2.8174,
+      "step": 5537
+    },
+    {
+      "epoch": 0.708864,
+      "grad_norm": 0.4153333902359009,
+      "learning_rate": 0.00020625834070675091,
+      "loss": 2.8055,
+      "step": 5538
+    },
+    {
+      "epoch": 0.708992,
+      "grad_norm": 0.3889099955558777,
+      "learning_rate": 0.00020609060228478015,
+      "loss": 2.7815,
+      "step": 5539
+    },
+    {
+      "epoch": 0.70912,
+      "grad_norm": 0.4242640733718872,
+      "learning_rate": 0.0002059229143892421,
+      "loss": 2.778,
+      "step": 5540
+    },
+    {
+      "epoch": 0.709248,
+      "grad_norm": 0.40455204248428345,
+      "learning_rate": 0.00020575527704896413,
+      "loss": 2.8513,
+      "step": 5541
+    },
+    {
+      "epoch": 0.709376,
+      "grad_norm": 0.3910122811794281,
+      "learning_rate": 0.00020558769029276535,
+      "loss": 2.7511,
+      "step": 5542
+    },
+    {
+      "epoch": 0.709504,
+      "grad_norm": 0.44274359941482544,
+      "learning_rate": 0.00020542015414945575,
+      "loss": 2.8631,
+      "step": 5543
+    },
+    {
+      "epoch": 0.709632,
+      "grad_norm": 0.3944084942340851,
+      "learning_rate": 0.00020525266864783676,
+      "loss": 2.7713,
+      "step": 5544
+    },
+    {
+      "epoch": 0.70976,
+      "grad_norm": 0.396205872297287,
+      "learning_rate": 0.0002050852338167008,
+      "loss": 2.8457,
+      "step": 5545
+    },
+    {
+      "epoch": 0.709888,
+      "grad_norm": 0.4138065576553345,
+      "learning_rate": 0.0002049178496848323,
+      "loss": 2.8852,
+      "step": 5546
+    },
+    {
+      "epoch": 0.710016,
+      "grad_norm": 0.4290752112865448,
+      "learning_rate": 0.00020475051628100633,
+      "loss": 2.7358,
+      "step": 5547
+    },
+    {
+      "epoch": 0.710144,
+      "grad_norm": 0.42084041237831116,
+      "learning_rate": 0.0002045832336339894,
+      "loss": 2.7392,
+      "step": 5548
+    },
+    {
+      "epoch": 0.710272,
+      "grad_norm": 0.4265095591545105,
+      "learning_rate": 0.00020441600177253938,
+      "loss": 2.8034,
+      "step": 5549
+    },
+    {
+      "epoch": 0.7104,
+      "grad_norm": 0.4011356830596924,
+      "learning_rate": 0.00020424882072540542,
+      "loss": 2.8337,
+      "step": 5550
+    },
+    {
+      "epoch": 0.710528,
+      "grad_norm": 0.431243896484375,
+      "learning_rate": 0.0002040816905213276,
+      "loss": 2.8502,
+      "step": 5551
+    },
+    {
+      "epoch": 0.710656,
+      "grad_norm": 0.42700234055519104,
+      "learning_rate": 0.00020391461118903786,
+      "loss": 2.771,
+      "step": 5552
+    },
+    {
+      "epoch": 0.710784,
+      "grad_norm": 0.37832438945770264,
+      "learning_rate": 0.0002037475827572589,
+      "loss": 2.8045,
+      "step": 5553
+    },
+    {
+      "epoch": 0.710912,
+      "grad_norm": 0.43035948276519775,
+      "learning_rate": 0.00020358060525470508,
+      "loss": 2.8235,
+      "step": 5554
+    },
+    {
+      "epoch": 0.71104,
+      "grad_norm": 0.4198765456676483,
+      "learning_rate": 0.00020341367871008154,
+      "loss": 2.8492,
+      "step": 5555
+    },
+    {
+      "epoch": 0.711168,
+      "grad_norm": 0.4173860251903534,
+      "learning_rate": 0.00020324680315208504,
+      "loss": 2.7722,
+      "step": 5556
+    },
+    {
+      "epoch": 0.711296,
+      "grad_norm": 0.4146764278411865,
+      "learning_rate": 0.0002030799786094033,
+      "loss": 2.7353,
+      "step": 5557
+    },
+    {
+      "epoch": 0.711424,
+      "grad_norm": 0.40512245893478394,
+      "learning_rate": 0.00020291320511071543,
+      "loss": 2.8861,
+      "step": 5558
+    },
+    {
+      "epoch": 0.711552,
+      "grad_norm": 0.36593788862228394,
+      "learning_rate": 0.00020274648268469155,
+      "loss": 2.8596,
+      "step": 5559
+    },
+    {
+      "epoch": 0.71168,
+      "grad_norm": 0.41018587350845337,
+      "learning_rate": 0.0002025798113599933,
+      "loss": 2.7674,
+      "step": 5560
+    },
+    {
+      "epoch": 0.711808,
+      "grad_norm": 0.39910927414894104,
+      "learning_rate": 0.00020241319116527374,
+      "loss": 2.8686,
+      "step": 5561
+    },
+    {
+      "epoch": 0.711936,
+      "grad_norm": 0.4076094627380371,
+      "learning_rate": 0.0002022466221291765,
+      "loss": 2.8925,
+      "step": 5562
+    },
+    {
+      "epoch": 0.712064,
+      "grad_norm": 0.4171661138534546,
+      "learning_rate": 0.00020208010428033674,
+      "loss": 2.7647,
+      "step": 5563
+    },
+    {
+      "epoch": 0.712192,
+      "grad_norm": 0.3852541744709015,
+      "learning_rate": 0.00020191363764738085,
+      "loss": 2.9246,
+      "step": 5564
+    },
+    {
+      "epoch": 0.71232,
+      "grad_norm": 0.4244108200073242,
+      "learning_rate": 0.00020174722225892638,
+      "loss": 2.7602,
+      "step": 5565
+    },
+    {
+      "epoch": 0.712448,
+      "grad_norm": 0.41147780418395996,
+      "learning_rate": 0.00020158085814358183,
+      "loss": 2.8337,
+      "step": 5566
+    },
+    {
+      "epoch": 0.712576,
+      "grad_norm": 0.4200011193752289,
+      "learning_rate": 0.0002014145453299476,
+      "loss": 2.845,
+      "step": 5567
+    },
+    {
+      "epoch": 0.712704,
+      "grad_norm": 0.38205811381340027,
+      "learning_rate": 0.0002012482838466145,
+      "loss": 2.816,
+      "step": 5568
+    },
+    {
+      "epoch": 0.712832,
+      "grad_norm": 0.3796984553337097,
+      "learning_rate": 0.00020108207372216457,
+      "loss": 2.9086,
+      "step": 5569
+    },
+    {
+      "epoch": 0.71296,
+      "grad_norm": 0.3922773003578186,
+      "learning_rate": 0.0002009159149851718,
+      "loss": 2.7869,
+      "step": 5570
+    },
+    {
+      "epoch": 0.713088,
+      "grad_norm": 0.40577468276023865,
+      "learning_rate": 0.00020074980766420048,
+      "loss": 2.8751,
+      "step": 5571
+    },
+    {
+      "epoch": 0.713216,
+      "grad_norm": 0.4266822636127472,
+      "learning_rate": 0.0002005837517878064,
+      "loss": 2.9376,
+      "step": 5572
+    },
+    {
+      "epoch": 0.713344,
+      "grad_norm": 0.3831930458545685,
+      "learning_rate": 0.0002004177473845366,
+      "loss": 2.823,
+      "step": 5573
+    },
+    {
+      "epoch": 0.713472,
+      "grad_norm": 0.40374428033828735,
+      "learning_rate": 0.00020025179448292886,
+      "loss": 2.8037,
+      "step": 5574
+    },
+    {
+      "epoch": 0.7136,
+      "grad_norm": 0.35885000228881836,
+      "learning_rate": 0.00020008589311151277,
+      "loss": 2.8283,
+      "step": 5575
+    },
+    {
+      "epoch": 0.713728,
+      "grad_norm": 0.6052151918411255,
+      "learning_rate": 0.00019992004329880857,
+      "loss": 2.8713,
+      "step": 5576
+    },
+    {
+      "epoch": 0.713856,
+      "grad_norm": 0.3735436201095581,
+      "learning_rate": 0.0001997542450733278,
+      "loss": 2.8306,
+      "step": 5577
+    },
+    {
+      "epoch": 0.713984,
+      "grad_norm": 0.38978826999664307,
+      "learning_rate": 0.00019958849846357285,
+      "loss": 2.7928,
+      "step": 5578
+    },
+    {
+      "epoch": 0.714112,
+      "grad_norm": 0.3864547312259674,
+      "learning_rate": 0.00019942280349803781,
+      "loss": 2.7148,
+      "step": 5579
+    },
+    {
+      "epoch": 0.71424,
+      "grad_norm": 0.39002901315689087,
+      "learning_rate": 0.00019925716020520752,
+      "loss": 2.8076,
+      "step": 5580
+    },
+    {
+      "epoch": 0.714368,
+      "grad_norm": 0.40254443883895874,
+      "learning_rate": 0.00019909156861355764,
+      "loss": 2.8557,
+      "step": 5581
+    },
+    {
+      "epoch": 0.714496,
+      "grad_norm": 0.4264690577983856,
+      "learning_rate": 0.0001989260287515558,
+      "loss": 2.8902,
+      "step": 5582
+    },
+    {
+      "epoch": 0.714624,
+      "grad_norm": 0.40972042083740234,
+      "learning_rate": 0.00019876054064765992,
+      "loss": 2.9533,
+      "step": 5583
+    },
+    {
+      "epoch": 0.714752,
+      "grad_norm": 0.43212366104125977,
+      "learning_rate": 0.0001985951043303194,
+      "loss": 2.8476,
+      "step": 5584
+    },
+    {
+      "epoch": 0.71488,
+      "grad_norm": 0.4348321557044983,
+      "learning_rate": 0.0001984297198279746,
+      "loss": 2.8031,
+      "step": 5585
+    },
+    {
+      "epoch": 0.715008,
+      "grad_norm": 0.39388126134872437,
+      "learning_rate": 0.0001982643871690571,
+      "loss": 2.8001,
+      "step": 5586
+    },
+    {
+      "epoch": 0.715136,
+      "grad_norm": 0.386217325925827,
+      "learning_rate": 0.00019809910638198924,
+      "loss": 2.6986,
+      "step": 5587
+    },
+    {
+      "epoch": 0.715264,
+      "grad_norm": 0.3903806507587433,
+      "learning_rate": 0.00019793387749518515,
+      "loss": 2.8801,
+      "step": 5588
+    },
+    {
+      "epoch": 0.715392,
+      "grad_norm": 0.43303659558296204,
+      "learning_rate": 0.00019776870053704915,
+      "loss": 2.7507,
+      "step": 5589
+    },
+    {
+      "epoch": 0.71552,
+      "grad_norm": 0.39700597524642944,
+      "learning_rate": 0.00019760357553597752,
+      "loss": 2.8063,
+      "step": 5590
+    },
+    {
+      "epoch": 0.715648,
+      "grad_norm": 0.42200082540512085,
+      "learning_rate": 0.0001974385025203569,
+      "loss": 2.7157,
+      "step": 5591
+    },
+    {
+      "epoch": 0.715776,
+      "grad_norm": 0.4264301061630249,
+      "learning_rate": 0.00019727348151856534,
+      "loss": 2.7545,
+      "step": 5592
+    },
+    {
+      "epoch": 0.715904,
+      "grad_norm": 0.4754263162612915,
+      "learning_rate": 0.00019710851255897173,
+      "loss": 2.9545,
+      "step": 5593
+    },
+    {
+      "epoch": 0.716032,
+      "grad_norm": 0.43189117312431335,
+      "learning_rate": 0.0001969435956699363,
+      "loss": 2.7738,
+      "step": 5594
+    },
+    {
+      "epoch": 0.71616,
+      "grad_norm": 0.39737826585769653,
+      "learning_rate": 0.0001967787308798099,
+      "loss": 2.847,
+      "step": 5595
+    },
+    {
+      "epoch": 0.716288,
+      "grad_norm": 0.39610207080841064,
+      "learning_rate": 0.00019661391821693507,
+      "loss": 2.7902,
+      "step": 5596
+    },
+    {
+      "epoch": 0.716416,
+      "grad_norm": 0.40401822328567505,
+      "learning_rate": 0.00019644915770964473,
+      "loss": 2.8609,
+      "step": 5597
+    },
+    {
+      "epoch": 0.716544,
+      "grad_norm": 0.4139217734336853,
+      "learning_rate": 0.00019628444938626333,
+      "loss": 2.8031,
+      "step": 5598
+    },
+    {
+      "epoch": 0.716672,
+      "grad_norm": 0.36782482266426086,
+      "learning_rate": 0.00019611979327510616,
+      "loss": 2.9363,
+      "step": 5599
+    },
+    {
+      "epoch": 0.7168,
+      "grad_norm": 0.42088085412979126,
+      "learning_rate": 0.00019595518940447931,
+      "loss": 2.8356,
+      "step": 5600
+    },
+    {
+      "epoch": 0.716928,
+      "grad_norm": 0.41331028938293457,
+      "learning_rate": 0.00019579063780268024,
+      "loss": 2.8632,
+      "step": 5601
+    },
+    {
+      "epoch": 0.717056,
+      "grad_norm": 0.531242311000824,
+      "learning_rate": 0.00019562613849799705,
+      "loss": 2.8795,
+      "step": 5602
+    },
+    {
+      "epoch": 0.717184,
+      "grad_norm": 0.3910711705684662,
+      "learning_rate": 0.00019546169151870942,
+      "loss": 2.7086,
+      "step": 5603
+    },
+    {
+      "epoch": 0.717312,
+      "grad_norm": 0.3968830406665802,
+      "learning_rate": 0.00019529729689308757,
+      "loss": 2.9039,
+      "step": 5604
+    },
+    {
+      "epoch": 0.71744,
+      "grad_norm": 0.34307682514190674,
+      "learning_rate": 0.00019513295464939273,
+      "loss": 2.6543,
+      "step": 5605
+    },
+    {
+      "epoch": 0.717568,
+      "grad_norm": 0.38861772418022156,
+      "learning_rate": 0.00019496866481587716,
+      "loss": 2.8568,
+      "step": 5606
+    },
+    {
+      "epoch": 0.717696,
+      "grad_norm": 0.391963928937912,
+      "learning_rate": 0.00019480442742078453,
+      "loss": 2.7584,
+      "step": 5607
+    },
+    {
+      "epoch": 0.717824,
+      "grad_norm": 0.4083961546421051,
+      "learning_rate": 0.00019464024249234897,
+      "loss": 2.8104,
+      "step": 5608
+    },
+    {
+      "epoch": 0.717952,
+      "grad_norm": 0.4260057210922241,
+      "learning_rate": 0.00019447611005879572,
+      "loss": 2.8719,
+      "step": 5609
+    },
+    {
+      "epoch": 0.71808,
+      "grad_norm": 0.4276992976665497,
+      "learning_rate": 0.00019431203014834092,
+      "loss": 2.8231,
+      "step": 5610
+    },
+    {
+      "epoch": 0.718208,
+      "grad_norm": 0.4170915186405182,
+      "learning_rate": 0.0001941480027891922,
+      "loss": 2.7256,
+      "step": 5611
+    },
+    {
+      "epoch": 0.718336,
+      "grad_norm": 0.4044041335582733,
+      "learning_rate": 0.00019398402800954745,
+      "loss": 2.8192,
+      "step": 5612
+    },
+    {
+      "epoch": 0.718464,
+      "grad_norm": 0.3991573750972748,
+      "learning_rate": 0.00019382010583759602,
+      "loss": 2.8394,
+      "step": 5613
+    },
+    {
+      "epoch": 0.718592,
+      "grad_norm": 0.40376946330070496,
+      "learning_rate": 0.0001936562363015179,
+      "loss": 2.772,
+      "step": 5614
+    },
+    {
+      "epoch": 0.71872,
+      "grad_norm": 0.4176672101020813,
+      "learning_rate": 0.00019349241942948403,
+      "loss": 2.6895,
+      "step": 5615
+    },
+    {
+      "epoch": 0.718848,
+      "grad_norm": 0.40193572640419006,
+      "learning_rate": 0.00019332865524965677,
+      "loss": 2.8171,
+      "step": 5616
+    },
+    {
+      "epoch": 0.718976,
+      "grad_norm": 0.4148348271846771,
+      "learning_rate": 0.00019316494379018872,
+      "loss": 2.716,
+      "step": 5617
+    },
+    {
+      "epoch": 0.719104,
+      "grad_norm": 0.4180174469947815,
+      "learning_rate": 0.00019300128507922416,
+      "loss": 2.9311,
+      "step": 5618
+    },
+    {
+      "epoch": 0.719232,
+      "grad_norm": 0.4115491211414337,
+      "learning_rate": 0.00019283767914489775,
+      "loss": 2.7734,
+      "step": 5619
+    },
+    {
+      "epoch": 0.71936,
+      "grad_norm": 0.4190838038921356,
+      "learning_rate": 0.00019267412601533518,
+      "loss": 2.9584,
+      "step": 5620
+    },
+    {
+      "epoch": 0.719488,
+      "grad_norm": 0.45079103112220764,
+      "learning_rate": 0.0001925106257186532,
+      "loss": 2.8535,
+      "step": 5621
+    },
+    {
+      "epoch": 0.719616,
+      "grad_norm": 0.38645362854003906,
+      "learning_rate": 0.00019234717828295938,
+      "loss": 2.8309,
+      "step": 5622
+    },
+    {
+      "epoch": 0.719744,
+      "grad_norm": 0.4080275297164917,
+      "learning_rate": 0.0001921837837363521,
+      "loss": 2.7567,
+      "step": 5623
+    },
+    {
+      "epoch": 0.719872,
+      "grad_norm": 0.4041535258293152,
+      "learning_rate": 0.000192020442106921,
+      "loss": 2.9163,
+      "step": 5624
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4225105345249176,
+      "learning_rate": 0.0001918571534227462,
+      "loss": 2.9396,
+      "step": 5625
+    },
+    {
+      "epoch": 0.720128,
+      "grad_norm": 0.4063555598258972,
+      "learning_rate": 0.00019169391771189914,
+      "loss": 2.8855,
+      "step": 5626
+    },
+    {
+      "epoch": 0.720256,
+      "grad_norm": 0.37957248091697693,
+      "learning_rate": 0.0001915307350024419,
+      "loss": 2.7546,
+      "step": 5627
+    },
+    {
+      "epoch": 0.720384,
+      "grad_norm": 0.41313159465789795,
+      "learning_rate": 0.00019136760532242731,
+      "loss": 2.7716,
+      "step": 5628
+    },
+    {
+      "epoch": 0.720512,
+      "grad_norm": 0.38547953963279724,
+      "learning_rate": 0.0001912045286998994,
+      "loss": 2.8025,
+      "step": 5629
+    },
+    {
+      "epoch": 0.72064,
+      "grad_norm": 0.396247535943985,
+      "learning_rate": 0.0001910415051628928,
+      "loss": 2.8041,
+      "step": 5630
+    },
+    {
+      "epoch": 0.720768,
+      "grad_norm": 0.3771476745605469,
+      "learning_rate": 0.0001908785347394331,
+      "loss": 2.827,
+      "step": 5631
+    },
+    {
+      "epoch": 0.720896,
+      "grad_norm": 0.41876915097236633,
+      "learning_rate": 0.00019071561745753713,
+      "loss": 2.9226,
+      "step": 5632
+    },
+    {
+      "epoch": 0.721024,
+      "grad_norm": 0.40431299805641174,
+      "learning_rate": 0.000190552753345212,
+      "loss": 2.7529,
+      "step": 5633
+    },
+    {
+      "epoch": 0.721152,
+      "grad_norm": 0.41509324312210083,
+      "learning_rate": 0.00019038994243045581,
+      "loss": 2.7702,
+      "step": 5634
+    },
+    {
+      "epoch": 0.72128,
+      "grad_norm": 0.40302619338035583,
+      "learning_rate": 0.000190227184741258,
+      "loss": 2.8952,
+      "step": 5635
+    },
+    {
+      "epoch": 0.721408,
+      "grad_norm": 0.4290706515312195,
+      "learning_rate": 0.00019006448030559831,
+      "loss": 2.7538,
+      "step": 5636
+    },
+    {
+      "epoch": 0.721536,
+      "grad_norm": 0.41448357701301575,
+      "learning_rate": 0.0001899018291514476,
+      "loss": 2.8162,
+      "step": 5637
+    },
+    {
+      "epoch": 0.721664,
+      "grad_norm": 0.39898091554641724,
+      "learning_rate": 0.0001897392313067672,
+      "loss": 2.8004,
+      "step": 5638
+    },
+    {
+      "epoch": 0.721792,
+      "grad_norm": 0.4179045557975769,
+      "learning_rate": 0.00018957668679950995,
+      "loss": 2.8276,
+      "step": 5639
+    },
+    {
+      "epoch": 0.72192,
+      "grad_norm": 0.4108937382698059,
+      "learning_rate": 0.00018941419565761904,
+      "loss": 2.8655,
+      "step": 5640
+    },
+    {
+      "epoch": 0.722048,
+      "grad_norm": 0.3865497410297394,
+      "learning_rate": 0.0001892517579090285,
+      "loss": 2.8274,
+      "step": 5641
+    },
+    {
+      "epoch": 0.722176,
+      "grad_norm": 0.4085240662097931,
+      "learning_rate": 0.00018908937358166322,
+      "loss": 2.9618,
+      "step": 5642
+    },
+    {
+      "epoch": 0.722304,
+      "grad_norm": 0.40878918766975403,
+      "learning_rate": 0.00018892704270343885,
+      "loss": 2.6917,
+      "step": 5643
+    },
+    {
+      "epoch": 0.722432,
+      "grad_norm": 0.41285261511802673,
+      "learning_rate": 0.00018876476530226233,
+      "loss": 2.7084,
+      "step": 5644
+    },
+    {
+      "epoch": 0.72256,
+      "grad_norm": 0.43436843156814575,
+      "learning_rate": 0.0001886025414060306,
+      "loss": 2.8481,
+      "step": 5645
+    },
+    {
+      "epoch": 0.722688,
+      "grad_norm": 0.3934682309627533,
+      "learning_rate": 0.00018844037104263223,
+      "loss": 2.8822,
+      "step": 5646
+    },
+    {
+      "epoch": 0.722816,
+      "grad_norm": 0.4593835175037384,
+      "learning_rate": 0.00018827825423994594,
+      "loss": 2.9108,
+      "step": 5647
+    },
+    {
+      "epoch": 0.722944,
+      "grad_norm": 0.41660261154174805,
+      "learning_rate": 0.00018811619102584155,
+      "loss": 2.7432,
+      "step": 5648
+    },
+    {
+      "epoch": 0.723072,
+      "grad_norm": 0.457335501909256,
+      "learning_rate": 0.0001879541814281796,
+      "loss": 2.9167,
+      "step": 5649
+    },
+    {
+      "epoch": 0.7232,
+      "grad_norm": 0.4252728521823883,
+      "learning_rate": 0.0001877922254748114,
+      "loss": 2.8962,
+      "step": 5650
+    },
+    {
+      "epoch": 0.723328,
+      "grad_norm": 0.4206703305244446,
+      "learning_rate": 0.0001876303231935791,
+      "loss": 2.8173,
+      "step": 5651
+    },
+    {
+      "epoch": 0.723456,
+      "grad_norm": 0.3872019052505493,
+      "learning_rate": 0.0001874684746123153,
+      "loss": 2.9488,
+      "step": 5652
+    },
+    {
+      "epoch": 0.723584,
+      "grad_norm": 0.4315073788166046,
+      "learning_rate": 0.00018730667975884398,
+      "loss": 2.7342,
+      "step": 5653
+    },
+    {
+      "epoch": 0.723712,
+      "grad_norm": 0.40667372941970825,
+      "learning_rate": 0.00018714493866097955,
+      "loss": 2.9591,
+      "step": 5654
+    },
+    {
+      "epoch": 0.72384,
+      "grad_norm": 0.40691274404525757,
+      "learning_rate": 0.0001869832513465271,
+      "loss": 2.7987,
+      "step": 5655
+    },
+    {
+      "epoch": 0.723968,
+      "grad_norm": 0.3999568223953247,
+      "learning_rate": 0.0001868216178432826,
+      "loss": 2.9089,
+      "step": 5656
+    },
+    {
+      "epoch": 0.724096,
+      "grad_norm": 0.3728209435939789,
+      "learning_rate": 0.00018666003817903267,
+      "loss": 2.857,
+      "step": 5657
+    },
+    {
+      "epoch": 0.724224,
+      "grad_norm": 0.3805883228778839,
+      "learning_rate": 0.00018649851238155464,
+      "loss": 2.9644,
+      "step": 5658
+    },
+    {
+      "epoch": 0.724352,
+      "grad_norm": 0.38581880927085876,
+      "learning_rate": 0.00018633704047861665,
+      "loss": 2.5898,
+      "step": 5659
+    },
+    {
+      "epoch": 0.72448,
+      "grad_norm": 0.37101730704307556,
+      "learning_rate": 0.00018617562249797787,
+      "loss": 2.7944,
+      "step": 5660
+    },
+    {
+      "epoch": 0.724608,
+      "grad_norm": 0.3754061758518219,
+      "learning_rate": 0.00018601425846738774,
+      "loss": 2.6942,
+      "step": 5661
+    },
+    {
+      "epoch": 0.724736,
+      "grad_norm": 0.38349273800849915,
+      "learning_rate": 0.0001858529484145864,
+      "loss": 2.6682,
+      "step": 5662
+    },
+    {
+      "epoch": 0.724864,
+      "grad_norm": 0.41331997513771057,
+      "learning_rate": 0.00018569169236730532,
+      "loss": 2.8582,
+      "step": 5663
+    },
+    {
+      "epoch": 0.724992,
+      "grad_norm": 0.5106680989265442,
+      "learning_rate": 0.00018553049035326612,
+      "loss": 2.7553,
+      "step": 5664
+    },
+    {
+      "epoch": 0.72512,
+      "grad_norm": 0.4231465458869934,
+      "learning_rate": 0.00018536934240018127,
+      "loss": 2.8244,
+      "step": 5665
+    },
+    {
+      "epoch": 0.725248,
+      "grad_norm": 0.39168453216552734,
+      "learning_rate": 0.0001852082485357538,
+      "loss": 2.7777,
+      "step": 5666
+    },
+    {
+      "epoch": 0.725376,
+      "grad_norm": 0.393441379070282,
+      "learning_rate": 0.00018504720878767793,
+      "loss": 2.7957,
+      "step": 5667
+    },
+    {
+      "epoch": 0.725504,
+      "grad_norm": 0.4090401530265808,
+      "learning_rate": 0.00018488622318363812,
+      "loss": 2.7931,
+      "step": 5668
+    },
+    {
+      "epoch": 0.725632,
+      "grad_norm": 0.4820912480354309,
+      "learning_rate": 0.0001847252917513097,
+      "loss": 2.8616,
+      "step": 5669
+    },
+    {
+      "epoch": 0.72576,
+      "grad_norm": 0.4030139446258545,
+      "learning_rate": 0.0001845644145183586,
+      "loss": 2.8172,
+      "step": 5670
+    },
+    {
+      "epoch": 0.725888,
+      "grad_norm": 0.3878277838230133,
+      "learning_rate": 0.0001844035915124413,
+      "loss": 2.8921,
+      "step": 5671
+    },
+    {
+      "epoch": 0.726016,
+      "grad_norm": 0.41612979769706726,
+      "learning_rate": 0.00018424282276120545,
+      "loss": 2.8753,
+      "step": 5672
+    },
+    {
+      "epoch": 0.726144,
+      "grad_norm": 0.3746958076953888,
+      "learning_rate": 0.000184082108292289,
+      "loss": 2.9296,
+      "step": 5673
+    },
+    {
+      "epoch": 0.726272,
+      "grad_norm": 0.40209609270095825,
+      "learning_rate": 0.0001839214481333204,
+      "loss": 2.745,
+      "step": 5674
+    },
+    {
+      "epoch": 0.7264,
+      "grad_norm": 0.38089117407798767,
+      "learning_rate": 0.0001837608423119193,
+      "loss": 2.8596,
+      "step": 5675
+    },
+    {
+      "epoch": 0.726528,
+      "grad_norm": 0.3986201584339142,
+      "learning_rate": 0.00018360029085569558,
+      "loss": 2.8292,
+      "step": 5676
+    },
+    {
+      "epoch": 0.726656,
+      "grad_norm": 0.4091886281967163,
+      "learning_rate": 0.0001834397937922499,
+      "loss": 2.9154,
+      "step": 5677
+    },
+    {
+      "epoch": 0.726784,
+      "grad_norm": 0.40659818053245544,
+      "learning_rate": 0.0001832793511491735,
+      "loss": 2.7026,
+      "step": 5678
+    },
+    {
+      "epoch": 0.726912,
+      "grad_norm": 0.38848069310188293,
+      "learning_rate": 0.00018311896295404839,
+      "loss": 2.7181,
+      "step": 5679
+    },
+    {
+      "epoch": 0.72704,
+      "grad_norm": 0.36186033487319946,
+      "learning_rate": 0.000182958629234447,
+      "loss": 2.9101,
+      "step": 5680
+    },
+    {
+      "epoch": 0.727168,
+      "grad_norm": 0.4290813207626343,
+      "learning_rate": 0.0001827983500179327,
+      "loss": 2.8375,
+      "step": 5681
+    },
+    {
+      "epoch": 0.727296,
+      "grad_norm": 0.4007698893547058,
+      "learning_rate": 0.00018263812533205953,
+      "loss": 2.8327,
+      "step": 5682
+    },
+    {
+      "epoch": 0.727424,
+      "grad_norm": 0.4003649353981018,
+      "learning_rate": 0.00018247795520437177,
+      "loss": 2.8954,
+      "step": 5683
+    },
+    {
+      "epoch": 0.727552,
+      "grad_norm": 0.4746381938457489,
+      "learning_rate": 0.00018231783966240456,
+      "loss": 2.8324,
+      "step": 5684
+    },
+    {
+      "epoch": 0.72768,
+      "grad_norm": 0.3691432774066925,
+      "learning_rate": 0.00018215777873368361,
+      "loss": 2.8142,
+      "step": 5685
+    },
+    {
+      "epoch": 0.727808,
+      "grad_norm": 0.4058607816696167,
+      "learning_rate": 0.00018199777244572523,
+      "loss": 2.8562,
+      "step": 5686
+    },
+    {
+      "epoch": 0.727936,
+      "grad_norm": 0.3954312205314636,
+      "learning_rate": 0.00018183782082603617,
+      "loss": 2.7864,
+      "step": 5687
+    },
+    {
+      "epoch": 0.728064,
+      "grad_norm": 0.39277786016464233,
+      "learning_rate": 0.00018167792390211436,
+      "loss": 2.7915,
+      "step": 5688
+    },
+    {
+      "epoch": 0.728192,
+      "grad_norm": 0.41055724024772644,
+      "learning_rate": 0.0001815180817014475,
+      "loss": 2.7997,
+      "step": 5689
+    },
+    {
+      "epoch": 0.72832,
+      "grad_norm": 0.43162333965301514,
+      "learning_rate": 0.0001813582942515148,
+      "loss": 2.8379,
+      "step": 5690
+    },
+    {
+      "epoch": 0.728448,
+      "grad_norm": 0.37094640731811523,
+      "learning_rate": 0.00018119856157978531,
+      "loss": 2.8875,
+      "step": 5691
+    },
+    {
+      "epoch": 0.728576,
+      "grad_norm": 0.39571839570999146,
+      "learning_rate": 0.00018103888371371897,
+      "loss": 2.8153,
+      "step": 5692
+    },
+    {
+      "epoch": 0.728704,
+      "grad_norm": 0.43145743012428284,
+      "learning_rate": 0.0001808792606807662,
+      "loss": 2.8642,
+      "step": 5693
+    },
+    {
+      "epoch": 0.728832,
+      "grad_norm": 0.42512065172195435,
+      "learning_rate": 0.00018071969250836812,
+      "loss": 2.9368,
+      "step": 5694
+    },
+    {
+      "epoch": 0.72896,
+      "grad_norm": 0.4043430984020233,
+      "learning_rate": 0.00018056017922395606,
+      "loss": 2.8677,
+      "step": 5695
+    },
+    {
+      "epoch": 0.729088,
+      "grad_norm": 0.4265734851360321,
+      "learning_rate": 0.00018040072085495274,
+      "loss": 2.8991,
+      "step": 5696
+    },
+    {
+      "epoch": 0.729216,
+      "grad_norm": 0.3890604078769684,
+      "learning_rate": 0.00018024131742877054,
+      "loss": 2.8843,
+      "step": 5697
+    },
+    {
+      "epoch": 0.729344,
+      "grad_norm": 0.4299130141735077,
+      "learning_rate": 0.00018008196897281286,
+      "loss": 2.833,
+      "step": 5698
+    },
+    {
+      "epoch": 0.729472,
+      "grad_norm": 0.4041121304035187,
+      "learning_rate": 0.00017992267551447333,
+      "loss": 2.8468,
+      "step": 5699
+    },
+    {
+      "epoch": 0.7296,
+      "grad_norm": 0.4180442988872528,
+      "learning_rate": 0.00017976343708113674,
+      "loss": 2.8284,
+      "step": 5700
+    },
+    {
+      "epoch": 0.729728,
+      "grad_norm": 0.3794522285461426,
+      "learning_rate": 0.0001796042537001778,
+      "loss": 2.7721,
+      "step": 5701
+    },
+    {
+      "epoch": 0.729856,
+      "grad_norm": 0.40512222051620483,
+      "learning_rate": 0.0001794451253989618,
+      "loss": 2.8332,
+      "step": 5702
+    },
+    {
+      "epoch": 0.729984,
+      "grad_norm": 0.4206134080886841,
+      "learning_rate": 0.00017928605220484512,
+      "loss": 2.6871,
+      "step": 5703
+    },
+    {
+      "epoch": 0.730112,
+      "grad_norm": 0.38419032096862793,
+      "learning_rate": 0.0001791270341451741,
+      "loss": 2.7083,
+      "step": 5704
+    },
+    {
+      "epoch": 0.73024,
+      "grad_norm": 0.4240459203720093,
+      "learning_rate": 0.00017896807124728582,
+      "loss": 2.8422,
+      "step": 5705
+    },
+    {
+      "epoch": 0.730368,
+      "grad_norm": 0.41638219356536865,
+      "learning_rate": 0.00017880916353850778,
+      "loss": 2.8687,
+      "step": 5706
+    },
+    {
+      "epoch": 0.730496,
+      "grad_norm": 0.5515574216842651,
+      "learning_rate": 0.00017865031104615807,
+      "loss": 2.8413,
+      "step": 5707
+    },
+    {
+      "epoch": 0.730624,
+      "grad_norm": 0.432660847902298,
+      "learning_rate": 0.0001784915137975452,
+      "loss": 2.6979,
+      "step": 5708
+    },
+    {
+      "epoch": 0.730752,
+      "grad_norm": 0.4233795404434204,
+      "learning_rate": 0.0001783327718199685,
+      "loss": 2.8143,
+      "step": 5709
+    },
+    {
+      "epoch": 0.73088,
+      "grad_norm": 0.38472211360931396,
+      "learning_rate": 0.0001781740851407172,
+      "loss": 2.8516,
+      "step": 5710
+    },
+    {
+      "epoch": 0.731008,
+      "grad_norm": 0.41738682985305786,
+      "learning_rate": 0.00017801545378707178,
+      "loss": 2.8349,
+      "step": 5711
+    },
+    {
+      "epoch": 0.731136,
+      "grad_norm": 0.4062075614929199,
+      "learning_rate": 0.00017785687778630266,
+      "loss": 2.8613,
+      "step": 5712
+    },
+    {
+      "epoch": 0.731264,
+      "grad_norm": 0.40037983655929565,
+      "learning_rate": 0.0001776983571656708,
+      "loss": 2.8923,
+      "step": 5713
+    },
+    {
+      "epoch": 0.731392,
+      "grad_norm": 0.36861124634742737,
+      "learning_rate": 0.00017753989195242782,
+      "loss": 2.8822,
+      "step": 5714
+    },
+    {
+      "epoch": 0.73152,
+      "grad_norm": 0.3789016008377075,
+      "learning_rate": 0.00017738148217381566,
+      "loss": 2.8404,
+      "step": 5715
+    },
+    {
+      "epoch": 0.731648,
+      "grad_norm": 0.4049930274486542,
+      "learning_rate": 0.00017722312785706668,
+      "loss": 2.8645,
+      "step": 5716
+    },
+    {
+      "epoch": 0.731776,
+      "grad_norm": 0.43416643142700195,
+      "learning_rate": 0.00017706482902940396,
+      "loss": 2.6555,
+      "step": 5717
+    },
+    {
+      "epoch": 0.731904,
+      "grad_norm": 0.4086271822452545,
+      "learning_rate": 0.00017690658571804107,
+      "loss": 2.7854,
+      "step": 5718
+    },
+    {
+      "epoch": 0.732032,
+      "grad_norm": 0.42517784237861633,
+      "learning_rate": 0.00017674839795018171,
+      "loss": 2.912,
+      "step": 5719
+    },
+    {
+      "epoch": 0.73216,
+      "grad_norm": 0.40207982063293457,
+      "learning_rate": 0.00017659026575302022,
+      "loss": 2.8573,
+      "step": 5720
+    },
+    {
+      "epoch": 0.732288,
+      "grad_norm": 0.466498464345932,
+      "learning_rate": 0.00017643218915374138,
+      "loss": 2.7642,
+      "step": 5721
+    },
+    {
+      "epoch": 0.732416,
+      "grad_norm": 0.4158949851989746,
+      "learning_rate": 0.0001762741681795203,
+      "loss": 3.0253,
+      "step": 5722
+    },
+    {
+      "epoch": 0.732544,
+      "grad_norm": 0.3934924304485321,
+      "learning_rate": 0.00017611620285752245,
+      "loss": 2.7517,
+      "step": 5723
+    },
+    {
+      "epoch": 0.732672,
+      "grad_norm": 0.40549302101135254,
+      "learning_rate": 0.00017595829321490435,
+      "loss": 2.795,
+      "step": 5724
+    },
+    {
+      "epoch": 0.7328,
+      "grad_norm": 0.3971658945083618,
+      "learning_rate": 0.00017580043927881224,
+      "loss": 2.8436,
+      "step": 5725
+    },
+    {
+      "epoch": 0.732928,
+      "grad_norm": 0.40559226274490356,
+      "learning_rate": 0.0001756426410763829,
+      "loss": 2.7919,
+      "step": 5726
+    },
+    {
+      "epoch": 0.733056,
+      "grad_norm": 0.40191179513931274,
+      "learning_rate": 0.00017548489863474393,
+      "loss": 2.7786,
+      "step": 5727
+    },
+    {
+      "epoch": 0.733184,
+      "grad_norm": 0.3893684446811676,
+      "learning_rate": 0.0001753272119810131,
+      "loss": 2.9255,
+      "step": 5728
+    },
+    {
+      "epoch": 0.733312,
+      "grad_norm": 0.4065633714199066,
+      "learning_rate": 0.00017516958114229835,
+      "loss": 2.7245,
+      "step": 5729
+    },
+    {
+      "epoch": 0.73344,
+      "grad_norm": 0.4112814664840698,
+      "learning_rate": 0.00017501200614569845,
+      "loss": 2.9015,
+      "step": 5730
+    },
+    {
+      "epoch": 0.733568,
+      "grad_norm": 0.422139972448349,
+      "learning_rate": 0.00017485448701830205,
+      "loss": 2.8889,
+      "step": 5731
+    },
+    {
+      "epoch": 0.733696,
+      "grad_norm": 0.39655381441116333,
+      "learning_rate": 0.00017469702378718893,
+      "loss": 2.7274,
+      "step": 5732
+    },
+    {
+      "epoch": 0.733824,
+      "grad_norm": 0.4026377499103546,
+      "learning_rate": 0.00017453961647942868,
+      "loss": 2.8529,
+      "step": 5733
+    },
+    {
+      "epoch": 0.733952,
+      "grad_norm": 0.4316220283508301,
+      "learning_rate": 0.00017438226512208138,
+      "loss": 2.8075,
+      "step": 5734
+    },
+    {
+      "epoch": 0.73408,
+      "grad_norm": 0.3839317560195923,
+      "learning_rate": 0.0001742249697421976,
+      "loss": 2.7565,
+      "step": 5735
+    },
+    {
+      "epoch": 0.734208,
+      "grad_norm": 0.39512956142425537,
+      "learning_rate": 0.00017406773036681805,
+      "loss": 2.9429,
+      "step": 5736
+    },
+    {
+      "epoch": 0.734336,
+      "grad_norm": 0.38711971044540405,
+      "learning_rate": 0.00017391054702297438,
+      "loss": 2.935,
+      "step": 5737
+    },
+    {
+      "epoch": 0.734464,
+      "grad_norm": 0.400461882352829,
+      "learning_rate": 0.00017375341973768787,
+      "loss": 2.7923,
+      "step": 5738
+    },
+    {
+      "epoch": 0.734592,
+      "grad_norm": 0.4130018353462219,
+      "learning_rate": 0.0001735963485379708,
+      "loss": 2.7738,
+      "step": 5739
+    },
+    {
+      "epoch": 0.73472,
+      "grad_norm": 0.4242972731590271,
+      "learning_rate": 0.00017343933345082547,
+      "loss": 2.8984,
+      "step": 5740
+    },
+    {
+      "epoch": 0.734848,
+      "grad_norm": 0.4131486415863037,
+      "learning_rate": 0.00017328237450324452,
+      "loss": 2.7868,
+      "step": 5741
+    },
+    {
+      "epoch": 0.734976,
+      "grad_norm": 0.4487167298793793,
+      "learning_rate": 0.00017312547172221098,
+      "loss": 2.9274,
+      "step": 5742
+    },
+    {
+      "epoch": 0.735104,
+      "grad_norm": 0.3840678632259369,
+      "learning_rate": 0.00017296862513469836,
+      "loss": 2.6998,
+      "step": 5743
+    },
+    {
+      "epoch": 0.735232,
+      "grad_norm": 0.4308314919471741,
+      "learning_rate": 0.00017281183476767015,
+      "loss": 2.8102,
+      "step": 5744
+    },
+    {
+      "epoch": 0.73536,
+      "grad_norm": 0.48931965231895447,
+      "learning_rate": 0.00017265510064808082,
+      "loss": 2.8088,
+      "step": 5745
+    },
+    {
+      "epoch": 0.735488,
+      "grad_norm": 0.3943510353565216,
+      "learning_rate": 0.00017249842280287443,
+      "loss": 2.6569,
+      "step": 5746
+    },
+    {
+      "epoch": 0.735616,
+      "grad_norm": 0.42107152938842773,
+      "learning_rate": 0.00017234180125898606,
+      "loss": 2.8223,
+      "step": 5747
+    },
+    {
+      "epoch": 0.735744,
+      "grad_norm": 0.4313999116420746,
+      "learning_rate": 0.00017218523604334057,
+      "loss": 2.8878,
+      "step": 5748
+    },
+    {
+      "epoch": 0.735872,
+      "grad_norm": 0.43895331025123596,
+      "learning_rate": 0.0001720287271828534,
+      "loss": 2.8736,
+      "step": 5749
+    },
+    {
+      "epoch": 0.736,
+      "grad_norm": 0.38427141308784485,
+      "learning_rate": 0.00017187227470443013,
+      "loss": 2.8581,
+      "step": 5750
+    },
+    {
+      "epoch": 0.736128,
+      "grad_norm": 0.436585009098053,
+      "learning_rate": 0.00017171587863496686,
+      "loss": 2.8017,
+      "step": 5751
+    },
+    {
+      "epoch": 0.736256,
+      "grad_norm": 0.40028420090675354,
+      "learning_rate": 0.00017155953900134969,
+      "loss": 2.8586,
+      "step": 5752
+    },
+    {
+      "epoch": 0.736384,
+      "grad_norm": 0.396611750125885,
+      "learning_rate": 0.00017140325583045552,
+      "loss": 3.0193,
+      "step": 5753
+    },
+    {
+      "epoch": 0.736512,
+      "grad_norm": 0.42958080768585205,
+      "learning_rate": 0.00017124702914915097,
+      "loss": 3.0051,
+      "step": 5754
+    },
+    {
+      "epoch": 0.73664,
+      "grad_norm": 0.4023631513118744,
+      "learning_rate": 0.00017109085898429343,
+      "loss": 2.8059,
+      "step": 5755
+    },
+    {
+      "epoch": 0.736768,
+      "grad_norm": 0.4360014498233795,
+      "learning_rate": 0.00017093474536273036,
+      "loss": 2.91,
+      "step": 5756
+    },
+    {
+      "epoch": 0.736896,
+      "grad_norm": 0.41664567589759827,
+      "learning_rate": 0.00017077868831129933,
+      "loss": 2.7752,
+      "step": 5757
+    },
+    {
+      "epoch": 0.737024,
+      "grad_norm": 0.4246987998485565,
+      "learning_rate": 0.00017062268785682848,
+      "loss": 2.8875,
+      "step": 5758
+    },
+    {
+      "epoch": 0.737152,
+      "grad_norm": 0.3726297914981842,
+      "learning_rate": 0.00017046674402613593,
+      "loss": 2.87,
+      "step": 5759
+    },
+    {
+      "epoch": 0.73728,
+      "grad_norm": 0.3859657943248749,
+      "learning_rate": 0.00017031085684603048,
+      "loss": 2.6714,
+      "step": 5760
+    },
+    {
+      "epoch": 0.737408,
+      "grad_norm": 0.39050784707069397,
+      "learning_rate": 0.00017015502634331082,
+      "loss": 2.8368,
+      "step": 5761
+    },
+    {
+      "epoch": 0.737536,
+      "grad_norm": 0.40209028124809265,
+      "learning_rate": 0.00016999925254476605,
+      "loss": 2.8939,
+      "step": 5762
+    },
+    {
+      "epoch": 0.737664,
+      "grad_norm": 0.43004485964775085,
+      "learning_rate": 0.0001698435354771753,
+      "loss": 2.8218,
+      "step": 5763
+    },
+    {
+      "epoch": 0.737792,
+      "grad_norm": 0.4191664159297943,
+      "learning_rate": 0.00016968787516730843,
+      "loss": 2.8641,
+      "step": 5764
+    },
+    {
+      "epoch": 0.73792,
+      "grad_norm": 0.4011708199977875,
+      "learning_rate": 0.00016953227164192514,
+      "loss": 2.771,
+      "step": 5765
+    },
+    {
+      "epoch": 0.738048,
+      "grad_norm": 0.42762047052383423,
+      "learning_rate": 0.00016937672492777544,
+      "loss": 2.9625,
+      "step": 5766
+    },
+    {
+      "epoch": 0.738176,
+      "grad_norm": 0.40987130999565125,
+      "learning_rate": 0.00016922123505159952,
+      "loss": 2.8465,
+      "step": 5767
+    },
+    {
+      "epoch": 0.738304,
+      "grad_norm": 0.4234412908554077,
+      "learning_rate": 0.00016906580204012817,
+      "loss": 2.7007,
+      "step": 5768
+    },
+    {
+      "epoch": 0.738432,
+      "grad_norm": 0.4016808569431305,
+      "learning_rate": 0.00016891042592008198,
+      "loss": 2.7461,
+      "step": 5769
+    },
+    {
+      "epoch": 0.73856,
+      "grad_norm": 0.40501871705055237,
+      "learning_rate": 0.00016875510671817185,
+      "loss": 2.7943,
+      "step": 5770
+    },
+    {
+      "epoch": 0.738688,
+      "grad_norm": 0.46846693754196167,
+      "learning_rate": 0.00016859984446109905,
+      "loss": 2.9159,
+      "step": 5771
+    },
+    {
+      "epoch": 0.738816,
+      "grad_norm": 0.388249009847641,
+      "learning_rate": 0.00016844463917555487,
+      "loss": 2.8774,
+      "step": 5772
+    },
+    {
+      "epoch": 0.738944,
+      "grad_norm": 0.40570926666259766,
+      "learning_rate": 0.00016828949088822077,
+      "loss": 2.8953,
+      "step": 5773
+    },
+    {
+      "epoch": 0.739072,
+      "grad_norm": 0.44169196486473083,
+      "learning_rate": 0.00016813439962576872,
+      "loss": 2.8591,
+      "step": 5774
+    },
+    {
+      "epoch": 0.7392,
+      "grad_norm": 0.41733258962631226,
+      "learning_rate": 0.00016797936541486082,
+      "loss": 2.8827,
+      "step": 5775
+    },
+    {
+      "epoch": 0.739328,
+      "grad_norm": 0.4051777124404907,
+      "learning_rate": 0.00016782438828214912,
+      "loss": 2.8314,
+      "step": 5776
+    },
+    {
+      "epoch": 0.739456,
+      "grad_norm": 0.4077097475528717,
+      "learning_rate": 0.00016766946825427605,
+      "loss": 2.7969,
+      "step": 5777
+    },
+    {
+      "epoch": 0.739584,
+      "grad_norm": 0.4007403254508972,
+      "learning_rate": 0.00016751460535787405,
+      "loss": 2.6269,
+      "step": 5778
+    },
+    {
+      "epoch": 0.739712,
+      "grad_norm": 0.3885023891925812,
+      "learning_rate": 0.00016735979961956588,
+      "loss": 2.8533,
+      "step": 5779
+    },
+    {
+      "epoch": 0.73984,
+      "grad_norm": 0.3929691016674042,
+      "learning_rate": 0.00016720505106596428,
+      "loss": 2.798,
+      "step": 5780
+    },
+    {
+      "epoch": 0.739968,
+      "grad_norm": 0.39759570360183716,
+      "learning_rate": 0.00016705035972367272,
+      "loss": 2.9286,
+      "step": 5781
+    },
+    {
+      "epoch": 0.740096,
+      "grad_norm": 0.40975135564804077,
+      "learning_rate": 0.00016689572561928395,
+      "loss": 2.7507,
+      "step": 5782
+    },
+    {
+      "epoch": 0.740224,
+      "grad_norm": 0.4421709179878235,
+      "learning_rate": 0.00016674114877938183,
+      "loss": 2.7654,
+      "step": 5783
+    },
+    {
+      "epoch": 0.740352,
+      "grad_norm": 0.3819759786128998,
+      "learning_rate": 0.00016658662923053975,
+      "loss": 2.8744,
+      "step": 5784
+    },
+    {
+      "epoch": 0.74048,
+      "grad_norm": 0.3879565894603729,
+      "learning_rate": 0.0001664321669993213,
+      "loss": 2.835,
+      "step": 5785
+    },
+    {
+      "epoch": 0.740608,
+      "grad_norm": 0.3724902868270874,
+      "learning_rate": 0.0001662777621122804,
+      "loss": 2.7326,
+      "step": 5786
+    },
+    {
+      "epoch": 0.740736,
+      "grad_norm": 0.4087950885295868,
+      "learning_rate": 0.0001661234145959611,
+      "loss": 2.8234,
+      "step": 5787
+    },
+    {
+      "epoch": 0.740864,
+      "grad_norm": 0.40751007199287415,
+      "learning_rate": 0.0001659691244768972,
+      "loss": 2.6421,
+      "step": 5788
+    },
+    {
+      "epoch": 0.740992,
+      "grad_norm": 0.3892669379711151,
+      "learning_rate": 0.00016581489178161347,
+      "loss": 2.8731,
+      "step": 5789
+    },
+    {
+      "epoch": 0.74112,
+      "grad_norm": 0.4162385165691376,
+      "learning_rate": 0.00016566071653662402,
+      "loss": 2.7956,
+      "step": 5790
+    },
+    {
+      "epoch": 0.741248,
+      "grad_norm": 0.4123213291168213,
+      "learning_rate": 0.00016550659876843316,
+      "loss": 2.8492,
+      "step": 5791
+    },
+    {
+      "epoch": 0.741376,
+      "grad_norm": 0.4261817932128906,
+      "learning_rate": 0.000165352538503536,
+      "loss": 2.8199,
+      "step": 5792
+    },
+    {
+      "epoch": 0.741504,
+      "grad_norm": 0.4053734838962555,
+      "learning_rate": 0.000165198535768417,
+      "loss": 2.7214,
+      "step": 5793
+    },
+    {
+      "epoch": 0.741632,
+      "grad_norm": 0.39749956130981445,
+      "learning_rate": 0.00016504459058955107,
+      "loss": 2.7235,
+      "step": 5794
+    },
+    {
+      "epoch": 0.74176,
+      "grad_norm": 0.38876229524612427,
+      "learning_rate": 0.00016489070299340297,
+      "loss": 2.6985,
+      "step": 5795
+    },
+    {
+      "epoch": 0.741888,
+      "grad_norm": 0.44426149129867554,
+      "learning_rate": 0.00016473687300642815,
+      "loss": 2.7805,
+      "step": 5796
+    },
+    {
+      "epoch": 0.742016,
+      "grad_norm": 0.44628140330314636,
+      "learning_rate": 0.00016458310065507154,
+      "loss": 2.8135,
+      "step": 5797
+    },
+    {
+      "epoch": 0.742144,
+      "grad_norm": 0.3979543447494507,
+      "learning_rate": 0.00016442938596576839,
+      "loss": 2.8954,
+      "step": 5798
+    },
+    {
+      "epoch": 0.742272,
+      "grad_norm": 0.4149665832519531,
+      "learning_rate": 0.0001642757289649441,
+      "loss": 2.8873,
+      "step": 5799
+    },
+    {
+      "epoch": 0.7424,
+      "grad_norm": 0.4052641987800598,
+      "learning_rate": 0.00016412212967901386,
+      "loss": 2.972,
+      "step": 5800
+    },
+    {
+      "epoch": 0.742528,
+      "grad_norm": 0.41587573289871216,
+      "learning_rate": 0.00016396858813438355,
+      "loss": 2.7488,
+      "step": 5801
+    },
+    {
+      "epoch": 0.742656,
+      "grad_norm": 0.4129316508769989,
+      "learning_rate": 0.0001638151043574485,
+      "loss": 2.98,
+      "step": 5802
+    },
+    {
+      "epoch": 0.742784,
+      "grad_norm": 0.4113772511482239,
+      "learning_rate": 0.00016366167837459427,
+      "loss": 2.8937,
+      "step": 5803
+    },
+    {
+      "epoch": 0.742912,
+      "grad_norm": 0.38557273149490356,
+      "learning_rate": 0.00016350831021219687,
+      "loss": 2.7684,
+      "step": 5804
+    },
+    {
+      "epoch": 0.74304,
+      "grad_norm": 0.40110984444618225,
+      "learning_rate": 0.00016335499989662185,
+      "loss": 2.7962,
+      "step": 5805
+    },
+    {
+      "epoch": 0.743168,
+      "grad_norm": 0.4084583520889282,
+      "learning_rate": 0.0001632017474542251,
+      "loss": 2.8156,
+      "step": 5806
+    },
+    {
+      "epoch": 0.743296,
+      "grad_norm": 0.4226151704788208,
+      "learning_rate": 0.00016304855291135245,
+      "loss": 2.8311,
+      "step": 5807
+    },
+    {
+      "epoch": 0.743424,
+      "grad_norm": 0.41061291098594666,
+      "learning_rate": 0.00016289541629433985,
+      "loss": 2.7025,
+      "step": 5808
+    },
+    {
+      "epoch": 0.743552,
+      "grad_norm": 0.3992769420146942,
+      "learning_rate": 0.00016274233762951306,
+      "loss": 2.8437,
+      "step": 5809
+    },
+    {
+      "epoch": 0.74368,
+      "grad_norm": 0.4023081958293915,
+      "learning_rate": 0.00016258931694318829,
+      "loss": 2.7575,
+      "step": 5810
+    },
+    {
+      "epoch": 0.743808,
+      "grad_norm": 0.3786858022212982,
+      "learning_rate": 0.0001624363542616717,
+      "loss": 2.7995,
+      "step": 5811
+    },
+    {
+      "epoch": 0.743936,
+      "grad_norm": 0.3958410918712616,
+      "learning_rate": 0.00016228344961125913,
+      "loss": 2.8538,
+      "step": 5812
+    },
+    {
+      "epoch": 0.744064,
+      "grad_norm": 0.40393826365470886,
+      "learning_rate": 0.00016213060301823674,
+      "loss": 2.8429,
+      "step": 5813
+    },
+    {
+      "epoch": 0.744192,
+      "grad_norm": 0.3950788676738739,
+      "learning_rate": 0.00016197781450888067,
+      "loss": 2.8512,
+      "step": 5814
+    },
+    {
+      "epoch": 0.74432,
+      "grad_norm": 0.38474878668785095,
+      "learning_rate": 0.0001618250841094569,
+      "loss": 2.7414,
+      "step": 5815
+    },
+    {
+      "epoch": 0.744448,
+      "grad_norm": 0.3970049023628235,
+      "learning_rate": 0.0001616724118462215,
+      "loss": 2.7681,
+      "step": 5816
+    },
+    {
+      "epoch": 0.744576,
+      "grad_norm": 0.39422428607940674,
+      "learning_rate": 0.00016151979774542085,
+      "loss": 2.852,
+      "step": 5817
+    },
+    {
+      "epoch": 0.744704,
+      "grad_norm": 0.42396023869514465,
+      "learning_rate": 0.00016136724183329104,
+      "loss": 2.7284,
+      "step": 5818
+    },
+    {
+      "epoch": 0.744832,
+      "grad_norm": 0.4110512137413025,
+      "learning_rate": 0.0001612147441360579,
+      "loss": 2.7746,
+      "step": 5819
+    },
+    {
+      "epoch": 0.74496,
+      "grad_norm": 0.3901733458042145,
+      "learning_rate": 0.000161062304679938,
+      "loss": 2.8838,
+      "step": 5820
+    },
+    {
+      "epoch": 0.745088,
+      "grad_norm": 0.4025753140449524,
+      "learning_rate": 0.0001609099234911372,
+      "loss": 2.7925,
+      "step": 5821
+    },
+    {
+      "epoch": 0.745216,
+      "grad_norm": 0.42084744572639465,
+      "learning_rate": 0.00016075760059585165,
+      "loss": 2.8182,
+      "step": 5822
+    },
+    {
+      "epoch": 0.745344,
+      "grad_norm": 0.36804306507110596,
+      "learning_rate": 0.00016060533602026733,
+      "loss": 2.8987,
+      "step": 5823
+    },
+    {
+      "epoch": 0.745472,
+      "grad_norm": 0.4400741159915924,
+      "learning_rate": 0.00016045312979056026,
+      "loss": 2.9293,
+      "step": 5824
+    },
+    {
+      "epoch": 0.7456,
+      "grad_norm": 0.39332351088523865,
+      "learning_rate": 0.00016030098193289667,
+      "loss": 2.8637,
+      "step": 5825
+    },
+    {
+      "epoch": 0.745728,
+      "grad_norm": 0.39441022276878357,
+      "learning_rate": 0.00016014889247343238,
+      "loss": 2.867,
+      "step": 5826
+    },
+    {
+      "epoch": 0.745856,
+      "grad_norm": 0.41240864992141724,
+      "learning_rate": 0.00015999686143831343,
+      "loss": 2.7298,
+      "step": 5827
+    },
+    {
+      "epoch": 0.745984,
+      "grad_norm": 0.39726778864860535,
+      "learning_rate": 0.00015984488885367542,
+      "loss": 2.9081,
+      "step": 5828
+    },
+    {
+      "epoch": 0.746112,
+      "grad_norm": 0.4081461429595947,
+      "learning_rate": 0.00015969297474564465,
+      "loss": 2.8983,
+      "step": 5829
+    },
+    {
+      "epoch": 0.74624,
+      "grad_norm": 0.41240212321281433,
+      "learning_rate": 0.0001595411191403367,
+      "loss": 2.75,
+      "step": 5830
+    },
+    {
+      "epoch": 0.746368,
+      "grad_norm": 0.44013431668281555,
+      "learning_rate": 0.00015938932206385715,
+      "loss": 2.7221,
+      "step": 5831
+    },
+    {
+      "epoch": 0.746496,
+      "grad_norm": 0.3772814869880676,
+      "learning_rate": 0.00015923758354230194,
+      "loss": 2.6736,
+      "step": 5832
+    },
+    {
+      "epoch": 0.746624,
+      "grad_norm": 0.38866764307022095,
+      "learning_rate": 0.00015908590360175662,
+      "loss": 2.787,
+      "step": 5833
+    },
+    {
+      "epoch": 0.746752,
+      "grad_norm": 0.37989377975463867,
+      "learning_rate": 0.00015893428226829671,
+      "loss": 2.8551,
+      "step": 5834
+    },
+    {
+      "epoch": 0.74688,
+      "grad_norm": 0.41493701934814453,
+      "learning_rate": 0.00015878271956798763,
+      "loss": 2.8161,
+      "step": 5835
+    },
+    {
+      "epoch": 0.747008,
+      "grad_norm": 0.4184514880180359,
+      "learning_rate": 0.00015863121552688476,
+      "loss": 2.8808,
+      "step": 5836
+    },
+    {
+      "epoch": 0.747136,
+      "grad_norm": 0.42070963978767395,
+      "learning_rate": 0.00015847977017103326,
+      "loss": 2.8951,
+      "step": 5837
+    },
+    {
+      "epoch": 0.747264,
+      "grad_norm": 0.4183414578437805,
+      "learning_rate": 0.00015832838352646856,
+      "loss": 2.8426,
+      "step": 5838
+    },
+    {
+      "epoch": 0.747392,
+      "grad_norm": 0.3963720202445984,
+      "learning_rate": 0.00015817705561921587,
+      "loss": 2.8446,
+      "step": 5839
+    },
+    {
+      "epoch": 0.74752,
+      "grad_norm": 0.4210982322692871,
+      "learning_rate": 0.00015802578647529004,
+      "loss": 2.8242,
+      "step": 5840
+    },
+    {
+      "epoch": 0.747648,
+      "grad_norm": 0.39354801177978516,
+      "learning_rate": 0.00015787457612069606,
+      "loss": 2.6436,
+      "step": 5841
+    },
+    {
+      "epoch": 0.747776,
+      "grad_norm": 0.4194756746292114,
+      "learning_rate": 0.00015772342458142868,
+      "loss": 2.8882,
+      "step": 5842
+    },
+    {
+      "epoch": 0.747904,
+      "grad_norm": 0.38502606749534607,
+      "learning_rate": 0.00015757233188347264,
+      "loss": 2.7367,
+      "step": 5843
+    },
+    {
+      "epoch": 0.748032,
+      "grad_norm": 0.4229998290538788,
+      "learning_rate": 0.00015742129805280247,
+      "loss": 2.7199,
+      "step": 5844
+    },
+    {
+      "epoch": 0.74816,
+      "grad_norm": 0.4104502201080322,
+      "learning_rate": 0.00015727032311538258,
+      "loss": 2.7668,
+      "step": 5845
+    },
+    {
+      "epoch": 0.748288,
+      "grad_norm": 0.43313491344451904,
+      "learning_rate": 0.00015711940709716755,
+      "loss": 2.8004,
+      "step": 5846
+    },
+    {
+      "epoch": 0.748416,
+      "grad_norm": 0.39337411522865295,
+      "learning_rate": 0.00015696855002410128,
+      "loss": 2.7803,
+      "step": 5847
+    },
+    {
+      "epoch": 0.748544,
+      "grad_norm": 0.38040661811828613,
+      "learning_rate": 0.00015681775192211817,
+      "loss": 2.7728,
+      "step": 5848
+    },
+    {
+      "epoch": 0.748672,
+      "grad_norm": 0.40119439363479614,
+      "learning_rate": 0.00015666701281714202,
+      "loss": 2.9151,
+      "step": 5849
+    },
+    {
+      "epoch": 0.7488,
+      "grad_norm": 0.42730769515037537,
+      "learning_rate": 0.00015651633273508665,
+      "loss": 2.8386,
+      "step": 5850
+    },
+    {
+      "epoch": 0.748928,
+      "grad_norm": 0.41976743936538696,
+      "learning_rate": 0.00015636571170185565,
+      "loss": 2.7051,
+      "step": 5851
+    },
+    {
+      "epoch": 0.749056,
+      "grad_norm": 0.4410417675971985,
+      "learning_rate": 0.00015621514974334245,
+      "loss": 2.7358,
+      "step": 5852
+    },
+    {
+      "epoch": 0.749184,
+      "grad_norm": 0.45395082235336304,
+      "learning_rate": 0.00015606464688543064,
+      "loss": 2.786,
+      "step": 5853
+    },
+    {
+      "epoch": 0.749312,
+      "grad_norm": 0.3996146619319916,
+      "learning_rate": 0.00015591420315399325,
+      "loss": 2.7037,
+      "step": 5854
+    },
+    {
+      "epoch": 0.74944,
+      "grad_norm": 0.4233349859714508,
+      "learning_rate": 0.00015576381857489337,
+      "loss": 2.8056,
+      "step": 5855
+    },
+    {
+      "epoch": 0.749568,
+      "grad_norm": 0.38889655470848083,
+      "learning_rate": 0.0001556134931739836,
+      "loss": 2.9415,
+      "step": 5856
+    },
+    {
+      "epoch": 0.749696,
+      "grad_norm": 0.4181360900402069,
+      "learning_rate": 0.00015546322697710698,
+      "loss": 2.8532,
+      "step": 5857
+    },
+    {
+      "epoch": 0.749824,
+      "grad_norm": 0.41981256008148193,
+      "learning_rate": 0.0001553130200100959,
+      "loss": 2.807,
+      "step": 5858
+    },
+    {
+      "epoch": 0.749952,
+      "grad_norm": 0.40848401188850403,
+      "learning_rate": 0.00015516287229877242,
+      "loss": 2.9469,
+      "step": 5859
+    },
+    {
+      "epoch": 0.75008,
+      "grad_norm": 0.38514742255210876,
+      "learning_rate": 0.00015501278386894908,
+      "loss": 2.9298,
+      "step": 5860
+    },
+    {
+      "epoch": 0.750208,
+      "grad_norm": 0.3661859333515167,
+      "learning_rate": 0.00015486275474642762,
+      "loss": 2.757,
+      "step": 5861
+    },
+    {
+      "epoch": 0.750336,
+      "grad_norm": 0.39096587896347046,
+      "learning_rate": 0.0001547127849569998,
+      "loss": 2.7705,
+      "step": 5862
+    },
+    {
+      "epoch": 0.750464,
+      "grad_norm": 0.42814722657203674,
+      "learning_rate": 0.00015456287452644718,
+      "loss": 2.8094,
+      "step": 5863
+    },
+    {
+      "epoch": 0.750592,
+      "grad_norm": 0.40801912546157837,
+      "learning_rate": 0.00015441302348054103,
+      "loss": 2.7564,
+      "step": 5864
+    },
+    {
+      "epoch": 0.75072,
+      "grad_norm": 0.4190700650215149,
+      "learning_rate": 0.00015426323184504243,
+      "loss": 2.8502,
+      "step": 5865
+    },
+    {
+      "epoch": 0.750848,
+      "grad_norm": 0.37980201840400696,
+      "learning_rate": 0.00015411349964570254,
+      "loss": 2.7793,
+      "step": 5866
+    },
+    {
+      "epoch": 0.750976,
+      "grad_norm": 0.4267800450325012,
+      "learning_rate": 0.00015396382690826172,
+      "loss": 2.7996,
+      "step": 5867
+    },
+    {
+      "epoch": 0.751104,
+      "grad_norm": 0.4189588129520416,
+      "learning_rate": 0.00015381421365845082,
+      "loss": 2.9177,
+      "step": 5868
+    },
+    {
+      "epoch": 0.751232,
+      "grad_norm": 0.3900931477546692,
+      "learning_rate": 0.00015366465992198996,
+      "loss": 2.7288,
+      "step": 5869
+    },
+    {
+      "epoch": 0.75136,
+      "grad_norm": 0.40210217237472534,
+      "learning_rate": 0.00015351516572458912,
+      "loss": 2.8252,
+      "step": 5870
+    },
+    {
+      "epoch": 0.751488,
+      "grad_norm": 0.4134489893913269,
+      "learning_rate": 0.00015336573109194808,
+      "loss": 2.745,
+      "step": 5871
+    },
+    {
+      "epoch": 0.751616,
+      "grad_norm": 0.4125532805919647,
+      "learning_rate": 0.00015321635604975637,
+      "loss": 2.8736,
+      "step": 5872
+    },
+    {
+      "epoch": 0.751744,
+      "grad_norm": 0.4300278127193451,
+      "learning_rate": 0.0001530670406236932,
+      "loss": 2.8141,
+      "step": 5873
+    },
+    {
+      "epoch": 0.751872,
+      "grad_norm": 0.3741309642791748,
+      "learning_rate": 0.0001529177848394277,
+      "loss": 2.7505,
+      "step": 5874
+    },
+    {
+      "epoch": 0.752,
+      "grad_norm": 0.40110012888908386,
+      "learning_rate": 0.0001527685887226189,
+      "loss": 2.7447,
+      "step": 5875
+    },
+    {
+      "epoch": 0.752128,
+      "grad_norm": 0.4271320700645447,
+      "learning_rate": 0.00015261945229891511,
+      "loss": 2.8765,
+      "step": 5876
+    },
+    {
+      "epoch": 0.752256,
+      "grad_norm": 0.49585211277008057,
+      "learning_rate": 0.00015247037559395466,
+      "loss": 2.9315,
+      "step": 5877
+    },
+    {
+      "epoch": 0.752384,
+      "grad_norm": 0.40537694096565247,
+      "learning_rate": 0.00015232135863336554,
+      "loss": 2.7986,
+      "step": 5878
+    },
+    {
+      "epoch": 0.752512,
+      "grad_norm": 0.4373496472835541,
+      "learning_rate": 0.00015217240144276556,
+      "loss": 2.8168,
+      "step": 5879
+    },
+    {
+      "epoch": 0.75264,
+      "grad_norm": 0.43734973669052124,
+      "learning_rate": 0.00015202350404776193,
+      "loss": 2.8707,
+      "step": 5880
+    },
+    {
+      "epoch": 0.752768,
+      "grad_norm": 0.4316136837005615,
+      "learning_rate": 0.00015187466647395226,
+      "loss": 2.8206,
+      "step": 5881
+    },
+    {
+      "epoch": 0.752896,
+      "grad_norm": 0.4474635422229767,
+      "learning_rate": 0.00015172588874692316,
+      "loss": 2.797,
+      "step": 5882
+    },
+    {
+      "epoch": 0.753024,
+      "grad_norm": 0.42602965235710144,
+      "learning_rate": 0.00015157717089225143,
+      "loss": 2.7356,
+      "step": 5883
+    },
+    {
+      "epoch": 0.753152,
+      "grad_norm": 0.39383816719055176,
+      "learning_rate": 0.00015142851293550302,
+      "loss": 2.8469,
+      "step": 5884
+    },
+    {
+      "epoch": 0.75328,
+      "grad_norm": 0.4420165717601776,
+      "learning_rate": 0.00015127991490223447,
+      "loss": 2.8203,
+      "step": 5885
+    },
+    {
+      "epoch": 0.753408,
+      "grad_norm": 0.410137414932251,
+      "learning_rate": 0.0001511313768179912,
+      "loss": 2.7825,
+      "step": 5886
+    },
+    {
+      "epoch": 0.753536,
+      "grad_norm": 0.40445080399513245,
+      "learning_rate": 0.00015098289870830867,
+      "loss": 2.9082,
+      "step": 5887
+    },
+    {
+      "epoch": 0.753664,
+      "grad_norm": 0.3894760310649872,
+      "learning_rate": 0.00015083448059871192,
+      "loss": 2.7581,
+      "step": 5888
+    },
+    {
+      "epoch": 0.753792,
+      "grad_norm": 0.48244160413742065,
+      "learning_rate": 0.00015068612251471593,
+      "loss": 2.7851,
+      "step": 5889
+    },
+    {
+      "epoch": 0.75392,
+      "grad_norm": 0.44337016344070435,
+      "learning_rate": 0.00015053782448182508,
+      "loss": 2.7612,
+      "step": 5890
+    },
+    {
+      "epoch": 0.754048,
+      "grad_norm": 0.42201873660087585,
+      "learning_rate": 0.00015038958652553353,
+      "loss": 2.9441,
+      "step": 5891
+    },
+    {
+      "epoch": 0.754176,
+      "grad_norm": 0.43038463592529297,
+      "learning_rate": 0.00015024140867132507,
+      "loss": 2.8962,
+      "step": 5892
+    },
+    {
+      "epoch": 0.754304,
+      "grad_norm": 0.40909913182258606,
+      "learning_rate": 0.0001500932909446731,
+      "loss": 2.8605,
+      "step": 5893
+    },
+    {
+      "epoch": 0.754432,
+      "grad_norm": 0.43432870507240295,
+      "learning_rate": 0.00014994523337104099,
+      "loss": 2.7879,
+      "step": 5894
+    },
+    {
+      "epoch": 0.75456,
+      "grad_norm": 0.39288169145584106,
+      "learning_rate": 0.0001497972359758813,
+      "loss": 2.8105,
+      "step": 5895
+    },
+    {
+      "epoch": 0.754688,
+      "grad_norm": 0.4388328492641449,
+      "learning_rate": 0.00014964929878463683,
+      "loss": 2.9277,
+      "step": 5896
+    },
+    {
+      "epoch": 0.754816,
+      "grad_norm": 0.3659157454967499,
+      "learning_rate": 0.00014950142182273946,
+      "loss": 2.8308,
+      "step": 5897
+    },
+    {
+      "epoch": 0.754944,
+      "grad_norm": 0.3966682553291321,
+      "learning_rate": 0.0001493536051156111,
+      "loss": 2.7164,
+      "step": 5898
+    },
+    {
+      "epoch": 0.755072,
+      "grad_norm": 0.412962943315506,
+      "learning_rate": 0.00014920584868866293,
+      "loss": 2.8503,
+      "step": 5899
+    },
+    {
+      "epoch": 0.7552,
+      "grad_norm": 0.49873483180999756,
+      "learning_rate": 0.0001490581525672962,
+      "loss": 2.8515,
+      "step": 5900
+    },
+    {
+      "epoch": 0.755328,
+      "grad_norm": 0.42087626457214355,
+      "learning_rate": 0.00014891051677690153,
+      "loss": 2.756,
+      "step": 5901
+    },
+    {
+      "epoch": 0.755456,
+      "grad_norm": 0.39853766560554504,
+      "learning_rate": 0.000148762941342859,
+      "loss": 2.7117,
+      "step": 5902
+    },
+    {
+      "epoch": 0.755584,
+      "grad_norm": 0.42309898138046265,
+      "learning_rate": 0.0001486154262905388,
+      "loss": 2.9443,
+      "step": 5903
+    },
+    {
+      "epoch": 0.755712,
+      "grad_norm": 0.4286082088947296,
+      "learning_rate": 0.00014846797164530051,
+      "loss": 2.7332,
+      "step": 5904
+    },
+    {
+      "epoch": 0.75584,
+      "grad_norm": 0.39802563190460205,
+      "learning_rate": 0.00014832057743249328,
+      "loss": 2.7384,
+      "step": 5905
+    },
+    {
+      "epoch": 0.755968,
+      "grad_norm": 0.4286649823188782,
+      "learning_rate": 0.00014817324367745572,
+      "loss": 2.8327,
+      "step": 5906
+    },
+    {
+      "epoch": 0.756096,
+      "grad_norm": 0.4134475290775299,
+      "learning_rate": 0.00014802597040551634,
+      "loss": 2.8021,
+      "step": 5907
+    },
+    {
+      "epoch": 0.756224,
+      "grad_norm": 0.417796790599823,
+      "learning_rate": 0.0001478787576419931,
+      "loss": 2.891,
+      "step": 5908
+    },
+    {
+      "epoch": 0.756352,
+      "grad_norm": 0.3942068815231323,
+      "learning_rate": 0.00014773160541219337,
+      "loss": 2.7497,
+      "step": 5909
+    },
+    {
+      "epoch": 0.75648,
+      "grad_norm": 0.4498046338558197,
+      "learning_rate": 0.00014758451374141467,
+      "loss": 2.8884,
+      "step": 5910
+    },
+    {
+      "epoch": 0.756608,
+      "grad_norm": 0.42494750022888184,
+      "learning_rate": 0.0001474374826549435,
+      "loss": 2.8214,
+      "step": 5911
+    },
+    {
+      "epoch": 0.756736,
+      "grad_norm": 0.4134047031402588,
+      "learning_rate": 0.00014729051217805646,
+      "loss": 2.7157,
+      "step": 5912
+    },
+    {
+      "epoch": 0.756864,
+      "grad_norm": 0.40388238430023193,
+      "learning_rate": 0.0001471436023360193,
+      "loss": 2.7818,
+      "step": 5913
+    },
+    {
+      "epoch": 0.756992,
+      "grad_norm": 0.3754107356071472,
+      "learning_rate": 0.00014699675315408756,
+      "loss": 2.8223,
+      "step": 5914
+    },
+    {
+      "epoch": 0.75712,
+      "grad_norm": 0.41532424092292786,
+      "learning_rate": 0.00014684996465750638,
+      "loss": 2.8722,
+      "step": 5915
+    },
+    {
+      "epoch": 0.757248,
+      "grad_norm": 0.4237290918827057,
+      "learning_rate": 0.0001467032368715101,
+      "loss": 2.8265,
+      "step": 5916
+    },
+    {
+      "epoch": 0.757376,
+      "grad_norm": 0.4301643371582031,
+      "learning_rate": 0.00014655656982132337,
+      "loss": 2.8573,
+      "step": 5917
+    },
+    {
+      "epoch": 0.757504,
+      "grad_norm": 0.40114426612854004,
+      "learning_rate": 0.0001464099635321598,
+      "loss": 2.8207,
+      "step": 5918
+    },
+    {
+      "epoch": 0.757632,
+      "grad_norm": 0.38689354062080383,
+      "learning_rate": 0.0001462634180292226,
+      "loss": 2.7961,
+      "step": 5919
+    },
+    {
+      "epoch": 0.75776,
+      "grad_norm": 0.3908846080303192,
+      "learning_rate": 0.0001461169333377047,
+      "loss": 2.8184,
+      "step": 5920
+    },
+    {
+      "epoch": 0.757888,
+      "grad_norm": 0.4077727496623993,
+      "learning_rate": 0.00014597050948278838,
+      "loss": 2.8911,
+      "step": 5921
+    },
+    {
+      "epoch": 0.758016,
+      "grad_norm": 0.4184965193271637,
+      "learning_rate": 0.00014582414648964593,
+      "loss": 2.8535,
+      "step": 5922
+    },
+    {
+      "epoch": 0.758144,
+      "grad_norm": 0.42283907532691956,
+      "learning_rate": 0.00014567784438343867,
+      "loss": 2.7909,
+      "step": 5923
+    },
+    {
+      "epoch": 0.758272,
+      "grad_norm": 0.3967253863811493,
+      "learning_rate": 0.0001455316031893174,
+      "loss": 2.6993,
+      "step": 5924
+    },
+    {
+      "epoch": 0.7584,
+      "grad_norm": 0.37783220410346985,
+      "learning_rate": 0.00014538542293242307,
+      "loss": 2.79,
+      "step": 5925
+    },
+    {
+      "epoch": 0.758528,
+      "grad_norm": 0.3978289067745209,
+      "learning_rate": 0.0001452393036378856,
+      "loss": 2.8905,
+      "step": 5926
+    },
+    {
+      "epoch": 0.758656,
+      "grad_norm": 0.40497732162475586,
+      "learning_rate": 0.00014509324533082459,
+      "loss": 2.7142,
+      "step": 5927
+    },
+    {
+      "epoch": 0.758784,
+      "grad_norm": 0.4095357060432434,
+      "learning_rate": 0.00014494724803634912,
+      "loss": 2.9394,
+      "step": 5928
+    },
+    {
+      "epoch": 0.758912,
+      "grad_norm": 0.4381791353225708,
+      "learning_rate": 0.0001448013117795578,
+      "loss": 2.8499,
+      "step": 5929
+    },
+    {
+      "epoch": 0.75904,
+      "grad_norm": 0.39493528008461,
+      "learning_rate": 0.0001446554365855387,
+      "loss": 2.774,
+      "step": 5930
+    },
+    {
+      "epoch": 0.759168,
+      "grad_norm": 0.4231765568256378,
+      "learning_rate": 0.0001445096224793695,
+      "loss": 2.8753,
+      "step": 5931
+    },
+    {
+      "epoch": 0.759296,
+      "grad_norm": 0.4194122552871704,
+      "learning_rate": 0.00014436386948611763,
+      "loss": 2.86,
+      "step": 5932
+    },
+    {
+      "epoch": 0.759424,
+      "grad_norm": 0.423678457736969,
+      "learning_rate": 0.0001442181776308394,
+      "loss": 2.8986,
+      "step": 5933
+    },
+    {
+      "epoch": 0.759552,
+      "grad_norm": 0.4010797441005707,
+      "learning_rate": 0.00014407254693858107,
+      "loss": 2.8284,
+      "step": 5934
+    },
+    {
+      "epoch": 0.75968,
+      "grad_norm": 0.3999691605567932,
+      "learning_rate": 0.00014392697743437815,
+      "loss": 2.8897,
+      "step": 5935
+    },
+    {
+      "epoch": 0.759808,
+      "grad_norm": 0.38598525524139404,
+      "learning_rate": 0.0001437814691432558,
+      "loss": 2.7034,
+      "step": 5936
+    },
+    {
+      "epoch": 0.759936,
+      "grad_norm": 0.4332312345504761,
+      "learning_rate": 0.00014363602209022835,
+      "loss": 2.9126,
+      "step": 5937
+    },
+    {
+      "epoch": 0.760064,
+      "grad_norm": 0.3868039548397064,
+      "learning_rate": 0.00014349063630030018,
+      "loss": 2.688,
+      "step": 5938
+    },
+    {
+      "epoch": 0.760192,
+      "grad_norm": 0.4024573862552643,
+      "learning_rate": 0.00014334531179846456,
+      "loss": 2.7601,
+      "step": 5939
+    },
+    {
+      "epoch": 0.76032,
+      "grad_norm": 0.39352214336395264,
+      "learning_rate": 0.0001432000486097046,
+      "loss": 2.824,
+      "step": 5940
+    },
+    {
+      "epoch": 0.760448,
+      "grad_norm": 0.40012630820274353,
+      "learning_rate": 0.00014305484675899272,
+      "loss": 2.867,
+      "step": 5941
+    },
+    {
+      "epoch": 0.760576,
+      "grad_norm": 0.40936949849128723,
+      "learning_rate": 0.00014290970627129073,
+      "loss": 2.8704,
+      "step": 5942
+    },
+    {
+      "epoch": 0.760704,
+      "grad_norm": 0.3880345821380615,
+      "learning_rate": 0.00014276462717154997,
+      "loss": 2.714,
+      "step": 5943
+    },
+    {
+      "epoch": 0.760832,
+      "grad_norm": 0.3912620544433594,
+      "learning_rate": 0.0001426196094847112,
+      "loss": 2.8394,
+      "step": 5944
+    },
+    {
+      "epoch": 0.76096,
+      "grad_norm": 0.39043429493904114,
+      "learning_rate": 0.0001424746532357046,
+      "loss": 2.8357,
+      "step": 5945
+    },
+    {
+      "epoch": 0.761088,
+      "grad_norm": 0.41549381613731384,
+      "learning_rate": 0.00014232975844944996,
+      "loss": 2.7772,
+      "step": 5946
+    },
+    {
+      "epoch": 0.761216,
+      "grad_norm": 0.4077136516571045,
+      "learning_rate": 0.00014218492515085635,
+      "loss": 2.8963,
+      "step": 5947
+    },
+    {
+      "epoch": 0.761344,
+      "grad_norm": 0.4121875464916229,
+      "learning_rate": 0.00014204015336482213,
+      "loss": 2.8955,
+      "step": 5948
+    },
+    {
+      "epoch": 0.761472,
+      "grad_norm": 0.402174711227417,
+      "learning_rate": 0.00014189544311623547,
+      "loss": 2.9282,
+      "step": 5949
+    },
+    {
+      "epoch": 0.7616,
+      "grad_norm": 0.3802185654640198,
+      "learning_rate": 0.00014175079442997369,
+      "loss": 2.6881,
+      "step": 5950
+    },
+    {
+      "epoch": 0.761728,
+      "grad_norm": 0.4174196720123291,
+      "learning_rate": 0.00014160620733090352,
+      "loss": 2.7804,
+      "step": 5951
+    },
+    {
+      "epoch": 0.761856,
+      "grad_norm": 0.4149150252342224,
+      "learning_rate": 0.00014146168184388098,
+      "loss": 2.7902,
+      "step": 5952
+    },
+    {
+      "epoch": 0.761984,
+      "grad_norm": 0.4356066584587097,
+      "learning_rate": 0.00014131721799375208,
+      "loss": 2.8318,
+      "step": 5953
+    },
+    {
+      "epoch": 0.762112,
+      "grad_norm": 0.3929494023323059,
+      "learning_rate": 0.0001411728158053516,
+      "loss": 2.8063,
+      "step": 5954
+    },
+    {
+      "epoch": 0.76224,
+      "grad_norm": 0.3987269103527069,
+      "learning_rate": 0.00014102847530350388,
+      "loss": 2.8373,
+      "step": 5955
+    },
+    {
+      "epoch": 0.762368,
+      "grad_norm": 0.38571178913116455,
+      "learning_rate": 0.00014088419651302287,
+      "loss": 2.8119,
+      "step": 5956
+    },
+    {
+      "epoch": 0.762496,
+      "grad_norm": 0.37068673968315125,
+      "learning_rate": 0.0001407399794587117,
+      "loss": 2.7742,
+      "step": 5957
+    },
+    {
+      "epoch": 0.762624,
+      "grad_norm": 0.3943469226360321,
+      "learning_rate": 0.0001405958241653628,
+      "loss": 2.8425,
+      "step": 5958
+    },
+    {
+      "epoch": 0.762752,
+      "grad_norm": 0.42290255427360535,
+      "learning_rate": 0.0001404517306577585,
+      "loss": 2.9782,
+      "step": 5959
+    },
+    {
+      "epoch": 0.76288,
+      "grad_norm": 0.39745959639549255,
+      "learning_rate": 0.00014030769896066975,
+      "loss": 2.7363,
+      "step": 5960
+    },
+    {
+      "epoch": 0.763008,
+      "grad_norm": 0.4134059548377991,
+      "learning_rate": 0.0001401637290988576,
+      "loss": 2.852,
+      "step": 5961
+    },
+    {
+      "epoch": 0.763136,
+      "grad_norm": 0.35809779167175293,
+      "learning_rate": 0.000140019821097072,
+      "loss": 2.5617,
+      "step": 5962
+    },
+    {
+      "epoch": 0.763264,
+      "grad_norm": 0.4200590252876282,
+      "learning_rate": 0.00013987597498005245,
+      "loss": 2.9247,
+      "step": 5963
+    },
+    {
+      "epoch": 0.763392,
+      "grad_norm": 0.38802197575569153,
+      "learning_rate": 0.00013973219077252768,
+      "loss": 2.8007,
+      "step": 5964
+    },
+    {
+      "epoch": 0.76352,
+      "grad_norm": 0.38669535517692566,
+      "learning_rate": 0.0001395884684992159,
+      "loss": 2.7901,
+      "step": 5965
+    },
+    {
+      "epoch": 0.763648,
+      "grad_norm": 0.431052029132843,
+      "learning_rate": 0.00013944480818482447,
+      "loss": 2.7414,
+      "step": 5966
+    },
+    {
+      "epoch": 0.763776,
+      "grad_norm": 0.4163278043270111,
+      "learning_rate": 0.0001393012098540505,
+      "loss": 2.7865,
+      "step": 5967
+    },
+    {
+      "epoch": 0.763904,
+      "grad_norm": 0.4216376841068268,
+      "learning_rate": 0.0001391576735315802,
+      "loss": 2.8057,
+      "step": 5968
+    },
+    {
+      "epoch": 0.764032,
+      "grad_norm": 0.3904905915260315,
+      "learning_rate": 0.00013901419924208908,
+      "loss": 2.8929,
+      "step": 5969
+    },
+    {
+      "epoch": 0.76416,
+      "grad_norm": 0.4145306646823883,
+      "learning_rate": 0.00013887078701024203,
+      "loss": 2.7639,
+      "step": 5970
+    },
+    {
+      "epoch": 0.764288,
+      "grad_norm": 0.4025112986564636,
+      "learning_rate": 0.00013872743686069327,
+      "loss": 2.8859,
+      "step": 5971
+    },
+    {
+      "epoch": 0.764416,
+      "grad_norm": 0.4066394567489624,
+      "learning_rate": 0.00013858414881808635,
+      "loss": 2.7264,
+      "step": 5972
+    },
+    {
+      "epoch": 0.764544,
+      "grad_norm": 0.39254534244537354,
+      "learning_rate": 0.00013844092290705396,
+      "loss": 2.8103,
+      "step": 5973
+    },
+    {
+      "epoch": 0.764672,
+      "grad_norm": 0.3794867694377899,
+      "learning_rate": 0.0001382977591522186,
+      "loss": 2.8375,
+      "step": 5974
+    },
+    {
+      "epoch": 0.7648,
+      "grad_norm": 0.39079567790031433,
+      "learning_rate": 0.00013815465757819175,
+      "loss": 2.8682,
+      "step": 5975
+    },
+    {
+      "epoch": 0.764928,
+      "grad_norm": 0.42680326104164124,
+      "learning_rate": 0.00013801161820957387,
+      "loss": 2.841,
+      "step": 5976
+    },
+    {
+      "epoch": 0.765056,
+      "grad_norm": 0.413564532995224,
+      "learning_rate": 0.0001378686410709556,
+      "loss": 2.7397,
+      "step": 5977
+    },
+    {
+      "epoch": 0.765184,
+      "grad_norm": 0.4080916941165924,
+      "learning_rate": 0.00013772572618691602,
+      "loss": 2.7479,
+      "step": 5978
+    },
+    {
+      "epoch": 0.765312,
+      "grad_norm": 0.44618213176727295,
+      "learning_rate": 0.00013758287358202405,
+      "loss": 2.9502,
+      "step": 5979
+    },
+    {
+      "epoch": 0.76544,
+      "grad_norm": 0.39370861649513245,
+      "learning_rate": 0.00013744008328083756,
+      "loss": 2.7587,
+      "step": 5980
+    },
+    {
+      "epoch": 0.765568,
+      "grad_norm": 0.4196154475212097,
+      "learning_rate": 0.00013729735530790377,
+      "loss": 2.8396,
+      "step": 5981
+    },
+    {
+      "epoch": 0.765696,
+      "grad_norm": 0.41972261667251587,
+      "learning_rate": 0.00013715468968775952,
+      "loss": 2.8481,
+      "step": 5982
+    },
+    {
+      "epoch": 0.765824,
+      "grad_norm": 0.4157065451145172,
+      "learning_rate": 0.00013701208644493064,
+      "loss": 2.6764,
+      "step": 5983
+    },
+    {
+      "epoch": 0.765952,
+      "grad_norm": 0.415440171957016,
+      "learning_rate": 0.00013686954560393217,
+      "loss": 2.8669,
+      "step": 5984
+    },
+    {
+      "epoch": 0.76608,
+      "grad_norm": 0.3725963532924652,
+      "learning_rate": 0.00013672706718926848,
+      "loss": 2.8249,
+      "step": 5985
+    },
+    {
+      "epoch": 0.766208,
+      "grad_norm": 0.3974570631980896,
+      "learning_rate": 0.00013658465122543345,
+      "loss": 2.8445,
+      "step": 5986
+    },
+    {
+      "epoch": 0.766336,
+      "grad_norm": 0.4035356044769287,
+      "learning_rate": 0.00013644229773690996,
+      "loss": 2.7157,
+      "step": 5987
+    },
+    {
+      "epoch": 0.766464,
+      "grad_norm": 0.40849825739860535,
+      "learning_rate": 0.00013630000674817012,
+      "loss": 2.8252,
+      "step": 5988
+    },
+    {
+      "epoch": 0.766592,
+      "grad_norm": 0.7513207197189331,
+      "learning_rate": 0.0001361577782836756,
+      "loss": 2.9826,
+      "step": 5989
+    },
+    {
+      "epoch": 0.76672,
+      "grad_norm": 0.41924604773521423,
+      "learning_rate": 0.00013601561236787702,
+      "loss": 2.8435,
+      "step": 5990
+    },
+    {
+      "epoch": 0.766848,
+      "grad_norm": 0.39821961522102356,
+      "learning_rate": 0.00013587350902521435,
+      "loss": 2.7814,
+      "step": 5991
+    },
+    {
+      "epoch": 0.766976,
+      "grad_norm": 0.39678099751472473,
+      "learning_rate": 0.0001357314682801168,
+      "loss": 2.7716,
+      "step": 5992
+    },
+    {
+      "epoch": 0.767104,
+      "grad_norm": 0.3956705331802368,
+      "learning_rate": 0.00013558949015700277,
+      "loss": 2.8408,
+      "step": 5993
+    },
+    {
+      "epoch": 0.767232,
+      "grad_norm": 0.38960060477256775,
+      "learning_rate": 0.00013544757468027984,
+      "loss": 2.7969,
+      "step": 5994
+    },
+    {
+      "epoch": 0.76736,
+      "grad_norm": 0.40965479612350464,
+      "learning_rate": 0.0001353057218743453,
+      "loss": 2.9156,
+      "step": 5995
+    },
+    {
+      "epoch": 0.767488,
+      "grad_norm": 0.4210774302482605,
+      "learning_rate": 0.00013516393176358487,
+      "loss": 2.8556,
+      "step": 5996
+    },
+    {
+      "epoch": 0.767616,
+      "grad_norm": 0.3650226294994354,
+      "learning_rate": 0.00013502220437237427,
+      "loss": 2.7651,
+      "step": 5997
+    },
+    {
+      "epoch": 0.767744,
+      "grad_norm": 0.4121313989162445,
+      "learning_rate": 0.0001348805397250779,
+      "loss": 2.8322,
+      "step": 5998
+    },
+    {
+      "epoch": 0.767872,
+      "grad_norm": 0.39417797327041626,
+      "learning_rate": 0.00013473893784604962,
+      "loss": 2.7728,
+      "step": 5999
+    },
+    {
+      "epoch": 0.768,
+      "grad_norm": 0.3893483877182007,
+      "learning_rate": 0.00013459739875963238,
+      "loss": 2.8388,
+      "step": 6000
+    },
+    {
+      "epoch": 0.768128,
+      "grad_norm": 0.39093202352523804,
+      "learning_rate": 0.0001344559224901584,
+      "loss": 2.6667,
+      "step": 6001
+    },
+    {
+      "epoch": 0.768256,
+      "grad_norm": 0.4114723205566406,
+      "learning_rate": 0.00013431450906194891,
+      "loss": 2.8044,
+      "step": 6002
+    },
+    {
+      "epoch": 0.768384,
+      "grad_norm": 0.4283005893230438,
+      "learning_rate": 0.00013417315849931493,
+      "loss": 2.7828,
+      "step": 6003
+    },
+    {
+      "epoch": 0.768512,
+      "grad_norm": 0.4123353660106659,
+      "learning_rate": 0.00013403187082655584,
+      "loss": 2.7977,
+      "step": 6004
+    },
+    {
+      "epoch": 0.76864,
+      "grad_norm": 0.4027843773365021,
+      "learning_rate": 0.00013389064606796096,
+      "loss": 2.8908,
+      "step": 6005
+    },
+    {
+      "epoch": 0.768768,
+      "grad_norm": 0.45087727904319763,
+      "learning_rate": 0.00013374948424780837,
+      "loss": 2.7851,
+      "step": 6006
+    },
+    {
+      "epoch": 0.768896,
+      "grad_norm": 0.39384204149246216,
+      "learning_rate": 0.0001336083853903653,
+      "loss": 2.8252,
+      "step": 6007
+    },
+    {
+      "epoch": 0.769024,
+      "grad_norm": 0.46835315227508545,
+      "learning_rate": 0.00013346734951988842,
+      "loss": 2.8402,
+      "step": 6008
+    },
+    {
+      "epoch": 0.769152,
+      "grad_norm": 0.402744859457016,
+      "learning_rate": 0.0001333263766606232,
+      "loss": 3.014,
+      "step": 6009
+    },
+    {
+      "epoch": 0.76928,
+      "grad_norm": 0.433898001909256,
+      "learning_rate": 0.00013318546683680483,
+      "loss": 2.82,
+      "step": 6010
+    },
+    {
+      "epoch": 0.769408,
+      "grad_norm": 0.40552106499671936,
+      "learning_rate": 0.00013304462007265717,
+      "loss": 2.9012,
+      "step": 6011
+    },
+    {
+      "epoch": 0.769536,
+      "grad_norm": 0.39013248682022095,
+      "learning_rate": 0.00013290383639239344,
+      "loss": 2.7678,
+      "step": 6012
+    },
+    {
+      "epoch": 0.769664,
+      "grad_norm": 0.41235145926475525,
+      "learning_rate": 0.00013276311582021582,
+      "loss": 2.7357,
+      "step": 6013
+    },
+    {
+      "epoch": 0.769792,
+      "grad_norm": 0.4105885922908783,
+      "learning_rate": 0.00013262245838031617,
+      "loss": 2.7397,
+      "step": 6014
+    },
+    {
+      "epoch": 0.76992,
+      "grad_norm": 0.3896887004375458,
+      "learning_rate": 0.0001324818640968749,
+      "loss": 2.7762,
+      "step": 6015
+    },
+    {
+      "epoch": 0.770048,
+      "grad_norm": 0.4382360577583313,
+      "learning_rate": 0.0001323413329940618,
+      "loss": 2.8311,
+      "step": 6016
+    },
+    {
+      "epoch": 0.770176,
+      "grad_norm": 0.3980408310890198,
+      "learning_rate": 0.00013220086509603567,
+      "loss": 2.8952,
+      "step": 6017
+    },
+    {
+      "epoch": 0.770304,
+      "grad_norm": 0.4382908344268799,
+      "learning_rate": 0.00013206046042694492,
+      "loss": 2.7661,
+      "step": 6018
+    },
+    {
+      "epoch": 0.770432,
+      "grad_norm": 0.41048333048820496,
+      "learning_rate": 0.00013192011901092655,
+      "loss": 2.752,
+      "step": 6019
+    },
+    {
+      "epoch": 0.77056,
+      "grad_norm": 0.44801875948905945,
+      "learning_rate": 0.0001317798408721068,
+      "loss": 2.6398,
+      "step": 6020
+    },
+    {
+      "epoch": 0.770688,
+      "grad_norm": 0.3827441930770874,
+      "learning_rate": 0.0001316396260346012,
+      "loss": 2.8601,
+      "step": 6021
+    },
+    {
+      "epoch": 0.770816,
+      "grad_norm": 0.39868560433387756,
+      "learning_rate": 0.0001314994745225142,
+      "loss": 2.7972,
+      "step": 6022
+    },
+    {
+      "epoch": 0.770944,
+      "grad_norm": 0.3910103142261505,
+      "learning_rate": 0.00013135938635993965,
+      "loss": 2.7192,
+      "step": 6023
+    },
+    {
+      "epoch": 0.771072,
+      "grad_norm": 0.41086655855178833,
+      "learning_rate": 0.0001312193615709601,
+      "loss": 2.8062,
+      "step": 6024
+    },
+    {
+      "epoch": 0.7712,
+      "grad_norm": 0.3916473984718323,
+      "learning_rate": 0.0001310794001796477,
+      "loss": 2.7557,
+      "step": 6025
+    },
+    {
+      "epoch": 0.771328,
+      "grad_norm": 0.41944679617881775,
+      "learning_rate": 0.00013093950221006328,
+      "loss": 2.7253,
+      "step": 6026
+    },
+    {
+      "epoch": 0.771456,
+      "grad_norm": 0.4071193337440491,
+      "learning_rate": 0.000130799667686257,
+      "loss": 2.8558,
+      "step": 6027
+    },
+    {
+      "epoch": 0.771584,
+      "grad_norm": 0.4105038344860077,
+      "learning_rate": 0.00013065989663226795,
+      "loss": 2.9638,
+      "step": 6028
+    },
+    {
+      "epoch": 0.771712,
+      "grad_norm": 0.439438134431839,
+      "learning_rate": 0.00013052018907212448,
+      "loss": 2.8166,
+      "step": 6029
+    },
+    {
+      "epoch": 0.77184,
+      "grad_norm": 0.4512619078159332,
+      "learning_rate": 0.0001303805450298437,
+      "loss": 2.8682,
+      "step": 6030
+    },
+    {
+      "epoch": 0.771968,
+      "grad_norm": 0.408778578042984,
+      "learning_rate": 0.00013024096452943234,
+      "loss": 2.7988,
+      "step": 6031
+    },
+    {
+      "epoch": 0.772096,
+      "grad_norm": 0.40609923005104065,
+      "learning_rate": 0.00013010144759488567,
+      "loss": 2.7875,
+      "step": 6032
+    },
+    {
+      "epoch": 0.772224,
+      "grad_norm": 0.38135001063346863,
+      "learning_rate": 0.00012996199425018855,
+      "loss": 2.8024,
+      "step": 6033
+    },
+    {
+      "epoch": 0.772352,
+      "grad_norm": 0.4277607500553131,
+      "learning_rate": 0.00012982260451931444,
+      "loss": 2.8445,
+      "step": 6034
+    },
+    {
+      "epoch": 0.77248,
+      "grad_norm": 0.40761274099349976,
+      "learning_rate": 0.00012968327842622612,
+      "loss": 2.7798,
+      "step": 6035
+    },
+    {
+      "epoch": 0.772608,
+      "grad_norm": 0.436018168926239,
+      "learning_rate": 0.0001295440159948753,
+      "loss": 2.769,
+      "step": 6036
+    },
+    {
+      "epoch": 0.772736,
+      "grad_norm": 0.4109160602092743,
+      "learning_rate": 0.00012940481724920284,
+      "loss": 2.7277,
+      "step": 6037
+    },
+    {
+      "epoch": 0.772864,
+      "grad_norm": 0.41091522574424744,
+      "learning_rate": 0.0001292656822131384,
+      "loss": 2.9243,
+      "step": 6038
+    },
+    {
+      "epoch": 0.772992,
+      "grad_norm": 0.4096376597881317,
+      "learning_rate": 0.0001291266109106013,
+      "loss": 2.941,
+      "step": 6039
+    },
+    {
+      "epoch": 0.77312,
+      "grad_norm": 0.4032200872898102,
+      "learning_rate": 0.0001289876033654993,
+      "loss": 2.7523,
+      "step": 6040
+    },
+    {
+      "epoch": 0.773248,
+      "grad_norm": 0.39408156275749207,
+      "learning_rate": 0.0001288486596017293,
+      "loss": 2.8583,
+      "step": 6041
+    },
+    {
+      "epoch": 0.773376,
+      "grad_norm": 0.3884097635746002,
+      "learning_rate": 0.0001287097796431776,
+      "loss": 2.817,
+      "step": 6042
+    },
+    {
+      "epoch": 0.773504,
+      "grad_norm": 0.4508708417415619,
+      "learning_rate": 0.00012857096351371917,
+      "loss": 2.851,
+      "step": 6043
+    },
+    {
+      "epoch": 0.773632,
+      "grad_norm": 0.4151083528995514,
+      "learning_rate": 0.00012843221123721803,
+      "loss": 2.7844,
+      "step": 6044
+    },
+    {
+      "epoch": 0.77376,
+      "grad_norm": 0.39166751503944397,
+      "learning_rate": 0.00012829352283752726,
+      "loss": 2.779,
+      "step": 6045
+    },
+    {
+      "epoch": 0.773888,
+      "grad_norm": 0.426025927066803,
+      "learning_rate": 0.00012815489833848927,
+      "loss": 2.8964,
+      "step": 6046
+    },
+    {
+      "epoch": 0.774016,
+      "grad_norm": 0.40899956226348877,
+      "learning_rate": 0.00012801633776393501,
+      "loss": 2.6457,
+      "step": 6047
+    },
+    {
+      "epoch": 0.774144,
+      "grad_norm": 0.37369462847709656,
+      "learning_rate": 0.0001278778411376847,
+      "loss": 2.7688,
+      "step": 6048
+    },
+    {
+      "epoch": 0.774272,
+      "grad_norm": 0.41010692715644836,
+      "learning_rate": 0.00012773940848354753,
+      "loss": 2.8428,
+      "step": 6049
+    },
+    {
+      "epoch": 0.7744,
+      "grad_norm": 0.4555782675743103,
+      "learning_rate": 0.0001276010398253215,
+      "loss": 2.8981,
+      "step": 6050
+    },
+    {
+      "epoch": 0.774528,
+      "grad_norm": 0.4050312340259552,
+      "learning_rate": 0.00012746273518679407,
+      "loss": 2.8893,
+      "step": 6051
+    },
+    {
+      "epoch": 0.774656,
+      "grad_norm": 0.42105546593666077,
+      "learning_rate": 0.00012732449459174134,
+      "loss": 2.6556,
+      "step": 6052
+    },
+    {
+      "epoch": 0.774784,
+      "grad_norm": 0.39184340834617615,
+      "learning_rate": 0.00012718631806392822,
+      "loss": 2.8033,
+      "step": 6053
+    },
+    {
+      "epoch": 0.774912,
+      "grad_norm": 0.43462711572647095,
+      "learning_rate": 0.00012704820562710922,
+      "loss": 2.7555,
+      "step": 6054
+    },
+    {
+      "epoch": 0.77504,
+      "grad_norm": 0.4140165448188782,
+      "learning_rate": 0.0001269101573050273,
+      "loss": 2.8916,
+      "step": 6055
+    },
+    {
+      "epoch": 0.775168,
+      "grad_norm": 0.3568967580795288,
+      "learning_rate": 0.00012677217312141454,
+      "loss": 2.7437,
+      "step": 6056
+    },
+    {
+      "epoch": 0.775296,
+      "grad_norm": 0.40950343012809753,
+      "learning_rate": 0.00012663425309999204,
+      "loss": 2.7527,
+      "step": 6057
+    },
+    {
+      "epoch": 0.775424,
+      "grad_norm": 0.4078839421272278,
+      "learning_rate": 0.00012649639726446994,
+      "loss": 2.8362,
+      "step": 6058
+    },
+    {
+      "epoch": 0.775552,
+      "grad_norm": 0.4016913175582886,
+      "learning_rate": 0.00012635860563854696,
+      "loss": 2.865,
+      "step": 6059
+    },
+    {
+      "epoch": 0.77568,
+      "grad_norm": 0.40343305468559265,
+      "learning_rate": 0.00012622087824591127,
+      "loss": 2.6962,
+      "step": 6060
+    },
+    {
+      "epoch": 0.775808,
+      "grad_norm": 0.4474470615386963,
+      "learning_rate": 0.00012608321511024006,
+      "loss": 2.8458,
+      "step": 6061
+    },
+    {
+      "epoch": 0.775936,
+      "grad_norm": 0.4026321470737457,
+      "learning_rate": 0.0001259456162551989,
+      "loss": 2.7819,
+      "step": 6062
+    },
+    {
+      "epoch": 0.776064,
+      "grad_norm": 0.39829570055007935,
+      "learning_rate": 0.0001258080817044428,
+      "loss": 2.8276,
+      "step": 6063
+    },
+    {
+      "epoch": 0.776192,
+      "grad_norm": 0.40230169892311096,
+      "learning_rate": 0.0001256706114816154,
+      "loss": 2.8437,
+      "step": 6064
+    },
+    {
+      "epoch": 0.77632,
+      "grad_norm": 0.4012216031551361,
+      "learning_rate": 0.00012553320561034954,
+      "loss": 2.7288,
+      "step": 6065
+    },
+    {
+      "epoch": 0.776448,
+      "grad_norm": 0.4057473838329315,
+      "learning_rate": 0.00012539586411426662,
+      "loss": 2.8523,
+      "step": 6066
+    },
+    {
+      "epoch": 0.776576,
+      "grad_norm": 0.4126920700073242,
+      "learning_rate": 0.00012525858701697761,
+      "loss": 2.7246,
+      "step": 6067
+    },
+    {
+      "epoch": 0.776704,
+      "grad_norm": 0.39727532863616943,
+      "learning_rate": 0.00012512137434208187,
+      "loss": 2.8513,
+      "step": 6068
+    },
+    {
+      "epoch": 0.776832,
+      "grad_norm": 0.4111011028289795,
+      "learning_rate": 0.00012498422611316768,
+      "loss": 2.6724,
+      "step": 6069
+    },
+    {
+      "epoch": 0.77696,
+      "grad_norm": 0.41258302330970764,
+      "learning_rate": 0.00012484714235381277,
+      "loss": 2.6564,
+      "step": 6070
+    },
+    {
+      "epoch": 0.777088,
+      "grad_norm": 0.4380761384963989,
+      "learning_rate": 0.00012471012308758322,
+      "loss": 2.7171,
+      "step": 6071
+    },
+    {
+      "epoch": 0.777216,
+      "grad_norm": 0.4151173532009125,
+      "learning_rate": 0.00012457316833803421,
+      "loss": 2.7186,
+      "step": 6072
+    },
+    {
+      "epoch": 0.777344,
+      "grad_norm": 0.4145059585571289,
+      "learning_rate": 0.00012443627812870988,
+      "loss": 2.7637,
+      "step": 6073
+    },
+    {
+      "epoch": 0.777472,
+      "grad_norm": 0.40431836247444153,
+      "learning_rate": 0.00012429945248314318,
+      "loss": 2.8762,
+      "step": 6074
+    },
+    {
+      "epoch": 0.7776,
+      "grad_norm": 0.3775599002838135,
+      "learning_rate": 0.00012416269142485615,
+      "loss": 2.6608,
+      "step": 6075
+    },
+    {
+      "epoch": 0.777728,
+      "grad_norm": 0.39286768436431885,
+      "learning_rate": 0.0001240259949773596,
+      "loss": 2.7184,
+      "step": 6076
+    },
+    {
+      "epoch": 0.777856,
+      "grad_norm": 0.4018600881099701,
+      "learning_rate": 0.00012388936316415317,
+      "loss": 2.8312,
+      "step": 6077
+    },
+    {
+      "epoch": 0.777984,
+      "grad_norm": 0.41900449991226196,
+      "learning_rate": 0.0001237527960087253,
+      "loss": 2.7903,
+      "step": 6078
+    },
+    {
+      "epoch": 0.778112,
+      "grad_norm": 0.379036545753479,
+      "learning_rate": 0.0001236162935345538,
+      "loss": 2.7989,
+      "step": 6079
+    },
+    {
+      "epoch": 0.77824,
+      "grad_norm": 0.42241859436035156,
+      "learning_rate": 0.0001234798557651048,
+      "loss": 2.9467,
+      "step": 6080
+    },
+    {
+      "epoch": 0.778368,
+      "grad_norm": 0.41191306710243225,
+      "learning_rate": 0.0001233434827238335,
+      "loss": 2.7819,
+      "step": 6081
+    },
+    {
+      "epoch": 0.778496,
+      "grad_norm": 0.3891913890838623,
+      "learning_rate": 0.00012320717443418423,
+      "loss": 2.6988,
+      "step": 6082
+    },
+    {
+      "epoch": 0.778624,
+      "grad_norm": 0.4111056923866272,
+      "learning_rate": 0.0001230709309195898,
+      "loss": 2.8466,
+      "step": 6083
+    },
+    {
+      "epoch": 0.778752,
+      "grad_norm": 0.36927661299705505,
+      "learning_rate": 0.00012293475220347212,
+      "loss": 2.8038,
+      "step": 6084
+    },
+    {
+      "epoch": 0.77888,
+      "grad_norm": 0.4115481674671173,
+      "learning_rate": 0.0001227986383092418,
+      "loss": 2.7679,
+      "step": 6085
+    },
+    {
+      "epoch": 0.779008,
+      "grad_norm": 0.4021357297897339,
+      "learning_rate": 0.0001226625892602985,
+      "loss": 2.8143,
+      "step": 6086
+    },
+    {
+      "epoch": 0.779136,
+      "grad_norm": 0.3970385789871216,
+      "learning_rate": 0.00012252660508003043,
+      "loss": 2.7802,
+      "step": 6087
+    },
+    {
+      "epoch": 0.779264,
+      "grad_norm": 0.5057937502861023,
+      "learning_rate": 0.00012239068579181494,
+      "loss": 2.629,
+      "step": 6088
+    },
+    {
+      "epoch": 0.779392,
+      "grad_norm": 0.39999157190322876,
+      "learning_rate": 0.00012225483141901837,
+      "loss": 2.8423,
+      "step": 6089
+    },
+    {
+      "epoch": 0.77952,
+      "grad_norm": 0.4192769527435303,
+      "learning_rate": 0.00012211904198499552,
+      "loss": 2.8029,
+      "step": 6090
+    },
+    {
+      "epoch": 0.779648,
+      "grad_norm": 0.49377232789993286,
+      "learning_rate": 0.00012198331751309005,
+      "loss": 2.8494,
+      "step": 6091
+    },
+    {
+      "epoch": 0.779776,
+      "grad_norm": 0.3840351700782776,
+      "learning_rate": 0.00012184765802663466,
+      "loss": 2.7177,
+      "step": 6092
+    },
+    {
+      "epoch": 0.779904,
+      "grad_norm": 0.44496962428092957,
+      "learning_rate": 0.00012171206354895081,
+      "loss": 2.8732,
+      "step": 6093
+    },
+    {
+      "epoch": 0.780032,
+      "grad_norm": 0.40121006965637207,
+      "learning_rate": 0.00012157653410334874,
+      "loss": 2.7388,
+      "step": 6094
+    },
+    {
+      "epoch": 0.78016,
+      "grad_norm": 0.417825847864151,
+      "learning_rate": 0.0001214410697131273,
+      "loss": 2.8195,
+      "step": 6095
+    },
+    {
+      "epoch": 0.780288,
+      "grad_norm": 0.41911542415618896,
+      "learning_rate": 0.00012130567040157465,
+      "loss": 2.7859,
+      "step": 6096
+    },
+    {
+      "epoch": 0.780416,
+      "grad_norm": 0.4534400701522827,
+      "learning_rate": 0.0001211703361919676,
+      "loss": 2.8468,
+      "step": 6097
+    },
+    {
+      "epoch": 0.780544,
+      "grad_norm": 0.42452070116996765,
+      "learning_rate": 0.00012103506710757156,
+      "loss": 2.6913,
+      "step": 6098
+    },
+    {
+      "epoch": 0.780672,
+      "grad_norm": 0.4037388861179352,
+      "learning_rate": 0.00012089986317164075,
+      "loss": 2.7828,
+      "step": 6099
+    },
+    {
+      "epoch": 0.7808,
+      "grad_norm": 0.4455913007259369,
+      "learning_rate": 0.00012076472440741842,
+      "loss": 2.8923,
+      "step": 6100
+    },
+    {
+      "epoch": 0.780928,
+      "grad_norm": 0.38897261023521423,
+      "learning_rate": 0.00012062965083813642,
+      "loss": 2.9451,
+      "step": 6101
+    },
+    {
+      "epoch": 0.781056,
+      "grad_norm": 0.38778096437454224,
+      "learning_rate": 0.00012049464248701536,
+      "loss": 2.9444,
+      "step": 6102
+    },
+    {
+      "epoch": 0.781184,
+      "grad_norm": 0.41741812229156494,
+      "learning_rate": 0.00012035969937726498,
+      "loss": 2.836,
+      "step": 6103
+    },
+    {
+      "epoch": 0.781312,
+      "grad_norm": 0.3800613284111023,
+      "learning_rate": 0.00012022482153208347,
+      "loss": 2.7303,
+      "step": 6104
+    },
+    {
+      "epoch": 0.78144,
+      "grad_norm": 0.4118509292602539,
+      "learning_rate": 0.00012009000897465783,
+      "loss": 2.7615,
+      "step": 6105
+    },
+    {
+      "epoch": 0.781568,
+      "grad_norm": 0.40536588430404663,
+      "learning_rate": 0.00011995526172816384,
+      "loss": 2.7406,
+      "step": 6106
+    },
+    {
+      "epoch": 0.781696,
+      "grad_norm": 0.37603074312210083,
+      "learning_rate": 0.00011982057981576632,
+      "loss": 2.8064,
+      "step": 6107
+    },
+    {
+      "epoch": 0.781824,
+      "grad_norm": 0.38122668862342834,
+      "learning_rate": 0.00011968596326061858,
+      "loss": 2.7362,
+      "step": 6108
+    },
+    {
+      "epoch": 0.781952,
+      "grad_norm": 0.3819859027862549,
+      "learning_rate": 0.00011955141208586257,
+      "loss": 2.8437,
+      "step": 6109
+    },
+    {
+      "epoch": 0.78208,
+      "grad_norm": 0.3772777020931244,
+      "learning_rate": 0.00011941692631462953,
+      "loss": 2.8402,
+      "step": 6110
+    },
+    {
+      "epoch": 0.782208,
+      "grad_norm": 0.36615708470344543,
+      "learning_rate": 0.00011928250597003892,
+      "loss": 2.7481,
+      "step": 6111
+    },
+    {
+      "epoch": 0.782336,
+      "grad_norm": 0.42943763732910156,
+      "learning_rate": 0.00011914815107519922,
+      "loss": 2.8346,
+      "step": 6112
+    },
+    {
+      "epoch": 0.782464,
+      "grad_norm": 0.41139236092567444,
+      "learning_rate": 0.00011901386165320755,
+      "loss": 2.8441,
+      "step": 6113
+    },
+    {
+      "epoch": 0.782592,
+      "grad_norm": 0.40139028429985046,
+      "learning_rate": 0.00011887963772714983,
+      "loss": 2.7225,
+      "step": 6114
+    },
+    {
+      "epoch": 0.78272,
+      "grad_norm": 0.39530378580093384,
+      "learning_rate": 0.00011874547932010055,
+      "loss": 2.8137,
+      "step": 6115
+    },
+    {
+      "epoch": 0.782848,
+      "grad_norm": 0.39510995149612427,
+      "learning_rate": 0.00011861138645512343,
+      "loss": 2.7124,
+      "step": 6116
+    },
+    {
+      "epoch": 0.782976,
+      "grad_norm": 0.4093014895915985,
+      "learning_rate": 0.00011847735915527025,
+      "loss": 2.821,
+      "step": 6117
+    },
+    {
+      "epoch": 0.783104,
+      "grad_norm": 0.391609787940979,
+      "learning_rate": 0.00011834339744358218,
+      "loss": 2.8655,
+      "step": 6118
+    },
+    {
+      "epoch": 0.783232,
+      "grad_norm": 0.3918015956878662,
+      "learning_rate": 0.0001182095013430886,
+      "loss": 2.7853,
+      "step": 6119
+    },
+    {
+      "epoch": 0.78336,
+      "grad_norm": 0.39427870512008667,
+      "learning_rate": 0.00011807567087680787,
+      "loss": 2.7295,
+      "step": 6120
+    },
+    {
+      "epoch": 0.783488,
+      "grad_norm": 0.39006420969963074,
+      "learning_rate": 0.00011794190606774697,
+      "loss": 2.6798,
+      "step": 6121
+    },
+    {
+      "epoch": 0.783616,
+      "grad_norm": 0.41265252232551575,
+      "learning_rate": 0.0001178082069389016,
+      "loss": 2.8502,
+      "step": 6122
+    },
+    {
+      "epoch": 0.783744,
+      "grad_norm": 0.37792232632637024,
+      "learning_rate": 0.00011767457351325605,
+      "loss": 2.7144,
+      "step": 6123
+    },
+    {
+      "epoch": 0.783872,
+      "grad_norm": 0.3761725127696991,
+      "learning_rate": 0.00011754100581378364,
+      "loss": 2.6613,
+      "step": 6124
+    },
+    {
+      "epoch": 0.784,
+      "grad_norm": 0.40166664123535156,
+      "learning_rate": 0.00011740750386344628,
+      "loss": 2.7601,
+      "step": 6125
+    },
+    {
+      "epoch": 0.784128,
+      "grad_norm": 0.4468143582344055,
+      "learning_rate": 0.00011727406768519445,
+      "loss": 2.8002,
+      "step": 6126
+    },
+    {
+      "epoch": 0.784256,
+      "grad_norm": 0.3967343866825104,
+      "learning_rate": 0.0001171406973019673,
+      "loss": 2.7028,
+      "step": 6127
+    },
+    {
+      "epoch": 0.784384,
+      "grad_norm": 0.3901252746582031,
+      "learning_rate": 0.00011700739273669275,
+      "loss": 2.6682,
+      "step": 6128
+    },
+    {
+      "epoch": 0.784512,
+      "grad_norm": 0.3900466561317444,
+      "learning_rate": 0.0001168741540122874,
+      "loss": 2.8066,
+      "step": 6129
+    },
+    {
+      "epoch": 0.78464,
+      "grad_norm": 0.4182237684726715,
+      "learning_rate": 0.00011674098115165643,
+      "loss": 2.8018,
+      "step": 6130
+    },
+    {
+      "epoch": 0.784768,
+      "grad_norm": 0.424528568983078,
+      "learning_rate": 0.00011660787417769409,
+      "loss": 2.7954,
+      "step": 6131
+    },
+    {
+      "epoch": 0.784896,
+      "grad_norm": 0.44193777441978455,
+      "learning_rate": 0.00011647483311328283,
+      "loss": 2.9693,
+      "step": 6132
+    },
+    {
+      "epoch": 0.785024,
+      "grad_norm": 0.40792712569236755,
+      "learning_rate": 0.00011634185798129382,
+      "loss": 2.8749,
+      "step": 6133
+    },
+    {
+      "epoch": 0.785152,
+      "grad_norm": 0.42117390036582947,
+      "learning_rate": 0.00011620894880458732,
+      "loss": 2.8323,
+      "step": 6134
+    },
+    {
+      "epoch": 0.78528,
+      "grad_norm": 0.412047803401947,
+      "learning_rate": 0.00011607610560601179,
+      "loss": 2.7701,
+      "step": 6135
+    },
+    {
+      "epoch": 0.785408,
+      "grad_norm": 0.41590726375579834,
+      "learning_rate": 0.00011594332840840455,
+      "loss": 2.7991,
+      "step": 6136
+    },
+    {
+      "epoch": 0.785536,
+      "grad_norm": 0.44487810134887695,
+      "learning_rate": 0.00011581061723459153,
+      "loss": 2.7009,
+      "step": 6137
+    },
+    {
+      "epoch": 0.785664,
+      "grad_norm": 0.38747504353523254,
+      "learning_rate": 0.00011567797210738712,
+      "loss": 2.7196,
+      "step": 6138
+    },
+    {
+      "epoch": 0.785792,
+      "grad_norm": 0.4208053946495056,
+      "learning_rate": 0.00011554539304959494,
+      "loss": 2.7709,
+      "step": 6139
+    },
+    {
+      "epoch": 0.78592,
+      "grad_norm": 0.39744043350219727,
+      "learning_rate": 0.00011541288008400663,
+      "loss": 2.7678,
+      "step": 6140
+    },
+    {
+      "epoch": 0.786048,
+      "grad_norm": 0.41169145703315735,
+      "learning_rate": 0.00011528043323340281,
+      "loss": 2.7851,
+      "step": 6141
+    },
+    {
+      "epoch": 0.786176,
+      "grad_norm": 0.38643744587898254,
+      "learning_rate": 0.0001151480525205525,
+      "loss": 2.7985,
+      "step": 6142
+    },
+    {
+      "epoch": 0.786304,
+      "grad_norm": 0.41244807839393616,
+      "learning_rate": 0.00011501573796821346,
+      "loss": 2.7896,
+      "step": 6143
+    },
+    {
+      "epoch": 0.786432,
+      "grad_norm": 0.40853631496429443,
+      "learning_rate": 0.0001148834895991323,
+      "loss": 2.7435,
+      "step": 6144
+    },
+    {
+      "epoch": 0.78656,
+      "grad_norm": 0.4178119897842407,
+      "learning_rate": 0.00011475130743604379,
+      "loss": 2.8107,
+      "step": 6145
+    },
+    {
+      "epoch": 0.786688,
+      "grad_norm": 0.3967326879501343,
+      "learning_rate": 0.00011461919150167188,
+      "loss": 2.7017,
+      "step": 6146
+    },
+    {
+      "epoch": 0.786816,
+      "grad_norm": 0.4049293100833893,
+      "learning_rate": 0.00011448714181872865,
+      "loss": 2.7421,
+      "step": 6147
+    },
+    {
+      "epoch": 0.786944,
+      "grad_norm": 0.4301431477069855,
+      "learning_rate": 0.000114355158409915,
+      "loss": 2.7241,
+      "step": 6148
+    },
+    {
+      "epoch": 0.787072,
+      "grad_norm": 0.39704757928848267,
+      "learning_rate": 0.00011422324129792039,
+      "loss": 2.8355,
+      "step": 6149
+    },
+    {
+      "epoch": 0.7872,
+      "grad_norm": 0.40377509593963623,
+      "learning_rate": 0.00011409139050542295,
+      "loss": 2.7554,
+      "step": 6150
+    },
+    {
+      "epoch": 0.787328,
+      "grad_norm": 0.40070194005966187,
+      "learning_rate": 0.00011395960605508915,
+      "loss": 2.8338,
+      "step": 6151
+    },
+    {
+      "epoch": 0.787456,
+      "grad_norm": 0.3852565586566925,
+      "learning_rate": 0.00011382788796957454,
+      "loss": 2.8142,
+      "step": 6152
+    },
+    {
+      "epoch": 0.787584,
+      "grad_norm": 0.40507224202156067,
+      "learning_rate": 0.00011369623627152276,
+      "loss": 2.8646,
+      "step": 6153
+    },
+    {
+      "epoch": 0.787712,
+      "grad_norm": 0.40590181946754456,
+      "learning_rate": 0.00011356465098356656,
+      "loss": 2.9535,
+      "step": 6154
+    },
+    {
+      "epoch": 0.78784,
+      "grad_norm": 0.41355717182159424,
+      "learning_rate": 0.00011343313212832674,
+      "loss": 2.7269,
+      "step": 6155
+    },
+    {
+      "epoch": 0.787968,
+      "grad_norm": 0.44469138979911804,
+      "learning_rate": 0.00011330167972841299,
+      "loss": 2.7507,
+      "step": 6156
+    },
+    {
+      "epoch": 0.788096,
+      "grad_norm": 0.4023699164390564,
+      "learning_rate": 0.00011317029380642352,
+      "loss": 2.8294,
+      "step": 6157
+    },
+    {
+      "epoch": 0.788224,
+      "grad_norm": 0.3976500630378723,
+      "learning_rate": 0.00011303897438494504,
+      "loss": 2.6333,
+      "step": 6158
+    },
+    {
+      "epoch": 0.788352,
+      "grad_norm": 0.4411526620388031,
+      "learning_rate": 0.00011290772148655276,
+      "loss": 2.7721,
+      "step": 6159
+    },
+    {
+      "epoch": 0.78848,
+      "grad_norm": 0.4419334828853607,
+      "learning_rate": 0.0001127765351338108,
+      "loss": 2.812,
+      "step": 6160
+    },
+    {
+      "epoch": 0.788608,
+      "grad_norm": 0.4114634692668915,
+      "learning_rate": 0.00011264541534927147,
+      "loss": 2.835,
+      "step": 6161
+    },
+    {
+      "epoch": 0.788736,
+      "grad_norm": 0.42156094312667847,
+      "learning_rate": 0.00011251436215547595,
+      "loss": 2.7908,
+      "step": 6162
+    },
+    {
+      "epoch": 0.788864,
+      "grad_norm": 0.39446473121643066,
+      "learning_rate": 0.0001123833755749537,
+      "loss": 2.8073,
+      "step": 6163
+    },
+    {
+      "epoch": 0.788992,
+      "grad_norm": 0.40755242109298706,
+      "learning_rate": 0.00011225245563022285,
+      "loss": 2.9206,
+      "step": 6164
+    },
+    {
+      "epoch": 0.78912,
+      "grad_norm": 0.46533462405204773,
+      "learning_rate": 0.00011212160234378999,
+      "loss": 2.8056,
+      "step": 6165
+    },
+    {
+      "epoch": 0.789248,
+      "grad_norm": 0.40233057737350464,
+      "learning_rate": 0.00011199081573815024,
+      "loss": 2.9496,
+      "step": 6166
+    },
+    {
+      "epoch": 0.789376,
+      "grad_norm": 0.400779664516449,
+      "learning_rate": 0.0001118600958357876,
+      "loss": 2.7748,
+      "step": 6167
+    },
+    {
+      "epoch": 0.789504,
+      "grad_norm": 0.3953477740287781,
+      "learning_rate": 0.00011172944265917417,
+      "loss": 2.7169,
+      "step": 6168
+    },
+    {
+      "epoch": 0.789632,
+      "grad_norm": 0.3545079231262207,
+      "learning_rate": 0.00011159885623077076,
+      "loss": 2.8313,
+      "step": 6169
+    },
+    {
+      "epoch": 0.78976,
+      "grad_norm": 0.43073102831840515,
+      "learning_rate": 0.00011146833657302658,
+      "loss": 2.9393,
+      "step": 6170
+    },
+    {
+      "epoch": 0.789888,
+      "grad_norm": 0.40910255908966064,
+      "learning_rate": 0.00011133788370837972,
+      "loss": 2.8773,
+      "step": 6171
+    },
+    {
+      "epoch": 0.790016,
+      "grad_norm": 0.431392103433609,
+      "learning_rate": 0.00011120749765925642,
+      "loss": 2.8361,
+      "step": 6172
+    },
+    {
+      "epoch": 0.790144,
+      "grad_norm": 0.43256497383117676,
+      "learning_rate": 0.00011107717844807153,
+      "loss": 2.7517,
+      "step": 6173
+    },
+    {
+      "epoch": 0.790272,
+      "grad_norm": 0.3892689347267151,
+      "learning_rate": 0.00011094692609722829,
+      "loss": 2.8561,
+      "step": 6174
+    },
+    {
+      "epoch": 0.7904,
+      "grad_norm": 0.4293498694896698,
+      "learning_rate": 0.0001108167406291189,
+      "loss": 2.8515,
+      "step": 6175
+    },
+    {
+      "epoch": 0.790528,
+      "grad_norm": 0.4012289047241211,
+      "learning_rate": 0.00011068662206612362,
+      "loss": 2.8088,
+      "step": 6176
+    },
+    {
+      "epoch": 0.790656,
+      "grad_norm": 0.4073731303215027,
+      "learning_rate": 0.00011055657043061124,
+      "loss": 2.8293,
+      "step": 6177
+    },
+    {
+      "epoch": 0.790784,
+      "grad_norm": 0.43461066484451294,
+      "learning_rate": 0.0001104265857449393,
+      "loss": 2.6929,
+      "step": 6178
+    },
+    {
+      "epoch": 0.790912,
+      "grad_norm": 0.3994751274585724,
+      "learning_rate": 0.00011029666803145355,
+      "loss": 2.7075,
+      "step": 6179
+    },
+    {
+      "epoch": 0.79104,
+      "grad_norm": 0.4171585440635681,
+      "learning_rate": 0.00011016681731248829,
+      "loss": 2.7312,
+      "step": 6180
+    },
+    {
+      "epoch": 0.791168,
+      "grad_norm": 0.41450217366218567,
+      "learning_rate": 0.00011003703361036643,
+      "loss": 2.86,
+      "step": 6181
+    },
+    {
+      "epoch": 0.791296,
+      "grad_norm": 0.424073189496994,
+      "learning_rate": 0.00010990731694739947,
+      "loss": 2.9106,
+      "step": 6182
+    },
+    {
+      "epoch": 0.791424,
+      "grad_norm": 0.4247032403945923,
+      "learning_rate": 0.00010977766734588706,
+      "loss": 2.8607,
+      "step": 6183
+    },
+    {
+      "epoch": 0.791552,
+      "grad_norm": 0.43495041131973267,
+      "learning_rate": 0.0001096480848281175,
+      "loss": 2.7134,
+      "step": 6184
+    },
+    {
+      "epoch": 0.79168,
+      "grad_norm": 0.39344313740730286,
+      "learning_rate": 0.00010951856941636751,
+      "loss": 2.7383,
+      "step": 6185
+    },
+    {
+      "epoch": 0.791808,
+      "grad_norm": 0.40216585993766785,
+      "learning_rate": 0.00010938912113290229,
+      "loss": 2.7885,
+      "step": 6186
+    },
+    {
+      "epoch": 0.791936,
+      "grad_norm": 0.4002363085746765,
+      "learning_rate": 0.00010925973999997535,
+      "loss": 2.7927,
+      "step": 6187
+    },
+    {
+      "epoch": 0.792064,
+      "grad_norm": 0.403379887342453,
+      "learning_rate": 0.00010913042603982909,
+      "loss": 2.8718,
+      "step": 6188
+    },
+    {
+      "epoch": 0.792192,
+      "grad_norm": 0.409394770860672,
+      "learning_rate": 0.00010900117927469383,
+      "loss": 2.8772,
+      "step": 6189
+    },
+    {
+      "epoch": 0.79232,
+      "grad_norm": 0.44204679131507874,
+      "learning_rate": 0.0001088719997267888,
+      "loss": 2.8745,
+      "step": 6190
+    },
+    {
+      "epoch": 0.792448,
+      "grad_norm": 0.4089730978012085,
+      "learning_rate": 0.00010874288741832139,
+      "loss": 2.8552,
+      "step": 6191
+    },
+    {
+      "epoch": 0.792576,
+      "grad_norm": 0.4090852737426758,
+      "learning_rate": 0.00010861384237148748,
+      "loss": 2.8743,
+      "step": 6192
+    },
+    {
+      "epoch": 0.792704,
+      "grad_norm": 0.39400237798690796,
+      "learning_rate": 0.00010848486460847134,
+      "loss": 2.7907,
+      "step": 6193
+    },
+    {
+      "epoch": 0.792832,
+      "grad_norm": 0.4023675322532654,
+      "learning_rate": 0.00010835595415144588,
+      "loss": 2.6943,
+      "step": 6194
+    },
+    {
+      "epoch": 0.79296,
+      "grad_norm": 0.43342387676239014,
+      "learning_rate": 0.00010822711102257204,
+      "loss": 2.7682,
+      "step": 6195
+    },
+    {
+      "epoch": 0.793088,
+      "grad_norm": 0.40816089510917664,
+      "learning_rate": 0.00010809833524399982,
+      "loss": 2.8751,
+      "step": 6196
+    },
+    {
+      "epoch": 0.793216,
+      "grad_norm": 0.40677696466445923,
+      "learning_rate": 0.00010796962683786709,
+      "loss": 2.8955,
+      "step": 6197
+    },
+    {
+      "epoch": 0.793344,
+      "grad_norm": 0.422405481338501,
+      "learning_rate": 0.00010784098582630015,
+      "loss": 2.7331,
+      "step": 6198
+    },
+    {
+      "epoch": 0.793472,
+      "grad_norm": 0.44124487042427063,
+      "learning_rate": 0.00010771241223141415,
+      "loss": 2.7427,
+      "step": 6199
+    },
+    {
+      "epoch": 0.7936,
+      "grad_norm": 0.40035033226013184,
+      "learning_rate": 0.00010758390607531232,
+      "loss": 2.7236,
+      "step": 6200
+    },
+    {
+      "epoch": 0.793728,
+      "grad_norm": 0.3903369605541229,
+      "learning_rate": 0.00010745546738008627,
+      "loss": 2.6236,
+      "step": 6201
+    },
+    {
+      "epoch": 0.793856,
+      "grad_norm": 0.4269874095916748,
+      "learning_rate": 0.00010732709616781606,
+      "loss": 2.7682,
+      "step": 6202
+    },
+    {
+      "epoch": 0.793984,
+      "grad_norm": 0.40401288866996765,
+      "learning_rate": 0.0001071987924605704,
+      "loss": 2.7663,
+      "step": 6203
+    },
+    {
+      "epoch": 0.794112,
+      "grad_norm": 0.3863646984100342,
+      "learning_rate": 0.00010707055628040602,
+      "loss": 2.7794,
+      "step": 6204
+    },
+    {
+      "epoch": 0.79424,
+      "grad_norm": 0.38241147994995117,
+      "learning_rate": 0.00010694238764936826,
+      "loss": 2.7144,
+      "step": 6205
+    },
+    {
+      "epoch": 0.794368,
+      "grad_norm": 0.40623533725738525,
+      "learning_rate": 0.00010681428658949082,
+      "loss": 2.7393,
+      "step": 6206
+    },
+    {
+      "epoch": 0.794496,
+      "grad_norm": 0.42680564522743225,
+      "learning_rate": 0.00010668625312279551,
+      "loss": 2.7837,
+      "step": 6207
+    },
+    {
+      "epoch": 0.794624,
+      "grad_norm": 0.3869225084781647,
+      "learning_rate": 0.00010655828727129318,
+      "loss": 2.9272,
+      "step": 6208
+    },
+    {
+      "epoch": 0.794752,
+      "grad_norm": 0.37734878063201904,
+      "learning_rate": 0.00010643038905698237,
+      "loss": 2.7159,
+      "step": 6209
+    },
+    {
+      "epoch": 0.79488,
+      "grad_norm": 0.4220265746116638,
+      "learning_rate": 0.00010630255850185022,
+      "loss": 2.8836,
+      "step": 6210
+    },
+    {
+      "epoch": 0.795008,
+      "grad_norm": 0.3850766718387604,
+      "learning_rate": 0.00010617479562787258,
+      "loss": 2.664,
+      "step": 6211
+    },
+    {
+      "epoch": 0.795136,
+      "grad_norm": 0.4060438275337219,
+      "learning_rate": 0.0001060471004570131,
+      "loss": 2.8146,
+      "step": 6212
+    },
+    {
+      "epoch": 0.795264,
+      "grad_norm": 0.4156434237957001,
+      "learning_rate": 0.00010591947301122424,
+      "loss": 2.8626,
+      "step": 6213
+    },
+    {
+      "epoch": 0.795392,
+      "grad_norm": 0.4240972101688385,
+      "learning_rate": 0.00010579191331244648,
+      "loss": 2.8374,
+      "step": 6214
+    },
+    {
+      "epoch": 0.79552,
+      "grad_norm": 0.6549705862998962,
+      "learning_rate": 0.00010566442138260884,
+      "loss": 2.9025,
+      "step": 6215
+    },
+    {
+      "epoch": 0.795648,
+      "grad_norm": 0.41877511143684387,
+      "learning_rate": 0.00010553699724362865,
+      "loss": 2.9188,
+      "step": 6216
+    },
+    {
+      "epoch": 0.795776,
+      "grad_norm": 0.38468965888023376,
+      "learning_rate": 0.00010540964091741157,
+      "loss": 2.7385,
+      "step": 6217
+    },
+    {
+      "epoch": 0.795904,
+      "grad_norm": 0.40132033824920654,
+      "learning_rate": 0.00010528235242585187,
+      "loss": 2.6227,
+      "step": 6218
+    },
+    {
+      "epoch": 0.796032,
+      "grad_norm": 0.378309965133667,
+      "learning_rate": 0.00010515513179083175,
+      "loss": 2.7893,
+      "step": 6219
+    },
+    {
+      "epoch": 0.79616,
+      "grad_norm": 0.42923951148986816,
+      "learning_rate": 0.0001050279790342219,
+      "loss": 2.9117,
+      "step": 6220
+    },
+    {
+      "epoch": 0.796288,
+      "grad_norm": 0.4053778052330017,
+      "learning_rate": 0.00010490089417788134,
+      "loss": 2.7977,
+      "step": 6221
+    },
+    {
+      "epoch": 0.796416,
+      "grad_norm": 0.398816853761673,
+      "learning_rate": 0.0001047738772436575,
+      "loss": 2.7296,
+      "step": 6222
+    },
+    {
+      "epoch": 0.796544,
+      "grad_norm": 0.398111492395401,
+      "learning_rate": 0.00010464692825338585,
+      "loss": 2.9482,
+      "step": 6223
+    },
+    {
+      "epoch": 0.796672,
+      "grad_norm": 0.41548076272010803,
+      "learning_rate": 0.00010452004722889069,
+      "loss": 2.962,
+      "step": 6224
+    },
+    {
+      "epoch": 0.7968,
+      "grad_norm": 0.41420844197273254,
+      "learning_rate": 0.00010439323419198421,
+      "loss": 2.7755,
+      "step": 6225
+    },
+    {
+      "epoch": 0.796928,
+      "grad_norm": 0.3828265964984894,
+      "learning_rate": 0.0001042664891644669,
+      "loss": 2.7829,
+      "step": 6226
+    },
+    {
+      "epoch": 0.797056,
+      "grad_norm": 0.4225963354110718,
+      "learning_rate": 0.00010413981216812801,
+      "loss": 2.8804,
+      "step": 6227
+    },
+    {
+      "epoch": 0.797184,
+      "grad_norm": 0.41914647817611694,
+      "learning_rate": 0.00010401320322474456,
+      "loss": 2.741,
+      "step": 6228
+    },
+    {
+      "epoch": 0.797312,
+      "grad_norm": 0.41094592213630676,
+      "learning_rate": 0.00010388666235608219,
+      "loss": 2.7747,
+      "step": 6229
+    },
+    {
+      "epoch": 0.79744,
+      "grad_norm": 0.39802664518356323,
+      "learning_rate": 0.00010376018958389472,
+      "loss": 2.8424,
+      "step": 6230
+    },
+    {
+      "epoch": 0.797568,
+      "grad_norm": 0.41074100136756897,
+      "learning_rate": 0.00010363378492992409,
+      "loss": 2.8328,
+      "step": 6231
+    },
+    {
+      "epoch": 0.797696,
+      "grad_norm": 0.43117064237594604,
+      "learning_rate": 0.00010350744841590103,
+      "loss": 2.7553,
+      "step": 6232
+    },
+    {
+      "epoch": 0.797824,
+      "grad_norm": 0.4362488389015198,
+      "learning_rate": 0.00010338118006354413,
+      "loss": 2.8492,
+      "step": 6233
+    },
+    {
+      "epoch": 0.797952,
+      "grad_norm": 0.41972899436950684,
+      "learning_rate": 0.0001032549798945604,
+      "loss": 2.8461,
+      "step": 6234
+    },
+    {
+      "epoch": 0.79808,
+      "grad_norm": 0.42310991883277893,
+      "learning_rate": 0.00010312884793064492,
+      "loss": 2.7977,
+      "step": 6235
+    },
+    {
+      "epoch": 0.798208,
+      "grad_norm": 0.3960663676261902,
+      "learning_rate": 0.00010300278419348158,
+      "loss": 2.8405,
+      "step": 6236
+    },
+    {
+      "epoch": 0.798336,
+      "grad_norm": 0.39331015944480896,
+      "learning_rate": 0.00010287678870474199,
+      "loss": 2.9171,
+      "step": 6237
+    },
+    {
+      "epoch": 0.798464,
+      "grad_norm": 0.3978608250617981,
+      "learning_rate": 0.00010275086148608614,
+      "loss": 2.8884,
+      "step": 6238
+    },
+    {
+      "epoch": 0.798592,
+      "grad_norm": 0.39447903633117676,
+      "learning_rate": 0.00010262500255916268,
+      "loss": 2.7681,
+      "step": 6239
+    },
+    {
+      "epoch": 0.79872,
+      "grad_norm": 0.41261228919029236,
+      "learning_rate": 0.00010249921194560802,
+      "loss": 2.834,
+      "step": 6240
+    },
+    {
+      "epoch": 0.798848,
+      "grad_norm": 0.39842668175697327,
+      "learning_rate": 0.00010237348966704707,
+      "loss": 2.9006,
+      "step": 6241
+    },
+    {
+      "epoch": 0.798976,
+      "grad_norm": 0.40993979573249817,
+      "learning_rate": 0.0001022478357450929,
+      "loss": 2.8452,
+      "step": 6242
+    },
+    {
+      "epoch": 0.799104,
+      "grad_norm": 0.41913068294525146,
+      "learning_rate": 0.00010212225020134691,
+      "loss": 2.7827,
+      "step": 6243
+    },
+    {
+      "epoch": 0.799232,
+      "grad_norm": 0.4090127646923065,
+      "learning_rate": 0.00010199673305739853,
+      "loss": 2.7503,
+      "step": 6244
+    },
+    {
+      "epoch": 0.79936,
+      "grad_norm": 0.42548665404319763,
+      "learning_rate": 0.000101871284334826,
+      "loss": 2.9402,
+      "step": 6245
+    },
+    {
+      "epoch": 0.799488,
+      "grad_norm": 0.4064464867115021,
+      "learning_rate": 0.00010174590405519501,
+      "loss": 2.8726,
+      "step": 6246
+    },
+    {
+      "epoch": 0.799616,
+      "grad_norm": 0.44287556409835815,
+      "learning_rate": 0.00010162059224006026,
+      "loss": 2.8059,
+      "step": 6247
+    },
+    {
+      "epoch": 0.799744,
+      "grad_norm": 0.4157921075820923,
+      "learning_rate": 0.00010149534891096408,
+      "loss": 2.8391,
+      "step": 6248
+    },
+    {
+      "epoch": 0.799872,
+      "grad_norm": 0.40464431047439575,
+      "learning_rate": 0.00010137017408943727,
+      "loss": 2.86,
+      "step": 6249
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.38962605595588684,
+      "learning_rate": 0.00010124506779699882,
+      "loss": 2.702,
+      "step": 6250
+    },
+    {
+      "epoch": 0.800128,
+      "grad_norm": 0.40357285737991333,
+      "learning_rate": 0.00010112003005515602,
+      "loss": 2.8592,
+      "step": 6251
+    },
+    {
+      "epoch": 0.800256,
+      "grad_norm": 0.3859737515449524,
+      "learning_rate": 0.00010099506088540417,
+      "loss": 2.7917,
+      "step": 6252
+    },
+    {
+      "epoch": 0.800384,
+      "grad_norm": 0.4180377423763275,
+      "learning_rate": 0.00010087016030922708,
+      "loss": 2.7968,
+      "step": 6253
+    },
+    {
+      "epoch": 0.800512,
+      "grad_norm": 0.44509023427963257,
+      "learning_rate": 0.00010074532834809646,
+      "loss": 2.718,
+      "step": 6254
+    },
+    {
+      "epoch": 0.80064,
+      "grad_norm": 0.4028811454772949,
+      "learning_rate": 0.00010062056502347256,
+      "loss": 2.6858,
+      "step": 6255
+    },
+    {
+      "epoch": 0.800768,
+      "grad_norm": 0.4147052466869354,
+      "learning_rate": 0.00010049587035680352,
+      "loss": 2.7598,
+      "step": 6256
+    },
+    {
+      "epoch": 0.800896,
+      "grad_norm": 0.42961445450782776,
+      "learning_rate": 0.00010037124436952588,
+      "loss": 2.8384,
+      "step": 6257
+    },
+    {
+      "epoch": 0.801024,
+      "grad_norm": 0.42213600873947144,
+      "learning_rate": 0.00010024668708306417,
+      "loss": 2.7082,
+      "step": 6258
+    },
+    {
+      "epoch": 0.801152,
+      "grad_norm": 0.3895009458065033,
+      "learning_rate": 0.0001001221985188312,
+      "loss": 2.9114,
+      "step": 6259
+    },
+    {
+      "epoch": 0.80128,
+      "grad_norm": 0.4187155067920685,
+      "learning_rate": 9.999777869822818e-05,
+      "loss": 2.7717,
+      "step": 6260
+    },
+    {
+      "epoch": 0.801408,
+      "grad_norm": 0.39875689148902893,
+      "learning_rate": 9.987342764264423e-05,
+      "loss": 2.7839,
+      "step": 6261
+    },
+    {
+      "epoch": 0.801536,
+      "grad_norm": 0.4285952150821686,
+      "learning_rate": 9.974914537345675e-05,
+      "loss": 2.8764,
+      "step": 6262
+    },
+    {
+      "epoch": 0.801664,
+      "grad_norm": 0.38977617025375366,
+      "learning_rate": 9.96249319120311e-05,
+      "loss": 2.8138,
+      "step": 6263
+    },
+    {
+      "epoch": 0.801792,
+      "grad_norm": 0.4116419553756714,
+      "learning_rate": 9.950078727972139e-05,
+      "loss": 2.7524,
+      "step": 6264
+    },
+    {
+      "epoch": 0.80192,
+      "grad_norm": 0.38551056385040283,
+      "learning_rate": 9.937671149786931e-05,
+      "loss": 2.7621,
+      "step": 6265
+    },
+    {
+      "epoch": 0.802048,
+      "grad_norm": 0.40352585911750793,
+      "learning_rate": 9.925270458780495e-05,
+      "loss": 2.9386,
+      "step": 6266
+    },
+    {
+      "epoch": 0.802176,
+      "grad_norm": 0.3983781635761261,
+      "learning_rate": 9.912876657084641e-05,
+      "loss": 2.8641,
+      "step": 6267
+    },
+    {
+      "epoch": 0.802304,
+      "grad_norm": 0.42702069878578186,
+      "learning_rate": 9.900489746830032e-05,
+      "loss": 2.7974,
+      "step": 6268
+    },
+    {
+      "epoch": 0.802432,
+      "grad_norm": 0.4021511673927307,
+      "learning_rate": 9.888109730146111e-05,
+      "loss": 2.7816,
+      "step": 6269
+    },
+    {
+      "epoch": 0.80256,
+      "grad_norm": 0.41109320521354675,
+      "learning_rate": 9.87573660916114e-05,
+      "loss": 2.8093,
+      "step": 6270
+    },
+    {
+      "epoch": 0.802688,
+      "grad_norm": 0.41892290115356445,
+      "learning_rate": 9.863370386002212e-05,
+      "loss": 2.8159,
+      "step": 6271
+    },
+    {
+      "epoch": 0.802816,
+      "grad_norm": 0.4278797507286072,
+      "learning_rate": 9.851011062795201e-05,
+      "loss": 2.9534,
+      "step": 6272
+    },
+    {
+      "epoch": 0.802944,
+      "grad_norm": 0.4007171094417572,
+      "learning_rate": 9.838658641664849e-05,
+      "loss": 2.9157,
+      "step": 6273
+    },
+    {
+      "epoch": 0.803072,
+      "grad_norm": 0.38895830512046814,
+      "learning_rate": 9.826313124734653e-05,
+      "loss": 2.8662,
+      "step": 6274
+    },
+    {
+      "epoch": 0.8032,
+      "grad_norm": 0.4464437961578369,
+      "learning_rate": 9.813974514126978e-05,
+      "loss": 2.8567,
+      "step": 6275
+    },
+    {
+      "epoch": 0.803328,
+      "grad_norm": 0.4071974754333496,
+      "learning_rate": 9.801642811962964e-05,
+      "loss": 2.8728,
+      "step": 6276
+    },
+    {
+      "epoch": 0.803456,
+      "grad_norm": 0.3949121832847595,
+      "learning_rate": 9.789318020362564e-05,
+      "loss": 2.7566,
+      "step": 6277
+    },
+    {
+      "epoch": 0.803584,
+      "grad_norm": 0.43998900055885315,
+      "learning_rate": 9.777000141444564e-05,
+      "loss": 2.852,
+      "step": 6278
+    },
+    {
+      "epoch": 0.803712,
+      "grad_norm": 0.39501601457595825,
+      "learning_rate": 9.764689177326541e-05,
+      "loss": 2.7217,
+      "step": 6279
+    },
+    {
+      "epoch": 0.80384,
+      "grad_norm": 0.4470687508583069,
+      "learning_rate": 9.75238513012488e-05,
+      "loss": 2.7293,
+      "step": 6280
+    },
+    {
+      "epoch": 0.803968,
+      "grad_norm": 0.37661466002464294,
+      "learning_rate": 9.740088001954811e-05,
+      "loss": 2.7238,
+      "step": 6281
+    },
+    {
+      "epoch": 0.804096,
+      "grad_norm": 0.39654895663261414,
+      "learning_rate": 9.72779779493036e-05,
+      "loss": 2.6475,
+      "step": 6282
+    },
+    {
+      "epoch": 0.804224,
+      "grad_norm": 0.4229336380958557,
+      "learning_rate": 9.715514511164342e-05,
+      "loss": 2.7026,
+      "step": 6283
+    },
+    {
+      "epoch": 0.804352,
+      "grad_norm": 0.40150025486946106,
+      "learning_rate": 9.703238152768401e-05,
+      "loss": 2.8594,
+      "step": 6284
+    },
+    {
+      "epoch": 0.80448,
+      "grad_norm": 0.38703954219818115,
+      "learning_rate": 9.690968721852978e-05,
+      "loss": 2.8291,
+      "step": 6285
+    },
+    {
+      "epoch": 0.804608,
+      "grad_norm": 0.4264048933982849,
+      "learning_rate": 9.678706220527333e-05,
+      "loss": 2.8574,
+      "step": 6286
+    },
+    {
+      "epoch": 0.804736,
+      "grad_norm": 0.39785218238830566,
+      "learning_rate": 9.666450650899533e-05,
+      "loss": 2.8134,
+      "step": 6287
+    },
+    {
+      "epoch": 0.804864,
+      "grad_norm": 0.41180238127708435,
+      "learning_rate": 9.654202015076441e-05,
+      "loss": 2.6971,
+      "step": 6288
+    },
+    {
+      "epoch": 0.804992,
+      "grad_norm": 0.39508870244026184,
+      "learning_rate": 9.641960315163767e-05,
+      "loss": 2.6879,
+      "step": 6289
+    },
+    {
+      "epoch": 0.80512,
+      "grad_norm": 0.43867212533950806,
+      "learning_rate": 9.629725553265983e-05,
+      "loss": 2.9161,
+      "step": 6290
+    },
+    {
+      "epoch": 0.805248,
+      "grad_norm": 0.4307636618614197,
+      "learning_rate": 9.617497731486374e-05,
+      "loss": 2.8518,
+      "step": 6291
+    },
+    {
+      "epoch": 0.805376,
+      "grad_norm": 0.42175477743148804,
+      "learning_rate": 9.605276851927075e-05,
+      "loss": 2.7691,
+      "step": 6292
+    },
+    {
+      "epoch": 0.805504,
+      "grad_norm": 0.3985667824745178,
+      "learning_rate": 9.593062916688982e-05,
+      "loss": 2.8163,
+      "step": 6293
+    },
+    {
+      "epoch": 0.805632,
+      "grad_norm": 0.41839835047721863,
+      "learning_rate": 9.580855927871807e-05,
+      "loss": 2.7579,
+      "step": 6294
+    },
+    {
+      "epoch": 0.80576,
+      "grad_norm": 0.4210363030433655,
+      "learning_rate": 9.568655887574068e-05,
+      "loss": 2.8129,
+      "step": 6295
+    },
+    {
+      "epoch": 0.805888,
+      "grad_norm": 0.4242975115776062,
+      "learning_rate": 9.556462797893111e-05,
+      "loss": 2.7963,
+      "step": 6296
+    },
+    {
+      "epoch": 0.806016,
+      "grad_norm": 0.4105023443698883,
+      "learning_rate": 9.544276660925066e-05,
+      "loss": 2.666,
+      "step": 6297
+    },
+    {
+      "epoch": 0.806144,
+      "grad_norm": 0.41321077942848206,
+      "learning_rate": 9.532097478764862e-05,
+      "loss": 2.8413,
+      "step": 6298
+    },
+    {
+      "epoch": 0.806272,
+      "grad_norm": 0.41338911652565,
+      "learning_rate": 9.519925253506245e-05,
+      "loss": 2.8363,
+      "step": 6299
+    },
+    {
+      "epoch": 0.8064,
+      "grad_norm": 0.4082629084587097,
+      "learning_rate": 9.507759987241754e-05,
+      "loss": 2.7056,
+      "step": 6300
+    },
+    {
+      "epoch": 0.806528,
+      "grad_norm": 0.4394904375076294,
+      "learning_rate": 9.495601682062755e-05,
+      "loss": 2.6813,
+      "step": 6301
+    },
+    {
+      "epoch": 0.806656,
+      "grad_norm": 0.412337064743042,
+      "learning_rate": 9.483450340059385e-05,
+      "loss": 2.8371,
+      "step": 6302
+    },
+    {
+      "epoch": 0.806784,
+      "grad_norm": 0.4108998477458954,
+      "learning_rate": 9.47130596332062e-05,
+      "loss": 2.8636,
+      "step": 6303
+    },
+    {
+      "epoch": 0.806912,
+      "grad_norm": 0.41016772389411926,
+      "learning_rate": 9.459168553934211e-05,
+      "loss": 2.7894,
+      "step": 6304
+    },
+    {
+      "epoch": 0.80704,
+      "grad_norm": 0.39784160256385803,
+      "learning_rate": 9.447038113986717e-05,
+      "loss": 2.79,
+      "step": 6305
+    },
+    {
+      "epoch": 0.807168,
+      "grad_norm": 0.3772314190864563,
+      "learning_rate": 9.434914645563497e-05,
+      "loss": 2.7958,
+      "step": 6306
+    },
+    {
+      "epoch": 0.807296,
+      "grad_norm": 0.435159295797348,
+      "learning_rate": 9.422798150748724e-05,
+      "loss": 2.7204,
+      "step": 6307
+    },
+    {
+      "epoch": 0.807424,
+      "grad_norm": 0.40642887353897095,
+      "learning_rate": 9.410688631625364e-05,
+      "loss": 2.8382,
+      "step": 6308
+    },
+    {
+      "epoch": 0.807552,
+      "grad_norm": 0.4203818142414093,
+      "learning_rate": 9.398586090275164e-05,
+      "loss": 2.9079,
+      "step": 6309
+    },
+    {
+      "epoch": 0.80768,
+      "grad_norm": 0.39524245262145996,
+      "learning_rate": 9.386490528778702e-05,
+      "loss": 2.8321,
+      "step": 6310
+    },
+    {
+      "epoch": 0.807808,
+      "grad_norm": 0.42568308115005493,
+      "learning_rate": 9.374401949215366e-05,
+      "loss": 2.8717,
+      "step": 6311
+    },
+    {
+      "epoch": 0.807936,
+      "grad_norm": 0.41659215092658997,
+      "learning_rate": 9.362320353663311e-05,
+      "loss": 2.9453,
+      "step": 6312
+    },
+    {
+      "epoch": 0.808064,
+      "grad_norm": 0.4123534560203552,
+      "learning_rate": 9.350245744199498e-05,
+      "loss": 2.7397,
+      "step": 6313
+    },
+    {
+      "epoch": 0.808192,
+      "grad_norm": 0.3820914924144745,
+      "learning_rate": 9.338178122899693e-05,
+      "loss": 2.7338,
+      "step": 6314
+    },
+    {
+      "epoch": 0.80832,
+      "grad_norm": 0.4211399257183075,
+      "learning_rate": 9.32611749183846e-05,
+      "loss": 2.9025,
+      "step": 6315
+    },
+    {
+      "epoch": 0.808448,
+      "grad_norm": 0.43430060148239136,
+      "learning_rate": 9.31406385308915e-05,
+      "loss": 2.755,
+      "step": 6316
+    },
+    {
+      "epoch": 0.808576,
+      "grad_norm": 0.44631633162498474,
+      "learning_rate": 9.30201720872395e-05,
+      "loss": 2.822,
+      "step": 6317
+    },
+    {
+      "epoch": 0.808704,
+      "grad_norm": 0.42229968309402466,
+      "learning_rate": 9.289977560813789e-05,
+      "loss": 2.8491,
+      "step": 6318
+    },
+    {
+      "epoch": 0.808832,
+      "grad_norm": 0.4112605154514313,
+      "learning_rate": 9.277944911428449e-05,
+      "loss": 2.7991,
+      "step": 6319
+    },
+    {
+      "epoch": 0.80896,
+      "grad_norm": 0.4035566747188568,
+      "learning_rate": 9.265919262636468e-05,
+      "loss": 2.7112,
+      "step": 6320
+    },
+    {
+      "epoch": 0.809088,
+      "grad_norm": 0.41175609827041626,
+      "learning_rate": 9.2539006165052e-05,
+      "loss": 2.7719,
+      "step": 6321
+    },
+    {
+      "epoch": 0.809216,
+      "grad_norm": 0.4218294620513916,
+      "learning_rate": 9.241888975100782e-05,
+      "loss": 2.8951,
+      "step": 6322
+    },
+    {
+      "epoch": 0.809344,
+      "grad_norm": 0.44411829113960266,
+      "learning_rate": 9.229884340488149e-05,
+      "loss": 2.6992,
+      "step": 6323
+    },
+    {
+      "epoch": 0.809472,
+      "grad_norm": 0.4142504334449768,
+      "learning_rate": 9.217886714731055e-05,
+      "loss": 2.7279,
+      "step": 6324
+    },
+    {
+      "epoch": 0.8096,
+      "grad_norm": 0.451833575963974,
+      "learning_rate": 9.205896099892019e-05,
+      "loss": 2.7951,
+      "step": 6325
+    },
+    {
+      "epoch": 0.809728,
+      "grad_norm": 0.38776570558547974,
+      "learning_rate": 9.193912498032375e-05,
+      "loss": 2.6814,
+      "step": 6326
+    },
+    {
+      "epoch": 0.809856,
+      "grad_norm": 0.4137299358844757,
+      "learning_rate": 9.18193591121223e-05,
+      "loss": 2.6892,
+      "step": 6327
+    },
+    {
+      "epoch": 0.809984,
+      "grad_norm": 0.4567056894302368,
+      "learning_rate": 9.169966341490499e-05,
+      "loss": 2.7798,
+      "step": 6328
+    },
+    {
+      "epoch": 0.810112,
+      "grad_norm": 0.40829551219940186,
+      "learning_rate": 9.158003790924906e-05,
+      "loss": 2.7701,
+      "step": 6329
+    },
+    {
+      "epoch": 0.81024,
+      "grad_norm": 0.40476807951927185,
+      "learning_rate": 9.146048261571944e-05,
+      "loss": 2.7246,
+      "step": 6330
+    },
+    {
+      "epoch": 0.810368,
+      "grad_norm": 0.43633773922920227,
+      "learning_rate": 9.134099755486891e-05,
+      "loss": 2.7284,
+      "step": 6331
+    },
+    {
+      "epoch": 0.810496,
+      "grad_norm": 0.40589866042137146,
+      "learning_rate": 9.122158274723863e-05,
+      "loss": 2.7763,
+      "step": 6332
+    },
+    {
+      "epoch": 0.810624,
+      "grad_norm": 0.4703599512577057,
+      "learning_rate": 9.110223821335723e-05,
+      "loss": 2.9567,
+      "step": 6333
+    },
+    {
+      "epoch": 0.810752,
+      "grad_norm": 0.3914400339126587,
+      "learning_rate": 9.098296397374145e-05,
+      "loss": 2.6282,
+      "step": 6334
+    },
+    {
+      "epoch": 0.81088,
+      "grad_norm": 0.42517101764678955,
+      "learning_rate": 9.08637600488959e-05,
+      "loss": 2.8113,
+      "step": 6335
+    },
+    {
+      "epoch": 0.811008,
+      "grad_norm": 0.44266965985298157,
+      "learning_rate": 9.074462645931308e-05,
+      "loss": 2.9034,
+      "step": 6336
+    },
+    {
+      "epoch": 0.811136,
+      "grad_norm": 0.4304594099521637,
+      "learning_rate": 9.062556322547333e-05,
+      "loss": 2.7554,
+      "step": 6337
+    },
+    {
+      "epoch": 0.811264,
+      "grad_norm": 0.42232954502105713,
+      "learning_rate": 9.050657036784515e-05,
+      "loss": 2.9583,
+      "step": 6338
+    },
+    {
+      "epoch": 0.811392,
+      "grad_norm": 0.4000717103481293,
+      "learning_rate": 9.038764790688492e-05,
+      "loss": 2.799,
+      "step": 6339
+    },
+    {
+      "epoch": 0.81152,
+      "grad_norm": 0.4131116271018982,
+      "learning_rate": 9.026879586303666e-05,
+      "loss": 2.8678,
+      "step": 6340
+    },
+    {
+      "epoch": 0.811648,
+      "grad_norm": 0.42350253462791443,
+      "learning_rate": 9.015001425673241e-05,
+      "loss": 2.7463,
+      "step": 6341
+    },
+    {
+      "epoch": 0.811776,
+      "grad_norm": 0.40651002526283264,
+      "learning_rate": 9.003130310839203e-05,
+      "loss": 2.8458,
+      "step": 6342
+    },
+    {
+      "epoch": 0.811904,
+      "grad_norm": 0.3976232707500458,
+      "learning_rate": 8.99126624384235e-05,
+      "loss": 2.8385,
+      "step": 6343
+    },
+    {
+      "epoch": 0.812032,
+      "grad_norm": 0.4117695093154907,
+      "learning_rate": 8.979409226722224e-05,
+      "loss": 2.9153,
+      "step": 6344
+    },
+    {
+      "epoch": 0.81216,
+      "grad_norm": 0.44803059101104736,
+      "learning_rate": 8.967559261517217e-05,
+      "loss": 2.8785,
+      "step": 6345
+    },
+    {
+      "epoch": 0.812288,
+      "grad_norm": 0.3883678913116455,
+      "learning_rate": 8.955716350264454e-05,
+      "loss": 2.6179,
+      "step": 6346
+    },
+    {
+      "epoch": 0.812416,
+      "grad_norm": 0.3924500644207001,
+      "learning_rate": 8.943880494999885e-05,
+      "loss": 2.7332,
+      "step": 6347
+    },
+    {
+      "epoch": 0.812544,
+      "grad_norm": 0.40052521228790283,
+      "learning_rate": 8.932051697758225e-05,
+      "loss": 2.756,
+      "step": 6348
+    },
+    {
+      "epoch": 0.812672,
+      "grad_norm": 0.4075007736682892,
+      "learning_rate": 8.920229960572973e-05,
+      "loss": 2.8879,
+      "step": 6349
+    },
+    {
+      "epoch": 0.8128,
+      "grad_norm": 0.4105590879917145,
+      "learning_rate": 8.908415285476434e-05,
+      "loss": 2.7744,
+      "step": 6350
+    },
+    {
+      "epoch": 0.812928,
+      "grad_norm": 0.40679478645324707,
+      "learning_rate": 8.896607674499679e-05,
+      "loss": 2.8765,
+      "step": 6351
+    },
+    {
+      "epoch": 0.813056,
+      "grad_norm": 0.43842384219169617,
+      "learning_rate": 8.884807129672567e-05,
+      "loss": 2.7749,
+      "step": 6352
+    },
+    {
+      "epoch": 0.813184,
+      "grad_norm": 0.3925323188304901,
+      "learning_rate": 8.873013653023765e-05,
+      "loss": 2.7871,
+      "step": 6353
+    },
+    {
+      "epoch": 0.813312,
+      "grad_norm": 0.4006810784339905,
+      "learning_rate": 8.861227246580705e-05,
+      "loss": 2.718,
+      "step": 6354
+    },
+    {
+      "epoch": 0.81344,
+      "grad_norm": 0.39854615926742554,
+      "learning_rate": 8.84944791236959e-05,
+      "loss": 2.8196,
+      "step": 6355
+    },
+    {
+      "epoch": 0.813568,
+      "grad_norm": 0.379446417093277,
+      "learning_rate": 8.83767565241545e-05,
+      "loss": 2.8209,
+      "step": 6356
+    },
+    {
+      "epoch": 0.813696,
+      "grad_norm": 0.4078393876552582,
+      "learning_rate": 8.82591046874206e-05,
+      "loss": 2.8755,
+      "step": 6357
+    },
+    {
+      "epoch": 0.813824,
+      "grad_norm": 0.4225532114505768,
+      "learning_rate": 8.814152363371992e-05,
+      "loss": 2.7287,
+      "step": 6358
+    },
+    {
+      "epoch": 0.813952,
+      "grad_norm": 0.42682400345802307,
+      "learning_rate": 8.802401338326582e-05,
+      "loss": 2.9176,
+      "step": 6359
+    },
+    {
+      "epoch": 0.81408,
+      "grad_norm": 0.38744431734085083,
+      "learning_rate": 8.790657395626001e-05,
+      "loss": 2.8236,
+      "step": 6360
+    },
+    {
+      "epoch": 0.814208,
+      "grad_norm": 0.38855916261672974,
+      "learning_rate": 8.778920537289154e-05,
+      "loss": 2.8216,
+      "step": 6361
+    },
+    {
+      "epoch": 0.814336,
+      "grad_norm": 0.42123064398765564,
+      "learning_rate": 8.767190765333744e-05,
+      "loss": 2.7098,
+      "step": 6362
+    },
+    {
+      "epoch": 0.814464,
+      "grad_norm": 0.41521748900413513,
+      "learning_rate": 8.755468081776253e-05,
+      "loss": 2.8139,
+      "step": 6363
+    },
+    {
+      "epoch": 0.814592,
+      "grad_norm": 0.4252949357032776,
+      "learning_rate": 8.743752488631945e-05,
+      "loss": 2.8844,
+      "step": 6364
+    },
+    {
+      "epoch": 0.81472,
+      "grad_norm": 0.4268285632133484,
+      "learning_rate": 8.732043987914857e-05,
+      "loss": 2.8126,
+      "step": 6365
+    },
+    {
+      "epoch": 0.814848,
+      "grad_norm": 0.4224264323711395,
+      "learning_rate": 8.720342581637836e-05,
+      "loss": 2.8149,
+      "step": 6366
+    },
+    {
+      "epoch": 0.814976,
+      "grad_norm": 0.4118562340736389,
+      "learning_rate": 8.70864827181247e-05,
+      "loss": 2.8408,
+      "step": 6367
+    },
+    {
+      "epoch": 0.815104,
+      "grad_norm": 0.4168948829174042,
+      "learning_rate": 8.696961060449165e-05,
+      "loss": 2.7458,
+      "step": 6368
+    },
+    {
+      "epoch": 0.815232,
+      "grad_norm": 0.39754167199134827,
+      "learning_rate": 8.685280949557084e-05,
+      "loss": 2.7415,
+      "step": 6369
+    },
+    {
+      "epoch": 0.81536,
+      "grad_norm": 0.42551249265670776,
+      "learning_rate": 8.673607941144169e-05,
+      "loss": 2.8776,
+      "step": 6370
+    },
+    {
+      "epoch": 0.815488,
+      "grad_norm": 0.37106308341026306,
+      "learning_rate": 8.66194203721714e-05,
+      "loss": 2.8035,
+      "step": 6371
+    },
+    {
+      "epoch": 0.815616,
+      "grad_norm": 0.40338441729545593,
+      "learning_rate": 8.650283239781515e-05,
+      "loss": 2.7614,
+      "step": 6372
+    },
+    {
+      "epoch": 0.815744,
+      "grad_norm": 0.4138956367969513,
+      "learning_rate": 8.638631550841552e-05,
+      "loss": 2.7649,
+      "step": 6373
+    },
+    {
+      "epoch": 0.815872,
+      "grad_norm": 0.3929102122783661,
+      "learning_rate": 8.626986972400325e-05,
+      "loss": 2.7781,
+      "step": 6374
+    },
+    {
+      "epoch": 0.816,
+      "grad_norm": 0.44573137164115906,
+      "learning_rate": 8.615349506459692e-05,
+      "loss": 2.9579,
+      "step": 6375
+    },
+    {
+      "epoch": 0.816128,
+      "grad_norm": 0.4179004430770874,
+      "learning_rate": 8.603719155020246e-05,
+      "loss": 2.6466,
+      "step": 6376
+    },
+    {
+      "epoch": 0.816256,
+      "grad_norm": 0.41491764783859253,
+      "learning_rate": 8.592095920081383e-05,
+      "loss": 2.8375,
+      "step": 6377
+    },
+    {
+      "epoch": 0.816384,
+      "grad_norm": 0.4214448928833008,
+      "learning_rate": 8.580479803641278e-05,
+      "loss": 2.8677,
+      "step": 6378
+    },
+    {
+      "epoch": 0.816512,
+      "grad_norm": 0.3984643518924713,
+      "learning_rate": 8.568870807696872e-05,
+      "loss": 2.7812,
+      "step": 6379
+    },
+    {
+      "epoch": 0.81664,
+      "grad_norm": 0.36198869347572327,
+      "learning_rate": 8.557268934243867e-05,
+      "loss": 2.7544,
+      "step": 6380
+    },
+    {
+      "epoch": 0.816768,
+      "grad_norm": 0.4011256694793701,
+      "learning_rate": 8.545674185276791e-05,
+      "loss": 2.6778,
+      "step": 6381
+    },
+    {
+      "epoch": 0.816896,
+      "grad_norm": 0.43435898423194885,
+      "learning_rate": 8.534086562788906e-05,
+      "loss": 2.823,
+      "step": 6382
+    },
+    {
+      "epoch": 0.817024,
+      "grad_norm": 0.3905279040336609,
+      "learning_rate": 8.52250606877224e-05,
+      "loss": 2.8617,
+      "step": 6383
+    },
+    {
+      "epoch": 0.817152,
+      "grad_norm": 0.3794632852077484,
+      "learning_rate": 8.510932705217644e-05,
+      "loss": 2.8516,
+      "step": 6384
+    },
+    {
+      "epoch": 0.81728,
+      "grad_norm": 0.44369086623191833,
+      "learning_rate": 8.499366474114695e-05,
+      "loss": 2.8222,
+      "step": 6385
+    },
+    {
+      "epoch": 0.817408,
+      "grad_norm": 0.4074725806713104,
+      "learning_rate": 8.487807377451767e-05,
+      "loss": 2.7385,
+      "step": 6386
+    },
+    {
+      "epoch": 0.817536,
+      "grad_norm": 0.40878990292549133,
+      "learning_rate": 8.476255417216005e-05,
+      "loss": 2.7469,
+      "step": 6387
+    },
+    {
+      "epoch": 0.817664,
+      "grad_norm": 0.3926815390586853,
+      "learning_rate": 8.464710595393305e-05,
+      "loss": 2.8717,
+      "step": 6388
+    },
+    {
+      "epoch": 0.817792,
+      "grad_norm": 0.3891645669937134,
+      "learning_rate": 8.453172913968382e-05,
+      "loss": 2.7291,
+      "step": 6389
+    },
+    {
+      "epoch": 0.81792,
+      "grad_norm": 0.3813799023628235,
+      "learning_rate": 8.44164237492469e-05,
+      "loss": 2.6938,
+      "step": 6390
+    },
+    {
+      "epoch": 0.818048,
+      "grad_norm": 0.3984074294567108,
+      "learning_rate": 8.430118980244461e-05,
+      "loss": 2.7729,
+      "step": 6391
+    },
+    {
+      "epoch": 0.818176,
+      "grad_norm": 0.41259172558784485,
+      "learning_rate": 8.418602731908686e-05,
+      "loss": 2.7326,
+      "step": 6392
+    },
+    {
+      "epoch": 0.818304,
+      "grad_norm": 0.4000777304172516,
+      "learning_rate": 8.407093631897167e-05,
+      "loss": 2.8199,
+      "step": 6393
+    },
+    {
+      "epoch": 0.818432,
+      "grad_norm": 0.39089617133140564,
+      "learning_rate": 8.395591682188442e-05,
+      "loss": 2.7542,
+      "step": 6394
+    },
+    {
+      "epoch": 0.81856,
+      "grad_norm": 0.4520619809627533,
+      "learning_rate": 8.384096884759807e-05,
+      "loss": 2.8881,
+      "step": 6395
+    },
+    {
+      "epoch": 0.818688,
+      "grad_norm": 0.4149484634399414,
+      "learning_rate": 8.372609241587387e-05,
+      "loss": 2.7859,
+      "step": 6396
+    },
+    {
+      "epoch": 0.818816,
+      "grad_norm": 0.4166579842567444,
+      "learning_rate": 8.361128754646024e-05,
+      "loss": 2.6646,
+      "step": 6397
+    },
+    {
+      "epoch": 0.818944,
+      "grad_norm": 0.4467387795448303,
+      "learning_rate": 8.349655425909347e-05,
+      "loss": 2.9191,
+      "step": 6398
+    },
+    {
+      "epoch": 0.819072,
+      "grad_norm": 0.40210139751434326,
+      "learning_rate": 8.338189257349754e-05,
+      "loss": 2.8064,
+      "step": 6399
+    },
+    {
+      "epoch": 0.8192,
+      "grad_norm": 0.3998998701572418,
+      "learning_rate": 8.326730250938413e-05,
+      "loss": 2.6699,
+      "step": 6400
+    },
+    {
+      "epoch": 0.819328,
+      "grad_norm": 0.4435540735721588,
+      "learning_rate": 8.31527840864525e-05,
+      "loss": 2.8448,
+      "step": 6401
+    },
+    {
+      "epoch": 0.819456,
+      "grad_norm": 0.38882145285606384,
+      "learning_rate": 8.303833732438987e-05,
+      "loss": 2.812,
+      "step": 6402
+    },
+    {
+      "epoch": 0.819584,
+      "grad_norm": 0.4324854612350464,
+      "learning_rate": 8.29239622428708e-05,
+      "loss": 2.8525,
+      "step": 6403
+    },
+    {
+      "epoch": 0.819712,
+      "grad_norm": 0.4176493287086487,
+      "learning_rate": 8.28096588615579e-05,
+      "loss": 2.8272,
+      "step": 6404
+    },
+    {
+      "epoch": 0.81984,
+      "grad_norm": 0.4246494770050049,
+      "learning_rate": 8.269542720010109e-05,
+      "loss": 2.6849,
+      "step": 6405
+    },
+    {
+      "epoch": 0.819968,
+      "grad_norm": 0.40853849053382874,
+      "learning_rate": 8.25812672781382e-05,
+      "loss": 2.7644,
+      "step": 6406
+    },
+    {
+      "epoch": 0.820096,
+      "grad_norm": 0.4398989677429199,
+      "learning_rate": 8.246717911529456e-05,
+      "loss": 2.7934,
+      "step": 6407
+    },
+    {
+      "epoch": 0.820224,
+      "grad_norm": 0.45662349462509155,
+      "learning_rate": 8.235316273118332e-05,
+      "loss": 2.8313,
+      "step": 6408
+    },
+    {
+      "epoch": 0.820352,
+      "grad_norm": 0.39702558517456055,
+      "learning_rate": 8.223921814540503e-05,
+      "loss": 2.6893,
+      "step": 6409
+    },
+    {
+      "epoch": 0.82048,
+      "grad_norm": 0.3795692026615143,
+      "learning_rate": 8.21253453775484e-05,
+      "loss": 2.7029,
+      "step": 6410
+    },
+    {
+      "epoch": 0.820608,
+      "grad_norm": 0.5672438740730286,
+      "learning_rate": 8.201154444718917e-05,
+      "loss": 2.8449,
+      "step": 6411
+    },
+    {
+      "epoch": 0.820736,
+      "grad_norm": 0.3926500380039215,
+      "learning_rate": 8.189781537389135e-05,
+      "loss": 2.8581,
+      "step": 6412
+    },
+    {
+      "epoch": 0.820864,
+      "grad_norm": 0.41469940543174744,
+      "learning_rate": 8.178415817720608e-05,
+      "loss": 2.7959,
+      "step": 6413
+    },
+    {
+      "epoch": 0.820992,
+      "grad_norm": 0.4103841185569763,
+      "learning_rate": 8.16705728766724e-05,
+      "loss": 2.7967,
+      "step": 6414
+    },
+    {
+      "epoch": 0.82112,
+      "grad_norm": 0.41807839274406433,
+      "learning_rate": 8.155705949181691e-05,
+      "loss": 2.8575,
+      "step": 6415
+    },
+    {
+      "epoch": 0.821248,
+      "grad_norm": 0.4147208333015442,
+      "learning_rate": 8.144361804215383e-05,
+      "loss": 2.8334,
+      "step": 6416
+    },
+    {
+      "epoch": 0.821376,
+      "grad_norm": 0.4193754196166992,
+      "learning_rate": 8.133024854718524e-05,
+      "loss": 2.7562,
+      "step": 6417
+    },
+    {
+      "epoch": 0.821504,
+      "grad_norm": 0.3973667025566101,
+      "learning_rate": 8.121695102640054e-05,
+      "loss": 2.7642,
+      "step": 6418
+    },
+    {
+      "epoch": 0.821632,
+      "grad_norm": 0.4274323880672455,
+      "learning_rate": 8.110372549927692e-05,
+      "loss": 2.7792,
+      "step": 6419
+    },
+    {
+      "epoch": 0.82176,
+      "grad_norm": 0.3834474980831146,
+      "learning_rate": 8.099057198527899e-05,
+      "loss": 2.8618,
+      "step": 6420
+    },
+    {
+      "epoch": 0.821888,
+      "grad_norm": 0.42203307151794434,
+      "learning_rate": 8.087749050385951e-05,
+      "loss": 2.777,
+      "step": 6421
+    },
+    {
+      "epoch": 0.822016,
+      "grad_norm": 0.4145670533180237,
+      "learning_rate": 8.076448107445822e-05,
+      "loss": 2.7717,
+      "step": 6422
+    },
+    {
+      "epoch": 0.822144,
+      "grad_norm": 0.38453733921051025,
+      "learning_rate": 8.065154371650285e-05,
+      "loss": 2.7175,
+      "step": 6423
+    },
+    {
+      "epoch": 0.822272,
+      "grad_norm": 0.38994100689888,
+      "learning_rate": 8.053867844940854e-05,
+      "loss": 2.721,
+      "step": 6424
+    },
+    {
+      "epoch": 0.8224,
+      "grad_norm": 0.43487119674682617,
+      "learning_rate": 8.042588529257828e-05,
+      "loss": 2.9093,
+      "step": 6425
+    },
+    {
+      "epoch": 0.822528,
+      "grad_norm": 0.41066858172416687,
+      "learning_rate": 8.031316426540252e-05,
+      "loss": 2.8784,
+      "step": 6426
+    },
+    {
+      "epoch": 0.822656,
+      "grad_norm": 0.4165816307067871,
+      "learning_rate": 8.02005153872592e-05,
+      "loss": 2.751,
+      "step": 6427
+    },
+    {
+      "epoch": 0.822784,
+      "grad_norm": 0.37504756450653076,
+      "learning_rate": 8.008793867751402e-05,
+      "loss": 2.6755,
+      "step": 6428
+    },
+    {
+      "epoch": 0.822912,
+      "grad_norm": 0.413067102432251,
+      "learning_rate": 7.997543415552011e-05,
+      "loss": 2.8048,
+      "step": 6429
+    },
+    {
+      "epoch": 0.82304,
+      "grad_norm": 0.39847639203071594,
+      "learning_rate": 7.986300184061856e-05,
+      "loss": 2.7865,
+      "step": 6430
+    },
+    {
+      "epoch": 0.823168,
+      "grad_norm": 0.4495554268360138,
+      "learning_rate": 7.975064175213748e-05,
+      "loss": 2.8567,
+      "step": 6431
+    },
+    {
+      "epoch": 0.823296,
+      "grad_norm": 0.38329777121543884,
+      "learning_rate": 7.963835390939316e-05,
+      "loss": 2.7364,
+      "step": 6432
+    },
+    {
+      "epoch": 0.823424,
+      "grad_norm": 0.38991114497184753,
+      "learning_rate": 7.952613833168909e-05,
+      "loss": 2.8483,
+      "step": 6433
+    },
+    {
+      "epoch": 0.823552,
+      "grad_norm": 0.3777281939983368,
+      "learning_rate": 7.941399503831637e-05,
+      "loss": 2.755,
+      "step": 6434
+    },
+    {
+      "epoch": 0.82368,
+      "grad_norm": 0.3889419436454773,
+      "learning_rate": 7.930192404855374e-05,
+      "loss": 2.7226,
+      "step": 6435
+    },
+    {
+      "epoch": 0.823808,
+      "grad_norm": 0.4205712676048279,
+      "learning_rate": 7.918992538166752e-05,
+      "loss": 2.9265,
+      "step": 6436
+    },
+    {
+      "epoch": 0.823936,
+      "grad_norm": 0.4140024185180664,
+      "learning_rate": 7.907799905691143e-05,
+      "loss": 2.8132,
+      "step": 6437
+    },
+    {
+      "epoch": 0.824064,
+      "grad_norm": 0.39202237129211426,
+      "learning_rate": 7.896614509352723e-05,
+      "loss": 2.7201,
+      "step": 6438
+    },
+    {
+      "epoch": 0.824192,
+      "grad_norm": 0.45322975516319275,
+      "learning_rate": 7.885436351074355e-05,
+      "loss": 2.7638,
+      "step": 6439
+    },
+    {
+      "epoch": 0.82432,
+      "grad_norm": 0.43222522735595703,
+      "learning_rate": 7.874265432777728e-05,
+      "loss": 2.7837,
+      "step": 6440
+    },
+    {
+      "epoch": 0.824448,
+      "grad_norm": 0.3907621204853058,
+      "learning_rate": 7.863101756383235e-05,
+      "loss": 2.7029,
+      "step": 6441
+    },
+    {
+      "epoch": 0.824576,
+      "grad_norm": 0.4315086305141449,
+      "learning_rate": 7.851945323810045e-05,
+      "loss": 2.8133,
+      "step": 6442
+    },
+    {
+      "epoch": 0.824704,
+      "grad_norm": 0.4195539653301239,
+      "learning_rate": 7.840796136976075e-05,
+      "loss": 2.7919,
+      "step": 6443
+    },
+    {
+      "epoch": 0.824832,
+      "grad_norm": 0.3921598196029663,
+      "learning_rate": 7.829654197797998e-05,
+      "loss": 2.858,
+      "step": 6444
+    },
+    {
+      "epoch": 0.82496,
+      "grad_norm": 0.3952738046646118,
+      "learning_rate": 7.818519508191235e-05,
+      "loss": 2.8133,
+      "step": 6445
+    },
+    {
+      "epoch": 0.825088,
+      "grad_norm": 0.38022229075431824,
+      "learning_rate": 7.807392070069991e-05,
+      "loss": 2.873,
+      "step": 6446
+    },
+    {
+      "epoch": 0.825216,
+      "grad_norm": 0.4111633002758026,
+      "learning_rate": 7.796271885347189e-05,
+      "loss": 2.8131,
+      "step": 6447
+    },
+    {
+      "epoch": 0.825344,
+      "grad_norm": 0.4143979251384735,
+      "learning_rate": 7.785158955934507e-05,
+      "loss": 2.861,
+      "step": 6448
+    },
+    {
+      "epoch": 0.825472,
+      "grad_norm": 0.4235285222530365,
+      "learning_rate": 7.774053283742406e-05,
+      "loss": 2.7059,
+      "step": 6449
+    },
+    {
+      "epoch": 0.8256,
+      "grad_norm": 0.40903136134147644,
+      "learning_rate": 7.762954870680067e-05,
+      "loss": 2.9307,
+      "step": 6450
+    },
+    {
+      "epoch": 0.825728,
+      "grad_norm": 0.3983044922351837,
+      "learning_rate": 7.751863718655444e-05,
+      "loss": 2.8554,
+      "step": 6451
+    },
+    {
+      "epoch": 0.825856,
+      "grad_norm": 0.40979716181755066,
+      "learning_rate": 7.740779829575217e-05,
+      "loss": 2.8461,
+      "step": 6452
+    },
+    {
+      "epoch": 0.825984,
+      "grad_norm": 0.39273661375045776,
+      "learning_rate": 7.729703205344863e-05,
+      "loss": 2.8508,
+      "step": 6453
+    },
+    {
+      "epoch": 0.826112,
+      "grad_norm": 0.4319429397583008,
+      "learning_rate": 7.718633847868567e-05,
+      "loss": 2.8491,
+      "step": 6454
+    },
+    {
+      "epoch": 0.82624,
+      "grad_norm": 0.422011137008667,
+      "learning_rate": 7.707571759049281e-05,
+      "loss": 2.7164,
+      "step": 6455
+    },
+    {
+      "epoch": 0.826368,
+      "grad_norm": 0.4050493836402893,
+      "learning_rate": 7.6965169407887e-05,
+      "loss": 2.7341,
+      "step": 6456
+    },
+    {
+      "epoch": 0.826496,
+      "grad_norm": 0.4322655498981476,
+      "learning_rate": 7.685469394987271e-05,
+      "loss": 2.7526,
+      "step": 6457
+    },
+    {
+      "epoch": 0.826624,
+      "grad_norm": 0.4072965979576111,
+      "learning_rate": 7.674429123544218e-05,
+      "loss": 2.889,
+      "step": 6458
+    },
+    {
+      "epoch": 0.826752,
+      "grad_norm": 0.451468288898468,
+      "learning_rate": 7.66339612835748e-05,
+      "loss": 2.8541,
+      "step": 6459
+    },
+    {
+      "epoch": 0.82688,
+      "grad_norm": 0.4187624752521515,
+      "learning_rate": 7.652370411323744e-05,
+      "loss": 2.7703,
+      "step": 6460
+    },
+    {
+      "epoch": 0.827008,
+      "grad_norm": 0.3749317526817322,
+      "learning_rate": 7.641351974338478e-05,
+      "loss": 2.6939,
+      "step": 6461
+    },
+    {
+      "epoch": 0.827136,
+      "grad_norm": 0.5535447597503662,
+      "learning_rate": 7.630340819295878e-05,
+      "loss": 2.7989,
+      "step": 6462
+    },
+    {
+      "epoch": 0.827264,
+      "grad_norm": 0.43238818645477295,
+      "learning_rate": 7.619336948088879e-05,
+      "loss": 2.7651,
+      "step": 6463
+    },
+    {
+      "epoch": 0.827392,
+      "grad_norm": 0.43169674277305603,
+      "learning_rate": 7.608340362609173e-05,
+      "loss": 2.8348,
+      "step": 6464
+    },
+    {
+      "epoch": 0.82752,
+      "grad_norm": 0.4379948377609253,
+      "learning_rate": 7.59735106474721e-05,
+      "loss": 2.7851,
+      "step": 6465
+    },
+    {
+      "epoch": 0.827648,
+      "grad_norm": 0.38960450887680054,
+      "learning_rate": 7.586369056392161e-05,
+      "loss": 2.7361,
+      "step": 6466
+    },
+    {
+      "epoch": 0.827776,
+      "grad_norm": 0.4070088267326355,
+      "learning_rate": 7.575394339431969e-05,
+      "loss": 2.8623,
+      "step": 6467
+    },
+    {
+      "epoch": 0.827904,
+      "grad_norm": 0.40492597222328186,
+      "learning_rate": 7.564426915753331e-05,
+      "loss": 2.7538,
+      "step": 6468
+    },
+    {
+      "epoch": 0.828032,
+      "grad_norm": 0.42275893688201904,
+      "learning_rate": 7.553466787241664e-05,
+      "loss": 2.7581,
+      "step": 6469
+    },
+    {
+      "epoch": 0.82816,
+      "grad_norm": 0.4216054081916809,
+      "learning_rate": 7.542513955781138e-05,
+      "loss": 2.7649,
+      "step": 6470
+    },
+    {
+      "epoch": 0.828288,
+      "grad_norm": 0.42190900444984436,
+      "learning_rate": 7.53156842325467e-05,
+      "loss": 2.7429,
+      "step": 6471
+    },
+    {
+      "epoch": 0.828416,
+      "grad_norm": 0.3981059491634369,
+      "learning_rate": 7.520630191543932e-05,
+      "loss": 2.7736,
+      "step": 6472
+    },
+    {
+      "epoch": 0.828544,
+      "grad_norm": 0.3897697627544403,
+      "learning_rate": 7.509699262529307e-05,
+      "loss": 2.8997,
+      "step": 6473
+    },
+    {
+      "epoch": 0.828672,
+      "grad_norm": 0.4167628288269043,
+      "learning_rate": 7.498775638089988e-05,
+      "loss": 2.9448,
+      "step": 6474
+    },
+    {
+      "epoch": 0.8288,
+      "grad_norm": 0.3865152895450592,
+      "learning_rate": 7.487859320103846e-05,
+      "loss": 2.8058,
+      "step": 6475
+    },
+    {
+      "epoch": 0.828928,
+      "grad_norm": 0.4414910674095154,
+      "learning_rate": 7.476950310447522e-05,
+      "loss": 2.7902,
+      "step": 6476
+    },
+    {
+      "epoch": 0.829056,
+      "grad_norm": 0.40097054839134216,
+      "learning_rate": 7.466048610996423e-05,
+      "loss": 2.8288,
+      "step": 6477
+    },
+    {
+      "epoch": 0.829184,
+      "grad_norm": 0.40266332030296326,
+      "learning_rate": 7.455154223624661e-05,
+      "loss": 2.7066,
+      "step": 6478
+    },
+    {
+      "epoch": 0.829312,
+      "grad_norm": 0.42418166995048523,
+      "learning_rate": 7.444267150205108e-05,
+      "loss": 2.8836,
+      "step": 6479
+    },
+    {
+      "epoch": 0.82944,
+      "grad_norm": 0.38977089524269104,
+      "learning_rate": 7.433387392609387e-05,
+      "loss": 2.7555,
+      "step": 6480
+    },
+    {
+      "epoch": 0.829568,
+      "grad_norm": 0.3858592212200165,
+      "learning_rate": 7.422514952707832e-05,
+      "loss": 2.8171,
+      "step": 6481
+    },
+    {
+      "epoch": 0.829696,
+      "grad_norm": 0.39577892422676086,
+      "learning_rate": 7.411649832369566e-05,
+      "loss": 2.7742,
+      "step": 6482
+    },
+    {
+      "epoch": 0.829824,
+      "grad_norm": 0.3723192811012268,
+      "learning_rate": 7.400792033462428e-05,
+      "loss": 2.6633,
+      "step": 6483
+    },
+    {
+      "epoch": 0.829952,
+      "grad_norm": 0.42295482754707336,
+      "learning_rate": 7.389941557852986e-05,
+      "loss": 2.7256,
+      "step": 6484
+    },
+    {
+      "epoch": 0.83008,
+      "grad_norm": 0.37720727920532227,
+      "learning_rate": 7.379098407406553e-05,
+      "loss": 2.7666,
+      "step": 6485
+    },
+    {
+      "epoch": 0.830208,
+      "grad_norm": 0.39416807889938354,
+      "learning_rate": 7.368262583987228e-05,
+      "loss": 2.7992,
+      "step": 6486
+    },
+    {
+      "epoch": 0.830336,
+      "grad_norm": 0.3967745304107666,
+      "learning_rate": 7.357434089457788e-05,
+      "loss": 2.8923,
+      "step": 6487
+    },
+    {
+      "epoch": 0.830464,
+      "grad_norm": 0.38864561915397644,
+      "learning_rate": 7.346612925679775e-05,
+      "loss": 2.7159,
+      "step": 6488
+    },
+    {
+      "epoch": 0.830592,
+      "grad_norm": 0.41957226395606995,
+      "learning_rate": 7.33579909451349e-05,
+      "loss": 2.8649,
+      "step": 6489
+    },
+    {
+      "epoch": 0.83072,
+      "grad_norm": 0.4291616976261139,
+      "learning_rate": 7.324992597817947e-05,
+      "loss": 2.9689,
+      "step": 6490
+    },
+    {
+      "epoch": 0.830848,
+      "grad_norm": 0.4100084602832794,
+      "learning_rate": 7.314193437450912e-05,
+      "loss": 2.8017,
+      "step": 6491
+    },
+    {
+      "epoch": 0.830976,
+      "grad_norm": 0.4021975100040436,
+      "learning_rate": 7.303401615268878e-05,
+      "loss": 2.8785,
+      "step": 6492
+    },
+    {
+      "epoch": 0.831104,
+      "grad_norm": 0.4174312651157379,
+      "learning_rate": 7.292617133127083e-05,
+      "loss": 2.8153,
+      "step": 6493
+    },
+    {
+      "epoch": 0.831232,
+      "grad_norm": 0.4363156259059906,
+      "learning_rate": 7.281839992879502e-05,
+      "loss": 2.8954,
+      "step": 6494
+    },
+    {
+      "epoch": 0.83136,
+      "grad_norm": 0.4124840497970581,
+      "learning_rate": 7.27107019637886e-05,
+      "loss": 2.6559,
+      "step": 6495
+    },
+    {
+      "epoch": 0.831488,
+      "grad_norm": 0.4032273292541504,
+      "learning_rate": 7.260307745476619e-05,
+      "loss": 2.9361,
+      "step": 6496
+    },
+    {
+      "epoch": 0.831616,
+      "grad_norm": 0.40921270847320557,
+      "learning_rate": 7.249552642022955e-05,
+      "loss": 2.747,
+      "step": 6497
+    },
+    {
+      "epoch": 0.831744,
+      "grad_norm": 0.41884079575538635,
+      "learning_rate": 7.238804887866796e-05,
+      "loss": 2.6878,
+      "step": 6498
+    },
+    {
+      "epoch": 0.831872,
+      "grad_norm": 0.43849366903305054,
+      "learning_rate": 7.228064484855806e-05,
+      "loss": 2.8146,
+      "step": 6499
+    },
+    {
+      "epoch": 0.832,
+      "grad_norm": 0.4023958444595337,
+      "learning_rate": 7.217331434836394e-05,
+      "loss": 2.7561,
+      "step": 6500
+    },
+    {
+      "epoch": 0.832128,
+      "grad_norm": 0.395170122385025,
+      "learning_rate": 7.206605739653682e-05,
+      "loss": 2.8068,
+      "step": 6501
+    },
+    {
+      "epoch": 0.832256,
+      "grad_norm": 0.4328382611274719,
+      "learning_rate": 7.195887401151535e-05,
+      "loss": 2.6927,
+      "step": 6502
+    },
+    {
+      "epoch": 0.832384,
+      "grad_norm": 0.41992807388305664,
+      "learning_rate": 7.185176421172573e-05,
+      "loss": 2.7644,
+      "step": 6503
+    },
+    {
+      "epoch": 0.832512,
+      "grad_norm": 0.4242829382419586,
+      "learning_rate": 7.174472801558146e-05,
+      "loss": 2.7252,
+      "step": 6504
+    },
+    {
+      "epoch": 0.83264,
+      "grad_norm": 0.45012959837913513,
+      "learning_rate": 7.16377654414832e-05,
+      "loss": 2.8134,
+      "step": 6505
+    },
+    {
+      "epoch": 0.832768,
+      "grad_norm": 0.40006500482559204,
+      "learning_rate": 7.153087650781909e-05,
+      "loss": 2.8622,
+      "step": 6506
+    },
+    {
+      "epoch": 0.832896,
+      "grad_norm": 0.42664411664009094,
+      "learning_rate": 7.142406123296453e-05,
+      "loss": 2.7883,
+      "step": 6507
+    },
+    {
+      "epoch": 0.833024,
+      "grad_norm": 0.3821522891521454,
+      "learning_rate": 7.131731963528232e-05,
+      "loss": 2.6798,
+      "step": 6508
+    },
+    {
+      "epoch": 0.833152,
+      "grad_norm": 0.4018377661705017,
+      "learning_rate": 7.121065173312253e-05,
+      "loss": 2.8553,
+      "step": 6509
+    },
+    {
+      "epoch": 0.83328,
+      "grad_norm": 0.4366210997104645,
+      "learning_rate": 7.110405754482268e-05,
+      "loss": 2.7658,
+      "step": 6510
+    },
+    {
+      "epoch": 0.833408,
+      "grad_norm": 0.42097243666648865,
+      "learning_rate": 7.09975370887076e-05,
+      "loss": 2.8409,
+      "step": 6511
+    },
+    {
+      "epoch": 0.833536,
+      "grad_norm": 0.39027512073516846,
+      "learning_rate": 7.089109038308927e-05,
+      "loss": 2.7506,
+      "step": 6512
+    },
+    {
+      "epoch": 0.833664,
+      "grad_norm": 0.4219440817832947,
+      "learning_rate": 7.078471744626707e-05,
+      "loss": 2.88,
+      "step": 6513
+    },
+    {
+      "epoch": 0.833792,
+      "grad_norm": 0.42795637249946594,
+      "learning_rate": 7.067841829652794e-05,
+      "loss": 2.7438,
+      "step": 6514
+    },
+    {
+      "epoch": 0.83392,
+      "grad_norm": 0.4283522963523865,
+      "learning_rate": 7.057219295214579e-05,
+      "loss": 2.7145,
+      "step": 6515
+    },
+    {
+      "epoch": 0.834048,
+      "grad_norm": 0.496569961309433,
+      "learning_rate": 7.046604143138197e-05,
+      "loss": 2.6984,
+      "step": 6516
+    },
+    {
+      "epoch": 0.834176,
+      "grad_norm": 0.4131828844547272,
+      "learning_rate": 7.035996375248527e-05,
+      "loss": 2.7841,
+      "step": 6517
+    },
+    {
+      "epoch": 0.834304,
+      "grad_norm": 0.4044988751411438,
+      "learning_rate": 7.025395993369165e-05,
+      "loss": 2.8893,
+      "step": 6518
+    },
+    {
+      "epoch": 0.834432,
+      "grad_norm": 0.41351380944252014,
+      "learning_rate": 7.01480299932244e-05,
+      "loss": 2.8209,
+      "step": 6519
+    },
+    {
+      "epoch": 0.83456,
+      "grad_norm": 0.42337438464164734,
+      "learning_rate": 7.004217394929402e-05,
+      "loss": 2.7412,
+      "step": 6520
+    },
+    {
+      "epoch": 0.834688,
+      "grad_norm": 0.3967662751674652,
+      "learning_rate": 6.993639182009842e-05,
+      "loss": 2.8332,
+      "step": 6521
+    },
+    {
+      "epoch": 0.834816,
+      "grad_norm": 0.4133740961551666,
+      "learning_rate": 6.983068362382273e-05,
+      "loss": 2.8026,
+      "step": 6522
+    },
+    {
+      "epoch": 0.834944,
+      "grad_norm": 0.41317278146743774,
+      "learning_rate": 6.972504937863955e-05,
+      "loss": 2.8158,
+      "step": 6523
+    },
+    {
+      "epoch": 0.835072,
+      "grad_norm": 0.4408941864967346,
+      "learning_rate": 6.961948910270843e-05,
+      "loss": 2.7166,
+      "step": 6524
+    },
+    {
+      "epoch": 0.8352,
+      "grad_norm": 0.4327394962310791,
+      "learning_rate": 6.951400281417669e-05,
+      "loss": 2.7606,
+      "step": 6525
+    },
+    {
+      "epoch": 0.835328,
+      "grad_norm": 0.3733211159706116,
+      "learning_rate": 6.940859053117842e-05,
+      "loss": 2.7066,
+      "step": 6526
+    },
+    {
+      "epoch": 0.835456,
+      "grad_norm": 0.38126513361930847,
+      "learning_rate": 6.930325227183538e-05,
+      "loss": 2.798,
+      "step": 6527
+    },
+    {
+      "epoch": 0.835584,
+      "grad_norm": 0.3819724917411804,
+      "learning_rate": 6.919798805425626e-05,
+      "loss": 2.7338,
+      "step": 6528
+    },
+    {
+      "epoch": 0.835712,
+      "grad_norm": 0.3976386487483978,
+      "learning_rate": 6.909279789653733e-05,
+      "loss": 2.8636,
+      "step": 6529
+    },
+    {
+      "epoch": 0.83584,
+      "grad_norm": 0.41866108775138855,
+      "learning_rate": 6.898768181676179e-05,
+      "loss": 2.8021,
+      "step": 6530
+    },
+    {
+      "epoch": 0.835968,
+      "grad_norm": 0.41314366459846497,
+      "learning_rate": 6.888263983300047e-05,
+      "loss": 2.7519,
+      "step": 6531
+    },
+    {
+      "epoch": 0.836096,
+      "grad_norm": 0.3954675793647766,
+      "learning_rate": 6.877767196331147e-05,
+      "loss": 2.7602,
+      "step": 6532
+    },
+    {
+      "epoch": 0.836224,
+      "grad_norm": 0.4231930673122406,
+      "learning_rate": 6.867277822573975e-05,
+      "loss": 2.7955,
+      "step": 6533
+    },
+    {
+      "epoch": 0.836352,
+      "grad_norm": 0.4536120295524597,
+      "learning_rate": 6.856795863831788e-05,
+      "loss": 2.7957,
+      "step": 6534
+    },
+    {
+      "epoch": 0.83648,
+      "grad_norm": 0.4156579375267029,
+      "learning_rate": 6.846321321906552e-05,
+      "loss": 2.7046,
+      "step": 6535
+    },
+    {
+      "epoch": 0.836608,
+      "grad_norm": 0.4437175691127777,
+      "learning_rate": 6.835854198598956e-05,
+      "loss": 2.7048,
+      "step": 6536
+    },
+    {
+      "epoch": 0.836736,
+      "grad_norm": 0.411540150642395,
+      "learning_rate": 6.825394495708415e-05,
+      "loss": 2.8324,
+      "step": 6537
+    },
+    {
+      "epoch": 0.836864,
+      "grad_norm": 0.423950731754303,
+      "learning_rate": 6.814942215033098e-05,
+      "loss": 2.8137,
+      "step": 6538
+    },
+    {
+      "epoch": 0.836992,
+      "grad_norm": 0.4087607264518738,
+      "learning_rate": 6.804497358369854e-05,
+      "loss": 2.7221,
+      "step": 6539
+    },
+    {
+      "epoch": 0.83712,
+      "grad_norm": 0.3874437212944031,
+      "learning_rate": 6.794059927514268e-05,
+      "loss": 2.7849,
+      "step": 6540
+    },
+    {
+      "epoch": 0.837248,
+      "grad_norm": 0.39564621448516846,
+      "learning_rate": 6.783629924260681e-05,
+      "loss": 2.6788,
+      "step": 6541
+    },
+    {
+      "epoch": 0.837376,
+      "grad_norm": 0.4207271337509155,
+      "learning_rate": 6.773207350402117e-05,
+      "loss": 2.8268,
+      "step": 6542
+    },
+    {
+      "epoch": 0.837504,
+      "grad_norm": 0.37767577171325684,
+      "learning_rate": 6.762792207730334e-05,
+      "loss": 2.713,
+      "step": 6543
+    },
+    {
+      "epoch": 0.837632,
+      "grad_norm": 0.41287654638290405,
+      "learning_rate": 6.752384498035824e-05,
+      "loss": 2.9474,
+      "step": 6544
+    },
+    {
+      "epoch": 0.83776,
+      "grad_norm": 0.40018975734710693,
+      "learning_rate": 6.741984223107773e-05,
+      "loss": 2.7556,
+      "step": 6545
+    },
+    {
+      "epoch": 0.837888,
+      "grad_norm": 0.3993525803089142,
+      "learning_rate": 6.731591384734137e-05,
+      "loss": 2.5989,
+      "step": 6546
+    },
+    {
+      "epoch": 0.838016,
+      "grad_norm": 0.43633896112442017,
+      "learning_rate": 6.72120598470155e-05,
+      "loss": 2.8449,
+      "step": 6547
+    },
+    {
+      "epoch": 0.838144,
+      "grad_norm": 0.3990640640258789,
+      "learning_rate": 6.710828024795391e-05,
+      "loss": 2.9155,
+      "step": 6548
+    },
+    {
+      "epoch": 0.838272,
+      "grad_norm": 0.3810189962387085,
+      "learning_rate": 6.70045750679974e-05,
+      "loss": 2.7793,
+      "step": 6549
+    },
+    {
+      "epoch": 0.8384,
+      "grad_norm": 0.40630918741226196,
+      "learning_rate": 6.690094432497407e-05,
+      "loss": 2.719,
+      "step": 6550
+    },
+    {
+      "epoch": 0.838528,
+      "grad_norm": 0.4205785393714905,
+      "learning_rate": 6.679738803669943e-05,
+      "loss": 2.7673,
+      "step": 6551
+    },
+    {
+      "epoch": 0.838656,
+      "grad_norm": 0.38597944378852844,
+      "learning_rate": 6.669390622097576e-05,
+      "loss": 2.8947,
+      "step": 6552
+    },
+    {
+      "epoch": 0.838784,
+      "grad_norm": 0.5584189295768738,
+      "learning_rate": 6.65904988955931e-05,
+      "loss": 2.7926,
+      "step": 6553
+    },
+    {
+      "epoch": 0.838912,
+      "grad_norm": 0.40467536449432373,
+      "learning_rate": 6.648716607832811e-05,
+      "loss": 2.8514,
+      "step": 6554
+    },
+    {
+      "epoch": 0.83904,
+      "grad_norm": 0.4191431403160095,
+      "learning_rate": 6.638390778694503e-05,
+      "loss": 2.8094,
+      "step": 6555
+    },
+    {
+      "epoch": 0.839168,
+      "grad_norm": 0.42346927523612976,
+      "learning_rate": 6.62807240391951e-05,
+      "loss": 2.8112,
+      "step": 6556
+    },
+    {
+      "epoch": 0.839296,
+      "grad_norm": 0.38982415199279785,
+      "learning_rate": 6.617761485281687e-05,
+      "loss": 2.6437,
+      "step": 6557
+    },
+    {
+      "epoch": 0.839424,
+      "grad_norm": 0.44957712292671204,
+      "learning_rate": 6.607458024553575e-05,
+      "loss": 2.7855,
+      "step": 6558
+    },
+    {
+      "epoch": 0.839552,
+      "grad_norm": 0.43692547082901,
+      "learning_rate": 6.597162023506492e-05,
+      "loss": 2.8561,
+      "step": 6559
+    },
+    {
+      "epoch": 0.83968,
+      "grad_norm": 0.3944839835166931,
+      "learning_rate": 6.586873483910416e-05,
+      "loss": 2.7049,
+      "step": 6560
+    },
+    {
+      "epoch": 0.839808,
+      "grad_norm": 0.40138429403305054,
+      "learning_rate": 6.576592407534088e-05,
+      "loss": 2.8478,
+      "step": 6561
+    },
+    {
+      "epoch": 0.839936,
+      "grad_norm": 0.40560606122016907,
+      "learning_rate": 6.566318796144933e-05,
+      "loss": 2.9931,
+      "step": 6562
+    },
+    {
+      "epoch": 0.840064,
+      "grad_norm": 0.4194161295890808,
+      "learning_rate": 6.556052651509104e-05,
+      "loss": 2.765,
+      "step": 6563
+    },
+    {
+      "epoch": 0.840192,
+      "grad_norm": 0.4358869791030884,
+      "learning_rate": 6.545793975391468e-05,
+      "loss": 2.8788,
+      "step": 6564
+    },
+    {
+      "epoch": 0.84032,
+      "grad_norm": 0.42047545313835144,
+      "learning_rate": 6.535542769555608e-05,
+      "loss": 2.9707,
+      "step": 6565
+    },
+    {
+      "epoch": 0.840448,
+      "grad_norm": 0.4216662049293518,
+      "learning_rate": 6.525299035763827e-05,
+      "loss": 2.7113,
+      "step": 6566
+    },
+    {
+      "epoch": 0.840576,
+      "grad_norm": 0.41034018993377686,
+      "learning_rate": 6.515062775777148e-05,
+      "loss": 2.8632,
+      "step": 6567
+    },
+    {
+      "epoch": 0.840704,
+      "grad_norm": 0.42741817235946655,
+      "learning_rate": 6.504833991355292e-05,
+      "loss": 2.8079,
+      "step": 6568
+    },
+    {
+      "epoch": 0.840832,
+      "grad_norm": 0.39183691143989563,
+      "learning_rate": 6.494612684256718e-05,
+      "loss": 2.7286,
+      "step": 6569
+    },
+    {
+      "epoch": 0.84096,
+      "grad_norm": 0.3994176685810089,
+      "learning_rate": 6.484398856238583e-05,
+      "loss": 2.9177,
+      "step": 6570
+    },
+    {
+      "epoch": 0.841088,
+      "grad_norm": 0.38768091797828674,
+      "learning_rate": 6.47419250905676e-05,
+      "loss": 2.7236,
+      "step": 6571
+    },
+    {
+      "epoch": 0.841216,
+      "grad_norm": 0.4285302758216858,
+      "learning_rate": 6.463993644465843e-05,
+      "loss": 2.7272,
+      "step": 6572
+    },
+    {
+      "epoch": 0.841344,
+      "grad_norm": 0.4281158447265625,
+      "learning_rate": 6.453802264219117e-05,
+      "loss": 2.7817,
+      "step": 6573
+    },
+    {
+      "epoch": 0.841472,
+      "grad_norm": 0.41462844610214233,
+      "learning_rate": 6.443618370068622e-05,
+      "loss": 2.8356,
+      "step": 6574
+    },
+    {
+      "epoch": 0.8416,
+      "grad_norm": 0.4131124019622803,
+      "learning_rate": 6.43344196376508e-05,
+      "loss": 2.8356,
+      "step": 6575
+    },
+    {
+      "epoch": 0.841728,
+      "grad_norm": 0.43159133195877075,
+      "learning_rate": 6.423273047057932e-05,
+      "loss": 2.7029,
+      "step": 6576
+    },
+    {
+      "epoch": 0.841856,
+      "grad_norm": 0.434068888425827,
+      "learning_rate": 6.413111621695323e-05,
+      "loss": 2.9109,
+      "step": 6577
+    },
+    {
+      "epoch": 0.841984,
+      "grad_norm": 0.42957985401153564,
+      "learning_rate": 6.402957689424138e-05,
+      "loss": 2.864,
+      "step": 6578
+    },
+    {
+      "epoch": 0.842112,
+      "grad_norm": 0.41976186633110046,
+      "learning_rate": 6.392811251989945e-05,
+      "loss": 2.7479,
+      "step": 6579
+    },
+    {
+      "epoch": 0.84224,
+      "grad_norm": 0.38536006212234497,
+      "learning_rate": 6.382672311137039e-05,
+      "loss": 2.7754,
+      "step": 6580
+    },
+    {
+      "epoch": 0.842368,
+      "grad_norm": 0.4470718204975128,
+      "learning_rate": 6.3725408686084e-05,
+      "loss": 2.7833,
+      "step": 6581
+    },
+    {
+      "epoch": 0.842496,
+      "grad_norm": 0.38824158906936646,
+      "learning_rate": 6.362416926145775e-05,
+      "loss": 2.8435,
+      "step": 6582
+    },
+    {
+      "epoch": 0.842624,
+      "grad_norm": 0.40862274169921875,
+      "learning_rate": 6.352300485489571e-05,
+      "loss": 2.9191,
+      "step": 6583
+    },
+    {
+      "epoch": 0.842752,
+      "grad_norm": 0.4084741473197937,
+      "learning_rate": 6.342191548378923e-05,
+      "loss": 2.7641,
+      "step": 6584
+    },
+    {
+      "epoch": 0.84288,
+      "grad_norm": 0.41008955240249634,
+      "learning_rate": 6.33209011655167e-05,
+      "loss": 2.7436,
+      "step": 6585
+    },
+    {
+      "epoch": 0.843008,
+      "grad_norm": 0.4489155411720276,
+      "learning_rate": 6.321996191744367e-05,
+      "loss": 2.8394,
+      "step": 6586
+    },
+    {
+      "epoch": 0.843136,
+      "grad_norm": 0.4014451801776886,
+      "learning_rate": 6.311909775692265e-05,
+      "loss": 2.6312,
+      "step": 6587
+    },
+    {
+      "epoch": 0.843264,
+      "grad_norm": 0.42964059114456177,
+      "learning_rate": 6.301830870129349e-05,
+      "loss": 2.8473,
+      "step": 6588
+    },
+    {
+      "epoch": 0.843392,
+      "grad_norm": 0.4214252829551697,
+      "learning_rate": 6.291759476788312e-05,
+      "loss": 2.8411,
+      "step": 6589
+    },
+    {
+      "epoch": 0.84352,
+      "grad_norm": 0.39669692516326904,
+      "learning_rate": 6.281695597400534e-05,
+      "loss": 2.6579,
+      "step": 6590
+    },
+    {
+      "epoch": 0.843648,
+      "grad_norm": 0.42938077449798584,
+      "learning_rate": 6.271639233696103e-05,
+      "loss": 2.7784,
+      "step": 6591
+    },
+    {
+      "epoch": 0.843776,
+      "grad_norm": 0.4156450629234314,
+      "learning_rate": 6.261590387403832e-05,
+      "loss": 2.7712,
+      "step": 6592
+    },
+    {
+      "epoch": 0.843904,
+      "grad_norm": 0.40384629368782043,
+      "learning_rate": 6.251549060251233e-05,
+      "loss": 2.8435,
+      "step": 6593
+    },
+    {
+      "epoch": 0.844032,
+      "grad_norm": 0.4309143126010895,
+      "learning_rate": 6.241515253964513e-05,
+      "loss": 2.8357,
+      "step": 6594
+    },
+    {
+      "epoch": 0.84416,
+      "grad_norm": 0.38664475083351135,
+      "learning_rate": 6.231488970268628e-05,
+      "loss": 2.6865,
+      "step": 6595
+    },
+    {
+      "epoch": 0.844288,
+      "grad_norm": 0.41510453820228577,
+      "learning_rate": 6.221470210887181e-05,
+      "loss": 2.8811,
+      "step": 6596
+    },
+    {
+      "epoch": 0.844416,
+      "grad_norm": 0.4016554653644562,
+      "learning_rate": 6.211458977542545e-05,
+      "loss": 2.9723,
+      "step": 6597
+    },
+    {
+      "epoch": 0.844544,
+      "grad_norm": 0.4282630681991577,
+      "learning_rate": 6.201455271955747e-05,
+      "loss": 2.9291,
+      "step": 6598
+    },
+    {
+      "epoch": 0.844672,
+      "grad_norm": 0.40689772367477417,
+      "learning_rate": 6.191459095846546e-05,
+      "loss": 2.7936,
+      "step": 6599
+    },
+    {
+      "epoch": 0.8448,
+      "grad_norm": 0.3734591603279114,
+      "learning_rate": 6.181470450933397e-05,
+      "loss": 2.8707,
+      "step": 6600
+    },
+    {
+      "epoch": 0.844928,
+      "grad_norm": 0.40506213903427124,
+      "learning_rate": 6.171489338933466e-05,
+      "loss": 2.6826,
+      "step": 6601
+    },
+    {
+      "epoch": 0.845056,
+      "grad_norm": 0.41944772005081177,
+      "learning_rate": 6.161515761562614e-05,
+      "loss": 2.927,
+      "step": 6602
+    },
+    {
+      "epoch": 0.845184,
+      "grad_norm": 0.3662762939929962,
+      "learning_rate": 6.151549720535432e-05,
+      "loss": 2.7417,
+      "step": 6603
+    },
+    {
+      "epoch": 0.845312,
+      "grad_norm": 0.4477834701538086,
+      "learning_rate": 6.141591217565185e-05,
+      "loss": 2.8423,
+      "step": 6604
+    },
+    {
+      "epoch": 0.84544,
+      "grad_norm": 0.4021969735622406,
+      "learning_rate": 6.131640254363846e-05,
+      "loss": 2.7282,
+      "step": 6605
+    },
+    {
+      "epoch": 0.845568,
+      "grad_norm": 0.4480055570602417,
+      "learning_rate": 6.121696832642126e-05,
+      "loss": 2.8311,
+      "step": 6606
+    },
+    {
+      "epoch": 0.845696,
+      "grad_norm": 0.4405057430267334,
+      "learning_rate": 6.111760954109402e-05,
+      "loss": 2.7205,
+      "step": 6607
+    },
+    {
+      "epoch": 0.845824,
+      "grad_norm": 0.4092563986778259,
+      "learning_rate": 6.101832620473763e-05,
+      "loss": 2.9055,
+      "step": 6608
+    },
+    {
+      "epoch": 0.845952,
+      "grad_norm": 0.4440437853336334,
+      "learning_rate": 6.0919118334419985e-05,
+      "loss": 2.921,
+      "step": 6609
+    },
+    {
+      "epoch": 0.84608,
+      "grad_norm": 0.40576520562171936,
+      "learning_rate": 6.081998594719629e-05,
+      "loss": 2.7884,
+      "step": 6610
+    },
+    {
+      "epoch": 0.846208,
+      "grad_norm": 0.40642642974853516,
+      "learning_rate": 6.07209290601084e-05,
+      "loss": 2.811,
+      "step": 6611
+    },
+    {
+      "epoch": 0.846336,
+      "grad_norm": 0.4652227759361267,
+      "learning_rate": 6.062194769018542e-05,
+      "loss": 2.9441,
+      "step": 6612
+    },
+    {
+      "epoch": 0.846464,
+      "grad_norm": 0.44354116916656494,
+      "learning_rate": 6.0523041854443295e-05,
+      "loss": 2.8898,
+      "step": 6613
+    },
+    {
+      "epoch": 0.846592,
+      "grad_norm": 0.3752554953098297,
+      "learning_rate": 6.042421156988498e-05,
+      "loss": 2.8775,
+      "step": 6614
+    },
+    {
+      "epoch": 0.84672,
+      "grad_norm": 0.45055776834487915,
+      "learning_rate": 6.032545685350088e-05,
+      "loss": 2.7842,
+      "step": 6615
+    },
+    {
+      "epoch": 0.846848,
+      "grad_norm": 0.44842129945755005,
+      "learning_rate": 6.0226777722267814e-05,
+      "loss": 2.9181,
+      "step": 6616
+    },
+    {
+      "epoch": 0.846976,
+      "grad_norm": 0.4071444869041443,
+      "learning_rate": 6.012817419314992e-05,
+      "loss": 2.6538,
+      "step": 6617
+    },
+    {
+      "epoch": 0.847104,
+      "grad_norm": 0.41164636611938477,
+      "learning_rate": 6.002964628309837e-05,
+      "loss": 2.7331,
+      "step": 6618
+    },
+    {
+      "epoch": 0.847232,
+      "grad_norm": 0.3925073742866516,
+      "learning_rate": 5.993119400905123e-05,
+      "loss": 2.7387,
+      "step": 6619
+    },
+    {
+      "epoch": 0.84736,
+      "grad_norm": 0.39554494619369507,
+      "learning_rate": 5.98328173879335e-05,
+      "loss": 2.8788,
+      "step": 6620
+    },
+    {
+      "epoch": 0.847488,
+      "grad_norm": 0.4241540729999542,
+      "learning_rate": 5.9734516436657335e-05,
+      "loss": 2.764,
+      "step": 6621
+    },
+    {
+      "epoch": 0.847616,
+      "grad_norm": 0.3848329484462738,
+      "learning_rate": 5.9636291172121824e-05,
+      "loss": 2.8828,
+      "step": 6622
+    },
+    {
+      "epoch": 0.847744,
+      "grad_norm": 0.3952842056751251,
+      "learning_rate": 5.95381416112129e-05,
+      "loss": 2.5863,
+      "step": 6623
+    },
+    {
+      "epoch": 0.847872,
+      "grad_norm": 0.4026731550693512,
+      "learning_rate": 5.944006777080363e-05,
+      "loss": 2.7991,
+      "step": 6624
+    },
+    {
+      "epoch": 0.848,
+      "grad_norm": 0.4067842662334442,
+      "learning_rate": 5.9342069667754295e-05,
+      "loss": 2.7939,
+      "step": 6625
+    },
+    {
+      "epoch": 0.848128,
+      "grad_norm": 0.39083048701286316,
+      "learning_rate": 5.92441473189117e-05,
+      "loss": 2.8641,
+      "step": 6626
+    },
+    {
+      "epoch": 0.848256,
+      "grad_norm": 0.4093102812767029,
+      "learning_rate": 5.9146300741109885e-05,
+      "loss": 2.7356,
+      "step": 6627
+    },
+    {
+      "epoch": 0.848384,
+      "grad_norm": 0.39713233709335327,
+      "learning_rate": 5.9048529951169775e-05,
+      "loss": 2.7529,
+      "step": 6628
+    },
+    {
+      "epoch": 0.848512,
+      "grad_norm": 0.38294360041618347,
+      "learning_rate": 5.8950834965899315e-05,
+      "loss": 2.7353,
+      "step": 6629
+    },
+    {
+      "epoch": 0.84864,
+      "grad_norm": 0.4432009756565094,
+      "learning_rate": 5.8853215802093244e-05,
+      "loss": 2.9324,
+      "step": 6630
+    },
+    {
+      "epoch": 0.848768,
+      "grad_norm": 0.37959012389183044,
+      "learning_rate": 5.875567247653374e-05,
+      "loss": 2.7027,
+      "step": 6631
+    },
+    {
+      "epoch": 0.848896,
+      "grad_norm": 0.4132591784000397,
+      "learning_rate": 5.8658205005989506e-05,
+      "loss": 2.8226,
+      "step": 6632
+    },
+    {
+      "epoch": 0.849024,
+      "grad_norm": 0.40788426995277405,
+      "learning_rate": 5.85608134072162e-05,
+      "loss": 2.8206,
+      "step": 6633
+    },
+    {
+      "epoch": 0.849152,
+      "grad_norm": 0.3922624886035919,
+      "learning_rate": 5.8463497696956746e-05,
+      "loss": 2.7027,
+      "step": 6634
+    },
+    {
+      "epoch": 0.84928,
+      "grad_norm": 0.39786043763160706,
+      "learning_rate": 5.836625789194078e-05,
+      "loss": 2.7403,
+      "step": 6635
+    },
+    {
+      "epoch": 0.849408,
+      "grad_norm": 0.3965446949005127,
+      "learning_rate": 5.826909400888491e-05,
+      "loss": 2.8329,
+      "step": 6636
+    },
+    {
+      "epoch": 0.849536,
+      "grad_norm": 0.4460756778717041,
+      "learning_rate": 5.81720060644928e-05,
+      "loss": 2.8596,
+      "step": 6637
+    },
+    {
+      "epoch": 0.849664,
+      "grad_norm": 0.39353257417678833,
+      "learning_rate": 5.807499407545486e-05,
+      "loss": 2.7678,
+      "step": 6638
+    },
+    {
+      "epoch": 0.849792,
+      "grad_norm": 0.3960173428058624,
+      "learning_rate": 5.797805805844875e-05,
+      "loss": 2.9202,
+      "step": 6639
+    },
+    {
+      "epoch": 0.84992,
+      "grad_norm": 0.4322003722190857,
+      "learning_rate": 5.7881198030138806e-05,
+      "loss": 2.704,
+      "step": 6640
+    },
+    {
+      "epoch": 0.850048,
+      "grad_norm": 0.40565335750579834,
+      "learning_rate": 5.7784414007176436e-05,
+      "loss": 2.6528,
+      "step": 6641
+    },
+    {
+      "epoch": 0.850176,
+      "grad_norm": 0.4208865761756897,
+      "learning_rate": 5.768770600619977e-05,
+      "loss": 2.9253,
+      "step": 6642
+    },
+    {
+      "epoch": 0.850304,
+      "grad_norm": 0.3830510973930359,
+      "learning_rate": 5.759107404383429e-05,
+      "loss": 2.7328,
+      "step": 6643
+    },
+    {
+      "epoch": 0.850432,
+      "grad_norm": 0.467758446931839,
+      "learning_rate": 5.749451813669204e-05,
+      "loss": 2.7526,
+      "step": 6644
+    },
+    {
+      "epoch": 0.85056,
+      "grad_norm": 0.4040319621562958,
+      "learning_rate": 5.7398038301371916e-05,
+      "loss": 2.7284,
+      "step": 6645
+    },
+    {
+      "epoch": 0.850688,
+      "grad_norm": 0.37753042578697205,
+      "learning_rate": 5.7301634554460246e-05,
+      "loss": 2.7522,
+      "step": 6646
+    },
+    {
+      "epoch": 0.850816,
+      "grad_norm": 0.37933897972106934,
+      "learning_rate": 5.7205306912529786e-05,
+      "loss": 2.7489,
+      "step": 6647
+    },
+    {
+      "epoch": 0.850944,
+      "grad_norm": 0.3946710228919983,
+      "learning_rate": 5.71090553921404e-05,
+      "loss": 2.8361,
+      "step": 6648
+    },
+    {
+      "epoch": 0.851072,
+      "grad_norm": 0.443515807390213,
+      "learning_rate": 5.701288000983884e-05,
+      "loss": 2.7614,
+      "step": 6649
+    },
+    {
+      "epoch": 0.8512,
+      "grad_norm": 0.4724649488925934,
+      "learning_rate": 5.691678078215873e-05,
+      "loss": 2.8522,
+      "step": 6650
+    },
+    {
+      "epoch": 0.851328,
+      "grad_norm": 0.38777095079421997,
+      "learning_rate": 5.682075772562051e-05,
+      "loss": 2.8148,
+      "step": 6651
+    },
+    {
+      "epoch": 0.851456,
+      "grad_norm": 0.40809592604637146,
+      "learning_rate": 5.672481085673198e-05,
+      "loss": 2.928,
+      "step": 6652
+    },
+    {
+      "epoch": 0.851584,
+      "grad_norm": 0.40649858117103577,
+      "learning_rate": 5.662894019198722e-05,
+      "loss": 2.7054,
+      "step": 6653
+    },
+    {
+      "epoch": 0.851712,
+      "grad_norm": 0.4034396708011627,
+      "learning_rate": 5.653314574786778e-05,
+      "loss": 2.7568,
+      "step": 6654
+    },
+    {
+      "epoch": 0.85184,
+      "grad_norm": 0.4127177596092224,
+      "learning_rate": 5.643742754084163e-05,
+      "loss": 2.9016,
+      "step": 6655
+    },
+    {
+      "epoch": 0.851968,
+      "grad_norm": 0.40980014204978943,
+      "learning_rate": 5.634178558736397e-05,
+      "loss": 2.839,
+      "step": 6656
+    },
+    {
+      "epoch": 0.852096,
+      "grad_norm": 0.41029250621795654,
+      "learning_rate": 5.624621990387668e-05,
+      "loss": 2.8038,
+      "step": 6657
+    },
+    {
+      "epoch": 0.852224,
+      "grad_norm": 0.40154600143432617,
+      "learning_rate": 5.615073050680858e-05,
+      "loss": 2.7018,
+      "step": 6658
+    },
+    {
+      "epoch": 0.852352,
+      "grad_norm": 0.4472983479499817,
+      "learning_rate": 5.6055317412575354e-05,
+      "loss": 2.9405,
+      "step": 6659
+    },
+    {
+      "epoch": 0.85248,
+      "grad_norm": 0.39646464586257935,
+      "learning_rate": 5.5959980637579844e-05,
+      "loss": 2.6415,
+      "step": 6660
+    },
+    {
+      "epoch": 0.852608,
+      "grad_norm": 0.4135877788066864,
+      "learning_rate": 5.5864720198211316e-05,
+      "loss": 2.8499,
+      "step": 6661
+    },
+    {
+      "epoch": 0.852736,
+      "grad_norm": 0.3947483003139496,
+      "learning_rate": 5.5769536110846344e-05,
+      "loss": 2.838,
+      "step": 6662
+    },
+    {
+      "epoch": 0.852864,
+      "grad_norm": 0.4264594614505768,
+      "learning_rate": 5.567442839184805e-05,
+      "loss": 2.719,
+      "step": 6663
+    },
+    {
+      "epoch": 0.852992,
+      "grad_norm": 0.4219519793987274,
+      "learning_rate": 5.5579397057566646e-05,
+      "loss": 2.7104,
+      "step": 6664
+    },
+    {
+      "epoch": 0.85312,
+      "grad_norm": 0.4125348925590515,
+      "learning_rate": 5.548444212433901e-05,
+      "loss": 2.8299,
+      "step": 6665
+    },
+    {
+      "epoch": 0.853248,
+      "grad_norm": 0.42722979187965393,
+      "learning_rate": 5.53895636084889e-05,
+      "loss": 2.8026,
+      "step": 6666
+    },
+    {
+      "epoch": 0.853376,
+      "grad_norm": 0.39301958680152893,
+      "learning_rate": 5.5294761526327345e-05,
+      "loss": 2.757,
+      "step": 6667
+    },
+    {
+      "epoch": 0.853504,
+      "grad_norm": 0.3947344720363617,
+      "learning_rate": 5.520003589415168e-05,
+      "loss": 2.6293,
+      "step": 6668
+    },
+    {
+      "epoch": 0.853632,
+      "grad_norm": 0.4279238283634186,
+      "learning_rate": 5.510538672824644e-05,
+      "loss": 2.7147,
+      "step": 6669
+    },
+    {
+      "epoch": 0.85376,
+      "grad_norm": 0.41515642404556274,
+      "learning_rate": 5.501081404488278e-05,
+      "loss": 2.7281,
+      "step": 6670
+    },
+    {
+      "epoch": 0.853888,
+      "grad_norm": 0.4357076585292816,
+      "learning_rate": 5.491631786031903e-05,
+      "loss": 2.7335,
+      "step": 6671
+    },
+    {
+      "epoch": 0.854016,
+      "grad_norm": 0.418525755405426,
+      "learning_rate": 5.482189819080013e-05,
+      "loss": 2.6762,
+      "step": 6672
+    },
+    {
+      "epoch": 0.854144,
+      "grad_norm": 0.4771140217781067,
+      "learning_rate": 5.472755505255783e-05,
+      "loss": 2.8236,
+      "step": 6673
+    },
+    {
+      "epoch": 0.854272,
+      "grad_norm": 0.41697192192077637,
+      "learning_rate": 5.463328846181082e-05,
+      "loss": 2.8563,
+      "step": 6674
+    },
+    {
+      "epoch": 0.8544,
+      "grad_norm": 0.4275633990764618,
+      "learning_rate": 5.453909843476468e-05,
+      "loss": 2.8623,
+      "step": 6675
+    },
+    {
+      "epoch": 0.854528,
+      "grad_norm": 0.39392387866973877,
+      "learning_rate": 5.44449849876118e-05,
+      "loss": 2.8004,
+      "step": 6676
+    },
+    {
+      "epoch": 0.854656,
+      "grad_norm": 0.4018356502056122,
+      "learning_rate": 5.4350948136531296e-05,
+      "loss": 2.8276,
+      "step": 6677
+    },
+    {
+      "epoch": 0.854784,
+      "grad_norm": 0.4768989682197571,
+      "learning_rate": 5.4256987897689234e-05,
+      "loss": 2.952,
+      "step": 6678
+    },
+    {
+      "epoch": 0.854912,
+      "grad_norm": 0.41779935359954834,
+      "learning_rate": 5.416310428723836e-05,
+      "loss": 2.6507,
+      "step": 6679
+    },
+    {
+      "epoch": 0.85504,
+      "grad_norm": 0.4069933593273163,
+      "learning_rate": 5.406929732131849e-05,
+      "loss": 2.8715,
+      "step": 6680
+    },
+    {
+      "epoch": 0.855168,
+      "grad_norm": 0.40739694237709045,
+      "learning_rate": 5.397556701605605e-05,
+      "loss": 2.7143,
+      "step": 6681
+    },
+    {
+      "epoch": 0.855296,
+      "grad_norm": 0.4157349169254303,
+      "learning_rate": 5.388191338756443e-05,
+      "loss": 2.8136,
+      "step": 6682
+    },
+    {
+      "epoch": 0.855424,
+      "grad_norm": 0.3940369486808777,
+      "learning_rate": 5.378833645194375e-05,
+      "loss": 2.7797,
+      "step": 6683
+    },
+    {
+      "epoch": 0.855552,
+      "grad_norm": 0.40367984771728516,
+      "learning_rate": 5.369483622528104e-05,
+      "loss": 2.7183,
+      "step": 6684
+    },
+    {
+      "epoch": 0.85568,
+      "grad_norm": 0.4290567934513092,
+      "learning_rate": 5.360141272364993e-05,
+      "loss": 2.7412,
+      "step": 6685
+    },
+    {
+      "epoch": 0.855808,
+      "grad_norm": 0.3803033232688904,
+      "learning_rate": 5.350806596311109e-05,
+      "loss": 2.7051,
+      "step": 6686
+    },
+    {
+      "epoch": 0.855936,
+      "grad_norm": 0.4197840690612793,
+      "learning_rate": 5.341479595971177e-05,
+      "loss": 2.7623,
+      "step": 6687
+    },
+    {
+      "epoch": 0.856064,
+      "grad_norm": 0.4183705747127533,
+      "learning_rate": 5.3321602729486275e-05,
+      "loss": 2.95,
+      "step": 6688
+    },
+    {
+      "epoch": 0.856192,
+      "grad_norm": 0.4130142331123352,
+      "learning_rate": 5.322848628845578e-05,
+      "loss": 2.7649,
+      "step": 6689
+    },
+    {
+      "epoch": 0.85632,
+      "grad_norm": 0.38634902238845825,
+      "learning_rate": 5.313544665262782e-05,
+      "loss": 2.892,
+      "step": 6690
+    },
+    {
+      "epoch": 0.856448,
+      "grad_norm": 0.4271065294742584,
+      "learning_rate": 5.3042483837997145e-05,
+      "loss": 2.7154,
+      "step": 6691
+    },
+    {
+      "epoch": 0.856576,
+      "grad_norm": 0.4173172414302826,
+      "learning_rate": 5.294959786054504e-05,
+      "loss": 2.7155,
+      "step": 6692
+    },
+    {
+      "epoch": 0.856704,
+      "grad_norm": 0.3992484509944916,
+      "learning_rate": 5.285678873623972e-05,
+      "loss": 2.6636,
+      "step": 6693
+    },
+    {
+      "epoch": 0.856832,
+      "grad_norm": 0.3883410394191742,
+      "learning_rate": 5.276405648103616e-05,
+      "loss": 2.8475,
+      "step": 6694
+    },
+    {
+      "epoch": 0.85696,
+      "grad_norm": 0.4200456440448761,
+      "learning_rate": 5.267140111087604e-05,
+      "loss": 2.8092,
+      "step": 6695
+    },
+    {
+      "epoch": 0.857088,
+      "grad_norm": 0.4250309467315674,
+      "learning_rate": 5.257882264168795e-05,
+      "loss": 2.7905,
+      "step": 6696
+    },
+    {
+      "epoch": 0.857216,
+      "grad_norm": 0.3978080153465271,
+      "learning_rate": 5.248632108938728e-05,
+      "loss": 2.7406,
+      "step": 6697
+    },
+    {
+      "epoch": 0.857344,
+      "grad_norm": 0.40402933955192566,
+      "learning_rate": 5.239389646987591e-05,
+      "loss": 2.7891,
+      "step": 6698
+    },
+    {
+      "epoch": 0.857472,
+      "grad_norm": 0.40753182768821716,
+      "learning_rate": 5.230154879904292e-05,
+      "loss": 2.7248,
+      "step": 6699
+    },
+    {
+      "epoch": 0.8576,
+      "grad_norm": 0.3978956937789917,
+      "learning_rate": 5.2209278092763824e-05,
+      "loss": 2.7745,
+      "step": 6700
+    },
+    {
+      "epoch": 0.857728,
+      "grad_norm": 0.4160376787185669,
+      "learning_rate": 5.21170843669011e-05,
+      "loss": 2.6945,
+      "step": 6701
+    },
+    {
+      "epoch": 0.857856,
+      "grad_norm": 0.4050270617008209,
+      "learning_rate": 5.20249676373038e-05,
+      "loss": 2.7097,
+      "step": 6702
+    },
+    {
+      "epoch": 0.857984,
+      "grad_norm": 0.40057095885276794,
+      "learning_rate": 5.1932927919808025e-05,
+      "loss": 2.8925,
+      "step": 6703
+    },
+    {
+      "epoch": 0.858112,
+      "grad_norm": 0.4531680643558502,
+      "learning_rate": 5.184096523023635e-05,
+      "loss": 2.8639,
+      "step": 6704
+    },
+    {
+      "epoch": 0.85824,
+      "grad_norm": 0.4299463927745819,
+      "learning_rate": 5.1749079584398284e-05,
+      "loss": 2.8575,
+      "step": 6705
+    },
+    {
+      "epoch": 0.858368,
+      "grad_norm": 0.4063490033149719,
+      "learning_rate": 5.165727099808998e-05,
+      "loss": 2.7288,
+      "step": 6706
+    },
+    {
+      "epoch": 0.858496,
+      "grad_norm": 0.41053998470306396,
+      "learning_rate": 5.156553948709436e-05,
+      "loss": 2.6943,
+      "step": 6707
+    },
+    {
+      "epoch": 0.858624,
+      "grad_norm": 0.4038132131099701,
+      "learning_rate": 5.147388506718126e-05,
+      "loss": 2.6678,
+      "step": 6708
+    },
+    {
+      "epoch": 0.858752,
+      "grad_norm": 0.4110935628414154,
+      "learning_rate": 5.138230775410713e-05,
+      "loss": 2.7634,
+      "step": 6709
+    },
+    {
+      "epoch": 0.85888,
+      "grad_norm": 0.40712764859199524,
+      "learning_rate": 5.1290807563614984e-05,
+      "loss": 2.8292,
+      "step": 6710
+    },
+    {
+      "epoch": 0.859008,
+      "grad_norm": 0.41197505593299866,
+      "learning_rate": 5.1199384511435023e-05,
+      "loss": 2.6984,
+      "step": 6711
+    },
+    {
+      "epoch": 0.859136,
+      "grad_norm": 0.41274333000183105,
+      "learning_rate": 5.110803861328378e-05,
+      "loss": 2.8417,
+      "step": 6712
+    },
+    {
+      "epoch": 0.859264,
+      "grad_norm": 0.41014406085014343,
+      "learning_rate": 5.10167698848647e-05,
+      "loss": 2.8109,
+      "step": 6713
+    },
+    {
+      "epoch": 0.859392,
+      "grad_norm": 0.4328247606754303,
+      "learning_rate": 5.092557834186795e-05,
+      "loss": 2.6696,
+      "step": 6714
+    },
+    {
+      "epoch": 0.85952,
+      "grad_norm": 0.48959219455718994,
+      "learning_rate": 5.083446399997044e-05,
+      "loss": 2.9866,
+      "step": 6715
+    },
+    {
+      "epoch": 0.859648,
+      "grad_norm": 0.4202902019023895,
+      "learning_rate": 5.0743426874835574e-05,
+      "loss": 2.7511,
+      "step": 6716
+    },
+    {
+      "epoch": 0.859776,
+      "grad_norm": 0.43326959013938904,
+      "learning_rate": 5.0652466982113896e-05,
+      "loss": 2.7829,
+      "step": 6717
+    },
+    {
+      "epoch": 0.859904,
+      "grad_norm": 0.4343530237674713,
+      "learning_rate": 5.0561584337442514e-05,
+      "loss": 2.7878,
+      "step": 6718
+    },
+    {
+      "epoch": 0.860032,
+      "grad_norm": 0.42366930842399597,
+      "learning_rate": 5.047077895644514e-05,
+      "loss": 2.9523,
+      "step": 6719
+    },
+    {
+      "epoch": 0.86016,
+      "grad_norm": 0.4285213351249695,
+      "learning_rate": 5.038005085473224e-05,
+      "loss": 2.6653,
+      "step": 6720
+    },
+    {
+      "epoch": 0.860288,
+      "grad_norm": 0.4114139676094055,
+      "learning_rate": 5.0289400047901055e-05,
+      "loss": 2.668,
+      "step": 6721
+    },
+    {
+      "epoch": 0.860416,
+      "grad_norm": 0.41421231627464294,
+      "learning_rate": 5.019882655153546e-05,
+      "loss": 2.8505,
+      "step": 6722
+    },
+    {
+      "epoch": 0.860544,
+      "grad_norm": 0.42042943835258484,
+      "learning_rate": 5.0108330381206044e-05,
+      "loss": 2.7414,
+      "step": 6723
+    },
+    {
+      "epoch": 0.860672,
+      "grad_norm": 0.3740580081939697,
+      "learning_rate": 5.001791155247032e-05,
+      "loss": 2.6555,
+      "step": 6724
+    },
+    {
+      "epoch": 0.8608,
+      "grad_norm": 0.405507355928421,
+      "learning_rate": 4.9927570080872176e-05,
+      "loss": 2.9211,
+      "step": 6725
+    },
+    {
+      "epoch": 0.860928,
+      "grad_norm": 0.3762024939060211,
+      "learning_rate": 4.983730598194247e-05,
+      "loss": 2.6803,
+      "step": 6726
+    },
+    {
+      "epoch": 0.861056,
+      "grad_norm": 0.38935616612434387,
+      "learning_rate": 4.974711927119863e-05,
+      "loss": 2.8033,
+      "step": 6727
+    },
+    {
+      "epoch": 0.861184,
+      "grad_norm": 0.40992265939712524,
+      "learning_rate": 4.9657009964144794e-05,
+      "loss": 2.7208,
+      "step": 6728
+    },
+    {
+      "epoch": 0.861312,
+      "grad_norm": 0.4049694538116455,
+      "learning_rate": 4.956697807627169e-05,
+      "loss": 2.8,
+      "step": 6729
+    },
+    {
+      "epoch": 0.86144,
+      "grad_norm": 0.38727641105651855,
+      "learning_rate": 4.9477023623057e-05,
+      "loss": 2.8189,
+      "step": 6730
+    },
+    {
+      "epoch": 0.861568,
+      "grad_norm": 0.42663678526878357,
+      "learning_rate": 4.9387146619964684e-05,
+      "loss": 2.8905,
+      "step": 6731
+    },
+    {
+      "epoch": 0.861696,
+      "grad_norm": 0.4253164529800415,
+      "learning_rate": 4.929734708244599e-05,
+      "loss": 2.8834,
+      "step": 6732
+    },
+    {
+      "epoch": 0.861824,
+      "grad_norm": 0.4141189455986023,
+      "learning_rate": 4.9207625025938285e-05,
+      "loss": 2.8573,
+      "step": 6733
+    },
+    {
+      "epoch": 0.861952,
+      "grad_norm": 0.42754098773002625,
+      "learning_rate": 4.911798046586591e-05,
+      "loss": 2.7763,
+      "step": 6734
+    },
+    {
+      "epoch": 0.86208,
+      "grad_norm": 0.41063645482063293,
+      "learning_rate": 4.902841341763964e-05,
+      "loss": 2.8423,
+      "step": 6735
+    },
+    {
+      "epoch": 0.862208,
+      "grad_norm": 0.40759560465812683,
+      "learning_rate": 4.893892389665727e-05,
+      "loss": 2.719,
+      "step": 6736
+    },
+    {
+      "epoch": 0.862336,
+      "grad_norm": 0.422067254781723,
+      "learning_rate": 4.8849511918303117e-05,
+      "loss": 2.7316,
+      "step": 6737
+    },
+    {
+      "epoch": 0.862464,
+      "grad_norm": 0.40484774112701416,
+      "learning_rate": 4.8760177497947866e-05,
+      "loss": 2.814,
+      "step": 6738
+    },
+    {
+      "epoch": 0.862592,
+      "grad_norm": 0.41862624883651733,
+      "learning_rate": 4.867092065094947e-05,
+      "loss": 2.8281,
+      "step": 6739
+    },
+    {
+      "epoch": 0.86272,
+      "grad_norm": 0.40887650847435,
+      "learning_rate": 4.858174139265209e-05,
+      "loss": 2.9576,
+      "step": 6740
+    },
+    {
+      "epoch": 0.862848,
+      "grad_norm": 0.4311162531375885,
+      "learning_rate": 4.849263973838664e-05,
+      "loss": 3.0433,
+      "step": 6741
+    },
+    {
+      "epoch": 0.862976,
+      "grad_norm": 0.44046536087989807,
+      "learning_rate": 4.840361570347068e-05,
+      "loss": 2.8842,
+      "step": 6742
+    },
+    {
+      "epoch": 0.863104,
+      "grad_norm": 0.4116315245628357,
+      "learning_rate": 4.8314669303208605e-05,
+      "loss": 2.7848,
+      "step": 6743
+    },
+    {
+      "epoch": 0.863232,
+      "grad_norm": 0.3963354527950287,
+      "learning_rate": 4.822580055289116e-05,
+      "loss": 2.7946,
+      "step": 6744
+    },
+    {
+      "epoch": 0.86336,
+      "grad_norm": 0.3921448588371277,
+      "learning_rate": 4.8137009467795976e-05,
+      "loss": 2.7641,
+      "step": 6745
+    },
+    {
+      "epoch": 0.863488,
+      "grad_norm": 0.4105132818222046,
+      "learning_rate": 4.8048296063187446e-05,
+      "loss": 2.703,
+      "step": 6746
+    },
+    {
+      "epoch": 0.863616,
+      "grad_norm": 0.41950738430023193,
+      "learning_rate": 4.795966035431626e-05,
+      "loss": 2.7887,
+      "step": 6747
+    },
+    {
+      "epoch": 0.863744,
+      "grad_norm": 0.38214248418807983,
+      "learning_rate": 4.787110235642001e-05,
+      "loss": 2.6815,
+      "step": 6748
+    },
+    {
+      "epoch": 0.863872,
+      "grad_norm": 0.39410412311553955,
+      "learning_rate": 4.778262208472273e-05,
+      "loss": 2.8068,
+      "step": 6749
+    },
+    {
+      "epoch": 0.864,
+      "grad_norm": 0.3965597152709961,
+      "learning_rate": 4.7694219554435316e-05,
+      "loss": 2.8751,
+      "step": 6750
+    },
+    {
+      "epoch": 0.864128,
+      "grad_norm": 0.3634563088417053,
+      "learning_rate": 4.760589478075511e-05,
+      "loss": 2.6281,
+      "step": 6751
+    },
+    {
+      "epoch": 0.864256,
+      "grad_norm": 0.41361531615257263,
+      "learning_rate": 4.751764777886614e-05,
+      "loss": 2.7105,
+      "step": 6752
+    },
+    {
+      "epoch": 0.864384,
+      "grad_norm": 0.39578092098236084,
+      "learning_rate": 4.742947856393909e-05,
+      "loss": 2.8153,
+      "step": 6753
+    },
+    {
+      "epoch": 0.864512,
+      "grad_norm": 0.38851773738861084,
+      "learning_rate": 4.734138715113145e-05,
+      "loss": 2.8048,
+      "step": 6754
+    },
+    {
+      "epoch": 0.86464,
+      "grad_norm": 0.42871639132499695,
+      "learning_rate": 4.7253373555587006e-05,
+      "loss": 2.7501,
+      "step": 6755
+    },
+    {
+      "epoch": 0.864768,
+      "grad_norm": 0.41018417477607727,
+      "learning_rate": 4.716543779243637e-05,
+      "loss": 2.7801,
+      "step": 6756
+    },
+    {
+      "epoch": 0.864896,
+      "grad_norm": 0.42612382769584656,
+      "learning_rate": 4.707757987679667e-05,
+      "loss": 2.9453,
+      "step": 6757
+    },
+    {
+      "epoch": 0.865024,
+      "grad_norm": 0.4636908769607544,
+      "learning_rate": 4.6989799823771714e-05,
+      "loss": 2.7519,
+      "step": 6758
+    },
+    {
+      "epoch": 0.865152,
+      "grad_norm": 0.4322591722011566,
+      "learning_rate": 4.6902097648451824e-05,
+      "loss": 2.7324,
+      "step": 6759
+    },
+    {
+      "epoch": 0.86528,
+      "grad_norm": 0.400112122297287,
+      "learning_rate": 4.681447336591421e-05,
+      "loss": 2.6816,
+      "step": 6760
+    },
+    {
+      "epoch": 0.865408,
+      "grad_norm": 0.383432537317276,
+      "learning_rate": 4.672692699122238e-05,
+      "loss": 2.6512,
+      "step": 6761
+    },
+    {
+      "epoch": 0.865536,
+      "grad_norm": 0.41137996315956116,
+      "learning_rate": 4.663945853942653e-05,
+      "loss": 2.65,
+      "step": 6762
+    },
+    {
+      "epoch": 0.865664,
+      "grad_norm": 0.41614627838134766,
+      "learning_rate": 4.655206802556361e-05,
+      "loss": 2.7382,
+      "step": 6763
+    },
+    {
+      "epoch": 0.865792,
+      "grad_norm": 0.4052010178565979,
+      "learning_rate": 4.646475546465706e-05,
+      "loss": 2.7893,
+      "step": 6764
+    },
+    {
+      "epoch": 0.86592,
+      "grad_norm": 0.40022939443588257,
+      "learning_rate": 4.637752087171687e-05,
+      "loss": 2.9726,
+      "step": 6765
+    },
+    {
+      "epoch": 0.866048,
+      "grad_norm": 0.39217624068260193,
+      "learning_rate": 4.6290364261739545e-05,
+      "loss": 2.8931,
+      "step": 6766
+    },
+    {
+      "epoch": 0.866176,
+      "grad_norm": 0.38439640402793884,
+      "learning_rate": 4.6203285649708603e-05,
+      "loss": 2.8785,
+      "step": 6767
+    },
+    {
+      "epoch": 0.866304,
+      "grad_norm": 0.3991905748844147,
+      "learning_rate": 4.611628505059373e-05,
+      "loss": 2.6616,
+      "step": 6768
+    },
+    {
+      "epoch": 0.866432,
+      "grad_norm": 0.3932170867919922,
+      "learning_rate": 4.60293624793513e-05,
+      "loss": 2.7974,
+      "step": 6769
+    },
+    {
+      "epoch": 0.86656,
+      "grad_norm": 0.3880203664302826,
+      "learning_rate": 4.59425179509243e-05,
+      "loss": 2.7653,
+      "step": 6770
+    },
+    {
+      "epoch": 0.866688,
+      "grad_norm": 0.40640613436698914,
+      "learning_rate": 4.58557514802424e-05,
+      "loss": 2.7301,
+      "step": 6771
+    },
+    {
+      "epoch": 0.866816,
+      "grad_norm": 0.40250253677368164,
+      "learning_rate": 4.576906308222162e-05,
+      "loss": 2.8099,
+      "step": 6772
+    },
+    {
+      "epoch": 0.866944,
+      "grad_norm": 0.417685866355896,
+      "learning_rate": 4.568245277176486e-05,
+      "loss": 3.0114,
+      "step": 6773
+    },
+    {
+      "epoch": 0.867072,
+      "grad_norm": 0.3843649625778198,
+      "learning_rate": 4.559592056376133e-05,
+      "loss": 2.8101,
+      "step": 6774
+    },
+    {
+      "epoch": 0.8672,
+      "grad_norm": 0.40598300099372864,
+      "learning_rate": 4.550946647308701e-05,
+      "loss": 2.8125,
+      "step": 6775
+    },
+    {
+      "epoch": 0.867328,
+      "grad_norm": 0.3954850733280182,
+      "learning_rate": 4.5423090514604304e-05,
+      "loss": 2.7099,
+      "step": 6776
+    },
+    {
+      "epoch": 0.867456,
+      "grad_norm": 0.41207921504974365,
+      "learning_rate": 4.533679270316227e-05,
+      "loss": 2.814,
+      "step": 6777
+    },
+    {
+      "epoch": 0.867584,
+      "grad_norm": 0.3848171830177307,
+      "learning_rate": 4.525057305359648e-05,
+      "loss": 2.7953,
+      "step": 6778
+    },
+    {
+      "epoch": 0.867712,
+      "grad_norm": 0.3875891864299774,
+      "learning_rate": 4.5164431580729084e-05,
+      "loss": 2.9646,
+      "step": 6779
+    },
+    {
+      "epoch": 0.86784,
+      "grad_norm": 0.40377140045166016,
+      "learning_rate": 4.5078368299368675e-05,
+      "loss": 2.7875,
+      "step": 6780
+    },
+    {
+      "epoch": 0.867968,
+      "grad_norm": 0.4000360071659088,
+      "learning_rate": 4.499238322431071e-05,
+      "loss": 2.8131,
+      "step": 6781
+    },
+    {
+      "epoch": 0.868096,
+      "grad_norm": 0.40190690755844116,
+      "learning_rate": 4.490647637033701e-05,
+      "loss": 2.8544,
+      "step": 6782
+    },
+    {
+      "epoch": 0.868224,
+      "grad_norm": 0.39437195658683777,
+      "learning_rate": 4.482064775221595e-05,
+      "loss": 2.8182,
+      "step": 6783
+    },
+    {
+      "epoch": 0.868352,
+      "grad_norm": 0.3711157441139221,
+      "learning_rate": 4.473489738470243e-05,
+      "loss": 2.79,
+      "step": 6784
+    },
+    {
+      "epoch": 0.86848,
+      "grad_norm": 0.4000413119792938,
+      "learning_rate": 4.464922528253795e-05,
+      "loss": 2.7412,
+      "step": 6785
+    },
+    {
+      "epoch": 0.868608,
+      "grad_norm": 0.3944229781627655,
+      "learning_rate": 4.4563631460450505e-05,
+      "loss": 2.897,
+      "step": 6786
+    },
+    {
+      "epoch": 0.868736,
+      "grad_norm": 0.4219585657119751,
+      "learning_rate": 4.447811593315459e-05,
+      "loss": 2.7613,
+      "step": 6787
+    },
+    {
+      "epoch": 0.868864,
+      "grad_norm": 0.43560704588890076,
+      "learning_rate": 4.439267871535152e-05,
+      "loss": 2.8286,
+      "step": 6788
+    },
+    {
+      "epoch": 0.868992,
+      "grad_norm": 0.3909452557563782,
+      "learning_rate": 4.430731982172886e-05,
+      "loss": 2.7691,
+      "step": 6789
+    },
+    {
+      "epoch": 0.86912,
+      "grad_norm": 0.41195759177207947,
+      "learning_rate": 4.422203926696061e-05,
+      "loss": 2.8067,
+      "step": 6790
+    },
+    {
+      "epoch": 0.869248,
+      "grad_norm": 0.3875870108604431,
+      "learning_rate": 4.4136837065707756e-05,
+      "loss": 2.7543,
+      "step": 6791
+    },
+    {
+      "epoch": 0.869376,
+      "grad_norm": 0.40576356649398804,
+      "learning_rate": 4.405171323261742e-05,
+      "loss": 2.6448,
+      "step": 6792
+    },
+    {
+      "epoch": 0.869504,
+      "grad_norm": 0.4288060665130615,
+      "learning_rate": 4.396666778232344e-05,
+      "loss": 2.7592,
+      "step": 6793
+    },
+    {
+      "epoch": 0.869632,
+      "grad_norm": 0.4187385141849518,
+      "learning_rate": 4.388170072944603e-05,
+      "loss": 2.7913,
+      "step": 6794
+    },
+    {
+      "epoch": 0.86976,
+      "grad_norm": 0.4398978352546692,
+      "learning_rate": 4.379681208859193e-05,
+      "loss": 2.767,
+      "step": 6795
+    },
+    {
+      "epoch": 0.869888,
+      "grad_norm": 0.3883199393749237,
+      "learning_rate": 4.37120018743547e-05,
+      "loss": 2.7947,
+      "step": 6796
+    },
+    {
+      "epoch": 0.870016,
+      "grad_norm": 0.415569543838501,
+      "learning_rate": 4.3627270101314064e-05,
+      "loss": 2.6891,
+      "step": 6797
+    },
+    {
+      "epoch": 0.870144,
+      "grad_norm": 0.40508151054382324,
+      "learning_rate": 4.3542616784036426e-05,
+      "loss": 2.7823,
+      "step": 6798
+    },
+    {
+      "epoch": 0.870272,
+      "grad_norm": 0.43731629848480225,
+      "learning_rate": 4.345804193707459e-05,
+      "loss": 2.8161,
+      "step": 6799
+    },
+    {
+      "epoch": 0.8704,
+      "grad_norm": 0.37825289368629456,
+      "learning_rate": 4.3373545574968144e-05,
+      "loss": 2.7693,
+      "step": 6800
+    },
+    {
+      "epoch": 0.870528,
+      "grad_norm": 0.3945426046848297,
+      "learning_rate": 4.328912771224286e-05,
+      "loss": 2.7775,
+      "step": 6801
+    },
+    {
+      "epoch": 0.870656,
+      "grad_norm": 0.39715880155563354,
+      "learning_rate": 4.320478836341113e-05,
+      "loss": 2.8454,
+      "step": 6802
+    },
+    {
+      "epoch": 0.870784,
+      "grad_norm": 0.40356338024139404,
+      "learning_rate": 4.312052754297197e-05,
+      "loss": 2.8602,
+      "step": 6803
+    },
+    {
+      "epoch": 0.870912,
+      "grad_norm": 0.456898033618927,
+      "learning_rate": 4.303634526541067e-05,
+      "loss": 2.8201,
+      "step": 6804
+    },
+    {
+      "epoch": 0.87104,
+      "grad_norm": 0.4042287766933441,
+      "learning_rate": 4.2952241545199225e-05,
+      "loss": 2.791,
+      "step": 6805
+    },
+    {
+      "epoch": 0.871168,
+      "grad_norm": 0.40027883648872375,
+      "learning_rate": 4.2868216396796055e-05,
+      "loss": 2.8302,
+      "step": 6806
+    },
+    {
+      "epoch": 0.871296,
+      "grad_norm": 0.4164825677871704,
+      "learning_rate": 4.278426983464595e-05,
+      "loss": 2.7443,
+      "step": 6807
+    },
+    {
+      "epoch": 0.871424,
+      "grad_norm": 0.43139946460723877,
+      "learning_rate": 4.2700401873180303e-05,
+      "loss": 2.7581,
+      "step": 6808
+    },
+    {
+      "epoch": 0.871552,
+      "grad_norm": 0.3829682767391205,
+      "learning_rate": 4.261661252681714e-05,
+      "loss": 2.807,
+      "step": 6809
+    },
+    {
+      "epoch": 0.87168,
+      "grad_norm": 0.3945547044277191,
+      "learning_rate": 4.253290180996067e-05,
+      "loss": 2.8183,
+      "step": 6810
+    },
+    {
+      "epoch": 0.871808,
+      "grad_norm": 0.40661776065826416,
+      "learning_rate": 4.2449269737001804e-05,
+      "loss": 2.8264,
+      "step": 6811
+    },
+    {
+      "epoch": 0.871936,
+      "grad_norm": 0.4471508264541626,
+      "learning_rate": 4.2365716322317894e-05,
+      "loss": 2.8529,
+      "step": 6812
+    },
+    {
+      "epoch": 0.872064,
+      "grad_norm": 0.4158177971839905,
+      "learning_rate": 4.228224158027272e-05,
+      "loss": 2.8457,
+      "step": 6813
+    },
+    {
+      "epoch": 0.872192,
+      "grad_norm": 0.40680423378944397,
+      "learning_rate": 4.2198845525216524e-05,
+      "loss": 2.7887,
+      "step": 6814
+    },
+    {
+      "epoch": 0.87232,
+      "grad_norm": 0.419357568025589,
+      "learning_rate": 4.211552817148606e-05,
+      "loss": 2.7451,
+      "step": 6815
+    },
+    {
+      "epoch": 0.872448,
+      "grad_norm": 0.416911780834198,
+      "learning_rate": 4.2032289533404434e-05,
+      "loss": 2.7679,
+      "step": 6816
+    },
+    {
+      "epoch": 0.872576,
+      "grad_norm": 0.41888153553009033,
+      "learning_rate": 4.194912962528158e-05,
+      "loss": 2.8201,
+      "step": 6817
+    },
+    {
+      "epoch": 0.872704,
+      "grad_norm": 0.39432671666145325,
+      "learning_rate": 4.186604846141345e-05,
+      "loss": 2.8259,
+      "step": 6818
+    },
+    {
+      "epoch": 0.872832,
+      "grad_norm": 0.40967047214508057,
+      "learning_rate": 4.17830460560828e-05,
+      "loss": 2.85,
+      "step": 6819
+    },
+    {
+      "epoch": 0.87296,
+      "grad_norm": 0.4248571991920471,
+      "learning_rate": 4.170012242355864e-05,
+      "loss": 2.7858,
+      "step": 6820
+    },
+    {
+      "epoch": 0.873088,
+      "grad_norm": 0.39520081877708435,
+      "learning_rate": 4.1617277578096526e-05,
+      "loss": 2.6958,
+      "step": 6821
+    },
+    {
+      "epoch": 0.873216,
+      "grad_norm": 0.42357221245765686,
+      "learning_rate": 4.153451153393839e-05,
+      "loss": 2.8739,
+      "step": 6822
+    },
+    {
+      "epoch": 0.873344,
+      "grad_norm": 0.39392492175102234,
+      "learning_rate": 4.1451824305312634e-05,
+      "loss": 2.7796,
+      "step": 6823
+    },
+    {
+      "epoch": 0.873472,
+      "grad_norm": 0.4194720387458801,
+      "learning_rate": 4.136921590643433e-05,
+      "loss": 2.9772,
+      "step": 6824
+    },
+    {
+      "epoch": 0.8736,
+      "grad_norm": 0.4259334206581116,
+      "learning_rate": 4.1286686351504735e-05,
+      "loss": 2.796,
+      "step": 6825
+    },
+    {
+      "epoch": 0.873728,
+      "grad_norm": 0.4316185414791107,
+      "learning_rate": 4.120423565471154e-05,
+      "loss": 2.8052,
+      "step": 6826
+    },
+    {
+      "epoch": 0.873856,
+      "grad_norm": 0.4099735915660858,
+      "learning_rate": 4.1121863830229036e-05,
+      "loss": 2.9852,
+      "step": 6827
+    },
+    {
+      "epoch": 0.873984,
+      "grad_norm": 0.40905454754829407,
+      "learning_rate": 4.103957089221799e-05,
+      "loss": 2.8292,
+      "step": 6828
+    },
+    {
+      "epoch": 0.874112,
+      "grad_norm": 0.3815423846244812,
+      "learning_rate": 4.0957356854825435e-05,
+      "loss": 2.8256,
+      "step": 6829
+    },
+    {
+      "epoch": 0.87424,
+      "grad_norm": 0.4104766249656677,
+      "learning_rate": 4.0875221732184884e-05,
+      "loss": 2.8182,
+      "step": 6830
+    },
+    {
+      "epoch": 0.874368,
+      "grad_norm": 0.4152514636516571,
+      "learning_rate": 4.0793165538416275e-05,
+      "loss": 2.8252,
+      "step": 6831
+    },
+    {
+      "epoch": 0.874496,
+      "grad_norm": 0.4230552613735199,
+      "learning_rate": 4.0711188287626156e-05,
+      "loss": 2.7113,
+      "step": 6832
+    },
+    {
+      "epoch": 0.874624,
+      "grad_norm": 0.4120693802833557,
+      "learning_rate": 4.062928999390736e-05,
+      "loss": 2.6987,
+      "step": 6833
+    },
+    {
+      "epoch": 0.874752,
+      "grad_norm": 0.39840877056121826,
+      "learning_rate": 4.054747067133907e-05,
+      "loss": 2.9111,
+      "step": 6834
+    },
+    {
+      "epoch": 0.87488,
+      "grad_norm": 0.4293572008609772,
+      "learning_rate": 4.0465730333986985e-05,
+      "loss": 2.7309,
+      "step": 6835
+    },
+    {
+      "epoch": 0.875008,
+      "grad_norm": 0.48787957429885864,
+      "learning_rate": 4.038406899590313e-05,
+      "loss": 2.792,
+      "step": 6836
+    },
+    {
+      "epoch": 0.875136,
+      "grad_norm": 0.4331625998020172,
+      "learning_rate": 4.030248667112629e-05,
+      "loss": 2.8425,
+      "step": 6837
+    },
+    {
+      "epoch": 0.875264,
+      "grad_norm": 0.39706695079803467,
+      "learning_rate": 4.022098337368113e-05,
+      "loss": 2.8726,
+      "step": 6838
+    },
+    {
+      "epoch": 0.875392,
+      "grad_norm": 0.4237077236175537,
+      "learning_rate": 4.013955911757922e-05,
+      "loss": 2.6362,
+      "step": 6839
+    },
+    {
+      "epoch": 0.87552,
+      "grad_norm": 0.4026554822921753,
+      "learning_rate": 4.0058213916818295e-05,
+      "loss": 2.7198,
+      "step": 6840
+    },
+    {
+      "epoch": 0.875648,
+      "grad_norm": 0.4035515785217285,
+      "learning_rate": 3.997694778538247e-05,
+      "loss": 2.8026,
+      "step": 6841
+    },
+    {
+      "epoch": 0.875776,
+      "grad_norm": 0.5746001601219177,
+      "learning_rate": 3.989576073724238e-05,
+      "loss": 2.7902,
+      "step": 6842
+    },
+    {
+      "epoch": 0.875904,
+      "grad_norm": 0.4237116873264313,
+      "learning_rate": 3.981465278635504e-05,
+      "loss": 2.8756,
+      "step": 6843
+    },
+    {
+      "epoch": 0.876032,
+      "grad_norm": 0.4437660872936249,
+      "learning_rate": 3.973362394666369e-05,
+      "loss": 2.8066,
+      "step": 6844
+    },
+    {
+      "epoch": 0.87616,
+      "grad_norm": 0.3983270227909088,
+      "learning_rate": 3.965267423209834e-05,
+      "loss": 2.9319,
+      "step": 6845
+    },
+    {
+      "epoch": 0.876288,
+      "grad_norm": 0.4172835648059845,
+      "learning_rate": 3.957180365657498e-05,
+      "loss": 2.8967,
+      "step": 6846
+    },
+    {
+      "epoch": 0.876416,
+      "grad_norm": 0.37722402811050415,
+      "learning_rate": 3.949101223399643e-05,
+      "loss": 2.6972,
+      "step": 6847
+    },
+    {
+      "epoch": 0.876544,
+      "grad_norm": 0.3836718201637268,
+      "learning_rate": 3.941029997825152e-05,
+      "loss": 2.7586,
+      "step": 6848
+    },
+    {
+      "epoch": 0.876672,
+      "grad_norm": 0.41036394238471985,
+      "learning_rate": 3.932966690321571e-05,
+      "loss": 2.7293,
+      "step": 6849
+    },
+    {
+      "epoch": 0.8768,
+      "grad_norm": 0.39531266689300537,
+      "learning_rate": 3.9249113022750694e-05,
+      "loss": 2.8251,
+      "step": 6850
+    },
+    {
+      "epoch": 0.876928,
+      "grad_norm": 0.399813711643219,
+      "learning_rate": 3.9168638350704614e-05,
+      "loss": 2.73,
+      "step": 6851
+    },
+    {
+      "epoch": 0.877056,
+      "grad_norm": 0.4146224856376648,
+      "learning_rate": 3.908824290091195e-05,
+      "loss": 2.6963,
+      "step": 6852
+    },
+    {
+      "epoch": 0.877184,
+      "grad_norm": 0.3800411820411682,
+      "learning_rate": 3.900792668719372e-05,
+      "loss": 2.878,
+      "step": 6853
+    },
+    {
+      "epoch": 0.877312,
+      "grad_norm": 0.40799155831336975,
+      "learning_rate": 3.892768972335725e-05,
+      "loss": 2.7652,
+      "step": 6854
+    },
+    {
+      "epoch": 0.87744,
+      "grad_norm": 0.3822653591632843,
+      "learning_rate": 3.884753202319596e-05,
+      "loss": 2.6497,
+      "step": 6855
+    },
+    {
+      "epoch": 0.877568,
+      "grad_norm": 0.38125482201576233,
+      "learning_rate": 3.8767453600490167e-05,
+      "loss": 2.7491,
+      "step": 6856
+    },
+    {
+      "epoch": 0.877696,
+      "grad_norm": 0.40978899598121643,
+      "learning_rate": 3.868745446900618e-05,
+      "loss": 2.824,
+      "step": 6857
+    },
+    {
+      "epoch": 0.877824,
+      "grad_norm": 0.3854672610759735,
+      "learning_rate": 3.8607534642496777e-05,
+      "loss": 2.7085,
+      "step": 6858
+    },
+    {
+      "epoch": 0.877952,
+      "grad_norm": 0.43170326948165894,
+      "learning_rate": 3.852769413470098e-05,
+      "loss": 2.818,
+      "step": 6859
+    },
+    {
+      "epoch": 0.87808,
+      "grad_norm": 0.4007253646850586,
+      "learning_rate": 3.844793295934451e-05,
+      "loss": 2.8846,
+      "step": 6860
+    },
+    {
+      "epoch": 0.878208,
+      "grad_norm": 0.4167078137397766,
+      "learning_rate": 3.8368251130139086e-05,
+      "loss": 2.7556,
+      "step": 6861
+    },
+    {
+      "epoch": 0.878336,
+      "grad_norm": 0.41342395544052124,
+      "learning_rate": 3.828864866078302e-05,
+      "loss": 2.7635,
+      "step": 6862
+    },
+    {
+      "epoch": 0.878464,
+      "grad_norm": 0.3984980881214142,
+      "learning_rate": 3.820912556496087e-05,
+      "loss": 2.75,
+      "step": 6863
+    },
+    {
+      "epoch": 0.878592,
+      "grad_norm": 0.4291757643222809,
+      "learning_rate": 3.812968185634352e-05,
+      "loss": 2.7601,
+      "step": 6864
+    },
+    {
+      "epoch": 0.87872,
+      "grad_norm": 0.42450547218322754,
+      "learning_rate": 3.805031754858834e-05,
+      "loss": 2.6968,
+      "step": 6865
+    },
+    {
+      "epoch": 0.878848,
+      "grad_norm": 0.42955103516578674,
+      "learning_rate": 3.7971032655339024e-05,
+      "loss": 2.8356,
+      "step": 6866
+    },
+    {
+      "epoch": 0.878976,
+      "grad_norm": 0.3717467188835144,
+      "learning_rate": 3.789182719022538e-05,
+      "loss": 2.6935,
+      "step": 6867
+    },
+    {
+      "epoch": 0.879104,
+      "grad_norm": 0.4085598289966583,
+      "learning_rate": 3.781270116686392e-05,
+      "loss": 2.6393,
+      "step": 6868
+    },
+    {
+      "epoch": 0.879232,
+      "grad_norm": 0.4069584906101227,
+      "learning_rate": 3.77336545988573e-05,
+      "loss": 2.7671,
+      "step": 6869
+    },
+    {
+      "epoch": 0.87936,
+      "grad_norm": 0.40315112471580505,
+      "learning_rate": 3.765468749979445e-05,
+      "loss": 2.8697,
+      "step": 6870
+    },
+    {
+      "epoch": 0.879488,
+      "grad_norm": 0.4174252152442932,
+      "learning_rate": 3.757579988325083e-05,
+      "loss": 2.8099,
+      "step": 6871
+    },
+    {
+      "epoch": 0.879616,
+      "grad_norm": 0.4108034670352936,
+      "learning_rate": 3.749699176278809e-05,
+      "loss": 2.7928,
+      "step": 6872
+    },
+    {
+      "epoch": 0.879744,
+      "grad_norm": 0.39325934648513794,
+      "learning_rate": 3.741826315195418e-05,
+      "loss": 2.7184,
+      "step": 6873
+    },
+    {
+      "epoch": 0.879872,
+      "grad_norm": 0.39533185958862305,
+      "learning_rate": 3.733961406428354e-05,
+      "loss": 2.7832,
+      "step": 6874
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.44735926389694214,
+      "learning_rate": 3.7261044513296926e-05,
+      "loss": 2.7042,
+      "step": 6875
+    },
+    {
+      "epoch": 0.880128,
+      "grad_norm": 0.40452417731285095,
+      "learning_rate": 3.7182554512501296e-05,
+      "loss": 2.6623,
+      "step": 6876
+    },
+    {
+      "epoch": 0.880256,
+      "grad_norm": 0.4243444800376892,
+      "learning_rate": 3.710414407538998e-05,
+      "loss": 2.7993,
+      "step": 6877
+    },
+    {
+      "epoch": 0.880384,
+      "grad_norm": 0.41359013319015503,
+      "learning_rate": 3.7025813215442684e-05,
+      "loss": 2.8125,
+      "step": 6878
+    },
+    {
+      "epoch": 0.880512,
+      "grad_norm": 0.4214875400066376,
+      "learning_rate": 3.694756194612531e-05,
+      "loss": 2.7362,
+      "step": 6879
+    },
+    {
+      "epoch": 0.88064,
+      "grad_norm": 0.3878746032714844,
+      "learning_rate": 3.686939028089015e-05,
+      "loss": 2.7494,
+      "step": 6880
+    },
+    {
+      "epoch": 0.880768,
+      "grad_norm": 0.4124908447265625,
+      "learning_rate": 3.679129823317595e-05,
+      "loss": 2.8716,
+      "step": 6881
+    },
+    {
+      "epoch": 0.880896,
+      "grad_norm": 0.4188864827156067,
+      "learning_rate": 3.6713285816407535e-05,
+      "loss": 2.6753,
+      "step": 6882
+    },
+    {
+      "epoch": 0.881024,
+      "grad_norm": 0.4181700646877289,
+      "learning_rate": 3.663535304399607e-05,
+      "loss": 2.7933,
+      "step": 6883
+    },
+    {
+      "epoch": 0.881152,
+      "grad_norm": 0.4064963757991791,
+      "learning_rate": 3.655749992933932e-05,
+      "loss": 2.8872,
+      "step": 6884
+    },
+    {
+      "epoch": 0.88128,
+      "grad_norm": 0.41051480174064636,
+      "learning_rate": 3.647972648582104e-05,
+      "loss": 2.8759,
+      "step": 6885
+    },
+    {
+      "epoch": 0.881408,
+      "grad_norm": 0.38654816150665283,
+      "learning_rate": 3.640203272681131e-05,
+      "loss": 2.8408,
+      "step": 6886
+    },
+    {
+      "epoch": 0.881536,
+      "grad_norm": 0.4362204670906067,
+      "learning_rate": 3.632441866566672e-05,
+      "loss": 2.8922,
+      "step": 6887
+    },
+    {
+      "epoch": 0.881664,
+      "grad_norm": 0.42553138732910156,
+      "learning_rate": 3.624688431572981e-05,
+      "loss": 2.6309,
+      "step": 6888
+    },
+    {
+      "epoch": 0.881792,
+      "grad_norm": 0.42558544874191284,
+      "learning_rate": 3.6169429690329825e-05,
+      "loss": 2.8412,
+      "step": 6889
+    },
+    {
+      "epoch": 0.88192,
+      "grad_norm": 0.4232051372528076,
+      "learning_rate": 3.60920548027821e-05,
+      "loss": 2.8588,
+      "step": 6890
+    },
+    {
+      "epoch": 0.882048,
+      "grad_norm": 0.4419994652271271,
+      "learning_rate": 3.6014759666388284e-05,
+      "loss": 2.6467,
+      "step": 6891
+    },
+    {
+      "epoch": 0.882176,
+      "grad_norm": 0.4427701532840729,
+      "learning_rate": 3.593754429443613e-05,
+      "loss": 2.8018,
+      "step": 6892
+    },
+    {
+      "epoch": 0.882304,
+      "grad_norm": 0.4117230176925659,
+      "learning_rate": 3.586040870020008e-05,
+      "loss": 2.8368,
+      "step": 6893
+    },
+    {
+      "epoch": 0.882432,
+      "grad_norm": 0.3999704420566559,
+      "learning_rate": 3.578335289694051e-05,
+      "loss": 2.7593,
+      "step": 6894
+    },
+    {
+      "epoch": 0.88256,
+      "grad_norm": 0.4605156183242798,
+      "learning_rate": 3.570637689790418e-05,
+      "loss": 2.7016,
+      "step": 6895
+    },
+    {
+      "epoch": 0.882688,
+      "grad_norm": 0.4085976183414459,
+      "learning_rate": 3.562948071632427e-05,
+      "loss": 2.8611,
+      "step": 6896
+    },
+    {
+      "epoch": 0.882816,
+      "grad_norm": 0.4350503087043762,
+      "learning_rate": 3.5552664365420094e-05,
+      "loss": 2.9266,
+      "step": 6897
+    },
+    {
+      "epoch": 0.882944,
+      "grad_norm": 0.45539504289627075,
+      "learning_rate": 3.547592785839726e-05,
+      "loss": 2.7775,
+      "step": 6898
+    },
+    {
+      "epoch": 0.883072,
+      "grad_norm": 0.4140052795410156,
+      "learning_rate": 3.539927120844766e-05,
+      "loss": 2.8979,
+      "step": 6899
+    },
+    {
+      "epoch": 0.8832,
+      "grad_norm": 0.42870891094207764,
+      "learning_rate": 3.532269442874941e-05,
+      "loss": 2.7775,
+      "step": 6900
+    },
+    {
+      "epoch": 0.883328,
+      "grad_norm": 0.4197491705417633,
+      "learning_rate": 3.524619753246694e-05,
+      "loss": 2.6971,
+      "step": 6901
+    },
+    {
+      "epoch": 0.883456,
+      "grad_norm": 0.4024498164653778,
+      "learning_rate": 3.5169780532751105e-05,
+      "loss": 2.9225,
+      "step": 6902
+    },
+    {
+      "epoch": 0.883584,
+      "grad_norm": 0.4439351260662079,
+      "learning_rate": 3.509344344273863e-05,
+      "loss": 2.9741,
+      "step": 6903
+    },
+    {
+      "epoch": 0.883712,
+      "grad_norm": 0.44953757524490356,
+      "learning_rate": 3.501718627555301e-05,
+      "loss": 2.831,
+      "step": 6904
+    },
+    {
+      "epoch": 0.88384,
+      "grad_norm": 0.40599435567855835,
+      "learning_rate": 3.4941009044303585e-05,
+      "loss": 2.8889,
+      "step": 6905
+    },
+    {
+      "epoch": 0.883968,
+      "grad_norm": 0.4028887450695038,
+      "learning_rate": 3.486491176208617e-05,
+      "loss": 2.8052,
+      "step": 6906
+    },
+    {
+      "epoch": 0.884096,
+      "grad_norm": 0.4125836491584778,
+      "learning_rate": 3.4788894441982685e-05,
+      "loss": 2.8699,
+      "step": 6907
+    },
+    {
+      "epoch": 0.884224,
+      "grad_norm": 0.393106073141098,
+      "learning_rate": 3.471295709706146e-05,
+      "loss": 2.7082,
+      "step": 6908
+    },
+    {
+      "epoch": 0.884352,
+      "grad_norm": 0.3960779309272766,
+      "learning_rate": 3.4637099740376935e-05,
+      "loss": 2.8432,
+      "step": 6909
+    },
+    {
+      "epoch": 0.88448,
+      "grad_norm": 0.45479726791381836,
+      "learning_rate": 3.456132238496984e-05,
+      "loss": 2.7504,
+      "step": 6910
+    },
+    {
+      "epoch": 0.884608,
+      "grad_norm": 0.4278700053691864,
+      "learning_rate": 3.448562504386738e-05,
+      "loss": 2.7718,
+      "step": 6911
+    },
+    {
+      "epoch": 0.884736,
+      "grad_norm": 0.4192088544368744,
+      "learning_rate": 3.4410007730082685e-05,
+      "loss": 2.7932,
+      "step": 6912
+    },
+    {
+      "epoch": 0.884864,
+      "grad_norm": 0.3983749747276306,
+      "learning_rate": 3.4334470456615254e-05,
+      "loss": 2.6966,
+      "step": 6913
+    },
+    {
+      "epoch": 0.884992,
+      "grad_norm": 0.3989611268043518,
+      "learning_rate": 3.425901323645075e-05,
+      "loss": 2.8491,
+      "step": 6914
+    },
+    {
+      "epoch": 0.88512,
+      "grad_norm": 0.3989495635032654,
+      "learning_rate": 3.418363608256125e-05,
+      "loss": 2.915,
+      "step": 6915
+    },
+    {
+      "epoch": 0.885248,
+      "grad_norm": 0.41082844138145447,
+      "learning_rate": 3.410833900790483e-05,
+      "loss": 2.8244,
+      "step": 6916
+    },
+    {
+      "epoch": 0.885376,
+      "grad_norm": 0.37322095036506653,
+      "learning_rate": 3.403312202542608e-05,
+      "loss": 2.7765,
+      "step": 6917
+    },
+    {
+      "epoch": 0.885504,
+      "grad_norm": 0.39516761898994446,
+      "learning_rate": 3.395798514805565e-05,
+      "loss": 2.8308,
+      "step": 6918
+    },
+    {
+      "epoch": 0.885632,
+      "grad_norm": 0.42023423314094543,
+      "learning_rate": 3.3882928388710376e-05,
+      "loss": 2.8732,
+      "step": 6919
+    },
+    {
+      "epoch": 0.88576,
+      "grad_norm": 0.38403281569480896,
+      "learning_rate": 3.3807951760293275e-05,
+      "loss": 2.7572,
+      "step": 6920
+    },
+    {
+      "epoch": 0.885888,
+      "grad_norm": 0.41202279925346375,
+      "learning_rate": 3.373305527569398e-05,
+      "loss": 2.7088,
+      "step": 6921
+    },
+    {
+      "epoch": 0.886016,
+      "grad_norm": 0.41739267110824585,
+      "learning_rate": 3.3658238947787855e-05,
+      "loss": 2.8671,
+      "step": 6922
+    },
+    {
+      "epoch": 0.886144,
+      "grad_norm": 0.4353700876235962,
+      "learning_rate": 3.358350278943678e-05,
+      "loss": 2.7184,
+      "step": 6923
+    },
+    {
+      "epoch": 0.886272,
+      "grad_norm": 0.40842410922050476,
+      "learning_rate": 3.3508846813488646e-05,
+      "loss": 2.8766,
+      "step": 6924
+    },
+    {
+      "epoch": 0.8864,
+      "grad_norm": 0.42620518803596497,
+      "learning_rate": 3.343427103277785e-05,
+      "loss": 2.7841,
+      "step": 6925
+    },
+    {
+      "epoch": 0.886528,
+      "grad_norm": 0.4129617214202881,
+      "learning_rate": 3.335977546012481e-05,
+      "loss": 2.7192,
+      "step": 6926
+    },
+    {
+      "epoch": 0.886656,
+      "grad_norm": 0.46160104870796204,
+      "learning_rate": 3.328536010833605e-05,
+      "loss": 2.9029,
+      "step": 6927
+    },
+    {
+      "epoch": 0.886784,
+      "grad_norm": 0.35625430941581726,
+      "learning_rate": 3.3211024990204565e-05,
+      "loss": 2.7074,
+      "step": 6928
+    },
+    {
+      "epoch": 0.886912,
+      "grad_norm": 0.4055804908275604,
+      "learning_rate": 3.313677011850924e-05,
+      "loss": 2.7361,
+      "step": 6929
+    },
+    {
+      "epoch": 0.88704,
+      "grad_norm": 0.4065127670764923,
+      "learning_rate": 3.306259550601559e-05,
+      "loss": 2.7765,
+      "step": 6930
+    },
+    {
+      "epoch": 0.887168,
+      "grad_norm": 0.3899782598018646,
+      "learning_rate": 3.298850116547486e-05,
+      "loss": 2.6979,
+      "step": 6931
+    },
+    {
+      "epoch": 0.887296,
+      "grad_norm": 0.4334196150302887,
+      "learning_rate": 3.291448710962497e-05,
+      "loss": 2.7222,
+      "step": 6932
+    },
+    {
+      "epoch": 0.887424,
+      "grad_norm": 0.4290858209133148,
+      "learning_rate": 3.284055335118963e-05,
+      "loss": 2.9348,
+      "step": 6933
+    },
+    {
+      "epoch": 0.887552,
+      "grad_norm": 0.44772645831108093,
+      "learning_rate": 3.276669990287895e-05,
+      "loss": 2.7405,
+      "step": 6934
+    },
+    {
+      "epoch": 0.88768,
+      "grad_norm": 0.42176827788352966,
+      "learning_rate": 3.269292677738922e-05,
+      "loss": 2.7656,
+      "step": 6935
+    },
+    {
+      "epoch": 0.887808,
+      "grad_norm": 0.4027102291584015,
+      "learning_rate": 3.261923398740285e-05,
+      "loss": 2.8083,
+      "step": 6936
+    },
+    {
+      "epoch": 0.887936,
+      "grad_norm": 0.43110325932502747,
+      "learning_rate": 3.2545621545588436e-05,
+      "loss": 2.7628,
+      "step": 6937
+    },
+    {
+      "epoch": 0.888064,
+      "grad_norm": 0.3973923325538635,
+      "learning_rate": 3.247208946460084e-05,
+      "loss": 2.8519,
+      "step": 6938
+    },
+    {
+      "epoch": 0.888192,
+      "grad_norm": 0.40758347511291504,
+      "learning_rate": 3.2398637757081185e-05,
+      "loss": 2.8387,
+      "step": 6939
+    },
+    {
+      "epoch": 0.88832,
+      "grad_norm": 0.38906505703926086,
+      "learning_rate": 3.232526643565664e-05,
+      "loss": 2.8033,
+      "step": 6940
+    },
+    {
+      "epoch": 0.888448,
+      "grad_norm": 0.3994798958301544,
+      "learning_rate": 3.2251975512940514e-05,
+      "loss": 2.6901,
+      "step": 6941
+    },
+    {
+      "epoch": 0.888576,
+      "grad_norm": 0.4096708595752716,
+      "learning_rate": 3.2178765001532386e-05,
+      "loss": 2.7785,
+      "step": 6942
+    },
+    {
+      "epoch": 0.888704,
+      "grad_norm": 0.393770307302475,
+      "learning_rate": 3.210563491401808e-05,
+      "loss": 2.7442,
+      "step": 6943
+    },
+    {
+      "epoch": 0.888832,
+      "grad_norm": 0.4440455734729767,
+      "learning_rate": 3.203258526296937e-05,
+      "loss": 2.794,
+      "step": 6944
+    },
+    {
+      "epoch": 0.88896,
+      "grad_norm": 0.4161309599876404,
+      "learning_rate": 3.195961606094433e-05,
+      "loss": 2.6082,
+      "step": 6945
+    },
+    {
+      "epoch": 0.889088,
+      "grad_norm": 0.40642860531806946,
+      "learning_rate": 3.188672732048731e-05,
+      "loss": 2.8139,
+      "step": 6946
+    },
+    {
+      "epoch": 0.889216,
+      "grad_norm": 0.3972402513027191,
+      "learning_rate": 3.181391905412867e-05,
+      "loss": 2.7538,
+      "step": 6947
+    },
+    {
+      "epoch": 0.889344,
+      "grad_norm": 0.4079653322696686,
+      "learning_rate": 3.174119127438507e-05,
+      "loss": 2.757,
+      "step": 6948
+    },
+    {
+      "epoch": 0.889472,
+      "grad_norm": 0.4150797128677368,
+      "learning_rate": 3.166854399375929e-05,
+      "loss": 2.6468,
+      "step": 6949
+    },
+    {
+      "epoch": 0.8896,
+      "grad_norm": 0.43969231843948364,
+      "learning_rate": 3.159597722474006e-05,
+      "loss": 2.7564,
+      "step": 6950
+    },
+    {
+      "epoch": 0.889728,
+      "grad_norm": 0.4270954132080078,
+      "learning_rate": 3.1523490979802614e-05,
+      "loss": 2.828,
+      "step": 6951
+    },
+    {
+      "epoch": 0.889856,
+      "grad_norm": 0.36761942505836487,
+      "learning_rate": 3.145108527140805e-05,
+      "loss": 2.686,
+      "step": 6952
+    },
+    {
+      "epoch": 0.889984,
+      "grad_norm": 0.4363018870353699,
+      "learning_rate": 3.1378760112003853e-05,
+      "loss": 2.9209,
+      "step": 6953
+    },
+    {
+      "epoch": 0.890112,
+      "grad_norm": 0.43821436166763306,
+      "learning_rate": 3.1306515514023527e-05,
+      "loss": 2.9194,
+      "step": 6954
+    },
+    {
+      "epoch": 0.89024,
+      "grad_norm": 0.43578243255615234,
+      "learning_rate": 3.1234351489886736e-05,
+      "loss": 2.8159,
+      "step": 6955
+    },
+    {
+      "epoch": 0.890368,
+      "grad_norm": 0.39462703466415405,
+      "learning_rate": 3.116226805199929e-05,
+      "loss": 2.824,
+      "step": 6956
+    },
+    {
+      "epoch": 0.890496,
+      "grad_norm": 0.4388856887817383,
+      "learning_rate": 3.109026521275321e-05,
+      "loss": 2.8499,
+      "step": 6957
+    },
+    {
+      "epoch": 0.890624,
+      "grad_norm": 0.40002986788749695,
+      "learning_rate": 3.1018342984526604e-05,
+      "loss": 2.7437,
+      "step": 6958
+    },
+    {
+      "epoch": 0.890752,
+      "grad_norm": 0.4126453101634979,
+      "learning_rate": 3.0946501379683686e-05,
+      "loss": 2.8126,
+      "step": 6959
+    },
+    {
+      "epoch": 0.89088,
+      "grad_norm": 0.42338117957115173,
+      "learning_rate": 3.087474041057503e-05,
+      "loss": 2.9638,
+      "step": 6960
+    },
+    {
+      "epoch": 0.891008,
+      "grad_norm": 0.406894326210022,
+      "learning_rate": 3.080306008953704e-05,
+      "loss": 2.736,
+      "step": 6961
+    },
+    {
+      "epoch": 0.891136,
+      "grad_norm": 0.43001723289489746,
+      "learning_rate": 3.073146042889241e-05,
+      "loss": 2.6895,
+      "step": 6962
+    },
+    {
+      "epoch": 0.891264,
+      "grad_norm": 0.43289750814437866,
+      "learning_rate": 3.065994144094997e-05,
+      "loss": 2.8498,
+      "step": 6963
+    },
+    {
+      "epoch": 0.891392,
+      "grad_norm": 0.402869313955307,
+      "learning_rate": 3.05885031380046e-05,
+      "loss": 2.8696,
+      "step": 6964
+    },
+    {
+      "epoch": 0.89152,
+      "grad_norm": 0.43561261892318726,
+      "learning_rate": 3.0517145532337476e-05,
+      "loss": 2.8166,
+      "step": 6965
+    },
+    {
+      "epoch": 0.891648,
+      "grad_norm": 0.4164077937602997,
+      "learning_rate": 3.0445868636215678e-05,
+      "loss": 2.9265,
+      "step": 6966
+    },
+    {
+      "epoch": 0.891776,
+      "grad_norm": 0.4243529140949249,
+      "learning_rate": 3.037467246189257e-05,
+      "loss": 2.7879,
+      "step": 6967
+    },
+    {
+      "epoch": 0.891904,
+      "grad_norm": 0.4544285535812378,
+      "learning_rate": 3.0303557021607754e-05,
+      "loss": 2.8777,
+      "step": 6968
+    },
+    {
+      "epoch": 0.892032,
+      "grad_norm": 0.39643898606300354,
+      "learning_rate": 3.0232522327586664e-05,
+      "loss": 2.6094,
+      "step": 6969
+    },
+    {
+      "epoch": 0.89216,
+      "grad_norm": 0.3827550709247589,
+      "learning_rate": 3.0161568392040982e-05,
+      "loss": 2.7178,
+      "step": 6970
+    },
+    {
+      "epoch": 0.892288,
+      "grad_norm": 0.4115031957626343,
+      "learning_rate": 3.0090695227168507e-05,
+      "loss": 2.9112,
+      "step": 6971
+    },
+    {
+      "epoch": 0.892416,
+      "grad_norm": 0.41958722472190857,
+      "learning_rate": 3.0019902845153214e-05,
+      "loss": 2.7769,
+      "step": 6972
+    },
+    {
+      "epoch": 0.892544,
+      "grad_norm": 0.4288786053657532,
+      "learning_rate": 2.9949191258164976e-05,
+      "loss": 2.8413,
+      "step": 6973
+    },
+    {
+      "epoch": 0.892672,
+      "grad_norm": 0.39965325593948364,
+      "learning_rate": 2.9878560478360183e-05,
+      "loss": 2.7204,
+      "step": 6974
+    },
+    {
+      "epoch": 0.8928,
+      "grad_norm": 0.39859431982040405,
+      "learning_rate": 2.9808010517880847e-05,
+      "loss": 2.8185,
+      "step": 6975
+    },
+    {
+      "epoch": 0.892928,
+      "grad_norm": 0.41201338171958923,
+      "learning_rate": 2.9737541388855536e-05,
+      "loss": 2.7716,
+      "step": 6976
+    },
+    {
+      "epoch": 0.893056,
+      "grad_norm": 0.3908863663673401,
+      "learning_rate": 2.966715310339857e-05,
+      "loss": 2.8211,
+      "step": 6977
+    },
+    {
+      "epoch": 0.893184,
+      "grad_norm": 0.3922537565231323,
+      "learning_rate": 2.9596845673610596e-05,
+      "loss": 2.9052,
+      "step": 6978
+    },
+    {
+      "epoch": 0.893312,
+      "grad_norm": 0.42297065258026123,
+      "learning_rate": 2.9526619111578222e-05,
+      "loss": 2.7627,
+      "step": 6979
+    },
+    {
+      "epoch": 0.89344,
+      "grad_norm": 0.4069047272205353,
+      "learning_rate": 2.945647342937413e-05,
+      "loss": 2.9885,
+      "step": 6980
+    },
+    {
+      "epoch": 0.893568,
+      "grad_norm": 0.3875328004360199,
+      "learning_rate": 2.9386408639057392e-05,
+      "loss": 2.6081,
+      "step": 6981
+    },
+    {
+      "epoch": 0.893696,
+      "grad_norm": 0.42956221103668213,
+      "learning_rate": 2.9316424752672767e-05,
+      "loss": 2.8191,
+      "step": 6982
+    },
+    {
+      "epoch": 0.893824,
+      "grad_norm": 0.410031259059906,
+      "learning_rate": 2.92465217822514e-05,
+      "loss": 2.9128,
+      "step": 6983
+    },
+    {
+      "epoch": 0.893952,
+      "grad_norm": 0.45070749521255493,
+      "learning_rate": 2.9176699739810298e-05,
+      "loss": 2.8816,
+      "step": 6984
+    },
+    {
+      "epoch": 0.89408,
+      "grad_norm": 0.39603760838508606,
+      "learning_rate": 2.9106958637352797e-05,
+      "loss": 2.8017,
+      "step": 6985
+    },
+    {
+      "epoch": 0.894208,
+      "grad_norm": 0.3971194922924042,
+      "learning_rate": 2.90372984868682e-05,
+      "loss": 2.8557,
+      "step": 6986
+    },
+    {
+      "epoch": 0.894336,
+      "grad_norm": 0.40735748410224915,
+      "learning_rate": 2.8967719300331875e-05,
+      "loss": 2.8925,
+      "step": 6987
+    },
+    {
+      "epoch": 0.894464,
+      "grad_norm": 0.3790603578090668,
+      "learning_rate": 2.889822108970519e-05,
+      "loss": 2.809,
+      "step": 6988
+    },
+    {
+      "epoch": 0.894592,
+      "grad_norm": 0.39002111554145813,
+      "learning_rate": 2.8828803866935815e-05,
+      "loss": 2.8902,
+      "step": 6989
+    },
+    {
+      "epoch": 0.89472,
+      "grad_norm": 0.39271244406700134,
+      "learning_rate": 2.8759467643957372e-05,
+      "loss": 2.8441,
+      "step": 6990
+    },
+    {
+      "epoch": 0.894848,
+      "grad_norm": 0.41369161009788513,
+      "learning_rate": 2.8690212432689544e-05,
+      "loss": 2.7772,
+      "step": 6991
+    },
+    {
+      "epoch": 0.894976,
+      "grad_norm": 0.49351412057876587,
+      "learning_rate": 2.862103824503809e-05,
+      "loss": 2.8622,
+      "step": 6992
+    },
+    {
+      "epoch": 0.895104,
+      "grad_norm": 0.4229761064052582,
+      "learning_rate": 2.8551945092894825e-05,
+      "loss": 2.854,
+      "step": 6993
+    },
+    {
+      "epoch": 0.895232,
+      "grad_norm": 0.39308997988700867,
+      "learning_rate": 2.8482932988137644e-05,
+      "loss": 2.7141,
+      "step": 6994
+    },
+    {
+      "epoch": 0.89536,
+      "grad_norm": 0.427569717168808,
+      "learning_rate": 2.841400194263055e-05,
+      "loss": 2.7062,
+      "step": 6995
+    },
+    {
+      "epoch": 0.895488,
+      "grad_norm": 0.4273299276828766,
+      "learning_rate": 2.8345151968223738e-05,
+      "loss": 2.7953,
+      "step": 6996
+    },
+    {
+      "epoch": 0.895616,
+      "grad_norm": 0.4211599826812744,
+      "learning_rate": 2.8276383076753175e-05,
+      "loss": 2.8058,
+      "step": 6997
+    },
+    {
+      "epoch": 0.895744,
+      "grad_norm": 0.3928157389163971,
+      "learning_rate": 2.8207695280041023e-05,
+      "loss": 2.7464,
+      "step": 6998
+    },
+    {
+      "epoch": 0.895872,
+      "grad_norm": 0.40235084295272827,
+      "learning_rate": 2.8139088589895557e-05,
+      "loss": 2.6831,
+      "step": 6999
+    },
+    {
+      "epoch": 0.896,
+      "grad_norm": 0.39827680587768555,
+      "learning_rate": 2.8070563018111063e-05,
+      "loss": 2.8152,
+      "step": 7000
+    },
+    {
+      "epoch": 0.896128,
+      "grad_norm": 0.4045201241970062,
+      "learning_rate": 2.8002118576467785e-05,
+      "loss": 2.8382,
+      "step": 7001
+    },
+    {
+      "epoch": 0.896256,
+      "grad_norm": 0.42470258474349976,
+      "learning_rate": 2.7933755276732254e-05,
+      "loss": 2.9823,
+      "step": 7002
+    },
+    {
+      "epoch": 0.896384,
+      "grad_norm": 0.43872204422950745,
+      "learning_rate": 2.7865473130656793e-05,
+      "loss": 2.7741,
+      "step": 7003
+    },
+    {
+      "epoch": 0.896512,
+      "grad_norm": 0.4113384783267975,
+      "learning_rate": 2.779727214997996e-05,
+      "loss": 2.8856,
+      "step": 7004
+    },
+    {
+      "epoch": 0.89664,
+      "grad_norm": 0.4409100115299225,
+      "learning_rate": 2.7729152346426366e-05,
+      "loss": 2.7888,
+      "step": 7005
+    },
+    {
+      "epoch": 0.896768,
+      "grad_norm": 0.4324277341365814,
+      "learning_rate": 2.7661113731706488e-05,
+      "loss": 2.745,
+      "step": 7006
+    },
+    {
+      "epoch": 0.896896,
+      "grad_norm": 0.4027922749519348,
+      "learning_rate": 2.7593156317516966e-05,
+      "loss": 2.7502,
+      "step": 7007
+    },
+    {
+      "epoch": 0.897024,
+      "grad_norm": 0.38401898741722107,
+      "learning_rate": 2.7525280115540506e-05,
+      "loss": 2.8013,
+      "step": 7008
+    },
+    {
+      "epoch": 0.897152,
+      "grad_norm": 0.3884783089160919,
+      "learning_rate": 2.7457485137445727e-05,
+      "loss": 2.7343,
+      "step": 7009
+    },
+    {
+      "epoch": 0.89728,
+      "grad_norm": 0.4106162488460541,
+      "learning_rate": 2.738977139488752e-05,
+      "loss": 2.7373,
+      "step": 7010
+    },
+    {
+      "epoch": 0.897408,
+      "grad_norm": 0.3941263258457184,
+      "learning_rate": 2.732213889950652e-05,
+      "loss": 2.7934,
+      "step": 7011
+    },
+    {
+      "epoch": 0.897536,
+      "grad_norm": 0.44603320956230164,
+      "learning_rate": 2.7254587662929543e-05,
+      "loss": 2.7906,
+      "step": 7012
+    },
+    {
+      "epoch": 0.897664,
+      "grad_norm": 0.3968757688999176,
+      "learning_rate": 2.718711769676957e-05,
+      "loss": 2.8569,
+      "step": 7013
+    },
+    {
+      "epoch": 0.897792,
+      "grad_norm": 0.4007909595966339,
+      "learning_rate": 2.7119729012625382e-05,
+      "loss": 2.6823,
+      "step": 7014
+    },
+    {
+      "epoch": 0.89792,
+      "grad_norm": 0.413675993680954,
+      "learning_rate": 2.705242162208188e-05,
+      "loss": 2.7102,
+      "step": 7015
+    },
+    {
+      "epoch": 0.898048,
+      "grad_norm": 0.42163369059562683,
+      "learning_rate": 2.6985195536709918e-05,
+      "loss": 2.9012,
+      "step": 7016
+    },
+    {
+      "epoch": 0.898176,
+      "grad_norm": 0.39900872111320496,
+      "learning_rate": 2.6918050768066528e-05,
+      "loss": 2.6561,
+      "step": 7017
+    },
+    {
+      "epoch": 0.898304,
+      "grad_norm": 0.40780001878738403,
+      "learning_rate": 2.6850987327694698e-05,
+      "loss": 2.6638,
+      "step": 7018
+    },
+    {
+      "epoch": 0.898432,
+      "grad_norm": 0.41159364581108093,
+      "learning_rate": 2.678400522712332e-05,
+      "loss": 2.7029,
+      "step": 7019
+    },
+    {
+      "epoch": 0.89856,
+      "grad_norm": 0.4096963703632355,
+      "learning_rate": 2.671710447786746e-05,
+      "loss": 2.7049,
+      "step": 7020
+    },
+    {
+      "epoch": 0.898688,
+      "grad_norm": 0.41576507687568665,
+      "learning_rate": 2.665028509142803e-05,
+      "loss": 2.7367,
+      "step": 7021
+    },
+    {
+      "epoch": 0.898816,
+      "grad_norm": 0.3890874683856964,
+      "learning_rate": 2.6583547079292224e-05,
+      "loss": 2.6744,
+      "step": 7022
+    },
+    {
+      "epoch": 0.898944,
+      "grad_norm": 0.4126293957233429,
+      "learning_rate": 2.6516890452932928e-05,
+      "loss": 2.7248,
+      "step": 7023
+    },
+    {
+      "epoch": 0.899072,
+      "grad_norm": 0.4598352909088135,
+      "learning_rate": 2.645031522380925e-05,
+      "loss": 2.8726,
+      "step": 7024
+    },
+    {
+      "epoch": 0.8992,
+      "grad_norm": 0.4295518100261688,
+      "learning_rate": 2.6383821403366258e-05,
+      "loss": 2.8203,
+      "step": 7025
+    },
+    {
+      "epoch": 0.899328,
+      "grad_norm": 0.43129631876945496,
+      "learning_rate": 2.6317409003035032e-05,
+      "loss": 2.7732,
+      "step": 7026
+    },
+    {
+      "epoch": 0.899456,
+      "grad_norm": 0.426800012588501,
+      "learning_rate": 2.62510780342326e-05,
+      "loss": 2.7938,
+      "step": 7027
+    },
+    {
+      "epoch": 0.899584,
+      "grad_norm": 0.6993824243545532,
+      "learning_rate": 2.6184828508362013e-05,
+      "loss": 2.7339,
+      "step": 7028
+    },
+    {
+      "epoch": 0.899712,
+      "grad_norm": 0.3830450773239136,
+      "learning_rate": 2.6118660436812324e-05,
+      "loss": 2.849,
+      "step": 7029
+    },
+    {
+      "epoch": 0.89984,
+      "grad_norm": 0.41029542684555054,
+      "learning_rate": 2.6052573830958547e-05,
+      "loss": 2.9071,
+      "step": 7030
+    },
+    {
+      "epoch": 0.899968,
+      "grad_norm": 0.42052197456359863,
+      "learning_rate": 2.5986568702161818e-05,
+      "loss": 2.7771,
+      "step": 7031
+    },
+    {
+      "epoch": 0.900096,
+      "grad_norm": 0.3978026807308197,
+      "learning_rate": 2.592064506176922e-05,
+      "loss": 2.8338,
+      "step": 7032
+    },
+    {
+      "epoch": 0.900224,
+      "grad_norm": 0.40983033180236816,
+      "learning_rate": 2.585480292111375e-05,
+      "loss": 2.7173,
+      "step": 7033
+    },
+    {
+      "epoch": 0.900352,
+      "grad_norm": 0.4297316074371338,
+      "learning_rate": 2.578904229151441e-05,
+      "loss": 2.8276,
+      "step": 7034
+    },
+    {
+      "epoch": 0.90048,
+      "grad_norm": 0.40545281767845154,
+      "learning_rate": 2.5723363184276206e-05,
+      "loss": 2.6547,
+      "step": 7035
+    },
+    {
+      "epoch": 0.900608,
+      "grad_norm": 0.42974191904067993,
+      "learning_rate": 2.565776561069022e-05,
+      "loss": 2.8255,
+      "step": 7036
+    },
+    {
+      "epoch": 0.900736,
+      "grad_norm": 0.38975006341934204,
+      "learning_rate": 2.559224958203321e-05,
+      "loss": 2.6979,
+      "step": 7037
+    },
+    {
+      "epoch": 0.900864,
+      "grad_norm": 0.42255353927612305,
+      "learning_rate": 2.5526815109568447e-05,
+      "loss": 2.7728,
+      "step": 7038
+    },
+    {
+      "epoch": 0.900992,
+      "grad_norm": 0.4280490577220917,
+      "learning_rate": 2.5461462204544704e-05,
+      "loss": 2.8362,
+      "step": 7039
+    },
+    {
+      "epoch": 0.90112,
+      "grad_norm": 0.3833889365196228,
+      "learning_rate": 2.539619087819689e-05,
+      "loss": 2.7645,
+      "step": 7040
+    },
+    {
+      "epoch": 0.901248,
+      "grad_norm": 0.4482387900352478,
+      "learning_rate": 2.5331001141745968e-05,
+      "loss": 2.7291,
+      "step": 7041
+    },
+    {
+      "epoch": 0.901376,
+      "grad_norm": 0.42637893557548523,
+      "learning_rate": 2.526589300639881e-05,
+      "loss": 2.8347,
+      "step": 7042
+    },
+    {
+      "epoch": 0.901504,
+      "grad_norm": 0.4354317784309387,
+      "learning_rate": 2.520086648334824e-05,
+      "loss": 2.7845,
+      "step": 7043
+    },
+    {
+      "epoch": 0.901632,
+      "grad_norm": 0.40458714962005615,
+      "learning_rate": 2.5135921583773036e-05,
+      "loss": 2.8533,
+      "step": 7044
+    },
+    {
+      "epoch": 0.90176,
+      "grad_norm": 0.40392860770225525,
+      "learning_rate": 2.5071058318837936e-05,
+      "loss": 2.8244,
+      "step": 7045
+    },
+    {
+      "epoch": 0.901888,
+      "grad_norm": 0.39721769094467163,
+      "learning_rate": 2.5006276699693852e-05,
+      "loss": 2.862,
+      "step": 7046
+    },
+    {
+      "epoch": 0.902016,
+      "grad_norm": 0.4257240891456604,
+      "learning_rate": 2.4941576737477434e-05,
+      "loss": 2.6641,
+      "step": 7047
+    },
+    {
+      "epoch": 0.902144,
+      "grad_norm": 0.4292222857475281,
+      "learning_rate": 2.4876958443311282e-05,
+      "loss": 2.7912,
+      "step": 7048
+    },
+    {
+      "epoch": 0.902272,
+      "grad_norm": 0.43063947558403015,
+      "learning_rate": 2.481242182830401e-05,
+      "loss": 2.6791,
+      "step": 7049
+    },
+    {
+      "epoch": 0.9024,
+      "grad_norm": 0.4001077711582184,
+      "learning_rate": 2.4747966903550358e-05,
+      "loss": 2.6967,
+      "step": 7050
+    },
+    {
+      "epoch": 0.902528,
+      "grad_norm": 0.4038490355014801,
+      "learning_rate": 2.4683593680130732e-05,
+      "loss": 2.8304,
+      "step": 7051
+    },
+    {
+      "epoch": 0.902656,
+      "grad_norm": 0.4329737722873688,
+      "learning_rate": 2.461930216911168e-05,
+      "loss": 2.8169,
+      "step": 7052
+    },
+    {
+      "epoch": 0.902784,
+      "grad_norm": 0.3931262493133545,
+      "learning_rate": 2.455509238154574e-05,
+      "loss": 2.7767,
+      "step": 7053
+    },
+    {
+      "epoch": 0.902912,
+      "grad_norm": 0.38865533471107483,
+      "learning_rate": 2.4490964328471254e-05,
+      "loss": 2.7954,
+      "step": 7054
+    },
+    {
+      "epoch": 0.90304,
+      "grad_norm": 0.39426887035369873,
+      "learning_rate": 2.442691802091257e-05,
+      "loss": 2.6547,
+      "step": 7055
+    },
+    {
+      "epoch": 0.903168,
+      "grad_norm": 0.4193617105484009,
+      "learning_rate": 2.4362953469879935e-05,
+      "loss": 2.7755,
+      "step": 7056
+    },
+    {
+      "epoch": 0.903296,
+      "grad_norm": 0.4308586120605469,
+      "learning_rate": 2.4299070686369717e-05,
+      "loss": 2.7908,
+      "step": 7057
+    },
+    {
+      "epoch": 0.903424,
+      "grad_norm": 0.3883199691772461,
+      "learning_rate": 2.423526968136397e-05,
+      "loss": 2.7301,
+      "step": 7058
+    },
+    {
+      "epoch": 0.903552,
+      "grad_norm": 0.39140820503234863,
+      "learning_rate": 2.417155046583097e-05,
+      "loss": 2.8501,
+      "step": 7059
+    },
+    {
+      "epoch": 0.90368,
+      "grad_norm": 0.4274643659591675,
+      "learning_rate": 2.4107913050724628e-05,
+      "loss": 2.8225,
+      "step": 7060
+    },
+    {
+      "epoch": 0.903808,
+      "grad_norm": 0.4105488657951355,
+      "learning_rate": 2.4044357446985133e-05,
+      "loss": 2.8899,
+      "step": 7061
+    },
+    {
+      "epoch": 0.903936,
+      "grad_norm": 0.4074006974697113,
+      "learning_rate": 2.3980883665538357e-05,
+      "loss": 2.8269,
+      "step": 7062
+    },
+    {
+      "epoch": 0.904064,
+      "grad_norm": 0.3989720344543457,
+      "learning_rate": 2.391749171729618e-05,
+      "loss": 2.7544,
+      "step": 7063
+    },
+    {
+      "epoch": 0.904192,
+      "grad_norm": 0.41034916043281555,
+      "learning_rate": 2.385418161315639e-05,
+      "loss": 2.7981,
+      "step": 7064
+    },
+    {
+      "epoch": 0.90432,
+      "grad_norm": 0.3993270993232727,
+      "learning_rate": 2.379095336400272e-05,
+      "loss": 2.7414,
+      "step": 7065
+    },
+    {
+      "epoch": 0.904448,
+      "grad_norm": 0.4075114130973816,
+      "learning_rate": 2.3727806980704758e-05,
+      "loss": 2.8868,
+      "step": 7066
+    },
+    {
+      "epoch": 0.904576,
+      "grad_norm": 0.3987666070461273,
+      "learning_rate": 2.3664742474118316e-05,
+      "loss": 2.6966,
+      "step": 7067
+    },
+    {
+      "epoch": 0.904704,
+      "grad_norm": 0.40569376945495605,
+      "learning_rate": 2.360175985508467e-05,
+      "loss": 2.7862,
+      "step": 7068
+    },
+    {
+      "epoch": 0.904832,
+      "grad_norm": 0.4228448271751404,
+      "learning_rate": 2.3538859134431546e-05,
+      "loss": 2.6657,
+      "step": 7069
+    },
+    {
+      "epoch": 0.90496,
+      "grad_norm": 0.39881983399391174,
+      "learning_rate": 2.347604032297207e-05,
+      "loss": 2.6262,
+      "step": 7070
+    },
+    {
+      "epoch": 0.905088,
+      "grad_norm": 0.3973422050476074,
+      "learning_rate": 2.3413303431505607e-05,
+      "loss": 2.783,
+      "step": 7071
+    },
+    {
+      "epoch": 0.905216,
+      "grad_norm": 0.40984949469566345,
+      "learning_rate": 2.3350648470817413e-05,
+      "loss": 2.8726,
+      "step": 7072
+    },
+    {
+      "epoch": 0.905344,
+      "grad_norm": 0.42449426651000977,
+      "learning_rate": 2.328807545167838e-05,
+      "loss": 2.8506,
+      "step": 7073
+    },
+    {
+      "epoch": 0.905472,
+      "grad_norm": 0.4082895517349243,
+      "learning_rate": 2.3225584384845845e-05,
+      "loss": 2.777,
+      "step": 7074
+    },
+    {
+      "epoch": 0.9056,
+      "grad_norm": 0.41488415002822876,
+      "learning_rate": 2.3163175281062544e-05,
+      "loss": 2.8437,
+      "step": 7075
+    },
+    {
+      "epoch": 0.905728,
+      "grad_norm": 0.4289550185203552,
+      "learning_rate": 2.3100848151057395e-05,
+      "loss": 2.6924,
+      "step": 7076
+    },
+    {
+      "epoch": 0.905856,
+      "grad_norm": 0.41728144884109497,
+      "learning_rate": 2.3038603005545113e-05,
+      "loss": 2.8135,
+      "step": 7077
+    },
+    {
+      "epoch": 0.905984,
+      "grad_norm": 0.4337783753871918,
+      "learning_rate": 2.2976439855226404e-05,
+      "loss": 2.8403,
+      "step": 7078
+    },
+    {
+      "epoch": 0.906112,
+      "grad_norm": 0.41463199257850647,
+      "learning_rate": 2.291435871078784e-05,
+      "loss": 2.8221,
+      "step": 7079
+    },
+    {
+      "epoch": 0.90624,
+      "grad_norm": 0.4096919298171997,
+      "learning_rate": 2.2852359582901826e-05,
+      "loss": 2.8477,
+      "step": 7080
+    },
+    {
+      "epoch": 0.906368,
+      "grad_norm": 0.40560415387153625,
+      "learning_rate": 2.2790442482226725e-05,
+      "loss": 2.7059,
+      "step": 7081
+    },
+    {
+      "epoch": 0.906496,
+      "grad_norm": 0.46525484323501587,
+      "learning_rate": 2.2728607419406965e-05,
+      "loss": 2.966,
+      "step": 7082
+    },
+    {
+      "epoch": 0.906624,
+      "grad_norm": 0.41826045513153076,
+      "learning_rate": 2.2666854405072547e-05,
+      "loss": 2.7623,
+      "step": 7083
+    },
+    {
+      "epoch": 0.906752,
+      "grad_norm": 0.3993058502674103,
+      "learning_rate": 2.2605183449839583e-05,
+      "loss": 2.9168,
+      "step": 7084
+    },
+    {
+      "epoch": 0.90688,
+      "grad_norm": 0.4144076704978943,
+      "learning_rate": 2.2543594564309988e-05,
+      "loss": 2.7913,
+      "step": 7085
+    },
+    {
+      "epoch": 0.907008,
+      "grad_norm": 0.42177870869636536,
+      "learning_rate": 2.2482087759071622e-05,
+      "loss": 2.6417,
+      "step": 7086
+    },
+    {
+      "epoch": 0.907136,
+      "grad_norm": 0.41166964173316956,
+      "learning_rate": 2.2420663044698253e-05,
+      "loss": 2.806,
+      "step": 7087
+    },
+    {
+      "epoch": 0.907264,
+      "grad_norm": 0.3960515260696411,
+      "learning_rate": 2.235932043174943e-05,
+      "loss": 2.733,
+      "step": 7088
+    },
+    {
+      "epoch": 0.907392,
+      "grad_norm": 0.39144355058670044,
+      "learning_rate": 2.229805993077083e-05,
+      "loss": 2.6831,
+      "step": 7089
+    },
+    {
+      "epoch": 0.90752,
+      "grad_norm": 0.43056443333625793,
+      "learning_rate": 2.2236881552293642e-05,
+      "loss": 2.7029,
+      "step": 7090
+    },
+    {
+      "epoch": 0.907648,
+      "grad_norm": 0.4176972508430481,
+      "learning_rate": 2.2175785306835284e-05,
+      "loss": 2.7003,
+      "step": 7091
+    },
+    {
+      "epoch": 0.907776,
+      "grad_norm": 0.41384202241897583,
+      "learning_rate": 2.21147712048988e-05,
+      "loss": 2.6699,
+      "step": 7092
+    },
+    {
+      "epoch": 0.907904,
+      "grad_norm": 0.4692571461200714,
+      "learning_rate": 2.2053839256973295e-05,
+      "loss": 2.7407,
+      "step": 7093
+    },
+    {
+      "epoch": 0.908032,
+      "grad_norm": 0.41165271401405334,
+      "learning_rate": 2.1992989473533563e-05,
+      "loss": 2.7041,
+      "step": 7094
+    },
+    {
+      "epoch": 0.90816,
+      "grad_norm": 0.4181007742881775,
+      "learning_rate": 2.193222186504057e-05,
+      "loss": 2.6405,
+      "step": 7095
+    },
+    {
+      "epoch": 0.908288,
+      "grad_norm": 0.38133928179740906,
+      "learning_rate": 2.1871536441940786e-05,
+      "loss": 2.7771,
+      "step": 7096
+    },
+    {
+      "epoch": 0.908416,
+      "grad_norm": 0.4051997661590576,
+      "learning_rate": 2.181093321466693e-05,
+      "loss": 2.8183,
+      "step": 7097
+    },
+    {
+      "epoch": 0.908544,
+      "grad_norm": 0.41303715109825134,
+      "learning_rate": 2.1750412193637214e-05,
+      "loss": 2.699,
+      "step": 7098
+    },
+    {
+      "epoch": 0.908672,
+      "grad_norm": 0.3939874768257141,
+      "learning_rate": 2.1689973389256044e-05,
+      "loss": 2.7715,
+      "step": 7099
+    },
+    {
+      "epoch": 0.9088,
+      "grad_norm": 0.39404669404029846,
+      "learning_rate": 2.16296168119135e-05,
+      "loss": 2.7823,
+      "step": 7100
+    },
+    {
+      "epoch": 0.908928,
+      "grad_norm": 0.41110822558403015,
+      "learning_rate": 2.1569342471985554e-05,
+      "loss": 2.7765,
+      "step": 7101
+    },
+    {
+      "epoch": 0.909056,
+      "grad_norm": 0.43581777811050415,
+      "learning_rate": 2.150915037983403e-05,
+      "loss": 2.7577,
+      "step": 7102
+    },
+    {
+      "epoch": 0.909184,
+      "grad_norm": 0.41508913040161133,
+      "learning_rate": 2.1449040545806764e-05,
+      "loss": 2.8568,
+      "step": 7103
+    },
+    {
+      "epoch": 0.909312,
+      "grad_norm": 0.44722750782966614,
+      "learning_rate": 2.1389012980237265e-05,
+      "loss": 2.7652,
+      "step": 7104
+    },
+    {
+      "epoch": 0.90944,
+      "grad_norm": 0.4826449453830719,
+      "learning_rate": 2.1329067693444893e-05,
+      "loss": 2.7066,
+      "step": 7105
+    },
+    {
+      "epoch": 0.909568,
+      "grad_norm": 0.3969660699367523,
+      "learning_rate": 2.1269204695735066e-05,
+      "loss": 2.7575,
+      "step": 7106
+    },
+    {
+      "epoch": 0.909696,
+      "grad_norm": 0.42177823185920715,
+      "learning_rate": 2.120942399739889e-05,
+      "loss": 2.6187,
+      "step": 7107
+    },
+    {
+      "epoch": 0.909824,
+      "grad_norm": 0.4250052571296692,
+      "learning_rate": 2.1149725608713367e-05,
+      "loss": 2.8753,
+      "step": 7108
+    },
+    {
+      "epoch": 0.909952,
+      "grad_norm": 0.43929198384284973,
+      "learning_rate": 2.1090109539941228e-05,
+      "loss": 2.9991,
+      "step": 7109
+    },
+    {
+      "epoch": 0.91008,
+      "grad_norm": 0.38322383165359497,
+      "learning_rate": 2.1030575801331332e-05,
+      "loss": 2.733,
+      "step": 7110
+    },
+    {
+      "epoch": 0.910208,
+      "grad_norm": 0.4153723120689392,
+      "learning_rate": 2.0971124403118157e-05,
+      "loss": 2.7635,
+      "step": 7111
+    },
+    {
+      "epoch": 0.910336,
+      "grad_norm": 0.3886135220527649,
+      "learning_rate": 2.091175535552209e-05,
+      "loss": 2.7616,
+      "step": 7112
+    },
+    {
+      "epoch": 0.910464,
+      "grad_norm": 0.4253280460834503,
+      "learning_rate": 2.0852468668749293e-05,
+      "loss": 2.7686,
+      "step": 7113
+    },
+    {
+      "epoch": 0.910592,
+      "grad_norm": 0.39962783455848694,
+      "learning_rate": 2.079326435299189e-05,
+      "loss": 2.7356,
+      "step": 7114
+    },
+    {
+      "epoch": 0.91072,
+      "grad_norm": 0.38672447204589844,
+      "learning_rate": 2.0734142418427805e-05,
+      "loss": 2.8434,
+      "step": 7115
+    },
+    {
+      "epoch": 0.910848,
+      "grad_norm": 0.4167063236236572,
+      "learning_rate": 2.067510287522073e-05,
+      "loss": 2.8536,
+      "step": 7116
+    },
+    {
+      "epoch": 0.910976,
+      "grad_norm": 0.43143826723098755,
+      "learning_rate": 2.0616145733520274e-05,
+      "loss": 2.7452,
+      "step": 7117
+    },
+    {
+      "epoch": 0.911104,
+      "grad_norm": 0.4103332757949829,
+      "learning_rate": 2.055727100346194e-05,
+      "loss": 2.8851,
+      "step": 7118
+    },
+    {
+      "epoch": 0.911232,
+      "grad_norm": 0.43618515133857727,
+      "learning_rate": 2.0498478695166855e-05,
+      "loss": 2.7644,
+      "step": 7119
+    },
+    {
+      "epoch": 0.91136,
+      "grad_norm": 0.42747169733047485,
+      "learning_rate": 2.0439768818742155e-05,
+      "loss": 2.7505,
+      "step": 7120
+    },
+    {
+      "epoch": 0.911488,
+      "grad_norm": 0.4305507242679596,
+      "learning_rate": 2.038114138428071e-05,
+      "loss": 2.7861,
+      "step": 7121
+    },
+    {
+      "epoch": 0.911616,
+      "grad_norm": 0.41690436005592346,
+      "learning_rate": 2.0322596401861294e-05,
+      "loss": 2.8686,
+      "step": 7122
+    },
+    {
+      "epoch": 0.911744,
+      "grad_norm": 0.38322651386260986,
+      "learning_rate": 2.0264133881548406e-05,
+      "loss": 2.7745,
+      "step": 7123
+    },
+    {
+      "epoch": 0.911872,
+      "grad_norm": 0.42225921154022217,
+      "learning_rate": 2.020575383339246e-05,
+      "loss": 2.78,
+      "step": 7124
+    },
+    {
+      "epoch": 0.912,
+      "grad_norm": 0.3863818049430847,
+      "learning_rate": 2.0147456267429755e-05,
+      "loss": 2.7185,
+      "step": 7125
+    },
+    {
+      "epoch": 0.912128,
+      "grad_norm": 0.38823267817497253,
+      "learning_rate": 2.0089241193682272e-05,
+      "loss": 2.7251,
+      "step": 7126
+    },
+    {
+      "epoch": 0.912256,
+      "grad_norm": 0.4257669448852539,
+      "learning_rate": 2.003110862215779e-05,
+      "loss": 2.8694,
+      "step": 7127
+    },
+    {
+      "epoch": 0.912384,
+      "grad_norm": 0.4594736099243164,
+      "learning_rate": 1.997305856285003e-05,
+      "loss": 2.8519,
+      "step": 7128
+    },
+    {
+      "epoch": 0.912512,
+      "grad_norm": 0.45102712512016296,
+      "learning_rate": 1.9915091025738462e-05,
+      "loss": 2.7818,
+      "step": 7129
+    },
+    {
+      "epoch": 0.91264,
+      "grad_norm": 0.4257560670375824,
+      "learning_rate": 1.9857206020788276e-05,
+      "loss": 2.7761,
+      "step": 7130
+    },
+    {
+      "epoch": 0.912768,
+      "grad_norm": 0.37905576825141907,
+      "learning_rate": 1.979940355795079e-05,
+      "loss": 2.6776,
+      "step": 7131
+    },
+    {
+      "epoch": 0.912896,
+      "grad_norm": 0.4240007698535919,
+      "learning_rate": 1.9741683647162724e-05,
+      "loss": 2.6867,
+      "step": 7132
+    },
+    {
+      "epoch": 0.913024,
+      "grad_norm": 0.4231319725513458,
+      "learning_rate": 1.9684046298346857e-05,
+      "loss": 2.881,
+      "step": 7133
+    },
+    {
+      "epoch": 0.913152,
+      "grad_norm": 0.42667970061302185,
+      "learning_rate": 1.9626491521411772e-05,
+      "loss": 2.8541,
+      "step": 7134
+    },
+    {
+      "epoch": 0.91328,
+      "grad_norm": 0.4071486294269562,
+      "learning_rate": 1.956901932625177e-05,
+      "loss": 2.7775,
+      "step": 7135
+    },
+    {
+      "epoch": 0.913408,
+      "grad_norm": 0.3966268002986908,
+      "learning_rate": 1.9511629722747004e-05,
+      "loss": 2.6413,
+      "step": 7136
+    },
+    {
+      "epoch": 0.913536,
+      "grad_norm": 0.39416077733039856,
+      "learning_rate": 1.9454322720763362e-05,
+      "loss": 2.8499,
+      "step": 7137
+    },
+    {
+      "epoch": 0.913664,
+      "grad_norm": 0.41405025124549866,
+      "learning_rate": 1.939709833015252e-05,
+      "loss": 2.7456,
+      "step": 7138
+    },
+    {
+      "epoch": 0.913792,
+      "grad_norm": 0.41042906045913696,
+      "learning_rate": 1.9339956560752214e-05,
+      "loss": 2.7561,
+      "step": 7139
+    },
+    {
+      "epoch": 0.91392,
+      "grad_norm": 0.43512094020843506,
+      "learning_rate": 1.9282897422385593e-05,
+      "loss": 2.6796,
+      "step": 7140
+    },
+    {
+      "epoch": 0.914048,
+      "grad_norm": 0.39255842566490173,
+      "learning_rate": 1.922592092486192e-05,
+      "loss": 2.7264,
+      "step": 7141
+    },
+    {
+      "epoch": 0.914176,
+      "grad_norm": 0.441133588552475,
+      "learning_rate": 1.9169027077975965e-05,
+      "loss": 2.7932,
+      "step": 7142
+    },
+    {
+      "epoch": 0.914304,
+      "grad_norm": 0.3981904685497284,
+      "learning_rate": 1.9112215891508632e-05,
+      "loss": 2.8212,
+      "step": 7143
+    },
+    {
+      "epoch": 0.914432,
+      "grad_norm": 0.41303545236587524,
+      "learning_rate": 1.9055487375226277e-05,
+      "loss": 2.7517,
+      "step": 7144
+    },
+    {
+      "epoch": 0.91456,
+      "grad_norm": 0.4455587863922119,
+      "learning_rate": 1.8998841538881152e-05,
+      "loss": 2.7617,
+      "step": 7145
+    },
+    {
+      "epoch": 0.914688,
+      "grad_norm": 0.43179696798324585,
+      "learning_rate": 1.8942278392211464e-05,
+      "loss": 2.7446,
+      "step": 7146
+    },
+    {
+      "epoch": 0.914816,
+      "grad_norm": 0.39787986874580383,
+      "learning_rate": 1.8885797944941052e-05,
+      "loss": 2.7688,
+      "step": 7147
+    },
+    {
+      "epoch": 0.914944,
+      "grad_norm": 0.4000432789325714,
+      "learning_rate": 1.8829400206779536e-05,
+      "loss": 2.7382,
+      "step": 7148
+    },
+    {
+      "epoch": 0.915072,
+      "grad_norm": 0.4270205497741699,
+      "learning_rate": 1.8773085187422323e-05,
+      "loss": 2.9054,
+      "step": 7149
+    },
+    {
+      "epoch": 0.9152,
+      "grad_norm": 0.3823355436325073,
+      "learning_rate": 1.8716852896550618e-05,
+      "loss": 2.7201,
+      "step": 7150
+    },
+    {
+      "epoch": 0.915328,
+      "grad_norm": 0.42487236857414246,
+      "learning_rate": 1.866070334383135e-05,
+      "loss": 2.8573,
+      "step": 7151
+    },
+    {
+      "epoch": 0.915456,
+      "grad_norm": 0.41847267746925354,
+      "learning_rate": 1.8604636538917364e-05,
+      "loss": 2.7226,
+      "step": 7152
+    },
+    {
+      "epoch": 0.915584,
+      "grad_norm": 0.38036295771598816,
+      "learning_rate": 1.8548652491447214e-05,
+      "loss": 2.6847,
+      "step": 7153
+    },
+    {
+      "epoch": 0.915712,
+      "grad_norm": 0.4161358177661896,
+      "learning_rate": 1.8492751211045157e-05,
+      "loss": 2.9169,
+      "step": 7154
+    },
+    {
+      "epoch": 0.91584,
+      "grad_norm": 0.4145375192165375,
+      "learning_rate": 1.8436932707321273e-05,
+      "loss": 2.83,
+      "step": 7155
+    },
+    {
+      "epoch": 0.915968,
+      "grad_norm": 0.3965677320957184,
+      "learning_rate": 1.8381196989871452e-05,
+      "loss": 2.6839,
+      "step": 7156
+    },
+    {
+      "epoch": 0.916096,
+      "grad_norm": 0.39585432410240173,
+      "learning_rate": 1.8325544068277245e-05,
+      "loss": 2.8159,
+      "step": 7157
+    },
+    {
+      "epoch": 0.916224,
+      "grad_norm": 0.37404412031173706,
+      "learning_rate": 1.8269973952106055e-05,
+      "loss": 2.6362,
+      "step": 7158
+    },
+    {
+      "epoch": 0.916352,
+      "grad_norm": 0.4056735038757324,
+      "learning_rate": 1.821448665091102e-05,
+      "loss": 2.815,
+      "step": 7159
+    },
+    {
+      "epoch": 0.91648,
+      "grad_norm": 0.44898033142089844,
+      "learning_rate": 1.8159082174231012e-05,
+      "loss": 2.799,
+      "step": 7160
+    },
+    {
+      "epoch": 0.916608,
+      "grad_norm": 0.44519680738449097,
+      "learning_rate": 1.8103760531590852e-05,
+      "loss": 2.8065,
+      "step": 7161
+    },
+    {
+      "epoch": 0.916736,
+      "grad_norm": 0.4142875373363495,
+      "learning_rate": 1.8048521732500878e-05,
+      "loss": 2.7332,
+      "step": 7162
+    },
+    {
+      "epoch": 0.916864,
+      "grad_norm": 0.42132917046546936,
+      "learning_rate": 1.7993365786457217e-05,
+      "loss": 2.6635,
+      "step": 7163
+    },
+    {
+      "epoch": 0.916992,
+      "grad_norm": 0.4064347445964813,
+      "learning_rate": 1.793829270294195e-05,
+      "loss": 2.8464,
+      "step": 7164
+    },
+    {
+      "epoch": 0.91712,
+      "grad_norm": 0.4211523234844208,
+      "learning_rate": 1.7883302491422672e-05,
+      "loss": 2.6938,
+      "step": 7165
+    },
+    {
+      "epoch": 0.917248,
+      "grad_norm": 0.39099180698394775,
+      "learning_rate": 1.7828395161352817e-05,
+      "loss": 2.7225,
+      "step": 7166
+    },
+    {
+      "epoch": 0.917376,
+      "grad_norm": 0.3896682560443878,
+      "learning_rate": 1.777357072217173e-05,
+      "loss": 2.7081,
+      "step": 7167
+    },
+    {
+      "epoch": 0.917504,
+      "grad_norm": 0.4372454285621643,
+      "learning_rate": 1.7718829183304252e-05,
+      "loss": 2.6845,
+      "step": 7168
+    },
+    {
+      "epoch": 0.917632,
+      "grad_norm": 0.4291950464248657,
+      "learning_rate": 1.7664170554161142e-05,
+      "loss": 2.672,
+      "step": 7169
+    },
+    {
+      "epoch": 0.91776,
+      "grad_norm": 0.41059446334838867,
+      "learning_rate": 1.7609594844138766e-05,
+      "loss": 2.8106,
+      "step": 7170
+    },
+    {
+      "epoch": 0.917888,
+      "grad_norm": 0.4217793047428131,
+      "learning_rate": 1.7555102062619455e-05,
+      "loss": 2.752,
+      "step": 7171
+    },
+    {
+      "epoch": 0.918016,
+      "grad_norm": 0.394521027803421,
+      "learning_rate": 1.750069221897105e-05,
+      "loss": 2.7026,
+      "step": 7172
+    },
+    {
+      "epoch": 0.918144,
+      "grad_norm": 0.43452125787734985,
+      "learning_rate": 1.744636532254723e-05,
+      "loss": 2.8548,
+      "step": 7173
+    },
+    {
+      "epoch": 0.918272,
+      "grad_norm": 0.4003255069255829,
+      "learning_rate": 1.739212138268753e-05,
+      "loss": 2.7666,
+      "step": 7174
+    },
+    {
+      "epoch": 0.9184,
+      "grad_norm": 0.4303540587425232,
+      "learning_rate": 1.733796040871699e-05,
+      "loss": 2.8978,
+      "step": 7175
+    },
+    {
+      "epoch": 0.918528,
+      "grad_norm": 0.4182627499103546,
+      "learning_rate": 1.7283882409946548e-05,
+      "loss": 2.8499,
+      "step": 7176
+    },
+    {
+      "epoch": 0.918656,
+      "grad_norm": 0.39889100193977356,
+      "learning_rate": 1.722988739567288e-05,
+      "loss": 2.8664,
+      "step": 7177
+    },
+    {
+      "epoch": 0.918784,
+      "grad_norm": 0.5440437197685242,
+      "learning_rate": 1.7175975375178344e-05,
+      "loss": 2.8087,
+      "step": 7178
+    },
+    {
+      "epoch": 0.918912,
+      "grad_norm": 0.41578084230422974,
+      "learning_rate": 1.7122146357730907e-05,
+      "loss": 2.8947,
+      "step": 7179
+    },
+    {
+      "epoch": 0.91904,
+      "grad_norm": 0.44754934310913086,
+      "learning_rate": 1.706840035258461e-05,
+      "loss": 2.8444,
+      "step": 7180
+    },
+    {
+      "epoch": 0.919168,
+      "grad_norm": 0.3931328058242798,
+      "learning_rate": 1.7014737368978793e-05,
+      "loss": 2.6685,
+      "step": 7181
+    },
+    {
+      "epoch": 0.919296,
+      "grad_norm": 0.3737390339374542,
+      "learning_rate": 1.696115741613902e-05,
+      "loss": 2.8333,
+      "step": 7182
+    },
+    {
+      "epoch": 0.919424,
+      "grad_norm": 0.44700300693511963,
+      "learning_rate": 1.690766050327608e-05,
+      "loss": 2.8473,
+      "step": 7183
+    },
+    {
+      "epoch": 0.919552,
+      "grad_norm": 0.4073667526245117,
+      "learning_rate": 1.685424663958679e-05,
+      "loss": 2.816,
+      "step": 7184
+    },
+    {
+      "epoch": 0.91968,
+      "grad_norm": 0.3893396258354187,
+      "learning_rate": 1.6800915834253637e-05,
+      "loss": 2.6965,
+      "step": 7185
+    },
+    {
+      "epoch": 0.919808,
+      "grad_norm": 0.4009106755256653,
+      "learning_rate": 1.674766809644479e-05,
+      "loss": 2.8635,
+      "step": 7186
+    },
+    {
+      "epoch": 0.919936,
+      "grad_norm": 0.43492624163627625,
+      "learning_rate": 1.6694503435314036e-05,
+      "loss": 2.8228,
+      "step": 7187
+    },
+    {
+      "epoch": 0.920064,
+      "grad_norm": 0.4108307957649231,
+      "learning_rate": 1.664142186000117e-05,
+      "loss": 2.7957,
+      "step": 7188
+    },
+    {
+      "epoch": 0.920192,
+      "grad_norm": 0.3924741744995117,
+      "learning_rate": 1.6588423379631458e-05,
+      "loss": 2.8201,
+      "step": 7189
+    },
+    {
+      "epoch": 0.92032,
+      "grad_norm": 0.38273847103118896,
+      "learning_rate": 1.6535508003315937e-05,
+      "loss": 2.6914,
+      "step": 7190
+    },
+    {
+      "epoch": 0.920448,
+      "grad_norm": 0.40600383281707764,
+      "learning_rate": 1.6482675740151442e-05,
+      "loss": 2.8135,
+      "step": 7191
+    },
+    {
+      "epoch": 0.920576,
+      "grad_norm": 0.4102696478366852,
+      "learning_rate": 1.6429926599220378e-05,
+      "loss": 2.9285,
+      "step": 7192
+    },
+    {
+      "epoch": 0.920704,
+      "grad_norm": 0.4533528983592987,
+      "learning_rate": 1.6377260589590938e-05,
+      "loss": 2.7375,
+      "step": 7193
+    },
+    {
+      "epoch": 0.920832,
+      "grad_norm": 0.3682515323162079,
+      "learning_rate": 1.632467772031704e-05,
+      "loss": 2.7502,
+      "step": 7194
+    },
+    {
+      "epoch": 0.92096,
+      "grad_norm": 0.39543160796165466,
+      "learning_rate": 1.627217800043829e-05,
+      "loss": 2.7883,
+      "step": 7195
+    },
+    {
+      "epoch": 0.921088,
+      "grad_norm": 0.3990235924720764,
+      "learning_rate": 1.621976143898002e-05,
+      "loss": 2.7269,
+      "step": 7196
+    },
+    {
+      "epoch": 0.921216,
+      "grad_norm": 0.4002995789051056,
+      "learning_rate": 1.6167428044953137e-05,
+      "loss": 2.7942,
+      "step": 7197
+    },
+    {
+      "epoch": 0.921344,
+      "grad_norm": 0.4004457890987396,
+      "learning_rate": 1.6115177827354554e-05,
+      "loss": 2.5687,
+      "step": 7198
+    },
+    {
+      "epoch": 0.921472,
+      "grad_norm": 0.3825281262397766,
+      "learning_rate": 1.6063010795166534e-05,
+      "loss": 2.7725,
+      "step": 7199
+    },
+    {
+      "epoch": 0.9216,
+      "grad_norm": 0.36551782488822937,
+      "learning_rate": 1.601092695735723e-05,
+      "loss": 2.8422,
+      "step": 7200
+    },
+    {
+      "epoch": 0.921728,
+      "grad_norm": 0.42287877202033997,
+      "learning_rate": 1.5958926322880484e-05,
+      "loss": 2.6411,
+      "step": 7201
+    },
+    {
+      "epoch": 0.921856,
+      "grad_norm": 0.39010685682296753,
+      "learning_rate": 1.5907008900675758e-05,
+      "loss": 2.8241,
+      "step": 7202
+    },
+    {
+      "epoch": 0.921984,
+      "grad_norm": 0.44504180550575256,
+      "learning_rate": 1.5855174699668296e-05,
+      "loss": 2.8007,
+      "step": 7203
+    },
+    {
+      "epoch": 0.922112,
+      "grad_norm": 0.3544329106807709,
+      "learning_rate": 1.580342372876903e-05,
+      "loss": 2.7792,
+      "step": 7204
+    },
+    {
+      "epoch": 0.92224,
+      "grad_norm": 0.39184871315956116,
+      "learning_rate": 1.5751755996874452e-05,
+      "loss": 2.7147,
+      "step": 7205
+    },
+    {
+      "epoch": 0.922368,
+      "grad_norm": 0.4221350848674774,
+      "learning_rate": 1.5700171512866957e-05,
+      "loss": 2.8158,
+      "step": 7206
+    },
+    {
+      "epoch": 0.922496,
+      "grad_norm": 0.4463709592819214,
+      "learning_rate": 1.5648670285614396e-05,
+      "loss": 2.8418,
+      "step": 7207
+    },
+    {
+      "epoch": 0.922624,
+      "grad_norm": 0.42480969429016113,
+      "learning_rate": 1.559725232397058e-05,
+      "loss": 2.6969,
+      "step": 7208
+    },
+    {
+      "epoch": 0.922752,
+      "grad_norm": 0.3931539058685303,
+      "learning_rate": 1.5545917636774653e-05,
+      "loss": 2.9302,
+      "step": 7209
+    },
+    {
+      "epoch": 0.92288,
+      "grad_norm": 0.42677679657936096,
+      "learning_rate": 1.5494666232851894e-05,
+      "loss": 2.9228,
+      "step": 7210
+    },
+    {
+      "epoch": 0.923008,
+      "grad_norm": 0.3918991982936859,
+      "learning_rate": 1.5443498121012812e-05,
+      "loss": 2.6645,
+      "step": 7211
+    },
+    {
+      "epoch": 0.923136,
+      "grad_norm": 0.4148939251899719,
+      "learning_rate": 1.5392413310053867e-05,
+      "loss": 2.9092,
+      "step": 7212
+    },
+    {
+      "epoch": 0.923264,
+      "grad_norm": 0.39794883131980896,
+      "learning_rate": 1.5341411808757143e-05,
+      "loss": 2.7219,
+      "step": 7213
+    },
+    {
+      "epoch": 0.923392,
+      "grad_norm": 0.4906368851661682,
+      "learning_rate": 1.529049362589041e-05,
+      "loss": 2.812,
+      "step": 7214
+    },
+    {
+      "epoch": 0.92352,
+      "grad_norm": 0.3929259181022644,
+      "learning_rate": 1.5239658770206943e-05,
+      "loss": 2.5863,
+      "step": 7215
+    },
+    {
+      "epoch": 0.923648,
+      "grad_norm": 0.3759031295776367,
+      "learning_rate": 1.5188907250446027e-05,
+      "loss": 2.6823,
+      "step": 7216
+    },
+    {
+      "epoch": 0.923776,
+      "grad_norm": 0.38156044483184814,
+      "learning_rate": 1.513823907533235e-05,
+      "loss": 2.8248,
+      "step": 7217
+    },
+    {
+      "epoch": 0.923904,
+      "grad_norm": 0.4121381640434265,
+      "learning_rate": 1.5087654253576388e-05,
+      "loss": 2.8337,
+      "step": 7218
+    },
+    {
+      "epoch": 0.924032,
+      "grad_norm": 0.39489078521728516,
+      "learning_rate": 1.5037152793874243e-05,
+      "loss": 2.6425,
+      "step": 7219
+    },
+    {
+      "epoch": 0.92416,
+      "grad_norm": 0.4143836498260498,
+      "learning_rate": 1.4986734704907745e-05,
+      "loss": 2.8274,
+      "step": 7220
+    },
+    {
+      "epoch": 0.924288,
+      "grad_norm": 0.4034320116043091,
+      "learning_rate": 1.4936399995344297e-05,
+      "loss": 2.6582,
+      "step": 7221
+    },
+    {
+      "epoch": 0.924416,
+      "grad_norm": 0.39877137541770935,
+      "learning_rate": 1.4886148673836974e-05,
+      "loss": 2.7966,
+      "step": 7222
+    },
+    {
+      "epoch": 0.924544,
+      "grad_norm": 0.4215317964553833,
+      "learning_rate": 1.4835980749024591e-05,
+      "loss": 2.8201,
+      "step": 7223
+    },
+    {
+      "epoch": 0.924672,
+      "grad_norm": 0.42389121651649475,
+      "learning_rate": 1.4785896229531693e-05,
+      "loss": 2.749,
+      "step": 7224
+    },
+    {
+      "epoch": 0.9248,
+      "grad_norm": 0.4027198553085327,
+      "learning_rate": 1.473589512396828e-05,
+      "loss": 2.8647,
+      "step": 7225
+    },
+    {
+      "epoch": 0.924928,
+      "grad_norm": 0.41305065155029297,
+      "learning_rate": 1.4685977440930199e-05,
+      "loss": 2.7717,
+      "step": 7226
+    },
+    {
+      "epoch": 0.925056,
+      "grad_norm": 0.39689087867736816,
+      "learning_rate": 1.4636143188998807e-05,
+      "loss": 2.8662,
+      "step": 7227
+    },
+    {
+      "epoch": 0.925184,
+      "grad_norm": 0.39557376503944397,
+      "learning_rate": 1.4586392376741254e-05,
+      "loss": 2.8219,
+      "step": 7228
+    },
+    {
+      "epoch": 0.925312,
+      "grad_norm": 0.42349973320961,
+      "learning_rate": 1.453672501271025e-05,
+      "loss": 2.8394,
+      "step": 7229
+    },
+    {
+      "epoch": 0.92544,
+      "grad_norm": 0.39367491006851196,
+      "learning_rate": 1.4487141105444134e-05,
+      "loss": 2.6939,
+      "step": 7230
+    },
+    {
+      "epoch": 0.925568,
+      "grad_norm": 0.4325856566429138,
+      "learning_rate": 1.4437640663467033e-05,
+      "loss": 2.7928,
+      "step": 7231
+    },
+    {
+      "epoch": 0.925696,
+      "grad_norm": 0.42057546973228455,
+      "learning_rate": 1.4388223695288694e-05,
+      "loss": 2.5821,
+      "step": 7232
+    },
+    {
+      "epoch": 0.925824,
+      "grad_norm": 0.3779714107513428,
+      "learning_rate": 1.433889020940432e-05,
+      "loss": 2.7333,
+      "step": 7233
+    },
+    {
+      "epoch": 0.925952,
+      "grad_norm": 0.45854809880256653,
+      "learning_rate": 1.4289640214294964e-05,
+      "loss": 2.8746,
+      "step": 7234
+    },
+    {
+      "epoch": 0.92608,
+      "grad_norm": 0.4408297538757324,
+      "learning_rate": 1.4240473718427348e-05,
+      "loss": 2.8842,
+      "step": 7235
+    },
+    {
+      "epoch": 0.926208,
+      "grad_norm": 0.40745607018470764,
+      "learning_rate": 1.4191390730253716e-05,
+      "loss": 2.7328,
+      "step": 7236
+    },
+    {
+      "epoch": 0.926336,
+      "grad_norm": 0.39507853984832764,
+      "learning_rate": 1.4142391258211984e-05,
+      "loss": 2.6508,
+      "step": 7237
+    },
+    {
+      "epoch": 0.926464,
+      "grad_norm": 0.42935624718666077,
+      "learning_rate": 1.409347531072569e-05,
+      "loss": 2.8717,
+      "step": 7238
+    },
+    {
+      "epoch": 0.926592,
+      "grad_norm": 0.4189510643482208,
+      "learning_rate": 1.4044642896204107e-05,
+      "loss": 2.7379,
+      "step": 7239
+    },
+    {
+      "epoch": 0.92672,
+      "grad_norm": 0.49101391434669495,
+      "learning_rate": 1.3995894023042133e-05,
+      "loss": 2.7999,
+      "step": 7240
+    },
+    {
+      "epoch": 0.926848,
+      "grad_norm": 0.39815565943717957,
+      "learning_rate": 1.394722869962023e-05,
+      "loss": 2.7484,
+      "step": 7241
+    },
+    {
+      "epoch": 0.926976,
+      "grad_norm": 0.4044831395149231,
+      "learning_rate": 1.3898646934304538e-05,
+      "loss": 2.7182,
+      "step": 7242
+    },
+    {
+      "epoch": 0.927104,
+      "grad_norm": 0.4353567361831665,
+      "learning_rate": 1.3850148735446766e-05,
+      "loss": 2.768,
+      "step": 7243
+    },
+    {
+      "epoch": 0.927232,
+      "grad_norm": 0.3993237018585205,
+      "learning_rate": 1.38017341113843e-05,
+      "loss": 2.7817,
+      "step": 7244
+    },
+    {
+      "epoch": 0.92736,
+      "grad_norm": 0.38064897060394287,
+      "learning_rate": 1.3753403070440262e-05,
+      "loss": 2.6521,
+      "step": 7245
+    },
+    {
+      "epoch": 0.927488,
+      "grad_norm": 0.41111186146736145,
+      "learning_rate": 1.3705155620923337e-05,
+      "loss": 2.9652,
+      "step": 7246
+    },
+    {
+      "epoch": 0.927616,
+      "grad_norm": 0.414444237947464,
+      "learning_rate": 1.365699177112778e-05,
+      "loss": 2.6796,
+      "step": 7247
+    },
+    {
+      "epoch": 0.927744,
+      "grad_norm": 0.4261646568775177,
+      "learning_rate": 1.3608911529333467e-05,
+      "loss": 2.971,
+      "step": 7248
+    },
+    {
+      "epoch": 0.927872,
+      "grad_norm": 0.4166247248649597,
+      "learning_rate": 1.3560914903806065e-05,
+      "loss": 2.659,
+      "step": 7249
+    },
+    {
+      "epoch": 0.928,
+      "grad_norm": 0.41265255212783813,
+      "learning_rate": 1.351300190279664e-05,
+      "loss": 2.8325,
+      "step": 7250
+    },
+    {
+      "epoch": 0.928128,
+      "grad_norm": 0.42533740401268005,
+      "learning_rate": 1.3465172534541937e-05,
+      "loss": 2.9426,
+      "step": 7251
+    },
+    {
+      "epoch": 0.928256,
+      "grad_norm": 0.4006561040878296,
+      "learning_rate": 1.3417426807264599e-05,
+      "loss": 2.8148,
+      "step": 7252
+    },
+    {
+      "epoch": 0.928384,
+      "grad_norm": 0.4050769805908203,
+      "learning_rate": 1.3369764729172451e-05,
+      "loss": 2.7802,
+      "step": 7253
+    },
+    {
+      "epoch": 0.928512,
+      "grad_norm": 0.41979560256004333,
+      "learning_rate": 1.3322186308459271e-05,
+      "loss": 2.7266,
+      "step": 7254
+    },
+    {
+      "epoch": 0.92864,
+      "grad_norm": 0.4357417821884155,
+      "learning_rate": 1.3274691553304353e-05,
+      "loss": 2.8426,
+      "step": 7255
+    },
+    {
+      "epoch": 0.928768,
+      "grad_norm": 0.42834293842315674,
+      "learning_rate": 1.322728047187255e-05,
+      "loss": 2.8868,
+      "step": 7256
+    },
+    {
+      "epoch": 0.928896,
+      "grad_norm": 0.4054630696773529,
+      "learning_rate": 1.3179953072314398e-05,
+      "loss": 2.7333,
+      "step": 7257
+    },
+    {
+      "epoch": 0.929024,
+      "grad_norm": 0.3943823575973511,
+      "learning_rate": 1.3132709362766004e-05,
+      "loss": 2.8149,
+      "step": 7258
+    },
+    {
+      "epoch": 0.929152,
+      "grad_norm": 0.43944865465164185,
+      "learning_rate": 1.3085549351349091e-05,
+      "loss": 2.619,
+      "step": 7259
+    },
+    {
+      "epoch": 0.92928,
+      "grad_norm": 0.397523432970047,
+      "learning_rate": 1.3038473046171061e-05,
+      "loss": 2.7444,
+      "step": 7260
+    },
+    {
+      "epoch": 0.929408,
+      "grad_norm": 0.39965319633483887,
+      "learning_rate": 1.299148045532489e-05,
+      "loss": 2.9066,
+      "step": 7261
+    },
+    {
+      "epoch": 0.929536,
+      "grad_norm": 0.428646057844162,
+      "learning_rate": 1.2944571586888997e-05,
+      "loss": 2.7273,
+      "step": 7262
+    },
+    {
+      "epoch": 0.929664,
+      "grad_norm": 0.3883942663669586,
+      "learning_rate": 1.2897746448927827e-05,
+      "loss": 2.7449,
+      "step": 7263
+    },
+    {
+      "epoch": 0.929792,
+      "grad_norm": 0.41792914271354675,
+      "learning_rate": 1.2851005049490939e-05,
+      "loss": 2.688,
+      "step": 7264
+    },
+    {
+      "epoch": 0.92992,
+      "grad_norm": 0.4304342567920685,
+      "learning_rate": 1.2804347396613847e-05,
+      "loss": 2.9051,
+      "step": 7265
+    },
+    {
+      "epoch": 0.930048,
+      "grad_norm": 0.48928460478782654,
+      "learning_rate": 1.2757773498317416e-05,
+      "loss": 2.7199,
+      "step": 7266
+    },
+    {
+      "epoch": 0.930176,
+      "grad_norm": 0.4157504141330719,
+      "learning_rate": 1.271128336260835e-05,
+      "loss": 2.8819,
+      "step": 7267
+    },
+    {
+      "epoch": 0.930304,
+      "grad_norm": 0.4267504811286926,
+      "learning_rate": 1.266487699747887e-05,
+      "loss": 2.8354,
+      "step": 7268
+    },
+    {
+      "epoch": 0.930432,
+      "grad_norm": 0.4346822500228882,
+      "learning_rate": 1.2618554410906646e-05,
+      "loss": 2.6995,
+      "step": 7269
+    },
+    {
+      "epoch": 0.93056,
+      "grad_norm": 0.3909984529018402,
+      "learning_rate": 1.25723156108552e-05,
+      "loss": 2.7812,
+      "step": 7270
+    },
+    {
+      "epoch": 0.930688,
+      "grad_norm": 0.4172215163707733,
+      "learning_rate": 1.252616060527334e-05,
+      "loss": 2.7454,
+      "step": 7271
+    },
+    {
+      "epoch": 0.930816,
+      "grad_norm": 0.41741228103637695,
+      "learning_rate": 1.248008940209583e-05,
+      "loss": 2.8511,
+      "step": 7272
+    },
+    {
+      "epoch": 0.930944,
+      "grad_norm": 0.40132737159729004,
+      "learning_rate": 1.243410200924272e-05,
+      "loss": 2.9383,
+      "step": 7273
+    },
+    {
+      "epoch": 0.931072,
+      "grad_norm": 0.41785162687301636,
+      "learning_rate": 1.2388198434619802e-05,
+      "loss": 2.7257,
+      "step": 7274
+    },
+    {
+      "epoch": 0.9312,
+      "grad_norm": 0.4163137972354889,
+      "learning_rate": 1.2342378686118538e-05,
+      "loss": 2.7916,
+      "step": 7275
+    },
+    {
+      "epoch": 0.931328,
+      "grad_norm": 0.42961323261260986,
+      "learning_rate": 1.229664277161574e-05,
+      "loss": 2.8974,
+      "step": 7276
+    },
+    {
+      "epoch": 0.931456,
+      "grad_norm": 0.40400633215904236,
+      "learning_rate": 1.2250990698974007e-05,
+      "loss": 2.8431,
+      "step": 7277
+    },
+    {
+      "epoch": 0.931584,
+      "grad_norm": 0.4653286635875702,
+      "learning_rate": 1.220542247604145e-05,
+      "loss": 2.693,
+      "step": 7278
+    },
+    {
+      "epoch": 0.931712,
+      "grad_norm": 0.4788395166397095,
+      "learning_rate": 1.2159938110651803e-05,
+      "loss": 2.7532,
+      "step": 7279
+    },
+    {
+      "epoch": 0.93184,
+      "grad_norm": 0.3919650614261627,
+      "learning_rate": 1.2114537610624254e-05,
+      "loss": 2.5447,
+      "step": 7280
+    },
+    {
+      "epoch": 0.931968,
+      "grad_norm": 0.4149761199951172,
+      "learning_rate": 1.2069220983763729e-05,
+      "loss": 2.8107,
+      "step": 7281
+    },
+    {
+      "epoch": 0.932096,
+      "grad_norm": 0.39184805750846863,
+      "learning_rate": 1.2023988237860717e-05,
+      "loss": 2.8202,
+      "step": 7282
+    },
+    {
+      "epoch": 0.932224,
+      "grad_norm": 0.4344351589679718,
+      "learning_rate": 1.1978839380691275e-05,
+      "loss": 2.7203,
+      "step": 7283
+    },
+    {
+      "epoch": 0.932352,
+      "grad_norm": 0.41809403896331787,
+      "learning_rate": 1.1933774420016974e-05,
+      "loss": 2.7934,
+      "step": 7284
+    },
+    {
+      "epoch": 0.93248,
+      "grad_norm": 0.3967108726501465,
+      "learning_rate": 1.1888793363584948e-05,
+      "loss": 2.7679,
+      "step": 7285
+    },
+    {
+      "epoch": 0.932608,
+      "grad_norm": 0.4224008321762085,
+      "learning_rate": 1.184389621912807e-05,
+      "loss": 2.8424,
+      "step": 7286
+    },
+    {
+      "epoch": 0.932736,
+      "grad_norm": 0.3797345757484436,
+      "learning_rate": 1.1799082994364551e-05,
+      "loss": 2.7463,
+      "step": 7287
+    },
+    {
+      "epoch": 0.932864,
+      "grad_norm": 0.4047430753707886,
+      "learning_rate": 1.17543536969984e-05,
+      "loss": 2.8112,
+      "step": 7288
+    },
+    {
+      "epoch": 0.932992,
+      "grad_norm": 0.4809977412223816,
+      "learning_rate": 1.1709708334719127e-05,
+      "loss": 2.7346,
+      "step": 7289
+    },
+    {
+      "epoch": 0.93312,
+      "grad_norm": 0.40083345770835876,
+      "learning_rate": 1.1665146915201653e-05,
+      "loss": 2.7335,
+      "step": 7290
+    },
+    {
+      "epoch": 0.933248,
+      "grad_norm": 0.41029614210128784,
+      "learning_rate": 1.1620669446106735e-05,
+      "loss": 2.9008,
+      "step": 7291
+    },
+    {
+      "epoch": 0.933376,
+      "grad_norm": 0.46394333243370056,
+      "learning_rate": 1.1576275935080538e-05,
+      "loss": 2.7147,
+      "step": 7292
+    },
+    {
+      "epoch": 0.933504,
+      "grad_norm": 0.39879706501960754,
+      "learning_rate": 1.1531966389754788e-05,
+      "loss": 2.7133,
+      "step": 7293
+    },
+    {
+      "epoch": 0.933632,
+      "grad_norm": 0.43449389934539795,
+      "learning_rate": 1.1487740817746839e-05,
+      "loss": 2.7375,
+      "step": 7294
+    },
+    {
+      "epoch": 0.93376,
+      "grad_norm": 0.41579702496528625,
+      "learning_rate": 1.1443599226659495e-05,
+      "loss": 2.6822,
+      "step": 7295
+    },
+    {
+      "epoch": 0.933888,
+      "grad_norm": 0.4261867105960846,
+      "learning_rate": 1.1399541624081355e-05,
+      "loss": 2.7793,
+      "step": 7296
+    },
+    {
+      "epoch": 0.934016,
+      "grad_norm": 0.42744457721710205,
+      "learning_rate": 1.1355568017586304e-05,
+      "loss": 2.7634,
+      "step": 7297
+    },
+    {
+      "epoch": 0.934144,
+      "grad_norm": 0.44169968366622925,
+      "learning_rate": 1.1311678414734016e-05,
+      "loss": 2.7998,
+      "step": 7298
+    },
+    {
+      "epoch": 0.934272,
+      "grad_norm": 0.4111073613166809,
+      "learning_rate": 1.1267872823069458e-05,
+      "loss": 2.7332,
+      "step": 7299
+    },
+    {
+      "epoch": 0.9344,
+      "grad_norm": 0.40997785329818726,
+      "learning_rate": 1.1224151250123548e-05,
+      "loss": 2.6228,
+      "step": 7300
+    },
+    {
+      "epoch": 0.934528,
+      "grad_norm": 0.41213101148605347,
+      "learning_rate": 1.1180513703412387e-05,
+      "loss": 2.8731,
+      "step": 7301
+    },
+    {
+      "epoch": 0.934656,
+      "grad_norm": 0.4095704257488251,
+      "learning_rate": 1.1136960190437751e-05,
+      "loss": 2.7947,
+      "step": 7302
+    },
+    {
+      "epoch": 0.934784,
+      "grad_norm": 0.43925514817237854,
+      "learning_rate": 1.1093490718687094e-05,
+      "loss": 2.69,
+      "step": 7303
+    },
+    {
+      "epoch": 0.934912,
+      "grad_norm": 0.39995554089546204,
+      "learning_rate": 1.1050105295633272e-05,
+      "loss": 2.9669,
+      "step": 7304
+    },
+    {
+      "epoch": 0.93504,
+      "grad_norm": 0.39841628074645996,
+      "learning_rate": 1.100680392873471e-05,
+      "loss": 2.8845,
+      "step": 7305
+    },
+    {
+      "epoch": 0.935168,
+      "grad_norm": 0.41223978996276855,
+      "learning_rate": 1.0963586625435506e-05,
+      "loss": 2.7856,
+      "step": 7306
+    },
+    {
+      "epoch": 0.935296,
+      "grad_norm": 0.44055646657943726,
+      "learning_rate": 1.092045339316511e-05,
+      "loss": 2.8157,
+      "step": 7307
+    },
+    {
+      "epoch": 0.935424,
+      "grad_norm": 0.3981113135814667,
+      "learning_rate": 1.087740423933864e-05,
+      "loss": 2.6995,
+      "step": 7308
+    },
+    {
+      "epoch": 0.935552,
+      "grad_norm": 0.4658939242362976,
+      "learning_rate": 1.0834439171356846e-05,
+      "loss": 2.86,
+      "step": 7309
+    },
+    {
+      "epoch": 0.93568,
+      "grad_norm": 0.38878440856933594,
+      "learning_rate": 1.0791558196605822e-05,
+      "loss": 2.8707,
+      "step": 7310
+    },
+    {
+      "epoch": 0.935808,
+      "grad_norm": 0.4584125578403473,
+      "learning_rate": 1.0748761322457335e-05,
+      "loss": 2.6576,
+      "step": 7311
+    },
+    {
+      "epoch": 0.935936,
+      "grad_norm": 0.4096093773841858,
+      "learning_rate": 1.0706048556268666e-05,
+      "loss": 2.806,
+      "step": 7312
+    },
+    {
+      "epoch": 0.936064,
+      "grad_norm": 0.4215119779109955,
+      "learning_rate": 1.0663419905382666e-05,
+      "loss": 2.6301,
+      "step": 7313
+    },
+    {
+      "epoch": 0.936192,
+      "grad_norm": 0.4039079546928406,
+      "learning_rate": 1.0620875377127637e-05,
+      "loss": 2.8709,
+      "step": 7314
+    },
+    {
+      "epoch": 0.93632,
+      "grad_norm": 0.48893171548843384,
+      "learning_rate": 1.0578414978817508e-05,
+      "loss": 2.7528,
+      "step": 7315
+    },
+    {
+      "epoch": 0.936448,
+      "grad_norm": 0.40473833680152893,
+      "learning_rate": 1.0536038717751606e-05,
+      "loss": 2.8432,
+      "step": 7316
+    },
+    {
+      "epoch": 0.936576,
+      "grad_norm": 0.442617267370224,
+      "learning_rate": 1.0493746601215104e-05,
+      "loss": 2.8151,
+      "step": 7317
+    },
+    {
+      "epoch": 0.936704,
+      "grad_norm": 0.41716238856315613,
+      "learning_rate": 1.0451538636478353e-05,
+      "loss": 2.8705,
+      "step": 7318
+    },
+    {
+      "epoch": 0.936832,
+      "grad_norm": 0.3930855393409729,
+      "learning_rate": 1.0409414830797493e-05,
+      "loss": 2.6863,
+      "step": 7319
+    },
+    {
+      "epoch": 0.93696,
+      "grad_norm": 0.43080079555511475,
+      "learning_rate": 1.0367375191414064e-05,
+      "loss": 2.7469,
+      "step": 7320
+    },
+    {
+      "epoch": 0.937088,
+      "grad_norm": 0.42484360933303833,
+      "learning_rate": 1.0325419725555119e-05,
+      "loss": 2.7907,
+      "step": 7321
+    },
+    {
+      "epoch": 0.937216,
+      "grad_norm": 0.4193482995033264,
+      "learning_rate": 1.0283548440433333e-05,
+      "loss": 2.8878,
+      "step": 7322
+    },
+    {
+      "epoch": 0.937344,
+      "grad_norm": 0.4071168899536133,
+      "learning_rate": 1.024176134324678e-05,
+      "loss": 2.7463,
+      "step": 7323
+    },
+    {
+      "epoch": 0.937472,
+      "grad_norm": 0.43503621220588684,
+      "learning_rate": 1.0200058441179271e-05,
+      "loss": 2.9101,
+      "step": 7324
+    },
+    {
+      "epoch": 0.9376,
+      "grad_norm": 0.4009076952934265,
+      "learning_rate": 1.015843974139996e-05,
+      "loss": 2.6843,
+      "step": 7325
+    },
+    {
+      "epoch": 0.937728,
+      "grad_norm": 0.4159998297691345,
+      "learning_rate": 1.0116905251063623e-05,
+      "loss": 2.8486,
+      "step": 7326
+    },
+    {
+      "epoch": 0.937856,
+      "grad_norm": 0.3933505713939667,
+      "learning_rate": 1.0075454977310383e-05,
+      "loss": 2.9473,
+      "step": 7327
+    },
+    {
+      "epoch": 0.937984,
+      "grad_norm": 0.40703582763671875,
+      "learning_rate": 1.0034088927266205e-05,
+      "loss": 2.808,
+      "step": 7328
+    },
+    {
+      "epoch": 0.938112,
+      "grad_norm": 0.4318026006221771,
+      "learning_rate": 9.99280710804229e-06,
+      "loss": 2.7499,
+      "step": 7329
+    },
+    {
+      "epoch": 0.93824,
+      "grad_norm": 0.43033817410469055,
+      "learning_rate": 9.951609526735517e-06,
+      "loss": 2.8434,
+      "step": 7330
+    },
+    {
+      "epoch": 0.938368,
+      "grad_norm": 0.4247799813747406,
+      "learning_rate": 9.910496190428163e-06,
+      "loss": 2.7173,
+      "step": 7331
+    },
+    {
+      "epoch": 0.938496,
+      "grad_norm": 0.3950340449810028,
+      "learning_rate": 9.869467106188135e-06,
+      "loss": 2.8572,
+      "step": 7332
+    },
+    {
+      "epoch": 0.938624,
+      "grad_norm": 0.45899397134780884,
+      "learning_rate": 9.828522281068786e-06,
+      "loss": 2.7605,
+      "step": 7333
+    },
+    {
+      "epoch": 0.938752,
+      "grad_norm": 0.42351770401000977,
+      "learning_rate": 9.787661722108987e-06,
+      "loss": 2.8032,
+      "step": 7334
+    },
+    {
+      "epoch": 0.93888,
+      "grad_norm": 0.39472585916519165,
+      "learning_rate": 9.746885436333231e-06,
+      "loss": 2.7079,
+      "step": 7335
+    },
+    {
+      "epoch": 0.939008,
+      "grad_norm": 0.43862205743789673,
+      "learning_rate": 9.706193430751298e-06,
+      "loss": 2.7947,
+      "step": 7336
+    },
+    {
+      "epoch": 0.939136,
+      "grad_norm": 0.4540042281150818,
+      "learning_rate": 9.665585712358705e-06,
+      "loss": 2.9644,
+      "step": 7337
+    },
+    {
+      "epoch": 0.939264,
+      "grad_norm": 0.4060978889465332,
+      "learning_rate": 9.625062288136366e-06,
+      "loss": 2.6784,
+      "step": 7338
+    },
+    {
+      "epoch": 0.939392,
+      "grad_norm": 0.41845664381980896,
+      "learning_rate": 9.584623165050766e-06,
+      "loss": 2.8003,
+      "step": 7339
+    },
+    {
+      "epoch": 0.93952,
+      "grad_norm": 0.41342583298683167,
+      "learning_rate": 9.544268350053842e-06,
+      "loss": 2.7982,
+      "step": 7340
+    },
+    {
+      "epoch": 0.939648,
+      "grad_norm": 0.40891024470329285,
+      "learning_rate": 9.50399785008299e-06,
+      "loss": 2.7671,
+      "step": 7341
+    },
+    {
+      "epoch": 0.939776,
+      "grad_norm": 0.419253408908844,
+      "learning_rate": 9.463811672061284e-06,
+      "loss": 2.7873,
+      "step": 7342
+    },
+    {
+      "epoch": 0.939904,
+      "grad_norm": 0.40253946185112,
+      "learning_rate": 9.423709822897086e-06,
+      "loss": 2.7504,
+      "step": 7343
+    },
+    {
+      "epoch": 0.940032,
+      "grad_norm": 0.42684200406074524,
+      "learning_rate": 9.383692309484381e-06,
+      "loss": 2.9041,
+      "step": 7344
+    },
+    {
+      "epoch": 0.94016,
+      "grad_norm": 0.42028701305389404,
+      "learning_rate": 9.343759138702724e-06,
+      "loss": 2.6949,
+      "step": 7345
+    },
+    {
+      "epoch": 0.940288,
+      "grad_norm": 0.4028686583042145,
+      "learning_rate": 9.303910317417064e-06,
+      "loss": 2.6938,
+      "step": 7346
+    },
+    {
+      "epoch": 0.940416,
+      "grad_norm": 0.3799770772457123,
+      "learning_rate": 9.264145852477868e-06,
+      "loss": 2.7727,
+      "step": 7347
+    },
+    {
+      "epoch": 0.940544,
+      "grad_norm": 0.3792549669742584,
+      "learning_rate": 9.224465750721056e-06,
+      "loss": 2.7597,
+      "step": 7348
+    },
+    {
+      "epoch": 0.940672,
+      "grad_norm": 0.4156973659992218,
+      "learning_rate": 9.184870018968172e-06,
+      "loss": 2.9149,
+      "step": 7349
+    },
+    {
+      "epoch": 0.9408,
+      "grad_norm": 0.42672547698020935,
+      "learning_rate": 9.14535866402616e-06,
+      "loss": 2.9487,
+      "step": 7350
+    },
+    {
+      "epoch": 0.940928,
+      "grad_norm": 0.41203394532203674,
+      "learning_rate": 9.105931692687419e-06,
+      "loss": 2.8865,
+      "step": 7351
+    },
+    {
+      "epoch": 0.941056,
+      "grad_norm": 0.41290971636772156,
+      "learning_rate": 9.066589111729972e-06,
+      "loss": 2.8488,
+      "step": 7352
+    },
+    {
+      "epoch": 0.941184,
+      "grad_norm": 0.5860098004341125,
+      "learning_rate": 9.027330927917243e-06,
+      "loss": 2.7735,
+      "step": 7353
+    },
+    {
+      "epoch": 0.941312,
+      "grad_norm": 0.41649916768074036,
+      "learning_rate": 8.988157147998222e-06,
+      "loss": 2.761,
+      "step": 7354
+    },
+    {
+      "epoch": 0.94144,
+      "grad_norm": 0.39867284893989563,
+      "learning_rate": 8.949067778707187e-06,
+      "loss": 2.7791,
+      "step": 7355
+    },
+    {
+      "epoch": 0.941568,
+      "grad_norm": 0.420900821685791,
+      "learning_rate": 8.910062826764208e-06,
+      "loss": 2.7932,
+      "step": 7356
+    },
+    {
+      "epoch": 0.941696,
+      "grad_norm": 0.44704166054725647,
+      "learning_rate": 8.871142298874646e-06,
+      "loss": 2.8349,
+      "step": 7357
+    },
+    {
+      "epoch": 0.941824,
+      "grad_norm": 0.412776380777359,
+      "learning_rate": 8.832306201729368e-06,
+      "loss": 2.7773,
+      "step": 7358
+    },
+    {
+      "epoch": 0.941952,
+      "grad_norm": 0.4216349124908447,
+      "learning_rate": 8.793554542004756e-06,
+      "loss": 2.7559,
+      "step": 7359
+    },
+    {
+      "epoch": 0.94208,
+      "grad_norm": 0.44509053230285645,
+      "learning_rate": 8.754887326362648e-06,
+      "loss": 2.9203,
+      "step": 7360
+    },
+    {
+      "epoch": 0.942208,
+      "grad_norm": 0.4118908941745758,
+      "learning_rate": 8.71630456145045e-06,
+      "loss": 2.7581,
+      "step": 7361
+    },
+    {
+      "epoch": 0.942336,
+      "grad_norm": 0.442282110452652,
+      "learning_rate": 8.677806253900967e-06,
+      "loss": 2.7675,
+      "step": 7362
+    },
+    {
+      "epoch": 0.942464,
+      "grad_norm": 0.40231284499168396,
+      "learning_rate": 8.639392410332403e-06,
+      "loss": 2.6432,
+      "step": 7363
+    },
+    {
+      "epoch": 0.942592,
+      "grad_norm": 0.4213030934333801,
+      "learning_rate": 8.601063037348644e-06,
+      "loss": 2.7352,
+      "step": 7364
+    },
+    {
+      "epoch": 0.94272,
+      "grad_norm": 0.40977272391319275,
+      "learning_rate": 8.562818141538975e-06,
+      "loss": 2.8889,
+      "step": 7365
+    },
+    {
+      "epoch": 0.942848,
+      "grad_norm": 0.4291535019874573,
+      "learning_rate": 8.52465772947808e-06,
+      "loss": 2.7687,
+      "step": 7366
+    },
+    {
+      "epoch": 0.942976,
+      "grad_norm": 0.3957488238811493,
+      "learning_rate": 8.486581807726157e-06,
+      "loss": 2.7233,
+      "step": 7367
+    },
+    {
+      "epoch": 0.943104,
+      "grad_norm": 0.38603368401527405,
+      "learning_rate": 8.448590382829024e-06,
+      "loss": 2.7347,
+      "step": 7368
+    },
+    {
+      "epoch": 0.943232,
+      "grad_norm": 0.7962201237678528,
+      "learning_rate": 8.410683461317681e-06,
+      "loss": 2.7835,
+      "step": 7369
+    },
+    {
+      "epoch": 0.94336,
+      "grad_norm": 0.4253509044647217,
+      "learning_rate": 8.372861049708858e-06,
+      "loss": 2.7303,
+      "step": 7370
+    },
+    {
+      "epoch": 0.943488,
+      "grad_norm": 0.4319707453250885,
+      "learning_rate": 8.335123154504687e-06,
+      "loss": 2.8695,
+      "step": 7371
+    },
+    {
+      "epoch": 0.943616,
+      "grad_norm": 0.4175657629966736,
+      "learning_rate": 8.297469782192702e-06,
+      "loss": 2.7632,
+      "step": 7372
+    },
+    {
+      "epoch": 0.943744,
+      "grad_norm": 0.427076518535614,
+      "learning_rate": 8.25990093924589e-06,
+      "loss": 2.9591,
+      "step": 7373
+    },
+    {
+      "epoch": 0.943872,
+      "grad_norm": 0.39819011092185974,
+      "learning_rate": 8.222416632122865e-06,
+      "loss": 2.7458,
+      "step": 7374
+    },
+    {
+      "epoch": 0.944,
+      "grad_norm": 0.4584044814109802,
+      "learning_rate": 8.185016867267691e-06,
+      "loss": 2.6726,
+      "step": 7375
+    },
+    {
+      "epoch": 0.944128,
+      "grad_norm": 0.42401599884033203,
+      "learning_rate": 8.147701651109674e-06,
+      "loss": 2.896,
+      "step": 7376
+    },
+    {
+      "epoch": 0.944256,
+      "grad_norm": 0.46189960837364197,
+      "learning_rate": 8.110470990063789e-06,
+      "loss": 2.6771,
+      "step": 7377
+    },
+    {
+      "epoch": 0.944384,
+      "grad_norm": 0.413958340883255,
+      "learning_rate": 8.07332489053042e-06,
+      "loss": 2.7618,
+      "step": 7378
+    },
+    {
+      "epoch": 0.944512,
+      "grad_norm": 0.4119345545768738,
+      "learning_rate": 8.036263358895402e-06,
+      "loss": 2.7581,
+      "step": 7379
+    },
+    {
+      "epoch": 0.94464,
+      "grad_norm": 0.4145139753818512,
+      "learning_rate": 7.99928640152997e-06,
+      "loss": 2.8818,
+      "step": 7380
+    },
+    {
+      "epoch": 0.944768,
+      "grad_norm": 0.39500296115875244,
+      "learning_rate": 7.962394024791042e-06,
+      "loss": 2.7428,
+      "step": 7381
+    },
+    {
+      "epoch": 0.944896,
+      "grad_norm": 0.4550384283065796,
+      "learning_rate": 7.925586235020766e-06,
+      "loss": 2.7972,
+      "step": 7382
+    },
+    {
+      "epoch": 0.945024,
+      "grad_norm": 0.4611116051673889,
+      "learning_rate": 7.888863038546801e-06,
+      "loss": 2.8276,
+      "step": 7383
+    },
+    {
+      "epoch": 0.945152,
+      "grad_norm": 0.4005953371524811,
+      "learning_rate": 7.852224441682376e-06,
+      "loss": 2.7771,
+      "step": 7384
+    },
+    {
+      "epoch": 0.94528,
+      "grad_norm": 0.4127246141433716,
+      "learning_rate": 7.81567045072601e-06,
+      "loss": 2.6315,
+      "step": 7385
+    },
+    {
+      "epoch": 0.945408,
+      "grad_norm": 0.41736242175102234,
+      "learning_rate": 7.779201071961784e-06,
+      "loss": 2.902,
+      "step": 7386
+    },
+    {
+      "epoch": 0.945536,
+      "grad_norm": 0.4266095757484436,
+      "learning_rate": 7.742816311659185e-06,
+      "loss": 2.779,
+      "step": 7387
+    },
+    {
+      "epoch": 0.945664,
+      "grad_norm": 0.4257403612136841,
+      "learning_rate": 7.706516176073208e-06,
+      "loss": 2.7808,
+      "step": 7388
+    },
+    {
+      "epoch": 0.945792,
+      "grad_norm": 0.39379993081092834,
+      "learning_rate": 7.670300671444308e-06,
+      "loss": 2.7095,
+      "step": 7389
+    },
+    {
+      "epoch": 0.94592,
+      "grad_norm": 0.41513267159461975,
+      "learning_rate": 7.634169803998226e-06,
+      "loss": 2.8513,
+      "step": 7390
+    },
+    {
+      "epoch": 0.946048,
+      "grad_norm": 0.4299713969230652,
+      "learning_rate": 7.598123579946381e-06,
+      "loss": 2.7522,
+      "step": 7391
+    },
+    {
+      "epoch": 0.946176,
+      "grad_norm": 0.40940579771995544,
+      "learning_rate": 7.562162005485485e-06,
+      "loss": 2.9811,
+      "step": 7392
+    },
+    {
+      "epoch": 0.946304,
+      "grad_norm": 0.40578699111938477,
+      "learning_rate": 7.5262850867978124e-06,
+      "loss": 2.7997,
+      "step": 7393
+    },
+    {
+      "epoch": 0.946432,
+      "grad_norm": 0.42870965600013733,
+      "learning_rate": 7.490492830050932e-06,
+      "loss": 2.935,
+      "step": 7394
+    },
+    {
+      "epoch": 0.94656,
+      "grad_norm": 0.40247130393981934,
+      "learning_rate": 7.454785241398032e-06,
+      "loss": 2.768,
+      "step": 7395
+    },
+    {
+      "epoch": 0.946688,
+      "grad_norm": 0.40258458256721497,
+      "learning_rate": 7.419162326977591e-06,
+      "loss": 2.7116,
+      "step": 7396
+    },
+    {
+      "epoch": 0.946816,
+      "grad_norm": 0.396810919046402,
+      "learning_rate": 7.383624092913654e-06,
+      "loss": 2.771,
+      "step": 7397
+    },
+    {
+      "epoch": 0.946944,
+      "grad_norm": 0.4090811610221863,
+      "learning_rate": 7.348170545315613e-06,
+      "loss": 2.8112,
+      "step": 7398
+    },
+    {
+      "epoch": 0.947072,
+      "grad_norm": 0.4231657385826111,
+      "learning_rate": 7.312801690278425e-06,
+      "loss": 2.7579,
+      "step": 7399
+    },
+    {
+      "epoch": 0.9472,
+      "grad_norm": 0.39769577980041504,
+      "learning_rate": 7.277517533882283e-06,
+      "loss": 2.7326,
+      "step": 7400
+    },
+    {
+      "epoch": 0.947328,
+      "grad_norm": 0.45373091101646423,
+      "learning_rate": 7.242318082193e-06,
+      "loss": 2.8453,
+      "step": 7401
+    },
+    {
+      "epoch": 0.947456,
+      "grad_norm": 0.40780624747276306,
+      "learning_rate": 7.207203341261792e-06,
+      "loss": 2.79,
+      "step": 7402
+    },
+    {
+      "epoch": 0.947584,
+      "grad_norm": 0.41133320331573486,
+      "learning_rate": 7.172173317125275e-06,
+      "loss": 2.8582,
+      "step": 7403
+    },
+    {
+      "epoch": 0.947712,
+      "grad_norm": 0.41070985794067383,
+      "learning_rate": 7.1372280158055196e-06,
+      "loss": 2.7651,
+      "step": 7404
+    },
+    {
+      "epoch": 0.94784,
+      "grad_norm": 0.4032180905342102,
+      "learning_rate": 7.102367443310054e-06,
+      "loss": 2.8263,
+      "step": 7405
+    },
+    {
+      "epoch": 0.947968,
+      "grad_norm": 0.4281095862388611,
+      "learning_rate": 7.067591605631751e-06,
+      "loss": 2.7224,
+      "step": 7406
+    },
+    {
+      "epoch": 0.948096,
+      "grad_norm": 0.3996802866458893,
+      "learning_rate": 7.032900508749051e-06,
+      "loss": 2.6509,
+      "step": 7407
+    },
+    {
+      "epoch": 0.948224,
+      "grad_norm": 0.3827175796031952,
+      "learning_rate": 6.998294158625684e-06,
+      "loss": 2.8228,
+      "step": 7408
+    },
+    {
+      "epoch": 0.948352,
+      "grad_norm": 0.4010069966316223,
+      "learning_rate": 6.96377256121089e-06,
+      "loss": 2.8194,
+      "step": 7409
+    },
+    {
+      "epoch": 0.94848,
+      "grad_norm": 0.39807942509651184,
+      "learning_rate": 6.9293357224393675e-06,
+      "loss": 2.7828,
+      "step": 7410
+    },
+    {
+      "epoch": 0.948608,
+      "grad_norm": 0.4328831434249878,
+      "learning_rate": 6.894983648231212e-06,
+      "loss": 2.9138,
+      "step": 7411
+    },
+    {
+      "epoch": 0.948736,
+      "grad_norm": 0.3947349190711975,
+      "learning_rate": 6.86071634449198e-06,
+      "loss": 2.7849,
+      "step": 7412
+    },
+    {
+      "epoch": 0.948864,
+      "grad_norm": 0.3990645706653595,
+      "learning_rate": 6.826533817112512e-06,
+      "loss": 2.8239,
+      "step": 7413
+    },
+    {
+      "epoch": 0.948992,
+      "grad_norm": 0.42281782627105713,
+      "learning_rate": 6.792436071969277e-06,
+      "loss": 2.8314,
+      "step": 7414
+    },
+    {
+      "epoch": 0.94912,
+      "grad_norm": 0.39910396933555603,
+      "learning_rate": 6.758423114924028e-06,
+      "loss": 2.8203,
+      "step": 7415
+    },
+    {
+      "epoch": 0.949248,
+      "grad_norm": 0.38140830397605896,
+      "learning_rate": 6.724494951823978e-06,
+      "loss": 2.7067,
+      "step": 7416
+    },
+    {
+      "epoch": 0.949376,
+      "grad_norm": 0.4393472671508789,
+      "learning_rate": 6.6906515885017945e-06,
+      "loss": 2.8642,
+      "step": 7417
+    },
+    {
+      "epoch": 0.949504,
+      "grad_norm": 0.41534239053726196,
+      "learning_rate": 6.656893030775546e-06,
+      "loss": 2.8112,
+      "step": 7418
+    },
+    {
+      "epoch": 0.949632,
+      "grad_norm": 0.41351842880249023,
+      "learning_rate": 6.623219284448645e-06,
+      "loss": 2.8048,
+      "step": 7419
+    },
+    {
+      "epoch": 0.94976,
+      "grad_norm": 0.4480913579463959,
+      "learning_rate": 6.589630355310128e-06,
+      "loss": 2.7092,
+      "step": 7420
+    },
+    {
+      "epoch": 0.949888,
+      "grad_norm": 0.38812255859375,
+      "learning_rate": 6.556126249134209e-06,
+      "loss": 2.7094,
+      "step": 7421
+    },
+    {
+      "epoch": 0.950016,
+      "grad_norm": 0.39170289039611816,
+      "learning_rate": 6.522706971680725e-06,
+      "loss": 2.8176,
+      "step": 7422
+    },
+    {
+      "epoch": 0.950144,
+      "grad_norm": 0.4044884741306305,
+      "learning_rate": 6.489372528694748e-06,
+      "loss": 2.848,
+      "step": 7423
+    },
+    {
+      "epoch": 0.950272,
+      "grad_norm": 0.42475321888923645,
+      "learning_rate": 6.456122925906971e-06,
+      "loss": 2.9364,
+      "step": 7424
+    },
+    {
+      "epoch": 0.9504,
+      "grad_norm": 0.44116756319999695,
+      "learning_rate": 6.422958169033266e-06,
+      "loss": 2.8513,
+      "step": 7425
+    },
+    {
+      "epoch": 0.950528,
+      "grad_norm": 0.43257007002830505,
+      "learning_rate": 6.389878263775128e-06,
+      "loss": 2.6852,
+      "step": 7426
+    },
+    {
+      "epoch": 0.950656,
+      "grad_norm": 0.39165666699409485,
+      "learning_rate": 6.3568832158192865e-06,
+      "loss": 2.7993,
+      "step": 7427
+    },
+    {
+      "epoch": 0.950784,
+      "grad_norm": 0.380571186542511,
+      "learning_rate": 6.323973030838037e-06,
+      "loss": 2.7798,
+      "step": 7428
+    },
+    {
+      "epoch": 0.950912,
+      "grad_norm": 0.4099527597427368,
+      "learning_rate": 6.291147714488965e-06,
+      "loss": 2.854,
+      "step": 7429
+    },
+    {
+      "epoch": 0.95104,
+      "grad_norm": 0.40075573325157166,
+      "learning_rate": 6.258407272415223e-06,
+      "loss": 2.78,
+      "step": 7430
+    },
+    {
+      "epoch": 0.951168,
+      "grad_norm": 0.4037899374961853,
+      "learning_rate": 6.2257517102451975e-06,
+      "loss": 2.7274,
+      "step": 7431
+    },
+    {
+      "epoch": 0.951296,
+      "grad_norm": 0.42156335711479187,
+      "learning_rate": 6.193181033592787e-06,
+      "loss": 2.8113,
+      "step": 7432
+    },
+    {
+      "epoch": 0.951424,
+      "grad_norm": 0.399564266204834,
+      "learning_rate": 6.1606952480572356e-06,
+      "loss": 2.6631,
+      "step": 7433
+    },
+    {
+      "epoch": 0.951552,
+      "grad_norm": 0.42383208870887756,
+      "learning_rate": 6.128294359223296e-06,
+      "loss": 2.7514,
+      "step": 7434
+    },
+    {
+      "epoch": 0.95168,
+      "grad_norm": 0.4180853068828583,
+      "learning_rate": 6.09597837266096e-06,
+      "loss": 2.842,
+      "step": 7435
+    },
+    {
+      "epoch": 0.951808,
+      "grad_norm": 0.4229358434677124,
+      "learning_rate": 6.063747293925781e-06,
+      "loss": 2.6826,
+      "step": 7436
+    },
+    {
+      "epoch": 0.951936,
+      "grad_norm": 0.3921695053577423,
+      "learning_rate": 6.031601128558606e-06,
+      "loss": 2.7544,
+      "step": 7437
+    },
+    {
+      "epoch": 0.952064,
+      "grad_norm": 0.39834630489349365,
+      "learning_rate": 5.999539882085791e-06,
+      "loss": 2.6944,
+      "step": 7438
+    },
+    {
+      "epoch": 0.952192,
+      "grad_norm": 0.3825263977050781,
+      "learning_rate": 5.967563560019041e-06,
+      "loss": 2.9867,
+      "step": 7439
+    },
+    {
+      "epoch": 0.95232,
+      "grad_norm": 0.3934604227542877,
+      "learning_rate": 5.935672167855455e-06,
+      "loss": 2.6795,
+      "step": 7440
+    },
+    {
+      "epoch": 0.952448,
+      "grad_norm": 0.392756849527359,
+      "learning_rate": 5.903865711077483e-06,
+      "loss": 2.8285,
+      "step": 7441
+    },
+    {
+      "epoch": 0.952576,
+      "grad_norm": 0.4342443346977234,
+      "learning_rate": 5.872144195153029e-06,
+      "loss": 2.7461,
+      "step": 7442
+    },
+    {
+      "epoch": 0.952704,
+      "grad_norm": 0.37794339656829834,
+      "learning_rate": 5.840507625535396e-06,
+      "loss": 2.8863,
+      "step": 7443
+    },
+    {
+      "epoch": 0.952832,
+      "grad_norm": 0.4097895920276642,
+      "learning_rate": 5.808956007663291e-06,
+      "loss": 2.7411,
+      "step": 7444
+    },
+    {
+      "epoch": 0.95296,
+      "grad_norm": 0.4010049104690552,
+      "learning_rate": 5.777489346960874e-06,
+      "loss": 2.8317,
+      "step": 7445
+    },
+    {
+      "epoch": 0.953088,
+      "grad_norm": 0.4099483788013458,
+      "learning_rate": 5.746107648837484e-06,
+      "loss": 2.8176,
+      "step": 7446
+    },
+    {
+      "epoch": 0.953216,
+      "grad_norm": 0.4138134717941284,
+      "learning_rate": 5.714810918688085e-06,
+      "loss": 2.7713,
+      "step": 7447
+    },
+    {
+      "epoch": 0.953344,
+      "grad_norm": 0.40161120891571045,
+      "learning_rate": 5.683599161892928e-06,
+      "loss": 2.7267,
+      "step": 7448
+    },
+    {
+      "epoch": 0.953472,
+      "grad_norm": 0.3722423315048218,
+      "learning_rate": 5.65247238381772e-06,
+      "loss": 2.719,
+      "step": 7449
+    },
+    {
+      "epoch": 0.9536,
+      "grad_norm": 0.43058279156684875,
+      "learning_rate": 5.621430589813459e-06,
+      "loss": 2.9181,
+      "step": 7450
+    },
+    {
+      "epoch": 0.953728,
+      "grad_norm": 0.421468049287796,
+      "learning_rate": 5.590473785216654e-06,
+      "loss": 2.692,
+      "step": 7451
+    },
+    {
+      "epoch": 0.953856,
+      "grad_norm": 0.42022329568862915,
+      "learning_rate": 5.559601975348994e-06,
+      "loss": 2.7708,
+      "step": 7452
+    },
+    {
+      "epoch": 0.953984,
+      "grad_norm": 0.3680902123451233,
+      "learning_rate": 5.528815165517842e-06,
+      "loss": 2.6975,
+      "step": 7453
+    },
+    {
+      "epoch": 0.954112,
+      "grad_norm": 0.40102800726890564,
+      "learning_rate": 5.498113361015799e-06,
+      "loss": 2.7781,
+      "step": 7454
+    },
+    {
+      "epoch": 0.95424,
+      "grad_norm": 0.4077575206756592,
+      "learning_rate": 5.467496567120811e-06,
+      "loss": 2.8372,
+      "step": 7455
+    },
+    {
+      "epoch": 0.954368,
+      "grad_norm": 0.44897377490997314,
+      "learning_rate": 5.436964789096221e-06,
+      "loss": 2.8405,
+      "step": 7456
+    },
+    {
+      "epoch": 0.954496,
+      "grad_norm": 0.39801156520843506,
+      "learning_rate": 5.406518032190943e-06,
+      "loss": 2.7783,
+      "step": 7457
+    },
+    {
+      "epoch": 0.954624,
+      "grad_norm": 0.4009719491004944,
+      "learning_rate": 5.376156301638957e-06,
+      "loss": 2.7182,
+      "step": 7458
+    },
+    {
+      "epoch": 0.954752,
+      "grad_norm": 0.4568454623222351,
+      "learning_rate": 5.34587960265992e-06,
+      "loss": 2.7614,
+      "step": 7459
+    },
+    {
+      "epoch": 0.95488,
+      "grad_norm": 0.4374779462814331,
+      "learning_rate": 5.315687940458669e-06,
+      "loss": 2.7912,
+      "step": 7460
+    },
+    {
+      "epoch": 0.955008,
+      "grad_norm": 0.3974648416042328,
+      "learning_rate": 5.285581320225552e-06,
+      "loss": 2.7396,
+      "step": 7461
+    },
+    {
+      "epoch": 0.955136,
+      "grad_norm": 0.4227934181690216,
+      "learning_rate": 5.255559747136263e-06,
+      "loss": 2.7356,
+      "step": 7462
+    },
+    {
+      "epoch": 0.955264,
+      "grad_norm": 0.4347386658191681,
+      "learning_rate": 5.2256232263517835e-06,
+      "loss": 2.7976,
+      "step": 7463
+    },
+    {
+      "epoch": 0.955392,
+      "grad_norm": 0.3952711820602417,
+      "learning_rate": 5.195771763018609e-06,
+      "loss": 2.754,
+      "step": 7464
+    },
+    {
+      "epoch": 0.95552,
+      "grad_norm": 0.4104423522949219,
+      "learning_rate": 5.166005362268467e-06,
+      "loss": 2.7705,
+      "step": 7465
+    },
+    {
+      "epoch": 0.955648,
+      "grad_norm": 0.4081437587738037,
+      "learning_rate": 5.136324029218653e-06,
+      "loss": 2.8432,
+      "step": 7466
+    },
+    {
+      "epoch": 0.955776,
+      "grad_norm": 0.4155314564704895,
+      "learning_rate": 5.106727768971697e-06,
+      "loss": 2.676,
+      "step": 7467
+    },
+    {
+      "epoch": 0.955904,
+      "grad_norm": 0.4259299337863922,
+      "learning_rate": 5.07721658661553e-06,
+      "loss": 2.8143,
+      "step": 7468
+    },
+    {
+      "epoch": 0.956032,
+      "grad_norm": 0.4132324159145355,
+      "learning_rate": 5.04779048722348e-06,
+      "loss": 2.8692,
+      "step": 7469
+    },
+    {
+      "epoch": 0.95616,
+      "grad_norm": 0.4249313771724701,
+      "learning_rate": 5.018449475854225e-06,
+      "loss": 2.7782,
+      "step": 7470
+    },
+    {
+      "epoch": 0.956288,
+      "grad_norm": 0.3840361535549164,
+      "learning_rate": 4.989193557551786e-06,
+      "loss": 2.7676,
+      "step": 7471
+    },
+    {
+      "epoch": 0.956416,
+      "grad_norm": 0.4212460219860077,
+      "learning_rate": 4.960022737345693e-06,
+      "loss": 2.8702,
+      "step": 7472
+    },
+    {
+      "epoch": 0.956544,
+      "grad_norm": 0.4365506172180176,
+      "learning_rate": 4.930937020250603e-06,
+      "loss": 2.8033,
+      "step": 7473
+    },
+    {
+      "epoch": 0.956672,
+      "grad_norm": 0.42247018218040466,
+      "learning_rate": 4.90193641126685e-06,
+      "loss": 2.8388,
+      "step": 7474
+    },
+    {
+      "epoch": 0.9568,
+      "grad_norm": 0.4761890172958374,
+      "learning_rate": 4.873020915379834e-06,
+      "loss": 2.8012,
+      "step": 7475
+    },
+    {
+      "epoch": 0.956928,
+      "grad_norm": 0.41149210929870605,
+      "learning_rate": 4.8441905375605774e-06,
+      "loss": 2.8246,
+      "step": 7476
+    },
+    {
+      "epoch": 0.957056,
+      "grad_norm": 0.4087129831314087,
+      "learning_rate": 4.815445282765285e-06,
+      "loss": 2.7579,
+      "step": 7477
+    },
+    {
+      "epoch": 0.957184,
+      "grad_norm": 0.40764153003692627,
+      "learning_rate": 4.786785155935669e-06,
+      "loss": 2.8164,
+      "step": 7478
+    },
+    {
+      "epoch": 0.957312,
+      "grad_norm": 0.39991438388824463,
+      "learning_rate": 4.758210161998622e-06,
+      "loss": 2.7859,
+      "step": 7479
+    },
+    {
+      "epoch": 0.95744,
+      "grad_norm": 0.4299394190311432,
+      "learning_rate": 4.729720305866603e-06,
+      "loss": 2.8027,
+      "step": 7480
+    },
+    {
+      "epoch": 0.957568,
+      "grad_norm": 0.4364003837108612,
+      "learning_rate": 4.70131559243736e-06,
+      "loss": 2.88,
+      "step": 7481
+    },
+    {
+      "epoch": 0.957696,
+      "grad_norm": 0.40597766637802124,
+      "learning_rate": 4.672996026593934e-06,
+      "loss": 2.7683,
+      "step": 7482
+    },
+    {
+      "epoch": 0.957824,
+      "grad_norm": 0.4130901098251343,
+      "learning_rate": 4.644761613204817e-06,
+      "loss": 2.7807,
+      "step": 7483
+    },
+    {
+      "epoch": 0.957952,
+      "grad_norm": 0.3977372646331787,
+      "learning_rate": 4.616612357123795e-06,
+      "loss": 2.8803,
+      "step": 7484
+    },
+    {
+      "epoch": 0.95808,
+      "grad_norm": 0.4228385090827942,
+      "learning_rate": 4.588548263190107e-06,
+      "loss": 2.8565,
+      "step": 7485
+    },
+    {
+      "epoch": 0.958208,
+      "grad_norm": 0.41030868887901306,
+      "learning_rate": 4.560569336228337e-06,
+      "loss": 2.7228,
+      "step": 7486
+    },
+    {
+      "epoch": 0.958336,
+      "grad_norm": 0.4129139482975006,
+      "learning_rate": 4.532675581048251e-06,
+      "loss": 2.8201,
+      "step": 7487
+    },
+    {
+      "epoch": 0.958464,
+      "grad_norm": 0.4102204442024231,
+      "learning_rate": 4.504867002445179e-06,
+      "loss": 2.6628,
+      "step": 7488
+    },
+    {
+      "epoch": 0.958592,
+      "grad_norm": 0.42799338698387146,
+      "learning_rate": 4.477143605199796e-06,
+      "loss": 2.8922,
+      "step": 7489
+    },
+    {
+      "epoch": 0.95872,
+      "grad_norm": 0.41426703333854675,
+      "learning_rate": 4.449505394078013e-06,
+      "loss": 2.7727,
+      "step": 7490
+    },
+    {
+      "epoch": 0.958848,
+      "grad_norm": 0.4427347481250763,
+      "learning_rate": 4.421952373831139e-06,
+      "loss": 2.7471,
+      "step": 7491
+    },
+    {
+      "epoch": 0.958976,
+      "grad_norm": 0.40322145819664,
+      "learning_rate": 4.394484549195887e-06,
+      "loss": 2.7913,
+      "step": 7492
+    },
+    {
+      "epoch": 0.959104,
+      "grad_norm": 0.3935280740261078,
+      "learning_rate": 4.3671019248943125e-06,
+      "loss": 2.9599,
+      "step": 7493
+    },
+    {
+      "epoch": 0.959232,
+      "grad_norm": 0.396615207195282,
+      "learning_rate": 4.3398045056337594e-06,
+      "loss": 2.6911,
+      "step": 7494
+    },
+    {
+      "epoch": 0.95936,
+      "grad_norm": 0.4067341387271881,
+      "learning_rate": 4.312592296106976e-06,
+      "loss": 2.8462,
+      "step": 7495
+    },
+    {
+      "epoch": 0.959488,
+      "grad_norm": 0.4267822504043579,
+      "learning_rate": 4.285465300992164e-06,
+      "loss": 2.7333,
+      "step": 7496
+    },
+    {
+      "epoch": 0.959616,
+      "grad_norm": 0.38999679684638977,
+      "learning_rate": 4.258423524952649e-06,
+      "loss": 2.76,
+      "step": 7497
+    },
+    {
+      "epoch": 0.959744,
+      "grad_norm": 0.4085739850997925,
+      "learning_rate": 4.23146697263721e-06,
+      "loss": 2.6974,
+      "step": 7498
+    },
+    {
+      "epoch": 0.959872,
+      "grad_norm": 0.41616928577423096,
+      "learning_rate": 4.204595648680088e-06,
+      "loss": 2.76,
+      "step": 7499
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4321580231189728,
+      "learning_rate": 4.177809557700751e-06,
+      "loss": 2.7594,
+      "step": 7500
+    },
+    {
+      "epoch": 0.960128,
+      "grad_norm": 0.45831453800201416,
+      "learning_rate": 4.151108704303963e-06,
+      "loss": 2.8521,
+      "step": 7501
+    },
+    {
+      "epoch": 0.960256,
+      "grad_norm": 0.4210914373397827,
+      "learning_rate": 4.12449309307994e-06,
+      "loss": 2.8211,
+      "step": 7502
+    },
+    {
+      "epoch": 0.960384,
+      "grad_norm": 0.38020461797714233,
+      "learning_rate": 4.097962728604299e-06,
+      "loss": 2.7691,
+      "step": 7503
+    },
+    {
+      "epoch": 0.960512,
+      "grad_norm": 0.3899320363998413,
+      "learning_rate": 4.071517615437836e-06,
+      "loss": 2.8216,
+      "step": 7504
+    },
+    {
+      "epoch": 0.96064,
+      "grad_norm": 0.38664382696151733,
+      "learning_rate": 4.045157758126749e-06,
+      "loss": 2.8149,
+      "step": 7505
+    },
+    {
+      "epoch": 0.960768,
+      "grad_norm": 0.41139915585517883,
+      "learning_rate": 4.018883161202691e-06,
+      "loss": 3.0054,
+      "step": 7506
+    },
+    {
+      "epoch": 0.960896,
+      "grad_norm": 0.4170616567134857,
+      "learning_rate": 3.992693829182548e-06,
+      "loss": 2.8038,
+      "step": 7507
+    },
+    {
+      "epoch": 0.961024,
+      "grad_norm": 0.38547492027282715,
+      "learning_rate": 3.966589766568496e-06,
+      "loss": 2.7831,
+      "step": 7508
+    },
+    {
+      "epoch": 0.961152,
+      "grad_norm": 0.3992699980735779,
+      "learning_rate": 3.940570977848168e-06,
+      "loss": 3.0033,
+      "step": 7509
+    },
+    {
+      "epoch": 0.96128,
+      "grad_norm": 0.4253869652748108,
+      "learning_rate": 3.914637467494542e-06,
+      "loss": 2.7604,
+      "step": 7510
+    },
+    {
+      "epoch": 0.961408,
+      "grad_norm": 0.4172957241535187,
+      "learning_rate": 3.888789239965884e-06,
+      "loss": 2.7327,
+      "step": 7511
+    },
+    {
+      "epoch": 0.961536,
+      "grad_norm": 0.4136348366737366,
+      "learning_rate": 3.863026299705696e-06,
+      "loss": 2.6421,
+      "step": 7512
+    },
+    {
+      "epoch": 0.961664,
+      "grad_norm": 0.4008583426475525,
+      "learning_rate": 3.837348651142991e-06,
+      "loss": 2.6725,
+      "step": 7513
+    },
+    {
+      "epoch": 0.961792,
+      "grad_norm": 0.42855483293533325,
+      "learning_rate": 3.811756298692126e-06,
+      "loss": 2.781,
+      "step": 7514
+    },
+    {
+      "epoch": 0.96192,
+      "grad_norm": 0.42376619577407837,
+      "learning_rate": 3.786249246752638e-06,
+      "loss": 2.824,
+      "step": 7515
+    },
+    {
+      "epoch": 0.962048,
+      "grad_norm": 0.3962092697620392,
+      "learning_rate": 3.7608274997095184e-06,
+      "loss": 2.7698,
+      "step": 7516
+    },
+    {
+      "epoch": 0.962176,
+      "grad_norm": 0.41270214319229126,
+      "learning_rate": 3.7354910619329942e-06,
+      "loss": 2.7929,
+      "step": 7517
+    },
+    {
+      "epoch": 0.962304,
+      "grad_norm": 0.42077794671058655,
+      "learning_rate": 3.710239937778803e-06,
+      "loss": 2.8111,
+      "step": 7518
+    },
+    {
+      "epoch": 0.962432,
+      "grad_norm": 0.5205722451210022,
+      "learning_rate": 3.685074131587862e-06,
+      "loss": 2.7348,
+      "step": 7519
+    },
+    {
+      "epoch": 0.96256,
+      "grad_norm": 0.4130792021751404,
+      "learning_rate": 3.6599936476864326e-06,
+      "loss": 2.7737,
+      "step": 7520
+    },
+    {
+      "epoch": 0.962688,
+      "grad_norm": 0.39363035559654236,
+      "learning_rate": 3.634998490386121e-06,
+      "loss": 2.7943,
+      "step": 7521
+    },
+    {
+      "epoch": 0.962816,
+      "grad_norm": 0.41831958293914795,
+      "learning_rate": 3.61008866398399e-06,
+      "loss": 2.8541,
+      "step": 7522
+    },
+    {
+      "epoch": 0.962944,
+      "grad_norm": 0.4250433146953583,
+      "learning_rate": 3.5852641727622257e-06,
+      "loss": 2.7566,
+      "step": 7523
+    },
+    {
+      "epoch": 0.963072,
+      "grad_norm": 0.40082722902297974,
+      "learning_rate": 3.5605250209885253e-06,
+      "loss": 2.7211,
+      "step": 7524
+    },
+    {
+      "epoch": 0.9632,
+      "grad_norm": 0.40332916378974915,
+      "learning_rate": 3.5358712129157642e-06,
+      "loss": 2.712,
+      "step": 7525
+    },
+    {
+      "epoch": 0.963328,
+      "grad_norm": 0.41891103982925415,
+      "learning_rate": 3.5113027527822192e-06,
+      "loss": 2.8062,
+      "step": 7526
+    },
+    {
+      "epoch": 0.963456,
+      "grad_norm": 0.41095125675201416,
+      "learning_rate": 3.4868196448115676e-06,
+      "loss": 2.8025,
+      "step": 7527
+    },
+    {
+      "epoch": 0.963584,
+      "grad_norm": 0.412178635597229,
+      "learning_rate": 3.4624218932127194e-06,
+      "loss": 2.6665,
+      "step": 7528
+    },
+    {
+      "epoch": 0.963712,
+      "grad_norm": 0.4042358994483948,
+      "learning_rate": 3.43810950217982e-06,
+      "loss": 2.8813,
+      "step": 7529
+    },
+    {
+      "epoch": 0.96384,
+      "grad_norm": 0.3830242156982422,
+      "learning_rate": 3.4138824758925822e-06,
+      "loss": 2.7015,
+      "step": 7530
+    },
+    {
+      "epoch": 0.963968,
+      "grad_norm": 0.3938251733779907,
+      "learning_rate": 3.389740818515841e-06,
+      "loss": 2.8357,
+      "step": 7531
+    },
+    {
+      "epoch": 0.964096,
+      "grad_norm": 0.4077300727367401,
+      "learning_rate": 3.3656845341998865e-06,
+      "loss": 2.7721,
+      "step": 7532
+    },
+    {
+      "epoch": 0.964224,
+      "grad_norm": 0.41337141394615173,
+      "learning_rate": 3.341713627080245e-06,
+      "loss": 2.7442,
+      "step": 7533
+    },
+    {
+      "epoch": 0.964352,
+      "grad_norm": 0.4123063087463379,
+      "learning_rate": 3.3178281012777866e-06,
+      "loss": 2.797,
+      "step": 7534
+    },
+    {
+      "epoch": 0.96448,
+      "grad_norm": 0.4150235056877136,
+      "learning_rate": 3.2940279608986714e-06,
+      "loss": 2.8552,
+      "step": 7535
+    },
+    {
+      "epoch": 0.964608,
+      "grad_norm": 0.4228116273880005,
+      "learning_rate": 3.2703132100345146e-06,
+      "loss": 2.7344,
+      "step": 7536
+    },
+    {
+      "epoch": 0.964736,
+      "grad_norm": 0.4046693742275238,
+      "learning_rate": 3.2466838527620557e-06,
+      "loss": 2.7356,
+      "step": 7537
+    },
+    {
+      "epoch": 0.964864,
+      "grad_norm": 0.4249413311481476,
+      "learning_rate": 3.22313989314349e-06,
+      "loss": 2.7965,
+      "step": 7538
+    },
+    {
+      "epoch": 0.964992,
+      "grad_norm": 0.4746173620223999,
+      "learning_rate": 3.1996813352263566e-06,
+      "loss": 2.8582,
+      "step": 7539
+    },
+    {
+      "epoch": 0.96512,
+      "grad_norm": 0.4036165475845337,
+      "learning_rate": 3.176308183043375e-06,
+      "loss": 2.6078,
+      "step": 7540
+    },
+    {
+      "epoch": 0.965248,
+      "grad_norm": 0.4236793518066406,
+      "learning_rate": 3.1530204406127192e-06,
+      "loss": 2.6796,
+      "step": 7541
+    },
+    {
+      "epoch": 0.965376,
+      "grad_norm": 0.3770029842853546,
+      "learning_rate": 3.1298181119377432e-06,
+      "loss": 2.7376,
+      "step": 7542
+    },
+    {
+      "epoch": 0.965504,
+      "grad_norm": 0.3981567621231079,
+      "learning_rate": 3.1067012010073114e-06,
+      "loss": 2.8573,
+      "step": 7543
+    },
+    {
+      "epoch": 0.965632,
+      "grad_norm": 0.41961055994033813,
+      "learning_rate": 3.0836697117954115e-06,
+      "loss": 2.7348,
+      "step": 7544
+    },
+    {
+      "epoch": 0.96576,
+      "grad_norm": 0.4306756854057312,
+      "learning_rate": 3.0607236482613766e-06,
+      "loss": 2.702,
+      "step": 7545
+    },
+    {
+      "epoch": 0.965888,
+      "grad_norm": 0.416497141122818,
+      "learning_rate": 3.0378630143499953e-06,
+      "loss": 2.7596,
+      "step": 7546
+    },
+    {
+      "epoch": 0.966016,
+      "grad_norm": 0.4046233296394348,
+      "learning_rate": 3.0150878139912906e-06,
+      "loss": 2.7061,
+      "step": 7547
+    },
+    {
+      "epoch": 0.966144,
+      "grad_norm": 0.41155678033828735,
+      "learning_rate": 2.9923980511004646e-06,
+      "loss": 2.8387,
+      "step": 7548
+    },
+    {
+      "epoch": 0.966272,
+      "grad_norm": 0.4397585988044739,
+      "learning_rate": 2.9697937295782295e-06,
+      "loss": 2.8683,
+      "step": 7549
+    },
+    {
+      "epoch": 0.9664,
+      "grad_norm": 0.40161439776420593,
+      "learning_rate": 2.9472748533105886e-06,
+      "loss": 2.7437,
+      "step": 7550
+    },
+    {
+      "epoch": 0.966528,
+      "grad_norm": 0.4130907654762268,
+      "learning_rate": 2.924841426168667e-06,
+      "loss": 2.716,
+      "step": 7551
+    },
+    {
+      "epoch": 0.966656,
+      "grad_norm": 0.39372122287750244,
+      "learning_rate": 2.9024934520091028e-06,
+      "loss": 2.8114,
+      "step": 7552
+    },
+    {
+      "epoch": 0.966784,
+      "grad_norm": 0.4379892349243164,
+      "learning_rate": 2.8802309346737666e-06,
+      "loss": 2.913,
+      "step": 7553
+    },
+    {
+      "epoch": 0.966912,
+      "grad_norm": 0.43348944187164307,
+      "learning_rate": 2.858053877989819e-06,
+      "loss": 2.8017,
+      "step": 7554
+    },
+    {
+      "epoch": 0.96704,
+      "grad_norm": 0.395609050989151,
+      "learning_rate": 2.835962285769822e-06,
+      "loss": 2.8098,
+      "step": 7555
+    },
+    {
+      "epoch": 0.967168,
+      "grad_norm": 0.40994560718536377,
+      "learning_rate": 2.813956161811515e-06,
+      "loss": 2.7561,
+      "step": 7556
+    },
+    {
+      "epoch": 0.967296,
+      "grad_norm": 0.3929232060909271,
+      "learning_rate": 2.7920355098979833e-06,
+      "loss": 2.9071,
+      "step": 7557
+    },
+    {
+      "epoch": 0.967424,
+      "grad_norm": 0.3983813524246216,
+      "learning_rate": 2.7702003337977123e-06,
+      "loss": 2.815,
+      "step": 7558
+    },
+    {
+      "epoch": 0.967552,
+      "grad_norm": 0.39699119329452515,
+      "learning_rate": 2.7484506372643658e-06,
+      "loss": 2.7502,
+      "step": 7559
+    },
+    {
+      "epoch": 0.96768,
+      "grad_norm": 0.4191446304321289,
+      "learning_rate": 2.7267864240369533e-06,
+      "loss": 2.8563,
+      "step": 7560
+    },
+    {
+      "epoch": 0.967808,
+      "grad_norm": 0.40774789452552795,
+      "learning_rate": 2.7052076978398842e-06,
+      "loss": 2.7956,
+      "step": 7561
+    },
+    {
+      "epoch": 0.967936,
+      "grad_norm": 0.43165335059165955,
+      "learning_rate": 2.683714462382747e-06,
+      "loss": 2.8264,
+      "step": 7562
+    },
+    {
+      "epoch": 0.968064,
+      "grad_norm": 0.37399888038635254,
+      "learning_rate": 2.6623067213605303e-06,
+      "loss": 2.7673,
+      "step": 7563
+    },
+    {
+      "epoch": 0.968192,
+      "grad_norm": 0.4281916320323944,
+      "learning_rate": 2.6409844784533456e-06,
+      "loss": 2.7581,
+      "step": 7564
+    },
+    {
+      "epoch": 0.96832,
+      "grad_norm": 0.3969825208187103,
+      "learning_rate": 2.619747737326872e-06,
+      "loss": 2.888,
+      "step": 7565
+    },
+    {
+      "epoch": 0.968448,
+      "grad_norm": 0.4149629473686218,
+      "learning_rate": 2.5985965016318e-06,
+      "loss": 2.768,
+      "step": 7566
+    },
+    {
+      "epoch": 0.968576,
+      "grad_norm": 0.4440431594848633,
+      "learning_rate": 2.5775307750043887e-06,
+      "loss": 2.9306,
+      "step": 7567
+    },
+    {
+      "epoch": 0.968704,
+      "grad_norm": 0.39394769072532654,
+      "learning_rate": 2.556550561066073e-06,
+      "loss": 2.7577,
+      "step": 7568
+    },
+    {
+      "epoch": 0.968832,
+      "grad_norm": 0.39998772740364075,
+      "learning_rate": 2.5356558634235806e-06,
+      "loss": 2.7022,
+      "step": 7569
+    },
+    {
+      "epoch": 0.96896,
+      "grad_norm": 0.40041384100914,
+      "learning_rate": 2.514846685668981e-06,
+      "loss": 2.8592,
+      "step": 7570
+    },
+    {
+      "epoch": 0.969088,
+      "grad_norm": 0.3943963646888733,
+      "learning_rate": 2.4941230313795247e-06,
+      "loss": 2.8032,
+      "step": 7571
+    },
+    {
+      "epoch": 0.969216,
+      "grad_norm": 0.41621309518814087,
+      "learning_rate": 2.4734849041179176e-06,
+      "loss": 2.8792,
+      "step": 7572
+    },
+    {
+      "epoch": 0.969344,
+      "grad_norm": 0.4032832682132721,
+      "learning_rate": 2.4529323074320984e-06,
+      "loss": 2.7989,
+      "step": 7573
+    },
+    {
+      "epoch": 0.969472,
+      "grad_norm": 0.4331413209438324,
+      "learning_rate": 2.432465244855242e-06,
+      "loss": 2.8867,
+      "step": 7574
+    },
+    {
+      "epoch": 0.9696,
+      "grad_norm": 0.40286073088645935,
+      "learning_rate": 2.412083719905922e-06,
+      "loss": 2.793,
+      "step": 7575
+    },
+    {
+      "epoch": 0.969728,
+      "grad_norm": 0.42065808176994324,
+      "learning_rate": 2.391787736087947e-06,
+      "loss": 2.7219,
+      "step": 7576
+    },
+    {
+      "epoch": 0.969856,
+      "grad_norm": 0.39660483598709106,
+      "learning_rate": 2.3715772968904147e-06,
+      "loss": 2.7705,
+      "step": 7577
+    },
+    {
+      "epoch": 0.969984,
+      "grad_norm": 0.43067866563796997,
+      "learning_rate": 2.3514524057877685e-06,
+      "loss": 2.8315,
+      "step": 7578
+    },
+    {
+      "epoch": 0.970112,
+      "grad_norm": 0.38596609234809875,
+      "learning_rate": 2.331413066239685e-06,
+      "loss": 2.61,
+      "step": 7579
+    },
+    {
+      "epoch": 0.97024,
+      "grad_norm": 0.38951170444488525,
+      "learning_rate": 2.3114592816911863e-06,
+      "loss": 2.7694,
+      "step": 7580
+    },
+    {
+      "epoch": 0.970368,
+      "grad_norm": 0.42816680669784546,
+      "learning_rate": 2.2915910555725283e-06,
+      "loss": 2.8267,
+      "step": 7581
+    },
+    {
+      "epoch": 0.970496,
+      "grad_norm": 0.4554072618484497,
+      "learning_rate": 2.2718083912992572e-06,
+      "loss": 2.738,
+      "step": 7582
+    },
+    {
+      "epoch": 0.970624,
+      "grad_norm": 0.39972421526908875,
+      "learning_rate": 2.2521112922723184e-06,
+      "loss": 2.8757,
+      "step": 7583
+    },
+    {
+      "epoch": 0.970752,
+      "grad_norm": 0.4425351023674011,
+      "learning_rate": 2.2324997618778375e-06,
+      "loss": 2.7702,
+      "step": 7584
+    },
+    {
+      "epoch": 0.97088,
+      "grad_norm": 0.42608925700187683,
+      "learning_rate": 2.212973803487284e-06,
+      "loss": 2.7267,
+      "step": 7585
+    },
+    {
+      "epoch": 0.971008,
+      "grad_norm": 0.39768266677856445,
+      "learning_rate": 2.1935334204573054e-06,
+      "loss": 2.7258,
+      "step": 7586
+    },
+    {
+      "epoch": 0.971136,
+      "grad_norm": 0.42298102378845215,
+      "learning_rate": 2.1741786161300626e-06,
+      "loss": 2.6493,
+      "step": 7587
+    },
+    {
+      "epoch": 0.971264,
+      "grad_norm": 0.4387364387512207,
+      "learning_rate": 2.1549093938327825e-06,
+      "loss": 2.7879,
+      "step": 7588
+    },
+    {
+      "epoch": 0.971392,
+      "grad_norm": 0.39813292026519775,
+      "learning_rate": 2.135725756878093e-06,
+      "loss": 2.909,
+      "step": 7589
+    },
+    {
+      "epoch": 0.97152,
+      "grad_norm": 0.39164644479751587,
+      "learning_rate": 2.1166277085639117e-06,
+      "loss": 2.8435,
+      "step": 7590
+    },
+    {
+      "epoch": 0.971648,
+      "grad_norm": 0.39229875802993774,
+      "learning_rate": 2.0976152521733905e-06,
+      "loss": 2.7705,
+      "step": 7591
+    },
+    {
+      "epoch": 0.971776,
+      "grad_norm": 0.42877814173698425,
+      "learning_rate": 2.0786883909749698e-06,
+      "loss": 2.8544,
+      "step": 7592
+    },
+    {
+      "epoch": 0.971904,
+      "grad_norm": 0.4041348099708557,
+      "learning_rate": 2.059847128222381e-06,
+      "loss": 2.8499,
+      "step": 7593
+    },
+    {
+      "epoch": 0.972032,
+      "grad_norm": 0.39145880937576294,
+      "learning_rate": 2.0410914671546986e-06,
+      "loss": 2.7651,
+      "step": 7594
+    },
+    {
+      "epoch": 0.97216,
+      "grad_norm": 0.3860267996788025,
+      "learning_rate": 2.0224214109962334e-06,
+      "loss": 2.8084,
+      "step": 7595
+    },
+    {
+      "epoch": 0.972288,
+      "grad_norm": 0.4040604829788208,
+      "learning_rate": 2.0038369629565846e-06,
+      "loss": 2.8392,
+      "step": 7596
+    },
+    {
+      "epoch": 0.972416,
+      "grad_norm": 0.4345647990703583,
+      "learning_rate": 1.98533812623064e-06,
+      "loss": 2.8532,
+      "step": 7597
+    },
+    {
+      "epoch": 0.972544,
+      "grad_norm": 0.38005390763282776,
+      "learning_rate": 1.9669249039985794e-06,
+      "loss": 2.7824,
+      "step": 7598
+    },
+    {
+      "epoch": 0.972672,
+      "grad_norm": 0.3936719596385956,
+      "learning_rate": 1.9485972994257585e-06,
+      "loss": 2.8159,
+      "step": 7599
+    },
+    {
+      "epoch": 0.9728,
+      "grad_norm": 0.43277212977409363,
+      "learning_rate": 1.930355315663046e-06,
+      "loss": 2.635,
+      "step": 7600
+    },
+    {
+      "epoch": 0.972928,
+      "grad_norm": 0.4292260408401489,
+      "learning_rate": 1.9121989558463225e-06,
+      "loss": 2.7822,
+      "step": 7601
+    },
+    {
+      "epoch": 0.973056,
+      "grad_norm": 0.39134371280670166,
+      "learning_rate": 1.8941282230969247e-06,
+      "loss": 2.7827,
+      "step": 7602
+    },
+    {
+      "epoch": 0.973184,
+      "grad_norm": 0.4156009256839752,
+      "learning_rate": 1.876143120521423e-06,
+      "loss": 2.7685,
+      "step": 7603
+    },
+    {
+      "epoch": 0.973312,
+      "grad_norm": 0.43048372864723206,
+      "learning_rate": 1.8582436512116775e-06,
+      "loss": 2.9625,
+      "step": 7604
+    },
+    {
+      "epoch": 0.97344,
+      "grad_norm": 0.3953916132450104,
+      "learning_rate": 1.840429818244782e-06,
+      "loss": 2.5815,
+      "step": 7605
+    },
+    {
+      "epoch": 0.973568,
+      "grad_norm": 0.40754303336143494,
+      "learning_rate": 1.8227016246831762e-06,
+      "loss": 2.7117,
+      "step": 7606
+    },
+    {
+      "epoch": 0.973696,
+      "grad_norm": 0.41868069767951965,
+      "learning_rate": 1.805059073574533e-06,
+      "loss": 2.8476,
+      "step": 7607
+    },
+    {
+      "epoch": 0.973824,
+      "grad_norm": 0.41999417543411255,
+      "learning_rate": 1.7875021679518155e-06,
+      "loss": 2.7138,
+      "step": 7608
+    },
+    {
+      "epoch": 0.973952,
+      "grad_norm": 0.3847648501396179,
+      "learning_rate": 1.7700309108332203e-06,
+      "loss": 2.7472,
+      "step": 7609
+    },
+    {
+      "epoch": 0.97408,
+      "grad_norm": 0.4758499562740326,
+      "learning_rate": 1.7526453052223445e-06,
+      "loss": 2.7662,
+      "step": 7610
+    },
+    {
+      "epoch": 0.974208,
+      "grad_norm": 0.41965749859809875,
+      "learning_rate": 1.7353453541078535e-06,
+      "loss": 2.813,
+      "step": 7611
+    },
+    {
+      "epoch": 0.974336,
+      "grad_norm": 0.39910799264907837,
+      "learning_rate": 1.7181310604639233e-06,
+      "loss": 2.8076,
+      "step": 7612
+    },
+    {
+      "epoch": 0.974464,
+      "grad_norm": 0.423800528049469,
+      "learning_rate": 1.7010024272498537e-06,
+      "loss": 2.8244,
+      "step": 7613
+    },
+    {
+      "epoch": 0.974592,
+      "grad_norm": 0.4008961319923401,
+      "learning_rate": 1.683959457410178e-06,
+      "loss": 2.7227,
+      "step": 7614
+    },
+    {
+      "epoch": 0.97472,
+      "grad_norm": 0.41595199704170227,
+      "learning_rate": 1.6670021538748859e-06,
+      "loss": 2.858,
+      "step": 7615
+    },
+    {
+      "epoch": 0.974848,
+      "grad_norm": 0.4170016050338745,
+      "learning_rate": 1.65013051955909e-06,
+      "loss": 2.8016,
+      "step": 7616
+    },
+    {
+      "epoch": 0.974976,
+      "grad_norm": 0.3875192105770111,
+      "learning_rate": 1.6333445573632478e-06,
+      "loss": 2.8291,
+      "step": 7617
+    },
+    {
+      "epoch": 0.975104,
+      "grad_norm": 0.40033677220344543,
+      "learning_rate": 1.6166442701730511e-06,
+      "loss": 2.7115,
+      "step": 7618
+    },
+    {
+      "epoch": 0.975232,
+      "grad_norm": 0.40940189361572266,
+      "learning_rate": 1.6000296608594256e-06,
+      "loss": 2.7809,
+      "step": 7619
+    },
+    {
+      "epoch": 0.97536,
+      "grad_norm": 0.4110143780708313,
+      "learning_rate": 1.5835007322786421e-06,
+      "loss": 2.7263,
+      "step": 7620
+    },
+    {
+      "epoch": 0.975488,
+      "grad_norm": 0.5542181134223938,
+      "learning_rate": 1.567057487272261e-06,
+      "loss": 2.8584,
+      "step": 7621
+    },
+    {
+      "epoch": 0.975616,
+      "grad_norm": 0.4390653669834137,
+      "learning_rate": 1.5506999286669654e-06,
+      "loss": 2.7579,
+      "step": 7622
+    },
+    {
+      "epoch": 0.975744,
+      "grad_norm": 0.4125896990299225,
+      "learning_rate": 1.534428059274895e-06,
+      "loss": 2.6767,
+      "step": 7623
+    },
+    {
+      "epoch": 0.975872,
+      "grad_norm": 0.4180055260658264,
+      "learning_rate": 1.5182418818933675e-06,
+      "loss": 2.8118,
+      "step": 7624
+    },
+    {
+      "epoch": 0.976,
+      "grad_norm": 0.41217243671417236,
+      "learning_rate": 1.502141399304935e-06,
+      "loss": 2.7926,
+      "step": 7625
+    },
+    {
+      "epoch": 0.976128,
+      "grad_norm": 0.4358762204647064,
+      "learning_rate": 1.4861266142775499e-06,
+      "loss": 2.8967,
+      "step": 7626
+    },
+    {
+      "epoch": 0.976256,
+      "grad_norm": 0.41720932722091675,
+      "learning_rate": 1.4701975295641768e-06,
+      "loss": 2.8003,
+      "step": 7627
+    },
+    {
+      "epoch": 0.976384,
+      "grad_norm": 0.4165570139884949,
+      "learning_rate": 1.4543541479033473e-06,
+      "loss": 2.7421,
+      "step": 7628
+    },
+    {
+      "epoch": 0.976512,
+      "grad_norm": 0.42214488983154297,
+      "learning_rate": 1.438596472018716e-06,
+      "loss": 2.7863,
+      "step": 7629
+    },
+    {
+      "epoch": 0.97664,
+      "grad_norm": 0.41833895444869995,
+      "learning_rate": 1.422924504619061e-06,
+      "loss": 2.5878,
+      "step": 7630
+    },
+    {
+      "epoch": 0.976768,
+      "grad_norm": 0.46618202328681946,
+      "learning_rate": 1.4073382483987817e-06,
+      "loss": 2.7392,
+      "step": 7631
+    },
+    {
+      "epoch": 0.976896,
+      "grad_norm": 0.40783461928367615,
+      "learning_rate": 1.39183770603718e-06,
+      "loss": 2.8646,
+      "step": 7632
+    },
+    {
+      "epoch": 0.977024,
+      "grad_norm": 0.46974778175354004,
+      "learning_rate": 1.3764228801990687e-06,
+      "loss": 2.7509,
+      "step": 7633
+    },
+    {
+      "epoch": 0.977152,
+      "grad_norm": 0.41759082674980164,
+      "learning_rate": 1.3610937735344386e-06,
+      "loss": 2.6784,
+      "step": 7634
+    },
+    {
+      "epoch": 0.97728,
+      "grad_norm": 0.42571502923965454,
+      "learning_rate": 1.3458503886784602e-06,
+      "loss": 2.786,
+      "step": 7635
+    },
+    {
+      "epoch": 0.977408,
+      "grad_norm": 0.407915323972702,
+      "learning_rate": 1.3306927282517034e-06,
+      "loss": 2.7676,
+      "step": 7636
+    },
+    {
+      "epoch": 0.977536,
+      "grad_norm": 0.3781096041202545,
+      "learning_rate": 1.3156207948599176e-06,
+      "loss": 2.7559,
+      "step": 7637
+    },
+    {
+      "epoch": 0.977664,
+      "grad_norm": 0.41817641258239746,
+      "learning_rate": 1.3006345910941963e-06,
+      "loss": 2.8616,
+      "step": 7638
+    },
+    {
+      "epoch": 0.977792,
+      "grad_norm": 0.39447641372680664,
+      "learning_rate": 1.2857341195308125e-06,
+      "loss": 2.7923,
+      "step": 7639
+    },
+    {
+      "epoch": 0.97792,
+      "grad_norm": 0.37216341495513916,
+      "learning_rate": 1.2709193827312725e-06,
+      "loss": 2.6967,
+      "step": 7640
+    },
+    {
+      "epoch": 0.978048,
+      "grad_norm": 0.43733322620391846,
+      "learning_rate": 1.2561903832424836e-06,
+      "loss": 2.7651,
+      "step": 7641
+    },
+    {
+      "epoch": 0.978176,
+      "grad_norm": 0.4213343858718872,
+      "learning_rate": 1.2415471235965315e-06,
+      "loss": 2.9162,
+      "step": 7642
+    },
+    {
+      "epoch": 0.978304,
+      "grad_norm": 0.4380703270435333,
+      "learning_rate": 1.2269896063107356e-06,
+      "loss": 2.731,
+      "step": 7643
+    },
+    {
+      "epoch": 0.978432,
+      "grad_norm": 0.4434877336025238,
+      "learning_rate": 1.2125178338876497e-06,
+      "loss": 2.8777,
+      "step": 7644
+    },
+    {
+      "epoch": 0.97856,
+      "grad_norm": 0.38695013523101807,
+      "learning_rate": 1.1981318088152281e-06,
+      "loss": 2.6829,
+      "step": 7645
+    },
+    {
+      "epoch": 0.978688,
+      "grad_norm": 0.4641690254211426,
+      "learning_rate": 1.1838315335664928e-06,
+      "loss": 2.777,
+      "step": 7646
+    },
+    {
+      "epoch": 0.978816,
+      "grad_norm": 0.3750092685222626,
+      "learning_rate": 1.169617010599977e-06,
+      "loss": 2.8117,
+      "step": 7647
+    },
+    {
+      "epoch": 0.978944,
+      "grad_norm": 0.41981834173202515,
+      "learning_rate": 1.1554882423591707e-06,
+      "loss": 2.8714,
+      "step": 7648
+    },
+    {
+      "epoch": 0.979072,
+      "grad_norm": 0.40872371196746826,
+      "learning_rate": 1.141445231273075e-06,
+      "loss": 2.7602,
+      "step": 7649
+    },
+    {
+      "epoch": 0.9792,
+      "grad_norm": 0.3993683159351349,
+      "learning_rate": 1.1274879797558146e-06,
+      "loss": 2.73,
+      "step": 7650
+    },
+    {
+      "epoch": 0.979328,
+      "grad_norm": 0.42088964581489563,
+      "learning_rate": 1.113616490206748e-06,
+      "loss": 2.8628,
+      "step": 7651
+    },
+    {
+      "epoch": 0.979456,
+      "grad_norm": 0.42027631402015686,
+      "learning_rate": 1.0998307650106342e-06,
+      "loss": 2.9051,
+      "step": 7652
+    },
+    {
+      "epoch": 0.979584,
+      "grad_norm": 0.4220849871635437,
+      "learning_rate": 1.0861308065373554e-06,
+      "loss": 2.7132,
+      "step": 7653
+    },
+    {
+      "epoch": 0.979712,
+      "grad_norm": 0.43303459882736206,
+      "learning_rate": 1.0725166171420831e-06,
+      "loss": 2.8081,
+      "step": 7654
+    },
+    {
+      "epoch": 0.97984,
+      "grad_norm": 0.434622585773468,
+      "learning_rate": 1.0589881991652784e-06,
+      "loss": 2.8504,
+      "step": 7655
+    },
+    {
+      "epoch": 0.979968,
+      "grad_norm": 0.5533398389816284,
+      "learning_rate": 1.0455455549326365e-06,
+      "loss": 2.6052,
+      "step": 7656
+    },
+    {
+      "epoch": 0.980096,
+      "grad_norm": 0.46554288268089294,
+      "learning_rate": 1.0321886867550868e-06,
+      "loss": 2.8378,
+      "step": 7657
+    },
+    {
+      "epoch": 0.980224,
+      "grad_norm": 0.43814224004745483,
+      "learning_rate": 1.0189175969287924e-06,
+      "loss": 2.7374,
+      "step": 7658
+    },
+    {
+      "epoch": 0.980352,
+      "grad_norm": 0.40809810161590576,
+      "learning_rate": 1.0057322877352615e-06,
+      "loss": 2.9278,
+      "step": 7659
+    },
+    {
+      "epoch": 0.98048,
+      "grad_norm": 0.4096071422100067,
+      "learning_rate": 9.926327614411811e-07,
+      "loss": 2.8225,
+      "step": 7660
+    },
+    {
+      "epoch": 0.980608,
+      "grad_norm": 0.38358479738235474,
+      "learning_rate": 9.796190202985832e-07,
+      "loss": 2.7425,
+      "step": 7661
+    },
+    {
+      "epoch": 0.980736,
+      "grad_norm": 0.4046532213687897,
+      "learning_rate": 9.666910665445672e-07,
+      "loss": 2.89,
+      "step": 7662
+    },
+    {
+      "epoch": 0.980864,
+      "grad_norm": 0.41400471329689026,
+      "learning_rate": 9.538489024016328e-07,
+      "loss": 2.7301,
+      "step": 7663
+    },
+    {
+      "epoch": 0.980992,
+      "grad_norm": 0.42196324467658997,
+      "learning_rate": 9.410925300775697e-07,
+      "loss": 3.0267,
+      "step": 7664
+    },
+    {
+      "epoch": 0.98112,
+      "grad_norm": 0.40495219826698303,
+      "learning_rate": 9.284219517652349e-07,
+      "loss": 2.8046,
+      "step": 7665
+    },
+    {
+      "epoch": 0.981248,
+      "grad_norm": 0.37828436493873596,
+      "learning_rate": 9.158371696428858e-07,
+      "loss": 2.7156,
+      "step": 7666
+    },
+    {
+      "epoch": 0.981376,
+      "grad_norm": 0.4120602607727051,
+      "learning_rate": 9.033381858740697e-07,
+      "loss": 2.7818,
+      "step": 7667
+    },
+    {
+      "epoch": 0.981504,
+      "grad_norm": 0.5060580968856812,
+      "learning_rate": 8.909250026074012e-07,
+      "loss": 2.5952,
+      "step": 7668
+    },
+    {
+      "epoch": 0.981632,
+      "grad_norm": 0.4025074243545532,
+      "learning_rate": 8.785976219768954e-07,
+      "loss": 2.8926,
+      "step": 7669
+    },
+    {
+      "epoch": 0.98176,
+      "grad_norm": 0.4291651248931885,
+      "learning_rate": 8.663560461018016e-07,
+      "loss": 2.8587,
+      "step": 7670
+    },
+    {
+      "epoch": 0.981888,
+      "grad_norm": 0.41276735067367554,
+      "learning_rate": 8.542002770865476e-07,
+      "loss": 2.9091,
+      "step": 7671
+    },
+    {
+      "epoch": 0.982016,
+      "grad_norm": 0.4233454465866089,
+      "learning_rate": 8.421303170208505e-07,
+      "loss": 2.6378,
+      "step": 7672
+    },
+    {
+      "epoch": 0.982144,
+      "grad_norm": 0.44946223497390747,
+      "learning_rate": 8.301461679796618e-07,
+      "loss": 2.7366,
+      "step": 7673
+    },
+    {
+      "epoch": 0.982272,
+      "grad_norm": 0.4310978949069977,
+      "learning_rate": 8.182478320232223e-07,
+      "loss": 2.7975,
+      "step": 7674
+    },
+    {
+      "epoch": 0.9824,
+      "grad_norm": 0.459465891122818,
+      "learning_rate": 8.064353111970068e-07,
+      "loss": 2.7977,
+      "step": 7675
+    },
+    {
+      "epoch": 0.982528,
+      "grad_norm": 0.40664732456207275,
+      "learning_rate": 7.947086075317245e-07,
+      "loss": 2.6993,
+      "step": 7676
+    },
+    {
+      "epoch": 0.982656,
+      "grad_norm": 0.47446268796920776,
+      "learning_rate": 7.830677230433181e-07,
+      "loss": 2.6813,
+      "step": 7677
+    },
+    {
+      "epoch": 0.982784,
+      "grad_norm": 0.3805856704711914,
+      "learning_rate": 7.715126597329647e-07,
+      "loss": 2.6493,
+      "step": 7678
+    },
+    {
+      "epoch": 0.982912,
+      "grad_norm": 0.373716801404953,
+      "learning_rate": 7.600434195871864e-07,
+      "loss": 2.7787,
+      "step": 7679
+    },
+    {
+      "epoch": 0.98304,
+      "grad_norm": 0.4138309359550476,
+      "learning_rate": 7.486600045775727e-07,
+      "loss": 2.7794,
+      "step": 7680
+    },
+    {
+      "epoch": 0.983168,
+      "grad_norm": 0.3977911174297333,
+      "learning_rate": 7.373624166611137e-07,
+      "loss": 2.7208,
+      "step": 7681
+    },
+    {
+      "epoch": 0.983296,
+      "grad_norm": 0.430889368057251,
+      "learning_rate": 7.261506577800892e-07,
+      "loss": 2.8263,
+      "step": 7682
+    },
+    {
+      "epoch": 0.983424,
+      "grad_norm": 0.412779301404953,
+      "learning_rate": 7.150247298617907e-07,
+      "loss": 2.8722,
+      "step": 7683
+    },
+    {
+      "epoch": 0.983552,
+      "grad_norm": 0.3816899061203003,
+      "learning_rate": 7.039846348189105e-07,
+      "loss": 2.8119,
+      "step": 7684
+    },
+    {
+      "epoch": 0.98368,
+      "grad_norm": 0.40839269757270813,
+      "learning_rate": 6.930303745494859e-07,
+      "loss": 2.7967,
+      "step": 7685
+    },
+    {
+      "epoch": 0.983808,
+      "grad_norm": 0.43114325404167175,
+      "learning_rate": 6.821619509365662e-07,
+      "loss": 2.8684,
+      "step": 7686
+    },
+    {
+      "epoch": 0.983936,
+      "grad_norm": 0.4007643461227417,
+      "learning_rate": 6.713793658486012e-07,
+      "loss": 2.7024,
+      "step": 7687
+    },
+    {
+      "epoch": 0.984064,
+      "grad_norm": 0.41818326711654663,
+      "learning_rate": 6.606826211392192e-07,
+      "loss": 2.7495,
+      "step": 7688
+    },
+    {
+      "epoch": 0.984192,
+      "grad_norm": 0.42143189907073975,
+      "learning_rate": 6.500717186473382e-07,
+      "loss": 2.7802,
+      "step": 7689
+    },
+    {
+      "epoch": 0.98432,
+      "grad_norm": 0.451842725276947,
+      "learning_rate": 6.3954666019711e-07,
+      "loss": 2.8253,
+      "step": 7690
+    },
+    {
+      "epoch": 0.984448,
+      "grad_norm": 0.3851945400238037,
+      "learning_rate": 6.291074475978653e-07,
+      "loss": 2.5816,
+      "step": 7691
+    },
+    {
+      "epoch": 0.984576,
+      "grad_norm": 0.40176764130592346,
+      "learning_rate": 6.187540826442795e-07,
+      "loss": 2.6632,
+      "step": 7692
+    },
+    {
+      "epoch": 0.984704,
+      "grad_norm": 0.38533180952072144,
+      "learning_rate": 6.084865671162065e-07,
+      "loss": 2.7813,
+      "step": 7693
+    },
+    {
+      "epoch": 0.984832,
+      "grad_norm": 0.393012672662735,
+      "learning_rate": 5.983049027786791e-07,
+      "loss": 2.9153,
+      "step": 7694
+    },
+    {
+      "epoch": 0.98496,
+      "grad_norm": 0.41062161326408386,
+      "learning_rate": 5.882090913821303e-07,
+      "loss": 2.8522,
+      "step": 7695
+    },
+    {
+      "epoch": 0.985088,
+      "grad_norm": 0.41157999634742737,
+      "learning_rate": 5.781991346621163e-07,
+      "loss": 2.817,
+      "step": 7696
+    },
+    {
+      "epoch": 0.985216,
+      "grad_norm": 0.41961047053337097,
+      "learning_rate": 5.682750343394826e-07,
+      "loss": 2.8129,
+      "step": 7697
+    },
+    {
+      "epoch": 0.985344,
+      "grad_norm": 0.4066692292690277,
+      "learning_rate": 5.584367921202538e-07,
+      "loss": 2.607,
+      "step": 7698
+    },
+    {
+      "epoch": 0.985472,
+      "grad_norm": 0.452188104391098,
+      "learning_rate": 5.486844096957988e-07,
+      "loss": 2.8588,
+      "step": 7699
+    },
+    {
+      "epoch": 0.9856,
+      "grad_norm": 0.42160680890083313,
+      "learning_rate": 5.3901788874261e-07,
+      "loss": 2.811,
+      "step": 7700
+    },
+    {
+      "epoch": 0.985728,
+      "grad_norm": 0.4019690454006195,
+      "learning_rate": 5.294372309224693e-07,
+      "loss": 2.6383,
+      "step": 7701
+    },
+    {
+      "epoch": 0.985856,
+      "grad_norm": 0.457445353269577,
+      "learning_rate": 5.19942437882448e-07,
+      "loss": 2.8002,
+      "step": 7702
+    },
+    {
+      "epoch": 0.985984,
+      "grad_norm": 0.40246522426605225,
+      "learning_rate": 5.10533511254796e-07,
+      "loss": 2.7796,
+      "step": 7703
+    },
+    {
+      "epoch": 0.986112,
+      "grad_norm": 0.38974031805992126,
+      "learning_rate": 5.012104526569972e-07,
+      "loss": 2.863,
+      "step": 7704
+    },
+    {
+      "epoch": 0.98624,
+      "grad_norm": 0.43098655343055725,
+      "learning_rate": 4.919732636918251e-07,
+      "loss": 2.8378,
+      "step": 7705
+    },
+    {
+      "epoch": 0.986368,
+      "grad_norm": 0.4006861746311188,
+      "learning_rate": 4.828219459472316e-07,
+      "loss": 2.9742,
+      "step": 7706
+    },
+    {
+      "epoch": 0.986496,
+      "grad_norm": 0.43582355976104736,
+      "learning_rate": 4.737565009964584e-07,
+      "loss": 2.865,
+      "step": 7707
+    },
+    {
+      "epoch": 0.986624,
+      "grad_norm": 0.40844815969467163,
+      "learning_rate": 4.647769303979255e-07,
+      "loss": 2.7895,
+      "step": 7708
+    },
+    {
+      "epoch": 0.986752,
+      "grad_norm": 0.41827887296676636,
+      "learning_rate": 4.558832356953979e-07,
+      "loss": 2.723,
+      "step": 7709
+    },
+    {
+      "epoch": 0.98688,
+      "grad_norm": 0.46037885546684265,
+      "learning_rate": 4.470754184177084e-07,
+      "loss": 2.8876,
+      "step": 7710
+    },
+    {
+      "epoch": 0.987008,
+      "grad_norm": 0.41751810908317566,
+      "learning_rate": 4.3835348007909004e-07,
+      "loss": 2.9055,
+      "step": 7711
+    },
+    {
+      "epoch": 0.987136,
+      "grad_norm": 0.40196916460990906,
+      "learning_rate": 4.297174221788991e-07,
+      "loss": 2.7571,
+      "step": 7712
+    },
+    {
+      "epoch": 0.987264,
+      "grad_norm": 0.39552897214889526,
+      "learning_rate": 4.2116724620183675e-07,
+      "loss": 2.7412,
+      "step": 7713
+    },
+    {
+      "epoch": 0.987392,
+      "grad_norm": 0.42975854873657227,
+      "learning_rate": 4.1270295361772737e-07,
+      "loss": 2.7532,
+      "step": 7714
+    },
+    {
+      "epoch": 0.98752,
+      "grad_norm": 0.41349369287490845,
+      "learning_rate": 4.043245458816847e-07,
+      "loss": 2.8442,
+      "step": 7715
+    },
+    {
+      "epoch": 0.987648,
+      "grad_norm": 0.3965969979763031,
+      "learning_rate": 3.960320244340565e-07,
+      "loss": 2.7164,
+      "step": 7716
+    },
+    {
+      "epoch": 0.987776,
+      "grad_norm": 0.4461132884025574,
+      "learning_rate": 3.878253907004248e-07,
+      "loss": 2.7685,
+      "step": 7717
+    },
+    {
+      "epoch": 0.987904,
+      "grad_norm": 0.42018136382102966,
+      "learning_rate": 3.7970464609160535e-07,
+      "loss": 2.7993,
+      "step": 7718
+    },
+    {
+      "epoch": 0.988032,
+      "grad_norm": 0.4150193929672241,
+      "learning_rate": 3.7166979200364826e-07,
+      "loss": 2.7478,
+      "step": 7719
+    },
+    {
+      "epoch": 0.98816,
+      "grad_norm": 0.4257791042327881,
+      "learning_rate": 3.637208298178929e-07,
+      "loss": 2.7997,
+      "step": 7720
+    },
+    {
+      "epoch": 0.988288,
+      "grad_norm": 0.4320071339607239,
+      "learning_rate": 3.558577609007463e-07,
+      "loss": 2.8544,
+      "step": 7721
+    },
+    {
+      "epoch": 0.988416,
+      "grad_norm": 0.4202640652656555,
+      "learning_rate": 3.480805866040715e-07,
+      "loss": 2.7461,
+      "step": 7722
+    },
+    {
+      "epoch": 0.988544,
+      "grad_norm": 0.38627561926841736,
+      "learning_rate": 3.403893082647991e-07,
+      "loss": 2.8419,
+      "step": 7723
+    },
+    {
+      "epoch": 0.988672,
+      "grad_norm": 0.432844877243042,
+      "learning_rate": 3.3278392720514915e-07,
+      "loss": 2.7579,
+      "step": 7724
+    },
+    {
+      "epoch": 0.9888,
+      "grad_norm": 0.4116506576538086,
+      "learning_rate": 3.252644447325759e-07,
+      "loss": 2.7548,
+      "step": 7725
+    },
+    {
+      "epoch": 0.988928,
+      "grad_norm": 0.39209485054016113,
+      "learning_rate": 3.178308621397674e-07,
+      "loss": 2.8065,
+      "step": 7726
+    },
+    {
+      "epoch": 0.989056,
+      "grad_norm": 0.3943476974964142,
+      "learning_rate": 3.10483180704646e-07,
+      "loss": 2.7812,
+      "step": 7727
+    },
+    {
+      "epoch": 0.989184,
+      "grad_norm": 0.4391293227672577,
+      "learning_rate": 3.0322140169036784e-07,
+      "loss": 2.8205,
+      "step": 7728
+    },
+    {
+      "epoch": 0.989312,
+      "grad_norm": 0.453749418258667,
+      "learning_rate": 2.9604552634532324e-07,
+      "loss": 2.8546,
+      "step": 7729
+    },
+    {
+      "epoch": 0.98944,
+      "grad_norm": 0.4463576674461365,
+      "learning_rate": 2.88955555903081e-07,
+      "loss": 2.7527,
+      "step": 7730
+    },
+    {
+      "epoch": 0.989568,
+      "grad_norm": 0.42400839924812317,
+      "learning_rate": 2.8195149158255497e-07,
+      "loss": 2.7581,
+      "step": 7731
+    },
+    {
+      "epoch": 0.989696,
+      "grad_norm": 0.39631354808807373,
+      "learning_rate": 2.75033334587782e-07,
+      "loss": 2.712,
+      "step": 7732
+    },
+    {
+      "epoch": 0.989824,
+      "grad_norm": 0.3739931583404541,
+      "learning_rate": 2.6820108610808855e-07,
+      "loss": 2.8684,
+      "step": 7733
+    },
+    {
+      "epoch": 0.989952,
+      "grad_norm": 0.3871767818927765,
+      "learning_rate": 2.614547473180351e-07,
+      "loss": 2.8882,
+      "step": 7734
+    },
+    {
+      "epoch": 0.99008,
+      "grad_norm": 0.4185982942581177,
+      "learning_rate": 2.5479431937736055e-07,
+      "loss": 2.7475,
+      "step": 7735
+    },
+    {
+      "epoch": 0.990208,
+      "grad_norm": 0.388759046792984,
+      "learning_rate": 2.482198034310934e-07,
+      "loss": 2.8511,
+      "step": 7736
+    },
+    {
+      "epoch": 0.990336,
+      "grad_norm": 0.452487587928772,
+      "learning_rate": 2.4173120060949626e-07,
+      "loss": 2.8186,
+      "step": 7737
+    },
+    {
+      "epoch": 0.990464,
+      "grad_norm": 0.4835643172264099,
+      "learning_rate": 2.353285120279547e-07,
+      "loss": 2.7591,
+      "step": 7738
+    },
+    {
+      "epoch": 0.990592,
+      "grad_norm": 0.43403810262680054,
+      "learning_rate": 2.2901173878725478e-07,
+      "loss": 2.7832,
+      "step": 7739
+    },
+    {
+      "epoch": 0.99072,
+      "grad_norm": 0.4090152680873871,
+      "learning_rate": 2.2278088197325018e-07,
+      "loss": 2.762,
+      "step": 7740
+    },
+    {
+      "epoch": 0.990848,
+      "grad_norm": 0.4667512774467468,
+      "learning_rate": 2.1663594265708407e-07,
+      "loss": 2.8493,
+      "step": 7741
+    },
+    {
+      "epoch": 0.990976,
+      "grad_norm": 0.4108279347419739,
+      "learning_rate": 2.1057692189524468e-07,
+      "loss": 2.8018,
+      "step": 7742
+    },
+    {
+      "epoch": 0.991104,
+      "grad_norm": 0.3998432159423828,
+      "learning_rate": 2.0460382072928775e-07,
+      "loss": 2.7033,
+      "step": 7743
+    },
+    {
+      "epoch": 0.991232,
+      "grad_norm": 0.415486603975296,
+      "learning_rate": 1.98716640186003e-07,
+      "loss": 2.7505,
+      "step": 7744
+    },
+    {
+      "epoch": 0.99136,
+      "grad_norm": 0.3870699405670166,
+      "learning_rate": 1.9291538127758078e-07,
+      "loss": 2.7838,
+      "step": 7745
+    },
+    {
+      "epoch": 0.991488,
+      "grad_norm": 0.40355563163757324,
+      "learning_rate": 1.8720004500122345e-07,
+      "loss": 2.7965,
+      "step": 7746
+    },
+    {
+      "epoch": 0.991616,
+      "grad_norm": 0.4022752642631531,
+      "learning_rate": 1.8157063233953387e-07,
+      "loss": 2.8348,
+      "step": 7747
+    },
+    {
+      "epoch": 0.991744,
+      "grad_norm": 0.42166921496391296,
+      "learning_rate": 1.7602714426023791e-07,
+      "loss": 2.7597,
+      "step": 7748
+    },
+    {
+      "epoch": 0.991872,
+      "grad_norm": 0.4102909564971924,
+      "learning_rate": 1.7056958171635108e-07,
+      "loss": 2.7222,
+      "step": 7749
+    },
+    {
+      "epoch": 0.992,
+      "grad_norm": 0.4390327036380768,
+      "learning_rate": 1.6519794564606727e-07,
+      "loss": 2.8775,
+      "step": 7750
+    },
+    {
+      "epoch": 0.992128,
+      "grad_norm": 0.4221785366535187,
+      "learning_rate": 1.5991223697281455e-07,
+      "loss": 2.7883,
+      "step": 7751
+    },
+    {
+      "epoch": 0.992256,
+      "grad_norm": 0.4124399721622467,
+      "learning_rate": 1.5471245660531042e-07,
+      "loss": 2.6867,
+      "step": 7752
+    },
+    {
+      "epoch": 0.992384,
+      "grad_norm": 0.398301362991333,
+      "learning_rate": 1.4959860543739546e-07,
+      "loss": 2.7359,
+      "step": 7753
+    },
+    {
+      "epoch": 0.992512,
+      "grad_norm": 0.4002898931503296,
+      "learning_rate": 1.4457068434831077e-07,
+      "loss": 2.8181,
+      "step": 7754
+    },
+    {
+      "epoch": 0.99264,
+      "grad_norm": 0.44058921933174133,
+      "learning_rate": 1.396286942023095e-07,
+      "loss": 2.6868,
+      "step": 7755
+    },
+    {
+      "epoch": 0.992768,
+      "grad_norm": 0.41495585441589355,
+      "learning_rate": 1.3477263584904532e-07,
+      "loss": 2.7087,
+      "step": 7756
+    },
+    {
+      "epoch": 0.992896,
+      "grad_norm": 0.4127732813358307,
+      "learning_rate": 1.3000251012323938e-07,
+      "loss": 2.7641,
+      "step": 7757
+    },
+    {
+      "epoch": 0.993024,
+      "grad_norm": 0.41915759444236755,
+      "learning_rate": 1.2531831784506897e-07,
+      "loss": 2.7087,
+      "step": 7758
+    },
+    {
+      "epoch": 0.993152,
+      "grad_norm": 0.41958460211753845,
+      "learning_rate": 1.2072005981966781e-07,
+      "loss": 2.7148,
+      "step": 7759
+    },
+    {
+      "epoch": 0.99328,
+      "grad_norm": 0.4300496280193329,
+      "learning_rate": 1.1620773683757024e-07,
+      "loss": 2.7817,
+      "step": 7760
+    },
+    {
+      "epoch": 0.993408,
+      "grad_norm": 0.4929201900959015,
+      "learning_rate": 1.1178134967454457e-07,
+      "loss": 2.8724,
+      "step": 7761
+    },
+    {
+      "epoch": 0.993536,
+      "grad_norm": 0.40605300664901733,
+      "learning_rate": 1.0744089909153765e-07,
+      "loss": 2.8789,
+      "step": 7762
+    },
+    {
+      "epoch": 0.993664,
+      "grad_norm": 0.40936776995658875,
+      "learning_rate": 1.0318638583467488e-07,
+      "loss": 2.7088,
+      "step": 7763
+    },
+    {
+      "epoch": 0.993792,
+      "grad_norm": 0.41030028462409973,
+      "learning_rate": 9.901781063531568e-08,
+      "loss": 2.772,
+      "step": 7764
+    },
+    {
+      "epoch": 0.99392,
+      "grad_norm": 0.4081823229789734,
+      "learning_rate": 9.493517421022002e-08,
+      "loss": 2.903,
+      "step": 7765
+    },
+    {
+      "epoch": 0.994048,
+      "grad_norm": 0.4112604558467865,
+      "learning_rate": 9.093847726110438e-08,
+      "loss": 2.8226,
+      "step": 7766
+    },
+    {
+      "epoch": 0.994176,
+      "grad_norm": 0.40158337354660034,
+      "learning_rate": 8.702772047514129e-08,
+      "loss": 2.8632,
+      "step": 7767
+    },
+    {
+      "epoch": 0.994304,
+      "grad_norm": 0.3741767406463623,
+      "learning_rate": 8.320290452462631e-08,
+      "loss": 2.8392,
+      "step": 7768
+    },
+    {
+      "epoch": 0.994432,
+      "grad_norm": 0.39240360260009766,
+      "learning_rate": 7.94640300670335e-08,
+      "loss": 2.8719,
+      "step": 7769
+    },
+    {
+      "epoch": 0.99456,
+      "grad_norm": 0.37618809938430786,
+      "learning_rate": 7.581109774512652e-08,
+      "loss": 2.7918,
+      "step": 7770
+    },
+    {
+      "epoch": 0.994688,
+      "grad_norm": 0.4048091471195221,
+      "learning_rate": 7.224410818695848e-08,
+      "loss": 2.7518,
+      "step": 7771
+    },
+    {
+      "epoch": 0.994816,
+      "grad_norm": 0.42344364523887634,
+      "learning_rate": 6.876306200565009e-08,
+      "loss": 2.8255,
+      "step": 7772
+    },
+    {
+      "epoch": 0.994944,
+      "grad_norm": 0.41678324341773987,
+      "learning_rate": 6.536795979966703e-08,
+      "loss": 2.7219,
+      "step": 7773
+    },
+    {
+      "epoch": 0.995072,
+      "grad_norm": 0.40984028577804565,
+      "learning_rate": 6.205880215270909e-08,
+      "loss": 2.7387,
+      "step": 7774
+    },
+    {
+      "epoch": 0.9952,
+      "grad_norm": 0.39151155948638916,
+      "learning_rate": 5.883558963359903e-08,
+      "loss": 2.7233,
+      "step": 7775
+    },
+    {
+      "epoch": 0.995328,
+      "grad_norm": 0.41418227553367615,
+      "learning_rate": 5.569832279644915e-08,
+      "loss": 2.7107,
+      "step": 7776
+    },
+    {
+      "epoch": 0.995456,
+      "grad_norm": 0.42729538679122925,
+      "learning_rate": 5.264700218066132e-08,
+      "loss": 2.7034,
+      "step": 7777
+    },
+    {
+      "epoch": 0.995584,
+      "grad_norm": 0.3867628574371338,
+      "learning_rate": 4.968162831070488e-08,
+      "loss": 2.6929,
+      "step": 7778
+    },
+    {
+      "epoch": 0.995712,
+      "grad_norm": 0.3937319815158844,
+      "learning_rate": 4.6802201696394264e-08,
+      "loss": 2.7338,
+      "step": 7779
+    },
+    {
+      "epoch": 0.99584,
+      "grad_norm": 0.38750770688056946,
+      "learning_rate": 4.400872283272239e-08,
+      "loss": 2.7726,
+      "step": 7780
+    },
+    {
+      "epoch": 0.995968,
+      "grad_norm": 0.42572107911109924,
+      "learning_rate": 4.130119219997175e-08,
+      "loss": 2.8096,
+      "step": 7781
+    },
+    {
+      "epoch": 0.996096,
+      "grad_norm": 0.43103939294815063,
+      "learning_rate": 3.8679610263547824e-08,
+      "loss": 2.8267,
+      "step": 7782
+    },
+    {
+      "epoch": 0.996224,
+      "grad_norm": 0.40234842896461487,
+      "learning_rate": 3.614397747420117e-08,
+      "loss": 2.7675,
+      "step": 7783
+    },
+    {
+      "epoch": 0.996352,
+      "grad_norm": 0.4140002429485321,
+      "learning_rate": 3.3694294267694326e-08,
+      "loss": 2.7705,
+      "step": 7784
+    },
+    {
+      "epoch": 0.99648,
+      "grad_norm": 0.41892483830451965,
+      "learning_rate": 3.133056106530141e-08,
+      "loss": 2.8533,
+      "step": 7785
+    },
+    {
+      "epoch": 0.996608,
+      "grad_norm": 0.4111126661300659,
+      "learning_rate": 2.9052778273308545e-08,
+      "loss": 2.7567,
+      "step": 7786
+    },
+    {
+      "epoch": 0.996736,
+      "grad_norm": 0.4092635214328766,
+      "learning_rate": 2.6860946283291388e-08,
+      "loss": 2.6598,
+      "step": 7787
+    },
+    {
+      "epoch": 0.996864,
+      "grad_norm": 0.4005700349807739,
+      "learning_rate": 2.4755065472115144e-08,
+      "loss": 2.8131,
+      "step": 7788
+    },
+    {
+      "epoch": 0.996992,
+      "grad_norm": 0.404610276222229,
+      "learning_rate": 2.2735136201712524e-08,
+      "loss": 2.7571,
+      "step": 7789
+    },
+    {
+      "epoch": 0.99712,
+      "grad_norm": 0.41123151779174805,
+      "learning_rate": 2.0801158819361287e-08,
+      "loss": 2.7643,
+      "step": 7790
+    },
+    {
+      "epoch": 0.997248,
+      "grad_norm": 0.4139678180217743,
+      "learning_rate": 1.895313365757323e-08,
+      "loss": 2.851,
+      "step": 7791
+    },
+    {
+      "epoch": 0.997376,
+      "grad_norm": 0.4951436221599579,
+      "learning_rate": 1.7191061033983158e-08,
+      "loss": 2.7068,
+      "step": 7792
+    },
+    {
+      "epoch": 0.997504,
+      "grad_norm": 0.4071565270423889,
+      "learning_rate": 1.5514941251570934e-08,
+      "loss": 2.7903,
+      "step": 7793
+    },
+    {
+      "epoch": 0.997632,
+      "grad_norm": 0.4379862844944,
+      "learning_rate": 1.3924774598439438e-08,
+      "loss": 2.7605,
+      "step": 7794
+    },
+    {
+      "epoch": 0.99776,
+      "grad_norm": 0.3959049880504608,
+      "learning_rate": 1.2420561348036597e-08,
+      "loss": 2.6612,
+      "step": 7795
+    },
+    {
+      "epoch": 0.997888,
+      "grad_norm": 0.45005959272384644,
+      "learning_rate": 1.100230175882233e-08,
+      "loss": 2.8513,
+      "step": 7796
+    },
+    {
+      "epoch": 0.998016,
+      "grad_norm": 0.42590606212615967,
+      "learning_rate": 9.669996074712639e-09,
+      "loss": 2.7712,
+      "step": 7797
+    },
+    {
+      "epoch": 0.998144,
+      "grad_norm": 0.49947336316108704,
+      "learning_rate": 8.423644524691021e-09,
+      "loss": 2.6874,
+      "step": 7798
+    },
+    {
+      "epoch": 0.998272,
+      "grad_norm": 0.39834997057914734,
+      "learning_rate": 7.263247323086031e-09,
+      "loss": 2.7302,
+      "step": 7799
+    },
+    {
+      "epoch": 0.9984,
+      "grad_norm": 0.42241621017456055,
+      "learning_rate": 6.188804669349235e-09,
+      "loss": 2.6042,
+      "step": 7800
+    },
+    {
+      "epoch": 0.998528,
+      "grad_norm": 0.395322322845459,
+      "learning_rate": 5.200316748110723e-09,
+      "loss": 2.8518,
+      "step": 7801
+    },
+    {
+      "epoch": 0.998656,
+      "grad_norm": 0.3784806728363037,
+      "learning_rate": 4.297783729456661e-09,
+      "loss": 2.7808,
+      "step": 7802
+    },
+    {
+      "epoch": 0.998784,
+      "grad_norm": 0.45947280526161194,
+      "learning_rate": 3.4812057684296962e-09,
+      "loss": 2.7958,
+      "step": 7803
+    },
+    {
+      "epoch": 0.998912,
+      "grad_norm": 0.4185103178024292,
+      "learning_rate": 2.7505830054175284e-09,
+      "loss": 2.8269,
+      "step": 7804
+    },
+    {
+      "epoch": 0.99904,
+      "grad_norm": 0.4232789874076843,
+      "learning_rate": 2.1059155660418936e-09,
+      "loss": 2.7998,
+      "step": 7805
+    },
+    {
+      "epoch": 0.999168,
+      "grad_norm": 0.4505091905593872,
+      "learning_rate": 1.547203561158561e-09,
+      "loss": 2.7368,
+      "step": 7806
+    },
+    {
+      "epoch": 0.999296,
+      "grad_norm": 0.38175198435783386,
+      "learning_rate": 1.074447086801822e-09,
+      "loss": 2.7399,
+      "step": 7807
+    },
+    {
+      "epoch": 0.999424,
+      "grad_norm": 0.3915434777736664,
+      "learning_rate": 6.876462241844905e-10,
+      "loss": 2.8112,
+      "step": 7808
+    },
+    {
+      "epoch": 0.999552,
+      "grad_norm": 0.42164915800094604,
+      "learning_rate": 3.8680103991994844e-10,
+      "loss": 2.6324,
+      "step": 7809
+    },
+    {
+      "epoch": 0.99968,
+      "grad_norm": 0.3984890282154083,
+      "learning_rate": 1.7191158563356623e-10,
+      "loss": 2.8101,
+      "step": 7810
+    },
+    {
+      "epoch": 0.999808,
+      "grad_norm": 0.4001835286617279,
+      "learning_rate": 4.297789824025955e-11,
+      "loss": 2.7234,
+      "step": 7811
+    },
+    {
+      "epoch": 0.999936,
+      "grad_norm": 0.4004862904548645,
+      "learning_rate": 0.0,
+      "loss": 2.8574,
+      "step": 7812
+    },
+    {
+      "epoch": 0.999936,
+      "step": 7812,
+      "total_flos": 3.9660414340327014e+18,
+      "train_loss": 2.9187875806644397,
+      "train_runtime": 58888.3895,
+      "train_samples_per_second": 33.963,
+      "train_steps_per_second": 0.133
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 7812,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 24000,
+  "total_flos": 3.9660414340327014e+18,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}