diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,59687 @@
+{
+  "best_metric": 1.2935380935668945,
+  "best_model_checkpoint": "/export/data/salmasia/tradutor/checkpoints/hf_llama3_lora/checkpoint-7000",
+  "epoch": 3.1645937801770776,
+  "eval_steps": 500,
+  "global_step": 8500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00037230515060906794,
+      "grad_norm": 1.4287878274917603,
+      "learning_rate": 2e-08,
+      "loss": 1.8524,
+      "step": 1
+    },
+    {
+      "epoch": 0.0007446103012181359,
+      "grad_norm": 1.438577651977539,
+      "learning_rate": 4e-08,
+      "loss": 1.8624,
+      "step": 2
+    },
+    {
+      "epoch": 0.001116915451827204,
+      "grad_norm": 1.4175422191619873,
+      "learning_rate": 6.000000000000001e-08,
+      "loss": 1.8535,
+      "step": 3
+    },
+    {
+      "epoch": 0.0014892206024362717,
+      "grad_norm": 1.3822880983352661,
+      "learning_rate": 8e-08,
+      "loss": 1.8581,
+      "step": 4
+    },
+    {
+      "epoch": 0.0018615257530453398,
+      "grad_norm": 1.4041341543197632,
+      "learning_rate": 1.0000000000000001e-07,
+      "loss": 1.8706,
+      "step": 5
+    },
+    {
+      "epoch": 0.002233830903654408,
+      "grad_norm": 1.352337121963501,
+      "learning_rate": 1.2000000000000002e-07,
+      "loss": 1.8438,
+      "step": 6
+    },
+    {
+      "epoch": 0.002606136054263476,
+      "grad_norm": 1.3975361585617065,
+      "learning_rate": 1.4e-07,
+      "loss": 1.8411,
+      "step": 7
+    },
+    {
+      "epoch": 0.0029784412048725435,
+      "grad_norm": 1.3485698699951172,
+      "learning_rate": 1.6e-07,
+      "loss": 1.8473,
+      "step": 8
+    },
+    {
+      "epoch": 0.0033507463554816115,
+      "grad_norm": 1.4282227754592896,
+      "learning_rate": 1.8e-07,
+      "loss": 1.8677,
+      "step": 9
+    },
+    {
+      "epoch": 0.0037230515060906796,
+      "grad_norm": 1.3968946933746338,
+      "learning_rate": 2.0000000000000002e-07,
+      "loss": 1.876,
+      "step": 10
+    },
+    {
+      "epoch": 0.004095356656699747,
+      "grad_norm": 1.4412802457809448,
+      "learning_rate": 2.2e-07,
+      "loss": 1.8736,
+      "step": 11
+    },
+    {
+      "epoch": 0.004467661807308816,
+      "grad_norm": 1.3732651472091675,
+      "learning_rate": 2.4000000000000003e-07,
+      "loss": 1.8566,
+      "step": 12
+    },
+    {
+      "epoch": 0.004839966957917883,
+      "grad_norm": 1.444575309753418,
+      "learning_rate": 2.6e-07,
+      "loss": 1.8654,
+      "step": 13
+    },
+    {
+      "epoch": 0.005212272108526952,
+      "grad_norm": 1.4188005924224854,
+      "learning_rate": 2.8e-07,
+      "loss": 1.8672,
+      "step": 14
+    },
+    {
+      "epoch": 0.005584577259136019,
+      "grad_norm": 1.3773771524429321,
+      "learning_rate": 3.0000000000000004e-07,
+      "loss": 1.8475,
+      "step": 15
+    },
+    {
+      "epoch": 0.005956882409745087,
+      "grad_norm": 1.4325660467147827,
+      "learning_rate": 3.2e-07,
+      "loss": 1.8845,
+      "step": 16
+    },
+    {
+      "epoch": 0.0063291875603541554,
+      "grad_norm": 1.366593599319458,
+      "learning_rate": 3.4000000000000003e-07,
+      "loss": 1.8627,
+      "step": 17
+    },
+    {
+      "epoch": 0.006701492710963223,
+      "grad_norm": 1.3906618356704712,
+      "learning_rate": 3.6e-07,
+      "loss": 1.8424,
+      "step": 18
+    },
+    {
+      "epoch": 0.0070737978615722915,
+      "grad_norm": 1.3560819625854492,
+      "learning_rate": 3.8e-07,
+      "loss": 1.8623,
+      "step": 19
+    },
+    {
+      "epoch": 0.007446103012181359,
+      "grad_norm": 1.3274669647216797,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 1.8461,
+      "step": 20
+    },
+    {
+      "epoch": 0.007818408162790428,
+      "grad_norm": 1.3316004276275635,
+      "learning_rate": 4.2000000000000006e-07,
+      "loss": 1.8471,
+      "step": 21
+    },
+    {
+      "epoch": 0.008190713313399494,
+      "grad_norm": 1.3909695148468018,
+      "learning_rate": 4.4e-07,
+      "loss": 1.8484,
+      "step": 22
+    },
+    {
+      "epoch": 0.008563018464008563,
+      "grad_norm": 1.3994210958480835,
+      "learning_rate": 4.6000000000000004e-07,
+      "loss": 1.8488,
+      "step": 23
+    },
+    {
+      "epoch": 0.008935323614617631,
+      "grad_norm": 1.4674007892608643,
+      "learning_rate": 4.800000000000001e-07,
+      "loss": 1.8548,
+      "step": 24
+    },
+    {
+      "epoch": 0.0093076287652267,
+      "grad_norm": 1.4008136987686157,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 1.8322,
+      "step": 25
+    },
+    {
+      "epoch": 0.009679933915835767,
+      "grad_norm": 1.386723279953003,
+      "learning_rate": 5.2e-07,
+      "loss": 1.8451,
+      "step": 26
+    },
+    {
+      "epoch": 0.010052239066444835,
+      "grad_norm": 1.398271918296814,
+      "learning_rate": 5.4e-07,
+      "loss": 1.8389,
+      "step": 27
+    },
+    {
+      "epoch": 0.010424544217053903,
+      "grad_norm": 1.4241478443145752,
+      "learning_rate": 5.6e-07,
+      "loss": 1.8636,
+      "step": 28
+    },
+    {
+      "epoch": 0.01079684936766297,
+      "grad_norm": 1.4040849208831787,
+      "learning_rate": 5.800000000000001e-07,
+      "loss": 1.8427,
+      "step": 29
+    },
+    {
+      "epoch": 0.011169154518272039,
+      "grad_norm": 1.3591426610946655,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 1.8354,
+      "step": 30
+    },
+    {
+      "epoch": 0.011541459668881107,
+      "grad_norm": 1.4355502128601074,
+      "learning_rate": 6.200000000000001e-07,
+      "loss": 1.8536,
+      "step": 31
+    },
+    {
+      "epoch": 0.011913764819490174,
+      "grad_norm": 1.4141122102737427,
+      "learning_rate": 6.4e-07,
+      "loss": 1.8295,
+      "step": 32
+    },
+    {
+      "epoch": 0.012286069970099242,
+      "grad_norm": 1.4399964809417725,
+      "learning_rate": 6.6e-07,
+      "loss": 1.8555,
+      "step": 33
+    },
+    {
+      "epoch": 0.012658375120708311,
+      "grad_norm": 1.3795485496520996,
+      "learning_rate": 6.800000000000001e-07,
+      "loss": 1.8304,
+      "step": 34
+    },
+    {
+      "epoch": 0.01303068027131738,
+      "grad_norm": 1.4039381742477417,
+      "learning_rate": 7.000000000000001e-07,
+      "loss": 1.8467,
+      "step": 35
+    },
+    {
+      "epoch": 0.013402985421926446,
+      "grad_norm": 1.4005100727081299,
+      "learning_rate": 7.2e-07,
+      "loss": 1.8253,
+      "step": 36
+    },
+    {
+      "epoch": 0.013775290572535515,
+      "grad_norm": 1.415635347366333,
+      "learning_rate": 7.4e-07,
+      "loss": 1.8458,
+      "step": 37
+    },
+    {
+      "epoch": 0.014147595723144583,
+      "grad_norm": 1.39545738697052,
+      "learning_rate": 7.6e-07,
+      "loss": 1.8264,
+      "step": 38
+    },
+    {
+      "epoch": 0.01451990087375365,
+      "grad_norm": 1.3957033157348633,
+      "learning_rate": 7.8e-07,
+      "loss": 1.8404,
+      "step": 39
+    },
+    {
+      "epoch": 0.014892206024362718,
+      "grad_norm": 1.3806068897247314,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 1.828,
+      "step": 40
+    },
+    {
+      "epoch": 0.015264511174971787,
+      "grad_norm": 1.387890338897705,
+      "learning_rate": 8.200000000000001e-07,
+      "loss": 1.8302,
+      "step": 41
+    },
+    {
+      "epoch": 0.015636816325580855,
+      "grad_norm": 1.3517177104949951,
+      "learning_rate": 8.400000000000001e-07,
+      "loss": 1.8021,
+      "step": 42
+    },
+    {
+      "epoch": 0.016009121476189922,
+      "grad_norm": 1.3489928245544434,
+      "learning_rate": 8.6e-07,
+      "loss": 1.8084,
+      "step": 43
+    },
+    {
+      "epoch": 0.01638142662679899,
+      "grad_norm": 1.352052092552185,
+      "learning_rate": 8.8e-07,
+      "loss": 1.8143,
+      "step": 44
+    },
+    {
+      "epoch": 0.01675373177740806,
+      "grad_norm": 1.324896216392517,
+      "learning_rate": 9.000000000000001e-07,
+      "loss": 1.7936,
+      "step": 45
+    },
+    {
+      "epoch": 0.017126036928017126,
+      "grad_norm": 1.2951364517211914,
+      "learning_rate": 9.200000000000001e-07,
+      "loss": 1.8092,
+      "step": 46
+    },
+    {
+      "epoch": 0.017498342078626192,
+      "grad_norm": 1.3121585845947266,
+      "learning_rate": 9.400000000000001e-07,
+      "loss": 1.8096,
+      "step": 47
+    },
+    {
+      "epoch": 0.017870647229235263,
+      "grad_norm": 1.359284520149231,
+      "learning_rate": 9.600000000000001e-07,
+      "loss": 1.822,
+      "step": 48
+    },
+    {
+      "epoch": 0.01824295237984433,
+      "grad_norm": 1.3160185813903809,
+      "learning_rate": 9.800000000000001e-07,
+      "loss": 1.8113,
+      "step": 49
+    },
+    {
+      "epoch": 0.0186152575304534,
+      "grad_norm": 1.2798627614974976,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 1.7942,
+      "step": 50
+    },
+    {
+      "epoch": 0.018987562681062466,
+      "grad_norm": 1.3182300329208374,
+      "learning_rate": 1.02e-06,
+      "loss": 1.8119,
+      "step": 51
+    },
+    {
+      "epoch": 0.019359867831671533,
+      "grad_norm": 1.3083176612854004,
+      "learning_rate": 1.04e-06,
+      "loss": 1.8189,
+      "step": 52
+    },
+    {
+      "epoch": 0.019732172982280603,
+      "grad_norm": 1.3400063514709473,
+      "learning_rate": 1.06e-06,
+      "loss": 1.837,
+      "step": 53
+    },
+    {
+      "epoch": 0.02010447813288967,
+      "grad_norm": 1.2751755714416504,
+      "learning_rate": 1.08e-06,
+      "loss": 1.7936,
+      "step": 54
+    },
+    {
+      "epoch": 0.020476783283498737,
+      "grad_norm": 1.223643183708191,
+      "learning_rate": 1.1e-06,
+      "loss": 1.7851,
+      "step": 55
+    },
+    {
+      "epoch": 0.020849088434107807,
+      "grad_norm": 1.2413800954818726,
+      "learning_rate": 1.12e-06,
+      "loss": 1.7893,
+      "step": 56
+    },
+    {
+      "epoch": 0.021221393584716874,
+      "grad_norm": 1.196629524230957,
+      "learning_rate": 1.14e-06,
+      "loss": 1.8031,
+      "step": 57
+    },
+    {
+      "epoch": 0.02159369873532594,
+      "grad_norm": 1.158254861831665,
+      "learning_rate": 1.1600000000000001e-06,
+      "loss": 1.79,
+      "step": 58
+    },
+    {
+      "epoch": 0.02196600388593501,
+      "grad_norm": 1.0098892450332642,
+      "learning_rate": 1.1800000000000001e-06,
+      "loss": 1.7772,
+      "step": 59
+    },
+    {
+      "epoch": 0.022338309036544077,
+      "grad_norm": 0.949759304523468,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 1.7582,
+      "step": 60
+    },
+    {
+      "epoch": 0.022710614187153144,
+      "grad_norm": 0.9904881119728088,
+      "learning_rate": 1.2200000000000002e-06,
+      "loss": 1.775,
+      "step": 61
+    },
+    {
+      "epoch": 0.023082919337762214,
+      "grad_norm": 0.9897575378417969,
+      "learning_rate": 1.2400000000000002e-06,
+      "loss": 1.7599,
+      "step": 62
+    },
+    {
+      "epoch": 0.02345522448837128,
+      "grad_norm": 0.9522385597229004,
+      "learning_rate": 1.26e-06,
+      "loss": 1.7643,
+      "step": 63
+    },
+    {
+      "epoch": 0.023827529638980348,
+      "grad_norm": 0.9590041637420654,
+      "learning_rate": 1.28e-06,
+      "loss": 1.7782,
+      "step": 64
+    },
+    {
+      "epoch": 0.024199834789589418,
+      "grad_norm": 0.910637378692627,
+      "learning_rate": 1.3e-06,
+      "loss": 1.7478,
+      "step": 65
+    },
+    {
+      "epoch": 0.024572139940198485,
+      "grad_norm": 0.9400556087493896,
+      "learning_rate": 1.32e-06,
+      "loss": 1.745,
+      "step": 66
+    },
+    {
+      "epoch": 0.02494444509080755,
+      "grad_norm": 0.9209262728691101,
+      "learning_rate": 1.34e-06,
+      "loss": 1.7594,
+      "step": 67
+    },
+    {
+      "epoch": 0.025316750241416622,
+      "grad_norm": 0.896427571773529,
+      "learning_rate": 1.3600000000000001e-06,
+      "loss": 1.7446,
+      "step": 68
+    },
+    {
+      "epoch": 0.02568905539202569,
+      "grad_norm": 0.895569920539856,
+      "learning_rate": 1.3800000000000001e-06,
+      "loss": 1.769,
+      "step": 69
+    },
+    {
+      "epoch": 0.02606136054263476,
+      "grad_norm": 0.7845183610916138,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 1.7196,
+      "step": 70
+    },
+    {
+      "epoch": 0.026433665693243825,
+      "grad_norm": 0.800564706325531,
+      "learning_rate": 1.42e-06,
+      "loss": 1.732,
+      "step": 71
+    },
+    {
+      "epoch": 0.026805970843852892,
+      "grad_norm": 0.6883242130279541,
+      "learning_rate": 1.44e-06,
+      "loss": 1.7251,
+      "step": 72
+    },
+    {
+      "epoch": 0.027178275994461962,
+      "grad_norm": 0.6776332259178162,
+      "learning_rate": 1.46e-06,
+      "loss": 1.7311,
+      "step": 73
+    },
+    {
+      "epoch": 0.02755058114507103,
+      "grad_norm": 0.6384090781211853,
+      "learning_rate": 1.48e-06,
+      "loss": 1.7215,
+      "step": 74
+    },
+    {
+      "epoch": 0.027922886295680096,
+      "grad_norm": 0.647678017616272,
+      "learning_rate": 1.5e-06,
+      "loss": 1.7038,
+      "step": 75
+    },
+    {
+      "epoch": 0.028295191446289166,
+      "grad_norm": 0.6253066658973694,
+      "learning_rate": 1.52e-06,
+      "loss": 1.7308,
+      "step": 76
+    },
+    {
+      "epoch": 0.028667496596898233,
+      "grad_norm": 0.599458634853363,
+      "learning_rate": 1.54e-06,
+      "loss": 1.6998,
+      "step": 77
+    },
+    {
+      "epoch": 0.0290398017475073,
+      "grad_norm": 0.5987157225608826,
+      "learning_rate": 1.56e-06,
+      "loss": 1.7092,
+      "step": 78
+    },
+    {
+      "epoch": 0.02941210689811637,
+      "grad_norm": 0.5856855511665344,
+      "learning_rate": 1.5800000000000001e-06,
+      "loss": 1.7001,
+      "step": 79
+    },
+    {
+      "epoch": 0.029784412048725437,
+      "grad_norm": 0.5748167037963867,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 1.6861,
+      "step": 80
+    },
+    {
+      "epoch": 0.030156717199334503,
+      "grad_norm": 0.5907240509986877,
+      "learning_rate": 1.6200000000000002e-06,
+      "loss": 1.7043,
+      "step": 81
+    },
+    {
+      "epoch": 0.030529022349943574,
+      "grad_norm": 0.5696431994438171,
+      "learning_rate": 1.6400000000000002e-06,
+      "loss": 1.6871,
+      "step": 82
+    },
+    {
+      "epoch": 0.03090132750055264,
+      "grad_norm": 0.5690540075302124,
+      "learning_rate": 1.6600000000000002e-06,
+      "loss": 1.6584,
+      "step": 83
+    },
+    {
+      "epoch": 0.03127363265116171,
+      "grad_norm": 0.5912604331970215,
+      "learning_rate": 1.6800000000000002e-06,
+      "loss": 1.6724,
+      "step": 84
+    },
+    {
+      "epoch": 0.031645937801770774,
+      "grad_norm": 0.595125138759613,
+      "learning_rate": 1.7000000000000002e-06,
+      "loss": 1.6629,
+      "step": 85
+    },
+    {
+      "epoch": 0.032018242952379844,
+      "grad_norm": 0.5880417227745056,
+      "learning_rate": 1.72e-06,
+      "loss": 1.6722,
+      "step": 86
+    },
+    {
+      "epoch": 0.032390548102988914,
+      "grad_norm": 0.621819019317627,
+      "learning_rate": 1.74e-06,
+      "loss": 1.6583,
+      "step": 87
+    },
+    {
+      "epoch": 0.03276285325359798,
+      "grad_norm": 0.6170995831489563,
+      "learning_rate": 1.76e-06,
+      "loss": 1.6434,
+      "step": 88
+    },
+    {
+      "epoch": 0.03313515840420705,
+      "grad_norm": 0.6138344407081604,
+      "learning_rate": 1.7800000000000001e-06,
+      "loss": 1.6368,
+      "step": 89
+    },
+    {
+      "epoch": 0.03350746355481612,
+      "grad_norm": 0.6477394700050354,
+      "learning_rate": 1.8000000000000001e-06,
+      "loss": 1.6606,
+      "step": 90
+    },
+    {
+      "epoch": 0.03387976870542518,
+      "grad_norm": 0.6119678616523743,
+      "learning_rate": 1.8200000000000002e-06,
+      "loss": 1.6266,
+      "step": 91
+    },
+    {
+      "epoch": 0.03425207385603425,
+      "grad_norm": 0.612330436706543,
+      "learning_rate": 1.8400000000000002e-06,
+      "loss": 1.618,
+      "step": 92
+    },
+    {
+      "epoch": 0.03462437900664332,
+      "grad_norm": 0.6225689053535461,
+      "learning_rate": 1.8600000000000002e-06,
+      "loss": 1.6234,
+      "step": 93
+    },
+    {
+      "epoch": 0.034996684157252385,
+      "grad_norm": 0.5868818759918213,
+      "learning_rate": 1.8800000000000002e-06,
+      "loss": 1.6091,
+      "step": 94
+    },
+    {
+      "epoch": 0.035368989307861455,
+      "grad_norm": 0.5659094452857971,
+      "learning_rate": 1.9000000000000002e-06,
+      "loss": 1.6211,
+      "step": 95
+    },
+    {
+      "epoch": 0.035741294458470525,
+      "grad_norm": 0.550031840801239,
+      "learning_rate": 1.9200000000000003e-06,
+      "loss": 1.6023,
+      "step": 96
+    },
+    {
+      "epoch": 0.03611359960907959,
+      "grad_norm": 0.5339946746826172,
+      "learning_rate": 1.94e-06,
+      "loss": 1.607,
+      "step": 97
+    },
+    {
+      "epoch": 0.03648590475968866,
+      "grad_norm": 0.5484851598739624,
+      "learning_rate": 1.9600000000000003e-06,
+      "loss": 1.6088,
+      "step": 98
+    },
+    {
+      "epoch": 0.03685820991029773,
+      "grad_norm": 0.5426100492477417,
+      "learning_rate": 1.98e-06,
+      "loss": 1.5946,
+      "step": 99
+    },
+    {
+      "epoch": 0.0372305150609068,
+      "grad_norm": 0.53022700548172,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 1.5923,
+      "step": 100
+    },
+    {
+      "epoch": 0.03760282021151586,
+      "grad_norm": 0.5056052207946777,
+      "learning_rate": 2.02e-06,
+      "loss": 1.5901,
+      "step": 101
+    },
+    {
+      "epoch": 0.03797512536212493,
+      "grad_norm": 0.5120390057563782,
+      "learning_rate": 2.04e-06,
+      "loss": 1.5924,
+      "step": 102
+    },
+    {
+      "epoch": 0.038347430512734,
+      "grad_norm": 0.5347453355789185,
+      "learning_rate": 2.06e-06,
+      "loss": 1.5924,
+      "step": 103
+    },
+    {
+      "epoch": 0.038719735663343066,
+      "grad_norm": 0.535159170627594,
+      "learning_rate": 2.08e-06,
+      "loss": 1.5821,
+      "step": 104
+    },
+    {
+      "epoch": 0.039092040813952136,
+      "grad_norm": 0.5138244032859802,
+      "learning_rate": 2.1000000000000002e-06,
+      "loss": 1.5981,
+      "step": 105
+    },
+    {
+      "epoch": 0.03946434596456121,
+      "grad_norm": 0.5143431425094604,
+      "learning_rate": 2.12e-06,
+      "loss": 1.5742,
+      "step": 106
+    },
+    {
+      "epoch": 0.03983665111517027,
+      "grad_norm": 0.5397112965583801,
+      "learning_rate": 2.1400000000000003e-06,
+      "loss": 1.5919,
+      "step": 107
+    },
+    {
+      "epoch": 0.04020895626577934,
+      "grad_norm": 0.503512978553772,
+      "learning_rate": 2.16e-06,
+      "loss": 1.5856,
+      "step": 108
+    },
+    {
+      "epoch": 0.04058126141638841,
+      "grad_norm": 0.5465199947357178,
+      "learning_rate": 2.1800000000000003e-06,
+      "loss": 1.5853,
+      "step": 109
+    },
+    {
+      "epoch": 0.040953566566997474,
+      "grad_norm": 0.561886727809906,
+      "learning_rate": 2.2e-06,
+      "loss": 1.5841,
+      "step": 110
+    },
+    {
+      "epoch": 0.041325871717606544,
+      "grad_norm": 0.5566477179527283,
+      "learning_rate": 2.2200000000000003e-06,
+      "loss": 1.5819,
+      "step": 111
+    },
+    {
+      "epoch": 0.041698176868215614,
+      "grad_norm": 0.5444602370262146,
+      "learning_rate": 2.24e-06,
+      "loss": 1.5702,
+      "step": 112
+    },
+    {
+      "epoch": 0.04207048201882468,
+      "grad_norm": 0.5601593852043152,
+      "learning_rate": 2.2600000000000004e-06,
+      "loss": 1.5667,
+      "step": 113
+    },
+    {
+      "epoch": 0.04244278716943375,
+      "grad_norm": 0.5660892128944397,
+      "learning_rate": 2.28e-06,
+      "loss": 1.5746,
+      "step": 114
+    },
+    {
+      "epoch": 0.04281509232004282,
+      "grad_norm": 0.5766957402229309,
+      "learning_rate": 2.3000000000000004e-06,
+      "loss": 1.5627,
+      "step": 115
+    },
+    {
+      "epoch": 0.04318739747065188,
+      "grad_norm": 0.6171875596046448,
+      "learning_rate": 2.3200000000000002e-06,
+      "loss": 1.5807,
+      "step": 116
+    },
+    {
+      "epoch": 0.04355970262126095,
+      "grad_norm": 0.5867440700531006,
+      "learning_rate": 2.3400000000000005e-06,
+      "loss": 1.5642,
+      "step": 117
+    },
+    {
+      "epoch": 0.04393200777187002,
+      "grad_norm": 0.5847681164741516,
+      "learning_rate": 2.3600000000000003e-06,
+      "loss": 1.5593,
+      "step": 118
+    },
+    {
+      "epoch": 0.044304312922479085,
+      "grad_norm": 0.5901392102241516,
+      "learning_rate": 2.38e-06,
+      "loss": 1.5697,
+      "step": 119
+    },
+    {
+      "epoch": 0.044676618073088155,
+      "grad_norm": 0.5898075699806213,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 1.5522,
+      "step": 120
+    },
+    {
+      "epoch": 0.045048923223697225,
+      "grad_norm": 0.5409351587295532,
+      "learning_rate": 2.42e-06,
+      "loss": 1.5498,
+      "step": 121
+    },
+    {
+      "epoch": 0.04542122837430629,
+      "grad_norm": 0.5224587321281433,
+      "learning_rate": 2.4400000000000004e-06,
+      "loss": 1.5567,
+      "step": 122
+    },
+    {
+      "epoch": 0.04579353352491536,
+      "grad_norm": 0.48012423515319824,
+      "learning_rate": 2.46e-06,
+      "loss": 1.5345,
+      "step": 123
+    },
+    {
+      "epoch": 0.04616583867552443,
+      "grad_norm": 0.44709786772727966,
+      "learning_rate": 2.4800000000000004e-06,
+      "loss": 1.5434,
+      "step": 124
+    },
+    {
+      "epoch": 0.04653814382613349,
+      "grad_norm": 0.43979910016059875,
+      "learning_rate": 2.5e-06,
+      "loss": 1.5461,
+      "step": 125
+    },
+    {
+      "epoch": 0.04691044897674256,
+      "grad_norm": 0.3758682608604431,
+      "learning_rate": 2.52e-06,
+      "loss": 1.5484,
+      "step": 126
+    },
+    {
+      "epoch": 0.04728275412735163,
+      "grad_norm": 0.34224575757980347,
+      "learning_rate": 2.5400000000000002e-06,
+      "loss": 1.5238,
+      "step": 127
+    },
+    {
+      "epoch": 0.047655059277960696,
+      "grad_norm": 0.3217172622680664,
+      "learning_rate": 2.56e-06,
+      "loss": 1.5567,
+      "step": 128
+    },
+    {
+      "epoch": 0.048027364428569766,
+      "grad_norm": 0.2725611925125122,
+      "learning_rate": 2.5800000000000003e-06,
+      "loss": 1.5349,
+      "step": 129
+    },
+    {
+      "epoch": 0.048399669579178836,
+      "grad_norm": 0.22641003131866455,
+      "learning_rate": 2.6e-06,
+      "loss": 1.533,
+      "step": 130
+    },
+    {
+      "epoch": 0.0487719747297879,
+      "grad_norm": 0.19425782561302185,
+      "learning_rate": 2.6200000000000003e-06,
+      "loss": 1.5486,
+      "step": 131
+    },
+    {
+      "epoch": 0.04914427988039697,
+      "grad_norm": 0.18164916336536407,
+      "learning_rate": 2.64e-06,
+      "loss": 1.5494,
+      "step": 132
+    },
+    {
+      "epoch": 0.04951658503100604,
+      "grad_norm": 0.16984394192695618,
+      "learning_rate": 2.6600000000000004e-06,
+      "loss": 1.5342,
+      "step": 133
+    },
+    {
+      "epoch": 0.0498888901816151,
+      "grad_norm": 0.15408554673194885,
+      "learning_rate": 2.68e-06,
+      "loss": 1.5349,
+      "step": 134
+    },
+    {
+      "epoch": 0.05026119533222417,
+      "grad_norm": 0.1477842777967453,
+      "learning_rate": 2.7000000000000004e-06,
+      "loss": 1.5492,
+      "step": 135
+    },
+    {
+      "epoch": 0.050633500482833244,
+      "grad_norm": 0.14385773241519928,
+      "learning_rate": 2.7200000000000002e-06,
+      "loss": 1.5415,
+      "step": 136
+    },
+    {
+      "epoch": 0.05100580563344231,
+      "grad_norm": 0.1423702985048294,
+      "learning_rate": 2.7400000000000004e-06,
+      "loss": 1.5231,
+      "step": 137
+    },
+    {
+      "epoch": 0.05137811078405138,
+      "grad_norm": 0.4125309884548187,
+      "learning_rate": 2.7600000000000003e-06,
+      "loss": 1.5381,
+      "step": 138
+    },
+    {
+      "epoch": 0.05175041593466045,
+      "grad_norm": 0.14331988990306854,
+      "learning_rate": 2.7800000000000005e-06,
+      "loss": 1.5438,
+      "step": 139
+    },
+    {
+      "epoch": 0.05212272108526952,
+      "grad_norm": 0.12616126239299774,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 1.5481,
+      "step": 140
+    },
+    {
+      "epoch": 0.05249502623587858,
+      "grad_norm": 0.1296859085559845,
+      "learning_rate": 2.82e-06,
+      "loss": 1.5316,
+      "step": 141
+    },
+    {
+      "epoch": 0.05286733138648765,
+      "grad_norm": 0.14372870326042175,
+      "learning_rate": 2.84e-06,
+      "loss": 1.5314,
+      "step": 142
+    },
+    {
+      "epoch": 0.05323963653709672,
+      "grad_norm": 0.12012794613838196,
+      "learning_rate": 2.86e-06,
+      "loss": 1.5256,
+      "step": 143
+    },
+    {
+      "epoch": 0.053611941687705784,
+      "grad_norm": 0.1238911896944046,
+      "learning_rate": 2.88e-06,
+      "loss": 1.5376,
+      "step": 144
+    },
+    {
+      "epoch": 0.053984246838314855,
+      "grad_norm": 0.11403171718120575,
+      "learning_rate": 2.9e-06,
+      "loss": 1.5313,
+      "step": 145
+    },
+    {
+      "epoch": 0.054356551988923925,
+      "grad_norm": 0.11582870781421661,
+      "learning_rate": 2.92e-06,
+      "loss": 1.5293,
+      "step": 146
+    },
+    {
+      "epoch": 0.05472885713953299,
+      "grad_norm": 0.11598813533782959,
+      "learning_rate": 2.9400000000000002e-06,
+      "loss": 1.5387,
+      "step": 147
+    },
+    {
+      "epoch": 0.05510116229014206,
+      "grad_norm": 0.11165577918291092,
+      "learning_rate": 2.96e-06,
+      "loss": 1.5195,
+      "step": 148
+    },
+    {
+      "epoch": 0.05547346744075113,
+      "grad_norm": 0.1125321313738823,
+      "learning_rate": 2.9800000000000003e-06,
+      "loss": 1.521,
+      "step": 149
+    },
+    {
+      "epoch": 0.05584577259136019,
+      "grad_norm": 0.11534541845321655,
+      "learning_rate": 3e-06,
+      "loss": 1.5368,
+      "step": 150
+    },
+    {
+      "epoch": 0.05621807774196926,
+      "grad_norm": 0.11158929765224457,
+      "learning_rate": 3.0200000000000003e-06,
+      "loss": 1.53,
+      "step": 151
+    },
+    {
+      "epoch": 0.05659038289257833,
+      "grad_norm": 0.11701493710279465,
+      "learning_rate": 3.04e-06,
+      "loss": 1.517,
+      "step": 152
+    },
+    {
+      "epoch": 0.056962688043187396,
+      "grad_norm": 0.10573873668909073,
+      "learning_rate": 3.0600000000000003e-06,
+      "loss": 1.5479,
+      "step": 153
+    },
+    {
+      "epoch": 0.057334993193796466,
+      "grad_norm": 0.10436706990003586,
+      "learning_rate": 3.08e-06,
+      "loss": 1.5376,
+      "step": 154
+    },
+    {
+      "epoch": 0.057707298344405536,
+      "grad_norm": 0.11538543552160263,
+      "learning_rate": 3.1000000000000004e-06,
+      "loss": 1.5405,
+      "step": 155
+    },
+    {
+      "epoch": 0.0580796034950146,
+      "grad_norm": 0.11722234636545181,
+      "learning_rate": 3.12e-06,
+      "loss": 1.5249,
+      "step": 156
+    },
+    {
+      "epoch": 0.05845190864562367,
+      "grad_norm": 0.11727150529623032,
+      "learning_rate": 3.1400000000000004e-06,
+      "loss": 1.5279,
+      "step": 157
+    },
+    {
+      "epoch": 0.05882421379623274,
+      "grad_norm": 0.10549698024988174,
+      "learning_rate": 3.1600000000000002e-06,
+      "loss": 1.5101,
+      "step": 158
+    },
+    {
+      "epoch": 0.0591965189468418,
+      "grad_norm": 0.10655630379915237,
+      "learning_rate": 3.1800000000000005e-06,
+      "loss": 1.523,
+      "step": 159
+    },
+    {
+      "epoch": 0.05956882409745087,
+      "grad_norm": 0.10541582852602005,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 1.5114,
+      "step": 160
+    },
+    {
+      "epoch": 0.05994112924805994,
+      "grad_norm": 0.1057916209101677,
+      "learning_rate": 3.2200000000000005e-06,
+      "loss": 1.5146,
+      "step": 161
+    },
+    {
+      "epoch": 0.06031343439866901,
+      "grad_norm": 0.10496512800455093,
+      "learning_rate": 3.2400000000000003e-06,
+      "loss": 1.5168,
+      "step": 162
+    },
+    {
+      "epoch": 0.06068573954927808,
+      "grad_norm": 0.09892502427101135,
+      "learning_rate": 3.2600000000000006e-06,
+      "loss": 1.5053,
+      "step": 163
+    },
+    {
+      "epoch": 0.06105804469988715,
+      "grad_norm": 0.10325302928686142,
+      "learning_rate": 3.2800000000000004e-06,
+      "loss": 1.5157,
+      "step": 164
+    },
+    {
+      "epoch": 0.06143034985049621,
+      "grad_norm": 0.11229792982339859,
+      "learning_rate": 3.3000000000000006e-06,
+      "loss": 1.4997,
+      "step": 165
+    },
+    {
+      "epoch": 0.06180265500110528,
+      "grad_norm": 0.10319443047046661,
+      "learning_rate": 3.3200000000000004e-06,
+      "loss": 1.5175,
+      "step": 166
+    },
+    {
+      "epoch": 0.06217496015171435,
+      "grad_norm": 0.10021348297595978,
+      "learning_rate": 3.3400000000000006e-06,
+      "loss": 1.5383,
+      "step": 167
+    },
+    {
+      "epoch": 0.06254726530232342,
+      "grad_norm": 0.09915069490671158,
+      "learning_rate": 3.3600000000000004e-06,
+      "loss": 1.5226,
+      "step": 168
+    },
+    {
+      "epoch": 0.06291957045293249,
+      "grad_norm": 0.10507778823375702,
+      "learning_rate": 3.3800000000000007e-06,
+      "loss": 1.5251,
+      "step": 169
+    },
+    {
+      "epoch": 0.06329187560354155,
+      "grad_norm": 0.10222301632165909,
+      "learning_rate": 3.4000000000000005e-06,
+      "loss": 1.5287,
+      "step": 170
+    },
+    {
+      "epoch": 0.06366418075415062,
+      "grad_norm": 0.10224229842424393,
+      "learning_rate": 3.4200000000000007e-06,
+      "loss": 1.5047,
+      "step": 171
+    },
+    {
+      "epoch": 0.06403648590475969,
+      "grad_norm": 0.09706564992666245,
+      "learning_rate": 3.44e-06,
+      "loss": 1.5136,
+      "step": 172
+    },
+    {
+      "epoch": 0.06440879105536876,
+      "grad_norm": 0.09412699937820435,
+      "learning_rate": 3.46e-06,
+      "loss": 1.5141,
+      "step": 173
+    },
+    {
+      "epoch": 0.06478109620597783,
+      "grad_norm": 0.09892858564853668,
+      "learning_rate": 3.48e-06,
+      "loss": 1.5198,
+      "step": 174
+    },
+    {
+      "epoch": 0.0651534013565869,
+      "grad_norm": 0.10099875926971436,
+      "learning_rate": 3.5e-06,
+      "loss": 1.5338,
+      "step": 175
+    },
+    {
+      "epoch": 0.06552570650719595,
+      "grad_norm": 0.0975039079785347,
+      "learning_rate": 3.52e-06,
+      "loss": 1.5202,
+      "step": 176
+    },
+    {
+      "epoch": 0.06589801165780503,
+      "grad_norm": 0.09733244776725769,
+      "learning_rate": 3.54e-06,
+      "loss": 1.5256,
+      "step": 177
+    },
+    {
+      "epoch": 0.0662703168084141,
+      "grad_norm": 0.10019107162952423,
+      "learning_rate": 3.5600000000000002e-06,
+      "loss": 1.5048,
+      "step": 178
+    },
+    {
+      "epoch": 0.06664262195902317,
+      "grad_norm": 0.10042434930801392,
+      "learning_rate": 3.58e-06,
+      "loss": 1.499,
+      "step": 179
+    },
+    {
+      "epoch": 0.06701492710963224,
+      "grad_norm": 0.10034509003162384,
+      "learning_rate": 3.6000000000000003e-06,
+      "loss": 1.5059,
+      "step": 180
+    },
+    {
+      "epoch": 0.0673872322602413,
+      "grad_norm": 0.09680823236703873,
+      "learning_rate": 3.62e-06,
+      "loss": 1.5074,
+      "step": 181
+    },
+    {
+      "epoch": 0.06775953741085036,
+      "grad_norm": 0.10564741492271423,
+      "learning_rate": 3.6400000000000003e-06,
+      "loss": 1.5014,
+      "step": 182
+    },
+    {
+      "epoch": 0.06813184256145943,
+      "grad_norm": 0.09912260621786118,
+      "learning_rate": 3.66e-06,
+      "loss": 1.5329,
+      "step": 183
+    },
+    {
+      "epoch": 0.0685041477120685,
+      "grad_norm": 0.09995097666978836,
+      "learning_rate": 3.6800000000000003e-06,
+      "loss": 1.4999,
+      "step": 184
+    },
+    {
+      "epoch": 0.06887645286267757,
+      "grad_norm": 0.10238537192344666,
+      "learning_rate": 3.7e-06,
+      "loss": 1.5197,
+      "step": 185
+    },
+    {
+      "epoch": 0.06924875801328664,
+      "grad_norm": 0.09943666309118271,
+      "learning_rate": 3.7200000000000004e-06,
+      "loss": 1.5234,
+      "step": 186
+    },
+    {
+      "epoch": 0.06962106316389571,
+      "grad_norm": 0.09201087057590485,
+      "learning_rate": 3.74e-06,
+      "loss": 1.5103,
+      "step": 187
+    },
+    {
+      "epoch": 0.06999336831450477,
+      "grad_norm": 0.09809333086013794,
+      "learning_rate": 3.7600000000000004e-06,
+      "loss": 1.5107,
+      "step": 188
+    },
+    {
+      "epoch": 0.07036567346511384,
+      "grad_norm": 0.09340554475784302,
+      "learning_rate": 3.7800000000000002e-06,
+      "loss": 1.517,
+      "step": 189
+    },
+    {
+      "epoch": 0.07073797861572291,
+      "grad_norm": 0.10037955641746521,
+      "learning_rate": 3.8000000000000005e-06,
+      "loss": 1.51,
+      "step": 190
+    },
+    {
+      "epoch": 0.07111028376633198,
+      "grad_norm": 0.0913090705871582,
+      "learning_rate": 3.820000000000001e-06,
+      "loss": 1.5033,
+      "step": 191
+    },
+    {
+      "epoch": 0.07148258891694105,
+      "grad_norm": 0.09793171286582947,
+      "learning_rate": 3.8400000000000005e-06,
+      "loss": 1.5268,
+      "step": 192
+    },
+    {
+      "epoch": 0.07185489406755012,
+      "grad_norm": 0.09348586201667786,
+      "learning_rate": 3.86e-06,
+      "loss": 1.5046,
+      "step": 193
+    },
+    {
+      "epoch": 0.07222719921815918,
+      "grad_norm": 0.09434423595666885,
+      "learning_rate": 3.88e-06,
+      "loss": 1.498,
+      "step": 194
+    },
+    {
+      "epoch": 0.07259950436876825,
+      "grad_norm": 0.08928900957107544,
+      "learning_rate": 3.900000000000001e-06,
+      "loss": 1.4918,
+      "step": 195
+    },
+    {
+      "epoch": 0.07297180951937732,
+      "grad_norm": 0.0911901518702507,
+      "learning_rate": 3.920000000000001e-06,
+      "loss": 1.4903,
+      "step": 196
+    },
+    {
+      "epoch": 0.07334411466998639,
+      "grad_norm": 0.08914012461900711,
+      "learning_rate": 3.94e-06,
+      "loss": 1.512,
+      "step": 197
+    },
+    {
+      "epoch": 0.07371641982059546,
+      "grad_norm": 0.0897861197590828,
+      "learning_rate": 3.96e-06,
+      "loss": 1.5127,
+      "step": 198
+    },
+    {
+      "epoch": 0.07408872497120453,
+      "grad_norm": 0.0919618159532547,
+      "learning_rate": 3.980000000000001e-06,
+      "loss": 1.5032,
+      "step": 199
+    },
+    {
+      "epoch": 0.0744610301218136,
+      "grad_norm": 0.0871221199631691,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.5092,
+      "step": 200
+    },
+    {
+      "epoch": 0.07483333527242265,
+      "grad_norm": 0.08722933381795883,
+      "learning_rate": 4.0200000000000005e-06,
+      "loss": 1.4905,
+      "step": 201
+    },
+    {
+      "epoch": 0.07520564042303172,
+      "grad_norm": 0.08796436339616776,
+      "learning_rate": 4.04e-06,
+      "loss": 1.4992,
+      "step": 202
+    },
+    {
+      "epoch": 0.0755779455736408,
+      "grad_norm": 0.09674139320850372,
+      "learning_rate": 4.060000000000001e-06,
+      "loss": 1.5032,
+      "step": 203
+    },
+    {
+      "epoch": 0.07595025072424987,
+      "grad_norm": 0.08799508959054947,
+      "learning_rate": 4.08e-06,
+      "loss": 1.5124,
+      "step": 204
+    },
+    {
+      "epoch": 0.07632255587485894,
+      "grad_norm": 0.08769244700670242,
+      "learning_rate": 4.1e-06,
+      "loss": 1.5087,
+      "step": 205
+    },
+    {
+      "epoch": 0.076694861025468,
+      "grad_norm": 0.08863134682178497,
+      "learning_rate": 4.12e-06,
+      "loss": 1.5129,
+      "step": 206
+    },
+    {
+      "epoch": 0.07706716617607706,
+      "grad_norm": 0.1002495214343071,
+      "learning_rate": 4.14e-06,
+      "loss": 1.5024,
+      "step": 207
+    },
+    {
+      "epoch": 0.07743947132668613,
+      "grad_norm": 0.09144977480173111,
+      "learning_rate": 4.16e-06,
+      "loss": 1.5006,
+      "step": 208
+    },
+    {
+      "epoch": 0.0778117764772952,
+      "grad_norm": 0.10466210544109344,
+      "learning_rate": 4.18e-06,
+      "loss": 1.5022,
+      "step": 209
+    },
+    {
+      "epoch": 0.07818408162790427,
+      "grad_norm": 0.08975500613451004,
+      "learning_rate": 4.2000000000000004e-06,
+      "loss": 1.4965,
+      "step": 210
+    },
+    {
+      "epoch": 0.07855638677851334,
+      "grad_norm": 0.08787556737661362,
+      "learning_rate": 4.22e-06,
+      "loss": 1.5042,
+      "step": 211
+    },
+    {
+      "epoch": 0.07892869192912241,
+      "grad_norm": 0.08773490786552429,
+      "learning_rate": 4.24e-06,
+      "loss": 1.5138,
+      "step": 212
+    },
+    {
+      "epoch": 0.07930099707973147,
+      "grad_norm": 0.08304975181818008,
+      "learning_rate": 4.26e-06,
+      "loss": 1.507,
+      "step": 213
+    },
+    {
+      "epoch": 0.07967330223034054,
+      "grad_norm": 0.08493588864803314,
+      "learning_rate": 4.2800000000000005e-06,
+      "loss": 1.5017,
+      "step": 214
+    },
+    {
+      "epoch": 0.08004560738094961,
+      "grad_norm": 0.08817660808563232,
+      "learning_rate": 4.3e-06,
+      "loss": 1.5068,
+      "step": 215
+    },
+    {
+      "epoch": 0.08041791253155868,
+      "grad_norm": 0.08963938057422638,
+      "learning_rate": 4.32e-06,
+      "loss": 1.5263,
+      "step": 216
+    },
+    {
+      "epoch": 0.08079021768216775,
+      "grad_norm": 0.08390969783067703,
+      "learning_rate": 4.34e-06,
+      "loss": 1.5,
+      "step": 217
+    },
+    {
+      "epoch": 0.08116252283277682,
+      "grad_norm": 0.08665873855352402,
+      "learning_rate": 4.360000000000001e-06,
+      "loss": 1.4882,
+      "step": 218
+    },
+    {
+      "epoch": 0.08153482798338588,
+      "grad_norm": 0.08230563998222351,
+      "learning_rate": 4.38e-06,
+      "loss": 1.4888,
+      "step": 219
+    },
+    {
+      "epoch": 0.08190713313399495,
+      "grad_norm": 0.08207309246063232,
+      "learning_rate": 4.4e-06,
+      "loss": 1.4954,
+      "step": 220
+    },
+    {
+      "epoch": 0.08227943828460402,
+      "grad_norm": 0.08314842730760574,
+      "learning_rate": 4.42e-06,
+      "loss": 1.5103,
+      "step": 221
+    },
+    {
+      "epoch": 0.08265174343521309,
+      "grad_norm": 0.08730518817901611,
+      "learning_rate": 4.440000000000001e-06,
+      "loss": 1.4845,
+      "step": 222
+    },
+    {
+      "epoch": 0.08302404858582216,
+      "grad_norm": 0.09334740787744522,
+      "learning_rate": 4.4600000000000005e-06,
+      "loss": 1.4726,
+      "step": 223
+    },
+    {
+      "epoch": 0.08339635373643123,
+      "grad_norm": 0.0923716351389885,
+      "learning_rate": 4.48e-06,
+      "loss": 1.5005,
+      "step": 224
+    },
+    {
+      "epoch": 0.08376865888704028,
+      "grad_norm": 0.08330400288105011,
+      "learning_rate": 4.5e-06,
+      "loss": 1.4939,
+      "step": 225
+    },
+    {
+      "epoch": 0.08414096403764935,
+      "grad_norm": 0.08828233182430267,
+      "learning_rate": 4.520000000000001e-06,
+      "loss": 1.4926,
+      "step": 226
+    },
+    {
+      "epoch": 0.08451326918825842,
+      "grad_norm": 0.08477775007486343,
+      "learning_rate": 4.540000000000001e-06,
+      "loss": 1.4872,
+      "step": 227
+    },
+    {
+      "epoch": 0.0848855743388675,
+      "grad_norm": 0.08313533663749695,
+      "learning_rate": 4.56e-06,
+      "loss": 1.488,
+      "step": 228
+    },
+    {
+      "epoch": 0.08525787948947657,
+      "grad_norm": 0.08070395141839981,
+      "learning_rate": 4.58e-06,
+      "loss": 1.498,
+      "step": 229
+    },
+    {
+      "epoch": 0.08563018464008564,
+      "grad_norm": 0.08496372401714325,
+      "learning_rate": 4.600000000000001e-06,
+      "loss": 1.4995,
+      "step": 230
+    },
+    {
+      "epoch": 0.08600248979069469,
+      "grad_norm": 0.08610887080430984,
+      "learning_rate": 4.620000000000001e-06,
+      "loss": 1.4861,
+      "step": 231
+    },
+    {
+      "epoch": 0.08637479494130376,
+      "grad_norm": 0.1109740138053894,
+      "learning_rate": 4.6400000000000005e-06,
+      "loss": 1.4997,
+      "step": 232
+    },
+    {
+      "epoch": 0.08674710009191283,
+      "grad_norm": 0.0828074961900711,
+      "learning_rate": 4.66e-06,
+      "loss": 1.4768,
+      "step": 233
+    },
+    {
+      "epoch": 0.0871194052425219,
+      "grad_norm": 0.10130389034748077,
+      "learning_rate": 4.680000000000001e-06,
+      "loss": 1.5143,
+      "step": 234
+    },
+    {
+      "epoch": 0.08749171039313097,
+      "grad_norm": 0.07845676690340042,
+      "learning_rate": 4.7e-06,
+      "loss": 1.4835,
+      "step": 235
+    },
+    {
+      "epoch": 0.08786401554374004,
+      "grad_norm": 0.0835128203034401,
+      "learning_rate": 4.7200000000000005e-06,
+      "loss": 1.487,
+      "step": 236
+    },
+    {
+      "epoch": 0.0882363206943491,
+      "grad_norm": 0.07999230176210403,
+      "learning_rate": 4.74e-06,
+      "loss": 1.4768,
+      "step": 237
+    },
+    {
+      "epoch": 0.08860862584495817,
+      "grad_norm": 0.08866851031780243,
+      "learning_rate": 4.76e-06,
+      "loss": 1.4877,
+      "step": 238
+    },
+    {
+      "epoch": 0.08898093099556724,
+      "grad_norm": 0.09091200679540634,
+      "learning_rate": 4.78e-06,
+      "loss": 1.4663,
+      "step": 239
+    },
+    {
+      "epoch": 0.08935323614617631,
+      "grad_norm": 0.09118737280368805,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 1.4818,
+      "step": 240
+    },
+    {
+      "epoch": 0.08972554129678538,
+      "grad_norm": 0.09198911488056183,
+      "learning_rate": 4.8200000000000004e-06,
+      "loss": 1.486,
+      "step": 241
+    },
+    {
+      "epoch": 0.09009784644739445,
+      "grad_norm": 0.08087292313575745,
+      "learning_rate": 4.84e-06,
+      "loss": 1.4839,
+      "step": 242
+    },
+    {
+      "epoch": 0.09047015159800352,
+      "grad_norm": 0.0856778621673584,
+      "learning_rate": 4.86e-06,
+      "loss": 1.4882,
+      "step": 243
+    },
+    {
+      "epoch": 0.09084245674861258,
+      "grad_norm": 0.09004831314086914,
+      "learning_rate": 4.880000000000001e-06,
+      "loss": 1.4977,
+      "step": 244
+    },
+    {
+      "epoch": 0.09121476189922165,
+      "grad_norm": 0.08808062225580215,
+      "learning_rate": 4.9000000000000005e-06,
+      "loss": 1.4961,
+      "step": 245
+    },
+    {
+      "epoch": 0.09158706704983072,
+      "grad_norm": 0.10897014290094376,
+      "learning_rate": 4.92e-06,
+      "loss": 1.4837,
+      "step": 246
+    },
+    {
+      "epoch": 0.09195937220043979,
+      "grad_norm": 0.12116125226020813,
+      "learning_rate": 4.94e-06,
+      "loss": 1.4899,
+      "step": 247
+    },
+    {
+      "epoch": 0.09233167735104886,
+      "grad_norm": 0.09493674337863922,
+      "learning_rate": 4.960000000000001e-06,
+      "loss": 1.4834,
+      "step": 248
+    },
+    {
+      "epoch": 0.09270398250165793,
+      "grad_norm": 0.13177277147769928,
+      "learning_rate": 4.980000000000001e-06,
+      "loss": 1.4721,
+      "step": 249
+    },
+    {
+      "epoch": 0.09307628765226698,
+      "grad_norm": 0.11521487683057785,
+      "learning_rate": 5e-06,
+      "loss": 1.5012,
+      "step": 250
+    },
+    {
+      "epoch": 0.09344859280287605,
+      "grad_norm": 0.106010802090168,
+      "learning_rate": 5.02e-06,
+      "loss": 1.5041,
+      "step": 251
+    },
+    {
+      "epoch": 0.09382089795348512,
+      "grad_norm": 0.08309640735387802,
+      "learning_rate": 5.04e-06,
+      "loss": 1.4651,
+      "step": 252
+    },
+    {
+      "epoch": 0.0941932031040942,
+      "grad_norm": 0.23099473118782043,
+      "learning_rate": 5.060000000000001e-06,
+      "loss": 1.4706,
+      "step": 253
+    },
+    {
+      "epoch": 0.09456550825470326,
+      "grad_norm": 0.08957033604383469,
+      "learning_rate": 5.0800000000000005e-06,
+      "loss": 1.4851,
+      "step": 254
+    },
+    {
+      "epoch": 0.09493781340531234,
+      "grad_norm": 0.08968747407197952,
+      "learning_rate": 5.1e-06,
+      "loss": 1.4615,
+      "step": 255
+    },
+    {
+      "epoch": 0.09531011855592139,
+      "grad_norm": 0.09505753219127655,
+      "learning_rate": 5.12e-06,
+      "loss": 1.4852,
+      "step": 256
+    },
+    {
+      "epoch": 0.09568242370653046,
+      "grad_norm": 0.08881967514753342,
+      "learning_rate": 5.140000000000001e-06,
+      "loss": 1.5098,
+      "step": 257
+    },
+    {
+      "epoch": 0.09605472885713953,
+      "grad_norm": 0.08623534440994263,
+      "learning_rate": 5.1600000000000006e-06,
+      "loss": 1.466,
+      "step": 258
+    },
+    {
+      "epoch": 0.0964270340077486,
+      "grad_norm": 0.08393881469964981,
+      "learning_rate": 5.18e-06,
+      "loss": 1.4932,
+      "step": 259
+    },
+    {
+      "epoch": 0.09679933915835767,
+      "grad_norm": 0.0790511816740036,
+      "learning_rate": 5.2e-06,
+      "loss": 1.489,
+      "step": 260
+    },
+    {
+      "epoch": 0.09717164430896674,
+      "grad_norm": 0.08720332384109497,
+      "learning_rate": 5.220000000000001e-06,
+      "loss": 1.4878,
+      "step": 261
+    },
+    {
+      "epoch": 0.0975439494595758,
+      "grad_norm": 0.09274443984031677,
+      "learning_rate": 5.240000000000001e-06,
+      "loss": 1.4629,
+      "step": 262
+    },
+    {
+      "epoch": 0.09791625461018487,
+      "grad_norm": 0.0823291540145874,
+      "learning_rate": 5.2600000000000005e-06,
+      "loss": 1.4798,
+      "step": 263
+    },
+    {
+      "epoch": 0.09828855976079394,
+      "grad_norm": 0.10442589223384857,
+      "learning_rate": 5.28e-06,
+      "loss": 1.4885,
+      "step": 264
+    },
+    {
+      "epoch": 0.09866086491140301,
+      "grad_norm": 0.08945530652999878,
+      "learning_rate": 5.300000000000001e-06,
+      "loss": 1.4802,
+      "step": 265
+    },
+    {
+      "epoch": 0.09903317006201208,
+      "grad_norm": 0.11119077354669571,
+      "learning_rate": 5.320000000000001e-06,
+      "loss": 1.4888,
+      "step": 266
+    },
+    {
+      "epoch": 0.09940547521262115,
+      "grad_norm": 0.09534472972154617,
+      "learning_rate": 5.3400000000000005e-06,
+      "loss": 1.4838,
+      "step": 267
+    },
+    {
+      "epoch": 0.0997777803632302,
+      "grad_norm": 0.08682861924171448,
+      "learning_rate": 5.36e-06,
+      "loss": 1.4617,
+      "step": 268
+    },
+    {
+      "epoch": 0.10015008551383928,
+      "grad_norm": 0.10255074501037598,
+      "learning_rate": 5.380000000000001e-06,
+      "loss": 1.4767,
+      "step": 269
+    },
+    {
+      "epoch": 0.10052239066444835,
+      "grad_norm": 0.09030777961015701,
+      "learning_rate": 5.400000000000001e-06,
+      "loss": 1.463,
+      "step": 270
+    },
+    {
+      "epoch": 0.10089469581505742,
+      "grad_norm": 0.09011770784854889,
+      "learning_rate": 5.420000000000001e-06,
+      "loss": 1.484,
+      "step": 271
+    },
+    {
+      "epoch": 0.10126700096566649,
+      "grad_norm": 0.08929470181465149,
+      "learning_rate": 5.4400000000000004e-06,
+      "loss": 1.4789,
+      "step": 272
+    },
+    {
+      "epoch": 0.10163930611627556,
+      "grad_norm": 0.09251522272825241,
+      "learning_rate": 5.460000000000001e-06,
+      "loss": 1.4682,
+      "step": 273
+    },
+    {
+      "epoch": 0.10201161126688461,
+      "grad_norm": 0.11165442317724228,
+      "learning_rate": 5.480000000000001e-06,
+      "loss": 1.4829,
+      "step": 274
+    },
+    {
+      "epoch": 0.10238391641749368,
+      "grad_norm": 0.09440817683935165,
+      "learning_rate": 5.500000000000001e-06,
+      "loss": 1.4769,
+      "step": 275
+    },
+    {
+      "epoch": 0.10275622156810275,
+      "grad_norm": 0.11706740409135818,
+      "learning_rate": 5.5200000000000005e-06,
+      "loss": 1.477,
+      "step": 276
+    },
+    {
+      "epoch": 0.10312852671871182,
+      "grad_norm": 0.08593502640724182,
+      "learning_rate": 5.540000000000001e-06,
+      "loss": 1.4727,
+      "step": 277
+    },
+    {
+      "epoch": 0.1035008318693209,
+      "grad_norm": 0.0881851464509964,
+      "learning_rate": 5.560000000000001e-06,
+      "loss": 1.4852,
+      "step": 278
+    },
+    {
+      "epoch": 0.10387313701992996,
+      "grad_norm": 0.11567908525466919,
+      "learning_rate": 5.580000000000001e-06,
+      "loss": 1.4608,
+      "step": 279
+    },
+    {
+      "epoch": 0.10424544217053903,
+      "grad_norm": 0.09296432882547379,
+      "learning_rate": 5.600000000000001e-06,
+      "loss": 1.4811,
+      "step": 280
+    },
+    {
+      "epoch": 0.10461774732114809,
+      "grad_norm": 0.10216089338064194,
+      "learning_rate": 5.620000000000001e-06,
+      "loss": 1.4682,
+      "step": 281
+    },
+    {
+      "epoch": 0.10499005247175716,
+      "grad_norm": 0.08564305305480957,
+      "learning_rate": 5.64e-06,
+      "loss": 1.4832,
+      "step": 282
+    },
+    {
+      "epoch": 0.10536235762236623,
+      "grad_norm": 0.11430171877145767,
+      "learning_rate": 5.66e-06,
+      "loss": 1.4754,
+      "step": 283
+    },
+    {
+      "epoch": 0.1057346627729753,
+      "grad_norm": 0.09230099618434906,
+      "learning_rate": 5.68e-06,
+      "loss": 1.4765,
+      "step": 284
+    },
+    {
+      "epoch": 0.10610696792358437,
+      "grad_norm": 0.14818595349788666,
+      "learning_rate": 5.7e-06,
+      "loss": 1.4642,
+      "step": 285
+    },
+    {
+      "epoch": 0.10647927307419344,
+      "grad_norm": 0.1206260696053505,
+      "learning_rate": 5.72e-06,
+      "loss": 1.462,
+      "step": 286
+    },
+    {
+      "epoch": 0.1068515782248025,
+      "grad_norm": 0.11496801674365997,
+      "learning_rate": 5.74e-06,
+      "loss": 1.4741,
+      "step": 287
+    },
+    {
+      "epoch": 0.10722388337541157,
+      "grad_norm": 0.09007902443408966,
+      "learning_rate": 5.76e-06,
+      "loss": 1.472,
+      "step": 288
+    },
+    {
+      "epoch": 0.10759618852602064,
+      "grad_norm": 0.0860317051410675,
+      "learning_rate": 5.78e-06,
+      "loss": 1.4814,
+      "step": 289
+    },
+    {
+      "epoch": 0.10796849367662971,
+      "grad_norm": 0.08847720175981522,
+      "learning_rate": 5.8e-06,
+      "loss": 1.473,
+      "step": 290
+    },
+    {
+      "epoch": 0.10834079882723878,
+      "grad_norm": 0.09280810505151749,
+      "learning_rate": 5.82e-06,
+      "loss": 1.4628,
+      "step": 291
+    },
+    {
+      "epoch": 0.10871310397784785,
+      "grad_norm": 0.0906905010342598,
+      "learning_rate": 5.84e-06,
+      "loss": 1.4637,
+      "step": 292
+    },
+    {
+      "epoch": 0.1090854091284569,
+      "grad_norm": 0.08728771656751633,
+      "learning_rate": 5.86e-06,
+      "loss": 1.4667,
+      "step": 293
+    },
+    {
+      "epoch": 0.10945771427906598,
+      "grad_norm": 0.09215451776981354,
+      "learning_rate": 5.8800000000000005e-06,
+      "loss": 1.4588,
+      "step": 294
+    },
+    {
+      "epoch": 0.10983001942967505,
+      "grad_norm": 0.09837444871664047,
+      "learning_rate": 5.9e-06,
+      "loss": 1.464,
+      "step": 295
+    },
+    {
+      "epoch": 0.11020232458028412,
+      "grad_norm": 0.08866247534751892,
+      "learning_rate": 5.92e-06,
+      "loss": 1.4585,
+      "step": 296
+    },
+    {
+      "epoch": 0.11057462973089319,
+      "grad_norm": 0.11951064318418503,
+      "learning_rate": 5.94e-06,
+      "loss": 1.4612,
+      "step": 297
+    },
+    {
+      "epoch": 0.11094693488150226,
+      "grad_norm": 0.08314000070095062,
+      "learning_rate": 5.9600000000000005e-06,
+      "loss": 1.4635,
+      "step": 298
+    },
+    {
+      "epoch": 0.11131924003211131,
+      "grad_norm": 0.12508852779865265,
+      "learning_rate": 5.98e-06,
+      "loss": 1.4722,
+      "step": 299
+    },
+    {
+      "epoch": 0.11169154518272038,
+      "grad_norm": 0.11367225646972656,
+      "learning_rate": 6e-06,
+      "loss": 1.4722,
+      "step": 300
+    },
+    {
+      "epoch": 0.11206385033332945,
+      "grad_norm": 0.1337917000055313,
+      "learning_rate": 6.02e-06,
+      "loss": 1.4531,
+      "step": 301
+    },
+    {
+      "epoch": 0.11243615548393852,
+      "grad_norm": 0.10296988487243652,
+      "learning_rate": 6.040000000000001e-06,
+      "loss": 1.4355,
+      "step": 302
+    },
+    {
+      "epoch": 0.1128084606345476,
+      "grad_norm": 0.09838810563087463,
+      "learning_rate": 6.0600000000000004e-06,
+      "loss": 1.4705,
+      "step": 303
+    },
+    {
+      "epoch": 0.11318076578515666,
+      "grad_norm": 0.14201873540878296,
+      "learning_rate": 6.08e-06,
+      "loss": 1.465,
+      "step": 304
+    },
+    {
+      "epoch": 0.11355307093576572,
+      "grad_norm": 0.10391878336668015,
+      "learning_rate": 6.1e-06,
+      "loss": 1.4691,
+      "step": 305
+    },
+    {
+      "epoch": 0.11392537608637479,
+      "grad_norm": 0.0908937081694603,
+      "learning_rate": 6.120000000000001e-06,
+      "loss": 1.4738,
+      "step": 306
+    },
+    {
+      "epoch": 0.11429768123698386,
+      "grad_norm": 0.1304592341184616,
+      "learning_rate": 6.1400000000000005e-06,
+      "loss": 1.4564,
+      "step": 307
+    },
+    {
+      "epoch": 0.11466998638759293,
+      "grad_norm": 0.11030007898807526,
+      "learning_rate": 6.16e-06,
+      "loss": 1.4681,
+      "step": 308
+    },
+    {
+      "epoch": 0.115042291538202,
+      "grad_norm": 0.1404293328523636,
+      "learning_rate": 6.18e-06,
+      "loss": 1.4489,
+      "step": 309
+    },
+    {
+      "epoch": 0.11541459668881107,
+      "grad_norm": 0.15717652440071106,
+      "learning_rate": 6.200000000000001e-06,
+      "loss": 1.4619,
+      "step": 310
+    },
+    {
+      "epoch": 0.11578690183942013,
+      "grad_norm": 0.08294162154197693,
+      "learning_rate": 6.220000000000001e-06,
+      "loss": 1.4564,
+      "step": 311
+    },
+    {
+      "epoch": 0.1161592069900292,
+      "grad_norm": 0.12932319939136505,
+      "learning_rate": 6.24e-06,
+      "loss": 1.4615,
+      "step": 312
+    },
+    {
+      "epoch": 0.11653151214063827,
+      "grad_norm": 0.1291813999414444,
+      "learning_rate": 6.26e-06,
+      "loss": 1.4736,
+      "step": 313
+    },
+    {
+      "epoch": 0.11690381729124734,
+      "grad_norm": 0.08296732604503632,
+      "learning_rate": 6.280000000000001e-06,
+      "loss": 1.4592,
+      "step": 314
+    },
+    {
+      "epoch": 0.11727612244185641,
+      "grad_norm": 0.1045786589384079,
+      "learning_rate": 6.300000000000001e-06,
+      "loss": 1.4473,
+      "step": 315
+    },
+    {
+      "epoch": 0.11764842759246548,
+      "grad_norm": 0.14720311760902405,
+      "learning_rate": 6.3200000000000005e-06,
+      "loss": 1.45,
+      "step": 316
+    },
+    {
+      "epoch": 0.11802073274307454,
+      "grad_norm": 0.08579614013433456,
+      "learning_rate": 6.34e-06,
+      "loss": 1.4448,
+      "step": 317
+    },
+    {
+      "epoch": 0.1183930378936836,
+      "grad_norm": 0.12133117765188217,
+      "learning_rate": 6.360000000000001e-06,
+      "loss": 1.4536,
+      "step": 318
+    },
+    {
+      "epoch": 0.11876534304429268,
+      "grad_norm": 0.09560643136501312,
+      "learning_rate": 6.380000000000001e-06,
+      "loss": 1.4501,
+      "step": 319
+    },
+    {
+      "epoch": 0.11913764819490175,
+      "grad_norm": 0.08440724015235901,
+      "learning_rate": 6.4000000000000006e-06,
+      "loss": 1.4516,
+      "step": 320
+    },
+    {
+      "epoch": 0.11950995334551082,
+      "grad_norm": 0.0895630419254303,
+      "learning_rate": 6.42e-06,
+      "loss": 1.4537,
+      "step": 321
+    },
+    {
+      "epoch": 0.11988225849611989,
+      "grad_norm": 0.10389460623264313,
+      "learning_rate": 6.440000000000001e-06,
+      "loss": 1.4552,
+      "step": 322
+    },
+    {
+      "epoch": 0.12025456364672896,
+      "grad_norm": 0.09369999170303345,
+      "learning_rate": 6.460000000000001e-06,
+      "loss": 1.4544,
+      "step": 323
+    },
+    {
+      "epoch": 0.12062686879733801,
+      "grad_norm": 0.099007248878479,
+      "learning_rate": 6.480000000000001e-06,
+      "loss": 1.4891,
+      "step": 324
+    },
+    {
+      "epoch": 0.12099917394794708,
+      "grad_norm": 0.09093964099884033,
+      "learning_rate": 6.5000000000000004e-06,
+      "loss": 1.4671,
+      "step": 325
+    },
+    {
+      "epoch": 0.12137147909855615,
+      "grad_norm": 0.09165001660585403,
+      "learning_rate": 6.520000000000001e-06,
+      "loss": 1.4577,
+      "step": 326
+    },
+    {
+      "epoch": 0.12174378424916522,
+      "grad_norm": 0.09465596079826355,
+      "learning_rate": 6.540000000000001e-06,
+      "loss": 1.4737,
+      "step": 327
+    },
+    {
+      "epoch": 0.1221160893997743,
+      "grad_norm": 0.09906231611967087,
+      "learning_rate": 6.560000000000001e-06,
+      "loss": 1.4587,
+      "step": 328
+    },
+    {
+      "epoch": 0.12248839455038336,
+      "grad_norm": 0.09170754998922348,
+      "learning_rate": 6.5800000000000005e-06,
+      "loss": 1.4604,
+      "step": 329
+    },
+    {
+      "epoch": 0.12286069970099242,
+      "grad_norm": 0.08741523325443268,
+      "learning_rate": 6.600000000000001e-06,
+      "loss": 1.4563,
+      "step": 330
+    },
+    {
+      "epoch": 0.12323300485160149,
+      "grad_norm": 0.1020127385854721,
+      "learning_rate": 6.620000000000001e-06,
+      "loss": 1.4656,
+      "step": 331
+    },
+    {
+      "epoch": 0.12360531000221056,
+      "grad_norm": 0.09506560862064362,
+      "learning_rate": 6.640000000000001e-06,
+      "loss": 1.4752,
+      "step": 332
+    },
+    {
+      "epoch": 0.12397761515281963,
+      "grad_norm": 0.1042810007929802,
+      "learning_rate": 6.660000000000001e-06,
+      "loss": 1.4593,
+      "step": 333
+    },
+    {
+      "epoch": 0.1243499203034287,
+      "grad_norm": 0.09513174742460251,
+      "learning_rate": 6.680000000000001e-06,
+      "loss": 1.4636,
+      "step": 334
+    },
+    {
+      "epoch": 0.12472222545403777,
+      "grad_norm": 0.09068495035171509,
+      "learning_rate": 6.700000000000001e-06,
+      "loss": 1.4729,
+      "step": 335
+    },
+    {
+      "epoch": 0.12509453060464684,
+      "grad_norm": 0.09143901616334915,
+      "learning_rate": 6.720000000000001e-06,
+      "loss": 1.4528,
+      "step": 336
+    },
+    {
+      "epoch": 0.1254668357552559,
+      "grad_norm": 0.08922934532165527,
+      "learning_rate": 6.740000000000001e-06,
+      "loss": 1.4668,
+      "step": 337
+    },
+    {
+      "epoch": 0.12583914090586498,
+      "grad_norm": 0.10561850666999817,
+      "learning_rate": 6.760000000000001e-06,
+      "loss": 1.4536,
+      "step": 338
+    },
+    {
+      "epoch": 0.12621144605647402,
+      "grad_norm": 0.09490080922842026,
+      "learning_rate": 6.780000000000001e-06,
+      "loss": 1.4604,
+      "step": 339
+    },
+    {
+      "epoch": 0.1265837512070831,
+      "grad_norm": 0.09219174832105637,
+      "learning_rate": 6.800000000000001e-06,
+      "loss": 1.4577,
+      "step": 340
+    },
+    {
+      "epoch": 0.12695605635769217,
+      "grad_norm": 0.11259440332651138,
+      "learning_rate": 6.820000000000001e-06,
+      "loss": 1.461,
+      "step": 341
+    },
+    {
+      "epoch": 0.12732836150830124,
+      "grad_norm": 0.10154417902231216,
+      "learning_rate": 6.8400000000000014e-06,
+      "loss": 1.4539,
+      "step": 342
+    },
+    {
+      "epoch": 0.1277006666589103,
+      "grad_norm": 0.08527655154466629,
+      "learning_rate": 6.860000000000001e-06,
+      "loss": 1.4516,
+      "step": 343
+    },
+    {
+      "epoch": 0.12807297180951938,
+      "grad_norm": 0.10697508603334427,
+      "learning_rate": 6.88e-06,
+      "loss": 1.4519,
+      "step": 344
+    },
+    {
+      "epoch": 0.12844527696012845,
+      "grad_norm": 0.1440954953432083,
+      "learning_rate": 6.9e-06,
+      "loss": 1.4394,
+      "step": 345
+    },
+    {
+      "epoch": 0.12881758211073752,
+      "grad_norm": 0.10360608994960785,
+      "learning_rate": 6.92e-06,
+      "loss": 1.4472,
+      "step": 346
+    },
+    {
+      "epoch": 0.1291898872613466,
+      "grad_norm": 0.09860967099666595,
+      "learning_rate": 6.9400000000000005e-06,
+      "loss": 1.4362,
+      "step": 347
+    },
+    {
+      "epoch": 0.12956219241195566,
+      "grad_norm": 0.10211774706840515,
+      "learning_rate": 6.96e-06,
+      "loss": 1.4497,
+      "step": 348
+    },
+    {
+      "epoch": 0.12993449756256473,
+      "grad_norm": 0.10965809226036072,
+      "learning_rate": 6.98e-06,
+      "loss": 1.457,
+      "step": 349
+    },
+    {
+      "epoch": 0.1303068027131738,
+      "grad_norm": 0.09159507602453232,
+      "learning_rate": 7e-06,
+      "loss": 1.4512,
+      "step": 350
+    },
+    {
+      "epoch": 0.13067910786378284,
+      "grad_norm": 0.11210039258003235,
+      "learning_rate": 7.0200000000000006e-06,
+      "loss": 1.4418,
+      "step": 351
+    },
+    {
+      "epoch": 0.1310514130143919,
+      "grad_norm": 0.0950743705034256,
+      "learning_rate": 7.04e-06,
+      "loss": 1.4443,
+      "step": 352
+    },
+    {
+      "epoch": 0.13142371816500098,
+      "grad_norm": 0.10520799458026886,
+      "learning_rate": 7.06e-06,
+      "loss": 1.4587,
+      "step": 353
+    },
+    {
+      "epoch": 0.13179602331561005,
+      "grad_norm": 0.09574756771326065,
+      "learning_rate": 7.08e-06,
+      "loss": 1.4353,
+      "step": 354
+    },
+    {
+      "epoch": 0.13216832846621912,
+      "grad_norm": 0.09506519138813019,
+      "learning_rate": 7.100000000000001e-06,
+      "loss": 1.4425,
+      "step": 355
+    },
+    {
+      "epoch": 0.1325406336168282,
+      "grad_norm": 0.09055452048778534,
+      "learning_rate": 7.1200000000000004e-06,
+      "loss": 1.4514,
+      "step": 356
+    },
+    {
+      "epoch": 0.13291293876743726,
+      "grad_norm": 0.08780631422996521,
+      "learning_rate": 7.14e-06,
+      "loss": 1.4378,
+      "step": 357
+    },
+    {
+      "epoch": 0.13328524391804633,
+      "grad_norm": 0.09684795886278152,
+      "learning_rate": 7.16e-06,
+      "loss": 1.4358,
+      "step": 358
+    },
+    {
+      "epoch": 0.1336575490686554,
+      "grad_norm": 0.09261483699083328,
+      "learning_rate": 7.180000000000001e-06,
+      "loss": 1.4521,
+      "step": 359
+    },
+    {
+      "epoch": 0.13402985421926447,
+      "grad_norm": 0.09582682698965073,
+      "learning_rate": 7.2000000000000005e-06,
+      "loss": 1.4393,
+      "step": 360
+    },
+    {
+      "epoch": 0.13440215936987354,
+      "grad_norm": 0.0970984697341919,
+      "learning_rate": 7.22e-06,
+      "loss": 1.4289,
+      "step": 361
+    },
+    {
+      "epoch": 0.1347744645204826,
+      "grad_norm": 0.10083375871181488,
+      "learning_rate": 7.24e-06,
+      "loss": 1.4493,
+      "step": 362
+    },
+    {
+      "epoch": 0.13514676967109168,
+      "grad_norm": 0.11022008955478668,
+      "learning_rate": 7.260000000000001e-06,
+      "loss": 1.4504,
+      "step": 363
+    },
+    {
+      "epoch": 0.13551907482170072,
+      "grad_norm": 0.09099752455949783,
+      "learning_rate": 7.280000000000001e-06,
+      "loss": 1.4439,
+      "step": 364
+    },
+    {
+      "epoch": 0.1358913799723098,
+      "grad_norm": 0.13994981348514557,
+      "learning_rate": 7.3e-06,
+      "loss": 1.4537,
+      "step": 365
+    },
+    {
+      "epoch": 0.13626368512291887,
+      "grad_norm": 0.0927768424153328,
+      "learning_rate": 7.32e-06,
+      "loss": 1.4577,
+      "step": 366
+    },
+    {
+      "epoch": 0.13663599027352794,
+      "grad_norm": 0.12550784647464752,
+      "learning_rate": 7.340000000000001e-06,
+      "loss": 1.4595,
+      "step": 367
+    },
+    {
+      "epoch": 0.137008295424137,
+      "grad_norm": 0.0998741090297699,
+      "learning_rate": 7.360000000000001e-06,
+      "loss": 1.4606,
+      "step": 368
+    },
+    {
+      "epoch": 0.13738060057474608,
+      "grad_norm": 0.09384217113256454,
+      "learning_rate": 7.3800000000000005e-06,
+      "loss": 1.4388,
+      "step": 369
+    },
+    {
+      "epoch": 0.13775290572535515,
+      "grad_norm": 0.10957197099924088,
+      "learning_rate": 7.4e-06,
+      "loss": 1.4463,
+      "step": 370
+    },
+    {
+      "epoch": 0.13812521087596422,
+      "grad_norm": 0.10382582992315292,
+      "learning_rate": 7.420000000000001e-06,
+      "loss": 1.4393,
+      "step": 371
+    },
+    {
+      "epoch": 0.1384975160265733,
+      "grad_norm": 0.09171084314584732,
+      "learning_rate": 7.440000000000001e-06,
+      "loss": 1.4493,
+      "step": 372
+    },
+    {
+      "epoch": 0.13886982117718236,
+      "grad_norm": 0.10419381409883499,
+      "learning_rate": 7.4600000000000006e-06,
+      "loss": 1.4589,
+      "step": 373
+    },
+    {
+      "epoch": 0.13924212632779143,
+      "grad_norm": 0.0996531993150711,
+      "learning_rate": 7.48e-06,
+      "loss": 1.4258,
+      "step": 374
+    },
+    {
+      "epoch": 0.1396144314784005,
+      "grad_norm": 0.09250695258378983,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 1.4458,
+      "step": 375
+    },
+    {
+      "epoch": 0.13998673662900954,
+      "grad_norm": 0.09015652537345886,
+      "learning_rate": 7.520000000000001e-06,
+      "loss": 1.452,
+      "step": 376
+    },
+    {
+      "epoch": 0.1403590417796186,
+      "grad_norm": 0.11032413691282272,
+      "learning_rate": 7.540000000000001e-06,
+      "loss": 1.4307,
+      "step": 377
+    },
+    {
+      "epoch": 0.14073134693022768,
+      "grad_norm": 0.09179755300283432,
+      "learning_rate": 7.5600000000000005e-06,
+      "loss": 1.4462,
+      "step": 378
+    },
+    {
+      "epoch": 0.14110365208083675,
+      "grad_norm": 0.09686073660850525,
+      "learning_rate": 7.58e-06,
+      "loss": 1.4362,
+      "step": 379
+    },
+    {
+      "epoch": 0.14147595723144582,
+      "grad_norm": 0.09349130094051361,
+      "learning_rate": 7.600000000000001e-06,
+      "loss": 1.4358,
+      "step": 380
+    },
+    {
+      "epoch": 0.1418482623820549,
+      "grad_norm": 0.10256840288639069,
+      "learning_rate": 7.620000000000001e-06,
+      "loss": 1.4469,
+      "step": 381
+    },
+    {
+      "epoch": 0.14222056753266396,
+      "grad_norm": 0.09944120049476624,
+      "learning_rate": 7.640000000000001e-06,
+      "loss": 1.4485,
+      "step": 382
+    },
+    {
+      "epoch": 0.14259287268327303,
+      "grad_norm": 0.09924148768186569,
+      "learning_rate": 7.660000000000001e-06,
+      "loss": 1.4382,
+      "step": 383
+    },
+    {
+      "epoch": 0.1429651778338821,
+      "grad_norm": 0.09792005270719528,
+      "learning_rate": 7.680000000000001e-06,
+      "loss": 1.4256,
+      "step": 384
+    },
+    {
+      "epoch": 0.14333748298449117,
+      "grad_norm": 0.08989827334880829,
+      "learning_rate": 7.7e-06,
+      "loss": 1.441,
+      "step": 385
+    },
+    {
+      "epoch": 0.14370978813510024,
+      "grad_norm": 0.13259711861610413,
+      "learning_rate": 7.72e-06,
+      "loss": 1.4604,
+      "step": 386
+    },
+    {
+      "epoch": 0.1440820932857093,
+      "grad_norm": 0.10013754665851593,
+      "learning_rate": 7.74e-06,
+      "loss": 1.4319,
+      "step": 387
+    },
+    {
+      "epoch": 0.14445439843631835,
+      "grad_norm": 0.10801331698894501,
+      "learning_rate": 7.76e-06,
+      "loss": 1.4358,
+      "step": 388
+    },
+    {
+      "epoch": 0.14482670358692742,
+      "grad_norm": 0.0908760130405426,
+      "learning_rate": 7.78e-06,
+      "loss": 1.4496,
+      "step": 389
+    },
+    {
+      "epoch": 0.1451990087375365,
+      "grad_norm": 0.09459855407476425,
+      "learning_rate": 7.800000000000002e-06,
+      "loss": 1.4482,
+      "step": 390
+    },
+    {
+      "epoch": 0.14557131388814556,
+      "grad_norm": 0.09493505209684372,
+      "learning_rate": 7.820000000000001e-06,
+      "loss": 1.443,
+      "step": 391
+    },
+    {
+      "epoch": 0.14594361903875464,
+      "grad_norm": 0.09393203258514404,
+      "learning_rate": 7.840000000000001e-06,
+      "loss": 1.4461,
+      "step": 392
+    },
+    {
+      "epoch": 0.1463159241893637,
+      "grad_norm": 0.0932602807879448,
+      "learning_rate": 7.860000000000001e-06,
+      "loss": 1.4193,
+      "step": 393
+    },
+    {
+      "epoch": 0.14668822933997278,
+      "grad_norm": 0.10115011781454086,
+      "learning_rate": 7.88e-06,
+      "loss": 1.4143,
+      "step": 394
+    },
+    {
+      "epoch": 0.14706053449058185,
+      "grad_norm": 0.10510238260030746,
+      "learning_rate": 7.9e-06,
+      "loss": 1.4377,
+      "step": 395
+    },
+    {
+      "epoch": 0.14743283964119092,
+      "grad_norm": 0.09857072681188583,
+      "learning_rate": 7.92e-06,
+      "loss": 1.4537,
+      "step": 396
+    },
+    {
+      "epoch": 0.14780514479179999,
+      "grad_norm": 0.11328350752592087,
+      "learning_rate": 7.94e-06,
+      "loss": 1.4274,
+      "step": 397
+    },
+    {
+      "epoch": 0.14817744994240906,
+      "grad_norm": 0.10213533043861389,
+      "learning_rate": 7.960000000000002e-06,
+      "loss": 1.4282,
+      "step": 398
+    },
+    {
+      "epoch": 0.14854975509301813,
+      "grad_norm": 0.1178865134716034,
+      "learning_rate": 7.980000000000002e-06,
+      "loss": 1.4333,
+      "step": 399
+    },
+    {
+      "epoch": 0.1489220602436272,
+      "grad_norm": 0.11789101362228394,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 1.4413,
+      "step": 400
+    },
+    {
+      "epoch": 0.14929436539423624,
+      "grad_norm": 0.10633790493011475,
+      "learning_rate": 8.020000000000001e-06,
+      "loss": 1.4281,
+      "step": 401
+    },
+    {
+      "epoch": 0.1496666705448453,
+      "grad_norm": 0.11100617051124573,
+      "learning_rate": 8.040000000000001e-06,
+      "loss": 1.4299,
+      "step": 402
+    },
+    {
+      "epoch": 0.15003897569545438,
+      "grad_norm": 0.09987955540418625,
+      "learning_rate": 8.06e-06,
+      "loss": 1.4396,
+      "step": 403
+    },
+    {
+      "epoch": 0.15041128084606345,
+      "grad_norm": 0.13840006291866302,
+      "learning_rate": 8.08e-06,
+      "loss": 1.4378,
+      "step": 404
+    },
+    {
+      "epoch": 0.15078358599667252,
+      "grad_norm": 0.10468387603759766,
+      "learning_rate": 8.1e-06,
+      "loss": 1.4361,
+      "step": 405
+    },
+    {
+      "epoch": 0.1511558911472816,
+      "grad_norm": 0.09684094786643982,
+      "learning_rate": 8.120000000000002e-06,
+      "loss": 1.4408,
+      "step": 406
+    },
+    {
+      "epoch": 0.15152819629789066,
+      "grad_norm": 0.1058192178606987,
+      "learning_rate": 8.14e-06,
+      "loss": 1.4353,
+      "step": 407
+    },
+    {
+      "epoch": 0.15190050144849973,
+      "grad_norm": 0.11632636189460754,
+      "learning_rate": 8.16e-06,
+      "loss": 1.4331,
+      "step": 408
+    },
+    {
+      "epoch": 0.1522728065991088,
+      "grad_norm": 0.10219406336545944,
+      "learning_rate": 8.18e-06,
+      "loss": 1.4527,
+      "step": 409
+    },
+    {
+      "epoch": 0.15264511174971787,
+      "grad_norm": 0.110807865858078,
+      "learning_rate": 8.2e-06,
+      "loss": 1.4389,
+      "step": 410
+    },
+    {
+      "epoch": 0.15301741690032694,
+      "grad_norm": 0.09530390799045563,
+      "learning_rate": 8.220000000000001e-06,
+      "loss": 1.4252,
+      "step": 411
+    },
+    {
+      "epoch": 0.153389722050936,
+      "grad_norm": 0.10499103367328644,
+      "learning_rate": 8.24e-06,
+      "loss": 1.4429,
+      "step": 412
+    },
+    {
+      "epoch": 0.15376202720154505,
+      "grad_norm": 0.1094844788312912,
+      "learning_rate": 8.26e-06,
+      "loss": 1.4172,
+      "step": 413
+    },
+    {
+      "epoch": 0.15413433235215412,
+      "grad_norm": 0.10519949346780777,
+      "learning_rate": 8.28e-06,
+      "loss": 1.4371,
+      "step": 414
+    },
+    {
+      "epoch": 0.1545066375027632,
+      "grad_norm": 0.10553234815597534,
+      "learning_rate": 8.3e-06,
+      "loss": 1.4409,
+      "step": 415
+    },
+    {
+      "epoch": 0.15487894265337226,
+      "grad_norm": 0.10840829461812973,
+      "learning_rate": 8.32e-06,
+      "loss": 1.4345,
+      "step": 416
+    },
+    {
+      "epoch": 0.15525124780398133,
+      "grad_norm": 0.09976733475923538,
+      "learning_rate": 8.34e-06,
+      "loss": 1.4508,
+      "step": 417
+    },
+    {
+      "epoch": 0.1556235529545904,
+      "grad_norm": 0.10259655117988586,
+      "learning_rate": 8.36e-06,
+      "loss": 1.4196,
+      "step": 418
+    },
+    {
+      "epoch": 0.15599585810519948,
+      "grad_norm": 0.10481464862823486,
+      "learning_rate": 8.380000000000001e-06,
+      "loss": 1.4367,
+      "step": 419
+    },
+    {
+      "epoch": 0.15636816325580855,
+      "grad_norm": 0.1130518987774849,
+      "learning_rate": 8.400000000000001e-06,
+      "loss": 1.4479,
+      "step": 420
+    },
+    {
+      "epoch": 0.15674046840641762,
+      "grad_norm": 0.11022792756557465,
+      "learning_rate": 8.42e-06,
+      "loss": 1.421,
+      "step": 421
+    },
+    {
+      "epoch": 0.15711277355702669,
+      "grad_norm": 0.0973302498459816,
+      "learning_rate": 8.44e-06,
+      "loss": 1.4445,
+      "step": 422
+    },
+    {
+      "epoch": 0.15748507870763576,
+      "grad_norm": 0.09701745212078094,
+      "learning_rate": 8.46e-06,
+      "loss": 1.42,
+      "step": 423
+    },
+    {
+      "epoch": 0.15785738385824483,
+      "grad_norm": 0.09979844093322754,
+      "learning_rate": 8.48e-06,
+      "loss": 1.4221,
+      "step": 424
+    },
+    {
+      "epoch": 0.15822968900885387,
+      "grad_norm": 0.099349245429039,
+      "learning_rate": 8.5e-06,
+      "loss": 1.4131,
+      "step": 425
+    },
+    {
+      "epoch": 0.15860199415946294,
+      "grad_norm": 0.10651430487632751,
+      "learning_rate": 8.52e-06,
+      "loss": 1.4249,
+      "step": 426
+    },
+    {
+      "epoch": 0.158974299310072,
+      "grad_norm": 0.11765170842409134,
+      "learning_rate": 8.540000000000001e-06,
+      "loss": 1.4503,
+      "step": 427
+    },
+    {
+      "epoch": 0.15934660446068108,
+      "grad_norm": 0.10456649959087372,
+      "learning_rate": 8.560000000000001e-06,
+      "loss": 1.4261,
+      "step": 428
+    },
+    {
+      "epoch": 0.15971890961129015,
+      "grad_norm": 0.13042724132537842,
+      "learning_rate": 8.580000000000001e-06,
+      "loss": 1.4238,
+      "step": 429
+    },
+    {
+      "epoch": 0.16009121476189922,
+      "grad_norm": 0.10613591969013214,
+      "learning_rate": 8.6e-06,
+      "loss": 1.4097,
+      "step": 430
+    },
+    {
+      "epoch": 0.1604635199125083,
+      "grad_norm": 0.10777163505554199,
+      "learning_rate": 8.62e-06,
+      "loss": 1.4327,
+      "step": 431
+    },
+    {
+      "epoch": 0.16083582506311736,
+      "grad_norm": 0.1047920435667038,
+      "learning_rate": 8.64e-06,
+      "loss": 1.4247,
+      "step": 432
+    },
+    {
+      "epoch": 0.16120813021372643,
+      "grad_norm": 0.09752184897661209,
+      "learning_rate": 8.66e-06,
+      "loss": 1.4327,
+      "step": 433
+    },
+    {
+      "epoch": 0.1615804353643355,
+      "grad_norm": 0.1117076650261879,
+      "learning_rate": 8.68e-06,
+      "loss": 1.4327,
+      "step": 434
+    },
+    {
+      "epoch": 0.16195274051494457,
+      "grad_norm": 0.11003684252500534,
+      "learning_rate": 8.700000000000001e-06,
+      "loss": 1.4211,
+      "step": 435
+    },
+    {
+      "epoch": 0.16232504566555364,
+      "grad_norm": 0.09675519168376923,
+      "learning_rate": 8.720000000000001e-06,
+      "loss": 1.435,
+      "step": 436
+    },
+    {
+      "epoch": 0.16269735081616268,
+      "grad_norm": 0.09906768053770065,
+      "learning_rate": 8.740000000000001e-06,
+      "loss": 1.4381,
+      "step": 437
+    },
+    {
+      "epoch": 0.16306965596677175,
+      "grad_norm": 0.10028046369552612,
+      "learning_rate": 8.76e-06,
+      "loss": 1.4251,
+      "step": 438
+    },
+    {
+      "epoch": 0.16344196111738082,
+      "grad_norm": 0.11393031477928162,
+      "learning_rate": 8.78e-06,
+      "loss": 1.4234,
+      "step": 439
+    },
+    {
+      "epoch": 0.1638142662679899,
+      "grad_norm": 0.10733836144208908,
+      "learning_rate": 8.8e-06,
+      "loss": 1.4168,
+      "step": 440
+    },
+    {
+      "epoch": 0.16418657141859896,
+      "grad_norm": 0.10900059342384338,
+      "learning_rate": 8.82e-06,
+      "loss": 1.4356,
+      "step": 441
+    },
+    {
+      "epoch": 0.16455887656920803,
+      "grad_norm": 0.10403525829315186,
+      "learning_rate": 8.84e-06,
+      "loss": 1.4058,
+      "step": 442
+    },
+    {
+      "epoch": 0.1649311817198171,
+      "grad_norm": 0.10996660590171814,
+      "learning_rate": 8.860000000000002e-06,
+      "loss": 1.4317,
+      "step": 443
+    },
+    {
+      "epoch": 0.16530348687042618,
+      "grad_norm": 0.1056893914937973,
+      "learning_rate": 8.880000000000001e-06,
+      "loss": 1.4258,
+      "step": 444
+    },
+    {
+      "epoch": 0.16567579202103525,
+      "grad_norm": 0.10371371358633041,
+      "learning_rate": 8.900000000000001e-06,
+      "loss": 1.4261,
+      "step": 445
+    },
+    {
+      "epoch": 0.16604809717164432,
+      "grad_norm": 0.10245388001203537,
+      "learning_rate": 8.920000000000001e-06,
+      "loss": 1.4237,
+      "step": 446
+    },
+    {
+      "epoch": 0.16642040232225339,
+      "grad_norm": 0.11346007138490677,
+      "learning_rate": 8.94e-06,
+      "loss": 1.4171,
+      "step": 447
+    },
+    {
+      "epoch": 0.16679270747286246,
+      "grad_norm": 0.1109023243188858,
+      "learning_rate": 8.96e-06,
+      "loss": 1.4311,
+      "step": 448
+    },
+    {
+      "epoch": 0.16716501262347153,
+      "grad_norm": 0.10533205419778824,
+      "learning_rate": 8.98e-06,
+      "loss": 1.4159,
+      "step": 449
+    },
+    {
+      "epoch": 0.16753731777408057,
+      "grad_norm": 0.1102309301495552,
+      "learning_rate": 9e-06,
+      "loss": 1.425,
+      "step": 450
+    },
+    {
+      "epoch": 0.16790962292468964,
+      "grad_norm": 0.10830759257078171,
+      "learning_rate": 9.020000000000002e-06,
+      "loss": 1.4188,
+      "step": 451
+    },
+    {
+      "epoch": 0.1682819280752987,
+      "grad_norm": 0.11227082461118698,
+      "learning_rate": 9.040000000000002e-06,
+      "loss": 1.4189,
+      "step": 452
+    },
+    {
+      "epoch": 0.16865423322590778,
+      "grad_norm": 0.11561018973588943,
+      "learning_rate": 9.060000000000001e-06,
+      "loss": 1.4225,
+      "step": 453
+    },
+    {
+      "epoch": 0.16902653837651685,
+      "grad_norm": 0.10718375444412231,
+      "learning_rate": 9.080000000000001e-06,
+      "loss": 1.437,
+      "step": 454
+    },
+    {
+      "epoch": 0.16939884352712592,
+      "grad_norm": 0.11734096705913544,
+      "learning_rate": 9.100000000000001e-06,
+      "loss": 1.4243,
+      "step": 455
+    },
+    {
+      "epoch": 0.169771148677735,
+      "grad_norm": 0.11969781666994095,
+      "learning_rate": 9.12e-06,
+      "loss": 1.4048,
+      "step": 456
+    },
+    {
+      "epoch": 0.17014345382834406,
+      "grad_norm": 0.1117396354675293,
+      "learning_rate": 9.14e-06,
+      "loss": 1.4281,
+      "step": 457
+    },
+    {
+      "epoch": 0.17051575897895313,
+      "grad_norm": 0.10435774177312851,
+      "learning_rate": 9.16e-06,
+      "loss": 1.429,
+      "step": 458
+    },
+    {
+      "epoch": 0.1708880641295622,
+      "grad_norm": 0.10865868628025055,
+      "learning_rate": 9.180000000000002e-06,
+      "loss": 1.4092,
+      "step": 459
+    },
+    {
+      "epoch": 0.17126036928017127,
+      "grad_norm": 0.1147746816277504,
+      "learning_rate": 9.200000000000002e-06,
+      "loss": 1.4197,
+      "step": 460
+    },
+    {
+      "epoch": 0.17163267443078034,
+      "grad_norm": 0.1095675453543663,
+      "learning_rate": 9.220000000000002e-06,
+      "loss": 1.425,
+      "step": 461
+    },
+    {
+      "epoch": 0.17200497958138938,
+      "grad_norm": 0.10761556029319763,
+      "learning_rate": 9.240000000000001e-06,
+      "loss": 1.4152,
+      "step": 462
+    },
+    {
+      "epoch": 0.17237728473199845,
+      "grad_norm": 0.11335011571645737,
+      "learning_rate": 9.260000000000001e-06,
+      "loss": 1.4224,
+      "step": 463
+    },
+    {
+      "epoch": 0.17274958988260752,
+      "grad_norm": 0.10845465958118439,
+      "learning_rate": 9.280000000000001e-06,
+      "loss": 1.4005,
+      "step": 464
+    },
+    {
+      "epoch": 0.1731218950332166,
+      "grad_norm": 0.1037868782877922,
+      "learning_rate": 9.3e-06,
+      "loss": 1.4277,
+      "step": 465
+    },
+    {
+      "epoch": 0.17349420018382566,
+      "grad_norm": 0.11700176447629929,
+      "learning_rate": 9.32e-06,
+      "loss": 1.4232,
+      "step": 466
+    },
+    {
+      "epoch": 0.17386650533443473,
+      "grad_norm": 0.10587536543607712,
+      "learning_rate": 9.340000000000002e-06,
+      "loss": 1.4399,
+      "step": 467
+    },
+    {
+      "epoch": 0.1742388104850438,
+      "grad_norm": 0.11157836019992828,
+      "learning_rate": 9.360000000000002e-06,
+      "loss": 1.4159,
+      "step": 468
+    },
+    {
+      "epoch": 0.17461111563565287,
+      "grad_norm": 0.10884707421064377,
+      "learning_rate": 9.38e-06,
+      "loss": 1.4088,
+      "step": 469
+    },
+    {
+      "epoch": 0.17498342078626195,
+      "grad_norm": 0.11198698729276657,
+      "learning_rate": 9.4e-06,
+      "loss": 1.3952,
+      "step": 470
+    },
+    {
+      "epoch": 0.17535572593687102,
+      "grad_norm": 0.1030043363571167,
+      "learning_rate": 9.42e-06,
+      "loss": 1.4294,
+      "step": 471
+    },
+    {
+      "epoch": 0.17572803108748009,
+      "grad_norm": 0.11066511273384094,
+      "learning_rate": 9.440000000000001e-06,
+      "loss": 1.4114,
+      "step": 472
+    },
+    {
+      "epoch": 0.17610033623808916,
+      "grad_norm": 0.1187884584069252,
+      "learning_rate": 9.460000000000001e-06,
+      "loss": 1.415,
+      "step": 473
+    },
+    {
+      "epoch": 0.1764726413886982,
+      "grad_norm": 0.09824799001216888,
+      "learning_rate": 9.48e-06,
+      "loss": 1.4166,
+      "step": 474
+    },
+    {
+      "epoch": 0.17684494653930727,
+      "grad_norm": 0.09938450902700424,
+      "learning_rate": 9.5e-06,
+      "loss": 1.4139,
+      "step": 475
+    },
+    {
+      "epoch": 0.17721725168991634,
+      "grad_norm": 0.11766441911458969,
+      "learning_rate": 9.52e-06,
+      "loss": 1.4234,
+      "step": 476
+    },
+    {
+      "epoch": 0.1775895568405254,
+      "grad_norm": 0.10825181007385254,
+      "learning_rate": 9.54e-06,
+      "loss": 1.4207,
+      "step": 477
+    },
+    {
+      "epoch": 0.17796186199113448,
+      "grad_norm": 0.10996980965137482,
+      "learning_rate": 9.56e-06,
+      "loss": 1.4228,
+      "step": 478
+    },
+    {
+      "epoch": 0.17833416714174355,
+      "grad_norm": 0.11518129706382751,
+      "learning_rate": 9.58e-06,
+      "loss": 1.4155,
+      "step": 479
+    },
+    {
+      "epoch": 0.17870647229235262,
+      "grad_norm": 0.1234855204820633,
+      "learning_rate": 9.600000000000001e-06,
+      "loss": 1.4309,
+      "step": 480
+    },
+    {
+      "epoch": 0.1790787774429617,
+      "grad_norm": 0.10811053216457367,
+      "learning_rate": 9.620000000000001e-06,
+      "loss": 1.4211,
+      "step": 481
+    },
+    {
+      "epoch": 0.17945108259357076,
+      "grad_norm": 0.10534091293811798,
+      "learning_rate": 9.640000000000001e-06,
+      "loss": 1.4175,
+      "step": 482
+    },
+    {
+      "epoch": 0.17982338774417983,
+      "grad_norm": 0.12702758610248566,
+      "learning_rate": 9.66e-06,
+      "loss": 1.4104,
+      "step": 483
+    },
+    {
+      "epoch": 0.1801956928947889,
+      "grad_norm": 0.10806774348020554,
+      "learning_rate": 9.68e-06,
+      "loss": 1.4136,
+      "step": 484
+    },
+    {
+      "epoch": 0.18056799804539797,
+      "grad_norm": 0.10667596012353897,
+      "learning_rate": 9.7e-06,
+      "loss": 1.4111,
+      "step": 485
+    },
+    {
+      "epoch": 0.18094030319600704,
+      "grad_norm": 0.1084335595369339,
+      "learning_rate": 9.72e-06,
+      "loss": 1.424,
+      "step": 486
+    },
+    {
+      "epoch": 0.18131260834661608,
+      "grad_norm": 0.10393060743808746,
+      "learning_rate": 9.74e-06,
+      "loss": 1.413,
+      "step": 487
+    },
+    {
+      "epoch": 0.18168491349722515,
+      "grad_norm": 0.11034102737903595,
+      "learning_rate": 9.760000000000001e-06,
+      "loss": 1.4215,
+      "step": 488
+    },
+    {
+      "epoch": 0.18205721864783422,
+      "grad_norm": 0.1163318008184433,
+      "learning_rate": 9.780000000000001e-06,
+      "loss": 1.4123,
+      "step": 489
+    },
+    {
+      "epoch": 0.1824295237984433,
+      "grad_norm": 0.11191318929195404,
+      "learning_rate": 9.800000000000001e-06,
+      "loss": 1.4292,
+      "step": 490
+    },
+    {
+      "epoch": 0.18280182894905236,
+      "grad_norm": 0.10725145041942596,
+      "learning_rate": 9.820000000000001e-06,
+      "loss": 1.4084,
+      "step": 491
+    },
+    {
+      "epoch": 0.18317413409966143,
+      "grad_norm": 0.11110781133174896,
+      "learning_rate": 9.84e-06,
+      "loss": 1.394,
+      "step": 492
+    },
+    {
+      "epoch": 0.1835464392502705,
+      "grad_norm": 0.10888979583978653,
+      "learning_rate": 9.86e-06,
+      "loss": 1.4129,
+      "step": 493
+    },
+    {
+      "epoch": 0.18391874440087957,
+      "grad_norm": 0.11001749336719513,
+      "learning_rate": 9.88e-06,
+      "loss": 1.4248,
+      "step": 494
+    },
+    {
+      "epoch": 0.18429104955148864,
+      "grad_norm": 0.10707477480173111,
+      "learning_rate": 9.9e-06,
+      "loss": 1.4088,
+      "step": 495
+    },
+    {
+      "epoch": 0.18466335470209772,
+      "grad_norm": 0.12006059288978577,
+      "learning_rate": 9.920000000000002e-06,
+      "loss": 1.4126,
+      "step": 496
+    },
+    {
+      "epoch": 0.18503565985270679,
+      "grad_norm": 0.10763873159885406,
+      "learning_rate": 9.940000000000001e-06,
+      "loss": 1.4219,
+      "step": 497
+    },
+    {
+      "epoch": 0.18540796500331586,
+      "grad_norm": 0.10750039666891098,
+      "learning_rate": 9.960000000000001e-06,
+      "loss": 1.3938,
+      "step": 498
+    },
+    {
+      "epoch": 0.1857802701539249,
+      "grad_norm": 0.10844654589891434,
+      "learning_rate": 9.980000000000001e-06,
+      "loss": 1.4257,
+      "step": 499
+    },
+    {
+      "epoch": 0.18615257530453397,
+      "grad_norm": 0.10777752101421356,
+      "learning_rate": 1e-05,
+      "loss": 1.4042,
+      "step": 500
+    },
+    {
+      "epoch": 0.18615257530453397,
+      "eval_loss": 1.3990447521209717,
+      "eval_runtime": 16.7137,
+      "eval_samples_per_second": 103.747,
+      "eval_steps_per_second": 5.205,
+      "step": 500
+    },
+    {
+      "epoch": 0.18652488045514304,
+      "grad_norm": 0.12805776298046112,
+      "learning_rate": 1.002e-05,
+      "loss": 1.3923,
+      "step": 501
+    },
+    {
+      "epoch": 0.1868971856057521,
+      "grad_norm": 0.11597350984811783,
+      "learning_rate": 1.004e-05,
+      "loss": 1.4246,
+      "step": 502
+    },
+    {
+      "epoch": 0.18726949075636118,
+      "grad_norm": 0.12692782282829285,
+      "learning_rate": 1.006e-05,
+      "loss": 1.4091,
+      "step": 503
+    },
+    {
+      "epoch": 0.18764179590697025,
+      "grad_norm": 0.11857876926660538,
+      "learning_rate": 1.008e-05,
+      "loss": 1.4025,
+      "step": 504
+    },
+    {
+      "epoch": 0.18801410105757932,
+      "grad_norm": 0.11153510212898254,
+      "learning_rate": 1.0100000000000002e-05,
+      "loss": 1.4069,
+      "step": 505
+    },
+    {
+      "epoch": 0.1883864062081884,
+      "grad_norm": 0.11478788405656815,
+      "learning_rate": 1.0120000000000001e-05,
+      "loss": 1.4043,
+      "step": 506
+    },
+    {
+      "epoch": 0.18875871135879746,
+      "grad_norm": 0.11333264410495758,
+      "learning_rate": 1.0140000000000001e-05,
+      "loss": 1.4,
+      "step": 507
+    },
+    {
+      "epoch": 0.18913101650940653,
+      "grad_norm": 0.12227758020162582,
+      "learning_rate": 1.0160000000000001e-05,
+      "loss": 1.4023,
+      "step": 508
+    },
+    {
+      "epoch": 0.1895033216600156,
+      "grad_norm": 0.12102984637022018,
+      "learning_rate": 1.018e-05,
+      "loss": 1.4056,
+      "step": 509
+    },
+    {
+      "epoch": 0.18987562681062467,
+      "grad_norm": 0.11923938989639282,
+      "learning_rate": 1.02e-05,
+      "loss": 1.4019,
+      "step": 510
+    },
+    {
+      "epoch": 0.1902479319612337,
+      "grad_norm": 0.12334270030260086,
+      "learning_rate": 1.022e-05,
+      "loss": 1.3959,
+      "step": 511
+    },
+    {
+      "epoch": 0.19062023711184278,
+      "grad_norm": 0.12032870948314667,
+      "learning_rate": 1.024e-05,
+      "loss": 1.3994,
+      "step": 512
+    },
+    {
+      "epoch": 0.19099254226245185,
+      "grad_norm": 0.1170472502708435,
+      "learning_rate": 1.0260000000000002e-05,
+      "loss": 1.3959,
+      "step": 513
+    },
+    {
+      "epoch": 0.19136484741306092,
+      "grad_norm": 0.10825354605913162,
+      "learning_rate": 1.0280000000000002e-05,
+      "loss": 1.4031,
+      "step": 514
+    },
+    {
+      "epoch": 0.19173715256367,
+      "grad_norm": 0.11196181178092957,
+      "learning_rate": 1.0300000000000001e-05,
+      "loss": 1.4048,
+      "step": 515
+    },
+    {
+      "epoch": 0.19210945771427906,
+      "grad_norm": 0.11114434897899628,
+      "learning_rate": 1.0320000000000001e-05,
+      "loss": 1.4064,
+      "step": 516
+    },
+    {
+      "epoch": 0.19248176286488813,
+      "grad_norm": 0.11616680771112442,
+      "learning_rate": 1.0340000000000001e-05,
+      "loss": 1.3967,
+      "step": 517
+    },
+    {
+      "epoch": 0.1928540680154972,
+      "grad_norm": 0.11657610535621643,
+      "learning_rate": 1.036e-05,
+      "loss": 1.4152,
+      "step": 518
+    },
+    {
+      "epoch": 0.19322637316610627,
+      "grad_norm": 0.10928953438997269,
+      "learning_rate": 1.038e-05,
+      "loss": 1.4026,
+      "step": 519
+    },
+    {
+      "epoch": 0.19359867831671534,
+      "grad_norm": 0.11354228109121323,
+      "learning_rate": 1.04e-05,
+      "loss": 1.4026,
+      "step": 520
+    },
+    {
+      "epoch": 0.19397098346732441,
+      "grad_norm": 0.12005306780338287,
+      "learning_rate": 1.0420000000000002e-05,
+      "loss": 1.4094,
+      "step": 521
+    },
+    {
+      "epoch": 0.19434328861793349,
+      "grad_norm": 0.12376662343740463,
+      "learning_rate": 1.0440000000000002e-05,
+      "loss": 1.4076,
+      "step": 522
+    },
+    {
+      "epoch": 0.19471559376854256,
+      "grad_norm": 0.1191818043589592,
+      "learning_rate": 1.0460000000000001e-05,
+      "loss": 1.4055,
+      "step": 523
+    },
+    {
+      "epoch": 0.1950878989191516,
+      "grad_norm": 0.11545392870903015,
+      "learning_rate": 1.0480000000000001e-05,
+      "loss": 1.3881,
+      "step": 524
+    },
+    {
+      "epoch": 0.19546020406976067,
+      "grad_norm": 0.1315152794122696,
+      "learning_rate": 1.0500000000000001e-05,
+      "loss": 1.4119,
+      "step": 525
+    },
+    {
+      "epoch": 0.19583250922036974,
+      "grad_norm": 0.11896664649248123,
+      "learning_rate": 1.0520000000000001e-05,
+      "loss": 1.4165,
+      "step": 526
+    },
+    {
+      "epoch": 0.1962048143709788,
+      "grad_norm": 0.11113154143095016,
+      "learning_rate": 1.054e-05,
+      "loss": 1.3992,
+      "step": 527
+    },
+    {
+      "epoch": 0.19657711952158788,
+      "grad_norm": 0.13099145889282227,
+      "learning_rate": 1.056e-05,
+      "loss": 1.3997,
+      "step": 528
+    },
+    {
+      "epoch": 0.19694942467219695,
+      "grad_norm": 0.11786054819822311,
+      "learning_rate": 1.0580000000000002e-05,
+      "loss": 1.3967,
+      "step": 529
+    },
+    {
+      "epoch": 0.19732172982280602,
+      "grad_norm": 0.1341516375541687,
+      "learning_rate": 1.0600000000000002e-05,
+      "loss": 1.3993,
+      "step": 530
+    },
+    {
+      "epoch": 0.1976940349734151,
+      "grad_norm": 0.12682707607746124,
+      "learning_rate": 1.0620000000000002e-05,
+      "loss": 1.4148,
+      "step": 531
+    },
+    {
+      "epoch": 0.19806634012402416,
+      "grad_norm": 0.1188652366399765,
+      "learning_rate": 1.0640000000000001e-05,
+      "loss": 1.3933,
+      "step": 532
+    },
+    {
+      "epoch": 0.19843864527463323,
+      "grad_norm": 0.11299144476652145,
+      "learning_rate": 1.0660000000000001e-05,
+      "loss": 1.4004,
+      "step": 533
+    },
+    {
+      "epoch": 0.1988109504252423,
+      "grad_norm": 0.12654365599155426,
+      "learning_rate": 1.0680000000000001e-05,
+      "loss": 1.3929,
+      "step": 534
+    },
+    {
+      "epoch": 0.19918325557585137,
+      "grad_norm": 0.12072401493787766,
+      "learning_rate": 1.0700000000000001e-05,
+      "loss": 1.3967,
+      "step": 535
+    },
+    {
+      "epoch": 0.1995555607264604,
+      "grad_norm": 0.12298454344272614,
+      "learning_rate": 1.072e-05,
+      "loss": 1.4072,
+      "step": 536
+    },
+    {
+      "epoch": 0.19992786587706948,
+      "grad_norm": 0.127255380153656,
+      "learning_rate": 1.0740000000000002e-05,
+      "loss": 1.408,
+      "step": 537
+    },
+    {
+      "epoch": 0.20030017102767855,
+      "grad_norm": 0.12813320755958557,
+      "learning_rate": 1.0760000000000002e-05,
+      "loss": 1.4025,
+      "step": 538
+    },
+    {
+      "epoch": 0.20067247617828762,
+      "grad_norm": 0.12207869440317154,
+      "learning_rate": 1.0780000000000002e-05,
+      "loss": 1.3844,
+      "step": 539
+    },
+    {
+      "epoch": 0.2010447813288967,
+      "grad_norm": 0.11978594213724136,
+      "learning_rate": 1.0800000000000002e-05,
+      "loss": 1.3953,
+      "step": 540
+    },
+    {
+      "epoch": 0.20141708647950576,
+      "grad_norm": 0.12205009162425995,
+      "learning_rate": 1.0820000000000001e-05,
+      "loss": 1.3904,
+      "step": 541
+    },
+    {
+      "epoch": 0.20178939163011483,
+      "grad_norm": 0.11987481266260147,
+      "learning_rate": 1.0840000000000001e-05,
+      "loss": 1.3985,
+      "step": 542
+    },
+    {
+      "epoch": 0.2021616967807239,
+      "grad_norm": 0.12077482044696808,
+      "learning_rate": 1.0860000000000001e-05,
+      "loss": 1.4081,
+      "step": 543
+    },
+    {
+      "epoch": 0.20253400193133297,
+      "grad_norm": 0.11721371859312057,
+      "learning_rate": 1.0880000000000001e-05,
+      "loss": 1.4001,
+      "step": 544
+    },
+    {
+      "epoch": 0.20290630708194204,
+      "grad_norm": 0.11542621999979019,
+      "learning_rate": 1.0900000000000002e-05,
+      "loss": 1.4109,
+      "step": 545
+    },
+    {
+      "epoch": 0.20327861223255111,
+      "grad_norm": 0.12170755863189697,
+      "learning_rate": 1.0920000000000002e-05,
+      "loss": 1.3943,
+      "step": 546
+    },
+    {
+      "epoch": 0.20365091738316018,
+      "grad_norm": 0.11331729590892792,
+      "learning_rate": 1.0940000000000002e-05,
+      "loss": 1.404,
+      "step": 547
+    },
+    {
+      "epoch": 0.20402322253376923,
+      "grad_norm": 0.12473474442958832,
+      "learning_rate": 1.0960000000000002e-05,
+      "loss": 1.399,
+      "step": 548
+    },
+    {
+      "epoch": 0.2043955276843783,
+      "grad_norm": 0.12112352252006531,
+      "learning_rate": 1.0980000000000002e-05,
+      "loss": 1.394,
+      "step": 549
+    },
+    {
+      "epoch": 0.20476783283498737,
+      "grad_norm": 0.11570420861244202,
+      "learning_rate": 1.1000000000000001e-05,
+      "loss": 1.3961,
+      "step": 550
+    },
+    {
+      "epoch": 0.20514013798559644,
+      "grad_norm": 0.12518157064914703,
+      "learning_rate": 1.1020000000000001e-05,
+      "loss": 1.3929,
+      "step": 551
+    },
+    {
+      "epoch": 0.2055124431362055,
+      "grad_norm": 0.12412535399198532,
+      "learning_rate": 1.1040000000000001e-05,
+      "loss": 1.3953,
+      "step": 552
+    },
+    {
+      "epoch": 0.20588474828681458,
+      "grad_norm": 0.12580837309360504,
+      "learning_rate": 1.1060000000000003e-05,
+      "loss": 1.3887,
+      "step": 553
+    },
+    {
+      "epoch": 0.20625705343742365,
+      "grad_norm": 0.12166325002908707,
+      "learning_rate": 1.1080000000000002e-05,
+      "loss": 1.4035,
+      "step": 554
+    },
+    {
+      "epoch": 0.20662935858803272,
+      "grad_norm": 0.11380494385957718,
+      "learning_rate": 1.1100000000000002e-05,
+      "loss": 1.3881,
+      "step": 555
+    },
+    {
+      "epoch": 0.2070016637386418,
+      "grad_norm": 0.12607711553573608,
+      "learning_rate": 1.1120000000000002e-05,
+      "loss": 1.3911,
+      "step": 556
+    },
+    {
+      "epoch": 0.20737396888925086,
+      "grad_norm": 0.1209443137049675,
+      "learning_rate": 1.1140000000000002e-05,
+      "loss": 1.4025,
+      "step": 557
+    },
+    {
+      "epoch": 0.20774627403985993,
+      "grad_norm": 0.11875788867473602,
+      "learning_rate": 1.1160000000000002e-05,
+      "loss": 1.382,
+      "step": 558
+    },
+    {
+      "epoch": 0.208118579190469,
+      "grad_norm": 0.12261742353439331,
+      "learning_rate": 1.1180000000000001e-05,
+      "loss": 1.39,
+      "step": 559
+    },
+    {
+      "epoch": 0.20849088434107807,
+      "grad_norm": 0.13006705045700073,
+      "learning_rate": 1.1200000000000001e-05,
+      "loss": 1.3926,
+      "step": 560
+    },
+    {
+      "epoch": 0.2088631894916871,
+      "grad_norm": 0.1237824410200119,
+      "learning_rate": 1.1220000000000003e-05,
+      "loss": 1.3977,
+      "step": 561
+    },
+    {
+      "epoch": 0.20923549464229618,
+      "grad_norm": 0.12436344474554062,
+      "learning_rate": 1.1240000000000002e-05,
+      "loss": 1.3924,
+      "step": 562
+    },
+    {
+      "epoch": 0.20960779979290525,
+      "grad_norm": 0.12768109142780304,
+      "learning_rate": 1.126e-05,
+      "loss": 1.4062,
+      "step": 563
+    },
+    {
+      "epoch": 0.20998010494351432,
+      "grad_norm": 0.12379388511180878,
+      "learning_rate": 1.128e-05,
+      "loss": 1.3951,
+      "step": 564
+    },
+    {
+      "epoch": 0.2103524100941234,
+      "grad_norm": 0.12182633578777313,
+      "learning_rate": 1.13e-05,
+      "loss": 1.3969,
+      "step": 565
+    },
+    {
+      "epoch": 0.21072471524473246,
+      "grad_norm": 0.12477164715528488,
+      "learning_rate": 1.132e-05,
+      "loss": 1.3954,
+      "step": 566
+    },
+    {
+      "epoch": 0.21109702039534153,
+      "grad_norm": 0.12322055548429489,
+      "learning_rate": 1.134e-05,
+      "loss": 1.4005,
+      "step": 567
+    },
+    {
+      "epoch": 0.2114693255459506,
+      "grad_norm": 0.12597453594207764,
+      "learning_rate": 1.136e-05,
+      "loss": 1.3974,
+      "step": 568
+    },
+    {
+      "epoch": 0.21184163069655967,
+      "grad_norm": 0.13351480662822723,
+      "learning_rate": 1.138e-05,
+      "loss": 1.3914,
+      "step": 569
+    },
+    {
+      "epoch": 0.21221393584716874,
+      "grad_norm": 0.13195490837097168,
+      "learning_rate": 1.14e-05,
+      "loss": 1.4162,
+      "step": 570
+    },
+    {
+      "epoch": 0.21258624099777781,
+      "grad_norm": 0.12694774568080902,
+      "learning_rate": 1.142e-05,
+      "loss": 1.3844,
+      "step": 571
+    },
+    {
+      "epoch": 0.21295854614838688,
+      "grad_norm": 0.12894363701343536,
+      "learning_rate": 1.144e-05,
+      "loss": 1.387,
+      "step": 572
+    },
+    {
+      "epoch": 0.21333085129899593,
+      "grad_norm": 0.11793356388807297,
+      "learning_rate": 1.146e-05,
+      "loss": 1.3881,
+      "step": 573
+    },
+    {
+      "epoch": 0.213703156449605,
+      "grad_norm": 0.13151799142360687,
+      "learning_rate": 1.148e-05,
+      "loss": 1.3872,
+      "step": 574
+    },
+    {
+      "epoch": 0.21407546160021407,
+      "grad_norm": 0.1286250352859497,
+      "learning_rate": 1.15e-05,
+      "loss": 1.3668,
+      "step": 575
+    },
+    {
+      "epoch": 0.21444776675082314,
+      "grad_norm": 0.13224467635154724,
+      "learning_rate": 1.152e-05,
+      "loss": 1.4045,
+      "step": 576
+    },
+    {
+      "epoch": 0.2148200719014322,
+      "grad_norm": 0.12700480222702026,
+      "learning_rate": 1.154e-05,
+      "loss": 1.3814,
+      "step": 577
+    },
+    {
+      "epoch": 0.21519237705204128,
+      "grad_norm": 0.13367070257663727,
+      "learning_rate": 1.156e-05,
+      "loss": 1.3868,
+      "step": 578
+    },
+    {
+      "epoch": 0.21556468220265035,
+      "grad_norm": 0.12190359085798264,
+      "learning_rate": 1.1580000000000001e-05,
+      "loss": 1.3981,
+      "step": 579
+    },
+    {
+      "epoch": 0.21593698735325942,
+      "grad_norm": 0.1240309402346611,
+      "learning_rate": 1.16e-05,
+      "loss": 1.3945,
+      "step": 580
+    },
+    {
+      "epoch": 0.2163092925038685,
+      "grad_norm": 0.15947964787483215,
+      "learning_rate": 1.162e-05,
+      "loss": 1.3889,
+      "step": 581
+    },
+    {
+      "epoch": 0.21668159765447756,
+      "grad_norm": 0.130709707736969,
+      "learning_rate": 1.164e-05,
+      "loss": 1.3987,
+      "step": 582
+    },
+    {
+      "epoch": 0.21705390280508663,
+      "grad_norm": 0.12745317816734314,
+      "learning_rate": 1.166e-05,
+      "loss": 1.3802,
+      "step": 583
+    },
+    {
+      "epoch": 0.2174262079556957,
+      "grad_norm": 0.12209773063659668,
+      "learning_rate": 1.168e-05,
+      "loss": 1.3798,
+      "step": 584
+    },
+    {
+      "epoch": 0.21779851310630474,
+      "grad_norm": 0.14180858433246613,
+      "learning_rate": 1.17e-05,
+      "loss": 1.3814,
+      "step": 585
+    },
+    {
+      "epoch": 0.2181708182569138,
+      "grad_norm": 0.11788824945688248,
+      "learning_rate": 1.172e-05,
+      "loss": 1.3758,
+      "step": 586
+    },
+    {
+      "epoch": 0.21854312340752288,
+      "grad_norm": 0.12975317239761353,
+      "learning_rate": 1.1740000000000001e-05,
+      "loss": 1.395,
+      "step": 587
+    },
+    {
+      "epoch": 0.21891542855813195,
+      "grad_norm": 0.12418822199106216,
+      "learning_rate": 1.1760000000000001e-05,
+      "loss": 1.4044,
+      "step": 588
+    },
+    {
+      "epoch": 0.21928773370874102,
+      "grad_norm": 0.12132619321346283,
+      "learning_rate": 1.178e-05,
+      "loss": 1.3866,
+      "step": 589
+    },
+    {
+      "epoch": 0.2196600388593501,
+      "grad_norm": 0.1325748711824417,
+      "learning_rate": 1.18e-05,
+      "loss": 1.3903,
+      "step": 590
+    },
+    {
+      "epoch": 0.22003234400995916,
+      "grad_norm": 0.1266520470380783,
+      "learning_rate": 1.182e-05,
+      "loss": 1.3894,
+      "step": 591
+    },
+    {
+      "epoch": 0.22040464916056823,
+      "grad_norm": 0.12683559954166412,
+      "learning_rate": 1.184e-05,
+      "loss": 1.3948,
+      "step": 592
+    },
+    {
+      "epoch": 0.2207769543111773,
+      "grad_norm": 0.12880323827266693,
+      "learning_rate": 1.186e-05,
+      "loss": 1.3756,
+      "step": 593
+    },
+    {
+      "epoch": 0.22114925946178637,
+      "grad_norm": 0.12198843061923981,
+      "learning_rate": 1.188e-05,
+      "loss": 1.3821,
+      "step": 594
+    },
+    {
+      "epoch": 0.22152156461239544,
+      "grad_norm": 0.13132554292678833,
+      "learning_rate": 1.1900000000000001e-05,
+      "loss": 1.3877,
+      "step": 595
+    },
+    {
+      "epoch": 0.22189386976300451,
+      "grad_norm": 0.1291092336177826,
+      "learning_rate": 1.1920000000000001e-05,
+      "loss": 1.4057,
+      "step": 596
+    },
+    {
+      "epoch": 0.22226617491361356,
+      "grad_norm": 0.13351795077323914,
+      "learning_rate": 1.1940000000000001e-05,
+      "loss": 1.3652,
+      "step": 597
+    },
+    {
+      "epoch": 0.22263848006422263,
+      "grad_norm": 0.12667742371559143,
+      "learning_rate": 1.196e-05,
+      "loss": 1.386,
+      "step": 598
+    },
+    {
+      "epoch": 0.2230107852148317,
+      "grad_norm": 0.13008779287338257,
+      "learning_rate": 1.198e-05,
+      "loss": 1.3887,
+      "step": 599
+    },
+    {
+      "epoch": 0.22338309036544077,
+      "grad_norm": 0.12079621851444244,
+      "learning_rate": 1.2e-05,
+      "loss": 1.3855,
+      "step": 600
+    },
+    {
+      "epoch": 0.22375539551604984,
+      "grad_norm": 0.13083826005458832,
+      "learning_rate": 1.202e-05,
+      "loss": 1.3819,
+      "step": 601
+    },
+    {
+      "epoch": 0.2241277006666589,
+      "grad_norm": 0.13189451396465302,
+      "learning_rate": 1.204e-05,
+      "loss": 1.3667,
+      "step": 602
+    },
+    {
+      "epoch": 0.22450000581726798,
+      "grad_norm": 0.1339392513036728,
+      "learning_rate": 1.2060000000000001e-05,
+      "loss": 1.3802,
+      "step": 603
+    },
+    {
+      "epoch": 0.22487231096787705,
+      "grad_norm": 0.12915685772895813,
+      "learning_rate": 1.2080000000000001e-05,
+      "loss": 1.3871,
+      "step": 604
+    },
+    {
+      "epoch": 0.22524461611848612,
+      "grad_norm": 0.13761506974697113,
+      "learning_rate": 1.2100000000000001e-05,
+      "loss": 1.392,
+      "step": 605
+    },
+    {
+      "epoch": 0.2256169212690952,
+      "grad_norm": 0.1275719255208969,
+      "learning_rate": 1.2120000000000001e-05,
+      "loss": 1.3715,
+      "step": 606
+    },
+    {
+      "epoch": 0.22598922641970426,
+      "grad_norm": 0.16112855076789856,
+      "learning_rate": 1.214e-05,
+      "loss": 1.3759,
+      "step": 607
+    },
+    {
+      "epoch": 0.22636153157031333,
+      "grad_norm": 0.13108204305171967,
+      "learning_rate": 1.216e-05,
+      "loss": 1.375,
+      "step": 608
+    },
+    {
+      "epoch": 0.2267338367209224,
+      "grad_norm": 0.13481353223323822,
+      "learning_rate": 1.218e-05,
+      "loss": 1.3704,
+      "step": 609
+    },
+    {
+      "epoch": 0.22710614187153144,
+      "grad_norm": 0.144792377948761,
+      "learning_rate": 1.22e-05,
+      "loss": 1.3865,
+      "step": 610
+    },
+    {
+      "epoch": 0.2274784470221405,
+      "grad_norm": 0.13363252580165863,
+      "learning_rate": 1.2220000000000002e-05,
+      "loss": 1.3826,
+      "step": 611
+    },
+    {
+      "epoch": 0.22785075217274958,
+      "grad_norm": 0.14733262360095978,
+      "learning_rate": 1.2240000000000001e-05,
+      "loss": 1.3875,
+      "step": 612
+    },
+    {
+      "epoch": 0.22822305732335865,
+      "grad_norm": 0.13364124298095703,
+      "learning_rate": 1.2260000000000001e-05,
+      "loss": 1.3885,
+      "step": 613
+    },
+    {
+      "epoch": 0.22859536247396772,
+      "grad_norm": 0.13215725123882294,
+      "learning_rate": 1.2280000000000001e-05,
+      "loss": 1.3644,
+      "step": 614
+    },
+    {
+      "epoch": 0.2289676676245768,
+      "grad_norm": 0.14032886922359467,
+      "learning_rate": 1.23e-05,
+      "loss": 1.3931,
+      "step": 615
+    },
+    {
+      "epoch": 0.22933997277518586,
+      "grad_norm": 0.15233881771564484,
+      "learning_rate": 1.232e-05,
+      "loss": 1.3771,
+      "step": 616
+    },
+    {
+      "epoch": 0.22971227792579493,
+      "grad_norm": 0.14122669398784637,
+      "learning_rate": 1.234e-05,
+      "loss": 1.3774,
+      "step": 617
+    },
+    {
+      "epoch": 0.230084583076404,
+      "grad_norm": 0.13818544149398804,
+      "learning_rate": 1.236e-05,
+      "loss": 1.374,
+      "step": 618
+    },
+    {
+      "epoch": 0.23045688822701307,
+      "grad_norm": 0.14192582666873932,
+      "learning_rate": 1.2380000000000002e-05,
+      "loss": 1.3773,
+      "step": 619
+    },
+    {
+      "epoch": 0.23082919337762214,
+      "grad_norm": 0.1523938775062561,
+      "learning_rate": 1.2400000000000002e-05,
+      "loss": 1.3832,
+      "step": 620
+    },
+    {
+      "epoch": 0.23120149852823121,
+      "grad_norm": 0.1431209146976471,
+      "learning_rate": 1.2420000000000001e-05,
+      "loss": 1.3765,
+      "step": 621
+    },
+    {
+      "epoch": 0.23157380367884026,
+      "grad_norm": 0.13690458238124847,
+      "learning_rate": 1.2440000000000001e-05,
+      "loss": 1.3791,
+      "step": 622
+    },
+    {
+      "epoch": 0.23194610882944933,
+      "grad_norm": 0.14077654480934143,
+      "learning_rate": 1.2460000000000001e-05,
+      "loss": 1.3785,
+      "step": 623
+    },
+    {
+      "epoch": 0.2323184139800584,
+      "grad_norm": 0.15121321380138397,
+      "learning_rate": 1.248e-05,
+      "loss": 1.3945,
+      "step": 624
+    },
+    {
+      "epoch": 0.23269071913066747,
+      "grad_norm": 0.1371421217918396,
+      "learning_rate": 1.25e-05,
+      "loss": 1.376,
+      "step": 625
+    },
+    {
+      "epoch": 0.23306302428127654,
+      "grad_norm": 0.14253154397010803,
+      "learning_rate": 1.252e-05,
+      "loss": 1.3823,
+      "step": 626
+    },
+    {
+      "epoch": 0.2334353294318856,
+      "grad_norm": 0.1420959234237671,
+      "learning_rate": 1.254e-05,
+      "loss": 1.3702,
+      "step": 627
+    },
+    {
+      "epoch": 0.23380763458249468,
+      "grad_norm": 0.15104436874389648,
+      "learning_rate": 1.2560000000000002e-05,
+      "loss": 1.385,
+      "step": 628
+    },
+    {
+      "epoch": 0.23417993973310375,
+      "grad_norm": 0.1486346274614334,
+      "learning_rate": 1.2580000000000002e-05,
+      "loss": 1.3792,
+      "step": 629
+    },
+    {
+      "epoch": 0.23455224488371282,
+      "grad_norm": 0.16748331487178802,
+      "learning_rate": 1.2600000000000001e-05,
+      "loss": 1.3916,
+      "step": 630
+    },
+    {
+      "epoch": 0.2349245500343219,
+      "grad_norm": 0.14780151844024658,
+      "learning_rate": 1.2620000000000001e-05,
+      "loss": 1.3827,
+      "step": 631
+    },
+    {
+      "epoch": 0.23529685518493096,
+      "grad_norm": 0.152265265583992,
+      "learning_rate": 1.2640000000000001e-05,
+      "loss": 1.3812,
+      "step": 632
+    },
+    {
+      "epoch": 0.23566916033554003,
+      "grad_norm": 0.1454855501651764,
+      "learning_rate": 1.266e-05,
+      "loss": 1.3907,
+      "step": 633
+    },
+    {
+      "epoch": 0.23604146548614907,
+      "grad_norm": 0.14732711017131805,
+      "learning_rate": 1.268e-05,
+      "loss": 1.3651,
+      "step": 634
+    },
+    {
+      "epoch": 0.23641377063675814,
+      "grad_norm": 0.1476028710603714,
+      "learning_rate": 1.27e-05,
+      "loss": 1.3723,
+      "step": 635
+    },
+    {
+      "epoch": 0.2367860757873672,
+      "grad_norm": 0.14366750419139862,
+      "learning_rate": 1.2720000000000002e-05,
+      "loss": 1.3904,
+      "step": 636
+    },
+    {
+      "epoch": 0.23715838093797628,
+      "grad_norm": 0.15820929408073425,
+      "learning_rate": 1.2740000000000002e-05,
+      "loss": 1.3852,
+      "step": 637
+    },
+    {
+      "epoch": 0.23753068608858535,
+      "grad_norm": 0.1551392823457718,
+      "learning_rate": 1.2760000000000001e-05,
+      "loss": 1.389,
+      "step": 638
+    },
+    {
+      "epoch": 0.23790299123919442,
+      "grad_norm": 0.13778699934482574,
+      "learning_rate": 1.2780000000000001e-05,
+      "loss": 1.3627,
+      "step": 639
+    },
+    {
+      "epoch": 0.2382752963898035,
+      "grad_norm": 0.13657543063163757,
+      "learning_rate": 1.2800000000000001e-05,
+      "loss": 1.3941,
+      "step": 640
+    },
+    {
+      "epoch": 0.23864760154041256,
+      "grad_norm": 0.13666628301143646,
+      "learning_rate": 1.2820000000000001e-05,
+      "loss": 1.3729,
+      "step": 641
+    },
+    {
+      "epoch": 0.23901990669102163,
+      "grad_norm": 0.14898766577243805,
+      "learning_rate": 1.284e-05,
+      "loss": 1.3823,
+      "step": 642
+    },
+    {
+      "epoch": 0.2393922118416307,
+      "grad_norm": 0.1644555926322937,
+      "learning_rate": 1.286e-05,
+      "loss": 1.3794,
+      "step": 643
+    },
+    {
+      "epoch": 0.23976451699223977,
+      "grad_norm": 0.1226273775100708,
+      "learning_rate": 1.2880000000000002e-05,
+      "loss": 1.3631,
+      "step": 644
+    },
+    {
+      "epoch": 0.24013682214284884,
+      "grad_norm": 0.16430190205574036,
+      "learning_rate": 1.2900000000000002e-05,
+      "loss": 1.3936,
+      "step": 645
+    },
+    {
+      "epoch": 0.2405091272934579,
+      "grad_norm": 0.15783290565013885,
+      "learning_rate": 1.2920000000000002e-05,
+      "loss": 1.3951,
+      "step": 646
+    },
+    {
+      "epoch": 0.24088143244406696,
+      "grad_norm": 0.14612354338169098,
+      "learning_rate": 1.2940000000000001e-05,
+      "loss": 1.3701,
+      "step": 647
+    },
+    {
+      "epoch": 0.24125373759467603,
+      "grad_norm": 0.15059353411197662,
+      "learning_rate": 1.2960000000000001e-05,
+      "loss": 1.3649,
+      "step": 648
+    },
+    {
+      "epoch": 0.2416260427452851,
+      "grad_norm": 0.15658129751682281,
+      "learning_rate": 1.2980000000000001e-05,
+      "loss": 1.3793,
+      "step": 649
+    },
+    {
+      "epoch": 0.24199834789589417,
+      "grad_norm": 0.14705069363117218,
+      "learning_rate": 1.3000000000000001e-05,
+      "loss": 1.3816,
+      "step": 650
+    },
+    {
+      "epoch": 0.24237065304650324,
+      "grad_norm": 0.13878506422042847,
+      "learning_rate": 1.302e-05,
+      "loss": 1.3881,
+      "step": 651
+    },
+    {
+      "epoch": 0.2427429581971123,
+      "grad_norm": 0.1551000475883484,
+      "learning_rate": 1.3040000000000002e-05,
+      "loss": 1.3759,
+      "step": 652
+    },
+    {
+      "epoch": 0.24311526334772138,
+      "grad_norm": 0.15696851909160614,
+      "learning_rate": 1.3060000000000002e-05,
+      "loss": 1.3744,
+      "step": 653
+    },
+    {
+      "epoch": 0.24348756849833045,
+      "grad_norm": 0.13662178814411163,
+      "learning_rate": 1.3080000000000002e-05,
+      "loss": 1.3699,
+      "step": 654
+    },
+    {
+      "epoch": 0.24385987364893952,
+      "grad_norm": 0.14424225687980652,
+      "learning_rate": 1.3100000000000002e-05,
+      "loss": 1.371,
+      "step": 655
+    },
+    {
+      "epoch": 0.2442321787995486,
+      "grad_norm": 0.17988137900829315,
+      "learning_rate": 1.3120000000000001e-05,
+      "loss": 1.3955,
+      "step": 656
+    },
+    {
+      "epoch": 0.24460448395015766,
+      "grad_norm": 0.1564609706401825,
+      "learning_rate": 1.3140000000000001e-05,
+      "loss": 1.3792,
+      "step": 657
+    },
+    {
+      "epoch": 0.24497678910076673,
+      "grad_norm": 0.1378934532403946,
+      "learning_rate": 1.3160000000000001e-05,
+      "loss": 1.3501,
+      "step": 658
+    },
+    {
+      "epoch": 0.24534909425137577,
+      "grad_norm": 0.14894217252731323,
+      "learning_rate": 1.3180000000000001e-05,
+      "loss": 1.3716,
+      "step": 659
+    },
+    {
+      "epoch": 0.24572139940198484,
+      "grad_norm": 0.1473313868045807,
+      "learning_rate": 1.3200000000000002e-05,
+      "loss": 1.3666,
+      "step": 660
+    },
+    {
+      "epoch": 0.2460937045525939,
+      "grad_norm": 0.14587174355983734,
+      "learning_rate": 1.3220000000000002e-05,
+      "loss": 1.3887,
+      "step": 661
+    },
+    {
+      "epoch": 0.24646600970320298,
+      "grad_norm": 0.14664749801158905,
+      "learning_rate": 1.3240000000000002e-05,
+      "loss": 1.3727,
+      "step": 662
+    },
+    {
+      "epoch": 0.24683831485381205,
+      "grad_norm": 0.1472824066877365,
+      "learning_rate": 1.3260000000000002e-05,
+      "loss": 1.3676,
+      "step": 663
+    },
+    {
+      "epoch": 0.24721062000442112,
+      "grad_norm": 0.1401338130235672,
+      "learning_rate": 1.3280000000000002e-05,
+      "loss": 1.3873,
+      "step": 664
+    },
+    {
+      "epoch": 0.2475829251550302,
+      "grad_norm": 0.13379812240600586,
+      "learning_rate": 1.3300000000000001e-05,
+      "loss": 1.3625,
+      "step": 665
+    },
+    {
+      "epoch": 0.24795523030563926,
+      "grad_norm": 0.14346960186958313,
+      "learning_rate": 1.3320000000000001e-05,
+      "loss": 1.359,
+      "step": 666
+    },
+    {
+      "epoch": 0.24832753545624833,
+      "grad_norm": 0.14162355661392212,
+      "learning_rate": 1.3340000000000001e-05,
+      "loss": 1.3702,
+      "step": 667
+    },
+    {
+      "epoch": 0.2486998406068574,
+      "grad_norm": 0.15658289194107056,
+      "learning_rate": 1.3360000000000003e-05,
+      "loss": 1.3698,
+      "step": 668
+    },
+    {
+      "epoch": 0.24907214575746647,
+      "grad_norm": 0.13803306221961975,
+      "learning_rate": 1.3380000000000002e-05,
+      "loss": 1.3588,
+      "step": 669
+    },
+    {
+      "epoch": 0.24944445090807554,
+      "grad_norm": 0.14932705461978912,
+      "learning_rate": 1.3400000000000002e-05,
+      "loss": 1.3748,
+      "step": 670
+    },
+    {
+      "epoch": 0.24981675605868459,
+      "grad_norm": 0.3856603503227234,
+      "learning_rate": 1.3420000000000002e-05,
+      "loss": 1.3876,
+      "step": 671
+    },
+    {
+      "epoch": 0.2501890612092937,
+      "grad_norm": 0.14300397038459778,
+      "learning_rate": 1.3440000000000002e-05,
+      "loss": 1.36,
+      "step": 672
+    },
+    {
+      "epoch": 0.2505613663599027,
+      "grad_norm": 0.15117305517196655,
+      "learning_rate": 1.3460000000000002e-05,
+      "loss": 1.3943,
+      "step": 673
+    },
+    {
+      "epoch": 0.2509336715105118,
+      "grad_norm": 0.15544290840625763,
+      "learning_rate": 1.3480000000000001e-05,
+      "loss": 1.3826,
+      "step": 674
+    },
+    {
+      "epoch": 0.25130597666112087,
+      "grad_norm": 0.14352373778820038,
+      "learning_rate": 1.3500000000000001e-05,
+      "loss": 1.3577,
+      "step": 675
+    },
+    {
+      "epoch": 0.25167828181172996,
+      "grad_norm": 0.14757125079631805,
+      "learning_rate": 1.3520000000000003e-05,
+      "loss": 1.3706,
+      "step": 676
+    },
+    {
+      "epoch": 0.252050586962339,
+      "grad_norm": 0.14953893423080444,
+      "learning_rate": 1.3540000000000003e-05,
+      "loss": 1.3656,
+      "step": 677
+    },
+    {
+      "epoch": 0.25242289211294805,
+      "grad_norm": 0.15957669913768768,
+      "learning_rate": 1.3560000000000002e-05,
+      "loss": 1.3569,
+      "step": 678
+    },
+    {
+      "epoch": 0.25279519726355715,
+      "grad_norm": 0.15265138447284698,
+      "learning_rate": 1.3580000000000002e-05,
+      "loss": 1.3636,
+      "step": 679
+    },
+    {
+      "epoch": 0.2531675024141662,
+      "grad_norm": 0.1513211727142334,
+      "learning_rate": 1.3600000000000002e-05,
+      "loss": 1.3717,
+      "step": 680
+    },
+    {
+      "epoch": 0.2535398075647753,
+      "grad_norm": 0.15139970183372498,
+      "learning_rate": 1.3620000000000002e-05,
+      "loss": 1.3776,
+      "step": 681
+    },
+    {
+      "epoch": 0.25391211271538433,
+      "grad_norm": 0.15319296717643738,
+      "learning_rate": 1.3640000000000002e-05,
+      "loss": 1.3674,
+      "step": 682
+    },
+    {
+      "epoch": 0.25428441786599343,
+      "grad_norm": 0.14396759867668152,
+      "learning_rate": 1.3660000000000001e-05,
+      "loss": 1.3739,
+      "step": 683
+    },
+    {
+      "epoch": 0.25465672301660247,
+      "grad_norm": 0.14723047614097595,
+      "learning_rate": 1.3680000000000003e-05,
+      "loss": 1.3777,
+      "step": 684
+    },
+    {
+      "epoch": 0.25502902816721157,
+      "grad_norm": 0.13962876796722412,
+      "learning_rate": 1.3700000000000003e-05,
+      "loss": 1.3676,
+      "step": 685
+    },
+    {
+      "epoch": 0.2554013333178206,
+      "grad_norm": 0.14297540485858917,
+      "learning_rate": 1.3720000000000002e-05,
+      "loss": 1.3725,
+      "step": 686
+    },
+    {
+      "epoch": 0.2557736384684297,
+      "grad_norm": 0.14365188777446747,
+      "learning_rate": 1.3740000000000002e-05,
+      "loss": 1.3547,
+      "step": 687
+    },
+    {
+      "epoch": 0.25614594361903875,
+      "grad_norm": 0.14428015053272247,
+      "learning_rate": 1.376e-05,
+      "loss": 1.3883,
+      "step": 688
+    },
+    {
+      "epoch": 0.25651824876964785,
+      "grad_norm": 0.14234168827533722,
+      "learning_rate": 1.378e-05,
+      "loss": 1.345,
+      "step": 689
+    },
+    {
+      "epoch": 0.2568905539202569,
+      "grad_norm": 0.14778351783752441,
+      "learning_rate": 1.38e-05,
+      "loss": 1.3773,
+      "step": 690
+    },
+    {
+      "epoch": 0.25726285907086593,
+      "grad_norm": 0.14804288744926453,
+      "learning_rate": 1.382e-05,
+      "loss": 1.3877,
+      "step": 691
+    },
+    {
+      "epoch": 0.25763516422147503,
+      "grad_norm": 0.1505001187324524,
+      "learning_rate": 1.384e-05,
+      "loss": 1.3638,
+      "step": 692
+    },
+    {
+      "epoch": 0.2580074693720841,
+      "grad_norm": 0.1481354534626007,
+      "learning_rate": 1.386e-05,
+      "loss": 1.3624,
+      "step": 693
+    },
+    {
+      "epoch": 0.2583797745226932,
+      "grad_norm": 0.1433618813753128,
+      "learning_rate": 1.3880000000000001e-05,
+      "loss": 1.3531,
+      "step": 694
+    },
+    {
+      "epoch": 0.2587520796733022,
+      "grad_norm": 0.14664116501808167,
+      "learning_rate": 1.39e-05,
+      "loss": 1.3705,
+      "step": 695
+    },
+    {
+      "epoch": 0.2591243848239113,
+      "grad_norm": 0.13810987770557404,
+      "learning_rate": 1.392e-05,
+      "loss": 1.3528,
+      "step": 696
+    },
+    {
+      "epoch": 0.25949668997452036,
+      "grad_norm": 0.14705303311347961,
+      "learning_rate": 1.394e-05,
+      "loss": 1.3598,
+      "step": 697
+    },
+    {
+      "epoch": 0.25986899512512945,
+      "grad_norm": 0.14440058171749115,
+      "learning_rate": 1.396e-05,
+      "loss": 1.3584,
+      "step": 698
+    },
+    {
+      "epoch": 0.2602413002757385,
+      "grad_norm": 0.14364704489707947,
+      "learning_rate": 1.398e-05,
+      "loss": 1.3734,
+      "step": 699
+    },
+    {
+      "epoch": 0.2606136054263476,
+      "grad_norm": 0.1593163013458252,
+      "learning_rate": 1.4e-05,
+      "loss": 1.3782,
+      "step": 700
+    },
+    {
+      "epoch": 0.26098591057695664,
+      "grad_norm": 0.15553325414657593,
+      "learning_rate": 1.402e-05,
+      "loss": 1.3768,
+      "step": 701
+    },
+    {
+      "epoch": 0.2613582157275657,
+      "grad_norm": 0.15635879337787628,
+      "learning_rate": 1.4040000000000001e-05,
+      "loss": 1.3541,
+      "step": 702
+    },
+    {
+      "epoch": 0.2617305208781748,
+      "grad_norm": 0.15832242369651794,
+      "learning_rate": 1.4060000000000001e-05,
+      "loss": 1.3801,
+      "step": 703
+    },
+    {
+      "epoch": 0.2621028260287838,
+      "grad_norm": 0.145589679479599,
+      "learning_rate": 1.408e-05,
+      "loss": 1.3701,
+      "step": 704
+    },
+    {
+      "epoch": 0.2624751311793929,
+      "grad_norm": 0.1448379009962082,
+      "learning_rate": 1.41e-05,
+      "loss": 1.3817,
+      "step": 705
+    },
+    {
+      "epoch": 0.26284743633000196,
+      "grad_norm": 0.15079927444458008,
+      "learning_rate": 1.412e-05,
+      "loss": 1.3642,
+      "step": 706
+    },
+    {
+      "epoch": 0.26321974148061106,
+      "grad_norm": 0.15496708452701569,
+      "learning_rate": 1.414e-05,
+      "loss": 1.383,
+      "step": 707
+    },
+    {
+      "epoch": 0.2635920466312201,
+      "grad_norm": 0.14300493896007538,
+      "learning_rate": 1.416e-05,
+      "loss": 1.3664,
+      "step": 708
+    },
+    {
+      "epoch": 0.2639643517818292,
+      "grad_norm": 0.15340474247932434,
+      "learning_rate": 1.418e-05,
+      "loss": 1.371,
+      "step": 709
+    },
+    {
+      "epoch": 0.26433665693243824,
+      "grad_norm": 0.14752480387687683,
+      "learning_rate": 1.4200000000000001e-05,
+      "loss": 1.3754,
+      "step": 710
+    },
+    {
+      "epoch": 0.26470896208304734,
+      "grad_norm": 0.15268085896968842,
+      "learning_rate": 1.4220000000000001e-05,
+      "loss": 1.3604,
+      "step": 711
+    },
+    {
+      "epoch": 0.2650812672336564,
+      "grad_norm": 0.15046533942222595,
+      "learning_rate": 1.4240000000000001e-05,
+      "loss": 1.3566,
+      "step": 712
+    },
+    {
+      "epoch": 0.2654535723842655,
+      "grad_norm": 0.14135941863059998,
+      "learning_rate": 1.426e-05,
+      "loss": 1.3729,
+      "step": 713
+    },
+    {
+      "epoch": 0.2658258775348745,
+      "grad_norm": 0.14512401819229126,
+      "learning_rate": 1.428e-05,
+      "loss": 1.3521,
+      "step": 714
+    },
+    {
+      "epoch": 0.26619818268548356,
+      "grad_norm": 0.15231768786907196,
+      "learning_rate": 1.43e-05,
+      "loss": 1.3648,
+      "step": 715
+    },
+    {
+      "epoch": 0.26657048783609266,
+      "grad_norm": 0.14620406925678253,
+      "learning_rate": 1.432e-05,
+      "loss": 1.3534,
+      "step": 716
+    },
+    {
+      "epoch": 0.2669427929867017,
+      "grad_norm": 0.15307819843292236,
+      "learning_rate": 1.434e-05,
+      "loss": 1.3661,
+      "step": 717
+    },
+    {
+      "epoch": 0.2673150981373108,
+      "grad_norm": 0.15813127160072327,
+      "learning_rate": 1.4360000000000001e-05,
+      "loss": 1.3647,
+      "step": 718
+    },
+    {
+      "epoch": 0.26768740328791985,
+      "grad_norm": 0.15037232637405396,
+      "learning_rate": 1.4380000000000001e-05,
+      "loss": 1.3595,
+      "step": 719
+    },
+    {
+      "epoch": 0.26805970843852894,
+      "grad_norm": 0.1518103927373886,
+      "learning_rate": 1.4400000000000001e-05,
+      "loss": 1.356,
+      "step": 720
+    },
+    {
+      "epoch": 0.268432013589138,
+      "grad_norm": 0.15428012609481812,
+      "learning_rate": 1.4420000000000001e-05,
+      "loss": 1.3636,
+      "step": 721
+    },
+    {
+      "epoch": 0.2688043187397471,
+      "grad_norm": 0.15426598489284515,
+      "learning_rate": 1.444e-05,
+      "loss": 1.355,
+      "step": 722
+    },
+    {
+      "epoch": 0.2691766238903561,
+      "grad_norm": 0.15186642110347748,
+      "learning_rate": 1.446e-05,
+      "loss": 1.3409,
+      "step": 723
+    },
+    {
+      "epoch": 0.2695489290409652,
+      "grad_norm": 0.15585435926914215,
+      "learning_rate": 1.448e-05,
+      "loss": 1.3503,
+      "step": 724
+    },
+    {
+      "epoch": 0.26992123419157427,
+      "grad_norm": 0.1601250320672989,
+      "learning_rate": 1.45e-05,
+      "loss": 1.3641,
+      "step": 725
+    },
+    {
+      "epoch": 0.27029353934218336,
+      "grad_norm": 0.15690714120864868,
+      "learning_rate": 1.4520000000000002e-05,
+      "loss": 1.3681,
+      "step": 726
+    },
+    {
+      "epoch": 0.2706658444927924,
+      "grad_norm": 0.1501239687204361,
+      "learning_rate": 1.4540000000000001e-05,
+      "loss": 1.361,
+      "step": 727
+    },
+    {
+      "epoch": 0.27103814964340145,
+      "grad_norm": 0.15437743067741394,
+      "learning_rate": 1.4560000000000001e-05,
+      "loss": 1.3699,
+      "step": 728
+    },
+    {
+      "epoch": 0.27141045479401055,
+      "grad_norm": 0.14239777624607086,
+      "learning_rate": 1.4580000000000001e-05,
+      "loss": 1.352,
+      "step": 729
+    },
+    {
+      "epoch": 0.2717827599446196,
+      "grad_norm": 0.15204055607318878,
+      "learning_rate": 1.46e-05,
+      "loss": 1.3609,
+      "step": 730
+    },
+    {
+      "epoch": 0.2721550650952287,
+      "grad_norm": 0.14352013170719147,
+      "learning_rate": 1.462e-05,
+      "loss": 1.3316,
+      "step": 731
+    },
+    {
+      "epoch": 0.27252737024583773,
+      "grad_norm": 0.1465366929769516,
+      "learning_rate": 1.464e-05,
+      "loss": 1.3621,
+      "step": 732
+    },
+    {
+      "epoch": 0.27289967539644683,
+      "grad_norm": 0.1439734250307083,
+      "learning_rate": 1.466e-05,
+      "loss": 1.3471,
+      "step": 733
+    },
+    {
+      "epoch": 0.27327198054705587,
+      "grad_norm": 0.16148164868354797,
+      "learning_rate": 1.4680000000000002e-05,
+      "loss": 1.3531,
+      "step": 734
+    },
+    {
+      "epoch": 0.27364428569766497,
+      "grad_norm": 0.15046028792858124,
+      "learning_rate": 1.4700000000000002e-05,
+      "loss": 1.357,
+      "step": 735
+    },
+    {
+      "epoch": 0.274016590848274,
+      "grad_norm": 0.1497603803873062,
+      "learning_rate": 1.4720000000000001e-05,
+      "loss": 1.3609,
+      "step": 736
+    },
+    {
+      "epoch": 0.2743888959988831,
+      "grad_norm": 0.1596868485212326,
+      "learning_rate": 1.4740000000000001e-05,
+      "loss": 1.3569,
+      "step": 737
+    },
+    {
+      "epoch": 0.27476120114949215,
+      "grad_norm": 0.1531444936990738,
+      "learning_rate": 1.4760000000000001e-05,
+      "loss": 1.3436,
+      "step": 738
+    },
+    {
+      "epoch": 0.2751335063001012,
+      "grad_norm": 0.14884676039218903,
+      "learning_rate": 1.478e-05,
+      "loss": 1.3583,
+      "step": 739
+    },
+    {
+      "epoch": 0.2755058114507103,
+      "grad_norm": 0.1622639298439026,
+      "learning_rate": 1.48e-05,
+      "loss": 1.3598,
+      "step": 740
+    },
+    {
+      "epoch": 0.27587811660131933,
+      "grad_norm": 0.16207291185855865,
+      "learning_rate": 1.482e-05,
+      "loss": 1.3771,
+      "step": 741
+    },
+    {
+      "epoch": 0.27625042175192843,
+      "grad_norm": 0.16403760015964508,
+      "learning_rate": 1.4840000000000002e-05,
+      "loss": 1.3618,
+      "step": 742
+    },
+    {
+      "epoch": 0.2766227269025375,
+      "grad_norm": 0.1647598147392273,
+      "learning_rate": 1.4860000000000002e-05,
+      "loss": 1.3542,
+      "step": 743
+    },
+    {
+      "epoch": 0.2769950320531466,
+      "grad_norm": 0.1558753401041031,
+      "learning_rate": 1.4880000000000002e-05,
+      "loss": 1.3527,
+      "step": 744
+    },
+    {
+      "epoch": 0.2773673372037556,
+      "grad_norm": 0.1733637899160385,
+      "learning_rate": 1.4900000000000001e-05,
+      "loss": 1.3538,
+      "step": 745
+    },
+    {
+      "epoch": 0.2777396423543647,
+      "grad_norm": 0.15772801637649536,
+      "learning_rate": 1.4920000000000001e-05,
+      "loss": 1.3492,
+      "step": 746
+    },
+    {
+      "epoch": 0.27811194750497376,
+      "grad_norm": 0.15663520991802216,
+      "learning_rate": 1.4940000000000001e-05,
+      "loss": 1.3565,
+      "step": 747
+    },
+    {
+      "epoch": 0.27848425265558285,
+      "grad_norm": 0.15726646780967712,
+      "learning_rate": 1.496e-05,
+      "loss": 1.3591,
+      "step": 748
+    },
+    {
+      "epoch": 0.2788565578061919,
+      "grad_norm": 0.17913725972175598,
+      "learning_rate": 1.498e-05,
+      "loss": 1.3504,
+      "step": 749
+    },
+    {
+      "epoch": 0.279228862956801,
+      "grad_norm": 0.15114524960517883,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 1.3499,
+      "step": 750
+    },
+    {
+      "epoch": 0.27960116810741004,
+      "grad_norm": 0.15915299952030182,
+      "learning_rate": 1.5020000000000002e-05,
+      "loss": 1.3364,
+      "step": 751
+    },
+    {
+      "epoch": 0.2799734732580191,
+      "grad_norm": 0.15501591563224792,
+      "learning_rate": 1.5040000000000002e-05,
+      "loss": 1.3473,
+      "step": 752
+    },
+    {
+      "epoch": 0.2803457784086282,
+      "grad_norm": 0.1513669490814209,
+      "learning_rate": 1.5060000000000001e-05,
+      "loss": 1.3547,
+      "step": 753
+    },
+    {
+      "epoch": 0.2807180835592372,
+      "grad_norm": 0.16978231072425842,
+      "learning_rate": 1.5080000000000001e-05,
+      "loss": 1.3447,
+      "step": 754
+    },
+    {
+      "epoch": 0.2810903887098463,
+      "grad_norm": 0.16618570685386658,
+      "learning_rate": 1.5100000000000001e-05,
+      "loss": 1.3493,
+      "step": 755
+    },
+    {
+      "epoch": 0.28146269386045536,
+      "grad_norm": 0.15305636823177338,
+      "learning_rate": 1.5120000000000001e-05,
+      "loss": 1.3675,
+      "step": 756
+    },
+    {
+      "epoch": 0.28183499901106446,
+      "grad_norm": 0.16489024460315704,
+      "learning_rate": 1.514e-05,
+      "loss": 1.3465,
+      "step": 757
+    },
+    {
+      "epoch": 0.2822073041616735,
+      "grad_norm": 0.1630878895521164,
+      "learning_rate": 1.516e-05,
+      "loss": 1.3551,
+      "step": 758
+    },
+    {
+      "epoch": 0.2825796093122826,
+      "grad_norm": 0.15557344257831573,
+      "learning_rate": 1.5180000000000002e-05,
+      "loss": 1.3452,
+      "step": 759
+    },
+    {
+      "epoch": 0.28295191446289164,
+      "grad_norm": 0.16549739241600037,
+      "learning_rate": 1.5200000000000002e-05,
+      "loss": 1.3515,
+      "step": 760
+    },
+    {
+      "epoch": 0.28332421961350074,
+      "grad_norm": 0.1618327498435974,
+      "learning_rate": 1.5220000000000002e-05,
+      "loss": 1.3422,
+      "step": 761
+    },
+    {
+      "epoch": 0.2836965247641098,
+      "grad_norm": 0.16406500339508057,
+      "learning_rate": 1.5240000000000001e-05,
+      "loss": 1.354,
+      "step": 762
+    },
+    {
+      "epoch": 0.2840688299147189,
+      "grad_norm": 0.1552378088235855,
+      "learning_rate": 1.5260000000000003e-05,
+      "loss": 1.3644,
+      "step": 763
+    },
+    {
+      "epoch": 0.2844411350653279,
+      "grad_norm": 0.16994218528270721,
+      "learning_rate": 1.5280000000000003e-05,
+      "loss": 1.3639,
+      "step": 764
+    },
+    {
+      "epoch": 0.28481344021593696,
+      "grad_norm": 0.15901988744735718,
+      "learning_rate": 1.5300000000000003e-05,
+      "loss": 1.3657,
+      "step": 765
+    },
+    {
+      "epoch": 0.28518574536654606,
+      "grad_norm": 0.164791077375412,
+      "learning_rate": 1.5320000000000002e-05,
+      "loss": 1.3518,
+      "step": 766
+    },
+    {
+      "epoch": 0.2855580505171551,
+      "grad_norm": 0.17213784158229828,
+      "learning_rate": 1.5340000000000002e-05,
+      "loss": 1.3479,
+      "step": 767
+    },
+    {
+      "epoch": 0.2859303556677642,
+      "grad_norm": 0.15734751522541046,
+      "learning_rate": 1.5360000000000002e-05,
+      "loss": 1.3589,
+      "step": 768
+    },
+    {
+      "epoch": 0.28630266081837324,
+      "grad_norm": 0.17351983487606049,
+      "learning_rate": 1.5380000000000002e-05,
+      "loss": 1.3469,
+      "step": 769
+    },
+    {
+      "epoch": 0.28667496596898234,
+      "grad_norm": 0.1618126779794693,
+      "learning_rate": 1.54e-05,
+      "loss": 1.3508,
+      "step": 770
+    },
+    {
+      "epoch": 0.2870472711195914,
+      "grad_norm": 0.15765798091888428,
+      "learning_rate": 1.542e-05,
+      "loss": 1.3704,
+      "step": 771
+    },
+    {
+      "epoch": 0.2874195762702005,
+      "grad_norm": 0.17241904139518738,
+      "learning_rate": 1.544e-05,
+      "loss": 1.3537,
+      "step": 772
+    },
+    {
+      "epoch": 0.2877918814208095,
+      "grad_norm": 0.15991517901420593,
+      "learning_rate": 1.546e-05,
+      "loss": 1.3434,
+      "step": 773
+    },
+    {
+      "epoch": 0.2881641865714186,
+      "grad_norm": 0.18322332203388214,
+      "learning_rate": 1.548e-05,
+      "loss": 1.3534,
+      "step": 774
+    },
+    {
+      "epoch": 0.28853649172202767,
+      "grad_norm": 0.1581200808286667,
+      "learning_rate": 1.55e-05,
+      "loss": 1.357,
+      "step": 775
+    },
+    {
+      "epoch": 0.2889087968726367,
+      "grad_norm": 0.16509777307510376,
+      "learning_rate": 1.552e-05,
+      "loss": 1.3659,
+      "step": 776
+    },
+    {
+      "epoch": 0.2892811020232458,
+      "grad_norm": 0.1753901094198227,
+      "learning_rate": 1.554e-05,
+      "loss": 1.3456,
+      "step": 777
+    },
+    {
+      "epoch": 0.28965340717385485,
+      "grad_norm": 0.16197755932807922,
+      "learning_rate": 1.556e-05,
+      "loss": 1.3508,
+      "step": 778
+    },
+    {
+      "epoch": 0.29002571232446395,
+      "grad_norm": 0.16059085726737976,
+      "learning_rate": 1.5580000000000003e-05,
+      "loss": 1.343,
+      "step": 779
+    },
+    {
+      "epoch": 0.290398017475073,
+      "grad_norm": 0.170423224568367,
+      "learning_rate": 1.5600000000000003e-05,
+      "loss": 1.3545,
+      "step": 780
+    },
+    {
+      "epoch": 0.2907703226256821,
+      "grad_norm": 0.1572631150484085,
+      "learning_rate": 1.5620000000000003e-05,
+      "loss": 1.3468,
+      "step": 781
+    },
+    {
+      "epoch": 0.29114262777629113,
+      "grad_norm": 0.16300170123577118,
+      "learning_rate": 1.5640000000000003e-05,
+      "loss": 1.3604,
+      "step": 782
+    },
+    {
+      "epoch": 0.2915149329269002,
+      "grad_norm": 0.1643751561641693,
+      "learning_rate": 1.5660000000000003e-05,
+      "loss": 1.3446,
+      "step": 783
+    },
+    {
+      "epoch": 0.29188723807750927,
+      "grad_norm": 0.16648028790950775,
+      "learning_rate": 1.5680000000000002e-05,
+      "loss": 1.3365,
+      "step": 784
+    },
+    {
+      "epoch": 0.29225954322811837,
+      "grad_norm": 0.1662505716085434,
+      "learning_rate": 1.5700000000000002e-05,
+      "loss": 1.3469,
+      "step": 785
+    },
+    {
+      "epoch": 0.2926318483787274,
+      "grad_norm": 0.16111180186271667,
+      "learning_rate": 1.5720000000000002e-05,
+      "loss": 1.3461,
+      "step": 786
+    },
+    {
+      "epoch": 0.2930041535293365,
+      "grad_norm": 0.16680561006069183,
+      "learning_rate": 1.5740000000000002e-05,
+      "loss": 1.3457,
+      "step": 787
+    },
+    {
+      "epoch": 0.29337645867994555,
+      "grad_norm": 0.18159611523151398,
+      "learning_rate": 1.576e-05,
+      "loss": 1.3601,
+      "step": 788
+    },
+    {
+      "epoch": 0.2937487638305546,
+      "grad_norm": 0.1730850487947464,
+      "learning_rate": 1.578e-05,
+      "loss": 1.3629,
+      "step": 789
+    },
+    {
+      "epoch": 0.2941210689811637,
+      "grad_norm": 0.17999385297298431,
+      "learning_rate": 1.58e-05,
+      "loss": 1.3558,
+      "step": 790
+    },
+    {
+      "epoch": 0.29449337413177273,
+      "grad_norm": 0.16653069853782654,
+      "learning_rate": 1.582e-05,
+      "loss": 1.3508,
+      "step": 791
+    },
+    {
+      "epoch": 0.29486567928238183,
+      "grad_norm": 0.16509103775024414,
+      "learning_rate": 1.584e-05,
+      "loss": 1.3411,
+      "step": 792
+    },
+    {
+      "epoch": 0.2952379844329909,
+      "grad_norm": 0.17233438789844513,
+      "learning_rate": 1.586e-05,
+      "loss": 1.3429,
+      "step": 793
+    },
+    {
+      "epoch": 0.29561028958359997,
+      "grad_norm": 0.16929791867733002,
+      "learning_rate": 1.588e-05,
+      "loss": 1.3628,
+      "step": 794
+    },
+    {
+      "epoch": 0.295982594734209,
+      "grad_norm": 0.16331572830677032,
+      "learning_rate": 1.5900000000000004e-05,
+      "loss": 1.346,
+      "step": 795
+    },
+    {
+      "epoch": 0.2963548998848181,
+      "grad_norm": 0.1649729311466217,
+      "learning_rate": 1.5920000000000003e-05,
+      "loss": 1.35,
+      "step": 796
+    },
+    {
+      "epoch": 0.29672720503542716,
+      "grad_norm": 0.16012988984584808,
+      "learning_rate": 1.5940000000000003e-05,
+      "loss": 1.3463,
+      "step": 797
+    },
+    {
+      "epoch": 0.29709951018603625,
+      "grad_norm": 0.16761143505573273,
+      "learning_rate": 1.5960000000000003e-05,
+      "loss": 1.3485,
+      "step": 798
+    },
+    {
+      "epoch": 0.2974718153366453,
+      "grad_norm": 0.17172127962112427,
+      "learning_rate": 1.5980000000000003e-05,
+      "loss": 1.351,
+      "step": 799
+    },
+    {
+      "epoch": 0.2978441204872544,
+      "grad_norm": 0.16175536811351776,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 1.3546,
+      "step": 800
+    },
+    {
+      "epoch": 0.29821642563786344,
+      "grad_norm": 0.17179609835147858,
+      "learning_rate": 1.6020000000000002e-05,
+      "loss": 1.3506,
+      "step": 801
+    },
+    {
+      "epoch": 0.2985887307884725,
+      "grad_norm": 0.1593533307313919,
+      "learning_rate": 1.6040000000000002e-05,
+      "loss": 1.3578,
+      "step": 802
+    },
+    {
+      "epoch": 0.2989610359390816,
+      "grad_norm": 0.17455071210861206,
+      "learning_rate": 1.6060000000000002e-05,
+      "loss": 1.3424,
+      "step": 803
+    },
+    {
+      "epoch": 0.2993333410896906,
+      "grad_norm": 0.1619621366262436,
+      "learning_rate": 1.6080000000000002e-05,
+      "loss": 1.3344,
+      "step": 804
+    },
+    {
+      "epoch": 0.2997056462402997,
+      "grad_norm": 0.1549704223871231,
+      "learning_rate": 1.6100000000000002e-05,
+      "loss": 1.3478,
+      "step": 805
+    },
+    {
+      "epoch": 0.30007795139090876,
+      "grad_norm": 0.16837717592716217,
+      "learning_rate": 1.612e-05,
+      "loss": 1.3585,
+      "step": 806
+    },
+    {
+      "epoch": 0.30045025654151786,
+      "grad_norm": 0.15680178999900818,
+      "learning_rate": 1.614e-05,
+      "loss": 1.3474,
+      "step": 807
+    },
+    {
+      "epoch": 0.3008225616921269,
+      "grad_norm": 0.16553205251693726,
+      "learning_rate": 1.616e-05,
+      "loss": 1.3576,
+      "step": 808
+    },
+    {
+      "epoch": 0.301194866842736,
+      "grad_norm": 0.16092002391815186,
+      "learning_rate": 1.618e-05,
+      "loss": 1.3503,
+      "step": 809
+    },
+    {
+      "epoch": 0.30156717199334504,
+      "grad_norm": 0.16862614452838898,
+      "learning_rate": 1.62e-05,
+      "loss": 1.3556,
+      "step": 810
+    },
+    {
+      "epoch": 0.30193947714395414,
+      "grad_norm": 0.16667672991752625,
+      "learning_rate": 1.6220000000000004e-05,
+      "loss": 1.3602,
+      "step": 811
+    },
+    {
+      "epoch": 0.3023117822945632,
+      "grad_norm": 0.16125406324863434,
+      "learning_rate": 1.6240000000000004e-05,
+      "loss": 1.3445,
+      "step": 812
+    },
+    {
+      "epoch": 0.3026840874451722,
+      "grad_norm": 0.16460643708705902,
+      "learning_rate": 1.626e-05,
+      "loss": 1.3546,
+      "step": 813
+    },
+    {
+      "epoch": 0.3030563925957813,
+      "grad_norm": 0.15730057656764984,
+      "learning_rate": 1.628e-05,
+      "loss": 1.3461,
+      "step": 814
+    },
+    {
+      "epoch": 0.30342869774639036,
+      "grad_norm": 0.16555914282798767,
+      "learning_rate": 1.63e-05,
+      "loss": 1.3506,
+      "step": 815
+    },
+    {
+      "epoch": 0.30380100289699946,
+      "grad_norm": 0.15877728164196014,
+      "learning_rate": 1.632e-05,
+      "loss": 1.3536,
+      "step": 816
+    },
+    {
+      "epoch": 0.3041733080476085,
+      "grad_norm": 0.1610872894525528,
+      "learning_rate": 1.634e-05,
+      "loss": 1.3461,
+      "step": 817
+    },
+    {
+      "epoch": 0.3045456131982176,
+      "grad_norm": 0.17202696204185486,
+      "learning_rate": 1.636e-05,
+      "loss": 1.3405,
+      "step": 818
+    },
+    {
+      "epoch": 0.30491791834882664,
+      "grad_norm": 0.16373762488365173,
+      "learning_rate": 1.638e-05,
+      "loss": 1.3514,
+      "step": 819
+    },
+    {
+      "epoch": 0.30529022349943574,
+      "grad_norm": 0.17125077545642853,
+      "learning_rate": 1.64e-05,
+      "loss": 1.3451,
+      "step": 820
+    },
+    {
+      "epoch": 0.3056625286500448,
+      "grad_norm": 0.18695010244846344,
+      "learning_rate": 1.6420000000000002e-05,
+      "loss": 1.3443,
+      "step": 821
+    },
+    {
+      "epoch": 0.3060348338006539,
+      "grad_norm": 0.1650739312171936,
+      "learning_rate": 1.6440000000000002e-05,
+      "loss": 1.3512,
+      "step": 822
+    },
+    {
+      "epoch": 0.3064071389512629,
+      "grad_norm": 0.18323633074760437,
+      "learning_rate": 1.646e-05,
+      "loss": 1.3426,
+      "step": 823
+    },
+    {
+      "epoch": 0.306779444101872,
+      "grad_norm": 0.1654973030090332,
+      "learning_rate": 1.648e-05,
+      "loss": 1.3452,
+      "step": 824
+    },
+    {
+      "epoch": 0.30715174925248107,
+      "grad_norm": 0.1673707365989685,
+      "learning_rate": 1.65e-05,
+      "loss": 1.3493,
+      "step": 825
+    },
+    {
+      "epoch": 0.3075240544030901,
+      "grad_norm": 0.16768096387386322,
+      "learning_rate": 1.652e-05,
+      "loss": 1.3484,
+      "step": 826
+    },
+    {
+      "epoch": 0.3078963595536992,
+      "grad_norm": 0.15851683914661407,
+      "learning_rate": 1.654e-05,
+      "loss": 1.3481,
+      "step": 827
+    },
+    {
+      "epoch": 0.30826866470430825,
+      "grad_norm": 0.1723686009645462,
+      "learning_rate": 1.656e-05,
+      "loss": 1.3566,
+      "step": 828
+    },
+    {
+      "epoch": 0.30864096985491735,
+      "grad_norm": 0.16892585158348083,
+      "learning_rate": 1.658e-05,
+      "loss": 1.3563,
+      "step": 829
+    },
+    {
+      "epoch": 0.3090132750055264,
+      "grad_norm": 0.16549967229366302,
+      "learning_rate": 1.66e-05,
+      "loss": 1.3514,
+      "step": 830
+    },
+    {
+      "epoch": 0.3093855801561355,
+      "grad_norm": 0.16105307638645172,
+      "learning_rate": 1.662e-05,
+      "loss": 1.3528,
+      "step": 831
+    },
+    {
+      "epoch": 0.30975788530674453,
+      "grad_norm": 0.1725122034549713,
+      "learning_rate": 1.664e-05,
+      "loss": 1.3408,
+      "step": 832
+    },
+    {
+      "epoch": 0.3101301904573536,
+      "grad_norm": 0.16419732570648193,
+      "learning_rate": 1.666e-05,
+      "loss": 1.3422,
+      "step": 833
+    },
+    {
+      "epoch": 0.31050249560796267,
+      "grad_norm": 0.16626691818237305,
+      "learning_rate": 1.668e-05,
+      "loss": 1.332,
+      "step": 834
+    },
+    {
+      "epoch": 0.31087480075857177,
+      "grad_norm": 0.16400444507598877,
+      "learning_rate": 1.67e-05,
+      "loss": 1.3378,
+      "step": 835
+    },
+    {
+      "epoch": 0.3112471059091808,
+      "grad_norm": 0.1672375500202179,
+      "learning_rate": 1.672e-05,
+      "loss": 1.334,
+      "step": 836
+    },
+    {
+      "epoch": 0.31161941105978985,
+      "grad_norm": 0.16108368337154388,
+      "learning_rate": 1.6740000000000002e-05,
+      "loss": 1.3525,
+      "step": 837
+    },
+    {
+      "epoch": 0.31199171621039895,
+      "grad_norm": 0.16484303772449493,
+      "learning_rate": 1.6760000000000002e-05,
+      "loss": 1.3541,
+      "step": 838
+    },
+    {
+      "epoch": 0.312364021361008,
+      "grad_norm": 0.16695603728294373,
+      "learning_rate": 1.6780000000000002e-05,
+      "loss": 1.3298,
+      "step": 839
+    },
+    {
+      "epoch": 0.3127363265116171,
+      "grad_norm": 0.1630071997642517,
+      "learning_rate": 1.6800000000000002e-05,
+      "loss": 1.3388,
+      "step": 840
+    },
+    {
+      "epoch": 0.31310863166222613,
+      "grad_norm": 0.17112383246421814,
+      "learning_rate": 1.682e-05,
+      "loss": 1.3464,
+      "step": 841
+    },
+    {
+      "epoch": 0.31348093681283523,
+      "grad_norm": 0.1732318103313446,
+      "learning_rate": 1.684e-05,
+      "loss": 1.3344,
+      "step": 842
+    },
+    {
+      "epoch": 0.3138532419634443,
+      "grad_norm": 0.17030829191207886,
+      "learning_rate": 1.686e-05,
+      "loss": 1.3411,
+      "step": 843
+    },
+    {
+      "epoch": 0.31422554711405337,
+      "grad_norm": 0.16124574840068817,
+      "learning_rate": 1.688e-05,
+      "loss": 1.3473,
+      "step": 844
+    },
+    {
+      "epoch": 0.3145978522646624,
+      "grad_norm": 0.17811204493045807,
+      "learning_rate": 1.69e-05,
+      "loss": 1.3522,
+      "step": 845
+    },
+    {
+      "epoch": 0.3149701574152715,
+      "grad_norm": 0.17002224922180176,
+      "learning_rate": 1.692e-05,
+      "loss": 1.3511,
+      "step": 846
+    },
+    {
+      "epoch": 0.31534246256588055,
+      "grad_norm": 0.19199255108833313,
+      "learning_rate": 1.694e-05,
+      "loss": 1.3263,
+      "step": 847
+    },
+    {
+      "epoch": 0.31571476771648965,
+      "grad_norm": 0.1713089942932129,
+      "learning_rate": 1.696e-05,
+      "loss": 1.3383,
+      "step": 848
+    },
+    {
+      "epoch": 0.3160870728670987,
+      "grad_norm": 0.17415201663970947,
+      "learning_rate": 1.698e-05,
+      "loss": 1.3336,
+      "step": 849
+    },
+    {
+      "epoch": 0.31645937801770774,
+      "grad_norm": 0.1600249856710434,
+      "learning_rate": 1.7e-05,
+      "loss": 1.3639,
+      "step": 850
+    },
+    {
+      "epoch": 0.31683168316831684,
+      "grad_norm": 0.17309750616550446,
+      "learning_rate": 1.702e-05,
+      "loss": 1.3371,
+      "step": 851
+    },
+    {
+      "epoch": 0.3172039883189259,
+      "grad_norm": 0.16656599938869476,
+      "learning_rate": 1.704e-05,
+      "loss": 1.3367,
+      "step": 852
+    },
+    {
+      "epoch": 0.317576293469535,
+      "grad_norm": 0.17623035609722137,
+      "learning_rate": 1.7060000000000003e-05,
+      "loss": 1.3414,
+      "step": 853
+    },
+    {
+      "epoch": 0.317948598620144,
+      "grad_norm": 0.16454781591892242,
+      "learning_rate": 1.7080000000000002e-05,
+      "loss": 1.3341,
+      "step": 854
+    },
+    {
+      "epoch": 0.3183209037707531,
+      "grad_norm": 0.16562296450138092,
+      "learning_rate": 1.7100000000000002e-05,
+      "loss": 1.3321,
+      "step": 855
+    },
+    {
+      "epoch": 0.31869320892136216,
+      "grad_norm": 0.16431719064712524,
+      "learning_rate": 1.7120000000000002e-05,
+      "loss": 1.3332,
+      "step": 856
+    },
+    {
+      "epoch": 0.31906551407197126,
+      "grad_norm": 0.16330453753471375,
+      "learning_rate": 1.7140000000000002e-05,
+      "loss": 1.3376,
+      "step": 857
+    },
+    {
+      "epoch": 0.3194378192225803,
+      "grad_norm": 0.1680067926645279,
+      "learning_rate": 1.7160000000000002e-05,
+      "loss": 1.3343,
+      "step": 858
+    },
+    {
+      "epoch": 0.3198101243731894,
+      "grad_norm": 0.1762932538986206,
+      "learning_rate": 1.718e-05,
+      "loss": 1.3387,
+      "step": 859
+    },
+    {
+      "epoch": 0.32018242952379844,
+      "grad_norm": 0.1647929847240448,
+      "learning_rate": 1.72e-05,
+      "loss": 1.3451,
+      "step": 860
+    },
+    {
+      "epoch": 0.32055473467440754,
+      "grad_norm": 0.16490478813648224,
+      "learning_rate": 1.722e-05,
+      "loss": 1.3486,
+      "step": 861
+    },
+    {
+      "epoch": 0.3209270398250166,
+      "grad_norm": 0.17381954193115234,
+      "learning_rate": 1.724e-05,
+      "loss": 1.3598,
+      "step": 862
+    },
+    {
+      "epoch": 0.3212993449756256,
+      "grad_norm": 0.17484751343727112,
+      "learning_rate": 1.726e-05,
+      "loss": 1.3432,
+      "step": 863
+    },
+    {
+      "epoch": 0.3216716501262347,
+      "grad_norm": 0.17640666663646698,
+      "learning_rate": 1.728e-05,
+      "loss": 1.3512,
+      "step": 864
+    },
+    {
+      "epoch": 0.32204395527684376,
+      "grad_norm": 0.17663167417049408,
+      "learning_rate": 1.73e-05,
+      "loss": 1.3646,
+      "step": 865
+    },
+    {
+      "epoch": 0.32241626042745286,
+      "grad_norm": 0.1713891476392746,
+      "learning_rate": 1.732e-05,
+      "loss": 1.3256,
+      "step": 866
+    },
+    {
+      "epoch": 0.3227885655780619,
+      "grad_norm": 0.17021964490413666,
+      "learning_rate": 1.734e-05,
+      "loss": 1.332,
+      "step": 867
+    },
+    {
+      "epoch": 0.323160870728671,
+      "grad_norm": 0.16920128464698792,
+      "learning_rate": 1.736e-05,
+      "loss": 1.3463,
+      "step": 868
+    },
+    {
+      "epoch": 0.32353317587928004,
+      "grad_norm": 0.18202915787696838,
+      "learning_rate": 1.7380000000000003e-05,
+      "loss": 1.3461,
+      "step": 869
+    },
+    {
+      "epoch": 0.32390548102988914,
+      "grad_norm": 0.16767475008964539,
+      "learning_rate": 1.7400000000000003e-05,
+      "loss": 1.33,
+      "step": 870
+    },
+    {
+      "epoch": 0.3242777861804982,
+      "grad_norm": 0.18158170580863953,
+      "learning_rate": 1.7420000000000003e-05,
+      "loss": 1.3504,
+      "step": 871
+    },
+    {
+      "epoch": 0.3246500913311073,
+      "grad_norm": 0.17010337114334106,
+      "learning_rate": 1.7440000000000002e-05,
+      "loss": 1.3229,
+      "step": 872
+    },
+    {
+      "epoch": 0.3250223964817163,
+      "grad_norm": 0.1630079299211502,
+      "learning_rate": 1.7460000000000002e-05,
+      "loss": 1.3142,
+      "step": 873
+    },
+    {
+      "epoch": 0.32539470163232537,
+      "grad_norm": 0.176998108625412,
+      "learning_rate": 1.7480000000000002e-05,
+      "loss": 1.3371,
+      "step": 874
+    },
+    {
+      "epoch": 0.32576700678293447,
+      "grad_norm": 0.18017543852329254,
+      "learning_rate": 1.7500000000000002e-05,
+      "loss": 1.3562,
+      "step": 875
+    },
+    {
+      "epoch": 0.3261393119335435,
+      "grad_norm": 0.17135824263095856,
+      "learning_rate": 1.752e-05,
+      "loss": 1.3387,
+      "step": 876
+    },
+    {
+      "epoch": 0.3265116170841526,
+      "grad_norm": 0.17937518656253815,
+      "learning_rate": 1.754e-05,
+      "loss": 1.3389,
+      "step": 877
+    },
+    {
+      "epoch": 0.32688392223476165,
+      "grad_norm": 0.1736384481191635,
+      "learning_rate": 1.756e-05,
+      "loss": 1.3556,
+      "step": 878
+    },
+    {
+      "epoch": 0.32725622738537075,
+      "grad_norm": 0.162176251411438,
+      "learning_rate": 1.758e-05,
+      "loss": 1.3499,
+      "step": 879
+    },
+    {
+      "epoch": 0.3276285325359798,
+      "grad_norm": 0.17646604776382446,
+      "learning_rate": 1.76e-05,
+      "loss": 1.3404,
+      "step": 880
+    },
+    {
+      "epoch": 0.3280008376865889,
+      "grad_norm": 0.18382136523723602,
+      "learning_rate": 1.762e-05,
+      "loss": 1.3429,
+      "step": 881
+    },
+    {
+      "epoch": 0.32837314283719793,
+      "grad_norm": 0.19566792249679565,
+      "learning_rate": 1.764e-05,
+      "loss": 1.3399,
+      "step": 882
+    },
+    {
+      "epoch": 0.328745447987807,
+      "grad_norm": 0.17790797352790833,
+      "learning_rate": 1.766e-05,
+      "loss": 1.3413,
+      "step": 883
+    },
+    {
+      "epoch": 0.32911775313841607,
+      "grad_norm": 0.16908185184001923,
+      "learning_rate": 1.768e-05,
+      "loss": 1.3363,
+      "step": 884
+    },
+    {
+      "epoch": 0.32949005828902517,
+      "grad_norm": 0.17043371498584747,
+      "learning_rate": 1.77e-05,
+      "loss": 1.3415,
+      "step": 885
+    },
+    {
+      "epoch": 0.3298623634396342,
+      "grad_norm": 0.18717624247074127,
+      "learning_rate": 1.7720000000000003e-05,
+      "loss": 1.3382,
+      "step": 886
+    },
+    {
+      "epoch": 0.33023466859024325,
+      "grad_norm": 0.17335492372512817,
+      "learning_rate": 1.7740000000000003e-05,
+      "loss": 1.3391,
+      "step": 887
+    },
+    {
+      "epoch": 0.33060697374085235,
+      "grad_norm": 0.1830221712589264,
+      "learning_rate": 1.7760000000000003e-05,
+      "loss": 1.3435,
+      "step": 888
+    },
+    {
+      "epoch": 0.3309792788914614,
+      "grad_norm": 0.17462217807769775,
+      "learning_rate": 1.7780000000000003e-05,
+      "loss": 1.3322,
+      "step": 889
+    },
+    {
+      "epoch": 0.3313515840420705,
+      "grad_norm": 0.17677493393421173,
+      "learning_rate": 1.7800000000000002e-05,
+      "loss": 1.3234,
+      "step": 890
+    },
+    {
+      "epoch": 0.33172388919267953,
+      "grad_norm": 0.18080869317054749,
+      "learning_rate": 1.7820000000000002e-05,
+      "loss": 1.3442,
+      "step": 891
+    },
+    {
+      "epoch": 0.33209619434328863,
+      "grad_norm": 0.1581638604402542,
+      "learning_rate": 1.7840000000000002e-05,
+      "loss": 1.3198,
+      "step": 892
+    },
+    {
+      "epoch": 0.3324684994938977,
+      "grad_norm": 0.18022367358207703,
+      "learning_rate": 1.7860000000000002e-05,
+      "loss": 1.3182,
+      "step": 893
+    },
+    {
+      "epoch": 0.33284080464450677,
+      "grad_norm": 0.19160951673984528,
+      "learning_rate": 1.788e-05,
+      "loss": 1.3365,
+      "step": 894
+    },
+    {
+      "epoch": 0.3332131097951158,
+      "grad_norm": 0.18681097030639648,
+      "learning_rate": 1.79e-05,
+      "loss": 1.337,
+      "step": 895
+    },
+    {
+      "epoch": 0.3335854149457249,
+      "grad_norm": 0.17079736292362213,
+      "learning_rate": 1.792e-05,
+      "loss": 1.3371,
+      "step": 896
+    },
+    {
+      "epoch": 0.33395772009633395,
+      "grad_norm": 0.17504505813121796,
+      "learning_rate": 1.794e-05,
+      "loss": 1.3492,
+      "step": 897
+    },
+    {
+      "epoch": 0.33433002524694305,
+      "grad_norm": 0.18182472884655,
+      "learning_rate": 1.796e-05,
+      "loss": 1.3276,
+      "step": 898
+    },
+    {
+      "epoch": 0.3347023303975521,
+      "grad_norm": 0.1663830280303955,
+      "learning_rate": 1.798e-05,
+      "loss": 1.3366,
+      "step": 899
+    },
+    {
+      "epoch": 0.33507463554816114,
+      "grad_norm": 0.18775950372219086,
+      "learning_rate": 1.8e-05,
+      "loss": 1.3478,
+      "step": 900
+    },
+    {
+      "epoch": 0.33544694069877024,
+      "grad_norm": 0.16797006130218506,
+      "learning_rate": 1.802e-05,
+      "loss": 1.3456,
+      "step": 901
+    },
+    {
+      "epoch": 0.3358192458493793,
+      "grad_norm": 0.16422320902347565,
+      "learning_rate": 1.8040000000000003e-05,
+      "loss": 1.327,
+      "step": 902
+    },
+    {
+      "epoch": 0.3361915509999884,
+      "grad_norm": 0.1987656205892563,
+      "learning_rate": 1.8060000000000003e-05,
+      "loss": 1.3367,
+      "step": 903
+    },
+    {
+      "epoch": 0.3365638561505974,
+      "grad_norm": 0.16781911253929138,
+      "learning_rate": 1.8080000000000003e-05,
+      "loss": 1.3366,
+      "step": 904
+    },
+    {
+      "epoch": 0.3369361613012065,
+      "grad_norm": 0.17309685051441193,
+      "learning_rate": 1.8100000000000003e-05,
+      "loss": 1.3398,
+      "step": 905
+    },
+    {
+      "epoch": 0.33730846645181556,
+      "grad_norm": 0.17799708247184753,
+      "learning_rate": 1.8120000000000003e-05,
+      "loss": 1.3504,
+      "step": 906
+    },
+    {
+      "epoch": 0.33768077160242466,
+      "grad_norm": 0.16965223848819733,
+      "learning_rate": 1.8140000000000003e-05,
+      "loss": 1.3352,
+      "step": 907
+    },
+    {
+      "epoch": 0.3380530767530337,
+      "grad_norm": 0.1705309897661209,
+      "learning_rate": 1.8160000000000002e-05,
+      "loss": 1.3119,
+      "step": 908
+    },
+    {
+      "epoch": 0.3384253819036428,
+      "grad_norm": 0.18711362779140472,
+      "learning_rate": 1.8180000000000002e-05,
+      "loss": 1.3077,
+      "step": 909
+    },
+    {
+      "epoch": 0.33879768705425184,
+      "grad_norm": 0.1738344430923462,
+      "learning_rate": 1.8200000000000002e-05,
+      "loss": 1.3292,
+      "step": 910
+    },
+    {
+      "epoch": 0.3391699922048609,
+      "grad_norm": 0.17799516022205353,
+      "learning_rate": 1.8220000000000002e-05,
+      "loss": 1.3294,
+      "step": 911
+    },
+    {
+      "epoch": 0.33954229735547,
+      "grad_norm": 0.18020783364772797,
+      "learning_rate": 1.824e-05,
+      "loss": 1.3383,
+      "step": 912
+    },
+    {
+      "epoch": 0.339914602506079,
+      "grad_norm": 0.16939756274223328,
+      "learning_rate": 1.826e-05,
+      "loss": 1.3291,
+      "step": 913
+    },
+    {
+      "epoch": 0.3402869076566881,
+      "grad_norm": 0.17973625659942627,
+      "learning_rate": 1.828e-05,
+      "loss": 1.3315,
+      "step": 914
+    },
+    {
+      "epoch": 0.34065921280729716,
+      "grad_norm": 0.1785053163766861,
+      "learning_rate": 1.83e-05,
+      "loss": 1.3386,
+      "step": 915
+    },
+    {
+      "epoch": 0.34103151795790626,
+      "grad_norm": 0.17944440245628357,
+      "learning_rate": 1.832e-05,
+      "loss": 1.3356,
+      "step": 916
+    },
+    {
+      "epoch": 0.3414038231085153,
+      "grad_norm": 0.1690181940793991,
+      "learning_rate": 1.834e-05,
+      "loss": 1.3428,
+      "step": 917
+    },
+    {
+      "epoch": 0.3417761282591244,
+      "grad_norm": 0.16925571858882904,
+      "learning_rate": 1.8360000000000004e-05,
+      "loss": 1.335,
+      "step": 918
+    },
+    {
+      "epoch": 0.34214843340973344,
+      "grad_norm": 0.17593559622764587,
+      "learning_rate": 1.8380000000000004e-05,
+      "loss": 1.3326,
+      "step": 919
+    },
+    {
+      "epoch": 0.34252073856034254,
+      "grad_norm": 0.1841055005788803,
+      "learning_rate": 1.8400000000000003e-05,
+      "loss": 1.3245,
+      "step": 920
+    },
+    {
+      "epoch": 0.3428930437109516,
+      "grad_norm": 0.18622370064258575,
+      "learning_rate": 1.8420000000000003e-05,
+      "loss": 1.3273,
+      "step": 921
+    },
+    {
+      "epoch": 0.3432653488615607,
+      "grad_norm": 0.17079919576644897,
+      "learning_rate": 1.8440000000000003e-05,
+      "loss": 1.3233,
+      "step": 922
+    },
+    {
+      "epoch": 0.3436376540121697,
+      "grad_norm": 0.17922057211399078,
+      "learning_rate": 1.8460000000000003e-05,
+      "loss": 1.3385,
+      "step": 923
+    },
+    {
+      "epoch": 0.34400995916277877,
+      "grad_norm": 0.1806870400905609,
+      "learning_rate": 1.8480000000000003e-05,
+      "loss": 1.3414,
+      "step": 924
+    },
+    {
+      "epoch": 0.34438226431338786,
+      "grad_norm": 0.1918867528438568,
+      "learning_rate": 1.8500000000000002e-05,
+      "loss": 1.3441,
+      "step": 925
+    },
+    {
+      "epoch": 0.3447545694639969,
+      "grad_norm": 0.18761901557445526,
+      "learning_rate": 1.8520000000000002e-05,
+      "loss": 1.3352,
+      "step": 926
+    },
+    {
+      "epoch": 0.345126874614606,
+      "grad_norm": 0.17143699526786804,
+      "learning_rate": 1.8540000000000002e-05,
+      "loss": 1.3225,
+      "step": 927
+    },
+    {
+      "epoch": 0.34549917976521505,
+      "grad_norm": 0.1826808899641037,
+      "learning_rate": 1.8560000000000002e-05,
+      "loss": 1.338,
+      "step": 928
+    },
+    {
+      "epoch": 0.34587148491582415,
+      "grad_norm": 0.1739514321088791,
+      "learning_rate": 1.858e-05,
+      "loss": 1.3572,
+      "step": 929
+    },
+    {
+      "epoch": 0.3462437900664332,
+      "grad_norm": 0.18177390098571777,
+      "learning_rate": 1.86e-05,
+      "loss": 1.3125,
+      "step": 930
+    },
+    {
+      "epoch": 0.3466160952170423,
+      "grad_norm": 0.17709952592849731,
+      "learning_rate": 1.862e-05,
+      "loss": 1.326,
+      "step": 931
+    },
+    {
+      "epoch": 0.34698840036765133,
+      "grad_norm": 0.1684853583574295,
+      "learning_rate": 1.864e-05,
+      "loss": 1.3176,
+      "step": 932
+    },
+    {
+      "epoch": 0.3473607055182604,
+      "grad_norm": 0.18772630393505096,
+      "learning_rate": 1.866e-05,
+      "loss": 1.3246,
+      "step": 933
+    },
+    {
+      "epoch": 0.34773301066886947,
+      "grad_norm": 0.1772524118423462,
+      "learning_rate": 1.8680000000000004e-05,
+      "loss": 1.3463,
+      "step": 934
+    },
+    {
+      "epoch": 0.34810531581947857,
+      "grad_norm": 0.1651849001646042,
+      "learning_rate": 1.8700000000000004e-05,
+      "loss": 1.3414,
+      "step": 935
+    },
+    {
+      "epoch": 0.3484776209700876,
+      "grad_norm": 0.19098713994026184,
+      "learning_rate": 1.8720000000000004e-05,
+      "loss": 1.3423,
+      "step": 936
+    },
+    {
+      "epoch": 0.34884992612069665,
+      "grad_norm": 0.18180294334888458,
+      "learning_rate": 1.8740000000000004e-05,
+      "loss": 1.324,
+      "step": 937
+    },
+    {
+      "epoch": 0.34922223127130575,
+      "grad_norm": 0.1721077859401703,
+      "learning_rate": 1.876e-05,
+      "loss": 1.3337,
+      "step": 938
+    },
+    {
+      "epoch": 0.3495945364219148,
+      "grad_norm": 0.1962614804506302,
+      "learning_rate": 1.878e-05,
+      "loss": 1.3268,
+      "step": 939
+    },
+    {
+      "epoch": 0.3499668415725239,
+      "grad_norm": 0.17538578808307648,
+      "learning_rate": 1.88e-05,
+      "loss": 1.3238,
+      "step": 940
+    },
+    {
+      "epoch": 0.35033914672313293,
+      "grad_norm": 0.17809061706066132,
+      "learning_rate": 1.882e-05,
+      "loss": 1.3135,
+      "step": 941
+    },
+    {
+      "epoch": 0.35071145187374203,
+      "grad_norm": 0.182732954621315,
+      "learning_rate": 1.884e-05,
+      "loss": 1.328,
+      "step": 942
+    },
+    {
+      "epoch": 0.3510837570243511,
+      "grad_norm": 0.1797078251838684,
+      "learning_rate": 1.886e-05,
+      "loss": 1.3315,
+      "step": 943
+    },
+    {
+      "epoch": 0.35145606217496017,
+      "grad_norm": 0.187151238322258,
+      "learning_rate": 1.8880000000000002e-05,
+      "loss": 1.3277,
+      "step": 944
+    },
+    {
+      "epoch": 0.3518283673255692,
+      "grad_norm": 0.16728267073631287,
+      "learning_rate": 1.8900000000000002e-05,
+      "loss": 1.3369,
+      "step": 945
+    },
+    {
+      "epoch": 0.3522006724761783,
+      "grad_norm": 0.1842220276594162,
+      "learning_rate": 1.8920000000000002e-05,
+      "loss": 1.3236,
+      "step": 946
+    },
+    {
+      "epoch": 0.35257297762678735,
+      "grad_norm": 0.18164734542369843,
+      "learning_rate": 1.894e-05,
+      "loss": 1.3348,
+      "step": 947
+    },
+    {
+      "epoch": 0.3529452827773964,
+      "grad_norm": 0.17997892200946808,
+      "learning_rate": 1.896e-05,
+      "loss": 1.3231,
+      "step": 948
+    },
+    {
+      "epoch": 0.3533175879280055,
+      "grad_norm": 0.19494079053401947,
+      "learning_rate": 1.898e-05,
+      "loss": 1.3258,
+      "step": 949
+    },
+    {
+      "epoch": 0.35368989307861454,
+      "grad_norm": 0.17961707711219788,
+      "learning_rate": 1.9e-05,
+      "loss": 1.3301,
+      "step": 950
+    },
+    {
+      "epoch": 0.35406219822922363,
+      "grad_norm": 0.17583470046520233,
+      "learning_rate": 1.902e-05,
+      "loss": 1.3331,
+      "step": 951
+    },
+    {
+      "epoch": 0.3544345033798327,
+      "grad_norm": 0.18307524919509888,
+      "learning_rate": 1.904e-05,
+      "loss": 1.3396,
+      "step": 952
+    },
+    {
+      "epoch": 0.3548068085304418,
+      "grad_norm": 0.1822507530450821,
+      "learning_rate": 1.906e-05,
+      "loss": 1.3372,
+      "step": 953
+    },
+    {
+      "epoch": 0.3551791136810508,
+      "grad_norm": 0.18464961647987366,
+      "learning_rate": 1.908e-05,
+      "loss": 1.3201,
+      "step": 954
+    },
+    {
+      "epoch": 0.3555514188316599,
+      "grad_norm": 0.1767146736383438,
+      "learning_rate": 1.91e-05,
+      "loss": 1.3194,
+      "step": 955
+    },
+    {
+      "epoch": 0.35592372398226896,
+      "grad_norm": 0.19084005057811737,
+      "learning_rate": 1.912e-05,
+      "loss": 1.3535,
+      "step": 956
+    },
+    {
+      "epoch": 0.35629602913287806,
+      "grad_norm": 0.19618546962738037,
+      "learning_rate": 1.914e-05,
+      "loss": 1.3393,
+      "step": 957
+    },
+    {
+      "epoch": 0.3566683342834871,
+      "grad_norm": 0.18410375714302063,
+      "learning_rate": 1.916e-05,
+      "loss": 1.3131,
+      "step": 958
+    },
+    {
+      "epoch": 0.3570406394340962,
+      "grad_norm": 0.20000839233398438,
+      "learning_rate": 1.918e-05,
+      "loss": 1.3226,
+      "step": 959
+    },
+    {
+      "epoch": 0.35741294458470524,
+      "grad_norm": 0.17728938162326813,
+      "learning_rate": 1.9200000000000003e-05,
+      "loss": 1.3195,
+      "step": 960
+    },
+    {
+      "epoch": 0.3577852497353143,
+      "grad_norm": 0.19262300431728363,
+      "learning_rate": 1.9220000000000002e-05,
+      "loss": 1.3395,
+      "step": 961
+    },
+    {
+      "epoch": 0.3581575548859234,
+      "grad_norm": 0.2089599370956421,
+      "learning_rate": 1.9240000000000002e-05,
+      "loss": 1.3296,
+      "step": 962
+    },
+    {
+      "epoch": 0.3585298600365324,
+      "grad_norm": 0.19653920829296112,
+      "learning_rate": 1.9260000000000002e-05,
+      "loss": 1.3266,
+      "step": 963
+    },
+    {
+      "epoch": 0.3589021651871415,
+      "grad_norm": 0.18377585709095,
+      "learning_rate": 1.9280000000000002e-05,
+      "loss": 1.3331,
+      "step": 964
+    },
+    {
+      "epoch": 0.35927447033775056,
+      "grad_norm": 0.18499638140201569,
+      "learning_rate": 1.93e-05,
+      "loss": 1.3195,
+      "step": 965
+    },
+    {
+      "epoch": 0.35964677548835966,
+      "grad_norm": 0.19310572743415833,
+      "learning_rate": 1.932e-05,
+      "loss": 1.3545,
+      "step": 966
+    },
+    {
+      "epoch": 0.3600190806389687,
+      "grad_norm": 0.19002506136894226,
+      "learning_rate": 1.934e-05,
+      "loss": 1.341,
+      "step": 967
+    },
+    {
+      "epoch": 0.3603913857895778,
+      "grad_norm": 0.1937621384859085,
+      "learning_rate": 1.936e-05,
+      "loss": 1.32,
+      "step": 968
+    },
+    {
+      "epoch": 0.36076369094018684,
+      "grad_norm": 0.1868428736925125,
+      "learning_rate": 1.938e-05,
+      "loss": 1.3172,
+      "step": 969
+    },
+    {
+      "epoch": 0.36113599609079594,
+      "grad_norm": 0.1846293956041336,
+      "learning_rate": 1.94e-05,
+      "loss": 1.3261,
+      "step": 970
+    },
+    {
+      "epoch": 0.361508301241405,
+      "grad_norm": 0.18833589553833008,
+      "learning_rate": 1.942e-05,
+      "loss": 1.3307,
+      "step": 971
+    },
+    {
+      "epoch": 0.3618806063920141,
+      "grad_norm": 0.1832055151462555,
+      "learning_rate": 1.944e-05,
+      "loss": 1.3167,
+      "step": 972
+    },
+    {
+      "epoch": 0.3622529115426231,
+      "grad_norm": 0.1855573207139969,
+      "learning_rate": 1.946e-05,
+      "loss": 1.3279,
+      "step": 973
+    },
+    {
+      "epoch": 0.36262521669323217,
+      "grad_norm": 0.1866430938243866,
+      "learning_rate": 1.948e-05,
+      "loss": 1.3361,
+      "step": 974
+    },
+    {
+      "epoch": 0.36299752184384126,
+      "grad_norm": 0.1905186027288437,
+      "learning_rate": 1.95e-05,
+      "loss": 1.3255,
+      "step": 975
+    },
+    {
+      "epoch": 0.3633698269944503,
+      "grad_norm": 0.1822662651538849,
+      "learning_rate": 1.9520000000000003e-05,
+      "loss": 1.314,
+      "step": 976
+    },
+    {
+      "epoch": 0.3637421321450594,
+      "grad_norm": 0.18091034889221191,
+      "learning_rate": 1.9540000000000003e-05,
+      "loss": 1.3267,
+      "step": 977
+    },
+    {
+      "epoch": 0.36411443729566845,
+      "grad_norm": 0.18915344774723053,
+      "learning_rate": 1.9560000000000002e-05,
+      "loss": 1.3161,
+      "step": 978
+    },
+    {
+      "epoch": 0.36448674244627755,
+      "grad_norm": 0.17233604192733765,
+      "learning_rate": 1.9580000000000002e-05,
+      "loss": 1.3453,
+      "step": 979
+    },
+    {
+      "epoch": 0.3648590475968866,
+      "grad_norm": 0.18806397914886475,
+      "learning_rate": 1.9600000000000002e-05,
+      "loss": 1.3442,
+      "step": 980
+    },
+    {
+      "epoch": 0.3652313527474957,
+      "grad_norm": 0.19142785668373108,
+      "learning_rate": 1.9620000000000002e-05,
+      "loss": 1.3218,
+      "step": 981
+    },
+    {
+      "epoch": 0.36560365789810473,
+      "grad_norm": 0.17856691777706146,
+      "learning_rate": 1.9640000000000002e-05,
+      "loss": 1.3334,
+      "step": 982
+    },
+    {
+      "epoch": 0.3659759630487138,
+      "grad_norm": 0.189533069729805,
+      "learning_rate": 1.966e-05,
+      "loss": 1.3201,
+      "step": 983
+    },
+    {
+      "epoch": 0.36634826819932287,
+      "grad_norm": 0.19316861033439636,
+      "learning_rate": 1.968e-05,
+      "loss": 1.3177,
+      "step": 984
+    },
+    {
+      "epoch": 0.3667205733499319,
+      "grad_norm": 0.18213561177253723,
+      "learning_rate": 1.97e-05,
+      "loss": 1.3317,
+      "step": 985
+    },
+    {
+      "epoch": 0.367092878500541,
+      "grad_norm": 0.21344400942325592,
+      "learning_rate": 1.972e-05,
+      "loss": 1.3263,
+      "step": 986
+    },
+    {
+      "epoch": 0.36746518365115005,
+      "grad_norm": 0.1816537231206894,
+      "learning_rate": 1.974e-05,
+      "loss": 1.3155,
+      "step": 987
+    },
+    {
+      "epoch": 0.36783748880175915,
+      "grad_norm": 0.18131639063358307,
+      "learning_rate": 1.976e-05,
+      "loss": 1.3265,
+      "step": 988
+    },
+    {
+      "epoch": 0.3682097939523682,
+      "grad_norm": 0.21126648783683777,
+      "learning_rate": 1.978e-05,
+      "loss": 1.3293,
+      "step": 989
+    },
+    {
+      "epoch": 0.3685820991029773,
+      "grad_norm": 0.17451252043247223,
+      "learning_rate": 1.98e-05,
+      "loss": 1.3293,
+      "step": 990
+    },
+    {
+      "epoch": 0.36895440425358633,
+      "grad_norm": 0.17249158024787903,
+      "learning_rate": 1.982e-05,
+      "loss": 1.3242,
+      "step": 991
+    },
+    {
+      "epoch": 0.36932670940419543,
+      "grad_norm": 0.18453699350357056,
+      "learning_rate": 1.9840000000000003e-05,
+      "loss": 1.335,
+      "step": 992
+    },
+    {
+      "epoch": 0.3696990145548045,
+      "grad_norm": 0.16820859909057617,
+      "learning_rate": 1.9860000000000003e-05,
+      "loss": 1.3249,
+      "step": 993
+    },
+    {
+      "epoch": 0.37007131970541357,
+      "grad_norm": 0.19031395018100739,
+      "learning_rate": 1.9880000000000003e-05,
+      "loss": 1.3255,
+      "step": 994
+    },
+    {
+      "epoch": 0.3704436248560226,
+      "grad_norm": 0.19492731988430023,
+      "learning_rate": 1.9900000000000003e-05,
+      "loss": 1.3323,
+      "step": 995
+    },
+    {
+      "epoch": 0.3708159300066317,
+      "grad_norm": 0.17573820054531097,
+      "learning_rate": 1.9920000000000002e-05,
+      "loss": 1.3371,
+      "step": 996
+    },
+    {
+      "epoch": 0.37118823515724075,
+      "grad_norm": 0.18719618022441864,
+      "learning_rate": 1.9940000000000002e-05,
+      "loss": 1.3166,
+      "step": 997
+    },
+    {
+      "epoch": 0.3715605403078498,
+      "grad_norm": 0.17977416515350342,
+      "learning_rate": 1.9960000000000002e-05,
+      "loss": 1.323,
+      "step": 998
+    },
+    {
+      "epoch": 0.3719328454584589,
+      "grad_norm": 0.1703476756811142,
+      "learning_rate": 1.9980000000000002e-05,
+      "loss": 1.3185,
+      "step": 999
+    },
+    {
+      "epoch": 0.37230515060906794,
+      "grad_norm": 0.1878376603126526,
+      "learning_rate": 2e-05,
+      "loss": 1.3445,
+      "step": 1000
+    },
+    {
+      "epoch": 0.37230515060906794,
+      "eval_loss": 1.3607902526855469,
+      "eval_runtime": 16.6559,
+      "eval_samples_per_second": 104.108,
+      "eval_steps_per_second": 5.223,
+      "step": 1000
+    },
+    {
+      "epoch": 0.37267745575967703,
+      "grad_norm": 0.18507210910320282,
+      "learning_rate": 1.9999999926150316e-05,
+      "loss": 1.3343,
+      "step": 1001
+    },
+    {
+      "epoch": 0.3730497609102861,
+      "grad_norm": 0.17154212296009064,
+      "learning_rate": 1.999999970460126e-05,
+      "loss": 1.3407,
+      "step": 1002
+    },
+    {
+      "epoch": 0.3734220660608952,
+      "grad_norm": 0.1845930516719818,
+      "learning_rate": 1.9999999335352835e-05,
+      "loss": 1.3236,
+      "step": 1003
+    },
+    {
+      "epoch": 0.3737943712115042,
+      "grad_norm": 0.1764208972454071,
+      "learning_rate": 1.9999998818405046e-05,
+      "loss": 1.3242,
+      "step": 1004
+    },
+    {
+      "epoch": 0.3741666763621133,
+      "grad_norm": 0.1841861754655838,
+      "learning_rate": 1.99999981537579e-05,
+      "loss": 1.3287,
+      "step": 1005
+    },
+    {
+      "epoch": 0.37453898151272236,
+      "grad_norm": 0.17870505154132843,
+      "learning_rate": 1.9999997341411412e-05,
+      "loss": 1.3134,
+      "step": 1006
+    },
+    {
+      "epoch": 0.37491128666333146,
+      "grad_norm": 0.18854598701000214,
+      "learning_rate": 1.999999638136559e-05,
+      "loss": 1.3089,
+      "step": 1007
+    },
+    {
+      "epoch": 0.3752835918139405,
+      "grad_norm": 0.19055992364883423,
+      "learning_rate": 1.9999995273620453e-05,
+      "loss": 1.3269,
+      "step": 1008
+    },
+    {
+      "epoch": 0.3756558969645496,
+      "grad_norm": 0.1775302290916443,
+      "learning_rate": 1.9999994018176008e-05,
+      "loss": 1.3235,
+      "step": 1009
+    },
+    {
+      "epoch": 0.37602820211515864,
+      "grad_norm": 0.17888285219669342,
+      "learning_rate": 1.999999261503228e-05,
+      "loss": 1.3249,
+      "step": 1010
+    },
+    {
+      "epoch": 0.3764005072657677,
+      "grad_norm": 0.17626729607582092,
+      "learning_rate": 1.999999106418929e-05,
+      "loss": 1.3312,
+      "step": 1011
+    },
+    {
+      "epoch": 0.3767728124163768,
+      "grad_norm": 0.1689792424440384,
+      "learning_rate": 1.999998936564706e-05,
+      "loss": 1.3235,
+      "step": 1012
+    },
+    {
+      "epoch": 0.3771451175669858,
+      "grad_norm": 0.18726208806037903,
+      "learning_rate": 1.9999987519405614e-05,
+      "loss": 1.3249,
+      "step": 1013
+    },
+    {
+      "epoch": 0.3775174227175949,
+      "grad_norm": 0.1736321747303009,
+      "learning_rate": 1.9999985525464977e-05,
+      "loss": 1.3251,
+      "step": 1014
+    },
+    {
+      "epoch": 0.37788972786820396,
+      "grad_norm": 0.1852688491344452,
+      "learning_rate": 1.9999983383825186e-05,
+      "loss": 1.3272,
+      "step": 1015
+    },
+    {
+      "epoch": 0.37826203301881306,
+      "grad_norm": 0.17666661739349365,
+      "learning_rate": 1.9999981094486264e-05,
+      "loss": 1.3341,
+      "step": 1016
+    },
+    {
+      "epoch": 0.3786343381694221,
+      "grad_norm": 0.17533712089061737,
+      "learning_rate": 1.9999978657448252e-05,
+      "loss": 1.3045,
+      "step": 1017
+    },
+    {
+      "epoch": 0.3790066433200312,
+      "grad_norm": 0.1879895031452179,
+      "learning_rate": 1.999997607271118e-05,
+      "loss": 1.3311,
+      "step": 1018
+    },
+    {
+      "epoch": 0.37937894847064024,
+      "grad_norm": 0.17512273788452148,
+      "learning_rate": 1.999997334027509e-05,
+      "loss": 1.3284,
+      "step": 1019
+    },
+    {
+      "epoch": 0.37975125362124934,
+      "grad_norm": 0.17723974585533142,
+      "learning_rate": 1.9999970460140023e-05,
+      "loss": 1.3109,
+      "step": 1020
+    },
+    {
+      "epoch": 0.3801235587718584,
+      "grad_norm": 0.1877744346857071,
+      "learning_rate": 1.999996743230602e-05,
+      "loss": 1.3204,
+      "step": 1021
+    },
+    {
+      "epoch": 0.3804958639224674,
+      "grad_norm": 0.17423878610134125,
+      "learning_rate": 1.9999964256773125e-05,
+      "loss": 1.3204,
+      "step": 1022
+    },
+    {
+      "epoch": 0.3808681690730765,
+      "grad_norm": 0.19012653827667236,
+      "learning_rate": 1.9999960933541383e-05,
+      "loss": 1.3345,
+      "step": 1023
+    },
+    {
+      "epoch": 0.38124047422368557,
+      "grad_norm": 0.19567571580410004,
+      "learning_rate": 1.999995746261085e-05,
+      "loss": 1.3195,
+      "step": 1024
+    },
+    {
+      "epoch": 0.38161277937429466,
+      "grad_norm": 0.17858469486236572,
+      "learning_rate": 1.999995384398157e-05,
+      "loss": 1.3161,
+      "step": 1025
+    },
+    {
+      "epoch": 0.3819850845249037,
+      "grad_norm": 0.20314441621303558,
+      "learning_rate": 1.9999950077653597e-05,
+      "loss": 1.3214,
+      "step": 1026
+    },
+    {
+      "epoch": 0.3823573896755128,
+      "grad_norm": 0.19132466614246368,
+      "learning_rate": 1.9999946163626993e-05,
+      "loss": 1.3227,
+      "step": 1027
+    },
+    {
+      "epoch": 0.38272969482612185,
+      "grad_norm": 0.18081142008304596,
+      "learning_rate": 1.999994210190181e-05,
+      "loss": 1.3226,
+      "step": 1028
+    },
+    {
+      "epoch": 0.38310199997673094,
+      "grad_norm": 0.1872538924217224,
+      "learning_rate": 1.999993789247811e-05,
+      "loss": 1.321,
+      "step": 1029
+    },
+    {
+      "epoch": 0.38347430512734,
+      "grad_norm": 0.16693538427352905,
+      "learning_rate": 1.9999933535355955e-05,
+      "loss": 1.32,
+      "step": 1030
+    },
+    {
+      "epoch": 0.3838466102779491,
+      "grad_norm": 0.18578219413757324,
+      "learning_rate": 1.9999929030535406e-05,
+      "loss": 1.3046,
+      "step": 1031
+    },
+    {
+      "epoch": 0.3842189154285581,
+      "grad_norm": 0.19395247101783752,
+      "learning_rate": 1.999992437801654e-05,
+      "loss": 1.3357,
+      "step": 1032
+    },
+    {
+      "epoch": 0.3845912205791672,
+      "grad_norm": 0.18756742775440216,
+      "learning_rate": 1.9999919577799415e-05,
+      "loss": 1.3373,
+      "step": 1033
+    },
+    {
+      "epoch": 0.38496352572977627,
+      "grad_norm": 0.18206077814102173,
+      "learning_rate": 1.9999914629884104e-05,
+      "loss": 1.3087,
+      "step": 1034
+    },
+    {
+      "epoch": 0.3853358308803853,
+      "grad_norm": 0.18908272683620453,
+      "learning_rate": 1.999990953427068e-05,
+      "loss": 1.3191,
+      "step": 1035
+    },
+    {
+      "epoch": 0.3857081360309944,
+      "grad_norm": 0.19302143156528473,
+      "learning_rate": 1.9999904290959225e-05,
+      "loss": 1.3283,
+      "step": 1036
+    },
+    {
+      "epoch": 0.38608044118160345,
+      "grad_norm": 0.1931566596031189,
+      "learning_rate": 1.999989889994981e-05,
+      "loss": 1.3368,
+      "step": 1037
+    },
+    {
+      "epoch": 0.38645274633221255,
+      "grad_norm": 0.18609251081943512,
+      "learning_rate": 1.9999893361242512e-05,
+      "loss": 1.3135,
+      "step": 1038
+    },
+    {
+      "epoch": 0.3868250514828216,
+      "grad_norm": 0.19373655319213867,
+      "learning_rate": 1.9999887674837416e-05,
+      "loss": 1.3423,
+      "step": 1039
+    },
+    {
+      "epoch": 0.3871973566334307,
+      "grad_norm": 0.2070687711238861,
+      "learning_rate": 1.9999881840734613e-05,
+      "loss": 1.3103,
+      "step": 1040
+    },
+    {
+      "epoch": 0.38756966178403973,
+      "grad_norm": 0.18242163956165314,
+      "learning_rate": 1.999987585893418e-05,
+      "loss": 1.3367,
+      "step": 1041
+    },
+    {
+      "epoch": 0.38794196693464883,
+      "grad_norm": 0.17910490930080414,
+      "learning_rate": 1.9999869729436205e-05,
+      "loss": 1.3158,
+      "step": 1042
+    },
+    {
+      "epoch": 0.38831427208525787,
+      "grad_norm": 0.20917744934558868,
+      "learning_rate": 1.9999863452240784e-05,
+      "loss": 1.3203,
+      "step": 1043
+    },
+    {
+      "epoch": 0.38868657723586697,
+      "grad_norm": 0.16768380999565125,
+      "learning_rate": 1.9999857027348008e-05,
+      "loss": 1.324,
+      "step": 1044
+    },
+    {
+      "epoch": 0.389058882386476,
+      "grad_norm": 0.1901809275150299,
+      "learning_rate": 1.999985045475797e-05,
+      "loss": 1.333,
+      "step": 1045
+    },
+    {
+      "epoch": 0.3894311875370851,
+      "grad_norm": 0.18755842745304108,
+      "learning_rate": 1.9999843734470768e-05,
+      "loss": 1.3174,
+      "step": 1046
+    },
+    {
+      "epoch": 0.38980349268769415,
+      "grad_norm": 0.18284977972507477,
+      "learning_rate": 1.9999836866486505e-05,
+      "loss": 1.3204,
+      "step": 1047
+    },
+    {
+      "epoch": 0.3901757978383032,
+      "grad_norm": 0.19717617332935333,
+      "learning_rate": 1.9999829850805273e-05,
+      "loss": 1.3158,
+      "step": 1048
+    },
+    {
+      "epoch": 0.3905481029889123,
+      "grad_norm": 0.19506654143333435,
+      "learning_rate": 1.9999822687427188e-05,
+      "loss": 1.3254,
+      "step": 1049
+    },
+    {
+      "epoch": 0.39092040813952134,
+      "grad_norm": 0.18811020255088806,
+      "learning_rate": 1.9999815376352346e-05,
+      "loss": 1.3195,
+      "step": 1050
+    },
+    {
+      "epoch": 0.39129271329013043,
+      "grad_norm": 0.19153869152069092,
+      "learning_rate": 1.9999807917580858e-05,
+      "loss": 1.3346,
+      "step": 1051
+    },
+    {
+      "epoch": 0.3916650184407395,
+      "grad_norm": 0.17620836198329926,
+      "learning_rate": 1.9999800311112838e-05,
+      "loss": 1.3085,
+      "step": 1052
+    },
+    {
+      "epoch": 0.3920373235913486,
+      "grad_norm": 0.18910714983940125,
+      "learning_rate": 1.999979255694839e-05,
+      "loss": 1.3138,
+      "step": 1053
+    },
+    {
+      "epoch": 0.3924096287419576,
+      "grad_norm": 0.1759803593158722,
+      "learning_rate": 1.999978465508764e-05,
+      "loss": 1.3205,
+      "step": 1054
+    },
+    {
+      "epoch": 0.3927819338925667,
+      "grad_norm": 0.1760639101266861,
+      "learning_rate": 1.9999776605530693e-05,
+      "loss": 1.3111,
+      "step": 1055
+    },
+    {
+      "epoch": 0.39315423904317576,
+      "grad_norm": 0.1799839735031128,
+      "learning_rate": 1.9999768408277674e-05,
+      "loss": 1.3128,
+      "step": 1056
+    },
+    {
+      "epoch": 0.39352654419378486,
+      "grad_norm": 0.18230627477169037,
+      "learning_rate": 1.9999760063328705e-05,
+      "loss": 1.3294,
+      "step": 1057
+    },
+    {
+      "epoch": 0.3938988493443939,
+      "grad_norm": 0.17548716068267822,
+      "learning_rate": 1.9999751570683905e-05,
+      "loss": 1.3224,
+      "step": 1058
+    },
+    {
+      "epoch": 0.39427115449500294,
+      "grad_norm": 0.17289309203624725,
+      "learning_rate": 1.9999742930343404e-05,
+      "loss": 1.3113,
+      "step": 1059
+    },
+    {
+      "epoch": 0.39464345964561204,
+      "grad_norm": 0.18272966146469116,
+      "learning_rate": 1.9999734142307326e-05,
+      "loss": 1.3166,
+      "step": 1060
+    },
+    {
+      "epoch": 0.3950157647962211,
+      "grad_norm": 0.1845044642686844,
+      "learning_rate": 1.99997252065758e-05,
+      "loss": 1.3138,
+      "step": 1061
+    },
+    {
+      "epoch": 0.3953880699468302,
+      "grad_norm": 0.1838754415512085,
+      "learning_rate": 1.9999716123148966e-05,
+      "loss": 1.312,
+      "step": 1062
+    },
+    {
+      "epoch": 0.3957603750974392,
+      "grad_norm": 0.18985752761363983,
+      "learning_rate": 1.999970689202695e-05,
+      "loss": 1.3226,
+      "step": 1063
+    },
+    {
+      "epoch": 0.3961326802480483,
+      "grad_norm": 0.1912066787481308,
+      "learning_rate": 1.999969751320989e-05,
+      "loss": 1.3408,
+      "step": 1064
+    },
+    {
+      "epoch": 0.39650498539865736,
+      "grad_norm": 0.1811860352754593,
+      "learning_rate": 1.9999687986697925e-05,
+      "loss": 1.3085,
+      "step": 1065
+    },
+    {
+      "epoch": 0.39687729054926646,
+      "grad_norm": 0.18986211717128754,
+      "learning_rate": 1.9999678312491194e-05,
+      "loss": 1.3329,
+      "step": 1066
+    },
+    {
+      "epoch": 0.3972495956998755,
+      "grad_norm": 0.19396282732486725,
+      "learning_rate": 1.9999668490589848e-05,
+      "loss": 1.3115,
+      "step": 1067
+    },
+    {
+      "epoch": 0.3976219008504846,
+      "grad_norm": 0.19043684005737305,
+      "learning_rate": 1.999965852099402e-05,
+      "loss": 1.3174,
+      "step": 1068
+    },
+    {
+      "epoch": 0.39799420600109364,
+      "grad_norm": 0.18925096094608307,
+      "learning_rate": 1.9999648403703867e-05,
+      "loss": 1.3015,
+      "step": 1069
+    },
+    {
+      "epoch": 0.39836651115170274,
+      "grad_norm": 0.19922588765621185,
+      "learning_rate": 1.9999638138719532e-05,
+      "loss": 1.3154,
+      "step": 1070
+    },
+    {
+      "epoch": 0.3987388163023118,
+      "grad_norm": 0.1778346300125122,
+      "learning_rate": 1.999962772604117e-05,
+      "loss": 1.314,
+      "step": 1071
+    },
+    {
+      "epoch": 0.3991111214529208,
+      "grad_norm": 0.18273292481899261,
+      "learning_rate": 1.9999617165668935e-05,
+      "loss": 1.2956,
+      "step": 1072
+    },
+    {
+      "epoch": 0.3994834266035299,
+      "grad_norm": 0.18473650515079498,
+      "learning_rate": 1.999960645760298e-05,
+      "loss": 1.3015,
+      "step": 1073
+    },
+    {
+      "epoch": 0.39985573175413897,
+      "grad_norm": 0.20028461515903473,
+      "learning_rate": 1.9999595601843466e-05,
+      "loss": 1.3271,
+      "step": 1074
+    },
+    {
+      "epoch": 0.40022803690474806,
+      "grad_norm": 0.1781882643699646,
+      "learning_rate": 1.9999584598390556e-05,
+      "loss": 1.3189,
+      "step": 1075
+    },
+    {
+      "epoch": 0.4006003420553571,
+      "grad_norm": 0.18099310994148254,
+      "learning_rate": 1.9999573447244404e-05,
+      "loss": 1.3319,
+      "step": 1076
+    },
+    {
+      "epoch": 0.4009726472059662,
+      "grad_norm": 0.18824267387390137,
+      "learning_rate": 1.9999562148405184e-05,
+      "loss": 1.315,
+      "step": 1077
+    },
+    {
+      "epoch": 0.40134495235657525,
+      "grad_norm": 0.18727093935012817,
+      "learning_rate": 1.9999550701873056e-05,
+      "loss": 1.3246,
+      "step": 1078
+    },
+    {
+      "epoch": 0.40171725750718434,
+      "grad_norm": 0.19821962714195251,
+      "learning_rate": 1.9999539107648195e-05,
+      "loss": 1.2926,
+      "step": 1079
+    },
+    {
+      "epoch": 0.4020895626577934,
+      "grad_norm": 0.18535029888153076,
+      "learning_rate": 1.9999527365730766e-05,
+      "loss": 1.3103,
+      "step": 1080
+    },
+    {
+      "epoch": 0.4024618678084025,
+      "grad_norm": 0.18775507807731628,
+      "learning_rate": 1.9999515476120945e-05,
+      "loss": 1.3188,
+      "step": 1081
+    },
+    {
+      "epoch": 0.4028341729590115,
+      "grad_norm": 0.18621505796909332,
+      "learning_rate": 1.999950343881891e-05,
+      "loss": 1.3179,
+      "step": 1082
+    },
+    {
+      "epoch": 0.4032064781096206,
+      "grad_norm": 0.18887226283550262,
+      "learning_rate": 1.9999491253824833e-05,
+      "loss": 1.3146,
+      "step": 1083
+    },
+    {
+      "epoch": 0.40357878326022967,
+      "grad_norm": 0.18072552978992462,
+      "learning_rate": 1.9999478921138898e-05,
+      "loss": 1.3137,
+      "step": 1084
+    },
+    {
+      "epoch": 0.4039510884108387,
+      "grad_norm": 0.18956084549427032,
+      "learning_rate": 1.999946644076129e-05,
+      "loss": 1.3064,
+      "step": 1085
+    },
+    {
+      "epoch": 0.4043233935614478,
+      "grad_norm": 0.19679097831249237,
+      "learning_rate": 1.999945381269219e-05,
+      "loss": 1.2912,
+      "step": 1086
+    },
+    {
+      "epoch": 0.40469569871205685,
+      "grad_norm": 0.18827444314956665,
+      "learning_rate": 1.9999441036931784e-05,
+      "loss": 1.328,
+      "step": 1087
+    },
+    {
+      "epoch": 0.40506800386266595,
+      "grad_norm": 0.1727646291255951,
+      "learning_rate": 1.999942811348026e-05,
+      "loss": 1.3148,
+      "step": 1088
+    },
+    {
+      "epoch": 0.405440309013275,
+      "grad_norm": 0.18676549196243286,
+      "learning_rate": 1.999941504233781e-05,
+      "loss": 1.3315,
+      "step": 1089
+    },
+    {
+      "epoch": 0.4058126141638841,
+      "grad_norm": 0.18589061498641968,
+      "learning_rate": 1.9999401823504628e-05,
+      "loss": 1.3326,
+      "step": 1090
+    },
+    {
+      "epoch": 0.40618491931449313,
+      "grad_norm": 0.19661076366901398,
+      "learning_rate": 1.999938845698091e-05,
+      "loss": 1.3221,
+      "step": 1091
+    },
+    {
+      "epoch": 0.40655722446510223,
+      "grad_norm": 0.1824602484703064,
+      "learning_rate": 1.9999374942766853e-05,
+      "loss": 1.3216,
+      "step": 1092
+    },
+    {
+      "epoch": 0.40692952961571127,
+      "grad_norm": 0.1898437738418579,
+      "learning_rate": 1.999936128086265e-05,
+      "loss": 1.3212,
+      "step": 1093
+    },
+    {
+      "epoch": 0.40730183476632037,
+      "grad_norm": 0.18662068247795105,
+      "learning_rate": 1.9999347471268517e-05,
+      "loss": 1.316,
+      "step": 1094
+    },
+    {
+      "epoch": 0.4076741399169294,
+      "grad_norm": 0.17926158010959625,
+      "learning_rate": 1.9999333513984644e-05,
+      "loss": 1.3243,
+      "step": 1095
+    },
+    {
+      "epoch": 0.40804644506753845,
+      "grad_norm": 0.18369993567466736,
+      "learning_rate": 1.9999319409011243e-05,
+      "loss": 1.3001,
+      "step": 1096
+    },
+    {
+      "epoch": 0.40841875021814755,
+      "grad_norm": 0.18747882544994354,
+      "learning_rate": 1.9999305156348523e-05,
+      "loss": 1.3137,
+      "step": 1097
+    },
+    {
+      "epoch": 0.4087910553687566,
+      "grad_norm": 0.17961914837360382,
+      "learning_rate": 1.999929075599669e-05,
+      "loss": 1.3138,
+      "step": 1098
+    },
+    {
+      "epoch": 0.4091633605193657,
+      "grad_norm": 0.1906268298625946,
+      "learning_rate": 1.9999276207955965e-05,
+      "loss": 1.3301,
+      "step": 1099
+    },
+    {
+      "epoch": 0.40953566566997474,
+      "grad_norm": 0.1943434178829193,
+      "learning_rate": 1.9999261512226556e-05,
+      "loss": 1.3254,
+      "step": 1100
+    },
+    {
+      "epoch": 0.40990797082058383,
+      "grad_norm": 0.17660486698150635,
+      "learning_rate": 1.999924666880868e-05,
+      "loss": 1.3178,
+      "step": 1101
+    },
+    {
+      "epoch": 0.4102802759711929,
+      "grad_norm": 0.1914491206407547,
+      "learning_rate": 1.999923167770256e-05,
+      "loss": 1.303,
+      "step": 1102
+    },
+    {
+      "epoch": 0.410652581121802,
+      "grad_norm": 0.18666186928749084,
+      "learning_rate": 1.9999216538908416e-05,
+      "loss": 1.3144,
+      "step": 1103
+    },
+    {
+      "epoch": 0.411024886272411,
+      "grad_norm": 0.18117761611938477,
+      "learning_rate": 1.9999201252426473e-05,
+      "loss": 1.3063,
+      "step": 1104
+    },
+    {
+      "epoch": 0.4113971914230201,
+      "grad_norm": 0.1789981722831726,
+      "learning_rate": 1.9999185818256953e-05,
+      "loss": 1.3212,
+      "step": 1105
+    },
+    {
+      "epoch": 0.41176949657362916,
+      "grad_norm": 0.18627247214317322,
+      "learning_rate": 1.9999170236400087e-05,
+      "loss": 1.3088,
+      "step": 1106
+    },
+    {
+      "epoch": 0.41214180172423825,
+      "grad_norm": 0.18642041087150574,
+      "learning_rate": 1.99991545068561e-05,
+      "loss": 1.3296,
+      "step": 1107
+    },
+    {
+      "epoch": 0.4125141068748473,
+      "grad_norm": 0.1862434297800064,
+      "learning_rate": 1.999913862962523e-05,
+      "loss": 1.3261,
+      "step": 1108
+    },
+    {
+      "epoch": 0.41288641202545634,
+      "grad_norm": 0.1830601990222931,
+      "learning_rate": 1.9999122604707714e-05,
+      "loss": 1.3262,
+      "step": 1109
+    },
+    {
+      "epoch": 0.41325871717606544,
+      "grad_norm": 0.1928444504737854,
+      "learning_rate": 1.9999106432103785e-05,
+      "loss": 1.3062,
+      "step": 1110
+    },
+    {
+      "epoch": 0.4136310223266745,
+      "grad_norm": 0.1889939159154892,
+      "learning_rate": 1.9999090111813674e-05,
+      "loss": 1.3028,
+      "step": 1111
+    },
+    {
+      "epoch": 0.4140033274772836,
+      "grad_norm": 0.18017908930778503,
+      "learning_rate": 1.9999073643837637e-05,
+      "loss": 1.3231,
+      "step": 1112
+    },
+    {
+      "epoch": 0.4143756326278926,
+      "grad_norm": 0.18042001128196716,
+      "learning_rate": 1.9999057028175906e-05,
+      "loss": 1.301,
+      "step": 1113
+    },
+    {
+      "epoch": 0.4147479377785017,
+      "grad_norm": 0.1881111115217209,
+      "learning_rate": 1.999904026482873e-05,
+      "loss": 1.3057,
+      "step": 1114
+    },
+    {
+      "epoch": 0.41512024292911076,
+      "grad_norm": 0.17957095801830292,
+      "learning_rate": 1.9999023353796357e-05,
+      "loss": 1.316,
+      "step": 1115
+    },
+    {
+      "epoch": 0.41549254807971986,
+      "grad_norm": 0.19074207544326782,
+      "learning_rate": 1.999900629507904e-05,
+      "loss": 1.3326,
+      "step": 1116
+    },
+    {
+      "epoch": 0.4158648532303289,
+      "grad_norm": 0.19031281769275665,
+      "learning_rate": 1.9998989088677027e-05,
+      "loss": 1.3277,
+      "step": 1117
+    },
+    {
+      "epoch": 0.416237158380938,
+      "grad_norm": 0.17994950711727142,
+      "learning_rate": 1.9998971734590567e-05,
+      "loss": 1.3236,
+      "step": 1118
+    },
+    {
+      "epoch": 0.41660946353154704,
+      "grad_norm": 0.17953146994113922,
+      "learning_rate": 1.9998954232819928e-05,
+      "loss": 1.3225,
+      "step": 1119
+    },
+    {
+      "epoch": 0.41698176868215614,
+      "grad_norm": 0.18217253684997559,
+      "learning_rate": 1.9998936583365358e-05,
+      "loss": 1.3234,
+      "step": 1120
+    },
+    {
+      "epoch": 0.4173540738327652,
+      "grad_norm": 0.1890113651752472,
+      "learning_rate": 1.9998918786227124e-05,
+      "loss": 1.3206,
+      "step": 1121
+    },
+    {
+      "epoch": 0.4177263789833742,
+      "grad_norm": 0.18318399786949158,
+      "learning_rate": 1.999890084140549e-05,
+      "loss": 1.3222,
+      "step": 1122
+    },
+    {
+      "epoch": 0.4180986841339833,
+      "grad_norm": 0.19415515661239624,
+      "learning_rate": 1.9998882748900714e-05,
+      "loss": 1.3069,
+      "step": 1123
+    },
+    {
+      "epoch": 0.41847098928459237,
+      "grad_norm": 0.1776365041732788,
+      "learning_rate": 1.9998864508713068e-05,
+      "loss": 1.3146,
+      "step": 1124
+    },
+    {
+      "epoch": 0.41884329443520146,
+      "grad_norm": 0.18183977901935577,
+      "learning_rate": 1.9998846120842824e-05,
+      "loss": 1.3096,
+      "step": 1125
+    },
+    {
+      "epoch": 0.4192155995858105,
+      "grad_norm": 0.19202455878257751,
+      "learning_rate": 1.9998827585290245e-05,
+      "loss": 1.2982,
+      "step": 1126
+    },
+    {
+      "epoch": 0.4195879047364196,
+      "grad_norm": 0.18983449041843414,
+      "learning_rate": 1.9998808902055616e-05,
+      "loss": 1.32,
+      "step": 1127
+    },
+    {
+      "epoch": 0.41996020988702865,
+      "grad_norm": 0.17936941981315613,
+      "learning_rate": 1.9998790071139202e-05,
+      "loss": 1.3271,
+      "step": 1128
+    },
+    {
+      "epoch": 0.42033251503763774,
+      "grad_norm": 0.18909084796905518,
+      "learning_rate": 1.9998771092541287e-05,
+      "loss": 1.3014,
+      "step": 1129
+    },
+    {
+      "epoch": 0.4207048201882468,
+      "grad_norm": 0.19434739649295807,
+      "learning_rate": 1.9998751966262154e-05,
+      "loss": 1.3097,
+      "step": 1130
+    },
+    {
+      "epoch": 0.4210771253388559,
+      "grad_norm": 0.17867596447467804,
+      "learning_rate": 1.9998732692302077e-05,
+      "loss": 1.3138,
+      "step": 1131
+    },
+    {
+      "epoch": 0.4214494304894649,
+      "grad_norm": 0.17590618133544922,
+      "learning_rate": 1.999871327066135e-05,
+      "loss": 1.3162,
+      "step": 1132
+    },
+    {
+      "epoch": 0.42182173564007397,
+      "grad_norm": 0.19735798239707947,
+      "learning_rate": 1.999869370134025e-05,
+      "loss": 1.3109,
+      "step": 1133
+    },
+    {
+      "epoch": 0.42219404079068307,
+      "grad_norm": 0.18292424082756042,
+      "learning_rate": 1.999867398433908e-05,
+      "loss": 1.3073,
+      "step": 1134
+    },
+    {
+      "epoch": 0.4225663459412921,
+      "grad_norm": 0.1984787881374359,
+      "learning_rate": 1.9998654119658115e-05,
+      "loss": 1.3127,
+      "step": 1135
+    },
+    {
+      "epoch": 0.4229386510919012,
+      "grad_norm": 0.19433364272117615,
+      "learning_rate": 1.999863410729766e-05,
+      "loss": 1.3199,
+      "step": 1136
+    },
+    {
+      "epoch": 0.42331095624251025,
+      "grad_norm": 0.1794789433479309,
+      "learning_rate": 1.9998613947258006e-05,
+      "loss": 1.3065,
+      "step": 1137
+    },
+    {
+      "epoch": 0.42368326139311935,
+      "grad_norm": 0.18570461869239807,
+      "learning_rate": 1.9998593639539453e-05,
+      "loss": 1.3106,
+      "step": 1138
+    },
+    {
+      "epoch": 0.4240555665437284,
+      "grad_norm": 0.18316258490085602,
+      "learning_rate": 1.9998573184142294e-05,
+      "loss": 1.3232,
+      "step": 1139
+    },
+    {
+      "epoch": 0.4244278716943375,
+      "grad_norm": 0.19057685136795044,
+      "learning_rate": 1.9998552581066842e-05,
+      "loss": 1.3216,
+      "step": 1140
+    },
+    {
+      "epoch": 0.42480017684494653,
+      "grad_norm": 0.17034713923931122,
+      "learning_rate": 1.9998531830313394e-05,
+      "loss": 1.3071,
+      "step": 1141
+    },
+    {
+      "epoch": 0.42517248199555563,
+      "grad_norm": 0.19503635168075562,
+      "learning_rate": 1.999851093188226e-05,
+      "loss": 1.3007,
+      "step": 1142
+    },
+    {
+      "epoch": 0.42554478714616467,
+      "grad_norm": 0.18633443117141724,
+      "learning_rate": 1.9998489885773746e-05,
+      "loss": 1.3197,
+      "step": 1143
+    },
+    {
+      "epoch": 0.42591709229677377,
+      "grad_norm": 0.16780251264572144,
+      "learning_rate": 1.999846869198816e-05,
+      "loss": 1.3104,
+      "step": 1144
+    },
+    {
+      "epoch": 0.4262893974473828,
+      "grad_norm": 0.1772105097770691,
+      "learning_rate": 1.9998447350525822e-05,
+      "loss": 1.3154,
+      "step": 1145
+    },
+    {
+      "epoch": 0.42666170259799185,
+      "grad_norm": 0.1707058846950531,
+      "learning_rate": 1.9998425861387045e-05,
+      "loss": 1.3108,
+      "step": 1146
+    },
+    {
+      "epoch": 0.42703400774860095,
+      "grad_norm": 0.1806851178407669,
+      "learning_rate": 1.9998404224572147e-05,
+      "loss": 1.3152,
+      "step": 1147
+    },
+    {
+      "epoch": 0.42740631289921,
+      "grad_norm": 0.19098347425460815,
+      "learning_rate": 1.9998382440081442e-05,
+      "loss": 1.3187,
+      "step": 1148
+    },
+    {
+      "epoch": 0.4277786180498191,
+      "grad_norm": 0.18459391593933105,
+      "learning_rate": 1.999836050791526e-05,
+      "loss": 1.3139,
+      "step": 1149
+    },
+    {
+      "epoch": 0.42815092320042814,
+      "grad_norm": 0.1884915679693222,
+      "learning_rate": 1.9998338428073916e-05,
+      "loss": 1.318,
+      "step": 1150
+    },
+    {
+      "epoch": 0.42852322835103723,
+      "grad_norm": 0.181873619556427,
+      "learning_rate": 1.9998316200557742e-05,
+      "loss": 1.3283,
+      "step": 1151
+    },
+    {
+      "epoch": 0.4288955335016463,
+      "grad_norm": 0.17732377350330353,
+      "learning_rate": 1.9998293825367066e-05,
+      "loss": 1.2947,
+      "step": 1152
+    },
+    {
+      "epoch": 0.4292678386522554,
+      "grad_norm": 0.17769268155097961,
+      "learning_rate": 1.999827130250222e-05,
+      "loss": 1.3118,
+      "step": 1153
+    },
+    {
+      "epoch": 0.4296401438028644,
+      "grad_norm": 0.1880146563053131,
+      "learning_rate": 1.9998248631963532e-05,
+      "loss": 1.3123,
+      "step": 1154
+    },
+    {
+      "epoch": 0.4300124489534735,
+      "grad_norm": 0.18913207948207855,
+      "learning_rate": 1.9998225813751338e-05,
+      "loss": 1.3172,
+      "step": 1155
+    },
+    {
+      "epoch": 0.43038475410408256,
+      "grad_norm": 0.19146108627319336,
+      "learning_rate": 1.999820284786598e-05,
+      "loss": 1.3207,
+      "step": 1156
+    },
+    {
+      "epoch": 0.43075705925469165,
+      "grad_norm": 0.16881105303764343,
+      "learning_rate": 1.9998179734307794e-05,
+      "loss": 1.3047,
+      "step": 1157
+    },
+    {
+      "epoch": 0.4311293644053007,
+      "grad_norm": 0.18325145542621613,
+      "learning_rate": 1.9998156473077114e-05,
+      "loss": 1.3047,
+      "step": 1158
+    },
+    {
+      "epoch": 0.43150166955590974,
+      "grad_norm": 0.19172656536102295,
+      "learning_rate": 1.9998133064174297e-05,
+      "loss": 1.2973,
+      "step": 1159
+    },
+    {
+      "epoch": 0.43187397470651884,
+      "grad_norm": 0.18118645250797272,
+      "learning_rate": 1.9998109507599678e-05,
+      "loss": 1.3079,
+      "step": 1160
+    },
+    {
+      "epoch": 0.4322462798571279,
+      "grad_norm": 0.1834815889596939,
+      "learning_rate": 1.999808580335361e-05,
+      "loss": 1.3397,
+      "step": 1161
+    },
+    {
+      "epoch": 0.432618585007737,
+      "grad_norm": 0.19053733348846436,
+      "learning_rate": 1.9998061951436444e-05,
+      "loss": 1.3065,
+      "step": 1162
+    },
+    {
+      "epoch": 0.432990890158346,
+      "grad_norm": 0.18071451783180237,
+      "learning_rate": 1.999803795184853e-05,
+      "loss": 1.3144,
+      "step": 1163
+    },
+    {
+      "epoch": 0.4333631953089551,
+      "grad_norm": 0.17997939884662628,
+      "learning_rate": 1.9998013804590223e-05,
+      "loss": 1.311,
+      "step": 1164
+    },
+    {
+      "epoch": 0.43373550045956416,
+      "grad_norm": 0.19499291479587555,
+      "learning_rate": 1.999798950966188e-05,
+      "loss": 1.3367,
+      "step": 1165
+    },
+    {
+      "epoch": 0.43410780561017326,
+      "grad_norm": 0.20583321154117584,
+      "learning_rate": 1.9997965067063856e-05,
+      "loss": 1.312,
+      "step": 1166
+    },
+    {
+      "epoch": 0.4344801107607823,
+      "grad_norm": 0.18382562696933746,
+      "learning_rate": 1.9997940476796516e-05,
+      "loss": 1.3215,
+      "step": 1167
+    },
+    {
+      "epoch": 0.4348524159113914,
+      "grad_norm": 0.19162672758102417,
+      "learning_rate": 1.9997915738860224e-05,
+      "loss": 1.3025,
+      "step": 1168
+    },
+    {
+      "epoch": 0.43522472106200044,
+      "grad_norm": 0.1860532909631729,
+      "learning_rate": 1.9997890853255346e-05,
+      "loss": 1.2959,
+      "step": 1169
+    },
+    {
+      "epoch": 0.4355970262126095,
+      "grad_norm": 0.1830555945634842,
+      "learning_rate": 1.9997865819982247e-05,
+      "loss": 1.3056,
+      "step": 1170
+    },
+    {
+      "epoch": 0.4359693313632186,
+      "grad_norm": 0.18392489850521088,
+      "learning_rate": 1.9997840639041293e-05,
+      "loss": 1.3165,
+      "step": 1171
+    },
+    {
+      "epoch": 0.4363416365138276,
+      "grad_norm": 0.18183907866477966,
+      "learning_rate": 1.9997815310432864e-05,
+      "loss": 1.3102,
+      "step": 1172
+    },
+    {
+      "epoch": 0.4367139416644367,
+      "grad_norm": 0.19381216168403625,
+      "learning_rate": 1.999778983415733e-05,
+      "loss": 1.2985,
+      "step": 1173
+    },
+    {
+      "epoch": 0.43708624681504576,
+      "grad_norm": 0.18290908634662628,
+      "learning_rate": 1.9997764210215067e-05,
+      "loss": 1.3021,
+      "step": 1174
+    },
+    {
+      "epoch": 0.43745855196565486,
+      "grad_norm": 0.19478839635849,
+      "learning_rate": 1.9997738438606454e-05,
+      "loss": 1.3154,
+      "step": 1175
+    },
+    {
+      "epoch": 0.4378308571162639,
+      "grad_norm": 0.1829633265733719,
+      "learning_rate": 1.999771251933187e-05,
+      "loss": 1.3105,
+      "step": 1176
+    },
+    {
+      "epoch": 0.438203162266873,
+      "grad_norm": 0.1946062445640564,
+      "learning_rate": 1.9997686452391703e-05,
+      "loss": 1.315,
+      "step": 1177
+    },
+    {
+      "epoch": 0.43857546741748205,
+      "grad_norm": 0.185636505484581,
+      "learning_rate": 1.999766023778633e-05,
+      "loss": 1.3065,
+      "step": 1178
+    },
+    {
+      "epoch": 0.43894777256809114,
+      "grad_norm": 0.1923389434814453,
+      "learning_rate": 1.9997633875516148e-05,
+      "loss": 1.306,
+      "step": 1179
+    },
+    {
+      "epoch": 0.4393200777187002,
+      "grad_norm": 0.18182942271232605,
+      "learning_rate": 1.999760736558154e-05,
+      "loss": 1.295,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4396923828693093,
+      "grad_norm": 0.1975245475769043,
+      "learning_rate": 1.9997580707982896e-05,
+      "loss": 1.3185,
+      "step": 1181
+    },
+    {
+      "epoch": 0.4400646880199183,
+      "grad_norm": 0.1868218183517456,
+      "learning_rate": 1.999755390272061e-05,
+      "loss": 1.3127,
+      "step": 1182
+    },
+    {
+      "epoch": 0.44043699317052737,
+      "grad_norm": 0.18964411318302155,
+      "learning_rate": 1.9997526949795087e-05,
+      "loss": 1.297,
+      "step": 1183
+    },
+    {
+      "epoch": 0.44080929832113647,
+      "grad_norm": 0.19626370072364807,
+      "learning_rate": 1.9997499849206715e-05,
+      "loss": 1.2892,
+      "step": 1184
+    },
+    {
+      "epoch": 0.4411816034717455,
+      "grad_norm": 0.1894233077764511,
+      "learning_rate": 1.99974726009559e-05,
+      "loss": 1.3238,
+      "step": 1185
+    },
+    {
+      "epoch": 0.4415539086223546,
+      "grad_norm": 0.20008660852909088,
+      "learning_rate": 1.9997445205043037e-05,
+      "loss": 1.3053,
+      "step": 1186
+    },
+    {
+      "epoch": 0.44192621377296365,
+      "grad_norm": 0.18232440948486328,
+      "learning_rate": 1.999741766146854e-05,
+      "loss": 1.3007,
+      "step": 1187
+    },
+    {
+      "epoch": 0.44229851892357275,
+      "grad_norm": 0.19060754776000977,
+      "learning_rate": 1.999738997023281e-05,
+      "loss": 1.2931,
+      "step": 1188
+    },
+    {
+      "epoch": 0.4426708240741818,
+      "grad_norm": 0.2020169347524643,
+      "learning_rate": 1.999736213133626e-05,
+      "loss": 1.3155,
+      "step": 1189
+    },
+    {
+      "epoch": 0.4430431292247909,
+      "grad_norm": 0.1948639452457428,
+      "learning_rate": 1.9997334144779295e-05,
+      "loss": 1.3057,
+      "step": 1190
+    },
+    {
+      "epoch": 0.44341543437539993,
+      "grad_norm": 0.1938532143831253,
+      "learning_rate": 1.9997306010562334e-05,
+      "loss": 1.2947,
+      "step": 1191
+    },
+    {
+      "epoch": 0.44378773952600903,
+      "grad_norm": 0.21005672216415405,
+      "learning_rate": 1.9997277728685788e-05,
+      "loss": 1.3209,
+      "step": 1192
+    },
+    {
+      "epoch": 0.44416004467661807,
+      "grad_norm": 0.20058314502239227,
+      "learning_rate": 1.9997249299150078e-05,
+      "loss": 1.2951,
+      "step": 1193
+    },
+    {
+      "epoch": 0.4445323498272271,
+      "grad_norm": 0.1923152059316635,
+      "learning_rate": 1.9997220721955627e-05,
+      "loss": 1.3118,
+      "step": 1194
+    },
+    {
+      "epoch": 0.4449046549778362,
+      "grad_norm": 0.19684681296348572,
+      "learning_rate": 1.9997191997102853e-05,
+      "loss": 1.3241,
+      "step": 1195
+    },
+    {
+      "epoch": 0.44527696012844525,
+      "grad_norm": 0.19519715011119843,
+      "learning_rate": 1.9997163124592175e-05,
+      "loss": 1.3147,
+      "step": 1196
+    },
+    {
+      "epoch": 0.44564926527905435,
+      "grad_norm": 0.19331727921962738,
+      "learning_rate": 1.9997134104424033e-05,
+      "loss": 1.3278,
+      "step": 1197
+    },
+    {
+      "epoch": 0.4460215704296634,
+      "grad_norm": 0.19848302006721497,
+      "learning_rate": 1.999710493659884e-05,
+      "loss": 1.307,
+      "step": 1198
+    },
+    {
+      "epoch": 0.4463938755802725,
+      "grad_norm": 0.19539180397987366,
+      "learning_rate": 1.999707562111704e-05,
+      "loss": 1.2933,
+      "step": 1199
+    },
+    {
+      "epoch": 0.44676618073088153,
+      "grad_norm": 0.18977102637290955,
+      "learning_rate": 1.999704615797906e-05,
+      "loss": 1.3087,
+      "step": 1200
+    },
+    {
+      "epoch": 0.44713848588149063,
+      "grad_norm": 0.2267744392156601,
+      "learning_rate": 1.9997016547185333e-05,
+      "loss": 1.3055,
+      "step": 1201
+    },
+    {
+      "epoch": 0.4475107910320997,
+      "grad_norm": 0.19115746021270752,
+      "learning_rate": 1.9996986788736298e-05,
+      "loss": 1.2991,
+      "step": 1202
+    },
+    {
+      "epoch": 0.4478830961827088,
+      "grad_norm": 0.1895446479320526,
+      "learning_rate": 1.99969568826324e-05,
+      "loss": 1.3006,
+      "step": 1203
+    },
+    {
+      "epoch": 0.4482554013333178,
+      "grad_norm": 0.18717582523822784,
+      "learning_rate": 1.999692682887407e-05,
+      "loss": 1.3014,
+      "step": 1204
+    },
+    {
+      "epoch": 0.4486277064839269,
+      "grad_norm": 0.19228947162628174,
+      "learning_rate": 1.9996896627461764e-05,
+      "loss": 1.3189,
+      "step": 1205
+    },
+    {
+      "epoch": 0.44900001163453596,
+      "grad_norm": 0.19240860641002655,
+      "learning_rate": 1.999686627839592e-05,
+      "loss": 1.2877,
+      "step": 1206
+    },
+    {
+      "epoch": 0.449372316785145,
+      "grad_norm": 0.18697458505630493,
+      "learning_rate": 1.999683578167699e-05,
+      "loss": 1.3058,
+      "step": 1207
+    },
+    {
+      "epoch": 0.4497446219357541,
+      "grad_norm": 0.21261201798915863,
+      "learning_rate": 1.999680513730542e-05,
+      "loss": 1.3055,
+      "step": 1208
+    },
+    {
+      "epoch": 0.45011692708636314,
+      "grad_norm": 0.19081415235996246,
+      "learning_rate": 1.9996774345281668e-05,
+      "loss": 1.2985,
+      "step": 1209
+    },
+    {
+      "epoch": 0.45048923223697224,
+      "grad_norm": 0.18838602304458618,
+      "learning_rate": 1.9996743405606188e-05,
+      "loss": 1.3032,
+      "step": 1210
+    },
+    {
+      "epoch": 0.4508615373875813,
+      "grad_norm": 0.19826094806194305,
+      "learning_rate": 1.999671231827943e-05,
+      "loss": 1.3116,
+      "step": 1211
+    },
+    {
+      "epoch": 0.4512338425381904,
+      "grad_norm": 0.1974056512117386,
+      "learning_rate": 1.999668108330186e-05,
+      "loss": 1.3211,
+      "step": 1212
+    },
+    {
+      "epoch": 0.4516061476887994,
+      "grad_norm": 0.1901020109653473,
+      "learning_rate": 1.999664970067394e-05,
+      "loss": 1.303,
+      "step": 1213
+    },
+    {
+      "epoch": 0.4519784528394085,
+      "grad_norm": 0.18617480993270874,
+      "learning_rate": 1.999661817039613e-05,
+      "loss": 1.3013,
+      "step": 1214
+    },
+    {
+      "epoch": 0.45235075799001756,
+      "grad_norm": 0.20741531252861023,
+      "learning_rate": 1.9996586492468895e-05,
+      "loss": 1.3031,
+      "step": 1215
+    },
+    {
+      "epoch": 0.45272306314062666,
+      "grad_norm": 0.18789881467819214,
+      "learning_rate": 1.999655466689271e-05,
+      "loss": 1.3058,
+      "step": 1216
+    },
+    {
+      "epoch": 0.4530953682912357,
+      "grad_norm": 0.1936379075050354,
+      "learning_rate": 1.9996522693668034e-05,
+      "loss": 1.311,
+      "step": 1217
+    },
+    {
+      "epoch": 0.4534676734418448,
+      "grad_norm": 0.20415879786014557,
+      "learning_rate": 1.9996490572795348e-05,
+      "loss": 1.32,
+      "step": 1218
+    },
+    {
+      "epoch": 0.45383997859245384,
+      "grad_norm": 0.19625741243362427,
+      "learning_rate": 1.9996458304275125e-05,
+      "loss": 1.3058,
+      "step": 1219
+    },
+    {
+      "epoch": 0.4542122837430629,
+      "grad_norm": 0.20820340514183044,
+      "learning_rate": 1.999642588810784e-05,
+      "loss": 1.3154,
+      "step": 1220
+    },
+    {
+      "epoch": 0.454584588893672,
+      "grad_norm": 0.2056087851524353,
+      "learning_rate": 1.9996393324293972e-05,
+      "loss": 1.3051,
+      "step": 1221
+    },
+    {
+      "epoch": 0.454956894044281,
+      "grad_norm": 0.19374904036521912,
+      "learning_rate": 1.9996360612833997e-05,
+      "loss": 1.3236,
+      "step": 1222
+    },
+    {
+      "epoch": 0.4553291991948901,
+      "grad_norm": 0.18062308430671692,
+      "learning_rate": 1.999632775372841e-05,
+      "loss": 1.292,
+      "step": 1223
+    },
+    {
+      "epoch": 0.45570150434549916,
+      "grad_norm": 0.20089417695999146,
+      "learning_rate": 1.9996294746977686e-05,
+      "loss": 1.3074,
+      "step": 1224
+    },
+    {
+      "epoch": 0.45607380949610826,
+      "grad_norm": 0.19090351462364197,
+      "learning_rate": 1.9996261592582312e-05,
+      "loss": 1.3025,
+      "step": 1225
+    },
+    {
+      "epoch": 0.4564461146467173,
+      "grad_norm": 0.19648805260658264,
+      "learning_rate": 1.9996228290542787e-05,
+      "loss": 1.3098,
+      "step": 1226
+    },
+    {
+      "epoch": 0.4568184197973264,
+      "grad_norm": 0.1995651125907898,
+      "learning_rate": 1.99961948408596e-05,
+      "loss": 1.3083,
+      "step": 1227
+    },
+    {
+      "epoch": 0.45719072494793545,
+      "grad_norm": 0.1871064007282257,
+      "learning_rate": 1.9996161243533238e-05,
+      "loss": 1.2835,
+      "step": 1228
+    },
+    {
+      "epoch": 0.45756303009854454,
+      "grad_norm": 0.19911763072013855,
+      "learning_rate": 1.9996127498564203e-05,
+      "loss": 1.3044,
+      "step": 1229
+    },
+    {
+      "epoch": 0.4579353352491536,
+      "grad_norm": 0.17743955552577972,
+      "learning_rate": 1.9996093605952992e-05,
+      "loss": 1.3067,
+      "step": 1230
+    },
+    {
+      "epoch": 0.45830764039976263,
+      "grad_norm": 0.19110862910747528,
+      "learning_rate": 1.9996059565700103e-05,
+      "loss": 1.2972,
+      "step": 1231
+    },
+    {
+      "epoch": 0.4586799455503717,
+      "grad_norm": 0.18435370922088623,
+      "learning_rate": 1.9996025377806044e-05,
+      "loss": 1.3045,
+      "step": 1232
+    },
+    {
+      "epoch": 0.45905225070098077,
+      "grad_norm": 0.18806539475917816,
+      "learning_rate": 1.999599104227132e-05,
+      "loss": 1.3115,
+      "step": 1233
+    },
+    {
+      "epoch": 0.45942455585158987,
+      "grad_norm": 0.18353897333145142,
+      "learning_rate": 1.9995956559096432e-05,
+      "loss": 1.3079,
+      "step": 1234
+    },
+    {
+      "epoch": 0.4597968610021989,
+      "grad_norm": 0.18023177981376648,
+      "learning_rate": 1.9995921928281893e-05,
+      "loss": 1.3037,
+      "step": 1235
+    },
+    {
+      "epoch": 0.460169166152808,
+      "grad_norm": 0.18274232745170593,
+      "learning_rate": 1.9995887149828216e-05,
+      "loss": 1.3211,
+      "step": 1236
+    },
+    {
+      "epoch": 0.46054147130341705,
+      "grad_norm": 0.18951238691806793,
+      "learning_rate": 1.9995852223735914e-05,
+      "loss": 1.299,
+      "step": 1237
+    },
+    {
+      "epoch": 0.46091377645402615,
+      "grad_norm": 0.1818682998418808,
+      "learning_rate": 1.9995817150005502e-05,
+      "loss": 1.2999,
+      "step": 1238
+    },
+    {
+      "epoch": 0.4612860816046352,
+      "grad_norm": 0.20125912129878998,
+      "learning_rate": 1.9995781928637494e-05,
+      "loss": 1.3045,
+      "step": 1239
+    },
+    {
+      "epoch": 0.4616583867552443,
+      "grad_norm": 0.16718147695064545,
+      "learning_rate": 1.9995746559632417e-05,
+      "loss": 1.2862,
+      "step": 1240
+    },
+    {
+      "epoch": 0.46203069190585333,
+      "grad_norm": 0.19009731709957123,
+      "learning_rate": 1.999571104299079e-05,
+      "loss": 1.3052,
+      "step": 1241
+    },
+    {
+      "epoch": 0.46240299705646243,
+      "grad_norm": 0.18542075157165527,
+      "learning_rate": 1.999567537871314e-05,
+      "loss": 1.2998,
+      "step": 1242
+    },
+    {
+      "epoch": 0.46277530220707147,
+      "grad_norm": 0.18148809671401978,
+      "learning_rate": 1.999563956679999e-05,
+      "loss": 1.3011,
+      "step": 1243
+    },
+    {
+      "epoch": 0.4631476073576805,
+      "grad_norm": 0.1700984239578247,
+      "learning_rate": 1.9995603607251873e-05,
+      "loss": 1.3053,
+      "step": 1244
+    },
+    {
+      "epoch": 0.4635199125082896,
+      "grad_norm": 0.19107471406459808,
+      "learning_rate": 1.9995567500069314e-05,
+      "loss": 1.3124,
+      "step": 1245
+    },
+    {
+      "epoch": 0.46389221765889865,
+      "grad_norm": 0.18561914563179016,
+      "learning_rate": 1.9995531245252854e-05,
+      "loss": 1.3055,
+      "step": 1246
+    },
+    {
+      "epoch": 0.46426452280950775,
+      "grad_norm": 0.18733622133731842,
+      "learning_rate": 1.9995494842803026e-05,
+      "loss": 1.3053,
+      "step": 1247
+    },
+    {
+      "epoch": 0.4646368279601168,
+      "grad_norm": 0.18674108386039734,
+      "learning_rate": 1.9995458292720364e-05,
+      "loss": 1.3043,
+      "step": 1248
+    },
+    {
+      "epoch": 0.4650091331107259,
+      "grad_norm": 0.17907080054283142,
+      "learning_rate": 1.9995421595005408e-05,
+      "loss": 1.3161,
+      "step": 1249
+    },
+    {
+      "epoch": 0.46538143826133493,
+      "grad_norm": 0.17312780022621155,
+      "learning_rate": 1.9995384749658705e-05,
+      "loss": 1.321,
+      "step": 1250
+    },
+    {
+      "epoch": 0.46575374341194403,
+      "grad_norm": 0.18137013912200928,
+      "learning_rate": 1.99953477566808e-05,
+      "loss": 1.3175,
+      "step": 1251
+    },
+    {
+      "epoch": 0.4661260485625531,
+      "grad_norm": 0.1776915341615677,
+      "learning_rate": 1.999531061607223e-05,
+      "loss": 1.3135,
+      "step": 1252
+    },
+    {
+      "epoch": 0.4664983537131622,
+      "grad_norm": 0.18204154074192047,
+      "learning_rate": 1.9995273327833553e-05,
+      "loss": 1.2955,
+      "step": 1253
+    },
+    {
+      "epoch": 0.4668706588637712,
+      "grad_norm": 0.18510939180850983,
+      "learning_rate": 1.999523589196531e-05,
+      "loss": 1.2945,
+      "step": 1254
+    },
+    {
+      "epoch": 0.4672429640143803,
+      "grad_norm": 0.17913533747196198,
+      "learning_rate": 1.999519830846807e-05,
+      "loss": 1.3177,
+      "step": 1255
+    },
+    {
+      "epoch": 0.46761526916498936,
+      "grad_norm": 0.18509091436862946,
+      "learning_rate": 1.9995160577342375e-05,
+      "loss": 1.3117,
+      "step": 1256
+    },
+    {
+      "epoch": 0.4679875743155984,
+      "grad_norm": 0.19263845682144165,
+      "learning_rate": 1.999512269858878e-05,
+      "loss": 1.2951,
+      "step": 1257
+    },
+    {
+      "epoch": 0.4683598794662075,
+      "grad_norm": 0.1835833191871643,
+      "learning_rate": 1.9995084672207855e-05,
+      "loss": 1.3027,
+      "step": 1258
+    },
+    {
+      "epoch": 0.46873218461681654,
+      "grad_norm": 0.18313245475292206,
+      "learning_rate": 1.9995046498200158e-05,
+      "loss": 1.2837,
+      "step": 1259
+    },
+    {
+      "epoch": 0.46910448976742564,
+      "grad_norm": 0.1853981614112854,
+      "learning_rate": 1.9995008176566247e-05,
+      "loss": 1.2973,
+      "step": 1260
+    },
+    {
+      "epoch": 0.4694767949180347,
+      "grad_norm": 0.20372353494167328,
+      "learning_rate": 1.9994969707306697e-05,
+      "loss": 1.3229,
+      "step": 1261
+    },
+    {
+      "epoch": 0.4698491000686438,
+      "grad_norm": 0.183163583278656,
+      "learning_rate": 1.9994931090422067e-05,
+      "loss": 1.3065,
+      "step": 1262
+    },
+    {
+      "epoch": 0.4702214052192528,
+      "grad_norm": 0.18896111845970154,
+      "learning_rate": 1.9994892325912937e-05,
+      "loss": 1.3135,
+      "step": 1263
+    },
+    {
+      "epoch": 0.4705937103698619,
+      "grad_norm": 0.18443109095096588,
+      "learning_rate": 1.999485341377987e-05,
+      "loss": 1.3002,
+      "step": 1264
+    },
+    {
+      "epoch": 0.47096601552047096,
+      "grad_norm": 0.19512967765331268,
+      "learning_rate": 1.9994814354023446e-05,
+      "loss": 1.2936,
+      "step": 1265
+    },
+    {
+      "epoch": 0.47133832067108006,
+      "grad_norm": 0.1836230754852295,
+      "learning_rate": 1.9994775146644245e-05,
+      "loss": 1.2921,
+      "step": 1266
+    },
+    {
+      "epoch": 0.4717106258216891,
+      "grad_norm": 0.20863986015319824,
+      "learning_rate": 1.999473579164284e-05,
+      "loss": 1.3075,
+      "step": 1267
+    },
+    {
+      "epoch": 0.47208293097229814,
+      "grad_norm": 0.19831794500350952,
+      "learning_rate": 1.999469628901981e-05,
+      "loss": 1.3106,
+      "step": 1268
+    },
+    {
+      "epoch": 0.47245523612290724,
+      "grad_norm": 0.18101930618286133,
+      "learning_rate": 1.999465663877575e-05,
+      "loss": 1.2915,
+      "step": 1269
+    },
+    {
+      "epoch": 0.4728275412735163,
+      "grad_norm": 0.20090216398239136,
+      "learning_rate": 1.9994616840911237e-05,
+      "loss": 1.2941,
+      "step": 1270
+    },
+    {
+      "epoch": 0.4731998464241254,
+      "grad_norm": 0.20153005421161652,
+      "learning_rate": 1.9994576895426858e-05,
+      "loss": 1.3137,
+      "step": 1271
+    },
+    {
+      "epoch": 0.4735721515747344,
+      "grad_norm": 0.18877728283405304,
+      "learning_rate": 1.999453680232321e-05,
+      "loss": 1.2897,
+      "step": 1272
+    },
+    {
+      "epoch": 0.4739444567253435,
+      "grad_norm": 0.19803088903427124,
+      "learning_rate": 1.9994496561600874e-05,
+      "loss": 1.2905,
+      "step": 1273
+    },
+    {
+      "epoch": 0.47431676187595256,
+      "grad_norm": 0.19023656845092773,
+      "learning_rate": 1.9994456173260457e-05,
+      "loss": 1.3019,
+      "step": 1274
+    },
+    {
+      "epoch": 0.47468906702656166,
+      "grad_norm": 0.17500866949558258,
+      "learning_rate": 1.9994415637302545e-05,
+      "loss": 1.3089,
+      "step": 1275
+    },
+    {
+      "epoch": 0.4750613721771707,
+      "grad_norm": 0.18865026533603668,
+      "learning_rate": 1.9994374953727747e-05,
+      "loss": 1.2977,
+      "step": 1276
+    },
+    {
+      "epoch": 0.4754336773277798,
+      "grad_norm": 0.176120787858963,
+      "learning_rate": 1.9994334122536654e-05,
+      "loss": 1.2825,
+      "step": 1277
+    },
+    {
+      "epoch": 0.47580598247838884,
+      "grad_norm": 0.18565769493579865,
+      "learning_rate": 1.9994293143729873e-05,
+      "loss": 1.3182,
+      "step": 1278
+    },
+    {
+      "epoch": 0.47617828762899794,
+      "grad_norm": 0.18790839612483978,
+      "learning_rate": 1.9994252017308012e-05,
+      "loss": 1.3215,
+      "step": 1279
+    },
+    {
+      "epoch": 0.476550592779607,
+      "grad_norm": 0.19500873982906342,
+      "learning_rate": 1.9994210743271675e-05,
+      "loss": 1.3114,
+      "step": 1280
+    },
+    {
+      "epoch": 0.476922897930216,
+      "grad_norm": 0.1855781525373459,
+      "learning_rate": 1.9994169321621474e-05,
+      "loss": 1.3051,
+      "step": 1281
+    },
+    {
+      "epoch": 0.4772952030808251,
+      "grad_norm": 0.1813666671514511,
+      "learning_rate": 1.9994127752358014e-05,
+      "loss": 1.2957,
+      "step": 1282
+    },
+    {
+      "epoch": 0.47766750823143417,
+      "grad_norm": 0.20357996225357056,
+      "learning_rate": 1.999408603548192e-05,
+      "loss": 1.2988,
+      "step": 1283
+    },
+    {
+      "epoch": 0.47803981338204327,
+      "grad_norm": 0.17496971786022186,
+      "learning_rate": 1.99940441709938e-05,
+      "loss": 1.3004,
+      "step": 1284
+    },
+    {
+      "epoch": 0.4784121185326523,
+      "grad_norm": 0.19494318962097168,
+      "learning_rate": 1.9994002158894274e-05,
+      "loss": 1.3138,
+      "step": 1285
+    },
+    {
+      "epoch": 0.4787844236832614,
+      "grad_norm": 0.19843032956123352,
+      "learning_rate": 1.9993959999183964e-05,
+      "loss": 1.3031,
+      "step": 1286
+    },
+    {
+      "epoch": 0.47915672883387045,
+      "grad_norm": 0.19136017560958862,
+      "learning_rate": 1.9993917691863493e-05,
+      "loss": 1.3004,
+      "step": 1287
+    },
+    {
+      "epoch": 0.47952903398447955,
+      "grad_norm": 0.18518052995204926,
+      "learning_rate": 1.9993875236933486e-05,
+      "loss": 1.2824,
+      "step": 1288
+    },
+    {
+      "epoch": 0.4799013391350886,
+      "grad_norm": 0.18933340907096863,
+      "learning_rate": 1.9993832634394564e-05,
+      "loss": 1.2983,
+      "step": 1289
+    },
+    {
+      "epoch": 0.4802736442856977,
+      "grad_norm": 0.1836833357810974,
+      "learning_rate": 1.9993789884247365e-05,
+      "loss": 1.2916,
+      "step": 1290
+    },
+    {
+      "epoch": 0.48064594943630673,
+      "grad_norm": 0.17370636761188507,
+      "learning_rate": 1.9993746986492515e-05,
+      "loss": 1.2841,
+      "step": 1291
+    },
+    {
+      "epoch": 0.4810182545869158,
+      "grad_norm": 0.19281500577926636,
+      "learning_rate": 1.999370394113065e-05,
+      "loss": 1.3149,
+      "step": 1292
+    },
+    {
+      "epoch": 0.48139055973752487,
+      "grad_norm": 0.1878812462091446,
+      "learning_rate": 1.999366074816241e-05,
+      "loss": 1.3104,
+      "step": 1293
+    },
+    {
+      "epoch": 0.4817628648881339,
+      "grad_norm": 0.18660883605480194,
+      "learning_rate": 1.999361740758842e-05,
+      "loss": 1.3022,
+      "step": 1294
+    },
+    {
+      "epoch": 0.482135170038743,
+      "grad_norm": 0.1894964575767517,
+      "learning_rate": 1.999357391940933e-05,
+      "loss": 1.3056,
+      "step": 1295
+    },
+    {
+      "epoch": 0.48250747518935205,
+      "grad_norm": 0.18290723860263824,
+      "learning_rate": 1.999353028362578e-05,
+      "loss": 1.3045,
+      "step": 1296
+    },
+    {
+      "epoch": 0.48287978033996115,
+      "grad_norm": 0.1781376600265503,
+      "learning_rate": 1.9993486500238417e-05,
+      "loss": 1.3114,
+      "step": 1297
+    },
+    {
+      "epoch": 0.4832520854905702,
+      "grad_norm": 0.1864173412322998,
+      "learning_rate": 1.9993442569247885e-05,
+      "loss": 1.3034,
+      "step": 1298
+    },
+    {
+      "epoch": 0.4836243906411793,
+      "grad_norm": 0.19244234263896942,
+      "learning_rate": 1.9993398490654835e-05,
+      "loss": 1.3071,
+      "step": 1299
+    },
+    {
+      "epoch": 0.48399669579178833,
+      "grad_norm": 0.18302254378795624,
+      "learning_rate": 1.9993354264459913e-05,
+      "loss": 1.3059,
+      "step": 1300
+    },
+    {
+      "epoch": 0.48436900094239743,
+      "grad_norm": 0.17523464560508728,
+      "learning_rate": 1.9993309890663775e-05,
+      "loss": 1.2919,
+      "step": 1301
+    },
+    {
+      "epoch": 0.4847413060930065,
+      "grad_norm": 0.18764130771160126,
+      "learning_rate": 1.999326536926708e-05,
+      "loss": 1.2931,
+      "step": 1302
+    },
+    {
+      "epoch": 0.4851136112436156,
+      "grad_norm": 0.1812962144613266,
+      "learning_rate": 1.9993220700270484e-05,
+      "loss": 1.289,
+      "step": 1303
+    },
+    {
+      "epoch": 0.4854859163942246,
+      "grad_norm": 0.1877674013376236,
+      "learning_rate": 1.9993175883674642e-05,
+      "loss": 1.2957,
+      "step": 1304
+    },
+    {
+      "epoch": 0.48585822154483366,
+      "grad_norm": 0.24576683342456818,
+      "learning_rate": 1.9993130919480223e-05,
+      "loss": 1.3056,
+      "step": 1305
+    },
+    {
+      "epoch": 0.48623052669544276,
+      "grad_norm": 0.185043066740036,
+      "learning_rate": 1.9993085807687883e-05,
+      "loss": 1.297,
+      "step": 1306
+    },
+    {
+      "epoch": 0.4866028318460518,
+      "grad_norm": 0.19536662101745605,
+      "learning_rate": 1.9993040548298297e-05,
+      "loss": 1.3053,
+      "step": 1307
+    },
+    {
+      "epoch": 0.4869751369966609,
+      "grad_norm": 0.17894597351551056,
+      "learning_rate": 1.9992995141312126e-05,
+      "loss": 1.3136,
+      "step": 1308
+    },
+    {
+      "epoch": 0.48734744214726994,
+      "grad_norm": 0.1845971643924713,
+      "learning_rate": 1.9992949586730046e-05,
+      "loss": 1.3001,
+      "step": 1309
+    },
+    {
+      "epoch": 0.48771974729787904,
+      "grad_norm": 0.18812014162540436,
+      "learning_rate": 1.9992903884552727e-05,
+      "loss": 1.2948,
+      "step": 1310
+    },
+    {
+      "epoch": 0.4880920524484881,
+      "grad_norm": 0.18313339352607727,
+      "learning_rate": 1.9992858034780848e-05,
+      "loss": 1.3007,
+      "step": 1311
+    },
+    {
+      "epoch": 0.4884643575990972,
+      "grad_norm": 0.18373283743858337,
+      "learning_rate": 1.9992812037415077e-05,
+      "loss": 1.2924,
+      "step": 1312
+    },
+    {
+      "epoch": 0.4888366627497062,
+      "grad_norm": 0.1760503053665161,
+      "learning_rate": 1.9992765892456102e-05,
+      "loss": 1.2953,
+      "step": 1313
+    },
+    {
+      "epoch": 0.4892089679003153,
+      "grad_norm": 0.18641230463981628,
+      "learning_rate": 1.99927195999046e-05,
+      "loss": 1.288,
+      "step": 1314
+    },
+    {
+      "epoch": 0.48958127305092436,
+      "grad_norm": 0.1766699105501175,
+      "learning_rate": 1.999267315976126e-05,
+      "loss": 1.2889,
+      "step": 1315
+    },
+    {
+      "epoch": 0.48995357820153346,
+      "grad_norm": 0.17779386043548584,
+      "learning_rate": 1.9992626572026764e-05,
+      "loss": 1.2969,
+      "step": 1316
+    },
+    {
+      "epoch": 0.4903258833521425,
+      "grad_norm": 0.1781335324048996,
+      "learning_rate": 1.9992579836701796e-05,
+      "loss": 1.2939,
+      "step": 1317
+    },
+    {
+      "epoch": 0.49069818850275154,
+      "grad_norm": 0.18048399686813354,
+      "learning_rate": 1.9992532953787057e-05,
+      "loss": 1.29,
+      "step": 1318
+    },
+    {
+      "epoch": 0.49107049365336064,
+      "grad_norm": 0.18843825161457062,
+      "learning_rate": 1.999248592328323e-05,
+      "loss": 1.3121,
+      "step": 1319
+    },
+    {
+      "epoch": 0.4914427988039697,
+      "grad_norm": 0.1733110100030899,
+      "learning_rate": 1.9992438745191017e-05,
+      "loss": 1.3121,
+      "step": 1320
+    },
+    {
+      "epoch": 0.4918151039545788,
+      "grad_norm": 0.18750248849391937,
+      "learning_rate": 1.999239141951111e-05,
+      "loss": 1.298,
+      "step": 1321
+    },
+    {
+      "epoch": 0.4921874091051878,
+      "grad_norm": 0.18104903399944305,
+      "learning_rate": 1.9992343946244205e-05,
+      "loss": 1.3,
+      "step": 1322
+    },
+    {
+      "epoch": 0.4925597142557969,
+      "grad_norm": 0.1810617595911026,
+      "learning_rate": 1.9992296325391004e-05,
+      "loss": 1.3075,
+      "step": 1323
+    },
+    {
+      "epoch": 0.49293201940640596,
+      "grad_norm": 0.18038879334926605,
+      "learning_rate": 1.999224855695222e-05,
+      "loss": 1.304,
+      "step": 1324
+    },
+    {
+      "epoch": 0.49330432455701506,
+      "grad_norm": 0.17841710150241852,
+      "learning_rate": 1.999220064092855e-05,
+      "loss": 1.3075,
+      "step": 1325
+    },
+    {
+      "epoch": 0.4936766297076241,
+      "grad_norm": 0.18616177141666412,
+      "learning_rate": 1.9992152577320706e-05,
+      "loss": 1.2902,
+      "step": 1326
+    },
+    {
+      "epoch": 0.4940489348582332,
+      "grad_norm": 0.18415139615535736,
+      "learning_rate": 1.999210436612939e-05,
+      "loss": 1.3003,
+      "step": 1327
+    },
+    {
+      "epoch": 0.49442124000884224,
+      "grad_norm": 0.2058713585138321,
+      "learning_rate": 1.9992056007355323e-05,
+      "loss": 1.3074,
+      "step": 1328
+    },
+    {
+      "epoch": 0.49479354515945134,
+      "grad_norm": 0.17545197904109955,
+      "learning_rate": 1.9992007500999216e-05,
+      "loss": 1.3167,
+      "step": 1329
+    },
+    {
+      "epoch": 0.4951658503100604,
+      "grad_norm": 0.1859792321920395,
+      "learning_rate": 1.9991958847061786e-05,
+      "loss": 1.3226,
+      "step": 1330
+    },
+    {
+      "epoch": 0.4955381554606694,
+      "grad_norm": 0.2005636990070343,
+      "learning_rate": 1.999191004554375e-05,
+      "loss": 1.3016,
+      "step": 1331
+    },
+    {
+      "epoch": 0.4959104606112785,
+      "grad_norm": 0.17621280252933502,
+      "learning_rate": 1.999186109644583e-05,
+      "loss": 1.2963,
+      "step": 1332
+    },
+    {
+      "epoch": 0.49628276576188757,
+      "grad_norm": 0.18345515429973602,
+      "learning_rate": 1.9991811999768747e-05,
+      "loss": 1.2984,
+      "step": 1333
+    },
+    {
+      "epoch": 0.49665507091249667,
+      "grad_norm": 0.191512331366539,
+      "learning_rate": 1.999176275551323e-05,
+      "loss": 1.2894,
+      "step": 1334
+    },
+    {
+      "epoch": 0.4970273760631057,
+      "grad_norm": 0.18425214290618896,
+      "learning_rate": 1.9991713363680002e-05,
+      "loss": 1.2952,
+      "step": 1335
+    },
+    {
+      "epoch": 0.4973996812137148,
+      "grad_norm": 0.4703981280326843,
+      "learning_rate": 1.9991663824269797e-05,
+      "loss": 1.3056,
+      "step": 1336
+    },
+    {
+      "epoch": 0.49777198636432385,
+      "grad_norm": 0.19318270683288574,
+      "learning_rate": 1.999161413728334e-05,
+      "loss": 1.2895,
+      "step": 1337
+    },
+    {
+      "epoch": 0.49814429151493295,
+      "grad_norm": 0.19595178961753845,
+      "learning_rate": 1.9991564302721374e-05,
+      "loss": 1.2952,
+      "step": 1338
+    },
+    {
+      "epoch": 0.498516596665542,
+      "grad_norm": 0.20309020578861237,
+      "learning_rate": 1.9991514320584628e-05,
+      "loss": 1.3124,
+      "step": 1339
+    },
+    {
+      "epoch": 0.4988889018161511,
+      "grad_norm": 0.19227808713912964,
+      "learning_rate": 1.9991464190873845e-05,
+      "loss": 1.2956,
+      "step": 1340
+    },
+    {
+      "epoch": 0.49926120696676013,
+      "grad_norm": 0.19931499660015106,
+      "learning_rate": 1.999141391358976e-05,
+      "loss": 1.2961,
+      "step": 1341
+    },
+    {
+      "epoch": 0.49963351211736917,
+      "grad_norm": 0.19872671365737915,
+      "learning_rate": 1.999136348873312e-05,
+      "loss": 1.2977,
+      "step": 1342
+    },
+    {
+      "epoch": 0.5000058172679782,
+      "grad_norm": 0.1944931149482727,
+      "learning_rate": 1.999131291630467e-05,
+      "loss": 1.2856,
+      "step": 1343
+    },
+    {
+      "epoch": 0.5003781224185874,
+      "grad_norm": 0.19236011803150177,
+      "learning_rate": 1.9991262196305153e-05,
+      "loss": 1.2963,
+      "step": 1344
+    },
+    {
+      "epoch": 0.5007504275691964,
+      "grad_norm": 0.19329576194286346,
+      "learning_rate": 1.999121132873532e-05,
+      "loss": 1.3028,
+      "step": 1345
+    },
+    {
+      "epoch": 0.5011227327198055,
+      "grad_norm": 0.17316564917564392,
+      "learning_rate": 1.9991160313595924e-05,
+      "loss": 1.2826,
+      "step": 1346
+    },
+    {
+      "epoch": 0.5014950378704145,
+      "grad_norm": 0.19512668251991272,
+      "learning_rate": 1.9991109150887715e-05,
+      "loss": 1.3035,
+      "step": 1347
+    },
+    {
+      "epoch": 0.5018673430210236,
+      "grad_norm": 0.19115027785301208,
+      "learning_rate": 1.9991057840611453e-05,
+      "loss": 1.3046,
+      "step": 1348
+    },
+    {
+      "epoch": 0.5022396481716327,
+      "grad_norm": 0.19715119898319244,
+      "learning_rate": 1.9991006382767892e-05,
+      "loss": 1.3129,
+      "step": 1349
+    },
+    {
+      "epoch": 0.5026119533222417,
+      "grad_norm": 0.18349628150463104,
+      "learning_rate": 1.9990954777357795e-05,
+      "loss": 1.3005,
+      "step": 1350
+    },
+    {
+      "epoch": 0.5029842584728508,
+      "grad_norm": 0.19362014532089233,
+      "learning_rate": 1.999090302438192e-05,
+      "loss": 1.3151,
+      "step": 1351
+    },
+    {
+      "epoch": 0.5033565636234599,
+      "grad_norm": 0.18228283524513245,
+      "learning_rate": 1.999085112384104e-05,
+      "loss": 1.3068,
+      "step": 1352
+    },
+    {
+      "epoch": 0.503728868774069,
+      "grad_norm": 0.193775936961174,
+      "learning_rate": 1.9990799075735912e-05,
+      "loss": 1.3139,
+      "step": 1353
+    },
+    {
+      "epoch": 0.504101173924678,
+      "grad_norm": 0.18167784810066223,
+      "learning_rate": 1.999074688006731e-05,
+      "loss": 1.2844,
+      "step": 1354
+    },
+    {
+      "epoch": 0.5044734790752871,
+      "grad_norm": 0.19863297045230865,
+      "learning_rate": 1.9990694536836002e-05,
+      "loss": 1.2932,
+      "step": 1355
+    },
+    {
+      "epoch": 0.5048457842258961,
+      "grad_norm": 0.189836323261261,
+      "learning_rate": 1.9990642046042766e-05,
+      "loss": 1.284,
+      "step": 1356
+    },
+    {
+      "epoch": 0.5052180893765053,
+      "grad_norm": 0.19672173261642456,
+      "learning_rate": 1.9990589407688373e-05,
+      "loss": 1.3062,
+      "step": 1357
+    },
+    {
+      "epoch": 0.5055903945271143,
+      "grad_norm": 0.18417510390281677,
+      "learning_rate": 1.99905366217736e-05,
+      "loss": 1.303,
+      "step": 1358
+    },
+    {
+      "epoch": 0.5059626996777233,
+      "grad_norm": 0.18591108918190002,
+      "learning_rate": 1.999048368829923e-05,
+      "loss": 1.291,
+      "step": 1359
+    },
+    {
+      "epoch": 0.5063350048283324,
+      "grad_norm": 0.1832229644060135,
+      "learning_rate": 1.9990430607266038e-05,
+      "loss": 1.2888,
+      "step": 1360
+    },
+    {
+      "epoch": 0.5067073099789415,
+      "grad_norm": 0.21367454528808594,
+      "learning_rate": 1.999037737867482e-05,
+      "loss": 1.307,
+      "step": 1361
+    },
+    {
+      "epoch": 0.5070796151295506,
+      "grad_norm": 0.18742305040359497,
+      "learning_rate": 1.999032400252635e-05,
+      "loss": 1.2955,
+      "step": 1362
+    },
+    {
+      "epoch": 0.5074519202801596,
+      "grad_norm": 0.18082669377326965,
+      "learning_rate": 1.9990270478821422e-05,
+      "loss": 1.2794,
+      "step": 1363
+    },
+    {
+      "epoch": 0.5078242254307687,
+      "grad_norm": 0.19011190533638,
+      "learning_rate": 1.9990216807560827e-05,
+      "loss": 1.2864,
+      "step": 1364
+    },
+    {
+      "epoch": 0.5081965305813778,
+      "grad_norm": 0.18790535628795624,
+      "learning_rate": 1.9990162988745357e-05,
+      "loss": 1.2974,
+      "step": 1365
+    },
+    {
+      "epoch": 0.5085688357319869,
+      "grad_norm": 0.19116294384002686,
+      "learning_rate": 1.9990109022375807e-05,
+      "loss": 1.2972,
+      "step": 1366
+    },
+    {
+      "epoch": 0.5089411408825959,
+      "grad_norm": 0.19198502600193024,
+      "learning_rate": 1.999005490845297e-05,
+      "loss": 1.3173,
+      "step": 1367
+    },
+    {
+      "epoch": 0.5093134460332049,
+      "grad_norm": 0.19381704926490784,
+      "learning_rate": 1.9990000646977653e-05,
+      "loss": 1.3012,
+      "step": 1368
+    },
+    {
+      "epoch": 0.509685751183814,
+      "grad_norm": 0.1835579127073288,
+      "learning_rate": 1.9989946237950653e-05,
+      "loss": 1.3028,
+      "step": 1369
+    },
+    {
+      "epoch": 0.5100580563344231,
+      "grad_norm": 0.18221747875213623,
+      "learning_rate": 1.998989168137277e-05,
+      "loss": 1.2992,
+      "step": 1370
+    },
+    {
+      "epoch": 0.5104303614850322,
+      "grad_norm": 0.18950197100639343,
+      "learning_rate": 1.998983697724482e-05,
+      "loss": 1.2997,
+      "step": 1371
+    },
+    {
+      "epoch": 0.5108026666356412,
+      "grad_norm": 0.18854863941669464,
+      "learning_rate": 1.99897821255676e-05,
+      "loss": 1.3083,
+      "step": 1372
+    },
+    {
+      "epoch": 0.5111749717862503,
+      "grad_norm": 0.1852334886789322,
+      "learning_rate": 1.9989727126341927e-05,
+      "loss": 1.2923,
+      "step": 1373
+    },
+    {
+      "epoch": 0.5115472769368594,
+      "grad_norm": 0.18260884284973145,
+      "learning_rate": 1.998967197956861e-05,
+      "loss": 1.2941,
+      "step": 1374
+    },
+    {
+      "epoch": 0.5119195820874685,
+      "grad_norm": 0.1905146837234497,
+      "learning_rate": 1.9989616685248468e-05,
+      "loss": 1.3061,
+      "step": 1375
+    },
+    {
+      "epoch": 0.5122918872380775,
+      "grad_norm": 0.17799590528011322,
+      "learning_rate": 1.9989561243382313e-05,
+      "loss": 1.305,
+      "step": 1376
+    },
+    {
+      "epoch": 0.5126641923886865,
+      "grad_norm": 0.18852990865707397,
+      "learning_rate": 1.9989505653970963e-05,
+      "loss": 1.3001,
+      "step": 1377
+    },
+    {
+      "epoch": 0.5130364975392957,
+      "grad_norm": 0.19264602661132812,
+      "learning_rate": 1.9989449917015242e-05,
+      "loss": 1.2991,
+      "step": 1378
+    },
+    {
+      "epoch": 0.5134088026899047,
+      "grad_norm": 0.1937609314918518,
+      "learning_rate": 1.9989394032515974e-05,
+      "loss": 1.2826,
+      "step": 1379
+    },
+    {
+      "epoch": 0.5137811078405138,
+      "grad_norm": 0.19785352051258087,
+      "learning_rate": 1.9989338000473982e-05,
+      "loss": 1.2972,
+      "step": 1380
+    },
+    {
+      "epoch": 0.5141534129911228,
+      "grad_norm": 0.18624554574489594,
+      "learning_rate": 1.9989281820890095e-05,
+      "loss": 1.2628,
+      "step": 1381
+    },
+    {
+      "epoch": 0.5145257181417319,
+      "grad_norm": 0.18756809830665588,
+      "learning_rate": 1.9989225493765144e-05,
+      "loss": 1.2842,
+      "step": 1382
+    },
+    {
+      "epoch": 0.514898023292341,
+      "grad_norm": 0.18710888922214508,
+      "learning_rate": 1.9989169019099956e-05,
+      "loss": 1.2887,
+      "step": 1383
+    },
+    {
+      "epoch": 0.5152703284429501,
+      "grad_norm": 0.17629897594451904,
+      "learning_rate": 1.9989112396895374e-05,
+      "loss": 1.2993,
+      "step": 1384
+    },
+    {
+      "epoch": 0.5156426335935591,
+      "grad_norm": 0.18918879330158234,
+      "learning_rate": 1.9989055627152222e-05,
+      "loss": 1.304,
+      "step": 1385
+    },
+    {
+      "epoch": 0.5160149387441682,
+      "grad_norm": 0.1894582360982895,
+      "learning_rate": 1.998899870987135e-05,
+      "loss": 1.2917,
+      "step": 1386
+    },
+    {
+      "epoch": 0.5163872438947773,
+      "grad_norm": 0.18650588393211365,
+      "learning_rate": 1.9988941645053594e-05,
+      "loss": 1.2985,
+      "step": 1387
+    },
+    {
+      "epoch": 0.5167595490453863,
+      "grad_norm": 0.17856687307357788,
+      "learning_rate": 1.9988884432699795e-05,
+      "loss": 1.2898,
+      "step": 1388
+    },
+    {
+      "epoch": 0.5171318541959954,
+      "grad_norm": 0.19840949773788452,
+      "learning_rate": 1.9988827072810798e-05,
+      "loss": 1.2982,
+      "step": 1389
+    },
+    {
+      "epoch": 0.5175041593466044,
+      "grad_norm": 0.18756093084812164,
+      "learning_rate": 1.9988769565387454e-05,
+      "loss": 1.294,
+      "step": 1390
+    },
+    {
+      "epoch": 0.5178764644972135,
+      "grad_norm": 0.17636418342590332,
+      "learning_rate": 1.9988711910430613e-05,
+      "loss": 1.291,
+      "step": 1391
+    },
+    {
+      "epoch": 0.5182487696478226,
+      "grad_norm": 0.18831880390644073,
+      "learning_rate": 1.998865410794112e-05,
+      "loss": 1.2831,
+      "step": 1392
+    },
+    {
+      "epoch": 0.5186210747984317,
+      "grad_norm": 0.17440542578697205,
+      "learning_rate": 1.9988596157919836e-05,
+      "loss": 1.305,
+      "step": 1393
+    },
+    {
+      "epoch": 0.5189933799490407,
+      "grad_norm": 0.18249642848968506,
+      "learning_rate": 1.9988538060367612e-05,
+      "loss": 1.2852,
+      "step": 1394
+    },
+    {
+      "epoch": 0.5193656850996498,
+      "grad_norm": 0.18289236724376678,
+      "learning_rate": 1.9988479815285308e-05,
+      "loss": 1.2902,
+      "step": 1395
+    },
+    {
+      "epoch": 0.5197379902502589,
+      "grad_norm": 0.18215063214302063,
+      "learning_rate": 1.998842142267378e-05,
+      "loss": 1.2896,
+      "step": 1396
+    },
+    {
+      "epoch": 0.520110295400868,
+      "grad_norm": 0.18357405066490173,
+      "learning_rate": 1.99883628825339e-05,
+      "loss": 1.2959,
+      "step": 1397
+    },
+    {
+      "epoch": 0.520482600551477,
+      "grad_norm": 0.18299397826194763,
+      "learning_rate": 1.9988304194866527e-05,
+      "loss": 1.3121,
+      "step": 1398
+    },
+    {
+      "epoch": 0.520854905702086,
+      "grad_norm": 0.18076153099536896,
+      "learning_rate": 1.9988245359672523e-05,
+      "loss": 1.3005,
+      "step": 1399
+    },
+    {
+      "epoch": 0.5212272108526952,
+      "grad_norm": 0.17696769535541534,
+      "learning_rate": 1.9988186376952766e-05,
+      "loss": 1.2815,
+      "step": 1400
+    },
+    {
+      "epoch": 0.5215995160033042,
+      "grad_norm": 0.18042761087417603,
+      "learning_rate": 1.998812724670812e-05,
+      "loss": 1.2893,
+      "step": 1401
+    },
+    {
+      "epoch": 0.5219718211539133,
+      "grad_norm": 0.1830262839794159,
+      "learning_rate": 1.9988067968939463e-05,
+      "loss": 1.2929,
+      "step": 1402
+    },
+    {
+      "epoch": 0.5223441263045223,
+      "grad_norm": 0.18442308902740479,
+      "learning_rate": 1.998800854364767e-05,
+      "loss": 1.2998,
+      "step": 1403
+    },
+    {
+      "epoch": 0.5227164314551314,
+      "grad_norm": 0.1778024435043335,
+      "learning_rate": 1.9987948970833616e-05,
+      "loss": 1.2871,
+      "step": 1404
+    },
+    {
+      "epoch": 0.5230887366057405,
+      "grad_norm": 0.17833255231380463,
+      "learning_rate": 1.9987889250498185e-05,
+      "loss": 1.2935,
+      "step": 1405
+    },
+    {
+      "epoch": 0.5234610417563496,
+      "grad_norm": 0.1755494475364685,
+      "learning_rate": 1.9987829382642253e-05,
+      "loss": 1.2908,
+      "step": 1406
+    },
+    {
+      "epoch": 0.5238333469069586,
+      "grad_norm": 0.18358327448368073,
+      "learning_rate": 1.998776936726671e-05,
+      "loss": 1.2984,
+      "step": 1407
+    },
+    {
+      "epoch": 0.5242056520575676,
+      "grad_norm": 0.1703406274318695,
+      "learning_rate": 1.998770920437244e-05,
+      "loss": 1.2906,
+      "step": 1408
+    },
+    {
+      "epoch": 0.5245779572081768,
+      "grad_norm": 0.1967104971408844,
+      "learning_rate": 1.9987648893960334e-05,
+      "loss": 1.293,
+      "step": 1409
+    },
+    {
+      "epoch": 0.5249502623587858,
+      "grad_norm": 0.19022947549819946,
+      "learning_rate": 1.998758843603128e-05,
+      "loss": 1.2919,
+      "step": 1410
+    },
+    {
+      "epoch": 0.5253225675093949,
+      "grad_norm": 0.17558811604976654,
+      "learning_rate": 1.9987527830586168e-05,
+      "loss": 1.2928,
+      "step": 1411
+    },
+    {
+      "epoch": 0.5256948726600039,
+      "grad_norm": 0.1828990876674652,
+      "learning_rate": 1.99874670776259e-05,
+      "loss": 1.2967,
+      "step": 1412
+    },
+    {
+      "epoch": 0.5260671778106131,
+      "grad_norm": 0.18160808086395264,
+      "learning_rate": 1.9987406177151368e-05,
+      "loss": 1.2936,
+      "step": 1413
+    },
+    {
+      "epoch": 0.5264394829612221,
+      "grad_norm": 0.1777198314666748,
+      "learning_rate": 1.9987345129163472e-05,
+      "loss": 1.2832,
+      "step": 1414
+    },
+    {
+      "epoch": 0.5268117881118312,
+      "grad_norm": 0.1877478063106537,
+      "learning_rate": 1.998728393366312e-05,
+      "loss": 1.3125,
+      "step": 1415
+    },
+    {
+      "epoch": 0.5271840932624402,
+      "grad_norm": 0.17702911794185638,
+      "learning_rate": 1.9987222590651206e-05,
+      "loss": 1.2747,
+      "step": 1416
+    },
+    {
+      "epoch": 0.5275563984130492,
+      "grad_norm": 0.19214600324630737,
+      "learning_rate": 1.9987161100128646e-05,
+      "loss": 1.2897,
+      "step": 1417
+    },
+    {
+      "epoch": 0.5279287035636584,
+      "grad_norm": 0.18091824650764465,
+      "learning_rate": 1.9987099462096342e-05,
+      "loss": 1.29,
+      "step": 1418
+    },
+    {
+      "epoch": 0.5283010087142674,
+      "grad_norm": 0.19012725353240967,
+      "learning_rate": 1.9987037676555205e-05,
+      "loss": 1.2834,
+      "step": 1419
+    },
+    {
+      "epoch": 0.5286733138648765,
+      "grad_norm": 0.18362641334533691,
+      "learning_rate": 1.9986975743506146e-05,
+      "loss": 1.2939,
+      "step": 1420
+    },
+    {
+      "epoch": 0.5290456190154855,
+      "grad_norm": 0.17870792746543884,
+      "learning_rate": 1.9986913662950084e-05,
+      "loss": 1.2931,
+      "step": 1421
+    },
+    {
+      "epoch": 0.5294179241660947,
+      "grad_norm": 0.16860565543174744,
+      "learning_rate": 1.9986851434887934e-05,
+      "loss": 1.3015,
+      "step": 1422
+    },
+    {
+      "epoch": 0.5297902293167037,
+      "grad_norm": 0.18669748306274414,
+      "learning_rate": 1.9986789059320614e-05,
+      "loss": 1.2966,
+      "step": 1423
+    },
+    {
+      "epoch": 0.5301625344673128,
+      "grad_norm": 0.1844332367181778,
+      "learning_rate": 1.998672653624905e-05,
+      "loss": 1.2936,
+      "step": 1424
+    },
+    {
+      "epoch": 0.5305348396179218,
+      "grad_norm": 0.18035277724266052,
+      "learning_rate": 1.998666386567416e-05,
+      "loss": 1.2964,
+      "step": 1425
+    },
+    {
+      "epoch": 0.530907144768531,
+      "grad_norm": 0.19097861647605896,
+      "learning_rate": 1.998660104759687e-05,
+      "loss": 1.305,
+      "step": 1426
+    },
+    {
+      "epoch": 0.53127944991914,
+      "grad_norm": 0.17276540398597717,
+      "learning_rate": 1.998653808201811e-05,
+      "loss": 1.3028,
+      "step": 1427
+    },
+    {
+      "epoch": 0.531651755069749,
+      "grad_norm": 0.19778695702552795,
+      "learning_rate": 1.9986474968938808e-05,
+      "loss": 1.2907,
+      "step": 1428
+    },
+    {
+      "epoch": 0.5320240602203581,
+      "grad_norm": 0.18655353784561157,
+      "learning_rate": 1.99864117083599e-05,
+      "loss": 1.2774,
+      "step": 1429
+    },
+    {
+      "epoch": 0.5323963653709671,
+      "grad_norm": 0.18784590065479279,
+      "learning_rate": 1.9986348300282318e-05,
+      "loss": 1.2953,
+      "step": 1430
+    },
+    {
+      "epoch": 0.5327686705215763,
+      "grad_norm": 0.17535433173179626,
+      "learning_rate": 1.9986284744706995e-05,
+      "loss": 1.2927,
+      "step": 1431
+    },
+    {
+      "epoch": 0.5331409756721853,
+      "grad_norm": 0.1903795450925827,
+      "learning_rate": 1.9986221041634874e-05,
+      "loss": 1.2778,
+      "step": 1432
+    },
+    {
+      "epoch": 0.5335132808227944,
+      "grad_norm": 0.1824638694524765,
+      "learning_rate": 1.9986157191066897e-05,
+      "loss": 1.2905,
+      "step": 1433
+    },
+    {
+      "epoch": 0.5338855859734034,
+      "grad_norm": 0.17660009860992432,
+      "learning_rate": 1.9986093193004005e-05,
+      "loss": 1.2794,
+      "step": 1434
+    },
+    {
+      "epoch": 0.5342578911240126,
+      "grad_norm": 0.18862681090831757,
+      "learning_rate": 1.998602904744714e-05,
+      "loss": 1.2997,
+      "step": 1435
+    },
+    {
+      "epoch": 0.5346301962746216,
+      "grad_norm": 0.18798676133155823,
+      "learning_rate": 1.9985964754397256e-05,
+      "loss": 1.2973,
+      "step": 1436
+    },
+    {
+      "epoch": 0.5350025014252306,
+      "grad_norm": 0.18016012012958527,
+      "learning_rate": 1.9985900313855297e-05,
+      "loss": 1.2902,
+      "step": 1437
+    },
+    {
+      "epoch": 0.5353748065758397,
+      "grad_norm": 0.18391664326190948,
+      "learning_rate": 1.9985835725822217e-05,
+      "loss": 1.2734,
+      "step": 1438
+    },
+    {
+      "epoch": 0.5357471117264488,
+      "grad_norm": 0.2144862860441208,
+      "learning_rate": 1.998577099029897e-05,
+      "loss": 1.2937,
+      "step": 1439
+    },
+    {
+      "epoch": 0.5361194168770579,
+      "grad_norm": 0.18601621687412262,
+      "learning_rate": 1.9985706107286515e-05,
+      "loss": 1.2836,
+      "step": 1440
+    },
+    {
+      "epoch": 0.5364917220276669,
+      "grad_norm": 0.18405480682849884,
+      "learning_rate": 1.9985641076785806e-05,
+      "loss": 1.2814,
+      "step": 1441
+    },
+    {
+      "epoch": 0.536864027178276,
+      "grad_norm": 0.19651199877262115,
+      "learning_rate": 1.9985575898797803e-05,
+      "loss": 1.2892,
+      "step": 1442
+    },
+    {
+      "epoch": 0.537236332328885,
+      "grad_norm": 0.18740807473659515,
+      "learning_rate": 1.9985510573323474e-05,
+      "loss": 1.3043,
+      "step": 1443
+    },
+    {
+      "epoch": 0.5376086374794942,
+      "grad_norm": 0.17624454200267792,
+      "learning_rate": 1.998544510036378e-05,
+      "loss": 1.2874,
+      "step": 1444
+    },
+    {
+      "epoch": 0.5379809426301032,
+      "grad_norm": 0.19030825793743134,
+      "learning_rate": 1.9985379479919685e-05,
+      "loss": 1.2915,
+      "step": 1445
+    },
+    {
+      "epoch": 0.5383532477807123,
+      "grad_norm": 0.18553490936756134,
+      "learning_rate": 1.9985313711992164e-05,
+      "loss": 1.3027,
+      "step": 1446
+    },
+    {
+      "epoch": 0.5387255529313213,
+      "grad_norm": 0.17176979780197144,
+      "learning_rate": 1.998524779658219e-05,
+      "loss": 1.2953,
+      "step": 1447
+    },
+    {
+      "epoch": 0.5390978580819304,
+      "grad_norm": 0.1822216957807541,
+      "learning_rate": 1.9985181733690728e-05,
+      "loss": 1.2795,
+      "step": 1448
+    },
+    {
+      "epoch": 0.5394701632325395,
+      "grad_norm": 0.18704013526439667,
+      "learning_rate": 1.9985115523318758e-05,
+      "loss": 1.2843,
+      "step": 1449
+    },
+    {
+      "epoch": 0.5398424683831485,
+      "grad_norm": 0.18013562262058258,
+      "learning_rate": 1.998504916546726e-05,
+      "loss": 1.2947,
+      "step": 1450
+    },
+    {
+      "epoch": 0.5402147735337576,
+      "grad_norm": 0.1960912048816681,
+      "learning_rate": 1.998498266013721e-05,
+      "loss": 1.3077,
+      "step": 1451
+    },
+    {
+      "epoch": 0.5405870786843667,
+      "grad_norm": 0.18962568044662476,
+      "learning_rate": 1.9984916007329596e-05,
+      "loss": 1.2998,
+      "step": 1452
+    },
+    {
+      "epoch": 0.5409593838349758,
+      "grad_norm": 0.17837461829185486,
+      "learning_rate": 1.99848492070454e-05,
+      "loss": 1.2742,
+      "step": 1453
+    },
+    {
+      "epoch": 0.5413316889855848,
+      "grad_norm": 0.17769278585910797,
+      "learning_rate": 1.9984782259285604e-05,
+      "loss": 1.29,
+      "step": 1454
+    },
+    {
+      "epoch": 0.5417039941361939,
+      "grad_norm": 0.19605344533920288,
+      "learning_rate": 1.9984715164051203e-05,
+      "loss": 1.2954,
+      "step": 1455
+    },
+    {
+      "epoch": 0.5420762992868029,
+      "grad_norm": 0.18324050307273865,
+      "learning_rate": 1.9984647921343185e-05,
+      "loss": 1.3072,
+      "step": 1456
+    },
+    {
+      "epoch": 0.542448604437412,
+      "grad_norm": 0.18658418953418732,
+      "learning_rate": 1.9984580531162544e-05,
+      "loss": 1.3097,
+      "step": 1457
+    },
+    {
+      "epoch": 0.5428209095880211,
+      "grad_norm": 0.1968865692615509,
+      "learning_rate": 1.9984512993510275e-05,
+      "loss": 1.2943,
+      "step": 1458
+    },
+    {
+      "epoch": 0.5431932147386301,
+      "grad_norm": 0.19187164306640625,
+      "learning_rate": 1.9984445308387377e-05,
+      "loss": 1.29,
+      "step": 1459
+    },
+    {
+      "epoch": 0.5435655198892392,
+      "grad_norm": 0.18562346696853638,
+      "learning_rate": 1.9984377475794847e-05,
+      "loss": 1.2895,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5439378250398483,
+      "grad_norm": 0.18785440921783447,
+      "learning_rate": 1.998430949573369e-05,
+      "loss": 1.3003,
+      "step": 1461
+    },
+    {
+      "epoch": 0.5443101301904574,
+      "grad_norm": 0.17857135832309723,
+      "learning_rate": 1.9984241368204907e-05,
+      "loss": 1.2806,
+      "step": 1462
+    },
+    {
+      "epoch": 0.5446824353410664,
+      "grad_norm": 0.18769045174121857,
+      "learning_rate": 1.99841730932095e-05,
+      "loss": 1.2866,
+      "step": 1463
+    },
+    {
+      "epoch": 0.5450547404916755,
+      "grad_norm": 0.1790071427822113,
+      "learning_rate": 1.9984104670748493e-05,
+      "loss": 1.2814,
+      "step": 1464
+    },
+    {
+      "epoch": 0.5454270456422845,
+      "grad_norm": 0.19664452970027924,
+      "learning_rate": 1.998403610082288e-05,
+      "loss": 1.2965,
+      "step": 1465
+    },
+    {
+      "epoch": 0.5457993507928937,
+      "grad_norm": 0.19964231550693512,
+      "learning_rate": 1.9983967383433685e-05,
+      "loss": 1.2875,
+      "step": 1466
+    },
+    {
+      "epoch": 0.5461716559435027,
+      "grad_norm": 0.1774752289056778,
+      "learning_rate": 1.9983898518581913e-05,
+      "loss": 1.2741,
+      "step": 1467
+    },
+    {
+      "epoch": 0.5465439610941117,
+      "grad_norm": 0.18312522768974304,
+      "learning_rate": 1.998382950626859e-05,
+      "loss": 1.2932,
+      "step": 1468
+    },
+    {
+      "epoch": 0.5469162662447208,
+      "grad_norm": 0.19389499723911285,
+      "learning_rate": 1.998376034649473e-05,
+      "loss": 1.2921,
+      "step": 1469
+    },
+    {
+      "epoch": 0.5472885713953299,
+      "grad_norm": 0.22545579075813293,
+      "learning_rate": 1.9983691039261358e-05,
+      "loss": 1.3011,
+      "step": 1470
+    },
+    {
+      "epoch": 0.547660876545939,
+      "grad_norm": 0.17456892132759094,
+      "learning_rate": 1.9983621584569496e-05,
+      "loss": 1.2762,
+      "step": 1471
+    },
+    {
+      "epoch": 0.548033181696548,
+      "grad_norm": 0.20221838355064392,
+      "learning_rate": 1.9983551982420168e-05,
+      "loss": 1.2822,
+      "step": 1472
+    },
+    {
+      "epoch": 0.5484054868471571,
+      "grad_norm": 0.17790015041828156,
+      "learning_rate": 1.9983482232814405e-05,
+      "loss": 1.2854,
+      "step": 1473
+    },
+    {
+      "epoch": 0.5487777919977662,
+      "grad_norm": 0.19563356041908264,
+      "learning_rate": 1.9983412335753237e-05,
+      "loss": 1.2901,
+      "step": 1474
+    },
+    {
+      "epoch": 0.5491500971483753,
+      "grad_norm": 0.17833010852336884,
+      "learning_rate": 1.9983342291237693e-05,
+      "loss": 1.2993,
+      "step": 1475
+    },
+    {
+      "epoch": 0.5495224022989843,
+      "grad_norm": 0.18136514723300934,
+      "learning_rate": 1.998327209926881e-05,
+      "loss": 1.2829,
+      "step": 1476
+    },
+    {
+      "epoch": 0.5498947074495933,
+      "grad_norm": 0.19292525947093964,
+      "learning_rate": 1.9983201759847627e-05,
+      "loss": 1.2967,
+      "step": 1477
+    },
+    {
+      "epoch": 0.5502670126002024,
+      "grad_norm": 0.1802646517753601,
+      "learning_rate": 1.9983131272975178e-05,
+      "loss": 1.2871,
+      "step": 1478
+    },
+    {
+      "epoch": 0.5506393177508115,
+      "grad_norm": 0.18988975882530212,
+      "learning_rate": 1.9983060638652507e-05,
+      "loss": 1.2881,
+      "step": 1479
+    },
+    {
+      "epoch": 0.5510116229014206,
+      "grad_norm": 0.18343466520309448,
+      "learning_rate": 1.9982989856880655e-05,
+      "loss": 1.2826,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5513839280520296,
+      "grad_norm": 0.18973736464977264,
+      "learning_rate": 1.9982918927660676e-05,
+      "loss": 1.277,
+      "step": 1481
+    },
+    {
+      "epoch": 0.5517562332026387,
+      "grad_norm": 0.1893860101699829,
+      "learning_rate": 1.9982847850993605e-05,
+      "loss": 1.2789,
+      "step": 1482
+    },
+    {
+      "epoch": 0.5521285383532478,
+      "grad_norm": 0.17867526412010193,
+      "learning_rate": 1.9982776626880498e-05,
+      "loss": 1.2916,
+      "step": 1483
+    },
+    {
+      "epoch": 0.5525008435038569,
+      "grad_norm": 0.1821223944425583,
+      "learning_rate": 1.998270525532241e-05,
+      "loss": 1.2678,
+      "step": 1484
+    },
+    {
+      "epoch": 0.5528731486544659,
+      "grad_norm": 0.18600653111934662,
+      "learning_rate": 1.998263373632039e-05,
+      "loss": 1.2878,
+      "step": 1485
+    },
+    {
+      "epoch": 0.553245453805075,
+      "grad_norm": 0.17727398872375488,
+      "learning_rate": 1.9982562069875495e-05,
+      "loss": 1.293,
+      "step": 1486
+    },
+    {
+      "epoch": 0.5536177589556841,
+      "grad_norm": 0.1803734004497528,
+      "learning_rate": 1.9982490255988786e-05,
+      "loss": 1.2873,
+      "step": 1487
+    },
+    {
+      "epoch": 0.5539900641062931,
+      "grad_norm": 0.1834772229194641,
+      "learning_rate": 1.9982418294661322e-05,
+      "loss": 1.2884,
+      "step": 1488
+    },
+    {
+      "epoch": 0.5543623692569022,
+      "grad_norm": 0.1872669756412506,
+      "learning_rate": 1.998234618589417e-05,
+      "loss": 1.2798,
+      "step": 1489
+    },
+    {
+      "epoch": 0.5547346744075112,
+      "grad_norm": 0.18299680948257446,
+      "learning_rate": 1.9982273929688384e-05,
+      "loss": 1.3008,
+      "step": 1490
+    },
+    {
+      "epoch": 0.5551069795581203,
+      "grad_norm": 0.1841609627008438,
+      "learning_rate": 1.9982201526045044e-05,
+      "loss": 1.2893,
+      "step": 1491
+    },
+    {
+      "epoch": 0.5554792847087294,
+      "grad_norm": 0.19560469686985016,
+      "learning_rate": 1.9982128974965215e-05,
+      "loss": 1.2846,
+      "step": 1492
+    },
+    {
+      "epoch": 0.5558515898593385,
+      "grad_norm": 0.17704376578330994,
+      "learning_rate": 1.998205627644996e-05,
+      "loss": 1.2972,
+      "step": 1493
+    },
+    {
+      "epoch": 0.5562238950099475,
+      "grad_norm": 0.18441899120807648,
+      "learning_rate": 1.9981983430500368e-05,
+      "loss": 1.3004,
+      "step": 1494
+    },
+    {
+      "epoch": 0.5565962001605566,
+      "grad_norm": 0.19001679122447968,
+      "learning_rate": 1.9981910437117502e-05,
+      "loss": 1.2903,
+      "step": 1495
+    },
+    {
+      "epoch": 0.5569685053111657,
+      "grad_norm": 0.19117416441440582,
+      "learning_rate": 1.998183729630245e-05,
+      "loss": 1.287,
+      "step": 1496
+    },
+    {
+      "epoch": 0.5573408104617747,
+      "grad_norm": 0.1872899830341339,
+      "learning_rate": 1.9981764008056283e-05,
+      "loss": 1.291,
+      "step": 1497
+    },
+    {
+      "epoch": 0.5577131156123838,
+      "grad_norm": 0.19934000074863434,
+      "learning_rate": 1.998169057238009e-05,
+      "loss": 1.2897,
+      "step": 1498
+    },
+    {
+      "epoch": 0.5580854207629928,
+      "grad_norm": 0.1839592605829239,
+      "learning_rate": 1.9981616989274955e-05,
+      "loss": 1.2772,
+      "step": 1499
+    },
+    {
+      "epoch": 0.558457725913602,
+      "grad_norm": 0.18648071587085724,
+      "learning_rate": 1.998154325874196e-05,
+      "loss": 1.291,
+      "step": 1500
+    },
+    {
+      "epoch": 0.558457725913602,
+      "eval_loss": 1.3492937088012695,
+      "eval_runtime": 16.1571,
+      "eval_samples_per_second": 107.321,
+      "eval_steps_per_second": 5.385,
+      "step": 1500
+    },
+    {
+      "epoch": 0.558830031064211,
+      "grad_norm": 0.18601854145526886,
+      "learning_rate": 1.9981469380782205e-05,
+      "loss": 1.2898,
+      "step": 1501
+    },
+    {
+      "epoch": 0.5592023362148201,
+      "grad_norm": 0.18247583508491516,
+      "learning_rate": 1.9981395355396764e-05,
+      "loss": 1.276,
+      "step": 1502
+    },
+    {
+      "epoch": 0.5595746413654291,
+      "grad_norm": 0.19537685811519623,
+      "learning_rate": 1.9981321182586746e-05,
+      "loss": 1.2851,
+      "step": 1503
+    },
+    {
+      "epoch": 0.5599469465160382,
+      "grad_norm": 0.17951083183288574,
+      "learning_rate": 1.998124686235324e-05,
+      "loss": 1.2892,
+      "step": 1504
+    },
+    {
+      "epoch": 0.5603192516666473,
+      "grad_norm": 0.1824089139699936,
+      "learning_rate": 1.998117239469734e-05,
+      "loss": 1.285,
+      "step": 1505
+    },
+    {
+      "epoch": 0.5606915568172564,
+      "grad_norm": 0.19962851703166962,
+      "learning_rate": 1.9981097779620156e-05,
+      "loss": 1.2866,
+      "step": 1506
+    },
+    {
+      "epoch": 0.5610638619678654,
+      "grad_norm": 0.17792317271232605,
+      "learning_rate": 1.998102301712278e-05,
+      "loss": 1.281,
+      "step": 1507
+    },
+    {
+      "epoch": 0.5614361671184744,
+      "grad_norm": 0.20243586599826813,
+      "learning_rate": 1.9980948107206323e-05,
+      "loss": 1.3045,
+      "step": 1508
+    },
+    {
+      "epoch": 0.5618084722690836,
+      "grad_norm": 0.1958632916212082,
+      "learning_rate": 1.998087304987189e-05,
+      "loss": 1.2902,
+      "step": 1509
+    },
+    {
+      "epoch": 0.5621807774196926,
+      "grad_norm": 0.17324163019657135,
+      "learning_rate": 1.9980797845120583e-05,
+      "loss": 1.2868,
+      "step": 1510
+    },
+    {
+      "epoch": 0.5625530825703017,
+      "grad_norm": 0.18130120635032654,
+      "learning_rate": 1.998072249295352e-05,
+      "loss": 1.289,
+      "step": 1511
+    },
+    {
+      "epoch": 0.5629253877209107,
+      "grad_norm": 0.18327681720256805,
+      "learning_rate": 1.9980646993371816e-05,
+      "loss": 1.269,
+      "step": 1512
+    },
+    {
+      "epoch": 0.5632976928715199,
+      "grad_norm": 0.1806684136390686,
+      "learning_rate": 1.998057134637658e-05,
+      "loss": 1.2989,
+      "step": 1513
+    },
+    {
+      "epoch": 0.5636699980221289,
+      "grad_norm": 0.18309220671653748,
+      "learning_rate": 1.998049555196893e-05,
+      "loss": 1.287,
+      "step": 1514
+    },
+    {
+      "epoch": 0.564042303172738,
+      "grad_norm": 0.19237829744815826,
+      "learning_rate": 1.998041961014999e-05,
+      "loss": 1.2894,
+      "step": 1515
+    },
+    {
+      "epoch": 0.564414608323347,
+      "grad_norm": 0.1768263727426529,
+      "learning_rate": 1.998034352092088e-05,
+      "loss": 1.2811,
+      "step": 1516
+    },
+    {
+      "epoch": 0.564786913473956,
+      "grad_norm": 0.17326796054840088,
+      "learning_rate": 1.9980267284282718e-05,
+      "loss": 1.2783,
+      "step": 1517
+    },
+    {
+      "epoch": 0.5651592186245652,
+      "grad_norm": 0.18072554469108582,
+      "learning_rate": 1.9980190900236637e-05,
+      "loss": 1.2742,
+      "step": 1518
+    },
+    {
+      "epoch": 0.5655315237751742,
+      "grad_norm": 0.17865929007530212,
+      "learning_rate": 1.998011436878376e-05,
+      "loss": 1.281,
+      "step": 1519
+    },
+    {
+      "epoch": 0.5659038289257833,
+      "grad_norm": 0.18478450179100037,
+      "learning_rate": 1.998003768992522e-05,
+      "loss": 1.2918,
+      "step": 1520
+    },
+    {
+      "epoch": 0.5662761340763923,
+      "grad_norm": 0.18366698920726776,
+      "learning_rate": 1.9979960863662155e-05,
+      "loss": 1.2721,
+      "step": 1521
+    },
+    {
+      "epoch": 0.5666484392270015,
+      "grad_norm": 0.17557835578918457,
+      "learning_rate": 1.997988388999569e-05,
+      "loss": 1.2887,
+      "step": 1522
+    },
+    {
+      "epoch": 0.5670207443776105,
+      "grad_norm": 0.19200293719768524,
+      "learning_rate": 1.997980676892697e-05,
+      "loss": 1.2778,
+      "step": 1523
+    },
+    {
+      "epoch": 0.5673930495282196,
+      "grad_norm": 0.18071214854717255,
+      "learning_rate": 1.9979729500457125e-05,
+      "loss": 1.2984,
+      "step": 1524
+    },
+    {
+      "epoch": 0.5677653546788286,
+      "grad_norm": 0.18595604598522186,
+      "learning_rate": 1.9979652084587305e-05,
+      "loss": 1.2879,
+      "step": 1525
+    },
+    {
+      "epoch": 0.5681376598294378,
+      "grad_norm": 0.17925933003425598,
+      "learning_rate": 1.9979574521318648e-05,
+      "loss": 1.2851,
+      "step": 1526
+    },
+    {
+      "epoch": 0.5685099649800468,
+      "grad_norm": 0.1706983745098114,
+      "learning_rate": 1.9979496810652303e-05,
+      "loss": 1.2867,
+      "step": 1527
+    },
+    {
+      "epoch": 0.5688822701306558,
+      "grad_norm": 0.1855979859828949,
+      "learning_rate": 1.9979418952589417e-05,
+      "loss": 1.3028,
+      "step": 1528
+    },
+    {
+      "epoch": 0.5692545752812649,
+      "grad_norm": 0.1966056227684021,
+      "learning_rate": 1.997934094713114e-05,
+      "loss": 1.2808,
+      "step": 1529
+    },
+    {
+      "epoch": 0.5696268804318739,
+      "grad_norm": 0.1873730570077896,
+      "learning_rate": 1.997926279427862e-05,
+      "loss": 1.2833,
+      "step": 1530
+    },
+    {
+      "epoch": 0.5699991855824831,
+      "grad_norm": 0.19508777558803558,
+      "learning_rate": 1.9979184494033016e-05,
+      "loss": 1.2821,
+      "step": 1531
+    },
+    {
+      "epoch": 0.5703714907330921,
+      "grad_norm": 0.17242860794067383,
+      "learning_rate": 1.9979106046395487e-05,
+      "loss": 1.2833,
+      "step": 1532
+    },
+    {
+      "epoch": 0.5707437958837012,
+      "grad_norm": 0.1809634268283844,
+      "learning_rate": 1.9979027451367185e-05,
+      "loss": 1.2796,
+      "step": 1533
+    },
+    {
+      "epoch": 0.5711161010343102,
+      "grad_norm": 0.17092421650886536,
+      "learning_rate": 1.9978948708949274e-05,
+      "loss": 1.2873,
+      "step": 1534
+    },
+    {
+      "epoch": 0.5714884061849194,
+      "grad_norm": 0.19138309359550476,
+      "learning_rate": 1.9978869819142915e-05,
+      "loss": 1.2923,
+      "step": 1535
+    },
+    {
+      "epoch": 0.5718607113355284,
+      "grad_norm": 0.18097564578056335,
+      "learning_rate": 1.997879078194928e-05,
+      "loss": 1.2754,
+      "step": 1536
+    },
+    {
+      "epoch": 0.5722330164861374,
+      "grad_norm": 0.1787206530570984,
+      "learning_rate": 1.9978711597369528e-05,
+      "loss": 1.2976,
+      "step": 1537
+    },
+    {
+      "epoch": 0.5726053216367465,
+      "grad_norm": 0.1908443570137024,
+      "learning_rate": 1.997863226540483e-05,
+      "loss": 1.2946,
+      "step": 1538
+    },
+    {
+      "epoch": 0.5729776267873555,
+      "grad_norm": 0.1926680952310562,
+      "learning_rate": 1.9978552786056364e-05,
+      "loss": 1.2935,
+      "step": 1539
+    },
+    {
+      "epoch": 0.5733499319379647,
+      "grad_norm": 0.18011966347694397,
+      "learning_rate": 1.9978473159325296e-05,
+      "loss": 1.2915,
+      "step": 1540
+    },
+    {
+      "epoch": 0.5737222370885737,
+      "grad_norm": 0.17944326996803284,
+      "learning_rate": 1.997839338521281e-05,
+      "loss": 1.2818,
+      "step": 1541
+    },
+    {
+      "epoch": 0.5740945422391828,
+      "grad_norm": 0.1845857799053192,
+      "learning_rate": 1.9978313463720073e-05,
+      "loss": 1.2721,
+      "step": 1542
+    },
+    {
+      "epoch": 0.5744668473897918,
+      "grad_norm": 0.18156015872955322,
+      "learning_rate": 1.997823339484828e-05,
+      "loss": 1.2868,
+      "step": 1543
+    },
+    {
+      "epoch": 0.574839152540401,
+      "grad_norm": 0.18522031605243683,
+      "learning_rate": 1.99781531785986e-05,
+      "loss": 1.2795,
+      "step": 1544
+    },
+    {
+      "epoch": 0.57521145769101,
+      "grad_norm": 0.20020869374275208,
+      "learning_rate": 1.9978072814972226e-05,
+      "loss": 1.3067,
+      "step": 1545
+    },
+    {
+      "epoch": 0.575583762841619,
+      "grad_norm": 0.17684873938560486,
+      "learning_rate": 1.9977992303970342e-05,
+      "loss": 1.2903,
+      "step": 1546
+    },
+    {
+      "epoch": 0.5759560679922281,
+      "grad_norm": 0.18267862498760223,
+      "learning_rate": 1.997791164559414e-05,
+      "loss": 1.2972,
+      "step": 1547
+    },
+    {
+      "epoch": 0.5763283731428372,
+      "grad_norm": 0.17601077258586884,
+      "learning_rate": 1.9977830839844808e-05,
+      "loss": 1.278,
+      "step": 1548
+    },
+    {
+      "epoch": 0.5767006782934463,
+      "grad_norm": 0.18240796029567719,
+      "learning_rate": 1.997774988672354e-05,
+      "loss": 1.2753,
+      "step": 1549
+    },
+    {
+      "epoch": 0.5770729834440553,
+      "grad_norm": 0.19379922747612,
+      "learning_rate": 1.9977668786231536e-05,
+      "loss": 1.2932,
+      "step": 1550
+    },
+    {
+      "epoch": 0.5774452885946644,
+      "grad_norm": 0.19803746044635773,
+      "learning_rate": 1.9977587538369985e-05,
+      "loss": 1.2876,
+      "step": 1551
+    },
+    {
+      "epoch": 0.5778175937452734,
+      "grad_norm": 0.19004493951797485,
+      "learning_rate": 1.9977506143140094e-05,
+      "loss": 1.2687,
+      "step": 1552
+    },
+    {
+      "epoch": 0.5781898988958826,
+      "grad_norm": 0.18594302237033844,
+      "learning_rate": 1.9977424600543065e-05,
+      "loss": 1.2875,
+      "step": 1553
+    },
+    {
+      "epoch": 0.5785622040464916,
+      "grad_norm": 0.193922758102417,
+      "learning_rate": 1.9977342910580097e-05,
+      "loss": 1.2779,
+      "step": 1554
+    },
+    {
+      "epoch": 0.5789345091971007,
+      "grad_norm": 0.19569186866283417,
+      "learning_rate": 1.9977261073252405e-05,
+      "loss": 1.2926,
+      "step": 1555
+    },
+    {
+      "epoch": 0.5793068143477097,
+      "grad_norm": 0.18837492167949677,
+      "learning_rate": 1.9977179088561193e-05,
+      "loss": 1.2963,
+      "step": 1556
+    },
+    {
+      "epoch": 0.5796791194983189,
+      "grad_norm": 0.17669405043125153,
+      "learning_rate": 1.9977096956507668e-05,
+      "loss": 1.2761,
+      "step": 1557
+    },
+    {
+      "epoch": 0.5800514246489279,
+      "grad_norm": 0.191486656665802,
+      "learning_rate": 1.997701467709305e-05,
+      "loss": 1.2913,
+      "step": 1558
+    },
+    {
+      "epoch": 0.5804237297995369,
+      "grad_norm": 0.18390990793704987,
+      "learning_rate": 1.997693225031855e-05,
+      "loss": 1.2948,
+      "step": 1559
+    },
+    {
+      "epoch": 0.580796034950146,
+      "grad_norm": 0.18185609579086304,
+      "learning_rate": 1.9976849676185384e-05,
+      "loss": 1.283,
+      "step": 1560
+    },
+    {
+      "epoch": 0.5811683401007551,
+      "grad_norm": 0.1948121190071106,
+      "learning_rate": 1.997676695469478e-05,
+      "loss": 1.2848,
+      "step": 1561
+    },
+    {
+      "epoch": 0.5815406452513642,
+      "grad_norm": 0.18750061094760895,
+      "learning_rate": 1.997668408584795e-05,
+      "loss": 1.2828,
+      "step": 1562
+    },
+    {
+      "epoch": 0.5819129504019732,
+      "grad_norm": 0.1849110871553421,
+      "learning_rate": 1.9976601069646126e-05,
+      "loss": 1.2883,
+      "step": 1563
+    },
+    {
+      "epoch": 0.5822852555525823,
+      "grad_norm": 0.18599045276641846,
+      "learning_rate": 1.9976517906090528e-05,
+      "loss": 1.2742,
+      "step": 1564
+    },
+    {
+      "epoch": 0.5826575607031913,
+      "grad_norm": 0.1824532300233841,
+      "learning_rate": 1.997643459518239e-05,
+      "loss": 1.2835,
+      "step": 1565
+    },
+    {
+      "epoch": 0.5830298658538005,
+      "grad_norm": 0.18060551583766937,
+      "learning_rate": 1.9976351136922934e-05,
+      "loss": 1.2756,
+      "step": 1566
+    },
+    {
+      "epoch": 0.5834021710044095,
+      "grad_norm": 0.18784838914871216,
+      "learning_rate": 1.99762675313134e-05,
+      "loss": 1.2935,
+      "step": 1567
+    },
+    {
+      "epoch": 0.5837744761550185,
+      "grad_norm": 0.25187990069389343,
+      "learning_rate": 1.9976183778355018e-05,
+      "loss": 1.2882,
+      "step": 1568
+    },
+    {
+      "epoch": 0.5841467813056276,
+      "grad_norm": 0.17487937211990356,
+      "learning_rate": 1.997609987804903e-05,
+      "loss": 1.2734,
+      "step": 1569
+    },
+    {
+      "epoch": 0.5845190864562367,
+      "grad_norm": 0.18456514179706573,
+      "learning_rate": 1.9976015830396676e-05,
+      "loss": 1.2816,
+      "step": 1570
+    },
+    {
+      "epoch": 0.5848913916068458,
+      "grad_norm": 0.18243633210659027,
+      "learning_rate": 1.997593163539919e-05,
+      "loss": 1.2809,
+      "step": 1571
+    },
+    {
+      "epoch": 0.5852636967574548,
+      "grad_norm": 0.17159947752952576,
+      "learning_rate": 1.9975847293057822e-05,
+      "loss": 1.274,
+      "step": 1572
+    },
+    {
+      "epoch": 0.5856360019080639,
+      "grad_norm": 0.18149283528327942,
+      "learning_rate": 1.9975762803373815e-05,
+      "loss": 1.2842,
+      "step": 1573
+    },
+    {
+      "epoch": 0.586008307058673,
+      "grad_norm": 0.19415758550167084,
+      "learning_rate": 1.9975678166348417e-05,
+      "loss": 1.2903,
+      "step": 1574
+    },
+    {
+      "epoch": 0.5863806122092821,
+      "grad_norm": 0.18272031843662262,
+      "learning_rate": 1.9975593381982877e-05,
+      "loss": 1.2931,
+      "step": 1575
+    },
+    {
+      "epoch": 0.5867529173598911,
+      "grad_norm": 0.19155685603618622,
+      "learning_rate": 1.997550845027845e-05,
+      "loss": 1.2876,
+      "step": 1576
+    },
+    {
+      "epoch": 0.5871252225105001,
+      "grad_norm": 0.17365863919258118,
+      "learning_rate": 1.9975423371236392e-05,
+      "loss": 1.2757,
+      "step": 1577
+    },
+    {
+      "epoch": 0.5874975276611092,
+      "grad_norm": 0.1818249225616455,
+      "learning_rate": 1.9975338144857954e-05,
+      "loss": 1.2853,
+      "step": 1578
+    },
+    {
+      "epoch": 0.5878698328117183,
+      "grad_norm": 0.17870758473873138,
+      "learning_rate": 1.99752527711444e-05,
+      "loss": 1.2682,
+      "step": 1579
+    },
+    {
+      "epoch": 0.5882421379623274,
+      "grad_norm": 0.18116502463817596,
+      "learning_rate": 1.9975167250096985e-05,
+      "loss": 1.2774,
+      "step": 1580
+    },
+    {
+      "epoch": 0.5886144431129364,
+      "grad_norm": 0.17226383090019226,
+      "learning_rate": 1.997508158171698e-05,
+      "loss": 1.2666,
+      "step": 1581
+    },
+    {
+      "epoch": 0.5889867482635455,
+      "grad_norm": 0.18359726667404175,
+      "learning_rate": 1.9974995766005644e-05,
+      "loss": 1.2794,
+      "step": 1582
+    },
+    {
+      "epoch": 0.5893590534141546,
+      "grad_norm": 0.21769051253795624,
+      "learning_rate": 1.9974909802964244e-05,
+      "loss": 1.2836,
+      "step": 1583
+    },
+    {
+      "epoch": 0.5897313585647637,
+      "grad_norm": 0.18334870040416718,
+      "learning_rate": 1.9974823692594054e-05,
+      "loss": 1.2834,
+      "step": 1584
+    },
+    {
+      "epoch": 0.5901036637153727,
+      "grad_norm": 0.18716122210025787,
+      "learning_rate": 1.9974737434896346e-05,
+      "loss": 1.2893,
+      "step": 1585
+    },
+    {
+      "epoch": 0.5904759688659817,
+      "grad_norm": 0.21679291129112244,
+      "learning_rate": 1.997465102987239e-05,
+      "loss": 1.264,
+      "step": 1586
+    },
+    {
+      "epoch": 0.5908482740165909,
+      "grad_norm": 0.1788294017314911,
+      "learning_rate": 1.9974564477523462e-05,
+      "loss": 1.2765,
+      "step": 1587
+    },
+    {
+      "epoch": 0.5912205791671999,
+      "grad_norm": 0.18102407455444336,
+      "learning_rate": 1.9974477777850847e-05,
+      "loss": 1.2636,
+      "step": 1588
+    },
+    {
+      "epoch": 0.591592884317809,
+      "grad_norm": 0.17994816601276398,
+      "learning_rate": 1.997439093085582e-05,
+      "loss": 1.2746,
+      "step": 1589
+    },
+    {
+      "epoch": 0.591965189468418,
+      "grad_norm": 0.18664704263210297,
+      "learning_rate": 1.9974303936539665e-05,
+      "loss": 1.2926,
+      "step": 1590
+    },
+    {
+      "epoch": 0.5923374946190271,
+      "grad_norm": 0.18071360886096954,
+      "learning_rate": 1.9974216794903666e-05,
+      "loss": 1.2776,
+      "step": 1591
+    },
+    {
+      "epoch": 0.5927097997696362,
+      "grad_norm": 0.19198505580425262,
+      "learning_rate": 1.9974129505949112e-05,
+      "loss": 1.2895,
+      "step": 1592
+    },
+    {
+      "epoch": 0.5930821049202453,
+      "grad_norm": 0.1870136260986328,
+      "learning_rate": 1.997404206967729e-05,
+      "loss": 1.2733,
+      "step": 1593
+    },
+    {
+      "epoch": 0.5934544100708543,
+      "grad_norm": 0.18357659876346588,
+      "learning_rate": 1.9973954486089494e-05,
+      "loss": 1.2859,
+      "step": 1594
+    },
+    {
+      "epoch": 0.5938267152214634,
+      "grad_norm": 0.20039206743240356,
+      "learning_rate": 1.9973866755187012e-05,
+      "loss": 1.2969,
+      "step": 1595
+    },
+    {
+      "epoch": 0.5941990203720725,
+      "grad_norm": 0.1847948431968689,
+      "learning_rate": 1.997377887697115e-05,
+      "loss": 1.2811,
+      "step": 1596
+    },
+    {
+      "epoch": 0.5945713255226815,
+      "grad_norm": 0.18077334761619568,
+      "learning_rate": 1.9973690851443198e-05,
+      "loss": 1.2874,
+      "step": 1597
+    },
+    {
+      "epoch": 0.5949436306732906,
+      "grad_norm": 0.17171575129032135,
+      "learning_rate": 1.9973602678604454e-05,
+      "loss": 1.2942,
+      "step": 1598
+    },
+    {
+      "epoch": 0.5953159358238996,
+      "grad_norm": 0.17874233424663544,
+      "learning_rate": 1.9973514358456228e-05,
+      "loss": 1.2938,
+      "step": 1599
+    },
+    {
+      "epoch": 0.5956882409745088,
+      "grad_norm": 0.19295063614845276,
+      "learning_rate": 1.997342589099982e-05,
+      "loss": 1.3016,
+      "step": 1600
+    },
+    {
+      "epoch": 0.5960605461251178,
+      "grad_norm": 0.18213194608688354,
+      "learning_rate": 1.9973337276236538e-05,
+      "loss": 1.2787,
+      "step": 1601
+    },
+    {
+      "epoch": 0.5964328512757269,
+      "grad_norm": 0.18399560451507568,
+      "learning_rate": 1.997324851416769e-05,
+      "loss": 1.2824,
+      "step": 1602
+    },
+    {
+      "epoch": 0.5968051564263359,
+      "grad_norm": 0.184067040681839,
+      "learning_rate": 1.9973159604794587e-05,
+      "loss": 1.2785,
+      "step": 1603
+    },
+    {
+      "epoch": 0.597177461576945,
+      "grad_norm": 0.18284447491168976,
+      "learning_rate": 1.9973070548118545e-05,
+      "loss": 1.261,
+      "step": 1604
+    },
+    {
+      "epoch": 0.5975497667275541,
+      "grad_norm": 0.18606549501419067,
+      "learning_rate": 1.9972981344140875e-05,
+      "loss": 1.2841,
+      "step": 1605
+    },
+    {
+      "epoch": 0.5979220718781632,
+      "grad_norm": 0.18835964798927307,
+      "learning_rate": 1.9972891992862895e-05,
+      "loss": 1.28,
+      "step": 1606
+    },
+    {
+      "epoch": 0.5982943770287722,
+      "grad_norm": 0.18669797480106354,
+      "learning_rate": 1.997280249428593e-05,
+      "loss": 1.2893,
+      "step": 1607
+    },
+    {
+      "epoch": 0.5986666821793812,
+      "grad_norm": 0.19121499359607697,
+      "learning_rate": 1.9972712848411292e-05,
+      "loss": 1.2847,
+      "step": 1608
+    },
+    {
+      "epoch": 0.5990389873299904,
+      "grad_norm": 0.18842531740665436,
+      "learning_rate": 1.9972623055240316e-05,
+      "loss": 1.2856,
+      "step": 1609
+    },
+    {
+      "epoch": 0.5994112924805994,
+      "grad_norm": 0.18126575648784637,
+      "learning_rate": 1.9972533114774322e-05,
+      "loss": 1.2767,
+      "step": 1610
+    },
+    {
+      "epoch": 0.5997835976312085,
+      "grad_norm": 0.17144346237182617,
+      "learning_rate": 1.997244302701464e-05,
+      "loss": 1.2738,
+      "step": 1611
+    },
+    {
+      "epoch": 0.6001559027818175,
+      "grad_norm": 0.17671994864940643,
+      "learning_rate": 1.99723527919626e-05,
+      "loss": 1.2693,
+      "step": 1612
+    },
+    {
+      "epoch": 0.6005282079324266,
+      "grad_norm": 0.17835721373558044,
+      "learning_rate": 1.9972262409619534e-05,
+      "loss": 1.2738,
+      "step": 1613
+    },
+    {
+      "epoch": 0.6009005130830357,
+      "grad_norm": 0.17789305746555328,
+      "learning_rate": 1.997217187998678e-05,
+      "loss": 1.2877,
+      "step": 1614
+    },
+    {
+      "epoch": 0.6012728182336448,
+      "grad_norm": 0.18692655861377716,
+      "learning_rate": 1.9972081203065672e-05,
+      "loss": 1.2754,
+      "step": 1615
+    },
+    {
+      "epoch": 0.6016451233842538,
+      "grad_norm": 0.1710839867591858,
+      "learning_rate": 1.997199037885755e-05,
+      "loss": 1.2831,
+      "step": 1616
+    },
+    {
+      "epoch": 0.6020174285348628,
+      "grad_norm": 0.1843128800392151,
+      "learning_rate": 1.9971899407363757e-05,
+      "loss": 1.281,
+      "step": 1617
+    },
+    {
+      "epoch": 0.602389733685472,
+      "grad_norm": 0.19229008257389069,
+      "learning_rate": 1.9971808288585636e-05,
+      "loss": 1.3007,
+      "step": 1618
+    },
+    {
+      "epoch": 0.602762038836081,
+      "grad_norm": 0.18288210034370422,
+      "learning_rate": 1.997171702252453e-05,
+      "loss": 1.2852,
+      "step": 1619
+    },
+    {
+      "epoch": 0.6031343439866901,
+      "grad_norm": 0.18607817590236664,
+      "learning_rate": 1.997162560918179e-05,
+      "loss": 1.2698,
+      "step": 1620
+    },
+    {
+      "epoch": 0.6035066491372991,
+      "grad_norm": 0.1832936704158783,
+      "learning_rate": 1.997153404855877e-05,
+      "loss": 1.2929,
+      "step": 1621
+    },
+    {
+      "epoch": 0.6038789542879083,
+      "grad_norm": 0.1910560131072998,
+      "learning_rate": 1.9971442340656812e-05,
+      "loss": 1.2891,
+      "step": 1622
+    },
+    {
+      "epoch": 0.6042512594385173,
+      "grad_norm": 0.19132542610168457,
+      "learning_rate": 1.997135048547728e-05,
+      "loss": 1.2862,
+      "step": 1623
+    },
+    {
+      "epoch": 0.6046235645891264,
+      "grad_norm": 0.18313977122306824,
+      "learning_rate": 1.9971258483021526e-05,
+      "loss": 1.278,
+      "step": 1624
+    },
+    {
+      "epoch": 0.6049958697397354,
+      "grad_norm": 0.17760074138641357,
+      "learning_rate": 1.997116633329091e-05,
+      "loss": 1.2783,
+      "step": 1625
+    },
+    {
+      "epoch": 0.6053681748903444,
+      "grad_norm": 0.18431836366653442,
+      "learning_rate": 1.997107403628679e-05,
+      "loss": 1.28,
+      "step": 1626
+    },
+    {
+      "epoch": 0.6057404800409536,
+      "grad_norm": 0.18512500822544098,
+      "learning_rate": 1.9970981592010538e-05,
+      "loss": 1.2817,
+      "step": 1627
+    },
+    {
+      "epoch": 0.6061127851915626,
+      "grad_norm": 0.18769600987434387,
+      "learning_rate": 1.9970889000463512e-05,
+      "loss": 1.2779,
+      "step": 1628
+    },
+    {
+      "epoch": 0.6064850903421717,
+      "grad_norm": 0.17421449720859528,
+      "learning_rate": 1.997079626164708e-05,
+      "loss": 1.2822,
+      "step": 1629
+    },
+    {
+      "epoch": 0.6068573954927807,
+      "grad_norm": 0.1772501915693283,
+      "learning_rate": 1.997070337556261e-05,
+      "loss": 1.2774,
+      "step": 1630
+    },
+    {
+      "epoch": 0.6072297006433899,
+      "grad_norm": 0.1793195903301239,
+      "learning_rate": 1.9970610342211484e-05,
+      "loss": 1.2789,
+      "step": 1631
+    },
+    {
+      "epoch": 0.6076020057939989,
+      "grad_norm": 0.18559563159942627,
+      "learning_rate": 1.9970517161595063e-05,
+      "loss": 1.2971,
+      "step": 1632
+    },
+    {
+      "epoch": 0.607974310944608,
+      "grad_norm": 0.18821711838245392,
+      "learning_rate": 1.997042383371473e-05,
+      "loss": 1.2901,
+      "step": 1633
+    },
+    {
+      "epoch": 0.608346616095217,
+      "grad_norm": 0.18289709091186523,
+      "learning_rate": 1.9970330358571862e-05,
+      "loss": 1.2772,
+      "step": 1634
+    },
+    {
+      "epoch": 0.6087189212458262,
+      "grad_norm": 0.18088442087173462,
+      "learning_rate": 1.9970236736167846e-05,
+      "loss": 1.2877,
+      "step": 1635
+    },
+    {
+      "epoch": 0.6090912263964352,
+      "grad_norm": 0.1836581975221634,
+      "learning_rate": 1.9970142966504053e-05,
+      "loss": 1.2779,
+      "step": 1636
+    },
+    {
+      "epoch": 0.6094635315470442,
+      "grad_norm": 0.18793676793575287,
+      "learning_rate": 1.9970049049581878e-05,
+      "loss": 1.2817,
+      "step": 1637
+    },
+    {
+      "epoch": 0.6098358366976533,
+      "grad_norm": 0.18134282529354095,
+      "learning_rate": 1.9969954985402702e-05,
+      "loss": 1.2842,
+      "step": 1638
+    },
+    {
+      "epoch": 0.6102081418482623,
+      "grad_norm": 0.1860257387161255,
+      "learning_rate": 1.9969860773967916e-05,
+      "loss": 1.292,
+      "step": 1639
+    },
+    {
+      "epoch": 0.6105804469988715,
+      "grad_norm": 0.1779407411813736,
+      "learning_rate": 1.9969766415278916e-05,
+      "loss": 1.2869,
+      "step": 1640
+    },
+    {
+      "epoch": 0.6109527521494805,
+      "grad_norm": 0.17907682061195374,
+      "learning_rate": 1.9969671909337086e-05,
+      "loss": 1.2777,
+      "step": 1641
+    },
+    {
+      "epoch": 0.6113250573000896,
+      "grad_norm": 0.21232067048549652,
+      "learning_rate": 1.996957725614383e-05,
+      "loss": 1.2694,
+      "step": 1642
+    },
+    {
+      "epoch": 0.6116973624506986,
+      "grad_norm": 0.1791219562292099,
+      "learning_rate": 1.9969482455700544e-05,
+      "loss": 1.2757,
+      "step": 1643
+    },
+    {
+      "epoch": 0.6120696676013078,
+      "grad_norm": 0.17560942471027374,
+      "learning_rate": 1.996938750800863e-05,
+      "loss": 1.2838,
+      "step": 1644
+    },
+    {
+      "epoch": 0.6124419727519168,
+      "grad_norm": 0.19401302933692932,
+      "learning_rate": 1.9969292413069485e-05,
+      "loss": 1.2896,
+      "step": 1645
+    },
+    {
+      "epoch": 0.6128142779025259,
+      "grad_norm": 0.1834617555141449,
+      "learning_rate": 1.9969197170884517e-05,
+      "loss": 1.2824,
+      "step": 1646
+    },
+    {
+      "epoch": 0.6131865830531349,
+      "grad_norm": 0.18403379619121552,
+      "learning_rate": 1.9969101781455132e-05,
+      "loss": 1.2783,
+      "step": 1647
+    },
+    {
+      "epoch": 0.613558888203744,
+      "grad_norm": 0.1884704977273941,
+      "learning_rate": 1.996900624478274e-05,
+      "loss": 1.2783,
+      "step": 1648
+    },
+    {
+      "epoch": 0.6139311933543531,
+      "grad_norm": 0.1792110949754715,
+      "learning_rate": 1.9968910560868757e-05,
+      "loss": 1.288,
+      "step": 1649
+    },
+    {
+      "epoch": 0.6143034985049621,
+      "grad_norm": 0.1805010288953781,
+      "learning_rate": 1.9968814729714584e-05,
+      "loss": 1.2694,
+      "step": 1650
+    },
+    {
+      "epoch": 0.6146758036555712,
+      "grad_norm": 0.17859011888504028,
+      "learning_rate": 1.9968718751321643e-05,
+      "loss": 1.2631,
+      "step": 1651
+    },
+    {
+      "epoch": 0.6150481088061802,
+      "grad_norm": 0.1796571910381317,
+      "learning_rate": 1.9968622625691353e-05,
+      "loss": 1.2819,
+      "step": 1652
+    },
+    {
+      "epoch": 0.6154204139567894,
+      "grad_norm": 0.1841122955083847,
+      "learning_rate": 1.9968526352825135e-05,
+      "loss": 1.27,
+      "step": 1653
+    },
+    {
+      "epoch": 0.6157927191073984,
+      "grad_norm": 0.17526361346244812,
+      "learning_rate": 1.9968429932724404e-05,
+      "loss": 1.2927,
+      "step": 1654
+    },
+    {
+      "epoch": 0.6161650242580075,
+      "grad_norm": 0.17394275963306427,
+      "learning_rate": 1.996833336539059e-05,
+      "loss": 1.2978,
+      "step": 1655
+    },
+    {
+      "epoch": 0.6165373294086165,
+      "grad_norm": 0.17003491520881653,
+      "learning_rate": 1.996823665082512e-05,
+      "loss": 1.2652,
+      "step": 1656
+    },
+    {
+      "epoch": 0.6169096345592257,
+      "grad_norm": 0.18178239464759827,
+      "learning_rate": 1.9968139789029418e-05,
+      "loss": 1.2768,
+      "step": 1657
+    },
+    {
+      "epoch": 0.6172819397098347,
+      "grad_norm": 0.17376887798309326,
+      "learning_rate": 1.9968042780004917e-05,
+      "loss": 1.2898,
+      "step": 1658
+    },
+    {
+      "epoch": 0.6176542448604437,
+      "grad_norm": 0.179343119263649,
+      "learning_rate": 1.9967945623753052e-05,
+      "loss": 1.2641,
+      "step": 1659
+    },
+    {
+      "epoch": 0.6180265500110528,
+      "grad_norm": 0.1855798065662384,
+      "learning_rate": 1.9967848320275253e-05,
+      "loss": 1.2841,
+      "step": 1660
+    },
+    {
+      "epoch": 0.6183988551616619,
+      "grad_norm": 0.17375251650810242,
+      "learning_rate": 1.996775086957296e-05,
+      "loss": 1.2759,
+      "step": 1661
+    },
+    {
+      "epoch": 0.618771160312271,
+      "grad_norm": 0.1763077974319458,
+      "learning_rate": 1.9967653271647613e-05,
+      "loss": 1.274,
+      "step": 1662
+    },
+    {
+      "epoch": 0.61914346546288,
+      "grad_norm": 0.17887422442436218,
+      "learning_rate": 1.9967555526500652e-05,
+      "loss": 1.2685,
+      "step": 1663
+    },
+    {
+      "epoch": 0.6195157706134891,
+      "grad_norm": 0.17712260782718658,
+      "learning_rate": 1.9967457634133524e-05,
+      "loss": 1.2755,
+      "step": 1664
+    },
+    {
+      "epoch": 0.6198880757640981,
+      "grad_norm": 0.1702347993850708,
+      "learning_rate": 1.996735959454767e-05,
+      "loss": 1.2645,
+      "step": 1665
+    },
+    {
+      "epoch": 0.6202603809147073,
+      "grad_norm": 0.17374765872955322,
+      "learning_rate": 1.996726140774454e-05,
+      "loss": 1.2772,
+      "step": 1666
+    },
+    {
+      "epoch": 0.6206326860653163,
+      "grad_norm": 0.17728720605373383,
+      "learning_rate": 1.9967163073725585e-05,
+      "loss": 1.2669,
+      "step": 1667
+    },
+    {
+      "epoch": 0.6210049912159253,
+      "grad_norm": 0.18125899136066437,
+      "learning_rate": 1.9967064592492258e-05,
+      "loss": 1.2798,
+      "step": 1668
+    },
+    {
+      "epoch": 0.6213772963665344,
+      "grad_norm": 0.17096486687660217,
+      "learning_rate": 1.996696596404601e-05,
+      "loss": 1.2666,
+      "step": 1669
+    },
+    {
+      "epoch": 0.6217496015171435,
+      "grad_norm": 0.18018490076065063,
+      "learning_rate": 1.9966867188388303e-05,
+      "loss": 1.2884,
+      "step": 1670
+    },
+    {
+      "epoch": 0.6221219066677526,
+      "grad_norm": 0.17836543917655945,
+      "learning_rate": 1.9966768265520593e-05,
+      "loss": 1.2874,
+      "step": 1671
+    },
+    {
+      "epoch": 0.6224942118183616,
+      "grad_norm": 0.17728979885578156,
+      "learning_rate": 1.996666919544434e-05,
+      "loss": 1.2756,
+      "step": 1672
+    },
+    {
+      "epoch": 0.6228665169689707,
+      "grad_norm": 0.18157729506492615,
+      "learning_rate": 1.9966569978161008e-05,
+      "loss": 1.2668,
+      "step": 1673
+    },
+    {
+      "epoch": 0.6232388221195797,
+      "grad_norm": 0.1761295050382614,
+      "learning_rate": 1.9966470613672064e-05,
+      "loss": 1.2999,
+      "step": 1674
+    },
+    {
+      "epoch": 0.6236111272701889,
+      "grad_norm": 0.18089498579502106,
+      "learning_rate": 1.9966371101978975e-05,
+      "loss": 1.2796,
+      "step": 1675
+    },
+    {
+      "epoch": 0.6239834324207979,
+      "grad_norm": 0.17708507180213928,
+      "learning_rate": 1.996627144308321e-05,
+      "loss": 1.2575,
+      "step": 1676
+    },
+    {
+      "epoch": 0.6243557375714069,
+      "grad_norm": 0.17729350924491882,
+      "learning_rate": 1.9966171636986238e-05,
+      "loss": 1.2783,
+      "step": 1677
+    },
+    {
+      "epoch": 0.624728042722016,
+      "grad_norm": 0.18605685234069824,
+      "learning_rate": 1.996607168368954e-05,
+      "loss": 1.271,
+      "step": 1678
+    },
+    {
+      "epoch": 0.6251003478726251,
+      "grad_norm": 0.1942148506641388,
+      "learning_rate": 1.9965971583194587e-05,
+      "loss": 1.2831,
+      "step": 1679
+    },
+    {
+      "epoch": 0.6254726530232342,
+      "grad_norm": 0.18195581436157227,
+      "learning_rate": 1.996587133550286e-05,
+      "loss": 1.278,
+      "step": 1680
+    },
+    {
+      "epoch": 0.6258449581738432,
+      "grad_norm": 0.17865653336048126,
+      "learning_rate": 1.996577094061584e-05,
+      "loss": 1.2773,
+      "step": 1681
+    },
+    {
+      "epoch": 0.6262172633244523,
+      "grad_norm": 0.1941101998090744,
+      "learning_rate": 1.9965670398535004e-05,
+      "loss": 1.2721,
+      "step": 1682
+    },
+    {
+      "epoch": 0.6265895684750614,
+      "grad_norm": 0.18968285620212555,
+      "learning_rate": 1.9965569709261845e-05,
+      "loss": 1.2871,
+      "step": 1683
+    },
+    {
+      "epoch": 0.6269618736256705,
+      "grad_norm": 0.18020257353782654,
+      "learning_rate": 1.996546887279785e-05,
+      "loss": 1.2824,
+      "step": 1684
+    },
+    {
+      "epoch": 0.6273341787762795,
+      "grad_norm": 0.1798500269651413,
+      "learning_rate": 1.99653678891445e-05,
+      "loss": 1.2753,
+      "step": 1685
+    },
+    {
+      "epoch": 0.6277064839268885,
+      "grad_norm": 0.16887503862380981,
+      "learning_rate": 1.9965266758303295e-05,
+      "loss": 1.2799,
+      "step": 1686
+    },
+    {
+      "epoch": 0.6280787890774976,
+      "grad_norm": 0.1764368861913681,
+      "learning_rate": 1.9965165480275722e-05,
+      "loss": 1.2649,
+      "step": 1687
+    },
+    {
+      "epoch": 0.6284510942281067,
+      "grad_norm": 0.19635643064975739,
+      "learning_rate": 1.9965064055063282e-05,
+      "loss": 1.2691,
+      "step": 1688
+    },
+    {
+      "epoch": 0.6288233993787158,
+      "grad_norm": 0.1750660389661789,
+      "learning_rate": 1.996496248266747e-05,
+      "loss": 1.2775,
+      "step": 1689
+    },
+    {
+      "epoch": 0.6291957045293248,
+      "grad_norm": 0.17725105583667755,
+      "learning_rate": 1.996486076308979e-05,
+      "loss": 1.2829,
+      "step": 1690
+    },
+    {
+      "epoch": 0.6295680096799339,
+      "grad_norm": 0.1825384497642517,
+      "learning_rate": 1.9964758896331743e-05,
+      "loss": 1.2577,
+      "step": 1691
+    },
+    {
+      "epoch": 0.629940314830543,
+      "grad_norm": 0.18121573328971863,
+      "learning_rate": 1.996465688239483e-05,
+      "loss": 1.2805,
+      "step": 1692
+    },
+    {
+      "epoch": 0.6303126199811521,
+      "grad_norm": 0.18279202282428741,
+      "learning_rate": 1.996455472128056e-05,
+      "loss": 1.2794,
+      "step": 1693
+    },
+    {
+      "epoch": 0.6306849251317611,
+      "grad_norm": 0.17859618365764618,
+      "learning_rate": 1.996445241299044e-05,
+      "loss": 1.2718,
+      "step": 1694
+    },
+    {
+      "epoch": 0.6310572302823702,
+      "grad_norm": 0.18363523483276367,
+      "learning_rate": 1.9964349957525988e-05,
+      "loss": 1.2731,
+      "step": 1695
+    },
+    {
+      "epoch": 0.6314295354329793,
+      "grad_norm": 0.1790492683649063,
+      "learning_rate": 1.9964247354888712e-05,
+      "loss": 1.2734,
+      "step": 1696
+    },
+    {
+      "epoch": 0.6318018405835883,
+      "grad_norm": 0.17761005461215973,
+      "learning_rate": 1.9964144605080125e-05,
+      "loss": 1.2536,
+      "step": 1697
+    },
+    {
+      "epoch": 0.6321741457341974,
+      "grad_norm": 0.18184557557106018,
+      "learning_rate": 1.9964041708101745e-05,
+      "loss": 1.2721,
+      "step": 1698
+    },
+    {
+      "epoch": 0.6325464508848064,
+      "grad_norm": 0.18488840758800507,
+      "learning_rate": 1.99639386639551e-05,
+      "loss": 1.2795,
+      "step": 1699
+    },
+    {
+      "epoch": 0.6329187560354155,
+      "grad_norm": 0.1846456080675125,
+      "learning_rate": 1.9963835472641704e-05,
+      "loss": 1.2744,
+      "step": 1700
+    },
+    {
+      "epoch": 0.6332910611860246,
+      "grad_norm": 0.1817891150712967,
+      "learning_rate": 1.9963732134163084e-05,
+      "loss": 1.2856,
+      "step": 1701
+    },
+    {
+      "epoch": 0.6336633663366337,
+      "grad_norm": 0.18094323575496674,
+      "learning_rate": 1.9963628648520767e-05,
+      "loss": 1.28,
+      "step": 1702
+    },
+    {
+      "epoch": 0.6340356714872427,
+      "grad_norm": 0.18917135894298553,
+      "learning_rate": 1.9963525015716277e-05,
+      "loss": 1.277,
+      "step": 1703
+    },
+    {
+      "epoch": 0.6344079766378518,
+      "grad_norm": 0.19249355792999268,
+      "learning_rate": 1.996342123575115e-05,
+      "loss": 1.2728,
+      "step": 1704
+    },
+    {
+      "epoch": 0.6347802817884609,
+      "grad_norm": 0.17361187934875488,
+      "learning_rate": 1.9963317308626916e-05,
+      "loss": 1.2699,
+      "step": 1705
+    },
+    {
+      "epoch": 0.63515258693907,
+      "grad_norm": 0.18167155981063843,
+      "learning_rate": 1.996321323434511e-05,
+      "loss": 1.2732,
+      "step": 1706
+    },
+    {
+      "epoch": 0.635524892089679,
+      "grad_norm": 0.18852338194847107,
+      "learning_rate": 1.9963109012907268e-05,
+      "loss": 1.2791,
+      "step": 1707
+    },
+    {
+      "epoch": 0.635897197240288,
+      "grad_norm": 0.17643553018569946,
+      "learning_rate": 1.9963004644314937e-05,
+      "loss": 1.2737,
+      "step": 1708
+    },
+    {
+      "epoch": 0.6362695023908972,
+      "grad_norm": 0.18324004113674164,
+      "learning_rate": 1.9962900128569645e-05,
+      "loss": 1.2779,
+      "step": 1709
+    },
+    {
+      "epoch": 0.6366418075415062,
+      "grad_norm": 0.18091078102588654,
+      "learning_rate": 1.9962795465672952e-05,
+      "loss": 1.2825,
+      "step": 1710
+    },
+    {
+      "epoch": 0.6370141126921153,
+      "grad_norm": 0.17854343354701996,
+      "learning_rate": 1.996269065562639e-05,
+      "loss": 1.2876,
+      "step": 1711
+    },
+    {
+      "epoch": 0.6373864178427243,
+      "grad_norm": 0.16912639141082764,
+      "learning_rate": 1.9962585698431513e-05,
+      "loss": 1.2463,
+      "step": 1712
+    },
+    {
+      "epoch": 0.6377587229933334,
+      "grad_norm": 0.18469521403312683,
+      "learning_rate": 1.9962480594089867e-05,
+      "loss": 1.2723,
+      "step": 1713
+    },
+    {
+      "epoch": 0.6381310281439425,
+      "grad_norm": 0.1734856516122818,
+      "learning_rate": 1.9962375342603013e-05,
+      "loss": 1.2732,
+      "step": 1714
+    },
+    {
+      "epoch": 0.6385033332945516,
+      "grad_norm": 0.17890629172325134,
+      "learning_rate": 1.99622699439725e-05,
+      "loss": 1.2711,
+      "step": 1715
+    },
+    {
+      "epoch": 0.6388756384451606,
+      "grad_norm": 0.18328309059143066,
+      "learning_rate": 1.996216439819988e-05,
+      "loss": 1.2765,
+      "step": 1716
+    },
+    {
+      "epoch": 0.6392479435957696,
+      "grad_norm": 0.18121285736560822,
+      "learning_rate": 1.9962058705286722e-05,
+      "loss": 1.2854,
+      "step": 1717
+    },
+    {
+      "epoch": 0.6396202487463788,
+      "grad_norm": 0.18126457929611206,
+      "learning_rate": 1.9961952865234582e-05,
+      "loss": 1.2748,
+      "step": 1718
+    },
+    {
+      "epoch": 0.6399925538969878,
+      "grad_norm": 0.17969095706939697,
+      "learning_rate": 1.9961846878045024e-05,
+      "loss": 1.2606,
+      "step": 1719
+    },
+    {
+      "epoch": 0.6403648590475969,
+      "grad_norm": 0.18220089375972748,
+      "learning_rate": 1.9961740743719612e-05,
+      "loss": 1.2719,
+      "step": 1720
+    },
+    {
+      "epoch": 0.6407371641982059,
+      "grad_norm": 0.1800280064344406,
+      "learning_rate": 1.996163446225991e-05,
+      "loss": 1.2857,
+      "step": 1721
+    },
+    {
+      "epoch": 0.6411094693488151,
+      "grad_norm": 0.17898349463939667,
+      "learning_rate": 1.9961528033667498e-05,
+      "loss": 1.27,
+      "step": 1722
+    },
+    {
+      "epoch": 0.6414817744994241,
+      "grad_norm": 0.18221056461334229,
+      "learning_rate": 1.996142145794394e-05,
+      "loss": 1.2665,
+      "step": 1723
+    },
+    {
+      "epoch": 0.6418540796500332,
+      "grad_norm": 0.18029160797595978,
+      "learning_rate": 1.996131473509081e-05,
+      "loss": 1.2915,
+      "step": 1724
+    },
+    {
+      "epoch": 0.6422263848006422,
+      "grad_norm": 0.17953678965568542,
+      "learning_rate": 1.9961207865109688e-05,
+      "loss": 1.2822,
+      "step": 1725
+    },
+    {
+      "epoch": 0.6425986899512512,
+      "grad_norm": 0.20553520321846008,
+      "learning_rate": 1.9961100848002154e-05,
+      "loss": 1.2923,
+      "step": 1726
+    },
+    {
+      "epoch": 0.6429709951018604,
+      "grad_norm": 0.1798718124628067,
+      "learning_rate": 1.996099368376978e-05,
+      "loss": 1.2898,
+      "step": 1727
+    },
+    {
+      "epoch": 0.6433433002524694,
+      "grad_norm": 0.18341365456581116,
+      "learning_rate": 1.996088637241416e-05,
+      "loss": 1.2693,
+      "step": 1728
+    },
+    {
+      "epoch": 0.6437156054030785,
+      "grad_norm": 0.18246109783649445,
+      "learning_rate": 1.996077891393687e-05,
+      "loss": 1.2783,
+      "step": 1729
+    },
+    {
+      "epoch": 0.6440879105536875,
+      "grad_norm": 0.1854100227355957,
+      "learning_rate": 1.9960671308339503e-05,
+      "loss": 1.2791,
+      "step": 1730
+    },
+    {
+      "epoch": 0.6444602157042967,
+      "grad_norm": 0.1951596736907959,
+      "learning_rate": 1.9960563555623644e-05,
+      "loss": 1.2851,
+      "step": 1731
+    },
+    {
+      "epoch": 0.6448325208549057,
+      "grad_norm": 0.20141799747943878,
+      "learning_rate": 1.9960455655790885e-05,
+      "loss": 1.277,
+      "step": 1732
+    },
+    {
+      "epoch": 0.6452048260055148,
+      "grad_norm": 0.1731090396642685,
+      "learning_rate": 1.996034760884282e-05,
+      "loss": 1.2755,
+      "step": 1733
+    },
+    {
+      "epoch": 0.6455771311561238,
+      "grad_norm": 0.2005818784236908,
+      "learning_rate": 1.9960239414781052e-05,
+      "loss": 1.2724,
+      "step": 1734
+    },
+    {
+      "epoch": 0.645949436306733,
+      "grad_norm": 0.1791912168264389,
+      "learning_rate": 1.996013107360717e-05,
+      "loss": 1.2689,
+      "step": 1735
+    },
+    {
+      "epoch": 0.646321741457342,
+      "grad_norm": 0.1755741536617279,
+      "learning_rate": 1.9960022585322774e-05,
+      "loss": 1.2772,
+      "step": 1736
+    },
+    {
+      "epoch": 0.646694046607951,
+      "grad_norm": 0.1764545440673828,
+      "learning_rate": 1.9959913949929474e-05,
+      "loss": 1.2745,
+      "step": 1737
+    },
+    {
+      "epoch": 0.6470663517585601,
+      "grad_norm": 0.1788182109594345,
+      "learning_rate": 1.9959805167428868e-05,
+      "loss": 1.2635,
+      "step": 1738
+    },
+    {
+      "epoch": 0.6474386569091691,
+      "grad_norm": 0.1874772310256958,
+      "learning_rate": 1.9959696237822566e-05,
+      "loss": 1.2883,
+      "step": 1739
+    },
+    {
+      "epoch": 0.6478109620597783,
+      "grad_norm": 0.18383602797985077,
+      "learning_rate": 1.9959587161112174e-05,
+      "loss": 1.2667,
+      "step": 1740
+    },
+    {
+      "epoch": 0.6481832672103873,
+      "grad_norm": 0.1755083203315735,
+      "learning_rate": 1.9959477937299305e-05,
+      "loss": 1.2857,
+      "step": 1741
+    },
+    {
+      "epoch": 0.6485555723609964,
+      "grad_norm": 0.1754198670387268,
+      "learning_rate": 1.995936856638557e-05,
+      "loss": 1.2582,
+      "step": 1742
+    },
+    {
+      "epoch": 0.6489278775116054,
+      "grad_norm": 0.18323983252048492,
+      "learning_rate": 1.9959259048372593e-05,
+      "loss": 1.2825,
+      "step": 1743
+    },
+    {
+      "epoch": 0.6493001826622146,
+      "grad_norm": 0.17415131628513336,
+      "learning_rate": 1.995914938326198e-05,
+      "loss": 1.2624,
+      "step": 1744
+    },
+    {
+      "epoch": 0.6496724878128236,
+      "grad_norm": 0.18525010347366333,
+      "learning_rate": 1.9959039571055356e-05,
+      "loss": 1.2923,
+      "step": 1745
+    },
+    {
+      "epoch": 0.6500447929634326,
+      "grad_norm": 0.1781807243824005,
+      "learning_rate": 1.995892961175434e-05,
+      "loss": 1.2821,
+      "step": 1746
+    },
+    {
+      "epoch": 0.6504170981140417,
+      "grad_norm": 0.1848139613866806,
+      "learning_rate": 1.995881950536056e-05,
+      "loss": 1.2953,
+      "step": 1747
+    },
+    {
+      "epoch": 0.6507894032646507,
+      "grad_norm": 0.17933423817157745,
+      "learning_rate": 1.9958709251875642e-05,
+      "loss": 1.2666,
+      "step": 1748
+    },
+    {
+      "epoch": 0.6511617084152599,
+      "grad_norm": 0.17203424870967865,
+      "learning_rate": 1.9958598851301218e-05,
+      "loss": 1.2691,
+      "step": 1749
+    },
+    {
+      "epoch": 0.6515340135658689,
+      "grad_norm": 0.1798200160264969,
+      "learning_rate": 1.995848830363891e-05,
+      "loss": 1.2672,
+      "step": 1750
+    },
+    {
+      "epoch": 0.651906318716478,
+      "grad_norm": 0.19163981080055237,
+      "learning_rate": 1.9958377608890348e-05,
+      "loss": 1.2615,
+      "step": 1751
+    },
+    {
+      "epoch": 0.652278623867087,
+      "grad_norm": 0.177472323179245,
+      "learning_rate": 1.9958266767057183e-05,
+      "loss": 1.2775,
+      "step": 1752
+    },
+    {
+      "epoch": 0.6526509290176962,
+      "grad_norm": 0.17973048985004425,
+      "learning_rate": 1.995815577814104e-05,
+      "loss": 1.2709,
+      "step": 1753
+    },
+    {
+      "epoch": 0.6530232341683052,
+      "grad_norm": 0.20184633135795593,
+      "learning_rate": 1.995804464214356e-05,
+      "loss": 1.2796,
+      "step": 1754
+    },
+    {
+      "epoch": 0.6533955393189143,
+      "grad_norm": 0.1837623566389084,
+      "learning_rate": 1.9957933359066385e-05,
+      "loss": 1.2679,
+      "step": 1755
+    },
+    {
+      "epoch": 0.6537678444695233,
+      "grad_norm": 0.17423506081104279,
+      "learning_rate": 1.995782192891116e-05,
+      "loss": 1.2671,
+      "step": 1756
+    },
+    {
+      "epoch": 0.6541401496201324,
+      "grad_norm": 0.1909656673669815,
+      "learning_rate": 1.9957710351679533e-05,
+      "loss": 1.2775,
+      "step": 1757
+    },
+    {
+      "epoch": 0.6545124547707415,
+      "grad_norm": 0.19637969136238098,
+      "learning_rate": 1.9957598627373145e-05,
+      "loss": 1.2737,
+      "step": 1758
+    },
+    {
+      "epoch": 0.6548847599213505,
+      "grad_norm": 0.18318361043930054,
+      "learning_rate": 1.995748675599365e-05,
+      "loss": 1.2772,
+      "step": 1759
+    },
+    {
+      "epoch": 0.6552570650719596,
+      "grad_norm": 0.1808813214302063,
+      "learning_rate": 1.9957374737542702e-05,
+      "loss": 1.2918,
+      "step": 1760
+    },
+    {
+      "epoch": 0.6556293702225686,
+      "grad_norm": 0.1866048276424408,
+      "learning_rate": 1.9957262572021955e-05,
+      "loss": 1.2677,
+      "step": 1761
+    },
+    {
+      "epoch": 0.6560016753731778,
+      "grad_norm": 0.18360257148742676,
+      "learning_rate": 1.9957150259433065e-05,
+      "loss": 1.2832,
+      "step": 1762
+    },
+    {
+      "epoch": 0.6563739805237868,
+      "grad_norm": 0.17542779445648193,
+      "learning_rate": 1.995703779977769e-05,
+      "loss": 1.2695,
+      "step": 1763
+    },
+    {
+      "epoch": 0.6567462856743959,
+      "grad_norm": 0.1828540414571762,
+      "learning_rate": 1.995692519305749e-05,
+      "loss": 1.2893,
+      "step": 1764
+    },
+    {
+      "epoch": 0.6571185908250049,
+      "grad_norm": 0.18941769003868103,
+      "learning_rate": 1.995681243927413e-05,
+      "loss": 1.2742,
+      "step": 1765
+    },
+    {
+      "epoch": 0.657490895975614,
+      "grad_norm": 0.1731724739074707,
+      "learning_rate": 1.9956699538429275e-05,
+      "loss": 1.2966,
+      "step": 1766
+    },
+    {
+      "epoch": 0.6578632011262231,
+      "grad_norm": 0.18366792798042297,
+      "learning_rate": 1.9956586490524596e-05,
+      "loss": 1.2779,
+      "step": 1767
+    },
+    {
+      "epoch": 0.6582355062768321,
+      "grad_norm": 0.18383246660232544,
+      "learning_rate": 1.9956473295561756e-05,
+      "loss": 1.267,
+      "step": 1768
+    },
+    {
+      "epoch": 0.6586078114274412,
+      "grad_norm": 0.19387759268283844,
+      "learning_rate": 1.9956359953542433e-05,
+      "loss": 1.2768,
+      "step": 1769
+    },
+    {
+      "epoch": 0.6589801165780503,
+      "grad_norm": 0.18272364139556885,
+      "learning_rate": 1.9956246464468294e-05,
+      "loss": 1.2654,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6593524217286594,
+      "grad_norm": 0.17363867163658142,
+      "learning_rate": 1.9956132828341022e-05,
+      "loss": 1.2724,
+      "step": 1771
+    },
+    {
+      "epoch": 0.6597247268792684,
+      "grad_norm": 0.18008361756801605,
+      "learning_rate": 1.9956019045162294e-05,
+      "loss": 1.2799,
+      "step": 1772
+    },
+    {
+      "epoch": 0.6600970320298775,
+      "grad_norm": 0.18820790946483612,
+      "learning_rate": 1.995590511493379e-05,
+      "loss": 1.2741,
+      "step": 1773
+    },
+    {
+      "epoch": 0.6604693371804865,
+      "grad_norm": 0.18072496354579926,
+      "learning_rate": 1.995579103765719e-05,
+      "loss": 1.2887,
+      "step": 1774
+    },
+    {
+      "epoch": 0.6608416423310957,
+      "grad_norm": 0.18533405661582947,
+      "learning_rate": 1.9955676813334182e-05,
+      "loss": 1.2798,
+      "step": 1775
+    },
+    {
+      "epoch": 0.6612139474817047,
+      "grad_norm": 0.19441775977611542,
+      "learning_rate": 1.9955562441966452e-05,
+      "loss": 1.2523,
+      "step": 1776
+    },
+    {
+      "epoch": 0.6615862526323137,
+      "grad_norm": 0.19145584106445312,
+      "learning_rate": 1.995544792355569e-05,
+      "loss": 1.2635,
+      "step": 1777
+    },
+    {
+      "epoch": 0.6619585577829228,
+      "grad_norm": 0.2026701271533966,
+      "learning_rate": 1.9955333258103586e-05,
+      "loss": 1.2747,
+      "step": 1778
+    },
+    {
+      "epoch": 0.6623308629335319,
+      "grad_norm": 0.18436765670776367,
+      "learning_rate": 1.9955218445611834e-05,
+      "loss": 1.2653,
+      "step": 1779
+    },
+    {
+      "epoch": 0.662703168084141,
+      "grad_norm": 0.18256279826164246,
+      "learning_rate": 1.9955103486082135e-05,
+      "loss": 1.2807,
+      "step": 1780
+    },
+    {
+      "epoch": 0.66307547323475,
+      "grad_norm": 0.1890406608581543,
+      "learning_rate": 1.9954988379516177e-05,
+      "loss": 1.2832,
+      "step": 1781
+    },
+    {
+      "epoch": 0.6634477783853591,
+      "grad_norm": 0.1824447363615036,
+      "learning_rate": 1.995487312591567e-05,
+      "loss": 1.2533,
+      "step": 1782
+    },
+    {
+      "epoch": 0.6638200835359682,
+      "grad_norm": 0.1869254857301712,
+      "learning_rate": 1.9954757725282308e-05,
+      "loss": 1.2811,
+      "step": 1783
+    },
+    {
+      "epoch": 0.6641923886865773,
+      "grad_norm": 0.18558120727539062,
+      "learning_rate": 1.99546421776178e-05,
+      "loss": 1.2687,
+      "step": 1784
+    },
+    {
+      "epoch": 0.6645646938371863,
+      "grad_norm": 0.18721050024032593,
+      "learning_rate": 1.995452648292385e-05,
+      "loss": 1.2834,
+      "step": 1785
+    },
+    {
+      "epoch": 0.6649369989877953,
+      "grad_norm": 0.1912430077791214,
+      "learning_rate": 1.9954410641202173e-05,
+      "loss": 1.2801,
+      "step": 1786
+    },
+    {
+      "epoch": 0.6653093041384044,
+      "grad_norm": 0.1814345419406891,
+      "learning_rate": 1.995429465245447e-05,
+      "loss": 1.2739,
+      "step": 1787
+    },
+    {
+      "epoch": 0.6656816092890135,
+      "grad_norm": 0.18435440957546234,
+      "learning_rate": 1.9954178516682464e-05,
+      "loss": 1.2652,
+      "step": 1788
+    },
+    {
+      "epoch": 0.6660539144396226,
+      "grad_norm": 0.18596956133842468,
+      "learning_rate": 1.9954062233887866e-05,
+      "loss": 1.2661,
+      "step": 1789
+    },
+    {
+      "epoch": 0.6664262195902316,
+      "grad_norm": 0.1784246861934662,
+      "learning_rate": 1.995394580407239e-05,
+      "loss": 1.2688,
+      "step": 1790
+    },
+    {
+      "epoch": 0.6667985247408407,
+      "grad_norm": 0.1948971003293991,
+      "learning_rate": 1.9953829227237762e-05,
+      "loss": 1.2825,
+      "step": 1791
+    },
+    {
+      "epoch": 0.6671708298914498,
+      "grad_norm": 0.17607876658439636,
+      "learning_rate": 1.9953712503385702e-05,
+      "loss": 1.2722,
+      "step": 1792
+    },
+    {
+      "epoch": 0.6675431350420589,
+      "grad_norm": 0.19248910248279572,
+      "learning_rate": 1.995359563251793e-05,
+      "loss": 1.2908,
+      "step": 1793
+    },
+    {
+      "epoch": 0.6679154401926679,
+      "grad_norm": 0.1824008673429489,
+      "learning_rate": 1.9953478614636178e-05,
+      "loss": 1.2737,
+      "step": 1794
+    },
+    {
+      "epoch": 0.668287745343277,
+      "grad_norm": 0.17767947912216187,
+      "learning_rate": 1.9953361449742167e-05,
+      "loss": 1.2698,
+      "step": 1795
+    },
+    {
+      "epoch": 0.6686600504938861,
+      "grad_norm": 0.20070941746234894,
+      "learning_rate": 1.995324413783764e-05,
+      "loss": 1.2716,
+      "step": 1796
+    },
+    {
+      "epoch": 0.6690323556444951,
+      "grad_norm": 0.1810212880373001,
+      "learning_rate": 1.9953126678924315e-05,
+      "loss": 1.2641,
+      "step": 1797
+    },
+    {
+      "epoch": 0.6694046607951042,
+      "grad_norm": 0.18314088881015778,
+      "learning_rate": 1.9953009073003935e-05,
+      "loss": 1.2713,
+      "step": 1798
+    },
+    {
+      "epoch": 0.6697769659457132,
+      "grad_norm": 0.1876356452703476,
+      "learning_rate": 1.9952891320078235e-05,
+      "loss": 1.279,
+      "step": 1799
+    },
+    {
+      "epoch": 0.6701492710963223,
+      "grad_norm": 0.18788988888263702,
+      "learning_rate": 1.9952773420148958e-05,
+      "loss": 1.2786,
+      "step": 1800
+    },
+    {
+      "epoch": 0.6705215762469314,
+      "grad_norm": 0.17897135019302368,
+      "learning_rate": 1.995265537321784e-05,
+      "loss": 1.2746,
+      "step": 1801
+    },
+    {
+      "epoch": 0.6708938813975405,
+      "grad_norm": 0.1953326016664505,
+      "learning_rate": 1.9952537179286623e-05,
+      "loss": 1.2815,
+      "step": 1802
+    },
+    {
+      "epoch": 0.6712661865481495,
+      "grad_norm": 0.19297561049461365,
+      "learning_rate": 1.995241883835706e-05,
+      "loss": 1.2698,
+      "step": 1803
+    },
+    {
+      "epoch": 0.6716384916987586,
+      "grad_norm": 0.18052466213703156,
+      "learning_rate": 1.99523003504309e-05,
+      "loss": 1.2744,
+      "step": 1804
+    },
+    {
+      "epoch": 0.6720107968493677,
+      "grad_norm": 0.1788993775844574,
+      "learning_rate": 1.9952181715509883e-05,
+      "loss": 1.2621,
+      "step": 1805
+    },
+    {
+      "epoch": 0.6723831019999768,
+      "grad_norm": 0.18485794961452484,
+      "learning_rate": 1.9952062933595765e-05,
+      "loss": 1.2767,
+      "step": 1806
+    },
+    {
+      "epoch": 0.6727554071505858,
+      "grad_norm": 0.18124793469905853,
+      "learning_rate": 1.9951944004690308e-05,
+      "loss": 1.2811,
+      "step": 1807
+    },
+    {
+      "epoch": 0.6731277123011948,
+      "grad_norm": 0.1836545318365097,
+      "learning_rate": 1.9951824928795255e-05,
+      "loss": 1.2692,
+      "step": 1808
+    },
+    {
+      "epoch": 0.673500017451804,
+      "grad_norm": 0.1730068475008011,
+      "learning_rate": 1.9951705705912377e-05,
+      "loss": 1.2753,
+      "step": 1809
+    },
+    {
+      "epoch": 0.673872322602413,
+      "grad_norm": 0.17720746994018555,
+      "learning_rate": 1.995158633604343e-05,
+      "loss": 1.2653,
+      "step": 1810
+    },
+    {
+      "epoch": 0.6742446277530221,
+      "grad_norm": 0.1764315515756607,
+      "learning_rate": 1.995146681919018e-05,
+      "loss": 1.2608,
+      "step": 1811
+    },
+    {
+      "epoch": 0.6746169329036311,
+      "grad_norm": 0.19442118704319,
+      "learning_rate": 1.9951347155354386e-05,
+      "loss": 1.2774,
+      "step": 1812
+    },
+    {
+      "epoch": 0.6749892380542402,
+      "grad_norm": 0.18689338862895966,
+      "learning_rate": 1.995122734453782e-05,
+      "loss": 1.2659,
+      "step": 1813
+    },
+    {
+      "epoch": 0.6753615432048493,
+      "grad_norm": 0.17291222512722015,
+      "learning_rate": 1.995110738674225e-05,
+      "loss": 1.2637,
+      "step": 1814
+    },
+    {
+      "epoch": 0.6757338483554584,
+      "grad_norm": 0.2008456140756607,
+      "learning_rate": 1.995098728196945e-05,
+      "loss": 1.2837,
+      "step": 1815
+    },
+    {
+      "epoch": 0.6761061535060674,
+      "grad_norm": 0.1726778745651245,
+      "learning_rate": 1.995086703022119e-05,
+      "loss": 1.2876,
+      "step": 1816
+    },
+    {
+      "epoch": 0.6764784586566764,
+      "grad_norm": 0.18805347383022308,
+      "learning_rate": 1.9950746631499252e-05,
+      "loss": 1.2486,
+      "step": 1817
+    },
+    {
+      "epoch": 0.6768507638072856,
+      "grad_norm": 0.1827612966299057,
+      "learning_rate": 1.9950626085805406e-05,
+      "loss": 1.2832,
+      "step": 1818
+    },
+    {
+      "epoch": 0.6772230689578946,
+      "grad_norm": 0.17770785093307495,
+      "learning_rate": 1.995050539314144e-05,
+      "loss": 1.2713,
+      "step": 1819
+    },
+    {
+      "epoch": 0.6775953741085037,
+      "grad_norm": 0.18160304427146912,
+      "learning_rate": 1.9950384553509134e-05,
+      "loss": 1.2836,
+      "step": 1820
+    },
+    {
+      "epoch": 0.6779676792591127,
+      "grad_norm": 0.1883334517478943,
+      "learning_rate": 1.9950263566910275e-05,
+      "loss": 1.2813,
+      "step": 1821
+    },
+    {
+      "epoch": 0.6783399844097218,
+      "grad_norm": 0.18117238581180573,
+      "learning_rate": 1.9950142433346642e-05,
+      "loss": 1.2715,
+      "step": 1822
+    },
+    {
+      "epoch": 0.6787122895603309,
+      "grad_norm": 0.1771121472120285,
+      "learning_rate": 1.9950021152820032e-05,
+      "loss": 1.266,
+      "step": 1823
+    },
+    {
+      "epoch": 0.67908459471094,
+      "grad_norm": 0.18538367748260498,
+      "learning_rate": 1.9949899725332233e-05,
+      "loss": 1.2638,
+      "step": 1824
+    },
+    {
+      "epoch": 0.679456899861549,
+      "grad_norm": 0.17744530737400055,
+      "learning_rate": 1.994977815088504e-05,
+      "loss": 1.2669,
+      "step": 1825
+    },
+    {
+      "epoch": 0.679829205012158,
+      "grad_norm": 0.17428846657276154,
+      "learning_rate": 1.9949656429480252e-05,
+      "loss": 1.2753,
+      "step": 1826
+    },
+    {
+      "epoch": 0.6802015101627672,
+      "grad_norm": 0.2017204761505127,
+      "learning_rate": 1.9949534561119658e-05,
+      "loss": 1.2723,
+      "step": 1827
+    },
+    {
+      "epoch": 0.6805738153133762,
+      "grad_norm": 0.18561388552188873,
+      "learning_rate": 1.9949412545805065e-05,
+      "loss": 1.2748,
+      "step": 1828
+    },
+    {
+      "epoch": 0.6809461204639853,
+      "grad_norm": 0.18007059395313263,
+      "learning_rate": 1.9949290383538272e-05,
+      "loss": 1.2579,
+      "step": 1829
+    },
+    {
+      "epoch": 0.6813184256145943,
+      "grad_norm": 0.17988115549087524,
+      "learning_rate": 1.9949168074321088e-05,
+      "loss": 1.2769,
+      "step": 1830
+    },
+    {
+      "epoch": 0.6816907307652035,
+      "grad_norm": 0.19387783110141754,
+      "learning_rate": 1.9949045618155312e-05,
+      "loss": 1.2746,
+      "step": 1831
+    },
+    {
+      "epoch": 0.6820630359158125,
+      "grad_norm": 0.1811383217573166,
+      "learning_rate": 1.994892301504276e-05,
+      "loss": 1.2784,
+      "step": 1832
+    },
+    {
+      "epoch": 0.6824353410664216,
+      "grad_norm": 0.17574800550937653,
+      "learning_rate": 1.9948800264985236e-05,
+      "loss": 1.2688,
+      "step": 1833
+    },
+    {
+      "epoch": 0.6828076462170306,
+      "grad_norm": 0.17047591507434845,
+      "learning_rate": 1.9948677367984558e-05,
+      "loss": 1.2783,
+      "step": 1834
+    },
+    {
+      "epoch": 0.6831799513676396,
+      "grad_norm": 0.18834684789180756,
+      "learning_rate": 1.994855432404254e-05,
+      "loss": 1.2685,
+      "step": 1835
+    },
+    {
+      "epoch": 0.6835522565182488,
+      "grad_norm": 0.17548666894435883,
+      "learning_rate": 1.9948431133160998e-05,
+      "loss": 1.269,
+      "step": 1836
+    },
+    {
+      "epoch": 0.6839245616688578,
+      "grad_norm": 0.17728163301944733,
+      "learning_rate": 1.9948307795341755e-05,
+      "loss": 1.2748,
+      "step": 1837
+    },
+    {
+      "epoch": 0.6842968668194669,
+      "grad_norm": 0.18654660880565643,
+      "learning_rate": 1.9948184310586625e-05,
+      "loss": 1.2839,
+      "step": 1838
+    },
+    {
+      "epoch": 0.6846691719700759,
+      "grad_norm": 0.18166908621788025,
+      "learning_rate": 1.9948060678897443e-05,
+      "loss": 1.2677,
+      "step": 1839
+    },
+    {
+      "epoch": 0.6850414771206851,
+      "grad_norm": 0.17448103427886963,
+      "learning_rate": 1.9947936900276023e-05,
+      "loss": 1.2738,
+      "step": 1840
+    },
+    {
+      "epoch": 0.6854137822712941,
+      "grad_norm": 0.17304863035678864,
+      "learning_rate": 1.9947812974724203e-05,
+      "loss": 1.2705,
+      "step": 1841
+    },
+    {
+      "epoch": 0.6857860874219032,
+      "grad_norm": 0.17068156599998474,
+      "learning_rate": 1.994768890224381e-05,
+      "loss": 1.2782,
+      "step": 1842
+    },
+    {
+      "epoch": 0.6861583925725122,
+      "grad_norm": 0.17635151743888855,
+      "learning_rate": 1.9947564682836678e-05,
+      "loss": 1.2707,
+      "step": 1843
+    },
+    {
+      "epoch": 0.6865306977231214,
+      "grad_norm": 0.17905697226524353,
+      "learning_rate": 1.9947440316504636e-05,
+      "loss": 1.2628,
+      "step": 1844
+    },
+    {
+      "epoch": 0.6869030028737304,
+      "grad_norm": 0.17490267753601074,
+      "learning_rate": 1.9947315803249525e-05,
+      "loss": 1.257,
+      "step": 1845
+    },
+    {
+      "epoch": 0.6872753080243394,
+      "grad_norm": 0.18009649217128754,
+      "learning_rate": 1.9947191143073185e-05,
+      "loss": 1.285,
+      "step": 1846
+    },
+    {
+      "epoch": 0.6876476131749485,
+      "grad_norm": 0.19091250002384186,
+      "learning_rate": 1.994706633597746e-05,
+      "loss": 1.2647,
+      "step": 1847
+    },
+    {
+      "epoch": 0.6880199183255575,
+      "grad_norm": 0.183569997549057,
+      "learning_rate": 1.994694138196418e-05,
+      "loss": 1.2662,
+      "step": 1848
+    },
+    {
+      "epoch": 0.6883922234761667,
+      "grad_norm": 0.18256776034832,
+      "learning_rate": 1.994681628103521e-05,
+      "loss": 1.2643,
+      "step": 1849
+    },
+    {
+      "epoch": 0.6887645286267757,
+      "grad_norm": 0.17568175494670868,
+      "learning_rate": 1.9946691033192384e-05,
+      "loss": 1.2724,
+      "step": 1850
+    },
+    {
+      "epoch": 0.6891368337773848,
+      "grad_norm": 0.18592911958694458,
+      "learning_rate": 1.9946565638437552e-05,
+      "loss": 1.2602,
+      "step": 1851
+    },
+    {
+      "epoch": 0.6895091389279938,
+      "grad_norm": 0.18410173058509827,
+      "learning_rate": 1.9946440096772574e-05,
+      "loss": 1.2539,
+      "step": 1852
+    },
+    {
+      "epoch": 0.689881444078603,
+      "grad_norm": 0.18612074851989746,
+      "learning_rate": 1.99463144081993e-05,
+      "loss": 1.265,
+      "step": 1853
+    },
+    {
+      "epoch": 0.690253749229212,
+      "grad_norm": 0.19011899828910828,
+      "learning_rate": 1.9946188572719585e-05,
+      "loss": 1.2938,
+      "step": 1854
+    },
+    {
+      "epoch": 0.690626054379821,
+      "grad_norm": 0.18528851866722107,
+      "learning_rate": 1.9946062590335287e-05,
+      "loss": 1.2757,
+      "step": 1855
+    },
+    {
+      "epoch": 0.6909983595304301,
+      "grad_norm": 0.1764160841703415,
+      "learning_rate": 1.9945936461048273e-05,
+      "loss": 1.2859,
+      "step": 1856
+    },
+    {
+      "epoch": 0.6913706646810392,
+      "grad_norm": 0.17735952138900757,
+      "learning_rate": 1.9945810184860396e-05,
+      "loss": 1.2706,
+      "step": 1857
+    },
+    {
+      "epoch": 0.6917429698316483,
+      "grad_norm": 0.18873561918735504,
+      "learning_rate": 1.9945683761773533e-05,
+      "loss": 1.2833,
+      "step": 1858
+    },
+    {
+      "epoch": 0.6921152749822573,
+      "grad_norm": 0.17699329555034637,
+      "learning_rate": 1.9945557191789543e-05,
+      "loss": 1.2709,
+      "step": 1859
+    },
+    {
+      "epoch": 0.6924875801328664,
+      "grad_norm": 0.18327003717422485,
+      "learning_rate": 1.9945430474910295e-05,
+      "loss": 1.2801,
+      "step": 1860
+    },
+    {
+      "epoch": 0.6928598852834754,
+      "grad_norm": 0.18743474781513214,
+      "learning_rate": 1.9945303611137665e-05,
+      "loss": 1.2724,
+      "step": 1861
+    },
+    {
+      "epoch": 0.6932321904340846,
+      "grad_norm": 0.17246921360492706,
+      "learning_rate": 1.9945176600473526e-05,
+      "loss": 1.252,
+      "step": 1862
+    },
+    {
+      "epoch": 0.6936044955846936,
+      "grad_norm": 0.17474091053009033,
+      "learning_rate": 1.994504944291975e-05,
+      "loss": 1.2746,
+      "step": 1863
+    },
+    {
+      "epoch": 0.6939768007353027,
+      "grad_norm": 0.19760233163833618,
+      "learning_rate": 1.994492213847822e-05,
+      "loss": 1.2925,
+      "step": 1864
+    },
+    {
+      "epoch": 0.6943491058859117,
+      "grad_norm": 0.32504546642303467,
+      "learning_rate": 1.9944794687150812e-05,
+      "loss": 1.2715,
+      "step": 1865
+    },
+    {
+      "epoch": 0.6947214110365209,
+      "grad_norm": 0.16915461421012878,
+      "learning_rate": 1.9944667088939414e-05,
+      "loss": 1.2777,
+      "step": 1866
+    },
+    {
+      "epoch": 0.6950937161871299,
+      "grad_norm": 0.18415702879428864,
+      "learning_rate": 1.9944539343845905e-05,
+      "loss": 1.2666,
+      "step": 1867
+    },
+    {
+      "epoch": 0.6954660213377389,
+      "grad_norm": 0.19284015893936157,
+      "learning_rate": 1.994441145187217e-05,
+      "loss": 1.2955,
+      "step": 1868
+    },
+    {
+      "epoch": 0.695838326488348,
+      "grad_norm": 0.1781989485025406,
+      "learning_rate": 1.994428341302011e-05,
+      "loss": 1.2754,
+      "step": 1869
+    },
+    {
+      "epoch": 0.6962106316389571,
+      "grad_norm": 0.18543171882629395,
+      "learning_rate": 1.9944155227291603e-05,
+      "loss": 1.2615,
+      "step": 1870
+    },
+    {
+      "epoch": 0.6965829367895662,
+      "grad_norm": 0.1806837022304535,
+      "learning_rate": 1.9944026894688547e-05,
+      "loss": 1.2585,
+      "step": 1871
+    },
+    {
+      "epoch": 0.6969552419401752,
+      "grad_norm": 0.18426816165447235,
+      "learning_rate": 1.9943898415212842e-05,
+      "loss": 1.2783,
+      "step": 1872
+    },
+    {
+      "epoch": 0.6973275470907843,
+      "grad_norm": 0.17620044946670532,
+      "learning_rate": 1.9943769788866377e-05,
+      "loss": 1.2815,
+      "step": 1873
+    },
+    {
+      "epoch": 0.6976998522413933,
+      "grad_norm": 0.1815589815378189,
+      "learning_rate": 1.9943641015651057e-05,
+      "loss": 1.2715,
+      "step": 1874
+    },
+    {
+      "epoch": 0.6980721573920025,
+      "grad_norm": 0.1777685433626175,
+      "learning_rate": 1.9943512095568785e-05,
+      "loss": 1.2717,
+      "step": 1875
+    },
+    {
+      "epoch": 0.6984444625426115,
+      "grad_norm": 0.19864802062511444,
+      "learning_rate": 1.9943383028621463e-05,
+      "loss": 1.2827,
+      "step": 1876
+    },
+    {
+      "epoch": 0.6988167676932205,
+      "grad_norm": 0.1869768351316452,
+      "learning_rate": 1.9943253814810998e-05,
+      "loss": 1.276,
+      "step": 1877
+    },
+    {
+      "epoch": 0.6991890728438296,
+      "grad_norm": 0.18127432465553284,
+      "learning_rate": 1.9943124454139298e-05,
+      "loss": 1.258,
+      "step": 1878
+    },
+    {
+      "epoch": 0.6995613779944387,
+      "grad_norm": 0.18956811726093292,
+      "learning_rate": 1.9942994946608273e-05,
+      "loss": 1.2536,
+      "step": 1879
+    },
+    {
+      "epoch": 0.6999336831450478,
+      "grad_norm": 0.19713161885738373,
+      "learning_rate": 1.9942865292219837e-05,
+      "loss": 1.272,
+      "step": 1880
+    },
+    {
+      "epoch": 0.7003059882956568,
+      "grad_norm": 0.1791374236345291,
+      "learning_rate": 1.9942735490975903e-05,
+      "loss": 1.2739,
+      "step": 1881
+    },
+    {
+      "epoch": 0.7006782934462659,
+      "grad_norm": 0.17911037802696228,
+      "learning_rate": 1.9942605542878393e-05,
+      "loss": 1.2725,
+      "step": 1882
+    },
+    {
+      "epoch": 0.701050598596875,
+      "grad_norm": 0.1858641803264618,
+      "learning_rate": 1.9942475447929223e-05,
+      "loss": 1.2588,
+      "step": 1883
+    },
+    {
+      "epoch": 0.7014229037474841,
+      "grad_norm": 0.1832643300294876,
+      "learning_rate": 1.9942345206130313e-05,
+      "loss": 1.2863,
+      "step": 1884
+    },
+    {
+      "epoch": 0.7017952088980931,
+      "grad_norm": 0.18985310196876526,
+      "learning_rate": 1.9942214817483588e-05,
+      "loss": 1.2711,
+      "step": 1885
+    },
+    {
+      "epoch": 0.7021675140487021,
+      "grad_norm": 0.8048028945922852,
+      "learning_rate": 1.9942084281990973e-05,
+      "loss": 1.2672,
+      "step": 1886
+    },
+    {
+      "epoch": 0.7025398191993112,
+      "grad_norm": 0.19335870444774628,
+      "learning_rate": 1.9941953599654398e-05,
+      "loss": 1.2665,
+      "step": 1887
+    },
+    {
+      "epoch": 0.7029121243499203,
+      "grad_norm": 0.17363718152046204,
+      "learning_rate": 1.9941822770475795e-05,
+      "loss": 1.2456,
+      "step": 1888
+    },
+    {
+      "epoch": 0.7032844295005294,
+      "grad_norm": 0.18014384806156158,
+      "learning_rate": 1.9941691794457088e-05,
+      "loss": 1.2661,
+      "step": 1889
+    },
+    {
+      "epoch": 0.7036567346511384,
+      "grad_norm": 0.17859698832035065,
+      "learning_rate": 1.9941560671600223e-05,
+      "loss": 1.2699,
+      "step": 1890
+    },
+    {
+      "epoch": 0.7040290398017475,
+      "grad_norm": 0.17710040509700775,
+      "learning_rate": 1.9941429401907126e-05,
+      "loss": 1.2629,
+      "step": 1891
+    },
+    {
+      "epoch": 0.7044013449523566,
+      "grad_norm": 0.186106339097023,
+      "learning_rate": 1.9941297985379747e-05,
+      "loss": 1.2573,
+      "step": 1892
+    },
+    {
+      "epoch": 0.7047736501029657,
+      "grad_norm": 0.1873396337032318,
+      "learning_rate": 1.9941166422020016e-05,
+      "loss": 1.2468,
+      "step": 1893
+    },
+    {
+      "epoch": 0.7051459552535747,
+      "grad_norm": 0.1858346313238144,
+      "learning_rate": 1.9941034711829878e-05,
+      "loss": 1.2776,
+      "step": 1894
+    },
+    {
+      "epoch": 0.7055182604041838,
+      "grad_norm": 0.18264545500278473,
+      "learning_rate": 1.9940902854811284e-05,
+      "loss": 1.278,
+      "step": 1895
+    },
+    {
+      "epoch": 0.7058905655547928,
+      "grad_norm": 0.18094655871391296,
+      "learning_rate": 1.9940770850966184e-05,
+      "loss": 1.2737,
+      "step": 1896
+    },
+    {
+      "epoch": 0.706262870705402,
+      "grad_norm": 0.18092148005962372,
+      "learning_rate": 1.9940638700296514e-05,
+      "loss": 1.2721,
+      "step": 1897
+    },
+    {
+      "epoch": 0.706635175856011,
+      "grad_norm": 0.18232271075248718,
+      "learning_rate": 1.994050640280424e-05,
+      "loss": 1.2564,
+      "step": 1898
+    },
+    {
+      "epoch": 0.70700748100662,
+      "grad_norm": 0.17807061970233917,
+      "learning_rate": 1.9940373958491308e-05,
+      "loss": 1.2714,
+      "step": 1899
+    },
+    {
+      "epoch": 0.7073797861572291,
+      "grad_norm": 0.17300982773303986,
+      "learning_rate": 1.9940241367359675e-05,
+      "loss": 1.255,
+      "step": 1900
+    },
+    {
+      "epoch": 0.7077520913078382,
+      "grad_norm": 0.18998445570468903,
+      "learning_rate": 1.9940108629411305e-05,
+      "loss": 1.2794,
+      "step": 1901
+    },
+    {
+      "epoch": 0.7081243964584473,
+      "grad_norm": 0.1696198284626007,
+      "learning_rate": 1.9939975744648152e-05,
+      "loss": 1.2784,
+      "step": 1902
+    },
+    {
+      "epoch": 0.7084967016090563,
+      "grad_norm": 0.1855718344449997,
+      "learning_rate": 1.993984271307218e-05,
+      "loss": 1.2728,
+      "step": 1903
+    },
+    {
+      "epoch": 0.7088690067596654,
+      "grad_norm": 0.18481917679309845,
+      "learning_rate": 1.9939709534685353e-05,
+      "loss": 1.2794,
+      "step": 1904
+    },
+    {
+      "epoch": 0.7092413119102745,
+      "grad_norm": 0.18903714418411255,
+      "learning_rate": 1.9939576209489648e-05,
+      "loss": 1.2656,
+      "step": 1905
+    },
+    {
+      "epoch": 0.7096136170608836,
+      "grad_norm": 0.17437592148780823,
+      "learning_rate": 1.993944273748702e-05,
+      "loss": 1.2737,
+      "step": 1906
+    },
+    {
+      "epoch": 0.7099859222114926,
+      "grad_norm": 0.17127926647663116,
+      "learning_rate": 1.9939309118679445e-05,
+      "loss": 1.2662,
+      "step": 1907
+    },
+    {
+      "epoch": 0.7103582273621016,
+      "grad_norm": 0.1836404949426651,
+      "learning_rate": 1.99391753530689e-05,
+      "loss": 1.2774,
+      "step": 1908
+    },
+    {
+      "epoch": 0.7107305325127107,
+      "grad_norm": 0.18096838891506195,
+      "learning_rate": 1.993904144065736e-05,
+      "loss": 1.2775,
+      "step": 1909
+    },
+    {
+      "epoch": 0.7111028376633198,
+      "grad_norm": 0.17942848801612854,
+      "learning_rate": 1.9938907381446802e-05,
+      "loss": 1.2671,
+      "step": 1910
+    },
+    {
+      "epoch": 0.7114751428139289,
+      "grad_norm": 0.1706017553806305,
+      "learning_rate": 1.9938773175439205e-05,
+      "loss": 1.2683,
+      "step": 1911
+    },
+    {
+      "epoch": 0.7118474479645379,
+      "grad_norm": 0.1811804175376892,
+      "learning_rate": 1.9938638822636555e-05,
+      "loss": 1.2771,
+      "step": 1912
+    },
+    {
+      "epoch": 0.712219753115147,
+      "grad_norm": 0.1840972751379013,
+      "learning_rate": 1.9938504323040826e-05,
+      "loss": 1.268,
+      "step": 1913
+    },
+    {
+      "epoch": 0.7125920582657561,
+      "grad_norm": 0.17353455722332,
+      "learning_rate": 1.9938369676654015e-05,
+      "loss": 1.2617,
+      "step": 1914
+    },
+    {
+      "epoch": 0.7129643634163652,
+      "grad_norm": 0.16893000900745392,
+      "learning_rate": 1.993823488347811e-05,
+      "loss": 1.2651,
+      "step": 1915
+    },
+    {
+      "epoch": 0.7133366685669742,
+      "grad_norm": 0.17910560965538025,
+      "learning_rate": 1.9938099943515098e-05,
+      "loss": 1.2633,
+      "step": 1916
+    },
+    {
+      "epoch": 0.7137089737175832,
+      "grad_norm": 0.18880413472652435,
+      "learning_rate": 1.9937964856766975e-05,
+      "loss": 1.2445,
+      "step": 1917
+    },
+    {
+      "epoch": 0.7140812788681924,
+      "grad_norm": 0.17777028679847717,
+      "learning_rate": 1.9937829623235733e-05,
+      "loss": 1.2701,
+      "step": 1918
+    },
+    {
+      "epoch": 0.7144535840188014,
+      "grad_norm": 0.17710170149803162,
+      "learning_rate": 1.993769424292337e-05,
+      "loss": 1.2697,
+      "step": 1919
+    },
+    {
+      "epoch": 0.7148258891694105,
+      "grad_norm": 0.17528566718101501,
+      "learning_rate": 1.993755871583189e-05,
+      "loss": 1.261,
+      "step": 1920
+    },
+    {
+      "epoch": 0.7151981943200195,
+      "grad_norm": 0.17900541424751282,
+      "learning_rate": 1.993742304196329e-05,
+      "loss": 1.2624,
+      "step": 1921
+    },
+    {
+      "epoch": 0.7155704994706286,
+      "grad_norm": 0.18231788277626038,
+      "learning_rate": 1.9937287221319576e-05,
+      "loss": 1.2659,
+      "step": 1922
+    },
+    {
+      "epoch": 0.7159428046212377,
+      "grad_norm": 0.1796240657567978,
+      "learning_rate": 1.993715125390275e-05,
+      "loss": 1.2658,
+      "step": 1923
+    },
+    {
+      "epoch": 0.7163151097718468,
+      "grad_norm": 0.16966049373149872,
+      "learning_rate": 1.9937015139714825e-05,
+      "loss": 1.2611,
+      "step": 1924
+    },
+    {
+      "epoch": 0.7166874149224558,
+      "grad_norm": 0.18244795501232147,
+      "learning_rate": 1.993687887875781e-05,
+      "loss": 1.2567,
+      "step": 1925
+    },
+    {
+      "epoch": 0.7170597200730648,
+      "grad_norm": 0.17813333868980408,
+      "learning_rate": 1.993674247103372e-05,
+      "loss": 1.2639,
+      "step": 1926
+    },
+    {
+      "epoch": 0.717432025223674,
+      "grad_norm": 0.17147675156593323,
+      "learning_rate": 1.9936605916544566e-05,
+      "loss": 1.2508,
+      "step": 1927
+    },
+    {
+      "epoch": 0.717804330374283,
+      "grad_norm": 0.1758221685886383,
+      "learning_rate": 1.9936469215292366e-05,
+      "loss": 1.2537,
+      "step": 1928
+    },
+    {
+      "epoch": 0.7181766355248921,
+      "grad_norm": 0.1679316908121109,
+      "learning_rate": 1.993633236727914e-05,
+      "loss": 1.2708,
+      "step": 1929
+    },
+    {
+      "epoch": 0.7185489406755011,
+      "grad_norm": 0.18029126524925232,
+      "learning_rate": 1.9936195372506906e-05,
+      "loss": 1.2628,
+      "step": 1930
+    },
+    {
+      "epoch": 0.7189212458261103,
+      "grad_norm": 0.1729169636964798,
+      "learning_rate": 1.9936058230977694e-05,
+      "loss": 1.2676,
+      "step": 1931
+    },
+    {
+      "epoch": 0.7192935509767193,
+      "grad_norm": 0.1810874193906784,
+      "learning_rate": 1.993592094269352e-05,
+      "loss": 1.2575,
+      "step": 1932
+    },
+    {
+      "epoch": 0.7196658561273284,
+      "grad_norm": 0.17737914621829987,
+      "learning_rate": 1.993578350765642e-05,
+      "loss": 1.2801,
+      "step": 1933
+    },
+    {
+      "epoch": 0.7200381612779374,
+      "grad_norm": 0.18004107475280762,
+      "learning_rate": 1.9935645925868424e-05,
+      "loss": 1.2596,
+      "step": 1934
+    },
+    {
+      "epoch": 0.7204104664285464,
+      "grad_norm": 0.17037642002105713,
+      "learning_rate": 1.9935508197331556e-05,
+      "loss": 1.2656,
+      "step": 1935
+    },
+    {
+      "epoch": 0.7207827715791556,
+      "grad_norm": 0.1718427985906601,
+      "learning_rate": 1.993537032204786e-05,
+      "loss": 1.2734,
+      "step": 1936
+    },
+    {
+      "epoch": 0.7211550767297646,
+      "grad_norm": 0.19909080862998962,
+      "learning_rate": 1.9935232300019364e-05,
+      "loss": 1.2584,
+      "step": 1937
+    },
+    {
+      "epoch": 0.7215273818803737,
+      "grad_norm": 0.17230063676834106,
+      "learning_rate": 1.9935094131248113e-05,
+      "loss": 1.2605,
+      "step": 1938
+    },
+    {
+      "epoch": 0.7218996870309827,
+      "grad_norm": 0.18916752934455872,
+      "learning_rate": 1.9934955815736145e-05,
+      "loss": 1.2758,
+      "step": 1939
+    },
+    {
+      "epoch": 0.7222719921815919,
+      "grad_norm": 0.1908046007156372,
+      "learning_rate": 1.99348173534855e-05,
+      "loss": 1.2634,
+      "step": 1940
+    },
+    {
+      "epoch": 0.7226442973322009,
+      "grad_norm": 0.17501883208751678,
+      "learning_rate": 1.9934678744498234e-05,
+      "loss": 1.2604,
+      "step": 1941
+    },
+    {
+      "epoch": 0.72301660248281,
+      "grad_norm": 0.17639294266700745,
+      "learning_rate": 1.993453998877638e-05,
+      "loss": 1.2641,
+      "step": 1942
+    },
+    {
+      "epoch": 0.723388907633419,
+      "grad_norm": 0.19345468282699585,
+      "learning_rate": 1.9934401086321995e-05,
+      "loss": 1.2709,
+      "step": 1943
+    },
+    {
+      "epoch": 0.7237612127840282,
+      "grad_norm": 0.18009069561958313,
+      "learning_rate": 1.9934262037137132e-05,
+      "loss": 1.2635,
+      "step": 1944
+    },
+    {
+      "epoch": 0.7241335179346372,
+      "grad_norm": 0.19029593467712402,
+      "learning_rate": 1.993412284122384e-05,
+      "loss": 1.2532,
+      "step": 1945
+    },
+    {
+      "epoch": 0.7245058230852462,
+      "grad_norm": 0.19446800649166107,
+      "learning_rate": 1.9933983498584175e-05,
+      "loss": 1.2771,
+      "step": 1946
+    },
+    {
+      "epoch": 0.7248781282358553,
+      "grad_norm": 0.17743223905563354,
+      "learning_rate": 1.99338440092202e-05,
+      "loss": 1.276,
+      "step": 1947
+    },
+    {
+      "epoch": 0.7252504333864643,
+      "grad_norm": 0.18026785552501678,
+      "learning_rate": 1.9933704373133967e-05,
+      "loss": 1.2737,
+      "step": 1948
+    },
+    {
+      "epoch": 0.7256227385370735,
+      "grad_norm": 0.191921204328537,
+      "learning_rate": 1.993356459032755e-05,
+      "loss": 1.2722,
+      "step": 1949
+    },
+    {
+      "epoch": 0.7259950436876825,
+      "grad_norm": 0.1790904849767685,
+      "learning_rate": 1.9933424660803006e-05,
+      "loss": 1.2661,
+      "step": 1950
+    },
+    {
+      "epoch": 0.7263673488382916,
+      "grad_norm": 0.19434896111488342,
+      "learning_rate": 1.99332845845624e-05,
+      "loss": 1.2543,
+      "step": 1951
+    },
+    {
+      "epoch": 0.7267396539889006,
+      "grad_norm": 0.17882947623729706,
+      "learning_rate": 1.993314436160781e-05,
+      "loss": 1.2631,
+      "step": 1952
+    },
+    {
+      "epoch": 0.7271119591395098,
+      "grad_norm": 0.18903744220733643,
+      "learning_rate": 1.99330039919413e-05,
+      "loss": 1.2721,
+      "step": 1953
+    },
+    {
+      "epoch": 0.7274842642901188,
+      "grad_norm": 0.17733576893806458,
+      "learning_rate": 1.993286347556494e-05,
+      "loss": 1.2782,
+      "step": 1954
+    },
+    {
+      "epoch": 0.7278565694407279,
+      "grad_norm": 0.1745980829000473,
+      "learning_rate": 1.9932722812480813e-05,
+      "loss": 1.2609,
+      "step": 1955
+    },
+    {
+      "epoch": 0.7282288745913369,
+      "grad_norm": 0.18172860145568848,
+      "learning_rate": 1.9932582002690993e-05,
+      "loss": 1.2645,
+      "step": 1956
+    },
+    {
+      "epoch": 0.728601179741946,
+      "grad_norm": 0.17440102994441986,
+      "learning_rate": 1.9932441046197558e-05,
+      "loss": 1.2739,
+      "step": 1957
+    },
+    {
+      "epoch": 0.7289734848925551,
+      "grad_norm": 0.17599335312843323,
+      "learning_rate": 1.9932299943002596e-05,
+      "loss": 1.2626,
+      "step": 1958
+    },
+    {
+      "epoch": 0.7293457900431641,
+      "grad_norm": 0.18249286711215973,
+      "learning_rate": 1.9932158693108183e-05,
+      "loss": 1.2629,
+      "step": 1959
+    },
+    {
+      "epoch": 0.7297180951937732,
+      "grad_norm": 0.19463662803173065,
+      "learning_rate": 1.9932017296516414e-05,
+      "loss": 1.2589,
+      "step": 1960
+    },
+    {
+      "epoch": 0.7300904003443822,
+      "grad_norm": 0.17670099437236786,
+      "learning_rate": 1.9931875753229367e-05,
+      "loss": 1.2893,
+      "step": 1961
+    },
+    {
+      "epoch": 0.7304627054949914,
+      "grad_norm": 0.181423619389534,
+      "learning_rate": 1.9931734063249143e-05,
+      "loss": 1.2583,
+      "step": 1962
+    },
+    {
+      "epoch": 0.7308350106456004,
+      "grad_norm": 0.19615116715431213,
+      "learning_rate": 1.993159222657783e-05,
+      "loss": 1.273,
+      "step": 1963
+    },
+    {
+      "epoch": 0.7312073157962095,
+      "grad_norm": 0.1963336318731308,
+      "learning_rate": 1.9931450243217522e-05,
+      "loss": 1.2694,
+      "step": 1964
+    },
+    {
+      "epoch": 0.7315796209468185,
+      "grad_norm": 0.18081746995449066,
+      "learning_rate": 1.993130811317032e-05,
+      "loss": 1.2604,
+      "step": 1965
+    },
+    {
+      "epoch": 0.7319519260974277,
+      "grad_norm": 0.19056904315948486,
+      "learning_rate": 1.9931165836438314e-05,
+      "loss": 1.2732,
+      "step": 1966
+    },
+    {
+      "epoch": 0.7323242312480367,
+      "grad_norm": 0.18734106421470642,
+      "learning_rate": 1.9931023413023615e-05,
+      "loss": 1.2747,
+      "step": 1967
+    },
+    {
+      "epoch": 0.7326965363986457,
+      "grad_norm": 0.17514047026634216,
+      "learning_rate": 1.9930880842928325e-05,
+      "loss": 1.2536,
+      "step": 1968
+    },
+    {
+      "epoch": 0.7330688415492548,
+      "grad_norm": 0.17114852368831635,
+      "learning_rate": 1.993073812615455e-05,
+      "loss": 1.2664,
+      "step": 1969
+    },
+    {
+      "epoch": 0.7334411466998638,
+      "grad_norm": 0.187078058719635,
+      "learning_rate": 1.993059526270439e-05,
+      "loss": 1.265,
+      "step": 1970
+    },
+    {
+      "epoch": 0.733813451850473,
+      "grad_norm": 0.1790546178817749,
+      "learning_rate": 1.9930452252579967e-05,
+      "loss": 1.2676,
+      "step": 1971
+    },
+    {
+      "epoch": 0.734185757001082,
+      "grad_norm": 0.18027950823307037,
+      "learning_rate": 1.9930309095783386e-05,
+      "loss": 1.2745,
+      "step": 1972
+    },
+    {
+      "epoch": 0.7345580621516911,
+      "grad_norm": 0.18280746042728424,
+      "learning_rate": 1.993016579231676e-05,
+      "loss": 1.2724,
+      "step": 1973
+    },
+    {
+      "epoch": 0.7349303673023001,
+      "grad_norm": 0.18688839673995972,
+      "learning_rate": 1.9930022342182213e-05,
+      "loss": 1.2695,
+      "step": 1974
+    },
+    {
+      "epoch": 0.7353026724529093,
+      "grad_norm": 0.17303088307380676,
+      "learning_rate": 1.9929878745381855e-05,
+      "loss": 1.2601,
+      "step": 1975
+    },
+    {
+      "epoch": 0.7356749776035183,
+      "grad_norm": 0.17247672379016876,
+      "learning_rate": 1.992973500191781e-05,
+      "loss": 1.2698,
+      "step": 1976
+    },
+    {
+      "epoch": 0.7360472827541273,
+      "grad_norm": 0.18683381378650665,
+      "learning_rate": 1.9929591111792206e-05,
+      "loss": 1.2856,
+      "step": 1977
+    },
+    {
+      "epoch": 0.7364195879047364,
+      "grad_norm": 0.17574839293956757,
+      "learning_rate": 1.9929447075007164e-05,
+      "loss": 1.2798,
+      "step": 1978
+    },
+    {
+      "epoch": 0.7367918930553455,
+      "grad_norm": 0.177045539021492,
+      "learning_rate": 1.992930289156481e-05,
+      "loss": 1.2671,
+      "step": 1979
+    },
+    {
+      "epoch": 0.7371641982059546,
+      "grad_norm": 0.1845403015613556,
+      "learning_rate": 1.9929158561467276e-05,
+      "loss": 1.2516,
+      "step": 1980
+    },
+    {
+      "epoch": 0.7375365033565636,
+      "grad_norm": 0.18033714592456818,
+      "learning_rate": 1.9929014084716695e-05,
+      "loss": 1.2661,
+      "step": 1981
+    },
+    {
+      "epoch": 0.7379088085071727,
+      "grad_norm": 0.17001290619373322,
+      "learning_rate": 1.9928869461315197e-05,
+      "loss": 1.2729,
+      "step": 1982
+    },
+    {
+      "epoch": 0.7382811136577817,
+      "grad_norm": 0.1728724241256714,
+      "learning_rate": 1.992872469126492e-05,
+      "loss": 1.2679,
+      "step": 1983
+    },
+    {
+      "epoch": 0.7386534188083909,
+      "grad_norm": 0.1804700642824173,
+      "learning_rate": 1.9928579774568005e-05,
+      "loss": 1.2785,
+      "step": 1984
+    },
+    {
+      "epoch": 0.7390257239589999,
+      "grad_norm": 0.174544095993042,
+      "learning_rate": 1.9928434711226586e-05,
+      "loss": 1.2484,
+      "step": 1985
+    },
+    {
+      "epoch": 0.739398029109609,
+      "grad_norm": 0.18159344792366028,
+      "learning_rate": 1.9928289501242812e-05,
+      "loss": 1.2658,
+      "step": 1986
+    },
+    {
+      "epoch": 0.739770334260218,
+      "grad_norm": 0.16609835624694824,
+      "learning_rate": 1.9928144144618824e-05,
+      "loss": 1.266,
+      "step": 1987
+    },
+    {
+      "epoch": 0.7401426394108271,
+      "grad_norm": 0.1732049435377121,
+      "learning_rate": 1.992799864135677e-05,
+      "loss": 1.2598,
+      "step": 1988
+    },
+    {
+      "epoch": 0.7405149445614362,
+      "grad_norm": 0.17838548123836517,
+      "learning_rate": 1.9927852991458802e-05,
+      "loss": 1.2866,
+      "step": 1989
+    },
+    {
+      "epoch": 0.7408872497120452,
+      "grad_norm": 0.17985506355762482,
+      "learning_rate": 1.9927707194927067e-05,
+      "loss": 1.2706,
+      "step": 1990
+    },
+    {
+      "epoch": 0.7412595548626543,
+      "grad_norm": 0.1812310814857483,
+      "learning_rate": 1.9927561251763717e-05,
+      "loss": 1.2751,
+      "step": 1991
+    },
+    {
+      "epoch": 0.7416318600132634,
+      "grad_norm": 0.1740981936454773,
+      "learning_rate": 1.9927415161970913e-05,
+      "loss": 1.2769,
+      "step": 1992
+    },
+    {
+      "epoch": 0.7420041651638725,
+      "grad_norm": 0.1680895835161209,
+      "learning_rate": 1.9927268925550808e-05,
+      "loss": 1.2502,
+      "step": 1993
+    },
+    {
+      "epoch": 0.7423764703144815,
+      "grad_norm": 0.17788243293762207,
+      "learning_rate": 1.992712254250557e-05,
+      "loss": 1.2735,
+      "step": 1994
+    },
+    {
+      "epoch": 0.7427487754650905,
+      "grad_norm": 0.17746172845363617,
+      "learning_rate": 1.9926976012837345e-05,
+      "loss": 1.2665,
+      "step": 1995
+    },
+    {
+      "epoch": 0.7431210806156996,
+      "grad_norm": 0.1807558834552765,
+      "learning_rate": 1.9926829336548314e-05,
+      "loss": 1.263,
+      "step": 1996
+    },
+    {
+      "epoch": 0.7434933857663087,
+      "grad_norm": 0.17683039605617523,
+      "learning_rate": 1.9926682513640634e-05,
+      "loss": 1.2665,
+      "step": 1997
+    },
+    {
+      "epoch": 0.7438656909169178,
+      "grad_norm": 0.1726188063621521,
+      "learning_rate": 1.992653554411648e-05,
+      "loss": 1.279,
+      "step": 1998
+    },
+    {
+      "epoch": 0.7442379960675268,
+      "grad_norm": 0.1840149462223053,
+      "learning_rate": 1.9926388427978016e-05,
+      "loss": 1.267,
+      "step": 1999
+    },
+    {
+      "epoch": 0.7446103012181359,
+      "grad_norm": 0.17595337331295013,
+      "learning_rate": 1.992624116522742e-05,
+      "loss": 1.264,
+      "step": 2000
+    },
+    {
+      "epoch": 0.7446103012181359,
+      "eval_loss": 1.3380753993988037,
+      "eval_runtime": 15.9929,
+      "eval_samples_per_second": 108.423,
+      "eval_steps_per_second": 5.44,
+      "step": 2000
+    },
+    {
+      "epoch": 0.744982606368745,
+      "grad_norm": 0.17696943879127502,
+      "learning_rate": 1.9926093755866862e-05,
+      "loss": 1.2773,
+      "step": 2001
+    },
+    {
+      "epoch": 0.7453549115193541,
+      "grad_norm": 0.1840478777885437,
+      "learning_rate": 1.9925946199898526e-05,
+      "loss": 1.2484,
+      "step": 2002
+    },
+    {
+      "epoch": 0.7457272166699631,
+      "grad_norm": 0.16878820955753326,
+      "learning_rate": 1.9925798497324583e-05,
+      "loss": 1.2478,
+      "step": 2003
+    },
+    {
+      "epoch": 0.7460995218205722,
+      "grad_norm": 0.17494863271713257,
+      "learning_rate": 1.992565064814722e-05,
+      "loss": 1.2547,
+      "step": 2004
+    },
+    {
+      "epoch": 0.7464718269711813,
+      "grad_norm": 0.17591069638729095,
+      "learning_rate": 1.992550265236862e-05,
+      "loss": 1.2644,
+      "step": 2005
+    },
+    {
+      "epoch": 0.7468441321217903,
+      "grad_norm": 0.17405180633068085,
+      "learning_rate": 1.992535450999097e-05,
+      "loss": 1.2685,
+      "step": 2006
+    },
+    {
+      "epoch": 0.7472164372723994,
+      "grad_norm": 0.17987771332263947,
+      "learning_rate": 1.9925206221016456e-05,
+      "loss": 1.2792,
+      "step": 2007
+    },
+    {
+      "epoch": 0.7475887424230084,
+      "grad_norm": 0.1718064546585083,
+      "learning_rate": 1.992505778544727e-05,
+      "loss": 1.2552,
+      "step": 2008
+    },
+    {
+      "epoch": 0.7479610475736175,
+      "grad_norm": 0.16629862785339355,
+      "learning_rate": 1.9924909203285604e-05,
+      "loss": 1.2621,
+      "step": 2009
+    },
+    {
+      "epoch": 0.7483333527242266,
+      "grad_norm": 0.16779156029224396,
+      "learning_rate": 1.9924760474533654e-05,
+      "loss": 1.2569,
+      "step": 2010
+    },
+    {
+      "epoch": 0.7487056578748357,
+      "grad_norm": 0.17635951936244965,
+      "learning_rate": 1.992461159919361e-05,
+      "loss": 1.2617,
+      "step": 2011
+    },
+    {
+      "epoch": 0.7490779630254447,
+      "grad_norm": 0.17284058034420013,
+      "learning_rate": 1.9924462577267676e-05,
+      "loss": 1.2622,
+      "step": 2012
+    },
+    {
+      "epoch": 0.7494502681760538,
+      "grad_norm": 0.16899628937244415,
+      "learning_rate": 1.9924313408758053e-05,
+      "loss": 1.2684,
+      "step": 2013
+    },
+    {
+      "epoch": 0.7498225733266629,
+      "grad_norm": 0.17767173051834106,
+      "learning_rate": 1.9924164093666946e-05,
+      "loss": 1.2624,
+      "step": 2014
+    },
+    {
+      "epoch": 0.750194878477272,
+      "grad_norm": 0.17249037325382233,
+      "learning_rate": 1.9924014631996557e-05,
+      "loss": 1.2647,
+      "step": 2015
+    },
+    {
+      "epoch": 0.750567183627881,
+      "grad_norm": 0.1744518280029297,
+      "learning_rate": 1.9923865023749095e-05,
+      "loss": 1.2541,
+      "step": 2016
+    },
+    {
+      "epoch": 0.75093948877849,
+      "grad_norm": 0.17807431519031525,
+      "learning_rate": 1.9923715268926765e-05,
+      "loss": 1.2731,
+      "step": 2017
+    },
+    {
+      "epoch": 0.7513117939290992,
+      "grad_norm": 0.17543937265872955,
+      "learning_rate": 1.992356536753179e-05,
+      "loss": 1.2463,
+      "step": 2018
+    },
+    {
+      "epoch": 0.7516840990797082,
+      "grad_norm": 0.17383424937725067,
+      "learning_rate": 1.9923415319566372e-05,
+      "loss": 1.2479,
+      "step": 2019
+    },
+    {
+      "epoch": 0.7520564042303173,
+      "grad_norm": 0.181121364235878,
+      "learning_rate": 1.9923265125032736e-05,
+      "loss": 1.2617,
+      "step": 2020
+    },
+    {
+      "epoch": 0.7524287093809263,
+      "grad_norm": 0.18264862895011902,
+      "learning_rate": 1.9923114783933096e-05,
+      "loss": 1.2607,
+      "step": 2021
+    },
+    {
+      "epoch": 0.7528010145315354,
+      "grad_norm": 0.1814924031496048,
+      "learning_rate": 1.9922964296269672e-05,
+      "loss": 1.2828,
+      "step": 2022
+    },
+    {
+      "epoch": 0.7531733196821445,
+      "grad_norm": 0.17525698244571686,
+      "learning_rate": 1.992281366204469e-05,
+      "loss": 1.2538,
+      "step": 2023
+    },
+    {
+      "epoch": 0.7535456248327536,
+      "grad_norm": 0.17692695558071136,
+      "learning_rate": 1.9922662881260374e-05,
+      "loss": 1.2701,
+      "step": 2024
+    },
+    {
+      "epoch": 0.7539179299833626,
+      "grad_norm": 0.17601527273654938,
+      "learning_rate": 1.9922511953918945e-05,
+      "loss": 1.2625,
+      "step": 2025
+    },
+    {
+      "epoch": 0.7542902351339716,
+      "grad_norm": 0.17493323981761932,
+      "learning_rate": 1.992236088002264e-05,
+      "loss": 1.256,
+      "step": 2026
+    },
+    {
+      "epoch": 0.7546625402845808,
+      "grad_norm": 0.1716911792755127,
+      "learning_rate": 1.992220965957369e-05,
+      "loss": 1.2375,
+      "step": 2027
+    },
+    {
+      "epoch": 0.7550348454351898,
+      "grad_norm": 0.1788312941789627,
+      "learning_rate": 1.9922058292574323e-05,
+      "loss": 1.2571,
+      "step": 2028
+    },
+    {
+      "epoch": 0.7554071505857989,
+      "grad_norm": 0.1795925498008728,
+      "learning_rate": 1.9921906779026775e-05,
+      "loss": 1.2737,
+      "step": 2029
+    },
+    {
+      "epoch": 0.7557794557364079,
+      "grad_norm": 0.17507243156433105,
+      "learning_rate": 1.9921755118933292e-05,
+      "loss": 1.2625,
+      "step": 2030
+    },
+    {
+      "epoch": 0.756151760887017,
+      "grad_norm": 0.17492863535881042,
+      "learning_rate": 1.992160331229611e-05,
+      "loss": 1.268,
+      "step": 2031
+    },
+    {
+      "epoch": 0.7565240660376261,
+      "grad_norm": 0.18614165484905243,
+      "learning_rate": 1.992145135911746e-05,
+      "loss": 1.2878,
+      "step": 2032
+    },
+    {
+      "epoch": 0.7568963711882352,
+      "grad_norm": 0.18001849949359894,
+      "learning_rate": 1.9921299259399604e-05,
+      "loss": 1.2622,
+      "step": 2033
+    },
+    {
+      "epoch": 0.7572686763388442,
+      "grad_norm": 0.18375787138938904,
+      "learning_rate": 1.9921147013144782e-05,
+      "loss": 1.2825,
+      "step": 2034
+    },
+    {
+      "epoch": 0.7576409814894532,
+      "grad_norm": 0.17793749272823334,
+      "learning_rate": 1.9920994620355236e-05,
+      "loss": 1.2546,
+      "step": 2035
+    },
+    {
+      "epoch": 0.7580132866400624,
+      "grad_norm": 0.18328694999217987,
+      "learning_rate": 1.9920842081033225e-05,
+      "loss": 1.2607,
+      "step": 2036
+    },
+    {
+      "epoch": 0.7583855917906714,
+      "grad_norm": 0.17626026272773743,
+      "learning_rate": 1.9920689395180996e-05,
+      "loss": 1.27,
+      "step": 2037
+    },
+    {
+      "epoch": 0.7587578969412805,
+      "grad_norm": 0.1846167892217636,
+      "learning_rate": 1.9920536562800808e-05,
+      "loss": 1.2657,
+      "step": 2038
+    },
+    {
+      "epoch": 0.7591302020918895,
+      "grad_norm": 0.1859884262084961,
+      "learning_rate": 1.9920383583894922e-05,
+      "loss": 1.2806,
+      "step": 2039
+    },
+    {
+      "epoch": 0.7595025072424987,
+      "grad_norm": 0.17855176329612732,
+      "learning_rate": 1.992023045846559e-05,
+      "loss": 1.2725,
+      "step": 2040
+    },
+    {
+      "epoch": 0.7598748123931077,
+      "grad_norm": 0.17177069187164307,
+      "learning_rate": 1.9920077186515076e-05,
+      "loss": 1.2623,
+      "step": 2041
+    },
+    {
+      "epoch": 0.7602471175437168,
+      "grad_norm": 0.18153630197048187,
+      "learning_rate": 1.9919923768045646e-05,
+      "loss": 1.2827,
+      "step": 2042
+    },
+    {
+      "epoch": 0.7606194226943258,
+      "grad_norm": 0.18789653480052948,
+      "learning_rate": 1.9919770203059564e-05,
+      "loss": 1.2675,
+      "step": 2043
+    },
+    {
+      "epoch": 0.7609917278449349,
+      "grad_norm": 0.18295817077159882,
+      "learning_rate": 1.99196164915591e-05,
+      "loss": 1.2635,
+      "step": 2044
+    },
+    {
+      "epoch": 0.761364032995544,
+      "grad_norm": 0.18244396150112152,
+      "learning_rate": 1.991946263354652e-05,
+      "loss": 1.2586,
+      "step": 2045
+    },
+    {
+      "epoch": 0.761736338146153,
+      "grad_norm": 0.19641557335853577,
+      "learning_rate": 1.99193086290241e-05,
+      "loss": 1.2838,
+      "step": 2046
+    },
+    {
+      "epoch": 0.7621086432967621,
+      "grad_norm": 0.1869594007730484,
+      "learning_rate": 1.9919154477994117e-05,
+      "loss": 1.262,
+      "step": 2047
+    },
+    {
+      "epoch": 0.7624809484473711,
+      "grad_norm": 0.1961170732975006,
+      "learning_rate": 1.991900018045884e-05,
+      "loss": 1.2574,
+      "step": 2048
+    },
+    {
+      "epoch": 0.7628532535979803,
+      "grad_norm": 0.18542712926864624,
+      "learning_rate": 1.9918845736420554e-05,
+      "loss": 1.2627,
+      "step": 2049
+    },
+    {
+      "epoch": 0.7632255587485893,
+      "grad_norm": 0.19344452023506165,
+      "learning_rate": 1.9918691145881542e-05,
+      "loss": 1.2757,
+      "step": 2050
+    },
+    {
+      "epoch": 0.7635978638991984,
+      "grad_norm": 0.19451677799224854,
+      "learning_rate": 1.9918536408844082e-05,
+      "loss": 1.265,
+      "step": 2051
+    },
+    {
+      "epoch": 0.7639701690498074,
+      "grad_norm": 0.19341467320919037,
+      "learning_rate": 1.9918381525310464e-05,
+      "loss": 1.2655,
+      "step": 2052
+    },
+    {
+      "epoch": 0.7643424742004166,
+      "grad_norm": 0.1805962473154068,
+      "learning_rate": 1.991822649528297e-05,
+      "loss": 1.2639,
+      "step": 2053
+    },
+    {
+      "epoch": 0.7647147793510256,
+      "grad_norm": 0.19268319010734558,
+      "learning_rate": 1.9918071318763898e-05,
+      "loss": 1.2563,
+      "step": 2054
+    },
+    {
+      "epoch": 0.7650870845016347,
+      "grad_norm": 0.1859290599822998,
+      "learning_rate": 1.991791599575553e-05,
+      "loss": 1.2654,
+      "step": 2055
+    },
+    {
+      "epoch": 0.7654593896522437,
+      "grad_norm": 0.18508413434028625,
+      "learning_rate": 1.991776052626017e-05,
+      "loss": 1.2577,
+      "step": 2056
+    },
+    {
+      "epoch": 0.7658316948028527,
+      "grad_norm": 0.19763630628585815,
+      "learning_rate": 1.9917604910280106e-05,
+      "loss": 1.2622,
+      "step": 2057
+    },
+    {
+      "epoch": 0.7662039999534619,
+      "grad_norm": 0.16519156098365784,
+      "learning_rate": 1.991744914781764e-05,
+      "loss": 1.2553,
+      "step": 2058
+    },
+    {
+      "epoch": 0.7665763051040709,
+      "grad_norm": 0.191755473613739,
+      "learning_rate": 1.991729323887507e-05,
+      "loss": 1.2623,
+      "step": 2059
+    },
+    {
+      "epoch": 0.76694861025468,
+      "grad_norm": 0.19937783479690552,
+      "learning_rate": 1.9917137183454706e-05,
+      "loss": 1.2631,
+      "step": 2060
+    },
+    {
+      "epoch": 0.767320915405289,
+      "grad_norm": 0.17880156636238098,
+      "learning_rate": 1.9916980981558846e-05,
+      "loss": 1.2731,
+      "step": 2061
+    },
+    {
+      "epoch": 0.7676932205558982,
+      "grad_norm": 0.17433685064315796,
+      "learning_rate": 1.99168246331898e-05,
+      "loss": 1.2617,
+      "step": 2062
+    },
+    {
+      "epoch": 0.7680655257065072,
+      "grad_norm": 0.19513197243213654,
+      "learning_rate": 1.9916668138349873e-05,
+      "loss": 1.2746,
+      "step": 2063
+    },
+    {
+      "epoch": 0.7684378308571163,
+      "grad_norm": 0.16553771495819092,
+      "learning_rate": 1.9916511497041388e-05,
+      "loss": 1.2607,
+      "step": 2064
+    },
+    {
+      "epoch": 0.7688101360077253,
+      "grad_norm": 0.1720885932445526,
+      "learning_rate": 1.9916354709266645e-05,
+      "loss": 1.2665,
+      "step": 2065
+    },
+    {
+      "epoch": 0.7691824411583345,
+      "grad_norm": 0.18258140981197357,
+      "learning_rate": 1.9916197775027967e-05,
+      "loss": 1.2586,
+      "step": 2066
+    },
+    {
+      "epoch": 0.7695547463089435,
+      "grad_norm": 0.18003836274147034,
+      "learning_rate": 1.991604069432767e-05,
+      "loss": 1.2594,
+      "step": 2067
+    },
+    {
+      "epoch": 0.7699270514595525,
+      "grad_norm": 0.1830645203590393,
+      "learning_rate": 1.991588346716807e-05,
+      "loss": 1.2621,
+      "step": 2068
+    },
+    {
+      "epoch": 0.7702993566101616,
+      "grad_norm": 0.1741950958967209,
+      "learning_rate": 1.9915726093551497e-05,
+      "loss": 1.2628,
+      "step": 2069
+    },
+    {
+      "epoch": 0.7706716617607706,
+      "grad_norm": 0.17035536468029022,
+      "learning_rate": 1.991556857348027e-05,
+      "loss": 1.2504,
+      "step": 2070
+    },
+    {
+      "epoch": 0.7710439669113798,
+      "grad_norm": 0.1875630021095276,
+      "learning_rate": 1.9915410906956723e-05,
+      "loss": 1.2616,
+      "step": 2071
+    },
+    {
+      "epoch": 0.7714162720619888,
+      "grad_norm": 0.1755642145872116,
+      "learning_rate": 1.9915253093983175e-05,
+      "loss": 1.2758,
+      "step": 2072
+    },
+    {
+      "epoch": 0.7717885772125979,
+      "grad_norm": 0.17393867671489716,
+      "learning_rate": 1.991509513456196e-05,
+      "loss": 1.2584,
+      "step": 2073
+    },
+    {
+      "epoch": 0.7721608823632069,
+      "grad_norm": 0.186118945479393,
+      "learning_rate": 1.9914937028695412e-05,
+      "loss": 1.2733,
+      "step": 2074
+    },
+    {
+      "epoch": 0.772533187513816,
+      "grad_norm": 0.1890277862548828,
+      "learning_rate": 1.991477877638587e-05,
+      "loss": 1.2652,
+      "step": 2075
+    },
+    {
+      "epoch": 0.7729054926644251,
+      "grad_norm": 0.18629412353038788,
+      "learning_rate": 1.9914620377635666e-05,
+      "loss": 1.2573,
+      "step": 2076
+    },
+    {
+      "epoch": 0.7732777978150341,
+      "grad_norm": 0.17759479582309723,
+      "learning_rate": 1.9914461832447142e-05,
+      "loss": 1.2551,
+      "step": 2077
+    },
+    {
+      "epoch": 0.7736501029656432,
+      "grad_norm": 0.1834801733493805,
+      "learning_rate": 1.9914303140822634e-05,
+      "loss": 1.252,
+      "step": 2078
+    },
+    {
+      "epoch": 0.7740224081162523,
+      "grad_norm": 0.17973852157592773,
+      "learning_rate": 1.9914144302764497e-05,
+      "loss": 1.2553,
+      "step": 2079
+    },
+    {
+      "epoch": 0.7743947132668614,
+      "grad_norm": 0.17896102368831635,
+      "learning_rate": 1.9913985318275068e-05,
+      "loss": 1.2553,
+      "step": 2080
+    },
+    {
+      "epoch": 0.7747670184174704,
+      "grad_norm": 0.18261843919754028,
+      "learning_rate": 1.99138261873567e-05,
+      "loss": 1.2491,
+      "step": 2081
+    },
+    {
+      "epoch": 0.7751393235680795,
+      "grad_norm": 0.19672666490077972,
+      "learning_rate": 1.9913666910011737e-05,
+      "loss": 1.2655,
+      "step": 2082
+    },
+    {
+      "epoch": 0.7755116287186885,
+      "grad_norm": 0.18677249550819397,
+      "learning_rate": 1.9913507486242537e-05,
+      "loss": 1.2749,
+      "step": 2083
+    },
+    {
+      "epoch": 0.7758839338692977,
+      "grad_norm": 0.18855483829975128,
+      "learning_rate": 1.9913347916051458e-05,
+      "loss": 1.2729,
+      "step": 2084
+    },
+    {
+      "epoch": 0.7762562390199067,
+      "grad_norm": 0.18166404962539673,
+      "learning_rate": 1.9913188199440848e-05,
+      "loss": 1.253,
+      "step": 2085
+    },
+    {
+      "epoch": 0.7766285441705157,
+      "grad_norm": 0.18028071522712708,
+      "learning_rate": 1.9913028336413074e-05,
+      "loss": 1.2587,
+      "step": 2086
+    },
+    {
+      "epoch": 0.7770008493211248,
+      "grad_norm": 0.18412554264068604,
+      "learning_rate": 1.991286832697049e-05,
+      "loss": 1.2503,
+      "step": 2087
+    },
+    {
+      "epoch": 0.7773731544717339,
+      "grad_norm": 0.18333709239959717,
+      "learning_rate": 1.9912708171115463e-05,
+      "loss": 1.2636,
+      "step": 2088
+    },
+    {
+      "epoch": 0.777745459622343,
+      "grad_norm": 0.18019062280654907,
+      "learning_rate": 1.991254786885036e-05,
+      "loss": 1.2577,
+      "step": 2089
+    },
+    {
+      "epoch": 0.778117764772952,
+      "grad_norm": 0.18314556777477264,
+      "learning_rate": 1.991238742017755e-05,
+      "loss": 1.2482,
+      "step": 2090
+    },
+    {
+      "epoch": 0.7784900699235611,
+      "grad_norm": 0.17646336555480957,
+      "learning_rate": 1.9912226825099395e-05,
+      "loss": 1.262,
+      "step": 2091
+    },
+    {
+      "epoch": 0.7788623750741702,
+      "grad_norm": 0.19445347785949707,
+      "learning_rate": 1.9912066083618275e-05,
+      "loss": 1.256,
+      "step": 2092
+    },
+    {
+      "epoch": 0.7792346802247793,
+      "grad_norm": 0.17373862862586975,
+      "learning_rate": 1.991190519573656e-05,
+      "loss": 1.2702,
+      "step": 2093
+    },
+    {
+      "epoch": 0.7796069853753883,
+      "grad_norm": 0.16923663020133972,
+      "learning_rate": 1.9911744161456624e-05,
+      "loss": 1.2482,
+      "step": 2094
+    },
+    {
+      "epoch": 0.7799792905259973,
+      "grad_norm": 0.1783125400543213,
+      "learning_rate": 1.9911582980780854e-05,
+      "loss": 1.2623,
+      "step": 2095
+    },
+    {
+      "epoch": 0.7803515956766064,
+      "grad_norm": 0.18715058267116547,
+      "learning_rate": 1.9911421653711624e-05,
+      "loss": 1.2634,
+      "step": 2096
+    },
+    {
+      "epoch": 0.7807239008272155,
+      "grad_norm": 0.17201177775859833,
+      "learning_rate": 1.9911260180251316e-05,
+      "loss": 1.2659,
+      "step": 2097
+    },
+    {
+      "epoch": 0.7810962059778246,
+      "grad_norm": 0.1753210425376892,
+      "learning_rate": 1.991109856040232e-05,
+      "loss": 1.2558,
+      "step": 2098
+    },
+    {
+      "epoch": 0.7814685111284336,
+      "grad_norm": 0.16961847245693207,
+      "learning_rate": 1.991093679416702e-05,
+      "loss": 1.2614,
+      "step": 2099
+    },
+    {
+      "epoch": 0.7818408162790427,
+      "grad_norm": 0.18109846115112305,
+      "learning_rate": 1.9910774881547803e-05,
+      "loss": 1.2597,
+      "step": 2100
+    },
+    {
+      "epoch": 0.7822131214296518,
+      "grad_norm": 0.16653625667095184,
+      "learning_rate": 1.9910612822547063e-05,
+      "loss": 1.2603,
+      "step": 2101
+    },
+    {
+      "epoch": 0.7825854265802609,
+      "grad_norm": 0.17125585675239563,
+      "learning_rate": 1.9910450617167198e-05,
+      "loss": 1.2597,
+      "step": 2102
+    },
+    {
+      "epoch": 0.7829577317308699,
+      "grad_norm": 0.18307803571224213,
+      "learning_rate": 1.9910288265410593e-05,
+      "loss": 1.2728,
+      "step": 2103
+    },
+    {
+      "epoch": 0.783330036881479,
+      "grad_norm": 0.17665451765060425,
+      "learning_rate": 1.9910125767279655e-05,
+      "loss": 1.2727,
+      "step": 2104
+    },
+    {
+      "epoch": 0.783702342032088,
+      "grad_norm": 0.188068225979805,
+      "learning_rate": 1.9909963122776785e-05,
+      "loss": 1.2526,
+      "step": 2105
+    },
+    {
+      "epoch": 0.7840746471826971,
+      "grad_norm": 0.1769012063741684,
+      "learning_rate": 1.9909800331904375e-05,
+      "loss": 1.2541,
+      "step": 2106
+    },
+    {
+      "epoch": 0.7844469523333062,
+      "grad_norm": 0.1765563189983368,
+      "learning_rate": 1.9909637394664842e-05,
+      "loss": 1.2499,
+      "step": 2107
+    },
+    {
+      "epoch": 0.7848192574839152,
+      "grad_norm": 0.19480668008327484,
+      "learning_rate": 1.9909474311060583e-05,
+      "loss": 1.2617,
+      "step": 2108
+    },
+    {
+      "epoch": 0.7851915626345243,
+      "grad_norm": 0.1799997091293335,
+      "learning_rate": 1.9909311081094012e-05,
+      "loss": 1.2636,
+      "step": 2109
+    },
+    {
+      "epoch": 0.7855638677851334,
+      "grad_norm": 0.16892112791538239,
+      "learning_rate": 1.9909147704767537e-05,
+      "loss": 1.2525,
+      "step": 2110
+    },
+    {
+      "epoch": 0.7859361729357425,
+      "grad_norm": 0.17401593923568726,
+      "learning_rate": 1.9908984182083574e-05,
+      "loss": 1.2497,
+      "step": 2111
+    },
+    {
+      "epoch": 0.7863084780863515,
+      "grad_norm": 0.18398383259773254,
+      "learning_rate": 1.9908820513044535e-05,
+      "loss": 1.2704,
+      "step": 2112
+    },
+    {
+      "epoch": 0.7866807832369606,
+      "grad_norm": 0.17597925662994385,
+      "learning_rate": 1.990865669765284e-05,
+      "loss": 1.2628,
+      "step": 2113
+    },
+    {
+      "epoch": 0.7870530883875697,
+      "grad_norm": 0.17537419497966766,
+      "learning_rate": 1.9908492735910907e-05,
+      "loss": 1.2674,
+      "step": 2114
+    },
+    {
+      "epoch": 0.7874253935381788,
+      "grad_norm": 0.198894664645195,
+      "learning_rate": 1.990832862782116e-05,
+      "loss": 1.273,
+      "step": 2115
+    },
+    {
+      "epoch": 0.7877976986887878,
+      "grad_norm": 0.18598125874996185,
+      "learning_rate": 1.9908164373386016e-05,
+      "loss": 1.2714,
+      "step": 2116
+    },
+    {
+      "epoch": 0.7881700038393968,
+      "grad_norm": 0.18410056829452515,
+      "learning_rate": 1.990799997260791e-05,
+      "loss": 1.2774,
+      "step": 2117
+    },
+    {
+      "epoch": 0.7885423089900059,
+      "grad_norm": 0.17665806412696838,
+      "learning_rate": 1.9907835425489263e-05,
+      "loss": 1.2697,
+      "step": 2118
+    },
+    {
+      "epoch": 0.788914614140615,
+      "grad_norm": 0.18189984560012817,
+      "learning_rate": 1.990767073203251e-05,
+      "loss": 1.2582,
+      "step": 2119
+    },
+    {
+      "epoch": 0.7892869192912241,
+      "grad_norm": 0.18040038645267487,
+      "learning_rate": 1.9907505892240084e-05,
+      "loss": 1.2587,
+      "step": 2120
+    },
+    {
+      "epoch": 0.7896592244418331,
+      "grad_norm": 0.18493278324604034,
+      "learning_rate": 1.9907340906114418e-05,
+      "loss": 1.2563,
+      "step": 2121
+    },
+    {
+      "epoch": 0.7900315295924422,
+      "grad_norm": 0.19034142792224884,
+      "learning_rate": 1.9907175773657945e-05,
+      "loss": 1.2633,
+      "step": 2122
+    },
+    {
+      "epoch": 0.7904038347430513,
+      "grad_norm": 0.18734246492385864,
+      "learning_rate": 1.990701049487311e-05,
+      "loss": 1.2665,
+      "step": 2123
+    },
+    {
+      "epoch": 0.7907761398936604,
+      "grad_norm": 0.18347962200641632,
+      "learning_rate": 1.9906845069762352e-05,
+      "loss": 1.2608,
+      "step": 2124
+    },
+    {
+      "epoch": 0.7911484450442694,
+      "grad_norm": 0.18983064591884613,
+      "learning_rate": 1.9906679498328114e-05,
+      "loss": 1.2524,
+      "step": 2125
+    },
+    {
+      "epoch": 0.7915207501948784,
+      "grad_norm": 0.17927458882331848,
+      "learning_rate": 1.990651378057284e-05,
+      "loss": 1.2826,
+      "step": 2126
+    },
+    {
+      "epoch": 0.7918930553454876,
+      "grad_norm": 0.1906890869140625,
+      "learning_rate": 1.990634791649898e-05,
+      "loss": 1.2494,
+      "step": 2127
+    },
+    {
+      "epoch": 0.7922653604960966,
+      "grad_norm": 0.19961805641651154,
+      "learning_rate": 1.9906181906108983e-05,
+      "loss": 1.2436,
+      "step": 2128
+    },
+    {
+      "epoch": 0.7926376656467057,
+      "grad_norm": 0.18504062294960022,
+      "learning_rate": 1.9906015749405302e-05,
+      "loss": 1.2439,
+      "step": 2129
+    },
+    {
+      "epoch": 0.7930099707973147,
+      "grad_norm": 0.18666766583919525,
+      "learning_rate": 1.9905849446390387e-05,
+      "loss": 1.2538,
+      "step": 2130
+    },
+    {
+      "epoch": 0.7933822759479238,
+      "grad_norm": 0.19338330626487732,
+      "learning_rate": 1.99056829970667e-05,
+      "loss": 1.2609,
+      "step": 2131
+    },
+    {
+      "epoch": 0.7937545810985329,
+      "grad_norm": 0.19320982694625854,
+      "learning_rate": 1.9905516401436698e-05,
+      "loss": 1.2667,
+      "step": 2132
+    },
+    {
+      "epoch": 0.794126886249142,
+      "grad_norm": 0.18429437279701233,
+      "learning_rate": 1.9905349659502836e-05,
+      "loss": 1.2662,
+      "step": 2133
+    },
+    {
+      "epoch": 0.794499191399751,
+      "grad_norm": 0.1833900511264801,
+      "learning_rate": 1.9905182771267583e-05,
+      "loss": 1.2662,
+      "step": 2134
+    },
+    {
+      "epoch": 0.79487149655036,
+      "grad_norm": 0.19020430743694305,
+      "learning_rate": 1.9905015736733406e-05,
+      "loss": 1.2428,
+      "step": 2135
+    },
+    {
+      "epoch": 0.7952438017009692,
+      "grad_norm": 0.18528495728969574,
+      "learning_rate": 1.9904848555902764e-05,
+      "loss": 1.2643,
+      "step": 2136
+    },
+    {
+      "epoch": 0.7956161068515782,
+      "grad_norm": 0.1845444291830063,
+      "learning_rate": 1.990468122877813e-05,
+      "loss": 1.2741,
+      "step": 2137
+    },
+    {
+      "epoch": 0.7959884120021873,
+      "grad_norm": 0.1760178506374359,
+      "learning_rate": 1.9904513755361978e-05,
+      "loss": 1.2555,
+      "step": 2138
+    },
+    {
+      "epoch": 0.7963607171527963,
+      "grad_norm": 0.1901838481426239,
+      "learning_rate": 1.990434613565678e-05,
+      "loss": 1.2593,
+      "step": 2139
+    },
+    {
+      "epoch": 0.7967330223034055,
+      "grad_norm": 0.19051726162433624,
+      "learning_rate": 1.990417836966501e-05,
+      "loss": 1.2585,
+      "step": 2140
+    },
+    {
+      "epoch": 0.7971053274540145,
+      "grad_norm": 0.17843446135520935,
+      "learning_rate": 1.9904010457389144e-05,
+      "loss": 1.2613,
+      "step": 2141
+    },
+    {
+      "epoch": 0.7974776326046236,
+      "grad_norm": 0.16943950951099396,
+      "learning_rate": 1.990384239883167e-05,
+      "loss": 1.2496,
+      "step": 2142
+    },
+    {
+      "epoch": 0.7978499377552326,
+      "grad_norm": 0.19439001381397247,
+      "learning_rate": 1.9903674193995064e-05,
+      "loss": 1.2617,
+      "step": 2143
+    },
+    {
+      "epoch": 0.7982222429058417,
+      "grad_norm": 0.17899306118488312,
+      "learning_rate": 1.990350584288181e-05,
+      "loss": 1.2463,
+      "step": 2144
+    },
+    {
+      "epoch": 0.7985945480564508,
+      "grad_norm": 0.1824941188097,
+      "learning_rate": 1.99033373454944e-05,
+      "loss": 1.243,
+      "step": 2145
+    },
+    {
+      "epoch": 0.7989668532070598,
+      "grad_norm": 0.1979120671749115,
+      "learning_rate": 1.9903168701835314e-05,
+      "loss": 1.2497,
+      "step": 2146
+    },
+    {
+      "epoch": 0.7993391583576689,
+      "grad_norm": 0.18462349474430084,
+      "learning_rate": 1.990299991190705e-05,
+      "loss": 1.2597,
+      "step": 2147
+    },
+    {
+      "epoch": 0.7997114635082779,
+      "grad_norm": 0.18362213671207428,
+      "learning_rate": 1.9902830975712096e-05,
+      "loss": 1.2657,
+      "step": 2148
+    },
+    {
+      "epoch": 0.8000837686588871,
+      "grad_norm": 0.18074361979961395,
+      "learning_rate": 1.9902661893252955e-05,
+      "loss": 1.263,
+      "step": 2149
+    },
+    {
+      "epoch": 0.8004560738094961,
+      "grad_norm": 0.1755630373954773,
+      "learning_rate": 1.9902492664532116e-05,
+      "loss": 1.2558,
+      "step": 2150
+    },
+    {
+      "epoch": 0.8008283789601052,
+      "grad_norm": 0.1772383749485016,
+      "learning_rate": 1.9902323289552084e-05,
+      "loss": 1.2668,
+      "step": 2151
+    },
+    {
+      "epoch": 0.8012006841107142,
+      "grad_norm": 0.18553152680397034,
+      "learning_rate": 1.9902153768315355e-05,
+      "loss": 1.2548,
+      "step": 2152
+    },
+    {
+      "epoch": 0.8015729892613234,
+      "grad_norm": 0.18078090250492096,
+      "learning_rate": 1.9901984100824442e-05,
+      "loss": 1.2535,
+      "step": 2153
+    },
+    {
+      "epoch": 0.8019452944119324,
+      "grad_norm": 0.19977954030036926,
+      "learning_rate": 1.990181428708184e-05,
+      "loss": 1.2606,
+      "step": 2154
+    },
+    {
+      "epoch": 0.8023175995625415,
+      "grad_norm": 0.18592652678489685,
+      "learning_rate": 1.9901644327090063e-05,
+      "loss": 1.2663,
+      "step": 2155
+    },
+    {
+      "epoch": 0.8026899047131505,
+      "grad_norm": 0.17959389090538025,
+      "learning_rate": 1.9901474220851625e-05,
+      "loss": 1.2591,
+      "step": 2156
+    },
+    {
+      "epoch": 0.8030622098637595,
+      "grad_norm": 0.19428277015686035,
+      "learning_rate": 1.9901303968369028e-05,
+      "loss": 1.2703,
+      "step": 2157
+    },
+    {
+      "epoch": 0.8034345150143687,
+      "grad_norm": 0.17926783859729767,
+      "learning_rate": 1.9901133569644794e-05,
+      "loss": 1.2682,
+      "step": 2158
+    },
+    {
+      "epoch": 0.8038068201649777,
+      "grad_norm": 0.18147458136081696,
+      "learning_rate": 1.9900963024681442e-05,
+      "loss": 1.2536,
+      "step": 2159
+    },
+    {
+      "epoch": 0.8041791253155868,
+      "grad_norm": 0.18848387897014618,
+      "learning_rate": 1.990079233348149e-05,
+      "loss": 1.2543,
+      "step": 2160
+    },
+    {
+      "epoch": 0.8045514304661958,
+      "grad_norm": 0.20593321323394775,
+      "learning_rate": 1.990062149604745e-05,
+      "loss": 1.2601,
+      "step": 2161
+    },
+    {
+      "epoch": 0.804923735616805,
+      "grad_norm": 0.17839789390563965,
+      "learning_rate": 1.990045051238185e-05,
+      "loss": 1.253,
+      "step": 2162
+    },
+    {
+      "epoch": 0.805296040767414,
+      "grad_norm": 0.18125569820404053,
+      "learning_rate": 1.9900279382487223e-05,
+      "loss": 1.2479,
+      "step": 2163
+    },
+    {
+      "epoch": 0.805668345918023,
+      "grad_norm": 0.1832691729068756,
+      "learning_rate": 1.990010810636609e-05,
+      "loss": 1.2563,
+      "step": 2164
+    },
+    {
+      "epoch": 0.8060406510686321,
+      "grad_norm": 0.18866664171218872,
+      "learning_rate": 1.9899936684020983e-05,
+      "loss": 1.2379,
+      "step": 2165
+    },
+    {
+      "epoch": 0.8064129562192413,
+      "grad_norm": 0.17657117545604706,
+      "learning_rate": 1.989976511545443e-05,
+      "loss": 1.2568,
+      "step": 2166
+    },
+    {
+      "epoch": 0.8067852613698503,
+      "grad_norm": 0.17251810431480408,
+      "learning_rate": 1.989959340066897e-05,
+      "loss": 1.2435,
+      "step": 2167
+    },
+    {
+      "epoch": 0.8071575665204593,
+      "grad_norm": 0.1733696609735489,
+      "learning_rate": 1.9899421539667132e-05,
+      "loss": 1.2434,
+      "step": 2168
+    },
+    {
+      "epoch": 0.8075298716710684,
+      "grad_norm": 0.1835404336452484,
+      "learning_rate": 1.989924953245146e-05,
+      "loss": 1.2721,
+      "step": 2169
+    },
+    {
+      "epoch": 0.8079021768216774,
+      "grad_norm": 0.18328076601028442,
+      "learning_rate": 1.9899077379024497e-05,
+      "loss": 1.2567,
+      "step": 2170
+    },
+    {
+      "epoch": 0.8082744819722866,
+      "grad_norm": 0.18229638040065765,
+      "learning_rate": 1.989890507938878e-05,
+      "loss": 1.2604,
+      "step": 2171
+    },
+    {
+      "epoch": 0.8086467871228956,
+      "grad_norm": 0.18881352245807648,
+      "learning_rate": 1.989873263354686e-05,
+      "loss": 1.2579,
+      "step": 2172
+    },
+    {
+      "epoch": 0.8090190922735047,
+      "grad_norm": 0.179013192653656,
+      "learning_rate": 1.9898560041501277e-05,
+      "loss": 1.2631,
+      "step": 2173
+    },
+    {
+      "epoch": 0.8093913974241137,
+      "grad_norm": 0.18358515202999115,
+      "learning_rate": 1.9898387303254584e-05,
+      "loss": 1.2431,
+      "step": 2174
+    },
+    {
+      "epoch": 0.8097637025747229,
+      "grad_norm": 0.179754838347435,
+      "learning_rate": 1.989821441880933e-05,
+      "loss": 1.2645,
+      "step": 2175
+    },
+    {
+      "epoch": 0.8101360077253319,
+      "grad_norm": 0.18313923478126526,
+      "learning_rate": 1.989804138816807e-05,
+      "loss": 1.2684,
+      "step": 2176
+    },
+    {
+      "epoch": 0.8105083128759409,
+      "grad_norm": 0.19836069643497467,
+      "learning_rate": 1.9897868211333362e-05,
+      "loss": 1.2612,
+      "step": 2177
+    },
+    {
+      "epoch": 0.81088061802655,
+      "grad_norm": 0.19085316359996796,
+      "learning_rate": 1.9897694888307763e-05,
+      "loss": 1.2471,
+      "step": 2178
+    },
+    {
+      "epoch": 0.811252923177159,
+      "grad_norm": 0.18687599897384644,
+      "learning_rate": 1.9897521419093828e-05,
+      "loss": 1.2446,
+      "step": 2179
+    },
+    {
+      "epoch": 0.8116252283277682,
+      "grad_norm": 0.1764868050813675,
+      "learning_rate": 1.9897347803694126e-05,
+      "loss": 1.2665,
+      "step": 2180
+    },
+    {
+      "epoch": 0.8119975334783772,
+      "grad_norm": 0.2134658545255661,
+      "learning_rate": 1.9897174042111214e-05,
+      "loss": 1.2447,
+      "step": 2181
+    },
+    {
+      "epoch": 0.8123698386289863,
+      "grad_norm": 0.1877439022064209,
+      "learning_rate": 1.9897000134347665e-05,
+      "loss": 1.2686,
+      "step": 2182
+    },
+    {
+      "epoch": 0.8127421437795953,
+      "grad_norm": 0.1730710119009018,
+      "learning_rate": 1.9896826080406046e-05,
+      "loss": 1.2661,
+      "step": 2183
+    },
+    {
+      "epoch": 0.8131144489302045,
+      "grad_norm": 0.2091267704963684,
+      "learning_rate": 1.9896651880288926e-05,
+      "loss": 1.2707,
+      "step": 2184
+    },
+    {
+      "epoch": 0.8134867540808135,
+      "grad_norm": 0.19189974665641785,
+      "learning_rate": 1.9896477533998883e-05,
+      "loss": 1.2617,
+      "step": 2185
+    },
+    {
+      "epoch": 0.8138590592314225,
+      "grad_norm": 0.18066149950027466,
+      "learning_rate": 1.989630304153848e-05,
+      "loss": 1.2609,
+      "step": 2186
+    },
+    {
+      "epoch": 0.8142313643820316,
+      "grad_norm": 0.2049744427204132,
+      "learning_rate": 1.9896128402910307e-05,
+      "loss": 1.2573,
+      "step": 2187
+    },
+    {
+      "epoch": 0.8146036695326407,
+      "grad_norm": 0.1748519241809845,
+      "learning_rate": 1.9895953618116935e-05,
+      "loss": 1.2437,
+      "step": 2188
+    },
+    {
+      "epoch": 0.8149759746832498,
+      "grad_norm": 0.17954249680042267,
+      "learning_rate": 1.9895778687160954e-05,
+      "loss": 1.2665,
+      "step": 2189
+    },
+    {
+      "epoch": 0.8153482798338588,
+      "grad_norm": 0.18017897009849548,
+      "learning_rate": 1.989560361004494e-05,
+      "loss": 1.2545,
+      "step": 2190
+    },
+    {
+      "epoch": 0.8157205849844679,
+      "grad_norm": 0.20069225132465363,
+      "learning_rate": 1.9895428386771482e-05,
+      "loss": 1.2736,
+      "step": 2191
+    },
+    {
+      "epoch": 0.8160928901350769,
+      "grad_norm": 0.17469745874404907,
+      "learning_rate": 1.989525301734317e-05,
+      "loss": 1.2449,
+      "step": 2192
+    },
+    {
+      "epoch": 0.8164651952856861,
+      "grad_norm": 0.18415339291095734,
+      "learning_rate": 1.9895077501762588e-05,
+      "loss": 1.2507,
+      "step": 2193
+    },
+    {
+      "epoch": 0.8168375004362951,
+      "grad_norm": 0.18301935493946075,
+      "learning_rate": 1.9894901840032336e-05,
+      "loss": 1.254,
+      "step": 2194
+    },
+    {
+      "epoch": 0.8172098055869041,
+      "grad_norm": 0.18746277689933777,
+      "learning_rate": 1.9894726032155e-05,
+      "loss": 1.2621,
+      "step": 2195
+    },
+    {
+      "epoch": 0.8175821107375132,
+      "grad_norm": 0.17902565002441406,
+      "learning_rate": 1.9894550078133186e-05,
+      "loss": 1.267,
+      "step": 2196
+    },
+    {
+      "epoch": 0.8179544158881223,
+      "grad_norm": 0.1712873876094818,
+      "learning_rate": 1.9894373977969486e-05,
+      "loss": 1.2545,
+      "step": 2197
+    },
+    {
+      "epoch": 0.8183267210387314,
+      "grad_norm": 0.1745031774044037,
+      "learning_rate": 1.9894197731666506e-05,
+      "loss": 1.244,
+      "step": 2198
+    },
+    {
+      "epoch": 0.8186990261893404,
+      "grad_norm": 0.18271346390247345,
+      "learning_rate": 1.9894021339226843e-05,
+      "loss": 1.263,
+      "step": 2199
+    },
+    {
+      "epoch": 0.8190713313399495,
+      "grad_norm": 0.17454905807971954,
+      "learning_rate": 1.989384480065311e-05,
+      "loss": 1.2636,
+      "step": 2200
+    },
+    {
+      "epoch": 0.8194436364905586,
+      "grad_norm": 0.17742374539375305,
+      "learning_rate": 1.9893668115947906e-05,
+      "loss": 1.2489,
+      "step": 2201
+    },
+    {
+      "epoch": 0.8198159416411677,
+      "grad_norm": 0.17203061282634735,
+      "learning_rate": 1.9893491285113845e-05,
+      "loss": 1.248,
+      "step": 2202
+    },
+    {
+      "epoch": 0.8201882467917767,
+      "grad_norm": 0.1837659776210785,
+      "learning_rate": 1.989331430815354e-05,
+      "loss": 1.2582,
+      "step": 2203
+    },
+    {
+      "epoch": 0.8205605519423858,
+      "grad_norm": 0.17900030314922333,
+      "learning_rate": 1.9893137185069603e-05,
+      "loss": 1.2472,
+      "step": 2204
+    },
+    {
+      "epoch": 0.8209328570929948,
+      "grad_norm": 0.17265020310878754,
+      "learning_rate": 1.9892959915864652e-05,
+      "loss": 1.2608,
+      "step": 2205
+    },
+    {
+      "epoch": 0.821305162243604,
+      "grad_norm": 0.1822979599237442,
+      "learning_rate": 1.9892782500541304e-05,
+      "loss": 1.2477,
+      "step": 2206
+    },
+    {
+      "epoch": 0.821677467394213,
+      "grad_norm": 0.18672309815883636,
+      "learning_rate": 1.9892604939102177e-05,
+      "loss": 1.2489,
+      "step": 2207
+    },
+    {
+      "epoch": 0.822049772544822,
+      "grad_norm": 0.18067359924316406,
+      "learning_rate": 1.9892427231549897e-05,
+      "loss": 1.2511,
+      "step": 2208
+    },
+    {
+      "epoch": 0.8224220776954311,
+      "grad_norm": 0.18508952856063843,
+      "learning_rate": 1.9892249377887086e-05,
+      "loss": 1.2522,
+      "step": 2209
+    },
+    {
+      "epoch": 0.8227943828460402,
+      "grad_norm": 0.17746296525001526,
+      "learning_rate": 1.9892071378116378e-05,
+      "loss": 1.2487,
+      "step": 2210
+    },
+    {
+      "epoch": 0.8231666879966493,
+      "grad_norm": 0.1730239987373352,
+      "learning_rate": 1.9891893232240394e-05,
+      "loss": 1.2631,
+      "step": 2211
+    },
+    {
+      "epoch": 0.8235389931472583,
+      "grad_norm": 0.1845473349094391,
+      "learning_rate": 1.9891714940261764e-05,
+      "loss": 1.2595,
+      "step": 2212
+    },
+    {
+      "epoch": 0.8239112982978674,
+      "grad_norm": 0.18520519137382507,
+      "learning_rate": 1.989153650218313e-05,
+      "loss": 1.2479,
+      "step": 2213
+    },
+    {
+      "epoch": 0.8242836034484765,
+      "grad_norm": 0.19345776736736298,
+      "learning_rate": 1.989135791800712e-05,
+      "loss": 1.2744,
+      "step": 2214
+    },
+    {
+      "epoch": 0.8246559085990856,
+      "grad_norm": 0.17609384655952454,
+      "learning_rate": 1.9891179187736375e-05,
+      "loss": 1.2465,
+      "step": 2215
+    },
+    {
+      "epoch": 0.8250282137496946,
+      "grad_norm": 0.1815994530916214,
+      "learning_rate": 1.9891000311373533e-05,
+      "loss": 1.2523,
+      "step": 2216
+    },
+    {
+      "epoch": 0.8254005189003036,
+      "grad_norm": 0.19329750537872314,
+      "learning_rate": 1.9890821288921238e-05,
+      "loss": 1.2651,
+      "step": 2217
+    },
+    {
+      "epoch": 0.8257728240509127,
+      "grad_norm": 0.18610511720180511,
+      "learning_rate": 1.9890642120382132e-05,
+      "loss": 1.2453,
+      "step": 2218
+    },
+    {
+      "epoch": 0.8261451292015218,
+      "grad_norm": 0.1695135533809662,
+      "learning_rate": 1.9890462805758863e-05,
+      "loss": 1.2745,
+      "step": 2219
+    },
+    {
+      "epoch": 0.8265174343521309,
+      "grad_norm": 0.18071943521499634,
+      "learning_rate": 1.9890283345054082e-05,
+      "loss": 1.2737,
+      "step": 2220
+    },
+    {
+      "epoch": 0.8268897395027399,
+      "grad_norm": 0.18198001384735107,
+      "learning_rate": 1.9890103738270433e-05,
+      "loss": 1.2495,
+      "step": 2221
+    },
+    {
+      "epoch": 0.827262044653349,
+      "grad_norm": 0.1641894280910492,
+      "learning_rate": 1.9889923985410576e-05,
+      "loss": 1.2556,
+      "step": 2222
+    },
+    {
+      "epoch": 0.8276343498039581,
+      "grad_norm": 0.18489384651184082,
+      "learning_rate": 1.9889744086477162e-05,
+      "loss": 1.2591,
+      "step": 2223
+    },
+    {
+      "epoch": 0.8280066549545672,
+      "grad_norm": 0.1797334849834442,
+      "learning_rate": 1.9889564041472846e-05,
+      "loss": 1.2376,
+      "step": 2224
+    },
+    {
+      "epoch": 0.8283789601051762,
+      "grad_norm": 0.16851265728473663,
+      "learning_rate": 1.988938385040029e-05,
+      "loss": 1.2442,
+      "step": 2225
+    },
+    {
+      "epoch": 0.8287512652557852,
+      "grad_norm": 0.17845316231250763,
+      "learning_rate": 1.9889203513262153e-05,
+      "loss": 1.2432,
+      "step": 2226
+    },
+    {
+      "epoch": 0.8291235704063944,
+      "grad_norm": 0.18235254287719727,
+      "learning_rate": 1.9889023030061106e-05,
+      "loss": 1.2593,
+      "step": 2227
+    },
+    {
+      "epoch": 0.8294958755570034,
+      "grad_norm": 0.1719173640012741,
+      "learning_rate": 1.9888842400799805e-05,
+      "loss": 1.2567,
+      "step": 2228
+    },
+    {
+      "epoch": 0.8298681807076125,
+      "grad_norm": 0.17865481972694397,
+      "learning_rate": 1.9888661625480927e-05,
+      "loss": 1.2712,
+      "step": 2229
+    },
+    {
+      "epoch": 0.8302404858582215,
+      "grad_norm": 0.1867910474538803,
+      "learning_rate": 1.9888480704107135e-05,
+      "loss": 1.2432,
+      "step": 2230
+    },
+    {
+      "epoch": 0.8306127910088306,
+      "grad_norm": 0.17481322586536407,
+      "learning_rate": 1.9888299636681105e-05,
+      "loss": 1.2538,
+      "step": 2231
+    },
+    {
+      "epoch": 0.8309850961594397,
+      "grad_norm": 0.18114839494228363,
+      "learning_rate": 1.9888118423205504e-05,
+      "loss": 1.2521,
+      "step": 2232
+    },
+    {
+      "epoch": 0.8313574013100488,
+      "grad_norm": 0.17833612859249115,
+      "learning_rate": 1.988793706368302e-05,
+      "loss": 1.2607,
+      "step": 2233
+    },
+    {
+      "epoch": 0.8317297064606578,
+      "grad_norm": 0.17668797075748444,
+      "learning_rate": 1.9887755558116324e-05,
+      "loss": 1.2573,
+      "step": 2234
+    },
+    {
+      "epoch": 0.8321020116112668,
+      "grad_norm": 0.1729530692100525,
+      "learning_rate": 1.9887573906508103e-05,
+      "loss": 1.2558,
+      "step": 2235
+    },
+    {
+      "epoch": 0.832474316761876,
+      "grad_norm": 0.19922944903373718,
+      "learning_rate": 1.988739210886103e-05,
+      "loss": 1.2743,
+      "step": 2236
+    },
+    {
+      "epoch": 0.832846621912485,
+      "grad_norm": 0.18021075427532196,
+      "learning_rate": 1.98872101651778e-05,
+      "loss": 1.2672,
+      "step": 2237
+    },
+    {
+      "epoch": 0.8332189270630941,
+      "grad_norm": 0.1811213493347168,
+      "learning_rate": 1.9887028075461096e-05,
+      "loss": 1.247,
+      "step": 2238
+    },
+    {
+      "epoch": 0.8335912322137031,
+      "grad_norm": 0.18018268048763275,
+      "learning_rate": 1.9886845839713604e-05,
+      "loss": 1.2451,
+      "step": 2239
+    },
+    {
+      "epoch": 0.8339635373643123,
+      "grad_norm": 0.1896362155675888,
+      "learning_rate": 1.9886663457938025e-05,
+      "loss": 1.2558,
+      "step": 2240
+    },
+    {
+      "epoch": 0.8343358425149213,
+      "grad_norm": 0.1847160905599594,
+      "learning_rate": 1.9886480930137046e-05,
+      "loss": 1.2619,
+      "step": 2241
+    },
+    {
+      "epoch": 0.8347081476655304,
+      "grad_norm": 0.17278259992599487,
+      "learning_rate": 1.988629825631336e-05,
+      "loss": 1.258,
+      "step": 2242
+    },
+    {
+      "epoch": 0.8350804528161394,
+      "grad_norm": 0.17939510941505432,
+      "learning_rate": 1.9886115436469674e-05,
+      "loss": 1.2665,
+      "step": 2243
+    },
+    {
+      "epoch": 0.8354527579667484,
+      "grad_norm": 0.17530454695224762,
+      "learning_rate": 1.9885932470608676e-05,
+      "loss": 1.2531,
+      "step": 2244
+    },
+    {
+      "epoch": 0.8358250631173576,
+      "grad_norm": 0.20562461018562317,
+      "learning_rate": 1.9885749358733086e-05,
+      "loss": 1.2467,
+      "step": 2245
+    },
+    {
+      "epoch": 0.8361973682679666,
+      "grad_norm": 0.17135387659072876,
+      "learning_rate": 1.988556610084559e-05,
+      "loss": 1.2486,
+      "step": 2246
+    },
+    {
+      "epoch": 0.8365696734185757,
+      "grad_norm": 0.18840156495571136,
+      "learning_rate": 1.9885382696948906e-05,
+      "loss": 1.2536,
+      "step": 2247
+    },
+    {
+      "epoch": 0.8369419785691847,
+      "grad_norm": 0.1828533560037613,
+      "learning_rate": 1.988519914704574e-05,
+      "loss": 1.2581,
+      "step": 2248
+    },
+    {
+      "epoch": 0.8373142837197939,
+      "grad_norm": 0.1792685091495514,
+      "learning_rate": 1.988501545113881e-05,
+      "loss": 1.2597,
+      "step": 2249
+    },
+    {
+      "epoch": 0.8376865888704029,
+      "grad_norm": 0.18444480001926422,
+      "learning_rate": 1.9884831609230813e-05,
+      "loss": 1.2414,
+      "step": 2250
+    },
+    {
+      "epoch": 0.838058894021012,
+      "grad_norm": 0.18098483979701996,
+      "learning_rate": 1.9884647621324475e-05,
+      "loss": 1.2448,
+      "step": 2251
+    },
+    {
+      "epoch": 0.838431199171621,
+      "grad_norm": 0.1816195845603943,
+      "learning_rate": 1.9884463487422515e-05,
+      "loss": 1.2499,
+      "step": 2252
+    },
+    {
+      "epoch": 0.83880350432223,
+      "grad_norm": 0.19393447041511536,
+      "learning_rate": 1.9884279207527647e-05,
+      "loss": 1.2584,
+      "step": 2253
+    },
+    {
+      "epoch": 0.8391758094728392,
+      "grad_norm": 0.18464012444019318,
+      "learning_rate": 1.9884094781642592e-05,
+      "loss": 1.2544,
+      "step": 2254
+    },
+    {
+      "epoch": 0.8395481146234482,
+      "grad_norm": 0.17956916987895966,
+      "learning_rate": 1.9883910209770083e-05,
+      "loss": 1.2602,
+      "step": 2255
+    },
+    {
+      "epoch": 0.8399204197740573,
+      "grad_norm": 0.18106333911418915,
+      "learning_rate": 1.988372549191284e-05,
+      "loss": 1.2553,
+      "step": 2256
+    },
+    {
+      "epoch": 0.8402927249246663,
+      "grad_norm": 0.18379664421081543,
+      "learning_rate": 1.9883540628073592e-05,
+      "loss": 1.2517,
+      "step": 2257
+    },
+    {
+      "epoch": 0.8406650300752755,
+      "grad_norm": 0.17581118643283844,
+      "learning_rate": 1.9883355618255068e-05,
+      "loss": 1.2534,
+      "step": 2258
+    },
+    {
+      "epoch": 0.8410373352258845,
+      "grad_norm": 0.17392081022262573,
+      "learning_rate": 1.988317046246e-05,
+      "loss": 1.2616,
+      "step": 2259
+    },
+    {
+      "epoch": 0.8414096403764936,
+      "grad_norm": 0.1733708381652832,
+      "learning_rate": 1.9882985160691127e-05,
+      "loss": 1.2549,
+      "step": 2260
+    },
+    {
+      "epoch": 0.8417819455271026,
+      "grad_norm": 0.1678503155708313,
+      "learning_rate": 1.9882799712951182e-05,
+      "loss": 1.2653,
+      "step": 2261
+    },
+    {
+      "epoch": 0.8421542506777118,
+      "grad_norm": 0.17663797736167908,
+      "learning_rate": 1.9882614119242906e-05,
+      "loss": 1.2556,
+      "step": 2262
+    },
+    {
+      "epoch": 0.8425265558283208,
+      "grad_norm": 0.17948319017887115,
+      "learning_rate": 1.988242837956904e-05,
+      "loss": 1.2459,
+      "step": 2263
+    },
+    {
+      "epoch": 0.8428988609789299,
+      "grad_norm": 0.17682886123657227,
+      "learning_rate": 1.9882242493932327e-05,
+      "loss": 1.2554,
+      "step": 2264
+    },
+    {
+      "epoch": 0.8432711661295389,
+      "grad_norm": 0.17694690823554993,
+      "learning_rate": 1.9882056462335513e-05,
+      "loss": 1.2536,
+      "step": 2265
+    },
+    {
+      "epoch": 0.8436434712801479,
+      "grad_norm": 0.18028870224952698,
+      "learning_rate": 1.9881870284781345e-05,
+      "loss": 1.2572,
+      "step": 2266
+    },
+    {
+      "epoch": 0.8440157764307571,
+      "grad_norm": 0.17367003858089447,
+      "learning_rate": 1.9881683961272572e-05,
+      "loss": 1.2496,
+      "step": 2267
+    },
+    {
+      "epoch": 0.8443880815813661,
+      "grad_norm": 0.18058733642101288,
+      "learning_rate": 1.988149749181195e-05,
+      "loss": 1.2648,
+      "step": 2268
+    },
+    {
+      "epoch": 0.8447603867319752,
+      "grad_norm": 0.1787886917591095,
+      "learning_rate": 1.9881310876402225e-05,
+      "loss": 1.2452,
+      "step": 2269
+    },
+    {
+      "epoch": 0.8451326918825842,
+      "grad_norm": 0.17933231592178345,
+      "learning_rate": 1.988112411504616e-05,
+      "loss": 1.2531,
+      "step": 2270
+    },
+    {
+      "epoch": 0.8455049970331934,
+      "grad_norm": 0.16929695010185242,
+      "learning_rate": 1.9880937207746515e-05,
+      "loss": 1.2432,
+      "step": 2271
+    },
+    {
+      "epoch": 0.8458773021838024,
+      "grad_norm": 0.17460580170154572,
+      "learning_rate": 1.9880750154506048e-05,
+      "loss": 1.2518,
+      "step": 2272
+    },
+    {
+      "epoch": 0.8462496073344115,
+      "grad_norm": 0.19186393916606903,
+      "learning_rate": 1.9880562955327516e-05,
+      "loss": 1.2534,
+      "step": 2273
+    },
+    {
+      "epoch": 0.8466219124850205,
+      "grad_norm": 0.17282183468341827,
+      "learning_rate": 1.9880375610213694e-05,
+      "loss": 1.241,
+      "step": 2274
+    },
+    {
+      "epoch": 0.8469942176356297,
+      "grad_norm": 0.16319245100021362,
+      "learning_rate": 1.9880188119167345e-05,
+      "loss": 1.2538,
+      "step": 2275
+    },
+    {
+      "epoch": 0.8473665227862387,
+      "grad_norm": 0.1815318912267685,
+      "learning_rate": 1.988000048219123e-05,
+      "loss": 1.2561,
+      "step": 2276
+    },
+    {
+      "epoch": 0.8477388279368477,
+      "grad_norm": 0.17304065823554993,
+      "learning_rate": 1.9879812699288136e-05,
+      "loss": 1.2545,
+      "step": 2277
+    },
+    {
+      "epoch": 0.8481111330874568,
+      "grad_norm": 0.17184028029441833,
+      "learning_rate": 1.9879624770460827e-05,
+      "loss": 1.2637,
+      "step": 2278
+    },
+    {
+      "epoch": 0.8484834382380658,
+      "grad_norm": 0.1777077466249466,
+      "learning_rate": 1.9879436695712076e-05,
+      "loss": 1.2599,
+      "step": 2279
+    },
+    {
+      "epoch": 0.848855743388675,
+      "grad_norm": 0.17770980298519135,
+      "learning_rate": 1.9879248475044668e-05,
+      "loss": 1.2289,
+      "step": 2280
+    },
+    {
+      "epoch": 0.849228048539284,
+      "grad_norm": 0.1785004436969757,
+      "learning_rate": 1.987906010846138e-05,
+      "loss": 1.2577,
+      "step": 2281
+    },
+    {
+      "epoch": 0.8496003536898931,
+      "grad_norm": 0.16683989763259888,
+      "learning_rate": 1.9878871595964993e-05,
+      "loss": 1.2454,
+      "step": 2282
+    },
+    {
+      "epoch": 0.8499726588405021,
+      "grad_norm": 0.18355853855609894,
+      "learning_rate": 1.9878682937558297e-05,
+      "loss": 1.284,
+      "step": 2283
+    },
+    {
+      "epoch": 0.8503449639911113,
+      "grad_norm": 0.1772661954164505,
+      "learning_rate": 1.987849413324407e-05,
+      "loss": 1.2512,
+      "step": 2284
+    },
+    {
+      "epoch": 0.8507172691417203,
+      "grad_norm": 0.17306406795978546,
+      "learning_rate": 1.9878305183025103e-05,
+      "loss": 1.2589,
+      "step": 2285
+    },
+    {
+      "epoch": 0.8510895742923293,
+      "grad_norm": 0.16759039461612701,
+      "learning_rate": 1.987811608690419e-05,
+      "loss": 1.2458,
+      "step": 2286
+    },
+    {
+      "epoch": 0.8514618794429384,
+      "grad_norm": 0.17323878407478333,
+      "learning_rate": 1.9877926844884126e-05,
+      "loss": 1.2559,
+      "step": 2287
+    },
+    {
+      "epoch": 0.8518341845935475,
+      "grad_norm": 0.16814987361431122,
+      "learning_rate": 1.9877737456967698e-05,
+      "loss": 1.2589,
+      "step": 2288
+    },
+    {
+      "epoch": 0.8522064897441566,
+      "grad_norm": 0.16971348226070404,
+      "learning_rate": 1.987754792315771e-05,
+      "loss": 1.2541,
+      "step": 2289
+    },
+    {
+      "epoch": 0.8525787948947656,
+      "grad_norm": 0.16793325543403625,
+      "learning_rate": 1.9877358243456956e-05,
+      "loss": 1.2716,
+      "step": 2290
+    },
+    {
+      "epoch": 0.8529511000453747,
+      "grad_norm": 0.17254911363124847,
+      "learning_rate": 1.9877168417868244e-05,
+      "loss": 1.253,
+      "step": 2291
+    },
+    {
+      "epoch": 0.8533234051959837,
+      "grad_norm": 0.17216616868972778,
+      "learning_rate": 1.9876978446394372e-05,
+      "loss": 1.2495,
+      "step": 2292
+    },
+    {
+      "epoch": 0.8536957103465929,
+      "grad_norm": 0.1849498599767685,
+      "learning_rate": 1.9876788329038147e-05,
+      "loss": 1.2736,
+      "step": 2293
+    },
+    {
+      "epoch": 0.8540680154972019,
+      "grad_norm": 0.17307497560977936,
+      "learning_rate": 1.987659806580238e-05,
+      "loss": 1.2549,
+      "step": 2294
+    },
+    {
+      "epoch": 0.854440320647811,
+      "grad_norm": 0.17321421205997467,
+      "learning_rate": 1.9876407656689883e-05,
+      "loss": 1.2634,
+      "step": 2295
+    },
+    {
+      "epoch": 0.85481262579842,
+      "grad_norm": 0.17435167729854584,
+      "learning_rate": 1.987621710170346e-05,
+      "loss": 1.2443,
+      "step": 2296
+    },
+    {
+      "epoch": 0.8551849309490291,
+      "grad_norm": 0.1697143167257309,
+      "learning_rate": 1.9876026400845933e-05,
+      "loss": 1.2621,
+      "step": 2297
+    },
+    {
+      "epoch": 0.8555572360996382,
+      "grad_norm": 0.18251660466194153,
+      "learning_rate": 1.9875835554120114e-05,
+      "loss": 1.2588,
+      "step": 2298
+    },
+    {
+      "epoch": 0.8559295412502472,
+      "grad_norm": 0.182033970952034,
+      "learning_rate": 1.9875644561528824e-05,
+      "loss": 1.2656,
+      "step": 2299
+    },
+    {
+      "epoch": 0.8563018464008563,
+      "grad_norm": 0.19597557187080383,
+      "learning_rate": 1.9875453423074883e-05,
+      "loss": 1.2545,
+      "step": 2300
+    },
+    {
+      "epoch": 0.8566741515514654,
+      "grad_norm": 0.17287831008434296,
+      "learning_rate": 1.9875262138761116e-05,
+      "loss": 1.2662,
+      "step": 2301
+    },
+    {
+      "epoch": 0.8570464567020745,
+      "grad_norm": 0.18034107983112335,
+      "learning_rate": 1.987507070859035e-05,
+      "loss": 1.2561,
+      "step": 2302
+    },
+    {
+      "epoch": 0.8574187618526835,
+      "grad_norm": 0.1734304428100586,
+      "learning_rate": 1.98748791325654e-05,
+      "loss": 1.2659,
+      "step": 2303
+    },
+    {
+      "epoch": 0.8577910670032926,
+      "grad_norm": 0.18945597112178802,
+      "learning_rate": 1.9874687410689114e-05,
+      "loss": 1.265,
+      "step": 2304
+    },
+    {
+      "epoch": 0.8581633721539016,
+      "grad_norm": 0.1864440143108368,
+      "learning_rate": 1.9874495542964308e-05,
+      "loss": 1.2634,
+      "step": 2305
+    },
+    {
+      "epoch": 0.8585356773045107,
+      "grad_norm": 0.1847706139087677,
+      "learning_rate": 1.987430352939383e-05,
+      "loss": 1.2572,
+      "step": 2306
+    },
+    {
+      "epoch": 0.8589079824551198,
+      "grad_norm": 0.18453647196292877,
+      "learning_rate": 1.98741113699805e-05,
+      "loss": 1.2501,
+      "step": 2307
+    },
+    {
+      "epoch": 0.8592802876057288,
+      "grad_norm": 0.17989476025104523,
+      "learning_rate": 1.9873919064727173e-05,
+      "loss": 1.2589,
+      "step": 2308
+    },
+    {
+      "epoch": 0.8596525927563379,
+      "grad_norm": 0.19050173461437225,
+      "learning_rate": 1.9873726613636678e-05,
+      "loss": 1.2489,
+      "step": 2309
+    },
+    {
+      "epoch": 0.860024897906947,
+      "grad_norm": 0.17572608590126038,
+      "learning_rate": 1.987353401671186e-05,
+      "loss": 1.2375,
+      "step": 2310
+    },
+    {
+      "epoch": 0.8603972030575561,
+      "grad_norm": 0.17381906509399414,
+      "learning_rate": 1.987334127395556e-05,
+      "loss": 1.2631,
+      "step": 2311
+    },
+    {
+      "epoch": 0.8607695082081651,
+      "grad_norm": 0.17476747930049896,
+      "learning_rate": 1.9873148385370635e-05,
+      "loss": 1.2418,
+      "step": 2312
+    },
+    {
+      "epoch": 0.8611418133587742,
+      "grad_norm": 0.1777612864971161,
+      "learning_rate": 1.9872955350959927e-05,
+      "loss": 1.2502,
+      "step": 2313
+    },
+    {
+      "epoch": 0.8615141185093833,
+      "grad_norm": 0.17522254586219788,
+      "learning_rate": 1.9872762170726284e-05,
+      "loss": 1.2653,
+      "step": 2314
+    },
+    {
+      "epoch": 0.8618864236599924,
+      "grad_norm": 0.17853045463562012,
+      "learning_rate": 1.9872568844672567e-05,
+      "loss": 1.2536,
+      "step": 2315
+    },
+    {
+      "epoch": 0.8622587288106014,
+      "grad_norm": 0.1815364956855774,
+      "learning_rate": 1.9872375372801627e-05,
+      "loss": 1.2518,
+      "step": 2316
+    },
+    {
+      "epoch": 0.8626310339612104,
+      "grad_norm": 0.1911124587059021,
+      "learning_rate": 1.9872181755116324e-05,
+      "loss": 1.2583,
+      "step": 2317
+    },
+    {
+      "epoch": 0.8630033391118195,
+      "grad_norm": 0.1666967123746872,
+      "learning_rate": 1.9871987991619516e-05,
+      "loss": 1.2509,
+      "step": 2318
+    },
+    {
+      "epoch": 0.8633756442624286,
+      "grad_norm": 0.19720950722694397,
+      "learning_rate": 1.9871794082314064e-05,
+      "loss": 1.2615,
+      "step": 2319
+    },
+    {
+      "epoch": 0.8637479494130377,
+      "grad_norm": 0.18802502751350403,
+      "learning_rate": 1.987160002720283e-05,
+      "loss": 1.2608,
+      "step": 2320
+    },
+    {
+      "epoch": 0.8641202545636467,
+      "grad_norm": 0.18227264285087585,
+      "learning_rate": 1.9871405826288685e-05,
+      "loss": 1.2452,
+      "step": 2321
+    },
+    {
+      "epoch": 0.8644925597142558,
+      "grad_norm": 0.1774458885192871,
+      "learning_rate": 1.9871211479574497e-05,
+      "loss": 1.2562,
+      "step": 2322
+    },
+    {
+      "epoch": 0.8648648648648649,
+      "grad_norm": 0.1868571639060974,
+      "learning_rate": 1.9871016987063133e-05,
+      "loss": 1.2485,
+      "step": 2323
+    },
+    {
+      "epoch": 0.865237170015474,
+      "grad_norm": 0.18438728153705597,
+      "learning_rate": 1.987082234875747e-05,
+      "loss": 1.2437,
+      "step": 2324
+    },
+    {
+      "epoch": 0.865609475166083,
+      "grad_norm": 0.17169415950775146,
+      "learning_rate": 1.987062756466038e-05,
+      "loss": 1.2487,
+      "step": 2325
+    },
+    {
+      "epoch": 0.865981780316692,
+      "grad_norm": 0.1830715835094452,
+      "learning_rate": 1.9870432634774737e-05,
+      "loss": 1.2454,
+      "step": 2326
+    },
+    {
+      "epoch": 0.8663540854673011,
+      "grad_norm": 0.18534304201602936,
+      "learning_rate": 1.9870237559103426e-05,
+      "loss": 1.2691,
+      "step": 2327
+    },
+    {
+      "epoch": 0.8667263906179102,
+      "grad_norm": 0.17846928536891937,
+      "learning_rate": 1.9870042337649328e-05,
+      "loss": 1.2524,
+      "step": 2328
+    },
+    {
+      "epoch": 0.8670986957685193,
+      "grad_norm": 0.17754697799682617,
+      "learning_rate": 1.9869846970415317e-05,
+      "loss": 1.2456,
+      "step": 2329
+    },
+    {
+      "epoch": 0.8674710009191283,
+      "grad_norm": 0.1744270771741867,
+      "learning_rate": 1.9869651457404293e-05,
+      "loss": 1.2393,
+      "step": 2330
+    },
+    {
+      "epoch": 0.8678433060697374,
+      "grad_norm": 0.18277184665203094,
+      "learning_rate": 1.9869455798619133e-05,
+      "loss": 1.2672,
+      "step": 2331
+    },
+    {
+      "epoch": 0.8682156112203465,
+      "grad_norm": 0.1796514391899109,
+      "learning_rate": 1.986925999406273e-05,
+      "loss": 1.247,
+      "step": 2332
+    },
+    {
+      "epoch": 0.8685879163709556,
+      "grad_norm": 0.17978057265281677,
+      "learning_rate": 1.9869064043737977e-05,
+      "loss": 1.258,
+      "step": 2333
+    },
+    {
+      "epoch": 0.8689602215215646,
+      "grad_norm": 0.1840755194425583,
+      "learning_rate": 1.9868867947647768e-05,
+      "loss": 1.2558,
+      "step": 2334
+    },
+    {
+      "epoch": 0.8693325266721736,
+      "grad_norm": 0.18063770234584808,
+      "learning_rate": 1.9868671705794997e-05,
+      "loss": 1.2623,
+      "step": 2335
+    },
+    {
+      "epoch": 0.8697048318227828,
+      "grad_norm": 0.16667230427265167,
+      "learning_rate": 1.9868475318182566e-05,
+      "loss": 1.2476,
+      "step": 2336
+    },
+    {
+      "epoch": 0.8700771369733918,
+      "grad_norm": 0.17838679254055023,
+      "learning_rate": 1.9868278784813374e-05,
+      "loss": 1.2663,
+      "step": 2337
+    },
+    {
+      "epoch": 0.8704494421240009,
+      "grad_norm": 0.1771472841501236,
+      "learning_rate": 1.9868082105690323e-05,
+      "loss": 1.2516,
+      "step": 2338
+    },
+    {
+      "epoch": 0.8708217472746099,
+      "grad_norm": 0.18569402396678925,
+      "learning_rate": 1.9867885280816317e-05,
+      "loss": 1.2496,
+      "step": 2339
+    },
+    {
+      "epoch": 0.871194052425219,
+      "grad_norm": 0.17394711077213287,
+      "learning_rate": 1.986768831019427e-05,
+      "loss": 1.2457,
+      "step": 2340
+    },
+    {
+      "epoch": 0.8715663575758281,
+      "grad_norm": 0.16605480015277863,
+      "learning_rate": 1.986749119382708e-05,
+      "loss": 1.2542,
+      "step": 2341
+    },
+    {
+      "epoch": 0.8719386627264372,
+      "grad_norm": 0.17628324031829834,
+      "learning_rate": 1.9867293931717664e-05,
+      "loss": 1.2544,
+      "step": 2342
+    },
+    {
+      "epoch": 0.8723109678770462,
+      "grad_norm": 0.18517057597637177,
+      "learning_rate": 1.986709652386894e-05,
+      "loss": 1.2576,
+      "step": 2343
+    },
+    {
+      "epoch": 0.8726832730276552,
+      "grad_norm": 0.17365185916423798,
+      "learning_rate": 1.9866898970283816e-05,
+      "loss": 1.2472,
+      "step": 2344
+    },
+    {
+      "epoch": 0.8730555781782644,
+      "grad_norm": 0.17242515087127686,
+      "learning_rate": 1.9866701270965217e-05,
+      "loss": 1.2536,
+      "step": 2345
+    },
+    {
+      "epoch": 0.8734278833288734,
+      "grad_norm": 0.1831977367401123,
+      "learning_rate": 1.986650342591606e-05,
+      "loss": 1.2538,
+      "step": 2346
+    },
+    {
+      "epoch": 0.8738001884794825,
+      "grad_norm": 0.1806264966726303,
+      "learning_rate": 1.986630543513926e-05,
+      "loss": 1.2417,
+      "step": 2347
+    },
+    {
+      "epoch": 0.8741724936300915,
+      "grad_norm": 0.1779744178056717,
+      "learning_rate": 1.9866107298637754e-05,
+      "loss": 1.2577,
+      "step": 2348
+    },
+    {
+      "epoch": 0.8745447987807007,
+      "grad_norm": 0.18145568668842316,
+      "learning_rate": 1.9865909016414462e-05,
+      "loss": 1.2498,
+      "step": 2349
+    },
+    {
+      "epoch": 0.8749171039313097,
+      "grad_norm": 0.1738026887178421,
+      "learning_rate": 1.9865710588472307e-05,
+      "loss": 1.2482,
+      "step": 2350
+    },
+    {
+      "epoch": 0.8752894090819188,
+      "grad_norm": 0.17503541707992554,
+      "learning_rate": 1.986551201481423e-05,
+      "loss": 1.2433,
+      "step": 2351
+    },
+    {
+      "epoch": 0.8756617142325278,
+      "grad_norm": 0.18008525669574738,
+      "learning_rate": 1.9865313295443164e-05,
+      "loss": 1.2692,
+      "step": 2352
+    },
+    {
+      "epoch": 0.8760340193831369,
+      "grad_norm": 0.17630065977573395,
+      "learning_rate": 1.9865114430362037e-05,
+      "loss": 1.256,
+      "step": 2353
+    },
+    {
+      "epoch": 0.876406324533746,
+      "grad_norm": 0.16775432229042053,
+      "learning_rate": 1.9864915419573787e-05,
+      "loss": 1.2493,
+      "step": 2354
+    },
+    {
+      "epoch": 0.876778629684355,
+      "grad_norm": 0.17783991992473602,
+      "learning_rate": 1.9864716263081356e-05,
+      "loss": 1.2344,
+      "step": 2355
+    },
+    {
+      "epoch": 0.8771509348349641,
+      "grad_norm": 0.1752803474664688,
+      "learning_rate": 1.9864516960887684e-05,
+      "loss": 1.2532,
+      "step": 2356
+    },
+    {
+      "epoch": 0.8775232399855731,
+      "grad_norm": 0.17517109215259552,
+      "learning_rate": 1.9864317512995718e-05,
+      "loss": 1.2508,
+      "step": 2357
+    },
+    {
+      "epoch": 0.8778955451361823,
+      "grad_norm": 0.19087643921375275,
+      "learning_rate": 1.9864117919408397e-05,
+      "loss": 1.2517,
+      "step": 2358
+    },
+    {
+      "epoch": 0.8782678502867913,
+      "grad_norm": 0.17448683083057404,
+      "learning_rate": 1.986391818012868e-05,
+      "loss": 1.2431,
+      "step": 2359
+    },
+    {
+      "epoch": 0.8786401554374004,
+      "grad_norm": 0.1711607575416565,
+      "learning_rate": 1.9863718295159503e-05,
+      "loss": 1.257,
+      "step": 2360
+    },
+    {
+      "epoch": 0.8790124605880094,
+      "grad_norm": 0.1857292354106903,
+      "learning_rate": 1.9863518264503832e-05,
+      "loss": 1.2477,
+      "step": 2361
+    },
+    {
+      "epoch": 0.8793847657386186,
+      "grad_norm": 0.18781475722789764,
+      "learning_rate": 1.9863318088164613e-05,
+      "loss": 1.2672,
+      "step": 2362
+    },
+    {
+      "epoch": 0.8797570708892276,
+      "grad_norm": 0.16976489126682281,
+      "learning_rate": 1.9863117766144807e-05,
+      "loss": 1.2528,
+      "step": 2363
+    },
+    {
+      "epoch": 0.8801293760398367,
+      "grad_norm": 0.17341898381710052,
+      "learning_rate": 1.9862917298447365e-05,
+      "loss": 1.2625,
+      "step": 2364
+    },
+    {
+      "epoch": 0.8805016811904457,
+      "grad_norm": 0.18177741765975952,
+      "learning_rate": 1.986271668507526e-05,
+      "loss": 1.2602,
+      "step": 2365
+    },
+    {
+      "epoch": 0.8808739863410547,
+      "grad_norm": 0.1771697849035263,
+      "learning_rate": 1.9862515926031444e-05,
+      "loss": 1.2413,
+      "step": 2366
+    },
+    {
+      "epoch": 0.8812462914916639,
+      "grad_norm": 0.17161764204502106,
+      "learning_rate": 1.9862315021318886e-05,
+      "loss": 1.2514,
+      "step": 2367
+    },
+    {
+      "epoch": 0.8816185966422729,
+      "grad_norm": 0.18902365863323212,
+      "learning_rate": 1.986211397094056e-05,
+      "loss": 1.2348,
+      "step": 2368
+    },
+    {
+      "epoch": 0.881990901792882,
+      "grad_norm": 0.17638647556304932,
+      "learning_rate": 1.9861912774899425e-05,
+      "loss": 1.2421,
+      "step": 2369
+    },
+    {
+      "epoch": 0.882363206943491,
+      "grad_norm": 0.17760038375854492,
+      "learning_rate": 1.9861711433198457e-05,
+      "loss": 1.2574,
+      "step": 2370
+    },
+    {
+      "epoch": 0.8827355120941002,
+      "grad_norm": 0.17845885455608368,
+      "learning_rate": 1.9861509945840632e-05,
+      "loss": 1.252,
+      "step": 2371
+    },
+    {
+      "epoch": 0.8831078172447092,
+      "grad_norm": 0.17632153630256653,
+      "learning_rate": 1.9861308312828923e-05,
+      "loss": 1.2648,
+      "step": 2372
+    },
+    {
+      "epoch": 0.8834801223953183,
+      "grad_norm": 0.17382317781448364,
+      "learning_rate": 1.9861106534166307e-05,
+      "loss": 1.2574,
+      "step": 2373
+    },
+    {
+      "epoch": 0.8838524275459273,
+      "grad_norm": 0.18132738769054413,
+      "learning_rate": 1.986090460985577e-05,
+      "loss": 1.2546,
+      "step": 2374
+    },
+    {
+      "epoch": 0.8842247326965365,
+      "grad_norm": 0.17880843579769135,
+      "learning_rate": 1.9860702539900288e-05,
+      "loss": 1.2542,
+      "step": 2375
+    },
+    {
+      "epoch": 0.8845970378471455,
+      "grad_norm": 0.17678992450237274,
+      "learning_rate": 1.9860500324302848e-05,
+      "loss": 1.258,
+      "step": 2376
+    },
+    {
+      "epoch": 0.8849693429977545,
+      "grad_norm": 0.17159947752952576,
+      "learning_rate": 1.986029796306644e-05,
+      "loss": 1.2545,
+      "step": 2377
+    },
+    {
+      "epoch": 0.8853416481483636,
+      "grad_norm": 0.17294186353683472,
+      "learning_rate": 1.9860095456194045e-05,
+      "loss": 1.2572,
+      "step": 2378
+    },
+    {
+      "epoch": 0.8857139532989726,
+      "grad_norm": 0.17959366738796234,
+      "learning_rate": 1.9859892803688666e-05,
+      "loss": 1.2406,
+      "step": 2379
+    },
+    {
+      "epoch": 0.8860862584495818,
+      "grad_norm": 0.1762249618768692,
+      "learning_rate": 1.9859690005553282e-05,
+      "loss": 1.2681,
+      "step": 2380
+    },
+    {
+      "epoch": 0.8864585636001908,
+      "grad_norm": 0.17363481223583221,
+      "learning_rate": 1.98594870617909e-05,
+      "loss": 1.2502,
+      "step": 2381
+    },
+    {
+      "epoch": 0.8868308687507999,
+      "grad_norm": 0.18337921798229218,
+      "learning_rate": 1.985928397240451e-05,
+      "loss": 1.2509,
+      "step": 2382
+    },
+    {
+      "epoch": 0.8872031739014089,
+      "grad_norm": 0.17734134197235107,
+      "learning_rate": 1.985908073739712e-05,
+      "loss": 1.2543,
+      "step": 2383
+    },
+    {
+      "epoch": 0.8875754790520181,
+      "grad_norm": 0.1737191379070282,
+      "learning_rate": 1.9858877356771722e-05,
+      "loss": 1.2543,
+      "step": 2384
+    },
+    {
+      "epoch": 0.8879477842026271,
+      "grad_norm": 0.18461737036705017,
+      "learning_rate": 1.9858673830531322e-05,
+      "loss": 1.2598,
+      "step": 2385
+    },
+    {
+      "epoch": 0.8883200893532361,
+      "grad_norm": 0.1724756807088852,
+      "learning_rate": 1.9858470158678932e-05,
+      "loss": 1.2436,
+      "step": 2386
+    },
+    {
+      "epoch": 0.8886923945038452,
+      "grad_norm": 0.16471019387245178,
+      "learning_rate": 1.9858266341217556e-05,
+      "loss": 1.2428,
+      "step": 2387
+    },
+    {
+      "epoch": 0.8890646996544542,
+      "grad_norm": 0.1689532995223999,
+      "learning_rate": 1.9858062378150204e-05,
+      "loss": 1.2496,
+      "step": 2388
+    },
+    {
+      "epoch": 0.8894370048050634,
+      "grad_norm": 0.17168131470680237,
+      "learning_rate": 1.9857858269479887e-05,
+      "loss": 1.2495,
+      "step": 2389
+    },
+    {
+      "epoch": 0.8898093099556724,
+      "grad_norm": 0.18134863674640656,
+      "learning_rate": 1.9857654015209627e-05,
+      "loss": 1.24,
+      "step": 2390
+    },
+    {
+      "epoch": 0.8901816151062815,
+      "grad_norm": 0.16769009828567505,
+      "learning_rate": 1.985744961534243e-05,
+      "loss": 1.244,
+      "step": 2391
+    },
+    {
+      "epoch": 0.8905539202568905,
+      "grad_norm": 0.16388821601867676,
+      "learning_rate": 1.9857245069881326e-05,
+      "loss": 1.2531,
+      "step": 2392
+    },
+    {
+      "epoch": 0.8909262254074997,
+      "grad_norm": 0.17608126997947693,
+      "learning_rate": 1.985704037882933e-05,
+      "loss": 1.2432,
+      "step": 2393
+    },
+    {
+      "epoch": 0.8912985305581087,
+      "grad_norm": 0.17434832453727722,
+      "learning_rate": 1.9856835542189464e-05,
+      "loss": 1.2555,
+      "step": 2394
+    },
+    {
+      "epoch": 0.8916708357087177,
+      "grad_norm": 0.16499166190624237,
+      "learning_rate": 1.9856630559964758e-05,
+      "loss": 1.2392,
+      "step": 2395
+    },
+    {
+      "epoch": 0.8920431408593268,
+      "grad_norm": 0.18076403439044952,
+      "learning_rate": 1.9856425432158236e-05,
+      "loss": 1.2435,
+      "step": 2396
+    },
+    {
+      "epoch": 0.8924154460099359,
+      "grad_norm": 0.17794503271579742,
+      "learning_rate": 1.9856220158772927e-05,
+      "loss": 1.2558,
+      "step": 2397
+    },
+    {
+      "epoch": 0.892787751160545,
+      "grad_norm": 0.17728637158870697,
+      "learning_rate": 1.9856014739811867e-05,
+      "loss": 1.2692,
+      "step": 2398
+    },
+    {
+      "epoch": 0.893160056311154,
+      "grad_norm": 0.17668506503105164,
+      "learning_rate": 1.9855809175278088e-05,
+      "loss": 1.2552,
+      "step": 2399
+    },
+    {
+      "epoch": 0.8935323614617631,
+      "grad_norm": 0.1653260886669159,
+      "learning_rate": 1.9855603465174623e-05,
+      "loss": 1.2485,
+      "step": 2400
+    },
+    {
+      "epoch": 0.8939046666123721,
+      "grad_norm": 0.17724184691905975,
+      "learning_rate": 1.9855397609504517e-05,
+      "loss": 1.2513,
+      "step": 2401
+    },
+    {
+      "epoch": 0.8942769717629813,
+      "grad_norm": 0.1668618619441986,
+      "learning_rate": 1.9855191608270807e-05,
+      "loss": 1.2387,
+      "step": 2402
+    },
+    {
+      "epoch": 0.8946492769135903,
+      "grad_norm": 0.16532456874847412,
+      "learning_rate": 1.9854985461476534e-05,
+      "loss": 1.2431,
+      "step": 2403
+    },
+    {
+      "epoch": 0.8950215820641994,
+      "grad_norm": 0.17286598682403564,
+      "learning_rate": 1.9854779169124745e-05,
+      "loss": 1.2322,
+      "step": 2404
+    },
+    {
+      "epoch": 0.8953938872148084,
+      "grad_norm": 0.17562846839427948,
+      "learning_rate": 1.9854572731218483e-05,
+      "loss": 1.2538,
+      "step": 2405
+    },
+    {
+      "epoch": 0.8957661923654175,
+      "grad_norm": 0.1721685379743576,
+      "learning_rate": 1.9854366147760803e-05,
+      "loss": 1.2421,
+      "step": 2406
+    },
+    {
+      "epoch": 0.8961384975160266,
+      "grad_norm": 0.16794408857822418,
+      "learning_rate": 1.9854159418754754e-05,
+      "loss": 1.2622,
+      "step": 2407
+    },
+    {
+      "epoch": 0.8965108026666356,
+      "grad_norm": 0.17553912103176117,
+      "learning_rate": 1.9853952544203387e-05,
+      "loss": 1.2526,
+      "step": 2408
+    },
+    {
+      "epoch": 0.8968831078172447,
+      "grad_norm": 0.17011193931102753,
+      "learning_rate": 1.985374552410976e-05,
+      "loss": 1.2588,
+      "step": 2409
+    },
+    {
+      "epoch": 0.8972554129678538,
+      "grad_norm": 0.17169992625713348,
+      "learning_rate": 1.9853538358476933e-05,
+      "loss": 1.2377,
+      "step": 2410
+    },
+    {
+      "epoch": 0.8976277181184629,
+      "grad_norm": 0.1703435331583023,
+      "learning_rate": 1.985333104730796e-05,
+      "loss": 1.2366,
+      "step": 2411
+    },
+    {
+      "epoch": 0.8980000232690719,
+      "grad_norm": 0.17589643597602844,
+      "learning_rate": 1.9853123590605904e-05,
+      "loss": 1.2371,
+      "step": 2412
+    },
+    {
+      "epoch": 0.898372328419681,
+      "grad_norm": 0.18404798209667206,
+      "learning_rate": 1.9852915988373834e-05,
+      "loss": 1.2483,
+      "step": 2413
+    },
+    {
+      "epoch": 0.89874463357029,
+      "grad_norm": 0.18587037920951843,
+      "learning_rate": 1.9852708240614812e-05,
+      "loss": 1.2575,
+      "step": 2414
+    },
+    {
+      "epoch": 0.8991169387208992,
+      "grad_norm": 0.16864776611328125,
+      "learning_rate": 1.9852500347331908e-05,
+      "loss": 1.2358,
+      "step": 2415
+    },
+    {
+      "epoch": 0.8994892438715082,
+      "grad_norm": 0.17366687953472137,
+      "learning_rate": 1.985229230852819e-05,
+      "loss": 1.2397,
+      "step": 2416
+    },
+    {
+      "epoch": 0.8998615490221172,
+      "grad_norm": 0.1771089732646942,
+      "learning_rate": 1.9852084124206735e-05,
+      "loss": 1.2628,
+      "step": 2417
+    },
+    {
+      "epoch": 0.9002338541727263,
+      "grad_norm": 0.16770784556865692,
+      "learning_rate": 1.9851875794370616e-05,
+      "loss": 1.2384,
+      "step": 2418
+    },
+    {
+      "epoch": 0.9006061593233354,
+      "grad_norm": 0.17862196266651154,
+      "learning_rate": 1.985166731902291e-05,
+      "loss": 1.2611,
+      "step": 2419
+    },
+    {
+      "epoch": 0.9009784644739445,
+      "grad_norm": 0.17286571860313416,
+      "learning_rate": 1.9851458698166693e-05,
+      "loss": 1.2643,
+      "step": 2420
+    },
+    {
+      "epoch": 0.9013507696245535,
+      "grad_norm": 0.17281019687652588,
+      "learning_rate": 1.9851249931805053e-05,
+      "loss": 1.2539,
+      "step": 2421
+    },
+    {
+      "epoch": 0.9017230747751626,
+      "grad_norm": 0.1699199378490448,
+      "learning_rate": 1.985104101994107e-05,
+      "loss": 1.2567,
+      "step": 2422
+    },
+    {
+      "epoch": 0.9020953799257717,
+      "grad_norm": 0.17170144617557526,
+      "learning_rate": 1.9850831962577824e-05,
+      "loss": 1.2658,
+      "step": 2423
+    },
+    {
+      "epoch": 0.9024676850763808,
+      "grad_norm": 0.17239661514759064,
+      "learning_rate": 1.9850622759718415e-05,
+      "loss": 1.2399,
+      "step": 2424
+    },
+    {
+      "epoch": 0.9028399902269898,
+      "grad_norm": 0.17295676469802856,
+      "learning_rate": 1.9850413411365923e-05,
+      "loss": 1.2494,
+      "step": 2425
+    },
+    {
+      "epoch": 0.9032122953775988,
+      "grad_norm": 0.17475001513957977,
+      "learning_rate": 1.985020391752344e-05,
+      "loss": 1.2487,
+      "step": 2426
+    },
+    {
+      "epoch": 0.9035846005282079,
+      "grad_norm": 0.16339634358882904,
+      "learning_rate": 1.9849994278194068e-05,
+      "loss": 1.2435,
+      "step": 2427
+    },
+    {
+      "epoch": 0.903956905678817,
+      "grad_norm": 0.1701515167951584,
+      "learning_rate": 1.9849784493380897e-05,
+      "loss": 1.2499,
+      "step": 2428
+    },
+    {
+      "epoch": 0.9043292108294261,
+      "grad_norm": 0.16852952539920807,
+      "learning_rate": 1.9849574563087025e-05,
+      "loss": 1.242,
+      "step": 2429
+    },
+    {
+      "epoch": 0.9047015159800351,
+      "grad_norm": 0.1811753362417221,
+      "learning_rate": 1.984936448731556e-05,
+      "loss": 1.2504,
+      "step": 2430
+    },
+    {
+      "epoch": 0.9050738211306442,
+      "grad_norm": 0.16984617710113525,
+      "learning_rate": 1.9849154266069597e-05,
+      "loss": 1.2338,
+      "step": 2431
+    },
+    {
+      "epoch": 0.9054461262812533,
+      "grad_norm": 0.18032081425189972,
+      "learning_rate": 1.984894389935224e-05,
+      "loss": 1.2283,
+      "step": 2432
+    },
+    {
+      "epoch": 0.9058184314318624,
+      "grad_norm": 0.1682289093732834,
+      "learning_rate": 1.9848733387166606e-05,
+      "loss": 1.2465,
+      "step": 2433
+    },
+    {
+      "epoch": 0.9061907365824714,
+      "grad_norm": 0.1777002066373825,
+      "learning_rate": 1.9848522729515794e-05,
+      "loss": 1.2515,
+      "step": 2434
+    },
+    {
+      "epoch": 0.9065630417330804,
+      "grad_norm": 0.17563197016716003,
+      "learning_rate": 1.984831192640292e-05,
+      "loss": 1.2523,
+      "step": 2435
+    },
+    {
+      "epoch": 0.9069353468836896,
+      "grad_norm": 0.17422862350940704,
+      "learning_rate": 1.9848100977831098e-05,
+      "loss": 1.2584,
+      "step": 2436
+    },
+    {
+      "epoch": 0.9073076520342986,
+      "grad_norm": 0.1725662648677826,
+      "learning_rate": 1.984788988380344e-05,
+      "loss": 1.2468,
+      "step": 2437
+    },
+    {
+      "epoch": 0.9076799571849077,
+      "grad_norm": 0.16899681091308594,
+      "learning_rate": 1.9847678644323068e-05,
+      "loss": 1.2547,
+      "step": 2438
+    },
+    {
+      "epoch": 0.9080522623355167,
+      "grad_norm": 0.1767120510339737,
+      "learning_rate": 1.9847467259393102e-05,
+      "loss": 1.2531,
+      "step": 2439
+    },
+    {
+      "epoch": 0.9084245674861258,
+      "grad_norm": 0.1795816719532013,
+      "learning_rate": 1.984725572901666e-05,
+      "loss": 1.2519,
+      "step": 2440
+    },
+    {
+      "epoch": 0.9087968726367349,
+      "grad_norm": 0.1791762113571167,
+      "learning_rate": 1.984704405319687e-05,
+      "loss": 1.2787,
+      "step": 2441
+    },
+    {
+      "epoch": 0.909169177787344,
+      "grad_norm": 0.1724022626876831,
+      "learning_rate": 1.9846832231936857e-05,
+      "loss": 1.259,
+      "step": 2442
+    },
+    {
+      "epoch": 0.909541482937953,
+      "grad_norm": 0.182962566614151,
+      "learning_rate": 1.984662026523975e-05,
+      "loss": 1.2568,
+      "step": 2443
+    },
+    {
+      "epoch": 0.909913788088562,
+      "grad_norm": 0.17582541704177856,
+      "learning_rate": 1.984640815310868e-05,
+      "loss": 1.237,
+      "step": 2444
+    },
+    {
+      "epoch": 0.9102860932391712,
+      "grad_norm": 0.17969557642936707,
+      "learning_rate": 1.984619589554678e-05,
+      "loss": 1.2662,
+      "step": 2445
+    },
+    {
+      "epoch": 0.9106583983897802,
+      "grad_norm": 0.17635543644428253,
+      "learning_rate": 1.9845983492557183e-05,
+      "loss": 1.2597,
+      "step": 2446
+    },
+    {
+      "epoch": 0.9110307035403893,
+      "grad_norm": 0.1814170926809311,
+      "learning_rate": 1.984577094414303e-05,
+      "loss": 1.243,
+      "step": 2447
+    },
+    {
+      "epoch": 0.9114030086909983,
+      "grad_norm": 0.16915488243103027,
+      "learning_rate": 1.984555825030746e-05,
+      "loss": 1.2477,
+      "step": 2448
+    },
+    {
+      "epoch": 0.9117753138416075,
+      "grad_norm": 0.1874271184206009,
+      "learning_rate": 1.9845345411053608e-05,
+      "loss": 1.2594,
+      "step": 2449
+    },
+    {
+      "epoch": 0.9121476189922165,
+      "grad_norm": 0.17666561901569366,
+      "learning_rate": 1.984513242638462e-05,
+      "loss": 1.2704,
+      "step": 2450
+    },
+    {
+      "epoch": 0.9125199241428256,
+      "grad_norm": 0.16944633424282074,
+      "learning_rate": 1.9844919296303647e-05,
+      "loss": 1.2457,
+      "step": 2451
+    },
+    {
+      "epoch": 0.9128922292934346,
+      "grad_norm": 0.18913982808589935,
+      "learning_rate": 1.9844706020813835e-05,
+      "loss": 1.242,
+      "step": 2452
+    },
+    {
+      "epoch": 0.9132645344440437,
+      "grad_norm": 0.17225903272628784,
+      "learning_rate": 1.9844492599918333e-05,
+      "loss": 1.2557,
+      "step": 2453
+    },
+    {
+      "epoch": 0.9136368395946528,
+      "grad_norm": 0.17968986928462982,
+      "learning_rate": 1.984427903362029e-05,
+      "loss": 1.2457,
+      "step": 2454
+    },
+    {
+      "epoch": 0.9140091447452618,
+      "grad_norm": 0.17983782291412354,
+      "learning_rate": 1.9844065321922867e-05,
+      "loss": 1.2449,
+      "step": 2455
+    },
+    {
+      "epoch": 0.9143814498958709,
+      "grad_norm": 0.17764170467853546,
+      "learning_rate": 1.9843851464829216e-05,
+      "loss": 1.241,
+      "step": 2456
+    },
+    {
+      "epoch": 0.9147537550464799,
+      "grad_norm": 0.17187097668647766,
+      "learning_rate": 1.9843637462342498e-05,
+      "loss": 1.2388,
+      "step": 2457
+    },
+    {
+      "epoch": 0.9151260601970891,
+      "grad_norm": 0.18232551217079163,
+      "learning_rate": 1.984342331446587e-05,
+      "loss": 1.2553,
+      "step": 2458
+    },
+    {
+      "epoch": 0.9154983653476981,
+      "grad_norm": 0.19238729774951935,
+      "learning_rate": 1.9843209021202496e-05,
+      "loss": 1.2357,
+      "step": 2459
+    },
+    {
+      "epoch": 0.9158706704983072,
+      "grad_norm": 0.1838630735874176,
+      "learning_rate": 1.9842994582555546e-05,
+      "loss": 1.2509,
+      "step": 2460
+    },
+    {
+      "epoch": 0.9162429756489162,
+      "grad_norm": 0.17731057107448578,
+      "learning_rate": 1.984277999852818e-05,
+      "loss": 1.2439,
+      "step": 2461
+    },
+    {
+      "epoch": 0.9166152807995253,
+      "grad_norm": 0.1820240616798401,
+      "learning_rate": 1.9842565269123577e-05,
+      "loss": 1.2467,
+      "step": 2462
+    },
+    {
+      "epoch": 0.9169875859501344,
+      "grad_norm": 0.1758500039577484,
+      "learning_rate": 1.9842350394344898e-05,
+      "loss": 1.2482,
+      "step": 2463
+    },
+    {
+      "epoch": 0.9173598911007435,
+      "grad_norm": 0.18491673469543457,
+      "learning_rate": 1.984213537419532e-05,
+      "loss": 1.2456,
+      "step": 2464
+    },
+    {
+      "epoch": 0.9177321962513525,
+      "grad_norm": 0.17699192464351654,
+      "learning_rate": 1.9841920208678024e-05,
+      "loss": 1.2429,
+      "step": 2465
+    },
+    {
+      "epoch": 0.9181045014019615,
+      "grad_norm": 0.17381340265274048,
+      "learning_rate": 1.9841704897796185e-05,
+      "loss": 1.2492,
+      "step": 2466
+    },
+    {
+      "epoch": 0.9184768065525707,
+      "grad_norm": 0.1726997047662735,
+      "learning_rate": 1.984148944155298e-05,
+      "loss": 1.2569,
+      "step": 2467
+    },
+    {
+      "epoch": 0.9188491117031797,
+      "grad_norm": 0.18082614243030548,
+      "learning_rate": 1.9841273839951595e-05,
+      "loss": 1.2426,
+      "step": 2468
+    },
+    {
+      "epoch": 0.9192214168537888,
+      "grad_norm": 0.17732475697994232,
+      "learning_rate": 1.984105809299521e-05,
+      "loss": 1.2449,
+      "step": 2469
+    },
+    {
+      "epoch": 0.9195937220043978,
+      "grad_norm": 0.17159555852413177,
+      "learning_rate": 1.9840842200687014e-05,
+      "loss": 1.2588,
+      "step": 2470
+    },
+    {
+      "epoch": 0.919966027155007,
+      "grad_norm": 0.17913275957107544,
+      "learning_rate": 1.9840626163030202e-05,
+      "loss": 1.2564,
+      "step": 2471
+    },
+    {
+      "epoch": 0.920338332305616,
+      "grad_norm": 0.17932996153831482,
+      "learning_rate": 1.9840409980027954e-05,
+      "loss": 1.2485,
+      "step": 2472
+    },
+    {
+      "epoch": 0.9207106374562251,
+      "grad_norm": 0.17403729259967804,
+      "learning_rate": 1.984019365168347e-05,
+      "loss": 1.2463,
+      "step": 2473
+    },
+    {
+      "epoch": 0.9210829426068341,
+      "grad_norm": 0.18577389419078827,
+      "learning_rate": 1.9839977177999942e-05,
+      "loss": 1.2486,
+      "step": 2474
+    },
+    {
+      "epoch": 0.9214552477574431,
+      "grad_norm": 0.1704840064048767,
+      "learning_rate": 1.9839760558980572e-05,
+      "loss": 1.2459,
+      "step": 2475
+    },
+    {
+      "epoch": 0.9218275529080523,
+      "grad_norm": 0.18757209181785583,
+      "learning_rate": 1.9839543794628553e-05,
+      "loss": 1.2486,
+      "step": 2476
+    },
+    {
+      "epoch": 0.9221998580586613,
+      "grad_norm": 0.18660812079906464,
+      "learning_rate": 1.9839326884947093e-05,
+      "loss": 1.2705,
+      "step": 2477
+    },
+    {
+      "epoch": 0.9225721632092704,
+      "grad_norm": 0.16449618339538574,
+      "learning_rate": 1.9839109829939388e-05,
+      "loss": 1.24,
+      "step": 2478
+    },
+    {
+      "epoch": 0.9229444683598794,
+      "grad_norm": 0.16529710590839386,
+      "learning_rate": 1.9838892629608652e-05,
+      "loss": 1.2578,
+      "step": 2479
+    },
+    {
+      "epoch": 0.9233167735104886,
+      "grad_norm": 0.17828698456287384,
+      "learning_rate": 1.9838675283958087e-05,
+      "loss": 1.256,
+      "step": 2480
+    },
+    {
+      "epoch": 0.9236890786610976,
+      "grad_norm": 0.17406156659126282,
+      "learning_rate": 1.9838457792990902e-05,
+      "loss": 1.2568,
+      "step": 2481
+    },
+    {
+      "epoch": 0.9240613838117067,
+      "grad_norm": 0.1763119399547577,
+      "learning_rate": 1.983824015671032e-05,
+      "loss": 1.2539,
+      "step": 2482
+    },
+    {
+      "epoch": 0.9244336889623157,
+      "grad_norm": 0.1707441657781601,
+      "learning_rate": 1.9838022375119544e-05,
+      "loss": 1.2391,
+      "step": 2483
+    },
+    {
+      "epoch": 0.9248059941129249,
+      "grad_norm": 0.1721871942281723,
+      "learning_rate": 1.9837804448221798e-05,
+      "loss": 1.2559,
+      "step": 2484
+    },
+    {
+      "epoch": 0.9251782992635339,
+      "grad_norm": 0.17778955399990082,
+      "learning_rate": 1.9837586376020293e-05,
+      "loss": 1.2563,
+      "step": 2485
+    },
+    {
+      "epoch": 0.9255506044141429,
+      "grad_norm": 0.17989608645439148,
+      "learning_rate": 1.983736815851826e-05,
+      "loss": 1.2358,
+      "step": 2486
+    },
+    {
+      "epoch": 0.925922909564752,
+      "grad_norm": 0.17223800718784332,
+      "learning_rate": 1.9837149795718913e-05,
+      "loss": 1.2619,
+      "step": 2487
+    },
+    {
+      "epoch": 0.926295214715361,
+      "grad_norm": 0.17492277920246124,
+      "learning_rate": 1.983693128762548e-05,
+      "loss": 1.24,
+      "step": 2488
+    },
+    {
+      "epoch": 0.9266675198659702,
+      "grad_norm": 0.17337973415851593,
+      "learning_rate": 1.9836712634241194e-05,
+      "loss": 1.2582,
+      "step": 2489
+    },
+    {
+      "epoch": 0.9270398250165792,
+      "grad_norm": 0.16389819979667664,
+      "learning_rate": 1.9836493835569278e-05,
+      "loss": 1.2379,
+      "step": 2490
+    },
+    {
+      "epoch": 0.9274121301671883,
+      "grad_norm": 0.17160442471504211,
+      "learning_rate": 1.9836274891612963e-05,
+      "loss": 1.2405,
+      "step": 2491
+    },
+    {
+      "epoch": 0.9277844353177973,
+      "grad_norm": 0.1783936619758606,
+      "learning_rate": 1.9836055802375488e-05,
+      "loss": 1.2555,
+      "step": 2492
+    },
+    {
+      "epoch": 0.9281567404684065,
+      "grad_norm": 0.17412415146827698,
+      "learning_rate": 1.9835836567860082e-05,
+      "loss": 1.2471,
+      "step": 2493
+    },
+    {
+      "epoch": 0.9285290456190155,
+      "grad_norm": 0.175662100315094,
+      "learning_rate": 1.983561718806999e-05,
+      "loss": 1.2335,
+      "step": 2494
+    },
+    {
+      "epoch": 0.9289013507696245,
+      "grad_norm": 0.1671876162290573,
+      "learning_rate": 1.983539766300845e-05,
+      "loss": 1.2543,
+      "step": 2495
+    },
+    {
+      "epoch": 0.9292736559202336,
+      "grad_norm": 0.16467541456222534,
+      "learning_rate": 1.9835177992678704e-05,
+      "loss": 1.2454,
+      "step": 2496
+    },
+    {
+      "epoch": 0.9296459610708427,
+      "grad_norm": 0.17586293816566467,
+      "learning_rate": 1.9834958177083995e-05,
+      "loss": 1.2568,
+      "step": 2497
+    },
+    {
+      "epoch": 0.9300182662214518,
+      "grad_norm": 0.170370951294899,
+      "learning_rate": 1.983473821622757e-05,
+      "loss": 1.2479,
+      "step": 2498
+    },
+    {
+      "epoch": 0.9303905713720608,
+      "grad_norm": 0.17328549921512604,
+      "learning_rate": 1.983451811011268e-05,
+      "loss": 1.2398,
+      "step": 2499
+    },
+    {
+      "epoch": 0.9307628765226699,
+      "grad_norm": 0.18288347125053406,
+      "learning_rate": 1.9834297858742574e-05,
+      "loss": 1.2438,
+      "step": 2500
+    },
+    {
+      "epoch": 0.9307628765226699,
+      "eval_loss": 1.3256843090057373,
+      "eval_runtime": 16.4567,
+      "eval_samples_per_second": 105.368,
+      "eval_steps_per_second": 5.287,
+      "step": 2500
+    },
+    {
+      "epoch": 0.9311351816732789,
+      "grad_norm": 0.17619635164737701,
+      "learning_rate": 1.9834077462120506e-05,
+      "loss": 1.2541,
+      "step": 2501
+    },
+    {
+      "epoch": 0.9315074868238881,
+      "grad_norm": 0.1839938461780548,
+      "learning_rate": 1.9833856920249733e-05,
+      "loss": 1.2747,
+      "step": 2502
+    },
+    {
+      "epoch": 0.9318797919744971,
+      "grad_norm": 0.18233221769332886,
+      "learning_rate": 1.9833636233133507e-05,
+      "loss": 1.2484,
+      "step": 2503
+    },
+    {
+      "epoch": 0.9322520971251061,
+      "grad_norm": 0.1671859472990036,
+      "learning_rate": 1.9833415400775092e-05,
+      "loss": 1.2481,
+      "step": 2504
+    },
+    {
+      "epoch": 0.9326244022757152,
+      "grad_norm": 0.17161886394023895,
+      "learning_rate": 1.9833194423177754e-05,
+      "loss": 1.2584,
+      "step": 2505
+    },
+    {
+      "epoch": 0.9329967074263243,
+      "grad_norm": 0.18463842570781708,
+      "learning_rate": 1.9832973300344745e-05,
+      "loss": 1.257,
+      "step": 2506
+    },
+    {
+      "epoch": 0.9333690125769334,
+      "grad_norm": 0.17637981474399567,
+      "learning_rate": 1.983275203227934e-05,
+      "loss": 1.2577,
+      "step": 2507
+    },
+    {
+      "epoch": 0.9337413177275424,
+      "grad_norm": 0.17521126568317413,
+      "learning_rate": 1.9832530618984802e-05,
+      "loss": 1.2564,
+      "step": 2508
+    },
+    {
+      "epoch": 0.9341136228781515,
+      "grad_norm": 0.17385391891002655,
+      "learning_rate": 1.9832309060464408e-05,
+      "loss": 1.245,
+      "step": 2509
+    },
+    {
+      "epoch": 0.9344859280287606,
+      "grad_norm": 0.16838274896144867,
+      "learning_rate": 1.9832087356721424e-05,
+      "loss": 1.2586,
+      "step": 2510
+    },
+    {
+      "epoch": 0.9348582331793697,
+      "grad_norm": 0.17618712782859802,
+      "learning_rate": 1.9831865507759125e-05,
+      "loss": 1.2478,
+      "step": 2511
+    },
+    {
+      "epoch": 0.9352305383299787,
+      "grad_norm": 0.17693477869033813,
+      "learning_rate": 1.983164351358079e-05,
+      "loss": 1.2573,
+      "step": 2512
+    },
+    {
+      "epoch": 0.9356028434805878,
+      "grad_norm": 0.17031066119670868,
+      "learning_rate": 1.9831421374189702e-05,
+      "loss": 1.2538,
+      "step": 2513
+    },
+    {
+      "epoch": 0.9359751486311968,
+      "grad_norm": 0.1740039438009262,
+      "learning_rate": 1.983119908958913e-05,
+      "loss": 1.2496,
+      "step": 2514
+    },
+    {
+      "epoch": 0.936347453781806,
+      "grad_norm": 0.18025629222393036,
+      "learning_rate": 1.983097665978237e-05,
+      "loss": 1.2503,
+      "step": 2515
+    },
+    {
+      "epoch": 0.936719758932415,
+      "grad_norm": 0.179831400513649,
+      "learning_rate": 1.98307540847727e-05,
+      "loss": 1.2372,
+      "step": 2516
+    },
+    {
+      "epoch": 0.937092064083024,
+      "grad_norm": 0.17471134662628174,
+      "learning_rate": 1.983053136456341e-05,
+      "loss": 1.2489,
+      "step": 2517
+    },
+    {
+      "epoch": 0.9374643692336331,
+      "grad_norm": 0.17015781998634338,
+      "learning_rate": 1.9830308499157787e-05,
+      "loss": 1.2523,
+      "step": 2518
+    },
+    {
+      "epoch": 0.9378366743842422,
+      "grad_norm": 0.17395399510860443,
+      "learning_rate": 1.9830085488559128e-05,
+      "loss": 1.2493,
+      "step": 2519
+    },
+    {
+      "epoch": 0.9382089795348513,
+      "grad_norm": 0.1732756495475769,
+      "learning_rate": 1.982986233277072e-05,
+      "loss": 1.2487,
+      "step": 2520
+    },
+    {
+      "epoch": 0.9385812846854603,
+      "grad_norm": 0.1738785058259964,
+      "learning_rate": 1.9829639031795862e-05,
+      "loss": 1.254,
+      "step": 2521
+    },
+    {
+      "epoch": 0.9389535898360694,
+      "grad_norm": 0.17151622474193573,
+      "learning_rate": 1.9829415585637853e-05,
+      "loss": 1.2453,
+      "step": 2522
+    },
+    {
+      "epoch": 0.9393258949866785,
+      "grad_norm": 0.16740648448467255,
+      "learning_rate": 1.982919199429999e-05,
+      "loss": 1.2497,
+      "step": 2523
+    },
+    {
+      "epoch": 0.9396982001372876,
+      "grad_norm": 0.1763676106929779,
+      "learning_rate": 1.9828968257785582e-05,
+      "loss": 1.2484,
+      "step": 2524
+    },
+    {
+      "epoch": 0.9400705052878966,
+      "grad_norm": 0.17714397609233856,
+      "learning_rate": 1.982874437609793e-05,
+      "loss": 1.2431,
+      "step": 2525
+    },
+    {
+      "epoch": 0.9404428104385056,
+      "grad_norm": 0.16998568177223206,
+      "learning_rate": 1.982852034924034e-05,
+      "loss": 1.2378,
+      "step": 2526
+    },
+    {
+      "epoch": 0.9408151155891147,
+      "grad_norm": 0.1771582067012787,
+      "learning_rate": 1.9828296177216118e-05,
+      "loss": 1.2533,
+      "step": 2527
+    },
+    {
+      "epoch": 0.9411874207397238,
+      "grad_norm": 0.18139605224132538,
+      "learning_rate": 1.9828071860028582e-05,
+      "loss": 1.2489,
+      "step": 2528
+    },
+    {
+      "epoch": 0.9415597258903329,
+      "grad_norm": 0.17635215818881989,
+      "learning_rate": 1.982784739768104e-05,
+      "loss": 1.2437,
+      "step": 2529
+    },
+    {
+      "epoch": 0.9419320310409419,
+      "grad_norm": 0.17219685018062592,
+      "learning_rate": 1.982762279017681e-05,
+      "loss": 1.2348,
+      "step": 2530
+    },
+    {
+      "epoch": 0.942304336191551,
+      "grad_norm": 0.18018695712089539,
+      "learning_rate": 1.9827398037519203e-05,
+      "loss": 1.2448,
+      "step": 2531
+    },
+    {
+      "epoch": 0.9426766413421601,
+      "grad_norm": 0.17004595696926117,
+      "learning_rate": 1.9827173139711547e-05,
+      "loss": 1.2489,
+      "step": 2532
+    },
+    {
+      "epoch": 0.9430489464927692,
+      "grad_norm": 0.17758417129516602,
+      "learning_rate": 1.9826948096757163e-05,
+      "loss": 1.2598,
+      "step": 2533
+    },
+    {
+      "epoch": 0.9434212516433782,
+      "grad_norm": 0.18373461067676544,
+      "learning_rate": 1.982672290865937e-05,
+      "loss": 1.2601,
+      "step": 2534
+    },
+    {
+      "epoch": 0.9437935567939872,
+      "grad_norm": 0.17354753613471985,
+      "learning_rate": 1.9826497575421498e-05,
+      "loss": 1.2353,
+      "step": 2535
+    },
+    {
+      "epoch": 0.9441658619445963,
+      "grad_norm": 0.179754376411438,
+      "learning_rate": 1.982627209704687e-05,
+      "loss": 1.2466,
+      "step": 2536
+    },
+    {
+      "epoch": 0.9445381670952054,
+      "grad_norm": 0.16391302645206451,
+      "learning_rate": 1.9826046473538823e-05,
+      "loss": 1.2386,
+      "step": 2537
+    },
+    {
+      "epoch": 0.9449104722458145,
+      "grad_norm": 0.1725778728723526,
+      "learning_rate": 1.9825820704900684e-05,
+      "loss": 1.2573,
+      "step": 2538
+    },
+    {
+      "epoch": 0.9452827773964235,
+      "grad_norm": 0.1723441183567047,
+      "learning_rate": 1.9825594791135792e-05,
+      "loss": 1.2355,
+      "step": 2539
+    },
+    {
+      "epoch": 0.9456550825470326,
+      "grad_norm": 0.16522756218910217,
+      "learning_rate": 1.982536873224748e-05,
+      "loss": 1.2386,
+      "step": 2540
+    },
+    {
+      "epoch": 0.9460273876976417,
+      "grad_norm": 0.1779457926750183,
+      "learning_rate": 1.982514252823909e-05,
+      "loss": 1.2522,
+      "step": 2541
+    },
+    {
+      "epoch": 0.9463996928482508,
+      "grad_norm": 0.17980381846427917,
+      "learning_rate": 1.982491617911396e-05,
+      "loss": 1.2639,
+      "step": 2542
+    },
+    {
+      "epoch": 0.9467719979988598,
+      "grad_norm": 0.17201921343803406,
+      "learning_rate": 1.9824689684875435e-05,
+      "loss": 1.2428,
+      "step": 2543
+    },
+    {
+      "epoch": 0.9471443031494688,
+      "grad_norm": 0.16897398233413696,
+      "learning_rate": 1.9824463045526857e-05,
+      "loss": 1.2488,
+      "step": 2544
+    },
+    {
+      "epoch": 0.947516608300078,
+      "grad_norm": 0.1801498979330063,
+      "learning_rate": 1.982423626107158e-05,
+      "loss": 1.2434,
+      "step": 2545
+    },
+    {
+      "epoch": 0.947888913450687,
+      "grad_norm": 0.1762707233428955,
+      "learning_rate": 1.982400933151295e-05,
+      "loss": 1.2467,
+      "step": 2546
+    },
+    {
+      "epoch": 0.9482612186012961,
+      "grad_norm": 0.1682814210653305,
+      "learning_rate": 1.982378225685432e-05,
+      "loss": 1.2472,
+      "step": 2547
+    },
+    {
+      "epoch": 0.9486335237519051,
+      "grad_norm": 0.17276160418987274,
+      "learning_rate": 1.982355503709904e-05,
+      "loss": 1.2439,
+      "step": 2548
+    },
+    {
+      "epoch": 0.9490058289025142,
+      "grad_norm": 0.16808001697063446,
+      "learning_rate": 1.982332767225047e-05,
+      "loss": 1.245,
+      "step": 2549
+    },
+    {
+      "epoch": 0.9493781340531233,
+      "grad_norm": 0.17079442739486694,
+      "learning_rate": 1.9823100162311967e-05,
+      "loss": 1.2439,
+      "step": 2550
+    },
+    {
+      "epoch": 0.9497504392037324,
+      "grad_norm": 0.17456983029842377,
+      "learning_rate": 1.982287250728689e-05,
+      "loss": 1.2551,
+      "step": 2551
+    },
+    {
+      "epoch": 0.9501227443543414,
+      "grad_norm": 0.1688481867313385,
+      "learning_rate": 1.98226447071786e-05,
+      "loss": 1.2487,
+      "step": 2552
+    },
+    {
+      "epoch": 0.9504950495049505,
+      "grad_norm": 0.1739954799413681,
+      "learning_rate": 1.982241676199047e-05,
+      "loss": 1.2556,
+      "step": 2553
+    },
+    {
+      "epoch": 0.9508673546555596,
+      "grad_norm": 0.1756342649459839,
+      "learning_rate": 1.9822188671725854e-05,
+      "loss": 1.2437,
+      "step": 2554
+    },
+    {
+      "epoch": 0.9512396598061686,
+      "grad_norm": 0.17892800271511078,
+      "learning_rate": 1.9821960436388134e-05,
+      "loss": 1.2401,
+      "step": 2555
+    },
+    {
+      "epoch": 0.9516119649567777,
+      "grad_norm": 0.17654842138290405,
+      "learning_rate": 1.9821732055980673e-05,
+      "loss": 1.2459,
+      "step": 2556
+    },
+    {
+      "epoch": 0.9519842701073867,
+      "grad_norm": 0.18467935919761658,
+      "learning_rate": 1.9821503530506843e-05,
+      "loss": 1.2376,
+      "step": 2557
+    },
+    {
+      "epoch": 0.9523565752579959,
+      "grad_norm": 0.1839521825313568,
+      "learning_rate": 1.9821274859970025e-05,
+      "loss": 1.2529,
+      "step": 2558
+    },
+    {
+      "epoch": 0.9527288804086049,
+      "grad_norm": 0.17149806022644043,
+      "learning_rate": 1.982104604437359e-05,
+      "loss": 1.2576,
+      "step": 2559
+    },
+    {
+      "epoch": 0.953101185559214,
+      "grad_norm": 0.17172785103321075,
+      "learning_rate": 1.9820817083720928e-05,
+      "loss": 1.2464,
+      "step": 2560
+    },
+    {
+      "epoch": 0.953473490709823,
+      "grad_norm": 0.17375504970550537,
+      "learning_rate": 1.9820587978015407e-05,
+      "loss": 1.2578,
+      "step": 2561
+    },
+    {
+      "epoch": 0.953845795860432,
+      "grad_norm": 0.17615382373332977,
+      "learning_rate": 1.982035872726042e-05,
+      "loss": 1.2555,
+      "step": 2562
+    },
+    {
+      "epoch": 0.9542181010110412,
+      "grad_norm": 0.18444810807704926,
+      "learning_rate": 1.9820129331459354e-05,
+      "loss": 1.2337,
+      "step": 2563
+    },
+    {
+      "epoch": 0.9545904061616503,
+      "grad_norm": 0.1793602854013443,
+      "learning_rate": 1.981989979061559e-05,
+      "loss": 1.2504,
+      "step": 2564
+    },
+    {
+      "epoch": 0.9549627113122593,
+      "grad_norm": 0.16998952627182007,
+      "learning_rate": 1.9819670104732528e-05,
+      "loss": 1.2424,
+      "step": 2565
+    },
+    {
+      "epoch": 0.9553350164628683,
+      "grad_norm": 0.16896101832389832,
+      "learning_rate": 1.981944027381355e-05,
+      "loss": 1.239,
+      "step": 2566
+    },
+    {
+      "epoch": 0.9557073216134775,
+      "grad_norm": 0.16461212933063507,
+      "learning_rate": 1.9819210297862055e-05,
+      "loss": 1.245,
+      "step": 2567
+    },
+    {
+      "epoch": 0.9560796267640865,
+      "grad_norm": 0.1697378307580948,
+      "learning_rate": 1.981898017688144e-05,
+      "loss": 1.2506,
+      "step": 2568
+    },
+    {
+      "epoch": 0.9564519319146956,
+      "grad_norm": 0.17256854474544525,
+      "learning_rate": 1.981874991087511e-05,
+      "loss": 1.2113,
+      "step": 2569
+    },
+    {
+      "epoch": 0.9568242370653046,
+      "grad_norm": 0.1716059297323227,
+      "learning_rate": 1.9818519499846457e-05,
+      "loss": 1.2552,
+      "step": 2570
+    },
+    {
+      "epoch": 0.9571965422159138,
+      "grad_norm": 0.16681870818138123,
+      "learning_rate": 1.9818288943798882e-05,
+      "loss": 1.2412,
+      "step": 2571
+    },
+    {
+      "epoch": 0.9575688473665228,
+      "grad_norm": 0.1805577576160431,
+      "learning_rate": 1.98180582427358e-05,
+      "loss": 1.2393,
+      "step": 2572
+    },
+    {
+      "epoch": 0.9579411525171319,
+      "grad_norm": 0.16774867475032806,
+      "learning_rate": 1.9817827396660615e-05,
+      "loss": 1.2327,
+      "step": 2573
+    },
+    {
+      "epoch": 0.9583134576677409,
+      "grad_norm": 0.1805572658777237,
+      "learning_rate": 1.9817596405576733e-05,
+      "loss": 1.2445,
+      "step": 2574
+    },
+    {
+      "epoch": 0.9586857628183499,
+      "grad_norm": 0.16827546060085297,
+      "learning_rate": 1.981736526948757e-05,
+      "loss": 1.2472,
+      "step": 2575
+    },
+    {
+      "epoch": 0.9590580679689591,
+      "grad_norm": 0.184850811958313,
+      "learning_rate": 1.9817133988396536e-05,
+      "loss": 1.2344,
+      "step": 2576
+    },
+    {
+      "epoch": 0.9594303731195681,
+      "grad_norm": 0.17866064608097076,
+      "learning_rate": 1.981690256230705e-05,
+      "loss": 1.247,
+      "step": 2577
+    },
+    {
+      "epoch": 0.9598026782701772,
+      "grad_norm": 0.17331889271736145,
+      "learning_rate": 1.981667099122253e-05,
+      "loss": 1.2627,
+      "step": 2578
+    },
+    {
+      "epoch": 0.9601749834207862,
+      "grad_norm": 0.17190296947956085,
+      "learning_rate": 1.9816439275146394e-05,
+      "loss": 1.2344,
+      "step": 2579
+    },
+    {
+      "epoch": 0.9605472885713954,
+      "grad_norm": 0.1726665496826172,
+      "learning_rate": 1.981620741408207e-05,
+      "loss": 1.247,
+      "step": 2580
+    },
+    {
+      "epoch": 0.9609195937220044,
+      "grad_norm": 0.1653529554605484,
+      "learning_rate": 1.9815975408032972e-05,
+      "loss": 1.2472,
+      "step": 2581
+    },
+    {
+      "epoch": 0.9612918988726135,
+      "grad_norm": 0.18165260553359985,
+      "learning_rate": 1.9815743257002537e-05,
+      "loss": 1.2457,
+      "step": 2582
+    },
+    {
+      "epoch": 0.9616642040232225,
+      "grad_norm": 0.18499329686164856,
+      "learning_rate": 1.981551096099419e-05,
+      "loss": 1.251,
+      "step": 2583
+    },
+    {
+      "epoch": 0.9620365091738317,
+      "grad_norm": 0.17018747329711914,
+      "learning_rate": 1.9815278520011364e-05,
+      "loss": 1.2342,
+      "step": 2584
+    },
+    {
+      "epoch": 0.9624088143244407,
+      "grad_norm": 0.17789021134376526,
+      "learning_rate": 1.9815045934057487e-05,
+      "loss": 1.2406,
+      "step": 2585
+    },
+    {
+      "epoch": 0.9627811194750497,
+      "grad_norm": 0.1729610115289688,
+      "learning_rate": 1.9814813203135998e-05,
+      "loss": 1.2466,
+      "step": 2586
+    },
+    {
+      "epoch": 0.9631534246256588,
+      "grad_norm": 0.17735299468040466,
+      "learning_rate": 1.9814580327250336e-05,
+      "loss": 1.2406,
+      "step": 2587
+    },
+    {
+      "epoch": 0.9635257297762678,
+      "grad_norm": 0.1712779998779297,
+      "learning_rate": 1.9814347306403936e-05,
+      "loss": 1.2408,
+      "step": 2588
+    },
+    {
+      "epoch": 0.963898034926877,
+      "grad_norm": 0.18011993169784546,
+      "learning_rate": 1.981411414060024e-05,
+      "loss": 1.2578,
+      "step": 2589
+    },
+    {
+      "epoch": 0.964270340077486,
+      "grad_norm": 0.17510679364204407,
+      "learning_rate": 1.9813880829842704e-05,
+      "loss": 1.2324,
+      "step": 2590
+    },
+    {
+      "epoch": 0.9646426452280951,
+      "grad_norm": 0.17391736805438995,
+      "learning_rate": 1.9813647374134756e-05,
+      "loss": 1.2433,
+      "step": 2591
+    },
+    {
+      "epoch": 0.9650149503787041,
+      "grad_norm": 0.17409949004650116,
+      "learning_rate": 1.9813413773479853e-05,
+      "loss": 1.2542,
+      "step": 2592
+    },
+    {
+      "epoch": 0.9653872555293133,
+      "grad_norm": 0.17249634861946106,
+      "learning_rate": 1.9813180027881445e-05,
+      "loss": 1.231,
+      "step": 2593
+    },
+    {
+      "epoch": 0.9657595606799223,
+      "grad_norm": 0.18046285212039948,
+      "learning_rate": 1.9812946137342984e-05,
+      "loss": 1.2408,
+      "step": 2594
+    },
+    {
+      "epoch": 0.9661318658305313,
+      "grad_norm": 0.16898088157176971,
+      "learning_rate": 1.9812712101867923e-05,
+      "loss": 1.2426,
+      "step": 2595
+    },
+    {
+      "epoch": 0.9665041709811404,
+      "grad_norm": 0.17144586145877838,
+      "learning_rate": 1.9812477921459724e-05,
+      "loss": 1.2358,
+      "step": 2596
+    },
+    {
+      "epoch": 0.9668764761317495,
+      "grad_norm": 0.17590296268463135,
+      "learning_rate": 1.981224359612184e-05,
+      "loss": 1.231,
+      "step": 2597
+    },
+    {
+      "epoch": 0.9672487812823586,
+      "grad_norm": 0.1714339703321457,
+      "learning_rate": 1.981200912585773e-05,
+      "loss": 1.2307,
+      "step": 2598
+    },
+    {
+      "epoch": 0.9676210864329676,
+      "grad_norm": 0.17280226945877075,
+      "learning_rate": 1.9811774510670866e-05,
+      "loss": 1.2432,
+      "step": 2599
+    },
+    {
+      "epoch": 0.9679933915835767,
+      "grad_norm": 0.19417522847652435,
+      "learning_rate": 1.9811539750564702e-05,
+      "loss": 1.2329,
+      "step": 2600
+    },
+    {
+      "epoch": 0.9683656967341857,
+      "grad_norm": 0.18655581772327423,
+      "learning_rate": 1.9811304845542717e-05,
+      "loss": 1.2364,
+      "step": 2601
+    },
+    {
+      "epoch": 0.9687380018847949,
+      "grad_norm": 0.1713135540485382,
+      "learning_rate": 1.9811069795608377e-05,
+      "loss": 1.266,
+      "step": 2602
+    },
+    {
+      "epoch": 0.9691103070354039,
+      "grad_norm": 0.18093515932559967,
+      "learning_rate": 1.9810834600765148e-05,
+      "loss": 1.2466,
+      "step": 2603
+    },
+    {
+      "epoch": 0.969482612186013,
+      "grad_norm": 0.17053551971912384,
+      "learning_rate": 1.9810599261016506e-05,
+      "loss": 1.2532,
+      "step": 2604
+    },
+    {
+      "epoch": 0.969854917336622,
+      "grad_norm": 0.1664905697107315,
+      "learning_rate": 1.9810363776365932e-05,
+      "loss": 1.232,
+      "step": 2605
+    },
+    {
+      "epoch": 0.9702272224872311,
+      "grad_norm": 0.1823788583278656,
+      "learning_rate": 1.98101281468169e-05,
+      "loss": 1.255,
+      "step": 2606
+    },
+    {
+      "epoch": 0.9705995276378402,
+      "grad_norm": 0.17830835282802582,
+      "learning_rate": 1.980989237237289e-05,
+      "loss": 1.2522,
+      "step": 2607
+    },
+    {
+      "epoch": 0.9709718327884492,
+      "grad_norm": 0.19234777987003326,
+      "learning_rate": 1.980965645303739e-05,
+      "loss": 1.2574,
+      "step": 2608
+    },
+    {
+      "epoch": 0.9713441379390583,
+      "grad_norm": 0.18148033320903778,
+      "learning_rate": 1.9809420388813874e-05,
+      "loss": 1.2511,
+      "step": 2609
+    },
+    {
+      "epoch": 0.9717164430896673,
+      "grad_norm": 0.1736968457698822,
+      "learning_rate": 1.9809184179705835e-05,
+      "loss": 1.2458,
+      "step": 2610
+    },
+    {
+      "epoch": 0.9720887482402765,
+      "grad_norm": 0.17815859615802765,
+      "learning_rate": 1.9808947825716768e-05,
+      "loss": 1.2548,
+      "step": 2611
+    },
+    {
+      "epoch": 0.9724610533908855,
+      "grad_norm": 0.1661943793296814,
+      "learning_rate": 1.980871132685015e-05,
+      "loss": 1.2295,
+      "step": 2612
+    },
+    {
+      "epoch": 0.9728333585414946,
+      "grad_norm": 0.16622133553028107,
+      "learning_rate": 1.980847468310948e-05,
+      "loss": 1.2403,
+      "step": 2613
+    },
+    {
+      "epoch": 0.9732056636921036,
+      "grad_norm": 0.17739859223365784,
+      "learning_rate": 1.980823789449826e-05,
+      "loss": 1.2407,
+      "step": 2614
+    },
+    {
+      "epoch": 0.9735779688427127,
+      "grad_norm": 0.17749415338039398,
+      "learning_rate": 1.980800096101998e-05,
+      "loss": 1.236,
+      "step": 2615
+    },
+    {
+      "epoch": 0.9739502739933218,
+      "grad_norm": 0.18621912598609924,
+      "learning_rate": 1.9807763882678143e-05,
+      "loss": 1.2393,
+      "step": 2616
+    },
+    {
+      "epoch": 0.9743225791439308,
+      "grad_norm": 0.1727140098810196,
+      "learning_rate": 1.9807526659476245e-05,
+      "loss": 1.2331,
+      "step": 2617
+    },
+    {
+      "epoch": 0.9746948842945399,
+      "grad_norm": 0.17797011137008667,
+      "learning_rate": 1.9807289291417795e-05,
+      "loss": 1.2603,
+      "step": 2618
+    },
+    {
+      "epoch": 0.975067189445149,
+      "grad_norm": 0.18820150196552277,
+      "learning_rate": 1.98070517785063e-05,
+      "loss": 1.2385,
+      "step": 2619
+    },
+    {
+      "epoch": 0.9754394945957581,
+      "grad_norm": 0.17448899149894714,
+      "learning_rate": 1.9806814120745265e-05,
+      "loss": 1.2361,
+      "step": 2620
+    },
+    {
+      "epoch": 0.9758117997463671,
+      "grad_norm": 0.1788555085659027,
+      "learning_rate": 1.9806576318138194e-05,
+      "loss": 1.2421,
+      "step": 2621
+    },
+    {
+      "epoch": 0.9761841048969762,
+      "grad_norm": 0.1790677309036255,
+      "learning_rate": 1.9806338370688615e-05,
+      "loss": 1.2379,
+      "step": 2622
+    },
+    {
+      "epoch": 0.9765564100475852,
+      "grad_norm": 0.17613591253757477,
+      "learning_rate": 1.980610027840003e-05,
+      "loss": 1.2557,
+      "step": 2623
+    },
+    {
+      "epoch": 0.9769287151981944,
+      "grad_norm": 0.19133977591991425,
+      "learning_rate": 1.9805862041275962e-05,
+      "loss": 1.2616,
+      "step": 2624
+    },
+    {
+      "epoch": 0.9773010203488034,
+      "grad_norm": 0.18040066957473755,
+      "learning_rate": 1.9805623659319924e-05,
+      "loss": 1.2437,
+      "step": 2625
+    },
+    {
+      "epoch": 0.9776733254994124,
+      "grad_norm": 0.17554713785648346,
+      "learning_rate": 1.980538513253544e-05,
+      "loss": 1.2483,
+      "step": 2626
+    },
+    {
+      "epoch": 0.9780456306500215,
+      "grad_norm": 0.17956022918224335,
+      "learning_rate": 1.9805146460926033e-05,
+      "loss": 1.2492,
+      "step": 2627
+    },
+    {
+      "epoch": 0.9784179358006306,
+      "grad_norm": 0.18192705512046814,
+      "learning_rate": 1.980490764449523e-05,
+      "loss": 1.2439,
+      "step": 2628
+    },
+    {
+      "epoch": 0.9787902409512397,
+      "grad_norm": 0.1905398815870285,
+      "learning_rate": 1.9804668683246556e-05,
+      "loss": 1.2384,
+      "step": 2629
+    },
+    {
+      "epoch": 0.9791625461018487,
+      "grad_norm": 0.17728392779827118,
+      "learning_rate": 1.980442957718354e-05,
+      "loss": 1.2492,
+      "step": 2630
+    },
+    {
+      "epoch": 0.9795348512524578,
+      "grad_norm": 0.17330513894557953,
+      "learning_rate": 1.9804190326309714e-05,
+      "loss": 1.2312,
+      "step": 2631
+    },
+    {
+      "epoch": 0.9799071564030669,
+      "grad_norm": 0.17659252882003784,
+      "learning_rate": 1.9803950930628616e-05,
+      "loss": 1.2429,
+      "step": 2632
+    },
+    {
+      "epoch": 0.980279461553676,
+      "grad_norm": 0.18615534901618958,
+      "learning_rate": 1.9803711390143774e-05,
+      "loss": 1.2442,
+      "step": 2633
+    },
+    {
+      "epoch": 0.980651766704285,
+      "grad_norm": 0.18511880934238434,
+      "learning_rate": 1.9803471704858733e-05,
+      "loss": 1.2489,
+      "step": 2634
+    },
+    {
+      "epoch": 0.981024071854894,
+      "grad_norm": 0.1814030408859253,
+      "learning_rate": 1.9803231874777025e-05,
+      "loss": 1.2361,
+      "step": 2635
+    },
+    {
+      "epoch": 0.9813963770055031,
+      "grad_norm": 0.1738506257534027,
+      "learning_rate": 1.98029918999022e-05,
+      "loss": 1.2462,
+      "step": 2636
+    },
+    {
+      "epoch": 0.9817686821561122,
+      "grad_norm": 0.1846051663160324,
+      "learning_rate": 1.98027517802378e-05,
+      "loss": 1.2436,
+      "step": 2637
+    },
+    {
+      "epoch": 0.9821409873067213,
+      "grad_norm": 0.19149626791477203,
+      "learning_rate": 1.9802511515787373e-05,
+      "loss": 1.2403,
+      "step": 2638
+    },
+    {
+      "epoch": 0.9825132924573303,
+      "grad_norm": 0.17990732192993164,
+      "learning_rate": 1.9802271106554464e-05,
+      "loss": 1.2479,
+      "step": 2639
+    },
+    {
+      "epoch": 0.9828855976079394,
+      "grad_norm": 0.18959732353687286,
+      "learning_rate": 1.9802030552542627e-05,
+      "loss": 1.238,
+      "step": 2640
+    },
+    {
+      "epoch": 0.9832579027585485,
+      "grad_norm": 0.1863657683134079,
+      "learning_rate": 1.9801789853755415e-05,
+      "loss": 1.2615,
+      "step": 2641
+    },
+    {
+      "epoch": 0.9836302079091576,
+      "grad_norm": 0.16739057004451752,
+      "learning_rate": 1.980154901019638e-05,
+      "loss": 1.2371,
+      "step": 2642
+    },
+    {
+      "epoch": 0.9840025130597666,
+      "grad_norm": 0.1758878231048584,
+      "learning_rate": 1.9801308021869084e-05,
+      "loss": 1.2301,
+      "step": 2643
+    },
+    {
+      "epoch": 0.9843748182103756,
+      "grad_norm": 0.18291743099689484,
+      "learning_rate": 1.9801066888777082e-05,
+      "loss": 1.2375,
+      "step": 2644
+    },
+    {
+      "epoch": 0.9847471233609848,
+      "grad_norm": 0.18052415549755096,
+      "learning_rate": 1.9800825610923937e-05,
+      "loss": 1.2571,
+      "step": 2645
+    },
+    {
+      "epoch": 0.9851194285115938,
+      "grad_norm": 0.17342287302017212,
+      "learning_rate": 1.980058418831321e-05,
+      "loss": 1.2584,
+      "step": 2646
+    },
+    {
+      "epoch": 0.9854917336622029,
+      "grad_norm": 0.17909403145313263,
+      "learning_rate": 1.9800342620948475e-05,
+      "loss": 1.2307,
+      "step": 2647
+    },
+    {
+      "epoch": 0.9858640388128119,
+      "grad_norm": 0.17488320171833038,
+      "learning_rate": 1.980010090883329e-05,
+      "loss": 1.2422,
+      "step": 2648
+    },
+    {
+      "epoch": 0.986236343963421,
+      "grad_norm": 0.17006643116474152,
+      "learning_rate": 1.9799859051971232e-05,
+      "loss": 1.2405,
+      "step": 2649
+    },
+    {
+      "epoch": 0.9866086491140301,
+      "grad_norm": 0.18193192780017853,
+      "learning_rate": 1.979961705036587e-05,
+      "loss": 1.2491,
+      "step": 2650
+    },
+    {
+      "epoch": 0.9869809542646392,
+      "grad_norm": 0.18248356878757477,
+      "learning_rate": 1.979937490402078e-05,
+      "loss": 1.2499,
+      "step": 2651
+    },
+    {
+      "epoch": 0.9873532594152482,
+      "grad_norm": 0.17710177600383759,
+      "learning_rate": 1.9799132612939535e-05,
+      "loss": 1.2346,
+      "step": 2652
+    },
+    {
+      "epoch": 0.9877255645658573,
+      "grad_norm": 0.17549732327461243,
+      "learning_rate": 1.979889017712572e-05,
+      "loss": 1.2341,
+      "step": 2653
+    },
+    {
+      "epoch": 0.9880978697164664,
+      "grad_norm": 0.17912618815898895,
+      "learning_rate": 1.979864759658291e-05,
+      "loss": 1.2562,
+      "step": 2654
+    },
+    {
+      "epoch": 0.9884701748670754,
+      "grad_norm": 0.17686498165130615,
+      "learning_rate": 1.979840487131469e-05,
+      "loss": 1.2422,
+      "step": 2655
+    },
+    {
+      "epoch": 0.9888424800176845,
+      "grad_norm": 0.1823192685842514,
+      "learning_rate": 1.9798162001324647e-05,
+      "loss": 1.2304,
+      "step": 2656
+    },
+    {
+      "epoch": 0.9892147851682935,
+      "grad_norm": 0.17622555792331696,
+      "learning_rate": 1.9797918986616362e-05,
+      "loss": 1.2241,
+      "step": 2657
+    },
+    {
+      "epoch": 0.9895870903189027,
+      "grad_norm": 0.1805817037820816,
+      "learning_rate": 1.979767582719343e-05,
+      "loss": 1.241,
+      "step": 2658
+    },
+    {
+      "epoch": 0.9899593954695117,
+      "grad_norm": 0.1678328961133957,
+      "learning_rate": 1.9797432523059442e-05,
+      "loss": 1.2248,
+      "step": 2659
+    },
+    {
+      "epoch": 0.9903317006201208,
+      "grad_norm": 0.17134131491184235,
+      "learning_rate": 1.9797189074217993e-05,
+      "loss": 1.2415,
+      "step": 2660
+    },
+    {
+      "epoch": 0.9907040057707298,
+      "grad_norm": 0.1789311319589615,
+      "learning_rate": 1.979694548067267e-05,
+      "loss": 1.2358,
+      "step": 2661
+    },
+    {
+      "epoch": 0.9910763109213389,
+      "grad_norm": 0.18833468854427338,
+      "learning_rate": 1.979670174242708e-05,
+      "loss": 1.2339,
+      "step": 2662
+    },
+    {
+      "epoch": 0.991448616071948,
+      "grad_norm": 0.18006418645381927,
+      "learning_rate": 1.9796457859484825e-05,
+      "loss": 1.2322,
+      "step": 2663
+    },
+    {
+      "epoch": 0.991820921222557,
+      "grad_norm": 0.1705029308795929,
+      "learning_rate": 1.9796213831849496e-05,
+      "loss": 1.2473,
+      "step": 2664
+    },
+    {
+      "epoch": 0.9921932263731661,
+      "grad_norm": 0.16797584295272827,
+      "learning_rate": 1.9795969659524705e-05,
+      "loss": 1.2422,
+      "step": 2665
+    },
+    {
+      "epoch": 0.9925655315237751,
+      "grad_norm": 0.19237132370471954,
+      "learning_rate": 1.9795725342514055e-05,
+      "loss": 1.2567,
+      "step": 2666
+    },
+    {
+      "epoch": 0.9929378366743843,
+      "grad_norm": 0.1747465878725052,
+      "learning_rate": 1.9795480880821162e-05,
+      "loss": 1.2496,
+      "step": 2667
+    },
+    {
+      "epoch": 0.9933101418249933,
+      "grad_norm": 0.17807847261428833,
+      "learning_rate": 1.9795236274449627e-05,
+      "loss": 1.2463,
+      "step": 2668
+    },
+    {
+      "epoch": 0.9936824469756024,
+      "grad_norm": 0.17453856766223907,
+      "learning_rate": 1.979499152340307e-05,
+      "loss": 1.2433,
+      "step": 2669
+    },
+    {
+      "epoch": 0.9940547521262114,
+      "grad_norm": 0.1748945713043213,
+      "learning_rate": 1.9794746627685097e-05,
+      "loss": 1.2363,
+      "step": 2670
+    },
+    {
+      "epoch": 0.9944270572768205,
+      "grad_norm": 0.1771155595779419,
+      "learning_rate": 1.9794501587299338e-05,
+      "loss": 1.236,
+      "step": 2671
+    },
+    {
+      "epoch": 0.9947993624274296,
+      "grad_norm": 0.17215952277183533,
+      "learning_rate": 1.9794256402249398e-05,
+      "loss": 1.233,
+      "step": 2672
+    },
+    {
+      "epoch": 0.9951716675780387,
+      "grad_norm": 0.1868884265422821,
+      "learning_rate": 1.979401107253891e-05,
+      "loss": 1.2446,
+      "step": 2673
+    },
+    {
+      "epoch": 0.9955439727286477,
+      "grad_norm": 0.18049085140228271,
+      "learning_rate": 1.9793765598171494e-05,
+      "loss": 1.2319,
+      "step": 2674
+    },
+    {
+      "epoch": 0.9959162778792567,
+      "grad_norm": 0.16808654367923737,
+      "learning_rate": 1.9793519979150773e-05,
+      "loss": 1.2382,
+      "step": 2675
+    },
+    {
+      "epoch": 0.9962885830298659,
+      "grad_norm": 0.18061494827270508,
+      "learning_rate": 1.9793274215480375e-05,
+      "loss": 1.2547,
+      "step": 2676
+    },
+    {
+      "epoch": 0.9966608881804749,
+      "grad_norm": 0.16883356869220734,
+      "learning_rate": 1.9793028307163937e-05,
+      "loss": 1.2517,
+      "step": 2677
+    },
+    {
+      "epoch": 0.997033193331084,
+      "grad_norm": 0.1701226681470871,
+      "learning_rate": 1.979278225420508e-05,
+      "loss": 1.2401,
+      "step": 2678
+    },
+    {
+      "epoch": 0.997405498481693,
+      "grad_norm": 0.17168228328227997,
+      "learning_rate": 1.9792536056607448e-05,
+      "loss": 1.2341,
+      "step": 2679
+    },
+    {
+      "epoch": 0.9977778036323022,
+      "grad_norm": 0.1792803853750229,
+      "learning_rate": 1.979228971437467e-05,
+      "loss": 1.2353,
+      "step": 2680
+    },
+    {
+      "epoch": 0.9981501087829112,
+      "grad_norm": 0.16425569355487823,
+      "learning_rate": 1.9792043227510387e-05,
+      "loss": 1.239,
+      "step": 2681
+    },
+    {
+      "epoch": 0.9985224139335203,
+      "grad_norm": 0.17416253685951233,
+      "learning_rate": 1.979179659601824e-05,
+      "loss": 1.2426,
+      "step": 2682
+    },
+    {
+      "epoch": 0.9988947190841293,
+      "grad_norm": 0.16776736080646515,
+      "learning_rate": 1.9791549819901875e-05,
+      "loss": 1.2377,
+      "step": 2683
+    },
+    {
+      "epoch": 0.9992670242347383,
+      "grad_norm": 0.16794565320014954,
+      "learning_rate": 1.9791302899164932e-05,
+      "loss": 1.2406,
+      "step": 2684
+    },
+    {
+      "epoch": 0.9996393293853475,
+      "grad_norm": 0.1647026389837265,
+      "learning_rate": 1.9791055833811056e-05,
+      "loss": 1.2427,
+      "step": 2685
+    },
+    {
+      "epoch": 1.0000116345359564,
+      "grad_norm": 0.19015128910541534,
+      "learning_rate": 1.9790808623843905e-05,
+      "loss": 1.2504,
+      "step": 2686
+    },
+    {
+      "epoch": 1.0003839396865657,
+      "grad_norm": 0.1643396019935608,
+      "learning_rate": 1.9790561269267122e-05,
+      "loss": 1.2387,
+      "step": 2687
+    },
+    {
+      "epoch": 1.0007562448371747,
+      "grad_norm": 0.178327739238739,
+      "learning_rate": 1.9790313770084363e-05,
+      "loss": 1.2417,
+      "step": 2688
+    },
+    {
+      "epoch": 1.0011285499877838,
+      "grad_norm": 0.18090787529945374,
+      "learning_rate": 1.9790066126299286e-05,
+      "loss": 1.2458,
+      "step": 2689
+    },
+    {
+      "epoch": 1.0015008551383928,
+      "grad_norm": 0.16783635318279266,
+      "learning_rate": 1.978981833791555e-05,
+      "loss": 1.2481,
+      "step": 2690
+    },
+    {
+      "epoch": 1.0018731602890019,
+      "grad_norm": 0.1700548678636551,
+      "learning_rate": 1.9789570404936805e-05,
+      "loss": 1.2429,
+      "step": 2691
+    },
+    {
+      "epoch": 1.002245465439611,
+      "grad_norm": 0.1691717803478241,
+      "learning_rate": 1.9789322327366722e-05,
+      "loss": 1.245,
+      "step": 2692
+    },
+    {
+      "epoch": 1.00261777059022,
+      "grad_norm": 0.16553466022014618,
+      "learning_rate": 1.9789074105208962e-05,
+      "loss": 1.2288,
+      "step": 2693
+    },
+    {
+      "epoch": 1.002990075740829,
+      "grad_norm": 0.17240577936172485,
+      "learning_rate": 1.9788825738467194e-05,
+      "loss": 1.2467,
+      "step": 2694
+    },
+    {
+      "epoch": 1.0033623808914383,
+      "grad_norm": 0.17340059578418732,
+      "learning_rate": 1.9788577227145084e-05,
+      "loss": 1.2392,
+      "step": 2695
+    },
+    {
+      "epoch": 1.0037346860420473,
+      "grad_norm": 0.1701386272907257,
+      "learning_rate": 1.97883285712463e-05,
+      "loss": 1.215,
+      "step": 2696
+    },
+    {
+      "epoch": 1.0041069911926563,
+      "grad_norm": 0.17179681360721588,
+      "learning_rate": 1.9788079770774517e-05,
+      "loss": 1.2226,
+      "step": 2697
+    },
+    {
+      "epoch": 1.0044792963432654,
+      "grad_norm": 0.1747562438249588,
+      "learning_rate": 1.9787830825733415e-05,
+      "loss": 1.2348,
+      "step": 2698
+    },
+    {
+      "epoch": 1.0048516014938744,
+      "grad_norm": 0.1735781580209732,
+      "learning_rate": 1.9787581736126663e-05,
+      "loss": 1.2438,
+      "step": 2699
+    },
+    {
+      "epoch": 1.0052239066444835,
+      "grad_norm": 0.16927596926689148,
+      "learning_rate": 1.9787332501957942e-05,
+      "loss": 1.2388,
+      "step": 2700
+    },
+    {
+      "epoch": 1.0055962117950925,
+      "grad_norm": 0.17900878190994263,
+      "learning_rate": 1.9787083123230933e-05,
+      "loss": 1.2479,
+      "step": 2701
+    },
+    {
+      "epoch": 1.0059685169457016,
+      "grad_norm": 0.1668340116739273,
+      "learning_rate": 1.9786833599949325e-05,
+      "loss": 1.2278,
+      "step": 2702
+    },
+    {
+      "epoch": 1.0063408220963106,
+      "grad_norm": 0.17158925533294678,
+      "learning_rate": 1.9786583932116795e-05,
+      "loss": 1.2364,
+      "step": 2703
+    },
+    {
+      "epoch": 1.0067131272469199,
+      "grad_norm": 0.17500467598438263,
+      "learning_rate": 1.9786334119737035e-05,
+      "loss": 1.2199,
+      "step": 2704
+    },
+    {
+      "epoch": 1.007085432397529,
+      "grad_norm": 0.1702401041984558,
+      "learning_rate": 1.9786084162813735e-05,
+      "loss": 1.252,
+      "step": 2705
+    },
+    {
+      "epoch": 1.007457737548138,
+      "grad_norm": 0.1826189160346985,
+      "learning_rate": 1.9785834061350585e-05,
+      "loss": 1.2441,
+      "step": 2706
+    },
+    {
+      "epoch": 1.007830042698747,
+      "grad_norm": 0.16807404160499573,
+      "learning_rate": 1.9785583815351285e-05,
+      "loss": 1.2245,
+      "step": 2707
+    },
+    {
+      "epoch": 1.008202347849356,
+      "grad_norm": 0.18120057880878448,
+      "learning_rate": 1.978533342481952e-05,
+      "loss": 1.2484,
+      "step": 2708
+    },
+    {
+      "epoch": 1.008574652999965,
+      "grad_norm": 0.16986186802387238,
+      "learning_rate": 1.9785082889759e-05,
+      "loss": 1.2615,
+      "step": 2709
+    },
+    {
+      "epoch": 1.0089469581505741,
+      "grad_norm": 0.18247941136360168,
+      "learning_rate": 1.9784832210173413e-05,
+      "loss": 1.2235,
+      "step": 2710
+    },
+    {
+      "epoch": 1.0093192633011832,
+      "grad_norm": 0.177805557847023,
+      "learning_rate": 1.9784581386066472e-05,
+      "loss": 1.2425,
+      "step": 2711
+    },
+    {
+      "epoch": 1.0096915684517922,
+      "grad_norm": 0.1815112829208374,
+      "learning_rate": 1.978433041744188e-05,
+      "loss": 1.2465,
+      "step": 2712
+    },
+    {
+      "epoch": 1.0100638736024015,
+      "grad_norm": 0.189849853515625,
+      "learning_rate": 1.9784079304303337e-05,
+      "loss": 1.2374,
+      "step": 2713
+    },
+    {
+      "epoch": 1.0104361787530105,
+      "grad_norm": 0.16993628442287445,
+      "learning_rate": 1.978382804665456e-05,
+      "loss": 1.2462,
+      "step": 2714
+    },
+    {
+      "epoch": 1.0108084839036195,
+      "grad_norm": 0.1785373091697693,
+      "learning_rate": 1.9783576644499257e-05,
+      "loss": 1.2423,
+      "step": 2715
+    },
+    {
+      "epoch": 1.0111807890542286,
+      "grad_norm": 0.19358791410923004,
+      "learning_rate": 1.978332509784114e-05,
+      "loss": 1.2281,
+      "step": 2716
+    },
+    {
+      "epoch": 1.0115530942048376,
+      "grad_norm": 0.16849210858345032,
+      "learning_rate": 1.9783073406683926e-05,
+      "loss": 1.2245,
+      "step": 2717
+    },
+    {
+      "epoch": 1.0119253993554467,
+      "grad_norm": 0.1804894357919693,
+      "learning_rate": 1.978282157103133e-05,
+      "loss": 1.2395,
+      "step": 2718
+    },
+    {
+      "epoch": 1.0122977045060557,
+      "grad_norm": 0.16764596104621887,
+      "learning_rate": 1.9782569590887075e-05,
+      "loss": 1.2343,
+      "step": 2719
+    },
+    {
+      "epoch": 1.0126700096566648,
+      "grad_norm": 0.17664223909378052,
+      "learning_rate": 1.978231746625488e-05,
+      "loss": 1.2432,
+      "step": 2720
+    },
+    {
+      "epoch": 1.0130423148072738,
+      "grad_norm": 0.17679232358932495,
+      "learning_rate": 1.978206519713847e-05,
+      "loss": 1.2355,
+      "step": 2721
+    },
+    {
+      "epoch": 1.013414619957883,
+      "grad_norm": 0.17845486104488373,
+      "learning_rate": 1.9781812783541574e-05,
+      "loss": 1.2529,
+      "step": 2722
+    },
+    {
+      "epoch": 1.013786925108492,
+      "grad_norm": 0.17717497050762177,
+      "learning_rate": 1.9781560225467913e-05,
+      "loss": 1.2383,
+      "step": 2723
+    },
+    {
+      "epoch": 1.0141592302591012,
+      "grad_norm": 0.17647191882133484,
+      "learning_rate": 1.9781307522921224e-05,
+      "loss": 1.2315,
+      "step": 2724
+    },
+    {
+      "epoch": 1.0145315354097102,
+      "grad_norm": 0.177278071641922,
+      "learning_rate": 1.9781054675905235e-05,
+      "loss": 1.2385,
+      "step": 2725
+    },
+    {
+      "epoch": 1.0149038405603192,
+      "grad_norm": 0.18225128948688507,
+      "learning_rate": 1.9780801684423684e-05,
+      "loss": 1.2356,
+      "step": 2726
+    },
+    {
+      "epoch": 1.0152761457109283,
+      "grad_norm": 0.1841052770614624,
+      "learning_rate": 1.9780548548480304e-05,
+      "loss": 1.2404,
+      "step": 2727
+    },
+    {
+      "epoch": 1.0156484508615373,
+      "grad_norm": 0.1847550868988037,
+      "learning_rate": 1.9780295268078834e-05,
+      "loss": 1.2343,
+      "step": 2728
+    },
+    {
+      "epoch": 1.0160207560121464,
+      "grad_norm": 0.18073725700378418,
+      "learning_rate": 1.9780041843223023e-05,
+      "loss": 1.2507,
+      "step": 2729
+    },
+    {
+      "epoch": 1.0163930611627556,
+      "grad_norm": 0.17755703628063202,
+      "learning_rate": 1.97797882739166e-05,
+      "loss": 1.2337,
+      "step": 2730
+    },
+    {
+      "epoch": 1.0167653663133647,
+      "grad_norm": 0.1812552660703659,
+      "learning_rate": 1.9779534560163324e-05,
+      "loss": 1.2212,
+      "step": 2731
+    },
+    {
+      "epoch": 1.0171376714639737,
+      "grad_norm": 0.18241752684116364,
+      "learning_rate": 1.9779280701966935e-05,
+      "loss": 1.2501,
+      "step": 2732
+    },
+    {
+      "epoch": 1.0175099766145828,
+      "grad_norm": 0.1742631196975708,
+      "learning_rate": 1.9779026699331183e-05,
+      "loss": 1.2476,
+      "step": 2733
+    },
+    {
+      "epoch": 1.0178822817651918,
+      "grad_norm": 0.17878419160842896,
+      "learning_rate": 1.9778772552259818e-05,
+      "loss": 1.2375,
+      "step": 2734
+    },
+    {
+      "epoch": 1.0182545869158008,
+      "grad_norm": 0.18265032768249512,
+      "learning_rate": 1.9778518260756602e-05,
+      "loss": 1.2443,
+      "step": 2735
+    },
+    {
+      "epoch": 1.0186268920664099,
+      "grad_norm": 0.17354628443717957,
+      "learning_rate": 1.977826382482528e-05,
+      "loss": 1.2458,
+      "step": 2736
+    },
+    {
+      "epoch": 1.018999197217019,
+      "grad_norm": 0.17573866248130798,
+      "learning_rate": 1.9778009244469617e-05,
+      "loss": 1.2469,
+      "step": 2737
+    },
+    {
+      "epoch": 1.019371502367628,
+      "grad_norm": 0.18138283491134644,
+      "learning_rate": 1.977775451969337e-05,
+      "loss": 1.2296,
+      "step": 2738
+    },
+    {
+      "epoch": 1.0197438075182372,
+      "grad_norm": 0.18564942479133606,
+      "learning_rate": 1.9777499650500303e-05,
+      "loss": 1.2492,
+      "step": 2739
+    },
+    {
+      "epoch": 1.0201161126688463,
+      "grad_norm": 0.18133315443992615,
+      "learning_rate": 1.977724463689418e-05,
+      "loss": 1.2432,
+      "step": 2740
+    },
+    {
+      "epoch": 1.0204884178194553,
+      "grad_norm": 0.18659467995166779,
+      "learning_rate": 1.9776989478878764e-05,
+      "loss": 1.258,
+      "step": 2741
+    },
+    {
+      "epoch": 1.0208607229700644,
+      "grad_norm": 0.2271583080291748,
+      "learning_rate": 1.9776734176457833e-05,
+      "loss": 1.236,
+      "step": 2742
+    },
+    {
+      "epoch": 1.0212330281206734,
+      "grad_norm": 0.1641337126493454,
+      "learning_rate": 1.9776478729635146e-05,
+      "loss": 1.2299,
+      "step": 2743
+    },
+    {
+      "epoch": 1.0216053332712824,
+      "grad_norm": 0.18119634687900543,
+      "learning_rate": 1.9776223138414486e-05,
+      "loss": 1.2252,
+      "step": 2744
+    },
+    {
+      "epoch": 1.0219776384218915,
+      "grad_norm": 0.17801545560359955,
+      "learning_rate": 1.977596740279962e-05,
+      "loss": 1.2272,
+      "step": 2745
+    },
+    {
+      "epoch": 1.0223499435725005,
+      "grad_norm": 0.18631921708583832,
+      "learning_rate": 1.9775711522794333e-05,
+      "loss": 1.252,
+      "step": 2746
+    },
+    {
+      "epoch": 1.0227222487231096,
+      "grad_norm": 0.18302686512470245,
+      "learning_rate": 1.97754554984024e-05,
+      "loss": 1.2308,
+      "step": 2747
+    },
+    {
+      "epoch": 1.0230945538737188,
+      "grad_norm": 0.178566575050354,
+      "learning_rate": 1.97751993296276e-05,
+      "loss": 1.2492,
+      "step": 2748
+    },
+    {
+      "epoch": 1.0234668590243279,
+      "grad_norm": 0.18399080634117126,
+      "learning_rate": 1.977494301647372e-05,
+      "loss": 1.2451,
+      "step": 2749
+    },
+    {
+      "epoch": 1.023839164174937,
+      "grad_norm": 0.18315501511096954,
+      "learning_rate": 1.9774686558944544e-05,
+      "loss": 1.2505,
+      "step": 2750
+    },
+    {
+      "epoch": 1.024211469325546,
+      "grad_norm": 0.184551402926445,
+      "learning_rate": 1.9774429957043866e-05,
+      "loss": 1.2294,
+      "step": 2751
+    },
+    {
+      "epoch": 1.024583774476155,
+      "grad_norm": 0.45721930265426636,
+      "learning_rate": 1.9774173210775466e-05,
+      "loss": 1.2194,
+      "step": 2752
+    },
+    {
+      "epoch": 1.024956079626764,
+      "grad_norm": 0.1807864010334015,
+      "learning_rate": 1.9773916320143144e-05,
+      "loss": 1.249,
+      "step": 2753
+    },
+    {
+      "epoch": 1.025328384777373,
+      "grad_norm": 0.18623942136764526,
+      "learning_rate": 1.977365928515069e-05,
+      "loss": 1.2281,
+      "step": 2754
+    },
+    {
+      "epoch": 1.0257006899279821,
+      "grad_norm": 0.18158674240112305,
+      "learning_rate": 1.97734021058019e-05,
+      "loss": 1.2333,
+      "step": 2755
+    },
+    {
+      "epoch": 1.0260729950785914,
+      "grad_norm": 0.1756281554698944,
+      "learning_rate": 1.9773144782100576e-05,
+      "loss": 1.2503,
+      "step": 2756
+    },
+    {
+      "epoch": 1.0264453002292004,
+      "grad_norm": 0.1774439960718155,
+      "learning_rate": 1.9772887314050516e-05,
+      "loss": 1.2371,
+      "step": 2757
+    },
+    {
+      "epoch": 1.0268176053798095,
+      "grad_norm": 0.17435887455940247,
+      "learning_rate": 1.9772629701655524e-05,
+      "loss": 1.2379,
+      "step": 2758
+    },
+    {
+      "epoch": 1.0271899105304185,
+      "grad_norm": 0.18114317953586578,
+      "learning_rate": 1.9772371944919406e-05,
+      "loss": 1.2389,
+      "step": 2759
+    },
+    {
+      "epoch": 1.0275622156810276,
+      "grad_norm": 0.1751965880393982,
+      "learning_rate": 1.9772114043845968e-05,
+      "loss": 1.2537,
+      "step": 2760
+    },
+    {
+      "epoch": 1.0279345208316366,
+      "grad_norm": 0.18120123445987701,
+      "learning_rate": 1.977185599843902e-05,
+      "loss": 1.2308,
+      "step": 2761
+    },
+    {
+      "epoch": 1.0283068259822457,
+      "grad_norm": 0.18496285378932953,
+      "learning_rate": 1.9771597808702366e-05,
+      "loss": 1.2399,
+      "step": 2762
+    },
+    {
+      "epoch": 1.0286791311328547,
+      "grad_norm": 0.17509673535823822,
+      "learning_rate": 1.9771339474639833e-05,
+      "loss": 1.2205,
+      "step": 2763
+    },
+    {
+      "epoch": 1.0290514362834637,
+      "grad_norm": 0.17925499379634857,
+      "learning_rate": 1.9771080996255226e-05,
+      "loss": 1.2414,
+      "step": 2764
+    },
+    {
+      "epoch": 1.029423741434073,
+      "grad_norm": 0.17115138471126556,
+      "learning_rate": 1.9770822373552362e-05,
+      "loss": 1.2282,
+      "step": 2765
+    },
+    {
+      "epoch": 1.029796046584682,
+      "grad_norm": 0.1737300306558609,
+      "learning_rate": 1.9770563606535068e-05,
+      "loss": 1.2417,
+      "step": 2766
+    },
+    {
+      "epoch": 1.030168351735291,
+      "grad_norm": 0.17681419849395752,
+      "learning_rate": 1.9770304695207164e-05,
+      "loss": 1.2274,
+      "step": 2767
+    },
+    {
+      "epoch": 1.0305406568859001,
+      "grad_norm": 0.1788710057735443,
+      "learning_rate": 1.9770045639572473e-05,
+      "loss": 1.2319,
+      "step": 2768
+    },
+    {
+      "epoch": 1.0309129620365092,
+      "grad_norm": 0.17474333941936493,
+      "learning_rate": 1.976978643963482e-05,
+      "loss": 1.2386,
+      "step": 2769
+    },
+    {
+      "epoch": 1.0312852671871182,
+      "grad_norm": 0.18028199672698975,
+      "learning_rate": 1.9769527095398033e-05,
+      "loss": 1.2426,
+      "step": 2770
+    },
+    {
+      "epoch": 1.0316575723377273,
+      "grad_norm": 0.17821404337882996,
+      "learning_rate": 1.9769267606865944e-05,
+      "loss": 1.256,
+      "step": 2771
+    },
+    {
+      "epoch": 1.0320298774883363,
+      "grad_norm": 0.17664535343647003,
+      "learning_rate": 1.976900797404239e-05,
+      "loss": 1.2397,
+      "step": 2772
+    },
+    {
+      "epoch": 1.0324021826389453,
+      "grad_norm": 0.17644058167934418,
+      "learning_rate": 1.9768748196931197e-05,
+      "loss": 1.259,
+      "step": 2773
+    },
+    {
+      "epoch": 1.0327744877895546,
+      "grad_norm": 0.18605953454971313,
+      "learning_rate": 1.976848827553621e-05,
+      "loss": 1.2336,
+      "step": 2774
+    },
+    {
+      "epoch": 1.0331467929401636,
+      "grad_norm": 0.17775902152061462,
+      "learning_rate": 1.9768228209861257e-05,
+      "loss": 1.2309,
+      "step": 2775
+    },
+    {
+      "epoch": 1.0335190980907727,
+      "grad_norm": 0.18057388067245483,
+      "learning_rate": 1.976796799991019e-05,
+      "loss": 1.231,
+      "step": 2776
+    },
+    {
+      "epoch": 1.0338914032413817,
+      "grad_norm": 0.17607726156711578,
+      "learning_rate": 1.9767707645686852e-05,
+      "loss": 1.2189,
+      "step": 2777
+    },
+    {
+      "epoch": 1.0342637083919908,
+      "grad_norm": 0.19016438722610474,
+      "learning_rate": 1.9767447147195083e-05,
+      "loss": 1.2416,
+      "step": 2778
+    },
+    {
+      "epoch": 1.0346360135425998,
+      "grad_norm": 0.1789761781692505,
+      "learning_rate": 1.976718650443873e-05,
+      "loss": 1.2498,
+      "step": 2779
+    },
+    {
+      "epoch": 1.0350083186932089,
+      "grad_norm": 0.1939614713191986,
+      "learning_rate": 1.9766925717421647e-05,
+      "loss": 1.2518,
+      "step": 2780
+    },
+    {
+      "epoch": 1.035380623843818,
+      "grad_norm": 0.1895962953567505,
+      "learning_rate": 1.9766664786147687e-05,
+      "loss": 1.2613,
+      "step": 2781
+    },
+    {
+      "epoch": 1.0357529289944272,
+      "grad_norm": 0.21478445827960968,
+      "learning_rate": 1.97664037106207e-05,
+      "loss": 1.235,
+      "step": 2782
+    },
+    {
+      "epoch": 1.0361252341450362,
+      "grad_norm": 0.17898200452327728,
+      "learning_rate": 1.976614249084454e-05,
+      "loss": 1.2397,
+      "step": 2783
+    },
+    {
+      "epoch": 1.0364975392956453,
+      "grad_norm": 0.18122392892837524,
+      "learning_rate": 1.976588112682307e-05,
+      "loss": 1.2253,
+      "step": 2784
+    },
+    {
+      "epoch": 1.0368698444462543,
+      "grad_norm": 0.17120665311813354,
+      "learning_rate": 1.9765619618560146e-05,
+      "loss": 1.2428,
+      "step": 2785
+    },
+    {
+      "epoch": 1.0372421495968633,
+      "grad_norm": 0.1834287941455841,
+      "learning_rate": 1.9765357966059638e-05,
+      "loss": 1.2253,
+      "step": 2786
+    },
+    {
+      "epoch": 1.0376144547474724,
+      "grad_norm": 0.1814526915550232,
+      "learning_rate": 1.9765096169325404e-05,
+      "loss": 1.242,
+      "step": 2787
+    },
+    {
+      "epoch": 1.0379867598980814,
+      "grad_norm": 0.18494722247123718,
+      "learning_rate": 1.9764834228361313e-05,
+      "loss": 1.2501,
+      "step": 2788
+    },
+    {
+      "epoch": 1.0383590650486905,
+      "grad_norm": 0.17375454306602478,
+      "learning_rate": 1.9764572143171232e-05,
+      "loss": 1.2292,
+      "step": 2789
+    },
+    {
+      "epoch": 1.0387313701992995,
+      "grad_norm": 0.18002989888191223,
+      "learning_rate": 1.9764309913759033e-05,
+      "loss": 1.2331,
+      "step": 2790
+    },
+    {
+      "epoch": 1.0391036753499088,
+      "grad_norm": 0.18305903673171997,
+      "learning_rate": 1.976404754012859e-05,
+      "loss": 1.2479,
+      "step": 2791
+    },
+    {
+      "epoch": 1.0394759805005178,
+      "grad_norm": 0.18125832080841064,
+      "learning_rate": 1.9763785022283778e-05,
+      "loss": 1.2238,
+      "step": 2792
+    },
+    {
+      "epoch": 1.0398482856511269,
+      "grad_norm": 0.1753983050584793,
+      "learning_rate": 1.976352236022847e-05,
+      "loss": 1.2198,
+      "step": 2793
+    },
+    {
+      "epoch": 1.040220590801736,
+      "grad_norm": 0.17953644692897797,
+      "learning_rate": 1.9763259553966553e-05,
+      "loss": 1.2477,
+      "step": 2794
+    },
+    {
+      "epoch": 1.040592895952345,
+      "grad_norm": 0.17996378242969513,
+      "learning_rate": 1.9762996603501908e-05,
+      "loss": 1.2696,
+      "step": 2795
+    },
+    {
+      "epoch": 1.040965201102954,
+      "grad_norm": 0.17698989808559418,
+      "learning_rate": 1.976273350883841e-05,
+      "loss": 1.248,
+      "step": 2796
+    },
+    {
+      "epoch": 1.041337506253563,
+      "grad_norm": 0.17306332290172577,
+      "learning_rate": 1.9762470269979955e-05,
+      "loss": 1.2555,
+      "step": 2797
+    },
+    {
+      "epoch": 1.041709811404172,
+      "grad_norm": 0.17857332527637482,
+      "learning_rate": 1.9762206886930423e-05,
+      "loss": 1.2386,
+      "step": 2798
+    },
+    {
+      "epoch": 1.0420821165547811,
+      "grad_norm": 0.18211491405963898,
+      "learning_rate": 1.9761943359693712e-05,
+      "loss": 1.2458,
+      "step": 2799
+    },
+    {
+      "epoch": 1.0424544217053904,
+      "grad_norm": 0.1814602017402649,
+      "learning_rate": 1.9761679688273708e-05,
+      "loss": 1.2442,
+      "step": 2800
+    },
+    {
+      "epoch": 1.0428267268559994,
+      "grad_norm": 0.1816301792860031,
+      "learning_rate": 1.976141587267431e-05,
+      "loss": 1.2377,
+      "step": 2801
+    },
+    {
+      "epoch": 1.0431990320066085,
+      "grad_norm": 0.1897687017917633,
+      "learning_rate": 1.976115191289941e-05,
+      "loss": 1.2532,
+      "step": 2802
+    },
+    {
+      "epoch": 1.0435713371572175,
+      "grad_norm": 0.174024760723114,
+      "learning_rate": 1.9760887808952908e-05,
+      "loss": 1.223,
+      "step": 2803
+    },
+    {
+      "epoch": 1.0439436423078265,
+      "grad_norm": 0.19265376031398773,
+      "learning_rate": 1.9760623560838707e-05,
+      "loss": 1.2442,
+      "step": 2804
+    },
+    {
+      "epoch": 1.0443159474584356,
+      "grad_norm": 0.17987599968910217,
+      "learning_rate": 1.9760359168560708e-05,
+      "loss": 1.2395,
+      "step": 2805
+    },
+    {
+      "epoch": 1.0446882526090446,
+      "grad_norm": 0.1701204478740692,
+      "learning_rate": 1.976009463212282e-05,
+      "loss": 1.2306,
+      "step": 2806
+    },
+    {
+      "epoch": 1.0450605577596537,
+      "grad_norm": 0.1814124584197998,
+      "learning_rate": 1.975982995152894e-05,
+      "loss": 1.2451,
+      "step": 2807
+    },
+    {
+      "epoch": 1.0454328629102627,
+      "grad_norm": 0.181508406996727,
+      "learning_rate": 1.9759565126782988e-05,
+      "loss": 1.2415,
+      "step": 2808
+    },
+    {
+      "epoch": 1.045805168060872,
+      "grad_norm": 0.167762890458107,
+      "learning_rate": 1.975930015788887e-05,
+      "loss": 1.2343,
+      "step": 2809
+    },
+    {
+      "epoch": 1.046177473211481,
+      "grad_norm": 0.17831118404865265,
+      "learning_rate": 1.9759035044850504e-05,
+      "loss": 1.2163,
+      "step": 2810
+    },
+    {
+      "epoch": 1.04654977836209,
+      "grad_norm": 0.18151675164699554,
+      "learning_rate": 1.9758769787671804e-05,
+      "loss": 1.2255,
+      "step": 2811
+    },
+    {
+      "epoch": 1.046922083512699,
+      "grad_norm": 0.18939217925071716,
+      "learning_rate": 1.9758504386356682e-05,
+      "loss": 1.2381,
+      "step": 2812
+    },
+    {
+      "epoch": 1.0472943886633082,
+      "grad_norm": 0.1697840541601181,
+      "learning_rate": 1.9758238840909066e-05,
+      "loss": 1.2406,
+      "step": 2813
+    },
+    {
+      "epoch": 1.0476666938139172,
+      "grad_norm": 0.17666485905647278,
+      "learning_rate": 1.975797315133287e-05,
+      "loss": 1.2312,
+      "step": 2814
+    },
+    {
+      "epoch": 1.0480389989645262,
+      "grad_norm": 0.1959092617034912,
+      "learning_rate": 1.975770731763203e-05,
+      "loss": 1.2444,
+      "step": 2815
+    },
+    {
+      "epoch": 1.0484113041151353,
+      "grad_norm": 0.17238794267177582,
+      "learning_rate": 1.9757441339810462e-05,
+      "loss": 1.227,
+      "step": 2816
+    },
+    {
+      "epoch": 1.0487836092657445,
+      "grad_norm": 0.17401093244552612,
+      "learning_rate": 1.9757175217872096e-05,
+      "loss": 1.2475,
+      "step": 2817
+    },
+    {
+      "epoch": 1.0491559144163536,
+      "grad_norm": 0.1765962839126587,
+      "learning_rate": 1.9756908951820866e-05,
+      "loss": 1.2348,
+      "step": 2818
+    },
+    {
+      "epoch": 1.0495282195669626,
+      "grad_norm": 0.17260397970676422,
+      "learning_rate": 1.9756642541660702e-05,
+      "loss": 1.2509,
+      "step": 2819
+    },
+    {
+      "epoch": 1.0499005247175717,
+      "grad_norm": 0.16606584191322327,
+      "learning_rate": 1.975637598739554e-05,
+      "loss": 1.2384,
+      "step": 2820
+    },
+    {
+      "epoch": 1.0502728298681807,
+      "grad_norm": 0.17037144303321838,
+      "learning_rate": 1.975610928902932e-05,
+      "loss": 1.2382,
+      "step": 2821
+    },
+    {
+      "epoch": 1.0506451350187898,
+      "grad_norm": 0.1816329061985016,
+      "learning_rate": 1.9755842446565975e-05,
+      "loss": 1.2329,
+      "step": 2822
+    },
+    {
+      "epoch": 1.0510174401693988,
+      "grad_norm": 0.18435317277908325,
+      "learning_rate": 1.975557546000945e-05,
+      "loss": 1.233,
+      "step": 2823
+    },
+    {
+      "epoch": 1.0513897453200078,
+      "grad_norm": 0.16666759550571442,
+      "learning_rate": 1.975530832936369e-05,
+      "loss": 1.2345,
+      "step": 2824
+    },
+    {
+      "epoch": 1.0517620504706169,
+      "grad_norm": 0.18911749124526978,
+      "learning_rate": 1.9755041054632634e-05,
+      "loss": 1.2416,
+      "step": 2825
+    },
+    {
+      "epoch": 1.0521343556212261,
+      "grad_norm": 0.17855370044708252,
+      "learning_rate": 1.9754773635820236e-05,
+      "loss": 1.246,
+      "step": 2826
+    },
+    {
+      "epoch": 1.0525066607718352,
+      "grad_norm": 0.17206569015979767,
+      "learning_rate": 1.9754506072930443e-05,
+      "loss": 1.2326,
+      "step": 2827
+    },
+    {
+      "epoch": 1.0528789659224442,
+      "grad_norm": 0.180022731423378,
+      "learning_rate": 1.9754238365967207e-05,
+      "loss": 1.2389,
+      "step": 2828
+    },
+    {
+      "epoch": 1.0532512710730533,
+      "grad_norm": 0.17861510813236237,
+      "learning_rate": 1.9753970514934485e-05,
+      "loss": 1.2355,
+      "step": 2829
+    },
+    {
+      "epoch": 1.0536235762236623,
+      "grad_norm": 0.17671099305152893,
+      "learning_rate": 1.9753702519836228e-05,
+      "loss": 1.2487,
+      "step": 2830
+    },
+    {
+      "epoch": 1.0539958813742714,
+      "grad_norm": 0.17241548001766205,
+      "learning_rate": 1.97534343806764e-05,
+      "loss": 1.2386,
+      "step": 2831
+    },
+    {
+      "epoch": 1.0543681865248804,
+      "grad_norm": 0.1827198565006256,
+      "learning_rate": 1.9753166097458957e-05,
+      "loss": 1.2326,
+      "step": 2832
+    },
+    {
+      "epoch": 1.0547404916754894,
+      "grad_norm": 0.17715848982334137,
+      "learning_rate": 1.975289767018786e-05,
+      "loss": 1.2346,
+      "step": 2833
+    },
+    {
+      "epoch": 1.0551127968260985,
+      "grad_norm": 0.17723973095417023,
+      "learning_rate": 1.975262909886708e-05,
+      "loss": 1.2426,
+      "step": 2834
+    },
+    {
+      "epoch": 1.0554851019767078,
+      "grad_norm": 0.1761004477739334,
+      "learning_rate": 1.975236038350058e-05,
+      "loss": 1.2411,
+      "step": 2835
+    },
+    {
+      "epoch": 1.0558574071273168,
+      "grad_norm": 0.18172328174114227,
+      "learning_rate": 1.9752091524092324e-05,
+      "loss": 1.2237,
+      "step": 2836
+    },
+    {
+      "epoch": 1.0562297122779258,
+      "grad_norm": 0.16737525165081024,
+      "learning_rate": 1.9751822520646297e-05,
+      "loss": 1.2503,
+      "step": 2837
+    },
+    {
+      "epoch": 1.0566020174285349,
+      "grad_norm": 0.1774168312549591,
+      "learning_rate": 1.9751553373166454e-05,
+      "loss": 1.2195,
+      "step": 2838
+    },
+    {
+      "epoch": 1.056974322579144,
+      "grad_norm": 0.17864681780338287,
+      "learning_rate": 1.9751284081656786e-05,
+      "loss": 1.2407,
+      "step": 2839
+    },
+    {
+      "epoch": 1.057346627729753,
+      "grad_norm": 0.1730041354894638,
+      "learning_rate": 1.975101464612126e-05,
+      "loss": 1.2441,
+      "step": 2840
+    },
+    {
+      "epoch": 1.057718932880362,
+      "grad_norm": 0.17690402269363403,
+      "learning_rate": 1.9750745066563864e-05,
+      "loss": 1.2445,
+      "step": 2841
+    },
+    {
+      "epoch": 1.058091238030971,
+      "grad_norm": 0.1674228459596634,
+      "learning_rate": 1.9750475342988572e-05,
+      "loss": 1.2313,
+      "step": 2842
+    },
+    {
+      "epoch": 1.0584635431815803,
+      "grad_norm": 0.1728244572877884,
+      "learning_rate": 1.9750205475399373e-05,
+      "loss": 1.2261,
+      "step": 2843
+    },
+    {
+      "epoch": 1.0588358483321894,
+      "grad_norm": 0.1766444444656372,
+      "learning_rate": 1.974993546380025e-05,
+      "loss": 1.2329,
+      "step": 2844
+    },
+    {
+      "epoch": 1.0592081534827984,
+      "grad_norm": 0.17906500399112701,
+      "learning_rate": 1.9749665308195194e-05,
+      "loss": 1.237,
+      "step": 2845
+    },
+    {
+      "epoch": 1.0595804586334074,
+      "grad_norm": 0.17800427973270416,
+      "learning_rate": 1.974939500858819e-05,
+      "loss": 1.2405,
+      "step": 2846
+    },
+    {
+      "epoch": 1.0599527637840165,
+      "grad_norm": 0.17567503452301025,
+      "learning_rate": 1.9749124564983237e-05,
+      "loss": 1.241,
+      "step": 2847
+    },
+    {
+      "epoch": 1.0603250689346255,
+      "grad_norm": 0.172315314412117,
+      "learning_rate": 1.9748853977384326e-05,
+      "loss": 1.2331,
+      "step": 2848
+    },
+    {
+      "epoch": 1.0606973740852346,
+      "grad_norm": 0.17983998358249664,
+      "learning_rate": 1.974858324579545e-05,
+      "loss": 1.2449,
+      "step": 2849
+    },
+    {
+      "epoch": 1.0610696792358436,
+      "grad_norm": 0.17935794591903687,
+      "learning_rate": 1.9748312370220613e-05,
+      "loss": 1.2357,
+      "step": 2850
+    },
+    {
+      "epoch": 1.0614419843864527,
+      "grad_norm": 0.1763681173324585,
+      "learning_rate": 1.9748041350663817e-05,
+      "loss": 1.2498,
+      "step": 2851
+    },
+    {
+      "epoch": 1.061814289537062,
+      "grad_norm": 0.17987823486328125,
+      "learning_rate": 1.9747770187129055e-05,
+      "loss": 1.2495,
+      "step": 2852
+    },
+    {
+      "epoch": 1.062186594687671,
+      "grad_norm": 0.17571550607681274,
+      "learning_rate": 1.9747498879620342e-05,
+      "loss": 1.2263,
+      "step": 2853
+    },
+    {
+      "epoch": 1.06255889983828,
+      "grad_norm": 0.17935210466384888,
+      "learning_rate": 1.9747227428141687e-05,
+      "loss": 1.2317,
+      "step": 2854
+    },
+    {
+      "epoch": 1.062931204988889,
+      "grad_norm": 0.16637402772903442,
+      "learning_rate": 1.9746955832697094e-05,
+      "loss": 1.2329,
+      "step": 2855
+    },
+    {
+      "epoch": 1.063303510139498,
+      "grad_norm": 0.17255236208438873,
+      "learning_rate": 1.974668409329057e-05,
+      "loss": 1.2353,
+      "step": 2856
+    },
+    {
+      "epoch": 1.0636758152901071,
+      "grad_norm": 0.17514565587043762,
+      "learning_rate": 1.9746412209926132e-05,
+      "loss": 1.2303,
+      "step": 2857
+    },
+    {
+      "epoch": 1.0640481204407162,
+      "grad_norm": 0.17487487196922302,
+      "learning_rate": 1.9746140182607805e-05,
+      "loss": 1.2199,
+      "step": 2858
+    },
+    {
+      "epoch": 1.0644204255913252,
+      "grad_norm": 0.16520456969738007,
+      "learning_rate": 1.9745868011339592e-05,
+      "loss": 1.2332,
+      "step": 2859
+    },
+    {
+      "epoch": 1.0647927307419343,
+      "grad_norm": 0.1774507462978363,
+      "learning_rate": 1.9745595696125526e-05,
+      "loss": 1.2407,
+      "step": 2860
+    },
+    {
+      "epoch": 1.0651650358925435,
+      "grad_norm": 0.16631199419498444,
+      "learning_rate": 1.974532323696962e-05,
+      "loss": 1.2345,
+      "step": 2861
+    },
+    {
+      "epoch": 1.0655373410431526,
+      "grad_norm": 0.17234042286872864,
+      "learning_rate": 1.97450506338759e-05,
+      "loss": 1.2222,
+      "step": 2862
+    },
+    {
+      "epoch": 1.0659096461937616,
+      "grad_norm": 0.18009594082832336,
+      "learning_rate": 1.97447778868484e-05,
+      "loss": 1.2411,
+      "step": 2863
+    },
+    {
+      "epoch": 1.0662819513443706,
+      "grad_norm": 0.1718275547027588,
+      "learning_rate": 1.9744504995891135e-05,
+      "loss": 1.2352,
+      "step": 2864
+    },
+    {
+      "epoch": 1.0666542564949797,
+      "grad_norm": 0.16826453804969788,
+      "learning_rate": 1.9744231961008147e-05,
+      "loss": 1.2314,
+      "step": 2865
+    },
+    {
+      "epoch": 1.0670265616455887,
+      "grad_norm": 0.17893622815608978,
+      "learning_rate": 1.9743958782203462e-05,
+      "loss": 1.2282,
+      "step": 2866
+    },
+    {
+      "epoch": 1.0673988667961978,
+      "grad_norm": 0.17840996384620667,
+      "learning_rate": 1.9743685459481118e-05,
+      "loss": 1.2473,
+      "step": 2867
+    },
+    {
+      "epoch": 1.0677711719468068,
+      "grad_norm": 0.1669299453496933,
+      "learning_rate": 1.974341199284515e-05,
+      "loss": 1.2345,
+      "step": 2868
+    },
+    {
+      "epoch": 1.0681434770974159,
+      "grad_norm": 0.1682414710521698,
+      "learning_rate": 1.97431383822996e-05,
+      "loss": 1.2212,
+      "step": 2869
+    },
+    {
+      "epoch": 1.0685157822480251,
+      "grad_norm": 0.17519889771938324,
+      "learning_rate": 1.974286462784851e-05,
+      "loss": 1.2377,
+      "step": 2870
+    },
+    {
+      "epoch": 1.0688880873986342,
+      "grad_norm": 0.16797272861003876,
+      "learning_rate": 1.9742590729495917e-05,
+      "loss": 1.2078,
+      "step": 2871
+    },
+    {
+      "epoch": 1.0692603925492432,
+      "grad_norm": 0.17429950833320618,
+      "learning_rate": 1.9742316687245873e-05,
+      "loss": 1.2334,
+      "step": 2872
+    },
+    {
+      "epoch": 1.0696326976998523,
+      "grad_norm": 0.1742522120475769,
+      "learning_rate": 1.974204250110242e-05,
+      "loss": 1.2474,
+      "step": 2873
+    },
+    {
+      "epoch": 1.0700050028504613,
+      "grad_norm": 0.17582035064697266,
+      "learning_rate": 1.9741768171069614e-05,
+      "loss": 1.241,
+      "step": 2874
+    },
+    {
+      "epoch": 1.0703773080010703,
+      "grad_norm": 0.17079758644104004,
+      "learning_rate": 1.9741493697151502e-05,
+      "loss": 1.2281,
+      "step": 2875
+    },
+    {
+      "epoch": 1.0707496131516794,
+      "grad_norm": 0.17140960693359375,
+      "learning_rate": 1.9741219079352142e-05,
+      "loss": 1.2393,
+      "step": 2876
+    },
+    {
+      "epoch": 1.0711219183022884,
+      "grad_norm": 0.1833873987197876,
+      "learning_rate": 1.9740944317675583e-05,
+      "loss": 1.2488,
+      "step": 2877
+    },
+    {
+      "epoch": 1.0714942234528975,
+      "grad_norm": 0.22053757309913635,
+      "learning_rate": 1.974066941212589e-05,
+      "loss": 1.2451,
+      "step": 2878
+    },
+    {
+      "epoch": 1.0718665286035067,
+      "grad_norm": 0.1896497905254364,
+      "learning_rate": 1.974039436270712e-05,
+      "loss": 1.2203,
+      "step": 2879
+    },
+    {
+      "epoch": 1.0722388337541158,
+      "grad_norm": 0.17092444002628326,
+      "learning_rate": 1.9740119169423337e-05,
+      "loss": 1.234,
+      "step": 2880
+    },
+    {
+      "epoch": 1.0726111389047248,
+      "grad_norm": 0.1865914762020111,
+      "learning_rate": 1.9739843832278604e-05,
+      "loss": 1.2252,
+      "step": 2881
+    },
+    {
+      "epoch": 1.0729834440553339,
+      "grad_norm": 0.1847282201051712,
+      "learning_rate": 1.973956835127699e-05,
+      "loss": 1.2307,
+      "step": 2882
+    },
+    {
+      "epoch": 1.073355749205943,
+      "grad_norm": 0.16904793679714203,
+      "learning_rate": 1.9739292726422565e-05,
+      "loss": 1.2399,
+      "step": 2883
+    },
+    {
+      "epoch": 1.073728054356552,
+      "grad_norm": 0.17725877463817596,
+      "learning_rate": 1.9739016957719393e-05,
+      "loss": 1.2425,
+      "step": 2884
+    },
+    {
+      "epoch": 1.074100359507161,
+      "grad_norm": 0.1759733408689499,
+      "learning_rate": 1.9738741045171556e-05,
+      "loss": 1.2151,
+      "step": 2885
+    },
+    {
+      "epoch": 1.07447266465777,
+      "grad_norm": 0.17010413110256195,
+      "learning_rate": 1.973846498878312e-05,
+      "loss": 1.2503,
+      "step": 2886
+    },
+    {
+      "epoch": 1.0748449698083793,
+      "grad_norm": 0.17372602224349976,
+      "learning_rate": 1.9738188788558174e-05,
+      "loss": 1.2469,
+      "step": 2887
+    },
+    {
+      "epoch": 1.0752172749589883,
+      "grad_norm": 0.17600518465042114,
+      "learning_rate": 1.9737912444500786e-05,
+      "loss": 1.2301,
+      "step": 2888
+    },
+    {
+      "epoch": 1.0755895801095974,
+      "grad_norm": 0.17310109734535217,
+      "learning_rate": 1.9737635956615044e-05,
+      "loss": 1.2356,
+      "step": 2889
+    },
+    {
+      "epoch": 1.0759618852602064,
+      "grad_norm": 0.1836235374212265,
+      "learning_rate": 1.973735932490503e-05,
+      "loss": 1.2501,
+      "step": 2890
+    },
+    {
+      "epoch": 1.0763341904108155,
+      "grad_norm": 0.16939985752105713,
+      "learning_rate": 1.9737082549374828e-05,
+      "loss": 1.2273,
+      "step": 2891
+    },
+    {
+      "epoch": 1.0767064955614245,
+      "grad_norm": 0.1679062396287918,
+      "learning_rate": 1.973680563002853e-05,
+      "loss": 1.2343,
+      "step": 2892
+    },
+    {
+      "epoch": 1.0770788007120335,
+      "grad_norm": 0.1816723644733429,
+      "learning_rate": 1.9736528566870223e-05,
+      "loss": 1.2332,
+      "step": 2893
+    },
+    {
+      "epoch": 1.0774511058626426,
+      "grad_norm": 0.17142869532108307,
+      "learning_rate": 1.9736251359904003e-05,
+      "loss": 1.2289,
+      "step": 2894
+    },
+    {
+      "epoch": 1.0778234110132519,
+      "grad_norm": 0.17428475618362427,
+      "learning_rate": 1.9735974009133957e-05,
+      "loss": 1.2244,
+      "step": 2895
+    },
+    {
+      "epoch": 1.078195716163861,
+      "grad_norm": 0.16701877117156982,
+      "learning_rate": 1.9735696514564188e-05,
+      "loss": 1.211,
+      "step": 2896
+    },
+    {
+      "epoch": 1.07856802131447,
+      "grad_norm": 0.17782089114189148,
+      "learning_rate": 1.9735418876198792e-05,
+      "loss": 1.2257,
+      "step": 2897
+    },
+    {
+      "epoch": 1.078940326465079,
+      "grad_norm": 0.1718931496143341,
+      "learning_rate": 1.973514109404187e-05,
+      "loss": 1.2393,
+      "step": 2898
+    },
+    {
+      "epoch": 1.079312631615688,
+      "grad_norm": 0.16768048703670502,
+      "learning_rate": 1.9734863168097526e-05,
+      "loss": 1.2331,
+      "step": 2899
+    },
+    {
+      "epoch": 1.079684936766297,
+      "grad_norm": 0.16998796164989471,
+      "learning_rate": 1.973458509836986e-05,
+      "loss": 1.226,
+      "step": 2900
+    },
+    {
+      "epoch": 1.080057241916906,
+      "grad_norm": 0.1749204695224762,
+      "learning_rate": 1.973430688486299e-05,
+      "loss": 1.2437,
+      "step": 2901
+    },
+    {
+      "epoch": 1.0804295470675152,
+      "grad_norm": 0.17514778673648834,
+      "learning_rate": 1.973402852758101e-05,
+      "loss": 1.2397,
+      "step": 2902
+    },
+    {
+      "epoch": 1.0808018522181242,
+      "grad_norm": 0.1695886254310608,
+      "learning_rate": 1.9733750026528046e-05,
+      "loss": 1.2261,
+      "step": 2903
+    },
+    {
+      "epoch": 1.0811741573687335,
+      "grad_norm": 0.16834037005901337,
+      "learning_rate": 1.9733471381708202e-05,
+      "loss": 1.2264,
+      "step": 2904
+    },
+    {
+      "epoch": 1.0815464625193425,
+      "grad_norm": 0.17187410593032837,
+      "learning_rate": 1.97331925931256e-05,
+      "loss": 1.229,
+      "step": 2905
+    },
+    {
+      "epoch": 1.0819187676699515,
+      "grad_norm": 0.17803004384040833,
+      "learning_rate": 1.973291366078435e-05,
+      "loss": 1.237,
+      "step": 2906
+    },
+    {
+      "epoch": 1.0822910728205606,
+      "grad_norm": 0.1769217699766159,
+      "learning_rate": 1.973263458468858e-05,
+      "loss": 1.2232,
+      "step": 2907
+    },
+    {
+      "epoch": 1.0826633779711696,
+      "grad_norm": 0.1921815276145935,
+      "learning_rate": 1.9732355364842404e-05,
+      "loss": 1.2533,
+      "step": 2908
+    },
+    {
+      "epoch": 1.0830356831217787,
+      "grad_norm": 0.17243653535842896,
+      "learning_rate": 1.973207600124995e-05,
+      "loss": 1.2333,
+      "step": 2909
+    },
+    {
+      "epoch": 1.0834079882723877,
+      "grad_norm": 0.1771405041217804,
+      "learning_rate": 1.9731796493915346e-05,
+      "loss": 1.2261,
+      "step": 2910
+    },
+    {
+      "epoch": 1.0837802934229968,
+      "grad_norm": 0.17933285236358643,
+      "learning_rate": 1.973151684284272e-05,
+      "loss": 1.243,
+      "step": 2911
+    },
+    {
+      "epoch": 1.0841525985736058,
+      "grad_norm": 0.16889271140098572,
+      "learning_rate": 1.9731237048036197e-05,
+      "loss": 1.227,
+      "step": 2912
+    },
+    {
+      "epoch": 1.084524903724215,
+      "grad_norm": 0.17501141130924225,
+      "learning_rate": 1.9730957109499917e-05,
+      "loss": 1.2439,
+      "step": 2913
+    },
+    {
+      "epoch": 1.084897208874824,
+      "grad_norm": 0.18403035402297974,
+      "learning_rate": 1.973067702723801e-05,
+      "loss": 1.2383,
+      "step": 2914
+    },
+    {
+      "epoch": 1.0852695140254331,
+      "grad_norm": 0.169046550989151,
+      "learning_rate": 1.9730396801254614e-05,
+      "loss": 1.2363,
+      "step": 2915
+    },
+    {
+      "epoch": 1.0856418191760422,
+      "grad_norm": 0.1726294606924057,
+      "learning_rate": 1.973011643155387e-05,
+      "loss": 1.2531,
+      "step": 2916
+    },
+    {
+      "epoch": 1.0860141243266512,
+      "grad_norm": 0.1682724505662918,
+      "learning_rate": 1.9729835918139914e-05,
+      "loss": 1.2225,
+      "step": 2917
+    },
+    {
+      "epoch": 1.0863864294772603,
+      "grad_norm": 0.2326967865228653,
+      "learning_rate": 1.9729555261016894e-05,
+      "loss": 1.2361,
+      "step": 2918
+    },
+    {
+      "epoch": 1.0867587346278693,
+      "grad_norm": 0.1621241271495819,
+      "learning_rate": 1.972927446018895e-05,
+      "loss": 1.2233,
+      "step": 2919
+    },
+    {
+      "epoch": 1.0871310397784784,
+      "grad_norm": 0.1713317185640335,
+      "learning_rate": 1.9728993515660236e-05,
+      "loss": 1.2284,
+      "step": 2920
+    },
+    {
+      "epoch": 1.0875033449290874,
+      "grad_norm": 0.1748856008052826,
+      "learning_rate": 1.9728712427434898e-05,
+      "loss": 1.2337,
+      "step": 2921
+    },
+    {
+      "epoch": 1.0878756500796967,
+      "grad_norm": 0.17016327381134033,
+      "learning_rate": 1.972843119551709e-05,
+      "loss": 1.2292,
+      "step": 2922
+    },
+    {
+      "epoch": 1.0882479552303057,
+      "grad_norm": 0.17471212148666382,
+      "learning_rate": 1.9728149819910958e-05,
+      "loss": 1.231,
+      "step": 2923
+    },
+    {
+      "epoch": 1.0886202603809148,
+      "grad_norm": 0.16408810019493103,
+      "learning_rate": 1.972786830062067e-05,
+      "loss": 1.2305,
+      "step": 2924
+    },
+    {
+      "epoch": 1.0889925655315238,
+      "grad_norm": 0.17399826645851135,
+      "learning_rate": 1.9727586637650373e-05,
+      "loss": 1.2377,
+      "step": 2925
+    },
+    {
+      "epoch": 1.0893648706821328,
+      "grad_norm": 0.18122164905071259,
+      "learning_rate": 1.9727304831004232e-05,
+      "loss": 1.2378,
+      "step": 2926
+    },
+    {
+      "epoch": 1.0897371758327419,
+      "grad_norm": 0.17349712550640106,
+      "learning_rate": 1.9727022880686413e-05,
+      "loss": 1.2237,
+      "step": 2927
+    },
+    {
+      "epoch": 1.090109480983351,
+      "grad_norm": 0.18894222378730774,
+      "learning_rate": 1.9726740786701075e-05,
+      "loss": 1.2254,
+      "step": 2928
+    },
+    {
+      "epoch": 1.09048178613396,
+      "grad_norm": 0.184165820479393,
+      "learning_rate": 1.9726458549052384e-05,
+      "loss": 1.2344,
+      "step": 2929
+    },
+    {
+      "epoch": 1.090854091284569,
+      "grad_norm": 0.18154264986515045,
+      "learning_rate": 1.9726176167744513e-05,
+      "loss": 1.225,
+      "step": 2930
+    },
+    {
+      "epoch": 1.0912263964351783,
+      "grad_norm": 0.17851291596889496,
+      "learning_rate": 1.972589364278163e-05,
+      "loss": 1.237,
+      "step": 2931
+    },
+    {
+      "epoch": 1.0915987015857873,
+      "grad_norm": 0.17904578149318695,
+      "learning_rate": 1.9725610974167907e-05,
+      "loss": 1.2378,
+      "step": 2932
+    },
+    {
+      "epoch": 1.0919710067363964,
+      "grad_norm": 0.16910214722156525,
+      "learning_rate": 1.972532816190752e-05,
+      "loss": 1.2342,
+      "step": 2933
+    },
+    {
+      "epoch": 1.0923433118870054,
+      "grad_norm": 0.17645606398582458,
+      "learning_rate": 1.972504520600465e-05,
+      "loss": 1.23,
+      "step": 2934
+    },
+    {
+      "epoch": 1.0927156170376144,
+      "grad_norm": 0.16807736456394196,
+      "learning_rate": 1.9724762106463467e-05,
+      "loss": 1.2305,
+      "step": 2935
+    },
+    {
+      "epoch": 1.0930879221882235,
+      "grad_norm": 0.20768019556999207,
+      "learning_rate": 1.972447886328816e-05,
+      "loss": 1.2291,
+      "step": 2936
+    },
+    {
+      "epoch": 1.0934602273388325,
+      "grad_norm": 0.1863759607076645,
+      "learning_rate": 1.9724195476482914e-05,
+      "loss": 1.2371,
+      "step": 2937
+    },
+    {
+      "epoch": 1.0938325324894416,
+      "grad_norm": 0.17857873439788818,
+      "learning_rate": 1.9723911946051905e-05,
+      "loss": 1.2179,
+      "step": 2938
+    },
+    {
+      "epoch": 1.0942048376400506,
+      "grad_norm": 0.17775695025920868,
+      "learning_rate": 1.972362827199933e-05,
+      "loss": 1.2498,
+      "step": 2939
+    },
+    {
+      "epoch": 1.0945771427906599,
+      "grad_norm": 0.1699213981628418,
+      "learning_rate": 1.9723344454329376e-05,
+      "loss": 1.2439,
+      "step": 2940
+    },
+    {
+      "epoch": 1.094949447941269,
+      "grad_norm": 0.1817527562379837,
+      "learning_rate": 1.9723060493046235e-05,
+      "loss": 1.2276,
+      "step": 2941
+    },
+    {
+      "epoch": 1.095321753091878,
+      "grad_norm": 0.17052580416202545,
+      "learning_rate": 1.97227763881541e-05,
+      "loss": 1.2161,
+      "step": 2942
+    },
+    {
+      "epoch": 1.095694058242487,
+      "grad_norm": 0.17107853293418884,
+      "learning_rate": 1.972249213965717e-05,
+      "loss": 1.2265,
+      "step": 2943
+    },
+    {
+      "epoch": 1.096066363393096,
+      "grad_norm": 0.17966510355472565,
+      "learning_rate": 1.9722207747559636e-05,
+      "loss": 1.2304,
+      "step": 2944
+    },
+    {
+      "epoch": 1.096438668543705,
+      "grad_norm": 0.1866060048341751,
+      "learning_rate": 1.972192321186571e-05,
+      "loss": 1.2266,
+      "step": 2945
+    },
+    {
+      "epoch": 1.0968109736943141,
+      "grad_norm": 0.17971013486385345,
+      "learning_rate": 1.9721638532579584e-05,
+      "loss": 1.2333,
+      "step": 2946
+    },
+    {
+      "epoch": 1.0971832788449232,
+      "grad_norm": 0.17302943766117096,
+      "learning_rate": 1.972135370970547e-05,
+      "loss": 1.2367,
+      "step": 2947
+    },
+    {
+      "epoch": 1.0975555839955324,
+      "grad_norm": 0.17640618979930878,
+      "learning_rate": 1.9721068743247568e-05,
+      "loss": 1.2385,
+      "step": 2948
+    },
+    {
+      "epoch": 1.0979278891461415,
+      "grad_norm": 0.17200317978858948,
+      "learning_rate": 1.972078363321009e-05,
+      "loss": 1.2277,
+      "step": 2949
+    },
+    {
+      "epoch": 1.0983001942967505,
+      "grad_norm": 0.1816646158695221,
+      "learning_rate": 1.9720498379597256e-05,
+      "loss": 1.2319,
+      "step": 2950
+    },
+    {
+      "epoch": 1.0986724994473596,
+      "grad_norm": 0.17960673570632935,
+      "learning_rate": 1.9720212982413262e-05,
+      "loss": 1.2235,
+      "step": 2951
+    },
+    {
+      "epoch": 1.0990448045979686,
+      "grad_norm": 0.17489486932754517,
+      "learning_rate": 1.971992744166234e-05,
+      "loss": 1.2377,
+      "step": 2952
+    },
+    {
+      "epoch": 1.0994171097485776,
+      "grad_norm": 0.17041970789432526,
+      "learning_rate": 1.9719641757348694e-05,
+      "loss": 1.226,
+      "step": 2953
+    },
+    {
+      "epoch": 1.0997894148991867,
+      "grad_norm": 0.18417948484420776,
+      "learning_rate": 1.9719355929476548e-05,
+      "loss": 1.237,
+      "step": 2954
+    },
+    {
+      "epoch": 1.1001617200497957,
+      "grad_norm": 0.17019295692443848,
+      "learning_rate": 1.971906995805013e-05,
+      "loss": 1.225,
+      "step": 2955
+    },
+    {
+      "epoch": 1.100534025200405,
+      "grad_norm": 0.17955681681632996,
+      "learning_rate": 1.9718783843073653e-05,
+      "loss": 1.2361,
+      "step": 2956
+    },
+    {
+      "epoch": 1.100906330351014,
+      "grad_norm": 0.3305249810218811,
+      "learning_rate": 1.9718497584551355e-05,
+      "loss": 1.2538,
+      "step": 2957
+    },
+    {
+      "epoch": 1.101278635501623,
+      "grad_norm": 0.18145957589149475,
+      "learning_rate": 1.9718211182487455e-05,
+      "loss": 1.2218,
+      "step": 2958
+    },
+    {
+      "epoch": 1.1016509406522321,
+      "grad_norm": 0.17651425302028656,
+      "learning_rate": 1.9717924636886186e-05,
+      "loss": 1.2455,
+      "step": 2959
+    },
+    {
+      "epoch": 1.1020232458028412,
+      "grad_norm": 0.17769207060337067,
+      "learning_rate": 1.971763794775178e-05,
+      "loss": 1.2381,
+      "step": 2960
+    },
+    {
+      "epoch": 1.1023955509534502,
+      "grad_norm": 0.17026957869529724,
+      "learning_rate": 1.971735111508847e-05,
+      "loss": 1.2176,
+      "step": 2961
+    },
+    {
+      "epoch": 1.1027678561040593,
+      "grad_norm": 0.17879103124141693,
+      "learning_rate": 1.9717064138900494e-05,
+      "loss": 1.2371,
+      "step": 2962
+    },
+    {
+      "epoch": 1.1031401612546683,
+      "grad_norm": 0.16946665942668915,
+      "learning_rate": 1.9716777019192087e-05,
+      "loss": 1.2431,
+      "step": 2963
+    },
+    {
+      "epoch": 1.1035124664052773,
+      "grad_norm": 0.16764768958091736,
+      "learning_rate": 1.97164897559675e-05,
+      "loss": 1.2293,
+      "step": 2964
+    },
+    {
+      "epoch": 1.1038847715558866,
+      "grad_norm": 0.1787741482257843,
+      "learning_rate": 1.9716202349230967e-05,
+      "loss": 1.2273,
+      "step": 2965
+    },
+    {
+      "epoch": 1.1042570767064956,
+      "grad_norm": 0.18097257614135742,
+      "learning_rate": 1.971591479898673e-05,
+      "loss": 1.2405,
+      "step": 2966
+    },
+    {
+      "epoch": 1.1046293818571047,
+      "grad_norm": 0.16577765345573425,
+      "learning_rate": 1.9715627105239048e-05,
+      "loss": 1.2213,
+      "step": 2967
+    },
+    {
+      "epoch": 1.1050016870077137,
+      "grad_norm": 0.17236430943012238,
+      "learning_rate": 1.9715339267992162e-05,
+      "loss": 1.2358,
+      "step": 2968
+    },
+    {
+      "epoch": 1.1053739921583228,
+      "grad_norm": 0.1793234795331955,
+      "learning_rate": 1.9715051287250322e-05,
+      "loss": 1.2465,
+      "step": 2969
+    },
+    {
+      "epoch": 1.1057462973089318,
+      "grad_norm": 0.17169257998466492,
+      "learning_rate": 1.9714763163017788e-05,
+      "loss": 1.2308,
+      "step": 2970
+    },
+    {
+      "epoch": 1.1061186024595409,
+      "grad_norm": 0.16924230754375458,
+      "learning_rate": 1.9714474895298807e-05,
+      "loss": 1.2452,
+      "step": 2971
+    },
+    {
+      "epoch": 1.10649090761015,
+      "grad_norm": 0.17970192432403564,
+      "learning_rate": 1.9714186484097646e-05,
+      "loss": 1.2216,
+      "step": 2972
+    },
+    {
+      "epoch": 1.106863212760759,
+      "grad_norm": 0.17152759432792664,
+      "learning_rate": 1.9713897929418556e-05,
+      "loss": 1.2237,
+      "step": 2973
+    },
+    {
+      "epoch": 1.1072355179113682,
+      "grad_norm": 0.18560026586055756,
+      "learning_rate": 1.9713609231265807e-05,
+      "loss": 1.2257,
+      "step": 2974
+    },
+    {
+      "epoch": 1.1076078230619772,
+      "grad_norm": 0.17931891977787018,
+      "learning_rate": 1.9713320389643658e-05,
+      "loss": 1.2455,
+      "step": 2975
+    },
+    {
+      "epoch": 1.1079801282125863,
+      "grad_norm": 0.17655859887599945,
+      "learning_rate": 1.9713031404556377e-05,
+      "loss": 1.2415,
+      "step": 2976
+    },
+    {
+      "epoch": 1.1083524333631953,
+      "grad_norm": 0.17913676798343658,
+      "learning_rate": 1.971274227600823e-05,
+      "loss": 1.2298,
+      "step": 2977
+    },
+    {
+      "epoch": 1.1087247385138044,
+      "grad_norm": 0.17090395092964172,
+      "learning_rate": 1.971245300400349e-05,
+      "loss": 1.2242,
+      "step": 2978
+    },
+    {
+      "epoch": 1.1090970436644134,
+      "grad_norm": 0.17184513807296753,
+      "learning_rate": 1.9712163588546426e-05,
+      "loss": 1.2221,
+      "step": 2979
+    },
+    {
+      "epoch": 1.1094693488150225,
+      "grad_norm": 0.17693102359771729,
+      "learning_rate": 1.971187402964132e-05,
+      "loss": 1.2345,
+      "step": 2980
+    },
+    {
+      "epoch": 1.1098416539656315,
+      "grad_norm": 0.17723120748996735,
+      "learning_rate": 1.9711584327292442e-05,
+      "loss": 1.2168,
+      "step": 2981
+    },
+    {
+      "epoch": 1.1102139591162405,
+      "grad_norm": 0.1746647208929062,
+      "learning_rate": 1.9711294481504074e-05,
+      "loss": 1.2297,
+      "step": 2982
+    },
+    {
+      "epoch": 1.1105862642668498,
+      "grad_norm": 0.17656634747982025,
+      "learning_rate": 1.9711004492280495e-05,
+      "loss": 1.2339,
+      "step": 2983
+    },
+    {
+      "epoch": 1.1109585694174589,
+      "grad_norm": 0.1806725114583969,
+      "learning_rate": 1.9710714359625987e-05,
+      "loss": 1.2305,
+      "step": 2984
+    },
+    {
+      "epoch": 1.111330874568068,
+      "grad_norm": 0.17355605959892273,
+      "learning_rate": 1.9710424083544837e-05,
+      "loss": 1.2275,
+      "step": 2985
+    },
+    {
+      "epoch": 1.111703179718677,
+      "grad_norm": 0.17555807530879974,
+      "learning_rate": 1.9710133664041338e-05,
+      "loss": 1.2219,
+      "step": 2986
+    },
+    {
+      "epoch": 1.112075484869286,
+      "grad_norm": 0.17484989762306213,
+      "learning_rate": 1.9709843101119772e-05,
+      "loss": 1.236,
+      "step": 2987
+    },
+    {
+      "epoch": 1.112447790019895,
+      "grad_norm": 0.1751967966556549,
+      "learning_rate": 1.970955239478443e-05,
+      "loss": 1.2551,
+      "step": 2988
+    },
+    {
+      "epoch": 1.112820095170504,
+      "grad_norm": 0.1683066189289093,
+      "learning_rate": 1.9709261545039614e-05,
+      "loss": 1.2307,
+      "step": 2989
+    },
+    {
+      "epoch": 1.113192400321113,
+      "grad_norm": 0.17595693469047546,
+      "learning_rate": 1.970897055188961e-05,
+      "loss": 1.247,
+      "step": 2990
+    },
+    {
+      "epoch": 1.1135647054717221,
+      "grad_norm": 0.17412763833999634,
+      "learning_rate": 1.9708679415338722e-05,
+      "loss": 1.2353,
+      "step": 2991
+    },
+    {
+      "epoch": 1.1139370106223314,
+      "grad_norm": 0.17140118777751923,
+      "learning_rate": 1.9708388135391247e-05,
+      "loss": 1.2259,
+      "step": 2992
+    },
+    {
+      "epoch": 1.1143093157729405,
+      "grad_norm": 0.17357899248600006,
+      "learning_rate": 1.970809671205149e-05,
+      "loss": 1.2314,
+      "step": 2993
+    },
+    {
+      "epoch": 1.1146816209235495,
+      "grad_norm": 0.16457659006118774,
+      "learning_rate": 1.970780514532375e-05,
+      "loss": 1.2371,
+      "step": 2994
+    },
+    {
+      "epoch": 1.1150539260741585,
+      "grad_norm": 0.17428340017795563,
+      "learning_rate": 1.970751343521234e-05,
+      "loss": 1.227,
+      "step": 2995
+    },
+    {
+      "epoch": 1.1154262312247676,
+      "grad_norm": 0.18031379580497742,
+      "learning_rate": 1.9707221581721568e-05,
+      "loss": 1.2366,
+      "step": 2996
+    },
+    {
+      "epoch": 1.1157985363753766,
+      "grad_norm": 0.16890005767345428,
+      "learning_rate": 1.9706929584855737e-05,
+      "loss": 1.2296,
+      "step": 2997
+    },
+    {
+      "epoch": 1.1161708415259857,
+      "grad_norm": 0.16828297078609467,
+      "learning_rate": 1.970663744461917e-05,
+      "loss": 1.2273,
+      "step": 2998
+    },
+    {
+      "epoch": 1.1165431466765947,
+      "grad_norm": 0.17010074853897095,
+      "learning_rate": 1.9706345161016177e-05,
+      "loss": 1.2263,
+      "step": 2999
+    },
+    {
+      "epoch": 1.116915451827204,
+      "grad_norm": 0.1739543378353119,
+      "learning_rate": 1.970605273405107e-05,
+      "loss": 1.2333,
+      "step": 3000
+    },
+    {
+      "epoch": 1.116915451827204,
+      "eval_loss": 1.324150562286377,
+      "eval_runtime": 16.2397,
+      "eval_samples_per_second": 106.776,
+      "eval_steps_per_second": 5.357,
+      "step": 3000
+    },
+    {
+      "epoch": 1.117287756977813,
+      "grad_norm": 0.1825607717037201,
+      "learning_rate": 1.970576016372818e-05,
+      "loss": 1.2399,
+      "step": 3001
+    },
+    {
+      "epoch": 1.117660062128422,
+      "grad_norm": 0.17324824631214142,
+      "learning_rate": 1.970546745005182e-05,
+      "loss": 1.2294,
+      "step": 3002
+    },
+    {
+      "epoch": 1.118032367279031,
+      "grad_norm": 0.17063994705677032,
+      "learning_rate": 1.970517459302631e-05,
+      "loss": 1.2403,
+      "step": 3003
+    },
+    {
+      "epoch": 1.1184046724296401,
+      "grad_norm": 0.16882003843784332,
+      "learning_rate": 1.9704881592655983e-05,
+      "loss": 1.2178,
+      "step": 3004
+    },
+    {
+      "epoch": 1.1187769775802492,
+      "grad_norm": 0.17386655509471893,
+      "learning_rate": 1.9704588448945167e-05,
+      "loss": 1.2296,
+      "step": 3005
+    },
+    {
+      "epoch": 1.1191492827308582,
+      "grad_norm": 0.16440868377685547,
+      "learning_rate": 1.9704295161898185e-05,
+      "loss": 1.2355,
+      "step": 3006
+    },
+    {
+      "epoch": 1.1195215878814673,
+      "grad_norm": 0.17162500321865082,
+      "learning_rate": 1.9704001731519374e-05,
+      "loss": 1.2373,
+      "step": 3007
+    },
+    {
+      "epoch": 1.1198938930320763,
+      "grad_norm": 0.1756156086921692,
+      "learning_rate": 1.9703708157813068e-05,
+      "loss": 1.2501,
+      "step": 3008
+    },
+    {
+      "epoch": 1.1202661981826856,
+      "grad_norm": 0.17049741744995117,
+      "learning_rate": 1.9703414440783596e-05,
+      "loss": 1.2334,
+      "step": 3009
+    },
+    {
+      "epoch": 1.1206385033332946,
+      "grad_norm": 0.17266800999641418,
+      "learning_rate": 1.9703120580435303e-05,
+      "loss": 1.2251,
+      "step": 3010
+    },
+    {
+      "epoch": 1.1210108084839037,
+      "grad_norm": 0.17048951983451843,
+      "learning_rate": 1.9702826576772535e-05,
+      "loss": 1.2344,
+      "step": 3011
+    },
+    {
+      "epoch": 1.1213831136345127,
+      "grad_norm": 0.17931020259857178,
+      "learning_rate": 1.9702532429799622e-05,
+      "loss": 1.2263,
+      "step": 3012
+    },
+    {
+      "epoch": 1.1217554187851217,
+      "grad_norm": 0.16861797869205475,
+      "learning_rate": 1.9702238139520912e-05,
+      "loss": 1.2332,
+      "step": 3013
+    },
+    {
+      "epoch": 1.1221277239357308,
+      "grad_norm": 0.1716170459985733,
+      "learning_rate": 1.9701943705940758e-05,
+      "loss": 1.2288,
+      "step": 3014
+    },
+    {
+      "epoch": 1.1225000290863398,
+      "grad_norm": 0.1761639565229416,
+      "learning_rate": 1.9701649129063503e-05,
+      "loss": 1.229,
+      "step": 3015
+    },
+    {
+      "epoch": 1.1228723342369489,
+      "grad_norm": 0.17630340158939362,
+      "learning_rate": 1.9701354408893495e-05,
+      "loss": 1.2453,
+      "step": 3016
+    },
+    {
+      "epoch": 1.1232446393875581,
+      "grad_norm": 0.17522114515304565,
+      "learning_rate": 1.9701059545435094e-05,
+      "loss": 1.2242,
+      "step": 3017
+    },
+    {
+      "epoch": 1.1236169445381672,
+      "grad_norm": 0.17642271518707275,
+      "learning_rate": 1.9700764538692657e-05,
+      "loss": 1.2354,
+      "step": 3018
+    },
+    {
+      "epoch": 1.1239892496887762,
+      "grad_norm": 0.17362841963768005,
+      "learning_rate": 1.970046938867053e-05,
+      "loss": 1.2199,
+      "step": 3019
+    },
+    {
+      "epoch": 1.1243615548393853,
+      "grad_norm": 0.17298756539821625,
+      "learning_rate": 1.970017409537308e-05,
+      "loss": 1.2324,
+      "step": 3020
+    },
+    {
+      "epoch": 1.1247338599899943,
+      "grad_norm": 0.17711667716503143,
+      "learning_rate": 1.9699878658804673e-05,
+      "loss": 1.2111,
+      "step": 3021
+    },
+    {
+      "epoch": 1.1251061651406034,
+      "grad_norm": 0.1722431480884552,
+      "learning_rate": 1.969958307896966e-05,
+      "loss": 1.2308,
+      "step": 3022
+    },
+    {
+      "epoch": 1.1254784702912124,
+      "grad_norm": 0.17593185603618622,
+      "learning_rate": 1.969928735587242e-05,
+      "loss": 1.2335,
+      "step": 3023
+    },
+    {
+      "epoch": 1.1258507754418214,
+      "grad_norm": 0.17119091749191284,
+      "learning_rate": 1.969899148951731e-05,
+      "loss": 1.2294,
+      "step": 3024
+    },
+    {
+      "epoch": 1.1262230805924305,
+      "grad_norm": 0.1774984896183014,
+      "learning_rate": 1.9698695479908706e-05,
+      "loss": 1.2279,
+      "step": 3025
+    },
+    {
+      "epoch": 1.1265953857430397,
+      "grad_norm": 0.17215633392333984,
+      "learning_rate": 1.9698399327050976e-05,
+      "loss": 1.227,
+      "step": 3026
+    },
+    {
+      "epoch": 1.1269676908936488,
+      "grad_norm": 0.1741963028907776,
+      "learning_rate": 1.96981030309485e-05,
+      "loss": 1.2408,
+      "step": 3027
+    },
+    {
+      "epoch": 1.1273399960442578,
+      "grad_norm": 0.1749897301197052,
+      "learning_rate": 1.969780659160565e-05,
+      "loss": 1.2416,
+      "step": 3028
+    },
+    {
+      "epoch": 1.1277123011948669,
+      "grad_norm": 0.16878561675548553,
+      "learning_rate": 1.9697510009026803e-05,
+      "loss": 1.2395,
+      "step": 3029
+    },
+    {
+      "epoch": 1.128084606345476,
+      "grad_norm": 0.1697298288345337,
+      "learning_rate": 1.9697213283216342e-05,
+      "loss": 1.2274,
+      "step": 3030
+    },
+    {
+      "epoch": 1.128456911496085,
+      "grad_norm": 0.166582852602005,
+      "learning_rate": 1.9696916414178652e-05,
+      "loss": 1.2283,
+      "step": 3031
+    },
+    {
+      "epoch": 1.128829216646694,
+      "grad_norm": 0.1713571697473526,
+      "learning_rate": 1.9696619401918112e-05,
+      "loss": 1.218,
+      "step": 3032
+    },
+    {
+      "epoch": 1.129201521797303,
+      "grad_norm": 0.1708502322435379,
+      "learning_rate": 1.9696322246439113e-05,
+      "loss": 1.2272,
+      "step": 3033
+    },
+    {
+      "epoch": 1.129573826947912,
+      "grad_norm": 0.16899080574512482,
+      "learning_rate": 1.9696024947746047e-05,
+      "loss": 1.2354,
+      "step": 3034
+    },
+    {
+      "epoch": 1.1299461320985213,
+      "grad_norm": 0.16772188246250153,
+      "learning_rate": 1.9695727505843298e-05,
+      "loss": 1.2415,
+      "step": 3035
+    },
+    {
+      "epoch": 1.1303184372491304,
+      "grad_norm": 0.17296922206878662,
+      "learning_rate": 1.969542992073526e-05,
+      "loss": 1.2145,
+      "step": 3036
+    },
+    {
+      "epoch": 1.1306907423997394,
+      "grad_norm": 0.16814911365509033,
+      "learning_rate": 1.9695132192426334e-05,
+      "loss": 1.2418,
+      "step": 3037
+    },
+    {
+      "epoch": 1.1310630475503485,
+      "grad_norm": 0.16937977075576782,
+      "learning_rate": 1.9694834320920912e-05,
+      "loss": 1.2245,
+      "step": 3038
+    },
+    {
+      "epoch": 1.1314353527009575,
+      "grad_norm": 0.17156411707401276,
+      "learning_rate": 1.9694536306223394e-05,
+      "loss": 1.2336,
+      "step": 3039
+    },
+    {
+      "epoch": 1.1318076578515666,
+      "grad_norm": 0.16524691879749298,
+      "learning_rate": 1.9694238148338186e-05,
+      "loss": 1.2354,
+      "step": 3040
+    },
+    {
+      "epoch": 1.1321799630021756,
+      "grad_norm": 0.1789710819721222,
+      "learning_rate": 1.9693939847269688e-05,
+      "loss": 1.2243,
+      "step": 3041
+    },
+    {
+      "epoch": 1.1325522681527846,
+      "grad_norm": 0.16184090077877045,
+      "learning_rate": 1.9693641403022308e-05,
+      "loss": 1.2194,
+      "step": 3042
+    },
+    {
+      "epoch": 1.1329245733033937,
+      "grad_norm": 0.17413978278636932,
+      "learning_rate": 1.969334281560045e-05,
+      "loss": 1.238,
+      "step": 3043
+    },
+    {
+      "epoch": 1.133296878454003,
+      "grad_norm": 0.17365868389606476,
+      "learning_rate": 1.969304408500853e-05,
+      "loss": 1.2253,
+      "step": 3044
+    },
+    {
+      "epoch": 1.133669183604612,
+      "grad_norm": 0.16915839910507202,
+      "learning_rate": 1.9692745211250957e-05,
+      "loss": 1.2234,
+      "step": 3045
+    },
+    {
+      "epoch": 1.134041488755221,
+      "grad_norm": 0.17093431949615479,
+      "learning_rate": 1.9692446194332144e-05,
+      "loss": 1.2275,
+      "step": 3046
+    },
+    {
+      "epoch": 1.13441379390583,
+      "grad_norm": 0.1693648397922516,
+      "learning_rate": 1.969214703425651e-05,
+      "loss": 1.2381,
+      "step": 3047
+    },
+    {
+      "epoch": 1.1347860990564391,
+      "grad_norm": 0.16665521264076233,
+      "learning_rate": 1.969184773102847e-05,
+      "loss": 1.2276,
+      "step": 3048
+    },
+    {
+      "epoch": 1.1351584042070482,
+      "grad_norm": 0.17075350880622864,
+      "learning_rate": 1.969154828465245e-05,
+      "loss": 1.2314,
+      "step": 3049
+    },
+    {
+      "epoch": 1.1355307093576572,
+      "grad_norm": 0.18020687997341156,
+      "learning_rate": 1.969124869513287e-05,
+      "loss": 1.2448,
+      "step": 3050
+    },
+    {
+      "epoch": 1.1359030145082663,
+      "grad_norm": 0.17107802629470825,
+      "learning_rate": 1.9690948962474152e-05,
+      "loss": 1.2441,
+      "step": 3051
+    },
+    {
+      "epoch": 1.1362753196588753,
+      "grad_norm": 0.1705516129732132,
+      "learning_rate": 1.969064908668073e-05,
+      "loss": 1.2429,
+      "step": 3052
+    },
+    {
+      "epoch": 1.1366476248094846,
+      "grad_norm": 0.16781964898109436,
+      "learning_rate": 1.9690349067757026e-05,
+      "loss": 1.2203,
+      "step": 3053
+    },
+    {
+      "epoch": 1.1370199299600936,
+      "grad_norm": 0.17965157330036163,
+      "learning_rate": 1.9690048905707476e-05,
+      "loss": 1.2352,
+      "step": 3054
+    },
+    {
+      "epoch": 1.1373922351107026,
+      "grad_norm": 0.16421709954738617,
+      "learning_rate": 1.9689748600536514e-05,
+      "loss": 1.2138,
+      "step": 3055
+    },
+    {
+      "epoch": 1.1377645402613117,
+      "grad_norm": 0.17832623422145844,
+      "learning_rate": 1.9689448152248568e-05,
+      "loss": 1.2288,
+      "step": 3056
+    },
+    {
+      "epoch": 1.1381368454119207,
+      "grad_norm": 0.16908565163612366,
+      "learning_rate": 1.9689147560848086e-05,
+      "loss": 1.2158,
+      "step": 3057
+    },
+    {
+      "epoch": 1.1385091505625298,
+      "grad_norm": 0.18035708367824554,
+      "learning_rate": 1.9688846826339498e-05,
+      "loss": 1.2292,
+      "step": 3058
+    },
+    {
+      "epoch": 1.1388814557131388,
+      "grad_norm": 0.18107391893863678,
+      "learning_rate": 1.9688545948727255e-05,
+      "loss": 1.2215,
+      "step": 3059
+    },
+    {
+      "epoch": 1.1392537608637479,
+      "grad_norm": 0.17275120317935944,
+      "learning_rate": 1.9688244928015795e-05,
+      "loss": 1.2262,
+      "step": 3060
+    },
+    {
+      "epoch": 1.139626066014357,
+      "grad_norm": 0.17838861048221588,
+      "learning_rate": 1.9687943764209564e-05,
+      "loss": 1.2246,
+      "step": 3061
+    },
+    {
+      "epoch": 1.1399983711649662,
+      "grad_norm": 0.17901575565338135,
+      "learning_rate": 1.968764245731301e-05,
+      "loss": 1.2299,
+      "step": 3062
+    },
+    {
+      "epoch": 1.1403706763155752,
+      "grad_norm": 0.17386050522327423,
+      "learning_rate": 1.9687341007330588e-05,
+      "loss": 1.2214,
+      "step": 3063
+    },
+    {
+      "epoch": 1.1407429814661842,
+      "grad_norm": 0.17698942124843597,
+      "learning_rate": 1.9687039414266745e-05,
+      "loss": 1.2363,
+      "step": 3064
+    },
+    {
+      "epoch": 1.1411152866167933,
+      "grad_norm": 0.16935193538665771,
+      "learning_rate": 1.968673767812594e-05,
+      "loss": 1.2189,
+      "step": 3065
+    },
+    {
+      "epoch": 1.1414875917674023,
+      "grad_norm": 0.18296149373054504,
+      "learning_rate": 1.9686435798912624e-05,
+      "loss": 1.2252,
+      "step": 3066
+    },
+    {
+      "epoch": 1.1418598969180114,
+      "grad_norm": 0.17983867228031158,
+      "learning_rate": 1.9686133776631263e-05,
+      "loss": 1.2338,
+      "step": 3067
+    },
+    {
+      "epoch": 1.1422322020686204,
+      "grad_norm": 0.17501530051231384,
+      "learning_rate": 1.9685831611286312e-05,
+      "loss": 1.2168,
+      "step": 3068
+    },
+    {
+      "epoch": 1.1426045072192297,
+      "grad_norm": 0.17005561292171478,
+      "learning_rate": 1.9685529302882237e-05,
+      "loss": 1.2367,
+      "step": 3069
+    },
+    {
+      "epoch": 1.1429768123698387,
+      "grad_norm": 0.1763869673013687,
+      "learning_rate": 1.9685226851423502e-05,
+      "loss": 1.2142,
+      "step": 3070
+    },
+    {
+      "epoch": 1.1433491175204478,
+      "grad_norm": 0.17087772488594055,
+      "learning_rate": 1.9684924256914574e-05,
+      "loss": 1.2228,
+      "step": 3071
+    },
+    {
+      "epoch": 1.1437214226710568,
+      "grad_norm": 0.167043536901474,
+      "learning_rate": 1.968462151935992e-05,
+      "loss": 1.2148,
+      "step": 3072
+    },
+    {
+      "epoch": 1.1440937278216659,
+      "grad_norm": 0.17006975412368774,
+      "learning_rate": 1.968431863876402e-05,
+      "loss": 1.2376,
+      "step": 3073
+    },
+    {
+      "epoch": 1.144466032972275,
+      "grad_norm": 0.175192192196846,
+      "learning_rate": 1.9684015615131336e-05,
+      "loss": 1.227,
+      "step": 3074
+    },
+    {
+      "epoch": 1.144838338122884,
+      "grad_norm": 0.1716625988483429,
+      "learning_rate": 1.968371244846635e-05,
+      "loss": 1.2399,
+      "step": 3075
+    },
+    {
+      "epoch": 1.145210643273493,
+      "grad_norm": 0.17771419882774353,
+      "learning_rate": 1.9683409138773538e-05,
+      "loss": 1.2311,
+      "step": 3076
+    },
+    {
+      "epoch": 1.145582948424102,
+      "grad_norm": 0.17725834250450134,
+      "learning_rate": 1.9683105686057383e-05,
+      "loss": 1.2195,
+      "step": 3077
+    },
+    {
+      "epoch": 1.1459552535747113,
+      "grad_norm": 0.17201553285121918,
+      "learning_rate": 1.9682802090322365e-05,
+      "loss": 1.2308,
+      "step": 3078
+    },
+    {
+      "epoch": 1.1463275587253203,
+      "grad_norm": 0.17508849501609802,
+      "learning_rate": 1.968249835157297e-05,
+      "loss": 1.2278,
+      "step": 3079
+    },
+    {
+      "epoch": 1.1466998638759294,
+      "grad_norm": 0.16712623834609985,
+      "learning_rate": 1.9682194469813675e-05,
+      "loss": 1.233,
+      "step": 3080
+    },
+    {
+      "epoch": 1.1470721690265384,
+      "grad_norm": 0.1693454533815384,
+      "learning_rate": 1.968189044504898e-05,
+      "loss": 1.2292,
+      "step": 3081
+    },
+    {
+      "epoch": 1.1474444741771475,
+      "grad_norm": 0.17314909398555756,
+      "learning_rate": 1.968158627728337e-05,
+      "loss": 1.2371,
+      "step": 3082
+    },
+    {
+      "epoch": 1.1478167793277565,
+      "grad_norm": 0.1748846024274826,
+      "learning_rate": 1.968128196652134e-05,
+      "loss": 1.2337,
+      "step": 3083
+    },
+    {
+      "epoch": 1.1481890844783655,
+      "grad_norm": 0.16931162774562836,
+      "learning_rate": 1.968097751276738e-05,
+      "loss": 1.2321,
+      "step": 3084
+    },
+    {
+      "epoch": 1.1485613896289746,
+      "grad_norm": 0.17732059955596924,
+      "learning_rate": 1.9680672916025993e-05,
+      "loss": 1.2152,
+      "step": 3085
+    },
+    {
+      "epoch": 1.1489336947795836,
+      "grad_norm": 0.17309656739234924,
+      "learning_rate": 1.9680368176301673e-05,
+      "loss": 1.235,
+      "step": 3086
+    },
+    {
+      "epoch": 1.149305999930193,
+      "grad_norm": 0.17718668282032013,
+      "learning_rate": 1.968006329359892e-05,
+      "loss": 1.2288,
+      "step": 3087
+    },
+    {
+      "epoch": 1.149678305080802,
+      "grad_norm": 0.17663928866386414,
+      "learning_rate": 1.9679758267922243e-05,
+      "loss": 1.2314,
+      "step": 3088
+    },
+    {
+      "epoch": 1.150050610231411,
+      "grad_norm": 0.16871769726276398,
+      "learning_rate": 1.9679453099276142e-05,
+      "loss": 1.2292,
+      "step": 3089
+    },
+    {
+      "epoch": 1.15042291538202,
+      "grad_norm": 0.177023783326149,
+      "learning_rate": 1.9679147787665128e-05,
+      "loss": 1.211,
+      "step": 3090
+    },
+    {
+      "epoch": 1.150795220532629,
+      "grad_norm": 0.17566607892513275,
+      "learning_rate": 1.9678842333093708e-05,
+      "loss": 1.2255,
+      "step": 3091
+    },
+    {
+      "epoch": 1.151167525683238,
+      "grad_norm": 0.17927557229995728,
+      "learning_rate": 1.9678536735566393e-05,
+      "loss": 1.2353,
+      "step": 3092
+    },
+    {
+      "epoch": 1.1515398308338471,
+      "grad_norm": 0.15841099619865417,
+      "learning_rate": 1.9678230995087696e-05,
+      "loss": 1.2296,
+      "step": 3093
+    },
+    {
+      "epoch": 1.1519121359844562,
+      "grad_norm": 0.17372049391269684,
+      "learning_rate": 1.9677925111662136e-05,
+      "loss": 1.2199,
+      "step": 3094
+    },
+    {
+      "epoch": 1.1522844411350652,
+      "grad_norm": 0.17242653667926788,
+      "learning_rate": 1.9677619085294234e-05,
+      "loss": 1.223,
+      "step": 3095
+    },
+    {
+      "epoch": 1.1526567462856745,
+      "grad_norm": 0.18381370604038239,
+      "learning_rate": 1.96773129159885e-05,
+      "loss": 1.2372,
+      "step": 3096
+    },
+    {
+      "epoch": 1.1530290514362835,
+      "grad_norm": 0.1743118166923523,
+      "learning_rate": 1.9677006603749463e-05,
+      "loss": 1.2133,
+      "step": 3097
+    },
+    {
+      "epoch": 1.1534013565868926,
+      "grad_norm": 0.18232518434524536,
+      "learning_rate": 1.967670014858165e-05,
+      "loss": 1.2328,
+      "step": 3098
+    },
+    {
+      "epoch": 1.1537736617375016,
+      "grad_norm": 0.17845921218395233,
+      "learning_rate": 1.9676393550489576e-05,
+      "loss": 1.2349,
+      "step": 3099
+    },
+    {
+      "epoch": 1.1541459668881107,
+      "grad_norm": 0.18814215064048767,
+      "learning_rate": 1.9676086809477778e-05,
+      "loss": 1.2375,
+      "step": 3100
+    },
+    {
+      "epoch": 1.1545182720387197,
+      "grad_norm": 0.17665289342403412,
+      "learning_rate": 1.9675779925550785e-05,
+      "loss": 1.216,
+      "step": 3101
+    },
+    {
+      "epoch": 1.1548905771893287,
+      "grad_norm": 0.1747797727584839,
+      "learning_rate": 1.9675472898713136e-05,
+      "loss": 1.2291,
+      "step": 3102
+    },
+    {
+      "epoch": 1.1552628823399378,
+      "grad_norm": 0.17490416765213013,
+      "learning_rate": 1.9675165728969353e-05,
+      "loss": 1.2187,
+      "step": 3103
+    },
+    {
+      "epoch": 1.1556351874905468,
+      "grad_norm": 0.17718598246574402,
+      "learning_rate": 1.967485841632398e-05,
+      "loss": 1.2374,
+      "step": 3104
+    },
+    {
+      "epoch": 1.156007492641156,
+      "grad_norm": 0.17890246212482452,
+      "learning_rate": 1.967455096078156e-05,
+      "loss": 1.2184,
+      "step": 3105
+    },
+    {
+      "epoch": 1.1563797977917651,
+      "grad_norm": 0.17774467170238495,
+      "learning_rate": 1.9674243362346624e-05,
+      "loss": 1.2251,
+      "step": 3106
+    },
+    {
+      "epoch": 1.1567521029423742,
+      "grad_norm": 0.1743806153535843,
+      "learning_rate": 1.9673935621023724e-05,
+      "loss": 1.2357,
+      "step": 3107
+    },
+    {
+      "epoch": 1.1571244080929832,
+      "grad_norm": 0.1773483157157898,
+      "learning_rate": 1.96736277368174e-05,
+      "loss": 1.2338,
+      "step": 3108
+    },
+    {
+      "epoch": 1.1574967132435923,
+      "grad_norm": 0.171565443277359,
+      "learning_rate": 1.9673319709732205e-05,
+      "loss": 1.2412,
+      "step": 3109
+    },
+    {
+      "epoch": 1.1578690183942013,
+      "grad_norm": 0.16667889058589935,
+      "learning_rate": 1.967301153977268e-05,
+      "loss": 1.227,
+      "step": 3110
+    },
+    {
+      "epoch": 1.1582413235448104,
+      "grad_norm": 0.18110859394073486,
+      "learning_rate": 1.9672703226943383e-05,
+      "loss": 1.247,
+      "step": 3111
+    },
+    {
+      "epoch": 1.1586136286954194,
+      "grad_norm": 0.18011483550071716,
+      "learning_rate": 1.9672394771248867e-05,
+      "loss": 1.23,
+      "step": 3112
+    },
+    {
+      "epoch": 1.1589859338460284,
+      "grad_norm": 0.17470338940620422,
+      "learning_rate": 1.967208617269369e-05,
+      "loss": 1.2272,
+      "step": 3113
+    },
+    {
+      "epoch": 1.1593582389966377,
+      "grad_norm": 0.16720262169837952,
+      "learning_rate": 1.9671777431282404e-05,
+      "loss": 1.2286,
+      "step": 3114
+    },
+    {
+      "epoch": 1.1597305441472467,
+      "grad_norm": 0.16974225640296936,
+      "learning_rate": 1.9671468547019575e-05,
+      "loss": 1.2358,
+      "step": 3115
+    },
+    {
+      "epoch": 1.1601028492978558,
+      "grad_norm": 0.17322564125061035,
+      "learning_rate": 1.9671159519909758e-05,
+      "loss": 1.2196,
+      "step": 3116
+    },
+    {
+      "epoch": 1.1604751544484648,
+      "grad_norm": 0.17480473220348358,
+      "learning_rate": 1.9670850349957525e-05,
+      "loss": 1.2199,
+      "step": 3117
+    },
+    {
+      "epoch": 1.1608474595990739,
+      "grad_norm": 0.17341922223567963,
+      "learning_rate": 1.967054103716744e-05,
+      "loss": 1.2197,
+      "step": 3118
+    },
+    {
+      "epoch": 1.161219764749683,
+      "grad_norm": 0.16872482001781464,
+      "learning_rate": 1.9670231581544068e-05,
+      "loss": 1.23,
+      "step": 3119
+    },
+    {
+      "epoch": 1.161592069900292,
+      "grad_norm": 0.16933618485927582,
+      "learning_rate": 1.9669921983091985e-05,
+      "loss": 1.2202,
+      "step": 3120
+    },
+    {
+      "epoch": 1.161964375050901,
+      "grad_norm": 0.164505273103714,
+      "learning_rate": 1.966961224181576e-05,
+      "loss": 1.2398,
+      "step": 3121
+    },
+    {
+      "epoch": 1.16233668020151,
+      "grad_norm": 0.1745142638683319,
+      "learning_rate": 1.9669302357719968e-05,
+      "loss": 1.2223,
+      "step": 3122
+    },
+    {
+      "epoch": 1.1627089853521193,
+      "grad_norm": 0.17189304530620575,
+      "learning_rate": 1.9668992330809187e-05,
+      "loss": 1.234,
+      "step": 3123
+    },
+    {
+      "epoch": 1.1630812905027283,
+      "grad_norm": 0.16852304339408875,
+      "learning_rate": 1.9668682161088e-05,
+      "loss": 1.219,
+      "step": 3124
+    },
+    {
+      "epoch": 1.1634535956533374,
+      "grad_norm": 0.17335280776023865,
+      "learning_rate": 1.966837184856098e-05,
+      "loss": 1.2107,
+      "step": 3125
+    },
+    {
+      "epoch": 1.1638259008039464,
+      "grad_norm": 0.17474551498889923,
+      "learning_rate": 1.9668061393232717e-05,
+      "loss": 1.2285,
+      "step": 3126
+    },
+    {
+      "epoch": 1.1641982059545555,
+      "grad_norm": 0.16944244503974915,
+      "learning_rate": 1.9667750795107793e-05,
+      "loss": 1.2281,
+      "step": 3127
+    },
+    {
+      "epoch": 1.1645705111051645,
+      "grad_norm": 0.1715431809425354,
+      "learning_rate": 1.9667440054190793e-05,
+      "loss": 1.2402,
+      "step": 3128
+    },
+    {
+      "epoch": 1.1649428162557736,
+      "grad_norm": 0.1723116636276245,
+      "learning_rate": 1.9667129170486315e-05,
+      "loss": 1.2295,
+      "step": 3129
+    },
+    {
+      "epoch": 1.1653151214063828,
+      "grad_norm": 0.172069251537323,
+      "learning_rate": 1.9666818143998944e-05,
+      "loss": 1.2467,
+      "step": 3130
+    },
+    {
+      "epoch": 1.1656874265569919,
+      "grad_norm": 0.16813892126083374,
+      "learning_rate": 1.9666506974733276e-05,
+      "loss": 1.2321,
+      "step": 3131
+    },
+    {
+      "epoch": 1.166059731707601,
+      "grad_norm": 0.17631420493125916,
+      "learning_rate": 1.9666195662693907e-05,
+      "loss": 1.2281,
+      "step": 3132
+    },
+    {
+      "epoch": 1.16643203685821,
+      "grad_norm": 0.1704123616218567,
+      "learning_rate": 1.9665884207885437e-05,
+      "loss": 1.2453,
+      "step": 3133
+    },
+    {
+      "epoch": 1.166804342008819,
+      "grad_norm": 0.17325717210769653,
+      "learning_rate": 1.966557261031246e-05,
+      "loss": 1.2172,
+      "step": 3134
+    },
+    {
+      "epoch": 1.167176647159428,
+      "grad_norm": 0.16483458876609802,
+      "learning_rate": 1.9665260869979585e-05,
+      "loss": 1.2209,
+      "step": 3135
+    },
+    {
+      "epoch": 1.167548952310037,
+      "grad_norm": 0.17396704852581024,
+      "learning_rate": 1.966494898689141e-05,
+      "loss": 1.2164,
+      "step": 3136
+    },
+    {
+      "epoch": 1.1679212574606461,
+      "grad_norm": 0.18191742897033691,
+      "learning_rate": 1.9664636961052547e-05,
+      "loss": 1.2402,
+      "step": 3137
+    },
+    {
+      "epoch": 1.1682935626112552,
+      "grad_norm": 0.18091581761837006,
+      "learning_rate": 1.9664324792467603e-05,
+      "loss": 1.2276,
+      "step": 3138
+    },
+    {
+      "epoch": 1.1686658677618644,
+      "grad_norm": 0.17447642982006073,
+      "learning_rate": 1.9664012481141185e-05,
+      "loss": 1.2262,
+      "step": 3139
+    },
+    {
+      "epoch": 1.1690381729124735,
+      "grad_norm": 0.17776253819465637,
+      "learning_rate": 1.9663700027077915e-05,
+      "loss": 1.2252,
+      "step": 3140
+    },
+    {
+      "epoch": 1.1694104780630825,
+      "grad_norm": 0.16708004474639893,
+      "learning_rate": 1.96633874302824e-05,
+      "loss": 1.2365,
+      "step": 3141
+    },
+    {
+      "epoch": 1.1697827832136916,
+      "grad_norm": 0.17193494737148285,
+      "learning_rate": 1.9663074690759255e-05,
+      "loss": 1.2224,
+      "step": 3142
+    },
+    {
+      "epoch": 1.1701550883643006,
+      "grad_norm": 0.16979169845581055,
+      "learning_rate": 1.9662761808513105e-05,
+      "loss": 1.2222,
+      "step": 3143
+    },
+    {
+      "epoch": 1.1705273935149096,
+      "grad_norm": 0.17003171145915985,
+      "learning_rate": 1.9662448783548575e-05,
+      "loss": 1.222,
+      "step": 3144
+    },
+    {
+      "epoch": 1.1708996986655187,
+      "grad_norm": 0.17789965867996216,
+      "learning_rate": 1.966213561587028e-05,
+      "loss": 1.2306,
+      "step": 3145
+    },
+    {
+      "epoch": 1.1712720038161277,
+      "grad_norm": 0.17377351224422455,
+      "learning_rate": 1.9661822305482845e-05,
+      "loss": 1.2268,
+      "step": 3146
+    },
+    {
+      "epoch": 1.1716443089667368,
+      "grad_norm": 0.17414116859436035,
+      "learning_rate": 1.96615088523909e-05,
+      "loss": 1.2161,
+      "step": 3147
+    },
+    {
+      "epoch": 1.172016614117346,
+      "grad_norm": 0.17168028652668,
+      "learning_rate": 1.966119525659908e-05,
+      "loss": 1.2287,
+      "step": 3148
+    },
+    {
+      "epoch": 1.172388919267955,
+      "grad_norm": 0.17371198534965515,
+      "learning_rate": 1.966088151811201e-05,
+      "loss": 1.2552,
+      "step": 3149
+    },
+    {
+      "epoch": 1.1727612244185641,
+      "grad_norm": 0.17159205675125122,
+      "learning_rate": 1.966056763693433e-05,
+      "loss": 1.2243,
+      "step": 3150
+    },
+    {
+      "epoch": 1.1731335295691732,
+      "grad_norm": 0.1716485470533371,
+      "learning_rate": 1.9660253613070667e-05,
+      "loss": 1.2413,
+      "step": 3151
+    },
+    {
+      "epoch": 1.1735058347197822,
+      "grad_norm": 0.17491815984249115,
+      "learning_rate": 1.9659939446525668e-05,
+      "loss": 1.2377,
+      "step": 3152
+    },
+    {
+      "epoch": 1.1738781398703912,
+      "grad_norm": 0.166838139295578,
+      "learning_rate": 1.965962513730397e-05,
+      "loss": 1.2133,
+      "step": 3153
+    },
+    {
+      "epoch": 1.1742504450210003,
+      "grad_norm": 0.16914509236812592,
+      "learning_rate": 1.9659310685410212e-05,
+      "loss": 1.2264,
+      "step": 3154
+    },
+    {
+      "epoch": 1.1746227501716093,
+      "grad_norm": 0.17503736913204193,
+      "learning_rate": 1.9658996090849042e-05,
+      "loss": 1.2341,
+      "step": 3155
+    },
+    {
+      "epoch": 1.1749950553222184,
+      "grad_norm": 0.17667143046855927,
+      "learning_rate": 1.9658681353625105e-05,
+      "loss": 1.231,
+      "step": 3156
+    },
+    {
+      "epoch": 1.1753673604728276,
+      "grad_norm": 0.17044112086296082,
+      "learning_rate": 1.9658366473743052e-05,
+      "loss": 1.2339,
+      "step": 3157
+    },
+    {
+      "epoch": 1.1757396656234367,
+      "grad_norm": 0.1663392037153244,
+      "learning_rate": 1.9658051451207536e-05,
+      "loss": 1.2352,
+      "step": 3158
+    },
+    {
+      "epoch": 1.1761119707740457,
+      "grad_norm": 0.18989932537078857,
+      "learning_rate": 1.96577362860232e-05,
+      "loss": 1.2266,
+      "step": 3159
+    },
+    {
+      "epoch": 1.1764842759246548,
+      "grad_norm": 0.18411101400852203,
+      "learning_rate": 1.965742097819471e-05,
+      "loss": 1.2343,
+      "step": 3160
+    },
+    {
+      "epoch": 1.1768565810752638,
+      "grad_norm": 0.17134809494018555,
+      "learning_rate": 1.9657105527726716e-05,
+      "loss": 1.2191,
+      "step": 3161
+    },
+    {
+      "epoch": 1.1772288862258729,
+      "grad_norm": 0.20970846712589264,
+      "learning_rate": 1.965678993462388e-05,
+      "loss": 1.2202,
+      "step": 3162
+    },
+    {
+      "epoch": 1.177601191376482,
+      "grad_norm": 0.20733484625816345,
+      "learning_rate": 1.9656474198890864e-05,
+      "loss": 1.2263,
+      "step": 3163
+    },
+    {
+      "epoch": 1.177973496527091,
+      "grad_norm": 0.18368762731552124,
+      "learning_rate": 1.965615832053233e-05,
+      "loss": 1.2363,
+      "step": 3164
+    },
+    {
+      "epoch": 1.1783458016777,
+      "grad_norm": 0.17400404810905457,
+      "learning_rate": 1.9655842299552938e-05,
+      "loss": 1.2301,
+      "step": 3165
+    },
+    {
+      "epoch": 1.1787181068283092,
+      "grad_norm": 0.19013762474060059,
+      "learning_rate": 1.9655526135957366e-05,
+      "loss": 1.2316,
+      "step": 3166
+    },
+    {
+      "epoch": 1.1790904119789183,
+      "grad_norm": 0.17968544363975525,
+      "learning_rate": 1.965520982975028e-05,
+      "loss": 1.2308,
+      "step": 3167
+    },
+    {
+      "epoch": 1.1794627171295273,
+      "grad_norm": 0.1678808182477951,
+      "learning_rate": 1.965489338093635e-05,
+      "loss": 1.2178,
+      "step": 3168
+    },
+    {
+      "epoch": 1.1798350222801364,
+      "grad_norm": 0.18025662004947662,
+      "learning_rate": 1.9654576789520248e-05,
+      "loss": 1.2261,
+      "step": 3169
+    },
+    {
+      "epoch": 1.1802073274307454,
+      "grad_norm": 0.1803680807352066,
+      "learning_rate": 1.965426005550665e-05,
+      "loss": 1.2291,
+      "step": 3170
+    },
+    {
+      "epoch": 1.1805796325813545,
+      "grad_norm": 0.1786152571439743,
+      "learning_rate": 1.9653943178900246e-05,
+      "loss": 1.2508,
+      "step": 3171
+    },
+    {
+      "epoch": 1.1809519377319635,
+      "grad_norm": 0.17993713915348053,
+      "learning_rate": 1.96536261597057e-05,
+      "loss": 1.224,
+      "step": 3172
+    },
+    {
+      "epoch": 1.1813242428825725,
+      "grad_norm": 0.18419890105724335,
+      "learning_rate": 1.9653308997927705e-05,
+      "loss": 1.236,
+      "step": 3173
+    },
+    {
+      "epoch": 1.1816965480331816,
+      "grad_norm": 0.17795118689537048,
+      "learning_rate": 1.9652991693570938e-05,
+      "loss": 1.2217,
+      "step": 3174
+    },
+    {
+      "epoch": 1.1820688531837908,
+      "grad_norm": 0.17739805579185486,
+      "learning_rate": 1.965267424664009e-05,
+      "loss": 1.2336,
+      "step": 3175
+    },
+    {
+      "epoch": 1.1824411583343999,
+      "grad_norm": 0.17250627279281616,
+      "learning_rate": 1.9652356657139854e-05,
+      "loss": 1.2117,
+      "step": 3176
+    },
+    {
+      "epoch": 1.182813463485009,
+      "grad_norm": 0.1707741916179657,
+      "learning_rate": 1.9652038925074916e-05,
+      "loss": 1.2457,
+      "step": 3177
+    },
+    {
+      "epoch": 1.183185768635618,
+      "grad_norm": 0.17577117681503296,
+      "learning_rate": 1.9651721050449964e-05,
+      "loss": 1.2213,
+      "step": 3178
+    },
+    {
+      "epoch": 1.183558073786227,
+      "grad_norm": 0.1711239516735077,
+      "learning_rate": 1.96514030332697e-05,
+      "loss": 1.2081,
+      "step": 3179
+    },
+    {
+      "epoch": 1.183930378936836,
+      "grad_norm": 0.16911734640598297,
+      "learning_rate": 1.9651084873538816e-05,
+      "loss": 1.2082,
+      "step": 3180
+    },
+    {
+      "epoch": 1.184302684087445,
+      "grad_norm": 0.1710880696773529,
+      "learning_rate": 1.965076657126202e-05,
+      "loss": 1.2193,
+      "step": 3181
+    },
+    {
+      "epoch": 1.1846749892380541,
+      "grad_norm": 0.1759801208972931,
+      "learning_rate": 1.9650448126444003e-05,
+      "loss": 1.2261,
+      "step": 3182
+    },
+    {
+      "epoch": 1.1850472943886632,
+      "grad_norm": 0.17261157929897308,
+      "learning_rate": 1.9650129539089477e-05,
+      "loss": 1.2322,
+      "step": 3183
+    },
+    {
+      "epoch": 1.1854195995392725,
+      "grad_norm": 0.17363622784614563,
+      "learning_rate": 1.9649810809203138e-05,
+      "loss": 1.2263,
+      "step": 3184
+    },
+    {
+      "epoch": 1.1857919046898815,
+      "grad_norm": 0.17991025745868683,
+      "learning_rate": 1.9649491936789702e-05,
+      "loss": 1.224,
+      "step": 3185
+    },
+    {
+      "epoch": 1.1861642098404905,
+      "grad_norm": 0.17815767228603363,
+      "learning_rate": 1.9649172921853873e-05,
+      "loss": 1.2289,
+      "step": 3186
+    },
+    {
+      "epoch": 1.1865365149910996,
+      "grad_norm": 0.1795058697462082,
+      "learning_rate": 1.964885376440037e-05,
+      "loss": 1.2335,
+      "step": 3187
+    },
+    {
+      "epoch": 1.1869088201417086,
+      "grad_norm": 0.16902460157871246,
+      "learning_rate": 1.9648534464433897e-05,
+      "loss": 1.2369,
+      "step": 3188
+    },
+    {
+      "epoch": 1.1872811252923177,
+      "grad_norm": 0.19541476666927338,
+      "learning_rate": 1.964821502195918e-05,
+      "loss": 1.2325,
+      "step": 3189
+    },
+    {
+      "epoch": 1.1876534304429267,
+      "grad_norm": 0.1734236776828766,
+      "learning_rate": 1.964789543698093e-05,
+      "loss": 1.2162,
+      "step": 3190
+    },
+    {
+      "epoch": 1.188025735593536,
+      "grad_norm": 0.26567402482032776,
+      "learning_rate": 1.9647575709503873e-05,
+      "loss": 1.2229,
+      "step": 3191
+    },
+    {
+      "epoch": 1.188398040744145,
+      "grad_norm": 0.1733405888080597,
+      "learning_rate": 1.9647255839532726e-05,
+      "loss": 1.2366,
+      "step": 3192
+    },
+    {
+      "epoch": 1.188770345894754,
+      "grad_norm": 0.17245125770568848,
+      "learning_rate": 1.9646935827072215e-05,
+      "loss": 1.2173,
+      "step": 3193
+    },
+    {
+      "epoch": 1.189142651045363,
+      "grad_norm": 0.16756393015384674,
+      "learning_rate": 1.9646615672127068e-05,
+      "loss": 1.2291,
+      "step": 3194
+    },
+    {
+      "epoch": 1.1895149561959721,
+      "grad_norm": 0.16170787811279297,
+      "learning_rate": 1.9646295374702014e-05,
+      "loss": 1.2338,
+      "step": 3195
+    },
+    {
+      "epoch": 1.1898872613465812,
+      "grad_norm": 0.1687183678150177,
+      "learning_rate": 1.964597493480178e-05,
+      "loss": 1.218,
+      "step": 3196
+    },
+    {
+      "epoch": 1.1902595664971902,
+      "grad_norm": 0.16018569469451904,
+      "learning_rate": 1.9645654352431105e-05,
+      "loss": 1.2158,
+      "step": 3197
+    },
+    {
+      "epoch": 1.1906318716477993,
+      "grad_norm": 0.1713458150625229,
+      "learning_rate": 1.9645333627594717e-05,
+      "loss": 1.2331,
+      "step": 3198
+    },
+    {
+      "epoch": 1.1910041767984083,
+      "grad_norm": 0.1839626431465149,
+      "learning_rate": 1.9645012760297358e-05,
+      "loss": 1.2246,
+      "step": 3199
+    },
+    {
+      "epoch": 1.1913764819490176,
+      "grad_norm": 0.16510577499866486,
+      "learning_rate": 1.964469175054377e-05,
+      "loss": 1.2169,
+      "step": 3200
+    },
+    {
+      "epoch": 1.1917487870996266,
+      "grad_norm": 0.16351880133152008,
+      "learning_rate": 1.9644370598338686e-05,
+      "loss": 1.237,
+      "step": 3201
+    },
+    {
+      "epoch": 1.1921210922502357,
+      "grad_norm": 0.16644950211048126,
+      "learning_rate": 1.9644049303686852e-05,
+      "loss": 1.2395,
+      "step": 3202
+    },
+    {
+      "epoch": 1.1924933974008447,
+      "grad_norm": 0.17122626304626465,
+      "learning_rate": 1.9643727866593015e-05,
+      "loss": 1.2404,
+      "step": 3203
+    },
+    {
+      "epoch": 1.1928657025514537,
+      "grad_norm": 0.1687421053647995,
+      "learning_rate": 1.9643406287061924e-05,
+      "loss": 1.227,
+      "step": 3204
+    },
+    {
+      "epoch": 1.1932380077020628,
+      "grad_norm": 0.17786358296871185,
+      "learning_rate": 1.964308456509833e-05,
+      "loss": 1.2271,
+      "step": 3205
+    },
+    {
+      "epoch": 1.1936103128526718,
+      "grad_norm": 0.16544969379901886,
+      "learning_rate": 1.964276270070698e-05,
+      "loss": 1.2288,
+      "step": 3206
+    },
+    {
+      "epoch": 1.1939826180032809,
+      "grad_norm": 0.1842457503080368,
+      "learning_rate": 1.964244069389263e-05,
+      "loss": 1.2244,
+      "step": 3207
+    },
+    {
+      "epoch": 1.19435492315389,
+      "grad_norm": 0.16619916260242462,
+      "learning_rate": 1.9642118544660036e-05,
+      "loss": 1.2357,
+      "step": 3208
+    },
+    {
+      "epoch": 1.1947272283044992,
+      "grad_norm": 0.1687583178281784,
+      "learning_rate": 1.9641796253013957e-05,
+      "loss": 1.2182,
+      "step": 3209
+    },
+    {
+      "epoch": 1.1950995334551082,
+      "grad_norm": 0.17150536179542542,
+      "learning_rate": 1.9641473818959153e-05,
+      "loss": 1.237,
+      "step": 3210
+    },
+    {
+      "epoch": 1.1954718386057173,
+      "grad_norm": 0.16135509312152863,
+      "learning_rate": 1.9641151242500383e-05,
+      "loss": 1.2405,
+      "step": 3211
+    },
+    {
+      "epoch": 1.1958441437563263,
+      "grad_norm": 0.17465688288211823,
+      "learning_rate": 1.9640828523642415e-05,
+      "loss": 1.2145,
+      "step": 3212
+    },
+    {
+      "epoch": 1.1962164489069353,
+      "grad_norm": 0.17879217863082886,
+      "learning_rate": 1.9640505662390017e-05,
+      "loss": 1.223,
+      "step": 3213
+    },
+    {
+      "epoch": 1.1965887540575444,
+      "grad_norm": 0.1738523691892624,
+      "learning_rate": 1.9640182658747955e-05,
+      "loss": 1.2235,
+      "step": 3214
+    },
+    {
+      "epoch": 1.1969610592081534,
+      "grad_norm": 0.17595960199832916,
+      "learning_rate": 1.9639859512721e-05,
+      "loss": 1.2419,
+      "step": 3215
+    },
+    {
+      "epoch": 1.1973333643587625,
+      "grad_norm": 0.17298002541065216,
+      "learning_rate": 1.963953622431392e-05,
+      "loss": 1.226,
+      "step": 3216
+    },
+    {
+      "epoch": 1.1977056695093715,
+      "grad_norm": 0.169357568025589,
+      "learning_rate": 1.96392127935315e-05,
+      "loss": 1.2276,
+      "step": 3217
+    },
+    {
+      "epoch": 1.1980779746599808,
+      "grad_norm": 0.17965613305568695,
+      "learning_rate": 1.963888922037851e-05,
+      "loss": 1.2232,
+      "step": 3218
+    },
+    {
+      "epoch": 1.1984502798105898,
+      "grad_norm": 0.16693131625652313,
+      "learning_rate": 1.9638565504859734e-05,
+      "loss": 1.2187,
+      "step": 3219
+    },
+    {
+      "epoch": 1.1988225849611989,
+      "grad_norm": 0.16887417435646057,
+      "learning_rate": 1.9638241646979947e-05,
+      "loss": 1.2175,
+      "step": 3220
+    },
+    {
+      "epoch": 1.199194890111808,
+      "grad_norm": 0.17603495717048645,
+      "learning_rate": 1.9637917646743937e-05,
+      "loss": 1.2293,
+      "step": 3221
+    },
+    {
+      "epoch": 1.199567195262417,
+      "grad_norm": 0.1772565096616745,
+      "learning_rate": 1.963759350415649e-05,
+      "loss": 1.2246,
+      "step": 3222
+    },
+    {
+      "epoch": 1.199939500413026,
+      "grad_norm": 0.16490373015403748,
+      "learning_rate": 1.963726921922239e-05,
+      "loss": 1.2226,
+      "step": 3223
+    },
+    {
+      "epoch": 1.200311805563635,
+      "grad_norm": 0.17405438423156738,
+      "learning_rate": 1.963694479194643e-05,
+      "loss": 1.2418,
+      "step": 3224
+    },
+    {
+      "epoch": 1.200684110714244,
+      "grad_norm": 0.1792418658733368,
+      "learning_rate": 1.9636620222333398e-05,
+      "loss": 1.2288,
+      "step": 3225
+    },
+    {
+      "epoch": 1.2010564158648531,
+      "grad_norm": 0.17188262939453125,
+      "learning_rate": 1.963629551038809e-05,
+      "loss": 1.2226,
+      "step": 3226
+    },
+    {
+      "epoch": 1.2014287210154624,
+      "grad_norm": 0.1634385883808136,
+      "learning_rate": 1.9635970656115303e-05,
+      "loss": 1.2126,
+      "step": 3227
+    },
+    {
+      "epoch": 1.2018010261660714,
+      "grad_norm": 0.17879730463027954,
+      "learning_rate": 1.9635645659519835e-05,
+      "loss": 1.226,
+      "step": 3228
+    },
+    {
+      "epoch": 1.2021733313166805,
+      "grad_norm": 0.19251345098018646,
+      "learning_rate": 1.9635320520606483e-05,
+      "loss": 1.228,
+      "step": 3229
+    },
+    {
+      "epoch": 1.2025456364672895,
+      "grad_norm": 0.16735433042049408,
+      "learning_rate": 1.9634995239380056e-05,
+      "loss": 1.2525,
+      "step": 3230
+    },
+    {
+      "epoch": 1.2029179416178986,
+      "grad_norm": 0.17983882129192352,
+      "learning_rate": 1.9634669815845352e-05,
+      "loss": 1.2331,
+      "step": 3231
+    },
+    {
+      "epoch": 1.2032902467685076,
+      "grad_norm": 0.17625576257705688,
+      "learning_rate": 1.9634344250007175e-05,
+      "loss": 1.2252,
+      "step": 3232
+    },
+    {
+      "epoch": 1.2036625519191166,
+      "grad_norm": 0.1711379885673523,
+      "learning_rate": 1.9634018541870342e-05,
+      "loss": 1.2268,
+      "step": 3233
+    },
+    {
+      "epoch": 1.2040348570697257,
+      "grad_norm": 0.17250944674015045,
+      "learning_rate": 1.9633692691439662e-05,
+      "loss": 1.214,
+      "step": 3234
+    },
+    {
+      "epoch": 1.2044071622203347,
+      "grad_norm": 0.17136184871196747,
+      "learning_rate": 1.963336669871994e-05,
+      "loss": 1.2315,
+      "step": 3235
+    },
+    {
+      "epoch": 1.204779467370944,
+      "grad_norm": 0.17495617270469666,
+      "learning_rate": 1.9633040563716e-05,
+      "loss": 1.2198,
+      "step": 3236
+    },
+    {
+      "epoch": 1.205151772521553,
+      "grad_norm": 0.16628341376781464,
+      "learning_rate": 1.9632714286432656e-05,
+      "loss": 1.2302,
+      "step": 3237
+    },
+    {
+      "epoch": 1.205524077672162,
+      "grad_norm": 0.17815084755420685,
+      "learning_rate": 1.963238786687472e-05,
+      "loss": 1.2172,
+      "step": 3238
+    },
+    {
+      "epoch": 1.2058963828227711,
+      "grad_norm": 0.18150483071804047,
+      "learning_rate": 1.9632061305047028e-05,
+      "loss": 1.2399,
+      "step": 3239
+    },
+    {
+      "epoch": 1.2062686879733802,
+      "grad_norm": 0.16711880266666412,
+      "learning_rate": 1.9631734600954392e-05,
+      "loss": 1.2162,
+      "step": 3240
+    },
+    {
+      "epoch": 1.2066409931239892,
+      "grad_norm": 0.16697241365909576,
+      "learning_rate": 1.963140775460164e-05,
+      "loss": 1.2246,
+      "step": 3241
+    },
+    {
+      "epoch": 1.2070132982745982,
+      "grad_norm": 0.17977376282215118,
+      "learning_rate": 1.96310807659936e-05,
+      "loss": 1.2305,
+      "step": 3242
+    },
+    {
+      "epoch": 1.2073856034252075,
+      "grad_norm": 0.1734996736049652,
+      "learning_rate": 1.9630753635135102e-05,
+      "loss": 1.229,
+      "step": 3243
+    },
+    {
+      "epoch": 1.2077579085758163,
+      "grad_norm": 0.17894618213176727,
+      "learning_rate": 1.9630426362030978e-05,
+      "loss": 1.2304,
+      "step": 3244
+    },
+    {
+      "epoch": 1.2081302137264256,
+      "grad_norm": 0.17757487297058105,
+      "learning_rate": 1.963009894668606e-05,
+      "loss": 1.2192,
+      "step": 3245
+    },
+    {
+      "epoch": 1.2085025188770346,
+      "grad_norm": 0.17967943847179413,
+      "learning_rate": 1.9629771389105185e-05,
+      "loss": 1.2279,
+      "step": 3246
+    },
+    {
+      "epoch": 1.2088748240276437,
+      "grad_norm": 0.18884268403053284,
+      "learning_rate": 1.962944368929319e-05,
+      "loss": 1.2305,
+      "step": 3247
+    },
+    {
+      "epoch": 1.2092471291782527,
+      "grad_norm": 0.17970938980579376,
+      "learning_rate": 1.9629115847254916e-05,
+      "loss": 1.2085,
+      "step": 3248
+    },
+    {
+      "epoch": 1.2096194343288618,
+      "grad_norm": 0.16718293726444244,
+      "learning_rate": 1.9628787862995207e-05,
+      "loss": 1.2305,
+      "step": 3249
+    },
+    {
+      "epoch": 1.2099917394794708,
+      "grad_norm": 0.17763985693454742,
+      "learning_rate": 1.9628459736518907e-05,
+      "loss": 1.2271,
+      "step": 3250
+    },
+    {
+      "epoch": 1.2103640446300798,
+      "grad_norm": 0.1723608523607254,
+      "learning_rate": 1.9628131467830856e-05,
+      "loss": 1.2224,
+      "step": 3251
+    },
+    {
+      "epoch": 1.2107363497806891,
+      "grad_norm": 0.16534040868282318,
+      "learning_rate": 1.9627803056935912e-05,
+      "loss": 1.2223,
+      "step": 3252
+    },
+    {
+      "epoch": 1.2111086549312982,
+      "grad_norm": 0.1700848489999771,
+      "learning_rate": 1.9627474503838918e-05,
+      "loss": 1.2291,
+      "step": 3253
+    },
+    {
+      "epoch": 1.2114809600819072,
+      "grad_norm": 0.18612729012966156,
+      "learning_rate": 1.9627145808544733e-05,
+      "loss": 1.2119,
+      "step": 3254
+    },
+    {
+      "epoch": 1.2118532652325162,
+      "grad_norm": 0.16971179842948914,
+      "learning_rate": 1.9626816971058205e-05,
+      "loss": 1.2249,
+      "step": 3255
+    },
+    {
+      "epoch": 1.2122255703831253,
+      "grad_norm": 0.1748899221420288,
+      "learning_rate": 1.9626487991384194e-05,
+      "loss": 1.2136,
+      "step": 3256
+    },
+    {
+      "epoch": 1.2125978755337343,
+      "grad_norm": 0.17875608801841736,
+      "learning_rate": 1.9626158869527564e-05,
+      "loss": 1.2189,
+      "step": 3257
+    },
+    {
+      "epoch": 1.2129701806843434,
+      "grad_norm": 0.17736589908599854,
+      "learning_rate": 1.962582960549317e-05,
+      "loss": 1.2349,
+      "step": 3258
+    },
+    {
+      "epoch": 1.2133424858349524,
+      "grad_norm": 0.17571696639060974,
+      "learning_rate": 1.9625500199285874e-05,
+      "loss": 1.2216,
+      "step": 3259
+    },
+    {
+      "epoch": 1.2137147909855615,
+      "grad_norm": 0.1696479171514511,
+      "learning_rate": 1.9625170650910547e-05,
+      "loss": 1.2141,
+      "step": 3260
+    },
+    {
+      "epoch": 1.2140870961361707,
+      "grad_norm": 0.17148007452487946,
+      "learning_rate": 1.9624840960372053e-05,
+      "loss": 1.2426,
+      "step": 3261
+    },
+    {
+      "epoch": 1.2144594012867798,
+      "grad_norm": 0.17675496637821198,
+      "learning_rate": 1.962451112767526e-05,
+      "loss": 1.2278,
+      "step": 3262
+    },
+    {
+      "epoch": 1.2148317064373888,
+      "grad_norm": 0.17009492218494415,
+      "learning_rate": 1.9624181152825044e-05,
+      "loss": 1.2285,
+      "step": 3263
+    },
+    {
+      "epoch": 1.2152040115879978,
+      "grad_norm": 0.1749381422996521,
+      "learning_rate": 1.9623851035826274e-05,
+      "loss": 1.2182,
+      "step": 3264
+    },
+    {
+      "epoch": 1.2155763167386069,
+      "grad_norm": 0.17395372688770294,
+      "learning_rate": 1.962352077668383e-05,
+      "loss": 1.2366,
+      "step": 3265
+    },
+    {
+      "epoch": 1.215948621889216,
+      "grad_norm": 0.17901504039764404,
+      "learning_rate": 1.962319037540259e-05,
+      "loss": 1.2295,
+      "step": 3266
+    },
+    {
+      "epoch": 1.216320927039825,
+      "grad_norm": 0.17020705342292786,
+      "learning_rate": 1.962285983198743e-05,
+      "loss": 1.239,
+      "step": 3267
+    },
+    {
+      "epoch": 1.216693232190434,
+      "grad_norm": 0.17261864244937897,
+      "learning_rate": 1.9622529146443235e-05,
+      "loss": 1.2242,
+      "step": 3268
+    },
+    {
+      "epoch": 1.217065537341043,
+      "grad_norm": 0.17174561321735382,
+      "learning_rate": 1.9622198318774884e-05,
+      "loss": 1.2337,
+      "step": 3269
+    },
+    {
+      "epoch": 1.2174378424916523,
+      "grad_norm": 0.1685078740119934,
+      "learning_rate": 1.9621867348987273e-05,
+      "loss": 1.2328,
+      "step": 3270
+    },
+    {
+      "epoch": 1.2178101476422614,
+      "grad_norm": 0.17111830413341522,
+      "learning_rate": 1.9621536237085285e-05,
+      "loss": 1.2164,
+      "step": 3271
+    },
+    {
+      "epoch": 1.2181824527928704,
+      "grad_norm": 0.1732606589794159,
+      "learning_rate": 1.9621204983073806e-05,
+      "loss": 1.2314,
+      "step": 3272
+    },
+    {
+      "epoch": 1.2185547579434794,
+      "grad_norm": 0.16949288547039032,
+      "learning_rate": 1.9620873586957735e-05,
+      "loss": 1.2231,
+      "step": 3273
+    },
+    {
+      "epoch": 1.2189270630940885,
+      "grad_norm": 0.17130224406719208,
+      "learning_rate": 1.962054204874197e-05,
+      "loss": 1.2315,
+      "step": 3274
+    },
+    {
+      "epoch": 1.2192993682446975,
+      "grad_norm": 0.16252587735652924,
+      "learning_rate": 1.96202103684314e-05,
+      "loss": 1.2077,
+      "step": 3275
+    },
+    {
+      "epoch": 1.2196716733953066,
+      "grad_norm": 0.1738114356994629,
+      "learning_rate": 1.961987854603092e-05,
+      "loss": 1.2344,
+      "step": 3276
+    },
+    {
+      "epoch": 1.2200439785459156,
+      "grad_norm": 0.16991567611694336,
+      "learning_rate": 1.9619546581545445e-05,
+      "loss": 1.2061,
+      "step": 3277
+    },
+    {
+      "epoch": 1.2204162836965247,
+      "grad_norm": 0.17030175030231476,
+      "learning_rate": 1.961921447497987e-05,
+      "loss": 1.2193,
+      "step": 3278
+    },
+    {
+      "epoch": 1.220788588847134,
+      "grad_norm": 0.16493943333625793,
+      "learning_rate": 1.9618882226339096e-05,
+      "loss": 1.2283,
+      "step": 3279
+    },
+    {
+      "epoch": 1.221160893997743,
+      "grad_norm": 0.17286229133605957,
+      "learning_rate": 1.961854983562804e-05,
+      "loss": 1.2392,
+      "step": 3280
+    },
+    {
+      "epoch": 1.221533199148352,
+      "grad_norm": 0.1747228503227234,
+      "learning_rate": 1.9618217302851607e-05,
+      "loss": 1.2216,
+      "step": 3281
+    },
+    {
+      "epoch": 1.221905504298961,
+      "grad_norm": 0.17101289331912994,
+      "learning_rate": 1.9617884628014707e-05,
+      "loss": 1.2203,
+      "step": 3282
+    },
+    {
+      "epoch": 1.22227780944957,
+      "grad_norm": 0.178693488240242,
+      "learning_rate": 1.9617551811122253e-05,
+      "loss": 1.2378,
+      "step": 3283
+    },
+    {
+      "epoch": 1.2226501146001791,
+      "grad_norm": 0.17429909110069275,
+      "learning_rate": 1.9617218852179162e-05,
+      "loss": 1.2383,
+      "step": 3284
+    },
+    {
+      "epoch": 1.2230224197507882,
+      "grad_norm": 0.17724157869815826,
+      "learning_rate": 1.9616885751190354e-05,
+      "loss": 1.2176,
+      "step": 3285
+    },
+    {
+      "epoch": 1.2233947249013972,
+      "grad_norm": 0.1739131063222885,
+      "learning_rate": 1.9616552508160747e-05,
+      "loss": 1.2302,
+      "step": 3286
+    },
+    {
+      "epoch": 1.2237670300520063,
+      "grad_norm": 0.1718931496143341,
+      "learning_rate": 1.9616219123095265e-05,
+      "loss": 1.2336,
+      "step": 3287
+    },
+    {
+      "epoch": 1.2241393352026155,
+      "grad_norm": 0.17403414845466614,
+      "learning_rate": 1.9615885595998825e-05,
+      "loss": 1.2193,
+      "step": 3288
+    },
+    {
+      "epoch": 1.2245116403532246,
+      "grad_norm": 0.16218788921833038,
+      "learning_rate": 1.9615551926876358e-05,
+      "loss": 1.2239,
+      "step": 3289
+    },
+    {
+      "epoch": 1.2248839455038336,
+      "grad_norm": 0.17476886510849,
+      "learning_rate": 1.9615218115732796e-05,
+      "loss": 1.2332,
+      "step": 3290
+    },
+    {
+      "epoch": 1.2252562506544427,
+      "grad_norm": 0.18384511768817902,
+      "learning_rate": 1.9614884162573067e-05,
+      "loss": 1.2243,
+      "step": 3291
+    },
+    {
+      "epoch": 1.2256285558050517,
+      "grad_norm": 0.174949049949646,
+      "learning_rate": 1.96145500674021e-05,
+      "loss": 1.2188,
+      "step": 3292
+    },
+    {
+      "epoch": 1.2260008609556607,
+      "grad_norm": 0.18955911695957184,
+      "learning_rate": 1.9614215830224832e-05,
+      "loss": 1.2269,
+      "step": 3293
+    },
+    {
+      "epoch": 1.2263731661062698,
+      "grad_norm": 0.17960596084594727,
+      "learning_rate": 1.96138814510462e-05,
+      "loss": 1.2313,
+      "step": 3294
+    },
+    {
+      "epoch": 1.2267454712568788,
+      "grad_norm": 0.18217206001281738,
+      "learning_rate": 1.961354692987114e-05,
+      "loss": 1.2345,
+      "step": 3295
+    },
+    {
+      "epoch": 1.2271177764074879,
+      "grad_norm": 0.1756083369255066,
+      "learning_rate": 1.9613212266704597e-05,
+      "loss": 1.2195,
+      "step": 3296
+    },
+    {
+      "epoch": 1.2274900815580971,
+      "grad_norm": 0.16914978623390198,
+      "learning_rate": 1.9612877461551516e-05,
+      "loss": 1.2131,
+      "step": 3297
+    },
+    {
+      "epoch": 1.2278623867087062,
+      "grad_norm": 0.1753888875246048,
+      "learning_rate": 1.9612542514416835e-05,
+      "loss": 1.24,
+      "step": 3298
+    },
+    {
+      "epoch": 1.2282346918593152,
+      "grad_norm": 0.17672811448574066,
+      "learning_rate": 1.96122074253055e-05,
+      "loss": 1.222,
+      "step": 3299
+    },
+    {
+      "epoch": 1.2286069970099243,
+      "grad_norm": 0.17020326852798462,
+      "learning_rate": 1.961187219422247e-05,
+      "loss": 1.2209,
+      "step": 3300
+    },
+    {
+      "epoch": 1.2289793021605333,
+      "grad_norm": 0.17728908360004425,
+      "learning_rate": 1.961153682117269e-05,
+      "loss": 1.2284,
+      "step": 3301
+    },
+    {
+      "epoch": 1.2293516073111423,
+      "grad_norm": 0.18490946292877197,
+      "learning_rate": 1.9611201306161115e-05,
+      "loss": 1.221,
+      "step": 3302
+    },
+    {
+      "epoch": 1.2297239124617514,
+      "grad_norm": 0.17374534904956818,
+      "learning_rate": 1.9610865649192695e-05,
+      "loss": 1.2393,
+      "step": 3303
+    },
+    {
+      "epoch": 1.2300962176123607,
+      "grad_norm": 0.17610304057598114,
+      "learning_rate": 1.96105298502724e-05,
+      "loss": 1.2163,
+      "step": 3304
+    },
+    {
+      "epoch": 1.2304685227629695,
+      "grad_norm": 0.17196281254291534,
+      "learning_rate": 1.961019390940518e-05,
+      "loss": 1.2256,
+      "step": 3305
+    },
+    {
+      "epoch": 1.2308408279135787,
+      "grad_norm": 0.16787520051002502,
+      "learning_rate": 1.9609857826595996e-05,
+      "loss": 1.2373,
+      "step": 3306
+    },
+    {
+      "epoch": 1.2312131330641878,
+      "grad_norm": 0.17314410209655762,
+      "learning_rate": 1.9609521601849815e-05,
+      "loss": 1.2213,
+      "step": 3307
+    },
+    {
+      "epoch": 1.2315854382147968,
+      "grad_norm": 0.1651289165019989,
+      "learning_rate": 1.9609185235171604e-05,
+      "loss": 1.2163,
+      "step": 3308
+    },
+    {
+      "epoch": 1.2319577433654059,
+      "grad_norm": 0.15858665108680725,
+      "learning_rate": 1.9608848726566328e-05,
+      "loss": 1.21,
+      "step": 3309
+    },
+    {
+      "epoch": 1.232330048516015,
+      "grad_norm": 0.1707463413476944,
+      "learning_rate": 1.9608512076038964e-05,
+      "loss": 1.216,
+      "step": 3310
+    },
+    {
+      "epoch": 1.232702353666624,
+      "grad_norm": 0.16724033653736115,
+      "learning_rate": 1.9608175283594476e-05,
+      "loss": 1.2289,
+      "step": 3311
+    },
+    {
+      "epoch": 1.233074658817233,
+      "grad_norm": 0.17127692699432373,
+      "learning_rate": 1.9607838349237847e-05,
+      "loss": 1.2233,
+      "step": 3312
+    },
+    {
+      "epoch": 1.2334469639678423,
+      "grad_norm": 0.17231450974941254,
+      "learning_rate": 1.9607501272974044e-05,
+      "loss": 1.2362,
+      "step": 3313
+    },
+    {
+      "epoch": 1.2338192691184513,
+      "grad_norm": 0.1658644676208496,
+      "learning_rate": 1.960716405480805e-05,
+      "loss": 1.2242,
+      "step": 3314
+    },
+    {
+      "epoch": 1.2341915742690603,
+      "grad_norm": 0.1748226135969162,
+      "learning_rate": 1.9606826694744847e-05,
+      "loss": 1.2307,
+      "step": 3315
+    },
+    {
+      "epoch": 1.2345638794196694,
+      "grad_norm": 0.1692144274711609,
+      "learning_rate": 1.9606489192789418e-05,
+      "loss": 1.2219,
+      "step": 3316
+    },
+    {
+      "epoch": 1.2349361845702784,
+      "grad_norm": 0.1659345030784607,
+      "learning_rate": 1.9606151548946744e-05,
+      "loss": 1.225,
+      "step": 3317
+    },
+    {
+      "epoch": 1.2353084897208875,
+      "grad_norm": 0.17193017899990082,
+      "learning_rate": 1.9605813763221816e-05,
+      "loss": 1.2318,
+      "step": 3318
+    },
+    {
+      "epoch": 1.2356807948714965,
+      "grad_norm": 0.17747105658054352,
+      "learning_rate": 1.960547583561962e-05,
+      "loss": 1.2413,
+      "step": 3319
+    },
+    {
+      "epoch": 1.2360531000221056,
+      "grad_norm": 0.1722479611635208,
+      "learning_rate": 1.9605137766145154e-05,
+      "loss": 1.2255,
+      "step": 3320
+    },
+    {
+      "epoch": 1.2364254051727146,
+      "grad_norm": 0.16848038136959076,
+      "learning_rate": 1.96047995548034e-05,
+      "loss": 1.2384,
+      "step": 3321
+    },
+    {
+      "epoch": 1.2367977103233239,
+      "grad_norm": 0.16798731684684753,
+      "learning_rate": 1.960446120159936e-05,
+      "loss": 1.222,
+      "step": 3322
+    },
+    {
+      "epoch": 1.237170015473933,
+      "grad_norm": 0.1701812744140625,
+      "learning_rate": 1.9604122706538033e-05,
+      "loss": 1.2218,
+      "step": 3323
+    },
+    {
+      "epoch": 1.237542320624542,
+      "grad_norm": 0.1743394136428833,
+      "learning_rate": 1.9603784069624417e-05,
+      "loss": 1.2381,
+      "step": 3324
+    },
+    {
+      "epoch": 1.237914625775151,
+      "grad_norm": 0.16606199741363525,
+      "learning_rate": 1.960344529086351e-05,
+      "loss": 1.2241,
+      "step": 3325
+    },
+    {
+      "epoch": 1.23828693092576,
+      "grad_norm": 0.17570117115974426,
+      "learning_rate": 1.960310637026032e-05,
+      "loss": 1.2356,
+      "step": 3326
+    },
+    {
+      "epoch": 1.238659236076369,
+      "grad_norm": 0.17124217748641968,
+      "learning_rate": 1.960276730781985e-05,
+      "loss": 1.2422,
+      "step": 3327
+    },
+    {
+      "epoch": 1.2390315412269781,
+      "grad_norm": 0.21202170848846436,
+      "learning_rate": 1.9602428103547112e-05,
+      "loss": 1.2334,
+      "step": 3328
+    },
+    {
+      "epoch": 1.2394038463775872,
+      "grad_norm": 0.17141880095005035,
+      "learning_rate": 1.9602088757447114e-05,
+      "loss": 1.2254,
+      "step": 3329
+    },
+    {
+      "epoch": 1.2397761515281962,
+      "grad_norm": 0.18297582864761353,
+      "learning_rate": 1.9601749269524867e-05,
+      "loss": 1.2451,
+      "step": 3330
+    },
+    {
+      "epoch": 1.2401484566788055,
+      "grad_norm": 0.1688864827156067,
+      "learning_rate": 1.9601409639785384e-05,
+      "loss": 1.2194,
+      "step": 3331
+    },
+    {
+      "epoch": 1.2405207618294145,
+      "grad_norm": 0.1724138855934143,
+      "learning_rate": 1.9601069868233687e-05,
+      "loss": 1.2278,
+      "step": 3332
+    },
+    {
+      "epoch": 1.2408930669800236,
+      "grad_norm": 0.174929678440094,
+      "learning_rate": 1.9600729954874786e-05,
+      "loss": 1.2341,
+      "step": 3333
+    },
+    {
+      "epoch": 1.2412653721306326,
+      "grad_norm": 0.17739325761795044,
+      "learning_rate": 1.9600389899713707e-05,
+      "loss": 1.2258,
+      "step": 3334
+    },
+    {
+      "epoch": 1.2416376772812416,
+      "grad_norm": 0.1830214411020279,
+      "learning_rate": 1.9600049702755473e-05,
+      "loss": 1.2313,
+      "step": 3335
+    },
+    {
+      "epoch": 1.2420099824318507,
+      "grad_norm": 0.16760197281837463,
+      "learning_rate": 1.9599709364005107e-05,
+      "loss": 1.2235,
+      "step": 3336
+    },
+    {
+      "epoch": 1.2423822875824597,
+      "grad_norm": 0.16950567066669464,
+      "learning_rate": 1.959936888346764e-05,
+      "loss": 1.2124,
+      "step": 3337
+    },
+    {
+      "epoch": 1.2427545927330688,
+      "grad_norm": 0.1735774725675583,
+      "learning_rate": 1.959902826114809e-05,
+      "loss": 1.213,
+      "step": 3338
+    },
+    {
+      "epoch": 1.2431268978836778,
+      "grad_norm": 0.17555302381515503,
+      "learning_rate": 1.9598687497051497e-05,
+      "loss": 1.2144,
+      "step": 3339
+    },
+    {
+      "epoch": 1.243499203034287,
+      "grad_norm": 0.17770478129386902,
+      "learning_rate": 1.9598346591182896e-05,
+      "loss": 1.2089,
+      "step": 3340
+    },
+    {
+      "epoch": 1.2438715081848961,
+      "grad_norm": 0.16483311355113983,
+      "learning_rate": 1.9598005543547315e-05,
+      "loss": 1.2113,
+      "step": 3341
+    },
+    {
+      "epoch": 1.2442438133355052,
+      "grad_norm": 0.17812606692314148,
+      "learning_rate": 1.9597664354149793e-05,
+      "loss": 1.2189,
+      "step": 3342
+    },
+    {
+      "epoch": 1.2446161184861142,
+      "grad_norm": 0.17400455474853516,
+      "learning_rate": 1.9597323022995375e-05,
+      "loss": 1.2183,
+      "step": 3343
+    },
+    {
+      "epoch": 1.2449884236367232,
+      "grad_norm": 0.17530007660388947,
+      "learning_rate": 1.95969815500891e-05,
+      "loss": 1.2222,
+      "step": 3344
+    },
+    {
+      "epoch": 1.2453607287873323,
+      "grad_norm": 0.16525647044181824,
+      "learning_rate": 1.9596639935436005e-05,
+      "loss": 1.2257,
+      "step": 3345
+    },
+    {
+      "epoch": 1.2457330339379413,
+      "grad_norm": 0.1811496466398239,
+      "learning_rate": 1.959629817904114e-05,
+      "loss": 1.2295,
+      "step": 3346
+    },
+    {
+      "epoch": 1.2461053390885504,
+      "grad_norm": 0.16356612741947174,
+      "learning_rate": 1.9595956280909552e-05,
+      "loss": 1.2251,
+      "step": 3347
+    },
+    {
+      "epoch": 1.2464776442391594,
+      "grad_norm": 0.1709991693496704,
+      "learning_rate": 1.9595614241046296e-05,
+      "loss": 1.2053,
+      "step": 3348
+    },
+    {
+      "epoch": 1.2468499493897687,
+      "grad_norm": 0.16833190619945526,
+      "learning_rate": 1.959527205945642e-05,
+      "loss": 1.2216,
+      "step": 3349
+    },
+    {
+      "epoch": 1.2472222545403777,
+      "grad_norm": 0.1702100783586502,
+      "learning_rate": 1.9594929736144978e-05,
+      "loss": 1.2207,
+      "step": 3350
+    },
+    {
+      "epoch": 1.2475945596909868,
+      "grad_norm": 0.1710902601480484,
+      "learning_rate": 1.9594587271117023e-05,
+      "loss": 1.2166,
+      "step": 3351
+    },
+    {
+      "epoch": 1.2479668648415958,
+      "grad_norm": 0.17266355454921722,
+      "learning_rate": 1.9594244664377617e-05,
+      "loss": 1.2489,
+      "step": 3352
+    },
+    {
+      "epoch": 1.2483391699922048,
+      "grad_norm": 0.16496752202510834,
+      "learning_rate": 1.959390191593182e-05,
+      "loss": 1.2185,
+      "step": 3353
+    },
+    {
+      "epoch": 1.2487114751428139,
+      "grad_norm": 0.16642600297927856,
+      "learning_rate": 1.9593559025784692e-05,
+      "loss": 1.2326,
+      "step": 3354
+    },
+    {
+      "epoch": 1.249083780293423,
+      "grad_norm": 0.17854921519756317,
+      "learning_rate": 1.95932159939413e-05,
+      "loss": 1.2324,
+      "step": 3355
+    },
+    {
+      "epoch": 1.249456085444032,
+      "grad_norm": 0.17069010436534882,
+      "learning_rate": 1.959287282040671e-05,
+      "loss": 1.2088,
+      "step": 3356
+    },
+    {
+      "epoch": 1.249828390594641,
+      "grad_norm": 0.18059507012367249,
+      "learning_rate": 1.9592529505185993e-05,
+      "loss": 1.2264,
+      "step": 3357
+    },
+    {
+      "epoch": 1.2502006957452503,
+      "grad_norm": 0.173674538731575,
+      "learning_rate": 1.9592186048284216e-05,
+      "loss": 1.2263,
+      "step": 3358
+    },
+    {
+      "epoch": 1.2505730008958593,
+      "grad_norm": 0.16623902320861816,
+      "learning_rate": 1.9591842449706454e-05,
+      "loss": 1.2264,
+      "step": 3359
+    },
+    {
+      "epoch": 1.2509453060464684,
+      "grad_norm": 0.1817052662372589,
+      "learning_rate": 1.9591498709457776e-05,
+      "loss": 1.2062,
+      "step": 3360
+    },
+    {
+      "epoch": 1.2513176111970774,
+      "grad_norm": 0.17939196527004242,
+      "learning_rate": 1.959115482754327e-05,
+      "loss": 1.2464,
+      "step": 3361
+    },
+    {
+      "epoch": 1.2516899163476864,
+      "grad_norm": 0.17410658299922943,
+      "learning_rate": 1.9590810803968003e-05,
+      "loss": 1.2175,
+      "step": 3362
+    },
+    {
+      "epoch": 1.2520622214982955,
+      "grad_norm": 0.19902339577674866,
+      "learning_rate": 1.9590466638737068e-05,
+      "loss": 1.2195,
+      "step": 3363
+    },
+    {
+      "epoch": 1.2524345266489045,
+      "grad_norm": 0.1737726330757141,
+      "learning_rate": 1.9590122331855543e-05,
+      "loss": 1.2097,
+      "step": 3364
+    },
+    {
+      "epoch": 1.2528068317995138,
+      "grad_norm": 0.17546404898166656,
+      "learning_rate": 1.9589777883328506e-05,
+      "loss": 1.2249,
+      "step": 3365
+    },
+    {
+      "epoch": 1.2531791369501226,
+      "grad_norm": 0.1736469715833664,
+      "learning_rate": 1.9589433293161057e-05,
+      "loss": 1.2135,
+      "step": 3366
+    },
+    {
+      "epoch": 1.2535514421007319,
+      "grad_norm": 0.17716355621814728,
+      "learning_rate": 1.9589088561358278e-05,
+      "loss": 1.2203,
+      "step": 3367
+    },
+    {
+      "epoch": 1.253923747251341,
+      "grad_norm": 0.18464305996894836,
+      "learning_rate": 1.9588743687925264e-05,
+      "loss": 1.2134,
+      "step": 3368
+    },
+    {
+      "epoch": 1.25429605240195,
+      "grad_norm": 0.17414681613445282,
+      "learning_rate": 1.9588398672867108e-05,
+      "loss": 1.2212,
+      "step": 3369
+    },
+    {
+      "epoch": 1.254668357552559,
+      "grad_norm": 0.17853331565856934,
+      "learning_rate": 1.9588053516188906e-05,
+      "loss": 1.2398,
+      "step": 3370
+    },
+    {
+      "epoch": 1.255040662703168,
+      "grad_norm": 0.17003083229064941,
+      "learning_rate": 1.958770821789575e-05,
+      "loss": 1.2234,
+      "step": 3371
+    },
+    {
+      "epoch": 1.255412967853777,
+      "grad_norm": 0.18236912786960602,
+      "learning_rate": 1.9587362777992753e-05,
+      "loss": 1.2246,
+      "step": 3372
+    },
+    {
+      "epoch": 1.2557852730043861,
+      "grad_norm": 0.18175008893013,
+      "learning_rate": 1.9587017196485007e-05,
+      "loss": 1.2313,
+      "step": 3373
+    },
+    {
+      "epoch": 1.2561575781549954,
+      "grad_norm": 0.18152983486652374,
+      "learning_rate": 1.9586671473377614e-05,
+      "loss": 1.222,
+      "step": 3374
+    },
+    {
+      "epoch": 1.2565298833056042,
+      "grad_norm": 0.1859092116355896,
+      "learning_rate": 1.9586325608675688e-05,
+      "loss": 1.231,
+      "step": 3375
+    },
+    {
+      "epoch": 1.2569021884562135,
+      "grad_norm": 0.19084875285625458,
+      "learning_rate": 1.9585979602384334e-05,
+      "loss": 1.2234,
+      "step": 3376
+    },
+    {
+      "epoch": 1.2572744936068225,
+      "grad_norm": 0.18165820837020874,
+      "learning_rate": 1.9585633454508665e-05,
+      "loss": 1.225,
+      "step": 3377
+    },
+    {
+      "epoch": 1.2576467987574316,
+      "grad_norm": 0.16506415605545044,
+      "learning_rate": 1.958528716505379e-05,
+      "loss": 1.2329,
+      "step": 3378
+    },
+    {
+      "epoch": 1.2580191039080406,
+      "grad_norm": 0.17367565631866455,
+      "learning_rate": 1.9584940734024826e-05,
+      "loss": 1.2263,
+      "step": 3379
+    },
+    {
+      "epoch": 1.2583914090586497,
+      "grad_norm": 0.16884395480155945,
+      "learning_rate": 1.9584594161426888e-05,
+      "loss": 1.2087,
+      "step": 3380
+    },
+    {
+      "epoch": 1.2587637142092587,
+      "grad_norm": 0.18630245327949524,
+      "learning_rate": 1.9584247447265095e-05,
+      "loss": 1.2279,
+      "step": 3381
+    },
+    {
+      "epoch": 1.2591360193598677,
+      "grad_norm": 0.16664794087409973,
+      "learning_rate": 1.958390059154457e-05,
+      "loss": 1.2317,
+      "step": 3382
+    },
+    {
+      "epoch": 1.259508324510477,
+      "grad_norm": 0.17144736647605896,
+      "learning_rate": 1.9583553594270433e-05,
+      "loss": 1.2157,
+      "step": 3383
+    },
+    {
+      "epoch": 1.259880629661086,
+      "grad_norm": 0.17621219158172607,
+      "learning_rate": 1.9583206455447812e-05,
+      "loss": 1.2179,
+      "step": 3384
+    },
+    {
+      "epoch": 1.260252934811695,
+      "grad_norm": 0.17258892953395844,
+      "learning_rate": 1.9582859175081835e-05,
+      "loss": 1.2349,
+      "step": 3385
+    },
+    {
+      "epoch": 1.2606252399623041,
+      "grad_norm": 0.16975384950637817,
+      "learning_rate": 1.9582511753177625e-05,
+      "loss": 1.2155,
+      "step": 3386
+    },
+    {
+      "epoch": 1.2609975451129132,
+      "grad_norm": 0.171031653881073,
+      "learning_rate": 1.9582164189740322e-05,
+      "loss": 1.2121,
+      "step": 3387
+    },
+    {
+      "epoch": 1.2613698502635222,
+      "grad_norm": 0.16449801623821259,
+      "learning_rate": 1.9581816484775055e-05,
+      "loss": 1.2154,
+      "step": 3388
+    },
+    {
+      "epoch": 1.2617421554141313,
+      "grad_norm": 0.17565850913524628,
+      "learning_rate": 1.9581468638286954e-05,
+      "loss": 1.2266,
+      "step": 3389
+    },
+    {
+      "epoch": 1.2621144605647403,
+      "grad_norm": 0.1752316802740097,
+      "learning_rate": 1.9581120650281166e-05,
+      "loss": 1.2237,
+      "step": 3390
+    },
+    {
+      "epoch": 1.2624867657153493,
+      "grad_norm": 0.16294018924236298,
+      "learning_rate": 1.958077252076283e-05,
+      "loss": 1.2161,
+      "step": 3391
+    },
+    {
+      "epoch": 1.2628590708659586,
+      "grad_norm": 0.16601049900054932,
+      "learning_rate": 1.9580424249737085e-05,
+      "loss": 1.2202,
+      "step": 3392
+    },
+    {
+      "epoch": 1.2632313760165677,
+      "grad_norm": 0.17357154190540314,
+      "learning_rate": 1.958007583720907e-05,
+      "loss": 1.2107,
+      "step": 3393
+    },
+    {
+      "epoch": 1.2636036811671767,
+      "grad_norm": 0.17299290001392365,
+      "learning_rate": 1.957972728318394e-05,
+      "loss": 1.2297,
+      "step": 3394
+    },
+    {
+      "epoch": 1.2639759863177857,
+      "grad_norm": 0.1687079221010208,
+      "learning_rate": 1.9579378587666838e-05,
+      "loss": 1.2087,
+      "step": 3395
+    },
+    {
+      "epoch": 1.2643482914683948,
+      "grad_norm": 0.16745884716510773,
+      "learning_rate": 1.9579029750662918e-05,
+      "loss": 1.2154,
+      "step": 3396
+    },
+    {
+      "epoch": 1.2647205966190038,
+      "grad_norm": 0.17740416526794434,
+      "learning_rate": 1.9578680772177327e-05,
+      "loss": 1.2304,
+      "step": 3397
+    },
+    {
+      "epoch": 1.2650929017696129,
+      "grad_norm": 0.16838476061820984,
+      "learning_rate": 1.9578331652215224e-05,
+      "loss": 1.2315,
+      "step": 3398
+    },
+    {
+      "epoch": 1.265465206920222,
+      "grad_norm": 0.1673501580953598,
+      "learning_rate": 1.9577982390781766e-05,
+      "loss": 1.2072,
+      "step": 3399
+    },
+    {
+      "epoch": 1.265837512070831,
+      "grad_norm": 0.1591964215040207,
+      "learning_rate": 1.9577632987882103e-05,
+      "loss": 1.2121,
+      "step": 3400
+    },
+    {
+      "epoch": 1.2662098172214402,
+      "grad_norm": 0.1804029494524002,
+      "learning_rate": 1.9577283443521403e-05,
+      "loss": 1.225,
+      "step": 3401
+    },
+    {
+      "epoch": 1.2665821223720493,
+      "grad_norm": 0.17372415959835052,
+      "learning_rate": 1.957693375770483e-05,
+      "loss": 1.212,
+      "step": 3402
+    },
+    {
+      "epoch": 1.2669544275226583,
+      "grad_norm": 0.17061223089694977,
+      "learning_rate": 1.9576583930437546e-05,
+      "loss": 1.2165,
+      "step": 3403
+    },
+    {
+      "epoch": 1.2673267326732673,
+      "grad_norm": 0.19941537082195282,
+      "learning_rate": 1.9576233961724716e-05,
+      "loss": 1.2322,
+      "step": 3404
+    },
+    {
+      "epoch": 1.2676990378238764,
+      "grad_norm": 0.1921522617340088,
+      "learning_rate": 1.9575883851571516e-05,
+      "loss": 1.2069,
+      "step": 3405
+    },
+    {
+      "epoch": 1.2680713429744854,
+      "grad_norm": 0.17495955526828766,
+      "learning_rate": 1.957553359998311e-05,
+      "loss": 1.227,
+      "step": 3406
+    },
+    {
+      "epoch": 1.2684436481250945,
+      "grad_norm": 0.18259906768798828,
+      "learning_rate": 1.9575183206964673e-05,
+      "loss": 1.2292,
+      "step": 3407
+    },
+    {
+      "epoch": 1.2688159532757037,
+      "grad_norm": 0.16548015177249908,
+      "learning_rate": 1.9574832672521384e-05,
+      "loss": 1.209,
+      "step": 3408
+    },
+    {
+      "epoch": 1.2691882584263126,
+      "grad_norm": 0.17932672798633575,
+      "learning_rate": 1.9574481996658412e-05,
+      "loss": 1.2247,
+      "step": 3409
+    },
+    {
+      "epoch": 1.2695605635769218,
+      "grad_norm": 0.16864344477653503,
+      "learning_rate": 1.9574131179380945e-05,
+      "loss": 1.209,
+      "step": 3410
+    },
+    {
+      "epoch": 1.2699328687275309,
+      "grad_norm": 0.16725954413414001,
+      "learning_rate": 1.957378022069416e-05,
+      "loss": 1.2094,
+      "step": 3411
+    },
+    {
+      "epoch": 1.27030517387814,
+      "grad_norm": 0.17758184671401978,
+      "learning_rate": 1.9573429120603245e-05,
+      "loss": 1.2177,
+      "step": 3412
+    },
+    {
+      "epoch": 1.270677479028749,
+      "grad_norm": 0.17803123593330383,
+      "learning_rate": 1.957307787911338e-05,
+      "loss": 1.2244,
+      "step": 3413
+    },
+    {
+      "epoch": 1.271049784179358,
+      "grad_norm": 0.1861790120601654,
+      "learning_rate": 1.9572726496229754e-05,
+      "loss": 1.2335,
+      "step": 3414
+    },
+    {
+      "epoch": 1.271422089329967,
+      "grad_norm": 0.17091241478919983,
+      "learning_rate": 1.9572374971957562e-05,
+      "loss": 1.2198,
+      "step": 3415
+    },
+    {
+      "epoch": 1.271794394480576,
+      "grad_norm": 0.16536393761634827,
+      "learning_rate": 1.957202330630199e-05,
+      "loss": 1.2227,
+      "step": 3416
+    },
+    {
+      "epoch": 1.2721666996311853,
+      "grad_norm": 0.1737341731786728,
+      "learning_rate": 1.9571671499268238e-05,
+      "loss": 1.2294,
+      "step": 3417
+    },
+    {
+      "epoch": 1.2725390047817942,
+      "grad_norm": 0.16454610228538513,
+      "learning_rate": 1.9571319550861493e-05,
+      "loss": 1.2099,
+      "step": 3418
+    },
+    {
+      "epoch": 1.2729113099324034,
+      "grad_norm": 0.17201176285743713,
+      "learning_rate": 1.957096746108696e-05,
+      "loss": 1.23,
+      "step": 3419
+    },
+    {
+      "epoch": 1.2732836150830125,
+      "grad_norm": 0.16498921811580658,
+      "learning_rate": 1.9570615229949844e-05,
+      "loss": 1.2081,
+      "step": 3420
+    },
+    {
+      "epoch": 1.2736559202336215,
+      "grad_norm": 0.16812345385551453,
+      "learning_rate": 1.9570262857455336e-05,
+      "loss": 1.2207,
+      "step": 3421
+    },
+    {
+      "epoch": 1.2740282253842306,
+      "grad_norm": 0.17763420939445496,
+      "learning_rate": 1.956991034360865e-05,
+      "loss": 1.233,
+      "step": 3422
+    },
+    {
+      "epoch": 1.2744005305348396,
+      "grad_norm": 0.18109717965126038,
+      "learning_rate": 1.9569557688414985e-05,
+      "loss": 1.2304,
+      "step": 3423
+    },
+    {
+      "epoch": 1.2747728356854486,
+      "grad_norm": 0.1638157218694687,
+      "learning_rate": 1.9569204891879554e-05,
+      "loss": 1.2216,
+      "step": 3424
+    },
+    {
+      "epoch": 1.2751451408360577,
+      "grad_norm": 0.17872877418994904,
+      "learning_rate": 1.956885195400757e-05,
+      "loss": 1.215,
+      "step": 3425
+    },
+    {
+      "epoch": 1.275517445986667,
+      "grad_norm": 0.1729113757610321,
+      "learning_rate": 1.956849887480424e-05,
+      "loss": 1.2073,
+      "step": 3426
+    },
+    {
+      "epoch": 1.2758897511372758,
+      "grad_norm": 0.16412028670310974,
+      "learning_rate": 1.9568145654274787e-05,
+      "loss": 1.2321,
+      "step": 3427
+    },
+    {
+      "epoch": 1.276262056287885,
+      "grad_norm": 0.17767532169818878,
+      "learning_rate": 1.9567792292424417e-05,
+      "loss": 1.2291,
+      "step": 3428
+    },
+    {
+      "epoch": 1.276634361438494,
+      "grad_norm": 0.17381222546100616,
+      "learning_rate": 1.956743878925836e-05,
+      "loss": 1.2315,
+      "step": 3429
+    },
+    {
+      "epoch": 1.2770066665891031,
+      "grad_norm": 0.17282582819461823,
+      "learning_rate": 1.956708514478183e-05,
+      "loss": 1.2208,
+      "step": 3430
+    },
+    {
+      "epoch": 1.2773789717397122,
+      "grad_norm": 0.16768918931484222,
+      "learning_rate": 1.9566731359000056e-05,
+      "loss": 1.2107,
+      "step": 3431
+    },
+    {
+      "epoch": 1.2777512768903212,
+      "grad_norm": 0.1659182608127594,
+      "learning_rate": 1.9566377431918258e-05,
+      "loss": 1.216,
+      "step": 3432
+    },
+    {
+      "epoch": 1.2781235820409302,
+      "grad_norm": 0.1740458458662033,
+      "learning_rate": 1.9566023363541664e-05,
+      "loss": 1.2357,
+      "step": 3433
+    },
+    {
+      "epoch": 1.2784958871915393,
+      "grad_norm": 0.167833611369133,
+      "learning_rate": 1.956566915387551e-05,
+      "loss": 1.2286,
+      "step": 3434
+    },
+    {
+      "epoch": 1.2788681923421485,
+      "grad_norm": 0.17230744659900665,
+      "learning_rate": 1.9565314802925017e-05,
+      "loss": 1.2251,
+      "step": 3435
+    },
+    {
+      "epoch": 1.2792404974927574,
+      "grad_norm": 0.16119259595870972,
+      "learning_rate": 1.9564960310695426e-05,
+      "loss": 1.2184,
+      "step": 3436
+    },
+    {
+      "epoch": 1.2796128026433666,
+      "grad_norm": 0.15957672894001007,
+      "learning_rate": 1.9564605677191975e-05,
+      "loss": 1.2133,
+      "step": 3437
+    },
+    {
+      "epoch": 1.2799851077939757,
+      "grad_norm": 0.16447031497955322,
+      "learning_rate": 1.9564250902419895e-05,
+      "loss": 1.2355,
+      "step": 3438
+    },
+    {
+      "epoch": 1.2803574129445847,
+      "grad_norm": 0.17678163945674896,
+      "learning_rate": 1.956389598638443e-05,
+      "loss": 1.2152,
+      "step": 3439
+    },
+    {
+      "epoch": 1.2807297180951938,
+      "grad_norm": 0.18602001667022705,
+      "learning_rate": 1.9563540929090825e-05,
+      "loss": 1.2069,
+      "step": 3440
+    },
+    {
+      "epoch": 1.2811020232458028,
+      "grad_norm": 0.16143512725830078,
+      "learning_rate": 1.9563185730544316e-05,
+      "loss": 1.2164,
+      "step": 3441
+    },
+    {
+      "epoch": 1.2814743283964118,
+      "grad_norm": 0.17867274582386017,
+      "learning_rate": 1.9562830390750157e-05,
+      "loss": 1.2228,
+      "step": 3442
+    },
+    {
+      "epoch": 1.2818466335470209,
+      "grad_norm": 0.16659069061279297,
+      "learning_rate": 1.9562474909713592e-05,
+      "loss": 1.2173,
+      "step": 3443
+    },
+    {
+      "epoch": 1.2822189386976302,
+      "grad_norm": 0.17568208277225494,
+      "learning_rate": 1.9562119287439874e-05,
+      "loss": 1.2199,
+      "step": 3444
+    },
+    {
+      "epoch": 1.2825912438482392,
+      "grad_norm": 0.18032683432102203,
+      "learning_rate": 1.956176352393425e-05,
+      "loss": 1.2246,
+      "step": 3445
+    },
+    {
+      "epoch": 1.2829635489988482,
+      "grad_norm": 0.1706835776567459,
+      "learning_rate": 1.956140761920198e-05,
+      "loss": 1.2079,
+      "step": 3446
+    },
+    {
+      "epoch": 1.2833358541494573,
+      "grad_norm": 0.17287270724773407,
+      "learning_rate": 1.9561051573248325e-05,
+      "loss": 1.2145,
+      "step": 3447
+    },
+    {
+      "epoch": 1.2837081593000663,
+      "grad_norm": 0.1678660660982132,
+      "learning_rate": 1.9560695386078536e-05,
+      "loss": 1.219,
+      "step": 3448
+    },
+    {
+      "epoch": 1.2840804644506754,
+      "grad_norm": 0.17408515512943268,
+      "learning_rate": 1.956033905769787e-05,
+      "loss": 1.2178,
+      "step": 3449
+    },
+    {
+      "epoch": 1.2844527696012844,
+      "grad_norm": 0.1688593178987503,
+      "learning_rate": 1.9559982588111604e-05,
+      "loss": 1.2167,
+      "step": 3450
+    },
+    {
+      "epoch": 1.2848250747518934,
+      "grad_norm": 0.16934001445770264,
+      "learning_rate": 1.9559625977324992e-05,
+      "loss": 1.2089,
+      "step": 3451
+    },
+    {
+      "epoch": 1.2851973799025025,
+      "grad_norm": 0.17669974267482758,
+      "learning_rate": 1.9559269225343303e-05,
+      "loss": 1.2276,
+      "step": 3452
+    },
+    {
+      "epoch": 1.2855696850531118,
+      "grad_norm": 0.18074657022953033,
+      "learning_rate": 1.955891233217181e-05,
+      "loss": 1.2208,
+      "step": 3453
+    },
+    {
+      "epoch": 1.2859419902037208,
+      "grad_norm": 0.16438442468643188,
+      "learning_rate": 1.955855529781578e-05,
+      "loss": 1.2096,
+      "step": 3454
+    },
+    {
+      "epoch": 1.2863142953543298,
+      "grad_norm": 0.20270679891109467,
+      "learning_rate": 1.9558198122280488e-05,
+      "loss": 1.2206,
+      "step": 3455
+    },
+    {
+      "epoch": 1.2866866005049389,
+      "grad_norm": 0.18252705037593842,
+      "learning_rate": 1.955784080557121e-05,
+      "loss": 1.2188,
+      "step": 3456
+    },
+    {
+      "epoch": 1.287058905655548,
+      "grad_norm": 0.17641377449035645,
+      "learning_rate": 1.9557483347693226e-05,
+      "loss": 1.2156,
+      "step": 3457
+    },
+    {
+      "epoch": 1.287431210806157,
+      "grad_norm": 0.16303670406341553,
+      "learning_rate": 1.955712574865181e-05,
+      "loss": 1.2378,
+      "step": 3458
+    },
+    {
+      "epoch": 1.287803515956766,
+      "grad_norm": 0.20156008005142212,
+      "learning_rate": 1.9556768008452245e-05,
+      "loss": 1.2284,
+      "step": 3459
+    },
+    {
+      "epoch": 1.288175821107375,
+      "grad_norm": 0.16495651006698608,
+      "learning_rate": 1.9556410127099817e-05,
+      "loss": 1.2097,
+      "step": 3460
+    },
+    {
+      "epoch": 1.288548126257984,
+      "grad_norm": 0.1750713586807251,
+      "learning_rate": 1.9556052104599813e-05,
+      "loss": 1.2175,
+      "step": 3461
+    },
+    {
+      "epoch": 1.2889204314085934,
+      "grad_norm": 0.17808473110198975,
+      "learning_rate": 1.9555693940957518e-05,
+      "loss": 1.2354,
+      "step": 3462
+    },
+    {
+      "epoch": 1.2892927365592024,
+      "grad_norm": 0.1678575724363327,
+      "learning_rate": 1.955533563617822e-05,
+      "loss": 1.2293,
+      "step": 3463
+    },
+    {
+      "epoch": 1.2896650417098114,
+      "grad_norm": 0.18347376585006714,
+      "learning_rate": 1.955497719026722e-05,
+      "loss": 1.2276,
+      "step": 3464
+    },
+    {
+      "epoch": 1.2900373468604205,
+      "grad_norm": 0.17216694355010986,
+      "learning_rate": 1.9554618603229804e-05,
+      "loss": 1.2365,
+      "step": 3465
+    },
+    {
+      "epoch": 1.2904096520110295,
+      "grad_norm": 0.16860246658325195,
+      "learning_rate": 1.9554259875071274e-05,
+      "loss": 1.2087,
+      "step": 3466
+    },
+    {
+      "epoch": 1.2907819571616386,
+      "grad_norm": 0.1744735687971115,
+      "learning_rate": 1.955390100579692e-05,
+      "loss": 1.2064,
+      "step": 3467
+    },
+    {
+      "epoch": 1.2911542623122476,
+      "grad_norm": 0.16828221082687378,
+      "learning_rate": 1.955354199541205e-05,
+      "loss": 1.2235,
+      "step": 3468
+    },
+    {
+      "epoch": 1.2915265674628569,
+      "grad_norm": 0.19916963577270508,
+      "learning_rate": 1.9553182843921963e-05,
+      "loss": 1.2269,
+      "step": 3469
+    },
+    {
+      "epoch": 1.2918988726134657,
+      "grad_norm": 0.17060431838035583,
+      "learning_rate": 1.9552823551331966e-05,
+      "loss": 1.2064,
+      "step": 3470
+    },
+    {
+      "epoch": 1.292271177764075,
+      "grad_norm": 0.17713354527950287,
+      "learning_rate": 1.9552464117647365e-05,
+      "loss": 1.2306,
+      "step": 3471
+    },
+    {
+      "epoch": 1.292643482914684,
+      "grad_norm": 0.17107143998146057,
+      "learning_rate": 1.955210454287347e-05,
+      "loss": 1.2413,
+      "step": 3472
+    },
+    {
+      "epoch": 1.293015788065293,
+      "grad_norm": 0.17678503692150116,
+      "learning_rate": 1.955174482701559e-05,
+      "loss": 1.2145,
+      "step": 3473
+    },
+    {
+      "epoch": 1.293388093215902,
+      "grad_norm": 0.17001782357692719,
+      "learning_rate": 1.955138497007904e-05,
+      "loss": 1.2357,
+      "step": 3474
+    },
+    {
+      "epoch": 1.2937603983665111,
+      "grad_norm": 0.1619400829076767,
+      "learning_rate": 1.9551024972069127e-05,
+      "loss": 1.2237,
+      "step": 3475
+    },
+    {
+      "epoch": 1.2941327035171202,
+      "grad_norm": 0.1685052067041397,
+      "learning_rate": 1.9550664832991178e-05,
+      "loss": 1.2217,
+      "step": 3476
+    },
+    {
+      "epoch": 1.2945050086677292,
+      "grad_norm": 0.1692323237657547,
+      "learning_rate": 1.9550304552850506e-05,
+      "loss": 1.2116,
+      "step": 3477
+    },
+    {
+      "epoch": 1.2948773138183385,
+      "grad_norm": 0.1704968959093094,
+      "learning_rate": 1.954994413165244e-05,
+      "loss": 1.2236,
+      "step": 3478
+    },
+    {
+      "epoch": 1.2952496189689473,
+      "grad_norm": 0.17128580808639526,
+      "learning_rate": 1.9549583569402297e-05,
+      "loss": 1.2362,
+      "step": 3479
+    },
+    {
+      "epoch": 1.2956219241195566,
+      "grad_norm": 0.16995707154273987,
+      "learning_rate": 1.95492228661054e-05,
+      "loss": 1.2178,
+      "step": 3480
+    },
+    {
+      "epoch": 1.2959942292701656,
+      "grad_norm": 0.16950835287570953,
+      "learning_rate": 1.9548862021767084e-05,
+      "loss": 1.2147,
+      "step": 3481
+    },
+    {
+      "epoch": 1.2963665344207747,
+      "grad_norm": 0.1846836656332016,
+      "learning_rate": 1.9548501036392676e-05,
+      "loss": 1.215,
+      "step": 3482
+    },
+    {
+      "epoch": 1.2967388395713837,
+      "grad_norm": 0.16860432922840118,
+      "learning_rate": 1.954813990998751e-05,
+      "loss": 1.2209,
+      "step": 3483
+    },
+    {
+      "epoch": 1.2971111447219927,
+      "grad_norm": 0.162691131234169,
+      "learning_rate": 1.9547778642556913e-05,
+      "loss": 1.1976,
+      "step": 3484
+    },
+    {
+      "epoch": 1.2974834498726018,
+      "grad_norm": 0.17532385885715485,
+      "learning_rate": 1.954741723410622e-05,
+      "loss": 1.226,
+      "step": 3485
+    },
+    {
+      "epoch": 1.2978557550232108,
+      "grad_norm": 0.18503205478191376,
+      "learning_rate": 1.954705568464078e-05,
+      "loss": 1.2223,
+      "step": 3486
+    },
+    {
+      "epoch": 1.29822806017382,
+      "grad_norm": 0.1707049161195755,
+      "learning_rate": 1.9546693994165922e-05,
+      "loss": 1.2179,
+      "step": 3487
+    },
+    {
+      "epoch": 1.298600365324429,
+      "grad_norm": 0.17060478031635284,
+      "learning_rate": 1.9546332162687e-05,
+      "loss": 1.2217,
+      "step": 3488
+    },
+    {
+      "epoch": 1.2989726704750382,
+      "grad_norm": 0.17310044169425964,
+      "learning_rate": 1.9545970190209346e-05,
+      "loss": 1.221,
+      "step": 3489
+    },
+    {
+      "epoch": 1.2993449756256472,
+      "grad_norm": 0.19877584278583527,
+      "learning_rate": 1.9545608076738312e-05,
+      "loss": 1.2391,
+      "step": 3490
+    },
+    {
+      "epoch": 1.2997172807762563,
+      "grad_norm": 0.16506221890449524,
+      "learning_rate": 1.9545245822279243e-05,
+      "loss": 1.2112,
+      "step": 3491
+    },
+    {
+      "epoch": 1.3000895859268653,
+      "grad_norm": 0.18883390724658966,
+      "learning_rate": 1.9544883426837497e-05,
+      "loss": 1.2221,
+      "step": 3492
+    },
+    {
+      "epoch": 1.3004618910774743,
+      "grad_norm": 0.18333663046360016,
+      "learning_rate": 1.954452089041842e-05,
+      "loss": 1.231,
+      "step": 3493
+    },
+    {
+      "epoch": 1.3008341962280834,
+      "grad_norm": 0.17763221263885498,
+      "learning_rate": 1.954415821302737e-05,
+      "loss": 1.2195,
+      "step": 3494
+    },
+    {
+      "epoch": 1.3012065013786924,
+      "grad_norm": 0.1757153868675232,
+      "learning_rate": 1.9543795394669696e-05,
+      "loss": 1.2183,
+      "step": 3495
+    },
+    {
+      "epoch": 1.3015788065293017,
+      "grad_norm": 0.1739683747291565,
+      "learning_rate": 1.954343243535077e-05,
+      "loss": 1.2039,
+      "step": 3496
+    },
+    {
+      "epoch": 1.3019511116799105,
+      "grad_norm": 0.17304451763629913,
+      "learning_rate": 1.9543069335075945e-05,
+      "loss": 1.2128,
+      "step": 3497
+    },
+    {
+      "epoch": 1.3023234168305198,
+      "grad_norm": 0.17375996708869934,
+      "learning_rate": 1.9542706093850585e-05,
+      "loss": 1.2332,
+      "step": 3498
+    },
+    {
+      "epoch": 1.3026957219811288,
+      "grad_norm": 0.18385480344295502,
+      "learning_rate": 1.9542342711680053e-05,
+      "loss": 1.2246,
+      "step": 3499
+    },
+    {
+      "epoch": 1.3030680271317379,
+      "grad_norm": 0.17756296694278717,
+      "learning_rate": 1.9541979188569717e-05,
+      "loss": 1.2084,
+      "step": 3500
+    },
+    {
+      "epoch": 1.3030680271317379,
+      "eval_loss": 1.3167004585266113,
+      "eval_runtime": 16.5219,
+      "eval_samples_per_second": 104.951,
+      "eval_steps_per_second": 5.266,
+      "step": 3500
+    },
+    {
+      "epoch": 1.303440332282347,
+      "grad_norm": 0.18115080893039703,
+      "learning_rate": 1.9541615524524946e-05,
+      "loss": 1.2279,
+      "step": 3501
+    },
+    {
+      "epoch": 1.303812637432956,
+      "grad_norm": 0.18096600472927094,
+      "learning_rate": 1.9541251719551116e-05,
+      "loss": 1.238,
+      "step": 3502
+    },
+    {
+      "epoch": 1.304184942583565,
+      "grad_norm": 0.18809261918067932,
+      "learning_rate": 1.9540887773653594e-05,
+      "loss": 1.2256,
+      "step": 3503
+    },
+    {
+      "epoch": 1.304557247734174,
+      "grad_norm": 0.19462642073631287,
+      "learning_rate": 1.954052368683776e-05,
+      "loss": 1.2191,
+      "step": 3504
+    },
+    {
+      "epoch": 1.3049295528847833,
+      "grad_norm": 0.17324475944042206,
+      "learning_rate": 1.9540159459108992e-05,
+      "loss": 1.218,
+      "step": 3505
+    },
+    {
+      "epoch": 1.3053018580353923,
+      "grad_norm": 0.18626105785369873,
+      "learning_rate": 1.9539795090472665e-05,
+      "loss": 1.2095,
+      "step": 3506
+    },
+    {
+      "epoch": 1.3056741631860014,
+      "grad_norm": 0.17831236124038696,
+      "learning_rate": 1.9539430580934162e-05,
+      "loss": 1.2298,
+      "step": 3507
+    },
+    {
+      "epoch": 1.3060464683366104,
+      "grad_norm": 0.17475959658622742,
+      "learning_rate": 1.953906593049887e-05,
+      "loss": 1.2262,
+      "step": 3508
+    },
+    {
+      "epoch": 1.3064187734872195,
+      "grad_norm": 0.1782606840133667,
+      "learning_rate": 1.9538701139172174e-05,
+      "loss": 1.2196,
+      "step": 3509
+    },
+    {
+      "epoch": 1.3067910786378285,
+      "grad_norm": 0.181804358959198,
+      "learning_rate": 1.9538336206959457e-05,
+      "loss": 1.2171,
+      "step": 3510
+    },
+    {
+      "epoch": 1.3071633837884375,
+      "grad_norm": 0.17738310992717743,
+      "learning_rate": 1.9537971133866116e-05,
+      "loss": 1.2304,
+      "step": 3511
+    },
+    {
+      "epoch": 1.3075356889390466,
+      "grad_norm": 0.16684716939926147,
+      "learning_rate": 1.953760591989754e-05,
+      "loss": 1.203,
+      "step": 3512
+    },
+    {
+      "epoch": 1.3079079940896556,
+      "grad_norm": 0.17668993771076202,
+      "learning_rate": 1.953724056505912e-05,
+      "loss": 1.2144,
+      "step": 3513
+    },
+    {
+      "epoch": 1.308280299240265,
+      "grad_norm": 0.17957065999507904,
+      "learning_rate": 1.953687506935626e-05,
+      "loss": 1.2201,
+      "step": 3514
+    },
+    {
+      "epoch": 1.308652604390874,
+      "grad_norm": 0.16784153878688812,
+      "learning_rate": 1.953650943279435e-05,
+      "loss": 1.2145,
+      "step": 3515
+    },
+    {
+      "epoch": 1.309024909541483,
+      "grad_norm": 0.17085857689380646,
+      "learning_rate": 1.9536143655378795e-05,
+      "loss": 1.2283,
+      "step": 3516
+    },
+    {
+      "epoch": 1.309397214692092,
+      "grad_norm": 0.1773957908153534,
+      "learning_rate": 1.9535777737114997e-05,
+      "loss": 1.2117,
+      "step": 3517
+    },
+    {
+      "epoch": 1.309769519842701,
+      "grad_norm": 0.1848251074552536,
+      "learning_rate": 1.953541167800836e-05,
+      "loss": 1.2149,
+      "step": 3518
+    },
+    {
+      "epoch": 1.3101418249933101,
+      "grad_norm": 0.1742524355649948,
+      "learning_rate": 1.9535045478064293e-05,
+      "loss": 1.2426,
+      "step": 3519
+    },
+    {
+      "epoch": 1.3105141301439192,
+      "grad_norm": 0.1685061901807785,
+      "learning_rate": 1.95346791372882e-05,
+      "loss": 1.2167,
+      "step": 3520
+    },
+    {
+      "epoch": 1.3108864352945282,
+      "grad_norm": 0.15838715434074402,
+      "learning_rate": 1.9534312655685497e-05,
+      "loss": 1.2213,
+      "step": 3521
+    },
+    {
+      "epoch": 1.3112587404451372,
+      "grad_norm": 0.1682320386171341,
+      "learning_rate": 1.9533946033261593e-05,
+      "loss": 1.2267,
+      "step": 3522
+    },
+    {
+      "epoch": 1.3116310455957465,
+      "grad_norm": 0.17389652132987976,
+      "learning_rate": 1.9533579270021904e-05,
+      "loss": 1.2222,
+      "step": 3523
+    },
+    {
+      "epoch": 1.3120033507463555,
+      "grad_norm": 0.17018966376781464,
+      "learning_rate": 1.9533212365971844e-05,
+      "loss": 1.2272,
+      "step": 3524
+    },
+    {
+      "epoch": 1.3123756558969646,
+      "grad_norm": 0.16955389082431793,
+      "learning_rate": 1.953284532111684e-05,
+      "loss": 1.2204,
+      "step": 3525
+    },
+    {
+      "epoch": 1.3127479610475736,
+      "grad_norm": 0.16709072887897491,
+      "learning_rate": 1.9532478135462313e-05,
+      "loss": 1.222,
+      "step": 3526
+    },
+    {
+      "epoch": 1.3131202661981827,
+      "grad_norm": 0.1752871423959732,
+      "learning_rate": 1.9532110809013676e-05,
+      "loss": 1.22,
+      "step": 3527
+    },
+    {
+      "epoch": 1.3134925713487917,
+      "grad_norm": 0.1721472591161728,
+      "learning_rate": 1.953174334177636e-05,
+      "loss": 1.2195,
+      "step": 3528
+    },
+    {
+      "epoch": 1.3138648764994008,
+      "grad_norm": 0.17040152847766876,
+      "learning_rate": 1.9531375733755795e-05,
+      "loss": 1.224,
+      "step": 3529
+    },
+    {
+      "epoch": 1.31423718165001,
+      "grad_norm": 0.17984364926815033,
+      "learning_rate": 1.9531007984957408e-05,
+      "loss": 1.2178,
+      "step": 3530
+    },
+    {
+      "epoch": 1.3146094868006188,
+      "grad_norm": 0.16760718822479248,
+      "learning_rate": 1.953064009538663e-05,
+      "loss": 1.2096,
+      "step": 3531
+    },
+    {
+      "epoch": 1.314981791951228,
+      "grad_norm": 0.1705697625875473,
+      "learning_rate": 1.9530272065048903e-05,
+      "loss": 1.2189,
+      "step": 3532
+    },
+    {
+      "epoch": 1.3153540971018371,
+      "grad_norm": 0.17441794276237488,
+      "learning_rate": 1.9529903893949647e-05,
+      "loss": 1.2218,
+      "step": 3533
+    },
+    {
+      "epoch": 1.3157264022524462,
+      "grad_norm": 0.17177151143550873,
+      "learning_rate": 1.9529535582094315e-05,
+      "loss": 1.219,
+      "step": 3534
+    },
+    {
+      "epoch": 1.3160987074030552,
+      "grad_norm": 0.18386155366897583,
+      "learning_rate": 1.9529167129488335e-05,
+      "loss": 1.2187,
+      "step": 3535
+    },
+    {
+      "epoch": 1.3164710125536643,
+      "grad_norm": 0.17924180626869202,
+      "learning_rate": 1.9528798536137157e-05,
+      "loss": 1.2303,
+      "step": 3536
+    },
+    {
+      "epoch": 1.3168433177042733,
+      "grad_norm": 0.16562217473983765,
+      "learning_rate": 1.9528429802046225e-05,
+      "loss": 1.2155,
+      "step": 3537
+    },
+    {
+      "epoch": 1.3172156228548824,
+      "grad_norm": 0.1981254667043686,
+      "learning_rate": 1.952806092722098e-05,
+      "loss": 1.203,
+      "step": 3538
+    },
+    {
+      "epoch": 1.3175879280054916,
+      "grad_norm": 0.17058147490024567,
+      "learning_rate": 1.9527691911666875e-05,
+      "loss": 1.2058,
+      "step": 3539
+    },
+    {
+      "epoch": 1.3179602331561004,
+      "grad_norm": 0.1780320107936859,
+      "learning_rate": 1.9527322755389355e-05,
+      "loss": 1.2242,
+      "step": 3540
+    },
+    {
+      "epoch": 1.3183325383067097,
+      "grad_norm": 0.17311114072799683,
+      "learning_rate": 1.9526953458393878e-05,
+      "loss": 1.2142,
+      "step": 3541
+    },
+    {
+      "epoch": 1.3187048434573188,
+      "grad_norm": 0.1679919809103012,
+      "learning_rate": 1.9526584020685896e-05,
+      "loss": 1.213,
+      "step": 3542
+    },
+    {
+      "epoch": 1.3190771486079278,
+      "grad_norm": 0.17678605020046234,
+      "learning_rate": 1.9526214442270865e-05,
+      "loss": 1.2275,
+      "step": 3543
+    },
+    {
+      "epoch": 1.3194494537585368,
+      "grad_norm": 0.16861461102962494,
+      "learning_rate": 1.9525844723154246e-05,
+      "loss": 1.2262,
+      "step": 3544
+    },
+    {
+      "epoch": 1.3198217589091459,
+      "grad_norm": 0.18046718835830688,
+      "learning_rate": 1.95254748633415e-05,
+      "loss": 1.2258,
+      "step": 3545
+    },
+    {
+      "epoch": 1.320194064059755,
+      "grad_norm": 0.16833364963531494,
+      "learning_rate": 1.9525104862838085e-05,
+      "loss": 1.2195,
+      "step": 3546
+    },
+    {
+      "epoch": 1.320566369210364,
+      "grad_norm": 0.16752612590789795,
+      "learning_rate": 1.952473472164947e-05,
+      "loss": 1.2227,
+      "step": 3547
+    },
+    {
+      "epoch": 1.3209386743609732,
+      "grad_norm": 0.16177985072135925,
+      "learning_rate": 1.952436443978112e-05,
+      "loss": 1.2036,
+      "step": 3548
+    },
+    {
+      "epoch": 1.321310979511582,
+      "grad_norm": 0.1741490513086319,
+      "learning_rate": 1.9523994017238505e-05,
+      "loss": 1.2292,
+      "step": 3549
+    },
+    {
+      "epoch": 1.3216832846621913,
+      "grad_norm": 0.17161637544631958,
+      "learning_rate": 1.9523623454027095e-05,
+      "loss": 1.2195,
+      "step": 3550
+    },
+    {
+      "epoch": 1.3220555898128004,
+      "grad_norm": 0.16780497133731842,
+      "learning_rate": 1.9523252750152367e-05,
+      "loss": 1.2235,
+      "step": 3551
+    },
+    {
+      "epoch": 1.3224278949634094,
+      "grad_norm": 0.17077326774597168,
+      "learning_rate": 1.9522881905619794e-05,
+      "loss": 1.2144,
+      "step": 3552
+    },
+    {
+      "epoch": 1.3228002001140184,
+      "grad_norm": 0.16986922919750214,
+      "learning_rate": 1.9522510920434853e-05,
+      "loss": 1.2222,
+      "step": 3553
+    },
+    {
+      "epoch": 1.3231725052646275,
+      "grad_norm": 0.17960046231746674,
+      "learning_rate": 1.9522139794603018e-05,
+      "loss": 1.2397,
+      "step": 3554
+    },
+    {
+      "epoch": 1.3235448104152365,
+      "grad_norm": 0.17524504661560059,
+      "learning_rate": 1.9521768528129782e-05,
+      "loss": 1.2136,
+      "step": 3555
+    },
+    {
+      "epoch": 1.3239171155658456,
+      "grad_norm": 0.16311313211917877,
+      "learning_rate": 1.952139712102062e-05,
+      "loss": 1.2182,
+      "step": 3556
+    },
+    {
+      "epoch": 1.3242894207164548,
+      "grad_norm": 0.2031218409538269,
+      "learning_rate": 1.952102557328102e-05,
+      "loss": 1.211,
+      "step": 3557
+    },
+    {
+      "epoch": 1.3246617258670637,
+      "grad_norm": 0.18743805587291718,
+      "learning_rate": 1.952065388491647e-05,
+      "loss": 1.2244,
+      "step": 3558
+    },
+    {
+      "epoch": 1.325034031017673,
+      "grad_norm": 0.17010486125946045,
+      "learning_rate": 1.952028205593246e-05,
+      "loss": 1.2096,
+      "step": 3559
+    },
+    {
+      "epoch": 1.325406336168282,
+      "grad_norm": 0.16542263329029083,
+      "learning_rate": 1.951991008633448e-05,
+      "loss": 1.233,
+      "step": 3560
+    },
+    {
+      "epoch": 1.325778641318891,
+      "grad_norm": 0.1798142045736313,
+      "learning_rate": 1.9519537976128025e-05,
+      "loss": 1.2222,
+      "step": 3561
+    },
+    {
+      "epoch": 1.3261509464695,
+      "grad_norm": 0.16815944015979767,
+      "learning_rate": 1.9519165725318594e-05,
+      "loss": 1.2215,
+      "step": 3562
+    },
+    {
+      "epoch": 1.326523251620109,
+      "grad_norm": 0.16989260911941528,
+      "learning_rate": 1.951879333391168e-05,
+      "loss": 1.2323,
+      "step": 3563
+    },
+    {
+      "epoch": 1.3268955567707181,
+      "grad_norm": 0.17635375261306763,
+      "learning_rate": 1.9518420801912787e-05,
+      "loss": 1.2118,
+      "step": 3564
+    },
+    {
+      "epoch": 1.3272678619213272,
+      "grad_norm": 0.1696915477514267,
+      "learning_rate": 1.951804812932742e-05,
+      "loss": 1.2136,
+      "step": 3565
+    },
+    {
+      "epoch": 1.3276401670719364,
+      "grad_norm": 0.17209488153457642,
+      "learning_rate": 1.9517675316161074e-05,
+      "loss": 1.2209,
+      "step": 3566
+    },
+    {
+      "epoch": 1.3280124722225455,
+      "grad_norm": 0.1687251627445221,
+      "learning_rate": 1.951730236241926e-05,
+      "loss": 1.2153,
+      "step": 3567
+    },
+    {
+      "epoch": 1.3283847773731545,
+      "grad_norm": 0.17375510931015015,
+      "learning_rate": 1.9516929268107492e-05,
+      "loss": 1.2257,
+      "step": 3568
+    },
+    {
+      "epoch": 1.3287570825237636,
+      "grad_norm": 0.17510513961315155,
+      "learning_rate": 1.9516556033231276e-05,
+      "loss": 1.2204,
+      "step": 3569
+    },
+    {
+      "epoch": 1.3291293876743726,
+      "grad_norm": 0.16425535082817078,
+      "learning_rate": 1.9516182657796123e-05,
+      "loss": 1.2278,
+      "step": 3570
+    },
+    {
+      "epoch": 1.3295016928249817,
+      "grad_norm": 0.17641210556030273,
+      "learning_rate": 1.9515809141807547e-05,
+      "loss": 1.2086,
+      "step": 3571
+    },
+    {
+      "epoch": 1.3298739979755907,
+      "grad_norm": 0.1740121841430664,
+      "learning_rate": 1.951543548527107e-05,
+      "loss": 1.2131,
+      "step": 3572
+    },
+    {
+      "epoch": 1.3302463031261997,
+      "grad_norm": 0.16927821934223175,
+      "learning_rate": 1.951506168819221e-05,
+      "loss": 1.2278,
+      "step": 3573
+    },
+    {
+      "epoch": 1.3306186082768088,
+      "grad_norm": 0.17070408165454865,
+      "learning_rate": 1.9514687750576483e-05,
+      "loss": 1.2097,
+      "step": 3574
+    },
+    {
+      "epoch": 1.330990913427418,
+      "grad_norm": 0.17762641608715057,
+      "learning_rate": 1.9514313672429414e-05,
+      "loss": 1.2277,
+      "step": 3575
+    },
+    {
+      "epoch": 1.331363218578027,
+      "grad_norm": 0.1731291264295578,
+      "learning_rate": 1.951393945375653e-05,
+      "loss": 1.2455,
+      "step": 3576
+    },
+    {
+      "epoch": 1.3317355237286361,
+      "grad_norm": 0.17669746279716492,
+      "learning_rate": 1.9513565094563358e-05,
+      "loss": 1.2225,
+      "step": 3577
+    },
+    {
+      "epoch": 1.3321078288792452,
+      "grad_norm": 0.16807028651237488,
+      "learning_rate": 1.9513190594855427e-05,
+      "loss": 1.2223,
+      "step": 3578
+    },
+    {
+      "epoch": 1.3324801340298542,
+      "grad_norm": 0.16439288854599,
+      "learning_rate": 1.9512815954638266e-05,
+      "loss": 1.2235,
+      "step": 3579
+    },
+    {
+      "epoch": 1.3328524391804633,
+      "grad_norm": 0.17088323831558228,
+      "learning_rate": 1.9512441173917415e-05,
+      "loss": 1.2066,
+      "step": 3580
+    },
+    {
+      "epoch": 1.3332247443310723,
+      "grad_norm": 0.1691887080669403,
+      "learning_rate": 1.95120662526984e-05,
+      "loss": 1.2274,
+      "step": 3581
+    },
+    {
+      "epoch": 1.3335970494816813,
+      "grad_norm": 0.17315596342086792,
+      "learning_rate": 1.9511691190986767e-05,
+      "loss": 1.2285,
+      "step": 3582
+    },
+    {
+      "epoch": 1.3339693546322904,
+      "grad_norm": 0.1639329493045807,
+      "learning_rate": 1.9511315988788046e-05,
+      "loss": 1.2236,
+      "step": 3583
+    },
+    {
+      "epoch": 1.3343416597828996,
+      "grad_norm": 0.16537517309188843,
+      "learning_rate": 1.951094064610779e-05,
+      "loss": 1.2046,
+      "step": 3584
+    },
+    {
+      "epoch": 1.3347139649335087,
+      "grad_norm": 0.173024520277977,
+      "learning_rate": 1.9510565162951538e-05,
+      "loss": 1.2142,
+      "step": 3585
+    },
+    {
+      "epoch": 1.3350862700841177,
+      "grad_norm": 0.1649821549654007,
+      "learning_rate": 1.9510189539324832e-05,
+      "loss": 1.225,
+      "step": 3586
+    },
+    {
+      "epoch": 1.3354585752347268,
+      "grad_norm": 0.1695072054862976,
+      "learning_rate": 1.9509813775233227e-05,
+      "loss": 1.2062,
+      "step": 3587
+    },
+    {
+      "epoch": 1.3358308803853358,
+      "grad_norm": 0.16297145187854767,
+      "learning_rate": 1.9509437870682268e-05,
+      "loss": 1.2252,
+      "step": 3588
+    },
+    {
+      "epoch": 1.3362031855359449,
+      "grad_norm": 0.1728506088256836,
+      "learning_rate": 1.9509061825677508e-05,
+      "loss": 1.2158,
+      "step": 3589
+    },
+    {
+      "epoch": 1.336575490686554,
+      "grad_norm": 0.16616937518119812,
+      "learning_rate": 1.95086856402245e-05,
+      "loss": 1.2066,
+      "step": 3590
+    },
+    {
+      "epoch": 1.3369477958371632,
+      "grad_norm": 0.16450245678424835,
+      "learning_rate": 1.9508309314328803e-05,
+      "loss": 1.2085,
+      "step": 3591
+    },
+    {
+      "epoch": 1.337320100987772,
+      "grad_norm": 0.17758624255657196,
+      "learning_rate": 1.9507932847995974e-05,
+      "loss": 1.2121,
+      "step": 3592
+    },
+    {
+      "epoch": 1.3376924061383813,
+      "grad_norm": 0.17107117176055908,
+      "learning_rate": 1.9507556241231574e-05,
+      "loss": 1.2223,
+      "step": 3593
+    },
+    {
+      "epoch": 1.3380647112889903,
+      "grad_norm": 0.17417475581169128,
+      "learning_rate": 1.9507179494041166e-05,
+      "loss": 1.2306,
+      "step": 3594
+    },
+    {
+      "epoch": 1.3384370164395993,
+      "grad_norm": 0.16709604859352112,
+      "learning_rate": 1.9506802606430314e-05,
+      "loss": 1.2262,
+      "step": 3595
+    },
+    {
+      "epoch": 1.3388093215902084,
+      "grad_norm": 0.1753034144639969,
+      "learning_rate": 1.950642557840458e-05,
+      "loss": 1.2329,
+      "step": 3596
+    },
+    {
+      "epoch": 1.3391816267408174,
+      "grad_norm": 0.16027675569057465,
+      "learning_rate": 1.950604840996954e-05,
+      "loss": 1.2315,
+      "step": 3597
+    },
+    {
+      "epoch": 1.3395539318914265,
+      "grad_norm": 0.17138755321502686,
+      "learning_rate": 1.950567110113076e-05,
+      "loss": 1.2045,
+      "step": 3598
+    },
+    {
+      "epoch": 1.3399262370420355,
+      "grad_norm": 0.1684652864933014,
+      "learning_rate": 1.9505293651893817e-05,
+      "loss": 1.2353,
+      "step": 3599
+    },
+    {
+      "epoch": 1.3402985421926448,
+      "grad_norm": 0.178995743393898,
+      "learning_rate": 1.9504916062264285e-05,
+      "loss": 1.2311,
+      "step": 3600
+    },
+    {
+      "epoch": 1.3406708473432536,
+      "grad_norm": 0.17695268988609314,
+      "learning_rate": 1.950453833224773e-05,
+      "loss": 1.2295,
+      "step": 3601
+    },
+    {
+      "epoch": 1.3410431524938629,
+      "grad_norm": 0.16872268915176392,
+      "learning_rate": 1.950416046184975e-05,
+      "loss": 1.2149,
+      "step": 3602
+    },
+    {
+      "epoch": 1.341415457644472,
+      "grad_norm": 0.1644962728023529,
+      "learning_rate": 1.9503782451075912e-05,
+      "loss": 1.2012,
+      "step": 3603
+    },
+    {
+      "epoch": 1.341787762795081,
+      "grad_norm": 0.16995151340961456,
+      "learning_rate": 1.9503404299931806e-05,
+      "loss": 1.2115,
+      "step": 3604
+    },
+    {
+      "epoch": 1.34216006794569,
+      "grad_norm": 0.1838231086730957,
+      "learning_rate": 1.950302600842301e-05,
+      "loss": 1.2096,
+      "step": 3605
+    },
+    {
+      "epoch": 1.342532373096299,
+      "grad_norm": 0.1684899926185608,
+      "learning_rate": 1.950264757655512e-05,
+      "loss": 1.2167,
+      "step": 3606
+    },
+    {
+      "epoch": 1.342904678246908,
+      "grad_norm": 0.17310784757137299,
+      "learning_rate": 1.9502269004333722e-05,
+      "loss": 1.2252,
+      "step": 3607
+    },
+    {
+      "epoch": 1.343276983397517,
+      "grad_norm": 0.17696700990200043,
+      "learning_rate": 1.950189029176441e-05,
+      "loss": 1.2212,
+      "step": 3608
+    },
+    {
+      "epoch": 1.3436492885481264,
+      "grad_norm": 0.16353021562099457,
+      "learning_rate": 1.9501511438852767e-05,
+      "loss": 1.2122,
+      "step": 3609
+    },
+    {
+      "epoch": 1.3440215936987352,
+      "grad_norm": 0.1800825595855713,
+      "learning_rate": 1.95011324456044e-05,
+      "loss": 1.222,
+      "step": 3610
+    },
+    {
+      "epoch": 1.3443938988493445,
+      "grad_norm": 0.17191185057163239,
+      "learning_rate": 1.9500753312024904e-05,
+      "loss": 1.2084,
+      "step": 3611
+    },
+    {
+      "epoch": 1.3447662039999535,
+      "grad_norm": 0.16857829689979553,
+      "learning_rate": 1.9500374038119877e-05,
+      "loss": 1.219,
+      "step": 3612
+    },
+    {
+      "epoch": 1.3451385091505625,
+      "grad_norm": 0.16424717009067535,
+      "learning_rate": 1.949999462389492e-05,
+      "loss": 1.2135,
+      "step": 3613
+    },
+    {
+      "epoch": 1.3455108143011716,
+      "grad_norm": 0.1720798909664154,
+      "learning_rate": 1.9499615069355644e-05,
+      "loss": 1.2162,
+      "step": 3614
+    },
+    {
+      "epoch": 1.3458831194517806,
+      "grad_norm": 0.1672721654176712,
+      "learning_rate": 1.9499235374507646e-05,
+      "loss": 1.2197,
+      "step": 3615
+    },
+    {
+      "epoch": 1.3462554246023897,
+      "grad_norm": 0.17355121672153473,
+      "learning_rate": 1.9498855539356534e-05,
+      "loss": 1.2289,
+      "step": 3616
+    },
+    {
+      "epoch": 1.3466277297529987,
+      "grad_norm": 0.17053773999214172,
+      "learning_rate": 1.9498475563907927e-05,
+      "loss": 1.2154,
+      "step": 3617
+    },
+    {
+      "epoch": 1.347000034903608,
+      "grad_norm": 0.17740602791309357,
+      "learning_rate": 1.9498095448167435e-05,
+      "loss": 1.2198,
+      "step": 3618
+    },
+    {
+      "epoch": 1.3473723400542168,
+      "grad_norm": 0.16917094588279724,
+      "learning_rate": 1.949771519214066e-05,
+      "loss": 1.2174,
+      "step": 3619
+    },
+    {
+      "epoch": 1.347744645204826,
+      "grad_norm": 0.20187613368034363,
+      "learning_rate": 1.9497334795833235e-05,
+      "loss": 1.213,
+      "step": 3620
+    },
+    {
+      "epoch": 1.348116950355435,
+      "grad_norm": 0.1762245148420334,
+      "learning_rate": 1.949695425925077e-05,
+      "loss": 1.1968,
+      "step": 3621
+    },
+    {
+      "epoch": 1.3484892555060441,
+      "grad_norm": 0.17733825743198395,
+      "learning_rate": 1.9496573582398884e-05,
+      "loss": 1.2316,
+      "step": 3622
+    },
+    {
+      "epoch": 1.3488615606566532,
+      "grad_norm": 0.16012229025363922,
+      "learning_rate": 1.9496192765283205e-05,
+      "loss": 1.2156,
+      "step": 3623
+    },
+    {
+      "epoch": 1.3492338658072622,
+      "grad_norm": 0.17193962633609772,
+      "learning_rate": 1.9495811807909353e-05,
+      "loss": 1.229,
+      "step": 3624
+    },
+    {
+      "epoch": 1.3496061709578713,
+      "grad_norm": 0.17869403958320618,
+      "learning_rate": 1.9495430710282956e-05,
+      "loss": 1.2101,
+      "step": 3625
+    },
+    {
+      "epoch": 1.3499784761084803,
+      "grad_norm": 0.1690213680267334,
+      "learning_rate": 1.9495049472409644e-05,
+      "loss": 1.2194,
+      "step": 3626
+    },
+    {
+      "epoch": 1.3503507812590896,
+      "grad_norm": 0.16633479297161102,
+      "learning_rate": 1.9494668094295046e-05,
+      "loss": 1.2069,
+      "step": 3627
+    },
+    {
+      "epoch": 1.3507230864096986,
+      "grad_norm": 0.17453880608081818,
+      "learning_rate": 1.94942865759448e-05,
+      "loss": 1.2156,
+      "step": 3628
+    },
+    {
+      "epoch": 1.3510953915603077,
+      "grad_norm": 0.176530122756958,
+      "learning_rate": 1.9493904917364533e-05,
+      "loss": 1.201,
+      "step": 3629
+    },
+    {
+      "epoch": 1.3514676967109167,
+      "grad_norm": 0.1691957712173462,
+      "learning_rate": 1.9493523118559888e-05,
+      "loss": 1.2361,
+      "step": 3630
+    },
+    {
+      "epoch": 1.3518400018615258,
+      "grad_norm": 0.16371867060661316,
+      "learning_rate": 1.9493141179536504e-05,
+      "loss": 1.2186,
+      "step": 3631
+    },
+    {
+      "epoch": 1.3522123070121348,
+      "grad_norm": 0.17232230305671692,
+      "learning_rate": 1.949275910030002e-05,
+      "loss": 1.2164,
+      "step": 3632
+    },
+    {
+      "epoch": 1.3525846121627438,
+      "grad_norm": 0.17034488916397095,
+      "learning_rate": 1.9492376880856075e-05,
+      "loss": 1.2146,
+      "step": 3633
+    },
+    {
+      "epoch": 1.3529569173133529,
+      "grad_norm": 0.16732533276081085,
+      "learning_rate": 1.9491994521210326e-05,
+      "loss": 1.2249,
+      "step": 3634
+    },
+    {
+      "epoch": 1.353329222463962,
+      "grad_norm": 0.16528116166591644,
+      "learning_rate": 1.949161202136841e-05,
+      "loss": 1.2163,
+      "step": 3635
+    },
+    {
+      "epoch": 1.3537015276145712,
+      "grad_norm": 0.1649235486984253,
+      "learning_rate": 1.9491229381335978e-05,
+      "loss": 1.2159,
+      "step": 3636
+    },
+    {
+      "epoch": 1.3540738327651802,
+      "grad_norm": 0.16926534473896027,
+      "learning_rate": 1.9490846601118685e-05,
+      "loss": 1.211,
+      "step": 3637
+    },
+    {
+      "epoch": 1.3544461379157893,
+      "grad_norm": 0.17798763513565063,
+      "learning_rate": 1.9490463680722183e-05,
+      "loss": 1.2198,
+      "step": 3638
+    },
+    {
+      "epoch": 1.3548184430663983,
+      "grad_norm": 0.1720110923051834,
+      "learning_rate": 1.9490080620152133e-05,
+      "loss": 1.2144,
+      "step": 3639
+    },
+    {
+      "epoch": 1.3551907482170074,
+      "grad_norm": 0.16815277934074402,
+      "learning_rate": 1.948969741941418e-05,
+      "loss": 1.2294,
+      "step": 3640
+    },
+    {
+      "epoch": 1.3555630533676164,
+      "grad_norm": 0.1696767359972,
+      "learning_rate": 1.9489314078514e-05,
+      "loss": 1.2161,
+      "step": 3641
+    },
+    {
+      "epoch": 1.3559353585182254,
+      "grad_norm": 0.17275430262088776,
+      "learning_rate": 1.9488930597457242e-05,
+      "loss": 1.2219,
+      "step": 3642
+    },
+    {
+      "epoch": 1.3563076636688347,
+      "grad_norm": 0.17057517170906067,
+      "learning_rate": 1.9488546976249572e-05,
+      "loss": 1.2106,
+      "step": 3643
+    },
+    {
+      "epoch": 1.3566799688194435,
+      "grad_norm": 0.16843107342720032,
+      "learning_rate": 1.9488163214896666e-05,
+      "loss": 1.2274,
+      "step": 3644
+    },
+    {
+      "epoch": 1.3570522739700528,
+      "grad_norm": 0.1690577119588852,
+      "learning_rate": 1.948777931340418e-05,
+      "loss": 1.2187,
+      "step": 3645
+    },
+    {
+      "epoch": 1.3574245791206618,
+      "grad_norm": 0.17254535853862762,
+      "learning_rate": 1.9487395271777787e-05,
+      "loss": 1.2228,
+      "step": 3646
+    },
+    {
+      "epoch": 1.3577968842712709,
+      "grad_norm": 0.17045117914676666,
+      "learning_rate": 1.9487011090023165e-05,
+      "loss": 1.195,
+      "step": 3647
+    },
+    {
+      "epoch": 1.35816918942188,
+      "grad_norm": 0.17185580730438232,
+      "learning_rate": 1.9486626768145986e-05,
+      "loss": 1.2162,
+      "step": 3648
+    },
+    {
+      "epoch": 1.358541494572489,
+      "grad_norm": 0.1752527505159378,
+      "learning_rate": 1.948624230615192e-05,
+      "loss": 1.2336,
+      "step": 3649
+    },
+    {
+      "epoch": 1.358913799723098,
+      "grad_norm": 0.1720474511384964,
+      "learning_rate": 1.9485857704046652e-05,
+      "loss": 1.2268,
+      "step": 3650
+    },
+    {
+      "epoch": 1.359286104873707,
+      "grad_norm": 0.1814902275800705,
+      "learning_rate": 1.9485472961835865e-05,
+      "loss": 1.2179,
+      "step": 3651
+    },
+    {
+      "epoch": 1.3596584100243163,
+      "grad_norm": 0.16393816471099854,
+      "learning_rate": 1.948508807952523e-05,
+      "loss": 1.2179,
+      "step": 3652
+    },
+    {
+      "epoch": 1.3600307151749251,
+      "grad_norm": 0.17267872393131256,
+      "learning_rate": 1.9484703057120444e-05,
+      "loss": 1.2152,
+      "step": 3653
+    },
+    {
+      "epoch": 1.3604030203255344,
+      "grad_norm": 0.1774880588054657,
+      "learning_rate": 1.948431789462719e-05,
+      "loss": 1.2323,
+      "step": 3654
+    },
+    {
+      "epoch": 1.3607753254761434,
+      "grad_norm": 0.17601603269577026,
+      "learning_rate": 1.9483932592051154e-05,
+      "loss": 1.2211,
+      "step": 3655
+    },
+    {
+      "epoch": 1.3611476306267525,
+      "grad_norm": 0.17215831577777863,
+      "learning_rate": 1.948354714939803e-05,
+      "loss": 1.2171,
+      "step": 3656
+    },
+    {
+      "epoch": 1.3615199357773615,
+      "grad_norm": 0.16573548316955566,
+      "learning_rate": 1.948316156667351e-05,
+      "loss": 1.2386,
+      "step": 3657
+    },
+    {
+      "epoch": 1.3618922409279706,
+      "grad_norm": 0.17070265114307404,
+      "learning_rate": 1.9482775843883287e-05,
+      "loss": 1.2306,
+      "step": 3658
+    },
+    {
+      "epoch": 1.3622645460785796,
+      "grad_norm": 0.17669783532619476,
+      "learning_rate": 1.948238998103306e-05,
+      "loss": 1.2186,
+      "step": 3659
+    },
+    {
+      "epoch": 1.3626368512291887,
+      "grad_norm": 0.17972193658351898,
+      "learning_rate": 1.948200397812853e-05,
+      "loss": 1.2208,
+      "step": 3660
+    },
+    {
+      "epoch": 1.363009156379798,
+      "grad_norm": 0.16259367763996124,
+      "learning_rate": 1.9481617835175394e-05,
+      "loss": 1.1995,
+      "step": 3661
+    },
+    {
+      "epoch": 1.3633814615304067,
+      "grad_norm": 0.16560673713684082,
+      "learning_rate": 1.948123155217936e-05,
+      "loss": 1.2124,
+      "step": 3662
+    },
+    {
+      "epoch": 1.363753766681016,
+      "grad_norm": 0.17198215425014496,
+      "learning_rate": 1.948084512914613e-05,
+      "loss": 1.223,
+      "step": 3663
+    },
+    {
+      "epoch": 1.364126071831625,
+      "grad_norm": 0.16872192919254303,
+      "learning_rate": 1.948045856608141e-05,
+      "loss": 1.213,
+      "step": 3664
+    },
+    {
+      "epoch": 1.364498376982234,
+      "grad_norm": 0.17418809235095978,
+      "learning_rate": 1.9480071862990917e-05,
+      "loss": 1.2059,
+      "step": 3665
+    },
+    {
+      "epoch": 1.3648706821328431,
+      "grad_norm": 0.1725587695837021,
+      "learning_rate": 1.9479685019880356e-05,
+      "loss": 1.2234,
+      "step": 3666
+    },
+    {
+      "epoch": 1.3652429872834522,
+      "grad_norm": 0.16984142363071442,
+      "learning_rate": 1.947929803675544e-05,
+      "loss": 1.2106,
+      "step": 3667
+    },
+    {
+      "epoch": 1.3656152924340612,
+      "grad_norm": 0.1837463080883026,
+      "learning_rate": 1.9478910913621888e-05,
+      "loss": 1.2143,
+      "step": 3668
+    },
+    {
+      "epoch": 1.3659875975846703,
+      "grad_norm": 0.1684904843568802,
+      "learning_rate": 1.9478523650485416e-05,
+      "loss": 1.2236,
+      "step": 3669
+    },
+    {
+      "epoch": 1.3663599027352795,
+      "grad_norm": 0.1728762686252594,
+      "learning_rate": 1.9478136247351745e-05,
+      "loss": 1.205,
+      "step": 3670
+    },
+    {
+      "epoch": 1.3667322078858883,
+      "grad_norm": 0.16816040873527527,
+      "learning_rate": 1.9477748704226597e-05,
+      "loss": 1.2231,
+      "step": 3671
+    },
+    {
+      "epoch": 1.3671045130364976,
+      "grad_norm": 0.17051219940185547,
+      "learning_rate": 1.9477361021115695e-05,
+      "loss": 1.2277,
+      "step": 3672
+    },
+    {
+      "epoch": 1.3674768181871066,
+      "grad_norm": 0.16704906523227692,
+      "learning_rate": 1.9476973198024766e-05,
+      "loss": 1.2166,
+      "step": 3673
+    },
+    {
+      "epoch": 1.3678491233377157,
+      "grad_norm": 0.16872693598270416,
+      "learning_rate": 1.9476585234959538e-05,
+      "loss": 1.2136,
+      "step": 3674
+    },
+    {
+      "epoch": 1.3682214284883247,
+      "grad_norm": 0.16978906095027924,
+      "learning_rate": 1.947619713192574e-05,
+      "loss": 1.209,
+      "step": 3675
+    },
+    {
+      "epoch": 1.3685937336389338,
+      "grad_norm": 0.16473500430583954,
+      "learning_rate": 1.9475808888929104e-05,
+      "loss": 1.2284,
+      "step": 3676
+    },
+    {
+      "epoch": 1.3689660387895428,
+      "grad_norm": 0.16459238529205322,
+      "learning_rate": 1.9475420505975366e-05,
+      "loss": 1.2103,
+      "step": 3677
+    },
+    {
+      "epoch": 1.3693383439401519,
+      "grad_norm": 0.1704069972038269,
+      "learning_rate": 1.9475031983070264e-05,
+      "loss": 1.2207,
+      "step": 3678
+    },
+    {
+      "epoch": 1.3697106490907611,
+      "grad_norm": 0.18187867105007172,
+      "learning_rate": 1.9474643320219534e-05,
+      "loss": 1.1893,
+      "step": 3679
+    },
+    {
+      "epoch": 1.3700829542413702,
+      "grad_norm": 0.1664721667766571,
+      "learning_rate": 1.9474254517428912e-05,
+      "loss": 1.23,
+      "step": 3680
+    },
+    {
+      "epoch": 1.3704552593919792,
+      "grad_norm": 0.17463192343711853,
+      "learning_rate": 1.947386557470415e-05,
+      "loss": 1.2172,
+      "step": 3681
+    },
+    {
+      "epoch": 1.3708275645425883,
+      "grad_norm": 0.17818517982959747,
+      "learning_rate": 1.9473476492050984e-05,
+      "loss": 1.2206,
+      "step": 3682
+    },
+    {
+      "epoch": 1.3711998696931973,
+      "grad_norm": 0.1713617891073227,
+      "learning_rate": 1.947308726947517e-05,
+      "loss": 1.2358,
+      "step": 3683
+    },
+    {
+      "epoch": 1.3715721748438063,
+      "grad_norm": 0.16868345439434052,
+      "learning_rate": 1.947269790698245e-05,
+      "loss": 1.2138,
+      "step": 3684
+    },
+    {
+      "epoch": 1.3719444799944154,
+      "grad_norm": 0.17152869701385498,
+      "learning_rate": 1.9472308404578574e-05,
+      "loss": 1.2142,
+      "step": 3685
+    },
+    {
+      "epoch": 1.3723167851450244,
+      "grad_norm": 0.16501954197883606,
+      "learning_rate": 1.9471918762269298e-05,
+      "loss": 1.2098,
+      "step": 3686
+    },
+    {
+      "epoch": 1.3726890902956335,
+      "grad_norm": 0.1698669195175171,
+      "learning_rate": 1.9471528980060378e-05,
+      "loss": 1.2238,
+      "step": 3687
+    },
+    {
+      "epoch": 1.3730613954462427,
+      "grad_norm": 0.17526181042194366,
+      "learning_rate": 1.9471139057957566e-05,
+      "loss": 1.226,
+      "step": 3688
+    },
+    {
+      "epoch": 1.3734337005968518,
+      "grad_norm": 0.1682656854391098,
+      "learning_rate": 1.947074899596663e-05,
+      "loss": 1.2115,
+      "step": 3689
+    },
+    {
+      "epoch": 1.3738060057474608,
+      "grad_norm": 0.16698940098285675,
+      "learning_rate": 1.947035879409332e-05,
+      "loss": 1.1977,
+      "step": 3690
+    },
+    {
+      "epoch": 1.3741783108980699,
+      "grad_norm": 0.17004626989364624,
+      "learning_rate": 1.9469968452343408e-05,
+      "loss": 1.2197,
+      "step": 3691
+    },
+    {
+      "epoch": 1.374550616048679,
+      "grad_norm": 0.17727632820606232,
+      "learning_rate": 1.9469577970722656e-05,
+      "loss": 1.2242,
+      "step": 3692
+    },
+    {
+      "epoch": 1.374922921199288,
+      "grad_norm": 0.16782982647418976,
+      "learning_rate": 1.9469187349236832e-05,
+      "loss": 1.2111,
+      "step": 3693
+    },
+    {
+      "epoch": 1.375295226349897,
+      "grad_norm": 0.17160236835479736,
+      "learning_rate": 1.9468796587891706e-05,
+      "loss": 1.2168,
+      "step": 3694
+    },
+    {
+      "epoch": 1.375667531500506,
+      "grad_norm": 0.16954834759235382,
+      "learning_rate": 1.9468405686693044e-05,
+      "loss": 1.2077,
+      "step": 3695
+    },
+    {
+      "epoch": 1.376039836651115,
+      "grad_norm": 0.168158158659935,
+      "learning_rate": 1.9468014645646628e-05,
+      "loss": 1.2249,
+      "step": 3696
+    },
+    {
+      "epoch": 1.3764121418017243,
+      "grad_norm": 0.16404445469379425,
+      "learning_rate": 1.946762346475823e-05,
+      "loss": 1.2158,
+      "step": 3697
+    },
+    {
+      "epoch": 1.3767844469523334,
+      "grad_norm": 0.1699007749557495,
+      "learning_rate": 1.946723214403363e-05,
+      "loss": 1.2043,
+      "step": 3698
+    },
+    {
+      "epoch": 1.3771567521029424,
+      "grad_norm": 0.17556270956993103,
+      "learning_rate": 1.9466840683478597e-05,
+      "loss": 1.2167,
+      "step": 3699
+    },
+    {
+      "epoch": 1.3775290572535515,
+      "grad_norm": 0.1715071052312851,
+      "learning_rate": 1.9466449083098927e-05,
+      "loss": 1.2228,
+      "step": 3700
+    },
+    {
+      "epoch": 1.3779013624041605,
+      "grad_norm": 0.1757829487323761,
+      "learning_rate": 1.94660573429004e-05,
+      "loss": 1.232,
+      "step": 3701
+    },
+    {
+      "epoch": 1.3782736675547695,
+      "grad_norm": 0.1744741052389145,
+      "learning_rate": 1.9465665462888798e-05,
+      "loss": 1.2026,
+      "step": 3702
+    },
+    {
+      "epoch": 1.3786459727053786,
+      "grad_norm": 0.17323531210422516,
+      "learning_rate": 1.9465273443069908e-05,
+      "loss": 1.2306,
+      "step": 3703
+    },
+    {
+      "epoch": 1.3790182778559879,
+      "grad_norm": 0.16859839856624603,
+      "learning_rate": 1.9464881283449525e-05,
+      "loss": 1.2169,
+      "step": 3704
+    },
+    {
+      "epoch": 1.3793905830065967,
+      "grad_norm": 0.1696728765964508,
+      "learning_rate": 1.9464488984033442e-05,
+      "loss": 1.2084,
+      "step": 3705
+    },
+    {
+      "epoch": 1.379762888157206,
+      "grad_norm": 0.16950103640556335,
+      "learning_rate": 1.946409654482745e-05,
+      "loss": 1.2233,
+      "step": 3706
+    },
+    {
+      "epoch": 1.380135193307815,
+      "grad_norm": 0.17010366916656494,
+      "learning_rate": 1.9463703965837344e-05,
+      "loss": 1.2042,
+      "step": 3707
+    },
+    {
+      "epoch": 1.380507498458424,
+      "grad_norm": 0.16800427436828613,
+      "learning_rate": 1.9463311247068927e-05,
+      "loss": 1.2012,
+      "step": 3708
+    },
+    {
+      "epoch": 1.380879803609033,
+      "grad_norm": 0.16539591550827026,
+      "learning_rate": 1.9462918388527994e-05,
+      "loss": 1.2236,
+      "step": 3709
+    },
+    {
+      "epoch": 1.381252108759642,
+      "grad_norm": 0.17388111352920532,
+      "learning_rate": 1.946252539022035e-05,
+      "loss": 1.2115,
+      "step": 3710
+    },
+    {
+      "epoch": 1.3816244139102511,
+      "grad_norm": 0.17521414160728455,
+      "learning_rate": 1.94621322521518e-05,
+      "loss": 1.2319,
+      "step": 3711
+    },
+    {
+      "epoch": 1.3819967190608602,
+      "grad_norm": 0.16813969612121582,
+      "learning_rate": 1.9461738974328156e-05,
+      "loss": 1.2158,
+      "step": 3712
+    },
+    {
+      "epoch": 1.3823690242114695,
+      "grad_norm": 0.16460061073303223,
+      "learning_rate": 1.9461345556755216e-05,
+      "loss": 1.2084,
+      "step": 3713
+    },
+    {
+      "epoch": 1.3827413293620783,
+      "grad_norm": 0.18938884139060974,
+      "learning_rate": 1.9460951999438798e-05,
+      "loss": 1.2169,
+      "step": 3714
+    },
+    {
+      "epoch": 1.3831136345126875,
+      "grad_norm": 0.19839580357074738,
+      "learning_rate": 1.946055830238471e-05,
+      "loss": 1.221,
+      "step": 3715
+    },
+    {
+      "epoch": 1.3834859396632966,
+      "grad_norm": 0.1847252994775772,
+      "learning_rate": 1.9460164465598773e-05,
+      "loss": 1.21,
+      "step": 3716
+    },
+    {
+      "epoch": 1.3838582448139056,
+      "grad_norm": 0.1856468766927719,
+      "learning_rate": 1.94597704890868e-05,
+      "loss": 1.222,
+      "step": 3717
+    },
+    {
+      "epoch": 1.3842305499645147,
+      "grad_norm": 0.1739731878042221,
+      "learning_rate": 1.945937637285461e-05,
+      "loss": 1.1989,
+      "step": 3718
+    },
+    {
+      "epoch": 1.3846028551151237,
+      "grad_norm": 0.16499097645282745,
+      "learning_rate": 1.945898211690802e-05,
+      "loss": 1.2297,
+      "step": 3719
+    },
+    {
+      "epoch": 1.3849751602657328,
+      "grad_norm": 0.17512036859989166,
+      "learning_rate": 1.9458587721252862e-05,
+      "loss": 1.2238,
+      "step": 3720
+    },
+    {
+      "epoch": 1.3853474654163418,
+      "grad_norm": 0.1747319996356964,
+      "learning_rate": 1.9458193185894957e-05,
+      "loss": 1.214,
+      "step": 3721
+    },
+    {
+      "epoch": 1.385719770566951,
+      "grad_norm": 0.17503361403942108,
+      "learning_rate": 1.9457798510840132e-05,
+      "loss": 1.225,
+      "step": 3722
+    },
+    {
+      "epoch": 1.3860920757175599,
+      "grad_norm": 0.16246047616004944,
+      "learning_rate": 1.9457403696094216e-05,
+      "loss": 1.2121,
+      "step": 3723
+    },
+    {
+      "epoch": 1.3864643808681691,
+      "grad_norm": 0.17145591974258423,
+      "learning_rate": 1.945700874166304e-05,
+      "loss": 1.1996,
+      "step": 3724
+    },
+    {
+      "epoch": 1.3868366860187782,
+      "grad_norm": 0.17188484966754913,
+      "learning_rate": 1.945661364755244e-05,
+      "loss": 1.217,
+      "step": 3725
+    },
+    {
+      "epoch": 1.3872089911693872,
+      "grad_norm": 0.17199209332466125,
+      "learning_rate": 1.945621841376825e-05,
+      "loss": 1.2215,
+      "step": 3726
+    },
+    {
+      "epoch": 1.3875812963199963,
+      "grad_norm": 0.17529122531414032,
+      "learning_rate": 1.945582304031631e-05,
+      "loss": 1.2076,
+      "step": 3727
+    },
+    {
+      "epoch": 1.3879536014706053,
+      "grad_norm": 0.16724155843257904,
+      "learning_rate": 1.945542752720245e-05,
+      "loss": 1.2258,
+      "step": 3728
+    },
+    {
+      "epoch": 1.3883259066212144,
+      "grad_norm": 0.16579319536685944,
+      "learning_rate": 1.9455031874432526e-05,
+      "loss": 1.2206,
+      "step": 3729
+    },
+    {
+      "epoch": 1.3886982117718234,
+      "grad_norm": 0.17617808282375336,
+      "learning_rate": 1.9454636082012373e-05,
+      "loss": 1.2258,
+      "step": 3730
+    },
+    {
+      "epoch": 1.3890705169224327,
+      "grad_norm": 0.17041444778442383,
+      "learning_rate": 1.9454240149947834e-05,
+      "loss": 1.2164,
+      "step": 3731
+    },
+    {
+      "epoch": 1.3894428220730415,
+      "grad_norm": 0.17237654328346252,
+      "learning_rate": 1.9453844078244767e-05,
+      "loss": 1.2238,
+      "step": 3732
+    },
+    {
+      "epoch": 1.3898151272236507,
+      "grad_norm": 0.17883872985839844,
+      "learning_rate": 1.9453447866909013e-05,
+      "loss": 1.2267,
+      "step": 3733
+    },
+    {
+      "epoch": 1.3901874323742598,
+      "grad_norm": 0.1674152910709381,
+      "learning_rate": 1.9453051515946428e-05,
+      "loss": 1.2181,
+      "step": 3734
+    },
+    {
+      "epoch": 1.3905597375248688,
+      "grad_norm": 0.16891363263130188,
+      "learning_rate": 1.9452655025362865e-05,
+      "loss": 1.2301,
+      "step": 3735
+    },
+    {
+      "epoch": 1.3909320426754779,
+      "grad_norm": 0.16902455687522888,
+      "learning_rate": 1.9452258395164182e-05,
+      "loss": 1.2097,
+      "step": 3736
+    },
+    {
+      "epoch": 1.391304347826087,
+      "grad_norm": 0.1644563525915146,
+      "learning_rate": 1.9451861625356235e-05,
+      "loss": 1.2291,
+      "step": 3737
+    },
+    {
+      "epoch": 1.391676652976696,
+      "grad_norm": 0.1637285202741623,
+      "learning_rate": 1.9451464715944885e-05,
+      "loss": 1.2201,
+      "step": 3738
+    },
+    {
+      "epoch": 1.392048958127305,
+      "grad_norm": 0.1716940551996231,
+      "learning_rate": 1.945106766693599e-05,
+      "loss": 1.2151,
+      "step": 3739
+    },
+    {
+      "epoch": 1.3924212632779143,
+      "grad_norm": 0.17560310661792755,
+      "learning_rate": 1.9450670478335424e-05,
+      "loss": 1.2121,
+      "step": 3740
+    },
+    {
+      "epoch": 1.3927935684285233,
+      "grad_norm": 0.164300337433815,
+      "learning_rate": 1.9450273150149047e-05,
+      "loss": 1.2194,
+      "step": 3741
+    },
+    {
+      "epoch": 1.3931658735791324,
+      "grad_norm": 0.17114891111850739,
+      "learning_rate": 1.944987568238273e-05,
+      "loss": 1.201,
+      "step": 3742
+    },
+    {
+      "epoch": 1.3935381787297414,
+      "grad_norm": 0.1734459102153778,
+      "learning_rate": 1.944947807504234e-05,
+      "loss": 1.221,
+      "step": 3743
+    },
+    {
+      "epoch": 1.3939104838803504,
+      "grad_norm": 0.16712883114814758,
+      "learning_rate": 1.944908032813375e-05,
+      "loss": 1.2168,
+      "step": 3744
+    },
+    {
+      "epoch": 1.3942827890309595,
+      "grad_norm": 0.17574042081832886,
+      "learning_rate": 1.944868244166284e-05,
+      "loss": 1.2226,
+      "step": 3745
+    },
+    {
+      "epoch": 1.3946550941815685,
+      "grad_norm": 0.167069673538208,
+      "learning_rate": 1.944828441563548e-05,
+      "loss": 1.1964,
+      "step": 3746
+    },
+    {
+      "epoch": 1.3950273993321776,
+      "grad_norm": 0.1734757423400879,
+      "learning_rate": 1.9447886250057556e-05,
+      "loss": 1.2265,
+      "step": 3747
+    },
+    {
+      "epoch": 1.3953997044827866,
+      "grad_norm": 0.1713225245475769,
+      "learning_rate": 1.944748794493494e-05,
+      "loss": 1.1987,
+      "step": 3748
+    },
+    {
+      "epoch": 1.3957720096333959,
+      "grad_norm": 0.18104128539562225,
+      "learning_rate": 1.9447089500273524e-05,
+      "loss": 1.212,
+      "step": 3749
+    },
+    {
+      "epoch": 1.396144314784005,
+      "grad_norm": 0.17931242287158966,
+      "learning_rate": 1.944669091607919e-05,
+      "loss": 1.218,
+      "step": 3750
+    },
+    {
+      "epoch": 1.396516619934614,
+      "grad_norm": 0.1705280989408493,
+      "learning_rate": 1.944629219235782e-05,
+      "loss": 1.2186,
+      "step": 3751
+    },
+    {
+      "epoch": 1.396888925085223,
+      "grad_norm": 0.1851729452610016,
+      "learning_rate": 1.944589332911531e-05,
+      "loss": 1.2086,
+      "step": 3752
+    },
+    {
+      "epoch": 1.397261230235832,
+      "grad_norm": 0.17898760735988617,
+      "learning_rate": 1.9445494326357548e-05,
+      "loss": 1.2299,
+      "step": 3753
+    },
+    {
+      "epoch": 1.397633535386441,
+      "grad_norm": 0.17653051018714905,
+      "learning_rate": 1.9445095184090428e-05,
+      "loss": 1.2135,
+      "step": 3754
+    },
+    {
+      "epoch": 1.3980058405370501,
+      "grad_norm": 0.17303214967250824,
+      "learning_rate": 1.9444695902319845e-05,
+      "loss": 1.2214,
+      "step": 3755
+    },
+    {
+      "epoch": 1.3983781456876592,
+      "grad_norm": 0.16854806244373322,
+      "learning_rate": 1.9444296481051697e-05,
+      "loss": 1.1992,
+      "step": 3756
+    },
+    {
+      "epoch": 1.3987504508382682,
+      "grad_norm": 0.17003561556339264,
+      "learning_rate": 1.944389692029188e-05,
+      "loss": 1.2143,
+      "step": 3757
+    },
+    {
+      "epoch": 1.3991227559888775,
+      "grad_norm": 0.17145571112632751,
+      "learning_rate": 1.9443497220046298e-05,
+      "loss": 1.2136,
+      "step": 3758
+    },
+    {
+      "epoch": 1.3994950611394865,
+      "grad_norm": 0.1773233562707901,
+      "learning_rate": 1.9443097380320855e-05,
+      "loss": 1.1977,
+      "step": 3759
+    },
+    {
+      "epoch": 1.3998673662900956,
+      "grad_norm": 0.1766585260629654,
+      "learning_rate": 1.944269740112146e-05,
+      "loss": 1.2133,
+      "step": 3760
+    },
+    {
+      "epoch": 1.4002396714407046,
+      "grad_norm": 0.18044057488441467,
+      "learning_rate": 1.9442297282454012e-05,
+      "loss": 1.206,
+      "step": 3761
+    },
+    {
+      "epoch": 1.4006119765913136,
+      "grad_norm": 0.1800515502691269,
+      "learning_rate": 1.9441897024324428e-05,
+      "loss": 1.2115,
+      "step": 3762
+    },
+    {
+      "epoch": 1.4009842817419227,
+      "grad_norm": 0.17421506345272064,
+      "learning_rate": 1.944149662673862e-05,
+      "loss": 1.2186,
+      "step": 3763
+    },
+    {
+      "epoch": 1.4013565868925317,
+      "grad_norm": 0.17194457352161407,
+      "learning_rate": 1.9441096089702495e-05,
+      "loss": 1.2109,
+      "step": 3764
+    },
+    {
+      "epoch": 1.401728892043141,
+      "grad_norm": 0.1739378422498703,
+      "learning_rate": 1.9440695413221975e-05,
+      "loss": 1.2066,
+      "step": 3765
+    },
+    {
+      "epoch": 1.4021011971937498,
+      "grad_norm": 0.17094124853610992,
+      "learning_rate": 1.9440294597302977e-05,
+      "loss": 1.217,
+      "step": 3766
+    },
+    {
+      "epoch": 1.402473502344359,
+      "grad_norm": 0.16742725670337677,
+      "learning_rate": 1.9439893641951418e-05,
+      "loss": 1.2153,
+      "step": 3767
+    },
+    {
+      "epoch": 1.4028458074949681,
+      "grad_norm": 0.17719261348247528,
+      "learning_rate": 1.9439492547173225e-05,
+      "loss": 1.2097,
+      "step": 3768
+    },
+    {
+      "epoch": 1.4032181126455772,
+      "grad_norm": 0.16668783128261566,
+      "learning_rate": 1.9439091312974317e-05,
+      "loss": 1.2193,
+      "step": 3769
+    },
+    {
+      "epoch": 1.4035904177961862,
+      "grad_norm": 0.17551898956298828,
+      "learning_rate": 1.9438689939360627e-05,
+      "loss": 1.2115,
+      "step": 3770
+    },
+    {
+      "epoch": 1.4039627229467952,
+      "grad_norm": 0.18291716277599335,
+      "learning_rate": 1.9438288426338073e-05,
+      "loss": 1.2085,
+      "step": 3771
+    },
+    {
+      "epoch": 1.4043350280974043,
+      "grad_norm": 0.16662460565567017,
+      "learning_rate": 1.9437886773912595e-05,
+      "loss": 1.2078,
+      "step": 3772
+    },
+    {
+      "epoch": 1.4047073332480133,
+      "grad_norm": 0.17178334295749664,
+      "learning_rate": 1.9437484982090122e-05,
+      "loss": 1.2195,
+      "step": 3773
+    },
+    {
+      "epoch": 1.4050796383986226,
+      "grad_norm": 0.17247992753982544,
+      "learning_rate": 1.9437083050876588e-05,
+      "loss": 1.2026,
+      "step": 3774
+    },
+    {
+      "epoch": 1.4054519435492314,
+      "grad_norm": 0.17090709507465363,
+      "learning_rate": 1.943668098027793e-05,
+      "loss": 1.2013,
+      "step": 3775
+    },
+    {
+      "epoch": 1.4058242486998407,
+      "grad_norm": 0.1711760014295578,
+      "learning_rate": 1.9436278770300082e-05,
+      "loss": 1.2116,
+      "step": 3776
+    },
+    {
+      "epoch": 1.4061965538504497,
+      "grad_norm": 0.17857705056667328,
+      "learning_rate": 1.943587642094899e-05,
+      "loss": 1.2237,
+      "step": 3777
+    },
+    {
+      "epoch": 1.4065688590010588,
+      "grad_norm": 0.18631206452846527,
+      "learning_rate": 1.9435473932230597e-05,
+      "loss": 1.2378,
+      "step": 3778
+    },
+    {
+      "epoch": 1.4069411641516678,
+      "grad_norm": 0.209711492061615,
+      "learning_rate": 1.9435071304150846e-05,
+      "loss": 1.2207,
+      "step": 3779
+    },
+    {
+      "epoch": 1.4073134693022769,
+      "grad_norm": 0.2716389298439026,
+      "learning_rate": 1.9434668536715686e-05,
+      "loss": 1.2311,
+      "step": 3780
+    },
+    {
+      "epoch": 1.407685774452886,
+      "grad_norm": 0.2101835310459137,
+      "learning_rate": 1.9434265629931063e-05,
+      "loss": 1.2202,
+      "step": 3781
+    },
+    {
+      "epoch": 1.408058079603495,
+      "grad_norm": 0.1864628791809082,
+      "learning_rate": 1.9433862583802927e-05,
+      "loss": 1.2012,
+      "step": 3782
+    },
+    {
+      "epoch": 1.4084303847541042,
+      "grad_norm": 0.1653883457183838,
+      "learning_rate": 1.9433459398337234e-05,
+      "loss": 1.2191,
+      "step": 3783
+    },
+    {
+      "epoch": 1.408802689904713,
+      "grad_norm": 0.17915792763233185,
+      "learning_rate": 1.9433056073539934e-05,
+      "loss": 1.2218,
+      "step": 3784
+    },
+    {
+      "epoch": 1.4091749950553223,
+      "grad_norm": 0.20036007463932037,
+      "learning_rate": 1.9432652609416993e-05,
+      "loss": 1.2227,
+      "step": 3785
+    },
+    {
+      "epoch": 1.4095473002059313,
+      "grad_norm": 0.18710802495479584,
+      "learning_rate": 1.943224900597436e-05,
+      "loss": 1.2126,
+      "step": 3786
+    },
+    {
+      "epoch": 1.4099196053565404,
+      "grad_norm": 0.18362727761268616,
+      "learning_rate": 1.9431845263218005e-05,
+      "loss": 1.219,
+      "step": 3787
+    },
+    {
+      "epoch": 1.4102919105071494,
+      "grad_norm": 0.18792852759361267,
+      "learning_rate": 1.943144138115389e-05,
+      "loss": 1.2113,
+      "step": 3788
+    },
+    {
+      "epoch": 1.4106642156577585,
+      "grad_norm": 0.1898011863231659,
+      "learning_rate": 1.943103735978797e-05,
+      "loss": 1.2172,
+      "step": 3789
+    },
+    {
+      "epoch": 1.4110365208083675,
+      "grad_norm": 0.1731170117855072,
+      "learning_rate": 1.9430633199126225e-05,
+      "loss": 1.2306,
+      "step": 3790
+    },
+    {
+      "epoch": 1.4114088259589765,
+      "grad_norm": 0.1677362322807312,
+      "learning_rate": 1.9430228899174617e-05,
+      "loss": 1.2239,
+      "step": 3791
+    },
+    {
+      "epoch": 1.4117811311095858,
+      "grad_norm": 0.16974781453609467,
+      "learning_rate": 1.9429824459939125e-05,
+      "loss": 1.2035,
+      "step": 3792
+    },
+    {
+      "epoch": 1.4121534362601946,
+      "grad_norm": 0.19337321817874908,
+      "learning_rate": 1.9429419881425713e-05,
+      "loss": 1.2091,
+      "step": 3793
+    },
+    {
+      "epoch": 1.412525741410804,
+      "grad_norm": 0.1791197508573532,
+      "learning_rate": 1.9429015163640363e-05,
+      "loss": 1.2148,
+      "step": 3794
+    },
+    {
+      "epoch": 1.412898046561413,
+      "grad_norm": 0.17587502300739288,
+      "learning_rate": 1.9428610306589047e-05,
+      "loss": 1.2287,
+      "step": 3795
+    },
+    {
+      "epoch": 1.413270351712022,
+      "grad_norm": 0.17841365933418274,
+      "learning_rate": 1.9428205310277752e-05,
+      "loss": 1.2108,
+      "step": 3796
+    },
+    {
+      "epoch": 1.413642656862631,
+      "grad_norm": 0.17350056767463684,
+      "learning_rate": 1.9427800174712455e-05,
+      "loss": 1.2035,
+      "step": 3797
+    },
+    {
+      "epoch": 1.41401496201324,
+      "grad_norm": 0.1747625321149826,
+      "learning_rate": 1.942739489989914e-05,
+      "loss": 1.2333,
+      "step": 3798
+    },
+    {
+      "epoch": 1.414387267163849,
+      "grad_norm": 0.17848855257034302,
+      "learning_rate": 1.9426989485843796e-05,
+      "loss": 1.2028,
+      "step": 3799
+    },
+    {
+      "epoch": 1.4147595723144581,
+      "grad_norm": 0.17129814624786377,
+      "learning_rate": 1.942658393255241e-05,
+      "loss": 1.228,
+      "step": 3800
+    },
+    {
+      "epoch": 1.4151318774650674,
+      "grad_norm": 0.1709514707326889,
+      "learning_rate": 1.942617824003097e-05,
+      "loss": 1.2262,
+      "step": 3801
+    },
+    {
+      "epoch": 1.4155041826156765,
+      "grad_norm": 0.16819478571414948,
+      "learning_rate": 1.942577240828547e-05,
+      "loss": 1.1978,
+      "step": 3802
+    },
+    {
+      "epoch": 1.4158764877662855,
+      "grad_norm": 0.17637069523334503,
+      "learning_rate": 1.94253664373219e-05,
+      "loss": 1.2166,
+      "step": 3803
+    },
+    {
+      "epoch": 1.4162487929168945,
+      "grad_norm": 0.1750773936510086,
+      "learning_rate": 1.942496032714626e-05,
+      "loss": 1.2239,
+      "step": 3804
+    },
+    {
+      "epoch": 1.4166210980675036,
+      "grad_norm": 0.16642986238002777,
+      "learning_rate": 1.9424554077764548e-05,
+      "loss": 1.2061,
+      "step": 3805
+    },
+    {
+      "epoch": 1.4169934032181126,
+      "grad_norm": 0.16567951440811157,
+      "learning_rate": 1.9424147689182765e-05,
+      "loss": 1.2207,
+      "step": 3806
+    },
+    {
+      "epoch": 1.4173657083687217,
+      "grad_norm": 0.17164906859397888,
+      "learning_rate": 1.942374116140691e-05,
+      "loss": 1.2136,
+      "step": 3807
+    },
+    {
+      "epoch": 1.4177380135193307,
+      "grad_norm": 0.1757207065820694,
+      "learning_rate": 1.942333449444299e-05,
+      "loss": 1.2229,
+      "step": 3808
+    },
+    {
+      "epoch": 1.4181103186699398,
+      "grad_norm": 0.1720336526632309,
+      "learning_rate": 1.9422927688297012e-05,
+      "loss": 1.2173,
+      "step": 3809
+    },
+    {
+      "epoch": 1.418482623820549,
+      "grad_norm": 0.16701281070709229,
+      "learning_rate": 1.9422520742974984e-05,
+      "loss": 1.2145,
+      "step": 3810
+    },
+    {
+      "epoch": 1.418854928971158,
+      "grad_norm": 0.16450679302215576,
+      "learning_rate": 1.9422113658482912e-05,
+      "loss": 1.2042,
+      "step": 3811
+    },
+    {
+      "epoch": 1.419227234121767,
+      "grad_norm": 0.16200435161590576,
+      "learning_rate": 1.942170643482682e-05,
+      "loss": 1.2152,
+      "step": 3812
+    },
+    {
+      "epoch": 1.4195995392723761,
+      "grad_norm": 0.1700204312801361,
+      "learning_rate": 1.9421299072012705e-05,
+      "loss": 1.2209,
+      "step": 3813
+    },
+    {
+      "epoch": 1.4199718444229852,
+      "grad_norm": 0.1691514402627945,
+      "learning_rate": 1.94208915700466e-05,
+      "loss": 1.2218,
+      "step": 3814
+    },
+    {
+      "epoch": 1.4203441495735942,
+      "grad_norm": 0.17180638015270233,
+      "learning_rate": 1.942048392893452e-05,
+      "loss": 1.2112,
+      "step": 3815
+    },
+    {
+      "epoch": 1.4207164547242033,
+      "grad_norm": 0.17067848145961761,
+      "learning_rate": 1.942007614868248e-05,
+      "loss": 1.2182,
+      "step": 3816
+    },
+    {
+      "epoch": 1.4210887598748123,
+      "grad_norm": 0.17079667747020721,
+      "learning_rate": 1.9419668229296507e-05,
+      "loss": 1.2248,
+      "step": 3817
+    },
+    {
+      "epoch": 1.4214610650254214,
+      "grad_norm": 0.16801148653030396,
+      "learning_rate": 1.9419260170782624e-05,
+      "loss": 1.217,
+      "step": 3818
+    },
+    {
+      "epoch": 1.4218333701760306,
+      "grad_norm": 0.16779641807079315,
+      "learning_rate": 1.9418851973146864e-05,
+      "loss": 1.2039,
+      "step": 3819
+    },
+    {
+      "epoch": 1.4222056753266397,
+      "grad_norm": 0.17368653416633606,
+      "learning_rate": 1.941844363639525e-05,
+      "loss": 1.2103,
+      "step": 3820
+    },
+    {
+      "epoch": 1.4225779804772487,
+      "grad_norm": 0.1701452136039734,
+      "learning_rate": 1.9418035160533813e-05,
+      "loss": 1.2069,
+      "step": 3821
+    },
+    {
+      "epoch": 1.4229502856278577,
+      "grad_norm": 0.17342039942741394,
+      "learning_rate": 1.9417626545568588e-05,
+      "loss": 1.222,
+      "step": 3822
+    },
+    {
+      "epoch": 1.4233225907784668,
+      "grad_norm": 0.173239603638649,
+      "learning_rate": 1.941721779150561e-05,
+      "loss": 1.2248,
+      "step": 3823
+    },
+    {
+      "epoch": 1.4236948959290758,
+      "grad_norm": 0.16423793137073517,
+      "learning_rate": 1.9416808898350915e-05,
+      "loss": 1.2084,
+      "step": 3824
+    },
+    {
+      "epoch": 1.4240672010796849,
+      "grad_norm": 0.16744939982891083,
+      "learning_rate": 1.9416399866110545e-05,
+      "loss": 1.2066,
+      "step": 3825
+    },
+    {
+      "epoch": 1.4244395062302941,
+      "grad_norm": 0.17226476967334747,
+      "learning_rate": 1.941599069479054e-05,
+      "loss": 1.2074,
+      "step": 3826
+    },
+    {
+      "epoch": 1.424811811380903,
+      "grad_norm": 0.16691239178180695,
+      "learning_rate": 1.9415581384396944e-05,
+      "loss": 1.2157,
+      "step": 3827
+    },
+    {
+      "epoch": 1.4251841165315122,
+      "grad_norm": 0.16759438812732697,
+      "learning_rate": 1.9415171934935798e-05,
+      "loss": 1.2214,
+      "step": 3828
+    },
+    {
+      "epoch": 1.4255564216821213,
+      "grad_norm": 0.1606573760509491,
+      "learning_rate": 1.9414762346413157e-05,
+      "loss": 1.202,
+      "step": 3829
+    },
+    {
+      "epoch": 1.4259287268327303,
+      "grad_norm": 0.16710183024406433,
+      "learning_rate": 1.9414352618835065e-05,
+      "loss": 1.226,
+      "step": 3830
+    },
+    {
+      "epoch": 1.4263010319833394,
+      "grad_norm": 0.16753865778446198,
+      "learning_rate": 1.9413942752207577e-05,
+      "loss": 1.2131,
+      "step": 3831
+    },
+    {
+      "epoch": 1.4266733371339484,
+      "grad_norm": 0.17130768299102783,
+      "learning_rate": 1.9413532746536744e-05,
+      "loss": 1.2127,
+      "step": 3832
+    },
+    {
+      "epoch": 1.4270456422845574,
+      "grad_norm": 0.1705525517463684,
+      "learning_rate": 1.9413122601828624e-05,
+      "loss": 1.2116,
+      "step": 3833
+    },
+    {
+      "epoch": 1.4274179474351665,
+      "grad_norm": 0.16685357689857483,
+      "learning_rate": 1.941271231808928e-05,
+      "loss": 1.2101,
+      "step": 3834
+    },
+    {
+      "epoch": 1.4277902525857757,
+      "grad_norm": 0.16948597133159637,
+      "learning_rate": 1.9412301895324755e-05,
+      "loss": 1.2146,
+      "step": 3835
+    },
+    {
+      "epoch": 1.4281625577363846,
+      "grad_norm": 0.16796061396598816,
+      "learning_rate": 1.941189133354113e-05,
+      "loss": 1.2168,
+      "step": 3836
+    },
+    {
+      "epoch": 1.4285348628869938,
+      "grad_norm": 0.1637044996023178,
+      "learning_rate": 1.9411480632744454e-05,
+      "loss": 1.2151,
+      "step": 3837
+    },
+    {
+      "epoch": 1.4289071680376029,
+      "grad_norm": 0.16353152692317963,
+      "learning_rate": 1.9411069792940803e-05,
+      "loss": 1.2055,
+      "step": 3838
+    },
+    {
+      "epoch": 1.429279473188212,
+      "grad_norm": 0.17246000468730927,
+      "learning_rate": 1.9410658814136243e-05,
+      "loss": 1.2039,
+      "step": 3839
+    },
+    {
+      "epoch": 1.429651778338821,
+      "grad_norm": 0.17234675586223602,
+      "learning_rate": 1.9410247696336842e-05,
+      "loss": 1.2052,
+      "step": 3840
+    },
+    {
+      "epoch": 1.43002408348943,
+      "grad_norm": 0.1646290272474289,
+      "learning_rate": 1.940983643954867e-05,
+      "loss": 1.203,
+      "step": 3841
+    },
+    {
+      "epoch": 1.430396388640039,
+      "grad_norm": 0.17096149921417236,
+      "learning_rate": 1.9409425043777806e-05,
+      "loss": 1.2062,
+      "step": 3842
+    },
+    {
+      "epoch": 1.430768693790648,
+      "grad_norm": 0.16843508183956146,
+      "learning_rate": 1.9409013509030327e-05,
+      "loss": 1.2088,
+      "step": 3843
+    },
+    {
+      "epoch": 1.4311409989412573,
+      "grad_norm": 0.17243371903896332,
+      "learning_rate": 1.940860183531231e-05,
+      "loss": 1.2183,
+      "step": 3844
+    },
+    {
+      "epoch": 1.4315133040918662,
+      "grad_norm": 0.16187606751918793,
+      "learning_rate": 1.940819002262983e-05,
+      "loss": 1.207,
+      "step": 3845
+    },
+    {
+      "epoch": 1.4318856092424754,
+      "grad_norm": 0.17250527441501617,
+      "learning_rate": 1.9407778070988978e-05,
+      "loss": 1.2233,
+      "step": 3846
+    },
+    {
+      "epoch": 1.4322579143930845,
+      "grad_norm": 0.17004723846912384,
+      "learning_rate": 1.9407365980395833e-05,
+      "loss": 1.2143,
+      "step": 3847
+    },
+    {
+      "epoch": 1.4326302195436935,
+      "grad_norm": 0.16374097764492035,
+      "learning_rate": 1.940695375085648e-05,
+      "loss": 1.2106,
+      "step": 3848
+    },
+    {
+      "epoch": 1.4330025246943026,
+      "grad_norm": 0.17270928621292114,
+      "learning_rate": 1.9406541382377012e-05,
+      "loss": 1.2206,
+      "step": 3849
+    },
+    {
+      "epoch": 1.4333748298449116,
+      "grad_norm": 0.17662329971790314,
+      "learning_rate": 1.940612887496352e-05,
+      "loss": 1.2189,
+      "step": 3850
+    },
+    {
+      "epoch": 1.4337471349955206,
+      "grad_norm": 0.1634707897901535,
+      "learning_rate": 1.9405716228622094e-05,
+      "loss": 1.2006,
+      "step": 3851
+    },
+    {
+      "epoch": 1.4341194401461297,
+      "grad_norm": 0.16489939391613007,
+      "learning_rate": 1.9405303443358827e-05,
+      "loss": 1.2142,
+      "step": 3852
+    },
+    {
+      "epoch": 1.434491745296739,
+      "grad_norm": 0.17095866799354553,
+      "learning_rate": 1.940489051917982e-05,
+      "loss": 1.2122,
+      "step": 3853
+    },
+    {
+      "epoch": 1.4348640504473478,
+      "grad_norm": 0.16927289962768555,
+      "learning_rate": 1.940447745609117e-05,
+      "loss": 1.2124,
+      "step": 3854
+    },
+    {
+      "epoch": 1.435236355597957,
+      "grad_norm": 0.17228350043296814,
+      "learning_rate": 1.940406425409898e-05,
+      "loss": 1.2221,
+      "step": 3855
+    },
+    {
+      "epoch": 1.435608660748566,
+      "grad_norm": 0.1686760038137436,
+      "learning_rate": 1.940365091320935e-05,
+      "loss": 1.2281,
+      "step": 3856
+    },
+    {
+      "epoch": 1.4359809658991751,
+      "grad_norm": 0.16984900832176208,
+      "learning_rate": 1.9403237433428384e-05,
+      "loss": 1.2348,
+      "step": 3857
+    },
+    {
+      "epoch": 1.4363532710497842,
+      "grad_norm": 0.16201676428318024,
+      "learning_rate": 1.9402823814762194e-05,
+      "loss": 1.2072,
+      "step": 3858
+    },
+    {
+      "epoch": 1.4367255762003932,
+      "grad_norm": 0.17006635665893555,
+      "learning_rate": 1.9402410057216886e-05,
+      "loss": 1.2149,
+      "step": 3859
+    },
+    {
+      "epoch": 1.4370978813510022,
+      "grad_norm": 0.17508037388324738,
+      "learning_rate": 1.9401996160798574e-05,
+      "loss": 1.2096,
+      "step": 3860
+    },
+    {
+      "epoch": 1.4374701865016113,
+      "grad_norm": 0.1723199486732483,
+      "learning_rate": 1.9401582125513364e-05,
+      "loss": 1.2175,
+      "step": 3861
+    },
+    {
+      "epoch": 1.4378424916522206,
+      "grad_norm": 0.16605544090270996,
+      "learning_rate": 1.9401167951367375e-05,
+      "loss": 1.2092,
+      "step": 3862
+    },
+    {
+      "epoch": 1.4382147968028296,
+      "grad_norm": 0.16768351197242737,
+      "learning_rate": 1.940075363836673e-05,
+      "loss": 1.229,
+      "step": 3863
+    },
+    {
+      "epoch": 1.4385871019534386,
+      "grad_norm": 0.17820565402507782,
+      "learning_rate": 1.9400339186517544e-05,
+      "loss": 1.2276,
+      "step": 3864
+    },
+    {
+      "epoch": 1.4389594071040477,
+      "grad_norm": 0.17079061269760132,
+      "learning_rate": 1.9399924595825936e-05,
+      "loss": 1.2158,
+      "step": 3865
+    },
+    {
+      "epoch": 1.4393317122546567,
+      "grad_norm": 0.16866520047187805,
+      "learning_rate": 1.939950986629803e-05,
+      "loss": 1.1979,
+      "step": 3866
+    },
+    {
+      "epoch": 1.4397040174052658,
+      "grad_norm": 0.16329576075077057,
+      "learning_rate": 1.9399094997939957e-05,
+      "loss": 1.2184,
+      "step": 3867
+    },
+    {
+      "epoch": 1.4400763225558748,
+      "grad_norm": 0.17052018642425537,
+      "learning_rate": 1.9398679990757837e-05,
+      "loss": 1.2144,
+      "step": 3868
+    },
+    {
+      "epoch": 1.4404486277064839,
+      "grad_norm": 0.16311021149158478,
+      "learning_rate": 1.9398264844757805e-05,
+      "loss": 1.206,
+      "step": 3869
+    },
+    {
+      "epoch": 1.440820932857093,
+      "grad_norm": 0.17270159721374512,
+      "learning_rate": 1.9397849559945993e-05,
+      "loss": 1.2159,
+      "step": 3870
+    },
+    {
+      "epoch": 1.4411932380077022,
+      "grad_norm": 0.17571993172168732,
+      "learning_rate": 1.939743413632853e-05,
+      "loss": 1.202,
+      "step": 3871
+    },
+    {
+      "epoch": 1.4415655431583112,
+      "grad_norm": 0.16022472083568573,
+      "learning_rate": 1.9397018573911558e-05,
+      "loss": 1.2048,
+      "step": 3872
+    },
+    {
+      "epoch": 1.4419378483089202,
+      "grad_norm": 0.1789269596338272,
+      "learning_rate": 1.9396602872701205e-05,
+      "loss": 1.2145,
+      "step": 3873
+    },
+    {
+      "epoch": 1.4423101534595293,
+      "grad_norm": 0.18328560888767242,
+      "learning_rate": 1.9396187032703624e-05,
+      "loss": 1.2088,
+      "step": 3874
+    },
+    {
+      "epoch": 1.4426824586101383,
+      "grad_norm": 0.16325511038303375,
+      "learning_rate": 1.939577105392495e-05,
+      "loss": 1.2033,
+      "step": 3875
+    },
+    {
+      "epoch": 1.4430547637607474,
+      "grad_norm": 0.16945096850395203,
+      "learning_rate": 1.9395354936371323e-05,
+      "loss": 1.2225,
+      "step": 3876
+    },
+    {
+      "epoch": 1.4434270689113564,
+      "grad_norm": 0.16546650230884552,
+      "learning_rate": 1.9394938680048893e-05,
+      "loss": 1.2079,
+      "step": 3877
+    },
+    {
+      "epoch": 1.4437993740619655,
+      "grad_norm": 0.1665334403514862,
+      "learning_rate": 1.9394522284963814e-05,
+      "loss": 1.2367,
+      "step": 3878
+    },
+    {
+      "epoch": 1.4441716792125745,
+      "grad_norm": 0.1650296449661255,
+      "learning_rate": 1.9394105751122226e-05,
+      "loss": 1.2172,
+      "step": 3879
+    },
+    {
+      "epoch": 1.4445439843631838,
+      "grad_norm": 0.16569402813911438,
+      "learning_rate": 1.9393689078530285e-05,
+      "loss": 1.2179,
+      "step": 3880
+    },
+    {
+      "epoch": 1.4449162895137928,
+      "grad_norm": 0.1602257788181305,
+      "learning_rate": 1.9393272267194144e-05,
+      "loss": 1.2111,
+      "step": 3881
+    },
+    {
+      "epoch": 1.4452885946644018,
+      "grad_norm": 0.16653552651405334,
+      "learning_rate": 1.9392855317119966e-05,
+      "loss": 1.2216,
+      "step": 3882
+    },
+    {
+      "epoch": 1.445660899815011,
+      "grad_norm": 0.16940967738628387,
+      "learning_rate": 1.93924382283139e-05,
+      "loss": 1.2129,
+      "step": 3883
+    },
+    {
+      "epoch": 1.44603320496562,
+      "grad_norm": 0.1640087068080902,
+      "learning_rate": 1.9392021000782114e-05,
+      "loss": 1.2136,
+      "step": 3884
+    },
+    {
+      "epoch": 1.446405510116229,
+      "grad_norm": 0.15895035862922668,
+      "learning_rate": 1.939160363453077e-05,
+      "loss": 1.197,
+      "step": 3885
+    },
+    {
+      "epoch": 1.446777815266838,
+      "grad_norm": 0.16390269994735718,
+      "learning_rate": 1.9391186129566025e-05,
+      "loss": 1.2022,
+      "step": 3886
+    },
+    {
+      "epoch": 1.4471501204174473,
+      "grad_norm": 0.16444234549999237,
+      "learning_rate": 1.939076848589405e-05,
+      "loss": 1.2089,
+      "step": 3887
+    },
+    {
+      "epoch": 1.447522425568056,
+      "grad_norm": 0.16005566716194153,
+      "learning_rate": 1.9390350703521015e-05,
+      "loss": 1.1956,
+      "step": 3888
+    },
+    {
+      "epoch": 1.4478947307186654,
+      "grad_norm": 0.16827602684497833,
+      "learning_rate": 1.938993278245309e-05,
+      "loss": 1.2111,
+      "step": 3889
+    },
+    {
+      "epoch": 1.4482670358692744,
+      "grad_norm": 0.16684898734092712,
+      "learning_rate": 1.938951472269645e-05,
+      "loss": 1.2087,
+      "step": 3890
+    },
+    {
+      "epoch": 1.4486393410198835,
+      "grad_norm": 0.1641646921634674,
+      "learning_rate": 1.9389096524257263e-05,
+      "loss": 1.2132,
+      "step": 3891
+    },
+    {
+      "epoch": 1.4490116461704925,
+      "grad_norm": 0.17110654711723328,
+      "learning_rate": 1.938867818714171e-05,
+      "loss": 1.2203,
+      "step": 3892
+    },
+    {
+      "epoch": 1.4493839513211015,
+      "grad_norm": 0.15883228182792664,
+      "learning_rate": 1.938825971135597e-05,
+      "loss": 1.2096,
+      "step": 3893
+    },
+    {
+      "epoch": 1.4497562564717106,
+      "grad_norm": 0.1638491004705429,
+      "learning_rate": 1.938784109690622e-05,
+      "loss": 1.2125,
+      "step": 3894
+    },
+    {
+      "epoch": 1.4501285616223196,
+      "grad_norm": 0.16501039266586304,
+      "learning_rate": 1.938742234379865e-05,
+      "loss": 1.2235,
+      "step": 3895
+    },
+    {
+      "epoch": 1.4505008667729289,
+      "grad_norm": 0.16907702386379242,
+      "learning_rate": 1.938700345203944e-05,
+      "loss": 1.2204,
+      "step": 3896
+    },
+    {
+      "epoch": 1.4508731719235377,
+      "grad_norm": 0.16508372128009796,
+      "learning_rate": 1.938658442163478e-05,
+      "loss": 1.2152,
+      "step": 3897
+    },
+    {
+      "epoch": 1.451245477074147,
+      "grad_norm": 0.1664949506521225,
+      "learning_rate": 1.9386165252590854e-05,
+      "loss": 1.2177,
+      "step": 3898
+    },
+    {
+      "epoch": 1.451617782224756,
+      "grad_norm": 0.1608046591281891,
+      "learning_rate": 1.9385745944913858e-05,
+      "loss": 1.2062,
+      "step": 3899
+    },
+    {
+      "epoch": 1.451990087375365,
+      "grad_norm": 0.16553208231925964,
+      "learning_rate": 1.9385326498609983e-05,
+      "loss": 1.2204,
+      "step": 3900
+    },
+    {
+      "epoch": 1.452362392525974,
+      "grad_norm": 0.16579049825668335,
+      "learning_rate": 1.9384906913685426e-05,
+      "loss": 1.2166,
+      "step": 3901
+    },
+    {
+      "epoch": 1.4527346976765831,
+      "grad_norm": 0.17063631117343903,
+      "learning_rate": 1.9384487190146383e-05,
+      "loss": 1.2075,
+      "step": 3902
+    },
+    {
+      "epoch": 1.4531070028271922,
+      "grad_norm": 0.16345682740211487,
+      "learning_rate": 1.938406732799905e-05,
+      "loss": 1.2176,
+      "step": 3903
+    },
+    {
+      "epoch": 1.4534793079778012,
+      "grad_norm": 0.16839192807674408,
+      "learning_rate": 1.9383647327249635e-05,
+      "loss": 1.223,
+      "step": 3904
+    },
+    {
+      "epoch": 1.4538516131284105,
+      "grad_norm": 0.1658104509115219,
+      "learning_rate": 1.9383227187904334e-05,
+      "loss": 1.2142,
+      "step": 3905
+    },
+    {
+      "epoch": 1.4542239182790193,
+      "grad_norm": 0.16820955276489258,
+      "learning_rate": 1.9382806909969362e-05,
+      "loss": 1.2302,
+      "step": 3906
+    },
+    {
+      "epoch": 1.4545962234296286,
+      "grad_norm": 0.16950318217277527,
+      "learning_rate": 1.9382386493450913e-05,
+      "loss": 1.2091,
+      "step": 3907
+    },
+    {
+      "epoch": 1.4549685285802376,
+      "grad_norm": 0.17140018939971924,
+      "learning_rate": 1.938196593835521e-05,
+      "loss": 1.2267,
+      "step": 3908
+    },
+    {
+      "epoch": 1.4553408337308467,
+      "grad_norm": 0.17573173344135284,
+      "learning_rate": 1.938154524468846e-05,
+      "loss": 1.211,
+      "step": 3909
+    },
+    {
+      "epoch": 1.4557131388814557,
+      "grad_norm": 0.16952942311763763,
+      "learning_rate": 1.938112441245687e-05,
+      "loss": 1.2111,
+      "step": 3910
+    },
+    {
+      "epoch": 1.4560854440320647,
+      "grad_norm": 0.17475372552871704,
+      "learning_rate": 1.9380703441666666e-05,
+      "loss": 1.1979,
+      "step": 3911
+    },
+    {
+      "epoch": 1.4564577491826738,
+      "grad_norm": 0.16706113517284393,
+      "learning_rate": 1.9380282332324055e-05,
+      "loss": 1.2209,
+      "step": 3912
+    },
+    {
+      "epoch": 1.4568300543332828,
+      "grad_norm": 0.1862281709909439,
+      "learning_rate": 1.937986108443527e-05,
+      "loss": 1.2208,
+      "step": 3913
+    },
+    {
+      "epoch": 1.457202359483892,
+      "grad_norm": 0.17388898134231567,
+      "learning_rate": 1.9379439698006522e-05,
+      "loss": 1.2103,
+      "step": 3914
+    },
+    {
+      "epoch": 1.457574664634501,
+      "grad_norm": 0.1705702692270279,
+      "learning_rate": 1.9379018173044038e-05,
+      "loss": 1.2244,
+      "step": 3915
+    },
+    {
+      "epoch": 1.4579469697851102,
+      "grad_norm": 0.16031381487846375,
+      "learning_rate": 1.9378596509554045e-05,
+      "loss": 1.2021,
+      "step": 3916
+    },
+    {
+      "epoch": 1.4583192749357192,
+      "grad_norm": 0.17074525356292725,
+      "learning_rate": 1.937817470754277e-05,
+      "loss": 1.2091,
+      "step": 3917
+    },
+    {
+      "epoch": 1.4586915800863283,
+      "grad_norm": 0.1819746345281601,
+      "learning_rate": 1.9377752767016443e-05,
+      "loss": 1.2182,
+      "step": 3918
+    },
+    {
+      "epoch": 1.4590638852369373,
+      "grad_norm": 0.16856756806373596,
+      "learning_rate": 1.9377330687981295e-05,
+      "loss": 1.2114,
+      "step": 3919
+    },
+    {
+      "epoch": 1.4594361903875464,
+      "grad_norm": 0.16812150180339813,
+      "learning_rate": 1.9376908470443562e-05,
+      "loss": 1.2114,
+      "step": 3920
+    },
+    {
+      "epoch": 1.4598084955381554,
+      "grad_norm": 0.17135708034038544,
+      "learning_rate": 1.937648611440948e-05,
+      "loss": 1.2128,
+      "step": 3921
+    },
+    {
+      "epoch": 1.4601808006887644,
+      "grad_norm": 0.17878791689872742,
+      "learning_rate": 1.9376063619885285e-05,
+      "loss": 1.2219,
+      "step": 3922
+    },
+    {
+      "epoch": 1.4605531058393737,
+      "grad_norm": 0.162892147898674,
+      "learning_rate": 1.937564098687722e-05,
+      "loss": 1.2149,
+      "step": 3923
+    },
+    {
+      "epoch": 1.4609254109899827,
+      "grad_norm": 0.16608810424804688,
+      "learning_rate": 1.9375218215391527e-05,
+      "loss": 1.1958,
+      "step": 3924
+    },
+    {
+      "epoch": 1.4612977161405918,
+      "grad_norm": 0.17034226655960083,
+      "learning_rate": 1.9374795305434446e-05,
+      "loss": 1.193,
+      "step": 3925
+    },
+    {
+      "epoch": 1.4616700212912008,
+      "grad_norm": 0.15888184309005737,
+      "learning_rate": 1.937437225701223e-05,
+      "loss": 1.2113,
+      "step": 3926
+    },
+    {
+      "epoch": 1.4620423264418099,
+      "grad_norm": 0.17904046177864075,
+      "learning_rate": 1.937394907013112e-05,
+      "loss": 1.2128,
+      "step": 3927
+    },
+    {
+      "epoch": 1.462414631592419,
+      "grad_norm": 0.16831357777118683,
+      "learning_rate": 1.9373525744797377e-05,
+      "loss": 1.2023,
+      "step": 3928
+    },
+    {
+      "epoch": 1.462786936743028,
+      "grad_norm": 0.17215365171432495,
+      "learning_rate": 1.9373102281017246e-05,
+      "loss": 1.21,
+      "step": 3929
+    },
+    {
+      "epoch": 1.463159241893637,
+      "grad_norm": 0.16621056199073792,
+      "learning_rate": 1.937267867879698e-05,
+      "loss": 1.2021,
+      "step": 3930
+    },
+    {
+      "epoch": 1.463531547044246,
+      "grad_norm": 0.17535530030727386,
+      "learning_rate": 1.9372254938142835e-05,
+      "loss": 1.2345,
+      "step": 3931
+    },
+    {
+      "epoch": 1.4639038521948553,
+      "grad_norm": 0.16179418563842773,
+      "learning_rate": 1.9371831059061078e-05,
+      "loss": 1.2097,
+      "step": 3932
+    },
+    {
+      "epoch": 1.4642761573454643,
+      "grad_norm": 0.17177635431289673,
+      "learning_rate": 1.937140704155796e-05,
+      "loss": 1.2181,
+      "step": 3933
+    },
+    {
+      "epoch": 1.4646484624960734,
+      "grad_norm": 0.17407308518886566,
+      "learning_rate": 1.9370982885639752e-05,
+      "loss": 1.215,
+      "step": 3934
+    },
+    {
+      "epoch": 1.4650207676466824,
+      "grad_norm": 0.1671421378850937,
+      "learning_rate": 1.937055859131271e-05,
+      "loss": 1.2185,
+      "step": 3935
+    },
+    {
+      "epoch": 1.4653930727972915,
+      "grad_norm": 0.16444651782512665,
+      "learning_rate": 1.937013415858311e-05,
+      "loss": 1.2247,
+      "step": 3936
+    },
+    {
+      "epoch": 1.4657653779479005,
+      "grad_norm": 0.1708529144525528,
+      "learning_rate": 1.9369709587457217e-05,
+      "loss": 1.1963,
+      "step": 3937
+    },
+    {
+      "epoch": 1.4661376830985096,
+      "grad_norm": 0.17232173681259155,
+      "learning_rate": 1.93692848779413e-05,
+      "loss": 1.2015,
+      "step": 3938
+    },
+    {
+      "epoch": 1.4665099882491186,
+      "grad_norm": 0.16869337856769562,
+      "learning_rate": 1.936886003004163e-05,
+      "loss": 1.2328,
+      "step": 3939
+    },
+    {
+      "epoch": 1.4668822933997276,
+      "grad_norm": 0.17547161877155304,
+      "learning_rate": 1.9368435043764493e-05,
+      "loss": 1.2094,
+      "step": 3940
+    },
+    {
+      "epoch": 1.467254598550337,
+      "grad_norm": 0.1809409260749817,
+      "learning_rate": 1.9368009919116152e-05,
+      "loss": 1.2065,
+      "step": 3941
+    },
+    {
+      "epoch": 1.467626903700946,
+      "grad_norm": 0.1913343220949173,
+      "learning_rate": 1.9367584656102895e-05,
+      "loss": 1.2309,
+      "step": 3942
+    },
+    {
+      "epoch": 1.467999208851555,
+      "grad_norm": 0.16980275511741638,
+      "learning_rate": 1.9367159254731e-05,
+      "loss": 1.2062,
+      "step": 3943
+    },
+    {
+      "epoch": 1.468371514002164,
+      "grad_norm": 0.17768965661525726,
+      "learning_rate": 1.936673371500675e-05,
+      "loss": 1.2134,
+      "step": 3944
+    },
+    {
+      "epoch": 1.468743819152773,
+      "grad_norm": 0.17502190172672272,
+      "learning_rate": 1.9366308036936433e-05,
+      "loss": 1.2255,
+      "step": 3945
+    },
+    {
+      "epoch": 1.4691161243033821,
+      "grad_norm": 0.176423579454422,
+      "learning_rate": 1.936588222052633e-05,
+      "loss": 1.2131,
+      "step": 3946
+    },
+    {
+      "epoch": 1.4694884294539912,
+      "grad_norm": 0.18307115137577057,
+      "learning_rate": 1.936545626578274e-05,
+      "loss": 1.2093,
+      "step": 3947
+    },
+    {
+      "epoch": 1.4698607346046004,
+      "grad_norm": 0.16688889265060425,
+      "learning_rate": 1.9365030172711946e-05,
+      "loss": 1.2074,
+      "step": 3948
+    },
+    {
+      "epoch": 1.4702330397552092,
+      "grad_norm": 0.16726486384868622,
+      "learning_rate": 1.9364603941320243e-05,
+      "loss": 1.1952,
+      "step": 3949
+    },
+    {
+      "epoch": 1.4706053449058185,
+      "grad_norm": 0.1644599884748459,
+      "learning_rate": 1.9364177571613927e-05,
+      "loss": 1.2164,
+      "step": 3950
+    },
+    {
+      "epoch": 1.4709776500564276,
+      "grad_norm": 0.17639389634132385,
+      "learning_rate": 1.9363751063599298e-05,
+      "loss": 1.2219,
+      "step": 3951
+    },
+    {
+      "epoch": 1.4713499552070366,
+      "grad_norm": 0.1668757051229477,
+      "learning_rate": 1.936332441728265e-05,
+      "loss": 1.2021,
+      "step": 3952
+    },
+    {
+      "epoch": 1.4717222603576456,
+      "grad_norm": 0.17846433818340302,
+      "learning_rate": 1.9362897632670293e-05,
+      "loss": 1.2116,
+      "step": 3953
+    },
+    {
+      "epoch": 1.4720945655082547,
+      "grad_norm": 0.18401432037353516,
+      "learning_rate": 1.936247070976852e-05,
+      "loss": 1.2167,
+      "step": 3954
+    },
+    {
+      "epoch": 1.4724668706588637,
+      "grad_norm": 0.16676369309425354,
+      "learning_rate": 1.9362043648583647e-05,
+      "loss": 1.2144,
+      "step": 3955
+    },
+    {
+      "epoch": 1.4728391758094728,
+      "grad_norm": 0.17086462676525116,
+      "learning_rate": 1.9361616449121973e-05,
+      "loss": 1.2163,
+      "step": 3956
+    },
+    {
+      "epoch": 1.473211480960082,
+      "grad_norm": 0.1699550896883011,
+      "learning_rate": 1.9361189111389817e-05,
+      "loss": 1.2083,
+      "step": 3957
+    },
+    {
+      "epoch": 1.4735837861106909,
+      "grad_norm": 0.16862007975578308,
+      "learning_rate": 1.936076163539348e-05,
+      "loss": 1.2129,
+      "step": 3958
+    },
+    {
+      "epoch": 1.4739560912613001,
+      "grad_norm": 0.168454110622406,
+      "learning_rate": 1.936033402113928e-05,
+      "loss": 1.2052,
+      "step": 3959
+    },
+    {
+      "epoch": 1.4743283964119092,
+      "grad_norm": 0.17576009035110474,
+      "learning_rate": 1.9359906268633542e-05,
+      "loss": 1.1988,
+      "step": 3960
+    },
+    {
+      "epoch": 1.4747007015625182,
+      "grad_norm": 0.17533783614635468,
+      "learning_rate": 1.9359478377882567e-05,
+      "loss": 1.2165,
+      "step": 3961
+    },
+    {
+      "epoch": 1.4750730067131272,
+      "grad_norm": 0.18363003432750702,
+      "learning_rate": 1.935905034889269e-05,
+      "loss": 1.2095,
+      "step": 3962
+    },
+    {
+      "epoch": 1.4754453118637363,
+      "grad_norm": 0.1739223450422287,
+      "learning_rate": 1.9358622181670225e-05,
+      "loss": 1.2043,
+      "step": 3963
+    },
+    {
+      "epoch": 1.4758176170143453,
+      "grad_norm": 0.1666637510061264,
+      "learning_rate": 1.9358193876221497e-05,
+      "loss": 1.2036,
+      "step": 3964
+    },
+    {
+      "epoch": 1.4761899221649544,
+      "grad_norm": 0.18564556539058685,
+      "learning_rate": 1.935776543255283e-05,
+      "loss": 1.2267,
+      "step": 3965
+    },
+    {
+      "epoch": 1.4765622273155636,
+      "grad_norm": 0.16822317242622375,
+      "learning_rate": 1.935733685067056e-05,
+      "loss": 1.2294,
+      "step": 3966
+    },
+    {
+      "epoch": 1.4769345324661725,
+      "grad_norm": 0.16911527514457703,
+      "learning_rate": 1.9356908130581008e-05,
+      "loss": 1.2128,
+      "step": 3967
+    },
+    {
+      "epoch": 1.4773068376167817,
+      "grad_norm": 0.19350941479206085,
+      "learning_rate": 1.9356479272290514e-05,
+      "loss": 1.2107,
+      "step": 3968
+    },
+    {
+      "epoch": 1.4776791427673908,
+      "grad_norm": 0.1877276450395584,
+      "learning_rate": 1.9356050275805406e-05,
+      "loss": 1.2059,
+      "step": 3969
+    },
+    {
+      "epoch": 1.4780514479179998,
+      "grad_norm": 0.17434152960777283,
+      "learning_rate": 1.9355621141132022e-05,
+      "loss": 1.2185,
+      "step": 3970
+    },
+    {
+      "epoch": 1.4784237530686088,
+      "grad_norm": 0.19158223271369934,
+      "learning_rate": 1.9355191868276702e-05,
+      "loss": 1.2107,
+      "step": 3971
+    },
+    {
+      "epoch": 1.478796058219218,
+      "grad_norm": 0.17890071868896484,
+      "learning_rate": 1.9354762457245782e-05,
+      "loss": 1.205,
+      "step": 3972
+    },
+    {
+      "epoch": 1.479168363369827,
+      "grad_norm": 0.16865649819374084,
+      "learning_rate": 1.935433290804561e-05,
+      "loss": 1.2112,
+      "step": 3973
+    },
+    {
+      "epoch": 1.479540668520436,
+      "grad_norm": 0.16355697810649872,
+      "learning_rate": 1.935390322068253e-05,
+      "loss": 1.2194,
+      "step": 3974
+    },
+    {
+      "epoch": 1.4799129736710452,
+      "grad_norm": 0.169992133975029,
+      "learning_rate": 1.9353473395162882e-05,
+      "loss": 1.2088,
+      "step": 3975
+    },
+    {
+      "epoch": 1.480285278821654,
+      "grad_norm": 0.16917189955711365,
+      "learning_rate": 1.9353043431493024e-05,
+      "loss": 1.2149,
+      "step": 3976
+    },
+    {
+      "epoch": 1.4806575839722633,
+      "grad_norm": 0.16801759600639343,
+      "learning_rate": 1.9352613329679298e-05,
+      "loss": 1.2251,
+      "step": 3977
+    },
+    {
+      "epoch": 1.4810298891228724,
+      "grad_norm": 0.16183248162269592,
+      "learning_rate": 1.935218308972806e-05,
+      "loss": 1.2019,
+      "step": 3978
+    },
+    {
+      "epoch": 1.4814021942734814,
+      "grad_norm": 0.16628491878509521,
+      "learning_rate": 1.935175271164567e-05,
+      "loss": 1.2186,
+      "step": 3979
+    },
+    {
+      "epoch": 1.4817744994240905,
+      "grad_norm": 0.167390838265419,
+      "learning_rate": 1.9351322195438472e-05,
+      "loss": 1.2196,
+      "step": 3980
+    },
+    {
+      "epoch": 1.4821468045746995,
+      "grad_norm": 0.17368678748607635,
+      "learning_rate": 1.9350891541112836e-05,
+      "loss": 1.2114,
+      "step": 3981
+    },
+    {
+      "epoch": 1.4825191097253085,
+      "grad_norm": 0.17052781581878662,
+      "learning_rate": 1.9350460748675117e-05,
+      "loss": 1.2013,
+      "step": 3982
+    },
+    {
+      "epoch": 1.4828914148759176,
+      "grad_norm": 0.16377271711826324,
+      "learning_rate": 1.935002981813168e-05,
+      "loss": 1.2114,
+      "step": 3983
+    },
+    {
+      "epoch": 1.4832637200265268,
+      "grad_norm": 0.1672280728816986,
+      "learning_rate": 1.934959874948889e-05,
+      "loss": 1.2214,
+      "step": 3984
+    },
+    {
+      "epoch": 1.4836360251771359,
+      "grad_norm": 0.16582052409648895,
+      "learning_rate": 1.9349167542753116e-05,
+      "loss": 1.2097,
+      "step": 3985
+    },
+    {
+      "epoch": 1.484008330327745,
+      "grad_norm": 0.15691331028938293,
+      "learning_rate": 1.934873619793072e-05,
+      "loss": 1.2001,
+      "step": 3986
+    },
+    {
+      "epoch": 1.484380635478354,
+      "grad_norm": 0.16252057254314423,
+      "learning_rate": 1.934830471502808e-05,
+      "loss": 1.2105,
+      "step": 3987
+    },
+    {
+      "epoch": 1.484752940628963,
+      "grad_norm": 0.18076485395431519,
+      "learning_rate": 1.9347873094051565e-05,
+      "loss": 1.1981,
+      "step": 3988
+    },
+    {
+      "epoch": 1.485125245779572,
+      "grad_norm": 0.18029817938804626,
+      "learning_rate": 1.9347441335007547e-05,
+      "loss": 1.2136,
+      "step": 3989
+    },
+    {
+      "epoch": 1.485497550930181,
+      "grad_norm": 0.17143037915229797,
+      "learning_rate": 1.9347009437902414e-05,
+      "loss": 1.1981,
+      "step": 3990
+    },
+    {
+      "epoch": 1.4858698560807901,
+      "grad_norm": 0.1707201451063156,
+      "learning_rate": 1.9346577402742532e-05,
+      "loss": 1.2137,
+      "step": 3991
+    },
+    {
+      "epoch": 1.4862421612313992,
+      "grad_norm": 0.17134663462638855,
+      "learning_rate": 1.9346145229534295e-05,
+      "loss": 1.2097,
+      "step": 3992
+    },
+    {
+      "epoch": 1.4866144663820084,
+      "grad_norm": 0.18079330027103424,
+      "learning_rate": 1.9345712918284074e-05,
+      "loss": 1.2135,
+      "step": 3993
+    },
+    {
+      "epoch": 1.4869867715326175,
+      "grad_norm": 0.16749143600463867,
+      "learning_rate": 1.934528046899826e-05,
+      "loss": 1.2145,
+      "step": 3994
+    },
+    {
+      "epoch": 1.4873590766832265,
+      "grad_norm": 0.17483735084533691,
+      "learning_rate": 1.9344847881683242e-05,
+      "loss": 1.2164,
+      "step": 3995
+    },
+    {
+      "epoch": 1.4877313818338356,
+      "grad_norm": 0.17424844205379486,
+      "learning_rate": 1.9344415156345407e-05,
+      "loss": 1.2007,
+      "step": 3996
+    },
+    {
+      "epoch": 1.4881036869844446,
+      "grad_norm": 0.18237757682800293,
+      "learning_rate": 1.9343982292991147e-05,
+      "loss": 1.1958,
+      "step": 3997
+    },
+    {
+      "epoch": 1.4884759921350537,
+      "grad_norm": 0.2026931643486023,
+      "learning_rate": 1.9343549291626853e-05,
+      "loss": 1.2126,
+      "step": 3998
+    },
+    {
+      "epoch": 1.4888482972856627,
+      "grad_norm": 0.2018553465604782,
+      "learning_rate": 1.9343116152258924e-05,
+      "loss": 1.2323,
+      "step": 3999
+    },
+    {
+      "epoch": 1.489220602436272,
+      "grad_norm": 0.17525236308574677,
+      "learning_rate": 1.9342682874893756e-05,
+      "loss": 1.2227,
+      "step": 4000
+    },
+    {
+      "epoch": 1.489220602436272,
+      "eval_loss": 1.3177570104599,
+      "eval_runtime": 16.441,
+      "eval_samples_per_second": 105.468,
+      "eval_steps_per_second": 5.292,
+      "step": 4000
+    },
+    {
+      "epoch": 1.4895929075868808,
+      "grad_norm": 0.20407749712467194,
+      "learning_rate": 1.9342249459537746e-05,
+      "loss": 1.2125,
+      "step": 4001
+    },
+    {
+      "epoch": 1.48996521273749,
+      "grad_norm": 0.1680436134338379,
+      "learning_rate": 1.93418159061973e-05,
+      "loss": 1.208,
+      "step": 4002
+    },
+    {
+      "epoch": 1.490337517888099,
+      "grad_norm": 0.16656237840652466,
+      "learning_rate": 1.934138221487882e-05,
+      "loss": 1.2182,
+      "step": 4003
+    },
+    {
+      "epoch": 1.4907098230387081,
+      "grad_norm": 0.17523828148841858,
+      "learning_rate": 1.9340948385588708e-05,
+      "loss": 1.2013,
+      "step": 4004
+    },
+    {
+      "epoch": 1.4910821281893172,
+      "grad_norm": 0.17147956788539886,
+      "learning_rate": 1.9340514418333375e-05,
+      "loss": 1.221,
+      "step": 4005
+    },
+    {
+      "epoch": 1.4914544333399262,
+      "grad_norm": 0.16101637482643127,
+      "learning_rate": 1.934008031311923e-05,
+      "loss": 1.2107,
+      "step": 4006
+    },
+    {
+      "epoch": 1.4918267384905353,
+      "grad_norm": 0.17050312459468842,
+      "learning_rate": 1.933964606995269e-05,
+      "loss": 1.216,
+      "step": 4007
+    },
+    {
+      "epoch": 1.4921990436411443,
+      "grad_norm": 0.17762567102909088,
+      "learning_rate": 1.933921168884016e-05,
+      "loss": 1.219,
+      "step": 4008
+    },
+    {
+      "epoch": 1.4925713487917536,
+      "grad_norm": 0.179819718003273,
+      "learning_rate": 1.9338777169788058e-05,
+      "loss": 1.2271,
+      "step": 4009
+    },
+    {
+      "epoch": 1.4929436539423624,
+      "grad_norm": 0.17511019110679626,
+      "learning_rate": 1.9338342512802805e-05,
+      "loss": 1.2128,
+      "step": 4010
+    },
+    {
+      "epoch": 1.4933159590929717,
+      "grad_norm": 0.17348961532115936,
+      "learning_rate": 1.9337907717890817e-05,
+      "loss": 1.2215,
+      "step": 4011
+    },
+    {
+      "epoch": 1.4936882642435807,
+      "grad_norm": 0.16590604186058044,
+      "learning_rate": 1.933747278505852e-05,
+      "loss": 1.2048,
+      "step": 4012
+    },
+    {
+      "epoch": 1.4940605693941897,
+      "grad_norm": 0.16932128369808197,
+      "learning_rate": 1.9337037714312337e-05,
+      "loss": 1.22,
+      "step": 4013
+    },
+    {
+      "epoch": 1.4944328745447988,
+      "grad_norm": 0.18030394613742828,
+      "learning_rate": 1.933660250565869e-05,
+      "loss": 1.2031,
+      "step": 4014
+    },
+    {
+      "epoch": 1.4948051796954078,
+      "grad_norm": 0.16804668307304382,
+      "learning_rate": 1.9336167159104012e-05,
+      "loss": 1.2167,
+      "step": 4015
+    },
+    {
+      "epoch": 1.4951774848460169,
+      "grad_norm": 0.17750367522239685,
+      "learning_rate": 1.9335731674654732e-05,
+      "loss": 1.2102,
+      "step": 4016
+    },
+    {
+      "epoch": 1.495549789996626,
+      "grad_norm": 0.1765013486146927,
+      "learning_rate": 1.9335296052317278e-05,
+      "loss": 1.2192,
+      "step": 4017
+    },
+    {
+      "epoch": 1.4959220951472352,
+      "grad_norm": 0.17468391358852386,
+      "learning_rate": 1.933486029209809e-05,
+      "loss": 1.2093,
+      "step": 4018
+    },
+    {
+      "epoch": 1.496294400297844,
+      "grad_norm": 0.16882555186748505,
+      "learning_rate": 1.93344243940036e-05,
+      "loss": 1.2092,
+      "step": 4019
+    },
+    {
+      "epoch": 1.4966667054484533,
+      "grad_norm": 0.16576802730560303,
+      "learning_rate": 1.9333988358040246e-05,
+      "loss": 1.2021,
+      "step": 4020
+    },
+    {
+      "epoch": 1.4970390105990623,
+      "grad_norm": 0.17148758471012115,
+      "learning_rate": 1.9333552184214473e-05,
+      "loss": 1.2202,
+      "step": 4021
+    },
+    {
+      "epoch": 1.4974113157496713,
+      "grad_norm": 0.16886599361896515,
+      "learning_rate": 1.933311587253272e-05,
+      "loss": 1.2053,
+      "step": 4022
+    },
+    {
+      "epoch": 1.4977836209002804,
+      "grad_norm": 0.16921773552894592,
+      "learning_rate": 1.9332679423001428e-05,
+      "loss": 1.2064,
+      "step": 4023
+    },
+    {
+      "epoch": 1.4981559260508894,
+      "grad_norm": 0.16829746961593628,
+      "learning_rate": 1.9332242835627048e-05,
+      "loss": 1.2097,
+      "step": 4024
+    },
+    {
+      "epoch": 1.4985282312014985,
+      "grad_norm": 0.16895058751106262,
+      "learning_rate": 1.9331806110416027e-05,
+      "loss": 1.2116,
+      "step": 4025
+    },
+    {
+      "epoch": 1.4989005363521075,
+      "grad_norm": 0.17444762587547302,
+      "learning_rate": 1.9331369247374815e-05,
+      "loss": 1.1994,
+      "step": 4026
+    },
+    {
+      "epoch": 1.4992728415027168,
+      "grad_norm": 0.17568188905715942,
+      "learning_rate": 1.9330932246509867e-05,
+      "loss": 1.1921,
+      "step": 4027
+    },
+    {
+      "epoch": 1.4996451466533256,
+      "grad_norm": 0.164281964302063,
+      "learning_rate": 1.9330495107827633e-05,
+      "loss": 1.221,
+      "step": 4028
+    },
+    {
+      "epoch": 1.5000174518039349,
+      "grad_norm": 0.1712205559015274,
+      "learning_rate": 1.9330057831334573e-05,
+      "loss": 1.2055,
+      "step": 4029
+    },
+    {
+      "epoch": 1.500389756954544,
+      "grad_norm": 0.17232446372509003,
+      "learning_rate": 1.9329620417037143e-05,
+      "loss": 1.2094,
+      "step": 4030
+    },
+    {
+      "epoch": 1.500762062105153,
+      "grad_norm": 0.16828641295433044,
+      "learning_rate": 1.9329182864941808e-05,
+      "loss": 1.2161,
+      "step": 4031
+    },
+    {
+      "epoch": 1.501134367255762,
+      "grad_norm": 0.1642870008945465,
+      "learning_rate": 1.9328745175055025e-05,
+      "loss": 1.2076,
+      "step": 4032
+    },
+    {
+      "epoch": 1.501506672406371,
+      "grad_norm": 0.16758432984352112,
+      "learning_rate": 1.932830734738326e-05,
+      "loss": 1.2149,
+      "step": 4033
+    },
+    {
+      "epoch": 1.5018789775569803,
+      "grad_norm": 0.16760718822479248,
+      "learning_rate": 1.9327869381932984e-05,
+      "loss": 1.221,
+      "step": 4034
+    },
+    {
+      "epoch": 1.5022512827075891,
+      "grad_norm": 0.16775357723236084,
+      "learning_rate": 1.932743127871066e-05,
+      "loss": 1.2106,
+      "step": 4035
+    },
+    {
+      "epoch": 1.5026235878581984,
+      "grad_norm": 0.16413766145706177,
+      "learning_rate": 1.9326993037722762e-05,
+      "loss": 1.2133,
+      "step": 4036
+    },
+    {
+      "epoch": 1.5029958930088072,
+      "grad_norm": 0.17524518072605133,
+      "learning_rate": 1.9326554658975766e-05,
+      "loss": 1.2083,
+      "step": 4037
+    },
+    {
+      "epoch": 1.5033681981594165,
+      "grad_norm": 0.17292781174182892,
+      "learning_rate": 1.9326116142476137e-05,
+      "loss": 1.2126,
+      "step": 4038
+    },
+    {
+      "epoch": 1.5037405033100255,
+      "grad_norm": 0.17386804521083832,
+      "learning_rate": 1.9325677488230364e-05,
+      "loss": 1.2151,
+      "step": 4039
+    },
+    {
+      "epoch": 1.5041128084606346,
+      "grad_norm": 0.1756100207567215,
+      "learning_rate": 1.9325238696244914e-05,
+      "loss": 1.2122,
+      "step": 4040
+    },
+    {
+      "epoch": 1.5044851136112436,
+      "grad_norm": 0.16365545988082886,
+      "learning_rate": 1.9324799766526276e-05,
+      "loss": 1.2159,
+      "step": 4041
+    },
+    {
+      "epoch": 1.5048574187618526,
+      "grad_norm": 0.1651315987110138,
+      "learning_rate": 1.932436069908093e-05,
+      "loss": 1.2278,
+      "step": 4042
+    },
+    {
+      "epoch": 1.505229723912462,
+      "grad_norm": 0.1678977757692337,
+      "learning_rate": 1.9323921493915364e-05,
+      "loss": 1.2131,
+      "step": 4043
+    },
+    {
+      "epoch": 1.5056020290630707,
+      "grad_norm": 0.16497032344341278,
+      "learning_rate": 1.932348215103606e-05,
+      "loss": 1.2105,
+      "step": 4044
+    },
+    {
+      "epoch": 1.50597433421368,
+      "grad_norm": 0.1694755107164383,
+      "learning_rate": 1.932304267044951e-05,
+      "loss": 1.2084,
+      "step": 4045
+    },
+    {
+      "epoch": 1.5063466393642888,
+      "grad_norm": 0.16757144033908844,
+      "learning_rate": 1.932260305216221e-05,
+      "loss": 1.2132,
+      "step": 4046
+    },
+    {
+      "epoch": 1.506718944514898,
+      "grad_norm": 0.16843421757221222,
+      "learning_rate": 1.932216329618064e-05,
+      "loss": 1.2041,
+      "step": 4047
+    },
+    {
+      "epoch": 1.5070912496655071,
+      "grad_norm": 0.17438024282455444,
+      "learning_rate": 1.9321723402511308e-05,
+      "loss": 1.2206,
+      "step": 4048
+    },
+    {
+      "epoch": 1.5074635548161162,
+      "grad_norm": 0.17273013293743134,
+      "learning_rate": 1.9321283371160704e-05,
+      "loss": 1.2153,
+      "step": 4049
+    },
+    {
+      "epoch": 1.5078358599667252,
+      "grad_norm": 0.16855518519878387,
+      "learning_rate": 1.9320843202135333e-05,
+      "loss": 1.2327,
+      "step": 4050
+    },
+    {
+      "epoch": 1.5082081651173342,
+      "grad_norm": 0.17092865705490112,
+      "learning_rate": 1.932040289544169e-05,
+      "loss": 1.2183,
+      "step": 4051
+    },
+    {
+      "epoch": 1.5085804702679435,
+      "grad_norm": 0.16333964467048645,
+      "learning_rate": 1.9319962451086282e-05,
+      "loss": 1.2078,
+      "step": 4052
+    },
+    {
+      "epoch": 1.5089527754185523,
+      "grad_norm": 0.16895681619644165,
+      "learning_rate": 1.9319521869075612e-05,
+      "loss": 1.2129,
+      "step": 4053
+    },
+    {
+      "epoch": 1.5093250805691616,
+      "grad_norm": 0.17089246213436127,
+      "learning_rate": 1.931908114941619e-05,
+      "loss": 1.1971,
+      "step": 4054
+    },
+    {
+      "epoch": 1.5096973857197704,
+      "grad_norm": 0.1670157015323639,
+      "learning_rate": 1.9318640292114526e-05,
+      "loss": 1.2041,
+      "step": 4055
+    },
+    {
+      "epoch": 1.5100696908703797,
+      "grad_norm": 0.16973476111888885,
+      "learning_rate": 1.9318199297177127e-05,
+      "loss": 1.2034,
+      "step": 4056
+    },
+    {
+      "epoch": 1.5104419960209887,
+      "grad_norm": 0.173712819814682,
+      "learning_rate": 1.931775816461051e-05,
+      "loss": 1.2121,
+      "step": 4057
+    },
+    {
+      "epoch": 1.5108143011715978,
+      "grad_norm": 0.1701354682445526,
+      "learning_rate": 1.931731689442119e-05,
+      "loss": 1.2193,
+      "step": 4058
+    },
+    {
+      "epoch": 1.5111866063222068,
+      "grad_norm": 0.16733410954475403,
+      "learning_rate": 1.9316875486615684e-05,
+      "loss": 1.2152,
+      "step": 4059
+    },
+    {
+      "epoch": 1.5115589114728158,
+      "grad_norm": 0.17362013459205627,
+      "learning_rate": 1.931643394120051e-05,
+      "loss": 1.2194,
+      "step": 4060
+    },
+    {
+      "epoch": 1.5119312166234251,
+      "grad_norm": 0.1801014095544815,
+      "learning_rate": 1.9315992258182196e-05,
+      "loss": 1.2112,
+      "step": 4061
+    },
+    {
+      "epoch": 1.512303521774034,
+      "grad_norm": 0.17193296551704407,
+      "learning_rate": 1.9315550437567258e-05,
+      "loss": 1.2177,
+      "step": 4062
+    },
+    {
+      "epoch": 1.5126758269246432,
+      "grad_norm": 0.17398405075073242,
+      "learning_rate": 1.9315108479362226e-05,
+      "loss": 1.2154,
+      "step": 4063
+    },
+    {
+      "epoch": 1.513048132075252,
+      "grad_norm": 0.18025220930576324,
+      "learning_rate": 1.9314666383573622e-05,
+      "loss": 1.2044,
+      "step": 4064
+    },
+    {
+      "epoch": 1.5134204372258613,
+      "grad_norm": 0.15889345109462738,
+      "learning_rate": 1.9314224150207986e-05,
+      "loss": 1.2138,
+      "step": 4065
+    },
+    {
+      "epoch": 1.5137927423764703,
+      "grad_norm": 0.1648905724287033,
+      "learning_rate": 1.9313781779271842e-05,
+      "loss": 1.2124,
+      "step": 4066
+    },
+    {
+      "epoch": 1.5141650475270794,
+      "grad_norm": 0.1780308037996292,
+      "learning_rate": 1.9313339270771724e-05,
+      "loss": 1.203,
+      "step": 4067
+    },
+    {
+      "epoch": 1.5145373526776884,
+      "grad_norm": 0.16824577748775482,
+      "learning_rate": 1.931289662471417e-05,
+      "loss": 1.2033,
+      "step": 4068
+    },
+    {
+      "epoch": 1.5149096578282975,
+      "grad_norm": 0.1748679131269455,
+      "learning_rate": 1.9312453841105716e-05,
+      "loss": 1.2034,
+      "step": 4069
+    },
+    {
+      "epoch": 1.5152819629789067,
+      "grad_norm": 0.16885723173618317,
+      "learning_rate": 1.9312010919952907e-05,
+      "loss": 1.2147,
+      "step": 4070
+    },
+    {
+      "epoch": 1.5156542681295155,
+      "grad_norm": 0.16560660302639008,
+      "learning_rate": 1.931156786126228e-05,
+      "loss": 1.219,
+      "step": 4071
+    },
+    {
+      "epoch": 1.5160265732801248,
+      "grad_norm": 0.16653484106063843,
+      "learning_rate": 1.9311124665040378e-05,
+      "loss": 1.2258,
+      "step": 4072
+    },
+    {
+      "epoch": 1.5163988784307338,
+      "grad_norm": 0.16899293661117554,
+      "learning_rate": 1.931068133129375e-05,
+      "loss": 1.208,
+      "step": 4073
+    },
+    {
+      "epoch": 1.5167711835813429,
+      "grad_norm": 0.17353177070617676,
+      "learning_rate": 1.931023786002894e-05,
+      "loss": 1.196,
+      "step": 4074
+    },
+    {
+      "epoch": 1.517143488731952,
+      "grad_norm": 0.16658303141593933,
+      "learning_rate": 1.9309794251252506e-05,
+      "loss": 1.2167,
+      "step": 4075
+    },
+    {
+      "epoch": 1.517515793882561,
+      "grad_norm": 0.17217494547367096,
+      "learning_rate": 1.9309350504970996e-05,
+      "loss": 1.2224,
+      "step": 4076
+    },
+    {
+      "epoch": 1.51788809903317,
+      "grad_norm": 0.1716787964105606,
+      "learning_rate": 1.930890662119096e-05,
+      "loss": 1.214,
+      "step": 4077
+    },
+    {
+      "epoch": 1.518260404183779,
+      "grad_norm": 0.16615904867649078,
+      "learning_rate": 1.9308462599918955e-05,
+      "loss": 1.2204,
+      "step": 4078
+    },
+    {
+      "epoch": 1.5186327093343883,
+      "grad_norm": 0.17430458962917328,
+      "learning_rate": 1.9308018441161547e-05,
+      "loss": 1.2237,
+      "step": 4079
+    },
+    {
+      "epoch": 1.5190050144849971,
+      "grad_norm": 0.17217344045639038,
+      "learning_rate": 1.9307574144925288e-05,
+      "loss": 1.22,
+      "step": 4080
+    },
+    {
+      "epoch": 1.5193773196356064,
+      "grad_norm": 0.1643909215927124,
+      "learning_rate": 1.930712971121674e-05,
+      "loss": 1.2008,
+      "step": 4081
+    },
+    {
+      "epoch": 1.5197496247862154,
+      "grad_norm": 0.17074188590049744,
+      "learning_rate": 1.9306685140042476e-05,
+      "loss": 1.2294,
+      "step": 4082
+    },
+    {
+      "epoch": 1.5201219299368245,
+      "grad_norm": 0.16520312428474426,
+      "learning_rate": 1.9306240431409056e-05,
+      "loss": 1.2173,
+      "step": 4083
+    },
+    {
+      "epoch": 1.5204942350874335,
+      "grad_norm": 0.16149955987930298,
+      "learning_rate": 1.9305795585323043e-05,
+      "loss": 1.2173,
+      "step": 4084
+    },
+    {
+      "epoch": 1.5208665402380426,
+      "grad_norm": 0.17141614854335785,
+      "learning_rate": 1.930535060179102e-05,
+      "loss": 1.2022,
+      "step": 4085
+    },
+    {
+      "epoch": 1.5212388453886518,
+      "grad_norm": 0.16703253984451294,
+      "learning_rate": 1.930490548081955e-05,
+      "loss": 1.2247,
+      "step": 4086
+    },
+    {
+      "epoch": 1.5216111505392607,
+      "grad_norm": 0.16546790301799774,
+      "learning_rate": 1.9304460222415207e-05,
+      "loss": 1.2045,
+      "step": 4087
+    },
+    {
+      "epoch": 1.52198345568987,
+      "grad_norm": 0.16973921656608582,
+      "learning_rate": 1.9304014826584578e-05,
+      "loss": 1.226,
+      "step": 4088
+    },
+    {
+      "epoch": 1.5223557608404787,
+      "grad_norm": 0.16389979422092438,
+      "learning_rate": 1.9303569293334226e-05,
+      "loss": 1.2052,
+      "step": 4089
+    },
+    {
+      "epoch": 1.522728065991088,
+      "grad_norm": 0.16464626789093018,
+      "learning_rate": 1.9303123622670743e-05,
+      "loss": 1.2047,
+      "step": 4090
+    },
+    {
+      "epoch": 1.523100371141697,
+      "grad_norm": 0.1665724664926529,
+      "learning_rate": 1.9302677814600707e-05,
+      "loss": 1.2025,
+      "step": 4091
+    },
+    {
+      "epoch": 1.523472676292306,
+      "grad_norm": 0.18270178139209747,
+      "learning_rate": 1.9302231869130703e-05,
+      "loss": 1.2181,
+      "step": 4092
+    },
+    {
+      "epoch": 1.5238449814429151,
+      "grad_norm": 0.1672508865594864,
+      "learning_rate": 1.9301785786267323e-05,
+      "loss": 1.2081,
+      "step": 4093
+    },
+    {
+      "epoch": 1.5242172865935242,
+      "grad_norm": 0.16370201110839844,
+      "learning_rate": 1.9301339566017144e-05,
+      "loss": 1.2036,
+      "step": 4094
+    },
+    {
+      "epoch": 1.5245895917441334,
+      "grad_norm": 0.17350997030735016,
+      "learning_rate": 1.930089320838677e-05,
+      "loss": 1.2072,
+      "step": 4095
+    },
+    {
+      "epoch": 1.5249618968947423,
+      "grad_norm": 0.17266206443309784,
+      "learning_rate": 1.930044671338278e-05,
+      "loss": 1.2169,
+      "step": 4096
+    },
+    {
+      "epoch": 1.5253342020453515,
+      "grad_norm": 0.16876213252544403,
+      "learning_rate": 1.9300000081011778e-05,
+      "loss": 1.2076,
+      "step": 4097
+    },
+    {
+      "epoch": 1.5257065071959603,
+      "grad_norm": 0.17879076302051544,
+      "learning_rate": 1.9299553311280358e-05,
+      "loss": 1.2003,
+      "step": 4098
+    },
+    {
+      "epoch": 1.5260788123465696,
+      "grad_norm": 0.16374212503433228,
+      "learning_rate": 1.9299106404195123e-05,
+      "loss": 1.2203,
+      "step": 4099
+    },
+    {
+      "epoch": 1.5264511174971787,
+      "grad_norm": 0.17050255835056305,
+      "learning_rate": 1.929865935976267e-05,
+      "loss": 1.2112,
+      "step": 4100
+    },
+    {
+      "epoch": 1.5268234226477877,
+      "grad_norm": 0.17496921122074127,
+      "learning_rate": 1.92982121779896e-05,
+      "loss": 1.2123,
+      "step": 4101
+    },
+    {
+      "epoch": 1.5271957277983967,
+      "grad_norm": 0.17470447719097137,
+      "learning_rate": 1.9297764858882516e-05,
+      "loss": 1.2115,
+      "step": 4102
+    },
+    {
+      "epoch": 1.5275680329490058,
+      "grad_norm": 0.166524276137352,
+      "learning_rate": 1.9297317402448032e-05,
+      "loss": 1.2235,
+      "step": 4103
+    },
+    {
+      "epoch": 1.527940338099615,
+      "grad_norm": 0.1667315661907196,
+      "learning_rate": 1.929686980869275e-05,
+      "loss": 1.2164,
+      "step": 4104
+    },
+    {
+      "epoch": 1.5283126432502239,
+      "grad_norm": 0.17172963917255402,
+      "learning_rate": 1.9296422077623293e-05,
+      "loss": 1.2133,
+      "step": 4105
+    },
+    {
+      "epoch": 1.5286849484008331,
+      "grad_norm": 0.16795189678668976,
+      "learning_rate": 1.9295974209246257e-05,
+      "loss": 1.2076,
+      "step": 4106
+    },
+    {
+      "epoch": 1.529057253551442,
+      "grad_norm": 0.1728687286376953,
+      "learning_rate": 1.9295526203568268e-05,
+      "loss": 1.214,
+      "step": 4107
+    },
+    {
+      "epoch": 1.5294295587020512,
+      "grad_norm": 0.16910706460475922,
+      "learning_rate": 1.9295078060595938e-05,
+      "loss": 1.2089,
+      "step": 4108
+    },
+    {
+      "epoch": 1.5298018638526603,
+      "grad_norm": 0.16839970648288727,
+      "learning_rate": 1.9294629780335895e-05,
+      "loss": 1.2014,
+      "step": 4109
+    },
+    {
+      "epoch": 1.5301741690032693,
+      "grad_norm": 0.17429544031620026,
+      "learning_rate": 1.9294181362794745e-05,
+      "loss": 1.2086,
+      "step": 4110
+    },
+    {
+      "epoch": 1.5305464741538783,
+      "grad_norm": 0.16927529871463776,
+      "learning_rate": 1.9293732807979127e-05,
+      "loss": 1.222,
+      "step": 4111
+    },
+    {
+      "epoch": 1.5309187793044874,
+      "grad_norm": 0.17815326154232025,
+      "learning_rate": 1.9293284115895656e-05,
+      "loss": 1.2107,
+      "step": 4112
+    },
+    {
+      "epoch": 1.5312910844550967,
+      "grad_norm": 0.17721591889858246,
+      "learning_rate": 1.929283528655096e-05,
+      "loss": 1.2062,
+      "step": 4113
+    },
+    {
+      "epoch": 1.5316633896057055,
+      "grad_norm": 0.16361045837402344,
+      "learning_rate": 1.929238631995167e-05,
+      "loss": 1.2002,
+      "step": 4114
+    },
+    {
+      "epoch": 1.5320356947563147,
+      "grad_norm": 0.17079226672649384,
+      "learning_rate": 1.929193721610442e-05,
+      "loss": 1.1955,
+      "step": 4115
+    },
+    {
+      "epoch": 1.5324079999069236,
+      "grad_norm": 0.17936839163303375,
+      "learning_rate": 1.9291487975015835e-05,
+      "loss": 1.2175,
+      "step": 4116
+    },
+    {
+      "epoch": 1.5327803050575328,
+      "grad_norm": 0.17177017033100128,
+      "learning_rate": 1.9291038596692562e-05,
+      "loss": 1.2078,
+      "step": 4117
+    },
+    {
+      "epoch": 1.5331526102081419,
+      "grad_norm": 0.1675311028957367,
+      "learning_rate": 1.929058908114123e-05,
+      "loss": 1.2289,
+      "step": 4118
+    },
+    {
+      "epoch": 1.533524915358751,
+      "grad_norm": 0.18256328999996185,
+      "learning_rate": 1.9290139428368482e-05,
+      "loss": 1.2118,
+      "step": 4119
+    },
+    {
+      "epoch": 1.53389722050936,
+      "grad_norm": 0.16636022925376892,
+      "learning_rate": 1.9289689638380956e-05,
+      "loss": 1.2061,
+      "step": 4120
+    },
+    {
+      "epoch": 1.534269525659969,
+      "grad_norm": 0.22456076741218567,
+      "learning_rate": 1.9289239711185293e-05,
+      "loss": 1.2134,
+      "step": 4121
+    },
+    {
+      "epoch": 1.5346418308105783,
+      "grad_norm": 0.19078871607780457,
+      "learning_rate": 1.928878964678815e-05,
+      "loss": 1.2185,
+      "step": 4122
+    },
+    {
+      "epoch": 1.535014135961187,
+      "grad_norm": 0.18136771023273468,
+      "learning_rate": 1.928833944519616e-05,
+      "loss": 1.2092,
+      "step": 4123
+    },
+    {
+      "epoch": 1.5353864411117963,
+      "grad_norm": 0.17033730447292328,
+      "learning_rate": 1.9287889106415983e-05,
+      "loss": 1.2218,
+      "step": 4124
+    },
+    {
+      "epoch": 1.5357587462624052,
+      "grad_norm": 0.17553742229938507,
+      "learning_rate": 1.9287438630454268e-05,
+      "loss": 1.2242,
+      "step": 4125
+    },
+    {
+      "epoch": 1.5361310514130144,
+      "grad_norm": 0.18240036070346832,
+      "learning_rate": 1.9286988017317664e-05,
+      "loss": 1.2029,
+      "step": 4126
+    },
+    {
+      "epoch": 1.5365033565636235,
+      "grad_norm": 0.17878971993923187,
+      "learning_rate": 1.928653726701283e-05,
+      "loss": 1.2057,
+      "step": 4127
+    },
+    {
+      "epoch": 1.5368756617142325,
+      "grad_norm": 0.16980788111686707,
+      "learning_rate": 1.9286086379546427e-05,
+      "loss": 1.2071,
+      "step": 4128
+    },
+    {
+      "epoch": 1.5372479668648416,
+      "grad_norm": 0.17709119617938995,
+      "learning_rate": 1.9285635354925107e-05,
+      "loss": 1.1857,
+      "step": 4129
+    },
+    {
+      "epoch": 1.5376202720154506,
+      "grad_norm": 0.1764722466468811,
+      "learning_rate": 1.9285184193155536e-05,
+      "loss": 1.2124,
+      "step": 4130
+    },
+    {
+      "epoch": 1.5379925771660599,
+      "grad_norm": 0.18060049414634705,
+      "learning_rate": 1.9284732894244378e-05,
+      "loss": 1.2073,
+      "step": 4131
+    },
+    {
+      "epoch": 1.5383648823166687,
+      "grad_norm": 0.16817817091941833,
+      "learning_rate": 1.92842814581983e-05,
+      "loss": 1.2128,
+      "step": 4132
+    },
+    {
+      "epoch": 1.538737187467278,
+      "grad_norm": 0.1727471500635147,
+      "learning_rate": 1.9283829885023967e-05,
+      "loss": 1.1974,
+      "step": 4133
+    },
+    {
+      "epoch": 1.539109492617887,
+      "grad_norm": 0.1839704066514969,
+      "learning_rate": 1.9283378174728046e-05,
+      "loss": 1.2003,
+      "step": 4134
+    },
+    {
+      "epoch": 1.539481797768496,
+      "grad_norm": 0.17027804255485535,
+      "learning_rate": 1.9282926327317213e-05,
+      "loss": 1.2135,
+      "step": 4135
+    },
+    {
+      "epoch": 1.539854102919105,
+      "grad_norm": 0.1670842319726944,
+      "learning_rate": 1.9282474342798143e-05,
+      "loss": 1.2023,
+      "step": 4136
+    },
+    {
+      "epoch": 1.5402264080697141,
+      "grad_norm": 0.1733679473400116,
+      "learning_rate": 1.9282022221177507e-05,
+      "loss": 1.2102,
+      "step": 4137
+    },
+    {
+      "epoch": 1.5405987132203232,
+      "grad_norm": 0.17217591404914856,
+      "learning_rate": 1.9281569962461986e-05,
+      "loss": 1.2237,
+      "step": 4138
+    },
+    {
+      "epoch": 1.5409710183709322,
+      "grad_norm": 0.174224391579628,
+      "learning_rate": 1.928111756665826e-05,
+      "loss": 1.2072,
+      "step": 4139
+    },
+    {
+      "epoch": 1.5413433235215415,
+      "grad_norm": 0.17632028460502625,
+      "learning_rate": 1.928066503377301e-05,
+      "loss": 1.2182,
+      "step": 4140
+    },
+    {
+      "epoch": 1.5417156286721503,
+      "grad_norm": 0.17433500289916992,
+      "learning_rate": 1.9280212363812918e-05,
+      "loss": 1.2297,
+      "step": 4141
+    },
+    {
+      "epoch": 1.5420879338227595,
+      "grad_norm": 0.1642346978187561,
+      "learning_rate": 1.9279759556784673e-05,
+      "loss": 1.2103,
+      "step": 4142
+    },
+    {
+      "epoch": 1.5424602389733686,
+      "grad_norm": 0.172978937625885,
+      "learning_rate": 1.9279306612694963e-05,
+      "loss": 1.2225,
+      "step": 4143
+    },
+    {
+      "epoch": 1.5428325441239776,
+      "grad_norm": 0.17284728586673737,
+      "learning_rate": 1.9278853531550475e-05,
+      "loss": 1.2209,
+      "step": 4144
+    },
+    {
+      "epoch": 1.5432048492745867,
+      "grad_norm": 0.16585254669189453,
+      "learning_rate": 1.9278400313357902e-05,
+      "loss": 1.2256,
+      "step": 4145
+    },
+    {
+      "epoch": 1.5435771544251957,
+      "grad_norm": 0.1627526581287384,
+      "learning_rate": 1.927794695812394e-05,
+      "loss": 1.2172,
+      "step": 4146
+    },
+    {
+      "epoch": 1.543949459575805,
+      "grad_norm": 0.17496748268604279,
+      "learning_rate": 1.9277493465855287e-05,
+      "loss": 1.197,
+      "step": 4147
+    },
+    {
+      "epoch": 1.5443217647264138,
+      "grad_norm": 0.1616668850183487,
+      "learning_rate": 1.9277039836558635e-05,
+      "loss": 1.2088,
+      "step": 4148
+    },
+    {
+      "epoch": 1.544694069877023,
+      "grad_norm": 0.16618521511554718,
+      "learning_rate": 1.9276586070240684e-05,
+      "loss": 1.215,
+      "step": 4149
+    },
+    {
+      "epoch": 1.5450663750276319,
+      "grad_norm": 0.16764888167381287,
+      "learning_rate": 1.927613216690814e-05,
+      "loss": 1.2012,
+      "step": 4150
+    },
+    {
+      "epoch": 1.5454386801782412,
+      "grad_norm": 0.166963592171669,
+      "learning_rate": 1.9275678126567707e-05,
+      "loss": 1.2234,
+      "step": 4151
+    },
+    {
+      "epoch": 1.5458109853288502,
+      "grad_norm": 0.166300967335701,
+      "learning_rate": 1.927522394922609e-05,
+      "loss": 1.2119,
+      "step": 4152
+    },
+    {
+      "epoch": 1.5461832904794592,
+      "grad_norm": 0.16552546620368958,
+      "learning_rate": 1.9274769634889997e-05,
+      "loss": 1.1968,
+      "step": 4153
+    },
+    {
+      "epoch": 1.5465555956300683,
+      "grad_norm": 0.1656341403722763,
+      "learning_rate": 1.927431518356614e-05,
+      "loss": 1.2084,
+      "step": 4154
+    },
+    {
+      "epoch": 1.5469279007806773,
+      "grad_norm": 0.16084742546081543,
+      "learning_rate": 1.9273860595261232e-05,
+      "loss": 1.2007,
+      "step": 4155
+    },
+    {
+      "epoch": 1.5473002059312866,
+      "grad_norm": 0.16208963096141815,
+      "learning_rate": 1.927340586998198e-05,
+      "loss": 1.2251,
+      "step": 4156
+    },
+    {
+      "epoch": 1.5476725110818954,
+      "grad_norm": 0.16170455515384674,
+      "learning_rate": 1.9272951007735108e-05,
+      "loss": 1.2208,
+      "step": 4157
+    },
+    {
+      "epoch": 1.5480448162325047,
+      "grad_norm": 0.159234881401062,
+      "learning_rate": 1.927249600852733e-05,
+      "loss": 1.2103,
+      "step": 4158
+    },
+    {
+      "epoch": 1.5484171213831135,
+      "grad_norm": 0.17041054368019104,
+      "learning_rate": 1.927204087236537e-05,
+      "loss": 1.2171,
+      "step": 4159
+    },
+    {
+      "epoch": 1.5487894265337228,
+      "grad_norm": 0.16318999230861664,
+      "learning_rate": 1.9271585599255945e-05,
+      "loss": 1.2101,
+      "step": 4160
+    },
+    {
+      "epoch": 1.5491617316843318,
+      "grad_norm": 0.15880700945854187,
+      "learning_rate": 1.9271130189205786e-05,
+      "loss": 1.1997,
+      "step": 4161
+    },
+    {
+      "epoch": 1.5495340368349408,
+      "grad_norm": 0.16643404960632324,
+      "learning_rate": 1.9270674642221614e-05,
+      "loss": 1.193,
+      "step": 4162
+    },
+    {
+      "epoch": 1.5499063419855499,
+      "grad_norm": 0.17623448371887207,
+      "learning_rate": 1.927021895831016e-05,
+      "loss": 1.2188,
+      "step": 4163
+    },
+    {
+      "epoch": 1.550278647136159,
+      "grad_norm": 0.1635187864303589,
+      "learning_rate": 1.9269763137478152e-05,
+      "loss": 1.1997,
+      "step": 4164
+    },
+    {
+      "epoch": 1.5506509522867682,
+      "grad_norm": 0.16812646389007568,
+      "learning_rate": 1.9269307179732325e-05,
+      "loss": 1.2083,
+      "step": 4165
+    },
+    {
+      "epoch": 1.551023257437377,
+      "grad_norm": 0.164559006690979,
+      "learning_rate": 1.9268851085079415e-05,
+      "loss": 1.1901,
+      "step": 4166
+    },
+    {
+      "epoch": 1.5513955625879863,
+      "grad_norm": 0.1598326563835144,
+      "learning_rate": 1.9268394853526156e-05,
+      "loss": 1.198,
+      "step": 4167
+    },
+    {
+      "epoch": 1.551767867738595,
+      "grad_norm": 0.16687574982643127,
+      "learning_rate": 1.9267938485079285e-05,
+      "loss": 1.2133,
+      "step": 4168
+    },
+    {
+      "epoch": 1.5521401728892044,
+      "grad_norm": 0.16783174872398376,
+      "learning_rate": 1.9267481979745544e-05,
+      "loss": 1.2217,
+      "step": 4169
+    },
+    {
+      "epoch": 1.5525124780398134,
+      "grad_norm": 0.1639779806137085,
+      "learning_rate": 1.9267025337531678e-05,
+      "loss": 1.2243,
+      "step": 4170
+    },
+    {
+      "epoch": 1.5528847831904224,
+      "grad_norm": 0.16327700018882751,
+      "learning_rate": 1.9266568558444426e-05,
+      "loss": 1.2213,
+      "step": 4171
+    },
+    {
+      "epoch": 1.5532570883410315,
+      "grad_norm": 0.16662442684173584,
+      "learning_rate": 1.926611164249054e-05,
+      "loss": 1.198,
+      "step": 4172
+    },
+    {
+      "epoch": 1.5536293934916405,
+      "grad_norm": 0.1651596873998642,
+      "learning_rate": 1.9265654589676767e-05,
+      "loss": 1.2113,
+      "step": 4173
+    },
+    {
+      "epoch": 1.5540016986422498,
+      "grad_norm": 0.16614238917827606,
+      "learning_rate": 1.9265197400009854e-05,
+      "loss": 1.2092,
+      "step": 4174
+    },
+    {
+      "epoch": 1.5543740037928586,
+      "grad_norm": 0.1736796498298645,
+      "learning_rate": 1.926474007349656e-05,
+      "loss": 1.2147,
+      "step": 4175
+    },
+    {
+      "epoch": 1.5547463089434679,
+      "grad_norm": 0.1689198911190033,
+      "learning_rate": 1.9264282610143638e-05,
+      "loss": 1.2119,
+      "step": 4176
+    },
+    {
+      "epoch": 1.5551186140940767,
+      "grad_norm": 0.16373711824417114,
+      "learning_rate": 1.926382500995784e-05,
+      "loss": 1.2077,
+      "step": 4177
+    },
+    {
+      "epoch": 1.555490919244686,
+      "grad_norm": 0.1627718210220337,
+      "learning_rate": 1.9263367272945927e-05,
+      "loss": 1.2128,
+      "step": 4178
+    },
+    {
+      "epoch": 1.555863224395295,
+      "grad_norm": 0.16306252777576447,
+      "learning_rate": 1.9262909399114663e-05,
+      "loss": 1.218,
+      "step": 4179
+    },
+    {
+      "epoch": 1.556235529545904,
+      "grad_norm": 0.16775202751159668,
+      "learning_rate": 1.9262451388470807e-05,
+      "loss": 1.2039,
+      "step": 4180
+    },
+    {
+      "epoch": 1.556607834696513,
+      "grad_norm": 0.1653129607439041,
+      "learning_rate": 1.926199324102113e-05,
+      "loss": 1.2081,
+      "step": 4181
+    },
+    {
+      "epoch": 1.5569801398471221,
+      "grad_norm": 0.1624881476163864,
+      "learning_rate": 1.9261534956772386e-05,
+      "loss": 1.2163,
+      "step": 4182
+    },
+    {
+      "epoch": 1.5573524449977314,
+      "grad_norm": 0.16071651875972748,
+      "learning_rate": 1.9261076535731356e-05,
+      "loss": 1.2008,
+      "step": 4183
+    },
+    {
+      "epoch": 1.5577247501483402,
+      "grad_norm": 0.17238472402095795,
+      "learning_rate": 1.926061797790481e-05,
+      "loss": 1.221,
+      "step": 4184
+    },
+    {
+      "epoch": 1.5580970552989495,
+      "grad_norm": 0.17098172008991241,
+      "learning_rate": 1.9260159283299514e-05,
+      "loss": 1.206,
+      "step": 4185
+    },
+    {
+      "epoch": 1.5584693604495583,
+      "grad_norm": 0.164502814412117,
+      "learning_rate": 1.9259700451922247e-05,
+      "loss": 1.2157,
+      "step": 4186
+    },
+    {
+      "epoch": 1.5588416656001676,
+      "grad_norm": 0.17409628629684448,
+      "learning_rate": 1.9259241483779787e-05,
+      "loss": 1.226,
+      "step": 4187
+    },
+    {
+      "epoch": 1.5592139707507766,
+      "grad_norm": 0.17351455986499786,
+      "learning_rate": 1.9258782378878908e-05,
+      "loss": 1.2167,
+      "step": 4188
+    },
+    {
+      "epoch": 1.5595862759013857,
+      "grad_norm": 0.16940516233444214,
+      "learning_rate": 1.92583231372264e-05,
+      "loss": 1.2044,
+      "step": 4189
+    },
+    {
+      "epoch": 1.5599585810519947,
+      "grad_norm": 0.16605758666992188,
+      "learning_rate": 1.9257863758829038e-05,
+      "loss": 1.2009,
+      "step": 4190
+    },
+    {
+      "epoch": 1.5603308862026037,
+      "grad_norm": 0.17315179109573364,
+      "learning_rate": 1.9257404243693606e-05,
+      "loss": 1.2139,
+      "step": 4191
+    },
+    {
+      "epoch": 1.560703191353213,
+      "grad_norm": 0.1688748598098755,
+      "learning_rate": 1.9256944591826893e-05,
+      "loss": 1.2212,
+      "step": 4192
+    },
+    {
+      "epoch": 1.5610754965038218,
+      "grad_norm": 0.16964520514011383,
+      "learning_rate": 1.9256484803235696e-05,
+      "loss": 1.2092,
+      "step": 4193
+    },
+    {
+      "epoch": 1.561447801654431,
+      "grad_norm": 0.16854505240917206,
+      "learning_rate": 1.9256024877926796e-05,
+      "loss": 1.2015,
+      "step": 4194
+    },
+    {
+      "epoch": 1.5618201068050401,
+      "grad_norm": 0.168868288397789,
+      "learning_rate": 1.9255564815906988e-05,
+      "loss": 1.2013,
+      "step": 4195
+    },
+    {
+      "epoch": 1.5621924119556492,
+      "grad_norm": 0.17103056609630585,
+      "learning_rate": 1.9255104617183068e-05,
+      "loss": 1.211,
+      "step": 4196
+    },
+    {
+      "epoch": 1.5625647171062582,
+      "grad_norm": 0.1685587465763092,
+      "learning_rate": 1.9254644281761838e-05,
+      "loss": 1.1984,
+      "step": 4197
+    },
+    {
+      "epoch": 1.5629370222568673,
+      "grad_norm": 0.17588046193122864,
+      "learning_rate": 1.925418380965009e-05,
+      "loss": 1.2144,
+      "step": 4198
+    },
+    {
+      "epoch": 1.5633093274074763,
+      "grad_norm": 0.16657619178295135,
+      "learning_rate": 1.925372320085463e-05,
+      "loss": 1.2011,
+      "step": 4199
+    },
+    {
+      "epoch": 1.5636816325580853,
+      "grad_norm": 0.16856853663921356,
+      "learning_rate": 1.9253262455382256e-05,
+      "loss": 1.2057,
+      "step": 4200
+    },
+    {
+      "epoch": 1.5640539377086946,
+      "grad_norm": 0.16726350784301758,
+      "learning_rate": 1.925280157323978e-05,
+      "loss": 1.2062,
+      "step": 4201
+    },
+    {
+      "epoch": 1.5644262428593034,
+      "grad_norm": 0.18542660772800446,
+      "learning_rate": 1.9252340554434003e-05,
+      "loss": 1.21,
+      "step": 4202
+    },
+    {
+      "epoch": 1.5647985480099127,
+      "grad_norm": 0.17337799072265625,
+      "learning_rate": 1.9251879398971733e-05,
+      "loss": 1.203,
+      "step": 4203
+    },
+    {
+      "epoch": 1.5651708531605217,
+      "grad_norm": 0.1717308908700943,
+      "learning_rate": 1.925141810685979e-05,
+      "loss": 1.22,
+      "step": 4204
+    },
+    {
+      "epoch": 1.5655431583111308,
+      "grad_norm": 0.1700487732887268,
+      "learning_rate": 1.925095667810498e-05,
+      "loss": 1.2018,
+      "step": 4205
+    },
+    {
+      "epoch": 1.5659154634617398,
+      "grad_norm": 0.1768191158771515,
+      "learning_rate": 1.9250495112714123e-05,
+      "loss": 1.1853,
+      "step": 4206
+    },
+    {
+      "epoch": 1.5662877686123489,
+      "grad_norm": 0.17097093164920807,
+      "learning_rate": 1.9250033410694032e-05,
+      "loss": 1.2047,
+      "step": 4207
+    },
+    {
+      "epoch": 1.5666600737629581,
+      "grad_norm": 0.1712087094783783,
+      "learning_rate": 1.9249571572051528e-05,
+      "loss": 1.2028,
+      "step": 4208
+    },
+    {
+      "epoch": 1.567032378913567,
+      "grad_norm": 0.16526564955711365,
+      "learning_rate": 1.924910959679343e-05,
+      "loss": 1.205,
+      "step": 4209
+    },
+    {
+      "epoch": 1.5674046840641762,
+      "grad_norm": 0.16260464489459991,
+      "learning_rate": 1.9248647484926568e-05,
+      "loss": 1.201,
+      "step": 4210
+    },
+    {
+      "epoch": 1.567776989214785,
+      "grad_norm": 0.1643940955400467,
+      "learning_rate": 1.924818523645776e-05,
+      "loss": 1.2147,
+      "step": 4211
+    },
+    {
+      "epoch": 1.5681492943653943,
+      "grad_norm": 0.17409846186637878,
+      "learning_rate": 1.9247722851393838e-05,
+      "loss": 1.2043,
+      "step": 4212
+    },
+    {
+      "epoch": 1.5685215995160033,
+      "grad_norm": 0.16997475922107697,
+      "learning_rate": 1.924726032974163e-05,
+      "loss": 1.2171,
+      "step": 4213
+    },
+    {
+      "epoch": 1.5688939046666124,
+      "grad_norm": 0.171822652220726,
+      "learning_rate": 1.9246797671507966e-05,
+      "loss": 1.2126,
+      "step": 4214
+    },
+    {
+      "epoch": 1.5692662098172214,
+      "grad_norm": 0.1666400134563446,
+      "learning_rate": 1.9246334876699682e-05,
+      "loss": 1.2152,
+      "step": 4215
+    },
+    {
+      "epoch": 1.5696385149678305,
+      "grad_norm": 0.17441390454769135,
+      "learning_rate": 1.924587194532361e-05,
+      "loss": 1.2002,
+      "step": 4216
+    },
+    {
+      "epoch": 1.5700108201184397,
+      "grad_norm": 0.18336564302444458,
+      "learning_rate": 1.9245408877386592e-05,
+      "loss": 1.2119,
+      "step": 4217
+    },
+    {
+      "epoch": 1.5703831252690486,
+      "grad_norm": 0.16609658300876617,
+      "learning_rate": 1.9244945672895464e-05,
+      "loss": 1.2084,
+      "step": 4218
+    },
+    {
+      "epoch": 1.5707554304196578,
+      "grad_norm": 0.17070931196212769,
+      "learning_rate": 1.924448233185707e-05,
+      "loss": 1.2045,
+      "step": 4219
+    },
+    {
+      "epoch": 1.5711277355702666,
+      "grad_norm": 0.16091009974479675,
+      "learning_rate": 1.924401885427825e-05,
+      "loss": 1.1902,
+      "step": 4220
+    },
+    {
+      "epoch": 1.571500040720876,
+      "grad_norm": 0.1693873405456543,
+      "learning_rate": 1.9243555240165855e-05,
+      "loss": 1.2213,
+      "step": 4221
+    },
+    {
+      "epoch": 1.571872345871485,
+      "grad_norm": 0.17035622894763947,
+      "learning_rate": 1.9243091489526728e-05,
+      "loss": 1.2,
+      "step": 4222
+    },
+    {
+      "epoch": 1.572244651022094,
+      "grad_norm": 0.17594295740127563,
+      "learning_rate": 1.9242627602367716e-05,
+      "loss": 1.2161,
+      "step": 4223
+    },
+    {
+      "epoch": 1.572616956172703,
+      "grad_norm": 0.16813448071479797,
+      "learning_rate": 1.924216357869568e-05,
+      "loss": 1.2029,
+      "step": 4224
+    },
+    {
+      "epoch": 1.572989261323312,
+      "grad_norm": 0.18052363395690918,
+      "learning_rate": 1.924169941851746e-05,
+      "loss": 1.2017,
+      "step": 4225
+    },
+    {
+      "epoch": 1.5733615664739213,
+      "grad_norm": 0.18110094964504242,
+      "learning_rate": 1.9241235121839927e-05,
+      "loss": 1.1987,
+      "step": 4226
+    },
+    {
+      "epoch": 1.5737338716245302,
+      "grad_norm": 0.17271529138088226,
+      "learning_rate": 1.9240770688669928e-05,
+      "loss": 1.2058,
+      "step": 4227
+    },
+    {
+      "epoch": 1.5741061767751394,
+      "grad_norm": 0.1713894158601761,
+      "learning_rate": 1.924030611901433e-05,
+      "loss": 1.2011,
+      "step": 4228
+    },
+    {
+      "epoch": 1.5744784819257482,
+      "grad_norm": 0.17483104765415192,
+      "learning_rate": 1.9239841412879983e-05,
+      "loss": 1.2166,
+      "step": 4229
+    },
+    {
+      "epoch": 1.5748507870763575,
+      "grad_norm": 0.17396202683448792,
+      "learning_rate": 1.9239376570273763e-05,
+      "loss": 1.2059,
+      "step": 4230
+    },
+    {
+      "epoch": 1.5752230922269665,
+      "grad_norm": 0.16693198680877686,
+      "learning_rate": 1.923891159120253e-05,
+      "loss": 1.2113,
+      "step": 4231
+    },
+    {
+      "epoch": 1.5755953973775756,
+      "grad_norm": 0.17103157937526703,
+      "learning_rate": 1.9238446475673155e-05,
+      "loss": 1.2166,
+      "step": 4232
+    },
+    {
+      "epoch": 1.5759677025281846,
+      "grad_norm": 0.1680799126625061,
+      "learning_rate": 1.92379812236925e-05,
+      "loss": 1.2189,
+      "step": 4233
+    },
+    {
+      "epoch": 1.5763400076787937,
+      "grad_norm": 0.18774433434009552,
+      "learning_rate": 1.9237515835267447e-05,
+      "loss": 1.2121,
+      "step": 4234
+    },
+    {
+      "epoch": 1.576712312829403,
+      "grad_norm": 0.16608703136444092,
+      "learning_rate": 1.9237050310404862e-05,
+      "loss": 1.1935,
+      "step": 4235
+    },
+    {
+      "epoch": 1.5770846179800118,
+      "grad_norm": 0.17058461904525757,
+      "learning_rate": 1.9236584649111628e-05,
+      "loss": 1.2014,
+      "step": 4236
+    },
+    {
+      "epoch": 1.577456923130621,
+      "grad_norm": 0.16331979632377625,
+      "learning_rate": 1.923611885139461e-05,
+      "loss": 1.2044,
+      "step": 4237
+    },
+    {
+      "epoch": 1.5778292282812298,
+      "grad_norm": 0.17175064980983734,
+      "learning_rate": 1.9235652917260705e-05,
+      "loss": 1.2032,
+      "step": 4238
+    },
+    {
+      "epoch": 1.578201533431839,
+      "grad_norm": 0.1720954328775406,
+      "learning_rate": 1.9235186846716784e-05,
+      "loss": 1.2092,
+      "step": 4239
+    },
+    {
+      "epoch": 1.5785738385824482,
+      "grad_norm": 0.16295336186885834,
+      "learning_rate": 1.923472063976973e-05,
+      "loss": 1.1955,
+      "step": 4240
+    },
+    {
+      "epoch": 1.5789461437330572,
+      "grad_norm": 0.16375908255577087,
+      "learning_rate": 1.923425429642643e-05,
+      "loss": 1.1909,
+      "step": 4241
+    },
+    {
+      "epoch": 1.5793184488836662,
+      "grad_norm": 0.1621014028787613,
+      "learning_rate": 1.923378781669378e-05,
+      "loss": 1.201,
+      "step": 4242
+    },
+    {
+      "epoch": 1.5796907540342753,
+      "grad_norm": 0.1648714393377304,
+      "learning_rate": 1.9233321200578657e-05,
+      "loss": 1.2119,
+      "step": 4243
+    },
+    {
+      "epoch": 1.5800630591848845,
+      "grad_norm": 0.16513188183307648,
+      "learning_rate": 1.9232854448087962e-05,
+      "loss": 1.2129,
+      "step": 4244
+    },
+    {
+      "epoch": 1.5804353643354934,
+      "grad_norm": 0.1719852089881897,
+      "learning_rate": 1.9232387559228587e-05,
+      "loss": 1.2095,
+      "step": 4245
+    },
+    {
+      "epoch": 1.5808076694861026,
+      "grad_norm": 0.16509710252285004,
+      "learning_rate": 1.9231920534007428e-05,
+      "loss": 1.2118,
+      "step": 4246
+    },
+    {
+      "epoch": 1.5811799746367114,
+      "grad_norm": 0.16327592730522156,
+      "learning_rate": 1.923145337243138e-05,
+      "loss": 1.2081,
+      "step": 4247
+    },
+    {
+      "epoch": 1.5815522797873207,
+      "grad_norm": 0.17513926327228546,
+      "learning_rate": 1.9230986074507347e-05,
+      "loss": 1.2215,
+      "step": 4248
+    },
+    {
+      "epoch": 1.5819245849379298,
+      "grad_norm": 0.17169274389743805,
+      "learning_rate": 1.9230518640242228e-05,
+      "loss": 1.2164,
+      "step": 4249
+    },
+    {
+      "epoch": 1.5822968900885388,
+      "grad_norm": 0.17580832540988922,
+      "learning_rate": 1.9230051069642927e-05,
+      "loss": 1.2066,
+      "step": 4250
+    },
+    {
+      "epoch": 1.5826691952391478,
+      "grad_norm": 0.16311529278755188,
+      "learning_rate": 1.922958336271635e-05,
+      "loss": 1.2061,
+      "step": 4251
+    },
+    {
+      "epoch": 1.5830415003897569,
+      "grad_norm": 0.18009954690933228,
+      "learning_rate": 1.922911551946941e-05,
+      "loss": 1.2161,
+      "step": 4252
+    },
+    {
+      "epoch": 1.5834138055403661,
+      "grad_norm": 0.17402635514736176,
+      "learning_rate": 1.922864753990901e-05,
+      "loss": 1.228,
+      "step": 4253
+    },
+    {
+      "epoch": 1.583786110690975,
+      "grad_norm": 0.17226436734199524,
+      "learning_rate": 1.9228179424042064e-05,
+      "loss": 1.209,
+      "step": 4254
+    },
+    {
+      "epoch": 1.5841584158415842,
+      "grad_norm": 0.1807158887386322,
+      "learning_rate": 1.9227711171875486e-05,
+      "loss": 1.202,
+      "step": 4255
+    },
+    {
+      "epoch": 1.5845307209921933,
+      "grad_norm": 0.16312603652477264,
+      "learning_rate": 1.92272427834162e-05,
+      "loss": 1.2014,
+      "step": 4256
+    },
+    {
+      "epoch": 1.5849030261428023,
+      "grad_norm": 0.16579662263393402,
+      "learning_rate": 1.9226774258671112e-05,
+      "loss": 1.208,
+      "step": 4257
+    },
+    {
+      "epoch": 1.5852753312934114,
+      "grad_norm": 0.1716030240058899,
+      "learning_rate": 1.9226305597647145e-05,
+      "loss": 1.2052,
+      "step": 4258
+    },
+    {
+      "epoch": 1.5856476364440204,
+      "grad_norm": 0.16442036628723145,
+      "learning_rate": 1.922583680035123e-05,
+      "loss": 1.2109,
+      "step": 4259
+    },
+    {
+      "epoch": 1.5860199415946294,
+      "grad_norm": 0.15719719231128693,
+      "learning_rate": 1.922536786679028e-05,
+      "loss": 1.2087,
+      "step": 4260
+    },
+    {
+      "epoch": 1.5863922467452385,
+      "grad_norm": 0.16500256955623627,
+      "learning_rate": 1.9224898796971224e-05,
+      "loss": 1.2206,
+      "step": 4261
+    },
+    {
+      "epoch": 1.5867645518958478,
+      "grad_norm": 0.17196062207221985,
+      "learning_rate": 1.9224429590900997e-05,
+      "loss": 1.206,
+      "step": 4262
+    },
+    {
+      "epoch": 1.5871368570464566,
+      "grad_norm": 0.17409005761146545,
+      "learning_rate": 1.9223960248586523e-05,
+      "loss": 1.2111,
+      "step": 4263
+    },
+    {
+      "epoch": 1.5875091621970658,
+      "grad_norm": 0.17429877817630768,
+      "learning_rate": 1.922349077003473e-05,
+      "loss": 1.2316,
+      "step": 4264
+    },
+    {
+      "epoch": 1.5878814673476749,
+      "grad_norm": 0.1687888205051422,
+      "learning_rate": 1.922302115525256e-05,
+      "loss": 1.2126,
+      "step": 4265
+    },
+    {
+      "epoch": 1.588253772498284,
+      "grad_norm": 0.1785813719034195,
+      "learning_rate": 1.922255140424695e-05,
+      "loss": 1.2094,
+      "step": 4266
+    },
+    {
+      "epoch": 1.588626077648893,
+      "grad_norm": 0.1841798722743988,
+      "learning_rate": 1.922208151702483e-05,
+      "loss": 1.2105,
+      "step": 4267
+    },
+    {
+      "epoch": 1.588998382799502,
+      "grad_norm": 0.16933433711528778,
+      "learning_rate": 1.9221611493593145e-05,
+      "loss": 1.2078,
+      "step": 4268
+    },
+    {
+      "epoch": 1.5893706879501113,
+      "grad_norm": 0.21182604134082794,
+      "learning_rate": 1.9221141333958837e-05,
+      "loss": 1.2086,
+      "step": 4269
+    },
+    {
+      "epoch": 1.58974299310072,
+      "grad_norm": 0.18182890117168427,
+      "learning_rate": 1.922067103812885e-05,
+      "loss": 1.2054,
+      "step": 4270
+    },
+    {
+      "epoch": 1.5901152982513294,
+      "grad_norm": 0.18493826687335968,
+      "learning_rate": 1.9220200606110132e-05,
+      "loss": 1.2018,
+      "step": 4271
+    },
+    {
+      "epoch": 1.5904876034019382,
+      "grad_norm": 0.17452014982700348,
+      "learning_rate": 1.921973003790963e-05,
+      "loss": 1.2287,
+      "step": 4272
+    },
+    {
+      "epoch": 1.5908599085525474,
+      "grad_norm": 0.16467060148715973,
+      "learning_rate": 1.9219259333534292e-05,
+      "loss": 1.2003,
+      "step": 4273
+    },
+    {
+      "epoch": 1.5912322137031565,
+      "grad_norm": 0.18403173983097076,
+      "learning_rate": 1.9218788492991075e-05,
+      "loss": 1.228,
+      "step": 4274
+    },
+    {
+      "epoch": 1.5916045188537655,
+      "grad_norm": 0.18004946410655975,
+      "learning_rate": 1.921831751628693e-05,
+      "loss": 1.202,
+      "step": 4275
+    },
+    {
+      "epoch": 1.5919768240043746,
+      "grad_norm": 0.16663858294487,
+      "learning_rate": 1.921784640342881e-05,
+      "loss": 1.2187,
+      "step": 4276
+    },
+    {
+      "epoch": 1.5923491291549836,
+      "grad_norm": 0.17807120084762573,
+      "learning_rate": 1.921737515442368e-05,
+      "loss": 1.2171,
+      "step": 4277
+    },
+    {
+      "epoch": 1.5927214343055929,
+      "grad_norm": 0.18053670227527618,
+      "learning_rate": 1.9216903769278498e-05,
+      "loss": 1.2175,
+      "step": 4278
+    },
+    {
+      "epoch": 1.5930937394562017,
+      "grad_norm": 0.17122212052345276,
+      "learning_rate": 1.9216432248000224e-05,
+      "loss": 1.2032,
+      "step": 4279
+    },
+    {
+      "epoch": 1.593466044606811,
+      "grad_norm": 0.16505327820777893,
+      "learning_rate": 1.9215960590595824e-05,
+      "loss": 1.2039,
+      "step": 4280
+    },
+    {
+      "epoch": 1.5938383497574198,
+      "grad_norm": 0.18405233323574066,
+      "learning_rate": 1.9215488797072267e-05,
+      "loss": 1.212,
+      "step": 4281
+    },
+    {
+      "epoch": 1.594210654908029,
+      "grad_norm": 0.17196938395500183,
+      "learning_rate": 1.9215016867436516e-05,
+      "loss": 1.2118,
+      "step": 4282
+    },
+    {
+      "epoch": 1.594582960058638,
+      "grad_norm": 0.16480191051959991,
+      "learning_rate": 1.9214544801695547e-05,
+      "loss": 1.2089,
+      "step": 4283
+    },
+    {
+      "epoch": 1.5949552652092471,
+      "grad_norm": 0.16730982065200806,
+      "learning_rate": 1.9214072599856326e-05,
+      "loss": 1.2027,
+      "step": 4284
+    },
+    {
+      "epoch": 1.5953275703598562,
+      "grad_norm": 0.17092877626419067,
+      "learning_rate": 1.9213600261925832e-05,
+      "loss": 1.2149,
+      "step": 4285
+    },
+    {
+      "epoch": 1.5956998755104652,
+      "grad_norm": 0.1654144823551178,
+      "learning_rate": 1.9213127787911045e-05,
+      "loss": 1.2017,
+      "step": 4286
+    },
+    {
+      "epoch": 1.5960721806610745,
+      "grad_norm": 0.17047090828418732,
+      "learning_rate": 1.9212655177818935e-05,
+      "loss": 1.1947,
+      "step": 4287
+    },
+    {
+      "epoch": 1.5964444858116833,
+      "grad_norm": 0.17292653024196625,
+      "learning_rate": 1.9212182431656487e-05,
+      "loss": 1.2226,
+      "step": 4288
+    },
+    {
+      "epoch": 1.5968167909622926,
+      "grad_norm": 0.16899065673351288,
+      "learning_rate": 1.9211709549430678e-05,
+      "loss": 1.196,
+      "step": 4289
+    },
+    {
+      "epoch": 1.5971890961129014,
+      "grad_norm": 0.16645705699920654,
+      "learning_rate": 1.92112365311485e-05,
+      "loss": 1.2066,
+      "step": 4290
+    },
+    {
+      "epoch": 1.5975614012635106,
+      "grad_norm": 0.1680358499288559,
+      "learning_rate": 1.921076337681694e-05,
+      "loss": 1.2134,
+      "step": 4291
+    },
+    {
+      "epoch": 1.5979337064141197,
+      "grad_norm": 0.16596710681915283,
+      "learning_rate": 1.9210290086442983e-05,
+      "loss": 1.2085,
+      "step": 4292
+    },
+    {
+      "epoch": 1.5983060115647287,
+      "grad_norm": 0.1577194184064865,
+      "learning_rate": 1.9209816660033613e-05,
+      "loss": 1.1921,
+      "step": 4293
+    },
+    {
+      "epoch": 1.5986783167153378,
+      "grad_norm": 0.1676270067691803,
+      "learning_rate": 1.9209343097595834e-05,
+      "loss": 1.2075,
+      "step": 4294
+    },
+    {
+      "epoch": 1.5990506218659468,
+      "grad_norm": 0.16375909745693207,
+      "learning_rate": 1.9208869399136633e-05,
+      "loss": 1.2086,
+      "step": 4295
+    },
+    {
+      "epoch": 1.599422927016556,
+      "grad_norm": 0.16282792389392853,
+      "learning_rate": 1.9208395564663012e-05,
+      "loss": 1.2115,
+      "step": 4296
+    },
+    {
+      "epoch": 1.599795232167165,
+      "grad_norm": 0.16538910567760468,
+      "learning_rate": 1.9207921594181964e-05,
+      "loss": 1.2063,
+      "step": 4297
+    },
+    {
+      "epoch": 1.6001675373177742,
+      "grad_norm": 0.1656702756881714,
+      "learning_rate": 1.9207447487700494e-05,
+      "loss": 1.2123,
+      "step": 4298
+    },
+    {
+      "epoch": 1.600539842468383,
+      "grad_norm": 0.16824668645858765,
+      "learning_rate": 1.92069732452256e-05,
+      "loss": 1.2085,
+      "step": 4299
+    },
+    {
+      "epoch": 1.6009121476189923,
+      "grad_norm": 0.16301900148391724,
+      "learning_rate": 1.920649886676429e-05,
+      "loss": 1.2227,
+      "step": 4300
+    },
+    {
+      "epoch": 1.6012844527696013,
+      "grad_norm": 0.173927441239357,
+      "learning_rate": 1.920602435232357e-05,
+      "loss": 1.2114,
+      "step": 4301
+    },
+    {
+      "epoch": 1.6016567579202103,
+      "grad_norm": 0.17178916931152344,
+      "learning_rate": 1.9205549701910445e-05,
+      "loss": 1.2076,
+      "step": 4302
+    },
+    {
+      "epoch": 1.6020290630708194,
+      "grad_norm": 0.16292187571525574,
+      "learning_rate": 1.920507491553193e-05,
+      "loss": 1.2231,
+      "step": 4303
+    },
+    {
+      "epoch": 1.6024013682214284,
+      "grad_norm": 0.1758165806531906,
+      "learning_rate": 1.9204599993195038e-05,
+      "loss": 1.2088,
+      "step": 4304
+    },
+    {
+      "epoch": 1.6027736733720377,
+      "grad_norm": 0.1699971705675125,
+      "learning_rate": 1.920412493490678e-05,
+      "loss": 1.2018,
+      "step": 4305
+    },
+    {
+      "epoch": 1.6031459785226465,
+      "grad_norm": 0.16813789308071136,
+      "learning_rate": 1.920364974067418e-05,
+      "loss": 1.2054,
+      "step": 4306
+    },
+    {
+      "epoch": 1.6035182836732558,
+      "grad_norm": 0.16365012526512146,
+      "learning_rate": 1.9203174410504243e-05,
+      "loss": 1.2084,
+      "step": 4307
+    },
+    {
+      "epoch": 1.6038905888238648,
+      "grad_norm": 0.16683803498744965,
+      "learning_rate": 1.9202698944404002e-05,
+      "loss": 1.2064,
+      "step": 4308
+    },
+    {
+      "epoch": 1.6042628939744739,
+      "grad_norm": 0.16585657000541687,
+      "learning_rate": 1.9202223342380475e-05,
+      "loss": 1.2156,
+      "step": 4309
+    },
+    {
+      "epoch": 1.604635199125083,
+      "grad_norm": 0.16558226943016052,
+      "learning_rate": 1.9201747604440686e-05,
+      "loss": 1.2181,
+      "step": 4310
+    },
+    {
+      "epoch": 1.605007504275692,
+      "grad_norm": 0.1677953600883484,
+      "learning_rate": 1.920127173059166e-05,
+      "loss": 1.2002,
+      "step": 4311
+    },
+    {
+      "epoch": 1.605379809426301,
+      "grad_norm": 0.16227871179580688,
+      "learning_rate": 1.920079572084043e-05,
+      "loss": 1.201,
+      "step": 4312
+    },
+    {
+      "epoch": 1.60575211457691,
+      "grad_norm": 0.16449685394763947,
+      "learning_rate": 1.9200319575194025e-05,
+      "loss": 1.2128,
+      "step": 4313
+    },
+    {
+      "epoch": 1.6061244197275193,
+      "grad_norm": 0.1711970865726471,
+      "learning_rate": 1.919984329365948e-05,
+      "loss": 1.2181,
+      "step": 4314
+    },
+    {
+      "epoch": 1.6064967248781281,
+      "grad_norm": 0.16856355965137482,
+      "learning_rate": 1.919936687624382e-05,
+      "loss": 1.2065,
+      "step": 4315
+    },
+    {
+      "epoch": 1.6068690300287374,
+      "grad_norm": 0.1796281933784485,
+      "learning_rate": 1.9198890322954092e-05,
+      "loss": 1.1957,
+      "step": 4316
+    },
+    {
+      "epoch": 1.6072413351793464,
+      "grad_norm": 0.17579133808612823,
+      "learning_rate": 1.9198413633797334e-05,
+      "loss": 1.2143,
+      "step": 4317
+    },
+    {
+      "epoch": 1.6076136403299555,
+      "grad_norm": 0.16579852998256683,
+      "learning_rate": 1.919793680878058e-05,
+      "loss": 1.2078,
+      "step": 4318
+    },
+    {
+      "epoch": 1.6079859454805645,
+      "grad_norm": 0.16327063739299774,
+      "learning_rate": 1.9197459847910878e-05,
+      "loss": 1.2,
+      "step": 4319
+    },
+    {
+      "epoch": 1.6083582506311735,
+      "grad_norm": 0.16912123560905457,
+      "learning_rate": 1.9196982751195272e-05,
+      "loss": 1.2125,
+      "step": 4320
+    },
+    {
+      "epoch": 1.6087305557817826,
+      "grad_norm": 0.16993340849876404,
+      "learning_rate": 1.9196505518640807e-05,
+      "loss": 1.2124,
+      "step": 4321
+    },
+    {
+      "epoch": 1.6091028609323916,
+      "grad_norm": 0.16973137855529785,
+      "learning_rate": 1.9196028150254535e-05,
+      "loss": 1.2082,
+      "step": 4322
+    },
+    {
+      "epoch": 1.609475166083001,
+      "grad_norm": 0.1582946479320526,
+      "learning_rate": 1.91955506460435e-05,
+      "loss": 1.2028,
+      "step": 4323
+    },
+    {
+      "epoch": 1.6098474712336097,
+      "grad_norm": 0.17046526074409485,
+      "learning_rate": 1.9195073006014762e-05,
+      "loss": 1.2113,
+      "step": 4324
+    },
+    {
+      "epoch": 1.610219776384219,
+      "grad_norm": 0.17357958853244781,
+      "learning_rate": 1.9194595230175373e-05,
+      "loss": 1.2014,
+      "step": 4325
+    },
+    {
+      "epoch": 1.610592081534828,
+      "grad_norm": 0.17264226078987122,
+      "learning_rate": 1.9194117318532387e-05,
+      "loss": 1.1983,
+      "step": 4326
+    },
+    {
+      "epoch": 1.610964386685437,
+      "grad_norm": 0.1646987348794937,
+      "learning_rate": 1.9193639271092866e-05,
+      "loss": 1.2043,
+      "step": 4327
+    },
+    {
+      "epoch": 1.611336691836046,
+      "grad_norm": 0.17559926211833954,
+      "learning_rate": 1.919316108786387e-05,
+      "loss": 1.2055,
+      "step": 4328
+    },
+    {
+      "epoch": 1.6117089969866552,
+      "grad_norm": 0.1712273806333542,
+      "learning_rate": 1.9192682768852464e-05,
+      "loss": 1.2072,
+      "step": 4329
+    },
+    {
+      "epoch": 1.6120813021372644,
+      "grad_norm": 0.16717827320098877,
+      "learning_rate": 1.919220431406571e-05,
+      "loss": 1.2085,
+      "step": 4330
+    },
+    {
+      "epoch": 1.6124536072878732,
+      "grad_norm": 0.1609518676996231,
+      "learning_rate": 1.919172572351067e-05,
+      "loss": 1.2075,
+      "step": 4331
+    },
+    {
+      "epoch": 1.6128259124384825,
+      "grad_norm": 0.1639975607395172,
+      "learning_rate": 1.9191246997194426e-05,
+      "loss": 1.2182,
+      "step": 4332
+    },
+    {
+      "epoch": 1.6131982175890913,
+      "grad_norm": 0.16888481378555298,
+      "learning_rate": 1.9190768135124034e-05,
+      "loss": 1.1964,
+      "step": 4333
+    },
+    {
+      "epoch": 1.6135705227397006,
+      "grad_norm": 0.17243389785289764,
+      "learning_rate": 1.9190289137306577e-05,
+      "loss": 1.2095,
+      "step": 4334
+    },
+    {
+      "epoch": 1.6139428278903096,
+      "grad_norm": 0.16882330179214478,
+      "learning_rate": 1.9189810003749125e-05,
+      "loss": 1.2119,
+      "step": 4335
+    },
+    {
+      "epoch": 1.6143151330409187,
+      "grad_norm": 0.16220669448375702,
+      "learning_rate": 1.9189330734458757e-05,
+      "loss": 1.2105,
+      "step": 4336
+    },
+    {
+      "epoch": 1.6146874381915277,
+      "grad_norm": 0.16079674661159515,
+      "learning_rate": 1.918885132944255e-05,
+      "loss": 1.2148,
+      "step": 4337
+    },
+    {
+      "epoch": 1.6150597433421368,
+      "grad_norm": 0.17351467907428741,
+      "learning_rate": 1.9188371788707585e-05,
+      "loss": 1.2062,
+      "step": 4338
+    },
+    {
+      "epoch": 1.615432048492746,
+      "grad_norm": 0.1602046936750412,
+      "learning_rate": 1.9187892112260944e-05,
+      "loss": 1.203,
+      "step": 4339
+    },
+    {
+      "epoch": 1.6158043536433548,
+      "grad_norm": 0.1660042405128479,
+      "learning_rate": 1.9187412300109714e-05,
+      "loss": 1.2168,
+      "step": 4340
+    },
+    {
+      "epoch": 1.616176658793964,
+      "grad_norm": 0.16898569464683533,
+      "learning_rate": 1.9186932352260984e-05,
+      "loss": 1.2069,
+      "step": 4341
+    },
+    {
+      "epoch": 1.616548963944573,
+      "grad_norm": 0.17615294456481934,
+      "learning_rate": 1.9186452268721838e-05,
+      "loss": 1.2044,
+      "step": 4342
+    },
+    {
+      "epoch": 1.6169212690951822,
+      "grad_norm": 0.17343638837337494,
+      "learning_rate": 1.9185972049499368e-05,
+      "loss": 1.2107,
+      "step": 4343
+    },
+    {
+      "epoch": 1.6172935742457912,
+      "grad_norm": 0.15935227274894714,
+      "learning_rate": 1.9185491694600668e-05,
+      "loss": 1.2153,
+      "step": 4344
+    },
+    {
+      "epoch": 1.6176658793964003,
+      "grad_norm": 0.1717858910560608,
+      "learning_rate": 1.9185011204032832e-05,
+      "loss": 1.211,
+      "step": 4345
+    },
+    {
+      "epoch": 1.6180381845470093,
+      "grad_norm": 0.17080077528953552,
+      "learning_rate": 1.9184530577802953e-05,
+      "loss": 1.1987,
+      "step": 4346
+    },
+    {
+      "epoch": 1.6184104896976184,
+      "grad_norm": 0.1667952537536621,
+      "learning_rate": 1.918404981591814e-05,
+      "loss": 1.2066,
+      "step": 4347
+    },
+    {
+      "epoch": 1.6187827948482276,
+      "grad_norm": 0.17135196924209595,
+      "learning_rate": 1.9183568918385484e-05,
+      "loss": 1.1975,
+      "step": 4348
+    },
+    {
+      "epoch": 1.6191550999988364,
+      "grad_norm": 0.16429933905601501,
+      "learning_rate": 1.918308788521209e-05,
+      "loss": 1.1992,
+      "step": 4349
+    },
+    {
+      "epoch": 1.6195274051494457,
+      "grad_norm": 0.16419640183448792,
+      "learning_rate": 1.918260671640507e-05,
+      "loss": 1.1938,
+      "step": 4350
+    },
+    {
+      "epoch": 1.6198997103000545,
+      "grad_norm": 0.1755063682794571,
+      "learning_rate": 1.9182125411971522e-05,
+      "loss": 1.2272,
+      "step": 4351
+    },
+    {
+      "epoch": 1.6202720154506638,
+      "grad_norm": 0.17690351605415344,
+      "learning_rate": 1.9181643971918557e-05,
+      "loss": 1.1957,
+      "step": 4352
+    },
+    {
+      "epoch": 1.6206443206012728,
+      "grad_norm": 0.16387419402599335,
+      "learning_rate": 1.9181162396253286e-05,
+      "loss": 1.1932,
+      "step": 4353
+    },
+    {
+      "epoch": 1.6210166257518819,
+      "grad_norm": 0.1648489236831665,
+      "learning_rate": 1.918068068498282e-05,
+      "loss": 1.2002,
+      "step": 4354
+    },
+    {
+      "epoch": 1.621388930902491,
+      "grad_norm": 0.171859472990036,
+      "learning_rate": 1.9180198838114284e-05,
+      "loss": 1.1869,
+      "step": 4355
+    },
+    {
+      "epoch": 1.6217612360531,
+      "grad_norm": 0.1747763305902481,
+      "learning_rate": 1.9179716855654783e-05,
+      "loss": 1.2189,
+      "step": 4356
+    },
+    {
+      "epoch": 1.6221335412037092,
+      "grad_norm": 0.1828392744064331,
+      "learning_rate": 1.917923473761144e-05,
+      "loss": 1.1953,
+      "step": 4357
+    },
+    {
+      "epoch": 1.622505846354318,
+      "grad_norm": 0.16498731076717377,
+      "learning_rate": 1.917875248399138e-05,
+      "loss": 1.2088,
+      "step": 4358
+    },
+    {
+      "epoch": 1.6228781515049273,
+      "grad_norm": 0.17804205417633057,
+      "learning_rate": 1.9178270094801713e-05,
+      "loss": 1.2074,
+      "step": 4359
+    },
+    {
+      "epoch": 1.6232504566555361,
+      "grad_norm": 0.1724364459514618,
+      "learning_rate": 1.917778757004958e-05,
+      "loss": 1.2236,
+      "step": 4360
+    },
+    {
+      "epoch": 1.6236227618061454,
+      "grad_norm": 0.17162750661373138,
+      "learning_rate": 1.91773049097421e-05,
+      "loss": 1.2016,
+      "step": 4361
+    },
+    {
+      "epoch": 1.6239950669567544,
+      "grad_norm": 0.17239919304847717,
+      "learning_rate": 1.91768221138864e-05,
+      "loss": 1.1871,
+      "step": 4362
+    },
+    {
+      "epoch": 1.6243673721073635,
+      "grad_norm": 0.1668175607919693,
+      "learning_rate": 1.9176339182489614e-05,
+      "loss": 1.1971,
+      "step": 4363
+    },
+    {
+      "epoch": 1.6247396772579725,
+      "grad_norm": 0.1689068228006363,
+      "learning_rate": 1.9175856115558876e-05,
+      "loss": 1.2086,
+      "step": 4364
+    },
+    {
+      "epoch": 1.6251119824085816,
+      "grad_norm": 0.16748180985450745,
+      "learning_rate": 1.9175372913101317e-05,
+      "loss": 1.2124,
+      "step": 4365
+    },
+    {
+      "epoch": 1.6254842875591908,
+      "grad_norm": 0.16516231000423431,
+      "learning_rate": 1.9174889575124077e-05,
+      "loss": 1.1986,
+      "step": 4366
+    },
+    {
+      "epoch": 1.6258565927097997,
+      "grad_norm": 0.16919684410095215,
+      "learning_rate": 1.9174406101634294e-05,
+      "loss": 1.2042,
+      "step": 4367
+    },
+    {
+      "epoch": 1.626228897860409,
+      "grad_norm": 0.1688244789838791,
+      "learning_rate": 1.917392249263911e-05,
+      "loss": 1.1963,
+      "step": 4368
+    },
+    {
+      "epoch": 1.626601203011018,
+      "grad_norm": 0.17357051372528076,
+      "learning_rate": 1.917343874814566e-05,
+      "loss": 1.2133,
+      "step": 4369
+    },
+    {
+      "epoch": 1.626973508161627,
+      "grad_norm": 0.16557662189006805,
+      "learning_rate": 1.9172954868161098e-05,
+      "loss": 1.1971,
+      "step": 4370
+    },
+    {
+      "epoch": 1.627345813312236,
+      "grad_norm": 0.16236409544944763,
+      "learning_rate": 1.9172470852692572e-05,
+      "loss": 1.1992,
+      "step": 4371
+    },
+    {
+      "epoch": 1.627718118462845,
+      "grad_norm": 0.16975420713424683,
+      "learning_rate": 1.9171986701747227e-05,
+      "loss": 1.1992,
+      "step": 4372
+    },
+    {
+      "epoch": 1.6280904236134541,
+      "grad_norm": 0.1642945110797882,
+      "learning_rate": 1.9171502415332214e-05,
+      "loss": 1.2119,
+      "step": 4373
+    },
+    {
+      "epoch": 1.6284627287640632,
+      "grad_norm": 0.15670865774154663,
+      "learning_rate": 1.9171017993454684e-05,
+      "loss": 1.1994,
+      "step": 4374
+    },
+    {
+      "epoch": 1.6288350339146724,
+      "grad_norm": 0.16495923697948456,
+      "learning_rate": 1.9170533436121793e-05,
+      "loss": 1.2089,
+      "step": 4375
+    },
+    {
+      "epoch": 1.6292073390652813,
+      "grad_norm": 0.16799496114253998,
+      "learning_rate": 1.9170048743340698e-05,
+      "loss": 1.1993,
+      "step": 4376
+    },
+    {
+      "epoch": 1.6295796442158905,
+      "grad_norm": 0.1674148440361023,
+      "learning_rate": 1.9169563915118562e-05,
+      "loss": 1.2032,
+      "step": 4377
+    },
+    {
+      "epoch": 1.6299519493664996,
+      "grad_norm": 0.16675646603107452,
+      "learning_rate": 1.9169078951462537e-05,
+      "loss": 1.2214,
+      "step": 4378
+    },
+    {
+      "epoch": 1.6303242545171086,
+      "grad_norm": 0.16630315780639648,
+      "learning_rate": 1.9168593852379798e-05,
+      "loss": 1.1958,
+      "step": 4379
+    },
+    {
+      "epoch": 1.6306965596677176,
+      "grad_norm": 0.16184255480766296,
+      "learning_rate": 1.91681086178775e-05,
+      "loss": 1.2152,
+      "step": 4380
+    },
+    {
+      "epoch": 1.6310688648183267,
+      "grad_norm": 0.1670718640089035,
+      "learning_rate": 1.9167623247962816e-05,
+      "loss": 1.2088,
+      "step": 4381
+    },
+    {
+      "epoch": 1.6314411699689357,
+      "grad_norm": 0.1772667020559311,
+      "learning_rate": 1.916713774264291e-05,
+      "loss": 1.2204,
+      "step": 4382
+    },
+    {
+      "epoch": 1.6318134751195448,
+      "grad_norm": 0.16231748461723328,
+      "learning_rate": 1.916665210192495e-05,
+      "loss": 1.1968,
+      "step": 4383
+    },
+    {
+      "epoch": 1.632185780270154,
+      "grad_norm": 0.1662776619195938,
+      "learning_rate": 1.916616632581612e-05,
+      "loss": 1.1995,
+      "step": 4384
+    },
+    {
+      "epoch": 1.6325580854207629,
+      "grad_norm": 0.16376206278800964,
+      "learning_rate": 1.9165680414323585e-05,
+      "loss": 1.207,
+      "step": 4385
+    },
+    {
+      "epoch": 1.6329303905713721,
+      "grad_norm": 0.16616666316986084,
+      "learning_rate": 1.916519436745453e-05,
+      "loss": 1.2061,
+      "step": 4386
+    },
+    {
+      "epoch": 1.6333026957219812,
+      "grad_norm": 0.16720803081989288,
+      "learning_rate": 1.9164708185216122e-05,
+      "loss": 1.2103,
+      "step": 4387
+    },
+    {
+      "epoch": 1.6336750008725902,
+      "grad_norm": 0.16905055940151215,
+      "learning_rate": 1.9164221867615556e-05,
+      "loss": 1.2003,
+      "step": 4388
+    },
+    {
+      "epoch": 1.6340473060231993,
+      "grad_norm": 0.17031989991664886,
+      "learning_rate": 1.9163735414660005e-05,
+      "loss": 1.1962,
+      "step": 4389
+    },
+    {
+      "epoch": 1.6344196111738083,
+      "grad_norm": 0.17323459684848785,
+      "learning_rate": 1.9163248826356657e-05,
+      "loss": 1.2219,
+      "step": 4390
+    },
+    {
+      "epoch": 1.6347919163244176,
+      "grad_norm": 0.17302441596984863,
+      "learning_rate": 1.91627621027127e-05,
+      "loss": 1.2199,
+      "step": 4391
+    },
+    {
+      "epoch": 1.6351642214750264,
+      "grad_norm": 0.19569987058639526,
+      "learning_rate": 1.916227524373532e-05,
+      "loss": 1.2103,
+      "step": 4392
+    },
+    {
+      "epoch": 1.6355365266256356,
+      "grad_norm": 0.16403131186962128,
+      "learning_rate": 1.916178824943171e-05,
+      "loss": 1.1942,
+      "step": 4393
+    },
+    {
+      "epoch": 1.6359088317762445,
+      "grad_norm": 0.17407037317752838,
+      "learning_rate": 1.9161301119809065e-05,
+      "loss": 1.2064,
+      "step": 4394
+    },
+    {
+      "epoch": 1.6362811369268537,
+      "grad_norm": 0.19399769604206085,
+      "learning_rate": 1.916081385487458e-05,
+      "loss": 1.2097,
+      "step": 4395
+    },
+    {
+      "epoch": 1.6366534420774628,
+      "grad_norm": 0.1640399843454361,
+      "learning_rate": 1.9160326454635442e-05,
+      "loss": 1.208,
+      "step": 4396
+    },
+    {
+      "epoch": 1.6370257472280718,
+      "grad_norm": 0.1667345017194748,
+      "learning_rate": 1.9159838919098862e-05,
+      "loss": 1.1953,
+      "step": 4397
+    },
+    {
+      "epoch": 1.6373980523786809,
+      "grad_norm": 0.16879738867282867,
+      "learning_rate": 1.9159351248272032e-05,
+      "loss": 1.2016,
+      "step": 4398
+    },
+    {
+      "epoch": 1.63777035752929,
+      "grad_norm": 0.19763416051864624,
+      "learning_rate": 1.9158863442162162e-05,
+      "loss": 1.2099,
+      "step": 4399
+    },
+    {
+      "epoch": 1.6381426626798992,
+      "grad_norm": 0.1749972701072693,
+      "learning_rate": 1.9158375500776454e-05,
+      "loss": 1.2072,
+      "step": 4400
+    },
+    {
+      "epoch": 1.638514967830508,
+      "grad_norm": 0.17299896478652954,
+      "learning_rate": 1.9157887424122112e-05,
+      "loss": 1.2046,
+      "step": 4401
+    },
+    {
+      "epoch": 1.6388872729811172,
+      "grad_norm": 0.16272982954978943,
+      "learning_rate": 1.915739921220635e-05,
+      "loss": 1.2075,
+      "step": 4402
+    },
+    {
+      "epoch": 1.639259578131726,
+      "grad_norm": 0.186982199549675,
+      "learning_rate": 1.9156910865036375e-05,
+      "loss": 1.2091,
+      "step": 4403
+    },
+    {
+      "epoch": 1.6396318832823353,
+      "grad_norm": 0.1684504896402359,
+      "learning_rate": 1.91564223826194e-05,
+      "loss": 1.212,
+      "step": 4404
+    },
+    {
+      "epoch": 1.6400041884329444,
+      "grad_norm": 0.17304614186286926,
+      "learning_rate": 1.9155933764962645e-05,
+      "loss": 1.2291,
+      "step": 4405
+    },
+    {
+      "epoch": 1.6403764935835534,
+      "grad_norm": 0.1625901609659195,
+      "learning_rate": 1.915544501207332e-05,
+      "loss": 1.2118,
+      "step": 4406
+    },
+    {
+      "epoch": 1.6407487987341625,
+      "grad_norm": 0.17311595380306244,
+      "learning_rate": 1.915495612395865e-05,
+      "loss": 1.2139,
+      "step": 4407
+    },
+    {
+      "epoch": 1.6411211038847715,
+      "grad_norm": 0.17164155840873718,
+      "learning_rate": 1.9154467100625848e-05,
+      "loss": 1.2048,
+      "step": 4408
+    },
+    {
+      "epoch": 1.6414934090353808,
+      "grad_norm": 0.169975146651268,
+      "learning_rate": 1.9153977942082143e-05,
+      "loss": 1.2059,
+      "step": 4409
+    },
+    {
+      "epoch": 1.6418657141859896,
+      "grad_norm": 0.17380517721176147,
+      "learning_rate": 1.915348864833476e-05,
+      "loss": 1.2203,
+      "step": 4410
+    },
+    {
+      "epoch": 1.6422380193365989,
+      "grad_norm": 0.16856248676776886,
+      "learning_rate": 1.9152999219390924e-05,
+      "loss": 1.2108,
+      "step": 4411
+    },
+    {
+      "epoch": 1.6426103244872077,
+      "grad_norm": 0.17079463601112366,
+      "learning_rate": 1.915250965525786e-05,
+      "loss": 1.207,
+      "step": 4412
+    },
+    {
+      "epoch": 1.642982629637817,
+      "grad_norm": 0.16879814863204956,
+      "learning_rate": 1.9152019955942808e-05,
+      "loss": 1.2071,
+      "step": 4413
+    },
+    {
+      "epoch": 1.643354934788426,
+      "grad_norm": 0.17056001722812653,
+      "learning_rate": 1.915153012145299e-05,
+      "loss": 1.2001,
+      "step": 4414
+    },
+    {
+      "epoch": 1.643727239939035,
+      "grad_norm": 0.1714567393064499,
+      "learning_rate": 1.915104015179565e-05,
+      "loss": 1.2208,
+      "step": 4415
+    },
+    {
+      "epoch": 1.644099545089644,
+      "grad_norm": 0.16902725398540497,
+      "learning_rate": 1.9150550046978022e-05,
+      "loss": 1.2072,
+      "step": 4416
+    },
+    {
+      "epoch": 1.644471850240253,
+      "grad_norm": 0.1708216667175293,
+      "learning_rate": 1.9150059807007343e-05,
+      "loss": 1.204,
+      "step": 4417
+    },
+    {
+      "epoch": 1.6448441553908624,
+      "grad_norm": 0.17180079221725464,
+      "learning_rate": 1.9149569431890854e-05,
+      "loss": 1.2099,
+      "step": 4418
+    },
+    {
+      "epoch": 1.6452164605414712,
+      "grad_norm": 0.17007555067539215,
+      "learning_rate": 1.91490789216358e-05,
+      "loss": 1.217,
+      "step": 4419
+    },
+    {
+      "epoch": 1.6455887656920805,
+      "grad_norm": 0.16884230077266693,
+      "learning_rate": 1.9148588276249423e-05,
+      "loss": 1.2053,
+      "step": 4420
+    },
+    {
+      "epoch": 1.6459610708426893,
+      "grad_norm": 0.16880488395690918,
+      "learning_rate": 1.9148097495738974e-05,
+      "loss": 1.1978,
+      "step": 4421
+    },
+    {
+      "epoch": 1.6463333759932985,
+      "grad_norm": 0.16282446682453156,
+      "learning_rate": 1.9147606580111696e-05,
+      "loss": 1.2012,
+      "step": 4422
+    },
+    {
+      "epoch": 1.6467056811439076,
+      "grad_norm": 0.17796848714351654,
+      "learning_rate": 1.9147115529374846e-05,
+      "loss": 1.2069,
+      "step": 4423
+    },
+    {
+      "epoch": 1.6470779862945166,
+      "grad_norm": 0.16863879561424255,
+      "learning_rate": 1.914662434353567e-05,
+      "loss": 1.2175,
+      "step": 4424
+    },
+    {
+      "epoch": 1.6474502914451257,
+      "grad_norm": 0.1625177413225174,
+      "learning_rate": 1.914613302260143e-05,
+      "loss": 1.1944,
+      "step": 4425
+    },
+    {
+      "epoch": 1.6478225965957347,
+      "grad_norm": 0.1714029610157013,
+      "learning_rate": 1.9145641566579377e-05,
+      "loss": 1.2129,
+      "step": 4426
+    },
+    {
+      "epoch": 1.648194901746344,
+      "grad_norm": 0.16849827766418457,
+      "learning_rate": 1.914514997547677e-05,
+      "loss": 1.2187,
+      "step": 4427
+    },
+    {
+      "epoch": 1.6485672068969528,
+      "grad_norm": 0.16397565603256226,
+      "learning_rate": 1.9144658249300877e-05,
+      "loss": 1.1988,
+      "step": 4428
+    },
+    {
+      "epoch": 1.648939512047562,
+      "grad_norm": 0.16494201123714447,
+      "learning_rate": 1.9144166388058952e-05,
+      "loss": 1.2045,
+      "step": 4429
+    },
+    {
+      "epoch": 1.649311817198171,
+      "grad_norm": 0.16765427589416504,
+      "learning_rate": 1.9143674391758264e-05,
+      "loss": 1.1967,
+      "step": 4430
+    },
+    {
+      "epoch": 1.6496841223487801,
+      "grad_norm": 0.16366161406040192,
+      "learning_rate": 1.914318226040608e-05,
+      "loss": 1.1777,
+      "step": 4431
+    },
+    {
+      "epoch": 1.6500564274993892,
+      "grad_norm": 0.16279412806034088,
+      "learning_rate": 1.9142689994009666e-05,
+      "loss": 1.2158,
+      "step": 4432
+    },
+    {
+      "epoch": 1.6504287326499982,
+      "grad_norm": 0.16934585571289062,
+      "learning_rate": 1.9142197592576294e-05,
+      "loss": 1.2082,
+      "step": 4433
+    },
+    {
+      "epoch": 1.6508010378006073,
+      "grad_norm": 0.17340950667858124,
+      "learning_rate": 1.914170505611324e-05,
+      "loss": 1.2204,
+      "step": 4434
+    },
+    {
+      "epoch": 1.6511733429512163,
+      "grad_norm": 0.16891925036907196,
+      "learning_rate": 1.9141212384627777e-05,
+      "loss": 1.2084,
+      "step": 4435
+    },
+    {
+      "epoch": 1.6515456481018256,
+      "grad_norm": 0.1604902744293213,
+      "learning_rate": 1.914071957812718e-05,
+      "loss": 1.1887,
+      "step": 4436
+    },
+    {
+      "epoch": 1.6519179532524344,
+      "grad_norm": 0.17385107278823853,
+      "learning_rate": 1.9140226636618726e-05,
+      "loss": 1.2132,
+      "step": 4437
+    },
+    {
+      "epoch": 1.6522902584030437,
+      "grad_norm": 0.17028431594371796,
+      "learning_rate": 1.91397335601097e-05,
+      "loss": 1.1935,
+      "step": 4438
+    },
+    {
+      "epoch": 1.6526625635536527,
+      "grad_norm": 0.17365415394306183,
+      "learning_rate": 1.913924034860738e-05,
+      "loss": 1.2097,
+      "step": 4439
+    },
+    {
+      "epoch": 1.6530348687042618,
+      "grad_norm": 0.1641881912946701,
+      "learning_rate": 1.913874700211906e-05,
+      "loss": 1.2043,
+      "step": 4440
+    },
+    {
+      "epoch": 1.6534071738548708,
+      "grad_norm": 0.17010696232318878,
+      "learning_rate": 1.9138253520652014e-05,
+      "loss": 1.1914,
+      "step": 4441
+    },
+    {
+      "epoch": 1.6537794790054798,
+      "grad_norm": 0.16985437273979187,
+      "learning_rate": 1.913775990421354e-05,
+      "loss": 1.2232,
+      "step": 4442
+    },
+    {
+      "epoch": 1.654151784156089,
+      "grad_norm": 0.16199250519275665,
+      "learning_rate": 1.9137266152810925e-05,
+      "loss": 1.1949,
+      "step": 4443
+    },
+    {
+      "epoch": 1.654524089306698,
+      "grad_norm": 0.16791431605815887,
+      "learning_rate": 1.9136772266451462e-05,
+      "loss": 1.2102,
+      "step": 4444
+    },
+    {
+      "epoch": 1.6548963944573072,
+      "grad_norm": 0.16607221961021423,
+      "learning_rate": 1.9136278245142446e-05,
+      "loss": 1.205,
+      "step": 4445
+    },
+    {
+      "epoch": 1.655268699607916,
+      "grad_norm": 0.16292716562747955,
+      "learning_rate": 1.9135784088891175e-05,
+      "loss": 1.2006,
+      "step": 4446
+    },
+    {
+      "epoch": 1.6556410047585253,
+      "grad_norm": 0.1653973013162613,
+      "learning_rate": 1.9135289797704946e-05,
+      "loss": 1.191,
+      "step": 4447
+    },
+    {
+      "epoch": 1.6560133099091343,
+      "grad_norm": 0.17572778463363647,
+      "learning_rate": 1.913479537159106e-05,
+      "loss": 1.2133,
+      "step": 4448
+    },
+    {
+      "epoch": 1.6563856150597434,
+      "grad_norm": 0.17195762693881989,
+      "learning_rate": 1.913430081055682e-05,
+      "loss": 1.1863,
+      "step": 4449
+    },
+    {
+      "epoch": 1.6567579202103524,
+      "grad_norm": 0.17644944787025452,
+      "learning_rate": 1.9133806114609527e-05,
+      "loss": 1.1969,
+      "step": 4450
+    },
+    {
+      "epoch": 1.6571302253609614,
+      "grad_norm": 0.16783183813095093,
+      "learning_rate": 1.9133311283756493e-05,
+      "loss": 1.2021,
+      "step": 4451
+    },
+    {
+      "epoch": 1.6575025305115707,
+      "grad_norm": 0.17917399108409882,
+      "learning_rate": 1.9132816318005026e-05,
+      "loss": 1.2142,
+      "step": 4452
+    },
+    {
+      "epoch": 1.6578748356621795,
+      "grad_norm": 0.16448262333869934,
+      "learning_rate": 1.9132321217362434e-05,
+      "loss": 1.205,
+      "step": 4453
+    },
+    {
+      "epoch": 1.6582471408127888,
+      "grad_norm": 0.1729598045349121,
+      "learning_rate": 1.913182598183603e-05,
+      "loss": 1.202,
+      "step": 4454
+    },
+    {
+      "epoch": 1.6586194459633976,
+      "grad_norm": 0.1698109656572342,
+      "learning_rate": 1.913133061143313e-05,
+      "loss": 1.1936,
+      "step": 4455
+    },
+    {
+      "epoch": 1.6589917511140069,
+      "grad_norm": 0.1744636744260788,
+      "learning_rate": 1.913083510616105e-05,
+      "loss": 1.1998,
+      "step": 4456
+    },
+    {
+      "epoch": 1.659364056264616,
+      "grad_norm": 0.1762547791004181,
+      "learning_rate": 1.9130339466027108e-05,
+      "loss": 1.2128,
+      "step": 4457
+    },
+    {
+      "epoch": 1.659736361415225,
+      "grad_norm": 0.16493569314479828,
+      "learning_rate": 1.9129843691038625e-05,
+      "loss": 1.205,
+      "step": 4458
+    },
+    {
+      "epoch": 1.660108666565834,
+      "grad_norm": 0.16134680807590485,
+      "learning_rate": 1.9129347781202924e-05,
+      "loss": 1.1999,
+      "step": 4459
+    },
+    {
+      "epoch": 1.660480971716443,
+      "grad_norm": 0.16520054638385773,
+      "learning_rate": 1.912885173652733e-05,
+      "loss": 1.2073,
+      "step": 4460
+    },
+    {
+      "epoch": 1.6608532768670523,
+      "grad_norm": 0.16811229288578033,
+      "learning_rate": 1.9128355557019168e-05,
+      "loss": 1.2041,
+      "step": 4461
+    },
+    {
+      "epoch": 1.6612255820176611,
+      "grad_norm": 0.16674669086933136,
+      "learning_rate": 1.912785924268577e-05,
+      "loss": 1.212,
+      "step": 4462
+    },
+    {
+      "epoch": 1.6615978871682704,
+      "grad_norm": 0.17780810594558716,
+      "learning_rate": 1.912736279353446e-05,
+      "loss": 1.2236,
+      "step": 4463
+    },
+    {
+      "epoch": 1.6619701923188792,
+      "grad_norm": 0.18256306648254395,
+      "learning_rate": 1.9126866209572575e-05,
+      "loss": 1.2028,
+      "step": 4464
+    },
+    {
+      "epoch": 1.6623424974694885,
+      "grad_norm": 0.18052802979946136,
+      "learning_rate": 1.912636949080745e-05,
+      "loss": 1.215,
+      "step": 4465
+    },
+    {
+      "epoch": 1.6627148026200975,
+      "grad_norm": 0.16654305160045624,
+      "learning_rate": 1.912587263724642e-05,
+      "loss": 1.2013,
+      "step": 4466
+    },
+    {
+      "epoch": 1.6630871077707066,
+      "grad_norm": 0.17217010259628296,
+      "learning_rate": 1.9125375648896823e-05,
+      "loss": 1.218,
+      "step": 4467
+    },
+    {
+      "epoch": 1.6634594129213156,
+      "grad_norm": 0.17608584463596344,
+      "learning_rate": 1.9124878525766002e-05,
+      "loss": 1.2081,
+      "step": 4468
+    },
+    {
+      "epoch": 1.6638317180719246,
+      "grad_norm": 0.16333593428134918,
+      "learning_rate": 1.9124381267861295e-05,
+      "loss": 1.2036,
+      "step": 4469
+    },
+    {
+      "epoch": 1.664204023222534,
+      "grad_norm": 0.1907689869403839,
+      "learning_rate": 1.9123883875190052e-05,
+      "loss": 1.202,
+      "step": 4470
+    },
+    {
+      "epoch": 1.6645763283731427,
+      "grad_norm": 0.18991775810718536,
+      "learning_rate": 1.9123386347759614e-05,
+      "loss": 1.2024,
+      "step": 4471
+    },
+    {
+      "epoch": 1.664948633523752,
+      "grad_norm": 0.17199549078941345,
+      "learning_rate": 1.9122888685577337e-05,
+      "loss": 1.2002,
+      "step": 4472
+    },
+    {
+      "epoch": 1.6653209386743608,
+      "grad_norm": 0.20347337424755096,
+      "learning_rate": 1.9122390888650564e-05,
+      "loss": 1.2212,
+      "step": 4473
+    },
+    {
+      "epoch": 1.66569324382497,
+      "grad_norm": 0.16607064008712769,
+      "learning_rate": 1.912189295698665e-05,
+      "loss": 1.2063,
+      "step": 4474
+    },
+    {
+      "epoch": 1.6660655489755791,
+      "grad_norm": 0.16356223821640015,
+      "learning_rate": 1.9121394890592948e-05,
+      "loss": 1.2087,
+      "step": 4475
+    },
+    {
+      "epoch": 1.6664378541261882,
+      "grad_norm": 0.1626313030719757,
+      "learning_rate": 1.9120896689476817e-05,
+      "loss": 1.1938,
+      "step": 4476
+    },
+    {
+      "epoch": 1.6668101592767972,
+      "grad_norm": 0.17197953164577484,
+      "learning_rate": 1.9120398353645615e-05,
+      "loss": 1.2206,
+      "step": 4477
+    },
+    {
+      "epoch": 1.6671824644274063,
+      "grad_norm": 0.16226613521575928,
+      "learning_rate": 1.9119899883106702e-05,
+      "loss": 1.2073,
+      "step": 4478
+    },
+    {
+      "epoch": 1.6675547695780155,
+      "grad_norm": 0.16794995963573456,
+      "learning_rate": 1.911940127786744e-05,
+      "loss": 1.2096,
+      "step": 4479
+    },
+    {
+      "epoch": 1.6679270747286243,
+      "grad_norm": 0.17185860872268677,
+      "learning_rate": 1.911890253793519e-05,
+      "loss": 1.1917,
+      "step": 4480
+    },
+    {
+      "epoch": 1.6682993798792336,
+      "grad_norm": 0.16930022835731506,
+      "learning_rate": 1.9118403663317323e-05,
+      "loss": 1.2207,
+      "step": 4481
+    },
+    {
+      "epoch": 1.6686716850298424,
+      "grad_norm": 0.16357828676700592,
+      "learning_rate": 1.911790465402121e-05,
+      "loss": 1.2031,
+      "step": 4482
+    },
+    {
+      "epoch": 1.6690439901804517,
+      "grad_norm": 0.16401416063308716,
+      "learning_rate": 1.9117405510054216e-05,
+      "loss": 1.1962,
+      "step": 4483
+    },
+    {
+      "epoch": 1.6694162953310607,
+      "grad_norm": 0.1795455366373062,
+      "learning_rate": 1.9116906231423712e-05,
+      "loss": 1.2093,
+      "step": 4484
+    },
+    {
+      "epoch": 1.6697886004816698,
+      "grad_norm": 0.17180180549621582,
+      "learning_rate": 1.9116406818137077e-05,
+      "loss": 1.2082,
+      "step": 4485
+    },
+    {
+      "epoch": 1.6701609056322788,
+      "grad_norm": 0.16501714289188385,
+      "learning_rate": 1.9115907270201684e-05,
+      "loss": 1.1998,
+      "step": 4486
+    },
+    {
+      "epoch": 1.6705332107828879,
+      "grad_norm": 0.16321603953838348,
+      "learning_rate": 1.9115407587624915e-05,
+      "loss": 1.2053,
+      "step": 4487
+    },
+    {
+      "epoch": 1.6709055159334971,
+      "grad_norm": 0.16696272790431976,
+      "learning_rate": 1.911490777041415e-05,
+      "loss": 1.2012,
+      "step": 4488
+    },
+    {
+      "epoch": 1.671277821084106,
+      "grad_norm": 0.16422761976718903,
+      "learning_rate": 1.9114407818576767e-05,
+      "loss": 1.1892,
+      "step": 4489
+    },
+    {
+      "epoch": 1.6716501262347152,
+      "grad_norm": 0.17487448453903198,
+      "learning_rate": 1.911390773212015e-05,
+      "loss": 1.2041,
+      "step": 4490
+    },
+    {
+      "epoch": 1.6720224313853242,
+      "grad_norm": 0.16392962634563446,
+      "learning_rate": 1.911340751105169e-05,
+      "loss": 1.2164,
+      "step": 4491
+    },
+    {
+      "epoch": 1.6723947365359333,
+      "grad_norm": 0.1618296504020691,
+      "learning_rate": 1.9112907155378772e-05,
+      "loss": 1.2109,
+      "step": 4492
+    },
+    {
+      "epoch": 1.6727670416865423,
+      "grad_norm": 0.17908450961112976,
+      "learning_rate": 1.911240666510879e-05,
+      "loss": 1.2248,
+      "step": 4493
+    },
+    {
+      "epoch": 1.6731393468371514,
+      "grad_norm": 0.17229048907756805,
+      "learning_rate": 1.9111906040249134e-05,
+      "loss": 1.2005,
+      "step": 4494
+    },
+    {
+      "epoch": 1.6735116519877604,
+      "grad_norm": 0.16625456511974335,
+      "learning_rate": 1.9111405280807192e-05,
+      "loss": 1.1994,
+      "step": 4495
+    },
+    {
+      "epoch": 1.6738839571383695,
+      "grad_norm": 0.18167874217033386,
+      "learning_rate": 1.911090438679037e-05,
+      "loss": 1.1959,
+      "step": 4496
+    },
+    {
+      "epoch": 1.6742562622889787,
+      "grad_norm": 0.16596947610378265,
+      "learning_rate": 1.911040335820606e-05,
+      "loss": 1.1882,
+      "step": 4497
+    },
+    {
+      "epoch": 1.6746285674395875,
+      "grad_norm": 0.17403769493103027,
+      "learning_rate": 1.9109902195061666e-05,
+      "loss": 1.2009,
+      "step": 4498
+    },
+    {
+      "epoch": 1.6750008725901968,
+      "grad_norm": 0.1789177507162094,
+      "learning_rate": 1.9109400897364584e-05,
+      "loss": 1.2039,
+      "step": 4499
+    },
+    {
+      "epoch": 1.6753731777408059,
+      "grad_norm": 0.1641203761100769,
+      "learning_rate": 1.9108899465122227e-05,
+      "loss": 1.2151,
+      "step": 4500
+    },
+    {
+      "epoch": 1.6753731777408059,
+      "eval_loss": 1.3091740608215332,
+      "eval_runtime": 17.081,
+      "eval_samples_per_second": 101.517,
+      "eval_steps_per_second": 5.093,
+      "step": 4500
+    },
+    {
+      "epoch": 1.675745482891415,
+      "grad_norm": 0.17204515635967255,
+      "learning_rate": 1.9108397898342e-05,
+      "loss": 1.2264,
+      "step": 4501
+    },
+    {
+      "epoch": 1.676117788042024,
+      "grad_norm": 0.16915835440158844,
+      "learning_rate": 1.9107896197031298e-05,
+      "loss": 1.2102,
+      "step": 4502
+    },
+    {
+      "epoch": 1.676490093192633,
+      "grad_norm": 0.22267740964889526,
+      "learning_rate": 1.9107394361197545e-05,
+      "loss": 1.2142,
+      "step": 4503
+    },
+    {
+      "epoch": 1.6768623983432422,
+      "grad_norm": 0.16702081263065338,
+      "learning_rate": 1.9106892390848154e-05,
+      "loss": 1.1892,
+      "step": 4504
+    },
+    {
+      "epoch": 1.677234703493851,
+      "grad_norm": 0.17088332772254944,
+      "learning_rate": 1.9106390285990527e-05,
+      "loss": 1.2173,
+      "step": 4505
+    },
+    {
+      "epoch": 1.6776070086444603,
+      "grad_norm": 0.16930538415908813,
+      "learning_rate": 1.9105888046632088e-05,
+      "loss": 1.1886,
+      "step": 4506
+    },
+    {
+      "epoch": 1.6779793137950691,
+      "grad_norm": 0.18196465075016022,
+      "learning_rate": 1.9105385672780256e-05,
+      "loss": 1.1952,
+      "step": 4507
+    },
+    {
+      "epoch": 1.6783516189456784,
+      "grad_norm": 0.17563000321388245,
+      "learning_rate": 1.910488316444245e-05,
+      "loss": 1.2135,
+      "step": 4508
+    },
+    {
+      "epoch": 1.6787239240962875,
+      "grad_norm": 0.17271769046783447,
+      "learning_rate": 1.910438052162609e-05,
+      "loss": 1.2188,
+      "step": 4509
+    },
+    {
+      "epoch": 1.6790962292468965,
+      "grad_norm": 0.21126636862754822,
+      "learning_rate": 1.91038777443386e-05,
+      "loss": 1.2002,
+      "step": 4510
+    },
+    {
+      "epoch": 1.6794685343975055,
+      "grad_norm": 0.24127759039402008,
+      "learning_rate": 1.9103374832587406e-05,
+      "loss": 1.1998,
+      "step": 4511
+    },
+    {
+      "epoch": 1.6798408395481146,
+      "grad_norm": 0.1943320333957672,
+      "learning_rate": 1.910287178637994e-05,
+      "loss": 1.2146,
+      "step": 4512
+    },
+    {
+      "epoch": 1.6802131446987238,
+      "grad_norm": 0.18076887726783752,
+      "learning_rate": 1.9102368605723626e-05,
+      "loss": 1.1915,
+      "step": 4513
+    },
+    {
+      "epoch": 1.6805854498493327,
+      "grad_norm": 0.2074926644563675,
+      "learning_rate": 1.9101865290625903e-05,
+      "loss": 1.1936,
+      "step": 4514
+    },
+    {
+      "epoch": 1.680957754999942,
+      "grad_norm": 0.17887428402900696,
+      "learning_rate": 1.91013618410942e-05,
+      "loss": 1.2016,
+      "step": 4515
+    },
+    {
+      "epoch": 1.6813300601505508,
+      "grad_norm": 0.1693534553050995,
+      "learning_rate": 1.910085825713595e-05,
+      "loss": 1.2022,
+      "step": 4516
+    },
+    {
+      "epoch": 1.68170236530116,
+      "grad_norm": 0.1669137179851532,
+      "learning_rate": 1.9100354538758598e-05,
+      "loss": 1.2027,
+      "step": 4517
+    },
+    {
+      "epoch": 1.682074670451769,
+      "grad_norm": 0.1691608726978302,
+      "learning_rate": 1.9099850685969578e-05,
+      "loss": 1.2078,
+      "step": 4518
+    },
+    {
+      "epoch": 1.682446975602378,
+      "grad_norm": 0.16599372029304504,
+      "learning_rate": 1.9099346698776338e-05,
+      "loss": 1.185,
+      "step": 4519
+    },
+    {
+      "epoch": 1.6828192807529871,
+      "grad_norm": 0.164974182844162,
+      "learning_rate": 1.9098842577186315e-05,
+      "loss": 1.2028,
+      "step": 4520
+    },
+    {
+      "epoch": 1.6831915859035962,
+      "grad_norm": 0.1712837517261505,
+      "learning_rate": 1.909833832120696e-05,
+      "loss": 1.215,
+      "step": 4521
+    },
+    {
+      "epoch": 1.6835638910542055,
+      "grad_norm": 0.16040773689746857,
+      "learning_rate": 1.9097833930845718e-05,
+      "loss": 1.2097,
+      "step": 4522
+    },
+    {
+      "epoch": 1.6839361962048143,
+      "grad_norm": 0.163836270570755,
+      "learning_rate": 1.9097329406110038e-05,
+      "loss": 1.2142,
+      "step": 4523
+    },
+    {
+      "epoch": 1.6843085013554235,
+      "grad_norm": 0.16300654411315918,
+      "learning_rate": 1.9096824747007378e-05,
+      "loss": 1.203,
+      "step": 4524
+    },
+    {
+      "epoch": 1.6846808065060324,
+      "grad_norm": 0.17333677411079407,
+      "learning_rate": 1.9096319953545186e-05,
+      "loss": 1.2009,
+      "step": 4525
+    },
+    {
+      "epoch": 1.6850531116566416,
+      "grad_norm": 0.1657482534646988,
+      "learning_rate": 1.9095815025730918e-05,
+      "loss": 1.224,
+      "step": 4526
+    },
+    {
+      "epoch": 1.6854254168072507,
+      "grad_norm": 0.17389807105064392,
+      "learning_rate": 1.9095309963572034e-05,
+      "loss": 1.208,
+      "step": 4527
+    },
+    {
+      "epoch": 1.6857977219578597,
+      "grad_norm": 0.16647979617118835,
+      "learning_rate": 1.909480476707599e-05,
+      "loss": 1.1943,
+      "step": 4528
+    },
+    {
+      "epoch": 1.6861700271084687,
+      "grad_norm": 0.16131357848644257,
+      "learning_rate": 1.9094299436250254e-05,
+      "loss": 1.1994,
+      "step": 4529
+    },
+    {
+      "epoch": 1.6865423322590778,
+      "grad_norm": 0.16403862833976746,
+      "learning_rate": 1.9093793971102282e-05,
+      "loss": 1.1896,
+      "step": 4530
+    },
+    {
+      "epoch": 1.686914637409687,
+      "grad_norm": 0.16016656160354614,
+      "learning_rate": 1.9093288371639547e-05,
+      "loss": 1.2013,
+      "step": 4531
+    },
+    {
+      "epoch": 1.6872869425602959,
+      "grad_norm": 0.1598949134349823,
+      "learning_rate": 1.9092782637869513e-05,
+      "loss": 1.1843,
+      "step": 4532
+    },
+    {
+      "epoch": 1.6876592477109051,
+      "grad_norm": 0.16211171448230743,
+      "learning_rate": 1.909227676979965e-05,
+      "loss": 1.1962,
+      "step": 4533
+    },
+    {
+      "epoch": 1.688031552861514,
+      "grad_norm": 0.1622130423784256,
+      "learning_rate": 1.9091770767437428e-05,
+      "loss": 1.1995,
+      "step": 4534
+    },
+    {
+      "epoch": 1.6884038580121232,
+      "grad_norm": 0.1642407476902008,
+      "learning_rate": 1.9091264630790324e-05,
+      "loss": 1.1968,
+      "step": 4535
+    },
+    {
+      "epoch": 1.6887761631627323,
+      "grad_norm": 0.16541729867458344,
+      "learning_rate": 1.9090758359865812e-05,
+      "loss": 1.2072,
+      "step": 4536
+    },
+    {
+      "epoch": 1.6891484683133413,
+      "grad_norm": 0.16923236846923828,
+      "learning_rate": 1.9090251954671372e-05,
+      "loss": 1.2137,
+      "step": 4537
+    },
+    {
+      "epoch": 1.6895207734639504,
+      "grad_norm": 0.16798345744609833,
+      "learning_rate": 1.9089745415214474e-05,
+      "loss": 1.1978,
+      "step": 4538
+    },
+    {
+      "epoch": 1.6898930786145594,
+      "grad_norm": 0.1650944948196411,
+      "learning_rate": 1.9089238741502614e-05,
+      "loss": 1.2105,
+      "step": 4539
+    },
+    {
+      "epoch": 1.6902653837651687,
+      "grad_norm": 0.17048072814941406,
+      "learning_rate": 1.9088731933543262e-05,
+      "loss": 1.208,
+      "step": 4540
+    },
+    {
+      "epoch": 1.6906376889157775,
+      "grad_norm": 0.15930227935314178,
+      "learning_rate": 1.9088224991343916e-05,
+      "loss": 1.217,
+      "step": 4541
+    },
+    {
+      "epoch": 1.6910099940663867,
+      "grad_norm": 0.163492351770401,
+      "learning_rate": 1.9087717914912054e-05,
+      "loss": 1.2102,
+      "step": 4542
+    },
+    {
+      "epoch": 1.6913822992169956,
+      "grad_norm": 0.16682982444763184,
+      "learning_rate": 1.908721070425517e-05,
+      "loss": 1.1899,
+      "step": 4543
+    },
+    {
+      "epoch": 1.6917546043676048,
+      "grad_norm": 0.16626925766468048,
+      "learning_rate": 1.908670335938075e-05,
+      "loss": 1.1931,
+      "step": 4544
+    },
+    {
+      "epoch": 1.6921269095182139,
+      "grad_norm": 0.1644056737422943,
+      "learning_rate": 1.9086195880296294e-05,
+      "loss": 1.2129,
+      "step": 4545
+    },
+    {
+      "epoch": 1.692499214668823,
+      "grad_norm": 0.16008804738521576,
+      "learning_rate": 1.9085688267009298e-05,
+      "loss": 1.2037,
+      "step": 4546
+    },
+    {
+      "epoch": 1.692871519819432,
+      "grad_norm": 0.16731880605220795,
+      "learning_rate": 1.9085180519527252e-05,
+      "loss": 1.1954,
+      "step": 4547
+    },
+    {
+      "epoch": 1.693243824970041,
+      "grad_norm": 0.1611679047346115,
+      "learning_rate": 1.9084672637857663e-05,
+      "loss": 1.2074,
+      "step": 4548
+    },
+    {
+      "epoch": 1.6936161301206503,
+      "grad_norm": 0.16875462234020233,
+      "learning_rate": 1.908416462200803e-05,
+      "loss": 1.2177,
+      "step": 4549
+    },
+    {
+      "epoch": 1.693988435271259,
+      "grad_norm": 0.15939076244831085,
+      "learning_rate": 1.9083656471985855e-05,
+      "loss": 1.2001,
+      "step": 4550
+    },
+    {
+      "epoch": 1.6943607404218683,
+      "grad_norm": 0.16312842071056366,
+      "learning_rate": 1.908314818779864e-05,
+      "loss": 1.1995,
+      "step": 4551
+    },
+    {
+      "epoch": 1.6947330455724774,
+      "grad_norm": 0.16439701616764069,
+      "learning_rate": 1.90826397694539e-05,
+      "loss": 1.2002,
+      "step": 4552
+    },
+    {
+      "epoch": 1.6951053507230864,
+      "grad_norm": 0.16645435988903046,
+      "learning_rate": 1.9082131216959137e-05,
+      "loss": 1.207,
+      "step": 4553
+    },
+    {
+      "epoch": 1.6954776558736955,
+      "grad_norm": 0.1708313226699829,
+      "learning_rate": 1.9081622530321874e-05,
+      "loss": 1.2047,
+      "step": 4554
+    },
+    {
+      "epoch": 1.6958499610243045,
+      "grad_norm": 0.16160647571086884,
+      "learning_rate": 1.908111370954961e-05,
+      "loss": 1.2187,
+      "step": 4555
+    },
+    {
+      "epoch": 1.6962222661749136,
+      "grad_norm": 0.15889696776866913,
+      "learning_rate": 1.9080604754649865e-05,
+      "loss": 1.2136,
+      "step": 4556
+    },
+    {
+      "epoch": 1.6965945713255226,
+      "grad_norm": 0.1612309068441391,
+      "learning_rate": 1.908009566563016e-05,
+      "loss": 1.1969,
+      "step": 4557
+    },
+    {
+      "epoch": 1.6969668764761319,
+      "grad_norm": 0.16933639347553253,
+      "learning_rate": 1.9079586442498016e-05,
+      "loss": 1.2167,
+      "step": 4558
+    },
+    {
+      "epoch": 1.6973391816267407,
+      "grad_norm": 0.1634131371974945,
+      "learning_rate": 1.9079077085260943e-05,
+      "loss": 1.216,
+      "step": 4559
+    },
+    {
+      "epoch": 1.69771148677735,
+      "grad_norm": 0.16775217652320862,
+      "learning_rate": 1.9078567593926472e-05,
+      "loss": 1.2064,
+      "step": 4560
+    },
+    {
+      "epoch": 1.698083791927959,
+      "grad_norm": 0.164699524641037,
+      "learning_rate": 1.9078057968502132e-05,
+      "loss": 1.2052,
+      "step": 4561
+    },
+    {
+      "epoch": 1.698456097078568,
+      "grad_norm": 0.1709076166152954,
+      "learning_rate": 1.9077548208995442e-05,
+      "loss": 1.2005,
+      "step": 4562
+    },
+    {
+      "epoch": 1.698828402229177,
+      "grad_norm": 0.1730164885520935,
+      "learning_rate": 1.907703831541394e-05,
+      "loss": 1.2262,
+      "step": 4563
+    },
+    {
+      "epoch": 1.6992007073797861,
+      "grad_norm": 0.1677834689617157,
+      "learning_rate": 1.9076528287765145e-05,
+      "loss": 1.1978,
+      "step": 4564
+    },
+    {
+      "epoch": 1.6995730125303954,
+      "grad_norm": 0.16488604247570038,
+      "learning_rate": 1.90760181260566e-05,
+      "loss": 1.1987,
+      "step": 4565
+    },
+    {
+      "epoch": 1.6999453176810042,
+      "grad_norm": 0.18518555164337158,
+      "learning_rate": 1.9075507830295837e-05,
+      "loss": 1.1877,
+      "step": 4566
+    },
+    {
+      "epoch": 1.7003176228316135,
+      "grad_norm": 0.16280607879161835,
+      "learning_rate": 1.9074997400490392e-05,
+      "loss": 1.2053,
+      "step": 4567
+    },
+    {
+      "epoch": 1.7006899279822223,
+      "grad_norm": 0.16857080161571503,
+      "learning_rate": 1.90744868366478e-05,
+      "loss": 1.2054,
+      "step": 4568
+    },
+    {
+      "epoch": 1.7010622331328316,
+      "grad_norm": 0.16795577108860016,
+      "learning_rate": 1.9073976138775613e-05,
+      "loss": 1.2102,
+      "step": 4569
+    },
+    {
+      "epoch": 1.7014345382834406,
+      "grad_norm": 0.17121511697769165,
+      "learning_rate": 1.907346530688137e-05,
+      "loss": 1.1977,
+      "step": 4570
+    },
+    {
+      "epoch": 1.7018068434340496,
+      "grad_norm": 0.1754191517829895,
+      "learning_rate": 1.907295434097261e-05,
+      "loss": 1.2035,
+      "step": 4571
+    },
+    {
+      "epoch": 1.7021791485846587,
+      "grad_norm": 0.1667300909757614,
+      "learning_rate": 1.9072443241056884e-05,
+      "loss": 1.2076,
+      "step": 4572
+    },
+    {
+      "epoch": 1.7025514537352677,
+      "grad_norm": 0.19166409969329834,
+      "learning_rate": 1.9071932007141742e-05,
+      "loss": 1.207,
+      "step": 4573
+    },
+    {
+      "epoch": 1.702923758885877,
+      "grad_norm": 0.16179704666137695,
+      "learning_rate": 1.907142063923473e-05,
+      "loss": 1.1928,
+      "step": 4574
+    },
+    {
+      "epoch": 1.7032960640364858,
+      "grad_norm": 0.17530782520771027,
+      "learning_rate": 1.907090913734341e-05,
+      "loss": 1.2057,
+      "step": 4575
+    },
+    {
+      "epoch": 1.703668369187095,
+      "grad_norm": 0.16238301992416382,
+      "learning_rate": 1.9070397501475327e-05,
+      "loss": 1.1965,
+      "step": 4576
+    },
+    {
+      "epoch": 1.704040674337704,
+      "grad_norm": 0.18561150133609772,
+      "learning_rate": 1.9069885731638045e-05,
+      "loss": 1.2201,
+      "step": 4577
+    },
+    {
+      "epoch": 1.7044129794883132,
+      "grad_norm": 0.17264258861541748,
+      "learning_rate": 1.9069373827839117e-05,
+      "loss": 1.1941,
+      "step": 4578
+    },
+    {
+      "epoch": 1.7047852846389222,
+      "grad_norm": 0.16849012672901154,
+      "learning_rate": 1.906886179008611e-05,
+      "loss": 1.2056,
+      "step": 4579
+    },
+    {
+      "epoch": 1.7051575897895312,
+      "grad_norm": 0.16359888017177582,
+      "learning_rate": 1.906834961838658e-05,
+      "loss": 1.2024,
+      "step": 4580
+    },
+    {
+      "epoch": 1.7055298949401403,
+      "grad_norm": 0.18342570960521698,
+      "learning_rate": 1.9067837312748097e-05,
+      "loss": 1.1985,
+      "step": 4581
+    },
+    {
+      "epoch": 1.7059022000907493,
+      "grad_norm": 0.16952671110630035,
+      "learning_rate": 1.9067324873178227e-05,
+      "loss": 1.2096,
+      "step": 4582
+    },
+    {
+      "epoch": 1.7062745052413586,
+      "grad_norm": 0.16725878417491913,
+      "learning_rate": 1.9066812299684537e-05,
+      "loss": 1.2007,
+      "step": 4583
+    },
+    {
+      "epoch": 1.7066468103919674,
+      "grad_norm": 0.1666475087404251,
+      "learning_rate": 1.9066299592274596e-05,
+      "loss": 1.2142,
+      "step": 4584
+    },
+    {
+      "epoch": 1.7070191155425767,
+      "grad_norm": 0.17125262320041656,
+      "learning_rate": 1.9065786750955983e-05,
+      "loss": 1.2146,
+      "step": 4585
+    },
+    {
+      "epoch": 1.7073914206931855,
+      "grad_norm": 0.16806922852993011,
+      "learning_rate": 1.9065273775736264e-05,
+      "loss": 1.1995,
+      "step": 4586
+    },
+    {
+      "epoch": 1.7077637258437948,
+      "grad_norm": 0.16687040030956268,
+      "learning_rate": 1.9064760666623025e-05,
+      "loss": 1.2051,
+      "step": 4587
+    },
+    {
+      "epoch": 1.7081360309944038,
+      "grad_norm": 0.1716591715812683,
+      "learning_rate": 1.9064247423623838e-05,
+      "loss": 1.204,
+      "step": 4588
+    },
+    {
+      "epoch": 1.7085083361450129,
+      "grad_norm": 0.1665405035018921,
+      "learning_rate": 1.9063734046746286e-05,
+      "loss": 1.1918,
+      "step": 4589
+    },
+    {
+      "epoch": 1.708880641295622,
+      "grad_norm": 0.17192888259887695,
+      "learning_rate": 1.906322053599795e-05,
+      "loss": 1.2096,
+      "step": 4590
+    },
+    {
+      "epoch": 1.709252946446231,
+      "grad_norm": 0.16220897436141968,
+      "learning_rate": 1.9062706891386414e-05,
+      "loss": 1.1828,
+      "step": 4591
+    },
+    {
+      "epoch": 1.7096252515968402,
+      "grad_norm": 0.17330990731716156,
+      "learning_rate": 1.9062193112919266e-05,
+      "loss": 1.1909,
+      "step": 4592
+    },
+    {
+      "epoch": 1.709997556747449,
+      "grad_norm": 0.17249375581741333,
+      "learning_rate": 1.9061679200604097e-05,
+      "loss": 1.1908,
+      "step": 4593
+    },
+    {
+      "epoch": 1.7103698618980583,
+      "grad_norm": 0.1671372950077057,
+      "learning_rate": 1.9061165154448496e-05,
+      "loss": 1.2069,
+      "step": 4594
+    },
+    {
+      "epoch": 1.710742167048667,
+      "grad_norm": 0.17133872210979462,
+      "learning_rate": 1.906065097446005e-05,
+      "loss": 1.1977,
+      "step": 4595
+    },
+    {
+      "epoch": 1.7111144721992764,
+      "grad_norm": 0.18286733329296112,
+      "learning_rate": 1.9060136660646362e-05,
+      "loss": 1.2198,
+      "step": 4596
+    },
+    {
+      "epoch": 1.7114867773498854,
+      "grad_norm": 0.17125152051448822,
+      "learning_rate": 1.9059622213015023e-05,
+      "loss": 1.2111,
+      "step": 4597
+    },
+    {
+      "epoch": 1.7118590825004945,
+      "grad_norm": 0.16331323981285095,
+      "learning_rate": 1.905910763157363e-05,
+      "loss": 1.212,
+      "step": 4598
+    },
+    {
+      "epoch": 1.7122313876511035,
+      "grad_norm": 0.17022307217121124,
+      "learning_rate": 1.905859291632979e-05,
+      "loss": 1.2115,
+      "step": 4599
+    },
+    {
+      "epoch": 1.7126036928017125,
+      "grad_norm": 0.17252519726753235,
+      "learning_rate": 1.9058078067291095e-05,
+      "loss": 1.2075,
+      "step": 4600
+    },
+    {
+      "epoch": 1.7129759979523218,
+      "grad_norm": 0.19809845089912415,
+      "learning_rate": 1.9057563084465157e-05,
+      "loss": 1.1885,
+      "step": 4601
+    },
+    {
+      "epoch": 1.7133483031029306,
+      "grad_norm": 0.1879017949104309,
+      "learning_rate": 1.9057047967859584e-05,
+      "loss": 1.2113,
+      "step": 4602
+    },
+    {
+      "epoch": 1.71372060825354,
+      "grad_norm": 0.17755410075187683,
+      "learning_rate": 1.905653271748198e-05,
+      "loss": 1.205,
+      "step": 4603
+    },
+    {
+      "epoch": 1.7140929134041487,
+      "grad_norm": 0.24146291613578796,
+      "learning_rate": 1.905601733333996e-05,
+      "loss": 1.1976,
+      "step": 4604
+    },
+    {
+      "epoch": 1.714465218554758,
+      "grad_norm": 0.2119484394788742,
+      "learning_rate": 1.9055501815441126e-05,
+      "loss": 1.1983,
+      "step": 4605
+    },
+    {
+      "epoch": 1.714837523705367,
+      "grad_norm": 0.18424339592456818,
+      "learning_rate": 1.90549861637931e-05,
+      "loss": 1.2067,
+      "step": 4606
+    },
+    {
+      "epoch": 1.715209828855976,
+      "grad_norm": 0.16043433547019958,
+      "learning_rate": 1.9054470378403495e-05,
+      "loss": 1.2139,
+      "step": 4607
+    },
+    {
+      "epoch": 1.715582134006585,
+      "grad_norm": 0.19521519541740417,
+      "learning_rate": 1.9053954459279934e-05,
+      "loss": 1.2078,
+      "step": 4608
+    },
+    {
+      "epoch": 1.7159544391571941,
+      "grad_norm": 0.1905895173549652,
+      "learning_rate": 1.905343840643003e-05,
+      "loss": 1.1947,
+      "step": 4609
+    },
+    {
+      "epoch": 1.7163267443078034,
+      "grad_norm": 0.17229844629764557,
+      "learning_rate": 1.9052922219861413e-05,
+      "loss": 1.2018,
+      "step": 4610
+    },
+    {
+      "epoch": 1.7166990494584122,
+      "grad_norm": 0.16754981875419617,
+      "learning_rate": 1.90524058995817e-05,
+      "loss": 1.2105,
+      "step": 4611
+    },
+    {
+      "epoch": 1.7170713546090215,
+      "grad_norm": 0.17675986886024475,
+      "learning_rate": 1.9051889445598524e-05,
+      "loss": 1.196,
+      "step": 4612
+    },
+    {
+      "epoch": 1.7174436597596305,
+      "grad_norm": 0.1735907346010208,
+      "learning_rate": 1.9051372857919505e-05,
+      "loss": 1.2019,
+      "step": 4613
+    },
+    {
+      "epoch": 1.7178159649102396,
+      "grad_norm": 0.16587558388710022,
+      "learning_rate": 1.905085613655228e-05,
+      "loss": 1.2019,
+      "step": 4614
+    },
+    {
+      "epoch": 1.7181882700608486,
+      "grad_norm": 0.17289356887340546,
+      "learning_rate": 1.9050339281504474e-05,
+      "loss": 1.1919,
+      "step": 4615
+    },
+    {
+      "epoch": 1.7185605752114577,
+      "grad_norm": 0.1681814342737198,
+      "learning_rate": 1.904982229278373e-05,
+      "loss": 1.1936,
+      "step": 4616
+    },
+    {
+      "epoch": 1.7189328803620667,
+      "grad_norm": 0.15795257687568665,
+      "learning_rate": 1.9049305170397673e-05,
+      "loss": 1.203,
+      "step": 4617
+    },
+    {
+      "epoch": 1.7193051855126757,
+      "grad_norm": 0.16355076432228088,
+      "learning_rate": 1.904878791435395e-05,
+      "loss": 1.2002,
+      "step": 4618
+    },
+    {
+      "epoch": 1.719677490663285,
+      "grad_norm": 0.16879984736442566,
+      "learning_rate": 1.9048270524660197e-05,
+      "loss": 1.2025,
+      "step": 4619
+    },
+    {
+      "epoch": 1.7200497958138938,
+      "grad_norm": 0.1663307398557663,
+      "learning_rate": 1.9047753001324057e-05,
+      "loss": 1.1998,
+      "step": 4620
+    },
+    {
+      "epoch": 1.720422100964503,
+      "grad_norm": 0.16687534749507904,
+      "learning_rate": 1.9047235344353173e-05,
+      "loss": 1.2093,
+      "step": 4621
+    },
+    {
+      "epoch": 1.7207944061151121,
+      "grad_norm": 0.1707238256931305,
+      "learning_rate": 1.9046717553755187e-05,
+      "loss": 1.2089,
+      "step": 4622
+    },
+    {
+      "epoch": 1.7211667112657212,
+      "grad_norm": 0.16789813339710236,
+      "learning_rate": 1.9046199629537754e-05,
+      "loss": 1.2204,
+      "step": 4623
+    },
+    {
+      "epoch": 1.7215390164163302,
+      "grad_norm": 0.16236470639705658,
+      "learning_rate": 1.9045681571708517e-05,
+      "loss": 1.2088,
+      "step": 4624
+    },
+    {
+      "epoch": 1.7219113215669393,
+      "grad_norm": 0.1650935858488083,
+      "learning_rate": 1.9045163380275134e-05,
+      "loss": 1.2009,
+      "step": 4625
+    },
+    {
+      "epoch": 1.7222836267175485,
+      "grad_norm": 0.17150568962097168,
+      "learning_rate": 1.9044645055245254e-05,
+      "loss": 1.2058,
+      "step": 4626
+    },
+    {
+      "epoch": 1.7226559318681574,
+      "grad_norm": 0.15954706072807312,
+      "learning_rate": 1.9044126596626536e-05,
+      "loss": 1.1959,
+      "step": 4627
+    },
+    {
+      "epoch": 1.7230282370187666,
+      "grad_norm": 0.16907548904418945,
+      "learning_rate": 1.9043608004426635e-05,
+      "loss": 1.2186,
+      "step": 4628
+    },
+    {
+      "epoch": 1.7234005421693754,
+      "grad_norm": 0.16456788778305054,
+      "learning_rate": 1.904308927865321e-05,
+      "loss": 1.2052,
+      "step": 4629
+    },
+    {
+      "epoch": 1.7237728473199847,
+      "grad_norm": 0.16648238897323608,
+      "learning_rate": 1.9042570419313927e-05,
+      "loss": 1.1964,
+      "step": 4630
+    },
+    {
+      "epoch": 1.7241451524705937,
+      "grad_norm": 0.16598086059093475,
+      "learning_rate": 1.904205142641644e-05,
+      "loss": 1.1976,
+      "step": 4631
+    },
+    {
+      "epoch": 1.7245174576212028,
+      "grad_norm": 0.16780155897140503,
+      "learning_rate": 1.9041532299968426e-05,
+      "loss": 1.202,
+      "step": 4632
+    },
+    {
+      "epoch": 1.7248897627718118,
+      "grad_norm": 0.1815885603427887,
+      "learning_rate": 1.9041013039977548e-05,
+      "loss": 1.2154,
+      "step": 4633
+    },
+    {
+      "epoch": 1.7252620679224209,
+      "grad_norm": 0.1682657152414322,
+      "learning_rate": 1.9040493646451472e-05,
+      "loss": 1.2037,
+      "step": 4634
+    },
+    {
+      "epoch": 1.7256343730730301,
+      "grad_norm": 0.16929581761360168,
+      "learning_rate": 1.9039974119397872e-05,
+      "loss": 1.2131,
+      "step": 4635
+    },
+    {
+      "epoch": 1.726006678223639,
+      "grad_norm": 0.15777097642421722,
+      "learning_rate": 1.9039454458824426e-05,
+      "loss": 1.1929,
+      "step": 4636
+    },
+    {
+      "epoch": 1.7263789833742482,
+      "grad_norm": 0.16508424282073975,
+      "learning_rate": 1.90389346647388e-05,
+      "loss": 1.2151,
+      "step": 4637
+    },
+    {
+      "epoch": 1.726751288524857,
+      "grad_norm": 0.16090057790279388,
+      "learning_rate": 1.903841473714868e-05,
+      "loss": 1.1994,
+      "step": 4638
+    },
+    {
+      "epoch": 1.7271235936754663,
+      "grad_norm": 0.16870178282260895,
+      "learning_rate": 1.903789467606174e-05,
+      "loss": 1.1991,
+      "step": 4639
+    },
+    {
+      "epoch": 1.7274958988260753,
+      "grad_norm": 0.17125073075294495,
+      "learning_rate": 1.903737448148566e-05,
+      "loss": 1.2113,
+      "step": 4640
+    },
+    {
+      "epoch": 1.7278682039766844,
+      "grad_norm": 0.17486359179019928,
+      "learning_rate": 1.903685415342813e-05,
+      "loss": 1.2241,
+      "step": 4641
+    },
+    {
+      "epoch": 1.7282405091272934,
+      "grad_norm": 0.16174235939979553,
+      "learning_rate": 1.903633369189683e-05,
+      "loss": 1.2012,
+      "step": 4642
+    },
+    {
+      "epoch": 1.7286128142779025,
+      "grad_norm": 0.1670721024274826,
+      "learning_rate": 1.9035813096899448e-05,
+      "loss": 1.2032,
+      "step": 4643
+    },
+    {
+      "epoch": 1.7289851194285117,
+      "grad_norm": 0.1667913943529129,
+      "learning_rate": 1.9035292368443674e-05,
+      "loss": 1.1978,
+      "step": 4644
+    },
+    {
+      "epoch": 1.7293574245791206,
+      "grad_norm": 0.16571226716041565,
+      "learning_rate": 1.90347715065372e-05,
+      "loss": 1.2055,
+      "step": 4645
+    },
+    {
+      "epoch": 1.7297297297297298,
+      "grad_norm": 0.16005724668502808,
+      "learning_rate": 1.9034250511187716e-05,
+      "loss": 1.1902,
+      "step": 4646
+    },
+    {
+      "epoch": 1.7301020348803386,
+      "grad_norm": 0.1715211570262909,
+      "learning_rate": 1.903372938240292e-05,
+      "loss": 1.2029,
+      "step": 4647
+    },
+    {
+      "epoch": 1.730474340030948,
+      "grad_norm": 0.16407069563865662,
+      "learning_rate": 1.9033208120190507e-05,
+      "loss": 1.1987,
+      "step": 4648
+    },
+    {
+      "epoch": 1.730846645181557,
+      "grad_norm": 0.16125743091106415,
+      "learning_rate": 1.9032686724558177e-05,
+      "loss": 1.1941,
+      "step": 4649
+    },
+    {
+      "epoch": 1.731218950332166,
+      "grad_norm": 0.17684829235076904,
+      "learning_rate": 1.9032165195513634e-05,
+      "loss": 1.2181,
+      "step": 4650
+    },
+    {
+      "epoch": 1.731591255482775,
+      "grad_norm": 0.1644441783428192,
+      "learning_rate": 1.9031643533064573e-05,
+      "loss": 1.2018,
+      "step": 4651
+    },
+    {
+      "epoch": 1.731963560633384,
+      "grad_norm": 0.1682443469762802,
+      "learning_rate": 1.9031121737218706e-05,
+      "loss": 1.197,
+      "step": 4652
+    },
+    {
+      "epoch": 1.7323358657839933,
+      "grad_norm": 0.16315114498138428,
+      "learning_rate": 1.9030599807983737e-05,
+      "loss": 1.2067,
+      "step": 4653
+    },
+    {
+      "epoch": 1.7327081709346022,
+      "grad_norm": 0.16800329089164734,
+      "learning_rate": 1.9030077745367377e-05,
+      "loss": 1.2031,
+      "step": 4654
+    },
+    {
+      "epoch": 1.7330804760852114,
+      "grad_norm": 0.16758573055267334,
+      "learning_rate": 1.9029555549377335e-05,
+      "loss": 1.1791,
+      "step": 4655
+    },
+    {
+      "epoch": 1.7334527812358203,
+      "grad_norm": 0.1693509966135025,
+      "learning_rate": 1.9029033220021325e-05,
+      "loss": 1.2177,
+      "step": 4656
+    },
+    {
+      "epoch": 1.7338250863864295,
+      "grad_norm": 0.15949083864688873,
+      "learning_rate": 1.902851075730706e-05,
+      "loss": 1.2019,
+      "step": 4657
+    },
+    {
+      "epoch": 1.7341973915370386,
+      "grad_norm": 0.16945861279964447,
+      "learning_rate": 1.9027988161242258e-05,
+      "loss": 1.1955,
+      "step": 4658
+    },
+    {
+      "epoch": 1.7345696966876476,
+      "grad_norm": 0.16500024497509003,
+      "learning_rate": 1.9027465431834637e-05,
+      "loss": 1.2125,
+      "step": 4659
+    },
+    {
+      "epoch": 1.7349420018382566,
+      "grad_norm": 0.1751050353050232,
+      "learning_rate": 1.9026942569091917e-05,
+      "loss": 1.1917,
+      "step": 4660
+    },
+    {
+      "epoch": 1.7353143069888657,
+      "grad_norm": 0.16122271120548248,
+      "learning_rate": 1.902641957302182e-05,
+      "loss": 1.1966,
+      "step": 4661
+    },
+    {
+      "epoch": 1.735686612139475,
+      "grad_norm": 0.16455887258052826,
+      "learning_rate": 1.9025896443632076e-05,
+      "loss": 1.1993,
+      "step": 4662
+    },
+    {
+      "epoch": 1.7360589172900838,
+      "grad_norm": 0.16358114778995514,
+      "learning_rate": 1.902537318093041e-05,
+      "loss": 1.2037,
+      "step": 4663
+    },
+    {
+      "epoch": 1.736431222440693,
+      "grad_norm": 0.16991184651851654,
+      "learning_rate": 1.9024849784924546e-05,
+      "loss": 1.2098,
+      "step": 4664
+    },
+    {
+      "epoch": 1.7368035275913019,
+      "grad_norm": 0.1610153317451477,
+      "learning_rate": 1.9024326255622215e-05,
+      "loss": 1.2134,
+      "step": 4665
+    },
+    {
+      "epoch": 1.7371758327419111,
+      "grad_norm": 0.16271455585956573,
+      "learning_rate": 1.9023802593031156e-05,
+      "loss": 1.2164,
+      "step": 4666
+    },
+    {
+      "epoch": 1.7375481378925202,
+      "grad_norm": 0.16444487869739532,
+      "learning_rate": 1.9023278797159096e-05,
+      "loss": 1.2124,
+      "step": 4667
+    },
+    {
+      "epoch": 1.7379204430431292,
+      "grad_norm": 0.16799621284008026,
+      "learning_rate": 1.9022754868013775e-05,
+      "loss": 1.2078,
+      "step": 4668
+    },
+    {
+      "epoch": 1.7382927481937382,
+      "grad_norm": 0.16734634339809418,
+      "learning_rate": 1.902223080560293e-05,
+      "loss": 1.2088,
+      "step": 4669
+    },
+    {
+      "epoch": 1.7386650533443473,
+      "grad_norm": 0.16235774755477905,
+      "learning_rate": 1.9021706609934305e-05,
+      "loss": 1.2093,
+      "step": 4670
+    },
+    {
+      "epoch": 1.7390373584949566,
+      "grad_norm": 0.1678633987903595,
+      "learning_rate": 1.9021182281015636e-05,
+      "loss": 1.2058,
+      "step": 4671
+    },
+    {
+      "epoch": 1.7394096636455654,
+      "grad_norm": 0.17294074594974518,
+      "learning_rate": 1.9020657818854673e-05,
+      "loss": 1.1982,
+      "step": 4672
+    },
+    {
+      "epoch": 1.7397819687961746,
+      "grad_norm": 0.17070721089839935,
+      "learning_rate": 1.902013322345916e-05,
+      "loss": 1.1967,
+      "step": 4673
+    },
+    {
+      "epoch": 1.7401542739467837,
+      "grad_norm": 0.16906553506851196,
+      "learning_rate": 1.9019608494836843e-05,
+      "loss": 1.2101,
+      "step": 4674
+    },
+    {
+      "epoch": 1.7405265790973927,
+      "grad_norm": 0.17362748086452484,
+      "learning_rate": 1.9019083632995476e-05,
+      "loss": 1.2191,
+      "step": 4675
+    },
+    {
+      "epoch": 1.7408988842480018,
+      "grad_norm": 0.17202383279800415,
+      "learning_rate": 1.9018558637942813e-05,
+      "loss": 1.2034,
+      "step": 4676
+    },
+    {
+      "epoch": 1.7412711893986108,
+      "grad_norm": 0.16142207384109497,
+      "learning_rate": 1.9018033509686603e-05,
+      "loss": 1.1993,
+      "step": 4677
+    },
+    {
+      "epoch": 1.7416434945492199,
+      "grad_norm": 0.16045573353767395,
+      "learning_rate": 1.9017508248234603e-05,
+      "loss": 1.2045,
+      "step": 4678
+    },
+    {
+      "epoch": 1.742015799699829,
+      "grad_norm": 0.16805481910705566,
+      "learning_rate": 1.901698285359457e-05,
+      "loss": 1.1967,
+      "step": 4679
+    },
+    {
+      "epoch": 1.7423881048504382,
+      "grad_norm": 0.16612930595874786,
+      "learning_rate": 1.9016457325774268e-05,
+      "loss": 1.1955,
+      "step": 4680
+    },
+    {
+      "epoch": 1.742760410001047,
+      "grad_norm": 0.16055123507976532,
+      "learning_rate": 1.901593166478146e-05,
+      "loss": 1.1979,
+      "step": 4681
+    },
+    {
+      "epoch": 1.7431327151516562,
+      "grad_norm": 0.16279493272304535,
+      "learning_rate": 1.90154058706239e-05,
+      "loss": 1.2181,
+      "step": 4682
+    },
+    {
+      "epoch": 1.7435050203022653,
+      "grad_norm": 0.18220022320747375,
+      "learning_rate": 1.9014879943309367e-05,
+      "loss": 1.2013,
+      "step": 4683
+    },
+    {
+      "epoch": 1.7438773254528743,
+      "grad_norm": 0.17012886703014374,
+      "learning_rate": 1.9014353882845626e-05,
+      "loss": 1.1997,
+      "step": 4684
+    },
+    {
+      "epoch": 1.7442496306034834,
+      "grad_norm": 0.16687701642513275,
+      "learning_rate": 1.9013827689240434e-05,
+      "loss": 1.1968,
+      "step": 4685
+    },
+    {
+      "epoch": 1.7446219357540924,
+      "grad_norm": 0.16534671187400818,
+      "learning_rate": 1.9013301362501583e-05,
+      "loss": 1.1906,
+      "step": 4686
+    },
+    {
+      "epoch": 1.7449942409047017,
+      "grad_norm": 0.16701330244541168,
+      "learning_rate": 1.901277490263683e-05,
+      "loss": 1.1929,
+      "step": 4687
+    },
+    {
+      "epoch": 1.7453665460553105,
+      "grad_norm": 0.17027342319488525,
+      "learning_rate": 1.901224830965396e-05,
+      "loss": 1.2095,
+      "step": 4688
+    },
+    {
+      "epoch": 1.7457388512059198,
+      "grad_norm": 0.17393416166305542,
+      "learning_rate": 1.9011721583560747e-05,
+      "loss": 1.2137,
+      "step": 4689
+    },
+    {
+      "epoch": 1.7461111563565286,
+      "grad_norm": 0.15819013118743896,
+      "learning_rate": 1.901119472436497e-05,
+      "loss": 1.2161,
+      "step": 4690
+    },
+    {
+      "epoch": 1.7464834615071378,
+      "grad_norm": 0.17175039649009705,
+      "learning_rate": 1.9010667732074415e-05,
+      "loss": 1.2104,
+      "step": 4691
+    },
+    {
+      "epoch": 1.746855766657747,
+      "grad_norm": 0.16250735521316528,
+      "learning_rate": 1.9010140606696865e-05,
+      "loss": 1.1984,
+      "step": 4692
+    },
+    {
+      "epoch": 1.747228071808356,
+      "grad_norm": 0.16283249855041504,
+      "learning_rate": 1.90096133482401e-05,
+      "loss": 1.1991,
+      "step": 4693
+    },
+    {
+      "epoch": 1.747600376958965,
+      "grad_norm": 0.16619625687599182,
+      "learning_rate": 1.9009085956711916e-05,
+      "loss": 1.2008,
+      "step": 4694
+    },
+    {
+      "epoch": 1.747972682109574,
+      "grad_norm": 0.17469939589500427,
+      "learning_rate": 1.9008558432120094e-05,
+      "loss": 1.2038,
+      "step": 4695
+    },
+    {
+      "epoch": 1.7483449872601833,
+      "grad_norm": 0.1646946370601654,
+      "learning_rate": 1.900803077447243e-05,
+      "loss": 1.1974,
+      "step": 4696
+    },
+    {
+      "epoch": 1.748717292410792,
+      "grad_norm": 0.17054390907287598,
+      "learning_rate": 1.9007502983776712e-05,
+      "loss": 1.2065,
+      "step": 4697
+    },
+    {
+      "epoch": 1.7490895975614014,
+      "grad_norm": 0.1633417159318924,
+      "learning_rate": 1.9006975060040746e-05,
+      "loss": 1.2084,
+      "step": 4698
+    },
+    {
+      "epoch": 1.7494619027120102,
+      "grad_norm": 0.18991155922412872,
+      "learning_rate": 1.9006447003272322e-05,
+      "loss": 1.1942,
+      "step": 4699
+    },
+    {
+      "epoch": 1.7498342078626195,
+      "grad_norm": 0.16292832791805267,
+      "learning_rate": 1.900591881347924e-05,
+      "loss": 1.1923,
+      "step": 4700
+    },
+    {
+      "epoch": 1.7502065130132285,
+      "grad_norm": 0.17191947996616364,
+      "learning_rate": 1.9005390490669305e-05,
+      "loss": 1.2142,
+      "step": 4701
+    },
+    {
+      "epoch": 1.7505788181638375,
+      "grad_norm": 0.17528562247753143,
+      "learning_rate": 1.9004862034850314e-05,
+      "loss": 1.2203,
+      "step": 4702
+    },
+    {
+      "epoch": 1.7509511233144466,
+      "grad_norm": 0.17579975724220276,
+      "learning_rate": 1.900433344603008e-05,
+      "loss": 1.2078,
+      "step": 4703
+    },
+    {
+      "epoch": 1.7513234284650556,
+      "grad_norm": 0.16660520434379578,
+      "learning_rate": 1.9003804724216402e-05,
+      "loss": 1.2123,
+      "step": 4704
+    },
+    {
+      "epoch": 1.7516957336156649,
+      "grad_norm": 0.16542848944664001,
+      "learning_rate": 1.90032758694171e-05,
+      "loss": 1.1933,
+      "step": 4705
+    },
+    {
+      "epoch": 1.7520680387662737,
+      "grad_norm": 0.16875770688056946,
+      "learning_rate": 1.9002746881639972e-05,
+      "loss": 1.1999,
+      "step": 4706
+    },
+    {
+      "epoch": 1.752440343916883,
+      "grad_norm": 0.1760881096124649,
+      "learning_rate": 1.900221776089284e-05,
+      "loss": 1.1932,
+      "step": 4707
+    },
+    {
+      "epoch": 1.7528126490674918,
+      "grad_norm": 0.17353412508964539,
+      "learning_rate": 1.9001688507183517e-05,
+      "loss": 1.2093,
+      "step": 4708
+    },
+    {
+      "epoch": 1.753184954218101,
+      "grad_norm": 0.16783374547958374,
+      "learning_rate": 1.9001159120519817e-05,
+      "loss": 1.2056,
+      "step": 4709
+    },
+    {
+      "epoch": 1.75355725936871,
+      "grad_norm": 0.1809961050748825,
+      "learning_rate": 1.9000629600909562e-05,
+      "loss": 1.21,
+      "step": 4710
+    },
+    {
+      "epoch": 1.7539295645193191,
+      "grad_norm": 0.1838044971227646,
+      "learning_rate": 1.9000099948360577e-05,
+      "loss": 1.1983,
+      "step": 4711
+    },
+    {
+      "epoch": 1.7543018696699282,
+      "grad_norm": 0.18065232038497925,
+      "learning_rate": 1.8999570162880676e-05,
+      "loss": 1.1836,
+      "step": 4712
+    },
+    {
+      "epoch": 1.7546741748205372,
+      "grad_norm": 0.17441041767597198,
+      "learning_rate": 1.899904024447769e-05,
+      "loss": 1.1929,
+      "step": 4713
+    },
+    {
+      "epoch": 1.7550464799711465,
+      "grad_norm": 0.16926120221614838,
+      "learning_rate": 1.8998510193159445e-05,
+      "loss": 1.2176,
+      "step": 4714
+    },
+    {
+      "epoch": 1.7554187851217553,
+      "grad_norm": 0.168484628200531,
+      "learning_rate": 1.8997980008933767e-05,
+      "loss": 1.2068,
+      "step": 4715
+    },
+    {
+      "epoch": 1.7557910902723646,
+      "grad_norm": 0.1681792289018631,
+      "learning_rate": 1.899744969180849e-05,
+      "loss": 1.1982,
+      "step": 4716
+    },
+    {
+      "epoch": 1.7561633954229734,
+      "grad_norm": 0.18074312806129456,
+      "learning_rate": 1.8996919241791446e-05,
+      "loss": 1.2009,
+      "step": 4717
+    },
+    {
+      "epoch": 1.7565357005735827,
+      "grad_norm": 0.17582470178604126,
+      "learning_rate": 1.899638865889047e-05,
+      "loss": 1.2103,
+      "step": 4718
+    },
+    {
+      "epoch": 1.7569080057241917,
+      "grad_norm": 0.17479169368743896,
+      "learning_rate": 1.89958579431134e-05,
+      "loss": 1.1962,
+      "step": 4719
+    },
+    {
+      "epoch": 1.7572803108748007,
+      "grad_norm": 0.21927672624588013,
+      "learning_rate": 1.899532709446807e-05,
+      "loss": 1.1914,
+      "step": 4720
+    },
+    {
+      "epoch": 1.7576526160254098,
+      "grad_norm": 0.18427489697933197,
+      "learning_rate": 1.8994796112962325e-05,
+      "loss": 1.189,
+      "step": 4721
+    },
+    {
+      "epoch": 1.7580249211760188,
+      "grad_norm": 0.17045900225639343,
+      "learning_rate": 1.8994264998604005e-05,
+      "loss": 1.192,
+      "step": 4722
+    },
+    {
+      "epoch": 1.758397226326628,
+      "grad_norm": 0.16136103868484497,
+      "learning_rate": 1.8993733751400953e-05,
+      "loss": 1.2078,
+      "step": 4723
+    },
+    {
+      "epoch": 1.758769531477237,
+      "grad_norm": 0.17973823845386505,
+      "learning_rate": 1.899320237136102e-05,
+      "loss": 1.2029,
+      "step": 4724
+    },
+    {
+      "epoch": 1.7591418366278462,
+      "grad_norm": 0.1666877269744873,
+      "learning_rate": 1.8992670858492053e-05,
+      "loss": 1.1991,
+      "step": 4725
+    },
+    {
+      "epoch": 1.7595141417784552,
+      "grad_norm": 0.16305620968341827,
+      "learning_rate": 1.89921392128019e-05,
+      "loss": 1.1932,
+      "step": 4726
+    },
+    {
+      "epoch": 1.7598864469290643,
+      "grad_norm": 0.1701706200838089,
+      "learning_rate": 1.899160743429842e-05,
+      "loss": 1.2117,
+      "step": 4727
+    },
+    {
+      "epoch": 1.7602587520796733,
+      "grad_norm": 0.17170703411102295,
+      "learning_rate": 1.8991075522989458e-05,
+      "loss": 1.1967,
+      "step": 4728
+    },
+    {
+      "epoch": 1.7606310572302823,
+      "grad_norm": 0.15608349442481995,
+      "learning_rate": 1.8990543478882875e-05,
+      "loss": 1.1778,
+      "step": 4729
+    },
+    {
+      "epoch": 1.7610033623808914,
+      "grad_norm": 0.17074252665042877,
+      "learning_rate": 1.899001130198653e-05,
+      "loss": 1.1975,
+      "step": 4730
+    },
+    {
+      "epoch": 1.7613756675315004,
+      "grad_norm": 0.168598935008049,
+      "learning_rate": 1.898947899230828e-05,
+      "loss": 1.2009,
+      "step": 4731
+    },
+    {
+      "epoch": 1.7617479726821097,
+      "grad_norm": 0.16988085210323334,
+      "learning_rate": 1.8988946549855994e-05,
+      "loss": 1.2064,
+      "step": 4732
+    },
+    {
+      "epoch": 1.7621202778327185,
+      "grad_norm": 0.16802845895290375,
+      "learning_rate": 1.8988413974637527e-05,
+      "loss": 1.1968,
+      "step": 4733
+    },
+    {
+      "epoch": 1.7624925829833278,
+      "grad_norm": 0.16767756640911102,
+      "learning_rate": 1.8987881266660754e-05,
+      "loss": 1.2002,
+      "step": 4734
+    },
+    {
+      "epoch": 1.7628648881339368,
+      "grad_norm": 0.16321119666099548,
+      "learning_rate": 1.8987348425933535e-05,
+      "loss": 1.2016,
+      "step": 4735
+    },
+    {
+      "epoch": 1.7632371932845459,
+      "grad_norm": 0.15938079357147217,
+      "learning_rate": 1.8986815452463747e-05,
+      "loss": 1.2024,
+      "step": 4736
+    },
+    {
+      "epoch": 1.763609498435155,
+      "grad_norm": 0.164838969707489,
+      "learning_rate": 1.8986282346259255e-05,
+      "loss": 1.1904,
+      "step": 4737
+    },
+    {
+      "epoch": 1.763981803585764,
+      "grad_norm": 0.16808170080184937,
+      "learning_rate": 1.898574910732794e-05,
+      "loss": 1.2027,
+      "step": 4738
+    },
+    {
+      "epoch": 1.764354108736373,
+      "grad_norm": 0.16249272227287292,
+      "learning_rate": 1.8985215735677673e-05,
+      "loss": 1.2025,
+      "step": 4739
+    },
+    {
+      "epoch": 1.764726413886982,
+      "grad_norm": 0.16439345479011536,
+      "learning_rate": 1.8984682231316335e-05,
+      "loss": 1.2028,
+      "step": 4740
+    },
+    {
+      "epoch": 1.7650987190375913,
+      "grad_norm": 0.16751708090305328,
+      "learning_rate": 1.89841485942518e-05,
+      "loss": 1.1973,
+      "step": 4741
+    },
+    {
+      "epoch": 1.7654710241882001,
+      "grad_norm": 0.15885969996452332,
+      "learning_rate": 1.8983614824491958e-05,
+      "loss": 1.1985,
+      "step": 4742
+    },
+    {
+      "epoch": 1.7658433293388094,
+      "grad_norm": 0.15906503796577454,
+      "learning_rate": 1.8983080922044687e-05,
+      "loss": 1.2051,
+      "step": 4743
+    },
+    {
+      "epoch": 1.7662156344894184,
+      "grad_norm": 0.16614754498004913,
+      "learning_rate": 1.8982546886917878e-05,
+      "loss": 1.2041,
+      "step": 4744
+    },
+    {
+      "epoch": 1.7665879396400275,
+      "grad_norm": 0.16927511990070343,
+      "learning_rate": 1.8982012719119414e-05,
+      "loss": 1.2092,
+      "step": 4745
+    },
+    {
+      "epoch": 1.7669602447906365,
+      "grad_norm": 0.16667069494724274,
+      "learning_rate": 1.8981478418657185e-05,
+      "loss": 1.2005,
+      "step": 4746
+    },
+    {
+      "epoch": 1.7673325499412456,
+      "grad_norm": 0.15757103264331818,
+      "learning_rate": 1.898094398553908e-05,
+      "loss": 1.2002,
+      "step": 4747
+    },
+    {
+      "epoch": 1.7677048550918548,
+      "grad_norm": 0.1850326508283615,
+      "learning_rate": 1.8980409419772998e-05,
+      "loss": 1.2049,
+      "step": 4748
+    },
+    {
+      "epoch": 1.7680771602424636,
+      "grad_norm": 0.1695263683795929,
+      "learning_rate": 1.8979874721366834e-05,
+      "loss": 1.1879,
+      "step": 4749
+    },
+    {
+      "epoch": 1.768449465393073,
+      "grad_norm": 0.17270156741142273,
+      "learning_rate": 1.8979339890328484e-05,
+      "loss": 1.2006,
+      "step": 4750
+    },
+    {
+      "epoch": 1.7688217705436817,
+      "grad_norm": 0.18274691700935364,
+      "learning_rate": 1.8978804926665848e-05,
+      "loss": 1.2109,
+      "step": 4751
+    },
+    {
+      "epoch": 1.769194075694291,
+      "grad_norm": 0.16077204048633575,
+      "learning_rate": 1.8978269830386825e-05,
+      "loss": 1.1919,
+      "step": 4752
+    },
+    {
+      "epoch": 1.7695663808449,
+      "grad_norm": 0.1679827719926834,
+      "learning_rate": 1.8977734601499322e-05,
+      "loss": 1.2057,
+      "step": 4753
+    },
+    {
+      "epoch": 1.769938685995509,
+      "grad_norm": 0.1700560301542282,
+      "learning_rate": 1.8977199240011237e-05,
+      "loss": 1.2149,
+      "step": 4754
+    },
+    {
+      "epoch": 1.7703109911461181,
+      "grad_norm": 0.17137514054775238,
+      "learning_rate": 1.8976663745930488e-05,
+      "loss": 1.2163,
+      "step": 4755
+    },
+    {
+      "epoch": 1.7706832962967272,
+      "grad_norm": 0.16699162125587463,
+      "learning_rate": 1.897612811926498e-05,
+      "loss": 1.1973,
+      "step": 4756
+    },
+    {
+      "epoch": 1.7710556014473364,
+      "grad_norm": 0.1748683899641037,
+      "learning_rate": 1.8975592360022616e-05,
+      "loss": 1.2049,
+      "step": 4757
+    },
+    {
+      "epoch": 1.7714279065979452,
+      "grad_norm": 0.17756815254688263,
+      "learning_rate": 1.8975056468211323e-05,
+      "loss": 1.1937,
+      "step": 4758
+    },
+    {
+      "epoch": 1.7718002117485545,
+      "grad_norm": 0.16323328018188477,
+      "learning_rate": 1.8974520443839007e-05,
+      "loss": 1.2013,
+      "step": 4759
+    },
+    {
+      "epoch": 1.7721725168991633,
+      "grad_norm": 0.19412948191165924,
+      "learning_rate": 1.8973984286913584e-05,
+      "loss": 1.2092,
+      "step": 4760
+    },
+    {
+      "epoch": 1.7725448220497726,
+      "grad_norm": 0.1670580357313156,
+      "learning_rate": 1.897344799744298e-05,
+      "loss": 1.2202,
+      "step": 4761
+    },
+    {
+      "epoch": 1.7729171272003816,
+      "grad_norm": 0.17248012125492096,
+      "learning_rate": 1.8972911575435112e-05,
+      "loss": 1.1881,
+      "step": 4762
+    },
+    {
+      "epoch": 1.7732894323509907,
+      "grad_norm": 0.17886854708194733,
+      "learning_rate": 1.8972375020897905e-05,
+      "loss": 1.2028,
+      "step": 4763
+    },
+    {
+      "epoch": 1.7736617375015997,
+      "grad_norm": 0.1719491332769394,
+      "learning_rate": 1.897183833383928e-05,
+      "loss": 1.2057,
+      "step": 4764
+    },
+    {
+      "epoch": 1.7740340426522088,
+      "grad_norm": 0.18928195536136627,
+      "learning_rate": 1.8971301514267162e-05,
+      "loss": 1.1843,
+      "step": 4765
+    },
+    {
+      "epoch": 1.774406347802818,
+      "grad_norm": 0.16735054552555084,
+      "learning_rate": 1.897076456218949e-05,
+      "loss": 1.1963,
+      "step": 4766
+    },
+    {
+      "epoch": 1.7747786529534268,
+      "grad_norm": 0.16620579361915588,
+      "learning_rate": 1.897022747761418e-05,
+      "loss": 1.2001,
+      "step": 4767
+    },
+    {
+      "epoch": 1.7751509581040361,
+      "grad_norm": 0.16270288825035095,
+      "learning_rate": 1.8969690260549183e-05,
+      "loss": 1.2023,
+      "step": 4768
+    },
+    {
+      "epoch": 1.775523263254645,
+      "grad_norm": 0.17014147341251373,
+      "learning_rate": 1.8969152911002417e-05,
+      "loss": 1.1932,
+      "step": 4769
+    },
+    {
+      "epoch": 1.7758955684052542,
+      "grad_norm": 0.1710130274295807,
+      "learning_rate": 1.896861542898183e-05,
+      "loss": 1.2129,
+      "step": 4770
+    },
+    {
+      "epoch": 1.7762678735558632,
+      "grad_norm": 0.16842330992221832,
+      "learning_rate": 1.8968077814495355e-05,
+      "loss": 1.195,
+      "step": 4771
+    },
+    {
+      "epoch": 1.7766401787064723,
+      "grad_norm": 0.16350851953029633,
+      "learning_rate": 1.896754006755093e-05,
+      "loss": 1.1914,
+      "step": 4772
+    },
+    {
+      "epoch": 1.7770124838570813,
+      "grad_norm": 0.18311436474323273,
+      "learning_rate": 1.8967002188156503e-05,
+      "loss": 1.1927,
+      "step": 4773
+    },
+    {
+      "epoch": 1.7773847890076904,
+      "grad_norm": 0.16735710203647614,
+      "learning_rate": 1.8966464176320015e-05,
+      "loss": 1.2019,
+      "step": 4774
+    },
+    {
+      "epoch": 1.7777570941582996,
+      "grad_norm": 0.16617412865161896,
+      "learning_rate": 1.8965926032049418e-05,
+      "loss": 1.2009,
+      "step": 4775
+    },
+    {
+      "epoch": 1.7781293993089085,
+      "grad_norm": 0.1925051510334015,
+      "learning_rate": 1.896538775535265e-05,
+      "loss": 1.1979,
+      "step": 4776
+    },
+    {
+      "epoch": 1.7785017044595177,
+      "grad_norm": 0.17179760336875916,
+      "learning_rate": 1.8964849346237676e-05,
+      "loss": 1.1963,
+      "step": 4777
+    },
+    {
+      "epoch": 1.7788740096101265,
+      "grad_norm": 0.172499880194664,
+      "learning_rate": 1.8964310804712435e-05,
+      "loss": 1.1916,
+      "step": 4778
+    },
+    {
+      "epoch": 1.7792463147607358,
+      "grad_norm": 0.1798262745141983,
+      "learning_rate": 1.8963772130784882e-05,
+      "loss": 1.2,
+      "step": 4779
+    },
+    {
+      "epoch": 1.7796186199113448,
+      "grad_norm": 0.16507485508918762,
+      "learning_rate": 1.8963233324462982e-05,
+      "loss": 1.1805,
+      "step": 4780
+    },
+    {
+      "epoch": 1.7799909250619539,
+      "grad_norm": 0.1641186773777008,
+      "learning_rate": 1.896269438575469e-05,
+      "loss": 1.1815,
+      "step": 4781
+    },
+    {
+      "epoch": 1.780363230212563,
+      "grad_norm": 0.16876539587974548,
+      "learning_rate": 1.896215531466796e-05,
+      "loss": 1.2031,
+      "step": 4782
+    },
+    {
+      "epoch": 1.780735535363172,
+      "grad_norm": 0.17245730757713318,
+      "learning_rate": 1.896161611121076e-05,
+      "loss": 1.1912,
+      "step": 4783
+    },
+    {
+      "epoch": 1.7811078405137812,
+      "grad_norm": 0.1740397810935974,
+      "learning_rate": 1.896107677539105e-05,
+      "loss": 1.2029,
+      "step": 4784
+    },
+    {
+      "epoch": 1.78148014566439,
+      "grad_norm": 0.1714610755443573,
+      "learning_rate": 1.8960537307216804e-05,
+      "loss": 1.208,
+      "step": 4785
+    },
+    {
+      "epoch": 1.7818524508149993,
+      "grad_norm": 0.17852343618869781,
+      "learning_rate": 1.895999770669598e-05,
+      "loss": 1.2083,
+      "step": 4786
+    },
+    {
+      "epoch": 1.7822247559656084,
+      "grad_norm": 0.16778838634490967,
+      "learning_rate": 1.8959457973836554e-05,
+      "loss": 1.2066,
+      "step": 4787
+    },
+    {
+      "epoch": 1.7825970611162174,
+      "grad_norm": 0.1763961762189865,
+      "learning_rate": 1.8958918108646495e-05,
+      "loss": 1.1975,
+      "step": 4788
+    },
+    {
+      "epoch": 1.7829693662668264,
+      "grad_norm": 0.16524644196033478,
+      "learning_rate": 1.8958378111133777e-05,
+      "loss": 1.2086,
+      "step": 4789
+    },
+    {
+      "epoch": 1.7833416714174355,
+      "grad_norm": 0.17384812235832214,
+      "learning_rate": 1.895783798130638e-05,
+      "loss": 1.1881,
+      "step": 4790
+    },
+    {
+      "epoch": 1.7837139765680445,
+      "grad_norm": 0.16518385708332062,
+      "learning_rate": 1.8957297719172278e-05,
+      "loss": 1.1994,
+      "step": 4791
+    },
+    {
+      "epoch": 1.7840862817186536,
+      "grad_norm": 0.16743482649326324,
+      "learning_rate": 1.8956757324739445e-05,
+      "loss": 1.199,
+      "step": 4792
+    },
+    {
+      "epoch": 1.7844585868692628,
+      "grad_norm": 0.1630256474018097,
+      "learning_rate": 1.8956216798015873e-05,
+      "loss": 1.2051,
+      "step": 4793
+    },
+    {
+      "epoch": 1.7848308920198717,
+      "grad_norm": 0.16784405708312988,
+      "learning_rate": 1.895567613900954e-05,
+      "loss": 1.2047,
+      "step": 4794
+    },
+    {
+      "epoch": 1.785203197170481,
+      "grad_norm": 0.16349439322948456,
+      "learning_rate": 1.8955135347728434e-05,
+      "loss": 1.1983,
+      "step": 4795
+    },
+    {
+      "epoch": 1.78557550232109,
+      "grad_norm": 0.16735652089118958,
+      "learning_rate": 1.895459442418054e-05,
+      "loss": 1.1949,
+      "step": 4796
+    },
+    {
+      "epoch": 1.785947807471699,
+      "grad_norm": 0.1649763435125351,
+      "learning_rate": 1.8954053368373846e-05,
+      "loss": 1.2026,
+      "step": 4797
+    },
+    {
+      "epoch": 1.786320112622308,
+      "grad_norm": 0.1692332774400711,
+      "learning_rate": 1.895351218031635e-05,
+      "loss": 1.2029,
+      "step": 4798
+    },
+    {
+      "epoch": 1.786692417772917,
+      "grad_norm": 0.1696692705154419,
+      "learning_rate": 1.895297086001604e-05,
+      "loss": 1.1935,
+      "step": 4799
+    },
+    {
+      "epoch": 1.7870647229235264,
+      "grad_norm": 0.1772766262292862,
+      "learning_rate": 1.8952429407480908e-05,
+      "loss": 1.1929,
+      "step": 4800
+    },
+    {
+      "epoch": 1.7874370280741352,
+      "grad_norm": 0.17916709184646606,
+      "learning_rate": 1.895188782271896e-05,
+      "loss": 1.2017,
+      "step": 4801
+    },
+    {
+      "epoch": 1.7878093332247444,
+      "grad_norm": 0.16637340188026428,
+      "learning_rate": 1.8951346105738188e-05,
+      "loss": 1.1952,
+      "step": 4802
+    },
+    {
+      "epoch": 1.7881816383753533,
+      "grad_norm": 0.20684456825256348,
+      "learning_rate": 1.8950804256546597e-05,
+      "loss": 1.2116,
+      "step": 4803
+    },
+    {
+      "epoch": 1.7885539435259625,
+      "grad_norm": 0.17692846059799194,
+      "learning_rate": 1.8950262275152188e-05,
+      "loss": 1.1984,
+      "step": 4804
+    },
+    {
+      "epoch": 1.7889262486765716,
+      "grad_norm": 0.17448139190673828,
+      "learning_rate": 1.8949720161562967e-05,
+      "loss": 1.1976,
+      "step": 4805
+    },
+    {
+      "epoch": 1.7892985538271806,
+      "grad_norm": 0.17009678483009338,
+      "learning_rate": 1.8949177915786942e-05,
+      "loss": 1.1925,
+      "step": 4806
+    },
+    {
+      "epoch": 1.7896708589777897,
+      "grad_norm": 0.17584814131259918,
+      "learning_rate": 1.894863553783212e-05,
+      "loss": 1.1907,
+      "step": 4807
+    },
+    {
+      "epoch": 1.7900431641283987,
+      "grad_norm": 0.17380961775779724,
+      "learning_rate": 1.8948093027706512e-05,
+      "loss": 1.2074,
+      "step": 4808
+    },
+    {
+      "epoch": 1.790415469279008,
+      "grad_norm": 0.1762208640575409,
+      "learning_rate": 1.8947550385418136e-05,
+      "loss": 1.214,
+      "step": 4809
+    },
+    {
+      "epoch": 1.7907877744296168,
+      "grad_norm": 0.16765017807483673,
+      "learning_rate": 1.8947007610974998e-05,
+      "loss": 1.2002,
+      "step": 4810
+    },
+    {
+      "epoch": 1.791160079580226,
+      "grad_norm": 0.197531059384346,
+      "learning_rate": 1.894646470438512e-05,
+      "loss": 1.1988,
+      "step": 4811
+    },
+    {
+      "epoch": 1.7915323847308349,
+      "grad_norm": 0.17233921587467194,
+      "learning_rate": 1.894592166565652e-05,
+      "loss": 1.1891,
+      "step": 4812
+    },
+    {
+      "epoch": 1.7919046898814441,
+      "grad_norm": 0.1792616844177246,
+      "learning_rate": 1.8945378494797216e-05,
+      "loss": 1.1974,
+      "step": 4813
+    },
+    {
+      "epoch": 1.7922769950320532,
+      "grad_norm": 0.16458769142627716,
+      "learning_rate": 1.8944835191815233e-05,
+      "loss": 1.2007,
+      "step": 4814
+    },
+    {
+      "epoch": 1.7926493001826622,
+      "grad_norm": 0.20892377197742462,
+      "learning_rate": 1.89442917567186e-05,
+      "loss": 1.1936,
+      "step": 4815
+    },
+    {
+      "epoch": 1.7930216053332713,
+      "grad_norm": 0.17339985072612762,
+      "learning_rate": 1.894374818951534e-05,
+      "loss": 1.1883,
+      "step": 4816
+    },
+    {
+      "epoch": 1.7933939104838803,
+      "grad_norm": 0.17067044973373413,
+      "learning_rate": 1.8943204490213474e-05,
+      "loss": 1.2031,
+      "step": 4817
+    },
+    {
+      "epoch": 1.7937662156344896,
+      "grad_norm": 0.15906625986099243,
+      "learning_rate": 1.8942660658821045e-05,
+      "loss": 1.2074,
+      "step": 4818
+    },
+    {
+      "epoch": 1.7941385207850984,
+      "grad_norm": 0.1729053556919098,
+      "learning_rate": 1.8942116695346073e-05,
+      "loss": 1.1915,
+      "step": 4819
+    },
+    {
+      "epoch": 1.7945108259357077,
+      "grad_norm": 0.16248184442520142,
+      "learning_rate": 1.8941572599796602e-05,
+      "loss": 1.1905,
+      "step": 4820
+    },
+    {
+      "epoch": 1.7948831310863165,
+      "grad_norm": 0.17368124425411224,
+      "learning_rate": 1.8941028372180667e-05,
+      "loss": 1.1896,
+      "step": 4821
+    },
+    {
+      "epoch": 1.7952554362369257,
+      "grad_norm": 0.1708829700946808,
+      "learning_rate": 1.8940484012506298e-05,
+      "loss": 1.2013,
+      "step": 4822
+    },
+    {
+      "epoch": 1.7956277413875348,
+      "grad_norm": 0.15987025201320648,
+      "learning_rate": 1.8939939520781546e-05,
+      "loss": 1.22,
+      "step": 4823
+    },
+    {
+      "epoch": 1.7960000465381438,
+      "grad_norm": 0.16917409002780914,
+      "learning_rate": 1.8939394897014448e-05,
+      "loss": 1.1861,
+      "step": 4824
+    },
+    {
+      "epoch": 1.7963723516887529,
+      "grad_norm": 0.1673855185508728,
+      "learning_rate": 1.893885014121305e-05,
+      "loss": 1.202,
+      "step": 4825
+    },
+    {
+      "epoch": 1.796744656839362,
+      "grad_norm": 0.1666531264781952,
+      "learning_rate": 1.8938305253385395e-05,
+      "loss": 1.2077,
+      "step": 4826
+    },
+    {
+      "epoch": 1.7971169619899712,
+      "grad_norm": 0.16617925465106964,
+      "learning_rate": 1.893776023353953e-05,
+      "loss": 1.2171,
+      "step": 4827
+    },
+    {
+      "epoch": 1.79748926714058,
+      "grad_norm": 0.16386044025421143,
+      "learning_rate": 1.893721508168351e-05,
+      "loss": 1.1957,
+      "step": 4828
+    },
+    {
+      "epoch": 1.7978615722911893,
+      "grad_norm": 0.19208215177059174,
+      "learning_rate": 1.8936669797825384e-05,
+      "loss": 1.2114,
+      "step": 4829
+    },
+    {
+      "epoch": 1.798233877441798,
+      "grad_norm": 0.16761602461338043,
+      "learning_rate": 1.8936124381973203e-05,
+      "loss": 1.194,
+      "step": 4830
+    },
+    {
+      "epoch": 1.7986061825924073,
+      "grad_norm": 0.1722363531589508,
+      "learning_rate": 1.893557883413503e-05,
+      "loss": 1.1789,
+      "step": 4831
+    },
+    {
+      "epoch": 1.7989784877430164,
+      "grad_norm": 0.1653546690940857,
+      "learning_rate": 1.8935033154318914e-05,
+      "loss": 1.2039,
+      "step": 4832
+    },
+    {
+      "epoch": 1.7993507928936254,
+      "grad_norm": 0.16508762538433075,
+      "learning_rate": 1.8934487342532925e-05,
+      "loss": 1.1938,
+      "step": 4833
+    },
+    {
+      "epoch": 1.7997230980442345,
+      "grad_norm": 0.17712639272212982,
+      "learning_rate": 1.893394139878511e-05,
+      "loss": 1.1873,
+      "step": 4834
+    },
+    {
+      "epoch": 1.8000954031948435,
+      "grad_norm": 0.17879736423492432,
+      "learning_rate": 1.8933395323083547e-05,
+      "loss": 1.2158,
+      "step": 4835
+    },
+    {
+      "epoch": 1.8004677083454528,
+      "grad_norm": 0.1629367172718048,
+      "learning_rate": 1.8932849115436296e-05,
+      "loss": 1.1977,
+      "step": 4836
+    },
+    {
+      "epoch": 1.8008400134960616,
+      "grad_norm": 0.16029439866542816,
+      "learning_rate": 1.8932302775851423e-05,
+      "loss": 1.2091,
+      "step": 4837
+    },
+    {
+      "epoch": 1.8012123186466709,
+      "grad_norm": 0.17447559535503387,
+      "learning_rate": 1.8931756304337e-05,
+      "loss": 1.2146,
+      "step": 4838
+    },
+    {
+      "epoch": 1.8015846237972797,
+      "grad_norm": 0.16988350450992584,
+      "learning_rate": 1.8931209700901096e-05,
+      "loss": 1.198,
+      "step": 4839
+    },
+    {
+      "epoch": 1.801956928947889,
+      "grad_norm": 0.1667841076850891,
+      "learning_rate": 1.8930662965551784e-05,
+      "loss": 1.1971,
+      "step": 4840
+    },
+    {
+      "epoch": 1.802329234098498,
+      "grad_norm": 0.17226542532444,
+      "learning_rate": 1.893011609829714e-05,
+      "loss": 1.1918,
+      "step": 4841
+    },
+    {
+      "epoch": 1.802701539249107,
+      "grad_norm": 0.16706976294517517,
+      "learning_rate": 1.892956909914524e-05,
+      "loss": 1.2056,
+      "step": 4842
+    },
+    {
+      "epoch": 1.803073844399716,
+      "grad_norm": 0.1757022887468338,
+      "learning_rate": 1.892902196810417e-05,
+      "loss": 1.211,
+      "step": 4843
+    },
+    {
+      "epoch": 1.8034461495503251,
+      "grad_norm": 0.17053164541721344,
+      "learning_rate": 1.8928474705182002e-05,
+      "loss": 1.219,
+      "step": 4844
+    },
+    {
+      "epoch": 1.8038184547009344,
+      "grad_norm": 0.1679467409849167,
+      "learning_rate": 1.8927927310386827e-05,
+      "loss": 1.1871,
+      "step": 4845
+    },
+    {
+      "epoch": 1.8041907598515432,
+      "grad_norm": 0.18097002804279327,
+      "learning_rate": 1.892737978372672e-05,
+      "loss": 1.2076,
+      "step": 4846
+    },
+    {
+      "epoch": 1.8045630650021525,
+      "grad_norm": 0.16347897052764893,
+      "learning_rate": 1.8926832125209776e-05,
+      "loss": 1.1959,
+      "step": 4847
+    },
+    {
+      "epoch": 1.8049353701527615,
+      "grad_norm": 0.16471904516220093,
+      "learning_rate": 1.8926284334844086e-05,
+      "loss": 1.1894,
+      "step": 4848
+    },
+    {
+      "epoch": 1.8053076753033706,
+      "grad_norm": 0.17791272699832916,
+      "learning_rate": 1.8925736412637734e-05,
+      "loss": 1.1922,
+      "step": 4849
+    },
+    {
+      "epoch": 1.8056799804539796,
+      "grad_norm": 0.16574202477931976,
+      "learning_rate": 1.8925188358598815e-05,
+      "loss": 1.2123,
+      "step": 4850
+    },
+    {
+      "epoch": 1.8060522856045886,
+      "grad_norm": 0.16811169683933258,
+      "learning_rate": 1.8924640172735423e-05,
+      "loss": 1.2018,
+      "step": 4851
+    },
+    {
+      "epoch": 1.8064245907551977,
+      "grad_norm": 0.1643866002559662,
+      "learning_rate": 1.8924091855055656e-05,
+      "loss": 1.1997,
+      "step": 4852
+    },
+    {
+      "epoch": 1.8067968959058067,
+      "grad_norm": 0.1829267144203186,
+      "learning_rate": 1.8923543405567612e-05,
+      "loss": 1.1984,
+      "step": 4853
+    },
+    {
+      "epoch": 1.807169201056416,
+      "grad_norm": 0.17171823978424072,
+      "learning_rate": 1.8922994824279394e-05,
+      "loss": 1.2109,
+      "step": 4854
+    },
+    {
+      "epoch": 1.8075415062070248,
+      "grad_norm": 0.1727183312177658,
+      "learning_rate": 1.89224461111991e-05,
+      "loss": 1.1947,
+      "step": 4855
+    },
+    {
+      "epoch": 1.807913811357634,
+      "grad_norm": 0.23297062516212463,
+      "learning_rate": 1.8921897266334837e-05,
+      "loss": 1.207,
+      "step": 4856
+    },
+    {
+      "epoch": 1.8082861165082431,
+      "grad_norm": 0.1721329540014267,
+      "learning_rate": 1.8921348289694713e-05,
+      "loss": 1.1893,
+      "step": 4857
+    },
+    {
+      "epoch": 1.8086584216588522,
+      "grad_norm": 0.17270666360855103,
+      "learning_rate": 1.8920799181286837e-05,
+      "loss": 1.2045,
+      "step": 4858
+    },
+    {
+      "epoch": 1.8090307268094612,
+      "grad_norm": 0.16383390128612518,
+      "learning_rate": 1.8920249941119313e-05,
+      "loss": 1.1815,
+      "step": 4859
+    },
+    {
+      "epoch": 1.8094030319600702,
+      "grad_norm": 0.16056787967681885,
+      "learning_rate": 1.891970056920026e-05,
+      "loss": 1.215,
+      "step": 4860
+    },
+    {
+      "epoch": 1.8097753371106795,
+      "grad_norm": 0.1632789522409439,
+      "learning_rate": 1.8919151065537788e-05,
+      "loss": 1.2041,
+      "step": 4861
+    },
+    {
+      "epoch": 1.8101476422612883,
+      "grad_norm": 0.17091669142246246,
+      "learning_rate": 1.8918601430140012e-05,
+      "loss": 1.2226,
+      "step": 4862
+    },
+    {
+      "epoch": 1.8105199474118976,
+      "grad_norm": 0.16258689761161804,
+      "learning_rate": 1.8918051663015057e-05,
+      "loss": 1.1869,
+      "step": 4863
+    },
+    {
+      "epoch": 1.8108922525625064,
+      "grad_norm": 0.1618296056985855,
+      "learning_rate": 1.8917501764171034e-05,
+      "loss": 1.211,
+      "step": 4864
+    },
+    {
+      "epoch": 1.8112645577131157,
+      "grad_norm": 0.1654951572418213,
+      "learning_rate": 1.8916951733616074e-05,
+      "loss": 1.2002,
+      "step": 4865
+    },
+    {
+      "epoch": 1.8116368628637247,
+      "grad_norm": 0.1661575883626938,
+      "learning_rate": 1.8916401571358292e-05,
+      "loss": 1.2041,
+      "step": 4866
+    },
+    {
+      "epoch": 1.8120091680143338,
+      "grad_norm": 0.16568347811698914,
+      "learning_rate": 1.8915851277405823e-05,
+      "loss": 1.2076,
+      "step": 4867
+    },
+    {
+      "epoch": 1.8123814731649428,
+      "grad_norm": 0.1610947996377945,
+      "learning_rate": 1.8915300851766786e-05,
+      "loss": 1.194,
+      "step": 4868
+    },
+    {
+      "epoch": 1.8127537783155518,
+      "grad_norm": 0.16845589876174927,
+      "learning_rate": 1.8914750294449317e-05,
+      "loss": 1.1927,
+      "step": 4869
+    },
+    {
+      "epoch": 1.813126083466161,
+      "grad_norm": 0.16183564066886902,
+      "learning_rate": 1.8914199605461546e-05,
+      "loss": 1.2055,
+      "step": 4870
+    },
+    {
+      "epoch": 1.81349838861677,
+      "grad_norm": 0.1744004338979721,
+      "learning_rate": 1.8913648784811607e-05,
+      "loss": 1.2018,
+      "step": 4871
+    },
+    {
+      "epoch": 1.8138706937673792,
+      "grad_norm": 0.1690053641796112,
+      "learning_rate": 1.8913097832507632e-05,
+      "loss": 1.199,
+      "step": 4872
+    },
+    {
+      "epoch": 1.814242998917988,
+      "grad_norm": 0.16530083119869232,
+      "learning_rate": 1.8912546748557762e-05,
+      "loss": 1.1827,
+      "step": 4873
+    },
+    {
+      "epoch": 1.8146153040685973,
+      "grad_norm": 0.17327569425106049,
+      "learning_rate": 1.891199553297014e-05,
+      "loss": 1.1972,
+      "step": 4874
+    },
+    {
+      "epoch": 1.8149876092192063,
+      "grad_norm": 0.163415789604187,
+      "learning_rate": 1.89114441857529e-05,
+      "loss": 1.1924,
+      "step": 4875
+    },
+    {
+      "epoch": 1.8153599143698154,
+      "grad_norm": 0.16614532470703125,
+      "learning_rate": 1.891089270691419e-05,
+      "loss": 1.1986,
+      "step": 4876
+    },
+    {
+      "epoch": 1.8157322195204244,
+      "grad_norm": 0.1664610356092453,
+      "learning_rate": 1.891034109646215e-05,
+      "loss": 1.2012,
+      "step": 4877
+    },
+    {
+      "epoch": 1.8161045246710334,
+      "grad_norm": 0.1688005030155182,
+      "learning_rate": 1.8909789354404934e-05,
+      "loss": 1.1825,
+      "step": 4878
+    },
+    {
+      "epoch": 1.8164768298216427,
+      "grad_norm": 0.16336117684841156,
+      "learning_rate": 1.890923748075069e-05,
+      "loss": 1.1941,
+      "step": 4879
+    },
+    {
+      "epoch": 1.8168491349722515,
+      "grad_norm": 0.17047706246376038,
+      "learning_rate": 1.8908685475507566e-05,
+      "loss": 1.1988,
+      "step": 4880
+    },
+    {
+      "epoch": 1.8172214401228608,
+      "grad_norm": 0.16434811055660248,
+      "learning_rate": 1.8908133338683715e-05,
+      "loss": 1.204,
+      "step": 4881
+    },
+    {
+      "epoch": 1.8175937452734696,
+      "grad_norm": 0.21952207386493683,
+      "learning_rate": 1.8907581070287295e-05,
+      "loss": 1.1928,
+      "step": 4882
+    },
+    {
+      "epoch": 1.8179660504240789,
+      "grad_norm": 0.1890069991350174,
+      "learning_rate": 1.8907028670326462e-05,
+      "loss": 1.2033,
+      "step": 4883
+    },
+    {
+      "epoch": 1.818338355574688,
+      "grad_norm": 0.17524456977844238,
+      "learning_rate": 1.8906476138809374e-05,
+      "loss": 1.1934,
+      "step": 4884
+    },
+    {
+      "epoch": 1.818710660725297,
+      "grad_norm": 0.17356589436531067,
+      "learning_rate": 1.890592347574419e-05,
+      "loss": 1.1977,
+      "step": 4885
+    },
+    {
+      "epoch": 1.819082965875906,
+      "grad_norm": 0.18048810958862305,
+      "learning_rate": 1.8905370681139083e-05,
+      "loss": 1.2109,
+      "step": 4886
+    },
+    {
+      "epoch": 1.819455271026515,
+      "grad_norm": 0.16388888657093048,
+      "learning_rate": 1.89048177550022e-05,
+      "loss": 1.1987,
+      "step": 4887
+    },
+    {
+      "epoch": 1.8198275761771243,
+      "grad_norm": 0.1702636033296585,
+      "learning_rate": 1.8904264697341723e-05,
+      "loss": 1.2081,
+      "step": 4888
+    },
+    {
+      "epoch": 1.8201998813277331,
+      "grad_norm": 0.16600562632083893,
+      "learning_rate": 1.8903711508165816e-05,
+      "loss": 1.1876,
+      "step": 4889
+    },
+    {
+      "epoch": 1.8205721864783424,
+      "grad_norm": 0.17300425469875336,
+      "learning_rate": 1.8903158187482646e-05,
+      "loss": 1.1927,
+      "step": 4890
+    },
+    {
+      "epoch": 1.8209444916289512,
+      "grad_norm": 0.18208463490009308,
+      "learning_rate": 1.890260473530039e-05,
+      "loss": 1.1933,
+      "step": 4891
+    },
+    {
+      "epoch": 1.8213167967795605,
+      "grad_norm": 0.16697552800178528,
+      "learning_rate": 1.890205115162722e-05,
+      "loss": 1.209,
+      "step": 4892
+    },
+    {
+      "epoch": 1.8216891019301695,
+      "grad_norm": 0.16606757044792175,
+      "learning_rate": 1.8901497436471314e-05,
+      "loss": 1.2104,
+      "step": 4893
+    },
+    {
+      "epoch": 1.8220614070807786,
+      "grad_norm": 0.16475386917591095,
+      "learning_rate": 1.890094358984085e-05,
+      "loss": 1.2086,
+      "step": 4894
+    },
+    {
+      "epoch": 1.8224337122313876,
+      "grad_norm": 0.16822832822799683,
+      "learning_rate": 1.890038961174401e-05,
+      "loss": 1.2009,
+      "step": 4895
+    },
+    {
+      "epoch": 1.8228060173819967,
+      "grad_norm": 0.1646614372730255,
+      "learning_rate": 1.889983550218897e-05,
+      "loss": 1.1945,
+      "step": 4896
+    },
+    {
+      "epoch": 1.823178322532606,
+      "grad_norm": 0.16594818234443665,
+      "learning_rate": 1.8899281261183916e-05,
+      "loss": 1.2005,
+      "step": 4897
+    },
+    {
+      "epoch": 1.8235506276832147,
+      "grad_norm": 0.15717746317386627,
+      "learning_rate": 1.889872688873704e-05,
+      "loss": 1.1948,
+      "step": 4898
+    },
+    {
+      "epoch": 1.823922932833824,
+      "grad_norm": 0.16720950603485107,
+      "learning_rate": 1.8898172384856526e-05,
+      "loss": 1.1799,
+      "step": 4899
+    },
+    {
+      "epoch": 1.8242952379844328,
+      "grad_norm": 0.17484848201274872,
+      "learning_rate": 1.8897617749550565e-05,
+      "loss": 1.1892,
+      "step": 4900
+    },
+    {
+      "epoch": 1.824667543135042,
+      "grad_norm": 0.16109175980091095,
+      "learning_rate": 1.8897062982827347e-05,
+      "loss": 1.1825,
+      "step": 4901
+    },
+    {
+      "epoch": 1.8250398482856511,
+      "grad_norm": 0.1625639796257019,
+      "learning_rate": 1.8896508084695068e-05,
+      "loss": 1.201,
+      "step": 4902
+    },
+    {
+      "epoch": 1.8254121534362602,
+      "grad_norm": 0.16512739658355713,
+      "learning_rate": 1.889595305516192e-05,
+      "loss": 1.2067,
+      "step": 4903
+    },
+    {
+      "epoch": 1.8257844585868692,
+      "grad_norm": 0.15775848925113678,
+      "learning_rate": 1.889539789423611e-05,
+      "loss": 1.1929,
+      "step": 4904
+    },
+    {
+      "epoch": 1.8261567637374783,
+      "grad_norm": 0.1584419310092926,
+      "learning_rate": 1.8894842601925823e-05,
+      "loss": 1.1877,
+      "step": 4905
+    },
+    {
+      "epoch": 1.8265290688880875,
+      "grad_norm": 0.16281954944133759,
+      "learning_rate": 1.8894287178239274e-05,
+      "loss": 1.1972,
+      "step": 4906
+    },
+    {
+      "epoch": 1.8269013740386963,
+      "grad_norm": 0.16380088031291962,
+      "learning_rate": 1.889373162318466e-05,
+      "loss": 1.18,
+      "step": 4907
+    },
+    {
+      "epoch": 1.8272736791893056,
+      "grad_norm": 0.17226645350456238,
+      "learning_rate": 1.8893175936770188e-05,
+      "loss": 1.2165,
+      "step": 4908
+    },
+    {
+      "epoch": 1.8276459843399147,
+      "grad_norm": 0.16510790586471558,
+      "learning_rate": 1.8892620119004067e-05,
+      "loss": 1.2002,
+      "step": 4909
+    },
+    {
+      "epoch": 1.8280182894905237,
+      "grad_norm": 0.18165461719036102,
+      "learning_rate": 1.8892064169894504e-05,
+      "loss": 1.1977,
+      "step": 4910
+    },
+    {
+      "epoch": 1.8283905946411327,
+      "grad_norm": 0.2052212953567505,
+      "learning_rate": 1.889150808944971e-05,
+      "loss": 1.1974,
+      "step": 4911
+    },
+    {
+      "epoch": 1.8287628997917418,
+      "grad_norm": 0.18859870731830597,
+      "learning_rate": 1.8890951877677903e-05,
+      "loss": 1.1985,
+      "step": 4912
+    },
+    {
+      "epoch": 1.8291352049423508,
+      "grad_norm": 0.16892032325267792,
+      "learning_rate": 1.889039553458729e-05,
+      "loss": 1.2059,
+      "step": 4913
+    },
+    {
+      "epoch": 1.8295075100929599,
+      "grad_norm": 0.16577079892158508,
+      "learning_rate": 1.8889839060186095e-05,
+      "loss": 1.1789,
+      "step": 4914
+    },
+    {
+      "epoch": 1.8298798152435691,
+      "grad_norm": 0.18138210475444794,
+      "learning_rate": 1.8889282454482538e-05,
+      "loss": 1.2013,
+      "step": 4915
+    },
+    {
+      "epoch": 1.830252120394178,
+      "grad_norm": 0.17776082456111908,
+      "learning_rate": 1.8888725717484834e-05,
+      "loss": 1.2137,
+      "step": 4916
+    },
+    {
+      "epoch": 1.8306244255447872,
+      "grad_norm": 0.1690797209739685,
+      "learning_rate": 1.888816884920121e-05,
+      "loss": 1.2019,
+      "step": 4917
+    },
+    {
+      "epoch": 1.8309967306953963,
+      "grad_norm": 0.17596592009067535,
+      "learning_rate": 1.888761184963989e-05,
+      "loss": 1.209,
+      "step": 4918
+    },
+    {
+      "epoch": 1.8313690358460053,
+      "grad_norm": 0.17080815136432648,
+      "learning_rate": 1.88870547188091e-05,
+      "loss": 1.1965,
+      "step": 4919
+    },
+    {
+      "epoch": 1.8317413409966143,
+      "grad_norm": 0.21370607614517212,
+      "learning_rate": 1.8886497456717073e-05,
+      "loss": 1.1991,
+      "step": 4920
+    },
+    {
+      "epoch": 1.8321136461472234,
+      "grad_norm": 0.18233619630336761,
+      "learning_rate": 1.888594006337203e-05,
+      "loss": 1.2039,
+      "step": 4921
+    },
+    {
+      "epoch": 1.8324859512978326,
+      "grad_norm": 0.1737797111272812,
+      "learning_rate": 1.8885382538782213e-05,
+      "loss": 1.1976,
+      "step": 4922
+    },
+    {
+      "epoch": 1.8328582564484415,
+      "grad_norm": 0.16964221000671387,
+      "learning_rate": 1.8884824882955853e-05,
+      "loss": 1.1955,
+      "step": 4923
+    },
+    {
+      "epoch": 1.8332305615990507,
+      "grad_norm": 0.18157334625720978,
+      "learning_rate": 1.888426709590119e-05,
+      "loss": 1.1951,
+      "step": 4924
+    },
+    {
+      "epoch": 1.8336028667496596,
+      "grad_norm": 0.1712358593940735,
+      "learning_rate": 1.8883709177626456e-05,
+      "loss": 1.1945,
+      "step": 4925
+    },
+    {
+      "epoch": 1.8339751719002688,
+      "grad_norm": 0.16328178346157074,
+      "learning_rate": 1.8883151128139898e-05,
+      "loss": 1.194,
+      "step": 4926
+    },
+    {
+      "epoch": 1.8343474770508779,
+      "grad_norm": 0.16288742423057556,
+      "learning_rate": 1.8882592947449753e-05,
+      "loss": 1.1876,
+      "step": 4927
+    },
+    {
+      "epoch": 1.834719782201487,
+      "grad_norm": 0.1691894680261612,
+      "learning_rate": 1.8882034635564266e-05,
+      "loss": 1.2148,
+      "step": 4928
+    },
+    {
+      "epoch": 1.835092087352096,
+      "grad_norm": 0.16904255747795105,
+      "learning_rate": 1.888147619249169e-05,
+      "loss": 1.1911,
+      "step": 4929
+    },
+    {
+      "epoch": 1.835464392502705,
+      "grad_norm": 0.1638314574956894,
+      "learning_rate": 1.8880917618240265e-05,
+      "loss": 1.1922,
+      "step": 4930
+    },
+    {
+      "epoch": 1.8358366976533143,
+      "grad_norm": 0.17079859972000122,
+      "learning_rate": 1.888035891281824e-05,
+      "loss": 1.1861,
+      "step": 4931
+    },
+    {
+      "epoch": 1.836209002803923,
+      "grad_norm": 0.1680002361536026,
+      "learning_rate": 1.8879800076233875e-05,
+      "loss": 1.2024,
+      "step": 4932
+    },
+    {
+      "epoch": 1.8365813079545323,
+      "grad_norm": 0.15936627984046936,
+      "learning_rate": 1.8879241108495423e-05,
+      "loss": 1.1981,
+      "step": 4933
+    },
+    {
+      "epoch": 1.8369536131051412,
+      "grad_norm": 0.16010630130767822,
+      "learning_rate": 1.8878682009611134e-05,
+      "loss": 1.1983,
+      "step": 4934
+    },
+    {
+      "epoch": 1.8373259182557504,
+      "grad_norm": 0.16550739109516144,
+      "learning_rate": 1.887812277958927e-05,
+      "loss": 1.2154,
+      "step": 4935
+    },
+    {
+      "epoch": 1.8376982234063595,
+      "grad_norm": 0.16892309486865997,
+      "learning_rate": 1.887756341843809e-05,
+      "loss": 1.2044,
+      "step": 4936
+    },
+    {
+      "epoch": 1.8380705285569685,
+      "grad_norm": 0.16828759014606476,
+      "learning_rate": 1.8877003926165852e-05,
+      "loss": 1.2074,
+      "step": 4937
+    },
+    {
+      "epoch": 1.8384428337075776,
+      "grad_norm": 0.16481132805347443,
+      "learning_rate": 1.8876444302780826e-05,
+      "loss": 1.2012,
+      "step": 4938
+    },
+    {
+      "epoch": 1.8388151388581866,
+      "grad_norm": 0.1642160415649414,
+      "learning_rate": 1.8875884548291274e-05,
+      "loss": 1.195,
+      "step": 4939
+    },
+    {
+      "epoch": 1.8391874440087959,
+      "grad_norm": 0.15789036452770233,
+      "learning_rate": 1.8875324662705467e-05,
+      "loss": 1.2151,
+      "step": 4940
+    },
+    {
+      "epoch": 1.8395597491594047,
+      "grad_norm": 0.1689959466457367,
+      "learning_rate": 1.8874764646031665e-05,
+      "loss": 1.1944,
+      "step": 4941
+    },
+    {
+      "epoch": 1.839932054310014,
+      "grad_norm": 0.17505736649036407,
+      "learning_rate": 1.8874204498278153e-05,
+      "loss": 1.2001,
+      "step": 4942
+    },
+    {
+      "epoch": 1.8403043594606228,
+      "grad_norm": 0.1680225431919098,
+      "learning_rate": 1.8873644219453194e-05,
+      "loss": 1.2081,
+      "step": 4943
+    },
+    {
+      "epoch": 1.840676664611232,
+      "grad_norm": 0.17328615486621857,
+      "learning_rate": 1.887308380956507e-05,
+      "loss": 1.2141,
+      "step": 4944
+    },
+    {
+      "epoch": 1.841048969761841,
+      "grad_norm": 0.16859564185142517,
+      "learning_rate": 1.887252326862205e-05,
+      "loss": 1.1836,
+      "step": 4945
+    },
+    {
+      "epoch": 1.8414212749124501,
+      "grad_norm": 0.17477625608444214,
+      "learning_rate": 1.887196259663242e-05,
+      "loss": 1.2045,
+      "step": 4946
+    },
+    {
+      "epoch": 1.8417935800630592,
+      "grad_norm": 0.16840267181396484,
+      "learning_rate": 1.8871401793604463e-05,
+      "loss": 1.2002,
+      "step": 4947
+    },
+    {
+      "epoch": 1.8421658852136682,
+      "grad_norm": 0.16595827043056488,
+      "learning_rate": 1.8870840859546455e-05,
+      "loss": 1.1985,
+      "step": 4948
+    },
+    {
+      "epoch": 1.8425381903642775,
+      "grad_norm": 0.17932920157909393,
+      "learning_rate": 1.8870279794466686e-05,
+      "loss": 1.1966,
+      "step": 4949
+    },
+    {
+      "epoch": 1.8429104955148863,
+      "grad_norm": 0.17470277845859528,
+      "learning_rate": 1.8869718598373438e-05,
+      "loss": 1.1887,
+      "step": 4950
+    },
+    {
+      "epoch": 1.8432828006654955,
+      "grad_norm": 0.16657747328281403,
+      "learning_rate": 1.8869157271275008e-05,
+      "loss": 1.191,
+      "step": 4951
+    },
+    {
+      "epoch": 1.8436551058161044,
+      "grad_norm": 0.1673271358013153,
+      "learning_rate": 1.8868595813179677e-05,
+      "loss": 1.192,
+      "step": 4952
+    },
+    {
+      "epoch": 1.8440274109667136,
+      "grad_norm": 0.1705188900232315,
+      "learning_rate": 1.8868034224095742e-05,
+      "loss": 1.2007,
+      "step": 4953
+    },
+    {
+      "epoch": 1.8443997161173227,
+      "grad_norm": 0.1590953916311264,
+      "learning_rate": 1.88674725040315e-05,
+      "loss": 1.2006,
+      "step": 4954
+    },
+    {
+      "epoch": 1.8447720212679317,
+      "grad_norm": 0.16473130881786346,
+      "learning_rate": 1.8866910652995244e-05,
+      "loss": 1.2069,
+      "step": 4955
+    },
+    {
+      "epoch": 1.8451443264185408,
+      "grad_norm": 0.16664549708366394,
+      "learning_rate": 1.886634867099528e-05,
+      "loss": 1.1985,
+      "step": 4956
+    },
+    {
+      "epoch": 1.8455166315691498,
+      "grad_norm": 0.1601409912109375,
+      "learning_rate": 1.8865786558039895e-05,
+      "loss": 1.1817,
+      "step": 4957
+    },
+    {
+      "epoch": 1.845888936719759,
+      "grad_norm": 0.16703009605407715,
+      "learning_rate": 1.8865224314137404e-05,
+      "loss": 1.1998,
+      "step": 4958
+    },
+    {
+      "epoch": 1.8462612418703679,
+      "grad_norm": 0.16715413331985474,
+      "learning_rate": 1.88646619392961e-05,
+      "loss": 1.1901,
+      "step": 4959
+    },
+    {
+      "epoch": 1.8466335470209772,
+      "grad_norm": 0.18171849846839905,
+      "learning_rate": 1.8864099433524302e-05,
+      "loss": 1.2008,
+      "step": 4960
+    },
+    {
+      "epoch": 1.847005852171586,
+      "grad_norm": 0.20484322309494019,
+      "learning_rate": 1.886353679683031e-05,
+      "loss": 1.1864,
+      "step": 4961
+    },
+    {
+      "epoch": 1.8473781573221952,
+      "grad_norm": 0.19610385596752167,
+      "learning_rate": 1.8862974029222438e-05,
+      "loss": 1.2133,
+      "step": 4962
+    },
+    {
+      "epoch": 1.8477504624728043,
+      "grad_norm": 0.18955834209918976,
+      "learning_rate": 1.8862411130708992e-05,
+      "loss": 1.2175,
+      "step": 4963
+    },
+    {
+      "epoch": 1.8481227676234133,
+      "grad_norm": 0.2106134295463562,
+      "learning_rate": 1.8861848101298287e-05,
+      "loss": 1.2053,
+      "step": 4964
+    },
+    {
+      "epoch": 1.8484950727740224,
+      "grad_norm": 0.16345474123954773,
+      "learning_rate": 1.8861284940998647e-05,
+      "loss": 1.2037,
+      "step": 4965
+    },
+    {
+      "epoch": 1.8488673779246314,
+      "grad_norm": 0.16316930949687958,
+      "learning_rate": 1.8860721649818383e-05,
+      "loss": 1.197,
+      "step": 4966
+    },
+    {
+      "epoch": 1.8492396830752407,
+      "grad_norm": 0.16572824120521545,
+      "learning_rate": 1.8860158227765816e-05,
+      "loss": 1.1979,
+      "step": 4967
+    },
+    {
+      "epoch": 1.8496119882258495,
+      "grad_norm": 0.16799062490463257,
+      "learning_rate": 1.8859594674849267e-05,
+      "loss": 1.1983,
+      "step": 4968
+    },
+    {
+      "epoch": 1.8499842933764588,
+      "grad_norm": 0.15991763770580292,
+      "learning_rate": 1.8859030991077062e-05,
+      "loss": 1.2133,
+      "step": 4969
+    },
+    {
+      "epoch": 1.8503565985270678,
+      "grad_norm": 0.1664821207523346,
+      "learning_rate": 1.8858467176457528e-05,
+      "loss": 1.2058,
+      "step": 4970
+    },
+    {
+      "epoch": 1.8507289036776768,
+      "grad_norm": 0.16562390327453613,
+      "learning_rate": 1.8857903230998986e-05,
+      "loss": 1.2027,
+      "step": 4971
+    },
+    {
+      "epoch": 1.8511012088282859,
+      "grad_norm": 0.16051267087459564,
+      "learning_rate": 1.885733915470977e-05,
+      "loss": 1.1803,
+      "step": 4972
+    },
+    {
+      "epoch": 1.851473513978895,
+      "grad_norm": 0.1623574048280716,
+      "learning_rate": 1.8856774947598212e-05,
+      "loss": 1.1947,
+      "step": 4973
+    },
+    {
+      "epoch": 1.851845819129504,
+      "grad_norm": 0.16899898648262024,
+      "learning_rate": 1.8856210609672643e-05,
+      "loss": 1.2068,
+      "step": 4974
+    },
+    {
+      "epoch": 1.852218124280113,
+      "grad_norm": 0.16852141916751862,
+      "learning_rate": 1.88556461409414e-05,
+      "loss": 1.1886,
+      "step": 4975
+    },
+    {
+      "epoch": 1.8525904294307223,
+      "grad_norm": 0.16997982561588287,
+      "learning_rate": 1.8855081541412814e-05,
+      "loss": 1.1986,
+      "step": 4976
+    },
+    {
+      "epoch": 1.852962734581331,
+      "grad_norm": 0.17644338309764862,
+      "learning_rate": 1.8854516811095234e-05,
+      "loss": 1.1934,
+      "step": 4977
+    },
+    {
+      "epoch": 1.8533350397319404,
+      "grad_norm": 0.18668413162231445,
+      "learning_rate": 1.8853951949996997e-05,
+      "loss": 1.1824,
+      "step": 4978
+    },
+    {
+      "epoch": 1.8537073448825494,
+      "grad_norm": 0.16832074522972107,
+      "learning_rate": 1.8853386958126444e-05,
+      "loss": 1.1913,
+      "step": 4979
+    },
+    {
+      "epoch": 1.8540796500331584,
+      "grad_norm": 0.18081939220428467,
+      "learning_rate": 1.885282183549192e-05,
+      "loss": 1.1971,
+      "step": 4980
+    },
+    {
+      "epoch": 1.8544519551837675,
+      "grad_norm": 0.17961086332798004,
+      "learning_rate": 1.8852256582101772e-05,
+      "loss": 1.206,
+      "step": 4981
+    },
+    {
+      "epoch": 1.8548242603343765,
+      "grad_norm": 0.16785657405853271,
+      "learning_rate": 1.8851691197964356e-05,
+      "loss": 1.1924,
+      "step": 4982
+    },
+    {
+      "epoch": 1.8551965654849858,
+      "grad_norm": 0.17671982944011688,
+      "learning_rate": 1.885112568308801e-05,
+      "loss": 1.1868,
+      "step": 4983
+    },
+    {
+      "epoch": 1.8555688706355946,
+      "grad_norm": 0.16151179373264313,
+      "learning_rate": 1.8850560037481095e-05,
+      "loss": 1.1999,
+      "step": 4984
+    },
+    {
+      "epoch": 1.8559411757862039,
+      "grad_norm": 0.18711332976818085,
+      "learning_rate": 1.8849994261151968e-05,
+      "loss": 1.1959,
+      "step": 4985
+    },
+    {
+      "epoch": 1.8563134809368127,
+      "grad_norm": 0.16686466336250305,
+      "learning_rate": 1.8849428354108977e-05,
+      "loss": 1.2055,
+      "step": 4986
+    },
+    {
+      "epoch": 1.856685786087422,
+      "grad_norm": 0.17078348994255066,
+      "learning_rate": 1.8848862316360485e-05,
+      "loss": 1.2076,
+      "step": 4987
+    },
+    {
+      "epoch": 1.857058091238031,
+      "grad_norm": 0.18692351877689362,
+      "learning_rate": 1.884829614791485e-05,
+      "loss": 1.2099,
+      "step": 4988
+    },
+    {
+      "epoch": 1.85743039638864,
+      "grad_norm": 0.16382160782814026,
+      "learning_rate": 1.884772984878044e-05,
+      "loss": 1.2002,
+      "step": 4989
+    },
+    {
+      "epoch": 1.857802701539249,
+      "grad_norm": 0.16507616639137268,
+      "learning_rate": 1.8847163418965613e-05,
+      "loss": 1.2038,
+      "step": 4990
+    },
+    {
+      "epoch": 1.8581750066898581,
+      "grad_norm": 0.1630743145942688,
+      "learning_rate": 1.884659685847874e-05,
+      "loss": 1.2044,
+      "step": 4991
+    },
+    {
+      "epoch": 1.8585473118404674,
+      "grad_norm": 0.18251408636569977,
+      "learning_rate": 1.884603016732818e-05,
+      "loss": 1.1941,
+      "step": 4992
+    },
+    {
+      "epoch": 1.8589196169910762,
+      "grad_norm": 0.1670679897069931,
+      "learning_rate": 1.8845463345522317e-05,
+      "loss": 1.1857,
+      "step": 4993
+    },
+    {
+      "epoch": 1.8592919221416855,
+      "grad_norm": 0.17508168518543243,
+      "learning_rate": 1.8844896393069514e-05,
+      "loss": 1.1886,
+      "step": 4994
+    },
+    {
+      "epoch": 1.8596642272922943,
+      "grad_norm": 0.16918566823005676,
+      "learning_rate": 1.8844329309978146e-05,
+      "loss": 1.2011,
+      "step": 4995
+    },
+    {
+      "epoch": 1.8600365324429036,
+      "grad_norm": 0.18971340358257294,
+      "learning_rate": 1.8843762096256587e-05,
+      "loss": 1.2074,
+      "step": 4996
+    },
+    {
+      "epoch": 1.8604088375935126,
+      "grad_norm": 0.1661243587732315,
+      "learning_rate": 1.8843194751913217e-05,
+      "loss": 1.194,
+      "step": 4997
+    },
+    {
+      "epoch": 1.8607811427441217,
+      "grad_norm": 0.1802155077457428,
+      "learning_rate": 1.8842627276956418e-05,
+      "loss": 1.1916,
+      "step": 4998
+    },
+    {
+      "epoch": 1.8611534478947307,
+      "grad_norm": 0.16783170402050018,
+      "learning_rate": 1.884205967139457e-05,
+      "loss": 1.1981,
+      "step": 4999
+    },
+    {
+      "epoch": 1.8615257530453397,
+      "grad_norm": 0.19096462428569794,
+      "learning_rate": 1.884149193523605e-05,
+      "loss": 1.2114,
+      "step": 5000
+    },
+    {
+      "epoch": 1.8615257530453397,
+      "eval_loss": 1.3059989213943481,
+      "eval_runtime": 16.6737,
+      "eval_samples_per_second": 103.996,
+      "eval_steps_per_second": 5.218,
+      "step": 5000
+    },
+    {
+      "epoch": 1.861898058195949,
+      "grad_norm": 0.16396373510360718,
+      "learning_rate": 1.884092406848925e-05,
+      "loss": 1.2119,
+      "step": 5001
+    },
+    {
+      "epoch": 1.8622703633465578,
+      "grad_norm": 0.17009267210960388,
+      "learning_rate": 1.8840356071162565e-05,
+      "loss": 1.1872,
+      "step": 5002
+    },
+    {
+      "epoch": 1.862642668497167,
+      "grad_norm": 0.16647803783416748,
+      "learning_rate": 1.8839787943264367e-05,
+      "loss": 1.199,
+      "step": 5003
+    },
+    {
+      "epoch": 1.863014973647776,
+      "grad_norm": 0.17351815104484558,
+      "learning_rate": 1.8839219684803057e-05,
+      "loss": 1.2002,
+      "step": 5004
+    },
+    {
+      "epoch": 1.8633872787983852,
+      "grad_norm": 0.17014597356319427,
+      "learning_rate": 1.8838651295787028e-05,
+      "loss": 1.205,
+      "step": 5005
+    },
+    {
+      "epoch": 1.8637595839489942,
+      "grad_norm": 0.17426498234272003,
+      "learning_rate": 1.8838082776224675e-05,
+      "loss": 1.2107,
+      "step": 5006
+    },
+    {
+      "epoch": 1.8641318890996033,
+      "grad_norm": 0.17104274034500122,
+      "learning_rate": 1.883751412612439e-05,
+      "loss": 1.2051,
+      "step": 5007
+    },
+    {
+      "epoch": 1.8645041942502123,
+      "grad_norm": 0.16714368760585785,
+      "learning_rate": 1.8836945345494584e-05,
+      "loss": 1.1975,
+      "step": 5008
+    },
+    {
+      "epoch": 1.8648764994008213,
+      "grad_norm": 0.1665792465209961,
+      "learning_rate": 1.8836376434343644e-05,
+      "loss": 1.2,
+      "step": 5009
+    },
+    {
+      "epoch": 1.8652488045514306,
+      "grad_norm": 0.16620507836341858,
+      "learning_rate": 1.8835807392679978e-05,
+      "loss": 1.1974,
+      "step": 5010
+    },
+    {
+      "epoch": 1.8656211097020394,
+      "grad_norm": 0.16727782785892487,
+      "learning_rate": 1.8835238220511997e-05,
+      "loss": 1.2209,
+      "step": 5011
+    },
+    {
+      "epoch": 1.8659934148526487,
+      "grad_norm": 0.17242635786533356,
+      "learning_rate": 1.8834668917848097e-05,
+      "loss": 1.2057,
+      "step": 5012
+    },
+    {
+      "epoch": 1.8663657200032575,
+      "grad_norm": 0.16240327060222626,
+      "learning_rate": 1.883409948469669e-05,
+      "loss": 1.1868,
+      "step": 5013
+    },
+    {
+      "epoch": 1.8667380251538668,
+      "grad_norm": 0.16446274518966675,
+      "learning_rate": 1.8833529921066193e-05,
+      "loss": 1.1848,
+      "step": 5014
+    },
+    {
+      "epoch": 1.8671103303044758,
+      "grad_norm": 0.16615699231624603,
+      "learning_rate": 1.883296022696501e-05,
+      "loss": 1.1968,
+      "step": 5015
+    },
+    {
+      "epoch": 1.8674826354550849,
+      "grad_norm": 0.15990613400936127,
+      "learning_rate": 1.883239040240156e-05,
+      "loss": 1.1975,
+      "step": 5016
+    },
+    {
+      "epoch": 1.867854940605694,
+      "grad_norm": 0.1651614010334015,
+      "learning_rate": 1.8831820447384256e-05,
+      "loss": 1.1967,
+      "step": 5017
+    },
+    {
+      "epoch": 1.868227245756303,
+      "grad_norm": 0.16867800056934357,
+      "learning_rate": 1.8831250361921522e-05,
+      "loss": 1.2167,
+      "step": 5018
+    },
+    {
+      "epoch": 1.8685995509069122,
+      "grad_norm": 0.1658807247877121,
+      "learning_rate": 1.8830680146021773e-05,
+      "loss": 1.1987,
+      "step": 5019
+    },
+    {
+      "epoch": 1.868971856057521,
+      "grad_norm": 0.1625361442565918,
+      "learning_rate": 1.8830109799693434e-05,
+      "loss": 1.2042,
+      "step": 5020
+    },
+    {
+      "epoch": 1.8693441612081303,
+      "grad_norm": 0.47724783420562744,
+      "learning_rate": 1.882953932294492e-05,
+      "loss": 1.2001,
+      "step": 5021
+    },
+    {
+      "epoch": 1.8697164663587391,
+      "grad_norm": 0.1839105486869812,
+      "learning_rate": 1.882896871578467e-05,
+      "loss": 1.1972,
+      "step": 5022
+    },
+    {
+      "epoch": 1.8700887715093484,
+      "grad_norm": 0.22842104732990265,
+      "learning_rate": 1.8828397978221108e-05,
+      "loss": 1.1797,
+      "step": 5023
+    },
+    {
+      "epoch": 1.8704610766599574,
+      "grad_norm": 0.18718920648097992,
+      "learning_rate": 1.882782711026266e-05,
+      "loss": 1.1971,
+      "step": 5024
+    },
+    {
+      "epoch": 1.8708333818105665,
+      "grad_norm": 0.17810329794883728,
+      "learning_rate": 1.8827256111917757e-05,
+      "loss": 1.1894,
+      "step": 5025
+    },
+    {
+      "epoch": 1.8712056869611755,
+      "grad_norm": 0.16336943209171295,
+      "learning_rate": 1.882668498319484e-05,
+      "loss": 1.2008,
+      "step": 5026
+    },
+    {
+      "epoch": 1.8715779921117845,
+      "grad_norm": 0.18069854378700256,
+      "learning_rate": 1.882611372410234e-05,
+      "loss": 1.198,
+      "step": 5027
+    },
+    {
+      "epoch": 1.8719502972623938,
+      "grad_norm": 0.17028245329856873,
+      "learning_rate": 1.8825542334648687e-05,
+      "loss": 1.1885,
+      "step": 5028
+    },
+    {
+      "epoch": 1.8723226024130026,
+      "grad_norm": 0.16441023349761963,
+      "learning_rate": 1.8824970814842332e-05,
+      "loss": 1.1954,
+      "step": 5029
+    },
+    {
+      "epoch": 1.872694907563612,
+      "grad_norm": 0.16318555176258087,
+      "learning_rate": 1.8824399164691712e-05,
+      "loss": 1.1845,
+      "step": 5030
+    },
+    {
+      "epoch": 1.873067212714221,
+      "grad_norm": 0.16445982456207275,
+      "learning_rate": 1.882382738420527e-05,
+      "loss": 1.1954,
+      "step": 5031
+    },
+    {
+      "epoch": 1.87343951786483,
+      "grad_norm": 0.16694188117980957,
+      "learning_rate": 1.8823255473391454e-05,
+      "loss": 1.2076,
+      "step": 5032
+    },
+    {
+      "epoch": 1.873811823015439,
+      "grad_norm": 0.17121298611164093,
+      "learning_rate": 1.8822683432258703e-05,
+      "loss": 1.2013,
+      "step": 5033
+    },
+    {
+      "epoch": 1.874184128166048,
+      "grad_norm": 0.16623573005199432,
+      "learning_rate": 1.8822111260815475e-05,
+      "loss": 1.1941,
+      "step": 5034
+    },
+    {
+      "epoch": 1.8745564333166571,
+      "grad_norm": 0.16473905742168427,
+      "learning_rate": 1.882153895907022e-05,
+      "loss": 1.1876,
+      "step": 5035
+    },
+    {
+      "epoch": 1.8749287384672662,
+      "grad_norm": 0.1598900854587555,
+      "learning_rate": 1.8820966527031383e-05,
+      "loss": 1.2061,
+      "step": 5036
+    },
+    {
+      "epoch": 1.8753010436178754,
+      "grad_norm": 0.16518044471740723,
+      "learning_rate": 1.8820393964707424e-05,
+      "loss": 1.1909,
+      "step": 5037
+    },
+    {
+      "epoch": 1.8756733487684842,
+      "grad_norm": 0.1635928601026535,
+      "learning_rate": 1.8819821272106803e-05,
+      "loss": 1.1977,
+      "step": 5038
+    },
+    {
+      "epoch": 1.8760456539190935,
+      "grad_norm": 0.16230987012386322,
+      "learning_rate": 1.8819248449237973e-05,
+      "loss": 1.1884,
+      "step": 5039
+    },
+    {
+      "epoch": 1.8764179590697025,
+      "grad_norm": 0.16847163438796997,
+      "learning_rate": 1.8818675496109398e-05,
+      "loss": 1.1951,
+      "step": 5040
+    },
+    {
+      "epoch": 1.8767902642203116,
+      "grad_norm": 0.16432876884937286,
+      "learning_rate": 1.8818102412729537e-05,
+      "loss": 1.1996,
+      "step": 5041
+    },
+    {
+      "epoch": 1.8771625693709206,
+      "grad_norm": 0.16060929000377655,
+      "learning_rate": 1.8817529199106858e-05,
+      "loss": 1.1812,
+      "step": 5042
+    },
+    {
+      "epoch": 1.8775348745215297,
+      "grad_norm": 0.15831544995307922,
+      "learning_rate": 1.8816955855249827e-05,
+      "loss": 1.2125,
+      "step": 5043
+    },
+    {
+      "epoch": 1.877907179672139,
+      "grad_norm": 0.1665830910205841,
+      "learning_rate": 1.8816382381166912e-05,
+      "loss": 1.1976,
+      "step": 5044
+    },
+    {
+      "epoch": 1.8782794848227478,
+      "grad_norm": 0.16302713751792908,
+      "learning_rate": 1.8815808776866583e-05,
+      "loss": 1.2007,
+      "step": 5045
+    },
+    {
+      "epoch": 1.878651789973357,
+      "grad_norm": 0.16040857136249542,
+      "learning_rate": 1.881523504235731e-05,
+      "loss": 1.2007,
+      "step": 5046
+    },
+    {
+      "epoch": 1.8790240951239658,
+      "grad_norm": 0.16229379177093506,
+      "learning_rate": 1.8814661177647567e-05,
+      "loss": 1.2095,
+      "step": 5047
+    },
+    {
+      "epoch": 1.879396400274575,
+      "grad_norm": 0.15823791921138763,
+      "learning_rate": 1.8814087182745835e-05,
+      "loss": 1.2027,
+      "step": 5048
+    },
+    {
+      "epoch": 1.8797687054251841,
+      "grad_norm": 0.16429072618484497,
+      "learning_rate": 1.8813513057660586e-05,
+      "loss": 1.1974,
+      "step": 5049
+    },
+    {
+      "epoch": 1.8801410105757932,
+      "grad_norm": 0.16185057163238525,
+      "learning_rate": 1.8812938802400303e-05,
+      "loss": 1.2046,
+      "step": 5050
+    },
+    {
+      "epoch": 1.8805133157264022,
+      "grad_norm": 0.16304458677768707,
+      "learning_rate": 1.8812364416973467e-05,
+      "loss": 1.1941,
+      "step": 5051
+    },
+    {
+      "epoch": 1.8808856208770113,
+      "grad_norm": 0.1603989452123642,
+      "learning_rate": 1.881178990138856e-05,
+      "loss": 1.2074,
+      "step": 5052
+    },
+    {
+      "epoch": 1.8812579260276205,
+      "grad_norm": 0.16219423711299896,
+      "learning_rate": 1.8811215255654074e-05,
+      "loss": 1.1982,
+      "step": 5053
+    },
+    {
+      "epoch": 1.8816302311782294,
+      "grad_norm": 0.16032250225543976,
+      "learning_rate": 1.8810640479778488e-05,
+      "loss": 1.1879,
+      "step": 5054
+    },
+    {
+      "epoch": 1.8820025363288386,
+      "grad_norm": 0.16628342866897583,
+      "learning_rate": 1.88100655737703e-05,
+      "loss": 1.201,
+      "step": 5055
+    },
+    {
+      "epoch": 1.8823748414794474,
+      "grad_norm": 0.15982067584991455,
+      "learning_rate": 1.8809490537637988e-05,
+      "loss": 1.1874,
+      "step": 5056
+    },
+    {
+      "epoch": 1.8827471466300567,
+      "grad_norm": 0.16088199615478516,
+      "learning_rate": 1.880891537139006e-05,
+      "loss": 1.1981,
+      "step": 5057
+    },
+    {
+      "epoch": 1.8831194517806658,
+      "grad_norm": 0.16720576584339142,
+      "learning_rate": 1.8808340075035e-05,
+      "loss": 1.1988,
+      "step": 5058
+    },
+    {
+      "epoch": 1.8834917569312748,
+      "grad_norm": 0.1663171947002411,
+      "learning_rate": 1.880776464858131e-05,
+      "loss": 1.2205,
+      "step": 5059
+    },
+    {
+      "epoch": 1.8838640620818838,
+      "grad_norm": 0.15910829603672028,
+      "learning_rate": 1.8807189092037494e-05,
+      "loss": 1.198,
+      "step": 5060
+    },
+    {
+      "epoch": 1.8842363672324929,
+      "grad_norm": 0.17297664284706116,
+      "learning_rate": 1.8806613405412045e-05,
+      "loss": 1.1931,
+      "step": 5061
+    },
+    {
+      "epoch": 1.8846086723831021,
+      "grad_norm": 0.18729183077812195,
+      "learning_rate": 1.880603758871347e-05,
+      "loss": 1.2086,
+      "step": 5062
+    },
+    {
+      "epoch": 1.884980977533711,
+      "grad_norm": 0.17871494591236115,
+      "learning_rate": 1.880546164195027e-05,
+      "loss": 1.202,
+      "step": 5063
+    },
+    {
+      "epoch": 1.8853532826843202,
+      "grad_norm": 0.16181787848472595,
+      "learning_rate": 1.8804885565130956e-05,
+      "loss": 1.1903,
+      "step": 5064
+    },
+    {
+      "epoch": 1.885725587834929,
+      "grad_norm": 0.16055096685886383,
+      "learning_rate": 1.8804309358264034e-05,
+      "loss": 1.1956,
+      "step": 5065
+    },
+    {
+      "epoch": 1.8860978929855383,
+      "grad_norm": 0.16259528696537018,
+      "learning_rate": 1.8803733021358015e-05,
+      "loss": 1.1927,
+      "step": 5066
+    },
+    {
+      "epoch": 1.8864701981361474,
+      "grad_norm": 0.16635549068450928,
+      "learning_rate": 1.880315655442141e-05,
+      "loss": 1.1886,
+      "step": 5067
+    },
+    {
+      "epoch": 1.8868425032867564,
+      "grad_norm": 0.1678260862827301,
+      "learning_rate": 1.8802579957462738e-05,
+      "loss": 1.216,
+      "step": 5068
+    },
+    {
+      "epoch": 1.8872148084373654,
+      "grad_norm": 0.1632586270570755,
+      "learning_rate": 1.880200323049051e-05,
+      "loss": 1.1899,
+      "step": 5069
+    },
+    {
+      "epoch": 1.8875871135879745,
+      "grad_norm": 0.17265328764915466,
+      "learning_rate": 1.880142637351325e-05,
+      "loss": 1.1961,
+      "step": 5070
+    },
+    {
+      "epoch": 1.8879594187385838,
+      "grad_norm": 0.19594860076904297,
+      "learning_rate": 1.8800849386539476e-05,
+      "loss": 1.2123,
+      "step": 5071
+    },
+    {
+      "epoch": 1.8883317238891926,
+      "grad_norm": 0.1857958883047104,
+      "learning_rate": 1.8800272269577706e-05,
+      "loss": 1.1959,
+      "step": 5072
+    },
+    {
+      "epoch": 1.8887040290398018,
+      "grad_norm": 0.17190240323543549,
+      "learning_rate": 1.8799695022636466e-05,
+      "loss": 1.1803,
+      "step": 5073
+    },
+    {
+      "epoch": 1.8890763341904107,
+      "grad_norm": 0.2080855667591095,
+      "learning_rate": 1.8799117645724282e-05,
+      "loss": 1.1875,
+      "step": 5074
+    },
+    {
+      "epoch": 1.88944863934102,
+      "grad_norm": 0.18487951159477234,
+      "learning_rate": 1.8798540138849685e-05,
+      "loss": 1.2109,
+      "step": 5075
+    },
+    {
+      "epoch": 1.889820944491629,
+      "grad_norm": 0.17496329545974731,
+      "learning_rate": 1.8797962502021203e-05,
+      "loss": 1.196,
+      "step": 5076
+    },
+    {
+      "epoch": 1.890193249642238,
+      "grad_norm": 0.16391214728355408,
+      "learning_rate": 1.8797384735247367e-05,
+      "loss": 1.2139,
+      "step": 5077
+    },
+    {
+      "epoch": 1.890565554792847,
+      "grad_norm": 0.1845993846654892,
+      "learning_rate": 1.8796806838536708e-05,
+      "loss": 1.208,
+      "step": 5078
+    },
+    {
+      "epoch": 1.890937859943456,
+      "grad_norm": 0.17227643728256226,
+      "learning_rate": 1.8796228811897764e-05,
+      "loss": 1.1862,
+      "step": 5079
+    },
+    {
+      "epoch": 1.8913101650940654,
+      "grad_norm": 0.16078023612499237,
+      "learning_rate": 1.8795650655339075e-05,
+      "loss": 1.1944,
+      "step": 5080
+    },
+    {
+      "epoch": 1.8916824702446742,
+      "grad_norm": 0.1699201762676239,
+      "learning_rate": 1.8795072368869176e-05,
+      "loss": 1.2018,
+      "step": 5081
+    },
+    {
+      "epoch": 1.8920547753952834,
+      "grad_norm": 0.17085964977741241,
+      "learning_rate": 1.879449395249661e-05,
+      "loss": 1.1928,
+      "step": 5082
+    },
+    {
+      "epoch": 1.8924270805458925,
+      "grad_norm": 0.19768460094928741,
+      "learning_rate": 1.8793915406229924e-05,
+      "loss": 1.2065,
+      "step": 5083
+    },
+    {
+      "epoch": 1.8927993856965015,
+      "grad_norm": 0.1722254604101181,
+      "learning_rate": 1.8793336730077657e-05,
+      "loss": 1.1908,
+      "step": 5084
+    },
+    {
+      "epoch": 1.8931716908471106,
+      "grad_norm": 0.17522525787353516,
+      "learning_rate": 1.8792757924048354e-05,
+      "loss": 1.1937,
+      "step": 5085
+    },
+    {
+      "epoch": 1.8935439959977196,
+      "grad_norm": 0.17701640725135803,
+      "learning_rate": 1.8792178988150574e-05,
+      "loss": 1.194,
+      "step": 5086
+    },
+    {
+      "epoch": 1.8939163011483287,
+      "grad_norm": 0.17206698656082153,
+      "learning_rate": 1.879159992239286e-05,
+      "loss": 1.2049,
+      "step": 5087
+    },
+    {
+      "epoch": 1.8942886062989377,
+      "grad_norm": 0.1732310801744461,
+      "learning_rate": 1.8791020726783767e-05,
+      "loss": 1.2098,
+      "step": 5088
+    },
+    {
+      "epoch": 1.894660911449547,
+      "grad_norm": 0.1696704626083374,
+      "learning_rate": 1.8790441401331848e-05,
+      "loss": 1.1892,
+      "step": 5089
+    },
+    {
+      "epoch": 1.8950332166001558,
+      "grad_norm": 0.17553074657917023,
+      "learning_rate": 1.8789861946045662e-05,
+      "loss": 1.1939,
+      "step": 5090
+    },
+    {
+      "epoch": 1.895405521750765,
+      "grad_norm": 0.16884693503379822,
+      "learning_rate": 1.8789282360933767e-05,
+      "loss": 1.1892,
+      "step": 5091
+    },
+    {
+      "epoch": 1.895777826901374,
+      "grad_norm": 0.17038117349147797,
+      "learning_rate": 1.8788702646004725e-05,
+      "loss": 1.2057,
+      "step": 5092
+    },
+    {
+      "epoch": 1.8961501320519831,
+      "grad_norm": 0.16676990687847137,
+      "learning_rate": 1.8788122801267094e-05,
+      "loss": 1.212,
+      "step": 5093
+    },
+    {
+      "epoch": 1.8965224372025922,
+      "grad_norm": 0.1800990104675293,
+      "learning_rate": 1.878754282672944e-05,
+      "loss": 1.2153,
+      "step": 5094
+    },
+    {
+      "epoch": 1.8968947423532012,
+      "grad_norm": 0.20043106377124786,
+      "learning_rate": 1.8786962722400334e-05,
+      "loss": 1.1969,
+      "step": 5095
+    },
+    {
+      "epoch": 1.8972670475038103,
+      "grad_norm": 0.1726304590702057,
+      "learning_rate": 1.878638248828834e-05,
+      "loss": 1.2091,
+      "step": 5096
+    },
+    {
+      "epoch": 1.8976393526544193,
+      "grad_norm": 0.17626087367534637,
+      "learning_rate": 1.8785802124402022e-05,
+      "loss": 1.1895,
+      "step": 5097
+    },
+    {
+      "epoch": 1.8980116578050286,
+      "grad_norm": 0.19044530391693115,
+      "learning_rate": 1.8785221630749964e-05,
+      "loss": 1.2022,
+      "step": 5098
+    },
+    {
+      "epoch": 1.8983839629556374,
+      "grad_norm": 0.17916062474250793,
+      "learning_rate": 1.878464100734073e-05,
+      "loss": 1.2064,
+      "step": 5099
+    },
+    {
+      "epoch": 1.8987562681062466,
+      "grad_norm": 0.17990046739578247,
+      "learning_rate": 1.8784060254182904e-05,
+      "loss": 1.1868,
+      "step": 5100
+    },
+    {
+      "epoch": 1.8991285732568557,
+      "grad_norm": 0.1693127304315567,
+      "learning_rate": 1.8783479371285054e-05,
+      "loss": 1.2112,
+      "step": 5101
+    },
+    {
+      "epoch": 1.8995008784074647,
+      "grad_norm": 0.1721697747707367,
+      "learning_rate": 1.8782898358655767e-05,
+      "loss": 1.1971,
+      "step": 5102
+    },
+    {
+      "epoch": 1.8998731835580738,
+      "grad_norm": 0.1762087643146515,
+      "learning_rate": 1.8782317216303624e-05,
+      "loss": 1.1838,
+      "step": 5103
+    },
+    {
+      "epoch": 1.9002454887086828,
+      "grad_norm": 0.2079368680715561,
+      "learning_rate": 1.8781735944237204e-05,
+      "loss": 1.1848,
+      "step": 5104
+    },
+    {
+      "epoch": 1.900617793859292,
+      "grad_norm": 0.18409818410873413,
+      "learning_rate": 1.87811545424651e-05,
+      "loss": 1.1941,
+      "step": 5105
+    },
+    {
+      "epoch": 1.900990099009901,
+      "grad_norm": 0.1844477504491806,
+      "learning_rate": 1.878057301099589e-05,
+      "loss": 1.1866,
+      "step": 5106
+    },
+    {
+      "epoch": 1.9013624041605102,
+      "grad_norm": 0.19511154294013977,
+      "learning_rate": 1.877999134983817e-05,
+      "loss": 1.2066,
+      "step": 5107
+    },
+    {
+      "epoch": 1.901734709311119,
+      "grad_norm": 0.16205884516239166,
+      "learning_rate": 1.877940955900053e-05,
+      "loss": 1.1894,
+      "step": 5108
+    },
+    {
+      "epoch": 1.9021070144617283,
+      "grad_norm": 0.1605387181043625,
+      "learning_rate": 1.8778827638491556e-05,
+      "loss": 1.2014,
+      "step": 5109
+    },
+    {
+      "epoch": 1.9024793196123373,
+      "grad_norm": 0.15819790959358215,
+      "learning_rate": 1.877824558831985e-05,
+      "loss": 1.2027,
+      "step": 5110
+    },
+    {
+      "epoch": 1.9028516247629463,
+      "grad_norm": 0.1695229858160019,
+      "learning_rate": 1.8777663408494013e-05,
+      "loss": 1.181,
+      "step": 5111
+    },
+    {
+      "epoch": 1.9032239299135554,
+      "grad_norm": 0.18196141719818115,
+      "learning_rate": 1.8777081099022636e-05,
+      "loss": 1.1962,
+      "step": 5112
+    },
+    {
+      "epoch": 1.9035962350641644,
+      "grad_norm": 0.18630284070968628,
+      "learning_rate": 1.877649865991432e-05,
+      "loss": 1.2019,
+      "step": 5113
+    },
+    {
+      "epoch": 1.9039685402147737,
+      "grad_norm": 0.16647082567214966,
+      "learning_rate": 1.8775916091177674e-05,
+      "loss": 1.1968,
+      "step": 5114
+    },
+    {
+      "epoch": 1.9043408453653825,
+      "grad_norm": 0.20907706022262573,
+      "learning_rate": 1.8775333392821294e-05,
+      "loss": 1.1929,
+      "step": 5115
+    },
+    {
+      "epoch": 1.9047131505159918,
+      "grad_norm": 0.19712580740451813,
+      "learning_rate": 1.877475056485379e-05,
+      "loss": 1.2029,
+      "step": 5116
+    },
+    {
+      "epoch": 1.9050854556666006,
+      "grad_norm": 0.17977239191532135,
+      "learning_rate": 1.8774167607283772e-05,
+      "loss": 1.198,
+      "step": 5117
+    },
+    {
+      "epoch": 1.9054577608172099,
+      "grad_norm": 0.1739930659532547,
+      "learning_rate": 1.8773584520119848e-05,
+      "loss": 1.1902,
+      "step": 5118
+    },
+    {
+      "epoch": 1.905830065967819,
+      "grad_norm": 0.21335825324058533,
+      "learning_rate": 1.8773001303370634e-05,
+      "loss": 1.1862,
+      "step": 5119
+    },
+    {
+      "epoch": 1.906202371118428,
+      "grad_norm": 0.15986791253089905,
+      "learning_rate": 1.8772417957044743e-05,
+      "loss": 1.2015,
+      "step": 5120
+    },
+    {
+      "epoch": 1.906574676269037,
+      "grad_norm": 0.1675948053598404,
+      "learning_rate": 1.8771834481150782e-05,
+      "loss": 1.1985,
+      "step": 5121
+    },
+    {
+      "epoch": 1.906946981419646,
+      "grad_norm": 0.16435672342777252,
+      "learning_rate": 1.8771250875697383e-05,
+      "loss": 1.2114,
+      "step": 5122
+    },
+    {
+      "epoch": 1.9073192865702553,
+      "grad_norm": 0.16337592899799347,
+      "learning_rate": 1.8770667140693155e-05,
+      "loss": 1.2001,
+      "step": 5123
+    },
+    {
+      "epoch": 1.907691591720864,
+      "grad_norm": 0.1674809753894806,
+      "learning_rate": 1.8770083276146726e-05,
+      "loss": 1.1977,
+      "step": 5124
+    },
+    {
+      "epoch": 1.9080638968714734,
+      "grad_norm": 0.17421855032444,
+      "learning_rate": 1.8769499282066716e-05,
+      "loss": 1.1829,
+      "step": 5125
+    },
+    {
+      "epoch": 1.9084362020220822,
+      "grad_norm": 0.16936717927455902,
+      "learning_rate": 1.8768915158461755e-05,
+      "loss": 1.2072,
+      "step": 5126
+    },
+    {
+      "epoch": 1.9088085071726915,
+      "grad_norm": 0.16133807599544525,
+      "learning_rate": 1.8768330905340462e-05,
+      "loss": 1.2025,
+      "step": 5127
+    },
+    {
+      "epoch": 1.9091808123233005,
+      "grad_norm": 0.1636780947446823,
+      "learning_rate": 1.8767746522711478e-05,
+      "loss": 1.2099,
+      "step": 5128
+    },
+    {
+      "epoch": 1.9095531174739095,
+      "grad_norm": 0.17051872611045837,
+      "learning_rate": 1.876716201058342e-05,
+      "loss": 1.1913,
+      "step": 5129
+    },
+    {
+      "epoch": 1.9099254226245186,
+      "grad_norm": 0.16685360670089722,
+      "learning_rate": 1.8766577368964937e-05,
+      "loss": 1.1992,
+      "step": 5130
+    },
+    {
+      "epoch": 1.9102977277751276,
+      "grad_norm": 0.1661144345998764,
+      "learning_rate": 1.8765992597864654e-05,
+      "loss": 1.1953,
+      "step": 5131
+    },
+    {
+      "epoch": 1.910670032925737,
+      "grad_norm": 0.16702596843242645,
+      "learning_rate": 1.8765407697291208e-05,
+      "loss": 1.2023,
+      "step": 5132
+    },
+    {
+      "epoch": 1.9110423380763457,
+      "grad_norm": 0.16298018395900726,
+      "learning_rate": 1.8764822667253244e-05,
+      "loss": 1.203,
+      "step": 5133
+    },
+    {
+      "epoch": 1.911414643226955,
+      "grad_norm": 0.16323602199554443,
+      "learning_rate": 1.87642375077594e-05,
+      "loss": 1.1987,
+      "step": 5134
+    },
+    {
+      "epoch": 1.9117869483775638,
+      "grad_norm": 0.1665439009666443,
+      "learning_rate": 1.8763652218818314e-05,
+      "loss": 1.1812,
+      "step": 5135
+    },
+    {
+      "epoch": 1.912159253528173,
+      "grad_norm": 0.16453301906585693,
+      "learning_rate": 1.8763066800438638e-05,
+      "loss": 1.2071,
+      "step": 5136
+    },
+    {
+      "epoch": 1.912531558678782,
+      "grad_norm": 0.1619269996881485,
+      "learning_rate": 1.8762481252629013e-05,
+      "loss": 1.2125,
+      "step": 5137
+    },
+    {
+      "epoch": 1.9129038638293911,
+      "grad_norm": 0.16248470544815063,
+      "learning_rate": 1.8761895575398094e-05,
+      "loss": 1.209,
+      "step": 5138
+    },
+    {
+      "epoch": 1.9132761689800002,
+      "grad_norm": 0.16813617944717407,
+      "learning_rate": 1.876130976875452e-05,
+      "loss": 1.1875,
+      "step": 5139
+    },
+    {
+      "epoch": 1.9136484741306092,
+      "grad_norm": 0.16750268638134003,
+      "learning_rate": 1.8760723832706955e-05,
+      "loss": 1.1921,
+      "step": 5140
+    },
+    {
+      "epoch": 1.9140207792812185,
+      "grad_norm": 0.16163413226604462,
+      "learning_rate": 1.8760137767264048e-05,
+      "loss": 1.18,
+      "step": 5141
+    },
+    {
+      "epoch": 1.9143930844318273,
+      "grad_norm": 0.16053326427936554,
+      "learning_rate": 1.875955157243446e-05,
+      "loss": 1.1977,
+      "step": 5142
+    },
+    {
+      "epoch": 1.9147653895824366,
+      "grad_norm": 0.16916437447071075,
+      "learning_rate": 1.8758965248226836e-05,
+      "loss": 1.2144,
+      "step": 5143
+    },
+    {
+      "epoch": 1.9151376947330456,
+      "grad_norm": 0.1655954271554947,
+      "learning_rate": 1.875837879464985e-05,
+      "loss": 1.2031,
+      "step": 5144
+    },
+    {
+      "epoch": 1.9155099998836547,
+      "grad_norm": 0.16790400445461273,
+      "learning_rate": 1.875779221171216e-05,
+      "loss": 1.1969,
+      "step": 5145
+    },
+    {
+      "epoch": 1.9158823050342637,
+      "grad_norm": 0.16594092547893524,
+      "learning_rate": 1.8757205499422428e-05,
+      "loss": 1.191,
+      "step": 5146
+    },
+    {
+      "epoch": 1.9162546101848728,
+      "grad_norm": 0.16297580301761627,
+      "learning_rate": 1.8756618657789322e-05,
+      "loss": 1.2073,
+      "step": 5147
+    },
+    {
+      "epoch": 1.9166269153354818,
+      "grad_norm": 0.16412940621376038,
+      "learning_rate": 1.8756031686821506e-05,
+      "loss": 1.1907,
+      "step": 5148
+    },
+    {
+      "epoch": 1.9169992204860908,
+      "grad_norm": 0.16388413310050964,
+      "learning_rate": 1.875544458652765e-05,
+      "loss": 1.182,
+      "step": 5149
+    },
+    {
+      "epoch": 1.9173715256367,
+      "grad_norm": 0.16367630660533905,
+      "learning_rate": 1.875485735691643e-05,
+      "loss": 1.2072,
+      "step": 5150
+    },
+    {
+      "epoch": 1.917743830787309,
+      "grad_norm": 0.16208967566490173,
+      "learning_rate": 1.8754269997996512e-05,
+      "loss": 1.1871,
+      "step": 5151
+    },
+    {
+      "epoch": 1.9181161359379182,
+      "grad_norm": 0.1770969182252884,
+      "learning_rate": 1.875368250977658e-05,
+      "loss": 1.2146,
+      "step": 5152
+    },
+    {
+      "epoch": 1.9184884410885272,
+      "grad_norm": 0.17094440758228302,
+      "learning_rate": 1.8753094892265308e-05,
+      "loss": 1.1936,
+      "step": 5153
+    },
+    {
+      "epoch": 1.9188607462391363,
+      "grad_norm": 0.16425853967666626,
+      "learning_rate": 1.875250714547137e-05,
+      "loss": 1.1972,
+      "step": 5154
+    },
+    {
+      "epoch": 1.9192330513897453,
+      "grad_norm": 0.16301047801971436,
+      "learning_rate": 1.875191926940345e-05,
+      "loss": 1.1832,
+      "step": 5155
+    },
+    {
+      "epoch": 1.9196053565403544,
+      "grad_norm": 0.16729892790317535,
+      "learning_rate": 1.8751331264070232e-05,
+      "loss": 1.2042,
+      "step": 5156
+    },
+    {
+      "epoch": 1.9199776616909636,
+      "grad_norm": 0.16945315897464752,
+      "learning_rate": 1.87507431294804e-05,
+      "loss": 1.1815,
+      "step": 5157
+    },
+    {
+      "epoch": 1.9203499668415724,
+      "grad_norm": 0.15983742475509644,
+      "learning_rate": 1.8750154865642644e-05,
+      "loss": 1.1798,
+      "step": 5158
+    },
+    {
+      "epoch": 1.9207222719921817,
+      "grad_norm": 0.1647757738828659,
+      "learning_rate": 1.874956647256565e-05,
+      "loss": 1.2057,
+      "step": 5159
+    },
+    {
+      "epoch": 1.9210945771427905,
+      "grad_norm": 0.16134053468704224,
+      "learning_rate": 1.8748977950258105e-05,
+      "loss": 1.2005,
+      "step": 5160
+    },
+    {
+      "epoch": 1.9214668822933998,
+      "grad_norm": 0.16527298092842102,
+      "learning_rate": 1.8748389298728708e-05,
+      "loss": 1.2001,
+      "step": 5161
+    },
+    {
+      "epoch": 1.9218391874440088,
+      "grad_norm": 0.15893900394439697,
+      "learning_rate": 1.8747800517986147e-05,
+      "loss": 1.2154,
+      "step": 5162
+    },
+    {
+      "epoch": 1.9222114925946179,
+      "grad_norm": 0.16622401773929596,
+      "learning_rate": 1.8747211608039124e-05,
+      "loss": 1.1916,
+      "step": 5163
+    },
+    {
+      "epoch": 1.922583797745227,
+      "grad_norm": 0.16279946267604828,
+      "learning_rate": 1.8746622568896334e-05,
+      "loss": 1.1944,
+      "step": 5164
+    },
+    {
+      "epoch": 1.922956102895836,
+      "grad_norm": 0.16495847702026367,
+      "learning_rate": 1.874603340056648e-05,
+      "loss": 1.2027,
+      "step": 5165
+    },
+    {
+      "epoch": 1.9233284080464452,
+      "grad_norm": 0.156333789229393,
+      "learning_rate": 1.8745444103058257e-05,
+      "loss": 1.1909,
+      "step": 5166
+    },
+    {
+      "epoch": 1.923700713197054,
+      "grad_norm": 0.16494396328926086,
+      "learning_rate": 1.8744854676380374e-05,
+      "loss": 1.193,
+      "step": 5167
+    },
+    {
+      "epoch": 1.9240730183476633,
+      "grad_norm": 0.1679079830646515,
+      "learning_rate": 1.8744265120541537e-05,
+      "loss": 1.2151,
+      "step": 5168
+    },
+    {
+      "epoch": 1.9244453234982721,
+      "grad_norm": 0.16192030906677246,
+      "learning_rate": 1.8743675435550453e-05,
+      "loss": 1.2061,
+      "step": 5169
+    },
+    {
+      "epoch": 1.9248176286488814,
+      "grad_norm": 0.16086839139461517,
+      "learning_rate": 1.8743085621415834e-05,
+      "loss": 1.1853,
+      "step": 5170
+    },
+    {
+      "epoch": 1.9251899337994904,
+      "grad_norm": 0.16591200232505798,
+      "learning_rate": 1.8742495678146384e-05,
+      "loss": 1.1983,
+      "step": 5171
+    },
+    {
+      "epoch": 1.9255622389500995,
+      "grad_norm": 0.16515636444091797,
+      "learning_rate": 1.8741905605750826e-05,
+      "loss": 1.1886,
+      "step": 5172
+    },
+    {
+      "epoch": 1.9259345441007085,
+      "grad_norm": 0.16431131958961487,
+      "learning_rate": 1.874131540423787e-05,
+      "loss": 1.2073,
+      "step": 5173
+    },
+    {
+      "epoch": 1.9263068492513176,
+      "grad_norm": 0.16539840400218964,
+      "learning_rate": 1.874072507361623e-05,
+      "loss": 1.1931,
+      "step": 5174
+    },
+    {
+      "epoch": 1.9266791544019268,
+      "grad_norm": 0.1654292792081833,
+      "learning_rate": 1.8740134613894633e-05,
+      "loss": 1.2022,
+      "step": 5175
+    },
+    {
+      "epoch": 1.9270514595525357,
+      "grad_norm": 0.16364841163158417,
+      "learning_rate": 1.8739544025081797e-05,
+      "loss": 1.1803,
+      "step": 5176
+    },
+    {
+      "epoch": 1.927423764703145,
+      "grad_norm": 0.16501003503799438,
+      "learning_rate": 1.8738953307186442e-05,
+      "loss": 1.1918,
+      "step": 5177
+    },
+    {
+      "epoch": 1.9277960698537537,
+      "grad_norm": 0.15340885519981384,
+      "learning_rate": 1.8738362460217296e-05,
+      "loss": 1.1955,
+      "step": 5178
+    },
+    {
+      "epoch": 1.928168375004363,
+      "grad_norm": 0.16795547306537628,
+      "learning_rate": 1.8737771484183084e-05,
+      "loss": 1.2006,
+      "step": 5179
+    },
+    {
+      "epoch": 1.928540680154972,
+      "grad_norm": 0.16092796623706818,
+      "learning_rate": 1.8737180379092536e-05,
+      "loss": 1.192,
+      "step": 5180
+    },
+    {
+      "epoch": 1.928912985305581,
+      "grad_norm": 0.1616433709859848,
+      "learning_rate": 1.8736589144954384e-05,
+      "loss": 1.1703,
+      "step": 5181
+    },
+    {
+      "epoch": 1.9292852904561901,
+      "grad_norm": 0.16441045701503754,
+      "learning_rate": 1.8735997781777356e-05,
+      "loss": 1.1818,
+      "step": 5182
+    },
+    {
+      "epoch": 1.9296575956067992,
+      "grad_norm": 0.16628822684288025,
+      "learning_rate": 1.8735406289570193e-05,
+      "loss": 1.1943,
+      "step": 5183
+    },
+    {
+      "epoch": 1.9300299007574084,
+      "grad_norm": 0.15827049314975739,
+      "learning_rate": 1.873481466834162e-05,
+      "loss": 1.1896,
+      "step": 5184
+    },
+    {
+      "epoch": 1.9304022059080173,
+      "grad_norm": 0.16958533227443695,
+      "learning_rate": 1.873422291810039e-05,
+      "loss": 1.1987,
+      "step": 5185
+    },
+    {
+      "epoch": 1.9307745110586265,
+      "grad_norm": 0.1640082448720932,
+      "learning_rate": 1.8733631038855232e-05,
+      "loss": 1.2068,
+      "step": 5186
+    },
+    {
+      "epoch": 1.9311468162092353,
+      "grad_norm": 0.16046150028705597,
+      "learning_rate": 1.8733039030614893e-05,
+      "loss": 1.1939,
+      "step": 5187
+    },
+    {
+      "epoch": 1.9315191213598446,
+      "grad_norm": 0.15839648246765137,
+      "learning_rate": 1.8732446893388116e-05,
+      "loss": 1.1919,
+      "step": 5188
+    },
+    {
+      "epoch": 1.9318914265104536,
+      "grad_norm": 0.15910884737968445,
+      "learning_rate": 1.8731854627183642e-05,
+      "loss": 1.2,
+      "step": 5189
+    },
+    {
+      "epoch": 1.9322637316610627,
+      "grad_norm": 0.17085225880146027,
+      "learning_rate": 1.8731262232010226e-05,
+      "loss": 1.1909,
+      "step": 5190
+    },
+    {
+      "epoch": 1.9326360368116717,
+      "grad_norm": 0.16858842968940735,
+      "learning_rate": 1.8730669707876617e-05,
+      "loss": 1.2053,
+      "step": 5191
+    },
+    {
+      "epoch": 1.9330083419622808,
+      "grad_norm": 0.16198118031024933,
+      "learning_rate": 1.8730077054791562e-05,
+      "loss": 1.1851,
+      "step": 5192
+    },
+    {
+      "epoch": 1.93338064711289,
+      "grad_norm": 0.16233724355697632,
+      "learning_rate": 1.8729484272763817e-05,
+      "loss": 1.1811,
+      "step": 5193
+    },
+    {
+      "epoch": 1.9337529522634989,
+      "grad_norm": 0.1625370979309082,
+      "learning_rate": 1.8728891361802136e-05,
+      "loss": 1.1946,
+      "step": 5194
+    },
+    {
+      "epoch": 1.9341252574141081,
+      "grad_norm": 0.16381089389324188,
+      "learning_rate": 1.872829832191528e-05,
+      "loss": 1.1884,
+      "step": 5195
+    },
+    {
+      "epoch": 1.934497562564717,
+      "grad_norm": 0.15602105855941772,
+      "learning_rate": 1.8727705153112007e-05,
+      "loss": 1.192,
+      "step": 5196
+    },
+    {
+      "epoch": 1.9348698677153262,
+      "grad_norm": 0.1663997620344162,
+      "learning_rate": 1.8727111855401073e-05,
+      "loss": 1.2075,
+      "step": 5197
+    },
+    {
+      "epoch": 1.9352421728659353,
+      "grad_norm": 0.16173984110355377,
+      "learning_rate": 1.8726518428791246e-05,
+      "loss": 1.1854,
+      "step": 5198
+    },
+    {
+      "epoch": 1.9356144780165443,
+      "grad_norm": 0.16155578196048737,
+      "learning_rate": 1.872592487329129e-05,
+      "loss": 1.1947,
+      "step": 5199
+    },
+    {
+      "epoch": 1.9359867831671533,
+      "grad_norm": 0.1606961041688919,
+      "learning_rate": 1.872533118890997e-05,
+      "loss": 1.1849,
+      "step": 5200
+    },
+    {
+      "epoch": 1.9363590883177624,
+      "grad_norm": 0.15817643702030182,
+      "learning_rate": 1.8724737375656054e-05,
+      "loss": 1.1949,
+      "step": 5201
+    },
+    {
+      "epoch": 1.9367313934683716,
+      "grad_norm": 0.16562969982624054,
+      "learning_rate": 1.8724143433538317e-05,
+      "loss": 1.1889,
+      "step": 5202
+    },
+    {
+      "epoch": 1.9371036986189805,
+      "grad_norm": 0.1643010526895523,
+      "learning_rate": 1.872354936256553e-05,
+      "loss": 1.1853,
+      "step": 5203
+    },
+    {
+      "epoch": 1.9374760037695897,
+      "grad_norm": 0.16269844770431519,
+      "learning_rate": 1.8722955162746465e-05,
+      "loss": 1.207,
+      "step": 5204
+    },
+    {
+      "epoch": 1.9378483089201988,
+      "grad_norm": 0.15886783599853516,
+      "learning_rate": 1.87223608340899e-05,
+      "loss": 1.1989,
+      "step": 5205
+    },
+    {
+      "epoch": 1.9382206140708078,
+      "grad_norm": 0.1662435382604599,
+      "learning_rate": 1.8721766376604612e-05,
+      "loss": 1.1928,
+      "step": 5206
+    },
+    {
+      "epoch": 1.9385929192214169,
+      "grad_norm": 0.16843284666538239,
+      "learning_rate": 1.8721171790299385e-05,
+      "loss": 1.1914,
+      "step": 5207
+    },
+    {
+      "epoch": 1.938965224372026,
+      "grad_norm": 0.16134469211101532,
+      "learning_rate": 1.8720577075182992e-05,
+      "loss": 1.1829,
+      "step": 5208
+    },
+    {
+      "epoch": 1.939337529522635,
+      "grad_norm": 0.16114822030067444,
+      "learning_rate": 1.8719982231264228e-05,
+      "loss": 1.2182,
+      "step": 5209
+    },
+    {
+      "epoch": 1.939709834673244,
+      "grad_norm": 0.15821397304534912,
+      "learning_rate": 1.8719387258551874e-05,
+      "loss": 1.1835,
+      "step": 5210
+    },
+    {
+      "epoch": 1.9400821398238532,
+      "grad_norm": 0.16827209293842316,
+      "learning_rate": 1.8718792157054714e-05,
+      "loss": 1.1927,
+      "step": 5211
+    },
+    {
+      "epoch": 1.940454444974462,
+      "grad_norm": 0.16307982802391052,
+      "learning_rate": 1.871819692678154e-05,
+      "loss": 1.1895,
+      "step": 5212
+    },
+    {
+      "epoch": 1.9408267501250713,
+      "grad_norm": 0.16273893415927887,
+      "learning_rate": 1.8717601567741147e-05,
+      "loss": 1.2002,
+      "step": 5213
+    },
+    {
+      "epoch": 1.9411990552756804,
+      "grad_norm": 0.16210372745990753,
+      "learning_rate": 1.871700607994233e-05,
+      "loss": 1.2093,
+      "step": 5214
+    },
+    {
+      "epoch": 1.9415713604262894,
+      "grad_norm": 0.16374051570892334,
+      "learning_rate": 1.8716410463393873e-05,
+      "loss": 1.2023,
+      "step": 5215
+    },
+    {
+      "epoch": 1.9419436655768985,
+      "grad_norm": 0.16489435732364655,
+      "learning_rate": 1.8715814718104585e-05,
+      "loss": 1.201,
+      "step": 5216
+    },
+    {
+      "epoch": 1.9423159707275075,
+      "grad_norm": 0.15813589096069336,
+      "learning_rate": 1.871521884408326e-05,
+      "loss": 1.1857,
+      "step": 5217
+    },
+    {
+      "epoch": 1.9426882758781168,
+      "grad_norm": 0.16261295974254608,
+      "learning_rate": 1.8714622841338696e-05,
+      "loss": 1.1985,
+      "step": 5218
+    },
+    {
+      "epoch": 1.9430605810287256,
+      "grad_norm": 0.15933023393154144,
+      "learning_rate": 1.8714026709879704e-05,
+      "loss": 1.2005,
+      "step": 5219
+    },
+    {
+      "epoch": 1.9434328861793349,
+      "grad_norm": 0.1632496863603592,
+      "learning_rate": 1.8713430449715086e-05,
+      "loss": 1.1946,
+      "step": 5220
+    },
+    {
+      "epoch": 1.9438051913299437,
+      "grad_norm": 0.16615080833435059,
+      "learning_rate": 1.8712834060853643e-05,
+      "loss": 1.2015,
+      "step": 5221
+    },
+    {
+      "epoch": 1.944177496480553,
+      "grad_norm": 0.1673406958580017,
+      "learning_rate": 1.871223754330419e-05,
+      "loss": 1.1935,
+      "step": 5222
+    },
+    {
+      "epoch": 1.944549801631162,
+      "grad_norm": 0.17133355140686035,
+      "learning_rate": 1.871164089707553e-05,
+      "loss": 1.198,
+      "step": 5223
+    },
+    {
+      "epoch": 1.944922106781771,
+      "grad_norm": 0.16196174919605255,
+      "learning_rate": 1.8711044122176484e-05,
+      "loss": 1.1964,
+      "step": 5224
+    },
+    {
+      "epoch": 1.94529441193238,
+      "grad_norm": 0.16071192920207977,
+      "learning_rate": 1.8710447218615865e-05,
+      "loss": 1.2009,
+      "step": 5225
+    },
+    {
+      "epoch": 1.945666717082989,
+      "grad_norm": 0.17128324508666992,
+      "learning_rate": 1.8709850186402487e-05,
+      "loss": 1.193,
+      "step": 5226
+    },
+    {
+      "epoch": 1.9460390222335984,
+      "grad_norm": 0.16837619245052338,
+      "learning_rate": 1.8709253025545167e-05,
+      "loss": 1.1982,
+      "step": 5227
+    },
+    {
+      "epoch": 1.9464113273842072,
+      "grad_norm": 0.16107183694839478,
+      "learning_rate": 1.8708655736052725e-05,
+      "loss": 1.1822,
+      "step": 5228
+    },
+    {
+      "epoch": 1.9467836325348165,
+      "grad_norm": 0.16632641851902008,
+      "learning_rate": 1.8708058317933986e-05,
+      "loss": 1.1895,
+      "step": 5229
+    },
+    {
+      "epoch": 1.9471559376854253,
+      "grad_norm": 0.16798974573612213,
+      "learning_rate": 1.8707460771197773e-05,
+      "loss": 1.2121,
+      "step": 5230
+    },
+    {
+      "epoch": 1.9475282428360345,
+      "grad_norm": 0.16306382417678833,
+      "learning_rate": 1.870686309585291e-05,
+      "loss": 1.1866,
+      "step": 5231
+    },
+    {
+      "epoch": 1.9479005479866436,
+      "grad_norm": 0.1670602411031723,
+      "learning_rate": 1.8706265291908226e-05,
+      "loss": 1.1971,
+      "step": 5232
+    },
+    {
+      "epoch": 1.9482728531372526,
+      "grad_norm": 0.1586066037416458,
+      "learning_rate": 1.870566735937255e-05,
+      "loss": 1.1894,
+      "step": 5233
+    },
+    {
+      "epoch": 1.9486451582878617,
+      "grad_norm": 0.16661204397678375,
+      "learning_rate": 1.870506929825471e-05,
+      "loss": 1.2033,
+      "step": 5234
+    },
+    {
+      "epoch": 1.9490174634384707,
+      "grad_norm": 0.16679902374744415,
+      "learning_rate": 1.870447110856355e-05,
+      "loss": 1.2003,
+      "step": 5235
+    },
+    {
+      "epoch": 1.94938976858908,
+      "grad_norm": 0.1620539128780365,
+      "learning_rate": 1.8703872790307892e-05,
+      "loss": 1.1742,
+      "step": 5236
+    },
+    {
+      "epoch": 1.9497620737396888,
+      "grad_norm": 0.16605715453624725,
+      "learning_rate": 1.870327434349658e-05,
+      "loss": 1.1964,
+      "step": 5237
+    },
+    {
+      "epoch": 1.950134378890298,
+      "grad_norm": 0.1599631905555725,
+      "learning_rate": 1.8702675768138453e-05,
+      "loss": 1.1964,
+      "step": 5238
+    },
+    {
+      "epoch": 1.9505066840409069,
+      "grad_norm": 0.1627977043390274,
+      "learning_rate": 1.870207706424235e-05,
+      "loss": 1.1941,
+      "step": 5239
+    },
+    {
+      "epoch": 1.9508789891915161,
+      "grad_norm": 0.16092990338802338,
+      "learning_rate": 1.8701478231817116e-05,
+      "loss": 1.2014,
+      "step": 5240
+    },
+    {
+      "epoch": 1.9512512943421252,
+      "grad_norm": 0.16339315474033356,
+      "learning_rate": 1.8700879270871594e-05,
+      "loss": 1.1798,
+      "step": 5241
+    },
+    {
+      "epoch": 1.9516235994927342,
+      "grad_norm": 0.16573093831539154,
+      "learning_rate": 1.8700280181414633e-05,
+      "loss": 1.1786,
+      "step": 5242
+    },
+    {
+      "epoch": 1.9519959046433433,
+      "grad_norm": 0.16669802367687225,
+      "learning_rate": 1.8699680963455076e-05,
+      "loss": 1.202,
+      "step": 5243
+    },
+    {
+      "epoch": 1.9523682097939523,
+      "grad_norm": 0.16122578084468842,
+      "learning_rate": 1.8699081617001784e-05,
+      "loss": 1.1927,
+      "step": 5244
+    },
+    {
+      "epoch": 1.9527405149445616,
+      "grad_norm": 0.1680309772491455,
+      "learning_rate": 1.8698482142063595e-05,
+      "loss": 1.1934,
+      "step": 5245
+    },
+    {
+      "epoch": 1.9531128200951704,
+      "grad_norm": 0.16212962567806244,
+      "learning_rate": 1.8697882538649373e-05,
+      "loss": 1.1842,
+      "step": 5246
+    },
+    {
+      "epoch": 1.9534851252457797,
+      "grad_norm": 0.16280242800712585,
+      "learning_rate": 1.8697282806767974e-05,
+      "loss": 1.2018,
+      "step": 5247
+    },
+    {
+      "epoch": 1.9538574303963885,
+      "grad_norm": 0.15737560391426086,
+      "learning_rate": 1.8696682946428253e-05,
+      "loss": 1.2001,
+      "step": 5248
+    },
+    {
+      "epoch": 1.9542297355469977,
+      "grad_norm": 0.1623457968235016,
+      "learning_rate": 1.8696082957639068e-05,
+      "loss": 1.1931,
+      "step": 5249
+    },
+    {
+      "epoch": 1.9546020406976068,
+      "grad_norm": 0.15832823514938354,
+      "learning_rate": 1.8695482840409287e-05,
+      "loss": 1.2063,
+      "step": 5250
+    },
+    {
+      "epoch": 1.9549743458482158,
+      "grad_norm": 0.15807972848415375,
+      "learning_rate": 1.869488259474777e-05,
+      "loss": 1.1896,
+      "step": 5251
+    },
+    {
+      "epoch": 1.9553466509988249,
+      "grad_norm": 0.16206707060337067,
+      "learning_rate": 1.869428222066338e-05,
+      "loss": 1.2002,
+      "step": 5252
+    },
+    {
+      "epoch": 1.955718956149434,
+      "grad_norm": 0.16285400092601776,
+      "learning_rate": 1.8693681718164987e-05,
+      "loss": 1.2033,
+      "step": 5253
+    },
+    {
+      "epoch": 1.9560912613000432,
+      "grad_norm": 0.1658763289451599,
+      "learning_rate": 1.8693081087261463e-05,
+      "loss": 1.1955,
+      "step": 5254
+    },
+    {
+      "epoch": 1.956463566450652,
+      "grad_norm": 0.15834468603134155,
+      "learning_rate": 1.8692480327961673e-05,
+      "loss": 1.1961,
+      "step": 5255
+    },
+    {
+      "epoch": 1.9568358716012613,
+      "grad_norm": 0.15797977149486542,
+      "learning_rate": 1.8691879440274498e-05,
+      "loss": 1.1883,
+      "step": 5256
+    },
+    {
+      "epoch": 1.95720817675187,
+      "grad_norm": 0.16849857568740845,
+      "learning_rate": 1.869127842420881e-05,
+      "loss": 1.1909,
+      "step": 5257
+    },
+    {
+      "epoch": 1.9575804819024794,
+      "grad_norm": 0.16109803318977356,
+      "learning_rate": 1.8690677279773482e-05,
+      "loss": 1.1815,
+      "step": 5258
+    },
+    {
+      "epoch": 1.9579527870530884,
+      "grad_norm": 0.15917837619781494,
+      "learning_rate": 1.86900760069774e-05,
+      "loss": 1.2017,
+      "step": 5259
+    },
+    {
+      "epoch": 1.9583250922036974,
+      "grad_norm": 0.1662922352552414,
+      "learning_rate": 1.8689474605829436e-05,
+      "loss": 1.2067,
+      "step": 5260
+    },
+    {
+      "epoch": 1.9586973973543065,
+      "grad_norm": 0.17103338241577148,
+      "learning_rate": 1.868887307633848e-05,
+      "loss": 1.196,
+      "step": 5261
+    },
+    {
+      "epoch": 1.9590697025049155,
+      "grad_norm": 0.1583465188741684,
+      "learning_rate": 1.8688271418513415e-05,
+      "loss": 1.1932,
+      "step": 5262
+    },
+    {
+      "epoch": 1.9594420076555248,
+      "grad_norm": 0.1627221256494522,
+      "learning_rate": 1.8687669632363122e-05,
+      "loss": 1.1863,
+      "step": 5263
+    },
+    {
+      "epoch": 1.9598143128061336,
+      "grad_norm": 0.1791868507862091,
+      "learning_rate": 1.86870677178965e-05,
+      "loss": 1.2089,
+      "step": 5264
+    },
+    {
+      "epoch": 1.9601866179567429,
+      "grad_norm": 0.17789161205291748,
+      "learning_rate": 1.8686465675122426e-05,
+      "loss": 1.1899,
+      "step": 5265
+    },
+    {
+      "epoch": 1.960558923107352,
+      "grad_norm": 0.16588279604911804,
+      "learning_rate": 1.8685863504049804e-05,
+      "loss": 1.1984,
+      "step": 5266
+    },
+    {
+      "epoch": 1.960931228257961,
+      "grad_norm": 0.16590848565101624,
+      "learning_rate": 1.8685261204687523e-05,
+      "loss": 1.2052,
+      "step": 5267
+    },
+    {
+      "epoch": 1.96130353340857,
+      "grad_norm": 0.22419647872447968,
+      "learning_rate": 1.8684658777044478e-05,
+      "loss": 1.1945,
+      "step": 5268
+    },
+    {
+      "epoch": 1.961675838559179,
+      "grad_norm": 0.20315192639827728,
+      "learning_rate": 1.8684056221129568e-05,
+      "loss": 1.2025,
+      "step": 5269
+    },
+    {
+      "epoch": 1.962048143709788,
+      "grad_norm": 0.17711172997951508,
+      "learning_rate": 1.868345353695169e-05,
+      "loss": 1.1777,
+      "step": 5270
+    },
+    {
+      "epoch": 1.9624204488603971,
+      "grad_norm": 0.16664499044418335,
+      "learning_rate": 1.8682850724519752e-05,
+      "loss": 1.1837,
+      "step": 5271
+    },
+    {
+      "epoch": 1.9627927540110064,
+      "grad_norm": 0.1918790340423584,
+      "learning_rate": 1.8682247783842654e-05,
+      "loss": 1.1987,
+      "step": 5272
+    },
+    {
+      "epoch": 1.9631650591616152,
+      "grad_norm": 0.17362017929553986,
+      "learning_rate": 1.86816447149293e-05,
+      "loss": 1.2091,
+      "step": 5273
+    },
+    {
+      "epoch": 1.9635373643122245,
+      "grad_norm": 0.1695503294467926,
+      "learning_rate": 1.8681041517788596e-05,
+      "loss": 1.1945,
+      "step": 5274
+    },
+    {
+      "epoch": 1.9639096694628335,
+      "grad_norm": 0.17335741221904755,
+      "learning_rate": 1.8680438192429455e-05,
+      "loss": 1.185,
+      "step": 5275
+    },
+    {
+      "epoch": 1.9642819746134426,
+      "grad_norm": 0.17371761798858643,
+      "learning_rate": 1.867983473886079e-05,
+      "loss": 1.1801,
+      "step": 5276
+    },
+    {
+      "epoch": 1.9646542797640516,
+      "grad_norm": 0.1588549017906189,
+      "learning_rate": 1.8679231157091507e-05,
+      "loss": 1.1924,
+      "step": 5277
+    },
+    {
+      "epoch": 1.9650265849146606,
+      "grad_norm": 0.16421490907669067,
+      "learning_rate": 1.8678627447130524e-05,
+      "loss": 1.1931,
+      "step": 5278
+    },
+    {
+      "epoch": 1.96539889006527,
+      "grad_norm": 0.17214372754096985,
+      "learning_rate": 1.8678023608986758e-05,
+      "loss": 1.1904,
+      "step": 5279
+    },
+    {
+      "epoch": 1.9657711952158787,
+      "grad_norm": 0.18433736264705658,
+      "learning_rate": 1.867741964266913e-05,
+      "loss": 1.1974,
+      "step": 5280
+    },
+    {
+      "epoch": 1.966143500366488,
+      "grad_norm": 0.15704011917114258,
+      "learning_rate": 1.867681554818656e-05,
+      "loss": 1.1743,
+      "step": 5281
+    },
+    {
+      "epoch": 1.9665158055170968,
+      "grad_norm": 0.16321733593940735,
+      "learning_rate": 1.8676211325547965e-05,
+      "loss": 1.204,
+      "step": 5282
+    },
+    {
+      "epoch": 1.966888110667706,
+      "grad_norm": 0.17481832206249237,
+      "learning_rate": 1.8675606974762274e-05,
+      "loss": 1.1891,
+      "step": 5283
+    },
+    {
+      "epoch": 1.9672604158183151,
+      "grad_norm": 0.16488084197044373,
+      "learning_rate": 1.8675002495838415e-05,
+      "loss": 1.2052,
+      "step": 5284
+    },
+    {
+      "epoch": 1.9676327209689242,
+      "grad_norm": 0.15766631066799164,
+      "learning_rate": 1.8674397888785312e-05,
+      "loss": 1.1902,
+      "step": 5285
+    },
+    {
+      "epoch": 1.9680050261195332,
+      "grad_norm": 0.17218059301376343,
+      "learning_rate": 1.8673793153611893e-05,
+      "loss": 1.1857,
+      "step": 5286
+    },
+    {
+      "epoch": 1.9683773312701422,
+      "grad_norm": 0.15942028164863586,
+      "learning_rate": 1.86731882903271e-05,
+      "loss": 1.1842,
+      "step": 5287
+    },
+    {
+      "epoch": 1.9687496364207515,
+      "grad_norm": 0.16559404134750366,
+      "learning_rate": 1.8672583298939857e-05,
+      "loss": 1.1907,
+      "step": 5288
+    },
+    {
+      "epoch": 1.9691219415713603,
+      "grad_norm": 0.16306178271770477,
+      "learning_rate": 1.8671978179459102e-05,
+      "loss": 1.1881,
+      "step": 5289
+    },
+    {
+      "epoch": 1.9694942467219696,
+      "grad_norm": 0.16512851417064667,
+      "learning_rate": 1.8671372931893775e-05,
+      "loss": 1.2026,
+      "step": 5290
+    },
+    {
+      "epoch": 1.9698665518725784,
+      "grad_norm": 0.1704392433166504,
+      "learning_rate": 1.8670767556252812e-05,
+      "loss": 1.207,
+      "step": 5291
+    },
+    {
+      "epoch": 1.9702388570231877,
+      "grad_norm": 0.160672128200531,
+      "learning_rate": 1.867016205254516e-05,
+      "loss": 1.196,
+      "step": 5292
+    },
+    {
+      "epoch": 1.9706111621737967,
+      "grad_norm": 0.16421304643154144,
+      "learning_rate": 1.866955642077976e-05,
+      "loss": 1.1896,
+      "step": 5293
+    },
+    {
+      "epoch": 1.9709834673244058,
+      "grad_norm": 0.1669030487537384,
+      "learning_rate": 1.866895066096555e-05,
+      "loss": 1.1982,
+      "step": 5294
+    },
+    {
+      "epoch": 1.9713557724750148,
+      "grad_norm": 0.171576127409935,
+      "learning_rate": 1.8668344773111483e-05,
+      "loss": 1.1997,
+      "step": 5295
+    },
+    {
+      "epoch": 1.9717280776256239,
+      "grad_norm": 0.16728414595127106,
+      "learning_rate": 1.866773875722651e-05,
+      "loss": 1.1875,
+      "step": 5296
+    },
+    {
+      "epoch": 1.9721003827762331,
+      "grad_norm": 0.15906740725040436,
+      "learning_rate": 1.866713261331958e-05,
+      "loss": 1.1797,
+      "step": 5297
+    },
+    {
+      "epoch": 1.972472687926842,
+      "grad_norm": 0.16088546812534332,
+      "learning_rate": 1.8666526341399644e-05,
+      "loss": 1.1836,
+      "step": 5298
+    },
+    {
+      "epoch": 1.9728449930774512,
+      "grad_norm": 0.16941684484481812,
+      "learning_rate": 1.866591994147566e-05,
+      "loss": 1.1921,
+      "step": 5299
+    },
+    {
+      "epoch": 1.97321729822806,
+      "grad_norm": 0.16476041078567505,
+      "learning_rate": 1.866531341355658e-05,
+      "loss": 1.1841,
+      "step": 5300
+    },
+    {
+      "epoch": 1.9735896033786693,
+      "grad_norm": 0.15881778299808502,
+      "learning_rate": 1.8664706757651365e-05,
+      "loss": 1.1853,
+      "step": 5301
+    },
+    {
+      "epoch": 1.9739619085292783,
+      "grad_norm": 0.164139986038208,
+      "learning_rate": 1.8664099973768975e-05,
+      "loss": 1.1816,
+      "step": 5302
+    },
+    {
+      "epoch": 1.9743342136798874,
+      "grad_norm": 0.16207018494606018,
+      "learning_rate": 1.8663493061918375e-05,
+      "loss": 1.1838,
+      "step": 5303
+    },
+    {
+      "epoch": 1.9747065188304964,
+      "grad_norm": 0.17509856820106506,
+      "learning_rate": 1.8662886022108524e-05,
+      "loss": 1.2028,
+      "step": 5304
+    },
+    {
+      "epoch": 1.9750788239811055,
+      "grad_norm": 0.17713604867458344,
+      "learning_rate": 1.8662278854348387e-05,
+      "loss": 1.2028,
+      "step": 5305
+    },
+    {
+      "epoch": 1.9754511291317147,
+      "grad_norm": 0.17039844393730164,
+      "learning_rate": 1.8661671558646938e-05,
+      "loss": 1.1883,
+      "step": 5306
+    },
+    {
+      "epoch": 1.9758234342823235,
+      "grad_norm": 0.17008256912231445,
+      "learning_rate": 1.8661064135013143e-05,
+      "loss": 1.1971,
+      "step": 5307
+    },
+    {
+      "epoch": 1.9761957394329328,
+      "grad_norm": 0.20797152817249298,
+      "learning_rate": 1.8660456583455974e-05,
+      "loss": 1.1907,
+      "step": 5308
+    },
+    {
+      "epoch": 1.9765680445835416,
+      "grad_norm": 0.20216019451618195,
+      "learning_rate": 1.865984890398441e-05,
+      "loss": 1.2112,
+      "step": 5309
+    },
+    {
+      "epoch": 1.976940349734151,
+      "grad_norm": 0.18128228187561035,
+      "learning_rate": 1.8659241096607416e-05,
+      "loss": 1.1903,
+      "step": 5310
+    },
+    {
+      "epoch": 1.97731265488476,
+      "grad_norm": 0.1619497686624527,
+      "learning_rate": 1.8658633161333974e-05,
+      "loss": 1.2063,
+      "step": 5311
+    },
+    {
+      "epoch": 1.977684960035369,
+      "grad_norm": 0.1790294498205185,
+      "learning_rate": 1.865802509817306e-05,
+      "loss": 1.197,
+      "step": 5312
+    },
+    {
+      "epoch": 1.978057265185978,
+      "grad_norm": 0.18129126727581024,
+      "learning_rate": 1.8657416907133668e-05,
+      "loss": 1.1738,
+      "step": 5313
+    },
+    {
+      "epoch": 1.978429570336587,
+      "grad_norm": 0.16873174905776978,
+      "learning_rate": 1.8656808588224767e-05,
+      "loss": 1.1904,
+      "step": 5314
+    },
+    {
+      "epoch": 1.9788018754871963,
+      "grad_norm": 0.1654161512851715,
+      "learning_rate": 1.8656200141455345e-05,
+      "loss": 1.1814,
+      "step": 5315
+    },
+    {
+      "epoch": 1.9791741806378051,
+      "grad_norm": 0.1696605384349823,
+      "learning_rate": 1.865559156683439e-05,
+      "loss": 1.1972,
+      "step": 5316
+    },
+    {
+      "epoch": 1.9795464857884144,
+      "grad_norm": 0.17438465356826782,
+      "learning_rate": 1.8654982864370893e-05,
+      "loss": 1.2101,
+      "step": 5317
+    },
+    {
+      "epoch": 1.9799187909390232,
+      "grad_norm": 0.16162490844726562,
+      "learning_rate": 1.8654374034073843e-05,
+      "loss": 1.1928,
+      "step": 5318
+    },
+    {
+      "epoch": 1.9802910960896325,
+      "grad_norm": 0.16213704645633698,
+      "learning_rate": 1.865376507595223e-05,
+      "loss": 1.1729,
+      "step": 5319
+    },
+    {
+      "epoch": 1.9806634012402415,
+      "grad_norm": 0.19320641458034515,
+      "learning_rate": 1.865315599001505e-05,
+      "loss": 1.1971,
+      "step": 5320
+    },
+    {
+      "epoch": 1.9810357063908506,
+      "grad_norm": 0.16562476754188538,
+      "learning_rate": 1.8652546776271297e-05,
+      "loss": 1.1873,
+      "step": 5321
+    },
+    {
+      "epoch": 1.9814080115414596,
+      "grad_norm": 0.16413205862045288,
+      "learning_rate": 1.8651937434729975e-05,
+      "loss": 1.1935,
+      "step": 5322
+    },
+    {
+      "epoch": 1.9817803166920687,
+      "grad_norm": 0.17290954291820526,
+      "learning_rate": 1.865132796540008e-05,
+      "loss": 1.1915,
+      "step": 5323
+    },
+    {
+      "epoch": 1.982152621842678,
+      "grad_norm": 0.17080329358577728,
+      "learning_rate": 1.8650718368290613e-05,
+      "loss": 1.1958,
+      "step": 5324
+    },
+    {
+      "epoch": 1.9825249269932868,
+      "grad_norm": 0.1644285023212433,
+      "learning_rate": 1.865010864341058e-05,
+      "loss": 1.1918,
+      "step": 5325
+    },
+    {
+      "epoch": 1.982897232143896,
+      "grad_norm": 0.1660122573375702,
+      "learning_rate": 1.864949879076898e-05,
+      "loss": 1.1808,
+      "step": 5326
+    },
+    {
+      "epoch": 1.983269537294505,
+      "grad_norm": 0.16145946085453033,
+      "learning_rate": 1.864888881037483e-05,
+      "loss": 1.2083,
+      "step": 5327
+    },
+    {
+      "epoch": 1.983641842445114,
+      "grad_norm": 0.16719859838485718,
+      "learning_rate": 1.8648278702237133e-05,
+      "loss": 1.1856,
+      "step": 5328
+    },
+    {
+      "epoch": 1.9840141475957231,
+      "grad_norm": 0.16676342487335205,
+      "learning_rate": 1.8647668466364903e-05,
+      "loss": 1.2016,
+      "step": 5329
+    },
+    {
+      "epoch": 1.9843864527463322,
+      "grad_norm": 0.16827064752578735,
+      "learning_rate": 1.864705810276715e-05,
+      "loss": 1.1768,
+      "step": 5330
+    },
+    {
+      "epoch": 1.9847587578969412,
+      "grad_norm": 0.1711677461862564,
+      "learning_rate": 1.8646447611452895e-05,
+      "loss": 1.1931,
+      "step": 5331
+    },
+    {
+      "epoch": 1.9851310630475503,
+      "grad_norm": 0.16668665409088135,
+      "learning_rate": 1.8645836992431152e-05,
+      "loss": 1.2098,
+      "step": 5332
+    },
+    {
+      "epoch": 1.9855033681981595,
+      "grad_norm": 0.1718050241470337,
+      "learning_rate": 1.8645226245710936e-05,
+      "loss": 1.2019,
+      "step": 5333
+    },
+    {
+      "epoch": 1.9858756733487684,
+      "grad_norm": 0.1724565029144287,
+      "learning_rate": 1.8644615371301275e-05,
+      "loss": 1.1771,
+      "step": 5334
+    },
+    {
+      "epoch": 1.9862479784993776,
+      "grad_norm": 0.1688944697380066,
+      "learning_rate": 1.8644004369211183e-05,
+      "loss": 1.1902,
+      "step": 5335
+    },
+    {
+      "epoch": 1.9866202836499867,
+      "grad_norm": 0.17193703353405,
+      "learning_rate": 1.864339323944969e-05,
+      "loss": 1.1893,
+      "step": 5336
+    },
+    {
+      "epoch": 1.9869925888005957,
+      "grad_norm": 0.166243776679039,
+      "learning_rate": 1.8642781982025827e-05,
+      "loss": 1.203,
+      "step": 5337
+    },
+    {
+      "epoch": 1.9873648939512047,
+      "grad_norm": 0.1623961478471756,
+      "learning_rate": 1.864217059694861e-05,
+      "loss": 1.1703,
+      "step": 5338
+    },
+    {
+      "epoch": 1.9877371991018138,
+      "grad_norm": 0.1668027937412262,
+      "learning_rate": 1.8641559084227078e-05,
+      "loss": 1.1886,
+      "step": 5339
+    },
+    {
+      "epoch": 1.988109504252423,
+      "grad_norm": 0.16468320786952972,
+      "learning_rate": 1.864094744387026e-05,
+      "loss": 1.2047,
+      "step": 5340
+    },
+    {
+      "epoch": 1.9884818094030319,
+      "grad_norm": 0.16658997535705566,
+      "learning_rate": 1.864033567588719e-05,
+      "loss": 1.1944,
+      "step": 5341
+    },
+    {
+      "epoch": 1.9888541145536411,
+      "grad_norm": 0.16415011882781982,
+      "learning_rate": 1.8639723780286903e-05,
+      "loss": 1.1896,
+      "step": 5342
+    },
+    {
+      "epoch": 1.98922641970425,
+      "grad_norm": 0.16742385923862457,
+      "learning_rate": 1.8639111757078444e-05,
+      "loss": 1.2031,
+      "step": 5343
+    },
+    {
+      "epoch": 1.9895987248548592,
+      "grad_norm": 0.16341787576675415,
+      "learning_rate": 1.863849960627084e-05,
+      "loss": 1.1822,
+      "step": 5344
+    },
+    {
+      "epoch": 1.9899710300054683,
+      "grad_norm": 0.16611303389072418,
+      "learning_rate": 1.863788732787314e-05,
+      "loss": 1.1853,
+      "step": 5345
+    },
+    {
+      "epoch": 1.9903433351560773,
+      "grad_norm": 0.17013627290725708,
+      "learning_rate": 1.863727492189439e-05,
+      "loss": 1.1891,
+      "step": 5346
+    },
+    {
+      "epoch": 1.9907156403066864,
+      "grad_norm": 0.17224200069904327,
+      "learning_rate": 1.863666238834363e-05,
+      "loss": 1.1946,
+      "step": 5347
+    },
+    {
+      "epoch": 1.9910879454572954,
+      "grad_norm": 0.1723671704530716,
+      "learning_rate": 1.8636049727229908e-05,
+      "loss": 1.1886,
+      "step": 5348
+    },
+    {
+      "epoch": 1.9914602506079047,
+      "grad_norm": 0.1694638729095459,
+      "learning_rate": 1.8635436938562273e-05,
+      "loss": 1.185,
+      "step": 5349
+    },
+    {
+      "epoch": 1.9918325557585135,
+      "grad_norm": 0.1634708046913147,
+      "learning_rate": 1.8634824022349773e-05,
+      "loss": 1.201,
+      "step": 5350
+    },
+    {
+      "epoch": 1.9922048609091227,
+      "grad_norm": 0.16697536408901215,
+      "learning_rate": 1.863421097860147e-05,
+      "loss": 1.1841,
+      "step": 5351
+    },
+    {
+      "epoch": 1.9925771660597316,
+      "grad_norm": 0.16353781521320343,
+      "learning_rate": 1.863359780732641e-05,
+      "loss": 1.189,
+      "step": 5352
+    },
+    {
+      "epoch": 1.9929494712103408,
+      "grad_norm": 0.17699691653251648,
+      "learning_rate": 1.8632984508533654e-05,
+      "loss": 1.1828,
+      "step": 5353
+    },
+    {
+      "epoch": 1.9933217763609499,
+      "grad_norm": 0.19866538047790527,
+      "learning_rate": 1.8632371082232254e-05,
+      "loss": 1.2084,
+      "step": 5354
+    },
+    {
+      "epoch": 1.993694081511559,
+      "grad_norm": 0.16737166047096252,
+      "learning_rate": 1.863175752843128e-05,
+      "loss": 1.204,
+      "step": 5355
+    },
+    {
+      "epoch": 1.994066386662168,
+      "grad_norm": 0.1831938922405243,
+      "learning_rate": 1.8631143847139785e-05,
+      "loss": 1.1924,
+      "step": 5356
+    },
+    {
+      "epoch": 1.994438691812777,
+      "grad_norm": 0.17462186515331268,
+      "learning_rate": 1.863053003836684e-05,
+      "loss": 1.2058,
+      "step": 5357
+    },
+    {
+      "epoch": 1.9948109969633863,
+      "grad_norm": 0.175477534532547,
+      "learning_rate": 1.8629916102121507e-05,
+      "loss": 1.1933,
+      "step": 5358
+    },
+    {
+      "epoch": 1.995183302113995,
+      "grad_norm": 0.22154709696769714,
+      "learning_rate": 1.8629302038412856e-05,
+      "loss": 1.1926,
+      "step": 5359
+    },
+    {
+      "epoch": 1.9955556072646043,
+      "grad_norm": 0.16697679460048676,
+      "learning_rate": 1.8628687847249955e-05,
+      "loss": 1.1954,
+      "step": 5360
+    },
+    {
+      "epoch": 1.9959279124152132,
+      "grad_norm": 0.17017914354801178,
+      "learning_rate": 1.8628073528641873e-05,
+      "loss": 1.203,
+      "step": 5361
+    },
+    {
+      "epoch": 1.9963002175658224,
+      "grad_norm": 0.1623312085866928,
+      "learning_rate": 1.862745908259769e-05,
+      "loss": 1.1899,
+      "step": 5362
+    },
+    {
+      "epoch": 1.9966725227164315,
+      "grad_norm": 0.16061881184577942,
+      "learning_rate": 1.8626844509126476e-05,
+      "loss": 1.1995,
+      "step": 5363
+    },
+    {
+      "epoch": 1.9970448278670405,
+      "grad_norm": 0.18128757178783417,
+      "learning_rate": 1.862622980823731e-05,
+      "loss": 1.2044,
+      "step": 5364
+    },
+    {
+      "epoch": 1.9974171330176496,
+      "grad_norm": 0.15951292216777802,
+      "learning_rate": 1.8625614979939273e-05,
+      "loss": 1.1834,
+      "step": 5365
+    },
+    {
+      "epoch": 1.9977894381682586,
+      "grad_norm": 0.15819518268108368,
+      "learning_rate": 1.8625000024241445e-05,
+      "loss": 1.1907,
+      "step": 5366
+    },
+    {
+      "epoch": 1.9981617433188679,
+      "grad_norm": 0.1662832796573639,
+      "learning_rate": 1.86243849411529e-05,
+      "loss": 1.1928,
+      "step": 5367
+    },
+    {
+      "epoch": 1.9985340484694767,
+      "grad_norm": 0.16362832486629486,
+      "learning_rate": 1.8623769730682738e-05,
+      "loss": 1.1923,
+      "step": 5368
+    },
+    {
+      "epoch": 1.998906353620086,
+      "grad_norm": 0.16274386644363403,
+      "learning_rate": 1.8623154392840036e-05,
+      "loss": 1.1892,
+      "step": 5369
+    },
+    {
+      "epoch": 1.9992786587706948,
+      "grad_norm": 0.2034921646118164,
+      "learning_rate": 1.8622538927633886e-05,
+      "loss": 1.1806,
+      "step": 5370
+    },
+    {
+      "epoch": 1.999650963921304,
+      "grad_norm": 0.1652044951915741,
+      "learning_rate": 1.8621923335073378e-05,
+      "loss": 1.2014,
+      "step": 5371
+    },
+    {
+      "epoch": 2.000023269071913,
+      "grad_norm": 0.16666065156459808,
+      "learning_rate": 1.86213076151676e-05,
+      "loss": 1.1979,
+      "step": 5372
+    },
+    {
+      "epoch": 2.000395574222522,
+      "grad_norm": 0.16674093902111053,
+      "learning_rate": 1.8620691767925655e-05,
+      "loss": 1.1733,
+      "step": 5373
+    },
+    {
+      "epoch": 2.0007678793731314,
+      "grad_norm": 0.16234448552131653,
+      "learning_rate": 1.862007579335663e-05,
+      "loss": 1.1845,
+      "step": 5374
+    },
+    {
+      "epoch": 2.00114018452374,
+      "grad_norm": 0.16867627203464508,
+      "learning_rate": 1.8619459691469625e-05,
+      "loss": 1.1914,
+      "step": 5375
+    },
+    {
+      "epoch": 2.0015124896743495,
+      "grad_norm": 0.17261743545532227,
+      "learning_rate": 1.8618843462273743e-05,
+      "loss": 1.1808,
+      "step": 5376
+    },
+    {
+      "epoch": 2.0018847948249583,
+      "grad_norm": 0.16891492903232574,
+      "learning_rate": 1.8618227105778086e-05,
+      "loss": 1.1928,
+      "step": 5377
+    },
+    {
+      "epoch": 2.0022570999755676,
+      "grad_norm": 0.17338630557060242,
+      "learning_rate": 1.8617610621991753e-05,
+      "loss": 1.181,
+      "step": 5378
+    },
+    {
+      "epoch": 2.0026294051261764,
+      "grad_norm": 0.15824103355407715,
+      "learning_rate": 1.861699401092385e-05,
+      "loss": 1.1842,
+      "step": 5379
+    },
+    {
+      "epoch": 2.0030017102767856,
+      "grad_norm": 0.19989395141601562,
+      "learning_rate": 1.861637727258349e-05,
+      "loss": 1.1842,
+      "step": 5380
+    },
+    {
+      "epoch": 2.0033740154273945,
+      "grad_norm": 0.1691095381975174,
+      "learning_rate": 1.8615760406979778e-05,
+      "loss": 1.1976,
+      "step": 5381
+    },
+    {
+      "epoch": 2.0037463205780037,
+      "grad_norm": 0.17154785990715027,
+      "learning_rate": 1.8615143414121823e-05,
+      "loss": 1.1795,
+      "step": 5382
+    },
+    {
+      "epoch": 2.004118625728613,
+      "grad_norm": 0.16465583443641663,
+      "learning_rate": 1.8614526294018743e-05,
+      "loss": 1.1805,
+      "step": 5383
+    },
+    {
+      "epoch": 2.004490930879222,
+      "grad_norm": 0.16546522080898285,
+      "learning_rate": 1.8613909046679646e-05,
+      "loss": 1.1785,
+      "step": 5384
+    },
+    {
+      "epoch": 2.004863236029831,
+      "grad_norm": 0.1838424950838089,
+      "learning_rate": 1.861329167211366e-05,
+      "loss": 1.1865,
+      "step": 5385
+    },
+    {
+      "epoch": 2.00523554118044,
+      "grad_norm": 0.16737298667430878,
+      "learning_rate": 1.861267417032989e-05,
+      "loss": 1.1896,
+      "step": 5386
+    },
+    {
+      "epoch": 2.005607846331049,
+      "grad_norm": 0.16740891337394714,
+      "learning_rate": 1.8612056541337466e-05,
+      "loss": 1.1955,
+      "step": 5387
+    },
+    {
+      "epoch": 2.005980151481658,
+      "grad_norm": 0.1717260479927063,
+      "learning_rate": 1.8611438785145508e-05,
+      "loss": 1.1835,
+      "step": 5388
+    },
+    {
+      "epoch": 2.0063524566322672,
+      "grad_norm": 0.16396482288837433,
+      "learning_rate": 1.8610820901763137e-05,
+      "loss": 1.1755,
+      "step": 5389
+    },
+    {
+      "epoch": 2.0067247617828765,
+      "grad_norm": 0.16121700406074524,
+      "learning_rate": 1.8610202891199484e-05,
+      "loss": 1.1849,
+      "step": 5390
+    },
+    {
+      "epoch": 2.0070970669334853,
+      "grad_norm": 0.18463963270187378,
+      "learning_rate": 1.8609584753463676e-05,
+      "loss": 1.1814,
+      "step": 5391
+    },
+    {
+      "epoch": 2.0074693720840946,
+      "grad_norm": 0.16385917365550995,
+      "learning_rate": 1.8608966488564838e-05,
+      "loss": 1.1919,
+      "step": 5392
+    },
+    {
+      "epoch": 2.0078416772347034,
+      "grad_norm": 0.1689513474702835,
+      "learning_rate": 1.8608348096512107e-05,
+      "loss": 1.2054,
+      "step": 5393
+    },
+    {
+      "epoch": 2.0082139823853127,
+      "grad_norm": 0.16833508014678955,
+      "learning_rate": 1.8607729577314615e-05,
+      "loss": 1.1865,
+      "step": 5394
+    },
+    {
+      "epoch": 2.0085862875359215,
+      "grad_norm": 0.16249039769172668,
+      "learning_rate": 1.8607110930981496e-05,
+      "loss": 1.1882,
+      "step": 5395
+    },
+    {
+      "epoch": 2.0089585926865308,
+      "grad_norm": 0.15944033861160278,
+      "learning_rate": 1.860649215752189e-05,
+      "loss": 1.1781,
+      "step": 5396
+    },
+    {
+      "epoch": 2.0093308978371396,
+      "grad_norm": 0.16826555132865906,
+      "learning_rate": 1.8605873256944934e-05,
+      "loss": 1.1748,
+      "step": 5397
+    },
+    {
+      "epoch": 2.009703202987749,
+      "grad_norm": 0.1765327900648117,
+      "learning_rate": 1.860525422925977e-05,
+      "loss": 1.1894,
+      "step": 5398
+    },
+    {
+      "epoch": 2.010075508138358,
+      "grad_norm": 0.1757960021495819,
+      "learning_rate": 1.8604635074475542e-05,
+      "loss": 1.1952,
+      "step": 5399
+    },
+    {
+      "epoch": 2.010447813288967,
+      "grad_norm": 0.17114783823490143,
+      "learning_rate": 1.8604015792601395e-05,
+      "loss": 1.1784,
+      "step": 5400
+    },
+    {
+      "epoch": 2.010820118439576,
+      "grad_norm": 0.16381992399692535,
+      "learning_rate": 1.860339638364647e-05,
+      "loss": 1.1807,
+      "step": 5401
+    },
+    {
+      "epoch": 2.011192423590185,
+      "grad_norm": 0.1801488697528839,
+      "learning_rate": 1.8602776847619926e-05,
+      "loss": 1.2051,
+      "step": 5402
+    },
+    {
+      "epoch": 2.0115647287407943,
+      "grad_norm": 0.17057554423809052,
+      "learning_rate": 1.8602157184530907e-05,
+      "loss": 1.1815,
+      "step": 5403
+    },
+    {
+      "epoch": 2.011937033891403,
+      "grad_norm": 0.175872340798378,
+      "learning_rate": 1.8601537394388565e-05,
+      "loss": 1.199,
+      "step": 5404
+    },
+    {
+      "epoch": 2.0123093390420124,
+      "grad_norm": 0.16850432753562927,
+      "learning_rate": 1.8600917477202055e-05,
+      "loss": 1.1914,
+      "step": 5405
+    },
+    {
+      "epoch": 2.012681644192621,
+      "grad_norm": 0.1771266758441925,
+      "learning_rate": 1.8600297432980533e-05,
+      "loss": 1.1708,
+      "step": 5406
+    },
+    {
+      "epoch": 2.0130539493432305,
+      "grad_norm": 0.17538318037986755,
+      "learning_rate": 1.859967726173316e-05,
+      "loss": 1.1903,
+      "step": 5407
+    },
+    {
+      "epoch": 2.0134262544938397,
+      "grad_norm": 0.1645265519618988,
+      "learning_rate": 1.859905696346909e-05,
+      "loss": 1.1905,
+      "step": 5408
+    },
+    {
+      "epoch": 2.0137985596444485,
+      "grad_norm": 0.16865003108978271,
+      "learning_rate": 1.8598436538197494e-05,
+      "loss": 1.1704,
+      "step": 5409
+    },
+    {
+      "epoch": 2.014170864795058,
+      "grad_norm": 0.1677810251712799,
+      "learning_rate": 1.8597815985927524e-05,
+      "loss": 1.2029,
+      "step": 5410
+    },
+    {
+      "epoch": 2.0145431699456666,
+      "grad_norm": 0.1736694723367691,
+      "learning_rate": 1.8597195306668355e-05,
+      "loss": 1.1831,
+      "step": 5411
+    },
+    {
+      "epoch": 2.014915475096276,
+      "grad_norm": 0.17751945555210114,
+      "learning_rate": 1.859657450042915e-05,
+      "loss": 1.19,
+      "step": 5412
+    },
+    {
+      "epoch": 2.0152877802468847,
+      "grad_norm": 0.17150509357452393,
+      "learning_rate": 1.8595953567219077e-05,
+      "loss": 1.2011,
+      "step": 5413
+    },
+    {
+      "epoch": 2.015660085397494,
+      "grad_norm": 0.16916759312152863,
+      "learning_rate": 1.859533250704731e-05,
+      "loss": 1.1708,
+      "step": 5414
+    },
+    {
+      "epoch": 2.016032390548103,
+      "grad_norm": 0.1699335128068924,
+      "learning_rate": 1.8594711319923026e-05,
+      "loss": 1.1795,
+      "step": 5415
+    },
+    {
+      "epoch": 2.016404695698712,
+      "grad_norm": 0.18379724025726318,
+      "learning_rate": 1.859409000585539e-05,
+      "loss": 1.1816,
+      "step": 5416
+    },
+    {
+      "epoch": 2.0167770008493213,
+      "grad_norm": 0.16987095773220062,
+      "learning_rate": 1.8593468564853587e-05,
+      "loss": 1.1684,
+      "step": 5417
+    },
+    {
+      "epoch": 2.01714930599993,
+      "grad_norm": 0.177865669131279,
+      "learning_rate": 1.8592846996926793e-05,
+      "loss": 1.1965,
+      "step": 5418
+    },
+    {
+      "epoch": 2.0175216111505394,
+      "grad_norm": 0.16508503258228302,
+      "learning_rate": 1.8592225302084187e-05,
+      "loss": 1.1869,
+      "step": 5419
+    },
+    {
+      "epoch": 2.0178939163011482,
+      "grad_norm": 0.21210977435112,
+      "learning_rate": 1.8591603480334947e-05,
+      "loss": 1.1911,
+      "step": 5420
+    },
+    {
+      "epoch": 2.0182662214517575,
+      "grad_norm": 0.16533805429935455,
+      "learning_rate": 1.859098153168827e-05,
+      "loss": 1.1882,
+      "step": 5421
+    },
+    {
+      "epoch": 2.0186385266023663,
+      "grad_norm": 0.17425382137298584,
+      "learning_rate": 1.8590359456153333e-05,
+      "loss": 1.1772,
+      "step": 5422
+    },
+    {
+      "epoch": 2.0190108317529756,
+      "grad_norm": 0.16604116559028625,
+      "learning_rate": 1.8589737253739325e-05,
+      "loss": 1.1949,
+      "step": 5423
+    },
+    {
+      "epoch": 2.0193831369035844,
+      "grad_norm": 0.16940464079380035,
+      "learning_rate": 1.8589114924455438e-05,
+      "loss": 1.1811,
+      "step": 5424
+    },
+    {
+      "epoch": 2.0197554420541937,
+      "grad_norm": 0.17113879323005676,
+      "learning_rate": 1.858849246831086e-05,
+      "loss": 1.1798,
+      "step": 5425
+    },
+    {
+      "epoch": 2.020127747204803,
+      "grad_norm": 0.1643594354391098,
+      "learning_rate": 1.8587869885314788e-05,
+      "loss": 1.1811,
+      "step": 5426
+    },
+    {
+      "epoch": 2.0205000523554117,
+      "grad_norm": 0.16450834274291992,
+      "learning_rate": 1.858724717547642e-05,
+      "loss": 1.1875,
+      "step": 5427
+    },
+    {
+      "epoch": 2.020872357506021,
+      "grad_norm": 0.1718573272228241,
+      "learning_rate": 1.8586624338804947e-05,
+      "loss": 1.1845,
+      "step": 5428
+    },
+    {
+      "epoch": 2.02124466265663,
+      "grad_norm": 0.16138458251953125,
+      "learning_rate": 1.8586001375309576e-05,
+      "loss": 1.1766,
+      "step": 5429
+    },
+    {
+      "epoch": 2.021616967807239,
+      "grad_norm": 0.20273058116436005,
+      "learning_rate": 1.85853782849995e-05,
+      "loss": 1.1965,
+      "step": 5430
+    },
+    {
+      "epoch": 2.021989272957848,
+      "grad_norm": 0.17914332449436188,
+      "learning_rate": 1.8584755067883923e-05,
+      "loss": 1.1734,
+      "step": 5431
+    },
+    {
+      "epoch": 2.022361578108457,
+      "grad_norm": 0.17309610545635223,
+      "learning_rate": 1.8584131723972055e-05,
+      "loss": 1.1775,
+      "step": 5432
+    },
+    {
+      "epoch": 2.022733883259066,
+      "grad_norm": 0.18708333373069763,
+      "learning_rate": 1.8583508253273098e-05,
+      "loss": 1.195,
+      "step": 5433
+    },
+    {
+      "epoch": 2.0231061884096753,
+      "grad_norm": 0.16678020358085632,
+      "learning_rate": 1.8582884655796266e-05,
+      "loss": 1.1987,
+      "step": 5434
+    },
+    {
+      "epoch": 2.0234784935602845,
+      "grad_norm": 0.16107121109962463,
+      "learning_rate": 1.8582260931550766e-05,
+      "loss": 1.1866,
+      "step": 5435
+    },
+    {
+      "epoch": 2.0238507987108934,
+      "grad_norm": 0.16838499903678894,
+      "learning_rate": 1.8581637080545813e-05,
+      "loss": 1.1874,
+      "step": 5436
+    },
+    {
+      "epoch": 2.0242231038615026,
+      "grad_norm": 0.16616487503051758,
+      "learning_rate": 1.8581013102790612e-05,
+      "loss": 1.1863,
+      "step": 5437
+    },
+    {
+      "epoch": 2.0245954090121114,
+      "grad_norm": 0.18162907660007477,
+      "learning_rate": 1.8580388998294393e-05,
+      "loss": 1.1745,
+      "step": 5438
+    },
+    {
+      "epoch": 2.0249677141627207,
+      "grad_norm": 0.16428324580192566,
+      "learning_rate": 1.8579764767066362e-05,
+      "loss": 1.1795,
+      "step": 5439
+    },
+    {
+      "epoch": 2.0253400193133295,
+      "grad_norm": 0.17782078683376312,
+      "learning_rate": 1.8579140409115744e-05,
+      "loss": 1.1999,
+      "step": 5440
+    },
+    {
+      "epoch": 2.025712324463939,
+      "grad_norm": 0.16830138862133026,
+      "learning_rate": 1.8578515924451765e-05,
+      "loss": 1.173,
+      "step": 5441
+    },
+    {
+      "epoch": 2.0260846296145476,
+      "grad_norm": 0.19714663922786713,
+      "learning_rate": 1.8577891313083637e-05,
+      "loss": 1.2026,
+      "step": 5442
+    },
+    {
+      "epoch": 2.026456934765157,
+      "grad_norm": 0.18386825919151306,
+      "learning_rate": 1.8577266575020598e-05,
+      "loss": 1.1741,
+      "step": 5443
+    },
+    {
+      "epoch": 2.026829239915766,
+      "grad_norm": 0.18373322486877441,
+      "learning_rate": 1.857664171027187e-05,
+      "loss": 1.171,
+      "step": 5444
+    },
+    {
+      "epoch": 2.027201545066375,
+      "grad_norm": 0.23901142179965973,
+      "learning_rate": 1.8576016718846678e-05,
+      "loss": 1.1837,
+      "step": 5445
+    },
+    {
+      "epoch": 2.027573850216984,
+      "grad_norm": 0.16936446726322174,
+      "learning_rate": 1.8575391600754266e-05,
+      "loss": 1.1687,
+      "step": 5446
+    },
+    {
+      "epoch": 2.027946155367593,
+      "grad_norm": 0.17396526038646698,
+      "learning_rate": 1.857476635600385e-05,
+      "loss": 1.1989,
+      "step": 5447
+    },
+    {
+      "epoch": 2.0283184605182023,
+      "grad_norm": 0.17292600870132446,
+      "learning_rate": 1.8574140984604672e-05,
+      "loss": 1.1754,
+      "step": 5448
+    },
+    {
+      "epoch": 2.028690765668811,
+      "grad_norm": 0.1673480123281479,
+      "learning_rate": 1.8573515486565976e-05,
+      "loss": 1.1891,
+      "step": 5449
+    },
+    {
+      "epoch": 2.0290630708194204,
+      "grad_norm": 0.16834761202335358,
+      "learning_rate": 1.8572889861896993e-05,
+      "loss": 1.1937,
+      "step": 5450
+    },
+    {
+      "epoch": 2.0294353759700297,
+      "grad_norm": 0.16884836554527283,
+      "learning_rate": 1.857226411060696e-05,
+      "loss": 1.1812,
+      "step": 5451
+    },
+    {
+      "epoch": 2.0298076811206385,
+      "grad_norm": 0.16621071100234985,
+      "learning_rate": 1.857163823270513e-05,
+      "loss": 1.1886,
+      "step": 5452
+    },
+    {
+      "epoch": 2.0301799862712477,
+      "grad_norm": 0.16866537928581238,
+      "learning_rate": 1.8571012228200737e-05,
+      "loss": 1.1831,
+      "step": 5453
+    },
+    {
+      "epoch": 2.0305522914218566,
+      "grad_norm": 0.16921500861644745,
+      "learning_rate": 1.8570386097103033e-05,
+      "loss": 1.1753,
+      "step": 5454
+    },
+    {
+      "epoch": 2.030924596572466,
+      "grad_norm": 0.1673291027545929,
+      "learning_rate": 1.8569759839421263e-05,
+      "loss": 1.1827,
+      "step": 5455
+    },
+    {
+      "epoch": 2.0312969017230746,
+      "grad_norm": 0.1631641387939453,
+      "learning_rate": 1.856913345516468e-05,
+      "loss": 1.1886,
+      "step": 5456
+    },
+    {
+      "epoch": 2.031669206873684,
+      "grad_norm": 0.16557154059410095,
+      "learning_rate": 1.8568506944342535e-05,
+      "loss": 1.1644,
+      "step": 5457
+    },
+    {
+      "epoch": 2.0320415120242927,
+      "grad_norm": 0.1667705774307251,
+      "learning_rate": 1.8567880306964077e-05,
+      "loss": 1.1846,
+      "step": 5458
+    },
+    {
+      "epoch": 2.032413817174902,
+      "grad_norm": 0.16573351621627808,
+      "learning_rate": 1.8567253543038564e-05,
+      "loss": 1.1714,
+      "step": 5459
+    },
+    {
+      "epoch": 2.0327861223255113,
+      "grad_norm": 0.17770934104919434,
+      "learning_rate": 1.8566626652575257e-05,
+      "loss": 1.1844,
+      "step": 5460
+    },
+    {
+      "epoch": 2.03315842747612,
+      "grad_norm": 0.1648428589105606,
+      "learning_rate": 1.856599963558341e-05,
+      "loss": 1.197,
+      "step": 5461
+    },
+    {
+      "epoch": 2.0335307326267293,
+      "grad_norm": 0.16865648329257965,
+      "learning_rate": 1.8565372492072288e-05,
+      "loss": 1.2002,
+      "step": 5462
+    },
+    {
+      "epoch": 2.033903037777338,
+      "grad_norm": 0.1726176142692566,
+      "learning_rate": 1.856474522205115e-05,
+      "loss": 1.1789,
+      "step": 5463
+    },
+    {
+      "epoch": 2.0342753429279474,
+      "grad_norm": 0.1659863144159317,
+      "learning_rate": 1.856411782552926e-05,
+      "loss": 1.1927,
+      "step": 5464
+    },
+    {
+      "epoch": 2.0346476480785562,
+      "grad_norm": 0.17737741768360138,
+      "learning_rate": 1.856349030251589e-05,
+      "loss": 1.1853,
+      "step": 5465
+    },
+    {
+      "epoch": 2.0350199532291655,
+      "grad_norm": 0.1588411182165146,
+      "learning_rate": 1.8562862653020306e-05,
+      "loss": 1.1897,
+      "step": 5466
+    },
+    {
+      "epoch": 2.0353922583797743,
+      "grad_norm": 0.200198695063591,
+      "learning_rate": 1.8562234877051778e-05,
+      "loss": 1.1845,
+      "step": 5467
+    },
+    {
+      "epoch": 2.0357645635303836,
+      "grad_norm": 0.18898078799247742,
+      "learning_rate": 1.8561606974619577e-05,
+      "loss": 1.1717,
+      "step": 5468
+    },
+    {
+      "epoch": 2.036136868680993,
+      "grad_norm": 0.17676281929016113,
+      "learning_rate": 1.856097894573298e-05,
+      "loss": 1.1838,
+      "step": 5469
+    },
+    {
+      "epoch": 2.0365091738316017,
+      "grad_norm": 0.23812207579612732,
+      "learning_rate": 1.856035079040126e-05,
+      "loss": 1.1728,
+      "step": 5470
+    },
+    {
+      "epoch": 2.036881478982211,
+      "grad_norm": 0.17304615676403046,
+      "learning_rate": 1.8559722508633698e-05,
+      "loss": 1.1913,
+      "step": 5471
+    },
+    {
+      "epoch": 2.0372537841328198,
+      "grad_norm": 0.17756541073322296,
+      "learning_rate": 1.8559094100439568e-05,
+      "loss": 1.1787,
+      "step": 5472
+    },
+    {
+      "epoch": 2.037626089283429,
+      "grad_norm": 0.16720567643642426,
+      "learning_rate": 1.8558465565828156e-05,
+      "loss": 1.1778,
+      "step": 5473
+    },
+    {
+      "epoch": 2.037998394434038,
+      "grad_norm": 0.19061878323554993,
+      "learning_rate": 1.855783690480875e-05,
+      "loss": 1.1798,
+      "step": 5474
+    },
+    {
+      "epoch": 2.038370699584647,
+      "grad_norm": 0.16539441049098969,
+      "learning_rate": 1.8557208117390626e-05,
+      "loss": 1.1939,
+      "step": 5475
+    },
+    {
+      "epoch": 2.038743004735256,
+      "grad_norm": 0.16110876202583313,
+      "learning_rate": 1.8556579203583075e-05,
+      "loss": 1.1937,
+      "step": 5476
+    },
+    {
+      "epoch": 2.039115309885865,
+      "grad_norm": 0.16478274762630463,
+      "learning_rate": 1.8555950163395383e-05,
+      "loss": 1.189,
+      "step": 5477
+    },
+    {
+      "epoch": 2.0394876150364745,
+      "grad_norm": 0.16724629700183868,
+      "learning_rate": 1.855532099683685e-05,
+      "loss": 1.1846,
+      "step": 5478
+    },
+    {
+      "epoch": 2.0398599201870833,
+      "grad_norm": 0.1901393085718155,
+      "learning_rate": 1.855469170391676e-05,
+      "loss": 1.1695,
+      "step": 5479
+    },
+    {
+      "epoch": 2.0402322253376926,
+      "grad_norm": 0.16991356015205383,
+      "learning_rate": 1.855406228464441e-05,
+      "loss": 1.2068,
+      "step": 5480
+    },
+    {
+      "epoch": 2.0406045304883014,
+      "grad_norm": 0.1755204200744629,
+      "learning_rate": 1.85534327390291e-05,
+      "loss": 1.178,
+      "step": 5481
+    },
+    {
+      "epoch": 2.0409768356389106,
+      "grad_norm": 0.17370069026947021,
+      "learning_rate": 1.855280306708012e-05,
+      "loss": 1.1727,
+      "step": 5482
+    },
+    {
+      "epoch": 2.0413491407895195,
+      "grad_norm": 0.17671510577201843,
+      "learning_rate": 1.8552173268806778e-05,
+      "loss": 1.1941,
+      "step": 5483
+    },
+    {
+      "epoch": 2.0417214459401287,
+      "grad_norm": 0.17848750948905945,
+      "learning_rate": 1.8551543344218372e-05,
+      "loss": 1.169,
+      "step": 5484
+    },
+    {
+      "epoch": 2.0420937510907375,
+      "grad_norm": 0.18597623705863953,
+      "learning_rate": 1.855091329332421e-05,
+      "loss": 1.1696,
+      "step": 5485
+    },
+    {
+      "epoch": 2.042466056241347,
+      "grad_norm": 0.19254527986049652,
+      "learning_rate": 1.8550283116133595e-05,
+      "loss": 1.1791,
+      "step": 5486
+    },
+    {
+      "epoch": 2.042838361391956,
+      "grad_norm": 0.17649273574352264,
+      "learning_rate": 1.8549652812655836e-05,
+      "loss": 1.1929,
+      "step": 5487
+    },
+    {
+      "epoch": 2.043210666542565,
+      "grad_norm": 0.16401410102844238,
+      "learning_rate": 1.854902238290024e-05,
+      "loss": 1.1727,
+      "step": 5488
+    },
+    {
+      "epoch": 2.043582971693174,
+      "grad_norm": 0.16937775909900665,
+      "learning_rate": 1.854839182687612e-05,
+      "loss": 1.1872,
+      "step": 5489
+    },
+    {
+      "epoch": 2.043955276843783,
+      "grad_norm": 0.16704395413398743,
+      "learning_rate": 1.854776114459279e-05,
+      "loss": 1.1857,
+      "step": 5490
+    },
+    {
+      "epoch": 2.0443275819943922,
+      "grad_norm": 0.16636380553245544,
+      "learning_rate": 1.8547130336059562e-05,
+      "loss": 1.1877,
+      "step": 5491
+    },
+    {
+      "epoch": 2.044699887145001,
+      "grad_norm": 0.17131738364696503,
+      "learning_rate": 1.8546499401285755e-05,
+      "loss": 1.1872,
+      "step": 5492
+    },
+    {
+      "epoch": 2.0450721922956103,
+      "grad_norm": 0.1786455512046814,
+      "learning_rate": 1.854586834028069e-05,
+      "loss": 1.1894,
+      "step": 5493
+    },
+    {
+      "epoch": 2.045444497446219,
+      "grad_norm": 0.17036451399326324,
+      "learning_rate": 1.8545237153053688e-05,
+      "loss": 1.1865,
+      "step": 5494
+    },
+    {
+      "epoch": 2.0458168025968284,
+      "grad_norm": 0.17519360780715942,
+      "learning_rate": 1.8544605839614066e-05,
+      "loss": 1.1922,
+      "step": 5495
+    },
+    {
+      "epoch": 2.0461891077474377,
+      "grad_norm": 0.17141428589820862,
+      "learning_rate": 1.8543974399971153e-05,
+      "loss": 1.1991,
+      "step": 5496
+    },
+    {
+      "epoch": 2.0465614128980465,
+      "grad_norm": 0.17702911794185638,
+      "learning_rate": 1.8543342834134276e-05,
+      "loss": 1.1764,
+      "step": 5497
+    },
+    {
+      "epoch": 2.0469337180486558,
+      "grad_norm": 0.16847530007362366,
+      "learning_rate": 1.854271114211276e-05,
+      "loss": 1.1831,
+      "step": 5498
+    },
+    {
+      "epoch": 2.0473060231992646,
+      "grad_norm": 0.17060035467147827,
+      "learning_rate": 1.8542079323915935e-05,
+      "loss": 1.2039,
+      "step": 5499
+    },
+    {
+      "epoch": 2.047678328349874,
+      "grad_norm": 0.1642797440290451,
+      "learning_rate": 1.8541447379553136e-05,
+      "loss": 1.1645,
+      "step": 5500
+    },
+    {
+      "epoch": 2.047678328349874,
+      "eval_loss": 1.3067960739135742,
+      "eval_runtime": 17.3329,
+      "eval_samples_per_second": 100.041,
+      "eval_steps_per_second": 5.019,
+      "step": 5500
+    },
+    {
+      "epoch": 2.0480506335004827,
+      "grad_norm": 0.2160232812166214,
+      "learning_rate": 1.8540815309033697e-05,
+      "loss": 1.2004,
+      "step": 5501
+    },
+    {
+      "epoch": 2.048422938651092,
+      "grad_norm": 0.1832730919122696,
+      "learning_rate": 1.854018311236695e-05,
+      "loss": 1.1862,
+      "step": 5502
+    },
+    {
+      "epoch": 2.0487952438017007,
+      "grad_norm": 0.18487650156021118,
+      "learning_rate": 1.8539550789562234e-05,
+      "loss": 1.2079,
+      "step": 5503
+    },
+    {
+      "epoch": 2.04916754895231,
+      "grad_norm": 0.16732271015644073,
+      "learning_rate": 1.853891834062889e-05,
+      "loss": 1.1865,
+      "step": 5504
+    },
+    {
+      "epoch": 2.0495398541029193,
+      "grad_norm": 0.21880120038986206,
+      "learning_rate": 1.853828576557626e-05,
+      "loss": 1.1793,
+      "step": 5505
+    },
+    {
+      "epoch": 2.049912159253528,
+      "grad_norm": 0.16671521961688995,
+      "learning_rate": 1.853765306441368e-05,
+      "loss": 1.1817,
+      "step": 5506
+    },
+    {
+      "epoch": 2.0502844644041374,
+      "grad_norm": 0.17051751911640167,
+      "learning_rate": 1.8537020237150503e-05,
+      "loss": 1.1646,
+      "step": 5507
+    },
+    {
+      "epoch": 2.050656769554746,
+      "grad_norm": 0.16859528422355652,
+      "learning_rate": 1.853638728379607e-05,
+      "loss": 1.1764,
+      "step": 5508
+    },
+    {
+      "epoch": 2.0510290747053554,
+      "grad_norm": 0.16496269404888153,
+      "learning_rate": 1.8535754204359737e-05,
+      "loss": 1.183,
+      "step": 5509
+    },
+    {
+      "epoch": 2.0514013798559643,
+      "grad_norm": 0.16721811890602112,
+      "learning_rate": 1.853512099885085e-05,
+      "loss": 1.2113,
+      "step": 5510
+    },
+    {
+      "epoch": 2.0517736850065735,
+      "grad_norm": 0.17551200091838837,
+      "learning_rate": 1.8534487667278757e-05,
+      "loss": 1.177,
+      "step": 5511
+    },
+    {
+      "epoch": 2.052145990157183,
+      "grad_norm": 0.16415713727474213,
+      "learning_rate": 1.853385420965282e-05,
+      "loss": 1.1782,
+      "step": 5512
+    },
+    {
+      "epoch": 2.0525182953077916,
+      "grad_norm": 0.16708432137966156,
+      "learning_rate": 1.8533220625982392e-05,
+      "loss": 1.2093,
+      "step": 5513
+    },
+    {
+      "epoch": 2.052890600458401,
+      "grad_norm": 0.16967132687568665,
+      "learning_rate": 1.8532586916276828e-05,
+      "loss": 1.18,
+      "step": 5514
+    },
+    {
+      "epoch": 2.0532629056090097,
+      "grad_norm": 0.17362266778945923,
+      "learning_rate": 1.8531953080545494e-05,
+      "loss": 1.1906,
+      "step": 5515
+    },
+    {
+      "epoch": 2.053635210759619,
+      "grad_norm": 0.1716122329235077,
+      "learning_rate": 1.853131911879775e-05,
+      "loss": 1.1809,
+      "step": 5516
+    },
+    {
+      "epoch": 2.054007515910228,
+      "grad_norm": 0.1693599373102188,
+      "learning_rate": 1.8530685031042952e-05,
+      "loss": 1.1829,
+      "step": 5517
+    },
+    {
+      "epoch": 2.054379821060837,
+      "grad_norm": 0.1685894876718521,
+      "learning_rate": 1.8530050817290477e-05,
+      "loss": 1.1855,
+      "step": 5518
+    },
+    {
+      "epoch": 2.054752126211446,
+      "grad_norm": 0.18109537661075592,
+      "learning_rate": 1.852941647754968e-05,
+      "loss": 1.1782,
+      "step": 5519
+    },
+    {
+      "epoch": 2.055124431362055,
+      "grad_norm": 0.17635756731033325,
+      "learning_rate": 1.8528782011829945e-05,
+      "loss": 1.1857,
+      "step": 5520
+    },
+    {
+      "epoch": 2.0554967365126644,
+      "grad_norm": 0.1745537966489792,
+      "learning_rate": 1.852814742014063e-05,
+      "loss": 1.1963,
+      "step": 5521
+    },
+    {
+      "epoch": 2.0558690416632732,
+      "grad_norm": 0.17796571552753448,
+      "learning_rate": 1.8527512702491116e-05,
+      "loss": 1.1818,
+      "step": 5522
+    },
+    {
+      "epoch": 2.0562413468138825,
+      "grad_norm": 0.17338848114013672,
+      "learning_rate": 1.852687785889077e-05,
+      "loss": 1.1864,
+      "step": 5523
+    },
+    {
+      "epoch": 2.0566136519644913,
+      "grad_norm": 0.18581651151180267,
+      "learning_rate": 1.8526242889348976e-05,
+      "loss": 1.1952,
+      "step": 5524
+    },
+    {
+      "epoch": 2.0569859571151006,
+      "grad_norm": 0.1682850867509842,
+      "learning_rate": 1.852560779387511e-05,
+      "loss": 1.1846,
+      "step": 5525
+    },
+    {
+      "epoch": 2.0573582622657094,
+      "grad_norm": 0.19426992535591125,
+      "learning_rate": 1.8524972572478554e-05,
+      "loss": 1.1823,
+      "step": 5526
+    },
+    {
+      "epoch": 2.0577305674163187,
+      "grad_norm": 0.17708083987236023,
+      "learning_rate": 1.8524337225168682e-05,
+      "loss": 1.1776,
+      "step": 5527
+    },
+    {
+      "epoch": 2.0581028725669275,
+      "grad_norm": 0.1654316484928131,
+      "learning_rate": 1.852370175195489e-05,
+      "loss": 1.1824,
+      "step": 5528
+    },
+    {
+      "epoch": 2.0584751777175367,
+      "grad_norm": 0.20494744181632996,
+      "learning_rate": 1.8523066152846552e-05,
+      "loss": 1.1866,
+      "step": 5529
+    },
+    {
+      "epoch": 2.058847482868146,
+      "grad_norm": 0.17771178483963013,
+      "learning_rate": 1.852243042785307e-05,
+      "loss": 1.1918,
+      "step": 5530
+    },
+    {
+      "epoch": 2.059219788018755,
+      "grad_norm": 0.17502543330192566,
+      "learning_rate": 1.852179457698382e-05,
+      "loss": 1.1821,
+      "step": 5531
+    },
+    {
+      "epoch": 2.059592093169364,
+      "grad_norm": 0.1928001195192337,
+      "learning_rate": 1.85211586002482e-05,
+      "loss": 1.1901,
+      "step": 5532
+    },
+    {
+      "epoch": 2.059964398319973,
+      "grad_norm": 0.16364991664886475,
+      "learning_rate": 1.85205224976556e-05,
+      "loss": 1.1747,
+      "step": 5533
+    },
+    {
+      "epoch": 2.060336703470582,
+      "grad_norm": 0.1756642460823059,
+      "learning_rate": 1.851988626921542e-05,
+      "loss": 1.1723,
+      "step": 5534
+    },
+    {
+      "epoch": 2.060709008621191,
+      "grad_norm": 0.19008730351924896,
+      "learning_rate": 1.8519249914937056e-05,
+      "loss": 1.1996,
+      "step": 5535
+    },
+    {
+      "epoch": 2.0610813137718003,
+      "grad_norm": 0.17250213027000427,
+      "learning_rate": 1.85186134348299e-05,
+      "loss": 1.1882,
+      "step": 5536
+    },
+    {
+      "epoch": 2.061453618922409,
+      "grad_norm": 0.1646471917629242,
+      "learning_rate": 1.8517976828903365e-05,
+      "loss": 1.1918,
+      "step": 5537
+    },
+    {
+      "epoch": 2.0618259240730183,
+      "grad_norm": 0.1910063773393631,
+      "learning_rate": 1.851734009716684e-05,
+      "loss": 1.187,
+      "step": 5538
+    },
+    {
+      "epoch": 2.0621982292236276,
+      "grad_norm": 0.19279128313064575,
+      "learning_rate": 1.8516703239629744e-05,
+      "loss": 1.1942,
+      "step": 5539
+    },
+    {
+      "epoch": 2.0625705343742364,
+      "grad_norm": 0.16666115820407867,
+      "learning_rate": 1.8516066256301468e-05,
+      "loss": 1.1883,
+      "step": 5540
+    },
+    {
+      "epoch": 2.0629428395248457,
+      "grad_norm": 0.20289574563503265,
+      "learning_rate": 1.8515429147191434e-05,
+      "loss": 1.1917,
+      "step": 5541
+    },
+    {
+      "epoch": 2.0633151446754545,
+      "grad_norm": 0.18597279489040375,
+      "learning_rate": 1.851479191230904e-05,
+      "loss": 1.1754,
+      "step": 5542
+    },
+    {
+      "epoch": 2.063687449826064,
+      "grad_norm": 0.17310966551303864,
+      "learning_rate": 1.851415455166371e-05,
+      "loss": 1.1892,
+      "step": 5543
+    },
+    {
+      "epoch": 2.0640597549766726,
+      "grad_norm": 0.20302632451057434,
+      "learning_rate": 1.851351706526485e-05,
+      "loss": 1.1921,
+      "step": 5544
+    },
+    {
+      "epoch": 2.064432060127282,
+      "grad_norm": 0.17062672972679138,
+      "learning_rate": 1.8512879453121874e-05,
+      "loss": 1.1821,
+      "step": 5545
+    },
+    {
+      "epoch": 2.0648043652778907,
+      "grad_norm": 0.16479294002056122,
+      "learning_rate": 1.8512241715244203e-05,
+      "loss": 1.1813,
+      "step": 5546
+    },
+    {
+      "epoch": 2.0651766704285,
+      "grad_norm": 0.17303155362606049,
+      "learning_rate": 1.8511603851641256e-05,
+      "loss": 1.1988,
+      "step": 5547
+    },
+    {
+      "epoch": 2.065548975579109,
+      "grad_norm": 0.17052410542964935,
+      "learning_rate": 1.8510965862322455e-05,
+      "loss": 1.1997,
+      "step": 5548
+    },
+    {
+      "epoch": 2.065921280729718,
+      "grad_norm": 0.15899476408958435,
+      "learning_rate": 1.8510327747297225e-05,
+      "loss": 1.1912,
+      "step": 5549
+    },
+    {
+      "epoch": 2.0662935858803273,
+      "grad_norm": 0.18540219962596893,
+      "learning_rate": 1.8509689506574986e-05,
+      "loss": 1.198,
+      "step": 5550
+    },
+    {
+      "epoch": 2.066665891030936,
+      "grad_norm": 0.1953384429216385,
+      "learning_rate": 1.8509051140165167e-05,
+      "loss": 1.1713,
+      "step": 5551
+    },
+    {
+      "epoch": 2.0670381961815454,
+      "grad_norm": 0.1603497862815857,
+      "learning_rate": 1.85084126480772e-05,
+      "loss": 1.1924,
+      "step": 5552
+    },
+    {
+      "epoch": 2.067410501332154,
+      "grad_norm": 0.3583022654056549,
+      "learning_rate": 1.8507774030320508e-05,
+      "loss": 1.1836,
+      "step": 5553
+    },
+    {
+      "epoch": 2.0677828064827635,
+      "grad_norm": 0.1765451580286026,
+      "learning_rate": 1.8507135286904527e-05,
+      "loss": 1.1841,
+      "step": 5554
+    },
+    {
+      "epoch": 2.0681551116333723,
+      "grad_norm": 0.17876610159873962,
+      "learning_rate": 1.8506496417838695e-05,
+      "loss": 1.1916,
+      "step": 5555
+    },
+    {
+      "epoch": 2.0685274167839816,
+      "grad_norm": 0.17126069962978363,
+      "learning_rate": 1.8505857423132447e-05,
+      "loss": 1.1818,
+      "step": 5556
+    },
+    {
+      "epoch": 2.068899721934591,
+      "grad_norm": 0.1569199413061142,
+      "learning_rate": 1.8505218302795213e-05,
+      "loss": 1.193,
+      "step": 5557
+    },
+    {
+      "epoch": 2.0692720270851996,
+      "grad_norm": 0.17595240473747253,
+      "learning_rate": 1.8504579056836437e-05,
+      "loss": 1.1999,
+      "step": 5558
+    },
+    {
+      "epoch": 2.069644332235809,
+      "grad_norm": 0.18248499929904938,
+      "learning_rate": 1.850393968526557e-05,
+      "loss": 1.1813,
+      "step": 5559
+    },
+    {
+      "epoch": 2.0700166373864177,
+      "grad_norm": 0.17279589176177979,
+      "learning_rate": 1.850330018809204e-05,
+      "loss": 1.1837,
+      "step": 5560
+    },
+    {
+      "epoch": 2.070388942537027,
+      "grad_norm": 0.16098688542842865,
+      "learning_rate": 1.8502660565325302e-05,
+      "loss": 1.1816,
+      "step": 5561
+    },
+    {
+      "epoch": 2.070761247687636,
+      "grad_norm": 0.16793614625930786,
+      "learning_rate": 1.8502020816974806e-05,
+      "loss": 1.1893,
+      "step": 5562
+    },
+    {
+      "epoch": 2.071133552838245,
+      "grad_norm": 0.17572426795959473,
+      "learning_rate": 1.850138094304999e-05,
+      "loss": 1.182,
+      "step": 5563
+    },
+    {
+      "epoch": 2.0715058579888543,
+      "grad_norm": 0.1719566136598587,
+      "learning_rate": 1.850074094356031e-05,
+      "loss": 1.191,
+      "step": 5564
+    },
+    {
+      "epoch": 2.071878163139463,
+      "grad_norm": 0.16841794550418854,
+      "learning_rate": 1.8500100818515224e-05,
+      "loss": 1.1896,
+      "step": 5565
+    },
+    {
+      "epoch": 2.0722504682900724,
+      "grad_norm": 0.16912922263145447,
+      "learning_rate": 1.8499460567924182e-05,
+      "loss": 1.1908,
+      "step": 5566
+    },
+    {
+      "epoch": 2.0726227734406812,
+      "grad_norm": 0.1644641011953354,
+      "learning_rate": 1.849882019179664e-05,
+      "loss": 1.1758,
+      "step": 5567
+    },
+    {
+      "epoch": 2.0729950785912905,
+      "grad_norm": 0.16581396758556366,
+      "learning_rate": 1.8498179690142057e-05,
+      "loss": 1.1841,
+      "step": 5568
+    },
+    {
+      "epoch": 2.0733673837418993,
+      "grad_norm": 0.165495365858078,
+      "learning_rate": 1.8497539062969893e-05,
+      "loss": 1.1847,
+      "step": 5569
+    },
+    {
+      "epoch": 2.0737396888925086,
+      "grad_norm": 0.16472892463207245,
+      "learning_rate": 1.849689831028961e-05,
+      "loss": 1.1978,
+      "step": 5570
+    },
+    {
+      "epoch": 2.0741119940431174,
+      "grad_norm": 0.1632337123155594,
+      "learning_rate": 1.8496257432110673e-05,
+      "loss": 1.1641,
+      "step": 5571
+    },
+    {
+      "epoch": 2.0744842991937267,
+      "grad_norm": 0.16693541407585144,
+      "learning_rate": 1.8495616428442546e-05,
+      "loss": 1.178,
+      "step": 5572
+    },
+    {
+      "epoch": 2.074856604344336,
+      "grad_norm": 0.17431595921516418,
+      "learning_rate": 1.84949752992947e-05,
+      "loss": 1.1958,
+      "step": 5573
+    },
+    {
+      "epoch": 2.0752289094949448,
+      "grad_norm": 0.1731480062007904,
+      "learning_rate": 1.84943340446766e-05,
+      "loss": 1.1769,
+      "step": 5574
+    },
+    {
+      "epoch": 2.075601214645554,
+      "grad_norm": 0.16712194681167603,
+      "learning_rate": 1.849369266459772e-05,
+      "loss": 1.1871,
+      "step": 5575
+    },
+    {
+      "epoch": 2.075973519796163,
+      "grad_norm": 0.16776369512081146,
+      "learning_rate": 1.849305115906753e-05,
+      "loss": 1.1734,
+      "step": 5576
+    },
+    {
+      "epoch": 2.076345824946772,
+      "grad_norm": 0.17154593765735626,
+      "learning_rate": 1.849240952809551e-05,
+      "loss": 1.1959,
+      "step": 5577
+    },
+    {
+      "epoch": 2.076718130097381,
+      "grad_norm": 0.16923579573631287,
+      "learning_rate": 1.8491767771691133e-05,
+      "loss": 1.1937,
+      "step": 5578
+    },
+    {
+      "epoch": 2.07709043524799,
+      "grad_norm": 0.17263515293598175,
+      "learning_rate": 1.849112588986388e-05,
+      "loss": 1.1912,
+      "step": 5579
+    },
+    {
+      "epoch": 2.077462740398599,
+      "grad_norm": 0.16480253636837006,
+      "learning_rate": 1.849048388262323e-05,
+      "loss": 1.1821,
+      "step": 5580
+    },
+    {
+      "epoch": 2.0778350455492083,
+      "grad_norm": 0.16835813224315643,
+      "learning_rate": 1.8489841749978668e-05,
+      "loss": 1.1754,
+      "step": 5581
+    },
+    {
+      "epoch": 2.0782073506998175,
+      "grad_norm": 0.17042113840579987,
+      "learning_rate": 1.848919949193967e-05,
+      "loss": 1.1767,
+      "step": 5582
+    },
+    {
+      "epoch": 2.0785796558504264,
+      "grad_norm": 0.17123256623744965,
+      "learning_rate": 1.8488557108515736e-05,
+      "loss": 1.1778,
+      "step": 5583
+    },
+    {
+      "epoch": 2.0789519610010356,
+      "grad_norm": 0.16714586317539215,
+      "learning_rate": 1.848791459971634e-05,
+      "loss": 1.183,
+      "step": 5584
+    },
+    {
+      "epoch": 2.0793242661516445,
+      "grad_norm": 0.16388732194900513,
+      "learning_rate": 1.848727196555098e-05,
+      "loss": 1.1881,
+      "step": 5585
+    },
+    {
+      "epoch": 2.0796965713022537,
+      "grad_norm": 0.1711377501487732,
+      "learning_rate": 1.8486629206029146e-05,
+      "loss": 1.1975,
+      "step": 5586
+    },
+    {
+      "epoch": 2.0800688764528625,
+      "grad_norm": 0.16601723432540894,
+      "learning_rate": 1.8485986321160335e-05,
+      "loss": 1.1854,
+      "step": 5587
+    },
+    {
+      "epoch": 2.080441181603472,
+      "grad_norm": 0.16255028545856476,
+      "learning_rate": 1.8485343310954033e-05,
+      "loss": 1.1762,
+      "step": 5588
+    },
+    {
+      "epoch": 2.0808134867540806,
+      "grad_norm": 0.1604170799255371,
+      "learning_rate": 1.8484700175419747e-05,
+      "loss": 1.1835,
+      "step": 5589
+    },
+    {
+      "epoch": 2.08118579190469,
+      "grad_norm": 0.16844023764133453,
+      "learning_rate": 1.8484056914566967e-05,
+      "loss": 1.189,
+      "step": 5590
+    },
+    {
+      "epoch": 2.081558097055299,
+      "grad_norm": 0.1678459644317627,
+      "learning_rate": 1.84834135284052e-05,
+      "loss": 1.1828,
+      "step": 5591
+    },
+    {
+      "epoch": 2.081930402205908,
+      "grad_norm": 0.16233506798744202,
+      "learning_rate": 1.8482770016943952e-05,
+      "loss": 1.1749,
+      "step": 5592
+    },
+    {
+      "epoch": 2.0823027073565172,
+      "grad_norm": 0.16579484939575195,
+      "learning_rate": 1.848212638019272e-05,
+      "loss": 1.1794,
+      "step": 5593
+    },
+    {
+      "epoch": 2.082675012507126,
+      "grad_norm": 0.16897903382778168,
+      "learning_rate": 1.8481482618161016e-05,
+      "loss": 1.2122,
+      "step": 5594
+    },
+    {
+      "epoch": 2.0830473176577353,
+      "grad_norm": 0.16713303327560425,
+      "learning_rate": 1.848083873085834e-05,
+      "loss": 1.1682,
+      "step": 5595
+    },
+    {
+      "epoch": 2.083419622808344,
+      "grad_norm": 0.1653788685798645,
+      "learning_rate": 1.8480194718294212e-05,
+      "loss": 1.183,
+      "step": 5596
+    },
+    {
+      "epoch": 2.0837919279589534,
+      "grad_norm": 0.15681540966033936,
+      "learning_rate": 1.847955058047814e-05,
+      "loss": 1.1812,
+      "step": 5597
+    },
+    {
+      "epoch": 2.0841642331095622,
+      "grad_norm": 0.16332903504371643,
+      "learning_rate": 1.8478906317419644e-05,
+      "loss": 1.1776,
+      "step": 5598
+    },
+    {
+      "epoch": 2.0845365382601715,
+      "grad_norm": 0.16454966366291046,
+      "learning_rate": 1.8478261929128226e-05,
+      "loss": 1.1807,
+      "step": 5599
+    },
+    {
+      "epoch": 2.0849088434107808,
+      "grad_norm": 0.16205929219722748,
+      "learning_rate": 1.8477617415613413e-05,
+      "loss": 1.1902,
+      "step": 5600
+    },
+    {
+      "epoch": 2.0852811485613896,
+      "grad_norm": 0.16746239364147186,
+      "learning_rate": 1.8476972776884724e-05,
+      "loss": 1.2076,
+      "step": 5601
+    },
+    {
+      "epoch": 2.085653453711999,
+      "grad_norm": 0.1641511619091034,
+      "learning_rate": 1.8476328012951677e-05,
+      "loss": 1.1847,
+      "step": 5602
+    },
+    {
+      "epoch": 2.0860257588626077,
+      "grad_norm": 0.16240178048610687,
+      "learning_rate": 1.84756831238238e-05,
+      "loss": 1.1733,
+      "step": 5603
+    },
+    {
+      "epoch": 2.086398064013217,
+      "grad_norm": 0.16634000837802887,
+      "learning_rate": 1.8475038109510612e-05,
+      "loss": 1.1749,
+      "step": 5604
+    },
+    {
+      "epoch": 2.0867703691638257,
+      "grad_norm": 0.17073014378547668,
+      "learning_rate": 1.847439297002165e-05,
+      "loss": 1.1751,
+      "step": 5605
+    },
+    {
+      "epoch": 2.087142674314435,
+      "grad_norm": 0.15842556953430176,
+      "learning_rate": 1.8473747705366427e-05,
+      "loss": 1.1722,
+      "step": 5606
+    },
+    {
+      "epoch": 2.087514979465044,
+      "grad_norm": 0.15961557626724243,
+      "learning_rate": 1.8473102315554484e-05,
+      "loss": 1.1957,
+      "step": 5607
+    },
+    {
+      "epoch": 2.087887284615653,
+      "grad_norm": 0.1633174568414688,
+      "learning_rate": 1.8472456800595355e-05,
+      "loss": 1.1652,
+      "step": 5608
+    },
+    {
+      "epoch": 2.0882595897662624,
+      "grad_norm": 0.17014336585998535,
+      "learning_rate": 1.847181116049857e-05,
+      "loss": 1.1878,
+      "step": 5609
+    },
+    {
+      "epoch": 2.088631894916871,
+      "grad_norm": 0.16674593091011047,
+      "learning_rate": 1.8471165395273666e-05,
+      "loss": 1.1844,
+      "step": 5610
+    },
+    {
+      "epoch": 2.0890042000674804,
+      "grad_norm": 0.16320446133613586,
+      "learning_rate": 1.8470519504930178e-05,
+      "loss": 1.1932,
+      "step": 5611
+    },
+    {
+      "epoch": 2.0893765052180893,
+      "grad_norm": 0.1698993742465973,
+      "learning_rate": 1.846987348947765e-05,
+      "loss": 1.1791,
+      "step": 5612
+    },
+    {
+      "epoch": 2.0897488103686985,
+      "grad_norm": 0.16859601438045502,
+      "learning_rate": 1.8469227348925624e-05,
+      "loss": 1.1977,
+      "step": 5613
+    },
+    {
+      "epoch": 2.0901211155193073,
+      "grad_norm": 0.16423748433589935,
+      "learning_rate": 1.846858108328364e-05,
+      "loss": 1.1899,
+      "step": 5614
+    },
+    {
+      "epoch": 2.0904934206699166,
+      "grad_norm": 0.15783381462097168,
+      "learning_rate": 1.8467934692561242e-05,
+      "loss": 1.1945,
+      "step": 5615
+    },
+    {
+      "epoch": 2.0908657258205254,
+      "grad_norm": 0.16708320379257202,
+      "learning_rate": 1.8467288176767986e-05,
+      "loss": 1.1789,
+      "step": 5616
+    },
+    {
+      "epoch": 2.0912380309711347,
+      "grad_norm": 0.16855353116989136,
+      "learning_rate": 1.846664153591341e-05,
+      "loss": 1.1807,
+      "step": 5617
+    },
+    {
+      "epoch": 2.091610336121744,
+      "grad_norm": 0.17245307564735413,
+      "learning_rate": 1.846599477000707e-05,
+      "loss": 1.1927,
+      "step": 5618
+    },
+    {
+      "epoch": 2.091982641272353,
+      "grad_norm": 0.1593155562877655,
+      "learning_rate": 1.8465347879058524e-05,
+      "loss": 1.1677,
+      "step": 5619
+    },
+    {
+      "epoch": 2.092354946422962,
+      "grad_norm": 0.16127386689186096,
+      "learning_rate": 1.8464700863077313e-05,
+      "loss": 1.1698,
+      "step": 5620
+    },
+    {
+      "epoch": 2.092727251573571,
+      "grad_norm": 0.16885069012641907,
+      "learning_rate": 1.846405372207301e-05,
+      "loss": 1.1863,
+      "step": 5621
+    },
+    {
+      "epoch": 2.09309955672418,
+      "grad_norm": 0.16606591641902924,
+      "learning_rate": 1.846340645605516e-05,
+      "loss": 1.1731,
+      "step": 5622
+    },
+    {
+      "epoch": 2.093471861874789,
+      "grad_norm": 0.16918626427650452,
+      "learning_rate": 1.8462759065033326e-05,
+      "loss": 1.1936,
+      "step": 5623
+    },
+    {
+      "epoch": 2.093844167025398,
+      "grad_norm": 0.16348494589328766,
+      "learning_rate": 1.8462111549017074e-05,
+      "loss": 1.1815,
+      "step": 5624
+    },
+    {
+      "epoch": 2.094216472176007,
+      "grad_norm": 0.16731733083724976,
+      "learning_rate": 1.8461463908015967e-05,
+      "loss": 1.2031,
+      "step": 5625
+    },
+    {
+      "epoch": 2.0945887773266163,
+      "grad_norm": 0.1636827141046524,
+      "learning_rate": 1.8460816142039566e-05,
+      "loss": 1.1862,
+      "step": 5626
+    },
+    {
+      "epoch": 2.0949610824772256,
+      "grad_norm": 0.1637611985206604,
+      "learning_rate": 1.8460168251097444e-05,
+      "loss": 1.1915,
+      "step": 5627
+    },
+    {
+      "epoch": 2.0953333876278344,
+      "grad_norm": 0.16247594356536865,
+      "learning_rate": 1.8459520235199165e-05,
+      "loss": 1.1826,
+      "step": 5628
+    },
+    {
+      "epoch": 2.0957056927784437,
+      "grad_norm": 0.1628650724887848,
+      "learning_rate": 1.8458872094354307e-05,
+      "loss": 1.1848,
+      "step": 5629
+    },
+    {
+      "epoch": 2.0960779979290525,
+      "grad_norm": 0.16687503457069397,
+      "learning_rate": 1.8458223828572435e-05,
+      "loss": 1.1787,
+      "step": 5630
+    },
+    {
+      "epoch": 2.0964503030796617,
+      "grad_norm": 0.16528525948524475,
+      "learning_rate": 1.8457575437863134e-05,
+      "loss": 1.1788,
+      "step": 5631
+    },
+    {
+      "epoch": 2.0968226082302706,
+      "grad_norm": 0.16622121632099152,
+      "learning_rate": 1.845692692223597e-05,
+      "loss": 1.1934,
+      "step": 5632
+    },
+    {
+      "epoch": 2.09719491338088,
+      "grad_norm": 0.1611912101507187,
+      "learning_rate": 1.8456278281700527e-05,
+      "loss": 1.1901,
+      "step": 5633
+    },
+    {
+      "epoch": 2.097567218531489,
+      "grad_norm": 0.1634882390499115,
+      "learning_rate": 1.845562951626638e-05,
+      "loss": 1.1897,
+      "step": 5634
+    },
+    {
+      "epoch": 2.097939523682098,
+      "grad_norm": 0.1702599674463272,
+      "learning_rate": 1.8454980625943122e-05,
+      "loss": 1.2001,
+      "step": 5635
+    },
+    {
+      "epoch": 2.098311828832707,
+      "grad_norm": 0.15951082110404968,
+      "learning_rate": 1.845433161074033e-05,
+      "loss": 1.1773,
+      "step": 5636
+    },
+    {
+      "epoch": 2.098684133983316,
+      "grad_norm": 0.16114339232444763,
+      "learning_rate": 1.845368247066759e-05,
+      "loss": 1.1853,
+      "step": 5637
+    },
+    {
+      "epoch": 2.0990564391339253,
+      "grad_norm": 0.17119161784648895,
+      "learning_rate": 1.845303320573449e-05,
+      "loss": 1.1879,
+      "step": 5638
+    },
+    {
+      "epoch": 2.099428744284534,
+      "grad_norm": 0.1637708991765976,
+      "learning_rate": 1.8452383815950616e-05,
+      "loss": 1.1693,
+      "step": 5639
+    },
+    {
+      "epoch": 2.0998010494351433,
+      "grad_norm": 0.16330254077911377,
+      "learning_rate": 1.845173430132557e-05,
+      "loss": 1.1775,
+      "step": 5640
+    },
+    {
+      "epoch": 2.100173354585752,
+      "grad_norm": 0.16474707424640656,
+      "learning_rate": 1.8451084661868936e-05,
+      "loss": 1.166,
+      "step": 5641
+    },
+    {
+      "epoch": 2.1005456597363614,
+      "grad_norm": 0.16186347603797913,
+      "learning_rate": 1.845043489759031e-05,
+      "loss": 1.1778,
+      "step": 5642
+    },
+    {
+      "epoch": 2.1009179648869707,
+      "grad_norm": 0.1683456301689148,
+      "learning_rate": 1.844978500849929e-05,
+      "loss": 1.1905,
+      "step": 5643
+    },
+    {
+      "epoch": 2.1012902700375795,
+      "grad_norm": 0.15968085825443268,
+      "learning_rate": 1.8449134994605483e-05,
+      "loss": 1.1779,
+      "step": 5644
+    },
+    {
+      "epoch": 2.1016625751881888,
+      "grad_norm": 0.15909917652606964,
+      "learning_rate": 1.8448484855918476e-05,
+      "loss": 1.1993,
+      "step": 5645
+    },
+    {
+      "epoch": 2.1020348803387976,
+      "grad_norm": 0.16778358817100525,
+      "learning_rate": 1.844783459244788e-05,
+      "loss": 1.2041,
+      "step": 5646
+    },
+    {
+      "epoch": 2.102407185489407,
+      "grad_norm": 0.1642913818359375,
+      "learning_rate": 1.8447184204203297e-05,
+      "loss": 1.1888,
+      "step": 5647
+    },
+    {
+      "epoch": 2.1027794906400157,
+      "grad_norm": 0.1615377813577652,
+      "learning_rate": 1.8446533691194332e-05,
+      "loss": 1.1834,
+      "step": 5648
+    },
+    {
+      "epoch": 2.103151795790625,
+      "grad_norm": 0.16255392134189606,
+      "learning_rate": 1.8445883053430597e-05,
+      "loss": 1.167,
+      "step": 5649
+    },
+    {
+      "epoch": 2.1035241009412338,
+      "grad_norm": 0.1646036058664322,
+      "learning_rate": 1.84452322909217e-05,
+      "loss": 1.1865,
+      "step": 5650
+    },
+    {
+      "epoch": 2.103896406091843,
+      "grad_norm": 0.17090143263339996,
+      "learning_rate": 1.844458140367725e-05,
+      "loss": 1.1907,
+      "step": 5651
+    },
+    {
+      "epoch": 2.1042687112424523,
+      "grad_norm": 0.17261987924575806,
+      "learning_rate": 1.8443930391706863e-05,
+      "loss": 1.194,
+      "step": 5652
+    },
+    {
+      "epoch": 2.104641016393061,
+      "grad_norm": 0.166230246424675,
+      "learning_rate": 1.8443279255020153e-05,
+      "loss": 1.188,
+      "step": 5653
+    },
+    {
+      "epoch": 2.1050133215436704,
+      "grad_norm": 0.168045312166214,
+      "learning_rate": 1.844262799362674e-05,
+      "loss": 1.1929,
+      "step": 5654
+    },
+    {
+      "epoch": 2.105385626694279,
+      "grad_norm": 0.1657107025384903,
+      "learning_rate": 1.844197660753624e-05,
+      "loss": 1.1787,
+      "step": 5655
+    },
+    {
+      "epoch": 2.1057579318448885,
+      "grad_norm": 0.1633072793483734,
+      "learning_rate": 1.8441325096758275e-05,
+      "loss": 1.1949,
+      "step": 5656
+    },
+    {
+      "epoch": 2.1061302369954973,
+      "grad_norm": 0.16375941038131714,
+      "learning_rate": 1.844067346130247e-05,
+      "loss": 1.1742,
+      "step": 5657
+    },
+    {
+      "epoch": 2.1065025421461065,
+      "grad_norm": 0.16451074182987213,
+      "learning_rate": 1.8440021701178445e-05,
+      "loss": 1.1873,
+      "step": 5658
+    },
+    {
+      "epoch": 2.1068748472967154,
+      "grad_norm": 0.1651698499917984,
+      "learning_rate": 1.843936981639583e-05,
+      "loss": 1.1857,
+      "step": 5659
+    },
+    {
+      "epoch": 2.1072471524473246,
+      "grad_norm": 0.1603275090456009,
+      "learning_rate": 1.8438717806964254e-05,
+      "loss": 1.1923,
+      "step": 5660
+    },
+    {
+      "epoch": 2.107619457597934,
+      "grad_norm": 0.16892802715301514,
+      "learning_rate": 1.8438065672893348e-05,
+      "loss": 1.1954,
+      "step": 5661
+    },
+    {
+      "epoch": 2.1079917627485427,
+      "grad_norm": 0.1669367104768753,
+      "learning_rate": 1.8437413414192734e-05,
+      "loss": 1.19,
+      "step": 5662
+    },
+    {
+      "epoch": 2.108364067899152,
+      "grad_norm": 0.16005094349384308,
+      "learning_rate": 1.843676103087206e-05,
+      "loss": 1.1905,
+      "step": 5663
+    },
+    {
+      "epoch": 2.108736373049761,
+      "grad_norm": 0.16192243993282318,
+      "learning_rate": 1.8436108522940953e-05,
+      "loss": 1.1865,
+      "step": 5664
+    },
+    {
+      "epoch": 2.10910867820037,
+      "grad_norm": 0.16024897992610931,
+      "learning_rate": 1.8435455890409054e-05,
+      "loss": 1.1641,
+      "step": 5665
+    },
+    {
+      "epoch": 2.109480983350979,
+      "grad_norm": 0.1648632138967514,
+      "learning_rate": 1.8434803133285998e-05,
+      "loss": 1.1792,
+      "step": 5666
+    },
+    {
+      "epoch": 2.109853288501588,
+      "grad_norm": 0.15843743085861206,
+      "learning_rate": 1.843415025158143e-05,
+      "loss": 1.1728,
+      "step": 5667
+    },
+    {
+      "epoch": 2.110225593652197,
+      "grad_norm": 0.1614643931388855,
+      "learning_rate": 1.843349724530499e-05,
+      "loss": 1.1726,
+      "step": 5668
+    },
+    {
+      "epoch": 2.1105978988028062,
+      "grad_norm": 0.16818486154079437,
+      "learning_rate": 1.843284411446633e-05,
+      "loss": 1.1832,
+      "step": 5669
+    },
+    {
+      "epoch": 2.1109702039534155,
+      "grad_norm": 0.16501553356647491,
+      "learning_rate": 1.8432190859075088e-05,
+      "loss": 1.1941,
+      "step": 5670
+    },
+    {
+      "epoch": 2.1113425091040243,
+      "grad_norm": 0.16194112598896027,
+      "learning_rate": 1.8431537479140916e-05,
+      "loss": 1.1829,
+      "step": 5671
+    },
+    {
+      "epoch": 2.1117148142546336,
+      "grad_norm": 0.1625380516052246,
+      "learning_rate": 1.8430883974673467e-05,
+      "loss": 1.1866,
+      "step": 5672
+    },
+    {
+      "epoch": 2.1120871194052424,
+      "grad_norm": 0.16628055274486542,
+      "learning_rate": 1.8430230345682393e-05,
+      "loss": 1.167,
+      "step": 5673
+    },
+    {
+      "epoch": 2.1124594245558517,
+      "grad_norm": 0.16523656249046326,
+      "learning_rate": 1.842957659217734e-05,
+      "loss": 1.1693,
+      "step": 5674
+    },
+    {
+      "epoch": 2.1128317297064605,
+      "grad_norm": 0.16235126554965973,
+      "learning_rate": 1.842892271416797e-05,
+      "loss": 1.1814,
+      "step": 5675
+    },
+    {
+      "epoch": 2.1132040348570698,
+      "grad_norm": 0.16224858164787292,
+      "learning_rate": 1.8428268711663943e-05,
+      "loss": 1.1778,
+      "step": 5676
+    },
+    {
+      "epoch": 2.113576340007679,
+      "grad_norm": 0.1565116047859192,
+      "learning_rate": 1.842761458467492e-05,
+      "loss": 1.1834,
+      "step": 5677
+    },
+    {
+      "epoch": 2.113948645158288,
+      "grad_norm": 0.1609666496515274,
+      "learning_rate": 1.842696033321055e-05,
+      "loss": 1.188,
+      "step": 5678
+    },
+    {
+      "epoch": 2.114320950308897,
+      "grad_norm": 0.17431986331939697,
+      "learning_rate": 1.842630595728051e-05,
+      "loss": 1.2025,
+      "step": 5679
+    },
+    {
+      "epoch": 2.114693255459506,
+      "grad_norm": 0.16619716584682465,
+      "learning_rate": 1.842565145689446e-05,
+      "loss": 1.1908,
+      "step": 5680
+    },
+    {
+      "epoch": 2.115065560610115,
+      "grad_norm": 0.16613732278347015,
+      "learning_rate": 1.8424996832062066e-05,
+      "loss": 1.1826,
+      "step": 5681
+    },
+    {
+      "epoch": 2.115437865760724,
+      "grad_norm": 0.16963297128677368,
+      "learning_rate": 1.8424342082792996e-05,
+      "loss": 1.1904,
+      "step": 5682
+    },
+    {
+      "epoch": 2.1158101709113333,
+      "grad_norm": 0.15998825430870056,
+      "learning_rate": 1.8423687209096926e-05,
+      "loss": 1.1873,
+      "step": 5683
+    },
+    {
+      "epoch": 2.116182476061942,
+      "grad_norm": 0.15529409050941467,
+      "learning_rate": 1.842303221098352e-05,
+      "loss": 1.1879,
+      "step": 5684
+    },
+    {
+      "epoch": 2.1165547812125514,
+      "grad_norm": 0.16346554458141327,
+      "learning_rate": 1.8422377088462456e-05,
+      "loss": 1.1873,
+      "step": 5685
+    },
+    {
+      "epoch": 2.1169270863631606,
+      "grad_norm": 0.16552503407001495,
+      "learning_rate": 1.8421721841543412e-05,
+      "loss": 1.1971,
+      "step": 5686
+    },
+    {
+      "epoch": 2.1172993915137694,
+      "grad_norm": 0.16993537545204163,
+      "learning_rate": 1.842106647023607e-05,
+      "loss": 1.2047,
+      "step": 5687
+    },
+    {
+      "epoch": 2.1176716966643787,
+      "grad_norm": 0.16615287959575653,
+      "learning_rate": 1.8420410974550103e-05,
+      "loss": 1.1826,
+      "step": 5688
+    },
+    {
+      "epoch": 2.1180440018149875,
+      "grad_norm": 0.16618463397026062,
+      "learning_rate": 1.8419755354495194e-05,
+      "loss": 1.1757,
+      "step": 5689
+    },
+    {
+      "epoch": 2.118416306965597,
+      "grad_norm": 0.16526402533054352,
+      "learning_rate": 1.8419099610081025e-05,
+      "loss": 1.1833,
+      "step": 5690
+    },
+    {
+      "epoch": 2.1187886121162056,
+      "grad_norm": 0.1653960645198822,
+      "learning_rate": 1.841844374131728e-05,
+      "loss": 1.176,
+      "step": 5691
+    },
+    {
+      "epoch": 2.119160917266815,
+      "grad_norm": 0.16341429948806763,
+      "learning_rate": 1.8417787748213655e-05,
+      "loss": 1.1807,
+      "step": 5692
+    },
+    {
+      "epoch": 2.1195332224174237,
+      "grad_norm": 0.15940816700458527,
+      "learning_rate": 1.8417131630779834e-05,
+      "loss": 1.1814,
+      "step": 5693
+    },
+    {
+      "epoch": 2.119905527568033,
+      "grad_norm": 0.1656249612569809,
+      "learning_rate": 1.84164753890255e-05,
+      "loss": 1.1895,
+      "step": 5694
+    },
+    {
+      "epoch": 2.1202778327186422,
+      "grad_norm": 0.1686115264892578,
+      "learning_rate": 1.841581902296036e-05,
+      "loss": 1.1888,
+      "step": 5695
+    },
+    {
+      "epoch": 2.120650137869251,
+      "grad_norm": 0.1635192185640335,
+      "learning_rate": 1.8415162532594096e-05,
+      "loss": 1.183,
+      "step": 5696
+    },
+    {
+      "epoch": 2.1210224430198603,
+      "grad_norm": 0.15994298458099365,
+      "learning_rate": 1.841450591793641e-05,
+      "loss": 1.1839,
+      "step": 5697
+    },
+    {
+      "epoch": 2.121394748170469,
+      "grad_norm": 0.16956914961338043,
+      "learning_rate": 1.8413849178997002e-05,
+      "loss": 1.1802,
+      "step": 5698
+    },
+    {
+      "epoch": 2.1217670533210784,
+      "grad_norm": 0.16558878123760223,
+      "learning_rate": 1.841319231578557e-05,
+      "loss": 1.2054,
+      "step": 5699
+    },
+    {
+      "epoch": 2.122139358471687,
+      "grad_norm": 0.16168220341205597,
+      "learning_rate": 1.8412535328311813e-05,
+      "loss": 1.1613,
+      "step": 5700
+    },
+    {
+      "epoch": 2.1225116636222965,
+      "grad_norm": 0.16287264227867126,
+      "learning_rate": 1.8411878216585436e-05,
+      "loss": 1.1889,
+      "step": 5701
+    },
+    {
+      "epoch": 2.1228839687729053,
+      "grad_norm": 0.1619829684495926,
+      "learning_rate": 1.841122098061615e-05,
+      "loss": 1.1877,
+      "step": 5702
+    },
+    {
+      "epoch": 2.1232562739235146,
+      "grad_norm": 0.16612021625041962,
+      "learning_rate": 1.8410563620413658e-05,
+      "loss": 1.1818,
+      "step": 5703
+    },
+    {
+      "epoch": 2.123628579074124,
+      "grad_norm": 0.16354532539844513,
+      "learning_rate": 1.8409906135987668e-05,
+      "loss": 1.1882,
+      "step": 5704
+    },
+    {
+      "epoch": 2.1240008842247327,
+      "grad_norm": 0.1652521938085556,
+      "learning_rate": 1.8409248527347888e-05,
+      "loss": 1.1854,
+      "step": 5705
+    },
+    {
+      "epoch": 2.124373189375342,
+      "grad_norm": 0.1664263755083084,
+      "learning_rate": 1.840859079450404e-05,
+      "loss": 1.1875,
+      "step": 5706
+    },
+    {
+      "epoch": 2.1247454945259507,
+      "grad_norm": 0.16455532610416412,
+      "learning_rate": 1.8407932937465835e-05,
+      "loss": 1.1903,
+      "step": 5707
+    },
+    {
+      "epoch": 2.12511779967656,
+      "grad_norm": 0.1682821661233902,
+      "learning_rate": 1.8407274956242983e-05,
+      "loss": 1.1965,
+      "step": 5708
+    },
+    {
+      "epoch": 2.125490104827169,
+      "grad_norm": 0.16767260432243347,
+      "learning_rate": 1.840661685084521e-05,
+      "loss": 1.178,
+      "step": 5709
+    },
+    {
+      "epoch": 2.125862409977778,
+      "grad_norm": 0.16613049805164337,
+      "learning_rate": 1.8405958621282232e-05,
+      "loss": 1.172,
+      "step": 5710
+    },
+    {
+      "epoch": 2.126234715128387,
+      "grad_norm": 0.16709165275096893,
+      "learning_rate": 1.8405300267563774e-05,
+      "loss": 1.1834,
+      "step": 5711
+    },
+    {
+      "epoch": 2.126607020278996,
+      "grad_norm": 0.16999532282352448,
+      "learning_rate": 1.840464178969956e-05,
+      "loss": 1.1879,
+      "step": 5712
+    },
+    {
+      "epoch": 2.1269793254296054,
+      "grad_norm": 0.1657310575246811,
+      "learning_rate": 1.840398318769931e-05,
+      "loss": 1.1959,
+      "step": 5713
+    },
+    {
+      "epoch": 2.1273516305802143,
+      "grad_norm": 0.16610091924667358,
+      "learning_rate": 1.8403324461572762e-05,
+      "loss": 1.1929,
+      "step": 5714
+    },
+    {
+      "epoch": 2.1277239357308235,
+      "grad_norm": 0.16511550545692444,
+      "learning_rate": 1.8402665611329635e-05,
+      "loss": 1.1819,
+      "step": 5715
+    },
+    {
+      "epoch": 2.1280962408814323,
+      "grad_norm": 0.1685413420200348,
+      "learning_rate": 1.8402006636979667e-05,
+      "loss": 1.1871,
+      "step": 5716
+    },
+    {
+      "epoch": 2.1284685460320416,
+      "grad_norm": 0.1703380048274994,
+      "learning_rate": 1.8401347538532585e-05,
+      "loss": 1.1795,
+      "step": 5717
+    },
+    {
+      "epoch": 2.1288408511826504,
+      "grad_norm": 0.16510550677776337,
+      "learning_rate": 1.8400688315998128e-05,
+      "loss": 1.1991,
+      "step": 5718
+    },
+    {
+      "epoch": 2.1292131563332597,
+      "grad_norm": 0.15952372550964355,
+      "learning_rate": 1.8400028969386033e-05,
+      "loss": 1.1844,
+      "step": 5719
+    },
+    {
+      "epoch": 2.1295854614838685,
+      "grad_norm": 0.1626126617193222,
+      "learning_rate": 1.8399369498706034e-05,
+      "loss": 1.189,
+      "step": 5720
+    },
+    {
+      "epoch": 2.129957766634478,
+      "grad_norm": 0.1677938997745514,
+      "learning_rate": 1.8398709903967878e-05,
+      "loss": 1.1905,
+      "step": 5721
+    },
+    {
+      "epoch": 2.130330071785087,
+      "grad_norm": 0.16606491804122925,
+      "learning_rate": 1.8398050185181303e-05,
+      "loss": 1.1932,
+      "step": 5722
+    },
+    {
+      "epoch": 2.130702376935696,
+      "grad_norm": 0.16483409702777863,
+      "learning_rate": 1.8397390342356054e-05,
+      "loss": 1.1819,
+      "step": 5723
+    },
+    {
+      "epoch": 2.131074682086305,
+      "grad_norm": 0.15964972972869873,
+      "learning_rate": 1.8396730375501877e-05,
+      "loss": 1.1887,
+      "step": 5724
+    },
+    {
+      "epoch": 2.131446987236914,
+      "grad_norm": 0.1669287085533142,
+      "learning_rate": 1.8396070284628518e-05,
+      "loss": 1.1811,
+      "step": 5725
+    },
+    {
+      "epoch": 2.131819292387523,
+      "grad_norm": 0.16600339114665985,
+      "learning_rate": 1.839541006974573e-05,
+      "loss": 1.187,
+      "step": 5726
+    },
+    {
+      "epoch": 2.132191597538132,
+      "grad_norm": 0.16649706661701202,
+      "learning_rate": 1.839474973086326e-05,
+      "loss": 1.1863,
+      "step": 5727
+    },
+    {
+      "epoch": 2.1325639026887413,
+      "grad_norm": 0.16561785340309143,
+      "learning_rate": 1.839408926799086e-05,
+      "loss": 1.1911,
+      "step": 5728
+    },
+    {
+      "epoch": 2.13293620783935,
+      "grad_norm": 0.16608673334121704,
+      "learning_rate": 1.8393428681138298e-05,
+      "loss": 1.1734,
+      "step": 5729
+    },
+    {
+      "epoch": 2.1333085129899594,
+      "grad_norm": 0.16552896797657013,
+      "learning_rate": 1.8392767970315314e-05,
+      "loss": 1.1823,
+      "step": 5730
+    },
+    {
+      "epoch": 2.1336808181405686,
+      "grad_norm": 0.16509748995304108,
+      "learning_rate": 1.8392107135531674e-05,
+      "loss": 1.1801,
+      "step": 5731
+    },
+    {
+      "epoch": 2.1340531232911775,
+      "grad_norm": 0.16142834722995758,
+      "learning_rate": 1.839144617679714e-05,
+      "loss": 1.174,
+      "step": 5732
+    },
+    {
+      "epoch": 2.1344254284417867,
+      "grad_norm": 0.17052648961544037,
+      "learning_rate": 1.839078509412147e-05,
+      "loss": 1.1891,
+      "step": 5733
+    },
+    {
+      "epoch": 2.1347977335923956,
+      "grad_norm": 0.15803445875644684,
+      "learning_rate": 1.8390123887514436e-05,
+      "loss": 1.1954,
+      "step": 5734
+    },
+    {
+      "epoch": 2.135170038743005,
+      "grad_norm": 0.16502295434474945,
+      "learning_rate": 1.8389462556985793e-05,
+      "loss": 1.1777,
+      "step": 5735
+    },
+    {
+      "epoch": 2.1355423438936136,
+      "grad_norm": 0.1639169305562973,
+      "learning_rate": 1.838880110254532e-05,
+      "loss": 1.1857,
+      "step": 5736
+    },
+    {
+      "epoch": 2.135914649044223,
+      "grad_norm": 0.16261015832424164,
+      "learning_rate": 1.8388139524202776e-05,
+      "loss": 1.1836,
+      "step": 5737
+    },
+    {
+      "epoch": 2.1362869541948317,
+      "grad_norm": 0.1698441356420517,
+      "learning_rate": 1.8387477821967938e-05,
+      "loss": 1.1822,
+      "step": 5738
+    },
+    {
+      "epoch": 2.136659259345441,
+      "grad_norm": 0.15911483764648438,
+      "learning_rate": 1.8386815995850584e-05,
+      "loss": 1.1733,
+      "step": 5739
+    },
+    {
+      "epoch": 2.1370315644960503,
+      "grad_norm": 0.16580499708652496,
+      "learning_rate": 1.838615404586048e-05,
+      "loss": 1.1809,
+      "step": 5740
+    },
+    {
+      "epoch": 2.137403869646659,
+      "grad_norm": 0.16274124383926392,
+      "learning_rate": 1.8385491972007408e-05,
+      "loss": 1.1737,
+      "step": 5741
+    },
+    {
+      "epoch": 2.1377761747972683,
+      "grad_norm": 0.1641577035188675,
+      "learning_rate": 1.8384829774301145e-05,
+      "loss": 1.1926,
+      "step": 5742
+    },
+    {
+      "epoch": 2.138148479947877,
+      "grad_norm": 0.159241184592247,
+      "learning_rate": 1.8384167452751473e-05,
+      "loss": 1.1794,
+      "step": 5743
+    },
+    {
+      "epoch": 2.1385207850984864,
+      "grad_norm": 0.17162783443927765,
+      "learning_rate": 1.8383505007368175e-05,
+      "loss": 1.1891,
+      "step": 5744
+    },
+    {
+      "epoch": 2.1388930902490952,
+      "grad_norm": 0.17057040333747864,
+      "learning_rate": 1.8382842438161034e-05,
+      "loss": 1.1981,
+      "step": 5745
+    },
+    {
+      "epoch": 2.1392653953997045,
+      "grad_norm": 0.16203658282756805,
+      "learning_rate": 1.8382179745139835e-05,
+      "loss": 1.1763,
+      "step": 5746
+    },
+    {
+      "epoch": 2.1396377005503133,
+      "grad_norm": 0.16306722164154053,
+      "learning_rate": 1.838151692831437e-05,
+      "loss": 1.1903,
+      "step": 5747
+    },
+    {
+      "epoch": 2.1400100057009226,
+      "grad_norm": 0.16811603307724,
+      "learning_rate": 1.838085398769442e-05,
+      "loss": 1.1941,
+      "step": 5748
+    },
+    {
+      "epoch": 2.140382310851532,
+      "grad_norm": 0.1626274734735489,
+      "learning_rate": 1.8380190923289785e-05,
+      "loss": 1.1709,
+      "step": 5749
+    },
+    {
+      "epoch": 2.1407546160021407,
+      "grad_norm": 0.16186201572418213,
+      "learning_rate": 1.837952773511026e-05,
+      "loss": 1.1861,
+      "step": 5750
+    },
+    {
+      "epoch": 2.14112692115275,
+      "grad_norm": 0.1650025099515915,
+      "learning_rate": 1.8378864423165632e-05,
+      "loss": 1.18,
+      "step": 5751
+    },
+    {
+      "epoch": 2.1414992263033588,
+      "grad_norm": 0.15938428044319153,
+      "learning_rate": 1.8378200987465704e-05,
+      "loss": 1.2017,
+      "step": 5752
+    },
+    {
+      "epoch": 2.141871531453968,
+      "grad_norm": 0.16518081724643707,
+      "learning_rate": 1.8377537428020273e-05,
+      "loss": 1.1805,
+      "step": 5753
+    },
+    {
+      "epoch": 2.142243836604577,
+      "grad_norm": 0.16796717047691345,
+      "learning_rate": 1.837687374483914e-05,
+      "loss": 1.1798,
+      "step": 5754
+    },
+    {
+      "epoch": 2.142616141755186,
+      "grad_norm": 0.1624600887298584,
+      "learning_rate": 1.837620993793211e-05,
+      "loss": 1.1888,
+      "step": 5755
+    },
+    {
+      "epoch": 2.142988446905795,
+      "grad_norm": 0.1630149781703949,
+      "learning_rate": 1.837554600730898e-05,
+      "loss": 1.1848,
+      "step": 5756
+    },
+    {
+      "epoch": 2.143360752056404,
+      "grad_norm": 0.16488610208034515,
+      "learning_rate": 1.837488195297956e-05,
+      "loss": 1.1831,
+      "step": 5757
+    },
+    {
+      "epoch": 2.1437330572070135,
+      "grad_norm": 0.1614954173564911,
+      "learning_rate": 1.8374217774953663e-05,
+      "loss": 1.1871,
+      "step": 5758
+    },
+    {
+      "epoch": 2.1441053623576223,
+      "grad_norm": 0.16131506860256195,
+      "learning_rate": 1.8373553473241097e-05,
+      "loss": 1.1734,
+      "step": 5759
+    },
+    {
+      "epoch": 2.1444776675082315,
+      "grad_norm": 0.16610832512378693,
+      "learning_rate": 1.837288904785167e-05,
+      "loss": 1.1945,
+      "step": 5760
+    },
+    {
+      "epoch": 2.1448499726588404,
+      "grad_norm": 0.16947484016418457,
+      "learning_rate": 1.8372224498795198e-05,
+      "loss": 1.1775,
+      "step": 5761
+    },
+    {
+      "epoch": 2.1452222778094496,
+      "grad_norm": 0.17164815962314606,
+      "learning_rate": 1.8371559826081492e-05,
+      "loss": 1.1955,
+      "step": 5762
+    },
+    {
+      "epoch": 2.1455945829600584,
+      "grad_norm": 0.16294875741004944,
+      "learning_rate": 1.8370895029720374e-05,
+      "loss": 1.1761,
+      "step": 5763
+    },
+    {
+      "epoch": 2.1459668881106677,
+      "grad_norm": 0.16515691578388214,
+      "learning_rate": 1.8370230109721664e-05,
+      "loss": 1.1913,
+      "step": 5764
+    },
+    {
+      "epoch": 2.146339193261277,
+      "grad_norm": 0.16761860251426697,
+      "learning_rate": 1.836956506609518e-05,
+      "loss": 1.183,
+      "step": 5765
+    },
+    {
+      "epoch": 2.146711498411886,
+      "grad_norm": 0.16894914209842682,
+      "learning_rate": 1.8368899898850744e-05,
+      "loss": 1.1825,
+      "step": 5766
+    },
+    {
+      "epoch": 2.147083803562495,
+      "grad_norm": 0.159805029630661,
+      "learning_rate": 1.836823460799818e-05,
+      "loss": 1.1854,
+      "step": 5767
+    },
+    {
+      "epoch": 2.147456108713104,
+      "grad_norm": 0.16420911252498627,
+      "learning_rate": 1.836756919354732e-05,
+      "loss": 1.1941,
+      "step": 5768
+    },
+    {
+      "epoch": 2.147828413863713,
+      "grad_norm": 0.1619480699300766,
+      "learning_rate": 1.8366903655507987e-05,
+      "loss": 1.1958,
+      "step": 5769
+    },
+    {
+      "epoch": 2.148200719014322,
+      "grad_norm": 0.16295601427555084,
+      "learning_rate": 1.836623799389001e-05,
+      "loss": 1.204,
+      "step": 5770
+    },
+    {
+      "epoch": 2.1485730241649312,
+      "grad_norm": 0.1661759316921234,
+      "learning_rate": 1.8365572208703225e-05,
+      "loss": 1.2004,
+      "step": 5771
+    },
+    {
+      "epoch": 2.14894532931554,
+      "grad_norm": 0.1644112765789032,
+      "learning_rate": 1.8364906299957464e-05,
+      "loss": 1.1852,
+      "step": 5772
+    },
+    {
+      "epoch": 2.1493176344661493,
+      "grad_norm": 0.16289396584033966,
+      "learning_rate": 1.836424026766256e-05,
+      "loss": 1.1738,
+      "step": 5773
+    },
+    {
+      "epoch": 2.1496899396167586,
+      "grad_norm": 0.16191107034683228,
+      "learning_rate": 1.836357411182835e-05,
+      "loss": 1.1809,
+      "step": 5774
+    },
+    {
+      "epoch": 2.1500622447673674,
+      "grad_norm": 0.16520683467388153,
+      "learning_rate": 1.8362907832464678e-05,
+      "loss": 1.179,
+      "step": 5775
+    },
+    {
+      "epoch": 2.1504345499179767,
+      "grad_norm": 0.16534850001335144,
+      "learning_rate": 1.8362241429581386e-05,
+      "loss": 1.1973,
+      "step": 5776
+    },
+    {
+      "epoch": 2.1508068550685855,
+      "grad_norm": 0.16141214966773987,
+      "learning_rate": 1.8361574903188307e-05,
+      "loss": 1.1926,
+      "step": 5777
+    },
+    {
+      "epoch": 2.1511791602191948,
+      "grad_norm": 0.15976232290267944,
+      "learning_rate": 1.8360908253295293e-05,
+      "loss": 1.1813,
+      "step": 5778
+    },
+    {
+      "epoch": 2.1515514653698036,
+      "grad_norm": 0.16501028835773468,
+      "learning_rate": 1.836024147991219e-05,
+      "loss": 1.1883,
+      "step": 5779
+    },
+    {
+      "epoch": 2.151923770520413,
+      "grad_norm": 0.16739144921302795,
+      "learning_rate": 1.8359574583048846e-05,
+      "loss": 1.1884,
+      "step": 5780
+    },
+    {
+      "epoch": 2.1522960756710217,
+      "grad_norm": 0.1655547022819519,
+      "learning_rate": 1.8358907562715104e-05,
+      "loss": 1.2032,
+      "step": 5781
+    },
+    {
+      "epoch": 2.152668380821631,
+      "grad_norm": 0.16336333751678467,
+      "learning_rate": 1.835824041892083e-05,
+      "loss": 1.1947,
+      "step": 5782
+    },
+    {
+      "epoch": 2.15304068597224,
+      "grad_norm": 0.16573452949523926,
+      "learning_rate": 1.8357573151675864e-05,
+      "loss": 1.1751,
+      "step": 5783
+    },
+    {
+      "epoch": 2.153412991122849,
+      "grad_norm": 0.1619141399860382,
+      "learning_rate": 1.835690576099007e-05,
+      "loss": 1.1685,
+      "step": 5784
+    },
+    {
+      "epoch": 2.1537852962734583,
+      "grad_norm": 0.16914691030979156,
+      "learning_rate": 1.8356238246873302e-05,
+      "loss": 1.1747,
+      "step": 5785
+    },
+    {
+      "epoch": 2.154157601424067,
+      "grad_norm": 0.16027076542377472,
+      "learning_rate": 1.8355570609335416e-05,
+      "loss": 1.1737,
+      "step": 5786
+    },
+    {
+      "epoch": 2.1545299065746764,
+      "grad_norm": 0.15901793539524078,
+      "learning_rate": 1.835490284838628e-05,
+      "loss": 1.1896,
+      "step": 5787
+    },
+    {
+      "epoch": 2.154902211725285,
+      "grad_norm": 0.17148235440254211,
+      "learning_rate": 1.8354234964035754e-05,
+      "loss": 1.192,
+      "step": 5788
+    },
+    {
+      "epoch": 2.1552745168758944,
+      "grad_norm": 0.16743379831314087,
+      "learning_rate": 1.83535669562937e-05,
+      "loss": 1.1747,
+      "step": 5789
+    },
+    {
+      "epoch": 2.1556468220265037,
+      "grad_norm": 0.1674523502588272,
+      "learning_rate": 1.8352898825169986e-05,
+      "loss": 1.1853,
+      "step": 5790
+    },
+    {
+      "epoch": 2.1560191271771125,
+      "grad_norm": 0.15756765007972717,
+      "learning_rate": 1.835223057067448e-05,
+      "loss": 1.1702,
+      "step": 5791
+    },
+    {
+      "epoch": 2.156391432327722,
+      "grad_norm": 0.16530895233154297,
+      "learning_rate": 1.8351562192817054e-05,
+      "loss": 1.1814,
+      "step": 5792
+    },
+    {
+      "epoch": 2.1567637374783306,
+      "grad_norm": 0.1682775914669037,
+      "learning_rate": 1.835089369160758e-05,
+      "loss": 1.1956,
+      "step": 5793
+    },
+    {
+      "epoch": 2.15713604262894,
+      "grad_norm": 0.1804119050502777,
+      "learning_rate": 1.8350225067055927e-05,
+      "loss": 1.175,
+      "step": 5794
+    },
+    {
+      "epoch": 2.1575083477795487,
+      "grad_norm": 0.1660710573196411,
+      "learning_rate": 1.8349556319171977e-05,
+      "loss": 1.1795,
+      "step": 5795
+    },
+    {
+      "epoch": 2.157880652930158,
+      "grad_norm": 0.16514445841312408,
+      "learning_rate": 1.83488874479656e-05,
+      "loss": 1.1923,
+      "step": 5796
+    },
+    {
+      "epoch": 2.158252958080767,
+      "grad_norm": 0.1635090708732605,
+      "learning_rate": 1.8348218453446685e-05,
+      "loss": 1.1853,
+      "step": 5797
+    },
+    {
+      "epoch": 2.158625263231376,
+      "grad_norm": 0.17032501101493835,
+      "learning_rate": 1.8347549335625104e-05,
+      "loss": 1.1996,
+      "step": 5798
+    },
+    {
+      "epoch": 2.1589975683819853,
+      "grad_norm": 0.17020241916179657,
+      "learning_rate": 1.834688009451074e-05,
+      "loss": 1.1841,
+      "step": 5799
+    },
+    {
+      "epoch": 2.159369873532594,
+      "grad_norm": 0.1642071157693863,
+      "learning_rate": 1.8346210730113484e-05,
+      "loss": 1.195,
+      "step": 5800
+    },
+    {
+      "epoch": 2.1597421786832034,
+      "grad_norm": 0.16940705478191376,
+      "learning_rate": 1.8345541242443223e-05,
+      "loss": 1.1909,
+      "step": 5801
+    },
+    {
+      "epoch": 2.160114483833812,
+      "grad_norm": 0.15778762102127075,
+      "learning_rate": 1.8344871631509837e-05,
+      "loss": 1.1872,
+      "step": 5802
+    },
+    {
+      "epoch": 2.1604867889844215,
+      "grad_norm": 0.1702834963798523,
+      "learning_rate": 1.834420189732322e-05,
+      "loss": 1.1847,
+      "step": 5803
+    },
+    {
+      "epoch": 2.1608590941350303,
+      "grad_norm": 0.16944074630737305,
+      "learning_rate": 1.834353203989327e-05,
+      "loss": 1.1858,
+      "step": 5804
+    },
+    {
+      "epoch": 2.1612313992856396,
+      "grad_norm": 0.17070455849170685,
+      "learning_rate": 1.834286205922987e-05,
+      "loss": 1.1943,
+      "step": 5805
+    },
+    {
+      "epoch": 2.1616037044362484,
+      "grad_norm": 0.16586799919605255,
+      "learning_rate": 1.8342191955342926e-05,
+      "loss": 1.1877,
+      "step": 5806
+    },
+    {
+      "epoch": 2.1619760095868576,
+      "grad_norm": 0.17121802270412445,
+      "learning_rate": 1.8341521728242324e-05,
+      "loss": 1.1779,
+      "step": 5807
+    },
+    {
+      "epoch": 2.162348314737467,
+      "grad_norm": 0.16471904516220093,
+      "learning_rate": 1.8340851377937975e-05,
+      "loss": 1.1963,
+      "step": 5808
+    },
+    {
+      "epoch": 2.1627206198880757,
+      "grad_norm": 0.16698746383190155,
+      "learning_rate": 1.834018090443977e-05,
+      "loss": 1.1695,
+      "step": 5809
+    },
+    {
+      "epoch": 2.163092925038685,
+      "grad_norm": 0.16752393543720245,
+      "learning_rate": 1.833951030775762e-05,
+      "loss": 1.1872,
+      "step": 5810
+    },
+    {
+      "epoch": 2.163465230189294,
+      "grad_norm": 0.16329661011695862,
+      "learning_rate": 1.8338839587901426e-05,
+      "loss": 1.1855,
+      "step": 5811
+    },
+    {
+      "epoch": 2.163837535339903,
+      "grad_norm": 0.16880545020103455,
+      "learning_rate": 1.833816874488109e-05,
+      "loss": 1.1856,
+      "step": 5812
+    },
+    {
+      "epoch": 2.164209840490512,
+      "grad_norm": 0.16855594515800476,
+      "learning_rate": 1.833749777870653e-05,
+      "loss": 1.1756,
+      "step": 5813
+    },
+    {
+      "epoch": 2.164582145641121,
+      "grad_norm": 0.16726450622081757,
+      "learning_rate": 1.833682668938765e-05,
+      "loss": 1.1606,
+      "step": 5814
+    },
+    {
+      "epoch": 2.16495445079173,
+      "grad_norm": 0.16840289533138275,
+      "learning_rate": 1.8336155476934365e-05,
+      "loss": 1.1811,
+      "step": 5815
+    },
+    {
+      "epoch": 2.1653267559423393,
+      "grad_norm": 0.16863827407360077,
+      "learning_rate": 1.8335484141356582e-05,
+      "loss": 1.1907,
+      "step": 5816
+    },
+    {
+      "epoch": 2.1656990610929485,
+      "grad_norm": 0.16578540205955505,
+      "learning_rate": 1.8334812682664224e-05,
+      "loss": 1.1792,
+      "step": 5817
+    },
+    {
+      "epoch": 2.1660713662435573,
+      "grad_norm": 0.16219614446163177,
+      "learning_rate": 1.8334141100867208e-05,
+      "loss": 1.1759,
+      "step": 5818
+    },
+    {
+      "epoch": 2.1664436713941666,
+      "grad_norm": 0.16222938895225525,
+      "learning_rate": 1.8333469395975446e-05,
+      "loss": 1.1734,
+      "step": 5819
+    },
+    {
+      "epoch": 2.1668159765447754,
+      "grad_norm": 0.17056572437286377,
+      "learning_rate": 1.8332797567998865e-05,
+      "loss": 1.1854,
+      "step": 5820
+    },
+    {
+      "epoch": 2.1671882816953847,
+      "grad_norm": 0.1574883610010147,
+      "learning_rate": 1.833212561694739e-05,
+      "loss": 1.1951,
+      "step": 5821
+    },
+    {
+      "epoch": 2.1675605868459935,
+      "grad_norm": 0.1614379733800888,
+      "learning_rate": 1.833145354283094e-05,
+      "loss": 1.1806,
+      "step": 5822
+    },
+    {
+      "epoch": 2.1679328919966028,
+      "grad_norm": 0.17167092859745026,
+      "learning_rate": 1.8330781345659447e-05,
+      "loss": 1.1928,
+      "step": 5823
+    },
+    {
+      "epoch": 2.1683051971472116,
+      "grad_norm": 0.17068855464458466,
+      "learning_rate": 1.8330109025442834e-05,
+      "loss": 1.184,
+      "step": 5824
+    },
+    {
+      "epoch": 2.168677502297821,
+      "grad_norm": 0.1654396653175354,
+      "learning_rate": 1.832943658219103e-05,
+      "loss": 1.1839,
+      "step": 5825
+    },
+    {
+      "epoch": 2.16904980744843,
+      "grad_norm": 0.16501565277576447,
+      "learning_rate": 1.832876401591398e-05,
+      "loss": 1.1828,
+      "step": 5826
+    },
+    {
+      "epoch": 2.169422112599039,
+      "grad_norm": 0.16627371311187744,
+      "learning_rate": 1.8328091326621597e-05,
+      "loss": 1.1904,
+      "step": 5827
+    },
+    {
+      "epoch": 2.169794417749648,
+      "grad_norm": 0.15978330373764038,
+      "learning_rate": 1.8327418514323833e-05,
+      "loss": 1.1706,
+      "step": 5828
+    },
+    {
+      "epoch": 2.170166722900257,
+      "grad_norm": 0.16594457626342773,
+      "learning_rate": 1.8326745579030623e-05,
+      "loss": 1.2023,
+      "step": 5829
+    },
+    {
+      "epoch": 2.1705390280508663,
+      "grad_norm": 0.17066100239753723,
+      "learning_rate": 1.83260725207519e-05,
+      "loss": 1.1809,
+      "step": 5830
+    },
+    {
+      "epoch": 2.170911333201475,
+      "grad_norm": 0.16543124616146088,
+      "learning_rate": 1.8325399339497608e-05,
+      "loss": 1.1749,
+      "step": 5831
+    },
+    {
+      "epoch": 2.1712836383520844,
+      "grad_norm": 0.16617171466350555,
+      "learning_rate": 1.8324726035277694e-05,
+      "loss": 1.1901,
+      "step": 5832
+    },
+    {
+      "epoch": 2.171655943502693,
+      "grad_norm": 0.17173443734645844,
+      "learning_rate": 1.8324052608102095e-05,
+      "loss": 1.1847,
+      "step": 5833
+    },
+    {
+      "epoch": 2.1720282486533025,
+      "grad_norm": 0.1652347892522812,
+      "learning_rate": 1.832337905798076e-05,
+      "loss": 1.1766,
+      "step": 5834
+    },
+    {
+      "epoch": 2.1724005538039117,
+      "grad_norm": 0.16114427149295807,
+      "learning_rate": 1.8322705384923644e-05,
+      "loss": 1.1835,
+      "step": 5835
+    },
+    {
+      "epoch": 2.1727728589545205,
+      "grad_norm": 0.17017604410648346,
+      "learning_rate": 1.8322031588940687e-05,
+      "loss": 1.1778,
+      "step": 5836
+    },
+    {
+      "epoch": 2.17314516410513,
+      "grad_norm": 0.16408562660217285,
+      "learning_rate": 1.8321357670041848e-05,
+      "loss": 1.1863,
+      "step": 5837
+    },
+    {
+      "epoch": 2.1735174692557386,
+      "grad_norm": 0.1657814234495163,
+      "learning_rate": 1.832068362823708e-05,
+      "loss": 1.1847,
+      "step": 5838
+    },
+    {
+      "epoch": 2.173889774406348,
+      "grad_norm": 0.15966399013996124,
+      "learning_rate": 1.8320009463536333e-05,
+      "loss": 1.1818,
+      "step": 5839
+    },
+    {
+      "epoch": 2.1742620795569567,
+      "grad_norm": 0.15668559074401855,
+      "learning_rate": 1.831933517594957e-05,
+      "loss": 1.1798,
+      "step": 5840
+    },
+    {
+      "epoch": 2.174634384707566,
+      "grad_norm": 0.1669110357761383,
+      "learning_rate": 1.831866076548675e-05,
+      "loss": 1.2053,
+      "step": 5841
+    },
+    {
+      "epoch": 2.175006689858175,
+      "grad_norm": 0.15878044068813324,
+      "learning_rate": 1.831798623215783e-05,
+      "loss": 1.1697,
+      "step": 5842
+    },
+    {
+      "epoch": 2.175378995008784,
+      "grad_norm": 0.16765941679477692,
+      "learning_rate": 1.8317311575972777e-05,
+      "loss": 1.1806,
+      "step": 5843
+    },
+    {
+      "epoch": 2.1757513001593933,
+      "grad_norm": 0.1708860546350479,
+      "learning_rate": 1.8316636796941555e-05,
+      "loss": 1.2002,
+      "step": 5844
+    },
+    {
+      "epoch": 2.176123605310002,
+      "grad_norm": 0.16999541223049164,
+      "learning_rate": 1.8315961895074127e-05,
+      "loss": 1.1778,
+      "step": 5845
+    },
+    {
+      "epoch": 2.1764959104606114,
+      "grad_norm": 0.1659235954284668,
+      "learning_rate": 1.8315286870380468e-05,
+      "loss": 1.1821,
+      "step": 5846
+    },
+    {
+      "epoch": 2.1768682156112202,
+      "grad_norm": 0.15911462903022766,
+      "learning_rate": 1.831461172287054e-05,
+      "loss": 1.197,
+      "step": 5847
+    },
+    {
+      "epoch": 2.1772405207618295,
+      "grad_norm": 0.15904010832309723,
+      "learning_rate": 1.8313936452554318e-05,
+      "loss": 1.1777,
+      "step": 5848
+    },
+    {
+      "epoch": 2.1776128259124383,
+      "grad_norm": 0.1634838581085205,
+      "learning_rate": 1.8313261059441777e-05,
+      "loss": 1.1752,
+      "step": 5849
+    },
+    {
+      "epoch": 2.1779851310630476,
+      "grad_norm": 0.16560928523540497,
+      "learning_rate": 1.8312585543542893e-05,
+      "loss": 1.1859,
+      "step": 5850
+    },
+    {
+      "epoch": 2.1783574362136564,
+      "grad_norm": 0.16026043891906738,
+      "learning_rate": 1.8311909904867643e-05,
+      "loss": 1.1715,
+      "step": 5851
+    },
+    {
+      "epoch": 2.1787297413642657,
+      "grad_norm": 0.1598140001296997,
+      "learning_rate": 1.8311234143426003e-05,
+      "loss": 1.1755,
+      "step": 5852
+    },
+    {
+      "epoch": 2.179102046514875,
+      "grad_norm": 0.1634136587381363,
+      "learning_rate": 1.831055825922796e-05,
+      "loss": 1.1806,
+      "step": 5853
+    },
+    {
+      "epoch": 2.1794743516654838,
+      "grad_norm": 0.16830392181873322,
+      "learning_rate": 1.830988225228349e-05,
+      "loss": 1.1638,
+      "step": 5854
+    },
+    {
+      "epoch": 2.179846656816093,
+      "grad_norm": 0.16452953219413757,
+      "learning_rate": 1.8309206122602582e-05,
+      "loss": 1.184,
+      "step": 5855
+    },
+    {
+      "epoch": 2.180218961966702,
+      "grad_norm": 0.16308681666851044,
+      "learning_rate": 1.830852987019522e-05,
+      "loss": 1.1898,
+      "step": 5856
+    },
+    {
+      "epoch": 2.180591267117311,
+      "grad_norm": 0.1646089106798172,
+      "learning_rate": 1.8307853495071394e-05,
+      "loss": 1.1717,
+      "step": 5857
+    },
+    {
+      "epoch": 2.18096357226792,
+      "grad_norm": 0.16182439029216766,
+      "learning_rate": 1.8307176997241092e-05,
+      "loss": 1.1833,
+      "step": 5858
+    },
+    {
+      "epoch": 2.181335877418529,
+      "grad_norm": 0.16360101103782654,
+      "learning_rate": 1.8306500376714307e-05,
+      "loss": 1.1876,
+      "step": 5859
+    },
+    {
+      "epoch": 2.181708182569138,
+      "grad_norm": 0.1654779016971588,
+      "learning_rate": 1.8305823633501033e-05,
+      "loss": 1.1686,
+      "step": 5860
+    },
+    {
+      "epoch": 2.1820804877197473,
+      "grad_norm": 0.16477808356285095,
+      "learning_rate": 1.8305146767611267e-05,
+      "loss": 1.1878,
+      "step": 5861
+    },
+    {
+      "epoch": 2.1824527928703565,
+      "grad_norm": 0.16140249371528625,
+      "learning_rate": 1.8304469779055e-05,
+      "loss": 1.1852,
+      "step": 5862
+    },
+    {
+      "epoch": 2.1828250980209654,
+      "grad_norm": 0.1675121784210205,
+      "learning_rate": 1.830379266784224e-05,
+      "loss": 1.1675,
+      "step": 5863
+    },
+    {
+      "epoch": 2.1831974031715746,
+      "grad_norm": 0.1668505072593689,
+      "learning_rate": 1.8303115433982984e-05,
+      "loss": 1.1938,
+      "step": 5864
+    },
+    {
+      "epoch": 2.1835697083221834,
+      "grad_norm": 0.16749152541160583,
+      "learning_rate": 1.830243807748723e-05,
+      "loss": 1.1722,
+      "step": 5865
+    },
+    {
+      "epoch": 2.1839420134727927,
+      "grad_norm": 0.16352511942386627,
+      "learning_rate": 1.830176059836499e-05,
+      "loss": 1.1873,
+      "step": 5866
+    },
+    {
+      "epoch": 2.1843143186234015,
+      "grad_norm": 0.16892147064208984,
+      "learning_rate": 1.8301082996626268e-05,
+      "loss": 1.1968,
+      "step": 5867
+    },
+    {
+      "epoch": 2.184686623774011,
+      "grad_norm": 0.1627473384141922,
+      "learning_rate": 1.830040527228107e-05,
+      "loss": 1.185,
+      "step": 5868
+    },
+    {
+      "epoch": 2.1850589289246196,
+      "grad_norm": 0.16038550436496735,
+      "learning_rate": 1.8299727425339405e-05,
+      "loss": 1.1742,
+      "step": 5869
+    },
+    {
+      "epoch": 2.185431234075229,
+      "grad_norm": 0.1608399599790573,
+      "learning_rate": 1.8299049455811285e-05,
+      "loss": 1.1837,
+      "step": 5870
+    },
+    {
+      "epoch": 2.185803539225838,
+      "grad_norm": 0.16327451169490814,
+      "learning_rate": 1.829837136370673e-05,
+      "loss": 1.1792,
+      "step": 5871
+    },
+    {
+      "epoch": 2.186175844376447,
+      "grad_norm": 0.16703878343105316,
+      "learning_rate": 1.829769314903575e-05,
+      "loss": 1.2048,
+      "step": 5872
+    },
+    {
+      "epoch": 2.1865481495270562,
+      "grad_norm": 0.16750751435756683,
+      "learning_rate": 1.8297014811808363e-05,
+      "loss": 1.1813,
+      "step": 5873
+    },
+    {
+      "epoch": 2.186920454677665,
+      "grad_norm": 0.16518592834472656,
+      "learning_rate": 1.8296336352034585e-05,
+      "loss": 1.1782,
+      "step": 5874
+    },
+    {
+      "epoch": 2.1872927598282743,
+      "grad_norm": 0.1652197390794754,
+      "learning_rate": 1.829565776972444e-05,
+      "loss": 1.1771,
+      "step": 5875
+    },
+    {
+      "epoch": 2.187665064978883,
+      "grad_norm": 0.16577941179275513,
+      "learning_rate": 1.829497906488795e-05,
+      "loss": 1.1896,
+      "step": 5876
+    },
+    {
+      "epoch": 2.1880373701294924,
+      "grad_norm": 0.16348230838775635,
+      "learning_rate": 1.8294300237535145e-05,
+      "loss": 1.185,
+      "step": 5877
+    },
+    {
+      "epoch": 2.188409675280101,
+      "grad_norm": 0.16359402239322662,
+      "learning_rate": 1.8293621287676043e-05,
+      "loss": 1.1909,
+      "step": 5878
+    },
+    {
+      "epoch": 2.1887819804307105,
+      "grad_norm": 0.16968275606632233,
+      "learning_rate": 1.8292942215320675e-05,
+      "loss": 1.1827,
+      "step": 5879
+    },
+    {
+      "epoch": 2.1891542855813197,
+      "grad_norm": 0.16383975744247437,
+      "learning_rate": 1.829226302047907e-05,
+      "loss": 1.1892,
+      "step": 5880
+    },
+    {
+      "epoch": 2.1895265907319286,
+      "grad_norm": 0.16417528688907623,
+      "learning_rate": 1.8291583703161263e-05,
+      "loss": 1.1879,
+      "step": 5881
+    },
+    {
+      "epoch": 2.189898895882538,
+      "grad_norm": 0.16152073442935944,
+      "learning_rate": 1.8290904263377284e-05,
+      "loss": 1.1755,
+      "step": 5882
+    },
+    {
+      "epoch": 2.1902712010331467,
+      "grad_norm": 0.16532781720161438,
+      "learning_rate": 1.8290224701137165e-05,
+      "loss": 1.1987,
+      "step": 5883
+    },
+    {
+      "epoch": 2.190643506183756,
+      "grad_norm": 0.1655869036912918,
+      "learning_rate": 1.8289545016450953e-05,
+      "loss": 1.187,
+      "step": 5884
+    },
+    {
+      "epoch": 2.1910158113343647,
+      "grad_norm": 0.16325949132442474,
+      "learning_rate": 1.828886520932868e-05,
+      "loss": 1.1926,
+      "step": 5885
+    },
+    {
+      "epoch": 2.191388116484974,
+      "grad_norm": 0.1616744101047516,
+      "learning_rate": 1.8288185279780388e-05,
+      "loss": 1.1851,
+      "step": 5886
+    },
+    {
+      "epoch": 2.1917604216355833,
+      "grad_norm": 0.15696094930171967,
+      "learning_rate": 1.828750522781612e-05,
+      "loss": 1.1899,
+      "step": 5887
+    },
+    {
+      "epoch": 2.192132726786192,
+      "grad_norm": 0.15940669178962708,
+      "learning_rate": 1.8286825053445916e-05,
+      "loss": 1.1872,
+      "step": 5888
+    },
+    {
+      "epoch": 2.1925050319368014,
+      "grad_norm": 0.16422483325004578,
+      "learning_rate": 1.828614475667983e-05,
+      "loss": 1.1823,
+      "step": 5889
+    },
+    {
+      "epoch": 2.19287733708741,
+      "grad_norm": 0.16130252182483673,
+      "learning_rate": 1.8285464337527906e-05,
+      "loss": 1.1784,
+      "step": 5890
+    },
+    {
+      "epoch": 2.1932496422380194,
+      "grad_norm": 0.1615862399339676,
+      "learning_rate": 1.8284783796000193e-05,
+      "loss": 1.185,
+      "step": 5891
+    },
+    {
+      "epoch": 2.1936219473886283,
+      "grad_norm": 0.15964102745056152,
+      "learning_rate": 1.8284103132106743e-05,
+      "loss": 1.1796,
+      "step": 5892
+    },
+    {
+      "epoch": 2.1939942525392375,
+      "grad_norm": 0.15634554624557495,
+      "learning_rate": 1.828342234585761e-05,
+      "loss": 1.1821,
+      "step": 5893
+    },
+    {
+      "epoch": 2.1943665576898463,
+      "grad_norm": 0.1648419350385666,
+      "learning_rate": 1.828274143726285e-05,
+      "loss": 1.1926,
+      "step": 5894
+    },
+    {
+      "epoch": 2.1947388628404556,
+      "grad_norm": 0.1677112877368927,
+      "learning_rate": 1.8282060406332513e-05,
+      "loss": 1.1815,
+      "step": 5895
+    },
+    {
+      "epoch": 2.195111167991065,
+      "grad_norm": 0.16209574043750763,
+      "learning_rate": 1.828137925307667e-05,
+      "loss": 1.1888,
+      "step": 5896
+    },
+    {
+      "epoch": 2.1954834731416737,
+      "grad_norm": 0.16768576204776764,
+      "learning_rate": 1.828069797750537e-05,
+      "loss": 1.1844,
+      "step": 5897
+    },
+    {
+      "epoch": 2.195855778292283,
+      "grad_norm": 0.15827785432338715,
+      "learning_rate": 1.8280016579628686e-05,
+      "loss": 1.1645,
+      "step": 5898
+    },
+    {
+      "epoch": 2.1962280834428918,
+      "grad_norm": 0.17181847989559174,
+      "learning_rate": 1.8279335059456673e-05,
+      "loss": 1.1777,
+      "step": 5899
+    },
+    {
+      "epoch": 2.196600388593501,
+      "grad_norm": 0.16436168551445007,
+      "learning_rate": 1.8278653416999402e-05,
+      "loss": 1.1708,
+      "step": 5900
+    },
+    {
+      "epoch": 2.19697269374411,
+      "grad_norm": 0.16275320947170258,
+      "learning_rate": 1.827797165226694e-05,
+      "loss": 1.1883,
+      "step": 5901
+    },
+    {
+      "epoch": 2.197344998894719,
+      "grad_norm": 0.15981367230415344,
+      "learning_rate": 1.8277289765269353e-05,
+      "loss": 1.1709,
+      "step": 5902
+    },
+    {
+      "epoch": 2.197717304045328,
+      "grad_norm": 0.17126475274562836,
+      "learning_rate": 1.8276607756016722e-05,
+      "loss": 1.1784,
+      "step": 5903
+    },
+    {
+      "epoch": 2.198089609195937,
+      "grad_norm": 0.16240163147449493,
+      "learning_rate": 1.827592562451911e-05,
+      "loss": 1.1772,
+      "step": 5904
+    },
+    {
+      "epoch": 2.1984619143465465,
+      "grad_norm": 0.16826151311397552,
+      "learning_rate": 1.8275243370786594e-05,
+      "loss": 1.1728,
+      "step": 5905
+    },
+    {
+      "epoch": 2.1988342194971553,
+      "grad_norm": 0.1712343990802765,
+      "learning_rate": 1.8274560994829256e-05,
+      "loss": 1.1836,
+      "step": 5906
+    },
+    {
+      "epoch": 2.1992065246477646,
+      "grad_norm": 0.166199192404747,
+      "learning_rate": 1.827387849665717e-05,
+      "loss": 1.1996,
+      "step": 5907
+    },
+    {
+      "epoch": 2.1995788297983734,
+      "grad_norm": 0.16163288056850433,
+      "learning_rate": 1.827319587628042e-05,
+      "loss": 1.1775,
+      "step": 5908
+    },
+    {
+      "epoch": 2.1999511349489826,
+      "grad_norm": 0.171270489692688,
+      "learning_rate": 1.827251313370908e-05,
+      "loss": 1.1866,
+      "step": 5909
+    },
+    {
+      "epoch": 2.2003234400995915,
+      "grad_norm": 0.1645098328590393,
+      "learning_rate": 1.8271830268953248e-05,
+      "loss": 1.1814,
+      "step": 5910
+    },
+    {
+      "epoch": 2.2006957452502007,
+      "grad_norm": 0.1669131964445114,
+      "learning_rate": 1.8271147282022998e-05,
+      "loss": 1.1928,
+      "step": 5911
+    },
+    {
+      "epoch": 2.20106805040081,
+      "grad_norm": 0.17018841207027435,
+      "learning_rate": 1.8270464172928423e-05,
+      "loss": 1.1825,
+      "step": 5912
+    },
+    {
+      "epoch": 2.201440355551419,
+      "grad_norm": 0.167917862534523,
+      "learning_rate": 1.826978094167961e-05,
+      "loss": 1.1882,
+      "step": 5913
+    },
+    {
+      "epoch": 2.201812660702028,
+      "grad_norm": 0.16571766138076782,
+      "learning_rate": 1.826909758828665e-05,
+      "loss": 1.1726,
+      "step": 5914
+    },
+    {
+      "epoch": 2.202184965852637,
+      "grad_norm": 0.16511200368404388,
+      "learning_rate": 1.826841411275964e-05,
+      "loss": 1.1786,
+      "step": 5915
+    },
+    {
+      "epoch": 2.202557271003246,
+      "grad_norm": 0.16730502247810364,
+      "learning_rate": 1.8267730515108674e-05,
+      "loss": 1.1868,
+      "step": 5916
+    },
+    {
+      "epoch": 2.202929576153855,
+      "grad_norm": 0.16795329749584198,
+      "learning_rate": 1.8267046795343845e-05,
+      "loss": 1.1757,
+      "step": 5917
+    },
+    {
+      "epoch": 2.2033018813044642,
+      "grad_norm": 0.17142654955387115,
+      "learning_rate": 1.8266362953475252e-05,
+      "loss": 1.1918,
+      "step": 5918
+    },
+    {
+      "epoch": 2.203674186455073,
+      "grad_norm": 0.1681024432182312,
+      "learning_rate": 1.8265678989513e-05,
+      "loss": 1.1676,
+      "step": 5919
+    },
+    {
+      "epoch": 2.2040464916056823,
+      "grad_norm": 0.1688128411769867,
+      "learning_rate": 1.8264994903467187e-05,
+      "loss": 1.1888,
+      "step": 5920
+    },
+    {
+      "epoch": 2.2044187967562916,
+      "grad_norm": 0.16253378987312317,
+      "learning_rate": 1.8264310695347918e-05,
+      "loss": 1.1793,
+      "step": 5921
+    },
+    {
+      "epoch": 2.2047911019069004,
+      "grad_norm": 0.16472946107387543,
+      "learning_rate": 1.8263626365165296e-05,
+      "loss": 1.1855,
+      "step": 5922
+    },
+    {
+      "epoch": 2.2051634070575097,
+      "grad_norm": 0.1619625687599182,
+      "learning_rate": 1.8262941912929434e-05,
+      "loss": 1.1799,
+      "step": 5923
+    },
+    {
+      "epoch": 2.2055357122081185,
+      "grad_norm": 0.16878865659236908,
+      "learning_rate": 1.826225733865044e-05,
+      "loss": 1.1956,
+      "step": 5924
+    },
+    {
+      "epoch": 2.2059080173587278,
+      "grad_norm": 0.16465233266353607,
+      "learning_rate": 1.8261572642338418e-05,
+      "loss": 1.1848,
+      "step": 5925
+    },
+    {
+      "epoch": 2.2062803225093366,
+      "grad_norm": 0.1674007624387741,
+      "learning_rate": 1.826088782400349e-05,
+      "loss": 1.1781,
+      "step": 5926
+    },
+    {
+      "epoch": 2.206652627659946,
+      "grad_norm": 0.16674551367759705,
+      "learning_rate": 1.8260202883655773e-05,
+      "loss": 1.1907,
+      "step": 5927
+    },
+    {
+      "epoch": 2.2070249328105547,
+      "grad_norm": 0.16409821808338165,
+      "learning_rate": 1.825951782130537e-05,
+      "loss": 1.1699,
+      "step": 5928
+    },
+    {
+      "epoch": 2.207397237961164,
+      "grad_norm": 0.16593073308467865,
+      "learning_rate": 1.825883263696241e-05,
+      "loss": 1.1953,
+      "step": 5929
+    },
+    {
+      "epoch": 2.207769543111773,
+      "grad_norm": 0.16026705503463745,
+      "learning_rate": 1.825814733063701e-05,
+      "loss": 1.1621,
+      "step": 5930
+    },
+    {
+      "epoch": 2.208141848262382,
+      "grad_norm": 0.16227182745933533,
+      "learning_rate": 1.825746190233929e-05,
+      "loss": 1.1973,
+      "step": 5931
+    },
+    {
+      "epoch": 2.2085141534129913,
+      "grad_norm": 0.16535073518753052,
+      "learning_rate": 1.8256776352079377e-05,
+      "loss": 1.1731,
+      "step": 5932
+    },
+    {
+      "epoch": 2.2088864585636,
+      "grad_norm": 0.16271446645259857,
+      "learning_rate": 1.82560906798674e-05,
+      "loss": 1.1718,
+      "step": 5933
+    },
+    {
+      "epoch": 2.2092587637142094,
+      "grad_norm": 0.16261382400989532,
+      "learning_rate": 1.8255404885713478e-05,
+      "loss": 1.1795,
+      "step": 5934
+    },
+    {
+      "epoch": 2.209631068864818,
+      "grad_norm": 0.16760127246379852,
+      "learning_rate": 1.825471896962774e-05,
+      "loss": 1.1688,
+      "step": 5935
+    },
+    {
+      "epoch": 2.2100033740154275,
+      "grad_norm": 0.16549386084079742,
+      "learning_rate": 1.8254032931620326e-05,
+      "loss": 1.1906,
+      "step": 5936
+    },
+    {
+      "epoch": 2.2103756791660363,
+      "grad_norm": 0.16376779973506927,
+      "learning_rate": 1.8253346771701363e-05,
+      "loss": 1.1764,
+      "step": 5937
+    },
+    {
+      "epoch": 2.2107479843166455,
+      "grad_norm": 0.1606384664773941,
+      "learning_rate": 1.8252660489880986e-05,
+      "loss": 1.1735,
+      "step": 5938
+    },
+    {
+      "epoch": 2.211120289467255,
+      "grad_norm": 0.16291670501232147,
+      "learning_rate": 1.8251974086169332e-05,
+      "loss": 1.178,
+      "step": 5939
+    },
+    {
+      "epoch": 2.2114925946178636,
+      "grad_norm": 0.16411124169826508,
+      "learning_rate": 1.8251287560576535e-05,
+      "loss": 1.1845,
+      "step": 5940
+    },
+    {
+      "epoch": 2.211864899768473,
+      "grad_norm": 0.15855379402637482,
+      "learning_rate": 1.8250600913112743e-05,
+      "loss": 1.1883,
+      "step": 5941
+    },
+    {
+      "epoch": 2.2122372049190817,
+      "grad_norm": 0.15984077751636505,
+      "learning_rate": 1.824991414378809e-05,
+      "loss": 1.1848,
+      "step": 5942
+    },
+    {
+      "epoch": 2.212609510069691,
+      "grad_norm": 0.16627465188503265,
+      "learning_rate": 1.8249227252612725e-05,
+      "loss": 1.1914,
+      "step": 5943
+    },
+    {
+      "epoch": 2.2129818152203,
+      "grad_norm": 0.1621500700712204,
+      "learning_rate": 1.824854023959679e-05,
+      "loss": 1.1999,
+      "step": 5944
+    },
+    {
+      "epoch": 2.213354120370909,
+      "grad_norm": 0.1592169553041458,
+      "learning_rate": 1.8247853104750433e-05,
+      "loss": 1.1838,
+      "step": 5945
+    },
+    {
+      "epoch": 2.213726425521518,
+      "grad_norm": 0.16420015692710876,
+      "learning_rate": 1.8247165848083805e-05,
+      "loss": 1.1777,
+      "step": 5946
+    },
+    {
+      "epoch": 2.214098730672127,
+      "grad_norm": 0.16197781264781952,
+      "learning_rate": 1.8246478469607055e-05,
+      "loss": 1.1815,
+      "step": 5947
+    },
+    {
+      "epoch": 2.2144710358227364,
+      "grad_norm": 0.1656760275363922,
+      "learning_rate": 1.8245790969330336e-05,
+      "loss": 1.185,
+      "step": 5948
+    },
+    {
+      "epoch": 2.2148433409733452,
+      "grad_norm": 0.16433504223823547,
+      "learning_rate": 1.82451033472638e-05,
+      "loss": 1.1878,
+      "step": 5949
+    },
+    {
+      "epoch": 2.2152156461239545,
+      "grad_norm": 0.16674500703811646,
+      "learning_rate": 1.8244415603417603e-05,
+      "loss": 1.1901,
+      "step": 5950
+    },
+    {
+      "epoch": 2.2155879512745633,
+      "grad_norm": 0.16727973520755768,
+      "learning_rate": 1.824372773780191e-05,
+      "loss": 1.1683,
+      "step": 5951
+    },
+    {
+      "epoch": 2.2159602564251726,
+      "grad_norm": 0.16087013483047485,
+      "learning_rate": 1.8243039750426872e-05,
+      "loss": 1.1982,
+      "step": 5952
+    },
+    {
+      "epoch": 2.2163325615757814,
+      "grad_norm": 0.1639922559261322,
+      "learning_rate": 1.8242351641302657e-05,
+      "loss": 1.1971,
+      "step": 5953
+    },
+    {
+      "epoch": 2.2167048667263907,
+      "grad_norm": 0.16302375495433807,
+      "learning_rate": 1.8241663410439424e-05,
+      "loss": 1.1748,
+      "step": 5954
+    },
+    {
+      "epoch": 2.2170771718769995,
+      "grad_norm": 0.16024988889694214,
+      "learning_rate": 1.8240975057847338e-05,
+      "loss": 1.1837,
+      "step": 5955
+    },
+    {
+      "epoch": 2.2174494770276088,
+      "grad_norm": 0.16668042540550232,
+      "learning_rate": 1.824028658353657e-05,
+      "loss": 1.1922,
+      "step": 5956
+    },
+    {
+      "epoch": 2.217821782178218,
+      "grad_norm": 0.1660262942314148,
+      "learning_rate": 1.8239597987517284e-05,
+      "loss": 1.1947,
+      "step": 5957
+    },
+    {
+      "epoch": 2.218194087328827,
+      "grad_norm": 0.15916001796722412,
+      "learning_rate": 1.8238909269799655e-05,
+      "loss": 1.1774,
+      "step": 5958
+    },
+    {
+      "epoch": 2.218566392479436,
+      "grad_norm": 0.16157934069633484,
+      "learning_rate": 1.8238220430393855e-05,
+      "loss": 1.1705,
+      "step": 5959
+    },
+    {
+      "epoch": 2.218938697630045,
+      "grad_norm": 0.17225240170955658,
+      "learning_rate": 1.8237531469310054e-05,
+      "loss": 1.1772,
+      "step": 5960
+    },
+    {
+      "epoch": 2.219311002780654,
+      "grad_norm": 0.16769270598888397,
+      "learning_rate": 1.823684238655843e-05,
+      "loss": 1.1843,
+      "step": 5961
+    },
+    {
+      "epoch": 2.219683307931263,
+      "grad_norm": 0.18250252306461334,
+      "learning_rate": 1.8236153182149158e-05,
+      "loss": 1.1767,
+      "step": 5962
+    },
+    {
+      "epoch": 2.2200556130818723,
+      "grad_norm": 0.16508784890174866,
+      "learning_rate": 1.8235463856092423e-05,
+      "loss": 1.178,
+      "step": 5963
+    },
+    {
+      "epoch": 2.220427918232481,
+      "grad_norm": 0.16602812707424164,
+      "learning_rate": 1.8234774408398405e-05,
+      "loss": 1.1758,
+      "step": 5964
+    },
+    {
+      "epoch": 2.2208002233830904,
+      "grad_norm": 0.16944126784801483,
+      "learning_rate": 1.8234084839077283e-05,
+      "loss": 1.1842,
+      "step": 5965
+    },
+    {
+      "epoch": 2.2211725285336996,
+      "grad_norm": 0.1692361980676651,
+      "learning_rate": 1.8233395148139246e-05,
+      "loss": 1.2078,
+      "step": 5966
+    },
+    {
+      "epoch": 2.2215448336843084,
+      "grad_norm": 0.175007626414299,
+      "learning_rate": 1.823270533559448e-05,
+      "loss": 1.1778,
+      "step": 5967
+    },
+    {
+      "epoch": 2.2219171388349177,
+      "grad_norm": 0.16647957265377045,
+      "learning_rate": 1.823201540145317e-05,
+      "loss": 1.1834,
+      "step": 5968
+    },
+    {
+      "epoch": 2.2222894439855265,
+      "grad_norm": 0.16229864954948425,
+      "learning_rate": 1.8231325345725514e-05,
+      "loss": 1.1748,
+      "step": 5969
+    },
+    {
+      "epoch": 2.222661749136136,
+      "grad_norm": 0.16611957550048828,
+      "learning_rate": 1.8230635168421694e-05,
+      "loss": 1.1781,
+      "step": 5970
+    },
+    {
+      "epoch": 2.2230340542867446,
+      "grad_norm": 0.17674987018108368,
+      "learning_rate": 1.8229944869551915e-05,
+      "loss": 1.1974,
+      "step": 5971
+    },
+    {
+      "epoch": 2.223406359437354,
+      "grad_norm": 0.16097386181354523,
+      "learning_rate": 1.8229254449126365e-05,
+      "loss": 1.1807,
+      "step": 5972
+    },
+    {
+      "epoch": 2.2237786645879627,
+      "grad_norm": 0.16559158265590668,
+      "learning_rate": 1.822856390715524e-05,
+      "loss": 1.1776,
+      "step": 5973
+    },
+    {
+      "epoch": 2.224150969738572,
+      "grad_norm": 0.16879995167255402,
+      "learning_rate": 1.8227873243648748e-05,
+      "loss": 1.1829,
+      "step": 5974
+    },
+    {
+      "epoch": 2.2245232748891812,
+      "grad_norm": 0.1685383915901184,
+      "learning_rate": 1.8227182458617076e-05,
+      "loss": 1.1909,
+      "step": 5975
+    },
+    {
+      "epoch": 2.22489558003979,
+      "grad_norm": 0.16316437721252441,
+      "learning_rate": 1.822649155207044e-05,
+      "loss": 1.1669,
+      "step": 5976
+    },
+    {
+      "epoch": 2.2252678851903993,
+      "grad_norm": 0.18334704637527466,
+      "learning_rate": 1.822580052401904e-05,
+      "loss": 1.1755,
+      "step": 5977
+    },
+    {
+      "epoch": 2.225640190341008,
+      "grad_norm": 0.16785620152950287,
+      "learning_rate": 1.8225109374473087e-05,
+      "loss": 1.1818,
+      "step": 5978
+    },
+    {
+      "epoch": 2.2260124954916174,
+      "grad_norm": 0.16493113338947296,
+      "learning_rate": 1.822441810344278e-05,
+      "loss": 1.2018,
+      "step": 5979
+    },
+    {
+      "epoch": 2.226384800642226,
+      "grad_norm": 0.16159646213054657,
+      "learning_rate": 1.822372671093833e-05,
+      "loss": 1.1709,
+      "step": 5980
+    },
+    {
+      "epoch": 2.2267571057928355,
+      "grad_norm": 0.16819043457508087,
+      "learning_rate": 1.822303519696996e-05,
+      "loss": 1.1913,
+      "step": 5981
+    },
+    {
+      "epoch": 2.2271294109434443,
+      "grad_norm": 0.16642168164253235,
+      "learning_rate": 1.8222343561547876e-05,
+      "loss": 1.1787,
+      "step": 5982
+    },
+    {
+      "epoch": 2.2275017160940536,
+      "grad_norm": 0.16357922554016113,
+      "learning_rate": 1.8221651804682287e-05,
+      "loss": 1.1764,
+      "step": 5983
+    },
+    {
+      "epoch": 2.227874021244663,
+      "grad_norm": 0.16216666996479034,
+      "learning_rate": 1.8220959926383422e-05,
+      "loss": 1.18,
+      "step": 5984
+    },
+    {
+      "epoch": 2.2282463263952716,
+      "grad_norm": 0.16932249069213867,
+      "learning_rate": 1.8220267926661494e-05,
+      "loss": 1.1882,
+      "step": 5985
+    },
+    {
+      "epoch": 2.228618631545881,
+      "grad_norm": 0.16690555214881897,
+      "learning_rate": 1.8219575805526723e-05,
+      "loss": 1.1948,
+      "step": 5986
+    },
+    {
+      "epoch": 2.2289909366964897,
+      "grad_norm": 0.16043587028980255,
+      "learning_rate": 1.8218883562989335e-05,
+      "loss": 1.1779,
+      "step": 5987
+    },
+    {
+      "epoch": 2.229363241847099,
+      "grad_norm": 0.16727250814437866,
+      "learning_rate": 1.8218191199059553e-05,
+      "loss": 1.1728,
+      "step": 5988
+    },
+    {
+      "epoch": 2.229735546997708,
+      "grad_norm": 0.17071470618247986,
+      "learning_rate": 1.82174987137476e-05,
+      "loss": 1.1852,
+      "step": 5989
+    },
+    {
+      "epoch": 2.230107852148317,
+      "grad_norm": 0.17446862161159515,
+      "learning_rate": 1.8216806107063705e-05,
+      "loss": 1.1967,
+      "step": 5990
+    },
+    {
+      "epoch": 2.230480157298926,
+      "grad_norm": 0.16440711915493011,
+      "learning_rate": 1.8216113379018105e-05,
+      "loss": 1.1864,
+      "step": 5991
+    },
+    {
+      "epoch": 2.230852462449535,
+      "grad_norm": 0.16457001864910126,
+      "learning_rate": 1.8215420529621025e-05,
+      "loss": 1.1746,
+      "step": 5992
+    },
+    {
+      "epoch": 2.2312247676001444,
+      "grad_norm": 0.16013966500759125,
+      "learning_rate": 1.82147275588827e-05,
+      "loss": 1.1715,
+      "step": 5993
+    },
+    {
+      "epoch": 2.2315970727507533,
+      "grad_norm": 0.1729276180267334,
+      "learning_rate": 1.821403446681336e-05,
+      "loss": 1.2026,
+      "step": 5994
+    },
+    {
+      "epoch": 2.2319693779013625,
+      "grad_norm": 0.16060994565486908,
+      "learning_rate": 1.8213341253423248e-05,
+      "loss": 1.1569,
+      "step": 5995
+    },
+    {
+      "epoch": 2.2323416830519713,
+      "grad_norm": 0.16661281883716583,
+      "learning_rate": 1.8212647918722605e-05,
+      "loss": 1.1791,
+      "step": 5996
+    },
+    {
+      "epoch": 2.2327139882025806,
+      "grad_norm": 0.16243353486061096,
+      "learning_rate": 1.8211954462721663e-05,
+      "loss": 1.1822,
+      "step": 5997
+    },
+    {
+      "epoch": 2.2330862933531894,
+      "grad_norm": 0.16968528926372528,
+      "learning_rate": 1.8211260885430672e-05,
+      "loss": 1.1697,
+      "step": 5998
+    },
+    {
+      "epoch": 2.2334585985037987,
+      "grad_norm": 0.16825567185878754,
+      "learning_rate": 1.821056718685987e-05,
+      "loss": 1.1784,
+      "step": 5999
+    },
+    {
+      "epoch": 2.233830903654408,
+      "grad_norm": 0.1593291014432907,
+      "learning_rate": 1.820987336701951e-05,
+      "loss": 1.1793,
+      "step": 6000
+    },
+    {
+      "epoch": 2.233830903654408,
+      "eval_loss": 1.3026354312896729,
+      "eval_runtime": 16.489,
+      "eval_samples_per_second": 105.161,
+      "eval_steps_per_second": 5.276,
+      "step": 6000
+    },
+    {
+      "epoch": 2.2342032088050168,
+      "grad_norm": 0.16981658339500427,
+      "learning_rate": 1.8209179425919832e-05,
+      "loss": 1.1754,
+      "step": 6001
+    },
+    {
+      "epoch": 2.234575513955626,
+      "grad_norm": 0.16528399288654327,
+      "learning_rate": 1.820848536357109e-05,
+      "loss": 1.1896,
+      "step": 6002
+    },
+    {
+      "epoch": 2.234947819106235,
+      "grad_norm": 0.16212095320224762,
+      "learning_rate": 1.8207791179983535e-05,
+      "loss": 1.1762,
+      "step": 6003
+    },
+    {
+      "epoch": 2.235320124256844,
+      "grad_norm": 0.16362003982067108,
+      "learning_rate": 1.8207096875167417e-05,
+      "loss": 1.1724,
+      "step": 6004
+    },
+    {
+      "epoch": 2.235692429407453,
+      "grad_norm": 0.16708360612392426,
+      "learning_rate": 1.8206402449132997e-05,
+      "loss": 1.1824,
+      "step": 6005
+    },
+    {
+      "epoch": 2.236064734558062,
+      "grad_norm": 0.16234560310840607,
+      "learning_rate": 1.8205707901890524e-05,
+      "loss": 1.178,
+      "step": 6006
+    },
+    {
+      "epoch": 2.236437039708671,
+      "grad_norm": 0.16618958115577698,
+      "learning_rate": 1.8205013233450268e-05,
+      "loss": 1.1672,
+      "step": 6007
+    },
+    {
+      "epoch": 2.2368093448592803,
+      "grad_norm": 0.17038699984550476,
+      "learning_rate": 1.8204318443822473e-05,
+      "loss": 1.1991,
+      "step": 6008
+    },
+    {
+      "epoch": 2.2371816500098896,
+      "grad_norm": 0.16078020632266998,
+      "learning_rate": 1.8203623533017413e-05,
+      "loss": 1.1836,
+      "step": 6009
+    },
+    {
+      "epoch": 2.2375539551604984,
+      "grad_norm": 0.16053731739521027,
+      "learning_rate": 1.8202928501045347e-05,
+      "loss": 1.1889,
+      "step": 6010
+    },
+    {
+      "epoch": 2.2379262603111076,
+      "grad_norm": 0.16496212780475616,
+      "learning_rate": 1.820223334791654e-05,
+      "loss": 1.1908,
+      "step": 6011
+    },
+    {
+      "epoch": 2.2382985654617165,
+      "grad_norm": 0.16259627044200897,
+      "learning_rate": 1.8201538073641264e-05,
+      "loss": 1.1845,
+      "step": 6012
+    },
+    {
+      "epoch": 2.2386708706123257,
+      "grad_norm": 0.16362962126731873,
+      "learning_rate": 1.8200842678229786e-05,
+      "loss": 1.1731,
+      "step": 6013
+    },
+    {
+      "epoch": 2.2390431757629345,
+      "grad_norm": 0.16907905042171478,
+      "learning_rate": 1.8200147161692373e-05,
+      "loss": 1.1932,
+      "step": 6014
+    },
+    {
+      "epoch": 2.239415480913544,
+      "grad_norm": 0.1673828512430191,
+      "learning_rate": 1.8199451524039308e-05,
+      "loss": 1.1848,
+      "step": 6015
+    },
+    {
+      "epoch": 2.2397877860641526,
+      "grad_norm": 0.16971193253993988,
+      "learning_rate": 1.819875576528085e-05,
+      "loss": 1.1777,
+      "step": 6016
+    },
+    {
+      "epoch": 2.240160091214762,
+      "grad_norm": 0.173600435256958,
+      "learning_rate": 1.819805988542729e-05,
+      "loss": 1.1947,
+      "step": 6017
+    },
+    {
+      "epoch": 2.240532396365371,
+      "grad_norm": 0.1610695868730545,
+      "learning_rate": 1.81973638844889e-05,
+      "loss": 1.1796,
+      "step": 6018
+    },
+    {
+      "epoch": 2.24090470151598,
+      "grad_norm": 0.16606231033802032,
+      "learning_rate": 1.8196667762475953e-05,
+      "loss": 1.1824,
+      "step": 6019
+    },
+    {
+      "epoch": 2.2412770066665892,
+      "grad_norm": 0.16269895434379578,
+      "learning_rate": 1.8195971519398744e-05,
+      "loss": 1.1918,
+      "step": 6020
+    },
+    {
+      "epoch": 2.241649311817198,
+      "grad_norm": 0.1672276109457016,
+      "learning_rate": 1.8195275155267546e-05,
+      "loss": 1.197,
+      "step": 6021
+    },
+    {
+      "epoch": 2.2420216169678073,
+      "grad_norm": 0.16324323415756226,
+      "learning_rate": 1.8194578670092654e-05,
+      "loss": 1.1799,
+      "step": 6022
+    },
+    {
+      "epoch": 2.242393922118416,
+      "grad_norm": 0.1610366255044937,
+      "learning_rate": 1.8193882063884346e-05,
+      "loss": 1.1787,
+      "step": 6023
+    },
+    {
+      "epoch": 2.2427662272690254,
+      "grad_norm": 0.163301020860672,
+      "learning_rate": 1.8193185336652912e-05,
+      "loss": 1.1851,
+      "step": 6024
+    },
+    {
+      "epoch": 2.2431385324196347,
+      "grad_norm": 0.16629861295223236,
+      "learning_rate": 1.819248848840865e-05,
+      "loss": 1.1963,
+      "step": 6025
+    },
+    {
+      "epoch": 2.2435108375702435,
+      "grad_norm": 0.1662205159664154,
+      "learning_rate": 1.819179151916184e-05,
+      "loss": 1.189,
+      "step": 6026
+    },
+    {
+      "epoch": 2.2438831427208528,
+      "grad_norm": 0.15960603952407837,
+      "learning_rate": 1.819109442892279e-05,
+      "loss": 1.1654,
+      "step": 6027
+    },
+    {
+      "epoch": 2.2442554478714616,
+      "grad_norm": 0.16344155371189117,
+      "learning_rate": 1.8190397217701785e-05,
+      "loss": 1.1841,
+      "step": 6028
+    },
+    {
+      "epoch": 2.244627753022071,
+      "grad_norm": 0.16831251978874207,
+      "learning_rate": 1.8189699885509128e-05,
+      "loss": 1.1615,
+      "step": 6029
+    },
+    {
+      "epoch": 2.2450000581726797,
+      "grad_norm": 0.1629910171031952,
+      "learning_rate": 1.818900243235512e-05,
+      "loss": 1.1961,
+      "step": 6030
+    },
+    {
+      "epoch": 2.245372363323289,
+      "grad_norm": 0.16238047182559967,
+      "learning_rate": 1.818830485825006e-05,
+      "loss": 1.1766,
+      "step": 6031
+    },
+    {
+      "epoch": 2.2457446684738978,
+      "grad_norm": 0.1695476919412613,
+      "learning_rate": 1.8187607163204246e-05,
+      "loss": 1.173,
+      "step": 6032
+    },
+    {
+      "epoch": 2.246116973624507,
+      "grad_norm": 0.16819293797016144,
+      "learning_rate": 1.8186909347227992e-05,
+      "loss": 1.1899,
+      "step": 6033
+    },
+    {
+      "epoch": 2.2464892787751163,
+      "grad_norm": 0.16680686175823212,
+      "learning_rate": 1.81862114103316e-05,
+      "loss": 1.1771,
+      "step": 6034
+    },
+    {
+      "epoch": 2.246861583925725,
+      "grad_norm": 0.1659124493598938,
+      "learning_rate": 1.818551335252538e-05,
+      "loss": 1.1693,
+      "step": 6035
+    },
+    {
+      "epoch": 2.2472338890763344,
+      "grad_norm": 0.15798041224479675,
+      "learning_rate": 1.818481517381964e-05,
+      "loss": 1.1757,
+      "step": 6036
+    },
+    {
+      "epoch": 2.247606194226943,
+      "grad_norm": 0.17841339111328125,
+      "learning_rate": 1.8184116874224695e-05,
+      "loss": 1.1828,
+      "step": 6037
+    },
+    {
+      "epoch": 2.2479784993775525,
+      "grad_norm": 0.1704692840576172,
+      "learning_rate": 1.818341845375086e-05,
+      "loss": 1.1874,
+      "step": 6038
+    },
+    {
+      "epoch": 2.2483508045281613,
+      "grad_norm": 0.16449272632598877,
+      "learning_rate": 1.818271991240844e-05,
+      "loss": 1.1822,
+      "step": 6039
+    },
+    {
+      "epoch": 2.2487231096787705,
+      "grad_norm": 0.15764431655406952,
+      "learning_rate": 1.818202125020777e-05,
+      "loss": 1.1709,
+      "step": 6040
+    },
+    {
+      "epoch": 2.2490954148293794,
+      "grad_norm": 0.18003995716571808,
+      "learning_rate": 1.8181322467159153e-05,
+      "loss": 1.1801,
+      "step": 6041
+    },
+    {
+      "epoch": 2.2494677199799886,
+      "grad_norm": 0.1611638069152832,
+      "learning_rate": 1.818062356327292e-05,
+      "loss": 1.1702,
+      "step": 6042
+    },
+    {
+      "epoch": 2.249840025130598,
+      "grad_norm": 0.17060458660125732,
+      "learning_rate": 1.8179924538559385e-05,
+      "loss": 1.1737,
+      "step": 6043
+    },
+    {
+      "epoch": 2.2502123302812067,
+      "grad_norm": 0.1664137840270996,
+      "learning_rate": 1.8179225393028883e-05,
+      "loss": 1.1984,
+      "step": 6044
+    },
+    {
+      "epoch": 2.250584635431816,
+      "grad_norm": 0.16480514407157898,
+      "learning_rate": 1.8178526126691734e-05,
+      "loss": 1.1802,
+      "step": 6045
+    },
+    {
+      "epoch": 2.250956940582425,
+      "grad_norm": 0.1663130819797516,
+      "learning_rate": 1.8177826739558268e-05,
+      "loss": 1.1844,
+      "step": 6046
+    },
+    {
+      "epoch": 2.251329245733034,
+      "grad_norm": 0.15989989042282104,
+      "learning_rate": 1.8177127231638815e-05,
+      "loss": 1.1773,
+      "step": 6047
+    },
+    {
+      "epoch": 2.251701550883643,
+      "grad_norm": 0.1636444330215454,
+      "learning_rate": 1.8176427602943705e-05,
+      "loss": 1.1707,
+      "step": 6048
+    },
+    {
+      "epoch": 2.252073856034252,
+      "grad_norm": 0.16399884223937988,
+      "learning_rate": 1.817572785348327e-05,
+      "loss": 1.1714,
+      "step": 6049
+    },
+    {
+      "epoch": 2.252446161184861,
+      "grad_norm": 0.1703033149242401,
+      "learning_rate": 1.817502798326785e-05,
+      "loss": 1.199,
+      "step": 6050
+    },
+    {
+      "epoch": 2.2528184663354702,
+      "grad_norm": 0.16354358196258545,
+      "learning_rate": 1.817432799230778e-05,
+      "loss": 1.1994,
+      "step": 6051
+    },
+    {
+      "epoch": 2.2531907714860795,
+      "grad_norm": 0.16181591153144836,
+      "learning_rate": 1.8173627880613394e-05,
+      "loss": 1.1794,
+      "step": 6052
+    },
+    {
+      "epoch": 2.2535630766366883,
+      "grad_norm": 0.1724006086587906,
+      "learning_rate": 1.8172927648195043e-05,
+      "loss": 1.1879,
+      "step": 6053
+    },
+    {
+      "epoch": 2.2539353817872976,
+      "grad_norm": 0.16579163074493408,
+      "learning_rate": 1.8172227295063062e-05,
+      "loss": 1.1827,
+      "step": 6054
+    },
+    {
+      "epoch": 2.2543076869379064,
+      "grad_norm": 0.16125434637069702,
+      "learning_rate": 1.81715268212278e-05,
+      "loss": 1.1758,
+      "step": 6055
+    },
+    {
+      "epoch": 2.2546799920885157,
+      "grad_norm": 0.16942323744297028,
+      "learning_rate": 1.8170826226699593e-05,
+      "loss": 1.1849,
+      "step": 6056
+    },
+    {
+      "epoch": 2.2550522972391245,
+      "grad_norm": 0.1689453423023224,
+      "learning_rate": 1.81701255114888e-05,
+      "loss": 1.1837,
+      "step": 6057
+    },
+    {
+      "epoch": 2.2554246023897337,
+      "grad_norm": 0.1727559119462967,
+      "learning_rate": 1.8169424675605766e-05,
+      "loss": 1.1916,
+      "step": 6058
+    },
+    {
+      "epoch": 2.2557969075403426,
+      "grad_norm": 0.17401790618896484,
+      "learning_rate": 1.816872371906084e-05,
+      "loss": 1.1693,
+      "step": 6059
+    },
+    {
+      "epoch": 2.256169212690952,
+      "grad_norm": 0.16698595881462097,
+      "learning_rate": 1.816802264186438e-05,
+      "loss": 1.18,
+      "step": 6060
+    },
+    {
+      "epoch": 2.256541517841561,
+      "grad_norm": 0.1707891970872879,
+      "learning_rate": 1.816732144402673e-05,
+      "loss": 1.1742,
+      "step": 6061
+    },
+    {
+      "epoch": 2.25691382299217,
+      "grad_norm": 0.16598151624202728,
+      "learning_rate": 1.8166620125558263e-05,
+      "loss": 1.188,
+      "step": 6062
+    },
+    {
+      "epoch": 2.257286128142779,
+      "grad_norm": 0.16978704929351807,
+      "learning_rate": 1.816591868646933e-05,
+      "loss": 1.1884,
+      "step": 6063
+    },
+    {
+      "epoch": 2.257658433293388,
+      "grad_norm": 0.16990995407104492,
+      "learning_rate": 1.8165217126770285e-05,
+      "loss": 1.1826,
+      "step": 6064
+    },
+    {
+      "epoch": 2.2580307384439973,
+      "grad_norm": 0.1661897599697113,
+      "learning_rate": 1.81645154464715e-05,
+      "loss": 1.1702,
+      "step": 6065
+    },
+    {
+      "epoch": 2.258403043594606,
+      "grad_norm": 0.16440676152706146,
+      "learning_rate": 1.816381364558333e-05,
+      "loss": 1.1769,
+      "step": 6066
+    },
+    {
+      "epoch": 2.2587753487452154,
+      "grad_norm": 0.18172229826450348,
+      "learning_rate": 1.816311172411615e-05,
+      "loss": 1.1747,
+      "step": 6067
+    },
+    {
+      "epoch": 2.259147653895824,
+      "grad_norm": 0.16828443109989166,
+      "learning_rate": 1.816240968208032e-05,
+      "loss": 1.2007,
+      "step": 6068
+    },
+    {
+      "epoch": 2.2595199590464334,
+      "grad_norm": 0.16745588183403015,
+      "learning_rate": 1.816170751948621e-05,
+      "loss": 1.1842,
+      "step": 6069
+    },
+    {
+      "epoch": 2.2598922641970427,
+      "grad_norm": 0.16287997364997864,
+      "learning_rate": 1.8161005236344193e-05,
+      "loss": 1.1852,
+      "step": 6070
+    },
+    {
+      "epoch": 2.2602645693476515,
+      "grad_norm": 0.1797943264245987,
+      "learning_rate": 1.816030283266464e-05,
+      "loss": 1.1838,
+      "step": 6071
+    },
+    {
+      "epoch": 2.260636874498261,
+      "grad_norm": 0.1775754690170288,
+      "learning_rate": 1.815960030845793e-05,
+      "loss": 1.1877,
+      "step": 6072
+    },
+    {
+      "epoch": 2.2610091796488696,
+      "grad_norm": 0.1645868569612503,
+      "learning_rate": 1.815889766373443e-05,
+      "loss": 1.184,
+      "step": 6073
+    },
+    {
+      "epoch": 2.261381484799479,
+      "grad_norm": 0.17757607996463776,
+      "learning_rate": 1.8158194898504526e-05,
+      "loss": 1.1702,
+      "step": 6074
+    },
+    {
+      "epoch": 2.2617537899500877,
+      "grad_norm": 0.16869355738162994,
+      "learning_rate": 1.8157492012778598e-05,
+      "loss": 1.1741,
+      "step": 6075
+    },
+    {
+      "epoch": 2.262126095100697,
+      "grad_norm": 0.1690586805343628,
+      "learning_rate": 1.8156789006567018e-05,
+      "loss": 1.1676,
+      "step": 6076
+    },
+    {
+      "epoch": 2.2624984002513058,
+      "grad_norm": 0.17510120570659637,
+      "learning_rate": 1.815608587988018e-05,
+      "loss": 1.1863,
+      "step": 6077
+    },
+    {
+      "epoch": 2.262870705401915,
+      "grad_norm": 0.1718439757823944,
+      "learning_rate": 1.8155382632728468e-05,
+      "loss": 1.1847,
+      "step": 6078
+    },
+    {
+      "epoch": 2.2632430105525243,
+      "grad_norm": 0.1764347404241562,
+      "learning_rate": 1.8154679265122265e-05,
+      "loss": 1.1627,
+      "step": 6079
+    },
+    {
+      "epoch": 2.263615315703133,
+      "grad_norm": 0.1741929054260254,
+      "learning_rate": 1.815397577707196e-05,
+      "loss": 1.186,
+      "step": 6080
+    },
+    {
+      "epoch": 2.2639876208537424,
+      "grad_norm": 0.162635937333107,
+      "learning_rate": 1.8153272168587947e-05,
+      "loss": 1.1809,
+      "step": 6081
+    },
+    {
+      "epoch": 2.264359926004351,
+      "grad_norm": 0.16348296403884888,
+      "learning_rate": 1.8152568439680612e-05,
+      "loss": 1.179,
+      "step": 6082
+    },
+    {
+      "epoch": 2.2647322311549605,
+      "grad_norm": 0.16939151287078857,
+      "learning_rate": 1.8151864590360354e-05,
+      "loss": 1.1605,
+      "step": 6083
+    },
+    {
+      "epoch": 2.2651045363055693,
+      "grad_norm": 0.17128928005695343,
+      "learning_rate": 1.815116062063757e-05,
+      "loss": 1.1947,
+      "step": 6084
+    },
+    {
+      "epoch": 2.2654768414561786,
+      "grad_norm": 0.1636468470096588,
+      "learning_rate": 1.8150456530522652e-05,
+      "loss": 1.181,
+      "step": 6085
+    },
+    {
+      "epoch": 2.2658491466067874,
+      "grad_norm": 0.1675485074520111,
+      "learning_rate": 1.8149752320026004e-05,
+      "loss": 1.1772,
+      "step": 6086
+    },
+    {
+      "epoch": 2.2662214517573966,
+      "grad_norm": 0.17299336194992065,
+      "learning_rate": 1.8149047989158026e-05,
+      "loss": 1.1853,
+      "step": 6087
+    },
+    {
+      "epoch": 2.266593756908006,
+      "grad_norm": 0.16324491798877716,
+      "learning_rate": 1.8148343537929118e-05,
+      "loss": 1.1806,
+      "step": 6088
+    },
+    {
+      "epoch": 2.2669660620586147,
+      "grad_norm": 0.16224730014801025,
+      "learning_rate": 1.8147638966349687e-05,
+      "loss": 1.1856,
+      "step": 6089
+    },
+    {
+      "epoch": 2.267338367209224,
+      "grad_norm": 0.16687463223934174,
+      "learning_rate": 1.8146934274430147e-05,
+      "loss": 1.2029,
+      "step": 6090
+    },
+    {
+      "epoch": 2.267710672359833,
+      "grad_norm": 0.16694878041744232,
+      "learning_rate": 1.814622946218089e-05,
+      "loss": 1.1644,
+      "step": 6091
+    },
+    {
+      "epoch": 2.268082977510442,
+      "grad_norm": 0.1973126232624054,
+      "learning_rate": 1.814552452961234e-05,
+      "loss": 1.1787,
+      "step": 6092
+    },
+    {
+      "epoch": 2.268455282661051,
+      "grad_norm": 0.20395469665527344,
+      "learning_rate": 1.81448194767349e-05,
+      "loss": 1.1916,
+      "step": 6093
+    },
+    {
+      "epoch": 2.26882758781166,
+      "grad_norm": 0.1700964868068695,
+      "learning_rate": 1.8144114303558993e-05,
+      "loss": 1.1889,
+      "step": 6094
+    },
+    {
+      "epoch": 2.269199892962269,
+      "grad_norm": 0.17254211008548737,
+      "learning_rate": 1.8143409010095028e-05,
+      "loss": 1.1644,
+      "step": 6095
+    },
+    {
+      "epoch": 2.2695721981128782,
+      "grad_norm": 0.19276972115039825,
+      "learning_rate": 1.814270359635342e-05,
+      "loss": 1.1883,
+      "step": 6096
+    },
+    {
+      "epoch": 2.2699445032634875,
+      "grad_norm": 0.1734069436788559,
+      "learning_rate": 1.814199806234459e-05,
+      "loss": 1.1843,
+      "step": 6097
+    },
+    {
+      "epoch": 2.2703168084140963,
+      "grad_norm": 0.1637657731771469,
+      "learning_rate": 1.8141292408078963e-05,
+      "loss": 1.1871,
+      "step": 6098
+    },
+    {
+      "epoch": 2.2706891135647056,
+      "grad_norm": 0.17510093748569489,
+      "learning_rate": 1.814058663356696e-05,
+      "loss": 1.1796,
+      "step": 6099
+    },
+    {
+      "epoch": 2.2710614187153144,
+      "grad_norm": 0.16966238617897034,
+      "learning_rate": 1.8139880738819e-05,
+      "loss": 1.1683,
+      "step": 6100
+    },
+    {
+      "epoch": 2.2714337238659237,
+      "grad_norm": 0.1668512225151062,
+      "learning_rate": 1.8139174723845513e-05,
+      "loss": 1.1993,
+      "step": 6101
+    },
+    {
+      "epoch": 2.2718060290165325,
+      "grad_norm": 0.17251057922840118,
+      "learning_rate": 1.8138468588656922e-05,
+      "loss": 1.19,
+      "step": 6102
+    },
+    {
+      "epoch": 2.2721783341671418,
+      "grad_norm": 0.1686212718486786,
+      "learning_rate": 1.8137762333263667e-05,
+      "loss": 1.1798,
+      "step": 6103
+    },
+    {
+      "epoch": 2.2725506393177506,
+      "grad_norm": 0.16368193924427032,
+      "learning_rate": 1.8137055957676172e-05,
+      "loss": 1.168,
+      "step": 6104
+    },
+    {
+      "epoch": 2.27292294446836,
+      "grad_norm": 0.1598677635192871,
+      "learning_rate": 1.8136349461904866e-05,
+      "loss": 1.1762,
+      "step": 6105
+    },
+    {
+      "epoch": 2.273295249618969,
+      "grad_norm": 0.16657836735248566,
+      "learning_rate": 1.8135642845960195e-05,
+      "loss": 1.179,
+      "step": 6106
+    },
+    {
+      "epoch": 2.273667554769578,
+      "grad_norm": 0.17238342761993408,
+      "learning_rate": 1.8134936109852587e-05,
+      "loss": 1.2088,
+      "step": 6107
+    },
+    {
+      "epoch": 2.274039859920187,
+      "grad_norm": 0.17234353721141815,
+      "learning_rate": 1.8134229253592485e-05,
+      "loss": 1.1965,
+      "step": 6108
+    },
+    {
+      "epoch": 2.274412165070796,
+      "grad_norm": 0.16358719766139984,
+      "learning_rate": 1.8133522277190324e-05,
+      "loss": 1.1843,
+      "step": 6109
+    },
+    {
+      "epoch": 2.2747844702214053,
+      "grad_norm": 0.17148621380329132,
+      "learning_rate": 1.8132815180656554e-05,
+      "loss": 1.1863,
+      "step": 6110
+    },
+    {
+      "epoch": 2.275156775372014,
+      "grad_norm": 0.17336560785770416,
+      "learning_rate": 1.813210796400161e-05,
+      "loss": 1.1859,
+      "step": 6111
+    },
+    {
+      "epoch": 2.2755290805226234,
+      "grad_norm": 0.17160752415657043,
+      "learning_rate": 1.813140062723594e-05,
+      "loss": 1.1925,
+      "step": 6112
+    },
+    {
+      "epoch": 2.275901385673232,
+      "grad_norm": 0.16290338337421417,
+      "learning_rate": 1.8130693170369998e-05,
+      "loss": 1.174,
+      "step": 6113
+    },
+    {
+      "epoch": 2.2762736908238415,
+      "grad_norm": 0.16730904579162598,
+      "learning_rate": 1.812998559341423e-05,
+      "loss": 1.1699,
+      "step": 6114
+    },
+    {
+      "epoch": 2.2766459959744507,
+      "grad_norm": 0.1715693324804306,
+      "learning_rate": 1.8129277896379077e-05,
+      "loss": 1.1877,
+      "step": 6115
+    },
+    {
+      "epoch": 2.2770183011250595,
+      "grad_norm": 0.16124042868614197,
+      "learning_rate": 1.812857007927501e-05,
+      "loss": 1.1604,
+      "step": 6116
+    },
+    {
+      "epoch": 2.277390606275669,
+      "grad_norm": 0.155757874250412,
+      "learning_rate": 1.8127862142112463e-05,
+      "loss": 1.1699,
+      "step": 6117
+    },
+    {
+      "epoch": 2.2777629114262776,
+      "grad_norm": 0.16820941865444183,
+      "learning_rate": 1.8127154084901906e-05,
+      "loss": 1.1702,
+      "step": 6118
+    },
+    {
+      "epoch": 2.278135216576887,
+      "grad_norm": 0.16459953784942627,
+      "learning_rate": 1.8126445907653797e-05,
+      "loss": 1.1873,
+      "step": 6119
+    },
+    {
+      "epoch": 2.2785075217274957,
+      "grad_norm": 0.16601146757602692,
+      "learning_rate": 1.8125737610378585e-05,
+      "loss": 1.1809,
+      "step": 6120
+    },
+    {
+      "epoch": 2.278879826878105,
+      "grad_norm": 0.16226966679096222,
+      "learning_rate": 1.8125029193086743e-05,
+      "loss": 1.1728,
+      "step": 6121
+    },
+    {
+      "epoch": 2.279252132028714,
+      "grad_norm": 0.16306665539741516,
+      "learning_rate": 1.812432065578873e-05,
+      "loss": 1.1887,
+      "step": 6122
+    },
+    {
+      "epoch": 2.279624437179323,
+      "grad_norm": 0.15871618688106537,
+      "learning_rate": 1.812361199849501e-05,
+      "loss": 1.1873,
+      "step": 6123
+    },
+    {
+      "epoch": 2.2799967423299323,
+      "grad_norm": 0.17421315610408783,
+      "learning_rate": 1.812290322121605e-05,
+      "loss": 1.1949,
+      "step": 6124
+    },
+    {
+      "epoch": 2.280369047480541,
+      "grad_norm": 0.16765080392360687,
+      "learning_rate": 1.8122194323962317e-05,
+      "loss": 1.1913,
+      "step": 6125
+    },
+    {
+      "epoch": 2.2807413526311504,
+      "grad_norm": 0.15896672010421753,
+      "learning_rate": 1.8121485306744286e-05,
+      "loss": 1.1887,
+      "step": 6126
+    },
+    {
+      "epoch": 2.2811136577817592,
+      "grad_norm": 0.16295817494392395,
+      "learning_rate": 1.8120776169572427e-05,
+      "loss": 1.1773,
+      "step": 6127
+    },
+    {
+      "epoch": 2.2814859629323685,
+      "grad_norm": 0.16846707463264465,
+      "learning_rate": 1.8120066912457216e-05,
+      "loss": 1.1831,
+      "step": 6128
+    },
+    {
+      "epoch": 2.2818582680829778,
+      "grad_norm": 0.1673799604177475,
+      "learning_rate": 1.811935753540912e-05,
+      "loss": 1.1824,
+      "step": 6129
+    },
+    {
+      "epoch": 2.2822305732335866,
+      "grad_norm": 0.1708277016878128,
+      "learning_rate": 1.8118648038438627e-05,
+      "loss": 1.1776,
+      "step": 6130
+    },
+    {
+      "epoch": 2.2826028783841954,
+      "grad_norm": 0.170358344912529,
+      "learning_rate": 1.811793842155621e-05,
+      "loss": 1.1843,
+      "step": 6131
+    },
+    {
+      "epoch": 2.2829751835348047,
+      "grad_norm": 0.17161040008068085,
+      "learning_rate": 1.8117228684772358e-05,
+      "loss": 1.1901,
+      "step": 6132
+    },
+    {
+      "epoch": 2.283347488685414,
+      "grad_norm": 0.16947078704833984,
+      "learning_rate": 1.811651882809754e-05,
+      "loss": 1.1788,
+      "step": 6133
+    },
+    {
+      "epoch": 2.2837197938360227,
+      "grad_norm": 0.1724851131439209,
+      "learning_rate": 1.8115808851542255e-05,
+      "loss": 1.1903,
+      "step": 6134
+    },
+    {
+      "epoch": 2.284092098986632,
+      "grad_norm": 0.17149105668067932,
+      "learning_rate": 1.8115098755116974e-05,
+      "loss": 1.1849,
+      "step": 6135
+    },
+    {
+      "epoch": 2.284464404137241,
+      "grad_norm": 0.16904261708259583,
+      "learning_rate": 1.81143885388322e-05,
+      "loss": 1.18,
+      "step": 6136
+    },
+    {
+      "epoch": 2.28483670928785,
+      "grad_norm": 0.16865547001361847,
+      "learning_rate": 1.811367820269842e-05,
+      "loss": 1.1773,
+      "step": 6137
+    },
+    {
+      "epoch": 2.2852090144384594,
+      "grad_norm": 0.16392925381660461,
+      "learning_rate": 1.811296774672611e-05,
+      "loss": 1.1895,
+      "step": 6138
+    },
+    {
+      "epoch": 2.285581319589068,
+      "grad_norm": 0.17182591557502747,
+      "learning_rate": 1.8112257170925785e-05,
+      "loss": 1.184,
+      "step": 6139
+    },
+    {
+      "epoch": 2.2859536247396774,
+      "grad_norm": 0.16225169599056244,
+      "learning_rate": 1.8111546475307927e-05,
+      "loss": 1.1836,
+      "step": 6140
+    },
+    {
+      "epoch": 2.2863259298902863,
+      "grad_norm": 0.16837048530578613,
+      "learning_rate": 1.811083565988304e-05,
+      "loss": 1.1813,
+      "step": 6141
+    },
+    {
+      "epoch": 2.2866982350408955,
+      "grad_norm": 0.16455179452896118,
+      "learning_rate": 1.8110124724661617e-05,
+      "loss": 1.1837,
+      "step": 6142
+    },
+    {
+      "epoch": 2.2870705401915044,
+      "grad_norm": 0.19278612732887268,
+      "learning_rate": 1.810941366965416e-05,
+      "loss": 1.177,
+      "step": 6143
+    },
+    {
+      "epoch": 2.2874428453421136,
+      "grad_norm": 0.18453273177146912,
+      "learning_rate": 1.8108702494871173e-05,
+      "loss": 1.1796,
+      "step": 6144
+    },
+    {
+      "epoch": 2.2878151504927224,
+      "grad_norm": 0.17446112632751465,
+      "learning_rate": 1.8107991200323162e-05,
+      "loss": 1.1892,
+      "step": 6145
+    },
+    {
+      "epoch": 2.2881874556433317,
+      "grad_norm": 0.15920548141002655,
+      "learning_rate": 1.8107279786020627e-05,
+      "loss": 1.1722,
+      "step": 6146
+    },
+    {
+      "epoch": 2.288559760793941,
+      "grad_norm": 0.21259953081607819,
+      "learning_rate": 1.8106568251974077e-05,
+      "loss": 1.1791,
+      "step": 6147
+    },
+    {
+      "epoch": 2.28893206594455,
+      "grad_norm": 0.16296431422233582,
+      "learning_rate": 1.8105856598194026e-05,
+      "loss": 1.1778,
+      "step": 6148
+    },
+    {
+      "epoch": 2.289304371095159,
+      "grad_norm": 0.1709262728691101,
+      "learning_rate": 1.8105144824690977e-05,
+      "loss": 1.179,
+      "step": 6149
+    },
+    {
+      "epoch": 2.289676676245768,
+      "grad_norm": 0.163625106215477,
+      "learning_rate": 1.8104432931475454e-05,
+      "loss": 1.1787,
+      "step": 6150
+    },
+    {
+      "epoch": 2.290048981396377,
+      "grad_norm": 0.16324341297149658,
+      "learning_rate": 1.810372091855796e-05,
+      "loss": 1.1714,
+      "step": 6151
+    },
+    {
+      "epoch": 2.290421286546986,
+      "grad_norm": 0.16584549844264984,
+      "learning_rate": 1.8103008785949015e-05,
+      "loss": 1.173,
+      "step": 6152
+    },
+    {
+      "epoch": 2.290793591697595,
+      "grad_norm": 0.17169234156608582,
+      "learning_rate": 1.8102296533659146e-05,
+      "loss": 1.1747,
+      "step": 6153
+    },
+    {
+      "epoch": 2.291165896848204,
+      "grad_norm": 0.17495130002498627,
+      "learning_rate": 1.810158416169886e-05,
+      "loss": 1.1814,
+      "step": 6154
+    },
+    {
+      "epoch": 2.2915382019988133,
+      "grad_norm": 0.16563518345355988,
+      "learning_rate": 1.8100871670078687e-05,
+      "loss": 1.1744,
+      "step": 6155
+    },
+    {
+      "epoch": 2.2919105071494226,
+      "grad_norm": 0.16511280834674835,
+      "learning_rate": 1.8100159058809146e-05,
+      "loss": 1.184,
+      "step": 6156
+    },
+    {
+      "epoch": 2.2922828123000314,
+      "grad_norm": 0.16747237741947174,
+      "learning_rate": 1.8099446327900766e-05,
+      "loss": 1.1883,
+      "step": 6157
+    },
+    {
+      "epoch": 2.2926551174506407,
+      "grad_norm": 0.16100740432739258,
+      "learning_rate": 1.809873347736407e-05,
+      "loss": 1.1764,
+      "step": 6158
+    },
+    {
+      "epoch": 2.2930274226012495,
+      "grad_norm": 0.16185326874256134,
+      "learning_rate": 1.809802050720959e-05,
+      "loss": 1.1846,
+      "step": 6159
+    },
+    {
+      "epoch": 2.2933997277518587,
+      "grad_norm": 0.1648186296224594,
+      "learning_rate": 1.8097307417447855e-05,
+      "loss": 1.1802,
+      "step": 6160
+    },
+    {
+      "epoch": 2.2937720329024676,
+      "grad_norm": 0.1759217381477356,
+      "learning_rate": 1.8096594208089397e-05,
+      "loss": 1.1826,
+      "step": 6161
+    },
+    {
+      "epoch": 2.294144338053077,
+      "grad_norm": 0.1690814197063446,
+      "learning_rate": 1.809588087914475e-05,
+      "loss": 1.1606,
+      "step": 6162
+    },
+    {
+      "epoch": 2.2945166432036856,
+      "grad_norm": 0.1603320837020874,
+      "learning_rate": 1.8095167430624454e-05,
+      "loss": 1.1739,
+      "step": 6163
+    },
+    {
+      "epoch": 2.294888948354295,
+      "grad_norm": 0.16544067859649658,
+      "learning_rate": 1.809445386253904e-05,
+      "loss": 1.1809,
+      "step": 6164
+    },
+    {
+      "epoch": 2.295261253504904,
+      "grad_norm": 0.16254539787769318,
+      "learning_rate": 1.809374017489905e-05,
+      "loss": 1.1904,
+      "step": 6165
+    },
+    {
+      "epoch": 2.295633558655513,
+      "grad_norm": 0.16620919108390808,
+      "learning_rate": 1.809302636771503e-05,
+      "loss": 1.1834,
+      "step": 6166
+    },
+    {
+      "epoch": 2.2960058638061223,
+      "grad_norm": 0.16282342374324799,
+      "learning_rate": 1.809231244099751e-05,
+      "loss": 1.1815,
+      "step": 6167
+    },
+    {
+      "epoch": 2.296378168956731,
+      "grad_norm": 0.157831609249115,
+      "learning_rate": 1.809159839475705e-05,
+      "loss": 1.1681,
+      "step": 6168
+    },
+    {
+      "epoch": 2.2967504741073403,
+      "grad_norm": 0.16010123491287231,
+      "learning_rate": 1.809088422900419e-05,
+      "loss": 1.1629,
+      "step": 6169
+    },
+    {
+      "epoch": 2.297122779257949,
+      "grad_norm": 0.16694015264511108,
+      "learning_rate": 1.8090169943749477e-05,
+      "loss": 1.1824,
+      "step": 6170
+    },
+    {
+      "epoch": 2.2974950844085584,
+      "grad_norm": 0.15992291271686554,
+      "learning_rate": 1.8089455539003457e-05,
+      "loss": 1.1718,
+      "step": 6171
+    },
+    {
+      "epoch": 2.2978673895591673,
+      "grad_norm": 0.1654921919107437,
+      "learning_rate": 1.808874101477669e-05,
+      "loss": 1.1947,
+      "step": 6172
+    },
+    {
+      "epoch": 2.2982396947097765,
+      "grad_norm": 0.1608392298221588,
+      "learning_rate": 1.8088026371079728e-05,
+      "loss": 1.1953,
+      "step": 6173
+    },
+    {
+      "epoch": 2.298611999860386,
+      "grad_norm": 0.1624198704957962,
+      "learning_rate": 1.8087311607923118e-05,
+      "loss": 1.1819,
+      "step": 6174
+    },
+    {
+      "epoch": 2.2989843050109946,
+      "grad_norm": 0.163405641913414,
+      "learning_rate": 1.808659672531743e-05,
+      "loss": 1.1877,
+      "step": 6175
+    },
+    {
+      "epoch": 2.299356610161604,
+      "grad_norm": 0.16340793669223785,
+      "learning_rate": 1.8085881723273215e-05,
+      "loss": 1.1771,
+      "step": 6176
+    },
+    {
+      "epoch": 2.2997289153122127,
+      "grad_norm": 0.16229262948036194,
+      "learning_rate": 1.808516660180103e-05,
+      "loss": 1.1862,
+      "step": 6177
+    },
+    {
+      "epoch": 2.300101220462822,
+      "grad_norm": 0.15974633395671844,
+      "learning_rate": 1.808445136091144e-05,
+      "loss": 1.1828,
+      "step": 6178
+    },
+    {
+      "epoch": 2.3004735256134308,
+      "grad_norm": 0.15938468277454376,
+      "learning_rate": 1.8083736000615017e-05,
+      "loss": 1.1667,
+      "step": 6179
+    },
+    {
+      "epoch": 2.30084583076404,
+      "grad_norm": 0.1704174429178238,
+      "learning_rate": 1.808302052092232e-05,
+      "loss": 1.1785,
+      "step": 6180
+    },
+    {
+      "epoch": 2.301218135914649,
+      "grad_norm": 0.15925031900405884,
+      "learning_rate": 1.8082304921843913e-05,
+      "loss": 1.1729,
+      "step": 6181
+    },
+    {
+      "epoch": 2.301590441065258,
+      "grad_norm": 0.17216730117797852,
+      "learning_rate": 1.8081589203390374e-05,
+      "loss": 1.1874,
+      "step": 6182
+    },
+    {
+      "epoch": 2.3019627462158674,
+      "grad_norm": 0.16319389641284943,
+      "learning_rate": 1.8080873365572265e-05,
+      "loss": 1.1853,
+      "step": 6183
+    },
+    {
+      "epoch": 2.302335051366476,
+      "grad_norm": 0.16606274247169495,
+      "learning_rate": 1.8080157408400167e-05,
+      "loss": 1.1777,
+      "step": 6184
+    },
+    {
+      "epoch": 2.3027073565170855,
+      "grad_norm": 0.16718682646751404,
+      "learning_rate": 1.807944133188465e-05,
+      "loss": 1.1713,
+      "step": 6185
+    },
+    {
+      "epoch": 2.3030796616676943,
+      "grad_norm": 0.16058477759361267,
+      "learning_rate": 1.8078725136036292e-05,
+      "loss": 1.171,
+      "step": 6186
+    },
+    {
+      "epoch": 2.3034519668183036,
+      "grad_norm": 0.1608082354068756,
+      "learning_rate": 1.8078008820865667e-05,
+      "loss": 1.1773,
+      "step": 6187
+    },
+    {
+      "epoch": 2.3038242719689124,
+      "grad_norm": 0.16686289012432098,
+      "learning_rate": 1.8077292386383364e-05,
+      "loss": 1.1741,
+      "step": 6188
+    },
+    {
+      "epoch": 2.3041965771195216,
+      "grad_norm": 0.16598117351531982,
+      "learning_rate": 1.8076575832599957e-05,
+      "loss": 1.1979,
+      "step": 6189
+    },
+    {
+      "epoch": 2.3045688822701305,
+      "grad_norm": 0.16313214600086212,
+      "learning_rate": 1.8075859159526033e-05,
+      "loss": 1.1774,
+      "step": 6190
+    },
+    {
+      "epoch": 2.3049411874207397,
+      "grad_norm": 0.16657821834087372,
+      "learning_rate": 1.8075142367172175e-05,
+      "loss": 1.1888,
+      "step": 6191
+    },
+    {
+      "epoch": 2.305313492571349,
+      "grad_norm": 0.16192109882831573,
+      "learning_rate": 1.8074425455548972e-05,
+      "loss": 1.1688,
+      "step": 6192
+    },
+    {
+      "epoch": 2.305685797721958,
+      "grad_norm": 0.16952742636203766,
+      "learning_rate": 1.807370842466701e-05,
+      "loss": 1.1592,
+      "step": 6193
+    },
+    {
+      "epoch": 2.306058102872567,
+      "grad_norm": 0.15994684398174286,
+      "learning_rate": 1.8072991274536883e-05,
+      "loss": 1.1903,
+      "step": 6194
+    },
+    {
+      "epoch": 2.306430408023176,
+      "grad_norm": 0.16636525094509125,
+      "learning_rate": 1.807227400516918e-05,
+      "loss": 1.1754,
+      "step": 6195
+    },
+    {
+      "epoch": 2.306802713173785,
+      "grad_norm": 0.16642355918884277,
+      "learning_rate": 1.8071556616574498e-05,
+      "loss": 1.182,
+      "step": 6196
+    },
+    {
+      "epoch": 2.307175018324394,
+      "grad_norm": 0.1568913608789444,
+      "learning_rate": 1.807083910876343e-05,
+      "loss": 1.1825,
+      "step": 6197
+    },
+    {
+      "epoch": 2.3075473234750032,
+      "grad_norm": 0.16373074054718018,
+      "learning_rate": 1.8070121481746576e-05,
+      "loss": 1.1873,
+      "step": 6198
+    },
+    {
+      "epoch": 2.307919628625612,
+      "grad_norm": 0.16669581830501556,
+      "learning_rate": 1.8069403735534533e-05,
+      "loss": 1.1804,
+      "step": 6199
+    },
+    {
+      "epoch": 2.3082919337762213,
+      "grad_norm": 0.1655244082212448,
+      "learning_rate": 1.8068685870137906e-05,
+      "loss": 1.194,
+      "step": 6200
+    },
+    {
+      "epoch": 2.3086642389268306,
+      "grad_norm": 0.16131706535816193,
+      "learning_rate": 1.8067967885567292e-05,
+      "loss": 1.172,
+      "step": 6201
+    },
+    {
+      "epoch": 2.3090365440774394,
+      "grad_norm": 0.1645621359348297,
+      "learning_rate": 1.80672497818333e-05,
+      "loss": 1.1778,
+      "step": 6202
+    },
+    {
+      "epoch": 2.3094088492280487,
+      "grad_norm": 0.1699766367673874,
+      "learning_rate": 1.8066531558946537e-05,
+      "loss": 1.1868,
+      "step": 6203
+    },
+    {
+      "epoch": 2.3097811543786575,
+      "grad_norm": 0.16291096806526184,
+      "learning_rate": 1.8065813216917604e-05,
+      "loss": 1.1798,
+      "step": 6204
+    },
+    {
+      "epoch": 2.3101534595292668,
+      "grad_norm": 0.16420194506645203,
+      "learning_rate": 1.806509475575712e-05,
+      "loss": 1.1724,
+      "step": 6205
+    },
+    {
+      "epoch": 2.3105257646798756,
+      "grad_norm": 0.1601489931344986,
+      "learning_rate": 1.806437617547569e-05,
+      "loss": 1.1648,
+      "step": 6206
+    },
+    {
+      "epoch": 2.310898069830485,
+      "grad_norm": 0.17182345688343048,
+      "learning_rate": 1.806365747608393e-05,
+      "loss": 1.1888,
+      "step": 6207
+    },
+    {
+      "epoch": 2.3112703749810937,
+      "grad_norm": 0.16343766450881958,
+      "learning_rate": 1.806293865759246e-05,
+      "loss": 1.17,
+      "step": 6208
+    },
+    {
+      "epoch": 2.311642680131703,
+      "grad_norm": 0.16800105571746826,
+      "learning_rate": 1.806221972001189e-05,
+      "loss": 1.1749,
+      "step": 6209
+    },
+    {
+      "epoch": 2.312014985282312,
+      "grad_norm": 0.16388358175754547,
+      "learning_rate": 1.806150066335284e-05,
+      "loss": 1.1791,
+      "step": 6210
+    },
+    {
+      "epoch": 2.312387290432921,
+      "grad_norm": 0.1601734161376953,
+      "learning_rate": 1.8060781487625927e-05,
+      "loss": 1.181,
+      "step": 6211
+    },
+    {
+      "epoch": 2.3127595955835303,
+      "grad_norm": 0.16823667287826538,
+      "learning_rate": 1.806006219284178e-05,
+      "loss": 1.176,
+      "step": 6212
+    },
+    {
+      "epoch": 2.313131900734139,
+      "grad_norm": 0.16742992401123047,
+      "learning_rate": 1.805934277901102e-05,
+      "loss": 1.175,
+      "step": 6213
+    },
+    {
+      "epoch": 2.3135042058847484,
+      "grad_norm": 0.16113218665122986,
+      "learning_rate": 1.8058623246144274e-05,
+      "loss": 1.1971,
+      "step": 6214
+    },
+    {
+      "epoch": 2.313876511035357,
+      "grad_norm": 0.1706179678440094,
+      "learning_rate": 1.805790359425217e-05,
+      "loss": 1.1806,
+      "step": 6215
+    },
+    {
+      "epoch": 2.3142488161859665,
+      "grad_norm": 0.16922074556350708,
+      "learning_rate": 1.8057183823345333e-05,
+      "loss": 1.184,
+      "step": 6216
+    },
+    {
+      "epoch": 2.3146211213365753,
+      "grad_norm": 0.16326811909675598,
+      "learning_rate": 1.80564639334344e-05,
+      "loss": 1.178,
+      "step": 6217
+    },
+    {
+      "epoch": 2.3149934264871845,
+      "grad_norm": 0.1662980020046234,
+      "learning_rate": 1.8055743924529996e-05,
+      "loss": 1.1884,
+      "step": 6218
+    },
+    {
+      "epoch": 2.315365731637794,
+      "grad_norm": 0.16735336184501648,
+      "learning_rate": 1.805502379664276e-05,
+      "loss": 1.1828,
+      "step": 6219
+    },
+    {
+      "epoch": 2.3157380367884026,
+      "grad_norm": 0.16329142451286316,
+      "learning_rate": 1.805430354978333e-05,
+      "loss": 1.1834,
+      "step": 6220
+    },
+    {
+      "epoch": 2.316110341939012,
+      "grad_norm": 0.16252319514751434,
+      "learning_rate": 1.8053583183962342e-05,
+      "loss": 1.1639,
+      "step": 6221
+    },
+    {
+      "epoch": 2.3164826470896207,
+      "grad_norm": 0.16563460230827332,
+      "learning_rate": 1.8052862699190435e-05,
+      "loss": 1.1942,
+      "step": 6222
+    },
+    {
+      "epoch": 2.31685495224023,
+      "grad_norm": 0.16497038304805756,
+      "learning_rate": 1.8052142095478253e-05,
+      "loss": 1.1846,
+      "step": 6223
+    },
+    {
+      "epoch": 2.317227257390839,
+      "grad_norm": 0.1694660484790802,
+      "learning_rate": 1.8051421372836438e-05,
+      "loss": 1.1886,
+      "step": 6224
+    },
+    {
+      "epoch": 2.317599562541448,
+      "grad_norm": 0.1612187922000885,
+      "learning_rate": 1.8050700531275632e-05,
+      "loss": 1.1802,
+      "step": 6225
+    },
+    {
+      "epoch": 2.317971867692057,
+      "grad_norm": 0.166747584939003,
+      "learning_rate": 1.8049979570806485e-05,
+      "loss": 1.1878,
+      "step": 6226
+    },
+    {
+      "epoch": 2.318344172842666,
+      "grad_norm": 0.17181634902954102,
+      "learning_rate": 1.8049258491439644e-05,
+      "loss": 1.1834,
+      "step": 6227
+    },
+    {
+      "epoch": 2.3187164779932754,
+      "grad_norm": 0.16277888417243958,
+      "learning_rate": 1.8048537293185763e-05,
+      "loss": 1.1808,
+      "step": 6228
+    },
+    {
+      "epoch": 2.3190887831438842,
+      "grad_norm": 0.16461539268493652,
+      "learning_rate": 1.804781597605549e-05,
+      "loss": 1.189,
+      "step": 6229
+    },
+    {
+      "epoch": 2.3194610882944935,
+      "grad_norm": 0.1649829000234604,
+      "learning_rate": 1.8047094540059478e-05,
+      "loss": 1.1877,
+      "step": 6230
+    },
+    {
+      "epoch": 2.3198333934451023,
+      "grad_norm": 0.16547247767448425,
+      "learning_rate": 1.804637298520839e-05,
+      "loss": 1.1789,
+      "step": 6231
+    },
+    {
+      "epoch": 2.3202056985957116,
+      "grad_norm": 0.15987227857112885,
+      "learning_rate": 1.804565131151287e-05,
+      "loss": 1.1692,
+      "step": 6232
+    },
+    {
+      "epoch": 2.3205780037463204,
+      "grad_norm": 0.16262219846248627,
+      "learning_rate": 1.8044929518983592e-05,
+      "loss": 1.1858,
+      "step": 6233
+    },
+    {
+      "epoch": 2.3209503088969297,
+      "grad_norm": 0.16427870094776154,
+      "learning_rate": 1.8044207607631206e-05,
+      "loss": 1.18,
+      "step": 6234
+    },
+    {
+      "epoch": 2.3213226140475385,
+      "grad_norm": 0.1622522473335266,
+      "learning_rate": 1.804348557746638e-05,
+      "loss": 1.1718,
+      "step": 6235
+    },
+    {
+      "epoch": 2.3216949191981477,
+      "grad_norm": 0.15820512175559998,
+      "learning_rate": 1.8042763428499777e-05,
+      "loss": 1.1816,
+      "step": 6236
+    },
+    {
+      "epoch": 2.322067224348757,
+      "grad_norm": 0.16606608033180237,
+      "learning_rate": 1.8042041160742062e-05,
+      "loss": 1.1708,
+      "step": 6237
+    },
+    {
+      "epoch": 2.322439529499366,
+      "grad_norm": 0.17201752960681915,
+      "learning_rate": 1.8041318774203908e-05,
+      "loss": 1.1814,
+      "step": 6238
+    },
+    {
+      "epoch": 2.322811834649975,
+      "grad_norm": 0.16523173451423645,
+      "learning_rate": 1.8040596268895973e-05,
+      "loss": 1.1717,
+      "step": 6239
+    },
+    {
+      "epoch": 2.323184139800584,
+      "grad_norm": 0.16738300025463104,
+      "learning_rate": 1.803987364482894e-05,
+      "loss": 1.1809,
+      "step": 6240
+    },
+    {
+      "epoch": 2.323556444951193,
+      "grad_norm": 0.1602403223514557,
+      "learning_rate": 1.8039150902013478e-05,
+      "loss": 1.1754,
+      "step": 6241
+    },
+    {
+      "epoch": 2.323928750101802,
+      "grad_norm": 0.16947504878044128,
+      "learning_rate": 1.803842804046026e-05,
+      "loss": 1.1869,
+      "step": 6242
+    },
+    {
+      "epoch": 2.3243010552524113,
+      "grad_norm": 0.16907165944576263,
+      "learning_rate": 1.8037705060179965e-05,
+      "loss": 1.1857,
+      "step": 6243
+    },
+    {
+      "epoch": 2.32467336040302,
+      "grad_norm": 0.17978952825069427,
+      "learning_rate": 1.8036981961183273e-05,
+      "loss": 1.1801,
+      "step": 6244
+    },
+    {
+      "epoch": 2.3250456655536293,
+      "grad_norm": 0.16338391602039337,
+      "learning_rate": 1.803625874348086e-05,
+      "loss": 1.1679,
+      "step": 6245
+    },
+    {
+      "epoch": 2.3254179707042386,
+      "grad_norm": 0.17273187637329102,
+      "learning_rate": 1.803553540708341e-05,
+      "loss": 1.1749,
+      "step": 6246
+    },
+    {
+      "epoch": 2.3257902758548474,
+      "grad_norm": 0.17095768451690674,
+      "learning_rate": 1.8034811952001602e-05,
+      "loss": 1.1663,
+      "step": 6247
+    },
+    {
+      "epoch": 2.3261625810054567,
+      "grad_norm": 0.16900064051151276,
+      "learning_rate": 1.803408837824613e-05,
+      "loss": 1.1831,
+      "step": 6248
+    },
+    {
+      "epoch": 2.3265348861560655,
+      "grad_norm": 0.17305012047290802,
+      "learning_rate": 1.8033364685827677e-05,
+      "loss": 1.1898,
+      "step": 6249
+    },
+    {
+      "epoch": 2.326907191306675,
+      "grad_norm": 0.1679958999156952,
+      "learning_rate": 1.8032640874756932e-05,
+      "loss": 1.1713,
+      "step": 6250
+    },
+    {
+      "epoch": 2.327279496457284,
+      "grad_norm": 0.1740444004535675,
+      "learning_rate": 1.8031916945044586e-05,
+      "loss": 1.194,
+      "step": 6251
+    },
+    {
+      "epoch": 2.327651801607893,
+      "grad_norm": 0.1582871377468109,
+      "learning_rate": 1.803119289670133e-05,
+      "loss": 1.1891,
+      "step": 6252
+    },
+    {
+      "epoch": 2.3280241067585017,
+      "grad_norm": 0.17062443494796753,
+      "learning_rate": 1.8030468729737856e-05,
+      "loss": 1.185,
+      "step": 6253
+    },
+    {
+      "epoch": 2.328396411909111,
+      "grad_norm": 0.16043491661548615,
+      "learning_rate": 1.802974444416487e-05,
+      "loss": 1.1798,
+      "step": 6254
+    },
+    {
+      "epoch": 2.32876871705972,
+      "grad_norm": 0.17315945029258728,
+      "learning_rate": 1.8029020039993055e-05,
+      "loss": 1.1855,
+      "step": 6255
+    },
+    {
+      "epoch": 2.329141022210329,
+      "grad_norm": 0.16167216002941132,
+      "learning_rate": 1.802829551723312e-05,
+      "loss": 1.1834,
+      "step": 6256
+    },
+    {
+      "epoch": 2.3295133273609383,
+      "grad_norm": 0.17070269584655762,
+      "learning_rate": 1.8027570875895762e-05,
+      "loss": 1.1805,
+      "step": 6257
+    },
+    {
+      "epoch": 2.329885632511547,
+      "grad_norm": 0.1713981032371521,
+      "learning_rate": 1.802684611599169e-05,
+      "loss": 1.1893,
+      "step": 6258
+    },
+    {
+      "epoch": 2.3302579376621564,
+      "grad_norm": 0.17081809043884277,
+      "learning_rate": 1.80261212375316e-05,
+      "loss": 1.1838,
+      "step": 6259
+    },
+    {
+      "epoch": 2.3306302428127657,
+      "grad_norm": 0.16853542625904083,
+      "learning_rate": 1.8025396240526208e-05,
+      "loss": 1.1858,
+      "step": 6260
+    },
+    {
+      "epoch": 2.3310025479633745,
+      "grad_norm": 0.1681629717350006,
+      "learning_rate": 1.8024671124986218e-05,
+      "loss": 1.177,
+      "step": 6261
+    },
+    {
+      "epoch": 2.3313748531139837,
+      "grad_norm": 0.1574486941099167,
+      "learning_rate": 1.8023945890922334e-05,
+      "loss": 1.1778,
+      "step": 6262
+    },
+    {
+      "epoch": 2.3317471582645926,
+      "grad_norm": 0.16776683926582336,
+      "learning_rate": 1.8023220538345276e-05,
+      "loss": 1.1888,
+      "step": 6263
+    },
+    {
+      "epoch": 2.332119463415202,
+      "grad_norm": 0.16034848988056183,
+      "learning_rate": 1.802249506726575e-05,
+      "loss": 1.1826,
+      "step": 6264
+    },
+    {
+      "epoch": 2.3324917685658106,
+      "grad_norm": 0.16641570627689362,
+      "learning_rate": 1.8021769477694482e-05,
+      "loss": 1.1863,
+      "step": 6265
+    },
+    {
+      "epoch": 2.33286407371642,
+      "grad_norm": 0.16648726165294647,
+      "learning_rate": 1.802104376964218e-05,
+      "loss": 1.1586,
+      "step": 6266
+    },
+    {
+      "epoch": 2.3332363788670287,
+      "grad_norm": 0.16356948018074036,
+      "learning_rate": 1.8020317943119563e-05,
+      "loss": 1.1854,
+      "step": 6267
+    },
+    {
+      "epoch": 2.333608684017638,
+      "grad_norm": 0.16005860269069672,
+      "learning_rate": 1.8019591998137355e-05,
+      "loss": 1.1887,
+      "step": 6268
+    },
+    {
+      "epoch": 2.3339809891682473,
+      "grad_norm": 0.1607915163040161,
+      "learning_rate": 1.8018865934706277e-05,
+      "loss": 1.1765,
+      "step": 6269
+    },
+    {
+      "epoch": 2.334353294318856,
+      "grad_norm": 0.1606864184141159,
+      "learning_rate": 1.801813975283705e-05,
+      "loss": 1.182,
+      "step": 6270
+    },
+    {
+      "epoch": 2.3347255994694653,
+      "grad_norm": 0.17700619995594025,
+      "learning_rate": 1.80174134525404e-05,
+      "loss": 1.1745,
+      "step": 6271
+    },
+    {
+      "epoch": 2.335097904620074,
+      "grad_norm": 0.1743892878293991,
+      "learning_rate": 1.801668703382706e-05,
+      "loss": 1.174,
+      "step": 6272
+    },
+    {
+      "epoch": 2.3354702097706834,
+      "grad_norm": 0.16445204615592957,
+      "learning_rate": 1.8015960496707756e-05,
+      "loss": 1.1716,
+      "step": 6273
+    },
+    {
+      "epoch": 2.3358425149212922,
+      "grad_norm": 0.1924527883529663,
+      "learning_rate": 1.8015233841193218e-05,
+      "loss": 1.1676,
+      "step": 6274
+    },
+    {
+      "epoch": 2.3362148200719015,
+      "grad_norm": 0.19679906964302063,
+      "learning_rate": 1.8014507067294177e-05,
+      "loss": 1.1849,
+      "step": 6275
+    },
+    {
+      "epoch": 2.3365871252225103,
+      "grad_norm": 0.16511952877044678,
+      "learning_rate": 1.8013780175021373e-05,
+      "loss": 1.2015,
+      "step": 6276
+    },
+    {
+      "epoch": 2.3369594303731196,
+      "grad_norm": 0.23659729957580566,
+      "learning_rate": 1.8013053164385538e-05,
+      "loss": 1.1712,
+      "step": 6277
+    },
+    {
+      "epoch": 2.337331735523729,
+      "grad_norm": 0.19166865944862366,
+      "learning_rate": 1.8012326035397407e-05,
+      "loss": 1.1822,
+      "step": 6278
+    },
+    {
+      "epoch": 2.3377040406743377,
+      "grad_norm": 0.1799168735742569,
+      "learning_rate": 1.8011598788067728e-05,
+      "loss": 1.1751,
+      "step": 6279
+    },
+    {
+      "epoch": 2.338076345824947,
+      "grad_norm": 0.16195322573184967,
+      "learning_rate": 1.8010871422407238e-05,
+      "loss": 1.1822,
+      "step": 6280
+    },
+    {
+      "epoch": 2.3384486509755558,
+      "grad_norm": 0.18496288359165192,
+      "learning_rate": 1.8010143938426674e-05,
+      "loss": 1.1882,
+      "step": 6281
+    },
+    {
+      "epoch": 2.338820956126165,
+      "grad_norm": 0.16737127304077148,
+      "learning_rate": 1.800941633613679e-05,
+      "loss": 1.1897,
+      "step": 6282
+    },
+    {
+      "epoch": 2.339193261276774,
+      "grad_norm": 0.16343113780021667,
+      "learning_rate": 1.800868861554833e-05,
+      "loss": 1.1761,
+      "step": 6283
+    },
+    {
+      "epoch": 2.339565566427383,
+      "grad_norm": 0.16952070593833923,
+      "learning_rate": 1.8007960776672043e-05,
+      "loss": 1.1742,
+      "step": 6284
+    },
+    {
+      "epoch": 2.339937871577992,
+      "grad_norm": 0.16918697953224182,
+      "learning_rate": 1.8007232819518675e-05,
+      "loss": 1.1809,
+      "step": 6285
+    },
+    {
+      "epoch": 2.340310176728601,
+      "grad_norm": 0.16563785076141357,
+      "learning_rate": 1.800650474409898e-05,
+      "loss": 1.1709,
+      "step": 6286
+    },
+    {
+      "epoch": 2.3406824818792105,
+      "grad_norm": 0.1735762506723404,
+      "learning_rate": 1.8005776550423718e-05,
+      "loss": 1.1761,
+      "step": 6287
+    },
+    {
+      "epoch": 2.3410547870298193,
+      "grad_norm": 0.16570459306240082,
+      "learning_rate": 1.8005048238503633e-05,
+      "loss": 1.1778,
+      "step": 6288
+    },
+    {
+      "epoch": 2.3414270921804285,
+      "grad_norm": 0.16612395644187927,
+      "learning_rate": 1.800431980834949e-05,
+      "loss": 1.1772,
+      "step": 6289
+    },
+    {
+      "epoch": 2.3417993973310374,
+      "grad_norm": 0.16212446987628937,
+      "learning_rate": 1.8003591259972047e-05,
+      "loss": 1.1854,
+      "step": 6290
+    },
+    {
+      "epoch": 2.3421717024816466,
+      "grad_norm": 0.1619550585746765,
+      "learning_rate": 1.8002862593382063e-05,
+      "loss": 1.1682,
+      "step": 6291
+    },
+    {
+      "epoch": 2.3425440076322555,
+      "grad_norm": 0.16271120309829712,
+      "learning_rate": 1.80021338085903e-05,
+      "loss": 1.1646,
+      "step": 6292
+    },
+    {
+      "epoch": 2.3429163127828647,
+      "grad_norm": 0.1650598645210266,
+      "learning_rate": 1.8001404905607523e-05,
+      "loss": 1.189,
+      "step": 6293
+    },
+    {
+      "epoch": 2.3432886179334735,
+      "grad_norm": 0.16364504396915436,
+      "learning_rate": 1.8000675884444495e-05,
+      "loss": 1.1838,
+      "step": 6294
+    },
+    {
+      "epoch": 2.343660923084083,
+      "grad_norm": 0.16529352962970734,
+      "learning_rate": 1.7999946745111993e-05,
+      "loss": 1.1908,
+      "step": 6295
+    },
+    {
+      "epoch": 2.344033228234692,
+      "grad_norm": 0.1709616482257843,
+      "learning_rate": 1.7999217487620773e-05,
+      "loss": 1.1731,
+      "step": 6296
+    },
+    {
+      "epoch": 2.344405533385301,
+      "grad_norm": 0.16353869438171387,
+      "learning_rate": 1.7998488111981616e-05,
+      "loss": 1.1748,
+      "step": 6297
+    },
+    {
+      "epoch": 2.34477783853591,
+      "grad_norm": 0.16740430891513824,
+      "learning_rate": 1.799775861820529e-05,
+      "loss": 1.1674,
+      "step": 6298
+    },
+    {
+      "epoch": 2.345150143686519,
+      "grad_norm": 0.1658317595720291,
+      "learning_rate": 1.7997029006302572e-05,
+      "loss": 1.1709,
+      "step": 6299
+    },
+    {
+      "epoch": 2.3455224488371282,
+      "grad_norm": 0.1644759625196457,
+      "learning_rate": 1.799629927628424e-05,
+      "loss": 1.184,
+      "step": 6300
+    },
+    {
+      "epoch": 2.345894753987737,
+      "grad_norm": 0.1665249615907669,
+      "learning_rate": 1.7995569428161066e-05,
+      "loss": 1.1826,
+      "step": 6301
+    },
+    {
+      "epoch": 2.3462670591383463,
+      "grad_norm": 0.1632251888513565,
+      "learning_rate": 1.7994839461943834e-05,
+      "loss": 1.1614,
+      "step": 6302
+    },
+    {
+      "epoch": 2.346639364288955,
+      "grad_norm": 0.16572819650173187,
+      "learning_rate": 1.7994109377643326e-05,
+      "loss": 1.1768,
+      "step": 6303
+    },
+    {
+      "epoch": 2.3470116694395644,
+      "grad_norm": 0.16670545935630798,
+      "learning_rate": 1.7993379175270323e-05,
+      "loss": 1.1751,
+      "step": 6304
+    },
+    {
+      "epoch": 2.3473839745901737,
+      "grad_norm": 0.16433420777320862,
+      "learning_rate": 1.7992648854835607e-05,
+      "loss": 1.1997,
+      "step": 6305
+    },
+    {
+      "epoch": 2.3477562797407825,
+      "grad_norm": 0.16783735156059265,
+      "learning_rate": 1.7991918416349977e-05,
+      "loss": 1.1837,
+      "step": 6306
+    },
+    {
+      "epoch": 2.3481285848913918,
+      "grad_norm": 0.166385218501091,
+      "learning_rate": 1.799118785982421e-05,
+      "loss": 1.196,
+      "step": 6307
+    },
+    {
+      "epoch": 2.3485008900420006,
+      "grad_norm": 0.1664101630449295,
+      "learning_rate": 1.79904571852691e-05,
+      "loss": 1.1839,
+      "step": 6308
+    },
+    {
+      "epoch": 2.34887319519261,
+      "grad_norm": 0.16921697556972504,
+      "learning_rate": 1.7989726392695438e-05,
+      "loss": 1.1826,
+      "step": 6309
+    },
+    {
+      "epoch": 2.3492455003432187,
+      "grad_norm": 0.16790980100631714,
+      "learning_rate": 1.798899548211402e-05,
+      "loss": 1.1739,
+      "step": 6310
+    },
+    {
+      "epoch": 2.349617805493828,
+      "grad_norm": 0.16884508728981018,
+      "learning_rate": 1.798826445353564e-05,
+      "loss": 1.1825,
+      "step": 6311
+    },
+    {
+      "epoch": 2.3499901106444367,
+      "grad_norm": 0.16576768457889557,
+      "learning_rate": 1.7987533306971093e-05,
+      "loss": 1.1503,
+      "step": 6312
+    },
+    {
+      "epoch": 2.350362415795046,
+      "grad_norm": 0.1681016981601715,
+      "learning_rate": 1.798680204243118e-05,
+      "loss": 1.1753,
+      "step": 6313
+    },
+    {
+      "epoch": 2.3507347209456553,
+      "grad_norm": 0.17938603460788727,
+      "learning_rate": 1.7986070659926705e-05,
+      "loss": 1.1865,
+      "step": 6314
+    },
+    {
+      "epoch": 2.351107026096264,
+      "grad_norm": 0.16416122019290924,
+      "learning_rate": 1.798533915946847e-05,
+      "loss": 1.1693,
+      "step": 6315
+    },
+    {
+      "epoch": 2.3514793312468734,
+      "grad_norm": 0.1787365823984146,
+      "learning_rate": 1.7984607541067272e-05,
+      "loss": 1.174,
+      "step": 6316
+    },
+    {
+      "epoch": 2.351851636397482,
+      "grad_norm": 0.16736140847206116,
+      "learning_rate": 1.7983875804733925e-05,
+      "loss": 1.1772,
+      "step": 6317
+    },
+    {
+      "epoch": 2.3522239415480914,
+      "grad_norm": 0.17561602592468262,
+      "learning_rate": 1.798314395047923e-05,
+      "loss": 1.1793,
+      "step": 6318
+    },
+    {
+      "epoch": 2.3525962466987003,
+      "grad_norm": 0.16405464708805084,
+      "learning_rate": 1.7982411978314e-05,
+      "loss": 1.1866,
+      "step": 6319
+    },
+    {
+      "epoch": 2.3529685518493095,
+      "grad_norm": 0.1719312071800232,
+      "learning_rate": 1.798167988824905e-05,
+      "loss": 1.1743,
+      "step": 6320
+    },
+    {
+      "epoch": 2.3533408569999184,
+      "grad_norm": 0.17220060527324677,
+      "learning_rate": 1.7980947680295187e-05,
+      "loss": 1.1719,
+      "step": 6321
+    },
+    {
+      "epoch": 2.3537131621505276,
+      "grad_norm": 0.1724272072315216,
+      "learning_rate": 1.7980215354463223e-05,
+      "loss": 1.1777,
+      "step": 6322
+    },
+    {
+      "epoch": 2.354085467301137,
+      "grad_norm": 0.17141182720661163,
+      "learning_rate": 1.7979482910763984e-05,
+      "loss": 1.1822,
+      "step": 6323
+    },
+    {
+      "epoch": 2.3544577724517457,
+      "grad_norm": 0.18284828960895538,
+      "learning_rate": 1.7978750349208284e-05,
+      "loss": 1.1885,
+      "step": 6324
+    },
+    {
+      "epoch": 2.354830077602355,
+      "grad_norm": 0.16586068272590637,
+      "learning_rate": 1.797801766980694e-05,
+      "loss": 1.1677,
+      "step": 6325
+    },
+    {
+      "epoch": 2.355202382752964,
+      "grad_norm": 0.1680738776922226,
+      "learning_rate": 1.7977284872570775e-05,
+      "loss": 1.1888,
+      "step": 6326
+    },
+    {
+      "epoch": 2.355574687903573,
+      "grad_norm": 0.16332261264324188,
+      "learning_rate": 1.7976551957510614e-05,
+      "loss": 1.1771,
+      "step": 6327
+    },
+    {
+      "epoch": 2.355946993054182,
+      "grad_norm": 0.16689446568489075,
+      "learning_rate": 1.797581892463728e-05,
+      "loss": 1.1698,
+      "step": 6328
+    },
+    {
+      "epoch": 2.356319298204791,
+      "grad_norm": 0.17073221504688263,
+      "learning_rate": 1.7975085773961606e-05,
+      "loss": 1.1931,
+      "step": 6329
+    },
+    {
+      "epoch": 2.3566916033554,
+      "grad_norm": 0.16630737483501434,
+      "learning_rate": 1.797435250549441e-05,
+      "loss": 1.1698,
+      "step": 6330
+    },
+    {
+      "epoch": 2.357063908506009,
+      "grad_norm": 0.19237372279167175,
+      "learning_rate": 1.797361911924653e-05,
+      "loss": 1.1976,
+      "step": 6331
+    },
+    {
+      "epoch": 2.3574362136566185,
+      "grad_norm": 0.1919546127319336,
+      "learning_rate": 1.7972885615228796e-05,
+      "loss": 1.1915,
+      "step": 6332
+    },
+    {
+      "epoch": 2.3578085188072273,
+      "grad_norm": 0.16632108390331268,
+      "learning_rate": 1.797215199345204e-05,
+      "loss": 1.187,
+      "step": 6333
+    },
+    {
+      "epoch": 2.3581808239578366,
+      "grad_norm": 0.26524388790130615,
+      "learning_rate": 1.7971418253927103e-05,
+      "loss": 1.1899,
+      "step": 6334
+    },
+    {
+      "epoch": 2.3585531291084454,
+      "grad_norm": 0.20467346906661987,
+      "learning_rate": 1.7970684396664814e-05,
+      "loss": 1.1828,
+      "step": 6335
+    },
+    {
+      "epoch": 2.3589254342590547,
+      "grad_norm": 0.18540099263191223,
+      "learning_rate": 1.7969950421676022e-05,
+      "loss": 1.1699,
+      "step": 6336
+    },
+    {
+      "epoch": 2.3592977394096635,
+      "grad_norm": 0.17078468203544617,
+      "learning_rate": 1.796921632897156e-05,
+      "loss": 1.1846,
+      "step": 6337
+    },
+    {
+      "epoch": 2.3596700445602727,
+      "grad_norm": 0.19237884879112244,
+      "learning_rate": 1.796848211856227e-05,
+      "loss": 1.1847,
+      "step": 6338
+    },
+    {
+      "epoch": 2.3600423497108816,
+      "grad_norm": 0.1824292242527008,
+      "learning_rate": 1.7967747790459006e-05,
+      "loss": 1.1699,
+      "step": 6339
+    },
+    {
+      "epoch": 2.360414654861491,
+      "grad_norm": 0.17438608407974243,
+      "learning_rate": 1.7967013344672602e-05,
+      "loss": 1.1848,
+      "step": 6340
+    },
+    {
+      "epoch": 2.3607869600121,
+      "grad_norm": 0.16449159383773804,
+      "learning_rate": 1.7966278781213914e-05,
+      "loss": 1.1741,
+      "step": 6341
+    },
+    {
+      "epoch": 2.361159265162709,
+      "grad_norm": 0.17680013179779053,
+      "learning_rate": 1.7965544100093785e-05,
+      "loss": 1.1611,
+      "step": 6342
+    },
+    {
+      "epoch": 2.361531570313318,
+      "grad_norm": 0.18226385116577148,
+      "learning_rate": 1.796480930132307e-05,
+      "loss": 1.1873,
+      "step": 6343
+    },
+    {
+      "epoch": 2.361903875463927,
+      "grad_norm": 0.1675233244895935,
+      "learning_rate": 1.7964074384912624e-05,
+      "loss": 1.1864,
+      "step": 6344
+    },
+    {
+      "epoch": 2.3622761806145363,
+      "grad_norm": 0.17054075002670288,
+      "learning_rate": 1.7963339350873296e-05,
+      "loss": 1.1863,
+      "step": 6345
+    },
+    {
+      "epoch": 2.362648485765145,
+      "grad_norm": 0.1669616848230362,
+      "learning_rate": 1.7962604199215946e-05,
+      "loss": 1.1782,
+      "step": 6346
+    },
+    {
+      "epoch": 2.3630207909157543,
+      "grad_norm": 0.1679152250289917,
+      "learning_rate": 1.7961868929951432e-05,
+      "loss": 1.1859,
+      "step": 6347
+    },
+    {
+      "epoch": 2.363393096066363,
+      "grad_norm": 0.1678704470396042,
+      "learning_rate": 1.7961133543090617e-05,
+      "loss": 1.1759,
+      "step": 6348
+    },
+    {
+      "epoch": 2.3637654012169724,
+      "grad_norm": 0.15599095821380615,
+      "learning_rate": 1.7960398038644356e-05,
+      "loss": 1.1686,
+      "step": 6349
+    },
+    {
+      "epoch": 2.3641377063675817,
+      "grad_norm": 0.16793273389339447,
+      "learning_rate": 1.7959662416623516e-05,
+      "loss": 1.1858,
+      "step": 6350
+    },
+    {
+      "epoch": 2.3645100115181905,
+      "grad_norm": 0.16774815320968628,
+      "learning_rate": 1.795892667703896e-05,
+      "loss": 1.1846,
+      "step": 6351
+    },
+    {
+      "epoch": 2.3648823166687998,
+      "grad_norm": 0.16648389399051666,
+      "learning_rate": 1.795819081990156e-05,
+      "loss": 1.1811,
+      "step": 6352
+    },
+    {
+      "epoch": 2.3652546218194086,
+      "grad_norm": 0.16627708077430725,
+      "learning_rate": 1.795745484522218e-05,
+      "loss": 1.1804,
+      "step": 6353
+    },
+    {
+      "epoch": 2.365626926970018,
+      "grad_norm": 0.16126348078250885,
+      "learning_rate": 1.795671875301169e-05,
+      "loss": 1.1786,
+      "step": 6354
+    },
+    {
+      "epoch": 2.3659992321206267,
+      "grad_norm": 0.16175805032253265,
+      "learning_rate": 1.7955982543280965e-05,
+      "loss": 1.1678,
+      "step": 6355
+    },
+    {
+      "epoch": 2.366371537271236,
+      "grad_norm": 0.16892601549625397,
+      "learning_rate": 1.7955246216040872e-05,
+      "loss": 1.1723,
+      "step": 6356
+    },
+    {
+      "epoch": 2.3667438424218448,
+      "grad_norm": 0.1636088341474533,
+      "learning_rate": 1.79545097713023e-05,
+      "loss": 1.1844,
+      "step": 6357
+    },
+    {
+      "epoch": 2.367116147572454,
+      "grad_norm": 0.16723226010799408,
+      "learning_rate": 1.795377320907611e-05,
+      "loss": 1.1886,
+      "step": 6358
+    },
+    {
+      "epoch": 2.3674884527230633,
+      "grad_norm": 0.17732274532318115,
+      "learning_rate": 1.7953036529373194e-05,
+      "loss": 1.1981,
+      "step": 6359
+    },
+    {
+      "epoch": 2.367860757873672,
+      "grad_norm": 0.15954697132110596,
+      "learning_rate": 1.795229973220443e-05,
+      "loss": 1.1924,
+      "step": 6360
+    },
+    {
+      "epoch": 2.3682330630242814,
+      "grad_norm": 0.1688387095928192,
+      "learning_rate": 1.795156281758069e-05,
+      "loss": 1.1782,
+      "step": 6361
+    },
+    {
+      "epoch": 2.36860536817489,
+      "grad_norm": 0.16727487742900848,
+      "learning_rate": 1.7950825785512873e-05,
+      "loss": 1.1868,
+      "step": 6362
+    },
+    {
+      "epoch": 2.3689776733254995,
+      "grad_norm": 0.1692897230386734,
+      "learning_rate": 1.7950088636011853e-05,
+      "loss": 1.1952,
+      "step": 6363
+    },
+    {
+      "epoch": 2.3693499784761083,
+      "grad_norm": 0.15988357365131378,
+      "learning_rate": 1.7949351369088526e-05,
+      "loss": 1.1724,
+      "step": 6364
+    },
+    {
+      "epoch": 2.3697222836267176,
+      "grad_norm": 0.16070933640003204,
+      "learning_rate": 1.7948613984753777e-05,
+      "loss": 1.1731,
+      "step": 6365
+    },
+    {
+      "epoch": 2.3700945887773264,
+      "grad_norm": 0.16724231839179993,
+      "learning_rate": 1.7947876483018498e-05,
+      "loss": 1.1779,
+      "step": 6366
+    },
+    {
+      "epoch": 2.3704668939279356,
+      "grad_norm": 0.16751320660114288,
+      "learning_rate": 1.7947138863893582e-05,
+      "loss": 1.1952,
+      "step": 6367
+    },
+    {
+      "epoch": 2.370839199078545,
+      "grad_norm": 0.16353389620780945,
+      "learning_rate": 1.7946401127389928e-05,
+      "loss": 1.1899,
+      "step": 6368
+    },
+    {
+      "epoch": 2.3712115042291537,
+      "grad_norm": 0.16663357615470886,
+      "learning_rate": 1.7945663273518423e-05,
+      "loss": 1.2004,
+      "step": 6369
+    },
+    {
+      "epoch": 2.371583809379763,
+      "grad_norm": 0.1695907711982727,
+      "learning_rate": 1.7944925302289972e-05,
+      "loss": 1.158,
+      "step": 6370
+    },
+    {
+      "epoch": 2.371956114530372,
+      "grad_norm": 0.16809728741645813,
+      "learning_rate": 1.794418721371547e-05,
+      "loss": 1.1689,
+      "step": 6371
+    },
+    {
+      "epoch": 2.372328419680981,
+      "grad_norm": 0.1606246680021286,
+      "learning_rate": 1.7943449007805824e-05,
+      "loss": 1.1615,
+      "step": 6372
+    },
+    {
+      "epoch": 2.3727007248315903,
+      "grad_norm": 0.1651766151189804,
+      "learning_rate": 1.7942710684571934e-05,
+      "loss": 1.183,
+      "step": 6373
+    },
+    {
+      "epoch": 2.373073029982199,
+      "grad_norm": 0.16688674688339233,
+      "learning_rate": 1.794197224402471e-05,
+      "loss": 1.1735,
+      "step": 6374
+    },
+    {
+      "epoch": 2.3734453351328084,
+      "grad_norm": 0.16931474208831787,
+      "learning_rate": 1.7941233686175052e-05,
+      "loss": 1.169,
+      "step": 6375
+    },
+    {
+      "epoch": 2.3738176402834172,
+      "grad_norm": 0.16354526579380035,
+      "learning_rate": 1.7940495011033866e-05,
+      "loss": 1.1828,
+      "step": 6376
+    },
+    {
+      "epoch": 2.3741899454340265,
+      "grad_norm": 0.16403120756149292,
+      "learning_rate": 1.7939756218612072e-05,
+      "loss": 1.1755,
+      "step": 6377
+    },
+    {
+      "epoch": 2.3745622505846353,
+      "grad_norm": 0.16923734545707703,
+      "learning_rate": 1.7939017308920575e-05,
+      "loss": 1.1704,
+      "step": 6378
+    },
+    {
+      "epoch": 2.3749345557352446,
+      "grad_norm": 0.16847112774848938,
+      "learning_rate": 1.793827828197029e-05,
+      "loss": 1.1739,
+      "step": 6379
+    },
+    {
+      "epoch": 2.3753068608858534,
+      "grad_norm": 0.17019587755203247,
+      "learning_rate": 1.7937539137772134e-05,
+      "loss": 1.1815,
+      "step": 6380
+    },
+    {
+      "epoch": 2.3756791660364627,
+      "grad_norm": 0.16363418102264404,
+      "learning_rate": 1.7936799876337022e-05,
+      "loss": 1.1707,
+      "step": 6381
+    },
+    {
+      "epoch": 2.376051471187072,
+      "grad_norm": 0.16532693803310394,
+      "learning_rate": 1.7936060497675875e-05,
+      "loss": 1.174,
+      "step": 6382
+    },
+    {
+      "epoch": 2.3764237763376808,
+      "grad_norm": 0.168051615357399,
+      "learning_rate": 1.793532100179961e-05,
+      "loss": 1.1609,
+      "step": 6383
+    },
+    {
+      "epoch": 2.37679608148829,
+      "grad_norm": 0.16365954279899597,
+      "learning_rate": 1.7934581388719158e-05,
+      "loss": 1.1779,
+      "step": 6384
+    },
+    {
+      "epoch": 2.377168386638899,
+      "grad_norm": 0.16921496391296387,
+      "learning_rate": 1.7933841658445432e-05,
+      "loss": 1.1967,
+      "step": 6385
+    },
+    {
+      "epoch": 2.377540691789508,
+      "grad_norm": 0.16775216162204742,
+      "learning_rate": 1.7933101810989363e-05,
+      "loss": 1.1698,
+      "step": 6386
+    },
+    {
+      "epoch": 2.377912996940117,
+      "grad_norm": 0.16097751259803772,
+      "learning_rate": 1.793236184636188e-05,
+      "loss": 1.1879,
+      "step": 6387
+    },
+    {
+      "epoch": 2.378285302090726,
+      "grad_norm": 0.1658683568239212,
+      "learning_rate": 1.7931621764573907e-05,
+      "loss": 1.1815,
+      "step": 6388
+    },
+    {
+      "epoch": 2.378657607241335,
+      "grad_norm": 0.16662012040615082,
+      "learning_rate": 1.793088156563638e-05,
+      "loss": 1.1767,
+      "step": 6389
+    },
+    {
+      "epoch": 2.3790299123919443,
+      "grad_norm": 0.16364309191703796,
+      "learning_rate": 1.7930141249560235e-05,
+      "loss": 1.1758,
+      "step": 6390
+    },
+    {
+      "epoch": 2.3794022175425535,
+      "grad_norm": 0.17829427123069763,
+      "learning_rate": 1.7929400816356394e-05,
+      "loss": 1.1832,
+      "step": 6391
+    },
+    {
+      "epoch": 2.3797745226931624,
+      "grad_norm": 0.16619321703910828,
+      "learning_rate": 1.7928660266035804e-05,
+      "loss": 1.1796,
+      "step": 6392
+    },
+    {
+      "epoch": 2.3801468278437716,
+      "grad_norm": 0.16665421426296234,
+      "learning_rate": 1.79279195986094e-05,
+      "loss": 1.1768,
+      "step": 6393
+    },
+    {
+      "epoch": 2.3805191329943804,
+      "grad_norm": 0.17309972643852234,
+      "learning_rate": 1.792717881408812e-05,
+      "loss": 1.1811,
+      "step": 6394
+    },
+    {
+      "epoch": 2.3808914381449897,
+      "grad_norm": 0.1681647002696991,
+      "learning_rate": 1.7926437912482905e-05,
+      "loss": 1.1738,
+      "step": 6395
+    },
+    {
+      "epoch": 2.3812637432955985,
+      "grad_norm": 0.17703039944171906,
+      "learning_rate": 1.79256968938047e-05,
+      "loss": 1.1759,
+      "step": 6396
+    },
+    {
+      "epoch": 2.381636048446208,
+      "grad_norm": 0.16433195769786835,
+      "learning_rate": 1.792495575806445e-05,
+      "loss": 1.1746,
+      "step": 6397
+    },
+    {
+      "epoch": 2.3820083535968166,
+      "grad_norm": 0.1649874895811081,
+      "learning_rate": 1.7924214505273102e-05,
+      "loss": 1.1785,
+      "step": 6398
+    },
+    {
+      "epoch": 2.382380658747426,
+      "grad_norm": 0.17650985717773438,
+      "learning_rate": 1.79234731354416e-05,
+      "loss": 1.1714,
+      "step": 6399
+    },
+    {
+      "epoch": 2.382752963898035,
+      "grad_norm": 0.19608263671398163,
+      "learning_rate": 1.7922731648580902e-05,
+      "loss": 1.1805,
+      "step": 6400
+    },
+    {
+      "epoch": 2.383125269048644,
+      "grad_norm": 0.18923339247703552,
+      "learning_rate": 1.7921990044701952e-05,
+      "loss": 1.1701,
+      "step": 6401
+    },
+    {
+      "epoch": 2.3834975741992532,
+      "grad_norm": 0.1707700490951538,
+      "learning_rate": 1.7921248323815703e-05,
+      "loss": 1.1829,
+      "step": 6402
+    },
+    {
+      "epoch": 2.383869879349862,
+      "grad_norm": 0.16345664858818054,
+      "learning_rate": 1.7920506485933117e-05,
+      "loss": 1.1881,
+      "step": 6403
+    },
+    {
+      "epoch": 2.3842421845004713,
+      "grad_norm": 0.166402205824852,
+      "learning_rate": 1.7919764531065147e-05,
+      "loss": 1.1603,
+      "step": 6404
+    },
+    {
+      "epoch": 2.38461448965108,
+      "grad_norm": 0.17079654335975647,
+      "learning_rate": 1.7919022459222754e-05,
+      "loss": 1.1882,
+      "step": 6405
+    },
+    {
+      "epoch": 2.3849867948016894,
+      "grad_norm": 0.16907550394535065,
+      "learning_rate": 1.791828027041689e-05,
+      "loss": 1.1921,
+      "step": 6406
+    },
+    {
+      "epoch": 2.3853590999522982,
+      "grad_norm": 0.15836122632026672,
+      "learning_rate": 1.791753796465853e-05,
+      "loss": 1.1894,
+      "step": 6407
+    },
+    {
+      "epoch": 2.3857314051029075,
+      "grad_norm": 0.1576787382364273,
+      "learning_rate": 1.791679554195863e-05,
+      "loss": 1.1631,
+      "step": 6408
+    },
+    {
+      "epoch": 2.3861037102535168,
+      "grad_norm": 0.1628400981426239,
+      "learning_rate": 1.7916053002328152e-05,
+      "loss": 1.1717,
+      "step": 6409
+    },
+    {
+      "epoch": 2.3864760154041256,
+      "grad_norm": 0.16520801186561584,
+      "learning_rate": 1.7915310345778072e-05,
+      "loss": 1.1745,
+      "step": 6410
+    },
+    {
+      "epoch": 2.386848320554735,
+      "grad_norm": 0.167642742395401,
+      "learning_rate": 1.7914567572319352e-05,
+      "loss": 1.1679,
+      "step": 6411
+    },
+    {
+      "epoch": 2.3872206257053437,
+      "grad_norm": 0.16473732888698578,
+      "learning_rate": 1.791382468196297e-05,
+      "loss": 1.17,
+      "step": 6412
+    },
+    {
+      "epoch": 2.387592930855953,
+      "grad_norm": 0.17128442227840424,
+      "learning_rate": 1.791308167471989e-05,
+      "loss": 1.1791,
+      "step": 6413
+    },
+    {
+      "epoch": 2.3879652360065617,
+      "grad_norm": 0.17826206982135773,
+      "learning_rate": 1.791233855060109e-05,
+      "loss": 1.1615,
+      "step": 6414
+    },
+    {
+      "epoch": 2.388337541157171,
+      "grad_norm": 0.24664366245269775,
+      "learning_rate": 1.7911595309617552e-05,
+      "loss": 1.1762,
+      "step": 6415
+    },
+    {
+      "epoch": 2.38870984630778,
+      "grad_norm": 0.2372390776872635,
+      "learning_rate": 1.7910851951780244e-05,
+      "loss": 1.1696,
+      "step": 6416
+    },
+    {
+      "epoch": 2.389082151458389,
+      "grad_norm": 0.18587949872016907,
+      "learning_rate": 1.791010847710015e-05,
+      "loss": 1.1784,
+      "step": 6417
+    },
+    {
+      "epoch": 2.3894544566089984,
+      "grad_norm": 0.16316647827625275,
+      "learning_rate": 1.7909364885588247e-05,
+      "loss": 1.1808,
+      "step": 6418
+    },
+    {
+      "epoch": 2.389826761759607,
+      "grad_norm": 0.19883513450622559,
+      "learning_rate": 1.7908621177255523e-05,
+      "loss": 1.1763,
+      "step": 6419
+    },
+    {
+      "epoch": 2.3901990669102164,
+      "grad_norm": 0.1844921112060547,
+      "learning_rate": 1.790787735211296e-05,
+      "loss": 1.1791,
+      "step": 6420
+    },
+    {
+      "epoch": 2.3905713720608253,
+      "grad_norm": 0.17219749093055725,
+      "learning_rate": 1.7907133410171548e-05,
+      "loss": 1.1735,
+      "step": 6421
+    },
+    {
+      "epoch": 2.3909436772114345,
+      "grad_norm": 0.16659492254257202,
+      "learning_rate": 1.790638935144227e-05,
+      "loss": 1.1776,
+      "step": 6422
+    },
+    {
+      "epoch": 2.3913159823620433,
+      "grad_norm": 0.1710551679134369,
+      "learning_rate": 1.7905645175936116e-05,
+      "loss": 1.18,
+      "step": 6423
+    },
+    {
+      "epoch": 2.3916882875126526,
+      "grad_norm": 0.17188216745853424,
+      "learning_rate": 1.7904900883664078e-05,
+      "loss": 1.1724,
+      "step": 6424
+    },
+    {
+      "epoch": 2.3920605926632614,
+      "grad_norm": 0.1644524782896042,
+      "learning_rate": 1.790415647463715e-05,
+      "loss": 1.1819,
+      "step": 6425
+    },
+    {
+      "epoch": 2.3924328978138707,
+      "grad_norm": 0.16760697960853577,
+      "learning_rate": 1.790341194886633e-05,
+      "loss": 1.1836,
+      "step": 6426
+    },
+    {
+      "epoch": 2.39280520296448,
+      "grad_norm": 0.16618818044662476,
+      "learning_rate": 1.790266730636261e-05,
+      "loss": 1.192,
+      "step": 6427
+    },
+    {
+      "epoch": 2.393177508115089,
+      "grad_norm": 0.17228281497955322,
+      "learning_rate": 1.790192254713699e-05,
+      "loss": 1.1891,
+      "step": 6428
+    },
+    {
+      "epoch": 2.393549813265698,
+      "grad_norm": 0.16444258391857147,
+      "learning_rate": 1.7901177671200472e-05,
+      "loss": 1.1802,
+      "step": 6429
+    },
+    {
+      "epoch": 2.393922118416307,
+      "grad_norm": 0.16610148549079895,
+      "learning_rate": 1.790043267856405e-05,
+      "loss": 1.1583,
+      "step": 6430
+    },
+    {
+      "epoch": 2.394294423566916,
+      "grad_norm": 0.16486294567584991,
+      "learning_rate": 1.789968756923874e-05,
+      "loss": 1.1835,
+      "step": 6431
+    },
+    {
+      "epoch": 2.394666728717525,
+      "grad_norm": 0.1681571900844574,
+      "learning_rate": 1.7898942343235535e-05,
+      "loss": 1.1903,
+      "step": 6432
+    },
+    {
+      "epoch": 2.395039033868134,
+      "grad_norm": 0.16249068081378937,
+      "learning_rate": 1.7898197000565448e-05,
+      "loss": 1.1848,
+      "step": 6433
+    },
+    {
+      "epoch": 2.395411339018743,
+      "grad_norm": 0.16723275184631348,
+      "learning_rate": 1.789745154123949e-05,
+      "loss": 1.1678,
+      "step": 6434
+    },
+    {
+      "epoch": 2.3957836441693523,
+      "grad_norm": 0.16536808013916016,
+      "learning_rate": 1.7896705965268668e-05,
+      "loss": 1.1744,
+      "step": 6435
+    },
+    {
+      "epoch": 2.3961559493199616,
+      "grad_norm": 0.16059263050556183,
+      "learning_rate": 1.7895960272663992e-05,
+      "loss": 1.1798,
+      "step": 6436
+    },
+    {
+      "epoch": 2.3965282544705704,
+      "grad_norm": 0.1607469618320465,
+      "learning_rate": 1.789521446343648e-05,
+      "loss": 1.1797,
+      "step": 6437
+    },
+    {
+      "epoch": 2.3969005596211796,
+      "grad_norm": 0.1656348556280136,
+      "learning_rate": 1.7894468537597146e-05,
+      "loss": 1.1781,
+      "step": 6438
+    },
+    {
+      "epoch": 2.3972728647717885,
+      "grad_norm": 0.16378162801265717,
+      "learning_rate": 1.7893722495157004e-05,
+      "loss": 1.1878,
+      "step": 6439
+    },
+    {
+      "epoch": 2.3976451699223977,
+      "grad_norm": 0.1643425077199936,
+      "learning_rate": 1.789297633612708e-05,
+      "loss": 1.1851,
+      "step": 6440
+    },
+    {
+      "epoch": 2.3980174750730066,
+      "grad_norm": 0.16220180690288544,
+      "learning_rate": 1.789223006051839e-05,
+      "loss": 1.1757,
+      "step": 6441
+    },
+    {
+      "epoch": 2.398389780223616,
+      "grad_norm": 0.16219714283943176,
+      "learning_rate": 1.7891483668341955e-05,
+      "loss": 1.1934,
+      "step": 6442
+    },
+    {
+      "epoch": 2.3987620853742246,
+      "grad_norm": 0.16055887937545776,
+      "learning_rate": 1.7890737159608803e-05,
+      "loss": 1.1712,
+      "step": 6443
+    },
+    {
+      "epoch": 2.399134390524834,
+      "grad_norm": 0.16199053823947906,
+      "learning_rate": 1.788999053432996e-05,
+      "loss": 1.1757,
+      "step": 6444
+    },
+    {
+      "epoch": 2.399506695675443,
+      "grad_norm": 0.1576344072818756,
+      "learning_rate": 1.7889243792516452e-05,
+      "loss": 1.1713,
+      "step": 6445
+    },
+    {
+      "epoch": 2.399879000826052,
+      "grad_norm": 0.17367145419120789,
+      "learning_rate": 1.7888496934179308e-05,
+      "loss": 1.1757,
+      "step": 6446
+    },
+    {
+      "epoch": 2.4002513059766613,
+      "grad_norm": 0.1603989452123642,
+      "learning_rate": 1.7887749959329555e-05,
+      "loss": 1.1654,
+      "step": 6447
+    },
+    {
+      "epoch": 2.40062361112727,
+      "grad_norm": 0.1615387499332428,
+      "learning_rate": 1.7887002867978234e-05,
+      "loss": 1.1664,
+      "step": 6448
+    },
+    {
+      "epoch": 2.4009959162778793,
+      "grad_norm": 0.1623525768518448,
+      "learning_rate": 1.7886255660136376e-05,
+      "loss": 1.1547,
+      "step": 6449
+    },
+    {
+      "epoch": 2.401368221428488,
+      "grad_norm": 0.16615155339241028,
+      "learning_rate": 1.7885508335815013e-05,
+      "loss": 1.1797,
+      "step": 6450
+    },
+    {
+      "epoch": 2.4017405265790974,
+      "grad_norm": 0.16319416463375092,
+      "learning_rate": 1.788476089502519e-05,
+      "loss": 1.1662,
+      "step": 6451
+    },
+    {
+      "epoch": 2.4021128317297062,
+      "grad_norm": 0.16049784421920776,
+      "learning_rate": 1.7884013337777944e-05,
+      "loss": 1.176,
+      "step": 6452
+    },
+    {
+      "epoch": 2.4024851368803155,
+      "grad_norm": 0.1557374745607376,
+      "learning_rate": 1.7883265664084317e-05,
+      "loss": 1.1815,
+      "step": 6453
+    },
+    {
+      "epoch": 2.4028574420309248,
+      "grad_norm": 0.17005103826522827,
+      "learning_rate": 1.7882517873955344e-05,
+      "loss": 1.1816,
+      "step": 6454
+    },
+    {
+      "epoch": 2.4032297471815336,
+      "grad_norm": 0.16482587158679962,
+      "learning_rate": 1.7881769967402083e-05,
+      "loss": 1.1879,
+      "step": 6455
+    },
+    {
+      "epoch": 2.403602052332143,
+      "grad_norm": 0.16888852417469025,
+      "learning_rate": 1.788102194443557e-05,
+      "loss": 1.1853,
+      "step": 6456
+    },
+    {
+      "epoch": 2.4039743574827517,
+      "grad_norm": 0.16261079907417297,
+      "learning_rate": 1.7880273805066864e-05,
+      "loss": 1.1675,
+      "step": 6457
+    },
+    {
+      "epoch": 2.404346662633361,
+      "grad_norm": 0.16388806700706482,
+      "learning_rate": 1.7879525549307004e-05,
+      "loss": 1.178,
+      "step": 6458
+    },
+    {
+      "epoch": 2.4047189677839698,
+      "grad_norm": 0.16773760318756104,
+      "learning_rate": 1.7878777177167047e-05,
+      "loss": 1.1691,
+      "step": 6459
+    },
+    {
+      "epoch": 2.405091272934579,
+      "grad_norm": 0.1617787927389145,
+      "learning_rate": 1.7878028688658044e-05,
+      "loss": 1.1715,
+      "step": 6460
+    },
+    {
+      "epoch": 2.405463578085188,
+      "grad_norm": 0.16519810259342194,
+      "learning_rate": 1.787728008379105e-05,
+      "loss": 1.189,
+      "step": 6461
+    },
+    {
+      "epoch": 2.405835883235797,
+      "grad_norm": 0.17165139317512512,
+      "learning_rate": 1.787653136257713e-05,
+      "loss": 1.1729,
+      "step": 6462
+    },
+    {
+      "epoch": 2.4062081883864064,
+      "grad_norm": 0.16290132701396942,
+      "learning_rate": 1.787578252502733e-05,
+      "loss": 1.1823,
+      "step": 6463
+    },
+    {
+      "epoch": 2.406580493537015,
+      "grad_norm": 0.16138114035129547,
+      "learning_rate": 1.7875033571152718e-05,
+      "loss": 1.1697,
+      "step": 6464
+    },
+    {
+      "epoch": 2.4069527986876245,
+      "grad_norm": 0.16568872332572937,
+      "learning_rate": 1.7874284500964357e-05,
+      "loss": 1.1734,
+      "step": 6465
+    },
+    {
+      "epoch": 2.4073251038382333,
+      "grad_norm": 0.1621939092874527,
+      "learning_rate": 1.7873535314473308e-05,
+      "loss": 1.1837,
+      "step": 6466
+    },
+    {
+      "epoch": 2.4076974089888425,
+      "grad_norm": 0.15826240181922913,
+      "learning_rate": 1.787278601169063e-05,
+      "loss": 1.1699,
+      "step": 6467
+    },
+    {
+      "epoch": 2.4080697141394514,
+      "grad_norm": 0.17099608480930328,
+      "learning_rate": 1.7872036592627404e-05,
+      "loss": 1.189,
+      "step": 6468
+    },
+    {
+      "epoch": 2.4084420192900606,
+      "grad_norm": 0.16510766744613647,
+      "learning_rate": 1.7871287057294688e-05,
+      "loss": 1.1759,
+      "step": 6469
+    },
+    {
+      "epoch": 2.4088143244406695,
+      "grad_norm": 0.1691225916147232,
+      "learning_rate": 1.7870537405703556e-05,
+      "loss": 1.1761,
+      "step": 6470
+    },
+    {
+      "epoch": 2.4091866295912787,
+      "grad_norm": 0.1651260256767273,
+      "learning_rate": 1.7869787637865084e-05,
+      "loss": 1.1737,
+      "step": 6471
+    },
+    {
+      "epoch": 2.409558934741888,
+      "grad_norm": 0.16575190424919128,
+      "learning_rate": 1.7869037753790343e-05,
+      "loss": 1.1768,
+      "step": 6472
+    },
+    {
+      "epoch": 2.409931239892497,
+      "grad_norm": 0.17237798869609833,
+      "learning_rate": 1.78682877534904e-05,
+      "loss": 1.1944,
+      "step": 6473
+    },
+    {
+      "epoch": 2.410303545043106,
+      "grad_norm": 0.15723305940628052,
+      "learning_rate": 1.7867537636976348e-05,
+      "loss": 1.1725,
+      "step": 6474
+    },
+    {
+      "epoch": 2.410675850193715,
+      "grad_norm": 0.16039879620075226,
+      "learning_rate": 1.7866787404259255e-05,
+      "loss": 1.1674,
+      "step": 6475
+    },
+    {
+      "epoch": 2.411048155344324,
+      "grad_norm": 0.1647602766752243,
+      "learning_rate": 1.786603705535021e-05,
+      "loss": 1.1616,
+      "step": 6476
+    },
+    {
+      "epoch": 2.411420460494933,
+      "grad_norm": 0.16092441976070404,
+      "learning_rate": 1.7865286590260288e-05,
+      "loss": 1.156,
+      "step": 6477
+    },
+    {
+      "epoch": 2.4117927656455422,
+      "grad_norm": 0.1571139693260193,
+      "learning_rate": 1.7864536009000575e-05,
+      "loss": 1.1713,
+      "step": 6478
+    },
+    {
+      "epoch": 2.412165070796151,
+      "grad_norm": 0.16552495956420898,
+      "learning_rate": 1.7863785311582163e-05,
+      "loss": 1.1648,
+      "step": 6479
+    },
+    {
+      "epoch": 2.4125373759467603,
+      "grad_norm": 0.16000515222549438,
+      "learning_rate": 1.7863034498016133e-05,
+      "loss": 1.1861,
+      "step": 6480
+    },
+    {
+      "epoch": 2.4129096810973696,
+      "grad_norm": 0.16232497990131378,
+      "learning_rate": 1.7862283568313578e-05,
+      "loss": 1.1737,
+      "step": 6481
+    },
+    {
+      "epoch": 2.4132819862479784,
+      "grad_norm": 0.1606409251689911,
+      "learning_rate": 1.7861532522485588e-05,
+      "loss": 1.1784,
+      "step": 6482
+    },
+    {
+      "epoch": 2.4136542913985877,
+      "grad_norm": 0.16612519323825836,
+      "learning_rate": 1.7860781360543255e-05,
+      "loss": 1.1825,
+      "step": 6483
+    },
+    {
+      "epoch": 2.4140265965491965,
+      "grad_norm": 0.16565221548080444,
+      "learning_rate": 1.7860030082497676e-05,
+      "loss": 1.1889,
+      "step": 6484
+    },
+    {
+      "epoch": 2.4143989016998058,
+      "grad_norm": 0.16412298381328583,
+      "learning_rate": 1.7859278688359946e-05,
+      "loss": 1.171,
+      "step": 6485
+    },
+    {
+      "epoch": 2.414771206850415,
+      "grad_norm": 0.1599871665239334,
+      "learning_rate": 1.7858527178141162e-05,
+      "loss": 1.1857,
+      "step": 6486
+    },
+    {
+      "epoch": 2.415143512001024,
+      "grad_norm": 0.16030682623386383,
+      "learning_rate": 1.7857775551852426e-05,
+      "loss": 1.1743,
+      "step": 6487
+    },
+    {
+      "epoch": 2.4155158171516327,
+      "grad_norm": 0.16047127544879913,
+      "learning_rate": 1.785702380950484e-05,
+      "loss": 1.1954,
+      "step": 6488
+    },
+    {
+      "epoch": 2.415888122302242,
+      "grad_norm": 0.16832852363586426,
+      "learning_rate": 1.78562719511095e-05,
+      "loss": 1.1862,
+      "step": 6489
+    },
+    {
+      "epoch": 2.416260427452851,
+      "grad_norm": 0.163087859749794,
+      "learning_rate": 1.785551997667752e-05,
+      "loss": 1.1849,
+      "step": 6490
+    },
+    {
+      "epoch": 2.41663273260346,
+      "grad_norm": 0.16009113192558289,
+      "learning_rate": 1.7854767886220002e-05,
+      "loss": 1.1717,
+      "step": 6491
+    },
+    {
+      "epoch": 2.4170050377540693,
+      "grad_norm": 0.1627979576587677,
+      "learning_rate": 1.7854015679748053e-05,
+      "loss": 1.1865,
+      "step": 6492
+    },
+    {
+      "epoch": 2.417377342904678,
+      "grad_norm": 0.16498367488384247,
+      "learning_rate": 1.785326335727279e-05,
+      "loss": 1.1613,
+      "step": 6493
+    },
+    {
+      "epoch": 2.4177496480552874,
+      "grad_norm": 0.16802792251110077,
+      "learning_rate": 1.7852510918805318e-05,
+      "loss": 1.1754,
+      "step": 6494
+    },
+    {
+      "epoch": 2.4181219532058966,
+      "grad_norm": 0.16168522834777832,
+      "learning_rate": 1.785175836435675e-05,
+      "loss": 1.1908,
+      "step": 6495
+    },
+    {
+      "epoch": 2.4184942583565054,
+      "grad_norm": 0.16734105348587036,
+      "learning_rate": 1.7851005693938206e-05,
+      "loss": 1.1671,
+      "step": 6496
+    },
+    {
+      "epoch": 2.4188665635071147,
+      "grad_norm": 0.1836169809103012,
+      "learning_rate": 1.78502529075608e-05,
+      "loss": 1.1681,
+      "step": 6497
+    },
+    {
+      "epoch": 2.4192388686577235,
+      "grad_norm": 0.20119720697402954,
+      "learning_rate": 1.7849500005235652e-05,
+      "loss": 1.1652,
+      "step": 6498
+    },
+    {
+      "epoch": 2.419611173808333,
+      "grad_norm": 0.19571252167224884,
+      "learning_rate": 1.7848746986973883e-05,
+      "loss": 1.163,
+      "step": 6499
+    },
+    {
+      "epoch": 2.4199834789589416,
+      "grad_norm": 0.19061268866062164,
+      "learning_rate": 1.7847993852786612e-05,
+      "loss": 1.1809,
+      "step": 6500
+    },
+    {
+      "epoch": 2.4199834789589416,
+      "eval_loss": 1.3013861179351807,
+      "eval_runtime": 17.1871,
+      "eval_samples_per_second": 100.889,
+      "eval_steps_per_second": 5.062,
+      "step": 6500
+    },
+    {
+      "epoch": 2.420355784109551,
+      "grad_norm": 0.17919202148914337,
+      "learning_rate": 1.7847240602684962e-05,
+      "loss": 1.1778,
+      "step": 6501
+    },
+    {
+      "epoch": 2.4207280892601597,
+      "grad_norm": 0.1730869710445404,
+      "learning_rate": 1.7846487236680064e-05,
+      "loss": 1.1851,
+      "step": 6502
+    },
+    {
+      "epoch": 2.421100394410769,
+      "grad_norm": 0.1636027991771698,
+      "learning_rate": 1.784573375478304e-05,
+      "loss": 1.1806,
+      "step": 6503
+    },
+    {
+      "epoch": 2.4214726995613782,
+      "grad_norm": 0.16488824784755707,
+      "learning_rate": 1.7844980157005022e-05,
+      "loss": 1.1554,
+      "step": 6504
+    },
+    {
+      "epoch": 2.421845004711987,
+      "grad_norm": 0.16549324989318848,
+      "learning_rate": 1.784422644335714e-05,
+      "loss": 1.1765,
+      "step": 6505
+    },
+    {
+      "epoch": 2.4222173098625963,
+      "grad_norm": 0.1688234508037567,
+      "learning_rate": 1.7843472613850523e-05,
+      "loss": 1.1575,
+      "step": 6506
+    },
+    {
+      "epoch": 2.422589615013205,
+      "grad_norm": 0.17091526091098785,
+      "learning_rate": 1.7842718668496312e-05,
+      "loss": 1.1692,
+      "step": 6507
+    },
+    {
+      "epoch": 2.4229619201638144,
+      "grad_norm": 0.1648736596107483,
+      "learning_rate": 1.7841964607305636e-05,
+      "loss": 1.1882,
+      "step": 6508
+    },
+    {
+      "epoch": 2.423334225314423,
+      "grad_norm": 0.18578572571277618,
+      "learning_rate": 1.7841210430289636e-05,
+      "loss": 1.1695,
+      "step": 6509
+    },
+    {
+      "epoch": 2.4237065304650325,
+      "grad_norm": 0.1957530379295349,
+      "learning_rate": 1.7840456137459447e-05,
+      "loss": 1.1688,
+      "step": 6510
+    },
+    {
+      "epoch": 2.4240788356156413,
+      "grad_norm": 0.16478918492794037,
+      "learning_rate": 1.7839701728826214e-05,
+      "loss": 1.1787,
+      "step": 6511
+    },
+    {
+      "epoch": 2.4244511407662506,
+      "grad_norm": 0.2944774329662323,
+      "learning_rate": 1.7838947204401083e-05,
+      "loss": 1.184,
+      "step": 6512
+    },
+    {
+      "epoch": 2.42482344591686,
+      "grad_norm": 0.18149353563785553,
+      "learning_rate": 1.7838192564195188e-05,
+      "loss": 1.1683,
+      "step": 6513
+    },
+    {
+      "epoch": 2.4251957510674687,
+      "grad_norm": 0.18618106842041016,
+      "learning_rate": 1.7837437808219683e-05,
+      "loss": 1.1592,
+      "step": 6514
+    },
+    {
+      "epoch": 2.425568056218078,
+      "grad_norm": 0.1710144430398941,
+      "learning_rate": 1.7836682936485714e-05,
+      "loss": 1.1993,
+      "step": 6515
+    },
+    {
+      "epoch": 2.4259403613686867,
+      "grad_norm": 0.17375528812408447,
+      "learning_rate": 1.783592794900443e-05,
+      "loss": 1.1793,
+      "step": 6516
+    },
+    {
+      "epoch": 2.426312666519296,
+      "grad_norm": 0.1787467747926712,
+      "learning_rate": 1.783517284578698e-05,
+      "loss": 1.1852,
+      "step": 6517
+    },
+    {
+      "epoch": 2.426684971669905,
+      "grad_norm": 0.17819075286388397,
+      "learning_rate": 1.783441762684452e-05,
+      "loss": 1.1715,
+      "step": 6518
+    },
+    {
+      "epoch": 2.427057276820514,
+      "grad_norm": 0.16626736521720886,
+      "learning_rate": 1.78336622921882e-05,
+      "loss": 1.1769,
+      "step": 6519
+    },
+    {
+      "epoch": 2.427429581971123,
+      "grad_norm": 0.1616727113723755,
+      "learning_rate": 1.7832906841829187e-05,
+      "loss": 1.1717,
+      "step": 6520
+    },
+    {
+      "epoch": 2.427801887121732,
+      "grad_norm": 0.1669609099626541,
+      "learning_rate": 1.7832151275778625e-05,
+      "loss": 1.1717,
+      "step": 6521
+    },
+    {
+      "epoch": 2.4281741922723414,
+      "grad_norm": 0.16761192679405212,
+      "learning_rate": 1.7831395594047682e-05,
+      "loss": 1.1763,
+      "step": 6522
+    },
+    {
+      "epoch": 2.4285464974229503,
+      "grad_norm": 0.1701916605234146,
+      "learning_rate": 1.783063979664752e-05,
+      "loss": 1.1905,
+      "step": 6523
+    },
+    {
+      "epoch": 2.4289188025735595,
+      "grad_norm": 0.16727663576602936,
+      "learning_rate": 1.7829883883589297e-05,
+      "loss": 1.1913,
+      "step": 6524
+    },
+    {
+      "epoch": 2.4292911077241683,
+      "grad_norm": 0.16369487345218658,
+      "learning_rate": 1.782912785488418e-05,
+      "loss": 1.1616,
+      "step": 6525
+    },
+    {
+      "epoch": 2.4296634128747776,
+      "grad_norm": 0.1752278357744217,
+      "learning_rate": 1.7828371710543336e-05,
+      "loss": 1.1921,
+      "step": 6526
+    },
+    {
+      "epoch": 2.4300357180253864,
+      "grad_norm": 0.16097818315029144,
+      "learning_rate": 1.7827615450577936e-05,
+      "loss": 1.1998,
+      "step": 6527
+    },
+    {
+      "epoch": 2.4304080231759957,
+      "grad_norm": 0.16219815611839294,
+      "learning_rate": 1.7826859074999145e-05,
+      "loss": 1.1806,
+      "step": 6528
+    },
+    {
+      "epoch": 2.4307803283266045,
+      "grad_norm": 0.16351665556430817,
+      "learning_rate": 1.7826102583818134e-05,
+      "loss": 1.1762,
+      "step": 6529
+    },
+    {
+      "epoch": 2.4311526334772138,
+      "grad_norm": 0.16075001657009125,
+      "learning_rate": 1.782534597704608e-05,
+      "loss": 1.1809,
+      "step": 6530
+    },
+    {
+      "epoch": 2.431524938627823,
+      "grad_norm": 0.19722290337085724,
+      "learning_rate": 1.7824589254694163e-05,
+      "loss": 1.1829,
+      "step": 6531
+    },
+    {
+      "epoch": 2.431897243778432,
+      "grad_norm": 0.1670161485671997,
+      "learning_rate": 1.782383241677355e-05,
+      "loss": 1.1921,
+      "step": 6532
+    },
+    {
+      "epoch": 2.432269548929041,
+      "grad_norm": 0.1666540503501892,
+      "learning_rate": 1.782307546329542e-05,
+      "loss": 1.1845,
+      "step": 6533
+    },
+    {
+      "epoch": 2.43264185407965,
+      "grad_norm": 0.16255459189414978,
+      "learning_rate": 1.782231839427096e-05,
+      "loss": 1.1682,
+      "step": 6534
+    },
+    {
+      "epoch": 2.433014159230259,
+      "grad_norm": 0.16188620030879974,
+      "learning_rate": 1.782156120971135e-05,
+      "loss": 1.1696,
+      "step": 6535
+    },
+    {
+      "epoch": 2.433386464380868,
+      "grad_norm": 0.17455708980560303,
+      "learning_rate": 1.7820803909627766e-05,
+      "loss": 1.1804,
+      "step": 6536
+    },
+    {
+      "epoch": 2.4337587695314773,
+      "grad_norm": 0.16326938569545746,
+      "learning_rate": 1.7820046494031405e-05,
+      "loss": 1.1794,
+      "step": 6537
+    },
+    {
+      "epoch": 2.434131074682086,
+      "grad_norm": 0.15706990659236908,
+      "learning_rate": 1.7819288962933442e-05,
+      "loss": 1.1619,
+      "step": 6538
+    },
+    {
+      "epoch": 2.4345033798326954,
+      "grad_norm": 0.16634133458137512,
+      "learning_rate": 1.7818531316345078e-05,
+      "loss": 1.1824,
+      "step": 6539
+    },
+    {
+      "epoch": 2.4348756849833046,
+      "grad_norm": 0.16253423690795898,
+      "learning_rate": 1.7817773554277493e-05,
+      "loss": 1.1833,
+      "step": 6540
+    },
+    {
+      "epoch": 2.4352479901339135,
+      "grad_norm": 0.16380086541175842,
+      "learning_rate": 1.7817015676741883e-05,
+      "loss": 1.1861,
+      "step": 6541
+    },
+    {
+      "epoch": 2.4356202952845227,
+      "grad_norm": 0.16374869644641876,
+      "learning_rate": 1.7816257683749444e-05,
+      "loss": 1.1758,
+      "step": 6542
+    },
+    {
+      "epoch": 2.4359926004351315,
+      "grad_norm": 0.1633332371711731,
+      "learning_rate": 1.7815499575311367e-05,
+      "loss": 1.1984,
+      "step": 6543
+    },
+    {
+      "epoch": 2.436364905585741,
+      "grad_norm": 0.16350342333316803,
+      "learning_rate": 1.7814741351438855e-05,
+      "loss": 1.1672,
+      "step": 6544
+    },
+    {
+      "epoch": 2.4367372107363496,
+      "grad_norm": 0.16379904747009277,
+      "learning_rate": 1.7813983012143104e-05,
+      "loss": 1.1895,
+      "step": 6545
+    },
+    {
+      "epoch": 2.437109515886959,
+      "grad_norm": 0.16727176308631897,
+      "learning_rate": 1.7813224557435313e-05,
+      "loss": 1.1741,
+      "step": 6546
+    },
+    {
+      "epoch": 2.4374818210375677,
+      "grad_norm": 0.1602688729763031,
+      "learning_rate": 1.7812465987326682e-05,
+      "loss": 1.1816,
+      "step": 6547
+    },
+    {
+      "epoch": 2.437854126188177,
+      "grad_norm": 0.16404597461223602,
+      "learning_rate": 1.7811707301828424e-05,
+      "loss": 1.1875,
+      "step": 6548
+    },
+    {
+      "epoch": 2.4382264313387862,
+      "grad_norm": 0.16207285225391388,
+      "learning_rate": 1.7810948500951738e-05,
+      "loss": 1.1627,
+      "step": 6549
+    },
+    {
+      "epoch": 2.438598736489395,
+      "grad_norm": 0.168754443526268,
+      "learning_rate": 1.7810189584707834e-05,
+      "loss": 1.1778,
+      "step": 6550
+    },
+    {
+      "epoch": 2.4389710416400043,
+      "grad_norm": 0.1652376502752304,
+      "learning_rate": 1.7809430553107915e-05,
+      "loss": 1.1854,
+      "step": 6551
+    },
+    {
+      "epoch": 2.439343346790613,
+      "grad_norm": 0.16541428864002228,
+      "learning_rate": 1.78086714061632e-05,
+      "loss": 1.1917,
+      "step": 6552
+    },
+    {
+      "epoch": 2.4397156519412224,
+      "grad_norm": 0.1626761257648468,
+      "learning_rate": 1.78079121438849e-05,
+      "loss": 1.1666,
+      "step": 6553
+    },
+    {
+      "epoch": 2.4400879570918312,
+      "grad_norm": 0.164178267121315,
+      "learning_rate": 1.7807152766284222e-05,
+      "loss": 1.1834,
+      "step": 6554
+    },
+    {
+      "epoch": 2.4404602622424405,
+      "grad_norm": 0.15936389565467834,
+      "learning_rate": 1.7806393273372396e-05,
+      "loss": 1.1668,
+      "step": 6555
+    },
+    {
+      "epoch": 2.4408325673930493,
+      "grad_norm": 0.16379861533641815,
+      "learning_rate": 1.7805633665160623e-05,
+      "loss": 1.1796,
+      "step": 6556
+    },
+    {
+      "epoch": 2.4412048725436586,
+      "grad_norm": 0.165420264005661,
+      "learning_rate": 1.7804873941660137e-05,
+      "loss": 1.1771,
+      "step": 6557
+    },
+    {
+      "epoch": 2.441577177694268,
+      "grad_norm": 0.16824427247047424,
+      "learning_rate": 1.780411410288215e-05,
+      "loss": 1.1844,
+      "step": 6558
+    },
+    {
+      "epoch": 2.4419494828448767,
+      "grad_norm": 0.16532348096370697,
+      "learning_rate": 1.780335414883789e-05,
+      "loss": 1.1887,
+      "step": 6559
+    },
+    {
+      "epoch": 2.442321787995486,
+      "grad_norm": 0.16297706961631775,
+      "learning_rate": 1.7802594079538574e-05,
+      "loss": 1.1702,
+      "step": 6560
+    },
+    {
+      "epoch": 2.4426940931460948,
+      "grad_norm": 0.16439802944660187,
+      "learning_rate": 1.7801833894995436e-05,
+      "loss": 1.1802,
+      "step": 6561
+    },
+    {
+      "epoch": 2.443066398296704,
+      "grad_norm": 0.16698960959911346,
+      "learning_rate": 1.7801073595219702e-05,
+      "loss": 1.183,
+      "step": 6562
+    },
+    {
+      "epoch": 2.443438703447313,
+      "grad_norm": 0.17252756655216217,
+      "learning_rate": 1.78003131802226e-05,
+      "loss": 1.187,
+      "step": 6563
+    },
+    {
+      "epoch": 2.443811008597922,
+      "grad_norm": 0.16879017651081085,
+      "learning_rate": 1.779955265001536e-05,
+      "loss": 1.1708,
+      "step": 6564
+    },
+    {
+      "epoch": 2.444183313748531,
+      "grad_norm": 0.16200561821460724,
+      "learning_rate": 1.779879200460922e-05,
+      "loss": 1.173,
+      "step": 6565
+    },
+    {
+      "epoch": 2.44455561889914,
+      "grad_norm": 0.16437043249607086,
+      "learning_rate": 1.7798031244015406e-05,
+      "loss": 1.1693,
+      "step": 6566
+    },
+    {
+      "epoch": 2.4449279240497495,
+      "grad_norm": 0.17090344429016113,
+      "learning_rate": 1.7797270368245166e-05,
+      "loss": 1.1667,
+      "step": 6567
+    },
+    {
+      "epoch": 2.4453002292003583,
+      "grad_norm": 0.1742093712091446,
+      "learning_rate": 1.7796509377309728e-05,
+      "loss": 1.1932,
+      "step": 6568
+    },
+    {
+      "epoch": 2.4456725343509675,
+      "grad_norm": 0.16742129623889923,
+      "learning_rate": 1.7795748271220337e-05,
+      "loss": 1.1628,
+      "step": 6569
+    },
+    {
+      "epoch": 2.4460448395015764,
+      "grad_norm": 0.16507847607135773,
+      "learning_rate": 1.779498704998823e-05,
+      "loss": 1.1942,
+      "step": 6570
+    },
+    {
+      "epoch": 2.4464171446521856,
+      "grad_norm": 0.16421516239643097,
+      "learning_rate": 1.7794225713624663e-05,
+      "loss": 1.1746,
+      "step": 6571
+    },
+    {
+      "epoch": 2.4467894498027944,
+      "grad_norm": 0.17189864814281464,
+      "learning_rate": 1.7793464262140864e-05,
+      "loss": 1.1843,
+      "step": 6572
+    },
+    {
+      "epoch": 2.4471617549534037,
+      "grad_norm": 0.16595551371574402,
+      "learning_rate": 1.7792702695548086e-05,
+      "loss": 1.187,
+      "step": 6573
+    },
+    {
+      "epoch": 2.4475340601040125,
+      "grad_norm": 0.16488155722618103,
+      "learning_rate": 1.779194101385758e-05,
+      "loss": 1.1726,
+      "step": 6574
+    },
+    {
+      "epoch": 2.447906365254622,
+      "grad_norm": 0.164886474609375,
+      "learning_rate": 1.7791179217080598e-05,
+      "loss": 1.1745,
+      "step": 6575
+    },
+    {
+      "epoch": 2.448278670405231,
+      "grad_norm": 0.17774739861488342,
+      "learning_rate": 1.779041730522838e-05,
+      "loss": 1.1738,
+      "step": 6576
+    },
+    {
+      "epoch": 2.44865097555584,
+      "grad_norm": 0.1659976691007614,
+      "learning_rate": 1.7789655278312198e-05,
+      "loss": 1.1803,
+      "step": 6577
+    },
+    {
+      "epoch": 2.449023280706449,
+      "grad_norm": 0.16639791429042816,
+      "learning_rate": 1.7788893136343288e-05,
+      "loss": 1.1628,
+      "step": 6578
+    },
+    {
+      "epoch": 2.449395585857058,
+      "grad_norm": 0.16745202243328094,
+      "learning_rate": 1.7788130879332918e-05,
+      "loss": 1.1697,
+      "step": 6579
+    },
+    {
+      "epoch": 2.4497678910076672,
+      "grad_norm": 0.16819866001605988,
+      "learning_rate": 1.7787368507292343e-05,
+      "loss": 1.1783,
+      "step": 6580
+    },
+    {
+      "epoch": 2.450140196158276,
+      "grad_norm": 0.16538161039352417,
+      "learning_rate": 1.7786606020232825e-05,
+      "loss": 1.1736,
+      "step": 6581
+    },
+    {
+      "epoch": 2.4505125013088853,
+      "grad_norm": 0.1632433533668518,
+      "learning_rate": 1.7785843418165624e-05,
+      "loss": 1.1779,
+      "step": 6582
+    },
+    {
+      "epoch": 2.450884806459494,
+      "grad_norm": 0.1659589260816574,
+      "learning_rate": 1.7785080701102003e-05,
+      "loss": 1.195,
+      "step": 6583
+    },
+    {
+      "epoch": 2.4512571116101034,
+      "grad_norm": 0.16399775445461273,
+      "learning_rate": 1.778431786905323e-05,
+      "loss": 1.1767,
+      "step": 6584
+    },
+    {
+      "epoch": 2.4516294167607127,
+      "grad_norm": 0.1630357950925827,
+      "learning_rate": 1.778355492203057e-05,
+      "loss": 1.1817,
+      "step": 6585
+    },
+    {
+      "epoch": 2.4520017219113215,
+      "grad_norm": 0.15688146650791168,
+      "learning_rate": 1.778279186004529e-05,
+      "loss": 1.1784,
+      "step": 6586
+    },
+    {
+      "epoch": 2.4523740270619308,
+      "grad_norm": 0.15879155695438385,
+      "learning_rate": 1.7782028683108667e-05,
+      "loss": 1.1713,
+      "step": 6587
+    },
+    {
+      "epoch": 2.4527463322125396,
+      "grad_norm": 0.16145865619182587,
+      "learning_rate": 1.7781265391231968e-05,
+      "loss": 1.186,
+      "step": 6588
+    },
+    {
+      "epoch": 2.453118637363149,
+      "grad_norm": 0.1632131040096283,
+      "learning_rate": 1.7780501984426465e-05,
+      "loss": 1.1699,
+      "step": 6589
+    },
+    {
+      "epoch": 2.4534909425137577,
+      "grad_norm": 0.15937484800815582,
+      "learning_rate": 1.777973846270344e-05,
+      "loss": 1.1633,
+      "step": 6590
+    },
+    {
+      "epoch": 2.453863247664367,
+      "grad_norm": 0.16633634269237518,
+      "learning_rate": 1.7778974826074163e-05,
+      "loss": 1.1933,
+      "step": 6591
+    },
+    {
+      "epoch": 2.4542355528149757,
+      "grad_norm": 0.1646929830312729,
+      "learning_rate": 1.7778211074549916e-05,
+      "loss": 1.1823,
+      "step": 6592
+    },
+    {
+      "epoch": 2.454607857965585,
+      "grad_norm": 0.1606590747833252,
+      "learning_rate": 1.777744720814198e-05,
+      "loss": 1.1797,
+      "step": 6593
+    },
+    {
+      "epoch": 2.4549801631161943,
+      "grad_norm": 0.16645345091819763,
+      "learning_rate": 1.7776683226861636e-05,
+      "loss": 1.1663,
+      "step": 6594
+    },
+    {
+      "epoch": 2.455352468266803,
+      "grad_norm": 0.16911546885967255,
+      "learning_rate": 1.777591913072017e-05,
+      "loss": 1.1717,
+      "step": 6595
+    },
+    {
+      "epoch": 2.4557247734174124,
+      "grad_norm": 0.16527560353279114,
+      "learning_rate": 1.777515491972887e-05,
+      "loss": 1.1854,
+      "step": 6596
+    },
+    {
+      "epoch": 2.456097078568021,
+      "grad_norm": 0.16510476171970367,
+      "learning_rate": 1.7774390593899014e-05,
+      "loss": 1.1735,
+      "step": 6597
+    },
+    {
+      "epoch": 2.4564693837186304,
+      "grad_norm": 0.16441304981708527,
+      "learning_rate": 1.7773626153241897e-05,
+      "loss": 1.1721,
+      "step": 6598
+    },
+    {
+      "epoch": 2.4568416888692393,
+      "grad_norm": 0.163100928068161,
+      "learning_rate": 1.7772861597768814e-05,
+      "loss": 1.1774,
+      "step": 6599
+    },
+    {
+      "epoch": 2.4572139940198485,
+      "grad_norm": 0.16607718169689178,
+      "learning_rate": 1.777209692749105e-05,
+      "loss": 1.171,
+      "step": 6600
+    },
+    {
+      "epoch": 2.4575862991704573,
+      "grad_norm": 0.1616518199443817,
+      "learning_rate": 1.77713321424199e-05,
+      "loss": 1.1722,
+      "step": 6601
+    },
+    {
+      "epoch": 2.4579586043210666,
+      "grad_norm": 0.1628957986831665,
+      "learning_rate": 1.7770567242566667e-05,
+      "loss": 1.1708,
+      "step": 6602
+    },
+    {
+      "epoch": 2.458330909471676,
+      "grad_norm": 0.1651378571987152,
+      "learning_rate": 1.776980222794264e-05,
+      "loss": 1.1775,
+      "step": 6603
+    },
+    {
+      "epoch": 2.4587032146222847,
+      "grad_norm": 0.16311919689178467,
+      "learning_rate": 1.7769037098559124e-05,
+      "loss": 1.1746,
+      "step": 6604
+    },
+    {
+      "epoch": 2.459075519772894,
+      "grad_norm": 0.16620442271232605,
+      "learning_rate": 1.7768271854427417e-05,
+      "loss": 1.177,
+      "step": 6605
+    },
+    {
+      "epoch": 2.459447824923503,
+      "grad_norm": 0.16138948500156403,
+      "learning_rate": 1.776750649555882e-05,
+      "loss": 1.1594,
+      "step": 6606
+    },
+    {
+      "epoch": 2.459820130074112,
+      "grad_norm": 0.16413603723049164,
+      "learning_rate": 1.776674102196464e-05,
+      "loss": 1.1901,
+      "step": 6607
+    },
+    {
+      "epoch": 2.4601924352247213,
+      "grad_norm": 0.1599273979663849,
+      "learning_rate": 1.7765975433656187e-05,
+      "loss": 1.1678,
+      "step": 6608
+    },
+    {
+      "epoch": 2.46056474037533,
+      "grad_norm": 0.15869589149951935,
+      "learning_rate": 1.776520973064476e-05,
+      "loss": 1.1647,
+      "step": 6609
+    },
+    {
+      "epoch": 2.460937045525939,
+      "grad_norm": 0.1629563570022583,
+      "learning_rate": 1.7764443912941675e-05,
+      "loss": 1.1711,
+      "step": 6610
+    },
+    {
+      "epoch": 2.461309350676548,
+      "grad_norm": 0.1599353700876236,
+      "learning_rate": 1.776367798055824e-05,
+      "loss": 1.1772,
+      "step": 6611
+    },
+    {
+      "epoch": 2.4616816558271575,
+      "grad_norm": 0.16407759487628937,
+      "learning_rate": 1.7762911933505767e-05,
+      "loss": 1.1826,
+      "step": 6612
+    },
+    {
+      "epoch": 2.4620539609777663,
+      "grad_norm": 0.1610613763332367,
+      "learning_rate": 1.776214577179557e-05,
+      "loss": 1.1704,
+      "step": 6613
+    },
+    {
+      "epoch": 2.4624262661283756,
+      "grad_norm": 0.16347752511501312,
+      "learning_rate": 1.776137949543897e-05,
+      "loss": 1.1712,
+      "step": 6614
+    },
+    {
+      "epoch": 2.4627985712789844,
+      "grad_norm": 0.15655462443828583,
+      "learning_rate": 1.7760613104447283e-05,
+      "loss": 1.1515,
+      "step": 6615
+    },
+    {
+      "epoch": 2.4631708764295936,
+      "grad_norm": 0.16429446637630463,
+      "learning_rate": 1.7759846598831827e-05,
+      "loss": 1.1793,
+      "step": 6616
+    },
+    {
+      "epoch": 2.463543181580203,
+      "grad_norm": 0.16493436694145203,
+      "learning_rate": 1.775907997860392e-05,
+      "loss": 1.1681,
+      "step": 6617
+    },
+    {
+      "epoch": 2.4639154867308117,
+      "grad_norm": 0.1644970327615738,
+      "learning_rate": 1.775831324377489e-05,
+      "loss": 1.1819,
+      "step": 6618
+    },
+    {
+      "epoch": 2.464287791881421,
+      "grad_norm": 0.16684550046920776,
+      "learning_rate": 1.7757546394356063e-05,
+      "loss": 1.1864,
+      "step": 6619
+    },
+    {
+      "epoch": 2.46466009703203,
+      "grad_norm": 0.16069693863391876,
+      "learning_rate": 1.775677943035876e-05,
+      "loss": 1.1721,
+      "step": 6620
+    },
+    {
+      "epoch": 2.465032402182639,
+      "grad_norm": 0.1637074053287506,
+      "learning_rate": 1.7756012351794315e-05,
+      "loss": 1.1697,
+      "step": 6621
+    },
+    {
+      "epoch": 2.465404707333248,
+      "grad_norm": 0.1612214297056198,
+      "learning_rate": 1.7755245158674054e-05,
+      "loss": 1.1658,
+      "step": 6622
+    },
+    {
+      "epoch": 2.465777012483857,
+      "grad_norm": 0.16873739659786224,
+      "learning_rate": 1.7754477851009307e-05,
+      "loss": 1.1943,
+      "step": 6623
+    },
+    {
+      "epoch": 2.466149317634466,
+      "grad_norm": 0.16475237905979156,
+      "learning_rate": 1.775371042881141e-05,
+      "loss": 1.1857,
+      "step": 6624
+    },
+    {
+      "epoch": 2.4665216227850753,
+      "grad_norm": 0.16475136578083038,
+      "learning_rate": 1.7752942892091694e-05,
+      "loss": 1.1782,
+      "step": 6625
+    },
+    {
+      "epoch": 2.4668939279356845,
+      "grad_norm": 0.16575996577739716,
+      "learning_rate": 1.7752175240861497e-05,
+      "loss": 1.1794,
+      "step": 6626
+    },
+    {
+      "epoch": 2.4672662330862933,
+      "grad_norm": 0.16450487077236176,
+      "learning_rate": 1.7751407475132164e-05,
+      "loss": 1.1768,
+      "step": 6627
+    },
+    {
+      "epoch": 2.4676385382369026,
+      "grad_norm": 0.1675347089767456,
+      "learning_rate": 1.7750639594915026e-05,
+      "loss": 1.1753,
+      "step": 6628
+    },
+    {
+      "epoch": 2.4680108433875114,
+      "grad_norm": 0.16457484662532806,
+      "learning_rate": 1.7749871600221426e-05,
+      "loss": 1.1841,
+      "step": 6629
+    },
+    {
+      "epoch": 2.4683831485381207,
+      "grad_norm": 0.17486611008644104,
+      "learning_rate": 1.774910349106271e-05,
+      "loss": 1.1963,
+      "step": 6630
+    },
+    {
+      "epoch": 2.4687554536887295,
+      "grad_norm": 0.1780180037021637,
+      "learning_rate": 1.774833526745022e-05,
+      "loss": 1.1998,
+      "step": 6631
+    },
+    {
+      "epoch": 2.4691277588393388,
+      "grad_norm": 0.168752983212471,
+      "learning_rate": 1.7747566929395307e-05,
+      "loss": 1.1728,
+      "step": 6632
+    },
+    {
+      "epoch": 2.4695000639899476,
+      "grad_norm": 0.1614418923854828,
+      "learning_rate": 1.7746798476909316e-05,
+      "loss": 1.1632,
+      "step": 6633
+    },
+    {
+      "epoch": 2.469872369140557,
+      "grad_norm": 0.16157226264476776,
+      "learning_rate": 1.7746029910003598e-05,
+      "loss": 1.1502,
+      "step": 6634
+    },
+    {
+      "epoch": 2.470244674291166,
+      "grad_norm": 0.16318723559379578,
+      "learning_rate": 1.7745261228689505e-05,
+      "loss": 1.1645,
+      "step": 6635
+    },
+    {
+      "epoch": 2.470616979441775,
+      "grad_norm": 0.17089669406414032,
+      "learning_rate": 1.7744492432978385e-05,
+      "loss": 1.1809,
+      "step": 6636
+    },
+    {
+      "epoch": 2.470989284592384,
+      "grad_norm": 0.1579371988773346,
+      "learning_rate": 1.7743723522881604e-05,
+      "loss": 1.1742,
+      "step": 6637
+    },
+    {
+      "epoch": 2.471361589742993,
+      "grad_norm": 0.16182205080986023,
+      "learning_rate": 1.774295449841051e-05,
+      "loss": 1.1868,
+      "step": 6638
+    },
+    {
+      "epoch": 2.4717338948936023,
+      "grad_norm": 0.16928228735923767,
+      "learning_rate": 1.7742185359576464e-05,
+      "loss": 1.1915,
+      "step": 6639
+    },
+    {
+      "epoch": 2.472106200044211,
+      "grad_norm": 0.15761636197566986,
+      "learning_rate": 1.7741416106390828e-05,
+      "loss": 1.1569,
+      "step": 6640
+    },
+    {
+      "epoch": 2.4724785051948204,
+      "grad_norm": 0.1611185520887375,
+      "learning_rate": 1.7740646738864956e-05,
+      "loss": 1.1806,
+      "step": 6641
+    },
+    {
+      "epoch": 2.472850810345429,
+      "grad_norm": 0.16303184628486633,
+      "learning_rate": 1.7739877257010226e-05,
+      "loss": 1.1706,
+      "step": 6642
+    },
+    {
+      "epoch": 2.4732231154960385,
+      "grad_norm": 0.1614006906747818,
+      "learning_rate": 1.7739107660837985e-05,
+      "loss": 1.1721,
+      "step": 6643
+    },
+    {
+      "epoch": 2.4735954206466477,
+      "grad_norm": 0.1621858775615692,
+      "learning_rate": 1.7738337950359617e-05,
+      "loss": 1.1672,
+      "step": 6644
+    },
+    {
+      "epoch": 2.4739677257972565,
+      "grad_norm": 0.15944872796535492,
+      "learning_rate": 1.7737568125586482e-05,
+      "loss": 1.1675,
+      "step": 6645
+    },
+    {
+      "epoch": 2.474340030947866,
+      "grad_norm": 0.16335934400558472,
+      "learning_rate": 1.7736798186529947e-05,
+      "loss": 1.1747,
+      "step": 6646
+    },
+    {
+      "epoch": 2.4747123360984746,
+      "grad_norm": 0.16249805688858032,
+      "learning_rate": 1.7736028133201394e-05,
+      "loss": 1.1767,
+      "step": 6647
+    },
+    {
+      "epoch": 2.475084641249084,
+      "grad_norm": 0.16741207242012024,
+      "learning_rate": 1.7735257965612188e-05,
+      "loss": 1.1809,
+      "step": 6648
+    },
+    {
+      "epoch": 2.4754569463996927,
+      "grad_norm": 0.167801633477211,
+      "learning_rate": 1.773448768377371e-05,
+      "loss": 1.174,
+      "step": 6649
+    },
+    {
+      "epoch": 2.475829251550302,
+      "grad_norm": 0.1614260971546173,
+      "learning_rate": 1.7733717287697328e-05,
+      "loss": 1.1734,
+      "step": 6650
+    },
+    {
+      "epoch": 2.476201556700911,
+      "grad_norm": 0.15728265047073364,
+      "learning_rate": 1.7732946777394432e-05,
+      "loss": 1.172,
+      "step": 6651
+    },
+    {
+      "epoch": 2.47657386185152,
+      "grad_norm": 0.16399987041950226,
+      "learning_rate": 1.77321761528764e-05,
+      "loss": 1.1902,
+      "step": 6652
+    },
+    {
+      "epoch": 2.4769461670021293,
+      "grad_norm": 0.1572953164577484,
+      "learning_rate": 1.7731405414154606e-05,
+      "loss": 1.1712,
+      "step": 6653
+    },
+    {
+      "epoch": 2.477318472152738,
+      "grad_norm": 0.16343964636325836,
+      "learning_rate": 1.7730634561240442e-05,
+      "loss": 1.178,
+      "step": 6654
+    },
+    {
+      "epoch": 2.4776907773033474,
+      "grad_norm": 0.1638653576374054,
+      "learning_rate": 1.7729863594145287e-05,
+      "loss": 1.1719,
+      "step": 6655
+    },
+    {
+      "epoch": 2.4780630824539562,
+      "grad_norm": 0.16249097883701324,
+      "learning_rate": 1.7729092512880534e-05,
+      "loss": 1.1715,
+      "step": 6656
+    },
+    {
+      "epoch": 2.4784353876045655,
+      "grad_norm": 0.15715673565864563,
+      "learning_rate": 1.7728321317457573e-05,
+      "loss": 1.1589,
+      "step": 6657
+    },
+    {
+      "epoch": 2.4788076927551743,
+      "grad_norm": 0.16444164514541626,
+      "learning_rate": 1.7727550007887787e-05,
+      "loss": 1.1729,
+      "step": 6658
+    },
+    {
+      "epoch": 2.4791799979057836,
+      "grad_norm": 0.16417501866817474,
+      "learning_rate": 1.7726778584182575e-05,
+      "loss": 1.1709,
+      "step": 6659
+    },
+    {
+      "epoch": 2.4795523030563924,
+      "grad_norm": 0.16987749934196472,
+      "learning_rate": 1.7726007046353328e-05,
+      "loss": 1.1784,
+      "step": 6660
+    },
+    {
+      "epoch": 2.4799246082070017,
+      "grad_norm": 0.16455167531967163,
+      "learning_rate": 1.772523539441144e-05,
+      "loss": 1.1882,
+      "step": 6661
+    },
+    {
+      "epoch": 2.480296913357611,
+      "grad_norm": 0.16781625151634216,
+      "learning_rate": 1.772446362836831e-05,
+      "loss": 1.1803,
+      "step": 6662
+    },
+    {
+      "epoch": 2.4806692185082198,
+      "grad_norm": 0.1608152836561203,
+      "learning_rate": 1.772369174823534e-05,
+      "loss": 1.1655,
+      "step": 6663
+    },
+    {
+      "epoch": 2.481041523658829,
+      "grad_norm": 0.16263465583324432,
+      "learning_rate": 1.7722919754023923e-05,
+      "loss": 1.1561,
+      "step": 6664
+    },
+    {
+      "epoch": 2.481413828809438,
+      "grad_norm": 0.16114389896392822,
+      "learning_rate": 1.772214764574547e-05,
+      "loss": 1.1724,
+      "step": 6665
+    },
+    {
+      "epoch": 2.481786133960047,
+      "grad_norm": 0.1621330976486206,
+      "learning_rate": 1.7721375423411378e-05,
+      "loss": 1.1724,
+      "step": 6666
+    },
+    {
+      "epoch": 2.482158439110656,
+      "grad_norm": 0.16913004219532013,
+      "learning_rate": 1.7720603087033058e-05,
+      "loss": 1.1843,
+      "step": 6667
+    },
+    {
+      "epoch": 2.482530744261265,
+      "grad_norm": 0.16739852726459503,
+      "learning_rate": 1.7719830636621914e-05,
+      "loss": 1.1717,
+      "step": 6668
+    },
+    {
+      "epoch": 2.482903049411874,
+      "grad_norm": 0.16293086111545563,
+      "learning_rate": 1.7719058072189355e-05,
+      "loss": 1.1743,
+      "step": 6669
+    },
+    {
+      "epoch": 2.4832753545624833,
+      "grad_norm": 0.16651244461536407,
+      "learning_rate": 1.771828539374679e-05,
+      "loss": 1.1892,
+      "step": 6670
+    },
+    {
+      "epoch": 2.4836476597130925,
+      "grad_norm": 0.16542814671993256,
+      "learning_rate": 1.771751260130564e-05,
+      "loss": 1.1692,
+      "step": 6671
+    },
+    {
+      "epoch": 2.4840199648637014,
+      "grad_norm": 0.16294194757938385,
+      "learning_rate": 1.771673969487731e-05,
+      "loss": 1.1784,
+      "step": 6672
+    },
+    {
+      "epoch": 2.4843922700143106,
+      "grad_norm": 0.16167296469211578,
+      "learning_rate": 1.771596667447322e-05,
+      "loss": 1.1642,
+      "step": 6673
+    },
+    {
+      "epoch": 2.4847645751649194,
+      "grad_norm": 0.1647670865058899,
+      "learning_rate": 1.771519354010479e-05,
+      "loss": 1.1843,
+      "step": 6674
+    },
+    {
+      "epoch": 2.4851368803155287,
+      "grad_norm": 0.16380488872528076,
+      "learning_rate": 1.771442029178343e-05,
+      "loss": 1.1939,
+      "step": 6675
+    },
+    {
+      "epoch": 2.4855091854661375,
+      "grad_norm": 0.16478745639324188,
+      "learning_rate": 1.7713646929520568e-05,
+      "loss": 1.174,
+      "step": 6676
+    },
+    {
+      "epoch": 2.485881490616747,
+      "grad_norm": 0.16080977022647858,
+      "learning_rate": 1.7712873453327626e-05,
+      "loss": 1.1729,
+      "step": 6677
+    },
+    {
+      "epoch": 2.4862537957673556,
+      "grad_norm": 0.16270096600055695,
+      "learning_rate": 1.7712099863216027e-05,
+      "loss": 1.1848,
+      "step": 6678
+    },
+    {
+      "epoch": 2.486626100917965,
+      "grad_norm": 0.16529610753059387,
+      "learning_rate": 1.7711326159197195e-05,
+      "loss": 1.1513,
+      "step": 6679
+    },
+    {
+      "epoch": 2.486998406068574,
+      "grad_norm": 0.16876338422298431,
+      "learning_rate": 1.771055234128256e-05,
+      "loss": 1.1805,
+      "step": 6680
+    },
+    {
+      "epoch": 2.487370711219183,
+      "grad_norm": 0.18239647150039673,
+      "learning_rate": 1.7709778409483554e-05,
+      "loss": 1.1735,
+      "step": 6681
+    },
+    {
+      "epoch": 2.4877430163697922,
+      "grad_norm": 0.16950790584087372,
+      "learning_rate": 1.7709004363811598e-05,
+      "loss": 1.1815,
+      "step": 6682
+    },
+    {
+      "epoch": 2.488115321520401,
+      "grad_norm": 0.16490155458450317,
+      "learning_rate": 1.770823020427814e-05,
+      "loss": 1.1648,
+      "step": 6683
+    },
+    {
+      "epoch": 2.4884876266710103,
+      "grad_norm": 0.17165668308734894,
+      "learning_rate": 1.7707455930894603e-05,
+      "loss": 1.178,
+      "step": 6684
+    },
+    {
+      "epoch": 2.488859931821619,
+      "grad_norm": 0.16048189997673035,
+      "learning_rate": 1.770668154367242e-05,
+      "loss": 1.1776,
+      "step": 6685
+    },
+    {
+      "epoch": 2.4892322369722284,
+      "grad_norm": 0.16602389514446259,
+      "learning_rate": 1.770590704262304e-05,
+      "loss": 1.1667,
+      "step": 6686
+    },
+    {
+      "epoch": 2.489604542122837,
+      "grad_norm": 0.1645582914352417,
+      "learning_rate": 1.7705132427757895e-05,
+      "loss": 1.1881,
+      "step": 6687
+    },
+    {
+      "epoch": 2.4899768472734465,
+      "grad_norm": 0.16345307230949402,
+      "learning_rate": 1.7704357699088426e-05,
+      "loss": 1.1757,
+      "step": 6688
+    },
+    {
+      "epoch": 2.4903491524240557,
+      "grad_norm": 0.1574520319700241,
+      "learning_rate": 1.770358285662608e-05,
+      "loss": 1.1658,
+      "step": 6689
+    },
+    {
+      "epoch": 2.4907214575746646,
+      "grad_norm": 0.16502508521080017,
+      "learning_rate": 1.7702807900382296e-05,
+      "loss": 1.1815,
+      "step": 6690
+    },
+    {
+      "epoch": 2.491093762725274,
+      "grad_norm": 0.16367071866989136,
+      "learning_rate": 1.7702032830368525e-05,
+      "loss": 1.1812,
+      "step": 6691
+    },
+    {
+      "epoch": 2.4914660678758827,
+      "grad_norm": 0.16245703399181366,
+      "learning_rate": 1.7701257646596212e-05,
+      "loss": 1.1684,
+      "step": 6692
+    },
+    {
+      "epoch": 2.491838373026492,
+      "grad_norm": 0.16035568714141846,
+      "learning_rate": 1.7700482349076808e-05,
+      "loss": 1.1712,
+      "step": 6693
+    },
+    {
+      "epoch": 2.4922106781771007,
+      "grad_norm": 0.16250720620155334,
+      "learning_rate": 1.769970693782176e-05,
+      "loss": 1.1736,
+      "step": 6694
+    },
+    {
+      "epoch": 2.49258298332771,
+      "grad_norm": 0.1739581972360611,
+      "learning_rate": 1.7698931412842526e-05,
+      "loss": 1.1765,
+      "step": 6695
+    },
+    {
+      "epoch": 2.492955288478319,
+      "grad_norm": 0.16992726922035217,
+      "learning_rate": 1.7698155774150553e-05,
+      "loss": 1.1814,
+      "step": 6696
+    },
+    {
+      "epoch": 2.493327593628928,
+      "grad_norm": 0.16158819198608398,
+      "learning_rate": 1.769738002175731e-05,
+      "loss": 1.1727,
+      "step": 6697
+    },
+    {
+      "epoch": 2.4936998987795373,
+      "grad_norm": 0.17658640444278717,
+      "learning_rate": 1.769660415567424e-05,
+      "loss": 1.1675,
+      "step": 6698
+    },
+    {
+      "epoch": 2.494072203930146,
+      "grad_norm": 0.21519418060779572,
+      "learning_rate": 1.7695828175912816e-05,
+      "loss": 1.181,
+      "step": 6699
+    },
+    {
+      "epoch": 2.4944445090807554,
+      "grad_norm": 0.1897740513086319,
+      "learning_rate": 1.7695052082484493e-05,
+      "loss": 1.168,
+      "step": 6700
+    },
+    {
+      "epoch": 2.4948168142313643,
+      "grad_norm": 0.1640034019947052,
+      "learning_rate": 1.769427587540073e-05,
+      "loss": 1.1741,
+      "step": 6701
+    },
+    {
+      "epoch": 2.4951891193819735,
+      "grad_norm": 0.17497296631336212,
+      "learning_rate": 1.7693499554672996e-05,
+      "loss": 1.1988,
+      "step": 6702
+    },
+    {
+      "epoch": 2.4955614245325823,
+      "grad_norm": 0.16799713671207428,
+      "learning_rate": 1.7692723120312757e-05,
+      "loss": 1.1981,
+      "step": 6703
+    },
+    {
+      "epoch": 2.4959337296831916,
+      "grad_norm": 0.16971002519130707,
+      "learning_rate": 1.7691946572331477e-05,
+      "loss": 1.198,
+      "step": 6704
+    },
+    {
+      "epoch": 2.4963060348338004,
+      "grad_norm": 0.17075687646865845,
+      "learning_rate": 1.7691169910740633e-05,
+      "loss": 1.1848,
+      "step": 6705
+    },
+    {
+      "epoch": 2.4966783399844097,
+      "grad_norm": 0.16614024341106415,
+      "learning_rate": 1.7690393135551692e-05,
+      "loss": 1.1759,
+      "step": 6706
+    },
+    {
+      "epoch": 2.497050645135019,
+      "grad_norm": 0.17377164959907532,
+      "learning_rate": 1.7689616246776125e-05,
+      "loss": 1.1758,
+      "step": 6707
+    },
+    {
+      "epoch": 2.4974229502856278,
+      "grad_norm": 0.1613984853029251,
+      "learning_rate": 1.768883924442541e-05,
+      "loss": 1.1545,
+      "step": 6708
+    },
+    {
+      "epoch": 2.497795255436237,
+      "grad_norm": 0.16565319895744324,
+      "learning_rate": 1.7688062128511023e-05,
+      "loss": 1.1795,
+      "step": 6709
+    },
+    {
+      "epoch": 2.498167560586846,
+      "grad_norm": 0.17493651807308197,
+      "learning_rate": 1.768728489904444e-05,
+      "loss": 1.1901,
+      "step": 6710
+    },
+    {
+      "epoch": 2.498539865737455,
+      "grad_norm": 0.16326481103897095,
+      "learning_rate": 1.7686507556037136e-05,
+      "loss": 1.1731,
+      "step": 6711
+    },
+    {
+      "epoch": 2.498912170888064,
+      "grad_norm": 0.1633228361606598,
+      "learning_rate": 1.7685730099500606e-05,
+      "loss": 1.1756,
+      "step": 6712
+    },
+    {
+      "epoch": 2.499284476038673,
+      "grad_norm": 0.16988112032413483,
+      "learning_rate": 1.768495252944632e-05,
+      "loss": 1.1778,
+      "step": 6713
+    },
+    {
+      "epoch": 2.499656781189282,
+      "grad_norm": 0.165721595287323,
+      "learning_rate": 1.7684174845885768e-05,
+      "loss": 1.178,
+      "step": 6714
+    },
+    {
+      "epoch": 2.5000290863398913,
+      "grad_norm": 0.1885932981967926,
+      "learning_rate": 1.768339704883044e-05,
+      "loss": 1.1743,
+      "step": 6715
+    },
+    {
+      "epoch": 2.5004013914905006,
+      "grad_norm": 0.16593773663043976,
+      "learning_rate": 1.768261913829181e-05,
+      "loss": 1.1713,
+      "step": 6716
+    },
+    {
+      "epoch": 2.5007736966411094,
+      "grad_norm": 0.16606302559375763,
+      "learning_rate": 1.7681841114281387e-05,
+      "loss": 1.1694,
+      "step": 6717
+    },
+    {
+      "epoch": 2.5011460017917186,
+      "grad_norm": 0.1713939607143402,
+      "learning_rate": 1.768106297681065e-05,
+      "loss": 1.167,
+      "step": 6718
+    },
+    {
+      "epoch": 2.5015183069423275,
+      "grad_norm": 0.172207772731781,
+      "learning_rate": 1.7680284725891095e-05,
+      "loss": 1.1824,
+      "step": 6719
+    },
+    {
+      "epoch": 2.5018906120929367,
+      "grad_norm": 0.1743270754814148,
+      "learning_rate": 1.7679506361534216e-05,
+      "loss": 1.1705,
+      "step": 6720
+    },
+    {
+      "epoch": 2.502262917243546,
+      "grad_norm": 0.16744069755077362,
+      "learning_rate": 1.7678727883751508e-05,
+      "loss": 1.1853,
+      "step": 6721
+    },
+    {
+      "epoch": 2.502635222394155,
+      "grad_norm": 0.18169431388378143,
+      "learning_rate": 1.7677949292554473e-05,
+      "loss": 1.1796,
+      "step": 6722
+    },
+    {
+      "epoch": 2.5030075275447636,
+      "grad_norm": 0.1659853309392929,
+      "learning_rate": 1.7677170587954607e-05,
+      "loss": 1.1716,
+      "step": 6723
+    },
+    {
+      "epoch": 2.503379832695373,
+      "grad_norm": 0.18055576086044312,
+      "learning_rate": 1.7676391769963416e-05,
+      "loss": 1.1826,
+      "step": 6724
+    },
+    {
+      "epoch": 2.503752137845982,
+      "grad_norm": 0.16379252076148987,
+      "learning_rate": 1.76756128385924e-05,
+      "loss": 1.1916,
+      "step": 6725
+    },
+    {
+      "epoch": 2.504124442996591,
+      "grad_norm": 0.16528604924678802,
+      "learning_rate": 1.7674833793853064e-05,
+      "loss": 1.1856,
+      "step": 6726
+    },
+    {
+      "epoch": 2.5044967481472002,
+      "grad_norm": 0.1737004965543747,
+      "learning_rate": 1.7674054635756914e-05,
+      "loss": 1.1808,
+      "step": 6727
+    },
+    {
+      "epoch": 2.504869053297809,
+      "grad_norm": 0.16210679709911346,
+      "learning_rate": 1.7673275364315458e-05,
+      "loss": 1.171,
+      "step": 6728
+    },
+    {
+      "epoch": 2.5052413584484183,
+      "grad_norm": 0.16479384899139404,
+      "learning_rate": 1.7672495979540207e-05,
+      "loss": 1.1741,
+      "step": 6729
+    },
+    {
+      "epoch": 2.5056136635990276,
+      "grad_norm": 0.16804920136928558,
+      "learning_rate": 1.7671716481442674e-05,
+      "loss": 1.1771,
+      "step": 6730
+    },
+    {
+      "epoch": 2.5059859687496364,
+      "grad_norm": 0.16524706780910492,
+      "learning_rate": 1.7670936870034366e-05,
+      "loss": 1.1762,
+      "step": 6731
+    },
+    {
+      "epoch": 2.5063582739002452,
+      "grad_norm": 0.163568913936615,
+      "learning_rate": 1.7670157145326806e-05,
+      "loss": 1.1838,
+      "step": 6732
+    },
+    {
+      "epoch": 2.5067305790508545,
+      "grad_norm": 0.15912410616874695,
+      "learning_rate": 1.7669377307331503e-05,
+      "loss": 1.1833,
+      "step": 6733
+    },
+    {
+      "epoch": 2.5071028842014638,
+      "grad_norm": 0.16776417195796967,
+      "learning_rate": 1.7668597356059977e-05,
+      "loss": 1.184,
+      "step": 6734
+    },
+    {
+      "epoch": 2.5074751893520726,
+      "grad_norm": 0.17335230112075806,
+      "learning_rate": 1.7667817291523753e-05,
+      "loss": 1.2042,
+      "step": 6735
+    },
+    {
+      "epoch": 2.507847494502682,
+      "grad_norm": 0.15524321794509888,
+      "learning_rate": 1.766703711373435e-05,
+      "loss": 1.1639,
+      "step": 6736
+    },
+    {
+      "epoch": 2.5082197996532907,
+      "grad_norm": 0.16183823347091675,
+      "learning_rate": 1.7666256822703288e-05,
+      "loss": 1.1863,
+      "step": 6737
+    },
+    {
+      "epoch": 2.5085921048039,
+      "grad_norm": 0.1681044101715088,
+      "learning_rate": 1.7665476418442092e-05,
+      "loss": 1.1777,
+      "step": 6738
+    },
+    {
+      "epoch": 2.508964409954509,
+      "grad_norm": 0.165101557970047,
+      "learning_rate": 1.766469590096229e-05,
+      "loss": 1.1829,
+      "step": 6739
+    },
+    {
+      "epoch": 2.509336715105118,
+      "grad_norm": 0.16544567048549652,
+      "learning_rate": 1.7663915270275413e-05,
+      "loss": 1.1665,
+      "step": 6740
+    },
+    {
+      "epoch": 2.509709020255727,
+      "grad_norm": 0.17958173155784607,
+      "learning_rate": 1.766313452639299e-05,
+      "loss": 1.1861,
+      "step": 6741
+    },
+    {
+      "epoch": 2.510081325406336,
+      "grad_norm": 0.19180700182914734,
+      "learning_rate": 1.766235366932655e-05,
+      "loss": 1.1774,
+      "step": 6742
+    },
+    {
+      "epoch": 2.5104536305569454,
+      "grad_norm": 0.1678982377052307,
+      "learning_rate": 1.7661572699087622e-05,
+      "loss": 1.1739,
+      "step": 6743
+    },
+    {
+      "epoch": 2.510825935707554,
+      "grad_norm": 0.16974776983261108,
+      "learning_rate": 1.7660791615687752e-05,
+      "loss": 1.1781,
+      "step": 6744
+    },
+    {
+      "epoch": 2.5111982408581635,
+      "grad_norm": 0.18585939705371857,
+      "learning_rate": 1.766001041913847e-05,
+      "loss": 1.1733,
+      "step": 6745
+    },
+    {
+      "epoch": 2.5115705460087723,
+      "grad_norm": 0.17436900734901428,
+      "learning_rate": 1.7659229109451312e-05,
+      "loss": 1.1856,
+      "step": 6746
+    },
+    {
+      "epoch": 2.5119428511593815,
+      "grad_norm": 0.1617736667394638,
+      "learning_rate": 1.765844768663782e-05,
+      "loss": 1.1664,
+      "step": 6747
+    },
+    {
+      "epoch": 2.512315156309991,
+      "grad_norm": 0.161626398563385,
+      "learning_rate": 1.765766615070954e-05,
+      "loss": 1.1708,
+      "step": 6748
+    },
+    {
+      "epoch": 2.5126874614605996,
+      "grad_norm": 0.16217051446437836,
+      "learning_rate": 1.7656884501678014e-05,
+      "loss": 1.1661,
+      "step": 6749
+    },
+    {
+      "epoch": 2.5130597666112084,
+      "grad_norm": 0.16591928899288177,
+      "learning_rate": 1.765610273955478e-05,
+      "loss": 1.1706,
+      "step": 6750
+    },
+    {
+      "epoch": 2.5134320717618177,
+      "grad_norm": 0.17161224782466888,
+      "learning_rate": 1.7655320864351388e-05,
+      "loss": 1.1799,
+      "step": 6751
+    },
+    {
+      "epoch": 2.513804376912427,
+      "grad_norm": 0.16334664821624756,
+      "learning_rate": 1.7654538876079387e-05,
+      "loss": 1.158,
+      "step": 6752
+    },
+    {
+      "epoch": 2.514176682063036,
+      "grad_norm": 0.15914195775985718,
+      "learning_rate": 1.7653756774750334e-05,
+      "loss": 1.1648,
+      "step": 6753
+    },
+    {
+      "epoch": 2.514548987213645,
+      "grad_norm": 0.18378858268260956,
+      "learning_rate": 1.7652974560375765e-05,
+      "loss": 1.1775,
+      "step": 6754
+    },
+    {
+      "epoch": 2.514921292364254,
+      "grad_norm": 0.21220822632312775,
+      "learning_rate": 1.7652192232967245e-05,
+      "loss": 1.1733,
+      "step": 6755
+    },
+    {
+      "epoch": 2.515293597514863,
+      "grad_norm": 0.1674444079399109,
+      "learning_rate": 1.7651409792536328e-05,
+      "loss": 1.1723,
+      "step": 6756
+    },
+    {
+      "epoch": 2.5156659026654724,
+      "grad_norm": 0.1930401474237442,
+      "learning_rate": 1.765062723909457e-05,
+      "loss": 1.1771,
+      "step": 6757
+    },
+    {
+      "epoch": 2.5160382078160812,
+      "grad_norm": 0.16456985473632812,
+      "learning_rate": 1.7649844572653523e-05,
+      "loss": 1.1803,
+      "step": 6758
+    },
+    {
+      "epoch": 2.51641051296669,
+      "grad_norm": 0.18256615102291107,
+      "learning_rate": 1.7649061793224752e-05,
+      "loss": 1.1701,
+      "step": 6759
+    },
+    {
+      "epoch": 2.5167828181172993,
+      "grad_norm": 0.16572584211826324,
+      "learning_rate": 1.7648278900819822e-05,
+      "loss": 1.1784,
+      "step": 6760
+    },
+    {
+      "epoch": 2.5171551232679086,
+      "grad_norm": 0.1717369556427002,
+      "learning_rate": 1.7647495895450292e-05,
+      "loss": 1.1871,
+      "step": 6761
+    },
+    {
+      "epoch": 2.5175274284185174,
+      "grad_norm": 0.16302919387817383,
+      "learning_rate": 1.7646712777127722e-05,
+      "loss": 1.1599,
+      "step": 6762
+    },
+    {
+      "epoch": 2.5178997335691267,
+      "grad_norm": 0.1645524501800537,
+      "learning_rate": 1.764592954586369e-05,
+      "loss": 1.1726,
+      "step": 6763
+    },
+    {
+      "epoch": 2.5182720387197355,
+      "grad_norm": 0.1693658083677292,
+      "learning_rate": 1.764514620166976e-05,
+      "loss": 1.1715,
+      "step": 6764
+    },
+    {
+      "epoch": 2.5186443438703447,
+      "grad_norm": 0.17292669415473938,
+      "learning_rate": 1.7644362744557498e-05,
+      "loss": 1.1747,
+      "step": 6765
+    },
+    {
+      "epoch": 2.519016649020954,
+      "grad_norm": 0.16915366053581238,
+      "learning_rate": 1.7643579174538475e-05,
+      "loss": 1.1728,
+      "step": 6766
+    },
+    {
+      "epoch": 2.519388954171563,
+      "grad_norm": 0.16110309958457947,
+      "learning_rate": 1.7642795491624268e-05,
+      "loss": 1.1649,
+      "step": 6767
+    },
+    {
+      "epoch": 2.519761259322172,
+      "grad_norm": 0.16794544458389282,
+      "learning_rate": 1.7642011695826455e-05,
+      "loss": 1.1718,
+      "step": 6768
+    },
+    {
+      "epoch": 2.520133564472781,
+      "grad_norm": 0.15866640210151672,
+      "learning_rate": 1.764122778715661e-05,
+      "loss": 1.1585,
+      "step": 6769
+    },
+    {
+      "epoch": 2.52050586962339,
+      "grad_norm": 0.16686029732227325,
+      "learning_rate": 1.7640443765626304e-05,
+      "loss": 1.1596,
+      "step": 6770
+    },
+    {
+      "epoch": 2.520878174773999,
+      "grad_norm": 0.16550125181674957,
+      "learning_rate": 1.7639659631247127e-05,
+      "loss": 1.1785,
+      "step": 6771
+    },
+    {
+      "epoch": 2.5212504799246083,
+      "grad_norm": 0.18188951909542084,
+      "learning_rate": 1.7638875384030654e-05,
+      "loss": 1.1607,
+      "step": 6772
+    },
+    {
+      "epoch": 2.521622785075217,
+      "grad_norm": 0.1796000748872757,
+      "learning_rate": 1.7638091023988473e-05,
+      "loss": 1.1778,
+      "step": 6773
+    },
+    {
+      "epoch": 2.5219950902258264,
+      "grad_norm": 0.17091524600982666,
+      "learning_rate": 1.7637306551132166e-05,
+      "loss": 1.1887,
+      "step": 6774
+    },
+    {
+      "epoch": 2.5223673953764356,
+      "grad_norm": 0.16658535599708557,
+      "learning_rate": 1.7636521965473324e-05,
+      "loss": 1.1695,
+      "step": 6775
+    },
+    {
+      "epoch": 2.5227397005270444,
+      "grad_norm": 0.16318367421627045,
+      "learning_rate": 1.7635737267023527e-05,
+      "loss": 1.1821,
+      "step": 6776
+    },
+    {
+      "epoch": 2.5231120056776537,
+      "grad_norm": 0.16290733218193054,
+      "learning_rate": 1.7634952455794373e-05,
+      "loss": 1.1683,
+      "step": 6777
+    },
+    {
+      "epoch": 2.5234843108282625,
+      "grad_norm": 0.16749358177185059,
+      "learning_rate": 1.7634167531797447e-05,
+      "loss": 1.1599,
+      "step": 6778
+    },
+    {
+      "epoch": 2.523856615978872,
+      "grad_norm": 0.16928349435329437,
+      "learning_rate": 1.7633382495044347e-05,
+      "loss": 1.1721,
+      "step": 6779
+    },
+    {
+      "epoch": 2.5242289211294806,
+      "grad_norm": 0.17672182619571686,
+      "learning_rate": 1.7632597345546667e-05,
+      "loss": 1.179,
+      "step": 6780
+    },
+    {
+      "epoch": 2.52460122628009,
+      "grad_norm": 0.21029046177864075,
+      "learning_rate": 1.7631812083316003e-05,
+      "loss": 1.1655,
+      "step": 6781
+    },
+    {
+      "epoch": 2.5249735314306987,
+      "grad_norm": 0.26718470454216003,
+      "learning_rate": 1.7631026708363956e-05,
+      "loss": 1.1803,
+      "step": 6782
+    },
+    {
+      "epoch": 2.525345836581308,
+      "grad_norm": 0.22429171204566956,
+      "learning_rate": 1.763024122070212e-05,
+      "loss": 1.1808,
+      "step": 6783
+    },
+    {
+      "epoch": 2.525718141731917,
+      "grad_norm": 0.1687416434288025,
+      "learning_rate": 1.76294556203421e-05,
+      "loss": 1.1771,
+      "step": 6784
+    },
+    {
+      "epoch": 2.526090446882526,
+      "grad_norm": 0.18023616075515747,
+      "learning_rate": 1.7628669907295504e-05,
+      "loss": 1.1736,
+      "step": 6785
+    },
+    {
+      "epoch": 2.5264627520331353,
+      "grad_norm": 0.17839936912059784,
+      "learning_rate": 1.762788408157393e-05,
+      "loss": 1.1743,
+      "step": 6786
+    },
+    {
+      "epoch": 2.526835057183744,
+      "grad_norm": 0.16387495398521423,
+      "learning_rate": 1.7627098143188982e-05,
+      "loss": 1.1711,
+      "step": 6787
+    },
+    {
+      "epoch": 2.5272073623343534,
+      "grad_norm": 0.17329815030097961,
+      "learning_rate": 1.762631209215228e-05,
+      "loss": 1.1782,
+      "step": 6788
+    },
+    {
+      "epoch": 2.527579667484962,
+      "grad_norm": 0.1764519065618515,
+      "learning_rate": 1.7625525928475424e-05,
+      "loss": 1.1734,
+      "step": 6789
+    },
+    {
+      "epoch": 2.5279519726355715,
+      "grad_norm": 0.1638839989900589,
+      "learning_rate": 1.762473965217003e-05,
+      "loss": 1.1666,
+      "step": 6790
+    },
+    {
+      "epoch": 2.5283242777861803,
+      "grad_norm": 0.16194722056388855,
+      "learning_rate": 1.7623953263247707e-05,
+      "loss": 1.1526,
+      "step": 6791
+    },
+    {
+      "epoch": 2.5286965829367896,
+      "grad_norm": 0.17311038076877594,
+      "learning_rate": 1.7623166761720075e-05,
+      "loss": 1.1802,
+      "step": 6792
+    },
+    {
+      "epoch": 2.529068888087399,
+      "grad_norm": 0.1649443656206131,
+      "learning_rate": 1.7622380147598745e-05,
+      "loss": 1.1698,
+      "step": 6793
+    },
+    {
+      "epoch": 2.5294411932380076,
+      "grad_norm": 0.16059640049934387,
+      "learning_rate": 1.7621593420895342e-05,
+      "loss": 1.1687,
+      "step": 6794
+    },
+    {
+      "epoch": 2.529813498388617,
+      "grad_norm": 0.17570586502552032,
+      "learning_rate": 1.762080658162148e-05,
+      "loss": 1.1768,
+      "step": 6795
+    },
+    {
+      "epoch": 2.5301858035392257,
+      "grad_norm": 0.1651303470134735,
+      "learning_rate": 1.7620019629788786e-05,
+      "loss": 1.1755,
+      "step": 6796
+    },
+    {
+      "epoch": 2.530558108689835,
+      "grad_norm": 0.16856788098812103,
+      "learning_rate": 1.7619232565408878e-05,
+      "loss": 1.1711,
+      "step": 6797
+    },
+    {
+      "epoch": 2.530930413840444,
+      "grad_norm": 0.16949835419654846,
+      "learning_rate": 1.7618445388493386e-05,
+      "loss": 1.1839,
+      "step": 6798
+    },
+    {
+      "epoch": 2.531302718991053,
+      "grad_norm": 0.17608442902565002,
+      "learning_rate": 1.761765809905393e-05,
+      "loss": 1.1857,
+      "step": 6799
+    },
+    {
+      "epoch": 2.531675024141662,
+      "grad_norm": 0.16439513862133026,
+      "learning_rate": 1.7616870697102144e-05,
+      "loss": 1.1789,
+      "step": 6800
+    },
+    {
+      "epoch": 2.532047329292271,
+      "grad_norm": 0.1684199720621109,
+      "learning_rate": 1.7616083182649654e-05,
+      "loss": 1.1857,
+      "step": 6801
+    },
+    {
+      "epoch": 2.5324196344428804,
+      "grad_norm": 0.16960062086582184,
+      "learning_rate": 1.7615295555708098e-05,
+      "loss": 1.1714,
+      "step": 6802
+    },
+    {
+      "epoch": 2.5327919395934892,
+      "grad_norm": 0.162268728017807,
+      "learning_rate": 1.7614507816289102e-05,
+      "loss": 1.1742,
+      "step": 6803
+    },
+    {
+      "epoch": 2.5331642447440985,
+      "grad_norm": 0.17141248285770416,
+      "learning_rate": 1.7613719964404303e-05,
+      "loss": 1.1886,
+      "step": 6804
+    },
+    {
+      "epoch": 2.5335365498947073,
+      "grad_norm": 0.17517106235027313,
+      "learning_rate": 1.7612932000065336e-05,
+      "loss": 1.1835,
+      "step": 6805
+    },
+    {
+      "epoch": 2.5339088550453166,
+      "grad_norm": 0.1663627028465271,
+      "learning_rate": 1.7612143923283844e-05,
+      "loss": 1.1616,
+      "step": 6806
+    },
+    {
+      "epoch": 2.5342811601959254,
+      "grad_norm": 0.1708042472600937,
+      "learning_rate": 1.7611355734071464e-05,
+      "loss": 1.1604,
+      "step": 6807
+    },
+    {
+      "epoch": 2.5346534653465347,
+      "grad_norm": 0.1797637939453125,
+      "learning_rate": 1.7610567432439834e-05,
+      "loss": 1.1802,
+      "step": 6808
+    },
+    {
+      "epoch": 2.5350257704971435,
+      "grad_norm": 0.16849200427532196,
+      "learning_rate": 1.7609779018400606e-05,
+      "loss": 1.1651,
+      "step": 6809
+    },
+    {
+      "epoch": 2.5353980756477528,
+      "grad_norm": 0.17425650358200073,
+      "learning_rate": 1.7608990491965416e-05,
+      "loss": 1.1741,
+      "step": 6810
+    },
+    {
+      "epoch": 2.535770380798362,
+      "grad_norm": 0.18015819787979126,
+      "learning_rate": 1.760820185314591e-05,
+      "loss": 1.1862,
+      "step": 6811
+    },
+    {
+      "epoch": 2.536142685948971,
+      "grad_norm": 0.16850873827934265,
+      "learning_rate": 1.7607413101953747e-05,
+      "loss": 1.1758,
+      "step": 6812
+    },
+    {
+      "epoch": 2.53651499109958,
+      "grad_norm": 0.1831638067960739,
+      "learning_rate": 1.7606624238400568e-05,
+      "loss": 1.1718,
+      "step": 6813
+    },
+    {
+      "epoch": 2.536887296250189,
+      "grad_norm": 0.16078107059001923,
+      "learning_rate": 1.7605835262498027e-05,
+      "loss": 1.1651,
+      "step": 6814
+    },
+    {
+      "epoch": 2.537259601400798,
+      "grad_norm": 0.17661617696285248,
+      "learning_rate": 1.7605046174257775e-05,
+      "loss": 1.1663,
+      "step": 6815
+    },
+    {
+      "epoch": 2.5376319065514075,
+      "grad_norm": 0.17243711650371552,
+      "learning_rate": 1.7604256973691468e-05,
+      "loss": 1.173,
+      "step": 6816
+    },
+    {
+      "epoch": 2.5380042117020163,
+      "grad_norm": 0.16753429174423218,
+      "learning_rate": 1.7603467660810763e-05,
+      "loss": 1.1635,
+      "step": 6817
+    },
+    {
+      "epoch": 2.538376516852625,
+      "grad_norm": 0.17135019600391388,
+      "learning_rate": 1.7602678235627317e-05,
+      "loss": 1.1634,
+      "step": 6818
+    },
+    {
+      "epoch": 2.5387488220032344,
+      "grad_norm": 0.16117660701274872,
+      "learning_rate": 1.7601888698152794e-05,
+      "loss": 1.1647,
+      "step": 6819
+    },
+    {
+      "epoch": 2.5391211271538436,
+      "grad_norm": 0.17489348351955414,
+      "learning_rate": 1.760109904839885e-05,
+      "loss": 1.1938,
+      "step": 6820
+    },
+    {
+      "epoch": 2.5394934323044525,
+      "grad_norm": 0.16629846394062042,
+      "learning_rate": 1.760030928637715e-05,
+      "loss": 1.1785,
+      "step": 6821
+    },
+    {
+      "epoch": 2.5398657374550617,
+      "grad_norm": 0.1682901829481125,
+      "learning_rate": 1.759951941209936e-05,
+      "loss": 1.1865,
+      "step": 6822
+    },
+    {
+      "epoch": 2.5402380426056705,
+      "grad_norm": 0.1646634191274643,
+      "learning_rate": 1.7598729425577143e-05,
+      "loss": 1.1787,
+      "step": 6823
+    },
+    {
+      "epoch": 2.54061034775628,
+      "grad_norm": 0.17023120820522308,
+      "learning_rate": 1.759793932682217e-05,
+      "loss": 1.1697,
+      "step": 6824
+    },
+    {
+      "epoch": 2.540982652906889,
+      "grad_norm": 0.17722590267658234,
+      "learning_rate": 1.759714911584611e-05,
+      "loss": 1.1666,
+      "step": 6825
+    },
+    {
+      "epoch": 2.541354958057498,
+      "grad_norm": 0.1645139455795288,
+      "learning_rate": 1.7596358792660633e-05,
+      "loss": 1.1907,
+      "step": 6826
+    },
+    {
+      "epoch": 2.5417272632081067,
+      "grad_norm": 0.1630629450082779,
+      "learning_rate": 1.7595568357277413e-05,
+      "loss": 1.1829,
+      "step": 6827
+    },
+    {
+      "epoch": 2.542099568358716,
+      "grad_norm": 0.16655808687210083,
+      "learning_rate": 1.759477780970813e-05,
+      "loss": 1.1751,
+      "step": 6828
+    },
+    {
+      "epoch": 2.5424718735093252,
+      "grad_norm": 0.19367186725139618,
+      "learning_rate": 1.759398714996445e-05,
+      "loss": 1.1613,
+      "step": 6829
+    },
+    {
+      "epoch": 2.542844178659934,
+      "grad_norm": 0.2233743518590927,
+      "learning_rate": 1.759319637805806e-05,
+      "loss": 1.1836,
+      "step": 6830
+    },
+    {
+      "epoch": 2.5432164838105433,
+      "grad_norm": 0.19930030405521393,
+      "learning_rate": 1.7592405494000635e-05,
+      "loss": 1.1862,
+      "step": 6831
+    },
+    {
+      "epoch": 2.543588788961152,
+      "grad_norm": 0.1649360954761505,
+      "learning_rate": 1.7591614497803856e-05,
+      "loss": 1.1586,
+      "step": 6832
+    },
+    {
+      "epoch": 2.5439610941117614,
+      "grad_norm": 0.1704995036125183,
+      "learning_rate": 1.7590823389479407e-05,
+      "loss": 1.1979,
+      "step": 6833
+    },
+    {
+      "epoch": 2.5443333992623707,
+      "grad_norm": 0.20266738533973694,
+      "learning_rate": 1.7590032169038974e-05,
+      "loss": 1.1811,
+      "step": 6834
+    },
+    {
+      "epoch": 2.5447057044129795,
+      "grad_norm": 0.1894606202840805,
+      "learning_rate": 1.7589240836494245e-05,
+      "loss": 1.1745,
+      "step": 6835
+    },
+    {
+      "epoch": 2.5450780095635883,
+      "grad_norm": 0.17341409623622894,
+      "learning_rate": 1.7588449391856903e-05,
+      "loss": 1.1793,
+      "step": 6836
+    },
+    {
+      "epoch": 2.5454503147141976,
+      "grad_norm": 0.20491334795951843,
+      "learning_rate": 1.758765783513864e-05,
+      "loss": 1.1762,
+      "step": 6837
+    },
+    {
+      "epoch": 2.545822619864807,
+      "grad_norm": 0.1708584725856781,
+      "learning_rate": 1.758686616635114e-05,
+      "loss": 1.1693,
+      "step": 6838
+    },
+    {
+      "epoch": 2.5461949250154157,
+      "grad_norm": 0.17428459227085114,
+      "learning_rate": 1.7586074385506114e-05,
+      "loss": 1.1827,
+      "step": 6839
+    },
+    {
+      "epoch": 2.546567230166025,
+      "grad_norm": 0.19591780006885529,
+      "learning_rate": 1.758528249261524e-05,
+      "loss": 1.1732,
+      "step": 6840
+    },
+    {
+      "epoch": 2.5469395353166338,
+      "grad_norm": 0.16241784393787384,
+      "learning_rate": 1.758449048769022e-05,
+      "loss": 1.1675,
+      "step": 6841
+    },
+    {
+      "epoch": 2.547311840467243,
+      "grad_norm": 0.17725640535354614,
+      "learning_rate": 1.758369837074275e-05,
+      "loss": 1.166,
+      "step": 6842
+    },
+    {
+      "epoch": 2.5476841456178523,
+      "grad_norm": 0.17130807042121887,
+      "learning_rate": 1.7582906141784534e-05,
+      "loss": 1.1961,
+      "step": 6843
+    },
+    {
+      "epoch": 2.548056450768461,
+      "grad_norm": 0.18036004900932312,
+      "learning_rate": 1.758211380082727e-05,
+      "loss": 1.1713,
+      "step": 6844
+    },
+    {
+      "epoch": 2.54842875591907,
+      "grad_norm": 0.16558697819709778,
+      "learning_rate": 1.7581321347882657e-05,
+      "loss": 1.1733,
+      "step": 6845
+    },
+    {
+      "epoch": 2.548801061069679,
+      "grad_norm": 0.17441225051879883,
+      "learning_rate": 1.7580528782962408e-05,
+      "loss": 1.1662,
+      "step": 6846
+    },
+    {
+      "epoch": 2.5491733662202885,
+      "grad_norm": 0.16816645860671997,
+      "learning_rate": 1.757973610607822e-05,
+      "loss": 1.1792,
+      "step": 6847
+    },
+    {
+      "epoch": 2.5495456713708973,
+      "grad_norm": 0.19029933214187622,
+      "learning_rate": 1.757894331724181e-05,
+      "loss": 1.1739,
+      "step": 6848
+    },
+    {
+      "epoch": 2.5499179765215065,
+      "grad_norm": 0.15939323604106903,
+      "learning_rate": 1.757815041646488e-05,
+      "loss": 1.1672,
+      "step": 6849
+    },
+    {
+      "epoch": 2.5502902816721154,
+      "grad_norm": 0.1666376292705536,
+      "learning_rate": 1.7577357403759147e-05,
+      "loss": 1.1632,
+      "step": 6850
+    },
+    {
+      "epoch": 2.5506625868227246,
+      "grad_norm": 0.16542813181877136,
+      "learning_rate": 1.7576564279136318e-05,
+      "loss": 1.1823,
+      "step": 6851
+    },
+    {
+      "epoch": 2.551034891973334,
+      "grad_norm": 0.19840499758720398,
+      "learning_rate": 1.757577104260811e-05,
+      "loss": 1.1901,
+      "step": 6852
+    },
+    {
+      "epoch": 2.5514071971239427,
+      "grad_norm": 0.17866215109825134,
+      "learning_rate": 1.757497769418624e-05,
+      "loss": 1.1641,
+      "step": 6853
+    },
+    {
+      "epoch": 2.5517795022745515,
+      "grad_norm": 0.16423308849334717,
+      "learning_rate": 1.7574184233882424e-05,
+      "loss": 1.176,
+      "step": 6854
+    },
+    {
+      "epoch": 2.552151807425161,
+      "grad_norm": 0.1964891403913498,
+      "learning_rate": 1.7573390661708386e-05,
+      "loss": 1.1694,
+      "step": 6855
+    },
+    {
+      "epoch": 2.55252411257577,
+      "grad_norm": 0.16654036939144135,
+      "learning_rate": 1.7572596977675837e-05,
+      "loss": 1.1754,
+      "step": 6856
+    },
+    {
+      "epoch": 2.552896417726379,
+      "grad_norm": 0.18395081162452698,
+      "learning_rate": 1.757180318179651e-05,
+      "loss": 1.1691,
+      "step": 6857
+    },
+    {
+      "epoch": 2.553268722876988,
+      "grad_norm": 0.18189731240272522,
+      "learning_rate": 1.7571009274082124e-05,
+      "loss": 1.181,
+      "step": 6858
+    },
+    {
+      "epoch": 2.553641028027597,
+      "grad_norm": 0.1674988567829132,
+      "learning_rate": 1.7570215254544406e-05,
+      "loss": 1.173,
+      "step": 6859
+    },
+    {
+      "epoch": 2.5540133331782062,
+      "grad_norm": 0.1818539947271347,
+      "learning_rate": 1.7569421123195086e-05,
+      "loss": 1.1811,
+      "step": 6860
+    },
+    {
+      "epoch": 2.5543856383288155,
+      "grad_norm": 0.1642785221338272,
+      "learning_rate": 1.7568626880045888e-05,
+      "loss": 1.1886,
+      "step": 6861
+    },
+    {
+      "epoch": 2.5547579434794243,
+      "grad_norm": 0.21301640570163727,
+      "learning_rate": 1.7567832525108547e-05,
+      "loss": 1.1762,
+      "step": 6862
+    },
+    {
+      "epoch": 2.555130248630033,
+      "grad_norm": 0.1925705522298813,
+      "learning_rate": 1.7567038058394797e-05,
+      "loss": 1.1685,
+      "step": 6863
+    },
+    {
+      "epoch": 2.5555025537806424,
+      "grad_norm": 0.1762634962797165,
+      "learning_rate": 1.7566243479916365e-05,
+      "loss": 1.1641,
+      "step": 6864
+    },
+    {
+      "epoch": 2.5558748589312517,
+      "grad_norm": 0.19311204552650452,
+      "learning_rate": 1.7565448789684996e-05,
+      "loss": 1.1789,
+      "step": 6865
+    },
+    {
+      "epoch": 2.5562471640818605,
+      "grad_norm": 0.17269863188266754,
+      "learning_rate": 1.756465398771242e-05,
+      "loss": 1.1702,
+      "step": 6866
+    },
+    {
+      "epoch": 2.5566194692324697,
+      "grad_norm": 0.1711881160736084,
+      "learning_rate": 1.7563859074010382e-05,
+      "loss": 1.1752,
+      "step": 6867
+    },
+    {
+      "epoch": 2.5569917743830786,
+      "grad_norm": 0.1725349724292755,
+      "learning_rate": 1.756306404859062e-05,
+      "loss": 1.1673,
+      "step": 6868
+    },
+    {
+      "epoch": 2.557364079533688,
+      "grad_norm": 0.1748928427696228,
+      "learning_rate": 1.7562268911464872e-05,
+      "loss": 1.1651,
+      "step": 6869
+    },
+    {
+      "epoch": 2.557736384684297,
+      "grad_norm": 0.17474377155303955,
+      "learning_rate": 1.7561473662644893e-05,
+      "loss": 1.1771,
+      "step": 6870
+    },
+    {
+      "epoch": 2.558108689834906,
+      "grad_norm": 0.17705856263637543,
+      "learning_rate": 1.7560678302142418e-05,
+      "loss": 1.1725,
+      "step": 6871
+    },
+    {
+      "epoch": 2.5584809949855147,
+      "grad_norm": 0.16561757028102875,
+      "learning_rate": 1.7559882829969203e-05,
+      "loss": 1.1827,
+      "step": 6872
+    },
+    {
+      "epoch": 2.558853300136124,
+      "grad_norm": 0.17302724719047546,
+      "learning_rate": 1.7559087246136987e-05,
+      "loss": 1.1834,
+      "step": 6873
+    },
+    {
+      "epoch": 2.5592256052867333,
+      "grad_norm": 0.16074450314044952,
+      "learning_rate": 1.755829155065753e-05,
+      "loss": 1.184,
+      "step": 6874
+    },
+    {
+      "epoch": 2.559597910437342,
+      "grad_norm": 0.16113080084323883,
+      "learning_rate": 1.7557495743542586e-05,
+      "loss": 1.1707,
+      "step": 6875
+    },
+    {
+      "epoch": 2.5599702155879513,
+      "grad_norm": 0.1629118025302887,
+      "learning_rate": 1.7556699824803897e-05,
+      "loss": 1.1578,
+      "step": 6876
+    },
+    {
+      "epoch": 2.56034252073856,
+      "grad_norm": 0.16696172952651978,
+      "learning_rate": 1.7555903794453232e-05,
+      "loss": 1.1842,
+      "step": 6877
+    },
+    {
+      "epoch": 2.5607148258891694,
+      "grad_norm": 0.16263677179813385,
+      "learning_rate": 1.7555107652502337e-05,
+      "loss": 1.1854,
+      "step": 6878
+    },
+    {
+      "epoch": 2.5610871310397787,
+      "grad_norm": 0.163103848695755,
+      "learning_rate": 1.7554311398962976e-05,
+      "loss": 1.173,
+      "step": 6879
+    },
+    {
+      "epoch": 2.5614594361903875,
+      "grad_norm": 0.18173319101333618,
+      "learning_rate": 1.7553515033846913e-05,
+      "loss": 1.1643,
+      "step": 6880
+    },
+    {
+      "epoch": 2.5618317413409963,
+      "grad_norm": 0.1870352178812027,
+      "learning_rate": 1.7552718557165907e-05,
+      "loss": 1.1847,
+      "step": 6881
+    },
+    {
+      "epoch": 2.5622040464916056,
+      "grad_norm": 0.166317880153656,
+      "learning_rate": 1.755192196893172e-05,
+      "loss": 1.1683,
+      "step": 6882
+    },
+    {
+      "epoch": 2.562576351642215,
+      "grad_norm": 0.2503323554992676,
+      "learning_rate": 1.755112526915612e-05,
+      "loss": 1.1824,
+      "step": 6883
+    },
+    {
+      "epoch": 2.5629486567928237,
+      "grad_norm": 0.2117423117160797,
+      "learning_rate": 1.7550328457850873e-05,
+      "loss": 1.1667,
+      "step": 6884
+    },
+    {
+      "epoch": 2.563320961943433,
+      "grad_norm": 0.19109854102134705,
+      "learning_rate": 1.754953153502775e-05,
+      "loss": 1.1603,
+      "step": 6885
+    },
+    {
+      "epoch": 2.5636932670940418,
+      "grad_norm": 0.1728496551513672,
+      "learning_rate": 1.754873450069852e-05,
+      "loss": 1.1538,
+      "step": 6886
+    },
+    {
+      "epoch": 2.564065572244651,
+      "grad_norm": 0.20083752274513245,
+      "learning_rate": 1.7547937354874953e-05,
+      "loss": 1.1703,
+      "step": 6887
+    },
+    {
+      "epoch": 2.5644378773952603,
+      "grad_norm": 0.1984606832265854,
+      "learning_rate": 1.7547140097568827e-05,
+      "loss": 1.1897,
+      "step": 6888
+    },
+    {
+      "epoch": 2.564810182545869,
+      "grad_norm": 0.1722983717918396,
+      "learning_rate": 1.7546342728791915e-05,
+      "loss": 1.1647,
+      "step": 6889
+    },
+    {
+      "epoch": 2.5651824876964784,
+      "grad_norm": 0.17819280922412872,
+      "learning_rate": 1.7545545248555994e-05,
+      "loss": 1.1718,
+      "step": 6890
+    },
+    {
+      "epoch": 2.565554792847087,
+      "grad_norm": 0.17654377222061157,
+      "learning_rate": 1.754474765687284e-05,
+      "loss": 1.1609,
+      "step": 6891
+    },
+    {
+      "epoch": 2.5659270979976965,
+      "grad_norm": 0.16665366291999817,
+      "learning_rate": 1.7543949953754244e-05,
+      "loss": 1.1773,
+      "step": 6892
+    },
+    {
+      "epoch": 2.5662994031483053,
+      "grad_norm": 0.16269910335540771,
+      "learning_rate": 1.7543152139211973e-05,
+      "loss": 1.1675,
+      "step": 6893
+    },
+    {
+      "epoch": 2.5666717082989146,
+      "grad_norm": 0.16960933804512024,
+      "learning_rate": 1.7542354213257825e-05,
+      "loss": 1.1802,
+      "step": 6894
+    },
+    {
+      "epoch": 2.5670440134495234,
+      "grad_norm": 0.18422453105449677,
+      "learning_rate": 1.7541556175903577e-05,
+      "loss": 1.1859,
+      "step": 6895
+    },
+    {
+      "epoch": 2.5674163186001326,
+      "grad_norm": 0.15753304958343506,
+      "learning_rate": 1.7540758027161014e-05,
+      "loss": 1.1712,
+      "step": 6896
+    },
+    {
+      "epoch": 2.567788623750742,
+      "grad_norm": 0.1667400747537613,
+      "learning_rate": 1.753995976704193e-05,
+      "loss": 1.1902,
+      "step": 6897
+    },
+    {
+      "epoch": 2.5681609289013507,
+      "grad_norm": 0.1718086153268814,
+      "learning_rate": 1.7539161395558115e-05,
+      "loss": 1.167,
+      "step": 6898
+    },
+    {
+      "epoch": 2.56853323405196,
+      "grad_norm": 0.1692952811717987,
+      "learning_rate": 1.7538362912721356e-05,
+      "loss": 1.1707,
+      "step": 6899
+    },
+    {
+      "epoch": 2.568905539202569,
+      "grad_norm": 0.16457752883434296,
+      "learning_rate": 1.7537564318543455e-05,
+      "loss": 1.1923,
+      "step": 6900
+    },
+    {
+      "epoch": 2.569277844353178,
+      "grad_norm": 0.16177764534950256,
+      "learning_rate": 1.7536765613036198e-05,
+      "loss": 1.1722,
+      "step": 6901
+    },
+    {
+      "epoch": 2.569650149503787,
+      "grad_norm": 0.17193439602851868,
+      "learning_rate": 1.7535966796211387e-05,
+      "loss": 1.1789,
+      "step": 6902
+    },
+    {
+      "epoch": 2.570022454654396,
+      "grad_norm": 0.1670045256614685,
+      "learning_rate": 1.753516786808082e-05,
+      "loss": 1.1715,
+      "step": 6903
+    },
+    {
+      "epoch": 2.570394759805005,
+      "grad_norm": 0.1643945276737213,
+      "learning_rate": 1.7534368828656295e-05,
+      "loss": 1.1788,
+      "step": 6904
+    },
+    {
+      "epoch": 2.5707670649556142,
+      "grad_norm": 0.1577821522951126,
+      "learning_rate": 1.7533569677949616e-05,
+      "loss": 1.1758,
+      "step": 6905
+    },
+    {
+      "epoch": 2.5711393701062235,
+      "grad_norm": 0.16875483095645905,
+      "learning_rate": 1.7532770415972585e-05,
+      "loss": 1.1799,
+      "step": 6906
+    },
+    {
+      "epoch": 2.5715116752568323,
+      "grad_norm": 0.1640816628932953,
+      "learning_rate": 1.7531971042737008e-05,
+      "loss": 1.1744,
+      "step": 6907
+    },
+    {
+      "epoch": 2.5718839804074416,
+      "grad_norm": 0.16457021236419678,
+      "learning_rate": 1.7531171558254692e-05,
+      "loss": 1.1772,
+      "step": 6908
+    },
+    {
+      "epoch": 2.5722562855580504,
+      "grad_norm": 0.16494521498680115,
+      "learning_rate": 1.7530371962537445e-05,
+      "loss": 1.1781,
+      "step": 6909
+    },
+    {
+      "epoch": 2.5726285907086597,
+      "grad_norm": 0.16351187229156494,
+      "learning_rate": 1.7529572255597077e-05,
+      "loss": 1.1792,
+      "step": 6910
+    },
+    {
+      "epoch": 2.5730008958592685,
+      "grad_norm": 0.1582973450422287,
+      "learning_rate": 1.7528772437445396e-05,
+      "loss": 1.1826,
+      "step": 6911
+    },
+    {
+      "epoch": 2.5733732010098778,
+      "grad_norm": 0.16494831442832947,
+      "learning_rate": 1.7527972508094223e-05,
+      "loss": 1.1892,
+      "step": 6912
+    },
+    {
+      "epoch": 2.5737455061604866,
+      "grad_norm": 0.16706949472427368,
+      "learning_rate": 1.7527172467555367e-05,
+      "loss": 1.1759,
+      "step": 6913
+    },
+    {
+      "epoch": 2.574117811311096,
+      "grad_norm": 0.16176964342594147,
+      "learning_rate": 1.752637231584064e-05,
+      "loss": 1.181,
+      "step": 6914
+    },
+    {
+      "epoch": 2.574490116461705,
+      "grad_norm": 0.1603250950574875,
+      "learning_rate": 1.7525572052961877e-05,
+      "loss": 1.1735,
+      "step": 6915
+    },
+    {
+      "epoch": 2.574862421612314,
+      "grad_norm": 0.1632986068725586,
+      "learning_rate": 1.752477167893088e-05,
+      "loss": 1.1669,
+      "step": 6916
+    },
+    {
+      "epoch": 2.575234726762923,
+      "grad_norm": 0.16188709437847137,
+      "learning_rate": 1.7523971193759482e-05,
+      "loss": 1.176,
+      "step": 6917
+    },
+    {
+      "epoch": 2.575607031913532,
+      "grad_norm": 0.16379517316818237,
+      "learning_rate": 1.7523170597459497e-05,
+      "loss": 1.1666,
+      "step": 6918
+    },
+    {
+      "epoch": 2.5759793370641413,
+      "grad_norm": 0.16542209684848785,
+      "learning_rate": 1.7522369890042755e-05,
+      "loss": 1.1477,
+      "step": 6919
+    },
+    {
+      "epoch": 2.57635164221475,
+      "grad_norm": 0.16362489759922028,
+      "learning_rate": 1.7521569071521084e-05,
+      "loss": 1.1709,
+      "step": 6920
+    },
+    {
+      "epoch": 2.5767239473653594,
+      "grad_norm": 0.18013128638267517,
+      "learning_rate": 1.752076814190631e-05,
+      "loss": 1.1737,
+      "step": 6921
+    },
+    {
+      "epoch": 2.577096252515968,
+      "grad_norm": 0.1668436974287033,
+      "learning_rate": 1.7519967101210264e-05,
+      "loss": 1.1797,
+      "step": 6922
+    },
+    {
+      "epoch": 2.5774685576665775,
+      "grad_norm": 0.17277368903160095,
+      "learning_rate": 1.751916594944477e-05,
+      "loss": 1.1708,
+      "step": 6923
+    },
+    {
+      "epoch": 2.5778408628171867,
+      "grad_norm": 0.1680234968662262,
+      "learning_rate": 1.751836468662167e-05,
+      "loss": 1.1707,
+      "step": 6924
+    },
+    {
+      "epoch": 2.5782131679677955,
+      "grad_norm": 0.17191176116466522,
+      "learning_rate": 1.7517563312752796e-05,
+      "loss": 1.1807,
+      "step": 6925
+    },
+    {
+      "epoch": 2.578585473118405,
+      "grad_norm": 0.16285885870456696,
+      "learning_rate": 1.7516761827849987e-05,
+      "loss": 1.1707,
+      "step": 6926
+    },
+    {
+      "epoch": 2.5789577782690136,
+      "grad_norm": 0.16699667274951935,
+      "learning_rate": 1.7515960231925072e-05,
+      "loss": 1.1765,
+      "step": 6927
+    },
+    {
+      "epoch": 2.579330083419623,
+      "grad_norm": 0.1778481900691986,
+      "learning_rate": 1.7515158524989896e-05,
+      "loss": 1.1785,
+      "step": 6928
+    },
+    {
+      "epoch": 2.579702388570232,
+      "grad_norm": 0.17215748131275177,
+      "learning_rate": 1.7514356707056303e-05,
+      "loss": 1.1653,
+      "step": 6929
+    },
+    {
+      "epoch": 2.580074693720841,
+      "grad_norm": 0.20714597404003143,
+      "learning_rate": 1.7513554778136133e-05,
+      "loss": 1.1778,
+      "step": 6930
+    },
+    {
+      "epoch": 2.58044699887145,
+      "grad_norm": 0.1802298128604889,
+      "learning_rate": 1.751275273824123e-05,
+      "loss": 1.1798,
+      "step": 6931
+    },
+    {
+      "epoch": 2.580819304022059,
+      "grad_norm": 0.18000389635562897,
+      "learning_rate": 1.7511950587383438e-05,
+      "loss": 1.1729,
+      "step": 6932
+    },
+    {
+      "epoch": 2.5811916091726683,
+      "grad_norm": 0.1696796715259552,
+      "learning_rate": 1.7511148325574613e-05,
+      "loss": 1.176,
+      "step": 6933
+    },
+    {
+      "epoch": 2.581563914323277,
+      "grad_norm": 0.1616394817829132,
+      "learning_rate": 1.7510345952826594e-05,
+      "loss": 1.1583,
+      "step": 6934
+    },
+    {
+      "epoch": 2.5819362194738864,
+      "grad_norm": 0.18179087340831757,
+      "learning_rate": 1.7509543469151234e-05,
+      "loss": 1.1809,
+      "step": 6935
+    },
+    {
+      "epoch": 2.5823085246244952,
+      "grad_norm": 0.17522992193698883,
+      "learning_rate": 1.7508740874560393e-05,
+      "loss": 1.1693,
+      "step": 6936
+    },
+    {
+      "epoch": 2.5826808297751045,
+      "grad_norm": 0.15960480272769928,
+      "learning_rate": 1.7507938169065922e-05,
+      "loss": 1.17,
+      "step": 6937
+    },
+    {
+      "epoch": 2.5830531349257138,
+      "grad_norm": 0.16696304082870483,
+      "learning_rate": 1.750713535267967e-05,
+      "loss": 1.1632,
+      "step": 6938
+    },
+    {
+      "epoch": 2.5834254400763226,
+      "grad_norm": 0.16897381842136383,
+      "learning_rate": 1.7506332425413505e-05,
+      "loss": 1.1773,
+      "step": 6939
+    },
+    {
+      "epoch": 2.5837977452269314,
+      "grad_norm": 0.16685079038143158,
+      "learning_rate": 1.750552938727928e-05,
+      "loss": 1.1664,
+      "step": 6940
+    },
+    {
+      "epoch": 2.5841700503775407,
+      "grad_norm": 0.1602102816104889,
+      "learning_rate": 1.7504726238288857e-05,
+      "loss": 1.174,
+      "step": 6941
+    },
+    {
+      "epoch": 2.58454235552815,
+      "grad_norm": 0.21326759457588196,
+      "learning_rate": 1.7503922978454094e-05,
+      "loss": 1.179,
+      "step": 6942
+    },
+    {
+      "epoch": 2.5849146606787587,
+      "grad_norm": 0.16109180450439453,
+      "learning_rate": 1.7503119607786865e-05,
+      "loss": 1.1676,
+      "step": 6943
+    },
+    {
+      "epoch": 2.585286965829368,
+      "grad_norm": 0.1658046394586563,
+      "learning_rate": 1.7502316126299027e-05,
+      "loss": 1.1825,
+      "step": 6944
+    },
+    {
+      "epoch": 2.585659270979977,
+      "grad_norm": 0.16446712613105774,
+      "learning_rate": 1.7501512534002453e-05,
+      "loss": 1.1782,
+      "step": 6945
+    },
+    {
+      "epoch": 2.586031576130586,
+      "grad_norm": 0.1652338057756424,
+      "learning_rate": 1.7500708830909006e-05,
+      "loss": 1.1615,
+      "step": 6946
+    },
+    {
+      "epoch": 2.5864038812811954,
+      "grad_norm": 0.16343404352664948,
+      "learning_rate": 1.7499905017030565e-05,
+      "loss": 1.1683,
+      "step": 6947
+    },
+    {
+      "epoch": 2.586776186431804,
+      "grad_norm": 0.16519500315189362,
+      "learning_rate": 1.7499101092378995e-05,
+      "loss": 1.175,
+      "step": 6948
+    },
+    {
+      "epoch": 2.587148491582413,
+      "grad_norm": 0.16351045668125153,
+      "learning_rate": 1.7498297056966174e-05,
+      "loss": 1.1709,
+      "step": 6949
+    },
+    {
+      "epoch": 2.5875207967330223,
+      "grad_norm": 0.1664671003818512,
+      "learning_rate": 1.7497492910803972e-05,
+      "loss": 1.169,
+      "step": 6950
+    },
+    {
+      "epoch": 2.5878931018836315,
+      "grad_norm": 0.16572503745555878,
+      "learning_rate": 1.7496688653904277e-05,
+      "loss": 1.1797,
+      "step": 6951
+    },
+    {
+      "epoch": 2.5882654070342404,
+      "grad_norm": 0.16367551684379578,
+      "learning_rate": 1.7495884286278955e-05,
+      "loss": 1.1695,
+      "step": 6952
+    },
+    {
+      "epoch": 2.5886377121848496,
+      "grad_norm": 0.16530726850032806,
+      "learning_rate": 1.7495079807939897e-05,
+      "loss": 1.1604,
+      "step": 6953
+    },
+    {
+      "epoch": 2.5890100173354584,
+      "grad_norm": 0.16979095339775085,
+      "learning_rate": 1.7494275218898976e-05,
+      "loss": 1.1693,
+      "step": 6954
+    },
+    {
+      "epoch": 2.5893823224860677,
+      "grad_norm": 0.1681201159954071,
+      "learning_rate": 1.749347051916808e-05,
+      "loss": 1.1655,
+      "step": 6955
+    },
+    {
+      "epoch": 2.589754627636677,
+      "grad_norm": 0.17620940506458282,
+      "learning_rate": 1.74926657087591e-05,
+      "loss": 1.1716,
+      "step": 6956
+    },
+    {
+      "epoch": 2.590126932787286,
+      "grad_norm": 0.16808199882507324,
+      "learning_rate": 1.7491860787683915e-05,
+      "loss": 1.1593,
+      "step": 6957
+    },
+    {
+      "epoch": 2.5904992379378946,
+      "grad_norm": 0.16996102035045624,
+      "learning_rate": 1.7491055755954418e-05,
+      "loss": 1.179,
+      "step": 6958
+    },
+    {
+      "epoch": 2.590871543088504,
+      "grad_norm": 0.16559867560863495,
+      "learning_rate": 1.7490250613582492e-05,
+      "loss": 1.1794,
+      "step": 6959
+    },
+    {
+      "epoch": 2.591243848239113,
+      "grad_norm": 0.16791416704654694,
+      "learning_rate": 1.748944536058004e-05,
+      "loss": 1.1791,
+      "step": 6960
+    },
+    {
+      "epoch": 2.591616153389722,
+      "grad_norm": 0.17107127606868744,
+      "learning_rate": 1.7488639996958952e-05,
+      "loss": 1.1862,
+      "step": 6961
+    },
+    {
+      "epoch": 2.591988458540331,
+      "grad_norm": 0.16655850410461426,
+      "learning_rate": 1.7487834522731115e-05,
+      "loss": 1.1807,
+      "step": 6962
+    },
+    {
+      "epoch": 2.59236076369094,
+      "grad_norm": 0.171476349234581,
+      "learning_rate": 1.7487028937908436e-05,
+      "loss": 1.1816,
+      "step": 6963
+    },
+    {
+      "epoch": 2.5927330688415493,
+      "grad_norm": 0.1685655415058136,
+      "learning_rate": 1.748622324250281e-05,
+      "loss": 1.1735,
+      "step": 6964
+    },
+    {
+      "epoch": 2.5931053739921586,
+      "grad_norm": 0.17578360438346863,
+      "learning_rate": 1.7485417436526134e-05,
+      "loss": 1.1802,
+      "step": 6965
+    },
+    {
+      "epoch": 2.5934776791427674,
+      "grad_norm": 0.17003393173217773,
+      "learning_rate": 1.748461151999031e-05,
+      "loss": 1.1937,
+      "step": 6966
+    },
+    {
+      "epoch": 2.593849984293376,
+      "grad_norm": 0.18038241565227509,
+      "learning_rate": 1.7483805492907246e-05,
+      "loss": 1.1826,
+      "step": 6967
+    },
+    {
+      "epoch": 2.5942222894439855,
+      "grad_norm": 0.23492062091827393,
+      "learning_rate": 1.7482999355288846e-05,
+      "loss": 1.1888,
+      "step": 6968
+    },
+    {
+      "epoch": 2.5945945945945947,
+      "grad_norm": 0.18728743493556976,
+      "learning_rate": 1.7482193107147012e-05,
+      "loss": 1.1753,
+      "step": 6969
+    },
+    {
+      "epoch": 2.5949668997452036,
+      "grad_norm": 0.1779962182044983,
+      "learning_rate": 1.7481386748493664e-05,
+      "loss": 1.1748,
+      "step": 6970
+    },
+    {
+      "epoch": 2.595339204895813,
+      "grad_norm": 0.1643301546573639,
+      "learning_rate": 1.7480580279340694e-05,
+      "loss": 1.17,
+      "step": 6971
+    },
+    {
+      "epoch": 2.5957115100464216,
+      "grad_norm": 0.25796541571617126,
+      "learning_rate": 1.7479773699700024e-05,
+      "loss": 1.1708,
+      "step": 6972
+    },
+    {
+      "epoch": 2.596083815197031,
+      "grad_norm": 0.1716795414686203,
+      "learning_rate": 1.747896700958357e-05,
+      "loss": 1.1742,
+      "step": 6973
+    },
+    {
+      "epoch": 2.59645612034764,
+      "grad_norm": 0.17768600583076477,
+      "learning_rate": 1.747816020900324e-05,
+      "loss": 1.1841,
+      "step": 6974
+    },
+    {
+      "epoch": 2.596828425498249,
+      "grad_norm": 0.164155513048172,
+      "learning_rate": 1.7477353297970952e-05,
+      "loss": 1.1678,
+      "step": 6975
+    },
+    {
+      "epoch": 2.597200730648858,
+      "grad_norm": 0.16183872520923615,
+      "learning_rate": 1.7476546276498625e-05,
+      "loss": 1.1626,
+      "step": 6976
+    },
+    {
+      "epoch": 2.597573035799467,
+      "grad_norm": 0.17142947018146515,
+      "learning_rate": 1.7475739144598183e-05,
+      "loss": 1.1773,
+      "step": 6977
+    },
+    {
+      "epoch": 2.5979453409500763,
+      "grad_norm": 0.16305896639823914,
+      "learning_rate": 1.7474931902281538e-05,
+      "loss": 1.182,
+      "step": 6978
+    },
+    {
+      "epoch": 2.598317646100685,
+      "grad_norm": 0.17089135944843292,
+      "learning_rate": 1.747412454956062e-05,
+      "loss": 1.1743,
+      "step": 6979
+    },
+    {
+      "epoch": 2.5986899512512944,
+      "grad_norm": 0.1611926555633545,
+      "learning_rate": 1.747331708644735e-05,
+      "loss": 1.1753,
+      "step": 6980
+    },
+    {
+      "epoch": 2.5990622564019032,
+      "grad_norm": 0.16074836254119873,
+      "learning_rate": 1.747250951295366e-05,
+      "loss": 1.1583,
+      "step": 6981
+    },
+    {
+      "epoch": 2.5994345615525125,
+      "grad_norm": 0.16934338212013245,
+      "learning_rate": 1.7471701829091468e-05,
+      "loss": 1.1796,
+      "step": 6982
+    },
+    {
+      "epoch": 2.5998068667031218,
+      "grad_norm": 0.16507279872894287,
+      "learning_rate": 1.747089403487271e-05,
+      "loss": 1.1742,
+      "step": 6983
+    },
+    {
+      "epoch": 2.6001791718537306,
+      "grad_norm": 0.16199171543121338,
+      "learning_rate": 1.747008613030932e-05,
+      "loss": 1.1715,
+      "step": 6984
+    },
+    {
+      "epoch": 2.6005514770043394,
+      "grad_norm": 0.15591688454151154,
+      "learning_rate": 1.7469278115413222e-05,
+      "loss": 1.171,
+      "step": 6985
+    },
+    {
+      "epoch": 2.6009237821549487,
+      "grad_norm": 0.16649380326271057,
+      "learning_rate": 1.7468469990196358e-05,
+      "loss": 1.1597,
+      "step": 6986
+    },
+    {
+      "epoch": 2.601296087305558,
+      "grad_norm": 0.1675834059715271,
+      "learning_rate": 1.746766175467066e-05,
+      "loss": 1.183,
+      "step": 6987
+    },
+    {
+      "epoch": 2.6016683924561668,
+      "grad_norm": 0.15882286429405212,
+      "learning_rate": 1.7466853408848067e-05,
+      "loss": 1.1885,
+      "step": 6988
+    },
+    {
+      "epoch": 2.602040697606776,
+      "grad_norm": 0.16068002581596375,
+      "learning_rate": 1.7466044952740517e-05,
+      "loss": 1.1896,
+      "step": 6989
+    },
+    {
+      "epoch": 2.602413002757385,
+      "grad_norm": 0.16341431438922882,
+      "learning_rate": 1.7465236386359952e-05,
+      "loss": 1.1716,
+      "step": 6990
+    },
+    {
+      "epoch": 2.602785307907994,
+      "grad_norm": 0.1623801589012146,
+      "learning_rate": 1.746442770971831e-05,
+      "loss": 1.1738,
+      "step": 6991
+    },
+    {
+      "epoch": 2.6031576130586034,
+      "grad_norm": 0.16583126783370972,
+      "learning_rate": 1.7463618922827545e-05,
+      "loss": 1.1788,
+      "step": 6992
+    },
+    {
+      "epoch": 2.603529918209212,
+      "grad_norm": 0.16922509670257568,
+      "learning_rate": 1.74628100256996e-05,
+      "loss": 1.1855,
+      "step": 6993
+    },
+    {
+      "epoch": 2.603902223359821,
+      "grad_norm": 0.16294334828853607,
+      "learning_rate": 1.7462001018346408e-05,
+      "loss": 1.178,
+      "step": 6994
+    },
+    {
+      "epoch": 2.6042745285104303,
+      "grad_norm": 0.16397880017757416,
+      "learning_rate": 1.7461191900779936e-05,
+      "loss": 1.1779,
+      "step": 6995
+    },
+    {
+      "epoch": 2.6046468336610396,
+      "grad_norm": 0.16646462678909302,
+      "learning_rate": 1.746038267301213e-05,
+      "loss": 1.1666,
+      "step": 6996
+    },
+    {
+      "epoch": 2.6050191388116484,
+      "grad_norm": 0.16266998648643494,
+      "learning_rate": 1.7459573335054935e-05,
+      "loss": 1.1792,
+      "step": 6997
+    },
+    {
+      "epoch": 2.6053914439622576,
+      "grad_norm": 0.16222761571407318,
+      "learning_rate": 1.745876388692031e-05,
+      "loss": 1.169,
+      "step": 6998
+    },
+    {
+      "epoch": 2.6057637491128665,
+      "grad_norm": 0.16887149214744568,
+      "learning_rate": 1.7457954328620217e-05,
+      "loss": 1.1724,
+      "step": 6999
+    },
+    {
+      "epoch": 2.6061360542634757,
+      "grad_norm": 0.16719955205917358,
+      "learning_rate": 1.74571446601666e-05,
+      "loss": 1.1934,
+      "step": 7000
+    },
+    {
+      "epoch": 2.6061360542634757,
+      "eval_loss": 1.2935380935668945,
+      "eval_runtime": 16.7437,
+      "eval_samples_per_second": 103.562,
+      "eval_steps_per_second": 5.196,
+      "step": 7000
+    },
+    {
+      "epoch": 2.606508359414085,
+      "grad_norm": 0.16188542544841766,
+      "learning_rate": 1.7456334881571428e-05,
+      "loss": 1.1968,
+      "step": 7001
+    },
+    {
+      "epoch": 2.606880664564694,
+      "grad_norm": 0.16819295287132263,
+      "learning_rate": 1.745552499284666e-05,
+      "loss": 1.1796,
+      "step": 7002
+    },
+    {
+      "epoch": 2.607252969715303,
+      "grad_norm": 0.1619422435760498,
+      "learning_rate": 1.745471499400425e-05,
+      "loss": 1.1739,
+      "step": 7003
+    },
+    {
+      "epoch": 2.607625274865912,
+      "grad_norm": 0.16731281578540802,
+      "learning_rate": 1.745390488505617e-05,
+      "loss": 1.1763,
+      "step": 7004
+    },
+    {
+      "epoch": 2.607997580016521,
+      "grad_norm": 0.16323542594909668,
+      "learning_rate": 1.745309466601438e-05,
+      "loss": 1.1742,
+      "step": 7005
+    },
+    {
+      "epoch": 2.60836988516713,
+      "grad_norm": 0.16577135026454926,
+      "learning_rate": 1.7452284336890853e-05,
+      "loss": 1.1684,
+      "step": 7006
+    },
+    {
+      "epoch": 2.6087421903177392,
+      "grad_norm": 0.16486932337284088,
+      "learning_rate": 1.7451473897697552e-05,
+      "loss": 1.168,
+      "step": 7007
+    },
+    {
+      "epoch": 2.609114495468348,
+      "grad_norm": 0.17325793206691742,
+      "learning_rate": 1.745066334844645e-05,
+      "loss": 1.1709,
+      "step": 7008
+    },
+    {
+      "epoch": 2.6094868006189573,
+      "grad_norm": 0.1616426408290863,
+      "learning_rate": 1.744985268914952e-05,
+      "loss": 1.1741,
+      "step": 7009
+    },
+    {
+      "epoch": 2.6098591057695666,
+      "grad_norm": 0.16339850425720215,
+      "learning_rate": 1.744904191981873e-05,
+      "loss": 1.1719,
+      "step": 7010
+    },
+    {
+      "epoch": 2.6102314109201754,
+      "grad_norm": 0.16253487765789032,
+      "learning_rate": 1.744823104046606e-05,
+      "loss": 1.1705,
+      "step": 7011
+    },
+    {
+      "epoch": 2.6106037160707847,
+      "grad_norm": 0.17588497698307037,
+      "learning_rate": 1.7447420051103483e-05,
+      "loss": 1.178,
+      "step": 7012
+    },
+    {
+      "epoch": 2.6109760212213935,
+      "grad_norm": 0.18001097440719604,
+      "learning_rate": 1.744660895174298e-05,
+      "loss": 1.1642,
+      "step": 7013
+    },
+    {
+      "epoch": 2.6113483263720028,
+      "grad_norm": 0.1719658225774765,
+      "learning_rate": 1.7445797742396535e-05,
+      "loss": 1.1675,
+      "step": 7014
+    },
+    {
+      "epoch": 2.6117206315226116,
+      "grad_norm": 0.20831653475761414,
+      "learning_rate": 1.7444986423076116e-05,
+      "loss": 1.168,
+      "step": 7015
+    },
+    {
+      "epoch": 2.612092936673221,
+      "grad_norm": 0.17336048185825348,
+      "learning_rate": 1.744417499379372e-05,
+      "loss": 1.1814,
+      "step": 7016
+    },
+    {
+      "epoch": 2.6124652418238297,
+      "grad_norm": 0.17226654291152954,
+      "learning_rate": 1.7443363454561327e-05,
+      "loss": 1.1761,
+      "step": 7017
+    },
+    {
+      "epoch": 2.612837546974439,
+      "grad_norm": 0.20160022377967834,
+      "learning_rate": 1.744255180539092e-05,
+      "loss": 1.1843,
+      "step": 7018
+    },
+    {
+      "epoch": 2.613209852125048,
+      "grad_norm": 0.1713937222957611,
+      "learning_rate": 1.7441740046294496e-05,
+      "loss": 1.1625,
+      "step": 7019
+    },
+    {
+      "epoch": 2.613582157275657,
+      "grad_norm": 0.1691260188817978,
+      "learning_rate": 1.744092817728403e-05,
+      "loss": 1.1839,
+      "step": 7020
+    },
+    {
+      "epoch": 2.6139544624262663,
+      "grad_norm": 0.16929064691066742,
+      "learning_rate": 1.7440116198371528e-05,
+      "loss": 1.1644,
+      "step": 7021
+    },
+    {
+      "epoch": 2.614326767576875,
+      "grad_norm": 0.16705340147018433,
+      "learning_rate": 1.7439304109568972e-05,
+      "loss": 1.1847,
+      "step": 7022
+    },
+    {
+      "epoch": 2.6146990727274844,
+      "grad_norm": 0.16961196064949036,
+      "learning_rate": 1.7438491910888367e-05,
+      "loss": 1.1534,
+      "step": 7023
+    },
+    {
+      "epoch": 2.615071377878093,
+      "grad_norm": 0.16219016909599304,
+      "learning_rate": 1.74376796023417e-05,
+      "loss": 1.1695,
+      "step": 7024
+    },
+    {
+      "epoch": 2.6154436830287024,
+      "grad_norm": 0.19066371023654938,
+      "learning_rate": 1.7436867183940972e-05,
+      "loss": 1.1739,
+      "step": 7025
+    },
+    {
+      "epoch": 2.6158159881793113,
+      "grad_norm": 0.16923125088214874,
+      "learning_rate": 1.7436054655698184e-05,
+      "loss": 1.1544,
+      "step": 7026
+    },
+    {
+      "epoch": 2.6161882933299205,
+      "grad_norm": 0.1667940318584442,
+      "learning_rate": 1.7435242017625333e-05,
+      "loss": 1.167,
+      "step": 7027
+    },
+    {
+      "epoch": 2.61656059848053,
+      "grad_norm": 0.18602821230888367,
+      "learning_rate": 1.7434429269734426e-05,
+      "loss": 1.1667,
+      "step": 7028
+    },
+    {
+      "epoch": 2.6169329036311386,
+      "grad_norm": 0.17603205144405365,
+      "learning_rate": 1.7433616412037462e-05,
+      "loss": 1.1759,
+      "step": 7029
+    },
+    {
+      "epoch": 2.617305208781748,
+      "grad_norm": 0.15942661464214325,
+      "learning_rate": 1.7432803444546454e-05,
+      "loss": 1.1731,
+      "step": 7030
+    },
+    {
+      "epoch": 2.6176775139323567,
+      "grad_norm": 0.19452649354934692,
+      "learning_rate": 1.7431990367273402e-05,
+      "loss": 1.1756,
+      "step": 7031
+    },
+    {
+      "epoch": 2.618049819082966,
+      "grad_norm": 0.1983923316001892,
+      "learning_rate": 1.7431177180230323e-05,
+      "loss": 1.1745,
+      "step": 7032
+    },
+    {
+      "epoch": 2.618422124233575,
+      "grad_norm": 0.16836190223693848,
+      "learning_rate": 1.7430363883429218e-05,
+      "loss": 1.1774,
+      "step": 7033
+    },
+    {
+      "epoch": 2.618794429384184,
+      "grad_norm": 0.29323258996009827,
+      "learning_rate": 1.742955047688211e-05,
+      "loss": 1.178,
+      "step": 7034
+    },
+    {
+      "epoch": 2.619166734534793,
+      "grad_norm": 0.19639234244823456,
+      "learning_rate": 1.7428736960601004e-05,
+      "loss": 1.1731,
+      "step": 7035
+    },
+    {
+      "epoch": 2.619539039685402,
+      "grad_norm": 0.1906968504190445,
+      "learning_rate": 1.7427923334597922e-05,
+      "loss": 1.1848,
+      "step": 7036
+    },
+    {
+      "epoch": 2.6199113448360114,
+      "grad_norm": 0.16878125071525574,
+      "learning_rate": 1.7427109598884877e-05,
+      "loss": 1.1764,
+      "step": 7037
+    },
+    {
+      "epoch": 2.6202836499866202,
+      "grad_norm": 0.16486749053001404,
+      "learning_rate": 1.742629575347389e-05,
+      "loss": 1.1691,
+      "step": 7038
+    },
+    {
+      "epoch": 2.6206559551372295,
+      "grad_norm": 0.17094659805297852,
+      "learning_rate": 1.742548179837698e-05,
+      "loss": 1.174,
+      "step": 7039
+    },
+    {
+      "epoch": 2.6210282602878383,
+      "grad_norm": 0.18062889575958252,
+      "learning_rate": 1.742466773360617e-05,
+      "loss": 1.1766,
+      "step": 7040
+    },
+    {
+      "epoch": 2.6214005654384476,
+      "grad_norm": 0.16564013063907623,
+      "learning_rate": 1.742385355917348e-05,
+      "loss": 1.1826,
+      "step": 7041
+    },
+    {
+      "epoch": 2.6217728705890564,
+      "grad_norm": 0.15717479586601257,
+      "learning_rate": 1.7423039275090947e-05,
+      "loss": 1.1718,
+      "step": 7042
+    },
+    {
+      "epoch": 2.6221451757396657,
+      "grad_norm": 0.16660241782665253,
+      "learning_rate": 1.7422224881370585e-05,
+      "loss": 1.1845,
+      "step": 7043
+    },
+    {
+      "epoch": 2.6225174808902745,
+      "grad_norm": 0.16383567452430725,
+      "learning_rate": 1.7421410378024428e-05,
+      "loss": 1.1843,
+      "step": 7044
+    },
+    {
+      "epoch": 2.6228897860408837,
+      "grad_norm": 0.1657799780368805,
+      "learning_rate": 1.7420595765064505e-05,
+      "loss": 1.1704,
+      "step": 7045
+    },
+    {
+      "epoch": 2.623262091191493,
+      "grad_norm": 0.16230309009552002,
+      "learning_rate": 1.7419781042502846e-05,
+      "loss": 1.1725,
+      "step": 7046
+    },
+    {
+      "epoch": 2.623634396342102,
+      "grad_norm": 0.16071511805057526,
+      "learning_rate": 1.7418966210351492e-05,
+      "loss": 1.1682,
+      "step": 7047
+    },
+    {
+      "epoch": 2.624006701492711,
+      "grad_norm": 0.18019315600395203,
+      "learning_rate": 1.741815126862247e-05,
+      "loss": 1.1601,
+      "step": 7048
+    },
+    {
+      "epoch": 2.62437900664332,
+      "grad_norm": 0.16016197204589844,
+      "learning_rate": 1.741733621732782e-05,
+      "loss": 1.1783,
+      "step": 7049
+    },
+    {
+      "epoch": 2.624751311793929,
+      "grad_norm": 0.15661844611167908,
+      "learning_rate": 1.7416521056479577e-05,
+      "loss": 1.1509,
+      "step": 7050
+    },
+    {
+      "epoch": 2.6251236169445384,
+      "grad_norm": 0.16214154660701752,
+      "learning_rate": 1.7415705786089784e-05,
+      "loss": 1.162,
+      "step": 7051
+    },
+    {
+      "epoch": 2.6254959220951473,
+      "grad_norm": 0.16358621418476105,
+      "learning_rate": 1.7414890406170487e-05,
+      "loss": 1.1727,
+      "step": 7052
+    },
+    {
+      "epoch": 2.625868227245756,
+      "grad_norm": 0.1643485575914383,
+      "learning_rate": 1.7414074916733715e-05,
+      "loss": 1.168,
+      "step": 7053
+    },
+    {
+      "epoch": 2.6262405323963653,
+      "grad_norm": 0.16986949741840363,
+      "learning_rate": 1.7413259317791528e-05,
+      "loss": 1.1664,
+      "step": 7054
+    },
+    {
+      "epoch": 2.6266128375469746,
+      "grad_norm": 0.17110756039619446,
+      "learning_rate": 1.7412443609355967e-05,
+      "loss": 1.163,
+      "step": 7055
+    },
+    {
+      "epoch": 2.6269851426975834,
+      "grad_norm": 0.16550353169441223,
+      "learning_rate": 1.7411627791439073e-05,
+      "loss": 1.1659,
+      "step": 7056
+    },
+    {
+      "epoch": 2.6273574478481927,
+      "grad_norm": 0.16365563869476318,
+      "learning_rate": 1.7410811864052908e-05,
+      "loss": 1.1871,
+      "step": 7057
+    },
+    {
+      "epoch": 2.6277297529988015,
+      "grad_norm": 0.17241568863391876,
+      "learning_rate": 1.7409995827209517e-05,
+      "loss": 1.1757,
+      "step": 7058
+    },
+    {
+      "epoch": 2.628102058149411,
+      "grad_norm": 0.16813285648822784,
+      "learning_rate": 1.7409179680920945e-05,
+      "loss": 1.1615,
+      "step": 7059
+    },
+    {
+      "epoch": 2.62847436330002,
+      "grad_norm": 0.15777957439422607,
+      "learning_rate": 1.740836342519926e-05,
+      "loss": 1.167,
+      "step": 7060
+    },
+    {
+      "epoch": 2.628846668450629,
+      "grad_norm": 0.1858019232749939,
+      "learning_rate": 1.7407547060056514e-05,
+      "loss": 1.1673,
+      "step": 7061
+    },
+    {
+      "epoch": 2.6292189736012377,
+      "grad_norm": 0.15979260206222534,
+      "learning_rate": 1.740673058550476e-05,
+      "loss": 1.178,
+      "step": 7062
+    },
+    {
+      "epoch": 2.629591278751847,
+      "grad_norm": 0.1698780208826065,
+      "learning_rate": 1.7405914001556058e-05,
+      "loss": 1.1721,
+      "step": 7063
+    },
+    {
+      "epoch": 2.629963583902456,
+      "grad_norm": 0.16836461424827576,
+      "learning_rate": 1.7405097308222474e-05,
+      "loss": 1.1717,
+      "step": 7064
+    },
+    {
+      "epoch": 2.630335889053065,
+      "grad_norm": 0.16282488405704498,
+      "learning_rate": 1.740428050551607e-05,
+      "loss": 1.1735,
+      "step": 7065
+    },
+    {
+      "epoch": 2.6307081942036743,
+      "grad_norm": 0.16781288385391235,
+      "learning_rate": 1.74034635934489e-05,
+      "loss": 1.2007,
+      "step": 7066
+    },
+    {
+      "epoch": 2.631080499354283,
+      "grad_norm": 0.16193710267543793,
+      "learning_rate": 1.7402646572033043e-05,
+      "loss": 1.1854,
+      "step": 7067
+    },
+    {
+      "epoch": 2.6314528045048924,
+      "grad_norm": 0.16660821437835693,
+      "learning_rate": 1.7401829441280563e-05,
+      "loss": 1.1706,
+      "step": 7068
+    },
+    {
+      "epoch": 2.6318251096555016,
+      "grad_norm": 0.16784422099590302,
+      "learning_rate": 1.740101220120352e-05,
+      "loss": 1.1672,
+      "step": 7069
+    },
+    {
+      "epoch": 2.6321974148061105,
+      "grad_norm": 0.17018967866897583,
+      "learning_rate": 1.7400194851813994e-05,
+      "loss": 1.1758,
+      "step": 7070
+    },
+    {
+      "epoch": 2.6325697199567193,
+      "grad_norm": 0.16101203858852386,
+      "learning_rate": 1.7399377393124056e-05,
+      "loss": 1.185,
+      "step": 7071
+    },
+    {
+      "epoch": 2.6329420251073286,
+      "grad_norm": 0.16601742804050446,
+      "learning_rate": 1.7398559825145776e-05,
+      "loss": 1.17,
+      "step": 7072
+    },
+    {
+      "epoch": 2.633314330257938,
+      "grad_norm": 0.16237667202949524,
+      "learning_rate": 1.7397742147891234e-05,
+      "loss": 1.1807,
+      "step": 7073
+    },
+    {
+      "epoch": 2.6336866354085466,
+      "grad_norm": 0.19661523401737213,
+      "learning_rate": 1.7396924361372504e-05,
+      "loss": 1.1696,
+      "step": 7074
+    },
+    {
+      "epoch": 2.634058940559156,
+      "grad_norm": 0.18306462466716766,
+      "learning_rate": 1.7396106465601662e-05,
+      "loss": 1.1693,
+      "step": 7075
+    },
+    {
+      "epoch": 2.6344312457097647,
+      "grad_norm": 0.16825948655605316,
+      "learning_rate": 1.7395288460590797e-05,
+      "loss": 1.1752,
+      "step": 7076
+    },
+    {
+      "epoch": 2.634803550860374,
+      "grad_norm": 0.18328233063220978,
+      "learning_rate": 1.739447034635198e-05,
+      "loss": 1.1778,
+      "step": 7077
+    },
+    {
+      "epoch": 2.6351758560109833,
+      "grad_norm": 0.1660008579492569,
+      "learning_rate": 1.7393652122897306e-05,
+      "loss": 1.1784,
+      "step": 7078
+    },
+    {
+      "epoch": 2.635548161161592,
+      "grad_norm": 0.1794942319393158,
+      "learning_rate": 1.7392833790238854e-05,
+      "loss": 1.1743,
+      "step": 7079
+    },
+    {
+      "epoch": 2.635920466312201,
+      "grad_norm": 0.16286969184875488,
+      "learning_rate": 1.7392015348388707e-05,
+      "loss": 1.1833,
+      "step": 7080
+    },
+    {
+      "epoch": 2.63629277146281,
+      "grad_norm": 0.18671861290931702,
+      "learning_rate": 1.7391196797358957e-05,
+      "loss": 1.1811,
+      "step": 7081
+    },
+    {
+      "epoch": 2.6366650766134194,
+      "grad_norm": 0.16873905062675476,
+      "learning_rate": 1.7390378137161694e-05,
+      "loss": 1.1764,
+      "step": 7082
+    },
+    {
+      "epoch": 2.6370373817640282,
+      "grad_norm": 0.16419054567813873,
+      "learning_rate": 1.7389559367809012e-05,
+      "loss": 1.1631,
+      "step": 7083
+    },
+    {
+      "epoch": 2.6374096869146375,
+      "grad_norm": 0.2097456008195877,
+      "learning_rate": 1.7388740489313e-05,
+      "loss": 1.1651,
+      "step": 7084
+    },
+    {
+      "epoch": 2.6377819920652463,
+      "grad_norm": 0.19423960149288177,
+      "learning_rate": 1.7387921501685757e-05,
+      "loss": 1.1651,
+      "step": 7085
+    },
+    {
+      "epoch": 2.6381542972158556,
+      "grad_norm": 0.17196422815322876,
+      "learning_rate": 1.7387102404939375e-05,
+      "loss": 1.1788,
+      "step": 7086
+    },
+    {
+      "epoch": 2.638526602366465,
+      "grad_norm": 0.2757280170917511,
+      "learning_rate": 1.7386283199085957e-05,
+      "loss": 1.182,
+      "step": 7087
+    },
+    {
+      "epoch": 2.6388989075170737,
+      "grad_norm": 0.16954492032527924,
+      "learning_rate": 1.73854638841376e-05,
+      "loss": 1.1732,
+      "step": 7088
+    },
+    {
+      "epoch": 2.6392712126676825,
+      "grad_norm": 0.1665467768907547,
+      "learning_rate": 1.7384644460106403e-05,
+      "loss": 1.1731,
+      "step": 7089
+    },
+    {
+      "epoch": 2.6396435178182918,
+      "grad_norm": 0.16451166570186615,
+      "learning_rate": 1.738382492700447e-05,
+      "loss": 1.1616,
+      "step": 7090
+    },
+    {
+      "epoch": 2.640015822968901,
+      "grad_norm": 0.16248999536037445,
+      "learning_rate": 1.7383005284843902e-05,
+      "loss": 1.1636,
+      "step": 7091
+    },
+    {
+      "epoch": 2.64038812811951,
+      "grad_norm": 0.1698039472103119,
+      "learning_rate": 1.7382185533636815e-05,
+      "loss": 1.1749,
+      "step": 7092
+    },
+    {
+      "epoch": 2.640760433270119,
+      "grad_norm": 0.17128440737724304,
+      "learning_rate": 1.738136567339531e-05,
+      "loss": 1.1702,
+      "step": 7093
+    },
+    {
+      "epoch": 2.641132738420728,
+      "grad_norm": 0.16965392231941223,
+      "learning_rate": 1.7380545704131496e-05,
+      "loss": 1.1698,
+      "step": 7094
+    },
+    {
+      "epoch": 2.641505043571337,
+      "grad_norm": 0.17016279697418213,
+      "learning_rate": 1.737972562585749e-05,
+      "loss": 1.1715,
+      "step": 7095
+    },
+    {
+      "epoch": 2.6418773487219465,
+      "grad_norm": 0.16740964353084564,
+      "learning_rate": 1.7378905438585394e-05,
+      "loss": 1.1743,
+      "step": 7096
+    },
+    {
+      "epoch": 2.6422496538725553,
+      "grad_norm": 0.16494162380695343,
+      "learning_rate": 1.737808514232733e-05,
+      "loss": 1.1626,
+      "step": 7097
+    },
+    {
+      "epoch": 2.642621959023164,
+      "grad_norm": 0.17924334108829498,
+      "learning_rate": 1.7377264737095408e-05,
+      "loss": 1.1731,
+      "step": 7098
+    },
+    {
+      "epoch": 2.6429942641737734,
+      "grad_norm": 0.16398359835147858,
+      "learning_rate": 1.7376444222901754e-05,
+      "loss": 1.1804,
+      "step": 7099
+    },
+    {
+      "epoch": 2.6433665693243826,
+      "grad_norm": 0.16579513251781464,
+      "learning_rate": 1.737562359975848e-05,
+      "loss": 1.1583,
+      "step": 7100
+    },
+    {
+      "epoch": 2.6437388744749915,
+      "grad_norm": 0.16405485570430756,
+      "learning_rate": 1.7374802867677706e-05,
+      "loss": 1.1651,
+      "step": 7101
+    },
+    {
+      "epoch": 2.6441111796256007,
+      "grad_norm": 0.16502594947814941,
+      "learning_rate": 1.7373982026671557e-05,
+      "loss": 1.1766,
+      "step": 7102
+    },
+    {
+      "epoch": 2.6444834847762095,
+      "grad_norm": 0.16717404127120972,
+      "learning_rate": 1.737316107675216e-05,
+      "loss": 1.1841,
+      "step": 7103
+    },
+    {
+      "epoch": 2.644855789926819,
+      "grad_norm": 0.16163359582424164,
+      "learning_rate": 1.7372340017931636e-05,
+      "loss": 1.1678,
+      "step": 7104
+    },
+    {
+      "epoch": 2.645228095077428,
+      "grad_norm": 0.17032460868358612,
+      "learning_rate": 1.737151885022211e-05,
+      "loss": 1.1817,
+      "step": 7105
+    },
+    {
+      "epoch": 2.645600400228037,
+      "grad_norm": 0.16142508387565613,
+      "learning_rate": 1.7370697573635714e-05,
+      "loss": 1.1869,
+      "step": 7106
+    },
+    {
+      "epoch": 2.6459727053786457,
+      "grad_norm": 0.15926913917064667,
+      "learning_rate": 1.7369876188184577e-05,
+      "loss": 1.1717,
+      "step": 7107
+    },
+    {
+      "epoch": 2.646345010529255,
+      "grad_norm": 0.1710928976535797,
+      "learning_rate": 1.7369054693880832e-05,
+      "loss": 1.1809,
+      "step": 7108
+    },
+    {
+      "epoch": 2.6467173156798642,
+      "grad_norm": 0.1716298758983612,
+      "learning_rate": 1.7368233090736613e-05,
+      "loss": 1.1709,
+      "step": 7109
+    },
+    {
+      "epoch": 2.647089620830473,
+      "grad_norm": 0.16648998856544495,
+      "learning_rate": 1.736741137876405e-05,
+      "loss": 1.1778,
+      "step": 7110
+    },
+    {
+      "epoch": 2.6474619259810823,
+      "grad_norm": 0.16086730360984802,
+      "learning_rate": 1.7366589557975287e-05,
+      "loss": 1.1909,
+      "step": 7111
+    },
+    {
+      "epoch": 2.647834231131691,
+      "grad_norm": 0.16966967284679413,
+      "learning_rate": 1.7365767628382456e-05,
+      "loss": 1.1799,
+      "step": 7112
+    },
+    {
+      "epoch": 2.6482065362823004,
+      "grad_norm": 0.16273948550224304,
+      "learning_rate": 1.7364945589997703e-05,
+      "loss": 1.1777,
+      "step": 7113
+    },
+    {
+      "epoch": 2.6485788414329097,
+      "grad_norm": 0.16940386593341827,
+      "learning_rate": 1.736412344283316e-05,
+      "loss": 1.1837,
+      "step": 7114
+    },
+    {
+      "epoch": 2.6489511465835185,
+      "grad_norm": 0.17343072593212128,
+      "learning_rate": 1.736330118690098e-05,
+      "loss": 1.173,
+      "step": 7115
+    },
+    {
+      "epoch": 2.6493234517341273,
+      "grad_norm": 0.16787442564964294,
+      "learning_rate": 1.73624788222133e-05,
+      "loss": 1.1676,
+      "step": 7116
+    },
+    {
+      "epoch": 2.6496957568847366,
+      "grad_norm": 0.1711404025554657,
+      "learning_rate": 1.7361656348782275e-05,
+      "loss": 1.1643,
+      "step": 7117
+    },
+    {
+      "epoch": 2.650068062035346,
+      "grad_norm": 0.17250101268291473,
+      "learning_rate": 1.7360833766620046e-05,
+      "loss": 1.1709,
+      "step": 7118
+    },
+    {
+      "epoch": 2.6504403671859547,
+      "grad_norm": 0.16760481894016266,
+      "learning_rate": 1.7360011075738762e-05,
+      "loss": 1.1867,
+      "step": 7119
+    },
+    {
+      "epoch": 2.650812672336564,
+      "grad_norm": 0.16904081404209137,
+      "learning_rate": 1.7359188276150578e-05,
+      "loss": 1.1658,
+      "step": 7120
+    },
+    {
+      "epoch": 2.6511849774871727,
+      "grad_norm": 0.16162091493606567,
+      "learning_rate": 1.7358365367867643e-05,
+      "loss": 1.1653,
+      "step": 7121
+    },
+    {
+      "epoch": 2.651557282637782,
+      "grad_norm": 0.166825532913208,
+      "learning_rate": 1.7357542350902114e-05,
+      "loss": 1.1697,
+      "step": 7122
+    },
+    {
+      "epoch": 2.6519295877883913,
+      "grad_norm": 0.16394229233264923,
+      "learning_rate": 1.7356719225266147e-05,
+      "loss": 1.1705,
+      "step": 7123
+    },
+    {
+      "epoch": 2.652301892939,
+      "grad_norm": 0.16806428134441376,
+      "learning_rate": 1.73558959909719e-05,
+      "loss": 1.1602,
+      "step": 7124
+    },
+    {
+      "epoch": 2.6526741980896094,
+      "grad_norm": 0.16216734051704407,
+      "learning_rate": 1.735507264803153e-05,
+      "loss": 1.1691,
+      "step": 7125
+    },
+    {
+      "epoch": 2.653046503240218,
+      "grad_norm": 0.17059317231178284,
+      "learning_rate": 1.7354249196457198e-05,
+      "loss": 1.1784,
+      "step": 7126
+    },
+    {
+      "epoch": 2.6534188083908274,
+      "grad_norm": 0.1767067313194275,
+      "learning_rate": 1.7353425636261067e-05,
+      "loss": 1.1686,
+      "step": 7127
+    },
+    {
+      "epoch": 2.6537911135414363,
+      "grad_norm": 0.16557256877422333,
+      "learning_rate": 1.7352601967455303e-05,
+      "loss": 1.1818,
+      "step": 7128
+    },
+    {
+      "epoch": 2.6541634186920455,
+      "grad_norm": 0.17525194585323334,
+      "learning_rate": 1.7351778190052067e-05,
+      "loss": 1.1759,
+      "step": 7129
+    },
+    {
+      "epoch": 2.6545357238426543,
+      "grad_norm": 0.16801874339580536,
+      "learning_rate": 1.7350954304063528e-05,
+      "loss": 1.1652,
+      "step": 7130
+    },
+    {
+      "epoch": 2.6549080289932636,
+      "grad_norm": 0.16785259544849396,
+      "learning_rate": 1.7350130309501855e-05,
+      "loss": 1.1728,
+      "step": 7131
+    },
+    {
+      "epoch": 2.655280334143873,
+      "grad_norm": 0.18316105008125305,
+      "learning_rate": 1.734930620637922e-05,
+      "loss": 1.1842,
+      "step": 7132
+    },
+    {
+      "epoch": 2.6556526392944817,
+      "grad_norm": 0.18117022514343262,
+      "learning_rate": 1.7348481994707795e-05,
+      "loss": 1.1717,
+      "step": 7133
+    },
+    {
+      "epoch": 2.656024944445091,
+      "grad_norm": 0.17637571692466736,
+      "learning_rate": 1.734765767449975e-05,
+      "loss": 1.1801,
+      "step": 7134
+    },
+    {
+      "epoch": 2.6563972495957,
+      "grad_norm": 0.1829034388065338,
+      "learning_rate": 1.7346833245767265e-05,
+      "loss": 1.1731,
+      "step": 7135
+    },
+    {
+      "epoch": 2.656769554746309,
+      "grad_norm": 0.17745697498321533,
+      "learning_rate": 1.734600870852251e-05,
+      "loss": 1.1639,
+      "step": 7136
+    },
+    {
+      "epoch": 2.657141859896918,
+      "grad_norm": 0.176448792219162,
+      "learning_rate": 1.7345184062777668e-05,
+      "loss": 1.1718,
+      "step": 7137
+    },
+    {
+      "epoch": 2.657514165047527,
+      "grad_norm": 0.1889655888080597,
+      "learning_rate": 1.734435930854492e-05,
+      "loss": 1.1717,
+      "step": 7138
+    },
+    {
+      "epoch": 2.657886470198136,
+      "grad_norm": 0.21692200005054474,
+      "learning_rate": 1.7343534445836446e-05,
+      "loss": 1.1726,
+      "step": 7139
+    },
+    {
+      "epoch": 2.658258775348745,
+      "grad_norm": 0.22923436760902405,
+      "learning_rate": 1.7342709474664426e-05,
+      "loss": 1.177,
+      "step": 7140
+    },
+    {
+      "epoch": 2.6586310804993545,
+      "grad_norm": 0.2060500830411911,
+      "learning_rate": 1.7341884395041052e-05,
+      "loss": 1.1685,
+      "step": 7141
+    },
+    {
+      "epoch": 2.6590033856499633,
+      "grad_norm": 0.16787934303283691,
+      "learning_rate": 1.7341059206978505e-05,
+      "loss": 1.1647,
+      "step": 7142
+    },
+    {
+      "epoch": 2.6593756908005726,
+      "grad_norm": 0.2028505802154541,
+      "learning_rate": 1.7340233910488973e-05,
+      "loss": 1.1583,
+      "step": 7143
+    },
+    {
+      "epoch": 2.6597479959511814,
+      "grad_norm": 0.18768154084682465,
+      "learning_rate": 1.7339408505584653e-05,
+      "loss": 1.1632,
+      "step": 7144
+    },
+    {
+      "epoch": 2.6601203011017907,
+      "grad_norm": 0.17588740587234497,
+      "learning_rate": 1.7338582992277723e-05,
+      "loss": 1.1676,
+      "step": 7145
+    },
+    {
+      "epoch": 2.6604926062523995,
+      "grad_norm": 0.19588559865951538,
+      "learning_rate": 1.7337757370580385e-05,
+      "loss": 1.1545,
+      "step": 7146
+    },
+    {
+      "epoch": 2.6608649114030087,
+      "grad_norm": 0.16614019870758057,
+      "learning_rate": 1.733693164050483e-05,
+      "loss": 1.1864,
+      "step": 7147
+    },
+    {
+      "epoch": 2.6612372165536176,
+      "grad_norm": 0.17815782129764557,
+      "learning_rate": 1.7336105802063255e-05,
+      "loss": 1.1624,
+      "step": 7148
+    },
+    {
+      "epoch": 2.661609521704227,
+      "grad_norm": 0.172567680478096,
+      "learning_rate": 1.7335279855267858e-05,
+      "loss": 1.1725,
+      "step": 7149
+    },
+    {
+      "epoch": 2.661981826854836,
+      "grad_norm": 0.16731174290180206,
+      "learning_rate": 1.733445380013084e-05,
+      "loss": 1.1608,
+      "step": 7150
+    },
+    {
+      "epoch": 2.662354132005445,
+      "grad_norm": 0.17305542528629303,
+      "learning_rate": 1.7333627636664397e-05,
+      "loss": 1.1803,
+      "step": 7151
+    },
+    {
+      "epoch": 2.662726437156054,
+      "grad_norm": 0.16189545392990112,
+      "learning_rate": 1.7332801364880734e-05,
+      "loss": 1.1735,
+      "step": 7152
+    },
+    {
+      "epoch": 2.663098742306663,
+      "grad_norm": 0.25279855728149414,
+      "learning_rate": 1.7331974984792056e-05,
+      "loss": 1.1793,
+      "step": 7153
+    },
+    {
+      "epoch": 2.6634710474572723,
+      "grad_norm": 0.1899125874042511,
+      "learning_rate": 1.733114849641057e-05,
+      "loss": 1.1931,
+      "step": 7154
+    },
+    {
+      "epoch": 2.663843352607881,
+      "grad_norm": 0.17817838490009308,
+      "learning_rate": 1.7330321899748476e-05,
+      "loss": 1.1718,
+      "step": 7155
+    },
+    {
+      "epoch": 2.6642156577584903,
+      "grad_norm": 0.16445818543434143,
+      "learning_rate": 1.732949519481799e-05,
+      "loss": 1.1743,
+      "step": 7156
+    },
+    {
+      "epoch": 2.664587962909099,
+      "grad_norm": 0.1586284637451172,
+      "learning_rate": 1.732866838163132e-05,
+      "loss": 1.173,
+      "step": 7157
+    },
+    {
+      "epoch": 2.6649602680597084,
+      "grad_norm": 0.1685306876897812,
+      "learning_rate": 1.7327841460200677e-05,
+      "loss": 1.1722,
+      "step": 7158
+    },
+    {
+      "epoch": 2.6653325732103177,
+      "grad_norm": 0.16867049038410187,
+      "learning_rate": 1.732701443053828e-05,
+      "loss": 1.1736,
+      "step": 7159
+    },
+    {
+      "epoch": 2.6657048783609265,
+      "grad_norm": 0.16906088590621948,
+      "learning_rate": 1.7326187292656332e-05,
+      "loss": 1.1659,
+      "step": 7160
+    },
+    {
+      "epoch": 2.6660771835115358,
+      "grad_norm": 0.16725754737854004,
+      "learning_rate": 1.7325360046567065e-05,
+      "loss": 1.1736,
+      "step": 7161
+    },
+    {
+      "epoch": 2.6664494886621446,
+      "grad_norm": 0.1647002398967743,
+      "learning_rate": 1.732453269228268e-05,
+      "loss": 1.1687,
+      "step": 7162
+    },
+    {
+      "epoch": 2.666821793812754,
+      "grad_norm": 0.16950398683547974,
+      "learning_rate": 1.7323705229815416e-05,
+      "loss": 1.1733,
+      "step": 7163
+    },
+    {
+      "epoch": 2.6671940989633627,
+      "grad_norm": 0.16586348414421082,
+      "learning_rate": 1.7322877659177482e-05,
+      "loss": 1.1649,
+      "step": 7164
+    },
+    {
+      "epoch": 2.667566404113972,
+      "grad_norm": 0.1677524447441101,
+      "learning_rate": 1.732204998038111e-05,
+      "loss": 1.1786,
+      "step": 7165
+    },
+    {
+      "epoch": 2.6679387092645808,
+      "grad_norm": 0.16064026951789856,
+      "learning_rate": 1.7321222193438513e-05,
+      "loss": 1.1817,
+      "step": 7166
+    },
+    {
+      "epoch": 2.66831101441519,
+      "grad_norm": 0.17599624395370483,
+      "learning_rate": 1.7320394298361926e-05,
+      "loss": 1.1767,
+      "step": 7167
+    },
+    {
+      "epoch": 2.6686833195657993,
+      "grad_norm": 0.16839216649532318,
+      "learning_rate": 1.7319566295163572e-05,
+      "loss": 1.165,
+      "step": 7168
+    },
+    {
+      "epoch": 2.669055624716408,
+      "grad_norm": 0.16207967698574066,
+      "learning_rate": 1.7318738183855685e-05,
+      "loss": 1.1772,
+      "step": 7169
+    },
+    {
+      "epoch": 2.6694279298670174,
+      "grad_norm": 0.1666906476020813,
+      "learning_rate": 1.7317909964450494e-05,
+      "loss": 1.1723,
+      "step": 7170
+    },
+    {
+      "epoch": 2.669800235017626,
+      "grad_norm": 0.1715075820684433,
+      "learning_rate": 1.731708163696023e-05,
+      "loss": 1.1788,
+      "step": 7171
+    },
+    {
+      "epoch": 2.6701725401682355,
+      "grad_norm": 0.1627563238143921,
+      "learning_rate": 1.7316253201397134e-05,
+      "loss": 1.1688,
+      "step": 7172
+    },
+    {
+      "epoch": 2.6705448453188447,
+      "grad_norm": 0.1687730997800827,
+      "learning_rate": 1.7315424657773433e-05,
+      "loss": 1.1657,
+      "step": 7173
+    },
+    {
+      "epoch": 2.6709171504694535,
+      "grad_norm": 0.16402725875377655,
+      "learning_rate": 1.7314596006101372e-05,
+      "loss": 1.1852,
+      "step": 7174
+    },
+    {
+      "epoch": 2.6712894556200624,
+      "grad_norm": 0.16662922501564026,
+      "learning_rate": 1.7313767246393184e-05,
+      "loss": 1.1782,
+      "step": 7175
+    },
+    {
+      "epoch": 2.6716617607706716,
+      "grad_norm": 0.1664755940437317,
+      "learning_rate": 1.7312938378661118e-05,
+      "loss": 1.1724,
+      "step": 7176
+    },
+    {
+      "epoch": 2.672034065921281,
+      "grad_norm": 0.15922842919826508,
+      "learning_rate": 1.7312109402917406e-05,
+      "loss": 1.1719,
+      "step": 7177
+    },
+    {
+      "epoch": 2.6724063710718897,
+      "grad_norm": 0.16002245247364044,
+      "learning_rate": 1.73112803191743e-05,
+      "loss": 1.176,
+      "step": 7178
+    },
+    {
+      "epoch": 2.672778676222499,
+      "grad_norm": 0.16558903455734253,
+      "learning_rate": 1.731045112744404e-05,
+      "loss": 1.1703,
+      "step": 7179
+    },
+    {
+      "epoch": 2.673150981373108,
+      "grad_norm": 0.16779029369354248,
+      "learning_rate": 1.7309621827738877e-05,
+      "loss": 1.1909,
+      "step": 7180
+    },
+    {
+      "epoch": 2.673523286523717,
+      "grad_norm": 0.16729268431663513,
+      "learning_rate": 1.730879242007106e-05,
+      "loss": 1.1688,
+      "step": 7181
+    },
+    {
+      "epoch": 2.6738955916743263,
+      "grad_norm": 0.1586335152387619,
+      "learning_rate": 1.7307962904452837e-05,
+      "loss": 1.1748,
+      "step": 7182
+    },
+    {
+      "epoch": 2.674267896824935,
+      "grad_norm": 0.16203922033309937,
+      "learning_rate": 1.730713328089646e-05,
+      "loss": 1.1755,
+      "step": 7183
+    },
+    {
+      "epoch": 2.674640201975544,
+      "grad_norm": 0.17098890244960785,
+      "learning_rate": 1.730630354941418e-05,
+      "loss": 1.1814,
+      "step": 7184
+    },
+    {
+      "epoch": 2.6750125071261532,
+      "grad_norm": 0.16195635497570038,
+      "learning_rate": 1.7305473710018258e-05,
+      "loss": 1.1654,
+      "step": 7185
+    },
+    {
+      "epoch": 2.6753848122767625,
+      "grad_norm": 0.16069965064525604,
+      "learning_rate": 1.730464376272095e-05,
+      "loss": 1.1661,
+      "step": 7186
+    },
+    {
+      "epoch": 2.6757571174273713,
+      "grad_norm": 0.1605072170495987,
+      "learning_rate": 1.7303813707534506e-05,
+      "loss": 1.1829,
+      "step": 7187
+    },
+    {
+      "epoch": 2.6761294225779806,
+      "grad_norm": 0.16202007234096527,
+      "learning_rate": 1.7302983544471197e-05,
+      "loss": 1.1668,
+      "step": 7188
+    },
+    {
+      "epoch": 2.6765017277285894,
+      "grad_norm": 0.16660577058792114,
+      "learning_rate": 1.7302153273543276e-05,
+      "loss": 1.1741,
+      "step": 7189
+    },
+    {
+      "epoch": 2.6768740328791987,
+      "grad_norm": 0.16862109303474426,
+      "learning_rate": 1.7301322894763013e-05,
+      "loss": 1.1775,
+      "step": 7190
+    },
+    {
+      "epoch": 2.677246338029808,
+      "grad_norm": 0.1640176773071289,
+      "learning_rate": 1.7300492408142666e-05,
+      "loss": 1.1671,
+      "step": 7191
+    },
+    {
+      "epoch": 2.6776186431804168,
+      "grad_norm": 0.1989293098449707,
+      "learning_rate": 1.729966181369451e-05,
+      "loss": 1.1844,
+      "step": 7192
+    },
+    {
+      "epoch": 2.6779909483310256,
+      "grad_norm": 0.21205829083919525,
+      "learning_rate": 1.72988311114308e-05,
+      "loss": 1.1789,
+      "step": 7193
+    },
+    {
+      "epoch": 2.678363253481635,
+      "grad_norm": 0.16825103759765625,
+      "learning_rate": 1.7298000301363815e-05,
+      "loss": 1.1659,
+      "step": 7194
+    },
+    {
+      "epoch": 2.678735558632244,
+      "grad_norm": 0.5028261542320251,
+      "learning_rate": 1.729716938350582e-05,
+      "loss": 1.1628,
+      "step": 7195
+    },
+    {
+      "epoch": 2.679107863782853,
+      "grad_norm": 0.1976155936717987,
+      "learning_rate": 1.72963383578691e-05,
+      "loss": 1.172,
+      "step": 7196
+    },
+    {
+      "epoch": 2.679480168933462,
+      "grad_norm": 0.18413805961608887,
+      "learning_rate": 1.7295507224465913e-05,
+      "loss": 1.1695,
+      "step": 7197
+    },
+    {
+      "epoch": 2.679852474084071,
+      "grad_norm": 0.17518746852874756,
+      "learning_rate": 1.7294675983308545e-05,
+      "loss": 1.1602,
+      "step": 7198
+    },
+    {
+      "epoch": 2.6802247792346803,
+      "grad_norm": 0.1705147624015808,
+      "learning_rate": 1.729384463440927e-05,
+      "loss": 1.1783,
+      "step": 7199
+    },
+    {
+      "epoch": 2.6805970843852895,
+      "grad_norm": 0.18338525295257568,
+      "learning_rate": 1.7293013177780368e-05,
+      "loss": 1.1786,
+      "step": 7200
+    },
+    {
+      "epoch": 2.6809693895358984,
+      "grad_norm": 0.1804172545671463,
+      "learning_rate": 1.7292181613434117e-05,
+      "loss": 1.1682,
+      "step": 7201
+    },
+    {
+      "epoch": 2.681341694686507,
+      "grad_norm": 0.17289164662361145,
+      "learning_rate": 1.7291349941382804e-05,
+      "loss": 1.1769,
+      "step": 7202
+    },
+    {
+      "epoch": 2.6817139998371164,
+      "grad_norm": 0.15843546390533447,
+      "learning_rate": 1.7290518161638707e-05,
+      "loss": 1.1736,
+      "step": 7203
+    },
+    {
+      "epoch": 2.6820863049877257,
+      "grad_norm": 0.16561047732830048,
+      "learning_rate": 1.7289686274214116e-05,
+      "loss": 1.1621,
+      "step": 7204
+    },
+    {
+      "epoch": 2.6824586101383345,
+      "grad_norm": 0.16593630611896515,
+      "learning_rate": 1.7288854279121318e-05,
+      "loss": 1.1724,
+      "step": 7205
+    },
+    {
+      "epoch": 2.682830915288944,
+      "grad_norm": 0.16903571784496307,
+      "learning_rate": 1.7288022176372597e-05,
+      "loss": 1.1679,
+      "step": 7206
+    },
+    {
+      "epoch": 2.6832032204395526,
+      "grad_norm": 0.16603899002075195,
+      "learning_rate": 1.7287189965980245e-05,
+      "loss": 1.1607,
+      "step": 7207
+    },
+    {
+      "epoch": 2.683575525590162,
+      "grad_norm": 0.16013038158416748,
+      "learning_rate": 1.728635764795656e-05,
+      "loss": 1.1641,
+      "step": 7208
+    },
+    {
+      "epoch": 2.683947830740771,
+      "grad_norm": 0.16311000287532806,
+      "learning_rate": 1.7285525222313823e-05,
+      "loss": 1.1721,
+      "step": 7209
+    },
+    {
+      "epoch": 2.68432013589138,
+      "grad_norm": 0.16615121066570282,
+      "learning_rate": 1.728469268906434e-05,
+      "loss": 1.181,
+      "step": 7210
+    },
+    {
+      "epoch": 2.684692441041989,
+      "grad_norm": 0.16762305796146393,
+      "learning_rate": 1.7283860048220403e-05,
+      "loss": 1.1645,
+      "step": 7211
+    },
+    {
+      "epoch": 2.685064746192598,
+      "grad_norm": 0.16307219862937927,
+      "learning_rate": 1.7283027299794306e-05,
+      "loss": 1.1769,
+      "step": 7212
+    },
+    {
+      "epoch": 2.6854370513432073,
+      "grad_norm": 0.16042660176753998,
+      "learning_rate": 1.7282194443798358e-05,
+      "loss": 1.1636,
+      "step": 7213
+    },
+    {
+      "epoch": 2.685809356493816,
+      "grad_norm": 0.16484393179416656,
+      "learning_rate": 1.7281361480244852e-05,
+      "loss": 1.1716,
+      "step": 7214
+    },
+    {
+      "epoch": 2.6861816616444254,
+      "grad_norm": 0.16563540697097778,
+      "learning_rate": 1.7280528409146097e-05,
+      "loss": 1.1618,
+      "step": 7215
+    },
+    {
+      "epoch": 2.686553966795034,
+      "grad_norm": 0.16255468130111694,
+      "learning_rate": 1.7279695230514392e-05,
+      "loss": 1.1771,
+      "step": 7216
+    },
+    {
+      "epoch": 2.6869262719456435,
+      "grad_norm": 0.16036321222782135,
+      "learning_rate": 1.7278861944362045e-05,
+      "loss": 1.1671,
+      "step": 7217
+    },
+    {
+      "epoch": 2.6872985770962527,
+      "grad_norm": 0.16216713190078735,
+      "learning_rate": 1.7278028550701364e-05,
+      "loss": 1.1772,
+      "step": 7218
+    },
+    {
+      "epoch": 2.6876708822468616,
+      "grad_norm": 0.16452451050281525,
+      "learning_rate": 1.727719504954466e-05,
+      "loss": 1.1639,
+      "step": 7219
+    },
+    {
+      "epoch": 2.6880431873974704,
+      "grad_norm": 0.16761431097984314,
+      "learning_rate": 1.727636144090424e-05,
+      "loss": 1.1635,
+      "step": 7220
+    },
+    {
+      "epoch": 2.6884154925480797,
+      "grad_norm": 0.16240081191062927,
+      "learning_rate": 1.7275527724792416e-05,
+      "loss": 1.1699,
+      "step": 7221
+    },
+    {
+      "epoch": 2.688787797698689,
+      "grad_norm": 0.16535256803035736,
+      "learning_rate": 1.7274693901221507e-05,
+      "loss": 1.1815,
+      "step": 7222
+    },
+    {
+      "epoch": 2.6891601028492977,
+      "grad_norm": 0.16197264194488525,
+      "learning_rate": 1.7273859970203825e-05,
+      "loss": 1.1708,
+      "step": 7223
+    },
+    {
+      "epoch": 2.689532407999907,
+      "grad_norm": 0.1600698083639145,
+      "learning_rate": 1.727302593175169e-05,
+      "loss": 1.1802,
+      "step": 7224
+    },
+    {
+      "epoch": 2.689904713150516,
+      "grad_norm": 0.1599569022655487,
+      "learning_rate": 1.7272191785877415e-05,
+      "loss": 1.1891,
+      "step": 7225
+    },
+    {
+      "epoch": 2.690277018301125,
+      "grad_norm": 0.16361677646636963,
+      "learning_rate": 1.7271357532593325e-05,
+      "loss": 1.1736,
+      "step": 7226
+    },
+    {
+      "epoch": 2.6906493234517344,
+      "grad_norm": 0.16064906120300293,
+      "learning_rate": 1.727052317191174e-05,
+      "loss": 1.168,
+      "step": 7227
+    },
+    {
+      "epoch": 2.691021628602343,
+      "grad_norm": 0.16545885801315308,
+      "learning_rate": 1.7269688703844984e-05,
+      "loss": 1.1683,
+      "step": 7228
+    },
+    {
+      "epoch": 2.691393933752952,
+      "grad_norm": 0.16295583546161652,
+      "learning_rate": 1.7268854128405384e-05,
+      "loss": 1.1605,
+      "step": 7229
+    },
+    {
+      "epoch": 2.6917662389035613,
+      "grad_norm": 0.16552822291851044,
+      "learning_rate": 1.7268019445605263e-05,
+      "loss": 1.161,
+      "step": 7230
+    },
+    {
+      "epoch": 2.6921385440541705,
+      "grad_norm": 0.16663306951522827,
+      "learning_rate": 1.726718465545695e-05,
+      "loss": 1.17,
+      "step": 7231
+    },
+    {
+      "epoch": 2.6925108492047793,
+      "grad_norm": 0.15905286371707916,
+      "learning_rate": 1.726634975797278e-05,
+      "loss": 1.1616,
+      "step": 7232
+    },
+    {
+      "epoch": 2.6928831543553886,
+      "grad_norm": 0.1620490998029709,
+      "learning_rate": 1.7265514753165075e-05,
+      "loss": 1.176,
+      "step": 7233
+    },
+    {
+      "epoch": 2.6932554595059974,
+      "grad_norm": 0.16328297555446625,
+      "learning_rate": 1.7264679641046176e-05,
+      "loss": 1.1792,
+      "step": 7234
+    },
+    {
+      "epoch": 2.6936277646566067,
+      "grad_norm": 0.1633196771144867,
+      "learning_rate": 1.7263844421628416e-05,
+      "loss": 1.1659,
+      "step": 7235
+    },
+    {
+      "epoch": 2.694000069807216,
+      "grad_norm": 0.1595584601163864,
+      "learning_rate": 1.7263009094924125e-05,
+      "loss": 1.1728,
+      "step": 7236
+    },
+    {
+      "epoch": 2.694372374957825,
+      "grad_norm": 0.16301463544368744,
+      "learning_rate": 1.7262173660945648e-05,
+      "loss": 1.1735,
+      "step": 7237
+    },
+    {
+      "epoch": 2.6947446801084336,
+      "grad_norm": 0.16339920461177826,
+      "learning_rate": 1.7261338119705323e-05,
+      "loss": 1.1772,
+      "step": 7238
+    },
+    {
+      "epoch": 2.695116985259043,
+      "grad_norm": 0.16579227149486542,
+      "learning_rate": 1.7260502471215488e-05,
+      "loss": 1.171,
+      "step": 7239
+    },
+    {
+      "epoch": 2.695489290409652,
+      "grad_norm": 0.164364293217659,
+      "learning_rate": 1.7259666715488487e-05,
+      "loss": 1.1683,
+      "step": 7240
+    },
+    {
+      "epoch": 2.695861595560261,
+      "grad_norm": 0.16661317646503448,
+      "learning_rate": 1.7258830852536666e-05,
+      "loss": 1.1794,
+      "step": 7241
+    },
+    {
+      "epoch": 2.69623390071087,
+      "grad_norm": 0.16931530833244324,
+      "learning_rate": 1.7257994882372368e-05,
+      "loss": 1.1703,
+      "step": 7242
+    },
+    {
+      "epoch": 2.696606205861479,
+      "grad_norm": 0.1614353507757187,
+      "learning_rate": 1.725715880500794e-05,
+      "loss": 1.1639,
+      "step": 7243
+    },
+    {
+      "epoch": 2.6969785110120883,
+      "grad_norm": 0.16335289180278778,
+      "learning_rate": 1.7256322620455733e-05,
+      "loss": 1.1744,
+      "step": 7244
+    },
+    {
+      "epoch": 2.6973508161626976,
+      "grad_norm": 0.16357359290122986,
+      "learning_rate": 1.7255486328728096e-05,
+      "loss": 1.1627,
+      "step": 7245
+    },
+    {
+      "epoch": 2.6977231213133064,
+      "grad_norm": 0.16286170482635498,
+      "learning_rate": 1.725464992983738e-05,
+      "loss": 1.1709,
+      "step": 7246
+    },
+    {
+      "epoch": 2.6980954264639156,
+      "grad_norm": 0.16695433855056763,
+      "learning_rate": 1.7253813423795943e-05,
+      "loss": 1.1667,
+      "step": 7247
+    },
+    {
+      "epoch": 2.6984677316145245,
+      "grad_norm": 0.15719039738178253,
+      "learning_rate": 1.7252976810616134e-05,
+      "loss": 1.1666,
+      "step": 7248
+    },
+    {
+      "epoch": 2.6988400367651337,
+      "grad_norm": 0.15986768901348114,
+      "learning_rate": 1.7252140090310314e-05,
+      "loss": 1.1727,
+      "step": 7249
+    },
+    {
+      "epoch": 2.6992123419157426,
+      "grad_norm": 0.15737248957157135,
+      "learning_rate": 1.7251303262890838e-05,
+      "loss": 1.181,
+      "step": 7250
+    },
+    {
+      "epoch": 2.699584647066352,
+      "grad_norm": 0.1623496413230896,
+      "learning_rate": 1.725046632837007e-05,
+      "loss": 1.1727,
+      "step": 7251
+    },
+    {
+      "epoch": 2.6999569522169606,
+      "grad_norm": 0.16519080102443695,
+      "learning_rate": 1.724962928676037e-05,
+      "loss": 1.1729,
+      "step": 7252
+    },
+    {
+      "epoch": 2.70032925736757,
+      "grad_norm": 0.16225138306617737,
+      "learning_rate": 1.72487921380741e-05,
+      "loss": 1.1672,
+      "step": 7253
+    },
+    {
+      "epoch": 2.700701562518179,
+      "grad_norm": 0.16735146939754486,
+      "learning_rate": 1.7247954882323622e-05,
+      "loss": 1.1627,
+      "step": 7254
+    },
+    {
+      "epoch": 2.701073867668788,
+      "grad_norm": 0.1635318249464035,
+      "learning_rate": 1.724711751952131e-05,
+      "loss": 1.166,
+      "step": 7255
+    },
+    {
+      "epoch": 2.7014461728193973,
+      "grad_norm": 0.1578553318977356,
+      "learning_rate": 1.7246280049679526e-05,
+      "loss": 1.155,
+      "step": 7256
+    },
+    {
+      "epoch": 2.701818477970006,
+      "grad_norm": 0.16123878955841064,
+      "learning_rate": 1.7245442472810638e-05,
+      "loss": 1.1906,
+      "step": 7257
+    },
+    {
+      "epoch": 2.7021907831206153,
+      "grad_norm": 0.16336768865585327,
+      "learning_rate": 1.724460478892702e-05,
+      "loss": 1.179,
+      "step": 7258
+    },
+    {
+      "epoch": 2.702563088271224,
+      "grad_norm": 0.16107387840747833,
+      "learning_rate": 1.7243766998041045e-05,
+      "loss": 1.1603,
+      "step": 7259
+    },
+    {
+      "epoch": 2.7029353934218334,
+      "grad_norm": 0.15997155010700226,
+      "learning_rate": 1.7242929100165085e-05,
+      "loss": 1.1739,
+      "step": 7260
+    },
+    {
+      "epoch": 2.7033076985724422,
+      "grad_norm": 0.16311846673488617,
+      "learning_rate": 1.7242091095311516e-05,
+      "loss": 1.1698,
+      "step": 7261
+    },
+    {
+      "epoch": 2.7036800037230515,
+      "grad_norm": 0.16336789727210999,
+      "learning_rate": 1.724125298349272e-05,
+      "loss": 1.1584,
+      "step": 7262
+    },
+    {
+      "epoch": 2.7040523088736608,
+      "grad_norm": 0.1627056896686554,
+      "learning_rate": 1.7240414764721067e-05,
+      "loss": 1.1787,
+      "step": 7263
+    },
+    {
+      "epoch": 2.7044246140242696,
+      "grad_norm": 0.1628088355064392,
+      "learning_rate": 1.7239576439008945e-05,
+      "loss": 1.185,
+      "step": 7264
+    },
+    {
+      "epoch": 2.704796919174879,
+      "grad_norm": 0.16446030139923096,
+      "learning_rate": 1.723873800636873e-05,
+      "loss": 1.1599,
+      "step": 7265
+    },
+    {
+      "epoch": 2.7051692243254877,
+      "grad_norm": 0.16059374809265137,
+      "learning_rate": 1.7237899466812814e-05,
+      "loss": 1.1611,
+      "step": 7266
+    },
+    {
+      "epoch": 2.705541529476097,
+      "grad_norm": 0.16304123401641846,
+      "learning_rate": 1.7237060820353573e-05,
+      "loss": 1.1764,
+      "step": 7267
+    },
+    {
+      "epoch": 2.7059138346267058,
+      "grad_norm": 0.16345947980880737,
+      "learning_rate": 1.7236222067003402e-05,
+      "loss": 1.1757,
+      "step": 7268
+    },
+    {
+      "epoch": 2.706286139777315,
+      "grad_norm": 0.15779638290405273,
+      "learning_rate": 1.7235383206774682e-05,
+      "loss": 1.1699,
+      "step": 7269
+    },
+    {
+      "epoch": 2.706658444927924,
+      "grad_norm": 0.1547388732433319,
+      "learning_rate": 1.7234544239679807e-05,
+      "loss": 1.1618,
+      "step": 7270
+    },
+    {
+      "epoch": 2.707030750078533,
+      "grad_norm": 0.16903281211853027,
+      "learning_rate": 1.723370516573117e-05,
+      "loss": 1.1696,
+      "step": 7271
+    },
+    {
+      "epoch": 2.7074030552291424,
+      "grad_norm": 0.16696353256702423,
+      "learning_rate": 1.7232865984941156e-05,
+      "loss": 1.1588,
+      "step": 7272
+    },
+    {
+      "epoch": 2.707775360379751,
+      "grad_norm": 0.1645137071609497,
+      "learning_rate": 1.723202669732217e-05,
+      "loss": 1.1645,
+      "step": 7273
+    },
+    {
+      "epoch": 2.7081476655303605,
+      "grad_norm": 0.16324058175086975,
+      "learning_rate": 1.72311873028866e-05,
+      "loss": 1.1723,
+      "step": 7274
+    },
+    {
+      "epoch": 2.7085199706809693,
+      "grad_norm": 0.16554492712020874,
+      "learning_rate": 1.723034780164685e-05,
+      "loss": 1.1936,
+      "step": 7275
+    },
+    {
+      "epoch": 2.7088922758315785,
+      "grad_norm": 0.16024906933307648,
+      "learning_rate": 1.7229508193615316e-05,
+      "loss": 1.1689,
+      "step": 7276
+    },
+    {
+      "epoch": 2.7092645809821874,
+      "grad_norm": 0.16716809570789337,
+      "learning_rate": 1.72286684788044e-05,
+      "loss": 1.1698,
+      "step": 7277
+    },
+    {
+      "epoch": 2.7096368861327966,
+      "grad_norm": 0.16631504893302917,
+      "learning_rate": 1.7227828657226506e-05,
+      "loss": 1.1789,
+      "step": 7278
+    },
+    {
+      "epoch": 2.7100091912834054,
+      "grad_norm": 0.1682237982749939,
+      "learning_rate": 1.7226988728894033e-05,
+      "loss": 1.1719,
+      "step": 7279
+    },
+    {
+      "epoch": 2.7103814964340147,
+      "grad_norm": 0.15563975274562836,
+      "learning_rate": 1.722614869381939e-05,
+      "loss": 1.166,
+      "step": 7280
+    },
+    {
+      "epoch": 2.710753801584624,
+      "grad_norm": 0.16396182775497437,
+      "learning_rate": 1.7225308552014988e-05,
+      "loss": 1.1777,
+      "step": 7281
+    },
+    {
+      "epoch": 2.711126106735233,
+      "grad_norm": 0.16292405128479004,
+      "learning_rate": 1.722446830349323e-05,
+      "loss": 1.1783,
+      "step": 7282
+    },
+    {
+      "epoch": 2.711498411885842,
+      "grad_norm": 0.15950658917427063,
+      "learning_rate": 1.7223627948266526e-05,
+      "loss": 1.1775,
+      "step": 7283
+    },
+    {
+      "epoch": 2.711870717036451,
+      "grad_norm": 0.16486436128616333,
+      "learning_rate": 1.7222787486347296e-05,
+      "loss": 1.1664,
+      "step": 7284
+    },
+    {
+      "epoch": 2.71224302218706,
+      "grad_norm": 0.16335855424404144,
+      "learning_rate": 1.7221946917747945e-05,
+      "loss": 1.1599,
+      "step": 7285
+    },
+    {
+      "epoch": 2.7126153273376694,
+      "grad_norm": 0.15834614634513855,
+      "learning_rate": 1.722110624248089e-05,
+      "loss": 1.1644,
+      "step": 7286
+    },
+    {
+      "epoch": 2.7129876324882782,
+      "grad_norm": 0.16369593143463135,
+      "learning_rate": 1.722026546055855e-05,
+      "loss": 1.1681,
+      "step": 7287
+    },
+    {
+      "epoch": 2.713359937638887,
+      "grad_norm": 0.16682885587215424,
+      "learning_rate": 1.7219424571993345e-05,
+      "loss": 1.1763,
+      "step": 7288
+    },
+    {
+      "epoch": 2.7137322427894963,
+      "grad_norm": 0.17135527729988098,
+      "learning_rate": 1.721858357679769e-05,
+      "loss": 1.1878,
+      "step": 7289
+    },
+    {
+      "epoch": 2.7141045479401056,
+      "grad_norm": 0.16507208347320557,
+      "learning_rate": 1.7217742474984006e-05,
+      "loss": 1.1619,
+      "step": 7290
+    },
+    {
+      "epoch": 2.7144768530907144,
+      "grad_norm": 0.16445507109165192,
+      "learning_rate": 1.721690126656472e-05,
+      "loss": 1.1794,
+      "step": 7291
+    },
+    {
+      "epoch": 2.7148491582413237,
+      "grad_norm": 0.1608470231294632,
+      "learning_rate": 1.7216059951552256e-05,
+      "loss": 1.1688,
+      "step": 7292
+    },
+    {
+      "epoch": 2.7152214633919325,
+      "grad_norm": 0.16081736981868744,
+      "learning_rate": 1.7215218529959042e-05,
+      "loss": 1.183,
+      "step": 7293
+    },
+    {
+      "epoch": 2.7155937685425418,
+      "grad_norm": 0.16180992126464844,
+      "learning_rate": 1.7214377001797498e-05,
+      "loss": 1.1553,
+      "step": 7294
+    },
+    {
+      "epoch": 2.715966073693151,
+      "grad_norm": 0.16335974633693695,
+      "learning_rate": 1.7213535367080064e-05,
+      "loss": 1.1667,
+      "step": 7295
+    },
+    {
+      "epoch": 2.71633837884376,
+      "grad_norm": 0.1620166003704071,
+      "learning_rate": 1.7212693625819163e-05,
+      "loss": 1.1695,
+      "step": 7296
+    },
+    {
+      "epoch": 2.7167106839943687,
+      "grad_norm": 0.15948939323425293,
+      "learning_rate": 1.7211851778027226e-05,
+      "loss": 1.1667,
+      "step": 7297
+    },
+    {
+      "epoch": 2.717082989144978,
+      "grad_norm": 0.15840508043766022,
+      "learning_rate": 1.7211009823716695e-05,
+      "loss": 1.1611,
+      "step": 7298
+    },
+    {
+      "epoch": 2.717455294295587,
+      "grad_norm": 0.1646668016910553,
+      "learning_rate": 1.72101677629e-05,
+      "loss": 1.1748,
+      "step": 7299
+    },
+    {
+      "epoch": 2.717827599446196,
+      "grad_norm": 0.16672910749912262,
+      "learning_rate": 1.720932559558958e-05,
+      "loss": 1.1637,
+      "step": 7300
+    },
+    {
+      "epoch": 2.7181999045968053,
+      "grad_norm": 0.16393324732780457,
+      "learning_rate": 1.7208483321797876e-05,
+      "loss": 1.1536,
+      "step": 7301
+    },
+    {
+      "epoch": 2.718572209747414,
+      "grad_norm": 0.16275498270988464,
+      "learning_rate": 1.720764094153732e-05,
+      "loss": 1.1675,
+      "step": 7302
+    },
+    {
+      "epoch": 2.7189445148980234,
+      "grad_norm": 0.167283296585083,
+      "learning_rate": 1.7206798454820366e-05,
+      "loss": 1.1612,
+      "step": 7303
+    },
+    {
+      "epoch": 2.7193168200486326,
+      "grad_norm": 0.16104766726493835,
+      "learning_rate": 1.7205955861659446e-05,
+      "loss": 1.1844,
+      "step": 7304
+    },
+    {
+      "epoch": 2.7196891251992414,
+      "grad_norm": 0.16111914813518524,
+      "learning_rate": 1.7205113162067013e-05,
+      "loss": 1.1702,
+      "step": 7305
+    },
+    {
+      "epoch": 2.7200614303498503,
+      "grad_norm": 0.16553427278995514,
+      "learning_rate": 1.720427035605551e-05,
+      "loss": 1.1664,
+      "step": 7306
+    },
+    {
+      "epoch": 2.7204337355004595,
+      "grad_norm": 0.16949054598808289,
+      "learning_rate": 1.7203427443637385e-05,
+      "loss": 1.1755,
+      "step": 7307
+    },
+    {
+      "epoch": 2.720806040651069,
+      "grad_norm": 0.16320055723190308,
+      "learning_rate": 1.7202584424825092e-05,
+      "loss": 1.1693,
+      "step": 7308
+    },
+    {
+      "epoch": 2.7211783458016776,
+      "grad_norm": 0.16071534156799316,
+      "learning_rate": 1.7201741299631075e-05,
+      "loss": 1.1695,
+      "step": 7309
+    },
+    {
+      "epoch": 2.721550650952287,
+      "grad_norm": 0.16326375305652618,
+      "learning_rate": 1.720089806806779e-05,
+      "loss": 1.1712,
+      "step": 7310
+    },
+    {
+      "epoch": 2.7219229561028957,
+      "grad_norm": 0.16622449457645416,
+      "learning_rate": 1.72000547301477e-05,
+      "loss": 1.1679,
+      "step": 7311
+    },
+    {
+      "epoch": 2.722295261253505,
+      "grad_norm": 0.1643485128879547,
+      "learning_rate": 1.7199211285883245e-05,
+      "loss": 1.1791,
+      "step": 7312
+    },
+    {
+      "epoch": 2.7226675664041142,
+      "grad_norm": 0.15881820023059845,
+      "learning_rate": 1.7198367735286897e-05,
+      "loss": 1.1625,
+      "step": 7313
+    },
+    {
+      "epoch": 2.723039871554723,
+      "grad_norm": 0.16432297229766846,
+      "learning_rate": 1.7197524078371105e-05,
+      "loss": 1.1736,
+      "step": 7314
+    },
+    {
+      "epoch": 2.723412176705332,
+      "grad_norm": 0.15992549061775208,
+      "learning_rate": 1.7196680315148335e-05,
+      "loss": 1.1772,
+      "step": 7315
+    },
+    {
+      "epoch": 2.723784481855941,
+      "grad_norm": 0.16613058745861053,
+      "learning_rate": 1.719583644563105e-05,
+      "loss": 1.1808,
+      "step": 7316
+    },
+    {
+      "epoch": 2.7241567870065504,
+      "grad_norm": 0.16339601576328278,
+      "learning_rate": 1.7194992469831712e-05,
+      "loss": 1.1593,
+      "step": 7317
+    },
+    {
+      "epoch": 2.724529092157159,
+      "grad_norm": 0.15978363156318665,
+      "learning_rate": 1.719414838776279e-05,
+      "loss": 1.1716,
+      "step": 7318
+    },
+    {
+      "epoch": 2.7249013973077685,
+      "grad_norm": 0.16619928181171417,
+      "learning_rate": 1.719330419943674e-05,
+      "loss": 1.1694,
+      "step": 7319
+    },
+    {
+      "epoch": 2.7252737024583773,
+      "grad_norm": 0.16087765991687775,
+      "learning_rate": 1.7192459904866042e-05,
+      "loss": 1.1529,
+      "step": 7320
+    },
+    {
+      "epoch": 2.7256460076089866,
+      "grad_norm": 0.16294820606708527,
+      "learning_rate": 1.7191615504063165e-05,
+      "loss": 1.1925,
+      "step": 7321
+    },
+    {
+      "epoch": 2.726018312759596,
+      "grad_norm": 0.1619005650281906,
+      "learning_rate": 1.7190770997040574e-05,
+      "loss": 1.1692,
+      "step": 7322
+    },
+    {
+      "epoch": 2.7263906179102047,
+      "grad_norm": 0.17229419946670532,
+      "learning_rate": 1.7189926383810755e-05,
+      "loss": 1.1677,
+      "step": 7323
+    },
+    {
+      "epoch": 2.7267629230608135,
+      "grad_norm": 0.16301006078720093,
+      "learning_rate": 1.7189081664386168e-05,
+      "loss": 1.1832,
+      "step": 7324
+    },
+    {
+      "epoch": 2.7271352282114227,
+      "grad_norm": 0.16268163919448853,
+      "learning_rate": 1.7188236838779297e-05,
+      "loss": 1.1718,
+      "step": 7325
+    },
+    {
+      "epoch": 2.727507533362032,
+      "grad_norm": 0.16183678805828094,
+      "learning_rate": 1.718739190700262e-05,
+      "loss": 1.1765,
+      "step": 7326
+    },
+    {
+      "epoch": 2.727879838512641,
+      "grad_norm": 0.16009078919887543,
+      "learning_rate": 1.7186546869068612e-05,
+      "loss": 1.1812,
+      "step": 7327
+    },
+    {
+      "epoch": 2.72825214366325,
+      "grad_norm": 0.16223828494548798,
+      "learning_rate": 1.718570172498976e-05,
+      "loss": 1.1905,
+      "step": 7328
+    },
+    {
+      "epoch": 2.728624448813859,
+      "grad_norm": 0.1691174954175949,
+      "learning_rate": 1.7184856474778543e-05,
+      "loss": 1.1672,
+      "step": 7329
+    },
+    {
+      "epoch": 2.728996753964468,
+      "grad_norm": 0.1607164889574051,
+      "learning_rate": 1.7184011118447448e-05,
+      "loss": 1.1904,
+      "step": 7330
+    },
+    {
+      "epoch": 2.7293690591150774,
+      "grad_norm": 0.16635142266750336,
+      "learning_rate": 1.718316565600896e-05,
+      "loss": 1.1633,
+      "step": 7331
+    },
+    {
+      "epoch": 2.7297413642656863,
+      "grad_norm": 0.16551473736763,
+      "learning_rate": 1.7182320087475567e-05,
+      "loss": 1.1649,
+      "step": 7332
+    },
+    {
+      "epoch": 2.730113669416295,
+      "grad_norm": 0.16264252364635468,
+      "learning_rate": 1.7181474412859756e-05,
+      "loss": 1.1627,
+      "step": 7333
+    },
+    {
+      "epoch": 2.7304859745669043,
+      "grad_norm": 0.16819992661476135,
+      "learning_rate": 1.718062863217402e-05,
+      "loss": 1.1749,
+      "step": 7334
+    },
+    {
+      "epoch": 2.7308582797175136,
+      "grad_norm": 0.16753548383712769,
+      "learning_rate": 1.717978274543085e-05,
+      "loss": 1.1715,
+      "step": 7335
+    },
+    {
+      "epoch": 2.7312305848681224,
+      "grad_norm": 0.16801391541957855,
+      "learning_rate": 1.7178936752642737e-05,
+      "loss": 1.175,
+      "step": 7336
+    },
+    {
+      "epoch": 2.7316028900187317,
+      "grad_norm": 0.16924819350242615,
+      "learning_rate": 1.717809065382218e-05,
+      "loss": 1.1764,
+      "step": 7337
+    },
+    {
+      "epoch": 2.7319751951693405,
+      "grad_norm": 0.1656593382358551,
+      "learning_rate": 1.7177244448981675e-05,
+      "loss": 1.1834,
+      "step": 7338
+    },
+    {
+      "epoch": 2.7323475003199498,
+      "grad_norm": 0.15879951417446136,
+      "learning_rate": 1.7176398138133718e-05,
+      "loss": 1.1816,
+      "step": 7339
+    },
+    {
+      "epoch": 2.732719805470559,
+      "grad_norm": 0.16423895955085754,
+      "learning_rate": 1.717555172129081e-05,
+      "loss": 1.1789,
+      "step": 7340
+    },
+    {
+      "epoch": 2.733092110621168,
+      "grad_norm": 0.16044852137565613,
+      "learning_rate": 1.7174705198465454e-05,
+      "loss": 1.158,
+      "step": 7341
+    },
+    {
+      "epoch": 2.7334644157717767,
+      "grad_norm": 0.16656622290611267,
+      "learning_rate": 1.7173858569670155e-05,
+      "loss": 1.1623,
+      "step": 7342
+    },
+    {
+      "epoch": 2.733836720922386,
+      "grad_norm": 0.16473737359046936,
+      "learning_rate": 1.7173011834917415e-05,
+      "loss": 1.1799,
+      "step": 7343
+    },
+    {
+      "epoch": 2.734209026072995,
+      "grad_norm": 0.16188117861747742,
+      "learning_rate": 1.7172164994219738e-05,
+      "loss": 1.1787,
+      "step": 7344
+    },
+    {
+      "epoch": 2.734581331223604,
+      "grad_norm": 0.17341145873069763,
+      "learning_rate": 1.7171318047589637e-05,
+      "loss": 1.178,
+      "step": 7345
+    },
+    {
+      "epoch": 2.7349536363742133,
+      "grad_norm": 0.1620875597000122,
+      "learning_rate": 1.7170470995039618e-05,
+      "loss": 1.1842,
+      "step": 7346
+    },
+    {
+      "epoch": 2.735325941524822,
+      "grad_norm": 0.16189490258693695,
+      "learning_rate": 1.716962383658219e-05,
+      "loss": 1.1682,
+      "step": 7347
+    },
+    {
+      "epoch": 2.7356982466754314,
+      "grad_norm": 0.16094739735126495,
+      "learning_rate": 1.716877657222987e-05,
+      "loss": 1.173,
+      "step": 7348
+    },
+    {
+      "epoch": 2.7360705518260406,
+      "grad_norm": 0.16400355100631714,
+      "learning_rate": 1.7167929201995167e-05,
+      "loss": 1.1718,
+      "step": 7349
+    },
+    {
+      "epoch": 2.7364428569766495,
+      "grad_norm": 0.1662655919790268,
+      "learning_rate": 1.7167081725890602e-05,
+      "loss": 1.1556,
+      "step": 7350
+    },
+    {
+      "epoch": 2.7368151621272583,
+      "grad_norm": 0.15632013976573944,
+      "learning_rate": 1.716623414392869e-05,
+      "loss": 1.1616,
+      "step": 7351
+    },
+    {
+      "epoch": 2.7371874672778675,
+      "grad_norm": 0.15825137495994568,
+      "learning_rate": 1.7165386456121948e-05,
+      "loss": 1.1631,
+      "step": 7352
+    },
+    {
+      "epoch": 2.737559772428477,
+      "grad_norm": 0.16320408880710602,
+      "learning_rate": 1.71645386624829e-05,
+      "loss": 1.1664,
+      "step": 7353
+    },
+    {
+      "epoch": 2.7379320775790856,
+      "grad_norm": 0.17327001690864563,
+      "learning_rate": 1.7163690763024063e-05,
+      "loss": 1.1686,
+      "step": 7354
+    },
+    {
+      "epoch": 2.738304382729695,
+      "grad_norm": 0.1668155938386917,
+      "learning_rate": 1.7162842757757964e-05,
+      "loss": 1.1665,
+      "step": 7355
+    },
+    {
+      "epoch": 2.7386766878803037,
+      "grad_norm": 0.15803512930870056,
+      "learning_rate": 1.716199464669713e-05,
+      "loss": 1.1703,
+      "step": 7356
+    },
+    {
+      "epoch": 2.739048993030913,
+      "grad_norm": 0.15973332524299622,
+      "learning_rate": 1.716114642985408e-05,
+      "loss": 1.1632,
+      "step": 7357
+    },
+    {
+      "epoch": 2.7394212981815222,
+      "grad_norm": 0.17077676951885223,
+      "learning_rate": 1.7160298107241347e-05,
+      "loss": 1.1578,
+      "step": 7358
+    },
+    {
+      "epoch": 2.739793603332131,
+      "grad_norm": 0.17319710552692413,
+      "learning_rate": 1.7159449678871463e-05,
+      "loss": 1.1749,
+      "step": 7359
+    },
+    {
+      "epoch": 2.7401659084827403,
+      "grad_norm": 0.16504226624965668,
+      "learning_rate": 1.7158601144756953e-05,
+      "loss": 1.1835,
+      "step": 7360
+    },
+    {
+      "epoch": 2.740538213633349,
+      "grad_norm": 0.1656225025653839,
+      "learning_rate": 1.715775250491036e-05,
+      "loss": 1.1673,
+      "step": 7361
+    },
+    {
+      "epoch": 2.7409105187839584,
+      "grad_norm": 0.16402550041675568,
+      "learning_rate": 1.7156903759344207e-05,
+      "loss": 1.1798,
+      "step": 7362
+    },
+    {
+      "epoch": 2.7412828239345672,
+      "grad_norm": 0.1609089970588684,
+      "learning_rate": 1.715605490807103e-05,
+      "loss": 1.1661,
+      "step": 7363
+    },
+    {
+      "epoch": 2.7416551290851765,
+      "grad_norm": 0.1700262427330017,
+      "learning_rate": 1.7155205951103378e-05,
+      "loss": 1.1807,
+      "step": 7364
+    },
+    {
+      "epoch": 2.7420274342357853,
+      "grad_norm": 0.16744990646839142,
+      "learning_rate": 1.715435688845378e-05,
+      "loss": 1.1824,
+      "step": 7365
+    },
+    {
+      "epoch": 2.7423997393863946,
+      "grad_norm": 0.1665429025888443,
+      "learning_rate": 1.715350772013478e-05,
+      "loss": 1.1701,
+      "step": 7366
+    },
+    {
+      "epoch": 2.742772044537004,
+      "grad_norm": 0.16256341338157654,
+      "learning_rate": 1.7152658446158926e-05,
+      "loss": 1.175,
+      "step": 7367
+    },
+    {
+      "epoch": 2.7431443496876127,
+      "grad_norm": 0.16586780548095703,
+      "learning_rate": 1.7151809066538746e-05,
+      "loss": 1.1882,
+      "step": 7368
+    },
+    {
+      "epoch": 2.743516654838222,
+      "grad_norm": 0.1687692552804947,
+      "learning_rate": 1.7150959581286802e-05,
+      "loss": 1.1726,
+      "step": 7369
+    },
+    {
+      "epoch": 2.7438889599888308,
+      "grad_norm": 0.1655583381652832,
+      "learning_rate": 1.715010999041563e-05,
+      "loss": 1.1742,
+      "step": 7370
+    },
+    {
+      "epoch": 2.74426126513944,
+      "grad_norm": 0.1582690328359604,
+      "learning_rate": 1.7149260293937782e-05,
+      "loss": 1.1841,
+      "step": 7371
+    },
+    {
+      "epoch": 2.744633570290049,
+      "grad_norm": 0.1673700511455536,
+      "learning_rate": 1.714841049186581e-05,
+      "loss": 1.1825,
+      "step": 7372
+    },
+    {
+      "epoch": 2.745005875440658,
+      "grad_norm": 0.1654442995786667,
+      "learning_rate": 1.7147560584212263e-05,
+      "loss": 1.1678,
+      "step": 7373
+    },
+    {
+      "epoch": 2.745378180591267,
+      "grad_norm": 0.16653363406658173,
+      "learning_rate": 1.7146710570989698e-05,
+      "loss": 1.1749,
+      "step": 7374
+    },
+    {
+      "epoch": 2.745750485741876,
+      "grad_norm": 0.15975458920001984,
+      "learning_rate": 1.7145860452210662e-05,
+      "loss": 1.18,
+      "step": 7375
+    },
+    {
+      "epoch": 2.7461227908924855,
+      "grad_norm": 0.16379772126674652,
+      "learning_rate": 1.7145010227887716e-05,
+      "loss": 1.1782,
+      "step": 7376
+    },
+    {
+      "epoch": 2.7464950960430943,
+      "grad_norm": 0.16204066574573517,
+      "learning_rate": 1.714415989803342e-05,
+      "loss": 1.1681,
+      "step": 7377
+    },
+    {
+      "epoch": 2.7468674011937035,
+      "grad_norm": 0.16692563891410828,
+      "learning_rate": 1.7143309462660326e-05,
+      "loss": 1.172,
+      "step": 7378
+    },
+    {
+      "epoch": 2.7472397063443124,
+      "grad_norm": 0.16398832201957703,
+      "learning_rate": 1.7142458921781002e-05,
+      "loss": 1.1752,
+      "step": 7379
+    },
+    {
+      "epoch": 2.7476120114949216,
+      "grad_norm": 0.16204705834388733,
+      "learning_rate": 1.714160827540801e-05,
+      "loss": 1.1742,
+      "step": 7380
+    },
+    {
+      "epoch": 2.7479843166455304,
+      "grad_norm": 0.16847161948680878,
+      "learning_rate": 1.7140757523553907e-05,
+      "loss": 1.1765,
+      "step": 7381
+    },
+    {
+      "epoch": 2.7483566217961397,
+      "grad_norm": 0.16786617040634155,
+      "learning_rate": 1.713990666623127e-05,
+      "loss": 1.1632,
+      "step": 7382
+    },
+    {
+      "epoch": 2.7487289269467485,
+      "grad_norm": 0.15807227790355682,
+      "learning_rate": 1.7139055703452653e-05,
+      "loss": 1.1683,
+      "step": 7383
+    },
+    {
+      "epoch": 2.749101232097358,
+      "grad_norm": 0.1643909215927124,
+      "learning_rate": 1.7138204635230637e-05,
+      "loss": 1.1715,
+      "step": 7384
+    },
+    {
+      "epoch": 2.749473537247967,
+      "grad_norm": 0.15912140905857086,
+      "learning_rate": 1.7137353461577785e-05,
+      "loss": 1.179,
+      "step": 7385
+    },
+    {
+      "epoch": 2.749845842398576,
+      "grad_norm": 0.16151286661624908,
+      "learning_rate": 1.713650218250667e-05,
+      "loss": 1.1741,
+      "step": 7386
+    },
+    {
+      "epoch": 2.750218147549185,
+      "grad_norm": 0.1597239375114441,
+      "learning_rate": 1.7135650798029864e-05,
+      "loss": 1.16,
+      "step": 7387
+    },
+    {
+      "epoch": 2.750590452699794,
+      "grad_norm": 0.16162365674972534,
+      "learning_rate": 1.7134799308159946e-05,
+      "loss": 1.1671,
+      "step": 7388
+    },
+    {
+      "epoch": 2.7509627578504032,
+      "grad_norm": 0.16579143702983856,
+      "learning_rate": 1.7133947712909485e-05,
+      "loss": 1.1645,
+      "step": 7389
+    },
+    {
+      "epoch": 2.751335063001012,
+      "grad_norm": 0.1629938781261444,
+      "learning_rate": 1.7133096012291067e-05,
+      "loss": 1.1764,
+      "step": 7390
+    },
+    {
+      "epoch": 2.7517073681516213,
+      "grad_norm": 0.15928851068019867,
+      "learning_rate": 1.7132244206317272e-05,
+      "loss": 1.1714,
+      "step": 7391
+    },
+    {
+      "epoch": 2.75207967330223,
+      "grad_norm": 0.16716843843460083,
+      "learning_rate": 1.7131392295000676e-05,
+      "loss": 1.1764,
+      "step": 7392
+    },
+    {
+      "epoch": 2.7524519784528394,
+      "grad_norm": 0.16544127464294434,
+      "learning_rate": 1.713054027835386e-05,
+      "loss": 1.1731,
+      "step": 7393
+    },
+    {
+      "epoch": 2.7528242836034487,
+      "grad_norm": 0.1620912104845047,
+      "learning_rate": 1.7129688156389414e-05,
+      "loss": 1.1723,
+      "step": 7394
+    },
+    {
+      "epoch": 2.7531965887540575,
+      "grad_norm": 0.1651720106601715,
+      "learning_rate": 1.7128835929119923e-05,
+      "loss": 1.1602,
+      "step": 7395
+    },
+    {
+      "epoch": 2.7535688939046667,
+      "grad_norm": 0.16467827558517456,
+      "learning_rate": 1.712798359655797e-05,
+      "loss": 1.1764,
+      "step": 7396
+    },
+    {
+      "epoch": 2.7539411990552756,
+      "grad_norm": 0.15525712072849274,
+      "learning_rate": 1.7127131158716145e-05,
+      "loss": 1.1681,
+      "step": 7397
+    },
+    {
+      "epoch": 2.754313504205885,
+      "grad_norm": 0.1612330675125122,
+      "learning_rate": 1.7126278615607045e-05,
+      "loss": 1.167,
+      "step": 7398
+    },
+    {
+      "epoch": 2.7546858093564937,
+      "grad_norm": 0.16370365023612976,
+      "learning_rate": 1.7125425967243252e-05,
+      "loss": 1.1789,
+      "step": 7399
+    },
+    {
+      "epoch": 2.755058114507103,
+      "grad_norm": 0.16268078982830048,
+      "learning_rate": 1.7124573213637367e-05,
+      "loss": 1.173,
+      "step": 7400
+    },
+    {
+      "epoch": 2.7554304196577117,
+      "grad_norm": 0.16411568224430084,
+      "learning_rate": 1.7123720354801984e-05,
+      "loss": 1.1595,
+      "step": 7401
+    },
+    {
+      "epoch": 2.755802724808321,
+      "grad_norm": 0.16271185874938965,
+      "learning_rate": 1.7122867390749697e-05,
+      "loss": 1.1676,
+      "step": 7402
+    },
+    {
+      "epoch": 2.7561750299589303,
+      "grad_norm": 0.16013269126415253,
+      "learning_rate": 1.7122014321493105e-05,
+      "loss": 1.1689,
+      "step": 7403
+    },
+    {
+      "epoch": 2.756547335109539,
+      "grad_norm": 0.168240487575531,
+      "learning_rate": 1.7121161147044813e-05,
+      "loss": 1.1712,
+      "step": 7404
+    },
+    {
+      "epoch": 2.7569196402601484,
+      "grad_norm": 0.16075722873210907,
+      "learning_rate": 1.7120307867417414e-05,
+      "loss": 1.1624,
+      "step": 7405
+    },
+    {
+      "epoch": 2.757291945410757,
+      "grad_norm": 0.16558417677879333,
+      "learning_rate": 1.7119454482623515e-05,
+      "loss": 1.1772,
+      "step": 7406
+    },
+    {
+      "epoch": 2.7576642505613664,
+      "grad_norm": 0.1621825397014618,
+      "learning_rate": 1.7118600992675718e-05,
+      "loss": 1.1814,
+      "step": 7407
+    },
+    {
+      "epoch": 2.7580365557119757,
+      "grad_norm": 0.16692297160625458,
+      "learning_rate": 1.7117747397586636e-05,
+      "loss": 1.1804,
+      "step": 7408
+    },
+    {
+      "epoch": 2.7584088608625845,
+      "grad_norm": 0.15905342996120453,
+      "learning_rate": 1.7116893697368866e-05,
+      "loss": 1.171,
+      "step": 7409
+    },
+    {
+      "epoch": 2.7587811660131933,
+      "grad_norm": 0.16342061758041382,
+      "learning_rate": 1.7116039892035025e-05,
+      "loss": 1.1518,
+      "step": 7410
+    },
+    {
+      "epoch": 2.7591534711638026,
+      "grad_norm": 0.161211296916008,
+      "learning_rate": 1.7115185981597725e-05,
+      "loss": 1.1736,
+      "step": 7411
+    },
+    {
+      "epoch": 2.759525776314412,
+      "grad_norm": 0.16845087707042694,
+      "learning_rate": 1.7114331966069572e-05,
+      "loss": 1.1743,
+      "step": 7412
+    },
+    {
+      "epoch": 2.7598980814650207,
+      "grad_norm": 0.16681751608848572,
+      "learning_rate": 1.7113477845463177e-05,
+      "loss": 1.1663,
+      "step": 7413
+    },
+    {
+      "epoch": 2.76027038661563,
+      "grad_norm": 0.16370663046836853,
+      "learning_rate": 1.7112623619791167e-05,
+      "loss": 1.1643,
+      "step": 7414
+    },
+    {
+      "epoch": 2.7606426917662388,
+      "grad_norm": 0.16601529717445374,
+      "learning_rate": 1.711176928906615e-05,
+      "loss": 1.1716,
+      "step": 7415
+    },
+    {
+      "epoch": 2.761014996916848,
+      "grad_norm": 0.1686096489429474,
+      "learning_rate": 1.7110914853300748e-05,
+      "loss": 1.1626,
+      "step": 7416
+    },
+    {
+      "epoch": 2.7613873020674573,
+      "grad_norm": 0.1649993360042572,
+      "learning_rate": 1.711006031250758e-05,
+      "loss": 1.1706,
+      "step": 7417
+    },
+    {
+      "epoch": 2.761759607218066,
+      "grad_norm": 0.16518589854240417,
+      "learning_rate": 1.710920566669927e-05,
+      "loss": 1.1622,
+      "step": 7418
+    },
+    {
+      "epoch": 2.762131912368675,
+      "grad_norm": 0.1586756408214569,
+      "learning_rate": 1.7108350915888432e-05,
+      "loss": 1.1615,
+      "step": 7419
+    },
+    {
+      "epoch": 2.762504217519284,
+      "grad_norm": 0.16265276074409485,
+      "learning_rate": 1.71074960600877e-05,
+      "loss": 1.1762,
+      "step": 7420
+    },
+    {
+      "epoch": 2.7628765226698935,
+      "grad_norm": 0.16679120063781738,
+      "learning_rate": 1.71066410993097e-05,
+      "loss": 1.1838,
+      "step": 7421
+    },
+    {
+      "epoch": 2.7632488278205023,
+      "grad_norm": 0.16883881390094757,
+      "learning_rate": 1.7105786033567055e-05,
+      "loss": 1.1721,
+      "step": 7422
+    },
+    {
+      "epoch": 2.7636211329711116,
+      "grad_norm": 0.15802286565303802,
+      "learning_rate": 1.7104930862872394e-05,
+      "loss": 1.1737,
+      "step": 7423
+    },
+    {
+      "epoch": 2.7639934381217204,
+      "grad_norm": 0.1627608686685562,
+      "learning_rate": 1.7104075587238353e-05,
+      "loss": 1.1678,
+      "step": 7424
+    },
+    {
+      "epoch": 2.7643657432723296,
+      "grad_norm": 0.1643580198287964,
+      "learning_rate": 1.710322020667756e-05,
+      "loss": 1.1746,
+      "step": 7425
+    },
+    {
+      "epoch": 2.764738048422939,
+      "grad_norm": 0.16312402486801147,
+      "learning_rate": 1.7102364721202655e-05,
+      "loss": 1.1812,
+      "step": 7426
+    },
+    {
+      "epoch": 2.7651103535735477,
+      "grad_norm": 0.15832361578941345,
+      "learning_rate": 1.710150913082626e-05,
+      "loss": 1.1886,
+      "step": 7427
+    },
+    {
+      "epoch": 2.7654826587241566,
+      "grad_norm": 0.16072019934654236,
+      "learning_rate": 1.7100653435561027e-05,
+      "loss": 1.1688,
+      "step": 7428
+    },
+    {
+      "epoch": 2.765854963874766,
+      "grad_norm": 0.16109511256217957,
+      "learning_rate": 1.7099797635419587e-05,
+      "loss": 1.184,
+      "step": 7429
+    },
+    {
+      "epoch": 2.766227269025375,
+      "grad_norm": 0.16431210935115814,
+      "learning_rate": 1.709894173041458e-05,
+      "loss": 1.1763,
+      "step": 7430
+    },
+    {
+      "epoch": 2.766599574175984,
+      "grad_norm": 0.16304095089435577,
+      "learning_rate": 1.7098085720558653e-05,
+      "loss": 1.1671,
+      "step": 7431
+    },
+    {
+      "epoch": 2.766971879326593,
+      "grad_norm": 0.16046518087387085,
+      "learning_rate": 1.709722960586444e-05,
+      "loss": 1.1659,
+      "step": 7432
+    },
+    {
+      "epoch": 2.767344184477202,
+      "grad_norm": 0.1624547243118286,
+      "learning_rate": 1.7096373386344596e-05,
+      "loss": 1.1645,
+      "step": 7433
+    },
+    {
+      "epoch": 2.7677164896278112,
+      "grad_norm": 0.1661730259656906,
+      "learning_rate": 1.709551706201176e-05,
+      "loss": 1.171,
+      "step": 7434
+    },
+    {
+      "epoch": 2.7680887947784205,
+      "grad_norm": 0.15459467470645905,
+      "learning_rate": 1.709466063287858e-05,
+      "loss": 1.1533,
+      "step": 7435
+    },
+    {
+      "epoch": 2.7684610999290293,
+      "grad_norm": 0.16066570580005646,
+      "learning_rate": 1.709380409895771e-05,
+      "loss": 1.1714,
+      "step": 7436
+    },
+    {
+      "epoch": 2.768833405079638,
+      "grad_norm": 0.1591673642396927,
+      "learning_rate": 1.7092947460261802e-05,
+      "loss": 1.1589,
+      "step": 7437
+    },
+    {
+      "epoch": 2.7692057102302474,
+      "grad_norm": 0.1647397130727768,
+      "learning_rate": 1.7092090716803503e-05,
+      "loss": 1.1783,
+      "step": 7438
+    },
+    {
+      "epoch": 2.7695780153808567,
+      "grad_norm": 0.16545167565345764,
+      "learning_rate": 1.7091233868595465e-05,
+      "loss": 1.1808,
+      "step": 7439
+    },
+    {
+      "epoch": 2.7699503205314655,
+      "grad_norm": 0.16354137659072876,
+      "learning_rate": 1.7090376915650354e-05,
+      "loss": 1.1713,
+      "step": 7440
+    },
+    {
+      "epoch": 2.7703226256820748,
+      "grad_norm": 0.1619860678911209,
+      "learning_rate": 1.708951985798082e-05,
+      "loss": 1.1803,
+      "step": 7441
+    },
+    {
+      "epoch": 2.7706949308326836,
+      "grad_norm": 0.1611848771572113,
+      "learning_rate": 1.7088662695599517e-05,
+      "loss": 1.1658,
+      "step": 7442
+    },
+    {
+      "epoch": 2.771067235983293,
+      "grad_norm": 0.1581423580646515,
+      "learning_rate": 1.7087805428519114e-05,
+      "loss": 1.1639,
+      "step": 7443
+    },
+    {
+      "epoch": 2.771439541133902,
+      "grad_norm": 0.1612246185541153,
+      "learning_rate": 1.708694805675227e-05,
+      "loss": 1.1588,
+      "step": 7444
+    },
+    {
+      "epoch": 2.771811846284511,
+      "grad_norm": 0.16077731549739838,
+      "learning_rate": 1.708609058031165e-05,
+      "loss": 1.1788,
+      "step": 7445
+    },
+    {
+      "epoch": 2.7721841514351198,
+      "grad_norm": 0.16849185526371002,
+      "learning_rate": 1.7085232999209915e-05,
+      "loss": 1.1797,
+      "step": 7446
+    },
+    {
+      "epoch": 2.772556456585729,
+      "grad_norm": 0.16407588124275208,
+      "learning_rate": 1.7084375313459735e-05,
+      "loss": 1.1638,
+      "step": 7447
+    },
+    {
+      "epoch": 2.7729287617363383,
+      "grad_norm": 0.1661745011806488,
+      "learning_rate": 1.7083517523073775e-05,
+      "loss": 1.1695,
+      "step": 7448
+    },
+    {
+      "epoch": 2.773301066886947,
+      "grad_norm": 0.1609290987253189,
+      "learning_rate": 1.7082659628064704e-05,
+      "loss": 1.1596,
+      "step": 7449
+    },
+    {
+      "epoch": 2.7736733720375564,
+      "grad_norm": 0.15832340717315674,
+      "learning_rate": 1.7081801628445195e-05,
+      "loss": 1.1536,
+      "step": 7450
+    },
+    {
+      "epoch": 2.774045677188165,
+      "grad_norm": 0.16378222405910492,
+      "learning_rate": 1.708094352422792e-05,
+      "loss": 1.1704,
+      "step": 7451
+    },
+    {
+      "epoch": 2.7744179823387745,
+      "grad_norm": 0.16052454710006714,
+      "learning_rate": 1.7080085315425557e-05,
+      "loss": 1.1643,
+      "step": 7452
+    },
+    {
+      "epoch": 2.7747902874893837,
+      "grad_norm": 0.16648857295513153,
+      "learning_rate": 1.7079227002050776e-05,
+      "loss": 1.1766,
+      "step": 7453
+    },
+    {
+      "epoch": 2.7751625926399925,
+      "grad_norm": 0.16049505770206451,
+      "learning_rate": 1.7078368584116256e-05,
+      "loss": 1.172,
+      "step": 7454
+    },
+    {
+      "epoch": 2.7755348977906014,
+      "grad_norm": 0.1630232185125351,
+      "learning_rate": 1.7077510061634675e-05,
+      "loss": 1.1696,
+      "step": 7455
+    },
+    {
+      "epoch": 2.7759072029412106,
+      "grad_norm": 0.15958517789840698,
+      "learning_rate": 1.707665143461872e-05,
+      "loss": 1.1703,
+      "step": 7456
+    },
+    {
+      "epoch": 2.77627950809182,
+      "grad_norm": 0.16538631916046143,
+      "learning_rate": 1.707579270308106e-05,
+      "loss": 1.1695,
+      "step": 7457
+    },
+    {
+      "epoch": 2.7766518132424287,
+      "grad_norm": 0.16825279593467712,
+      "learning_rate": 1.7074933867034392e-05,
+      "loss": 1.1668,
+      "step": 7458
+    },
+    {
+      "epoch": 2.777024118393038,
+      "grad_norm": 0.15979918837547302,
+      "learning_rate": 1.7074074926491392e-05,
+      "loss": 1.1692,
+      "step": 7459
+    },
+    {
+      "epoch": 2.777396423543647,
+      "grad_norm": 0.1648228019475937,
+      "learning_rate": 1.707321588146475e-05,
+      "loss": 1.1818,
+      "step": 7460
+    },
+    {
+      "epoch": 2.777768728694256,
+      "grad_norm": 0.16858553886413574,
+      "learning_rate": 1.7072356731967152e-05,
+      "loss": 1.1717,
+      "step": 7461
+    },
+    {
+      "epoch": 2.7781410338448653,
+      "grad_norm": 0.16324035823345184,
+      "learning_rate": 1.707149747801129e-05,
+      "loss": 1.1706,
+      "step": 7462
+    },
+    {
+      "epoch": 2.778513338995474,
+      "grad_norm": 0.16237075626850128,
+      "learning_rate": 1.7070638119609854e-05,
+      "loss": 1.1689,
+      "step": 7463
+    },
+    {
+      "epoch": 2.778885644146083,
+      "grad_norm": 0.1649467796087265,
+      "learning_rate": 1.706977865677554e-05,
+      "loss": 1.1741,
+      "step": 7464
+    },
+    {
+      "epoch": 2.7792579492966922,
+      "grad_norm": 0.16637516021728516,
+      "learning_rate": 1.7068919089521032e-05,
+      "loss": 1.1881,
+      "step": 7465
+    },
+    {
+      "epoch": 2.7796302544473015,
+      "grad_norm": 0.1721596121788025,
+      "learning_rate": 1.7068059417859037e-05,
+      "loss": 1.1669,
+      "step": 7466
+    },
+    {
+      "epoch": 2.7800025595979103,
+      "grad_norm": 0.16336818039417267,
+      "learning_rate": 1.7067199641802247e-05,
+      "loss": 1.1616,
+      "step": 7467
+    },
+    {
+      "epoch": 2.7803748647485196,
+      "grad_norm": 0.1582047939300537,
+      "learning_rate": 1.7066339761363364e-05,
+      "loss": 1.1887,
+      "step": 7468
+    },
+    {
+      "epoch": 2.7807471698991284,
+      "grad_norm": 0.16283096373081207,
+      "learning_rate": 1.7065479776555083e-05,
+      "loss": 1.1633,
+      "step": 7469
+    },
+    {
+      "epoch": 2.7811194750497377,
+      "grad_norm": 0.16766688227653503,
+      "learning_rate": 1.7064619687390108e-05,
+      "loss": 1.1501,
+      "step": 7470
+    },
+    {
+      "epoch": 2.781491780200347,
+      "grad_norm": 0.16073793172836304,
+      "learning_rate": 1.706375949388115e-05,
+      "loss": 1.166,
+      "step": 7471
+    },
+    {
+      "epoch": 2.7818640853509558,
+      "grad_norm": 0.1599031388759613,
+      "learning_rate": 1.7062899196040903e-05,
+      "loss": 1.1751,
+      "step": 7472
+    },
+    {
+      "epoch": 2.7822363905015646,
+      "grad_norm": 0.16344262659549713,
+      "learning_rate": 1.7062038793882078e-05,
+      "loss": 1.1671,
+      "step": 7473
+    },
+    {
+      "epoch": 2.782608695652174,
+      "grad_norm": 0.16053412854671478,
+      "learning_rate": 1.7061178287417383e-05,
+      "loss": 1.1696,
+      "step": 7474
+    },
+    {
+      "epoch": 2.782981000802783,
+      "grad_norm": 0.16248364746570587,
+      "learning_rate": 1.706031767665953e-05,
+      "loss": 1.1705,
+      "step": 7475
+    },
+    {
+      "epoch": 2.783353305953392,
+      "grad_norm": 0.16297762095928192,
+      "learning_rate": 1.7059456961621226e-05,
+      "loss": 1.1607,
+      "step": 7476
+    },
+    {
+      "epoch": 2.783725611104001,
+      "grad_norm": 0.16271525621414185,
+      "learning_rate": 1.7058596142315185e-05,
+      "loss": 1.1568,
+      "step": 7477
+    },
+    {
+      "epoch": 2.78409791625461,
+      "grad_norm": 0.1527474820613861,
+      "learning_rate": 1.7057735218754126e-05,
+      "loss": 1.1806,
+      "step": 7478
+    },
+    {
+      "epoch": 2.7844702214052193,
+      "grad_norm": 0.15945225954055786,
+      "learning_rate": 1.705687419095076e-05,
+      "loss": 1.1747,
+      "step": 7479
+    },
+    {
+      "epoch": 2.7848425265558285,
+      "grad_norm": 0.16087423264980316,
+      "learning_rate": 1.7056013058917802e-05,
+      "loss": 1.1723,
+      "step": 7480
+    },
+    {
+      "epoch": 2.7852148317064374,
+      "grad_norm": 0.16305401921272278,
+      "learning_rate": 1.7055151822667975e-05,
+      "loss": 1.1841,
+      "step": 7481
+    },
+    {
+      "epoch": 2.7855871368570466,
+      "grad_norm": 0.16334961354732513,
+      "learning_rate": 1.7054290482213996e-05,
+      "loss": 1.1602,
+      "step": 7482
+    },
+    {
+      "epoch": 2.7859594420076554,
+      "grad_norm": 0.16361406445503235,
+      "learning_rate": 1.7053429037568596e-05,
+      "loss": 1.1759,
+      "step": 7483
+    },
+    {
+      "epoch": 2.7863317471582647,
+      "grad_norm": 0.1611858606338501,
+      "learning_rate": 1.7052567488744485e-05,
+      "loss": 1.1622,
+      "step": 7484
+    },
+    {
+      "epoch": 2.7867040523088735,
+      "grad_norm": 0.1641787737607956,
+      "learning_rate": 1.7051705835754394e-05,
+      "loss": 1.1794,
+      "step": 7485
+    },
+    {
+      "epoch": 2.787076357459483,
+      "grad_norm": 0.15844056010246277,
+      "learning_rate": 1.7050844078611058e-05,
+      "loss": 1.1711,
+      "step": 7486
+    },
+    {
+      "epoch": 2.7874486626100916,
+      "grad_norm": 0.15609781444072723,
+      "learning_rate": 1.7049982217327192e-05,
+      "loss": 1.1585,
+      "step": 7487
+    },
+    {
+      "epoch": 2.787820967760701,
+      "grad_norm": 0.16171182692050934,
+      "learning_rate": 1.704912025191553e-05,
+      "loss": 1.1798,
+      "step": 7488
+    },
+    {
+      "epoch": 2.78819327291131,
+      "grad_norm": 0.16126160323619843,
+      "learning_rate": 1.704825818238881e-05,
+      "loss": 1.158,
+      "step": 7489
+    },
+    {
+      "epoch": 2.788565578061919,
+      "grad_norm": 0.16148677468299866,
+      "learning_rate": 1.7047396008759755e-05,
+      "loss": 1.168,
+      "step": 7490
+    },
+    {
+      "epoch": 2.7889378832125282,
+      "grad_norm": 0.1633322536945343,
+      "learning_rate": 1.7046533731041103e-05,
+      "loss": 1.1784,
+      "step": 7491
+    },
+    {
+      "epoch": 2.789310188363137,
+      "grad_norm": 0.16195450723171234,
+      "learning_rate": 1.7045671349245588e-05,
+      "loss": 1.1889,
+      "step": 7492
+    },
+    {
+      "epoch": 2.7896824935137463,
+      "grad_norm": 0.16481108963489532,
+      "learning_rate": 1.7044808863385953e-05,
+      "loss": 1.1693,
+      "step": 7493
+    },
+    {
+      "epoch": 2.790054798664355,
+      "grad_norm": 0.16438551247119904,
+      "learning_rate": 1.7043946273474935e-05,
+      "loss": 1.1582,
+      "step": 7494
+    },
+    {
+      "epoch": 2.7904271038149644,
+      "grad_norm": 0.16081902384757996,
+      "learning_rate": 1.704308357952527e-05,
+      "loss": 1.1666,
+      "step": 7495
+    },
+    {
+      "epoch": 2.790799408965573,
+      "grad_norm": 0.16997969150543213,
+      "learning_rate": 1.7042220781549703e-05,
+      "loss": 1.1741,
+      "step": 7496
+    },
+    {
+      "epoch": 2.7911717141161825,
+      "grad_norm": 0.16140542924404144,
+      "learning_rate": 1.7041357879560972e-05,
+      "loss": 1.171,
+      "step": 7497
+    },
+    {
+      "epoch": 2.7915440192667917,
+      "grad_norm": 0.1633981466293335,
+      "learning_rate": 1.7040494873571832e-05,
+      "loss": 1.1596,
+      "step": 7498
+    },
+    {
+      "epoch": 2.7919163244174006,
+      "grad_norm": 0.17155370116233826,
+      "learning_rate": 1.7039631763595025e-05,
+      "loss": 1.1825,
+      "step": 7499
+    },
+    {
+      "epoch": 2.79228862956801,
+      "grad_norm": 0.17078445851802826,
+      "learning_rate": 1.7038768549643297e-05,
+      "loss": 1.175,
+      "step": 7500
+    },
+    {
+      "epoch": 2.79228862956801,
+      "eval_loss": 1.2953369617462158,
+      "eval_runtime": 16.7431,
+      "eval_samples_per_second": 103.565,
+      "eval_steps_per_second": 5.196,
+      "step": 7500
+    },
+    {
+      "epoch": 2.7926609347186186,
+      "grad_norm": 0.16718538105487823,
+      "learning_rate": 1.7037905231729402e-05,
+      "loss": 1.1802,
+      "step": 7501
+    },
+    {
+      "epoch": 2.793033239869228,
+      "grad_norm": 0.16334514319896698,
+      "learning_rate": 1.7037041809866085e-05,
+      "loss": 1.1619,
+      "step": 7502
+    },
+    {
+      "epoch": 2.7934055450198367,
+      "grad_norm": 0.1652374416589737,
+      "learning_rate": 1.7036178284066103e-05,
+      "loss": 1.1584,
+      "step": 7503
+    },
+    {
+      "epoch": 2.793777850170446,
+      "grad_norm": 0.17066913843154907,
+      "learning_rate": 1.703531465434221e-05,
+      "loss": 1.1792,
+      "step": 7504
+    },
+    {
+      "epoch": 2.794150155321055,
+      "grad_norm": 0.1763291358947754,
+      "learning_rate": 1.7034450920707162e-05,
+      "loss": 1.1846,
+      "step": 7505
+    },
+    {
+      "epoch": 2.794522460471664,
+      "grad_norm": 0.16274365782737732,
+      "learning_rate": 1.7033587083173713e-05,
+      "loss": 1.1639,
+      "step": 7506
+    },
+    {
+      "epoch": 2.7948947656222733,
+      "grad_norm": 0.16275723278522491,
+      "learning_rate": 1.7032723141754626e-05,
+      "loss": 1.1731,
+      "step": 7507
+    },
+    {
+      "epoch": 2.795267070772882,
+      "grad_norm": 0.16381405293941498,
+      "learning_rate": 1.703185909646266e-05,
+      "loss": 1.1696,
+      "step": 7508
+    },
+    {
+      "epoch": 2.7956393759234914,
+      "grad_norm": 0.1678345650434494,
+      "learning_rate": 1.7030994947310576e-05,
+      "loss": 1.1755,
+      "step": 7509
+    },
+    {
+      "epoch": 2.7960116810741003,
+      "grad_norm": 0.16982942819595337,
+      "learning_rate": 1.703013069431114e-05,
+      "loss": 1.1687,
+      "step": 7510
+    },
+    {
+      "epoch": 2.7963839862247095,
+      "grad_norm": 0.16046252846717834,
+      "learning_rate": 1.7029266337477106e-05,
+      "loss": 1.1686,
+      "step": 7511
+    },
+    {
+      "epoch": 2.7967562913753183,
+      "grad_norm": 0.16178347170352936,
+      "learning_rate": 1.7028401876821257e-05,
+      "loss": 1.1689,
+      "step": 7512
+    },
+    {
+      "epoch": 2.7971285965259276,
+      "grad_norm": 0.16835175454616547,
+      "learning_rate": 1.7027537312356353e-05,
+      "loss": 1.1482,
+      "step": 7513
+    },
+    {
+      "epoch": 2.7975009016765364,
+      "grad_norm": 0.16404716670513153,
+      "learning_rate": 1.702667264409516e-05,
+      "loss": 1.157,
+      "step": 7514
+    },
+    {
+      "epoch": 2.7978732068271457,
+      "grad_norm": 0.15957726538181305,
+      "learning_rate": 1.7025807872050456e-05,
+      "loss": 1.1719,
+      "step": 7515
+    },
+    {
+      "epoch": 2.798245511977755,
+      "grad_norm": 0.16722844541072845,
+      "learning_rate": 1.702494299623501e-05,
+      "loss": 1.174,
+      "step": 7516
+    },
+    {
+      "epoch": 2.7986178171283638,
+      "grad_norm": 0.16469328105449677,
+      "learning_rate": 1.7024078016661597e-05,
+      "loss": 1.1758,
+      "step": 7517
+    },
+    {
+      "epoch": 2.798990122278973,
+      "grad_norm": 0.16519738733768463,
+      "learning_rate": 1.7023212933342995e-05,
+      "loss": 1.1714,
+      "step": 7518
+    },
+    {
+      "epoch": 2.799362427429582,
+      "grad_norm": 0.16509708762168884,
+      "learning_rate": 1.7022347746291975e-05,
+      "loss": 1.1692,
+      "step": 7519
+    },
+    {
+      "epoch": 2.799734732580191,
+      "grad_norm": 0.15529975295066833,
+      "learning_rate": 1.7021482455521323e-05,
+      "loss": 1.1697,
+      "step": 7520
+    },
+    {
+      "epoch": 2.8001070377308,
+      "grad_norm": 0.16663393378257751,
+      "learning_rate": 1.7020617061043815e-05,
+      "loss": 1.1656,
+      "step": 7521
+    },
+    {
+      "epoch": 2.800479342881409,
+      "grad_norm": 0.16711021959781647,
+      "learning_rate": 1.701975156287223e-05,
+      "loss": 1.1635,
+      "step": 7522
+    },
+    {
+      "epoch": 2.800851648032018,
+      "grad_norm": 0.1636432558298111,
+      "learning_rate": 1.7018885961019356e-05,
+      "loss": 1.1626,
+      "step": 7523
+    },
+    {
+      "epoch": 2.8012239531826273,
+      "grad_norm": 0.16135211288928986,
+      "learning_rate": 1.701802025549798e-05,
+      "loss": 1.1764,
+      "step": 7524
+    },
+    {
+      "epoch": 2.8015962583332366,
+      "grad_norm": 0.161915123462677,
+      "learning_rate": 1.7017154446320882e-05,
+      "loss": 1.1893,
+      "step": 7525
+    },
+    {
+      "epoch": 2.8019685634838454,
+      "grad_norm": 0.1613880842924118,
+      "learning_rate": 1.7016288533500855e-05,
+      "loss": 1.1716,
+      "step": 7526
+    },
+    {
+      "epoch": 2.8023408686344546,
+      "grad_norm": 0.1624586284160614,
+      "learning_rate": 1.7015422517050686e-05,
+      "loss": 1.1683,
+      "step": 7527
+    },
+    {
+      "epoch": 2.8027131737850635,
+      "grad_norm": 0.1543359011411667,
+      "learning_rate": 1.7014556396983168e-05,
+      "loss": 1.1611,
+      "step": 7528
+    },
+    {
+      "epoch": 2.8030854789356727,
+      "grad_norm": 0.16166086494922638,
+      "learning_rate": 1.701369017331109e-05,
+      "loss": 1.1609,
+      "step": 7529
+    },
+    {
+      "epoch": 2.803457784086282,
+      "grad_norm": 0.1620410680770874,
+      "learning_rate": 1.7012823846047252e-05,
+      "loss": 1.1608,
+      "step": 7530
+    },
+    {
+      "epoch": 2.803830089236891,
+      "grad_norm": 0.168722003698349,
+      "learning_rate": 1.7011957415204443e-05,
+      "loss": 1.1798,
+      "step": 7531
+    },
+    {
+      "epoch": 2.8042023943874996,
+      "grad_norm": 0.16052375733852386,
+      "learning_rate": 1.7011090880795463e-05,
+      "loss": 1.1571,
+      "step": 7532
+    },
+    {
+      "epoch": 2.804574699538109,
+      "grad_norm": 0.1640620082616806,
+      "learning_rate": 1.701022424283311e-05,
+      "loss": 1.1875,
+      "step": 7533
+    },
+    {
+      "epoch": 2.804947004688718,
+      "grad_norm": 0.16312499344348907,
+      "learning_rate": 1.7009357501330188e-05,
+      "loss": 1.1748,
+      "step": 7534
+    },
+    {
+      "epoch": 2.805319309839327,
+      "grad_norm": 0.1676671802997589,
+      "learning_rate": 1.7008490656299492e-05,
+      "loss": 1.1347,
+      "step": 7535
+    },
+    {
+      "epoch": 2.8056916149899362,
+      "grad_norm": 0.15957991778850555,
+      "learning_rate": 1.700762370775383e-05,
+      "loss": 1.1705,
+      "step": 7536
+    },
+    {
+      "epoch": 2.806063920140545,
+      "grad_norm": 0.1587969958782196,
+      "learning_rate": 1.700675665570601e-05,
+      "loss": 1.1683,
+      "step": 7537
+    },
+    {
+      "epoch": 2.8064362252911543,
+      "grad_norm": 0.15972483158111572,
+      "learning_rate": 1.7005889500168828e-05,
+      "loss": 1.172,
+      "step": 7538
+    },
+    {
+      "epoch": 2.8068085304417636,
+      "grad_norm": 0.16055303812026978,
+      "learning_rate": 1.70050222411551e-05,
+      "loss": 1.1702,
+      "step": 7539
+    },
+    {
+      "epoch": 2.8071808355923724,
+      "grad_norm": 0.16454675793647766,
+      "learning_rate": 1.7004154878677634e-05,
+      "loss": 1.173,
+      "step": 7540
+    },
+    {
+      "epoch": 2.8075531407429812,
+      "grad_norm": 0.16008242964744568,
+      "learning_rate": 1.7003287412749236e-05,
+      "loss": 1.1688,
+      "step": 7541
+    },
+    {
+      "epoch": 2.8079254458935905,
+      "grad_norm": 0.16654878854751587,
+      "learning_rate": 1.7002419843382724e-05,
+      "loss": 1.1581,
+      "step": 7542
+    },
+    {
+      "epoch": 2.8082977510441998,
+      "grad_norm": 0.15987439453601837,
+      "learning_rate": 1.7001552170590913e-05,
+      "loss": 1.1766,
+      "step": 7543
+    },
+    {
+      "epoch": 2.8086700561948086,
+      "grad_norm": 0.15910540521144867,
+      "learning_rate": 1.7000684394386615e-05,
+      "loss": 1.1669,
+      "step": 7544
+    },
+    {
+      "epoch": 2.809042361345418,
+      "grad_norm": 0.16279643774032593,
+      "learning_rate": 1.6999816514782647e-05,
+      "loss": 1.1597,
+      "step": 7545
+    },
+    {
+      "epoch": 2.8094146664960267,
+      "grad_norm": 0.16432224214076996,
+      "learning_rate": 1.699894853179183e-05,
+      "loss": 1.1671,
+      "step": 7546
+    },
+    {
+      "epoch": 2.809786971646636,
+      "grad_norm": 0.16405612230300903,
+      "learning_rate": 1.699808044542698e-05,
+      "loss": 1.1825,
+      "step": 7547
+    },
+    {
+      "epoch": 2.810159276797245,
+      "grad_norm": 0.16156992316246033,
+      "learning_rate": 1.6997212255700924e-05,
+      "loss": 1.1532,
+      "step": 7548
+    },
+    {
+      "epoch": 2.810531581947854,
+      "grad_norm": 0.1572588086128235,
+      "learning_rate": 1.699634396262648e-05,
+      "loss": 1.1612,
+      "step": 7549
+    },
+    {
+      "epoch": 2.810903887098463,
+      "grad_norm": 0.16491782665252686,
+      "learning_rate": 1.6995475566216475e-05,
+      "loss": 1.1627,
+      "step": 7550
+    },
+    {
+      "epoch": 2.811276192249072,
+      "grad_norm": 0.1590576469898224,
+      "learning_rate": 1.6994607066483735e-05,
+      "loss": 1.159,
+      "step": 7551
+    },
+    {
+      "epoch": 2.8116484973996814,
+      "grad_norm": 0.16059501469135284,
+      "learning_rate": 1.6993738463441087e-05,
+      "loss": 1.1792,
+      "step": 7552
+    },
+    {
+      "epoch": 2.81202080255029,
+      "grad_norm": 0.16398665308952332,
+      "learning_rate": 1.6992869757101362e-05,
+      "loss": 1.1696,
+      "step": 7553
+    },
+    {
+      "epoch": 2.8123931077008995,
+      "grad_norm": 0.16270703077316284,
+      "learning_rate": 1.6992000947477386e-05,
+      "loss": 1.1745,
+      "step": 7554
+    },
+    {
+      "epoch": 2.8127654128515083,
+      "grad_norm": 0.1659878045320511,
+      "learning_rate": 1.6991132034582e-05,
+      "loss": 1.1604,
+      "step": 7555
+    },
+    {
+      "epoch": 2.8131377180021175,
+      "grad_norm": 0.16921178996562958,
+      "learning_rate": 1.699026301842803e-05,
+      "loss": 1.1652,
+      "step": 7556
+    },
+    {
+      "epoch": 2.813510023152727,
+      "grad_norm": 0.16774874925613403,
+      "learning_rate": 1.6989393899028313e-05,
+      "loss": 1.1847,
+      "step": 7557
+    },
+    {
+      "epoch": 2.8138823283033356,
+      "grad_norm": 0.15916696190834045,
+      "learning_rate": 1.698852467639569e-05,
+      "loss": 1.1682,
+      "step": 7558
+    },
+    {
+      "epoch": 2.8142546334539444,
+      "grad_norm": 0.1583160012960434,
+      "learning_rate": 1.6987655350542993e-05,
+      "loss": 1.1811,
+      "step": 7559
+    },
+    {
+      "epoch": 2.8146269386045537,
+      "grad_norm": 0.16362488269805908,
+      "learning_rate": 1.6986785921483068e-05,
+      "loss": 1.1675,
+      "step": 7560
+    },
+    {
+      "epoch": 2.814999243755163,
+      "grad_norm": 0.16386942565441132,
+      "learning_rate": 1.6985916389228746e-05,
+      "loss": 1.173,
+      "step": 7561
+    },
+    {
+      "epoch": 2.815371548905772,
+      "grad_norm": 0.16522067785263062,
+      "learning_rate": 1.6985046753792885e-05,
+      "loss": 1.1626,
+      "step": 7562
+    },
+    {
+      "epoch": 2.815743854056381,
+      "grad_norm": 0.15882590413093567,
+      "learning_rate": 1.698417701518832e-05,
+      "loss": 1.1694,
+      "step": 7563
+    },
+    {
+      "epoch": 2.81611615920699,
+      "grad_norm": 0.1628119945526123,
+      "learning_rate": 1.69833071734279e-05,
+      "loss": 1.1819,
+      "step": 7564
+    },
+    {
+      "epoch": 2.816488464357599,
+      "grad_norm": 0.1661859154701233,
+      "learning_rate": 1.6982437228524468e-05,
+      "loss": 1.1754,
+      "step": 7565
+    },
+    {
+      "epoch": 2.8168607695082084,
+      "grad_norm": 0.16389884054660797,
+      "learning_rate": 1.698156718049088e-05,
+      "loss": 1.1741,
+      "step": 7566
+    },
+    {
+      "epoch": 2.8172330746588172,
+      "grad_norm": 0.16429540514945984,
+      "learning_rate": 1.6980697029339978e-05,
+      "loss": 1.1648,
+      "step": 7567
+    },
+    {
+      "epoch": 2.817605379809426,
+      "grad_norm": 0.16381768882274628,
+      "learning_rate": 1.6979826775084624e-05,
+      "loss": 1.1656,
+      "step": 7568
+    },
+    {
+      "epoch": 2.8179776849600353,
+      "grad_norm": 0.1680193990468979,
+      "learning_rate": 1.6978956417737663e-05,
+      "loss": 1.167,
+      "step": 7569
+    },
+    {
+      "epoch": 2.8183499901106446,
+      "grad_norm": 0.16285154223442078,
+      "learning_rate": 1.6978085957311956e-05,
+      "loss": 1.1723,
+      "step": 7570
+    },
+    {
+      "epoch": 2.8187222952612534,
+      "grad_norm": 0.1649884134531021,
+      "learning_rate": 1.6977215393820357e-05,
+      "loss": 1.1937,
+      "step": 7571
+    },
+    {
+      "epoch": 2.8190946004118627,
+      "grad_norm": 0.1563154011964798,
+      "learning_rate": 1.6976344727275725e-05,
+      "loss": 1.177,
+      "step": 7572
+    },
+    {
+      "epoch": 2.8194669055624715,
+      "grad_norm": 0.1616252064704895,
+      "learning_rate": 1.6975473957690917e-05,
+      "loss": 1.1679,
+      "step": 7573
+    },
+    {
+      "epoch": 2.8198392107130807,
+      "grad_norm": 0.16457171738147736,
+      "learning_rate": 1.6974603085078798e-05,
+      "loss": 1.1668,
+      "step": 7574
+    },
+    {
+      "epoch": 2.82021151586369,
+      "grad_norm": 0.1589018553495407,
+      "learning_rate": 1.697373210945223e-05,
+      "loss": 1.1653,
+      "step": 7575
+    },
+    {
+      "epoch": 2.820583821014299,
+      "grad_norm": 0.16056711971759796,
+      "learning_rate": 1.6972861030824072e-05,
+      "loss": 1.169,
+      "step": 7576
+    },
+    {
+      "epoch": 2.8209561261649077,
+      "grad_norm": 0.1676499992609024,
+      "learning_rate": 1.69719898492072e-05,
+      "loss": 1.1786,
+      "step": 7577
+    },
+    {
+      "epoch": 2.821328431315517,
+      "grad_norm": 0.16967107355594635,
+      "learning_rate": 1.6971118564614473e-05,
+      "loss": 1.158,
+      "step": 7578
+    },
+    {
+      "epoch": 2.821700736466126,
+      "grad_norm": 0.17274503409862518,
+      "learning_rate": 1.697024717705876e-05,
+      "loss": 1.1906,
+      "step": 7579
+    },
+    {
+      "epoch": 2.822073041616735,
+      "grad_norm": 0.16663110256195068,
+      "learning_rate": 1.696937568655294e-05,
+      "loss": 1.17,
+      "step": 7580
+    },
+    {
+      "epoch": 2.8224453467673443,
+      "grad_norm": 0.16743509471416473,
+      "learning_rate": 1.696850409310987e-05,
+      "loss": 1.1783,
+      "step": 7581
+    },
+    {
+      "epoch": 2.822817651917953,
+      "grad_norm": 0.16072916984558105,
+      "learning_rate": 1.6967632396742434e-05,
+      "loss": 1.1684,
+      "step": 7582
+    },
+    {
+      "epoch": 2.8231899570685624,
+      "grad_norm": 0.16104640066623688,
+      "learning_rate": 1.696676059746351e-05,
+      "loss": 1.1607,
+      "step": 7583
+    },
+    {
+      "epoch": 2.8235622622191716,
+      "grad_norm": 0.1597306877374649,
+      "learning_rate": 1.6965888695285965e-05,
+      "loss": 1.1765,
+      "step": 7584
+    },
+    {
+      "epoch": 2.8239345673697804,
+      "grad_norm": 0.16316664218902588,
+      "learning_rate": 1.6965016690222685e-05,
+      "loss": 1.1683,
+      "step": 7585
+    },
+    {
+      "epoch": 2.8243068725203893,
+      "grad_norm": 0.1655420958995819,
+      "learning_rate": 1.696414458228654e-05,
+      "loss": 1.1851,
+      "step": 7586
+    },
+    {
+      "epoch": 2.8246791776709985,
+      "grad_norm": 0.16307014226913452,
+      "learning_rate": 1.696327237149042e-05,
+      "loss": 1.174,
+      "step": 7587
+    },
+    {
+      "epoch": 2.825051482821608,
+      "grad_norm": 0.16215850412845612,
+      "learning_rate": 1.6962400057847202e-05,
+      "loss": 1.1768,
+      "step": 7588
+    },
+    {
+      "epoch": 2.8254237879722166,
+      "grad_norm": 0.16659583151340485,
+      "learning_rate": 1.6961527641369774e-05,
+      "loss": 1.1698,
+      "step": 7589
+    },
+    {
+      "epoch": 2.825796093122826,
+      "grad_norm": 0.16407622396945953,
+      "learning_rate": 1.6960655122071023e-05,
+      "loss": 1.1619,
+      "step": 7590
+    },
+    {
+      "epoch": 2.8261683982734347,
+      "grad_norm": 0.15823231637477875,
+      "learning_rate": 1.6959782499963827e-05,
+      "loss": 1.1612,
+      "step": 7591
+    },
+    {
+      "epoch": 2.826540703424044,
+      "grad_norm": 0.16647450625896454,
+      "learning_rate": 1.6958909775061082e-05,
+      "loss": 1.1726,
+      "step": 7592
+    },
+    {
+      "epoch": 2.826913008574653,
+      "grad_norm": 0.16879020631313324,
+      "learning_rate": 1.6958036947375676e-05,
+      "loss": 1.1824,
+      "step": 7593
+    },
+    {
+      "epoch": 2.827285313725262,
+      "grad_norm": 0.16589350998401642,
+      "learning_rate": 1.69571640169205e-05,
+      "loss": 1.1716,
+      "step": 7594
+    },
+    {
+      "epoch": 2.827657618875871,
+      "grad_norm": 0.1656552106142044,
+      "learning_rate": 1.695629098370845e-05,
+      "loss": 1.1689,
+      "step": 7595
+    },
+    {
+      "epoch": 2.82802992402648,
+      "grad_norm": 0.16472189128398895,
+      "learning_rate": 1.6955417847752417e-05,
+      "loss": 1.1785,
+      "step": 7596
+    },
+    {
+      "epoch": 2.8284022291770894,
+      "grad_norm": 0.16501384973526,
+      "learning_rate": 1.69545446090653e-05,
+      "loss": 1.1715,
+      "step": 7597
+    },
+    {
+      "epoch": 2.828774534327698,
+      "grad_norm": 0.16673643887043,
+      "learning_rate": 1.6953671267659996e-05,
+      "loss": 1.184,
+      "step": 7598
+    },
+    {
+      "epoch": 2.8291468394783075,
+      "grad_norm": 0.16138608753681183,
+      "learning_rate": 1.6952797823549406e-05,
+      "loss": 1.164,
+      "step": 7599
+    },
+    {
+      "epoch": 2.8295191446289163,
+      "grad_norm": 0.15768542885780334,
+      "learning_rate": 1.6951924276746425e-05,
+      "loss": 1.1794,
+      "step": 7600
+    },
+    {
+      "epoch": 2.8298914497795256,
+      "grad_norm": 0.1626758873462677,
+      "learning_rate": 1.6951050627263958e-05,
+      "loss": 1.1608,
+      "step": 7601
+    },
+    {
+      "epoch": 2.830263754930135,
+      "grad_norm": 0.16393102705478668,
+      "learning_rate": 1.695017687511491e-05,
+      "loss": 1.1826,
+      "step": 7602
+    },
+    {
+      "epoch": 2.8306360600807436,
+      "grad_norm": 0.16286106407642365,
+      "learning_rate": 1.6949303020312188e-05,
+      "loss": 1.1822,
+      "step": 7603
+    },
+    {
+      "epoch": 2.831008365231353,
+      "grad_norm": 0.16217343509197235,
+      "learning_rate": 1.6948429062868697e-05,
+      "loss": 1.1658,
+      "step": 7604
+    },
+    {
+      "epoch": 2.8313806703819617,
+      "grad_norm": 0.17708547413349152,
+      "learning_rate": 1.6947555002797344e-05,
+      "loss": 1.1522,
+      "step": 7605
+    },
+    {
+      "epoch": 2.831752975532571,
+      "grad_norm": 0.18956957757472992,
+      "learning_rate": 1.6946680840111035e-05,
+      "loss": 1.17,
+      "step": 7606
+    },
+    {
+      "epoch": 2.83212528068318,
+      "grad_norm": 0.1878315806388855,
+      "learning_rate": 1.6945806574822693e-05,
+      "loss": 1.1833,
+      "step": 7607
+    },
+    {
+      "epoch": 2.832497585833789,
+      "grad_norm": 0.1767759919166565,
+      "learning_rate": 1.6944932206945218e-05,
+      "loss": 1.1772,
+      "step": 7608
+    },
+    {
+      "epoch": 2.832869890984398,
+      "grad_norm": 0.1732354611158371,
+      "learning_rate": 1.6944057736491534e-05,
+      "loss": 1.1649,
+      "step": 7609
+    },
+    {
+      "epoch": 2.833242196135007,
+      "grad_norm": 0.16452713310718536,
+      "learning_rate": 1.694318316347455e-05,
+      "loss": 1.1707,
+      "step": 7610
+    },
+    {
+      "epoch": 2.8336145012856164,
+      "grad_norm": 0.167094424366951,
+      "learning_rate": 1.6942308487907187e-05,
+      "loss": 1.1749,
+      "step": 7611
+    },
+    {
+      "epoch": 2.8339868064362252,
+      "grad_norm": 0.17273786664009094,
+      "learning_rate": 1.694143370980237e-05,
+      "loss": 1.1577,
+      "step": 7612
+    },
+    {
+      "epoch": 2.8343591115868345,
+      "grad_norm": 0.16662479937076569,
+      "learning_rate": 1.6940558829173004e-05,
+      "loss": 1.1733,
+      "step": 7613
+    },
+    {
+      "epoch": 2.8347314167374433,
+      "grad_norm": 0.16592276096343994,
+      "learning_rate": 1.6939683846032022e-05,
+      "loss": 1.1724,
+      "step": 7614
+    },
+    {
+      "epoch": 2.8351037218880526,
+      "grad_norm": 0.16113296151161194,
+      "learning_rate": 1.6938808760392346e-05,
+      "loss": 1.1643,
+      "step": 7615
+    },
+    {
+      "epoch": 2.8354760270386614,
+      "grad_norm": 0.16023372113704681,
+      "learning_rate": 1.69379335722669e-05,
+      "loss": 1.1684,
+      "step": 7616
+    },
+    {
+      "epoch": 2.8358483321892707,
+      "grad_norm": 0.16699042916297913,
+      "learning_rate": 1.693705828166861e-05,
+      "loss": 1.1701,
+      "step": 7617
+    },
+    {
+      "epoch": 2.8362206373398795,
+      "grad_norm": 0.1670234203338623,
+      "learning_rate": 1.693618288861041e-05,
+      "loss": 1.1809,
+      "step": 7618
+    },
+    {
+      "epoch": 2.8365929424904888,
+      "grad_norm": 0.19244331121444702,
+      "learning_rate": 1.6935307393105215e-05,
+      "loss": 1.153,
+      "step": 7619
+    },
+    {
+      "epoch": 2.836965247641098,
+      "grad_norm": 0.16699756681919098,
+      "learning_rate": 1.6934431795165972e-05,
+      "loss": 1.1687,
+      "step": 7620
+    },
+    {
+      "epoch": 2.837337552791707,
+      "grad_norm": 0.1908612698316574,
+      "learning_rate": 1.6933556094805602e-05,
+      "loss": 1.1881,
+      "step": 7621
+    },
+    {
+      "epoch": 2.837709857942316,
+      "grad_norm": 0.17475947737693787,
+      "learning_rate": 1.6932680292037045e-05,
+      "loss": 1.1511,
+      "step": 7622
+    },
+    {
+      "epoch": 2.838082163092925,
+      "grad_norm": 0.16836613416671753,
+      "learning_rate": 1.6931804386873232e-05,
+      "loss": 1.1599,
+      "step": 7623
+    },
+    {
+      "epoch": 2.838454468243534,
+      "grad_norm": 0.17344890534877777,
+      "learning_rate": 1.693092837932711e-05,
+      "loss": 1.1767,
+      "step": 7624
+    },
+    {
+      "epoch": 2.838826773394143,
+      "grad_norm": 0.17366820573806763,
+      "learning_rate": 1.693005226941161e-05,
+      "loss": 1.1815,
+      "step": 7625
+    },
+    {
+      "epoch": 2.8391990785447523,
+      "grad_norm": 0.16873829066753387,
+      "learning_rate": 1.692917605713967e-05,
+      "loss": 1.1751,
+      "step": 7626
+    },
+    {
+      "epoch": 2.839571383695361,
+      "grad_norm": 0.1660252809524536,
+      "learning_rate": 1.6928299742524236e-05,
+      "loss": 1.1723,
+      "step": 7627
+    },
+    {
+      "epoch": 2.8399436888459704,
+      "grad_norm": 0.17038235068321228,
+      "learning_rate": 1.6927423325578248e-05,
+      "loss": 1.1829,
+      "step": 7628
+    },
+    {
+      "epoch": 2.8403159939965796,
+      "grad_norm": 0.1674618273973465,
+      "learning_rate": 1.692654680631465e-05,
+      "loss": 1.1787,
+      "step": 7629
+    },
+    {
+      "epoch": 2.8406882991471885,
+      "grad_norm": 0.16933341324329376,
+      "learning_rate": 1.692567018474639e-05,
+      "loss": 1.169,
+      "step": 7630
+    },
+    {
+      "epoch": 2.8410606042977977,
+      "grad_norm": 0.16508066654205322,
+      "learning_rate": 1.6924793460886424e-05,
+      "loss": 1.1711,
+      "step": 7631
+    },
+    {
+      "epoch": 2.8414329094484065,
+      "grad_norm": 0.1625910848379135,
+      "learning_rate": 1.692391663474769e-05,
+      "loss": 1.1557,
+      "step": 7632
+    },
+    {
+      "epoch": 2.841805214599016,
+      "grad_norm": 0.16758020222187042,
+      "learning_rate": 1.692303970634314e-05,
+      "loss": 1.1555,
+      "step": 7633
+    },
+    {
+      "epoch": 2.8421775197496246,
+      "grad_norm": 0.1603182852268219,
+      "learning_rate": 1.6922162675685725e-05,
+      "loss": 1.1575,
+      "step": 7634
+    },
+    {
+      "epoch": 2.842549824900234,
+      "grad_norm": 0.16572555899620056,
+      "learning_rate": 1.6921285542788405e-05,
+      "loss": 1.1687,
+      "step": 7635
+    },
+    {
+      "epoch": 2.8429221300508427,
+      "grad_norm": 0.16787642240524292,
+      "learning_rate": 1.6920408307664132e-05,
+      "loss": 1.1629,
+      "step": 7636
+    },
+    {
+      "epoch": 2.843294435201452,
+      "grad_norm": 0.16438321769237518,
+      "learning_rate": 1.6919530970325865e-05,
+      "loss": 1.1628,
+      "step": 7637
+    },
+    {
+      "epoch": 2.8436667403520612,
+      "grad_norm": 0.16906222701072693,
+      "learning_rate": 1.6918653530786555e-05,
+      "loss": 1.1589,
+      "step": 7638
+    },
+    {
+      "epoch": 2.84403904550267,
+      "grad_norm": 0.1649426817893982,
+      "learning_rate": 1.6917775989059167e-05,
+      "loss": 1.1705,
+      "step": 7639
+    },
+    {
+      "epoch": 2.8444113506532793,
+      "grad_norm": 0.17663908004760742,
+      "learning_rate": 1.6916898345156668e-05,
+      "loss": 1.1638,
+      "step": 7640
+    },
+    {
+      "epoch": 2.844783655803888,
+      "grad_norm": 0.16318029165267944,
+      "learning_rate": 1.6916020599092007e-05,
+      "loss": 1.1592,
+      "step": 7641
+    },
+    {
+      "epoch": 2.8451559609544974,
+      "grad_norm": 0.17562070488929749,
+      "learning_rate": 1.691514275087816e-05,
+      "loss": 1.1606,
+      "step": 7642
+    },
+    {
+      "epoch": 2.8455282661051067,
+      "grad_norm": 0.1714848279953003,
+      "learning_rate": 1.6914264800528087e-05,
+      "loss": 1.1647,
+      "step": 7643
+    },
+    {
+      "epoch": 2.8459005712557155,
+      "grad_norm": 0.1683172881603241,
+      "learning_rate": 1.6913386748054757e-05,
+      "loss": 1.175,
+      "step": 7644
+    },
+    {
+      "epoch": 2.8462728764063243,
+      "grad_norm": 0.18591049313545227,
+      "learning_rate": 1.6912508593471137e-05,
+      "loss": 1.1501,
+      "step": 7645
+    },
+    {
+      "epoch": 2.8466451815569336,
+      "grad_norm": 0.16155651211738586,
+      "learning_rate": 1.69116303367902e-05,
+      "loss": 1.1764,
+      "step": 7646
+    },
+    {
+      "epoch": 2.847017486707543,
+      "grad_norm": 0.1677398383617401,
+      "learning_rate": 1.691075197802492e-05,
+      "loss": 1.1649,
+      "step": 7647
+    },
+    {
+      "epoch": 2.8473897918581517,
+      "grad_norm": 0.16577932238578796,
+      "learning_rate": 1.6909873517188263e-05,
+      "loss": 1.1761,
+      "step": 7648
+    },
+    {
+      "epoch": 2.847762097008761,
+      "grad_norm": 0.161788210272789,
+      "learning_rate": 1.690899495429321e-05,
+      "loss": 1.1662,
+      "step": 7649
+    },
+    {
+      "epoch": 2.8481344021593697,
+      "grad_norm": 0.20268653333187103,
+      "learning_rate": 1.6908116289352735e-05,
+      "loss": 1.1871,
+      "step": 7650
+    },
+    {
+      "epoch": 2.848506707309979,
+      "grad_norm": 0.18162259459495544,
+      "learning_rate": 1.6907237522379816e-05,
+      "loss": 1.1549,
+      "step": 7651
+    },
+    {
+      "epoch": 2.8488790124605883,
+      "grad_norm": 0.16151395440101624,
+      "learning_rate": 1.690635865338743e-05,
+      "loss": 1.1815,
+      "step": 7652
+    },
+    {
+      "epoch": 2.849251317611197,
+      "grad_norm": 0.1718040108680725,
+      "learning_rate": 1.6905479682388565e-05,
+      "loss": 1.1562,
+      "step": 7653
+    },
+    {
+      "epoch": 2.849623622761806,
+      "grad_norm": 0.165380597114563,
+      "learning_rate": 1.690460060939619e-05,
+      "loss": 1.1713,
+      "step": 7654
+    },
+    {
+      "epoch": 2.849995927912415,
+      "grad_norm": 0.18731936812400818,
+      "learning_rate": 1.6903721434423306e-05,
+      "loss": 1.1732,
+      "step": 7655
+    },
+    {
+      "epoch": 2.8503682330630244,
+      "grad_norm": 0.16210754215717316,
+      "learning_rate": 1.6902842157482885e-05,
+      "loss": 1.1712,
+      "step": 7656
+    },
+    {
+      "epoch": 2.8507405382136333,
+      "grad_norm": 0.1721230000257492,
+      "learning_rate": 1.690196277858792e-05,
+      "loss": 1.1572,
+      "step": 7657
+    },
+    {
+      "epoch": 2.8511128433642425,
+      "grad_norm": 0.16690939664840698,
+      "learning_rate": 1.6901083297751397e-05,
+      "loss": 1.1777,
+      "step": 7658
+    },
+    {
+      "epoch": 2.8514851485148514,
+      "grad_norm": 0.16586491465568542,
+      "learning_rate": 1.6900203714986307e-05,
+      "loss": 1.1747,
+      "step": 7659
+    },
+    {
+      "epoch": 2.8518574536654606,
+      "grad_norm": 0.17058177292346954,
+      "learning_rate": 1.689932403030564e-05,
+      "loss": 1.1755,
+      "step": 7660
+    },
+    {
+      "epoch": 2.85222975881607,
+      "grad_norm": 0.16409049928188324,
+      "learning_rate": 1.6898444243722395e-05,
+      "loss": 1.1629,
+      "step": 7661
+    },
+    {
+      "epoch": 2.8526020639666787,
+      "grad_norm": 0.16023750603199005,
+      "learning_rate": 1.6897564355249556e-05,
+      "loss": 1.1642,
+      "step": 7662
+    },
+    {
+      "epoch": 2.8529743691172875,
+      "grad_norm": 0.17429199814796448,
+      "learning_rate": 1.6896684364900127e-05,
+      "loss": 1.1746,
+      "step": 7663
+    },
+    {
+      "epoch": 2.853346674267897,
+      "grad_norm": 0.1672469824552536,
+      "learning_rate": 1.68958042726871e-05,
+      "loss": 1.1697,
+      "step": 7664
+    },
+    {
+      "epoch": 2.853718979418506,
+      "grad_norm": 0.16756999492645264,
+      "learning_rate": 1.689492407862348e-05,
+      "loss": 1.1722,
+      "step": 7665
+    },
+    {
+      "epoch": 2.854091284569115,
+      "grad_norm": 0.16315622627735138,
+      "learning_rate": 1.689404378272226e-05,
+      "loss": 1.1669,
+      "step": 7666
+    },
+    {
+      "epoch": 2.854463589719724,
+      "grad_norm": 0.16442465782165527,
+      "learning_rate": 1.6893163384996453e-05,
+      "loss": 1.1892,
+      "step": 7667
+    },
+    {
+      "epoch": 2.854835894870333,
+      "grad_norm": 0.16709788143634796,
+      "learning_rate": 1.689228288545905e-05,
+      "loss": 1.1729,
+      "step": 7668
+    },
+    {
+      "epoch": 2.855208200020942,
+      "grad_norm": 0.17159411311149597,
+      "learning_rate": 1.689140228412306e-05,
+      "loss": 1.1619,
+      "step": 7669
+    },
+    {
+      "epoch": 2.8555805051715515,
+      "grad_norm": 0.16300630569458008,
+      "learning_rate": 1.68905215810015e-05,
+      "loss": 1.1589,
+      "step": 7670
+    },
+    {
+      "epoch": 2.8559528103221603,
+      "grad_norm": 0.17560981214046478,
+      "learning_rate": 1.6889640776107356e-05,
+      "loss": 1.1749,
+      "step": 7671
+    },
+    {
+      "epoch": 2.856325115472769,
+      "grad_norm": 0.16075876355171204,
+      "learning_rate": 1.688875986945366e-05,
+      "loss": 1.1659,
+      "step": 7672
+    },
+    {
+      "epoch": 2.8566974206233784,
+      "grad_norm": 0.16523872315883636,
+      "learning_rate": 1.6887878861053407e-05,
+      "loss": 1.175,
+      "step": 7673
+    },
+    {
+      "epoch": 2.8570697257739877,
+      "grad_norm": 0.1663205772638321,
+      "learning_rate": 1.688699775091962e-05,
+      "loss": 1.1689,
+      "step": 7674
+    },
+    {
+      "epoch": 2.8574420309245965,
+      "grad_norm": 0.17149294912815094,
+      "learning_rate": 1.68861165390653e-05,
+      "loss": 1.1637,
+      "step": 7675
+    },
+    {
+      "epoch": 2.8578143360752057,
+      "grad_norm": 0.15809527039527893,
+      "learning_rate": 1.688523522550348e-05,
+      "loss": 1.1621,
+      "step": 7676
+    },
+    {
+      "epoch": 2.8581866412258146,
+      "grad_norm": 0.16826729476451874,
+      "learning_rate": 1.6884353810247166e-05,
+      "loss": 1.1671,
+      "step": 7677
+    },
+    {
+      "epoch": 2.858558946376424,
+      "grad_norm": 0.1609870195388794,
+      "learning_rate": 1.6883472293309375e-05,
+      "loss": 1.1648,
+      "step": 7678
+    },
+    {
+      "epoch": 2.858931251527033,
+      "grad_norm": 0.16899265348911285,
+      "learning_rate": 1.6882590674703135e-05,
+      "loss": 1.1532,
+      "step": 7679
+    },
+    {
+      "epoch": 2.859303556677642,
+      "grad_norm": 0.1640852838754654,
+      "learning_rate": 1.6881708954441458e-05,
+      "loss": 1.1584,
+      "step": 7680
+    },
+    {
+      "epoch": 2.8596758618282507,
+      "grad_norm": 0.16974326968193054,
+      "learning_rate": 1.6880827132537373e-05,
+      "loss": 1.1733,
+      "step": 7681
+    },
+    {
+      "epoch": 2.86004816697886,
+      "grad_norm": 0.16964828968048096,
+      "learning_rate": 1.6879945209003903e-05,
+      "loss": 1.1698,
+      "step": 7682
+    },
+    {
+      "epoch": 2.8604204721294693,
+      "grad_norm": 0.17090067267417908,
+      "learning_rate": 1.6879063183854076e-05,
+      "loss": 1.1671,
+      "step": 7683
+    },
+    {
+      "epoch": 2.860792777280078,
+      "grad_norm": 0.1564168483018875,
+      "learning_rate": 1.6878181057100915e-05,
+      "loss": 1.1604,
+      "step": 7684
+    },
+    {
+      "epoch": 2.8611650824306873,
+      "grad_norm": 0.1685391068458557,
+      "learning_rate": 1.6877298828757452e-05,
+      "loss": 1.1765,
+      "step": 7685
+    },
+    {
+      "epoch": 2.861537387581296,
+      "grad_norm": 0.16214828193187714,
+      "learning_rate": 1.6876416498836717e-05,
+      "loss": 1.166,
+      "step": 7686
+    },
+    {
+      "epoch": 2.8619096927319054,
+      "grad_norm": 0.16030354797840118,
+      "learning_rate": 1.6875534067351744e-05,
+      "loss": 1.1634,
+      "step": 7687
+    },
+    {
+      "epoch": 2.8622819978825147,
+      "grad_norm": 0.17029841244220734,
+      "learning_rate": 1.687465153431556e-05,
+      "loss": 1.1719,
+      "step": 7688
+    },
+    {
+      "epoch": 2.8626543030331235,
+      "grad_norm": 0.1636890470981598,
+      "learning_rate": 1.6873768899741212e-05,
+      "loss": 1.1582,
+      "step": 7689
+    },
+    {
+      "epoch": 2.8630266081837323,
+      "grad_norm": 0.16757601499557495,
+      "learning_rate": 1.687288616364172e-05,
+      "loss": 1.1557,
+      "step": 7690
+    },
+    {
+      "epoch": 2.8633989133343416,
+      "grad_norm": 0.18567247688770294,
+      "learning_rate": 1.6872003326030136e-05,
+      "loss": 1.1699,
+      "step": 7691
+    },
+    {
+      "epoch": 2.863771218484951,
+      "grad_norm": 0.1703827977180481,
+      "learning_rate": 1.6871120386919493e-05,
+      "loss": 1.1579,
+      "step": 7692
+    },
+    {
+      "epoch": 2.8641435236355597,
+      "grad_norm": 0.16367879509925842,
+      "learning_rate": 1.6870237346322832e-05,
+      "loss": 1.156,
+      "step": 7693
+    },
+    {
+      "epoch": 2.864515828786169,
+      "grad_norm": 0.1624721884727478,
+      "learning_rate": 1.6869354204253195e-05,
+      "loss": 1.1659,
+      "step": 7694
+    },
+    {
+      "epoch": 2.8648881339367778,
+      "grad_norm": 0.16088762879371643,
+      "learning_rate": 1.686847096072363e-05,
+      "loss": 1.1753,
+      "step": 7695
+    },
+    {
+      "epoch": 2.865260439087387,
+      "grad_norm": 0.17531664669513702,
+      "learning_rate": 1.6867587615747184e-05,
+      "loss": 1.183,
+      "step": 7696
+    },
+    {
+      "epoch": 2.8656327442379963,
+      "grad_norm": 0.18076026439666748,
+      "learning_rate": 1.6866704169336895e-05,
+      "loss": 1.1681,
+      "step": 7697
+    },
+    {
+      "epoch": 2.866005049388605,
+      "grad_norm": 0.20938239991664886,
+      "learning_rate": 1.6865820621505812e-05,
+      "loss": 1.1816,
+      "step": 7698
+    },
+    {
+      "epoch": 2.866377354539214,
+      "grad_norm": 0.23691917955875397,
+      "learning_rate": 1.6864936972266996e-05,
+      "loss": 1.1581,
+      "step": 7699
+    },
+    {
+      "epoch": 2.866749659689823,
+      "grad_norm": 0.19211941957473755,
+      "learning_rate": 1.686405322163349e-05,
+      "loss": 1.1664,
+      "step": 7700
+    },
+    {
+      "epoch": 2.8671219648404325,
+      "grad_norm": 0.17298676073551178,
+      "learning_rate": 1.686316936961835e-05,
+      "loss": 1.1663,
+      "step": 7701
+    },
+    {
+      "epoch": 2.8674942699910413,
+      "grad_norm": 0.17014887928962708,
+      "learning_rate": 1.6862285416234628e-05,
+      "loss": 1.162,
+      "step": 7702
+    },
+    {
+      "epoch": 2.8678665751416506,
+      "grad_norm": 0.1860615760087967,
+      "learning_rate": 1.686140136149538e-05,
+      "loss": 1.1707,
+      "step": 7703
+    },
+    {
+      "epoch": 2.8682388802922594,
+      "grad_norm": 0.1729809045791626,
+      "learning_rate": 1.6860517205413667e-05,
+      "loss": 1.1786,
+      "step": 7704
+    },
+    {
+      "epoch": 2.8686111854428686,
+      "grad_norm": 0.16746124625205994,
+      "learning_rate": 1.6859632948002542e-05,
+      "loss": 1.164,
+      "step": 7705
+    },
+    {
+      "epoch": 2.868983490593478,
+      "grad_norm": 0.176571786403656,
+      "learning_rate": 1.685874858927507e-05,
+      "loss": 1.152,
+      "step": 7706
+    },
+    {
+      "epoch": 2.8693557957440867,
+      "grad_norm": 0.17990010976791382,
+      "learning_rate": 1.6857864129244314e-05,
+      "loss": 1.1736,
+      "step": 7707
+    },
+    {
+      "epoch": 2.8697281008946955,
+      "grad_norm": 0.16373847424983978,
+      "learning_rate": 1.6856979567923333e-05,
+      "loss": 1.1595,
+      "step": 7708
+    },
+    {
+      "epoch": 2.870100406045305,
+      "grad_norm": 0.16468669474124908,
+      "learning_rate": 1.6856094905325195e-05,
+      "loss": 1.1652,
+      "step": 7709
+    },
+    {
+      "epoch": 2.870472711195914,
+      "grad_norm": 0.1700790673494339,
+      "learning_rate": 1.6855210141462964e-05,
+      "loss": 1.1577,
+      "step": 7710
+    },
+    {
+      "epoch": 2.870845016346523,
+      "grad_norm": 0.17594699561595917,
+      "learning_rate": 1.685432527634971e-05,
+      "loss": 1.1672,
+      "step": 7711
+    },
+    {
+      "epoch": 2.871217321497132,
+      "grad_norm": 0.16046394407749176,
+      "learning_rate": 1.68534403099985e-05,
+      "loss": 1.1777,
+      "step": 7712
+    },
+    {
+      "epoch": 2.871589626647741,
+      "grad_norm": 0.1609521210193634,
+      "learning_rate": 1.685255524242241e-05,
+      "loss": 1.173,
+      "step": 7713
+    },
+    {
+      "epoch": 2.8719619317983502,
+      "grad_norm": 0.17106173932552338,
+      "learning_rate": 1.6851670073634513e-05,
+      "loss": 1.1632,
+      "step": 7714
+    },
+    {
+      "epoch": 2.8723342369489595,
+      "grad_norm": 0.16623558104038239,
+      "learning_rate": 1.685078480364787e-05,
+      "loss": 1.1556,
+      "step": 7715
+    },
+    {
+      "epoch": 2.8727065420995683,
+      "grad_norm": 0.1585904061794281,
+      "learning_rate": 1.684989943247557e-05,
+      "loss": 1.1718,
+      "step": 7716
+    },
+    {
+      "epoch": 2.8730788472501776,
+      "grad_norm": 0.1625899374485016,
+      "learning_rate": 1.6849013960130687e-05,
+      "loss": 1.1638,
+      "step": 7717
+    },
+    {
+      "epoch": 2.8734511524007864,
+      "grad_norm": 0.1680610626935959,
+      "learning_rate": 1.6848128386626297e-05,
+      "loss": 1.175,
+      "step": 7718
+    },
+    {
+      "epoch": 2.8738234575513957,
+      "grad_norm": 0.16707110404968262,
+      "learning_rate": 1.6847242711975477e-05,
+      "loss": 1.1724,
+      "step": 7719
+    },
+    {
+      "epoch": 2.8741957627020045,
+      "grad_norm": 0.16035796701908112,
+      "learning_rate": 1.684635693619131e-05,
+      "loss": 1.1504,
+      "step": 7720
+    },
+    {
+      "epoch": 2.8745680678526138,
+      "grad_norm": 0.1676149070262909,
+      "learning_rate": 1.684547105928689e-05,
+      "loss": 1.1657,
+      "step": 7721
+    },
+    {
+      "epoch": 2.8749403730032226,
+      "grad_norm": 0.17349836230278015,
+      "learning_rate": 1.6844585081275285e-05,
+      "loss": 1.1534,
+      "step": 7722
+    },
+    {
+      "epoch": 2.875312678153832,
+      "grad_norm": 0.16714976727962494,
+      "learning_rate": 1.684369900216959e-05,
+      "loss": 1.1686,
+      "step": 7723
+    },
+    {
+      "epoch": 2.875684983304441,
+      "grad_norm": 0.168970987200737,
+      "learning_rate": 1.684281282198289e-05,
+      "loss": 1.1644,
+      "step": 7724
+    },
+    {
+      "epoch": 2.87605728845505,
+      "grad_norm": 0.16321352124214172,
+      "learning_rate": 1.6841926540728276e-05,
+      "loss": 1.1585,
+      "step": 7725
+    },
+    {
+      "epoch": 2.876429593605659,
+      "grad_norm": 0.16680431365966797,
+      "learning_rate": 1.684104015841883e-05,
+      "loss": 1.1496,
+      "step": 7726
+    },
+    {
+      "epoch": 2.876801898756268,
+      "grad_norm": 0.1670006364583969,
+      "learning_rate": 1.6840153675067658e-05,
+      "loss": 1.1815,
+      "step": 7727
+    },
+    {
+      "epoch": 2.8771742039068773,
+      "grad_norm": 0.1647917628288269,
+      "learning_rate": 1.683926709068784e-05,
+      "loss": 1.1795,
+      "step": 7728
+    },
+    {
+      "epoch": 2.877546509057486,
+      "grad_norm": 0.16818176209926605,
+      "learning_rate": 1.683838040529248e-05,
+      "loss": 1.181,
+      "step": 7729
+    },
+    {
+      "epoch": 2.8779188142080954,
+      "grad_norm": 0.16495051980018616,
+      "learning_rate": 1.6837493618894666e-05,
+      "loss": 1.151,
+      "step": 7730
+    },
+    {
+      "epoch": 2.878291119358704,
+      "grad_norm": 0.15997083485126495,
+      "learning_rate": 1.6836606731507506e-05,
+      "loss": 1.1581,
+      "step": 7731
+    },
+    {
+      "epoch": 2.8786634245093135,
+      "grad_norm": 0.16573216021060944,
+      "learning_rate": 1.683571974314409e-05,
+      "loss": 1.1754,
+      "step": 7732
+    },
+    {
+      "epoch": 2.8790357296599227,
+      "grad_norm": 0.16260533034801483,
+      "learning_rate": 1.683483265381752e-05,
+      "loss": 1.1807,
+      "step": 7733
+    },
+    {
+      "epoch": 2.8794080348105315,
+      "grad_norm": 0.16376620531082153,
+      "learning_rate": 1.6833945463540906e-05,
+      "loss": 1.1666,
+      "step": 7734
+    },
+    {
+      "epoch": 2.879780339961141,
+      "grad_norm": 0.1692168414592743,
+      "learning_rate": 1.6833058172327344e-05,
+      "loss": 1.1674,
+      "step": 7735
+    },
+    {
+      "epoch": 2.8801526451117496,
+      "grad_norm": 0.16318558156490326,
+      "learning_rate": 1.683217078018994e-05,
+      "loss": 1.1756,
+      "step": 7736
+    },
+    {
+      "epoch": 2.880524950262359,
+      "grad_norm": 0.1628124862909317,
+      "learning_rate": 1.6831283287141807e-05,
+      "loss": 1.1631,
+      "step": 7737
+    },
+    {
+      "epoch": 2.8808972554129677,
+      "grad_norm": 0.1683143526315689,
+      "learning_rate": 1.6830395693196043e-05,
+      "loss": 1.1762,
+      "step": 7738
+    },
+    {
+      "epoch": 2.881269560563577,
+      "grad_norm": 0.16499009728431702,
+      "learning_rate": 1.682950799836577e-05,
+      "loss": 1.1689,
+      "step": 7739
+    },
+    {
+      "epoch": 2.881641865714186,
+      "grad_norm": 0.16890759766101837,
+      "learning_rate": 1.6828620202664086e-05,
+      "loss": 1.1914,
+      "step": 7740
+    },
+    {
+      "epoch": 2.882014170864795,
+      "grad_norm": 0.16511119902133942,
+      "learning_rate": 1.6827732306104113e-05,
+      "loss": 1.1627,
+      "step": 7741
+    },
+    {
+      "epoch": 2.8823864760154043,
+      "grad_norm": 0.16963976621627808,
+      "learning_rate": 1.6826844308698962e-05,
+      "loss": 1.1626,
+      "step": 7742
+    },
+    {
+      "epoch": 2.882758781166013,
+      "grad_norm": 0.17037193477153778,
+      "learning_rate": 1.682595621046175e-05,
+      "loss": 1.1666,
+      "step": 7743
+    },
+    {
+      "epoch": 2.8831310863166224,
+      "grad_norm": 0.16597189009189606,
+      "learning_rate": 1.682506801140559e-05,
+      "loss": 1.1835,
+      "step": 7744
+    },
+    {
+      "epoch": 2.8835033914672312,
+      "grad_norm": 0.2021397054195404,
+      "learning_rate": 1.6824179711543607e-05,
+      "loss": 1.1815,
+      "step": 7745
+    },
+    {
+      "epoch": 2.8838756966178405,
+      "grad_norm": 0.1731014996767044,
+      "learning_rate": 1.6823291310888916e-05,
+      "loss": 1.1642,
+      "step": 7746
+    },
+    {
+      "epoch": 2.8842480017684493,
+      "grad_norm": 0.17293405532836914,
+      "learning_rate": 1.682240280945464e-05,
+      "loss": 1.1538,
+      "step": 7747
+    },
+    {
+      "epoch": 2.8846203069190586,
+      "grad_norm": 0.1841578483581543,
+      "learning_rate": 1.6821514207253905e-05,
+      "loss": 1.1458,
+      "step": 7748
+    },
+    {
+      "epoch": 2.8849926120696674,
+      "grad_norm": 0.16726046800613403,
+      "learning_rate": 1.6820625504299833e-05,
+      "loss": 1.1729,
+      "step": 7749
+    },
+    {
+      "epoch": 2.8853649172202767,
+      "grad_norm": 0.1666422039270401,
+      "learning_rate": 1.6819736700605548e-05,
+      "loss": 1.16,
+      "step": 7750
+    },
+    {
+      "epoch": 2.885737222370886,
+      "grad_norm": 0.1761844903230667,
+      "learning_rate": 1.6818847796184185e-05,
+      "loss": 1.1787,
+      "step": 7751
+    },
+    {
+      "epoch": 2.8861095275214947,
+      "grad_norm": 0.16153545677661896,
+      "learning_rate": 1.6817958791048866e-05,
+      "loss": 1.1775,
+      "step": 7752
+    },
+    {
+      "epoch": 2.886481832672104,
+      "grad_norm": 0.16363756358623505,
+      "learning_rate": 1.6817069685212717e-05,
+      "loss": 1.1761,
+      "step": 7753
+    },
+    {
+      "epoch": 2.886854137822713,
+      "grad_norm": 0.176415354013443,
+      "learning_rate": 1.6816180478688885e-05,
+      "loss": 1.1782,
+      "step": 7754
+    },
+    {
+      "epoch": 2.887226442973322,
+      "grad_norm": 0.17358264327049255,
+      "learning_rate": 1.681529117149049e-05,
+      "loss": 1.1663,
+      "step": 7755
+    },
+    {
+      "epoch": 2.887598748123931,
+      "grad_norm": 0.1688726395368576,
+      "learning_rate": 1.6814401763630674e-05,
+      "loss": 1.1724,
+      "step": 7756
+    },
+    {
+      "epoch": 2.88797105327454,
+      "grad_norm": 0.20944684743881226,
+      "learning_rate": 1.6813512255122573e-05,
+      "loss": 1.1587,
+      "step": 7757
+    },
+    {
+      "epoch": 2.888343358425149,
+      "grad_norm": 0.20509681105613708,
+      "learning_rate": 1.681262264597932e-05,
+      "loss": 1.1472,
+      "step": 7758
+    },
+    {
+      "epoch": 2.8887156635757583,
+      "grad_norm": 0.17619788646697998,
+      "learning_rate": 1.6811732936214063e-05,
+      "loss": 1.1787,
+      "step": 7759
+    },
+    {
+      "epoch": 2.8890879687263675,
+      "grad_norm": 0.16343960165977478,
+      "learning_rate": 1.6810843125839934e-05,
+      "loss": 1.1768,
+      "step": 7760
+    },
+    {
+      "epoch": 2.8894602738769763,
+      "grad_norm": 0.1722916066646576,
+      "learning_rate": 1.680995321487008e-05,
+      "loss": 1.1698,
+      "step": 7761
+    },
+    {
+      "epoch": 2.8898325790275856,
+      "grad_norm": 0.1792430579662323,
+      "learning_rate": 1.6809063203317645e-05,
+      "loss": 1.1665,
+      "step": 7762
+    },
+    {
+      "epoch": 2.8902048841781944,
+      "grad_norm": 0.17468638718128204,
+      "learning_rate": 1.6808173091195774e-05,
+      "loss": 1.1726,
+      "step": 7763
+    },
+    {
+      "epoch": 2.8905771893288037,
+      "grad_norm": 0.1607234925031662,
+      "learning_rate": 1.6807282878517614e-05,
+      "loss": 1.1714,
+      "step": 7764
+    },
+    {
+      "epoch": 2.890949494479413,
+      "grad_norm": 0.15706735849380493,
+      "learning_rate": 1.680639256529631e-05,
+      "loss": 1.1607,
+      "step": 7765
+    },
+    {
+      "epoch": 2.891321799630022,
+      "grad_norm": 0.15637610852718353,
+      "learning_rate": 1.6805502151545022e-05,
+      "loss": 1.1658,
+      "step": 7766
+    },
+    {
+      "epoch": 2.8916941047806306,
+      "grad_norm": 0.16825786232948303,
+      "learning_rate": 1.6804611637276888e-05,
+      "loss": 1.1706,
+      "step": 7767
+    },
+    {
+      "epoch": 2.89206640993124,
+      "grad_norm": 0.18059858679771423,
+      "learning_rate": 1.680372102250507e-05,
+      "loss": 1.1694,
+      "step": 7768
+    },
+    {
+      "epoch": 2.892438715081849,
+      "grad_norm": 0.18463246524333954,
+      "learning_rate": 1.6802830307242716e-05,
+      "loss": 1.1592,
+      "step": 7769
+    },
+    {
+      "epoch": 2.892811020232458,
+      "grad_norm": 0.16244830191135406,
+      "learning_rate": 1.680193949150299e-05,
+      "loss": 1.1693,
+      "step": 7770
+    },
+    {
+      "epoch": 2.893183325383067,
+      "grad_norm": 0.1961405873298645,
+      "learning_rate": 1.680104857529904e-05,
+      "loss": 1.184,
+      "step": 7771
+    },
+    {
+      "epoch": 2.893555630533676,
+      "grad_norm": 0.19858300685882568,
+      "learning_rate": 1.6800157558644034e-05,
+      "loss": 1.1683,
+      "step": 7772
+    },
+    {
+      "epoch": 2.8939279356842853,
+      "grad_norm": 0.18224839866161346,
+      "learning_rate": 1.6799266441551124e-05,
+      "loss": 1.1785,
+      "step": 7773
+    },
+    {
+      "epoch": 2.8943002408348946,
+      "grad_norm": 0.2373683899641037,
+      "learning_rate": 1.679837522403348e-05,
+      "loss": 1.1623,
+      "step": 7774
+    },
+    {
+      "epoch": 2.8946725459855034,
+      "grad_norm": 0.16270266473293304,
+      "learning_rate": 1.6797483906104256e-05,
+      "loss": 1.1622,
+      "step": 7775
+    },
+    {
+      "epoch": 2.895044851136112,
+      "grad_norm": 0.16937381029129028,
+      "learning_rate": 1.679659248777662e-05,
+      "loss": 1.1632,
+      "step": 7776
+    },
+    {
+      "epoch": 2.8954171562867215,
+      "grad_norm": 0.16399821639060974,
+      "learning_rate": 1.6795700969063743e-05,
+      "loss": 1.1614,
+      "step": 7777
+    },
+    {
+      "epoch": 2.8957894614373307,
+      "grad_norm": 0.16400204598903656,
+      "learning_rate": 1.679480934997879e-05,
+      "loss": 1.1718,
+      "step": 7778
+    },
+    {
+      "epoch": 2.8961617665879396,
+      "grad_norm": 0.16825588047504425,
+      "learning_rate": 1.679391763053493e-05,
+      "loss": 1.1615,
+      "step": 7779
+    },
+    {
+      "epoch": 2.896534071738549,
+      "grad_norm": 0.16895346343517303,
+      "learning_rate": 1.679302581074533e-05,
+      "loss": 1.1646,
+      "step": 7780
+    },
+    {
+      "epoch": 2.8969063768891576,
+      "grad_norm": 0.16163942217826843,
+      "learning_rate": 1.6792133890623162e-05,
+      "loss": 1.1582,
+      "step": 7781
+    },
+    {
+      "epoch": 2.897278682039767,
+      "grad_norm": 0.1691490262746811,
+      "learning_rate": 1.6791241870181607e-05,
+      "loss": 1.1644,
+      "step": 7782
+    },
+    {
+      "epoch": 2.897650987190376,
+      "grad_norm": 0.16696293652057648,
+      "learning_rate": 1.6790349749433835e-05,
+      "loss": 1.1835,
+      "step": 7783
+    },
+    {
+      "epoch": 2.898023292340985,
+      "grad_norm": 0.16553637385368347,
+      "learning_rate": 1.678945752839302e-05,
+      "loss": 1.175,
+      "step": 7784
+    },
+    {
+      "epoch": 2.898395597491594,
+      "grad_norm": 0.1722661852836609,
+      "learning_rate": 1.678856520707235e-05,
+      "loss": 1.1799,
+      "step": 7785
+    },
+    {
+      "epoch": 2.898767902642203,
+      "grad_norm": 0.16318213939666748,
+      "learning_rate": 1.6787672785484998e-05,
+      "loss": 1.1511,
+      "step": 7786
+    },
+    {
+      "epoch": 2.8991402077928123,
+      "grad_norm": 0.16159263253211975,
+      "learning_rate": 1.678678026364414e-05,
+      "loss": 1.1715,
+      "step": 7787
+    },
+    {
+      "epoch": 2.899512512943421,
+      "grad_norm": 0.16238828003406525,
+      "learning_rate": 1.6785887641562967e-05,
+      "loss": 1.1655,
+      "step": 7788
+    },
+    {
+      "epoch": 2.8998848180940304,
+      "grad_norm": 0.16337160766124725,
+      "learning_rate": 1.6784994919254652e-05,
+      "loss": 1.1697,
+      "step": 7789
+    },
+    {
+      "epoch": 2.9002571232446392,
+      "grad_norm": 0.16441784799098969,
+      "learning_rate": 1.67841020967324e-05,
+      "loss": 1.1774,
+      "step": 7790
+    },
+    {
+      "epoch": 2.9006294283952485,
+      "grad_norm": 0.16245967149734497,
+      "learning_rate": 1.6783209174009377e-05,
+      "loss": 1.1751,
+      "step": 7791
+    },
+    {
+      "epoch": 2.9010017335458578,
+      "grad_norm": 0.17327164113521576,
+      "learning_rate": 1.678231615109878e-05,
+      "loss": 1.1608,
+      "step": 7792
+    },
+    {
+      "epoch": 2.9013740386964666,
+      "grad_norm": 0.16266658902168274,
+      "learning_rate": 1.6781423028013803e-05,
+      "loss": 1.1599,
+      "step": 7793
+    },
+    {
+      "epoch": 2.9017463438470754,
+      "grad_norm": 0.17774546146392822,
+      "learning_rate": 1.6780529804767632e-05,
+      "loss": 1.1655,
+      "step": 7794
+    },
+    {
+      "epoch": 2.9021186489976847,
+      "grad_norm": 0.18671812117099762,
+      "learning_rate": 1.6779636481373462e-05,
+      "loss": 1.1759,
+      "step": 7795
+    },
+    {
+      "epoch": 2.902490954148294,
+      "grad_norm": 0.1647380143404007,
+      "learning_rate": 1.6778743057844487e-05,
+      "loss": 1.1736,
+      "step": 7796
+    },
+    {
+      "epoch": 2.9028632592989028,
+      "grad_norm": 0.23281651735305786,
+      "learning_rate": 1.67778495341939e-05,
+      "loss": 1.1769,
+      "step": 7797
+    },
+    {
+      "epoch": 2.903235564449512,
+      "grad_norm": 0.1884678155183792,
+      "learning_rate": 1.67769559104349e-05,
+      "loss": 1.1746,
+      "step": 7798
+    },
+    {
+      "epoch": 2.903607869600121,
+      "grad_norm": 0.18494202196598053,
+      "learning_rate": 1.677606218658069e-05,
+      "loss": 1.1601,
+      "step": 7799
+    },
+    {
+      "epoch": 2.90398017475073,
+      "grad_norm": 0.1650589555501938,
+      "learning_rate": 1.6775168362644465e-05,
+      "loss": 1.1903,
+      "step": 7800
+    },
+    {
+      "epoch": 2.9043524799013394,
+      "grad_norm": 0.2014658898115158,
+      "learning_rate": 1.677427443863943e-05,
+      "loss": 1.1632,
+      "step": 7801
+    },
+    {
+      "epoch": 2.904724785051948,
+      "grad_norm": 0.16059400141239166,
+      "learning_rate": 1.6773380414578785e-05,
+      "loss": 1.1651,
+      "step": 7802
+    },
+    {
+      "epoch": 2.905097090202557,
+      "grad_norm": 0.1685784012079239,
+      "learning_rate": 1.6772486290475737e-05,
+      "loss": 1.1869,
+      "step": 7803
+    },
+    {
+      "epoch": 2.9054693953531663,
+      "grad_norm": 0.16995343565940857,
+      "learning_rate": 1.677159206634349e-05,
+      "loss": 1.1793,
+      "step": 7804
+    },
+    {
+      "epoch": 2.9058417005037755,
+      "grad_norm": 0.16553527116775513,
+      "learning_rate": 1.6770697742195256e-05,
+      "loss": 1.151,
+      "step": 7805
+    },
+    {
+      "epoch": 2.9062140056543844,
+      "grad_norm": 0.16535979509353638,
+      "learning_rate": 1.676980331804424e-05,
+      "loss": 1.1719,
+      "step": 7806
+    },
+    {
+      "epoch": 2.9065863108049936,
+      "grad_norm": 0.172554612159729,
+      "learning_rate": 1.6768908793903653e-05,
+      "loss": 1.1577,
+      "step": 7807
+    },
+    {
+      "epoch": 2.9069586159556025,
+      "grad_norm": 0.1674085408449173,
+      "learning_rate": 1.6768014169786712e-05,
+      "loss": 1.1716,
+      "step": 7808
+    },
+    {
+      "epoch": 2.9073309211062117,
+      "grad_norm": 0.17498281598091125,
+      "learning_rate": 1.676711944570662e-05,
+      "loss": 1.1595,
+      "step": 7809
+    },
+    {
+      "epoch": 2.907703226256821,
+      "grad_norm": 0.1705126017332077,
+      "learning_rate": 1.67662246216766e-05,
+      "loss": 1.1731,
+      "step": 7810
+    },
+    {
+      "epoch": 2.90807553140743,
+      "grad_norm": 0.1687796264886856,
+      "learning_rate": 1.676532969770987e-05,
+      "loss": 1.18,
+      "step": 7811
+    },
+    {
+      "epoch": 2.9084478365580386,
+      "grad_norm": 0.17635595798492432,
+      "learning_rate": 1.6764434673819644e-05,
+      "loss": 1.1541,
+      "step": 7812
+    },
+    {
+      "epoch": 2.908820141708648,
+      "grad_norm": 0.17208701372146606,
+      "learning_rate": 1.6763539550019143e-05,
+      "loss": 1.1639,
+      "step": 7813
+    },
+    {
+      "epoch": 2.909192446859257,
+      "grad_norm": 0.172771617770195,
+      "learning_rate": 1.6762644326321586e-05,
+      "loss": 1.1729,
+      "step": 7814
+    },
+    {
+      "epoch": 2.909564752009866,
+      "grad_norm": 0.16505715250968933,
+      "learning_rate": 1.6761749002740195e-05,
+      "loss": 1.1502,
+      "step": 7815
+    },
+    {
+      "epoch": 2.9099370571604752,
+      "grad_norm": 0.17262002825737,
+      "learning_rate": 1.6760853579288196e-05,
+      "loss": 1.1625,
+      "step": 7816
+    },
+    {
+      "epoch": 2.910309362311084,
+      "grad_norm": 0.17353248596191406,
+      "learning_rate": 1.675995805597882e-05,
+      "loss": 1.1678,
+      "step": 7817
+    },
+    {
+      "epoch": 2.9106816674616933,
+      "grad_norm": 0.18139515817165375,
+      "learning_rate": 1.675906243282528e-05,
+      "loss": 1.169,
+      "step": 7818
+    },
+    {
+      "epoch": 2.9110539726123026,
+      "grad_norm": 0.1768987476825714,
+      "learning_rate": 1.6758166709840815e-05,
+      "loss": 1.1685,
+      "step": 7819
+    },
+    {
+      "epoch": 2.9114262777629114,
+      "grad_norm": 0.15893658995628357,
+      "learning_rate": 1.6757270887038653e-05,
+      "loss": 1.1696,
+      "step": 7820
+    },
+    {
+      "epoch": 2.9117985829135202,
+      "grad_norm": 0.19786134362220764,
+      "learning_rate": 1.6756374964432022e-05,
+      "loss": 1.1686,
+      "step": 7821
+    },
+    {
+      "epoch": 2.9121708880641295,
+      "grad_norm": 0.16756854951381683,
+      "learning_rate": 1.675547894203416e-05,
+      "loss": 1.1556,
+      "step": 7822
+    },
+    {
+      "epoch": 2.9125431932147388,
+      "grad_norm": 0.17230384051799774,
+      "learning_rate": 1.6754582819858295e-05,
+      "loss": 1.19,
+      "step": 7823
+    },
+    {
+      "epoch": 2.9129154983653476,
+      "grad_norm": 0.17054490745067596,
+      "learning_rate": 1.6753686597917668e-05,
+      "loss": 1.161,
+      "step": 7824
+    },
+    {
+      "epoch": 2.913287803515957,
+      "grad_norm": 0.16075944900512695,
+      "learning_rate": 1.675279027622551e-05,
+      "loss": 1.1755,
+      "step": 7825
+    },
+    {
+      "epoch": 2.9136601086665657,
+      "grad_norm": 0.17305156588554382,
+      "learning_rate": 1.6751893854795068e-05,
+      "loss": 1.1624,
+      "step": 7826
+    },
+    {
+      "epoch": 2.914032413817175,
+      "grad_norm": 0.17909590899944305,
+      "learning_rate": 1.6750997333639574e-05,
+      "loss": 1.1666,
+      "step": 7827
+    },
+    {
+      "epoch": 2.914404718967784,
+      "grad_norm": 0.16497522592544556,
+      "learning_rate": 1.6750100712772276e-05,
+      "loss": 1.1747,
+      "step": 7828
+    },
+    {
+      "epoch": 2.914777024118393,
+      "grad_norm": 0.17517150938510895,
+      "learning_rate": 1.6749203992206412e-05,
+      "loss": 1.1543,
+      "step": 7829
+    },
+    {
+      "epoch": 2.915149329269002,
+      "grad_norm": 0.17387013137340546,
+      "learning_rate": 1.6748307171955226e-05,
+      "loss": 1.1769,
+      "step": 7830
+    },
+    {
+      "epoch": 2.915521634419611,
+      "grad_norm": 0.1653532087802887,
+      "learning_rate": 1.674741025203197e-05,
+      "loss": 1.1586,
+      "step": 7831
+    },
+    {
+      "epoch": 2.9158939395702204,
+      "grad_norm": 0.17375896871089935,
+      "learning_rate": 1.6746513232449888e-05,
+      "loss": 1.1712,
+      "step": 7832
+    },
+    {
+      "epoch": 2.916266244720829,
+      "grad_norm": 0.16700343787670135,
+      "learning_rate": 1.6745616113222228e-05,
+      "loss": 1.1565,
+      "step": 7833
+    },
+    {
+      "epoch": 2.9166385498714384,
+      "grad_norm": 0.15995582938194275,
+      "learning_rate": 1.6744718894362243e-05,
+      "loss": 1.1726,
+      "step": 7834
+    },
+    {
+      "epoch": 2.9170108550220473,
+      "grad_norm": 0.16936053335666656,
+      "learning_rate": 1.674382157588318e-05,
+      "loss": 1.1607,
+      "step": 7835
+    },
+    {
+      "epoch": 2.9173831601726565,
+      "grad_norm": 0.16181506216526031,
+      "learning_rate": 1.6742924157798302e-05,
+      "loss": 1.1635,
+      "step": 7836
+    },
+    {
+      "epoch": 2.917755465323266,
+      "grad_norm": 0.1655145138502121,
+      "learning_rate": 1.674202664012085e-05,
+      "loss": 1.1594,
+      "step": 7837
+    },
+    {
+      "epoch": 2.9181277704738746,
+      "grad_norm": 0.1637190282344818,
+      "learning_rate": 1.674112902286409e-05,
+      "loss": 1.1595,
+      "step": 7838
+    },
+    {
+      "epoch": 2.918500075624484,
+      "grad_norm": 0.16163207590579987,
+      "learning_rate": 1.674023130604128e-05,
+      "loss": 1.1678,
+      "step": 7839
+    },
+    {
+      "epoch": 2.9188723807750927,
+      "grad_norm": 0.1664661020040512,
+      "learning_rate": 1.6739333489665672e-05,
+      "loss": 1.1553,
+      "step": 7840
+    },
+    {
+      "epoch": 2.919244685925702,
+      "grad_norm": 0.1619381159543991,
+      "learning_rate": 1.6738435573750535e-05,
+      "loss": 1.1539,
+      "step": 7841
+    },
+    {
+      "epoch": 2.919616991076311,
+      "grad_norm": 0.16180214285850525,
+      "learning_rate": 1.6737537558309128e-05,
+      "loss": 1.1754,
+      "step": 7842
+    },
+    {
+      "epoch": 2.91998929622692,
+      "grad_norm": 0.17024828493595123,
+      "learning_rate": 1.6736639443354712e-05,
+      "loss": 1.1597,
+      "step": 7843
+    },
+    {
+      "epoch": 2.920361601377529,
+      "grad_norm": 0.16076843440532684,
+      "learning_rate": 1.6735741228900556e-05,
+      "loss": 1.169,
+      "step": 7844
+    },
+    {
+      "epoch": 2.920733906528138,
+      "grad_norm": 0.16919641196727753,
+      "learning_rate": 1.673484291495992e-05,
+      "loss": 1.179,
+      "step": 7845
+    },
+    {
+      "epoch": 2.9211062116787474,
+      "grad_norm": 0.1667402684688568,
+      "learning_rate": 1.673394450154608e-05,
+      "loss": 1.1683,
+      "step": 7846
+    },
+    {
+      "epoch": 2.921478516829356,
+      "grad_norm": 0.16661867499351501,
+      "learning_rate": 1.6733045988672306e-05,
+      "loss": 1.1734,
+      "step": 7847
+    },
+    {
+      "epoch": 2.9218508219799655,
+      "grad_norm": 0.16996459662914276,
+      "learning_rate": 1.673214737635186e-05,
+      "loss": 1.169,
+      "step": 7848
+    },
+    {
+      "epoch": 2.9222231271305743,
+      "grad_norm": 0.16732630133628845,
+      "learning_rate": 1.6731248664598023e-05,
+      "loss": 1.157,
+      "step": 7849
+    },
+    {
+      "epoch": 2.9225954322811836,
+      "grad_norm": 0.1656404733657837,
+      "learning_rate": 1.6730349853424064e-05,
+      "loss": 1.1684,
+      "step": 7850
+    },
+    {
+      "epoch": 2.9229677374317924,
+      "grad_norm": 0.16698972880840302,
+      "learning_rate": 1.6729450942843256e-05,
+      "loss": 1.1607,
+      "step": 7851
+    },
+    {
+      "epoch": 2.9233400425824017,
+      "grad_norm": 0.18779774010181427,
+      "learning_rate": 1.6728551932868885e-05,
+      "loss": 1.1719,
+      "step": 7852
+    },
+    {
+      "epoch": 2.9237123477330105,
+      "grad_norm": 0.1793741136789322,
+      "learning_rate": 1.6727652823514225e-05,
+      "loss": 1.1573,
+      "step": 7853
+    },
+    {
+      "epoch": 2.9240846528836197,
+      "grad_norm": 0.17160984873771667,
+      "learning_rate": 1.6726753614792555e-05,
+      "loss": 1.1786,
+      "step": 7854
+    },
+    {
+      "epoch": 2.924456958034229,
+      "grad_norm": 0.16810734570026398,
+      "learning_rate": 1.6725854306717155e-05,
+      "loss": 1.1771,
+      "step": 7855
+    },
+    {
+      "epoch": 2.924829263184838,
+      "grad_norm": 0.1664050966501236,
+      "learning_rate": 1.6724954899301308e-05,
+      "loss": 1.1654,
+      "step": 7856
+    },
+    {
+      "epoch": 2.925201568335447,
+      "grad_norm": 0.19625960290431976,
+      "learning_rate": 1.6724055392558302e-05,
+      "loss": 1.1529,
+      "step": 7857
+    },
+    {
+      "epoch": 2.925573873486056,
+      "grad_norm": 0.20255877077579498,
+      "learning_rate": 1.6723155786501414e-05,
+      "loss": 1.1741,
+      "step": 7858
+    },
+    {
+      "epoch": 2.925946178636665,
+      "grad_norm": 0.1825208216905594,
+      "learning_rate": 1.672225608114394e-05,
+      "loss": 1.1686,
+      "step": 7859
+    },
+    {
+      "epoch": 2.926318483787274,
+      "grad_norm": 0.3680337071418762,
+      "learning_rate": 1.672135627649917e-05,
+      "loss": 1.1604,
+      "step": 7860
+    },
+    {
+      "epoch": 2.9266907889378833,
+      "grad_norm": 0.18671971559524536,
+      "learning_rate": 1.672045637258038e-05,
+      "loss": 1.166,
+      "step": 7861
+    },
+    {
+      "epoch": 2.927063094088492,
+      "grad_norm": 0.1835385113954544,
+      "learning_rate": 1.671955636940088e-05,
+      "loss": 1.1751,
+      "step": 7862
+    },
+    {
+      "epoch": 2.9274353992391013,
+      "grad_norm": 0.15858596563339233,
+      "learning_rate": 1.6718656266973952e-05,
+      "loss": 1.1597,
+      "step": 7863
+    },
+    {
+      "epoch": 2.9278077043897106,
+      "grad_norm": 0.16256222128868103,
+      "learning_rate": 1.6717756065312892e-05,
+      "loss": 1.1723,
+      "step": 7864
+    },
+    {
+      "epoch": 2.9281800095403194,
+      "grad_norm": 0.17425408959388733,
+      "learning_rate": 1.6716855764430995e-05,
+      "loss": 1.1597,
+      "step": 7865
+    },
+    {
+      "epoch": 2.9285523146909287,
+      "grad_norm": 0.16741712391376495,
+      "learning_rate": 1.6715955364341563e-05,
+      "loss": 1.168,
+      "step": 7866
+    },
+    {
+      "epoch": 2.9289246198415375,
+      "grad_norm": 0.1664438098669052,
+      "learning_rate": 1.671505486505789e-05,
+      "loss": 1.1784,
+      "step": 7867
+    },
+    {
+      "epoch": 2.9292969249921468,
+      "grad_norm": 0.15664342045783997,
+      "learning_rate": 1.6714154266593277e-05,
+      "loss": 1.1586,
+      "step": 7868
+    },
+    {
+      "epoch": 2.9296692301427556,
+      "grad_norm": 0.1611543446779251,
+      "learning_rate": 1.671325356896103e-05,
+      "loss": 1.1643,
+      "step": 7869
+    },
+    {
+      "epoch": 2.930041535293365,
+      "grad_norm": 0.167099729180336,
+      "learning_rate": 1.6712352772174444e-05,
+      "loss": 1.1622,
+      "step": 7870
+    },
+    {
+      "epoch": 2.9304138404439737,
+      "grad_norm": 0.16537703573703766,
+      "learning_rate": 1.6711451876246833e-05,
+      "loss": 1.1692,
+      "step": 7871
+    },
+    {
+      "epoch": 2.930786145594583,
+      "grad_norm": 0.16481894254684448,
+      "learning_rate": 1.6710550881191498e-05,
+      "loss": 1.1521,
+      "step": 7872
+    },
+    {
+      "epoch": 2.931158450745192,
+      "grad_norm": 0.1629692018032074,
+      "learning_rate": 1.6709649787021748e-05,
+      "loss": 1.1602,
+      "step": 7873
+    },
+    {
+      "epoch": 2.931530755895801,
+      "grad_norm": 0.16847184300422668,
+      "learning_rate": 1.6708748593750888e-05,
+      "loss": 1.1644,
+      "step": 7874
+    },
+    {
+      "epoch": 2.9319030610464103,
+      "grad_norm": 0.1654757410287857,
+      "learning_rate": 1.6707847301392237e-05,
+      "loss": 1.1596,
+      "step": 7875
+    },
+    {
+      "epoch": 2.932275366197019,
+      "grad_norm": 0.16103285551071167,
+      "learning_rate": 1.67069459099591e-05,
+      "loss": 1.1816,
+      "step": 7876
+    },
+    {
+      "epoch": 2.9326476713476284,
+      "grad_norm": 0.15957196056842804,
+      "learning_rate": 1.6706044419464792e-05,
+      "loss": 1.1675,
+      "step": 7877
+    },
+    {
+      "epoch": 2.933019976498237,
+      "grad_norm": 0.16220717132091522,
+      "learning_rate": 1.670514282992263e-05,
+      "loss": 1.1758,
+      "step": 7878
+    },
+    {
+      "epoch": 2.9333922816488465,
+      "grad_norm": 0.16329532861709595,
+      "learning_rate": 1.670424114134593e-05,
+      "loss": 1.1742,
+      "step": 7879
+    },
+    {
+      "epoch": 2.9337645867994553,
+      "grad_norm": 0.16387712955474854,
+      "learning_rate": 1.6703339353748006e-05,
+      "loss": 1.1657,
+      "step": 7880
+    },
+    {
+      "epoch": 2.9341368919500646,
+      "grad_norm": 0.15758801996707916,
+      "learning_rate": 1.6702437467142186e-05,
+      "loss": 1.1537,
+      "step": 7881
+    },
+    {
+      "epoch": 2.934509197100674,
+      "grad_norm": 0.16278088092803955,
+      "learning_rate": 1.6701535481541783e-05,
+      "loss": 1.1738,
+      "step": 7882
+    },
+    {
+      "epoch": 2.9348815022512826,
+      "grad_norm": 0.16199669241905212,
+      "learning_rate": 1.670063339696012e-05,
+      "loss": 1.1741,
+      "step": 7883
+    },
+    {
+      "epoch": 2.935253807401892,
+      "grad_norm": 0.164675772190094,
+      "learning_rate": 1.6699731213410524e-05,
+      "loss": 1.1584,
+      "step": 7884
+    },
+    {
+      "epoch": 2.9356261125525007,
+      "grad_norm": 0.16404102742671967,
+      "learning_rate": 1.6698828930906316e-05,
+      "loss": 1.1697,
+      "step": 7885
+    },
+    {
+      "epoch": 2.93599841770311,
+      "grad_norm": 0.16526629030704498,
+      "learning_rate": 1.6697926549460826e-05,
+      "loss": 1.1623,
+      "step": 7886
+    },
+    {
+      "epoch": 2.9363707228537193,
+      "grad_norm": 0.15449416637420654,
+      "learning_rate": 1.669702406908738e-05,
+      "loss": 1.1646,
+      "step": 7887
+    },
+    {
+      "epoch": 2.936743028004328,
+      "grad_norm": 0.1667950302362442,
+      "learning_rate": 1.6696121489799314e-05,
+      "loss": 1.1745,
+      "step": 7888
+    },
+    {
+      "epoch": 2.937115333154937,
+      "grad_norm": 0.16033466160297394,
+      "learning_rate": 1.669521881160995e-05,
+      "loss": 1.1658,
+      "step": 7889
+    },
+    {
+      "epoch": 2.937487638305546,
+      "grad_norm": 0.1629643738269806,
+      "learning_rate": 1.6694316034532626e-05,
+      "loss": 1.177,
+      "step": 7890
+    },
+    {
+      "epoch": 2.9378599434561554,
+      "grad_norm": 0.15993858873844147,
+      "learning_rate": 1.6693413158580672e-05,
+      "loss": 1.1587,
+      "step": 7891
+    },
+    {
+      "epoch": 2.9382322486067642,
+      "grad_norm": 0.16693513095378876,
+      "learning_rate": 1.6692510183767424e-05,
+      "loss": 1.1738,
+      "step": 7892
+    },
+    {
+      "epoch": 2.9386045537573735,
+      "grad_norm": 0.1641031950712204,
+      "learning_rate": 1.6691607110106223e-05,
+      "loss": 1.1536,
+      "step": 7893
+    },
+    {
+      "epoch": 2.9389768589079823,
+      "grad_norm": 0.15620005130767822,
+      "learning_rate": 1.6690703937610406e-05,
+      "loss": 1.1768,
+      "step": 7894
+    },
+    {
+      "epoch": 2.9393491640585916,
+      "grad_norm": 0.16095943748950958,
+      "learning_rate": 1.668980066629331e-05,
+      "loss": 1.1693,
+      "step": 7895
+    },
+    {
+      "epoch": 2.939721469209201,
+      "grad_norm": 0.1613335907459259,
+      "learning_rate": 1.668889729616828e-05,
+      "loss": 1.1716,
+      "step": 7896
+    },
+    {
+      "epoch": 2.9400937743598097,
+      "grad_norm": 0.1632033735513687,
+      "learning_rate": 1.6687993827248657e-05,
+      "loss": 1.1618,
+      "step": 7897
+    },
+    {
+      "epoch": 2.9404660795104185,
+      "grad_norm": 0.16224397718906403,
+      "learning_rate": 1.6687090259547782e-05,
+      "loss": 1.1716,
+      "step": 7898
+    },
+    {
+      "epoch": 2.9408383846610278,
+      "grad_norm": 0.16363045573234558,
+      "learning_rate": 1.6686186593079003e-05,
+      "loss": 1.1671,
+      "step": 7899
+    },
+    {
+      "epoch": 2.941210689811637,
+      "grad_norm": 0.1638343632221222,
+      "learning_rate": 1.6685282827855672e-05,
+      "loss": 1.1756,
+      "step": 7900
+    },
+    {
+      "epoch": 2.941582994962246,
+      "grad_norm": 0.16511334478855133,
+      "learning_rate": 1.668437896389113e-05,
+      "loss": 1.1802,
+      "step": 7901
+    },
+    {
+      "epoch": 2.941955300112855,
+      "grad_norm": 0.16726627945899963,
+      "learning_rate": 1.6683475001198733e-05,
+      "loss": 1.1729,
+      "step": 7902
+    },
+    {
+      "epoch": 2.942327605263464,
+      "grad_norm": 0.15701860189437866,
+      "learning_rate": 1.6682570939791827e-05,
+      "loss": 1.1546,
+      "step": 7903
+    },
+    {
+      "epoch": 2.942699910414073,
+      "grad_norm": 0.16227418184280396,
+      "learning_rate": 1.668166677968377e-05,
+      "loss": 1.1613,
+      "step": 7904
+    },
+    {
+      "epoch": 2.9430722155646825,
+      "grad_norm": 0.16237196326255798,
+      "learning_rate": 1.668076252088791e-05,
+      "loss": 1.1529,
+      "step": 7905
+    },
+    {
+      "epoch": 2.9434445207152913,
+      "grad_norm": 0.15888628363609314,
+      "learning_rate": 1.6679858163417607e-05,
+      "loss": 1.1825,
+      "step": 7906
+    },
+    {
+      "epoch": 2.9438168258659,
+      "grad_norm": 0.17086133360862732,
+      "learning_rate": 1.667895370728622e-05,
+      "loss": 1.1891,
+      "step": 7907
+    },
+    {
+      "epoch": 2.9441891310165094,
+      "grad_norm": 0.16408465802669525,
+      "learning_rate": 1.667804915250711e-05,
+      "loss": 1.1658,
+      "step": 7908
+    },
+    {
+      "epoch": 2.9445614361671186,
+      "grad_norm": 0.15986572206020355,
+      "learning_rate": 1.6677144499093626e-05,
+      "loss": 1.1581,
+      "step": 7909
+    },
+    {
+      "epoch": 2.9449337413177274,
+      "grad_norm": 0.164417564868927,
+      "learning_rate": 1.667623974705914e-05,
+      "loss": 1.1553,
+      "step": 7910
+    },
+    {
+      "epoch": 2.9453060464683367,
+      "grad_norm": 0.16091735661029816,
+      "learning_rate": 1.6675334896417014e-05,
+      "loss": 1.1755,
+      "step": 7911
+    },
+    {
+      "epoch": 2.9456783516189455,
+      "grad_norm": 0.16419173777103424,
+      "learning_rate": 1.6674429947180607e-05,
+      "loss": 1.169,
+      "step": 7912
+    },
+    {
+      "epoch": 2.946050656769555,
+      "grad_norm": 0.17175257205963135,
+      "learning_rate": 1.667352489936329e-05,
+      "loss": 1.1745,
+      "step": 7913
+    },
+    {
+      "epoch": 2.946422961920164,
+      "grad_norm": 0.16759838163852692,
+      "learning_rate": 1.6672619752978428e-05,
+      "loss": 1.1677,
+      "step": 7914
+    },
+    {
+      "epoch": 2.946795267070773,
+      "grad_norm": 0.16285474598407745,
+      "learning_rate": 1.6671714508039394e-05,
+      "loss": 1.1712,
+      "step": 7915
+    },
+    {
+      "epoch": 2.9471675722213817,
+      "grad_norm": 0.16049782931804657,
+      "learning_rate": 1.6670809164559553e-05,
+      "loss": 1.1513,
+      "step": 7916
+    },
+    {
+      "epoch": 2.947539877371991,
+      "grad_norm": 0.16227254271507263,
+      "learning_rate": 1.666990372255228e-05,
+      "loss": 1.1794,
+      "step": 7917
+    },
+    {
+      "epoch": 2.9479121825226002,
+      "grad_norm": 0.16385634243488312,
+      "learning_rate": 1.6668998182030945e-05,
+      "loss": 1.179,
+      "step": 7918
+    },
+    {
+      "epoch": 2.948284487673209,
+      "grad_norm": 0.15982592105865479,
+      "learning_rate": 1.666809254300893e-05,
+      "loss": 1.1648,
+      "step": 7919
+    },
+    {
+      "epoch": 2.9486567928238183,
+      "grad_norm": 0.16695238649845123,
+      "learning_rate": 1.6667186805499605e-05,
+      "loss": 1.1793,
+      "step": 7920
+    },
+    {
+      "epoch": 2.949029097974427,
+      "grad_norm": 0.1569126844406128,
+      "learning_rate": 1.666628096951635e-05,
+      "loss": 1.165,
+      "step": 7921
+    },
+    {
+      "epoch": 2.9494014031250364,
+      "grad_norm": 0.1580849289894104,
+      "learning_rate": 1.666537503507254e-05,
+      "loss": 1.1664,
+      "step": 7922
+    },
+    {
+      "epoch": 2.9497737082756457,
+      "grad_norm": 0.16597579419612885,
+      "learning_rate": 1.6664469002181562e-05,
+      "loss": 1.1624,
+      "step": 7923
+    },
+    {
+      "epoch": 2.9501460134262545,
+      "grad_norm": 0.16480037569999695,
+      "learning_rate": 1.6663562870856793e-05,
+      "loss": 1.1796,
+      "step": 7924
+    },
+    {
+      "epoch": 2.9505183185768633,
+      "grad_norm": 0.16459214687347412,
+      "learning_rate": 1.6662656641111623e-05,
+      "loss": 1.1583,
+      "step": 7925
+    },
+    {
+      "epoch": 2.9508906237274726,
+      "grad_norm": 0.1645277440547943,
+      "learning_rate": 1.666175031295943e-05,
+      "loss": 1.173,
+      "step": 7926
+    },
+    {
+      "epoch": 2.951262928878082,
+      "grad_norm": 0.16501055657863617,
+      "learning_rate": 1.66608438864136e-05,
+      "loss": 1.1613,
+      "step": 7927
+    },
+    {
+      "epoch": 2.9516352340286907,
+      "grad_norm": 0.16335533559322357,
+      "learning_rate": 1.6659937361487527e-05,
+      "loss": 1.1732,
+      "step": 7928
+    },
+    {
+      "epoch": 2.9520075391793,
+      "grad_norm": 0.16024471819400787,
+      "learning_rate": 1.6659030738194594e-05,
+      "loss": 1.1738,
+      "step": 7929
+    },
+    {
+      "epoch": 2.9523798443299087,
+      "grad_norm": 0.15945962071418762,
+      "learning_rate": 1.66581240165482e-05,
+      "loss": 1.1485,
+      "step": 7930
+    },
+    {
+      "epoch": 2.952752149480518,
+      "grad_norm": 0.15847419202327728,
+      "learning_rate": 1.6657217196561727e-05,
+      "loss": 1.1606,
+      "step": 7931
+    },
+    {
+      "epoch": 2.9531244546311273,
+      "grad_norm": 0.1631278246641159,
+      "learning_rate": 1.6656310278248577e-05,
+      "loss": 1.1659,
+      "step": 7932
+    },
+    {
+      "epoch": 2.953496759781736,
+      "grad_norm": 0.1575775444507599,
+      "learning_rate": 1.6655403261622143e-05,
+      "loss": 1.1682,
+      "step": 7933
+    },
+    {
+      "epoch": 2.953869064932345,
+      "grad_norm": 0.16149340569972992,
+      "learning_rate": 1.6654496146695817e-05,
+      "loss": 1.1578,
+      "step": 7934
+    },
+    {
+      "epoch": 2.954241370082954,
+      "grad_norm": 0.16209939122200012,
+      "learning_rate": 1.6653588933483003e-05,
+      "loss": 1.1566,
+      "step": 7935
+    },
+    {
+      "epoch": 2.9546136752335634,
+      "grad_norm": 0.15919779241085052,
+      "learning_rate": 1.6652681621997095e-05,
+      "loss": 1.1714,
+      "step": 7936
+    },
+    {
+      "epoch": 2.9549859803841723,
+      "grad_norm": 0.15968568623065948,
+      "learning_rate": 1.66517742122515e-05,
+      "loss": 1.1729,
+      "step": 7937
+    },
+    {
+      "epoch": 2.9553582855347815,
+      "grad_norm": 0.16550806164741516,
+      "learning_rate": 1.6650866704259615e-05,
+      "loss": 1.1591,
+      "step": 7938
+    },
+    {
+      "epoch": 2.9557305906853903,
+      "grad_norm": 0.16333597898483276,
+      "learning_rate": 1.664995909803485e-05,
+      "loss": 1.1571,
+      "step": 7939
+    },
+    {
+      "epoch": 2.9561028958359996,
+      "grad_norm": 0.1608911156654358,
+      "learning_rate": 1.6649051393590605e-05,
+      "loss": 1.1608,
+      "step": 7940
+    },
+    {
+      "epoch": 2.956475200986609,
+      "grad_norm": 0.1644187867641449,
+      "learning_rate": 1.6648143590940286e-05,
+      "loss": 1.1719,
+      "step": 7941
+    },
+    {
+      "epoch": 2.9568475061372177,
+      "grad_norm": 0.1601906716823578,
+      "learning_rate": 1.6647235690097303e-05,
+      "loss": 1.1627,
+      "step": 7942
+    },
+    {
+      "epoch": 2.9572198112878265,
+      "grad_norm": 0.15772992372512817,
+      "learning_rate": 1.6646327691075067e-05,
+      "loss": 1.1456,
+      "step": 7943
+    },
+    {
+      "epoch": 2.957592116438436,
+      "grad_norm": 0.15940575301647186,
+      "learning_rate": 1.664541959388699e-05,
+      "loss": 1.1709,
+      "step": 7944
+    },
+    {
+      "epoch": 2.957964421589045,
+      "grad_norm": 0.16511191427707672,
+      "learning_rate": 1.664451139854648e-05,
+      "loss": 1.1746,
+      "step": 7945
+    },
+    {
+      "epoch": 2.958336726739654,
+      "grad_norm": 0.15787945687770844,
+      "learning_rate": 1.6643603105066955e-05,
+      "loss": 1.1639,
+      "step": 7946
+    },
+    {
+      "epoch": 2.958709031890263,
+      "grad_norm": 0.15846529603004456,
+      "learning_rate": 1.664269471346183e-05,
+      "loss": 1.1778,
+      "step": 7947
+    },
+    {
+      "epoch": 2.959081337040872,
+      "grad_norm": 0.16118839383125305,
+      "learning_rate": 1.6641786223744518e-05,
+      "loss": 1.1771,
+      "step": 7948
+    },
+    {
+      "epoch": 2.959453642191481,
+      "grad_norm": 0.16540107131004333,
+      "learning_rate": 1.6640877635928446e-05,
+      "loss": 1.1726,
+      "step": 7949
+    },
+    {
+      "epoch": 2.9598259473420905,
+      "grad_norm": 0.15974678099155426,
+      "learning_rate": 1.6639968950027023e-05,
+      "loss": 1.1715,
+      "step": 7950
+    },
+    {
+      "epoch": 2.9601982524926993,
+      "grad_norm": 0.15811410546302795,
+      "learning_rate": 1.663906016605368e-05,
+      "loss": 1.1567,
+      "step": 7951
+    },
+    {
+      "epoch": 2.960570557643308,
+      "grad_norm": 0.16194650530815125,
+      "learning_rate": 1.6638151284021828e-05,
+      "loss": 1.1735,
+      "step": 7952
+    },
+    {
+      "epoch": 2.9609428627939174,
+      "grad_norm": 0.1610521525144577,
+      "learning_rate": 1.66372423039449e-05,
+      "loss": 1.1847,
+      "step": 7953
+    },
+    {
+      "epoch": 2.9613151679445266,
+      "grad_norm": 0.16063009202480316,
+      "learning_rate": 1.6636333225836323e-05,
+      "loss": 1.1642,
+      "step": 7954
+    },
+    {
+      "epoch": 2.9616874730951355,
+      "grad_norm": 0.16281422972679138,
+      "learning_rate": 1.663542404970952e-05,
+      "loss": 1.1756,
+      "step": 7955
+    },
+    {
+      "epoch": 2.9620597782457447,
+      "grad_norm": 0.15930494666099548,
+      "learning_rate": 1.663451477557792e-05,
+      "loss": 1.1409,
+      "step": 7956
+    },
+    {
+      "epoch": 2.9624320833963536,
+      "grad_norm": 0.16165930032730103,
+      "learning_rate": 1.6633605403454952e-05,
+      "loss": 1.1662,
+      "step": 7957
+    },
+    {
+      "epoch": 2.962804388546963,
+      "grad_norm": 0.16098619997501373,
+      "learning_rate": 1.6632695933354052e-05,
+      "loss": 1.1678,
+      "step": 7958
+    },
+    {
+      "epoch": 2.963176693697572,
+      "grad_norm": 0.16247297823429108,
+      "learning_rate": 1.663178636528864e-05,
+      "loss": 1.1676,
+      "step": 7959
+    },
+    {
+      "epoch": 2.963548998848181,
+      "grad_norm": 0.16938212513923645,
+      "learning_rate": 1.663087669927217e-05,
+      "loss": 1.1518,
+      "step": 7960
+    },
+    {
+      "epoch": 2.96392130399879,
+      "grad_norm": 0.16098153591156006,
+      "learning_rate": 1.6629966935318062e-05,
+      "loss": 1.1678,
+      "step": 7961
+    },
+    {
+      "epoch": 2.964293609149399,
+      "grad_norm": 0.16554594039916992,
+      "learning_rate": 1.662905707343976e-05,
+      "loss": 1.1628,
+      "step": 7962
+    },
+    {
+      "epoch": 2.9646659143000083,
+      "grad_norm": 0.16836988925933838,
+      "learning_rate": 1.6628147113650703e-05,
+      "loss": 1.1883,
+      "step": 7963
+    },
+    {
+      "epoch": 2.965038219450617,
+      "grad_norm": 0.16309542953968048,
+      "learning_rate": 1.6627237055964324e-05,
+      "loss": 1.1766,
+      "step": 7964
+    },
+    {
+      "epoch": 2.9654105246012263,
+      "grad_norm": 0.16153202950954437,
+      "learning_rate": 1.6626326900394073e-05,
+      "loss": 1.1642,
+      "step": 7965
+    },
+    {
+      "epoch": 2.965782829751835,
+      "grad_norm": 0.16180694103240967,
+      "learning_rate": 1.662541664695339e-05,
+      "loss": 1.1684,
+      "step": 7966
+    },
+    {
+      "epoch": 2.9661551349024444,
+      "grad_norm": 0.16211725771427155,
+      "learning_rate": 1.662450629565572e-05,
+      "loss": 1.1634,
+      "step": 7967
+    },
+    {
+      "epoch": 2.9665274400530537,
+      "grad_norm": 0.16405652463436127,
+      "learning_rate": 1.6623595846514503e-05,
+      "loss": 1.1675,
+      "step": 7968
+    },
+    {
+      "epoch": 2.9668997452036625,
+      "grad_norm": 0.1562814861536026,
+      "learning_rate": 1.662268529954319e-05,
+      "loss": 1.1668,
+      "step": 7969
+    },
+    {
+      "epoch": 2.9672720503542718,
+      "grad_norm": 0.161887064576149,
+      "learning_rate": 1.6621774654755238e-05,
+      "loss": 1.1572,
+      "step": 7970
+    },
+    {
+      "epoch": 2.9676443555048806,
+      "grad_norm": 0.16800396144390106,
+      "learning_rate": 1.6620863912164086e-05,
+      "loss": 1.1691,
+      "step": 7971
+    },
+    {
+      "epoch": 2.96801666065549,
+      "grad_norm": 0.16673104465007782,
+      "learning_rate": 1.661995307178319e-05,
+      "loss": 1.1746,
+      "step": 7972
+    },
+    {
+      "epoch": 2.9683889658060987,
+      "grad_norm": 0.17055685818195343,
+      "learning_rate": 1.6619042133626003e-05,
+      "loss": 1.1713,
+      "step": 7973
+    },
+    {
+      "epoch": 2.968761270956708,
+      "grad_norm": 0.16628359258174896,
+      "learning_rate": 1.661813109770598e-05,
+      "loss": 1.1812,
+      "step": 7974
+    },
+    {
+      "epoch": 2.9691335761073168,
+      "grad_norm": 0.16214534640312195,
+      "learning_rate": 1.6617219964036572e-05,
+      "loss": 1.1605,
+      "step": 7975
+    },
+    {
+      "epoch": 2.969505881257926,
+      "grad_norm": 0.16540850698947906,
+      "learning_rate": 1.6616308732631245e-05,
+      "loss": 1.1694,
+      "step": 7976
+    },
+    {
+      "epoch": 2.9698781864085353,
+      "grad_norm": 0.16304931044578552,
+      "learning_rate": 1.6615397403503452e-05,
+      "loss": 1.1558,
+      "step": 7977
+    },
+    {
+      "epoch": 2.970250491559144,
+      "grad_norm": 0.16456028819084167,
+      "learning_rate": 1.661448597666665e-05,
+      "loss": 1.1587,
+      "step": 7978
+    },
+    {
+      "epoch": 2.9706227967097534,
+      "grad_norm": 0.1654767096042633,
+      "learning_rate": 1.6613574452134314e-05,
+      "loss": 1.1568,
+      "step": 7979
+    },
+    {
+      "epoch": 2.970995101860362,
+      "grad_norm": 0.16024528443813324,
+      "learning_rate": 1.6612662829919894e-05,
+      "loss": 1.1774,
+      "step": 7980
+    },
+    {
+      "epoch": 2.9713674070109715,
+      "grad_norm": 0.1553923636674881,
+      "learning_rate": 1.6611751110036856e-05,
+      "loss": 1.1664,
+      "step": 7981
+    },
+    {
+      "epoch": 2.9717397121615803,
+      "grad_norm": 0.16233742237091064,
+      "learning_rate": 1.661083929249867e-05,
+      "loss": 1.166,
+      "step": 7982
+    },
+    {
+      "epoch": 2.9721120173121895,
+      "grad_norm": 0.16320884227752686,
+      "learning_rate": 1.6609927377318804e-05,
+      "loss": 1.1617,
+      "step": 7983
+    },
+    {
+      "epoch": 2.9724843224627984,
+      "grad_norm": 0.16230836510658264,
+      "learning_rate": 1.6609015364510726e-05,
+      "loss": 1.157,
+      "step": 7984
+    },
+    {
+      "epoch": 2.9728566276134076,
+      "grad_norm": 0.16287636756896973,
+      "learning_rate": 1.6608103254087905e-05,
+      "loss": 1.1753,
+      "step": 7985
+    },
+    {
+      "epoch": 2.973228932764017,
+      "grad_norm": 0.16399414837360382,
+      "learning_rate": 1.6607191046063815e-05,
+      "loss": 1.1643,
+      "step": 7986
+    },
+    {
+      "epoch": 2.9736012379146257,
+      "grad_norm": 0.158194899559021,
+      "learning_rate": 1.6606278740451927e-05,
+      "loss": 1.1791,
+      "step": 7987
+    },
+    {
+      "epoch": 2.973973543065235,
+      "grad_norm": 0.16402703523635864,
+      "learning_rate": 1.660536633726572e-05,
+      "loss": 1.1695,
+      "step": 7988
+    },
+    {
+      "epoch": 2.974345848215844,
+      "grad_norm": 0.1630365401506424,
+      "learning_rate": 1.6604453836518658e-05,
+      "loss": 1.1654,
+      "step": 7989
+    },
+    {
+      "epoch": 2.974718153366453,
+      "grad_norm": 0.16490447521209717,
+      "learning_rate": 1.6603541238224235e-05,
+      "loss": 1.1718,
+      "step": 7990
+    },
+    {
+      "epoch": 2.975090458517062,
+      "grad_norm": 0.15666000545024872,
+      "learning_rate": 1.660262854239592e-05,
+      "loss": 1.1588,
+      "step": 7991
+    },
+    {
+      "epoch": 2.975462763667671,
+      "grad_norm": 0.16078460216522217,
+      "learning_rate": 1.6601715749047195e-05,
+      "loss": 1.1669,
+      "step": 7992
+    },
+    {
+      "epoch": 2.97583506881828,
+      "grad_norm": 0.16264961659908295,
+      "learning_rate": 1.6600802858191543e-05,
+      "loss": 1.1733,
+      "step": 7993
+    },
+    {
+      "epoch": 2.9762073739688892,
+      "grad_norm": 0.16248923540115356,
+      "learning_rate": 1.6599889869842447e-05,
+      "loss": 1.1615,
+      "step": 7994
+    },
+    {
+      "epoch": 2.9765796791194985,
+      "grad_norm": 0.16207122802734375,
+      "learning_rate": 1.6598976784013394e-05,
+      "loss": 1.1652,
+      "step": 7995
+    },
+    {
+      "epoch": 2.9769519842701073,
+      "grad_norm": 0.1700456589460373,
+      "learning_rate": 1.6598063600717865e-05,
+      "loss": 1.1734,
+      "step": 7996
+    },
+    {
+      "epoch": 2.9773242894207166,
+      "grad_norm": 0.1640198975801468,
+      "learning_rate": 1.659715031996935e-05,
+      "loss": 1.1636,
+      "step": 7997
+    },
+    {
+      "epoch": 2.9776965945713254,
+      "grad_norm": 0.16763059794902802,
+      "learning_rate": 1.6596236941781342e-05,
+      "loss": 1.1581,
+      "step": 7998
+    },
+    {
+      "epoch": 2.9780688997219347,
+      "grad_norm": 0.16505372524261475,
+      "learning_rate": 1.6595323466167327e-05,
+      "loss": 1.1632,
+      "step": 7999
+    },
+    {
+      "epoch": 2.978441204872544,
+      "grad_norm": 0.16153225302696228,
+      "learning_rate": 1.6594409893140796e-05,
+      "loss": 1.1629,
+      "step": 8000
+    },
+    {
+      "epoch": 2.978441204872544,
+      "eval_loss": 1.2953892946243286,
+      "eval_runtime": 16.6947,
+      "eval_samples_per_second": 103.865,
+      "eval_steps_per_second": 5.211,
+      "step": 8000
+    },
+    {
+      "epoch": 2.9788135100231528,
+      "grad_norm": 0.163248211145401,
+      "learning_rate": 1.659349622271525e-05,
+      "loss": 1.1506,
+      "step": 8001
+    },
+    {
+      "epoch": 2.9791858151737616,
+      "grad_norm": 0.16736340522766113,
+      "learning_rate": 1.6592582454904176e-05,
+      "loss": 1.164,
+      "step": 8002
+    },
+    {
+      "epoch": 2.979558120324371,
+      "grad_norm": 0.164361834526062,
+      "learning_rate": 1.659166858972107e-05,
+      "loss": 1.1693,
+      "step": 8003
+    },
+    {
+      "epoch": 2.97993042547498,
+      "grad_norm": 0.16543278098106384,
+      "learning_rate": 1.6590754627179438e-05,
+      "loss": 1.1647,
+      "step": 8004
+    },
+    {
+      "epoch": 2.980302730625589,
+      "grad_norm": 0.1678323745727539,
+      "learning_rate": 1.658984056729277e-05,
+      "loss": 1.1628,
+      "step": 8005
+    },
+    {
+      "epoch": 2.980675035776198,
+      "grad_norm": 0.16372528672218323,
+      "learning_rate": 1.6588926410074573e-05,
+      "loss": 1.1664,
+      "step": 8006
+    },
+    {
+      "epoch": 2.981047340926807,
+      "grad_norm": 0.1635918915271759,
+      "learning_rate": 1.6588012155538343e-05,
+      "loss": 1.1571,
+      "step": 8007
+    },
+    {
+      "epoch": 2.9814196460774163,
+      "grad_norm": 0.1657022386789322,
+      "learning_rate": 1.658709780369759e-05,
+      "loss": 1.174,
+      "step": 8008
+    },
+    {
+      "epoch": 2.9817919512280255,
+      "grad_norm": 0.1680508553981781,
+      "learning_rate": 1.6586183354565814e-05,
+      "loss": 1.1591,
+      "step": 8009
+    },
+    {
+      "epoch": 2.9821642563786344,
+      "grad_norm": 0.16020500659942627,
+      "learning_rate": 1.658526880815652e-05,
+      "loss": 1.1726,
+      "step": 8010
+    },
+    {
+      "epoch": 2.982536561529243,
+      "grad_norm": 0.16022956371307373,
+      "learning_rate": 1.6584354164483225e-05,
+      "loss": 1.1781,
+      "step": 8011
+    },
+    {
+      "epoch": 2.9829088666798524,
+      "grad_norm": 0.16404716670513153,
+      "learning_rate": 1.658343942355943e-05,
+      "loss": 1.1735,
+      "step": 8012
+    },
+    {
+      "epoch": 2.9832811718304617,
+      "grad_norm": 0.1646728813648224,
+      "learning_rate": 1.6582524585398647e-05,
+      "loss": 1.1829,
+      "step": 8013
+    },
+    {
+      "epoch": 2.9836534769810705,
+      "grad_norm": 0.16547395288944244,
+      "learning_rate": 1.658160965001439e-05,
+      "loss": 1.1801,
+      "step": 8014
+    },
+    {
+      "epoch": 2.98402578213168,
+      "grad_norm": 0.16899387538433075,
+      "learning_rate": 1.6580694617420173e-05,
+      "loss": 1.1598,
+      "step": 8015
+    },
+    {
+      "epoch": 2.9843980872822886,
+      "grad_norm": 0.1721050888299942,
+      "learning_rate": 1.6579779487629508e-05,
+      "loss": 1.165,
+      "step": 8016
+    },
+    {
+      "epoch": 2.984770392432898,
+      "grad_norm": 0.166593998670578,
+      "learning_rate": 1.657886426065591e-05,
+      "loss": 1.1621,
+      "step": 8017
+    },
+    {
+      "epoch": 2.985142697583507,
+      "grad_norm": 0.16592232882976532,
+      "learning_rate": 1.6577948936512905e-05,
+      "loss": 1.1688,
+      "step": 8018
+    },
+    {
+      "epoch": 2.985515002734116,
+      "grad_norm": 0.16835105419158936,
+      "learning_rate": 1.6577033515214e-05,
+      "loss": 1.1917,
+      "step": 8019
+    },
+    {
+      "epoch": 2.985887307884725,
+      "grad_norm": 0.16211645305156708,
+      "learning_rate": 1.6576117996772728e-05,
+      "loss": 1.1697,
+      "step": 8020
+    },
+    {
+      "epoch": 2.986259613035334,
+      "grad_norm": 0.16470582783222198,
+      "learning_rate": 1.657520238120261e-05,
+      "loss": 1.1691,
+      "step": 8021
+    },
+    {
+      "epoch": 2.9866319181859433,
+      "grad_norm": 0.16480398178100586,
+      "learning_rate": 1.6574286668517155e-05,
+      "loss": 1.1564,
+      "step": 8022
+    },
+    {
+      "epoch": 2.987004223336552,
+      "grad_norm": 0.15897324681282043,
+      "learning_rate": 1.6573370858729907e-05,
+      "loss": 1.1565,
+      "step": 8023
+    },
+    {
+      "epoch": 2.9873765284871614,
+      "grad_norm": 0.1609337478876114,
+      "learning_rate": 1.657245495185438e-05,
+      "loss": 1.1762,
+      "step": 8024
+    },
+    {
+      "epoch": 2.98774883363777,
+      "grad_norm": 0.16375748813152313,
+      "learning_rate": 1.6571538947904105e-05,
+      "loss": 1.172,
+      "step": 8025
+    },
+    {
+      "epoch": 2.9881211387883795,
+      "grad_norm": 0.16461491584777832,
+      "learning_rate": 1.6570622846892615e-05,
+      "loss": 1.1717,
+      "step": 8026
+    },
+    {
+      "epoch": 2.9884934439389887,
+      "grad_norm": 0.1679835021495819,
+      "learning_rate": 1.6569706648833436e-05,
+      "loss": 1.1607,
+      "step": 8027
+    },
+    {
+      "epoch": 2.9888657490895976,
+      "grad_norm": 0.1627042144536972,
+      "learning_rate": 1.6568790353740104e-05,
+      "loss": 1.1534,
+      "step": 8028
+    },
+    {
+      "epoch": 2.9892380542402064,
+      "grad_norm": 0.1635371893644333,
+      "learning_rate": 1.656787396162615e-05,
+      "loss": 1.1718,
+      "step": 8029
+    },
+    {
+      "epoch": 2.9896103593908157,
+      "grad_norm": 0.16765522956848145,
+      "learning_rate": 1.656695747250511e-05,
+      "loss": 1.1629,
+      "step": 8030
+    },
+    {
+      "epoch": 2.989982664541425,
+      "grad_norm": 0.16241970658302307,
+      "learning_rate": 1.656604088639052e-05,
+      "loss": 1.1618,
+      "step": 8031
+    },
+    {
+      "epoch": 2.9903549696920337,
+      "grad_norm": 0.16690513491630554,
+      "learning_rate": 1.6565124203295918e-05,
+      "loss": 1.1722,
+      "step": 8032
+    },
+    {
+      "epoch": 2.990727274842643,
+      "grad_norm": 0.15712475776672363,
+      "learning_rate": 1.656420742323484e-05,
+      "loss": 1.1419,
+      "step": 8033
+    },
+    {
+      "epoch": 2.991099579993252,
+      "grad_norm": 0.16157348453998566,
+      "learning_rate": 1.6563290546220835e-05,
+      "loss": 1.1554,
+      "step": 8034
+    },
+    {
+      "epoch": 2.991471885143861,
+      "grad_norm": 0.16115577518939972,
+      "learning_rate": 1.6562373572267438e-05,
+      "loss": 1.167,
+      "step": 8035
+    },
+    {
+      "epoch": 2.9918441902944704,
+      "grad_norm": 0.16360346972942352,
+      "learning_rate": 1.6561456501388197e-05,
+      "loss": 1.1597,
+      "step": 8036
+    },
+    {
+      "epoch": 2.992216495445079,
+      "grad_norm": 0.15973906219005585,
+      "learning_rate": 1.6560539333596657e-05,
+      "loss": 1.1556,
+      "step": 8037
+    },
+    {
+      "epoch": 2.992588800595688,
+      "grad_norm": 0.16520273685455322,
+      "learning_rate": 1.6559622068906357e-05,
+      "loss": 1.1799,
+      "step": 8038
+    },
+    {
+      "epoch": 2.9929611057462973,
+      "grad_norm": 0.1652621626853943,
+      "learning_rate": 1.6558704707330857e-05,
+      "loss": 1.1635,
+      "step": 8039
+    },
+    {
+      "epoch": 2.9933334108969065,
+      "grad_norm": 0.16183042526245117,
+      "learning_rate": 1.6557787248883698e-05,
+      "loss": 1.1567,
+      "step": 8040
+    },
+    {
+      "epoch": 2.9937057160475153,
+      "grad_norm": 0.1610204577445984,
+      "learning_rate": 1.655686969357843e-05,
+      "loss": 1.1709,
+      "step": 8041
+    },
+    {
+      "epoch": 2.9940780211981246,
+      "grad_norm": 0.162316232919693,
+      "learning_rate": 1.655595204142861e-05,
+      "loss": 1.165,
+      "step": 8042
+    },
+    {
+      "epoch": 2.9944503263487334,
+      "grad_norm": 0.1625785082578659,
+      "learning_rate": 1.655503429244779e-05,
+      "loss": 1.1799,
+      "step": 8043
+    },
+    {
+      "epoch": 2.9948226314993427,
+      "grad_norm": 0.15980488061904907,
+      "learning_rate": 1.6554116446649528e-05,
+      "loss": 1.1653,
+      "step": 8044
+    },
+    {
+      "epoch": 2.995194936649952,
+      "grad_norm": 0.16220425069332123,
+      "learning_rate": 1.655319850404737e-05,
+      "loss": 1.1635,
+      "step": 8045
+    },
+    {
+      "epoch": 2.9955672418005608,
+      "grad_norm": 0.1602453589439392,
+      "learning_rate": 1.6552280464654888e-05,
+      "loss": 1.1805,
+      "step": 8046
+    },
+    {
+      "epoch": 2.9959395469511696,
+      "grad_norm": 0.16323374211788177,
+      "learning_rate": 1.6551362328485633e-05,
+      "loss": 1.1555,
+      "step": 8047
+    },
+    {
+      "epoch": 2.996311852101779,
+      "grad_norm": 0.16577033698558807,
+      "learning_rate": 1.6550444095553167e-05,
+      "loss": 1.1722,
+      "step": 8048
+    },
+    {
+      "epoch": 2.996684157252388,
+      "grad_norm": 0.15781056880950928,
+      "learning_rate": 1.654952576587105e-05,
+      "loss": 1.1596,
+      "step": 8049
+    },
+    {
+      "epoch": 2.997056462402997,
+      "grad_norm": 0.16249702870845795,
+      "learning_rate": 1.6548607339452853e-05,
+      "loss": 1.1499,
+      "step": 8050
+    },
+    {
+      "epoch": 2.997428767553606,
+      "grad_norm": 0.16227367520332336,
+      "learning_rate": 1.6547688816312134e-05,
+      "loss": 1.1898,
+      "step": 8051
+    },
+    {
+      "epoch": 2.997801072704215,
+      "grad_norm": 0.16416963934898376,
+      "learning_rate": 1.6546770196462462e-05,
+      "loss": 1.1614,
+      "step": 8052
+    },
+    {
+      "epoch": 2.9981733778548243,
+      "grad_norm": 0.1586388200521469,
+      "learning_rate": 1.654585147991741e-05,
+      "loss": 1.175,
+      "step": 8053
+    },
+    {
+      "epoch": 2.9985456830054336,
+      "grad_norm": 0.16161870956420898,
+      "learning_rate": 1.6544932666690538e-05,
+      "loss": 1.1595,
+      "step": 8054
+    },
+    {
+      "epoch": 2.9989179881560424,
+      "grad_norm": 0.16111552715301514,
+      "learning_rate": 1.654401375679542e-05,
+      "loss": 1.1749,
+      "step": 8055
+    },
+    {
+      "epoch": 2.999290293306651,
+      "grad_norm": 0.16045445203781128,
+      "learning_rate": 1.654309475024563e-05,
+      "loss": 1.1532,
+      "step": 8056
+    },
+    {
+      "epoch": 2.9996625984572605,
+      "grad_norm": 0.16429480910301208,
+      "learning_rate": 1.654217564705474e-05,
+      "loss": 1.1617,
+      "step": 8057
+    },
+    {
+      "epoch": 3.0000349036078697,
+      "grad_norm": 0.16927365958690643,
+      "learning_rate": 1.654125644723633e-05,
+      "loss": 1.1758,
+      "step": 8058
+    },
+    {
+      "epoch": 3.0004072087584785,
+      "grad_norm": 0.15990622341632843,
+      "learning_rate": 1.654033715080397e-05,
+      "loss": 1.1513,
+      "step": 8059
+    },
+    {
+      "epoch": 3.000779513909088,
+      "grad_norm": 0.16184860467910767,
+      "learning_rate": 1.6539417757771246e-05,
+      "loss": 1.1631,
+      "step": 8060
+    },
+    {
+      "epoch": 3.0011518190596966,
+      "grad_norm": 0.16141143441200256,
+      "learning_rate": 1.6538498268151728e-05,
+      "loss": 1.153,
+      "step": 8061
+    },
+    {
+      "epoch": 3.001524124210306,
+      "grad_norm": 0.16540926694869995,
+      "learning_rate": 1.6537578681958998e-05,
+      "loss": 1.1469,
+      "step": 8062
+    },
+    {
+      "epoch": 3.0018964293609147,
+      "grad_norm": 0.18034324049949646,
+      "learning_rate": 1.6536658999206643e-05,
+      "loss": 1.1609,
+      "step": 8063
+    },
+    {
+      "epoch": 3.002268734511524,
+      "grad_norm": 0.15943139791488647,
+      "learning_rate": 1.653573921990825e-05,
+      "loss": 1.1609,
+      "step": 8064
+    },
+    {
+      "epoch": 3.0026410396621332,
+      "grad_norm": 0.16149209439754486,
+      "learning_rate": 1.6534819344077392e-05,
+      "loss": 1.1507,
+      "step": 8065
+    },
+    {
+      "epoch": 3.003013344812742,
+      "grad_norm": 0.1729980707168579,
+      "learning_rate": 1.6533899371727668e-05,
+      "loss": 1.1611,
+      "step": 8066
+    },
+    {
+      "epoch": 3.0033856499633513,
+      "grad_norm": 0.16380807757377625,
+      "learning_rate": 1.6532979302872654e-05,
+      "loss": 1.1694,
+      "step": 8067
+    },
+    {
+      "epoch": 3.00375795511396,
+      "grad_norm": 0.16485844552516937,
+      "learning_rate": 1.653205913752595e-05,
+      "loss": 1.1572,
+      "step": 8068
+    },
+    {
+      "epoch": 3.0041302602645694,
+      "grad_norm": 0.1658172309398651,
+      "learning_rate": 1.6531138875701142e-05,
+      "loss": 1.1508,
+      "step": 8069
+    },
+    {
+      "epoch": 3.0045025654151782,
+      "grad_norm": 0.15986262261867523,
+      "learning_rate": 1.6530218517411823e-05,
+      "loss": 1.1413,
+      "step": 8070
+    },
+    {
+      "epoch": 3.0048748705657875,
+      "grad_norm": 0.16682755947113037,
+      "learning_rate": 1.6529298062671587e-05,
+      "loss": 1.1601,
+      "step": 8071
+    },
+    {
+      "epoch": 3.0052471757163968,
+      "grad_norm": 0.16259485483169556,
+      "learning_rate": 1.6528377511494028e-05,
+      "loss": 1.1513,
+      "step": 8072
+    },
+    {
+      "epoch": 3.0056194808670056,
+      "grad_norm": 0.1639915257692337,
+      "learning_rate": 1.652745686389274e-05,
+      "loss": 1.1732,
+      "step": 8073
+    },
+    {
+      "epoch": 3.005991786017615,
+      "grad_norm": 0.16997599601745605,
+      "learning_rate": 1.6526536119881325e-05,
+      "loss": 1.1559,
+      "step": 8074
+    },
+    {
+      "epoch": 3.0063640911682237,
+      "grad_norm": 0.17218388617038727,
+      "learning_rate": 1.6525615279473385e-05,
+      "loss": 1.1643,
+      "step": 8075
+    },
+    {
+      "epoch": 3.006736396318833,
+      "grad_norm": 0.16318227350711823,
+      "learning_rate": 1.652469434268251e-05,
+      "loss": 1.1595,
+      "step": 8076
+    },
+    {
+      "epoch": 3.0071087014694418,
+      "grad_norm": 0.1637173295021057,
+      "learning_rate": 1.6523773309522314e-05,
+      "loss": 1.1586,
+      "step": 8077
+    },
+    {
+      "epoch": 3.007481006620051,
+      "grad_norm": 0.1644509881734848,
+      "learning_rate": 1.6522852180006396e-05,
+      "loss": 1.168,
+      "step": 8078
+    },
+    {
+      "epoch": 3.00785331177066,
+      "grad_norm": 0.16260574758052826,
+      "learning_rate": 1.6521930954148358e-05,
+      "loss": 1.1534,
+      "step": 8079
+    },
+    {
+      "epoch": 3.008225616921269,
+      "grad_norm": 0.16555775701999664,
+      "learning_rate": 1.652100963196181e-05,
+      "loss": 1.1544,
+      "step": 8080
+    },
+    {
+      "epoch": 3.0085979220718784,
+      "grad_norm": 0.1621171087026596,
+      "learning_rate": 1.652008821346036e-05,
+      "loss": 1.1601,
+      "step": 8081
+    },
+    {
+      "epoch": 3.008970227222487,
+      "grad_norm": 0.15849411487579346,
+      "learning_rate": 1.6519166698657616e-05,
+      "loss": 1.1589,
+      "step": 8082
+    },
+    {
+      "epoch": 3.0093425323730965,
+      "grad_norm": 0.16716186702251434,
+      "learning_rate": 1.6518245087567188e-05,
+      "loss": 1.154,
+      "step": 8083
+    },
+    {
+      "epoch": 3.0097148375237053,
+      "grad_norm": 0.1690070927143097,
+      "learning_rate": 1.6517323380202693e-05,
+      "loss": 1.1483,
+      "step": 8084
+    },
+    {
+      "epoch": 3.0100871426743145,
+      "grad_norm": 0.16490869224071503,
+      "learning_rate": 1.6516401576577736e-05,
+      "loss": 1.1596,
+      "step": 8085
+    },
+    {
+      "epoch": 3.0104594478249234,
+      "grad_norm": 0.16512571275234222,
+      "learning_rate": 1.6515479676705935e-05,
+      "loss": 1.1574,
+      "step": 8086
+    },
+    {
+      "epoch": 3.0108317529755326,
+      "grad_norm": 0.17184092104434967,
+      "learning_rate": 1.6514557680600912e-05,
+      "loss": 1.1579,
+      "step": 8087
+    },
+    {
+      "epoch": 3.0112040581261414,
+      "grad_norm": 0.17025062441825867,
+      "learning_rate": 1.651363558827628e-05,
+      "loss": 1.1661,
+      "step": 8088
+    },
+    {
+      "epoch": 3.0115763632767507,
+      "grad_norm": 0.17025823891162872,
+      "learning_rate": 1.651271339974566e-05,
+      "loss": 1.1601,
+      "step": 8089
+    },
+    {
+      "epoch": 3.01194866842736,
+      "grad_norm": 0.16634193062782288,
+      "learning_rate": 1.6511791115022672e-05,
+      "loss": 1.1519,
+      "step": 8090
+    },
+    {
+      "epoch": 3.012320973577969,
+      "grad_norm": 0.17068715393543243,
+      "learning_rate": 1.6510868734120935e-05,
+      "loss": 1.171,
+      "step": 8091
+    },
+    {
+      "epoch": 3.012693278728578,
+      "grad_norm": 0.1623985916376114,
+      "learning_rate": 1.6509946257054078e-05,
+      "loss": 1.1518,
+      "step": 8092
+    },
+    {
+      "epoch": 3.013065583879187,
+      "grad_norm": 0.16179263591766357,
+      "learning_rate": 1.650902368383572e-05,
+      "loss": 1.1498,
+      "step": 8093
+    },
+    {
+      "epoch": 3.013437889029796,
+      "grad_norm": 0.1658586859703064,
+      "learning_rate": 1.6508101014479494e-05,
+      "loss": 1.1601,
+      "step": 8094
+    },
+    {
+      "epoch": 3.013810194180405,
+      "grad_norm": 0.16824504733085632,
+      "learning_rate": 1.6507178248999026e-05,
+      "loss": 1.1453,
+      "step": 8095
+    },
+    {
+      "epoch": 3.0141824993310142,
+      "grad_norm": 0.16324131190776825,
+      "learning_rate": 1.6506255387407942e-05,
+      "loss": 1.16,
+      "step": 8096
+    },
+    {
+      "epoch": 3.014554804481623,
+      "grad_norm": 0.16698867082595825,
+      "learning_rate": 1.6505332429719872e-05,
+      "loss": 1.1527,
+      "step": 8097
+    },
+    {
+      "epoch": 3.0149271096322323,
+      "grad_norm": 0.16737176477909088,
+      "learning_rate": 1.650440937594845e-05,
+      "loss": 1.162,
+      "step": 8098
+    },
+    {
+      "epoch": 3.0152994147828416,
+      "grad_norm": 0.16450172662734985,
+      "learning_rate": 1.650348622610731e-05,
+      "loss": 1.1524,
+      "step": 8099
+    },
+    {
+      "epoch": 3.0156717199334504,
+      "grad_norm": 0.16533337533473969,
+      "learning_rate": 1.650256298021009e-05,
+      "loss": 1.1577,
+      "step": 8100
+    },
+    {
+      "epoch": 3.0160440250840597,
+      "grad_norm": 0.16812501847743988,
+      "learning_rate": 1.650163963827042e-05,
+      "loss": 1.1594,
+      "step": 8101
+    },
+    {
+      "epoch": 3.0164163302346685,
+      "grad_norm": 0.16462981700897217,
+      "learning_rate": 1.6500716200301943e-05,
+      "loss": 1.1554,
+      "step": 8102
+    },
+    {
+      "epoch": 3.0167886353852778,
+      "grad_norm": 0.1637783944606781,
+      "learning_rate": 1.6499792666318294e-05,
+      "loss": 1.1647,
+      "step": 8103
+    },
+    {
+      "epoch": 3.0171609405358866,
+      "grad_norm": 0.16835294663906097,
+      "learning_rate": 1.6498869036333116e-05,
+      "loss": 1.1474,
+      "step": 8104
+    },
+    {
+      "epoch": 3.017533245686496,
+      "grad_norm": 0.16287805140018463,
+      "learning_rate": 1.649794531036005e-05,
+      "loss": 1.1588,
+      "step": 8105
+    },
+    {
+      "epoch": 3.0179055508371047,
+      "grad_norm": 0.1584957391023636,
+      "learning_rate": 1.649702148841274e-05,
+      "loss": 1.1543,
+      "step": 8106
+    },
+    {
+      "epoch": 3.018277855987714,
+      "grad_norm": 0.15747535228729248,
+      "learning_rate": 1.6496097570504826e-05,
+      "loss": 1.1669,
+      "step": 8107
+    },
+    {
+      "epoch": 3.018650161138323,
+      "grad_norm": 0.16498741507530212,
+      "learning_rate": 1.6495173556649965e-05,
+      "loss": 1.1697,
+      "step": 8108
+    },
+    {
+      "epoch": 3.019022466288932,
+      "grad_norm": 0.16772598028182983,
+      "learning_rate": 1.6494249446861795e-05,
+      "loss": 1.1731,
+      "step": 8109
+    },
+    {
+      "epoch": 3.0193947714395413,
+      "grad_norm": 0.16565538942813873,
+      "learning_rate": 1.6493325241153968e-05,
+      "loss": 1.1572,
+      "step": 8110
+    },
+    {
+      "epoch": 3.01976707659015,
+      "grad_norm": 0.16811548173427582,
+      "learning_rate": 1.6492400939540134e-05,
+      "loss": 1.1409,
+      "step": 8111
+    },
+    {
+      "epoch": 3.0201393817407594,
+      "grad_norm": 0.17630524933338165,
+      "learning_rate": 1.6491476542033948e-05,
+      "loss": 1.1761,
+      "step": 8112
+    },
+    {
+      "epoch": 3.020511686891368,
+      "grad_norm": 0.17578953504562378,
+      "learning_rate": 1.649055204864906e-05,
+      "loss": 1.1435,
+      "step": 8113
+    },
+    {
+      "epoch": 3.0208839920419774,
+      "grad_norm": 0.17409630119800568,
+      "learning_rate": 1.6489627459399123e-05,
+      "loss": 1.1627,
+      "step": 8114
+    },
+    {
+      "epoch": 3.0212562971925863,
+      "grad_norm": 0.16359230875968933,
+      "learning_rate": 1.64887027742978e-05,
+      "loss": 1.1507,
+      "step": 8115
+    },
+    {
+      "epoch": 3.0216286023431955,
+      "grad_norm": 0.16680704057216644,
+      "learning_rate": 1.648777799335874e-05,
+      "loss": 1.1732,
+      "step": 8116
+    },
+    {
+      "epoch": 3.022000907493805,
+      "grad_norm": 0.1649051010608673,
+      "learning_rate": 1.6486853116595608e-05,
+      "loss": 1.1658,
+      "step": 8117
+    },
+    {
+      "epoch": 3.0223732126444136,
+      "grad_norm": 0.1658775508403778,
+      "learning_rate": 1.6485928144022066e-05,
+      "loss": 1.1569,
+      "step": 8118
+    },
+    {
+      "epoch": 3.022745517795023,
+      "grad_norm": 0.17610371112823486,
+      "learning_rate": 1.648500307565177e-05,
+      "loss": 1.1667,
+      "step": 8119
+    },
+    {
+      "epoch": 3.0231178229456317,
+      "grad_norm": 0.17497166991233826,
+      "learning_rate": 1.6484077911498383e-05,
+      "loss": 1.146,
+      "step": 8120
+    },
+    {
+      "epoch": 3.023490128096241,
+      "grad_norm": 0.17234422266483307,
+      "learning_rate": 1.6483152651575575e-05,
+      "loss": 1.1553,
+      "step": 8121
+    },
+    {
+      "epoch": 3.02386243324685,
+      "grad_norm": 0.166042760014534,
+      "learning_rate": 1.6482227295897008e-05,
+      "loss": 1.161,
+      "step": 8122
+    },
+    {
+      "epoch": 3.024234738397459,
+      "grad_norm": 0.16150951385498047,
+      "learning_rate": 1.648130184447635e-05,
+      "loss": 1.1666,
+      "step": 8123
+    },
+    {
+      "epoch": 3.024607043548068,
+      "grad_norm": 0.16276511549949646,
+      "learning_rate": 1.648037629732727e-05,
+      "loss": 1.1454,
+      "step": 8124
+    },
+    {
+      "epoch": 3.024979348698677,
+      "grad_norm": 0.16008317470550537,
+      "learning_rate": 1.6479450654463443e-05,
+      "loss": 1.152,
+      "step": 8125
+    },
+    {
+      "epoch": 3.0253516538492864,
+      "grad_norm": 0.1709553301334381,
+      "learning_rate": 1.6478524915898532e-05,
+      "loss": 1.1614,
+      "step": 8126
+    },
+    {
+      "epoch": 3.025723958999895,
+      "grad_norm": 0.16601483523845673,
+      "learning_rate": 1.6477599081646217e-05,
+      "loss": 1.1536,
+      "step": 8127
+    },
+    {
+      "epoch": 3.0260962641505045,
+      "grad_norm": 0.16305498778820038,
+      "learning_rate": 1.647667315172017e-05,
+      "loss": 1.1581,
+      "step": 8128
+    },
+    {
+      "epoch": 3.0264685693011133,
+      "grad_norm": 0.17328190803527832,
+      "learning_rate": 1.6475747126134066e-05,
+      "loss": 1.1502,
+      "step": 8129
+    },
+    {
+      "epoch": 3.0268408744517226,
+      "grad_norm": 0.20106814801692963,
+      "learning_rate": 1.647482100490158e-05,
+      "loss": 1.1654,
+      "step": 8130
+    },
+    {
+      "epoch": 3.0272131796023314,
+      "grad_norm": 0.18390576541423798,
+      "learning_rate": 1.64738947880364e-05,
+      "loss": 1.1635,
+      "step": 8131
+    },
+    {
+      "epoch": 3.0275854847529406,
+      "grad_norm": 0.16346316039562225,
+      "learning_rate": 1.6472968475552197e-05,
+      "loss": 1.177,
+      "step": 8132
+    },
+    {
+      "epoch": 3.02795778990355,
+      "grad_norm": 0.18944934010505676,
+      "learning_rate": 1.647204206746266e-05,
+      "loss": 1.1598,
+      "step": 8133
+    },
+    {
+      "epoch": 3.0283300950541587,
+      "grad_norm": 0.1777556836605072,
+      "learning_rate": 1.6471115563781467e-05,
+      "loss": 1.1578,
+      "step": 8134
+    },
+    {
+      "epoch": 3.028702400204768,
+      "grad_norm": 0.17949633300304413,
+      "learning_rate": 1.6470188964522296e-05,
+      "loss": 1.1698,
+      "step": 8135
+    },
+    {
+      "epoch": 3.029074705355377,
+      "grad_norm": 0.18838725984096527,
+      "learning_rate": 1.6469262269698846e-05,
+      "loss": 1.1585,
+      "step": 8136
+    },
+    {
+      "epoch": 3.029447010505986,
+      "grad_norm": 0.1662994623184204,
+      "learning_rate": 1.6468335479324796e-05,
+      "loss": 1.1597,
+      "step": 8137
+    },
+    {
+      "epoch": 3.029819315656595,
+      "grad_norm": 0.172524094581604,
+      "learning_rate": 1.646740859341384e-05,
+      "loss": 1.1683,
+      "step": 8138
+    },
+    {
+      "epoch": 3.030191620807204,
+      "grad_norm": 0.17046119272708893,
+      "learning_rate": 1.6466481611979665e-05,
+      "loss": 1.1582,
+      "step": 8139
+    },
+    {
+      "epoch": 3.030563925957813,
+      "grad_norm": 0.18434128165245056,
+      "learning_rate": 1.646555453503596e-05,
+      "loss": 1.1576,
+      "step": 8140
+    },
+    {
+      "epoch": 3.0309362311084223,
+      "grad_norm": 0.16541865468025208,
+      "learning_rate": 1.646462736259642e-05,
+      "loss": 1.1453,
+      "step": 8141
+    },
+    {
+      "epoch": 3.0313085362590315,
+      "grad_norm": 0.1774558275938034,
+      "learning_rate": 1.646370009467474e-05,
+      "loss": 1.1619,
+      "step": 8142
+    },
+    {
+      "epoch": 3.0316808414096403,
+      "grad_norm": 0.16616548597812653,
+      "learning_rate": 1.6462772731284615e-05,
+      "loss": 1.1659,
+      "step": 8143
+    },
+    {
+      "epoch": 3.0320531465602496,
+      "grad_norm": 0.17070108652114868,
+      "learning_rate": 1.6461845272439743e-05,
+      "loss": 1.1528,
+      "step": 8144
+    },
+    {
+      "epoch": 3.0324254517108584,
+      "grad_norm": 0.17470189929008484,
+      "learning_rate": 1.646091771815382e-05,
+      "loss": 1.1824,
+      "step": 8145
+    },
+    {
+      "epoch": 3.0327977568614677,
+      "grad_norm": 0.17306390404701233,
+      "learning_rate": 1.645999006844055e-05,
+      "loss": 1.1623,
+      "step": 8146
+    },
+    {
+      "epoch": 3.0331700620120765,
+      "grad_norm": 0.16506531834602356,
+      "learning_rate": 1.6459062323313634e-05,
+      "loss": 1.1725,
+      "step": 8147
+    },
+    {
+      "epoch": 3.0335423671626858,
+      "grad_norm": 0.17168426513671875,
+      "learning_rate": 1.645813448278677e-05,
+      "loss": 1.1475,
+      "step": 8148
+    },
+    {
+      "epoch": 3.0339146723132946,
+      "grad_norm": 0.16225890815258026,
+      "learning_rate": 1.6457206546873665e-05,
+      "loss": 1.1614,
+      "step": 8149
+    },
+    {
+      "epoch": 3.034286977463904,
+      "grad_norm": 0.16611915826797485,
+      "learning_rate": 1.6456278515588023e-05,
+      "loss": 1.1497,
+      "step": 8150
+    },
+    {
+      "epoch": 3.034659282614513,
+      "grad_norm": 0.174469992518425,
+      "learning_rate": 1.6455350388943555e-05,
+      "loss": 1.17,
+      "step": 8151
+    },
+    {
+      "epoch": 3.035031587765122,
+      "grad_norm": 0.16496939957141876,
+      "learning_rate": 1.6454422166953968e-05,
+      "loss": 1.1392,
+      "step": 8152
+    },
+    {
+      "epoch": 3.035403892915731,
+      "grad_norm": 0.17337270081043243,
+      "learning_rate": 1.6453493849632968e-05,
+      "loss": 1.1587,
+      "step": 8153
+    },
+    {
+      "epoch": 3.03577619806634,
+      "grad_norm": 0.1641514003276825,
+      "learning_rate": 1.6452565436994272e-05,
+      "loss": 1.1646,
+      "step": 8154
+    },
+    {
+      "epoch": 3.0361485032169493,
+      "grad_norm": 0.17081360518932343,
+      "learning_rate": 1.6451636929051587e-05,
+      "loss": 1.145,
+      "step": 8155
+    },
+    {
+      "epoch": 3.036520808367558,
+      "grad_norm": 0.16759049892425537,
+      "learning_rate": 1.645070832581863e-05,
+      "loss": 1.171,
+      "step": 8156
+    },
+    {
+      "epoch": 3.0368931135181674,
+      "grad_norm": 0.170254647731781,
+      "learning_rate": 1.6449779627309113e-05,
+      "loss": 1.142,
+      "step": 8157
+    },
+    {
+      "epoch": 3.037265418668776,
+      "grad_norm": 0.16546176373958588,
+      "learning_rate": 1.644885083353676e-05,
+      "loss": 1.1524,
+      "step": 8158
+    },
+    {
+      "epoch": 3.0376377238193855,
+      "grad_norm": 0.16441422700881958,
+      "learning_rate": 1.6447921944515285e-05,
+      "loss": 1.1519,
+      "step": 8159
+    },
+    {
+      "epoch": 3.0380100289699947,
+      "grad_norm": 0.17892083525657654,
+      "learning_rate": 1.6446992960258404e-05,
+      "loss": 1.169,
+      "step": 8160
+    },
+    {
+      "epoch": 3.0383823341206035,
+      "grad_norm": 0.1632370948791504,
+      "learning_rate": 1.6446063880779845e-05,
+      "loss": 1.1567,
+      "step": 8161
+    },
+    {
+      "epoch": 3.038754639271213,
+      "grad_norm": 0.17492267489433289,
+      "learning_rate": 1.6445134706093325e-05,
+      "loss": 1.1518,
+      "step": 8162
+    },
+    {
+      "epoch": 3.0391269444218216,
+      "grad_norm": 0.17203693091869354,
+      "learning_rate": 1.6444205436212567e-05,
+      "loss": 1.1601,
+      "step": 8163
+    },
+    {
+      "epoch": 3.039499249572431,
+      "grad_norm": 0.17056483030319214,
+      "learning_rate": 1.6443276071151303e-05,
+      "loss": 1.1711,
+      "step": 8164
+    },
+    {
+      "epoch": 3.0398715547230397,
+      "grad_norm": 0.17935492098331451,
+      "learning_rate": 1.6442346610923258e-05,
+      "loss": 1.1439,
+      "step": 8165
+    },
+    {
+      "epoch": 3.040243859873649,
+      "grad_norm": 0.16036418080329895,
+      "learning_rate": 1.6441417055542154e-05,
+      "loss": 1.1444,
+      "step": 8166
+    },
+    {
+      "epoch": 3.040616165024258,
+      "grad_norm": 0.1764068752527237,
+      "learning_rate": 1.6440487405021727e-05,
+      "loss": 1.1411,
+      "step": 8167
+    },
+    {
+      "epoch": 3.040988470174867,
+      "grad_norm": 0.16799296438694,
+      "learning_rate": 1.6439557659375705e-05,
+      "loss": 1.1551,
+      "step": 8168
+    },
+    {
+      "epoch": 3.0413607753254763,
+      "grad_norm": 0.17432555556297302,
+      "learning_rate": 1.6438627818617817e-05,
+      "loss": 1.1627,
+      "step": 8169
+    },
+    {
+      "epoch": 3.041733080476085,
+      "grad_norm": 0.18049269914627075,
+      "learning_rate": 1.6437697882761802e-05,
+      "loss": 1.1532,
+      "step": 8170
+    },
+    {
+      "epoch": 3.0421053856266944,
+      "grad_norm": 0.16729703545570374,
+      "learning_rate": 1.6436767851821395e-05,
+      "loss": 1.1795,
+      "step": 8171
+    },
+    {
+      "epoch": 3.0424776907773032,
+      "grad_norm": 0.1749987155199051,
+      "learning_rate": 1.6435837725810326e-05,
+      "loss": 1.1612,
+      "step": 8172
+    },
+    {
+      "epoch": 3.0428499959279125,
+      "grad_norm": 0.16448847949504852,
+      "learning_rate": 1.6434907504742342e-05,
+      "loss": 1.1616,
+      "step": 8173
+    },
+    {
+      "epoch": 3.0432223010785213,
+      "grad_norm": 0.1609332412481308,
+      "learning_rate": 1.6433977188631177e-05,
+      "loss": 1.1641,
+      "step": 8174
+    },
+    {
+      "epoch": 3.0435946062291306,
+      "grad_norm": 0.1693718433380127,
+      "learning_rate": 1.6433046777490576e-05,
+      "loss": 1.1719,
+      "step": 8175
+    },
+    {
+      "epoch": 3.0439669113797394,
+      "grad_norm": 0.16598600149154663,
+      "learning_rate": 1.643211627133427e-05,
+      "loss": 1.1562,
+      "step": 8176
+    },
+    {
+      "epoch": 3.0443392165303487,
+      "grad_norm": 0.16353359818458557,
+      "learning_rate": 1.6431185670176017e-05,
+      "loss": 1.1626,
+      "step": 8177
+    },
+    {
+      "epoch": 3.044711521680958,
+      "grad_norm": 0.1780644953250885,
+      "learning_rate": 1.6430254974029554e-05,
+      "loss": 1.1502,
+      "step": 8178
+    },
+    {
+      "epoch": 3.0450838268315668,
+      "grad_norm": 0.1959240734577179,
+      "learning_rate": 1.6429324182908628e-05,
+      "loss": 1.1409,
+      "step": 8179
+    },
+    {
+      "epoch": 3.045456131982176,
+      "grad_norm": 0.18989701569080353,
+      "learning_rate": 1.6428393296826987e-05,
+      "loss": 1.1457,
+      "step": 8180
+    },
+    {
+      "epoch": 3.045828437132785,
+      "grad_norm": 0.16791072487831116,
+      "learning_rate": 1.642746231579838e-05,
+      "loss": 1.1521,
+      "step": 8181
+    },
+    {
+      "epoch": 3.046200742283394,
+      "grad_norm": 0.17417898774147034,
+      "learning_rate": 1.642653123983656e-05,
+      "loss": 1.1476,
+      "step": 8182
+    },
+    {
+      "epoch": 3.046573047434003,
+      "grad_norm": 0.20077449083328247,
+      "learning_rate": 1.6425600068955272e-05,
+      "loss": 1.1731,
+      "step": 8183
+    },
+    {
+      "epoch": 3.046945352584612,
+      "grad_norm": 0.1830679327249527,
+      "learning_rate": 1.642466880316828e-05,
+      "loss": 1.166,
+      "step": 8184
+    },
+    {
+      "epoch": 3.0473176577352215,
+      "grad_norm": 0.16485272347927094,
+      "learning_rate": 1.642373744248933e-05,
+      "loss": 1.1681,
+      "step": 8185
+    },
+    {
+      "epoch": 3.0476899628858303,
+      "grad_norm": 0.1660563200712204,
+      "learning_rate": 1.6422805986932184e-05,
+      "loss": 1.1524,
+      "step": 8186
+    },
+    {
+      "epoch": 3.0480622680364395,
+      "grad_norm": 0.16424550116062164,
+      "learning_rate": 1.642187443651059e-05,
+      "loss": 1.1539,
+      "step": 8187
+    },
+    {
+      "epoch": 3.0484345731870484,
+      "grad_norm": 0.1737472116947174,
+      "learning_rate": 1.642094279123832e-05,
+      "loss": 1.1597,
+      "step": 8188
+    },
+    {
+      "epoch": 3.0488068783376576,
+      "grad_norm": 0.1679793894290924,
+      "learning_rate": 1.6420011051129127e-05,
+      "loss": 1.1495,
+      "step": 8189
+    },
+    {
+      "epoch": 3.0491791834882664,
+      "grad_norm": 0.16718783974647522,
+      "learning_rate": 1.641907921619677e-05,
+      "loss": 1.1653,
+      "step": 8190
+    },
+    {
+      "epoch": 3.0495514886388757,
+      "grad_norm": 0.17275018990039825,
+      "learning_rate": 1.641814728645502e-05,
+      "loss": 1.1521,
+      "step": 8191
+    },
+    {
+      "epoch": 3.0499237937894845,
+      "grad_norm": 0.17378568649291992,
+      "learning_rate": 1.6417215261917638e-05,
+      "loss": 1.1681,
+      "step": 8192
+    },
+    {
+      "epoch": 3.050296098940094,
+      "grad_norm": 0.19507519900798798,
+      "learning_rate": 1.6416283142598387e-05,
+      "loss": 1.1686,
+      "step": 8193
+    },
+    {
+      "epoch": 3.050668404090703,
+      "grad_norm": 0.1943422257900238,
+      "learning_rate": 1.6415350928511037e-05,
+      "loss": 1.1595,
+      "step": 8194
+    },
+    {
+      "epoch": 3.051040709241312,
+      "grad_norm": 0.16781660914421082,
+      "learning_rate": 1.6414418619669354e-05,
+      "loss": 1.1471,
+      "step": 8195
+    },
+    {
+      "epoch": 3.051413014391921,
+      "grad_norm": 0.1694183200597763,
+      "learning_rate": 1.6413486216087114e-05,
+      "loss": 1.1439,
+      "step": 8196
+    },
+    {
+      "epoch": 3.05178531954253,
+      "grad_norm": 0.17168377339839935,
+      "learning_rate": 1.6412553717778085e-05,
+      "loss": 1.159,
+      "step": 8197
+    },
+    {
+      "epoch": 3.0521576246931392,
+      "grad_norm": 0.16123619675636292,
+      "learning_rate": 1.6411621124756035e-05,
+      "loss": 1.1616,
+      "step": 8198
+    },
+    {
+      "epoch": 3.052529929843748,
+      "grad_norm": 0.18184486031532288,
+      "learning_rate": 1.641068843703475e-05,
+      "loss": 1.1613,
+      "step": 8199
+    },
+    {
+      "epoch": 3.0529022349943573,
+      "grad_norm": 0.2111905813217163,
+      "learning_rate": 1.6409755654627994e-05,
+      "loss": 1.165,
+      "step": 8200
+    },
+    {
+      "epoch": 3.053274540144966,
+      "grad_norm": 0.1873953491449356,
+      "learning_rate": 1.6408822777549552e-05,
+      "loss": 1.1603,
+      "step": 8201
+    },
+    {
+      "epoch": 3.0536468452955754,
+      "grad_norm": 0.1657257080078125,
+      "learning_rate": 1.64078898058132e-05,
+      "loss": 1.1615,
+      "step": 8202
+    },
+    {
+      "epoch": 3.0540191504461847,
+      "grad_norm": 0.18995431065559387,
+      "learning_rate": 1.6406956739432716e-05,
+      "loss": 1.1451,
+      "step": 8203
+    },
+    {
+      "epoch": 3.0543914555967935,
+      "grad_norm": 0.1653064638376236,
+      "learning_rate": 1.6406023578421884e-05,
+      "loss": 1.1514,
+      "step": 8204
+    },
+    {
+      "epoch": 3.0547637607474027,
+      "grad_norm": 0.19428938627243042,
+      "learning_rate": 1.6405090322794484e-05,
+      "loss": 1.1483,
+      "step": 8205
+    },
+    {
+      "epoch": 3.0551360658980116,
+      "grad_norm": 0.18280021846294403,
+      "learning_rate": 1.6404156972564305e-05,
+      "loss": 1.1456,
+      "step": 8206
+    },
+    {
+      "epoch": 3.055508371048621,
+      "grad_norm": 0.18493548035621643,
+      "learning_rate": 1.6403223527745127e-05,
+      "loss": 1.152,
+      "step": 8207
+    },
+    {
+      "epoch": 3.0558806761992297,
+      "grad_norm": 0.16537559032440186,
+      "learning_rate": 1.6402289988350742e-05,
+      "loss": 1.1563,
+      "step": 8208
+    },
+    {
+      "epoch": 3.056252981349839,
+      "grad_norm": 0.19156776368618011,
+      "learning_rate": 1.6401356354394934e-05,
+      "loss": 1.1445,
+      "step": 8209
+    },
+    {
+      "epoch": 3.0566252865004477,
+      "grad_norm": 0.17017942667007446,
+      "learning_rate": 1.6400422625891493e-05,
+      "loss": 1.1643,
+      "step": 8210
+    },
+    {
+      "epoch": 3.056997591651057,
+      "grad_norm": 0.1759144514799118,
+      "learning_rate": 1.6399488802854214e-05,
+      "loss": 1.1651,
+      "step": 8211
+    },
+    {
+      "epoch": 3.0573698968016663,
+      "grad_norm": 0.17706505954265594,
+      "learning_rate": 1.6398554885296888e-05,
+      "loss": 1.1491,
+      "step": 8212
+    },
+    {
+      "epoch": 3.057742201952275,
+      "grad_norm": 0.16222470998764038,
+      "learning_rate": 1.6397620873233304e-05,
+      "loss": 1.1636,
+      "step": 8213
+    },
+    {
+      "epoch": 3.0581145071028843,
+      "grad_norm": 0.21946337819099426,
+      "learning_rate": 1.6396686766677263e-05,
+      "loss": 1.1623,
+      "step": 8214
+    },
+    {
+      "epoch": 3.058486812253493,
+      "grad_norm": 0.16888217628002167,
+      "learning_rate": 1.639575256564256e-05,
+      "loss": 1.1594,
+      "step": 8215
+    },
+    {
+      "epoch": 3.0588591174041024,
+      "grad_norm": 0.16997283697128296,
+      "learning_rate": 1.6394818270142995e-05,
+      "loss": 1.1507,
+      "step": 8216
+    },
+    {
+      "epoch": 3.0592314225547113,
+      "grad_norm": 0.16864870488643646,
+      "learning_rate": 1.6393883880192362e-05,
+      "loss": 1.1613,
+      "step": 8217
+    },
+    {
+      "epoch": 3.0596037277053205,
+      "grad_norm": 0.16163980960845947,
+      "learning_rate": 1.6392949395804464e-05,
+      "loss": 1.1675,
+      "step": 8218
+    },
+    {
+      "epoch": 3.0599760328559293,
+      "grad_norm": 0.16427217423915863,
+      "learning_rate": 1.639201481699311e-05,
+      "loss": 1.1476,
+      "step": 8219
+    },
+    {
+      "epoch": 3.0603483380065386,
+      "grad_norm": 0.16542796790599823,
+      "learning_rate": 1.6391080143772094e-05,
+      "loss": 1.1446,
+      "step": 8220
+    },
+    {
+      "epoch": 3.060720643157148,
+      "grad_norm": 0.1682531237602234,
+      "learning_rate": 1.639014537615523e-05,
+      "loss": 1.1564,
+      "step": 8221
+    },
+    {
+      "epoch": 3.0610929483077567,
+      "grad_norm": 0.16446895897388458,
+      "learning_rate": 1.6389210514156317e-05,
+      "loss": 1.1661,
+      "step": 8222
+    },
+    {
+      "epoch": 3.061465253458366,
+      "grad_norm": 0.16606532037258148,
+      "learning_rate": 1.6388275557789165e-05,
+      "loss": 1.1658,
+      "step": 8223
+    },
+    {
+      "epoch": 3.0618375586089748,
+      "grad_norm": 0.16892418265342712,
+      "learning_rate": 1.6387340507067584e-05,
+      "loss": 1.1503,
+      "step": 8224
+    },
+    {
+      "epoch": 3.062209863759584,
+      "grad_norm": 0.16336201131343842,
+      "learning_rate": 1.6386405362005385e-05,
+      "loss": 1.1647,
+      "step": 8225
+    },
+    {
+      "epoch": 3.062582168910193,
+      "grad_norm": 0.15720415115356445,
+      "learning_rate": 1.638547012261638e-05,
+      "loss": 1.151,
+      "step": 8226
+    },
+    {
+      "epoch": 3.062954474060802,
+      "grad_norm": 0.1673574596643448,
+      "learning_rate": 1.6384534788914383e-05,
+      "loss": 1.1508,
+      "step": 8227
+    },
+    {
+      "epoch": 3.063326779211411,
+      "grad_norm": 0.1641821265220642,
+      "learning_rate": 1.6383599360913204e-05,
+      "loss": 1.1522,
+      "step": 8228
+    },
+    {
+      "epoch": 3.06369908436202,
+      "grad_norm": 0.17019106447696686,
+      "learning_rate": 1.6382663838626667e-05,
+      "loss": 1.1652,
+      "step": 8229
+    },
+    {
+      "epoch": 3.0640713895126295,
+      "grad_norm": 0.17099004983901978,
+      "learning_rate": 1.6381728222068585e-05,
+      "loss": 1.1716,
+      "step": 8230
+    },
+    {
+      "epoch": 3.0644436946632383,
+      "grad_norm": 0.16094860434532166,
+      "learning_rate": 1.6380792511252775e-05,
+      "loss": 1.1498,
+      "step": 8231
+    },
+    {
+      "epoch": 3.0648159998138476,
+      "grad_norm": 0.16481037437915802,
+      "learning_rate": 1.6379856706193064e-05,
+      "loss": 1.1479,
+      "step": 8232
+    },
+    {
+      "epoch": 3.0651883049644564,
+      "grad_norm": 0.16703465580940247,
+      "learning_rate": 1.6378920806903265e-05,
+      "loss": 1.1656,
+      "step": 8233
+    },
+    {
+      "epoch": 3.0655606101150656,
+      "grad_norm": 0.17205815017223358,
+      "learning_rate": 1.6377984813397212e-05,
+      "loss": 1.1599,
+      "step": 8234
+    },
+    {
+      "epoch": 3.0659329152656745,
+      "grad_norm": 0.1638091653585434,
+      "learning_rate": 1.637704872568872e-05,
+      "loss": 1.1548,
+      "step": 8235
+    },
+    {
+      "epoch": 3.0663052204162837,
+      "grad_norm": 0.16245143115520477,
+      "learning_rate": 1.6376112543791622e-05,
+      "loss": 1.1547,
+      "step": 8236
+    },
+    {
+      "epoch": 3.0666775255668925,
+      "grad_norm": 0.17442138493061066,
+      "learning_rate": 1.6375176267719735e-05,
+      "loss": 1.1526,
+      "step": 8237
+    },
+    {
+      "epoch": 3.067049830717502,
+      "grad_norm": 0.1698945164680481,
+      "learning_rate": 1.63742398974869e-05,
+      "loss": 1.1514,
+      "step": 8238
+    },
+    {
+      "epoch": 3.067422135868111,
+      "grad_norm": 0.17009849846363068,
+      "learning_rate": 1.6373303433106936e-05,
+      "loss": 1.1399,
+      "step": 8239
+    },
+    {
+      "epoch": 3.06779444101872,
+      "grad_norm": 0.1673266440629959,
+      "learning_rate": 1.6372366874593688e-05,
+      "loss": 1.1445,
+      "step": 8240
+    },
+    {
+      "epoch": 3.068166746169329,
+      "grad_norm": 0.16925260424613953,
+      "learning_rate": 1.6371430221960975e-05,
+      "loss": 1.1528,
+      "step": 8241
+    },
+    {
+      "epoch": 3.068539051319938,
+      "grad_norm": 0.17465613782405853,
+      "learning_rate": 1.637049347522264e-05,
+      "loss": 1.159,
+      "step": 8242
+    },
+    {
+      "epoch": 3.0689113564705472,
+      "grad_norm": 0.17698585987091064,
+      "learning_rate": 1.636955663439252e-05,
+      "loss": 1.1602,
+      "step": 8243
+    },
+    {
+      "epoch": 3.069283661621156,
+      "grad_norm": 0.17667868733406067,
+      "learning_rate": 1.6368619699484446e-05,
+      "loss": 1.158,
+      "step": 8244
+    },
+    {
+      "epoch": 3.0696559667717653,
+      "grad_norm": 0.17002210021018982,
+      "learning_rate": 1.6367682670512253e-05,
+      "loss": 1.1616,
+      "step": 8245
+    },
+    {
+      "epoch": 3.070028271922374,
+      "grad_norm": 0.18032929301261902,
+      "learning_rate": 1.636674554748979e-05,
+      "loss": 1.1825,
+      "step": 8246
+    },
+    {
+      "epoch": 3.0704005770729834,
+      "grad_norm": 0.16729597747325897,
+      "learning_rate": 1.6365808330430897e-05,
+      "loss": 1.161,
+      "step": 8247
+    },
+    {
+      "epoch": 3.0707728822235927,
+      "grad_norm": 0.188649520277977,
+      "learning_rate": 1.6364871019349414e-05,
+      "loss": 1.1605,
+      "step": 8248
+    },
+    {
+      "epoch": 3.0711451873742015,
+      "grad_norm": 0.1881103366613388,
+      "learning_rate": 1.6363933614259184e-05,
+      "loss": 1.1416,
+      "step": 8249
+    },
+    {
+      "epoch": 3.0715174925248108,
+      "grad_norm": 0.1650039553642273,
+      "learning_rate": 1.6362996115174056e-05,
+      "loss": 1.1556,
+      "step": 8250
+    },
+    {
+      "epoch": 3.0718897976754196,
+      "grad_norm": 0.19066570699214935,
+      "learning_rate": 1.6362058522107872e-05,
+      "loss": 1.1541,
+      "step": 8251
+    },
+    {
+      "epoch": 3.072262102826029,
+      "grad_norm": 0.19495618343353271,
+      "learning_rate": 1.6361120835074485e-05,
+      "loss": 1.1702,
+      "step": 8252
+    },
+    {
+      "epoch": 3.0726344079766377,
+      "grad_norm": 0.1677856743335724,
+      "learning_rate": 1.636018305408774e-05,
+      "loss": 1.1482,
+      "step": 8253
+    },
+    {
+      "epoch": 3.073006713127247,
+      "grad_norm": 0.22052864730358124,
+      "learning_rate": 1.6359245179161492e-05,
+      "loss": 1.1543,
+      "step": 8254
+    },
+    {
+      "epoch": 3.073379018277856,
+      "grad_norm": 0.17813050746917725,
+      "learning_rate": 1.6358307210309595e-05,
+      "loss": 1.1654,
+      "step": 8255
+    },
+    {
+      "epoch": 3.073751323428465,
+      "grad_norm": 0.18808187544345856,
+      "learning_rate": 1.6357369147545894e-05,
+      "loss": 1.165,
+      "step": 8256
+    },
+    {
+      "epoch": 3.0741236285790743,
+      "grad_norm": 0.16268914937973022,
+      "learning_rate": 1.635643099088425e-05,
+      "loss": 1.1691,
+      "step": 8257
+    },
+    {
+      "epoch": 3.074495933729683,
+      "grad_norm": 0.19553907215595245,
+      "learning_rate": 1.6355492740338523e-05,
+      "loss": 1.1465,
+      "step": 8258
+    },
+    {
+      "epoch": 3.0748682388802924,
+      "grad_norm": 0.1695629507303238,
+      "learning_rate": 1.6354554395922564e-05,
+      "loss": 1.1672,
+      "step": 8259
+    },
+    {
+      "epoch": 3.075240544030901,
+      "grad_norm": 0.1807146817445755,
+      "learning_rate": 1.635361595765024e-05,
+      "loss": 1.1486,
+      "step": 8260
+    },
+    {
+      "epoch": 3.0756128491815105,
+      "grad_norm": 0.1665555089712143,
+      "learning_rate": 1.63526774255354e-05,
+      "loss": 1.1478,
+      "step": 8261
+    },
+    {
+      "epoch": 3.0759851543321193,
+      "grad_norm": 0.17835910618305206,
+      "learning_rate": 1.6351738799591918e-05,
+      "loss": 1.1594,
+      "step": 8262
+    },
+    {
+      "epoch": 3.0763574594827285,
+      "grad_norm": 0.17160485684871674,
+      "learning_rate": 1.635080007983365e-05,
+      "loss": 1.1574,
+      "step": 8263
+    },
+    {
+      "epoch": 3.076729764633338,
+      "grad_norm": 0.16415171325206757,
+      "learning_rate": 1.6349861266274467e-05,
+      "loss": 1.1511,
+      "step": 8264
+    },
+    {
+      "epoch": 3.0771020697839466,
+      "grad_norm": 0.16749624907970428,
+      "learning_rate": 1.6348922358928228e-05,
+      "loss": 1.148,
+      "step": 8265
+    },
+    {
+      "epoch": 3.077474374934556,
+      "grad_norm": 0.1690567582845688,
+      "learning_rate": 1.6347983357808804e-05,
+      "loss": 1.1539,
+      "step": 8266
+    },
+    {
+      "epoch": 3.0778466800851647,
+      "grad_norm": 0.17391081154346466,
+      "learning_rate": 1.6347044262930067e-05,
+      "loss": 1.1471,
+      "step": 8267
+    },
+    {
+      "epoch": 3.078218985235774,
+      "grad_norm": 0.17175255715847015,
+      "learning_rate": 1.6346105074305884e-05,
+      "loss": 1.1618,
+      "step": 8268
+    },
+    {
+      "epoch": 3.078591290386383,
+      "grad_norm": 0.1740272045135498,
+      "learning_rate": 1.6345165791950125e-05,
+      "loss": 1.1542,
+      "step": 8269
+    },
+    {
+      "epoch": 3.078963595536992,
+      "grad_norm": 0.19044634699821472,
+      "learning_rate": 1.634422641587667e-05,
+      "loss": 1.1599,
+      "step": 8270
+    },
+    {
+      "epoch": 3.079335900687601,
+      "grad_norm": 0.1639656275510788,
+      "learning_rate": 1.6343286946099385e-05,
+      "loss": 1.1623,
+      "step": 8271
+    },
+    {
+      "epoch": 3.07970820583821,
+      "grad_norm": 0.17656481266021729,
+      "learning_rate": 1.6342347382632155e-05,
+      "loss": 1.1523,
+      "step": 8272
+    },
+    {
+      "epoch": 3.0800805109888194,
+      "grad_norm": 0.15982165932655334,
+      "learning_rate": 1.6341407725488844e-05,
+      "loss": 1.1475,
+      "step": 8273
+    },
+    {
+      "epoch": 3.0804528161394282,
+      "grad_norm": 0.17611384391784668,
+      "learning_rate": 1.6340467974683344e-05,
+      "loss": 1.1676,
+      "step": 8274
+    },
+    {
+      "epoch": 3.0808251212900375,
+      "grad_norm": 0.17255373299121857,
+      "learning_rate": 1.633952813022953e-05,
+      "loss": 1.1819,
+      "step": 8275
+    },
+    {
+      "epoch": 3.0811974264406463,
+      "grad_norm": 0.17750908434391022,
+      "learning_rate": 1.633858819214128e-05,
+      "loss": 1.1745,
+      "step": 8276
+    },
+    {
+      "epoch": 3.0815697315912556,
+      "grad_norm": 0.21422599256038666,
+      "learning_rate": 1.6337648160432484e-05,
+      "loss": 1.1535,
+      "step": 8277
+    },
+    {
+      "epoch": 3.0819420367418644,
+      "grad_norm": 0.17324034869670868,
+      "learning_rate": 1.633670803511702e-05,
+      "loss": 1.1591,
+      "step": 8278
+    },
+    {
+      "epoch": 3.0823143418924737,
+      "grad_norm": 0.18860980868339539,
+      "learning_rate": 1.6335767816208775e-05,
+      "loss": 1.1505,
+      "step": 8279
+    },
+    {
+      "epoch": 3.0826866470430825,
+      "grad_norm": 0.20710186660289764,
+      "learning_rate": 1.633482750372164e-05,
+      "loss": 1.1695,
+      "step": 8280
+    },
+    {
+      "epoch": 3.0830589521936917,
+      "grad_norm": 0.16789516806602478,
+      "learning_rate": 1.63338870976695e-05,
+      "loss": 1.1677,
+      "step": 8281
+    },
+    {
+      "epoch": 3.083431257344301,
+      "grad_norm": 0.1649201661348343,
+      "learning_rate": 1.6332946598066244e-05,
+      "loss": 1.1482,
+      "step": 8282
+    },
+    {
+      "epoch": 3.08380356249491,
+      "grad_norm": 0.16981996595859528,
+      "learning_rate": 1.6332006004925763e-05,
+      "loss": 1.1704,
+      "step": 8283
+    },
+    {
+      "epoch": 3.084175867645519,
+      "grad_norm": 0.16521765291690826,
+      "learning_rate": 1.6331065318261955e-05,
+      "loss": 1.1556,
+      "step": 8284
+    },
+    {
+      "epoch": 3.084548172796128,
+      "grad_norm": 0.16958005726337433,
+      "learning_rate": 1.6330124538088705e-05,
+      "loss": 1.1542,
+      "step": 8285
+    },
+    {
+      "epoch": 3.084920477946737,
+      "grad_norm": 0.1606731116771698,
+      "learning_rate": 1.6329183664419918e-05,
+      "loss": 1.1508,
+      "step": 8286
+    },
+    {
+      "epoch": 3.085292783097346,
+      "grad_norm": 0.166421040892601,
+      "learning_rate": 1.6328242697269478e-05,
+      "loss": 1.1642,
+      "step": 8287
+    },
+    {
+      "epoch": 3.0856650882479553,
+      "grad_norm": 0.1679898500442505,
+      "learning_rate": 1.6327301636651296e-05,
+      "loss": 1.1504,
+      "step": 8288
+    },
+    {
+      "epoch": 3.086037393398564,
+      "grad_norm": 0.1639619916677475,
+      "learning_rate": 1.6326360482579265e-05,
+      "loss": 1.158,
+      "step": 8289
+    },
+    {
+      "epoch": 3.0864096985491734,
+      "grad_norm": 0.16645826399326324,
+      "learning_rate": 1.6325419235067286e-05,
+      "loss": 1.151,
+      "step": 8290
+    },
+    {
+      "epoch": 3.0867820036997826,
+      "grad_norm": 0.1687537580728531,
+      "learning_rate": 1.6324477894129263e-05,
+      "loss": 1.1579,
+      "step": 8291
+    },
+    {
+      "epoch": 3.0871543088503914,
+      "grad_norm": 0.16648247838020325,
+      "learning_rate": 1.6323536459779098e-05,
+      "loss": 1.1513,
+      "step": 8292
+    },
+    {
+      "epoch": 3.0875266140010007,
+      "grad_norm": 0.16747455298900604,
+      "learning_rate": 1.6322594932030697e-05,
+      "loss": 1.1528,
+      "step": 8293
+    },
+    {
+      "epoch": 3.0878989191516095,
+      "grad_norm": 0.20737069845199585,
+      "learning_rate": 1.632165331089796e-05,
+      "loss": 1.1452,
+      "step": 8294
+    },
+    {
+      "epoch": 3.088271224302219,
+      "grad_norm": 0.3523947298526764,
+      "learning_rate": 1.6320711596394805e-05,
+      "loss": 1.1623,
+      "step": 8295
+    },
+    {
+      "epoch": 3.0886435294528276,
+      "grad_norm": 0.21294717490673065,
+      "learning_rate": 1.6319769788535135e-05,
+      "loss": 1.1556,
+      "step": 8296
+    },
+    {
+      "epoch": 3.089015834603437,
+      "grad_norm": 0.18315008282661438,
+      "learning_rate": 1.6318827887332865e-05,
+      "loss": 1.1677,
+      "step": 8297
+    },
+    {
+      "epoch": 3.089388139754046,
+      "grad_norm": 0.1660303920507431,
+      "learning_rate": 1.6317885892801902e-05,
+      "loss": 1.164,
+      "step": 8298
+    },
+    {
+      "epoch": 3.089760444904655,
+      "grad_norm": 0.1686762571334839,
+      "learning_rate": 1.631694380495616e-05,
+      "loss": 1.1713,
+      "step": 8299
+    },
+    {
+      "epoch": 3.090132750055264,
+      "grad_norm": 0.1759401112794876,
+      "learning_rate": 1.6316001623809557e-05,
+      "loss": 1.1509,
+      "step": 8300
+    },
+    {
+      "epoch": 3.090505055205873,
+      "grad_norm": 0.17869412899017334,
+      "learning_rate": 1.6315059349376002e-05,
+      "loss": 1.1608,
+      "step": 8301
+    },
+    {
+      "epoch": 3.0908773603564823,
+      "grad_norm": 0.17332401871681213,
+      "learning_rate": 1.6314116981669418e-05,
+      "loss": 1.1478,
+      "step": 8302
+    },
+    {
+      "epoch": 3.091249665507091,
+      "grad_norm": 0.1675678789615631,
+      "learning_rate": 1.6313174520703727e-05,
+      "loss": 1.166,
+      "step": 8303
+    },
+    {
+      "epoch": 3.0916219706577004,
+      "grad_norm": 0.1689613312482834,
+      "learning_rate": 1.631223196649284e-05,
+      "loss": 1.1671,
+      "step": 8304
+    },
+    {
+      "epoch": 3.091994275808309,
+      "grad_norm": 0.16656659543514252,
+      "learning_rate": 1.631128931905068e-05,
+      "loss": 1.1671,
+      "step": 8305
+    },
+    {
+      "epoch": 3.0923665809589185,
+      "grad_norm": 0.16922014951705933,
+      "learning_rate": 1.631034657839118e-05,
+      "loss": 1.1645,
+      "step": 8306
+    },
+    {
+      "epoch": 3.0927388861095277,
+      "grad_norm": 0.169945627450943,
+      "learning_rate": 1.6309403744528254e-05,
+      "loss": 1.1565,
+      "step": 8307
+    },
+    {
+      "epoch": 3.0931111912601366,
+      "grad_norm": 0.1612582504749298,
+      "learning_rate": 1.630846081747583e-05,
+      "loss": 1.1604,
+      "step": 8308
+    },
+    {
+      "epoch": 3.093483496410746,
+      "grad_norm": 0.16429011523723602,
+      "learning_rate": 1.6307517797247836e-05,
+      "loss": 1.1472,
+      "step": 8309
+    },
+    {
+      "epoch": 3.0938558015613546,
+      "grad_norm": 0.1718926727771759,
+      "learning_rate": 1.63065746838582e-05,
+      "loss": 1.1568,
+      "step": 8310
+    },
+    {
+      "epoch": 3.094228106711964,
+      "grad_norm": 0.16393272578716278,
+      "learning_rate": 1.6305631477320853e-05,
+      "loss": 1.1553,
+      "step": 8311
+    },
+    {
+      "epoch": 3.0946004118625727,
+      "grad_norm": 0.166838601231575,
+      "learning_rate": 1.6304688177649725e-05,
+      "loss": 1.1584,
+      "step": 8312
+    },
+    {
+      "epoch": 3.094972717013182,
+      "grad_norm": 0.16786356270313263,
+      "learning_rate": 1.6303744784858745e-05,
+      "loss": 1.1563,
+      "step": 8313
+    },
+    {
+      "epoch": 3.095345022163791,
+      "grad_norm": 0.16941292583942413,
+      "learning_rate": 1.6302801298961853e-05,
+      "loss": 1.158,
+      "step": 8314
+    },
+    {
+      "epoch": 3.0957173273144,
+      "grad_norm": 0.16948345303535461,
+      "learning_rate": 1.6301857719972977e-05,
+      "loss": 1.1635,
+      "step": 8315
+    },
+    {
+      "epoch": 3.0960896324650093,
+      "grad_norm": 0.16545285284519196,
+      "learning_rate": 1.6300914047906063e-05,
+      "loss": 1.1388,
+      "step": 8316
+    },
+    {
+      "epoch": 3.096461937615618,
+      "grad_norm": 0.1651238054037094,
+      "learning_rate": 1.6299970282775046e-05,
+      "loss": 1.1709,
+      "step": 8317
+    },
+    {
+      "epoch": 3.0968342427662274,
+      "grad_norm": 0.16604574024677277,
+      "learning_rate": 1.6299026424593858e-05,
+      "loss": 1.1495,
+      "step": 8318
+    },
+    {
+      "epoch": 3.0972065479168363,
+      "grad_norm": 0.16288162767887115,
+      "learning_rate": 1.6298082473376444e-05,
+      "loss": 1.149,
+      "step": 8319
+    },
+    {
+      "epoch": 3.0975788530674455,
+      "grad_norm": 0.16391494870185852,
+      "learning_rate": 1.629713842913675e-05,
+      "loss": 1.1536,
+      "step": 8320
+    },
+    {
+      "epoch": 3.0979511582180543,
+      "grad_norm": 0.16908983886241913,
+      "learning_rate": 1.6296194291888718e-05,
+      "loss": 1.1626,
+      "step": 8321
+    },
+    {
+      "epoch": 3.0983234633686636,
+      "grad_norm": 0.16984610259532928,
+      "learning_rate": 1.6295250061646292e-05,
+      "loss": 1.1611,
+      "step": 8322
+    },
+    {
+      "epoch": 3.0986957685192724,
+      "grad_norm": 0.16279767453670502,
+      "learning_rate": 1.6294305738423413e-05,
+      "loss": 1.1429,
+      "step": 8323
+    },
+    {
+      "epoch": 3.0990680736698817,
+      "grad_norm": 0.16995452344417572,
+      "learning_rate": 1.6293361322234036e-05,
+      "loss": 1.154,
+      "step": 8324
+    },
+    {
+      "epoch": 3.099440378820491,
+      "grad_norm": 0.1628962606191635,
+      "learning_rate": 1.6292416813092107e-05,
+      "loss": 1.1576,
+      "step": 8325
+    },
+    {
+      "epoch": 3.0998126839710998,
+      "grad_norm": 0.16195490956306458,
+      "learning_rate": 1.6291472211011575e-05,
+      "loss": 1.1549,
+      "step": 8326
+    },
+    {
+      "epoch": 3.100184989121709,
+      "grad_norm": 0.16598109900951385,
+      "learning_rate": 1.6290527516006396e-05,
+      "loss": 1.1605,
+      "step": 8327
+    },
+    {
+      "epoch": 3.100557294272318,
+      "grad_norm": 0.16446641087532043,
+      "learning_rate": 1.628958272809052e-05,
+      "loss": 1.163,
+      "step": 8328
+    },
+    {
+      "epoch": 3.100929599422927,
+      "grad_norm": 0.1585415005683899,
+      "learning_rate": 1.62886378472779e-05,
+      "loss": 1.1575,
+      "step": 8329
+    },
+    {
+      "epoch": 3.101301904573536,
+      "grad_norm": 0.16187165677547455,
+      "learning_rate": 1.6287692873582495e-05,
+      "loss": 1.1461,
+      "step": 8330
+    },
+    {
+      "epoch": 3.101674209724145,
+      "grad_norm": 0.1652093529701233,
+      "learning_rate": 1.628674780701826e-05,
+      "loss": 1.1509,
+      "step": 8331
+    },
+    {
+      "epoch": 3.102046514874754,
+      "grad_norm": 0.16078272461891174,
+      "learning_rate": 1.6285802647599156e-05,
+      "loss": 1.1402,
+      "step": 8332
+    },
+    {
+      "epoch": 3.1024188200253633,
+      "grad_norm": 0.1659494936466217,
+      "learning_rate": 1.6284857395339143e-05,
+      "loss": 1.1555,
+      "step": 8333
+    },
+    {
+      "epoch": 3.1027911251759726,
+      "grad_norm": 0.16813765466213226,
+      "learning_rate": 1.6283912050252176e-05,
+      "loss": 1.1623,
+      "step": 8334
+    },
+    {
+      "epoch": 3.1031634303265814,
+      "grad_norm": 0.16813194751739502,
+      "learning_rate": 1.6282966612352224e-05,
+      "loss": 1.1508,
+      "step": 8335
+    },
+    {
+      "epoch": 3.1035357354771906,
+      "grad_norm": 0.16411879658699036,
+      "learning_rate": 1.628202108165325e-05,
+      "loss": 1.1743,
+      "step": 8336
+    },
+    {
+      "epoch": 3.1039080406277995,
+      "grad_norm": 0.17190620303153992,
+      "learning_rate": 1.628107545816922e-05,
+      "loss": 1.1595,
+      "step": 8337
+    },
+    {
+      "epoch": 3.1042803457784087,
+      "grad_norm": 0.16098536550998688,
+      "learning_rate": 1.6280129741914098e-05,
+      "loss": 1.1521,
+      "step": 8338
+    },
+    {
+      "epoch": 3.1046526509290175,
+      "grad_norm": 0.16388416290283203,
+      "learning_rate": 1.6279183932901853e-05,
+      "loss": 1.164,
+      "step": 8339
+    },
+    {
+      "epoch": 3.105024956079627,
+      "grad_norm": 0.16218237578868866,
+      "learning_rate": 1.627823803114646e-05,
+      "loss": 1.1579,
+      "step": 8340
+    },
+    {
+      "epoch": 3.1053972612302356,
+      "grad_norm": 0.16982564330101013,
+      "learning_rate": 1.627729203666188e-05,
+      "loss": 1.1598,
+      "step": 8341
+    },
+    {
+      "epoch": 3.105769566380845,
+      "grad_norm": 0.16258504986763,
+      "learning_rate": 1.627634594946209e-05,
+      "loss": 1.1694,
+      "step": 8342
+    },
+    {
+      "epoch": 3.106141871531454,
+      "grad_norm": 0.16323330998420715,
+      "learning_rate": 1.6275399769561068e-05,
+      "loss": 1.1582,
+      "step": 8343
+    },
+    {
+      "epoch": 3.106514176682063,
+      "grad_norm": 0.16780774295330048,
+      "learning_rate": 1.6274453496972783e-05,
+      "loss": 1.1605,
+      "step": 8344
+    },
+    {
+      "epoch": 3.1068864818326722,
+      "grad_norm": 0.1661432832479477,
+      "learning_rate": 1.6273507131711216e-05,
+      "loss": 1.1553,
+      "step": 8345
+    },
+    {
+      "epoch": 3.107258786983281,
+      "grad_norm": 0.16441033780574799,
+      "learning_rate": 1.627256067379034e-05,
+      "loss": 1.1619,
+      "step": 8346
+    },
+    {
+      "epoch": 3.1076310921338903,
+      "grad_norm": 0.16335833072662354,
+      "learning_rate": 1.6271614123224137e-05,
+      "loss": 1.151,
+      "step": 8347
+    },
+    {
+      "epoch": 3.108003397284499,
+      "grad_norm": 0.16474972665309906,
+      "learning_rate": 1.6270667480026588e-05,
+      "loss": 1.1616,
+      "step": 8348
+    },
+    {
+      "epoch": 3.1083757024351084,
+      "grad_norm": 0.17309875786304474,
+      "learning_rate": 1.6269720744211675e-05,
+      "loss": 1.1736,
+      "step": 8349
+    },
+    {
+      "epoch": 3.1087480075857172,
+      "grad_norm": 0.16838182508945465,
+      "learning_rate": 1.6268773915793376e-05,
+      "loss": 1.1496,
+      "step": 8350
+    },
+    {
+      "epoch": 3.1091203127363265,
+      "grad_norm": 0.1622922271490097,
+      "learning_rate": 1.6267826994785683e-05,
+      "loss": 1.1562,
+      "step": 8351
+    },
+    {
+      "epoch": 3.1094926178869358,
+      "grad_norm": 0.16335055232048035,
+      "learning_rate": 1.6266879981202577e-05,
+      "loss": 1.1565,
+      "step": 8352
+    },
+    {
+      "epoch": 3.1098649230375446,
+      "grad_norm": 0.16523607075214386,
+      "learning_rate": 1.626593287505805e-05,
+      "loss": 1.1497,
+      "step": 8353
+    },
+    {
+      "epoch": 3.110237228188154,
+      "grad_norm": 0.16475920379161835,
+      "learning_rate": 1.6264985676366085e-05,
+      "loss": 1.1481,
+      "step": 8354
+    },
+    {
+      "epoch": 3.1106095333387627,
+      "grad_norm": 0.16444498300552368,
+      "learning_rate": 1.6264038385140676e-05,
+      "loss": 1.1676,
+      "step": 8355
+    },
+    {
+      "epoch": 3.110981838489372,
+      "grad_norm": 0.16481232643127441,
+      "learning_rate": 1.6263091001395808e-05,
+      "loss": 1.1595,
+      "step": 8356
+    },
+    {
+      "epoch": 3.1113541436399808,
+      "grad_norm": 0.16143718361854553,
+      "learning_rate": 1.6262143525145485e-05,
+      "loss": 1.143,
+      "step": 8357
+    },
+    {
+      "epoch": 3.11172644879059,
+      "grad_norm": 0.16786842048168182,
+      "learning_rate": 1.6261195956403694e-05,
+      "loss": 1.159,
+      "step": 8358
+    },
+    {
+      "epoch": 3.112098753941199,
+      "grad_norm": 0.163263738155365,
+      "learning_rate": 1.626024829518443e-05,
+      "loss": 1.1516,
+      "step": 8359
+    },
+    {
+      "epoch": 3.112471059091808,
+      "grad_norm": 0.16883710026741028,
+      "learning_rate": 1.6259300541501694e-05,
+      "loss": 1.1651,
+      "step": 8360
+    },
+    {
+      "epoch": 3.1128433642424174,
+      "grad_norm": 0.16147491335868835,
+      "learning_rate": 1.6258352695369478e-05,
+      "loss": 1.1586,
+      "step": 8361
+    },
+    {
+      "epoch": 3.113215669393026,
+      "grad_norm": 0.15874330699443817,
+      "learning_rate": 1.625740475680179e-05,
+      "loss": 1.1513,
+      "step": 8362
+    },
+    {
+      "epoch": 3.1135879745436355,
+      "grad_norm": 0.16957154870033264,
+      "learning_rate": 1.6256456725812625e-05,
+      "loss": 1.1636,
+      "step": 8363
+    },
+    {
+      "epoch": 3.1139602796942443,
+      "grad_norm": 0.1629595011472702,
+      "learning_rate": 1.6255508602415987e-05,
+      "loss": 1.1588,
+      "step": 8364
+    },
+    {
+      "epoch": 3.1143325848448535,
+      "grad_norm": 0.1647290140390396,
+      "learning_rate": 1.6254560386625874e-05,
+      "loss": 1.1537,
+      "step": 8365
+    },
+    {
+      "epoch": 3.1147048899954624,
+      "grad_norm": 0.1659194529056549,
+      "learning_rate": 1.6253612078456304e-05,
+      "loss": 1.1602,
+      "step": 8366
+    },
+    {
+      "epoch": 3.1150771951460716,
+      "grad_norm": 0.16474999487400055,
+      "learning_rate": 1.625266367792127e-05,
+      "loss": 1.152,
+      "step": 8367
+    },
+    {
+      "epoch": 3.1154495002966804,
+      "grad_norm": 0.1607973873615265,
+      "learning_rate": 1.6251715185034795e-05,
+      "loss": 1.1555,
+      "step": 8368
+    },
+    {
+      "epoch": 3.1158218054472897,
+      "grad_norm": 0.16143812239170074,
+      "learning_rate": 1.625076659981087e-05,
+      "loss": 1.149,
+      "step": 8369
+    },
+    {
+      "epoch": 3.116194110597899,
+      "grad_norm": 0.16376273334026337,
+      "learning_rate": 1.6249817922263518e-05,
+      "loss": 1.1621,
+      "step": 8370
+    },
+    {
+      "epoch": 3.116566415748508,
+      "grad_norm": 0.1660854071378708,
+      "learning_rate": 1.6248869152406745e-05,
+      "loss": 1.1475,
+      "step": 8371
+    },
+    {
+      "epoch": 3.116938720899117,
+      "grad_norm": 0.1652403473854065,
+      "learning_rate": 1.624792029025457e-05,
+      "loss": 1.1496,
+      "step": 8372
+    },
+    {
+      "epoch": 3.117311026049726,
+      "grad_norm": 0.1686633676290512,
+      "learning_rate": 1.6246971335821004e-05,
+      "loss": 1.1589,
+      "step": 8373
+    },
+    {
+      "epoch": 3.117683331200335,
+      "grad_norm": 0.1678876131772995,
+      "learning_rate": 1.6246022289120063e-05,
+      "loss": 1.1533,
+      "step": 8374
+    },
+    {
+      "epoch": 3.118055636350944,
+      "grad_norm": 0.1635279506444931,
+      "learning_rate": 1.6245073150165766e-05,
+      "loss": 1.156,
+      "step": 8375
+    },
+    {
+      "epoch": 3.1184279415015532,
+      "grad_norm": 0.16207703948020935,
+      "learning_rate": 1.624412391897213e-05,
+      "loss": 1.1726,
+      "step": 8376
+    },
+    {
+      "epoch": 3.1188002466521625,
+      "grad_norm": 0.1617921143770218,
+      "learning_rate": 1.6243174595553174e-05,
+      "loss": 1.1607,
+      "step": 8377
+    },
+    {
+      "epoch": 3.1191725518027713,
+      "grad_norm": 0.16754458844661713,
+      "learning_rate": 1.624222517992292e-05,
+      "loss": 1.163,
+      "step": 8378
+    },
+    {
+      "epoch": 3.1195448569533806,
+      "grad_norm": 0.15687395632266998,
+      "learning_rate": 1.6241275672095397e-05,
+      "loss": 1.1651,
+      "step": 8379
+    },
+    {
+      "epoch": 3.1199171621039894,
+      "grad_norm": 0.1605893075466156,
+      "learning_rate": 1.6240326072084617e-05,
+      "loss": 1.1457,
+      "step": 8380
+    },
+    {
+      "epoch": 3.1202894672545987,
+      "grad_norm": 0.16542315483093262,
+      "learning_rate": 1.6239376379904618e-05,
+      "loss": 1.165,
+      "step": 8381
+    },
+    {
+      "epoch": 3.1206617724052075,
+      "grad_norm": 0.16830827295780182,
+      "learning_rate": 1.623842659556942e-05,
+      "loss": 1.1516,
+      "step": 8382
+    },
+    {
+      "epoch": 3.1210340775558167,
+      "grad_norm": 0.1646735519170761,
+      "learning_rate": 1.623747671909305e-05,
+      "loss": 1.1631,
+      "step": 8383
+    },
+    {
+      "epoch": 3.1214063827064256,
+      "grad_norm": 0.16516734659671783,
+      "learning_rate": 1.6236526750489542e-05,
+      "loss": 1.1675,
+      "step": 8384
+    },
+    {
+      "epoch": 3.121778687857035,
+      "grad_norm": 0.16348835825920105,
+      "learning_rate": 1.6235576689772927e-05,
+      "loss": 1.1596,
+      "step": 8385
+    },
+    {
+      "epoch": 3.122150993007644,
+      "grad_norm": 0.17139267921447754,
+      "learning_rate": 1.6234626536957235e-05,
+      "loss": 1.1434,
+      "step": 8386
+    },
+    {
+      "epoch": 3.122523298158253,
+      "grad_norm": 0.16735519468784332,
+      "learning_rate": 1.62336762920565e-05,
+      "loss": 1.1433,
+      "step": 8387
+    },
+    {
+      "epoch": 3.122895603308862,
+      "grad_norm": 0.16513751447200775,
+      "learning_rate": 1.6232725955084756e-05,
+      "loss": 1.1627,
+      "step": 8388
+    },
+    {
+      "epoch": 3.123267908459471,
+      "grad_norm": 0.16176214814186096,
+      "learning_rate": 1.6231775526056044e-05,
+      "loss": 1.1773,
+      "step": 8389
+    },
+    {
+      "epoch": 3.1236402136100803,
+      "grad_norm": 0.16931404173374176,
+      "learning_rate": 1.6230825004984395e-05,
+      "loss": 1.1647,
+      "step": 8390
+    },
+    {
+      "epoch": 3.124012518760689,
+      "grad_norm": 0.16403864324092865,
+      "learning_rate": 1.6229874391883856e-05,
+      "loss": 1.1753,
+      "step": 8391
+    },
+    {
+      "epoch": 3.1243848239112983,
+      "grad_norm": 0.16192291676998138,
+      "learning_rate": 1.6228923686768458e-05,
+      "loss": 1.1377,
+      "step": 8392
+    },
+    {
+      "epoch": 3.124757129061907,
+      "grad_norm": 0.16515354812145233,
+      "learning_rate": 1.622797288965225e-05,
+      "loss": 1.1599,
+      "step": 8393
+    },
+    {
+      "epoch": 3.1251294342125164,
+      "grad_norm": 0.1676434427499771,
+      "learning_rate": 1.6227022000549276e-05,
+      "loss": 1.163,
+      "step": 8394
+    },
+    {
+      "epoch": 3.1255017393631257,
+      "grad_norm": 0.16849081218242645,
+      "learning_rate": 1.6226071019473577e-05,
+      "loss": 1.1566,
+      "step": 8395
+    },
+    {
+      "epoch": 3.1258740445137345,
+      "grad_norm": 0.16453030705451965,
+      "learning_rate": 1.6225119946439196e-05,
+      "loss": 1.1531,
+      "step": 8396
+    },
+    {
+      "epoch": 3.126246349664344,
+      "grad_norm": 0.166966512799263,
+      "learning_rate": 1.622416878146019e-05,
+      "loss": 1.1641,
+      "step": 8397
+    },
+    {
+      "epoch": 3.1266186548149526,
+      "grad_norm": 0.1629679799079895,
+      "learning_rate": 1.6223217524550595e-05,
+      "loss": 1.16,
+      "step": 8398
+    },
+    {
+      "epoch": 3.126990959965562,
+      "grad_norm": 0.16159994900226593,
+      "learning_rate": 1.6222266175724472e-05,
+      "loss": 1.1775,
+      "step": 8399
+    },
+    {
+      "epoch": 3.1273632651161707,
+      "grad_norm": 0.16298353672027588,
+      "learning_rate": 1.6221314734995867e-05,
+      "loss": 1.1664,
+      "step": 8400
+    },
+    {
+      "epoch": 3.12773557026678,
+      "grad_norm": 0.16081856191158295,
+      "learning_rate": 1.622036320237883e-05,
+      "loss": 1.1433,
+      "step": 8401
+    },
+    {
+      "epoch": 3.1281078754173888,
+      "grad_norm": 0.17165522277355194,
+      "learning_rate": 1.6219411577887428e-05,
+      "loss": 1.1592,
+      "step": 8402
+    },
+    {
+      "epoch": 3.128480180567998,
+      "grad_norm": 0.16671797633171082,
+      "learning_rate": 1.62184598615357e-05,
+      "loss": 1.1508,
+      "step": 8403
+    },
+    {
+      "epoch": 3.1288524857186073,
+      "grad_norm": 0.16210219264030457,
+      "learning_rate": 1.6217508053337713e-05,
+      "loss": 1.1595,
+      "step": 8404
+    },
+    {
+      "epoch": 3.129224790869216,
+      "grad_norm": 0.17077292501926422,
+      "learning_rate": 1.6216556153307518e-05,
+      "loss": 1.1484,
+      "step": 8405
+    },
+    {
+      "epoch": 3.1295970960198254,
+      "grad_norm": 0.16143833100795746,
+      "learning_rate": 1.6215604161459183e-05,
+      "loss": 1.1785,
+      "step": 8406
+    },
+    {
+      "epoch": 3.129969401170434,
+      "grad_norm": 0.18213441967964172,
+      "learning_rate": 1.6214652077806764e-05,
+      "loss": 1.1635,
+      "step": 8407
+    },
+    {
+      "epoch": 3.1303417063210435,
+      "grad_norm": 0.16605854034423828,
+      "learning_rate": 1.6213699902364325e-05,
+      "loss": 1.1576,
+      "step": 8408
+    },
+    {
+      "epoch": 3.1307140114716523,
+      "grad_norm": 0.17813484370708466,
+      "learning_rate": 1.6212747635145928e-05,
+      "loss": 1.1615,
+      "step": 8409
+    },
+    {
+      "epoch": 3.1310863166222616,
+      "grad_norm": 0.16783683001995087,
+      "learning_rate": 1.6211795276165635e-05,
+      "loss": 1.1513,
+      "step": 8410
+    },
+    {
+      "epoch": 3.131458621772871,
+      "grad_norm": 0.1840137392282486,
+      "learning_rate": 1.621084282543752e-05,
+      "loss": 1.1596,
+      "step": 8411
+    },
+    {
+      "epoch": 3.1318309269234796,
+      "grad_norm": 0.1757200062274933,
+      "learning_rate": 1.6209890282975644e-05,
+      "loss": 1.1677,
+      "step": 8412
+    },
+    {
+      "epoch": 3.132203232074089,
+      "grad_norm": 0.17295728623867035,
+      "learning_rate": 1.6208937648794076e-05,
+      "loss": 1.1536,
+      "step": 8413
+    },
+    {
+      "epoch": 3.1325755372246977,
+      "grad_norm": 0.1669415831565857,
+      "learning_rate": 1.6207984922906893e-05,
+      "loss": 1.1625,
+      "step": 8414
+    },
+    {
+      "epoch": 3.132947842375307,
+      "grad_norm": 0.18558120727539062,
+      "learning_rate": 1.620703210532816e-05,
+      "loss": 1.1613,
+      "step": 8415
+    },
+    {
+      "epoch": 3.133320147525916,
+      "grad_norm": 0.16660122573375702,
+      "learning_rate": 1.6206079196071952e-05,
+      "loss": 1.1513,
+      "step": 8416
+    },
+    {
+      "epoch": 3.133692452676525,
+      "grad_norm": 0.17779971659183502,
+      "learning_rate": 1.6205126195152345e-05,
+      "loss": 1.1553,
+      "step": 8417
+    },
+    {
+      "epoch": 3.134064757827134,
+      "grad_norm": 0.17245428264141083,
+      "learning_rate": 1.620417310258341e-05,
+      "loss": 1.1591,
+      "step": 8418
+    },
+    {
+      "epoch": 3.134437062977743,
+      "grad_norm": 0.179290309548378,
+      "learning_rate": 1.620321991837923e-05,
+      "loss": 1.1569,
+      "step": 8419
+    },
+    {
+      "epoch": 3.1348093681283524,
+      "grad_norm": 0.16428595781326294,
+      "learning_rate": 1.6202266642553884e-05,
+      "loss": 1.1522,
+      "step": 8420
+    },
+    {
+      "epoch": 3.1351816732789612,
+      "grad_norm": 0.1769377738237381,
+      "learning_rate": 1.6201313275121447e-05,
+      "loss": 1.1554,
+      "step": 8421
+    },
+    {
+      "epoch": 3.1355539784295705,
+      "grad_norm": 0.16289122402668,
+      "learning_rate": 1.6200359816096e-05,
+      "loss": 1.1597,
+      "step": 8422
+    },
+    {
+      "epoch": 3.1359262835801793,
+      "grad_norm": 0.1719742715358734,
+      "learning_rate": 1.619940626549163e-05,
+      "loss": 1.1637,
+      "step": 8423
+    },
+    {
+      "epoch": 3.1362985887307886,
+      "grad_norm": 0.16820377111434937,
+      "learning_rate": 1.619845262332242e-05,
+      "loss": 1.1467,
+      "step": 8424
+    },
+    {
+      "epoch": 3.1366708938813974,
+      "grad_norm": 0.16198331117630005,
+      "learning_rate": 1.619749888960245e-05,
+      "loss": 1.1575,
+      "step": 8425
+    },
+    {
+      "epoch": 3.1370431990320067,
+      "grad_norm": 0.17046979069709778,
+      "learning_rate": 1.6196545064345813e-05,
+      "loss": 1.1583,
+      "step": 8426
+    },
+    {
+      "epoch": 3.1374155041826155,
+      "grad_norm": 0.15980608761310577,
+      "learning_rate": 1.619559114756659e-05,
+      "loss": 1.1496,
+      "step": 8427
+    },
+    {
+      "epoch": 3.1377878093332248,
+      "grad_norm": 0.1674344688653946,
+      "learning_rate": 1.619463713927888e-05,
+      "loss": 1.1611,
+      "step": 8428
+    },
+    {
+      "epoch": 3.138160114483834,
+      "grad_norm": 0.17162348330020905,
+      "learning_rate": 1.6193683039496768e-05,
+      "loss": 1.1675,
+      "step": 8429
+    },
+    {
+      "epoch": 3.138532419634443,
+      "grad_norm": 0.16896787285804749,
+      "learning_rate": 1.6192728848234343e-05,
+      "loss": 1.1662,
+      "step": 8430
+    },
+    {
+      "epoch": 3.138904724785052,
+      "grad_norm": 0.17117677628993988,
+      "learning_rate": 1.6191774565505703e-05,
+      "loss": 1.1653,
+      "step": 8431
+    },
+    {
+      "epoch": 3.139277029935661,
+      "grad_norm": 0.16469299793243408,
+      "learning_rate": 1.619082019132494e-05,
+      "loss": 1.1709,
+      "step": 8432
+    },
+    {
+      "epoch": 3.13964933508627,
+      "grad_norm": 0.18532411754131317,
+      "learning_rate": 1.618986572570615e-05,
+      "loss": 1.1446,
+      "step": 8433
+    },
+    {
+      "epoch": 3.140021640236879,
+      "grad_norm": 0.16900207102298737,
+      "learning_rate": 1.6188911168663433e-05,
+      "loss": 1.1662,
+      "step": 8434
+    },
+    {
+      "epoch": 3.1403939453874883,
+      "grad_norm": 0.17637377977371216,
+      "learning_rate": 1.6187956520210893e-05,
+      "loss": 1.1558,
+      "step": 8435
+    },
+    {
+      "epoch": 3.140766250538097,
+      "grad_norm": 0.16088640689849854,
+      "learning_rate": 1.6187001780362613e-05,
+      "loss": 1.1539,
+      "step": 8436
+    },
+    {
+      "epoch": 3.1411385556887064,
+      "grad_norm": 0.18611860275268555,
+      "learning_rate": 1.6186046949132713e-05,
+      "loss": 1.1575,
+      "step": 8437
+    },
+    {
+      "epoch": 3.1415108608393156,
+      "grad_norm": 0.17658965289592743,
+      "learning_rate": 1.6185092026535286e-05,
+      "loss": 1.1652,
+      "step": 8438
+    },
+    {
+      "epoch": 3.1418831659899245,
+      "grad_norm": 0.19135300815105438,
+      "learning_rate": 1.6184137012584434e-05,
+      "loss": 1.1432,
+      "step": 8439
+    },
+    {
+      "epoch": 3.1422554711405337,
+      "grad_norm": 0.16602368652820587,
+      "learning_rate": 1.618318190729427e-05,
+      "loss": 1.1599,
+      "step": 8440
+    },
+    {
+      "epoch": 3.1426277762911425,
+      "grad_norm": 0.20455126464366913,
+      "learning_rate": 1.6182226710678898e-05,
+      "loss": 1.1602,
+      "step": 8441
+    },
+    {
+      "epoch": 3.143000081441752,
+      "grad_norm": 0.18047231435775757,
+      "learning_rate": 1.6181271422752424e-05,
+      "loss": 1.1531,
+      "step": 8442
+    },
+    {
+      "epoch": 3.1433723865923606,
+      "grad_norm": 0.17739862203598022,
+      "learning_rate": 1.6180316043528957e-05,
+      "loss": 1.1533,
+      "step": 8443
+    },
+    {
+      "epoch": 3.14374469174297,
+      "grad_norm": 0.16755236685276031,
+      "learning_rate": 1.617936057302261e-05,
+      "loss": 1.1491,
+      "step": 8444
+    },
+    {
+      "epoch": 3.1441169968935787,
+      "grad_norm": 0.18111351132392883,
+      "learning_rate": 1.61784050112475e-05,
+      "loss": 1.156,
+      "step": 8445
+    },
+    {
+      "epoch": 3.144489302044188,
+      "grad_norm": 0.1638672947883606,
+      "learning_rate": 1.617744935821773e-05,
+      "loss": 1.1649,
+      "step": 8446
+    },
+    {
+      "epoch": 3.1448616071947972,
+      "grad_norm": 0.17401021718978882,
+      "learning_rate": 1.6176493613947425e-05,
+      "loss": 1.1624,
+      "step": 8447
+    },
+    {
+      "epoch": 3.145233912345406,
+      "grad_norm": 0.16585645079612732,
+      "learning_rate": 1.6175537778450694e-05,
+      "loss": 1.1537,
+      "step": 8448
+    },
+    {
+      "epoch": 3.1456062174960153,
+      "grad_norm": 0.18722714483737946,
+      "learning_rate": 1.6174581851741658e-05,
+      "loss": 1.1663,
+      "step": 8449
+    },
+    {
+      "epoch": 3.145978522646624,
+      "grad_norm": 0.16669632494449615,
+      "learning_rate": 1.6173625833834438e-05,
+      "loss": 1.143,
+      "step": 8450
+    },
+    {
+      "epoch": 3.1463508277972334,
+      "grad_norm": 0.1624676138162613,
+      "learning_rate": 1.617266972474315e-05,
+      "loss": 1.1485,
+      "step": 8451
+    },
+    {
+      "epoch": 3.1467231329478422,
+      "grad_norm": 0.16883505880832672,
+      "learning_rate": 1.617171352448192e-05,
+      "loss": 1.1538,
+      "step": 8452
+    },
+    {
+      "epoch": 3.1470954380984515,
+      "grad_norm": 0.16919681429862976,
+      "learning_rate": 1.6170757233064863e-05,
+      "loss": 1.1483,
+      "step": 8453
+    },
+    {
+      "epoch": 3.1474677432490603,
+      "grad_norm": 0.18415765464305878,
+      "learning_rate": 1.6169800850506113e-05,
+      "loss": 1.1692,
+      "step": 8454
+    },
+    {
+      "epoch": 3.1478400483996696,
+      "grad_norm": 0.16374878585338593,
+      "learning_rate": 1.616884437681979e-05,
+      "loss": 1.146,
+      "step": 8455
+    },
+    {
+      "epoch": 3.148212353550279,
+      "grad_norm": 0.16538135707378387,
+      "learning_rate": 1.6167887812020023e-05,
+      "loss": 1.1619,
+      "step": 8456
+    },
+    {
+      "epoch": 3.1485846587008877,
+      "grad_norm": 0.1735272854566574,
+      "learning_rate": 1.6166931156120943e-05,
+      "loss": 1.1598,
+      "step": 8457
+    },
+    {
+      "epoch": 3.148956963851497,
+      "grad_norm": 0.16497787833213806,
+      "learning_rate": 1.6165974409136673e-05,
+      "loss": 1.1705,
+      "step": 8458
+    },
+    {
+      "epoch": 3.1493292690021057,
+      "grad_norm": 0.1695416122674942,
+      "learning_rate": 1.6165017571081348e-05,
+      "loss": 1.1711,
+      "step": 8459
+    },
+    {
+      "epoch": 3.149701574152715,
+      "grad_norm": 0.1687697172164917,
+      "learning_rate": 1.6164060641969104e-05,
+      "loss": 1.1484,
+      "step": 8460
+    },
+    {
+      "epoch": 3.150073879303324,
+      "grad_norm": 0.16305190324783325,
+      "learning_rate": 1.6163103621814065e-05,
+      "loss": 1.1711,
+      "step": 8461
+    },
+    {
+      "epoch": 3.150446184453933,
+      "grad_norm": 0.16032063961029053,
+      "learning_rate": 1.616214651063038e-05,
+      "loss": 1.1599,
+      "step": 8462
+    },
+    {
+      "epoch": 3.150818489604542,
+      "grad_norm": 0.17928291857242584,
+      "learning_rate": 1.6161189308432174e-05,
+      "loss": 1.1642,
+      "step": 8463
+    },
+    {
+      "epoch": 3.151190794755151,
+      "grad_norm": 0.170723557472229,
+      "learning_rate": 1.616023201523359e-05,
+      "loss": 1.1557,
+      "step": 8464
+    },
+    {
+      "epoch": 3.1515630999057604,
+      "grad_norm": 0.17493624985218048,
+      "learning_rate": 1.6159274631048763e-05,
+      "loss": 1.137,
+      "step": 8465
+    },
+    {
+      "epoch": 3.1519354050563693,
+      "grad_norm": 0.19886091351509094,
+      "learning_rate": 1.6158317155891837e-05,
+      "loss": 1.1471,
+      "step": 8466
+    },
+    {
+      "epoch": 3.1523077102069785,
+      "grad_norm": 0.1778782457113266,
+      "learning_rate": 1.6157359589776952e-05,
+      "loss": 1.1636,
+      "step": 8467
+    },
+    {
+      "epoch": 3.1526800153575874,
+      "grad_norm": 0.16123712062835693,
+      "learning_rate": 1.6156401932718258e-05,
+      "loss": 1.1593,
+      "step": 8468
+    },
+    {
+      "epoch": 3.1530523205081966,
+      "grad_norm": 0.17915739119052887,
+      "learning_rate": 1.6155444184729888e-05,
+      "loss": 1.1455,
+      "step": 8469
+    },
+    {
+      "epoch": 3.1534246256588054,
+      "grad_norm": 0.19462330639362335,
+      "learning_rate": 1.6154486345825996e-05,
+      "loss": 1.1657,
+      "step": 8470
+    },
+    {
+      "epoch": 3.1537969308094147,
+      "grad_norm": 0.1797885149717331,
+      "learning_rate": 1.6153528416020724e-05,
+      "loss": 1.1494,
+      "step": 8471
+    },
+    {
+      "epoch": 3.1541692359600235,
+      "grad_norm": 0.16456390917301178,
+      "learning_rate": 1.6152570395328227e-05,
+      "loss": 1.1628,
+      "step": 8472
+    },
+    {
+      "epoch": 3.154541541110633,
+      "grad_norm": 0.16568942368030548,
+      "learning_rate": 1.6151612283762653e-05,
+      "loss": 1.162,
+      "step": 8473
+    },
+    {
+      "epoch": 3.154913846261242,
+      "grad_norm": 0.16583888232707977,
+      "learning_rate": 1.6150654081338143e-05,
+      "loss": 1.1659,
+      "step": 8474
+    },
+    {
+      "epoch": 3.155286151411851,
+      "grad_norm": 0.15852290391921997,
+      "learning_rate": 1.6149695788068868e-05,
+      "loss": 1.1591,
+      "step": 8475
+    },
+    {
+      "epoch": 3.15565845656246,
+      "grad_norm": 0.16142380237579346,
+      "learning_rate": 1.6148737403968967e-05,
+      "loss": 1.1588,
+      "step": 8476
+    },
+    {
+      "epoch": 3.156030761713069,
+      "grad_norm": 0.17836064100265503,
+      "learning_rate": 1.6147778929052602e-05,
+      "loss": 1.1708,
+      "step": 8477
+    },
+    {
+      "epoch": 3.156403066863678,
+      "grad_norm": 0.1994626820087433,
+      "learning_rate": 1.614682036333393e-05,
+      "loss": 1.153,
+      "step": 8478
+    },
+    {
+      "epoch": 3.156775372014287,
+      "grad_norm": 0.21441636979579926,
+      "learning_rate": 1.6145861706827104e-05,
+      "loss": 1.1498,
+      "step": 8479
+    },
+    {
+      "epoch": 3.1571476771648963,
+      "grad_norm": 0.19782859086990356,
+      "learning_rate": 1.6144902959546286e-05,
+      "loss": 1.1527,
+      "step": 8480
+    },
+    {
+      "epoch": 3.157519982315505,
+      "grad_norm": 0.16160625219345093,
+      "learning_rate": 1.614394412150564e-05,
+      "loss": 1.1526,
+      "step": 8481
+    },
+    {
+      "epoch": 3.1578922874661144,
+      "grad_norm": 0.20167799293994904,
+      "learning_rate": 1.614298519271932e-05,
+      "loss": 1.155,
+      "step": 8482
+    },
+    {
+      "epoch": 3.1582645926167237,
+      "grad_norm": 0.21661166846752167,
+      "learning_rate": 1.61420261732015e-05,
+      "loss": 1.1462,
+      "step": 8483
+    },
+    {
+      "epoch": 3.1586368977673325,
+      "grad_norm": 0.16825565695762634,
+      "learning_rate": 1.6141067062966332e-05,
+      "loss": 1.1623,
+      "step": 8484
+    },
+    {
+      "epoch": 3.1590092029179417,
+      "grad_norm": 0.2019083946943283,
+      "learning_rate": 1.6140107862027993e-05,
+      "loss": 1.1671,
+      "step": 8485
+    },
+    {
+      "epoch": 3.1593815080685506,
+      "grad_norm": 0.17671798169612885,
+      "learning_rate": 1.6139148570400647e-05,
+      "loss": 1.1589,
+      "step": 8486
+    },
+    {
+      "epoch": 3.15975381321916,
+      "grad_norm": 0.17492499947547913,
+      "learning_rate": 1.6138189188098463e-05,
+      "loss": 1.1591,
+      "step": 8487
+    },
+    {
+      "epoch": 3.1601261183697686,
+      "grad_norm": 0.16377635300159454,
+      "learning_rate": 1.6137229715135604e-05,
+      "loss": 1.1596,
+      "step": 8488
+    },
+    {
+      "epoch": 3.160498423520378,
+      "grad_norm": 0.17293529212474823,
+      "learning_rate": 1.6136270151526254e-05,
+      "loss": 1.1644,
+      "step": 8489
+    },
+    {
+      "epoch": 3.1608707286709867,
+      "grad_norm": 0.15997327864170074,
+      "learning_rate": 1.6135310497284575e-05,
+      "loss": 1.141,
+      "step": 8490
+    },
+    {
+      "epoch": 3.161243033821596,
+      "grad_norm": 0.16692161560058594,
+      "learning_rate": 1.6134350752424746e-05,
+      "loss": 1.1528,
+      "step": 8491
+    },
+    {
+      "epoch": 3.1616153389722053,
+      "grad_norm": 0.17296627163887024,
+      "learning_rate": 1.613339091696094e-05,
+      "loss": 1.1472,
+      "step": 8492
+    },
+    {
+      "epoch": 3.161987644122814,
+      "grad_norm": 0.1687869429588318,
+      "learning_rate": 1.6132430990907338e-05,
+      "loss": 1.1503,
+      "step": 8493
+    },
+    {
+      "epoch": 3.1623599492734233,
+      "grad_norm": 0.18529605865478516,
+      "learning_rate": 1.6131470974278114e-05,
+      "loss": 1.1723,
+      "step": 8494
+    },
+    {
+      "epoch": 3.162732254424032,
+      "grad_norm": 0.1638706773519516,
+      "learning_rate": 1.6130510867087447e-05,
+      "loss": 1.1436,
+      "step": 8495
+    },
+    {
+      "epoch": 3.1631045595746414,
+      "grad_norm": 0.18358993530273438,
+      "learning_rate": 1.612955066934952e-05,
+      "loss": 1.1605,
+      "step": 8496
+    },
+    {
+      "epoch": 3.1634768647252502,
+      "grad_norm": 0.16975806653499603,
+      "learning_rate": 1.6128590381078516e-05,
+      "loss": 1.1495,
+      "step": 8497
+    },
+    {
+      "epoch": 3.1638491698758595,
+      "grad_norm": 0.1780412644147873,
+      "learning_rate": 1.6127630002288615e-05,
+      "loss": 1.1628,
+      "step": 8498
+    },
+    {
+      "epoch": 3.1642214750264688,
+      "grad_norm": 0.17131397128105164,
+      "learning_rate": 1.6126669532994003e-05,
+      "loss": 1.1578,
+      "step": 8499
+    },
+    {
+      "epoch": 3.1645937801770776,
+      "grad_norm": 0.17910616099834442,
+      "learning_rate": 1.6125708973208868e-05,
+      "loss": 1.1559,
+      "step": 8500
+    },
+    {
+      "epoch": 3.1645937801770776,
+      "eval_loss": 1.297155499458313,
+      "eval_runtime": 16.9178,
+      "eval_samples_per_second": 102.496,
+      "eval_steps_per_second": 5.143,
+      "step": 8500
+    },
+    {
+      "epoch": 3.1645937801770776,
+      "step": 8500,
+      "total_flos": 1.2178098370248075e+20,
+      "train_loss": 1.2383465781352099,
+      "train_runtime": 245543.4426,
+      "train_samples_per_second": 70.008,
+      "train_steps_per_second": 0.109
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 26850,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 10,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "EarlyStoppingCallback": {
+      "args": {
+        "early_stopping_patience": 3,
+        "early_stopping_threshold": 0.0
+      },
+      "attributes": {
+        "early_stopping_patience_counter": 0
+      }
+    },
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.2178098370248075e+20,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}