diff --git "a/checkpoint-3000/trainer_state.json" "b/checkpoint-3000/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-3000/trainer_state.json"
@@ -0,0 +1,21033 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.8152173913043478,
+  "eval_steps": 500,
+  "global_step": 3000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0002717391304347826,
+      "grad_norm": Infinity,
+      "learning_rate": 0.0,
+      "loss": 3.8119,
+      "step": 1
+    },
+    {
+      "epoch": 0.0005434782608695652,
+      "grad_norm": Infinity,
+      "learning_rate": 0.0,
+      "loss": 5.6312,
+      "step": 2
+    },
+    {
+      "epoch": 0.0008152173913043478,
+      "grad_norm": 3.606865644454956,
+      "learning_rate": 1.0000000000000001e-07,
+      "loss": 3.6085,
+      "step": 3
+    },
+    {
+      "epoch": 0.0010869565217391304,
+      "grad_norm": 5.723954677581787,
+      "learning_rate": 2.0000000000000002e-07,
+      "loss": 4.5497,
+      "step": 4
+    },
+    {
+      "epoch": 0.001358695652173913,
+      "grad_norm": 8.396666526794434,
+      "learning_rate": 3.0000000000000004e-07,
+      "loss": 5.6655,
+      "step": 5
+    },
+    {
+      "epoch": 0.0016304347826086956,
+      "grad_norm": 6.602017402648926,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 4.9113,
+      "step": 6
+    },
+    {
+      "epoch": 0.0019021739130434783,
+      "grad_norm": 2.8019049167633057,
+      "learning_rate": 5e-07,
+      "loss": 3.0843,
+      "step": 7
+    },
+    {
+      "epoch": 0.002173913043478261,
+      "grad_norm": 5.959436893463135,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 4.4235,
+      "step": 8
+    },
+    {
+      "epoch": 0.0024456521739130437,
+      "grad_norm": 5.623807430267334,
+      "learning_rate": 7.000000000000001e-07,
+      "loss": 4.5667,
+      "step": 9
+    },
+    {
+      "epoch": 0.002717391304347826,
+      "grad_norm": 5.9944844245910645,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 4.6548,
+      "step": 10
+    },
+    {
+      "epoch": 0.002989130434782609,
+      "grad_norm": 7.308701038360596,
+      "learning_rate": 9e-07,
+      "loss": 5.1422,
+      "step": 11
+    },
+    {
+      "epoch": 0.003260869565217391,
+      "grad_norm": 5.805826187133789,
+      "learning_rate": 1e-06,
+      "loss": 4.3458,
+      "step": 12
+    },
+    {
+      "epoch": 0.003532608695652174,
+      "grad_norm": 5.096138000488281,
+      "learning_rate": 1.1e-06,
+      "loss": 4.0761,
+      "step": 13
+    },
+    {
+      "epoch": 0.0038043478260869567,
+      "grad_norm": 4.222947120666504,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 3.7963,
+      "step": 14
+    },
+    {
+      "epoch": 0.004076086956521739,
+      "grad_norm": 8.304177284240723,
+      "learning_rate": 1.3e-06,
+      "loss": 5.4697,
+      "step": 15
+    },
+    {
+      "epoch": 0.004347826086956522,
+      "grad_norm": 6.131305694580078,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 4.5078,
+      "step": 16
+    },
+    {
+      "epoch": 0.004619565217391305,
+      "grad_norm": 5.1205315589904785,
+      "learning_rate": 1.5e-06,
+      "loss": 4.1935,
+      "step": 17
+    },
+    {
+      "epoch": 0.004891304347826087,
+      "grad_norm": 3.348874568939209,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 3.165,
+      "step": 18
+    },
+    {
+      "epoch": 0.005163043478260869,
+      "grad_norm": 11.620071411132812,
+      "learning_rate": 1.7e-06,
+      "loss": 6.6692,
+      "step": 19
+    },
+    {
+      "epoch": 0.005434782608695652,
+      "grad_norm": 5.148247241973877,
+      "learning_rate": 1.8e-06,
+      "loss": 3.7967,
+      "step": 20
+    },
+    {
+      "epoch": 0.005706521739130435,
+      "grad_norm": 6.738089561462402,
+      "learning_rate": 1.9000000000000002e-06,
+      "loss": 4.6053,
+      "step": 21
+    },
+    {
+      "epoch": 0.005978260869565218,
+      "grad_norm": 7.383264541625977,
+      "learning_rate": 2e-06,
+      "loss": 4.7862,
+      "step": 22
+    },
+    {
+      "epoch": 0.00625,
+      "grad_norm": 9.380542755126953,
+      "learning_rate": 2.1000000000000002e-06,
+      "loss": 5.5961,
+      "step": 23
+    },
+    {
+      "epoch": 0.006521739130434782,
+      "grad_norm": 5.481351852416992,
+      "learning_rate": 2.2e-06,
+      "loss": 3.9489,
+      "step": 24
+    },
+    {
+      "epoch": 0.006793478260869565,
+      "grad_norm": 3.769026756286621,
+      "learning_rate": 2.3e-06,
+      "loss": 3.3576,
+      "step": 25
+    },
+    {
+      "epoch": 0.007065217391304348,
+      "grad_norm": 6.088125705718994,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 4.3102,
+      "step": 26
+    },
+    {
+      "epoch": 0.007336956521739131,
+      "grad_norm": 3.7228612899780273,
+      "learning_rate": 2.4999999999999998e-06,
+      "loss": 3.2898,
+      "step": 27
+    },
+    {
+      "epoch": 0.007608695652173913,
+      "grad_norm": 5.163967609405518,
+      "learning_rate": 2.6e-06,
+      "loss": 3.9394,
+      "step": 28
+    },
+    {
+      "epoch": 0.007880434782608695,
+      "grad_norm": 3.577744960784912,
+      "learning_rate": 2.7e-06,
+      "loss": 3.4044,
+      "step": 29
+    },
+    {
+      "epoch": 0.008152173913043478,
+      "grad_norm": 5.421440124511719,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 4.2156,
+      "step": 30
+    },
+    {
+      "epoch": 0.00842391304347826,
+      "grad_norm": 4.284877777099609,
+      "learning_rate": 2.9e-06,
+      "loss": 3.519,
+      "step": 31
+    },
+    {
+      "epoch": 0.008695652173913044,
+      "grad_norm": 6.597541809082031,
+      "learning_rate": 3e-06,
+      "loss": 4.5028,
+      "step": 32
+    },
+    {
+      "epoch": 0.008967391304347826,
+      "grad_norm": 7.851593494415283,
+      "learning_rate": 3.1e-06,
+      "loss": 5.1247,
+      "step": 33
+    },
+    {
+      "epoch": 0.00923913043478261,
+      "grad_norm": 2.644531726837158,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 2.9878,
+      "step": 34
+    },
+    {
+      "epoch": 0.009510869565217392,
+      "grad_norm": 3.316633462905884,
+      "learning_rate": 3.3e-06,
+      "loss": 3.4112,
+      "step": 35
+    },
+    {
+      "epoch": 0.009782608695652175,
+      "grad_norm": 3.0373375415802,
+      "learning_rate": 3.4e-06,
+      "loss": 3.301,
+      "step": 36
+    },
+    {
+      "epoch": 0.010054347826086956,
+      "grad_norm": 5.997847080230713,
+      "learning_rate": 3.5e-06,
+      "loss": 4.9763,
+      "step": 37
+    },
+    {
+      "epoch": 0.010326086956521738,
+      "grad_norm": 5.2729268074035645,
+      "learning_rate": 3.6e-06,
+      "loss": 4.7279,
+      "step": 38
+    },
+    {
+      "epoch": 0.010597826086956521,
+      "grad_norm": 2.7055575847625732,
+      "learning_rate": 3.7e-06,
+      "loss": 3.4656,
+      "step": 39
+    },
+    {
+      "epoch": 0.010869565217391304,
+      "grad_norm": 3.5352964401245117,
+      "learning_rate": 3.8000000000000005e-06,
+      "loss": 3.9642,
+      "step": 40
+    },
+    {
+      "epoch": 0.011141304347826087,
+      "grad_norm": 2.8491668701171875,
+      "learning_rate": 3.9e-06,
+      "loss": 3.8092,
+      "step": 41
+    },
+    {
+      "epoch": 0.01141304347826087,
+      "grad_norm": 1.8910856246948242,
+      "learning_rate": 4e-06,
+      "loss": 3.0861,
+      "step": 42
+    },
+    {
+      "epoch": 0.011684782608695652,
+      "grad_norm": 2.873784303665161,
+      "learning_rate": 4.1e-06,
+      "loss": 3.788,
+      "step": 43
+    },
+    {
+      "epoch": 0.011956521739130435,
+      "grad_norm": 3.121163845062256,
+      "learning_rate": 4.2000000000000004e-06,
+      "loss": 4.0837,
+      "step": 44
+    },
+    {
+      "epoch": 0.012228260869565218,
+      "grad_norm": 2.977182388305664,
+      "learning_rate": 4.3e-06,
+      "loss": 4.0382,
+      "step": 45
+    },
+    {
+      "epoch": 0.0125,
+      "grad_norm": 3.935685396194458,
+      "learning_rate": 4.4e-06,
+      "loss": 5.0603,
+      "step": 46
+    },
+    {
+      "epoch": 0.012771739130434783,
+      "grad_norm": 1.5644341707229614,
+      "learning_rate": 4.5e-06,
+      "loss": 3.0085,
+      "step": 47
+    },
+    {
+      "epoch": 0.013043478260869565,
+      "grad_norm": 2.2639412879943848,
+      "learning_rate": 4.6e-06,
+      "loss": 3.5606,
+      "step": 48
+    },
+    {
+      "epoch": 0.013315217391304347,
+      "grad_norm": 3.8248438835144043,
+      "learning_rate": 4.700000000000001e-06,
+      "loss": 5.1055,
+      "step": 49
+    },
+    {
+      "epoch": 0.01358695652173913,
+      "grad_norm": 2.0189621448516846,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 3.6357,
+      "step": 50
+    },
+    {
+      "epoch": 0.013858695652173913,
+      "grad_norm": 1.7053027153015137,
+      "learning_rate": 4.9e-06,
+      "loss": 3.3706,
+      "step": 51
+    },
+    {
+      "epoch": 0.014130434782608696,
+      "grad_norm": 2.996800184249878,
+      "learning_rate": 4.9999999999999996e-06,
+      "loss": 4.2643,
+      "step": 52
+    },
+    {
+      "epoch": 0.014402173913043478,
+      "grad_norm": 2.2045717239379883,
+      "learning_rate": 5.1e-06,
+      "loss": 3.4676,
+      "step": 53
+    },
+    {
+      "epoch": 0.014673913043478261,
+      "grad_norm": 1.9880719184875488,
+      "learning_rate": 5.2e-06,
+      "loss": 3.7901,
+      "step": 54
+    },
+    {
+      "epoch": 0.014945652173913044,
+      "grad_norm": 1.4993855953216553,
+      "learning_rate": 5.3e-06,
+      "loss": 3.3175,
+      "step": 55
+    },
+    {
+      "epoch": 0.015217391304347827,
+      "grad_norm": 1.3958892822265625,
+      "learning_rate": 5.4e-06,
+      "loss": 3.0166,
+      "step": 56
+    },
+    {
+      "epoch": 0.01548913043478261,
+      "grad_norm": 1.5511572360992432,
+      "learning_rate": 5.5e-06,
+      "loss": 3.4503,
+      "step": 57
+    },
+    {
+      "epoch": 0.01576086956521739,
+      "grad_norm": 2.274249792098999,
+      "learning_rate": 5.600000000000001e-06,
+      "loss": 4.1636,
+      "step": 58
+    },
+    {
+      "epoch": 0.016032608695652175,
+      "grad_norm": 1.7447724342346191,
+      "learning_rate": 5.7000000000000005e-06,
+      "loss": 3.7056,
+      "step": 59
+    },
+    {
+      "epoch": 0.016304347826086956,
+      "grad_norm": 2.4315755367279053,
+      "learning_rate": 5.8e-06,
+      "loss": 3.9477,
+      "step": 60
+    },
+    {
+      "epoch": 0.01657608695652174,
+      "grad_norm": 1.5694116353988647,
+      "learning_rate": 5.899999999999999e-06,
+      "loss": 3.321,
+      "step": 61
+    },
+    {
+      "epoch": 0.01684782608695652,
+      "grad_norm": 1.856467366218567,
+      "learning_rate": 6e-06,
+      "loss": 3.8384,
+      "step": 62
+    },
+    {
+      "epoch": 0.017119565217391303,
+      "grad_norm": 1.3913215398788452,
+      "learning_rate": 6.1e-06,
+      "loss": 3.1717,
+      "step": 63
+    },
+    {
+      "epoch": 0.017391304347826087,
+      "grad_norm": 1.1280888319015503,
+      "learning_rate": 6.2e-06,
+      "loss": 3.0487,
+      "step": 64
+    },
+    {
+      "epoch": 0.017663043478260868,
+      "grad_norm": 1.3553597927093506,
+      "learning_rate": 6.3e-06,
+      "loss": 3.4274,
+      "step": 65
+    },
+    {
+      "epoch": 0.017934782608695653,
+      "grad_norm": 1.5680698156356812,
+      "learning_rate": 6.4000000000000006e-06,
+      "loss": 3.2402,
+      "step": 66
+    },
+    {
+      "epoch": 0.018206521739130434,
+      "grad_norm": 1.2775722742080688,
+      "learning_rate": 6.5000000000000004e-06,
+      "loss": 3.1252,
+      "step": 67
+    },
+    {
+      "epoch": 0.01847826086956522,
+      "grad_norm": 1.547057867050171,
+      "learning_rate": 6.6e-06,
+      "loss": 3.6654,
+      "step": 68
+    },
+    {
+      "epoch": 0.01875,
+      "grad_norm": 1.6398184299468994,
+      "learning_rate": 6.7e-06,
+      "loss": 3.4637,
+      "step": 69
+    },
+    {
+      "epoch": 0.019021739130434784,
+      "grad_norm": 0.8923302292823792,
+      "learning_rate": 6.8e-06,
+      "loss": 2.6619,
+      "step": 70
+    },
+    {
+      "epoch": 0.019293478260869565,
+      "grad_norm": 1.1862719058990479,
+      "learning_rate": 6.900000000000001e-06,
+      "loss": 3.1319,
+      "step": 71
+    },
+    {
+      "epoch": 0.01956521739130435,
+      "grad_norm": 1.0350522994995117,
+      "learning_rate": 7e-06,
+      "loss": 2.8874,
+      "step": 72
+    },
+    {
+      "epoch": 0.01983695652173913,
+      "grad_norm": 1.7756403684616089,
+      "learning_rate": 7.1e-06,
+      "loss": 3.7356,
+      "step": 73
+    },
+    {
+      "epoch": 0.02010869565217391,
+      "grad_norm": 1.2788705825805664,
+      "learning_rate": 7.2e-06,
+      "loss": 3.2409,
+      "step": 74
+    },
+    {
+      "epoch": 0.020380434782608696,
+      "grad_norm": 1.4206314086914062,
+      "learning_rate": 7.3e-06,
+      "loss": 3.5884,
+      "step": 75
+    },
+    {
+      "epoch": 0.020652173913043477,
+      "grad_norm": 2.0677270889282227,
+      "learning_rate": 7.4e-06,
+      "loss": 4.3218,
+      "step": 76
+    },
+    {
+      "epoch": 0.02092391304347826,
+      "grad_norm": 0.700586199760437,
+      "learning_rate": 7.5e-06,
+      "loss": 2.4596,
+      "step": 77
+    },
+    {
+      "epoch": 0.021195652173913043,
+      "grad_norm": 2.1819796562194824,
+      "learning_rate": 7.600000000000001e-06,
+      "loss": 3.9885,
+      "step": 78
+    },
+    {
+      "epoch": 0.021467391304347827,
+      "grad_norm": 1.0155264139175415,
+      "learning_rate": 7.699999999999999e-06,
+      "loss": 2.7548,
+      "step": 79
+    },
+    {
+      "epoch": 0.021739130434782608,
+      "grad_norm": 1.2022877931594849,
+      "learning_rate": 7.8e-06,
+      "loss": 3.1796,
+      "step": 80
+    },
+    {
+      "epoch": 0.022010869565217393,
+      "grad_norm": 1.1962783336639404,
+      "learning_rate": 7.899999999999999e-06,
+      "loss": 2.907,
+      "step": 81
+    },
+    {
+      "epoch": 0.022282608695652174,
+      "grad_norm": 1.0273329019546509,
+      "learning_rate": 8e-06,
+      "loss": 2.7207,
+      "step": 82
+    },
+    {
+      "epoch": 0.022554347826086958,
+      "grad_norm": 2.3930468559265137,
+      "learning_rate": 8.1e-06,
+      "loss": 3.9644,
+      "step": 83
+    },
+    {
+      "epoch": 0.02282608695652174,
+      "grad_norm": 2.244040012359619,
+      "learning_rate": 8.2e-06,
+      "loss": 3.5003,
+      "step": 84
+    },
+    {
+      "epoch": 0.02309782608695652,
+      "grad_norm": 1.685024380683899,
+      "learning_rate": 8.3e-06,
+      "loss": 3.0757,
+      "step": 85
+    },
+    {
+      "epoch": 0.023369565217391305,
+      "grad_norm": 3.1130995750427246,
+      "learning_rate": 8.400000000000001e-06,
+      "loss": 3.9588,
+      "step": 86
+    },
+    {
+      "epoch": 0.023641304347826086,
+      "grad_norm": 2.6688392162323,
+      "learning_rate": 8.5e-06,
+      "loss": 3.6427,
+      "step": 87
+    },
+    {
+      "epoch": 0.02391304347826087,
+      "grad_norm": 1.3578976392745972,
+      "learning_rate": 8.6e-06,
+      "loss": 2.9824,
+      "step": 88
+    },
+    {
+      "epoch": 0.02418478260869565,
+      "grad_norm": 1.2437134981155396,
+      "learning_rate": 8.7e-06,
+      "loss": 3.1183,
+      "step": 89
+    },
+    {
+      "epoch": 0.024456521739130436,
+      "grad_norm": 2.143594741821289,
+      "learning_rate": 8.8e-06,
+      "loss": 3.6833,
+      "step": 90
+    },
+    {
+      "epoch": 0.024728260869565217,
+      "grad_norm": 1.2158911228179932,
+      "learning_rate": 8.900000000000001e-06,
+      "loss": 2.7814,
+      "step": 91
+    },
+    {
+      "epoch": 0.025,
+      "grad_norm": 2.4209673404693604,
+      "learning_rate": 9e-06,
+      "loss": 4.3185,
+      "step": 92
+    },
+    {
+      "epoch": 0.025271739130434782,
+      "grad_norm": 3.090715169906616,
+      "learning_rate": 9.100000000000001e-06,
+      "loss": 4.0217,
+      "step": 93
+    },
+    {
+      "epoch": 0.025543478260869567,
+      "grad_norm": 3.1287500858306885,
+      "learning_rate": 9.2e-06,
+      "loss": 3.8665,
+      "step": 94
+    },
+    {
+      "epoch": 0.025815217391304348,
+      "grad_norm": 2.6370797157287598,
+      "learning_rate": 9.3e-06,
+      "loss": 3.7443,
+      "step": 95
+    },
+    {
+      "epoch": 0.02608695652173913,
+      "grad_norm": 4.120901107788086,
+      "learning_rate": 9.400000000000001e-06,
+      "loss": 4.3482,
+      "step": 96
+    },
+    {
+      "epoch": 0.026358695652173914,
+      "grad_norm": NaN,
+      "learning_rate": 9.400000000000001e-06,
+      "loss": 3.7633,
+      "step": 97
+    },
+    {
+      "epoch": 0.026630434782608695,
+      "grad_norm": 2.1682493686676025,
+      "learning_rate": 9.5e-06,
+      "loss": 3.135,
+      "step": 98
+    },
+    {
+      "epoch": 0.02690217391304348,
+      "grad_norm": 9.943031311035156,
+      "learning_rate": 9.600000000000001e-06,
+      "loss": 4.4992,
+      "step": 99
+    },
+    {
+      "epoch": 0.02717391304347826,
+      "grad_norm": 9.851699829101562,
+      "learning_rate": 9.699999999999999e-06,
+      "loss": 3.3346,
+      "step": 100
+    },
+    {
+      "epoch": 0.027445652173913045,
+      "grad_norm": 9.068572044372559,
+      "learning_rate": 9.8e-06,
+      "loss": 3.4352,
+      "step": 101
+    },
+    {
+      "epoch": 0.027717391304347826,
+      "grad_norm": 16.46036148071289,
+      "learning_rate": 9.9e-06,
+      "loss": 4.2803,
+      "step": 102
+    },
+    {
+      "epoch": 0.02798913043478261,
+      "grad_norm": 4.916048049926758,
+      "learning_rate": 9.999999999999999e-06,
+      "loss": 5.2283,
+      "step": 103
+    },
+    {
+      "epoch": 0.02826086956521739,
+      "grad_norm": 3.9112675189971924,
+      "learning_rate": 1.01e-05,
+      "loss": 3.9085,
+      "step": 104
+    },
+    {
+      "epoch": 0.028532608695652172,
+      "grad_norm": 2.9353702068328857,
+      "learning_rate": 1.02e-05,
+      "loss": 3.436,
+      "step": 105
+    },
+    {
+      "epoch": 0.028804347826086957,
+      "grad_norm": 2.1756107807159424,
+      "learning_rate": 1.03e-05,
+      "loss": 4.2118,
+      "step": 106
+    },
+    {
+      "epoch": 0.029076086956521738,
+      "grad_norm": 2.4335989952087402,
+      "learning_rate": 1.04e-05,
+      "loss": 4.3144,
+      "step": 107
+    },
+    {
+      "epoch": 0.029347826086956522,
+      "grad_norm": 2.6821694374084473,
+      "learning_rate": 1.05e-05,
+      "loss": 4.461,
+      "step": 108
+    },
+    {
+      "epoch": 0.029619565217391303,
+      "grad_norm": 1.9325740337371826,
+      "learning_rate": 1.06e-05,
+      "loss": 4.1079,
+      "step": 109
+    },
+    {
+      "epoch": 0.029891304347826088,
+      "grad_norm": 3.598806381225586,
+      "learning_rate": 1.0700000000000001e-05,
+      "loss": 3.6908,
+      "step": 110
+    },
+    {
+      "epoch": 0.03016304347826087,
+      "grad_norm": 1.7592302560806274,
+      "learning_rate": 1.08e-05,
+      "loss": 3.2482,
+      "step": 111
+    },
+    {
+      "epoch": 0.030434782608695653,
+      "grad_norm": 2.4772744178771973,
+      "learning_rate": 1.09e-05,
+      "loss": 4.3941,
+      "step": 112
+    },
+    {
+      "epoch": 0.030706521739130434,
+      "grad_norm": 1.8405948877334595,
+      "learning_rate": 1.1e-05,
+      "loss": 3.6201,
+      "step": 113
+    },
+    {
+      "epoch": 0.03097826086956522,
+      "grad_norm": 4.303748607635498,
+      "learning_rate": 1.11e-05,
+      "loss": 4.6995,
+      "step": 114
+    },
+    {
+      "epoch": 0.03125,
+      "grad_norm": 2.199484348297119,
+      "learning_rate": 1.1200000000000001e-05,
+      "loss": 3.7969,
+      "step": 115
+    },
+    {
+      "epoch": 0.03152173913043478,
+      "grad_norm": 1.0349135398864746,
+      "learning_rate": 1.13e-05,
+      "loss": 2.8461,
+      "step": 116
+    },
+    {
+      "epoch": 0.03179347826086956,
+      "grad_norm": 2.374260425567627,
+      "learning_rate": 1.1400000000000001e-05,
+      "loss": 4.2862,
+      "step": 117
+    },
+    {
+      "epoch": 0.03206521739130435,
+      "grad_norm": 1.283584713935852,
+      "learning_rate": 1.1500000000000002e-05,
+      "loss": 3.1288,
+      "step": 118
+    },
+    {
+      "epoch": 0.03233695652173913,
+      "grad_norm": 3.549565076828003,
+      "learning_rate": 1.16e-05,
+      "loss": 4.1137,
+      "step": 119
+    },
+    {
+      "epoch": 0.03260869565217391,
+      "grad_norm": 1.3312267065048218,
+      "learning_rate": 1.1700000000000001e-05,
+      "loss": 3.6003,
+      "step": 120
+    },
+    {
+      "epoch": 0.03288043478260869,
+      "grad_norm": 1.655480146408081,
+      "learning_rate": 1.1799999999999999e-05,
+      "loss": 3.7798,
+      "step": 121
+    },
+    {
+      "epoch": 0.03315217391304348,
+      "grad_norm": 1.6580840349197388,
+      "learning_rate": 1.19e-05,
+      "loss": 3.5654,
+      "step": 122
+    },
+    {
+      "epoch": 0.03342391304347826,
+      "grad_norm": 2.8110294342041016,
+      "learning_rate": 1.2e-05,
+      "loss": 3.9673,
+      "step": 123
+    },
+    {
+      "epoch": 0.03369565217391304,
+      "grad_norm": 1.2965394258499146,
+      "learning_rate": 1.21e-05,
+      "loss": 3.26,
+      "step": 124
+    },
+    {
+      "epoch": 0.033967391304347824,
+      "grad_norm": 1.914302945137024,
+      "learning_rate": 1.22e-05,
+      "loss": 3.95,
+      "step": 125
+    },
+    {
+      "epoch": 0.034239130434782605,
+      "grad_norm": 1.3706483840942383,
+      "learning_rate": 1.2299999999999999e-05,
+      "loss": 3.5317,
+      "step": 126
+    },
+    {
+      "epoch": 0.03451086956521739,
+      "grad_norm": 2.016113042831421,
+      "learning_rate": 1.24e-05,
+      "loss": 3.2582,
+      "step": 127
+    },
+    {
+      "epoch": 0.034782608695652174,
+      "grad_norm": 2.5129711627960205,
+      "learning_rate": 1.25e-05,
+      "loss": 3.3608,
+      "step": 128
+    },
+    {
+      "epoch": 0.035054347826086955,
+      "grad_norm": Infinity,
+      "learning_rate": 1.25e-05,
+      "loss": 4.3213,
+      "step": 129
+    },
+    {
+      "epoch": 0.035326086956521736,
+      "grad_norm": 3.073050022125244,
+      "learning_rate": 1.26e-05,
+      "loss": 4.4668,
+      "step": 130
+    },
+    {
+      "epoch": 0.035597826086956524,
+      "grad_norm": 1.8033742904663086,
+      "learning_rate": 1.27e-05,
+      "loss": 3.5871,
+      "step": 131
+    },
+    {
+      "epoch": 0.035869565217391305,
+      "grad_norm": 3.148634195327759,
+      "learning_rate": 1.2800000000000001e-05,
+      "loss": 3.7611,
+      "step": 132
+    },
+    {
+      "epoch": 0.036141304347826086,
+      "grad_norm": 2.898167610168457,
+      "learning_rate": 1.29e-05,
+      "loss": 3.6928,
+      "step": 133
+    },
+    {
+      "epoch": 0.03641304347826087,
+      "grad_norm": 1.9879661798477173,
+      "learning_rate": 1.3000000000000001e-05,
+      "loss": 3.607,
+      "step": 134
+    },
+    {
+      "epoch": 0.036684782608695655,
+      "grad_norm": 2.0479657649993896,
+      "learning_rate": 1.31e-05,
+      "loss": 3.2456,
+      "step": 135
+    },
+    {
+      "epoch": 0.03695652173913044,
+      "grad_norm": 1.5060527324676514,
+      "learning_rate": 1.32e-05,
+      "loss": 3.4124,
+      "step": 136
+    },
+    {
+      "epoch": 0.03722826086956522,
+      "grad_norm": 2.2012810707092285,
+      "learning_rate": 1.3300000000000001e-05,
+      "loss": 4.1008,
+      "step": 137
+    },
+    {
+      "epoch": 0.0375,
+      "grad_norm": 2.2929439544677734,
+      "learning_rate": 1.34e-05,
+      "loss": 4.5483,
+      "step": 138
+    },
+    {
+      "epoch": 0.03777173913043478,
+      "grad_norm": 1.777136206626892,
+      "learning_rate": 1.3500000000000001e-05,
+      "loss": 3.9796,
+      "step": 139
+    },
+    {
+      "epoch": 0.03804347826086957,
+      "grad_norm": 3.1867823600769043,
+      "learning_rate": 1.36e-05,
+      "loss": 4.13,
+      "step": 140
+    },
+    {
+      "epoch": 0.03831521739130435,
+      "grad_norm": 2.070611000061035,
+      "learning_rate": 1.3700000000000001e-05,
+      "loss": 3.3474,
+      "step": 141
+    },
+    {
+      "epoch": 0.03858695652173913,
+      "grad_norm": 1.9961167573928833,
+      "learning_rate": 1.3800000000000002e-05,
+      "loss": 4.0889,
+      "step": 142
+    },
+    {
+      "epoch": 0.03885869565217391,
+      "grad_norm": 1.709995985031128,
+      "learning_rate": 1.39e-05,
+      "loss": 3.9576,
+      "step": 143
+    },
+    {
+      "epoch": 0.0391304347826087,
+      "grad_norm": 1.793494462966919,
+      "learning_rate": 1.4e-05,
+      "loss": 4.419,
+      "step": 144
+    },
+    {
+      "epoch": 0.03940217391304348,
+      "grad_norm": 2.1228878498077393,
+      "learning_rate": 1.4099999999999999e-05,
+      "loss": 3.7677,
+      "step": 145
+    },
+    {
+      "epoch": 0.03967391304347826,
+      "grad_norm": 1.5662310123443604,
+      "learning_rate": 1.42e-05,
+      "loss": 3.5038,
+      "step": 146
+    },
+    {
+      "epoch": 0.03994565217391304,
+      "grad_norm": 1.1859204769134521,
+      "learning_rate": 1.43e-05,
+      "loss": 3.0831,
+      "step": 147
+    },
+    {
+      "epoch": 0.04021739130434782,
+      "grad_norm": 1.998694896697998,
+      "learning_rate": 1.44e-05,
+      "loss": 4.06,
+      "step": 148
+    },
+    {
+      "epoch": 0.04048913043478261,
+      "grad_norm": 2.2709462642669678,
+      "learning_rate": 1.45e-05,
+      "loss": 3.4234,
+      "step": 149
+    },
+    {
+      "epoch": 0.04076086956521739,
+      "grad_norm": 3.572169065475464,
+      "learning_rate": 1.46e-05,
+      "loss": 4.1884,
+      "step": 150
+    },
+    {
+      "epoch": 0.04103260869565217,
+      "grad_norm": 2.4503207206726074,
+      "learning_rate": 1.47e-05,
+      "loss": 4.0506,
+      "step": 151
+    },
+    {
+      "epoch": 0.041304347826086954,
+      "grad_norm": 2.66367244720459,
+      "learning_rate": 1.48e-05,
+      "loss": 4.6858,
+      "step": 152
+    },
+    {
+      "epoch": 0.04157608695652174,
+      "grad_norm": 2.1236956119537354,
+      "learning_rate": 1.49e-05,
+      "loss": 3.2571,
+      "step": 153
+    },
+    {
+      "epoch": 0.04184782608695652,
+      "grad_norm": 4.9067206382751465,
+      "learning_rate": 1.5e-05,
+      "loss": 5.32,
+      "step": 154
+    },
+    {
+      "epoch": 0.042119565217391304,
+      "grad_norm": 2.0733673572540283,
+      "learning_rate": 1.51e-05,
+      "loss": 3.7548,
+      "step": 155
+    },
+    {
+      "epoch": 0.042391304347826085,
+      "grad_norm": 1.186569333076477,
+      "learning_rate": 1.5200000000000002e-05,
+      "loss": 3.3953,
+      "step": 156
+    },
+    {
+      "epoch": 0.042663043478260866,
+      "grad_norm": 2.5303616523742676,
+      "learning_rate": 1.53e-05,
+      "loss": 4.0742,
+      "step": 157
+    },
+    {
+      "epoch": 0.042934782608695654,
+      "grad_norm": 2.3837692737579346,
+      "learning_rate": 1.5399999999999998e-05,
+      "loss": 4.3684,
+      "step": 158
+    },
+    {
+      "epoch": 0.043206521739130435,
+      "grad_norm": 2.3704850673675537,
+      "learning_rate": 1.55e-05,
+      "loss": 4.1658,
+      "step": 159
+    },
+    {
+      "epoch": 0.043478260869565216,
+      "grad_norm": 3.8034274578094482,
+      "learning_rate": 1.56e-05,
+      "loss": 4.0919,
+      "step": 160
+    },
+    {
+      "epoch": 0.04375,
+      "grad_norm": 1.2763932943344116,
+      "learning_rate": 1.57e-05,
+      "loss": 3.8752,
+      "step": 161
+    },
+    {
+      "epoch": 0.044021739130434785,
+      "grad_norm": 1.7155462503433228,
+      "learning_rate": 1.5799999999999998e-05,
+      "loss": 4.317,
+      "step": 162
+    },
+    {
+      "epoch": 0.044293478260869566,
+      "grad_norm": 2.481580972671509,
+      "learning_rate": 1.59e-05,
+      "loss": 4.1314,
+      "step": 163
+    },
+    {
+      "epoch": 0.04456521739130435,
+      "grad_norm": 2.3853437900543213,
+      "learning_rate": 1.6e-05,
+      "loss": 3.6408,
+      "step": 164
+    },
+    {
+      "epoch": 0.04483695652173913,
+      "grad_norm": 1.7092500925064087,
+      "learning_rate": 1.61e-05,
+      "loss": 3.833,
+      "step": 165
+    },
+    {
+      "epoch": 0.045108695652173916,
+      "grad_norm": 2.1334292888641357,
+      "learning_rate": 1.62e-05,
+      "loss": 4.5951,
+      "step": 166
+    },
+    {
+      "epoch": 0.0453804347826087,
+      "grad_norm": 1.352221965789795,
+      "learning_rate": 1.63e-05,
+      "loss": 3.804,
+      "step": 167
+    },
+    {
+      "epoch": 0.04565217391304348,
+      "grad_norm": 2.178356409072876,
+      "learning_rate": 1.64e-05,
+      "loss": 3.7594,
+      "step": 168
+    },
+    {
+      "epoch": 0.04592391304347826,
+      "grad_norm": 2.9925286769866943,
+      "learning_rate": 1.65e-05,
+      "loss": 3.8819,
+      "step": 169
+    },
+    {
+      "epoch": 0.04619565217391304,
+      "grad_norm": 3.0277819633483887,
+      "learning_rate": 1.66e-05,
+      "loss": 4.4896,
+      "step": 170
+    },
+    {
+      "epoch": 0.04646739130434783,
+      "grad_norm": 1.459825873374939,
+      "learning_rate": 1.67e-05,
+      "loss": 4.1701,
+      "step": 171
+    },
+    {
+      "epoch": 0.04673913043478261,
+      "grad_norm": 1.6663670539855957,
+      "learning_rate": 1.6800000000000002e-05,
+      "loss": 3.6651,
+      "step": 172
+    },
+    {
+      "epoch": 0.04701086956521739,
+      "grad_norm": 1.7755922079086304,
+      "learning_rate": 1.69e-05,
+      "loss": 3.3178,
+      "step": 173
+    },
+    {
+      "epoch": 0.04728260869565217,
+      "grad_norm": 1.4884556531906128,
+      "learning_rate": 1.7e-05,
+      "loss": 3.6467,
+      "step": 174
+    },
+    {
+      "epoch": 0.04755434782608696,
+      "grad_norm": 2.701298713684082,
+      "learning_rate": 1.71e-05,
+      "loss": 4.773,
+      "step": 175
+    },
+    {
+      "epoch": 0.04782608695652174,
+      "grad_norm": 3.169405937194824,
+      "learning_rate": 1.72e-05,
+      "loss": 4.1844,
+      "step": 176
+    },
+    {
+      "epoch": 0.04809782608695652,
+      "grad_norm": 2.4385275840759277,
+      "learning_rate": 1.73e-05,
+      "loss": 3.8691,
+      "step": 177
+    },
+    {
+      "epoch": 0.0483695652173913,
+      "grad_norm": 3.446741819381714,
+      "learning_rate": 1.74e-05,
+      "loss": 3.6733,
+      "step": 178
+    },
+    {
+      "epoch": 0.048641304347826084,
+      "grad_norm": 4.095767974853516,
+      "learning_rate": 1.7500000000000002e-05,
+      "loss": 3.927,
+      "step": 179
+    },
+    {
+      "epoch": 0.04891304347826087,
+      "grad_norm": 3.42309308052063,
+      "learning_rate": 1.76e-05,
+      "loss": 4.152,
+      "step": 180
+    },
+    {
+      "epoch": 0.04918478260869565,
+      "grad_norm": 1.8505162000656128,
+      "learning_rate": 1.77e-05,
+      "loss": 3.5489,
+      "step": 181
+    },
+    {
+      "epoch": 0.049456521739130434,
+      "grad_norm": 3.3934953212738037,
+      "learning_rate": 1.7800000000000002e-05,
+      "loss": 3.9161,
+      "step": 182
+    },
+    {
+      "epoch": 0.049728260869565215,
+      "grad_norm": 3.2479214668273926,
+      "learning_rate": 1.79e-05,
+      "loss": 3.3177,
+      "step": 183
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.14975643157959,
+      "learning_rate": 1.8e-05,
+      "loss": 4.0072,
+      "step": 184
+    },
+    {
+      "epoch": 0.050271739130434784,
+      "grad_norm": 4.540154457092285,
+      "learning_rate": 1.8100000000000003e-05,
+      "loss": 3.878,
+      "step": 185
+    },
+    {
+      "epoch": 0.050543478260869565,
+      "grad_norm": 2.0964109897613525,
+      "learning_rate": 1.8200000000000002e-05,
+      "loss": 3.6273,
+      "step": 186
+    },
+    {
+      "epoch": 0.050815217391304346,
+      "grad_norm": 2.119013547897339,
+      "learning_rate": 1.83e-05,
+      "loss": 3.5527,
+      "step": 187
+    },
+    {
+      "epoch": 0.051086956521739134,
+      "grad_norm": 4.674685478210449,
+      "learning_rate": 1.84e-05,
+      "loss": 4.1131,
+      "step": 188
+    },
+    {
+      "epoch": 0.051358695652173915,
+      "grad_norm": 7.127579689025879,
+      "learning_rate": 1.8500000000000002e-05,
+      "loss": 4.0545,
+      "step": 189
+    },
+    {
+      "epoch": 0.051630434782608696,
+      "grad_norm": 3.285233497619629,
+      "learning_rate": 1.86e-05,
+      "loss": 3.6065,
+      "step": 190
+    },
+    {
+      "epoch": 0.05190217391304348,
+      "grad_norm": 1.405059576034546,
+      "learning_rate": 1.87e-05,
+      "loss": 3.4746,
+      "step": 191
+    },
+    {
+      "epoch": 0.05217391304347826,
+      "grad_norm": 3.002393960952759,
+      "learning_rate": 1.8800000000000003e-05,
+      "loss": 3.9168,
+      "step": 192
+    },
+    {
+      "epoch": 0.052445652173913046,
+      "grad_norm": 2.253774404525757,
+      "learning_rate": 1.8900000000000002e-05,
+      "loss": 3.0605,
+      "step": 193
+    },
+    {
+      "epoch": 0.05271739130434783,
+      "grad_norm": 3.4486026763916016,
+      "learning_rate": 1.9e-05,
+      "loss": 3.6955,
+      "step": 194
+    },
+    {
+      "epoch": 0.05298913043478261,
+      "grad_norm": 2.6076107025146484,
+      "learning_rate": 1.9100000000000003e-05,
+      "loss": 4.7168,
+      "step": 195
+    },
+    {
+      "epoch": 0.05326086956521739,
+      "grad_norm": 2.173379421234131,
+      "learning_rate": 1.9200000000000003e-05,
+      "loss": 3.9563,
+      "step": 196
+    },
+    {
+      "epoch": 0.05353260869565218,
+      "grad_norm": 1.8571799993515015,
+      "learning_rate": 1.9299999999999998e-05,
+      "loss": 3.531,
+      "step": 197
+    },
+    {
+      "epoch": 0.05380434782608696,
+      "grad_norm": 2.940171957015991,
+      "learning_rate": 1.9399999999999997e-05,
+      "loss": 3.9534,
+      "step": 198
+    },
+    {
+      "epoch": 0.05407608695652174,
+      "grad_norm": 1.6403425931930542,
+      "learning_rate": 1.95e-05,
+      "loss": 3.7979,
+      "step": 199
+    },
+    {
+      "epoch": 0.05434782608695652,
+      "grad_norm": 1.9086182117462158,
+      "learning_rate": 1.96e-05,
+      "loss": 3.3714,
+      "step": 200
+    },
+    {
+      "epoch": 0.0546195652173913,
+      "grad_norm": 2.42441463470459,
+      "learning_rate": 1.9699999999999998e-05,
+      "loss": 3.6953,
+      "step": 201
+    },
+    {
+      "epoch": 0.05489130434782609,
+      "grad_norm": 1.7879725694656372,
+      "learning_rate": 1.98e-05,
+      "loss": 3.424,
+      "step": 202
+    },
+    {
+      "epoch": 0.05516304347826087,
+      "grad_norm": 1.951125144958496,
+      "learning_rate": 1.99e-05,
+      "loss": 4.0087,
+      "step": 203
+    },
+    {
+      "epoch": 0.05543478260869565,
+      "grad_norm": 1.8883895874023438,
+      "learning_rate": 1.9999999999999998e-05,
+      "loss": 3.4857,
+      "step": 204
+    },
+    {
+      "epoch": 0.05570652173913043,
+      "grad_norm": 2.1979916095733643,
+      "learning_rate": 2.01e-05,
+      "loss": 3.8898,
+      "step": 205
+    },
+    {
+      "epoch": 0.05597826086956522,
+      "grad_norm": 1.6748876571655273,
+      "learning_rate": 2.02e-05,
+      "loss": 3.5928,
+      "step": 206
+    },
+    {
+      "epoch": 0.05625,
+      "grad_norm": 4.129452705383301,
+      "learning_rate": 2.03e-05,
+      "loss": 4.2804,
+      "step": 207
+    },
+    {
+      "epoch": 0.05652173913043478,
+      "grad_norm": 2.1873629093170166,
+      "learning_rate": 2.04e-05,
+      "loss": 3.5242,
+      "step": 208
+    },
+    {
+      "epoch": 0.05679347826086956,
+      "grad_norm": 3.064663887023926,
+      "learning_rate": 2.05e-05,
+      "loss": 3.379,
+      "step": 209
+    },
+    {
+      "epoch": 0.057065217391304345,
+      "grad_norm": 2.8077127933502197,
+      "learning_rate": 2.06e-05,
+      "loss": 3.4032,
+      "step": 210
+    },
+    {
+      "epoch": 0.05733695652173913,
+      "grad_norm": 2.902247905731201,
+      "learning_rate": 2.07e-05,
+      "loss": 3.9496,
+      "step": 211
+    },
+    {
+      "epoch": 0.057608695652173914,
+      "grad_norm": 2.3877575397491455,
+      "learning_rate": 2.08e-05,
+      "loss": 3.5266,
+      "step": 212
+    },
+    {
+      "epoch": 0.057880434782608695,
+      "grad_norm": 4.157260417938232,
+      "learning_rate": 2.09e-05,
+      "loss": 4.0795,
+      "step": 213
+    },
+    {
+      "epoch": 0.058152173913043476,
+      "grad_norm": 4.117393970489502,
+      "learning_rate": 2.1e-05,
+      "loss": 3.3584,
+      "step": 214
+    },
+    {
+      "epoch": 0.058423913043478264,
+      "grad_norm": 3.1118719577789307,
+      "learning_rate": 2.11e-05,
+      "loss": 3.5,
+      "step": 215
+    },
+    {
+      "epoch": 0.058695652173913045,
+      "grad_norm": 3.3953042030334473,
+      "learning_rate": 2.12e-05,
+      "loss": 3.6948,
+      "step": 216
+    },
+    {
+      "epoch": 0.058967391304347826,
+      "grad_norm": 4.036997318267822,
+      "learning_rate": 2.13e-05,
+      "loss": 3.6871,
+      "step": 217
+    },
+    {
+      "epoch": 0.05923913043478261,
+      "grad_norm": 4.334664821624756,
+      "learning_rate": 2.1400000000000002e-05,
+      "loss": 4.0403,
+      "step": 218
+    },
+    {
+      "epoch": 0.059510869565217395,
+      "grad_norm": 3.2473344802856445,
+      "learning_rate": 2.15e-05,
+      "loss": 2.8208,
+      "step": 219
+    },
+    {
+      "epoch": 0.059782608695652176,
+      "grad_norm": 5.333037376403809,
+      "learning_rate": 2.16e-05,
+      "loss": 3.4915,
+      "step": 220
+    },
+    {
+      "epoch": 0.06005434782608696,
+      "grad_norm": 4.594522953033447,
+      "learning_rate": 2.1700000000000002e-05,
+      "loss": 3.773,
+      "step": 221
+    },
+    {
+      "epoch": 0.06032608695652174,
+      "grad_norm": 2.8913493156433105,
+      "learning_rate": 2.18e-05,
+      "loss": 3.5039,
+      "step": 222
+    },
+    {
+      "epoch": 0.06059782608695652,
+      "grad_norm": 3.1354737281799316,
+      "learning_rate": 2.19e-05,
+      "loss": 3.5823,
+      "step": 223
+    },
+    {
+      "epoch": 0.06086956521739131,
+      "grad_norm": 4.392479419708252,
+      "learning_rate": 2.2e-05,
+      "loss": 3.8093,
+      "step": 224
+    },
+    {
+      "epoch": 0.06114130434782609,
+      "grad_norm": 3.4441139698028564,
+      "learning_rate": 2.2100000000000002e-05,
+      "loss": 3.9232,
+      "step": 225
+    },
+    {
+      "epoch": 0.06141304347826087,
+      "grad_norm": 2.799862861633301,
+      "learning_rate": 2.22e-05,
+      "loss": 3.9281,
+      "step": 226
+    },
+    {
+      "epoch": 0.06168478260869565,
+      "grad_norm": 2.910454750061035,
+      "learning_rate": 2.23e-05,
+      "loss": 3.635,
+      "step": 227
+    },
+    {
+      "epoch": 0.06195652173913044,
+      "grad_norm": 3.9778056144714355,
+      "learning_rate": 2.2400000000000002e-05,
+      "loss": 3.5988,
+      "step": 228
+    },
+    {
+      "epoch": 0.06222826086956522,
+      "grad_norm": 3.500877857208252,
+      "learning_rate": 2.25e-05,
+      "loss": 3.2595,
+      "step": 229
+    },
+    {
+      "epoch": 0.0625,
+      "grad_norm": 3.3998258113861084,
+      "learning_rate": 2.26e-05,
+      "loss": 3.9116,
+      "step": 230
+    },
+    {
+      "epoch": 0.06277173913043478,
+      "grad_norm": 2.585266351699829,
+      "learning_rate": 2.2700000000000003e-05,
+      "loss": 3.6087,
+      "step": 231
+    },
+    {
+      "epoch": 0.06304347826086956,
+      "grad_norm": 2.313265085220337,
+      "learning_rate": 2.2800000000000002e-05,
+      "loss": 3.8507,
+      "step": 232
+    },
+    {
+      "epoch": 0.06331521739130434,
+      "grad_norm": 6.233187198638916,
+      "learning_rate": 2.29e-05,
+      "loss": 4.6271,
+      "step": 233
+    },
+    {
+      "epoch": 0.06358695652173912,
+      "grad_norm": 4.387447357177734,
+      "learning_rate": 2.3000000000000003e-05,
+      "loss": 3.7921,
+      "step": 234
+    },
+    {
+      "epoch": 0.06385869565217392,
+      "grad_norm": 3.573700189590454,
+      "learning_rate": 2.3100000000000002e-05,
+      "loss": 3.6567,
+      "step": 235
+    },
+    {
+      "epoch": 0.0641304347826087,
+      "grad_norm": 1.9834128618240356,
+      "learning_rate": 2.32e-05,
+      "loss": 3.715,
+      "step": 236
+    },
+    {
+      "epoch": 0.06440217391304348,
+      "grad_norm": 3.0220143795013428,
+      "learning_rate": 2.33e-05,
+      "loss": 4.302,
+      "step": 237
+    },
+    {
+      "epoch": 0.06467391304347826,
+      "grad_norm": 5.85452938079834,
+      "learning_rate": 2.3400000000000003e-05,
+      "loss": 4.3755,
+      "step": 238
+    },
+    {
+      "epoch": 0.06494565217391304,
+      "grad_norm": 6.263114929199219,
+      "learning_rate": 2.3500000000000002e-05,
+      "loss": 4.0045,
+      "step": 239
+    },
+    {
+      "epoch": 0.06521739130434782,
+      "grad_norm": 4.6512322425842285,
+      "learning_rate": 2.3599999999999998e-05,
+      "loss": 4.202,
+      "step": 240
+    },
+    {
+      "epoch": 0.0654891304347826,
+      "grad_norm": 1.589857578277588,
+      "learning_rate": 2.37e-05,
+      "loss": 3.0368,
+      "step": 241
+    },
+    {
+      "epoch": 0.06576086956521739,
+      "grad_norm": 3.0660102367401123,
+      "learning_rate": 2.38e-05,
+      "loss": 3.959,
+      "step": 242
+    },
+    {
+      "epoch": 0.06603260869565217,
+      "grad_norm": 3.7164723873138428,
+      "learning_rate": 2.3899999999999998e-05,
+      "loss": 3.5374,
+      "step": 243
+    },
+    {
+      "epoch": 0.06630434782608696,
+      "grad_norm": 3.772892713546753,
+      "learning_rate": 2.4e-05,
+      "loss": 3.3464,
+      "step": 244
+    },
+    {
+      "epoch": 0.06657608695652174,
+      "grad_norm": 3.2034425735473633,
+      "learning_rate": 2.41e-05,
+      "loss": 3.3077,
+      "step": 245
+    },
+    {
+      "epoch": 0.06684782608695652,
+      "grad_norm": 4.294682502746582,
+      "learning_rate": 2.42e-05,
+      "loss": 3.813,
+      "step": 246
+    },
+    {
+      "epoch": 0.0671195652173913,
+      "grad_norm": 3.805955171585083,
+      "learning_rate": 2.43e-05,
+      "loss": 3.6854,
+      "step": 247
+    },
+    {
+      "epoch": 0.06739130434782609,
+      "grad_norm": 4.587597370147705,
+      "learning_rate": 2.44e-05,
+      "loss": 3.3058,
+      "step": 248
+    },
+    {
+      "epoch": 0.06766304347826087,
+      "grad_norm": 5.513325214385986,
+      "learning_rate": 2.45e-05,
+      "loss": 3.6714,
+      "step": 249
+    },
+    {
+      "epoch": 0.06793478260869565,
+      "grad_norm": 5.820981025695801,
+      "learning_rate": 2.4599999999999998e-05,
+      "loss": 3.6253,
+      "step": 250
+    },
+    {
+      "epoch": 0.06820652173913043,
+      "grad_norm": 4.236602306365967,
+      "learning_rate": 2.47e-05,
+      "loss": 3.605,
+      "step": 251
+    },
+    {
+      "epoch": 0.06847826086956521,
+      "grad_norm": 3.524596929550171,
+      "learning_rate": 2.48e-05,
+      "loss": 3.6746,
+      "step": 252
+    },
+    {
+      "epoch": 0.06875,
+      "grad_norm": 4.912044525146484,
+      "learning_rate": 2.49e-05,
+      "loss": 3.6485,
+      "step": 253
+    },
+    {
+      "epoch": 0.06902173913043479,
+      "grad_norm": 5.730941295623779,
+      "learning_rate": 2.5e-05,
+      "loss": 3.7662,
+      "step": 254
+    },
+    {
+      "epoch": 0.06929347826086957,
+      "grad_norm": 7.801654815673828,
+      "learning_rate": 2.51e-05,
+      "loss": 4.8335,
+      "step": 255
+    },
+    {
+      "epoch": 0.06956521739130435,
+      "grad_norm": 4.116402626037598,
+      "learning_rate": 2.52e-05,
+      "loss": 3.9456,
+      "step": 256
+    },
+    {
+      "epoch": 0.06983695652173913,
+      "grad_norm": 6.192750930786133,
+      "learning_rate": 2.5300000000000002e-05,
+      "loss": 4.0458,
+      "step": 257
+    },
+    {
+      "epoch": 0.07010869565217391,
+      "grad_norm": 4.565896511077881,
+      "learning_rate": 2.54e-05,
+      "loss": 3.3915,
+      "step": 258
+    },
+    {
+      "epoch": 0.07038043478260869,
+      "grad_norm": 4.502013206481934,
+      "learning_rate": 2.55e-05,
+      "loss": 3.4612,
+      "step": 259
+    },
+    {
+      "epoch": 0.07065217391304347,
+      "grad_norm": 3.8033788204193115,
+      "learning_rate": 2.5600000000000002e-05,
+      "loss": 2.9393,
+      "step": 260
+    },
+    {
+      "epoch": 0.07092391304347827,
+      "grad_norm": 4.34993839263916,
+      "learning_rate": 2.57e-05,
+      "loss": 3.6575,
+      "step": 261
+    },
+    {
+      "epoch": 0.07119565217391305,
+      "grad_norm": 3.9064600467681885,
+      "learning_rate": 2.58e-05,
+      "loss": 3.6109,
+      "step": 262
+    },
+    {
+      "epoch": 0.07146739130434783,
+      "grad_norm": 3.737865447998047,
+      "learning_rate": 2.59e-05,
+      "loss": 3.7551,
+      "step": 263
+    },
+    {
+      "epoch": 0.07173913043478261,
+      "grad_norm": 4.92514181137085,
+      "learning_rate": 2.6000000000000002e-05,
+      "loss": 3.9133,
+      "step": 264
+    },
+    {
+      "epoch": 0.07201086956521739,
+      "grad_norm": 4.838226318359375,
+      "learning_rate": 2.61e-05,
+      "loss": 3.4457,
+      "step": 265
+    },
+    {
+      "epoch": 0.07228260869565217,
+      "grad_norm": 5.465858459472656,
+      "learning_rate": 2.62e-05,
+      "loss": 3.3492,
+      "step": 266
+    },
+    {
+      "epoch": 0.07255434782608695,
+      "grad_norm": 4.452864646911621,
+      "learning_rate": 2.6300000000000002e-05,
+      "loss": 4.0503,
+      "step": 267
+    },
+    {
+      "epoch": 0.07282608695652174,
+      "grad_norm": 2.657294750213623,
+      "learning_rate": 2.64e-05,
+      "loss": 3.5952,
+      "step": 268
+    },
+    {
+      "epoch": 0.07309782608695652,
+      "grad_norm": 3.738781213760376,
+      "learning_rate": 2.65e-05,
+      "loss": 3.2922,
+      "step": 269
+    },
+    {
+      "epoch": 0.07336956521739131,
+      "grad_norm": 4.5520524978637695,
+      "learning_rate": 2.6600000000000003e-05,
+      "loss": 3.4737,
+      "step": 270
+    },
+    {
+      "epoch": 0.07364130434782609,
+      "grad_norm": 6.709564685821533,
+      "learning_rate": 2.6700000000000002e-05,
+      "loss": 3.8822,
+      "step": 271
+    },
+    {
+      "epoch": 0.07391304347826087,
+      "grad_norm": 5.555806636810303,
+      "learning_rate": 2.68e-05,
+      "loss": 4.1245,
+      "step": 272
+    },
+    {
+      "epoch": 0.07418478260869565,
+      "grad_norm": 3.99271821975708,
+      "learning_rate": 2.69e-05,
+      "loss": 3.535,
+      "step": 273
+    },
+    {
+      "epoch": 0.07445652173913044,
+      "grad_norm": 3.0163462162017822,
+      "learning_rate": 2.7000000000000002e-05,
+      "loss": 3.4727,
+      "step": 274
+    },
+    {
+      "epoch": 0.07472826086956522,
+      "grad_norm": 5.363269329071045,
+      "learning_rate": 2.71e-05,
+      "loss": 4.1682,
+      "step": 275
+    },
+    {
+      "epoch": 0.075,
+      "grad_norm": 4.282686710357666,
+      "learning_rate": 2.72e-05,
+      "loss": 3.1355,
+      "step": 276
+    },
+    {
+      "epoch": 0.07527173913043478,
+      "grad_norm": 5.983767509460449,
+      "learning_rate": 2.7300000000000003e-05,
+      "loss": 3.8695,
+      "step": 277
+    },
+    {
+      "epoch": 0.07554347826086956,
+      "grad_norm": 5.345359802246094,
+      "learning_rate": 2.7400000000000002e-05,
+      "loss": 4.1882,
+      "step": 278
+    },
+    {
+      "epoch": 0.07581521739130435,
+      "grad_norm": 2.5548171997070312,
+      "learning_rate": 2.75e-05,
+      "loss": 3.2401,
+      "step": 279
+    },
+    {
+      "epoch": 0.07608695652173914,
+      "grad_norm": 3.790065050125122,
+      "learning_rate": 2.7600000000000003e-05,
+      "loss": 3.8057,
+      "step": 280
+    },
+    {
+      "epoch": 0.07635869565217392,
+      "grad_norm": 4.037810325622559,
+      "learning_rate": 2.7700000000000002e-05,
+      "loss": 3.8397,
+      "step": 281
+    },
+    {
+      "epoch": 0.0766304347826087,
+      "grad_norm": 4.571537017822266,
+      "learning_rate": 2.78e-05,
+      "loss": 3.1902,
+      "step": 282
+    },
+    {
+      "epoch": 0.07690217391304348,
+      "grad_norm": 4.4513702392578125,
+      "learning_rate": 2.79e-05,
+      "loss": 3.6552,
+      "step": 283
+    },
+    {
+      "epoch": 0.07717391304347826,
+      "grad_norm": 2.919421911239624,
+      "learning_rate": 2.8e-05,
+      "loss": 3.0449,
+      "step": 284
+    },
+    {
+      "epoch": 0.07744565217391304,
+      "grad_norm": 3.9303693771362305,
+      "learning_rate": 2.81e-05,
+      "loss": 3.8186,
+      "step": 285
+    },
+    {
+      "epoch": 0.07771739130434782,
+      "grad_norm": 3.8404746055603027,
+      "learning_rate": 2.8199999999999998e-05,
+      "loss": 3.2603,
+      "step": 286
+    },
+    {
+      "epoch": 0.0779891304347826,
+      "grad_norm": 3.2469253540039062,
+      "learning_rate": 2.83e-05,
+      "loss": 2.917,
+      "step": 287
+    },
+    {
+      "epoch": 0.0782608695652174,
+      "grad_norm": 2.8970375061035156,
+      "learning_rate": 2.84e-05,
+      "loss": 3.6446,
+      "step": 288
+    },
+    {
+      "epoch": 0.07853260869565218,
+      "grad_norm": 4.029104232788086,
+      "learning_rate": 2.8499999999999998e-05,
+      "loss": 4.1754,
+      "step": 289
+    },
+    {
+      "epoch": 0.07880434782608696,
+      "grad_norm": 3.967705726623535,
+      "learning_rate": 2.86e-05,
+      "loss": 3.3913,
+      "step": 290
+    },
+    {
+      "epoch": 0.07907608695652174,
+      "grad_norm": 3.084944009780884,
+      "learning_rate": 2.87e-05,
+      "loss": 3.4598,
+      "step": 291
+    },
+    {
+      "epoch": 0.07934782608695652,
+      "grad_norm": 3.2562334537506104,
+      "learning_rate": 2.88e-05,
+      "loss": 3.7326,
+      "step": 292
+    },
+    {
+      "epoch": 0.0796195652173913,
+      "grad_norm": 3.9668397903442383,
+      "learning_rate": 2.89e-05,
+      "loss": 3.8312,
+      "step": 293
+    },
+    {
+      "epoch": 0.07989130434782608,
+      "grad_norm": 3.7607946395874023,
+      "learning_rate": 2.9e-05,
+      "loss": 3.9309,
+      "step": 294
+    },
+    {
+      "epoch": 0.08016304347826086,
+      "grad_norm": 4.704547882080078,
+      "learning_rate": 2.91e-05,
+      "loss": 3.9181,
+      "step": 295
+    },
+    {
+      "epoch": 0.08043478260869565,
+      "grad_norm": 5.68735408782959,
+      "learning_rate": 2.92e-05,
+      "loss": 3.6405,
+      "step": 296
+    },
+    {
+      "epoch": 0.08070652173913044,
+      "grad_norm": 4.839725971221924,
+      "learning_rate": 2.93e-05,
+      "loss": 3.6735,
+      "step": 297
+    },
+    {
+      "epoch": 0.08097826086956522,
+      "grad_norm": 3.6157710552215576,
+      "learning_rate": 2.94e-05,
+      "loss": 4.1132,
+      "step": 298
+    },
+    {
+      "epoch": 0.08125,
+      "grad_norm": 3.6725997924804688,
+      "learning_rate": 2.95e-05,
+      "loss": 3.9299,
+      "step": 299
+    },
+    {
+      "epoch": 0.08152173913043478,
+      "grad_norm": 4.408240795135498,
+      "learning_rate": 2.96e-05,
+      "loss": 3.4141,
+      "step": 300
+    },
+    {
+      "epoch": 0.08179347826086956,
+      "grad_norm": 3.8871631622314453,
+      "learning_rate": 2.97e-05,
+      "loss": 2.7057,
+      "step": 301
+    },
+    {
+      "epoch": 0.08206521739130435,
+      "grad_norm": 5.148252010345459,
+      "learning_rate": 2.98e-05,
+      "loss": 3.0604,
+      "step": 302
+    },
+    {
+      "epoch": 0.08233695652173913,
+      "grad_norm": 3.8631784915924072,
+      "learning_rate": 2.9900000000000002e-05,
+      "loss": 3.4581,
+      "step": 303
+    },
+    {
+      "epoch": 0.08260869565217391,
+      "grad_norm": 2.67698073387146,
+      "learning_rate": 3e-05,
+      "loss": 3.1465,
+      "step": 304
+    },
+    {
+      "epoch": 0.08288043478260869,
+      "grad_norm": 2.6791391372680664,
+      "learning_rate": 3e-05,
+      "loss": 2.8593,
+      "step": 305
+    },
+    {
+      "epoch": 0.08315217391304348,
+      "grad_norm": 3.844003677368164,
+      "learning_rate": 3e-05,
+      "loss": 3.4371,
+      "step": 306
+    },
+    {
+      "epoch": 0.08342391304347826,
+      "grad_norm": 5.624002933502197,
+      "learning_rate": 3e-05,
+      "loss": 3.8372,
+      "step": 307
+    },
+    {
+      "epoch": 0.08369565217391305,
+      "grad_norm": 4.992491245269775,
+      "learning_rate": 3e-05,
+      "loss": 3.6231,
+      "step": 308
+    },
+    {
+      "epoch": 0.08396739130434783,
+      "grad_norm": 4.717808246612549,
+      "learning_rate": 3e-05,
+      "loss": 3.9652,
+      "step": 309
+    },
+    {
+      "epoch": 0.08423913043478261,
+      "grad_norm": 3.6492786407470703,
+      "learning_rate": 3e-05,
+      "loss": 3.7065,
+      "step": 310
+    },
+    {
+      "epoch": 0.08451086956521739,
+      "grad_norm": 4.453199863433838,
+      "learning_rate": 3e-05,
+      "loss": 3.7317,
+      "step": 311
+    },
+    {
+      "epoch": 0.08478260869565217,
+      "grad_norm": 5.009824275970459,
+      "learning_rate": 3e-05,
+      "loss": 3.7651,
+      "step": 312
+    },
+    {
+      "epoch": 0.08505434782608695,
+      "grad_norm": 6.800715446472168,
+      "learning_rate": 3e-05,
+      "loss": 4.0778,
+      "step": 313
+    },
+    {
+      "epoch": 0.08532608695652173,
+      "grad_norm": 6.225503444671631,
+      "learning_rate": 3e-05,
+      "loss": 4.3679,
+      "step": 314
+    },
+    {
+      "epoch": 0.08559782608695653,
+      "grad_norm": 2.9776594638824463,
+      "learning_rate": 3e-05,
+      "loss": 3.1434,
+      "step": 315
+    },
+    {
+      "epoch": 0.08586956521739131,
+      "grad_norm": 3.1179118156433105,
+      "learning_rate": 3e-05,
+      "loss": 3.5896,
+      "step": 316
+    },
+    {
+      "epoch": 0.08614130434782609,
+      "grad_norm": 2.81669282913208,
+      "learning_rate": 3e-05,
+      "loss": 2.9054,
+      "step": 317
+    },
+    {
+      "epoch": 0.08641304347826087,
+      "grad_norm": 3.894665002822876,
+      "learning_rate": 3e-05,
+      "loss": 3.2122,
+      "step": 318
+    },
+    {
+      "epoch": 0.08668478260869565,
+      "grad_norm": 5.43796443939209,
+      "learning_rate": 3e-05,
+      "loss": 3.8039,
+      "step": 319
+    },
+    {
+      "epoch": 0.08695652173913043,
+      "grad_norm": 3.9177141189575195,
+      "learning_rate": 3e-05,
+      "loss": 3.6393,
+      "step": 320
+    },
+    {
+      "epoch": 0.08722826086956521,
+      "grad_norm": 3.0576508045196533,
+      "learning_rate": 3e-05,
+      "loss": 3.67,
+      "step": 321
+    },
+    {
+      "epoch": 0.0875,
+      "grad_norm": 5.208736896514893,
+      "learning_rate": 3e-05,
+      "loss": 3.9719,
+      "step": 322
+    },
+    {
+      "epoch": 0.08777173913043479,
+      "grad_norm": 4.839489936828613,
+      "learning_rate": 3e-05,
+      "loss": 3.7197,
+      "step": 323
+    },
+    {
+      "epoch": 0.08804347826086957,
+      "grad_norm": 4.195175647735596,
+      "learning_rate": 3e-05,
+      "loss": 3.3963,
+      "step": 324
+    },
+    {
+      "epoch": 0.08831521739130435,
+      "grad_norm": 3.565690279006958,
+      "learning_rate": 3e-05,
+      "loss": 3.177,
+      "step": 325
+    },
+    {
+      "epoch": 0.08858695652173913,
+      "grad_norm": 2.6973624229431152,
+      "learning_rate": 3e-05,
+      "loss": 3.0416,
+      "step": 326
+    },
+    {
+      "epoch": 0.08885869565217391,
+      "grad_norm": 2.620365619659424,
+      "learning_rate": 3e-05,
+      "loss": 3.3155,
+      "step": 327
+    },
+    {
+      "epoch": 0.0891304347826087,
+      "grad_norm": 3.7461090087890625,
+      "learning_rate": 3e-05,
+      "loss": 3.6553,
+      "step": 328
+    },
+    {
+      "epoch": 0.08940217391304348,
+      "grad_norm": 6.702042102813721,
+      "learning_rate": 3e-05,
+      "loss": 4.2574,
+      "step": 329
+    },
+    {
+      "epoch": 0.08967391304347826,
+      "grad_norm": 5.888139247894287,
+      "learning_rate": 3e-05,
+      "loss": 4.148,
+      "step": 330
+    },
+    {
+      "epoch": 0.08994565217391304,
+      "grad_norm": 4.737085342407227,
+      "learning_rate": 3e-05,
+      "loss": 3.8761,
+      "step": 331
+    },
+    {
+      "epoch": 0.09021739130434783,
+      "grad_norm": 3.9981937408447266,
+      "learning_rate": 3e-05,
+      "loss": 4.4605,
+      "step": 332
+    },
+    {
+      "epoch": 0.09048913043478261,
+      "grad_norm": 5.0993266105651855,
+      "learning_rate": 3e-05,
+      "loss": 4.095,
+      "step": 333
+    },
+    {
+      "epoch": 0.0907608695652174,
+      "grad_norm": 4.450422763824463,
+      "learning_rate": 3e-05,
+      "loss": 2.8801,
+      "step": 334
+    },
+    {
+      "epoch": 0.09103260869565218,
+      "grad_norm": 6.846070766448975,
+      "learning_rate": 3e-05,
+      "loss": 3.5946,
+      "step": 335
+    },
+    {
+      "epoch": 0.09130434782608696,
+      "grad_norm": 5.887484550476074,
+      "learning_rate": 3e-05,
+      "loss": 3.4395,
+      "step": 336
+    },
+    {
+      "epoch": 0.09157608695652174,
+      "grad_norm": 3.8128914833068848,
+      "learning_rate": 3e-05,
+      "loss": 3.0863,
+      "step": 337
+    },
+    {
+      "epoch": 0.09184782608695652,
+      "grad_norm": 2.493584394454956,
+      "learning_rate": 3e-05,
+      "loss": 3.0949,
+      "step": 338
+    },
+    {
+      "epoch": 0.0921195652173913,
+      "grad_norm": 3.304899215698242,
+      "learning_rate": 3e-05,
+      "loss": 3.3348,
+      "step": 339
+    },
+    {
+      "epoch": 0.09239130434782608,
+      "grad_norm": 4.633394241333008,
+      "learning_rate": 3e-05,
+      "loss": 3.5233,
+      "step": 340
+    },
+    {
+      "epoch": 0.09266304347826088,
+      "grad_norm": 5.405637264251709,
+      "learning_rate": 3e-05,
+      "loss": 3.776,
+      "step": 341
+    },
+    {
+      "epoch": 0.09293478260869566,
+      "grad_norm": 5.505592346191406,
+      "learning_rate": 3e-05,
+      "loss": 3.9894,
+      "step": 342
+    },
+    {
+      "epoch": 0.09320652173913044,
+      "grad_norm": 3.901547431945801,
+      "learning_rate": 3e-05,
+      "loss": 4.1865,
+      "step": 343
+    },
+    {
+      "epoch": 0.09347826086956522,
+      "grad_norm": 3.3303725719451904,
+      "learning_rate": 3e-05,
+      "loss": 3.3656,
+      "step": 344
+    },
+    {
+      "epoch": 0.09375,
+      "grad_norm": 5.3323493003845215,
+      "learning_rate": 3e-05,
+      "loss": 4.2369,
+      "step": 345
+    },
+    {
+      "epoch": 0.09402173913043478,
+      "grad_norm": 5.352675914764404,
+      "learning_rate": 3e-05,
+      "loss": 3.7106,
+      "step": 346
+    },
+    {
+      "epoch": 0.09429347826086956,
+      "grad_norm": 5.6517839431762695,
+      "learning_rate": 3e-05,
+      "loss": 3.4388,
+      "step": 347
+    },
+    {
+      "epoch": 0.09456521739130434,
+      "grad_norm": 4.7645111083984375,
+      "learning_rate": 3e-05,
+      "loss": 3.4332,
+      "step": 348
+    },
+    {
+      "epoch": 0.09483695652173912,
+      "grad_norm": 3.140958309173584,
+      "learning_rate": 3e-05,
+      "loss": 3.2139,
+      "step": 349
+    },
+    {
+      "epoch": 0.09510869565217392,
+      "grad_norm": 2.692319393157959,
+      "learning_rate": 3e-05,
+      "loss": 3.3015,
+      "step": 350
+    },
+    {
+      "epoch": 0.0953804347826087,
+      "grad_norm": 4.1542134284973145,
+      "learning_rate": 3e-05,
+      "loss": 3.29,
+      "step": 351
+    },
+    {
+      "epoch": 0.09565217391304348,
+      "grad_norm": 4.946714401245117,
+      "learning_rate": 3e-05,
+      "loss": 3.2509,
+      "step": 352
+    },
+    {
+      "epoch": 0.09592391304347826,
+      "grad_norm": 5.529204368591309,
+      "learning_rate": 3e-05,
+      "loss": 4.0578,
+      "step": 353
+    },
+    {
+      "epoch": 0.09619565217391304,
+      "grad_norm": 3.837975025177002,
+      "learning_rate": 3e-05,
+      "loss": 3.7921,
+      "step": 354
+    },
+    {
+      "epoch": 0.09646739130434782,
+      "grad_norm": 3.0146334171295166,
+      "learning_rate": 3e-05,
+      "loss": 4.0369,
+      "step": 355
+    },
+    {
+      "epoch": 0.0967391304347826,
+      "grad_norm": 4.283068656921387,
+      "learning_rate": 3e-05,
+      "loss": 3.9506,
+      "step": 356
+    },
+    {
+      "epoch": 0.09701086956521739,
+      "grad_norm": 4.30133581161499,
+      "learning_rate": 3e-05,
+      "loss": 3.4218,
+      "step": 357
+    },
+    {
+      "epoch": 0.09728260869565217,
+      "grad_norm": 6.228667736053467,
+      "learning_rate": 3e-05,
+      "loss": 4.0054,
+      "step": 358
+    },
+    {
+      "epoch": 0.09755434782608696,
+      "grad_norm": 3.927131414413452,
+      "learning_rate": 3e-05,
+      "loss": 2.9401,
+      "step": 359
+    },
+    {
+      "epoch": 0.09782608695652174,
+      "grad_norm": 4.374425888061523,
+      "learning_rate": 3e-05,
+      "loss": 3.5279,
+      "step": 360
+    },
+    {
+      "epoch": 0.09809782608695652,
+      "grad_norm": 2.559403419494629,
+      "learning_rate": 3e-05,
+      "loss": 2.8306,
+      "step": 361
+    },
+    {
+      "epoch": 0.0983695652173913,
+      "grad_norm": 2.9556596279144287,
+      "learning_rate": 3e-05,
+      "loss": 3.4532,
+      "step": 362
+    },
+    {
+      "epoch": 0.09864130434782609,
+      "grad_norm": 4.841480731964111,
+      "learning_rate": 3e-05,
+      "loss": 3.474,
+      "step": 363
+    },
+    {
+      "epoch": 0.09891304347826087,
+      "grad_norm": 5.785233497619629,
+      "learning_rate": 3e-05,
+      "loss": 3.6698,
+      "step": 364
+    },
+    {
+      "epoch": 0.09918478260869565,
+      "grad_norm": 5.180464267730713,
+      "learning_rate": 3e-05,
+      "loss": 3.4199,
+      "step": 365
+    },
+    {
+      "epoch": 0.09945652173913043,
+      "grad_norm": 3.5438666343688965,
+      "learning_rate": 3e-05,
+      "loss": 3.5603,
+      "step": 366
+    },
+    {
+      "epoch": 0.09972826086956521,
+      "grad_norm": 3.1015937328338623,
+      "learning_rate": 3e-05,
+      "loss": 2.9937,
+      "step": 367
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.174754619598389,
+      "learning_rate": 3e-05,
+      "loss": 3.5994,
+      "step": 368
+    },
+    {
+      "epoch": 0.10027173913043479,
+      "grad_norm": 4.678737163543701,
+      "learning_rate": 3e-05,
+      "loss": 4.0816,
+      "step": 369
+    },
+    {
+      "epoch": 0.10054347826086957,
+      "grad_norm": 3.884875774383545,
+      "learning_rate": 3e-05,
+      "loss": 2.8268,
+      "step": 370
+    },
+    {
+      "epoch": 0.10081521739130435,
+      "grad_norm": 4.060309886932373,
+      "learning_rate": 3e-05,
+      "loss": 3.5038,
+      "step": 371
+    },
+    {
+      "epoch": 0.10108695652173913,
+      "grad_norm": 4.1803812980651855,
+      "learning_rate": 3e-05,
+      "loss": 3.8122,
+      "step": 372
+    },
+    {
+      "epoch": 0.10135869565217391,
+      "grad_norm": 3.1725456714630127,
+      "learning_rate": 3e-05,
+      "loss": 3.7945,
+      "step": 373
+    },
+    {
+      "epoch": 0.10163043478260869,
+      "grad_norm": 3.1784756183624268,
+      "learning_rate": 3e-05,
+      "loss": 3.0676,
+      "step": 374
+    },
+    {
+      "epoch": 0.10190217391304347,
+      "grad_norm": 4.812568187713623,
+      "learning_rate": 3e-05,
+      "loss": 3.593,
+      "step": 375
+    },
+    {
+      "epoch": 0.10217391304347827,
+      "grad_norm": 5.944811820983887,
+      "learning_rate": 3e-05,
+      "loss": 3.843,
+      "step": 376
+    },
+    {
+      "epoch": 0.10244565217391305,
+      "grad_norm": 4.13525915145874,
+      "learning_rate": 3e-05,
+      "loss": 3.9103,
+      "step": 377
+    },
+    {
+      "epoch": 0.10271739130434783,
+      "grad_norm": 5.860477447509766,
+      "learning_rate": 3e-05,
+      "loss": 4.9384,
+      "step": 378
+    },
+    {
+      "epoch": 0.10298913043478261,
+      "grad_norm": 3.2515783309936523,
+      "learning_rate": 3e-05,
+      "loss": 3.0586,
+      "step": 379
+    },
+    {
+      "epoch": 0.10326086956521739,
+      "grad_norm": 4.5452494621276855,
+      "learning_rate": 3e-05,
+      "loss": 3.7046,
+      "step": 380
+    },
+    {
+      "epoch": 0.10353260869565217,
+      "grad_norm": 3.8891849517822266,
+      "learning_rate": 3e-05,
+      "loss": 3.34,
+      "step": 381
+    },
+    {
+      "epoch": 0.10380434782608695,
+      "grad_norm": 4.083732604980469,
+      "learning_rate": 3e-05,
+      "loss": 3.3112,
+      "step": 382
+    },
+    {
+      "epoch": 0.10407608695652174,
+      "grad_norm": 3.438572406768799,
+      "learning_rate": 3e-05,
+      "loss": 2.9908,
+      "step": 383
+    },
+    {
+      "epoch": 0.10434782608695652,
+      "grad_norm": 3.0750012397766113,
+      "learning_rate": 3e-05,
+      "loss": 3.3847,
+      "step": 384
+    },
+    {
+      "epoch": 0.10461956521739131,
+      "grad_norm": 3.0833847522735596,
+      "learning_rate": 3e-05,
+      "loss": 3.2274,
+      "step": 385
+    },
+    {
+      "epoch": 0.10489130434782609,
+      "grad_norm": 4.356601238250732,
+      "learning_rate": 3e-05,
+      "loss": 3.9247,
+      "step": 386
+    },
+    {
+      "epoch": 0.10516304347826087,
+      "grad_norm": 4.894799709320068,
+      "learning_rate": 3e-05,
+      "loss": 3.6005,
+      "step": 387
+    },
+    {
+      "epoch": 0.10543478260869565,
+      "grad_norm": 4.887790203094482,
+      "learning_rate": 3e-05,
+      "loss": 4.3258,
+      "step": 388
+    },
+    {
+      "epoch": 0.10570652173913044,
+      "grad_norm": 4.861374855041504,
+      "learning_rate": 3e-05,
+      "loss": 3.6857,
+      "step": 389
+    },
+    {
+      "epoch": 0.10597826086956522,
+      "grad_norm": 4.775892734527588,
+      "learning_rate": 3e-05,
+      "loss": 4.3695,
+      "step": 390
+    },
+    {
+      "epoch": 0.10625,
+      "grad_norm": 4.2715582847595215,
+      "learning_rate": 3e-05,
+      "loss": 3.7657,
+      "step": 391
+    },
+    {
+      "epoch": 0.10652173913043478,
+      "grad_norm": 3.8440310955047607,
+      "learning_rate": 3e-05,
+      "loss": 3.4856,
+      "step": 392
+    },
+    {
+      "epoch": 0.10679347826086956,
+      "grad_norm": 4.285994529724121,
+      "learning_rate": 3e-05,
+      "loss": 3.4081,
+      "step": 393
+    },
+    {
+      "epoch": 0.10706521739130435,
+      "grad_norm": 2.1318747997283936,
+      "learning_rate": 3e-05,
+      "loss": 2.9345,
+      "step": 394
+    },
+    {
+      "epoch": 0.10733695652173914,
+      "grad_norm": 2.4292728900909424,
+      "learning_rate": 3e-05,
+      "loss": 3.091,
+      "step": 395
+    },
+    {
+      "epoch": 0.10760869565217392,
+      "grad_norm": 2.3496248722076416,
+      "learning_rate": 3e-05,
+      "loss": 3.1994,
+      "step": 396
+    },
+    {
+      "epoch": 0.1078804347826087,
+      "grad_norm": 5.604576110839844,
+      "learning_rate": 3e-05,
+      "loss": 4.6392,
+      "step": 397
+    },
+    {
+      "epoch": 0.10815217391304348,
+      "grad_norm": 5.448110580444336,
+      "learning_rate": 3e-05,
+      "loss": 4.4596,
+      "step": 398
+    },
+    {
+      "epoch": 0.10842391304347826,
+      "grad_norm": 3.2921388149261475,
+      "learning_rate": 3e-05,
+      "loss": 3.5248,
+      "step": 399
+    },
+    {
+      "epoch": 0.10869565217391304,
+      "grad_norm": 3.1395716667175293,
+      "learning_rate": 3e-05,
+      "loss": 4.1658,
+      "step": 400
+    },
+    {
+      "epoch": 0.10896739130434782,
+      "grad_norm": 2.3313381671905518,
+      "learning_rate": 3e-05,
+      "loss": 3.7592,
+      "step": 401
+    },
+    {
+      "epoch": 0.1092391304347826,
+      "grad_norm": 2.9619717597961426,
+      "learning_rate": 3e-05,
+      "loss": 3.5212,
+      "step": 402
+    },
+    {
+      "epoch": 0.1095108695652174,
+      "grad_norm": 3.2616071701049805,
+      "learning_rate": 3e-05,
+      "loss": 3.2052,
+      "step": 403
+    },
+    {
+      "epoch": 0.10978260869565218,
+      "grad_norm": 4.842064380645752,
+      "learning_rate": 3e-05,
+      "loss": 3.907,
+      "step": 404
+    },
+    {
+      "epoch": 0.11005434782608696,
+      "grad_norm": 3.6140100955963135,
+      "learning_rate": 3e-05,
+      "loss": 4.061,
+      "step": 405
+    },
+    {
+      "epoch": 0.11032608695652174,
+      "grad_norm": 1.4534454345703125,
+      "learning_rate": 3e-05,
+      "loss": 3.0512,
+      "step": 406
+    },
+    {
+      "epoch": 0.11059782608695652,
+      "grad_norm": 2.4747161865234375,
+      "learning_rate": 3e-05,
+      "loss": 3.7008,
+      "step": 407
+    },
+    {
+      "epoch": 0.1108695652173913,
+      "grad_norm": 3.0130977630615234,
+      "learning_rate": 3e-05,
+      "loss": 3.9765,
+      "step": 408
+    },
+    {
+      "epoch": 0.11114130434782608,
+      "grad_norm": 2.256558656692505,
+      "learning_rate": 3e-05,
+      "loss": 3.2158,
+      "step": 409
+    },
+    {
+      "epoch": 0.11141304347826086,
+      "grad_norm": 4.283812522888184,
+      "learning_rate": 3e-05,
+      "loss": 4.4811,
+      "step": 410
+    },
+    {
+      "epoch": 0.11168478260869565,
+      "grad_norm": 2.566425323486328,
+      "learning_rate": 3e-05,
+      "loss": 3.9403,
+      "step": 411
+    },
+    {
+      "epoch": 0.11195652173913044,
+      "grad_norm": 2.4623820781707764,
+      "learning_rate": 3e-05,
+      "loss": 3.2361,
+      "step": 412
+    },
+    {
+      "epoch": 0.11222826086956522,
+      "grad_norm": 2.934722661972046,
+      "learning_rate": 3e-05,
+      "loss": 3.0554,
+      "step": 413
+    },
+    {
+      "epoch": 0.1125,
+      "grad_norm": 5.524335861206055,
+      "learning_rate": 3e-05,
+      "loss": 4.0103,
+      "step": 414
+    },
+    {
+      "epoch": 0.11277173913043478,
+      "grad_norm": 3.285830020904541,
+      "learning_rate": 3e-05,
+      "loss": 3.6652,
+      "step": 415
+    },
+    {
+      "epoch": 0.11304347826086956,
+      "grad_norm": 2.5553982257843018,
+      "learning_rate": 3e-05,
+      "loss": 3.8115,
+      "step": 416
+    },
+    {
+      "epoch": 0.11331521739130435,
+      "grad_norm": 1.8953039646148682,
+      "learning_rate": 3e-05,
+      "loss": 3.6769,
+      "step": 417
+    },
+    {
+      "epoch": 0.11358695652173913,
+      "grad_norm": 2.4014382362365723,
+      "learning_rate": 3e-05,
+      "loss": 3.6343,
+      "step": 418
+    },
+    {
+      "epoch": 0.11385869565217391,
+      "grad_norm": 3.2362847328186035,
+      "learning_rate": 3e-05,
+      "loss": 3.722,
+      "step": 419
+    },
+    {
+      "epoch": 0.11413043478260869,
+      "grad_norm": 2.525406837463379,
+      "learning_rate": 3e-05,
+      "loss": 3.2201,
+      "step": 420
+    },
+    {
+      "epoch": 0.11440217391304348,
+      "grad_norm": 2.492753028869629,
+      "learning_rate": 3e-05,
+      "loss": 3.3864,
+      "step": 421
+    },
+    {
+      "epoch": 0.11467391304347826,
+      "grad_norm": 1.6389284133911133,
+      "learning_rate": 3e-05,
+      "loss": 3.1505,
+      "step": 422
+    },
+    {
+      "epoch": 0.11494565217391305,
+      "grad_norm": 2.721895933151245,
+      "learning_rate": 3e-05,
+      "loss": 4.02,
+      "step": 423
+    },
+    {
+      "epoch": 0.11521739130434783,
+      "grad_norm": 2.1239542961120605,
+      "learning_rate": 3e-05,
+      "loss": 3.3395,
+      "step": 424
+    },
+    {
+      "epoch": 0.11548913043478261,
+      "grad_norm": 2.3277628421783447,
+      "learning_rate": 3e-05,
+      "loss": 4.3061,
+      "step": 425
+    },
+    {
+      "epoch": 0.11576086956521739,
+      "grad_norm": 1.7965201139450073,
+      "learning_rate": 3e-05,
+      "loss": 3.8778,
+      "step": 426
+    },
+    {
+      "epoch": 0.11603260869565217,
+      "grad_norm": 2.9034996032714844,
+      "learning_rate": 3e-05,
+      "loss": 3.8348,
+      "step": 427
+    },
+    {
+      "epoch": 0.11630434782608695,
+      "grad_norm": 2.0713295936584473,
+      "learning_rate": 3e-05,
+      "loss": 3.6901,
+      "step": 428
+    },
+    {
+      "epoch": 0.11657608695652173,
+      "grad_norm": 2.612128257751465,
+      "learning_rate": 3e-05,
+      "loss": 3.1308,
+      "step": 429
+    },
+    {
+      "epoch": 0.11684782608695653,
+      "grad_norm": 3.535771369934082,
+      "learning_rate": 3e-05,
+      "loss": 3.7629,
+      "step": 430
+    },
+    {
+      "epoch": 0.11711956521739131,
+      "grad_norm": 2.6418893337249756,
+      "learning_rate": 3e-05,
+      "loss": 3.351,
+      "step": 431
+    },
+    {
+      "epoch": 0.11739130434782609,
+      "grad_norm": 1.9122227430343628,
+      "learning_rate": 3e-05,
+      "loss": 2.9162,
+      "step": 432
+    },
+    {
+      "epoch": 0.11766304347826087,
+      "grad_norm": 3.12231707572937,
+      "learning_rate": 3e-05,
+      "loss": 3.6215,
+      "step": 433
+    },
+    {
+      "epoch": 0.11793478260869565,
+      "grad_norm": 2.5034008026123047,
+      "learning_rate": 3e-05,
+      "loss": 3.6675,
+      "step": 434
+    },
+    {
+      "epoch": 0.11820652173913043,
+      "grad_norm": 2.69254732131958,
+      "learning_rate": 3e-05,
+      "loss": 3.8847,
+      "step": 435
+    },
+    {
+      "epoch": 0.11847826086956521,
+      "grad_norm": 2.026611566543579,
+      "learning_rate": 3e-05,
+      "loss": 3.9927,
+      "step": 436
+    },
+    {
+      "epoch": 0.11875,
+      "grad_norm": 3.05465030670166,
+      "learning_rate": 3e-05,
+      "loss": 3.4363,
+      "step": 437
+    },
+    {
+      "epoch": 0.11902173913043479,
+      "grad_norm": 2.800114631652832,
+      "learning_rate": 3e-05,
+      "loss": 4.0161,
+      "step": 438
+    },
+    {
+      "epoch": 0.11929347826086957,
+      "grad_norm": 2.452852725982666,
+      "learning_rate": 3e-05,
+      "loss": 3.8825,
+      "step": 439
+    },
+    {
+      "epoch": 0.11956521739130435,
+      "grad_norm": 2.425046920776367,
+      "learning_rate": 3e-05,
+      "loss": 3.8342,
+      "step": 440
+    },
+    {
+      "epoch": 0.11983695652173913,
+      "grad_norm": 1.6898943185806274,
+      "learning_rate": 3e-05,
+      "loss": 3.4458,
+      "step": 441
+    },
+    {
+      "epoch": 0.12010869565217391,
+      "grad_norm": 2.265521764755249,
+      "learning_rate": 3e-05,
+      "loss": 3.498,
+      "step": 442
+    },
+    {
+      "epoch": 0.1203804347826087,
+      "grad_norm": 3.3182766437530518,
+      "learning_rate": 3e-05,
+      "loss": 4.0802,
+      "step": 443
+    },
+    {
+      "epoch": 0.12065217391304348,
+      "grad_norm": 3.441931962966919,
+      "learning_rate": 3e-05,
+      "loss": 3.7694,
+      "step": 444
+    },
+    {
+      "epoch": 0.12092391304347826,
+      "grad_norm": 2.1982293128967285,
+      "learning_rate": 3e-05,
+      "loss": 3.3313,
+      "step": 445
+    },
+    {
+      "epoch": 0.12119565217391304,
+      "grad_norm": 1.986126184463501,
+      "learning_rate": 3e-05,
+      "loss": 3.3758,
+      "step": 446
+    },
+    {
+      "epoch": 0.12146739130434783,
+      "grad_norm": 1.6694939136505127,
+      "learning_rate": 3e-05,
+      "loss": 3.1291,
+      "step": 447
+    },
+    {
+      "epoch": 0.12173913043478261,
+      "grad_norm": 3.418254852294922,
+      "learning_rate": 3e-05,
+      "loss": 3.9384,
+      "step": 448
+    },
+    {
+      "epoch": 0.1220108695652174,
+      "grad_norm": 2.4987125396728516,
+      "learning_rate": 3e-05,
+      "loss": 4.0415,
+      "step": 449
+    },
+    {
+      "epoch": 0.12228260869565218,
+      "grad_norm": 2.4183948040008545,
+      "learning_rate": 3e-05,
+      "loss": 3.4816,
+      "step": 450
+    },
+    {
+      "epoch": 0.12255434782608696,
+      "grad_norm": 2.753169059753418,
+      "learning_rate": 3e-05,
+      "loss": 3.3146,
+      "step": 451
+    },
+    {
+      "epoch": 0.12282608695652174,
+      "grad_norm": 2.2018933296203613,
+      "learning_rate": 3e-05,
+      "loss": 3.5041,
+      "step": 452
+    },
+    {
+      "epoch": 0.12309782608695652,
+      "grad_norm": 1.910427451133728,
+      "learning_rate": 3e-05,
+      "loss": 3.1185,
+      "step": 453
+    },
+    {
+      "epoch": 0.1233695652173913,
+      "grad_norm": 2.2100231647491455,
+      "learning_rate": 3e-05,
+      "loss": 3.1726,
+      "step": 454
+    },
+    {
+      "epoch": 0.12364130434782608,
+      "grad_norm": 2.6990854740142822,
+      "learning_rate": 3e-05,
+      "loss": 3.7998,
+      "step": 455
+    },
+    {
+      "epoch": 0.12391304347826088,
+      "grad_norm": 3.1403119564056396,
+      "learning_rate": 3e-05,
+      "loss": 3.3689,
+      "step": 456
+    },
+    {
+      "epoch": 0.12418478260869566,
+      "grad_norm": 1.9951990842819214,
+      "learning_rate": 3e-05,
+      "loss": 3.0753,
+      "step": 457
+    },
+    {
+      "epoch": 0.12445652173913044,
+      "grad_norm": 2.348395347595215,
+      "learning_rate": 3e-05,
+      "loss": 3.8548,
+      "step": 458
+    },
+    {
+      "epoch": 0.12472826086956522,
+      "grad_norm": 2.324969530105591,
+      "learning_rate": 3e-05,
+      "loss": 3.3703,
+      "step": 459
+    },
+    {
+      "epoch": 0.125,
+      "grad_norm": 2.897873640060425,
+      "learning_rate": 3e-05,
+      "loss": 3.4589,
+      "step": 460
+    },
+    {
+      "epoch": 0.1252717391304348,
+      "grad_norm": 2.81925892829895,
+      "learning_rate": 3e-05,
+      "loss": 3.4818,
+      "step": 461
+    },
+    {
+      "epoch": 0.12554347826086956,
+      "grad_norm": 4.320977687835693,
+      "learning_rate": 3e-05,
+      "loss": 4.3366,
+      "step": 462
+    },
+    {
+      "epoch": 0.12581521739130436,
+      "grad_norm": 3.2348084449768066,
+      "learning_rate": 3e-05,
+      "loss": 4.1263,
+      "step": 463
+    },
+    {
+      "epoch": 0.12608695652173912,
+      "grad_norm": 2.6263833045959473,
+      "learning_rate": 3e-05,
+      "loss": 2.9206,
+      "step": 464
+    },
+    {
+      "epoch": 0.12635869565217392,
+      "grad_norm": 4.796882629394531,
+      "learning_rate": 3e-05,
+      "loss": 3.8013,
+      "step": 465
+    },
+    {
+      "epoch": 0.1266304347826087,
+      "grad_norm": 5.0804619789123535,
+      "learning_rate": 3e-05,
+      "loss": 3.8332,
+      "step": 466
+    },
+    {
+      "epoch": 0.12690217391304348,
+      "grad_norm": 3.0131075382232666,
+      "learning_rate": 3e-05,
+      "loss": 3.5667,
+      "step": 467
+    },
+    {
+      "epoch": 0.12717391304347825,
+      "grad_norm": 1.6071690320968628,
+      "learning_rate": 3e-05,
+      "loss": 2.8863,
+      "step": 468
+    },
+    {
+      "epoch": 0.12744565217391304,
+      "grad_norm": 2.211411476135254,
+      "learning_rate": 3e-05,
+      "loss": 3.2423,
+      "step": 469
+    },
+    {
+      "epoch": 0.12771739130434784,
+      "grad_norm": 2.4000439643859863,
+      "learning_rate": 3e-05,
+      "loss": 3.3725,
+      "step": 470
+    },
+    {
+      "epoch": 0.1279891304347826,
+      "grad_norm": 2.421668529510498,
+      "learning_rate": 3e-05,
+      "loss": 3.6009,
+      "step": 471
+    },
+    {
+      "epoch": 0.1282608695652174,
+      "grad_norm": 1.906244158744812,
+      "learning_rate": 3e-05,
+      "loss": 3.4958,
+      "step": 472
+    },
+    {
+      "epoch": 0.12853260869565217,
+      "grad_norm": 1.7820132970809937,
+      "learning_rate": 3e-05,
+      "loss": 3.0321,
+      "step": 473
+    },
+    {
+      "epoch": 0.12880434782608696,
+      "grad_norm": 2.929561138153076,
+      "learning_rate": 3e-05,
+      "loss": 3.788,
+      "step": 474
+    },
+    {
+      "epoch": 0.12907608695652173,
+      "grad_norm": 2.740144968032837,
+      "learning_rate": 3e-05,
+      "loss": 3.6582,
+      "step": 475
+    },
+    {
+      "epoch": 0.12934782608695652,
+      "grad_norm": 2.153079032897949,
+      "learning_rate": 3e-05,
+      "loss": 3.6739,
+      "step": 476
+    },
+    {
+      "epoch": 0.1296195652173913,
+      "grad_norm": 1.366892695426941,
+      "learning_rate": 3e-05,
+      "loss": 3.5163,
+      "step": 477
+    },
+    {
+      "epoch": 0.1298913043478261,
+      "grad_norm": 2.3881099224090576,
+      "learning_rate": 3e-05,
+      "loss": 3.7265,
+      "step": 478
+    },
+    {
+      "epoch": 0.13016304347826088,
+      "grad_norm": 1.72128164768219,
+      "learning_rate": 3e-05,
+      "loss": 2.9679,
+      "step": 479
+    },
+    {
+      "epoch": 0.13043478260869565,
+      "grad_norm": 1.8024574518203735,
+      "learning_rate": 3e-05,
+      "loss": 3.6759,
+      "step": 480
+    },
+    {
+      "epoch": 0.13070652173913044,
+      "grad_norm": 2.3349437713623047,
+      "learning_rate": 3e-05,
+      "loss": 3.9231,
+      "step": 481
+    },
+    {
+      "epoch": 0.1309782608695652,
+      "grad_norm": 2.60579776763916,
+      "learning_rate": 3e-05,
+      "loss": 3.5484,
+      "step": 482
+    },
+    {
+      "epoch": 0.13125,
+      "grad_norm": 1.823648452758789,
+      "learning_rate": 3e-05,
+      "loss": 3.409,
+      "step": 483
+    },
+    {
+      "epoch": 0.13152173913043477,
+      "grad_norm": 1.939772367477417,
+      "learning_rate": 3e-05,
+      "loss": 3.7185,
+      "step": 484
+    },
+    {
+      "epoch": 0.13179347826086957,
+      "grad_norm": 1.7059251070022583,
+      "learning_rate": 3e-05,
+      "loss": 3.2223,
+      "step": 485
+    },
+    {
+      "epoch": 0.13206521739130433,
+      "grad_norm": 2.1309432983398438,
+      "learning_rate": 3e-05,
+      "loss": 3.5143,
+      "step": 486
+    },
+    {
+      "epoch": 0.13233695652173913,
+      "grad_norm": 2.5169553756713867,
+      "learning_rate": 3e-05,
+      "loss": 3.9375,
+      "step": 487
+    },
+    {
+      "epoch": 0.13260869565217392,
+      "grad_norm": 1.6783947944641113,
+      "learning_rate": 3e-05,
+      "loss": 3.0069,
+      "step": 488
+    },
+    {
+      "epoch": 0.1328804347826087,
+      "grad_norm": 2.5426528453826904,
+      "learning_rate": 3e-05,
+      "loss": 3.4032,
+      "step": 489
+    },
+    {
+      "epoch": 0.1331521739130435,
+      "grad_norm": 2.8521511554718018,
+      "learning_rate": 3e-05,
+      "loss": 3.1236,
+      "step": 490
+    },
+    {
+      "epoch": 0.13342391304347825,
+      "grad_norm": 2.7477147579193115,
+      "learning_rate": 3e-05,
+      "loss": 3.0802,
+      "step": 491
+    },
+    {
+      "epoch": 0.13369565217391305,
+      "grad_norm": 2.785457134246826,
+      "learning_rate": 3e-05,
+      "loss": 3.8152,
+      "step": 492
+    },
+    {
+      "epoch": 0.13396739130434782,
+      "grad_norm": 3.6600492000579834,
+      "learning_rate": 3e-05,
+      "loss": 3.6564,
+      "step": 493
+    },
+    {
+      "epoch": 0.1342391304347826,
+      "grad_norm": 4.777431488037109,
+      "learning_rate": 3e-05,
+      "loss": 3.7449,
+      "step": 494
+    },
+    {
+      "epoch": 0.13451086956521738,
+      "grad_norm": 4.060612201690674,
+      "learning_rate": 3e-05,
+      "loss": 4.1282,
+      "step": 495
+    },
+    {
+      "epoch": 0.13478260869565217,
+      "grad_norm": 4.2701592445373535,
+      "learning_rate": 3e-05,
+      "loss": 4.6185,
+      "step": 496
+    },
+    {
+      "epoch": 0.13505434782608697,
+      "grad_norm": 3.9511125087738037,
+      "learning_rate": 3e-05,
+      "loss": 3.2463,
+      "step": 497
+    },
+    {
+      "epoch": 0.13532608695652174,
+      "grad_norm": 3.8475091457366943,
+      "learning_rate": 3e-05,
+      "loss": 3.0689,
+      "step": 498
+    },
+    {
+      "epoch": 0.13559782608695653,
+      "grad_norm": 5.655093669891357,
+      "learning_rate": 3e-05,
+      "loss": 4.024,
+      "step": 499
+    },
+    {
+      "epoch": 0.1358695652173913,
+      "grad_norm": 3.6835744380950928,
+      "learning_rate": 3e-05,
+      "loss": 3.1939,
+      "step": 500
+    },
+    {
+      "epoch": 0.1361413043478261,
+      "grad_norm": 3.2142796516418457,
+      "learning_rate": 3e-05,
+      "loss": 4.0483,
+      "step": 501
+    },
+    {
+      "epoch": 0.13641304347826086,
+      "grad_norm": 3.1059064865112305,
+      "learning_rate": 3e-05,
+      "loss": 3.8777,
+      "step": 502
+    },
+    {
+      "epoch": 0.13668478260869565,
+      "grad_norm": 2.3298041820526123,
+      "learning_rate": 3e-05,
+      "loss": 3.2728,
+      "step": 503
+    },
+    {
+      "epoch": 0.13695652173913042,
+      "grad_norm": 3.318361759185791,
+      "learning_rate": 3e-05,
+      "loss": 4.0421,
+      "step": 504
+    },
+    {
+      "epoch": 0.13722826086956522,
+      "grad_norm": 2.8279385566711426,
+      "learning_rate": 3e-05,
+      "loss": 3.49,
+      "step": 505
+    },
+    {
+      "epoch": 0.1375,
+      "grad_norm": 2.233304023742676,
+      "learning_rate": 3e-05,
+      "loss": 3.3737,
+      "step": 506
+    },
+    {
+      "epoch": 0.13777173913043478,
+      "grad_norm": 2.943725824356079,
+      "learning_rate": 3e-05,
+      "loss": 3.5774,
+      "step": 507
+    },
+    {
+      "epoch": 0.13804347826086957,
+      "grad_norm": 2.764371871948242,
+      "learning_rate": 3e-05,
+      "loss": 3.7041,
+      "step": 508
+    },
+    {
+      "epoch": 0.13831521739130434,
+      "grad_norm": 2.647728443145752,
+      "learning_rate": 3e-05,
+      "loss": 3.0545,
+      "step": 509
+    },
+    {
+      "epoch": 0.13858695652173914,
+      "grad_norm": 3.1184072494506836,
+      "learning_rate": 3e-05,
+      "loss": 3.2394,
+      "step": 510
+    },
+    {
+      "epoch": 0.1388586956521739,
+      "grad_norm": 1.9611910581588745,
+      "learning_rate": 3e-05,
+      "loss": 3.4171,
+      "step": 511
+    },
+    {
+      "epoch": 0.1391304347826087,
+      "grad_norm": 1.5122188329696655,
+      "learning_rate": 3e-05,
+      "loss": 3.1593,
+      "step": 512
+    },
+    {
+      "epoch": 0.13940217391304346,
+      "grad_norm": 1.9166432619094849,
+      "learning_rate": 3e-05,
+      "loss": 3.21,
+      "step": 513
+    },
+    {
+      "epoch": 0.13967391304347826,
+      "grad_norm": 2.1466691493988037,
+      "learning_rate": 3e-05,
+      "loss": 3.4333,
+      "step": 514
+    },
+    {
+      "epoch": 0.13994565217391305,
+      "grad_norm": 2.217186689376831,
+      "learning_rate": 3e-05,
+      "loss": 3.3885,
+      "step": 515
+    },
+    {
+      "epoch": 0.14021739130434782,
+      "grad_norm": 2.03560733795166,
+      "learning_rate": 3e-05,
+      "loss": 3.9298,
+      "step": 516
+    },
+    {
+      "epoch": 0.14048913043478262,
+      "grad_norm": 2.214279890060425,
+      "learning_rate": 3e-05,
+      "loss": 3.2791,
+      "step": 517
+    },
+    {
+      "epoch": 0.14076086956521738,
+      "grad_norm": 3.428405523300171,
+      "learning_rate": 3e-05,
+      "loss": 3.668,
+      "step": 518
+    },
+    {
+      "epoch": 0.14103260869565218,
+      "grad_norm": 3.077150344848633,
+      "learning_rate": 3e-05,
+      "loss": 3.7187,
+      "step": 519
+    },
+    {
+      "epoch": 0.14130434782608695,
+      "grad_norm": 1.3983957767486572,
+      "learning_rate": 3e-05,
+      "loss": 3.3129,
+      "step": 520
+    },
+    {
+      "epoch": 0.14157608695652174,
+      "grad_norm": 2.0241615772247314,
+      "learning_rate": 3e-05,
+      "loss": 3.6655,
+      "step": 521
+    },
+    {
+      "epoch": 0.14184782608695654,
+      "grad_norm": 2.20249080657959,
+      "learning_rate": 3e-05,
+      "loss": 3.2375,
+      "step": 522
+    },
+    {
+      "epoch": 0.1421195652173913,
+      "grad_norm": 2.2980117797851562,
+      "learning_rate": 3e-05,
+      "loss": 3.7075,
+      "step": 523
+    },
+    {
+      "epoch": 0.1423913043478261,
+      "grad_norm": 1.2177653312683105,
+      "learning_rate": 3e-05,
+      "loss": 3.2919,
+      "step": 524
+    },
+    {
+      "epoch": 0.14266304347826086,
+      "grad_norm": 1.0807304382324219,
+      "learning_rate": 3e-05,
+      "loss": 3.2761,
+      "step": 525
+    },
+    {
+      "epoch": 0.14293478260869566,
+      "grad_norm": 1.4918198585510254,
+      "learning_rate": 3e-05,
+      "loss": 3.6412,
+      "step": 526
+    },
+    {
+      "epoch": 0.14320652173913043,
+      "grad_norm": 1.3037655353546143,
+      "learning_rate": 3e-05,
+      "loss": 3.4596,
+      "step": 527
+    },
+    {
+      "epoch": 0.14347826086956522,
+      "grad_norm": 1.1929515600204468,
+      "learning_rate": 3e-05,
+      "loss": 3.4554,
+      "step": 528
+    },
+    {
+      "epoch": 0.14375,
+      "grad_norm": 1.7983393669128418,
+      "learning_rate": 3e-05,
+      "loss": 3.0941,
+      "step": 529
+    },
+    {
+      "epoch": 0.14402173913043478,
+      "grad_norm": 1.4893770217895508,
+      "learning_rate": 3e-05,
+      "loss": 3.688,
+      "step": 530
+    },
+    {
+      "epoch": 0.14429347826086958,
+      "grad_norm": 1.4558385610580444,
+      "learning_rate": 3e-05,
+      "loss": 3.6866,
+      "step": 531
+    },
+    {
+      "epoch": 0.14456521739130435,
+      "grad_norm": 1.4489972591400146,
+      "learning_rate": 3e-05,
+      "loss": 3.1238,
+      "step": 532
+    },
+    {
+      "epoch": 0.14483695652173914,
+      "grad_norm": 1.603597640991211,
+      "learning_rate": 3e-05,
+      "loss": 2.9605,
+      "step": 533
+    },
+    {
+      "epoch": 0.1451086956521739,
+      "grad_norm": 1.1347497701644897,
+      "learning_rate": 3e-05,
+      "loss": 3.2066,
+      "step": 534
+    },
+    {
+      "epoch": 0.1453804347826087,
+      "grad_norm": 1.9727925062179565,
+      "learning_rate": 3e-05,
+      "loss": 3.6686,
+      "step": 535
+    },
+    {
+      "epoch": 0.14565217391304347,
+      "grad_norm": 2.5670254230499268,
+      "learning_rate": 3e-05,
+      "loss": 3.3345,
+      "step": 536
+    },
+    {
+      "epoch": 0.14592391304347826,
+      "grad_norm": 2.017517328262329,
+      "learning_rate": 3e-05,
+      "loss": 3.7092,
+      "step": 537
+    },
+    {
+      "epoch": 0.14619565217391303,
+      "grad_norm": 2.1940996646881104,
+      "learning_rate": 3e-05,
+      "loss": 3.807,
+      "step": 538
+    },
+    {
+      "epoch": 0.14646739130434783,
+      "grad_norm": 1.7565933465957642,
+      "learning_rate": 3e-05,
+      "loss": 3.6645,
+      "step": 539
+    },
+    {
+      "epoch": 0.14673913043478262,
+      "grad_norm": 2.248276948928833,
+      "learning_rate": 3e-05,
+      "loss": 3.979,
+      "step": 540
+    },
+    {
+      "epoch": 0.1470108695652174,
+      "grad_norm": 2.3259811401367188,
+      "learning_rate": 3e-05,
+      "loss": 3.3435,
+      "step": 541
+    },
+    {
+      "epoch": 0.14728260869565218,
+      "grad_norm": 2.7300784587860107,
+      "learning_rate": 3e-05,
+      "loss": 3.7696,
+      "step": 542
+    },
+    {
+      "epoch": 0.14755434782608695,
+      "grad_norm": 1.406214952468872,
+      "learning_rate": 3e-05,
+      "loss": 2.6787,
+      "step": 543
+    },
+    {
+      "epoch": 0.14782608695652175,
+      "grad_norm": 2.6304099559783936,
+      "learning_rate": 3e-05,
+      "loss": 3.7273,
+      "step": 544
+    },
+    {
+      "epoch": 0.1480978260869565,
+      "grad_norm": 4.7364912033081055,
+      "learning_rate": 3e-05,
+      "loss": 4.4759,
+      "step": 545
+    },
+    {
+      "epoch": 0.1483695652173913,
+      "grad_norm": 5.363766193389893,
+      "learning_rate": 3e-05,
+      "loss": 4.1787,
+      "step": 546
+    },
+    {
+      "epoch": 0.14864130434782608,
+      "grad_norm": 2.660940408706665,
+      "learning_rate": 3e-05,
+      "loss": 3.2534,
+      "step": 547
+    },
+    {
+      "epoch": 0.14891304347826087,
+      "grad_norm": 3.186602830886841,
+      "learning_rate": 3e-05,
+      "loss": 3.5762,
+      "step": 548
+    },
+    {
+      "epoch": 0.14918478260869567,
+      "grad_norm": 3.962216377258301,
+      "learning_rate": 3e-05,
+      "loss": 3.5002,
+      "step": 549
+    },
+    {
+      "epoch": 0.14945652173913043,
+      "grad_norm": 5.419135093688965,
+      "learning_rate": 3e-05,
+      "loss": 3.6202,
+      "step": 550
+    },
+    {
+      "epoch": 0.14972826086956523,
+      "grad_norm": 4.870014667510986,
+      "learning_rate": 3e-05,
+      "loss": 3.5467,
+      "step": 551
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.175389289855957,
+      "learning_rate": 3e-05,
+      "loss": 3.9385,
+      "step": 552
+    },
+    {
+      "epoch": 0.1502717391304348,
+      "grad_norm": 2.783637046813965,
+      "learning_rate": 3e-05,
+      "loss": 3.8554,
+      "step": 553
+    },
+    {
+      "epoch": 0.15054347826086956,
+      "grad_norm": 2.6941978931427,
+      "learning_rate": 3e-05,
+      "loss": 3.211,
+      "step": 554
+    },
+    {
+      "epoch": 0.15081521739130435,
+      "grad_norm": 3.185209274291992,
+      "learning_rate": 3e-05,
+      "loss": 3.3476,
+      "step": 555
+    },
+    {
+      "epoch": 0.15108695652173912,
+      "grad_norm": 2.922362804412842,
+      "learning_rate": 3e-05,
+      "loss": 3.0941,
+      "step": 556
+    },
+    {
+      "epoch": 0.1513586956521739,
+      "grad_norm": 2.7833430767059326,
+      "learning_rate": 3e-05,
+      "loss": 3.3851,
+      "step": 557
+    },
+    {
+      "epoch": 0.1516304347826087,
+      "grad_norm": 1.4858049154281616,
+      "learning_rate": 3e-05,
+      "loss": 2.7326,
+      "step": 558
+    },
+    {
+      "epoch": 0.15190217391304348,
+      "grad_norm": 2.9602174758911133,
+      "learning_rate": 3e-05,
+      "loss": 3.2037,
+      "step": 559
+    },
+    {
+      "epoch": 0.15217391304347827,
+      "grad_norm": 3.0462303161621094,
+      "learning_rate": 3e-05,
+      "loss": 3.2449,
+      "step": 560
+    },
+    {
+      "epoch": 0.15244565217391304,
+      "grad_norm": 3.038954019546509,
+      "learning_rate": 3e-05,
+      "loss": 3.3796,
+      "step": 561
+    },
+    {
+      "epoch": 0.15271739130434783,
+      "grad_norm": 1.572611689567566,
+      "learning_rate": 3e-05,
+      "loss": 3.1521,
+      "step": 562
+    },
+    {
+      "epoch": 0.1529891304347826,
+      "grad_norm": 2.4640631675720215,
+      "learning_rate": 3e-05,
+      "loss": 3.9771,
+      "step": 563
+    },
+    {
+      "epoch": 0.1532608695652174,
+      "grad_norm": 1.757712483406067,
+      "learning_rate": 3e-05,
+      "loss": 2.9893,
+      "step": 564
+    },
+    {
+      "epoch": 0.15353260869565216,
+      "grad_norm": 2.606873035430908,
+      "learning_rate": 3e-05,
+      "loss": 3.702,
+      "step": 565
+    },
+    {
+      "epoch": 0.15380434782608696,
+      "grad_norm": 1.677304983139038,
+      "learning_rate": 3e-05,
+      "loss": 3.2607,
+      "step": 566
+    },
+    {
+      "epoch": 0.15407608695652175,
+      "grad_norm": 2.4729526042938232,
+      "learning_rate": 3e-05,
+      "loss": 3.9865,
+      "step": 567
+    },
+    {
+      "epoch": 0.15434782608695652,
+      "grad_norm": 2.162191867828369,
+      "learning_rate": 3e-05,
+      "loss": 3.6349,
+      "step": 568
+    },
+    {
+      "epoch": 0.1546195652173913,
+      "grad_norm": 3.027984380722046,
+      "learning_rate": 3e-05,
+      "loss": 3.8996,
+      "step": 569
+    },
+    {
+      "epoch": 0.15489130434782608,
+      "grad_norm": 1.9820539951324463,
+      "learning_rate": 3e-05,
+      "loss": 3.3207,
+      "step": 570
+    },
+    {
+      "epoch": 0.15516304347826088,
+      "grad_norm": 1.7880254983901978,
+      "learning_rate": 3e-05,
+      "loss": 3.6082,
+      "step": 571
+    },
+    {
+      "epoch": 0.15543478260869564,
+      "grad_norm": 1.9113103151321411,
+      "learning_rate": 3e-05,
+      "loss": 3.2639,
+      "step": 572
+    },
+    {
+      "epoch": 0.15570652173913044,
+      "grad_norm": 1.7418012619018555,
+      "learning_rate": 3e-05,
+      "loss": 3.4052,
+      "step": 573
+    },
+    {
+      "epoch": 0.1559782608695652,
+      "grad_norm": 1.6183768510818481,
+      "learning_rate": 3e-05,
+      "loss": 3.8061,
+      "step": 574
+    },
+    {
+      "epoch": 0.15625,
+      "grad_norm": 1.33986496925354,
+      "learning_rate": 3e-05,
+      "loss": 3.3393,
+      "step": 575
+    },
+    {
+      "epoch": 0.1565217391304348,
+      "grad_norm": 1.4486945867538452,
+      "learning_rate": 3e-05,
+      "loss": 3.7729,
+      "step": 576
+    },
+    {
+      "epoch": 0.15679347826086956,
+      "grad_norm": 1.7886120080947876,
+      "learning_rate": 3e-05,
+      "loss": 3.8346,
+      "step": 577
+    },
+    {
+      "epoch": 0.15706521739130436,
+      "grad_norm": 1.2352441549301147,
+      "learning_rate": 3e-05,
+      "loss": 3.2718,
+      "step": 578
+    },
+    {
+      "epoch": 0.15733695652173912,
+      "grad_norm": 1.8679944276809692,
+      "learning_rate": 3e-05,
+      "loss": 3.4247,
+      "step": 579
+    },
+    {
+      "epoch": 0.15760869565217392,
+      "grad_norm": 1.689501404762268,
+      "learning_rate": 3e-05,
+      "loss": 4.0971,
+      "step": 580
+    },
+    {
+      "epoch": 0.1578804347826087,
+      "grad_norm": 1.6779184341430664,
+      "learning_rate": 3e-05,
+      "loss": 3.8803,
+      "step": 581
+    },
+    {
+      "epoch": 0.15815217391304348,
+      "grad_norm": 2.0760598182678223,
+      "learning_rate": 3e-05,
+      "loss": 3.5429,
+      "step": 582
+    },
+    {
+      "epoch": 0.15842391304347825,
+      "grad_norm": 1.8308002948760986,
+      "learning_rate": 3e-05,
+      "loss": 3.4797,
+      "step": 583
+    },
+    {
+      "epoch": 0.15869565217391304,
+      "grad_norm": 2.4416396617889404,
+      "learning_rate": 3e-05,
+      "loss": 4.2491,
+      "step": 584
+    },
+    {
+      "epoch": 0.15896739130434784,
+      "grad_norm": 2.365509271621704,
+      "learning_rate": 3e-05,
+      "loss": 3.4862,
+      "step": 585
+    },
+    {
+      "epoch": 0.1592391304347826,
+      "grad_norm": 4.06919002532959,
+      "learning_rate": 3e-05,
+      "loss": 4.0302,
+      "step": 586
+    },
+    {
+      "epoch": 0.1595108695652174,
+      "grad_norm": 2.430508613586426,
+      "learning_rate": 3e-05,
+      "loss": 3.8575,
+      "step": 587
+    },
+    {
+      "epoch": 0.15978260869565217,
+      "grad_norm": 2.2944397926330566,
+      "learning_rate": 3e-05,
+      "loss": 4.3022,
+      "step": 588
+    },
+    {
+      "epoch": 0.16005434782608696,
+      "grad_norm": 4.096346855163574,
+      "learning_rate": 3e-05,
+      "loss": 4.0663,
+      "step": 589
+    },
+    {
+      "epoch": 0.16032608695652173,
+      "grad_norm": 3.8741347789764404,
+      "learning_rate": 3e-05,
+      "loss": 3.3398,
+      "step": 590
+    },
+    {
+      "epoch": 0.16059782608695652,
+      "grad_norm": 3.7039456367492676,
+      "learning_rate": 3e-05,
+      "loss": 3.1813,
+      "step": 591
+    },
+    {
+      "epoch": 0.1608695652173913,
+      "grad_norm": 2.129674196243286,
+      "learning_rate": 3e-05,
+      "loss": 3.0778,
+      "step": 592
+    },
+    {
+      "epoch": 0.1611413043478261,
+      "grad_norm": 2.404423952102661,
+      "learning_rate": 3e-05,
+      "loss": 3.7275,
+      "step": 593
+    },
+    {
+      "epoch": 0.16141304347826088,
+      "grad_norm": 2.341529607772827,
+      "learning_rate": 3e-05,
+      "loss": 2.9763,
+      "step": 594
+    },
+    {
+      "epoch": 0.16168478260869565,
+      "grad_norm": 4.927234649658203,
+      "learning_rate": 3e-05,
+      "loss": 4.1287,
+      "step": 595
+    },
+    {
+      "epoch": 0.16195652173913044,
+      "grad_norm": 4.529046535491943,
+      "learning_rate": 3e-05,
+      "loss": 3.7712,
+      "step": 596
+    },
+    {
+      "epoch": 0.1622282608695652,
+      "grad_norm": 4.280128479003906,
+      "learning_rate": 3e-05,
+      "loss": 3.7557,
+      "step": 597
+    },
+    {
+      "epoch": 0.1625,
+      "grad_norm": 2.845654249191284,
+      "learning_rate": 3e-05,
+      "loss": 3.4817,
+      "step": 598
+    },
+    {
+      "epoch": 0.16277173913043477,
+      "grad_norm": 2.5575075149536133,
+      "learning_rate": 3e-05,
+      "loss": 3.0366,
+      "step": 599
+    },
+    {
+      "epoch": 0.16304347826086957,
+      "grad_norm": 3.475111722946167,
+      "learning_rate": 3e-05,
+      "loss": 3.2081,
+      "step": 600
+    },
+    {
+      "epoch": 0.16331521739130433,
+      "grad_norm": 3.926103353500366,
+      "learning_rate": 3e-05,
+      "loss": 2.8455,
+      "step": 601
+    },
+    {
+      "epoch": 0.16358695652173913,
+      "grad_norm": 3.796124219894409,
+      "learning_rate": 3e-05,
+      "loss": 3.4857,
+      "step": 602
+    },
+    {
+      "epoch": 0.16385869565217392,
+      "grad_norm": 2.974625825881958,
+      "learning_rate": 3e-05,
+      "loss": 3.3609,
+      "step": 603
+    },
+    {
+      "epoch": 0.1641304347826087,
+      "grad_norm": 1.9188250303268433,
+      "learning_rate": 3e-05,
+      "loss": 3.0467,
+      "step": 604
+    },
+    {
+      "epoch": 0.1644021739130435,
+      "grad_norm": 2.922881841659546,
+      "learning_rate": 3e-05,
+      "loss": 3.969,
+      "step": 605
+    },
+    {
+      "epoch": 0.16467391304347825,
+      "grad_norm": 2.9243340492248535,
+      "learning_rate": 3e-05,
+      "loss": 3.418,
+      "step": 606
+    },
+    {
+      "epoch": 0.16494565217391305,
+      "grad_norm": 2.3607518672943115,
+      "learning_rate": 3e-05,
+      "loss": 3.3568,
+      "step": 607
+    },
+    {
+      "epoch": 0.16521739130434782,
+      "grad_norm": 1.4109339714050293,
+      "learning_rate": 3e-05,
+      "loss": 3.1942,
+      "step": 608
+    },
+    {
+      "epoch": 0.1654891304347826,
+      "grad_norm": 2.339900016784668,
+      "learning_rate": 3e-05,
+      "loss": 3.6621,
+      "step": 609
+    },
+    {
+      "epoch": 0.16576086956521738,
+      "grad_norm": 3.0392491817474365,
+      "learning_rate": 3e-05,
+      "loss": 3.6538,
+      "step": 610
+    },
+    {
+      "epoch": 0.16603260869565217,
+      "grad_norm": 2.5297353267669678,
+      "learning_rate": 3e-05,
+      "loss": 3.72,
+      "step": 611
+    },
+    {
+      "epoch": 0.16630434782608697,
+      "grad_norm": 2.1570205688476562,
+      "learning_rate": 3e-05,
+      "loss": 3.1765,
+      "step": 612
+    },
+    {
+      "epoch": 0.16657608695652174,
+      "grad_norm": 1.5125519037246704,
+      "learning_rate": 3e-05,
+      "loss": 3.5076,
+      "step": 613
+    },
+    {
+      "epoch": 0.16684782608695653,
+      "grad_norm": 1.5017513036727905,
+      "learning_rate": 3e-05,
+      "loss": 3.5962,
+      "step": 614
+    },
+    {
+      "epoch": 0.1671195652173913,
+      "grad_norm": 2.1280760765075684,
+      "learning_rate": 3e-05,
+      "loss": 3.4612,
+      "step": 615
+    },
+    {
+      "epoch": 0.1673913043478261,
+      "grad_norm": 1.6511037349700928,
+      "learning_rate": 3e-05,
+      "loss": 3.3281,
+      "step": 616
+    },
+    {
+      "epoch": 0.16766304347826086,
+      "grad_norm": 1.842786431312561,
+      "learning_rate": 3e-05,
+      "loss": 2.876,
+      "step": 617
+    },
+    {
+      "epoch": 0.16793478260869565,
+      "grad_norm": 2.008168935775757,
+      "learning_rate": 3e-05,
+      "loss": 3.7421,
+      "step": 618
+    },
+    {
+      "epoch": 0.16820652173913042,
+      "grad_norm": 2.965365409851074,
+      "learning_rate": 3e-05,
+      "loss": 3.9864,
+      "step": 619
+    },
+    {
+      "epoch": 0.16847826086956522,
+      "grad_norm": 3.619001626968384,
+      "learning_rate": 3e-05,
+      "loss": 4.1181,
+      "step": 620
+    },
+    {
+      "epoch": 0.16875,
+      "grad_norm": 3.3966426849365234,
+      "learning_rate": 3e-05,
+      "loss": 3.2355,
+      "step": 621
+    },
+    {
+      "epoch": 0.16902173913043478,
+      "grad_norm": 2.4561333656311035,
+      "learning_rate": 3e-05,
+      "loss": 3.0077,
+      "step": 622
+    },
+    {
+      "epoch": 0.16929347826086957,
+      "grad_norm": 3.7762365341186523,
+      "learning_rate": 3e-05,
+      "loss": 3.8245,
+      "step": 623
+    },
+    {
+      "epoch": 0.16956521739130434,
+      "grad_norm": 4.286952495574951,
+      "learning_rate": 3e-05,
+      "loss": 3.4895,
+      "step": 624
+    },
+    {
+      "epoch": 0.16983695652173914,
+      "grad_norm": 3.7213828563690186,
+      "learning_rate": 3e-05,
+      "loss": 2.9799,
+      "step": 625
+    },
+    {
+      "epoch": 0.1701086956521739,
+      "grad_norm": 3.109027862548828,
+      "learning_rate": 3e-05,
+      "loss": 3.8767,
+      "step": 626
+    },
+    {
+      "epoch": 0.1703804347826087,
+      "grad_norm": 2.4418489933013916,
+      "learning_rate": 3e-05,
+      "loss": 3.7539,
+      "step": 627
+    },
+    {
+      "epoch": 0.17065217391304346,
+      "grad_norm": 2.3365840911865234,
+      "learning_rate": 3e-05,
+      "loss": 3.0382,
+      "step": 628
+    },
+    {
+      "epoch": 0.17092391304347826,
+      "grad_norm": 2.649116039276123,
+      "learning_rate": 3e-05,
+      "loss": 3.0011,
+      "step": 629
+    },
+    {
+      "epoch": 0.17119565217391305,
+      "grad_norm": 4.0966105461120605,
+      "learning_rate": 3e-05,
+      "loss": 3.5889,
+      "step": 630
+    },
+    {
+      "epoch": 0.17146739130434782,
+      "grad_norm": 4.70590877532959,
+      "learning_rate": 3e-05,
+      "loss": 3.4292,
+      "step": 631
+    },
+    {
+      "epoch": 0.17173913043478262,
+      "grad_norm": 4.246743202209473,
+      "learning_rate": 3e-05,
+      "loss": 3.3927,
+      "step": 632
+    },
+    {
+      "epoch": 0.17201086956521738,
+      "grad_norm": 3.4085919857025146,
+      "learning_rate": 3e-05,
+      "loss": 3.5786,
+      "step": 633
+    },
+    {
+      "epoch": 0.17228260869565218,
+      "grad_norm": 3.164703607559204,
+      "learning_rate": 3e-05,
+      "loss": 3.4236,
+      "step": 634
+    },
+    {
+      "epoch": 0.17255434782608695,
+      "grad_norm": 4.914327144622803,
+      "learning_rate": 3e-05,
+      "loss": 4.2673,
+      "step": 635
+    },
+    {
+      "epoch": 0.17282608695652174,
+      "grad_norm": 4.016328811645508,
+      "learning_rate": 3e-05,
+      "loss": 3.5927,
+      "step": 636
+    },
+    {
+      "epoch": 0.17309782608695654,
+      "grad_norm": 3.3188321590423584,
+      "learning_rate": 3e-05,
+      "loss": 2.9687,
+      "step": 637
+    },
+    {
+      "epoch": 0.1733695652173913,
+      "grad_norm": 3.4671499729156494,
+      "learning_rate": 3e-05,
+      "loss": 3.3144,
+      "step": 638
+    },
+    {
+      "epoch": 0.1736413043478261,
+      "grad_norm": 2.273728609085083,
+      "learning_rate": 3e-05,
+      "loss": 3.2453,
+      "step": 639
+    },
+    {
+      "epoch": 0.17391304347826086,
+      "grad_norm": 2.1210081577301025,
+      "learning_rate": 3e-05,
+      "loss": 3.0778,
+      "step": 640
+    },
+    {
+      "epoch": 0.17418478260869566,
+      "grad_norm": 3.6642115116119385,
+      "learning_rate": 3e-05,
+      "loss": 3.7503,
+      "step": 641
+    },
+    {
+      "epoch": 0.17445652173913043,
+      "grad_norm": 3.753002166748047,
+      "learning_rate": 3e-05,
+      "loss": 3.9504,
+      "step": 642
+    },
+    {
+      "epoch": 0.17472826086956522,
+      "grad_norm": 2.683023691177368,
+      "learning_rate": 3e-05,
+      "loss": 3.5103,
+      "step": 643
+    },
+    {
+      "epoch": 0.175,
+      "grad_norm": 2.7764129638671875,
+      "learning_rate": 3e-05,
+      "loss": 3.8473,
+      "step": 644
+    },
+    {
+      "epoch": 0.17527173913043478,
+      "grad_norm": 2.340895652770996,
+      "learning_rate": 3e-05,
+      "loss": 3.3381,
+      "step": 645
+    },
+    {
+      "epoch": 0.17554347826086958,
+      "grad_norm": 3.224985122680664,
+      "learning_rate": 3e-05,
+      "loss": 3.7218,
+      "step": 646
+    },
+    {
+      "epoch": 0.17581521739130435,
+      "grad_norm": 4.798543930053711,
+      "learning_rate": 3e-05,
+      "loss": 4.2621,
+      "step": 647
+    },
+    {
+      "epoch": 0.17608695652173914,
+      "grad_norm": 3.2494759559631348,
+      "learning_rate": 3e-05,
+      "loss": 3.4175,
+      "step": 648
+    },
+    {
+      "epoch": 0.1763586956521739,
+      "grad_norm": 1.9543368816375732,
+      "learning_rate": 3e-05,
+      "loss": 2.9133,
+      "step": 649
+    },
+    {
+      "epoch": 0.1766304347826087,
+      "grad_norm": 1.7534743547439575,
+      "learning_rate": 3e-05,
+      "loss": 3.2838,
+      "step": 650
+    },
+    {
+      "epoch": 0.17690217391304347,
+      "grad_norm": 1.8143278360366821,
+      "learning_rate": 3e-05,
+      "loss": 3.015,
+      "step": 651
+    },
+    {
+      "epoch": 0.17717391304347826,
+      "grad_norm": 3.4449610710144043,
+      "learning_rate": 3e-05,
+      "loss": 3.7868,
+      "step": 652
+    },
+    {
+      "epoch": 0.17744565217391303,
+      "grad_norm": 2.250290632247925,
+      "learning_rate": 3e-05,
+      "loss": 3.6572,
+      "step": 653
+    },
+    {
+      "epoch": 0.17771739130434783,
+      "grad_norm": 2.3306338787078857,
+      "learning_rate": 3e-05,
+      "loss": 3.6463,
+      "step": 654
+    },
+    {
+      "epoch": 0.17798913043478262,
+      "grad_norm": 1.6108341217041016,
+      "learning_rate": 3e-05,
+      "loss": 3.3977,
+      "step": 655
+    },
+    {
+      "epoch": 0.1782608695652174,
+      "grad_norm": 1.7349133491516113,
+      "learning_rate": 3e-05,
+      "loss": 3.5826,
+      "step": 656
+    },
+    {
+      "epoch": 0.17853260869565218,
+      "grad_norm": 2.5024263858795166,
+      "learning_rate": 3e-05,
+      "loss": 4.2271,
+      "step": 657
+    },
+    {
+      "epoch": 0.17880434782608695,
+      "grad_norm": 2.2973721027374268,
+      "learning_rate": 3e-05,
+      "loss": 3.8238,
+      "step": 658
+    },
+    {
+      "epoch": 0.17907608695652175,
+      "grad_norm": 2.167056083679199,
+      "learning_rate": 3e-05,
+      "loss": 3.5719,
+      "step": 659
+    },
+    {
+      "epoch": 0.1793478260869565,
+      "grad_norm": 1.8604989051818848,
+      "learning_rate": 3e-05,
+      "loss": 3.9091,
+      "step": 660
+    },
+    {
+      "epoch": 0.1796195652173913,
+      "grad_norm": 1.7601895332336426,
+      "learning_rate": 3e-05,
+      "loss": 3.8582,
+      "step": 661
+    },
+    {
+      "epoch": 0.17989130434782608,
+      "grad_norm": 1.5462661981582642,
+      "learning_rate": 3e-05,
+      "loss": 2.9604,
+      "step": 662
+    },
+    {
+      "epoch": 0.18016304347826087,
+      "grad_norm": 1.9906418323516846,
+      "learning_rate": 3e-05,
+      "loss": 3.8877,
+      "step": 663
+    },
+    {
+      "epoch": 0.18043478260869567,
+      "grad_norm": 2.1721136569976807,
+      "learning_rate": 3e-05,
+      "loss": 4.241,
+      "step": 664
+    },
+    {
+      "epoch": 0.18070652173913043,
+      "grad_norm": 1.7870231866836548,
+      "learning_rate": 3e-05,
+      "loss": 3.0484,
+      "step": 665
+    },
+    {
+      "epoch": 0.18097826086956523,
+      "grad_norm": 1.543925166130066,
+      "learning_rate": 3e-05,
+      "loss": 3.1115,
+      "step": 666
+    },
+    {
+      "epoch": 0.18125,
+      "grad_norm": 1.9159563779830933,
+      "learning_rate": 3e-05,
+      "loss": 3.3634,
+      "step": 667
+    },
+    {
+      "epoch": 0.1815217391304348,
+      "grad_norm": 2.8006904125213623,
+      "learning_rate": 3e-05,
+      "loss": 3.7725,
+      "step": 668
+    },
+    {
+      "epoch": 0.18179347826086956,
+      "grad_norm": 2.8094773292541504,
+      "learning_rate": 3e-05,
+      "loss": 3.5421,
+      "step": 669
+    },
+    {
+      "epoch": 0.18206521739130435,
+      "grad_norm": 2.0162885189056396,
+      "learning_rate": 3e-05,
+      "loss": 3.4805,
+      "step": 670
+    },
+    {
+      "epoch": 0.18233695652173912,
+      "grad_norm": 2.175633430480957,
+      "learning_rate": 3e-05,
+      "loss": 3.3937,
+      "step": 671
+    },
+    {
+      "epoch": 0.1826086956521739,
+      "grad_norm": 2.4267466068267822,
+      "learning_rate": 3e-05,
+      "loss": 3.1429,
+      "step": 672
+    },
+    {
+      "epoch": 0.1828804347826087,
+      "grad_norm": 4.546562194824219,
+      "learning_rate": 3e-05,
+      "loss": 4.1168,
+      "step": 673
+    },
+    {
+      "epoch": 0.18315217391304348,
+      "grad_norm": 3.3911025524139404,
+      "learning_rate": 3e-05,
+      "loss": 3.7966,
+      "step": 674
+    },
+    {
+      "epoch": 0.18342391304347827,
+      "grad_norm": 1.7894442081451416,
+      "learning_rate": 3e-05,
+      "loss": 3.8492,
+      "step": 675
+    },
+    {
+      "epoch": 0.18369565217391304,
+      "grad_norm": 2.7028391361236572,
+      "learning_rate": 3e-05,
+      "loss": 3.5555,
+      "step": 676
+    },
+    {
+      "epoch": 0.18396739130434783,
+      "grad_norm": 3.052639961242676,
+      "learning_rate": 3e-05,
+      "loss": 3.9254,
+      "step": 677
+    },
+    {
+      "epoch": 0.1842391304347826,
+      "grad_norm": 2.4930381774902344,
+      "learning_rate": 3e-05,
+      "loss": 3.4053,
+      "step": 678
+    },
+    {
+      "epoch": 0.1845108695652174,
+      "grad_norm": 1.9123059511184692,
+      "learning_rate": 3e-05,
+      "loss": 3.5101,
+      "step": 679
+    },
+    {
+      "epoch": 0.18478260869565216,
+      "grad_norm": 1.5995222330093384,
+      "learning_rate": 3e-05,
+      "loss": 3.1762,
+      "step": 680
+    },
+    {
+      "epoch": 0.18505434782608696,
+      "grad_norm": 4.272251129150391,
+      "learning_rate": 3e-05,
+      "loss": 4.397,
+      "step": 681
+    },
+    {
+      "epoch": 0.18532608695652175,
+      "grad_norm": 3.0316965579986572,
+      "learning_rate": 3e-05,
+      "loss": 3.8721,
+      "step": 682
+    },
+    {
+      "epoch": 0.18559782608695652,
+      "grad_norm": 2.0125246047973633,
+      "learning_rate": 3e-05,
+      "loss": 3.3782,
+      "step": 683
+    },
+    {
+      "epoch": 0.1858695652173913,
+      "grad_norm": 1.8079962730407715,
+      "learning_rate": 3e-05,
+      "loss": 3.4851,
+      "step": 684
+    },
+    {
+      "epoch": 0.18614130434782608,
+      "grad_norm": 1.8277610540390015,
+      "learning_rate": 3e-05,
+      "loss": 2.8756,
+      "step": 685
+    },
+    {
+      "epoch": 0.18641304347826088,
+      "grad_norm": 2.4962170124053955,
+      "learning_rate": 3e-05,
+      "loss": 4.2237,
+      "step": 686
+    },
+    {
+      "epoch": 0.18668478260869564,
+      "grad_norm": 1.3268901109695435,
+      "learning_rate": 3e-05,
+      "loss": 3.1177,
+      "step": 687
+    },
+    {
+      "epoch": 0.18695652173913044,
+      "grad_norm": 2.3978965282440186,
+      "learning_rate": 3e-05,
+      "loss": 3.583,
+      "step": 688
+    },
+    {
+      "epoch": 0.1872282608695652,
+      "grad_norm": 1.2693482637405396,
+      "learning_rate": 3e-05,
+      "loss": 3.3973,
+      "step": 689
+    },
+    {
+      "epoch": 0.1875,
+      "grad_norm": 2.0299911499023438,
+      "learning_rate": 3e-05,
+      "loss": 3.6627,
+      "step": 690
+    },
+    {
+      "epoch": 0.1877717391304348,
+      "grad_norm": 1.2860682010650635,
+      "learning_rate": 3e-05,
+      "loss": 3.2345,
+      "step": 691
+    },
+    {
+      "epoch": 0.18804347826086956,
+      "grad_norm": 2.264256477355957,
+      "learning_rate": 3e-05,
+      "loss": 4.2411,
+      "step": 692
+    },
+    {
+      "epoch": 0.18831521739130436,
+      "grad_norm": 1.5267597436904907,
+      "learning_rate": 3e-05,
+      "loss": 3.1942,
+      "step": 693
+    },
+    {
+      "epoch": 0.18858695652173912,
+      "grad_norm": 1.2684136629104614,
+      "learning_rate": 3e-05,
+      "loss": 3.192,
+      "step": 694
+    },
+    {
+      "epoch": 0.18885869565217392,
+      "grad_norm": 1.6705724000930786,
+      "learning_rate": 3e-05,
+      "loss": 3.6082,
+      "step": 695
+    },
+    {
+      "epoch": 0.1891304347826087,
+      "grad_norm": 2.0569725036621094,
+      "learning_rate": 3e-05,
+      "loss": 3.9019,
+      "step": 696
+    },
+    {
+      "epoch": 0.18940217391304348,
+      "grad_norm": 2.03778076171875,
+      "learning_rate": 3e-05,
+      "loss": 4.0118,
+      "step": 697
+    },
+    {
+      "epoch": 0.18967391304347825,
+      "grad_norm": 1.7906421422958374,
+      "learning_rate": 3e-05,
+      "loss": 3.7659,
+      "step": 698
+    },
+    {
+      "epoch": 0.18994565217391304,
+      "grad_norm": 1.6202226877212524,
+      "learning_rate": 3e-05,
+      "loss": 3.3228,
+      "step": 699
+    },
+    {
+      "epoch": 0.19021739130434784,
+      "grad_norm": 1.568548560142517,
+      "learning_rate": 3e-05,
+      "loss": 3.3992,
+      "step": 700
+    },
+    {
+      "epoch": 0.1904891304347826,
+      "grad_norm": 1.3486888408660889,
+      "learning_rate": 3e-05,
+      "loss": 2.7886,
+      "step": 701
+    },
+    {
+      "epoch": 0.1907608695652174,
+      "grad_norm": 2.013223171234131,
+      "learning_rate": 3e-05,
+      "loss": 3.365,
+      "step": 702
+    },
+    {
+      "epoch": 0.19103260869565217,
+      "grad_norm": 1.7130358219146729,
+      "learning_rate": 3e-05,
+      "loss": 3.7516,
+      "step": 703
+    },
+    {
+      "epoch": 0.19130434782608696,
+      "grad_norm": 1.7324903011322021,
+      "learning_rate": 3e-05,
+      "loss": 3.1275,
+      "step": 704
+    },
+    {
+      "epoch": 0.19157608695652173,
+      "grad_norm": 1.7227182388305664,
+      "learning_rate": 3e-05,
+      "loss": 3.1415,
+      "step": 705
+    },
+    {
+      "epoch": 0.19184782608695652,
+      "grad_norm": 1.3472079038619995,
+      "learning_rate": 3e-05,
+      "loss": 3.1949,
+      "step": 706
+    },
+    {
+      "epoch": 0.1921195652173913,
+      "grad_norm": 1.1573243141174316,
+      "learning_rate": 3e-05,
+      "loss": 2.8314,
+      "step": 707
+    },
+    {
+      "epoch": 0.1923913043478261,
+      "grad_norm": 1.6137231588363647,
+      "learning_rate": 3e-05,
+      "loss": 3.6169,
+      "step": 708
+    },
+    {
+      "epoch": 0.19266304347826088,
+      "grad_norm": 1.4710439443588257,
+      "learning_rate": 3e-05,
+      "loss": 3.4097,
+      "step": 709
+    },
+    {
+      "epoch": 0.19293478260869565,
+      "grad_norm": 2.347607374191284,
+      "learning_rate": 3e-05,
+      "loss": 4.1638,
+      "step": 710
+    },
+    {
+      "epoch": 0.19320652173913044,
+      "grad_norm": 1.4299064874649048,
+      "learning_rate": 3e-05,
+      "loss": 3.811,
+      "step": 711
+    },
+    {
+      "epoch": 0.1934782608695652,
+      "grad_norm": 1.5702636241912842,
+      "learning_rate": 3e-05,
+      "loss": 3.5396,
+      "step": 712
+    },
+    {
+      "epoch": 0.19375,
+      "grad_norm": 1.8417882919311523,
+      "learning_rate": 3e-05,
+      "loss": 3.5613,
+      "step": 713
+    },
+    {
+      "epoch": 0.19402173913043477,
+      "grad_norm": 1.4125804901123047,
+      "learning_rate": 3e-05,
+      "loss": 3.1808,
+      "step": 714
+    },
+    {
+      "epoch": 0.19429347826086957,
+      "grad_norm": 1.3733165264129639,
+      "learning_rate": 3e-05,
+      "loss": 3.3127,
+      "step": 715
+    },
+    {
+      "epoch": 0.19456521739130433,
+      "grad_norm": 2.1651833057403564,
+      "learning_rate": 3e-05,
+      "loss": 4.1732,
+      "step": 716
+    },
+    {
+      "epoch": 0.19483695652173913,
+      "grad_norm": 1.4096370935440063,
+      "learning_rate": 3e-05,
+      "loss": 3.5565,
+      "step": 717
+    },
+    {
+      "epoch": 0.19510869565217392,
+      "grad_norm": 1.7251945734024048,
+      "learning_rate": 3e-05,
+      "loss": 3.7342,
+      "step": 718
+    },
+    {
+      "epoch": 0.1953804347826087,
+      "grad_norm": 1.643470287322998,
+      "learning_rate": 3e-05,
+      "loss": 2.9678,
+      "step": 719
+    },
+    {
+      "epoch": 0.1956521739130435,
+      "grad_norm": 1.5827432870864868,
+      "learning_rate": 3e-05,
+      "loss": 3.1044,
+      "step": 720
+    },
+    {
+      "epoch": 0.19592391304347825,
+      "grad_norm": 1.4647548198699951,
+      "learning_rate": 3e-05,
+      "loss": 2.8413,
+      "step": 721
+    },
+    {
+      "epoch": 0.19619565217391305,
+      "grad_norm": 2.1561598777770996,
+      "learning_rate": 3e-05,
+      "loss": 3.4492,
+      "step": 722
+    },
+    {
+      "epoch": 0.19646739130434782,
+      "grad_norm": 1.6970568895339966,
+      "learning_rate": 3e-05,
+      "loss": 3.1848,
+      "step": 723
+    },
+    {
+      "epoch": 0.1967391304347826,
+      "grad_norm": 2.248117685317993,
+      "learning_rate": 3e-05,
+      "loss": 3.6296,
+      "step": 724
+    },
+    {
+      "epoch": 0.19701086956521738,
+      "grad_norm": 3.168339252471924,
+      "learning_rate": 3e-05,
+      "loss": 4.291,
+      "step": 725
+    },
+    {
+      "epoch": 0.19728260869565217,
+      "grad_norm": 3.802591323852539,
+      "learning_rate": 3e-05,
+      "loss": 3.9495,
+      "step": 726
+    },
+    {
+      "epoch": 0.19755434782608697,
+      "grad_norm": 3.848099708557129,
+      "learning_rate": 3e-05,
+      "loss": 3.6353,
+      "step": 727
+    },
+    {
+      "epoch": 0.19782608695652174,
+      "grad_norm": 3.534794569015503,
+      "learning_rate": 3e-05,
+      "loss": 3.9035,
+      "step": 728
+    },
+    {
+      "epoch": 0.19809782608695653,
+      "grad_norm": 1.7165888547897339,
+      "learning_rate": 3e-05,
+      "loss": 3.3266,
+      "step": 729
+    },
+    {
+      "epoch": 0.1983695652173913,
+      "grad_norm": 2.5262253284454346,
+      "learning_rate": 3e-05,
+      "loss": 3.6359,
+      "step": 730
+    },
+    {
+      "epoch": 0.1986413043478261,
+      "grad_norm": 2.5349607467651367,
+      "learning_rate": 3e-05,
+      "loss": 3.129,
+      "step": 731
+    },
+    {
+      "epoch": 0.19891304347826086,
+      "grad_norm": 5.680181503295898,
+      "learning_rate": 3e-05,
+      "loss": 4.8488,
+      "step": 732
+    },
+    {
+      "epoch": 0.19918478260869565,
+      "grad_norm": 3.0405654907226562,
+      "learning_rate": 3e-05,
+      "loss": 3.3313,
+      "step": 733
+    },
+    {
+      "epoch": 0.19945652173913042,
+      "grad_norm": 1.6890621185302734,
+      "learning_rate": 3e-05,
+      "loss": 3.2893,
+      "step": 734
+    },
+    {
+      "epoch": 0.19972826086956522,
+      "grad_norm": 2.204172372817993,
+      "learning_rate": 3e-05,
+      "loss": 3.5067,
+      "step": 735
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.858398914337158,
+      "learning_rate": 3e-05,
+      "loss": 3.4036,
+      "step": 736
+    },
+    {
+      "epoch": 0.20027173913043478,
+      "grad_norm": 2.4763529300689697,
+      "learning_rate": 3e-05,
+      "loss": 3.5171,
+      "step": 737
+    },
+    {
+      "epoch": 0.20054347826086957,
+      "grad_norm": 2.957118034362793,
+      "learning_rate": 3e-05,
+      "loss": 4.1728,
+      "step": 738
+    },
+    {
+      "epoch": 0.20081521739130434,
+      "grad_norm": 1.8309071063995361,
+      "learning_rate": 3e-05,
+      "loss": 2.9663,
+      "step": 739
+    },
+    {
+      "epoch": 0.20108695652173914,
+      "grad_norm": 2.270033121109009,
+      "learning_rate": 3e-05,
+      "loss": 3.6115,
+      "step": 740
+    },
+    {
+      "epoch": 0.2013586956521739,
+      "grad_norm": 2.1604442596435547,
+      "learning_rate": 3e-05,
+      "loss": 3.2511,
+      "step": 741
+    },
+    {
+      "epoch": 0.2016304347826087,
+      "grad_norm": 3.5561349391937256,
+      "learning_rate": 3e-05,
+      "loss": 3.8886,
+      "step": 742
+    },
+    {
+      "epoch": 0.20190217391304346,
+      "grad_norm": 2.510963201522827,
+      "learning_rate": 3e-05,
+      "loss": 3.4812,
+      "step": 743
+    },
+    {
+      "epoch": 0.20217391304347826,
+      "grad_norm": 1.5206657648086548,
+      "learning_rate": 3e-05,
+      "loss": 3.0231,
+      "step": 744
+    },
+    {
+      "epoch": 0.20244565217391305,
+      "grad_norm": 1.645155906677246,
+      "learning_rate": 3e-05,
+      "loss": 3.0951,
+      "step": 745
+    },
+    {
+      "epoch": 0.20271739130434782,
+      "grad_norm": 2.075399160385132,
+      "learning_rate": 3e-05,
+      "loss": 3.4057,
+      "step": 746
+    },
+    {
+      "epoch": 0.20298913043478262,
+      "grad_norm": 2.36306095123291,
+      "learning_rate": 3e-05,
+      "loss": 3.6449,
+      "step": 747
+    },
+    {
+      "epoch": 0.20326086956521738,
+      "grad_norm": 1.9054253101348877,
+      "learning_rate": 3e-05,
+      "loss": 3.3983,
+      "step": 748
+    },
+    {
+      "epoch": 0.20353260869565218,
+      "grad_norm": 1.7470569610595703,
+      "learning_rate": 3e-05,
+      "loss": 3.7903,
+      "step": 749
+    },
+    {
+      "epoch": 0.20380434782608695,
+      "grad_norm": 1.6297615766525269,
+      "learning_rate": 3e-05,
+      "loss": 3.0304,
+      "step": 750
+    },
+    {
+      "epoch": 0.20407608695652174,
+      "grad_norm": 3.2830233573913574,
+      "learning_rate": 3e-05,
+      "loss": 4.2781,
+      "step": 751
+    },
+    {
+      "epoch": 0.20434782608695654,
+      "grad_norm": 1.361449122428894,
+      "learning_rate": 3e-05,
+      "loss": 3.2224,
+      "step": 752
+    },
+    {
+      "epoch": 0.2046195652173913,
+      "grad_norm": 1.4894417524337769,
+      "learning_rate": 3e-05,
+      "loss": 3.262,
+      "step": 753
+    },
+    {
+      "epoch": 0.2048913043478261,
+      "grad_norm": 1.7271361351013184,
+      "learning_rate": 3e-05,
+      "loss": 3.3017,
+      "step": 754
+    },
+    {
+      "epoch": 0.20516304347826086,
+      "grad_norm": 2.033215284347534,
+      "learning_rate": 3e-05,
+      "loss": 2.9264,
+      "step": 755
+    },
+    {
+      "epoch": 0.20543478260869566,
+      "grad_norm": 2.808903455734253,
+      "learning_rate": 3e-05,
+      "loss": 3.7223,
+      "step": 756
+    },
+    {
+      "epoch": 0.20570652173913043,
+      "grad_norm": 1.450661540031433,
+      "learning_rate": 3e-05,
+      "loss": 4.0346,
+      "step": 757
+    },
+    {
+      "epoch": 0.20597826086956522,
+      "grad_norm": 1.9796857833862305,
+      "learning_rate": 3e-05,
+      "loss": 3.2978,
+      "step": 758
+    },
+    {
+      "epoch": 0.20625,
+      "grad_norm": 2.9408633708953857,
+      "learning_rate": 3e-05,
+      "loss": 3.7247,
+      "step": 759
+    },
+    {
+      "epoch": 0.20652173913043478,
+      "grad_norm": 1.4558497667312622,
+      "learning_rate": 3e-05,
+      "loss": 3.0258,
+      "step": 760
+    },
+    {
+      "epoch": 0.20679347826086958,
+      "grad_norm": 1.1229232549667358,
+      "learning_rate": 3e-05,
+      "loss": 2.9328,
+      "step": 761
+    },
+    {
+      "epoch": 0.20706521739130435,
+      "grad_norm": 1.663891077041626,
+      "learning_rate": 3e-05,
+      "loss": 3.2935,
+      "step": 762
+    },
+    {
+      "epoch": 0.20733695652173914,
+      "grad_norm": 1.8489412069320679,
+      "learning_rate": 3e-05,
+      "loss": 2.8728,
+      "step": 763
+    },
+    {
+      "epoch": 0.2076086956521739,
+      "grad_norm": 1.359864592552185,
+      "learning_rate": 3e-05,
+      "loss": 3.6552,
+      "step": 764
+    },
+    {
+      "epoch": 0.2078804347826087,
+      "grad_norm": 1.5933263301849365,
+      "learning_rate": 3e-05,
+      "loss": 3.2384,
+      "step": 765
+    },
+    {
+      "epoch": 0.20815217391304347,
+      "grad_norm": 3.0126466751098633,
+      "learning_rate": 3e-05,
+      "loss": 3.4475,
+      "step": 766
+    },
+    {
+      "epoch": 0.20842391304347826,
+      "grad_norm": 2.608242988586426,
+      "learning_rate": 3e-05,
+      "loss": 3.2701,
+      "step": 767
+    },
+    {
+      "epoch": 0.20869565217391303,
+      "grad_norm": 2.7247979640960693,
+      "learning_rate": 3e-05,
+      "loss": 3.96,
+      "step": 768
+    },
+    {
+      "epoch": 0.20896739130434783,
+      "grad_norm": 1.8706130981445312,
+      "learning_rate": 3e-05,
+      "loss": 3.0889,
+      "step": 769
+    },
+    {
+      "epoch": 0.20923913043478262,
+      "grad_norm": 2.668093681335449,
+      "learning_rate": 3e-05,
+      "loss": 3.4351,
+      "step": 770
+    },
+    {
+      "epoch": 0.2095108695652174,
+      "grad_norm": 3.017181396484375,
+      "learning_rate": 3e-05,
+      "loss": 3.2463,
+      "step": 771
+    },
+    {
+      "epoch": 0.20978260869565218,
+      "grad_norm": 3.0812008380889893,
+      "learning_rate": 3e-05,
+      "loss": 3.3998,
+      "step": 772
+    },
+    {
+      "epoch": 0.21005434782608695,
+      "grad_norm": 2.7510030269622803,
+      "learning_rate": 3e-05,
+      "loss": 4.2353,
+      "step": 773
+    },
+    {
+      "epoch": 0.21032608695652175,
+      "grad_norm": 2.4355452060699463,
+      "learning_rate": 3e-05,
+      "loss": 3.8853,
+      "step": 774
+    },
+    {
+      "epoch": 0.2105978260869565,
+      "grad_norm": 4.041244983673096,
+      "learning_rate": 3e-05,
+      "loss": 4.0771,
+      "step": 775
+    },
+    {
+      "epoch": 0.2108695652173913,
+      "grad_norm": 3.403510332107544,
+      "learning_rate": 3e-05,
+      "loss": 3.6963,
+      "step": 776
+    },
+    {
+      "epoch": 0.21114130434782608,
+      "grad_norm": 3.216482400894165,
+      "learning_rate": 3e-05,
+      "loss": 3.6608,
+      "step": 777
+    },
+    {
+      "epoch": 0.21141304347826087,
+      "grad_norm": 2.5989348888397217,
+      "learning_rate": 3e-05,
+      "loss": 3.689,
+      "step": 778
+    },
+    {
+      "epoch": 0.21168478260869567,
+      "grad_norm": 2.0239574909210205,
+      "learning_rate": 3e-05,
+      "loss": 3.681,
+      "step": 779
+    },
+    {
+      "epoch": 0.21195652173913043,
+      "grad_norm": 3.3878705501556396,
+      "learning_rate": 3e-05,
+      "loss": 3.9301,
+      "step": 780
+    },
+    {
+      "epoch": 0.21222826086956523,
+      "grad_norm": 4.106480598449707,
+      "learning_rate": 3e-05,
+      "loss": 3.2929,
+      "step": 781
+    },
+    {
+      "epoch": 0.2125,
+      "grad_norm": 3.4818196296691895,
+      "learning_rate": 3e-05,
+      "loss": 2.8562,
+      "step": 782
+    },
+    {
+      "epoch": 0.2127717391304348,
+      "grad_norm": 3.4106242656707764,
+      "learning_rate": 3e-05,
+      "loss": 3.6078,
+      "step": 783
+    },
+    {
+      "epoch": 0.21304347826086956,
+      "grad_norm": 2.401315450668335,
+      "learning_rate": 3e-05,
+      "loss": 3.693,
+      "step": 784
+    },
+    {
+      "epoch": 0.21331521739130435,
+      "grad_norm": 3.092789649963379,
+      "learning_rate": 3e-05,
+      "loss": 3.6907,
+      "step": 785
+    },
+    {
+      "epoch": 0.21358695652173912,
+      "grad_norm": 2.8016254901885986,
+      "learning_rate": 3e-05,
+      "loss": 3.7602,
+      "step": 786
+    },
+    {
+      "epoch": 0.2138586956521739,
+      "grad_norm": 2.281668186187744,
+      "learning_rate": 3e-05,
+      "loss": 2.9451,
+      "step": 787
+    },
+    {
+      "epoch": 0.2141304347826087,
+      "grad_norm": 2.3392529487609863,
+      "learning_rate": 3e-05,
+      "loss": 3.1969,
+      "step": 788
+    },
+    {
+      "epoch": 0.21440217391304348,
+      "grad_norm": 1.7437257766723633,
+      "learning_rate": 3e-05,
+      "loss": 3.2468,
+      "step": 789
+    },
+    {
+      "epoch": 0.21467391304347827,
+      "grad_norm": 2.128056049346924,
+      "learning_rate": 3e-05,
+      "loss": 3.5476,
+      "step": 790
+    },
+    {
+      "epoch": 0.21494565217391304,
+      "grad_norm": 2.1912710666656494,
+      "learning_rate": 3e-05,
+      "loss": 3.7689,
+      "step": 791
+    },
+    {
+      "epoch": 0.21521739130434783,
+      "grad_norm": 2.5414609909057617,
+      "learning_rate": 3e-05,
+      "loss": 3.4014,
+      "step": 792
+    },
+    {
+      "epoch": 0.2154891304347826,
+      "grad_norm": 2.6355223655700684,
+      "learning_rate": 3e-05,
+      "loss": 3.3787,
+      "step": 793
+    },
+    {
+      "epoch": 0.2157608695652174,
+      "grad_norm": 2.1927592754364014,
+      "learning_rate": 3e-05,
+      "loss": 3.2381,
+      "step": 794
+    },
+    {
+      "epoch": 0.21603260869565216,
+      "grad_norm": 1.7231847047805786,
+      "learning_rate": 3e-05,
+      "loss": 3.4627,
+      "step": 795
+    },
+    {
+      "epoch": 0.21630434782608696,
+      "grad_norm": 1.8754600286483765,
+      "learning_rate": 3e-05,
+      "loss": 3.7409,
+      "step": 796
+    },
+    {
+      "epoch": 0.21657608695652175,
+      "grad_norm": 1.7413159608840942,
+      "learning_rate": 3e-05,
+      "loss": 3.6012,
+      "step": 797
+    },
+    {
+      "epoch": 0.21684782608695652,
+      "grad_norm": 1.9197379350662231,
+      "learning_rate": 3e-05,
+      "loss": 3.15,
+      "step": 798
+    },
+    {
+      "epoch": 0.2171195652173913,
+      "grad_norm": 2.100706100463867,
+      "learning_rate": 3e-05,
+      "loss": 2.8748,
+      "step": 799
+    },
+    {
+      "epoch": 0.21739130434782608,
+      "grad_norm": 2.7185757160186768,
+      "learning_rate": 3e-05,
+      "loss": 3.7859,
+      "step": 800
+    },
+    {
+      "epoch": 0.21766304347826088,
+      "grad_norm": 3.241926908493042,
+      "learning_rate": 3e-05,
+      "loss": 3.7198,
+      "step": 801
+    },
+    {
+      "epoch": 0.21793478260869564,
+      "grad_norm": 2.365151882171631,
+      "learning_rate": 3e-05,
+      "loss": 3.7607,
+      "step": 802
+    },
+    {
+      "epoch": 0.21820652173913044,
+      "grad_norm": 2.0234482288360596,
+      "learning_rate": 3e-05,
+      "loss": 3.1843,
+      "step": 803
+    },
+    {
+      "epoch": 0.2184782608695652,
+      "grad_norm": 3.084533929824829,
+      "learning_rate": 3e-05,
+      "loss": 4.1945,
+      "step": 804
+    },
+    {
+      "epoch": 0.21875,
+      "grad_norm": 2.927222490310669,
+      "learning_rate": 3e-05,
+      "loss": 3.3531,
+      "step": 805
+    },
+    {
+      "epoch": 0.2190217391304348,
+      "grad_norm": 2.6623260974884033,
+      "learning_rate": 3e-05,
+      "loss": 3.5994,
+      "step": 806
+    },
+    {
+      "epoch": 0.21929347826086956,
+      "grad_norm": 2.8344593048095703,
+      "learning_rate": 3e-05,
+      "loss": 3.6001,
+      "step": 807
+    },
+    {
+      "epoch": 0.21956521739130436,
+      "grad_norm": 1.8181391954421997,
+      "learning_rate": 3e-05,
+      "loss": 2.9434,
+      "step": 808
+    },
+    {
+      "epoch": 0.21983695652173912,
+      "grad_norm": 2.103512763977051,
+      "learning_rate": 3e-05,
+      "loss": 3.8034,
+      "step": 809
+    },
+    {
+      "epoch": 0.22010869565217392,
+      "grad_norm": 1.961503505706787,
+      "learning_rate": 3e-05,
+      "loss": 3.0883,
+      "step": 810
+    },
+    {
+      "epoch": 0.2203804347826087,
+      "grad_norm": 2.3394432067871094,
+      "learning_rate": 3e-05,
+      "loss": 2.8568,
+      "step": 811
+    },
+    {
+      "epoch": 0.22065217391304348,
+      "grad_norm": 2.6501801013946533,
+      "learning_rate": 3e-05,
+      "loss": 3.2469,
+      "step": 812
+    },
+    {
+      "epoch": 0.22092391304347825,
+      "grad_norm": 2.7663819789886475,
+      "learning_rate": 3e-05,
+      "loss": 3.3066,
+      "step": 813
+    },
+    {
+      "epoch": 0.22119565217391304,
+      "grad_norm": 2.4567878246307373,
+      "learning_rate": 3e-05,
+      "loss": 3.0416,
+      "step": 814
+    },
+    {
+      "epoch": 0.22146739130434784,
+      "grad_norm": 2.378188371658325,
+      "learning_rate": 3e-05,
+      "loss": 3.6954,
+      "step": 815
+    },
+    {
+      "epoch": 0.2217391304347826,
+      "grad_norm": 2.611842632293701,
+      "learning_rate": 3e-05,
+      "loss": 3.3243,
+      "step": 816
+    },
+    {
+      "epoch": 0.2220108695652174,
+      "grad_norm": 2.1804463863372803,
+      "learning_rate": 3e-05,
+      "loss": 2.9858,
+      "step": 817
+    },
+    {
+      "epoch": 0.22228260869565217,
+      "grad_norm": 2.03049373626709,
+      "learning_rate": 3e-05,
+      "loss": 2.787,
+      "step": 818
+    },
+    {
+      "epoch": 0.22255434782608696,
+      "grad_norm": 1.8013122081756592,
+      "learning_rate": 3e-05,
+      "loss": 2.9299,
+      "step": 819
+    },
+    {
+      "epoch": 0.22282608695652173,
+      "grad_norm": 2.039745569229126,
+      "learning_rate": 3e-05,
+      "loss": 3.2247,
+      "step": 820
+    },
+    {
+      "epoch": 0.22309782608695652,
+      "grad_norm": 1.6249157190322876,
+      "learning_rate": 3e-05,
+      "loss": 2.8942,
+      "step": 821
+    },
+    {
+      "epoch": 0.2233695652173913,
+      "grad_norm": 1.8277668952941895,
+      "learning_rate": 3e-05,
+      "loss": 3.5634,
+      "step": 822
+    },
+    {
+      "epoch": 0.2236413043478261,
+      "grad_norm": 2.2749369144439697,
+      "learning_rate": 3e-05,
+      "loss": 3.6253,
+      "step": 823
+    },
+    {
+      "epoch": 0.22391304347826088,
+      "grad_norm": 2.3847649097442627,
+      "learning_rate": 3e-05,
+      "loss": 3.3725,
+      "step": 824
+    },
+    {
+      "epoch": 0.22418478260869565,
+      "grad_norm": 1.9465465545654297,
+      "learning_rate": 3e-05,
+      "loss": 3.4556,
+      "step": 825
+    },
+    {
+      "epoch": 0.22445652173913044,
+      "grad_norm": 2.793731927871704,
+      "learning_rate": 3e-05,
+      "loss": 4.0801,
+      "step": 826
+    },
+    {
+      "epoch": 0.2247282608695652,
+      "grad_norm": 2.605842113494873,
+      "learning_rate": 3e-05,
+      "loss": 4.0396,
+      "step": 827
+    },
+    {
+      "epoch": 0.225,
+      "grad_norm": 2.200025796890259,
+      "learning_rate": 3e-05,
+      "loss": 3.7514,
+      "step": 828
+    },
+    {
+      "epoch": 0.22527173913043477,
+      "grad_norm": 1.8318731784820557,
+      "learning_rate": 3e-05,
+      "loss": 3.3619,
+      "step": 829
+    },
+    {
+      "epoch": 0.22554347826086957,
+      "grad_norm": 1.7559170722961426,
+      "learning_rate": 3e-05,
+      "loss": 3.3115,
+      "step": 830
+    },
+    {
+      "epoch": 0.22581521739130433,
+      "grad_norm": 1.7032020092010498,
+      "learning_rate": 3e-05,
+      "loss": 3.113,
+      "step": 831
+    },
+    {
+      "epoch": 0.22608695652173913,
+      "grad_norm": 1.8972278833389282,
+      "learning_rate": 3e-05,
+      "loss": 3.1965,
+      "step": 832
+    },
+    {
+      "epoch": 0.22635869565217392,
+      "grad_norm": 1.9449539184570312,
+      "learning_rate": 3e-05,
+      "loss": 3.2422,
+      "step": 833
+    },
+    {
+      "epoch": 0.2266304347826087,
+      "grad_norm": 1.6739164590835571,
+      "learning_rate": 3e-05,
+      "loss": 3.2255,
+      "step": 834
+    },
+    {
+      "epoch": 0.2269021739130435,
+      "grad_norm": 1.7071807384490967,
+      "learning_rate": 3e-05,
+      "loss": 3.4032,
+      "step": 835
+    },
+    {
+      "epoch": 0.22717391304347825,
+      "grad_norm": 3.5912764072418213,
+      "learning_rate": 3e-05,
+      "loss": 4.1757,
+      "step": 836
+    },
+    {
+      "epoch": 0.22744565217391305,
+      "grad_norm": 1.969997763633728,
+      "learning_rate": 3e-05,
+      "loss": 3.1138,
+      "step": 837
+    },
+    {
+      "epoch": 0.22771739130434782,
+      "grad_norm": 1.9781277179718018,
+      "learning_rate": 3e-05,
+      "loss": 3.5066,
+      "step": 838
+    },
+    {
+      "epoch": 0.2279891304347826,
+      "grad_norm": 1.6115202903747559,
+      "learning_rate": 3e-05,
+      "loss": 2.9759,
+      "step": 839
+    },
+    {
+      "epoch": 0.22826086956521738,
+      "grad_norm": 1.3473846912384033,
+      "learning_rate": 3e-05,
+      "loss": 3.0617,
+      "step": 840
+    },
+    {
+      "epoch": 0.22853260869565217,
+      "grad_norm": 1.6643739938735962,
+      "learning_rate": 3e-05,
+      "loss": 3.4268,
+      "step": 841
+    },
+    {
+      "epoch": 0.22880434782608697,
+      "grad_norm": 1.9249311685562134,
+      "learning_rate": 3e-05,
+      "loss": 3.6207,
+      "step": 842
+    },
+    {
+      "epoch": 0.22907608695652174,
+      "grad_norm": 2.079634666442871,
+      "learning_rate": 3e-05,
+      "loss": 3.534,
+      "step": 843
+    },
+    {
+      "epoch": 0.22934782608695653,
+      "grad_norm": 1.8367491960525513,
+      "learning_rate": 3e-05,
+      "loss": 3.9664,
+      "step": 844
+    },
+    {
+      "epoch": 0.2296195652173913,
+      "grad_norm": 1.7043336629867554,
+      "learning_rate": 3e-05,
+      "loss": 3.4275,
+      "step": 845
+    },
+    {
+      "epoch": 0.2298913043478261,
+      "grad_norm": 1.6460905075073242,
+      "learning_rate": 3e-05,
+      "loss": 3.1771,
+      "step": 846
+    },
+    {
+      "epoch": 0.23016304347826086,
+      "grad_norm": 2.0371439456939697,
+      "learning_rate": 3e-05,
+      "loss": 3.3044,
+      "step": 847
+    },
+    {
+      "epoch": 0.23043478260869565,
+      "grad_norm": 1.5710325241088867,
+      "learning_rate": 3e-05,
+      "loss": 3.0044,
+      "step": 848
+    },
+    {
+      "epoch": 0.23070652173913042,
+      "grad_norm": 1.6866766214370728,
+      "learning_rate": 3e-05,
+      "loss": 2.9916,
+      "step": 849
+    },
+    {
+      "epoch": 0.23097826086956522,
+      "grad_norm": 2.127194404602051,
+      "learning_rate": 3e-05,
+      "loss": 3.5753,
+      "step": 850
+    },
+    {
+      "epoch": 0.23125,
+      "grad_norm": 2.0354788303375244,
+      "learning_rate": 3e-05,
+      "loss": 2.8992,
+      "step": 851
+    },
+    {
+      "epoch": 0.23152173913043478,
+      "grad_norm": 1.5567448139190674,
+      "learning_rate": 3e-05,
+      "loss": 3.4486,
+      "step": 852
+    },
+    {
+      "epoch": 0.23179347826086957,
+      "grad_norm": 1.3115332126617432,
+      "learning_rate": 3e-05,
+      "loss": 3.1354,
+      "step": 853
+    },
+    {
+      "epoch": 0.23206521739130434,
+      "grad_norm": 1.3727189302444458,
+      "learning_rate": 3e-05,
+      "loss": 3.0521,
+      "step": 854
+    },
+    {
+      "epoch": 0.23233695652173914,
+      "grad_norm": 1.3423751592636108,
+      "learning_rate": 3e-05,
+      "loss": 2.908,
+      "step": 855
+    },
+    {
+      "epoch": 0.2326086956521739,
+      "grad_norm": 1.61993408203125,
+      "learning_rate": 3e-05,
+      "loss": 3.2873,
+      "step": 856
+    },
+    {
+      "epoch": 0.2328804347826087,
+      "grad_norm": 1.8327138423919678,
+      "learning_rate": 3e-05,
+      "loss": 3.6789,
+      "step": 857
+    },
+    {
+      "epoch": 0.23315217391304346,
+      "grad_norm": 1.626006007194519,
+      "learning_rate": 3e-05,
+      "loss": 3.0027,
+      "step": 858
+    },
+    {
+      "epoch": 0.23342391304347826,
+      "grad_norm": 2.236844539642334,
+      "learning_rate": 3e-05,
+      "loss": 4.083,
+      "step": 859
+    },
+    {
+      "epoch": 0.23369565217391305,
+      "grad_norm": 2.1321239471435547,
+      "learning_rate": 3e-05,
+      "loss": 3.455,
+      "step": 860
+    },
+    {
+      "epoch": 0.23396739130434782,
+      "grad_norm": 2.1679024696350098,
+      "learning_rate": 3e-05,
+      "loss": 3.8644,
+      "step": 861
+    },
+    {
+      "epoch": 0.23423913043478262,
+      "grad_norm": 1.9821666479110718,
+      "learning_rate": 3e-05,
+      "loss": 3.5024,
+      "step": 862
+    },
+    {
+      "epoch": 0.23451086956521738,
+      "grad_norm": 2.8352460861206055,
+      "learning_rate": 3e-05,
+      "loss": 4.0082,
+      "step": 863
+    },
+    {
+      "epoch": 0.23478260869565218,
+      "grad_norm": 2.2241246700286865,
+      "learning_rate": 3e-05,
+      "loss": 3.8031,
+      "step": 864
+    },
+    {
+      "epoch": 0.23505434782608695,
+      "grad_norm": 2.1706037521362305,
+      "learning_rate": 3e-05,
+      "loss": 4.0421,
+      "step": 865
+    },
+    {
+      "epoch": 0.23532608695652174,
+      "grad_norm": 1.695703148841858,
+      "learning_rate": 3e-05,
+      "loss": 2.9672,
+      "step": 866
+    },
+    {
+      "epoch": 0.23559782608695654,
+      "grad_norm": 1.994674563407898,
+      "learning_rate": 3e-05,
+      "loss": 3.3449,
+      "step": 867
+    },
+    {
+      "epoch": 0.2358695652173913,
+      "grad_norm": 2.692945957183838,
+      "learning_rate": 3e-05,
+      "loss": 4.1761,
+      "step": 868
+    },
+    {
+      "epoch": 0.2361413043478261,
+      "grad_norm": 2.5395121574401855,
+      "learning_rate": 3e-05,
+      "loss": 3.921,
+      "step": 869
+    },
+    {
+      "epoch": 0.23641304347826086,
+      "grad_norm": 1.991248607635498,
+      "learning_rate": 3e-05,
+      "loss": 2.9,
+      "step": 870
+    },
+    {
+      "epoch": 0.23668478260869566,
+      "grad_norm": 2.2419328689575195,
+      "learning_rate": 3e-05,
+      "loss": 3.501,
+      "step": 871
+    },
+    {
+      "epoch": 0.23695652173913043,
+      "grad_norm": 1.9551500082015991,
+      "learning_rate": 3e-05,
+      "loss": 3.8966,
+      "step": 872
+    },
+    {
+      "epoch": 0.23722826086956522,
+      "grad_norm": 1.9755557775497437,
+      "learning_rate": 3e-05,
+      "loss": 3.7852,
+      "step": 873
+    },
+    {
+      "epoch": 0.2375,
+      "grad_norm": 1.964316487312317,
+      "learning_rate": 3e-05,
+      "loss": 3.4841,
+      "step": 874
+    },
+    {
+      "epoch": 0.23777173913043478,
+      "grad_norm": 1.667006015777588,
+      "learning_rate": 3e-05,
+      "loss": 3.3479,
+      "step": 875
+    },
+    {
+      "epoch": 0.23804347826086958,
+      "grad_norm": 1.3252580165863037,
+      "learning_rate": 3e-05,
+      "loss": 3.1133,
+      "step": 876
+    },
+    {
+      "epoch": 0.23831521739130435,
+      "grad_norm": 1.904585361480713,
+      "learning_rate": 3e-05,
+      "loss": 3.7604,
+      "step": 877
+    },
+    {
+      "epoch": 0.23858695652173914,
+      "grad_norm": 1.581666350364685,
+      "learning_rate": 3e-05,
+      "loss": 3.3015,
+      "step": 878
+    },
+    {
+      "epoch": 0.2388586956521739,
+      "grad_norm": 1.2517045736312866,
+      "learning_rate": 3e-05,
+      "loss": 3.2687,
+      "step": 879
+    },
+    {
+      "epoch": 0.2391304347826087,
+      "grad_norm": 1.5604926347732544,
+      "learning_rate": 3e-05,
+      "loss": 3.0638,
+      "step": 880
+    },
+    {
+      "epoch": 0.23940217391304347,
+      "grad_norm": 1.9475339651107788,
+      "learning_rate": 3e-05,
+      "loss": 3.8968,
+      "step": 881
+    },
+    {
+      "epoch": 0.23967391304347826,
+      "grad_norm": 1.528891921043396,
+      "learning_rate": 3e-05,
+      "loss": 3.0379,
+      "step": 882
+    },
+    {
+      "epoch": 0.23994565217391303,
+      "grad_norm": 1.838294506072998,
+      "learning_rate": 3e-05,
+      "loss": 3.7645,
+      "step": 883
+    },
+    {
+      "epoch": 0.24021739130434783,
+      "grad_norm": 1.5220407247543335,
+      "learning_rate": 3e-05,
+      "loss": 3.9417,
+      "step": 884
+    },
+    {
+      "epoch": 0.24048913043478262,
+      "grad_norm": 1.4615991115570068,
+      "learning_rate": 3e-05,
+      "loss": 3.4286,
+      "step": 885
+    },
+    {
+      "epoch": 0.2407608695652174,
+      "grad_norm": 2.569244623184204,
+      "learning_rate": 3e-05,
+      "loss": 3.9001,
+      "step": 886
+    },
+    {
+      "epoch": 0.24103260869565218,
+      "grad_norm": 2.3717269897460938,
+      "learning_rate": 3e-05,
+      "loss": 4.2076,
+      "step": 887
+    },
+    {
+      "epoch": 0.24130434782608695,
+      "grad_norm": 1.8509140014648438,
+      "learning_rate": 3e-05,
+      "loss": 3.2101,
+      "step": 888
+    },
+    {
+      "epoch": 0.24157608695652175,
+      "grad_norm": 2.0046916007995605,
+      "learning_rate": 3e-05,
+      "loss": 2.9609,
+      "step": 889
+    },
+    {
+      "epoch": 0.2418478260869565,
+      "grad_norm": 2.2359912395477295,
+      "learning_rate": 3e-05,
+      "loss": 3.6416,
+      "step": 890
+    },
+    {
+      "epoch": 0.2421195652173913,
+      "grad_norm": 1.8625085353851318,
+      "learning_rate": 3e-05,
+      "loss": 3.3007,
+      "step": 891
+    },
+    {
+      "epoch": 0.24239130434782608,
+      "grad_norm": 2.0958662033081055,
+      "learning_rate": 3e-05,
+      "loss": 2.7472,
+      "step": 892
+    },
+    {
+      "epoch": 0.24266304347826087,
+      "grad_norm": 2.5501413345336914,
+      "learning_rate": 3e-05,
+      "loss": 3.0367,
+      "step": 893
+    },
+    {
+      "epoch": 0.24293478260869567,
+      "grad_norm": 3.020587921142578,
+      "learning_rate": 3e-05,
+      "loss": 3.7357,
+      "step": 894
+    },
+    {
+      "epoch": 0.24320652173913043,
+      "grad_norm": 2.0165812969207764,
+      "learning_rate": 3e-05,
+      "loss": 3.6235,
+      "step": 895
+    },
+    {
+      "epoch": 0.24347826086956523,
+      "grad_norm": 3.1172547340393066,
+      "learning_rate": 3e-05,
+      "loss": 4.1981,
+      "step": 896
+    },
+    {
+      "epoch": 0.24375,
+      "grad_norm": 2.381237268447876,
+      "learning_rate": 3e-05,
+      "loss": 3.104,
+      "step": 897
+    },
+    {
+      "epoch": 0.2440217391304348,
+      "grad_norm": 3.345118522644043,
+      "learning_rate": 3e-05,
+      "loss": 4.1412,
+      "step": 898
+    },
+    {
+      "epoch": 0.24429347826086956,
+      "grad_norm": 2.8453972339630127,
+      "learning_rate": 3e-05,
+      "loss": 3.3134,
+      "step": 899
+    },
+    {
+      "epoch": 0.24456521739130435,
+      "grad_norm": 1.9727962017059326,
+      "learning_rate": 3e-05,
+      "loss": 3.1409,
+      "step": 900
+    },
+    {
+      "epoch": 0.24483695652173912,
+      "grad_norm": 1.3924411535263062,
+      "learning_rate": 3e-05,
+      "loss": 3.2174,
+      "step": 901
+    },
+    {
+      "epoch": 0.2451086956521739,
+      "grad_norm": 1.8885339498519897,
+      "learning_rate": 3e-05,
+      "loss": 3.4122,
+      "step": 902
+    },
+    {
+      "epoch": 0.2453804347826087,
+      "grad_norm": 2.3274450302124023,
+      "learning_rate": 3e-05,
+      "loss": 3.2905,
+      "step": 903
+    },
+    {
+      "epoch": 0.24565217391304348,
+      "grad_norm": 1.877142310142517,
+      "learning_rate": 3e-05,
+      "loss": 3.4118,
+      "step": 904
+    },
+    {
+      "epoch": 0.24592391304347827,
+      "grad_norm": 1.8108350038528442,
+      "learning_rate": 3e-05,
+      "loss": 3.206,
+      "step": 905
+    },
+    {
+      "epoch": 0.24619565217391304,
+      "grad_norm": 1.6787278652191162,
+      "learning_rate": 3e-05,
+      "loss": 3.2236,
+      "step": 906
+    },
+    {
+      "epoch": 0.24646739130434783,
+      "grad_norm": 2.007154703140259,
+      "learning_rate": 3e-05,
+      "loss": 4.1134,
+      "step": 907
+    },
+    {
+      "epoch": 0.2467391304347826,
+      "grad_norm": 1.9029121398925781,
+      "learning_rate": 3e-05,
+      "loss": 3.2066,
+      "step": 908
+    },
+    {
+      "epoch": 0.2470108695652174,
+      "grad_norm": 1.5358937978744507,
+      "learning_rate": 3e-05,
+      "loss": 3.3406,
+      "step": 909
+    },
+    {
+      "epoch": 0.24728260869565216,
+      "grad_norm": 1.706233263015747,
+      "learning_rate": 3e-05,
+      "loss": 3.368,
+      "step": 910
+    },
+    {
+      "epoch": 0.24755434782608696,
+      "grad_norm": 2.283039093017578,
+      "learning_rate": 3e-05,
+      "loss": 4.1815,
+      "step": 911
+    },
+    {
+      "epoch": 0.24782608695652175,
+      "grad_norm": 1.6811976432800293,
+      "learning_rate": 3e-05,
+      "loss": 3.8339,
+      "step": 912
+    },
+    {
+      "epoch": 0.24809782608695652,
+      "grad_norm": 1.6838414669036865,
+      "learning_rate": 3e-05,
+      "loss": 3.6558,
+      "step": 913
+    },
+    {
+      "epoch": 0.2483695652173913,
+      "grad_norm": 1.451900839805603,
+      "learning_rate": 3e-05,
+      "loss": 3.2279,
+      "step": 914
+    },
+    {
+      "epoch": 0.24864130434782608,
+      "grad_norm": 2.3503222465515137,
+      "learning_rate": 3e-05,
+      "loss": 3.9484,
+      "step": 915
+    },
+    {
+      "epoch": 0.24891304347826088,
+      "grad_norm": 2.001974582672119,
+      "learning_rate": 3e-05,
+      "loss": 3.3028,
+      "step": 916
+    },
+    {
+      "epoch": 0.24918478260869564,
+      "grad_norm": 1.5036247968673706,
+      "learning_rate": 3e-05,
+      "loss": 3.2459,
+      "step": 917
+    },
+    {
+      "epoch": 0.24945652173913044,
+      "grad_norm": 2.033236026763916,
+      "learning_rate": 3e-05,
+      "loss": 3.4323,
+      "step": 918
+    },
+    {
+      "epoch": 0.2497282608695652,
+      "grad_norm": 1.4043093919754028,
+      "learning_rate": 3e-05,
+      "loss": 3.0965,
+      "step": 919
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.203482985496521,
+      "learning_rate": 3e-05,
+      "loss": 3.1118,
+      "step": 920
+    },
+    {
+      "epoch": 0.25027173913043477,
+      "grad_norm": 1.4571930170059204,
+      "learning_rate": 3e-05,
+      "loss": 3.351,
+      "step": 921
+    },
+    {
+      "epoch": 0.2505434782608696,
+      "grad_norm": 0.9827572703361511,
+      "learning_rate": 3e-05,
+      "loss": 2.8122,
+      "step": 922
+    },
+    {
+      "epoch": 0.25081521739130436,
+      "grad_norm": 1.520359754562378,
+      "learning_rate": 3e-05,
+      "loss": 3.6678,
+      "step": 923
+    },
+    {
+      "epoch": 0.2510869565217391,
+      "grad_norm": 1.277106761932373,
+      "learning_rate": 3e-05,
+      "loss": 2.9916,
+      "step": 924
+    },
+    {
+      "epoch": 0.2513586956521739,
+      "grad_norm": 1.7534111738204956,
+      "learning_rate": 3e-05,
+      "loss": 3.9959,
+      "step": 925
+    },
+    {
+      "epoch": 0.2516304347826087,
+      "grad_norm": 1.031008005142212,
+      "learning_rate": 3e-05,
+      "loss": 2.9768,
+      "step": 926
+    },
+    {
+      "epoch": 0.2519021739130435,
+      "grad_norm": 1.1608198881149292,
+      "learning_rate": 3e-05,
+      "loss": 3.456,
+      "step": 927
+    },
+    {
+      "epoch": 0.25217391304347825,
+      "grad_norm": 1.4016661643981934,
+      "learning_rate": 3e-05,
+      "loss": 3.4423,
+      "step": 928
+    },
+    {
+      "epoch": 0.25244565217391307,
+      "grad_norm": 1.1200820207595825,
+      "learning_rate": 3e-05,
+      "loss": 2.897,
+      "step": 929
+    },
+    {
+      "epoch": 0.25271739130434784,
+      "grad_norm": 1.4193557500839233,
+      "learning_rate": 3e-05,
+      "loss": 3.7882,
+      "step": 930
+    },
+    {
+      "epoch": 0.2529891304347826,
+      "grad_norm": 1.2631326913833618,
+      "learning_rate": 3e-05,
+      "loss": 3.8642,
+      "step": 931
+    },
+    {
+      "epoch": 0.2532608695652174,
+      "grad_norm": 1.6645845174789429,
+      "learning_rate": 3e-05,
+      "loss": 3.7794,
+      "step": 932
+    },
+    {
+      "epoch": 0.2535326086956522,
+      "grad_norm": 1.7016172409057617,
+      "learning_rate": 3e-05,
+      "loss": 3.3343,
+      "step": 933
+    },
+    {
+      "epoch": 0.25380434782608696,
+      "grad_norm": 1.1917381286621094,
+      "learning_rate": 3e-05,
+      "loss": 3.3563,
+      "step": 934
+    },
+    {
+      "epoch": 0.25407608695652173,
+      "grad_norm": 1.4421312808990479,
+      "learning_rate": 3e-05,
+      "loss": 3.6694,
+      "step": 935
+    },
+    {
+      "epoch": 0.2543478260869565,
+      "grad_norm": 1.2393507957458496,
+      "learning_rate": 3e-05,
+      "loss": 2.8572,
+      "step": 936
+    },
+    {
+      "epoch": 0.2546195652173913,
+      "grad_norm": 1.4665427207946777,
+      "learning_rate": 3e-05,
+      "loss": 3.5944,
+      "step": 937
+    },
+    {
+      "epoch": 0.2548913043478261,
+      "grad_norm": 1.3151808977127075,
+      "learning_rate": 3e-05,
+      "loss": 3.1136,
+      "step": 938
+    },
+    {
+      "epoch": 0.25516304347826085,
+      "grad_norm": 1.3733018636703491,
+      "learning_rate": 3e-05,
+      "loss": 3.1559,
+      "step": 939
+    },
+    {
+      "epoch": 0.2554347826086957,
+      "grad_norm": 1.638494610786438,
+      "learning_rate": 3e-05,
+      "loss": 3.5616,
+      "step": 940
+    },
+    {
+      "epoch": 0.25570652173913044,
+      "grad_norm": 1.5186861753463745,
+      "learning_rate": 3e-05,
+      "loss": 3.7529,
+      "step": 941
+    },
+    {
+      "epoch": 0.2559782608695652,
+      "grad_norm": 1.7179397344589233,
+      "learning_rate": 3e-05,
+      "loss": 3.8634,
+      "step": 942
+    },
+    {
+      "epoch": 0.25625,
+      "grad_norm": 1.9366356134414673,
+      "learning_rate": 3e-05,
+      "loss": 3.785,
+      "step": 943
+    },
+    {
+      "epoch": 0.2565217391304348,
+      "grad_norm": 2.2738037109375,
+      "learning_rate": 3e-05,
+      "loss": 3.7442,
+      "step": 944
+    },
+    {
+      "epoch": 0.25679347826086957,
+      "grad_norm": 2.67051100730896,
+      "learning_rate": 3e-05,
+      "loss": 3.8186,
+      "step": 945
+    },
+    {
+      "epoch": 0.25706521739130433,
+      "grad_norm": 3.3741190433502197,
+      "learning_rate": 3e-05,
+      "loss": 4.2118,
+      "step": 946
+    },
+    {
+      "epoch": 0.25733695652173916,
+      "grad_norm": 1.6262043714523315,
+      "learning_rate": 3e-05,
+      "loss": 3.0208,
+      "step": 947
+    },
+    {
+      "epoch": 0.2576086956521739,
+      "grad_norm": 1.7342244386672974,
+      "learning_rate": 3e-05,
+      "loss": 3.1897,
+      "step": 948
+    },
+    {
+      "epoch": 0.2578804347826087,
+      "grad_norm": 2.0924177169799805,
+      "learning_rate": 3e-05,
+      "loss": 2.7324,
+      "step": 949
+    },
+    {
+      "epoch": 0.25815217391304346,
+      "grad_norm": 2.21530818939209,
+      "learning_rate": 3e-05,
+      "loss": 3.1098,
+      "step": 950
+    },
+    {
+      "epoch": 0.2584239130434783,
+      "grad_norm": 1.642945647239685,
+      "learning_rate": 3e-05,
+      "loss": 3.58,
+      "step": 951
+    },
+    {
+      "epoch": 0.25869565217391305,
+      "grad_norm": 1.4641486406326294,
+      "learning_rate": 3e-05,
+      "loss": 3.4771,
+      "step": 952
+    },
+    {
+      "epoch": 0.2589673913043478,
+      "grad_norm": 1.4831671714782715,
+      "learning_rate": 3e-05,
+      "loss": 3.0967,
+      "step": 953
+    },
+    {
+      "epoch": 0.2592391304347826,
+      "grad_norm": 1.605242133140564,
+      "learning_rate": 3e-05,
+      "loss": 3.3319,
+      "step": 954
+    },
+    {
+      "epoch": 0.2595108695652174,
+      "grad_norm": 2.312819480895996,
+      "learning_rate": 3e-05,
+      "loss": 4.214,
+      "step": 955
+    },
+    {
+      "epoch": 0.2597826086956522,
+      "grad_norm": 1.751943588256836,
+      "learning_rate": 3e-05,
+      "loss": 3.3116,
+      "step": 956
+    },
+    {
+      "epoch": 0.26005434782608694,
+      "grad_norm": 1.951372504234314,
+      "learning_rate": 3e-05,
+      "loss": 3.4478,
+      "step": 957
+    },
+    {
+      "epoch": 0.26032608695652176,
+      "grad_norm": 2.3981149196624756,
+      "learning_rate": 3e-05,
+      "loss": 3.4107,
+      "step": 958
+    },
+    {
+      "epoch": 0.26059782608695653,
+      "grad_norm": 2.448025703430176,
+      "learning_rate": 3e-05,
+      "loss": 4.0686,
+      "step": 959
+    },
+    {
+      "epoch": 0.2608695652173913,
+      "grad_norm": 2.1298911571502686,
+      "learning_rate": 3e-05,
+      "loss": 3.4676,
+      "step": 960
+    },
+    {
+      "epoch": 0.26114130434782606,
+      "grad_norm": 1.9788066148757935,
+      "learning_rate": 3e-05,
+      "loss": 3.4239,
+      "step": 961
+    },
+    {
+      "epoch": 0.2614130434782609,
+      "grad_norm": 1.6203289031982422,
+      "learning_rate": 3e-05,
+      "loss": 3.2623,
+      "step": 962
+    },
+    {
+      "epoch": 0.26168478260869565,
+      "grad_norm": 1.7709441184997559,
+      "learning_rate": 3e-05,
+      "loss": 3.5855,
+      "step": 963
+    },
+    {
+      "epoch": 0.2619565217391304,
+      "grad_norm": 1.4671415090560913,
+      "learning_rate": 3e-05,
+      "loss": 3.2081,
+      "step": 964
+    },
+    {
+      "epoch": 0.26222826086956524,
+      "grad_norm": 1.2123912572860718,
+      "learning_rate": 3e-05,
+      "loss": 3.0287,
+      "step": 965
+    },
+    {
+      "epoch": 0.2625,
+      "grad_norm": 1.9638895988464355,
+      "learning_rate": 3e-05,
+      "loss": 3.4101,
+      "step": 966
+    },
+    {
+      "epoch": 0.2627717391304348,
+      "grad_norm": 1.6470272541046143,
+      "learning_rate": 3e-05,
+      "loss": 3.1761,
+      "step": 967
+    },
+    {
+      "epoch": 0.26304347826086955,
+      "grad_norm": 1.7481547594070435,
+      "learning_rate": 3e-05,
+      "loss": 3.7102,
+      "step": 968
+    },
+    {
+      "epoch": 0.26331521739130437,
+      "grad_norm": 1.589220404624939,
+      "learning_rate": 3e-05,
+      "loss": 2.9376,
+      "step": 969
+    },
+    {
+      "epoch": 0.26358695652173914,
+      "grad_norm": 2.317216634750366,
+      "learning_rate": 3e-05,
+      "loss": 3.888,
+      "step": 970
+    },
+    {
+      "epoch": 0.2638586956521739,
+      "grad_norm": 1.9831353425979614,
+      "learning_rate": 3e-05,
+      "loss": 3.831,
+      "step": 971
+    },
+    {
+      "epoch": 0.26413043478260867,
+      "grad_norm": 2.201885223388672,
+      "learning_rate": 3e-05,
+      "loss": 3.5633,
+      "step": 972
+    },
+    {
+      "epoch": 0.2644021739130435,
+      "grad_norm": 1.755304217338562,
+      "learning_rate": 3e-05,
+      "loss": 3.9027,
+      "step": 973
+    },
+    {
+      "epoch": 0.26467391304347826,
+      "grad_norm": 1.7514090538024902,
+      "learning_rate": 3e-05,
+      "loss": 3.3256,
+      "step": 974
+    },
+    {
+      "epoch": 0.264945652173913,
+      "grad_norm": 2.1359901428222656,
+      "learning_rate": 3e-05,
+      "loss": 3.9904,
+      "step": 975
+    },
+    {
+      "epoch": 0.26521739130434785,
+      "grad_norm": 1.5768872499465942,
+      "learning_rate": 3e-05,
+      "loss": 3.3242,
+      "step": 976
+    },
+    {
+      "epoch": 0.2654891304347826,
+      "grad_norm": 1.8486355543136597,
+      "learning_rate": 3e-05,
+      "loss": 3.7133,
+      "step": 977
+    },
+    {
+      "epoch": 0.2657608695652174,
+      "grad_norm": 1.5333871841430664,
+      "learning_rate": 3e-05,
+      "loss": 3.4191,
+      "step": 978
+    },
+    {
+      "epoch": 0.26603260869565215,
+      "grad_norm": 1.6924866437911987,
+      "learning_rate": 3e-05,
+      "loss": 3.3763,
+      "step": 979
+    },
+    {
+      "epoch": 0.266304347826087,
+      "grad_norm": 2.1424450874328613,
+      "learning_rate": 3e-05,
+      "loss": 3.8962,
+      "step": 980
+    },
+    {
+      "epoch": 0.26657608695652174,
+      "grad_norm": 1.5438109636306763,
+      "learning_rate": 3e-05,
+      "loss": 3.2176,
+      "step": 981
+    },
+    {
+      "epoch": 0.2668478260869565,
+      "grad_norm": 1.5958982706069946,
+      "learning_rate": 3e-05,
+      "loss": 3.8708,
+      "step": 982
+    },
+    {
+      "epoch": 0.26711956521739133,
+      "grad_norm": 1.129310965538025,
+      "learning_rate": 3e-05,
+      "loss": 2.8674,
+      "step": 983
+    },
+    {
+      "epoch": 0.2673913043478261,
+      "grad_norm": 1.6280345916748047,
+      "learning_rate": 3e-05,
+      "loss": 4.165,
+      "step": 984
+    },
+    {
+      "epoch": 0.26766304347826086,
+      "grad_norm": 1.499475121498108,
+      "learning_rate": 3e-05,
+      "loss": 2.9953,
+      "step": 985
+    },
+    {
+      "epoch": 0.26793478260869563,
+      "grad_norm": 1.253732442855835,
+      "learning_rate": 3e-05,
+      "loss": 3.4718,
+      "step": 986
+    },
+    {
+      "epoch": 0.26820652173913045,
+      "grad_norm": 1.113315463066101,
+      "learning_rate": 3e-05,
+      "loss": 3.3101,
+      "step": 987
+    },
+    {
+      "epoch": 0.2684782608695652,
+      "grad_norm": 1.4947681427001953,
+      "learning_rate": 3e-05,
+      "loss": 3.217,
+      "step": 988
+    },
+    {
+      "epoch": 0.26875,
+      "grad_norm": 1.9959630966186523,
+      "learning_rate": 3e-05,
+      "loss": 4.0953,
+      "step": 989
+    },
+    {
+      "epoch": 0.26902173913043476,
+      "grad_norm": 1.6297112703323364,
+      "learning_rate": 3e-05,
+      "loss": 3.4108,
+      "step": 990
+    },
+    {
+      "epoch": 0.2692934782608696,
+      "grad_norm": 1.9028682708740234,
+      "learning_rate": 3e-05,
+      "loss": 4.1301,
+      "step": 991
+    },
+    {
+      "epoch": 0.26956521739130435,
+      "grad_norm": 1.6655460596084595,
+      "learning_rate": 3e-05,
+      "loss": 3.5511,
+      "step": 992
+    },
+    {
+      "epoch": 0.2698369565217391,
+      "grad_norm": 1.8980700969696045,
+      "learning_rate": 3e-05,
+      "loss": 3.5143,
+      "step": 993
+    },
+    {
+      "epoch": 0.27010869565217394,
+      "grad_norm": 1.5208625793457031,
+      "learning_rate": 3e-05,
+      "loss": 3.5514,
+      "step": 994
+    },
+    {
+      "epoch": 0.2703804347826087,
+      "grad_norm": 1.1632534265518188,
+      "learning_rate": 3e-05,
+      "loss": 3.2167,
+      "step": 995
+    },
+    {
+      "epoch": 0.27065217391304347,
+      "grad_norm": 2.094667673110962,
+      "learning_rate": 3e-05,
+      "loss": 3.0362,
+      "step": 996
+    },
+    {
+      "epoch": 0.27092391304347824,
+      "grad_norm": 1.4889837503433228,
+      "learning_rate": 3e-05,
+      "loss": 3.4279,
+      "step": 997
+    },
+    {
+      "epoch": 0.27119565217391306,
+      "grad_norm": 1.9965155124664307,
+      "learning_rate": 3e-05,
+      "loss": 3.774,
+      "step": 998
+    },
+    {
+      "epoch": 0.2714673913043478,
+      "grad_norm": 1.5410970449447632,
+      "learning_rate": 3e-05,
+      "loss": 3.2647,
+      "step": 999
+    },
+    {
+      "epoch": 0.2717391304347826,
+      "grad_norm": 1.9614113569259644,
+      "learning_rate": 3e-05,
+      "loss": 3.5648,
+      "step": 1000
+    },
+    {
+      "epoch": 0.2720108695652174,
+      "grad_norm": 1.3656377792358398,
+      "learning_rate": 3e-05,
+      "loss": 3.3263,
+      "step": 1001
+    },
+    {
+      "epoch": 0.2722826086956522,
+      "grad_norm": 2.057823419570923,
+      "learning_rate": 3e-05,
+      "loss": 3.1952,
+      "step": 1002
+    },
+    {
+      "epoch": 0.27255434782608695,
+      "grad_norm": 1.9812930822372437,
+      "learning_rate": 3e-05,
+      "loss": 3.5128,
+      "step": 1003
+    },
+    {
+      "epoch": 0.2728260869565217,
+      "grad_norm": 1.443973183631897,
+      "learning_rate": 3e-05,
+      "loss": 3.5108,
+      "step": 1004
+    },
+    {
+      "epoch": 0.27309782608695654,
+      "grad_norm": 1.1789878606796265,
+      "learning_rate": 3e-05,
+      "loss": 3.142,
+      "step": 1005
+    },
+    {
+      "epoch": 0.2733695652173913,
+      "grad_norm": 1.5677335262298584,
+      "learning_rate": 3e-05,
+      "loss": 3.0946,
+      "step": 1006
+    },
+    {
+      "epoch": 0.2736413043478261,
+      "grad_norm": 1.6985656023025513,
+      "learning_rate": 3e-05,
+      "loss": 3.8017,
+      "step": 1007
+    },
+    {
+      "epoch": 0.27391304347826084,
+      "grad_norm": 0.9420762658119202,
+      "learning_rate": 3e-05,
+      "loss": 2.8754,
+      "step": 1008
+    },
+    {
+      "epoch": 0.27418478260869567,
+      "grad_norm": 1.1210711002349854,
+      "learning_rate": 3e-05,
+      "loss": 3.1734,
+      "step": 1009
+    },
+    {
+      "epoch": 0.27445652173913043,
+      "grad_norm": 0.9131266474723816,
+      "learning_rate": 3e-05,
+      "loss": 2.5223,
+      "step": 1010
+    },
+    {
+      "epoch": 0.2747282608695652,
+      "grad_norm": 1.4077900648117065,
+      "learning_rate": 3e-05,
+      "loss": 3.2054,
+      "step": 1011
+    },
+    {
+      "epoch": 0.275,
+      "grad_norm": 1.4042190313339233,
+      "learning_rate": 3e-05,
+      "loss": 3.3916,
+      "step": 1012
+    },
+    {
+      "epoch": 0.2752717391304348,
+      "grad_norm": 1.1815489530563354,
+      "learning_rate": 3e-05,
+      "loss": 3.2108,
+      "step": 1013
+    },
+    {
+      "epoch": 0.27554347826086956,
+      "grad_norm": 1.007114052772522,
+      "learning_rate": 3e-05,
+      "loss": 2.8035,
+      "step": 1014
+    },
+    {
+      "epoch": 0.2758152173913043,
+      "grad_norm": 1.6630133390426636,
+      "learning_rate": 3e-05,
+      "loss": 3.5364,
+      "step": 1015
+    },
+    {
+      "epoch": 0.27608695652173915,
+      "grad_norm": 1.3151155710220337,
+      "learning_rate": 3e-05,
+      "loss": 3.8982,
+      "step": 1016
+    },
+    {
+      "epoch": 0.2763586956521739,
+      "grad_norm": 1.3191099166870117,
+      "learning_rate": 3e-05,
+      "loss": 3.3812,
+      "step": 1017
+    },
+    {
+      "epoch": 0.2766304347826087,
+      "grad_norm": 0.9553658366203308,
+      "learning_rate": 3e-05,
+      "loss": 3.2285,
+      "step": 1018
+    },
+    {
+      "epoch": 0.2769021739130435,
+      "grad_norm": 1.2714715003967285,
+      "learning_rate": 3e-05,
+      "loss": 4.1331,
+      "step": 1019
+    },
+    {
+      "epoch": 0.27717391304347827,
+      "grad_norm": 1.4878464937210083,
+      "learning_rate": 3e-05,
+      "loss": 3.5607,
+      "step": 1020
+    },
+    {
+      "epoch": 0.27744565217391304,
+      "grad_norm": 1.3467168807983398,
+      "learning_rate": 3e-05,
+      "loss": 3.4943,
+      "step": 1021
+    },
+    {
+      "epoch": 0.2777173913043478,
+      "grad_norm": 1.052966833114624,
+      "learning_rate": 3e-05,
+      "loss": 2.9399,
+      "step": 1022
+    },
+    {
+      "epoch": 0.2779891304347826,
+      "grad_norm": 1.1948199272155762,
+      "learning_rate": 3e-05,
+      "loss": 3.0152,
+      "step": 1023
+    },
+    {
+      "epoch": 0.2782608695652174,
+      "grad_norm": 1.1960163116455078,
+      "learning_rate": 3e-05,
+      "loss": 2.9338,
+      "step": 1024
+    },
+    {
+      "epoch": 0.27853260869565216,
+      "grad_norm": 1.2721116542816162,
+      "learning_rate": 3e-05,
+      "loss": 3.2096,
+      "step": 1025
+    },
+    {
+      "epoch": 0.27880434782608693,
+      "grad_norm": 1.620215654373169,
+      "learning_rate": 3e-05,
+      "loss": 3.3828,
+      "step": 1026
+    },
+    {
+      "epoch": 0.27907608695652175,
+      "grad_norm": 1.4918383359909058,
+      "learning_rate": 3e-05,
+      "loss": 3.5746,
+      "step": 1027
+    },
+    {
+      "epoch": 0.2793478260869565,
+      "grad_norm": 1.9811660051345825,
+      "learning_rate": 3e-05,
+      "loss": 4.0559,
+      "step": 1028
+    },
+    {
+      "epoch": 0.2796195652173913,
+      "grad_norm": 1.3812586069107056,
+      "learning_rate": 3e-05,
+      "loss": 3.1452,
+      "step": 1029
+    },
+    {
+      "epoch": 0.2798913043478261,
+      "grad_norm": 1.53925359249115,
+      "learning_rate": 3e-05,
+      "loss": 3.9887,
+      "step": 1030
+    },
+    {
+      "epoch": 0.2801630434782609,
+      "grad_norm": 2.2425036430358887,
+      "learning_rate": 3e-05,
+      "loss": 4.2516,
+      "step": 1031
+    },
+    {
+      "epoch": 0.28043478260869564,
+      "grad_norm": 1.2638776302337646,
+      "learning_rate": 3e-05,
+      "loss": 3.5207,
+      "step": 1032
+    },
+    {
+      "epoch": 0.2807065217391304,
+      "grad_norm": 1.1238954067230225,
+      "learning_rate": 3e-05,
+      "loss": 3.4072,
+      "step": 1033
+    },
+    {
+      "epoch": 0.28097826086956523,
+      "grad_norm": 1.2834945917129517,
+      "learning_rate": 3e-05,
+      "loss": 3.0585,
+      "step": 1034
+    },
+    {
+      "epoch": 0.28125,
+      "grad_norm": 1.5284252166748047,
+      "learning_rate": 3e-05,
+      "loss": 3.1547,
+      "step": 1035
+    },
+    {
+      "epoch": 0.28152173913043477,
+      "grad_norm": 1.8825403451919556,
+      "learning_rate": 3e-05,
+      "loss": 3.892,
+      "step": 1036
+    },
+    {
+      "epoch": 0.2817934782608696,
+      "grad_norm": 1.620376467704773,
+      "learning_rate": 3e-05,
+      "loss": 2.9044,
+      "step": 1037
+    },
+    {
+      "epoch": 0.28206521739130436,
+      "grad_norm": 2.1728601455688477,
+      "learning_rate": 3e-05,
+      "loss": 3.1266,
+      "step": 1038
+    },
+    {
+      "epoch": 0.2823369565217391,
+      "grad_norm": 2.090522050857544,
+      "learning_rate": 3e-05,
+      "loss": 4.4403,
+      "step": 1039
+    },
+    {
+      "epoch": 0.2826086956521739,
+      "grad_norm": 2.0362656116485596,
+      "learning_rate": 3e-05,
+      "loss": 3.6416,
+      "step": 1040
+    },
+    {
+      "epoch": 0.2828804347826087,
+      "grad_norm": 1.493579387664795,
+      "learning_rate": 3e-05,
+      "loss": 3.2111,
+      "step": 1041
+    },
+    {
+      "epoch": 0.2831521739130435,
+      "grad_norm": 1.2144675254821777,
+      "learning_rate": 3e-05,
+      "loss": 2.911,
+      "step": 1042
+    },
+    {
+      "epoch": 0.28342391304347825,
+      "grad_norm": 1.833510398864746,
+      "learning_rate": 3e-05,
+      "loss": 3.0022,
+      "step": 1043
+    },
+    {
+      "epoch": 0.28369565217391307,
+      "grad_norm": 2.0041656494140625,
+      "learning_rate": 3e-05,
+      "loss": 3.1412,
+      "step": 1044
+    },
+    {
+      "epoch": 0.28396739130434784,
+      "grad_norm": 1.2218122482299805,
+      "learning_rate": 3e-05,
+      "loss": 3.4581,
+      "step": 1045
+    },
+    {
+      "epoch": 0.2842391304347826,
+      "grad_norm": 1.6633751392364502,
+      "learning_rate": 3e-05,
+      "loss": 3.6047,
+      "step": 1046
+    },
+    {
+      "epoch": 0.2845108695652174,
+      "grad_norm": 2.4305849075317383,
+      "learning_rate": 3e-05,
+      "loss": 3.3797,
+      "step": 1047
+    },
+    {
+      "epoch": 0.2847826086956522,
+      "grad_norm": 2.309483289718628,
+      "learning_rate": 3e-05,
+      "loss": 3.6304,
+      "step": 1048
+    },
+    {
+      "epoch": 0.28505434782608696,
+      "grad_norm": 1.678916335105896,
+      "learning_rate": 3e-05,
+      "loss": 3.4073,
+      "step": 1049
+    },
+    {
+      "epoch": 0.28532608695652173,
+      "grad_norm": 1.9873173236846924,
+      "learning_rate": 3e-05,
+      "loss": 3.9298,
+      "step": 1050
+    },
+    {
+      "epoch": 0.2855978260869565,
+      "grad_norm": 2.085996389389038,
+      "learning_rate": 3e-05,
+      "loss": 4.4263,
+      "step": 1051
+    },
+    {
+      "epoch": 0.2858695652173913,
+      "grad_norm": 3.049558639526367,
+      "learning_rate": 3e-05,
+      "loss": 3.6507,
+      "step": 1052
+    },
+    {
+      "epoch": 0.2861413043478261,
+      "grad_norm": 1.4257125854492188,
+      "learning_rate": 3e-05,
+      "loss": 3.3861,
+      "step": 1053
+    },
+    {
+      "epoch": 0.28641304347826085,
+      "grad_norm": 1.5436605215072632,
+      "learning_rate": 3e-05,
+      "loss": 3.1414,
+      "step": 1054
+    },
+    {
+      "epoch": 0.2866847826086957,
+      "grad_norm": 3.0967252254486084,
+      "learning_rate": 3e-05,
+      "loss": 3.9067,
+      "step": 1055
+    },
+    {
+      "epoch": 0.28695652173913044,
+      "grad_norm": 1.6564278602600098,
+      "learning_rate": 3e-05,
+      "loss": 3.5354,
+      "step": 1056
+    },
+    {
+      "epoch": 0.2872282608695652,
+      "grad_norm": 2.498901605606079,
+      "learning_rate": 3e-05,
+      "loss": 3.3153,
+      "step": 1057
+    },
+    {
+      "epoch": 0.2875,
+      "grad_norm": 2.2996490001678467,
+      "learning_rate": 3e-05,
+      "loss": 3.4023,
+      "step": 1058
+    },
+    {
+      "epoch": 0.2877717391304348,
+      "grad_norm": 2.0972297191619873,
+      "learning_rate": 3e-05,
+      "loss": 3.038,
+      "step": 1059
+    },
+    {
+      "epoch": 0.28804347826086957,
+      "grad_norm": 1.221664547920227,
+      "learning_rate": 3e-05,
+      "loss": 3.0761,
+      "step": 1060
+    },
+    {
+      "epoch": 0.28831521739130433,
+      "grad_norm": 2.2344908714294434,
+      "learning_rate": 3e-05,
+      "loss": 2.8314,
+      "step": 1061
+    },
+    {
+      "epoch": 0.28858695652173916,
+      "grad_norm": 1.9931285381317139,
+      "learning_rate": 3e-05,
+      "loss": 3.3899,
+      "step": 1062
+    },
+    {
+      "epoch": 0.2888586956521739,
+      "grad_norm": 1.924612045288086,
+      "learning_rate": 3e-05,
+      "loss": 3.4548,
+      "step": 1063
+    },
+    {
+      "epoch": 0.2891304347826087,
+      "grad_norm": 1.9507229328155518,
+      "learning_rate": 3e-05,
+      "loss": 3.9302,
+      "step": 1064
+    },
+    {
+      "epoch": 0.28940217391304346,
+      "grad_norm": 1.8761188983917236,
+      "learning_rate": 3e-05,
+      "loss": 3.195,
+      "step": 1065
+    },
+    {
+      "epoch": 0.2896739130434783,
+      "grad_norm": 2.233293294906616,
+      "learning_rate": 3e-05,
+      "loss": 3.1672,
+      "step": 1066
+    },
+    {
+      "epoch": 0.28994565217391305,
+      "grad_norm": 1.7154324054718018,
+      "learning_rate": 3e-05,
+      "loss": 3.1413,
+      "step": 1067
+    },
+    {
+      "epoch": 0.2902173913043478,
+      "grad_norm": 1.8050798177719116,
+      "learning_rate": 3e-05,
+      "loss": 3.4341,
+      "step": 1068
+    },
+    {
+      "epoch": 0.2904891304347826,
+      "grad_norm": 1.838165044784546,
+      "learning_rate": 3e-05,
+      "loss": 3.3827,
+      "step": 1069
+    },
+    {
+      "epoch": 0.2907608695652174,
+      "grad_norm": 2.7229554653167725,
+      "learning_rate": 3e-05,
+      "loss": 3.3938,
+      "step": 1070
+    },
+    {
+      "epoch": 0.2910326086956522,
+      "grad_norm": 3.150635242462158,
+      "learning_rate": 3e-05,
+      "loss": 3.8396,
+      "step": 1071
+    },
+    {
+      "epoch": 0.29130434782608694,
+      "grad_norm": 2.2429847717285156,
+      "learning_rate": 3e-05,
+      "loss": 3.9437,
+      "step": 1072
+    },
+    {
+      "epoch": 0.29157608695652176,
+      "grad_norm": 1.305802345275879,
+      "learning_rate": 3e-05,
+      "loss": 3.5962,
+      "step": 1073
+    },
+    {
+      "epoch": 0.29184782608695653,
+      "grad_norm": 1.4509013891220093,
+      "learning_rate": 3e-05,
+      "loss": 3.2975,
+      "step": 1074
+    },
+    {
+      "epoch": 0.2921195652173913,
+      "grad_norm": 2.059382200241089,
+      "learning_rate": 3e-05,
+      "loss": 4.21,
+      "step": 1075
+    },
+    {
+      "epoch": 0.29239130434782606,
+      "grad_norm": 0.9824801087379456,
+      "learning_rate": 3e-05,
+      "loss": 2.6315,
+      "step": 1076
+    },
+    {
+      "epoch": 0.2926630434782609,
+      "grad_norm": 1.0867946147918701,
+      "learning_rate": 3e-05,
+      "loss": 2.6838,
+      "step": 1077
+    },
+    {
+      "epoch": 0.29293478260869565,
+      "grad_norm": 1.992459774017334,
+      "learning_rate": 3e-05,
+      "loss": 4.2482,
+      "step": 1078
+    },
+    {
+      "epoch": 0.2932065217391304,
+      "grad_norm": 1.6963497400283813,
+      "learning_rate": 3e-05,
+      "loss": 3.5684,
+      "step": 1079
+    },
+    {
+      "epoch": 0.29347826086956524,
+      "grad_norm": 1.2850168943405151,
+      "learning_rate": 3e-05,
+      "loss": 3.399,
+      "step": 1080
+    },
+    {
+      "epoch": 0.29375,
+      "grad_norm": 1.652712106704712,
+      "learning_rate": 3e-05,
+      "loss": 3.3213,
+      "step": 1081
+    },
+    {
+      "epoch": 0.2940217391304348,
+      "grad_norm": 1.733959436416626,
+      "learning_rate": 3e-05,
+      "loss": 3.1942,
+      "step": 1082
+    },
+    {
+      "epoch": 0.29429347826086955,
+      "grad_norm": 1.8719747066497803,
+      "learning_rate": 3e-05,
+      "loss": 3.8789,
+      "step": 1083
+    },
+    {
+      "epoch": 0.29456521739130437,
+      "grad_norm": 1.5426567792892456,
+      "learning_rate": 3e-05,
+      "loss": 3.1319,
+      "step": 1084
+    },
+    {
+      "epoch": 0.29483695652173914,
+      "grad_norm": 2.0286598205566406,
+      "learning_rate": 3e-05,
+      "loss": 3.9002,
+      "step": 1085
+    },
+    {
+      "epoch": 0.2951086956521739,
+      "grad_norm": 1.5031648874282837,
+      "learning_rate": 3e-05,
+      "loss": 3.3979,
+      "step": 1086
+    },
+    {
+      "epoch": 0.29538043478260867,
+      "grad_norm": 1.556738257408142,
+      "learning_rate": 3e-05,
+      "loss": 3.9843,
+      "step": 1087
+    },
+    {
+      "epoch": 0.2956521739130435,
+      "grad_norm": 1.8954527378082275,
+      "learning_rate": 3e-05,
+      "loss": 4.6858,
+      "step": 1088
+    },
+    {
+      "epoch": 0.29592391304347826,
+      "grad_norm": 1.7330039739608765,
+      "learning_rate": 3e-05,
+      "loss": 2.8431,
+      "step": 1089
+    },
+    {
+      "epoch": 0.296195652173913,
+      "grad_norm": 2.0785205364227295,
+      "learning_rate": 3e-05,
+      "loss": 3.6168,
+      "step": 1090
+    },
+    {
+      "epoch": 0.29646739130434785,
+      "grad_norm": 1.708628535270691,
+      "learning_rate": 3e-05,
+      "loss": 3.6387,
+      "step": 1091
+    },
+    {
+      "epoch": 0.2967391304347826,
+      "grad_norm": 2.5519232749938965,
+      "learning_rate": 3e-05,
+      "loss": 4.0809,
+      "step": 1092
+    },
+    {
+      "epoch": 0.2970108695652174,
+      "grad_norm": 1.5967369079589844,
+      "learning_rate": 3e-05,
+      "loss": 3.5908,
+      "step": 1093
+    },
+    {
+      "epoch": 0.29728260869565215,
+      "grad_norm": 1.345123052597046,
+      "learning_rate": 3e-05,
+      "loss": 3.297,
+      "step": 1094
+    },
+    {
+      "epoch": 0.297554347826087,
+      "grad_norm": 2.123605251312256,
+      "learning_rate": 3e-05,
+      "loss": 4.1267,
+      "step": 1095
+    },
+    {
+      "epoch": 0.29782608695652174,
+      "grad_norm": 1.6863003969192505,
+      "learning_rate": 3e-05,
+      "loss": 3.4321,
+      "step": 1096
+    },
+    {
+      "epoch": 0.2980978260869565,
+      "grad_norm": 2.0726163387298584,
+      "learning_rate": 3e-05,
+      "loss": 3.386,
+      "step": 1097
+    },
+    {
+      "epoch": 0.29836956521739133,
+      "grad_norm": 1.5322779417037964,
+      "learning_rate": 3e-05,
+      "loss": 2.9161,
+      "step": 1098
+    },
+    {
+      "epoch": 0.2986413043478261,
+      "grad_norm": 1.1131296157836914,
+      "learning_rate": 3e-05,
+      "loss": 2.7584,
+      "step": 1099
+    },
+    {
+      "epoch": 0.29891304347826086,
+      "grad_norm": 1.9430760145187378,
+      "learning_rate": 3e-05,
+      "loss": 3.4661,
+      "step": 1100
+    },
+    {
+      "epoch": 0.29918478260869563,
+      "grad_norm": 2.6745107173919678,
+      "learning_rate": 3e-05,
+      "loss": 3.0976,
+      "step": 1101
+    },
+    {
+      "epoch": 0.29945652173913045,
+      "grad_norm": 3.127723217010498,
+      "learning_rate": 3e-05,
+      "loss": 3.7013,
+      "step": 1102
+    },
+    {
+      "epoch": 0.2997282608695652,
+      "grad_norm": 1.894990086555481,
+      "learning_rate": 3e-05,
+      "loss": 3.4841,
+      "step": 1103
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9465398788452148,
+      "learning_rate": 3e-05,
+      "loss": 3.48,
+      "step": 1104
+    },
+    {
+      "epoch": 0.30027173913043476,
+      "grad_norm": 2.5459132194519043,
+      "learning_rate": 3e-05,
+      "loss": 3.239,
+      "step": 1105
+    },
+    {
+      "epoch": 0.3005434782608696,
+      "grad_norm": 2.382801055908203,
+      "learning_rate": 3e-05,
+      "loss": 3.2947,
+      "step": 1106
+    },
+    {
+      "epoch": 0.30081521739130435,
+      "grad_norm": 1.913734793663025,
+      "learning_rate": 3e-05,
+      "loss": 3.0162,
+      "step": 1107
+    },
+    {
+      "epoch": 0.3010869565217391,
+      "grad_norm": 1.1874632835388184,
+      "learning_rate": 3e-05,
+      "loss": 2.6723,
+      "step": 1108
+    },
+    {
+      "epoch": 0.30135869565217394,
+      "grad_norm": 1.6929433345794678,
+      "learning_rate": 3e-05,
+      "loss": 3.1864,
+      "step": 1109
+    },
+    {
+      "epoch": 0.3016304347826087,
+      "grad_norm": 2.1046361923217773,
+      "learning_rate": 3e-05,
+      "loss": 3.442,
+      "step": 1110
+    },
+    {
+      "epoch": 0.30190217391304347,
+      "grad_norm": 2.196091413497925,
+      "learning_rate": 3e-05,
+      "loss": 3.2791,
+      "step": 1111
+    },
+    {
+      "epoch": 0.30217391304347824,
+      "grad_norm": 2.6548757553100586,
+      "learning_rate": 3e-05,
+      "loss": 4.0965,
+      "step": 1112
+    },
+    {
+      "epoch": 0.30244565217391306,
+      "grad_norm": 1.7033164501190186,
+      "learning_rate": 3e-05,
+      "loss": 3.5031,
+      "step": 1113
+    },
+    {
+      "epoch": 0.3027173913043478,
+      "grad_norm": 1.5866395235061646,
+      "learning_rate": 3e-05,
+      "loss": 2.9015,
+      "step": 1114
+    },
+    {
+      "epoch": 0.3029891304347826,
+      "grad_norm": 2.0246646404266357,
+      "learning_rate": 3e-05,
+      "loss": 3.1981,
+      "step": 1115
+    },
+    {
+      "epoch": 0.3032608695652174,
+      "grad_norm": 2.0733206272125244,
+      "learning_rate": 3e-05,
+      "loss": 3.3524,
+      "step": 1116
+    },
+    {
+      "epoch": 0.3035326086956522,
+      "grad_norm": 1.6320792436599731,
+      "learning_rate": 3e-05,
+      "loss": 3.6667,
+      "step": 1117
+    },
+    {
+      "epoch": 0.30380434782608695,
+      "grad_norm": 1.4059102535247803,
+      "learning_rate": 3e-05,
+      "loss": 3.4399,
+      "step": 1118
+    },
+    {
+      "epoch": 0.3040760869565217,
+      "grad_norm": 1.6570987701416016,
+      "learning_rate": 3e-05,
+      "loss": 3.2745,
+      "step": 1119
+    },
+    {
+      "epoch": 0.30434782608695654,
+      "grad_norm": 1.9712703227996826,
+      "learning_rate": 3e-05,
+      "loss": 3.751,
+      "step": 1120
+    },
+    {
+      "epoch": 0.3046195652173913,
+      "grad_norm": 1.4697049856185913,
+      "learning_rate": 3e-05,
+      "loss": 3.4402,
+      "step": 1121
+    },
+    {
+      "epoch": 0.3048913043478261,
+      "grad_norm": 1.8021398782730103,
+      "learning_rate": 3e-05,
+      "loss": 3.8907,
+      "step": 1122
+    },
+    {
+      "epoch": 0.30516304347826084,
+      "grad_norm": 1.5529664754867554,
+      "learning_rate": 3e-05,
+      "loss": 3.7741,
+      "step": 1123
+    },
+    {
+      "epoch": 0.30543478260869567,
+      "grad_norm": 1.5865482091903687,
+      "learning_rate": 3e-05,
+      "loss": 3.7243,
+      "step": 1124
+    },
+    {
+      "epoch": 0.30570652173913043,
+      "grad_norm": 2.221069097518921,
+      "learning_rate": 3e-05,
+      "loss": 3.65,
+      "step": 1125
+    },
+    {
+      "epoch": 0.3059782608695652,
+      "grad_norm": 2.1728017330169678,
+      "learning_rate": 3e-05,
+      "loss": 3.0357,
+      "step": 1126
+    },
+    {
+      "epoch": 0.30625,
+      "grad_norm": 1.6755625009536743,
+      "learning_rate": 3e-05,
+      "loss": 3.2292,
+      "step": 1127
+    },
+    {
+      "epoch": 0.3065217391304348,
+      "grad_norm": 2.135910987854004,
+      "learning_rate": 3e-05,
+      "loss": 4.0064,
+      "step": 1128
+    },
+    {
+      "epoch": 0.30679347826086956,
+      "grad_norm": 1.7224137783050537,
+      "learning_rate": 3e-05,
+      "loss": 3.6076,
+      "step": 1129
+    },
+    {
+      "epoch": 0.3070652173913043,
+      "grad_norm": 1.5196242332458496,
+      "learning_rate": 3e-05,
+      "loss": 3.2132,
+      "step": 1130
+    },
+    {
+      "epoch": 0.30733695652173915,
+      "grad_norm": 2.016763210296631,
+      "learning_rate": 3e-05,
+      "loss": 4.1373,
+      "step": 1131
+    },
+    {
+      "epoch": 0.3076086956521739,
+      "grad_norm": 1.3778748512268066,
+      "learning_rate": 3e-05,
+      "loss": 3.4815,
+      "step": 1132
+    },
+    {
+      "epoch": 0.3078804347826087,
+      "grad_norm": 1.5161099433898926,
+      "learning_rate": 3e-05,
+      "loss": 3.2094,
+      "step": 1133
+    },
+    {
+      "epoch": 0.3081521739130435,
+      "grad_norm": 1.2191746234893799,
+      "learning_rate": 3e-05,
+      "loss": 3.0117,
+      "step": 1134
+    },
+    {
+      "epoch": 0.30842391304347827,
+      "grad_norm": 1.6270561218261719,
+      "learning_rate": 3e-05,
+      "loss": 3.8738,
+      "step": 1135
+    },
+    {
+      "epoch": 0.30869565217391304,
+      "grad_norm": 1.3209682703018188,
+      "learning_rate": 3e-05,
+      "loss": 3.1873,
+      "step": 1136
+    },
+    {
+      "epoch": 0.3089673913043478,
+      "grad_norm": 1.7960654497146606,
+      "learning_rate": 3e-05,
+      "loss": 3.9211,
+      "step": 1137
+    },
+    {
+      "epoch": 0.3092391304347826,
+      "grad_norm": 1.5645345449447632,
+      "learning_rate": 3e-05,
+      "loss": 3.8176,
+      "step": 1138
+    },
+    {
+      "epoch": 0.3095108695652174,
+      "grad_norm": 1.2274796962738037,
+      "learning_rate": 3e-05,
+      "loss": 3.0552,
+      "step": 1139
+    },
+    {
+      "epoch": 0.30978260869565216,
+      "grad_norm": 3.1637630462646484,
+      "learning_rate": 3e-05,
+      "loss": 4.0908,
+      "step": 1140
+    },
+    {
+      "epoch": 0.31005434782608693,
+      "grad_norm": 1.7452456951141357,
+      "learning_rate": 3e-05,
+      "loss": 3.2379,
+      "step": 1141
+    },
+    {
+      "epoch": 0.31032608695652175,
+      "grad_norm": 1.3035019636154175,
+      "learning_rate": 3e-05,
+      "loss": 3.1619,
+      "step": 1142
+    },
+    {
+      "epoch": 0.3105978260869565,
+      "grad_norm": 1.5721486806869507,
+      "learning_rate": 3e-05,
+      "loss": 3.3092,
+      "step": 1143
+    },
+    {
+      "epoch": 0.3108695652173913,
+      "grad_norm": 1.5714532136917114,
+      "learning_rate": 3e-05,
+      "loss": 3.7343,
+      "step": 1144
+    },
+    {
+      "epoch": 0.3111413043478261,
+      "grad_norm": 1.4483362436294556,
+      "learning_rate": 3e-05,
+      "loss": 3.5843,
+      "step": 1145
+    },
+    {
+      "epoch": 0.3114130434782609,
+      "grad_norm": 1.2886680364608765,
+      "learning_rate": 3e-05,
+      "loss": 3.4743,
+      "step": 1146
+    },
+    {
+      "epoch": 0.31168478260869564,
+      "grad_norm": 1.5609358549118042,
+      "learning_rate": 3e-05,
+      "loss": 3.7947,
+      "step": 1147
+    },
+    {
+      "epoch": 0.3119565217391304,
+      "grad_norm": 1.7903861999511719,
+      "learning_rate": 3e-05,
+      "loss": 3.3165,
+      "step": 1148
+    },
+    {
+      "epoch": 0.31222826086956523,
+      "grad_norm": 1.917932152748108,
+      "learning_rate": 3e-05,
+      "loss": 3.3985,
+      "step": 1149
+    },
+    {
+      "epoch": 0.3125,
+      "grad_norm": 1.5185085535049438,
+      "learning_rate": 3e-05,
+      "loss": 3.571,
+      "step": 1150
+    },
+    {
+      "epoch": 0.31277173913043477,
+      "grad_norm": 1.3498207330703735,
+      "learning_rate": 3e-05,
+      "loss": 3.3921,
+      "step": 1151
+    },
+    {
+      "epoch": 0.3130434782608696,
+      "grad_norm": 1.6622629165649414,
+      "learning_rate": 3e-05,
+      "loss": 3.4101,
+      "step": 1152
+    },
+    {
+      "epoch": 0.31331521739130436,
+      "grad_norm": 1.516265630722046,
+      "learning_rate": 3e-05,
+      "loss": 3.1228,
+      "step": 1153
+    },
+    {
+      "epoch": 0.3135869565217391,
+      "grad_norm": 1.5660970211029053,
+      "learning_rate": 3e-05,
+      "loss": 3.3751,
+      "step": 1154
+    },
+    {
+      "epoch": 0.3138586956521739,
+      "grad_norm": 1.8406776189804077,
+      "learning_rate": 3e-05,
+      "loss": 3.3181,
+      "step": 1155
+    },
+    {
+      "epoch": 0.3141304347826087,
+      "grad_norm": 1.8162765502929688,
+      "learning_rate": 3e-05,
+      "loss": 4.0117,
+      "step": 1156
+    },
+    {
+      "epoch": 0.3144021739130435,
+      "grad_norm": 1.2983214855194092,
+      "learning_rate": 3e-05,
+      "loss": 3.2412,
+      "step": 1157
+    },
+    {
+      "epoch": 0.31467391304347825,
+      "grad_norm": 1.1409294605255127,
+      "learning_rate": 3e-05,
+      "loss": 3.0329,
+      "step": 1158
+    },
+    {
+      "epoch": 0.31494565217391307,
+      "grad_norm": 1.754345178604126,
+      "learning_rate": 3e-05,
+      "loss": 3.6711,
+      "step": 1159
+    },
+    {
+      "epoch": 0.31521739130434784,
+      "grad_norm": 1.868580937385559,
+      "learning_rate": 3e-05,
+      "loss": 3.6684,
+      "step": 1160
+    },
+    {
+      "epoch": 0.3154891304347826,
+      "grad_norm": 1.8012367486953735,
+      "learning_rate": 3e-05,
+      "loss": 3.2884,
+      "step": 1161
+    },
+    {
+      "epoch": 0.3157608695652174,
+      "grad_norm": 1.7688064575195312,
+      "learning_rate": 3e-05,
+      "loss": 3.3512,
+      "step": 1162
+    },
+    {
+      "epoch": 0.3160326086956522,
+      "grad_norm": 1.5674185752868652,
+      "learning_rate": 3e-05,
+      "loss": 3.443,
+      "step": 1163
+    },
+    {
+      "epoch": 0.31630434782608696,
+      "grad_norm": 1.5843913555145264,
+      "learning_rate": 3e-05,
+      "loss": 3.8519,
+      "step": 1164
+    },
+    {
+      "epoch": 0.31657608695652173,
+      "grad_norm": 1.4522455930709839,
+      "learning_rate": 3e-05,
+      "loss": 3.123,
+      "step": 1165
+    },
+    {
+      "epoch": 0.3168478260869565,
+      "grad_norm": 1.5610785484313965,
+      "learning_rate": 3e-05,
+      "loss": 3.1963,
+      "step": 1166
+    },
+    {
+      "epoch": 0.3171195652173913,
+      "grad_norm": 1.4327051639556885,
+      "learning_rate": 3e-05,
+      "loss": 2.9382,
+      "step": 1167
+    },
+    {
+      "epoch": 0.3173913043478261,
+      "grad_norm": 1.6065129041671753,
+      "learning_rate": 3e-05,
+      "loss": 3.4928,
+      "step": 1168
+    },
+    {
+      "epoch": 0.31766304347826085,
+      "grad_norm": 1.9382498264312744,
+      "learning_rate": 3e-05,
+      "loss": 4.3174,
+      "step": 1169
+    },
+    {
+      "epoch": 0.3179347826086957,
+      "grad_norm": 1.164689064025879,
+      "learning_rate": 3e-05,
+      "loss": 3.0678,
+      "step": 1170
+    },
+    {
+      "epoch": 0.31820652173913044,
+      "grad_norm": 1.084055781364441,
+      "learning_rate": 3e-05,
+      "loss": 3.0878,
+      "step": 1171
+    },
+    {
+      "epoch": 0.3184782608695652,
+      "grad_norm": 1.4881168603897095,
+      "learning_rate": 3e-05,
+      "loss": 3.5084,
+      "step": 1172
+    },
+    {
+      "epoch": 0.31875,
+      "grad_norm": 1.5508403778076172,
+      "learning_rate": 3e-05,
+      "loss": 3.6359,
+      "step": 1173
+    },
+    {
+      "epoch": 0.3190217391304348,
+      "grad_norm": 1.7135038375854492,
+      "learning_rate": 3e-05,
+      "loss": 2.819,
+      "step": 1174
+    },
+    {
+      "epoch": 0.31929347826086957,
+      "grad_norm": 1.9360867738723755,
+      "learning_rate": 3e-05,
+      "loss": 3.711,
+      "step": 1175
+    },
+    {
+      "epoch": 0.31956521739130433,
+      "grad_norm": 1.3898394107818604,
+      "learning_rate": 3e-05,
+      "loss": 3.0356,
+      "step": 1176
+    },
+    {
+      "epoch": 0.31983695652173916,
+      "grad_norm": 1.37100088596344,
+      "learning_rate": 3e-05,
+      "loss": 2.9441,
+      "step": 1177
+    },
+    {
+      "epoch": 0.3201086956521739,
+      "grad_norm": 1.7066446542739868,
+      "learning_rate": 3e-05,
+      "loss": 3.8696,
+      "step": 1178
+    },
+    {
+      "epoch": 0.3203804347826087,
+      "grad_norm": 1.4085134267807007,
+      "learning_rate": 3e-05,
+      "loss": 3.4616,
+      "step": 1179
+    },
+    {
+      "epoch": 0.32065217391304346,
+      "grad_norm": 1.5345913171768188,
+      "learning_rate": 3e-05,
+      "loss": 3.6033,
+      "step": 1180
+    },
+    {
+      "epoch": 0.3209239130434783,
+      "grad_norm": 1.5240819454193115,
+      "learning_rate": 3e-05,
+      "loss": 3.1863,
+      "step": 1181
+    },
+    {
+      "epoch": 0.32119565217391305,
+      "grad_norm": 1.147559404373169,
+      "learning_rate": 3e-05,
+      "loss": 3.2533,
+      "step": 1182
+    },
+    {
+      "epoch": 0.3214673913043478,
+      "grad_norm": 1.3267796039581299,
+      "learning_rate": 3e-05,
+      "loss": 3.3342,
+      "step": 1183
+    },
+    {
+      "epoch": 0.3217391304347826,
+      "grad_norm": 1.567126989364624,
+      "learning_rate": 3e-05,
+      "loss": 3.5055,
+      "step": 1184
+    },
+    {
+      "epoch": 0.3220108695652174,
+      "grad_norm": 2.085658073425293,
+      "learning_rate": 3e-05,
+      "loss": 3.1383,
+      "step": 1185
+    },
+    {
+      "epoch": 0.3222826086956522,
+      "grad_norm": 1.7624791860580444,
+      "learning_rate": 3e-05,
+      "loss": 3.2409,
+      "step": 1186
+    },
+    {
+      "epoch": 0.32255434782608694,
+      "grad_norm": 1.4368826150894165,
+      "learning_rate": 3e-05,
+      "loss": 3.7424,
+      "step": 1187
+    },
+    {
+      "epoch": 0.32282608695652176,
+      "grad_norm": 1.5510128736495972,
+      "learning_rate": 3e-05,
+      "loss": 3.3895,
+      "step": 1188
+    },
+    {
+      "epoch": 0.32309782608695653,
+      "grad_norm": 1.471490740776062,
+      "learning_rate": 3e-05,
+      "loss": 3.6772,
+      "step": 1189
+    },
+    {
+      "epoch": 0.3233695652173913,
+      "grad_norm": 1.2040995359420776,
+      "learning_rate": 3e-05,
+      "loss": 3.3302,
+      "step": 1190
+    },
+    {
+      "epoch": 0.32364130434782606,
+      "grad_norm": 1.4731234312057495,
+      "learning_rate": 3e-05,
+      "loss": 3.3842,
+      "step": 1191
+    },
+    {
+      "epoch": 0.3239130434782609,
+      "grad_norm": 1.3904873132705688,
+      "learning_rate": 3e-05,
+      "loss": 3.8537,
+      "step": 1192
+    },
+    {
+      "epoch": 0.32418478260869565,
+      "grad_norm": 1.187960147857666,
+      "learning_rate": 3e-05,
+      "loss": 3.0656,
+      "step": 1193
+    },
+    {
+      "epoch": 0.3244565217391304,
+      "grad_norm": 1.1819937229156494,
+      "learning_rate": 3e-05,
+      "loss": 2.7234,
+      "step": 1194
+    },
+    {
+      "epoch": 0.32472826086956524,
+      "grad_norm": 1.6428884267807007,
+      "learning_rate": 3e-05,
+      "loss": 3.4446,
+      "step": 1195
+    },
+    {
+      "epoch": 0.325,
+      "grad_norm": 1.6877864599227905,
+      "learning_rate": 3e-05,
+      "loss": 3.2674,
+      "step": 1196
+    },
+    {
+      "epoch": 0.3252717391304348,
+      "grad_norm": 1.596166729927063,
+      "learning_rate": 3e-05,
+      "loss": 3.371,
+      "step": 1197
+    },
+    {
+      "epoch": 0.32554347826086955,
+      "grad_norm": 1.3886607885360718,
+      "learning_rate": 3e-05,
+      "loss": 3.3149,
+      "step": 1198
+    },
+    {
+      "epoch": 0.32581521739130437,
+      "grad_norm": 1.4295148849487305,
+      "learning_rate": 3e-05,
+      "loss": 3.0519,
+      "step": 1199
+    },
+    {
+      "epoch": 0.32608695652173914,
+      "grad_norm": 1.5317387580871582,
+      "learning_rate": 3e-05,
+      "loss": 3.3816,
+      "step": 1200
+    },
+    {
+      "epoch": 0.3263586956521739,
+      "grad_norm": 1.1524282693862915,
+      "learning_rate": 3e-05,
+      "loss": 3.2564,
+      "step": 1201
+    },
+    {
+      "epoch": 0.32663043478260867,
+      "grad_norm": 1.3997209072113037,
+      "learning_rate": 3e-05,
+      "loss": 3.3173,
+      "step": 1202
+    },
+    {
+      "epoch": 0.3269021739130435,
+      "grad_norm": 1.357279896736145,
+      "learning_rate": 3e-05,
+      "loss": 3.7316,
+      "step": 1203
+    },
+    {
+      "epoch": 0.32717391304347826,
+      "grad_norm": 1.1962755918502808,
+      "learning_rate": 3e-05,
+      "loss": 3.5203,
+      "step": 1204
+    },
+    {
+      "epoch": 0.327445652173913,
+      "grad_norm": 1.4180536270141602,
+      "learning_rate": 3e-05,
+      "loss": 3.7679,
+      "step": 1205
+    },
+    {
+      "epoch": 0.32771739130434785,
+      "grad_norm": 1.4997798204421997,
+      "learning_rate": 3e-05,
+      "loss": 3.9842,
+      "step": 1206
+    },
+    {
+      "epoch": 0.3279891304347826,
+      "grad_norm": 1.5224746465682983,
+      "learning_rate": 3e-05,
+      "loss": 3.723,
+      "step": 1207
+    },
+    {
+      "epoch": 0.3282608695652174,
+      "grad_norm": 1.4814677238464355,
+      "learning_rate": 3e-05,
+      "loss": 3.3444,
+      "step": 1208
+    },
+    {
+      "epoch": 0.32853260869565215,
+      "grad_norm": 1.8869359493255615,
+      "learning_rate": 3e-05,
+      "loss": 3.8663,
+      "step": 1209
+    },
+    {
+      "epoch": 0.328804347826087,
+      "grad_norm": 1.2305731773376465,
+      "learning_rate": 3e-05,
+      "loss": 3.1855,
+      "step": 1210
+    },
+    {
+      "epoch": 0.32907608695652174,
+      "grad_norm": 1.3822002410888672,
+      "learning_rate": 3e-05,
+      "loss": 3.9625,
+      "step": 1211
+    },
+    {
+      "epoch": 0.3293478260869565,
+      "grad_norm": 1.2895426750183105,
+      "learning_rate": 3e-05,
+      "loss": 2.9574,
+      "step": 1212
+    },
+    {
+      "epoch": 0.32961956521739133,
+      "grad_norm": 1.2724965810775757,
+      "learning_rate": 3e-05,
+      "loss": 3.2984,
+      "step": 1213
+    },
+    {
+      "epoch": 0.3298913043478261,
+      "grad_norm": 1.569366455078125,
+      "learning_rate": 3e-05,
+      "loss": 3.4928,
+      "step": 1214
+    },
+    {
+      "epoch": 0.33016304347826086,
+      "grad_norm": 1.538090467453003,
+      "learning_rate": 3e-05,
+      "loss": 3.8567,
+      "step": 1215
+    },
+    {
+      "epoch": 0.33043478260869563,
+      "grad_norm": 1.3484816551208496,
+      "learning_rate": 3e-05,
+      "loss": 3.1096,
+      "step": 1216
+    },
+    {
+      "epoch": 0.33070652173913045,
+      "grad_norm": 1.1450049877166748,
+      "learning_rate": 3e-05,
+      "loss": 2.773,
+      "step": 1217
+    },
+    {
+      "epoch": 0.3309782608695652,
+      "grad_norm": 1.5292000770568848,
+      "learning_rate": 3e-05,
+      "loss": 3.3972,
+      "step": 1218
+    },
+    {
+      "epoch": 0.33125,
+      "grad_norm": 1.3615137338638306,
+      "learning_rate": 3e-05,
+      "loss": 3.2462,
+      "step": 1219
+    },
+    {
+      "epoch": 0.33152173913043476,
+      "grad_norm": 1.7036170959472656,
+      "learning_rate": 3e-05,
+      "loss": 3.7338,
+      "step": 1220
+    },
+    {
+      "epoch": 0.3317934782608696,
+      "grad_norm": 1.7132785320281982,
+      "learning_rate": 3e-05,
+      "loss": 3.3378,
+      "step": 1221
+    },
+    {
+      "epoch": 0.33206521739130435,
+      "grad_norm": 1.9641152620315552,
+      "learning_rate": 3e-05,
+      "loss": 3.8121,
+      "step": 1222
+    },
+    {
+      "epoch": 0.3323369565217391,
+      "grad_norm": 1.3222547769546509,
+      "learning_rate": 3e-05,
+      "loss": 3.2911,
+      "step": 1223
+    },
+    {
+      "epoch": 0.33260869565217394,
+      "grad_norm": 1.3695377111434937,
+      "learning_rate": 3e-05,
+      "loss": 3.3004,
+      "step": 1224
+    },
+    {
+      "epoch": 0.3328804347826087,
+      "grad_norm": 1.8578277826309204,
+      "learning_rate": 3e-05,
+      "loss": 4.1831,
+      "step": 1225
+    },
+    {
+      "epoch": 0.33315217391304347,
+      "grad_norm": 1.7893558740615845,
+      "learning_rate": 3e-05,
+      "loss": 3.359,
+      "step": 1226
+    },
+    {
+      "epoch": 0.33342391304347824,
+      "grad_norm": 1.474753499031067,
+      "learning_rate": 3e-05,
+      "loss": 2.9094,
+      "step": 1227
+    },
+    {
+      "epoch": 0.33369565217391306,
+      "grad_norm": 1.973402738571167,
+      "learning_rate": 3e-05,
+      "loss": 4.0505,
+      "step": 1228
+    },
+    {
+      "epoch": 0.3339673913043478,
+      "grad_norm": 1.7613177299499512,
+      "learning_rate": 3e-05,
+      "loss": 3.3559,
+      "step": 1229
+    },
+    {
+      "epoch": 0.3342391304347826,
+      "grad_norm": 2.031470775604248,
+      "learning_rate": 3e-05,
+      "loss": 3.3066,
+      "step": 1230
+    },
+    {
+      "epoch": 0.3345108695652174,
+      "grad_norm": 2.1547582149505615,
+      "learning_rate": 3e-05,
+      "loss": 3.6043,
+      "step": 1231
+    },
+    {
+      "epoch": 0.3347826086956522,
+      "grad_norm": 2.2331910133361816,
+      "learning_rate": 3e-05,
+      "loss": 2.9871,
+      "step": 1232
+    },
+    {
+      "epoch": 0.33505434782608695,
+      "grad_norm": 2.156770944595337,
+      "learning_rate": 3e-05,
+      "loss": 3.6777,
+      "step": 1233
+    },
+    {
+      "epoch": 0.3353260869565217,
+      "grad_norm": 1.717960000038147,
+      "learning_rate": 3e-05,
+      "loss": 3.6268,
+      "step": 1234
+    },
+    {
+      "epoch": 0.33559782608695654,
+      "grad_norm": 1.649511694908142,
+      "learning_rate": 3e-05,
+      "loss": 3.3226,
+      "step": 1235
+    },
+    {
+      "epoch": 0.3358695652173913,
+      "grad_norm": 1.8623225688934326,
+      "learning_rate": 3e-05,
+      "loss": 3.9329,
+      "step": 1236
+    },
+    {
+      "epoch": 0.3361413043478261,
+      "grad_norm": 2.167483329772949,
+      "learning_rate": 3e-05,
+      "loss": 3.445,
+      "step": 1237
+    },
+    {
+      "epoch": 0.33641304347826084,
+      "grad_norm": 1.7240395545959473,
+      "learning_rate": 3e-05,
+      "loss": 3.493,
+      "step": 1238
+    },
+    {
+      "epoch": 0.33668478260869567,
+      "grad_norm": 1.3711011409759521,
+      "learning_rate": 3e-05,
+      "loss": 2.7254,
+      "step": 1239
+    },
+    {
+      "epoch": 0.33695652173913043,
+      "grad_norm": 2.091745615005493,
+      "learning_rate": 3e-05,
+      "loss": 4.0388,
+      "step": 1240
+    },
+    {
+      "epoch": 0.3372282608695652,
+      "grad_norm": 1.5960465669631958,
+      "learning_rate": 3e-05,
+      "loss": 3.5036,
+      "step": 1241
+    },
+    {
+      "epoch": 0.3375,
+      "grad_norm": 1.1491891145706177,
+      "learning_rate": 3e-05,
+      "loss": 3.2282,
+      "step": 1242
+    },
+    {
+      "epoch": 0.3377717391304348,
+      "grad_norm": 1.1655139923095703,
+      "learning_rate": 3e-05,
+      "loss": 3.3578,
+      "step": 1243
+    },
+    {
+      "epoch": 0.33804347826086956,
+      "grad_norm": 1.4522464275360107,
+      "learning_rate": 3e-05,
+      "loss": 3.4827,
+      "step": 1244
+    },
+    {
+      "epoch": 0.3383152173913043,
+      "grad_norm": 1.494310736656189,
+      "learning_rate": 3e-05,
+      "loss": 3.7874,
+      "step": 1245
+    },
+    {
+      "epoch": 0.33858695652173915,
+      "grad_norm": 1.98121976852417,
+      "learning_rate": 3e-05,
+      "loss": 3.2323,
+      "step": 1246
+    },
+    {
+      "epoch": 0.3388586956521739,
+      "grad_norm": 1.772510290145874,
+      "learning_rate": 3e-05,
+      "loss": 2.9783,
+      "step": 1247
+    },
+    {
+      "epoch": 0.3391304347826087,
+      "grad_norm": 1.3050928115844727,
+      "learning_rate": 3e-05,
+      "loss": 3.3902,
+      "step": 1248
+    },
+    {
+      "epoch": 0.3394021739130435,
+      "grad_norm": 1.590279221534729,
+      "learning_rate": 3e-05,
+      "loss": 3.0477,
+      "step": 1249
+    },
+    {
+      "epoch": 0.33967391304347827,
+      "grad_norm": 2.1541528701782227,
+      "learning_rate": 3e-05,
+      "loss": 3.5042,
+      "step": 1250
+    },
+    {
+      "epoch": 0.33994565217391304,
+      "grad_norm": 2.138265609741211,
+      "learning_rate": 3e-05,
+      "loss": 3.704,
+      "step": 1251
+    },
+    {
+      "epoch": 0.3402173913043478,
+      "grad_norm": 1.5365831851959229,
+      "learning_rate": 3e-05,
+      "loss": 3.3562,
+      "step": 1252
+    },
+    {
+      "epoch": 0.3404891304347826,
+      "grad_norm": 1.3827742338180542,
+      "learning_rate": 3e-05,
+      "loss": 3.2202,
+      "step": 1253
+    },
+    {
+      "epoch": 0.3407608695652174,
+      "grad_norm": 1.4450342655181885,
+      "learning_rate": 3e-05,
+      "loss": 3.4887,
+      "step": 1254
+    },
+    {
+      "epoch": 0.34103260869565216,
+      "grad_norm": 1.2322391271591187,
+      "learning_rate": 3e-05,
+      "loss": 2.7866,
+      "step": 1255
+    },
+    {
+      "epoch": 0.34130434782608693,
+      "grad_norm": 1.3767555952072144,
+      "learning_rate": 3e-05,
+      "loss": 2.8691,
+      "step": 1256
+    },
+    {
+      "epoch": 0.34157608695652175,
+      "grad_norm": 1.9902418851852417,
+      "learning_rate": 3e-05,
+      "loss": 3.5811,
+      "step": 1257
+    },
+    {
+      "epoch": 0.3418478260869565,
+      "grad_norm": 1.7038296461105347,
+      "learning_rate": 3e-05,
+      "loss": 3.363,
+      "step": 1258
+    },
+    {
+      "epoch": 0.3421195652173913,
+      "grad_norm": 1.82673978805542,
+      "learning_rate": 3e-05,
+      "loss": 3.0324,
+      "step": 1259
+    },
+    {
+      "epoch": 0.3423913043478261,
+      "grad_norm": 1.4924451112747192,
+      "learning_rate": 3e-05,
+      "loss": 3.5862,
+      "step": 1260
+    },
+    {
+      "epoch": 0.3426630434782609,
+      "grad_norm": 1.1290408372879028,
+      "learning_rate": 3e-05,
+      "loss": 3.0983,
+      "step": 1261
+    },
+    {
+      "epoch": 0.34293478260869564,
+      "grad_norm": 1.3652050495147705,
+      "learning_rate": 3e-05,
+      "loss": 3.2806,
+      "step": 1262
+    },
+    {
+      "epoch": 0.3432065217391304,
+      "grad_norm": 1.3618534803390503,
+      "learning_rate": 3e-05,
+      "loss": 3.0708,
+      "step": 1263
+    },
+    {
+      "epoch": 0.34347826086956523,
+      "grad_norm": 1.3616082668304443,
+      "learning_rate": 3e-05,
+      "loss": 2.8319,
+      "step": 1264
+    },
+    {
+      "epoch": 0.34375,
+      "grad_norm": 1.3398898839950562,
+      "learning_rate": 3e-05,
+      "loss": 3.1065,
+      "step": 1265
+    },
+    {
+      "epoch": 0.34402173913043477,
+      "grad_norm": 1.6687146425247192,
+      "learning_rate": 3e-05,
+      "loss": 3.3584,
+      "step": 1266
+    },
+    {
+      "epoch": 0.3442934782608696,
+      "grad_norm": 1.2948143482208252,
+      "learning_rate": 3e-05,
+      "loss": 3.2489,
+      "step": 1267
+    },
+    {
+      "epoch": 0.34456521739130436,
+      "grad_norm": 1.7158524990081787,
+      "learning_rate": 3e-05,
+      "loss": 3.3558,
+      "step": 1268
+    },
+    {
+      "epoch": 0.3448369565217391,
+      "grad_norm": 1.6835163831710815,
+      "learning_rate": 3e-05,
+      "loss": 2.9967,
+      "step": 1269
+    },
+    {
+      "epoch": 0.3451086956521739,
+      "grad_norm": 1.971177101135254,
+      "learning_rate": 3e-05,
+      "loss": 3.829,
+      "step": 1270
+    },
+    {
+      "epoch": 0.3453804347826087,
+      "grad_norm": 1.660911202430725,
+      "learning_rate": 3e-05,
+      "loss": 2.9975,
+      "step": 1271
+    },
+    {
+      "epoch": 0.3456521739130435,
+      "grad_norm": 1.8289529085159302,
+      "learning_rate": 3e-05,
+      "loss": 3.4466,
+      "step": 1272
+    },
+    {
+      "epoch": 0.34592391304347825,
+      "grad_norm": 1.9580225944519043,
+      "learning_rate": 3e-05,
+      "loss": 3.4431,
+      "step": 1273
+    },
+    {
+      "epoch": 0.34619565217391307,
+      "grad_norm": 1.9081965684890747,
+      "learning_rate": 3e-05,
+      "loss": 3.1295,
+      "step": 1274
+    },
+    {
+      "epoch": 0.34646739130434784,
+      "grad_norm": 2.8098397254943848,
+      "learning_rate": 3e-05,
+      "loss": 3.9265,
+      "step": 1275
+    },
+    {
+      "epoch": 0.3467391304347826,
+      "grad_norm": 1.8996559381484985,
+      "learning_rate": 3e-05,
+      "loss": 2.9118,
+      "step": 1276
+    },
+    {
+      "epoch": 0.3470108695652174,
+      "grad_norm": 1.8447978496551514,
+      "learning_rate": 3e-05,
+      "loss": 3.8908,
+      "step": 1277
+    },
+    {
+      "epoch": 0.3472826086956522,
+      "grad_norm": 1.450376033782959,
+      "learning_rate": 3e-05,
+      "loss": 3.1262,
+      "step": 1278
+    },
+    {
+      "epoch": 0.34755434782608696,
+      "grad_norm": 2.3735806941986084,
+      "learning_rate": 3e-05,
+      "loss": 3.4959,
+      "step": 1279
+    },
+    {
+      "epoch": 0.34782608695652173,
+      "grad_norm": 2.2202954292297363,
+      "learning_rate": 3e-05,
+      "loss": 3.2127,
+      "step": 1280
+    },
+    {
+      "epoch": 0.3480978260869565,
+      "grad_norm": 1.6313713788986206,
+      "learning_rate": 3e-05,
+      "loss": 2.7593,
+      "step": 1281
+    },
+    {
+      "epoch": 0.3483695652173913,
+      "grad_norm": 1.8261613845825195,
+      "learning_rate": 3e-05,
+      "loss": 3.5207,
+      "step": 1282
+    },
+    {
+      "epoch": 0.3486413043478261,
+      "grad_norm": 1.6101502180099487,
+      "learning_rate": 3e-05,
+      "loss": 3.1187,
+      "step": 1283
+    },
+    {
+      "epoch": 0.34891304347826085,
+      "grad_norm": 1.4293962717056274,
+      "learning_rate": 3e-05,
+      "loss": 3.3783,
+      "step": 1284
+    },
+    {
+      "epoch": 0.3491847826086957,
+      "grad_norm": 1.335721731185913,
+      "learning_rate": 3e-05,
+      "loss": 3.8365,
+      "step": 1285
+    },
+    {
+      "epoch": 0.34945652173913044,
+      "grad_norm": 1.5614105463027954,
+      "learning_rate": 3e-05,
+      "loss": 3.7661,
+      "step": 1286
+    },
+    {
+      "epoch": 0.3497282608695652,
+      "grad_norm": 1.38046133518219,
+      "learning_rate": 3e-05,
+      "loss": 3.1717,
+      "step": 1287
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3673638105392456,
+      "learning_rate": 3e-05,
+      "loss": 3.7002,
+      "step": 1288
+    },
+    {
+      "epoch": 0.3502717391304348,
+      "grad_norm": 1.2321789264678955,
+      "learning_rate": 3e-05,
+      "loss": 2.8849,
+      "step": 1289
+    },
+    {
+      "epoch": 0.35054347826086957,
+      "grad_norm": 1.3166979551315308,
+      "learning_rate": 3e-05,
+      "loss": 3.1094,
+      "step": 1290
+    },
+    {
+      "epoch": 0.35081521739130433,
+      "grad_norm": 1.5346661806106567,
+      "learning_rate": 3e-05,
+      "loss": 3.2696,
+      "step": 1291
+    },
+    {
+      "epoch": 0.35108695652173916,
+      "grad_norm": 1.550195336341858,
+      "learning_rate": 3e-05,
+      "loss": 3.3044,
+      "step": 1292
+    },
+    {
+      "epoch": 0.3513586956521739,
+      "grad_norm": 1.7930152416229248,
+      "learning_rate": 3e-05,
+      "loss": 3.7782,
+      "step": 1293
+    },
+    {
+      "epoch": 0.3516304347826087,
+      "grad_norm": 1.9929698705673218,
+      "learning_rate": 3e-05,
+      "loss": 3.12,
+      "step": 1294
+    },
+    {
+      "epoch": 0.35190217391304346,
+      "grad_norm": 1.244729995727539,
+      "learning_rate": 3e-05,
+      "loss": 2.9368,
+      "step": 1295
+    },
+    {
+      "epoch": 0.3521739130434783,
+      "grad_norm": 1.3026173114776611,
+      "learning_rate": 3e-05,
+      "loss": 3.1947,
+      "step": 1296
+    },
+    {
+      "epoch": 0.35244565217391305,
+      "grad_norm": 1.297813892364502,
+      "learning_rate": 3e-05,
+      "loss": 2.5978,
+      "step": 1297
+    },
+    {
+      "epoch": 0.3527173913043478,
+      "grad_norm": 1.4547146558761597,
+      "learning_rate": 3e-05,
+      "loss": 3.4649,
+      "step": 1298
+    },
+    {
+      "epoch": 0.3529891304347826,
+      "grad_norm": 1.2365776300430298,
+      "learning_rate": 3e-05,
+      "loss": 3.1037,
+      "step": 1299
+    },
+    {
+      "epoch": 0.3532608695652174,
+      "grad_norm": 1.2477210760116577,
+      "learning_rate": 3e-05,
+      "loss": 3.1045,
+      "step": 1300
+    },
+    {
+      "epoch": 0.3535326086956522,
+      "grad_norm": 1.4197901487350464,
+      "learning_rate": 3e-05,
+      "loss": 3.2411,
+      "step": 1301
+    },
+    {
+      "epoch": 0.35380434782608694,
+      "grad_norm": 1.1791913509368896,
+      "learning_rate": 3e-05,
+      "loss": 3.4775,
+      "step": 1302
+    },
+    {
+      "epoch": 0.35407608695652176,
+      "grad_norm": 1.1935291290283203,
+      "learning_rate": 3e-05,
+      "loss": 2.9038,
+      "step": 1303
+    },
+    {
+      "epoch": 0.35434782608695653,
+      "grad_norm": 1.515292763710022,
+      "learning_rate": 3e-05,
+      "loss": 3.6569,
+      "step": 1304
+    },
+    {
+      "epoch": 0.3546195652173913,
+      "grad_norm": 1.6459676027297974,
+      "learning_rate": 3e-05,
+      "loss": 3.638,
+      "step": 1305
+    },
+    {
+      "epoch": 0.35489130434782606,
+      "grad_norm": 1.2799147367477417,
+      "learning_rate": 3e-05,
+      "loss": 3.3,
+      "step": 1306
+    },
+    {
+      "epoch": 0.3551630434782609,
+      "grad_norm": 1.1415566205978394,
+      "learning_rate": 3e-05,
+      "loss": 2.7568,
+      "step": 1307
+    },
+    {
+      "epoch": 0.35543478260869565,
+      "grad_norm": 1.4160751104354858,
+      "learning_rate": 3e-05,
+      "loss": 3.5389,
+      "step": 1308
+    },
+    {
+      "epoch": 0.3557065217391304,
+      "grad_norm": 1.5589027404785156,
+      "learning_rate": 3e-05,
+      "loss": 3.5721,
+      "step": 1309
+    },
+    {
+      "epoch": 0.35597826086956524,
+      "grad_norm": 1.4486863613128662,
+      "learning_rate": 3e-05,
+      "loss": 3.5145,
+      "step": 1310
+    },
+    {
+      "epoch": 0.35625,
+      "grad_norm": 1.1204112768173218,
+      "learning_rate": 3e-05,
+      "loss": 3.2109,
+      "step": 1311
+    },
+    {
+      "epoch": 0.3565217391304348,
+      "grad_norm": 1.4160523414611816,
+      "learning_rate": 3e-05,
+      "loss": 3.9734,
+      "step": 1312
+    },
+    {
+      "epoch": 0.35679347826086955,
+      "grad_norm": 1.359877347946167,
+      "learning_rate": 3e-05,
+      "loss": 3.1197,
+      "step": 1313
+    },
+    {
+      "epoch": 0.35706521739130437,
+      "grad_norm": 1.3469575643539429,
+      "learning_rate": 3e-05,
+      "loss": 2.7031,
+      "step": 1314
+    },
+    {
+      "epoch": 0.35733695652173914,
+      "grad_norm": 1.877429485321045,
+      "learning_rate": 3e-05,
+      "loss": 3.6433,
+      "step": 1315
+    },
+    {
+      "epoch": 0.3576086956521739,
+      "grad_norm": 1.6881115436553955,
+      "learning_rate": 3e-05,
+      "loss": 3.8429,
+      "step": 1316
+    },
+    {
+      "epoch": 0.35788043478260867,
+      "grad_norm": 0.9885056018829346,
+      "learning_rate": 3e-05,
+      "loss": 2.9943,
+      "step": 1317
+    },
+    {
+      "epoch": 0.3581521739130435,
+      "grad_norm": 1.1146260499954224,
+      "learning_rate": 3e-05,
+      "loss": 3.0583,
+      "step": 1318
+    },
+    {
+      "epoch": 0.35842391304347826,
+      "grad_norm": 1.0930460691452026,
+      "learning_rate": 3e-05,
+      "loss": 3.1491,
+      "step": 1319
+    },
+    {
+      "epoch": 0.358695652173913,
+      "grad_norm": 1.7721790075302124,
+      "learning_rate": 3e-05,
+      "loss": 3.682,
+      "step": 1320
+    },
+    {
+      "epoch": 0.35896739130434785,
+      "grad_norm": 1.8233963251113892,
+      "learning_rate": 3e-05,
+      "loss": 4.0548,
+      "step": 1321
+    },
+    {
+      "epoch": 0.3592391304347826,
+      "grad_norm": 2.3196828365325928,
+      "learning_rate": 3e-05,
+      "loss": 3.8065,
+      "step": 1322
+    },
+    {
+      "epoch": 0.3595108695652174,
+      "grad_norm": 2.0975351333618164,
+      "learning_rate": 3e-05,
+      "loss": 3.3098,
+      "step": 1323
+    },
+    {
+      "epoch": 0.35978260869565215,
+      "grad_norm": 1.4240121841430664,
+      "learning_rate": 3e-05,
+      "loss": 3.327,
+      "step": 1324
+    },
+    {
+      "epoch": 0.360054347826087,
+      "grad_norm": 1.4401570558547974,
+      "learning_rate": 3e-05,
+      "loss": 2.9099,
+      "step": 1325
+    },
+    {
+      "epoch": 0.36032608695652174,
+      "grad_norm": 1.7879765033721924,
+      "learning_rate": 3e-05,
+      "loss": 2.9776,
+      "step": 1326
+    },
+    {
+      "epoch": 0.3605978260869565,
+      "grad_norm": 1.8819643259048462,
+      "learning_rate": 3e-05,
+      "loss": 3.3526,
+      "step": 1327
+    },
+    {
+      "epoch": 0.36086956521739133,
+      "grad_norm": 1.8750115633010864,
+      "learning_rate": 3e-05,
+      "loss": 3.7411,
+      "step": 1328
+    },
+    {
+      "epoch": 0.3611413043478261,
+      "grad_norm": 1.802832007408142,
+      "learning_rate": 3e-05,
+      "loss": 2.8997,
+      "step": 1329
+    },
+    {
+      "epoch": 0.36141304347826086,
+      "grad_norm": 1.9432928562164307,
+      "learning_rate": 3e-05,
+      "loss": 2.9711,
+      "step": 1330
+    },
+    {
+      "epoch": 0.36168478260869563,
+      "grad_norm": 2.222259998321533,
+      "learning_rate": 3e-05,
+      "loss": 3.5956,
+      "step": 1331
+    },
+    {
+      "epoch": 0.36195652173913045,
+      "grad_norm": 1.0173529386520386,
+      "learning_rate": 3e-05,
+      "loss": 3.1368,
+      "step": 1332
+    },
+    {
+      "epoch": 0.3622282608695652,
+      "grad_norm": 1.3840457201004028,
+      "learning_rate": 3e-05,
+      "loss": 3.3336,
+      "step": 1333
+    },
+    {
+      "epoch": 0.3625,
+      "grad_norm": 2.199570417404175,
+      "learning_rate": 3e-05,
+      "loss": 3.1846,
+      "step": 1334
+    },
+    {
+      "epoch": 0.36277173913043476,
+      "grad_norm": 2.3063576221466064,
+      "learning_rate": 3e-05,
+      "loss": 3.1996,
+      "step": 1335
+    },
+    {
+      "epoch": 0.3630434782608696,
+      "grad_norm": 2.0329015254974365,
+      "learning_rate": 3e-05,
+      "loss": 3.6631,
+      "step": 1336
+    },
+    {
+      "epoch": 0.36331521739130435,
+      "grad_norm": 0.9080404043197632,
+      "learning_rate": 3e-05,
+      "loss": 2.8746,
+      "step": 1337
+    },
+    {
+      "epoch": 0.3635869565217391,
+      "grad_norm": 2.173773765563965,
+      "learning_rate": 3e-05,
+      "loss": 3.8486,
+      "step": 1338
+    },
+    {
+      "epoch": 0.36385869565217394,
+      "grad_norm": 2.493887186050415,
+      "learning_rate": 3e-05,
+      "loss": 3.5707,
+      "step": 1339
+    },
+    {
+      "epoch": 0.3641304347826087,
+      "grad_norm": 1.8862491846084595,
+      "learning_rate": 3e-05,
+      "loss": 3.8585,
+      "step": 1340
+    },
+    {
+      "epoch": 0.36440217391304347,
+      "grad_norm": 1.23379647731781,
+      "learning_rate": 3e-05,
+      "loss": 3.5806,
+      "step": 1341
+    },
+    {
+      "epoch": 0.36467391304347824,
+      "grad_norm": 1.3684219121932983,
+      "learning_rate": 3e-05,
+      "loss": 3.1318,
+      "step": 1342
+    },
+    {
+      "epoch": 0.36494565217391306,
+      "grad_norm": 1.2184761762619019,
+      "learning_rate": 3e-05,
+      "loss": 3.4761,
+      "step": 1343
+    },
+    {
+      "epoch": 0.3652173913043478,
+      "grad_norm": 1.230040431022644,
+      "learning_rate": 3e-05,
+      "loss": 4.1053,
+      "step": 1344
+    },
+    {
+      "epoch": 0.3654891304347826,
+      "grad_norm": 1.4862629175186157,
+      "learning_rate": 3e-05,
+      "loss": 3.4211,
+      "step": 1345
+    },
+    {
+      "epoch": 0.3657608695652174,
+      "grad_norm": 1.2832411527633667,
+      "learning_rate": 3e-05,
+      "loss": 2.7821,
+      "step": 1346
+    },
+    {
+      "epoch": 0.3660326086956522,
+      "grad_norm": 1.1333979368209839,
+      "learning_rate": 3e-05,
+      "loss": 3.4129,
+      "step": 1347
+    },
+    {
+      "epoch": 0.36630434782608695,
+      "grad_norm": 1.9462121725082397,
+      "learning_rate": 3e-05,
+      "loss": 3.8102,
+      "step": 1348
+    },
+    {
+      "epoch": 0.3665760869565217,
+      "grad_norm": 1.314512848854065,
+      "learning_rate": 3e-05,
+      "loss": 3.5802,
+      "step": 1349
+    },
+    {
+      "epoch": 0.36684782608695654,
+      "grad_norm": 0.9899564981460571,
+      "learning_rate": 3e-05,
+      "loss": 3.2397,
+      "step": 1350
+    },
+    {
+      "epoch": 0.3671195652173913,
+      "grad_norm": 0.9129654765129089,
+      "learning_rate": 3e-05,
+      "loss": 3.0704,
+      "step": 1351
+    },
+    {
+      "epoch": 0.3673913043478261,
+      "grad_norm": 1.2658374309539795,
+      "learning_rate": 3e-05,
+      "loss": 2.9416,
+      "step": 1352
+    },
+    {
+      "epoch": 0.36766304347826084,
+      "grad_norm": 1.156530499458313,
+      "learning_rate": 3e-05,
+      "loss": 3.2254,
+      "step": 1353
+    },
+    {
+      "epoch": 0.36793478260869567,
+      "grad_norm": 1.4215689897537231,
+      "learning_rate": 3e-05,
+      "loss": 3.5928,
+      "step": 1354
+    },
+    {
+      "epoch": 0.36820652173913043,
+      "grad_norm": 1.4695452451705933,
+      "learning_rate": 3e-05,
+      "loss": 3.5298,
+      "step": 1355
+    },
+    {
+      "epoch": 0.3684782608695652,
+      "grad_norm": 1.185398817062378,
+      "learning_rate": 3e-05,
+      "loss": 2.9786,
+      "step": 1356
+    },
+    {
+      "epoch": 0.36875,
+      "grad_norm": 1.6181942224502563,
+      "learning_rate": 3e-05,
+      "loss": 3.8328,
+      "step": 1357
+    },
+    {
+      "epoch": 0.3690217391304348,
+      "grad_norm": 1.7642178535461426,
+      "learning_rate": 3e-05,
+      "loss": 3.6084,
+      "step": 1358
+    },
+    {
+      "epoch": 0.36929347826086956,
+      "grad_norm": 1.3343443870544434,
+      "learning_rate": 3e-05,
+      "loss": 4.0491,
+      "step": 1359
+    },
+    {
+      "epoch": 0.3695652173913043,
+      "grad_norm": 2.167877435684204,
+      "learning_rate": 3e-05,
+      "loss": 3.6741,
+      "step": 1360
+    },
+    {
+      "epoch": 0.36983695652173915,
+      "grad_norm": 2.0322484970092773,
+      "learning_rate": 3e-05,
+      "loss": 4.1298,
+      "step": 1361
+    },
+    {
+      "epoch": 0.3701086956521739,
+      "grad_norm": 1.1205507516860962,
+      "learning_rate": 3e-05,
+      "loss": 3.2478,
+      "step": 1362
+    },
+    {
+      "epoch": 0.3703804347826087,
+      "grad_norm": 1.4173895120620728,
+      "learning_rate": 3e-05,
+      "loss": 3.2024,
+      "step": 1363
+    },
+    {
+      "epoch": 0.3706521739130435,
+      "grad_norm": 1.4186296463012695,
+      "learning_rate": 3e-05,
+      "loss": 2.9463,
+      "step": 1364
+    },
+    {
+      "epoch": 0.37092391304347827,
+      "grad_norm": 1.4897836446762085,
+      "learning_rate": 3e-05,
+      "loss": 3.4138,
+      "step": 1365
+    },
+    {
+      "epoch": 0.37119565217391304,
+      "grad_norm": 0.9213505983352661,
+      "learning_rate": 3e-05,
+      "loss": 2.7726,
+      "step": 1366
+    },
+    {
+      "epoch": 0.3714673913043478,
+      "grad_norm": 1.4468334913253784,
+      "learning_rate": 3e-05,
+      "loss": 3.3552,
+      "step": 1367
+    },
+    {
+      "epoch": 0.3717391304347826,
+      "grad_norm": 1.6020596027374268,
+      "learning_rate": 3e-05,
+      "loss": 3.3098,
+      "step": 1368
+    },
+    {
+      "epoch": 0.3720108695652174,
+      "grad_norm": 1.2568944692611694,
+      "learning_rate": 3e-05,
+      "loss": 2.9896,
+      "step": 1369
+    },
+    {
+      "epoch": 0.37228260869565216,
+      "grad_norm": 1.2092691659927368,
+      "learning_rate": 3e-05,
+      "loss": 3.2886,
+      "step": 1370
+    },
+    {
+      "epoch": 0.37255434782608693,
+      "grad_norm": 1.0227844715118408,
+      "learning_rate": 3e-05,
+      "loss": 2.9555,
+      "step": 1371
+    },
+    {
+      "epoch": 0.37282608695652175,
+      "grad_norm": 1.400106430053711,
+      "learning_rate": 3e-05,
+      "loss": 3.2697,
+      "step": 1372
+    },
+    {
+      "epoch": 0.3730978260869565,
+      "grad_norm": 1.3000365495681763,
+      "learning_rate": 3e-05,
+      "loss": 3.1748,
+      "step": 1373
+    },
+    {
+      "epoch": 0.3733695652173913,
+      "grad_norm": 1.197072148323059,
+      "learning_rate": 3e-05,
+      "loss": 3.3346,
+      "step": 1374
+    },
+    {
+      "epoch": 0.3736413043478261,
+      "grad_norm": 1.1778289079666138,
+      "learning_rate": 3e-05,
+      "loss": 3.089,
+      "step": 1375
+    },
+    {
+      "epoch": 0.3739130434782609,
+      "grad_norm": 1.23442542552948,
+      "learning_rate": 3e-05,
+      "loss": 3.0972,
+      "step": 1376
+    },
+    {
+      "epoch": 0.37418478260869564,
+      "grad_norm": 1.135533094406128,
+      "learning_rate": 3e-05,
+      "loss": 2.9638,
+      "step": 1377
+    },
+    {
+      "epoch": 0.3744565217391304,
+      "grad_norm": 1.4869073629379272,
+      "learning_rate": 3e-05,
+      "loss": 3.7439,
+      "step": 1378
+    },
+    {
+      "epoch": 0.37472826086956523,
+      "grad_norm": 1.5528923273086548,
+      "learning_rate": 3e-05,
+      "loss": 2.8187,
+      "step": 1379
+    },
+    {
+      "epoch": 0.375,
+      "grad_norm": 1.0053822994232178,
+      "learning_rate": 3e-05,
+      "loss": 3.2647,
+      "step": 1380
+    },
+    {
+      "epoch": 0.37527173913043477,
+      "grad_norm": 1.2217366695404053,
+      "learning_rate": 3e-05,
+      "loss": 3.5844,
+      "step": 1381
+    },
+    {
+      "epoch": 0.3755434782608696,
+      "grad_norm": 1.245322823524475,
+      "learning_rate": 3e-05,
+      "loss": 2.9917,
+      "step": 1382
+    },
+    {
+      "epoch": 0.37581521739130436,
+      "grad_norm": 1.3003109693527222,
+      "learning_rate": 3e-05,
+      "loss": 3.7163,
+      "step": 1383
+    },
+    {
+      "epoch": 0.3760869565217391,
+      "grad_norm": 2.398679733276367,
+      "learning_rate": 3e-05,
+      "loss": 4.2985,
+      "step": 1384
+    },
+    {
+      "epoch": 0.3763586956521739,
+      "grad_norm": 1.2174817323684692,
+      "learning_rate": 3e-05,
+      "loss": 3.3311,
+      "step": 1385
+    },
+    {
+      "epoch": 0.3766304347826087,
+      "grad_norm": 1.2073396444320679,
+      "learning_rate": 3e-05,
+      "loss": 3.4313,
+      "step": 1386
+    },
+    {
+      "epoch": 0.3769021739130435,
+      "grad_norm": 1.2685997486114502,
+      "learning_rate": 3e-05,
+      "loss": 3.1358,
+      "step": 1387
+    },
+    {
+      "epoch": 0.37717391304347825,
+      "grad_norm": 1.3566850423812866,
+      "learning_rate": 3e-05,
+      "loss": 3.0956,
+      "step": 1388
+    },
+    {
+      "epoch": 0.37744565217391307,
+      "grad_norm": 1.6564099788665771,
+      "learning_rate": 3e-05,
+      "loss": 3.7791,
+      "step": 1389
+    },
+    {
+      "epoch": 0.37771739130434784,
+      "grad_norm": 1.3698561191558838,
+      "learning_rate": 3e-05,
+      "loss": 3.3446,
+      "step": 1390
+    },
+    {
+      "epoch": 0.3779891304347826,
+      "grad_norm": 1.2702010869979858,
+      "learning_rate": 3e-05,
+      "loss": 4.0951,
+      "step": 1391
+    },
+    {
+      "epoch": 0.3782608695652174,
+      "grad_norm": 1.4307619333267212,
+      "learning_rate": 3e-05,
+      "loss": 3.3821,
+      "step": 1392
+    },
+    {
+      "epoch": 0.3785326086956522,
+      "grad_norm": 1.207396149635315,
+      "learning_rate": 3e-05,
+      "loss": 3.6105,
+      "step": 1393
+    },
+    {
+      "epoch": 0.37880434782608696,
+      "grad_norm": 1.1929562091827393,
+      "learning_rate": 3e-05,
+      "loss": 2.8538,
+      "step": 1394
+    },
+    {
+      "epoch": 0.37907608695652173,
+      "grad_norm": 1.2328208684921265,
+      "learning_rate": 3e-05,
+      "loss": 3.0622,
+      "step": 1395
+    },
+    {
+      "epoch": 0.3793478260869565,
+      "grad_norm": 1.4275541305541992,
+      "learning_rate": 3e-05,
+      "loss": 3.4482,
+      "step": 1396
+    },
+    {
+      "epoch": 0.3796195652173913,
+      "grad_norm": 0.9579852223396301,
+      "learning_rate": 3e-05,
+      "loss": 2.8646,
+      "step": 1397
+    },
+    {
+      "epoch": 0.3798913043478261,
+      "grad_norm": 1.6573379039764404,
+      "learning_rate": 3e-05,
+      "loss": 3.4417,
+      "step": 1398
+    },
+    {
+      "epoch": 0.38016304347826085,
+      "grad_norm": 1.5388156175613403,
+      "learning_rate": 3e-05,
+      "loss": 3.2041,
+      "step": 1399
+    },
+    {
+      "epoch": 0.3804347826086957,
+      "grad_norm": 1.4631450176239014,
+      "learning_rate": 3e-05,
+      "loss": 3.8626,
+      "step": 1400
+    },
+    {
+      "epoch": 0.38070652173913044,
+      "grad_norm": 1.3570584058761597,
+      "learning_rate": 3e-05,
+      "loss": 3.441,
+      "step": 1401
+    },
+    {
+      "epoch": 0.3809782608695652,
+      "grad_norm": 1.8016875982284546,
+      "learning_rate": 3e-05,
+      "loss": 3.1616,
+      "step": 1402
+    },
+    {
+      "epoch": 0.38125,
+      "grad_norm": 1.4279495477676392,
+      "learning_rate": 3e-05,
+      "loss": 3.6129,
+      "step": 1403
+    },
+    {
+      "epoch": 0.3815217391304348,
+      "grad_norm": 1.3305083513259888,
+      "learning_rate": 3e-05,
+      "loss": 3.3956,
+      "step": 1404
+    },
+    {
+      "epoch": 0.38179347826086957,
+      "grad_norm": 1.7289437055587769,
+      "learning_rate": 3e-05,
+      "loss": 3.6499,
+      "step": 1405
+    },
+    {
+      "epoch": 0.38206521739130433,
+      "grad_norm": 1.75969398021698,
+      "learning_rate": 3e-05,
+      "loss": 3.104,
+      "step": 1406
+    },
+    {
+      "epoch": 0.38233695652173916,
+      "grad_norm": 1.545464277267456,
+      "learning_rate": 3e-05,
+      "loss": 4.1157,
+      "step": 1407
+    },
+    {
+      "epoch": 0.3826086956521739,
+      "grad_norm": 1.5299899578094482,
+      "learning_rate": 3e-05,
+      "loss": 2.9975,
+      "step": 1408
+    },
+    {
+      "epoch": 0.3828804347826087,
+      "grad_norm": 1.7116879224777222,
+      "learning_rate": 3e-05,
+      "loss": 3.0605,
+      "step": 1409
+    },
+    {
+      "epoch": 0.38315217391304346,
+      "grad_norm": 1.4728469848632812,
+      "learning_rate": 3e-05,
+      "loss": 3.2079,
+      "step": 1410
+    },
+    {
+      "epoch": 0.3834239130434783,
+      "grad_norm": 2.216775894165039,
+      "learning_rate": 3e-05,
+      "loss": 3.8937,
+      "step": 1411
+    },
+    {
+      "epoch": 0.38369565217391305,
+      "grad_norm": 1.6279774904251099,
+      "learning_rate": 3e-05,
+      "loss": 3.7957,
+      "step": 1412
+    },
+    {
+      "epoch": 0.3839673913043478,
+      "grad_norm": 0.9390795230865479,
+      "learning_rate": 3e-05,
+      "loss": 2.8752,
+      "step": 1413
+    },
+    {
+      "epoch": 0.3842391304347826,
+      "grad_norm": 1.3048573732376099,
+      "learning_rate": 3e-05,
+      "loss": 3.0132,
+      "step": 1414
+    },
+    {
+      "epoch": 0.3845108695652174,
+      "grad_norm": 1.132607340812683,
+      "learning_rate": 3e-05,
+      "loss": 2.965,
+      "step": 1415
+    },
+    {
+      "epoch": 0.3847826086956522,
+      "grad_norm": 1.6570241451263428,
+      "learning_rate": 3e-05,
+      "loss": 3.1474,
+      "step": 1416
+    },
+    {
+      "epoch": 0.38505434782608694,
+      "grad_norm": 1.5501550436019897,
+      "learning_rate": 3e-05,
+      "loss": 3.4049,
+      "step": 1417
+    },
+    {
+      "epoch": 0.38532608695652176,
+      "grad_norm": 1.1150649785995483,
+      "learning_rate": 3e-05,
+      "loss": 3.4114,
+      "step": 1418
+    },
+    {
+      "epoch": 0.38559782608695653,
+      "grad_norm": 1.3886046409606934,
+      "learning_rate": 3e-05,
+      "loss": 3.7872,
+      "step": 1419
+    },
+    {
+      "epoch": 0.3858695652173913,
+      "grad_norm": 1.178165078163147,
+      "learning_rate": 3e-05,
+      "loss": 3.1258,
+      "step": 1420
+    },
+    {
+      "epoch": 0.38614130434782606,
+      "grad_norm": 1.2557275295257568,
+      "learning_rate": 3e-05,
+      "loss": 3.4292,
+      "step": 1421
+    },
+    {
+      "epoch": 0.3864130434782609,
+      "grad_norm": 0.9806421995162964,
+      "learning_rate": 3e-05,
+      "loss": 2.8931,
+      "step": 1422
+    },
+    {
+      "epoch": 0.38668478260869565,
+      "grad_norm": 1.6238960027694702,
+      "learning_rate": 3e-05,
+      "loss": 3.5711,
+      "step": 1423
+    },
+    {
+      "epoch": 0.3869565217391304,
+      "grad_norm": 1.3578873872756958,
+      "learning_rate": 3e-05,
+      "loss": 2.9015,
+      "step": 1424
+    },
+    {
+      "epoch": 0.38722826086956524,
+      "grad_norm": 1.2328130006790161,
+      "learning_rate": 3e-05,
+      "loss": 3.2193,
+      "step": 1425
+    },
+    {
+      "epoch": 0.3875,
+      "grad_norm": 2.146702289581299,
+      "learning_rate": 3e-05,
+      "loss": 4.0931,
+      "step": 1426
+    },
+    {
+      "epoch": 0.3877717391304348,
+      "grad_norm": 1.9415775537490845,
+      "learning_rate": 3e-05,
+      "loss": 3.0855,
+      "step": 1427
+    },
+    {
+      "epoch": 0.38804347826086955,
+      "grad_norm": 1.7980471849441528,
+      "learning_rate": 3e-05,
+      "loss": 3.3307,
+      "step": 1428
+    },
+    {
+      "epoch": 0.38831521739130437,
+      "grad_norm": 1.429101824760437,
+      "learning_rate": 3e-05,
+      "loss": 3.1628,
+      "step": 1429
+    },
+    {
+      "epoch": 0.38858695652173914,
+      "grad_norm": 2.225107431411743,
+      "learning_rate": 3e-05,
+      "loss": 3.5771,
+      "step": 1430
+    },
+    {
+      "epoch": 0.3888586956521739,
+      "grad_norm": 2.46109938621521,
+      "learning_rate": 3e-05,
+      "loss": 3.7408,
+      "step": 1431
+    },
+    {
+      "epoch": 0.38913043478260867,
+      "grad_norm": 2.695655345916748,
+      "learning_rate": 3e-05,
+      "loss": 3.4776,
+      "step": 1432
+    },
+    {
+      "epoch": 0.3894021739130435,
+      "grad_norm": 2.3022725582122803,
+      "learning_rate": 3e-05,
+      "loss": 3.4633,
+      "step": 1433
+    },
+    {
+      "epoch": 0.38967391304347826,
+      "grad_norm": 1.5871728658676147,
+      "learning_rate": 3e-05,
+      "loss": 2.8911,
+      "step": 1434
+    },
+    {
+      "epoch": 0.389945652173913,
+      "grad_norm": 1.3922131061553955,
+      "learning_rate": 3e-05,
+      "loss": 2.7701,
+      "step": 1435
+    },
+    {
+      "epoch": 0.39021739130434785,
+      "grad_norm": 2.131774663925171,
+      "learning_rate": 3e-05,
+      "loss": 3.6811,
+      "step": 1436
+    },
+    {
+      "epoch": 0.3904891304347826,
+      "grad_norm": 1.8943536281585693,
+      "learning_rate": 3e-05,
+      "loss": 2.6957,
+      "step": 1437
+    },
+    {
+      "epoch": 0.3907608695652174,
+      "grad_norm": 1.7963645458221436,
+      "learning_rate": 3e-05,
+      "loss": 3.4574,
+      "step": 1438
+    },
+    {
+      "epoch": 0.39103260869565215,
+      "grad_norm": 1.613253116607666,
+      "learning_rate": 3e-05,
+      "loss": 3.5328,
+      "step": 1439
+    },
+    {
+      "epoch": 0.391304347826087,
+      "grad_norm": 1.8886051177978516,
+      "learning_rate": 3e-05,
+      "loss": 2.7514,
+      "step": 1440
+    },
+    {
+      "epoch": 0.39157608695652174,
+      "grad_norm": 2.6512999534606934,
+      "learning_rate": 3e-05,
+      "loss": 3.9306,
+      "step": 1441
+    },
+    {
+      "epoch": 0.3918478260869565,
+      "grad_norm": 1.6623197793960571,
+      "learning_rate": 3e-05,
+      "loss": 3.4983,
+      "step": 1442
+    },
+    {
+      "epoch": 0.39211956521739133,
+      "grad_norm": 1.2289677858352661,
+      "learning_rate": 3e-05,
+      "loss": 3.1967,
+      "step": 1443
+    },
+    {
+      "epoch": 0.3923913043478261,
+      "grad_norm": 0.9220126867294312,
+      "learning_rate": 3e-05,
+      "loss": 2.8288,
+      "step": 1444
+    },
+    {
+      "epoch": 0.39266304347826086,
+      "grad_norm": 1.853948950767517,
+      "learning_rate": 3e-05,
+      "loss": 3.2876,
+      "step": 1445
+    },
+    {
+      "epoch": 0.39293478260869563,
+      "grad_norm": 2.040001392364502,
+      "learning_rate": 3e-05,
+      "loss": 3.6805,
+      "step": 1446
+    },
+    {
+      "epoch": 0.39320652173913045,
+      "grad_norm": 1.4386035203933716,
+      "learning_rate": 3e-05,
+      "loss": 3.1173,
+      "step": 1447
+    },
+    {
+      "epoch": 0.3934782608695652,
+      "grad_norm": 1.48452889919281,
+      "learning_rate": 3e-05,
+      "loss": 3.3276,
+      "step": 1448
+    },
+    {
+      "epoch": 0.39375,
+      "grad_norm": 1.058858871459961,
+      "learning_rate": 3e-05,
+      "loss": 2.5619,
+      "step": 1449
+    },
+    {
+      "epoch": 0.39402173913043476,
+      "grad_norm": 2.131944179534912,
+      "learning_rate": 3e-05,
+      "loss": 3.7079,
+      "step": 1450
+    },
+    {
+      "epoch": 0.3942934782608696,
+      "grad_norm": 1.5783636569976807,
+      "learning_rate": 3e-05,
+      "loss": 3.5856,
+      "step": 1451
+    },
+    {
+      "epoch": 0.39456521739130435,
+      "grad_norm": 1.249036192893982,
+      "learning_rate": 3e-05,
+      "loss": 3.0789,
+      "step": 1452
+    },
+    {
+      "epoch": 0.3948369565217391,
+      "grad_norm": 1.6562423706054688,
+      "learning_rate": 3e-05,
+      "loss": 3.9025,
+      "step": 1453
+    },
+    {
+      "epoch": 0.39510869565217394,
+      "grad_norm": 1.6861488819122314,
+      "learning_rate": 3e-05,
+      "loss": 3.5286,
+      "step": 1454
+    },
+    {
+      "epoch": 0.3953804347826087,
+      "grad_norm": 1.503812313079834,
+      "learning_rate": 3e-05,
+      "loss": 3.0909,
+      "step": 1455
+    },
+    {
+      "epoch": 0.39565217391304347,
+      "grad_norm": 1.8524895906448364,
+      "learning_rate": 3e-05,
+      "loss": 3.4152,
+      "step": 1456
+    },
+    {
+      "epoch": 0.39592391304347824,
+      "grad_norm": 1.3550132513046265,
+      "learning_rate": 3e-05,
+      "loss": 3.5444,
+      "step": 1457
+    },
+    {
+      "epoch": 0.39619565217391306,
+      "grad_norm": 1.454416036605835,
+      "learning_rate": 3e-05,
+      "loss": 3.6969,
+      "step": 1458
+    },
+    {
+      "epoch": 0.3964673913043478,
+      "grad_norm": 1.5491242408752441,
+      "learning_rate": 3e-05,
+      "loss": 3.4632,
+      "step": 1459
+    },
+    {
+      "epoch": 0.3967391304347826,
+      "grad_norm": 1.177980661392212,
+      "learning_rate": 3e-05,
+      "loss": 3.2844,
+      "step": 1460
+    },
+    {
+      "epoch": 0.3970108695652174,
+      "grad_norm": 1.2985824346542358,
+      "learning_rate": 3e-05,
+      "loss": 3.6743,
+      "step": 1461
+    },
+    {
+      "epoch": 0.3972826086956522,
+      "grad_norm": 1.545882225036621,
+      "learning_rate": 3e-05,
+      "loss": 3.4353,
+      "step": 1462
+    },
+    {
+      "epoch": 0.39755434782608695,
+      "grad_norm": 1.3580163717269897,
+      "learning_rate": 3e-05,
+      "loss": 3.7926,
+      "step": 1463
+    },
+    {
+      "epoch": 0.3978260869565217,
+      "grad_norm": 1.4254475831985474,
+      "learning_rate": 3e-05,
+      "loss": 3.7358,
+      "step": 1464
+    },
+    {
+      "epoch": 0.39809782608695654,
+      "grad_norm": 1.2237005233764648,
+      "learning_rate": 3e-05,
+      "loss": 3.3169,
+      "step": 1465
+    },
+    {
+      "epoch": 0.3983695652173913,
+      "grad_norm": 1.1622488498687744,
+      "learning_rate": 3e-05,
+      "loss": 3.1261,
+      "step": 1466
+    },
+    {
+      "epoch": 0.3986413043478261,
+      "grad_norm": 0.9447959065437317,
+      "learning_rate": 3e-05,
+      "loss": 2.8757,
+      "step": 1467
+    },
+    {
+      "epoch": 0.39891304347826084,
+      "grad_norm": 1.5178827047348022,
+      "learning_rate": 3e-05,
+      "loss": 3.7686,
+      "step": 1468
+    },
+    {
+      "epoch": 0.39918478260869567,
+      "grad_norm": 1.2081400156021118,
+      "learning_rate": 3e-05,
+      "loss": 2.7978,
+      "step": 1469
+    },
+    {
+      "epoch": 0.39945652173913043,
+      "grad_norm": 1.3333414793014526,
+      "learning_rate": 3e-05,
+      "loss": 3.8005,
+      "step": 1470
+    },
+    {
+      "epoch": 0.3997282608695652,
+      "grad_norm": 1.2284449338912964,
+      "learning_rate": 3e-05,
+      "loss": 3.2384,
+      "step": 1471
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6261060237884521,
+      "learning_rate": 3e-05,
+      "loss": 3.6663,
+      "step": 1472
+    },
+    {
+      "epoch": 0.4002717391304348,
+      "grad_norm": 1.2063846588134766,
+      "learning_rate": 3e-05,
+      "loss": 3.2808,
+      "step": 1473
+    },
+    {
+      "epoch": 0.40054347826086956,
+      "grad_norm": 1.1668663024902344,
+      "learning_rate": 3e-05,
+      "loss": 3.6527,
+      "step": 1474
+    },
+    {
+      "epoch": 0.4008152173913043,
+      "grad_norm": 1.1682161092758179,
+      "learning_rate": 3e-05,
+      "loss": 3.0487,
+      "step": 1475
+    },
+    {
+      "epoch": 0.40108695652173915,
+      "grad_norm": 1.2479826211929321,
+      "learning_rate": 3e-05,
+      "loss": 3.8231,
+      "step": 1476
+    },
+    {
+      "epoch": 0.4013586956521739,
+      "grad_norm": 1.3903083801269531,
+      "learning_rate": 3e-05,
+      "loss": 3.3589,
+      "step": 1477
+    },
+    {
+      "epoch": 0.4016304347826087,
+      "grad_norm": 2.756995677947998,
+      "learning_rate": 3e-05,
+      "loss": 3.5956,
+      "step": 1478
+    },
+    {
+      "epoch": 0.4019021739130435,
+      "grad_norm": 1.1125861406326294,
+      "learning_rate": 3e-05,
+      "loss": 2.9482,
+      "step": 1479
+    },
+    {
+      "epoch": 0.40217391304347827,
+      "grad_norm": 1.7494776248931885,
+      "learning_rate": 3e-05,
+      "loss": 3.6373,
+      "step": 1480
+    },
+    {
+      "epoch": 0.40244565217391304,
+      "grad_norm": 1.7383618354797363,
+      "learning_rate": 3e-05,
+      "loss": 3.5196,
+      "step": 1481
+    },
+    {
+      "epoch": 0.4027173913043478,
+      "grad_norm": 2.0332143306732178,
+      "learning_rate": 3e-05,
+      "loss": 3.4659,
+      "step": 1482
+    },
+    {
+      "epoch": 0.4029891304347826,
+      "grad_norm": 1.7549830675125122,
+      "learning_rate": 3e-05,
+      "loss": 3.6027,
+      "step": 1483
+    },
+    {
+      "epoch": 0.4032608695652174,
+      "grad_norm": 1.4020575284957886,
+      "learning_rate": 3e-05,
+      "loss": 3.1451,
+      "step": 1484
+    },
+    {
+      "epoch": 0.40353260869565216,
+      "grad_norm": 1.134787917137146,
+      "learning_rate": 3e-05,
+      "loss": 2.6124,
+      "step": 1485
+    },
+    {
+      "epoch": 0.40380434782608693,
+      "grad_norm": 1.6569262742996216,
+      "learning_rate": 3e-05,
+      "loss": 3.8979,
+      "step": 1486
+    },
+    {
+      "epoch": 0.40407608695652175,
+      "grad_norm": 1.7637388706207275,
+      "learning_rate": 3e-05,
+      "loss": 3.8109,
+      "step": 1487
+    },
+    {
+      "epoch": 0.4043478260869565,
+      "grad_norm": 1.2442704439163208,
+      "learning_rate": 3e-05,
+      "loss": 3.6988,
+      "step": 1488
+    },
+    {
+      "epoch": 0.4046195652173913,
+      "grad_norm": 1.0942063331604004,
+      "learning_rate": 3e-05,
+      "loss": 3.0464,
+      "step": 1489
+    },
+    {
+      "epoch": 0.4048913043478261,
+      "grad_norm": 1.337386965751648,
+      "learning_rate": 3e-05,
+      "loss": 3.8127,
+      "step": 1490
+    },
+    {
+      "epoch": 0.4051630434782609,
+      "grad_norm": 1.7307548522949219,
+      "learning_rate": 3e-05,
+      "loss": 3.3854,
+      "step": 1491
+    },
+    {
+      "epoch": 0.40543478260869564,
+      "grad_norm": 3.1193931102752686,
+      "learning_rate": 3e-05,
+      "loss": 4.9053,
+      "step": 1492
+    },
+    {
+      "epoch": 0.4057065217391304,
+      "grad_norm": 1.5950829982757568,
+      "learning_rate": 3e-05,
+      "loss": 3.6687,
+      "step": 1493
+    },
+    {
+      "epoch": 0.40597826086956523,
+      "grad_norm": 1.6609991788864136,
+      "learning_rate": 3e-05,
+      "loss": 3.4985,
+      "step": 1494
+    },
+    {
+      "epoch": 0.40625,
+      "grad_norm": 1.824634075164795,
+      "learning_rate": 3e-05,
+      "loss": 2.7576,
+      "step": 1495
+    },
+    {
+      "epoch": 0.40652173913043477,
+      "grad_norm": 1.8175928592681885,
+      "learning_rate": 3e-05,
+      "loss": 3.7615,
+      "step": 1496
+    },
+    {
+      "epoch": 0.4067934782608696,
+      "grad_norm": 1.3160499334335327,
+      "learning_rate": 3e-05,
+      "loss": 3.1647,
+      "step": 1497
+    },
+    {
+      "epoch": 0.40706521739130436,
+      "grad_norm": 1.7744853496551514,
+      "learning_rate": 3e-05,
+      "loss": 3.0396,
+      "step": 1498
+    },
+    {
+      "epoch": 0.4073369565217391,
+      "grad_norm": 2.0241026878356934,
+      "learning_rate": 3e-05,
+      "loss": 3.8311,
+      "step": 1499
+    },
+    {
+      "epoch": 0.4076086956521739,
+      "grad_norm": 1.7789679765701294,
+      "learning_rate": 3e-05,
+      "loss": 2.9538,
+      "step": 1500
+    },
+    {
+      "epoch": 0.4078804347826087,
+      "grad_norm": 1.4469610452651978,
+      "learning_rate": 3e-05,
+      "loss": 3.3083,
+      "step": 1501
+    },
+    {
+      "epoch": 0.4081521739130435,
+      "grad_norm": 1.570556879043579,
+      "learning_rate": 3e-05,
+      "loss": 3.0793,
+      "step": 1502
+    },
+    {
+      "epoch": 0.40842391304347825,
+      "grad_norm": 1.3964000940322876,
+      "learning_rate": 3e-05,
+      "loss": 2.9603,
+      "step": 1503
+    },
+    {
+      "epoch": 0.40869565217391307,
+      "grad_norm": 1.1800280809402466,
+      "learning_rate": 3e-05,
+      "loss": 3.2305,
+      "step": 1504
+    },
+    {
+      "epoch": 0.40896739130434784,
+      "grad_norm": 1.762847661972046,
+      "learning_rate": 3e-05,
+      "loss": 3.7564,
+      "step": 1505
+    },
+    {
+      "epoch": 0.4092391304347826,
+      "grad_norm": 1.5233442783355713,
+      "learning_rate": 3e-05,
+      "loss": 3.2744,
+      "step": 1506
+    },
+    {
+      "epoch": 0.4095108695652174,
+      "grad_norm": 1.9167475700378418,
+      "learning_rate": 3e-05,
+      "loss": 3.6962,
+      "step": 1507
+    },
+    {
+      "epoch": 0.4097826086956522,
+      "grad_norm": 1.949634075164795,
+      "learning_rate": 3e-05,
+      "loss": 3.1084,
+      "step": 1508
+    },
+    {
+      "epoch": 0.41005434782608696,
+      "grad_norm": 1.9483072757720947,
+      "learning_rate": 3e-05,
+      "loss": 3.1895,
+      "step": 1509
+    },
+    {
+      "epoch": 0.41032608695652173,
+      "grad_norm": 1.368509292602539,
+      "learning_rate": 3e-05,
+      "loss": 2.6804,
+      "step": 1510
+    },
+    {
+      "epoch": 0.4105978260869565,
+      "grad_norm": 1.6476247310638428,
+      "learning_rate": 3e-05,
+      "loss": 3.1565,
+      "step": 1511
+    },
+    {
+      "epoch": 0.4108695652173913,
+      "grad_norm": 2.14841890335083,
+      "learning_rate": 3e-05,
+      "loss": 3.5723,
+      "step": 1512
+    },
+    {
+      "epoch": 0.4111413043478261,
+      "grad_norm": 2.1899356842041016,
+      "learning_rate": 3e-05,
+      "loss": 3.9324,
+      "step": 1513
+    },
+    {
+      "epoch": 0.41141304347826085,
+      "grad_norm": 1.971421480178833,
+      "learning_rate": 3e-05,
+      "loss": 3.3302,
+      "step": 1514
+    },
+    {
+      "epoch": 0.4116847826086957,
+      "grad_norm": 1.555216908454895,
+      "learning_rate": 3e-05,
+      "loss": 2.8559,
+      "step": 1515
+    },
+    {
+      "epoch": 0.41195652173913044,
+      "grad_norm": 2.025975227355957,
+      "learning_rate": 3e-05,
+      "loss": 3.8605,
+      "step": 1516
+    },
+    {
+      "epoch": 0.4122282608695652,
+      "grad_norm": 1.609168291091919,
+      "learning_rate": 3e-05,
+      "loss": 3.521,
+      "step": 1517
+    },
+    {
+      "epoch": 0.4125,
+      "grad_norm": 1.4011493921279907,
+      "learning_rate": 3e-05,
+      "loss": 2.905,
+      "step": 1518
+    },
+    {
+      "epoch": 0.4127717391304348,
+      "grad_norm": 1.813499927520752,
+      "learning_rate": 3e-05,
+      "loss": 3.9861,
+      "step": 1519
+    },
+    {
+      "epoch": 0.41304347826086957,
+      "grad_norm": 2.023961067199707,
+      "learning_rate": 3e-05,
+      "loss": 3.3814,
+      "step": 1520
+    },
+    {
+      "epoch": 0.41331521739130433,
+      "grad_norm": 1.9356021881103516,
+      "learning_rate": 3e-05,
+      "loss": 2.9743,
+      "step": 1521
+    },
+    {
+      "epoch": 0.41358695652173916,
+      "grad_norm": 1.6765419244766235,
+      "learning_rate": 3e-05,
+      "loss": 2.9504,
+      "step": 1522
+    },
+    {
+      "epoch": 0.4138586956521739,
+      "grad_norm": 1.954119324684143,
+      "learning_rate": 3e-05,
+      "loss": 3.7254,
+      "step": 1523
+    },
+    {
+      "epoch": 0.4141304347826087,
+      "grad_norm": 1.5637805461883545,
+      "learning_rate": 3e-05,
+      "loss": 3.8543,
+      "step": 1524
+    },
+    {
+      "epoch": 0.41440217391304346,
+      "grad_norm": 1.175187587738037,
+      "learning_rate": 3e-05,
+      "loss": 3.2871,
+      "step": 1525
+    },
+    {
+      "epoch": 0.4146739130434783,
+      "grad_norm": 1.3228040933609009,
+      "learning_rate": 3e-05,
+      "loss": 2.9662,
+      "step": 1526
+    },
+    {
+      "epoch": 0.41494565217391305,
+      "grad_norm": 1.4776461124420166,
+      "learning_rate": 3e-05,
+      "loss": 2.9617,
+      "step": 1527
+    },
+    {
+      "epoch": 0.4152173913043478,
+      "grad_norm": 1.5169886350631714,
+      "learning_rate": 3e-05,
+      "loss": 3.6288,
+      "step": 1528
+    },
+    {
+      "epoch": 0.4154891304347826,
+      "grad_norm": 1.4785473346710205,
+      "learning_rate": 3e-05,
+      "loss": 3.0807,
+      "step": 1529
+    },
+    {
+      "epoch": 0.4157608695652174,
+      "grad_norm": 1.4565258026123047,
+      "learning_rate": 3e-05,
+      "loss": 3.3099,
+      "step": 1530
+    },
+    {
+      "epoch": 0.4160326086956522,
+      "grad_norm": 1.2732715606689453,
+      "learning_rate": 3e-05,
+      "loss": 3.2948,
+      "step": 1531
+    },
+    {
+      "epoch": 0.41630434782608694,
+      "grad_norm": 1.0217174291610718,
+      "learning_rate": 3e-05,
+      "loss": 3.0084,
+      "step": 1532
+    },
+    {
+      "epoch": 0.41657608695652176,
+      "grad_norm": 1.2822792530059814,
+      "learning_rate": 3e-05,
+      "loss": 2.9284,
+      "step": 1533
+    },
+    {
+      "epoch": 0.41684782608695653,
+      "grad_norm": 1.641827940940857,
+      "learning_rate": 3e-05,
+      "loss": 3.4684,
+      "step": 1534
+    },
+    {
+      "epoch": 0.4171195652173913,
+      "grad_norm": 1.5296562910079956,
+      "learning_rate": 3e-05,
+      "loss": 2.5836,
+      "step": 1535
+    },
+    {
+      "epoch": 0.41739130434782606,
+      "grad_norm": 1.3570102453231812,
+      "learning_rate": 3e-05,
+      "loss": 3.1323,
+      "step": 1536
+    },
+    {
+      "epoch": 0.4176630434782609,
+      "grad_norm": 0.9901162981987,
+      "learning_rate": 3e-05,
+      "loss": 2.7749,
+      "step": 1537
+    },
+    {
+      "epoch": 0.41793478260869565,
+      "grad_norm": 1.2664462327957153,
+      "learning_rate": 3e-05,
+      "loss": 3.1213,
+      "step": 1538
+    },
+    {
+      "epoch": 0.4182065217391304,
+      "grad_norm": 1.682426929473877,
+      "learning_rate": 3e-05,
+      "loss": 3.8292,
+      "step": 1539
+    },
+    {
+      "epoch": 0.41847826086956524,
+      "grad_norm": 1.3441531658172607,
+      "learning_rate": 3e-05,
+      "loss": 2.8989,
+      "step": 1540
+    },
+    {
+      "epoch": 0.41875,
+      "grad_norm": 1.6413555145263672,
+      "learning_rate": 3e-05,
+      "loss": 3.4333,
+      "step": 1541
+    },
+    {
+      "epoch": 0.4190217391304348,
+      "grad_norm": 1.2601666450500488,
+      "learning_rate": 3e-05,
+      "loss": 3.2815,
+      "step": 1542
+    },
+    {
+      "epoch": 0.41929347826086955,
+      "grad_norm": 1.5582549571990967,
+      "learning_rate": 3e-05,
+      "loss": 3.5814,
+      "step": 1543
+    },
+    {
+      "epoch": 0.41956521739130437,
+      "grad_norm": 1.0220012664794922,
+      "learning_rate": 3e-05,
+      "loss": 2.7242,
+      "step": 1544
+    },
+    {
+      "epoch": 0.41983695652173914,
+      "grad_norm": 1.710802674293518,
+      "learning_rate": 3e-05,
+      "loss": 3.5271,
+      "step": 1545
+    },
+    {
+      "epoch": 0.4201086956521739,
+      "grad_norm": 2.3148810863494873,
+      "learning_rate": 3e-05,
+      "loss": 4.0931,
+      "step": 1546
+    },
+    {
+      "epoch": 0.42038043478260867,
+      "grad_norm": 1.4409066438674927,
+      "learning_rate": 3e-05,
+      "loss": 3.4806,
+      "step": 1547
+    },
+    {
+      "epoch": 0.4206521739130435,
+      "grad_norm": 1.4869251251220703,
+      "learning_rate": 3e-05,
+      "loss": 3.8039,
+      "step": 1548
+    },
+    {
+      "epoch": 0.42092391304347826,
+      "grad_norm": 2.2359511852264404,
+      "learning_rate": 3e-05,
+      "loss": 4.1842,
+      "step": 1549
+    },
+    {
+      "epoch": 0.421195652173913,
+      "grad_norm": 2.2962846755981445,
+      "learning_rate": 3e-05,
+      "loss": 3.8559,
+      "step": 1550
+    },
+    {
+      "epoch": 0.42146739130434785,
+      "grad_norm": 1.5050580501556396,
+      "learning_rate": 3e-05,
+      "loss": 3.6678,
+      "step": 1551
+    },
+    {
+      "epoch": 0.4217391304347826,
+      "grad_norm": 1.2851321697235107,
+      "learning_rate": 3e-05,
+      "loss": 3.4175,
+      "step": 1552
+    },
+    {
+      "epoch": 0.4220108695652174,
+      "grad_norm": 1.2671363353729248,
+      "learning_rate": 3e-05,
+      "loss": 3.0412,
+      "step": 1553
+    },
+    {
+      "epoch": 0.42228260869565215,
+      "grad_norm": 1.2936915159225464,
+      "learning_rate": 3e-05,
+      "loss": 3.0918,
+      "step": 1554
+    },
+    {
+      "epoch": 0.422554347826087,
+      "grad_norm": 1.4296749830245972,
+      "learning_rate": 3e-05,
+      "loss": 3.3257,
+      "step": 1555
+    },
+    {
+      "epoch": 0.42282608695652174,
+      "grad_norm": 1.2652101516723633,
+      "learning_rate": 3e-05,
+      "loss": 3.5451,
+      "step": 1556
+    },
+    {
+      "epoch": 0.4230978260869565,
+      "grad_norm": 1.4265552759170532,
+      "learning_rate": 3e-05,
+      "loss": 3.4145,
+      "step": 1557
+    },
+    {
+      "epoch": 0.42336956521739133,
+      "grad_norm": 1.3775297403335571,
+      "learning_rate": 3e-05,
+      "loss": 3.3248,
+      "step": 1558
+    },
+    {
+      "epoch": 0.4236413043478261,
+      "grad_norm": 1.8461849689483643,
+      "learning_rate": 3e-05,
+      "loss": 3.9212,
+      "step": 1559
+    },
+    {
+      "epoch": 0.42391304347826086,
+      "grad_norm": 1.412374496459961,
+      "learning_rate": 3e-05,
+      "loss": 3.4737,
+      "step": 1560
+    },
+    {
+      "epoch": 0.42418478260869563,
+      "grad_norm": 1.6272127628326416,
+      "learning_rate": 3e-05,
+      "loss": 3.4379,
+      "step": 1561
+    },
+    {
+      "epoch": 0.42445652173913045,
+      "grad_norm": 1.754745364189148,
+      "learning_rate": 3e-05,
+      "loss": 3.4164,
+      "step": 1562
+    },
+    {
+      "epoch": 0.4247282608695652,
+      "grad_norm": 1.2676855325698853,
+      "learning_rate": 3e-05,
+      "loss": 3.3106,
+      "step": 1563
+    },
+    {
+      "epoch": 0.425,
+      "grad_norm": 1.7627284526824951,
+      "learning_rate": 3e-05,
+      "loss": 3.841,
+      "step": 1564
+    },
+    {
+      "epoch": 0.42527173913043476,
+      "grad_norm": 1.143531084060669,
+      "learning_rate": 3e-05,
+      "loss": 2.82,
+      "step": 1565
+    },
+    {
+      "epoch": 0.4255434782608696,
+      "grad_norm": 1.0254485607147217,
+      "learning_rate": 3e-05,
+      "loss": 2.9556,
+      "step": 1566
+    },
+    {
+      "epoch": 0.42581521739130435,
+      "grad_norm": 1.3668675422668457,
+      "learning_rate": 3e-05,
+      "loss": 3.7105,
+      "step": 1567
+    },
+    {
+      "epoch": 0.4260869565217391,
+      "grad_norm": 2.208592176437378,
+      "learning_rate": 3e-05,
+      "loss": 3.4152,
+      "step": 1568
+    },
+    {
+      "epoch": 0.42635869565217394,
+      "grad_norm": 1.1941901445388794,
+      "learning_rate": 3e-05,
+      "loss": 3.3036,
+      "step": 1569
+    },
+    {
+      "epoch": 0.4266304347826087,
+      "grad_norm": 1.536787748336792,
+      "learning_rate": 3e-05,
+      "loss": 3.5194,
+      "step": 1570
+    },
+    {
+      "epoch": 0.42690217391304347,
+      "grad_norm": 1.4054354429244995,
+      "learning_rate": 3e-05,
+      "loss": 3.263,
+      "step": 1571
+    },
+    {
+      "epoch": 0.42717391304347824,
+      "grad_norm": 1.6471023559570312,
+      "learning_rate": 3e-05,
+      "loss": 3.4855,
+      "step": 1572
+    },
+    {
+      "epoch": 0.42744565217391306,
+      "grad_norm": 0.9407289028167725,
+      "learning_rate": 3e-05,
+      "loss": 3.1382,
+      "step": 1573
+    },
+    {
+      "epoch": 0.4277173913043478,
+      "grad_norm": 1.244971513748169,
+      "learning_rate": 3e-05,
+      "loss": 3.4904,
+      "step": 1574
+    },
+    {
+      "epoch": 0.4279891304347826,
+      "grad_norm": 1.305245041847229,
+      "learning_rate": 3e-05,
+      "loss": 3.3049,
+      "step": 1575
+    },
+    {
+      "epoch": 0.4282608695652174,
+      "grad_norm": 1.2542614936828613,
+      "learning_rate": 3e-05,
+      "loss": 3.4931,
+      "step": 1576
+    },
+    {
+      "epoch": 0.4285326086956522,
+      "grad_norm": 0.9759632349014282,
+      "learning_rate": 3e-05,
+      "loss": 3.4818,
+      "step": 1577
+    },
+    {
+      "epoch": 0.42880434782608695,
+      "grad_norm": 1.4484260082244873,
+      "learning_rate": 3e-05,
+      "loss": 3.2934,
+      "step": 1578
+    },
+    {
+      "epoch": 0.4290760869565217,
+      "grad_norm": 1.1898603439331055,
+      "learning_rate": 3e-05,
+      "loss": 3.4601,
+      "step": 1579
+    },
+    {
+      "epoch": 0.42934782608695654,
+      "grad_norm": 1.486471176147461,
+      "learning_rate": 3e-05,
+      "loss": 3.762,
+      "step": 1580
+    },
+    {
+      "epoch": 0.4296195652173913,
+      "grad_norm": 1.2455788850784302,
+      "learning_rate": 3e-05,
+      "loss": 3.721,
+      "step": 1581
+    },
+    {
+      "epoch": 0.4298913043478261,
+      "grad_norm": 1.1194945573806763,
+      "learning_rate": 3e-05,
+      "loss": 3.3573,
+      "step": 1582
+    },
+    {
+      "epoch": 0.43016304347826084,
+      "grad_norm": 1.666195273399353,
+      "learning_rate": 3e-05,
+      "loss": 3.531,
+      "step": 1583
+    },
+    {
+      "epoch": 0.43043478260869567,
+      "grad_norm": 1.192292332649231,
+      "learning_rate": 3e-05,
+      "loss": 3.1313,
+      "step": 1584
+    },
+    {
+      "epoch": 0.43070652173913043,
+      "grad_norm": 1.1080567836761475,
+      "learning_rate": 3e-05,
+      "loss": 3.2448,
+      "step": 1585
+    },
+    {
+      "epoch": 0.4309782608695652,
+      "grad_norm": 1.7732088565826416,
+      "learning_rate": 3e-05,
+      "loss": 3.4527,
+      "step": 1586
+    },
+    {
+      "epoch": 0.43125,
+      "grad_norm": 1.2465471029281616,
+      "learning_rate": 3e-05,
+      "loss": 2.8915,
+      "step": 1587
+    },
+    {
+      "epoch": 0.4315217391304348,
+      "grad_norm": 1.43854820728302,
+      "learning_rate": 3e-05,
+      "loss": 3.2366,
+      "step": 1588
+    },
+    {
+      "epoch": 0.43179347826086956,
+      "grad_norm": 1.3088682889938354,
+      "learning_rate": 3e-05,
+      "loss": 3.3654,
+      "step": 1589
+    },
+    {
+      "epoch": 0.4320652173913043,
+      "grad_norm": 1.0148249864578247,
+      "learning_rate": 3e-05,
+      "loss": 3.083,
+      "step": 1590
+    },
+    {
+      "epoch": 0.43233695652173915,
+      "grad_norm": 1.2219303846359253,
+      "learning_rate": 3e-05,
+      "loss": 3.0298,
+      "step": 1591
+    },
+    {
+      "epoch": 0.4326086956521739,
+      "grad_norm": 1.081207275390625,
+      "learning_rate": 3e-05,
+      "loss": 3.2116,
+      "step": 1592
+    },
+    {
+      "epoch": 0.4328804347826087,
+      "grad_norm": 1.515478491783142,
+      "learning_rate": 3e-05,
+      "loss": 3.4368,
+      "step": 1593
+    },
+    {
+      "epoch": 0.4331521739130435,
+      "grad_norm": 1.073887586593628,
+      "learning_rate": 3e-05,
+      "loss": 3.1832,
+      "step": 1594
+    },
+    {
+      "epoch": 0.43342391304347827,
+      "grad_norm": 1.1452287435531616,
+      "learning_rate": 3e-05,
+      "loss": 3.5617,
+      "step": 1595
+    },
+    {
+      "epoch": 0.43369565217391304,
+      "grad_norm": 1.512242078781128,
+      "learning_rate": 3e-05,
+      "loss": 3.4468,
+      "step": 1596
+    },
+    {
+      "epoch": 0.4339673913043478,
+      "grad_norm": 1.1862598657608032,
+      "learning_rate": 3e-05,
+      "loss": 3.3795,
+      "step": 1597
+    },
+    {
+      "epoch": 0.4342391304347826,
+      "grad_norm": 1.4696996212005615,
+      "learning_rate": 3e-05,
+      "loss": 3.5239,
+      "step": 1598
+    },
+    {
+      "epoch": 0.4345108695652174,
+      "grad_norm": 1.5985097885131836,
+      "learning_rate": 3e-05,
+      "loss": 3.8709,
+      "step": 1599
+    },
+    {
+      "epoch": 0.43478260869565216,
+      "grad_norm": 1.3134711980819702,
+      "learning_rate": 3e-05,
+      "loss": 3.4582,
+      "step": 1600
+    },
+    {
+      "epoch": 0.43505434782608693,
+      "grad_norm": 1.2656311988830566,
+      "learning_rate": 3e-05,
+      "loss": 3.7025,
+      "step": 1601
+    },
+    {
+      "epoch": 0.43532608695652175,
+      "grad_norm": 1.1780740022659302,
+      "learning_rate": 3e-05,
+      "loss": 3.3486,
+      "step": 1602
+    },
+    {
+      "epoch": 0.4355978260869565,
+      "grad_norm": 1.2048178911209106,
+      "learning_rate": 3e-05,
+      "loss": 3.3115,
+      "step": 1603
+    },
+    {
+      "epoch": 0.4358695652173913,
+      "grad_norm": 1.3589476346969604,
+      "learning_rate": 3e-05,
+      "loss": 3.2494,
+      "step": 1604
+    },
+    {
+      "epoch": 0.4361413043478261,
+      "grad_norm": 1.3171757459640503,
+      "learning_rate": 3e-05,
+      "loss": 2.8635,
+      "step": 1605
+    },
+    {
+      "epoch": 0.4364130434782609,
+      "grad_norm": 1.1358094215393066,
+      "learning_rate": 3e-05,
+      "loss": 2.9426,
+      "step": 1606
+    },
+    {
+      "epoch": 0.43668478260869564,
+      "grad_norm": 1.376583456993103,
+      "learning_rate": 3e-05,
+      "loss": 3.2611,
+      "step": 1607
+    },
+    {
+      "epoch": 0.4369565217391304,
+      "grad_norm": 1.7452890872955322,
+      "learning_rate": 3e-05,
+      "loss": 3.7445,
+      "step": 1608
+    },
+    {
+      "epoch": 0.43722826086956523,
+      "grad_norm": 1.715166687965393,
+      "learning_rate": 3e-05,
+      "loss": 3.4322,
+      "step": 1609
+    },
+    {
+      "epoch": 0.4375,
+      "grad_norm": 1.3614475727081299,
+      "learning_rate": 3e-05,
+      "loss": 3.5131,
+      "step": 1610
+    },
+    {
+      "epoch": 0.43777173913043477,
+      "grad_norm": 1.4811841249465942,
+      "learning_rate": 3e-05,
+      "loss": 3.714,
+      "step": 1611
+    },
+    {
+      "epoch": 0.4380434782608696,
+      "grad_norm": 1.071926236152649,
+      "learning_rate": 3e-05,
+      "loss": 3.0734,
+      "step": 1612
+    },
+    {
+      "epoch": 0.43831521739130436,
+      "grad_norm": 1.5618953704833984,
+      "learning_rate": 3e-05,
+      "loss": 3.876,
+      "step": 1613
+    },
+    {
+      "epoch": 0.4385869565217391,
+      "grad_norm": 1.1414976119995117,
+      "learning_rate": 3e-05,
+      "loss": 3.5318,
+      "step": 1614
+    },
+    {
+      "epoch": 0.4388586956521739,
+      "grad_norm": 1.0475927591323853,
+      "learning_rate": 3e-05,
+      "loss": 3.293,
+      "step": 1615
+    },
+    {
+      "epoch": 0.4391304347826087,
+      "grad_norm": 1.392609715461731,
+      "learning_rate": 3e-05,
+      "loss": 2.9974,
+      "step": 1616
+    },
+    {
+      "epoch": 0.4394021739130435,
+      "grad_norm": 0.906802237033844,
+      "learning_rate": 3e-05,
+      "loss": 2.9825,
+      "step": 1617
+    },
+    {
+      "epoch": 0.43967391304347825,
+      "grad_norm": 1.4019746780395508,
+      "learning_rate": 3e-05,
+      "loss": 3.8698,
+      "step": 1618
+    },
+    {
+      "epoch": 0.43994565217391307,
+      "grad_norm": 1.321829915046692,
+      "learning_rate": 3e-05,
+      "loss": 3.7109,
+      "step": 1619
+    },
+    {
+      "epoch": 0.44021739130434784,
+      "grad_norm": 1.0838751792907715,
+      "learning_rate": 3e-05,
+      "loss": 2.6205,
+      "step": 1620
+    },
+    {
+      "epoch": 0.4404891304347826,
+      "grad_norm": 2.383323907852173,
+      "learning_rate": 3e-05,
+      "loss": 4.2267,
+      "step": 1621
+    },
+    {
+      "epoch": 0.4407608695652174,
+      "grad_norm": 1.0130791664123535,
+      "learning_rate": 3e-05,
+      "loss": 2.788,
+      "step": 1622
+    },
+    {
+      "epoch": 0.4410326086956522,
+      "grad_norm": 1.3985093832015991,
+      "learning_rate": 3e-05,
+      "loss": 3.1002,
+      "step": 1623
+    },
+    {
+      "epoch": 0.44130434782608696,
+      "grad_norm": 1.3819383382797241,
+      "learning_rate": 3e-05,
+      "loss": 3.2219,
+      "step": 1624
+    },
+    {
+      "epoch": 0.44157608695652173,
+      "grad_norm": 1.4096386432647705,
+      "learning_rate": 3e-05,
+      "loss": 3.6886,
+      "step": 1625
+    },
+    {
+      "epoch": 0.4418478260869565,
+      "grad_norm": 1.4819231033325195,
+      "learning_rate": 3e-05,
+      "loss": 3.2191,
+      "step": 1626
+    },
+    {
+      "epoch": 0.4421195652173913,
+      "grad_norm": 1.3344846963882446,
+      "learning_rate": 3e-05,
+      "loss": 3.3927,
+      "step": 1627
+    },
+    {
+      "epoch": 0.4423913043478261,
+      "grad_norm": 1.469277024269104,
+      "learning_rate": 3e-05,
+      "loss": 3.4069,
+      "step": 1628
+    },
+    {
+      "epoch": 0.44266304347826085,
+      "grad_norm": 1.0877147912979126,
+      "learning_rate": 3e-05,
+      "loss": 2.9891,
+      "step": 1629
+    },
+    {
+      "epoch": 0.4429347826086957,
+      "grad_norm": 1.1775012016296387,
+      "learning_rate": 3e-05,
+      "loss": 3.3463,
+      "step": 1630
+    },
+    {
+      "epoch": 0.44320652173913044,
+      "grad_norm": 1.240936517715454,
+      "learning_rate": 3e-05,
+      "loss": 3.5751,
+      "step": 1631
+    },
+    {
+      "epoch": 0.4434782608695652,
+      "grad_norm": 0.877865195274353,
+      "learning_rate": 3e-05,
+      "loss": 2.7317,
+      "step": 1632
+    },
+    {
+      "epoch": 0.44375,
+      "grad_norm": 1.1812602281570435,
+      "learning_rate": 3e-05,
+      "loss": 3.0307,
+      "step": 1633
+    },
+    {
+      "epoch": 0.4440217391304348,
+      "grad_norm": 1.7113842964172363,
+      "learning_rate": 3e-05,
+      "loss": 4.1471,
+      "step": 1634
+    },
+    {
+      "epoch": 0.44429347826086957,
+      "grad_norm": 1.1820313930511475,
+      "learning_rate": 3e-05,
+      "loss": 3.362,
+      "step": 1635
+    },
+    {
+      "epoch": 0.44456521739130433,
+      "grad_norm": 1.3378487825393677,
+      "learning_rate": 3e-05,
+      "loss": 3.5815,
+      "step": 1636
+    },
+    {
+      "epoch": 0.44483695652173916,
+      "grad_norm": 0.8756458163261414,
+      "learning_rate": 3e-05,
+      "loss": 2.6882,
+      "step": 1637
+    },
+    {
+      "epoch": 0.4451086956521739,
+      "grad_norm": 1.020820140838623,
+      "learning_rate": 3e-05,
+      "loss": 3.1006,
+      "step": 1638
+    },
+    {
+      "epoch": 0.4453804347826087,
+      "grad_norm": 1.5075011253356934,
+      "learning_rate": 3e-05,
+      "loss": 4.0147,
+      "step": 1639
+    },
+    {
+      "epoch": 0.44565217391304346,
+      "grad_norm": 1.1874289512634277,
+      "learning_rate": 3e-05,
+      "loss": 3.0811,
+      "step": 1640
+    },
+    {
+      "epoch": 0.4459239130434783,
+      "grad_norm": 1.5765689611434937,
+      "learning_rate": 3e-05,
+      "loss": 3.4697,
+      "step": 1641
+    },
+    {
+      "epoch": 0.44619565217391305,
+      "grad_norm": 1.3812084197998047,
+      "learning_rate": 3e-05,
+      "loss": 3.6468,
+      "step": 1642
+    },
+    {
+      "epoch": 0.4464673913043478,
+      "grad_norm": 1.1477484703063965,
+      "learning_rate": 3e-05,
+      "loss": 3.2347,
+      "step": 1643
+    },
+    {
+      "epoch": 0.4467391304347826,
+      "grad_norm": 1.431220531463623,
+      "learning_rate": 3e-05,
+      "loss": 3.8073,
+      "step": 1644
+    },
+    {
+      "epoch": 0.4470108695652174,
+      "grad_norm": 1.0271748304367065,
+      "learning_rate": 3e-05,
+      "loss": 3.2165,
+      "step": 1645
+    },
+    {
+      "epoch": 0.4472826086956522,
+      "grad_norm": 1.2913216352462769,
+      "learning_rate": 3e-05,
+      "loss": 3.2766,
+      "step": 1646
+    },
+    {
+      "epoch": 0.44755434782608694,
+      "grad_norm": 1.341670274734497,
+      "learning_rate": 3e-05,
+      "loss": 3.8426,
+      "step": 1647
+    },
+    {
+      "epoch": 0.44782608695652176,
+      "grad_norm": 1.1871728897094727,
+      "learning_rate": 3e-05,
+      "loss": 3.1251,
+      "step": 1648
+    },
+    {
+      "epoch": 0.44809782608695653,
+      "grad_norm": 1.2151528596878052,
+      "learning_rate": 3e-05,
+      "loss": 3.3811,
+      "step": 1649
+    },
+    {
+      "epoch": 0.4483695652173913,
+      "grad_norm": 1.7120990753173828,
+      "learning_rate": 3e-05,
+      "loss": 3.7093,
+      "step": 1650
+    },
+    {
+      "epoch": 0.44864130434782606,
+      "grad_norm": 1.1526025533676147,
+      "learning_rate": 3e-05,
+      "loss": 3.1806,
+      "step": 1651
+    },
+    {
+      "epoch": 0.4489130434782609,
+      "grad_norm": 1.5061256885528564,
+      "learning_rate": 3e-05,
+      "loss": 3.7246,
+      "step": 1652
+    },
+    {
+      "epoch": 0.44918478260869565,
+      "grad_norm": 1.442798376083374,
+      "learning_rate": 3e-05,
+      "loss": 2.8975,
+      "step": 1653
+    },
+    {
+      "epoch": 0.4494565217391304,
+      "grad_norm": 1.5733429193496704,
+      "learning_rate": 3e-05,
+      "loss": 3.7047,
+      "step": 1654
+    },
+    {
+      "epoch": 0.44972826086956524,
+      "grad_norm": 1.5891181230545044,
+      "learning_rate": 3e-05,
+      "loss": 3.5587,
+      "step": 1655
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2529733180999756,
+      "learning_rate": 3e-05,
+      "loss": 3.079,
+      "step": 1656
+    },
+    {
+      "epoch": 0.4502717391304348,
+      "grad_norm": 2.397771120071411,
+      "learning_rate": 3e-05,
+      "loss": 3.2517,
+      "step": 1657
+    },
+    {
+      "epoch": 0.45054347826086955,
+      "grad_norm": 1.5895618200302124,
+      "learning_rate": 3e-05,
+      "loss": 3.289,
+      "step": 1658
+    },
+    {
+      "epoch": 0.45081521739130437,
+      "grad_norm": 1.3885931968688965,
+      "learning_rate": 3e-05,
+      "loss": 3.8401,
+      "step": 1659
+    },
+    {
+      "epoch": 0.45108695652173914,
+      "grad_norm": 1.848649263381958,
+      "learning_rate": 3e-05,
+      "loss": 3.2375,
+      "step": 1660
+    },
+    {
+      "epoch": 0.4513586956521739,
+      "grad_norm": 2.288536787033081,
+      "learning_rate": 3e-05,
+      "loss": 3.0737,
+      "step": 1661
+    },
+    {
+      "epoch": 0.45163043478260867,
+      "grad_norm": 1.5626355409622192,
+      "learning_rate": 3e-05,
+      "loss": 3.6127,
+      "step": 1662
+    },
+    {
+      "epoch": 0.4519021739130435,
+      "grad_norm": 1.2159314155578613,
+      "learning_rate": 3e-05,
+      "loss": 3.1681,
+      "step": 1663
+    },
+    {
+      "epoch": 0.45217391304347826,
+      "grad_norm": 2.7480335235595703,
+      "learning_rate": 3e-05,
+      "loss": 3.3753,
+      "step": 1664
+    },
+    {
+      "epoch": 0.452445652173913,
+      "grad_norm": 1.5772864818572998,
+      "learning_rate": 3e-05,
+      "loss": 3.4648,
+      "step": 1665
+    },
+    {
+      "epoch": 0.45271739130434785,
+      "grad_norm": 2.1393795013427734,
+      "learning_rate": 3e-05,
+      "loss": 4.067,
+      "step": 1666
+    },
+    {
+      "epoch": 0.4529891304347826,
+      "grad_norm": 1.48007071018219,
+      "learning_rate": 3e-05,
+      "loss": 3.0416,
+      "step": 1667
+    },
+    {
+      "epoch": 0.4532608695652174,
+      "grad_norm": 1.3958046436309814,
+      "learning_rate": 3e-05,
+      "loss": 2.9871,
+      "step": 1668
+    },
+    {
+      "epoch": 0.45353260869565215,
+      "grad_norm": 1.6530804634094238,
+      "learning_rate": 3e-05,
+      "loss": 3.1665,
+      "step": 1669
+    },
+    {
+      "epoch": 0.453804347826087,
+      "grad_norm": 1.5180555582046509,
+      "learning_rate": 3e-05,
+      "loss": 3.1545,
+      "step": 1670
+    },
+    {
+      "epoch": 0.45407608695652174,
+      "grad_norm": 1.3946945667266846,
+      "learning_rate": 3e-05,
+      "loss": 3.0606,
+      "step": 1671
+    },
+    {
+      "epoch": 0.4543478260869565,
+      "grad_norm": 1.655504584312439,
+      "learning_rate": 3e-05,
+      "loss": 3.5141,
+      "step": 1672
+    },
+    {
+      "epoch": 0.45461956521739133,
+      "grad_norm": 1.8746256828308105,
+      "learning_rate": 3e-05,
+      "loss": 3.2537,
+      "step": 1673
+    },
+    {
+      "epoch": 0.4548913043478261,
+      "grad_norm": 1.6152852773666382,
+      "learning_rate": 3e-05,
+      "loss": 3.1725,
+      "step": 1674
+    },
+    {
+      "epoch": 0.45516304347826086,
+      "grad_norm": 2.030758857727051,
+      "learning_rate": 3e-05,
+      "loss": 4.1911,
+      "step": 1675
+    },
+    {
+      "epoch": 0.45543478260869563,
+      "grad_norm": 1.573011040687561,
+      "learning_rate": 3e-05,
+      "loss": 2.7004,
+      "step": 1676
+    },
+    {
+      "epoch": 0.45570652173913045,
+      "grad_norm": 1.7782936096191406,
+      "learning_rate": 3e-05,
+      "loss": 3.4129,
+      "step": 1677
+    },
+    {
+      "epoch": 0.4559782608695652,
+      "grad_norm": 1.6364409923553467,
+      "learning_rate": 3e-05,
+      "loss": 3.7828,
+      "step": 1678
+    },
+    {
+      "epoch": 0.45625,
+      "grad_norm": 1.095005750656128,
+      "learning_rate": 3e-05,
+      "loss": 2.9575,
+      "step": 1679
+    },
+    {
+      "epoch": 0.45652173913043476,
+      "grad_norm": 1.4323863983154297,
+      "learning_rate": 3e-05,
+      "loss": 2.9855,
+      "step": 1680
+    },
+    {
+      "epoch": 0.4567934782608696,
+      "grad_norm": 1.8281795978546143,
+      "learning_rate": 3e-05,
+      "loss": 3.2274,
+      "step": 1681
+    },
+    {
+      "epoch": 0.45706521739130435,
+      "grad_norm": 1.553568959236145,
+      "learning_rate": 3e-05,
+      "loss": 3.4245,
+      "step": 1682
+    },
+    {
+      "epoch": 0.4573369565217391,
+      "grad_norm": 1.229791522026062,
+      "learning_rate": 3e-05,
+      "loss": 2.8749,
+      "step": 1683
+    },
+    {
+      "epoch": 0.45760869565217394,
+      "grad_norm": 1.4264166355133057,
+      "learning_rate": 3e-05,
+      "loss": 3.0987,
+      "step": 1684
+    },
+    {
+      "epoch": 0.4578804347826087,
+      "grad_norm": 1.4009737968444824,
+      "learning_rate": 3e-05,
+      "loss": 3.2024,
+      "step": 1685
+    },
+    {
+      "epoch": 0.45815217391304347,
+      "grad_norm": 1.2461766004562378,
+      "learning_rate": 3e-05,
+      "loss": 3.1414,
+      "step": 1686
+    },
+    {
+      "epoch": 0.45842391304347824,
+      "grad_norm": 1.1765496730804443,
+      "learning_rate": 3e-05,
+      "loss": 3.7237,
+      "step": 1687
+    },
+    {
+      "epoch": 0.45869565217391306,
+      "grad_norm": 1.2187762260437012,
+      "learning_rate": 3e-05,
+      "loss": 3.5997,
+      "step": 1688
+    },
+    {
+      "epoch": 0.4589673913043478,
+      "grad_norm": 2.117086410522461,
+      "learning_rate": 3e-05,
+      "loss": 3.5497,
+      "step": 1689
+    },
+    {
+      "epoch": 0.4592391304347826,
+      "grad_norm": 1.2980027198791504,
+      "learning_rate": 3e-05,
+      "loss": 3.6001,
+      "step": 1690
+    },
+    {
+      "epoch": 0.4595108695652174,
+      "grad_norm": 1.1031181812286377,
+      "learning_rate": 3e-05,
+      "loss": 2.814,
+      "step": 1691
+    },
+    {
+      "epoch": 0.4597826086956522,
+      "grad_norm": 1.201127052307129,
+      "learning_rate": 3e-05,
+      "loss": 3.4627,
+      "step": 1692
+    },
+    {
+      "epoch": 0.46005434782608695,
+      "grad_norm": 1.9975850582122803,
+      "learning_rate": 3e-05,
+      "loss": 3.8157,
+      "step": 1693
+    },
+    {
+      "epoch": 0.4603260869565217,
+      "grad_norm": 1.4874250888824463,
+      "learning_rate": 3e-05,
+      "loss": 3.5901,
+      "step": 1694
+    },
+    {
+      "epoch": 0.46059782608695654,
+      "grad_norm": 1.9747381210327148,
+      "learning_rate": 3e-05,
+      "loss": 3.8642,
+      "step": 1695
+    },
+    {
+      "epoch": 0.4608695652173913,
+      "grad_norm": 1.405049443244934,
+      "learning_rate": 3e-05,
+      "loss": 3.3367,
+      "step": 1696
+    },
+    {
+      "epoch": 0.4611413043478261,
+      "grad_norm": 1.4599591493606567,
+      "learning_rate": 3e-05,
+      "loss": 3.6499,
+      "step": 1697
+    },
+    {
+      "epoch": 0.46141304347826084,
+      "grad_norm": 1.6565916538238525,
+      "learning_rate": 3e-05,
+      "loss": 3.5368,
+      "step": 1698
+    },
+    {
+      "epoch": 0.46168478260869567,
+      "grad_norm": 1.327285647392273,
+      "learning_rate": 3e-05,
+      "loss": 2.7457,
+      "step": 1699
+    },
+    {
+      "epoch": 0.46195652173913043,
+      "grad_norm": 1.317681074142456,
+      "learning_rate": 3e-05,
+      "loss": 3.3931,
+      "step": 1700
+    },
+    {
+      "epoch": 0.4622282608695652,
+      "grad_norm": 1.7210628986358643,
+      "learning_rate": 3e-05,
+      "loss": 4.0996,
+      "step": 1701
+    },
+    {
+      "epoch": 0.4625,
+      "grad_norm": 1.4575893878936768,
+      "learning_rate": 3e-05,
+      "loss": 3.5176,
+      "step": 1702
+    },
+    {
+      "epoch": 0.4627717391304348,
+      "grad_norm": 0.8817341327667236,
+      "learning_rate": 3e-05,
+      "loss": 2.9138,
+      "step": 1703
+    },
+    {
+      "epoch": 0.46304347826086956,
+      "grad_norm": 1.6268318891525269,
+      "learning_rate": 3e-05,
+      "loss": 3.3346,
+      "step": 1704
+    },
+    {
+      "epoch": 0.4633152173913043,
+      "grad_norm": 1.281497836112976,
+      "learning_rate": 3e-05,
+      "loss": 2.776,
+      "step": 1705
+    },
+    {
+      "epoch": 0.46358695652173915,
+      "grad_norm": 1.6672241687774658,
+      "learning_rate": 3e-05,
+      "loss": 3.5205,
+      "step": 1706
+    },
+    {
+      "epoch": 0.4638586956521739,
+      "grad_norm": 1.4224839210510254,
+      "learning_rate": 3e-05,
+      "loss": 3.7986,
+      "step": 1707
+    },
+    {
+      "epoch": 0.4641304347826087,
+      "grad_norm": 1.1987775564193726,
+      "learning_rate": 3e-05,
+      "loss": 3.2364,
+      "step": 1708
+    },
+    {
+      "epoch": 0.4644021739130435,
+      "grad_norm": 1.4384889602661133,
+      "learning_rate": 3e-05,
+      "loss": 3.6339,
+      "step": 1709
+    },
+    {
+      "epoch": 0.46467391304347827,
+      "grad_norm": 1.1359198093414307,
+      "learning_rate": 3e-05,
+      "loss": 3.2827,
+      "step": 1710
+    },
+    {
+      "epoch": 0.46494565217391304,
+      "grad_norm": 1.3200854063034058,
+      "learning_rate": 3e-05,
+      "loss": 3.2209,
+      "step": 1711
+    },
+    {
+      "epoch": 0.4652173913043478,
+      "grad_norm": 1.613688588142395,
+      "learning_rate": 3e-05,
+      "loss": 3.7578,
+      "step": 1712
+    },
+    {
+      "epoch": 0.4654891304347826,
+      "grad_norm": 1.7866336107254028,
+      "learning_rate": 3e-05,
+      "loss": 3.9267,
+      "step": 1713
+    },
+    {
+      "epoch": 0.4657608695652174,
+      "grad_norm": 2.0430612564086914,
+      "learning_rate": 3e-05,
+      "loss": 3.6663,
+      "step": 1714
+    },
+    {
+      "epoch": 0.46603260869565216,
+      "grad_norm": 1.848181128501892,
+      "learning_rate": 3e-05,
+      "loss": 3.3196,
+      "step": 1715
+    },
+    {
+      "epoch": 0.46630434782608693,
+      "grad_norm": 1.4686863422393799,
+      "learning_rate": 3e-05,
+      "loss": 3.352,
+      "step": 1716
+    },
+    {
+      "epoch": 0.46657608695652175,
+      "grad_norm": 1.1621712446212769,
+      "learning_rate": 3e-05,
+      "loss": 3.1954,
+      "step": 1717
+    },
+    {
+      "epoch": 0.4668478260869565,
+      "grad_norm": 1.2042642831802368,
+      "learning_rate": 3e-05,
+      "loss": 2.7059,
+      "step": 1718
+    },
+    {
+      "epoch": 0.4671195652173913,
+      "grad_norm": 2.4052507877349854,
+      "learning_rate": 3e-05,
+      "loss": 3.6323,
+      "step": 1719
+    },
+    {
+      "epoch": 0.4673913043478261,
+      "grad_norm": 0.8568060994148254,
+      "learning_rate": 3e-05,
+      "loss": 2.5241,
+      "step": 1720
+    },
+    {
+      "epoch": 0.4676630434782609,
+      "grad_norm": 0.9677852988243103,
+      "learning_rate": 3e-05,
+      "loss": 2.5949,
+      "step": 1721
+    },
+    {
+      "epoch": 0.46793478260869564,
+      "grad_norm": 1.2894710302352905,
+      "learning_rate": 3e-05,
+      "loss": 2.989,
+      "step": 1722
+    },
+    {
+      "epoch": 0.4682065217391304,
+      "grad_norm": 1.5006341934204102,
+      "learning_rate": 3e-05,
+      "loss": 3.6919,
+      "step": 1723
+    },
+    {
+      "epoch": 0.46847826086956523,
+      "grad_norm": 1.2537362575531006,
+      "learning_rate": 3e-05,
+      "loss": 3.3816,
+      "step": 1724
+    },
+    {
+      "epoch": 0.46875,
+      "grad_norm": 1.333892822265625,
+      "learning_rate": 3e-05,
+      "loss": 3.3992,
+      "step": 1725
+    },
+    {
+      "epoch": 0.46902173913043477,
+      "grad_norm": 1.1863936185836792,
+      "learning_rate": 3e-05,
+      "loss": 3.1168,
+      "step": 1726
+    },
+    {
+      "epoch": 0.4692934782608696,
+      "grad_norm": 1.2577567100524902,
+      "learning_rate": 3e-05,
+      "loss": 3.4444,
+      "step": 1727
+    },
+    {
+      "epoch": 0.46956521739130436,
+      "grad_norm": 1.2792398929595947,
+      "learning_rate": 3e-05,
+      "loss": 2.8464,
+      "step": 1728
+    },
+    {
+      "epoch": 0.4698369565217391,
+      "grad_norm": 2.50295090675354,
+      "learning_rate": 3e-05,
+      "loss": 3.2523,
+      "step": 1729
+    },
+    {
+      "epoch": 0.4701086956521739,
+      "grad_norm": 1.842974066734314,
+      "learning_rate": 3e-05,
+      "loss": 3.3193,
+      "step": 1730
+    },
+    {
+      "epoch": 0.4703804347826087,
+      "grad_norm": 1.1899735927581787,
+      "learning_rate": 3e-05,
+      "loss": 3.4586,
+      "step": 1731
+    },
+    {
+      "epoch": 0.4706521739130435,
+      "grad_norm": 1.7522422075271606,
+      "learning_rate": 3e-05,
+      "loss": 3.44,
+      "step": 1732
+    },
+    {
+      "epoch": 0.47092391304347825,
+      "grad_norm": 2.0558695793151855,
+      "learning_rate": 3e-05,
+      "loss": 3.3303,
+      "step": 1733
+    },
+    {
+      "epoch": 0.47119565217391307,
+      "grad_norm": 1.5954774618148804,
+      "learning_rate": 3e-05,
+      "loss": 3.3366,
+      "step": 1734
+    },
+    {
+      "epoch": 0.47146739130434784,
+      "grad_norm": 1.4976911544799805,
+      "learning_rate": 3e-05,
+      "loss": 3.7435,
+      "step": 1735
+    },
+    {
+      "epoch": 0.4717391304347826,
+      "grad_norm": 1.4843627214431763,
+      "learning_rate": 3e-05,
+      "loss": 3.0997,
+      "step": 1736
+    },
+    {
+      "epoch": 0.4720108695652174,
+      "grad_norm": 1.7294570207595825,
+      "learning_rate": 3e-05,
+      "loss": 3.4683,
+      "step": 1737
+    },
+    {
+      "epoch": 0.4722826086956522,
+      "grad_norm": 1.7918407917022705,
+      "learning_rate": 3e-05,
+      "loss": 3.5634,
+      "step": 1738
+    },
+    {
+      "epoch": 0.47255434782608696,
+      "grad_norm": 1.458388090133667,
+      "learning_rate": 3e-05,
+      "loss": 3.724,
+      "step": 1739
+    },
+    {
+      "epoch": 0.47282608695652173,
+      "grad_norm": 1.7183854579925537,
+      "learning_rate": 3e-05,
+      "loss": 3.7768,
+      "step": 1740
+    },
+    {
+      "epoch": 0.4730978260869565,
+      "grad_norm": 1.6272084712982178,
+      "learning_rate": 3e-05,
+      "loss": 3.1361,
+      "step": 1741
+    },
+    {
+      "epoch": 0.4733695652173913,
+      "grad_norm": 1.2989143133163452,
+      "learning_rate": 3e-05,
+      "loss": 3.1645,
+      "step": 1742
+    },
+    {
+      "epoch": 0.4736413043478261,
+      "grad_norm": 1.1352357864379883,
+      "learning_rate": 3e-05,
+      "loss": 2.9178,
+      "step": 1743
+    },
+    {
+      "epoch": 0.47391304347826085,
+      "grad_norm": 1.6457839012145996,
+      "learning_rate": 3e-05,
+      "loss": 3.8608,
+      "step": 1744
+    },
+    {
+      "epoch": 0.4741847826086957,
+      "grad_norm": 1.3702744245529175,
+      "learning_rate": 3e-05,
+      "loss": 2.8434,
+      "step": 1745
+    },
+    {
+      "epoch": 0.47445652173913044,
+      "grad_norm": 1.2821577787399292,
+      "learning_rate": 3e-05,
+      "loss": 3.0827,
+      "step": 1746
+    },
+    {
+      "epoch": 0.4747282608695652,
+      "grad_norm": 1.4484645128250122,
+      "learning_rate": 3e-05,
+      "loss": 3.222,
+      "step": 1747
+    },
+    {
+      "epoch": 0.475,
+      "grad_norm": 1.7668582201004028,
+      "learning_rate": 3e-05,
+      "loss": 3.2441,
+      "step": 1748
+    },
+    {
+      "epoch": 0.4752717391304348,
+      "grad_norm": 1.764919400215149,
+      "learning_rate": 3e-05,
+      "loss": 3.7584,
+      "step": 1749
+    },
+    {
+      "epoch": 0.47554347826086957,
+      "grad_norm": 1.640993356704712,
+      "learning_rate": 3e-05,
+      "loss": 3.3807,
+      "step": 1750
+    },
+    {
+      "epoch": 0.47581521739130433,
+      "grad_norm": 1.6005651950836182,
+      "learning_rate": 3e-05,
+      "loss": 4.2452,
+      "step": 1751
+    },
+    {
+      "epoch": 0.47608695652173916,
+      "grad_norm": 1.421935796737671,
+      "learning_rate": 3e-05,
+      "loss": 3.0033,
+      "step": 1752
+    },
+    {
+      "epoch": 0.4763586956521739,
+      "grad_norm": 1.7603181600570679,
+      "learning_rate": 3e-05,
+      "loss": 3.2585,
+      "step": 1753
+    },
+    {
+      "epoch": 0.4766304347826087,
+      "grad_norm": 1.378020167350769,
+      "learning_rate": 3e-05,
+      "loss": 3.0198,
+      "step": 1754
+    },
+    {
+      "epoch": 0.47690217391304346,
+      "grad_norm": 1.5630027055740356,
+      "learning_rate": 3e-05,
+      "loss": 3.0409,
+      "step": 1755
+    },
+    {
+      "epoch": 0.4771739130434783,
+      "grad_norm": 1.1110471487045288,
+      "learning_rate": 3e-05,
+      "loss": 3.2521,
+      "step": 1756
+    },
+    {
+      "epoch": 0.47744565217391305,
+      "grad_norm": 1.403874397277832,
+      "learning_rate": 3e-05,
+      "loss": 2.9834,
+      "step": 1757
+    },
+    {
+      "epoch": 0.4777173913043478,
+      "grad_norm": 1.2942219972610474,
+      "learning_rate": 3e-05,
+      "loss": 3.1724,
+      "step": 1758
+    },
+    {
+      "epoch": 0.4779891304347826,
+      "grad_norm": 1.152891993522644,
+      "learning_rate": 3e-05,
+      "loss": 3.2869,
+      "step": 1759
+    },
+    {
+      "epoch": 0.4782608695652174,
+      "grad_norm": 1.1676504611968994,
+      "learning_rate": 3e-05,
+      "loss": 3.1582,
+      "step": 1760
+    },
+    {
+      "epoch": 0.4785326086956522,
+      "grad_norm": 1.151185154914856,
+      "learning_rate": 3e-05,
+      "loss": 3.3322,
+      "step": 1761
+    },
+    {
+      "epoch": 0.47880434782608694,
+      "grad_norm": 1.4192887544631958,
+      "learning_rate": 3e-05,
+      "loss": 3.3659,
+      "step": 1762
+    },
+    {
+      "epoch": 0.47907608695652176,
+      "grad_norm": 1.5391749143600464,
+      "learning_rate": 3e-05,
+      "loss": 3.2556,
+      "step": 1763
+    },
+    {
+      "epoch": 0.47934782608695653,
+      "grad_norm": 1.4667121171951294,
+      "learning_rate": 3e-05,
+      "loss": 3.8051,
+      "step": 1764
+    },
+    {
+      "epoch": 0.4796195652173913,
+      "grad_norm": 1.3416023254394531,
+      "learning_rate": 3e-05,
+      "loss": 3.4739,
+      "step": 1765
+    },
+    {
+      "epoch": 0.47989130434782606,
+      "grad_norm": 1.7379714250564575,
+      "learning_rate": 3e-05,
+      "loss": 3.2991,
+      "step": 1766
+    },
+    {
+      "epoch": 0.4801630434782609,
+      "grad_norm": 1.6660994291305542,
+      "learning_rate": 3e-05,
+      "loss": 3.1885,
+      "step": 1767
+    },
+    {
+      "epoch": 0.48043478260869565,
+      "grad_norm": 1.9886304140090942,
+      "learning_rate": 3e-05,
+      "loss": 3.3295,
+      "step": 1768
+    },
+    {
+      "epoch": 0.4807065217391304,
+      "grad_norm": 2.345217227935791,
+      "learning_rate": 3e-05,
+      "loss": 3.6796,
+      "step": 1769
+    },
+    {
+      "epoch": 0.48097826086956524,
+      "grad_norm": 1.6064239740371704,
+      "learning_rate": 3e-05,
+      "loss": 2.977,
+      "step": 1770
+    },
+    {
+      "epoch": 0.48125,
+      "grad_norm": 1.7928065061569214,
+      "learning_rate": 3e-05,
+      "loss": 3.5848,
+      "step": 1771
+    },
+    {
+      "epoch": 0.4815217391304348,
+      "grad_norm": 1.5981751680374146,
+      "learning_rate": 3e-05,
+      "loss": 3.1049,
+      "step": 1772
+    },
+    {
+      "epoch": 0.48179347826086955,
+      "grad_norm": 1.6446112394332886,
+      "learning_rate": 3e-05,
+      "loss": 2.9694,
+      "step": 1773
+    },
+    {
+      "epoch": 0.48206521739130437,
+      "grad_norm": 1.8991427421569824,
+      "learning_rate": 3e-05,
+      "loss": 3.3139,
+      "step": 1774
+    },
+    {
+      "epoch": 0.48233695652173914,
+      "grad_norm": 2.3716299533843994,
+      "learning_rate": 3e-05,
+      "loss": 3.9361,
+      "step": 1775
+    },
+    {
+      "epoch": 0.4826086956521739,
+      "grad_norm": 1.8573355674743652,
+      "learning_rate": 3e-05,
+      "loss": 3.1394,
+      "step": 1776
+    },
+    {
+      "epoch": 0.48288043478260867,
+      "grad_norm": 1.8625982999801636,
+      "learning_rate": 3e-05,
+      "loss": 3.8017,
+      "step": 1777
+    },
+    {
+      "epoch": 0.4831521739130435,
+      "grad_norm": 2.2275497913360596,
+      "learning_rate": 3e-05,
+      "loss": 3.4565,
+      "step": 1778
+    },
+    {
+      "epoch": 0.48342391304347826,
+      "grad_norm": 1.9856013059616089,
+      "learning_rate": 3e-05,
+      "loss": 3.5517,
+      "step": 1779
+    },
+    {
+      "epoch": 0.483695652173913,
+      "grad_norm": 1.3510781526565552,
+      "learning_rate": 3e-05,
+      "loss": 3.1976,
+      "step": 1780
+    },
+    {
+      "epoch": 0.48396739130434785,
+      "grad_norm": 1.0482627153396606,
+      "learning_rate": 3e-05,
+      "loss": 3.0025,
+      "step": 1781
+    },
+    {
+      "epoch": 0.4842391304347826,
+      "grad_norm": 1.3749173879623413,
+      "learning_rate": 3e-05,
+      "loss": 3.2774,
+      "step": 1782
+    },
+    {
+      "epoch": 0.4845108695652174,
+      "grad_norm": 1.405247449874878,
+      "learning_rate": 3e-05,
+      "loss": 3.4099,
+      "step": 1783
+    },
+    {
+      "epoch": 0.48478260869565215,
+      "grad_norm": 1.2630794048309326,
+      "learning_rate": 3e-05,
+      "loss": 3.7092,
+      "step": 1784
+    },
+    {
+      "epoch": 0.485054347826087,
+      "grad_norm": 1.1290115118026733,
+      "learning_rate": 3e-05,
+      "loss": 3.432,
+      "step": 1785
+    },
+    {
+      "epoch": 0.48532608695652174,
+      "grad_norm": 1.2523903846740723,
+      "learning_rate": 3e-05,
+      "loss": 3.1431,
+      "step": 1786
+    },
+    {
+      "epoch": 0.4855978260869565,
+      "grad_norm": 1.218867540359497,
+      "learning_rate": 3e-05,
+      "loss": 3.4881,
+      "step": 1787
+    },
+    {
+      "epoch": 0.48586956521739133,
+      "grad_norm": 1.6144737005233765,
+      "learning_rate": 3e-05,
+      "loss": 4.0674,
+      "step": 1788
+    },
+    {
+      "epoch": 0.4861413043478261,
+      "grad_norm": 1.4506441354751587,
+      "learning_rate": 3e-05,
+      "loss": 3.5498,
+      "step": 1789
+    },
+    {
+      "epoch": 0.48641304347826086,
+      "grad_norm": 1.5907642841339111,
+      "learning_rate": 3e-05,
+      "loss": 3.1341,
+      "step": 1790
+    },
+    {
+      "epoch": 0.48668478260869563,
+      "grad_norm": 1.18953275680542,
+      "learning_rate": 3e-05,
+      "loss": 3.0503,
+      "step": 1791
+    },
+    {
+      "epoch": 0.48695652173913045,
+      "grad_norm": 1.2186987400054932,
+      "learning_rate": 3e-05,
+      "loss": 2.7508,
+      "step": 1792
+    },
+    {
+      "epoch": 0.4872282608695652,
+      "grad_norm": 1.111574649810791,
+      "learning_rate": 3e-05,
+      "loss": 2.8275,
+      "step": 1793
+    },
+    {
+      "epoch": 0.4875,
+      "grad_norm": 1.275896430015564,
+      "learning_rate": 3e-05,
+      "loss": 3.4576,
+      "step": 1794
+    },
+    {
+      "epoch": 0.48777173913043476,
+      "grad_norm": 1.182626485824585,
+      "learning_rate": 3e-05,
+      "loss": 2.9678,
+      "step": 1795
+    },
+    {
+      "epoch": 0.4880434782608696,
+      "grad_norm": 1.098120927810669,
+      "learning_rate": 3e-05,
+      "loss": 3.1544,
+      "step": 1796
+    },
+    {
+      "epoch": 0.48831521739130435,
+      "grad_norm": 0.9524686932563782,
+      "learning_rate": 3e-05,
+      "loss": 3.0739,
+      "step": 1797
+    },
+    {
+      "epoch": 0.4885869565217391,
+      "grad_norm": 1.302868366241455,
+      "learning_rate": 3e-05,
+      "loss": 3.3036,
+      "step": 1798
+    },
+    {
+      "epoch": 0.48885869565217394,
+      "grad_norm": 1.124603271484375,
+      "learning_rate": 3e-05,
+      "loss": 3.2132,
+      "step": 1799
+    },
+    {
+      "epoch": 0.4891304347826087,
+      "grad_norm": 1.3126977682113647,
+      "learning_rate": 3e-05,
+      "loss": 3.6832,
+      "step": 1800
+    },
+    {
+      "epoch": 0.48940217391304347,
+      "grad_norm": 1.436482310295105,
+      "learning_rate": 3e-05,
+      "loss": 3.9607,
+      "step": 1801
+    },
+    {
+      "epoch": 0.48967391304347824,
+      "grad_norm": 1.6329675912857056,
+      "learning_rate": 3e-05,
+      "loss": 2.6751,
+      "step": 1802
+    },
+    {
+      "epoch": 0.48994565217391306,
+      "grad_norm": 1.2873584032058716,
+      "learning_rate": 3e-05,
+      "loss": 3.3793,
+      "step": 1803
+    },
+    {
+      "epoch": 0.4902173913043478,
+      "grad_norm": 1.0539244413375854,
+      "learning_rate": 3e-05,
+      "loss": 3.0761,
+      "step": 1804
+    },
+    {
+      "epoch": 0.4904891304347826,
+      "grad_norm": 1.7557803392410278,
+      "learning_rate": 3e-05,
+      "loss": 3.0488,
+      "step": 1805
+    },
+    {
+      "epoch": 0.4907608695652174,
+      "grad_norm": 1.8407940864562988,
+      "learning_rate": 3e-05,
+      "loss": 3.1567,
+      "step": 1806
+    },
+    {
+      "epoch": 0.4910326086956522,
+      "grad_norm": 1.3472580909729004,
+      "learning_rate": 3e-05,
+      "loss": 3.2331,
+      "step": 1807
+    },
+    {
+      "epoch": 0.49130434782608695,
+      "grad_norm": 1.093360424041748,
+      "learning_rate": 3e-05,
+      "loss": 3.1883,
+      "step": 1808
+    },
+    {
+      "epoch": 0.4915760869565217,
+      "grad_norm": 1.650109887123108,
+      "learning_rate": 3e-05,
+      "loss": 3.7661,
+      "step": 1809
+    },
+    {
+      "epoch": 0.49184782608695654,
+      "grad_norm": 1.454372525215149,
+      "learning_rate": 3e-05,
+      "loss": 3.1205,
+      "step": 1810
+    },
+    {
+      "epoch": 0.4921195652173913,
+      "grad_norm": 1.4370765686035156,
+      "learning_rate": 3e-05,
+      "loss": 3.8767,
+      "step": 1811
+    },
+    {
+      "epoch": 0.4923913043478261,
+      "grad_norm": 2.6512532234191895,
+      "learning_rate": 3e-05,
+      "loss": 4.0366,
+      "step": 1812
+    },
+    {
+      "epoch": 0.49266304347826084,
+      "grad_norm": 2.1347718238830566,
+      "learning_rate": 3e-05,
+      "loss": 3.5459,
+      "step": 1813
+    },
+    {
+      "epoch": 0.49293478260869567,
+      "grad_norm": 1.5831135511398315,
+      "learning_rate": 3e-05,
+      "loss": 3.4828,
+      "step": 1814
+    },
+    {
+      "epoch": 0.49320652173913043,
+      "grad_norm": 1.0648609399795532,
+      "learning_rate": 3e-05,
+      "loss": 3.0868,
+      "step": 1815
+    },
+    {
+      "epoch": 0.4934782608695652,
+      "grad_norm": 1.467032551765442,
+      "learning_rate": 3e-05,
+      "loss": 3.6537,
+      "step": 1816
+    },
+    {
+      "epoch": 0.49375,
+      "grad_norm": 1.2471240758895874,
+      "learning_rate": 3e-05,
+      "loss": 3.2671,
+      "step": 1817
+    },
+    {
+      "epoch": 0.4940217391304348,
+      "grad_norm": 0.9745358228683472,
+      "learning_rate": 3e-05,
+      "loss": 3.4916,
+      "step": 1818
+    },
+    {
+      "epoch": 0.49429347826086956,
+      "grad_norm": 1.3291053771972656,
+      "learning_rate": 3e-05,
+      "loss": 3.7441,
+      "step": 1819
+    },
+    {
+      "epoch": 0.4945652173913043,
+      "grad_norm": 1.3420668840408325,
+      "learning_rate": 3e-05,
+      "loss": 3.3106,
+      "step": 1820
+    },
+    {
+      "epoch": 0.49483695652173915,
+      "grad_norm": 1.0097370147705078,
+      "learning_rate": 3e-05,
+      "loss": 3.1215,
+      "step": 1821
+    },
+    {
+      "epoch": 0.4951086956521739,
+      "grad_norm": 1.1169931888580322,
+      "learning_rate": 3e-05,
+      "loss": 2.9595,
+      "step": 1822
+    },
+    {
+      "epoch": 0.4953804347826087,
+      "grad_norm": 1.1370187997817993,
+      "learning_rate": 3e-05,
+      "loss": 3.5224,
+      "step": 1823
+    },
+    {
+      "epoch": 0.4956521739130435,
+      "grad_norm": 1.5470448732376099,
+      "learning_rate": 3e-05,
+      "loss": 3.9269,
+      "step": 1824
+    },
+    {
+      "epoch": 0.49592391304347827,
+      "grad_norm": 1.000895380973816,
+      "learning_rate": 3e-05,
+      "loss": 2.9649,
+      "step": 1825
+    },
+    {
+      "epoch": 0.49619565217391304,
+      "grad_norm": 1.2407017946243286,
+      "learning_rate": 3e-05,
+      "loss": 3.5568,
+      "step": 1826
+    },
+    {
+      "epoch": 0.4964673913043478,
+      "grad_norm": 1.226485013961792,
+      "learning_rate": 3e-05,
+      "loss": 2.5796,
+      "step": 1827
+    },
+    {
+      "epoch": 0.4967391304347826,
+      "grad_norm": 1.1970787048339844,
+      "learning_rate": 3e-05,
+      "loss": 2.8772,
+      "step": 1828
+    },
+    {
+      "epoch": 0.4970108695652174,
+      "grad_norm": 1.2727575302124023,
+      "learning_rate": 3e-05,
+      "loss": 3.6501,
+      "step": 1829
+    },
+    {
+      "epoch": 0.49728260869565216,
+      "grad_norm": 1.2288799285888672,
+      "learning_rate": 3e-05,
+      "loss": 3.0078,
+      "step": 1830
+    },
+    {
+      "epoch": 0.49755434782608693,
+      "grad_norm": 1.688232660293579,
+      "learning_rate": 3e-05,
+      "loss": 3.2542,
+      "step": 1831
+    },
+    {
+      "epoch": 0.49782608695652175,
+      "grad_norm": 1.6907511949539185,
+      "learning_rate": 3e-05,
+      "loss": 3.6193,
+      "step": 1832
+    },
+    {
+      "epoch": 0.4980978260869565,
+      "grad_norm": 1.1391464471817017,
+      "learning_rate": 3e-05,
+      "loss": 3.3688,
+      "step": 1833
+    },
+    {
+      "epoch": 0.4983695652173913,
+      "grad_norm": 1.2691822052001953,
+      "learning_rate": 3e-05,
+      "loss": 3.2763,
+      "step": 1834
+    },
+    {
+      "epoch": 0.4986413043478261,
+      "grad_norm": 1.1176376342773438,
+      "learning_rate": 3e-05,
+      "loss": 3.0038,
+      "step": 1835
+    },
+    {
+      "epoch": 0.4989130434782609,
+      "grad_norm": 0.9173082709312439,
+      "learning_rate": 3e-05,
+      "loss": 2.7976,
+      "step": 1836
+    },
+    {
+      "epoch": 0.49918478260869564,
+      "grad_norm": 1.0989121198654175,
+      "learning_rate": 3e-05,
+      "loss": 3.0254,
+      "step": 1837
+    },
+    {
+      "epoch": 0.4994565217391304,
+      "grad_norm": 1.2885278463363647,
+      "learning_rate": 3e-05,
+      "loss": 3.2596,
+      "step": 1838
+    },
+    {
+      "epoch": 0.49972826086956523,
+      "grad_norm": 1.1294320821762085,
+      "learning_rate": 3e-05,
+      "loss": 3.0833,
+      "step": 1839
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9139266014099121,
+      "learning_rate": 3e-05,
+      "loss": 3.043,
+      "step": 1840
+    },
+    {
+      "epoch": 0.5002717391304348,
+      "grad_norm": 0.9999741315841675,
+      "learning_rate": 3e-05,
+      "loss": 2.6746,
+      "step": 1841
+    },
+    {
+      "epoch": 0.5005434782608695,
+      "grad_norm": 2.3439218997955322,
+      "learning_rate": 3e-05,
+      "loss": 4.0552,
+      "step": 1842
+    },
+    {
+      "epoch": 0.5008152173913043,
+      "grad_norm": 1.3185632228851318,
+      "learning_rate": 3e-05,
+      "loss": 3.4046,
+      "step": 1843
+    },
+    {
+      "epoch": 0.5010869565217392,
+      "grad_norm": 1.4254107475280762,
+      "learning_rate": 3e-05,
+      "loss": 3.5323,
+      "step": 1844
+    },
+    {
+      "epoch": 0.501358695652174,
+      "grad_norm": 1.0706696510314941,
+      "learning_rate": 3e-05,
+      "loss": 3.0672,
+      "step": 1845
+    },
+    {
+      "epoch": 0.5016304347826087,
+      "grad_norm": 1.316158413887024,
+      "learning_rate": 3e-05,
+      "loss": 3.3709,
+      "step": 1846
+    },
+    {
+      "epoch": 0.5019021739130435,
+      "grad_norm": 1.2480392456054688,
+      "learning_rate": 3e-05,
+      "loss": 3.673,
+      "step": 1847
+    },
+    {
+      "epoch": 0.5021739130434782,
+      "grad_norm": 1.596869707107544,
+      "learning_rate": 3e-05,
+      "loss": 3.3651,
+      "step": 1848
+    },
+    {
+      "epoch": 0.502445652173913,
+      "grad_norm": 1.6036261320114136,
+      "learning_rate": 3e-05,
+      "loss": 3.68,
+      "step": 1849
+    },
+    {
+      "epoch": 0.5027173913043478,
+      "grad_norm": 1.2420178651809692,
+      "learning_rate": 3e-05,
+      "loss": 2.9173,
+      "step": 1850
+    },
+    {
+      "epoch": 0.5029891304347827,
+      "grad_norm": 1.588339924812317,
+      "learning_rate": 3e-05,
+      "loss": 3.7931,
+      "step": 1851
+    },
+    {
+      "epoch": 0.5032608695652174,
+      "grad_norm": 1.2808369398117065,
+      "learning_rate": 3e-05,
+      "loss": 3.05,
+      "step": 1852
+    },
+    {
+      "epoch": 0.5035326086956522,
+      "grad_norm": 1.234026551246643,
+      "learning_rate": 3e-05,
+      "loss": 3.3401,
+      "step": 1853
+    },
+    {
+      "epoch": 0.503804347826087,
+      "grad_norm": 1.6384221315383911,
+      "learning_rate": 3e-05,
+      "loss": 3.5597,
+      "step": 1854
+    },
+    {
+      "epoch": 0.5040760869565217,
+      "grad_norm": 1.8604109287261963,
+      "learning_rate": 3e-05,
+      "loss": 3.6333,
+      "step": 1855
+    },
+    {
+      "epoch": 0.5043478260869565,
+      "grad_norm": 1.4895154237747192,
+      "learning_rate": 3e-05,
+      "loss": 3.4255,
+      "step": 1856
+    },
+    {
+      "epoch": 0.5046195652173913,
+      "grad_norm": 1.3410334587097168,
+      "learning_rate": 3e-05,
+      "loss": 2.7753,
+      "step": 1857
+    },
+    {
+      "epoch": 0.5048913043478261,
+      "grad_norm": 1.446807861328125,
+      "learning_rate": 3e-05,
+      "loss": 4.1229,
+      "step": 1858
+    },
+    {
+      "epoch": 0.5051630434782609,
+      "grad_norm": 1.188097596168518,
+      "learning_rate": 3e-05,
+      "loss": 3.2846,
+      "step": 1859
+    },
+    {
+      "epoch": 0.5054347826086957,
+      "grad_norm": 1.2120987176895142,
+      "learning_rate": 3e-05,
+      "loss": 3.147,
+      "step": 1860
+    },
+    {
+      "epoch": 0.5057065217391304,
+      "grad_norm": 1.5114277601242065,
+      "learning_rate": 3e-05,
+      "loss": 3.0952,
+      "step": 1861
+    },
+    {
+      "epoch": 0.5059782608695652,
+      "grad_norm": 1.360628366470337,
+      "learning_rate": 3e-05,
+      "loss": 2.907,
+      "step": 1862
+    },
+    {
+      "epoch": 0.50625,
+      "grad_norm": 1.4083906412124634,
+      "learning_rate": 3e-05,
+      "loss": 3.4222,
+      "step": 1863
+    },
+    {
+      "epoch": 0.5065217391304347,
+      "grad_norm": 1.245968222618103,
+      "learning_rate": 3e-05,
+      "loss": 3.5063,
+      "step": 1864
+    },
+    {
+      "epoch": 0.5067934782608695,
+      "grad_norm": 1.3542323112487793,
+      "learning_rate": 3e-05,
+      "loss": 3.981,
+      "step": 1865
+    },
+    {
+      "epoch": 0.5070652173913044,
+      "grad_norm": 2.4021403789520264,
+      "learning_rate": 3e-05,
+      "loss": 3.9367,
+      "step": 1866
+    },
+    {
+      "epoch": 0.5073369565217392,
+      "grad_norm": 1.4897315502166748,
+      "learning_rate": 3e-05,
+      "loss": 2.7825,
+      "step": 1867
+    },
+    {
+      "epoch": 0.5076086956521739,
+      "grad_norm": 1.7353469133377075,
+      "learning_rate": 3e-05,
+      "loss": 3.5709,
+      "step": 1868
+    },
+    {
+      "epoch": 0.5078804347826087,
+      "grad_norm": 1.216436505317688,
+      "learning_rate": 3e-05,
+      "loss": 3.1314,
+      "step": 1869
+    },
+    {
+      "epoch": 0.5081521739130435,
+      "grad_norm": 1.2940754890441895,
+      "learning_rate": 3e-05,
+      "loss": 3.3858,
+      "step": 1870
+    },
+    {
+      "epoch": 0.5084239130434782,
+      "grad_norm": 0.9794735312461853,
+      "learning_rate": 3e-05,
+      "loss": 2.823,
+      "step": 1871
+    },
+    {
+      "epoch": 0.508695652173913,
+      "grad_norm": 1.4511076211929321,
+      "learning_rate": 3e-05,
+      "loss": 4.0647,
+      "step": 1872
+    },
+    {
+      "epoch": 0.5089673913043479,
+      "grad_norm": 1.559597373008728,
+      "learning_rate": 3e-05,
+      "loss": 3.093,
+      "step": 1873
+    },
+    {
+      "epoch": 0.5092391304347826,
+      "grad_norm": 1.7877241373062134,
+      "learning_rate": 3e-05,
+      "loss": 3.742,
+      "step": 1874
+    },
+    {
+      "epoch": 0.5095108695652174,
+      "grad_norm": 1.585822582244873,
+      "learning_rate": 3e-05,
+      "loss": 3.631,
+      "step": 1875
+    },
+    {
+      "epoch": 0.5097826086956522,
+      "grad_norm": 1.284661054611206,
+      "learning_rate": 3e-05,
+      "loss": 3.228,
+      "step": 1876
+    },
+    {
+      "epoch": 0.5100543478260869,
+      "grad_norm": 1.295775294303894,
+      "learning_rate": 3e-05,
+      "loss": 3.7922,
+      "step": 1877
+    },
+    {
+      "epoch": 0.5103260869565217,
+      "grad_norm": 1.2768347263336182,
+      "learning_rate": 3e-05,
+      "loss": 3.0369,
+      "step": 1878
+    },
+    {
+      "epoch": 0.5105978260869565,
+      "grad_norm": 1.4773716926574707,
+      "learning_rate": 3e-05,
+      "loss": 3.3407,
+      "step": 1879
+    },
+    {
+      "epoch": 0.5108695652173914,
+      "grad_norm": 1.2089797258377075,
+      "learning_rate": 3e-05,
+      "loss": 3.0556,
+      "step": 1880
+    },
+    {
+      "epoch": 0.5111413043478261,
+      "grad_norm": 1.4668608903884888,
+      "learning_rate": 3e-05,
+      "loss": 3.4665,
+      "step": 1881
+    },
+    {
+      "epoch": 0.5114130434782609,
+      "grad_norm": 1.998399257659912,
+      "learning_rate": 3e-05,
+      "loss": 3.8042,
+      "step": 1882
+    },
+    {
+      "epoch": 0.5116847826086957,
+      "grad_norm": 1.448439598083496,
+      "learning_rate": 3e-05,
+      "loss": 3.7303,
+      "step": 1883
+    },
+    {
+      "epoch": 0.5119565217391304,
+      "grad_norm": 0.9362940192222595,
+      "learning_rate": 3e-05,
+      "loss": 2.665,
+      "step": 1884
+    },
+    {
+      "epoch": 0.5122282608695652,
+      "grad_norm": 1.6077910661697388,
+      "learning_rate": 3e-05,
+      "loss": 3.4826,
+      "step": 1885
+    },
+    {
+      "epoch": 0.5125,
+      "grad_norm": 1.5588127374649048,
+      "learning_rate": 3e-05,
+      "loss": 2.837,
+      "step": 1886
+    },
+    {
+      "epoch": 0.5127717391304348,
+      "grad_norm": 1.473461389541626,
+      "learning_rate": 3e-05,
+      "loss": 3.2511,
+      "step": 1887
+    },
+    {
+      "epoch": 0.5130434782608696,
+      "grad_norm": 1.6276698112487793,
+      "learning_rate": 3e-05,
+      "loss": 3.5821,
+      "step": 1888
+    },
+    {
+      "epoch": 0.5133152173913044,
+      "grad_norm": 1.6201423406600952,
+      "learning_rate": 3e-05,
+      "loss": 3.195,
+      "step": 1889
+    },
+    {
+      "epoch": 0.5135869565217391,
+      "grad_norm": 1.6411088705062866,
+      "learning_rate": 3e-05,
+      "loss": 3.1111,
+      "step": 1890
+    },
+    {
+      "epoch": 0.5138586956521739,
+      "grad_norm": 1.2973185777664185,
+      "learning_rate": 3e-05,
+      "loss": 2.6324,
+      "step": 1891
+    },
+    {
+      "epoch": 0.5141304347826087,
+      "grad_norm": 2.3508846759796143,
+      "learning_rate": 3e-05,
+      "loss": 3.7621,
+      "step": 1892
+    },
+    {
+      "epoch": 0.5144021739130434,
+      "grad_norm": 1.718246340751648,
+      "learning_rate": 3e-05,
+      "loss": 3.759,
+      "step": 1893
+    },
+    {
+      "epoch": 0.5146739130434783,
+      "grad_norm": 1.5278515815734863,
+      "learning_rate": 3e-05,
+      "loss": 3.2879,
+      "step": 1894
+    },
+    {
+      "epoch": 0.5149456521739131,
+      "grad_norm": 1.308929681777954,
+      "learning_rate": 3e-05,
+      "loss": 3.1075,
+      "step": 1895
+    },
+    {
+      "epoch": 0.5152173913043478,
+      "grad_norm": 1.40326988697052,
+      "learning_rate": 3e-05,
+      "loss": 3.5211,
+      "step": 1896
+    },
+    {
+      "epoch": 0.5154891304347826,
+      "grad_norm": 1.5129557847976685,
+      "learning_rate": 3e-05,
+      "loss": 3.2549,
+      "step": 1897
+    },
+    {
+      "epoch": 0.5157608695652174,
+      "grad_norm": 1.1121063232421875,
+      "learning_rate": 3e-05,
+      "loss": 2.9461,
+      "step": 1898
+    },
+    {
+      "epoch": 0.5160326086956522,
+      "grad_norm": 1.4302382469177246,
+      "learning_rate": 3e-05,
+      "loss": 3.627,
+      "step": 1899
+    },
+    {
+      "epoch": 0.5163043478260869,
+      "grad_norm": 1.6184173822402954,
+      "learning_rate": 3e-05,
+      "loss": 2.9157,
+      "step": 1900
+    },
+    {
+      "epoch": 0.5165760869565217,
+      "grad_norm": 1.1715192794799805,
+      "learning_rate": 3e-05,
+      "loss": 2.9425,
+      "step": 1901
+    },
+    {
+      "epoch": 0.5168478260869566,
+      "grad_norm": 2.065067768096924,
+      "learning_rate": 3e-05,
+      "loss": 3.3198,
+      "step": 1902
+    },
+    {
+      "epoch": 0.5171195652173913,
+      "grad_norm": 1.6942062377929688,
+      "learning_rate": 3e-05,
+      "loss": 3.4998,
+      "step": 1903
+    },
+    {
+      "epoch": 0.5173913043478261,
+      "grad_norm": 1.4067577123641968,
+      "learning_rate": 3e-05,
+      "loss": 3.501,
+      "step": 1904
+    },
+    {
+      "epoch": 0.5176630434782609,
+      "grad_norm": 1.5794352293014526,
+      "learning_rate": 3e-05,
+      "loss": 3.1307,
+      "step": 1905
+    },
+    {
+      "epoch": 0.5179347826086956,
+      "grad_norm": 1.5043370723724365,
+      "learning_rate": 3e-05,
+      "loss": 3.8714,
+      "step": 1906
+    },
+    {
+      "epoch": 0.5182065217391304,
+      "grad_norm": 1.2884905338287354,
+      "learning_rate": 3e-05,
+      "loss": 3.4459,
+      "step": 1907
+    },
+    {
+      "epoch": 0.5184782608695652,
+      "grad_norm": 1.3515671491622925,
+      "learning_rate": 3e-05,
+      "loss": 3.5685,
+      "step": 1908
+    },
+    {
+      "epoch": 0.51875,
+      "grad_norm": 1.2115263938903809,
+      "learning_rate": 3e-05,
+      "loss": 3.2234,
+      "step": 1909
+    },
+    {
+      "epoch": 0.5190217391304348,
+      "grad_norm": 1.3323924541473389,
+      "learning_rate": 3e-05,
+      "loss": 3.42,
+      "step": 1910
+    },
+    {
+      "epoch": 0.5192934782608696,
+      "grad_norm": 1.2837159633636475,
+      "learning_rate": 3e-05,
+      "loss": 3.0186,
+      "step": 1911
+    },
+    {
+      "epoch": 0.5195652173913043,
+      "grad_norm": 1.1012318134307861,
+      "learning_rate": 3e-05,
+      "loss": 2.9993,
+      "step": 1912
+    },
+    {
+      "epoch": 0.5198369565217391,
+      "grad_norm": 1.253250002861023,
+      "learning_rate": 3e-05,
+      "loss": 3.371,
+      "step": 1913
+    },
+    {
+      "epoch": 0.5201086956521739,
+      "grad_norm": 0.9867514967918396,
+      "learning_rate": 3e-05,
+      "loss": 2.7761,
+      "step": 1914
+    },
+    {
+      "epoch": 0.5203804347826086,
+      "grad_norm": 1.0671703815460205,
+      "learning_rate": 3e-05,
+      "loss": 3.5288,
+      "step": 1915
+    },
+    {
+      "epoch": 0.5206521739130435,
+      "grad_norm": 0.9514213800430298,
+      "learning_rate": 3e-05,
+      "loss": 2.7831,
+      "step": 1916
+    },
+    {
+      "epoch": 0.5209239130434783,
+      "grad_norm": 1.3238492012023926,
+      "learning_rate": 3e-05,
+      "loss": 3.1874,
+      "step": 1917
+    },
+    {
+      "epoch": 0.5211956521739131,
+      "grad_norm": 1.5408774614334106,
+      "learning_rate": 3e-05,
+      "loss": 3.1997,
+      "step": 1918
+    },
+    {
+      "epoch": 0.5214673913043478,
+      "grad_norm": 1.1520451307296753,
+      "learning_rate": 3e-05,
+      "loss": 3.1287,
+      "step": 1919
+    },
+    {
+      "epoch": 0.5217391304347826,
+      "grad_norm": 1.5356744527816772,
+      "learning_rate": 3e-05,
+      "loss": 3.5281,
+      "step": 1920
+    },
+    {
+      "epoch": 0.5220108695652174,
+      "grad_norm": 1.0727752447128296,
+      "learning_rate": 3e-05,
+      "loss": 3.0281,
+      "step": 1921
+    },
+    {
+      "epoch": 0.5222826086956521,
+      "grad_norm": 1.4737282991409302,
+      "learning_rate": 3e-05,
+      "loss": 4.0763,
+      "step": 1922
+    },
+    {
+      "epoch": 0.522554347826087,
+      "grad_norm": 1.2212142944335938,
+      "learning_rate": 3e-05,
+      "loss": 3.6583,
+      "step": 1923
+    },
+    {
+      "epoch": 0.5228260869565218,
+      "grad_norm": 1.42307710647583,
+      "learning_rate": 3e-05,
+      "loss": 3.9908,
+      "step": 1924
+    },
+    {
+      "epoch": 0.5230978260869565,
+      "grad_norm": 1.2455291748046875,
+      "learning_rate": 3e-05,
+      "loss": 2.7686,
+      "step": 1925
+    },
+    {
+      "epoch": 0.5233695652173913,
+      "grad_norm": 1.1880292892456055,
+      "learning_rate": 3e-05,
+      "loss": 2.8165,
+      "step": 1926
+    },
+    {
+      "epoch": 0.5236413043478261,
+      "grad_norm": 1.329270362854004,
+      "learning_rate": 3e-05,
+      "loss": 3.421,
+      "step": 1927
+    },
+    {
+      "epoch": 0.5239130434782608,
+      "grad_norm": 1.8960901498794556,
+      "learning_rate": 3e-05,
+      "loss": 4.0437,
+      "step": 1928
+    },
+    {
+      "epoch": 0.5241847826086956,
+      "grad_norm": 1.223769187927246,
+      "learning_rate": 3e-05,
+      "loss": 3.3202,
+      "step": 1929
+    },
+    {
+      "epoch": 0.5244565217391305,
+      "grad_norm": 1.5670809745788574,
+      "learning_rate": 3e-05,
+      "loss": 3.6805,
+      "step": 1930
+    },
+    {
+      "epoch": 0.5247282608695653,
+      "grad_norm": 1.9276920557022095,
+      "learning_rate": 3e-05,
+      "loss": 3.4752,
+      "step": 1931
+    },
+    {
+      "epoch": 0.525,
+      "grad_norm": 1.7917790412902832,
+      "learning_rate": 3e-05,
+      "loss": 3.4745,
+      "step": 1932
+    },
+    {
+      "epoch": 0.5252717391304348,
+      "grad_norm": 1.1239231824874878,
+      "learning_rate": 3e-05,
+      "loss": 2.9622,
+      "step": 1933
+    },
+    {
+      "epoch": 0.5255434782608696,
+      "grad_norm": 1.2177050113677979,
+      "learning_rate": 3e-05,
+      "loss": 2.691,
+      "step": 1934
+    },
+    {
+      "epoch": 0.5258152173913043,
+      "grad_norm": 1.1526474952697754,
+      "learning_rate": 3e-05,
+      "loss": 2.7313,
+      "step": 1935
+    },
+    {
+      "epoch": 0.5260869565217391,
+      "grad_norm": 1.3403104543685913,
+      "learning_rate": 3e-05,
+      "loss": 3.215,
+      "step": 1936
+    },
+    {
+      "epoch": 0.5263586956521739,
+      "grad_norm": 1.1823958158493042,
+      "learning_rate": 3e-05,
+      "loss": 3.0382,
+      "step": 1937
+    },
+    {
+      "epoch": 0.5266304347826087,
+      "grad_norm": 1.2578941583633423,
+      "learning_rate": 3e-05,
+      "loss": 2.9962,
+      "step": 1938
+    },
+    {
+      "epoch": 0.5269021739130435,
+      "grad_norm": 1.2342561483383179,
+      "learning_rate": 3e-05,
+      "loss": 3.6362,
+      "step": 1939
+    },
+    {
+      "epoch": 0.5271739130434783,
+      "grad_norm": 1.0608114004135132,
+      "learning_rate": 3e-05,
+      "loss": 3.2318,
+      "step": 1940
+    },
+    {
+      "epoch": 0.527445652173913,
+      "grad_norm": 1.2502753734588623,
+      "learning_rate": 3e-05,
+      "loss": 3.1968,
+      "step": 1941
+    },
+    {
+      "epoch": 0.5277173913043478,
+      "grad_norm": 0.9138715267181396,
+      "learning_rate": 3e-05,
+      "loss": 3.0353,
+      "step": 1942
+    },
+    {
+      "epoch": 0.5279891304347826,
+      "grad_norm": 1.257987380027771,
+      "learning_rate": 3e-05,
+      "loss": 2.5984,
+      "step": 1943
+    },
+    {
+      "epoch": 0.5282608695652173,
+      "grad_norm": 1.5247408151626587,
+      "learning_rate": 3e-05,
+      "loss": 3.4343,
+      "step": 1944
+    },
+    {
+      "epoch": 0.5285326086956522,
+      "grad_norm": 1.1517291069030762,
+      "learning_rate": 3e-05,
+      "loss": 2.7655,
+      "step": 1945
+    },
+    {
+      "epoch": 0.528804347826087,
+      "grad_norm": 1.5516606569290161,
+      "learning_rate": 3e-05,
+      "loss": 3.9976,
+      "step": 1946
+    },
+    {
+      "epoch": 0.5290760869565218,
+      "grad_norm": 1.6808713674545288,
+      "learning_rate": 3e-05,
+      "loss": 3.6858,
+      "step": 1947
+    },
+    {
+      "epoch": 0.5293478260869565,
+      "grad_norm": 1.0476493835449219,
+      "learning_rate": 3e-05,
+      "loss": 2.8933,
+      "step": 1948
+    },
+    {
+      "epoch": 0.5296195652173913,
+      "grad_norm": 1.7297437191009521,
+      "learning_rate": 3e-05,
+      "loss": 3.1754,
+      "step": 1949
+    },
+    {
+      "epoch": 0.529891304347826,
+      "grad_norm": 1.4908910989761353,
+      "learning_rate": 3e-05,
+      "loss": 3.9229,
+      "step": 1950
+    },
+    {
+      "epoch": 0.5301630434782608,
+      "grad_norm": 1.3128184080123901,
+      "learning_rate": 3e-05,
+      "loss": 3.5215,
+      "step": 1951
+    },
+    {
+      "epoch": 0.5304347826086957,
+      "grad_norm": 1.7025582790374756,
+      "learning_rate": 3e-05,
+      "loss": 3.5113,
+      "step": 1952
+    },
+    {
+      "epoch": 0.5307065217391305,
+      "grad_norm": 1.4770921468734741,
+      "learning_rate": 3e-05,
+      "loss": 3.0171,
+      "step": 1953
+    },
+    {
+      "epoch": 0.5309782608695652,
+      "grad_norm": 1.6477012634277344,
+      "learning_rate": 3e-05,
+      "loss": 3.5914,
+      "step": 1954
+    },
+    {
+      "epoch": 0.53125,
+      "grad_norm": 1.5680129528045654,
+      "learning_rate": 3e-05,
+      "loss": 3.6162,
+      "step": 1955
+    },
+    {
+      "epoch": 0.5315217391304348,
+      "grad_norm": 2.2198116779327393,
+      "learning_rate": 3e-05,
+      "loss": 3.2198,
+      "step": 1956
+    },
+    {
+      "epoch": 0.5317934782608695,
+      "grad_norm": 1.9690414667129517,
+      "learning_rate": 3e-05,
+      "loss": 3.0914,
+      "step": 1957
+    },
+    {
+      "epoch": 0.5320652173913043,
+      "grad_norm": 1.4856075048446655,
+      "learning_rate": 3e-05,
+      "loss": 3.206,
+      "step": 1958
+    },
+    {
+      "epoch": 0.5323369565217392,
+      "grad_norm": 1.1297136545181274,
+      "learning_rate": 3e-05,
+      "loss": 2.8479,
+      "step": 1959
+    },
+    {
+      "epoch": 0.532608695652174,
+      "grad_norm": 1.859328269958496,
+      "learning_rate": 3e-05,
+      "loss": 3.3152,
+      "step": 1960
+    },
+    {
+      "epoch": 0.5328804347826087,
+      "grad_norm": 1.939894199371338,
+      "learning_rate": 3e-05,
+      "loss": 3.2557,
+      "step": 1961
+    },
+    {
+      "epoch": 0.5331521739130435,
+      "grad_norm": 2.148735523223877,
+      "learning_rate": 3e-05,
+      "loss": 3.6686,
+      "step": 1962
+    },
+    {
+      "epoch": 0.5334239130434782,
+      "grad_norm": 1.4653475284576416,
+      "learning_rate": 3e-05,
+      "loss": 3.8618,
+      "step": 1963
+    },
+    {
+      "epoch": 0.533695652173913,
+      "grad_norm": 2.1480512619018555,
+      "learning_rate": 3e-05,
+      "loss": 3.4809,
+      "step": 1964
+    },
+    {
+      "epoch": 0.5339673913043478,
+      "grad_norm": 2.2585768699645996,
+      "learning_rate": 3e-05,
+      "loss": 3.1685,
+      "step": 1965
+    },
+    {
+      "epoch": 0.5342391304347827,
+      "grad_norm": 2.143669605255127,
+      "learning_rate": 3e-05,
+      "loss": 3.8206,
+      "step": 1966
+    },
+    {
+      "epoch": 0.5345108695652174,
+      "grad_norm": 1.5735530853271484,
+      "learning_rate": 3e-05,
+      "loss": 3.5483,
+      "step": 1967
+    },
+    {
+      "epoch": 0.5347826086956522,
+      "grad_norm": 1.5427213907241821,
+      "learning_rate": 3e-05,
+      "loss": 3.469,
+      "step": 1968
+    },
+    {
+      "epoch": 0.535054347826087,
+      "grad_norm": 1.7961771488189697,
+      "learning_rate": 3e-05,
+      "loss": 3.2229,
+      "step": 1969
+    },
+    {
+      "epoch": 0.5353260869565217,
+      "grad_norm": 1.5997658967971802,
+      "learning_rate": 3e-05,
+      "loss": 3.6575,
+      "step": 1970
+    },
+    {
+      "epoch": 0.5355978260869565,
+      "grad_norm": 1.2979601621627808,
+      "learning_rate": 3e-05,
+      "loss": 3.7012,
+      "step": 1971
+    },
+    {
+      "epoch": 0.5358695652173913,
+      "grad_norm": 1.5162733793258667,
+      "learning_rate": 3e-05,
+      "loss": 3.1471,
+      "step": 1972
+    },
+    {
+      "epoch": 0.5361413043478261,
+      "grad_norm": 1.5043525695800781,
+      "learning_rate": 3e-05,
+      "loss": 2.8777,
+      "step": 1973
+    },
+    {
+      "epoch": 0.5364130434782609,
+      "grad_norm": 1.1250759363174438,
+      "learning_rate": 3e-05,
+      "loss": 2.9937,
+      "step": 1974
+    },
+    {
+      "epoch": 0.5366847826086957,
+      "grad_norm": 1.2006231546401978,
+      "learning_rate": 3e-05,
+      "loss": 3.5737,
+      "step": 1975
+    },
+    {
+      "epoch": 0.5369565217391304,
+      "grad_norm": 0.9950990080833435,
+      "learning_rate": 3e-05,
+      "loss": 2.9763,
+      "step": 1976
+    },
+    {
+      "epoch": 0.5372282608695652,
+      "grad_norm": 1.2111644744873047,
+      "learning_rate": 3e-05,
+      "loss": 3.5871,
+      "step": 1977
+    },
+    {
+      "epoch": 0.5375,
+      "grad_norm": 1.4334638118743896,
+      "learning_rate": 3e-05,
+      "loss": 3.645,
+      "step": 1978
+    },
+    {
+      "epoch": 0.5377717391304347,
+      "grad_norm": 1.376451849937439,
+      "learning_rate": 3e-05,
+      "loss": 3.5722,
+      "step": 1979
+    },
+    {
+      "epoch": 0.5380434782608695,
+      "grad_norm": 1.1852439641952515,
+      "learning_rate": 3e-05,
+      "loss": 3.2733,
+      "step": 1980
+    },
+    {
+      "epoch": 0.5383152173913044,
+      "grad_norm": 1.208321452140808,
+      "learning_rate": 3e-05,
+      "loss": 3.4223,
+      "step": 1981
+    },
+    {
+      "epoch": 0.5385869565217392,
+      "grad_norm": 1.2566808462142944,
+      "learning_rate": 3e-05,
+      "loss": 3.227,
+      "step": 1982
+    },
+    {
+      "epoch": 0.5388586956521739,
+      "grad_norm": 1.1662538051605225,
+      "learning_rate": 3e-05,
+      "loss": 3.9731,
+      "step": 1983
+    },
+    {
+      "epoch": 0.5391304347826087,
+      "grad_norm": 0.9991174340248108,
+      "learning_rate": 3e-05,
+      "loss": 3.2856,
+      "step": 1984
+    },
+    {
+      "epoch": 0.5394021739130435,
+      "grad_norm": 1.6792123317718506,
+      "learning_rate": 3e-05,
+      "loss": 4.0315,
+      "step": 1985
+    },
+    {
+      "epoch": 0.5396739130434782,
+      "grad_norm": 1.3554564714431763,
+      "learning_rate": 3e-05,
+      "loss": 2.7884,
+      "step": 1986
+    },
+    {
+      "epoch": 0.539945652173913,
+      "grad_norm": 2.0495409965515137,
+      "learning_rate": 3e-05,
+      "loss": 4.0834,
+      "step": 1987
+    },
+    {
+      "epoch": 0.5402173913043479,
+      "grad_norm": 1.3076777458190918,
+      "learning_rate": 3e-05,
+      "loss": 2.9288,
+      "step": 1988
+    },
+    {
+      "epoch": 0.5404891304347826,
+      "grad_norm": 1.2214155197143555,
+      "learning_rate": 3e-05,
+      "loss": 2.9785,
+      "step": 1989
+    },
+    {
+      "epoch": 0.5407608695652174,
+      "grad_norm": 1.6654658317565918,
+      "learning_rate": 3e-05,
+      "loss": 3.6795,
+      "step": 1990
+    },
+    {
+      "epoch": 0.5410326086956522,
+      "grad_norm": 1.224678874015808,
+      "learning_rate": 3e-05,
+      "loss": 2.9736,
+      "step": 1991
+    },
+    {
+      "epoch": 0.5413043478260869,
+      "grad_norm": 1.128450870513916,
+      "learning_rate": 3e-05,
+      "loss": 3.0371,
+      "step": 1992
+    },
+    {
+      "epoch": 0.5415760869565217,
+      "grad_norm": 1.1452875137329102,
+      "learning_rate": 3e-05,
+      "loss": 2.989,
+      "step": 1993
+    },
+    {
+      "epoch": 0.5418478260869565,
+      "grad_norm": 1.2154463529586792,
+      "learning_rate": 3e-05,
+      "loss": 2.9404,
+      "step": 1994
+    },
+    {
+      "epoch": 0.5421195652173914,
+      "grad_norm": 1.1099878549575806,
+      "learning_rate": 3e-05,
+      "loss": 3.1758,
+      "step": 1995
+    },
+    {
+      "epoch": 0.5423913043478261,
+      "grad_norm": 0.9395751953125,
+      "learning_rate": 3e-05,
+      "loss": 2.9722,
+      "step": 1996
+    },
+    {
+      "epoch": 0.5426630434782609,
+      "grad_norm": 1.1049753427505493,
+      "learning_rate": 3e-05,
+      "loss": 3.0526,
+      "step": 1997
+    },
+    {
+      "epoch": 0.5429347826086957,
+      "grad_norm": 1.45954430103302,
+      "learning_rate": 3e-05,
+      "loss": 3.7409,
+      "step": 1998
+    },
+    {
+      "epoch": 0.5432065217391304,
+      "grad_norm": 1.2864134311676025,
+      "learning_rate": 3e-05,
+      "loss": 3.4088,
+      "step": 1999
+    },
+    {
+      "epoch": 0.5434782608695652,
+      "grad_norm": 1.2165127992630005,
+      "learning_rate": 3e-05,
+      "loss": 3.2604,
+      "step": 2000
+    },
+    {
+      "epoch": 0.54375,
+      "grad_norm": 1.272796869277954,
+      "learning_rate": 3e-05,
+      "loss": 3.6574,
+      "step": 2001
+    },
+    {
+      "epoch": 0.5440217391304348,
+      "grad_norm": 1.3245208263397217,
+      "learning_rate": 3e-05,
+      "loss": 3.1406,
+      "step": 2002
+    },
+    {
+      "epoch": 0.5442934782608696,
+      "grad_norm": 1.4110616445541382,
+      "learning_rate": 3e-05,
+      "loss": 2.8426,
+      "step": 2003
+    },
+    {
+      "epoch": 0.5445652173913044,
+      "grad_norm": 1.7510126829147339,
+      "learning_rate": 3e-05,
+      "loss": 3.926,
+      "step": 2004
+    },
+    {
+      "epoch": 0.5448369565217391,
+      "grad_norm": 1.226434588432312,
+      "learning_rate": 3e-05,
+      "loss": 3.5155,
+      "step": 2005
+    },
+    {
+      "epoch": 0.5451086956521739,
+      "grad_norm": 1.5343637466430664,
+      "learning_rate": 3e-05,
+      "loss": 2.7504,
+      "step": 2006
+    },
+    {
+      "epoch": 0.5453804347826087,
+      "grad_norm": 1.3316307067871094,
+      "learning_rate": 3e-05,
+      "loss": 3.0848,
+      "step": 2007
+    },
+    {
+      "epoch": 0.5456521739130434,
+      "grad_norm": 1.2327113151550293,
+      "learning_rate": 3e-05,
+      "loss": 3.0306,
+      "step": 2008
+    },
+    {
+      "epoch": 0.5459239130434783,
+      "grad_norm": 1.180682897567749,
+      "learning_rate": 3e-05,
+      "loss": 3.0591,
+      "step": 2009
+    },
+    {
+      "epoch": 0.5461956521739131,
+      "grad_norm": 1.3480994701385498,
+      "learning_rate": 3e-05,
+      "loss": 3.3818,
+      "step": 2010
+    },
+    {
+      "epoch": 0.5464673913043478,
+      "grad_norm": 1.246237874031067,
+      "learning_rate": 3e-05,
+      "loss": 3.3416,
+      "step": 2011
+    },
+    {
+      "epoch": 0.5467391304347826,
+      "grad_norm": 0.8937397003173828,
+      "learning_rate": 3e-05,
+      "loss": 3.0593,
+      "step": 2012
+    },
+    {
+      "epoch": 0.5470108695652174,
+      "grad_norm": 1.3019452095031738,
+      "learning_rate": 3e-05,
+      "loss": 3.2755,
+      "step": 2013
+    },
+    {
+      "epoch": 0.5472826086956522,
+      "grad_norm": 1.408234715461731,
+      "learning_rate": 3e-05,
+      "loss": 3.4283,
+      "step": 2014
+    },
+    {
+      "epoch": 0.5475543478260869,
+      "grad_norm": 1.5709846019744873,
+      "learning_rate": 3e-05,
+      "loss": 3.2933,
+      "step": 2015
+    },
+    {
+      "epoch": 0.5478260869565217,
+      "grad_norm": 1.6085715293884277,
+      "learning_rate": 3e-05,
+      "loss": 3.2472,
+      "step": 2016
+    },
+    {
+      "epoch": 0.5480978260869566,
+      "grad_norm": 1.2559678554534912,
+      "learning_rate": 3e-05,
+      "loss": 2.8599,
+      "step": 2017
+    },
+    {
+      "epoch": 0.5483695652173913,
+      "grad_norm": 1.272606372833252,
+      "learning_rate": 3e-05,
+      "loss": 3.7284,
+      "step": 2018
+    },
+    {
+      "epoch": 0.5486413043478261,
+      "grad_norm": 1.098386287689209,
+      "learning_rate": 3e-05,
+      "loss": 3.0488,
+      "step": 2019
+    },
+    {
+      "epoch": 0.5489130434782609,
+      "grad_norm": 1.707297444343567,
+      "learning_rate": 3e-05,
+      "loss": 3.122,
+      "step": 2020
+    },
+    {
+      "epoch": 0.5491847826086956,
+      "grad_norm": 1.7398796081542969,
+      "learning_rate": 3e-05,
+      "loss": 3.9982,
+      "step": 2021
+    },
+    {
+      "epoch": 0.5494565217391304,
+      "grad_norm": 1.1505324840545654,
+      "learning_rate": 3e-05,
+      "loss": 3.6328,
+      "step": 2022
+    },
+    {
+      "epoch": 0.5497282608695652,
+      "grad_norm": 0.978933572769165,
+      "learning_rate": 3e-05,
+      "loss": 2.8985,
+      "step": 2023
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2565099000930786,
+      "learning_rate": 3e-05,
+      "loss": 3.1667,
+      "step": 2024
+    },
+    {
+      "epoch": 0.5502717391304348,
+      "grad_norm": 1.2417691946029663,
+      "learning_rate": 3e-05,
+      "loss": 3.5036,
+      "step": 2025
+    },
+    {
+      "epoch": 0.5505434782608696,
+      "grad_norm": 1.2185314893722534,
+      "learning_rate": 3e-05,
+      "loss": 3.1685,
+      "step": 2026
+    },
+    {
+      "epoch": 0.5508152173913043,
+      "grad_norm": 1.4619181156158447,
+      "learning_rate": 3e-05,
+      "loss": 4.2336,
+      "step": 2027
+    },
+    {
+      "epoch": 0.5510869565217391,
+      "grad_norm": 1.2455370426177979,
+      "learning_rate": 3e-05,
+      "loss": 3.6007,
+      "step": 2028
+    },
+    {
+      "epoch": 0.5513586956521739,
+      "grad_norm": 1.4226831197738647,
+      "learning_rate": 3e-05,
+      "loss": 3.7129,
+      "step": 2029
+    },
+    {
+      "epoch": 0.5516304347826086,
+      "grad_norm": 1.2961459159851074,
+      "learning_rate": 3e-05,
+      "loss": 3.2041,
+      "step": 2030
+    },
+    {
+      "epoch": 0.5519021739130435,
+      "grad_norm": 1.259477138519287,
+      "learning_rate": 3e-05,
+      "loss": 3.1745,
+      "step": 2031
+    },
+    {
+      "epoch": 0.5521739130434783,
+      "grad_norm": 1.691286325454712,
+      "learning_rate": 3e-05,
+      "loss": 3.5502,
+      "step": 2032
+    },
+    {
+      "epoch": 0.5524456521739131,
+      "grad_norm": 1.3523718118667603,
+      "learning_rate": 3e-05,
+      "loss": 3.4055,
+      "step": 2033
+    },
+    {
+      "epoch": 0.5527173913043478,
+      "grad_norm": 1.2605832815170288,
+      "learning_rate": 3e-05,
+      "loss": 2.7262,
+      "step": 2034
+    },
+    {
+      "epoch": 0.5529891304347826,
+      "grad_norm": 1.440855860710144,
+      "learning_rate": 3e-05,
+      "loss": 2.9385,
+      "step": 2035
+    },
+    {
+      "epoch": 0.5532608695652174,
+      "grad_norm": 2.026885986328125,
+      "learning_rate": 3e-05,
+      "loss": 4.3039,
+      "step": 2036
+    },
+    {
+      "epoch": 0.5535326086956521,
+      "grad_norm": 1.1887251138687134,
+      "learning_rate": 3e-05,
+      "loss": 2.8882,
+      "step": 2037
+    },
+    {
+      "epoch": 0.553804347826087,
+      "grad_norm": 1.322061538696289,
+      "learning_rate": 3e-05,
+      "loss": 3.2703,
+      "step": 2038
+    },
+    {
+      "epoch": 0.5540760869565218,
+      "grad_norm": 1.3264415264129639,
+      "learning_rate": 3e-05,
+      "loss": 3.2199,
+      "step": 2039
+    },
+    {
+      "epoch": 0.5543478260869565,
+      "grad_norm": 1.6366621255874634,
+      "learning_rate": 3e-05,
+      "loss": 3.6402,
+      "step": 2040
+    },
+    {
+      "epoch": 0.5546195652173913,
+      "grad_norm": 1.592647910118103,
+      "learning_rate": 3e-05,
+      "loss": 3.2358,
+      "step": 2041
+    },
+    {
+      "epoch": 0.5548913043478261,
+      "grad_norm": 1.4537502527236938,
+      "learning_rate": 3e-05,
+      "loss": 3.7269,
+      "step": 2042
+    },
+    {
+      "epoch": 0.5551630434782608,
+      "grad_norm": 1.3483185768127441,
+      "learning_rate": 3e-05,
+      "loss": 3.2456,
+      "step": 2043
+    },
+    {
+      "epoch": 0.5554347826086956,
+      "grad_norm": 1.271835207939148,
+      "learning_rate": 3e-05,
+      "loss": 3.5684,
+      "step": 2044
+    },
+    {
+      "epoch": 0.5557065217391305,
+      "grad_norm": 1.2505134344100952,
+      "learning_rate": 3e-05,
+      "loss": 3.2272,
+      "step": 2045
+    },
+    {
+      "epoch": 0.5559782608695653,
+      "grad_norm": 1.3669381141662598,
+      "learning_rate": 3e-05,
+      "loss": 2.8729,
+      "step": 2046
+    },
+    {
+      "epoch": 0.55625,
+      "grad_norm": 1.1156485080718994,
+      "learning_rate": 3e-05,
+      "loss": 3.003,
+      "step": 2047
+    },
+    {
+      "epoch": 0.5565217391304348,
+      "grad_norm": 1.458754301071167,
+      "learning_rate": 3e-05,
+      "loss": 3.0448,
+      "step": 2048
+    },
+    {
+      "epoch": 0.5567934782608696,
+      "grad_norm": 1.1323773860931396,
+      "learning_rate": 3e-05,
+      "loss": 3.1986,
+      "step": 2049
+    },
+    {
+      "epoch": 0.5570652173913043,
+      "grad_norm": 1.1786192655563354,
+      "learning_rate": 3e-05,
+      "loss": 3.6913,
+      "step": 2050
+    },
+    {
+      "epoch": 0.5573369565217391,
+      "grad_norm": 1.2700916528701782,
+      "learning_rate": 3e-05,
+      "loss": 2.9963,
+      "step": 2051
+    },
+    {
+      "epoch": 0.5576086956521739,
+      "grad_norm": 1.21451735496521,
+      "learning_rate": 3e-05,
+      "loss": 3.0128,
+      "step": 2052
+    },
+    {
+      "epoch": 0.5578804347826087,
+      "grad_norm": 1.5701782703399658,
+      "learning_rate": 3e-05,
+      "loss": 2.8858,
+      "step": 2053
+    },
+    {
+      "epoch": 0.5581521739130435,
+      "grad_norm": 1.7219531536102295,
+      "learning_rate": 3e-05,
+      "loss": 2.8876,
+      "step": 2054
+    },
+    {
+      "epoch": 0.5584239130434783,
+      "grad_norm": 1.694474220275879,
+      "learning_rate": 3e-05,
+      "loss": 3.1169,
+      "step": 2055
+    },
+    {
+      "epoch": 0.558695652173913,
+      "grad_norm": 1.684638500213623,
+      "learning_rate": 3e-05,
+      "loss": 3.9718,
+      "step": 2056
+    },
+    {
+      "epoch": 0.5589673913043478,
+      "grad_norm": 1.8228230476379395,
+      "learning_rate": 3e-05,
+      "loss": 3.0664,
+      "step": 2057
+    },
+    {
+      "epoch": 0.5592391304347826,
+      "grad_norm": 1.8484151363372803,
+      "learning_rate": 3e-05,
+      "loss": 3.148,
+      "step": 2058
+    },
+    {
+      "epoch": 0.5595108695652173,
+      "grad_norm": 1.6520676612854004,
+      "learning_rate": 3e-05,
+      "loss": 3.6549,
+      "step": 2059
+    },
+    {
+      "epoch": 0.5597826086956522,
+      "grad_norm": 1.4072345495224,
+      "learning_rate": 3e-05,
+      "loss": 2.7737,
+      "step": 2060
+    },
+    {
+      "epoch": 0.560054347826087,
+      "grad_norm": 1.459381341934204,
+      "learning_rate": 3e-05,
+      "loss": 3.2269,
+      "step": 2061
+    },
+    {
+      "epoch": 0.5603260869565218,
+      "grad_norm": 1.6240458488464355,
+      "learning_rate": 3e-05,
+      "loss": 3.2216,
+      "step": 2062
+    },
+    {
+      "epoch": 0.5605978260869565,
+      "grad_norm": 1.3474336862564087,
+      "learning_rate": 3e-05,
+      "loss": 3.191,
+      "step": 2063
+    },
+    {
+      "epoch": 0.5608695652173913,
+      "grad_norm": 1.2936657667160034,
+      "learning_rate": 3e-05,
+      "loss": 3.1124,
+      "step": 2064
+    },
+    {
+      "epoch": 0.561141304347826,
+      "grad_norm": 1.4614254236221313,
+      "learning_rate": 3e-05,
+      "loss": 3.5676,
+      "step": 2065
+    },
+    {
+      "epoch": 0.5614130434782608,
+      "grad_norm": 1.4121429920196533,
+      "learning_rate": 3e-05,
+      "loss": 2.9918,
+      "step": 2066
+    },
+    {
+      "epoch": 0.5616847826086957,
+      "grad_norm": 1.8185652494430542,
+      "learning_rate": 3e-05,
+      "loss": 3.2873,
+      "step": 2067
+    },
+    {
+      "epoch": 0.5619565217391305,
+      "grad_norm": 1.7184209823608398,
+      "learning_rate": 3e-05,
+      "loss": 3.3629,
+      "step": 2068
+    },
+    {
+      "epoch": 0.5622282608695652,
+      "grad_norm": 1.2696248292922974,
+      "learning_rate": 3e-05,
+      "loss": 2.8962,
+      "step": 2069
+    },
+    {
+      "epoch": 0.5625,
+      "grad_norm": 1.2504596710205078,
+      "learning_rate": 3e-05,
+      "loss": 3.111,
+      "step": 2070
+    },
+    {
+      "epoch": 0.5627717391304348,
+      "grad_norm": 1.2336483001708984,
+      "learning_rate": 3e-05,
+      "loss": 3.1268,
+      "step": 2071
+    },
+    {
+      "epoch": 0.5630434782608695,
+      "grad_norm": 1.420301914215088,
+      "learning_rate": 3e-05,
+      "loss": 3.0844,
+      "step": 2072
+    },
+    {
+      "epoch": 0.5633152173913043,
+      "grad_norm": 1.4836974143981934,
+      "learning_rate": 3e-05,
+      "loss": 3.305,
+      "step": 2073
+    },
+    {
+      "epoch": 0.5635869565217392,
+      "grad_norm": 1.4846380949020386,
+      "learning_rate": 3e-05,
+      "loss": 2.9409,
+      "step": 2074
+    },
+    {
+      "epoch": 0.563858695652174,
+      "grad_norm": 1.5037497282028198,
+      "learning_rate": 3e-05,
+      "loss": 3.691,
+      "step": 2075
+    },
+    {
+      "epoch": 0.5641304347826087,
+      "grad_norm": 1.1565651893615723,
+      "learning_rate": 3e-05,
+      "loss": 3.678,
+      "step": 2076
+    },
+    {
+      "epoch": 0.5644021739130435,
+      "grad_norm": 1.3678765296936035,
+      "learning_rate": 3e-05,
+      "loss": 3.7943,
+      "step": 2077
+    },
+    {
+      "epoch": 0.5646739130434782,
+      "grad_norm": 1.6229230165481567,
+      "learning_rate": 3e-05,
+      "loss": 3.6795,
+      "step": 2078
+    },
+    {
+      "epoch": 0.564945652173913,
+      "grad_norm": 1.3019323348999023,
+      "learning_rate": 3e-05,
+      "loss": 3.2213,
+      "step": 2079
+    },
+    {
+      "epoch": 0.5652173913043478,
+      "grad_norm": 1.5816229581832886,
+      "learning_rate": 3e-05,
+      "loss": 3.9553,
+      "step": 2080
+    },
+    {
+      "epoch": 0.5654891304347827,
+      "grad_norm": 1.5422357320785522,
+      "learning_rate": 3e-05,
+      "loss": 4.4725,
+      "step": 2081
+    },
+    {
+      "epoch": 0.5657608695652174,
+      "grad_norm": 1.4155722856521606,
+      "learning_rate": 3e-05,
+      "loss": 3.6369,
+      "step": 2082
+    },
+    {
+      "epoch": 0.5660326086956522,
+      "grad_norm": 1.52280592918396,
+      "learning_rate": 3e-05,
+      "loss": 3.205,
+      "step": 2083
+    },
+    {
+      "epoch": 0.566304347826087,
+      "grad_norm": 1.4019720554351807,
+      "learning_rate": 3e-05,
+      "loss": 3.5101,
+      "step": 2084
+    },
+    {
+      "epoch": 0.5665760869565217,
+      "grad_norm": 1.2434489727020264,
+      "learning_rate": 3e-05,
+      "loss": 3.5674,
+      "step": 2085
+    },
+    {
+      "epoch": 0.5668478260869565,
+      "grad_norm": 1.4792324304580688,
+      "learning_rate": 3e-05,
+      "loss": 2.8302,
+      "step": 2086
+    },
+    {
+      "epoch": 0.5671195652173913,
+      "grad_norm": 2.289886236190796,
+      "learning_rate": 3e-05,
+      "loss": 3.8648,
+      "step": 2087
+    },
+    {
+      "epoch": 0.5673913043478261,
+      "grad_norm": 1.3174787759780884,
+      "learning_rate": 3e-05,
+      "loss": 3.2811,
+      "step": 2088
+    },
+    {
+      "epoch": 0.5676630434782609,
+      "grad_norm": 1.2132010459899902,
+      "learning_rate": 3e-05,
+      "loss": 3.6173,
+      "step": 2089
+    },
+    {
+      "epoch": 0.5679347826086957,
+      "grad_norm": 1.528756856918335,
+      "learning_rate": 3e-05,
+      "loss": 3.5117,
+      "step": 2090
+    },
+    {
+      "epoch": 0.5682065217391304,
+      "grad_norm": 1.3807978630065918,
+      "learning_rate": 3e-05,
+      "loss": 3.0206,
+      "step": 2091
+    },
+    {
+      "epoch": 0.5684782608695652,
+      "grad_norm": 1.0497171878814697,
+      "learning_rate": 3e-05,
+      "loss": 3.2278,
+      "step": 2092
+    },
+    {
+      "epoch": 0.56875,
+      "grad_norm": 1.6013227701187134,
+      "learning_rate": 3e-05,
+      "loss": 3.4125,
+      "step": 2093
+    },
+    {
+      "epoch": 0.5690217391304347,
+      "grad_norm": 0.9287932515144348,
+      "learning_rate": 3e-05,
+      "loss": 3.1837,
+      "step": 2094
+    },
+    {
+      "epoch": 0.5692934782608695,
+      "grad_norm": 1.1170742511749268,
+      "learning_rate": 3e-05,
+      "loss": 3.5987,
+      "step": 2095
+    },
+    {
+      "epoch": 0.5695652173913044,
+      "grad_norm": 0.9773249626159668,
+      "learning_rate": 3e-05,
+      "loss": 3.4219,
+      "step": 2096
+    },
+    {
+      "epoch": 0.5698369565217392,
+      "grad_norm": 0.9633996486663818,
+      "learning_rate": 3e-05,
+      "loss": 3.3764,
+      "step": 2097
+    },
+    {
+      "epoch": 0.5701086956521739,
+      "grad_norm": 1.2055789232254028,
+      "learning_rate": 3e-05,
+      "loss": 3.3378,
+      "step": 2098
+    },
+    {
+      "epoch": 0.5703804347826087,
+      "grad_norm": 0.9966376423835754,
+      "learning_rate": 3e-05,
+      "loss": 2.8607,
+      "step": 2099
+    },
+    {
+      "epoch": 0.5706521739130435,
+      "grad_norm": 1.2070221900939941,
+      "learning_rate": 3e-05,
+      "loss": 2.7183,
+      "step": 2100
+    },
+    {
+      "epoch": 0.5709239130434782,
+      "grad_norm": 1.4068926572799683,
+      "learning_rate": 3e-05,
+      "loss": 3.8481,
+      "step": 2101
+    },
+    {
+      "epoch": 0.571195652173913,
+      "grad_norm": 1.5947731733322144,
+      "learning_rate": 3e-05,
+      "loss": 2.9881,
+      "step": 2102
+    },
+    {
+      "epoch": 0.5714673913043479,
+      "grad_norm": 1.2691866159439087,
+      "learning_rate": 3e-05,
+      "loss": 3.4337,
+      "step": 2103
+    },
+    {
+      "epoch": 0.5717391304347826,
+      "grad_norm": 1.850817322731018,
+      "learning_rate": 3e-05,
+      "loss": 3.4459,
+      "step": 2104
+    },
+    {
+      "epoch": 0.5720108695652174,
+      "grad_norm": 1.5024229288101196,
+      "learning_rate": 3e-05,
+      "loss": 3.1217,
+      "step": 2105
+    },
+    {
+      "epoch": 0.5722826086956522,
+      "grad_norm": 1.167541742324829,
+      "learning_rate": 3e-05,
+      "loss": 3.584,
+      "step": 2106
+    },
+    {
+      "epoch": 0.5725543478260869,
+      "grad_norm": 1.89936363697052,
+      "learning_rate": 3e-05,
+      "loss": 3.7818,
+      "step": 2107
+    },
+    {
+      "epoch": 0.5728260869565217,
+      "grad_norm": 1.5929173231124878,
+      "learning_rate": 3e-05,
+      "loss": 3.4213,
+      "step": 2108
+    },
+    {
+      "epoch": 0.5730978260869565,
+      "grad_norm": 1.122673749923706,
+      "learning_rate": 3e-05,
+      "loss": 3.6769,
+      "step": 2109
+    },
+    {
+      "epoch": 0.5733695652173914,
+      "grad_norm": 1.0892047882080078,
+      "learning_rate": 3e-05,
+      "loss": 3.2629,
+      "step": 2110
+    },
+    {
+      "epoch": 0.5736413043478261,
+      "grad_norm": 0.9661053419113159,
+      "learning_rate": 3e-05,
+      "loss": 3.0303,
+      "step": 2111
+    },
+    {
+      "epoch": 0.5739130434782609,
+      "grad_norm": 1.2852338552474976,
+      "learning_rate": 3e-05,
+      "loss": 3.6586,
+      "step": 2112
+    },
+    {
+      "epoch": 0.5741847826086957,
+      "grad_norm": 1.211471676826477,
+      "learning_rate": 3e-05,
+      "loss": 3.1448,
+      "step": 2113
+    },
+    {
+      "epoch": 0.5744565217391304,
+      "grad_norm": 1.1296757459640503,
+      "learning_rate": 3e-05,
+      "loss": 2.9875,
+      "step": 2114
+    },
+    {
+      "epoch": 0.5747282608695652,
+      "grad_norm": 1.3020541667938232,
+      "learning_rate": 3e-05,
+      "loss": 3.111,
+      "step": 2115
+    },
+    {
+      "epoch": 0.575,
+      "grad_norm": 1.0208520889282227,
+      "learning_rate": 3e-05,
+      "loss": 2.7882,
+      "step": 2116
+    },
+    {
+      "epoch": 0.5752717391304348,
+      "grad_norm": 1.4555299282073975,
+      "learning_rate": 3e-05,
+      "loss": 3.3721,
+      "step": 2117
+    },
+    {
+      "epoch": 0.5755434782608696,
+      "grad_norm": 2.195035934448242,
+      "learning_rate": 3e-05,
+      "loss": 4.1556,
+      "step": 2118
+    },
+    {
+      "epoch": 0.5758152173913044,
+      "grad_norm": 1.4949889183044434,
+      "learning_rate": 3e-05,
+      "loss": 2.6564,
+      "step": 2119
+    },
+    {
+      "epoch": 0.5760869565217391,
+      "grad_norm": 1.1324836015701294,
+      "learning_rate": 3e-05,
+      "loss": 3.1073,
+      "step": 2120
+    },
+    {
+      "epoch": 0.5763586956521739,
+      "grad_norm": 1.3076951503753662,
+      "learning_rate": 3e-05,
+      "loss": 3.6178,
+      "step": 2121
+    },
+    {
+      "epoch": 0.5766304347826087,
+      "grad_norm": 1.322385311126709,
+      "learning_rate": 3e-05,
+      "loss": 3.9284,
+      "step": 2122
+    },
+    {
+      "epoch": 0.5769021739130434,
+      "grad_norm": 1.567959189414978,
+      "learning_rate": 3e-05,
+      "loss": 4.0229,
+      "step": 2123
+    },
+    {
+      "epoch": 0.5771739130434783,
+      "grad_norm": 1.4464517831802368,
+      "learning_rate": 3e-05,
+      "loss": 3.3669,
+      "step": 2124
+    },
+    {
+      "epoch": 0.5774456521739131,
+      "grad_norm": 1.5971894264221191,
+      "learning_rate": 3e-05,
+      "loss": 4.1792,
+      "step": 2125
+    },
+    {
+      "epoch": 0.5777173913043478,
+      "grad_norm": 1.4091331958770752,
+      "learning_rate": 3e-05,
+      "loss": 3.3196,
+      "step": 2126
+    },
+    {
+      "epoch": 0.5779891304347826,
+      "grad_norm": 1.4950571060180664,
+      "learning_rate": 3e-05,
+      "loss": 3.6238,
+      "step": 2127
+    },
+    {
+      "epoch": 0.5782608695652174,
+      "grad_norm": 1.0344115495681763,
+      "learning_rate": 3e-05,
+      "loss": 2.8622,
+      "step": 2128
+    },
+    {
+      "epoch": 0.5785326086956522,
+      "grad_norm": 1.1362147331237793,
+      "learning_rate": 3e-05,
+      "loss": 2.7986,
+      "step": 2129
+    },
+    {
+      "epoch": 0.5788043478260869,
+      "grad_norm": 1.1396723985671997,
+      "learning_rate": 3e-05,
+      "loss": 2.7951,
+      "step": 2130
+    },
+    {
+      "epoch": 0.5790760869565217,
+      "grad_norm": 1.355010747909546,
+      "learning_rate": 3e-05,
+      "loss": 3.1708,
+      "step": 2131
+    },
+    {
+      "epoch": 0.5793478260869566,
+      "grad_norm": 1.5592600107192993,
+      "learning_rate": 3e-05,
+      "loss": 3.5758,
+      "step": 2132
+    },
+    {
+      "epoch": 0.5796195652173913,
+      "grad_norm": 1.56844961643219,
+      "learning_rate": 3e-05,
+      "loss": 3.4122,
+      "step": 2133
+    },
+    {
+      "epoch": 0.5798913043478261,
+      "grad_norm": 1.3906667232513428,
+      "learning_rate": 3e-05,
+      "loss": 3.7639,
+      "step": 2134
+    },
+    {
+      "epoch": 0.5801630434782609,
+      "grad_norm": 1.131856918334961,
+      "learning_rate": 3e-05,
+      "loss": 3.3024,
+      "step": 2135
+    },
+    {
+      "epoch": 0.5804347826086956,
+      "grad_norm": 1.2349239587783813,
+      "learning_rate": 3e-05,
+      "loss": 3.2338,
+      "step": 2136
+    },
+    {
+      "epoch": 0.5807065217391304,
+      "grad_norm": 1.320646047592163,
+      "learning_rate": 3e-05,
+      "loss": 3.2089,
+      "step": 2137
+    },
+    {
+      "epoch": 0.5809782608695652,
+      "grad_norm": 1.0962297916412354,
+      "learning_rate": 3e-05,
+      "loss": 2.9312,
+      "step": 2138
+    },
+    {
+      "epoch": 0.58125,
+      "grad_norm": 1.2109733819961548,
+      "learning_rate": 3e-05,
+      "loss": 2.6762,
+      "step": 2139
+    },
+    {
+      "epoch": 0.5815217391304348,
+      "grad_norm": 1.3556448221206665,
+      "learning_rate": 3e-05,
+      "loss": 3.6022,
+      "step": 2140
+    },
+    {
+      "epoch": 0.5817934782608696,
+      "grad_norm": 1.216602087020874,
+      "learning_rate": 3e-05,
+      "loss": 3.9107,
+      "step": 2141
+    },
+    {
+      "epoch": 0.5820652173913043,
+      "grad_norm": 1.2305996417999268,
+      "learning_rate": 3e-05,
+      "loss": 3.2447,
+      "step": 2142
+    },
+    {
+      "epoch": 0.5823369565217391,
+      "grad_norm": 1.2739956378936768,
+      "learning_rate": 3e-05,
+      "loss": 3.4232,
+      "step": 2143
+    },
+    {
+      "epoch": 0.5826086956521739,
+      "grad_norm": 1.2404019832611084,
+      "learning_rate": 3e-05,
+      "loss": 3.467,
+      "step": 2144
+    },
+    {
+      "epoch": 0.5828804347826086,
+      "grad_norm": 1.3793765306472778,
+      "learning_rate": 3e-05,
+      "loss": 3.4627,
+      "step": 2145
+    },
+    {
+      "epoch": 0.5831521739130435,
+      "grad_norm": 1.6223127841949463,
+      "learning_rate": 3e-05,
+      "loss": 3.2725,
+      "step": 2146
+    },
+    {
+      "epoch": 0.5834239130434783,
+      "grad_norm": 1.055362343788147,
+      "learning_rate": 3e-05,
+      "loss": 3.2517,
+      "step": 2147
+    },
+    {
+      "epoch": 0.5836956521739131,
+      "grad_norm": 1.2135682106018066,
+      "learning_rate": 3e-05,
+      "loss": 3.0423,
+      "step": 2148
+    },
+    {
+      "epoch": 0.5839673913043478,
+      "grad_norm": 1.7276960611343384,
+      "learning_rate": 3e-05,
+      "loss": 4.2823,
+      "step": 2149
+    },
+    {
+      "epoch": 0.5842391304347826,
+      "grad_norm": 1.1459511518478394,
+      "learning_rate": 3e-05,
+      "loss": 3.129,
+      "step": 2150
+    },
+    {
+      "epoch": 0.5845108695652174,
+      "grad_norm": 1.7933846712112427,
+      "learning_rate": 3e-05,
+      "loss": 3.6957,
+      "step": 2151
+    },
+    {
+      "epoch": 0.5847826086956521,
+      "grad_norm": 1.3888658285140991,
+      "learning_rate": 3e-05,
+      "loss": 4.2824,
+      "step": 2152
+    },
+    {
+      "epoch": 0.585054347826087,
+      "grad_norm": 1.18975031375885,
+      "learning_rate": 3e-05,
+      "loss": 2.9272,
+      "step": 2153
+    },
+    {
+      "epoch": 0.5853260869565218,
+      "grad_norm": 1.0910744667053223,
+      "learning_rate": 3e-05,
+      "loss": 3.4617,
+      "step": 2154
+    },
+    {
+      "epoch": 0.5855978260869565,
+      "grad_norm": 1.3888293504714966,
+      "learning_rate": 3e-05,
+      "loss": 2.9697,
+      "step": 2155
+    },
+    {
+      "epoch": 0.5858695652173913,
+      "grad_norm": 1.1816877126693726,
+      "learning_rate": 3e-05,
+      "loss": 3.5301,
+      "step": 2156
+    },
+    {
+      "epoch": 0.5861413043478261,
+      "grad_norm": 1.2659231424331665,
+      "learning_rate": 3e-05,
+      "loss": 3.5646,
+      "step": 2157
+    },
+    {
+      "epoch": 0.5864130434782608,
+      "grad_norm": 1.5958353281021118,
+      "learning_rate": 3e-05,
+      "loss": 3.378,
+      "step": 2158
+    },
+    {
+      "epoch": 0.5866847826086956,
+      "grad_norm": 1.2585597038269043,
+      "learning_rate": 3e-05,
+      "loss": 3.0523,
+      "step": 2159
+    },
+    {
+      "epoch": 0.5869565217391305,
+      "grad_norm": 1.0975054502487183,
+      "learning_rate": 3e-05,
+      "loss": 3.0933,
+      "step": 2160
+    },
+    {
+      "epoch": 0.5872282608695653,
+      "grad_norm": 1.5233417749404907,
+      "learning_rate": 3e-05,
+      "loss": 3.2665,
+      "step": 2161
+    },
+    {
+      "epoch": 0.5875,
+      "grad_norm": 2.02248215675354,
+      "learning_rate": 3e-05,
+      "loss": 3.754,
+      "step": 2162
+    },
+    {
+      "epoch": 0.5877717391304348,
+      "grad_norm": 0.9959807395935059,
+      "learning_rate": 3e-05,
+      "loss": 2.9276,
+      "step": 2163
+    },
+    {
+      "epoch": 0.5880434782608696,
+      "grad_norm": 1.1323211193084717,
+      "learning_rate": 3e-05,
+      "loss": 3.0789,
+      "step": 2164
+    },
+    {
+      "epoch": 0.5883152173913043,
+      "grad_norm": 1.0242100954055786,
+      "learning_rate": 3e-05,
+      "loss": 3.3039,
+      "step": 2165
+    },
+    {
+      "epoch": 0.5885869565217391,
+      "grad_norm": 1.0082273483276367,
+      "learning_rate": 3e-05,
+      "loss": 3.3631,
+      "step": 2166
+    },
+    {
+      "epoch": 0.5888586956521739,
+      "grad_norm": 1.1713404655456543,
+      "learning_rate": 3e-05,
+      "loss": 2.7448,
+      "step": 2167
+    },
+    {
+      "epoch": 0.5891304347826087,
+      "grad_norm": 1.1318339109420776,
+      "learning_rate": 3e-05,
+      "loss": 3.3552,
+      "step": 2168
+    },
+    {
+      "epoch": 0.5894021739130435,
+      "grad_norm": 1.3056354522705078,
+      "learning_rate": 3e-05,
+      "loss": 3.8349,
+      "step": 2169
+    },
+    {
+      "epoch": 0.5896739130434783,
+      "grad_norm": 1.1058902740478516,
+      "learning_rate": 3e-05,
+      "loss": 2.9931,
+      "step": 2170
+    },
+    {
+      "epoch": 0.589945652173913,
+      "grad_norm": 1.3379881381988525,
+      "learning_rate": 3e-05,
+      "loss": 3.0709,
+      "step": 2171
+    },
+    {
+      "epoch": 0.5902173913043478,
+      "grad_norm": 1.3458565473556519,
+      "learning_rate": 3e-05,
+      "loss": 3.5521,
+      "step": 2172
+    },
+    {
+      "epoch": 0.5904891304347826,
+      "grad_norm": 1.2749676704406738,
+      "learning_rate": 3e-05,
+      "loss": 3.3535,
+      "step": 2173
+    },
+    {
+      "epoch": 0.5907608695652173,
+      "grad_norm": 1.3268214464187622,
+      "learning_rate": 3e-05,
+      "loss": 3.6918,
+      "step": 2174
+    },
+    {
+      "epoch": 0.5910326086956522,
+      "grad_norm": 1.4857879877090454,
+      "learning_rate": 3e-05,
+      "loss": 3.6046,
+      "step": 2175
+    },
+    {
+      "epoch": 0.591304347826087,
+      "grad_norm": 1.009945034980774,
+      "learning_rate": 3e-05,
+      "loss": 3.435,
+      "step": 2176
+    },
+    {
+      "epoch": 0.5915760869565218,
+      "grad_norm": 1.1983013153076172,
+      "learning_rate": 3e-05,
+      "loss": 3.5871,
+      "step": 2177
+    },
+    {
+      "epoch": 0.5918478260869565,
+      "grad_norm": 0.9718300700187683,
+      "learning_rate": 3e-05,
+      "loss": 3.3072,
+      "step": 2178
+    },
+    {
+      "epoch": 0.5921195652173913,
+      "grad_norm": 1.101165771484375,
+      "learning_rate": 3e-05,
+      "loss": 3.2276,
+      "step": 2179
+    },
+    {
+      "epoch": 0.592391304347826,
+      "grad_norm": 1.012749433517456,
+      "learning_rate": 3e-05,
+      "loss": 3.1112,
+      "step": 2180
+    },
+    {
+      "epoch": 0.5926630434782608,
+      "grad_norm": 0.9688409566879272,
+      "learning_rate": 3e-05,
+      "loss": 3.0918,
+      "step": 2181
+    },
+    {
+      "epoch": 0.5929347826086957,
+      "grad_norm": 0.9854469895362854,
+      "learning_rate": 3e-05,
+      "loss": 3.1777,
+      "step": 2182
+    },
+    {
+      "epoch": 0.5932065217391305,
+      "grad_norm": 0.9278433918952942,
+      "learning_rate": 3e-05,
+      "loss": 3.1989,
+      "step": 2183
+    },
+    {
+      "epoch": 0.5934782608695652,
+      "grad_norm": 0.983065664768219,
+      "learning_rate": 3e-05,
+      "loss": 3.3761,
+      "step": 2184
+    },
+    {
+      "epoch": 0.59375,
+      "grad_norm": 1.473644733428955,
+      "learning_rate": 3e-05,
+      "loss": 3.6095,
+      "step": 2185
+    },
+    {
+      "epoch": 0.5940217391304348,
+      "grad_norm": 1.2839797735214233,
+      "learning_rate": 3e-05,
+      "loss": 2.9857,
+      "step": 2186
+    },
+    {
+      "epoch": 0.5942934782608695,
+      "grad_norm": 1.7509822845458984,
+      "learning_rate": 3e-05,
+      "loss": 3.7711,
+      "step": 2187
+    },
+    {
+      "epoch": 0.5945652173913043,
+      "grad_norm": 1.171937346458435,
+      "learning_rate": 3e-05,
+      "loss": 3.6335,
+      "step": 2188
+    },
+    {
+      "epoch": 0.5948369565217392,
+      "grad_norm": 1.4512766599655151,
+      "learning_rate": 3e-05,
+      "loss": 2.9629,
+      "step": 2189
+    },
+    {
+      "epoch": 0.595108695652174,
+      "grad_norm": 1.0719918012619019,
+      "learning_rate": 3e-05,
+      "loss": 2.8943,
+      "step": 2190
+    },
+    {
+      "epoch": 0.5953804347826087,
+      "grad_norm": 1.11574387550354,
+      "learning_rate": 3e-05,
+      "loss": 3.7779,
+      "step": 2191
+    },
+    {
+      "epoch": 0.5956521739130435,
+      "grad_norm": 1.6633427143096924,
+      "learning_rate": 3e-05,
+      "loss": 3.7736,
+      "step": 2192
+    },
+    {
+      "epoch": 0.5959239130434782,
+      "grad_norm": 1.1903231143951416,
+      "learning_rate": 3e-05,
+      "loss": 2.9944,
+      "step": 2193
+    },
+    {
+      "epoch": 0.596195652173913,
+      "grad_norm": 1.2331459522247314,
+      "learning_rate": 3e-05,
+      "loss": 3.639,
+      "step": 2194
+    },
+    {
+      "epoch": 0.5964673913043478,
+      "grad_norm": 1.8560791015625,
+      "learning_rate": 3e-05,
+      "loss": 4.0684,
+      "step": 2195
+    },
+    {
+      "epoch": 0.5967391304347827,
+      "grad_norm": 1.2733112573623657,
+      "learning_rate": 3e-05,
+      "loss": 3.1905,
+      "step": 2196
+    },
+    {
+      "epoch": 0.5970108695652174,
+      "grad_norm": 1.156823992729187,
+      "learning_rate": 3e-05,
+      "loss": 3.3605,
+      "step": 2197
+    },
+    {
+      "epoch": 0.5972826086956522,
+      "grad_norm": 1.0290706157684326,
+      "learning_rate": 3e-05,
+      "loss": 3.2214,
+      "step": 2198
+    },
+    {
+      "epoch": 0.597554347826087,
+      "grad_norm": 1.1795909404754639,
+      "learning_rate": 3e-05,
+      "loss": 2.9137,
+      "step": 2199
+    },
+    {
+      "epoch": 0.5978260869565217,
+      "grad_norm": 1.3558768033981323,
+      "learning_rate": 3e-05,
+      "loss": 3.6129,
+      "step": 2200
+    },
+    {
+      "epoch": 0.5980978260869565,
+      "grad_norm": 0.9279618859291077,
+      "learning_rate": 3e-05,
+      "loss": 2.5655,
+      "step": 2201
+    },
+    {
+      "epoch": 0.5983695652173913,
+      "grad_norm": 1.2650532722473145,
+      "learning_rate": 3e-05,
+      "loss": 2.8838,
+      "step": 2202
+    },
+    {
+      "epoch": 0.5986413043478261,
+      "grad_norm": 1.2165402173995972,
+      "learning_rate": 3e-05,
+      "loss": 3.5675,
+      "step": 2203
+    },
+    {
+      "epoch": 0.5989130434782609,
+      "grad_norm": 1.1845487356185913,
+      "learning_rate": 3e-05,
+      "loss": 2.9909,
+      "step": 2204
+    },
+    {
+      "epoch": 0.5991847826086957,
+      "grad_norm": 1.1865789890289307,
+      "learning_rate": 3e-05,
+      "loss": 3.6638,
+      "step": 2205
+    },
+    {
+      "epoch": 0.5994565217391304,
+      "grad_norm": 0.9880366921424866,
+      "learning_rate": 3e-05,
+      "loss": 3.023,
+      "step": 2206
+    },
+    {
+      "epoch": 0.5997282608695652,
+      "grad_norm": 1.1242929697036743,
+      "learning_rate": 3e-05,
+      "loss": 3.1062,
+      "step": 2207
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1712896823883057,
+      "learning_rate": 3e-05,
+      "loss": 3.029,
+      "step": 2208
+    },
+    {
+      "epoch": 0.6002717391304347,
+      "grad_norm": 1.5733115673065186,
+      "learning_rate": 3e-05,
+      "loss": 3.7851,
+      "step": 2209
+    },
+    {
+      "epoch": 0.6005434782608695,
+      "grad_norm": 1.2813576459884644,
+      "learning_rate": 3e-05,
+      "loss": 3.4965,
+      "step": 2210
+    },
+    {
+      "epoch": 0.6008152173913044,
+      "grad_norm": 1.3263084888458252,
+      "learning_rate": 3e-05,
+      "loss": 4.1356,
+      "step": 2211
+    },
+    {
+      "epoch": 0.6010869565217392,
+      "grad_norm": 1.1176859140396118,
+      "learning_rate": 3e-05,
+      "loss": 3.4671,
+      "step": 2212
+    },
+    {
+      "epoch": 0.6013586956521739,
+      "grad_norm": 1.2956467866897583,
+      "learning_rate": 3e-05,
+      "loss": 3.8482,
+      "step": 2213
+    },
+    {
+      "epoch": 0.6016304347826087,
+      "grad_norm": 1.8947595357894897,
+      "learning_rate": 3e-05,
+      "loss": 3.2963,
+      "step": 2214
+    },
+    {
+      "epoch": 0.6019021739130435,
+      "grad_norm": 1.3561947345733643,
+      "learning_rate": 3e-05,
+      "loss": 3.1028,
+      "step": 2215
+    },
+    {
+      "epoch": 0.6021739130434782,
+      "grad_norm": 1.1945207118988037,
+      "learning_rate": 3e-05,
+      "loss": 3.2949,
+      "step": 2216
+    },
+    {
+      "epoch": 0.602445652173913,
+      "grad_norm": 1.3198587894439697,
+      "learning_rate": 3e-05,
+      "loss": 3.5389,
+      "step": 2217
+    },
+    {
+      "epoch": 0.6027173913043479,
+      "grad_norm": 1.6522905826568604,
+      "learning_rate": 3e-05,
+      "loss": 3.2054,
+      "step": 2218
+    },
+    {
+      "epoch": 0.6029891304347826,
+      "grad_norm": 1.2735859155654907,
+      "learning_rate": 3e-05,
+      "loss": 3.4955,
+      "step": 2219
+    },
+    {
+      "epoch": 0.6032608695652174,
+      "grad_norm": 1.2329673767089844,
+      "learning_rate": 3e-05,
+      "loss": 3.7713,
+      "step": 2220
+    },
+    {
+      "epoch": 0.6035326086956522,
+      "grad_norm": 1.1366288661956787,
+      "learning_rate": 3e-05,
+      "loss": 3.333,
+      "step": 2221
+    },
+    {
+      "epoch": 0.6038043478260869,
+      "grad_norm": 1.0295037031173706,
+      "learning_rate": 3e-05,
+      "loss": 3.2991,
+      "step": 2222
+    },
+    {
+      "epoch": 0.6040760869565217,
+      "grad_norm": 1.0760974884033203,
+      "learning_rate": 3e-05,
+      "loss": 3.6266,
+      "step": 2223
+    },
+    {
+      "epoch": 0.6043478260869565,
+      "grad_norm": 1.0044188499450684,
+      "learning_rate": 3e-05,
+      "loss": 3.1854,
+      "step": 2224
+    },
+    {
+      "epoch": 0.6046195652173914,
+      "grad_norm": 0.79954993724823,
+      "learning_rate": 3e-05,
+      "loss": 2.5833,
+      "step": 2225
+    },
+    {
+      "epoch": 0.6048913043478261,
+      "grad_norm": 1.3376729488372803,
+      "learning_rate": 3e-05,
+      "loss": 3.8299,
+      "step": 2226
+    },
+    {
+      "epoch": 0.6051630434782609,
+      "grad_norm": 1.3323661088943481,
+      "learning_rate": 3e-05,
+      "loss": 3.5493,
+      "step": 2227
+    },
+    {
+      "epoch": 0.6054347826086957,
+      "grad_norm": 0.9946414828300476,
+      "learning_rate": 3e-05,
+      "loss": 3.0806,
+      "step": 2228
+    },
+    {
+      "epoch": 0.6057065217391304,
+      "grad_norm": 1.486099123954773,
+      "learning_rate": 3e-05,
+      "loss": 3.5046,
+      "step": 2229
+    },
+    {
+      "epoch": 0.6059782608695652,
+      "grad_norm": 1.4763749837875366,
+      "learning_rate": 3e-05,
+      "loss": 3.2249,
+      "step": 2230
+    },
+    {
+      "epoch": 0.60625,
+      "grad_norm": 1.1273298263549805,
+      "learning_rate": 3e-05,
+      "loss": 3.097,
+      "step": 2231
+    },
+    {
+      "epoch": 0.6065217391304348,
+      "grad_norm": 1.273410677909851,
+      "learning_rate": 3e-05,
+      "loss": 3.3963,
+      "step": 2232
+    },
+    {
+      "epoch": 0.6067934782608696,
+      "grad_norm": 1.6567277908325195,
+      "learning_rate": 3e-05,
+      "loss": 3.4074,
+      "step": 2233
+    },
+    {
+      "epoch": 0.6070652173913044,
+      "grad_norm": 1.272204875946045,
+      "learning_rate": 3e-05,
+      "loss": 3.5062,
+      "step": 2234
+    },
+    {
+      "epoch": 0.6073369565217391,
+      "grad_norm": 0.9623717665672302,
+      "learning_rate": 3e-05,
+      "loss": 3.348,
+      "step": 2235
+    },
+    {
+      "epoch": 0.6076086956521739,
+      "grad_norm": 1.7306931018829346,
+      "learning_rate": 3e-05,
+      "loss": 3.5976,
+      "step": 2236
+    },
+    {
+      "epoch": 0.6078804347826087,
+      "grad_norm": 1.2716137170791626,
+      "learning_rate": 3e-05,
+      "loss": 3.8244,
+      "step": 2237
+    },
+    {
+      "epoch": 0.6081521739130434,
+      "grad_norm": 1.3230034112930298,
+      "learning_rate": 3e-05,
+      "loss": 3.4558,
+      "step": 2238
+    },
+    {
+      "epoch": 0.6084239130434783,
+      "grad_norm": 1.733973503112793,
+      "learning_rate": 3e-05,
+      "loss": 3.0608,
+      "step": 2239
+    },
+    {
+      "epoch": 0.6086956521739131,
+      "grad_norm": 1.2597112655639648,
+      "learning_rate": 3e-05,
+      "loss": 2.8695,
+      "step": 2240
+    },
+    {
+      "epoch": 0.6089673913043478,
+      "grad_norm": 1.033078670501709,
+      "learning_rate": 3e-05,
+      "loss": 2.528,
+      "step": 2241
+    },
+    {
+      "epoch": 0.6092391304347826,
+      "grad_norm": 1.4419554471969604,
+      "learning_rate": 3e-05,
+      "loss": 3.5569,
+      "step": 2242
+    },
+    {
+      "epoch": 0.6095108695652174,
+      "grad_norm": 1.376265287399292,
+      "learning_rate": 3e-05,
+      "loss": 3.4915,
+      "step": 2243
+    },
+    {
+      "epoch": 0.6097826086956522,
+      "grad_norm": 1.1557855606079102,
+      "learning_rate": 3e-05,
+      "loss": 3.4314,
+      "step": 2244
+    },
+    {
+      "epoch": 0.6100543478260869,
+      "grad_norm": 1.3031344413757324,
+      "learning_rate": 3e-05,
+      "loss": 3.5682,
+      "step": 2245
+    },
+    {
+      "epoch": 0.6103260869565217,
+      "grad_norm": 1.2761789560317993,
+      "learning_rate": 3e-05,
+      "loss": 3.4629,
+      "step": 2246
+    },
+    {
+      "epoch": 0.6105978260869566,
+      "grad_norm": 1.1963956356048584,
+      "learning_rate": 3e-05,
+      "loss": 3.2122,
+      "step": 2247
+    },
+    {
+      "epoch": 0.6108695652173913,
+      "grad_norm": 1.357904314994812,
+      "learning_rate": 3e-05,
+      "loss": 3.0868,
+      "step": 2248
+    },
+    {
+      "epoch": 0.6111413043478261,
+      "grad_norm": 1.3014525175094604,
+      "learning_rate": 3e-05,
+      "loss": 2.9693,
+      "step": 2249
+    },
+    {
+      "epoch": 0.6114130434782609,
+      "grad_norm": 1.3314582109451294,
+      "learning_rate": 3e-05,
+      "loss": 3.3714,
+      "step": 2250
+    },
+    {
+      "epoch": 0.6116847826086956,
+      "grad_norm": 1.3849120140075684,
+      "learning_rate": 3e-05,
+      "loss": 2.7755,
+      "step": 2251
+    },
+    {
+      "epoch": 0.6119565217391304,
+      "grad_norm": 0.9548225402832031,
+      "learning_rate": 3e-05,
+      "loss": 3.0885,
+      "step": 2252
+    },
+    {
+      "epoch": 0.6122282608695652,
+      "grad_norm": 1.4415459632873535,
+      "learning_rate": 3e-05,
+      "loss": 3.8783,
+      "step": 2253
+    },
+    {
+      "epoch": 0.6125,
+      "grad_norm": 1.7835477590560913,
+      "learning_rate": 3e-05,
+      "loss": 3.6217,
+      "step": 2254
+    },
+    {
+      "epoch": 0.6127717391304348,
+      "grad_norm": 1.4011732339859009,
+      "learning_rate": 3e-05,
+      "loss": 3.9328,
+      "step": 2255
+    },
+    {
+      "epoch": 0.6130434782608696,
+      "grad_norm": 1.0685231685638428,
+      "learning_rate": 3e-05,
+      "loss": 3.2949,
+      "step": 2256
+    },
+    {
+      "epoch": 0.6133152173913043,
+      "grad_norm": 1.2185416221618652,
+      "learning_rate": 3e-05,
+      "loss": 3.5132,
+      "step": 2257
+    },
+    {
+      "epoch": 0.6135869565217391,
+      "grad_norm": 1.0910768508911133,
+      "learning_rate": 3e-05,
+      "loss": 3.0972,
+      "step": 2258
+    },
+    {
+      "epoch": 0.6138586956521739,
+      "grad_norm": 1.0861088037490845,
+      "learning_rate": 3e-05,
+      "loss": 3.3816,
+      "step": 2259
+    },
+    {
+      "epoch": 0.6141304347826086,
+      "grad_norm": 1.0227428674697876,
+      "learning_rate": 3e-05,
+      "loss": 3.1733,
+      "step": 2260
+    },
+    {
+      "epoch": 0.6144021739130435,
+      "grad_norm": 1.3343944549560547,
+      "learning_rate": 3e-05,
+      "loss": 3.2618,
+      "step": 2261
+    },
+    {
+      "epoch": 0.6146739130434783,
+      "grad_norm": 1.3029199838638306,
+      "learning_rate": 3e-05,
+      "loss": 3.4307,
+      "step": 2262
+    },
+    {
+      "epoch": 0.6149456521739131,
+      "grad_norm": 1.152552843093872,
+      "learning_rate": 3e-05,
+      "loss": 3.4878,
+      "step": 2263
+    },
+    {
+      "epoch": 0.6152173913043478,
+      "grad_norm": 1.5189865827560425,
+      "learning_rate": 3e-05,
+      "loss": 3.4252,
+      "step": 2264
+    },
+    {
+      "epoch": 0.6154891304347826,
+      "grad_norm": 1.1164250373840332,
+      "learning_rate": 3e-05,
+      "loss": 3.5027,
+      "step": 2265
+    },
+    {
+      "epoch": 0.6157608695652174,
+      "grad_norm": 0.9906913638114929,
+      "learning_rate": 3e-05,
+      "loss": 2.9876,
+      "step": 2266
+    },
+    {
+      "epoch": 0.6160326086956521,
+      "grad_norm": 1.1681506633758545,
+      "learning_rate": 3e-05,
+      "loss": 3.2722,
+      "step": 2267
+    },
+    {
+      "epoch": 0.616304347826087,
+      "grad_norm": 1.183396816253662,
+      "learning_rate": 3e-05,
+      "loss": 3.2224,
+      "step": 2268
+    },
+    {
+      "epoch": 0.6165760869565218,
+      "grad_norm": 1.0938737392425537,
+      "learning_rate": 3e-05,
+      "loss": 3.2871,
+      "step": 2269
+    },
+    {
+      "epoch": 0.6168478260869565,
+      "grad_norm": 1.5161495208740234,
+      "learning_rate": 3e-05,
+      "loss": 3.7379,
+      "step": 2270
+    },
+    {
+      "epoch": 0.6171195652173913,
+      "grad_norm": 1.2433841228485107,
+      "learning_rate": 3e-05,
+      "loss": 2.9899,
+      "step": 2271
+    },
+    {
+      "epoch": 0.6173913043478261,
+      "grad_norm": 1.092914342880249,
+      "learning_rate": 3e-05,
+      "loss": 2.7804,
+      "step": 2272
+    },
+    {
+      "epoch": 0.6176630434782608,
+      "grad_norm": 1.2224652767181396,
+      "learning_rate": 3e-05,
+      "loss": 3.628,
+      "step": 2273
+    },
+    {
+      "epoch": 0.6179347826086956,
+      "grad_norm": 1.3725236654281616,
+      "learning_rate": 3e-05,
+      "loss": 3.6826,
+      "step": 2274
+    },
+    {
+      "epoch": 0.6182065217391305,
+      "grad_norm": 1.0092295408248901,
+      "learning_rate": 3e-05,
+      "loss": 2.9988,
+      "step": 2275
+    },
+    {
+      "epoch": 0.6184782608695653,
+      "grad_norm": 1.259446144104004,
+      "learning_rate": 3e-05,
+      "loss": 3.4234,
+      "step": 2276
+    },
+    {
+      "epoch": 0.61875,
+      "grad_norm": 1.220897912979126,
+      "learning_rate": 3e-05,
+      "loss": 3.2565,
+      "step": 2277
+    },
+    {
+      "epoch": 0.6190217391304348,
+      "grad_norm": 0.9857214093208313,
+      "learning_rate": 3e-05,
+      "loss": 3.2311,
+      "step": 2278
+    },
+    {
+      "epoch": 0.6192934782608696,
+      "grad_norm": 1.256536841392517,
+      "learning_rate": 3e-05,
+      "loss": 3.4595,
+      "step": 2279
+    },
+    {
+      "epoch": 0.6195652173913043,
+      "grad_norm": 1.0857690572738647,
+      "learning_rate": 3e-05,
+      "loss": 2.8696,
+      "step": 2280
+    },
+    {
+      "epoch": 0.6198369565217391,
+      "grad_norm": 0.9030542969703674,
+      "learning_rate": 3e-05,
+      "loss": 2.931,
+      "step": 2281
+    },
+    {
+      "epoch": 0.6201086956521739,
+      "grad_norm": 1.245595097541809,
+      "learning_rate": 3e-05,
+      "loss": 3.0565,
+      "step": 2282
+    },
+    {
+      "epoch": 0.6203804347826087,
+      "grad_norm": 1.130018949508667,
+      "learning_rate": 3e-05,
+      "loss": 3.6569,
+      "step": 2283
+    },
+    {
+      "epoch": 0.6206521739130435,
+      "grad_norm": 1.1791630983352661,
+      "learning_rate": 3e-05,
+      "loss": 2.9299,
+      "step": 2284
+    },
+    {
+      "epoch": 0.6209239130434783,
+      "grad_norm": 1.6365814208984375,
+      "learning_rate": 3e-05,
+      "loss": 3.7201,
+      "step": 2285
+    },
+    {
+      "epoch": 0.621195652173913,
+      "grad_norm": 1.023141622543335,
+      "learning_rate": 3e-05,
+      "loss": 3.4364,
+      "step": 2286
+    },
+    {
+      "epoch": 0.6214673913043478,
+      "grad_norm": 1.3121358156204224,
+      "learning_rate": 3e-05,
+      "loss": 3.4811,
+      "step": 2287
+    },
+    {
+      "epoch": 0.6217391304347826,
+      "grad_norm": 1.009194254875183,
+      "learning_rate": 3e-05,
+      "loss": 3.0576,
+      "step": 2288
+    },
+    {
+      "epoch": 0.6220108695652173,
+      "grad_norm": 1.0715727806091309,
+      "learning_rate": 3e-05,
+      "loss": 3.193,
+      "step": 2289
+    },
+    {
+      "epoch": 0.6222826086956522,
+      "grad_norm": 1.0925318002700806,
+      "learning_rate": 3e-05,
+      "loss": 3.1963,
+      "step": 2290
+    },
+    {
+      "epoch": 0.622554347826087,
+      "grad_norm": 1.6540546417236328,
+      "learning_rate": 3e-05,
+      "loss": 3.7859,
+      "step": 2291
+    },
+    {
+      "epoch": 0.6228260869565218,
+      "grad_norm": 1.0642199516296387,
+      "learning_rate": 3e-05,
+      "loss": 3.8112,
+      "step": 2292
+    },
+    {
+      "epoch": 0.6230978260869565,
+      "grad_norm": 1.2560967206954956,
+      "learning_rate": 3e-05,
+      "loss": 3.528,
+      "step": 2293
+    },
+    {
+      "epoch": 0.6233695652173913,
+      "grad_norm": 1.2816760540008545,
+      "learning_rate": 3e-05,
+      "loss": 3.1706,
+      "step": 2294
+    },
+    {
+      "epoch": 0.623641304347826,
+      "grad_norm": 0.9735274314880371,
+      "learning_rate": 3e-05,
+      "loss": 3.2275,
+      "step": 2295
+    },
+    {
+      "epoch": 0.6239130434782608,
+      "grad_norm": 1.1143537759780884,
+      "learning_rate": 3e-05,
+      "loss": 3.2471,
+      "step": 2296
+    },
+    {
+      "epoch": 0.6241847826086957,
+      "grad_norm": 1.2723908424377441,
+      "learning_rate": 3e-05,
+      "loss": 2.8132,
+      "step": 2297
+    },
+    {
+      "epoch": 0.6244565217391305,
+      "grad_norm": 1.269571304321289,
+      "learning_rate": 3e-05,
+      "loss": 3.2307,
+      "step": 2298
+    },
+    {
+      "epoch": 0.6247282608695652,
+      "grad_norm": 1.0772123336791992,
+      "learning_rate": 3e-05,
+      "loss": 3.2448,
+      "step": 2299
+    },
+    {
+      "epoch": 0.625,
+      "grad_norm": 1.5265069007873535,
+      "learning_rate": 3e-05,
+      "loss": 3.2314,
+      "step": 2300
+    },
+    {
+      "epoch": 0.6252717391304348,
+      "grad_norm": 1.133500576019287,
+      "learning_rate": 3e-05,
+      "loss": 3.1259,
+      "step": 2301
+    },
+    {
+      "epoch": 0.6255434782608695,
+      "grad_norm": 1.0310444831848145,
+      "learning_rate": 3e-05,
+      "loss": 3.1374,
+      "step": 2302
+    },
+    {
+      "epoch": 0.6258152173913043,
+      "grad_norm": 1.789737582206726,
+      "learning_rate": 3e-05,
+      "loss": 3.657,
+      "step": 2303
+    },
+    {
+      "epoch": 0.6260869565217392,
+      "grad_norm": 1.028426170349121,
+      "learning_rate": 3e-05,
+      "loss": 3.0235,
+      "step": 2304
+    },
+    {
+      "epoch": 0.626358695652174,
+      "grad_norm": 1.0289982557296753,
+      "learning_rate": 3e-05,
+      "loss": 3.0216,
+      "step": 2305
+    },
+    {
+      "epoch": 0.6266304347826087,
+      "grad_norm": 1.2136385440826416,
+      "learning_rate": 3e-05,
+      "loss": 3.2615,
+      "step": 2306
+    },
+    {
+      "epoch": 0.6269021739130435,
+      "grad_norm": 1.1237839460372925,
+      "learning_rate": 3e-05,
+      "loss": 3.1519,
+      "step": 2307
+    },
+    {
+      "epoch": 0.6271739130434782,
+      "grad_norm": 1.0916386842727661,
+      "learning_rate": 3e-05,
+      "loss": 3.404,
+      "step": 2308
+    },
+    {
+      "epoch": 0.627445652173913,
+      "grad_norm": 1.1526830196380615,
+      "learning_rate": 3e-05,
+      "loss": 3.671,
+      "step": 2309
+    },
+    {
+      "epoch": 0.6277173913043478,
+      "grad_norm": 1.0757122039794922,
+      "learning_rate": 3e-05,
+      "loss": 3.069,
+      "step": 2310
+    },
+    {
+      "epoch": 0.6279891304347827,
+      "grad_norm": 1.0339561700820923,
+      "learning_rate": 3e-05,
+      "loss": 2.9702,
+      "step": 2311
+    },
+    {
+      "epoch": 0.6282608695652174,
+      "grad_norm": 1.2000614404678345,
+      "learning_rate": 3e-05,
+      "loss": 2.9536,
+      "step": 2312
+    },
+    {
+      "epoch": 0.6285326086956522,
+      "grad_norm": 1.3104127645492554,
+      "learning_rate": 3e-05,
+      "loss": 3.1653,
+      "step": 2313
+    },
+    {
+      "epoch": 0.628804347826087,
+      "grad_norm": 1.2230424880981445,
+      "learning_rate": 3e-05,
+      "loss": 3.464,
+      "step": 2314
+    },
+    {
+      "epoch": 0.6290760869565217,
+      "grad_norm": 1.258482813835144,
+      "learning_rate": 3e-05,
+      "loss": 3.4273,
+      "step": 2315
+    },
+    {
+      "epoch": 0.6293478260869565,
+      "grad_norm": 1.1824918985366821,
+      "learning_rate": 3e-05,
+      "loss": 3.5311,
+      "step": 2316
+    },
+    {
+      "epoch": 0.6296195652173913,
+      "grad_norm": 0.9511439800262451,
+      "learning_rate": 3e-05,
+      "loss": 2.979,
+      "step": 2317
+    },
+    {
+      "epoch": 0.6298913043478261,
+      "grad_norm": 1.0357087850570679,
+      "learning_rate": 3e-05,
+      "loss": 2.8735,
+      "step": 2318
+    },
+    {
+      "epoch": 0.6301630434782609,
+      "grad_norm": 1.392822504043579,
+      "learning_rate": 3e-05,
+      "loss": 3.3224,
+      "step": 2319
+    },
+    {
+      "epoch": 0.6304347826086957,
+      "grad_norm": 1.4284133911132812,
+      "learning_rate": 3e-05,
+      "loss": 3.8771,
+      "step": 2320
+    },
+    {
+      "epoch": 0.6307065217391304,
+      "grad_norm": 1.2343987226486206,
+      "learning_rate": 3e-05,
+      "loss": 3.4731,
+      "step": 2321
+    },
+    {
+      "epoch": 0.6309782608695652,
+      "grad_norm": 1.3236418962478638,
+      "learning_rate": 3e-05,
+      "loss": 3.0388,
+      "step": 2322
+    },
+    {
+      "epoch": 0.63125,
+      "grad_norm": 1.332114577293396,
+      "learning_rate": 3e-05,
+      "loss": 3.1262,
+      "step": 2323
+    },
+    {
+      "epoch": 0.6315217391304347,
+      "grad_norm": 1.0655372142791748,
+      "learning_rate": 3e-05,
+      "loss": 3.2109,
+      "step": 2324
+    },
+    {
+      "epoch": 0.6317934782608695,
+      "grad_norm": 1.0099713802337646,
+      "learning_rate": 3e-05,
+      "loss": 3.2809,
+      "step": 2325
+    },
+    {
+      "epoch": 0.6320652173913044,
+      "grad_norm": 1.0744123458862305,
+      "learning_rate": 3e-05,
+      "loss": 2.862,
+      "step": 2326
+    },
+    {
+      "epoch": 0.6323369565217392,
+      "grad_norm": 1.051435112953186,
+      "learning_rate": 3e-05,
+      "loss": 2.8705,
+      "step": 2327
+    },
+    {
+      "epoch": 0.6326086956521739,
+      "grad_norm": 1.400758147239685,
+      "learning_rate": 3e-05,
+      "loss": 3.7773,
+      "step": 2328
+    },
+    {
+      "epoch": 0.6328804347826087,
+      "grad_norm": 1.1519436836242676,
+      "learning_rate": 3e-05,
+      "loss": 2.9399,
+      "step": 2329
+    },
+    {
+      "epoch": 0.6331521739130435,
+      "grad_norm": 1.264049768447876,
+      "learning_rate": 3e-05,
+      "loss": 3.7764,
+      "step": 2330
+    },
+    {
+      "epoch": 0.6334239130434782,
+      "grad_norm": 1.0223599672317505,
+      "learning_rate": 3e-05,
+      "loss": 2.9021,
+      "step": 2331
+    },
+    {
+      "epoch": 0.633695652173913,
+      "grad_norm": 1.255913257598877,
+      "learning_rate": 3e-05,
+      "loss": 3.6456,
+      "step": 2332
+    },
+    {
+      "epoch": 0.6339673913043479,
+      "grad_norm": 1.7114654779434204,
+      "learning_rate": 3e-05,
+      "loss": 4.0404,
+      "step": 2333
+    },
+    {
+      "epoch": 0.6342391304347826,
+      "grad_norm": 2.052917718887329,
+      "learning_rate": 3e-05,
+      "loss": 4.1781,
+      "step": 2334
+    },
+    {
+      "epoch": 0.6345108695652174,
+      "grad_norm": 1.376478672027588,
+      "learning_rate": 3e-05,
+      "loss": 3.4097,
+      "step": 2335
+    },
+    {
+      "epoch": 0.6347826086956522,
+      "grad_norm": 1.3323421478271484,
+      "learning_rate": 3e-05,
+      "loss": 3.0153,
+      "step": 2336
+    },
+    {
+      "epoch": 0.6350543478260869,
+      "grad_norm": 1.7771837711334229,
+      "learning_rate": 3e-05,
+      "loss": 3.7671,
+      "step": 2337
+    },
+    {
+      "epoch": 0.6353260869565217,
+      "grad_norm": 2.060122489929199,
+      "learning_rate": 3e-05,
+      "loss": 3.5854,
+      "step": 2338
+    },
+    {
+      "epoch": 0.6355978260869565,
+      "grad_norm": 1.3856672048568726,
+      "learning_rate": 3e-05,
+      "loss": 3.0497,
+      "step": 2339
+    },
+    {
+      "epoch": 0.6358695652173914,
+      "grad_norm": 1.3146904706954956,
+      "learning_rate": 3e-05,
+      "loss": 3.1533,
+      "step": 2340
+    },
+    {
+      "epoch": 0.6361413043478261,
+      "grad_norm": 1.4848661422729492,
+      "learning_rate": 3e-05,
+      "loss": 3.5613,
+      "step": 2341
+    },
+    {
+      "epoch": 0.6364130434782609,
+      "grad_norm": 1.4037156105041504,
+      "learning_rate": 3e-05,
+      "loss": 3.8576,
+      "step": 2342
+    },
+    {
+      "epoch": 0.6366847826086957,
+      "grad_norm": 1.288294792175293,
+      "learning_rate": 3e-05,
+      "loss": 3.5626,
+      "step": 2343
+    },
+    {
+      "epoch": 0.6369565217391304,
+      "grad_norm": 1.080474853515625,
+      "learning_rate": 3e-05,
+      "loss": 3.2761,
+      "step": 2344
+    },
+    {
+      "epoch": 0.6372282608695652,
+      "grad_norm": 1.1792863607406616,
+      "learning_rate": 3e-05,
+      "loss": 2.9453,
+      "step": 2345
+    },
+    {
+      "epoch": 0.6375,
+      "grad_norm": 1.4226300716400146,
+      "learning_rate": 3e-05,
+      "loss": 2.8206,
+      "step": 2346
+    },
+    {
+      "epoch": 0.6377717391304348,
+      "grad_norm": 1.5433789491653442,
+      "learning_rate": 3e-05,
+      "loss": 3.4793,
+      "step": 2347
+    },
+    {
+      "epoch": 0.6380434782608696,
+      "grad_norm": 1.0570566654205322,
+      "learning_rate": 3e-05,
+      "loss": 3.0796,
+      "step": 2348
+    },
+    {
+      "epoch": 0.6383152173913044,
+      "grad_norm": 1.4246288537979126,
+      "learning_rate": 3e-05,
+      "loss": 3.259,
+      "step": 2349
+    },
+    {
+      "epoch": 0.6385869565217391,
+      "grad_norm": 2.0091404914855957,
+      "learning_rate": 3e-05,
+      "loss": 3.2464,
+      "step": 2350
+    },
+    {
+      "epoch": 0.6388586956521739,
+      "grad_norm": 1.7661538124084473,
+      "learning_rate": 3e-05,
+      "loss": 3.0348,
+      "step": 2351
+    },
+    {
+      "epoch": 0.6391304347826087,
+      "grad_norm": 1.6644985675811768,
+      "learning_rate": 3e-05,
+      "loss": 3.2656,
+      "step": 2352
+    },
+    {
+      "epoch": 0.6394021739130434,
+      "grad_norm": 1.192597508430481,
+      "learning_rate": 3e-05,
+      "loss": 3.2846,
+      "step": 2353
+    },
+    {
+      "epoch": 0.6396739130434783,
+      "grad_norm": 1.234485149383545,
+      "learning_rate": 3e-05,
+      "loss": 3.2553,
+      "step": 2354
+    },
+    {
+      "epoch": 0.6399456521739131,
+      "grad_norm": 1.2841020822525024,
+      "learning_rate": 3e-05,
+      "loss": 2.8951,
+      "step": 2355
+    },
+    {
+      "epoch": 0.6402173913043478,
+      "grad_norm": 1.119943618774414,
+      "learning_rate": 3e-05,
+      "loss": 2.7422,
+      "step": 2356
+    },
+    {
+      "epoch": 0.6404891304347826,
+      "grad_norm": 1.5436328649520874,
+      "learning_rate": 3e-05,
+      "loss": 3.9712,
+      "step": 2357
+    },
+    {
+      "epoch": 0.6407608695652174,
+      "grad_norm": 1.1272965669631958,
+      "learning_rate": 3e-05,
+      "loss": 3.1857,
+      "step": 2358
+    },
+    {
+      "epoch": 0.6410326086956522,
+      "grad_norm": 1.340259075164795,
+      "learning_rate": 3e-05,
+      "loss": 3.7488,
+      "step": 2359
+    },
+    {
+      "epoch": 0.6413043478260869,
+      "grad_norm": 1.486802577972412,
+      "learning_rate": 3e-05,
+      "loss": 3.8155,
+      "step": 2360
+    },
+    {
+      "epoch": 0.6415760869565217,
+      "grad_norm": 0.9581712484359741,
+      "learning_rate": 3e-05,
+      "loss": 3.117,
+      "step": 2361
+    },
+    {
+      "epoch": 0.6418478260869566,
+      "grad_norm": 1.012144923210144,
+      "learning_rate": 3e-05,
+      "loss": 3.0744,
+      "step": 2362
+    },
+    {
+      "epoch": 0.6421195652173913,
+      "grad_norm": 1.1179242134094238,
+      "learning_rate": 3e-05,
+      "loss": 3.2825,
+      "step": 2363
+    },
+    {
+      "epoch": 0.6423913043478261,
+      "grad_norm": 1.3681845664978027,
+      "learning_rate": 3e-05,
+      "loss": 3.1833,
+      "step": 2364
+    },
+    {
+      "epoch": 0.6426630434782609,
+      "grad_norm": 1.192631721496582,
+      "learning_rate": 3e-05,
+      "loss": 3.3842,
+      "step": 2365
+    },
+    {
+      "epoch": 0.6429347826086956,
+      "grad_norm": 1.2317907810211182,
+      "learning_rate": 3e-05,
+      "loss": 3.1083,
+      "step": 2366
+    },
+    {
+      "epoch": 0.6432065217391304,
+      "grad_norm": 1.0752184391021729,
+      "learning_rate": 3e-05,
+      "loss": 3.2455,
+      "step": 2367
+    },
+    {
+      "epoch": 0.6434782608695652,
+      "grad_norm": 1.035239338874817,
+      "learning_rate": 3e-05,
+      "loss": 2.9436,
+      "step": 2368
+    },
+    {
+      "epoch": 0.64375,
+      "grad_norm": 1.3296703100204468,
+      "learning_rate": 3e-05,
+      "loss": 3.2708,
+      "step": 2369
+    },
+    {
+      "epoch": 0.6440217391304348,
+      "grad_norm": 1.2996970415115356,
+      "learning_rate": 3e-05,
+      "loss": 3.8312,
+      "step": 2370
+    },
+    {
+      "epoch": 0.6442934782608696,
+      "grad_norm": 0.9831019639968872,
+      "learning_rate": 3e-05,
+      "loss": 3.2478,
+      "step": 2371
+    },
+    {
+      "epoch": 0.6445652173913043,
+      "grad_norm": 0.9067811965942383,
+      "learning_rate": 3e-05,
+      "loss": 3.1556,
+      "step": 2372
+    },
+    {
+      "epoch": 0.6448369565217391,
+      "grad_norm": 1.2001372575759888,
+      "learning_rate": 3e-05,
+      "loss": 3.7225,
+      "step": 2373
+    },
+    {
+      "epoch": 0.6451086956521739,
+      "grad_norm": 1.1935441493988037,
+      "learning_rate": 3e-05,
+      "loss": 3.3109,
+      "step": 2374
+    },
+    {
+      "epoch": 0.6453804347826086,
+      "grad_norm": 0.9141666889190674,
+      "learning_rate": 3e-05,
+      "loss": 3.0955,
+      "step": 2375
+    },
+    {
+      "epoch": 0.6456521739130435,
+      "grad_norm": 1.0943185091018677,
+      "learning_rate": 3e-05,
+      "loss": 3.3094,
+      "step": 2376
+    },
+    {
+      "epoch": 0.6459239130434783,
+      "grad_norm": 1.1921093463897705,
+      "learning_rate": 3e-05,
+      "loss": 3.3715,
+      "step": 2377
+    },
+    {
+      "epoch": 0.6461956521739131,
+      "grad_norm": 1.0566505193710327,
+      "learning_rate": 3e-05,
+      "loss": 3.1073,
+      "step": 2378
+    },
+    {
+      "epoch": 0.6464673913043478,
+      "grad_norm": 1.5691936016082764,
+      "learning_rate": 3e-05,
+      "loss": 3.4321,
+      "step": 2379
+    },
+    {
+      "epoch": 0.6467391304347826,
+      "grad_norm": 1.041152000427246,
+      "learning_rate": 3e-05,
+      "loss": 2.6529,
+      "step": 2380
+    },
+    {
+      "epoch": 0.6470108695652174,
+      "grad_norm": 1.2687933444976807,
+      "learning_rate": 3e-05,
+      "loss": 2.9547,
+      "step": 2381
+    },
+    {
+      "epoch": 0.6472826086956521,
+      "grad_norm": 1.074779748916626,
+      "learning_rate": 3e-05,
+      "loss": 3.0635,
+      "step": 2382
+    },
+    {
+      "epoch": 0.647554347826087,
+      "grad_norm": 1.0904431343078613,
+      "learning_rate": 3e-05,
+      "loss": 2.8928,
+      "step": 2383
+    },
+    {
+      "epoch": 0.6478260869565218,
+      "grad_norm": 1.3329259157180786,
+      "learning_rate": 3e-05,
+      "loss": 3.6509,
+      "step": 2384
+    },
+    {
+      "epoch": 0.6480978260869565,
+      "grad_norm": 1.2522311210632324,
+      "learning_rate": 3e-05,
+      "loss": 3.6357,
+      "step": 2385
+    },
+    {
+      "epoch": 0.6483695652173913,
+      "grad_norm": 1.2754337787628174,
+      "learning_rate": 3e-05,
+      "loss": 3.7422,
+      "step": 2386
+    },
+    {
+      "epoch": 0.6486413043478261,
+      "grad_norm": 1.338080883026123,
+      "learning_rate": 3e-05,
+      "loss": 3.3859,
+      "step": 2387
+    },
+    {
+      "epoch": 0.6489130434782608,
+      "grad_norm": 1.1422319412231445,
+      "learning_rate": 3e-05,
+      "loss": 3.8218,
+      "step": 2388
+    },
+    {
+      "epoch": 0.6491847826086956,
+      "grad_norm": 1.647072196006775,
+      "learning_rate": 3e-05,
+      "loss": 3.6864,
+      "step": 2389
+    },
+    {
+      "epoch": 0.6494565217391305,
+      "grad_norm": 1.1428362131118774,
+      "learning_rate": 3e-05,
+      "loss": 2.8359,
+      "step": 2390
+    },
+    {
+      "epoch": 0.6497282608695653,
+      "grad_norm": 1.1382925510406494,
+      "learning_rate": 3e-05,
+      "loss": 3.1163,
+      "step": 2391
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0694186687469482,
+      "learning_rate": 3e-05,
+      "loss": 3.1842,
+      "step": 2392
+    },
+    {
+      "epoch": 0.6502717391304348,
+      "grad_norm": 1.413272738456726,
+      "learning_rate": 3e-05,
+      "loss": 3.7837,
+      "step": 2393
+    },
+    {
+      "epoch": 0.6505434782608696,
+      "grad_norm": 1.353119134902954,
+      "learning_rate": 3e-05,
+      "loss": 3.2845,
+      "step": 2394
+    },
+    {
+      "epoch": 0.6508152173913043,
+      "grad_norm": 1.542140245437622,
+      "learning_rate": 3e-05,
+      "loss": 3.4894,
+      "step": 2395
+    },
+    {
+      "epoch": 0.6510869565217391,
+      "grad_norm": 1.6058558225631714,
+      "learning_rate": 3e-05,
+      "loss": 3.4778,
+      "step": 2396
+    },
+    {
+      "epoch": 0.6513586956521739,
+      "grad_norm": 1.719277262687683,
+      "learning_rate": 3e-05,
+      "loss": 3.5328,
+      "step": 2397
+    },
+    {
+      "epoch": 0.6516304347826087,
+      "grad_norm": 1.7350586652755737,
+      "learning_rate": 3e-05,
+      "loss": 3.5851,
+      "step": 2398
+    },
+    {
+      "epoch": 0.6519021739130435,
+      "grad_norm": 1.4615671634674072,
+      "learning_rate": 3e-05,
+      "loss": 4.099,
+      "step": 2399
+    },
+    {
+      "epoch": 0.6521739130434783,
+      "grad_norm": 1.194122314453125,
+      "learning_rate": 3e-05,
+      "loss": 3.353,
+      "step": 2400
+    },
+    {
+      "epoch": 0.652445652173913,
+      "grad_norm": 1.4127830266952515,
+      "learning_rate": 3e-05,
+      "loss": 3.8697,
+      "step": 2401
+    },
+    {
+      "epoch": 0.6527173913043478,
+      "grad_norm": 1.6707136631011963,
+      "learning_rate": 3e-05,
+      "loss": 3.9621,
+      "step": 2402
+    },
+    {
+      "epoch": 0.6529891304347826,
+      "grad_norm": 1.4606815576553345,
+      "learning_rate": 3e-05,
+      "loss": 3.7677,
+      "step": 2403
+    },
+    {
+      "epoch": 0.6532608695652173,
+      "grad_norm": 1.2474243640899658,
+      "learning_rate": 3e-05,
+      "loss": 3.5581,
+      "step": 2404
+    },
+    {
+      "epoch": 0.6535326086956522,
+      "grad_norm": 1.1824653148651123,
+      "learning_rate": 3e-05,
+      "loss": 2.9751,
+      "step": 2405
+    },
+    {
+      "epoch": 0.653804347826087,
+      "grad_norm": 1.22133469581604,
+      "learning_rate": 3e-05,
+      "loss": 3.3986,
+      "step": 2406
+    },
+    {
+      "epoch": 0.6540760869565218,
+      "grad_norm": 1.160557746887207,
+      "learning_rate": 3e-05,
+      "loss": 3.2008,
+      "step": 2407
+    },
+    {
+      "epoch": 0.6543478260869565,
+      "grad_norm": 1.2002159357070923,
+      "learning_rate": 3e-05,
+      "loss": 3.6029,
+      "step": 2408
+    },
+    {
+      "epoch": 0.6546195652173913,
+      "grad_norm": 1.1110812425613403,
+      "learning_rate": 3e-05,
+      "loss": 3.2051,
+      "step": 2409
+    },
+    {
+      "epoch": 0.654891304347826,
+      "grad_norm": 0.9706496596336365,
+      "learning_rate": 3e-05,
+      "loss": 2.9527,
+      "step": 2410
+    },
+    {
+      "epoch": 0.6551630434782608,
+      "grad_norm": 1.0990405082702637,
+      "learning_rate": 3e-05,
+      "loss": 2.6553,
+      "step": 2411
+    },
+    {
+      "epoch": 0.6554347826086957,
+      "grad_norm": 1.065306544303894,
+      "learning_rate": 3e-05,
+      "loss": 3.3728,
+      "step": 2412
+    },
+    {
+      "epoch": 0.6557065217391305,
+      "grad_norm": 0.956051766872406,
+      "learning_rate": 3e-05,
+      "loss": 2.9055,
+      "step": 2413
+    },
+    {
+      "epoch": 0.6559782608695652,
+      "grad_norm": 1.1310175657272339,
+      "learning_rate": 3e-05,
+      "loss": 3.8269,
+      "step": 2414
+    },
+    {
+      "epoch": 0.65625,
+      "grad_norm": 0.9513333439826965,
+      "learning_rate": 3e-05,
+      "loss": 3.2811,
+      "step": 2415
+    },
+    {
+      "epoch": 0.6565217391304348,
+      "grad_norm": 1.301697015762329,
+      "learning_rate": 3e-05,
+      "loss": 3.5579,
+      "step": 2416
+    },
+    {
+      "epoch": 0.6567934782608695,
+      "grad_norm": 0.9810994267463684,
+      "learning_rate": 3e-05,
+      "loss": 2.9914,
+      "step": 2417
+    },
+    {
+      "epoch": 0.6570652173913043,
+      "grad_norm": 1.3848271369934082,
+      "learning_rate": 3e-05,
+      "loss": 3.8109,
+      "step": 2418
+    },
+    {
+      "epoch": 0.6573369565217392,
+      "grad_norm": 1.169350504875183,
+      "learning_rate": 3e-05,
+      "loss": 2.8291,
+      "step": 2419
+    },
+    {
+      "epoch": 0.657608695652174,
+      "grad_norm": 1.3478142023086548,
+      "learning_rate": 3e-05,
+      "loss": 3.5609,
+      "step": 2420
+    },
+    {
+      "epoch": 0.6578804347826087,
+      "grad_norm": 1.1918085813522339,
+      "learning_rate": 3e-05,
+      "loss": 3.5339,
+      "step": 2421
+    },
+    {
+      "epoch": 0.6581521739130435,
+      "grad_norm": 1.4257116317749023,
+      "learning_rate": 3e-05,
+      "loss": 3.4636,
+      "step": 2422
+    },
+    {
+      "epoch": 0.6584239130434782,
+      "grad_norm": 1.2269481420516968,
+      "learning_rate": 3e-05,
+      "loss": 3.2741,
+      "step": 2423
+    },
+    {
+      "epoch": 0.658695652173913,
+      "grad_norm": 1.2066056728363037,
+      "learning_rate": 3e-05,
+      "loss": 3.2177,
+      "step": 2424
+    },
+    {
+      "epoch": 0.6589673913043478,
+      "grad_norm": 1.3328955173492432,
+      "learning_rate": 3e-05,
+      "loss": 3.527,
+      "step": 2425
+    },
+    {
+      "epoch": 0.6592391304347827,
+      "grad_norm": 1.1746604442596436,
+      "learning_rate": 3e-05,
+      "loss": 3.1194,
+      "step": 2426
+    },
+    {
+      "epoch": 0.6595108695652174,
+      "grad_norm": 1.028620958328247,
+      "learning_rate": 3e-05,
+      "loss": 3.1119,
+      "step": 2427
+    },
+    {
+      "epoch": 0.6597826086956522,
+      "grad_norm": 1.0080275535583496,
+      "learning_rate": 3e-05,
+      "loss": 2.7994,
+      "step": 2428
+    },
+    {
+      "epoch": 0.660054347826087,
+      "grad_norm": 0.8145027756690979,
+      "learning_rate": 3e-05,
+      "loss": 2.9428,
+      "step": 2429
+    },
+    {
+      "epoch": 0.6603260869565217,
+      "grad_norm": 1.0437175035476685,
+      "learning_rate": 3e-05,
+      "loss": 2.9878,
+      "step": 2430
+    },
+    {
+      "epoch": 0.6605978260869565,
+      "grad_norm": 1.130467414855957,
+      "learning_rate": 3e-05,
+      "loss": 3.7007,
+      "step": 2431
+    },
+    {
+      "epoch": 0.6608695652173913,
+      "grad_norm": 1.0865567922592163,
+      "learning_rate": 3e-05,
+      "loss": 3.1175,
+      "step": 2432
+    },
+    {
+      "epoch": 0.6611413043478261,
+      "grad_norm": 1.1606049537658691,
+      "learning_rate": 3e-05,
+      "loss": 3.3582,
+      "step": 2433
+    },
+    {
+      "epoch": 0.6614130434782609,
+      "grad_norm": 0.9023473262786865,
+      "learning_rate": 3e-05,
+      "loss": 2.9861,
+      "step": 2434
+    },
+    {
+      "epoch": 0.6616847826086957,
+      "grad_norm": 0.856779932975769,
+      "learning_rate": 3e-05,
+      "loss": 3.0402,
+      "step": 2435
+    },
+    {
+      "epoch": 0.6619565217391304,
+      "grad_norm": 1.0253337621688843,
+      "learning_rate": 3e-05,
+      "loss": 3.0909,
+      "step": 2436
+    },
+    {
+      "epoch": 0.6622282608695652,
+      "grad_norm": 1.0018707513809204,
+      "learning_rate": 3e-05,
+      "loss": 2.9996,
+      "step": 2437
+    },
+    {
+      "epoch": 0.6625,
+      "grad_norm": 1.0722147226333618,
+      "learning_rate": 3e-05,
+      "loss": 3.3102,
+      "step": 2438
+    },
+    {
+      "epoch": 0.6627717391304347,
+      "grad_norm": 1.0798076391220093,
+      "learning_rate": 3e-05,
+      "loss": 2.9907,
+      "step": 2439
+    },
+    {
+      "epoch": 0.6630434782608695,
+      "grad_norm": 1.1593507528305054,
+      "learning_rate": 3e-05,
+      "loss": 3.3741,
+      "step": 2440
+    },
+    {
+      "epoch": 0.6633152173913044,
+      "grad_norm": 1.0947400331497192,
+      "learning_rate": 3e-05,
+      "loss": 3.0087,
+      "step": 2441
+    },
+    {
+      "epoch": 0.6635869565217392,
+      "grad_norm": 1.2882115840911865,
+      "learning_rate": 3e-05,
+      "loss": 3.1978,
+      "step": 2442
+    },
+    {
+      "epoch": 0.6638586956521739,
+      "grad_norm": 1.2732001543045044,
+      "learning_rate": 3e-05,
+      "loss": 3.8371,
+      "step": 2443
+    },
+    {
+      "epoch": 0.6641304347826087,
+      "grad_norm": 0.9186998605728149,
+      "learning_rate": 3e-05,
+      "loss": 2.643,
+      "step": 2444
+    },
+    {
+      "epoch": 0.6644021739130435,
+      "grad_norm": 1.0889346599578857,
+      "learning_rate": 3e-05,
+      "loss": 2.9247,
+      "step": 2445
+    },
+    {
+      "epoch": 0.6646739130434782,
+      "grad_norm": 1.094017505645752,
+      "learning_rate": 3e-05,
+      "loss": 3.1954,
+      "step": 2446
+    },
+    {
+      "epoch": 0.664945652173913,
+      "grad_norm": 0.9859854578971863,
+      "learning_rate": 3e-05,
+      "loss": 2.7335,
+      "step": 2447
+    },
+    {
+      "epoch": 0.6652173913043479,
+      "grad_norm": 1.8823328018188477,
+      "learning_rate": 3e-05,
+      "loss": 3.4934,
+      "step": 2448
+    },
+    {
+      "epoch": 0.6654891304347826,
+      "grad_norm": 0.9667043685913086,
+      "learning_rate": 3e-05,
+      "loss": 3.0711,
+      "step": 2449
+    },
+    {
+      "epoch": 0.6657608695652174,
+      "grad_norm": 1.0037490129470825,
+      "learning_rate": 3e-05,
+      "loss": 3.1836,
+      "step": 2450
+    },
+    {
+      "epoch": 0.6660326086956522,
+      "grad_norm": 1.0636874437332153,
+      "learning_rate": 3e-05,
+      "loss": 3.3269,
+      "step": 2451
+    },
+    {
+      "epoch": 0.6663043478260869,
+      "grad_norm": 1.02604079246521,
+      "learning_rate": 3e-05,
+      "loss": 3.4928,
+      "step": 2452
+    },
+    {
+      "epoch": 0.6665760869565217,
+      "grad_norm": 1.050881028175354,
+      "learning_rate": 3e-05,
+      "loss": 3.3983,
+      "step": 2453
+    },
+    {
+      "epoch": 0.6668478260869565,
+      "grad_norm": 1.1651184558868408,
+      "learning_rate": 3e-05,
+      "loss": 3.573,
+      "step": 2454
+    },
+    {
+      "epoch": 0.6671195652173914,
+      "grad_norm": 0.9464408755302429,
+      "learning_rate": 3e-05,
+      "loss": 3.4415,
+      "step": 2455
+    },
+    {
+      "epoch": 0.6673913043478261,
+      "grad_norm": 0.9894149899482727,
+      "learning_rate": 3e-05,
+      "loss": 3.4124,
+      "step": 2456
+    },
+    {
+      "epoch": 0.6676630434782609,
+      "grad_norm": 1.1918611526489258,
+      "learning_rate": 3e-05,
+      "loss": 3.1563,
+      "step": 2457
+    },
+    {
+      "epoch": 0.6679347826086957,
+      "grad_norm": 1.1251368522644043,
+      "learning_rate": 3e-05,
+      "loss": 3.0224,
+      "step": 2458
+    },
+    {
+      "epoch": 0.6682065217391304,
+      "grad_norm": 0.9575350880622864,
+      "learning_rate": 3e-05,
+      "loss": 3.1342,
+      "step": 2459
+    },
+    {
+      "epoch": 0.6684782608695652,
+      "grad_norm": 1.1674573421478271,
+      "learning_rate": 3e-05,
+      "loss": 3.5287,
+      "step": 2460
+    },
+    {
+      "epoch": 0.66875,
+      "grad_norm": 1.3104429244995117,
+      "learning_rate": 3e-05,
+      "loss": 3.3732,
+      "step": 2461
+    },
+    {
+      "epoch": 0.6690217391304348,
+      "grad_norm": 1.0205787420272827,
+      "learning_rate": 3e-05,
+      "loss": 3.3107,
+      "step": 2462
+    },
+    {
+      "epoch": 0.6692934782608696,
+      "grad_norm": 1.1491562128067017,
+      "learning_rate": 3e-05,
+      "loss": 3.1578,
+      "step": 2463
+    },
+    {
+      "epoch": 0.6695652173913044,
+      "grad_norm": 1.2550066709518433,
+      "learning_rate": 3e-05,
+      "loss": 2.9625,
+      "step": 2464
+    },
+    {
+      "epoch": 0.6698369565217391,
+      "grad_norm": 1.396528720855713,
+      "learning_rate": 3e-05,
+      "loss": 3.4526,
+      "step": 2465
+    },
+    {
+      "epoch": 0.6701086956521739,
+      "grad_norm": 1.3484258651733398,
+      "learning_rate": 3e-05,
+      "loss": 3.0718,
+      "step": 2466
+    },
+    {
+      "epoch": 0.6703804347826087,
+      "grad_norm": 1.2837543487548828,
+      "learning_rate": 3e-05,
+      "loss": 3.9352,
+      "step": 2467
+    },
+    {
+      "epoch": 0.6706521739130434,
+      "grad_norm": 1.126395583152771,
+      "learning_rate": 3e-05,
+      "loss": 3.2975,
+      "step": 2468
+    },
+    {
+      "epoch": 0.6709239130434783,
+      "grad_norm": 1.0975403785705566,
+      "learning_rate": 3e-05,
+      "loss": 2.9911,
+      "step": 2469
+    },
+    {
+      "epoch": 0.6711956521739131,
+      "grad_norm": 1.260852336883545,
+      "learning_rate": 3e-05,
+      "loss": 3.4067,
+      "step": 2470
+    },
+    {
+      "epoch": 0.6714673913043478,
+      "grad_norm": 1.3457390069961548,
+      "learning_rate": 3e-05,
+      "loss": 3.3277,
+      "step": 2471
+    },
+    {
+      "epoch": 0.6717391304347826,
+      "grad_norm": 1.6091570854187012,
+      "learning_rate": 3e-05,
+      "loss": 3.8467,
+      "step": 2472
+    },
+    {
+      "epoch": 0.6720108695652174,
+      "grad_norm": 1.3068939447402954,
+      "learning_rate": 3e-05,
+      "loss": 3.5756,
+      "step": 2473
+    },
+    {
+      "epoch": 0.6722826086956522,
+      "grad_norm": 1.532850742340088,
+      "learning_rate": 3e-05,
+      "loss": 3.6188,
+      "step": 2474
+    },
+    {
+      "epoch": 0.6725543478260869,
+      "grad_norm": 1.179315447807312,
+      "learning_rate": 3e-05,
+      "loss": 3.3408,
+      "step": 2475
+    },
+    {
+      "epoch": 0.6728260869565217,
+      "grad_norm": 1.278639793395996,
+      "learning_rate": 3e-05,
+      "loss": 3.9328,
+      "step": 2476
+    },
+    {
+      "epoch": 0.6730978260869566,
+      "grad_norm": 0.9712943434715271,
+      "learning_rate": 3e-05,
+      "loss": 2.7786,
+      "step": 2477
+    },
+    {
+      "epoch": 0.6733695652173913,
+      "grad_norm": 1.1749228239059448,
+      "learning_rate": 3e-05,
+      "loss": 2.7182,
+      "step": 2478
+    },
+    {
+      "epoch": 0.6736413043478261,
+      "grad_norm": 1.3901921510696411,
+      "learning_rate": 3e-05,
+      "loss": 3.2455,
+      "step": 2479
+    },
+    {
+      "epoch": 0.6739130434782609,
+      "grad_norm": 1.3088897466659546,
+      "learning_rate": 3e-05,
+      "loss": 3.4134,
+      "step": 2480
+    },
+    {
+      "epoch": 0.6741847826086956,
+      "grad_norm": 1.2072702646255493,
+      "learning_rate": 3e-05,
+      "loss": 3.7264,
+      "step": 2481
+    },
+    {
+      "epoch": 0.6744565217391304,
+      "grad_norm": 1.3006356954574585,
+      "learning_rate": 3e-05,
+      "loss": 3.1591,
+      "step": 2482
+    },
+    {
+      "epoch": 0.6747282608695652,
+      "grad_norm": 2.3080663681030273,
+      "learning_rate": 3e-05,
+      "loss": 3.6425,
+      "step": 2483
+    },
+    {
+      "epoch": 0.675,
+      "grad_norm": 1.3426471948623657,
+      "learning_rate": 3e-05,
+      "loss": 2.9936,
+      "step": 2484
+    },
+    {
+      "epoch": 0.6752717391304348,
+      "grad_norm": 1.1594306230545044,
+      "learning_rate": 3e-05,
+      "loss": 3.3481,
+      "step": 2485
+    },
+    {
+      "epoch": 0.6755434782608696,
+      "grad_norm": 1.0936838388442993,
+      "learning_rate": 3e-05,
+      "loss": 3.0749,
+      "step": 2486
+    },
+    {
+      "epoch": 0.6758152173913043,
+      "grad_norm": 1.3417373895645142,
+      "learning_rate": 3e-05,
+      "loss": 3.2727,
+      "step": 2487
+    },
+    {
+      "epoch": 0.6760869565217391,
+      "grad_norm": 1.6536362171173096,
+      "learning_rate": 3e-05,
+      "loss": 3.3898,
+      "step": 2488
+    },
+    {
+      "epoch": 0.6763586956521739,
+      "grad_norm": 1.2044962644577026,
+      "learning_rate": 3e-05,
+      "loss": 3.06,
+      "step": 2489
+    },
+    {
+      "epoch": 0.6766304347826086,
+      "grad_norm": 1.6772176027297974,
+      "learning_rate": 3e-05,
+      "loss": 3.6963,
+      "step": 2490
+    },
+    {
+      "epoch": 0.6769021739130435,
+      "grad_norm": 1.2547991275787354,
+      "learning_rate": 3e-05,
+      "loss": 2.9258,
+      "step": 2491
+    },
+    {
+      "epoch": 0.6771739130434783,
+      "grad_norm": 1.2318508625030518,
+      "learning_rate": 3e-05,
+      "loss": 3.4305,
+      "step": 2492
+    },
+    {
+      "epoch": 0.6774456521739131,
+      "grad_norm": 1.4965567588806152,
+      "learning_rate": 3e-05,
+      "loss": 3.4216,
+      "step": 2493
+    },
+    {
+      "epoch": 0.6777173913043478,
+      "grad_norm": 1.6711788177490234,
+      "learning_rate": 3e-05,
+      "loss": 2.8026,
+      "step": 2494
+    },
+    {
+      "epoch": 0.6779891304347826,
+      "grad_norm": 1.285457968711853,
+      "learning_rate": 3e-05,
+      "loss": 2.9369,
+      "step": 2495
+    },
+    {
+      "epoch": 0.6782608695652174,
+      "grad_norm": 1.1391198635101318,
+      "learning_rate": 3e-05,
+      "loss": 2.6831,
+      "step": 2496
+    },
+    {
+      "epoch": 0.6785326086956521,
+      "grad_norm": 1.4873690605163574,
+      "learning_rate": 3e-05,
+      "loss": 3.5054,
+      "step": 2497
+    },
+    {
+      "epoch": 0.678804347826087,
+      "grad_norm": 1.5623383522033691,
+      "learning_rate": 3e-05,
+      "loss": 3.5831,
+      "step": 2498
+    },
+    {
+      "epoch": 0.6790760869565218,
+      "grad_norm": 1.0868808031082153,
+      "learning_rate": 3e-05,
+      "loss": 3.5378,
+      "step": 2499
+    },
+    {
+      "epoch": 0.6793478260869565,
+      "grad_norm": 1.5260769128799438,
+      "learning_rate": 3e-05,
+      "loss": 3.6442,
+      "step": 2500
+    },
+    {
+      "epoch": 0.6796195652173913,
+      "grad_norm": 1.5094234943389893,
+      "learning_rate": 3e-05,
+      "loss": 2.8721,
+      "step": 2501
+    },
+    {
+      "epoch": 0.6798913043478261,
+      "grad_norm": 1.03408944606781,
+      "learning_rate": 3e-05,
+      "loss": 3.2489,
+      "step": 2502
+    },
+    {
+      "epoch": 0.6801630434782608,
+      "grad_norm": 1.5139715671539307,
+      "learning_rate": 3e-05,
+      "loss": 3.1632,
+      "step": 2503
+    },
+    {
+      "epoch": 0.6804347826086956,
+      "grad_norm": 0.8952838778495789,
+      "learning_rate": 3e-05,
+      "loss": 2.678,
+      "step": 2504
+    },
+    {
+      "epoch": 0.6807065217391305,
+      "grad_norm": 0.886978268623352,
+      "learning_rate": 3e-05,
+      "loss": 2.9663,
+      "step": 2505
+    },
+    {
+      "epoch": 0.6809782608695653,
+      "grad_norm": 1.4015369415283203,
+      "learning_rate": 3e-05,
+      "loss": 3.4536,
+      "step": 2506
+    },
+    {
+      "epoch": 0.68125,
+      "grad_norm": 1.1252248287200928,
+      "learning_rate": 3e-05,
+      "loss": 3.1302,
+      "step": 2507
+    },
+    {
+      "epoch": 0.6815217391304348,
+      "grad_norm": 1.2304083108901978,
+      "learning_rate": 3e-05,
+      "loss": 3.5204,
+      "step": 2508
+    },
+    {
+      "epoch": 0.6817934782608696,
+      "grad_norm": 1.2945035696029663,
+      "learning_rate": 3e-05,
+      "loss": 3.7562,
+      "step": 2509
+    },
+    {
+      "epoch": 0.6820652173913043,
+      "grad_norm": 1.1755324602127075,
+      "learning_rate": 3e-05,
+      "loss": 2.787,
+      "step": 2510
+    },
+    {
+      "epoch": 0.6823369565217391,
+      "grad_norm": 1.413987159729004,
+      "learning_rate": 3e-05,
+      "loss": 3.5305,
+      "step": 2511
+    },
+    {
+      "epoch": 0.6826086956521739,
+      "grad_norm": 1.0807994604110718,
+      "learning_rate": 3e-05,
+      "loss": 3.2698,
+      "step": 2512
+    },
+    {
+      "epoch": 0.6828804347826087,
+      "grad_norm": 1.5065345764160156,
+      "learning_rate": 3e-05,
+      "loss": 3.3808,
+      "step": 2513
+    },
+    {
+      "epoch": 0.6831521739130435,
+      "grad_norm": 1.2206026315689087,
+      "learning_rate": 3e-05,
+      "loss": 2.9492,
+      "step": 2514
+    },
+    {
+      "epoch": 0.6834239130434783,
+      "grad_norm": 1.1851751804351807,
+      "learning_rate": 3e-05,
+      "loss": 3.425,
+      "step": 2515
+    },
+    {
+      "epoch": 0.683695652173913,
+      "grad_norm": 1.1727910041809082,
+      "learning_rate": 3e-05,
+      "loss": 3.6993,
+      "step": 2516
+    },
+    {
+      "epoch": 0.6839673913043478,
+      "grad_norm": 1.5769599676132202,
+      "learning_rate": 3e-05,
+      "loss": 3.3359,
+      "step": 2517
+    },
+    {
+      "epoch": 0.6842391304347826,
+      "grad_norm": 1.5909152030944824,
+      "learning_rate": 3e-05,
+      "loss": 3.1606,
+      "step": 2518
+    },
+    {
+      "epoch": 0.6845108695652173,
+      "grad_norm": 1.094890832901001,
+      "learning_rate": 3e-05,
+      "loss": 3.2958,
+      "step": 2519
+    },
+    {
+      "epoch": 0.6847826086956522,
+      "grad_norm": 1.5951069593429565,
+      "learning_rate": 3e-05,
+      "loss": 3.6897,
+      "step": 2520
+    },
+    {
+      "epoch": 0.685054347826087,
+      "grad_norm": 1.3260122537612915,
+      "learning_rate": 3e-05,
+      "loss": 3.5941,
+      "step": 2521
+    },
+    {
+      "epoch": 0.6853260869565218,
+      "grad_norm": 1.431256651878357,
+      "learning_rate": 3e-05,
+      "loss": 2.8581,
+      "step": 2522
+    },
+    {
+      "epoch": 0.6855978260869565,
+      "grad_norm": 1.3032617568969727,
+      "learning_rate": 3e-05,
+      "loss": 3.3358,
+      "step": 2523
+    },
+    {
+      "epoch": 0.6858695652173913,
+      "grad_norm": 1.0501641035079956,
+      "learning_rate": 3e-05,
+      "loss": 2.9872,
+      "step": 2524
+    },
+    {
+      "epoch": 0.686141304347826,
+      "grad_norm": 1.1217620372772217,
+      "learning_rate": 3e-05,
+      "loss": 3.1064,
+      "step": 2525
+    },
+    {
+      "epoch": 0.6864130434782608,
+      "grad_norm": 1.0204346179962158,
+      "learning_rate": 3e-05,
+      "loss": 3.13,
+      "step": 2526
+    },
+    {
+      "epoch": 0.6866847826086957,
+      "grad_norm": 1.479447603225708,
+      "learning_rate": 3e-05,
+      "loss": 3.3739,
+      "step": 2527
+    },
+    {
+      "epoch": 0.6869565217391305,
+      "grad_norm": 1.3281017541885376,
+      "learning_rate": 3e-05,
+      "loss": 3.1517,
+      "step": 2528
+    },
+    {
+      "epoch": 0.6872282608695652,
+      "grad_norm": 1.358286738395691,
+      "learning_rate": 3e-05,
+      "loss": 3.215,
+      "step": 2529
+    },
+    {
+      "epoch": 0.6875,
+      "grad_norm": 1.169724464416504,
+      "learning_rate": 3e-05,
+      "loss": 2.9665,
+      "step": 2530
+    },
+    {
+      "epoch": 0.6877717391304348,
+      "grad_norm": 1.373205542564392,
+      "learning_rate": 3e-05,
+      "loss": 3.3161,
+      "step": 2531
+    },
+    {
+      "epoch": 0.6880434782608695,
+      "grad_norm": 1.0364346504211426,
+      "learning_rate": 3e-05,
+      "loss": 3.135,
+      "step": 2532
+    },
+    {
+      "epoch": 0.6883152173913043,
+      "grad_norm": 1.2910298109054565,
+      "learning_rate": 3e-05,
+      "loss": 3.2629,
+      "step": 2533
+    },
+    {
+      "epoch": 0.6885869565217392,
+      "grad_norm": 1.7945650815963745,
+      "learning_rate": 3e-05,
+      "loss": 3.5541,
+      "step": 2534
+    },
+    {
+      "epoch": 0.688858695652174,
+      "grad_norm": 1.3837659358978271,
+      "learning_rate": 3e-05,
+      "loss": 3.5515,
+      "step": 2535
+    },
+    {
+      "epoch": 0.6891304347826087,
+      "grad_norm": 1.3154077529907227,
+      "learning_rate": 3e-05,
+      "loss": 3.6571,
+      "step": 2536
+    },
+    {
+      "epoch": 0.6894021739130435,
+      "grad_norm": 1.3466911315917969,
+      "learning_rate": 3e-05,
+      "loss": 3.0786,
+      "step": 2537
+    },
+    {
+      "epoch": 0.6896739130434782,
+      "grad_norm": 1.3170411586761475,
+      "learning_rate": 3e-05,
+      "loss": 2.9075,
+      "step": 2538
+    },
+    {
+      "epoch": 0.689945652173913,
+      "grad_norm": 1.1219418048858643,
+      "learning_rate": 3e-05,
+      "loss": 3.1026,
+      "step": 2539
+    },
+    {
+      "epoch": 0.6902173913043478,
+      "grad_norm": 1.15764319896698,
+      "learning_rate": 3e-05,
+      "loss": 3.45,
+      "step": 2540
+    },
+    {
+      "epoch": 0.6904891304347827,
+      "grad_norm": 1.196807861328125,
+      "learning_rate": 3e-05,
+      "loss": 3.185,
+      "step": 2541
+    },
+    {
+      "epoch": 0.6907608695652174,
+      "grad_norm": 1.187324047088623,
+      "learning_rate": 3e-05,
+      "loss": 3.4707,
+      "step": 2542
+    },
+    {
+      "epoch": 0.6910326086956522,
+      "grad_norm": 1.0559121370315552,
+      "learning_rate": 3e-05,
+      "loss": 2.9871,
+      "step": 2543
+    },
+    {
+      "epoch": 0.691304347826087,
+      "grad_norm": 1.1243149042129517,
+      "learning_rate": 3e-05,
+      "loss": 3.5086,
+      "step": 2544
+    },
+    {
+      "epoch": 0.6915760869565217,
+      "grad_norm": 1.1120831966400146,
+      "learning_rate": 3e-05,
+      "loss": 3.1197,
+      "step": 2545
+    },
+    {
+      "epoch": 0.6918478260869565,
+      "grad_norm": 1.294692039489746,
+      "learning_rate": 3e-05,
+      "loss": 3.9801,
+      "step": 2546
+    },
+    {
+      "epoch": 0.6921195652173913,
+      "grad_norm": 1.3371599912643433,
+      "learning_rate": 3e-05,
+      "loss": 3.8999,
+      "step": 2547
+    },
+    {
+      "epoch": 0.6923913043478261,
+      "grad_norm": 1.219671368598938,
+      "learning_rate": 3e-05,
+      "loss": 3.3713,
+      "step": 2548
+    },
+    {
+      "epoch": 0.6926630434782609,
+      "grad_norm": 0.9593479037284851,
+      "learning_rate": 3e-05,
+      "loss": 2.8705,
+      "step": 2549
+    },
+    {
+      "epoch": 0.6929347826086957,
+      "grad_norm": 1.2361105680465698,
+      "learning_rate": 3e-05,
+      "loss": 3.6656,
+      "step": 2550
+    },
+    {
+      "epoch": 0.6932065217391304,
+      "grad_norm": 0.9579165577888489,
+      "learning_rate": 3e-05,
+      "loss": 2.9437,
+      "step": 2551
+    },
+    {
+      "epoch": 0.6934782608695652,
+      "grad_norm": 1.0255228281021118,
+      "learning_rate": 3e-05,
+      "loss": 3.0341,
+      "step": 2552
+    },
+    {
+      "epoch": 0.69375,
+      "grad_norm": 1.2201510667800903,
+      "learning_rate": 3e-05,
+      "loss": 3.6407,
+      "step": 2553
+    },
+    {
+      "epoch": 0.6940217391304347,
+      "grad_norm": 1.1329454183578491,
+      "learning_rate": 3e-05,
+      "loss": 3.4387,
+      "step": 2554
+    },
+    {
+      "epoch": 0.6942934782608695,
+      "grad_norm": 1.0753947496414185,
+      "learning_rate": 3e-05,
+      "loss": 3.1663,
+      "step": 2555
+    },
+    {
+      "epoch": 0.6945652173913044,
+      "grad_norm": 1.8046010732650757,
+      "learning_rate": 3e-05,
+      "loss": 3.624,
+      "step": 2556
+    },
+    {
+      "epoch": 0.6948369565217392,
+      "grad_norm": 1.0631359815597534,
+      "learning_rate": 3e-05,
+      "loss": 3.2537,
+      "step": 2557
+    },
+    {
+      "epoch": 0.6951086956521739,
+      "grad_norm": 1.0694079399108887,
+      "learning_rate": 3e-05,
+      "loss": 2.8613,
+      "step": 2558
+    },
+    {
+      "epoch": 0.6953804347826087,
+      "grad_norm": 1.2790333032608032,
+      "learning_rate": 3e-05,
+      "loss": 3.0732,
+      "step": 2559
+    },
+    {
+      "epoch": 0.6956521739130435,
+      "grad_norm": 1.253812551498413,
+      "learning_rate": 3e-05,
+      "loss": 3.0818,
+      "step": 2560
+    },
+    {
+      "epoch": 0.6959239130434782,
+      "grad_norm": 1.2076342105865479,
+      "learning_rate": 3e-05,
+      "loss": 2.7011,
+      "step": 2561
+    },
+    {
+      "epoch": 0.696195652173913,
+      "grad_norm": 1.2318416833877563,
+      "learning_rate": 3e-05,
+      "loss": 2.7658,
+      "step": 2562
+    },
+    {
+      "epoch": 0.6964673913043479,
+      "grad_norm": 1.442963719367981,
+      "learning_rate": 3e-05,
+      "loss": 3.7564,
+      "step": 2563
+    },
+    {
+      "epoch": 0.6967391304347826,
+      "grad_norm": 1.3947499990463257,
+      "learning_rate": 3e-05,
+      "loss": 3.4476,
+      "step": 2564
+    },
+    {
+      "epoch": 0.6970108695652174,
+      "grad_norm": 1.8033784627914429,
+      "learning_rate": 3e-05,
+      "loss": 3.7798,
+      "step": 2565
+    },
+    {
+      "epoch": 0.6972826086956522,
+      "grad_norm": 1.3678721189498901,
+      "learning_rate": 3e-05,
+      "loss": 3.3788,
+      "step": 2566
+    },
+    {
+      "epoch": 0.6975543478260869,
+      "grad_norm": 1.0561424493789673,
+      "learning_rate": 3e-05,
+      "loss": 3.1749,
+      "step": 2567
+    },
+    {
+      "epoch": 0.6978260869565217,
+      "grad_norm": 1.2744587659835815,
+      "learning_rate": 3e-05,
+      "loss": 3.3584,
+      "step": 2568
+    },
+    {
+      "epoch": 0.6980978260869565,
+      "grad_norm": 1.2167631387710571,
+      "learning_rate": 3e-05,
+      "loss": 3.4043,
+      "step": 2569
+    },
+    {
+      "epoch": 0.6983695652173914,
+      "grad_norm": 1.1461526155471802,
+      "learning_rate": 3e-05,
+      "loss": 3.381,
+      "step": 2570
+    },
+    {
+      "epoch": 0.6986413043478261,
+      "grad_norm": 1.173651933670044,
+      "learning_rate": 3e-05,
+      "loss": 2.8328,
+      "step": 2571
+    },
+    {
+      "epoch": 0.6989130434782609,
+      "grad_norm": 1.3456881046295166,
+      "learning_rate": 3e-05,
+      "loss": 3.9301,
+      "step": 2572
+    },
+    {
+      "epoch": 0.6991847826086957,
+      "grad_norm": 1.6022355556488037,
+      "learning_rate": 3e-05,
+      "loss": 4.0052,
+      "step": 2573
+    },
+    {
+      "epoch": 0.6994565217391304,
+      "grad_norm": 1.4133738279342651,
+      "learning_rate": 3e-05,
+      "loss": 3.6391,
+      "step": 2574
+    },
+    {
+      "epoch": 0.6997282608695652,
+      "grad_norm": 1.2311711311340332,
+      "learning_rate": 3e-05,
+      "loss": 3.6895,
+      "step": 2575
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0733444690704346,
+      "learning_rate": 3e-05,
+      "loss": 2.685,
+      "step": 2576
+    },
+    {
+      "epoch": 0.7002717391304348,
+      "grad_norm": 1.596375823020935,
+      "learning_rate": 3e-05,
+      "loss": 3.2422,
+      "step": 2577
+    },
+    {
+      "epoch": 0.7005434782608696,
+      "grad_norm": 1.098339319229126,
+      "learning_rate": 3e-05,
+      "loss": 3.0258,
+      "step": 2578
+    },
+    {
+      "epoch": 0.7008152173913044,
+      "grad_norm": 1.215551495552063,
+      "learning_rate": 3e-05,
+      "loss": 3.6583,
+      "step": 2579
+    },
+    {
+      "epoch": 0.7010869565217391,
+      "grad_norm": 1.175678014755249,
+      "learning_rate": 3e-05,
+      "loss": 2.9082,
+      "step": 2580
+    },
+    {
+      "epoch": 0.7013586956521739,
+      "grad_norm": 1.285050868988037,
+      "learning_rate": 3e-05,
+      "loss": 3.0373,
+      "step": 2581
+    },
+    {
+      "epoch": 0.7016304347826087,
+      "grad_norm": 1.0280994176864624,
+      "learning_rate": 3e-05,
+      "loss": 3.027,
+      "step": 2582
+    },
+    {
+      "epoch": 0.7019021739130434,
+      "grad_norm": 1.1611984968185425,
+      "learning_rate": 3e-05,
+      "loss": 2.9426,
+      "step": 2583
+    },
+    {
+      "epoch": 0.7021739130434783,
+      "grad_norm": 1.475341558456421,
+      "learning_rate": 3e-05,
+      "loss": 3.366,
+      "step": 2584
+    },
+    {
+      "epoch": 0.7024456521739131,
+      "grad_norm": 1.6377277374267578,
+      "learning_rate": 3e-05,
+      "loss": 3.4074,
+      "step": 2585
+    },
+    {
+      "epoch": 0.7027173913043478,
+      "grad_norm": 1.4452388286590576,
+      "learning_rate": 3e-05,
+      "loss": 2.9369,
+      "step": 2586
+    },
+    {
+      "epoch": 0.7029891304347826,
+      "grad_norm": 1.533266305923462,
+      "learning_rate": 3e-05,
+      "loss": 3.9352,
+      "step": 2587
+    },
+    {
+      "epoch": 0.7032608695652174,
+      "grad_norm": 1.6853458881378174,
+      "learning_rate": 3e-05,
+      "loss": 3.4523,
+      "step": 2588
+    },
+    {
+      "epoch": 0.7035326086956522,
+      "grad_norm": 1.3552666902542114,
+      "learning_rate": 3e-05,
+      "loss": 2.7867,
+      "step": 2589
+    },
+    {
+      "epoch": 0.7038043478260869,
+      "grad_norm": 1.5494905710220337,
+      "learning_rate": 3e-05,
+      "loss": 3.4699,
+      "step": 2590
+    },
+    {
+      "epoch": 0.7040760869565217,
+      "grad_norm": 1.74055016040802,
+      "learning_rate": 3e-05,
+      "loss": 3.3874,
+      "step": 2591
+    },
+    {
+      "epoch": 0.7043478260869566,
+      "grad_norm": 1.250354290008545,
+      "learning_rate": 3e-05,
+      "loss": 3.1795,
+      "step": 2592
+    },
+    {
+      "epoch": 0.7046195652173913,
+      "grad_norm": 1.302441954612732,
+      "learning_rate": 3e-05,
+      "loss": 3.9465,
+      "step": 2593
+    },
+    {
+      "epoch": 0.7048913043478261,
+      "grad_norm": 1.7971372604370117,
+      "learning_rate": 3e-05,
+      "loss": 3.6084,
+      "step": 2594
+    },
+    {
+      "epoch": 0.7051630434782609,
+      "grad_norm": 1.188605546951294,
+      "learning_rate": 3e-05,
+      "loss": 3.6716,
+      "step": 2595
+    },
+    {
+      "epoch": 0.7054347826086956,
+      "grad_norm": 0.9954262971878052,
+      "learning_rate": 3e-05,
+      "loss": 3.1982,
+      "step": 2596
+    },
+    {
+      "epoch": 0.7057065217391304,
+      "grad_norm": 1.174482822418213,
+      "learning_rate": 3e-05,
+      "loss": 3.1708,
+      "step": 2597
+    },
+    {
+      "epoch": 0.7059782608695652,
+      "grad_norm": 1.039506435394287,
+      "learning_rate": 3e-05,
+      "loss": 3.1857,
+      "step": 2598
+    },
+    {
+      "epoch": 0.70625,
+      "grad_norm": 0.9666784405708313,
+      "learning_rate": 3e-05,
+      "loss": 3.6005,
+      "step": 2599
+    },
+    {
+      "epoch": 0.7065217391304348,
+      "grad_norm": 0.921465277671814,
+      "learning_rate": 3e-05,
+      "loss": 3.0565,
+      "step": 2600
+    },
+    {
+      "epoch": 0.7067934782608696,
+      "grad_norm": 1.018864393234253,
+      "learning_rate": 3e-05,
+      "loss": 2.7895,
+      "step": 2601
+    },
+    {
+      "epoch": 0.7070652173913043,
+      "grad_norm": 1.0360243320465088,
+      "learning_rate": 3e-05,
+      "loss": 3.125,
+      "step": 2602
+    },
+    {
+      "epoch": 0.7073369565217391,
+      "grad_norm": 0.9393570423126221,
+      "learning_rate": 3e-05,
+      "loss": 3.1148,
+      "step": 2603
+    },
+    {
+      "epoch": 0.7076086956521739,
+      "grad_norm": 1.1420108079910278,
+      "learning_rate": 3e-05,
+      "loss": 3.1415,
+      "step": 2604
+    },
+    {
+      "epoch": 0.7078804347826086,
+      "grad_norm": 1.1641017198562622,
+      "learning_rate": 3e-05,
+      "loss": 3.0936,
+      "step": 2605
+    },
+    {
+      "epoch": 0.7081521739130435,
+      "grad_norm": 1.6107722520828247,
+      "learning_rate": 3e-05,
+      "loss": 4.061,
+      "step": 2606
+    },
+    {
+      "epoch": 0.7084239130434783,
+      "grad_norm": 1.2036430835723877,
+      "learning_rate": 3e-05,
+      "loss": 3.1156,
+      "step": 2607
+    },
+    {
+      "epoch": 0.7086956521739131,
+      "grad_norm": 1.2490811347961426,
+      "learning_rate": 3e-05,
+      "loss": 3.0078,
+      "step": 2608
+    },
+    {
+      "epoch": 0.7089673913043478,
+      "grad_norm": 1.0128928422927856,
+      "learning_rate": 3e-05,
+      "loss": 2.8499,
+      "step": 2609
+    },
+    {
+      "epoch": 0.7092391304347826,
+      "grad_norm": 1.363193154335022,
+      "learning_rate": 3e-05,
+      "loss": 3.9375,
+      "step": 2610
+    },
+    {
+      "epoch": 0.7095108695652174,
+      "grad_norm": 0.8343818187713623,
+      "learning_rate": 3e-05,
+      "loss": 3.1519,
+      "step": 2611
+    },
+    {
+      "epoch": 0.7097826086956521,
+      "grad_norm": 1.0047922134399414,
+      "learning_rate": 3e-05,
+      "loss": 2.8608,
+      "step": 2612
+    },
+    {
+      "epoch": 0.710054347826087,
+      "grad_norm": 1.18521249294281,
+      "learning_rate": 3e-05,
+      "loss": 3.7399,
+      "step": 2613
+    },
+    {
+      "epoch": 0.7103260869565218,
+      "grad_norm": 1.3591742515563965,
+      "learning_rate": 3e-05,
+      "loss": 3.5885,
+      "step": 2614
+    },
+    {
+      "epoch": 0.7105978260869565,
+      "grad_norm": 1.1936426162719727,
+      "learning_rate": 3e-05,
+      "loss": 3.7192,
+      "step": 2615
+    },
+    {
+      "epoch": 0.7108695652173913,
+      "grad_norm": 1.114531397819519,
+      "learning_rate": 3e-05,
+      "loss": 3.22,
+      "step": 2616
+    },
+    {
+      "epoch": 0.7111413043478261,
+      "grad_norm": 1.1929512023925781,
+      "learning_rate": 3e-05,
+      "loss": 2.751,
+      "step": 2617
+    },
+    {
+      "epoch": 0.7114130434782608,
+      "grad_norm": 1.3775954246520996,
+      "learning_rate": 3e-05,
+      "loss": 3.6234,
+      "step": 2618
+    },
+    {
+      "epoch": 0.7116847826086956,
+      "grad_norm": 1.313889980316162,
+      "learning_rate": 3e-05,
+      "loss": 3.2312,
+      "step": 2619
+    },
+    {
+      "epoch": 0.7119565217391305,
+      "grad_norm": 1.4679901599884033,
+      "learning_rate": 3e-05,
+      "loss": 3.2006,
+      "step": 2620
+    },
+    {
+      "epoch": 0.7122282608695653,
+      "grad_norm": 0.9835330843925476,
+      "learning_rate": 3e-05,
+      "loss": 3.2365,
+      "step": 2621
+    },
+    {
+      "epoch": 0.7125,
+      "grad_norm": 1.9934345483779907,
+      "learning_rate": 3e-05,
+      "loss": 4.4249,
+      "step": 2622
+    },
+    {
+      "epoch": 0.7127717391304348,
+      "grad_norm": 1.2943999767303467,
+      "learning_rate": 3e-05,
+      "loss": 3.517,
+      "step": 2623
+    },
+    {
+      "epoch": 0.7130434782608696,
+      "grad_norm": 1.1526103019714355,
+      "learning_rate": 3e-05,
+      "loss": 3.6404,
+      "step": 2624
+    },
+    {
+      "epoch": 0.7133152173913043,
+      "grad_norm": 1.3137702941894531,
+      "learning_rate": 3e-05,
+      "loss": 3.305,
+      "step": 2625
+    },
+    {
+      "epoch": 0.7135869565217391,
+      "grad_norm": 1.2102550268173218,
+      "learning_rate": 3e-05,
+      "loss": 3.0545,
+      "step": 2626
+    },
+    {
+      "epoch": 0.7138586956521739,
+      "grad_norm": 1.0959724187850952,
+      "learning_rate": 3e-05,
+      "loss": 3.0383,
+      "step": 2627
+    },
+    {
+      "epoch": 0.7141304347826087,
+      "grad_norm": 1.7450772523880005,
+      "learning_rate": 3e-05,
+      "loss": 3.5891,
+      "step": 2628
+    },
+    {
+      "epoch": 0.7144021739130435,
+      "grad_norm": 1.3901242017745972,
+      "learning_rate": 3e-05,
+      "loss": 3.5405,
+      "step": 2629
+    },
+    {
+      "epoch": 0.7146739130434783,
+      "grad_norm": 1.2243647575378418,
+      "learning_rate": 3e-05,
+      "loss": 3.4307,
+      "step": 2630
+    },
+    {
+      "epoch": 0.714945652173913,
+      "grad_norm": 2.3953914642333984,
+      "learning_rate": 3e-05,
+      "loss": 4.3341,
+      "step": 2631
+    },
+    {
+      "epoch": 0.7152173913043478,
+      "grad_norm": 1.8682715892791748,
+      "learning_rate": 3e-05,
+      "loss": 3.0704,
+      "step": 2632
+    },
+    {
+      "epoch": 0.7154891304347826,
+      "grad_norm": 1.4622515439987183,
+      "learning_rate": 3e-05,
+      "loss": 3.3376,
+      "step": 2633
+    },
+    {
+      "epoch": 0.7157608695652173,
+      "grad_norm": 1.2892646789550781,
+      "learning_rate": 3e-05,
+      "loss": 3.312,
+      "step": 2634
+    },
+    {
+      "epoch": 0.7160326086956522,
+      "grad_norm": 1.4950429201126099,
+      "learning_rate": 3e-05,
+      "loss": 3.3923,
+      "step": 2635
+    },
+    {
+      "epoch": 0.716304347826087,
+      "grad_norm": 1.380149245262146,
+      "learning_rate": 3e-05,
+      "loss": 3.5931,
+      "step": 2636
+    },
+    {
+      "epoch": 0.7165760869565218,
+      "grad_norm": 1.433959722518921,
+      "learning_rate": 3e-05,
+      "loss": 2.9042,
+      "step": 2637
+    },
+    {
+      "epoch": 0.7168478260869565,
+      "grad_norm": 1.2683796882629395,
+      "learning_rate": 3e-05,
+      "loss": 3.3012,
+      "step": 2638
+    },
+    {
+      "epoch": 0.7171195652173913,
+      "grad_norm": 1.0387176275253296,
+      "learning_rate": 3e-05,
+      "loss": 2.9117,
+      "step": 2639
+    },
+    {
+      "epoch": 0.717391304347826,
+      "grad_norm": 1.5096675157546997,
+      "learning_rate": 3e-05,
+      "loss": 3.125,
+      "step": 2640
+    },
+    {
+      "epoch": 0.7176630434782608,
+      "grad_norm": 1.1815725564956665,
+      "learning_rate": 3e-05,
+      "loss": 2.9721,
+      "step": 2641
+    },
+    {
+      "epoch": 0.7179347826086957,
+      "grad_norm": 1.1765364408493042,
+      "learning_rate": 3e-05,
+      "loss": 3.3088,
+      "step": 2642
+    },
+    {
+      "epoch": 0.7182065217391305,
+      "grad_norm": 1.087518572807312,
+      "learning_rate": 3e-05,
+      "loss": 3.1549,
+      "step": 2643
+    },
+    {
+      "epoch": 0.7184782608695652,
+      "grad_norm": 1.1884835958480835,
+      "learning_rate": 3e-05,
+      "loss": 3.2079,
+      "step": 2644
+    },
+    {
+      "epoch": 0.71875,
+      "grad_norm": 1.1198290586471558,
+      "learning_rate": 3e-05,
+      "loss": 2.9726,
+      "step": 2645
+    },
+    {
+      "epoch": 0.7190217391304348,
+      "grad_norm": 1.2286672592163086,
+      "learning_rate": 3e-05,
+      "loss": 3.2783,
+      "step": 2646
+    },
+    {
+      "epoch": 0.7192934782608695,
+      "grad_norm": 1.3148912191390991,
+      "learning_rate": 3e-05,
+      "loss": 3.914,
+      "step": 2647
+    },
+    {
+      "epoch": 0.7195652173913043,
+      "grad_norm": 1.1227190494537354,
+      "learning_rate": 3e-05,
+      "loss": 3.1269,
+      "step": 2648
+    },
+    {
+      "epoch": 0.7198369565217392,
+      "grad_norm": 1.3971885442733765,
+      "learning_rate": 3e-05,
+      "loss": 3.3385,
+      "step": 2649
+    },
+    {
+      "epoch": 0.720108695652174,
+      "grad_norm": 1.0572166442871094,
+      "learning_rate": 3e-05,
+      "loss": 2.8119,
+      "step": 2650
+    },
+    {
+      "epoch": 0.7203804347826087,
+      "grad_norm": 1.3299329280853271,
+      "learning_rate": 3e-05,
+      "loss": 3.5842,
+      "step": 2651
+    },
+    {
+      "epoch": 0.7206521739130435,
+      "grad_norm": 1.2632654905319214,
+      "learning_rate": 3e-05,
+      "loss": 2.8276,
+      "step": 2652
+    },
+    {
+      "epoch": 0.7209239130434782,
+      "grad_norm": 1.2086924314498901,
+      "learning_rate": 3e-05,
+      "loss": 3.3622,
+      "step": 2653
+    },
+    {
+      "epoch": 0.721195652173913,
+      "grad_norm": 1.1867645978927612,
+      "learning_rate": 3e-05,
+      "loss": 3.1258,
+      "step": 2654
+    },
+    {
+      "epoch": 0.7214673913043478,
+      "grad_norm": 1.123406171798706,
+      "learning_rate": 3e-05,
+      "loss": 3.3561,
+      "step": 2655
+    },
+    {
+      "epoch": 0.7217391304347827,
+      "grad_norm": 1.3678390979766846,
+      "learning_rate": 3e-05,
+      "loss": 3.4183,
+      "step": 2656
+    },
+    {
+      "epoch": 0.7220108695652174,
+      "grad_norm": 1.21256685256958,
+      "learning_rate": 3e-05,
+      "loss": 3.5207,
+      "step": 2657
+    },
+    {
+      "epoch": 0.7222826086956522,
+      "grad_norm": 1.1133040189743042,
+      "learning_rate": 3e-05,
+      "loss": 2.8218,
+      "step": 2658
+    },
+    {
+      "epoch": 0.722554347826087,
+      "grad_norm": 1.0846513509750366,
+      "learning_rate": 3e-05,
+      "loss": 2.9519,
+      "step": 2659
+    },
+    {
+      "epoch": 0.7228260869565217,
+      "grad_norm": 1.0870620012283325,
+      "learning_rate": 3e-05,
+      "loss": 2.8142,
+      "step": 2660
+    },
+    {
+      "epoch": 0.7230978260869565,
+      "grad_norm": 1.2407505512237549,
+      "learning_rate": 3e-05,
+      "loss": 3.0965,
+      "step": 2661
+    },
+    {
+      "epoch": 0.7233695652173913,
+      "grad_norm": 1.0677931308746338,
+      "learning_rate": 3e-05,
+      "loss": 2.841,
+      "step": 2662
+    },
+    {
+      "epoch": 0.7236413043478261,
+      "grad_norm": 1.1866375207901,
+      "learning_rate": 3e-05,
+      "loss": 3.2199,
+      "step": 2663
+    },
+    {
+      "epoch": 0.7239130434782609,
+      "grad_norm": 1.4401320219039917,
+      "learning_rate": 3e-05,
+      "loss": 4.1391,
+      "step": 2664
+    },
+    {
+      "epoch": 0.7241847826086957,
+      "grad_norm": 1.8525235652923584,
+      "learning_rate": 3e-05,
+      "loss": 2.954,
+      "step": 2665
+    },
+    {
+      "epoch": 0.7244565217391304,
+      "grad_norm": 1.4307048320770264,
+      "learning_rate": 3e-05,
+      "loss": 3.6121,
+      "step": 2666
+    },
+    {
+      "epoch": 0.7247282608695652,
+      "grad_norm": 1.4132686853408813,
+      "learning_rate": 3e-05,
+      "loss": 3.2657,
+      "step": 2667
+    },
+    {
+      "epoch": 0.725,
+      "grad_norm": 1.1710587739944458,
+      "learning_rate": 3e-05,
+      "loss": 3.5075,
+      "step": 2668
+    },
+    {
+      "epoch": 0.7252717391304347,
+      "grad_norm": 1.4402532577514648,
+      "learning_rate": 3e-05,
+      "loss": 3.5403,
+      "step": 2669
+    },
+    {
+      "epoch": 0.7255434782608695,
+      "grad_norm": 1.0295389890670776,
+      "learning_rate": 3e-05,
+      "loss": 3.2313,
+      "step": 2670
+    },
+    {
+      "epoch": 0.7258152173913044,
+      "grad_norm": 1.9836853742599487,
+      "learning_rate": 3e-05,
+      "loss": 3.7605,
+      "step": 2671
+    },
+    {
+      "epoch": 0.7260869565217392,
+      "grad_norm": 1.313320279121399,
+      "learning_rate": 3e-05,
+      "loss": 2.892,
+      "step": 2672
+    },
+    {
+      "epoch": 0.7263586956521739,
+      "grad_norm": 1.227516770362854,
+      "learning_rate": 3e-05,
+      "loss": 3.4072,
+      "step": 2673
+    },
+    {
+      "epoch": 0.7266304347826087,
+      "grad_norm": 1.343525767326355,
+      "learning_rate": 3e-05,
+      "loss": 3.1093,
+      "step": 2674
+    },
+    {
+      "epoch": 0.7269021739130435,
+      "grad_norm": 1.6472978591918945,
+      "learning_rate": 3e-05,
+      "loss": 3.2588,
+      "step": 2675
+    },
+    {
+      "epoch": 0.7271739130434782,
+      "grad_norm": 1.313114047050476,
+      "learning_rate": 3e-05,
+      "loss": 3.0819,
+      "step": 2676
+    },
+    {
+      "epoch": 0.727445652173913,
+      "grad_norm": 1.3546169996261597,
+      "learning_rate": 3e-05,
+      "loss": 3.3884,
+      "step": 2677
+    },
+    {
+      "epoch": 0.7277173913043479,
+      "grad_norm": 1.4308792352676392,
+      "learning_rate": 3e-05,
+      "loss": 3.2228,
+      "step": 2678
+    },
+    {
+      "epoch": 0.7279891304347826,
+      "grad_norm": 1.2303006649017334,
+      "learning_rate": 3e-05,
+      "loss": 2.9464,
+      "step": 2679
+    },
+    {
+      "epoch": 0.7282608695652174,
+      "grad_norm": 2.12964129447937,
+      "learning_rate": 3e-05,
+      "loss": 4.1515,
+      "step": 2680
+    },
+    {
+      "epoch": 0.7285326086956522,
+      "grad_norm": 1.1187695264816284,
+      "learning_rate": 3e-05,
+      "loss": 3.2245,
+      "step": 2681
+    },
+    {
+      "epoch": 0.7288043478260869,
+      "grad_norm": 1.440588355064392,
+      "learning_rate": 3e-05,
+      "loss": 4.0875,
+      "step": 2682
+    },
+    {
+      "epoch": 0.7290760869565217,
+      "grad_norm": 1.2928602695465088,
+      "learning_rate": 3e-05,
+      "loss": 3.4652,
+      "step": 2683
+    },
+    {
+      "epoch": 0.7293478260869565,
+      "grad_norm": 1.1873844861984253,
+      "learning_rate": 3e-05,
+      "loss": 2.8352,
+      "step": 2684
+    },
+    {
+      "epoch": 0.7296195652173914,
+      "grad_norm": 1.1879850625991821,
+      "learning_rate": 3e-05,
+      "loss": 2.8956,
+      "step": 2685
+    },
+    {
+      "epoch": 0.7298913043478261,
+      "grad_norm": 1.206374168395996,
+      "learning_rate": 3e-05,
+      "loss": 3.3039,
+      "step": 2686
+    },
+    {
+      "epoch": 0.7301630434782609,
+      "grad_norm": 1.1491773128509521,
+      "learning_rate": 3e-05,
+      "loss": 3.0314,
+      "step": 2687
+    },
+    {
+      "epoch": 0.7304347826086957,
+      "grad_norm": 1.3866368532180786,
+      "learning_rate": 3e-05,
+      "loss": 3.055,
+      "step": 2688
+    },
+    {
+      "epoch": 0.7307065217391304,
+      "grad_norm": 1.5674668550491333,
+      "learning_rate": 3e-05,
+      "loss": 3.1808,
+      "step": 2689
+    },
+    {
+      "epoch": 0.7309782608695652,
+      "grad_norm": 1.562813639640808,
+      "learning_rate": 3e-05,
+      "loss": 3.2706,
+      "step": 2690
+    },
+    {
+      "epoch": 0.73125,
+      "grad_norm": 1.6883511543273926,
+      "learning_rate": 3e-05,
+      "loss": 3.2192,
+      "step": 2691
+    },
+    {
+      "epoch": 0.7315217391304348,
+      "grad_norm": 1.1254132986068726,
+      "learning_rate": 3e-05,
+      "loss": 3.1529,
+      "step": 2692
+    },
+    {
+      "epoch": 0.7317934782608696,
+      "grad_norm": 1.0081984996795654,
+      "learning_rate": 3e-05,
+      "loss": 3.0864,
+      "step": 2693
+    },
+    {
+      "epoch": 0.7320652173913044,
+      "grad_norm": 1.453481674194336,
+      "learning_rate": 3e-05,
+      "loss": 3.2169,
+      "step": 2694
+    },
+    {
+      "epoch": 0.7323369565217391,
+      "grad_norm": 1.3893227577209473,
+      "learning_rate": 3e-05,
+      "loss": 3.3761,
+      "step": 2695
+    },
+    {
+      "epoch": 0.7326086956521739,
+      "grad_norm": 1.7900173664093018,
+      "learning_rate": 3e-05,
+      "loss": 3.4638,
+      "step": 2696
+    },
+    {
+      "epoch": 0.7328804347826087,
+      "grad_norm": 1.1031737327575684,
+      "learning_rate": 3e-05,
+      "loss": 2.6962,
+      "step": 2697
+    },
+    {
+      "epoch": 0.7331521739130434,
+      "grad_norm": 1.3984167575836182,
+      "learning_rate": 3e-05,
+      "loss": 3.3052,
+      "step": 2698
+    },
+    {
+      "epoch": 0.7334239130434783,
+      "grad_norm": 1.6182535886764526,
+      "learning_rate": 3e-05,
+      "loss": 3.4213,
+      "step": 2699
+    },
+    {
+      "epoch": 0.7336956521739131,
+      "grad_norm": 1.7064732313156128,
+      "learning_rate": 3e-05,
+      "loss": 4.0381,
+      "step": 2700
+    },
+    {
+      "epoch": 0.7339673913043478,
+      "grad_norm": 1.357704520225525,
+      "learning_rate": 3e-05,
+      "loss": 3.6925,
+      "step": 2701
+    },
+    {
+      "epoch": 0.7342391304347826,
+      "grad_norm": 1.6365474462509155,
+      "learning_rate": 3e-05,
+      "loss": 3.2891,
+      "step": 2702
+    },
+    {
+      "epoch": 0.7345108695652174,
+      "grad_norm": 1.8574589490890503,
+      "learning_rate": 3e-05,
+      "loss": 3.3637,
+      "step": 2703
+    },
+    {
+      "epoch": 0.7347826086956522,
+      "grad_norm": 1.3467528820037842,
+      "learning_rate": 3e-05,
+      "loss": 3.1417,
+      "step": 2704
+    },
+    {
+      "epoch": 0.7350543478260869,
+      "grad_norm": 1.4658093452453613,
+      "learning_rate": 3e-05,
+      "loss": 3.3401,
+      "step": 2705
+    },
+    {
+      "epoch": 0.7353260869565217,
+      "grad_norm": 1.2771729230880737,
+      "learning_rate": 3e-05,
+      "loss": 3.3731,
+      "step": 2706
+    },
+    {
+      "epoch": 0.7355978260869566,
+      "grad_norm": 1.0597422122955322,
+      "learning_rate": 3e-05,
+      "loss": 3.1315,
+      "step": 2707
+    },
+    {
+      "epoch": 0.7358695652173913,
+      "grad_norm": 1.1597671508789062,
+      "learning_rate": 3e-05,
+      "loss": 2.6791,
+      "step": 2708
+    },
+    {
+      "epoch": 0.7361413043478261,
+      "grad_norm": 1.7157272100448608,
+      "learning_rate": 3e-05,
+      "loss": 3.8816,
+      "step": 2709
+    },
+    {
+      "epoch": 0.7364130434782609,
+      "grad_norm": 1.3123220205307007,
+      "learning_rate": 3e-05,
+      "loss": 3.6255,
+      "step": 2710
+    },
+    {
+      "epoch": 0.7366847826086956,
+      "grad_norm": 1.2993775606155396,
+      "learning_rate": 3e-05,
+      "loss": 3.6012,
+      "step": 2711
+    },
+    {
+      "epoch": 0.7369565217391304,
+      "grad_norm": 1.0929709672927856,
+      "learning_rate": 3e-05,
+      "loss": 3.3147,
+      "step": 2712
+    },
+    {
+      "epoch": 0.7372282608695652,
+      "grad_norm": 1.2103468179702759,
+      "learning_rate": 3e-05,
+      "loss": 3.3148,
+      "step": 2713
+    },
+    {
+      "epoch": 0.7375,
+      "grad_norm": 0.9965810179710388,
+      "learning_rate": 3e-05,
+      "loss": 3.2745,
+      "step": 2714
+    },
+    {
+      "epoch": 0.7377717391304348,
+      "grad_norm": 1.7804696559906006,
+      "learning_rate": 3e-05,
+      "loss": 3.6753,
+      "step": 2715
+    },
+    {
+      "epoch": 0.7380434782608696,
+      "grad_norm": 1.2880659103393555,
+      "learning_rate": 3e-05,
+      "loss": 3.4252,
+      "step": 2716
+    },
+    {
+      "epoch": 0.7383152173913043,
+      "grad_norm": 0.9781152606010437,
+      "learning_rate": 3e-05,
+      "loss": 2.9325,
+      "step": 2717
+    },
+    {
+      "epoch": 0.7385869565217391,
+      "grad_norm": 1.724921464920044,
+      "learning_rate": 3e-05,
+      "loss": 3.6383,
+      "step": 2718
+    },
+    {
+      "epoch": 0.7388586956521739,
+      "grad_norm": 1.3509818315505981,
+      "learning_rate": 3e-05,
+      "loss": 3.2334,
+      "step": 2719
+    },
+    {
+      "epoch": 0.7391304347826086,
+      "grad_norm": 1.2037723064422607,
+      "learning_rate": 3e-05,
+      "loss": 3.242,
+      "step": 2720
+    },
+    {
+      "epoch": 0.7394021739130435,
+      "grad_norm": 1.337058424949646,
+      "learning_rate": 3e-05,
+      "loss": 3.5805,
+      "step": 2721
+    },
+    {
+      "epoch": 0.7396739130434783,
+      "grad_norm": 1.3748770952224731,
+      "learning_rate": 3e-05,
+      "loss": 2.9626,
+      "step": 2722
+    },
+    {
+      "epoch": 0.7399456521739131,
+      "grad_norm": 1.3539892435073853,
+      "learning_rate": 3e-05,
+      "loss": 3.0298,
+      "step": 2723
+    },
+    {
+      "epoch": 0.7402173913043478,
+      "grad_norm": 0.979083776473999,
+      "learning_rate": 3e-05,
+      "loss": 2.7295,
+      "step": 2724
+    },
+    {
+      "epoch": 0.7404891304347826,
+      "grad_norm": 1.7237200736999512,
+      "learning_rate": 3e-05,
+      "loss": 3.5329,
+      "step": 2725
+    },
+    {
+      "epoch": 0.7407608695652174,
+      "grad_norm": 1.4173531532287598,
+      "learning_rate": 3e-05,
+      "loss": 3.5202,
+      "step": 2726
+    },
+    {
+      "epoch": 0.7410326086956521,
+      "grad_norm": 1.3420840501785278,
+      "learning_rate": 3e-05,
+      "loss": 3.3239,
+      "step": 2727
+    },
+    {
+      "epoch": 0.741304347826087,
+      "grad_norm": 1.3799097537994385,
+      "learning_rate": 3e-05,
+      "loss": 3.4807,
+      "step": 2728
+    },
+    {
+      "epoch": 0.7415760869565218,
+      "grad_norm": 1.2215850353240967,
+      "learning_rate": 3e-05,
+      "loss": 3.5144,
+      "step": 2729
+    },
+    {
+      "epoch": 0.7418478260869565,
+      "grad_norm": 1.2649718523025513,
+      "learning_rate": 3e-05,
+      "loss": 3.4898,
+      "step": 2730
+    },
+    {
+      "epoch": 0.7421195652173913,
+      "grad_norm": 1.1176096200942993,
+      "learning_rate": 3e-05,
+      "loss": 3.4977,
+      "step": 2731
+    },
+    {
+      "epoch": 0.7423913043478261,
+      "grad_norm": 1.0005884170532227,
+      "learning_rate": 3e-05,
+      "loss": 3.1848,
+      "step": 2732
+    },
+    {
+      "epoch": 0.7426630434782608,
+      "grad_norm": 1.6306564807891846,
+      "learning_rate": 3e-05,
+      "loss": 3.8107,
+      "step": 2733
+    },
+    {
+      "epoch": 0.7429347826086956,
+      "grad_norm": 1.6788973808288574,
+      "learning_rate": 3e-05,
+      "loss": 3.985,
+      "step": 2734
+    },
+    {
+      "epoch": 0.7432065217391305,
+      "grad_norm": 1.16837739944458,
+      "learning_rate": 3e-05,
+      "loss": 2.8648,
+      "step": 2735
+    },
+    {
+      "epoch": 0.7434782608695653,
+      "grad_norm": 1.324316382408142,
+      "learning_rate": 3e-05,
+      "loss": 3.5445,
+      "step": 2736
+    },
+    {
+      "epoch": 0.74375,
+      "grad_norm": 1.2631101608276367,
+      "learning_rate": 3e-05,
+      "loss": 3.2311,
+      "step": 2737
+    },
+    {
+      "epoch": 0.7440217391304348,
+      "grad_norm": 1.030339241027832,
+      "learning_rate": 3e-05,
+      "loss": 2.8981,
+      "step": 2738
+    },
+    {
+      "epoch": 0.7442934782608696,
+      "grad_norm": 1.5260175466537476,
+      "learning_rate": 3e-05,
+      "loss": 3.2335,
+      "step": 2739
+    },
+    {
+      "epoch": 0.7445652173913043,
+      "grad_norm": 1.610879898071289,
+      "learning_rate": 3e-05,
+      "loss": 3.3666,
+      "step": 2740
+    },
+    {
+      "epoch": 0.7448369565217391,
+      "grad_norm": 1.290809988975525,
+      "learning_rate": 3e-05,
+      "loss": 2.9172,
+      "step": 2741
+    },
+    {
+      "epoch": 0.7451086956521739,
+      "grad_norm": 1.2643457651138306,
+      "learning_rate": 3e-05,
+      "loss": 3.1168,
+      "step": 2742
+    },
+    {
+      "epoch": 0.7453804347826087,
+      "grad_norm": 1.2268829345703125,
+      "learning_rate": 3e-05,
+      "loss": 3.2188,
+      "step": 2743
+    },
+    {
+      "epoch": 0.7456521739130435,
+      "grad_norm": 1.022884726524353,
+      "learning_rate": 3e-05,
+      "loss": 2.884,
+      "step": 2744
+    },
+    {
+      "epoch": 0.7459239130434783,
+      "grad_norm": 1.9909472465515137,
+      "learning_rate": 3e-05,
+      "loss": 4.1557,
+      "step": 2745
+    },
+    {
+      "epoch": 0.746195652173913,
+      "grad_norm": 2.2413318157196045,
+      "learning_rate": 3e-05,
+      "loss": 4.0246,
+      "step": 2746
+    },
+    {
+      "epoch": 0.7464673913043478,
+      "grad_norm": 1.0311791896820068,
+      "learning_rate": 3e-05,
+      "loss": 2.8403,
+      "step": 2747
+    },
+    {
+      "epoch": 0.7467391304347826,
+      "grad_norm": 1.0483710765838623,
+      "learning_rate": 3e-05,
+      "loss": 2.7621,
+      "step": 2748
+    },
+    {
+      "epoch": 0.7470108695652173,
+      "grad_norm": 0.9712722897529602,
+      "learning_rate": 3e-05,
+      "loss": 2.7713,
+      "step": 2749
+    },
+    {
+      "epoch": 0.7472826086956522,
+      "grad_norm": 1.5642272233963013,
+      "learning_rate": 3e-05,
+      "loss": 3.2916,
+      "step": 2750
+    },
+    {
+      "epoch": 0.747554347826087,
+      "grad_norm": 1.0362199544906616,
+      "learning_rate": 3e-05,
+      "loss": 3.0884,
+      "step": 2751
+    },
+    {
+      "epoch": 0.7478260869565218,
+      "grad_norm": 1.2575013637542725,
+      "learning_rate": 3e-05,
+      "loss": 3.3603,
+      "step": 2752
+    },
+    {
+      "epoch": 0.7480978260869565,
+      "grad_norm": 1.2161589860916138,
+      "learning_rate": 3e-05,
+      "loss": 3.3827,
+      "step": 2753
+    },
+    {
+      "epoch": 0.7483695652173913,
+      "grad_norm": 1.0151880979537964,
+      "learning_rate": 3e-05,
+      "loss": 2.9711,
+      "step": 2754
+    },
+    {
+      "epoch": 0.748641304347826,
+      "grad_norm": 1.3773422241210938,
+      "learning_rate": 3e-05,
+      "loss": 3.9046,
+      "step": 2755
+    },
+    {
+      "epoch": 0.7489130434782608,
+      "grad_norm": 0.8108665943145752,
+      "learning_rate": 3e-05,
+      "loss": 2.9052,
+      "step": 2756
+    },
+    {
+      "epoch": 0.7491847826086957,
+      "grad_norm": 0.8243381381034851,
+      "learning_rate": 3e-05,
+      "loss": 2.8174,
+      "step": 2757
+    },
+    {
+      "epoch": 0.7494565217391305,
+      "grad_norm": 1.3421036005020142,
+      "learning_rate": 3e-05,
+      "loss": 3.3869,
+      "step": 2758
+    },
+    {
+      "epoch": 0.7497282608695652,
+      "grad_norm": 1.1268045902252197,
+      "learning_rate": 3e-05,
+      "loss": 3.4369,
+      "step": 2759
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.982552707195282,
+      "learning_rate": 3e-05,
+      "loss": 3.0077,
+      "step": 2760
+    },
+    {
+      "epoch": 0.7502717391304348,
+      "grad_norm": 1.548680305480957,
+      "learning_rate": 3e-05,
+      "loss": 2.9879,
+      "step": 2761
+    },
+    {
+      "epoch": 0.7505434782608695,
+      "grad_norm": 1.1556179523468018,
+      "learning_rate": 3e-05,
+      "loss": 2.9654,
+      "step": 2762
+    },
+    {
+      "epoch": 0.7508152173913043,
+      "grad_norm": 1.310678482055664,
+      "learning_rate": 3e-05,
+      "loss": 3.6287,
+      "step": 2763
+    },
+    {
+      "epoch": 0.7510869565217392,
+      "grad_norm": 1.4762641191482544,
+      "learning_rate": 3e-05,
+      "loss": 3.6961,
+      "step": 2764
+    },
+    {
+      "epoch": 0.751358695652174,
+      "grad_norm": 1.3265812397003174,
+      "learning_rate": 3e-05,
+      "loss": 3.2219,
+      "step": 2765
+    },
+    {
+      "epoch": 0.7516304347826087,
+      "grad_norm": 1.2346985340118408,
+      "learning_rate": 3e-05,
+      "loss": 3.259,
+      "step": 2766
+    },
+    {
+      "epoch": 0.7519021739130435,
+      "grad_norm": 0.9740216135978699,
+      "learning_rate": 3e-05,
+      "loss": 2.854,
+      "step": 2767
+    },
+    {
+      "epoch": 0.7521739130434782,
+      "grad_norm": 1.2616466283798218,
+      "learning_rate": 3e-05,
+      "loss": 2.9113,
+      "step": 2768
+    },
+    {
+      "epoch": 0.752445652173913,
+      "grad_norm": 1.4515007734298706,
+      "learning_rate": 3e-05,
+      "loss": 4.4309,
+      "step": 2769
+    },
+    {
+      "epoch": 0.7527173913043478,
+      "grad_norm": 1.2112640142440796,
+      "learning_rate": 3e-05,
+      "loss": 3.1847,
+      "step": 2770
+    },
+    {
+      "epoch": 0.7529891304347827,
+      "grad_norm": 1.123841643333435,
+      "learning_rate": 3e-05,
+      "loss": 3.3345,
+      "step": 2771
+    },
+    {
+      "epoch": 0.7532608695652174,
+      "grad_norm": 1.1541368961334229,
+      "learning_rate": 3e-05,
+      "loss": 2.8234,
+      "step": 2772
+    },
+    {
+      "epoch": 0.7535326086956522,
+      "grad_norm": 1.0209699869155884,
+      "learning_rate": 3e-05,
+      "loss": 3.1726,
+      "step": 2773
+    },
+    {
+      "epoch": 0.753804347826087,
+      "grad_norm": 1.2185944318771362,
+      "learning_rate": 3e-05,
+      "loss": 3.8311,
+      "step": 2774
+    },
+    {
+      "epoch": 0.7540760869565217,
+      "grad_norm": 1.553025722503662,
+      "learning_rate": 3e-05,
+      "loss": 3.2896,
+      "step": 2775
+    },
+    {
+      "epoch": 0.7543478260869565,
+      "grad_norm": 1.5103728771209717,
+      "learning_rate": 3e-05,
+      "loss": 3.0133,
+      "step": 2776
+    },
+    {
+      "epoch": 0.7546195652173913,
+      "grad_norm": 1.0733397006988525,
+      "learning_rate": 3e-05,
+      "loss": 3.0122,
+      "step": 2777
+    },
+    {
+      "epoch": 0.7548913043478261,
+      "grad_norm": 1.4341579675674438,
+      "learning_rate": 3e-05,
+      "loss": 3.5646,
+      "step": 2778
+    },
+    {
+      "epoch": 0.7551630434782609,
+      "grad_norm": 1.6829897165298462,
+      "learning_rate": 3e-05,
+      "loss": 3.6338,
+      "step": 2779
+    },
+    {
+      "epoch": 0.7554347826086957,
+      "grad_norm": 1.0816172361373901,
+      "learning_rate": 3e-05,
+      "loss": 3.1679,
+      "step": 2780
+    },
+    {
+      "epoch": 0.7557065217391304,
+      "grad_norm": 1.3089653253555298,
+      "learning_rate": 3e-05,
+      "loss": 3.6569,
+      "step": 2781
+    },
+    {
+      "epoch": 0.7559782608695652,
+      "grad_norm": 1.329774022102356,
+      "learning_rate": 3e-05,
+      "loss": 3.0589,
+      "step": 2782
+    },
+    {
+      "epoch": 0.75625,
+      "grad_norm": 1.5232934951782227,
+      "learning_rate": 3e-05,
+      "loss": 3.5649,
+      "step": 2783
+    },
+    {
+      "epoch": 0.7565217391304347,
+      "grad_norm": 1.501591444015503,
+      "learning_rate": 3e-05,
+      "loss": 3.7119,
+      "step": 2784
+    },
+    {
+      "epoch": 0.7567934782608695,
+      "grad_norm": 0.89922696352005,
+      "learning_rate": 3e-05,
+      "loss": 2.9755,
+      "step": 2785
+    },
+    {
+      "epoch": 0.7570652173913044,
+      "grad_norm": 1.0538735389709473,
+      "learning_rate": 3e-05,
+      "loss": 3.4823,
+      "step": 2786
+    },
+    {
+      "epoch": 0.7573369565217392,
+      "grad_norm": 0.8761924505233765,
+      "learning_rate": 3e-05,
+      "loss": 3.1571,
+      "step": 2787
+    },
+    {
+      "epoch": 0.7576086956521739,
+      "grad_norm": 1.2133067846298218,
+      "learning_rate": 3e-05,
+      "loss": 3.3668,
+      "step": 2788
+    },
+    {
+      "epoch": 0.7578804347826087,
+      "grad_norm": 1.0436729192733765,
+      "learning_rate": 3e-05,
+      "loss": 3.338,
+      "step": 2789
+    },
+    {
+      "epoch": 0.7581521739130435,
+      "grad_norm": 1.0040472745895386,
+      "learning_rate": 3e-05,
+      "loss": 3.314,
+      "step": 2790
+    },
+    {
+      "epoch": 0.7584239130434782,
+      "grad_norm": 1.019578456878662,
+      "learning_rate": 3e-05,
+      "loss": 3.3716,
+      "step": 2791
+    },
+    {
+      "epoch": 0.758695652173913,
+      "grad_norm": 1.209367036819458,
+      "learning_rate": 3e-05,
+      "loss": 3.7626,
+      "step": 2792
+    },
+    {
+      "epoch": 0.7589673913043479,
+      "grad_norm": 1.0667492151260376,
+      "learning_rate": 3e-05,
+      "loss": 3.4984,
+      "step": 2793
+    },
+    {
+      "epoch": 0.7592391304347826,
+      "grad_norm": 1.200366497039795,
+      "learning_rate": 3e-05,
+      "loss": 3.228,
+      "step": 2794
+    },
+    {
+      "epoch": 0.7595108695652174,
+      "grad_norm": 1.2726914882659912,
+      "learning_rate": 3e-05,
+      "loss": 2.9393,
+      "step": 2795
+    },
+    {
+      "epoch": 0.7597826086956522,
+      "grad_norm": 1.244170904159546,
+      "learning_rate": 3e-05,
+      "loss": 2.9552,
+      "step": 2796
+    },
+    {
+      "epoch": 0.7600543478260869,
+      "grad_norm": 1.113420844078064,
+      "learning_rate": 3e-05,
+      "loss": 3.1033,
+      "step": 2797
+    },
+    {
+      "epoch": 0.7603260869565217,
+      "grad_norm": 1.9228954315185547,
+      "learning_rate": 3e-05,
+      "loss": 3.7877,
+      "step": 2798
+    },
+    {
+      "epoch": 0.7605978260869565,
+      "grad_norm": 1.1893031597137451,
+      "learning_rate": 3e-05,
+      "loss": 3.0852,
+      "step": 2799
+    },
+    {
+      "epoch": 0.7608695652173914,
+      "grad_norm": 1.524493932723999,
+      "learning_rate": 3e-05,
+      "loss": 4.1925,
+      "step": 2800
+    },
+    {
+      "epoch": 0.7611413043478261,
+      "grad_norm": 1.5600111484527588,
+      "learning_rate": 3e-05,
+      "loss": 3.4646,
+      "step": 2801
+    },
+    {
+      "epoch": 0.7614130434782609,
+      "grad_norm": 0.9608951210975647,
+      "learning_rate": 3e-05,
+      "loss": 2.8357,
+      "step": 2802
+    },
+    {
+      "epoch": 0.7616847826086957,
+      "grad_norm": 1.2598545551300049,
+      "learning_rate": 3e-05,
+      "loss": 3.2298,
+      "step": 2803
+    },
+    {
+      "epoch": 0.7619565217391304,
+      "grad_norm": 1.4729243516921997,
+      "learning_rate": 3e-05,
+      "loss": 3.2004,
+      "step": 2804
+    },
+    {
+      "epoch": 0.7622282608695652,
+      "grad_norm": 1.04935622215271,
+      "learning_rate": 3e-05,
+      "loss": 2.9192,
+      "step": 2805
+    },
+    {
+      "epoch": 0.7625,
+      "grad_norm": 1.1317492723464966,
+      "learning_rate": 3e-05,
+      "loss": 3.5182,
+      "step": 2806
+    },
+    {
+      "epoch": 0.7627717391304348,
+      "grad_norm": 0.8651010394096375,
+      "learning_rate": 3e-05,
+      "loss": 2.7038,
+      "step": 2807
+    },
+    {
+      "epoch": 0.7630434782608696,
+      "grad_norm": 1.4115127325057983,
+      "learning_rate": 3e-05,
+      "loss": 3.225,
+      "step": 2808
+    },
+    {
+      "epoch": 0.7633152173913044,
+      "grad_norm": 1.0244847536087036,
+      "learning_rate": 3e-05,
+      "loss": 3.1254,
+      "step": 2809
+    },
+    {
+      "epoch": 0.7635869565217391,
+      "grad_norm": 1.4117242097854614,
+      "learning_rate": 3e-05,
+      "loss": 3.4427,
+      "step": 2810
+    },
+    {
+      "epoch": 0.7638586956521739,
+      "grad_norm": 1.3557825088500977,
+      "learning_rate": 3e-05,
+      "loss": 3.2669,
+      "step": 2811
+    },
+    {
+      "epoch": 0.7641304347826087,
+      "grad_norm": 1.2034603357315063,
+      "learning_rate": 3e-05,
+      "loss": 3.77,
+      "step": 2812
+    },
+    {
+      "epoch": 0.7644021739130434,
+      "grad_norm": 1.2670207023620605,
+      "learning_rate": 3e-05,
+      "loss": 3.1149,
+      "step": 2813
+    },
+    {
+      "epoch": 0.7646739130434783,
+      "grad_norm": 0.9181430339813232,
+      "learning_rate": 3e-05,
+      "loss": 3.0427,
+      "step": 2814
+    },
+    {
+      "epoch": 0.7649456521739131,
+      "grad_norm": 1.2132922410964966,
+      "learning_rate": 3e-05,
+      "loss": 3.4703,
+      "step": 2815
+    },
+    {
+      "epoch": 0.7652173913043478,
+      "grad_norm": 0.9925717115402222,
+      "learning_rate": 3e-05,
+      "loss": 3.1396,
+      "step": 2816
+    },
+    {
+      "epoch": 0.7654891304347826,
+      "grad_norm": 1.0748741626739502,
+      "learning_rate": 3e-05,
+      "loss": 2.8827,
+      "step": 2817
+    },
+    {
+      "epoch": 0.7657608695652174,
+      "grad_norm": 1.0484018325805664,
+      "learning_rate": 3e-05,
+      "loss": 3.1837,
+      "step": 2818
+    },
+    {
+      "epoch": 0.7660326086956522,
+      "grad_norm": 1.1004911661148071,
+      "learning_rate": 3e-05,
+      "loss": 3.135,
+      "step": 2819
+    },
+    {
+      "epoch": 0.7663043478260869,
+      "grad_norm": 1.3505610227584839,
+      "learning_rate": 3e-05,
+      "loss": 3.3415,
+      "step": 2820
+    },
+    {
+      "epoch": 0.7665760869565217,
+      "grad_norm": 1.1511911153793335,
+      "learning_rate": 3e-05,
+      "loss": 3.3407,
+      "step": 2821
+    },
+    {
+      "epoch": 0.7668478260869566,
+      "grad_norm": 1.2488864660263062,
+      "learning_rate": 3e-05,
+      "loss": 3.0697,
+      "step": 2822
+    },
+    {
+      "epoch": 0.7671195652173913,
+      "grad_norm": 0.970241129398346,
+      "learning_rate": 3e-05,
+      "loss": 2.6801,
+      "step": 2823
+    },
+    {
+      "epoch": 0.7673913043478261,
+      "grad_norm": 1.3036609888076782,
+      "learning_rate": 3e-05,
+      "loss": 3.7613,
+      "step": 2824
+    },
+    {
+      "epoch": 0.7676630434782609,
+      "grad_norm": 1.1077839136123657,
+      "learning_rate": 3e-05,
+      "loss": 3.829,
+      "step": 2825
+    },
+    {
+      "epoch": 0.7679347826086956,
+      "grad_norm": 1.1794723272323608,
+      "learning_rate": 3e-05,
+      "loss": 3.1147,
+      "step": 2826
+    },
+    {
+      "epoch": 0.7682065217391304,
+      "grad_norm": 1.460880994796753,
+      "learning_rate": 3e-05,
+      "loss": 3.7514,
+      "step": 2827
+    },
+    {
+      "epoch": 0.7684782608695652,
+      "grad_norm": 1.2111482620239258,
+      "learning_rate": 3e-05,
+      "loss": 3.5524,
+      "step": 2828
+    },
+    {
+      "epoch": 0.76875,
+      "grad_norm": 1.426535725593567,
+      "learning_rate": 3e-05,
+      "loss": 4.0174,
+      "step": 2829
+    },
+    {
+      "epoch": 0.7690217391304348,
+      "grad_norm": 1.4052951335906982,
+      "learning_rate": 3e-05,
+      "loss": 3.3475,
+      "step": 2830
+    },
+    {
+      "epoch": 0.7692934782608696,
+      "grad_norm": 0.9852263331413269,
+      "learning_rate": 3e-05,
+      "loss": 2.9218,
+      "step": 2831
+    },
+    {
+      "epoch": 0.7695652173913043,
+      "grad_norm": 1.2342568635940552,
+      "learning_rate": 3e-05,
+      "loss": 3.1937,
+      "step": 2832
+    },
+    {
+      "epoch": 0.7698369565217391,
+      "grad_norm": 2.569143772125244,
+      "learning_rate": 3e-05,
+      "loss": 3.8627,
+      "step": 2833
+    },
+    {
+      "epoch": 0.7701086956521739,
+      "grad_norm": 1.07598876953125,
+      "learning_rate": 3e-05,
+      "loss": 2.8528,
+      "step": 2834
+    },
+    {
+      "epoch": 0.7703804347826086,
+      "grad_norm": 1.0930018424987793,
+      "learning_rate": 3e-05,
+      "loss": 3.6242,
+      "step": 2835
+    },
+    {
+      "epoch": 0.7706521739130435,
+      "grad_norm": 1.571661353111267,
+      "learning_rate": 3e-05,
+      "loss": 3.2027,
+      "step": 2836
+    },
+    {
+      "epoch": 0.7709239130434783,
+      "grad_norm": 1.623907446861267,
+      "learning_rate": 3e-05,
+      "loss": 3.9722,
+      "step": 2837
+    },
+    {
+      "epoch": 0.7711956521739131,
+      "grad_norm": 1.4214915037155151,
+      "learning_rate": 3e-05,
+      "loss": 3.483,
+      "step": 2838
+    },
+    {
+      "epoch": 0.7714673913043478,
+      "grad_norm": 1.1177326440811157,
+      "learning_rate": 3e-05,
+      "loss": 3.2388,
+      "step": 2839
+    },
+    {
+      "epoch": 0.7717391304347826,
+      "grad_norm": 1.2823110818862915,
+      "learning_rate": 3e-05,
+      "loss": 3.2644,
+      "step": 2840
+    },
+    {
+      "epoch": 0.7720108695652174,
+      "grad_norm": 1.169673204421997,
+      "learning_rate": 3e-05,
+      "loss": 3.8629,
+      "step": 2841
+    },
+    {
+      "epoch": 0.7722826086956521,
+      "grad_norm": 1.5543997287750244,
+      "learning_rate": 3e-05,
+      "loss": 3.6887,
+      "step": 2842
+    },
+    {
+      "epoch": 0.772554347826087,
+      "grad_norm": 0.8735353946685791,
+      "learning_rate": 3e-05,
+      "loss": 2.7277,
+      "step": 2843
+    },
+    {
+      "epoch": 0.7728260869565218,
+      "grad_norm": 0.8694882988929749,
+      "learning_rate": 3e-05,
+      "loss": 2.9491,
+      "step": 2844
+    },
+    {
+      "epoch": 0.7730978260869565,
+      "grad_norm": 1.2447638511657715,
+      "learning_rate": 3e-05,
+      "loss": 3.473,
+      "step": 2845
+    },
+    {
+      "epoch": 0.7733695652173913,
+      "grad_norm": 0.869049072265625,
+      "learning_rate": 3e-05,
+      "loss": 3.0047,
+      "step": 2846
+    },
+    {
+      "epoch": 0.7736413043478261,
+      "grad_norm": 1.1336791515350342,
+      "learning_rate": 3e-05,
+      "loss": 3.5196,
+      "step": 2847
+    },
+    {
+      "epoch": 0.7739130434782608,
+      "grad_norm": 1.0776828527450562,
+      "learning_rate": 3e-05,
+      "loss": 3.2469,
+      "step": 2848
+    },
+    {
+      "epoch": 0.7741847826086956,
+      "grad_norm": 1.1838096380233765,
+      "learning_rate": 3e-05,
+      "loss": 3.5876,
+      "step": 2849
+    },
+    {
+      "epoch": 0.7744565217391305,
+      "grad_norm": 0.9228270649909973,
+      "learning_rate": 3e-05,
+      "loss": 3.0038,
+      "step": 2850
+    },
+    {
+      "epoch": 0.7747282608695653,
+      "grad_norm": 0.9622864127159119,
+      "learning_rate": 3e-05,
+      "loss": 2.7338,
+      "step": 2851
+    },
+    {
+      "epoch": 0.775,
+      "grad_norm": 1.3739656209945679,
+      "learning_rate": 3e-05,
+      "loss": 3.3167,
+      "step": 2852
+    },
+    {
+      "epoch": 0.7752717391304348,
+      "grad_norm": 1.2962409257888794,
+      "learning_rate": 3e-05,
+      "loss": 2.9095,
+      "step": 2853
+    },
+    {
+      "epoch": 0.7755434782608696,
+      "grad_norm": 1.3018954992294312,
+      "learning_rate": 3e-05,
+      "loss": 3.4756,
+      "step": 2854
+    },
+    {
+      "epoch": 0.7758152173913043,
+      "grad_norm": 1.2301543951034546,
+      "learning_rate": 3e-05,
+      "loss": 3.658,
+      "step": 2855
+    },
+    {
+      "epoch": 0.7760869565217391,
+      "grad_norm": 0.9971374869346619,
+      "learning_rate": 3e-05,
+      "loss": 2.8921,
+      "step": 2856
+    },
+    {
+      "epoch": 0.7763586956521739,
+      "grad_norm": 1.0504611730575562,
+      "learning_rate": 3e-05,
+      "loss": 2.907,
+      "step": 2857
+    },
+    {
+      "epoch": 0.7766304347826087,
+      "grad_norm": 0.9376839399337769,
+      "learning_rate": 3e-05,
+      "loss": 2.9993,
+      "step": 2858
+    },
+    {
+      "epoch": 0.7769021739130435,
+      "grad_norm": 1.1995012760162354,
+      "learning_rate": 3e-05,
+      "loss": 3.505,
+      "step": 2859
+    },
+    {
+      "epoch": 0.7771739130434783,
+      "grad_norm": 1.2391998767852783,
+      "learning_rate": 3e-05,
+      "loss": 3.5183,
+      "step": 2860
+    },
+    {
+      "epoch": 0.777445652173913,
+      "grad_norm": 1.0271795988082886,
+      "learning_rate": 3e-05,
+      "loss": 2.764,
+      "step": 2861
+    },
+    {
+      "epoch": 0.7777173913043478,
+      "grad_norm": 1.1367555856704712,
+      "learning_rate": 3e-05,
+      "loss": 3.4225,
+      "step": 2862
+    },
+    {
+      "epoch": 0.7779891304347826,
+      "grad_norm": 1.258692979812622,
+      "learning_rate": 3e-05,
+      "loss": 3.9128,
+      "step": 2863
+    },
+    {
+      "epoch": 0.7782608695652173,
+      "grad_norm": 1.050280213356018,
+      "learning_rate": 3e-05,
+      "loss": 3.1644,
+      "step": 2864
+    },
+    {
+      "epoch": 0.7785326086956522,
+      "grad_norm": 1.4861702919006348,
+      "learning_rate": 3e-05,
+      "loss": 3.651,
+      "step": 2865
+    },
+    {
+      "epoch": 0.778804347826087,
+      "grad_norm": 1.9443000555038452,
+      "learning_rate": 3e-05,
+      "loss": 3.1064,
+      "step": 2866
+    },
+    {
+      "epoch": 0.7790760869565218,
+      "grad_norm": 1.2355011701583862,
+      "learning_rate": 3e-05,
+      "loss": 3.2686,
+      "step": 2867
+    },
+    {
+      "epoch": 0.7793478260869565,
+      "grad_norm": 1.06037175655365,
+      "learning_rate": 3e-05,
+      "loss": 3.4073,
+      "step": 2868
+    },
+    {
+      "epoch": 0.7796195652173913,
+      "grad_norm": 1.3563976287841797,
+      "learning_rate": 3e-05,
+      "loss": 3.9419,
+      "step": 2869
+    },
+    {
+      "epoch": 0.779891304347826,
+      "grad_norm": 1.411892294883728,
+      "learning_rate": 3e-05,
+      "loss": 3.4305,
+      "step": 2870
+    },
+    {
+      "epoch": 0.7801630434782608,
+      "grad_norm": 2.1324245929718018,
+      "learning_rate": 3e-05,
+      "loss": 3.5014,
+      "step": 2871
+    },
+    {
+      "epoch": 0.7804347826086957,
+      "grad_norm": 1.3413655757904053,
+      "learning_rate": 3e-05,
+      "loss": 3.4645,
+      "step": 2872
+    },
+    {
+      "epoch": 0.7807065217391305,
+      "grad_norm": 1.2617512941360474,
+      "learning_rate": 3e-05,
+      "loss": 3.1992,
+      "step": 2873
+    },
+    {
+      "epoch": 0.7809782608695652,
+      "grad_norm": 1.1358786821365356,
+      "learning_rate": 3e-05,
+      "loss": 3.223,
+      "step": 2874
+    },
+    {
+      "epoch": 0.78125,
+      "grad_norm": 1.284077525138855,
+      "learning_rate": 3e-05,
+      "loss": 3.2483,
+      "step": 2875
+    },
+    {
+      "epoch": 0.7815217391304348,
+      "grad_norm": 2.2102537155151367,
+      "learning_rate": 3e-05,
+      "loss": 3.3587,
+      "step": 2876
+    },
+    {
+      "epoch": 0.7817934782608695,
+      "grad_norm": 1.2199740409851074,
+      "learning_rate": 3e-05,
+      "loss": 3.0096,
+      "step": 2877
+    },
+    {
+      "epoch": 0.7820652173913043,
+      "grad_norm": 1.0134555101394653,
+      "learning_rate": 3e-05,
+      "loss": 3.0385,
+      "step": 2878
+    },
+    {
+      "epoch": 0.7823369565217392,
+      "grad_norm": 0.9114274978637695,
+      "learning_rate": 3e-05,
+      "loss": 2.7133,
+      "step": 2879
+    },
+    {
+      "epoch": 0.782608695652174,
+      "grad_norm": 1.0371575355529785,
+      "learning_rate": 3e-05,
+      "loss": 3.1674,
+      "step": 2880
+    },
+    {
+      "epoch": 0.7828804347826087,
+      "grad_norm": 1.2628506422042847,
+      "learning_rate": 3e-05,
+      "loss": 3.5175,
+      "step": 2881
+    },
+    {
+      "epoch": 0.7831521739130435,
+      "grad_norm": 1.2910369634628296,
+      "learning_rate": 3e-05,
+      "loss": 3.1921,
+      "step": 2882
+    },
+    {
+      "epoch": 0.7834239130434782,
+      "grad_norm": 1.460605263710022,
+      "learning_rate": 3e-05,
+      "loss": 3.2101,
+      "step": 2883
+    },
+    {
+      "epoch": 0.783695652173913,
+      "grad_norm": 1.4789795875549316,
+      "learning_rate": 3e-05,
+      "loss": 3.5643,
+      "step": 2884
+    },
+    {
+      "epoch": 0.7839673913043478,
+      "grad_norm": 1.3687986135482788,
+      "learning_rate": 3e-05,
+      "loss": 3.5187,
+      "step": 2885
+    },
+    {
+      "epoch": 0.7842391304347827,
+      "grad_norm": 1.4191471338272095,
+      "learning_rate": 3e-05,
+      "loss": 3.8682,
+      "step": 2886
+    },
+    {
+      "epoch": 0.7845108695652174,
+      "grad_norm": 1.3148168325424194,
+      "learning_rate": 3e-05,
+      "loss": 3.0195,
+      "step": 2887
+    },
+    {
+      "epoch": 0.7847826086956522,
+      "grad_norm": 1.15743088722229,
+      "learning_rate": 3e-05,
+      "loss": 2.9205,
+      "step": 2888
+    },
+    {
+      "epoch": 0.785054347826087,
+      "grad_norm": 1.167377233505249,
+      "learning_rate": 3e-05,
+      "loss": 3.3717,
+      "step": 2889
+    },
+    {
+      "epoch": 0.7853260869565217,
+      "grad_norm": 1.369253158569336,
+      "learning_rate": 3e-05,
+      "loss": 3.3943,
+      "step": 2890
+    },
+    {
+      "epoch": 0.7855978260869565,
+      "grad_norm": 1.3081823587417603,
+      "learning_rate": 3e-05,
+      "loss": 3.1458,
+      "step": 2891
+    },
+    {
+      "epoch": 0.7858695652173913,
+      "grad_norm": 1.0637850761413574,
+      "learning_rate": 3e-05,
+      "loss": 3.5853,
+      "step": 2892
+    },
+    {
+      "epoch": 0.7861413043478261,
+      "grad_norm": 1.325531005859375,
+      "learning_rate": 3e-05,
+      "loss": 3.6246,
+      "step": 2893
+    },
+    {
+      "epoch": 0.7864130434782609,
+      "grad_norm": 1.2909876108169556,
+      "learning_rate": 3e-05,
+      "loss": 2.7209,
+      "step": 2894
+    },
+    {
+      "epoch": 0.7866847826086957,
+      "grad_norm": 1.072770357131958,
+      "learning_rate": 3e-05,
+      "loss": 2.7364,
+      "step": 2895
+    },
+    {
+      "epoch": 0.7869565217391304,
+      "grad_norm": 1.2623767852783203,
+      "learning_rate": 3e-05,
+      "loss": 3.193,
+      "step": 2896
+    },
+    {
+      "epoch": 0.7872282608695652,
+      "grad_norm": 1.262104868888855,
+      "learning_rate": 3e-05,
+      "loss": 3.4054,
+      "step": 2897
+    },
+    {
+      "epoch": 0.7875,
+      "grad_norm": 0.9568426012992859,
+      "learning_rate": 3e-05,
+      "loss": 3.0418,
+      "step": 2898
+    },
+    {
+      "epoch": 0.7877717391304347,
+      "grad_norm": 1.0304343700408936,
+      "learning_rate": 3e-05,
+      "loss": 3.1849,
+      "step": 2899
+    },
+    {
+      "epoch": 0.7880434782608695,
+      "grad_norm": 1.1511303186416626,
+      "learning_rate": 3e-05,
+      "loss": 3.3258,
+      "step": 2900
+    },
+    {
+      "epoch": 0.7883152173913044,
+      "grad_norm": 0.9511053562164307,
+      "learning_rate": 3e-05,
+      "loss": 2.9477,
+      "step": 2901
+    },
+    {
+      "epoch": 0.7885869565217392,
+      "grad_norm": 1.1826517581939697,
+      "learning_rate": 3e-05,
+      "loss": 2.9749,
+      "step": 2902
+    },
+    {
+      "epoch": 0.7888586956521739,
+      "grad_norm": 0.9371961355209351,
+      "learning_rate": 3e-05,
+      "loss": 3.0142,
+      "step": 2903
+    },
+    {
+      "epoch": 0.7891304347826087,
+      "grad_norm": 1.3846261501312256,
+      "learning_rate": 3e-05,
+      "loss": 3.3297,
+      "step": 2904
+    },
+    {
+      "epoch": 0.7894021739130435,
+      "grad_norm": 1.09294593334198,
+      "learning_rate": 3e-05,
+      "loss": 3.1767,
+      "step": 2905
+    },
+    {
+      "epoch": 0.7896739130434782,
+      "grad_norm": 1.326731562614441,
+      "learning_rate": 3e-05,
+      "loss": 3.8467,
+      "step": 2906
+    },
+    {
+      "epoch": 0.789945652173913,
+      "grad_norm": 1.1063183546066284,
+      "learning_rate": 3e-05,
+      "loss": 3.0861,
+      "step": 2907
+    },
+    {
+      "epoch": 0.7902173913043479,
+      "grad_norm": 1.174807071685791,
+      "learning_rate": 3e-05,
+      "loss": 3.0249,
+      "step": 2908
+    },
+    {
+      "epoch": 0.7904891304347826,
+      "grad_norm": 0.9046750068664551,
+      "learning_rate": 3e-05,
+      "loss": 3.0051,
+      "step": 2909
+    },
+    {
+      "epoch": 0.7907608695652174,
+      "grad_norm": 2.0610415935516357,
+      "learning_rate": 3e-05,
+      "loss": 3.9009,
+      "step": 2910
+    },
+    {
+      "epoch": 0.7910326086956522,
+      "grad_norm": 1.4573287963867188,
+      "learning_rate": 3e-05,
+      "loss": 3.1832,
+      "step": 2911
+    },
+    {
+      "epoch": 0.7913043478260869,
+      "grad_norm": 1.1538214683532715,
+      "learning_rate": 3e-05,
+      "loss": 2.7802,
+      "step": 2912
+    },
+    {
+      "epoch": 0.7915760869565217,
+      "grad_norm": 1.0094918012619019,
+      "learning_rate": 3e-05,
+      "loss": 3.0695,
+      "step": 2913
+    },
+    {
+      "epoch": 0.7918478260869565,
+      "grad_norm": 1.1341625452041626,
+      "learning_rate": 3e-05,
+      "loss": 3.4901,
+      "step": 2914
+    },
+    {
+      "epoch": 0.7921195652173914,
+      "grad_norm": 1.0618107318878174,
+      "learning_rate": 3e-05,
+      "loss": 3.1211,
+      "step": 2915
+    },
+    {
+      "epoch": 0.7923913043478261,
+      "grad_norm": 1.4313887357711792,
+      "learning_rate": 3e-05,
+      "loss": 3.8207,
+      "step": 2916
+    },
+    {
+      "epoch": 0.7926630434782609,
+      "grad_norm": 1.4587217569351196,
+      "learning_rate": 3e-05,
+      "loss": 3.5706,
+      "step": 2917
+    },
+    {
+      "epoch": 0.7929347826086957,
+      "grad_norm": 1.4848532676696777,
+      "learning_rate": 3e-05,
+      "loss": 3.6406,
+      "step": 2918
+    },
+    {
+      "epoch": 0.7932065217391304,
+      "grad_norm": 1.006595253944397,
+      "learning_rate": 3e-05,
+      "loss": 3.0846,
+      "step": 2919
+    },
+    {
+      "epoch": 0.7934782608695652,
+      "grad_norm": 1.184501051902771,
+      "learning_rate": 3e-05,
+      "loss": 3.3719,
+      "step": 2920
+    },
+    {
+      "epoch": 0.79375,
+      "grad_norm": 1.3015398979187012,
+      "learning_rate": 3e-05,
+      "loss": 2.9707,
+      "step": 2921
+    },
+    {
+      "epoch": 0.7940217391304348,
+      "grad_norm": 1.001300573348999,
+      "learning_rate": 3e-05,
+      "loss": 2.9032,
+      "step": 2922
+    },
+    {
+      "epoch": 0.7942934782608696,
+      "grad_norm": 1.4572199583053589,
+      "learning_rate": 3e-05,
+      "loss": 3.7695,
+      "step": 2923
+    },
+    {
+      "epoch": 0.7945652173913044,
+      "grad_norm": 1.3405942916870117,
+      "learning_rate": 3e-05,
+      "loss": 2.8765,
+      "step": 2924
+    },
+    {
+      "epoch": 0.7948369565217391,
+      "grad_norm": 1.6628684997558594,
+      "learning_rate": 3e-05,
+      "loss": 3.6848,
+      "step": 2925
+    },
+    {
+      "epoch": 0.7951086956521739,
+      "grad_norm": 1.3848426342010498,
+      "learning_rate": 3e-05,
+      "loss": 3.3296,
+      "step": 2926
+    },
+    {
+      "epoch": 0.7953804347826087,
+      "grad_norm": 1.5421236753463745,
+      "learning_rate": 3e-05,
+      "loss": 3.363,
+      "step": 2927
+    },
+    {
+      "epoch": 0.7956521739130434,
+      "grad_norm": 1.3111261129379272,
+      "learning_rate": 3e-05,
+      "loss": 2.8346,
+      "step": 2928
+    },
+    {
+      "epoch": 0.7959239130434783,
+      "grad_norm": 1.0846003293991089,
+      "learning_rate": 3e-05,
+      "loss": 3.3071,
+      "step": 2929
+    },
+    {
+      "epoch": 0.7961956521739131,
+      "grad_norm": 0.9776524305343628,
+      "learning_rate": 3e-05,
+      "loss": 2.9533,
+      "step": 2930
+    },
+    {
+      "epoch": 0.7964673913043478,
+      "grad_norm": 1.3840489387512207,
+      "learning_rate": 3e-05,
+      "loss": 3.4739,
+      "step": 2931
+    },
+    {
+      "epoch": 0.7967391304347826,
+      "grad_norm": 1.2059979438781738,
+      "learning_rate": 3e-05,
+      "loss": 2.9474,
+      "step": 2932
+    },
+    {
+      "epoch": 0.7970108695652174,
+      "grad_norm": 1.2110824584960938,
+      "learning_rate": 3e-05,
+      "loss": 3.2581,
+      "step": 2933
+    },
+    {
+      "epoch": 0.7972826086956522,
+      "grad_norm": 1.0525647401809692,
+      "learning_rate": 3e-05,
+      "loss": 2.9519,
+      "step": 2934
+    },
+    {
+      "epoch": 0.7975543478260869,
+      "grad_norm": 1.2633483409881592,
+      "learning_rate": 3e-05,
+      "loss": 3.1686,
+      "step": 2935
+    },
+    {
+      "epoch": 0.7978260869565217,
+      "grad_norm": 0.8925679922103882,
+      "learning_rate": 3e-05,
+      "loss": 2.8761,
+      "step": 2936
+    },
+    {
+      "epoch": 0.7980978260869566,
+      "grad_norm": 0.9802154898643494,
+      "learning_rate": 3e-05,
+      "loss": 3.2334,
+      "step": 2937
+    },
+    {
+      "epoch": 0.7983695652173913,
+      "grad_norm": 1.3679046630859375,
+      "learning_rate": 3e-05,
+      "loss": 3.2836,
+      "step": 2938
+    },
+    {
+      "epoch": 0.7986413043478261,
+      "grad_norm": 1.0334969758987427,
+      "learning_rate": 3e-05,
+      "loss": 3.1355,
+      "step": 2939
+    },
+    {
+      "epoch": 0.7989130434782609,
+      "grad_norm": 1.0484052896499634,
+      "learning_rate": 3e-05,
+      "loss": 3.1652,
+      "step": 2940
+    },
+    {
+      "epoch": 0.7991847826086956,
+      "grad_norm": 1.0501008033752441,
+      "learning_rate": 3e-05,
+      "loss": 2.9963,
+      "step": 2941
+    },
+    {
+      "epoch": 0.7994565217391304,
+      "grad_norm": 1.2090955972671509,
+      "learning_rate": 3e-05,
+      "loss": 3.3553,
+      "step": 2942
+    },
+    {
+      "epoch": 0.7997282608695652,
+      "grad_norm": 1.1819127798080444,
+      "learning_rate": 3e-05,
+      "loss": 3.1212,
+      "step": 2943
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1051278114318848,
+      "learning_rate": 3e-05,
+      "loss": 3.2308,
+      "step": 2944
+    },
+    {
+      "epoch": 0.8002717391304348,
+      "grad_norm": 0.9563131928443909,
+      "learning_rate": 3e-05,
+      "loss": 2.7561,
+      "step": 2945
+    },
+    {
+      "epoch": 0.8005434782608696,
+      "grad_norm": 1.1029945611953735,
+      "learning_rate": 3e-05,
+      "loss": 3.6245,
+      "step": 2946
+    },
+    {
+      "epoch": 0.8008152173913043,
+      "grad_norm": 1.3487423658370972,
+      "learning_rate": 3e-05,
+      "loss": 3.4707,
+      "step": 2947
+    },
+    {
+      "epoch": 0.8010869565217391,
+      "grad_norm": 1.7481412887573242,
+      "learning_rate": 3e-05,
+      "loss": 3.6706,
+      "step": 2948
+    },
+    {
+      "epoch": 0.8013586956521739,
+      "grad_norm": 1.0554581880569458,
+      "learning_rate": 3e-05,
+      "loss": 3.0406,
+      "step": 2949
+    },
+    {
+      "epoch": 0.8016304347826086,
+      "grad_norm": 1.0953891277313232,
+      "learning_rate": 3e-05,
+      "loss": 2.7938,
+      "step": 2950
+    },
+    {
+      "epoch": 0.8019021739130435,
+      "grad_norm": 1.3923141956329346,
+      "learning_rate": 3e-05,
+      "loss": 3.0691,
+      "step": 2951
+    },
+    {
+      "epoch": 0.8021739130434783,
+      "grad_norm": 1.1058765649795532,
+      "learning_rate": 3e-05,
+      "loss": 2.9356,
+      "step": 2952
+    },
+    {
+      "epoch": 0.8024456521739131,
+      "grad_norm": 1.1275757551193237,
+      "learning_rate": 3e-05,
+      "loss": 2.8867,
+      "step": 2953
+    },
+    {
+      "epoch": 0.8027173913043478,
+      "grad_norm": 1.2130517959594727,
+      "learning_rate": 3e-05,
+      "loss": 3.1133,
+      "step": 2954
+    },
+    {
+      "epoch": 0.8029891304347826,
+      "grad_norm": 1.6094809770584106,
+      "learning_rate": 3e-05,
+      "loss": 3.8152,
+      "step": 2955
+    },
+    {
+      "epoch": 0.8032608695652174,
+      "grad_norm": 0.9870564341545105,
+      "learning_rate": 3e-05,
+      "loss": 2.721,
+      "step": 2956
+    },
+    {
+      "epoch": 0.8035326086956521,
+      "grad_norm": 1.2996047735214233,
+      "learning_rate": 3e-05,
+      "loss": 3.6848,
+      "step": 2957
+    },
+    {
+      "epoch": 0.803804347826087,
+      "grad_norm": 1.2500205039978027,
+      "learning_rate": 3e-05,
+      "loss": 3.3352,
+      "step": 2958
+    },
+    {
+      "epoch": 0.8040760869565218,
+      "grad_norm": 1.3211798667907715,
+      "learning_rate": 3e-05,
+      "loss": 3.7284,
+      "step": 2959
+    },
+    {
+      "epoch": 0.8043478260869565,
+      "grad_norm": 1.514905571937561,
+      "learning_rate": 3e-05,
+      "loss": 3.2425,
+      "step": 2960
+    },
+    {
+      "epoch": 0.8046195652173913,
+      "grad_norm": 1.2107200622558594,
+      "learning_rate": 3e-05,
+      "loss": 3.4771,
+      "step": 2961
+    },
+    {
+      "epoch": 0.8048913043478261,
+      "grad_norm": 1.707463264465332,
+      "learning_rate": 3e-05,
+      "loss": 3.6518,
+      "step": 2962
+    },
+    {
+      "epoch": 0.8051630434782608,
+      "grad_norm": 0.9443992376327515,
+      "learning_rate": 3e-05,
+      "loss": 2.8471,
+      "step": 2963
+    },
+    {
+      "epoch": 0.8054347826086956,
+      "grad_norm": 1.3588606119155884,
+      "learning_rate": 3e-05,
+      "loss": 3.3993,
+      "step": 2964
+    },
+    {
+      "epoch": 0.8057065217391305,
+      "grad_norm": 1.2722697257995605,
+      "learning_rate": 3e-05,
+      "loss": 3.2316,
+      "step": 2965
+    },
+    {
+      "epoch": 0.8059782608695653,
+      "grad_norm": 1.6636009216308594,
+      "learning_rate": 3e-05,
+      "loss": 3.5648,
+      "step": 2966
+    },
+    {
+      "epoch": 0.80625,
+      "grad_norm": 1.3926972150802612,
+      "learning_rate": 3e-05,
+      "loss": 3.6403,
+      "step": 2967
+    },
+    {
+      "epoch": 0.8065217391304348,
+      "grad_norm": 1.2489240169525146,
+      "learning_rate": 3e-05,
+      "loss": 3.6428,
+      "step": 2968
+    },
+    {
+      "epoch": 0.8067934782608696,
+      "grad_norm": 1.3040508031845093,
+      "learning_rate": 3e-05,
+      "loss": 3.3698,
+      "step": 2969
+    },
+    {
+      "epoch": 0.8070652173913043,
+      "grad_norm": 1.024529218673706,
+      "learning_rate": 3e-05,
+      "loss": 3.1765,
+      "step": 2970
+    },
+    {
+      "epoch": 0.8073369565217391,
+      "grad_norm": 1.127133846282959,
+      "learning_rate": 3e-05,
+      "loss": 3.4167,
+      "step": 2971
+    },
+    {
+      "epoch": 0.8076086956521739,
+      "grad_norm": 1.5928646326065063,
+      "learning_rate": 3e-05,
+      "loss": 3.2939,
+      "step": 2972
+    },
+    {
+      "epoch": 0.8078804347826087,
+      "grad_norm": 1.099817156791687,
+      "learning_rate": 3e-05,
+      "loss": 3.3166,
+      "step": 2973
+    },
+    {
+      "epoch": 0.8081521739130435,
+      "grad_norm": 0.9949479699134827,
+      "learning_rate": 3e-05,
+      "loss": 3.4637,
+      "step": 2974
+    },
+    {
+      "epoch": 0.8084239130434783,
+      "grad_norm": 1.3012653589248657,
+      "learning_rate": 3e-05,
+      "loss": 3.3314,
+      "step": 2975
+    },
+    {
+      "epoch": 0.808695652173913,
+      "grad_norm": 1.1788009405136108,
+      "learning_rate": 3e-05,
+      "loss": 2.9917,
+      "step": 2976
+    },
+    {
+      "epoch": 0.8089673913043478,
+      "grad_norm": 1.5235824584960938,
+      "learning_rate": 3e-05,
+      "loss": 3.0784,
+      "step": 2977
+    },
+    {
+      "epoch": 0.8092391304347826,
+      "grad_norm": 1.4393073320388794,
+      "learning_rate": 3e-05,
+      "loss": 3.8917,
+      "step": 2978
+    },
+    {
+      "epoch": 0.8095108695652173,
+      "grad_norm": 1.432785987854004,
+      "learning_rate": 3e-05,
+      "loss": 3.8753,
+      "step": 2979
+    },
+    {
+      "epoch": 0.8097826086956522,
+      "grad_norm": 1.4102272987365723,
+      "learning_rate": 3e-05,
+      "loss": 3.0096,
+      "step": 2980
+    },
+    {
+      "epoch": 0.810054347826087,
+      "grad_norm": 1.5349915027618408,
+      "learning_rate": 3e-05,
+      "loss": 3.2429,
+      "step": 2981
+    },
+    {
+      "epoch": 0.8103260869565218,
+      "grad_norm": 0.9741056561470032,
+      "learning_rate": 3e-05,
+      "loss": 3.2216,
+      "step": 2982
+    },
+    {
+      "epoch": 0.8105978260869565,
+      "grad_norm": 1.3699191808700562,
+      "learning_rate": 3e-05,
+      "loss": 2.8594,
+      "step": 2983
+    },
+    {
+      "epoch": 0.8108695652173913,
+      "grad_norm": 1.0808976888656616,
+      "learning_rate": 3e-05,
+      "loss": 3.1742,
+      "step": 2984
+    },
+    {
+      "epoch": 0.811141304347826,
+      "grad_norm": 0.9032256007194519,
+      "learning_rate": 3e-05,
+      "loss": 3.0112,
+      "step": 2985
+    },
+    {
+      "epoch": 0.8114130434782608,
+      "grad_norm": 1.119470477104187,
+      "learning_rate": 3e-05,
+      "loss": 3.1507,
+      "step": 2986
+    },
+    {
+      "epoch": 0.8116847826086957,
+      "grad_norm": 1.3185955286026,
+      "learning_rate": 3e-05,
+      "loss": 3.7403,
+      "step": 2987
+    },
+    {
+      "epoch": 0.8119565217391305,
+      "grad_norm": 1.283615231513977,
+      "learning_rate": 3e-05,
+      "loss": 3.2814,
+      "step": 2988
+    },
+    {
+      "epoch": 0.8122282608695652,
+      "grad_norm": 1.2075817584991455,
+      "learning_rate": 3e-05,
+      "loss": 3.5796,
+      "step": 2989
+    },
+    {
+      "epoch": 0.8125,
+      "grad_norm": 1.1944336891174316,
+      "learning_rate": 3e-05,
+      "loss": 3.3505,
+      "step": 2990
+    },
+    {
+      "epoch": 0.8127717391304348,
+      "grad_norm": 1.1111395359039307,
+      "learning_rate": 3e-05,
+      "loss": 3.2605,
+      "step": 2991
+    },
+    {
+      "epoch": 0.8130434782608695,
+      "grad_norm": 1.2699017524719238,
+      "learning_rate": 3e-05,
+      "loss": 3.4753,
+      "step": 2992
+    },
+    {
+      "epoch": 0.8133152173913043,
+      "grad_norm": 1.1877753734588623,
+      "learning_rate": 3e-05,
+      "loss": 3.0174,
+      "step": 2993
+    },
+    {
+      "epoch": 0.8135869565217392,
+      "grad_norm": 1.198492407798767,
+      "learning_rate": 3e-05,
+      "loss": 3.4792,
+      "step": 2994
+    },
+    {
+      "epoch": 0.813858695652174,
+      "grad_norm": 1.2535799741744995,
+      "learning_rate": 3e-05,
+      "loss": 3.4246,
+      "step": 2995
+    },
+    {
+      "epoch": 0.8141304347826087,
+      "grad_norm": 1.0552626848220825,
+      "learning_rate": 3e-05,
+      "loss": 2.9214,
+      "step": 2996
+    },
+    {
+      "epoch": 0.8144021739130435,
+      "grad_norm": 1.1369494199752808,
+      "learning_rate": 3e-05,
+      "loss": 2.9704,
+      "step": 2997
+    },
+    {
+      "epoch": 0.8146739130434782,
+      "grad_norm": 1.04477858543396,
+      "learning_rate": 3e-05,
+      "loss": 2.916,
+      "step": 2998
+    },
+    {
+      "epoch": 0.814945652173913,
+      "grad_norm": 0.9620361328125,
+      "learning_rate": 3e-05,
+      "loss": 2.6156,
+      "step": 2999
+    },
+    {
+      "epoch": 0.8152173913043478,
+      "grad_norm": 1.0974937677383423,
+      "learning_rate": 3e-05,
+      "loss": 3.362,
+      "step": 3000
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 3680,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.262770368118784e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}