pt1 / trainer_state.json
szang18's picture
Upload folder using huggingface_hub
46047a6 verified
Raw
History Blame Contribute Delete
497 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 19.0,
"eval_steps": 500,
"global_step": 3059,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.006230529595015576,
"grad_norm": 22.5,
"learning_rate": 0.0,
"loss": 4.3003,
"step": 1
},
{
"epoch": 0.012461059190031152,
"grad_norm": 22.125,
"learning_rate": 1.5527950310559006e-07,
"loss": 4.3092,
"step": 2
},
{
"epoch": 0.018691588785046728,
"grad_norm": 21.875,
"learning_rate": 3.1055900621118013e-07,
"loss": 4.3271,
"step": 3
},
{
"epoch": 0.024922118380062305,
"grad_norm": 22.25,
"learning_rate": 4.6583850931677024e-07,
"loss": 4.4428,
"step": 4
},
{
"epoch": 0.03115264797507788,
"grad_norm": 21.25,
"learning_rate": 6.211180124223603e-07,
"loss": 4.269,
"step": 5
},
{
"epoch": 0.037383177570093455,
"grad_norm": 21.375,
"learning_rate": 7.763975155279503e-07,
"loss": 4.272,
"step": 6
},
{
"epoch": 0.04361370716510903,
"grad_norm": 22.125,
"learning_rate": 9.316770186335405e-07,
"loss": 4.3826,
"step": 7
},
{
"epoch": 0.04984423676012461,
"grad_norm": 21.625,
"learning_rate": 1.0869565217391306e-06,
"loss": 4.4407,
"step": 8
},
{
"epoch": 0.056074766355140186,
"grad_norm": 22.125,
"learning_rate": 1.2422360248447205e-06,
"loss": 4.2015,
"step": 9
},
{
"epoch": 0.06230529595015576,
"grad_norm": 21.75,
"learning_rate": 1.3975155279503107e-06,
"loss": 4.3222,
"step": 10
},
{
"epoch": 0.06853582554517133,
"grad_norm": 21.25,
"learning_rate": 1.5527950310559006e-06,
"loss": 4.2423,
"step": 11
},
{
"epoch": 0.07476635514018691,
"grad_norm": 21.5,
"learning_rate": 1.7080745341614908e-06,
"loss": 4.4348,
"step": 12
},
{
"epoch": 0.08099688473520249,
"grad_norm": 20.875,
"learning_rate": 1.863354037267081e-06,
"loss": 4.2771,
"step": 13
},
{
"epoch": 0.08722741433021806,
"grad_norm": 22.5,
"learning_rate": 2.018633540372671e-06,
"loss": 4.4241,
"step": 14
},
{
"epoch": 0.09345794392523364,
"grad_norm": 21.25,
"learning_rate": 2.173913043478261e-06,
"loss": 4.29,
"step": 15
},
{
"epoch": 0.09968847352024922,
"grad_norm": 21.375,
"learning_rate": 2.329192546583851e-06,
"loss": 4.155,
"step": 16
},
{
"epoch": 0.1059190031152648,
"grad_norm": 23.625,
"learning_rate": 2.484472049689441e-06,
"loss": 4.3643,
"step": 17
},
{
"epoch": 0.11214953271028037,
"grad_norm": 21.0,
"learning_rate": 2.639751552795031e-06,
"loss": 4.2453,
"step": 18
},
{
"epoch": 0.11838006230529595,
"grad_norm": 20.375,
"learning_rate": 2.7950310559006214e-06,
"loss": 4.2387,
"step": 19
},
{
"epoch": 0.12461059190031153,
"grad_norm": 20.625,
"learning_rate": 2.9503105590062115e-06,
"loss": 4.2028,
"step": 20
},
{
"epoch": 0.1308411214953271,
"grad_norm": 32.75,
"learning_rate": 3.1055900621118013e-06,
"loss": 4.2556,
"step": 21
},
{
"epoch": 0.13707165109034267,
"grad_norm": 20.75,
"learning_rate": 3.2608695652173914e-06,
"loss": 4.1559,
"step": 22
},
{
"epoch": 0.14330218068535824,
"grad_norm": 21.125,
"learning_rate": 3.4161490683229816e-06,
"loss": 4.3193,
"step": 23
},
{
"epoch": 0.14953271028037382,
"grad_norm": 23.25,
"learning_rate": 3.5714285714285714e-06,
"loss": 4.1798,
"step": 24
},
{
"epoch": 0.1557632398753894,
"grad_norm": 20.375,
"learning_rate": 3.726708074534162e-06,
"loss": 4.1705,
"step": 25
},
{
"epoch": 0.16199376947040497,
"grad_norm": 20.125,
"learning_rate": 3.881987577639752e-06,
"loss": 4.1772,
"step": 26
},
{
"epoch": 0.16822429906542055,
"grad_norm": 20.5,
"learning_rate": 4.037267080745342e-06,
"loss": 4.0817,
"step": 27
},
{
"epoch": 0.17445482866043613,
"grad_norm": 20.5,
"learning_rate": 4.192546583850932e-06,
"loss": 4.0005,
"step": 28
},
{
"epoch": 0.1806853582554517,
"grad_norm": 21.75,
"learning_rate": 4.347826086956522e-06,
"loss": 3.8614,
"step": 29
},
{
"epoch": 0.18691588785046728,
"grad_norm": 19.375,
"learning_rate": 4.503105590062112e-06,
"loss": 4.039,
"step": 30
},
{
"epoch": 0.19314641744548286,
"grad_norm": 19.125,
"learning_rate": 4.658385093167702e-06,
"loss": 3.9061,
"step": 31
},
{
"epoch": 0.19937694704049844,
"grad_norm": 18.375,
"learning_rate": 4.813664596273292e-06,
"loss": 3.8094,
"step": 32
},
{
"epoch": 0.205607476635514,
"grad_norm": 19.875,
"learning_rate": 4.968944099378882e-06,
"loss": 3.8203,
"step": 33
},
{
"epoch": 0.2118380062305296,
"grad_norm": 17.875,
"learning_rate": 5.124223602484472e-06,
"loss": 3.8233,
"step": 34
},
{
"epoch": 0.21806853582554517,
"grad_norm": 17.5,
"learning_rate": 5.279503105590062e-06,
"loss": 3.8332,
"step": 35
},
{
"epoch": 0.22429906542056074,
"grad_norm": 18.0,
"learning_rate": 5.4347826086956525e-06,
"loss": 3.7435,
"step": 36
},
{
"epoch": 0.23052959501557632,
"grad_norm": 17.75,
"learning_rate": 5.590062111801243e-06,
"loss": 3.6507,
"step": 37
},
{
"epoch": 0.2367601246105919,
"grad_norm": 16.75,
"learning_rate": 5.745341614906832e-06,
"loss": 3.6827,
"step": 38
},
{
"epoch": 0.24299065420560748,
"grad_norm": 17.625,
"learning_rate": 5.900621118012423e-06,
"loss": 3.6929,
"step": 39
},
{
"epoch": 0.24922118380062305,
"grad_norm": 16.25,
"learning_rate": 6.055900621118013e-06,
"loss": 3.5796,
"step": 40
},
{
"epoch": 0.2554517133956386,
"grad_norm": 16.25,
"learning_rate": 6.2111801242236025e-06,
"loss": 3.5128,
"step": 41
},
{
"epoch": 0.2616822429906542,
"grad_norm": 16.0,
"learning_rate": 6.366459627329193e-06,
"loss": 3.5517,
"step": 42
},
{
"epoch": 0.26791277258566976,
"grad_norm": 15.4375,
"learning_rate": 6.521739130434783e-06,
"loss": 3.4609,
"step": 43
},
{
"epoch": 0.27414330218068533,
"grad_norm": 15.875,
"learning_rate": 6.677018633540373e-06,
"loss": 3.4698,
"step": 44
},
{
"epoch": 0.2803738317757009,
"grad_norm": 14.5,
"learning_rate": 6.832298136645963e-06,
"loss": 3.4448,
"step": 45
},
{
"epoch": 0.2866043613707165,
"grad_norm": 13.875,
"learning_rate": 6.9875776397515525e-06,
"loss": 3.2967,
"step": 46
},
{
"epoch": 0.29283489096573206,
"grad_norm": 14.3125,
"learning_rate": 7.142857142857143e-06,
"loss": 3.2206,
"step": 47
},
{
"epoch": 0.29906542056074764,
"grad_norm": 13.75,
"learning_rate": 7.298136645962733e-06,
"loss": 3.2208,
"step": 48
},
{
"epoch": 0.3052959501557632,
"grad_norm": 13.0625,
"learning_rate": 7.453416149068324e-06,
"loss": 3.2302,
"step": 49
},
{
"epoch": 0.3115264797507788,
"grad_norm": 12.5625,
"learning_rate": 7.608695652173914e-06,
"loss": 3.1471,
"step": 50
},
{
"epoch": 0.3177570093457944,
"grad_norm": 11.6875,
"learning_rate": 7.763975155279503e-06,
"loss": 3.0274,
"step": 51
},
{
"epoch": 0.32398753894080995,
"grad_norm": 12.375,
"learning_rate": 7.919254658385093e-06,
"loss": 3.0628,
"step": 52
},
{
"epoch": 0.3302180685358255,
"grad_norm": 10.8125,
"learning_rate": 8.074534161490684e-06,
"loss": 2.9704,
"step": 53
},
{
"epoch": 0.3364485981308411,
"grad_norm": 9.75,
"learning_rate": 8.229813664596275e-06,
"loss": 2.8166,
"step": 54
},
{
"epoch": 0.3426791277258567,
"grad_norm": 8.9375,
"learning_rate": 8.385093167701864e-06,
"loss": 2.8457,
"step": 55
},
{
"epoch": 0.34890965732087226,
"grad_norm": 8.0625,
"learning_rate": 8.540372670807453e-06,
"loss": 2.8161,
"step": 56
},
{
"epoch": 0.35514018691588783,
"grad_norm": 7.78125,
"learning_rate": 8.695652173913044e-06,
"loss": 2.7495,
"step": 57
},
{
"epoch": 0.3613707165109034,
"grad_norm": 7.375,
"learning_rate": 8.850931677018634e-06,
"loss": 2.6288,
"step": 58
},
{
"epoch": 0.367601246105919,
"grad_norm": 6.28125,
"learning_rate": 9.006211180124225e-06,
"loss": 2.688,
"step": 59
},
{
"epoch": 0.37383177570093457,
"grad_norm": 6.25,
"learning_rate": 9.161490683229814e-06,
"loss": 2.6383,
"step": 60
},
{
"epoch": 0.38006230529595014,
"grad_norm": 5.4375,
"learning_rate": 9.316770186335403e-06,
"loss": 2.6195,
"step": 61
},
{
"epoch": 0.3862928348909657,
"grad_norm": 5.8125,
"learning_rate": 9.472049689440994e-06,
"loss": 2.5998,
"step": 62
},
{
"epoch": 0.3925233644859813,
"grad_norm": 5.65625,
"learning_rate": 9.627329192546584e-06,
"loss": 2.5223,
"step": 63
},
{
"epoch": 0.3987538940809969,
"grad_norm": 4.75,
"learning_rate": 9.782608695652175e-06,
"loss": 2.5611,
"step": 64
},
{
"epoch": 0.40498442367601245,
"grad_norm": 5.03125,
"learning_rate": 9.937888198757764e-06,
"loss": 2.3835,
"step": 65
},
{
"epoch": 0.411214953271028,
"grad_norm": 3.84375,
"learning_rate": 1.0093167701863353e-05,
"loss": 2.4907,
"step": 66
},
{
"epoch": 0.4174454828660436,
"grad_norm": 4.125,
"learning_rate": 1.0248447204968944e-05,
"loss": 2.4563,
"step": 67
},
{
"epoch": 0.4236760124610592,
"grad_norm": 3.921875,
"learning_rate": 1.0403726708074535e-05,
"loss": 2.4542,
"step": 68
},
{
"epoch": 0.42990654205607476,
"grad_norm": 3.625,
"learning_rate": 1.0559006211180125e-05,
"loss": 2.4394,
"step": 69
},
{
"epoch": 0.43613707165109034,
"grad_norm": 3.5625,
"learning_rate": 1.0714285714285714e-05,
"loss": 2.3642,
"step": 70
},
{
"epoch": 0.4423676012461059,
"grad_norm": 4.25,
"learning_rate": 1.0869565217391305e-05,
"loss": 2.4301,
"step": 71
},
{
"epoch": 0.4485981308411215,
"grad_norm": 4.21875,
"learning_rate": 1.1024844720496894e-05,
"loss": 2.4077,
"step": 72
},
{
"epoch": 0.45482866043613707,
"grad_norm": 4.0,
"learning_rate": 1.1180124223602485e-05,
"loss": 2.4943,
"step": 73
},
{
"epoch": 0.46105919003115264,
"grad_norm": 3.734375,
"learning_rate": 1.1335403726708076e-05,
"loss": 2.4186,
"step": 74
},
{
"epoch": 0.4672897196261682,
"grad_norm": 3.75,
"learning_rate": 1.1490683229813664e-05,
"loss": 2.3254,
"step": 75
},
{
"epoch": 0.4735202492211838,
"grad_norm": 3.3125,
"learning_rate": 1.1645962732919255e-05,
"loss": 2.3334,
"step": 76
},
{
"epoch": 0.4797507788161994,
"grad_norm": 3.359375,
"learning_rate": 1.1801242236024846e-05,
"loss": 2.3132,
"step": 77
},
{
"epoch": 0.48598130841121495,
"grad_norm": 3.984375,
"learning_rate": 1.1956521739130435e-05,
"loss": 2.2571,
"step": 78
},
{
"epoch": 0.49221183800623053,
"grad_norm": 3.5625,
"learning_rate": 1.2111801242236026e-05,
"loss": 2.3271,
"step": 79
},
{
"epoch": 0.4984423676012461,
"grad_norm": 4.15625,
"learning_rate": 1.2267080745341616e-05,
"loss": 2.4829,
"step": 80
},
{
"epoch": 0.5046728971962616,
"grad_norm": 5.53125,
"learning_rate": 1.2422360248447205e-05,
"loss": 2.4702,
"step": 81
},
{
"epoch": 0.5109034267912772,
"grad_norm": 3.359375,
"learning_rate": 1.2577639751552794e-05,
"loss": 2.4282,
"step": 82
},
{
"epoch": 0.5171339563862928,
"grad_norm": 3.375,
"learning_rate": 1.2732919254658385e-05,
"loss": 2.286,
"step": 83
},
{
"epoch": 0.5233644859813084,
"grad_norm": 3.828125,
"learning_rate": 1.2888198757763975e-05,
"loss": 2.403,
"step": 84
},
{
"epoch": 0.5295950155763239,
"grad_norm": 3.921875,
"learning_rate": 1.3043478260869566e-05,
"loss": 2.3953,
"step": 85
},
{
"epoch": 0.5358255451713395,
"grad_norm": 3.75,
"learning_rate": 1.3198757763975155e-05,
"loss": 2.3857,
"step": 86
},
{
"epoch": 0.5420560747663551,
"grad_norm": 3.3125,
"learning_rate": 1.3354037267080746e-05,
"loss": 2.3896,
"step": 87
},
{
"epoch": 0.5482866043613707,
"grad_norm": 3.546875,
"learning_rate": 1.3509316770186337e-05,
"loss": 2.2358,
"step": 88
},
{
"epoch": 0.5545171339563862,
"grad_norm": 3.5625,
"learning_rate": 1.3664596273291926e-05,
"loss": 2.3736,
"step": 89
},
{
"epoch": 0.5607476635514018,
"grad_norm": 3.859375,
"learning_rate": 1.3819875776397517e-05,
"loss": 2.4358,
"step": 90
},
{
"epoch": 0.5669781931464174,
"grad_norm": 3.9375,
"learning_rate": 1.3975155279503105e-05,
"loss": 2.3269,
"step": 91
},
{
"epoch": 0.573208722741433,
"grad_norm": 3.8125,
"learning_rate": 1.4130434782608694e-05,
"loss": 2.3836,
"step": 92
},
{
"epoch": 0.5794392523364486,
"grad_norm": 3.734375,
"learning_rate": 1.4285714285714285e-05,
"loss": 2.354,
"step": 93
},
{
"epoch": 0.5856697819314641,
"grad_norm": 3.6875,
"learning_rate": 1.4440993788819876e-05,
"loss": 2.3433,
"step": 94
},
{
"epoch": 0.5919003115264797,
"grad_norm": 3.953125,
"learning_rate": 1.4596273291925466e-05,
"loss": 2.2974,
"step": 95
},
{
"epoch": 0.5981308411214953,
"grad_norm": 3.5625,
"learning_rate": 1.4751552795031057e-05,
"loss": 2.2944,
"step": 96
},
{
"epoch": 0.6043613707165109,
"grad_norm": 3.828125,
"learning_rate": 1.4906832298136648e-05,
"loss": 2.3776,
"step": 97
},
{
"epoch": 0.6105919003115264,
"grad_norm": 3.859375,
"learning_rate": 1.5062111801242237e-05,
"loss": 2.3543,
"step": 98
},
{
"epoch": 0.616822429906542,
"grad_norm": 4.59375,
"learning_rate": 1.5217391304347828e-05,
"loss": 2.3162,
"step": 99
},
{
"epoch": 0.6230529595015576,
"grad_norm": 4.1875,
"learning_rate": 1.537267080745342e-05,
"loss": 2.3072,
"step": 100
},
{
"epoch": 0.6292834890965732,
"grad_norm": 3.46875,
"learning_rate": 1.5527950310559007e-05,
"loss": 2.2608,
"step": 101
},
{
"epoch": 0.6355140186915887,
"grad_norm": 3.5625,
"learning_rate": 1.5683229813664594e-05,
"loss": 2.1772,
"step": 102
},
{
"epoch": 0.6417445482866043,
"grad_norm": 4.625,
"learning_rate": 1.5838509316770185e-05,
"loss": 2.2931,
"step": 103
},
{
"epoch": 0.6479750778816199,
"grad_norm": 4.25,
"learning_rate": 1.5993788819875776e-05,
"loss": 2.3164,
"step": 104
},
{
"epoch": 0.6542056074766355,
"grad_norm": 4.9375,
"learning_rate": 1.6149068322981367e-05,
"loss": 2.2621,
"step": 105
},
{
"epoch": 0.660436137071651,
"grad_norm": 5.5,
"learning_rate": 1.630434782608696e-05,
"loss": 2.2707,
"step": 106
},
{
"epoch": 0.6666666666666666,
"grad_norm": 3.578125,
"learning_rate": 1.645962732919255e-05,
"loss": 2.3164,
"step": 107
},
{
"epoch": 0.6728971962616822,
"grad_norm": 5.09375,
"learning_rate": 1.6614906832298137e-05,
"loss": 2.2146,
"step": 108
},
{
"epoch": 0.6791277258566978,
"grad_norm": 5.21875,
"learning_rate": 1.6770186335403728e-05,
"loss": 2.1987,
"step": 109
},
{
"epoch": 0.6853582554517134,
"grad_norm": 5.125,
"learning_rate": 1.6925465838509316e-05,
"loss": 2.2464,
"step": 110
},
{
"epoch": 0.6915887850467289,
"grad_norm": 8.75,
"learning_rate": 1.7080745341614907e-05,
"loss": 2.2376,
"step": 111
},
{
"epoch": 0.6978193146417445,
"grad_norm": 4.03125,
"learning_rate": 1.7236024844720498e-05,
"loss": 2.1998,
"step": 112
},
{
"epoch": 0.7040498442367601,
"grad_norm": 6.625,
"learning_rate": 1.739130434782609e-05,
"loss": 2.209,
"step": 113
},
{
"epoch": 0.7102803738317757,
"grad_norm": 4.6875,
"learning_rate": 1.7546583850931676e-05,
"loss": 2.1031,
"step": 114
},
{
"epoch": 0.7165109034267912,
"grad_norm": 3.453125,
"learning_rate": 1.7701863354037267e-05,
"loss": 2.1656,
"step": 115
},
{
"epoch": 0.7227414330218068,
"grad_norm": 4.46875,
"learning_rate": 1.785714285714286e-05,
"loss": 2.1367,
"step": 116
},
{
"epoch": 0.7289719626168224,
"grad_norm": 5.3125,
"learning_rate": 1.801242236024845e-05,
"loss": 2.1261,
"step": 117
},
{
"epoch": 0.735202492211838,
"grad_norm": 3.671875,
"learning_rate": 1.816770186335404e-05,
"loss": 2.1594,
"step": 118
},
{
"epoch": 0.7414330218068536,
"grad_norm": 4.75,
"learning_rate": 1.8322981366459628e-05,
"loss": 2.1456,
"step": 119
},
{
"epoch": 0.7476635514018691,
"grad_norm": 4.40625,
"learning_rate": 1.8478260869565216e-05,
"loss": 2.1589,
"step": 120
},
{
"epoch": 0.7538940809968847,
"grad_norm": 3.53125,
"learning_rate": 1.8633540372670807e-05,
"loss": 2.0812,
"step": 121
},
{
"epoch": 0.7601246105919003,
"grad_norm": 3.6875,
"learning_rate": 1.8788819875776398e-05,
"loss": 2.1571,
"step": 122
},
{
"epoch": 0.7663551401869159,
"grad_norm": 3.203125,
"learning_rate": 1.894409937888199e-05,
"loss": 2.1413,
"step": 123
},
{
"epoch": 0.7725856697819314,
"grad_norm": 3.421875,
"learning_rate": 1.909937888198758e-05,
"loss": 2.0845,
"step": 124
},
{
"epoch": 0.778816199376947,
"grad_norm": 3.125,
"learning_rate": 1.9254658385093167e-05,
"loss": 2.1658,
"step": 125
},
{
"epoch": 0.7850467289719626,
"grad_norm": 3.09375,
"learning_rate": 1.940993788819876e-05,
"loss": 2.1945,
"step": 126
},
{
"epoch": 0.7912772585669782,
"grad_norm": 3.0,
"learning_rate": 1.956521739130435e-05,
"loss": 2.1947,
"step": 127
},
{
"epoch": 0.7975077881619937,
"grad_norm": 3.9375,
"learning_rate": 1.972049689440994e-05,
"loss": 2.2302,
"step": 128
},
{
"epoch": 0.8037383177570093,
"grad_norm": 3.53125,
"learning_rate": 1.9875776397515528e-05,
"loss": 2.139,
"step": 129
},
{
"epoch": 0.8099688473520249,
"grad_norm": 4.4375,
"learning_rate": 2.003105590062112e-05,
"loss": 2.1387,
"step": 130
},
{
"epoch": 0.8161993769470405,
"grad_norm": 6.65625,
"learning_rate": 2.0186335403726707e-05,
"loss": 2.1359,
"step": 131
},
{
"epoch": 0.822429906542056,
"grad_norm": 3.0625,
"learning_rate": 2.0341614906832298e-05,
"loss": 2.0794,
"step": 132
},
{
"epoch": 0.8286604361370716,
"grad_norm": 3.921875,
"learning_rate": 2.049689440993789e-05,
"loss": 2.1403,
"step": 133
},
{
"epoch": 0.8348909657320872,
"grad_norm": 3.015625,
"learning_rate": 2.065217391304348e-05,
"loss": 2.1147,
"step": 134
},
{
"epoch": 0.8411214953271028,
"grad_norm": 3.796875,
"learning_rate": 2.080745341614907e-05,
"loss": 2.1486,
"step": 135
},
{
"epoch": 0.8473520249221184,
"grad_norm": 3.1875,
"learning_rate": 2.096273291925466e-05,
"loss": 2.1352,
"step": 136
},
{
"epoch": 0.8535825545171339,
"grad_norm": 3.203125,
"learning_rate": 2.111801242236025e-05,
"loss": 2.1537,
"step": 137
},
{
"epoch": 0.8598130841121495,
"grad_norm": 3.046875,
"learning_rate": 2.127329192546584e-05,
"loss": 2.1042,
"step": 138
},
{
"epoch": 0.8660436137071651,
"grad_norm": 3.28125,
"learning_rate": 2.1428571428571428e-05,
"loss": 2.1917,
"step": 139
},
{
"epoch": 0.8722741433021807,
"grad_norm": 3.703125,
"learning_rate": 2.158385093167702e-05,
"loss": 2.1075,
"step": 140
},
{
"epoch": 0.8785046728971962,
"grad_norm": 3.34375,
"learning_rate": 2.173913043478261e-05,
"loss": 2.0968,
"step": 141
},
{
"epoch": 0.8847352024922118,
"grad_norm": 3.25,
"learning_rate": 2.1894409937888198e-05,
"loss": 2.1369,
"step": 142
},
{
"epoch": 0.8909657320872274,
"grad_norm": 3.390625,
"learning_rate": 2.204968944099379e-05,
"loss": 2.1224,
"step": 143
},
{
"epoch": 0.897196261682243,
"grad_norm": 3.25,
"learning_rate": 2.220496894409938e-05,
"loss": 2.0634,
"step": 144
},
{
"epoch": 0.9034267912772586,
"grad_norm": 4.1875,
"learning_rate": 2.236024844720497e-05,
"loss": 2.1143,
"step": 145
},
{
"epoch": 0.9096573208722741,
"grad_norm": 3.15625,
"learning_rate": 2.2515527950310562e-05,
"loss": 2.1619,
"step": 146
},
{
"epoch": 0.9158878504672897,
"grad_norm": 3.765625,
"learning_rate": 2.2670807453416153e-05,
"loss": 2.158,
"step": 147
},
{
"epoch": 0.9221183800623053,
"grad_norm": 3.46875,
"learning_rate": 2.282608695652174e-05,
"loss": 2.1727,
"step": 148
},
{
"epoch": 0.9283489096573209,
"grad_norm": 3.453125,
"learning_rate": 2.2981366459627328e-05,
"loss": 2.1031,
"step": 149
},
{
"epoch": 0.9345794392523364,
"grad_norm": 3.03125,
"learning_rate": 2.313664596273292e-05,
"loss": 2.1154,
"step": 150
},
{
"epoch": 0.940809968847352,
"grad_norm": 2.859375,
"learning_rate": 2.329192546583851e-05,
"loss": 2.0786,
"step": 151
},
{
"epoch": 0.9470404984423676,
"grad_norm": 3.171875,
"learning_rate": 2.34472049689441e-05,
"loss": 2.0949,
"step": 152
},
{
"epoch": 0.9532710280373832,
"grad_norm": 3.40625,
"learning_rate": 2.3602484472049692e-05,
"loss": 2.1483,
"step": 153
},
{
"epoch": 0.9595015576323987,
"grad_norm": 3.3125,
"learning_rate": 2.375776397515528e-05,
"loss": 2.079,
"step": 154
},
{
"epoch": 0.9657320872274143,
"grad_norm": 2.9375,
"learning_rate": 2.391304347826087e-05,
"loss": 2.0745,
"step": 155
},
{
"epoch": 0.9719626168224299,
"grad_norm": 3.328125,
"learning_rate": 2.4068322981366462e-05,
"loss": 2.0387,
"step": 156
},
{
"epoch": 0.9781931464174455,
"grad_norm": 3.15625,
"learning_rate": 2.4223602484472053e-05,
"loss": 2.1282,
"step": 157
},
{
"epoch": 0.9844236760124611,
"grad_norm": 3.109375,
"learning_rate": 2.437888198757764e-05,
"loss": 2.1016,
"step": 158
},
{
"epoch": 0.9906542056074766,
"grad_norm": 3.046875,
"learning_rate": 2.453416149068323e-05,
"loss": 2.0856,
"step": 159
},
{
"epoch": 0.9968847352024922,
"grad_norm": 2.859375,
"learning_rate": 2.468944099378882e-05,
"loss": 2.05,
"step": 160
},
{
"epoch": 1.0,
"grad_norm": 2.296875,
"learning_rate": 2.484472049689441e-05,
"loss": 1.0603,
"step": 161
},
{
"epoch": 1.0062305295950156,
"grad_norm": 3.328125,
"learning_rate": 2.5e-05,
"loss": 2.1419,
"step": 162
},
{
"epoch": 1.0124610591900312,
"grad_norm": 3.859375,
"learning_rate": 2.515527950310559e-05,
"loss": 2.0533,
"step": 163
},
{
"epoch": 1.0186915887850467,
"grad_norm": 3.359375,
"learning_rate": 2.5310559006211183e-05,
"loss": 2.0753,
"step": 164
},
{
"epoch": 1.0249221183800623,
"grad_norm": 3.296875,
"learning_rate": 2.546583850931677e-05,
"loss": 2.0986,
"step": 165
},
{
"epoch": 1.0311526479750779,
"grad_norm": 3.40625,
"learning_rate": 2.5621118012422362e-05,
"loss": 2.0875,
"step": 166
},
{
"epoch": 1.0373831775700935,
"grad_norm": 3.40625,
"learning_rate": 2.577639751552795e-05,
"loss": 2.0334,
"step": 167
},
{
"epoch": 1.043613707165109,
"grad_norm": 2.640625,
"learning_rate": 2.5931677018633544e-05,
"loss": 2.0787,
"step": 168
},
{
"epoch": 1.0498442367601246,
"grad_norm": 3.09375,
"learning_rate": 2.608695652173913e-05,
"loss": 2.0918,
"step": 169
},
{
"epoch": 1.0560747663551402,
"grad_norm": 2.796875,
"learning_rate": 2.6242236024844723e-05,
"loss": 2.0837,
"step": 170
},
{
"epoch": 1.0623052959501558,
"grad_norm": 3.21875,
"learning_rate": 2.639751552795031e-05,
"loss": 2.1337,
"step": 171
},
{
"epoch": 1.0685358255451713,
"grad_norm": 3.234375,
"learning_rate": 2.6552795031055898e-05,
"loss": 2.1146,
"step": 172
},
{
"epoch": 1.074766355140187,
"grad_norm": 2.859375,
"learning_rate": 2.6708074534161492e-05,
"loss": 2.0637,
"step": 173
},
{
"epoch": 1.0809968847352025,
"grad_norm": 3.0625,
"learning_rate": 2.686335403726708e-05,
"loss": 2.0964,
"step": 174
},
{
"epoch": 1.087227414330218,
"grad_norm": 3.40625,
"learning_rate": 2.7018633540372674e-05,
"loss": 2.0987,
"step": 175
},
{
"epoch": 1.0934579439252337,
"grad_norm": 3.140625,
"learning_rate": 2.7173913043478262e-05,
"loss": 2.1179,
"step": 176
},
{
"epoch": 1.0996884735202492,
"grad_norm": 3.53125,
"learning_rate": 2.7329192546583853e-05,
"loss": 2.0785,
"step": 177
},
{
"epoch": 1.1059190031152648,
"grad_norm": 3.109375,
"learning_rate": 2.748447204968944e-05,
"loss": 2.0614,
"step": 178
},
{
"epoch": 1.1121495327102804,
"grad_norm": 2.84375,
"learning_rate": 2.7639751552795035e-05,
"loss": 2.0249,
"step": 179
},
{
"epoch": 1.118380062305296,
"grad_norm": 2.75,
"learning_rate": 2.7795031055900623e-05,
"loss": 2.0923,
"step": 180
},
{
"epoch": 1.1246105919003115,
"grad_norm": 3.234375,
"learning_rate": 2.795031055900621e-05,
"loss": 2.0288,
"step": 181
},
{
"epoch": 1.1308411214953271,
"grad_norm": 3.34375,
"learning_rate": 2.8105590062111805e-05,
"loss": 2.0697,
"step": 182
},
{
"epoch": 1.1370716510903427,
"grad_norm": 3.109375,
"learning_rate": 2.826086956521739e-05,
"loss": 2.0496,
"step": 183
},
{
"epoch": 1.1433021806853583,
"grad_norm": 3.15625,
"learning_rate": 2.8416149068322983e-05,
"loss": 2.0604,
"step": 184
},
{
"epoch": 1.1495327102803738,
"grad_norm": 2.609375,
"learning_rate": 2.857142857142857e-05,
"loss": 2.0665,
"step": 185
},
{
"epoch": 1.1557632398753894,
"grad_norm": 3.046875,
"learning_rate": 2.8726708074534165e-05,
"loss": 2.0806,
"step": 186
},
{
"epoch": 1.161993769470405,
"grad_norm": 3.734375,
"learning_rate": 2.8881987577639753e-05,
"loss": 2.0135,
"step": 187
},
{
"epoch": 1.1682242990654206,
"grad_norm": 3.125,
"learning_rate": 2.9037267080745344e-05,
"loss": 2.0919,
"step": 188
},
{
"epoch": 1.1744548286604362,
"grad_norm": 3.09375,
"learning_rate": 2.919254658385093e-05,
"loss": 2.0705,
"step": 189
},
{
"epoch": 1.1806853582554517,
"grad_norm": 3.125,
"learning_rate": 2.9347826086956526e-05,
"loss": 2.0987,
"step": 190
},
{
"epoch": 1.1869158878504673,
"grad_norm": 3.046875,
"learning_rate": 2.9503105590062114e-05,
"loss": 2.1217,
"step": 191
},
{
"epoch": 1.1931464174454829,
"grad_norm": 3.390625,
"learning_rate": 2.96583850931677e-05,
"loss": 2.0846,
"step": 192
},
{
"epoch": 1.1993769470404985,
"grad_norm": 3.0625,
"learning_rate": 2.9813664596273296e-05,
"loss": 2.1126,
"step": 193
},
{
"epoch": 1.205607476635514,
"grad_norm": 3.78125,
"learning_rate": 2.9968944099378883e-05,
"loss": 2.0716,
"step": 194
},
{
"epoch": 1.2118380062305296,
"grad_norm": 3.125,
"learning_rate": 3.0124223602484474e-05,
"loss": 2.0856,
"step": 195
},
{
"epoch": 1.2180685358255452,
"grad_norm": 3.453125,
"learning_rate": 3.0279503105590062e-05,
"loss": 2.0801,
"step": 196
},
{
"epoch": 1.2242990654205608,
"grad_norm": 3.234375,
"learning_rate": 3.0434782608695656e-05,
"loss": 2.1112,
"step": 197
},
{
"epoch": 1.2305295950155763,
"grad_norm": 3.609375,
"learning_rate": 3.059006211180124e-05,
"loss": 2.0616,
"step": 198
},
{
"epoch": 1.236760124610592,
"grad_norm": 4.40625,
"learning_rate": 3.074534161490684e-05,
"loss": 2.0934,
"step": 199
},
{
"epoch": 1.2429906542056075,
"grad_norm": 3.140625,
"learning_rate": 3.090062111801242e-05,
"loss": 2.162,
"step": 200
},
{
"epoch": 1.249221183800623,
"grad_norm": 2.828125,
"learning_rate": 3.1055900621118014e-05,
"loss": 2.0811,
"step": 201
},
{
"epoch": 1.2554517133956387,
"grad_norm": 3.09375,
"learning_rate": 3.1211180124223605e-05,
"loss": 2.13,
"step": 202
},
{
"epoch": 1.2616822429906542,
"grad_norm": 3.375,
"learning_rate": 3.136645962732919e-05,
"loss": 2.1482,
"step": 203
},
{
"epoch": 1.2679127725856698,
"grad_norm": 2.78125,
"learning_rate": 3.152173913043479e-05,
"loss": 2.0389,
"step": 204
},
{
"epoch": 1.2741433021806854,
"grad_norm": 3.96875,
"learning_rate": 3.167701863354037e-05,
"loss": 2.0823,
"step": 205
},
{
"epoch": 1.280373831775701,
"grad_norm": 2.921875,
"learning_rate": 3.183229813664597e-05,
"loss": 2.0425,
"step": 206
},
{
"epoch": 1.2866043613707165,
"grad_norm": 3.046875,
"learning_rate": 3.198757763975155e-05,
"loss": 2.0558,
"step": 207
},
{
"epoch": 1.2928348909657321,
"grad_norm": 3.03125,
"learning_rate": 3.2142857142857144e-05,
"loss": 1.9776,
"step": 208
},
{
"epoch": 1.2990654205607477,
"grad_norm": 3.015625,
"learning_rate": 3.2298136645962735e-05,
"loss": 2.0971,
"step": 209
},
{
"epoch": 1.3052959501557633,
"grad_norm": 3.203125,
"learning_rate": 3.245341614906832e-05,
"loss": 2.1358,
"step": 210
},
{
"epoch": 1.3115264797507789,
"grad_norm": 2.953125,
"learning_rate": 3.260869565217392e-05,
"loss": 2.1331,
"step": 211
},
{
"epoch": 1.3177570093457944,
"grad_norm": 3.078125,
"learning_rate": 3.27639751552795e-05,
"loss": 2.0813,
"step": 212
},
{
"epoch": 1.32398753894081,
"grad_norm": 2.953125,
"learning_rate": 3.29192546583851e-05,
"loss": 2.0251,
"step": 213
},
{
"epoch": 1.3302180685358256,
"grad_norm": 3.125,
"learning_rate": 3.307453416149068e-05,
"loss": 2.0913,
"step": 214
},
{
"epoch": 1.3364485981308412,
"grad_norm": 3.28125,
"learning_rate": 3.3229813664596274e-05,
"loss": 1.9863,
"step": 215
},
{
"epoch": 1.3426791277258567,
"grad_norm": 2.625,
"learning_rate": 3.3385093167701865e-05,
"loss": 2.054,
"step": 216
},
{
"epoch": 1.3489096573208723,
"grad_norm": 2.984375,
"learning_rate": 3.3540372670807456e-05,
"loss": 2.0374,
"step": 217
},
{
"epoch": 1.355140186915888,
"grad_norm": 2.75,
"learning_rate": 3.369565217391305e-05,
"loss": 2.0634,
"step": 218
},
{
"epoch": 1.3613707165109035,
"grad_norm": 2.953125,
"learning_rate": 3.385093167701863e-05,
"loss": 2.1201,
"step": 219
},
{
"epoch": 1.367601246105919,
"grad_norm": 2.765625,
"learning_rate": 3.400621118012422e-05,
"loss": 2.0579,
"step": 220
},
{
"epoch": 1.3738317757009346,
"grad_norm": 2.875,
"learning_rate": 3.4161490683229814e-05,
"loss": 2.0824,
"step": 221
},
{
"epoch": 1.3800623052959502,
"grad_norm": 3.265625,
"learning_rate": 3.4316770186335405e-05,
"loss": 2.0402,
"step": 222
},
{
"epoch": 1.3862928348909658,
"grad_norm": 3.0625,
"learning_rate": 3.4472049689440996e-05,
"loss": 2.0295,
"step": 223
},
{
"epoch": 1.3925233644859814,
"grad_norm": 2.84375,
"learning_rate": 3.462732919254659e-05,
"loss": 2.0156,
"step": 224
},
{
"epoch": 1.398753894080997,
"grad_norm": 2.640625,
"learning_rate": 3.478260869565218e-05,
"loss": 2.0881,
"step": 225
},
{
"epoch": 1.4049844236760125,
"grad_norm": 3.21875,
"learning_rate": 3.493788819875777e-05,
"loss": 2.0587,
"step": 226
},
{
"epoch": 1.411214953271028,
"grad_norm": 3.25,
"learning_rate": 3.509316770186335e-05,
"loss": 2.0893,
"step": 227
},
{
"epoch": 1.4174454828660437,
"grad_norm": 3.109375,
"learning_rate": 3.524844720496895e-05,
"loss": 2.1113,
"step": 228
},
{
"epoch": 1.4236760124610592,
"grad_norm": 3.34375,
"learning_rate": 3.5403726708074535e-05,
"loss": 2.0261,
"step": 229
},
{
"epoch": 1.4299065420560748,
"grad_norm": 4.90625,
"learning_rate": 3.5559006211180126e-05,
"loss": 2.0266,
"step": 230
},
{
"epoch": 1.4361370716510904,
"grad_norm": 3.0625,
"learning_rate": 3.571428571428572e-05,
"loss": 2.0307,
"step": 231
},
{
"epoch": 1.442367601246106,
"grad_norm": 3.40625,
"learning_rate": 3.58695652173913e-05,
"loss": 2.1302,
"step": 232
},
{
"epoch": 1.4485981308411215,
"grad_norm": 3.484375,
"learning_rate": 3.60248447204969e-05,
"loss": 2.0454,
"step": 233
},
{
"epoch": 1.4548286604361371,
"grad_norm": 2.859375,
"learning_rate": 3.618012422360248e-05,
"loss": 2.0382,
"step": 234
},
{
"epoch": 1.4610591900311527,
"grad_norm": 3.53125,
"learning_rate": 3.633540372670808e-05,
"loss": 2.077,
"step": 235
},
{
"epoch": 1.4672897196261683,
"grad_norm": 4.4375,
"learning_rate": 3.6490683229813665e-05,
"loss": 1.9795,
"step": 236
},
{
"epoch": 1.4735202492211839,
"grad_norm": 3.90625,
"learning_rate": 3.6645962732919256e-05,
"loss": 2.0261,
"step": 237
},
{
"epoch": 1.4797507788161994,
"grad_norm": 2.984375,
"learning_rate": 3.680124223602485e-05,
"loss": 2.0306,
"step": 238
},
{
"epoch": 1.485981308411215,
"grad_norm": 2.875,
"learning_rate": 3.695652173913043e-05,
"loss": 2.0003,
"step": 239
},
{
"epoch": 1.4922118380062306,
"grad_norm": 2.953125,
"learning_rate": 3.711180124223603e-05,
"loss": 2.0031,
"step": 240
},
{
"epoch": 1.4984423676012462,
"grad_norm": 3.171875,
"learning_rate": 3.7267080745341614e-05,
"loss": 2.0533,
"step": 241
},
{
"epoch": 1.5046728971962615,
"grad_norm": 3.421875,
"learning_rate": 3.742236024844721e-05,
"loss": 2.0356,
"step": 242
},
{
"epoch": 1.5109034267912773,
"grad_norm": 3.296875,
"learning_rate": 3.7577639751552796e-05,
"loss": 2.0567,
"step": 243
},
{
"epoch": 1.5171339563862927,
"grad_norm": 3.21875,
"learning_rate": 3.773291925465839e-05,
"loss": 2.0419,
"step": 244
},
{
"epoch": 1.5233644859813085,
"grad_norm": 3.328125,
"learning_rate": 3.788819875776398e-05,
"loss": 2.031,
"step": 245
},
{
"epoch": 1.5295950155763238,
"grad_norm": 3.109375,
"learning_rate": 3.804347826086957e-05,
"loss": 1.9832,
"step": 246
},
{
"epoch": 1.5358255451713396,
"grad_norm": 2.96875,
"learning_rate": 3.819875776397516e-05,
"loss": 2.018,
"step": 247
},
{
"epoch": 1.542056074766355,
"grad_norm": 2.46875,
"learning_rate": 3.8354037267080744e-05,
"loss": 2.0755,
"step": 248
},
{
"epoch": 1.5482866043613708,
"grad_norm": 3.046875,
"learning_rate": 3.8509316770186335e-05,
"loss": 1.9826,
"step": 249
},
{
"epoch": 1.5545171339563861,
"grad_norm": 3.875,
"learning_rate": 3.8664596273291926e-05,
"loss": 2.0682,
"step": 250
},
{
"epoch": 1.560747663551402,
"grad_norm": 3.15625,
"learning_rate": 3.881987577639752e-05,
"loss": 2.0588,
"step": 251
},
{
"epoch": 1.5669781931464173,
"grad_norm": 3.484375,
"learning_rate": 3.897515527950311e-05,
"loss": 2.0599,
"step": 252
},
{
"epoch": 1.573208722741433,
"grad_norm": 3.640625,
"learning_rate": 3.91304347826087e-05,
"loss": 1.9605,
"step": 253
},
{
"epoch": 1.5794392523364484,
"grad_norm": 3.0,
"learning_rate": 3.928571428571429e-05,
"loss": 2.0104,
"step": 254
},
{
"epoch": 1.5856697819314642,
"grad_norm": 3.265625,
"learning_rate": 3.944099378881988e-05,
"loss": 2.0551,
"step": 255
},
{
"epoch": 1.5919003115264796,
"grad_norm": 3.125,
"learning_rate": 3.9596273291925465e-05,
"loss": 2.0468,
"step": 256
},
{
"epoch": 1.5981308411214954,
"grad_norm": 3.421875,
"learning_rate": 3.9751552795031056e-05,
"loss": 2.02,
"step": 257
},
{
"epoch": 1.6043613707165107,
"grad_norm": 3.21875,
"learning_rate": 3.990683229813665e-05,
"loss": 2.0149,
"step": 258
},
{
"epoch": 1.6105919003115265,
"grad_norm": 3.671875,
"learning_rate": 4.006211180124224e-05,
"loss": 2.0387,
"step": 259
},
{
"epoch": 1.616822429906542,
"grad_norm": 2.921875,
"learning_rate": 4.021739130434783e-05,
"loss": 1.9553,
"step": 260
},
{
"epoch": 1.6230529595015577,
"grad_norm": 2.9375,
"learning_rate": 4.0372670807453414e-05,
"loss": 2.0293,
"step": 261
},
{
"epoch": 1.629283489096573,
"grad_norm": 3.0625,
"learning_rate": 4.052795031055901e-05,
"loss": 1.9883,
"step": 262
},
{
"epoch": 1.6355140186915889,
"grad_norm": 3.203125,
"learning_rate": 4.0683229813664596e-05,
"loss": 2.0252,
"step": 263
},
{
"epoch": 1.6417445482866042,
"grad_norm": 3.15625,
"learning_rate": 4.0838509316770193e-05,
"loss": 1.9663,
"step": 264
},
{
"epoch": 1.64797507788162,
"grad_norm": 3.4375,
"learning_rate": 4.099378881987578e-05,
"loss": 2.0354,
"step": 265
},
{
"epoch": 1.6542056074766354,
"grad_norm": 3.90625,
"learning_rate": 4.114906832298137e-05,
"loss": 2.0284,
"step": 266
},
{
"epoch": 1.6604361370716512,
"grad_norm": 4.1875,
"learning_rate": 4.130434782608696e-05,
"loss": 2.0113,
"step": 267
},
{
"epoch": 1.6666666666666665,
"grad_norm": 3.9375,
"learning_rate": 4.1459627329192544e-05,
"loss": 2.0592,
"step": 268
},
{
"epoch": 1.6728971962616823,
"grad_norm": 3.34375,
"learning_rate": 4.161490683229814e-05,
"loss": 2.023,
"step": 269
},
{
"epoch": 1.6791277258566977,
"grad_norm": 3.8125,
"learning_rate": 4.1770186335403726e-05,
"loss": 1.9804,
"step": 270
},
{
"epoch": 1.6853582554517135,
"grad_norm": 3.328125,
"learning_rate": 4.192546583850932e-05,
"loss": 1.9465,
"step": 271
},
{
"epoch": 1.6915887850467288,
"grad_norm": 3.484375,
"learning_rate": 4.208074534161491e-05,
"loss": 1.9713,
"step": 272
},
{
"epoch": 1.6978193146417446,
"grad_norm": 3.46875,
"learning_rate": 4.22360248447205e-05,
"loss": 2.0509,
"step": 273
},
{
"epoch": 1.70404984423676,
"grad_norm": 3.3125,
"learning_rate": 4.239130434782609e-05,
"loss": 1.997,
"step": 274
},
{
"epoch": 1.7102803738317758,
"grad_norm": 3.328125,
"learning_rate": 4.254658385093168e-05,
"loss": 2.0374,
"step": 275
},
{
"epoch": 1.7165109034267911,
"grad_norm": 3.625,
"learning_rate": 4.270186335403727e-05,
"loss": 2.0804,
"step": 276
},
{
"epoch": 1.722741433021807,
"grad_norm": 3.1875,
"learning_rate": 4.2857142857142856e-05,
"loss": 1.9943,
"step": 277
},
{
"epoch": 1.7289719626168223,
"grad_norm": 3.171875,
"learning_rate": 4.301242236024845e-05,
"loss": 2.0213,
"step": 278
},
{
"epoch": 1.735202492211838,
"grad_norm": 3.3125,
"learning_rate": 4.316770186335404e-05,
"loss": 2.0074,
"step": 279
},
{
"epoch": 1.7414330218068534,
"grad_norm": 3.578125,
"learning_rate": 4.332298136645963e-05,
"loss": 2.0221,
"step": 280
},
{
"epoch": 1.7476635514018692,
"grad_norm": 3.703125,
"learning_rate": 4.347826086956522e-05,
"loss": 2.0185,
"step": 281
},
{
"epoch": 1.7538940809968846,
"grad_norm": 3.421875,
"learning_rate": 4.363354037267081e-05,
"loss": 2.0309,
"step": 282
},
{
"epoch": 1.7601246105919004,
"grad_norm": 3.609375,
"learning_rate": 4.3788819875776396e-05,
"loss": 1.9929,
"step": 283
},
{
"epoch": 1.7663551401869158,
"grad_norm": 3.65625,
"learning_rate": 4.3944099378881993e-05,
"loss": 1.8966,
"step": 284
},
{
"epoch": 1.7725856697819315,
"grad_norm": 3.21875,
"learning_rate": 4.409937888198758e-05,
"loss": 1.9699,
"step": 285
},
{
"epoch": 1.778816199376947,
"grad_norm": 3.9375,
"learning_rate": 4.425465838509317e-05,
"loss": 1.9846,
"step": 286
},
{
"epoch": 1.7850467289719627,
"grad_norm": 3.3125,
"learning_rate": 4.440993788819876e-05,
"loss": 2.0081,
"step": 287
},
{
"epoch": 1.791277258566978,
"grad_norm": 3.015625,
"learning_rate": 4.456521739130435e-05,
"loss": 1.9469,
"step": 288
},
{
"epoch": 1.7975077881619939,
"grad_norm": 3.9375,
"learning_rate": 4.472049689440994e-05,
"loss": 1.9745,
"step": 289
},
{
"epoch": 1.8037383177570092,
"grad_norm": 3.671875,
"learning_rate": 4.4875776397515526e-05,
"loss": 2.019,
"step": 290
},
{
"epoch": 1.809968847352025,
"grad_norm": 3.984375,
"learning_rate": 4.5031055900621124e-05,
"loss": 1.915,
"step": 291
},
{
"epoch": 1.8161993769470404,
"grad_norm": 3.375,
"learning_rate": 4.518633540372671e-05,
"loss": 1.9832,
"step": 292
},
{
"epoch": 1.8224299065420562,
"grad_norm": 3.96875,
"learning_rate": 4.5341614906832306e-05,
"loss": 1.9723,
"step": 293
},
{
"epoch": 1.8286604361370715,
"grad_norm": 3.515625,
"learning_rate": 4.549689440993789e-05,
"loss": 1.9366,
"step": 294
},
{
"epoch": 1.8348909657320873,
"grad_norm": 3.765625,
"learning_rate": 4.565217391304348e-05,
"loss": 2.0083,
"step": 295
},
{
"epoch": 1.8411214953271027,
"grad_norm": 3.796875,
"learning_rate": 4.580745341614907e-05,
"loss": 1.9305,
"step": 296
},
{
"epoch": 1.8473520249221185,
"grad_norm": 4.21875,
"learning_rate": 4.5962732919254656e-05,
"loss": 1.951,
"step": 297
},
{
"epoch": 1.8535825545171338,
"grad_norm": 3.625,
"learning_rate": 4.6118012422360254e-05,
"loss": 1.9282,
"step": 298
},
{
"epoch": 1.8598130841121496,
"grad_norm": 3.671875,
"learning_rate": 4.627329192546584e-05,
"loss": 1.9599,
"step": 299
},
{
"epoch": 1.866043613707165,
"grad_norm": 3.875,
"learning_rate": 4.642857142857143e-05,
"loss": 1.9801,
"step": 300
},
{
"epoch": 1.8722741433021808,
"grad_norm": 4.40625,
"learning_rate": 4.658385093167702e-05,
"loss": 1.9462,
"step": 301
},
{
"epoch": 1.8785046728971961,
"grad_norm": 3.765625,
"learning_rate": 4.673913043478261e-05,
"loss": 2.0157,
"step": 302
},
{
"epoch": 1.884735202492212,
"grad_norm": 3.765625,
"learning_rate": 4.68944099378882e-05,
"loss": 1.939,
"step": 303
},
{
"epoch": 1.8909657320872273,
"grad_norm": 4.78125,
"learning_rate": 4.7049689440993793e-05,
"loss": 1.9591,
"step": 304
},
{
"epoch": 1.897196261682243,
"grad_norm": 6.53125,
"learning_rate": 4.7204968944099384e-05,
"loss": 1.9056,
"step": 305
},
{
"epoch": 1.9034267912772584,
"grad_norm": 4.03125,
"learning_rate": 4.736024844720497e-05,
"loss": 1.9685,
"step": 306
},
{
"epoch": 1.9096573208722742,
"grad_norm": 5.5625,
"learning_rate": 4.751552795031056e-05,
"loss": 1.9851,
"step": 307
},
{
"epoch": 1.9158878504672896,
"grad_norm": 3.9375,
"learning_rate": 4.767080745341615e-05,
"loss": 1.9225,
"step": 308
},
{
"epoch": 1.9221183800623054,
"grad_norm": 4.15625,
"learning_rate": 4.782608695652174e-05,
"loss": 1.9633,
"step": 309
},
{
"epoch": 1.9283489096573208,
"grad_norm": 4.40625,
"learning_rate": 4.798136645962733e-05,
"loss": 1.8659,
"step": 310
},
{
"epoch": 1.9345794392523366,
"grad_norm": 4.09375,
"learning_rate": 4.8136645962732924e-05,
"loss": 1.9911,
"step": 311
},
{
"epoch": 1.940809968847352,
"grad_norm": 4.09375,
"learning_rate": 4.829192546583851e-05,
"loss": 1.8502,
"step": 312
},
{
"epoch": 1.9470404984423677,
"grad_norm": 4.28125,
"learning_rate": 4.8447204968944106e-05,
"loss": 1.9252,
"step": 313
},
{
"epoch": 1.953271028037383,
"grad_norm": 3.90625,
"learning_rate": 4.860248447204969e-05,
"loss": 1.8731,
"step": 314
},
{
"epoch": 1.9595015576323989,
"grad_norm": 3.640625,
"learning_rate": 4.875776397515528e-05,
"loss": 1.8565,
"step": 315
},
{
"epoch": 1.9657320872274142,
"grad_norm": 5.75,
"learning_rate": 4.891304347826087e-05,
"loss": 1.8848,
"step": 316
},
{
"epoch": 1.97196261682243,
"grad_norm": 4.09375,
"learning_rate": 4.906832298136646e-05,
"loss": 1.9044,
"step": 317
},
{
"epoch": 1.9781931464174454,
"grad_norm": 4.375,
"learning_rate": 4.9223602484472054e-05,
"loss": 1.8283,
"step": 318
},
{
"epoch": 1.9844236760124612,
"grad_norm": 3.84375,
"learning_rate": 4.937888198757764e-05,
"loss": 1.8427,
"step": 319
},
{
"epoch": 1.9906542056074765,
"grad_norm": 4.21875,
"learning_rate": 4.9534161490683236e-05,
"loss": 1.9034,
"step": 320
},
{
"epoch": 1.9968847352024923,
"grad_norm": 4.5,
"learning_rate": 4.968944099378882e-05,
"loss": 1.8454,
"step": 321
},
{
"epoch": 2.0,
"grad_norm": 2.796875,
"learning_rate": 4.984472049689442e-05,
"loss": 0.917,
"step": 322
},
{
"epoch": 2.0062305295950154,
"grad_norm": 4.125,
"learning_rate": 5e-05,
"loss": 1.8594,
"step": 323
},
{
"epoch": 2.012461059190031,
"grad_norm": 4.84375,
"learning_rate": 4.99999998785975e-05,
"loss": 1.8636,
"step": 324
},
{
"epoch": 2.0186915887850465,
"grad_norm": 5.09375,
"learning_rate": 4.999999951439e-05,
"loss": 1.8519,
"step": 325
},
{
"epoch": 2.0249221183800623,
"grad_norm": 5.5,
"learning_rate": 4.999999890737752e-05,
"loss": 1.9552,
"step": 326
},
{
"epoch": 2.0311526479750777,
"grad_norm": 4.65625,
"learning_rate": 4.9999998057560046e-05,
"loss": 1.8387,
"step": 327
},
{
"epoch": 2.0373831775700935,
"grad_norm": 4.46875,
"learning_rate": 4.999999696493759e-05,
"loss": 1.8041,
"step": 328
},
{
"epoch": 2.043613707165109,
"grad_norm": 4.96875,
"learning_rate": 4.999999562951018e-05,
"loss": 1.7945,
"step": 329
},
{
"epoch": 2.0498442367601246,
"grad_norm": 4.46875,
"learning_rate": 4.99999940512778e-05,
"loss": 1.8101,
"step": 330
},
{
"epoch": 2.05607476635514,
"grad_norm": 4.84375,
"learning_rate": 4.9999992230240476e-05,
"loss": 1.8855,
"step": 331
},
{
"epoch": 2.0623052959501558,
"grad_norm": 4.71875,
"learning_rate": 4.999999016639825e-05,
"loss": 1.7228,
"step": 332
},
{
"epoch": 2.068535825545171,
"grad_norm": 4.21875,
"learning_rate": 4.9999987859751106e-05,
"loss": 1.7644,
"step": 333
},
{
"epoch": 2.074766355140187,
"grad_norm": 7.46875,
"learning_rate": 4.999998531029908e-05,
"loss": 1.8284,
"step": 334
},
{
"epoch": 2.0809968847352023,
"grad_norm": 4.90625,
"learning_rate": 4.999998251804221e-05,
"loss": 1.7605,
"step": 335
},
{
"epoch": 2.087227414330218,
"grad_norm": 4.53125,
"learning_rate": 4.999997948298051e-05,
"loss": 1.7437,
"step": 336
},
{
"epoch": 2.0934579439252334,
"grad_norm": 4.59375,
"learning_rate": 4.999997620511401e-05,
"loss": 1.694,
"step": 337
},
{
"epoch": 2.0996884735202492,
"grad_norm": 5.65625,
"learning_rate": 4.999997268444275e-05,
"loss": 1.7563,
"step": 338
},
{
"epoch": 2.1059190031152646,
"grad_norm": 5.3125,
"learning_rate": 4.999996892096675e-05,
"loss": 1.715,
"step": 339
},
{
"epoch": 2.1121495327102804,
"grad_norm": 4.71875,
"learning_rate": 4.999996491468606e-05,
"loss": 1.7149,
"step": 340
},
{
"epoch": 2.1183800623052957,
"grad_norm": 4.75,
"learning_rate": 4.999996066560071e-05,
"loss": 1.7624,
"step": 341
},
{
"epoch": 2.1246105919003115,
"grad_norm": 4.5,
"learning_rate": 4.999995617371074e-05,
"loss": 1.7721,
"step": 342
},
{
"epoch": 2.130841121495327,
"grad_norm": 4.90625,
"learning_rate": 4.99999514390162e-05,
"loss": 1.7928,
"step": 343
},
{
"epoch": 2.1370716510903427,
"grad_norm": 5.09375,
"learning_rate": 4.9999946461517144e-05,
"loss": 1.6423,
"step": 344
},
{
"epoch": 2.143302180685358,
"grad_norm": 5.34375,
"learning_rate": 4.99999412412136e-05,
"loss": 1.6847,
"step": 345
},
{
"epoch": 2.149532710280374,
"grad_norm": 5.25,
"learning_rate": 4.999993577810563e-05,
"loss": 1.6455,
"step": 346
},
{
"epoch": 2.155763239875389,
"grad_norm": 5.40625,
"learning_rate": 4.999993007219329e-05,
"loss": 1.6867,
"step": 347
},
{
"epoch": 2.161993769470405,
"grad_norm": 6.90625,
"learning_rate": 4.9999924123476636e-05,
"loss": 1.6675,
"step": 348
},
{
"epoch": 2.1682242990654204,
"grad_norm": 5.5625,
"learning_rate": 4.999991793195572e-05,
"loss": 1.6571,
"step": 349
},
{
"epoch": 2.174454828660436,
"grad_norm": 5.71875,
"learning_rate": 4.9999911497630595e-05,
"loss": 1.6956,
"step": 350
},
{
"epoch": 2.1806853582554515,
"grad_norm": 5.96875,
"learning_rate": 4.999990482050134e-05,
"loss": 1.6936,
"step": 351
},
{
"epoch": 2.1869158878504673,
"grad_norm": 5.0,
"learning_rate": 4.9999897900568005e-05,
"loss": 1.641,
"step": 352
},
{
"epoch": 2.1931464174454827,
"grad_norm": 5.28125,
"learning_rate": 4.999989073783067e-05,
"loss": 1.4915,
"step": 353
},
{
"epoch": 2.1993769470404985,
"grad_norm": 5.09375,
"learning_rate": 4.99998833322894e-05,
"loss": 1.4548,
"step": 354
},
{
"epoch": 2.205607476635514,
"grad_norm": 5.25,
"learning_rate": 4.999987568394426e-05,
"loss": 1.5147,
"step": 355
},
{
"epoch": 2.2118380062305296,
"grad_norm": 5.25,
"learning_rate": 4.999986779279533e-05,
"loss": 1.6349,
"step": 356
},
{
"epoch": 2.218068535825545,
"grad_norm": 5.71875,
"learning_rate": 4.9999859658842694e-05,
"loss": 1.5987,
"step": 357
},
{
"epoch": 2.2242990654205608,
"grad_norm": 6.21875,
"learning_rate": 4.999985128208642e-05,
"loss": 1.5206,
"step": 358
},
{
"epoch": 2.230529595015576,
"grad_norm": 5.34375,
"learning_rate": 4.9999842662526596e-05,
"loss": 1.636,
"step": 359
},
{
"epoch": 2.236760124610592,
"grad_norm": 5.59375,
"learning_rate": 4.99998338001633e-05,
"loss": 1.6139,
"step": 360
},
{
"epoch": 2.2429906542056073,
"grad_norm": 5.96875,
"learning_rate": 4.999982469499662e-05,
"loss": 1.6658,
"step": 361
},
{
"epoch": 2.249221183800623,
"grad_norm": 5.40625,
"learning_rate": 4.9999815347026644e-05,
"loss": 1.5206,
"step": 362
},
{
"epoch": 2.2554517133956384,
"grad_norm": 5.5625,
"learning_rate": 4.999980575625346e-05,
"loss": 1.4304,
"step": 363
},
{
"epoch": 2.2616822429906542,
"grad_norm": 7.71875,
"learning_rate": 4.999979592267717e-05,
"loss": 1.6273,
"step": 364
},
{
"epoch": 2.2679127725856696,
"grad_norm": 5.6875,
"learning_rate": 4.999978584629787e-05,
"loss": 1.3524,
"step": 365
},
{
"epoch": 2.2741433021806854,
"grad_norm": 6.90625,
"learning_rate": 4.9999775527115644e-05,
"loss": 1.3856,
"step": 366
},
{
"epoch": 2.2803738317757007,
"grad_norm": 6.875,
"learning_rate": 4.99997649651306e-05,
"loss": 1.3359,
"step": 367
},
{
"epoch": 2.2866043613707165,
"grad_norm": 6.75,
"learning_rate": 4.999975416034285e-05,
"loss": 1.42,
"step": 368
},
{
"epoch": 2.292834890965732,
"grad_norm": 5.8125,
"learning_rate": 4.9999743112752485e-05,
"loss": 1.3416,
"step": 369
},
{
"epoch": 2.2990654205607477,
"grad_norm": 5.875,
"learning_rate": 4.9999731822359616e-05,
"loss": 1.3344,
"step": 370
},
{
"epoch": 2.305295950155763,
"grad_norm": 5.71875,
"learning_rate": 4.999972028916436e-05,
"loss": 1.2344,
"step": 371
},
{
"epoch": 2.311526479750779,
"grad_norm": 6.03125,
"learning_rate": 4.9999708513166824e-05,
"loss": 1.4772,
"step": 372
},
{
"epoch": 2.317757009345794,
"grad_norm": 6.25,
"learning_rate": 4.999969649436711e-05,
"loss": 1.5601,
"step": 373
},
{
"epoch": 2.32398753894081,
"grad_norm": 5.75,
"learning_rate": 4.9999684232765357e-05,
"loss": 1.3735,
"step": 374
},
{
"epoch": 2.3302180685358254,
"grad_norm": 7.96875,
"learning_rate": 4.999967172836167e-05,
"loss": 1.323,
"step": 375
},
{
"epoch": 2.336448598130841,
"grad_norm": 6.0625,
"learning_rate": 4.999965898115617e-05,
"loss": 1.227,
"step": 376
},
{
"epoch": 2.3426791277258565,
"grad_norm": 5.53125,
"learning_rate": 4.9999645991148994e-05,
"loss": 1.2507,
"step": 377
},
{
"epoch": 2.3489096573208723,
"grad_norm": 6.40625,
"learning_rate": 4.999963275834025e-05,
"loss": 1.4622,
"step": 378
},
{
"epoch": 2.3551401869158877,
"grad_norm": 5.71875,
"learning_rate": 4.9999619282730084e-05,
"loss": 1.2182,
"step": 379
},
{
"epoch": 2.3613707165109035,
"grad_norm": 5.71875,
"learning_rate": 4.9999605564318606e-05,
"loss": 1.4055,
"step": 380
},
{
"epoch": 2.367601246105919,
"grad_norm": 5.3125,
"learning_rate": 4.999959160310597e-05,
"loss": 1.1462,
"step": 381
},
{
"epoch": 2.3738317757009346,
"grad_norm": 5.84375,
"learning_rate": 4.9999577399092304e-05,
"loss": 1.1671,
"step": 382
},
{
"epoch": 2.38006230529595,
"grad_norm": 5.65625,
"learning_rate": 4.999956295227774e-05,
"loss": 1.1832,
"step": 383
},
{
"epoch": 2.3862928348909658,
"grad_norm": 5.78125,
"learning_rate": 4.999954826266242e-05,
"loss": 1.3127,
"step": 384
},
{
"epoch": 2.392523364485981,
"grad_norm": 5.625,
"learning_rate": 4.99995333302465e-05,
"loss": 1.2608,
"step": 385
},
{
"epoch": 2.398753894080997,
"grad_norm": 6.28125,
"learning_rate": 4.999951815503011e-05,
"loss": 1.3688,
"step": 386
},
{
"epoch": 2.4049844236760123,
"grad_norm": 5.46875,
"learning_rate": 4.99995027370134e-05,
"loss": 1.2532,
"step": 387
},
{
"epoch": 2.411214953271028,
"grad_norm": 7.46875,
"learning_rate": 4.999948707619653e-05,
"loss": 1.6053,
"step": 388
},
{
"epoch": 2.4174454828660434,
"grad_norm": 6.03125,
"learning_rate": 4.999947117257964e-05,
"loss": 1.1827,
"step": 389
},
{
"epoch": 2.4236760124610592,
"grad_norm": 5.8125,
"learning_rate": 4.999945502616289e-05,
"loss": 1.3497,
"step": 390
},
{
"epoch": 2.4299065420560746,
"grad_norm": 7.09375,
"learning_rate": 4.9999438636946427e-05,
"loss": 1.2837,
"step": 391
},
{
"epoch": 2.4361370716510904,
"grad_norm": 6.15625,
"learning_rate": 4.999942200493043e-05,
"loss": 1.061,
"step": 392
},
{
"epoch": 2.4423676012461057,
"grad_norm": 5.78125,
"learning_rate": 4.999940513011504e-05,
"loss": 1.2833,
"step": 393
},
{
"epoch": 2.4485981308411215,
"grad_norm": 5.8125,
"learning_rate": 4.999938801250044e-05,
"loss": 1.0304,
"step": 394
},
{
"epoch": 2.454828660436137,
"grad_norm": 5.59375,
"learning_rate": 4.9999370652086775e-05,
"loss": 1.2047,
"step": 395
},
{
"epoch": 2.4610591900311527,
"grad_norm": 6.75,
"learning_rate": 4.9999353048874234e-05,
"loss": 1.1018,
"step": 396
},
{
"epoch": 2.467289719626168,
"grad_norm": 6.0,
"learning_rate": 4.999933520286297e-05,
"loss": 1.2562,
"step": 397
},
{
"epoch": 2.473520249221184,
"grad_norm": 5.65625,
"learning_rate": 4.999931711405318e-05,
"loss": 1.1375,
"step": 398
},
{
"epoch": 2.479750778816199,
"grad_norm": 6.59375,
"learning_rate": 4.9999298782445007e-05,
"loss": 1.1745,
"step": 399
},
{
"epoch": 2.485981308411215,
"grad_norm": 6.125,
"learning_rate": 4.999928020803866e-05,
"loss": 1.0651,
"step": 400
},
{
"epoch": 2.4922118380062304,
"grad_norm": 5.4375,
"learning_rate": 4.99992613908343e-05,
"loss": 1.1111,
"step": 401
},
{
"epoch": 2.498442367601246,
"grad_norm": 6.8125,
"learning_rate": 4.999924233083212e-05,
"loss": 1.1664,
"step": 402
},
{
"epoch": 2.5046728971962615,
"grad_norm": 5.1875,
"learning_rate": 4.9999223028032295e-05,
"loss": 0.9779,
"step": 403
},
{
"epoch": 2.5109034267912773,
"grad_norm": 6.0,
"learning_rate": 4.9999203482435024e-05,
"loss": 1.0657,
"step": 404
},
{
"epoch": 2.5171339563862927,
"grad_norm": 5.6875,
"learning_rate": 4.999918369404049e-05,
"loss": 1.1241,
"step": 405
},
{
"epoch": 2.5233644859813085,
"grad_norm": 6.15625,
"learning_rate": 4.999916366284889e-05,
"loss": 1.2425,
"step": 406
},
{
"epoch": 2.529595015576324,
"grad_norm": 5.40625,
"learning_rate": 4.999914338886042e-05,
"loss": 1.163,
"step": 407
},
{
"epoch": 2.5358255451713396,
"grad_norm": 5.125,
"learning_rate": 4.999912287207525e-05,
"loss": 1.0524,
"step": 408
},
{
"epoch": 2.542056074766355,
"grad_norm": 5.84375,
"learning_rate": 4.9999102112493625e-05,
"loss": 1.1676,
"step": 409
},
{
"epoch": 2.5482866043613708,
"grad_norm": 5.625,
"learning_rate": 4.999908111011571e-05,
"loss": 1.0219,
"step": 410
},
{
"epoch": 2.554517133956386,
"grad_norm": 5.625,
"learning_rate": 4.9999059864941726e-05,
"loss": 1.0881,
"step": 411
},
{
"epoch": 2.560747663551402,
"grad_norm": 6.25,
"learning_rate": 4.9999038376971875e-05,
"loss": 1.1802,
"step": 412
},
{
"epoch": 2.5669781931464173,
"grad_norm": 5.84375,
"learning_rate": 4.9999016646206366e-05,
"loss": 1.1766,
"step": 413
},
{
"epoch": 2.573208722741433,
"grad_norm": 5.59375,
"learning_rate": 4.9998994672645415e-05,
"loss": 1.0766,
"step": 414
},
{
"epoch": 2.5794392523364484,
"grad_norm": 5.46875,
"learning_rate": 4.9998972456289226e-05,
"loss": 1.0961,
"step": 415
},
{
"epoch": 2.5856697819314642,
"grad_norm": 5.4375,
"learning_rate": 4.9998949997138015e-05,
"loss": 1.0148,
"step": 416
},
{
"epoch": 2.5919003115264796,
"grad_norm": 5.21875,
"learning_rate": 4.999892729519201e-05,
"loss": 1.0775,
"step": 417
},
{
"epoch": 2.5981308411214954,
"grad_norm": 6.125,
"learning_rate": 4.999890435045142e-05,
"loss": 0.8921,
"step": 418
},
{
"epoch": 2.6043613707165107,
"grad_norm": 5.9375,
"learning_rate": 4.999888116291649e-05,
"loss": 1.0866,
"step": 419
},
{
"epoch": 2.6105919003115265,
"grad_norm": 5.28125,
"learning_rate": 4.9998857732587414e-05,
"loss": 0.8889,
"step": 420
},
{
"epoch": 2.616822429906542,
"grad_norm": 5.65625,
"learning_rate": 4.9998834059464436e-05,
"loss": 1.0615,
"step": 421
},
{
"epoch": 2.6230529595015577,
"grad_norm": 5.46875,
"learning_rate": 4.999881014354779e-05,
"loss": 1.0774,
"step": 422
},
{
"epoch": 2.629283489096573,
"grad_norm": 6.0625,
"learning_rate": 4.99987859848377e-05,
"loss": 1.134,
"step": 423
},
{
"epoch": 2.635514018691589,
"grad_norm": 5.53125,
"learning_rate": 4.99987615833344e-05,
"loss": 0.9071,
"step": 424
},
{
"epoch": 2.641744548286604,
"grad_norm": 6.1875,
"learning_rate": 4.999873693903814e-05,
"loss": 0.9826,
"step": 425
},
{
"epoch": 2.64797507788162,
"grad_norm": 5.71875,
"learning_rate": 4.999871205194914e-05,
"loss": 1.154,
"step": 426
},
{
"epoch": 2.6542056074766354,
"grad_norm": 6.78125,
"learning_rate": 4.999868692206765e-05,
"loss": 1.031,
"step": 427
},
{
"epoch": 2.660436137071651,
"grad_norm": 5.71875,
"learning_rate": 4.9998661549393924e-05,
"loss": 0.9096,
"step": 428
},
{
"epoch": 2.6666666666666665,
"grad_norm": 5.21875,
"learning_rate": 4.9998635933928196e-05,
"loss": 0.8355,
"step": 429
},
{
"epoch": 2.6728971962616823,
"grad_norm": 7.125,
"learning_rate": 4.999861007567072e-05,
"loss": 1.2212,
"step": 430
},
{
"epoch": 2.6791277258566977,
"grad_norm": 5.59375,
"learning_rate": 4.9998583974621736e-05,
"loss": 0.9916,
"step": 431
},
{
"epoch": 2.6853582554517135,
"grad_norm": 5.625,
"learning_rate": 4.9998557630781514e-05,
"loss": 1.0815,
"step": 432
},
{
"epoch": 2.691588785046729,
"grad_norm": 5.75,
"learning_rate": 4.999853104415031e-05,
"loss": 0.8154,
"step": 433
},
{
"epoch": 2.6978193146417446,
"grad_norm": 6.03125,
"learning_rate": 4.999850421472837e-05,
"loss": 1.0478,
"step": 434
},
{
"epoch": 2.70404984423676,
"grad_norm": 6.0625,
"learning_rate": 4.999847714251596e-05,
"loss": 0.8922,
"step": 435
},
{
"epoch": 2.710280373831776,
"grad_norm": 6.21875,
"learning_rate": 4.9998449827513344e-05,
"loss": 0.9921,
"step": 436
},
{
"epoch": 2.716510903426791,
"grad_norm": 5.375,
"learning_rate": 4.999842226972078e-05,
"loss": 1.012,
"step": 437
},
{
"epoch": 2.722741433021807,
"grad_norm": 4.8125,
"learning_rate": 4.999839446913855e-05,
"loss": 0.8552,
"step": 438
},
{
"epoch": 2.7289719626168223,
"grad_norm": 5.71875,
"learning_rate": 4.999836642576692e-05,
"loss": 0.9812,
"step": 439
},
{
"epoch": 2.735202492211838,
"grad_norm": 5.09375,
"learning_rate": 4.999833813960615e-05,
"loss": 0.9391,
"step": 440
},
{
"epoch": 2.7414330218068534,
"grad_norm": 5.3125,
"learning_rate": 4.999830961065652e-05,
"loss": 0.9238,
"step": 441
},
{
"epoch": 2.7476635514018692,
"grad_norm": 5.03125,
"learning_rate": 4.999828083891831e-05,
"loss": 0.9852,
"step": 442
},
{
"epoch": 2.7538940809968846,
"grad_norm": 5.34375,
"learning_rate": 4.999825182439181e-05,
"loss": 0.9494,
"step": 443
},
{
"epoch": 2.7601246105919004,
"grad_norm": 4.6875,
"learning_rate": 4.999822256707728e-05,
"loss": 0.8689,
"step": 444
},
{
"epoch": 2.7663551401869158,
"grad_norm": 4.34375,
"learning_rate": 4.9998193066975016e-05,
"loss": 0.8148,
"step": 445
},
{
"epoch": 2.7725856697819315,
"grad_norm": 5.1875,
"learning_rate": 4.999816332408531e-05,
"loss": 0.9029,
"step": 446
},
{
"epoch": 2.778816199376947,
"grad_norm": 5.3125,
"learning_rate": 4.999813333840845e-05,
"loss": 0.9183,
"step": 447
},
{
"epoch": 2.7850467289719627,
"grad_norm": 5.28125,
"learning_rate": 4.9998103109944714e-05,
"loss": 1.0838,
"step": 448
},
{
"epoch": 2.791277258566978,
"grad_norm": 4.9375,
"learning_rate": 4.99980726386944e-05,
"loss": 0.7835,
"step": 449
},
{
"epoch": 2.797507788161994,
"grad_norm": 6.21875,
"learning_rate": 4.9998041924657815e-05,
"loss": 0.9563,
"step": 450
},
{
"epoch": 2.803738317757009,
"grad_norm": 5.9375,
"learning_rate": 4.999801096783524e-05,
"loss": 1.0686,
"step": 451
},
{
"epoch": 2.809968847352025,
"grad_norm": 5.125,
"learning_rate": 4.9997979768227e-05,
"loss": 0.7887,
"step": 452
},
{
"epoch": 2.8161993769470404,
"grad_norm": 4.375,
"learning_rate": 4.999794832583337e-05,
"loss": 0.797,
"step": 453
},
{
"epoch": 2.822429906542056,
"grad_norm": 5.25,
"learning_rate": 4.999791664065467e-05,
"loss": 0.9853,
"step": 454
},
{
"epoch": 2.8286604361370715,
"grad_norm": 5.28125,
"learning_rate": 4.999788471269122e-05,
"loss": 0.9075,
"step": 455
},
{
"epoch": 2.8348909657320873,
"grad_norm": 5.40625,
"learning_rate": 4.99978525419433e-05,
"loss": 0.845,
"step": 456
},
{
"epoch": 2.8411214953271027,
"grad_norm": 5.78125,
"learning_rate": 4.999782012841126e-05,
"loss": 0.7564,
"step": 457
},
{
"epoch": 2.8473520249221185,
"grad_norm": 4.625,
"learning_rate": 4.999778747209537e-05,
"loss": 0.7361,
"step": 458
},
{
"epoch": 2.853582554517134,
"grad_norm": 5.625,
"learning_rate": 4.999775457299598e-05,
"loss": 0.9825,
"step": 459
},
{
"epoch": 2.8598130841121496,
"grad_norm": 5.90625,
"learning_rate": 4.999772143111341e-05,
"loss": 0.913,
"step": 460
},
{
"epoch": 2.866043613707165,
"grad_norm": 5.3125,
"learning_rate": 4.999768804644796e-05,
"loss": 0.8219,
"step": 461
},
{
"epoch": 2.872274143302181,
"grad_norm": 5.375,
"learning_rate": 4.999765441899997e-05,
"loss": 0.8614,
"step": 462
},
{
"epoch": 2.878504672897196,
"grad_norm": 4.46875,
"learning_rate": 4.999762054876977e-05,
"loss": 0.7095,
"step": 463
},
{
"epoch": 2.884735202492212,
"grad_norm": 5.5625,
"learning_rate": 4.999758643575768e-05,
"loss": 0.8056,
"step": 464
},
{
"epoch": 2.8909657320872273,
"grad_norm": 6.0625,
"learning_rate": 4.999755207996403e-05,
"loss": 1.027,
"step": 465
},
{
"epoch": 2.897196261682243,
"grad_norm": 5.4375,
"learning_rate": 4.9997517481389156e-05,
"loss": 0.8827,
"step": 466
},
{
"epoch": 2.9034267912772584,
"grad_norm": 3.9375,
"learning_rate": 4.99974826400334e-05,
"loss": 0.6149,
"step": 467
},
{
"epoch": 2.9096573208722742,
"grad_norm": 7.5,
"learning_rate": 4.999744755589709e-05,
"loss": 1.0117,
"step": 468
},
{
"epoch": 2.9158878504672896,
"grad_norm": 5.1875,
"learning_rate": 4.999741222898058e-05,
"loss": 0.7164,
"step": 469
},
{
"epoch": 2.9221183800623054,
"grad_norm": 4.90625,
"learning_rate": 4.99973766592842e-05,
"loss": 0.7376,
"step": 470
},
{
"epoch": 2.9283489096573208,
"grad_norm": 5.625,
"learning_rate": 4.99973408468083e-05,
"loss": 0.8718,
"step": 471
},
{
"epoch": 2.9345794392523366,
"grad_norm": 4.8125,
"learning_rate": 4.999730479155323e-05,
"loss": 0.6672,
"step": 472
},
{
"epoch": 2.940809968847352,
"grad_norm": 5.375,
"learning_rate": 4.9997268493519345e-05,
"loss": 0.8156,
"step": 473
},
{
"epoch": 2.9470404984423677,
"grad_norm": 4.28125,
"learning_rate": 4.999723195270698e-05,
"loss": 0.6679,
"step": 474
},
{
"epoch": 2.953271028037383,
"grad_norm": 4.75,
"learning_rate": 4.9997195169116514e-05,
"loss": 0.7508,
"step": 475
},
{
"epoch": 2.959501557632399,
"grad_norm": 4.9375,
"learning_rate": 4.999715814274828e-05,
"loss": 0.7646,
"step": 476
},
{
"epoch": 2.965732087227414,
"grad_norm": 5.25,
"learning_rate": 4.9997120873602654e-05,
"loss": 0.9933,
"step": 477
},
{
"epoch": 2.97196261682243,
"grad_norm": 5.09375,
"learning_rate": 4.9997083361679995e-05,
"loss": 0.8115,
"step": 478
},
{
"epoch": 2.9781931464174454,
"grad_norm": 4.75,
"learning_rate": 4.999704560698067e-05,
"loss": 0.7211,
"step": 479
},
{
"epoch": 2.984423676012461,
"grad_norm": 5.875,
"learning_rate": 4.999700760950503e-05,
"loss": 0.7356,
"step": 480
},
{
"epoch": 2.9906542056074765,
"grad_norm": 5.25,
"learning_rate": 4.999696936925347e-05,
"loss": 0.758,
"step": 481
},
{
"epoch": 2.9968847352024923,
"grad_norm": 4.625,
"learning_rate": 4.999693088622633e-05,
"loss": 0.7358,
"step": 482
},
{
"epoch": 3.0,
"grad_norm": 3.046875,
"learning_rate": 4.9996892160424016e-05,
"loss": 0.4621,
"step": 483
},
{
"epoch": 3.0062305295950154,
"grad_norm": 5.65625,
"learning_rate": 4.9996853191846885e-05,
"loss": 1.1231,
"step": 484
},
{
"epoch": 3.012461059190031,
"grad_norm": 6.21875,
"learning_rate": 4.9996813980495305e-05,
"loss": 0.9528,
"step": 485
},
{
"epoch": 3.0186915887850465,
"grad_norm": 5.15625,
"learning_rate": 4.999677452636968e-05,
"loss": 0.7979,
"step": 486
},
{
"epoch": 3.0249221183800623,
"grad_norm": 4.25,
"learning_rate": 4.999673482947039e-05,
"loss": 0.6507,
"step": 487
},
{
"epoch": 3.0311526479750777,
"grad_norm": 4.25,
"learning_rate": 4.999669488979781e-05,
"loss": 0.7202,
"step": 488
},
{
"epoch": 3.0373831775700935,
"grad_norm": 5.0625,
"learning_rate": 4.999665470735233e-05,
"loss": 0.6574,
"step": 489
},
{
"epoch": 3.043613707165109,
"grad_norm": 4.59375,
"learning_rate": 4.999661428213435e-05,
"loss": 0.6182,
"step": 490
},
{
"epoch": 3.0498442367601246,
"grad_norm": 4.71875,
"learning_rate": 4.999657361414425e-05,
"loss": 0.835,
"step": 491
},
{
"epoch": 3.05607476635514,
"grad_norm": 4.9375,
"learning_rate": 4.999653270338242e-05,
"loss": 0.6736,
"step": 492
},
{
"epoch": 3.0623052959501558,
"grad_norm": 4.53125,
"learning_rate": 4.9996491549849286e-05,
"loss": 0.7084,
"step": 493
},
{
"epoch": 3.068535825545171,
"grad_norm": 4.8125,
"learning_rate": 4.999645015354522e-05,
"loss": 0.6486,
"step": 494
},
{
"epoch": 3.074766355140187,
"grad_norm": 3.765625,
"learning_rate": 4.999640851447064e-05,
"loss": 0.5048,
"step": 495
},
{
"epoch": 3.0809968847352023,
"grad_norm": 4.34375,
"learning_rate": 4.999636663262594e-05,
"loss": 0.5721,
"step": 496
},
{
"epoch": 3.087227414330218,
"grad_norm": 5.53125,
"learning_rate": 4.999632450801153e-05,
"loss": 0.8099,
"step": 497
},
{
"epoch": 3.0934579439252334,
"grad_norm": 5.15625,
"learning_rate": 4.999628214062782e-05,
"loss": 0.7127,
"step": 498
},
{
"epoch": 3.0996884735202492,
"grad_norm": 4.59375,
"learning_rate": 4.999623953047522e-05,
"loss": 0.696,
"step": 499
},
{
"epoch": 3.1059190031152646,
"grad_norm": 3.765625,
"learning_rate": 4.999619667755414e-05,
"loss": 0.5194,
"step": 500
},
{
"epoch": 3.1121495327102804,
"grad_norm": 4.40625,
"learning_rate": 4.9996153581865006e-05,
"loss": 0.6775,
"step": 501
},
{
"epoch": 3.1183800623052957,
"grad_norm": 4.90625,
"learning_rate": 4.9996110243408234e-05,
"loss": 0.7582,
"step": 502
},
{
"epoch": 3.1246105919003115,
"grad_norm": 4.40625,
"learning_rate": 4.999606666218424e-05,
"loss": 0.5244,
"step": 503
},
{
"epoch": 3.130841121495327,
"grad_norm": 4.125,
"learning_rate": 4.9996022838193454e-05,
"loss": 0.5733,
"step": 504
},
{
"epoch": 3.1370716510903427,
"grad_norm": 4.34375,
"learning_rate": 4.9995978771436294e-05,
"loss": 0.6864,
"step": 505
},
{
"epoch": 3.143302180685358,
"grad_norm": 5.03125,
"learning_rate": 4.999593446191319e-05,
"loss": 0.6933,
"step": 506
},
{
"epoch": 3.149532710280374,
"grad_norm": 4.40625,
"learning_rate": 4.999588990962458e-05,
"loss": 0.568,
"step": 507
},
{
"epoch": 3.155763239875389,
"grad_norm": 4.90625,
"learning_rate": 4.9995845114570886e-05,
"loss": 0.7914,
"step": 508
},
{
"epoch": 3.161993769470405,
"grad_norm": 4.84375,
"learning_rate": 4.999580007675255e-05,
"loss": 0.7657,
"step": 509
},
{
"epoch": 3.1682242990654204,
"grad_norm": 4.75,
"learning_rate": 4.999575479617e-05,
"loss": 0.6046,
"step": 510
},
{
"epoch": 3.174454828660436,
"grad_norm": 6.09375,
"learning_rate": 4.9995709272823686e-05,
"loss": 0.9041,
"step": 511
},
{
"epoch": 3.1806853582554515,
"grad_norm": 4.28125,
"learning_rate": 4.9995663506714054e-05,
"loss": 0.5564,
"step": 512
},
{
"epoch": 3.1869158878504673,
"grad_norm": 4.5,
"learning_rate": 4.999561749784154e-05,
"loss": 0.6658,
"step": 513
},
{
"epoch": 3.1931464174454827,
"grad_norm": 4.1875,
"learning_rate": 4.9995571246206586e-05,
"loss": 0.4249,
"step": 514
},
{
"epoch": 3.1993769470404985,
"grad_norm": 4.625,
"learning_rate": 4.999552475180965e-05,
"loss": 0.5835,
"step": 515
},
{
"epoch": 3.205607476635514,
"grad_norm": 4.34375,
"learning_rate": 4.999547801465119e-05,
"loss": 0.7145,
"step": 516
},
{
"epoch": 3.2118380062305296,
"grad_norm": 3.78125,
"learning_rate": 4.999543103473164e-05,
"loss": 0.5384,
"step": 517
},
{
"epoch": 3.218068535825545,
"grad_norm": 4.25,
"learning_rate": 4.999538381205147e-05,
"loss": 0.5949,
"step": 518
},
{
"epoch": 3.2242990654205608,
"grad_norm": 5.09375,
"learning_rate": 4.999533634661113e-05,
"loss": 0.7683,
"step": 519
},
{
"epoch": 3.230529595015576,
"grad_norm": 5.0,
"learning_rate": 4.99952886384111e-05,
"loss": 0.7697,
"step": 520
},
{
"epoch": 3.236760124610592,
"grad_norm": 3.359375,
"learning_rate": 4.999524068745182e-05,
"loss": 0.4189,
"step": 521
},
{
"epoch": 3.2429906542056073,
"grad_norm": 4.84375,
"learning_rate": 4.999519249373377e-05,
"loss": 0.6684,
"step": 522
},
{
"epoch": 3.249221183800623,
"grad_norm": 4.46875,
"learning_rate": 4.9995144057257413e-05,
"loss": 0.7399,
"step": 523
},
{
"epoch": 3.2554517133956384,
"grad_norm": 4.125,
"learning_rate": 4.999509537802322e-05,
"loss": 0.4795,
"step": 524
},
{
"epoch": 3.2616822429906542,
"grad_norm": 4.125,
"learning_rate": 4.999504645603167e-05,
"loss": 0.5085,
"step": 525
},
{
"epoch": 3.2679127725856696,
"grad_norm": 3.5625,
"learning_rate": 4.999499729128322e-05,
"loss": 0.5397,
"step": 526
},
{
"epoch": 3.2741433021806854,
"grad_norm": 4.1875,
"learning_rate": 4.9994947883778375e-05,
"loss": 0.6497,
"step": 527
},
{
"epoch": 3.2803738317757007,
"grad_norm": 5.28125,
"learning_rate": 4.999489823351759e-05,
"loss": 0.8145,
"step": 528
},
{
"epoch": 3.2866043613707165,
"grad_norm": 4.71875,
"learning_rate": 4.999484834050136e-05,
"loss": 0.7202,
"step": 529
},
{
"epoch": 3.292834890965732,
"grad_norm": 4.375,
"learning_rate": 4.9994798204730166e-05,
"loss": 0.715,
"step": 530
},
{
"epoch": 3.2990654205607477,
"grad_norm": 6.0625,
"learning_rate": 4.999474782620449e-05,
"loss": 0.7715,
"step": 531
},
{
"epoch": 3.305295950155763,
"grad_norm": 4.0625,
"learning_rate": 4.999469720492483e-05,
"loss": 0.5498,
"step": 532
},
{
"epoch": 3.311526479750779,
"grad_norm": 4.9375,
"learning_rate": 4.999464634089168e-05,
"loss": 0.7726,
"step": 533
},
{
"epoch": 3.317757009345794,
"grad_norm": 4.875,
"learning_rate": 4.999459523410553e-05,
"loss": 0.8293,
"step": 534
},
{
"epoch": 3.32398753894081,
"grad_norm": 5.78125,
"learning_rate": 4.999454388456687e-05,
"loss": 0.7182,
"step": 535
},
{
"epoch": 3.3302180685358254,
"grad_norm": 5.125,
"learning_rate": 4.99944922922762e-05,
"loss": 0.4778,
"step": 536
},
{
"epoch": 3.336448598130841,
"grad_norm": 6.15625,
"learning_rate": 4.999444045723403e-05,
"loss": 0.8836,
"step": 537
},
{
"epoch": 3.3426791277258565,
"grad_norm": 3.734375,
"learning_rate": 4.9994388379440854e-05,
"loss": 0.5658,
"step": 538
},
{
"epoch": 3.3489096573208723,
"grad_norm": 5.09375,
"learning_rate": 4.9994336058897184e-05,
"loss": 0.7233,
"step": 539
},
{
"epoch": 3.3551401869158877,
"grad_norm": 4.46875,
"learning_rate": 4.999428349560352e-05,
"loss": 0.6664,
"step": 540
},
{
"epoch": 3.3613707165109035,
"grad_norm": 4.875,
"learning_rate": 4.9994230689560387e-05,
"loss": 0.7333,
"step": 541
},
{
"epoch": 3.367601246105919,
"grad_norm": 5.84375,
"learning_rate": 4.999417764076829e-05,
"loss": 0.7767,
"step": 542
},
{
"epoch": 3.3738317757009346,
"grad_norm": 3.8125,
"learning_rate": 4.9994124349227736e-05,
"loss": 0.4648,
"step": 543
},
{
"epoch": 3.38006230529595,
"grad_norm": 4.6875,
"learning_rate": 4.9994070814939256e-05,
"loss": 0.6658,
"step": 544
},
{
"epoch": 3.3862928348909658,
"grad_norm": 4.90625,
"learning_rate": 4.9994017037903365e-05,
"loss": 0.8765,
"step": 545
},
{
"epoch": 3.392523364485981,
"grad_norm": 3.703125,
"learning_rate": 4.999396301812058e-05,
"loss": 0.495,
"step": 546
},
{
"epoch": 3.398753894080997,
"grad_norm": 4.09375,
"learning_rate": 4.999390875559143e-05,
"loss": 0.5036,
"step": 547
},
{
"epoch": 3.4049844236760123,
"grad_norm": 4.03125,
"learning_rate": 4.999385425031644e-05,
"loss": 0.5652,
"step": 548
},
{
"epoch": 3.411214953271028,
"grad_norm": 4.25,
"learning_rate": 4.999379950229615e-05,
"loss": 0.5252,
"step": 549
},
{
"epoch": 3.4174454828660434,
"grad_norm": 4.09375,
"learning_rate": 4.999374451153107e-05,
"loss": 0.5549,
"step": 550
},
{
"epoch": 3.4236760124610592,
"grad_norm": 4.0625,
"learning_rate": 4.999368927802176e-05,
"loss": 0.4714,
"step": 551
},
{
"epoch": 3.4299065420560746,
"grad_norm": 4.9375,
"learning_rate": 4.999363380176874e-05,
"loss": 0.6734,
"step": 552
},
{
"epoch": 3.4361370716510904,
"grad_norm": 4.40625,
"learning_rate": 4.999357808277255e-05,
"loss": 0.6682,
"step": 553
},
{
"epoch": 3.4423676012461057,
"grad_norm": 3.328125,
"learning_rate": 4.999352212103373e-05,
"loss": 0.5223,
"step": 554
},
{
"epoch": 3.4485981308411215,
"grad_norm": 5.03125,
"learning_rate": 4.999346591655284e-05,
"loss": 0.9254,
"step": 555
},
{
"epoch": 3.454828660436137,
"grad_norm": 4.9375,
"learning_rate": 4.99934094693304e-05,
"loss": 0.7136,
"step": 556
},
{
"epoch": 3.4610591900311527,
"grad_norm": 3.84375,
"learning_rate": 4.999335277936698e-05,
"loss": 0.6735,
"step": 557
},
{
"epoch": 3.467289719626168,
"grad_norm": 4.0625,
"learning_rate": 4.999329584666311e-05,
"loss": 0.5275,
"step": 558
},
{
"epoch": 3.473520249221184,
"grad_norm": 4.5625,
"learning_rate": 4.999323867121936e-05,
"loss": 0.6179,
"step": 559
},
{
"epoch": 3.479750778816199,
"grad_norm": 3.421875,
"learning_rate": 4.999318125303628e-05,
"loss": 0.4797,
"step": 560
},
{
"epoch": 3.485981308411215,
"grad_norm": 3.546875,
"learning_rate": 4.9993123592114435e-05,
"loss": 0.4278,
"step": 561
},
{
"epoch": 3.4922118380062304,
"grad_norm": 3.765625,
"learning_rate": 4.9993065688454365e-05,
"loss": 0.5626,
"step": 562
},
{
"epoch": 3.498442367601246,
"grad_norm": 3.53125,
"learning_rate": 4.999300754205666e-05,
"loss": 0.4351,
"step": 563
},
{
"epoch": 3.5046728971962615,
"grad_norm": 4.6875,
"learning_rate": 4.999294915292186e-05,
"loss": 0.6529,
"step": 564
},
{
"epoch": 3.5109034267912773,
"grad_norm": 3.984375,
"learning_rate": 4.999289052105054e-05,
"loss": 0.6563,
"step": 565
},
{
"epoch": 3.5171339563862927,
"grad_norm": 4.9375,
"learning_rate": 4.999283164644328e-05,
"loss": 0.7045,
"step": 566
},
{
"epoch": 3.5233644859813085,
"grad_norm": 4.03125,
"learning_rate": 4.999277252910063e-05,
"loss": 0.5543,
"step": 567
},
{
"epoch": 3.529595015576324,
"grad_norm": 4.75,
"learning_rate": 4.9992713169023184e-05,
"loss": 0.656,
"step": 568
},
{
"epoch": 3.5358255451713396,
"grad_norm": 3.953125,
"learning_rate": 4.9992653566211525e-05,
"loss": 0.6504,
"step": 569
},
{
"epoch": 3.542056074766355,
"grad_norm": 3.859375,
"learning_rate": 4.99925937206662e-05,
"loss": 0.4315,
"step": 570
},
{
"epoch": 3.5482866043613708,
"grad_norm": 4.96875,
"learning_rate": 4.9992533632387815e-05,
"loss": 0.6828,
"step": 571
},
{
"epoch": 3.554517133956386,
"grad_norm": 4.625,
"learning_rate": 4.999247330137695e-05,
"loss": 0.773,
"step": 572
},
{
"epoch": 3.560747663551402,
"grad_norm": 5.3125,
"learning_rate": 4.9992412727634186e-05,
"loss": 0.5821,
"step": 573
},
{
"epoch": 3.5669781931464173,
"grad_norm": 4.28125,
"learning_rate": 4.999235191116011e-05,
"loss": 0.6647,
"step": 574
},
{
"epoch": 3.573208722741433,
"grad_norm": 4.3125,
"learning_rate": 4.9992290851955325e-05,
"loss": 0.6922,
"step": 575
},
{
"epoch": 3.5794392523364484,
"grad_norm": 4.75,
"learning_rate": 4.999222955002041e-05,
"loss": 0.7323,
"step": 576
},
{
"epoch": 3.5856697819314642,
"grad_norm": 4.03125,
"learning_rate": 4.9992168005355963e-05,
"loss": 0.5439,
"step": 577
},
{
"epoch": 3.5919003115264796,
"grad_norm": 4.09375,
"learning_rate": 4.999210621796259e-05,
"loss": 0.5479,
"step": 578
},
{
"epoch": 3.5981308411214954,
"grad_norm": 3.453125,
"learning_rate": 4.999204418784088e-05,
"loss": 0.4272,
"step": 579
},
{
"epoch": 3.6043613707165107,
"grad_norm": 4.65625,
"learning_rate": 4.999198191499145e-05,
"loss": 0.7121,
"step": 580
},
{
"epoch": 3.6105919003115265,
"grad_norm": 4.125,
"learning_rate": 4.9991919399414887e-05,
"loss": 0.5861,
"step": 581
},
{
"epoch": 3.616822429906542,
"grad_norm": 3.671875,
"learning_rate": 4.999185664111181e-05,
"loss": 0.435,
"step": 582
},
{
"epoch": 3.6230529595015577,
"grad_norm": 3.90625,
"learning_rate": 4.999179364008283e-05,
"loss": 0.4602,
"step": 583
},
{
"epoch": 3.629283489096573,
"grad_norm": 3.546875,
"learning_rate": 4.999173039632855e-05,
"loss": 0.5044,
"step": 584
},
{
"epoch": 3.635514018691589,
"grad_norm": 3.484375,
"learning_rate": 4.9991666909849586e-05,
"loss": 0.584,
"step": 585
},
{
"epoch": 3.641744548286604,
"grad_norm": 5.03125,
"learning_rate": 4.999160318064656e-05,
"loss": 0.6684,
"step": 586
},
{
"epoch": 3.64797507788162,
"grad_norm": 4.28125,
"learning_rate": 4.9991539208720096e-05,
"loss": 0.6246,
"step": 587
},
{
"epoch": 3.6542056074766354,
"grad_norm": 3.28125,
"learning_rate": 4.99914749940708e-05,
"loss": 0.3903,
"step": 588
},
{
"epoch": 3.660436137071651,
"grad_norm": 3.484375,
"learning_rate": 4.9991410536699304e-05,
"loss": 0.3893,
"step": 589
},
{
"epoch": 3.6666666666666665,
"grad_norm": 3.65625,
"learning_rate": 4.999134583660623e-05,
"loss": 0.5107,
"step": 590
},
{
"epoch": 3.6728971962616823,
"grad_norm": 3.453125,
"learning_rate": 4.999128089379221e-05,
"loss": 0.4047,
"step": 591
},
{
"epoch": 3.6791277258566977,
"grad_norm": 3.796875,
"learning_rate": 4.999121570825788e-05,
"loss": 0.4711,
"step": 592
},
{
"epoch": 3.6853582554517135,
"grad_norm": 5.375,
"learning_rate": 4.9991150280003864e-05,
"loss": 0.7124,
"step": 593
},
{
"epoch": 3.691588785046729,
"grad_norm": 3.96875,
"learning_rate": 4.99910846090308e-05,
"loss": 0.4646,
"step": 594
},
{
"epoch": 3.6978193146417446,
"grad_norm": 3.484375,
"learning_rate": 4.999101869533932e-05,
"loss": 0.4874,
"step": 595
},
{
"epoch": 3.70404984423676,
"grad_norm": 3.546875,
"learning_rate": 4.999095253893008e-05,
"loss": 0.4,
"step": 596
},
{
"epoch": 3.710280373831776,
"grad_norm": 4.15625,
"learning_rate": 4.9990886139803714e-05,
"loss": 0.527,
"step": 597
},
{
"epoch": 3.716510903426791,
"grad_norm": 4.78125,
"learning_rate": 4.999081949796086e-05,
"loss": 0.7171,
"step": 598
},
{
"epoch": 3.722741433021807,
"grad_norm": 3.921875,
"learning_rate": 4.999075261340218e-05,
"loss": 0.3611,
"step": 599
},
{
"epoch": 3.7289719626168223,
"grad_norm": 4.875,
"learning_rate": 4.999068548612831e-05,
"loss": 0.6283,
"step": 600
},
{
"epoch": 3.735202492211838,
"grad_norm": 3.953125,
"learning_rate": 4.9990618116139904e-05,
"loss": 0.5929,
"step": 601
},
{
"epoch": 3.7414330218068534,
"grad_norm": 4.6875,
"learning_rate": 4.999055050343762e-05,
"loss": 0.8102,
"step": 602
},
{
"epoch": 3.7476635514018692,
"grad_norm": 3.9375,
"learning_rate": 4.999048264802212e-05,
"loss": 0.7025,
"step": 603
},
{
"epoch": 3.7538940809968846,
"grad_norm": 4.4375,
"learning_rate": 4.999041454989405e-05,
"loss": 0.752,
"step": 604
},
{
"epoch": 3.7601246105919004,
"grad_norm": 4.125,
"learning_rate": 4.999034620905408e-05,
"loss": 0.4942,
"step": 605
},
{
"epoch": 3.7663551401869158,
"grad_norm": 3.015625,
"learning_rate": 4.9990277625502876e-05,
"loss": 0.3325,
"step": 606
},
{
"epoch": 3.7725856697819315,
"grad_norm": 3.890625,
"learning_rate": 4.99902087992411e-05,
"loss": 0.6548,
"step": 607
},
{
"epoch": 3.778816199376947,
"grad_norm": 3.71875,
"learning_rate": 4.999013973026942e-05,
"loss": 0.5297,
"step": 608
},
{
"epoch": 3.7850467289719627,
"grad_norm": 3.9375,
"learning_rate": 4.999007041858851e-05,
"loss": 0.5369,
"step": 609
},
{
"epoch": 3.791277258566978,
"grad_norm": 4.59375,
"learning_rate": 4.999000086419903e-05,
"loss": 0.5929,
"step": 610
},
{
"epoch": 3.797507788161994,
"grad_norm": 4.21875,
"learning_rate": 4.998993106710168e-05,
"loss": 0.5623,
"step": 611
},
{
"epoch": 3.803738317757009,
"grad_norm": 3.9375,
"learning_rate": 4.9989861027297115e-05,
"loss": 0.692,
"step": 612
},
{
"epoch": 3.809968847352025,
"grad_norm": 4.03125,
"learning_rate": 4.998979074478603e-05,
"loss": 0.6394,
"step": 613
},
{
"epoch": 3.8161993769470404,
"grad_norm": 4.1875,
"learning_rate": 4.998972021956909e-05,
"loss": 0.5574,
"step": 614
},
{
"epoch": 3.822429906542056,
"grad_norm": 3.125,
"learning_rate": 4.9989649451647e-05,
"loss": 0.3449,
"step": 615
},
{
"epoch": 3.8286604361370715,
"grad_norm": 3.625,
"learning_rate": 4.998957844102044e-05,
"loss": 0.415,
"step": 616
},
{
"epoch": 3.8348909657320873,
"grad_norm": 4.03125,
"learning_rate": 4.9989507187690106e-05,
"loss": 0.5575,
"step": 617
},
{
"epoch": 3.8411214953271027,
"grad_norm": 3.984375,
"learning_rate": 4.9989435691656674e-05,
"loss": 0.6489,
"step": 618
},
{
"epoch": 3.8473520249221185,
"grad_norm": 4.125,
"learning_rate": 4.9989363952920846e-05,
"loss": 0.6339,
"step": 619
},
{
"epoch": 3.853582554517134,
"grad_norm": 3.390625,
"learning_rate": 4.9989291971483325e-05,
"loss": 0.3011,
"step": 620
},
{
"epoch": 3.8598130841121496,
"grad_norm": 4.46875,
"learning_rate": 4.9989219747344804e-05,
"loss": 0.6336,
"step": 621
},
{
"epoch": 3.866043613707165,
"grad_norm": 3.8125,
"learning_rate": 4.998914728050598e-05,
"loss": 0.4203,
"step": 622
},
{
"epoch": 3.872274143302181,
"grad_norm": 3.921875,
"learning_rate": 4.998907457096757e-05,
"loss": 0.5101,
"step": 623
},
{
"epoch": 3.878504672897196,
"grad_norm": 3.4375,
"learning_rate": 4.998900161873027e-05,
"loss": 0.4257,
"step": 624
},
{
"epoch": 3.884735202492212,
"grad_norm": 3.578125,
"learning_rate": 4.998892842379479e-05,
"loss": 0.3759,
"step": 625
},
{
"epoch": 3.8909657320872273,
"grad_norm": 3.5625,
"learning_rate": 4.998885498616184e-05,
"loss": 0.5193,
"step": 626
},
{
"epoch": 3.897196261682243,
"grad_norm": 3.671875,
"learning_rate": 4.998878130583214e-05,
"loss": 0.5978,
"step": 627
},
{
"epoch": 3.9034267912772584,
"grad_norm": 4.3125,
"learning_rate": 4.99887073828064e-05,
"loss": 0.5659,
"step": 628
},
{
"epoch": 3.9096573208722742,
"grad_norm": 4.0625,
"learning_rate": 4.998863321708534e-05,
"loss": 0.5093,
"step": 629
},
{
"epoch": 3.9158878504672896,
"grad_norm": 3.546875,
"learning_rate": 4.998855880866967e-05,
"loss": 0.5278,
"step": 630
},
{
"epoch": 3.9221183800623054,
"grad_norm": 3.5625,
"learning_rate": 4.9988484157560136e-05,
"loss": 0.4751,
"step": 631
},
{
"epoch": 3.9283489096573208,
"grad_norm": 3.390625,
"learning_rate": 4.9988409263757436e-05,
"loss": 0.438,
"step": 632
},
{
"epoch": 3.9345794392523366,
"grad_norm": 4.0,
"learning_rate": 4.998833412726231e-05,
"loss": 0.5687,
"step": 633
},
{
"epoch": 3.940809968847352,
"grad_norm": 3.5625,
"learning_rate": 4.99882587480755e-05,
"loss": 0.3689,
"step": 634
},
{
"epoch": 3.9470404984423677,
"grad_norm": 3.640625,
"learning_rate": 4.998818312619772e-05,
"loss": 0.4029,
"step": 635
},
{
"epoch": 3.953271028037383,
"grad_norm": 3.359375,
"learning_rate": 4.99881072616297e-05,
"loss": 0.3715,
"step": 636
},
{
"epoch": 3.959501557632399,
"grad_norm": 4.53125,
"learning_rate": 4.99880311543722e-05,
"loss": 0.523,
"step": 637
},
{
"epoch": 3.965732087227414,
"grad_norm": 4.4375,
"learning_rate": 4.998795480442595e-05,
"loss": 0.5642,
"step": 638
},
{
"epoch": 3.97196261682243,
"grad_norm": 4.21875,
"learning_rate": 4.9987878211791674e-05,
"loss": 0.5477,
"step": 639
},
{
"epoch": 3.9781931464174454,
"grad_norm": 3.84375,
"learning_rate": 4.998780137647014e-05,
"loss": 0.5899,
"step": 640
},
{
"epoch": 3.984423676012461,
"grad_norm": 3.78125,
"learning_rate": 4.998772429846208e-05,
"loss": 0.4752,
"step": 641
},
{
"epoch": 3.9906542056074765,
"grad_norm": 4.21875,
"learning_rate": 4.998764697776824e-05,
"loss": 0.5676,
"step": 642
},
{
"epoch": 3.9968847352024923,
"grad_norm": 3.296875,
"learning_rate": 4.998756941438939e-05,
"loss": 0.4421,
"step": 643
},
{
"epoch": 4.0,
"grad_norm": 2.3125,
"learning_rate": 4.9987491608326267e-05,
"loss": 0.2777,
"step": 644
},
{
"epoch": 4.006230529595015,
"grad_norm": 4.65625,
"learning_rate": 4.9987413559579636e-05,
"loss": 0.6684,
"step": 645
},
{
"epoch": 4.012461059190031,
"grad_norm": 4.3125,
"learning_rate": 4.998733526815024e-05,
"loss": 0.4495,
"step": 646
},
{
"epoch": 4.018691588785047,
"grad_norm": 3.1875,
"learning_rate": 4.998725673403885e-05,
"loss": 0.353,
"step": 647
},
{
"epoch": 4.024922118380062,
"grad_norm": 3.78125,
"learning_rate": 4.998717795724624e-05,
"loss": 0.6157,
"step": 648
},
{
"epoch": 4.031152647975078,
"grad_norm": 3.453125,
"learning_rate": 4.998709893777315e-05,
"loss": 0.5152,
"step": 649
},
{
"epoch": 4.037383177570093,
"grad_norm": 3.8125,
"learning_rate": 4.998701967562035e-05,
"loss": 0.4455,
"step": 650
},
{
"epoch": 4.043613707165109,
"grad_norm": 4.25,
"learning_rate": 4.9986940170788646e-05,
"loss": 0.5454,
"step": 651
},
{
"epoch": 4.049844236760125,
"grad_norm": 3.015625,
"learning_rate": 4.998686042327877e-05,
"loss": 0.3405,
"step": 652
},
{
"epoch": 4.05607476635514,
"grad_norm": 2.6875,
"learning_rate": 4.998678043309151e-05,
"loss": 0.308,
"step": 653
},
{
"epoch": 4.062305295950155,
"grad_norm": 3.765625,
"learning_rate": 4.9986700200227645e-05,
"loss": 0.5059,
"step": 654
},
{
"epoch": 4.068535825545172,
"grad_norm": 5.0,
"learning_rate": 4.9986619724687954e-05,
"loss": 0.6773,
"step": 655
},
{
"epoch": 4.074766355140187,
"grad_norm": 4.125,
"learning_rate": 4.9986539006473215e-05,
"loss": 0.508,
"step": 656
},
{
"epoch": 4.080996884735202,
"grad_norm": 3.40625,
"learning_rate": 4.9986458045584215e-05,
"loss": 0.4089,
"step": 657
},
{
"epoch": 4.087227414330218,
"grad_norm": 4.15625,
"learning_rate": 4.9986376842021745e-05,
"loss": 0.4406,
"step": 658
},
{
"epoch": 4.093457943925234,
"grad_norm": 4.03125,
"learning_rate": 4.9986295395786586e-05,
"loss": 0.4039,
"step": 659
},
{
"epoch": 4.099688473520249,
"grad_norm": 3.46875,
"learning_rate": 4.998621370687952e-05,
"loss": 0.4581,
"step": 660
},
{
"epoch": 4.105919003115265,
"grad_norm": 3.625,
"learning_rate": 4.998613177530137e-05,
"loss": 0.5185,
"step": 661
},
{
"epoch": 4.11214953271028,
"grad_norm": 4.03125,
"learning_rate": 4.998604960105291e-05,
"loss": 0.5209,
"step": 662
},
{
"epoch": 4.118380062305296,
"grad_norm": 3.703125,
"learning_rate": 4.9985967184134936e-05,
"loss": 0.4484,
"step": 663
},
{
"epoch": 4.1246105919003115,
"grad_norm": 3.34375,
"learning_rate": 4.9985884524548254e-05,
"loss": 0.3238,
"step": 664
},
{
"epoch": 4.130841121495327,
"grad_norm": 3.0,
"learning_rate": 4.9985801622293674e-05,
"loss": 0.3502,
"step": 665
},
{
"epoch": 4.137071651090342,
"grad_norm": 3.59375,
"learning_rate": 4.998571847737198e-05,
"loss": 0.4413,
"step": 666
},
{
"epoch": 4.1433021806853585,
"grad_norm": 3.421875,
"learning_rate": 4.9985635089784e-05,
"loss": 0.381,
"step": 667
},
{
"epoch": 4.149532710280374,
"grad_norm": 2.875,
"learning_rate": 4.998555145953054e-05,
"loss": 0.3024,
"step": 668
},
{
"epoch": 4.155763239875389,
"grad_norm": 2.953125,
"learning_rate": 4.998546758661241e-05,
"loss": 0.3351,
"step": 669
},
{
"epoch": 4.161993769470405,
"grad_norm": 4.34375,
"learning_rate": 4.9985383471030425e-05,
"loss": 0.4435,
"step": 670
},
{
"epoch": 4.168224299065421,
"grad_norm": 3.75,
"learning_rate": 4.9985299112785406e-05,
"loss": 0.4076,
"step": 671
},
{
"epoch": 4.174454828660436,
"grad_norm": 3.609375,
"learning_rate": 4.998521451187816e-05,
"loss": 0.4847,
"step": 672
},
{
"epoch": 4.1806853582554515,
"grad_norm": 4.28125,
"learning_rate": 4.998512966830951e-05,
"loss": 0.6744,
"step": 673
},
{
"epoch": 4.186915887850467,
"grad_norm": 3.84375,
"learning_rate": 4.99850445820803e-05,
"loss": 0.4352,
"step": 674
},
{
"epoch": 4.193146417445483,
"grad_norm": 3.015625,
"learning_rate": 4.998495925319134e-05,
"loss": 0.324,
"step": 675
},
{
"epoch": 4.1993769470404985,
"grad_norm": 3.78125,
"learning_rate": 4.998487368164346e-05,
"loss": 0.5384,
"step": 676
},
{
"epoch": 4.205607476635514,
"grad_norm": 4.40625,
"learning_rate": 4.998478786743749e-05,
"loss": 0.5151,
"step": 677
},
{
"epoch": 4.211838006230529,
"grad_norm": 4.3125,
"learning_rate": 4.998470181057426e-05,
"loss": 0.4478,
"step": 678
},
{
"epoch": 4.218068535825545,
"grad_norm": 3.875,
"learning_rate": 4.998461551105462e-05,
"loss": 0.4707,
"step": 679
},
{
"epoch": 4.224299065420561,
"grad_norm": 4.5625,
"learning_rate": 4.9984528968879394e-05,
"loss": 0.5456,
"step": 680
},
{
"epoch": 4.230529595015576,
"grad_norm": 5.0,
"learning_rate": 4.998444218404944e-05,
"loss": 0.5539,
"step": 681
},
{
"epoch": 4.2367601246105915,
"grad_norm": 3.3125,
"learning_rate": 4.9984355156565576e-05,
"loss": 0.5032,
"step": 682
},
{
"epoch": 4.242990654205608,
"grad_norm": 3.90625,
"learning_rate": 4.998426788642867e-05,
"loss": 0.4802,
"step": 683
},
{
"epoch": 4.249221183800623,
"grad_norm": 3.375,
"learning_rate": 4.9984180373639555e-05,
"loss": 0.3989,
"step": 684
},
{
"epoch": 4.255451713395638,
"grad_norm": 2.96875,
"learning_rate": 4.998409261819909e-05,
"loss": 0.3357,
"step": 685
},
{
"epoch": 4.261682242990654,
"grad_norm": 3.34375,
"learning_rate": 4.998400462010812e-05,
"loss": 0.4396,
"step": 686
},
{
"epoch": 4.26791277258567,
"grad_norm": 3.96875,
"learning_rate": 4.99839163793675e-05,
"loss": 0.4657,
"step": 687
},
{
"epoch": 4.274143302180685,
"grad_norm": 3.390625,
"learning_rate": 4.998382789597809e-05,
"loss": 0.3952,
"step": 688
},
{
"epoch": 4.280373831775701,
"grad_norm": 3.5625,
"learning_rate": 4.9983739169940756e-05,
"loss": 0.3831,
"step": 689
},
{
"epoch": 4.286604361370716,
"grad_norm": 3.796875,
"learning_rate": 4.998365020125635e-05,
"loss": 0.4033,
"step": 690
},
{
"epoch": 4.292834890965732,
"grad_norm": 3.171875,
"learning_rate": 4.9983560989925736e-05,
"loss": 0.3334,
"step": 691
},
{
"epoch": 4.299065420560748,
"grad_norm": 3.109375,
"learning_rate": 4.998347153594979e-05,
"loss": 0.4003,
"step": 692
},
{
"epoch": 4.305295950155763,
"grad_norm": 2.8125,
"learning_rate": 4.998338183932937e-05,
"loss": 0.3331,
"step": 693
},
{
"epoch": 4.311526479750778,
"grad_norm": 3.984375,
"learning_rate": 4.998329190006535e-05,
"loss": 0.5233,
"step": 694
},
{
"epoch": 4.317757009345795,
"grad_norm": 3.203125,
"learning_rate": 4.99832017181586e-05,
"loss": 0.5107,
"step": 695
},
{
"epoch": 4.32398753894081,
"grad_norm": 3.609375,
"learning_rate": 4.9983111293610015e-05,
"loss": 0.4391,
"step": 696
},
{
"epoch": 4.330218068535825,
"grad_norm": 4.375,
"learning_rate": 4.998302062642045e-05,
"loss": 0.6024,
"step": 697
},
{
"epoch": 4.336448598130841,
"grad_norm": 3.28125,
"learning_rate": 4.99829297165908e-05,
"loss": 0.4184,
"step": 698
},
{
"epoch": 4.342679127725857,
"grad_norm": 2.71875,
"learning_rate": 4.998283856412194e-05,
"loss": 0.3263,
"step": 699
},
{
"epoch": 4.348909657320872,
"grad_norm": 3.109375,
"learning_rate": 4.998274716901475e-05,
"loss": 0.3787,
"step": 700
},
{
"epoch": 4.355140186915888,
"grad_norm": 4.03125,
"learning_rate": 4.998265553127013e-05,
"loss": 0.3787,
"step": 701
},
{
"epoch": 4.361370716510903,
"grad_norm": 4.125,
"learning_rate": 4.998256365088897e-05,
"loss": 0.4806,
"step": 702
},
{
"epoch": 4.367601246105919,
"grad_norm": 3.03125,
"learning_rate": 4.998247152787216e-05,
"loss": 0.313,
"step": 703
},
{
"epoch": 4.373831775700935,
"grad_norm": 4.21875,
"learning_rate": 4.9982379162220586e-05,
"loss": 0.5431,
"step": 704
},
{
"epoch": 4.38006230529595,
"grad_norm": 3.75,
"learning_rate": 4.998228655393515e-05,
"loss": 0.6257,
"step": 705
},
{
"epoch": 4.386292834890965,
"grad_norm": 4.03125,
"learning_rate": 4.9982193703016765e-05,
"loss": 0.4019,
"step": 706
},
{
"epoch": 4.392523364485982,
"grad_norm": 2.984375,
"learning_rate": 4.998210060946631e-05,
"loss": 0.3264,
"step": 707
},
{
"epoch": 4.398753894080997,
"grad_norm": 3.65625,
"learning_rate": 4.99820072732847e-05,
"loss": 0.4703,
"step": 708
},
{
"epoch": 4.404984423676012,
"grad_norm": 3.1875,
"learning_rate": 4.9981913694472846e-05,
"loss": 0.3716,
"step": 709
},
{
"epoch": 4.411214953271028,
"grad_norm": 4.1875,
"learning_rate": 4.9981819873031656e-05,
"loss": 0.5235,
"step": 710
},
{
"epoch": 4.417445482866044,
"grad_norm": 3.40625,
"learning_rate": 4.998172580896203e-05,
"loss": 0.3997,
"step": 711
},
{
"epoch": 4.423676012461059,
"grad_norm": 3.75,
"learning_rate": 4.998163150226489e-05,
"loss": 0.4582,
"step": 712
},
{
"epoch": 4.429906542056075,
"grad_norm": 2.546875,
"learning_rate": 4.998153695294116e-05,
"loss": 0.3264,
"step": 713
},
{
"epoch": 4.43613707165109,
"grad_norm": 2.671875,
"learning_rate": 4.998144216099174e-05,
"loss": 0.2587,
"step": 714
},
{
"epoch": 4.442367601246106,
"grad_norm": 3.671875,
"learning_rate": 4.9981347126417564e-05,
"loss": 0.4466,
"step": 715
},
{
"epoch": 4.4485981308411215,
"grad_norm": 2.75,
"learning_rate": 4.998125184921955e-05,
"loss": 0.3148,
"step": 716
},
{
"epoch": 4.454828660436137,
"grad_norm": 3.421875,
"learning_rate": 4.9981156329398626e-05,
"loss": 0.3908,
"step": 717
},
{
"epoch": 4.461059190031152,
"grad_norm": 3.34375,
"learning_rate": 4.9981060566955715e-05,
"loss": 0.4605,
"step": 718
},
{
"epoch": 4.4672897196261685,
"grad_norm": 3.515625,
"learning_rate": 4.998096456189175e-05,
"loss": 0.328,
"step": 719
},
{
"epoch": 4.473520249221184,
"grad_norm": 3.21875,
"learning_rate": 4.9980868314207665e-05,
"loss": 0.4464,
"step": 720
},
{
"epoch": 4.479750778816199,
"grad_norm": 3.203125,
"learning_rate": 4.9980771823904396e-05,
"loss": 0.4604,
"step": 721
},
{
"epoch": 4.485981308411215,
"grad_norm": 3.359375,
"learning_rate": 4.998067509098288e-05,
"loss": 0.4413,
"step": 722
},
{
"epoch": 4.492211838006231,
"grad_norm": 4.875,
"learning_rate": 4.9980578115444045e-05,
"loss": 0.6606,
"step": 723
},
{
"epoch": 4.498442367601246,
"grad_norm": 4.09375,
"learning_rate": 4.998048089728884e-05,
"loss": 0.656,
"step": 724
},
{
"epoch": 4.5046728971962615,
"grad_norm": 3.5,
"learning_rate": 4.998038343651822e-05,
"loss": 0.447,
"step": 725
},
{
"epoch": 4.510903426791277,
"grad_norm": 3.5,
"learning_rate": 4.998028573313312e-05,
"loss": 0.396,
"step": 726
},
{
"epoch": 4.517133956386293,
"grad_norm": 3.765625,
"learning_rate": 4.9980187787134495e-05,
"loss": 0.4968,
"step": 727
},
{
"epoch": 4.5233644859813085,
"grad_norm": 4.28125,
"learning_rate": 4.998008959852328e-05,
"loss": 0.6456,
"step": 728
},
{
"epoch": 4.529595015576324,
"grad_norm": 4.03125,
"learning_rate": 4.9979991167300453e-05,
"loss": 0.6197,
"step": 729
},
{
"epoch": 4.535825545171339,
"grad_norm": 3.265625,
"learning_rate": 4.997989249346695e-05,
"loss": 0.3573,
"step": 730
},
{
"epoch": 4.542056074766355,
"grad_norm": 2.59375,
"learning_rate": 4.9979793577023746e-05,
"loss": 0.2501,
"step": 731
},
{
"epoch": 4.548286604361371,
"grad_norm": 3.265625,
"learning_rate": 4.997969441797179e-05,
"loss": 0.3663,
"step": 732
},
{
"epoch": 4.554517133956386,
"grad_norm": 3.609375,
"learning_rate": 4.997959501631205e-05,
"loss": 0.4846,
"step": 733
},
{
"epoch": 4.5607476635514015,
"grad_norm": 3.640625,
"learning_rate": 4.997949537204549e-05,
"loss": 0.3935,
"step": 734
},
{
"epoch": 4.566978193146418,
"grad_norm": 3.015625,
"learning_rate": 4.997939548517308e-05,
"loss": 0.3709,
"step": 735
},
{
"epoch": 4.573208722741433,
"grad_norm": 2.953125,
"learning_rate": 4.9979295355695776e-05,
"loss": 0.4009,
"step": 736
},
{
"epoch": 4.579439252336448,
"grad_norm": 3.328125,
"learning_rate": 4.997919498361457e-05,
"loss": 0.4188,
"step": 737
},
{
"epoch": 4.585669781931464,
"grad_norm": 4.0625,
"learning_rate": 4.997909436893043e-05,
"loss": 0.4907,
"step": 738
},
{
"epoch": 4.59190031152648,
"grad_norm": 3.8125,
"learning_rate": 4.9978993511644334e-05,
"loss": 0.4445,
"step": 739
},
{
"epoch": 4.598130841121495,
"grad_norm": 2.90625,
"learning_rate": 4.997889241175726e-05,
"loss": 0.2928,
"step": 740
},
{
"epoch": 4.604361370716511,
"grad_norm": 2.59375,
"learning_rate": 4.9978791069270184e-05,
"loss": 0.2715,
"step": 741
},
{
"epoch": 4.610591900311526,
"grad_norm": 3.359375,
"learning_rate": 4.9978689484184104e-05,
"loss": 0.4514,
"step": 742
},
{
"epoch": 4.616822429906542,
"grad_norm": 3.375,
"learning_rate": 4.9978587656499995e-05,
"loss": 0.3046,
"step": 743
},
{
"epoch": 4.623052959501558,
"grad_norm": 2.9375,
"learning_rate": 4.9978485586218846e-05,
"loss": 0.3143,
"step": 744
},
{
"epoch": 4.629283489096573,
"grad_norm": 3.28125,
"learning_rate": 4.9978383273341655e-05,
"loss": 0.3901,
"step": 745
},
{
"epoch": 4.635514018691588,
"grad_norm": 3.5,
"learning_rate": 4.997828071786941e-05,
"loss": 0.4213,
"step": 746
},
{
"epoch": 4.641744548286605,
"grad_norm": 4.0625,
"learning_rate": 4.997817791980311e-05,
"loss": 0.5713,
"step": 747
},
{
"epoch": 4.64797507788162,
"grad_norm": 3.25,
"learning_rate": 4.9978074879143754e-05,
"loss": 0.3691,
"step": 748
},
{
"epoch": 4.654205607476635,
"grad_norm": 3.375,
"learning_rate": 4.997797159589235e-05,
"loss": 0.3337,
"step": 749
},
{
"epoch": 4.660436137071651,
"grad_norm": 2.9375,
"learning_rate": 4.997786807004988e-05,
"loss": 0.3845,
"step": 750
},
{
"epoch": 4.666666666666667,
"grad_norm": 3.5625,
"learning_rate": 4.997776430161737e-05,
"loss": 0.5719,
"step": 751
},
{
"epoch": 4.672897196261682,
"grad_norm": 2.90625,
"learning_rate": 4.997766029059582e-05,
"loss": 0.3318,
"step": 752
},
{
"epoch": 4.679127725856698,
"grad_norm": 2.8125,
"learning_rate": 4.9977556036986236e-05,
"loss": 0.2992,
"step": 753
},
{
"epoch": 4.685358255451713,
"grad_norm": 4.28125,
"learning_rate": 4.9977451540789635e-05,
"loss": 0.4576,
"step": 754
},
{
"epoch": 4.691588785046729,
"grad_norm": 2.953125,
"learning_rate": 4.997734680200703e-05,
"loss": 0.4092,
"step": 755
},
{
"epoch": 4.697819314641745,
"grad_norm": 3.0,
"learning_rate": 4.997724182063945e-05,
"loss": 0.3353,
"step": 756
},
{
"epoch": 4.70404984423676,
"grad_norm": 2.984375,
"learning_rate": 4.99771365966879e-05,
"loss": 0.3915,
"step": 757
},
{
"epoch": 4.710280373831775,
"grad_norm": 2.9375,
"learning_rate": 4.997703113015339e-05,
"loss": 0.3059,
"step": 758
},
{
"epoch": 4.716510903426792,
"grad_norm": 2.96875,
"learning_rate": 4.9976925421036986e-05,
"loss": 0.4154,
"step": 759
},
{
"epoch": 4.722741433021807,
"grad_norm": 4.25,
"learning_rate": 4.997681946933967e-05,
"loss": 0.5411,
"step": 760
},
{
"epoch": 4.728971962616822,
"grad_norm": 3.140625,
"learning_rate": 4.9976713275062506e-05,
"loss": 0.4312,
"step": 761
},
{
"epoch": 4.735202492211838,
"grad_norm": 3.0625,
"learning_rate": 4.9976606838206506e-05,
"loss": 0.4505,
"step": 762
},
{
"epoch": 4.741433021806854,
"grad_norm": 3.078125,
"learning_rate": 4.9976500158772706e-05,
"loss": 0.378,
"step": 763
},
{
"epoch": 4.747663551401869,
"grad_norm": 2.328125,
"learning_rate": 4.997639323676214e-05,
"loss": 0.2747,
"step": 764
},
{
"epoch": 4.753894080996885,
"grad_norm": 3.515625,
"learning_rate": 4.997628607217585e-05,
"loss": 0.5593,
"step": 765
},
{
"epoch": 4.7601246105919,
"grad_norm": 3.1875,
"learning_rate": 4.997617866501488e-05,
"loss": 0.4726,
"step": 766
},
{
"epoch": 4.766355140186916,
"grad_norm": 2.890625,
"learning_rate": 4.9976071015280276e-05,
"loss": 0.2882,
"step": 767
},
{
"epoch": 4.7725856697819315,
"grad_norm": 3.890625,
"learning_rate": 4.9975963122973066e-05,
"loss": 0.5086,
"step": 768
},
{
"epoch": 4.778816199376947,
"grad_norm": 3.09375,
"learning_rate": 4.9975854988094316e-05,
"loss": 0.4216,
"step": 769
},
{
"epoch": 4.785046728971962,
"grad_norm": 2.65625,
"learning_rate": 4.9975746610645075e-05,
"loss": 0.2954,
"step": 770
},
{
"epoch": 4.7912772585669785,
"grad_norm": 2.765625,
"learning_rate": 4.997563799062638e-05,
"loss": 0.2664,
"step": 771
},
{
"epoch": 4.797507788161994,
"grad_norm": 3.25,
"learning_rate": 4.9975529128039305e-05,
"loss": 0.434,
"step": 772
},
{
"epoch": 4.803738317757009,
"grad_norm": 2.765625,
"learning_rate": 4.9975420022884896e-05,
"loss": 0.3033,
"step": 773
},
{
"epoch": 4.809968847352025,
"grad_norm": 2.984375,
"learning_rate": 4.997531067516421e-05,
"loss": 0.4063,
"step": 774
},
{
"epoch": 4.816199376947041,
"grad_norm": 2.90625,
"learning_rate": 4.997520108487832e-05,
"loss": 0.272,
"step": 775
},
{
"epoch": 4.822429906542056,
"grad_norm": 3.09375,
"learning_rate": 4.997509125202828e-05,
"loss": 0.3964,
"step": 776
},
{
"epoch": 4.8286604361370715,
"grad_norm": 3.5,
"learning_rate": 4.997498117661517e-05,
"loss": 0.4103,
"step": 777
},
{
"epoch": 4.834890965732087,
"grad_norm": 3.71875,
"learning_rate": 4.997487085864004e-05,
"loss": 0.4903,
"step": 778
},
{
"epoch": 4.841121495327103,
"grad_norm": 3.03125,
"learning_rate": 4.997476029810398e-05,
"loss": 0.4039,
"step": 779
},
{
"epoch": 4.8473520249221185,
"grad_norm": 3.390625,
"learning_rate": 4.9974649495008055e-05,
"loss": 0.509,
"step": 780
},
{
"epoch": 4.853582554517134,
"grad_norm": 2.765625,
"learning_rate": 4.9974538449353335e-05,
"loss": 0.3442,
"step": 781
},
{
"epoch": 4.859813084112149,
"grad_norm": 3.265625,
"learning_rate": 4.997442716114091e-05,
"loss": 0.3326,
"step": 782
},
{
"epoch": 4.866043613707165,
"grad_norm": 2.125,
"learning_rate": 4.9974315630371855e-05,
"loss": 0.2269,
"step": 783
},
{
"epoch": 4.872274143302181,
"grad_norm": 2.859375,
"learning_rate": 4.997420385704725e-05,
"loss": 0.3225,
"step": 784
},
{
"epoch": 4.878504672897196,
"grad_norm": 3.4375,
"learning_rate": 4.9974091841168195e-05,
"loss": 0.3084,
"step": 785
},
{
"epoch": 4.8847352024922115,
"grad_norm": 3.234375,
"learning_rate": 4.997397958273576e-05,
"loss": 0.3902,
"step": 786
},
{
"epoch": 4.890965732087228,
"grad_norm": 3.296875,
"learning_rate": 4.997386708175105e-05,
"loss": 0.3667,
"step": 787
},
{
"epoch": 4.897196261682243,
"grad_norm": 2.78125,
"learning_rate": 4.997375433821514e-05,
"loss": 0.2716,
"step": 788
},
{
"epoch": 4.9034267912772584,
"grad_norm": 2.8125,
"learning_rate": 4.9973641352129144e-05,
"loss": 0.3484,
"step": 789
},
{
"epoch": 4.909657320872274,
"grad_norm": 2.796875,
"learning_rate": 4.9973528123494146e-05,
"loss": 0.3118,
"step": 790
},
{
"epoch": 4.91588785046729,
"grad_norm": 3.34375,
"learning_rate": 4.997341465231126e-05,
"loss": 0.3984,
"step": 791
},
{
"epoch": 4.922118380062305,
"grad_norm": 3.5,
"learning_rate": 4.997330093858157e-05,
"loss": 0.4269,
"step": 792
},
{
"epoch": 4.928348909657321,
"grad_norm": 3.875,
"learning_rate": 4.997318698230619e-05,
"loss": 0.4527,
"step": 793
},
{
"epoch": 4.934579439252336,
"grad_norm": 3.609375,
"learning_rate": 4.9973072783486226e-05,
"loss": 0.4197,
"step": 794
},
{
"epoch": 4.940809968847352,
"grad_norm": 3.578125,
"learning_rate": 4.997295834212279e-05,
"loss": 0.3834,
"step": 795
},
{
"epoch": 4.947040498442368,
"grad_norm": 3.375,
"learning_rate": 4.997284365821699e-05,
"loss": 0.4308,
"step": 796
},
{
"epoch": 4.953271028037383,
"grad_norm": 3.875,
"learning_rate": 4.997272873176994e-05,
"loss": 0.4949,
"step": 797
},
{
"epoch": 4.959501557632398,
"grad_norm": 3.234375,
"learning_rate": 4.9972613562782756e-05,
"loss": 0.3454,
"step": 798
},
{
"epoch": 4.965732087227415,
"grad_norm": 4.125,
"learning_rate": 4.9972498151256555e-05,
"loss": 0.3727,
"step": 799
},
{
"epoch": 4.97196261682243,
"grad_norm": 4.25,
"learning_rate": 4.9972382497192464e-05,
"loss": 0.6099,
"step": 800
},
{
"epoch": 4.978193146417445,
"grad_norm": 3.015625,
"learning_rate": 4.99722666005916e-05,
"loss": 0.337,
"step": 801
},
{
"epoch": 4.984423676012461,
"grad_norm": 2.21875,
"learning_rate": 4.997215046145509e-05,
"loss": 0.2683,
"step": 802
},
{
"epoch": 4.990654205607477,
"grad_norm": 3.03125,
"learning_rate": 4.997203407978407e-05,
"loss": 0.2881,
"step": 803
},
{
"epoch": 4.996884735202492,
"grad_norm": 2.640625,
"learning_rate": 4.9971917455579654e-05,
"loss": 0.2627,
"step": 804
},
{
"epoch": 5.0,
"grad_norm": 2.140625,
"learning_rate": 4.997180058884299e-05,
"loss": 0.2039,
"step": 805
},
{
"epoch": 5.006230529595015,
"grad_norm": 3.140625,
"learning_rate": 4.99716834795752e-05,
"loss": 0.3558,
"step": 806
},
{
"epoch": 5.012461059190031,
"grad_norm": 3.203125,
"learning_rate": 4.9971566127777436e-05,
"loss": 0.3786,
"step": 807
},
{
"epoch": 5.018691588785047,
"grad_norm": 2.109375,
"learning_rate": 4.9971448533450825e-05,
"loss": 0.2912,
"step": 808
},
{
"epoch": 5.024922118380062,
"grad_norm": 2.015625,
"learning_rate": 4.997133069659652e-05,
"loss": 0.2317,
"step": 809
},
{
"epoch": 5.031152647975078,
"grad_norm": 3.328125,
"learning_rate": 4.997121261721565e-05,
"loss": 0.4452,
"step": 810
},
{
"epoch": 5.037383177570093,
"grad_norm": 3.34375,
"learning_rate": 4.997109429530938e-05,
"loss": 0.4679,
"step": 811
},
{
"epoch": 5.043613707165109,
"grad_norm": 3.34375,
"learning_rate": 4.997097573087884e-05,
"loss": 0.4239,
"step": 812
},
{
"epoch": 5.049844236760125,
"grad_norm": 3.546875,
"learning_rate": 4.997085692392521e-05,
"loss": 0.522,
"step": 813
},
{
"epoch": 5.05607476635514,
"grad_norm": 3.265625,
"learning_rate": 4.9970737874449615e-05,
"loss": 0.4041,
"step": 814
},
{
"epoch": 5.062305295950155,
"grad_norm": 3.140625,
"learning_rate": 4.997061858245322e-05,
"loss": 0.3211,
"step": 815
},
{
"epoch": 5.068535825545172,
"grad_norm": 4.125,
"learning_rate": 4.997049904793719e-05,
"loss": 0.603,
"step": 816
},
{
"epoch": 5.074766355140187,
"grad_norm": 3.09375,
"learning_rate": 4.997037927090268e-05,
"loss": 0.3718,
"step": 817
},
{
"epoch": 5.080996884735202,
"grad_norm": 3.03125,
"learning_rate": 4.997025925135086e-05,
"loss": 0.3187,
"step": 818
},
{
"epoch": 5.087227414330218,
"grad_norm": 3.734375,
"learning_rate": 4.997013898928288e-05,
"loss": 0.4453,
"step": 819
},
{
"epoch": 5.093457943925234,
"grad_norm": 3.71875,
"learning_rate": 4.997001848469993e-05,
"loss": 0.4849,
"step": 820
},
{
"epoch": 5.099688473520249,
"grad_norm": 3.109375,
"learning_rate": 4.996989773760316e-05,
"loss": 0.3236,
"step": 821
},
{
"epoch": 5.105919003115265,
"grad_norm": 3.203125,
"learning_rate": 4.996977674799376e-05,
"loss": 0.4022,
"step": 822
},
{
"epoch": 5.11214953271028,
"grad_norm": 2.453125,
"learning_rate": 4.9969655515872895e-05,
"loss": 0.3246,
"step": 823
},
{
"epoch": 5.118380062305296,
"grad_norm": 3.328125,
"learning_rate": 4.996953404124174e-05,
"loss": 0.4188,
"step": 824
},
{
"epoch": 5.1246105919003115,
"grad_norm": 2.828125,
"learning_rate": 4.996941232410148e-05,
"loss": 0.3976,
"step": 825
},
{
"epoch": 5.130841121495327,
"grad_norm": 3.796875,
"learning_rate": 4.996929036445329e-05,
"loss": 0.4302,
"step": 826
},
{
"epoch": 5.137071651090342,
"grad_norm": 3.140625,
"learning_rate": 4.996916816229837e-05,
"loss": 0.3995,
"step": 827
},
{
"epoch": 5.1433021806853585,
"grad_norm": 3.0,
"learning_rate": 4.9969045717637896e-05,
"loss": 0.353,
"step": 828
},
{
"epoch": 5.149532710280374,
"grad_norm": 2.890625,
"learning_rate": 4.996892303047306e-05,
"loss": 0.3389,
"step": 829
},
{
"epoch": 5.155763239875389,
"grad_norm": 3.34375,
"learning_rate": 4.9968800100805054e-05,
"loss": 0.4003,
"step": 830
},
{
"epoch": 5.161993769470405,
"grad_norm": 3.09375,
"learning_rate": 4.996867692863507e-05,
"loss": 0.3559,
"step": 831
},
{
"epoch": 5.168224299065421,
"grad_norm": 4.46875,
"learning_rate": 4.99685535139643e-05,
"loss": 0.6878,
"step": 832
},
{
"epoch": 5.174454828660436,
"grad_norm": 2.171875,
"learning_rate": 4.9968429856793945e-05,
"loss": 0.2246,
"step": 833
},
{
"epoch": 5.1806853582554515,
"grad_norm": 2.625,
"learning_rate": 4.9968305957125215e-05,
"loss": 0.2521,
"step": 834
},
{
"epoch": 5.186915887850467,
"grad_norm": 3.109375,
"learning_rate": 4.99681818149593e-05,
"loss": 0.4201,
"step": 835
},
{
"epoch": 5.193146417445483,
"grad_norm": 3.03125,
"learning_rate": 4.9968057430297415e-05,
"loss": 0.3427,
"step": 836
},
{
"epoch": 5.1993769470404985,
"grad_norm": 3.265625,
"learning_rate": 4.9967932803140764e-05,
"loss": 0.3246,
"step": 837
},
{
"epoch": 5.205607476635514,
"grad_norm": 2.921875,
"learning_rate": 4.996780793349056e-05,
"loss": 0.2956,
"step": 838
},
{
"epoch": 5.211838006230529,
"grad_norm": 4.15625,
"learning_rate": 4.996768282134802e-05,
"loss": 0.4981,
"step": 839
},
{
"epoch": 5.218068535825545,
"grad_norm": 2.15625,
"learning_rate": 4.996755746671434e-05,
"loss": 0.2081,
"step": 840
},
{
"epoch": 5.224299065420561,
"grad_norm": 3.078125,
"learning_rate": 4.9967431869590766e-05,
"loss": 0.3338,
"step": 841
},
{
"epoch": 5.230529595015576,
"grad_norm": 2.84375,
"learning_rate": 4.9967306029978497e-05,
"loss": 0.3563,
"step": 842
},
{
"epoch": 5.2367601246105915,
"grad_norm": 3.859375,
"learning_rate": 4.996717994787876e-05,
"loss": 0.3779,
"step": 843
},
{
"epoch": 5.242990654205608,
"grad_norm": 3.890625,
"learning_rate": 4.996705362329278e-05,
"loss": 0.5422,
"step": 844
},
{
"epoch": 5.249221183800623,
"grad_norm": 2.34375,
"learning_rate": 4.996692705622178e-05,
"loss": 0.2982,
"step": 845
},
{
"epoch": 5.255451713395638,
"grad_norm": 3.109375,
"learning_rate": 4.9966800246667e-05,
"loss": 0.44,
"step": 846
},
{
"epoch": 5.261682242990654,
"grad_norm": 2.640625,
"learning_rate": 4.996667319462966e-05,
"loss": 0.2964,
"step": 847
},
{
"epoch": 5.26791277258567,
"grad_norm": 3.53125,
"learning_rate": 4.9966545900111006e-05,
"loss": 0.4021,
"step": 848
},
{
"epoch": 5.274143302180685,
"grad_norm": 3.140625,
"learning_rate": 4.996641836311226e-05,
"loss": 0.3612,
"step": 849
},
{
"epoch": 5.280373831775701,
"grad_norm": 3.296875,
"learning_rate": 4.996629058363468e-05,
"loss": 0.3233,
"step": 850
},
{
"epoch": 5.286604361370716,
"grad_norm": 2.625,
"learning_rate": 4.9966162561679486e-05,
"loss": 0.3235,
"step": 851
},
{
"epoch": 5.292834890965732,
"grad_norm": 3.03125,
"learning_rate": 4.996603429724793e-05,
"loss": 0.3976,
"step": 852
},
{
"epoch": 5.299065420560748,
"grad_norm": 1.9375,
"learning_rate": 4.996590579034126e-05,
"loss": 0.2009,
"step": 853
},
{
"epoch": 5.305295950155763,
"grad_norm": 2.0625,
"learning_rate": 4.996577704096072e-05,
"loss": 0.2654,
"step": 854
},
{
"epoch": 5.311526479750778,
"grad_norm": 3.125,
"learning_rate": 4.9965648049107575e-05,
"loss": 0.4302,
"step": 855
},
{
"epoch": 5.317757009345795,
"grad_norm": 2.828125,
"learning_rate": 4.996551881478305e-05,
"loss": 0.4216,
"step": 856
},
{
"epoch": 5.32398753894081,
"grad_norm": 3.359375,
"learning_rate": 4.996538933798843e-05,
"loss": 0.4452,
"step": 857
},
{
"epoch": 5.330218068535825,
"grad_norm": 4.125,
"learning_rate": 4.996525961872495e-05,
"loss": 0.735,
"step": 858
},
{
"epoch": 5.336448598130841,
"grad_norm": 3.375,
"learning_rate": 4.9965129656993875e-05,
"loss": 0.4283,
"step": 859
},
{
"epoch": 5.342679127725857,
"grad_norm": 2.109375,
"learning_rate": 4.996499945279648e-05,
"loss": 0.2183,
"step": 860
},
{
"epoch": 5.348909657320872,
"grad_norm": 3.859375,
"learning_rate": 4.9964869006134004e-05,
"loss": 0.4985,
"step": 861
},
{
"epoch": 5.355140186915888,
"grad_norm": 3.046875,
"learning_rate": 4.996473831700774e-05,
"loss": 0.2995,
"step": 862
},
{
"epoch": 5.361370716510903,
"grad_norm": 4.6875,
"learning_rate": 4.996460738541895e-05,
"loss": 0.6549,
"step": 863
},
{
"epoch": 5.367601246105919,
"grad_norm": 2.96875,
"learning_rate": 4.996447621136891e-05,
"loss": 0.2947,
"step": 864
},
{
"epoch": 5.373831775700935,
"grad_norm": 2.8125,
"learning_rate": 4.996434479485886e-05,
"loss": 0.2825,
"step": 865
},
{
"epoch": 5.38006230529595,
"grad_norm": 2.859375,
"learning_rate": 4.996421313589013e-05,
"loss": 0.3065,
"step": 866
},
{
"epoch": 5.386292834890965,
"grad_norm": 4.09375,
"learning_rate": 4.996408123446396e-05,
"loss": 0.3484,
"step": 867
},
{
"epoch": 5.392523364485982,
"grad_norm": 3.203125,
"learning_rate": 4.996394909058164e-05,
"loss": 0.2996,
"step": 868
},
{
"epoch": 5.398753894080997,
"grad_norm": 3.796875,
"learning_rate": 4.996381670424446e-05,
"loss": 0.5246,
"step": 869
},
{
"epoch": 5.404984423676012,
"grad_norm": 3.71875,
"learning_rate": 4.99636840754537e-05,
"loss": 0.3496,
"step": 870
},
{
"epoch": 5.411214953271028,
"grad_norm": 2.4375,
"learning_rate": 4.9963551204210655e-05,
"loss": 0.2159,
"step": 871
},
{
"epoch": 5.417445482866044,
"grad_norm": 2.375,
"learning_rate": 4.996341809051661e-05,
"loss": 0.2329,
"step": 872
},
{
"epoch": 5.423676012461059,
"grad_norm": 3.484375,
"learning_rate": 4.9963284734372855e-05,
"loss": 0.4562,
"step": 873
},
{
"epoch": 5.429906542056075,
"grad_norm": 3.203125,
"learning_rate": 4.996315113578069e-05,
"loss": 0.3015,
"step": 874
},
{
"epoch": 5.43613707165109,
"grad_norm": 2.9375,
"learning_rate": 4.9963017294741407e-05,
"loss": 0.3876,
"step": 875
},
{
"epoch": 5.442367601246106,
"grad_norm": 3.015625,
"learning_rate": 4.9962883211256315e-05,
"loss": 0.2761,
"step": 876
},
{
"epoch": 5.4485981308411215,
"grad_norm": 2.859375,
"learning_rate": 4.996274888532672e-05,
"loss": 0.374,
"step": 877
},
{
"epoch": 5.454828660436137,
"grad_norm": 3.078125,
"learning_rate": 4.99626143169539e-05,
"loss": 0.3615,
"step": 878
},
{
"epoch": 5.461059190031152,
"grad_norm": 2.53125,
"learning_rate": 4.9962479506139195e-05,
"loss": 0.2846,
"step": 879
},
{
"epoch": 5.4672897196261685,
"grad_norm": 3.4375,
"learning_rate": 4.996234445288389e-05,
"loss": 0.4174,
"step": 880
},
{
"epoch": 5.473520249221184,
"grad_norm": 2.859375,
"learning_rate": 4.996220915718931e-05,
"loss": 0.2457,
"step": 881
},
{
"epoch": 5.479750778816199,
"grad_norm": 3.21875,
"learning_rate": 4.996207361905676e-05,
"loss": 0.3794,
"step": 882
},
{
"epoch": 5.485981308411215,
"grad_norm": 2.71875,
"learning_rate": 4.996193783848756e-05,
"loss": 0.4012,
"step": 883
},
{
"epoch": 5.492211838006231,
"grad_norm": 2.46875,
"learning_rate": 4.996180181548303e-05,
"loss": 0.3452,
"step": 884
},
{
"epoch": 5.498442367601246,
"grad_norm": 2.75,
"learning_rate": 4.99616655500445e-05,
"loss": 0.2451,
"step": 885
},
{
"epoch": 5.5046728971962615,
"grad_norm": 2.5,
"learning_rate": 4.996152904217328e-05,
"loss": 0.2936,
"step": 886
},
{
"epoch": 5.510903426791277,
"grad_norm": 2.65625,
"learning_rate": 4.99613922918707e-05,
"loss": 0.3443,
"step": 887
},
{
"epoch": 5.517133956386293,
"grad_norm": 3.453125,
"learning_rate": 4.996125529913809e-05,
"loss": 0.3565,
"step": 888
},
{
"epoch": 5.5233644859813085,
"grad_norm": 2.90625,
"learning_rate": 4.996111806397677e-05,
"loss": 0.401,
"step": 889
},
{
"epoch": 5.529595015576324,
"grad_norm": 4.46875,
"learning_rate": 4.996098058638809e-05,
"loss": 0.5297,
"step": 890
},
{
"epoch": 5.535825545171339,
"grad_norm": 3.6875,
"learning_rate": 4.996084286637338e-05,
"loss": 0.4182,
"step": 891
},
{
"epoch": 5.542056074766355,
"grad_norm": 3.328125,
"learning_rate": 4.996070490393397e-05,
"loss": 0.469,
"step": 892
},
{
"epoch": 5.548286604361371,
"grad_norm": 2.59375,
"learning_rate": 4.9960566699071204e-05,
"loss": 0.2162,
"step": 893
},
{
"epoch": 5.554517133956386,
"grad_norm": 2.859375,
"learning_rate": 4.9960428251786426e-05,
"loss": 0.3344,
"step": 894
},
{
"epoch": 5.5607476635514015,
"grad_norm": 3.484375,
"learning_rate": 4.9960289562080974e-05,
"loss": 0.3314,
"step": 895
},
{
"epoch": 5.566978193146418,
"grad_norm": 2.21875,
"learning_rate": 4.996015062995621e-05,
"loss": 0.245,
"step": 896
},
{
"epoch": 5.573208722741433,
"grad_norm": 3.109375,
"learning_rate": 4.996001145541347e-05,
"loss": 0.3681,
"step": 897
},
{
"epoch": 5.579439252336448,
"grad_norm": 3.09375,
"learning_rate": 4.99598720384541e-05,
"loss": 0.3845,
"step": 898
},
{
"epoch": 5.585669781931464,
"grad_norm": 4.25,
"learning_rate": 4.9959732379079475e-05,
"loss": 0.4653,
"step": 899
},
{
"epoch": 5.59190031152648,
"grad_norm": 2.40625,
"learning_rate": 4.9959592477290936e-05,
"loss": 0.2978,
"step": 900
},
{
"epoch": 5.598130841121495,
"grad_norm": 3.359375,
"learning_rate": 4.995945233308985e-05,
"loss": 0.4309,
"step": 901
},
{
"epoch": 5.604361370716511,
"grad_norm": 3.125,
"learning_rate": 4.995931194647757e-05,
"loss": 0.2908,
"step": 902
},
{
"epoch": 5.610591900311526,
"grad_norm": 4.71875,
"learning_rate": 4.9959171317455455e-05,
"loss": 0.4873,
"step": 903
},
{
"epoch": 5.616822429906542,
"grad_norm": 2.453125,
"learning_rate": 4.995903044602489e-05,
"loss": 0.2486,
"step": 904
},
{
"epoch": 5.623052959501558,
"grad_norm": 3.25,
"learning_rate": 4.995888933218723e-05,
"loss": 0.3601,
"step": 905
},
{
"epoch": 5.629283489096573,
"grad_norm": 2.5625,
"learning_rate": 4.995874797594384e-05,
"loss": 0.2908,
"step": 906
},
{
"epoch": 5.635514018691588,
"grad_norm": 3.5625,
"learning_rate": 4.995860637729611e-05,
"loss": 0.3601,
"step": 907
},
{
"epoch": 5.641744548286605,
"grad_norm": 2.046875,
"learning_rate": 4.9958464536245406e-05,
"loss": 0.2128,
"step": 908
},
{
"epoch": 5.64797507788162,
"grad_norm": 2.859375,
"learning_rate": 4.9958322452793096e-05,
"loss": 0.2758,
"step": 909
},
{
"epoch": 5.654205607476635,
"grad_norm": 2.96875,
"learning_rate": 4.9958180126940577e-05,
"loss": 0.3776,
"step": 910
},
{
"epoch": 5.660436137071651,
"grad_norm": 3.84375,
"learning_rate": 4.9958037558689216e-05,
"loss": 0.4962,
"step": 911
},
{
"epoch": 5.666666666666667,
"grad_norm": 2.703125,
"learning_rate": 4.995789474804042e-05,
"loss": 0.2658,
"step": 912
},
{
"epoch": 5.672897196261682,
"grad_norm": 2.921875,
"learning_rate": 4.9957751694995545e-05,
"loss": 0.4234,
"step": 913
},
{
"epoch": 5.679127725856698,
"grad_norm": 2.671875,
"learning_rate": 4.995760839955601e-05,
"loss": 0.2653,
"step": 914
},
{
"epoch": 5.685358255451713,
"grad_norm": 2.640625,
"learning_rate": 4.995746486172318e-05,
"loss": 0.315,
"step": 915
},
{
"epoch": 5.691588785046729,
"grad_norm": 2.546875,
"learning_rate": 4.9957321081498475e-05,
"loss": 0.2417,
"step": 916
},
{
"epoch": 5.697819314641745,
"grad_norm": 2.609375,
"learning_rate": 4.995717705888327e-05,
"loss": 0.2503,
"step": 917
},
{
"epoch": 5.70404984423676,
"grad_norm": 3.359375,
"learning_rate": 4.995703279387898e-05,
"loss": 0.5599,
"step": 918
},
{
"epoch": 5.710280373831775,
"grad_norm": 3.296875,
"learning_rate": 4.995688828648699e-05,
"loss": 0.5631,
"step": 919
},
{
"epoch": 5.716510903426792,
"grad_norm": 2.9375,
"learning_rate": 4.995674353670872e-05,
"loss": 0.3231,
"step": 920
},
{
"epoch": 5.722741433021807,
"grad_norm": 4.0625,
"learning_rate": 4.9956598544545566e-05,
"loss": 0.4601,
"step": 921
},
{
"epoch": 5.728971962616822,
"grad_norm": 2.03125,
"learning_rate": 4.995645330999893e-05,
"loss": 0.1966,
"step": 922
},
{
"epoch": 5.735202492211838,
"grad_norm": 3.34375,
"learning_rate": 4.995630783307024e-05,
"loss": 0.4002,
"step": 923
},
{
"epoch": 5.741433021806854,
"grad_norm": 3.25,
"learning_rate": 4.995616211376089e-05,
"loss": 0.4733,
"step": 924
},
{
"epoch": 5.747663551401869,
"grad_norm": 4.75,
"learning_rate": 4.9956016152072316e-05,
"loss": 0.658,
"step": 925
},
{
"epoch": 5.753894080996885,
"grad_norm": 2.46875,
"learning_rate": 4.995586994800592e-05,
"loss": 0.2404,
"step": 926
},
{
"epoch": 5.7601246105919,
"grad_norm": 3.875,
"learning_rate": 4.995572350156313e-05,
"loss": 0.547,
"step": 927
},
{
"epoch": 5.766355140186916,
"grad_norm": 2.984375,
"learning_rate": 4.9955576812745366e-05,
"loss": 0.3325,
"step": 928
},
{
"epoch": 5.7725856697819315,
"grad_norm": 3.515625,
"learning_rate": 4.995542988155404e-05,
"loss": 0.421,
"step": 929
},
{
"epoch": 5.778816199376947,
"grad_norm": 2.578125,
"learning_rate": 4.99552827079906e-05,
"loss": 0.3212,
"step": 930
},
{
"epoch": 5.785046728971962,
"grad_norm": 3.8125,
"learning_rate": 4.995513529205647e-05,
"loss": 0.6084,
"step": 931
},
{
"epoch": 5.7912772585669785,
"grad_norm": 4.4375,
"learning_rate": 4.995498763375306e-05,
"loss": 0.4221,
"step": 932
},
{
"epoch": 5.797507788161994,
"grad_norm": 3.078125,
"learning_rate": 4.995483973308184e-05,
"loss": 0.3393,
"step": 933
},
{
"epoch": 5.803738317757009,
"grad_norm": 2.234375,
"learning_rate": 4.9954691590044215e-05,
"loss": 0.2374,
"step": 934
},
{
"epoch": 5.809968847352025,
"grad_norm": 3.390625,
"learning_rate": 4.9954543204641644e-05,
"loss": 0.3846,
"step": 935
},
{
"epoch": 5.816199376947041,
"grad_norm": 3.0,
"learning_rate": 4.9954394576875564e-05,
"loss": 0.3901,
"step": 936
},
{
"epoch": 5.822429906542056,
"grad_norm": 3.140625,
"learning_rate": 4.9954245706747405e-05,
"loss": 0.3586,
"step": 937
},
{
"epoch": 5.8286604361370715,
"grad_norm": 3.453125,
"learning_rate": 4.9954096594258637e-05,
"loss": 0.2853,
"step": 938
},
{
"epoch": 5.834890965732087,
"grad_norm": 2.21875,
"learning_rate": 4.995394723941068e-05,
"loss": 0.2042,
"step": 939
},
{
"epoch": 5.841121495327103,
"grad_norm": 2.40625,
"learning_rate": 4.995379764220501e-05,
"loss": 0.2389,
"step": 940
},
{
"epoch": 5.8473520249221185,
"grad_norm": 2.296875,
"learning_rate": 4.9953647802643065e-05,
"loss": 0.2265,
"step": 941
},
{
"epoch": 5.853582554517134,
"grad_norm": 3.296875,
"learning_rate": 4.995349772072631e-05,
"loss": 0.336,
"step": 942
},
{
"epoch": 5.859813084112149,
"grad_norm": 3.21875,
"learning_rate": 4.9953347396456194e-05,
"loss": 0.4726,
"step": 943
},
{
"epoch": 5.866043613707165,
"grad_norm": 2.875,
"learning_rate": 4.995319682983418e-05,
"loss": 0.3356,
"step": 944
},
{
"epoch": 5.872274143302181,
"grad_norm": 2.296875,
"learning_rate": 4.9953046020861724e-05,
"loss": 0.2246,
"step": 945
},
{
"epoch": 5.878504672897196,
"grad_norm": 2.703125,
"learning_rate": 4.9952894969540296e-05,
"loss": 0.2844,
"step": 946
},
{
"epoch": 5.8847352024922115,
"grad_norm": 2.359375,
"learning_rate": 4.995274367587137e-05,
"loss": 0.1962,
"step": 947
},
{
"epoch": 5.890965732087228,
"grad_norm": 3.75,
"learning_rate": 4.9952592139856413e-05,
"loss": 0.438,
"step": 948
},
{
"epoch": 5.897196261682243,
"grad_norm": 3.0,
"learning_rate": 4.9952440361496886e-05,
"loss": 0.322,
"step": 949
},
{
"epoch": 5.9034267912772584,
"grad_norm": 3.0,
"learning_rate": 4.995228834079428e-05,
"loss": 0.3377,
"step": 950
},
{
"epoch": 5.909657320872274,
"grad_norm": 2.4375,
"learning_rate": 4.995213607775005e-05,
"loss": 0.2984,
"step": 951
},
{
"epoch": 5.91588785046729,
"grad_norm": 2.84375,
"learning_rate": 4.9951983572365694e-05,
"loss": 0.2857,
"step": 952
},
{
"epoch": 5.922118380062305,
"grad_norm": 2.90625,
"learning_rate": 4.995183082464269e-05,
"loss": 0.4088,
"step": 953
},
{
"epoch": 5.928348909657321,
"grad_norm": 2.765625,
"learning_rate": 4.9951677834582514e-05,
"loss": 0.2862,
"step": 954
},
{
"epoch": 5.934579439252336,
"grad_norm": 2.4375,
"learning_rate": 4.995152460218665e-05,
"loss": 0.2485,
"step": 955
},
{
"epoch": 5.940809968847352,
"grad_norm": 3.09375,
"learning_rate": 4.995137112745659e-05,
"loss": 0.4405,
"step": 956
},
{
"epoch": 5.947040498442368,
"grad_norm": 3.125,
"learning_rate": 4.995121741039383e-05,
"loss": 0.3973,
"step": 957
},
{
"epoch": 5.953271028037383,
"grad_norm": 3.125,
"learning_rate": 4.9951063450999856e-05,
"loss": 0.4436,
"step": 958
},
{
"epoch": 5.959501557632398,
"grad_norm": 3.015625,
"learning_rate": 4.9950909249276166e-05,
"loss": 0.3593,
"step": 959
},
{
"epoch": 5.965732087227415,
"grad_norm": 3.125,
"learning_rate": 4.9950754805224265e-05,
"loss": 0.4188,
"step": 960
},
{
"epoch": 5.97196261682243,
"grad_norm": 2.90625,
"learning_rate": 4.9950600118845644e-05,
"loss": 0.3854,
"step": 961
},
{
"epoch": 5.978193146417445,
"grad_norm": 2.34375,
"learning_rate": 4.9950445190141805e-05,
"loss": 0.2303,
"step": 962
},
{
"epoch": 5.984423676012461,
"grad_norm": 3.359375,
"learning_rate": 4.9950290019114255e-05,
"loss": 0.3306,
"step": 963
},
{
"epoch": 5.990654205607477,
"grad_norm": 2.75,
"learning_rate": 4.99501346057645e-05,
"loss": 0.2787,
"step": 964
},
{
"epoch": 5.996884735202492,
"grad_norm": 2.609375,
"learning_rate": 4.994997895009404e-05,
"loss": 0.3304,
"step": 965
},
{
"epoch": 6.0,
"grad_norm": 2.0625,
"learning_rate": 4.9949823052104416e-05,
"loss": 0.1785,
"step": 966
},
{
"epoch": 6.006230529595015,
"grad_norm": 1.7578125,
"learning_rate": 4.994966691179711e-05,
"loss": 0.1604,
"step": 967
},
{
"epoch": 6.012461059190031,
"grad_norm": 2.71875,
"learning_rate": 4.994951052917366e-05,
"loss": 0.287,
"step": 968
},
{
"epoch": 6.018691588785047,
"grad_norm": 2.421875,
"learning_rate": 4.9949353904235574e-05,
"loss": 0.2618,
"step": 969
},
{
"epoch": 6.024922118380062,
"grad_norm": 3.125,
"learning_rate": 4.994919703698438e-05,
"loss": 0.3903,
"step": 970
},
{
"epoch": 6.031152647975078,
"grad_norm": 2.640625,
"learning_rate": 4.9949039927421584e-05,
"loss": 0.2531,
"step": 971
},
{
"epoch": 6.037383177570093,
"grad_norm": 2.5,
"learning_rate": 4.9948882575548726e-05,
"loss": 0.3004,
"step": 972
},
{
"epoch": 6.043613707165109,
"grad_norm": 3.828125,
"learning_rate": 4.994872498136734e-05,
"loss": 0.4174,
"step": 973
},
{
"epoch": 6.049844236760125,
"grad_norm": 2.28125,
"learning_rate": 4.9948567144878956e-05,
"loss": 0.2079,
"step": 974
},
{
"epoch": 6.05607476635514,
"grad_norm": 3.34375,
"learning_rate": 4.9948409066085086e-05,
"loss": 0.449,
"step": 975
},
{
"epoch": 6.062305295950155,
"grad_norm": 1.6328125,
"learning_rate": 4.994825074498729e-05,
"loss": 0.1914,
"step": 976
},
{
"epoch": 6.068535825545172,
"grad_norm": 1.890625,
"learning_rate": 4.9948092181587095e-05,
"loss": 0.1871,
"step": 977
},
{
"epoch": 6.074766355140187,
"grad_norm": 2.640625,
"learning_rate": 4.994793337588605e-05,
"loss": 0.2913,
"step": 978
},
{
"epoch": 6.080996884735202,
"grad_norm": 3.46875,
"learning_rate": 4.994777432788568e-05,
"loss": 0.5842,
"step": 979
},
{
"epoch": 6.087227414330218,
"grad_norm": 2.6875,
"learning_rate": 4.994761503758753e-05,
"loss": 0.3218,
"step": 980
},
{
"epoch": 6.093457943925234,
"grad_norm": 3.328125,
"learning_rate": 4.994745550499317e-05,
"loss": 0.3007,
"step": 981
},
{
"epoch": 6.099688473520249,
"grad_norm": 3.78125,
"learning_rate": 4.994729573010413e-05,
"loss": 0.4734,
"step": 982
},
{
"epoch": 6.105919003115265,
"grad_norm": 3.0,
"learning_rate": 4.994713571292196e-05,
"loss": 0.4021,
"step": 983
},
{
"epoch": 6.11214953271028,
"grad_norm": 2.203125,
"learning_rate": 4.994697545344823e-05,
"loss": 0.1996,
"step": 984
},
{
"epoch": 6.118380062305296,
"grad_norm": 2.75,
"learning_rate": 4.9946814951684484e-05,
"loss": 0.2544,
"step": 985
},
{
"epoch": 6.1246105919003115,
"grad_norm": 3.046875,
"learning_rate": 4.994665420763229e-05,
"loss": 0.2599,
"step": 986
},
{
"epoch": 6.130841121495327,
"grad_norm": 2.65625,
"learning_rate": 4.994649322129319e-05,
"loss": 0.2825,
"step": 987
},
{
"epoch": 6.137071651090342,
"grad_norm": 2.921875,
"learning_rate": 4.9946331992668775e-05,
"loss": 0.3433,
"step": 988
},
{
"epoch": 6.1433021806853585,
"grad_norm": 3.96875,
"learning_rate": 4.994617052176059e-05,
"loss": 0.5342,
"step": 989
},
{
"epoch": 6.149532710280374,
"grad_norm": 1.734375,
"learning_rate": 4.994600880857022e-05,
"loss": 0.1754,
"step": 990
},
{
"epoch": 6.155763239875389,
"grad_norm": 1.6953125,
"learning_rate": 4.994584685309921e-05,
"loss": 0.1571,
"step": 991
},
{
"epoch": 6.161993769470405,
"grad_norm": 3.140625,
"learning_rate": 4.994568465534916e-05,
"loss": 0.4553,
"step": 992
},
{
"epoch": 6.168224299065421,
"grad_norm": 2.875,
"learning_rate": 4.9945522215321635e-05,
"loss": 0.3567,
"step": 993
},
{
"epoch": 6.174454828660436,
"grad_norm": 2.046875,
"learning_rate": 4.994535953301821e-05,
"loss": 0.2049,
"step": 994
},
{
"epoch": 6.1806853582554515,
"grad_norm": 3.046875,
"learning_rate": 4.9945196608440456e-05,
"loss": 0.299,
"step": 995
},
{
"epoch": 6.186915887850467,
"grad_norm": 2.53125,
"learning_rate": 4.994503344158998e-05,
"loss": 0.3303,
"step": 996
},
{
"epoch": 6.193146417445483,
"grad_norm": 2.1875,
"learning_rate": 4.9944870032468346e-05,
"loss": 0.2206,
"step": 997
},
{
"epoch": 6.1993769470404985,
"grad_norm": 2.8125,
"learning_rate": 4.9944706381077154e-05,
"loss": 0.2574,
"step": 998
},
{
"epoch": 6.205607476635514,
"grad_norm": 2.90625,
"learning_rate": 4.994454248741798e-05,
"loss": 0.3475,
"step": 999
},
{
"epoch": 6.211838006230529,
"grad_norm": 4.1875,
"learning_rate": 4.9944378351492424e-05,
"loss": 0.5048,
"step": 1000
},
{
"epoch": 6.218068535825545,
"grad_norm": 3.015625,
"learning_rate": 4.9944213973302084e-05,
"loss": 0.3856,
"step": 1001
},
{
"epoch": 6.224299065420561,
"grad_norm": 1.9140625,
"learning_rate": 4.994404935284855e-05,
"loss": 0.1851,
"step": 1002
},
{
"epoch": 6.230529595015576,
"grad_norm": 3.0,
"learning_rate": 4.994388449013342e-05,
"loss": 0.2835,
"step": 1003
},
{
"epoch": 6.2367601246105915,
"grad_norm": 2.984375,
"learning_rate": 4.99437193851583e-05,
"loss": 0.373,
"step": 1004
},
{
"epoch": 6.242990654205608,
"grad_norm": 3.71875,
"learning_rate": 4.994355403792479e-05,
"loss": 0.4645,
"step": 1005
},
{
"epoch": 6.249221183800623,
"grad_norm": 2.65625,
"learning_rate": 4.994338844843449e-05,
"loss": 0.3053,
"step": 1006
},
{
"epoch": 6.255451713395638,
"grad_norm": 3.234375,
"learning_rate": 4.9943222616689025e-05,
"loss": 0.414,
"step": 1007
},
{
"epoch": 6.261682242990654,
"grad_norm": 3.546875,
"learning_rate": 4.994305654268999e-05,
"loss": 0.5521,
"step": 1008
},
{
"epoch": 6.26791277258567,
"grad_norm": 3.96875,
"learning_rate": 4.994289022643901e-05,
"loss": 0.3424,
"step": 1009
},
{
"epoch": 6.274143302180685,
"grad_norm": 3.546875,
"learning_rate": 4.99427236679377e-05,
"loss": 0.2711,
"step": 1010
},
{
"epoch": 6.280373831775701,
"grad_norm": 2.296875,
"learning_rate": 4.9942556867187654e-05,
"loss": 0.2164,
"step": 1011
},
{
"epoch": 6.286604361370716,
"grad_norm": 3.09375,
"learning_rate": 4.994238982419052e-05,
"loss": 0.4428,
"step": 1012
},
{
"epoch": 6.292834890965732,
"grad_norm": 2.578125,
"learning_rate": 4.994222253894791e-05,
"loss": 0.2959,
"step": 1013
},
{
"epoch": 6.299065420560748,
"grad_norm": 1.7265625,
"learning_rate": 4.9942055011461455e-05,
"loss": 0.1635,
"step": 1014
},
{
"epoch": 6.305295950155763,
"grad_norm": 1.875,
"learning_rate": 4.994188724173276e-05,
"loss": 0.1931,
"step": 1015
},
{
"epoch": 6.311526479750778,
"grad_norm": 2.125,
"learning_rate": 4.994171922976348e-05,
"loss": 0.2264,
"step": 1016
},
{
"epoch": 6.317757009345795,
"grad_norm": 3.015625,
"learning_rate": 4.994155097555524e-05,
"loss": 0.3373,
"step": 1017
},
{
"epoch": 6.32398753894081,
"grad_norm": 2.546875,
"learning_rate": 4.9941382479109665e-05,
"loss": 0.2805,
"step": 1018
},
{
"epoch": 6.330218068535825,
"grad_norm": 2.5,
"learning_rate": 4.99412137404284e-05,
"loss": 0.2219,
"step": 1019
},
{
"epoch": 6.336448598130841,
"grad_norm": 2.4375,
"learning_rate": 4.994104475951308e-05,
"loss": 0.2394,
"step": 1020
},
{
"epoch": 6.342679127725857,
"grad_norm": 2.40625,
"learning_rate": 4.994087553636535e-05,
"loss": 0.2434,
"step": 1021
},
{
"epoch": 6.348909657320872,
"grad_norm": 3.203125,
"learning_rate": 4.9940706070986854e-05,
"loss": 0.5657,
"step": 1022
},
{
"epoch": 6.355140186915888,
"grad_norm": 1.78125,
"learning_rate": 4.9940536363379234e-05,
"loss": 0.1665,
"step": 1023
},
{
"epoch": 6.361370716510903,
"grad_norm": 2.921875,
"learning_rate": 4.994036641354413e-05,
"loss": 0.3261,
"step": 1024
},
{
"epoch": 6.367601246105919,
"grad_norm": 2.15625,
"learning_rate": 4.9940196221483214e-05,
"loss": 0.2207,
"step": 1025
},
{
"epoch": 6.373831775700935,
"grad_norm": 3.125,
"learning_rate": 4.994002578719812e-05,
"loss": 0.4089,
"step": 1026
},
{
"epoch": 6.38006230529595,
"grad_norm": 3.359375,
"learning_rate": 4.993985511069051e-05,
"loss": 0.4658,
"step": 1027
},
{
"epoch": 6.386292834890965,
"grad_norm": 2.609375,
"learning_rate": 4.993968419196204e-05,
"loss": 0.2922,
"step": 1028
},
{
"epoch": 6.392523364485982,
"grad_norm": 2.46875,
"learning_rate": 4.9939513031014376e-05,
"loss": 0.2515,
"step": 1029
},
{
"epoch": 6.398753894080997,
"grad_norm": 3.0,
"learning_rate": 4.9939341627849176e-05,
"loss": 0.33,
"step": 1030
},
{
"epoch": 6.404984423676012,
"grad_norm": 2.28125,
"learning_rate": 4.993916998246811e-05,
"loss": 0.2433,
"step": 1031
},
{
"epoch": 6.411214953271028,
"grad_norm": 1.484375,
"learning_rate": 4.993899809487284e-05,
"loss": 0.1538,
"step": 1032
},
{
"epoch": 6.417445482866044,
"grad_norm": 2.78125,
"learning_rate": 4.993882596506502e-05,
"loss": 0.3554,
"step": 1033
},
{
"epoch": 6.423676012461059,
"grad_norm": 3.0625,
"learning_rate": 4.9938653593046345e-05,
"loss": 0.2663,
"step": 1034
},
{
"epoch": 6.429906542056075,
"grad_norm": 2.40625,
"learning_rate": 4.993848097881849e-05,
"loss": 0.2282,
"step": 1035
},
{
"epoch": 6.43613707165109,
"grad_norm": 3.1875,
"learning_rate": 4.993830812238311e-05,
"loss": 0.377,
"step": 1036
},
{
"epoch": 6.442367601246106,
"grad_norm": 3.109375,
"learning_rate": 4.9938135023741904e-05,
"loss": 0.308,
"step": 1037
},
{
"epoch": 6.4485981308411215,
"grad_norm": 2.046875,
"learning_rate": 4.993796168289654e-05,
"loss": 0.2159,
"step": 1038
},
{
"epoch": 6.454828660436137,
"grad_norm": 3.03125,
"learning_rate": 4.9937788099848713e-05,
"loss": 0.2375,
"step": 1039
},
{
"epoch": 6.461059190031152,
"grad_norm": 3.3125,
"learning_rate": 4.99376142746001e-05,
"loss": 0.3617,
"step": 1040
},
{
"epoch": 6.4672897196261685,
"grad_norm": 2.59375,
"learning_rate": 4.99374402071524e-05,
"loss": 0.3013,
"step": 1041
},
{
"epoch": 6.473520249221184,
"grad_norm": 1.546875,
"learning_rate": 4.993726589750728e-05,
"loss": 0.16,
"step": 1042
},
{
"epoch": 6.479750778816199,
"grad_norm": 3.453125,
"learning_rate": 4.993709134566646e-05,
"loss": 0.451,
"step": 1043
},
{
"epoch": 6.485981308411215,
"grad_norm": 2.40625,
"learning_rate": 4.993691655163162e-05,
"loss": 0.2965,
"step": 1044
},
{
"epoch": 6.492211838006231,
"grad_norm": 1.734375,
"learning_rate": 4.993674151540446e-05,
"loss": 0.1681,
"step": 1045
},
{
"epoch": 6.498442367601246,
"grad_norm": 2.640625,
"learning_rate": 4.993656623698669e-05,
"loss": 0.3733,
"step": 1046
},
{
"epoch": 6.5046728971962615,
"grad_norm": 3.515625,
"learning_rate": 4.993639071638e-05,
"loss": 0.503,
"step": 1047
},
{
"epoch": 6.510903426791277,
"grad_norm": 2.109375,
"learning_rate": 4.993621495358609e-05,
"loss": 0.2667,
"step": 1048
},
{
"epoch": 6.517133956386293,
"grad_norm": 2.328125,
"learning_rate": 4.993603894860669e-05,
"loss": 0.2469,
"step": 1049
},
{
"epoch": 6.5233644859813085,
"grad_norm": 2.1875,
"learning_rate": 4.993586270144349e-05,
"loss": 0.2132,
"step": 1050
},
{
"epoch": 6.529595015576324,
"grad_norm": 3.921875,
"learning_rate": 4.99356862120982e-05,
"loss": 0.4015,
"step": 1051
},
{
"epoch": 6.535825545171339,
"grad_norm": 3.203125,
"learning_rate": 4.993550948057255e-05,
"loss": 0.3572,
"step": 1052
},
{
"epoch": 6.542056074766355,
"grad_norm": 2.75,
"learning_rate": 4.993533250686824e-05,
"loss": 0.2925,
"step": 1053
},
{
"epoch": 6.548286604361371,
"grad_norm": 2.625,
"learning_rate": 4.9935155290987e-05,
"loss": 0.2946,
"step": 1054
},
{
"epoch": 6.554517133956386,
"grad_norm": 2.734375,
"learning_rate": 4.9934977832930546e-05,
"loss": 0.3983,
"step": 1055
},
{
"epoch": 6.5607476635514015,
"grad_norm": 2.84375,
"learning_rate": 4.99348001327006e-05,
"loss": 0.4005,
"step": 1056
},
{
"epoch": 6.566978193146418,
"grad_norm": 2.953125,
"learning_rate": 4.9934622190298896e-05,
"loss": 0.387,
"step": 1057
},
{
"epoch": 6.573208722741433,
"grad_norm": 2.90625,
"learning_rate": 4.993444400572716e-05,
"loss": 0.3036,
"step": 1058
},
{
"epoch": 6.579439252336448,
"grad_norm": 2.859375,
"learning_rate": 4.993426557898711e-05,
"loss": 0.3134,
"step": 1059
},
{
"epoch": 6.585669781931464,
"grad_norm": 2.171875,
"learning_rate": 4.9934086910080505e-05,
"loss": 0.1983,
"step": 1060
},
{
"epoch": 6.59190031152648,
"grad_norm": 2.8125,
"learning_rate": 4.993390799900904e-05,
"loss": 0.3526,
"step": 1061
},
{
"epoch": 6.598130841121495,
"grad_norm": 3.859375,
"learning_rate": 4.9933728845774496e-05,
"loss": 0.456,
"step": 1062
},
{
"epoch": 6.604361370716511,
"grad_norm": 2.75,
"learning_rate": 4.993354945037858e-05,
"loss": 0.2574,
"step": 1063
},
{
"epoch": 6.610591900311526,
"grad_norm": 1.8984375,
"learning_rate": 4.993336981282306e-05,
"loss": 0.1942,
"step": 1064
},
{
"epoch": 6.616822429906542,
"grad_norm": 3.359375,
"learning_rate": 4.9933189933109657e-05,
"loss": 0.4002,
"step": 1065
},
{
"epoch": 6.623052959501558,
"grad_norm": 3.1875,
"learning_rate": 4.9933009811240136e-05,
"loss": 0.2731,
"step": 1066
},
{
"epoch": 6.629283489096573,
"grad_norm": 3.09375,
"learning_rate": 4.993282944721624e-05,
"loss": 0.308,
"step": 1067
},
{
"epoch": 6.635514018691588,
"grad_norm": 2.109375,
"learning_rate": 4.9932648841039716e-05,
"loss": 0.2025,
"step": 1068
},
{
"epoch": 6.641744548286605,
"grad_norm": 2.796875,
"learning_rate": 4.993246799271233e-05,
"loss": 0.216,
"step": 1069
},
{
"epoch": 6.64797507788162,
"grad_norm": 2.640625,
"learning_rate": 4.9932286902235824e-05,
"loss": 0.2641,
"step": 1070
},
{
"epoch": 6.654205607476635,
"grad_norm": 2.328125,
"learning_rate": 4.993210556961197e-05,
"loss": 0.1993,
"step": 1071
},
{
"epoch": 6.660436137071651,
"grad_norm": 2.1875,
"learning_rate": 4.993192399484251e-05,
"loss": 0.195,
"step": 1072
},
{
"epoch": 6.666666666666667,
"grad_norm": 1.7890625,
"learning_rate": 4.993174217792923e-05,
"loss": 0.1737,
"step": 1073
},
{
"epoch": 6.672897196261682,
"grad_norm": 3.21875,
"learning_rate": 4.993156011887389e-05,
"loss": 0.4515,
"step": 1074
},
{
"epoch": 6.679127725856698,
"grad_norm": 2.75,
"learning_rate": 4.9931377817678246e-05,
"loss": 0.3959,
"step": 1075
},
{
"epoch": 6.685358255451713,
"grad_norm": 3.53125,
"learning_rate": 4.993119527434408e-05,
"loss": 0.3409,
"step": 1076
},
{
"epoch": 6.691588785046729,
"grad_norm": 2.9375,
"learning_rate": 4.993101248887316e-05,
"loss": 0.3399,
"step": 1077
},
{
"epoch": 6.697819314641745,
"grad_norm": 2.734375,
"learning_rate": 4.993082946126727e-05,
"loss": 0.3476,
"step": 1078
},
{
"epoch": 6.70404984423676,
"grad_norm": 2.96875,
"learning_rate": 4.9930646191528175e-05,
"loss": 0.2943,
"step": 1079
},
{
"epoch": 6.710280373831775,
"grad_norm": 2.046875,
"learning_rate": 4.993046267965766e-05,
"loss": 0.1799,
"step": 1080
},
{
"epoch": 6.716510903426792,
"grad_norm": 3.109375,
"learning_rate": 4.993027892565751e-05,
"loss": 0.4703,
"step": 1081
},
{
"epoch": 6.722741433021807,
"grad_norm": 2.609375,
"learning_rate": 4.9930094929529506e-05,
"loss": 0.2351,
"step": 1082
},
{
"epoch": 6.728971962616822,
"grad_norm": 2.515625,
"learning_rate": 4.992991069127544e-05,
"loss": 0.225,
"step": 1083
},
{
"epoch": 6.735202492211838,
"grad_norm": 2.59375,
"learning_rate": 4.992972621089709e-05,
"loss": 0.2461,
"step": 1084
},
{
"epoch": 6.741433021806854,
"grad_norm": 2.609375,
"learning_rate": 4.9929541488396266e-05,
"loss": 0.34,
"step": 1085
},
{
"epoch": 6.747663551401869,
"grad_norm": 2.609375,
"learning_rate": 4.9929356523774745e-05,
"loss": 0.2672,
"step": 1086
},
{
"epoch": 6.753894080996885,
"grad_norm": 2.640625,
"learning_rate": 4.992917131703433e-05,
"loss": 0.2723,
"step": 1087
},
{
"epoch": 6.7601246105919,
"grad_norm": 1.5390625,
"learning_rate": 4.992898586817682e-05,
"loss": 0.1486,
"step": 1088
},
{
"epoch": 6.766355140186916,
"grad_norm": 2.3125,
"learning_rate": 4.9928800177204025e-05,
"loss": 0.2868,
"step": 1089
},
{
"epoch": 6.7725856697819315,
"grad_norm": 2.359375,
"learning_rate": 4.9928614244117735e-05,
"loss": 0.2993,
"step": 1090
},
{
"epoch": 6.778816199376947,
"grad_norm": 3.3125,
"learning_rate": 4.9928428068919755e-05,
"loss": 0.4439,
"step": 1091
},
{
"epoch": 6.785046728971962,
"grad_norm": 3.53125,
"learning_rate": 4.99282416516119e-05,
"loss": 0.4064,
"step": 1092
},
{
"epoch": 6.7912772585669785,
"grad_norm": 2.921875,
"learning_rate": 4.9928054992195985e-05,
"loss": 0.335,
"step": 1093
},
{
"epoch": 6.797507788161994,
"grad_norm": 3.015625,
"learning_rate": 4.992786809067381e-05,
"loss": 0.4055,
"step": 1094
},
{
"epoch": 6.803738317757009,
"grad_norm": 1.9765625,
"learning_rate": 4.99276809470472e-05,
"loss": 0.1942,
"step": 1095
},
{
"epoch": 6.809968847352025,
"grad_norm": 1.71875,
"learning_rate": 4.9927493561317974e-05,
"loss": 0.1657,
"step": 1096
},
{
"epoch": 6.816199376947041,
"grad_norm": 3.125,
"learning_rate": 4.992730593348794e-05,
"loss": 0.3272,
"step": 1097
},
{
"epoch": 6.822429906542056,
"grad_norm": 2.96875,
"learning_rate": 4.992711806355893e-05,
"loss": 0.3497,
"step": 1098
},
{
"epoch": 6.8286604361370715,
"grad_norm": 2.671875,
"learning_rate": 4.992692995153276e-05,
"loss": 0.3556,
"step": 1099
},
{
"epoch": 6.834890965732087,
"grad_norm": 2.609375,
"learning_rate": 4.992674159741128e-05,
"loss": 0.3238,
"step": 1100
},
{
"epoch": 6.841121495327103,
"grad_norm": 2.03125,
"learning_rate": 4.992655300119629e-05,
"loss": 0.168,
"step": 1101
},
{
"epoch": 6.8473520249221185,
"grad_norm": 3.125,
"learning_rate": 4.992636416288964e-05,
"loss": 0.2816,
"step": 1102
},
{
"epoch": 6.853582554517134,
"grad_norm": 1.3125,
"learning_rate": 4.992617508249316e-05,
"loss": 0.1358,
"step": 1103
},
{
"epoch": 6.859813084112149,
"grad_norm": 4.1875,
"learning_rate": 4.992598576000868e-05,
"loss": 0.455,
"step": 1104
},
{
"epoch": 6.866043613707165,
"grad_norm": 1.953125,
"learning_rate": 4.9925796195438044e-05,
"loss": 0.2512,
"step": 1105
},
{
"epoch": 6.872274143302181,
"grad_norm": 2.328125,
"learning_rate": 4.99256063887831e-05,
"loss": 0.2027,
"step": 1106
},
{
"epoch": 6.878504672897196,
"grad_norm": 2.9375,
"learning_rate": 4.992541634004567e-05,
"loss": 0.3096,
"step": 1107
},
{
"epoch": 6.8847352024922115,
"grad_norm": 1.6875,
"learning_rate": 4.992522604922762e-05,
"loss": 0.1803,
"step": 1108
},
{
"epoch": 6.890965732087228,
"grad_norm": 2.5625,
"learning_rate": 4.9925035516330796e-05,
"loss": 0.2641,
"step": 1109
},
{
"epoch": 6.897196261682243,
"grad_norm": 2.609375,
"learning_rate": 4.992484474135705e-05,
"loss": 0.3003,
"step": 1110
},
{
"epoch": 6.9034267912772584,
"grad_norm": 3.0,
"learning_rate": 4.9924653724308225e-05,
"loss": 0.2993,
"step": 1111
},
{
"epoch": 6.909657320872274,
"grad_norm": 2.140625,
"learning_rate": 4.9924462465186174e-05,
"loss": 0.2023,
"step": 1112
},
{
"epoch": 6.91588785046729,
"grad_norm": 2.21875,
"learning_rate": 4.992427096399277e-05,
"loss": 0.2736,
"step": 1113
},
{
"epoch": 6.922118380062305,
"grad_norm": 2.25,
"learning_rate": 4.9924079220729855e-05,
"loss": 0.2094,
"step": 1114
},
{
"epoch": 6.928348909657321,
"grad_norm": 1.5390625,
"learning_rate": 4.992388723539931e-05,
"loss": 0.169,
"step": 1115
},
{
"epoch": 6.934579439252336,
"grad_norm": 2.546875,
"learning_rate": 4.992369500800299e-05,
"loss": 0.4083,
"step": 1116
},
{
"epoch": 6.940809968847352,
"grad_norm": 2.578125,
"learning_rate": 4.992350253854276e-05,
"loss": 0.2591,
"step": 1117
},
{
"epoch": 6.947040498442368,
"grad_norm": 3.484375,
"learning_rate": 4.992330982702048e-05,
"loss": 0.3221,
"step": 1118
},
{
"epoch": 6.953271028037383,
"grad_norm": 2.25,
"learning_rate": 4.992311687343804e-05,
"loss": 0.2375,
"step": 1119
},
{
"epoch": 6.959501557632398,
"grad_norm": 2.140625,
"learning_rate": 4.992292367779731e-05,
"loss": 0.2425,
"step": 1120
},
{
"epoch": 6.965732087227415,
"grad_norm": 2.125,
"learning_rate": 4.992273024010016e-05,
"loss": 0.2376,
"step": 1121
},
{
"epoch": 6.97196261682243,
"grad_norm": 3.484375,
"learning_rate": 4.9922536560348473e-05,
"loss": 0.4243,
"step": 1122
},
{
"epoch": 6.978193146417445,
"grad_norm": 2.125,
"learning_rate": 4.9922342638544125e-05,
"loss": 0.1816,
"step": 1123
},
{
"epoch": 6.984423676012461,
"grad_norm": 3.40625,
"learning_rate": 4.992214847468901e-05,
"loss": 0.3521,
"step": 1124
},
{
"epoch": 6.990654205607477,
"grad_norm": 2.546875,
"learning_rate": 4.992195406878501e-05,
"loss": 0.3053,
"step": 1125
},
{
"epoch": 6.996884735202492,
"grad_norm": 2.03125,
"learning_rate": 4.9921759420834e-05,
"loss": 0.1958,
"step": 1126
},
{
"epoch": 7.0,
"grad_norm": 2.4375,
"learning_rate": 4.9921564530837894e-05,
"loss": 0.2773,
"step": 1127
},
{
"epoch": 7.006230529595015,
"grad_norm": 3.015625,
"learning_rate": 4.992136939879856e-05,
"loss": 0.3725,
"step": 1128
},
{
"epoch": 7.012461059190031,
"grad_norm": 3.15625,
"learning_rate": 4.992117402471791e-05,
"loss": 0.3355,
"step": 1129
},
{
"epoch": 7.018691588785047,
"grad_norm": 2.90625,
"learning_rate": 4.9920978408597837e-05,
"loss": 0.2413,
"step": 1130
},
{
"epoch": 7.024922118380062,
"grad_norm": 3.828125,
"learning_rate": 4.9920782550440236e-05,
"loss": 0.3873,
"step": 1131
},
{
"epoch": 7.031152647975078,
"grad_norm": 1.75,
"learning_rate": 4.992058645024702e-05,
"loss": 0.1634,
"step": 1132
},
{
"epoch": 7.037383177570093,
"grad_norm": 2.296875,
"learning_rate": 4.9920390108020075e-05,
"loss": 0.2301,
"step": 1133
},
{
"epoch": 7.043613707165109,
"grad_norm": 1.8125,
"learning_rate": 4.992019352376133e-05,
"loss": 0.1679,
"step": 1134
},
{
"epoch": 7.049844236760125,
"grad_norm": 2.125,
"learning_rate": 4.9919996697472684e-05,
"loss": 0.2263,
"step": 1135
},
{
"epoch": 7.05607476635514,
"grad_norm": 3.046875,
"learning_rate": 4.9919799629156036e-05,
"loss": 0.3071,
"step": 1136
},
{
"epoch": 7.062305295950155,
"grad_norm": 3.515625,
"learning_rate": 4.991960231881333e-05,
"loss": 0.3405,
"step": 1137
},
{
"epoch": 7.068535825545172,
"grad_norm": 2.5625,
"learning_rate": 4.9919404766446456e-05,
"loss": 0.2193,
"step": 1138
},
{
"epoch": 7.074766355140187,
"grad_norm": 2.390625,
"learning_rate": 4.991920697205734e-05,
"loss": 0.2261,
"step": 1139
},
{
"epoch": 7.080996884735202,
"grad_norm": 2.96875,
"learning_rate": 4.9919008935647904e-05,
"loss": 0.3618,
"step": 1140
},
{
"epoch": 7.087227414330218,
"grad_norm": 2.765625,
"learning_rate": 4.9918810657220075e-05,
"loss": 0.2416,
"step": 1141
},
{
"epoch": 7.093457943925234,
"grad_norm": 2.359375,
"learning_rate": 4.9918612136775776e-05,
"loss": 0.1814,
"step": 1142
},
{
"epoch": 7.099688473520249,
"grad_norm": 3.09375,
"learning_rate": 4.991841337431694e-05,
"loss": 0.3431,
"step": 1143
},
{
"epoch": 7.105919003115265,
"grad_norm": 3.625,
"learning_rate": 4.991821436984548e-05,
"loss": 0.3351,
"step": 1144
},
{
"epoch": 7.11214953271028,
"grad_norm": 3.8125,
"learning_rate": 4.9918015123363355e-05,
"loss": 0.3975,
"step": 1145
},
{
"epoch": 7.118380062305296,
"grad_norm": 2.9375,
"learning_rate": 4.991781563487248e-05,
"loss": 0.3844,
"step": 1146
},
{
"epoch": 7.1246105919003115,
"grad_norm": 2.171875,
"learning_rate": 4.991761590437479e-05,
"loss": 0.1916,
"step": 1147
},
{
"epoch": 7.130841121495327,
"grad_norm": 2.890625,
"learning_rate": 4.991741593187225e-05,
"loss": 0.4398,
"step": 1148
},
{
"epoch": 7.137071651090342,
"grad_norm": 3.15625,
"learning_rate": 4.991721571736677e-05,
"loss": 0.4154,
"step": 1149
},
{
"epoch": 7.1433021806853585,
"grad_norm": 2.078125,
"learning_rate": 4.991701526086032e-05,
"loss": 0.1655,
"step": 1150
},
{
"epoch": 7.149532710280374,
"grad_norm": 2.3125,
"learning_rate": 4.991681456235483e-05,
"loss": 0.2062,
"step": 1151
},
{
"epoch": 7.155763239875389,
"grad_norm": 1.96875,
"learning_rate": 4.9916613621852265e-05,
"loss": 0.1777,
"step": 1152
},
{
"epoch": 7.161993769470405,
"grad_norm": 2.734375,
"learning_rate": 4.991641243935457e-05,
"loss": 0.2217,
"step": 1153
},
{
"epoch": 7.168224299065421,
"grad_norm": 2.875,
"learning_rate": 4.991621101486368e-05,
"loss": 0.2871,
"step": 1154
},
{
"epoch": 7.174454828660436,
"grad_norm": 2.921875,
"learning_rate": 4.9916009348381574e-05,
"loss": 0.2615,
"step": 1155
},
{
"epoch": 7.1806853582554515,
"grad_norm": 3.078125,
"learning_rate": 4.9915807439910213e-05,
"loss": 0.34,
"step": 1156
},
{
"epoch": 7.186915887850467,
"grad_norm": 3.046875,
"learning_rate": 4.991560528945154e-05,
"loss": 0.2948,
"step": 1157
},
{
"epoch": 7.193146417445483,
"grad_norm": 2.640625,
"learning_rate": 4.991540289700754e-05,
"loss": 0.3547,
"step": 1158
},
{
"epoch": 7.1993769470404985,
"grad_norm": 2.28125,
"learning_rate": 4.991520026258015e-05,
"loss": 0.2638,
"step": 1159
},
{
"epoch": 7.205607476635514,
"grad_norm": 1.84375,
"learning_rate": 4.991499738617137e-05,
"loss": 0.1681,
"step": 1160
},
{
"epoch": 7.211838006230529,
"grad_norm": 2.453125,
"learning_rate": 4.9914794267783146e-05,
"loss": 0.2091,
"step": 1161
},
{
"epoch": 7.218068535825545,
"grad_norm": 2.40625,
"learning_rate": 4.991459090741746e-05,
"loss": 0.2552,
"step": 1162
},
{
"epoch": 7.224299065420561,
"grad_norm": 1.5625,
"learning_rate": 4.991438730507629e-05,
"loss": 0.1431,
"step": 1163
},
{
"epoch": 7.230529595015576,
"grad_norm": 2.65625,
"learning_rate": 4.991418346076161e-05,
"loss": 0.2409,
"step": 1164
},
{
"epoch": 7.2367601246105915,
"grad_norm": 2.734375,
"learning_rate": 4.991397937447541e-05,
"loss": 0.3428,
"step": 1165
},
{
"epoch": 7.242990654205608,
"grad_norm": 4.03125,
"learning_rate": 4.991377504621965e-05,
"loss": 0.5163,
"step": 1166
},
{
"epoch": 7.249221183800623,
"grad_norm": 3.25,
"learning_rate": 4.991357047599632e-05,
"loss": 0.2896,
"step": 1167
},
{
"epoch": 7.255451713395638,
"grad_norm": 2.21875,
"learning_rate": 4.991336566380742e-05,
"loss": 0.301,
"step": 1168
},
{
"epoch": 7.261682242990654,
"grad_norm": 2.46875,
"learning_rate": 4.991316060965494e-05,
"loss": 0.2329,
"step": 1169
},
{
"epoch": 7.26791277258567,
"grad_norm": 3.15625,
"learning_rate": 4.991295531354085e-05,
"loss": 0.3114,
"step": 1170
},
{
"epoch": 7.274143302180685,
"grad_norm": 2.390625,
"learning_rate": 4.991274977546717e-05,
"loss": 0.321,
"step": 1171
},
{
"epoch": 7.280373831775701,
"grad_norm": 2.328125,
"learning_rate": 4.9912543995435875e-05,
"loss": 0.2386,
"step": 1172
},
{
"epoch": 7.286604361370716,
"grad_norm": 2.328125,
"learning_rate": 4.9912337973448984e-05,
"loss": 0.2128,
"step": 1173
},
{
"epoch": 7.292834890965732,
"grad_norm": 3.171875,
"learning_rate": 4.991213170950848e-05,
"loss": 0.281,
"step": 1174
},
{
"epoch": 7.299065420560748,
"grad_norm": 2.28125,
"learning_rate": 4.9911925203616374e-05,
"loss": 0.2791,
"step": 1175
},
{
"epoch": 7.305295950155763,
"grad_norm": 3.15625,
"learning_rate": 4.991171845577467e-05,
"loss": 0.2705,
"step": 1176
},
{
"epoch": 7.311526479750778,
"grad_norm": 4.125,
"learning_rate": 4.9911511465985386e-05,
"loss": 0.4424,
"step": 1177
},
{
"epoch": 7.317757009345795,
"grad_norm": 2.25,
"learning_rate": 4.9911304234250514e-05,
"loss": 0.2078,
"step": 1178
},
{
"epoch": 7.32398753894081,
"grad_norm": 2.78125,
"learning_rate": 4.9911096760572066e-05,
"loss": 0.2911,
"step": 1179
},
{
"epoch": 7.330218068535825,
"grad_norm": 2.9375,
"learning_rate": 4.991088904495208e-05,
"loss": 0.2789,
"step": 1180
},
{
"epoch": 7.336448598130841,
"grad_norm": 2.671875,
"learning_rate": 4.9910681087392554e-05,
"loss": 0.2471,
"step": 1181
},
{
"epoch": 7.342679127725857,
"grad_norm": 2.453125,
"learning_rate": 4.9910472887895524e-05,
"loss": 0.221,
"step": 1182
},
{
"epoch": 7.348909657320872,
"grad_norm": 3.90625,
"learning_rate": 4.9910264446462986e-05,
"loss": 0.5766,
"step": 1183
},
{
"epoch": 7.355140186915888,
"grad_norm": 2.71875,
"learning_rate": 4.991005576309699e-05,
"loss": 0.2471,
"step": 1184
},
{
"epoch": 7.361370716510903,
"grad_norm": 3.515625,
"learning_rate": 4.990984683779955e-05,
"loss": 0.3868,
"step": 1185
},
{
"epoch": 7.367601246105919,
"grad_norm": 3.515625,
"learning_rate": 4.99096376705727e-05,
"loss": 0.4757,
"step": 1186
},
{
"epoch": 7.373831775700935,
"grad_norm": 2.359375,
"learning_rate": 4.990942826141847e-05,
"loss": 0.2152,
"step": 1187
},
{
"epoch": 7.38006230529595,
"grad_norm": 2.234375,
"learning_rate": 4.990921861033889e-05,
"loss": 0.1922,
"step": 1188
},
{
"epoch": 7.386292834890965,
"grad_norm": 3.515625,
"learning_rate": 4.9909008717336e-05,
"loss": 0.3685,
"step": 1189
},
{
"epoch": 7.392523364485982,
"grad_norm": 2.546875,
"learning_rate": 4.9908798582411845e-05,
"loss": 0.2018,
"step": 1190
},
{
"epoch": 7.398753894080997,
"grad_norm": 1.5859375,
"learning_rate": 4.990858820556845e-05,
"loss": 0.147,
"step": 1191
},
{
"epoch": 7.404984423676012,
"grad_norm": 3.421875,
"learning_rate": 4.990837758680787e-05,
"loss": 0.4266,
"step": 1192
},
{
"epoch": 7.411214953271028,
"grad_norm": 2.59375,
"learning_rate": 4.990816672613216e-05,
"loss": 0.2978,
"step": 1193
},
{
"epoch": 7.417445482866044,
"grad_norm": 2.96875,
"learning_rate": 4.990795562354333e-05,
"loss": 0.3059,
"step": 1194
},
{
"epoch": 7.423676012461059,
"grad_norm": 2.40625,
"learning_rate": 4.9907744279043475e-05,
"loss": 0.2049,
"step": 1195
},
{
"epoch": 7.429906542056075,
"grad_norm": 2.796875,
"learning_rate": 4.990753269263463e-05,
"loss": 0.2724,
"step": 1196
},
{
"epoch": 7.43613707165109,
"grad_norm": 2.4375,
"learning_rate": 4.990732086431884e-05,
"loss": 0.3533,
"step": 1197
},
{
"epoch": 7.442367601246106,
"grad_norm": 3.09375,
"learning_rate": 4.9907108794098176e-05,
"loss": 0.3799,
"step": 1198
},
{
"epoch": 7.4485981308411215,
"grad_norm": 3.0,
"learning_rate": 4.990689648197469e-05,
"loss": 0.2508,
"step": 1199
},
{
"epoch": 7.454828660436137,
"grad_norm": 3.140625,
"learning_rate": 4.9906683927950446e-05,
"loss": 0.2877,
"step": 1200
},
{
"epoch": 7.461059190031152,
"grad_norm": 2.953125,
"learning_rate": 4.990647113202751e-05,
"loss": 0.3226,
"step": 1201
},
{
"epoch": 7.4672897196261685,
"grad_norm": 3.046875,
"learning_rate": 4.9906258094207945e-05,
"loss": 0.4022,
"step": 1202
},
{
"epoch": 7.473520249221184,
"grad_norm": 1.7265625,
"learning_rate": 4.9906044814493824e-05,
"loss": 0.1485,
"step": 1203
},
{
"epoch": 7.479750778816199,
"grad_norm": 2.296875,
"learning_rate": 4.990583129288722e-05,
"loss": 0.1904,
"step": 1204
},
{
"epoch": 7.485981308411215,
"grad_norm": 2.546875,
"learning_rate": 4.9905617529390203e-05,
"loss": 0.2834,
"step": 1205
},
{
"epoch": 7.492211838006231,
"grad_norm": 2.984375,
"learning_rate": 4.9905403524004844e-05,
"loss": 0.3213,
"step": 1206
},
{
"epoch": 7.498442367601246,
"grad_norm": 3.03125,
"learning_rate": 4.990518927673324e-05,
"loss": 0.2763,
"step": 1207
},
{
"epoch": 7.5046728971962615,
"grad_norm": 3.25,
"learning_rate": 4.9904974787577455e-05,
"loss": 0.3346,
"step": 1208
},
{
"epoch": 7.510903426791277,
"grad_norm": 2.46875,
"learning_rate": 4.9904760056539565e-05,
"loss": 0.2467,
"step": 1209
},
{
"epoch": 7.517133956386293,
"grad_norm": 3.5625,
"learning_rate": 4.990454508362168e-05,
"loss": 0.5957,
"step": 1210
},
{
"epoch": 7.5233644859813085,
"grad_norm": 2.53125,
"learning_rate": 4.9904329868825874e-05,
"loss": 0.3154,
"step": 1211
},
{
"epoch": 7.529595015576324,
"grad_norm": 2.578125,
"learning_rate": 4.990411441215423e-05,
"loss": 0.2026,
"step": 1212
},
{
"epoch": 7.535825545171339,
"grad_norm": 3.734375,
"learning_rate": 4.9903898713608856e-05,
"loss": 0.5806,
"step": 1213
},
{
"epoch": 7.542056074766355,
"grad_norm": 1.5859375,
"learning_rate": 4.990368277319183e-05,
"loss": 0.1491,
"step": 1214
},
{
"epoch": 7.548286604361371,
"grad_norm": 2.40625,
"learning_rate": 4.990346659090527e-05,
"loss": 0.2684,
"step": 1215
},
{
"epoch": 7.554517133956386,
"grad_norm": 3.3125,
"learning_rate": 4.990325016675126e-05,
"loss": 0.4056,
"step": 1216
},
{
"epoch": 7.5607476635514015,
"grad_norm": 2.015625,
"learning_rate": 4.9903033500731906e-05,
"loss": 0.1955,
"step": 1217
},
{
"epoch": 7.566978193146418,
"grad_norm": 2.328125,
"learning_rate": 4.990281659284931e-05,
"loss": 0.299,
"step": 1218
},
{
"epoch": 7.573208722741433,
"grad_norm": 3.296875,
"learning_rate": 4.9902599443105584e-05,
"loss": 0.448,
"step": 1219
},
{
"epoch": 7.579439252336448,
"grad_norm": 1.7265625,
"learning_rate": 4.990238205150284e-05,
"loss": 0.155,
"step": 1220
},
{
"epoch": 7.585669781931464,
"grad_norm": 3.8125,
"learning_rate": 4.990216441804318e-05,
"loss": 0.5479,
"step": 1221
},
{
"epoch": 7.59190031152648,
"grad_norm": 1.625,
"learning_rate": 4.9901946542728714e-05,
"loss": 0.1643,
"step": 1222
},
{
"epoch": 7.598130841121495,
"grad_norm": 3.125,
"learning_rate": 4.9901728425561564e-05,
"loss": 0.3774,
"step": 1223
},
{
"epoch": 7.604361370716511,
"grad_norm": 2.859375,
"learning_rate": 4.990151006654386e-05,
"loss": 0.35,
"step": 1224
},
{
"epoch": 7.610591900311526,
"grad_norm": 3.34375,
"learning_rate": 4.9901291465677706e-05,
"loss": 0.392,
"step": 1225
},
{
"epoch": 7.616822429906542,
"grad_norm": 2.4375,
"learning_rate": 4.990107262296524e-05,
"loss": 0.2132,
"step": 1226
},
{
"epoch": 7.623052959501558,
"grad_norm": 2.25,
"learning_rate": 4.990085353840857e-05,
"loss": 0.2004,
"step": 1227
},
{
"epoch": 7.629283489096573,
"grad_norm": 2.46875,
"learning_rate": 4.9900634212009834e-05,
"loss": 0.2055,
"step": 1228
},
{
"epoch": 7.635514018691588,
"grad_norm": 4.21875,
"learning_rate": 4.990041464377117e-05,
"loss": 0.5485,
"step": 1229
},
{
"epoch": 7.641744548286605,
"grad_norm": 1.7109375,
"learning_rate": 4.990019483369469e-05,
"loss": 0.1708,
"step": 1230
},
{
"epoch": 7.64797507788162,
"grad_norm": 3.015625,
"learning_rate": 4.9899974781782546e-05,
"loss": 0.2506,
"step": 1231
},
{
"epoch": 7.654205607476635,
"grad_norm": 2.25,
"learning_rate": 4.9899754488036864e-05,
"loss": 0.2005,
"step": 1232
},
{
"epoch": 7.660436137071651,
"grad_norm": 2.109375,
"learning_rate": 4.98995339524598e-05,
"loss": 0.1867,
"step": 1233
},
{
"epoch": 7.666666666666667,
"grad_norm": 2.421875,
"learning_rate": 4.9899313175053474e-05,
"loss": 0.2405,
"step": 1234
},
{
"epoch": 7.672897196261682,
"grad_norm": 1.84375,
"learning_rate": 4.989909215582005e-05,
"loss": 0.1721,
"step": 1235
},
{
"epoch": 7.679127725856698,
"grad_norm": 3.453125,
"learning_rate": 4.9898870894761664e-05,
"loss": 0.4021,
"step": 1236
},
{
"epoch": 7.685358255451713,
"grad_norm": 2.140625,
"learning_rate": 4.989864939188046e-05,
"loss": 0.1931,
"step": 1237
},
{
"epoch": 7.691588785046729,
"grad_norm": 1.5546875,
"learning_rate": 4.9898427647178604e-05,
"loss": 0.1477,
"step": 1238
},
{
"epoch": 7.697819314641745,
"grad_norm": 2.71875,
"learning_rate": 4.989820566065824e-05,
"loss": 0.3643,
"step": 1239
},
{
"epoch": 7.70404984423676,
"grad_norm": 2.515625,
"learning_rate": 4.989798343232151e-05,
"loss": 0.2022,
"step": 1240
},
{
"epoch": 7.710280373831775,
"grad_norm": 2.375,
"learning_rate": 4.989776096217061e-05,
"loss": 0.1852,
"step": 1241
},
{
"epoch": 7.716510903426792,
"grad_norm": 2.34375,
"learning_rate": 4.989753825020766e-05,
"loss": 0.2476,
"step": 1242
},
{
"epoch": 7.722741433021807,
"grad_norm": 3.203125,
"learning_rate": 4.989731529643486e-05,
"loss": 0.5024,
"step": 1243
},
{
"epoch": 7.728971962616822,
"grad_norm": 3.34375,
"learning_rate": 4.989709210085435e-05,
"loss": 0.3194,
"step": 1244
},
{
"epoch": 7.735202492211838,
"grad_norm": 2.546875,
"learning_rate": 4.9896868663468296e-05,
"loss": 0.239,
"step": 1245
},
{
"epoch": 7.741433021806854,
"grad_norm": 3.015625,
"learning_rate": 4.989664498427889e-05,
"loss": 0.2533,
"step": 1246
},
{
"epoch": 7.747663551401869,
"grad_norm": 1.9140625,
"learning_rate": 4.9896421063288286e-05,
"loss": 0.1771,
"step": 1247
},
{
"epoch": 7.753894080996885,
"grad_norm": 2.0625,
"learning_rate": 4.989619690049866e-05,
"loss": 0.2893,
"step": 1248
},
{
"epoch": 7.7601246105919,
"grad_norm": 2.59375,
"learning_rate": 4.98959724959122e-05,
"loss": 0.2171,
"step": 1249
},
{
"epoch": 7.766355140186916,
"grad_norm": 2.3125,
"learning_rate": 4.989574784953107e-05,
"loss": 0.1915,
"step": 1250
},
{
"epoch": 7.7725856697819315,
"grad_norm": 2.84375,
"learning_rate": 4.989552296135747e-05,
"loss": 0.3356,
"step": 1251
},
{
"epoch": 7.778816199376947,
"grad_norm": 2.453125,
"learning_rate": 4.989529783139356e-05,
"loss": 0.227,
"step": 1252
},
{
"epoch": 7.785046728971962,
"grad_norm": 1.9296875,
"learning_rate": 4.9895072459641554e-05,
"loss": 0.1737,
"step": 1253
},
{
"epoch": 7.7912772585669785,
"grad_norm": 2.21875,
"learning_rate": 4.9894846846103625e-05,
"loss": 0.2202,
"step": 1254
},
{
"epoch": 7.797507788161994,
"grad_norm": 1.984375,
"learning_rate": 4.989462099078197e-05,
"loss": 0.1684,
"step": 1255
},
{
"epoch": 7.803738317757009,
"grad_norm": 2.484375,
"learning_rate": 4.989439489367877e-05,
"loss": 0.1992,
"step": 1256
},
{
"epoch": 7.809968847352025,
"grad_norm": 2.015625,
"learning_rate": 4.989416855479624e-05,
"loss": 0.2001,
"step": 1257
},
{
"epoch": 7.816199376947041,
"grad_norm": 2.28125,
"learning_rate": 4.989394197413656e-05,
"loss": 0.2384,
"step": 1258
},
{
"epoch": 7.822429906542056,
"grad_norm": 2.3125,
"learning_rate": 4.989371515170195e-05,
"loss": 0.1858,
"step": 1259
},
{
"epoch": 7.8286604361370715,
"grad_norm": 1.8046875,
"learning_rate": 4.989348808749459e-05,
"loss": 0.1784,
"step": 1260
},
{
"epoch": 7.834890965732087,
"grad_norm": 3.4375,
"learning_rate": 4.989326078151671e-05,
"loss": 0.4752,
"step": 1261
},
{
"epoch": 7.841121495327103,
"grad_norm": 1.6640625,
"learning_rate": 4.98930332337705e-05,
"loss": 0.1378,
"step": 1262
},
{
"epoch": 7.8473520249221185,
"grad_norm": 3.109375,
"learning_rate": 4.9892805444258174e-05,
"loss": 0.2625,
"step": 1263
},
{
"epoch": 7.853582554517134,
"grad_norm": 1.84375,
"learning_rate": 4.9892577412981946e-05,
"loss": 0.1518,
"step": 1264
},
{
"epoch": 7.859813084112149,
"grad_norm": 2.90625,
"learning_rate": 4.989234913994403e-05,
"loss": 0.2871,
"step": 1265
},
{
"epoch": 7.866043613707165,
"grad_norm": 1.765625,
"learning_rate": 4.989212062514664e-05,
"loss": 0.1656,
"step": 1266
},
{
"epoch": 7.872274143302181,
"grad_norm": 1.90625,
"learning_rate": 4.989189186859201e-05,
"loss": 0.1925,
"step": 1267
},
{
"epoch": 7.878504672897196,
"grad_norm": 2.25,
"learning_rate": 4.989166287028234e-05,
"loss": 0.2037,
"step": 1268
},
{
"epoch": 7.8847352024922115,
"grad_norm": 1.671875,
"learning_rate": 4.989143363021987e-05,
"loss": 0.1539,
"step": 1269
},
{
"epoch": 7.890965732087228,
"grad_norm": 2.46875,
"learning_rate": 4.989120414840681e-05,
"loss": 0.2164,
"step": 1270
},
{
"epoch": 7.897196261682243,
"grad_norm": 2.90625,
"learning_rate": 4.989097442484542e-05,
"loss": 0.3652,
"step": 1271
},
{
"epoch": 7.9034267912772584,
"grad_norm": 2.46875,
"learning_rate": 4.989074445953789e-05,
"loss": 0.2309,
"step": 1272
},
{
"epoch": 7.909657320872274,
"grad_norm": 1.6171875,
"learning_rate": 4.9890514252486486e-05,
"loss": 0.1515,
"step": 1273
},
{
"epoch": 7.91588785046729,
"grad_norm": 2.609375,
"learning_rate": 4.989028380369343e-05,
"loss": 0.2918,
"step": 1274
},
{
"epoch": 7.922118380062305,
"grad_norm": 3.09375,
"learning_rate": 4.989005311316096e-05,
"loss": 0.4187,
"step": 1275
},
{
"epoch": 7.928348909657321,
"grad_norm": 3.0,
"learning_rate": 4.9889822180891324e-05,
"loss": 0.2939,
"step": 1276
},
{
"epoch": 7.934579439252336,
"grad_norm": 1.7265625,
"learning_rate": 4.988959100688675e-05,
"loss": 0.1672,
"step": 1277
},
{
"epoch": 7.940809968847352,
"grad_norm": 2.28125,
"learning_rate": 4.9889359591149496e-05,
"loss": 0.2157,
"step": 1278
},
{
"epoch": 7.947040498442368,
"grad_norm": 2.296875,
"learning_rate": 4.988912793368181e-05,
"loss": 0.2388,
"step": 1279
},
{
"epoch": 7.953271028037383,
"grad_norm": 2.890625,
"learning_rate": 4.9888896034485935e-05,
"loss": 0.3006,
"step": 1280
},
{
"epoch": 7.959501557632398,
"grad_norm": 3.46875,
"learning_rate": 4.988866389356413e-05,
"loss": 0.3776,
"step": 1281
},
{
"epoch": 7.965732087227415,
"grad_norm": 2.296875,
"learning_rate": 4.9888431510918645e-05,
"loss": 0.2468,
"step": 1282
},
{
"epoch": 7.97196261682243,
"grad_norm": 2.578125,
"learning_rate": 4.988819888655174e-05,
"loss": 0.2838,
"step": 1283
},
{
"epoch": 7.978193146417445,
"grad_norm": 2.234375,
"learning_rate": 4.988796602046567e-05,
"loss": 0.2038,
"step": 1284
},
{
"epoch": 7.984423676012461,
"grad_norm": 2.125,
"learning_rate": 4.98877329126627e-05,
"loss": 0.1826,
"step": 1285
},
{
"epoch": 7.990654205607477,
"grad_norm": 2.9375,
"learning_rate": 4.988749956314509e-05,
"loss": 0.3048,
"step": 1286
},
{
"epoch": 7.996884735202492,
"grad_norm": 2.484375,
"learning_rate": 4.988726597191511e-05,
"loss": 0.2232,
"step": 1287
},
{
"epoch": 8.0,
"grad_norm": 2.640625,
"learning_rate": 4.988703213897503e-05,
"loss": 0.2823,
"step": 1288
},
{
"epoch": 8.006230529595015,
"grad_norm": 2.375,
"learning_rate": 4.988679806432712e-05,
"loss": 0.2045,
"step": 1289
},
{
"epoch": 8.01246105919003,
"grad_norm": 2.828125,
"learning_rate": 4.988656374797365e-05,
"loss": 0.2906,
"step": 1290
},
{
"epoch": 8.018691588785046,
"grad_norm": 3.09375,
"learning_rate": 4.98863291899169e-05,
"loss": 0.3193,
"step": 1291
},
{
"epoch": 8.024922118380061,
"grad_norm": 3.203125,
"learning_rate": 4.9886094390159144e-05,
"loss": 0.2674,
"step": 1292
},
{
"epoch": 8.031152647975079,
"grad_norm": 3.34375,
"learning_rate": 4.9885859348702665e-05,
"loss": 0.3455,
"step": 1293
},
{
"epoch": 8.037383177570094,
"grad_norm": 1.8125,
"learning_rate": 4.9885624065549747e-05,
"loss": 0.1661,
"step": 1294
},
{
"epoch": 8.04361370716511,
"grad_norm": 3.078125,
"learning_rate": 4.988538854070267e-05,
"loss": 0.3376,
"step": 1295
},
{
"epoch": 8.049844236760125,
"grad_norm": 3.4375,
"learning_rate": 4.988515277416373e-05,
"loss": 0.3291,
"step": 1296
},
{
"epoch": 8.05607476635514,
"grad_norm": 3.296875,
"learning_rate": 4.9884916765935205e-05,
"loss": 0.2704,
"step": 1297
},
{
"epoch": 8.062305295950155,
"grad_norm": 2.03125,
"learning_rate": 4.98846805160194e-05,
"loss": 0.1677,
"step": 1298
},
{
"epoch": 8.06853582554517,
"grad_norm": 1.8203125,
"learning_rate": 4.988444402441861e-05,
"loss": 0.1654,
"step": 1299
},
{
"epoch": 8.074766355140186,
"grad_norm": 2.453125,
"learning_rate": 4.988420729113511e-05,
"loss": 0.2511,
"step": 1300
},
{
"epoch": 8.080996884735203,
"grad_norm": 3.25,
"learning_rate": 4.9883970316171226e-05,
"loss": 0.3582,
"step": 1301
},
{
"epoch": 8.087227414330219,
"grad_norm": 3.015625,
"learning_rate": 4.988373309952924e-05,
"loss": 0.3642,
"step": 1302
},
{
"epoch": 8.093457943925234,
"grad_norm": 3.5625,
"learning_rate": 4.988349564121147e-05,
"loss": 0.2871,
"step": 1303
},
{
"epoch": 8.09968847352025,
"grad_norm": 2.28125,
"learning_rate": 4.988325794122022e-05,
"loss": 0.1829,
"step": 1304
},
{
"epoch": 8.105919003115265,
"grad_norm": 1.6640625,
"learning_rate": 4.9883019999557786e-05,
"loss": 0.1333,
"step": 1305
},
{
"epoch": 8.11214953271028,
"grad_norm": 3.640625,
"learning_rate": 4.9882781816226485e-05,
"loss": 0.3194,
"step": 1306
},
{
"epoch": 8.118380062305295,
"grad_norm": 2.46875,
"learning_rate": 4.988254339122864e-05,
"loss": 0.2146,
"step": 1307
},
{
"epoch": 8.12461059190031,
"grad_norm": 2.90625,
"learning_rate": 4.9882304724566565e-05,
"loss": 0.3194,
"step": 1308
},
{
"epoch": 8.130841121495328,
"grad_norm": 3.578125,
"learning_rate": 4.988206581624256e-05,
"loss": 0.3612,
"step": 1309
},
{
"epoch": 8.137071651090343,
"grad_norm": 3.40625,
"learning_rate": 4.988182666625897e-05,
"loss": 0.2823,
"step": 1310
},
{
"epoch": 8.143302180685358,
"grad_norm": 1.484375,
"learning_rate": 4.988158727461809e-05,
"loss": 0.1518,
"step": 1311
},
{
"epoch": 8.149532710280374,
"grad_norm": 3.03125,
"learning_rate": 4.9881347641322277e-05,
"loss": 0.2846,
"step": 1312
},
{
"epoch": 8.15576323987539,
"grad_norm": 1.859375,
"learning_rate": 4.988110776637383e-05,
"loss": 0.1508,
"step": 1313
},
{
"epoch": 8.161993769470405,
"grad_norm": 1.734375,
"learning_rate": 4.9880867649775096e-05,
"loss": 0.1447,
"step": 1314
},
{
"epoch": 8.16822429906542,
"grad_norm": 2.65625,
"learning_rate": 4.9880627291528403e-05,
"loss": 0.2412,
"step": 1315
},
{
"epoch": 8.174454828660435,
"grad_norm": 1.390625,
"learning_rate": 4.9880386691636086e-05,
"loss": 0.1281,
"step": 1316
},
{
"epoch": 8.180685358255452,
"grad_norm": 2.265625,
"learning_rate": 4.988014585010048e-05,
"loss": 0.1747,
"step": 1317
},
{
"epoch": 8.186915887850468,
"grad_norm": 2.25,
"learning_rate": 4.987990476692392e-05,
"loss": 0.2533,
"step": 1318
},
{
"epoch": 8.193146417445483,
"grad_norm": 3.0625,
"learning_rate": 4.987966344210875e-05,
"loss": 0.4199,
"step": 1319
},
{
"epoch": 8.199376947040498,
"grad_norm": 2.78125,
"learning_rate": 4.9879421875657314e-05,
"loss": 0.3077,
"step": 1320
},
{
"epoch": 8.205607476635514,
"grad_norm": 3.140625,
"learning_rate": 4.9879180067571966e-05,
"loss": 0.2582,
"step": 1321
},
{
"epoch": 8.21183800623053,
"grad_norm": 2.546875,
"learning_rate": 4.9878938017855046e-05,
"loss": 0.2435,
"step": 1322
},
{
"epoch": 8.218068535825545,
"grad_norm": 3.328125,
"learning_rate": 4.987869572650891e-05,
"loss": 0.4615,
"step": 1323
},
{
"epoch": 8.22429906542056,
"grad_norm": 2.515625,
"learning_rate": 4.98784531935359e-05,
"loss": 0.2269,
"step": 1324
},
{
"epoch": 8.230529595015577,
"grad_norm": 3.0625,
"learning_rate": 4.987821041893839e-05,
"loss": 0.3083,
"step": 1325
},
{
"epoch": 8.236760124610592,
"grad_norm": 2.484375,
"learning_rate": 4.987796740271872e-05,
"loss": 0.2325,
"step": 1326
},
{
"epoch": 8.242990654205608,
"grad_norm": 3.296875,
"learning_rate": 4.987772414487926e-05,
"loss": 0.4787,
"step": 1327
},
{
"epoch": 8.249221183800623,
"grad_norm": 2.625,
"learning_rate": 4.9877480645422365e-05,
"loss": 0.2342,
"step": 1328
},
{
"epoch": 8.255451713395638,
"grad_norm": 1.9453125,
"learning_rate": 4.987723690435041e-05,
"loss": 0.163,
"step": 1329
},
{
"epoch": 8.261682242990654,
"grad_norm": 2.625,
"learning_rate": 4.9876992921665754e-05,
"loss": 0.2721,
"step": 1330
},
{
"epoch": 8.26791277258567,
"grad_norm": 2.0625,
"learning_rate": 4.987674869737077e-05,
"loss": 0.2707,
"step": 1331
},
{
"epoch": 8.274143302180685,
"grad_norm": 2.4375,
"learning_rate": 4.987650423146783e-05,
"loss": 0.2195,
"step": 1332
},
{
"epoch": 8.280373831775702,
"grad_norm": 2.140625,
"learning_rate": 4.987625952395931e-05,
"loss": 0.1863,
"step": 1333
},
{
"epoch": 8.286604361370717,
"grad_norm": 2.921875,
"learning_rate": 4.987601457484758e-05,
"loss": 0.383,
"step": 1334
},
{
"epoch": 8.292834890965732,
"grad_norm": 1.9765625,
"learning_rate": 4.987576938413504e-05,
"loss": 0.1629,
"step": 1335
},
{
"epoch": 8.299065420560748,
"grad_norm": 2.859375,
"learning_rate": 4.9875523951824045e-05,
"loss": 0.3721,
"step": 1336
},
{
"epoch": 8.305295950155763,
"grad_norm": 3.15625,
"learning_rate": 4.987527827791698e-05,
"loss": 0.3861,
"step": 1337
},
{
"epoch": 8.311526479750778,
"grad_norm": 2.90625,
"learning_rate": 4.9875032362416254e-05,
"loss": 0.3514,
"step": 1338
},
{
"epoch": 8.317757009345794,
"grad_norm": 2.875,
"learning_rate": 4.987478620532423e-05,
"loss": 0.2971,
"step": 1339
},
{
"epoch": 8.32398753894081,
"grad_norm": 2.34375,
"learning_rate": 4.987453980664332e-05,
"loss": 0.2939,
"step": 1340
},
{
"epoch": 8.330218068535826,
"grad_norm": 2.53125,
"learning_rate": 4.98742931663759e-05,
"loss": 0.3928,
"step": 1341
},
{
"epoch": 8.336448598130842,
"grad_norm": 2.765625,
"learning_rate": 4.987404628452438e-05,
"loss": 0.3009,
"step": 1342
},
{
"epoch": 8.342679127725857,
"grad_norm": 2.96875,
"learning_rate": 4.987379916109114e-05,
"loss": 0.3627,
"step": 1343
},
{
"epoch": 8.348909657320872,
"grad_norm": 3.25,
"learning_rate": 4.987355179607859e-05,
"loss": 0.4399,
"step": 1344
},
{
"epoch": 8.355140186915888,
"grad_norm": 3.1875,
"learning_rate": 4.9873304189489144e-05,
"loss": 0.3667,
"step": 1345
},
{
"epoch": 8.361370716510903,
"grad_norm": 2.515625,
"learning_rate": 4.987305634132519e-05,
"loss": 0.2301,
"step": 1346
},
{
"epoch": 8.367601246105918,
"grad_norm": 2.21875,
"learning_rate": 4.9872808251589144e-05,
"loss": 0.2264,
"step": 1347
},
{
"epoch": 8.373831775700934,
"grad_norm": 2.609375,
"learning_rate": 4.98725599202834e-05,
"loss": 0.2753,
"step": 1348
},
{
"epoch": 8.38006230529595,
"grad_norm": 1.7421875,
"learning_rate": 4.98723113474104e-05,
"loss": 0.1794,
"step": 1349
},
{
"epoch": 8.386292834890966,
"grad_norm": 2.328125,
"learning_rate": 4.987206253297253e-05,
"loss": 0.2642,
"step": 1350
},
{
"epoch": 8.392523364485982,
"grad_norm": 1.4765625,
"learning_rate": 4.987181347697221e-05,
"loss": 0.1595,
"step": 1351
},
{
"epoch": 8.398753894080997,
"grad_norm": 2.8125,
"learning_rate": 4.987156417941188e-05,
"loss": 0.2435,
"step": 1352
},
{
"epoch": 8.404984423676012,
"grad_norm": 1.4921875,
"learning_rate": 4.987131464029394e-05,
"loss": 0.1443,
"step": 1353
},
{
"epoch": 8.411214953271028,
"grad_norm": 2.90625,
"learning_rate": 4.987106485962082e-05,
"loss": 0.3003,
"step": 1354
},
{
"epoch": 8.417445482866043,
"grad_norm": 2.375,
"learning_rate": 4.987081483739495e-05,
"loss": 0.1905,
"step": 1355
},
{
"epoch": 8.423676012461058,
"grad_norm": 2.453125,
"learning_rate": 4.9870564573618747e-05,
"loss": 0.2298,
"step": 1356
},
{
"epoch": 8.429906542056075,
"grad_norm": 3.390625,
"learning_rate": 4.9870314068294655e-05,
"loss": 0.4301,
"step": 1357
},
{
"epoch": 8.43613707165109,
"grad_norm": 2.3125,
"learning_rate": 4.9870063321425105e-05,
"loss": 0.2245,
"step": 1358
},
{
"epoch": 8.442367601246106,
"grad_norm": 1.9609375,
"learning_rate": 4.986981233301252e-05,
"loss": 0.1616,
"step": 1359
},
{
"epoch": 8.448598130841122,
"grad_norm": 2.484375,
"learning_rate": 4.9869561103059354e-05,
"loss": 0.2401,
"step": 1360
},
{
"epoch": 8.454828660436137,
"grad_norm": 3.15625,
"learning_rate": 4.986930963156804e-05,
"loss": 0.2952,
"step": 1361
},
{
"epoch": 8.461059190031152,
"grad_norm": 3.515625,
"learning_rate": 4.986905791854101e-05,
"loss": 0.2627,
"step": 1362
},
{
"epoch": 8.467289719626168,
"grad_norm": 3.265625,
"learning_rate": 4.9868805963980726e-05,
"loss": 0.4909,
"step": 1363
},
{
"epoch": 8.473520249221183,
"grad_norm": 2.984375,
"learning_rate": 4.986855376788963e-05,
"loss": 0.3883,
"step": 1364
},
{
"epoch": 8.4797507788162,
"grad_norm": 2.609375,
"learning_rate": 4.986830133027016e-05,
"loss": 0.2517,
"step": 1365
},
{
"epoch": 8.485981308411215,
"grad_norm": 3.59375,
"learning_rate": 4.986804865112478e-05,
"loss": 0.4105,
"step": 1366
},
{
"epoch": 8.49221183800623,
"grad_norm": 1.6640625,
"learning_rate": 4.986779573045594e-05,
"loss": 0.1677,
"step": 1367
},
{
"epoch": 8.498442367601246,
"grad_norm": 2.78125,
"learning_rate": 4.98675425682661e-05,
"loss": 0.2406,
"step": 1368
},
{
"epoch": 8.504672897196262,
"grad_norm": 2.984375,
"learning_rate": 4.986728916455771e-05,
"loss": 0.3213,
"step": 1369
},
{
"epoch": 8.510903426791277,
"grad_norm": 1.125,
"learning_rate": 4.986703551933324e-05,
"loss": 0.1278,
"step": 1370
},
{
"epoch": 8.517133956386292,
"grad_norm": 1.9140625,
"learning_rate": 4.986678163259515e-05,
"loss": 0.1879,
"step": 1371
},
{
"epoch": 8.523364485981308,
"grad_norm": 3.1875,
"learning_rate": 4.986652750434591e-05,
"loss": 0.3841,
"step": 1372
},
{
"epoch": 8.529595015576325,
"grad_norm": 1.8828125,
"learning_rate": 4.986627313458798e-05,
"loss": 0.1701,
"step": 1373
},
{
"epoch": 8.53582554517134,
"grad_norm": 2.375,
"learning_rate": 4.986601852332383e-05,
"loss": 0.2421,
"step": 1374
},
{
"epoch": 8.542056074766355,
"grad_norm": 1.4609375,
"learning_rate": 4.986576367055594e-05,
"loss": 0.1515,
"step": 1375
},
{
"epoch": 8.54828660436137,
"grad_norm": 2.296875,
"learning_rate": 4.986550857628679e-05,
"loss": 0.2912,
"step": 1376
},
{
"epoch": 8.554517133956386,
"grad_norm": 2.671875,
"learning_rate": 4.986525324051885e-05,
"loss": 0.2301,
"step": 1377
},
{
"epoch": 8.560747663551401,
"grad_norm": 2.5,
"learning_rate": 4.986499766325459e-05,
"loss": 0.268,
"step": 1378
},
{
"epoch": 8.566978193146417,
"grad_norm": 2.84375,
"learning_rate": 4.9864741844496506e-05,
"loss": 0.2818,
"step": 1379
},
{
"epoch": 8.573208722741432,
"grad_norm": 2.859375,
"learning_rate": 4.9864485784247086e-05,
"loss": 0.2566,
"step": 1380
},
{
"epoch": 8.57943925233645,
"grad_norm": 1.8359375,
"learning_rate": 4.9864229482508804e-05,
"loss": 0.1607,
"step": 1381
},
{
"epoch": 8.585669781931465,
"grad_norm": 1.515625,
"learning_rate": 4.986397293928415e-05,
"loss": 0.1462,
"step": 1382
},
{
"epoch": 8.59190031152648,
"grad_norm": 2.4375,
"learning_rate": 4.986371615457564e-05,
"loss": 0.1928,
"step": 1383
},
{
"epoch": 8.598130841121495,
"grad_norm": 4.4375,
"learning_rate": 4.986345912838573e-05,
"loss": 0.4604,
"step": 1384
},
{
"epoch": 8.60436137071651,
"grad_norm": 1.890625,
"learning_rate": 4.986320186071694e-05,
"loss": 0.1611,
"step": 1385
},
{
"epoch": 8.610591900311526,
"grad_norm": 1.671875,
"learning_rate": 4.986294435157176e-05,
"loss": 0.1434,
"step": 1386
},
{
"epoch": 8.616822429906541,
"grad_norm": 1.8984375,
"learning_rate": 4.986268660095269e-05,
"loss": 0.1932,
"step": 1387
},
{
"epoch": 8.623052959501557,
"grad_norm": 2.75,
"learning_rate": 4.9862428608862253e-05,
"loss": 0.3594,
"step": 1388
},
{
"epoch": 8.629283489096574,
"grad_norm": 2.796875,
"learning_rate": 4.986217037530293e-05,
"loss": 0.2659,
"step": 1389
},
{
"epoch": 8.63551401869159,
"grad_norm": 2.75,
"learning_rate": 4.9861911900277236e-05,
"loss": 0.2814,
"step": 1390
},
{
"epoch": 8.641744548286605,
"grad_norm": 1.2578125,
"learning_rate": 4.986165318378769e-05,
"loss": 0.1063,
"step": 1391
},
{
"epoch": 8.64797507788162,
"grad_norm": 3.4375,
"learning_rate": 4.98613942258368e-05,
"loss": 0.4202,
"step": 1392
},
{
"epoch": 8.654205607476635,
"grad_norm": 2.8125,
"learning_rate": 4.9861135026427074e-05,
"loss": 0.3191,
"step": 1393
},
{
"epoch": 8.66043613707165,
"grad_norm": 3.453125,
"learning_rate": 4.986087558556104e-05,
"loss": 0.3234,
"step": 1394
},
{
"epoch": 8.666666666666666,
"grad_norm": 2.375,
"learning_rate": 4.98606159032412e-05,
"loss": 0.2226,
"step": 1395
},
{
"epoch": 8.672897196261681,
"grad_norm": 3.5,
"learning_rate": 4.986035597947011e-05,
"loss": 0.4417,
"step": 1396
},
{
"epoch": 8.679127725856699,
"grad_norm": 2.5625,
"learning_rate": 4.986009581425026e-05,
"loss": 0.2359,
"step": 1397
},
{
"epoch": 8.685358255451714,
"grad_norm": 2.484375,
"learning_rate": 4.9859835407584186e-05,
"loss": 0.2071,
"step": 1398
},
{
"epoch": 8.69158878504673,
"grad_norm": 2.953125,
"learning_rate": 4.985957475947443e-05,
"loss": 0.3075,
"step": 1399
},
{
"epoch": 8.697819314641745,
"grad_norm": 2.546875,
"learning_rate": 4.985931386992351e-05,
"loss": 0.2536,
"step": 1400
},
{
"epoch": 8.70404984423676,
"grad_norm": 1.9453125,
"learning_rate": 4.9859052738933966e-05,
"loss": 0.1969,
"step": 1401
},
{
"epoch": 8.710280373831775,
"grad_norm": 2.53125,
"learning_rate": 4.985879136650834e-05,
"loss": 0.2291,
"step": 1402
},
{
"epoch": 8.71651090342679,
"grad_norm": 2.4375,
"learning_rate": 4.985852975264915e-05,
"loss": 0.2319,
"step": 1403
},
{
"epoch": 8.722741433021806,
"grad_norm": 3.015625,
"learning_rate": 4.9858267897358956e-05,
"loss": 0.2537,
"step": 1404
},
{
"epoch": 8.728971962616823,
"grad_norm": 2.296875,
"learning_rate": 4.985800580064029e-05,
"loss": 0.1959,
"step": 1405
},
{
"epoch": 8.735202492211839,
"grad_norm": 1.9765625,
"learning_rate": 4.985774346249571e-05,
"loss": 0.2649,
"step": 1406
},
{
"epoch": 8.741433021806854,
"grad_norm": 3.671875,
"learning_rate": 4.985748088292775e-05,
"loss": 0.4753,
"step": 1407
},
{
"epoch": 8.74766355140187,
"grad_norm": 3.46875,
"learning_rate": 4.9857218061938976e-05,
"loss": 0.3041,
"step": 1408
},
{
"epoch": 8.753894080996885,
"grad_norm": 3.109375,
"learning_rate": 4.985695499953193e-05,
"loss": 0.3963,
"step": 1409
},
{
"epoch": 8.7601246105919,
"grad_norm": 2.40625,
"learning_rate": 4.985669169570916e-05,
"loss": 0.2326,
"step": 1410
},
{
"epoch": 8.766355140186915,
"grad_norm": 2.359375,
"learning_rate": 4.985642815047323e-05,
"loss": 0.3253,
"step": 1411
},
{
"epoch": 8.77258566978193,
"grad_norm": 2.6875,
"learning_rate": 4.985616436382671e-05,
"loss": 0.3197,
"step": 1412
},
{
"epoch": 8.778816199376948,
"grad_norm": 2.140625,
"learning_rate": 4.985590033577215e-05,
"loss": 0.1772,
"step": 1413
},
{
"epoch": 8.785046728971963,
"grad_norm": 2.71875,
"learning_rate": 4.985563606631212e-05,
"loss": 0.3099,
"step": 1414
},
{
"epoch": 8.791277258566979,
"grad_norm": 2.59375,
"learning_rate": 4.9855371555449185e-05,
"loss": 0.2429,
"step": 1415
},
{
"epoch": 8.797507788161994,
"grad_norm": 1.7734375,
"learning_rate": 4.98551068031859e-05,
"loss": 0.1783,
"step": 1416
},
{
"epoch": 8.80373831775701,
"grad_norm": 3.53125,
"learning_rate": 4.9854841809524866e-05,
"loss": 0.3142,
"step": 1417
},
{
"epoch": 8.809968847352025,
"grad_norm": 1.921875,
"learning_rate": 4.985457657446863e-05,
"loss": 0.1786,
"step": 1418
},
{
"epoch": 8.81619937694704,
"grad_norm": 2.3125,
"learning_rate": 4.985431109801978e-05,
"loss": 0.2253,
"step": 1419
},
{
"epoch": 8.822429906542055,
"grad_norm": 2.03125,
"learning_rate": 4.985404538018089e-05,
"loss": 0.2112,
"step": 1420
},
{
"epoch": 8.828660436137072,
"grad_norm": 2.375,
"learning_rate": 4.985377942095455e-05,
"loss": 0.273,
"step": 1421
},
{
"epoch": 8.834890965732088,
"grad_norm": 2.4375,
"learning_rate": 4.985351322034333e-05,
"loss": 0.2831,
"step": 1422
},
{
"epoch": 8.841121495327103,
"grad_norm": 2.703125,
"learning_rate": 4.985324677834983e-05,
"loss": 0.3214,
"step": 1423
},
{
"epoch": 8.847352024922118,
"grad_norm": 2.546875,
"learning_rate": 4.985298009497662e-05,
"loss": 0.3224,
"step": 1424
},
{
"epoch": 8.853582554517134,
"grad_norm": 4.125,
"learning_rate": 4.9852713170226294e-05,
"loss": 0.5239,
"step": 1425
},
{
"epoch": 8.85981308411215,
"grad_norm": 1.7578125,
"learning_rate": 4.985244600410146e-05,
"loss": 0.1608,
"step": 1426
},
{
"epoch": 8.866043613707165,
"grad_norm": 2.359375,
"learning_rate": 4.9852178596604705e-05,
"loss": 0.2016,
"step": 1427
},
{
"epoch": 8.87227414330218,
"grad_norm": 1.8515625,
"learning_rate": 4.985191094773861e-05,
"loss": 0.2506,
"step": 1428
},
{
"epoch": 8.878504672897197,
"grad_norm": 2.171875,
"learning_rate": 4.9851643057505805e-05,
"loss": 0.1966,
"step": 1429
},
{
"epoch": 8.884735202492212,
"grad_norm": 1.234375,
"learning_rate": 4.985137492590887e-05,
"loss": 0.1261,
"step": 1430
},
{
"epoch": 8.890965732087228,
"grad_norm": 2.953125,
"learning_rate": 4.9851106552950404e-05,
"loss": 0.2757,
"step": 1431
},
{
"epoch": 8.897196261682243,
"grad_norm": 2.90625,
"learning_rate": 4.985083793863303e-05,
"loss": 0.3261,
"step": 1432
},
{
"epoch": 8.903426791277258,
"grad_norm": 3.046875,
"learning_rate": 4.985056908295935e-05,
"loss": 0.3119,
"step": 1433
},
{
"epoch": 8.909657320872274,
"grad_norm": 2.734375,
"learning_rate": 4.985029998593198e-05,
"loss": 0.2861,
"step": 1434
},
{
"epoch": 8.91588785046729,
"grad_norm": 3.015625,
"learning_rate": 4.9850030647553525e-05,
"loss": 0.2788,
"step": 1435
},
{
"epoch": 8.922118380062305,
"grad_norm": 3.4375,
"learning_rate": 4.984976106782661e-05,
"loss": 0.3219,
"step": 1436
},
{
"epoch": 8.928348909657322,
"grad_norm": 2.046875,
"learning_rate": 4.984949124675383e-05,
"loss": 0.1623,
"step": 1437
},
{
"epoch": 8.934579439252337,
"grad_norm": 1.8359375,
"learning_rate": 4.9849221184337843e-05,
"loss": 0.1646,
"step": 1438
},
{
"epoch": 8.940809968847352,
"grad_norm": 1.8984375,
"learning_rate": 4.9848950880581246e-05,
"loss": 0.18,
"step": 1439
},
{
"epoch": 8.947040498442368,
"grad_norm": 1.703125,
"learning_rate": 4.984868033548667e-05,
"loss": 0.1656,
"step": 1440
},
{
"epoch": 8.953271028037383,
"grad_norm": 3.125,
"learning_rate": 4.9848409549056745e-05,
"loss": 0.2464,
"step": 1441
},
{
"epoch": 8.959501557632398,
"grad_norm": 3.21875,
"learning_rate": 4.984813852129409e-05,
"loss": 0.2689,
"step": 1442
},
{
"epoch": 8.965732087227414,
"grad_norm": 1.2734375,
"learning_rate": 4.984786725220136e-05,
"loss": 0.1151,
"step": 1443
},
{
"epoch": 8.97196261682243,
"grad_norm": 2.515625,
"learning_rate": 4.9847595741781175e-05,
"loss": 0.2327,
"step": 1444
},
{
"epoch": 8.978193146417446,
"grad_norm": 2.96875,
"learning_rate": 4.9847323990036174e-05,
"loss": 0.3466,
"step": 1445
},
{
"epoch": 8.984423676012462,
"grad_norm": 2.640625,
"learning_rate": 4.984705199696899e-05,
"loss": 0.2398,
"step": 1446
},
{
"epoch": 8.990654205607477,
"grad_norm": 2.140625,
"learning_rate": 4.984677976258227e-05,
"loss": 0.1734,
"step": 1447
},
{
"epoch": 8.996884735202492,
"grad_norm": 3.265625,
"learning_rate": 4.9846507286878665e-05,
"loss": 0.3352,
"step": 1448
},
{
"epoch": 9.0,
"grad_norm": 2.140625,
"learning_rate": 4.9846234569860806e-05,
"loss": 0.1757,
"step": 1449
},
{
"epoch": 9.006230529595015,
"grad_norm": 2.0625,
"learning_rate": 4.984596161153136e-05,
"loss": 0.1674,
"step": 1450
},
{
"epoch": 9.01246105919003,
"grad_norm": 2.890625,
"learning_rate": 4.9845688411892956e-05,
"loss": 0.2776,
"step": 1451
},
{
"epoch": 9.018691588785046,
"grad_norm": 1.9140625,
"learning_rate": 4.984541497094827e-05,
"loss": 0.2397,
"step": 1452
},
{
"epoch": 9.024922118380061,
"grad_norm": 2.578125,
"learning_rate": 4.984514128869994e-05,
"loss": 0.3013,
"step": 1453
},
{
"epoch": 9.031152647975079,
"grad_norm": 3.3125,
"learning_rate": 4.9844867365150636e-05,
"loss": 0.2638,
"step": 1454
},
{
"epoch": 9.037383177570094,
"grad_norm": 2.203125,
"learning_rate": 4.984459320030301e-05,
"loss": 0.2161,
"step": 1455
},
{
"epoch": 9.04361370716511,
"grad_norm": 1.6953125,
"learning_rate": 4.984431879415973e-05,
"loss": 0.1441,
"step": 1456
},
{
"epoch": 9.049844236760125,
"grad_norm": 2.75,
"learning_rate": 4.984404414672346e-05,
"loss": 0.2828,
"step": 1457
},
{
"epoch": 9.05607476635514,
"grad_norm": 2.421875,
"learning_rate": 4.984376925799687e-05,
"loss": 0.202,
"step": 1458
},
{
"epoch": 9.062305295950155,
"grad_norm": 1.859375,
"learning_rate": 4.984349412798263e-05,
"loss": 0.1418,
"step": 1459
},
{
"epoch": 9.06853582554517,
"grad_norm": 2.8125,
"learning_rate": 4.98432187566834e-05,
"loss": 0.3252,
"step": 1460
},
{
"epoch": 9.074766355140186,
"grad_norm": 3.296875,
"learning_rate": 4.984294314410186e-05,
"loss": 0.3058,
"step": 1461
},
{
"epoch": 9.080996884735203,
"grad_norm": 2.96875,
"learning_rate": 4.98426672902407e-05,
"loss": 0.2499,
"step": 1462
},
{
"epoch": 9.087227414330219,
"grad_norm": 2.75,
"learning_rate": 4.9842391195102585e-05,
"loss": 0.324,
"step": 1463
},
{
"epoch": 9.093457943925234,
"grad_norm": 3.140625,
"learning_rate": 4.98421148586902e-05,
"loss": 0.2756,
"step": 1464
},
{
"epoch": 9.09968847352025,
"grad_norm": 1.515625,
"learning_rate": 4.9841838281006235e-05,
"loss": 0.1286,
"step": 1465
},
{
"epoch": 9.105919003115265,
"grad_norm": 1.546875,
"learning_rate": 4.984156146205336e-05,
"loss": 0.136,
"step": 1466
},
{
"epoch": 9.11214953271028,
"grad_norm": 1.859375,
"learning_rate": 4.984128440183429e-05,
"loss": 0.1575,
"step": 1467
},
{
"epoch": 9.118380062305295,
"grad_norm": 1.515625,
"learning_rate": 4.984100710035169e-05,
"loss": 0.1421,
"step": 1468
},
{
"epoch": 9.12461059190031,
"grad_norm": 1.90625,
"learning_rate": 4.9840729557608265e-05,
"loss": 0.1751,
"step": 1469
},
{
"epoch": 9.130841121495328,
"grad_norm": 3.171875,
"learning_rate": 4.984045177360671e-05,
"loss": 0.2774,
"step": 1470
},
{
"epoch": 9.137071651090343,
"grad_norm": 2.609375,
"learning_rate": 4.9840173748349716e-05,
"loss": 0.2859,
"step": 1471
},
{
"epoch": 9.143302180685358,
"grad_norm": 1.8828125,
"learning_rate": 4.9839895481839994e-05,
"loss": 0.1415,
"step": 1472
},
{
"epoch": 9.149532710280374,
"grad_norm": 1.1640625,
"learning_rate": 4.9839616974080246e-05,
"loss": 0.1095,
"step": 1473
},
{
"epoch": 9.15576323987539,
"grad_norm": 2.265625,
"learning_rate": 4.983933822507316e-05,
"loss": 0.2008,
"step": 1474
},
{
"epoch": 9.161993769470405,
"grad_norm": 2.296875,
"learning_rate": 4.983905923482147e-05,
"loss": 0.2,
"step": 1475
},
{
"epoch": 9.16822429906542,
"grad_norm": 1.8671875,
"learning_rate": 4.983878000332787e-05,
"loss": 0.1659,
"step": 1476
},
{
"epoch": 9.174454828660435,
"grad_norm": 2.453125,
"learning_rate": 4.9838500530595066e-05,
"loss": 0.2421,
"step": 1477
},
{
"epoch": 9.180685358255452,
"grad_norm": 1.171875,
"learning_rate": 4.983822081662578e-05,
"loss": 0.1029,
"step": 1478
},
{
"epoch": 9.186915887850468,
"grad_norm": 2.59375,
"learning_rate": 4.983794086142273e-05,
"loss": 0.2531,
"step": 1479
},
{
"epoch": 9.193146417445483,
"grad_norm": 2.015625,
"learning_rate": 4.9837660664988636e-05,
"loss": 0.1444,
"step": 1480
},
{
"epoch": 9.199376947040498,
"grad_norm": 2.296875,
"learning_rate": 4.983738022732621e-05,
"loss": 0.3452,
"step": 1481
},
{
"epoch": 9.205607476635514,
"grad_norm": 1.9140625,
"learning_rate": 4.9837099548438195e-05,
"loss": 0.1781,
"step": 1482
},
{
"epoch": 9.21183800623053,
"grad_norm": 2.28125,
"learning_rate": 4.98368186283273e-05,
"loss": 0.2138,
"step": 1483
},
{
"epoch": 9.218068535825545,
"grad_norm": 1.4765625,
"learning_rate": 4.9836537466996244e-05,
"loss": 0.1293,
"step": 1484
},
{
"epoch": 9.22429906542056,
"grad_norm": 1.40625,
"learning_rate": 4.983625606444779e-05,
"loss": 0.1329,
"step": 1485
},
{
"epoch": 9.230529595015577,
"grad_norm": 2.046875,
"learning_rate": 4.983597442068464e-05,
"loss": 0.1709,
"step": 1486
},
{
"epoch": 9.236760124610592,
"grad_norm": 1.9921875,
"learning_rate": 4.983569253570955e-05,
"loss": 0.1538,
"step": 1487
},
{
"epoch": 9.242990654205608,
"grad_norm": 2.390625,
"learning_rate": 4.9835410409525246e-05,
"loss": 0.2351,
"step": 1488
},
{
"epoch": 9.249221183800623,
"grad_norm": 2.984375,
"learning_rate": 4.983512804213447e-05,
"loss": 0.3049,
"step": 1489
},
{
"epoch": 9.255451713395638,
"grad_norm": 2.640625,
"learning_rate": 4.9834845433539966e-05,
"loss": 0.2533,
"step": 1490
},
{
"epoch": 9.261682242990654,
"grad_norm": 1.8203125,
"learning_rate": 4.983456258374448e-05,
"loss": 0.1464,
"step": 1491
},
{
"epoch": 9.26791277258567,
"grad_norm": 1.59375,
"learning_rate": 4.983427949275076e-05,
"loss": 0.2184,
"step": 1492
},
{
"epoch": 9.274143302180685,
"grad_norm": 3.015625,
"learning_rate": 4.983399616056155e-05,
"loss": 0.2733,
"step": 1493
},
{
"epoch": 9.280373831775702,
"grad_norm": 2.828125,
"learning_rate": 4.983371258717961e-05,
"loss": 0.2039,
"step": 1494
},
{
"epoch": 9.286604361370717,
"grad_norm": 3.75,
"learning_rate": 4.9833428772607685e-05,
"loss": 0.3428,
"step": 1495
},
{
"epoch": 9.292834890965732,
"grad_norm": 1.40625,
"learning_rate": 4.983314471684853e-05,
"loss": 0.1198,
"step": 1496
},
{
"epoch": 9.299065420560748,
"grad_norm": 1.953125,
"learning_rate": 4.983286041990492e-05,
"loss": 0.2077,
"step": 1497
},
{
"epoch": 9.305295950155763,
"grad_norm": 2.625,
"learning_rate": 4.98325758817796e-05,
"loss": 0.2469,
"step": 1498
},
{
"epoch": 9.311526479750778,
"grad_norm": 3.390625,
"learning_rate": 4.983229110247534e-05,
"loss": 0.4669,
"step": 1499
},
{
"epoch": 9.317757009345794,
"grad_norm": 2.6875,
"learning_rate": 4.9832006081994907e-05,
"loss": 0.2665,
"step": 1500
},
{
"epoch": 9.32398753894081,
"grad_norm": 3.046875,
"learning_rate": 4.983172082034106e-05,
"loss": 0.3867,
"step": 1501
},
{
"epoch": 9.330218068535826,
"grad_norm": 2.609375,
"learning_rate": 4.983143531751659e-05,
"loss": 0.2946,
"step": 1502
},
{
"epoch": 9.336448598130842,
"grad_norm": 3.046875,
"learning_rate": 4.983114957352425e-05,
"loss": 0.2423,
"step": 1503
},
{
"epoch": 9.342679127725857,
"grad_norm": 1.8984375,
"learning_rate": 4.9830863588366815e-05,
"loss": 0.1348,
"step": 1504
},
{
"epoch": 9.348909657320872,
"grad_norm": 1.5625,
"learning_rate": 4.9830577362047076e-05,
"loss": 0.1351,
"step": 1505
},
{
"epoch": 9.355140186915888,
"grad_norm": 2.984375,
"learning_rate": 4.98302908945678e-05,
"loss": 0.2836,
"step": 1506
},
{
"epoch": 9.361370716510903,
"grad_norm": 1.578125,
"learning_rate": 4.983000418593179e-05,
"loss": 0.1351,
"step": 1507
},
{
"epoch": 9.367601246105918,
"grad_norm": 1.2578125,
"learning_rate": 4.98297172361418e-05,
"loss": 0.1174,
"step": 1508
},
{
"epoch": 9.373831775700934,
"grad_norm": 1.5625,
"learning_rate": 4.982943004520064e-05,
"loss": 0.1565,
"step": 1509
},
{
"epoch": 9.38006230529595,
"grad_norm": 2.46875,
"learning_rate": 4.982914261311109e-05,
"loss": 0.2583,
"step": 1510
},
{
"epoch": 9.386292834890966,
"grad_norm": 1.90625,
"learning_rate": 4.9828854939875945e-05,
"loss": 0.1722,
"step": 1511
},
{
"epoch": 9.392523364485982,
"grad_norm": 2.140625,
"learning_rate": 4.9828567025498e-05,
"loss": 0.2332,
"step": 1512
},
{
"epoch": 9.398753894080997,
"grad_norm": 2.1875,
"learning_rate": 4.9828278869980044e-05,
"loss": 0.1997,
"step": 1513
},
{
"epoch": 9.404984423676012,
"grad_norm": 1.78125,
"learning_rate": 4.982799047332488e-05,
"loss": 0.1589,
"step": 1514
},
{
"epoch": 9.411214953271028,
"grad_norm": 2.25,
"learning_rate": 4.9827701835535315e-05,
"loss": 0.1945,
"step": 1515
},
{
"epoch": 9.417445482866043,
"grad_norm": 2.828125,
"learning_rate": 4.9827412956614145e-05,
"loss": 0.3032,
"step": 1516
},
{
"epoch": 9.423676012461058,
"grad_norm": 1.796875,
"learning_rate": 4.9827123836564175e-05,
"loss": 0.1585,
"step": 1517
},
{
"epoch": 9.429906542056075,
"grad_norm": 1.640625,
"learning_rate": 4.982683447538822e-05,
"loss": 0.1467,
"step": 1518
},
{
"epoch": 9.43613707165109,
"grad_norm": 2.765625,
"learning_rate": 4.982654487308908e-05,
"loss": 0.26,
"step": 1519
},
{
"epoch": 9.442367601246106,
"grad_norm": 1.7890625,
"learning_rate": 4.9826255029669577e-05,
"loss": 0.1785,
"step": 1520
},
{
"epoch": 9.448598130841122,
"grad_norm": 2.703125,
"learning_rate": 4.982596494513252e-05,
"loss": 0.2496,
"step": 1521
},
{
"epoch": 9.454828660436137,
"grad_norm": 1.640625,
"learning_rate": 4.982567461948073e-05,
"loss": 0.1353,
"step": 1522
},
{
"epoch": 9.461059190031152,
"grad_norm": 2.703125,
"learning_rate": 4.982538405271703e-05,
"loss": 0.3908,
"step": 1523
},
{
"epoch": 9.467289719626168,
"grad_norm": 1.5078125,
"learning_rate": 4.982509324484424e-05,
"loss": 0.1318,
"step": 1524
},
{
"epoch": 9.473520249221183,
"grad_norm": 2.359375,
"learning_rate": 4.9824802195865164e-05,
"loss": 0.2077,
"step": 1525
},
{
"epoch": 9.4797507788162,
"grad_norm": 2.140625,
"learning_rate": 4.982451090578266e-05,
"loss": 0.2335,
"step": 1526
},
{
"epoch": 9.485981308411215,
"grad_norm": 2.09375,
"learning_rate": 4.9824219374599543e-05,
"loss": 0.1829,
"step": 1527
},
{
"epoch": 9.49221183800623,
"grad_norm": 3.21875,
"learning_rate": 4.982392760231864e-05,
"loss": 0.334,
"step": 1528
},
{
"epoch": 9.498442367601246,
"grad_norm": 2.265625,
"learning_rate": 4.98236355889428e-05,
"loss": 0.2206,
"step": 1529
},
{
"epoch": 9.504672897196262,
"grad_norm": 2.515625,
"learning_rate": 4.9823343334474844e-05,
"loss": 0.2845,
"step": 1530
},
{
"epoch": 9.510903426791277,
"grad_norm": 3.03125,
"learning_rate": 4.982305083891761e-05,
"loss": 0.3204,
"step": 1531
},
{
"epoch": 9.517133956386292,
"grad_norm": 1.7421875,
"learning_rate": 4.982275810227395e-05,
"loss": 0.1653,
"step": 1532
},
{
"epoch": 9.523364485981308,
"grad_norm": 1.59375,
"learning_rate": 4.9822465124546705e-05,
"loss": 0.1311,
"step": 1533
},
{
"epoch": 9.529595015576325,
"grad_norm": 3.046875,
"learning_rate": 4.982217190573871e-05,
"loss": 0.2715,
"step": 1534
},
{
"epoch": 9.53582554517134,
"grad_norm": 1.984375,
"learning_rate": 4.982187844585282e-05,
"loss": 0.1534,
"step": 1535
},
{
"epoch": 9.542056074766355,
"grad_norm": 1.5859375,
"learning_rate": 4.982158474489189e-05,
"loss": 0.143,
"step": 1536
},
{
"epoch": 9.54828660436137,
"grad_norm": 1.3125,
"learning_rate": 4.982129080285876e-05,
"loss": 0.1215,
"step": 1537
},
{
"epoch": 9.554517133956386,
"grad_norm": 2.40625,
"learning_rate": 4.98209966197563e-05,
"loss": 0.2063,
"step": 1538
},
{
"epoch": 9.560747663551401,
"grad_norm": 1.296875,
"learning_rate": 4.9820702195587356e-05,
"loss": 0.1286,
"step": 1539
},
{
"epoch": 9.566978193146417,
"grad_norm": 2.203125,
"learning_rate": 4.9820407530354786e-05,
"loss": 0.2058,
"step": 1540
},
{
"epoch": 9.573208722741432,
"grad_norm": 2.296875,
"learning_rate": 4.982011262406145e-05,
"loss": 0.1924,
"step": 1541
},
{
"epoch": 9.57943925233645,
"grad_norm": 2.53125,
"learning_rate": 4.981981747671024e-05,
"loss": 0.2313,
"step": 1542
},
{
"epoch": 9.585669781931465,
"grad_norm": 3.0625,
"learning_rate": 4.981952208830399e-05,
"loss": 0.3645,
"step": 1543
},
{
"epoch": 9.59190031152648,
"grad_norm": 2.078125,
"learning_rate": 4.981922645884558e-05,
"loss": 0.2067,
"step": 1544
},
{
"epoch": 9.598130841121495,
"grad_norm": 1.84375,
"learning_rate": 4.981893058833788e-05,
"loss": 0.1623,
"step": 1545
},
{
"epoch": 9.60436137071651,
"grad_norm": 2.109375,
"learning_rate": 4.9818634476783764e-05,
"loss": 0.2569,
"step": 1546
},
{
"epoch": 9.610591900311526,
"grad_norm": 2.9375,
"learning_rate": 4.981833812418612e-05,
"loss": 0.3088,
"step": 1547
},
{
"epoch": 9.616822429906541,
"grad_norm": 2.203125,
"learning_rate": 4.9818041530547796e-05,
"loss": 0.1853,
"step": 1548
},
{
"epoch": 9.623052959501557,
"grad_norm": 2.90625,
"learning_rate": 4.98177446958717e-05,
"loss": 0.3672,
"step": 1549
},
{
"epoch": 9.629283489096574,
"grad_norm": 2.875,
"learning_rate": 4.981744762016072e-05,
"loss": 0.3835,
"step": 1550
},
{
"epoch": 9.63551401869159,
"grad_norm": 2.21875,
"learning_rate": 4.9817150303417716e-05,
"loss": 0.1835,
"step": 1551
},
{
"epoch": 9.641744548286605,
"grad_norm": 3.59375,
"learning_rate": 4.981685274564558e-05,
"loss": 0.3706,
"step": 1552
},
{
"epoch": 9.64797507788162,
"grad_norm": 2.375,
"learning_rate": 4.9816554946847216e-05,
"loss": 0.2154,
"step": 1553
},
{
"epoch": 9.654205607476635,
"grad_norm": 1.6640625,
"learning_rate": 4.981625690702552e-05,
"loss": 0.1465,
"step": 1554
},
{
"epoch": 9.66043613707165,
"grad_norm": 3.0,
"learning_rate": 4.981595862618336e-05,
"loss": 0.297,
"step": 1555
},
{
"epoch": 9.666666666666666,
"grad_norm": 2.375,
"learning_rate": 4.9815660104323656e-05,
"loss": 0.1785,
"step": 1556
},
{
"epoch": 9.672897196261681,
"grad_norm": 2.109375,
"learning_rate": 4.98153613414493e-05,
"loss": 0.1738,
"step": 1557
},
{
"epoch": 9.679127725856699,
"grad_norm": 2.296875,
"learning_rate": 4.98150623375632e-05,
"loss": 0.2032,
"step": 1558
},
{
"epoch": 9.685358255451714,
"grad_norm": 1.796875,
"learning_rate": 4.9814763092668245e-05,
"loss": 0.1519,
"step": 1559
},
{
"epoch": 9.69158878504673,
"grad_norm": 1.921875,
"learning_rate": 4.981446360676736e-05,
"loss": 0.1471,
"step": 1560
},
{
"epoch": 9.697819314641745,
"grad_norm": 1.9453125,
"learning_rate": 4.9814163879863436e-05,
"loss": 0.1645,
"step": 1561
},
{
"epoch": 9.70404984423676,
"grad_norm": 2.5,
"learning_rate": 4.981386391195939e-05,
"loss": 0.2428,
"step": 1562
},
{
"epoch": 9.710280373831775,
"grad_norm": 1.2578125,
"learning_rate": 4.981356370305815e-05,
"loss": 0.1163,
"step": 1563
},
{
"epoch": 9.71651090342679,
"grad_norm": 2.1875,
"learning_rate": 4.981326325316261e-05,
"loss": 0.1772,
"step": 1564
},
{
"epoch": 9.722741433021806,
"grad_norm": 1.40625,
"learning_rate": 4.981296256227569e-05,
"loss": 0.1311,
"step": 1565
},
{
"epoch": 9.728971962616823,
"grad_norm": 1.671875,
"learning_rate": 4.981266163040033e-05,
"loss": 0.1337,
"step": 1566
},
{
"epoch": 9.735202492211839,
"grad_norm": 1.6171875,
"learning_rate": 4.9812360457539444e-05,
"loss": 0.152,
"step": 1567
},
{
"epoch": 9.741433021806854,
"grad_norm": 2.609375,
"learning_rate": 4.981205904369594e-05,
"loss": 0.2644,
"step": 1568
},
{
"epoch": 9.74766355140187,
"grad_norm": 2.375,
"learning_rate": 4.9811757388872764e-05,
"loss": 0.2263,
"step": 1569
},
{
"epoch": 9.753894080996885,
"grad_norm": 1.8359375,
"learning_rate": 4.981145549307285e-05,
"loss": 0.1543,
"step": 1570
},
{
"epoch": 9.7601246105919,
"grad_norm": 2.734375,
"learning_rate": 4.98111533562991e-05,
"loss": 0.2581,
"step": 1571
},
{
"epoch": 9.766355140186915,
"grad_norm": 1.453125,
"learning_rate": 4.981085097855449e-05,
"loss": 0.1439,
"step": 1572
},
{
"epoch": 9.77258566978193,
"grad_norm": 2.40625,
"learning_rate": 4.981054835984192e-05,
"loss": 0.2774,
"step": 1573
},
{
"epoch": 9.778816199376948,
"grad_norm": 2.109375,
"learning_rate": 4.981024550016435e-05,
"loss": 0.1812,
"step": 1574
},
{
"epoch": 9.785046728971963,
"grad_norm": 2.0625,
"learning_rate": 4.980994239952472e-05,
"loss": 0.1607,
"step": 1575
},
{
"epoch": 9.791277258566979,
"grad_norm": 1.890625,
"learning_rate": 4.9809639057925966e-05,
"loss": 0.1684,
"step": 1576
},
{
"epoch": 9.797507788161994,
"grad_norm": 2.1875,
"learning_rate": 4.9809335475371045e-05,
"loss": 0.2217,
"step": 1577
},
{
"epoch": 9.80373831775701,
"grad_norm": 2.3125,
"learning_rate": 4.980903165186289e-05,
"loss": 0.2152,
"step": 1578
},
{
"epoch": 9.809968847352025,
"grad_norm": 2.140625,
"learning_rate": 4.9808727587404464e-05,
"loss": 0.1967,
"step": 1579
},
{
"epoch": 9.81619937694704,
"grad_norm": 2.015625,
"learning_rate": 4.980842328199872e-05,
"loss": 0.2031,
"step": 1580
},
{
"epoch": 9.822429906542055,
"grad_norm": 1.9375,
"learning_rate": 4.98081187356486e-05,
"loss": 0.1544,
"step": 1581
},
{
"epoch": 9.828660436137072,
"grad_norm": 1.9765625,
"learning_rate": 4.980781394835707e-05,
"loss": 0.1629,
"step": 1582
},
{
"epoch": 9.834890965732088,
"grad_norm": 2.609375,
"learning_rate": 4.980750892012711e-05,
"loss": 0.3513,
"step": 1583
},
{
"epoch": 9.841121495327103,
"grad_norm": 2.328125,
"learning_rate": 4.9807203650961654e-05,
"loss": 0.2037,
"step": 1584
},
{
"epoch": 9.847352024922118,
"grad_norm": 2.6875,
"learning_rate": 4.980689814086368e-05,
"loss": 0.1899,
"step": 1585
},
{
"epoch": 9.853582554517134,
"grad_norm": 2.46875,
"learning_rate": 4.980659238983615e-05,
"loss": 0.223,
"step": 1586
},
{
"epoch": 9.85981308411215,
"grad_norm": 1.9765625,
"learning_rate": 4.980628639788203e-05,
"loss": 0.1712,
"step": 1587
},
{
"epoch": 9.866043613707165,
"grad_norm": 1.8203125,
"learning_rate": 4.9805980165004304e-05,
"loss": 0.1325,
"step": 1588
},
{
"epoch": 9.87227414330218,
"grad_norm": 3.03125,
"learning_rate": 4.980567369120594e-05,
"loss": 0.3035,
"step": 1589
},
{
"epoch": 9.878504672897197,
"grad_norm": 1.6015625,
"learning_rate": 4.9805366976489915e-05,
"loss": 0.1334,
"step": 1590
},
{
"epoch": 9.884735202492212,
"grad_norm": 2.609375,
"learning_rate": 4.980506002085921e-05,
"loss": 0.2982,
"step": 1591
},
{
"epoch": 9.890965732087228,
"grad_norm": 2.71875,
"learning_rate": 4.9804752824316794e-05,
"loss": 0.2385,
"step": 1592
},
{
"epoch": 9.897196261682243,
"grad_norm": 2.9375,
"learning_rate": 4.980444538686567e-05,
"loss": 0.3384,
"step": 1593
},
{
"epoch": 9.903426791277258,
"grad_norm": 2.921875,
"learning_rate": 4.9804137708508814e-05,
"loss": 0.2634,
"step": 1594
},
{
"epoch": 9.909657320872274,
"grad_norm": 2.78125,
"learning_rate": 4.9803829789249205e-05,
"loss": 0.2041,
"step": 1595
},
{
"epoch": 9.91588785046729,
"grad_norm": 2.25,
"learning_rate": 4.980352162908985e-05,
"loss": 0.2806,
"step": 1596
},
{
"epoch": 9.922118380062305,
"grad_norm": 1.75,
"learning_rate": 4.980321322803374e-05,
"loss": 0.1654,
"step": 1597
},
{
"epoch": 9.928348909657322,
"grad_norm": 2.125,
"learning_rate": 4.9802904586083856e-05,
"loss": 0.2492,
"step": 1598
},
{
"epoch": 9.934579439252337,
"grad_norm": 1.734375,
"learning_rate": 4.9802595703243205e-05,
"loss": 0.1504,
"step": 1599
},
{
"epoch": 9.940809968847352,
"grad_norm": 2.28125,
"learning_rate": 4.9802286579514787e-05,
"loss": 0.2094,
"step": 1600
},
{
"epoch": 9.947040498442368,
"grad_norm": 1.5546875,
"learning_rate": 4.980197721490161e-05,
"loss": 0.1167,
"step": 1601
},
{
"epoch": 9.953271028037383,
"grad_norm": 2.484375,
"learning_rate": 4.980166760940667e-05,
"loss": 0.1977,
"step": 1602
},
{
"epoch": 9.959501557632398,
"grad_norm": 2.078125,
"learning_rate": 4.980135776303297e-05,
"loss": 0.2364,
"step": 1603
},
{
"epoch": 9.965732087227414,
"grad_norm": 3.3125,
"learning_rate": 4.980104767578353e-05,
"loss": 0.334,
"step": 1604
},
{
"epoch": 9.97196261682243,
"grad_norm": 3.0,
"learning_rate": 4.980073734766136e-05,
"loss": 0.231,
"step": 1605
},
{
"epoch": 9.978193146417446,
"grad_norm": 2.09375,
"learning_rate": 4.980042677866947e-05,
"loss": 0.163,
"step": 1606
},
{
"epoch": 9.984423676012462,
"grad_norm": 2.65625,
"learning_rate": 4.980011596881087e-05,
"loss": 0.2123,
"step": 1607
},
{
"epoch": 9.990654205607477,
"grad_norm": 2.84375,
"learning_rate": 4.97998049180886e-05,
"loss": 0.2517,
"step": 1608
},
{
"epoch": 9.996884735202492,
"grad_norm": 1.8359375,
"learning_rate": 4.9799493626505653e-05,
"loss": 0.1461,
"step": 1609
},
{
"epoch": 10.0,
"grad_norm": 1.2109375,
"learning_rate": 4.9799182094065075e-05,
"loss": 0.0846,
"step": 1610
},
{
"epoch": 10.006230529595015,
"grad_norm": 3.40625,
"learning_rate": 4.9798870320769886e-05,
"loss": 0.3162,
"step": 1611
},
{
"epoch": 10.01246105919003,
"grad_norm": 2.890625,
"learning_rate": 4.979855830662311e-05,
"loss": 0.1844,
"step": 1612
},
{
"epoch": 10.018691588785046,
"grad_norm": 2.640625,
"learning_rate": 4.9798246051627776e-05,
"loss": 0.2943,
"step": 1613
},
{
"epoch": 10.024922118380061,
"grad_norm": 4.1875,
"learning_rate": 4.979793355578691e-05,
"loss": 0.4716,
"step": 1614
},
{
"epoch": 10.031152647975079,
"grad_norm": 2.90625,
"learning_rate": 4.9797620819103574e-05,
"loss": 0.3448,
"step": 1615
},
{
"epoch": 10.037383177570094,
"grad_norm": 3.546875,
"learning_rate": 4.979730784158078e-05,
"loss": 0.299,
"step": 1616
},
{
"epoch": 10.04361370716511,
"grad_norm": 2.453125,
"learning_rate": 4.979699462322157e-05,
"loss": 0.2789,
"step": 1617
},
{
"epoch": 10.049844236760125,
"grad_norm": 3.234375,
"learning_rate": 4.9796681164029e-05,
"loss": 0.3696,
"step": 1618
},
{
"epoch": 10.05607476635514,
"grad_norm": 2.765625,
"learning_rate": 4.97963674640061e-05,
"loss": 0.2211,
"step": 1619
},
{
"epoch": 10.062305295950155,
"grad_norm": 2.171875,
"learning_rate": 4.979605352315592e-05,
"loss": 0.157,
"step": 1620
},
{
"epoch": 10.06853582554517,
"grad_norm": 2.96875,
"learning_rate": 4.979573934148152e-05,
"loss": 0.2228,
"step": 1621
},
{
"epoch": 10.074766355140186,
"grad_norm": 2.8125,
"learning_rate": 4.9795424918985936e-05,
"loss": 0.3504,
"step": 1622
},
{
"epoch": 10.080996884735203,
"grad_norm": 1.953125,
"learning_rate": 4.979511025567223e-05,
"loss": 0.1943,
"step": 1623
},
{
"epoch": 10.087227414330219,
"grad_norm": 1.4140625,
"learning_rate": 4.979479535154346e-05,
"loss": 0.1135,
"step": 1624
},
{
"epoch": 10.093457943925234,
"grad_norm": 1.3046875,
"learning_rate": 4.979448020660268e-05,
"loss": 0.1203,
"step": 1625
},
{
"epoch": 10.09968847352025,
"grad_norm": 2.4375,
"learning_rate": 4.979416482085295e-05,
"loss": 0.1933,
"step": 1626
},
{
"epoch": 10.105919003115265,
"grad_norm": 1.6171875,
"learning_rate": 4.9793849194297335e-05,
"loss": 0.1373,
"step": 1627
},
{
"epoch": 10.11214953271028,
"grad_norm": 2.6875,
"learning_rate": 4.97935333269389e-05,
"loss": 0.2248,
"step": 1628
},
{
"epoch": 10.118380062305295,
"grad_norm": 2.5,
"learning_rate": 4.979321721878072e-05,
"loss": 0.2399,
"step": 1629
},
{
"epoch": 10.12461059190031,
"grad_norm": 2.640625,
"learning_rate": 4.9792900869825854e-05,
"loss": 0.3285,
"step": 1630
},
{
"epoch": 10.130841121495328,
"grad_norm": 3.140625,
"learning_rate": 4.979258428007738e-05,
"loss": 0.2709,
"step": 1631
},
{
"epoch": 10.137071651090343,
"grad_norm": 2.328125,
"learning_rate": 4.979226744953837e-05,
"loss": 0.2379,
"step": 1632
},
{
"epoch": 10.143302180685358,
"grad_norm": 1.6484375,
"learning_rate": 4.97919503782119e-05,
"loss": 0.1562,
"step": 1633
},
{
"epoch": 10.149532710280374,
"grad_norm": 2.234375,
"learning_rate": 4.979163306610105e-05,
"loss": 0.2213,
"step": 1634
},
{
"epoch": 10.15576323987539,
"grad_norm": 3.765625,
"learning_rate": 4.979131551320891e-05,
"loss": 0.3077,
"step": 1635
},
{
"epoch": 10.161993769470405,
"grad_norm": 1.5390625,
"learning_rate": 4.9790997719538554e-05,
"loss": 0.1196,
"step": 1636
},
{
"epoch": 10.16822429906542,
"grad_norm": 3.09375,
"learning_rate": 4.9790679685093075e-05,
"loss": 0.2924,
"step": 1637
},
{
"epoch": 10.174454828660435,
"grad_norm": 2.484375,
"learning_rate": 4.979036140987556e-05,
"loss": 0.2639,
"step": 1638
},
{
"epoch": 10.180685358255452,
"grad_norm": 3.828125,
"learning_rate": 4.97900428938891e-05,
"loss": 0.3353,
"step": 1639
},
{
"epoch": 10.186915887850468,
"grad_norm": 1.9765625,
"learning_rate": 4.978972413713678e-05,
"loss": 0.2401,
"step": 1640
},
{
"epoch": 10.193146417445483,
"grad_norm": 1.7421875,
"learning_rate": 4.9789405139621714e-05,
"loss": 0.1426,
"step": 1641
},
{
"epoch": 10.199376947040498,
"grad_norm": 2.21875,
"learning_rate": 4.978908590134699e-05,
"loss": 0.187,
"step": 1642
},
{
"epoch": 10.205607476635514,
"grad_norm": 2.40625,
"learning_rate": 4.97887664223157e-05,
"loss": 0.1985,
"step": 1643
},
{
"epoch": 10.21183800623053,
"grad_norm": 1.4296875,
"learning_rate": 4.978844670253096e-05,
"loss": 0.1227,
"step": 1644
},
{
"epoch": 10.218068535825545,
"grad_norm": 2.21875,
"learning_rate": 4.978812674199587e-05,
"loss": 0.2002,
"step": 1645
},
{
"epoch": 10.22429906542056,
"grad_norm": 2.421875,
"learning_rate": 4.9787806540713546e-05,
"loss": 0.2242,
"step": 1646
},
{
"epoch": 10.230529595015577,
"grad_norm": 1.6015625,
"learning_rate": 4.9787486098687076e-05,
"loss": 0.1358,
"step": 1647
},
{
"epoch": 10.236760124610592,
"grad_norm": 2.484375,
"learning_rate": 4.97871654159196e-05,
"loss": 0.1952,
"step": 1648
},
{
"epoch": 10.242990654205608,
"grad_norm": 2.390625,
"learning_rate": 4.978684449241421e-05,
"loss": 0.234,
"step": 1649
},
{
"epoch": 10.249221183800623,
"grad_norm": 1.953125,
"learning_rate": 4.978652332817403e-05,
"loss": 0.155,
"step": 1650
},
{
"epoch": 10.255451713395638,
"grad_norm": 2.90625,
"learning_rate": 4.978620192320218e-05,
"loss": 0.2544,
"step": 1651
},
{
"epoch": 10.261682242990654,
"grad_norm": 2.40625,
"learning_rate": 4.978588027750178e-05,
"loss": 0.1945,
"step": 1652
},
{
"epoch": 10.26791277258567,
"grad_norm": 2.109375,
"learning_rate": 4.978555839107597e-05,
"loss": 0.1795,
"step": 1653
},
{
"epoch": 10.274143302180685,
"grad_norm": 1.8359375,
"learning_rate": 4.9785236263927854e-05,
"loss": 0.1228,
"step": 1654
},
{
"epoch": 10.280373831775702,
"grad_norm": 1.984375,
"learning_rate": 4.978491389606057e-05,
"loss": 0.1917,
"step": 1655
},
{
"epoch": 10.286604361370717,
"grad_norm": 1.9296875,
"learning_rate": 4.978459128747724e-05,
"loss": 0.1515,
"step": 1656
},
{
"epoch": 10.292834890965732,
"grad_norm": 2.453125,
"learning_rate": 4.9784268438181016e-05,
"loss": 0.337,
"step": 1657
},
{
"epoch": 10.299065420560748,
"grad_norm": 1.71875,
"learning_rate": 4.978394534817502e-05,
"loss": 0.1445,
"step": 1658
},
{
"epoch": 10.305295950155763,
"grad_norm": 3.046875,
"learning_rate": 4.978362201746239e-05,
"loss": 0.2806,
"step": 1659
},
{
"epoch": 10.311526479750778,
"grad_norm": 2.21875,
"learning_rate": 4.978329844604627e-05,
"loss": 0.1788,
"step": 1660
},
{
"epoch": 10.317757009345794,
"grad_norm": 2.703125,
"learning_rate": 4.97829746339298e-05,
"loss": 0.3297,
"step": 1661
},
{
"epoch": 10.32398753894081,
"grad_norm": 3.8125,
"learning_rate": 4.978265058111613e-05,
"loss": 0.3753,
"step": 1662
},
{
"epoch": 10.330218068535826,
"grad_norm": 3.171875,
"learning_rate": 4.978232628760841e-05,
"loss": 0.3624,
"step": 1663
},
{
"epoch": 10.336448598130842,
"grad_norm": 1.9765625,
"learning_rate": 4.978200175340977e-05,
"loss": 0.23,
"step": 1664
},
{
"epoch": 10.342679127725857,
"grad_norm": 1.8203125,
"learning_rate": 4.978167697852338e-05,
"loss": 0.1557,
"step": 1665
},
{
"epoch": 10.348909657320872,
"grad_norm": 1.9453125,
"learning_rate": 4.9781351962952384e-05,
"loss": 0.1902,
"step": 1666
},
{
"epoch": 10.355140186915888,
"grad_norm": 1.78125,
"learning_rate": 4.978102670669995e-05,
"loss": 0.1413,
"step": 1667
},
{
"epoch": 10.361370716510903,
"grad_norm": 2.421875,
"learning_rate": 4.978070120976923e-05,
"loss": 0.2219,
"step": 1668
},
{
"epoch": 10.367601246105918,
"grad_norm": 1.7265625,
"learning_rate": 4.9780375472163385e-05,
"loss": 0.1963,
"step": 1669
},
{
"epoch": 10.373831775700934,
"grad_norm": 2.8125,
"learning_rate": 4.978004949388558e-05,
"loss": 0.2523,
"step": 1670
},
{
"epoch": 10.38006230529595,
"grad_norm": 2.5625,
"learning_rate": 4.977972327493899e-05,
"loss": 0.225,
"step": 1671
},
{
"epoch": 10.386292834890966,
"grad_norm": 2.703125,
"learning_rate": 4.9779396815326765e-05,
"loss": 0.2594,
"step": 1672
},
{
"epoch": 10.392523364485982,
"grad_norm": 2.046875,
"learning_rate": 4.977907011505208e-05,
"loss": 0.1579,
"step": 1673
},
{
"epoch": 10.398753894080997,
"grad_norm": 1.6796875,
"learning_rate": 4.977874317411813e-05,
"loss": 0.1428,
"step": 1674
},
{
"epoch": 10.404984423676012,
"grad_norm": 3.296875,
"learning_rate": 4.977841599252806e-05,
"loss": 0.2216,
"step": 1675
},
{
"epoch": 10.411214953271028,
"grad_norm": 2.953125,
"learning_rate": 4.977808857028506e-05,
"loss": 0.2573,
"step": 1676
},
{
"epoch": 10.417445482866043,
"grad_norm": 3.0625,
"learning_rate": 4.977776090739231e-05,
"loss": 0.2386,
"step": 1677
},
{
"epoch": 10.423676012461058,
"grad_norm": 1.5,
"learning_rate": 4.9777433003853e-05,
"loss": 0.1202,
"step": 1678
},
{
"epoch": 10.429906542056075,
"grad_norm": 1.453125,
"learning_rate": 4.977710485967031e-05,
"loss": 0.1286,
"step": 1679
},
{
"epoch": 10.43613707165109,
"grad_norm": 2.6875,
"learning_rate": 4.977677647484741e-05,
"loss": 0.3034,
"step": 1680
},
{
"epoch": 10.442367601246106,
"grad_norm": 1.984375,
"learning_rate": 4.977644784938752e-05,
"loss": 0.1777,
"step": 1681
},
{
"epoch": 10.448598130841122,
"grad_norm": 2.3125,
"learning_rate": 4.9776118983293805e-05,
"loss": 0.2191,
"step": 1682
},
{
"epoch": 10.454828660436137,
"grad_norm": 1.59375,
"learning_rate": 4.9775789876569476e-05,
"loss": 0.1258,
"step": 1683
},
{
"epoch": 10.461059190031152,
"grad_norm": 1.28125,
"learning_rate": 4.9775460529217724e-05,
"loss": 0.1072,
"step": 1684
},
{
"epoch": 10.467289719626168,
"grad_norm": 2.046875,
"learning_rate": 4.9775130941241747e-05,
"loss": 0.2662,
"step": 1685
},
{
"epoch": 10.473520249221183,
"grad_norm": 1.8359375,
"learning_rate": 4.977480111264474e-05,
"loss": 0.1421,
"step": 1686
},
{
"epoch": 10.4797507788162,
"grad_norm": 2.265625,
"learning_rate": 4.977447104342992e-05,
"loss": 0.2492,
"step": 1687
},
{
"epoch": 10.485981308411215,
"grad_norm": 2.375,
"learning_rate": 4.9774140733600475e-05,
"loss": 0.2886,
"step": 1688
},
{
"epoch": 10.49221183800623,
"grad_norm": 3.09375,
"learning_rate": 4.977381018315963e-05,
"loss": 0.2295,
"step": 1689
},
{
"epoch": 10.498442367601246,
"grad_norm": 1.375,
"learning_rate": 4.977347939211058e-05,
"loss": 0.1041,
"step": 1690
},
{
"epoch": 10.504672897196262,
"grad_norm": 2.40625,
"learning_rate": 4.9773148360456555e-05,
"loss": 0.2903,
"step": 1691
},
{
"epoch": 10.510903426791277,
"grad_norm": 3.0625,
"learning_rate": 4.977281708820076e-05,
"loss": 0.2962,
"step": 1692
},
{
"epoch": 10.517133956386292,
"grad_norm": 3.65625,
"learning_rate": 4.977248557534641e-05,
"loss": 0.461,
"step": 1693
},
{
"epoch": 10.523364485981308,
"grad_norm": 2.859375,
"learning_rate": 4.9772153821896725e-05,
"loss": 0.4168,
"step": 1694
},
{
"epoch": 10.529595015576325,
"grad_norm": 2.84375,
"learning_rate": 4.977182182785494e-05,
"loss": 0.2739,
"step": 1695
},
{
"epoch": 10.53582554517134,
"grad_norm": 1.6484375,
"learning_rate": 4.977148959322426e-05,
"loss": 0.136,
"step": 1696
},
{
"epoch": 10.542056074766355,
"grad_norm": 2.3125,
"learning_rate": 4.977115711800793e-05,
"loss": 0.2081,
"step": 1697
},
{
"epoch": 10.54828660436137,
"grad_norm": 2.53125,
"learning_rate": 4.977082440220917e-05,
"loss": 0.2135,
"step": 1698
},
{
"epoch": 10.554517133956386,
"grad_norm": 2.6875,
"learning_rate": 4.977049144583121e-05,
"loss": 0.2211,
"step": 1699
},
{
"epoch": 10.560747663551401,
"grad_norm": 1.953125,
"learning_rate": 4.9770158248877286e-05,
"loss": 0.1578,
"step": 1700
},
{
"epoch": 10.566978193146417,
"grad_norm": 4.21875,
"learning_rate": 4.976982481135063e-05,
"loss": 0.4271,
"step": 1701
},
{
"epoch": 10.573208722741432,
"grad_norm": 2.703125,
"learning_rate": 4.9769491133254485e-05,
"loss": 0.2162,
"step": 1702
},
{
"epoch": 10.57943925233645,
"grad_norm": 1.6171875,
"learning_rate": 4.976915721459209e-05,
"loss": 0.1122,
"step": 1703
},
{
"epoch": 10.585669781931465,
"grad_norm": 3.640625,
"learning_rate": 4.976882305536669e-05,
"loss": 0.3504,
"step": 1704
},
{
"epoch": 10.59190031152648,
"grad_norm": 1.8671875,
"learning_rate": 4.976848865558153e-05,
"loss": 0.1443,
"step": 1705
},
{
"epoch": 10.598130841121495,
"grad_norm": 2.59375,
"learning_rate": 4.9768154015239854e-05,
"loss": 0.2229,
"step": 1706
},
{
"epoch": 10.60436137071651,
"grad_norm": 2.75,
"learning_rate": 4.9767819134344914e-05,
"loss": 0.2412,
"step": 1707
},
{
"epoch": 10.610591900311526,
"grad_norm": 2.5625,
"learning_rate": 4.976748401289997e-05,
"loss": 0.2515,
"step": 1708
},
{
"epoch": 10.616822429906541,
"grad_norm": 1.796875,
"learning_rate": 4.976714865090827e-05,
"loss": 0.1439,
"step": 1709
},
{
"epoch": 10.623052959501557,
"grad_norm": 2.4375,
"learning_rate": 4.976681304837307e-05,
"loss": 0.2239,
"step": 1710
},
{
"epoch": 10.629283489096574,
"grad_norm": 2.5625,
"learning_rate": 4.9766477205297634e-05,
"loss": 0.248,
"step": 1711
},
{
"epoch": 10.63551401869159,
"grad_norm": 1.6171875,
"learning_rate": 4.976614112168522e-05,
"loss": 0.1344,
"step": 1712
},
{
"epoch": 10.641744548286605,
"grad_norm": 1.671875,
"learning_rate": 4.9765804797539086e-05,
"loss": 0.1276,
"step": 1713
},
{
"epoch": 10.64797507788162,
"grad_norm": 1.421875,
"learning_rate": 4.9765468232862505e-05,
"loss": 0.1291,
"step": 1714
},
{
"epoch": 10.654205607476635,
"grad_norm": 2.578125,
"learning_rate": 4.976513142765875e-05,
"loss": 0.2585,
"step": 1715
},
{
"epoch": 10.66043613707165,
"grad_norm": 2.0625,
"learning_rate": 4.976479438193109e-05,
"loss": 0.1774,
"step": 1716
},
{
"epoch": 10.666666666666666,
"grad_norm": 2.0625,
"learning_rate": 4.976445709568279e-05,
"loss": 0.1887,
"step": 1717
},
{
"epoch": 10.672897196261681,
"grad_norm": 2.921875,
"learning_rate": 4.976411956891714e-05,
"loss": 0.3024,
"step": 1718
},
{
"epoch": 10.679127725856699,
"grad_norm": 2.46875,
"learning_rate": 4.976378180163741e-05,
"loss": 0.2105,
"step": 1719
},
{
"epoch": 10.685358255451714,
"grad_norm": 2.796875,
"learning_rate": 4.976344379384688e-05,
"loss": 0.2695,
"step": 1720
},
{
"epoch": 10.69158878504673,
"grad_norm": 1.8828125,
"learning_rate": 4.976310554554883e-05,
"loss": 0.163,
"step": 1721
},
{
"epoch": 10.697819314641745,
"grad_norm": 2.59375,
"learning_rate": 4.976276705674655e-05,
"loss": 0.2065,
"step": 1722
},
{
"epoch": 10.70404984423676,
"grad_norm": 2.21875,
"learning_rate": 4.976242832744332e-05,
"loss": 0.2196,
"step": 1723
},
{
"epoch": 10.710280373831775,
"grad_norm": 2.46875,
"learning_rate": 4.976208935764245e-05,
"loss": 0.2938,
"step": 1724
},
{
"epoch": 10.71651090342679,
"grad_norm": 2.375,
"learning_rate": 4.976175014734721e-05,
"loss": 0.2685,
"step": 1725
},
{
"epoch": 10.722741433021806,
"grad_norm": 1.84375,
"learning_rate": 4.976141069656091e-05,
"loss": 0.2164,
"step": 1726
},
{
"epoch": 10.728971962616823,
"grad_norm": 1.6640625,
"learning_rate": 4.976107100528683e-05,
"loss": 0.13,
"step": 1727
},
{
"epoch": 10.735202492211839,
"grad_norm": 3.453125,
"learning_rate": 4.9760731073528285e-05,
"loss": 0.3868,
"step": 1728
},
{
"epoch": 10.741433021806854,
"grad_norm": 1.2421875,
"learning_rate": 4.976039090128857e-05,
"loss": 0.1038,
"step": 1729
},
{
"epoch": 10.74766355140187,
"grad_norm": 2.59375,
"learning_rate": 4.9760050488570984e-05,
"loss": 0.23,
"step": 1730
},
{
"epoch": 10.753894080996885,
"grad_norm": 2.046875,
"learning_rate": 4.975970983537884e-05,
"loss": 0.1765,
"step": 1731
},
{
"epoch": 10.7601246105919,
"grad_norm": 2.265625,
"learning_rate": 4.9759368941715445e-05,
"loss": 0.1723,
"step": 1732
},
{
"epoch": 10.766355140186915,
"grad_norm": 1.4609375,
"learning_rate": 4.975902780758411e-05,
"loss": 0.1267,
"step": 1733
},
{
"epoch": 10.77258566978193,
"grad_norm": 1.7265625,
"learning_rate": 4.9758686432988155e-05,
"loss": 0.1594,
"step": 1734
},
{
"epoch": 10.778816199376948,
"grad_norm": 1.59375,
"learning_rate": 4.975834481793088e-05,
"loss": 0.1331,
"step": 1735
},
{
"epoch": 10.785046728971963,
"grad_norm": 3.46875,
"learning_rate": 4.975800296241561e-05,
"loss": 0.4277,
"step": 1736
},
{
"epoch": 10.791277258566979,
"grad_norm": 2.1875,
"learning_rate": 4.975766086644567e-05,
"loss": 0.2175,
"step": 1737
},
{
"epoch": 10.797507788161994,
"grad_norm": 3.421875,
"learning_rate": 4.9757318530024376e-05,
"loss": 0.2801,
"step": 1738
},
{
"epoch": 10.80373831775701,
"grad_norm": 2.5,
"learning_rate": 4.975697595315506e-05,
"loss": 0.2282,
"step": 1739
},
{
"epoch": 10.809968847352025,
"grad_norm": 1.4453125,
"learning_rate": 4.975663313584104e-05,
"loss": 0.1272,
"step": 1740
},
{
"epoch": 10.81619937694704,
"grad_norm": 1.2890625,
"learning_rate": 4.975629007808565e-05,
"loss": 0.1181,
"step": 1741
},
{
"epoch": 10.822429906542055,
"grad_norm": 1.2421875,
"learning_rate": 4.9755946779892225e-05,
"loss": 0.1028,
"step": 1742
},
{
"epoch": 10.828660436137072,
"grad_norm": 3.953125,
"learning_rate": 4.97556032412641e-05,
"loss": 0.4362,
"step": 1743
},
{
"epoch": 10.834890965732088,
"grad_norm": 2.703125,
"learning_rate": 4.9755259462204604e-05,
"loss": 0.2065,
"step": 1744
},
{
"epoch": 10.841121495327103,
"grad_norm": 2.59375,
"learning_rate": 4.975491544271708e-05,
"loss": 0.2354,
"step": 1745
},
{
"epoch": 10.847352024922118,
"grad_norm": 1.890625,
"learning_rate": 4.975457118280487e-05,
"loss": 0.156,
"step": 1746
},
{
"epoch": 10.853582554517134,
"grad_norm": 2.609375,
"learning_rate": 4.975422668247131e-05,
"loss": 0.2213,
"step": 1747
},
{
"epoch": 10.85981308411215,
"grad_norm": 2.421875,
"learning_rate": 4.975388194171975e-05,
"loss": 0.2686,
"step": 1748
},
{
"epoch": 10.866043613707165,
"grad_norm": 1.703125,
"learning_rate": 4.9753536960553545e-05,
"loss": 0.1357,
"step": 1749
},
{
"epoch": 10.87227414330218,
"grad_norm": 2.296875,
"learning_rate": 4.9753191738976045e-05,
"loss": 0.1998,
"step": 1750
},
{
"epoch": 10.878504672897197,
"grad_norm": 1.6640625,
"learning_rate": 4.97528462769906e-05,
"loss": 0.1478,
"step": 1751
},
{
"epoch": 10.884735202492212,
"grad_norm": 2.671875,
"learning_rate": 4.9752500574600556e-05,
"loss": 0.222,
"step": 1752
},
{
"epoch": 10.890965732087228,
"grad_norm": 1.046875,
"learning_rate": 4.975215463180929e-05,
"loss": 0.0979,
"step": 1753
},
{
"epoch": 10.897196261682243,
"grad_norm": 1.515625,
"learning_rate": 4.975180844862014e-05,
"loss": 0.1128,
"step": 1754
},
{
"epoch": 10.903426791277258,
"grad_norm": 1.4296875,
"learning_rate": 4.975146202503648e-05,
"loss": 0.1341,
"step": 1755
},
{
"epoch": 10.909657320872274,
"grad_norm": 2.71875,
"learning_rate": 4.975111536106167e-05,
"loss": 0.2905,
"step": 1756
},
{
"epoch": 10.91588785046729,
"grad_norm": 1.59375,
"learning_rate": 4.9750768456699095e-05,
"loss": 0.1381,
"step": 1757
},
{
"epoch": 10.922118380062305,
"grad_norm": 1.78125,
"learning_rate": 4.975042131195209e-05,
"loss": 0.1477,
"step": 1758
},
{
"epoch": 10.928348909657322,
"grad_norm": 2.390625,
"learning_rate": 4.975007392682406e-05,
"loss": 0.216,
"step": 1759
},
{
"epoch": 10.934579439252337,
"grad_norm": 2.5,
"learning_rate": 4.974972630131836e-05,
"loss": 0.2315,
"step": 1760
},
{
"epoch": 10.940809968847352,
"grad_norm": 2.078125,
"learning_rate": 4.974937843543836e-05,
"loss": 0.1682,
"step": 1761
},
{
"epoch": 10.947040498442368,
"grad_norm": 1.484375,
"learning_rate": 4.9749030329187466e-05,
"loss": 0.1408,
"step": 1762
},
{
"epoch": 10.953271028037383,
"grad_norm": 2.234375,
"learning_rate": 4.9748681982569035e-05,
"loss": 0.1862,
"step": 1763
},
{
"epoch": 10.959501557632398,
"grad_norm": 2.859375,
"learning_rate": 4.974833339558645e-05,
"loss": 0.3211,
"step": 1764
},
{
"epoch": 10.965732087227414,
"grad_norm": 2.109375,
"learning_rate": 4.974798456824312e-05,
"loss": 0.2061,
"step": 1765
},
{
"epoch": 10.97196261682243,
"grad_norm": 2.1875,
"learning_rate": 4.974763550054241e-05,
"loss": 0.2293,
"step": 1766
},
{
"epoch": 10.978193146417446,
"grad_norm": 1.9140625,
"learning_rate": 4.974728619248772e-05,
"loss": 0.1504,
"step": 1767
},
{
"epoch": 10.984423676012462,
"grad_norm": 2.0625,
"learning_rate": 4.974693664408243e-05,
"loss": 0.1593,
"step": 1768
},
{
"epoch": 10.990654205607477,
"grad_norm": 3.078125,
"learning_rate": 4.9746586855329944e-05,
"loss": 0.279,
"step": 1769
},
{
"epoch": 10.996884735202492,
"grad_norm": 2.15625,
"learning_rate": 4.9746236826233674e-05,
"loss": 0.1967,
"step": 1770
},
{
"epoch": 11.0,
"grad_norm": 0.75,
"learning_rate": 4.9745886556796995e-05,
"loss": 0.0474,
"step": 1771
},
{
"epoch": 11.006230529595015,
"grad_norm": 2.640625,
"learning_rate": 4.9745536047023324e-05,
"loss": 0.2473,
"step": 1772
},
{
"epoch": 11.01246105919003,
"grad_norm": 2.390625,
"learning_rate": 4.9745185296916054e-05,
"loss": 0.1999,
"step": 1773
},
{
"epoch": 11.018691588785046,
"grad_norm": 1.6953125,
"learning_rate": 4.97448343064786e-05,
"loss": 0.1439,
"step": 1774
},
{
"epoch": 11.024922118380061,
"grad_norm": 2.546875,
"learning_rate": 4.974448307571437e-05,
"loss": 0.2885,
"step": 1775
},
{
"epoch": 11.031152647975079,
"grad_norm": 2.078125,
"learning_rate": 4.974413160462678e-05,
"loss": 0.1661,
"step": 1776
},
{
"epoch": 11.037383177570094,
"grad_norm": 2.4375,
"learning_rate": 4.974377989321923e-05,
"loss": 0.2039,
"step": 1777
},
{
"epoch": 11.04361370716511,
"grad_norm": 2.5,
"learning_rate": 4.974342794149515e-05,
"loss": 0.1695,
"step": 1778
},
{
"epoch": 11.049844236760125,
"grad_norm": 1.3515625,
"learning_rate": 4.974307574945794e-05,
"loss": 0.1093,
"step": 1779
},
{
"epoch": 11.05607476635514,
"grad_norm": 2.5625,
"learning_rate": 4.974272331711104e-05,
"loss": 0.3766,
"step": 1780
},
{
"epoch": 11.062305295950155,
"grad_norm": 1.40625,
"learning_rate": 4.974237064445787e-05,
"loss": 0.1116,
"step": 1781
},
{
"epoch": 11.06853582554517,
"grad_norm": 1.1953125,
"learning_rate": 4.974201773150184e-05,
"loss": 0.1074,
"step": 1782
},
{
"epoch": 11.074766355140186,
"grad_norm": 2.0,
"learning_rate": 4.9741664578246394e-05,
"loss": 0.1674,
"step": 1783
},
{
"epoch": 11.080996884735203,
"grad_norm": 1.8359375,
"learning_rate": 4.974131118469495e-05,
"loss": 0.1637,
"step": 1784
},
{
"epoch": 11.087227414330219,
"grad_norm": 1.28125,
"learning_rate": 4.9740957550850956e-05,
"loss": 0.1055,
"step": 1785
},
{
"epoch": 11.093457943925234,
"grad_norm": 1.65625,
"learning_rate": 4.974060367671783e-05,
"loss": 0.1346,
"step": 1786
},
{
"epoch": 11.09968847352025,
"grad_norm": 1.8515625,
"learning_rate": 4.974024956229901e-05,
"loss": 0.1722,
"step": 1787
},
{
"epoch": 11.105919003115265,
"grad_norm": 2.265625,
"learning_rate": 4.973989520759795e-05,
"loss": 0.2293,
"step": 1788
},
{
"epoch": 11.11214953271028,
"grad_norm": 2.421875,
"learning_rate": 4.973954061261809e-05,
"loss": 0.2891,
"step": 1789
},
{
"epoch": 11.118380062305295,
"grad_norm": 3.25,
"learning_rate": 4.973918577736285e-05,
"loss": 0.3023,
"step": 1790
},
{
"epoch": 11.12461059190031,
"grad_norm": 2.0,
"learning_rate": 4.97388307018357e-05,
"loss": 0.2183,
"step": 1791
},
{
"epoch": 11.130841121495328,
"grad_norm": 3.109375,
"learning_rate": 4.973847538604008e-05,
"loss": 0.3105,
"step": 1792
},
{
"epoch": 11.137071651090343,
"grad_norm": 2.40625,
"learning_rate": 4.973811982997944e-05,
"loss": 0.3001,
"step": 1793
},
{
"epoch": 11.143302180685358,
"grad_norm": 2.265625,
"learning_rate": 4.973776403365724e-05,
"loss": 0.1905,
"step": 1794
},
{
"epoch": 11.149532710280374,
"grad_norm": 3.234375,
"learning_rate": 4.973740799707692e-05,
"loss": 0.2825,
"step": 1795
},
{
"epoch": 11.15576323987539,
"grad_norm": 2.171875,
"learning_rate": 4.973705172024196e-05,
"loss": 0.1509,
"step": 1796
},
{
"epoch": 11.161993769470405,
"grad_norm": 1.25,
"learning_rate": 4.9736695203155805e-05,
"loss": 0.1087,
"step": 1797
},
{
"epoch": 11.16822429906542,
"grad_norm": 3.296875,
"learning_rate": 4.9736338445821925e-05,
"loss": 0.412,
"step": 1798
},
{
"epoch": 11.174454828660435,
"grad_norm": 1.609375,
"learning_rate": 4.973598144824377e-05,
"loss": 0.1208,
"step": 1799
},
{
"epoch": 11.180685358255452,
"grad_norm": 2.640625,
"learning_rate": 4.973562421042483e-05,
"loss": 0.2741,
"step": 1800
},
{
"epoch": 11.186915887850468,
"grad_norm": 2.59375,
"learning_rate": 4.973526673236856e-05,
"loss": 0.2072,
"step": 1801
},
{
"epoch": 11.193146417445483,
"grad_norm": 1.65625,
"learning_rate": 4.9734909014078435e-05,
"loss": 0.2099,
"step": 1802
},
{
"epoch": 11.199376947040498,
"grad_norm": 1.7265625,
"learning_rate": 4.973455105555793e-05,
"loss": 0.1198,
"step": 1803
},
{
"epoch": 11.205607476635514,
"grad_norm": 2.796875,
"learning_rate": 4.973419285681052e-05,
"loss": 0.2748,
"step": 1804
},
{
"epoch": 11.21183800623053,
"grad_norm": 1.6953125,
"learning_rate": 4.973383441783969e-05,
"loss": 0.1523,
"step": 1805
},
{
"epoch": 11.218068535825545,
"grad_norm": 2.796875,
"learning_rate": 4.973347573864891e-05,
"loss": 0.2345,
"step": 1806
},
{
"epoch": 11.22429906542056,
"grad_norm": 2.125,
"learning_rate": 4.9733116819241664e-05,
"loss": 0.1693,
"step": 1807
},
{
"epoch": 11.230529595015577,
"grad_norm": 2.484375,
"learning_rate": 4.973275765962145e-05,
"loss": 0.2228,
"step": 1808
},
{
"epoch": 11.236760124610592,
"grad_norm": 2.25,
"learning_rate": 4.973239825979175e-05,
"loss": 0.2332,
"step": 1809
},
{
"epoch": 11.242990654205608,
"grad_norm": 2.453125,
"learning_rate": 4.973203861975605e-05,
"loss": 0.1706,
"step": 1810
},
{
"epoch": 11.249221183800623,
"grad_norm": 1.890625,
"learning_rate": 4.973167873951785e-05,
"loss": 0.1327,
"step": 1811
},
{
"epoch": 11.255451713395638,
"grad_norm": 2.21875,
"learning_rate": 4.973131861908064e-05,
"loss": 0.1717,
"step": 1812
},
{
"epoch": 11.261682242990654,
"grad_norm": 2.96875,
"learning_rate": 4.973095825844793e-05,
"loss": 0.228,
"step": 1813
},
{
"epoch": 11.26791277258567,
"grad_norm": 1.7109375,
"learning_rate": 4.973059765762319e-05,
"loss": 0.1332,
"step": 1814
},
{
"epoch": 11.274143302180685,
"grad_norm": 2.734375,
"learning_rate": 4.973023681660996e-05,
"loss": 0.2524,
"step": 1815
},
{
"epoch": 11.280373831775702,
"grad_norm": 1.6796875,
"learning_rate": 4.972987573541172e-05,
"loss": 0.1292,
"step": 1816
},
{
"epoch": 11.286604361370717,
"grad_norm": 1.4375,
"learning_rate": 4.9729514414031985e-05,
"loss": 0.1136,
"step": 1817
},
{
"epoch": 11.292834890965732,
"grad_norm": 2.65625,
"learning_rate": 4.972915285247426e-05,
"loss": 0.2002,
"step": 1818
},
{
"epoch": 11.299065420560748,
"grad_norm": 1.3515625,
"learning_rate": 4.9728791050742064e-05,
"loss": 0.0982,
"step": 1819
},
{
"epoch": 11.305295950155763,
"grad_norm": 1.3828125,
"learning_rate": 4.9728429008838906e-05,
"loss": 0.107,
"step": 1820
},
{
"epoch": 11.311526479750778,
"grad_norm": 1.3203125,
"learning_rate": 4.97280667267683e-05,
"loss": 0.1005,
"step": 1821
},
{
"epoch": 11.317757009345794,
"grad_norm": 2.109375,
"learning_rate": 4.972770420453376e-05,
"loss": 0.1847,
"step": 1822
},
{
"epoch": 11.32398753894081,
"grad_norm": 2.75,
"learning_rate": 4.972734144213882e-05,
"loss": 0.2239,
"step": 1823
},
{
"epoch": 11.330218068535826,
"grad_norm": 1.765625,
"learning_rate": 4.9726978439587e-05,
"loss": 0.123,
"step": 1824
},
{
"epoch": 11.336448598130842,
"grad_norm": 2.953125,
"learning_rate": 4.972661519688182e-05,
"loss": 0.2348,
"step": 1825
},
{
"epoch": 11.342679127725857,
"grad_norm": 1.3359375,
"learning_rate": 4.9726251714026804e-05,
"loss": 0.1172,
"step": 1826
},
{
"epoch": 11.348909657320872,
"grad_norm": 2.15625,
"learning_rate": 4.97258879910255e-05,
"loss": 0.2203,
"step": 1827
},
{
"epoch": 11.355140186915888,
"grad_norm": 1.359375,
"learning_rate": 4.972552402788142e-05,
"loss": 0.1102,
"step": 1828
},
{
"epoch": 11.361370716510903,
"grad_norm": 1.9140625,
"learning_rate": 4.972515982459812e-05,
"loss": 0.1631,
"step": 1829
},
{
"epoch": 11.367601246105918,
"grad_norm": 3.265625,
"learning_rate": 4.972479538117911e-05,
"loss": 0.3766,
"step": 1830
},
{
"epoch": 11.373831775700934,
"grad_norm": 2.40625,
"learning_rate": 4.972443069762795e-05,
"loss": 0.17,
"step": 1831
},
{
"epoch": 11.38006230529595,
"grad_norm": 2.984375,
"learning_rate": 4.9724065773948185e-05,
"loss": 0.2224,
"step": 1832
},
{
"epoch": 11.386292834890966,
"grad_norm": 2.859375,
"learning_rate": 4.9723700610143344e-05,
"loss": 0.2957,
"step": 1833
},
{
"epoch": 11.392523364485982,
"grad_norm": 2.546875,
"learning_rate": 4.972333520621698e-05,
"loss": 0.1669,
"step": 1834
},
{
"epoch": 11.398753894080997,
"grad_norm": 1.859375,
"learning_rate": 4.972296956217265e-05,
"loss": 0.1273,
"step": 1835
},
{
"epoch": 11.404984423676012,
"grad_norm": 2.53125,
"learning_rate": 4.972260367801389e-05,
"loss": 0.1724,
"step": 1836
},
{
"epoch": 11.411214953271028,
"grad_norm": 2.71875,
"learning_rate": 4.972223755374426e-05,
"loss": 0.3126,
"step": 1837
},
{
"epoch": 11.417445482866043,
"grad_norm": 2.828125,
"learning_rate": 4.972187118936732e-05,
"loss": 0.311,
"step": 1838
},
{
"epoch": 11.423676012461058,
"grad_norm": 3.5625,
"learning_rate": 4.9721504584886624e-05,
"loss": 0.2452,
"step": 1839
},
{
"epoch": 11.429906542056075,
"grad_norm": 2.171875,
"learning_rate": 4.9721137740305734e-05,
"loss": 0.1502,
"step": 1840
},
{
"epoch": 11.43613707165109,
"grad_norm": 3.0625,
"learning_rate": 4.972077065562821e-05,
"loss": 0.2773,
"step": 1841
},
{
"epoch": 11.442367601246106,
"grad_norm": 1.703125,
"learning_rate": 4.972040333085763e-05,
"loss": 0.1326,
"step": 1842
},
{
"epoch": 11.448598130841122,
"grad_norm": 2.453125,
"learning_rate": 4.972003576599754e-05,
"loss": 0.2429,
"step": 1843
},
{
"epoch": 11.454828660436137,
"grad_norm": 2.890625,
"learning_rate": 4.971966796105153e-05,
"loss": 0.2395,
"step": 1844
},
{
"epoch": 11.461059190031152,
"grad_norm": 2.234375,
"learning_rate": 4.971929991602315e-05,
"loss": 0.1979,
"step": 1845
},
{
"epoch": 11.467289719626168,
"grad_norm": 3.140625,
"learning_rate": 4.9718931630916e-05,
"loss": 0.2384,
"step": 1846
},
{
"epoch": 11.473520249221183,
"grad_norm": 2.234375,
"learning_rate": 4.971856310573365e-05,
"loss": 0.2112,
"step": 1847
},
{
"epoch": 11.4797507788162,
"grad_norm": 1.6953125,
"learning_rate": 4.9718194340479663e-05,
"loss": 0.136,
"step": 1848
},
{
"epoch": 11.485981308411215,
"grad_norm": 1.9140625,
"learning_rate": 4.971782533515763e-05,
"loss": 0.1602,
"step": 1849
},
{
"epoch": 11.49221183800623,
"grad_norm": 2.15625,
"learning_rate": 4.971745608977114e-05,
"loss": 0.1366,
"step": 1850
},
{
"epoch": 11.498442367601246,
"grad_norm": 3.015625,
"learning_rate": 4.971708660432378e-05,
"loss": 0.2083,
"step": 1851
},
{
"epoch": 11.504672897196262,
"grad_norm": 2.15625,
"learning_rate": 4.971671687881913e-05,
"loss": 0.2385,
"step": 1852
},
{
"epoch": 11.510903426791277,
"grad_norm": 2.1875,
"learning_rate": 4.971634691326079e-05,
"loss": 0.1667,
"step": 1853
},
{
"epoch": 11.517133956386292,
"grad_norm": 1.5859375,
"learning_rate": 4.971597670765234e-05,
"loss": 0.1173,
"step": 1854
},
{
"epoch": 11.523364485981308,
"grad_norm": 1.6328125,
"learning_rate": 4.9715606261997386e-05,
"loss": 0.1275,
"step": 1855
},
{
"epoch": 11.529595015576325,
"grad_norm": 2.3125,
"learning_rate": 4.9715235576299523e-05,
"loss": 0.1873,
"step": 1856
},
{
"epoch": 11.53582554517134,
"grad_norm": 2.421875,
"learning_rate": 4.9714864650562356e-05,
"loss": 0.2615,
"step": 1857
},
{
"epoch": 11.542056074766355,
"grad_norm": 1.640625,
"learning_rate": 4.9714493484789476e-05,
"loss": 0.1258,
"step": 1858
},
{
"epoch": 11.54828660436137,
"grad_norm": 2.703125,
"learning_rate": 4.971412207898451e-05,
"loss": 0.1977,
"step": 1859
},
{
"epoch": 11.554517133956386,
"grad_norm": 1.8359375,
"learning_rate": 4.971375043315103e-05,
"loss": 0.1286,
"step": 1860
},
{
"epoch": 11.560747663551401,
"grad_norm": 2.28125,
"learning_rate": 4.971337854729268e-05,
"loss": 0.1769,
"step": 1861
},
{
"epoch": 11.566978193146417,
"grad_norm": 3.578125,
"learning_rate": 4.9713006421413047e-05,
"loss": 0.2715,
"step": 1862
},
{
"epoch": 11.573208722741432,
"grad_norm": 2.625,
"learning_rate": 4.971263405551576e-05,
"loss": 0.2011,
"step": 1863
},
{
"epoch": 11.57943925233645,
"grad_norm": 2.265625,
"learning_rate": 4.971226144960443e-05,
"loss": 0.2537,
"step": 1864
},
{
"epoch": 11.585669781931465,
"grad_norm": 1.7734375,
"learning_rate": 4.971188860368268e-05,
"loss": 0.1248,
"step": 1865
},
{
"epoch": 11.59190031152648,
"grad_norm": 1.9453125,
"learning_rate": 4.971151551775413e-05,
"loss": 0.1326,
"step": 1866
},
{
"epoch": 11.598130841121495,
"grad_norm": 1.2734375,
"learning_rate": 4.971114219182239e-05,
"loss": 0.119,
"step": 1867
},
{
"epoch": 11.60436137071651,
"grad_norm": 2.28125,
"learning_rate": 4.9710768625891114e-05,
"loss": 0.248,
"step": 1868
},
{
"epoch": 11.610591900311526,
"grad_norm": 1.7109375,
"learning_rate": 4.97103948199639e-05,
"loss": 0.1094,
"step": 1869
},
{
"epoch": 11.616822429906541,
"grad_norm": 1.84375,
"learning_rate": 4.9710020774044394e-05,
"loss": 0.1232,
"step": 1870
},
{
"epoch": 11.623052959501557,
"grad_norm": 1.4921875,
"learning_rate": 4.970964648813623e-05,
"loss": 0.1095,
"step": 1871
},
{
"epoch": 11.629283489096574,
"grad_norm": 2.671875,
"learning_rate": 4.9709271962243044e-05,
"loss": 0.2553,
"step": 1872
},
{
"epoch": 11.63551401869159,
"grad_norm": 1.4609375,
"learning_rate": 4.970889719636847e-05,
"loss": 0.1222,
"step": 1873
},
{
"epoch": 11.641744548286605,
"grad_norm": 1.2578125,
"learning_rate": 4.9708522190516135e-05,
"loss": 0.1078,
"step": 1874
},
{
"epoch": 11.64797507788162,
"grad_norm": 2.140625,
"learning_rate": 4.9708146944689705e-05,
"loss": 0.1601,
"step": 1875
},
{
"epoch": 11.654205607476635,
"grad_norm": 2.40625,
"learning_rate": 4.9707771458892804e-05,
"loss": 0.208,
"step": 1876
},
{
"epoch": 11.66043613707165,
"grad_norm": 2.515625,
"learning_rate": 4.970739573312909e-05,
"loss": 0.2072,
"step": 1877
},
{
"epoch": 11.666666666666666,
"grad_norm": 2.140625,
"learning_rate": 4.970701976740222e-05,
"loss": 0.1701,
"step": 1878
},
{
"epoch": 11.672897196261681,
"grad_norm": 2.265625,
"learning_rate": 4.970664356171583e-05,
"loss": 0.248,
"step": 1879
},
{
"epoch": 11.679127725856699,
"grad_norm": 1.7421875,
"learning_rate": 4.9706267116073565e-05,
"loss": 0.1338,
"step": 1880
},
{
"epoch": 11.685358255451714,
"grad_norm": 2.296875,
"learning_rate": 4.970589043047911e-05,
"loss": 0.1946,
"step": 1881
},
{
"epoch": 11.69158878504673,
"grad_norm": 2.40625,
"learning_rate": 4.970551350493611e-05,
"loss": 0.1582,
"step": 1882
},
{
"epoch": 11.697819314641745,
"grad_norm": 2.125,
"learning_rate": 4.9705136339448214e-05,
"loss": 0.1436,
"step": 1883
},
{
"epoch": 11.70404984423676,
"grad_norm": 1.625,
"learning_rate": 4.97047589340191e-05,
"loss": 0.1507,
"step": 1884
},
{
"epoch": 11.710280373831775,
"grad_norm": 1.671875,
"learning_rate": 4.9704381288652436e-05,
"loss": 0.1237,
"step": 1885
},
{
"epoch": 11.71651090342679,
"grad_norm": 1.3984375,
"learning_rate": 4.970400340335187e-05,
"loss": 0.1227,
"step": 1886
},
{
"epoch": 11.722741433021806,
"grad_norm": 2.359375,
"learning_rate": 4.970362527812109e-05,
"loss": 0.2665,
"step": 1887
},
{
"epoch": 11.728971962616823,
"grad_norm": 1.8984375,
"learning_rate": 4.9703246912963764e-05,
"loss": 0.1566,
"step": 1888
},
{
"epoch": 11.735202492211839,
"grad_norm": 2.34375,
"learning_rate": 4.970286830788357e-05,
"loss": 0.2407,
"step": 1889
},
{
"epoch": 11.741433021806854,
"grad_norm": 3.375,
"learning_rate": 4.970248946288417e-05,
"loss": 0.3571,
"step": 1890
},
{
"epoch": 11.74766355140187,
"grad_norm": 4.21875,
"learning_rate": 4.970211037796927e-05,
"loss": 0.4333,
"step": 1891
},
{
"epoch": 11.753894080996885,
"grad_norm": 2.171875,
"learning_rate": 4.970173105314252e-05,
"loss": 0.1818,
"step": 1892
},
{
"epoch": 11.7601246105919,
"grad_norm": 1.71875,
"learning_rate": 4.970135148840763e-05,
"loss": 0.1232,
"step": 1893
},
{
"epoch": 11.766355140186915,
"grad_norm": 2.953125,
"learning_rate": 4.970097168376827e-05,
"loss": 0.1983,
"step": 1894
},
{
"epoch": 11.77258566978193,
"grad_norm": 2.171875,
"learning_rate": 4.9700591639228144e-05,
"loss": 0.1569,
"step": 1895
},
{
"epoch": 11.778816199376948,
"grad_norm": 2.90625,
"learning_rate": 4.970021135479093e-05,
"loss": 0.2414,
"step": 1896
},
{
"epoch": 11.785046728971963,
"grad_norm": 3.0,
"learning_rate": 4.9699830830460324e-05,
"loss": 0.2723,
"step": 1897
},
{
"epoch": 11.791277258566979,
"grad_norm": 2.359375,
"learning_rate": 4.969945006624003e-05,
"loss": 0.1635,
"step": 1898
},
{
"epoch": 11.797507788161994,
"grad_norm": 2.765625,
"learning_rate": 4.969906906213373e-05,
"loss": 0.1922,
"step": 1899
},
{
"epoch": 11.80373831775701,
"grad_norm": 1.59375,
"learning_rate": 4.969868781814514e-05,
"loss": 0.1005,
"step": 1900
},
{
"epoch": 11.809968847352025,
"grad_norm": 2.765625,
"learning_rate": 4.9698306334277956e-05,
"loss": 0.2874,
"step": 1901
},
{
"epoch": 11.81619937694704,
"grad_norm": 1.9140625,
"learning_rate": 4.969792461053588e-05,
"loss": 0.2374,
"step": 1902
},
{
"epoch": 11.822429906542055,
"grad_norm": 1.9921875,
"learning_rate": 4.969754264692264e-05,
"loss": 0.1584,
"step": 1903
},
{
"epoch": 11.828660436137072,
"grad_norm": 2.375,
"learning_rate": 4.969716044344191e-05,
"loss": 0.1864,
"step": 1904
},
{
"epoch": 11.834890965732088,
"grad_norm": 2.328125,
"learning_rate": 4.9696778000097424e-05,
"loss": 0.2914,
"step": 1905
},
{
"epoch": 11.841121495327103,
"grad_norm": 2.4375,
"learning_rate": 4.9696395316892895e-05,
"loss": 0.19,
"step": 1906
},
{
"epoch": 11.847352024922118,
"grad_norm": 1.6484375,
"learning_rate": 4.9696012393832034e-05,
"loss": 0.1257,
"step": 1907
},
{
"epoch": 11.853582554517134,
"grad_norm": 2.578125,
"learning_rate": 4.9695629230918574e-05,
"loss": 0.1815,
"step": 1908
},
{
"epoch": 11.85981308411215,
"grad_norm": 1.78125,
"learning_rate": 4.969524582815622e-05,
"loss": 0.1471,
"step": 1909
},
{
"epoch": 11.866043613707165,
"grad_norm": 0.90234375,
"learning_rate": 4.969486218554871e-05,
"loss": 0.0855,
"step": 1910
},
{
"epoch": 11.87227414330218,
"grad_norm": 1.96875,
"learning_rate": 4.969447830309975e-05,
"loss": 0.1667,
"step": 1911
},
{
"epoch": 11.878504672897197,
"grad_norm": 3.078125,
"learning_rate": 4.969409418081309e-05,
"loss": 0.3011,
"step": 1912
},
{
"epoch": 11.884735202492212,
"grad_norm": 1.109375,
"learning_rate": 4.9693709818692444e-05,
"loss": 0.0922,
"step": 1913
},
{
"epoch": 11.890965732087228,
"grad_norm": 2.6875,
"learning_rate": 4.969332521674155e-05,
"loss": 0.2498,
"step": 1914
},
{
"epoch": 11.897196261682243,
"grad_norm": 2.375,
"learning_rate": 4.9692940374964155e-05,
"loss": 0.2104,
"step": 1915
},
{
"epoch": 11.903426791277258,
"grad_norm": 3.28125,
"learning_rate": 4.969255529336398e-05,
"loss": 0.2758,
"step": 1916
},
{
"epoch": 11.909657320872274,
"grad_norm": 1.46875,
"learning_rate": 4.9692169971944774e-05,
"loss": 0.1073,
"step": 1917
},
{
"epoch": 11.91588785046729,
"grad_norm": 3.359375,
"learning_rate": 4.969178441071028e-05,
"loss": 0.2595,
"step": 1918
},
{
"epoch": 11.922118380062305,
"grad_norm": 1.53125,
"learning_rate": 4.9691398609664233e-05,
"loss": 0.1194,
"step": 1919
},
{
"epoch": 11.928348909657322,
"grad_norm": 1.4765625,
"learning_rate": 4.969101256881039e-05,
"loss": 0.1236,
"step": 1920
},
{
"epoch": 11.934579439252337,
"grad_norm": 1.90625,
"learning_rate": 4.9690626288152504e-05,
"loss": 0.1435,
"step": 1921
},
{
"epoch": 11.940809968847352,
"grad_norm": 1.1171875,
"learning_rate": 4.969023976769431e-05,
"loss": 0.0951,
"step": 1922
},
{
"epoch": 11.947040498442368,
"grad_norm": 1.1171875,
"learning_rate": 4.9689853007439577e-05,
"loss": 0.0989,
"step": 1923
},
{
"epoch": 11.953271028037383,
"grad_norm": 2.96875,
"learning_rate": 4.968946600739206e-05,
"loss": 0.2474,
"step": 1924
},
{
"epoch": 11.959501557632398,
"grad_norm": 1.7734375,
"learning_rate": 4.9689078767555505e-05,
"loss": 0.1261,
"step": 1925
},
{
"epoch": 11.965732087227414,
"grad_norm": 1.78125,
"learning_rate": 4.968869128793369e-05,
"loss": 0.1464,
"step": 1926
},
{
"epoch": 11.97196261682243,
"grad_norm": 2.671875,
"learning_rate": 4.968830356853037e-05,
"loss": 0.3305,
"step": 1927
},
{
"epoch": 11.978193146417446,
"grad_norm": 2.171875,
"learning_rate": 4.9687915609349305e-05,
"loss": 0.1653,
"step": 1928
},
{
"epoch": 11.984423676012462,
"grad_norm": 2.203125,
"learning_rate": 4.9687527410394275e-05,
"loss": 0.1674,
"step": 1929
},
{
"epoch": 11.990654205607477,
"grad_norm": 2.4375,
"learning_rate": 4.968713897166903e-05,
"loss": 0.2533,
"step": 1930
},
{
"epoch": 11.996884735202492,
"grad_norm": 2.40625,
"learning_rate": 4.968675029317738e-05,
"loss": 0.1899,
"step": 1931
},
{
"epoch": 12.0,
"grad_norm": 1.203125,
"learning_rate": 4.968636137492305e-05,
"loss": 0.071,
"step": 1932
},
{
"epoch": 12.006230529595015,
"grad_norm": 1.8671875,
"learning_rate": 4.968597221690986e-05,
"loss": 0.1285,
"step": 1933
},
{
"epoch": 12.01246105919003,
"grad_norm": 2.25,
"learning_rate": 4.968558281914158e-05,
"loss": 0.1731,
"step": 1934
},
{
"epoch": 12.018691588785046,
"grad_norm": 1.6171875,
"learning_rate": 4.968519318162196e-05,
"loss": 0.1186,
"step": 1935
},
{
"epoch": 12.024922118380061,
"grad_norm": 3.203125,
"learning_rate": 4.968480330435483e-05,
"loss": 0.2754,
"step": 1936
},
{
"epoch": 12.031152647975079,
"grad_norm": 3.03125,
"learning_rate": 4.9684413187343946e-05,
"loss": 0.2454,
"step": 1937
},
{
"epoch": 12.037383177570094,
"grad_norm": 2.25,
"learning_rate": 4.9684022830593116e-05,
"loss": 0.1727,
"step": 1938
},
{
"epoch": 12.04361370716511,
"grad_norm": 1.078125,
"learning_rate": 4.968363223410612e-05,
"loss": 0.0814,
"step": 1939
},
{
"epoch": 12.049844236760125,
"grad_norm": 1.90625,
"learning_rate": 4.9683241397886745e-05,
"loss": 0.1505,
"step": 1940
},
{
"epoch": 12.05607476635514,
"grad_norm": 2.40625,
"learning_rate": 4.96828503219388e-05,
"loss": 0.1951,
"step": 1941
},
{
"epoch": 12.062305295950155,
"grad_norm": 2.15625,
"learning_rate": 4.968245900626608e-05,
"loss": 0.1806,
"step": 1942
},
{
"epoch": 12.06853582554517,
"grad_norm": 1.8515625,
"learning_rate": 4.9682067450872385e-05,
"loss": 0.1461,
"step": 1943
},
{
"epoch": 12.074766355140186,
"grad_norm": 1.2421875,
"learning_rate": 4.968167565576151e-05,
"loss": 0.0957,
"step": 1944
},
{
"epoch": 12.080996884735203,
"grad_norm": 1.6171875,
"learning_rate": 4.968128362093728e-05,
"loss": 0.1233,
"step": 1945
},
{
"epoch": 12.087227414330219,
"grad_norm": 1.3515625,
"learning_rate": 4.968089134640348e-05,
"loss": 0.1056,
"step": 1946
},
{
"epoch": 12.093457943925234,
"grad_norm": 2.8125,
"learning_rate": 4.968049883216393e-05,
"loss": 0.2707,
"step": 1947
},
{
"epoch": 12.09968847352025,
"grad_norm": 2.640625,
"learning_rate": 4.9680106078222444e-05,
"loss": 0.2636,
"step": 1948
},
{
"epoch": 12.105919003115265,
"grad_norm": 3.375,
"learning_rate": 4.967971308458283e-05,
"loss": 0.2248,
"step": 1949
},
{
"epoch": 12.11214953271028,
"grad_norm": 2.3125,
"learning_rate": 4.967931985124892e-05,
"loss": 0.1862,
"step": 1950
},
{
"epoch": 12.118380062305295,
"grad_norm": 2.265625,
"learning_rate": 4.967892637822451e-05,
"loss": 0.1725,
"step": 1951
},
{
"epoch": 12.12461059190031,
"grad_norm": 2.59375,
"learning_rate": 4.9678532665513444e-05,
"loss": 0.2522,
"step": 1952
},
{
"epoch": 12.130841121495328,
"grad_norm": 1.8828125,
"learning_rate": 4.9678138713119524e-05,
"loss": 0.1413,
"step": 1953
},
{
"epoch": 12.137071651090343,
"grad_norm": 3.03125,
"learning_rate": 4.96777445210466e-05,
"loss": 0.2449,
"step": 1954
},
{
"epoch": 12.143302180685358,
"grad_norm": 3.203125,
"learning_rate": 4.967735008929848e-05,
"loss": 0.2671,
"step": 1955
},
{
"epoch": 12.149532710280374,
"grad_norm": 1.78125,
"learning_rate": 4.967695541787901e-05,
"loss": 0.1358,
"step": 1956
},
{
"epoch": 12.15576323987539,
"grad_norm": 3.0625,
"learning_rate": 4.967656050679201e-05,
"loss": 0.26,
"step": 1957
},
{
"epoch": 12.161993769470405,
"grad_norm": 1.9140625,
"learning_rate": 4.967616535604133e-05,
"loss": 0.1636,
"step": 1958
},
{
"epoch": 12.16822429906542,
"grad_norm": 2.125,
"learning_rate": 4.967576996563079e-05,
"loss": 0.1557,
"step": 1959
},
{
"epoch": 12.174454828660435,
"grad_norm": 2.46875,
"learning_rate": 4.967537433556424e-05,
"loss": 0.206,
"step": 1960
},
{
"epoch": 12.180685358255452,
"grad_norm": 1.3359375,
"learning_rate": 4.967497846584552e-05,
"loss": 0.1071,
"step": 1961
},
{
"epoch": 12.186915887850468,
"grad_norm": 1.140625,
"learning_rate": 4.967458235647849e-05,
"loss": 0.0928,
"step": 1962
},
{
"epoch": 12.193146417445483,
"grad_norm": 2.796875,
"learning_rate": 4.967418600746697e-05,
"loss": 0.2263,
"step": 1963
},
{
"epoch": 12.199376947040498,
"grad_norm": 3.25,
"learning_rate": 4.967378941881483e-05,
"loss": 0.2713,
"step": 1964
},
{
"epoch": 12.205607476635514,
"grad_norm": 2.453125,
"learning_rate": 4.9673392590525915e-05,
"loss": 0.15,
"step": 1965
},
{
"epoch": 12.21183800623053,
"grad_norm": 2.765625,
"learning_rate": 4.9672995522604085e-05,
"loss": 0.3008,
"step": 1966
},
{
"epoch": 12.218068535825545,
"grad_norm": 1.578125,
"learning_rate": 4.967259821505318e-05,
"loss": 0.11,
"step": 1967
},
{
"epoch": 12.22429906542056,
"grad_norm": 2.40625,
"learning_rate": 4.9672200667877065e-05,
"loss": 0.1646,
"step": 1968
},
{
"epoch": 12.230529595015577,
"grad_norm": 2.578125,
"learning_rate": 4.967180288107961e-05,
"loss": 0.2224,
"step": 1969
},
{
"epoch": 12.236760124610592,
"grad_norm": 2.328125,
"learning_rate": 4.9671404854664685e-05,
"loss": 0.1958,
"step": 1970
},
{
"epoch": 12.242990654205608,
"grad_norm": 1.359375,
"learning_rate": 4.967100658863613e-05,
"loss": 0.1128,
"step": 1971
},
{
"epoch": 12.249221183800623,
"grad_norm": 2.359375,
"learning_rate": 4.967060808299784e-05,
"loss": 0.2898,
"step": 1972
},
{
"epoch": 12.255451713395638,
"grad_norm": 1.28125,
"learning_rate": 4.967020933775366e-05,
"loss": 0.1113,
"step": 1973
},
{
"epoch": 12.261682242990654,
"grad_norm": 3.265625,
"learning_rate": 4.966981035290748e-05,
"loss": 0.3731,
"step": 1974
},
{
"epoch": 12.26791277258567,
"grad_norm": 1.9765625,
"learning_rate": 4.966941112846317e-05,
"loss": 0.2185,
"step": 1975
},
{
"epoch": 12.274143302180685,
"grad_norm": 2.03125,
"learning_rate": 4.9669011664424616e-05,
"loss": 0.1424,
"step": 1976
},
{
"epoch": 12.280373831775702,
"grad_norm": 2.609375,
"learning_rate": 4.9668611960795683e-05,
"loss": 0.3395,
"step": 1977
},
{
"epoch": 12.286604361370717,
"grad_norm": 2.53125,
"learning_rate": 4.966821201758025e-05,
"loss": 0.1658,
"step": 1978
},
{
"epoch": 12.292834890965732,
"grad_norm": 1.2421875,
"learning_rate": 4.9667811834782224e-05,
"loss": 0.0983,
"step": 1979
},
{
"epoch": 12.299065420560748,
"grad_norm": 1.96875,
"learning_rate": 4.9667411412405477e-05,
"loss": 0.1618,
"step": 1980
},
{
"epoch": 12.305295950155763,
"grad_norm": 1.71875,
"learning_rate": 4.9667010750453896e-05,
"loss": 0.1476,
"step": 1981
},
{
"epoch": 12.311526479750778,
"grad_norm": 3.03125,
"learning_rate": 4.966660984893138e-05,
"loss": 0.3753,
"step": 1982
},
{
"epoch": 12.317757009345794,
"grad_norm": 3.25,
"learning_rate": 4.966620870784181e-05,
"loss": 0.3067,
"step": 1983
},
{
"epoch": 12.32398753894081,
"grad_norm": 2.171875,
"learning_rate": 4.9665807327189096e-05,
"loss": 0.1515,
"step": 1984
},
{
"epoch": 12.330218068535826,
"grad_norm": 1.046875,
"learning_rate": 4.966540570697713e-05,
"loss": 0.0807,
"step": 1985
},
{
"epoch": 12.336448598130842,
"grad_norm": 2.078125,
"learning_rate": 4.966500384720981e-05,
"loss": 0.1917,
"step": 1986
},
{
"epoch": 12.342679127725857,
"grad_norm": 1.0859375,
"learning_rate": 4.966460174789105e-05,
"loss": 0.0943,
"step": 1987
},
{
"epoch": 12.348909657320872,
"grad_norm": 2.265625,
"learning_rate": 4.9664199409024745e-05,
"loss": 0.1662,
"step": 1988
},
{
"epoch": 12.355140186915888,
"grad_norm": 1.7578125,
"learning_rate": 4.96637968306148e-05,
"loss": 0.1299,
"step": 1989
},
{
"epoch": 12.361370716510903,
"grad_norm": 2.453125,
"learning_rate": 4.9663394012665135e-05,
"loss": 0.1783,
"step": 1990
},
{
"epoch": 12.367601246105918,
"grad_norm": 2.515625,
"learning_rate": 4.9662990955179656e-05,
"loss": 0.1909,
"step": 1991
},
{
"epoch": 12.373831775700934,
"grad_norm": 2.265625,
"learning_rate": 4.966258765816227e-05,
"loss": 0.2347,
"step": 1992
},
{
"epoch": 12.38006230529595,
"grad_norm": 2.625,
"learning_rate": 4.966218412161692e-05,
"loss": 0.1937,
"step": 1993
},
{
"epoch": 12.386292834890966,
"grad_norm": 1.2734375,
"learning_rate": 4.9661780345547495e-05,
"loss": 0.0965,
"step": 1994
},
{
"epoch": 12.392523364485982,
"grad_norm": 2.53125,
"learning_rate": 4.966137632995793e-05,
"loss": 0.1904,
"step": 1995
},
{
"epoch": 12.398753894080997,
"grad_norm": 1.75,
"learning_rate": 4.9660972074852154e-05,
"loss": 0.1324,
"step": 1996
},
{
"epoch": 12.404984423676012,
"grad_norm": 1.0390625,
"learning_rate": 4.966056758023408e-05,
"loss": 0.101,
"step": 1997
},
{
"epoch": 12.411214953271028,
"grad_norm": 2.8125,
"learning_rate": 4.9660162846107654e-05,
"loss": 0.3173,
"step": 1998
},
{
"epoch": 12.417445482866043,
"grad_norm": 2.28125,
"learning_rate": 4.96597578724768e-05,
"loss": 0.2041,
"step": 1999
},
{
"epoch": 12.423676012461058,
"grad_norm": 2.046875,
"learning_rate": 4.965935265934544e-05,
"loss": 0.1545,
"step": 2000
},
{
"epoch": 12.429906542056075,
"grad_norm": 2.515625,
"learning_rate": 4.965894720671751e-05,
"loss": 0.261,
"step": 2001
},
{
"epoch": 12.43613707165109,
"grad_norm": 2.4375,
"learning_rate": 4.965854151459697e-05,
"loss": 0.1961,
"step": 2002
},
{
"epoch": 12.442367601246106,
"grad_norm": 2.734375,
"learning_rate": 4.965813558298773e-05,
"loss": 0.3065,
"step": 2003
},
{
"epoch": 12.448598130841122,
"grad_norm": 2.15625,
"learning_rate": 4.965772941189376e-05,
"loss": 0.1613,
"step": 2004
},
{
"epoch": 12.454828660436137,
"grad_norm": 1.3125,
"learning_rate": 4.965732300131899e-05,
"loss": 0.1152,
"step": 2005
},
{
"epoch": 12.461059190031152,
"grad_norm": 1.9765625,
"learning_rate": 4.9656916351267375e-05,
"loss": 0.2444,
"step": 2006
},
{
"epoch": 12.467289719626168,
"grad_norm": 1.609375,
"learning_rate": 4.965650946174285e-05,
"loss": 0.1182,
"step": 2007
},
{
"epoch": 12.473520249221183,
"grad_norm": 1.2265625,
"learning_rate": 4.965610233274939e-05,
"loss": 0.0953,
"step": 2008
},
{
"epoch": 12.4797507788162,
"grad_norm": 1.5,
"learning_rate": 4.965569496429092e-05,
"loss": 0.1148,
"step": 2009
},
{
"epoch": 12.485981308411215,
"grad_norm": 2.21875,
"learning_rate": 4.9655287356371416e-05,
"loss": 0.2934,
"step": 2010
},
{
"epoch": 12.49221183800623,
"grad_norm": 2.296875,
"learning_rate": 4.965487950899484e-05,
"loss": 0.1496,
"step": 2011
},
{
"epoch": 12.498442367601246,
"grad_norm": 1.890625,
"learning_rate": 4.9654471422165135e-05,
"loss": 0.1332,
"step": 2012
},
{
"epoch": 12.504672897196262,
"grad_norm": 1.34375,
"learning_rate": 4.9654063095886286e-05,
"loss": 0.1027,
"step": 2013
},
{
"epoch": 12.510903426791277,
"grad_norm": 2.8125,
"learning_rate": 4.965365453016224e-05,
"loss": 0.3397,
"step": 2014
},
{
"epoch": 12.517133956386292,
"grad_norm": 2.15625,
"learning_rate": 4.965324572499698e-05,
"loss": 0.166,
"step": 2015
},
{
"epoch": 12.523364485981308,
"grad_norm": 1.6640625,
"learning_rate": 4.965283668039447e-05,
"loss": 0.1177,
"step": 2016
},
{
"epoch": 12.529595015576325,
"grad_norm": 2.234375,
"learning_rate": 4.965242739635867e-05,
"loss": 0.1905,
"step": 2017
},
{
"epoch": 12.53582554517134,
"grad_norm": 1.3046875,
"learning_rate": 4.965201787289358e-05,
"loss": 0.1161,
"step": 2018
},
{
"epoch": 12.542056074766355,
"grad_norm": 1.96875,
"learning_rate": 4.965160811000316e-05,
"loss": 0.1643,
"step": 2019
},
{
"epoch": 12.54828660436137,
"grad_norm": 2.5,
"learning_rate": 4.96511981076914e-05,
"loss": 0.3163,
"step": 2020
},
{
"epoch": 12.554517133956386,
"grad_norm": 2.203125,
"learning_rate": 4.965078786596227e-05,
"loss": 0.1629,
"step": 2021
},
{
"epoch": 12.560747663551401,
"grad_norm": 1.2890625,
"learning_rate": 4.965037738481976e-05,
"loss": 0.0894,
"step": 2022
},
{
"epoch": 12.566978193146417,
"grad_norm": 1.15625,
"learning_rate": 4.9649966664267856e-05,
"loss": 0.0857,
"step": 2023
},
{
"epoch": 12.573208722741432,
"grad_norm": 1.3671875,
"learning_rate": 4.964955570431055e-05,
"loss": 0.1114,
"step": 2024
},
{
"epoch": 12.57943925233645,
"grad_norm": 1.828125,
"learning_rate": 4.964914450495183e-05,
"loss": 0.1416,
"step": 2025
},
{
"epoch": 12.585669781931465,
"grad_norm": 1.5703125,
"learning_rate": 4.9648733066195694e-05,
"loss": 0.1427,
"step": 2026
},
{
"epoch": 12.59190031152648,
"grad_norm": 1.7265625,
"learning_rate": 4.9648321388046137e-05,
"loss": 0.1332,
"step": 2027
},
{
"epoch": 12.598130841121495,
"grad_norm": 3.265625,
"learning_rate": 4.964790947050715e-05,
"loss": 0.2923,
"step": 2028
},
{
"epoch": 12.60436137071651,
"grad_norm": 2.1875,
"learning_rate": 4.9647497313582744e-05,
"loss": 0.1732,
"step": 2029
},
{
"epoch": 12.610591900311526,
"grad_norm": 2.484375,
"learning_rate": 4.964708491727692e-05,
"loss": 0.1688,
"step": 2030
},
{
"epoch": 12.616822429906541,
"grad_norm": 2.953125,
"learning_rate": 4.9646672281593674e-05,
"loss": 0.2158,
"step": 2031
},
{
"epoch": 12.623052959501557,
"grad_norm": 3.03125,
"learning_rate": 4.964625940653702e-05,
"loss": 0.2393,
"step": 2032
},
{
"epoch": 12.629283489096574,
"grad_norm": 1.734375,
"learning_rate": 4.964584629211098e-05,
"loss": 0.1239,
"step": 2033
},
{
"epoch": 12.63551401869159,
"grad_norm": 1.859375,
"learning_rate": 4.9645432938319536e-05,
"loss": 0.1423,
"step": 2034
},
{
"epoch": 12.641744548286605,
"grad_norm": 2.6875,
"learning_rate": 4.9645019345166734e-05,
"loss": 0.2177,
"step": 2035
},
{
"epoch": 12.64797507788162,
"grad_norm": 1.984375,
"learning_rate": 4.964460551265657e-05,
"loss": 0.181,
"step": 2036
},
{
"epoch": 12.654205607476635,
"grad_norm": 1.7734375,
"learning_rate": 4.9644191440793076e-05,
"loss": 0.1221,
"step": 2037
},
{
"epoch": 12.66043613707165,
"grad_norm": 1.65625,
"learning_rate": 4.9643777129580263e-05,
"loss": 0.1423,
"step": 2038
},
{
"epoch": 12.666666666666666,
"grad_norm": 1.9296875,
"learning_rate": 4.964336257902217e-05,
"loss": 0.1632,
"step": 2039
},
{
"epoch": 12.672897196261681,
"grad_norm": 1.5390625,
"learning_rate": 4.964294778912281e-05,
"loss": 0.1266,
"step": 2040
},
{
"epoch": 12.679127725856699,
"grad_norm": 2.078125,
"learning_rate": 4.9642532759886205e-05,
"loss": 0.1526,
"step": 2041
},
{
"epoch": 12.685358255451714,
"grad_norm": 2.65625,
"learning_rate": 4.96421174913164e-05,
"loss": 0.315,
"step": 2042
},
{
"epoch": 12.69158878504673,
"grad_norm": 1.6484375,
"learning_rate": 4.9641701983417434e-05,
"loss": 0.1251,
"step": 2043
},
{
"epoch": 12.697819314641745,
"grad_norm": 2.1875,
"learning_rate": 4.964128623619332e-05,
"loss": 0.2172,
"step": 2044
},
{
"epoch": 12.70404984423676,
"grad_norm": 2.421875,
"learning_rate": 4.964087024964812e-05,
"loss": 0.1798,
"step": 2045
},
{
"epoch": 12.710280373831775,
"grad_norm": 1.4453125,
"learning_rate": 4.964045402378585e-05,
"loss": 0.104,
"step": 2046
},
{
"epoch": 12.71651090342679,
"grad_norm": 2.21875,
"learning_rate": 4.964003755861057e-05,
"loss": 0.1685,
"step": 2047
},
{
"epoch": 12.722741433021806,
"grad_norm": 2.671875,
"learning_rate": 4.9639620854126326e-05,
"loss": 0.2008,
"step": 2048
},
{
"epoch": 12.728971962616823,
"grad_norm": 2.203125,
"learning_rate": 4.963920391033715e-05,
"loss": 0.1804,
"step": 2049
},
{
"epoch": 12.735202492211839,
"grad_norm": 2.359375,
"learning_rate": 4.963878672724711e-05,
"loss": 0.2025,
"step": 2050
},
{
"epoch": 12.741433021806854,
"grad_norm": 2.703125,
"learning_rate": 4.963836930486025e-05,
"loss": 0.1814,
"step": 2051
},
{
"epoch": 12.74766355140187,
"grad_norm": 2.0,
"learning_rate": 4.9637951643180605e-05,
"loss": 0.1513,
"step": 2052
},
{
"epoch": 12.753894080996885,
"grad_norm": 2.46875,
"learning_rate": 4.963753374221226e-05,
"loss": 0.1391,
"step": 2053
},
{
"epoch": 12.7601246105919,
"grad_norm": 2.28125,
"learning_rate": 4.963711560195926e-05,
"loss": 0.1667,
"step": 2054
},
{
"epoch": 12.766355140186915,
"grad_norm": 1.6484375,
"learning_rate": 4.963669722242567e-05,
"loss": 0.1139,
"step": 2055
},
{
"epoch": 12.77258566978193,
"grad_norm": 2.296875,
"learning_rate": 4.963627860361555e-05,
"loss": 0.1806,
"step": 2056
},
{
"epoch": 12.778816199376948,
"grad_norm": 1.5078125,
"learning_rate": 4.9635859745532974e-05,
"loss": 0.1127,
"step": 2057
},
{
"epoch": 12.785046728971963,
"grad_norm": 2.125,
"learning_rate": 4.9635440648181994e-05,
"loss": 0.1342,
"step": 2058
},
{
"epoch": 12.791277258566979,
"grad_norm": 2.078125,
"learning_rate": 4.9635021311566696e-05,
"loss": 0.1587,
"step": 2059
},
{
"epoch": 12.797507788161994,
"grad_norm": 2.359375,
"learning_rate": 4.963460173569114e-05,
"loss": 0.2998,
"step": 2060
},
{
"epoch": 12.80373831775701,
"grad_norm": 1.5078125,
"learning_rate": 4.963418192055942e-05,
"loss": 0.1202,
"step": 2061
},
{
"epoch": 12.809968847352025,
"grad_norm": 2.484375,
"learning_rate": 4.9633761866175586e-05,
"loss": 0.2342,
"step": 2062
},
{
"epoch": 12.81619937694704,
"grad_norm": 3.109375,
"learning_rate": 4.9633341572543745e-05,
"loss": 0.2401,
"step": 2063
},
{
"epoch": 12.822429906542055,
"grad_norm": 3.140625,
"learning_rate": 4.963292103966796e-05,
"loss": 0.1826,
"step": 2064
},
{
"epoch": 12.828660436137072,
"grad_norm": 1.1953125,
"learning_rate": 4.963250026755232e-05,
"loss": 0.0961,
"step": 2065
},
{
"epoch": 12.834890965732088,
"grad_norm": 1.875,
"learning_rate": 4.963207925620092e-05,
"loss": 0.2327,
"step": 2066
},
{
"epoch": 12.841121495327103,
"grad_norm": 1.7890625,
"learning_rate": 4.963165800561784e-05,
"loss": 0.1251,
"step": 2067
},
{
"epoch": 12.847352024922118,
"grad_norm": 3.453125,
"learning_rate": 4.963123651580717e-05,
"loss": 0.2411,
"step": 2068
},
{
"epoch": 12.853582554517134,
"grad_norm": 1.578125,
"learning_rate": 4.963081478677301e-05,
"loss": 0.1117,
"step": 2069
},
{
"epoch": 12.85981308411215,
"grad_norm": 3.046875,
"learning_rate": 4.9630392818519457e-05,
"loss": 0.2683,
"step": 2070
},
{
"epoch": 12.866043613707165,
"grad_norm": 2.109375,
"learning_rate": 4.96299706110506e-05,
"loss": 0.18,
"step": 2071
},
{
"epoch": 12.87227414330218,
"grad_norm": 2.21875,
"learning_rate": 4.9629548164370546e-05,
"loss": 0.1834,
"step": 2072
},
{
"epoch": 12.878504672897197,
"grad_norm": 2.78125,
"learning_rate": 4.962912547848339e-05,
"loss": 0.2004,
"step": 2073
},
{
"epoch": 12.884735202492212,
"grad_norm": 3.75,
"learning_rate": 4.9628702553393255e-05,
"loss": 0.3019,
"step": 2074
},
{
"epoch": 12.890965732087228,
"grad_norm": 2.1875,
"learning_rate": 4.962827938910424e-05,
"loss": 0.1473,
"step": 2075
},
{
"epoch": 12.897196261682243,
"grad_norm": 1.6015625,
"learning_rate": 4.962785598562044e-05,
"loss": 0.1212,
"step": 2076
},
{
"epoch": 12.903426791277258,
"grad_norm": 2.25,
"learning_rate": 4.962743234294599e-05,
"loss": 0.152,
"step": 2077
},
{
"epoch": 12.909657320872274,
"grad_norm": 2.5625,
"learning_rate": 4.9627008461084987e-05,
"loss": 0.1896,
"step": 2078
},
{
"epoch": 12.91588785046729,
"grad_norm": 3.0,
"learning_rate": 4.962658434004156e-05,
"loss": 0.2605,
"step": 2079
},
{
"epoch": 12.922118380062305,
"grad_norm": 1.796875,
"learning_rate": 4.9626159979819826e-05,
"loss": 0.2373,
"step": 2080
},
{
"epoch": 12.928348909657322,
"grad_norm": 3.296875,
"learning_rate": 4.96257353804239e-05,
"loss": 0.2523,
"step": 2081
},
{
"epoch": 12.934579439252337,
"grad_norm": 2.328125,
"learning_rate": 4.962531054185791e-05,
"loss": 0.1939,
"step": 2082
},
{
"epoch": 12.940809968847352,
"grad_norm": 2.03125,
"learning_rate": 4.962488546412598e-05,
"loss": 0.1306,
"step": 2083
},
{
"epoch": 12.947040498442368,
"grad_norm": 2.46875,
"learning_rate": 4.962446014723224e-05,
"loss": 0.2382,
"step": 2084
},
{
"epoch": 12.953271028037383,
"grad_norm": 1.375,
"learning_rate": 4.9624034591180815e-05,
"loss": 0.105,
"step": 2085
},
{
"epoch": 12.959501557632398,
"grad_norm": 2.546875,
"learning_rate": 4.962360879597585e-05,
"loss": 0.1863,
"step": 2086
},
{
"epoch": 12.965732087227414,
"grad_norm": 2.59375,
"learning_rate": 4.962318276162148e-05,
"loss": 0.1856,
"step": 2087
},
{
"epoch": 12.97196261682243,
"grad_norm": 2.234375,
"learning_rate": 4.9622756488121826e-05,
"loss": 0.1664,
"step": 2088
},
{
"epoch": 12.978193146417446,
"grad_norm": 1.1953125,
"learning_rate": 4.962232997548105e-05,
"loss": 0.0791,
"step": 2089
},
{
"epoch": 12.984423676012462,
"grad_norm": 3.046875,
"learning_rate": 4.962190322370327e-05,
"loss": 0.2587,
"step": 2090
},
{
"epoch": 12.990654205607477,
"grad_norm": 1.59375,
"learning_rate": 4.9621476232792654e-05,
"loss": 0.1095,
"step": 2091
},
{
"epoch": 12.996884735202492,
"grad_norm": 1.59375,
"learning_rate": 4.962104900275334e-05,
"loss": 0.1152,
"step": 2092
},
{
"epoch": 13.0,
"grad_norm": 1.828125,
"learning_rate": 4.962062153358947e-05,
"loss": 0.115,
"step": 2093
},
{
"epoch": 13.006230529595015,
"grad_norm": 2.6875,
"learning_rate": 4.962019382530521e-05,
"loss": 0.1679,
"step": 2094
},
{
"epoch": 13.01246105919003,
"grad_norm": 1.6015625,
"learning_rate": 4.9619765877904696e-05,
"loss": 0.1122,
"step": 2095
},
{
"epoch": 13.018691588785046,
"grad_norm": 2.484375,
"learning_rate": 4.9619337691392096e-05,
"loss": 0.1612,
"step": 2096
},
{
"epoch": 13.024922118380061,
"grad_norm": 2.015625,
"learning_rate": 4.9618909265771574e-05,
"loss": 0.1393,
"step": 2097
},
{
"epoch": 13.031152647975079,
"grad_norm": 2.8125,
"learning_rate": 4.961848060104728e-05,
"loss": 0.2185,
"step": 2098
},
{
"epoch": 13.037383177570094,
"grad_norm": 2.421875,
"learning_rate": 4.961805169722338e-05,
"loss": 0.2454,
"step": 2099
},
{
"epoch": 13.04361370716511,
"grad_norm": 2.140625,
"learning_rate": 4.9617622554304044e-05,
"loss": 0.1812,
"step": 2100
},
{
"epoch": 13.049844236760125,
"grad_norm": 1.7734375,
"learning_rate": 4.961719317229344e-05,
"loss": 0.1226,
"step": 2101
},
{
"epoch": 13.05607476635514,
"grad_norm": 1.6953125,
"learning_rate": 4.961676355119572e-05,
"loss": 0.1079,
"step": 2102
},
{
"epoch": 13.062305295950155,
"grad_norm": 2.25,
"learning_rate": 4.961633369101509e-05,
"loss": 0.1443,
"step": 2103
},
{
"epoch": 13.06853582554517,
"grad_norm": 1.8046875,
"learning_rate": 4.96159035917557e-05,
"loss": 0.1399,
"step": 2104
},
{
"epoch": 13.074766355140186,
"grad_norm": 2.171875,
"learning_rate": 4.9615473253421727e-05,
"loss": 0.235,
"step": 2105
},
{
"epoch": 13.080996884735203,
"grad_norm": 2.546875,
"learning_rate": 4.961504267601737e-05,
"loss": 0.1857,
"step": 2106
},
{
"epoch": 13.087227414330219,
"grad_norm": 1.3671875,
"learning_rate": 4.961461185954678e-05,
"loss": 0.0883,
"step": 2107
},
{
"epoch": 13.093457943925234,
"grad_norm": 1.5546875,
"learning_rate": 4.961418080401418e-05,
"loss": 0.1132,
"step": 2108
},
{
"epoch": 13.09968847352025,
"grad_norm": 2.515625,
"learning_rate": 4.961374950942372e-05,
"loss": 0.1926,
"step": 2109
},
{
"epoch": 13.105919003115265,
"grad_norm": 2.328125,
"learning_rate": 4.9613317975779605e-05,
"loss": 0.2349,
"step": 2110
},
{
"epoch": 13.11214953271028,
"grad_norm": 2.546875,
"learning_rate": 4.961288620308603e-05,
"loss": 0.1921,
"step": 2111
},
{
"epoch": 13.118380062305295,
"grad_norm": 1.890625,
"learning_rate": 4.961245419134719e-05,
"loss": 0.1321,
"step": 2112
},
{
"epoch": 13.12461059190031,
"grad_norm": 1.78125,
"learning_rate": 4.961202194056727e-05,
"loss": 0.1347,
"step": 2113
},
{
"epoch": 13.130841121495328,
"grad_norm": 3.921875,
"learning_rate": 4.9611589450750475e-05,
"loss": 0.352,
"step": 2114
},
{
"epoch": 13.137071651090343,
"grad_norm": 2.1875,
"learning_rate": 4.9611156721901e-05,
"loss": 0.1502,
"step": 2115
},
{
"epoch": 13.143302180685358,
"grad_norm": 2.484375,
"learning_rate": 4.961072375402305e-05,
"loss": 0.3023,
"step": 2116
},
{
"epoch": 13.149532710280374,
"grad_norm": 1.375,
"learning_rate": 4.9610290547120835e-05,
"loss": 0.1038,
"step": 2117
},
{
"epoch": 13.15576323987539,
"grad_norm": 3.09375,
"learning_rate": 4.960985710119855e-05,
"loss": 0.387,
"step": 2118
},
{
"epoch": 13.161993769470405,
"grad_norm": 1.3828125,
"learning_rate": 4.9609423416260426e-05,
"loss": 0.1019,
"step": 2119
},
{
"epoch": 13.16822429906542,
"grad_norm": 1.7421875,
"learning_rate": 4.960898949231066e-05,
"loss": 0.1297,
"step": 2120
},
{
"epoch": 13.174454828660435,
"grad_norm": 2.875,
"learning_rate": 4.960855532935347e-05,
"loss": 0.3049,
"step": 2121
},
{
"epoch": 13.180685358255452,
"grad_norm": 3.203125,
"learning_rate": 4.960812092739306e-05,
"loss": 0.3853,
"step": 2122
},
{
"epoch": 13.186915887850468,
"grad_norm": 1.7890625,
"learning_rate": 4.9607686286433676e-05,
"loss": 0.1216,
"step": 2123
},
{
"epoch": 13.193146417445483,
"grad_norm": 2.09375,
"learning_rate": 4.960725140647952e-05,
"loss": 0.2009,
"step": 2124
},
{
"epoch": 13.199376947040498,
"grad_norm": 4.0,
"learning_rate": 4.9606816287534805e-05,
"loss": 0.341,
"step": 2125
},
{
"epoch": 13.205607476635514,
"grad_norm": 2.46875,
"learning_rate": 4.9606380929603784e-05,
"loss": 0.1833,
"step": 2126
},
{
"epoch": 13.21183800623053,
"grad_norm": 1.5078125,
"learning_rate": 4.960594533269067e-05,
"loss": 0.1223,
"step": 2127
},
{
"epoch": 13.218068535825545,
"grad_norm": 1.515625,
"learning_rate": 4.960550949679969e-05,
"loss": 0.1182,
"step": 2128
},
{
"epoch": 13.22429906542056,
"grad_norm": 2.59375,
"learning_rate": 4.96050734219351e-05,
"loss": 0.2841,
"step": 2129
},
{
"epoch": 13.230529595015577,
"grad_norm": 3.40625,
"learning_rate": 4.960463710810111e-05,
"loss": 0.2849,
"step": 2130
},
{
"epoch": 13.236760124610592,
"grad_norm": 2.1875,
"learning_rate": 4.960420055530196e-05,
"loss": 0.1444,
"step": 2131
},
{
"epoch": 13.242990654205608,
"grad_norm": 2.21875,
"learning_rate": 4.96037637635419e-05,
"loss": 0.2168,
"step": 2132
},
{
"epoch": 13.249221183800623,
"grad_norm": 1.0703125,
"learning_rate": 4.9603326732825175e-05,
"loss": 0.0826,
"step": 2133
},
{
"epoch": 13.255451713395638,
"grad_norm": 3.859375,
"learning_rate": 4.960288946315601e-05,
"loss": 0.3266,
"step": 2134
},
{
"epoch": 13.261682242990654,
"grad_norm": 3.109375,
"learning_rate": 4.960245195453868e-05,
"loss": 0.3227,
"step": 2135
},
{
"epoch": 13.26791277258567,
"grad_norm": 1.859375,
"learning_rate": 4.96020142069774e-05,
"loss": 0.1374,
"step": 2136
},
{
"epoch": 13.274143302180685,
"grad_norm": 2.640625,
"learning_rate": 4.960157622047645e-05,
"loss": 0.1961,
"step": 2137
},
{
"epoch": 13.280373831775702,
"grad_norm": 2.515625,
"learning_rate": 4.960113799504008e-05,
"loss": 0.1846,
"step": 2138
},
{
"epoch": 13.286604361370717,
"grad_norm": 2.96875,
"learning_rate": 4.9600699530672536e-05,
"loss": 0.2135,
"step": 2139
},
{
"epoch": 13.292834890965732,
"grad_norm": 3.265625,
"learning_rate": 4.9600260827378074e-05,
"loss": 0.3625,
"step": 2140
},
{
"epoch": 13.299065420560748,
"grad_norm": 1.1640625,
"learning_rate": 4.9599821885160966e-05,
"loss": 0.1039,
"step": 2141
},
{
"epoch": 13.305295950155763,
"grad_norm": 1.7109375,
"learning_rate": 4.959938270402547e-05,
"loss": 0.1208,
"step": 2142
},
{
"epoch": 13.311526479750778,
"grad_norm": 1.859375,
"learning_rate": 4.9598943283975854e-05,
"loss": 0.1211,
"step": 2143
},
{
"epoch": 13.317757009345794,
"grad_norm": 1.890625,
"learning_rate": 4.959850362501638e-05,
"loss": 0.1403,
"step": 2144
},
{
"epoch": 13.32398753894081,
"grad_norm": 1.78125,
"learning_rate": 4.9598063727151325e-05,
"loss": 0.1476,
"step": 2145
},
{
"epoch": 13.330218068535826,
"grad_norm": 1.171875,
"learning_rate": 4.9597623590384956e-05,
"loss": 0.0902,
"step": 2146
},
{
"epoch": 13.336448598130842,
"grad_norm": 2.640625,
"learning_rate": 4.959718321472155e-05,
"loss": 0.2299,
"step": 2147
},
{
"epoch": 13.342679127725857,
"grad_norm": 2.296875,
"learning_rate": 4.959674260016538e-05,
"loss": 0.194,
"step": 2148
},
{
"epoch": 13.348909657320872,
"grad_norm": 1.9296875,
"learning_rate": 4.9596301746720734e-05,
"loss": 0.1634,
"step": 2149
},
{
"epoch": 13.355140186915888,
"grad_norm": 2.546875,
"learning_rate": 4.959586065439189e-05,
"loss": 0.1896,
"step": 2150
},
{
"epoch": 13.361370716510903,
"grad_norm": 1.3203125,
"learning_rate": 4.9595419323183125e-05,
"loss": 0.1025,
"step": 2151
},
{
"epoch": 13.367601246105918,
"grad_norm": 1.875,
"learning_rate": 4.9594977753098734e-05,
"loss": 0.1618,
"step": 2152
},
{
"epoch": 13.373831775700934,
"grad_norm": 1.625,
"learning_rate": 4.959453594414301e-05,
"loss": 0.1266,
"step": 2153
},
{
"epoch": 13.38006230529595,
"grad_norm": 2.375,
"learning_rate": 4.959409389632022e-05,
"loss": 0.1937,
"step": 2154
},
{
"epoch": 13.386292834890966,
"grad_norm": 1.28125,
"learning_rate": 4.9593651609634696e-05,
"loss": 0.1015,
"step": 2155
},
{
"epoch": 13.392523364485982,
"grad_norm": 1.59375,
"learning_rate": 4.9593209084090696e-05,
"loss": 0.1184,
"step": 2156
},
{
"epoch": 13.398753894080997,
"grad_norm": 1.359375,
"learning_rate": 4.959276631969253e-05,
"loss": 0.1061,
"step": 2157
},
{
"epoch": 13.404984423676012,
"grad_norm": 2.359375,
"learning_rate": 4.959232331644451e-05,
"loss": 0.1946,
"step": 2158
},
{
"epoch": 13.411214953271028,
"grad_norm": 1.9921875,
"learning_rate": 4.9591880074350926e-05,
"loss": 0.1477,
"step": 2159
},
{
"epoch": 13.417445482866043,
"grad_norm": 1.6796875,
"learning_rate": 4.959143659341609e-05,
"loss": 0.1138,
"step": 2160
},
{
"epoch": 13.423676012461058,
"grad_norm": 2.90625,
"learning_rate": 4.95909928736443e-05,
"loss": 0.3002,
"step": 2161
},
{
"epoch": 13.429906542056075,
"grad_norm": 1.421875,
"learning_rate": 4.959054891503988e-05,
"loss": 0.1127,
"step": 2162
},
{
"epoch": 13.43613707165109,
"grad_norm": 3.734375,
"learning_rate": 4.9590104717607135e-05,
"loss": 0.4809,
"step": 2163
},
{
"epoch": 13.442367601246106,
"grad_norm": 1.7734375,
"learning_rate": 4.958966028135037e-05,
"loss": 0.2498,
"step": 2164
},
{
"epoch": 13.448598130841122,
"grad_norm": 1.828125,
"learning_rate": 4.9589215606273905e-05,
"loss": 0.124,
"step": 2165
},
{
"epoch": 13.454828660436137,
"grad_norm": 1.34375,
"learning_rate": 4.958877069238208e-05,
"loss": 0.1094,
"step": 2166
},
{
"epoch": 13.461059190031152,
"grad_norm": 2.59375,
"learning_rate": 4.9588325539679185e-05,
"loss": 0.2169,
"step": 2167
},
{
"epoch": 13.467289719626168,
"grad_norm": 1.6484375,
"learning_rate": 4.9587880148169554e-05,
"loss": 0.1068,
"step": 2168
},
{
"epoch": 13.473520249221183,
"grad_norm": 1.3984375,
"learning_rate": 4.9587434517857525e-05,
"loss": 0.1003,
"step": 2169
},
{
"epoch": 13.4797507788162,
"grad_norm": 2.328125,
"learning_rate": 4.958698864874741e-05,
"loss": 0.2141,
"step": 2170
},
{
"epoch": 13.485981308411215,
"grad_norm": 2.6875,
"learning_rate": 4.958654254084355e-05,
"loss": 0.1754,
"step": 2171
},
{
"epoch": 13.49221183800623,
"grad_norm": 1.5390625,
"learning_rate": 4.958609619415028e-05,
"loss": 0.0904,
"step": 2172
},
{
"epoch": 13.498442367601246,
"grad_norm": 1.875,
"learning_rate": 4.958564960867192e-05,
"loss": 0.1232,
"step": 2173
},
{
"epoch": 13.504672897196262,
"grad_norm": 1.4375,
"learning_rate": 4.958520278441282e-05,
"loss": 0.0999,
"step": 2174
},
{
"epoch": 13.510903426791277,
"grad_norm": 1.6796875,
"learning_rate": 4.958475572137732e-05,
"loss": 0.1586,
"step": 2175
},
{
"epoch": 13.517133956386292,
"grad_norm": 1.2265625,
"learning_rate": 4.9584308419569755e-05,
"loss": 0.0995,
"step": 2176
},
{
"epoch": 13.523364485981308,
"grad_norm": 2.21875,
"learning_rate": 4.958386087899447e-05,
"loss": 0.1848,
"step": 2177
},
{
"epoch": 13.529595015576325,
"grad_norm": 1.703125,
"learning_rate": 4.958341309965582e-05,
"loss": 0.1239,
"step": 2178
},
{
"epoch": 13.53582554517134,
"grad_norm": 0.88671875,
"learning_rate": 4.958296508155814e-05,
"loss": 0.0753,
"step": 2179
},
{
"epoch": 13.542056074766355,
"grad_norm": 0.95703125,
"learning_rate": 4.9582516824705795e-05,
"loss": 0.0721,
"step": 2180
},
{
"epoch": 13.54828660436137,
"grad_norm": 1.234375,
"learning_rate": 4.9582068329103125e-05,
"loss": 0.1031,
"step": 2181
},
{
"epoch": 13.554517133956386,
"grad_norm": 2.4375,
"learning_rate": 4.95816195947545e-05,
"loss": 0.1995,
"step": 2182
},
{
"epoch": 13.560747663551401,
"grad_norm": 1.734375,
"learning_rate": 4.958117062166427e-05,
"loss": 0.1181,
"step": 2183
},
{
"epoch": 13.566978193146417,
"grad_norm": 1.6796875,
"learning_rate": 4.958072140983679e-05,
"loss": 0.1307,
"step": 2184
},
{
"epoch": 13.573208722741432,
"grad_norm": 1.5625,
"learning_rate": 4.9580271959276445e-05,
"loss": 0.1299,
"step": 2185
},
{
"epoch": 13.57943925233645,
"grad_norm": 1.328125,
"learning_rate": 4.9579822269987574e-05,
"loss": 0.0977,
"step": 2186
},
{
"epoch": 13.585669781931465,
"grad_norm": 3.15625,
"learning_rate": 4.957937234197456e-05,
"loss": 0.2871,
"step": 2187
},
{
"epoch": 13.59190031152648,
"grad_norm": 1.90625,
"learning_rate": 4.957892217524177e-05,
"loss": 0.123,
"step": 2188
},
{
"epoch": 13.598130841121495,
"grad_norm": 1.2265625,
"learning_rate": 4.957847176979357e-05,
"loss": 0.0895,
"step": 2189
},
{
"epoch": 13.60436137071651,
"grad_norm": 2.234375,
"learning_rate": 4.957802112563434e-05,
"loss": 0.1984,
"step": 2190
},
{
"epoch": 13.610591900311526,
"grad_norm": 2.359375,
"learning_rate": 4.957757024276846e-05,
"loss": 0.1557,
"step": 2191
},
{
"epoch": 13.616822429906541,
"grad_norm": 2.25,
"learning_rate": 4.957711912120031e-05,
"loss": 0.2248,
"step": 2192
},
{
"epoch": 13.623052959501557,
"grad_norm": 1.1171875,
"learning_rate": 4.9576667760934256e-05,
"loss": 0.0801,
"step": 2193
},
{
"epoch": 13.629283489096574,
"grad_norm": 2.109375,
"learning_rate": 4.957621616197469e-05,
"loss": 0.1884,
"step": 2194
},
{
"epoch": 13.63551401869159,
"grad_norm": 1.6484375,
"learning_rate": 4.9575764324326004e-05,
"loss": 0.1033,
"step": 2195
},
{
"epoch": 13.641744548286605,
"grad_norm": 3.46875,
"learning_rate": 4.9575312247992584e-05,
"loss": 0.2806,
"step": 2196
},
{
"epoch": 13.64797507788162,
"grad_norm": 1.1015625,
"learning_rate": 4.957485993297882e-05,
"loss": 0.1043,
"step": 2197
},
{
"epoch": 13.654205607476635,
"grad_norm": 2.171875,
"learning_rate": 4.95744073792891e-05,
"loss": 0.1607,
"step": 2198
},
{
"epoch": 13.66043613707165,
"grad_norm": 1.7890625,
"learning_rate": 4.957395458692782e-05,
"loss": 0.1128,
"step": 2199
},
{
"epoch": 13.666666666666666,
"grad_norm": 1.0078125,
"learning_rate": 4.9573501555899394e-05,
"loss": 0.0806,
"step": 2200
},
{
"epoch": 13.672897196261681,
"grad_norm": 1.640625,
"learning_rate": 4.95730482862082e-05,
"loss": 0.1113,
"step": 2201
},
{
"epoch": 13.679127725856699,
"grad_norm": 2.75,
"learning_rate": 4.9572594777858644e-05,
"loss": 0.3248,
"step": 2202
},
{
"epoch": 13.685358255451714,
"grad_norm": 2.171875,
"learning_rate": 4.957214103085514e-05,
"loss": 0.149,
"step": 2203
},
{
"epoch": 13.69158878504673,
"grad_norm": 2.90625,
"learning_rate": 4.957168704520209e-05,
"loss": 0.2427,
"step": 2204
},
{
"epoch": 13.697819314641745,
"grad_norm": 2.40625,
"learning_rate": 4.9571232820903903e-05,
"loss": 0.1843,
"step": 2205
},
{
"epoch": 13.70404984423676,
"grad_norm": 1.515625,
"learning_rate": 4.9570778357965e-05,
"loss": 0.105,
"step": 2206
},
{
"epoch": 13.710280373831775,
"grad_norm": 2.25,
"learning_rate": 4.9570323656389776e-05,
"loss": 0.2104,
"step": 2207
},
{
"epoch": 13.71651090342679,
"grad_norm": 2.421875,
"learning_rate": 4.956986871618266e-05,
"loss": 0.1988,
"step": 2208
},
{
"epoch": 13.722741433021806,
"grad_norm": 1.6875,
"learning_rate": 4.956941353734807e-05,
"loss": 0.125,
"step": 2209
},
{
"epoch": 13.728971962616823,
"grad_norm": 2.140625,
"learning_rate": 4.9568958119890417e-05,
"loss": 0.189,
"step": 2210
},
{
"epoch": 13.735202492211839,
"grad_norm": 1.71875,
"learning_rate": 4.956850246381414e-05,
"loss": 0.1327,
"step": 2211
},
{
"epoch": 13.741433021806854,
"grad_norm": 1.8515625,
"learning_rate": 4.956804656912365e-05,
"loss": 0.1685,
"step": 2212
},
{
"epoch": 13.74766355140187,
"grad_norm": 1.1484375,
"learning_rate": 4.9567590435823383e-05,
"loss": 0.0943,
"step": 2213
},
{
"epoch": 13.753894080996885,
"grad_norm": 4.21875,
"learning_rate": 4.956713406391777e-05,
"loss": 0.4163,
"step": 2214
},
{
"epoch": 13.7601246105919,
"grad_norm": 2.546875,
"learning_rate": 4.956667745341124e-05,
"loss": 0.2747,
"step": 2215
},
{
"epoch": 13.766355140186915,
"grad_norm": 2.390625,
"learning_rate": 4.9566220604308224e-05,
"loss": 0.171,
"step": 2216
},
{
"epoch": 13.77258566978193,
"grad_norm": 2.140625,
"learning_rate": 4.9565763516613165e-05,
"loss": 0.1366,
"step": 2217
},
{
"epoch": 13.778816199376948,
"grad_norm": 2.125,
"learning_rate": 4.956530619033049e-05,
"loss": 0.222,
"step": 2218
},
{
"epoch": 13.785046728971963,
"grad_norm": 2.140625,
"learning_rate": 4.956484862546467e-05,
"loss": 0.2334,
"step": 2219
},
{
"epoch": 13.791277258566979,
"grad_norm": 1.8046875,
"learning_rate": 4.9564390822020114e-05,
"loss": 0.1356,
"step": 2220
},
{
"epoch": 13.797507788161994,
"grad_norm": 1.5390625,
"learning_rate": 4.956393278000129e-05,
"loss": 0.1173,
"step": 2221
},
{
"epoch": 13.80373831775701,
"grad_norm": 1.6015625,
"learning_rate": 4.956347449941264e-05,
"loss": 0.1092,
"step": 2222
},
{
"epoch": 13.809968847352025,
"grad_norm": 3.203125,
"learning_rate": 4.956301598025861e-05,
"loss": 0.2296,
"step": 2223
},
{
"epoch": 13.81619937694704,
"grad_norm": 2.65625,
"learning_rate": 4.956255722254367e-05,
"loss": 0.2327,
"step": 2224
},
{
"epoch": 13.822429906542055,
"grad_norm": 2.265625,
"learning_rate": 4.956209822627226e-05,
"loss": 0.2213,
"step": 2225
},
{
"epoch": 13.828660436137072,
"grad_norm": 2.078125,
"learning_rate": 4.956163899144884e-05,
"loss": 0.1561,
"step": 2226
},
{
"epoch": 13.834890965732088,
"grad_norm": 2.4375,
"learning_rate": 4.9561179518077875e-05,
"loss": 0.1604,
"step": 2227
},
{
"epoch": 13.841121495327103,
"grad_norm": 2.015625,
"learning_rate": 4.956071980616382e-05,
"loss": 0.2162,
"step": 2228
},
{
"epoch": 13.847352024922118,
"grad_norm": 2.078125,
"learning_rate": 4.956025985571115e-05,
"loss": 0.1325,
"step": 2229
},
{
"epoch": 13.853582554517134,
"grad_norm": 2.1875,
"learning_rate": 4.955979966672432e-05,
"loss": 0.2623,
"step": 2230
},
{
"epoch": 13.85981308411215,
"grad_norm": 2.34375,
"learning_rate": 4.9559339239207815e-05,
"loss": 0.2084,
"step": 2231
},
{
"epoch": 13.866043613707165,
"grad_norm": 2.875,
"learning_rate": 4.955887857316609e-05,
"loss": 0.2823,
"step": 2232
},
{
"epoch": 13.87227414330218,
"grad_norm": 2.328125,
"learning_rate": 4.955841766860363e-05,
"loss": 0.1658,
"step": 2233
},
{
"epoch": 13.878504672897197,
"grad_norm": 2.171875,
"learning_rate": 4.9557956525524904e-05,
"loss": 0.137,
"step": 2234
},
{
"epoch": 13.884735202492212,
"grad_norm": 2.265625,
"learning_rate": 4.9557495143934405e-05,
"loss": 0.183,
"step": 2235
},
{
"epoch": 13.890965732087228,
"grad_norm": 1.4765625,
"learning_rate": 4.95570335238366e-05,
"loss": 0.099,
"step": 2236
},
{
"epoch": 13.897196261682243,
"grad_norm": 1.7109375,
"learning_rate": 4.955657166523597e-05,
"loss": 0.1268,
"step": 2237
},
{
"epoch": 13.903426791277258,
"grad_norm": 2.578125,
"learning_rate": 4.955610956813702e-05,
"loss": 0.2679,
"step": 2238
},
{
"epoch": 13.909657320872274,
"grad_norm": 1.96875,
"learning_rate": 4.9555647232544214e-05,
"loss": 0.1234,
"step": 2239
},
{
"epoch": 13.91588785046729,
"grad_norm": 1.3359375,
"learning_rate": 4.955518465846206e-05,
"loss": 0.1004,
"step": 2240
},
{
"epoch": 13.922118380062305,
"grad_norm": 1.7421875,
"learning_rate": 4.955472184589505e-05,
"loss": 0.1144,
"step": 2241
},
{
"epoch": 13.928348909657322,
"grad_norm": 1.921875,
"learning_rate": 4.955425879484766e-05,
"loss": 0.1389,
"step": 2242
},
{
"epoch": 13.934579439252337,
"grad_norm": 1.8046875,
"learning_rate": 4.9553795505324406e-05,
"loss": 0.1458,
"step": 2243
},
{
"epoch": 13.940809968847352,
"grad_norm": 1.703125,
"learning_rate": 4.955333197732978e-05,
"loss": 0.1422,
"step": 2244
},
{
"epoch": 13.947040498442368,
"grad_norm": 3.53125,
"learning_rate": 4.955286821086829e-05,
"loss": 0.361,
"step": 2245
},
{
"epoch": 13.953271028037383,
"grad_norm": 2.28125,
"learning_rate": 4.955240420594443e-05,
"loss": 0.1405,
"step": 2246
},
{
"epoch": 13.959501557632398,
"grad_norm": 2.15625,
"learning_rate": 4.9551939962562714e-05,
"loss": 0.1537,
"step": 2247
},
{
"epoch": 13.965732087227414,
"grad_norm": 3.09375,
"learning_rate": 4.955147548072765e-05,
"loss": 0.2403,
"step": 2248
},
{
"epoch": 13.97196261682243,
"grad_norm": 2.984375,
"learning_rate": 4.955101076044375e-05,
"loss": 0.2678,
"step": 2249
},
{
"epoch": 13.978193146417446,
"grad_norm": 1.171875,
"learning_rate": 4.955054580171553e-05,
"loss": 0.0793,
"step": 2250
},
{
"epoch": 13.984423676012462,
"grad_norm": 2.328125,
"learning_rate": 4.955008060454749e-05,
"loss": 0.2029,
"step": 2251
},
{
"epoch": 13.990654205607477,
"grad_norm": 2.9375,
"learning_rate": 4.954961516894417e-05,
"loss": 0.2074,
"step": 2252
},
{
"epoch": 13.996884735202492,
"grad_norm": 2.484375,
"learning_rate": 4.954914949491007e-05,
"loss": 0.1587,
"step": 2253
},
{
"epoch": 14.0,
"grad_norm": 1.171875,
"learning_rate": 4.9548683582449733e-05,
"loss": 0.0668,
"step": 2254
},
{
"epoch": 14.006230529595015,
"grad_norm": 2.4375,
"learning_rate": 4.9548217431567665e-05,
"loss": 0.1623,
"step": 2255
},
{
"epoch": 14.01246105919003,
"grad_norm": 3.140625,
"learning_rate": 4.954775104226841e-05,
"loss": 0.2489,
"step": 2256
},
{
"epoch": 14.018691588785046,
"grad_norm": 1.9609375,
"learning_rate": 4.9547284414556487e-05,
"loss": 0.1395,
"step": 2257
},
{
"epoch": 14.024922118380061,
"grad_norm": 1.34375,
"learning_rate": 4.954681754843643e-05,
"loss": 0.1017,
"step": 2258
},
{
"epoch": 14.031152647975079,
"grad_norm": 3.1875,
"learning_rate": 4.9546350443912783e-05,
"loss": 0.3952,
"step": 2259
},
{
"epoch": 14.037383177570094,
"grad_norm": 3.34375,
"learning_rate": 4.954588310099007e-05,
"loss": 0.4087,
"step": 2260
},
{
"epoch": 14.04361370716511,
"grad_norm": 2.359375,
"learning_rate": 4.954541551967282e-05,
"loss": 0.1689,
"step": 2261
},
{
"epoch": 14.049844236760125,
"grad_norm": 2.0625,
"learning_rate": 4.9544947699965594e-05,
"loss": 0.1435,
"step": 2262
},
{
"epoch": 14.05607476635514,
"grad_norm": 3.890625,
"learning_rate": 4.954447964187294e-05,
"loss": 0.2587,
"step": 2263
},
{
"epoch": 14.062305295950155,
"grad_norm": 1.2578125,
"learning_rate": 4.9544011345399384e-05,
"loss": 0.1013,
"step": 2264
},
{
"epoch": 14.06853582554517,
"grad_norm": 2.46875,
"learning_rate": 4.954354281054948e-05,
"loss": 0.1757,
"step": 2265
},
{
"epoch": 14.074766355140186,
"grad_norm": 2.421875,
"learning_rate": 4.95430740373278e-05,
"loss": 0.2627,
"step": 2266
},
{
"epoch": 14.080996884735203,
"grad_norm": 1.8671875,
"learning_rate": 4.954260502573886e-05,
"loss": 0.1441,
"step": 2267
},
{
"epoch": 14.087227414330219,
"grad_norm": 2.078125,
"learning_rate": 4.954213577578724e-05,
"loss": 0.118,
"step": 2268
},
{
"epoch": 14.093457943925234,
"grad_norm": 1.9765625,
"learning_rate": 4.9541666287477494e-05,
"loss": 0.1404,
"step": 2269
},
{
"epoch": 14.09968847352025,
"grad_norm": 2.625,
"learning_rate": 4.9541196560814173e-05,
"loss": 0.229,
"step": 2270
},
{
"epoch": 14.105919003115265,
"grad_norm": 1.890625,
"learning_rate": 4.954072659580185e-05,
"loss": 0.1985,
"step": 2271
},
{
"epoch": 14.11214953271028,
"grad_norm": 1.46875,
"learning_rate": 4.9540256392445086e-05,
"loss": 0.0883,
"step": 2272
},
{
"epoch": 14.118380062305295,
"grad_norm": 1.5546875,
"learning_rate": 4.953978595074845e-05,
"loss": 0.1026,
"step": 2273
},
{
"epoch": 14.12461059190031,
"grad_norm": 2.625,
"learning_rate": 4.95393152707165e-05,
"loss": 0.246,
"step": 2274
},
{
"epoch": 14.130841121495328,
"grad_norm": 2.6875,
"learning_rate": 4.953884435235382e-05,
"loss": 0.2968,
"step": 2275
},
{
"epoch": 14.137071651090343,
"grad_norm": 2.578125,
"learning_rate": 4.953837319566497e-05,
"loss": 0.1816,
"step": 2276
},
{
"epoch": 14.143302180685358,
"grad_norm": 3.03125,
"learning_rate": 4.953790180065454e-05,
"loss": 0.2489,
"step": 2277
},
{
"epoch": 14.149532710280374,
"grad_norm": 2.46875,
"learning_rate": 4.95374301673271e-05,
"loss": 0.2565,
"step": 2278
},
{
"epoch": 14.15576323987539,
"grad_norm": 3.453125,
"learning_rate": 4.9536958295687244e-05,
"loss": 0.3585,
"step": 2279
},
{
"epoch": 14.161993769470405,
"grad_norm": 1.5546875,
"learning_rate": 4.953648618573953e-05,
"loss": 0.098,
"step": 2280
},
{
"epoch": 14.16822429906542,
"grad_norm": 2.125,
"learning_rate": 4.953601383748857e-05,
"loss": 0.1662,
"step": 2281
},
{
"epoch": 14.174454828660435,
"grad_norm": 1.625,
"learning_rate": 4.953554125093893e-05,
"loss": 0.1042,
"step": 2282
},
{
"epoch": 14.180685358255452,
"grad_norm": 1.4453125,
"learning_rate": 4.953506842609521e-05,
"loss": 0.0843,
"step": 2283
},
{
"epoch": 14.186915887850468,
"grad_norm": 2.09375,
"learning_rate": 4.9534595362962e-05,
"loss": 0.1144,
"step": 2284
},
{
"epoch": 14.193146417445483,
"grad_norm": 2.234375,
"learning_rate": 4.95341220615439e-05,
"loss": 0.1528,
"step": 2285
},
{
"epoch": 14.199376947040498,
"grad_norm": 2.84375,
"learning_rate": 4.95336485218455e-05,
"loss": 0.2706,
"step": 2286
},
{
"epoch": 14.205607476635514,
"grad_norm": 2.203125,
"learning_rate": 4.95331747438714e-05,
"loss": 0.1558,
"step": 2287
},
{
"epoch": 14.21183800623053,
"grad_norm": 3.3125,
"learning_rate": 4.9532700727626205e-05,
"loss": 0.2397,
"step": 2288
},
{
"epoch": 14.218068535825545,
"grad_norm": 2.40625,
"learning_rate": 4.953222647311452e-05,
"loss": 0.1892,
"step": 2289
},
{
"epoch": 14.22429906542056,
"grad_norm": 1.765625,
"learning_rate": 4.953175198034094e-05,
"loss": 0.1214,
"step": 2290
},
{
"epoch": 14.230529595015577,
"grad_norm": 2.421875,
"learning_rate": 4.953127724931008e-05,
"loss": 0.1614,
"step": 2291
},
{
"epoch": 14.236760124610592,
"grad_norm": 3.390625,
"learning_rate": 4.953080228002656e-05,
"loss": 0.3075,
"step": 2292
},
{
"epoch": 14.242990654205608,
"grad_norm": 1.5,
"learning_rate": 4.9530327072494985e-05,
"loss": 0.1058,
"step": 2293
},
{
"epoch": 14.249221183800623,
"grad_norm": 2.46875,
"learning_rate": 4.952985162671996e-05,
"loss": 0.2046,
"step": 2294
},
{
"epoch": 14.255451713395638,
"grad_norm": 2.015625,
"learning_rate": 4.9529375942706124e-05,
"loss": 0.1407,
"step": 2295
},
{
"epoch": 14.261682242990654,
"grad_norm": 1.4296875,
"learning_rate": 4.952890002045808e-05,
"loss": 0.1016,
"step": 2296
},
{
"epoch": 14.26791277258567,
"grad_norm": 2.796875,
"learning_rate": 4.952842385998046e-05,
"loss": 0.3037,
"step": 2297
},
{
"epoch": 14.274143302180685,
"grad_norm": 1.40625,
"learning_rate": 4.952794746127787e-05,
"loss": 0.1105,
"step": 2298
},
{
"epoch": 14.280373831775702,
"grad_norm": 2.1875,
"learning_rate": 4.952747082435496e-05,
"loss": 0.1667,
"step": 2299
},
{
"epoch": 14.286604361370717,
"grad_norm": 2.21875,
"learning_rate": 4.9526993949216355e-05,
"loss": 0.1938,
"step": 2300
},
{
"epoch": 14.292834890965732,
"grad_norm": 2.609375,
"learning_rate": 4.952651683586668e-05,
"loss": 0.233,
"step": 2301
},
{
"epoch": 14.299065420560748,
"grad_norm": 2.28125,
"learning_rate": 4.952603948431057e-05,
"loss": 0.1874,
"step": 2302
},
{
"epoch": 14.305295950155763,
"grad_norm": 2.71875,
"learning_rate": 4.952556189455266e-05,
"loss": 0.1896,
"step": 2303
},
{
"epoch": 14.311526479750778,
"grad_norm": 3.203125,
"learning_rate": 4.9525084066597594e-05,
"loss": 0.3146,
"step": 2304
},
{
"epoch": 14.317757009345794,
"grad_norm": 1.828125,
"learning_rate": 4.9524606000450005e-05,
"loss": 0.1348,
"step": 2305
},
{
"epoch": 14.32398753894081,
"grad_norm": 1.8125,
"learning_rate": 4.952412769611454e-05,
"loss": 0.1153,
"step": 2306
},
{
"epoch": 14.330218068535826,
"grad_norm": 1.7109375,
"learning_rate": 4.952364915359584e-05,
"loss": 0.1408,
"step": 2307
},
{
"epoch": 14.336448598130842,
"grad_norm": 1.1875,
"learning_rate": 4.952317037289857e-05,
"loss": 0.1055,
"step": 2308
},
{
"epoch": 14.342679127725857,
"grad_norm": 2.109375,
"learning_rate": 4.952269135402736e-05,
"loss": 0.1572,
"step": 2309
},
{
"epoch": 14.348909657320872,
"grad_norm": 2.890625,
"learning_rate": 4.952221209698687e-05,
"loss": 0.2005,
"step": 2310
},
{
"epoch": 14.355140186915888,
"grad_norm": 2.25,
"learning_rate": 4.952173260178176e-05,
"loss": 0.2004,
"step": 2311
},
{
"epoch": 14.361370716510903,
"grad_norm": 2.75,
"learning_rate": 4.9521252868416665e-05,
"loss": 0.2251,
"step": 2312
},
{
"epoch": 14.367601246105918,
"grad_norm": 1.0546875,
"learning_rate": 4.952077289689627e-05,
"loss": 0.0754,
"step": 2313
},
{
"epoch": 14.373831775700934,
"grad_norm": 1.609375,
"learning_rate": 4.952029268722523e-05,
"loss": 0.1176,
"step": 2314
},
{
"epoch": 14.38006230529595,
"grad_norm": 1.1171875,
"learning_rate": 4.951981223940821e-05,
"loss": 0.0916,
"step": 2315
},
{
"epoch": 14.386292834890966,
"grad_norm": 1.609375,
"learning_rate": 4.951933155344987e-05,
"loss": 0.1912,
"step": 2316
},
{
"epoch": 14.392523364485982,
"grad_norm": 2.109375,
"learning_rate": 4.951885062935487e-05,
"loss": 0.1519,
"step": 2317
},
{
"epoch": 14.398753894080997,
"grad_norm": 1.609375,
"learning_rate": 4.95183694671279e-05,
"loss": 0.1079,
"step": 2318
},
{
"epoch": 14.404984423676012,
"grad_norm": 1.1484375,
"learning_rate": 4.951788806677362e-05,
"loss": 0.0837,
"step": 2319
},
{
"epoch": 14.411214953271028,
"grad_norm": 2.078125,
"learning_rate": 4.951740642829671e-05,
"loss": 0.1761,
"step": 2320
},
{
"epoch": 14.417445482866043,
"grad_norm": 1.9609375,
"learning_rate": 4.951692455170185e-05,
"loss": 0.1441,
"step": 2321
},
{
"epoch": 14.423676012461058,
"grad_norm": 3.109375,
"learning_rate": 4.9516442436993724e-05,
"loss": 0.4294,
"step": 2322
},
{
"epoch": 14.429906542056075,
"grad_norm": 3.25,
"learning_rate": 4.9515960084177005e-05,
"loss": 0.2206,
"step": 2323
},
{
"epoch": 14.43613707165109,
"grad_norm": 2.28125,
"learning_rate": 4.951547749325638e-05,
"loss": 0.1432,
"step": 2324
},
{
"epoch": 14.442367601246106,
"grad_norm": 2.0625,
"learning_rate": 4.9514994664236535e-05,
"loss": 0.2368,
"step": 2325
},
{
"epoch": 14.448598130841122,
"grad_norm": 2.234375,
"learning_rate": 4.951451159712216e-05,
"loss": 0.1537,
"step": 2326
},
{
"epoch": 14.454828660436137,
"grad_norm": 2.453125,
"learning_rate": 4.951402829191795e-05,
"loss": 0.1773,
"step": 2327
},
{
"epoch": 14.461059190031152,
"grad_norm": 2.109375,
"learning_rate": 4.9513544748628596e-05,
"loss": 0.1453,
"step": 2328
},
{
"epoch": 14.467289719626168,
"grad_norm": 1.9921875,
"learning_rate": 4.9513060967258794e-05,
"loss": 0.1226,
"step": 2329
},
{
"epoch": 14.473520249221183,
"grad_norm": 3.046875,
"learning_rate": 4.951257694781325e-05,
"loss": 0.188,
"step": 2330
},
{
"epoch": 14.4797507788162,
"grad_norm": 1.5703125,
"learning_rate": 4.951209269029665e-05,
"loss": 0.1101,
"step": 2331
},
{
"epoch": 14.485981308411215,
"grad_norm": 1.734375,
"learning_rate": 4.951160819471372e-05,
"loss": 0.1174,
"step": 2332
},
{
"epoch": 14.49221183800623,
"grad_norm": 2.921875,
"learning_rate": 4.951112346106914e-05,
"loss": 0.2202,
"step": 2333
},
{
"epoch": 14.498442367601246,
"grad_norm": 2.453125,
"learning_rate": 4.951063848936763e-05,
"loss": 0.1959,
"step": 2334
},
{
"epoch": 14.504672897196262,
"grad_norm": 2.84375,
"learning_rate": 4.95101532796139e-05,
"loss": 0.2395,
"step": 2335
},
{
"epoch": 14.510903426791277,
"grad_norm": 1.796875,
"learning_rate": 4.950966783181267e-05,
"loss": 0.1354,
"step": 2336
},
{
"epoch": 14.517133956386292,
"grad_norm": 2.109375,
"learning_rate": 4.950918214596863e-05,
"loss": 0.1376,
"step": 2337
},
{
"epoch": 14.523364485981308,
"grad_norm": 2.03125,
"learning_rate": 4.950869622208653e-05,
"loss": 0.1354,
"step": 2338
},
{
"epoch": 14.529595015576325,
"grad_norm": 1.90625,
"learning_rate": 4.950821006017107e-05,
"loss": 0.1661,
"step": 2339
},
{
"epoch": 14.53582554517134,
"grad_norm": 2.09375,
"learning_rate": 4.950772366022697e-05,
"loss": 0.1398,
"step": 2340
},
{
"epoch": 14.542056074766355,
"grad_norm": 1.0078125,
"learning_rate": 4.9507237022258956e-05,
"loss": 0.0732,
"step": 2341
},
{
"epoch": 14.54828660436137,
"grad_norm": 2.46875,
"learning_rate": 4.950675014627176e-05,
"loss": 0.2891,
"step": 2342
},
{
"epoch": 14.554517133956386,
"grad_norm": 1.6171875,
"learning_rate": 4.950626303227011e-05,
"loss": 0.115,
"step": 2343
},
{
"epoch": 14.560747663551401,
"grad_norm": 2.296875,
"learning_rate": 4.9505775680258734e-05,
"loss": 0.2081,
"step": 2344
},
{
"epoch": 14.566978193146417,
"grad_norm": 2.328125,
"learning_rate": 4.9505288090242364e-05,
"loss": 0.1588,
"step": 2345
},
{
"epoch": 14.573208722741432,
"grad_norm": 2.25,
"learning_rate": 4.950480026222574e-05,
"loss": 0.1862,
"step": 2346
},
{
"epoch": 14.57943925233645,
"grad_norm": 1.6875,
"learning_rate": 4.9504312196213596e-05,
"loss": 0.1291,
"step": 2347
},
{
"epoch": 14.585669781931465,
"grad_norm": 1.6796875,
"learning_rate": 4.950382389221068e-05,
"loss": 0.1083,
"step": 2348
},
{
"epoch": 14.59190031152648,
"grad_norm": 2.9375,
"learning_rate": 4.9503335350221717e-05,
"loss": 0.2017,
"step": 2349
},
{
"epoch": 14.598130841121495,
"grad_norm": 2.890625,
"learning_rate": 4.950284657025147e-05,
"loss": 0.2735,
"step": 2350
},
{
"epoch": 14.60436137071651,
"grad_norm": 1.296875,
"learning_rate": 4.950235755230468e-05,
"loss": 0.0889,
"step": 2351
},
{
"epoch": 14.610591900311526,
"grad_norm": 1.78125,
"learning_rate": 4.950186829638609e-05,
"loss": 0.2318,
"step": 2352
},
{
"epoch": 14.616822429906541,
"grad_norm": 1.015625,
"learning_rate": 4.950137880250046e-05,
"loss": 0.0792,
"step": 2353
},
{
"epoch": 14.623052959501557,
"grad_norm": 2.609375,
"learning_rate": 4.9500889070652545e-05,
"loss": 0.2732,
"step": 2354
},
{
"epoch": 14.629283489096574,
"grad_norm": 1.4296875,
"learning_rate": 4.9500399100847094e-05,
"loss": 0.095,
"step": 2355
},
{
"epoch": 14.63551401869159,
"grad_norm": 1.8828125,
"learning_rate": 4.949990889308887e-05,
"loss": 0.1278,
"step": 2356
},
{
"epoch": 14.641744548286605,
"grad_norm": 3.09375,
"learning_rate": 4.949941844738263e-05,
"loss": 0.2483,
"step": 2357
},
{
"epoch": 14.64797507788162,
"grad_norm": 1.546875,
"learning_rate": 4.949892776373315e-05,
"loss": 0.0991,
"step": 2358
},
{
"epoch": 14.654205607476635,
"grad_norm": 2.1875,
"learning_rate": 4.949843684214518e-05,
"loss": 0.1755,
"step": 2359
},
{
"epoch": 14.66043613707165,
"grad_norm": 1.890625,
"learning_rate": 4.9497945682623495e-05,
"loss": 0.1838,
"step": 2360
},
{
"epoch": 14.666666666666666,
"grad_norm": 2.140625,
"learning_rate": 4.9497454285172865e-05,
"loss": 0.1312,
"step": 2361
},
{
"epoch": 14.672897196261681,
"grad_norm": 3.375,
"learning_rate": 4.949696264979806e-05,
"loss": 0.2611,
"step": 2362
},
{
"epoch": 14.679127725856699,
"grad_norm": 1.3828125,
"learning_rate": 4.9496470776503855e-05,
"loss": 0.0988,
"step": 2363
},
{
"epoch": 14.685358255451714,
"grad_norm": 2.21875,
"learning_rate": 4.949597866529503e-05,
"loss": 0.1741,
"step": 2364
},
{
"epoch": 14.69158878504673,
"grad_norm": 2.21875,
"learning_rate": 4.949548631617637e-05,
"loss": 0.1745,
"step": 2365
},
{
"epoch": 14.697819314641745,
"grad_norm": 2.09375,
"learning_rate": 4.949499372915264e-05,
"loss": 0.1379,
"step": 2366
},
{
"epoch": 14.70404984423676,
"grad_norm": 2.265625,
"learning_rate": 4.949450090422865e-05,
"loss": 0.133,
"step": 2367
},
{
"epoch": 14.710280373831775,
"grad_norm": 2.453125,
"learning_rate": 4.949400784140915e-05,
"loss": 0.1887,
"step": 2368
},
{
"epoch": 14.71651090342679,
"grad_norm": 0.92578125,
"learning_rate": 4.9493514540698963e-05,
"loss": 0.0758,
"step": 2369
},
{
"epoch": 14.722741433021806,
"grad_norm": 3.859375,
"learning_rate": 4.949302100210287e-05,
"loss": 0.5057,
"step": 2370
},
{
"epoch": 14.728971962616823,
"grad_norm": 2.625,
"learning_rate": 4.949252722562565e-05,
"loss": 0.1903,
"step": 2371
},
{
"epoch": 14.735202492211839,
"grad_norm": 1.28125,
"learning_rate": 4.949203321127212e-05,
"loss": 0.0895,
"step": 2372
},
{
"epoch": 14.741433021806854,
"grad_norm": 2.96875,
"learning_rate": 4.9491538959047053e-05,
"loss": 0.1946,
"step": 2373
},
{
"epoch": 14.74766355140187,
"grad_norm": 2.484375,
"learning_rate": 4.949104446895528e-05,
"loss": 0.2978,
"step": 2374
},
{
"epoch": 14.753894080996885,
"grad_norm": 1.234375,
"learning_rate": 4.949054974100158e-05,
"loss": 0.0981,
"step": 2375
},
{
"epoch": 14.7601246105919,
"grad_norm": 1.9921875,
"learning_rate": 4.949005477519077e-05,
"loss": 0.136,
"step": 2376
},
{
"epoch": 14.766355140186915,
"grad_norm": 1.5,
"learning_rate": 4.9489559571527644e-05,
"loss": 0.1013,
"step": 2377
},
{
"epoch": 14.77258566978193,
"grad_norm": 1.6015625,
"learning_rate": 4.9489064130017025e-05,
"loss": 0.1048,
"step": 2378
},
{
"epoch": 14.778816199376948,
"grad_norm": 2.640625,
"learning_rate": 4.948856845066372e-05,
"loss": 0.1993,
"step": 2379
},
{
"epoch": 14.785046728971963,
"grad_norm": 2.375,
"learning_rate": 4.9488072533472546e-05,
"loss": 0.1703,
"step": 2380
},
{
"epoch": 14.791277258566979,
"grad_norm": 1.8125,
"learning_rate": 4.948757637844831e-05,
"loss": 0.1169,
"step": 2381
},
{
"epoch": 14.797507788161994,
"grad_norm": 1.21875,
"learning_rate": 4.948707998559584e-05,
"loss": 0.0839,
"step": 2382
},
{
"epoch": 14.80373831775701,
"grad_norm": 2.703125,
"learning_rate": 4.948658335491996e-05,
"loss": 0.1837,
"step": 2383
},
{
"epoch": 14.809968847352025,
"grad_norm": 2.28125,
"learning_rate": 4.9486086486425484e-05,
"loss": 0.1983,
"step": 2384
},
{
"epoch": 14.81619937694704,
"grad_norm": 2.453125,
"learning_rate": 4.948558938011724e-05,
"loss": 0.2198,
"step": 2385
},
{
"epoch": 14.822429906542055,
"grad_norm": 2.671875,
"learning_rate": 4.9485092036000055e-05,
"loss": 0.1923,
"step": 2386
},
{
"epoch": 14.828660436137072,
"grad_norm": 2.421875,
"learning_rate": 4.9484594454078765e-05,
"loss": 0.1707,
"step": 2387
},
{
"epoch": 14.834890965732088,
"grad_norm": 2.359375,
"learning_rate": 4.94840966343582e-05,
"loss": 0.2324,
"step": 2388
},
{
"epoch": 14.841121495327103,
"grad_norm": 2.109375,
"learning_rate": 4.9483598576843196e-05,
"loss": 0.17,
"step": 2389
},
{
"epoch": 14.847352024922118,
"grad_norm": 2.0625,
"learning_rate": 4.9483100281538573e-05,
"loss": 0.1374,
"step": 2390
},
{
"epoch": 14.853582554517134,
"grad_norm": 1.8828125,
"learning_rate": 4.9482601748449204e-05,
"loss": 0.1407,
"step": 2391
},
{
"epoch": 14.85981308411215,
"grad_norm": 2.875,
"learning_rate": 4.9482102977579905e-05,
"loss": 0.2734,
"step": 2392
},
{
"epoch": 14.866043613707165,
"grad_norm": 3.71875,
"learning_rate": 4.948160396893553e-05,
"loss": 0.4588,
"step": 2393
},
{
"epoch": 14.87227414330218,
"grad_norm": 2.265625,
"learning_rate": 4.948110472252092e-05,
"loss": 0.1357,
"step": 2394
},
{
"epoch": 14.878504672897197,
"grad_norm": 2.046875,
"learning_rate": 4.948060523834094e-05,
"loss": 0.1343,
"step": 2395
},
{
"epoch": 14.884735202492212,
"grad_norm": 1.7578125,
"learning_rate": 4.948010551640041e-05,
"loss": 0.1993,
"step": 2396
},
{
"epoch": 14.890965732087228,
"grad_norm": 3.0625,
"learning_rate": 4.947960555670421e-05,
"loss": 0.2678,
"step": 2397
},
{
"epoch": 14.897196261682243,
"grad_norm": 1.765625,
"learning_rate": 4.947910535925719e-05,
"loss": 0.1344,
"step": 2398
},
{
"epoch": 14.903426791277258,
"grad_norm": 1.890625,
"learning_rate": 4.947860492406421e-05,
"loss": 0.1364,
"step": 2399
},
{
"epoch": 14.909657320872274,
"grad_norm": 2.28125,
"learning_rate": 4.947810425113011e-05,
"loss": 0.1274,
"step": 2400
},
{
"epoch": 14.91588785046729,
"grad_norm": 2.046875,
"learning_rate": 4.947760334045978e-05,
"loss": 0.151,
"step": 2401
},
{
"epoch": 14.922118380062305,
"grad_norm": 2.53125,
"learning_rate": 4.947710219205808e-05,
"loss": 0.168,
"step": 2402
},
{
"epoch": 14.928348909657322,
"grad_norm": 2.921875,
"learning_rate": 4.947660080592986e-05,
"loss": 0.2284,
"step": 2403
},
{
"epoch": 14.934579439252337,
"grad_norm": 2.21875,
"learning_rate": 4.947609918208e-05,
"loss": 0.1551,
"step": 2404
},
{
"epoch": 14.940809968847352,
"grad_norm": 2.09375,
"learning_rate": 4.9475597320513374e-05,
"loss": 0.2174,
"step": 2405
},
{
"epoch": 14.947040498442368,
"grad_norm": 1.34375,
"learning_rate": 4.947509522123485e-05,
"loss": 0.0929,
"step": 2406
},
{
"epoch": 14.953271028037383,
"grad_norm": 1.3125,
"learning_rate": 4.947459288424932e-05,
"loss": 0.0919,
"step": 2407
},
{
"epoch": 14.959501557632398,
"grad_norm": 1.34375,
"learning_rate": 4.9474090309561635e-05,
"loss": 0.0911,
"step": 2408
},
{
"epoch": 14.965732087227414,
"grad_norm": 2.90625,
"learning_rate": 4.947358749717671e-05,
"loss": 0.213,
"step": 2409
},
{
"epoch": 14.97196261682243,
"grad_norm": 2.46875,
"learning_rate": 4.9473084447099404e-05,
"loss": 0.199,
"step": 2410
},
{
"epoch": 14.978193146417446,
"grad_norm": 1.6171875,
"learning_rate": 4.947258115933461e-05,
"loss": 0.1083,
"step": 2411
},
{
"epoch": 14.984423676012462,
"grad_norm": 2.34375,
"learning_rate": 4.947207763388722e-05,
"loss": 0.2469,
"step": 2412
},
{
"epoch": 14.990654205607477,
"grad_norm": 2.03125,
"learning_rate": 4.94715738707621e-05,
"loss": 0.1449,
"step": 2413
},
{
"epoch": 14.996884735202492,
"grad_norm": 1.921875,
"learning_rate": 4.947106986996418e-05,
"loss": 0.1315,
"step": 2414
},
{
"epoch": 15.0,
"grad_norm": 1.796875,
"learning_rate": 4.947056563149834e-05,
"loss": 0.0868,
"step": 2415
},
{
"epoch": 15.006230529595015,
"grad_norm": 2.65625,
"learning_rate": 4.947006115536947e-05,
"loss": 0.2952,
"step": 2416
},
{
"epoch": 15.01246105919003,
"grad_norm": 2.546875,
"learning_rate": 4.946955644158248e-05,
"loss": 0.2069,
"step": 2417
},
{
"epoch": 15.018691588785046,
"grad_norm": 2.015625,
"learning_rate": 4.946905149014226e-05,
"loss": 0.1491,
"step": 2418
},
{
"epoch": 15.024922118380061,
"grad_norm": 1.78125,
"learning_rate": 4.946854630105372e-05,
"loss": 0.1217,
"step": 2419
},
{
"epoch": 15.031152647975079,
"grad_norm": 3.515625,
"learning_rate": 4.946804087432177e-05,
"loss": 0.318,
"step": 2420
},
{
"epoch": 15.037383177570094,
"grad_norm": 1.625,
"learning_rate": 4.9467535209951315e-05,
"loss": 0.1107,
"step": 2421
},
{
"epoch": 15.04361370716511,
"grad_norm": 3.125,
"learning_rate": 4.946702930794726e-05,
"loss": 0.2134,
"step": 2422
},
{
"epoch": 15.049844236760125,
"grad_norm": 1.28125,
"learning_rate": 4.9466523168314535e-05,
"loss": 0.1083,
"step": 2423
},
{
"epoch": 15.05607476635514,
"grad_norm": 2.53125,
"learning_rate": 4.946601679105805e-05,
"loss": 0.2567,
"step": 2424
},
{
"epoch": 15.062305295950155,
"grad_norm": 1.9609375,
"learning_rate": 4.9465510176182714e-05,
"loss": 0.1608,
"step": 2425
},
{
"epoch": 15.06853582554517,
"grad_norm": 2.375,
"learning_rate": 4.946500332369345e-05,
"loss": 0.1786,
"step": 2426
},
{
"epoch": 15.074766355140186,
"grad_norm": 1.2734375,
"learning_rate": 4.946449623359519e-05,
"loss": 0.0994,
"step": 2427
},
{
"epoch": 15.080996884735203,
"grad_norm": 2.28125,
"learning_rate": 4.9463988905892847e-05,
"loss": 0.189,
"step": 2428
},
{
"epoch": 15.087227414330219,
"grad_norm": 2.140625,
"learning_rate": 4.946348134059136e-05,
"loss": 0.2117,
"step": 2429
},
{
"epoch": 15.093457943925234,
"grad_norm": 1.2578125,
"learning_rate": 4.946297353769564e-05,
"loss": 0.0869,
"step": 2430
},
{
"epoch": 15.09968847352025,
"grad_norm": 1.84375,
"learning_rate": 4.946246549721064e-05,
"loss": 0.1497,
"step": 2431
},
{
"epoch": 15.105919003115265,
"grad_norm": 1.4921875,
"learning_rate": 4.946195721914128e-05,
"loss": 0.1156,
"step": 2432
},
{
"epoch": 15.11214953271028,
"grad_norm": 2.015625,
"learning_rate": 4.94614487034925e-05,
"loss": 0.1944,
"step": 2433
},
{
"epoch": 15.118380062305295,
"grad_norm": 2.390625,
"learning_rate": 4.946093995026925e-05,
"loss": 0.2263,
"step": 2434
},
{
"epoch": 15.12461059190031,
"grad_norm": 1.484375,
"learning_rate": 4.9460430959476456e-05,
"loss": 0.1073,
"step": 2435
},
{
"epoch": 15.130841121495328,
"grad_norm": 1.2578125,
"learning_rate": 4.9459921731119065e-05,
"loss": 0.0895,
"step": 2436
},
{
"epoch": 15.137071651090343,
"grad_norm": 1.0,
"learning_rate": 4.945941226520203e-05,
"loss": 0.0758,
"step": 2437
},
{
"epoch": 15.143302180685358,
"grad_norm": 1.8828125,
"learning_rate": 4.945890256173029e-05,
"loss": 0.156,
"step": 2438
},
{
"epoch": 15.149532710280374,
"grad_norm": 2.1875,
"learning_rate": 4.945839262070879e-05,
"loss": 0.1432,
"step": 2439
},
{
"epoch": 15.15576323987539,
"grad_norm": 2.375,
"learning_rate": 4.945788244214251e-05,
"loss": 0.2473,
"step": 2440
},
{
"epoch": 15.161993769470405,
"grad_norm": 1.328125,
"learning_rate": 4.945737202603637e-05,
"loss": 0.0824,
"step": 2441
},
{
"epoch": 15.16822429906542,
"grad_norm": 2.1875,
"learning_rate": 4.9456861372395356e-05,
"loss": 0.1497,
"step": 2442
},
{
"epoch": 15.174454828660435,
"grad_norm": 1.40625,
"learning_rate": 4.9456350481224415e-05,
"loss": 0.0949,
"step": 2443
},
{
"epoch": 15.180685358255452,
"grad_norm": 2.5625,
"learning_rate": 4.94558393525285e-05,
"loss": 0.1969,
"step": 2444
},
{
"epoch": 15.186915887850468,
"grad_norm": 1.5390625,
"learning_rate": 4.9455327986312595e-05,
"loss": 0.1143,
"step": 2445
},
{
"epoch": 15.193146417445483,
"grad_norm": 2.140625,
"learning_rate": 4.9454816382581645e-05,
"loss": 0.1605,
"step": 2446
},
{
"epoch": 15.199376947040498,
"grad_norm": 2.953125,
"learning_rate": 4.945430454134064e-05,
"loss": 0.2256,
"step": 2447
},
{
"epoch": 15.205607476635514,
"grad_norm": 1.7890625,
"learning_rate": 4.945379246259454e-05,
"loss": 0.094,
"step": 2448
},
{
"epoch": 15.21183800623053,
"grad_norm": 2.90625,
"learning_rate": 4.9453280146348315e-05,
"loss": 0.3187,
"step": 2449
},
{
"epoch": 15.218068535825545,
"grad_norm": 1.4296875,
"learning_rate": 4.945276759260695e-05,
"loss": 0.0903,
"step": 2450
},
{
"epoch": 15.22429906542056,
"grad_norm": 1.921875,
"learning_rate": 4.9452254801375407e-05,
"loss": 0.1233,
"step": 2451
},
{
"epoch": 15.230529595015577,
"grad_norm": 2.859375,
"learning_rate": 4.945174177265869e-05,
"loss": 0.226,
"step": 2452
},
{
"epoch": 15.236760124610592,
"grad_norm": 3.046875,
"learning_rate": 4.945122850646176e-05,
"loss": 0.3301,
"step": 2453
},
{
"epoch": 15.242990654205608,
"grad_norm": 2.34375,
"learning_rate": 4.945071500278962e-05,
"loss": 0.1578,
"step": 2454
},
{
"epoch": 15.249221183800623,
"grad_norm": 1.921875,
"learning_rate": 4.945020126164724e-05,
"loss": 0.1195,
"step": 2455
},
{
"epoch": 15.255451713395638,
"grad_norm": 1.4296875,
"learning_rate": 4.944968728303962e-05,
"loss": 0.0846,
"step": 2456
},
{
"epoch": 15.261682242990654,
"grad_norm": 1.140625,
"learning_rate": 4.944917306697175e-05,
"loss": 0.0834,
"step": 2457
},
{
"epoch": 15.26791277258567,
"grad_norm": 2.109375,
"learning_rate": 4.944865861344863e-05,
"loss": 0.1498,
"step": 2458
},
{
"epoch": 15.274143302180685,
"grad_norm": 0.97265625,
"learning_rate": 4.944814392247525e-05,
"loss": 0.073,
"step": 2459
},
{
"epoch": 15.280373831775702,
"grad_norm": 2.71875,
"learning_rate": 4.9447628994056604e-05,
"loss": 0.1863,
"step": 2460
},
{
"epoch": 15.286604361370717,
"grad_norm": 2.59375,
"learning_rate": 4.94471138281977e-05,
"loss": 0.2542,
"step": 2461
},
{
"epoch": 15.292834890965732,
"grad_norm": 1.375,
"learning_rate": 4.944659842490354e-05,
"loss": 0.0899,
"step": 2462
},
{
"epoch": 15.299065420560748,
"grad_norm": 2.40625,
"learning_rate": 4.944608278417912e-05,
"loss": 0.1904,
"step": 2463
},
{
"epoch": 15.305295950155763,
"grad_norm": 2.21875,
"learning_rate": 4.944556690602947e-05,
"loss": 0.2043,
"step": 2464
},
{
"epoch": 15.311526479750778,
"grad_norm": 3.734375,
"learning_rate": 4.944505079045958e-05,
"loss": 0.3343,
"step": 2465
},
{
"epoch": 15.317757009345794,
"grad_norm": 2.078125,
"learning_rate": 4.9444534437474476e-05,
"loss": 0.1488,
"step": 2466
},
{
"epoch": 15.32398753894081,
"grad_norm": 1.890625,
"learning_rate": 4.944401784707916e-05,
"loss": 0.1443,
"step": 2467
},
{
"epoch": 15.330218068535826,
"grad_norm": 2.125,
"learning_rate": 4.9443501019278664e-05,
"loss": 0.206,
"step": 2468
},
{
"epoch": 15.336448598130842,
"grad_norm": 1.921875,
"learning_rate": 4.944298395407799e-05,
"loss": 0.1245,
"step": 2469
},
{
"epoch": 15.342679127725857,
"grad_norm": 1.765625,
"learning_rate": 4.9442466651482184e-05,
"loss": 0.1243,
"step": 2470
},
{
"epoch": 15.348909657320872,
"grad_norm": 1.203125,
"learning_rate": 4.944194911149625e-05,
"loss": 0.0918,
"step": 2471
},
{
"epoch": 15.355140186915888,
"grad_norm": 2.984375,
"learning_rate": 4.944143133412521e-05,
"loss": 0.2446,
"step": 2472
},
{
"epoch": 15.361370716510903,
"grad_norm": 0.94140625,
"learning_rate": 4.944091331937412e-05,
"loss": 0.0718,
"step": 2473
},
{
"epoch": 15.367601246105918,
"grad_norm": 1.1484375,
"learning_rate": 4.944039506724798e-05,
"loss": 0.0948,
"step": 2474
},
{
"epoch": 15.373831775700934,
"grad_norm": 1.890625,
"learning_rate": 4.943987657775185e-05,
"loss": 0.2077,
"step": 2475
},
{
"epoch": 15.38006230529595,
"grad_norm": 1.1171875,
"learning_rate": 4.9439357850890744e-05,
"loss": 0.0882,
"step": 2476
},
{
"epoch": 15.386292834890966,
"grad_norm": 2.28125,
"learning_rate": 4.9438838886669714e-05,
"loss": 0.1763,
"step": 2477
},
{
"epoch": 15.392523364485982,
"grad_norm": 1.1875,
"learning_rate": 4.943831968509379e-05,
"loss": 0.0882,
"step": 2478
},
{
"epoch": 15.398753894080997,
"grad_norm": 1.34375,
"learning_rate": 4.9437800246168025e-05,
"loss": 0.1045,
"step": 2479
},
{
"epoch": 15.404984423676012,
"grad_norm": 4.28125,
"learning_rate": 4.943728056989746e-05,
"loss": 0.4443,
"step": 2480
},
{
"epoch": 15.411214953271028,
"grad_norm": 1.8515625,
"learning_rate": 4.943676065628714e-05,
"loss": 0.1241,
"step": 2481
},
{
"epoch": 15.417445482866043,
"grad_norm": 1.5859375,
"learning_rate": 4.9436240505342115e-05,
"loss": 0.1107,
"step": 2482
},
{
"epoch": 15.423676012461058,
"grad_norm": 2.453125,
"learning_rate": 4.9435720117067434e-05,
"loss": 0.1894,
"step": 2483
},
{
"epoch": 15.429906542056075,
"grad_norm": 2.0625,
"learning_rate": 4.943519949146816e-05,
"loss": 0.1477,
"step": 2484
},
{
"epoch": 15.43613707165109,
"grad_norm": 1.78125,
"learning_rate": 4.943467862854934e-05,
"loss": 0.1033,
"step": 2485
},
{
"epoch": 15.442367601246106,
"grad_norm": 3.046875,
"learning_rate": 4.9434157528316036e-05,
"loss": 0.2204,
"step": 2486
},
{
"epoch": 15.448598130841122,
"grad_norm": 2.5625,
"learning_rate": 4.943363619077331e-05,
"loss": 0.1966,
"step": 2487
},
{
"epoch": 15.454828660436137,
"grad_norm": 1.640625,
"learning_rate": 4.943311461592623e-05,
"loss": 0.2042,
"step": 2488
},
{
"epoch": 15.461059190031152,
"grad_norm": 1.5390625,
"learning_rate": 4.943259280377986e-05,
"loss": 0.0998,
"step": 2489
},
{
"epoch": 15.467289719626168,
"grad_norm": 2.5,
"learning_rate": 4.9432070754339254e-05,
"loss": 0.2184,
"step": 2490
},
{
"epoch": 15.473520249221183,
"grad_norm": 3.265625,
"learning_rate": 4.94315484676095e-05,
"loss": 0.2549,
"step": 2491
},
{
"epoch": 15.4797507788162,
"grad_norm": 2.265625,
"learning_rate": 4.943102594359566e-05,
"loss": 0.1666,
"step": 2492
},
{
"epoch": 15.485981308411215,
"grad_norm": 1.2578125,
"learning_rate": 4.943050318230281e-05,
"loss": 0.089,
"step": 2493
},
{
"epoch": 15.49221183800623,
"grad_norm": 3.40625,
"learning_rate": 4.9429980183736034e-05,
"loss": 0.3957,
"step": 2494
},
{
"epoch": 15.498442367601246,
"grad_norm": 1.1171875,
"learning_rate": 4.942945694790041e-05,
"loss": 0.0787,
"step": 2495
},
{
"epoch": 15.504672897196262,
"grad_norm": 1.640625,
"learning_rate": 4.942893347480101e-05,
"loss": 0.1021,
"step": 2496
},
{
"epoch": 15.510903426791277,
"grad_norm": 2.375,
"learning_rate": 4.9428409764442926e-05,
"loss": 0.2503,
"step": 2497
},
{
"epoch": 15.517133956386292,
"grad_norm": 1.1015625,
"learning_rate": 4.9427885816831256e-05,
"loss": 0.0893,
"step": 2498
},
{
"epoch": 15.523364485981308,
"grad_norm": 2.03125,
"learning_rate": 4.942736163197106e-05,
"loss": 0.1286,
"step": 2499
},
{
"epoch": 15.529595015576325,
"grad_norm": 2.140625,
"learning_rate": 4.942683720986745e-05,
"loss": 0.1459,
"step": 2500
},
{
"epoch": 15.53582554517134,
"grad_norm": 1.6640625,
"learning_rate": 4.942631255052551e-05,
"loss": 0.1192,
"step": 2501
},
{
"epoch": 15.542056074766355,
"grad_norm": 2.8125,
"learning_rate": 4.942578765395035e-05,
"loss": 0.2442,
"step": 2502
},
{
"epoch": 15.54828660436137,
"grad_norm": 1.765625,
"learning_rate": 4.9425262520147055e-05,
"loss": 0.1134,
"step": 2503
},
{
"epoch": 15.554517133956386,
"grad_norm": 2.296875,
"learning_rate": 4.942473714912072e-05,
"loss": 0.1884,
"step": 2504
},
{
"epoch": 15.560747663551401,
"grad_norm": 1.40625,
"learning_rate": 4.9424211540876456e-05,
"loss": 0.0947,
"step": 2505
},
{
"epoch": 15.566978193146417,
"grad_norm": 2.765625,
"learning_rate": 4.9423685695419374e-05,
"loss": 0.311,
"step": 2506
},
{
"epoch": 15.573208722741432,
"grad_norm": 1.703125,
"learning_rate": 4.942315961275457e-05,
"loss": 0.126,
"step": 2507
},
{
"epoch": 15.57943925233645,
"grad_norm": 2.6875,
"learning_rate": 4.942263329288716e-05,
"loss": 0.206,
"step": 2508
},
{
"epoch": 15.585669781931465,
"grad_norm": 3.765625,
"learning_rate": 4.942210673582226e-05,
"loss": 0.2744,
"step": 2509
},
{
"epoch": 15.59190031152648,
"grad_norm": 2.03125,
"learning_rate": 4.9421579941564965e-05,
"loss": 0.1416,
"step": 2510
},
{
"epoch": 15.598130841121495,
"grad_norm": 0.94140625,
"learning_rate": 4.942105291012041e-05,
"loss": 0.0709,
"step": 2511
},
{
"epoch": 15.60436137071651,
"grad_norm": 3.203125,
"learning_rate": 4.94205256414937e-05,
"loss": 0.2996,
"step": 2512
},
{
"epoch": 15.610591900311526,
"grad_norm": 1.921875,
"learning_rate": 4.9419998135689974e-05,
"loss": 0.1162,
"step": 2513
},
{
"epoch": 15.616822429906541,
"grad_norm": 2.453125,
"learning_rate": 4.941947039271434e-05,
"loss": 0.163,
"step": 2514
},
{
"epoch": 15.623052959501557,
"grad_norm": 1.5703125,
"learning_rate": 4.941894241257193e-05,
"loss": 0.1185,
"step": 2515
},
{
"epoch": 15.629283489096574,
"grad_norm": 2.40625,
"learning_rate": 4.9418414195267873e-05,
"loss": 0.1889,
"step": 2516
},
{
"epoch": 15.63551401869159,
"grad_norm": 1.8046875,
"learning_rate": 4.941788574080729e-05,
"loss": 0.138,
"step": 2517
},
{
"epoch": 15.641744548286605,
"grad_norm": 2.609375,
"learning_rate": 4.941735704919532e-05,
"loss": 0.1769,
"step": 2518
},
{
"epoch": 15.64797507788162,
"grad_norm": 2.0625,
"learning_rate": 4.9416828120437095e-05,
"loss": 0.1496,
"step": 2519
},
{
"epoch": 15.654205607476635,
"grad_norm": 2.84375,
"learning_rate": 4.9416298954537763e-05,
"loss": 0.1836,
"step": 2520
},
{
"epoch": 15.66043613707165,
"grad_norm": 1.625,
"learning_rate": 4.941576955150245e-05,
"loss": 0.1119,
"step": 2521
},
{
"epoch": 15.666666666666666,
"grad_norm": 1.0703125,
"learning_rate": 4.94152399113363e-05,
"loss": 0.0756,
"step": 2522
},
{
"epoch": 15.672897196261681,
"grad_norm": 2.09375,
"learning_rate": 4.9414710034044464e-05,
"loss": 0.1476,
"step": 2523
},
{
"epoch": 15.679127725856699,
"grad_norm": 1.5625,
"learning_rate": 4.941417991963208e-05,
"loss": 0.1061,
"step": 2524
},
{
"epoch": 15.685358255451714,
"grad_norm": 1.109375,
"learning_rate": 4.9413649568104305e-05,
"loss": 0.0776,
"step": 2525
},
{
"epoch": 15.69158878504673,
"grad_norm": 1.1328125,
"learning_rate": 4.941311897946628e-05,
"loss": 0.0908,
"step": 2526
},
{
"epoch": 15.697819314641745,
"grad_norm": 1.7421875,
"learning_rate": 4.9412588153723174e-05,
"loss": 0.2062,
"step": 2527
},
{
"epoch": 15.70404984423676,
"grad_norm": 2.984375,
"learning_rate": 4.941205709088011e-05,
"loss": 0.2497,
"step": 2528
},
{
"epoch": 15.710280373831775,
"grad_norm": 2.015625,
"learning_rate": 4.941152579094229e-05,
"loss": 0.1273,
"step": 2529
},
{
"epoch": 15.71651090342679,
"grad_norm": 3.234375,
"learning_rate": 4.9410994253914845e-05,
"loss": 0.2807,
"step": 2530
},
{
"epoch": 15.722741433021806,
"grad_norm": 1.546875,
"learning_rate": 4.9410462479802945e-05,
"loss": 0.101,
"step": 2531
},
{
"epoch": 15.728971962616823,
"grad_norm": 2.046875,
"learning_rate": 4.940993046861174e-05,
"loss": 0.1781,
"step": 2532
},
{
"epoch": 15.735202492211839,
"grad_norm": 1.6328125,
"learning_rate": 4.940939822034643e-05,
"loss": 0.1151,
"step": 2533
},
{
"epoch": 15.741433021806854,
"grad_norm": 1.21875,
"learning_rate": 4.9408865735012154e-05,
"loss": 0.0849,
"step": 2534
},
{
"epoch": 15.74766355140187,
"grad_norm": 1.328125,
"learning_rate": 4.94083330126141e-05,
"loss": 0.0857,
"step": 2535
},
{
"epoch": 15.753894080996885,
"grad_norm": 1.7421875,
"learning_rate": 4.940780005315744e-05,
"loss": 0.0971,
"step": 2536
},
{
"epoch": 15.7601246105919,
"grad_norm": 1.1953125,
"learning_rate": 4.940726685664734e-05,
"loss": 0.0792,
"step": 2537
},
{
"epoch": 15.766355140186915,
"grad_norm": 2.359375,
"learning_rate": 4.940673342308898e-05,
"loss": 0.2538,
"step": 2538
},
{
"epoch": 15.77258566978193,
"grad_norm": 2.625,
"learning_rate": 4.940619975248756e-05,
"loss": 0.1804,
"step": 2539
},
{
"epoch": 15.778816199376948,
"grad_norm": 2.78125,
"learning_rate": 4.940566584484824e-05,
"loss": 0.2148,
"step": 2540
},
{
"epoch": 15.785046728971963,
"grad_norm": 2.6875,
"learning_rate": 4.9405131700176215e-05,
"loss": 0.2542,
"step": 2541
},
{
"epoch": 15.791277258566979,
"grad_norm": 2.828125,
"learning_rate": 4.940459731847667e-05,
"loss": 0.2405,
"step": 2542
},
{
"epoch": 15.797507788161994,
"grad_norm": 2.140625,
"learning_rate": 4.94040626997548e-05,
"loss": 0.1655,
"step": 2543
},
{
"epoch": 15.80373831775701,
"grad_norm": 4.78125,
"learning_rate": 4.9403527844015805e-05,
"loss": 0.4451,
"step": 2544
},
{
"epoch": 15.809968847352025,
"grad_norm": 2.09375,
"learning_rate": 4.940299275126486e-05,
"loss": 0.1201,
"step": 2545
},
{
"epoch": 15.81619937694704,
"grad_norm": 1.875,
"learning_rate": 4.940245742150718e-05,
"loss": 0.1444,
"step": 2546
},
{
"epoch": 15.822429906542055,
"grad_norm": 3.40625,
"learning_rate": 4.940192185474794e-05,
"loss": 0.1884,
"step": 2547
},
{
"epoch": 15.828660436137072,
"grad_norm": 1.3515625,
"learning_rate": 4.9401386050992365e-05,
"loss": 0.0933,
"step": 2548
},
{
"epoch": 15.834890965732088,
"grad_norm": 1.96875,
"learning_rate": 4.9400850010245656e-05,
"loss": 0.1312,
"step": 2549
},
{
"epoch": 15.841121495327103,
"grad_norm": 0.84375,
"learning_rate": 4.9400313732513006e-05,
"loss": 0.0673,
"step": 2550
},
{
"epoch": 15.847352024922118,
"grad_norm": 2.21875,
"learning_rate": 4.939977721779964e-05,
"loss": 0.1941,
"step": 2551
},
{
"epoch": 15.853582554517134,
"grad_norm": 1.5703125,
"learning_rate": 4.9399240466110754e-05,
"loss": 0.0904,
"step": 2552
},
{
"epoch": 15.85981308411215,
"grad_norm": 1.6015625,
"learning_rate": 4.939870347745156e-05,
"loss": 0.1016,
"step": 2553
},
{
"epoch": 15.866043613707165,
"grad_norm": 1.25,
"learning_rate": 4.939816625182729e-05,
"loss": 0.0914,
"step": 2554
},
{
"epoch": 15.87227414330218,
"grad_norm": 2.0,
"learning_rate": 4.9397628789243155e-05,
"loss": 0.1305,
"step": 2555
},
{
"epoch": 15.878504672897197,
"grad_norm": 1.9375,
"learning_rate": 4.9397091089704364e-05,
"loss": 0.1613,
"step": 2556
},
{
"epoch": 15.884735202492212,
"grad_norm": 2.953125,
"learning_rate": 4.939655315321615e-05,
"loss": 0.2042,
"step": 2557
},
{
"epoch": 15.890965732087228,
"grad_norm": 1.5703125,
"learning_rate": 4.9396014979783736e-05,
"loss": 0.1105,
"step": 2558
},
{
"epoch": 15.897196261682243,
"grad_norm": 1.1015625,
"learning_rate": 4.9395476569412355e-05,
"loss": 0.0731,
"step": 2559
},
{
"epoch": 15.903426791277258,
"grad_norm": 2.734375,
"learning_rate": 4.9394937922107224e-05,
"loss": 0.2014,
"step": 2560
},
{
"epoch": 15.909657320872274,
"grad_norm": 3.125,
"learning_rate": 4.9394399037873574e-05,
"loss": 0.368,
"step": 2561
},
{
"epoch": 15.91588785046729,
"grad_norm": 1.84375,
"learning_rate": 4.939385991671664e-05,
"loss": 0.1235,
"step": 2562
},
{
"epoch": 15.922118380062305,
"grad_norm": 2.875,
"learning_rate": 4.9393320558641676e-05,
"loss": 0.1915,
"step": 2563
},
{
"epoch": 15.928348909657322,
"grad_norm": 2.921875,
"learning_rate": 4.9392780963653905e-05,
"loss": 0.2945,
"step": 2564
},
{
"epoch": 15.934579439252337,
"grad_norm": 1.28125,
"learning_rate": 4.939224113175857e-05,
"loss": 0.1083,
"step": 2565
},
{
"epoch": 15.940809968847352,
"grad_norm": 2.203125,
"learning_rate": 4.939170106296091e-05,
"loss": 0.2401,
"step": 2566
},
{
"epoch": 15.947040498442368,
"grad_norm": 3.21875,
"learning_rate": 4.9391160757266164e-05,
"loss": 0.238,
"step": 2567
},
{
"epoch": 15.953271028037383,
"grad_norm": 2.625,
"learning_rate": 4.9390620214679595e-05,
"loss": 0.2065,
"step": 2568
},
{
"epoch": 15.959501557632398,
"grad_norm": 2.265625,
"learning_rate": 4.9390079435206446e-05,
"loss": 0.2199,
"step": 2569
},
{
"epoch": 15.965732087227414,
"grad_norm": 1.9375,
"learning_rate": 4.9389538418851974e-05,
"loss": 0.1635,
"step": 2570
},
{
"epoch": 15.97196261682243,
"grad_norm": 2.875,
"learning_rate": 4.938899716562143e-05,
"loss": 0.1873,
"step": 2571
},
{
"epoch": 15.978193146417446,
"grad_norm": 1.9765625,
"learning_rate": 4.938845567552007e-05,
"loss": 0.2317,
"step": 2572
},
{
"epoch": 15.984423676012462,
"grad_norm": 2.140625,
"learning_rate": 4.938791394855316e-05,
"loss": 0.1591,
"step": 2573
},
{
"epoch": 15.990654205607477,
"grad_norm": 2.359375,
"learning_rate": 4.938737198472594e-05,
"loss": 0.2144,
"step": 2574
},
{
"epoch": 15.996884735202492,
"grad_norm": 2.03125,
"learning_rate": 4.93868297840437e-05,
"loss": 0.1866,
"step": 2575
},
{
"epoch": 16.0,
"grad_norm": 1.4375,
"learning_rate": 4.938628734651168e-05,
"loss": 0.084,
"step": 2576
},
{
"epoch": 16.006230529595015,
"grad_norm": 1.9765625,
"learning_rate": 4.938574467213518e-05,
"loss": 0.1411,
"step": 2577
},
{
"epoch": 16.01246105919003,
"grad_norm": 2.9375,
"learning_rate": 4.938520176091944e-05,
"loss": 0.1947,
"step": 2578
},
{
"epoch": 16.018691588785046,
"grad_norm": 1.6015625,
"learning_rate": 4.9384658612869754e-05,
"loss": 0.1015,
"step": 2579
},
{
"epoch": 16.02492211838006,
"grad_norm": 2.890625,
"learning_rate": 4.938411522799139e-05,
"loss": 0.218,
"step": 2580
},
{
"epoch": 16.031152647975077,
"grad_norm": 1.5234375,
"learning_rate": 4.938357160628962e-05,
"loss": 0.1119,
"step": 2581
},
{
"epoch": 16.037383177570092,
"grad_norm": 1.6015625,
"learning_rate": 4.938302774776973e-05,
"loss": 0.1244,
"step": 2582
},
{
"epoch": 16.043613707165107,
"grad_norm": 1.078125,
"learning_rate": 4.9382483652437e-05,
"loss": 0.0891,
"step": 2583
},
{
"epoch": 16.049844236760123,
"grad_norm": 2.921875,
"learning_rate": 4.938193932029671e-05,
"loss": 0.2269,
"step": 2584
},
{
"epoch": 16.05607476635514,
"grad_norm": 3.9375,
"learning_rate": 4.938139475135416e-05,
"loss": 0.4413,
"step": 2585
},
{
"epoch": 16.062305295950157,
"grad_norm": 3.046875,
"learning_rate": 4.938084994561463e-05,
"loss": 0.2602,
"step": 2586
},
{
"epoch": 16.068535825545172,
"grad_norm": 2.546875,
"learning_rate": 4.9380304903083404e-05,
"loss": 0.1564,
"step": 2587
},
{
"epoch": 16.074766355140188,
"grad_norm": 1.0859375,
"learning_rate": 4.937975962376579e-05,
"loss": 0.0812,
"step": 2588
},
{
"epoch": 16.080996884735203,
"grad_norm": 2.3125,
"learning_rate": 4.937921410766707e-05,
"loss": 0.2546,
"step": 2589
},
{
"epoch": 16.08722741433022,
"grad_norm": 1.7578125,
"learning_rate": 4.937866835479256e-05,
"loss": 0.0966,
"step": 2590
},
{
"epoch": 16.093457943925234,
"grad_norm": 1.984375,
"learning_rate": 4.9378122365147536e-05,
"loss": 0.1163,
"step": 2591
},
{
"epoch": 16.09968847352025,
"grad_norm": 1.8125,
"learning_rate": 4.937757613873732e-05,
"loss": 0.125,
"step": 2592
},
{
"epoch": 16.105919003115265,
"grad_norm": 2.296875,
"learning_rate": 4.937702967556722e-05,
"loss": 0.1388,
"step": 2593
},
{
"epoch": 16.11214953271028,
"grad_norm": 1.2265625,
"learning_rate": 4.937648297564252e-05,
"loss": 0.0814,
"step": 2594
},
{
"epoch": 16.118380062305295,
"grad_norm": 1.765625,
"learning_rate": 4.9375936038968565e-05,
"loss": 0.1268,
"step": 2595
},
{
"epoch": 16.12461059190031,
"grad_norm": 1.5,
"learning_rate": 4.9375388865550635e-05,
"loss": 0.1082,
"step": 2596
},
{
"epoch": 16.130841121495326,
"grad_norm": 1.8125,
"learning_rate": 4.937484145539405e-05,
"loss": 0.1739,
"step": 2597
},
{
"epoch": 16.13707165109034,
"grad_norm": 0.94140625,
"learning_rate": 4.9374293808504136e-05,
"loss": 0.0801,
"step": 2598
},
{
"epoch": 16.143302180685357,
"grad_norm": 1.265625,
"learning_rate": 4.937374592488621e-05,
"loss": 0.1023,
"step": 2599
},
{
"epoch": 16.149532710280372,
"grad_norm": 1.9375,
"learning_rate": 4.937319780454559e-05,
"loss": 0.2277,
"step": 2600
},
{
"epoch": 16.15576323987539,
"grad_norm": 2.453125,
"learning_rate": 4.9372649447487604e-05,
"loss": 0.2075,
"step": 2601
},
{
"epoch": 16.161993769470406,
"grad_norm": 2.875,
"learning_rate": 4.9372100853717577e-05,
"loss": 0.3201,
"step": 2602
},
{
"epoch": 16.16822429906542,
"grad_norm": 2.734375,
"learning_rate": 4.937155202324083e-05,
"loss": 0.2867,
"step": 2603
},
{
"epoch": 16.174454828660437,
"grad_norm": 1.8515625,
"learning_rate": 4.9371002956062694e-05,
"loss": 0.1388,
"step": 2604
},
{
"epoch": 16.180685358255452,
"grad_norm": 2.078125,
"learning_rate": 4.937045365218851e-05,
"loss": 0.1387,
"step": 2605
},
{
"epoch": 16.186915887850468,
"grad_norm": 1.65625,
"learning_rate": 4.936990411162361e-05,
"loss": 0.107,
"step": 2606
},
{
"epoch": 16.193146417445483,
"grad_norm": 1.6875,
"learning_rate": 4.936935433437333e-05,
"loss": 0.1121,
"step": 2607
},
{
"epoch": 16.1993769470405,
"grad_norm": 1.1953125,
"learning_rate": 4.9368804320443004e-05,
"loss": 0.0863,
"step": 2608
},
{
"epoch": 16.205607476635514,
"grad_norm": 2.15625,
"learning_rate": 4.936825406983798e-05,
"loss": 0.1574,
"step": 2609
},
{
"epoch": 16.21183800623053,
"grad_norm": 2.609375,
"learning_rate": 4.9367703582563606e-05,
"loss": 0.1748,
"step": 2610
},
{
"epoch": 16.218068535825545,
"grad_norm": 2.40625,
"learning_rate": 4.936715285862523e-05,
"loss": 0.1667,
"step": 2611
},
{
"epoch": 16.22429906542056,
"grad_norm": 3.140625,
"learning_rate": 4.936660189802818e-05,
"loss": 0.2833,
"step": 2612
},
{
"epoch": 16.230529595015575,
"grad_norm": 1.515625,
"learning_rate": 4.936605070077783e-05,
"loss": 0.1013,
"step": 2613
},
{
"epoch": 16.23676012461059,
"grad_norm": 2.453125,
"learning_rate": 4.936549926687952e-05,
"loss": 0.1514,
"step": 2614
},
{
"epoch": 16.242990654205606,
"grad_norm": 2.078125,
"learning_rate": 4.936494759633862e-05,
"loss": 0.1339,
"step": 2615
},
{
"epoch": 16.24922118380062,
"grad_norm": 2.03125,
"learning_rate": 4.936439568916047e-05,
"loss": 0.1428,
"step": 2616
},
{
"epoch": 16.25545171339564,
"grad_norm": 2.84375,
"learning_rate": 4.936384354535044e-05,
"loss": 0.1974,
"step": 2617
},
{
"epoch": 16.261682242990656,
"grad_norm": 2.3125,
"learning_rate": 4.93632911649139e-05,
"loss": 0.1534,
"step": 2618
},
{
"epoch": 16.26791277258567,
"grad_norm": 2.140625,
"learning_rate": 4.936273854785619e-05,
"loss": 0.1432,
"step": 2619
},
{
"epoch": 16.274143302180686,
"grad_norm": 2.34375,
"learning_rate": 4.9362185694182706e-05,
"loss": 0.1757,
"step": 2620
},
{
"epoch": 16.2803738317757,
"grad_norm": 2.25,
"learning_rate": 4.93616326038988e-05,
"loss": 0.1375,
"step": 2621
},
{
"epoch": 16.286604361370717,
"grad_norm": 1.7109375,
"learning_rate": 4.936107927700985e-05,
"loss": 0.1011,
"step": 2622
},
{
"epoch": 16.292834890965732,
"grad_norm": 2.40625,
"learning_rate": 4.936052571352122e-05,
"loss": 0.1807,
"step": 2623
},
{
"epoch": 16.299065420560748,
"grad_norm": 2.46875,
"learning_rate": 4.93599719134383e-05,
"loss": 0.167,
"step": 2624
},
{
"epoch": 16.305295950155763,
"grad_norm": 2.265625,
"learning_rate": 4.935941787676647e-05,
"loss": 0.1602,
"step": 2625
},
{
"epoch": 16.31152647975078,
"grad_norm": 2.84375,
"learning_rate": 4.9358863603511095e-05,
"loss": 0.1903,
"step": 2626
},
{
"epoch": 16.317757009345794,
"grad_norm": 2.578125,
"learning_rate": 4.935830909367757e-05,
"loss": 0.1582,
"step": 2627
},
{
"epoch": 16.32398753894081,
"grad_norm": 1.1953125,
"learning_rate": 4.935775434727128e-05,
"loss": 0.0844,
"step": 2628
},
{
"epoch": 16.330218068535824,
"grad_norm": 3.6875,
"learning_rate": 4.9357199364297615e-05,
"loss": 0.3161,
"step": 2629
},
{
"epoch": 16.33644859813084,
"grad_norm": 3.609375,
"learning_rate": 4.935664414476196e-05,
"loss": 0.3987,
"step": 2630
},
{
"epoch": 16.342679127725855,
"grad_norm": 1.1015625,
"learning_rate": 4.9356088688669696e-05,
"loss": 0.0864,
"step": 2631
},
{
"epoch": 16.34890965732087,
"grad_norm": 2.765625,
"learning_rate": 4.9355532996026246e-05,
"loss": 0.3497,
"step": 2632
},
{
"epoch": 16.35514018691589,
"grad_norm": 1.1328125,
"learning_rate": 4.9354977066836986e-05,
"loss": 0.0782,
"step": 2633
},
{
"epoch": 16.361370716510905,
"grad_norm": 1.9375,
"learning_rate": 4.935442090110731e-05,
"loss": 0.1348,
"step": 2634
},
{
"epoch": 16.36760124610592,
"grad_norm": 0.94140625,
"learning_rate": 4.9353864498842636e-05,
"loss": 0.0672,
"step": 2635
},
{
"epoch": 16.373831775700936,
"grad_norm": 1.3203125,
"learning_rate": 4.935330786004837e-05,
"loss": 0.1163,
"step": 2636
},
{
"epoch": 16.38006230529595,
"grad_norm": 1.828125,
"learning_rate": 4.93527509847299e-05,
"loss": 0.1283,
"step": 2637
},
{
"epoch": 16.386292834890966,
"grad_norm": 3.09375,
"learning_rate": 4.935219387289265e-05,
"loss": 0.2545,
"step": 2638
},
{
"epoch": 16.39252336448598,
"grad_norm": 2.71875,
"learning_rate": 4.935163652454202e-05,
"loss": 0.2124,
"step": 2639
},
{
"epoch": 16.398753894080997,
"grad_norm": 2.75,
"learning_rate": 4.9351078939683426e-05,
"loss": 0.1845,
"step": 2640
},
{
"epoch": 16.404984423676012,
"grad_norm": 1.890625,
"learning_rate": 4.935052111832229e-05,
"loss": 0.1399,
"step": 2641
},
{
"epoch": 16.411214953271028,
"grad_norm": 1.1953125,
"learning_rate": 4.934996306046403e-05,
"loss": 0.0892,
"step": 2642
},
{
"epoch": 16.417445482866043,
"grad_norm": 2.359375,
"learning_rate": 4.9349404766114063e-05,
"loss": 0.175,
"step": 2643
},
{
"epoch": 16.42367601246106,
"grad_norm": 3.46875,
"learning_rate": 4.9348846235277804e-05,
"loss": 0.2778,
"step": 2644
},
{
"epoch": 16.429906542056074,
"grad_norm": 1.953125,
"learning_rate": 4.934828746796067e-05,
"loss": 0.134,
"step": 2645
},
{
"epoch": 16.43613707165109,
"grad_norm": 2.65625,
"learning_rate": 4.934772846416812e-05,
"loss": 0.2072,
"step": 2646
},
{
"epoch": 16.442367601246104,
"grad_norm": 1.9609375,
"learning_rate": 4.934716922390556e-05,
"loss": 0.2143,
"step": 2647
},
{
"epoch": 16.44859813084112,
"grad_norm": 1.8515625,
"learning_rate": 4.9346609747178426e-05,
"loss": 0.111,
"step": 2648
},
{
"epoch": 16.45482866043614,
"grad_norm": 2.484375,
"learning_rate": 4.9346050033992145e-05,
"loss": 0.2356,
"step": 2649
},
{
"epoch": 16.461059190031154,
"grad_norm": 2.28125,
"learning_rate": 4.9345490084352165e-05,
"loss": 0.1535,
"step": 2650
},
{
"epoch": 16.46728971962617,
"grad_norm": 2.6875,
"learning_rate": 4.9344929898263914e-05,
"loss": 0.2256,
"step": 2651
},
{
"epoch": 16.473520249221185,
"grad_norm": 1.3828125,
"learning_rate": 4.9344369475732846e-05,
"loss": 0.0936,
"step": 2652
},
{
"epoch": 16.4797507788162,
"grad_norm": 1.1640625,
"learning_rate": 4.934380881676439e-05,
"loss": 0.0845,
"step": 2653
},
{
"epoch": 16.485981308411215,
"grad_norm": 1.3359375,
"learning_rate": 4.934324792136399e-05,
"loss": 0.0886,
"step": 2654
},
{
"epoch": 16.49221183800623,
"grad_norm": 2.609375,
"learning_rate": 4.934268678953711e-05,
"loss": 0.2096,
"step": 2655
},
{
"epoch": 16.498442367601246,
"grad_norm": 1.4921875,
"learning_rate": 4.934212542128919e-05,
"loss": 0.1013,
"step": 2656
},
{
"epoch": 16.50467289719626,
"grad_norm": 1.9140625,
"learning_rate": 4.934156381662567e-05,
"loss": 0.1374,
"step": 2657
},
{
"epoch": 16.510903426791277,
"grad_norm": 1.578125,
"learning_rate": 4.9341001975552027e-05,
"loss": 0.0965,
"step": 2658
},
{
"epoch": 16.517133956386292,
"grad_norm": 1.6171875,
"learning_rate": 4.934043989807371e-05,
"loss": 0.1076,
"step": 2659
},
{
"epoch": 16.523364485981308,
"grad_norm": 1.5546875,
"learning_rate": 4.933987758419617e-05,
"loss": 0.1018,
"step": 2660
},
{
"epoch": 16.529595015576323,
"grad_norm": 2.265625,
"learning_rate": 4.9339315033924874e-05,
"loss": 0.131,
"step": 2661
},
{
"epoch": 16.53582554517134,
"grad_norm": 2.265625,
"learning_rate": 4.933875224726529e-05,
"loss": 0.1832,
"step": 2662
},
{
"epoch": 16.542056074766354,
"grad_norm": 1.1796875,
"learning_rate": 4.933818922422287e-05,
"loss": 0.0891,
"step": 2663
},
{
"epoch": 16.54828660436137,
"grad_norm": 2.640625,
"learning_rate": 4.933762596480309e-05,
"loss": 0.2423,
"step": 2664
},
{
"epoch": 16.554517133956388,
"grad_norm": 1.953125,
"learning_rate": 4.933706246901143e-05,
"loss": 0.146,
"step": 2665
},
{
"epoch": 16.560747663551403,
"grad_norm": 1.609375,
"learning_rate": 4.9336498736853347e-05,
"loss": 0.1043,
"step": 2666
},
{
"epoch": 16.56697819314642,
"grad_norm": 1.96875,
"learning_rate": 4.9335934768334326e-05,
"loss": 0.1436,
"step": 2667
},
{
"epoch": 16.573208722741434,
"grad_norm": 1.8125,
"learning_rate": 4.933537056345985e-05,
"loss": 0.1148,
"step": 2668
},
{
"epoch": 16.57943925233645,
"grad_norm": 2.109375,
"learning_rate": 4.9334806122235376e-05,
"loss": 0.2526,
"step": 2669
},
{
"epoch": 16.585669781931465,
"grad_norm": 2.953125,
"learning_rate": 4.933424144466641e-05,
"loss": 0.1763,
"step": 2670
},
{
"epoch": 16.59190031152648,
"grad_norm": 2.890625,
"learning_rate": 4.933367653075843e-05,
"loss": 0.2067,
"step": 2671
},
{
"epoch": 16.598130841121495,
"grad_norm": 1.7421875,
"learning_rate": 4.9333111380516914e-05,
"loss": 0.1091,
"step": 2672
},
{
"epoch": 16.60436137071651,
"grad_norm": 2.078125,
"learning_rate": 4.9332545993947355e-05,
"loss": 0.1834,
"step": 2673
},
{
"epoch": 16.610591900311526,
"grad_norm": 1.578125,
"learning_rate": 4.9331980371055244e-05,
"loss": 0.1175,
"step": 2674
},
{
"epoch": 16.61682242990654,
"grad_norm": 2.3125,
"learning_rate": 4.933141451184608e-05,
"loss": 0.206,
"step": 2675
},
{
"epoch": 16.623052959501557,
"grad_norm": 1.5,
"learning_rate": 4.933084841632535e-05,
"loss": 0.1213,
"step": 2676
},
{
"epoch": 16.629283489096572,
"grad_norm": 2.734375,
"learning_rate": 4.933028208449856e-05,
"loss": 0.1978,
"step": 2677
},
{
"epoch": 16.635514018691588,
"grad_norm": 2.828125,
"learning_rate": 4.932971551637121e-05,
"loss": 0.2284,
"step": 2678
},
{
"epoch": 16.641744548286603,
"grad_norm": 1.2109375,
"learning_rate": 4.932914871194879e-05,
"loss": 0.0816,
"step": 2679
},
{
"epoch": 16.64797507788162,
"grad_norm": 1.03125,
"learning_rate": 4.9328581671236824e-05,
"loss": 0.081,
"step": 2680
},
{
"epoch": 16.654205607476637,
"grad_norm": 3.046875,
"learning_rate": 4.9328014394240804e-05,
"loss": 0.2364,
"step": 2681
},
{
"epoch": 16.660436137071652,
"grad_norm": 1.125,
"learning_rate": 4.932744688096625e-05,
"loss": 0.0859,
"step": 2682
},
{
"epoch": 16.666666666666668,
"grad_norm": 1.8671875,
"learning_rate": 4.9326879131418666e-05,
"loss": 0.1326,
"step": 2683
},
{
"epoch": 16.672897196261683,
"grad_norm": 1.25,
"learning_rate": 4.932631114560357e-05,
"loss": 0.071,
"step": 2684
},
{
"epoch": 16.6791277258567,
"grad_norm": 3.03125,
"learning_rate": 4.9325742923526476e-05,
"loss": 0.2021,
"step": 2685
},
{
"epoch": 16.685358255451714,
"grad_norm": 2.4375,
"learning_rate": 4.93251744651929e-05,
"loss": 0.1693,
"step": 2686
},
{
"epoch": 16.69158878504673,
"grad_norm": 2.453125,
"learning_rate": 4.9324605770608375e-05,
"loss": 0.2069,
"step": 2687
},
{
"epoch": 16.697819314641745,
"grad_norm": 2.15625,
"learning_rate": 4.932403683977842e-05,
"loss": 0.1299,
"step": 2688
},
{
"epoch": 16.70404984423676,
"grad_norm": 1.71875,
"learning_rate": 4.9323467672708545e-05,
"loss": 0.0985,
"step": 2689
},
{
"epoch": 16.710280373831775,
"grad_norm": 1.3203125,
"learning_rate": 4.9322898269404293e-05,
"loss": 0.1026,
"step": 2690
},
{
"epoch": 16.71651090342679,
"grad_norm": 1.7421875,
"learning_rate": 4.932232862987119e-05,
"loss": 0.1111,
"step": 2691
},
{
"epoch": 16.722741433021806,
"grad_norm": 2.734375,
"learning_rate": 4.932175875411478e-05,
"loss": 0.1578,
"step": 2692
},
{
"epoch": 16.72897196261682,
"grad_norm": 2.265625,
"learning_rate": 4.9321188642140575e-05,
"loss": 0.1288,
"step": 2693
},
{
"epoch": 16.735202492211837,
"grad_norm": 2.71875,
"learning_rate": 4.932061829395413e-05,
"loss": 0.213,
"step": 2694
},
{
"epoch": 16.741433021806852,
"grad_norm": 1.6953125,
"learning_rate": 4.932004770956098e-05,
"loss": 0.1252,
"step": 2695
},
{
"epoch": 16.747663551401867,
"grad_norm": 2.0,
"learning_rate": 4.931947688896666e-05,
"loss": 0.1508,
"step": 2696
},
{
"epoch": 16.753894080996886,
"grad_norm": 1.796875,
"learning_rate": 4.931890583217672e-05,
"loss": 0.1077,
"step": 2697
},
{
"epoch": 16.7601246105919,
"grad_norm": 1.4765625,
"learning_rate": 4.931833453919671e-05,
"loss": 0.0881,
"step": 2698
},
{
"epoch": 16.766355140186917,
"grad_norm": 2.703125,
"learning_rate": 4.931776301003217e-05,
"loss": 0.1994,
"step": 2699
},
{
"epoch": 16.772585669781932,
"grad_norm": 1.046875,
"learning_rate": 4.931719124468865e-05,
"loss": 0.0779,
"step": 2700
},
{
"epoch": 16.778816199376948,
"grad_norm": 1.203125,
"learning_rate": 4.9316619243171714e-05,
"loss": 0.0984,
"step": 2701
},
{
"epoch": 16.785046728971963,
"grad_norm": 1.859375,
"learning_rate": 4.931604700548691e-05,
"loss": 0.1235,
"step": 2702
},
{
"epoch": 16.79127725856698,
"grad_norm": 2.140625,
"learning_rate": 4.93154745316398e-05,
"loss": 0.1616,
"step": 2703
},
{
"epoch": 16.797507788161994,
"grad_norm": 2.546875,
"learning_rate": 4.931490182163593e-05,
"loss": 0.203,
"step": 2704
},
{
"epoch": 16.80373831775701,
"grad_norm": 2.296875,
"learning_rate": 4.931432887548089e-05,
"loss": 0.2078,
"step": 2705
},
{
"epoch": 16.809968847352025,
"grad_norm": 3.828125,
"learning_rate": 4.931375569318021e-05,
"loss": 0.2342,
"step": 2706
},
{
"epoch": 16.81619937694704,
"grad_norm": 2.0,
"learning_rate": 4.931318227473949e-05,
"loss": 0.1159,
"step": 2707
},
{
"epoch": 16.822429906542055,
"grad_norm": 1.90625,
"learning_rate": 4.931260862016427e-05,
"loss": 0.1137,
"step": 2708
},
{
"epoch": 16.82866043613707,
"grad_norm": 1.3671875,
"learning_rate": 4.9312034729460144e-05,
"loss": 0.09,
"step": 2709
},
{
"epoch": 16.834890965732086,
"grad_norm": 1.3125,
"learning_rate": 4.931146060263268e-05,
"loss": 0.0876,
"step": 2710
},
{
"epoch": 16.8411214953271,
"grad_norm": 2.15625,
"learning_rate": 4.931088623968744e-05,
"loss": 0.1846,
"step": 2711
},
{
"epoch": 16.847352024922117,
"grad_norm": 2.75,
"learning_rate": 4.9310311640630017e-05,
"loss": 0.1676,
"step": 2712
},
{
"epoch": 16.853582554517136,
"grad_norm": 1.265625,
"learning_rate": 4.930973680546599e-05,
"loss": 0.0911,
"step": 2713
},
{
"epoch": 16.85981308411215,
"grad_norm": 1.609375,
"learning_rate": 4.9309161734200934e-05,
"loss": 0.0959,
"step": 2714
},
{
"epoch": 16.866043613707166,
"grad_norm": 2.140625,
"learning_rate": 4.9308586426840454e-05,
"loss": 0.1888,
"step": 2715
},
{
"epoch": 16.87227414330218,
"grad_norm": 2.65625,
"learning_rate": 4.930801088339011e-05,
"loss": 0.2061,
"step": 2716
},
{
"epoch": 16.878504672897197,
"grad_norm": 1.453125,
"learning_rate": 4.9307435103855507e-05,
"loss": 0.0877,
"step": 2717
},
{
"epoch": 16.884735202492212,
"grad_norm": 1.921875,
"learning_rate": 4.930685908824224e-05,
"loss": 0.1183,
"step": 2718
},
{
"epoch": 16.890965732087228,
"grad_norm": 1.7421875,
"learning_rate": 4.930628283655589e-05,
"loss": 0.1244,
"step": 2719
},
{
"epoch": 16.897196261682243,
"grad_norm": 2.34375,
"learning_rate": 4.930570634880207e-05,
"loss": 0.1471,
"step": 2720
},
{
"epoch": 16.90342679127726,
"grad_norm": 1.546875,
"learning_rate": 4.930512962498638e-05,
"loss": 0.0924,
"step": 2721
},
{
"epoch": 16.909657320872274,
"grad_norm": 1.6875,
"learning_rate": 4.93045526651144e-05,
"loss": 0.101,
"step": 2722
},
{
"epoch": 16.91588785046729,
"grad_norm": 1.984375,
"learning_rate": 4.930397546919174e-05,
"loss": 0.1957,
"step": 2723
},
{
"epoch": 16.922118380062305,
"grad_norm": 2.515625,
"learning_rate": 4.930339803722402e-05,
"loss": 0.3333,
"step": 2724
},
{
"epoch": 16.92834890965732,
"grad_norm": 3.78125,
"learning_rate": 4.930282036921685e-05,
"loss": 0.3328,
"step": 2725
},
{
"epoch": 16.934579439252335,
"grad_norm": 1.375,
"learning_rate": 4.9302242465175816e-05,
"loss": 0.0865,
"step": 2726
},
{
"epoch": 16.94080996884735,
"grad_norm": 2.9375,
"learning_rate": 4.930166432510656e-05,
"loss": 0.238,
"step": 2727
},
{
"epoch": 16.947040498442366,
"grad_norm": 1.9609375,
"learning_rate": 4.930108594901467e-05,
"loss": 0.1185,
"step": 2728
},
{
"epoch": 16.953271028037385,
"grad_norm": 0.7265625,
"learning_rate": 4.9300507336905775e-05,
"loss": 0.0647,
"step": 2729
},
{
"epoch": 16.9595015576324,
"grad_norm": 0.984375,
"learning_rate": 4.92999284887855e-05,
"loss": 0.0722,
"step": 2730
},
{
"epoch": 16.965732087227416,
"grad_norm": 0.9765625,
"learning_rate": 4.929934940465946e-05,
"loss": 0.0812,
"step": 2731
},
{
"epoch": 16.97196261682243,
"grad_norm": 1.4296875,
"learning_rate": 4.9298770084533286e-05,
"loss": 0.091,
"step": 2732
},
{
"epoch": 16.978193146417446,
"grad_norm": 2.1875,
"learning_rate": 4.9298190528412594e-05,
"loss": 0.1497,
"step": 2733
},
{
"epoch": 16.98442367601246,
"grad_norm": 1.1171875,
"learning_rate": 4.929761073630302e-05,
"loss": 0.0821,
"step": 2734
},
{
"epoch": 16.990654205607477,
"grad_norm": 2.203125,
"learning_rate": 4.929703070821019e-05,
"loss": 0.1693,
"step": 2735
},
{
"epoch": 16.996884735202492,
"grad_norm": 1.6015625,
"learning_rate": 4.929645044413975e-05,
"loss": 0.1103,
"step": 2736
},
{
"epoch": 17.0,
"grad_norm": 0.875,
"learning_rate": 4.929586994409732e-05,
"loss": 0.0589,
"step": 2737
},
{
"epoch": 17.006230529595015,
"grad_norm": 1.703125,
"learning_rate": 4.929528920808854e-05,
"loss": 0.1228,
"step": 2738
},
{
"epoch": 17.01246105919003,
"grad_norm": 1.3984375,
"learning_rate": 4.929470823611906e-05,
"loss": 0.098,
"step": 2739
},
{
"epoch": 17.018691588785046,
"grad_norm": 1.21875,
"learning_rate": 4.929412702819453e-05,
"loss": 0.0843,
"step": 2740
},
{
"epoch": 17.02492211838006,
"grad_norm": 2.71875,
"learning_rate": 4.9293545584320564e-05,
"loss": 0.1952,
"step": 2741
},
{
"epoch": 17.031152647975077,
"grad_norm": 0.9609375,
"learning_rate": 4.9292963904502834e-05,
"loss": 0.077,
"step": 2742
},
{
"epoch": 17.037383177570092,
"grad_norm": 1.4375,
"learning_rate": 4.929238198874698e-05,
"loss": 0.0928,
"step": 2743
},
{
"epoch": 17.043613707165107,
"grad_norm": 1.0859375,
"learning_rate": 4.929179983705866e-05,
"loss": 0.0927,
"step": 2744
},
{
"epoch": 17.049844236760123,
"grad_norm": 2.0,
"learning_rate": 4.9291217449443515e-05,
"loss": 0.1819,
"step": 2745
},
{
"epoch": 17.05607476635514,
"grad_norm": 0.9375,
"learning_rate": 4.929063482590722e-05,
"loss": 0.0669,
"step": 2746
},
{
"epoch": 17.062305295950157,
"grad_norm": 1.1171875,
"learning_rate": 4.929005196645542e-05,
"loss": 0.0849,
"step": 2747
},
{
"epoch": 17.068535825545172,
"grad_norm": 3.0625,
"learning_rate": 4.928946887109378e-05,
"loss": 0.2486,
"step": 2748
},
{
"epoch": 17.074766355140188,
"grad_norm": 2.21875,
"learning_rate": 4.928888553982797e-05,
"loss": 0.1577,
"step": 2749
},
{
"epoch": 17.080996884735203,
"grad_norm": 1.8671875,
"learning_rate": 4.928830197266364e-05,
"loss": 0.1394,
"step": 2750
},
{
"epoch": 17.08722741433022,
"grad_norm": 2.25,
"learning_rate": 4.9287718169606464e-05,
"loss": 0.1591,
"step": 2751
},
{
"epoch": 17.093457943925234,
"grad_norm": 2.296875,
"learning_rate": 4.9287134130662114e-05,
"loss": 0.1629,
"step": 2752
},
{
"epoch": 17.09968847352025,
"grad_norm": 1.7734375,
"learning_rate": 4.928654985583627e-05,
"loss": 0.1078,
"step": 2753
},
{
"epoch": 17.105919003115265,
"grad_norm": 1.5234375,
"learning_rate": 4.92859653451346e-05,
"loss": 0.1091,
"step": 2754
},
{
"epoch": 17.11214953271028,
"grad_norm": 1.8203125,
"learning_rate": 4.928538059856277e-05,
"loss": 0.1338,
"step": 2755
},
{
"epoch": 17.118380062305295,
"grad_norm": 2.6875,
"learning_rate": 4.928479561612648e-05,
"loss": 0.2153,
"step": 2756
},
{
"epoch": 17.12461059190031,
"grad_norm": 2.453125,
"learning_rate": 4.9284210397831384e-05,
"loss": 0.1461,
"step": 2757
},
{
"epoch": 17.130841121495326,
"grad_norm": 2.390625,
"learning_rate": 4.9283624943683204e-05,
"loss": 0.1431,
"step": 2758
},
{
"epoch": 17.13707165109034,
"grad_norm": 3.59375,
"learning_rate": 4.928303925368759e-05,
"loss": 0.2683,
"step": 2759
},
{
"epoch": 17.143302180685357,
"grad_norm": 1.921875,
"learning_rate": 4.928245332785025e-05,
"loss": 0.1221,
"step": 2760
},
{
"epoch": 17.149532710280372,
"grad_norm": 2.359375,
"learning_rate": 4.928186716617686e-05,
"loss": 0.2195,
"step": 2761
},
{
"epoch": 17.15576323987539,
"grad_norm": 1.9140625,
"learning_rate": 4.9281280768673134e-05,
"loss": 0.1481,
"step": 2762
},
{
"epoch": 17.161993769470406,
"grad_norm": 1.796875,
"learning_rate": 4.928069413534475e-05,
"loss": 0.126,
"step": 2763
},
{
"epoch": 17.16822429906542,
"grad_norm": 2.3125,
"learning_rate": 4.928010726619742e-05,
"loss": 0.142,
"step": 2764
},
{
"epoch": 17.174454828660437,
"grad_norm": 1.265625,
"learning_rate": 4.9279520161236824e-05,
"loss": 0.1001,
"step": 2765
},
{
"epoch": 17.180685358255452,
"grad_norm": 2.59375,
"learning_rate": 4.9278932820468674e-05,
"loss": 0.1823,
"step": 2766
},
{
"epoch": 17.186915887850468,
"grad_norm": 1.4609375,
"learning_rate": 4.927834524389868e-05,
"loss": 0.1054,
"step": 2767
},
{
"epoch": 17.193146417445483,
"grad_norm": 1.703125,
"learning_rate": 4.927775743153254e-05,
"loss": 0.1215,
"step": 2768
},
{
"epoch": 17.1993769470405,
"grad_norm": 2.0625,
"learning_rate": 4.9277169383375974e-05,
"loss": 0.1252,
"step": 2769
},
{
"epoch": 17.205607476635514,
"grad_norm": 1.5625,
"learning_rate": 4.9276581099434685e-05,
"loss": 0.1154,
"step": 2770
},
{
"epoch": 17.21183800623053,
"grad_norm": 2.421875,
"learning_rate": 4.927599257971438e-05,
"loss": 0.2041,
"step": 2771
},
{
"epoch": 17.218068535825545,
"grad_norm": 1.3984375,
"learning_rate": 4.927540382422079e-05,
"loss": 0.102,
"step": 2772
},
{
"epoch": 17.22429906542056,
"grad_norm": 2.21875,
"learning_rate": 4.9274814832959627e-05,
"loss": 0.1488,
"step": 2773
},
{
"epoch": 17.230529595015575,
"grad_norm": 1.3203125,
"learning_rate": 4.92742256059366e-05,
"loss": 0.0863,
"step": 2774
},
{
"epoch": 17.23676012461059,
"grad_norm": 1.328125,
"learning_rate": 4.9273636143157445e-05,
"loss": 0.0768,
"step": 2775
},
{
"epoch": 17.242990654205606,
"grad_norm": 1.890625,
"learning_rate": 4.927304644462789e-05,
"loss": 0.1124,
"step": 2776
},
{
"epoch": 17.24922118380062,
"grad_norm": 1.34375,
"learning_rate": 4.927245651035365e-05,
"loss": 0.092,
"step": 2777
},
{
"epoch": 17.25545171339564,
"grad_norm": 2.421875,
"learning_rate": 4.927186634034046e-05,
"loss": 0.1642,
"step": 2778
},
{
"epoch": 17.261682242990656,
"grad_norm": 2.3125,
"learning_rate": 4.927127593459405e-05,
"loss": 0.1416,
"step": 2779
},
{
"epoch": 17.26791277258567,
"grad_norm": 2.171875,
"learning_rate": 4.9270685293120164e-05,
"loss": 0.1354,
"step": 2780
},
{
"epoch": 17.274143302180686,
"grad_norm": 1.4765625,
"learning_rate": 4.927009441592453e-05,
"loss": 0.094,
"step": 2781
},
{
"epoch": 17.2803738317757,
"grad_norm": 2.875,
"learning_rate": 4.926950330301288e-05,
"loss": 0.2124,
"step": 2782
},
{
"epoch": 17.286604361370717,
"grad_norm": 1.2109375,
"learning_rate": 4.9268911954390974e-05,
"loss": 0.0905,
"step": 2783
},
{
"epoch": 17.292834890965732,
"grad_norm": 1.03125,
"learning_rate": 4.926832037006453e-05,
"loss": 0.0715,
"step": 2784
},
{
"epoch": 17.299065420560748,
"grad_norm": 1.9140625,
"learning_rate": 4.9267728550039316e-05,
"loss": 0.1507,
"step": 2785
},
{
"epoch": 17.305295950155763,
"grad_norm": 2.5625,
"learning_rate": 4.926713649432107e-05,
"loss": 0.1968,
"step": 2786
},
{
"epoch": 17.31152647975078,
"grad_norm": 1.0,
"learning_rate": 4.9266544202915545e-05,
"loss": 0.0776,
"step": 2787
},
{
"epoch": 17.317757009345794,
"grad_norm": 2.109375,
"learning_rate": 4.926595167582849e-05,
"loss": 0.1607,
"step": 2788
},
{
"epoch": 17.32398753894081,
"grad_norm": 1.6953125,
"learning_rate": 4.926535891306566e-05,
"loss": 0.1354,
"step": 2789
},
{
"epoch": 17.330218068535824,
"grad_norm": 2.421875,
"learning_rate": 4.926476591463282e-05,
"loss": 0.1545,
"step": 2790
},
{
"epoch": 17.33644859813084,
"grad_norm": 3.453125,
"learning_rate": 4.926417268053571e-05,
"loss": 0.2485,
"step": 2791
},
{
"epoch": 17.342679127725855,
"grad_norm": 2.1875,
"learning_rate": 4.926357921078012e-05,
"loss": 0.1236,
"step": 2792
},
{
"epoch": 17.34890965732087,
"grad_norm": 1.75,
"learning_rate": 4.926298550537179e-05,
"loss": 0.11,
"step": 2793
},
{
"epoch": 17.35514018691589,
"grad_norm": 2.78125,
"learning_rate": 4.92623915643165e-05,
"loss": 0.2444,
"step": 2794
},
{
"epoch": 17.361370716510905,
"grad_norm": 2.0,
"learning_rate": 4.926179738762e-05,
"loss": 0.2227,
"step": 2795
},
{
"epoch": 17.36760124610592,
"grad_norm": 1.8046875,
"learning_rate": 4.926120297528809e-05,
"loss": 0.1093,
"step": 2796
},
{
"epoch": 17.373831775700936,
"grad_norm": 1.921875,
"learning_rate": 4.926060832732652e-05,
"loss": 0.122,
"step": 2797
},
{
"epoch": 17.38006230529595,
"grad_norm": 3.390625,
"learning_rate": 4.9260013443741074e-05,
"loss": 0.3062,
"step": 2798
},
{
"epoch": 17.386292834890966,
"grad_norm": 2.28125,
"learning_rate": 4.9259418324537523e-05,
"loss": 0.2088,
"step": 2799
},
{
"epoch": 17.39252336448598,
"grad_norm": 3.296875,
"learning_rate": 4.9258822969721655e-05,
"loss": 0.193,
"step": 2800
},
{
"epoch": 17.398753894080997,
"grad_norm": 1.625,
"learning_rate": 4.925822737929925e-05,
"loss": 0.0916,
"step": 2801
},
{
"epoch": 17.404984423676012,
"grad_norm": 2.421875,
"learning_rate": 4.925763155327609e-05,
"loss": 0.1465,
"step": 2802
},
{
"epoch": 17.411214953271028,
"grad_norm": 1.8046875,
"learning_rate": 4.9257035491657965e-05,
"loss": 0.1135,
"step": 2803
},
{
"epoch": 17.417445482866043,
"grad_norm": 1.40625,
"learning_rate": 4.925643919445067e-05,
"loss": 0.0883,
"step": 2804
},
{
"epoch": 17.42367601246106,
"grad_norm": 2.0625,
"learning_rate": 4.9255842661659975e-05,
"loss": 0.1644,
"step": 2805
},
{
"epoch": 17.429906542056074,
"grad_norm": 2.5,
"learning_rate": 4.925524589329169e-05,
"loss": 0.1803,
"step": 2806
},
{
"epoch": 17.43613707165109,
"grad_norm": 2.03125,
"learning_rate": 4.925464888935162e-05,
"loss": 0.2052,
"step": 2807
},
{
"epoch": 17.442367601246104,
"grad_norm": 3.421875,
"learning_rate": 4.925405164984553e-05,
"loss": 0.2667,
"step": 2808
},
{
"epoch": 17.44859813084112,
"grad_norm": 2.140625,
"learning_rate": 4.9253454174779265e-05,
"loss": 0.1334,
"step": 2809
},
{
"epoch": 17.45482866043614,
"grad_norm": 1.6015625,
"learning_rate": 4.9252856464158595e-05,
"loss": 0.0982,
"step": 2810
},
{
"epoch": 17.461059190031154,
"grad_norm": 2.328125,
"learning_rate": 4.925225851798934e-05,
"loss": 0.1825,
"step": 2811
},
{
"epoch": 17.46728971962617,
"grad_norm": 1.4765625,
"learning_rate": 4.9251660336277296e-05,
"loss": 0.118,
"step": 2812
},
{
"epoch": 17.473520249221185,
"grad_norm": 2.359375,
"learning_rate": 4.9251061919028275e-05,
"loss": 0.1676,
"step": 2813
},
{
"epoch": 17.4797507788162,
"grad_norm": 2.109375,
"learning_rate": 4.92504632662481e-05,
"loss": 0.1501,
"step": 2814
},
{
"epoch": 17.485981308411215,
"grad_norm": 1.8203125,
"learning_rate": 4.9249864377942576e-05,
"loss": 0.1197,
"step": 2815
},
{
"epoch": 17.49221183800623,
"grad_norm": 0.921875,
"learning_rate": 4.924926525411752e-05,
"loss": 0.074,
"step": 2816
},
{
"epoch": 17.498442367601246,
"grad_norm": 1.9140625,
"learning_rate": 4.924866589477876e-05,
"loss": 0.1412,
"step": 2817
},
{
"epoch": 17.50467289719626,
"grad_norm": 1.6015625,
"learning_rate": 4.92480662999321e-05,
"loss": 0.1121,
"step": 2818
},
{
"epoch": 17.510903426791277,
"grad_norm": 2.0625,
"learning_rate": 4.924746646958338e-05,
"loss": 0.1459,
"step": 2819
},
{
"epoch": 17.517133956386292,
"grad_norm": 1.7734375,
"learning_rate": 4.9246866403738415e-05,
"loss": 0.1152,
"step": 2820
},
{
"epoch": 17.523364485981308,
"grad_norm": 1.96875,
"learning_rate": 4.924626610240304e-05,
"loss": 0.1358,
"step": 2821
},
{
"epoch": 17.529595015576323,
"grad_norm": 1.9140625,
"learning_rate": 4.9245665565583075e-05,
"loss": 0.1248,
"step": 2822
},
{
"epoch": 17.53582554517134,
"grad_norm": 1.3828125,
"learning_rate": 4.9245064793284365e-05,
"loss": 0.1045,
"step": 2823
},
{
"epoch": 17.542056074766354,
"grad_norm": 1.4765625,
"learning_rate": 4.924446378551274e-05,
"loss": 0.1083,
"step": 2824
},
{
"epoch": 17.54828660436137,
"grad_norm": 3.078125,
"learning_rate": 4.924386254227403e-05,
"loss": 0.2128,
"step": 2825
},
{
"epoch": 17.554517133956388,
"grad_norm": 2.0625,
"learning_rate": 4.924326106357409e-05,
"loss": 0.1199,
"step": 2826
},
{
"epoch": 17.560747663551403,
"grad_norm": 1.578125,
"learning_rate": 4.9242659349418746e-05,
"loss": 0.1055,
"step": 2827
},
{
"epoch": 17.56697819314642,
"grad_norm": 1.6796875,
"learning_rate": 4.9242057399813845e-05,
"loss": 0.1238,
"step": 2828
},
{
"epoch": 17.573208722741434,
"grad_norm": 1.2890625,
"learning_rate": 4.924145521476525e-05,
"loss": 0.0897,
"step": 2829
},
{
"epoch": 17.57943925233645,
"grad_norm": 1.6953125,
"learning_rate": 4.924085279427879e-05,
"loss": 0.1071,
"step": 2830
},
{
"epoch": 17.585669781931465,
"grad_norm": 2.59375,
"learning_rate": 4.924025013836031e-05,
"loss": 0.1823,
"step": 2831
},
{
"epoch": 17.59190031152648,
"grad_norm": 2.71875,
"learning_rate": 4.923964724701568e-05,
"loss": 0.2738,
"step": 2832
},
{
"epoch": 17.598130841121495,
"grad_norm": 3.234375,
"learning_rate": 4.923904412025077e-05,
"loss": 0.2473,
"step": 2833
},
{
"epoch": 17.60436137071651,
"grad_norm": 1.84375,
"learning_rate": 4.92384407580714e-05,
"loss": 0.1155,
"step": 2834
},
{
"epoch": 17.610591900311526,
"grad_norm": 1.8359375,
"learning_rate": 4.923783716048345e-05,
"loss": 0.1239,
"step": 2835
},
{
"epoch": 17.61682242990654,
"grad_norm": 2.09375,
"learning_rate": 4.923723332749278e-05,
"loss": 0.14,
"step": 2836
},
{
"epoch": 17.623052959501557,
"grad_norm": 1.2109375,
"learning_rate": 4.923662925910526e-05,
"loss": 0.0819,
"step": 2837
},
{
"epoch": 17.629283489096572,
"grad_norm": 2.671875,
"learning_rate": 4.923602495532675e-05,
"loss": 0.1952,
"step": 2838
},
{
"epoch": 17.635514018691588,
"grad_norm": 2.921875,
"learning_rate": 4.923542041616312e-05,
"loss": 0.2535,
"step": 2839
},
{
"epoch": 17.641744548286603,
"grad_norm": 1.6015625,
"learning_rate": 4.923481564162025e-05,
"loss": 0.1336,
"step": 2840
},
{
"epoch": 17.64797507788162,
"grad_norm": 2.4375,
"learning_rate": 4.923421063170399e-05,
"loss": 0.1551,
"step": 2841
},
{
"epoch": 17.654205607476637,
"grad_norm": 1.390625,
"learning_rate": 4.923360538642025e-05,
"loss": 0.0952,
"step": 2842
},
{
"epoch": 17.660436137071652,
"grad_norm": 1.390625,
"learning_rate": 4.923299990577488e-05,
"loss": 0.0952,
"step": 2843
},
{
"epoch": 17.666666666666668,
"grad_norm": 2.53125,
"learning_rate": 4.9232394189773764e-05,
"loss": 0.2359,
"step": 2844
},
{
"epoch": 17.672897196261683,
"grad_norm": 2.53125,
"learning_rate": 4.9231788238422806e-05,
"loss": 0.1424,
"step": 2845
},
{
"epoch": 17.6791277258567,
"grad_norm": 1.0859375,
"learning_rate": 4.923118205172786e-05,
"loss": 0.0731,
"step": 2846
},
{
"epoch": 17.685358255451714,
"grad_norm": 3.25,
"learning_rate": 4.9230575629694844e-05,
"loss": 0.3522,
"step": 2847
},
{
"epoch": 17.69158878504673,
"grad_norm": 2.359375,
"learning_rate": 4.922996897232963e-05,
"loss": 0.1536,
"step": 2848
},
{
"epoch": 17.697819314641745,
"grad_norm": 1.4453125,
"learning_rate": 4.922936207963812e-05,
"loss": 0.0867,
"step": 2849
},
{
"epoch": 17.70404984423676,
"grad_norm": 2.078125,
"learning_rate": 4.9228754951626195e-05,
"loss": 0.1164,
"step": 2850
},
{
"epoch": 17.710280373831775,
"grad_norm": 2.140625,
"learning_rate": 4.922814758829975e-05,
"loss": 0.1344,
"step": 2851
},
{
"epoch": 17.71651090342679,
"grad_norm": 3.40625,
"learning_rate": 4.9227539989664706e-05,
"loss": 0.2818,
"step": 2852
},
{
"epoch": 17.722741433021806,
"grad_norm": 1.4296875,
"learning_rate": 4.922693215572695e-05,
"loss": 0.0936,
"step": 2853
},
{
"epoch": 17.72897196261682,
"grad_norm": 2.328125,
"learning_rate": 4.922632408649238e-05,
"loss": 0.2038,
"step": 2854
},
{
"epoch": 17.735202492211837,
"grad_norm": 2.015625,
"learning_rate": 4.922571578196692e-05,
"loss": 0.1268,
"step": 2855
},
{
"epoch": 17.741433021806852,
"grad_norm": 1.171875,
"learning_rate": 4.922510724215646e-05,
"loss": 0.0836,
"step": 2856
},
{
"epoch": 17.747663551401867,
"grad_norm": 2.40625,
"learning_rate": 4.922449846706691e-05,
"loss": 0.2362,
"step": 2857
},
{
"epoch": 17.753894080996886,
"grad_norm": 1.4140625,
"learning_rate": 4.9223889456704185e-05,
"loss": 0.0838,
"step": 2858
},
{
"epoch": 17.7601246105919,
"grad_norm": 1.515625,
"learning_rate": 4.9223280211074206e-05,
"loss": 0.0965,
"step": 2859
},
{
"epoch": 17.766355140186917,
"grad_norm": 2.234375,
"learning_rate": 4.92226707301829e-05,
"loss": 0.1423,
"step": 2860
},
{
"epoch": 17.772585669781932,
"grad_norm": 1.375,
"learning_rate": 4.922206101403616e-05,
"loss": 0.0944,
"step": 2861
},
{
"epoch": 17.778816199376948,
"grad_norm": 3.15625,
"learning_rate": 4.9221451062639925e-05,
"loss": 0.3236,
"step": 2862
},
{
"epoch": 17.785046728971963,
"grad_norm": 2.53125,
"learning_rate": 4.9220840876000116e-05,
"loss": 0.1945,
"step": 2863
},
{
"epoch": 17.79127725856698,
"grad_norm": 2.109375,
"learning_rate": 4.922023045412266e-05,
"loss": 0.1205,
"step": 2864
},
{
"epoch": 17.797507788161994,
"grad_norm": 2.09375,
"learning_rate": 4.921961979701348e-05,
"loss": 0.1354,
"step": 2865
},
{
"epoch": 17.80373831775701,
"grad_norm": 1.3359375,
"learning_rate": 4.921900890467852e-05,
"loss": 0.0919,
"step": 2866
},
{
"epoch": 17.809968847352025,
"grad_norm": 1.359375,
"learning_rate": 4.92183977771237e-05,
"loss": 0.0985,
"step": 2867
},
{
"epoch": 17.81619937694704,
"grad_norm": 2.875,
"learning_rate": 4.921778641435496e-05,
"loss": 0.2089,
"step": 2868
},
{
"epoch": 17.822429906542055,
"grad_norm": 1.578125,
"learning_rate": 4.921717481637823e-05,
"loss": 0.0977,
"step": 2869
},
{
"epoch": 17.82866043613707,
"grad_norm": 1.703125,
"learning_rate": 4.9216562983199465e-05,
"loss": 0.1198,
"step": 2870
},
{
"epoch": 17.834890965732086,
"grad_norm": 2.703125,
"learning_rate": 4.921595091482461e-05,
"loss": 0.1844,
"step": 2871
},
{
"epoch": 17.8411214953271,
"grad_norm": 2.109375,
"learning_rate": 4.921533861125958e-05,
"loss": 0.1428,
"step": 2872
},
{
"epoch": 17.847352024922117,
"grad_norm": 1.9375,
"learning_rate": 4.921472607251034e-05,
"loss": 0.1227,
"step": 2873
},
{
"epoch": 17.853582554517136,
"grad_norm": 3.0,
"learning_rate": 4.921411329858285e-05,
"loss": 0.1976,
"step": 2874
},
{
"epoch": 17.85981308411215,
"grad_norm": 1.9375,
"learning_rate": 4.921350028948305e-05,
"loss": 0.1719,
"step": 2875
},
{
"epoch": 17.866043613707166,
"grad_norm": 2.75,
"learning_rate": 4.921288704521689e-05,
"loss": 0.2152,
"step": 2876
},
{
"epoch": 17.87227414330218,
"grad_norm": 3.65625,
"learning_rate": 4.9212273565790344e-05,
"loss": 0.2334,
"step": 2877
},
{
"epoch": 17.878504672897197,
"grad_norm": 3.796875,
"learning_rate": 4.9211659851209344e-05,
"loss": 0.3051,
"step": 2878
},
{
"epoch": 17.884735202492212,
"grad_norm": 1.796875,
"learning_rate": 4.921104590147987e-05,
"loss": 0.1633,
"step": 2879
},
{
"epoch": 17.890965732087228,
"grad_norm": 1.4296875,
"learning_rate": 4.9210431716607875e-05,
"loss": 0.1015,
"step": 2880
},
{
"epoch": 17.897196261682243,
"grad_norm": 2.703125,
"learning_rate": 4.920981729659933e-05,
"loss": 0.1998,
"step": 2881
},
{
"epoch": 17.90342679127726,
"grad_norm": 0.90625,
"learning_rate": 4.92092026414602e-05,
"loss": 0.0677,
"step": 2882
},
{
"epoch": 17.909657320872274,
"grad_norm": 1.7265625,
"learning_rate": 4.9208587751196454e-05,
"loss": 0.1073,
"step": 2883
},
{
"epoch": 17.91588785046729,
"grad_norm": 2.328125,
"learning_rate": 4.920797262581406e-05,
"loss": 0.2212,
"step": 2884
},
{
"epoch": 17.922118380062305,
"grad_norm": 1.140625,
"learning_rate": 4.9207357265319005e-05,
"loss": 0.0839,
"step": 2885
},
{
"epoch": 17.92834890965732,
"grad_norm": 1.21875,
"learning_rate": 4.920674166971725e-05,
"loss": 0.0796,
"step": 2886
},
{
"epoch": 17.934579439252335,
"grad_norm": 2.359375,
"learning_rate": 4.920612583901479e-05,
"loss": 0.1875,
"step": 2887
},
{
"epoch": 17.94080996884735,
"grad_norm": 1.78125,
"learning_rate": 4.920550977321759e-05,
"loss": 0.1221,
"step": 2888
},
{
"epoch": 17.947040498442366,
"grad_norm": 1.421875,
"learning_rate": 4.920489347233164e-05,
"loss": 0.0953,
"step": 2889
},
{
"epoch": 17.953271028037385,
"grad_norm": 1.4140625,
"learning_rate": 4.9204276936362926e-05,
"loss": 0.101,
"step": 2890
},
{
"epoch": 17.9595015576324,
"grad_norm": 1.09375,
"learning_rate": 4.920366016531744e-05,
"loss": 0.0782,
"step": 2891
},
{
"epoch": 17.965732087227416,
"grad_norm": 3.921875,
"learning_rate": 4.9203043159201164e-05,
"loss": 0.3337,
"step": 2892
},
{
"epoch": 17.97196261682243,
"grad_norm": 1.984375,
"learning_rate": 4.92024259180201e-05,
"loss": 0.1305,
"step": 2893
},
{
"epoch": 17.978193146417446,
"grad_norm": 2.46875,
"learning_rate": 4.920180844178024e-05,
"loss": 0.2344,
"step": 2894
},
{
"epoch": 17.98442367601246,
"grad_norm": 1.1875,
"learning_rate": 4.920119073048757e-05,
"loss": 0.0926,
"step": 2895
},
{
"epoch": 17.990654205607477,
"grad_norm": 3.03125,
"learning_rate": 4.9200572784148104e-05,
"loss": 0.31,
"step": 2896
},
{
"epoch": 17.996884735202492,
"grad_norm": 1.1484375,
"learning_rate": 4.9199954602767826e-05,
"loss": 0.0727,
"step": 2897
},
{
"epoch": 18.0,
"grad_norm": 2.3125,
"learning_rate": 4.9199336186352764e-05,
"loss": 0.146,
"step": 2898
},
{
"epoch": 18.006230529595015,
"grad_norm": 2.6875,
"learning_rate": 4.919871753490891e-05,
"loss": 0.1747,
"step": 2899
},
{
"epoch": 18.01246105919003,
"grad_norm": 2.390625,
"learning_rate": 4.919809864844227e-05,
"loss": 0.1406,
"step": 2900
},
{
"epoch": 18.018691588785046,
"grad_norm": 1.40625,
"learning_rate": 4.919747952695886e-05,
"loss": 0.0971,
"step": 2901
},
{
"epoch": 18.02492211838006,
"grad_norm": 1.1796875,
"learning_rate": 4.91968601704647e-05,
"loss": 0.0716,
"step": 2902
},
{
"epoch": 18.031152647975077,
"grad_norm": 1.515625,
"learning_rate": 4.919624057896579e-05,
"loss": 0.0877,
"step": 2903
},
{
"epoch": 18.037383177570092,
"grad_norm": 2.859375,
"learning_rate": 4.919562075246815e-05,
"loss": 0.2071,
"step": 2904
},
{
"epoch": 18.043613707165107,
"grad_norm": 2.484375,
"learning_rate": 4.919500069097781e-05,
"loss": 0.1832,
"step": 2905
},
{
"epoch": 18.049844236760123,
"grad_norm": 2.140625,
"learning_rate": 4.919438039450078e-05,
"loss": 0.2429,
"step": 2906
},
{
"epoch": 18.05607476635514,
"grad_norm": 2.90625,
"learning_rate": 4.91937598630431e-05,
"loss": 0.2335,
"step": 2907
},
{
"epoch": 18.062305295950157,
"grad_norm": 1.75,
"learning_rate": 4.9193139096610784e-05,
"loss": 0.1136,
"step": 2908
},
{
"epoch": 18.068535825545172,
"grad_norm": 2.34375,
"learning_rate": 4.9192518095209874e-05,
"loss": 0.1628,
"step": 2909
},
{
"epoch": 18.074766355140188,
"grad_norm": 1.109375,
"learning_rate": 4.9191896858846375e-05,
"loss": 0.0737,
"step": 2910
},
{
"epoch": 18.080996884735203,
"grad_norm": 1.7578125,
"learning_rate": 4.9191275387526345e-05,
"loss": 0.2249,
"step": 2911
},
{
"epoch": 18.08722741433022,
"grad_norm": 1.1640625,
"learning_rate": 4.919065368125582e-05,
"loss": 0.0782,
"step": 2912
},
{
"epoch": 18.093457943925234,
"grad_norm": 2.296875,
"learning_rate": 4.919003174004083e-05,
"loss": 0.139,
"step": 2913
},
{
"epoch": 18.09968847352025,
"grad_norm": 1.7265625,
"learning_rate": 4.918940956388741e-05,
"loss": 0.1042,
"step": 2914
},
{
"epoch": 18.105919003115265,
"grad_norm": 1.7421875,
"learning_rate": 4.918878715280161e-05,
"loss": 0.1181,
"step": 2915
},
{
"epoch": 18.11214953271028,
"grad_norm": 1.5390625,
"learning_rate": 4.9188164506789475e-05,
"loss": 0.1011,
"step": 2916
},
{
"epoch": 18.118380062305295,
"grad_norm": 1.9765625,
"learning_rate": 4.918754162585705e-05,
"loss": 0.1316,
"step": 2917
},
{
"epoch": 18.12461059190031,
"grad_norm": 2.328125,
"learning_rate": 4.918691851001039e-05,
"loss": 0.1425,
"step": 2918
},
{
"epoch": 18.130841121495326,
"grad_norm": 2.84375,
"learning_rate": 4.9186295159255537e-05,
"loss": 0.2212,
"step": 2919
},
{
"epoch": 18.13707165109034,
"grad_norm": 3.21875,
"learning_rate": 4.9185671573598555e-05,
"loss": 0.2307,
"step": 2920
},
{
"epoch": 18.143302180685357,
"grad_norm": 1.609375,
"learning_rate": 4.9185047753045486e-05,
"loss": 0.103,
"step": 2921
},
{
"epoch": 18.149532710280372,
"grad_norm": 2.84375,
"learning_rate": 4.918442369760241e-05,
"loss": 0.2106,
"step": 2922
},
{
"epoch": 18.15576323987539,
"grad_norm": 2.3125,
"learning_rate": 4.9183799407275375e-05,
"loss": 0.1504,
"step": 2923
},
{
"epoch": 18.161993769470406,
"grad_norm": 2.78125,
"learning_rate": 4.918317488207045e-05,
"loss": 0.1985,
"step": 2924
},
{
"epoch": 18.16822429906542,
"grad_norm": 2.359375,
"learning_rate": 4.9182550121993685e-05,
"loss": 0.1639,
"step": 2925
},
{
"epoch": 18.174454828660437,
"grad_norm": 1.984375,
"learning_rate": 4.918192512705117e-05,
"loss": 0.1403,
"step": 2926
},
{
"epoch": 18.180685358255452,
"grad_norm": 1.796875,
"learning_rate": 4.918129989724896e-05,
"loss": 0.1828,
"step": 2927
},
{
"epoch": 18.186915887850468,
"grad_norm": 1.5703125,
"learning_rate": 4.918067443259313e-05,
"loss": 0.0947,
"step": 2928
},
{
"epoch": 18.193146417445483,
"grad_norm": 1.5078125,
"learning_rate": 4.9180048733089755e-05,
"loss": 0.0982,
"step": 2929
},
{
"epoch": 18.1993769470405,
"grad_norm": 1.4609375,
"learning_rate": 4.917942279874491e-05,
"loss": 0.0912,
"step": 2930
},
{
"epoch": 18.205607476635514,
"grad_norm": 1.4609375,
"learning_rate": 4.917879662956469e-05,
"loss": 0.1005,
"step": 2931
},
{
"epoch": 18.21183800623053,
"grad_norm": 2.015625,
"learning_rate": 4.917817022555515e-05,
"loss": 0.1386,
"step": 2932
},
{
"epoch": 18.218068535825545,
"grad_norm": 1.6015625,
"learning_rate": 4.9177543586722395e-05,
"loss": 0.1053,
"step": 2933
},
{
"epoch": 18.22429906542056,
"grad_norm": 1.3203125,
"learning_rate": 4.917691671307251e-05,
"loss": 0.0908,
"step": 2934
},
{
"epoch": 18.230529595015575,
"grad_norm": 2.8125,
"learning_rate": 4.917628960461157e-05,
"loss": 0.285,
"step": 2935
},
{
"epoch": 18.23676012461059,
"grad_norm": 2.1875,
"learning_rate": 4.917566226134567e-05,
"loss": 0.1795,
"step": 2936
},
{
"epoch": 18.242990654205606,
"grad_norm": 1.5859375,
"learning_rate": 4.917503468328091e-05,
"loss": 0.0899,
"step": 2937
},
{
"epoch": 18.24922118380062,
"grad_norm": 2.59375,
"learning_rate": 4.917440687042338e-05,
"loss": 0.2031,
"step": 2938
},
{
"epoch": 18.25545171339564,
"grad_norm": 1.3203125,
"learning_rate": 4.917377882277917e-05,
"loss": 0.0838,
"step": 2939
},
{
"epoch": 18.261682242990656,
"grad_norm": 2.703125,
"learning_rate": 4.91731505403544e-05,
"loss": 0.2015,
"step": 2940
},
{
"epoch": 18.26791277258567,
"grad_norm": 1.5234375,
"learning_rate": 4.9172522023155154e-05,
"loss": 0.1105,
"step": 2941
},
{
"epoch": 18.274143302180686,
"grad_norm": 2.109375,
"learning_rate": 4.917189327118754e-05,
"loss": 0.1444,
"step": 2942
},
{
"epoch": 18.2803738317757,
"grad_norm": 1.9921875,
"learning_rate": 4.9171264284457675e-05,
"loss": 0.1599,
"step": 2943
},
{
"epoch": 18.286604361370717,
"grad_norm": 1.953125,
"learning_rate": 4.917063506297165e-05,
"loss": 0.1429,
"step": 2944
},
{
"epoch": 18.292834890965732,
"grad_norm": 1.4765625,
"learning_rate": 4.9170005606735594e-05,
"loss": 0.0879,
"step": 2945
},
{
"epoch": 18.299065420560748,
"grad_norm": 2.96875,
"learning_rate": 4.9169375915755606e-05,
"loss": 0.323,
"step": 2946
},
{
"epoch": 18.305295950155763,
"grad_norm": 1.8125,
"learning_rate": 4.916874599003781e-05,
"loss": 0.1159,
"step": 2947
},
{
"epoch": 18.31152647975078,
"grad_norm": 3.0,
"learning_rate": 4.9168115829588326e-05,
"loss": 0.2416,
"step": 2948
},
{
"epoch": 18.317757009345794,
"grad_norm": 1.46875,
"learning_rate": 4.916748543441326e-05,
"loss": 0.0823,
"step": 2949
},
{
"epoch": 18.32398753894081,
"grad_norm": 2.75,
"learning_rate": 4.916685480451875e-05,
"loss": 0.2693,
"step": 2950
},
{
"epoch": 18.330218068535824,
"grad_norm": 1.84375,
"learning_rate": 4.9166223939910916e-05,
"loss": 0.1241,
"step": 2951
},
{
"epoch": 18.33644859813084,
"grad_norm": 1.1015625,
"learning_rate": 4.9165592840595884e-05,
"loss": 0.0755,
"step": 2952
},
{
"epoch": 18.342679127725855,
"grad_norm": 1.5859375,
"learning_rate": 4.916496150657978e-05,
"loss": 0.1007,
"step": 2953
},
{
"epoch": 18.34890965732087,
"grad_norm": 1.90625,
"learning_rate": 4.916432993786874e-05,
"loss": 0.1263,
"step": 2954
},
{
"epoch": 18.35514018691589,
"grad_norm": 3.828125,
"learning_rate": 4.9163698134468906e-05,
"loss": 0.2476,
"step": 2955
},
{
"epoch": 18.361370716510905,
"grad_norm": 0.9921875,
"learning_rate": 4.91630660963864e-05,
"loss": 0.0653,
"step": 2956
},
{
"epoch": 18.36760124610592,
"grad_norm": 3.453125,
"learning_rate": 4.9162433823627366e-05,
"loss": 0.3513,
"step": 2957
},
{
"epoch": 18.373831775700936,
"grad_norm": 2.03125,
"learning_rate": 4.9161801316197945e-05,
"loss": 0.2188,
"step": 2958
},
{
"epoch": 18.38006230529595,
"grad_norm": 1.40625,
"learning_rate": 4.916116857410427e-05,
"loss": 0.0948,
"step": 2959
},
{
"epoch": 18.386292834890966,
"grad_norm": 3.140625,
"learning_rate": 4.916053559735251e-05,
"loss": 0.244,
"step": 2960
},
{
"epoch": 18.39252336448598,
"grad_norm": 2.84375,
"learning_rate": 4.915990238594879e-05,
"loss": 0.1912,
"step": 2961
},
{
"epoch": 18.398753894080997,
"grad_norm": 3.0,
"learning_rate": 4.915926893989927e-05,
"loss": 0.2879,
"step": 2962
},
{
"epoch": 18.404984423676012,
"grad_norm": 2.3125,
"learning_rate": 4.91586352592101e-05,
"loss": 0.1693,
"step": 2963
},
{
"epoch": 18.411214953271028,
"grad_norm": 1.28125,
"learning_rate": 4.915800134388744e-05,
"loss": 0.0848,
"step": 2964
},
{
"epoch": 18.417445482866043,
"grad_norm": 2.046875,
"learning_rate": 4.9157367193937444e-05,
"loss": 0.1198,
"step": 2965
},
{
"epoch": 18.42367601246106,
"grad_norm": 1.5390625,
"learning_rate": 4.9156732809366264e-05,
"loss": 0.0948,
"step": 2966
},
{
"epoch": 18.429906542056074,
"grad_norm": 1.0703125,
"learning_rate": 4.915609819018007e-05,
"loss": 0.0867,
"step": 2967
},
{
"epoch": 18.43613707165109,
"grad_norm": 1.9921875,
"learning_rate": 4.915546333638501e-05,
"loss": 0.1309,
"step": 2968
},
{
"epoch": 18.442367601246104,
"grad_norm": 2.125,
"learning_rate": 4.9154828247987275e-05,
"loss": 0.2049,
"step": 2969
},
{
"epoch": 18.44859813084112,
"grad_norm": 1.046875,
"learning_rate": 4.9154192924993015e-05,
"loss": 0.0739,
"step": 2970
},
{
"epoch": 18.45482866043614,
"grad_norm": 1.8515625,
"learning_rate": 4.915355736740841e-05,
"loss": 0.1857,
"step": 2971
},
{
"epoch": 18.461059190031154,
"grad_norm": 1.1015625,
"learning_rate": 4.9152921575239616e-05,
"loss": 0.0758,
"step": 2972
},
{
"epoch": 18.46728971962617,
"grad_norm": 1.7109375,
"learning_rate": 4.915228554849283e-05,
"loss": 0.1135,
"step": 2973
},
{
"epoch": 18.473520249221185,
"grad_norm": 1.890625,
"learning_rate": 4.915164928717421e-05,
"loss": 0.1466,
"step": 2974
},
{
"epoch": 18.4797507788162,
"grad_norm": 1.546875,
"learning_rate": 4.915101279128995e-05,
"loss": 0.1016,
"step": 2975
},
{
"epoch": 18.485981308411215,
"grad_norm": 2.09375,
"learning_rate": 4.915037606084622e-05,
"loss": 0.1708,
"step": 2976
},
{
"epoch": 18.49221183800623,
"grad_norm": 2.421875,
"learning_rate": 4.914973909584922e-05,
"loss": 0.1668,
"step": 2977
},
{
"epoch": 18.498442367601246,
"grad_norm": 2.734375,
"learning_rate": 4.9149101896305114e-05,
"loss": 0.1949,
"step": 2978
},
{
"epoch": 18.50467289719626,
"grad_norm": 2.1875,
"learning_rate": 4.914846446222012e-05,
"loss": 0.2074,
"step": 2979
},
{
"epoch": 18.510903426791277,
"grad_norm": 1.734375,
"learning_rate": 4.9147826793600396e-05,
"loss": 0.1917,
"step": 2980
},
{
"epoch": 18.517133956386292,
"grad_norm": 1.4921875,
"learning_rate": 4.9147188890452164e-05,
"loss": 0.0978,
"step": 2981
},
{
"epoch": 18.523364485981308,
"grad_norm": 1.6796875,
"learning_rate": 4.91465507527816e-05,
"loss": 0.1213,
"step": 2982
},
{
"epoch": 18.529595015576323,
"grad_norm": 1.078125,
"learning_rate": 4.914591238059491e-05,
"loss": 0.0818,
"step": 2983
},
{
"epoch": 18.53582554517134,
"grad_norm": 2.234375,
"learning_rate": 4.9145273773898284e-05,
"loss": 0.1419,
"step": 2984
},
{
"epoch": 18.542056074766354,
"grad_norm": 0.98046875,
"learning_rate": 4.914463493269794e-05,
"loss": 0.0748,
"step": 2985
},
{
"epoch": 18.54828660436137,
"grad_norm": 1.640625,
"learning_rate": 4.9143995857000074e-05,
"loss": 0.1274,
"step": 2986
},
{
"epoch": 18.554517133956388,
"grad_norm": 1.703125,
"learning_rate": 4.9143356546810895e-05,
"loss": 0.1108,
"step": 2987
},
{
"epoch": 18.560747663551403,
"grad_norm": 1.9609375,
"learning_rate": 4.914271700213661e-05,
"loss": 0.1472,
"step": 2988
},
{
"epoch": 18.56697819314642,
"grad_norm": 1.9140625,
"learning_rate": 4.914207722298343e-05,
"loss": 0.1093,
"step": 2989
},
{
"epoch": 18.573208722741434,
"grad_norm": 1.6796875,
"learning_rate": 4.9141437209357574e-05,
"loss": 0.0979,
"step": 2990
},
{
"epoch": 18.57943925233645,
"grad_norm": 2.25,
"learning_rate": 4.914079696126526e-05,
"loss": 0.15,
"step": 2991
},
{
"epoch": 18.585669781931465,
"grad_norm": 0.94140625,
"learning_rate": 4.9140156478712685e-05,
"loss": 0.0735,
"step": 2992
},
{
"epoch": 18.59190031152648,
"grad_norm": 1.9140625,
"learning_rate": 4.913951576170609e-05,
"loss": 0.1178,
"step": 2993
},
{
"epoch": 18.598130841121495,
"grad_norm": 2.8125,
"learning_rate": 4.91388748102517e-05,
"loss": 0.3484,
"step": 2994
},
{
"epoch": 18.60436137071651,
"grad_norm": 1.75,
"learning_rate": 4.9138233624355724e-05,
"loss": 0.1056,
"step": 2995
},
{
"epoch": 18.610591900311526,
"grad_norm": 2.734375,
"learning_rate": 4.91375922040244e-05,
"loss": 0.2017,
"step": 2996
},
{
"epoch": 18.61682242990654,
"grad_norm": 2.625,
"learning_rate": 4.913695054926396e-05,
"loss": 0.2184,
"step": 2997
},
{
"epoch": 18.623052959501557,
"grad_norm": 1.3515625,
"learning_rate": 4.913630866008063e-05,
"loss": 0.1045,
"step": 2998
},
{
"epoch": 18.629283489096572,
"grad_norm": 2.21875,
"learning_rate": 4.913566653648064e-05,
"loss": 0.1464,
"step": 2999
},
{
"epoch": 18.635514018691588,
"grad_norm": 2.734375,
"learning_rate": 4.913502417847024e-05,
"loss": 0.2774,
"step": 3000
},
{
"epoch": 18.641744548286603,
"grad_norm": 0.98046875,
"learning_rate": 4.9134381586055655e-05,
"loss": 0.0673,
"step": 3001
},
{
"epoch": 18.64797507788162,
"grad_norm": 1.9765625,
"learning_rate": 4.913373875924313e-05,
"loss": 0.1971,
"step": 3002
},
{
"epoch": 18.654205607476637,
"grad_norm": 2.140625,
"learning_rate": 4.913309569803891e-05,
"loss": 0.1754,
"step": 3003
},
{
"epoch": 18.660436137071652,
"grad_norm": 4.03125,
"learning_rate": 4.9132452402449246e-05,
"loss": 0.3549,
"step": 3004
},
{
"epoch": 18.666666666666668,
"grad_norm": 1.859375,
"learning_rate": 4.9131808872480376e-05,
"loss": 0.1306,
"step": 3005
},
{
"epoch": 18.672897196261683,
"grad_norm": 2.640625,
"learning_rate": 4.9131165108138554e-05,
"loss": 0.1728,
"step": 3006
},
{
"epoch": 18.6791277258567,
"grad_norm": 2.296875,
"learning_rate": 4.913052110943004e-05,
"loss": 0.1658,
"step": 3007
},
{
"epoch": 18.685358255451714,
"grad_norm": 3.515625,
"learning_rate": 4.9129876876361076e-05,
"loss": 0.4628,
"step": 3008
},
{
"epoch": 18.69158878504673,
"grad_norm": 1.46875,
"learning_rate": 4.9129232408937917e-05,
"loss": 0.114,
"step": 3009
},
{
"epoch": 18.697819314641745,
"grad_norm": 2.234375,
"learning_rate": 4.912858770716684e-05,
"loss": 0.1319,
"step": 3010
},
{
"epoch": 18.70404984423676,
"grad_norm": 1.890625,
"learning_rate": 4.9127942771054095e-05,
"loss": 0.2005,
"step": 3011
},
{
"epoch": 18.710280373831775,
"grad_norm": 1.1875,
"learning_rate": 4.912729760060595e-05,
"loss": 0.0797,
"step": 3012
},
{
"epoch": 18.71651090342679,
"grad_norm": 2.34375,
"learning_rate": 4.912665219582866e-05,
"loss": 0.1564,
"step": 3013
},
{
"epoch": 18.722741433021806,
"grad_norm": 2.21875,
"learning_rate": 4.91260065567285e-05,
"loss": 0.1439,
"step": 3014
},
{
"epoch": 18.72897196261682,
"grad_norm": 1.046875,
"learning_rate": 4.9125360683311744e-05,
"loss": 0.0727,
"step": 3015
},
{
"epoch": 18.735202492211837,
"grad_norm": 3.59375,
"learning_rate": 4.9124714575584664e-05,
"loss": 0.3317,
"step": 3016
},
{
"epoch": 18.741433021806852,
"grad_norm": 2.8125,
"learning_rate": 4.912406823355353e-05,
"loss": 0.1868,
"step": 3017
},
{
"epoch": 18.747663551401867,
"grad_norm": 1.296875,
"learning_rate": 4.9123421657224636e-05,
"loss": 0.0873,
"step": 3018
},
{
"epoch": 18.753894080996886,
"grad_norm": 1.5078125,
"learning_rate": 4.912277484660424e-05,
"loss": 0.112,
"step": 3019
},
{
"epoch": 18.7601246105919,
"grad_norm": 1.875,
"learning_rate": 4.912212780169863e-05,
"loss": 0.1352,
"step": 3020
},
{
"epoch": 18.766355140186917,
"grad_norm": 2.71875,
"learning_rate": 4.9121480522514085e-05,
"loss": 0.213,
"step": 3021
},
{
"epoch": 18.772585669781932,
"grad_norm": 2.546875,
"learning_rate": 4.912083300905691e-05,
"loss": 0.1436,
"step": 3022
},
{
"epoch": 18.778816199376948,
"grad_norm": 2.171875,
"learning_rate": 4.9120185261333385e-05,
"loss": 0.1102,
"step": 3023
},
{
"epoch": 18.785046728971963,
"grad_norm": 2.71875,
"learning_rate": 4.911953727934979e-05,
"loss": 0.1621,
"step": 3024
},
{
"epoch": 18.79127725856698,
"grad_norm": 2.28125,
"learning_rate": 4.9118889063112437e-05,
"loss": 0.1596,
"step": 3025
},
{
"epoch": 18.797507788161994,
"grad_norm": 2.203125,
"learning_rate": 4.911824061262761e-05,
"loss": 0.1691,
"step": 3026
},
{
"epoch": 18.80373831775701,
"grad_norm": 1.3671875,
"learning_rate": 4.911759192790161e-05,
"loss": 0.0837,
"step": 3027
},
{
"epoch": 18.809968847352025,
"grad_norm": 3.296875,
"learning_rate": 4.911694300894073e-05,
"loss": 0.2527,
"step": 3028
},
{
"epoch": 18.81619937694704,
"grad_norm": 1.296875,
"learning_rate": 4.9116293855751285e-05,
"loss": 0.0896,
"step": 3029
},
{
"epoch": 18.822429906542055,
"grad_norm": 2.890625,
"learning_rate": 4.9115644468339575e-05,
"loss": 0.1951,
"step": 3030
},
{
"epoch": 18.82866043613707,
"grad_norm": 2.15625,
"learning_rate": 4.91149948467119e-05,
"loss": 0.2217,
"step": 3031
},
{
"epoch": 18.834890965732086,
"grad_norm": 1.2578125,
"learning_rate": 4.911434499087457e-05,
"loss": 0.0769,
"step": 3032
},
{
"epoch": 18.8411214953271,
"grad_norm": 2.921875,
"learning_rate": 4.9113694900833906e-05,
"loss": 0.1989,
"step": 3033
},
{
"epoch": 18.847352024922117,
"grad_norm": 3.296875,
"learning_rate": 4.911304457659622e-05,
"loss": 0.2276,
"step": 3034
},
{
"epoch": 18.853582554517136,
"grad_norm": 2.3125,
"learning_rate": 4.9112394018167825e-05,
"loss": 0.1323,
"step": 3035
},
{
"epoch": 18.85981308411215,
"grad_norm": 0.86328125,
"learning_rate": 4.911174322555504e-05,
"loss": 0.0558,
"step": 3036
},
{
"epoch": 18.866043613707166,
"grad_norm": 2.546875,
"learning_rate": 4.911109219876417e-05,
"loss": 0.2169,
"step": 3037
},
{
"epoch": 18.87227414330218,
"grad_norm": 1.015625,
"learning_rate": 4.911044093780157e-05,
"loss": 0.0777,
"step": 3038
},
{
"epoch": 18.878504672897197,
"grad_norm": 1.6015625,
"learning_rate": 4.9109789442673545e-05,
"loss": 0.1026,
"step": 3039
},
{
"epoch": 18.884735202492212,
"grad_norm": 2.75,
"learning_rate": 4.9109137713386424e-05,
"loss": 0.1577,
"step": 3040
},
{
"epoch": 18.890965732087228,
"grad_norm": 2.375,
"learning_rate": 4.9108485749946544e-05,
"loss": 0.1425,
"step": 3041
},
{
"epoch": 18.897196261682243,
"grad_norm": 1.6953125,
"learning_rate": 4.9107833552360215e-05,
"loss": 0.1076,
"step": 3042
},
{
"epoch": 18.90342679127726,
"grad_norm": 1.8515625,
"learning_rate": 4.910718112063381e-05,
"loss": 0.1178,
"step": 3043
},
{
"epoch": 18.909657320872274,
"grad_norm": 2.25,
"learning_rate": 4.910652845477363e-05,
"loss": 0.1779,
"step": 3044
},
{
"epoch": 18.91588785046729,
"grad_norm": 0.87109375,
"learning_rate": 4.910587555478602e-05,
"loss": 0.0626,
"step": 3045
},
{
"epoch": 18.922118380062305,
"grad_norm": 2.484375,
"learning_rate": 4.910522242067733e-05,
"loss": 0.1883,
"step": 3046
},
{
"epoch": 18.92834890965732,
"grad_norm": 3.578125,
"learning_rate": 4.910456905245391e-05,
"loss": 0.3525,
"step": 3047
},
{
"epoch": 18.934579439252335,
"grad_norm": 1.859375,
"learning_rate": 4.9103915450122094e-05,
"loss": 0.1351,
"step": 3048
},
{
"epoch": 18.94080996884735,
"grad_norm": 2.5,
"learning_rate": 4.910326161368823e-05,
"loss": 0.1599,
"step": 3049
},
{
"epoch": 18.947040498442366,
"grad_norm": 1.390625,
"learning_rate": 4.910260754315868e-05,
"loss": 0.0815,
"step": 3050
},
{
"epoch": 18.953271028037385,
"grad_norm": 3.359375,
"learning_rate": 4.9101953238539775e-05,
"loss": 0.3449,
"step": 3051
},
{
"epoch": 18.9595015576324,
"grad_norm": 1.375,
"learning_rate": 4.910129869983789e-05,
"loss": 0.0942,
"step": 3052
},
{
"epoch": 18.965732087227416,
"grad_norm": 2.390625,
"learning_rate": 4.910064392705937e-05,
"loss": 0.1526,
"step": 3053
},
{
"epoch": 18.97196261682243,
"grad_norm": 1.53125,
"learning_rate": 4.9099988920210585e-05,
"loss": 0.1035,
"step": 3054
},
{
"epoch": 18.978193146417446,
"grad_norm": 2.84375,
"learning_rate": 4.909933367929789e-05,
"loss": 0.1923,
"step": 3055
},
{
"epoch": 18.98442367601246,
"grad_norm": 3.0,
"learning_rate": 4.909867820432764e-05,
"loss": 0.2797,
"step": 3056
},
{
"epoch": 18.990654205607477,
"grad_norm": 2.109375,
"learning_rate": 4.909802249530622e-05,
"loss": 0.1259,
"step": 3057
},
{
"epoch": 18.996884735202492,
"grad_norm": 1.609375,
"learning_rate": 4.909736655223998e-05,
"loss": 0.1066,
"step": 3058
},
{
"epoch": 19.0,
"grad_norm": 2.703125,
"learning_rate": 4.90967103751353e-05,
"loss": 0.1976,
"step": 3059
}
],
"logging_steps": 1,
"max_steps": 32200,
"num_input_tokens_seen": 0,
"num_train_epochs": 200,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 0.0,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}