1b-200b_nor / trainer_state.json
Lie24's picture
Upload model from 200b_nor
a822632 verified
Raw
History Blame Contribute Delete
99.7 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"eval_steps": 500,
"global_step": 606,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0049504950495049506,
"learning_rate": 0,
"loss": 2.0345,
"step": 1
},
{
"epoch": 0.009900990099009901,
"learning_rate": 0,
"loss": 2.0585,
"step": 2
},
{
"epoch": 0.01485148514851485,
"learning_rate": 0,
"loss": 2.0398,
"step": 3
},
{
"epoch": 0.019801980198019802,
"grad_norm": 4.209187030792236,
"learning_rate": 0.0,
"loss": 2.0229,
"step": 4
},
{
"epoch": 0.024752475247524754,
"grad_norm": 4.209187030792236,
"learning_rate": 0.0,
"loss": 2.0908,
"step": 5
},
{
"epoch": 0.0297029702970297,
"grad_norm": 4.209187030792236,
"learning_rate": 0.0,
"loss": 2.0517,
"step": 6
},
{
"epoch": 0.034653465346534656,
"grad_norm": 4.209187030792236,
"learning_rate": 0.0,
"loss": 2.0646,
"step": 7
},
{
"epoch": 0.039603960396039604,
"grad_norm": 4.125631332397461,
"learning_rate": 3.3722619737900224e-06,
"loss": 2.0672,
"step": 8
},
{
"epoch": 0.04455445544554455,
"grad_norm": 4.125631332397461,
"learning_rate": 3.3722619737900224e-06,
"loss": 2.1191,
"step": 9
},
{
"epoch": 0.04950495049504951,
"grad_norm": 4.125631332397461,
"learning_rate": 3.3722619737900224e-06,
"loss": 2.0464,
"step": 10
},
{
"epoch": 0.054455445544554455,
"grad_norm": 4.125631332397461,
"learning_rate": 3.3722619737900224e-06,
"loss": 2.0722,
"step": 11
},
{
"epoch": 0.0594059405940594,
"grad_norm": 4.173640251159668,
"learning_rate": 5.344908771065097e-06,
"loss": 2.0743,
"step": 12
},
{
"epoch": 0.06435643564356436,
"grad_norm": 4.173640251159668,
"learning_rate": 5.344908771065097e-06,
"loss": 2.0854,
"step": 13
},
{
"epoch": 0.06930693069306931,
"grad_norm": 4.173640251159668,
"learning_rate": 5.344908771065097e-06,
"loss": 2.0599,
"step": 14
},
{
"epoch": 0.07425742574257425,
"grad_norm": 4.173640251159668,
"learning_rate": 5.344908771065097e-06,
"loss": 2.1657,
"step": 15
},
{
"epoch": 0.07920792079207921,
"grad_norm": 4.176037788391113,
"learning_rate": 6.744523947580045e-06,
"loss": 2.0968,
"step": 16
},
{
"epoch": 0.08415841584158416,
"grad_norm": 4.176037788391113,
"learning_rate": 6.744523947580045e-06,
"loss": 1.9771,
"step": 17
},
{
"epoch": 0.0891089108910891,
"grad_norm": 4.176037788391113,
"learning_rate": 6.744523947580045e-06,
"loss": 2.0003,
"step": 18
},
{
"epoch": 0.09405940594059406,
"grad_norm": 4.176037788391113,
"learning_rate": 6.744523947580045e-06,
"loss": 2.0625,
"step": 19
},
{
"epoch": 0.09900990099009901,
"grad_norm": 3.837010383605957,
"learning_rate": 7.830149820263363e-06,
"loss": 2.0683,
"step": 20
},
{
"epoch": 0.10396039603960396,
"grad_norm": 3.837010383605957,
"learning_rate": 7.830149820263363e-06,
"loss": 2.0997,
"step": 21
},
{
"epoch": 0.10891089108910891,
"grad_norm": 3.837010383605957,
"learning_rate": 7.830149820263363e-06,
"loss": 2.0397,
"step": 22
},
{
"epoch": 0.11386138613861387,
"grad_norm": 3.837010383605957,
"learning_rate": 7.830149820263363e-06,
"loss": 1.9807,
"step": 23
},
{
"epoch": 0.1188118811881188,
"grad_norm": 3.8453454971313477,
"learning_rate": 8.717170744855119e-06,
"loss": 2.1141,
"step": 24
},
{
"epoch": 0.12376237623762376,
"grad_norm": 3.8453454971313477,
"learning_rate": 8.717170744855119e-06,
"loss": 2.0142,
"step": 25
},
{
"epoch": 0.12871287128712872,
"grad_norm": 3.8453454971313477,
"learning_rate": 8.717170744855119e-06,
"loss": 2.0366,
"step": 26
},
{
"epoch": 0.13366336633663367,
"grad_norm": 3.8453454971313477,
"learning_rate": 8.717170744855119e-06,
"loss": 1.9011,
"step": 27
},
{
"epoch": 0.13861386138613863,
"grad_norm": 3.6541080474853516,
"learning_rate": 9.46713625058711e-06,
"loss": 1.9517,
"step": 28
},
{
"epoch": 0.14356435643564355,
"grad_norm": 3.6541080474853516,
"learning_rate": 9.46713625058711e-06,
"loss": 1.754,
"step": 29
},
{
"epoch": 0.1485148514851485,
"grad_norm": 3.6541080474853516,
"learning_rate": 9.46713625058711e-06,
"loss": 1.8566,
"step": 30
},
{
"epoch": 0.15346534653465346,
"grad_norm": 3.6541080474853516,
"learning_rate": 9.46713625058711e-06,
"loss": 1.8216,
"step": 31
},
{
"epoch": 0.15841584158415842,
"grad_norm": 4.683382987976074,
"learning_rate": 1.0116785921370066e-05,
"loss": 1.9257,
"step": 32
},
{
"epoch": 0.16336633663366337,
"grad_norm": 4.683382987976074,
"learning_rate": 1.0116785921370066e-05,
"loss": 1.7506,
"step": 33
},
{
"epoch": 0.16831683168316833,
"grad_norm": 4.683382987976074,
"learning_rate": 1.0116785921370066e-05,
"loss": 1.8564,
"step": 34
},
{
"epoch": 0.17326732673267325,
"grad_norm": 4.683382987976074,
"learning_rate": 1.0116785921370066e-05,
"loss": 1.8363,
"step": 35
},
{
"epoch": 0.1782178217821782,
"grad_norm": 3.723945140838623,
"learning_rate": 1.0689817542130194e-05,
"loss": 1.7481,
"step": 36
},
{
"epoch": 0.18316831683168316,
"grad_norm": 3.723945140838623,
"learning_rate": 1.0689817542130194e-05,
"loss": 1.7635,
"step": 37
},
{
"epoch": 0.18811881188118812,
"grad_norm": 3.723945140838623,
"learning_rate": 1.0689817542130194e-05,
"loss": 1.8008,
"step": 38
},
{
"epoch": 0.19306930693069307,
"grad_norm": 3.723945140838623,
"learning_rate": 1.0689817542130194e-05,
"loss": 1.8132,
"step": 39
},
{
"epoch": 0.19801980198019803,
"grad_norm": 2.6204545497894287,
"learning_rate": 1.1202411794053388e-05,
"loss": 1.779,
"step": 40
},
{
"epoch": 0.20297029702970298,
"grad_norm": 2.6204545497894287,
"learning_rate": 1.1202411794053388e-05,
"loss": 1.6519,
"step": 41
},
{
"epoch": 0.2079207920792079,
"grad_norm": 2.6204545497894287,
"learning_rate": 1.1202411794053388e-05,
"loss": 1.7333,
"step": 42
},
{
"epoch": 0.21287128712871287,
"grad_norm": 2.6204545497894287,
"learning_rate": 1.1202411794053388e-05,
"loss": 1.7226,
"step": 43
},
{
"epoch": 0.21782178217821782,
"grad_norm": 2.1774990558624268,
"learning_rate": 1.1666109698457423e-05,
"loss": 1.6555,
"step": 44
},
{
"epoch": 0.22277227722772278,
"grad_norm": 2.1774990558624268,
"learning_rate": 1.1666109698457423e-05,
"loss": 1.7738,
"step": 45
},
{
"epoch": 0.22772277227722773,
"grad_norm": 2.1774990558624268,
"learning_rate": 1.1666109698457423e-05,
"loss": 1.7065,
"step": 46
},
{
"epoch": 0.23267326732673269,
"grad_norm": 2.1774990558624268,
"learning_rate": 1.1666109698457423e-05,
"loss": 1.6668,
"step": 47
},
{
"epoch": 0.2376237623762376,
"grad_norm": 2.0205671787261963,
"learning_rate": 1.2089432718645142e-05,
"loss": 1.6546,
"step": 48
},
{
"epoch": 0.24257425742574257,
"grad_norm": 2.0205671787261963,
"learning_rate": 1.2089432718645142e-05,
"loss": 1.634,
"step": 49
},
{
"epoch": 0.24752475247524752,
"grad_norm": 2.0205671787261963,
"learning_rate": 1.2089432718645142e-05,
"loss": 1.6475,
"step": 50
},
{
"epoch": 0.2524752475247525,
"grad_norm": 2.0205671787261963,
"learning_rate": 1.2089432718645142e-05,
"loss": 1.7894,
"step": 51
},
{
"epoch": 0.25742574257425743,
"grad_norm": 1.4178664684295654,
"learning_rate": 1.2478852147789474e-05,
"loss": 1.6676,
"step": 52
},
{
"epoch": 0.2623762376237624,
"grad_norm": 1.4178664684295654,
"learning_rate": 1.2478852147789474e-05,
"loss": 1.6544,
"step": 53
},
{
"epoch": 0.26732673267326734,
"grad_norm": 1.4178664684295654,
"learning_rate": 1.2478852147789474e-05,
"loss": 1.6889,
"step": 54
},
{
"epoch": 0.2722772277227723,
"grad_norm": 1.4178664684295654,
"learning_rate": 1.2478852147789474e-05,
"loss": 1.6452,
"step": 55
},
{
"epoch": 0.27722772277227725,
"grad_norm": 1.224704384803772,
"learning_rate": 1.2839398224377133e-05,
"loss": 1.6775,
"step": 56
},
{
"epoch": 0.28217821782178215,
"grad_norm": 1.224704384803772,
"learning_rate": 1.2839398224377133e-05,
"loss": 1.6098,
"step": 57
},
{
"epoch": 0.2871287128712871,
"grad_norm": 1.224704384803772,
"learning_rate": 1.2839398224377133e-05,
"loss": 1.5452,
"step": 58
},
{
"epoch": 0.29207920792079206,
"grad_norm": 1.224704384803772,
"learning_rate": 1.2839398224377133e-05,
"loss": 1.6561,
"step": 59
},
{
"epoch": 0.297029702970297,
"grad_norm": 1.1560372114181519,
"learning_rate": 1.317505859132846e-05,
"loss": 1.5946,
"step": 60
},
{
"epoch": 0.30198019801980197,
"grad_norm": 1.1560372114181519,
"learning_rate": 1.317505859132846e-05,
"loss": 1.5241,
"step": 61
},
{
"epoch": 0.3069306930693069,
"grad_norm": 1.1560372114181519,
"learning_rate": 1.317505859132846e-05,
"loss": 1.6005,
"step": 62
},
{
"epoch": 0.3118811881188119,
"grad_norm": 1.1560372114181519,
"learning_rate": 1.317505859132846e-05,
"loss": 1.6085,
"step": 63
},
{
"epoch": 0.31683168316831684,
"grad_norm": 1.0169198513031006,
"learning_rate": 1.348904789516009e-05,
"loss": 1.5602,
"step": 64
},
{
"epoch": 0.3217821782178218,
"grad_norm": 1.0169198513031006,
"learning_rate": 1.348904789516009e-05,
"loss": 1.6285,
"step": 65
},
{
"epoch": 0.32673267326732675,
"grad_norm": 1.0169198513031006,
"learning_rate": 1.348904789516009e-05,
"loss": 1.6125,
"step": 66
},
{
"epoch": 0.3316831683168317,
"grad_norm": 1.0169198513031006,
"learning_rate": 1.348904789516009e-05,
"loss": 1.5793,
"step": 67
},
{
"epoch": 0.33663366336633666,
"grad_norm": 1.8628435134887695,
"learning_rate": 1.3783995508828243e-05,
"loss": 1.6558,
"step": 68
},
{
"epoch": 0.3415841584158416,
"grad_norm": 1.8628435134887695,
"learning_rate": 1.3783995508828243e-05,
"loss": 1.559,
"step": 69
},
{
"epoch": 0.3465346534653465,
"grad_norm": 1.8628435134887695,
"learning_rate": 1.3783995508828243e-05,
"loss": 1.6103,
"step": 70
},
{
"epoch": 0.35148514851485146,
"grad_norm": 1.8628435134887695,
"learning_rate": 1.3783995508828243e-05,
"loss": 1.5236,
"step": 71
},
{
"epoch": 0.3564356435643564,
"grad_norm": 1.1173871755599976,
"learning_rate": 1.4062079515920212e-05,
"loss": 1.5886,
"step": 72
},
{
"epoch": 0.3613861386138614,
"grad_norm": 1.1173871755599976,
"learning_rate": 1.4062079515920212e-05,
"loss": 1.5753,
"step": 73
},
{
"epoch": 0.36633663366336633,
"grad_norm": 1.1173871755599976,
"learning_rate": 1.4062079515920212e-05,
"loss": 1.5629,
"step": 74
},
{
"epoch": 0.3712871287128713,
"grad_norm": 1.1173871755599976,
"learning_rate": 1.4062079515920212e-05,
"loss": 1.5504,
"step": 75
},
{
"epoch": 0.37623762376237624,
"grad_norm": 0.9223718047142029,
"learning_rate": 1.4325124421002207e-05,
"loss": 1.5459,
"step": 76
},
{
"epoch": 0.3811881188118812,
"grad_norm": 0.9223718047142029,
"learning_rate": 1.4325124421002207e-05,
"loss": 1.5124,
"step": 77
},
{
"epoch": 0.38613861386138615,
"grad_norm": 0.9223718047142029,
"learning_rate": 1.4325124421002207e-05,
"loss": 1.5781,
"step": 78
},
{
"epoch": 0.3910891089108911,
"grad_norm": 0.9223718047142029,
"learning_rate": 1.4325124421002207e-05,
"loss": 1.496,
"step": 79
},
{
"epoch": 0.39603960396039606,
"grad_norm": 0.8484737873077393,
"learning_rate": 1.4574673767843407e-05,
"loss": 1.5254,
"step": 80
},
{
"epoch": 0.400990099009901,
"grad_norm": 0.8484737873077393,
"learning_rate": 1.4574673767843407e-05,
"loss": 1.5498,
"step": 81
},
{
"epoch": 0.40594059405940597,
"grad_norm": 0.8484737873077393,
"learning_rate": 1.4574673767843407e-05,
"loss": 1.5435,
"step": 82
},
{
"epoch": 0.41089108910891087,
"grad_norm": 0.8484737873077393,
"learning_rate": 1.4574673767843407e-05,
"loss": 1.5388,
"step": 83
},
{
"epoch": 0.4158415841584158,
"grad_norm": 0.7764194011688232,
"learning_rate": 1.4812045021652206e-05,
"loss": 1.5761,
"step": 84
},
{
"epoch": 0.4207920792079208,
"grad_norm": 0.7764194011688232,
"learning_rate": 1.4812045021652206e-05,
"loss": 1.5303,
"step": 85
},
{
"epoch": 0.42574257425742573,
"grad_norm": 0.7764194011688232,
"learning_rate": 1.4812045021652206e-05,
"loss": 1.4934,
"step": 86
},
{
"epoch": 0.4306930693069307,
"grad_norm": 0.7764194011688232,
"learning_rate": 1.4812045021652206e-05,
"loss": 1.5334,
"step": 87
},
{
"epoch": 0.43564356435643564,
"grad_norm": 0.7228842377662659,
"learning_rate": 1.5038371672247447e-05,
"loss": 1.5522,
"step": 88
},
{
"epoch": 0.4405940594059406,
"grad_norm": 0.7228842377662659,
"learning_rate": 1.5038371672247447e-05,
"loss": 1.5202,
"step": 89
},
{
"epoch": 0.44554455445544555,
"grad_norm": 0.7228842377662659,
"learning_rate": 1.5038371672247447e-05,
"loss": 1.5,
"step": 90
},
{
"epoch": 0.4504950495049505,
"grad_norm": 0.7228842377662659,
"learning_rate": 1.5038371672247447e-05,
"loss": 1.541,
"step": 91
},
{
"epoch": 0.45544554455445546,
"grad_norm": 0.6627932786941528,
"learning_rate": 1.5254635970494278e-05,
"loss": 1.529,
"step": 92
},
{
"epoch": 0.4603960396039604,
"grad_norm": 0.6627932786941528,
"learning_rate": 1.5254635970494278e-05,
"loss": 1.5146,
"step": 93
},
{
"epoch": 0.46534653465346537,
"grad_norm": 0.6627932786941528,
"learning_rate": 1.5254635970494278e-05,
"loss": 1.4557,
"step": 94
},
{
"epoch": 0.47029702970297027,
"grad_norm": 0.6627932786941528,
"learning_rate": 1.5254635970494278e-05,
"loss": 1.5658,
"step": 95
},
{
"epoch": 0.4752475247524752,
"grad_norm": 0.5558151602745056,
"learning_rate": 1.5461694692435165e-05,
"loss": 1.5039,
"step": 96
},
{
"epoch": 0.4801980198019802,
"grad_norm": 0.5558151602745056,
"learning_rate": 1.5461694692435165e-05,
"loss": 1.426,
"step": 97
},
{
"epoch": 0.48514851485148514,
"grad_norm": 0.5558151602745056,
"learning_rate": 1.5461694692435165e-05,
"loss": 1.4679,
"step": 98
},
{
"epoch": 0.4900990099009901,
"grad_norm": 0.5558151602745056,
"learning_rate": 1.5461694692435165e-05,
"loss": 1.5091,
"step": 99
},
{
"epoch": 0.49504950495049505,
"grad_norm": 0.5453949570655823,
"learning_rate": 1.5660299640526725e-05,
"loss": 1.5156,
"step": 100
},
{
"epoch": 0.5,
"grad_norm": 0.5453949570655823,
"learning_rate": 1.5660299640526725e-05,
"loss": 1.5273,
"step": 101
},
{
"epoch": 0.504950495049505,
"grad_norm": 0.5453949570655823,
"learning_rate": 1.5660299640526725e-05,
"loss": 1.4772,
"step": 102
},
{
"epoch": 0.5099009900990099,
"grad_norm": 0.5453949570655823,
"learning_rate": 1.5660299640526725e-05,
"loss": 1.5542,
"step": 103
},
{
"epoch": 0.5148514851485149,
"grad_norm": 0.5719484090805054,
"learning_rate": 1.5851114121579497e-05,
"loss": 1.4709,
"step": 104
},
{
"epoch": 0.5198019801980198,
"grad_norm": 0.5719484090805054,
"learning_rate": 1.5851114121579497e-05,
"loss": 1.4731,
"step": 105
},
{
"epoch": 0.5247524752475248,
"grad_norm": 0.5719484090805054,
"learning_rate": 1.5851114121579497e-05,
"loss": 1.5327,
"step": 106
},
{
"epoch": 0.5297029702970297,
"grad_norm": 0.5719484090805054,
"learning_rate": 1.5851114121579497e-05,
"loss": 1.4846,
"step": 107
},
{
"epoch": 0.5346534653465347,
"grad_norm": 0.5558347702026367,
"learning_rate": 1.603472631319529e-05,
"loss": 1.4792,
"step": 108
},
{
"epoch": 0.5396039603960396,
"grad_norm": 0.5558347702026367,
"learning_rate": 1.603472631319529e-05,
"loss": 1.4207,
"step": 109
},
{
"epoch": 0.5445544554455446,
"grad_norm": 0.5558347702026367,
"learning_rate": 1.603472631319529e-05,
"loss": 1.5215,
"step": 110
},
{
"epoch": 0.5495049504950495,
"grad_norm": 0.5558347702026367,
"learning_rate": 1.603472631319529e-05,
"loss": 1.445,
"step": 111
},
{
"epoch": 0.5544554455445545,
"grad_norm": 0.5420742034912109,
"learning_rate": 1.6211660198167156e-05,
"loss": 1.5001,
"step": 112
},
{
"epoch": 0.5594059405940595,
"grad_norm": 0.5420742034912109,
"learning_rate": 1.6211660198167156e-05,
"loss": 1.422,
"step": 113
},
{
"epoch": 0.5643564356435643,
"grad_norm": 0.5420742034912109,
"learning_rate": 1.6211660198167156e-05,
"loss": 1.4553,
"step": 114
},
{
"epoch": 0.5693069306930693,
"grad_norm": 0.5420742034912109,
"learning_rate": 1.6211660198167156e-05,
"loss": 1.4299,
"step": 115
},
{
"epoch": 0.5742574257425742,
"grad_norm": 0.5059153437614441,
"learning_rate": 1.6382384579263326e-05,
"loss": 1.4914,
"step": 116
},
{
"epoch": 0.5792079207920792,
"grad_norm": 0.5059153437614441,
"learning_rate": 1.6382384579263326e-05,
"loss": 1.475,
"step": 117
},
{
"epoch": 0.5841584158415841,
"grad_norm": 0.5059153437614441,
"learning_rate": 1.6382384579263326e-05,
"loss": 1.4929,
"step": 118
},
{
"epoch": 0.5891089108910891,
"grad_norm": 0.5059153437614441,
"learning_rate": 1.6382384579263326e-05,
"loss": 1.4906,
"step": 119
},
{
"epoch": 0.594059405940594,
"grad_norm": 0.5310883522033691,
"learning_rate": 1.6547320565118483e-05,
"loss": 1.4862,
"step": 120
},
{
"epoch": 0.599009900990099,
"grad_norm": 0.5310883522033691,
"learning_rate": 1.6547320565118483e-05,
"loss": 1.4575,
"step": 121
},
{
"epoch": 0.6039603960396039,
"grad_norm": 0.5310883522033691,
"learning_rate": 1.6547320565118483e-05,
"loss": 1.4348,
"step": 122
},
{
"epoch": 0.6089108910891089,
"grad_norm": 0.5310883522033691,
"learning_rate": 1.6547320565118483e-05,
"loss": 1.5009,
"step": 123
},
{
"epoch": 0.6138613861386139,
"grad_norm": 0.5079181790351868,
"learning_rate": 1.670684782820849e-05,
"loss": 1.457,
"step": 124
},
{
"epoch": 0.6188118811881188,
"grad_norm": 0.5079181790351868,
"learning_rate": 1.670684782820849e-05,
"loss": 1.4546,
"step": 125
},
{
"epoch": 0.6237623762376238,
"grad_norm": 0.5079181790351868,
"learning_rate": 1.670684782820849e-05,
"loss": 1.4618,
"step": 126
},
{
"epoch": 0.6287128712871287,
"grad_norm": 0.5079181790351868,
"learning_rate": 1.670684782820849e-05,
"loss": 1.4523,
"step": 127
},
{
"epoch": 0.6336633663366337,
"grad_norm": 0.47183218598365784,
"learning_rate": 1.6861309868950113e-05,
"loss": 1.431,
"step": 128
},
{
"epoch": 0.6386138613861386,
"grad_norm": 0.47183218598365784,
"learning_rate": 1.6861309868950113e-05,
"loss": 1.4504,
"step": 129
},
{
"epoch": 0.6435643564356436,
"grad_norm": 0.47183218598365784,
"learning_rate": 1.6861309868950113e-05,
"loss": 1.4339,
"step": 130
},
{
"epoch": 0.6485148514851485,
"grad_norm": 0.47183218598365784,
"learning_rate": 1.6861309868950113e-05,
"loss": 1.3721,
"step": 131
},
{
"epoch": 0.6534653465346535,
"grad_norm": 0.4818020462989807,
"learning_rate": 1.701101846952252e-05,
"loss": 1.4547,
"step": 132
},
{
"epoch": 0.6584158415841584,
"grad_norm": 0.4818020462989807,
"learning_rate": 1.701101846952252e-05,
"loss": 1.4792,
"step": 133
},
{
"epoch": 0.6633663366336634,
"grad_norm": 0.4818020462989807,
"learning_rate": 1.701101846952252e-05,
"loss": 1.4623,
"step": 134
},
{
"epoch": 0.6683168316831684,
"grad_norm": 0.4818020462989807,
"learning_rate": 1.701101846952252e-05,
"loss": 1.4307,
"step": 135
},
{
"epoch": 0.6732673267326733,
"grad_norm": 0.4701865017414093,
"learning_rate": 1.7156257482618265e-05,
"loss": 1.4044,
"step": 136
},
{
"epoch": 0.6782178217821783,
"grad_norm": 0.4701865017414093,
"learning_rate": 1.7156257482618265e-05,
"loss": 1.4443,
"step": 137
},
{
"epoch": 0.6831683168316832,
"grad_norm": 0.4701865017414093,
"learning_rate": 1.7156257482618265e-05,
"loss": 1.4075,
"step": 138
},
{
"epoch": 0.6881188118811881,
"grad_norm": 0.4701865017414093,
"learning_rate": 1.7156257482618265e-05,
"loss": 1.4285,
"step": 139
},
{
"epoch": 0.693069306930693,
"grad_norm": 0.4632120728492737,
"learning_rate": 1.7297286070850474e-05,
"loss": 1.4193,
"step": 140
},
{
"epoch": 0.698019801980198,
"grad_norm": 0.4632120728492737,
"learning_rate": 1.7297286070850474e-05,
"loss": 1.4308,
"step": 141
},
{
"epoch": 0.7029702970297029,
"grad_norm": 0.4632120728492737,
"learning_rate": 1.7297286070850474e-05,
"loss": 1.345,
"step": 142
},
{
"epoch": 0.7079207920792079,
"grad_norm": 0.4632120728492737,
"learning_rate": 1.7297286070850474e-05,
"loss": 1.4194,
"step": 143
},
{
"epoch": 0.7128712871287128,
"grad_norm": 0.42748597264289856,
"learning_rate": 1.7434341489710237e-05,
"loss": 1.3935,
"step": 144
},
{
"epoch": 0.7178217821782178,
"grad_norm": 0.42748597264289856,
"learning_rate": 1.7434341489710237e-05,
"loss": 1.4309,
"step": 145
},
{
"epoch": 0.7227722772277227,
"grad_norm": 0.42748597264289856,
"learning_rate": 1.7434341489710237e-05,
"loss": 1.4268,
"step": 146
},
{
"epoch": 0.7277227722772277,
"grad_norm": 0.42748597264289856,
"learning_rate": 1.7434341489710237e-05,
"loss": 1.3828,
"step": 147
},
{
"epoch": 0.7326732673267327,
"grad_norm": 0.4540037214756012,
"learning_rate": 1.7567641489142956e-05,
"loss": 1.3705,
"step": 148
},
{
"epoch": 0.7376237623762376,
"grad_norm": 0.4540037214756012,
"learning_rate": 1.7567641489142956e-05,
"loss": 1.4615,
"step": 149
},
{
"epoch": 0.7425742574257426,
"grad_norm": 0.4540037214756012,
"learning_rate": 1.7567641489142956e-05,
"loss": 1.4167,
"step": 150
},
{
"epoch": 0.7475247524752475,
"grad_norm": 0.4540037214756012,
"learning_rate": 1.7567641489142956e-05,
"loss": 1.4241,
"step": 151
},
{
"epoch": 0.7524752475247525,
"grad_norm": 0.41410332918167114,
"learning_rate": 1.769738639479223e-05,
"loss": 1.4645,
"step": 152
},
{
"epoch": 0.7574257425742574,
"grad_norm": 0.41410332918167114,
"learning_rate": 1.769738639479223e-05,
"loss": 1.3894,
"step": 153
},
{
"epoch": 0.7623762376237624,
"grad_norm": 0.41410332918167114,
"learning_rate": 1.769738639479223e-05,
"loss": 1.4719,
"step": 154
},
{
"epoch": 0.7673267326732673,
"grad_norm": 0.41410332918167114,
"learning_rate": 1.769738639479223e-05,
"loss": 1.3803,
"step": 155
},
{
"epoch": 0.7722772277227723,
"grad_norm": 0.43414077162742615,
"learning_rate": 1.782376091885457e-05,
"loss": 1.4036,
"step": 156
},
{
"epoch": 0.7772277227722773,
"grad_norm": 0.43414077162742615,
"learning_rate": 1.782376091885457e-05,
"loss": 1.5118,
"step": 157
},
{
"epoch": 0.7821782178217822,
"grad_norm": 0.43414077162742615,
"learning_rate": 1.782376091885457e-05,
"loss": 1.4621,
"step": 158
},
{
"epoch": 0.7871287128712872,
"grad_norm": 0.43414077162742615,
"learning_rate": 1.782376091885457e-05,
"loss": 1.4165,
"step": 159
},
{
"epoch": 0.7920792079207921,
"grad_norm": 0.4301641285419464,
"learning_rate": 1.794693574163343e-05,
"loss": 1.4183,
"step": 160
},
{
"epoch": 0.7970297029702971,
"grad_norm": 0.4301641285419464,
"learning_rate": 1.794693574163343e-05,
"loss": 1.3872,
"step": 161
},
{
"epoch": 0.801980198019802,
"grad_norm": 0.4301641285419464,
"learning_rate": 1.794693574163343e-05,
"loss": 1.4385,
"step": 162
},
{
"epoch": 0.806930693069307,
"grad_norm": 0.4301641285419464,
"learning_rate": 1.794693574163343e-05,
"loss": 1.3858,
"step": 163
},
{
"epoch": 0.8118811881188119,
"grad_norm": 0.4241538345813751,
"learning_rate": 1.8067068897776073e-05,
"loss": 1.4738,
"step": 164
},
{
"epoch": 0.8168316831683168,
"grad_norm": 0.4241538345813751,
"learning_rate": 1.8067068897776073e-05,
"loss": 1.4518,
"step": 165
},
{
"epoch": 0.8217821782178217,
"grad_norm": 0.4241538345813751,
"learning_rate": 1.8067068897776073e-05,
"loss": 1.427,
"step": 166
},
{
"epoch": 0.8267326732673267,
"grad_norm": 0.4241538345813751,
"learning_rate": 1.8067068897776073e-05,
"loss": 1.4738,
"step": 167
},
{
"epoch": 0.8316831683168316,
"grad_norm": 0.4305919110774994,
"learning_rate": 1.8184306995442228e-05,
"loss": 1.3938,
"step": 168
},
{
"epoch": 0.8366336633663366,
"grad_norm": 0.4305919110774994,
"learning_rate": 1.8184306995442228e-05,
"loss": 1.418,
"step": 169
},
{
"epoch": 0.8415841584158416,
"grad_norm": 0.4305919110774994,
"learning_rate": 1.8184306995442228e-05,
"loss": 1.4442,
"step": 170
},
{
"epoch": 0.8465346534653465,
"grad_norm": 0.4305919110774994,
"learning_rate": 1.8184306995442228e-05,
"loss": 1.4044,
"step": 171
},
{
"epoch": 0.8514851485148515,
"grad_norm": 0.4117210805416107,
"learning_rate": 1.829878629199893e-05,
"loss": 1.4607,
"step": 172
},
{
"epoch": 0.8564356435643564,
"grad_norm": 0.4117210805416107,
"learning_rate": 1.829878629199893e-05,
"loss": 1.4324,
"step": 173
},
{
"epoch": 0.8613861386138614,
"grad_norm": 0.4117210805416107,
"learning_rate": 1.829878629199893e-05,
"loss": 1.4398,
"step": 174
},
{
"epoch": 0.8663366336633663,
"grad_norm": 0.4117210805416107,
"learning_rate": 1.829878629199893e-05,
"loss": 1.3948,
"step": 175
},
{
"epoch": 0.8712871287128713,
"grad_norm": 0.4110773801803589,
"learning_rate": 1.8410633646037467e-05,
"loss": 1.4956,
"step": 176
},
{
"epoch": 0.8762376237623762,
"grad_norm": 0.4110773801803589,
"learning_rate": 1.8410633646037467e-05,
"loss": 1.376,
"step": 177
},
{
"epoch": 0.8811881188118812,
"grad_norm": 0.4110773801803589,
"learning_rate": 1.8410633646037467e-05,
"loss": 1.4004,
"step": 178
},
{
"epoch": 0.8861386138613861,
"grad_norm": 0.4110773801803589,
"learning_rate": 1.8410633646037467e-05,
"loss": 1.3954,
"step": 179
},
{
"epoch": 0.8910891089108911,
"grad_norm": 0.43696391582489014,
"learning_rate": 1.8519967362393555e-05,
"loss": 1.4219,
"step": 180
},
{
"epoch": 0.8960396039603961,
"grad_norm": 0.43696391582489014,
"learning_rate": 1.8519967362393555e-05,
"loss": 1.3707,
"step": 181
},
{
"epoch": 0.900990099009901,
"grad_norm": 0.43696391582489014,
"learning_rate": 1.8519967362393555e-05,
"loss": 1.3917,
"step": 182
},
{
"epoch": 0.905940594059406,
"grad_norm": 0.43696391582489014,
"learning_rate": 1.8519967362393555e-05,
"loss": 1.338,
"step": 183
},
{
"epoch": 0.9108910891089109,
"grad_norm": 0.41487598419189453,
"learning_rate": 1.86268979442843e-05,
"loss": 1.45,
"step": 184
},
{
"epoch": 0.9158415841584159,
"grad_norm": 0.41487598419189453,
"learning_rate": 1.86268979442843e-05,
"loss": 1.3708,
"step": 185
},
{
"epoch": 0.9207920792079208,
"grad_norm": 0.41487598419189453,
"learning_rate": 1.86268979442843e-05,
"loss": 1.3827,
"step": 186
},
{
"epoch": 0.9257425742574258,
"grad_norm": 0.41487598419189453,
"learning_rate": 1.86268979442843e-05,
"loss": 1.4042,
"step": 187
},
{
"epoch": 0.9306930693069307,
"grad_norm": 0.4020172655582428,
"learning_rate": 1.8731528764550483e-05,
"loss": 1.3495,
"step": 188
},
{
"epoch": 0.9356435643564357,
"grad_norm": 0.4020172655582428,
"learning_rate": 1.8731528764550483e-05,
"loss": 1.4029,
"step": 189
},
{
"epoch": 0.9405940594059405,
"grad_norm": 0.4020172655582428,
"learning_rate": 1.8731528764550483e-05,
"loss": 1.3869,
"step": 190
},
{
"epoch": 0.9455445544554455,
"grad_norm": 0.4020172655582428,
"learning_rate": 1.8731528764550483e-05,
"loss": 1.3952,
"step": 191
},
{
"epoch": 0.9504950495049505,
"grad_norm": 0.42964646220207214,
"learning_rate": 1.883395666622519e-05,
"loss": 1.4665,
"step": 192
},
{
"epoch": 0.9554455445544554,
"grad_norm": 0.42964646220207214,
"learning_rate": 1.883395666622519e-05,
"loss": 1.3636,
"step": 193
},
{
"epoch": 0.9603960396039604,
"grad_norm": 0.42964646220207214,
"learning_rate": 1.883395666622519e-05,
"loss": 1.4287,
"step": 194
},
{
"epoch": 0.9653465346534653,
"grad_norm": 0.42964646220207214,
"learning_rate": 1.883395666622519e-05,
"loss": 1.4016,
"step": 195
},
{
"epoch": 0.9702970297029703,
"grad_norm": 0.4380471110343933,
"learning_rate": 1.893427250117422e-05,
"loss": 1.4009,
"step": 196
},
{
"epoch": 0.9752475247524752,
"grad_norm": 0.4380471110343933,
"learning_rate": 1.893427250117422e-05,
"loss": 1.3946,
"step": 197
},
{
"epoch": 0.9801980198019802,
"grad_norm": 0.4380471110343933,
"learning_rate": 1.893427250117422e-05,
"loss": 1.4028,
"step": 198
},
{
"epoch": 0.9851485148514851,
"grad_norm": 0.4380471110343933,
"learning_rate": 1.893427250117422e-05,
"loss": 1.4349,
"step": 199
},
{
"epoch": 0.9900990099009901,
"grad_norm": 0.4167586863040924,
"learning_rate": 1.903256161431675e-05,
"loss": 1.4017,
"step": 200
},
{
"epoch": 0.995049504950495,
"grad_norm": 0.4167586863040924,
"learning_rate": 1.903256161431675e-05,
"loss": 1.4089,
"step": 201
},
{
"epoch": 1.0,
"grad_norm": 0.4167586863040924,
"learning_rate": 1.903256161431675e-05,
"loss": 1.3633,
"step": 202
},
{
"epoch": 1.004950495049505,
"grad_norm": 0.4167586863040924,
"learning_rate": 1.903256161431675e-05,
"loss": 1.3055,
"step": 203
},
{
"epoch": 1.00990099009901,
"grad_norm": 0.4184738099575043,
"learning_rate": 1.912890427989334e-05,
"loss": 1.3383,
"step": 204
},
{
"epoch": 1.0148514851485149,
"grad_norm": 0.4184738099575043,
"learning_rate": 1.912890427989334e-05,
"loss": 1.3695,
"step": 205
},
{
"epoch": 1.0198019801980198,
"grad_norm": 0.4184738099575043,
"learning_rate": 1.912890427989334e-05,
"loss": 1.3554,
"step": 206
},
{
"epoch": 1.0247524752475248,
"grad_norm": 0.4184738099575043,
"learning_rate": 1.912890427989334e-05,
"loss": 1.3308,
"step": 207
},
{
"epoch": 1.0297029702970297,
"grad_norm": 0.3985500633716583,
"learning_rate": 1.922337609536952e-05,
"loss": 1.4412,
"step": 208
},
{
"epoch": 1.0346534653465347,
"grad_norm": 0.3985500633716583,
"learning_rate": 1.922337609536952e-05,
"loss": 1.3089,
"step": 209
},
{
"epoch": 1.0396039603960396,
"grad_norm": 0.3985500633716583,
"learning_rate": 1.922337609536952e-05,
"loss": 1.357,
"step": 210
},
{
"epoch": 1.0445544554455446,
"grad_norm": 0.3985500633716583,
"learning_rate": 1.922337609536952e-05,
"loss": 1.3764,
"step": 211
},
{
"epoch": 1.0495049504950495,
"grad_norm": 0.41849395632743835,
"learning_rate": 1.9316048337817536e-05,
"loss": 1.3597,
"step": 212
},
{
"epoch": 1.0544554455445545,
"grad_norm": 0.41849395632743835,
"learning_rate": 1.9316048337817536e-05,
"loss": 1.3121,
"step": 213
},
{
"epoch": 1.0594059405940595,
"grad_norm": 0.41849395632743835,
"learning_rate": 1.9316048337817536e-05,
"loss": 1.3736,
"step": 214
},
{
"epoch": 1.0643564356435644,
"grad_norm": 0.41849395632743835,
"learning_rate": 1.9316048337817536e-05,
"loss": 1.3734,
"step": 215
},
{
"epoch": 1.0693069306930694,
"grad_norm": 0.418398380279541,
"learning_rate": 1.9406988286985313e-05,
"loss": 1.3206,
"step": 216
},
{
"epoch": 1.0742574257425743,
"grad_norm": 0.418398380279541,
"learning_rate": 1.9406988286985313e-05,
"loss": 1.3278,
"step": 217
},
{
"epoch": 1.0792079207920793,
"grad_norm": 0.418398380279541,
"learning_rate": 1.9406988286985313e-05,
"loss": 1.3299,
"step": 218
},
{
"epoch": 1.0841584158415842,
"grad_norm": 0.418398380279541,
"learning_rate": 1.9406988286985313e-05,
"loss": 1.2563,
"step": 219
},
{
"epoch": 1.0891089108910892,
"grad_norm": 0.4052264392375946,
"learning_rate": 1.9496259518720788e-05,
"loss": 1.3457,
"step": 220
},
{
"epoch": 1.0940594059405941,
"grad_norm": 0.4052264392375946,
"learning_rate": 1.9496259518720788e-05,
"loss": 1.3533,
"step": 221
},
{
"epoch": 1.099009900990099,
"grad_norm": 0.4052264392375946,
"learning_rate": 1.9496259518720788e-05,
"loss": 1.3337,
"step": 222
},
{
"epoch": 1.103960396039604,
"grad_norm": 0.4052264392375946,
"learning_rate": 1.9496259518720788e-05,
"loss": 1.376,
"step": 223
},
{
"epoch": 1.108910891089109,
"grad_norm": 0.4087579548358917,
"learning_rate": 1.958392217195718e-05,
"loss": 1.2935,
"step": 224
},
{
"epoch": 1.113861386138614,
"grad_norm": 0.4087579548358917,
"learning_rate": 1.958392217195718e-05,
"loss": 1.3463,
"step": 225
},
{
"epoch": 1.118811881188119,
"grad_norm": 0.4087579548358917,
"learning_rate": 1.958392217195718e-05,
"loss": 1.3692,
"step": 226
},
{
"epoch": 1.1237623762376239,
"grad_norm": 0.4087579548358917,
"learning_rate": 1.958392217195718e-05,
"loss": 1.3841,
"step": 227
},
{
"epoch": 1.1287128712871288,
"grad_norm": 0.4138057231903076,
"learning_rate": 1.9670033192067303e-05,
"loss": 1.3293,
"step": 228
},
{
"epoch": 1.1336633663366338,
"grad_norm": 0.4138057231903076,
"learning_rate": 1.9670033192067303e-05,
"loss": 1.3184,
"step": 229
},
{
"epoch": 1.1386138613861387,
"grad_norm": 0.4138057231903076,
"learning_rate": 1.9670033192067303e-05,
"loss": 1.324,
"step": 230
},
{
"epoch": 1.1435643564356435,
"grad_norm": 0.4138057231903076,
"learning_rate": 1.9670033192067303e-05,
"loss": 1.4038,
"step": 231
},
{
"epoch": 1.1485148514851484,
"grad_norm": 0.4029700458049774,
"learning_rate": 1.9754646553053346e-05,
"loss": 1.3516,
"step": 232
},
{
"epoch": 1.1534653465346534,
"grad_norm": 0.4029700458049774,
"learning_rate": 1.9754646553053346e-05,
"loss": 1.3441,
"step": 233
},
{
"epoch": 1.1584158415841583,
"grad_norm": 0.4029700458049774,
"learning_rate": 1.9754646553053346e-05,
"loss": 1.4112,
"step": 234
},
{
"epoch": 1.1633663366336633,
"grad_norm": 0.4029700458049774,
"learning_rate": 1.9754646553053346e-05,
"loss": 1.3665,
"step": 235
},
{
"epoch": 1.1683168316831682,
"grad_norm": 0.42000120878219604,
"learning_rate": 1.983781346074312e-05,
"loss": 1.3685,
"step": 236
},
{
"epoch": 1.1732673267326732,
"grad_norm": 0.42000120878219604,
"learning_rate": 1.983781346074312e-05,
"loss": 1.2953,
"step": 237
},
{
"epoch": 1.1782178217821782,
"grad_norm": 0.42000120878219604,
"learning_rate": 1.983781346074312e-05,
"loss": 1.3794,
"step": 238
},
{
"epoch": 1.183168316831683,
"grad_norm": 0.42000120878219604,
"learning_rate": 1.983781346074312e-05,
"loss": 1.3798,
"step": 239
},
{
"epoch": 1.188118811881188,
"grad_norm": 0.4064106047153473,
"learning_rate": 1.9919582538908503e-05,
"loss": 1.3141,
"step": 240
},
{
"epoch": 1.193069306930693,
"grad_norm": 0.4064106047153473,
"learning_rate": 1.9919582538908503e-05,
"loss": 1.3422,
"step": 241
},
{
"epoch": 1.198019801980198,
"grad_norm": 0.4064106047153473,
"learning_rate": 1.9919582538908503e-05,
"loss": 1.2989,
"step": 242
},
{
"epoch": 1.202970297029703,
"grad_norm": 0.4064106047153473,
"learning_rate": 1.9919582538908503e-05,
"loss": 1.3844,
"step": 243
},
{
"epoch": 1.2079207920792079,
"grad_norm": 0.41598373651504517,
"learning_rate": 2e-05,
"loss": 1.3988,
"step": 244
},
{
"epoch": 1.2128712871287128,
"grad_norm": 0.41598373651504517,
"learning_rate": 2e-05,
"loss": 1.3686,
"step": 245
},
{
"epoch": 1.2178217821782178,
"grad_norm": 0.41598373651504517,
"learning_rate": 2e-05,
"loss": 1.3098,
"step": 246
},
{
"epoch": 1.2227722772277227,
"grad_norm": 0.41598373651504517,
"learning_rate": 2e-05,
"loss": 1.3163,
"step": 247
},
{
"epoch": 1.2277227722772277,
"grad_norm": 0.39052268862724304,
"learning_rate": 2e-05,
"loss": 1.3101,
"step": 248
},
{
"epoch": 1.2326732673267327,
"grad_norm": 0.39052268862724304,
"learning_rate": 2e-05,
"loss": 1.3795,
"step": 249
},
{
"epoch": 1.2376237623762376,
"grad_norm": 0.39052268862724304,
"learning_rate": 2e-05,
"loss": 1.3273,
"step": 250
},
{
"epoch": 1.2425742574257426,
"grad_norm": 0.39052268862724304,
"learning_rate": 2e-05,
"loss": 1.4051,
"step": 251
},
{
"epoch": 1.2475247524752475,
"grad_norm": 0.4185222089290619,
"learning_rate": 2e-05,
"loss": 1.3809,
"step": 252
},
{
"epoch": 1.2524752475247525,
"grad_norm": 0.4185222089290619,
"learning_rate": 2e-05,
"loss": 1.3717,
"step": 253
},
{
"epoch": 1.2574257425742574,
"grad_norm": 0.4185222089290619,
"learning_rate": 2e-05,
"loss": 1.3269,
"step": 254
},
{
"epoch": 1.2623762376237624,
"grad_norm": 0.4185222089290619,
"learning_rate": 2e-05,
"loss": 1.3118,
"step": 255
},
{
"epoch": 1.2673267326732673,
"grad_norm": 0.39412784576416016,
"learning_rate": 2e-05,
"loss": 1.3335,
"step": 256
},
{
"epoch": 1.2722772277227723,
"grad_norm": 0.39412784576416016,
"learning_rate": 2e-05,
"loss": 1.423,
"step": 257
},
{
"epoch": 1.2772277227722773,
"grad_norm": 0.39412784576416016,
"learning_rate": 2e-05,
"loss": 1.3877,
"step": 258
},
{
"epoch": 1.2821782178217822,
"grad_norm": 0.39412784576416016,
"learning_rate": 2e-05,
"loss": 1.3321,
"step": 259
},
{
"epoch": 1.2871287128712872,
"grad_norm": 0.4140811264514923,
"learning_rate": 2e-05,
"loss": 1.3841,
"step": 260
},
{
"epoch": 1.2920792079207921,
"grad_norm": 0.4140811264514923,
"learning_rate": 2e-05,
"loss": 1.2695,
"step": 261
},
{
"epoch": 1.297029702970297,
"grad_norm": 0.4140811264514923,
"learning_rate": 2e-05,
"loss": 1.3277,
"step": 262
},
{
"epoch": 1.301980198019802,
"grad_norm": 0.4140811264514923,
"learning_rate": 2e-05,
"loss": 1.3214,
"step": 263
},
{
"epoch": 1.306930693069307,
"grad_norm": 0.4246792495250702,
"learning_rate": 2e-05,
"loss": 1.3414,
"step": 264
},
{
"epoch": 1.311881188118812,
"grad_norm": 0.4246792495250702,
"learning_rate": 2e-05,
"loss": 1.3104,
"step": 265
},
{
"epoch": 1.316831683168317,
"grad_norm": 0.4246792495250702,
"learning_rate": 2e-05,
"loss": 1.3042,
"step": 266
},
{
"epoch": 1.3217821782178218,
"grad_norm": 0.4246792495250702,
"learning_rate": 2e-05,
"loss": 1.4332,
"step": 267
},
{
"epoch": 1.3267326732673268,
"grad_norm": 0.39743533730506897,
"learning_rate": 2e-05,
"loss": 1.3768,
"step": 268
},
{
"epoch": 1.3316831683168318,
"grad_norm": 0.39743533730506897,
"learning_rate": 2e-05,
"loss": 1.3958,
"step": 269
},
{
"epoch": 1.3366336633663367,
"grad_norm": 0.39743533730506897,
"learning_rate": 2e-05,
"loss": 1.3518,
"step": 270
},
{
"epoch": 1.3415841584158417,
"grad_norm": 0.39743533730506897,
"learning_rate": 2e-05,
"loss": 1.2891,
"step": 271
},
{
"epoch": 1.3465346534653464,
"grad_norm": 0.4067614674568176,
"learning_rate": 2e-05,
"loss": 1.3414,
"step": 272
},
{
"epoch": 1.3514851485148514,
"grad_norm": 0.4067614674568176,
"learning_rate": 2e-05,
"loss": 1.3099,
"step": 273
},
{
"epoch": 1.3564356435643563,
"grad_norm": 0.4067614674568176,
"learning_rate": 2e-05,
"loss": 1.34,
"step": 274
},
{
"epoch": 1.3613861386138613,
"grad_norm": 0.4067614674568176,
"learning_rate": 2e-05,
"loss": 1.2508,
"step": 275
},
{
"epoch": 1.3663366336633662,
"grad_norm": 0.39786049723625183,
"learning_rate": 2e-05,
"loss": 1.3004,
"step": 276
},
{
"epoch": 1.3712871287128712,
"grad_norm": 0.39786049723625183,
"learning_rate": 2e-05,
"loss": 1.2944,
"step": 277
},
{
"epoch": 1.3762376237623761,
"grad_norm": 0.39786049723625183,
"learning_rate": 2e-05,
"loss": 1.3135,
"step": 278
},
{
"epoch": 1.381188118811881,
"grad_norm": 0.39786049723625183,
"learning_rate": 2e-05,
"loss": 1.3066,
"step": 279
},
{
"epoch": 1.386138613861386,
"grad_norm": 0.41155368089675903,
"learning_rate": 2e-05,
"loss": 1.3806,
"step": 280
},
{
"epoch": 1.391089108910891,
"grad_norm": 0.41155368089675903,
"learning_rate": 2e-05,
"loss": 1.3244,
"step": 281
},
{
"epoch": 1.396039603960396,
"grad_norm": 0.41155368089675903,
"learning_rate": 2e-05,
"loss": 1.3038,
"step": 282
},
{
"epoch": 1.400990099009901,
"grad_norm": 0.41155368089675903,
"learning_rate": 2e-05,
"loss": 1.3682,
"step": 283
},
{
"epoch": 1.4059405940594059,
"grad_norm": 0.4219290018081665,
"learning_rate": 2e-05,
"loss": 1.3352,
"step": 284
},
{
"epoch": 1.4108910891089108,
"grad_norm": 0.4219290018081665,
"learning_rate": 2e-05,
"loss": 1.3082,
"step": 285
},
{
"epoch": 1.4158415841584158,
"grad_norm": 0.4219290018081665,
"learning_rate": 2e-05,
"loss": 1.3582,
"step": 286
},
{
"epoch": 1.4207920792079207,
"grad_norm": 0.4219290018081665,
"learning_rate": 2e-05,
"loss": 1.349,
"step": 287
},
{
"epoch": 1.4257425742574257,
"grad_norm": 0.4278065264225006,
"learning_rate": 2e-05,
"loss": 1.2147,
"step": 288
},
{
"epoch": 1.4306930693069306,
"grad_norm": 0.4278065264225006,
"learning_rate": 2e-05,
"loss": 1.3291,
"step": 289
},
{
"epoch": 1.4356435643564356,
"grad_norm": 0.4278065264225006,
"learning_rate": 2e-05,
"loss": 1.3212,
"step": 290
},
{
"epoch": 1.4405940594059405,
"grad_norm": 0.4278065264225006,
"learning_rate": 2e-05,
"loss": 1.3336,
"step": 291
},
{
"epoch": 1.4455445544554455,
"grad_norm": 0.4144158363342285,
"learning_rate": 2e-05,
"loss": 1.3113,
"step": 292
},
{
"epoch": 1.4504950495049505,
"grad_norm": 0.4144158363342285,
"learning_rate": 2e-05,
"loss": 1.3461,
"step": 293
},
{
"epoch": 1.4554455445544554,
"grad_norm": 0.4144158363342285,
"learning_rate": 2e-05,
"loss": 1.2883,
"step": 294
},
{
"epoch": 1.4603960396039604,
"grad_norm": 0.4144158363342285,
"learning_rate": 2e-05,
"loss": 1.3363,
"step": 295
},
{
"epoch": 1.4653465346534653,
"grad_norm": 0.41483309864997864,
"learning_rate": 2e-05,
"loss": 1.3176,
"step": 296
},
{
"epoch": 1.4702970297029703,
"grad_norm": 0.41483309864997864,
"learning_rate": 2e-05,
"loss": 1.3304,
"step": 297
},
{
"epoch": 1.4752475247524752,
"grad_norm": 0.41483309864997864,
"learning_rate": 2e-05,
"loss": 1.3265,
"step": 298
},
{
"epoch": 1.4801980198019802,
"grad_norm": 0.41483309864997864,
"learning_rate": 2e-05,
"loss": 1.3423,
"step": 299
},
{
"epoch": 1.4851485148514851,
"grad_norm": 0.4375357925891876,
"learning_rate": 2e-05,
"loss": 1.3133,
"step": 300
},
{
"epoch": 1.49009900990099,
"grad_norm": 0.4375357925891876,
"learning_rate": 2e-05,
"loss": 1.2606,
"step": 301
},
{
"epoch": 1.495049504950495,
"grad_norm": 0.4375357925891876,
"learning_rate": 2e-05,
"loss": 1.3678,
"step": 302
},
{
"epoch": 1.5,
"grad_norm": 0.4375357925891876,
"learning_rate": 2e-05,
"loss": 1.3469,
"step": 303
},
{
"epoch": 1.504950495049505,
"grad_norm": 0.41463586688041687,
"learning_rate": 2e-05,
"loss": 1.348,
"step": 304
},
{
"epoch": 1.50990099009901,
"grad_norm": 0.41463586688041687,
"learning_rate": 2e-05,
"loss": 1.3122,
"step": 305
},
{
"epoch": 1.5148514851485149,
"grad_norm": 0.41463586688041687,
"learning_rate": 2e-05,
"loss": 1.3219,
"step": 306
},
{
"epoch": 1.5198019801980198,
"grad_norm": 0.41463586688041687,
"learning_rate": 2e-05,
"loss": 1.3452,
"step": 307
},
{
"epoch": 1.5247524752475248,
"grad_norm": 0.41508471965789795,
"learning_rate": 2e-05,
"loss": 1.3678,
"step": 308
},
{
"epoch": 1.5297029702970297,
"grad_norm": 0.41508471965789795,
"learning_rate": 2e-05,
"loss": 1.2469,
"step": 309
},
{
"epoch": 1.5346534653465347,
"grad_norm": 0.41508471965789795,
"learning_rate": 2e-05,
"loss": 1.3395,
"step": 310
},
{
"epoch": 1.5396039603960396,
"grad_norm": 0.41508471965789795,
"learning_rate": 2e-05,
"loss": 1.3736,
"step": 311
},
{
"epoch": 1.5445544554455446,
"grad_norm": 0.40919822454452515,
"learning_rate": 2e-05,
"loss": 1.3368,
"step": 312
},
{
"epoch": 1.5495049504950495,
"grad_norm": 0.40919822454452515,
"learning_rate": 2e-05,
"loss": 1.2967,
"step": 313
},
{
"epoch": 1.5544554455445545,
"grad_norm": 0.40919822454452515,
"learning_rate": 2e-05,
"loss": 1.3359,
"step": 314
},
{
"epoch": 1.5594059405940595,
"grad_norm": 0.40919822454452515,
"learning_rate": 2e-05,
"loss": 1.2565,
"step": 315
},
{
"epoch": 1.5643564356435644,
"grad_norm": 0.4025939702987671,
"learning_rate": 2e-05,
"loss": 1.3794,
"step": 316
},
{
"epoch": 1.5693069306930694,
"grad_norm": 0.4025939702987671,
"learning_rate": 2e-05,
"loss": 1.3351,
"step": 317
},
{
"epoch": 1.5742574257425743,
"grad_norm": 0.4025939702987671,
"learning_rate": 2e-05,
"loss": 1.3199,
"step": 318
},
{
"epoch": 1.5792079207920793,
"grad_norm": 0.4025939702987671,
"learning_rate": 2e-05,
"loss": 1.2823,
"step": 319
},
{
"epoch": 1.5841584158415842,
"grad_norm": 0.3928684890270233,
"learning_rate": 2e-05,
"loss": 1.2881,
"step": 320
},
{
"epoch": 1.5891089108910892,
"grad_norm": 0.3928684890270233,
"learning_rate": 2e-05,
"loss": 1.3232,
"step": 321
},
{
"epoch": 1.5940594059405941,
"grad_norm": 0.3928684890270233,
"learning_rate": 2e-05,
"loss": 1.3568,
"step": 322
},
{
"epoch": 1.599009900990099,
"grad_norm": 0.3928684890270233,
"learning_rate": 2e-05,
"loss": 1.364,
"step": 323
},
{
"epoch": 1.603960396039604,
"grad_norm": 0.4311462342739105,
"learning_rate": 2e-05,
"loss": 1.2694,
"step": 324
},
{
"epoch": 1.608910891089109,
"grad_norm": 0.4311462342739105,
"learning_rate": 2e-05,
"loss": 1.2885,
"step": 325
},
{
"epoch": 1.613861386138614,
"grad_norm": 0.4311462342739105,
"learning_rate": 2e-05,
"loss": 1.3271,
"step": 326
},
{
"epoch": 1.618811881188119,
"grad_norm": 0.4311462342739105,
"learning_rate": 2e-05,
"loss": 1.2827,
"step": 327
},
{
"epoch": 1.6237623762376239,
"grad_norm": 0.43089422583580017,
"learning_rate": 2e-05,
"loss": 1.3469,
"step": 328
},
{
"epoch": 1.6287128712871288,
"grad_norm": 0.43089422583580017,
"learning_rate": 2e-05,
"loss": 1.3216,
"step": 329
},
{
"epoch": 1.6336633663366338,
"grad_norm": 0.43089422583580017,
"learning_rate": 2e-05,
"loss": 1.3085,
"step": 330
},
{
"epoch": 1.6386138613861387,
"grad_norm": 0.43089422583580017,
"learning_rate": 2e-05,
"loss": 1.2395,
"step": 331
},
{
"epoch": 1.6435643564356437,
"grad_norm": 0.4424372613430023,
"learning_rate": 2e-05,
"loss": 1.322,
"step": 332
},
{
"epoch": 1.6485148514851486,
"grad_norm": 0.4424372613430023,
"learning_rate": 2e-05,
"loss": 1.3302,
"step": 333
},
{
"epoch": 1.6534653465346536,
"grad_norm": 0.4424372613430023,
"learning_rate": 2e-05,
"loss": 1.2655,
"step": 334
},
{
"epoch": 1.6584158415841586,
"grad_norm": 0.4424372613430023,
"learning_rate": 2e-05,
"loss": 1.3223,
"step": 335
},
{
"epoch": 1.6633663366336635,
"grad_norm": 0.41536450386047363,
"learning_rate": 2e-05,
"loss": 1.2872,
"step": 336
},
{
"epoch": 1.6683168316831685,
"grad_norm": 0.41536450386047363,
"learning_rate": 2e-05,
"loss": 1.3872,
"step": 337
},
{
"epoch": 1.6732673267326734,
"grad_norm": 0.41536450386047363,
"learning_rate": 2e-05,
"loss": 1.2959,
"step": 338
},
{
"epoch": 1.6782178217821784,
"grad_norm": 0.41536450386047363,
"learning_rate": 2e-05,
"loss": 1.3195,
"step": 339
},
{
"epoch": 1.6831683168316833,
"grad_norm": 0.42268532514572144,
"learning_rate": 2e-05,
"loss": 1.3125,
"step": 340
},
{
"epoch": 1.688118811881188,
"grad_norm": 0.42268532514572144,
"learning_rate": 2e-05,
"loss": 1.2278,
"step": 341
},
{
"epoch": 1.693069306930693,
"grad_norm": 0.42268532514572144,
"learning_rate": 2e-05,
"loss": 1.2874,
"step": 342
},
{
"epoch": 1.698019801980198,
"grad_norm": 0.42268532514572144,
"learning_rate": 2e-05,
"loss": 1.2696,
"step": 343
},
{
"epoch": 1.702970297029703,
"grad_norm": 0.4271681010723114,
"learning_rate": 2e-05,
"loss": 1.3482,
"step": 344
},
{
"epoch": 1.7079207920792079,
"grad_norm": 0.4271681010723114,
"learning_rate": 2e-05,
"loss": 1.319,
"step": 345
},
{
"epoch": 1.7128712871287128,
"grad_norm": 0.4271681010723114,
"learning_rate": 2e-05,
"loss": 1.3505,
"step": 346
},
{
"epoch": 1.7178217821782178,
"grad_norm": 0.4271681010723114,
"learning_rate": 2e-05,
"loss": 1.2825,
"step": 347
},
{
"epoch": 1.7227722772277227,
"grad_norm": 0.438308984041214,
"learning_rate": 2e-05,
"loss": 1.3104,
"step": 348
},
{
"epoch": 1.7277227722772277,
"grad_norm": 0.438308984041214,
"learning_rate": 2e-05,
"loss": 1.3456,
"step": 349
},
{
"epoch": 1.7326732673267327,
"grad_norm": 0.438308984041214,
"learning_rate": 2e-05,
"loss": 1.3879,
"step": 350
},
{
"epoch": 1.7376237623762376,
"grad_norm": 0.438308984041214,
"learning_rate": 2e-05,
"loss": 1.2799,
"step": 351
},
{
"epoch": 1.7425742574257426,
"grad_norm": 0.3902476727962494,
"learning_rate": 2e-05,
"loss": 1.3175,
"step": 352
},
{
"epoch": 1.7475247524752475,
"grad_norm": 0.3902476727962494,
"learning_rate": 2e-05,
"loss": 1.337,
"step": 353
},
{
"epoch": 1.7524752475247525,
"grad_norm": 0.3902476727962494,
"learning_rate": 2e-05,
"loss": 1.3281,
"step": 354
},
{
"epoch": 1.7574257425742574,
"grad_norm": 0.3902476727962494,
"learning_rate": 2e-05,
"loss": 1.2937,
"step": 355
},
{
"epoch": 1.7623762376237624,
"grad_norm": 0.4025382101535797,
"learning_rate": 2e-05,
"loss": 1.3155,
"step": 356
},
{
"epoch": 1.7673267326732673,
"grad_norm": 0.4025382101535797,
"learning_rate": 2e-05,
"loss": 1.3147,
"step": 357
},
{
"epoch": 1.7722772277227723,
"grad_norm": 0.4025382101535797,
"learning_rate": 2e-05,
"loss": 1.3334,
"step": 358
},
{
"epoch": 1.7772277227722773,
"grad_norm": 0.4025382101535797,
"learning_rate": 2e-05,
"loss": 1.2924,
"step": 359
},
{
"epoch": 1.7821782178217822,
"grad_norm": 0.4225155711174011,
"learning_rate": 2e-05,
"loss": 1.2439,
"step": 360
},
{
"epoch": 1.7871287128712872,
"grad_norm": 0.4225155711174011,
"learning_rate": 2e-05,
"loss": 1.2926,
"step": 361
},
{
"epoch": 1.7920792079207921,
"grad_norm": 0.4225155711174011,
"learning_rate": 2e-05,
"loss": 1.2872,
"step": 362
},
{
"epoch": 1.797029702970297,
"grad_norm": 0.4225155711174011,
"learning_rate": 2e-05,
"loss": 1.2972,
"step": 363
},
{
"epoch": 1.801980198019802,
"grad_norm": 0.4173654615879059,
"learning_rate": 2e-05,
"loss": 1.3434,
"step": 364
},
{
"epoch": 1.806930693069307,
"grad_norm": 0.4173654615879059,
"learning_rate": 2e-05,
"loss": 1.3229,
"step": 365
},
{
"epoch": 1.811881188118812,
"grad_norm": 0.4173654615879059,
"learning_rate": 2e-05,
"loss": 1.315,
"step": 366
},
{
"epoch": 1.8168316831683167,
"grad_norm": 0.4173654615879059,
"learning_rate": 2e-05,
"loss": 1.3057,
"step": 367
},
{
"epoch": 1.8217821782178216,
"grad_norm": 0.4116384983062744,
"learning_rate": 2e-05,
"loss": 1.3509,
"step": 368
},
{
"epoch": 1.8267326732673266,
"grad_norm": 0.4116384983062744,
"learning_rate": 2e-05,
"loss": 1.2946,
"step": 369
},
{
"epoch": 1.8316831683168315,
"grad_norm": 0.4116384983062744,
"learning_rate": 2e-05,
"loss": 1.2919,
"step": 370
},
{
"epoch": 1.8366336633663365,
"grad_norm": 0.4116384983062744,
"learning_rate": 2e-05,
"loss": 1.3235,
"step": 371
},
{
"epoch": 1.8415841584158414,
"grad_norm": 0.41547831892967224,
"learning_rate": 2e-05,
"loss": 1.2645,
"step": 372
},
{
"epoch": 1.8465346534653464,
"grad_norm": 0.41547831892967224,
"learning_rate": 2e-05,
"loss": 1.3242,
"step": 373
},
{
"epoch": 1.8514851485148514,
"grad_norm": 0.41547831892967224,
"learning_rate": 2e-05,
"loss": 1.2963,
"step": 374
},
{
"epoch": 1.8564356435643563,
"grad_norm": 0.41547831892967224,
"learning_rate": 2e-05,
"loss": 1.2947,
"step": 375
},
{
"epoch": 1.8613861386138613,
"grad_norm": 0.4250006675720215,
"learning_rate": 2e-05,
"loss": 1.3529,
"step": 376
},
{
"epoch": 1.8663366336633662,
"grad_norm": 0.4250006675720215,
"learning_rate": 2e-05,
"loss": 1.3274,
"step": 377
},
{
"epoch": 1.8712871287128712,
"grad_norm": 0.4250006675720215,
"learning_rate": 2e-05,
"loss": 1.2877,
"step": 378
},
{
"epoch": 1.8762376237623761,
"grad_norm": 0.4250006675720215,
"learning_rate": 2e-05,
"loss": 1.3131,
"step": 379
},
{
"epoch": 1.881188118811881,
"grad_norm": 0.4326990842819214,
"learning_rate": 2e-05,
"loss": 1.2957,
"step": 380
},
{
"epoch": 1.886138613861386,
"grad_norm": 0.4326990842819214,
"learning_rate": 2e-05,
"loss": 1.3127,
"step": 381
},
{
"epoch": 1.891089108910891,
"grad_norm": 0.4326990842819214,
"learning_rate": 2e-05,
"loss": 1.3008,
"step": 382
},
{
"epoch": 1.896039603960396,
"grad_norm": 0.4326990842819214,
"learning_rate": 2e-05,
"loss": 1.2995,
"step": 383
},
{
"epoch": 1.900990099009901,
"grad_norm": 0.42278966307640076,
"learning_rate": 2e-05,
"loss": 1.3595,
"step": 384
},
{
"epoch": 1.9059405940594059,
"grad_norm": 0.42278966307640076,
"learning_rate": 2e-05,
"loss": 1.3013,
"step": 385
},
{
"epoch": 1.9108910891089108,
"grad_norm": 0.42278966307640076,
"learning_rate": 2e-05,
"loss": 1.3213,
"step": 386
},
{
"epoch": 1.9158415841584158,
"grad_norm": 0.42278966307640076,
"learning_rate": 2e-05,
"loss": 1.2492,
"step": 387
},
{
"epoch": 1.9207920792079207,
"grad_norm": 0.40483537316322327,
"learning_rate": 2e-05,
"loss": 1.2576,
"step": 388
},
{
"epoch": 1.9257425742574257,
"grad_norm": 0.40483537316322327,
"learning_rate": 2e-05,
"loss": 1.285,
"step": 389
},
{
"epoch": 1.9306930693069306,
"grad_norm": 0.40483537316322327,
"learning_rate": 2e-05,
"loss": 1.3289,
"step": 390
},
{
"epoch": 1.9356435643564356,
"grad_norm": 0.40483537316322327,
"learning_rate": 2e-05,
"loss": 1.286,
"step": 391
},
{
"epoch": 1.9405940594059405,
"grad_norm": 0.433361291885376,
"learning_rate": 2e-05,
"loss": 1.3486,
"step": 392
},
{
"epoch": 1.9455445544554455,
"grad_norm": 0.433361291885376,
"learning_rate": 2e-05,
"loss": 1.2726,
"step": 393
},
{
"epoch": 1.9504950495049505,
"grad_norm": 0.433361291885376,
"learning_rate": 2e-05,
"loss": 1.3015,
"step": 394
},
{
"epoch": 1.9554455445544554,
"grad_norm": 0.433361291885376,
"learning_rate": 2e-05,
"loss": 1.2891,
"step": 395
},
{
"epoch": 1.9603960396039604,
"grad_norm": 0.43291258811950684,
"learning_rate": 2e-05,
"loss": 1.3412,
"step": 396
},
{
"epoch": 1.9653465346534653,
"grad_norm": 0.43291258811950684,
"learning_rate": 2e-05,
"loss": 1.2877,
"step": 397
},
{
"epoch": 1.9702970297029703,
"grad_norm": 0.43291258811950684,
"learning_rate": 2e-05,
"loss": 1.3146,
"step": 398
},
{
"epoch": 1.9752475247524752,
"grad_norm": 0.43291258811950684,
"learning_rate": 2e-05,
"loss": 1.319,
"step": 399
},
{
"epoch": 1.9801980198019802,
"grad_norm": 0.4065687954425812,
"learning_rate": 2e-05,
"loss": 1.2858,
"step": 400
},
{
"epoch": 1.9851485148514851,
"grad_norm": 0.4065687954425812,
"learning_rate": 2e-05,
"loss": 1.3487,
"step": 401
},
{
"epoch": 1.99009900990099,
"grad_norm": 0.4065687954425812,
"learning_rate": 2e-05,
"loss": 1.2867,
"step": 402
},
{
"epoch": 1.995049504950495,
"grad_norm": 0.4065687954425812,
"learning_rate": 2e-05,
"loss": 1.3239,
"step": 403
},
{
"epoch": 2.0,
"grad_norm": 0.42932483553886414,
"learning_rate": 2e-05,
"loss": 1.1878,
"step": 404
},
{
"epoch": 2.004950495049505,
"grad_norm": 0.42932483553886414,
"learning_rate": 2e-05,
"loss": 1.2622,
"step": 405
},
{
"epoch": 2.00990099009901,
"grad_norm": 0.42932483553886414,
"learning_rate": 2e-05,
"loss": 1.1849,
"step": 406
},
{
"epoch": 2.014851485148515,
"grad_norm": 0.42932483553886414,
"learning_rate": 2e-05,
"loss": 1.2635,
"step": 407
},
{
"epoch": 2.01980198019802,
"grad_norm": 0.45234864950180054,
"learning_rate": 2e-05,
"loss": 1.2452,
"step": 408
},
{
"epoch": 2.0247524752475248,
"grad_norm": 0.45234864950180054,
"learning_rate": 2e-05,
"loss": 1.2034,
"step": 409
},
{
"epoch": 2.0297029702970297,
"grad_norm": 0.45234864950180054,
"learning_rate": 2e-05,
"loss": 1.2569,
"step": 410
},
{
"epoch": 2.0346534653465347,
"grad_norm": 0.45234864950180054,
"learning_rate": 2e-05,
"loss": 1.2518,
"step": 411
},
{
"epoch": 2.0396039603960396,
"grad_norm": 0.442615270614624,
"learning_rate": 2e-05,
"loss": 1.2477,
"step": 412
},
{
"epoch": 2.0445544554455446,
"grad_norm": 0.442615270614624,
"learning_rate": 2e-05,
"loss": 1.273,
"step": 413
},
{
"epoch": 2.0495049504950495,
"grad_norm": 0.442615270614624,
"learning_rate": 2e-05,
"loss": 1.2756,
"step": 414
},
{
"epoch": 2.0544554455445545,
"grad_norm": 0.442615270614624,
"learning_rate": 2e-05,
"loss": 1.2095,
"step": 415
},
{
"epoch": 2.0594059405940595,
"grad_norm": 0.45337530970573425,
"learning_rate": 2e-05,
"loss": 1.287,
"step": 416
},
{
"epoch": 2.0643564356435644,
"grad_norm": 0.45337530970573425,
"learning_rate": 2e-05,
"loss": 1.295,
"step": 417
},
{
"epoch": 2.0693069306930694,
"grad_norm": 0.45337530970573425,
"learning_rate": 2e-05,
"loss": 1.2793,
"step": 418
},
{
"epoch": 2.0742574257425743,
"grad_norm": 0.45337530970573425,
"learning_rate": 2e-05,
"loss": 1.2571,
"step": 419
},
{
"epoch": 2.0792079207920793,
"grad_norm": 0.42007654905319214,
"learning_rate": 2e-05,
"loss": 1.2914,
"step": 420
},
{
"epoch": 2.0841584158415842,
"grad_norm": 0.42007654905319214,
"learning_rate": 2e-05,
"loss": 1.257,
"step": 421
},
{
"epoch": 2.089108910891089,
"grad_norm": 0.42007654905319214,
"learning_rate": 2e-05,
"loss": 1.2842,
"step": 422
},
{
"epoch": 2.094059405940594,
"grad_norm": 0.42007654905319214,
"learning_rate": 2e-05,
"loss": 1.1936,
"step": 423
},
{
"epoch": 2.099009900990099,
"grad_norm": 0.4102495014667511,
"learning_rate": 2e-05,
"loss": 1.2335,
"step": 424
},
{
"epoch": 2.103960396039604,
"grad_norm": 0.4102495014667511,
"learning_rate": 2e-05,
"loss": 1.2233,
"step": 425
},
{
"epoch": 2.108910891089109,
"grad_norm": 0.4102495014667511,
"learning_rate": 2e-05,
"loss": 1.2249,
"step": 426
},
{
"epoch": 2.113861386138614,
"grad_norm": 0.4102495014667511,
"learning_rate": 2e-05,
"loss": 1.2346,
"step": 427
},
{
"epoch": 2.118811881188119,
"grad_norm": 0.44734662771224976,
"learning_rate": 2e-05,
"loss": 1.2588,
"step": 428
},
{
"epoch": 2.123762376237624,
"grad_norm": 0.44734662771224976,
"learning_rate": 2e-05,
"loss": 1.2399,
"step": 429
},
{
"epoch": 2.128712871287129,
"grad_norm": 0.44734662771224976,
"learning_rate": 2e-05,
"loss": 1.2648,
"step": 430
},
{
"epoch": 2.133663366336634,
"grad_norm": 0.44734662771224976,
"learning_rate": 2e-05,
"loss": 1.2579,
"step": 431
},
{
"epoch": 2.1386138613861387,
"grad_norm": 0.43320950865745544,
"learning_rate": 2e-05,
"loss": 1.2439,
"step": 432
},
{
"epoch": 2.1435643564356437,
"grad_norm": 0.43320950865745544,
"learning_rate": 2e-05,
"loss": 1.2623,
"step": 433
},
{
"epoch": 2.1485148514851486,
"grad_norm": 0.43320950865745544,
"learning_rate": 2e-05,
"loss": 1.2533,
"step": 434
},
{
"epoch": 2.1534653465346536,
"grad_norm": 0.43320950865745544,
"learning_rate": 2e-05,
"loss": 1.2041,
"step": 435
},
{
"epoch": 2.1584158415841586,
"grad_norm": 0.41689327359199524,
"learning_rate": 2e-05,
"loss": 1.2712,
"step": 436
},
{
"epoch": 2.1633663366336635,
"grad_norm": 0.41689327359199524,
"learning_rate": 2e-05,
"loss": 1.2137,
"step": 437
},
{
"epoch": 2.1683168316831685,
"grad_norm": 0.41689327359199524,
"learning_rate": 2e-05,
"loss": 1.2734,
"step": 438
},
{
"epoch": 2.1732673267326734,
"grad_norm": 0.41689327359199524,
"learning_rate": 2e-05,
"loss": 1.2999,
"step": 439
},
{
"epoch": 2.1782178217821784,
"grad_norm": 0.43515822291374207,
"learning_rate": 2e-05,
"loss": 1.2454,
"step": 440
},
{
"epoch": 2.1831683168316833,
"grad_norm": 0.43515822291374207,
"learning_rate": 2e-05,
"loss": 1.238,
"step": 441
},
{
"epoch": 2.1881188118811883,
"grad_norm": 0.43515822291374207,
"learning_rate": 2e-05,
"loss": 1.2441,
"step": 442
},
{
"epoch": 2.1930693069306932,
"grad_norm": 0.43515822291374207,
"learning_rate": 2e-05,
"loss": 1.2308,
"step": 443
},
{
"epoch": 2.198019801980198,
"grad_norm": 0.41985857486724854,
"learning_rate": 2e-05,
"loss": 1.2215,
"step": 444
},
{
"epoch": 2.202970297029703,
"grad_norm": 0.41985857486724854,
"learning_rate": 2e-05,
"loss": 1.2789,
"step": 445
},
{
"epoch": 2.207920792079208,
"grad_norm": 0.41985857486724854,
"learning_rate": 2e-05,
"loss": 1.2652,
"step": 446
},
{
"epoch": 2.212871287128713,
"grad_norm": 0.41985857486724854,
"learning_rate": 2e-05,
"loss": 1.2621,
"step": 447
},
{
"epoch": 2.217821782178218,
"grad_norm": 0.4170067608356476,
"learning_rate": 2e-05,
"loss": 1.2432,
"step": 448
},
{
"epoch": 2.222772277227723,
"grad_norm": 0.4170067608356476,
"learning_rate": 2e-05,
"loss": 1.1784,
"step": 449
},
{
"epoch": 2.227722772277228,
"grad_norm": 0.4170067608356476,
"learning_rate": 2e-05,
"loss": 1.1838,
"step": 450
},
{
"epoch": 2.232673267326733,
"grad_norm": 0.4170067608356476,
"learning_rate": 2e-05,
"loss": 1.2554,
"step": 451
},
{
"epoch": 2.237623762376238,
"grad_norm": 0.4227045476436615,
"learning_rate": 2e-05,
"loss": 1.1805,
"step": 452
},
{
"epoch": 2.2425742574257423,
"grad_norm": 0.4227045476436615,
"learning_rate": 2e-05,
"loss": 1.2748,
"step": 453
},
{
"epoch": 2.2475247524752477,
"grad_norm": 0.4227045476436615,
"learning_rate": 2e-05,
"loss": 1.2344,
"step": 454
},
{
"epoch": 2.2524752475247523,
"grad_norm": 0.4227045476436615,
"learning_rate": 2e-05,
"loss": 1.2355,
"step": 455
},
{
"epoch": 2.2574257425742577,
"grad_norm": 0.43470829725265503,
"learning_rate": 2e-05,
"loss": 1.2363,
"step": 456
},
{
"epoch": 2.262376237623762,
"grad_norm": 0.43470829725265503,
"learning_rate": 2e-05,
"loss": 1.265,
"step": 457
},
{
"epoch": 2.2673267326732676,
"grad_norm": 0.43470829725265503,
"learning_rate": 2e-05,
"loss": 1.2154,
"step": 458
},
{
"epoch": 2.272277227722772,
"grad_norm": 0.43470829725265503,
"learning_rate": 2e-05,
"loss": 1.1905,
"step": 459
},
{
"epoch": 2.2772277227722775,
"grad_norm": 0.4438628554344177,
"learning_rate": 2e-05,
"loss": 1.2347,
"step": 460
},
{
"epoch": 2.282178217821782,
"grad_norm": 0.4438628554344177,
"learning_rate": 2e-05,
"loss": 1.1932,
"step": 461
},
{
"epoch": 2.287128712871287,
"grad_norm": 0.4438628554344177,
"learning_rate": 2e-05,
"loss": 1.2726,
"step": 462
},
{
"epoch": 2.292079207920792,
"grad_norm": 0.4438628554344177,
"learning_rate": 2e-05,
"loss": 1.2243,
"step": 463
},
{
"epoch": 2.297029702970297,
"grad_norm": 0.42613375186920166,
"learning_rate": 2e-05,
"loss": 1.2906,
"step": 464
},
{
"epoch": 2.301980198019802,
"grad_norm": 0.42613375186920166,
"learning_rate": 2e-05,
"loss": 1.2565,
"step": 465
},
{
"epoch": 2.3069306930693068,
"grad_norm": 0.42613375186920166,
"learning_rate": 2e-05,
"loss": 1.2877,
"step": 466
},
{
"epoch": 2.3118811881188117,
"grad_norm": 0.42613375186920166,
"learning_rate": 2e-05,
"loss": 1.2683,
"step": 467
},
{
"epoch": 2.3168316831683167,
"grad_norm": 0.4208451509475708,
"learning_rate": 2e-05,
"loss": 1.243,
"step": 468
},
{
"epoch": 2.3217821782178216,
"grad_norm": 0.4208451509475708,
"learning_rate": 2e-05,
"loss": 1.2013,
"step": 469
},
{
"epoch": 2.3267326732673266,
"grad_norm": 0.4208451509475708,
"learning_rate": 2e-05,
"loss": 1.2882,
"step": 470
},
{
"epoch": 2.3316831683168315,
"grad_norm": 0.4208451509475708,
"learning_rate": 2e-05,
"loss": 1.2326,
"step": 471
},
{
"epoch": 2.3366336633663365,
"grad_norm": 0.4349665343761444,
"learning_rate": 2e-05,
"loss": 1.266,
"step": 472
},
{
"epoch": 2.3415841584158414,
"grad_norm": 0.4349665343761444,
"learning_rate": 2e-05,
"loss": 1.2149,
"step": 473
},
{
"epoch": 2.3465346534653464,
"grad_norm": 0.4349665343761444,
"learning_rate": 2e-05,
"loss": 1.2356,
"step": 474
},
{
"epoch": 2.3514851485148514,
"grad_norm": 0.4349665343761444,
"learning_rate": 2e-05,
"loss": 1.2391,
"step": 475
},
{
"epoch": 2.3564356435643563,
"grad_norm": 0.4321076571941376,
"learning_rate": 2e-05,
"loss": 1.2513,
"step": 476
},
{
"epoch": 2.3613861386138613,
"grad_norm": 0.4321076571941376,
"learning_rate": 2e-05,
"loss": 1.269,
"step": 477
},
{
"epoch": 2.366336633663366,
"grad_norm": 0.4321076571941376,
"learning_rate": 2e-05,
"loss": 1.2416,
"step": 478
},
{
"epoch": 2.371287128712871,
"grad_norm": 0.4321076571941376,
"learning_rate": 2e-05,
"loss": 1.2774,
"step": 479
},
{
"epoch": 2.376237623762376,
"grad_norm": 0.42193785309791565,
"learning_rate": 2e-05,
"loss": 1.2511,
"step": 480
},
{
"epoch": 2.381188118811881,
"grad_norm": 0.42193785309791565,
"learning_rate": 2e-05,
"loss": 1.239,
"step": 481
},
{
"epoch": 2.386138613861386,
"grad_norm": 0.42193785309791565,
"learning_rate": 2e-05,
"loss": 1.2129,
"step": 482
},
{
"epoch": 2.391089108910891,
"grad_norm": 0.42193785309791565,
"learning_rate": 2e-05,
"loss": 1.2611,
"step": 483
},
{
"epoch": 2.396039603960396,
"grad_norm": 0.43613630533218384,
"learning_rate": 2e-05,
"loss": 1.2512,
"step": 484
},
{
"epoch": 2.400990099009901,
"grad_norm": 0.43613630533218384,
"learning_rate": 2e-05,
"loss": 1.2196,
"step": 485
},
{
"epoch": 2.405940594059406,
"grad_norm": 0.43613630533218384,
"learning_rate": 2e-05,
"loss": 1.1889,
"step": 486
},
{
"epoch": 2.410891089108911,
"grad_norm": 0.43613630533218384,
"learning_rate": 2e-05,
"loss": 1.2203,
"step": 487
},
{
"epoch": 2.4158415841584158,
"grad_norm": 0.4264240562915802,
"learning_rate": 2e-05,
"loss": 1.2234,
"step": 488
},
{
"epoch": 2.4207920792079207,
"grad_norm": 0.4264240562915802,
"learning_rate": 2e-05,
"loss": 1.1956,
"step": 489
},
{
"epoch": 2.4257425742574257,
"grad_norm": 0.4264240562915802,
"learning_rate": 2e-05,
"loss": 1.2189,
"step": 490
},
{
"epoch": 2.4306930693069306,
"grad_norm": 0.4264240562915802,
"learning_rate": 2e-05,
"loss": 1.2528,
"step": 491
},
{
"epoch": 2.4356435643564356,
"grad_norm": 0.44283366203308105,
"learning_rate": 2e-05,
"loss": 1.2116,
"step": 492
},
{
"epoch": 2.4405940594059405,
"grad_norm": 0.44283366203308105,
"learning_rate": 2e-05,
"loss": 1.1898,
"step": 493
},
{
"epoch": 2.4455445544554455,
"grad_norm": 0.44283366203308105,
"learning_rate": 2e-05,
"loss": 1.2933,
"step": 494
},
{
"epoch": 2.4504950495049505,
"grad_norm": 0.44283366203308105,
"learning_rate": 2e-05,
"loss": 1.1879,
"step": 495
},
{
"epoch": 2.4554455445544554,
"grad_norm": 0.4315781593322754,
"learning_rate": 2e-05,
"loss": 1.1529,
"step": 496
},
{
"epoch": 2.4603960396039604,
"grad_norm": 0.4315781593322754,
"learning_rate": 2e-05,
"loss": 1.3047,
"step": 497
},
{
"epoch": 2.4653465346534653,
"grad_norm": 0.4315781593322754,
"learning_rate": 2e-05,
"loss": 1.222,
"step": 498
},
{
"epoch": 2.4702970297029703,
"grad_norm": 0.4315781593322754,
"learning_rate": 2e-05,
"loss": 1.2699,
"step": 499
},
{
"epoch": 2.4752475247524752,
"grad_norm": 0.441769540309906,
"learning_rate": 2e-05,
"loss": 1.2127,
"step": 500
},
{
"epoch": 2.48019801980198,
"grad_norm": 0.441769540309906,
"learning_rate": 2e-05,
"loss": 1.223,
"step": 501
},
{
"epoch": 2.485148514851485,
"grad_norm": 0.441769540309906,
"learning_rate": 2e-05,
"loss": 1.2389,
"step": 502
},
{
"epoch": 2.49009900990099,
"grad_norm": 0.441769540309906,
"learning_rate": 2e-05,
"loss": 1.1949,
"step": 503
},
{
"epoch": 2.495049504950495,
"grad_norm": 0.40025854110717773,
"learning_rate": 2e-05,
"loss": 1.2149,
"step": 504
},
{
"epoch": 2.5,
"grad_norm": 0.40025854110717773,
"learning_rate": 2e-05,
"loss": 1.2519,
"step": 505
},
{
"epoch": 2.504950495049505,
"grad_norm": 0.40025854110717773,
"learning_rate": 2e-05,
"loss": 1.2644,
"step": 506
},
{
"epoch": 2.50990099009901,
"grad_norm": 0.40025854110717773,
"learning_rate": 2e-05,
"loss": 1.2348,
"step": 507
},
{
"epoch": 2.514851485148515,
"grad_norm": 0.409300297498703,
"learning_rate": 2e-05,
"loss": 1.1972,
"step": 508
},
{
"epoch": 2.51980198019802,
"grad_norm": 0.409300297498703,
"learning_rate": 2e-05,
"loss": 1.2407,
"step": 509
},
{
"epoch": 2.5247524752475248,
"grad_norm": 0.409300297498703,
"learning_rate": 2e-05,
"loss": 1.2038,
"step": 510
},
{
"epoch": 2.5297029702970297,
"grad_norm": 0.409300297498703,
"learning_rate": 2e-05,
"loss": 1.2143,
"step": 511
},
{
"epoch": 2.5346534653465347,
"grad_norm": 0.4438338875770569,
"learning_rate": 2e-05,
"loss": 1.2121,
"step": 512
},
{
"epoch": 2.5396039603960396,
"grad_norm": 0.4438338875770569,
"learning_rate": 2e-05,
"loss": 1.1522,
"step": 513
},
{
"epoch": 2.5445544554455446,
"grad_norm": 0.4438338875770569,
"learning_rate": 2e-05,
"loss": 1.1879,
"step": 514
},
{
"epoch": 2.5495049504950495,
"grad_norm": 0.4438338875770569,
"learning_rate": 2e-05,
"loss": 1.2499,
"step": 515
},
{
"epoch": 2.5544554455445545,
"grad_norm": 0.45127996802330017,
"learning_rate": 2e-05,
"loss": 1.2534,
"step": 516
},
{
"epoch": 2.5594059405940595,
"grad_norm": 0.45127996802330017,
"learning_rate": 2e-05,
"loss": 1.2211,
"step": 517
},
{
"epoch": 2.5643564356435644,
"grad_norm": 0.45127996802330017,
"learning_rate": 2e-05,
"loss": 1.1961,
"step": 518
},
{
"epoch": 2.5693069306930694,
"grad_norm": 0.45127996802330017,
"learning_rate": 2e-05,
"loss": 1.1957,
"step": 519
},
{
"epoch": 2.5742574257425743,
"grad_norm": 0.4387412965297699,
"learning_rate": 2e-05,
"loss": 1.2454,
"step": 520
},
{
"epoch": 2.5792079207920793,
"grad_norm": 0.4387412965297699,
"learning_rate": 2e-05,
"loss": 1.2698,
"step": 521
},
{
"epoch": 2.5841584158415842,
"grad_norm": 0.4387412965297699,
"learning_rate": 2e-05,
"loss": 1.2219,
"step": 522
},
{
"epoch": 2.589108910891089,
"grad_norm": 0.4387412965297699,
"learning_rate": 2e-05,
"loss": 1.196,
"step": 523
},
{
"epoch": 2.594059405940594,
"grad_norm": 0.41671085357666016,
"learning_rate": 2e-05,
"loss": 1.2303,
"step": 524
},
{
"epoch": 2.599009900990099,
"grad_norm": 0.41671085357666016,
"learning_rate": 2e-05,
"loss": 1.2923,
"step": 525
},
{
"epoch": 2.603960396039604,
"grad_norm": 0.41671085357666016,
"learning_rate": 2e-05,
"loss": 1.2663,
"step": 526
},
{
"epoch": 2.608910891089109,
"grad_norm": 0.41671085357666016,
"learning_rate": 2e-05,
"loss": 1.1659,
"step": 527
},
{
"epoch": 2.613861386138614,
"grad_norm": 0.42053425312042236,
"learning_rate": 2e-05,
"loss": 1.164,
"step": 528
},
{
"epoch": 2.618811881188119,
"grad_norm": 0.42053425312042236,
"learning_rate": 2e-05,
"loss": 1.1659,
"step": 529
},
{
"epoch": 2.623762376237624,
"grad_norm": 0.42053425312042236,
"learning_rate": 2e-05,
"loss": 1.1622,
"step": 530
},
{
"epoch": 2.628712871287129,
"grad_norm": 0.42053425312042236,
"learning_rate": 2e-05,
"loss": 1.193,
"step": 531
},
{
"epoch": 2.633663366336634,
"grad_norm": 0.4210812449455261,
"learning_rate": 2e-05,
"loss": 1.2281,
"step": 532
},
{
"epoch": 2.6386138613861387,
"grad_norm": 0.4210812449455261,
"learning_rate": 2e-05,
"loss": 1.2056,
"step": 533
},
{
"epoch": 2.6435643564356437,
"grad_norm": 0.4210812449455261,
"learning_rate": 2e-05,
"loss": 1.1654,
"step": 534
},
{
"epoch": 2.6485148514851486,
"grad_norm": 0.4210812449455261,
"learning_rate": 2e-05,
"loss": 1.232,
"step": 535
},
{
"epoch": 2.6534653465346536,
"grad_norm": 0.4091170132160187,
"learning_rate": 2e-05,
"loss": 1.1129,
"step": 536
},
{
"epoch": 2.6584158415841586,
"grad_norm": 0.4091170132160187,
"learning_rate": 2e-05,
"loss": 1.1746,
"step": 537
},
{
"epoch": 2.6633663366336635,
"grad_norm": 0.4091170132160187,
"learning_rate": 2e-05,
"loss": 1.2418,
"step": 538
},
{
"epoch": 2.6683168316831685,
"grad_norm": 0.4091170132160187,
"learning_rate": 2e-05,
"loss": 1.2067,
"step": 539
},
{
"epoch": 2.6732673267326734,
"grad_norm": 0.40510034561157227,
"learning_rate": 2e-05,
"loss": 1.2064,
"step": 540
},
{
"epoch": 2.6782178217821784,
"grad_norm": 0.40510034561157227,
"learning_rate": 2e-05,
"loss": 1.2531,
"step": 541
},
{
"epoch": 2.6831683168316833,
"grad_norm": 0.40510034561157227,
"learning_rate": 2e-05,
"loss": 1.2307,
"step": 542
},
{
"epoch": 2.6881188118811883,
"grad_norm": 0.40510034561157227,
"learning_rate": 2e-05,
"loss": 1.1908,
"step": 543
},
{
"epoch": 2.693069306930693,
"grad_norm": 0.41868939995765686,
"learning_rate": 2e-05,
"loss": 1.2277,
"step": 544
},
{
"epoch": 2.698019801980198,
"grad_norm": 0.41868939995765686,
"learning_rate": 2e-05,
"loss": 1.225,
"step": 545
},
{
"epoch": 2.7029702970297027,
"grad_norm": 0.41868939995765686,
"learning_rate": 2e-05,
"loss": 1.266,
"step": 546
},
{
"epoch": 2.707920792079208,
"grad_norm": 0.41868939995765686,
"learning_rate": 2e-05,
"loss": 1.2184,
"step": 547
},
{
"epoch": 2.7128712871287126,
"grad_norm": 0.4461022615432739,
"learning_rate": 2e-05,
"loss": 1.2893,
"step": 548
},
{
"epoch": 2.717821782178218,
"grad_norm": 0.4461022615432739,
"learning_rate": 2e-05,
"loss": 1.182,
"step": 549
},
{
"epoch": 2.7227722772277225,
"grad_norm": 0.4461022615432739,
"learning_rate": 2e-05,
"loss": 1.2444,
"step": 550
},
{
"epoch": 2.727722772277228,
"grad_norm": 0.4461022615432739,
"learning_rate": 2e-05,
"loss": 1.2448,
"step": 551
},
{
"epoch": 2.7326732673267324,
"grad_norm": 0.4156385362148285,
"learning_rate": 2e-05,
"loss": 1.3042,
"step": 552
},
{
"epoch": 2.737623762376238,
"grad_norm": 0.4156385362148285,
"learning_rate": 2e-05,
"loss": 1.217,
"step": 553
},
{
"epoch": 2.7425742574257423,
"grad_norm": 0.4156385362148285,
"learning_rate": 2e-05,
"loss": 1.2732,
"step": 554
},
{
"epoch": 2.7475247524752477,
"grad_norm": 0.4156385362148285,
"learning_rate": 2e-05,
"loss": 1.182,
"step": 555
},
{
"epoch": 2.7524752475247523,
"grad_norm": 0.4390084147453308,
"learning_rate": 2e-05,
"loss": 1.2262,
"step": 556
},
{
"epoch": 2.7574257425742577,
"grad_norm": 0.4390084147453308,
"learning_rate": 2e-05,
"loss": 1.2321,
"step": 557
},
{
"epoch": 2.762376237623762,
"grad_norm": 0.4390084147453308,
"learning_rate": 2e-05,
"loss": 1.1932,
"step": 558
},
{
"epoch": 2.7673267326732676,
"grad_norm": 0.4390084147453308,
"learning_rate": 2e-05,
"loss": 1.1887,
"step": 559
},
{
"epoch": 2.772277227722772,
"grad_norm": 0.4339062571525574,
"learning_rate": 2e-05,
"loss": 1.1702,
"step": 560
},
{
"epoch": 2.7772277227722775,
"grad_norm": 0.4339062571525574,
"learning_rate": 2e-05,
"loss": 1.2032,
"step": 561
},
{
"epoch": 2.782178217821782,
"grad_norm": 0.4339062571525574,
"learning_rate": 2e-05,
"loss": 1.2205,
"step": 562
},
{
"epoch": 2.7871287128712874,
"grad_norm": 0.4339062571525574,
"learning_rate": 2e-05,
"loss": 1.1772,
"step": 563
},
{
"epoch": 2.792079207920792,
"grad_norm": 0.4191722869873047,
"learning_rate": 2e-05,
"loss": 1.2586,
"step": 564
},
{
"epoch": 2.7970297029702973,
"grad_norm": 0.4191722869873047,
"learning_rate": 2e-05,
"loss": 1.2074,
"step": 565
},
{
"epoch": 2.801980198019802,
"grad_norm": 0.4191722869873047,
"learning_rate": 2e-05,
"loss": 1.2108,
"step": 566
},
{
"epoch": 2.806930693069307,
"grad_norm": 0.4191722869873047,
"learning_rate": 2e-05,
"loss": 1.2453,
"step": 567
},
{
"epoch": 2.8118811881188117,
"grad_norm": 0.4224371612071991,
"learning_rate": 2e-05,
"loss": 1.2207,
"step": 568
},
{
"epoch": 2.8168316831683167,
"grad_norm": 0.4224371612071991,
"learning_rate": 2e-05,
"loss": 1.2493,
"step": 569
},
{
"epoch": 2.8217821782178216,
"grad_norm": 0.4224371612071991,
"learning_rate": 2e-05,
"loss": 1.2622,
"step": 570
},
{
"epoch": 2.8267326732673266,
"grad_norm": 0.4224371612071991,
"learning_rate": 2e-05,
"loss": 1.2023,
"step": 571
},
{
"epoch": 2.8316831683168315,
"grad_norm": 0.44747769832611084,
"learning_rate": 2e-05,
"loss": 1.2241,
"step": 572
},
{
"epoch": 2.8366336633663365,
"grad_norm": 0.44747769832611084,
"learning_rate": 2e-05,
"loss": 1.2751,
"step": 573
},
{
"epoch": 2.8415841584158414,
"grad_norm": 0.44747769832611084,
"learning_rate": 2e-05,
"loss": 1.2416,
"step": 574
},
{
"epoch": 2.8465346534653464,
"grad_norm": 0.44747769832611084,
"learning_rate": 2e-05,
"loss": 1.192,
"step": 575
},
{
"epoch": 2.8514851485148514,
"grad_norm": 0.4167119562625885,
"learning_rate": 2e-05,
"loss": 1.1984,
"step": 576
},
{
"epoch": 2.8564356435643563,
"grad_norm": 0.4167119562625885,
"learning_rate": 2e-05,
"loss": 1.1947,
"step": 577
},
{
"epoch": 2.8613861386138613,
"grad_norm": 0.4167119562625885,
"learning_rate": 2e-05,
"loss": 1.2046,
"step": 578
},
{
"epoch": 2.866336633663366,
"grad_norm": 0.4167119562625885,
"learning_rate": 2e-05,
"loss": 1.2331,
"step": 579
},
{
"epoch": 2.871287128712871,
"grad_norm": 0.4322595000267029,
"learning_rate": 2e-05,
"loss": 1.231,
"step": 580
},
{
"epoch": 2.876237623762376,
"grad_norm": 0.4322595000267029,
"learning_rate": 2e-05,
"loss": 1.2479,
"step": 581
},
{
"epoch": 2.881188118811881,
"grad_norm": 0.4322595000267029,
"learning_rate": 2e-05,
"loss": 1.2034,
"step": 582
},
{
"epoch": 2.886138613861386,
"grad_norm": 0.4322595000267029,
"learning_rate": 2e-05,
"loss": 1.2375,
"step": 583
},
{
"epoch": 2.891089108910891,
"grad_norm": 0.43500980734825134,
"learning_rate": 2e-05,
"loss": 1.2637,
"step": 584
},
{
"epoch": 2.896039603960396,
"grad_norm": 0.43500980734825134,
"learning_rate": 2e-05,
"loss": 1.2457,
"step": 585
},
{
"epoch": 2.900990099009901,
"grad_norm": 0.43500980734825134,
"learning_rate": 2e-05,
"loss": 1.1687,
"step": 586
},
{
"epoch": 2.905940594059406,
"grad_norm": 0.43500980734825134,
"learning_rate": 2e-05,
"loss": 1.2414,
"step": 587
},
{
"epoch": 2.910891089108911,
"grad_norm": 0.4270128905773163,
"learning_rate": 2e-05,
"loss": 1.21,
"step": 588
},
{
"epoch": 2.9158415841584158,
"grad_norm": 0.4270128905773163,
"learning_rate": 2e-05,
"loss": 1.1977,
"step": 589
},
{
"epoch": 2.9207920792079207,
"grad_norm": 0.4270128905773163,
"learning_rate": 2e-05,
"loss": 1.2328,
"step": 590
},
{
"epoch": 2.9257425742574257,
"grad_norm": 0.4270128905773163,
"learning_rate": 2e-05,
"loss": 1.2134,
"step": 591
},
{
"epoch": 2.9306930693069306,
"grad_norm": 0.46624571084976196,
"learning_rate": 2e-05,
"loss": 1.2181,
"step": 592
},
{
"epoch": 2.9356435643564356,
"grad_norm": 0.46624571084976196,
"learning_rate": 2e-05,
"loss": 1.1702,
"step": 593
},
{
"epoch": 2.9405940594059405,
"grad_norm": 0.46624571084976196,
"learning_rate": 2e-05,
"loss": 1.2488,
"step": 594
},
{
"epoch": 2.9455445544554455,
"grad_norm": 0.46624571084976196,
"learning_rate": 2e-05,
"loss": 1.2098,
"step": 595
},
{
"epoch": 2.9504950495049505,
"grad_norm": 0.4665633738040924,
"learning_rate": 2e-05,
"loss": 1.2354,
"step": 596
},
{
"epoch": 2.9554455445544554,
"grad_norm": 0.4665633738040924,
"learning_rate": 2e-05,
"loss": 1.2353,
"step": 597
},
{
"epoch": 2.9603960396039604,
"grad_norm": 0.4665633738040924,
"learning_rate": 2e-05,
"loss": 1.2094,
"step": 598
},
{
"epoch": 2.9653465346534653,
"grad_norm": 0.4665633738040924,
"learning_rate": 2e-05,
"loss": 1.2296,
"step": 599
},
{
"epoch": 2.9702970297029703,
"grad_norm": 0.42361822724342346,
"learning_rate": 2e-05,
"loss": 1.1949,
"step": 600
},
{
"epoch": 2.9752475247524752,
"grad_norm": 0.42361822724342346,
"learning_rate": 2e-05,
"loss": 1.2012,
"step": 601
},
{
"epoch": 2.98019801980198,
"grad_norm": 0.42361822724342346,
"learning_rate": 2e-05,
"loss": 1.2655,
"step": 602
},
{
"epoch": 2.985148514851485,
"grad_norm": 0.42361822724342346,
"learning_rate": 2e-05,
"loss": 1.1577,
"step": 603
},
{
"epoch": 2.99009900990099,
"grad_norm": 0.464103639125824,
"learning_rate": 2e-05,
"loss": 1.2296,
"step": 604
},
{
"epoch": 2.995049504950495,
"grad_norm": 0.464103639125824,
"learning_rate": 2e-05,
"loss": 1.2255,
"step": 605
},
{
"epoch": 3.0,
"grad_norm": 0.464103639125824,
"learning_rate": 2e-05,
"loss": 1.1357,
"step": 606
},
{
"epoch": 3.0,
"step": 606,
"total_flos": 3.877969000755364e+17,
"train_loss": 1.3801187025045012,
"train_runtime": 382.7506,
"train_samples_per_second": 403.508,
"train_steps_per_second": 1.583
}
],
"logging_steps": 1.0,
"max_steps": 606,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 3000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 3.877969000755364e+17,
"train_batch_size": 32,
"trial_name": null,
"trial_params": null
}