DeQA-Score-Mix3 / trainer_state.json
zhiyuanyou
add script files
d4dad89
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"eval_steps": 500,
"global_step": 567,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.01,
"learning_rate": 1.111111111111111e-06,
"loss": 0.4065,
"step": 1
},
{
"epoch": 0.01,
"learning_rate": 2.222222222222222e-06,
"loss": 0.4126,
"step": 2
},
{
"epoch": 0.02,
"learning_rate": 3.3333333333333333e-06,
"loss": 0.406,
"step": 3
},
{
"epoch": 0.02,
"learning_rate": 4.444444444444444e-06,
"loss": 0.3757,
"step": 4
},
{
"epoch": 0.03,
"learning_rate": 5.555555555555557e-06,
"loss": 0.2993,
"step": 5
},
{
"epoch": 0.03,
"learning_rate": 6.666666666666667e-06,
"loss": 0.2178,
"step": 6
},
{
"epoch": 0.04,
"learning_rate": 7.77777777777778e-06,
"loss": 0.1332,
"step": 7
},
{
"epoch": 0.04,
"learning_rate": 8.888888888888888e-06,
"loss": 0.1436,
"step": 8
},
{
"epoch": 0.05,
"learning_rate": 1e-05,
"loss": 0.1261,
"step": 9
},
{
"epoch": 0.05,
"learning_rate": 1.1111111111111113e-05,
"loss": 0.1156,
"step": 10
},
{
"epoch": 0.06,
"learning_rate": 1.2222222222222224e-05,
"loss": 0.0961,
"step": 11
},
{
"epoch": 0.06,
"learning_rate": 1.3333333333333333e-05,
"loss": 0.0911,
"step": 12
},
{
"epoch": 0.07,
"learning_rate": 1.4444444444444446e-05,
"loss": 0.0706,
"step": 13
},
{
"epoch": 0.07,
"learning_rate": 1.555555555555556e-05,
"loss": 0.1345,
"step": 14
},
{
"epoch": 0.08,
"learning_rate": 1.6666666666666667e-05,
"loss": 0.0871,
"step": 15
},
{
"epoch": 0.08,
"learning_rate": 1.7777777777777777e-05,
"loss": 0.0744,
"step": 16
},
{
"epoch": 0.09,
"learning_rate": 1.888888888888889e-05,
"loss": 0.1035,
"step": 17
},
{
"epoch": 0.1,
"learning_rate": 2e-05,
"loss": 0.0768,
"step": 18
},
{
"epoch": 0.1,
"learning_rate": 1.9999836271653566e-05,
"loss": 0.0895,
"step": 19
},
{
"epoch": 0.11,
"learning_rate": 1.9999345091975652e-05,
"loss": 0.0998,
"step": 20
},
{
"epoch": 0.11,
"learning_rate": 1.999852647705027e-05,
"loss": 0.1071,
"step": 21
},
{
"epoch": 0.12,
"learning_rate": 1.9997380453683513e-05,
"loss": 0.0921,
"step": 22
},
{
"epoch": 0.12,
"learning_rate": 1.999590705940268e-05,
"loss": 0.0948,
"step": 23
},
{
"epoch": 0.13,
"learning_rate": 1.9994106342455053e-05,
"loss": 0.095,
"step": 24
},
{
"epoch": 0.13,
"learning_rate": 1.999197836180632e-05,
"loss": 0.0547,
"step": 25
},
{
"epoch": 0.14,
"learning_rate": 1.9989523187138623e-05,
"loss": 0.0719,
"step": 26
},
{
"epoch": 0.14,
"learning_rate": 1.9986740898848306e-05,
"loss": 0.078,
"step": 27
},
{
"epoch": 0.15,
"learning_rate": 1.998363158804326e-05,
"loss": 0.0674,
"step": 28
},
{
"epoch": 0.15,
"learning_rate": 1.9980195356539944e-05,
"loss": 0.0602,
"step": 29
},
{
"epoch": 0.16,
"learning_rate": 1.9976432316860065e-05,
"loss": 0.055,
"step": 30
},
{
"epoch": 0.16,
"learning_rate": 1.9972342592226873e-05,
"loss": 0.0587,
"step": 31
},
{
"epoch": 0.17,
"learning_rate": 1.9967926316561136e-05,
"loss": 0.0601,
"step": 32
},
{
"epoch": 0.17,
"learning_rate": 1.9963183634476757e-05,
"loss": 0.0464,
"step": 33
},
{
"epoch": 0.18,
"learning_rate": 1.995811470127604e-05,
"loss": 0.0599,
"step": 34
},
{
"epoch": 0.19,
"learning_rate": 1.9952719682944588e-05,
"loss": 0.0483,
"step": 35
},
{
"epoch": 0.19,
"learning_rate": 1.9946998756145894e-05,
"loss": 0.0439,
"step": 36
},
{
"epoch": 0.2,
"learning_rate": 1.9940952108215526e-05,
"loss": 0.0474,
"step": 37
},
{
"epoch": 0.2,
"learning_rate": 1.993457993715503e-05,
"loss": 0.0598,
"step": 38
},
{
"epoch": 0.21,
"learning_rate": 1.99278824516254e-05,
"loss": 0.0459,
"step": 39
},
{
"epoch": 0.21,
"learning_rate": 1.9920859870940292e-05,
"loss": 0.0493,
"step": 40
},
{
"epoch": 0.22,
"learning_rate": 1.9913512425058803e-05,
"loss": 0.0404,
"step": 41
},
{
"epoch": 0.22,
"learning_rate": 1.990584035457797e-05,
"loss": 0.0405,
"step": 42
},
{
"epoch": 0.23,
"learning_rate": 1.9897843910724877e-05,
"loss": 0.0479,
"step": 43
},
{
"epoch": 0.23,
"learning_rate": 1.9889523355348427e-05,
"loss": 0.0446,
"step": 44
},
{
"epoch": 0.24,
"learning_rate": 1.9880878960910772e-05,
"loss": 0.0359,
"step": 45
},
{
"epoch": 0.24,
"learning_rate": 1.98719110104784e-05,
"loss": 0.0374,
"step": 46
},
{
"epoch": 0.25,
"learning_rate": 1.9862619797712845e-05,
"loss": 0.0383,
"step": 47
},
{
"epoch": 0.25,
"learning_rate": 1.985300562686109e-05,
"loss": 0.0384,
"step": 48
},
{
"epoch": 0.26,
"learning_rate": 1.9843068812745595e-05,
"loss": 0.0403,
"step": 49
},
{
"epoch": 0.26,
"learning_rate": 1.9832809680753985e-05,
"loss": 0.0313,
"step": 50
},
{
"epoch": 0.27,
"learning_rate": 1.982222856682841e-05,
"loss": 0.0382,
"step": 51
},
{
"epoch": 0.28,
"learning_rate": 1.9811325817454516e-05,
"loss": 0.0397,
"step": 52
},
{
"epoch": 0.28,
"learning_rate": 1.980010178965014e-05,
"loss": 0.0405,
"step": 53
},
{
"epoch": 0.29,
"learning_rate": 1.978855685095358e-05,
"loss": 0.0396,
"step": 54
},
{
"epoch": 0.29,
"learning_rate": 1.977669137941158e-05,
"loss": 0.0333,
"step": 55
},
{
"epoch": 0.3,
"learning_rate": 1.9764505763566945e-05,
"loss": 0.0414,
"step": 56
},
{
"epoch": 0.3,
"learning_rate": 1.9752000402445824e-05,
"loss": 0.0384,
"step": 57
},
{
"epoch": 0.31,
"learning_rate": 1.973917570554464e-05,
"loss": 0.0419,
"step": 58
},
{
"epoch": 0.31,
"learning_rate": 1.9726032092816672e-05,
"loss": 0.0366,
"step": 59
},
{
"epoch": 0.32,
"learning_rate": 1.9712569994658315e-05,
"loss": 0.0363,
"step": 60
},
{
"epoch": 0.32,
"learning_rate": 1.9698789851894986e-05,
"loss": 0.0358,
"step": 61
},
{
"epoch": 0.33,
"learning_rate": 1.9684692115766683e-05,
"loss": 0.0344,
"step": 62
},
{
"epoch": 0.33,
"learning_rate": 1.9670277247913205e-05,
"loss": 0.0373,
"step": 63
},
{
"epoch": 0.34,
"learning_rate": 1.9655545720359056e-05,
"loss": 0.0302,
"step": 64
},
{
"epoch": 0.34,
"learning_rate": 1.9640498015497956e-05,
"loss": 0.0455,
"step": 65
},
{
"epoch": 0.35,
"learning_rate": 1.9625134626077084e-05,
"loss": 0.0337,
"step": 66
},
{
"epoch": 0.35,
"learning_rate": 1.96094560551809e-05,
"loss": 0.0277,
"step": 67
},
{
"epoch": 0.36,
"learning_rate": 1.9593462816214698e-05,
"loss": 0.0352,
"step": 68
},
{
"epoch": 0.37,
"learning_rate": 1.9577155432887805e-05,
"loss": 0.031,
"step": 69
},
{
"epoch": 0.37,
"learning_rate": 1.956053443919639e-05,
"loss": 0.0298,
"step": 70
},
{
"epoch": 0.38,
"learning_rate": 1.9543600379406027e-05,
"loss": 0.0254,
"step": 71
},
{
"epoch": 0.38,
"learning_rate": 1.9526353808033827e-05,
"loss": 0.0312,
"step": 72
},
{
"epoch": 0.39,
"learning_rate": 1.950879528983032e-05,
"loss": 0.035,
"step": 73
},
{
"epoch": 0.39,
"learning_rate": 1.9490925399760928e-05,
"loss": 0.0298,
"step": 74
},
{
"epoch": 0.4,
"learning_rate": 1.947274472298717e-05,
"loss": 0.0368,
"step": 75
},
{
"epoch": 0.4,
"learning_rate": 1.9454253854847472e-05,
"loss": 0.0346,
"step": 76
},
{
"epoch": 0.41,
"learning_rate": 1.9435453400837683e-05,
"loss": 0.0361,
"step": 77
},
{
"epoch": 0.41,
"learning_rate": 1.941634397659126e-05,
"loss": 0.034,
"step": 78
},
{
"epoch": 0.42,
"learning_rate": 1.9396926207859085e-05,
"loss": 0.0361,
"step": 79
},
{
"epoch": 0.42,
"learning_rate": 1.937720073048899e-05,
"loss": 0.0359,
"step": 80
},
{
"epoch": 0.43,
"learning_rate": 1.9357168190404937e-05,
"loss": 0.0323,
"step": 81
},
{
"epoch": 0.43,
"learning_rate": 1.9336829243585856e-05,
"loss": 0.0435,
"step": 82
},
{
"epoch": 0.44,
"learning_rate": 1.9316184556044176e-05,
"loss": 0.0344,
"step": 83
},
{
"epoch": 0.44,
"learning_rate": 1.9295234803804005e-05,
"loss": 0.0272,
"step": 84
},
{
"epoch": 0.45,
"learning_rate": 1.9273980672879e-05,
"loss": 0.0387,
"step": 85
},
{
"epoch": 0.46,
"learning_rate": 1.925242285924991e-05,
"loss": 0.0331,
"step": 86
},
{
"epoch": 0.46,
"learning_rate": 1.9230562068841764e-05,
"loss": 0.0322,
"step": 87
},
{
"epoch": 0.47,
"learning_rate": 1.9208399017500773e-05,
"loss": 0.0351,
"step": 88
},
{
"epoch": 0.47,
"learning_rate": 1.9185934430970897e-05,
"loss": 0.0285,
"step": 89
},
{
"epoch": 0.48,
"learning_rate": 1.916316904487005e-05,
"loss": 0.0328,
"step": 90
},
{
"epoch": 0.48,
"learning_rate": 1.9140103604666035e-05,
"loss": 0.0374,
"step": 91
},
{
"epoch": 0.49,
"learning_rate": 1.9116738865652134e-05,
"loss": 0.036,
"step": 92
},
{
"epoch": 0.49,
"learning_rate": 1.909307559292236e-05,
"loss": 0.0313,
"step": 93
},
{
"epoch": 0.5,
"learning_rate": 1.906911456134642e-05,
"loss": 0.0331,
"step": 94
},
{
"epoch": 0.5,
"learning_rate": 1.9044856555544323e-05,
"loss": 0.0322,
"step": 95
},
{
"epoch": 0.51,
"learning_rate": 1.9020302369860708e-05,
"loss": 0.0301,
"step": 96
},
{
"epoch": 0.51,
"learning_rate": 1.8995452808338822e-05,
"loss": 0.0266,
"step": 97
},
{
"epoch": 0.52,
"learning_rate": 1.8970308684694186e-05,
"loss": 0.0274,
"step": 98
},
{
"epoch": 0.52,
"learning_rate": 1.8944870822287957e-05,
"loss": 0.0305,
"step": 99
},
{
"epoch": 0.53,
"learning_rate": 1.8919140054099966e-05,
"loss": 0.0273,
"step": 100
},
{
"epoch": 0.53,
"learning_rate": 1.8893117222701435e-05,
"loss": 0.0264,
"step": 101
},
{
"epoch": 0.54,
"learning_rate": 1.8866803180227403e-05,
"loss": 0.0319,
"step": 102
},
{
"epoch": 0.54,
"learning_rate": 1.8840198788348795e-05,
"loss": 0.03,
"step": 103
},
{
"epoch": 0.55,
"learning_rate": 1.881330491824423e-05,
"loss": 0.0268,
"step": 104
},
{
"epoch": 0.56,
"learning_rate": 1.8786122450571485e-05,
"loss": 0.0311,
"step": 105
},
{
"epoch": 0.56,
"learning_rate": 1.8758652275438657e-05,
"loss": 0.0269,
"step": 106
},
{
"epoch": 0.57,
"learning_rate": 1.8730895292375018e-05,
"loss": 0.0298,
"step": 107
},
{
"epoch": 0.57,
"learning_rate": 1.8702852410301556e-05,
"loss": 0.0275,
"step": 108
},
{
"epoch": 0.58,
"learning_rate": 1.8674524547501207e-05,
"loss": 0.0287,
"step": 109
},
{
"epoch": 0.58,
"learning_rate": 1.8645912631588806e-05,
"loss": 0.0314,
"step": 110
},
{
"epoch": 0.59,
"learning_rate": 1.861701759948068e-05,
"loss": 0.0307,
"step": 111
},
{
"epoch": 0.59,
"learning_rate": 1.8587840397364007e-05,
"loss": 0.0322,
"step": 112
},
{
"epoch": 0.6,
"learning_rate": 1.855838198066579e-05,
"loss": 0.0236,
"step": 113
},
{
"epoch": 0.6,
"learning_rate": 1.85286433140216e-05,
"loss": 0.0309,
"step": 114
},
{
"epoch": 0.61,
"learning_rate": 1.8498625371243978e-05,
"loss": 0.0273,
"step": 115
},
{
"epoch": 0.61,
"learning_rate": 1.8468329135290555e-05,
"loss": 0.0362,
"step": 116
},
{
"epoch": 0.62,
"learning_rate": 1.8437755598231857e-05,
"loss": 0.0238,
"step": 117
},
{
"epoch": 0.62,
"learning_rate": 1.8406905761218815e-05,
"loss": 0.0271,
"step": 118
},
{
"epoch": 0.63,
"learning_rate": 1.837578063444998e-05,
"loss": 0.0309,
"step": 119
},
{
"epoch": 0.63,
"learning_rate": 1.8344381237138473e-05,
"loss": 0.0305,
"step": 120
},
{
"epoch": 0.64,
"learning_rate": 1.831270859747857e-05,
"loss": 0.025,
"step": 121
},
{
"epoch": 0.65,
"learning_rate": 1.8280763752612052e-05,
"loss": 0.028,
"step": 122
},
{
"epoch": 0.65,
"learning_rate": 1.8248547748594246e-05,
"loss": 0.0257,
"step": 123
},
{
"epoch": 0.66,
"learning_rate": 1.8216061640359764e-05,
"loss": 0.0255,
"step": 124
},
{
"epoch": 0.66,
"learning_rate": 1.8183306491687968e-05,
"loss": 0.0288,
"step": 125
},
{
"epoch": 0.67,
"learning_rate": 1.8150283375168112e-05,
"loss": 0.0318,
"step": 126
},
{
"epoch": 0.67,
"learning_rate": 1.8116993372164265e-05,
"loss": 0.0282,
"step": 127
},
{
"epoch": 0.68,
"learning_rate": 1.8083437572779842e-05,
"loss": 0.0302,
"step": 128
},
{
"epoch": 0.68,
"learning_rate": 1.8049617075821962e-05,
"loss": 0.0264,
"step": 129
},
{
"epoch": 0.69,
"learning_rate": 1.8015532988765427e-05,
"loss": 0.027,
"step": 130
},
{
"epoch": 0.69,
"learning_rate": 1.7981186427716478e-05,
"loss": 0.0251,
"step": 131
},
{
"epoch": 0.7,
"learning_rate": 1.794657851737625e-05,
"loss": 0.0339,
"step": 132
},
{
"epoch": 0.7,
"learning_rate": 1.791171039100393e-05,
"loss": 0.0276,
"step": 133
},
{
"epoch": 0.71,
"learning_rate": 1.787658319037965e-05,
"loss": 0.0252,
"step": 134
},
{
"epoch": 0.71,
"learning_rate": 1.7841198065767107e-05,
"loss": 0.0262,
"step": 135
},
{
"epoch": 0.72,
"learning_rate": 1.7805556175875886e-05,
"loss": 0.0243,
"step": 136
},
{
"epoch": 0.72,
"learning_rate": 1.7769658687823525e-05,
"loss": 0.03,
"step": 137
},
{
"epoch": 0.73,
"learning_rate": 1.77335067770973e-05,
"loss": 0.0271,
"step": 138
},
{
"epoch": 0.74,
"learning_rate": 1.7697101627515722e-05,
"loss": 0.0296,
"step": 139
},
{
"epoch": 0.74,
"learning_rate": 1.766044443118978e-05,
"loss": 0.0221,
"step": 140
},
{
"epoch": 0.75,
"learning_rate": 1.7623536388483902e-05,
"loss": 0.0272,
"step": 141
},
{
"epoch": 0.75,
"learning_rate": 1.758637870797665e-05,
"loss": 0.0257,
"step": 142
},
{
"epoch": 0.76,
"learning_rate": 1.754897260642114e-05,
"loss": 0.0321,
"step": 143
},
{
"epoch": 0.76,
"learning_rate": 1.7511319308705198e-05,
"loss": 0.0276,
"step": 144
},
{
"epoch": 0.77,
"learning_rate": 1.747342004781127e-05,
"loss": 0.0314,
"step": 145
},
{
"epoch": 0.77,
"learning_rate": 1.7435276064776004e-05,
"loss": 0.0249,
"step": 146
},
{
"epoch": 0.78,
"learning_rate": 1.7396888608649673e-05,
"loss": 0.0289,
"step": 147
},
{
"epoch": 0.78,
"learning_rate": 1.7358258936455203e-05,
"loss": 0.0271,
"step": 148
},
{
"epoch": 0.79,
"learning_rate": 1.7319388313147067e-05,
"loss": 0.0231,
"step": 149
},
{
"epoch": 0.79,
"learning_rate": 1.7280278011569848e-05,
"loss": 0.0272,
"step": 150
},
{
"epoch": 0.8,
"learning_rate": 1.7240929312416545e-05,
"loss": 0.0307,
"step": 151
},
{
"epoch": 0.8,
"learning_rate": 1.7201343504186646e-05,
"loss": 0.0264,
"step": 152
},
{
"epoch": 0.81,
"learning_rate": 1.7161521883143936e-05,
"loss": 0.0295,
"step": 153
},
{
"epoch": 0.81,
"learning_rate": 1.7121465753274047e-05,
"loss": 0.0259,
"step": 154
},
{
"epoch": 0.82,
"learning_rate": 1.708117642624176e-05,
"loss": 0.0299,
"step": 155
},
{
"epoch": 0.83,
"learning_rate": 1.7040655221348057e-05,
"loss": 0.0261,
"step": 156
},
{
"epoch": 0.83,
"learning_rate": 1.6999903465486913e-05,
"loss": 0.0273,
"step": 157
},
{
"epoch": 0.84,
"learning_rate": 1.6958922493101844e-05,
"loss": 0.0269,
"step": 158
},
{
"epoch": 0.84,
"learning_rate": 1.6917713646142222e-05,
"loss": 0.0291,
"step": 159
},
{
"epoch": 0.85,
"learning_rate": 1.687627827401932e-05,
"loss": 0.0236,
"step": 160
},
{
"epoch": 0.85,
"learning_rate": 1.683461773356213e-05,
"loss": 0.0222,
"step": 161
},
{
"epoch": 0.86,
"learning_rate": 1.679273338897293e-05,
"loss": 0.0297,
"step": 162
},
{
"epoch": 0.86,
"learning_rate": 1.6750626611782624e-05,
"loss": 0.0275,
"step": 163
},
{
"epoch": 0.87,
"learning_rate": 1.6708298780805808e-05,
"loss": 0.0244,
"step": 164
},
{
"epoch": 0.87,
"learning_rate": 1.6665751282095634e-05,
"loss": 0.0226,
"step": 165
},
{
"epoch": 0.88,
"learning_rate": 1.6622985508898427e-05,
"loss": 0.0293,
"step": 166
},
{
"epoch": 0.88,
"learning_rate": 1.6580002861608053e-05,
"loss": 0.0307,
"step": 167
},
{
"epoch": 0.89,
"learning_rate": 1.653680474772006e-05,
"loss": 0.0267,
"step": 168
},
{
"epoch": 0.89,
"learning_rate": 1.6493392581785608e-05,
"loss": 0.0295,
"step": 169
},
{
"epoch": 0.9,
"learning_rate": 1.644976778536512e-05,
"loss": 0.0274,
"step": 170
},
{
"epoch": 0.9,
"learning_rate": 1.6405931786981753e-05,
"loss": 0.0282,
"step": 171
},
{
"epoch": 0.91,
"learning_rate": 1.6361886022074612e-05,
"loss": 0.0245,
"step": 172
},
{
"epoch": 0.92,
"learning_rate": 1.6317631932951754e-05,
"loss": 0.0302,
"step": 173
},
{
"epoch": 0.92,
"learning_rate": 1.6273170968742942e-05,
"loss": 0.0271,
"step": 174
},
{
"epoch": 0.93,
"learning_rate": 1.62285045853522e-05,
"loss": 0.0267,
"step": 175
},
{
"epoch": 0.93,
"learning_rate": 1.618363424541016e-05,
"loss": 0.0201,
"step": 176
},
{
"epoch": 0.94,
"learning_rate": 1.613856141822612e-05,
"loss": 0.0324,
"step": 177
},
{
"epoch": 0.94,
"learning_rate": 1.6093287579739983e-05,
"loss": 0.0289,
"step": 178
},
{
"epoch": 0.95,
"learning_rate": 1.604781421247389e-05,
"loss": 0.0288,
"step": 179
},
{
"epoch": 0.95,
"learning_rate": 1.6002142805483686e-05,
"loss": 0.029,
"step": 180
},
{
"epoch": 0.96,
"learning_rate": 1.5956274854310157e-05,
"loss": 0.0242,
"step": 181
},
{
"epoch": 0.96,
"learning_rate": 1.5910211860930063e-05,
"loss": 0.0278,
"step": 182
},
{
"epoch": 0.97,
"learning_rate": 1.586395533370696e-05,
"loss": 0.0319,
"step": 183
},
{
"epoch": 0.97,
"learning_rate": 1.5817506787341782e-05,
"loss": 0.0247,
"step": 184
},
{
"epoch": 0.98,
"learning_rate": 1.5770867742823268e-05,
"loss": 0.0246,
"step": 185
},
{
"epoch": 0.98,
"learning_rate": 1.572403972737815e-05,
"loss": 0.0273,
"step": 186
},
{
"epoch": 0.99,
"learning_rate": 1.567702427442113e-05,
"loss": 0.0281,
"step": 187
},
{
"epoch": 0.99,
"learning_rate": 1.5629822923504692e-05,
"loss": 0.0289,
"step": 188
},
{
"epoch": 1.0,
"learning_rate": 1.5582437220268648e-05,
"loss": 0.026,
"step": 189
},
{
"epoch": 1.01,
"learning_rate": 1.553486871638958e-05,
"loss": 0.0183,
"step": 190
},
{
"epoch": 1.01,
"learning_rate": 1.5487118969529973e-05,
"loss": 0.0248,
"step": 191
},
{
"epoch": 1.02,
"learning_rate": 1.5439189543287247e-05,
"loss": 0.0309,
"step": 192
},
{
"epoch": 1.02,
"learning_rate": 1.539108200714255e-05,
"loss": 0.0274,
"step": 193
},
{
"epoch": 1.03,
"learning_rate": 1.534279793640935e-05,
"loss": 0.0235,
"step": 194
},
{
"epoch": 1.03,
"learning_rate": 1.529433891218185e-05,
"loss": 0.0276,
"step": 195
},
{
"epoch": 1.04,
"learning_rate": 1.5245706521283246e-05,
"loss": 0.0228,
"step": 196
},
{
"epoch": 1.04,
"learning_rate": 1.5196902356213715e-05,
"loss": 0.0213,
"step": 197
},
{
"epoch": 1.05,
"learning_rate": 1.5147928015098309e-05,
"loss": 0.0185,
"step": 198
},
{
"epoch": 1.05,
"learning_rate": 1.5098785101634605e-05,
"loss": 0.0231,
"step": 199
},
{
"epoch": 1.06,
"learning_rate": 1.5049475225040202e-05,
"loss": 0.0202,
"step": 200
},
{
"epoch": 1.06,
"learning_rate": 1.5000000000000002e-05,
"loss": 0.0236,
"step": 201
},
{
"epoch": 1.07,
"learning_rate": 1.4950361046613367e-05,
"loss": 0.0242,
"step": 202
},
{
"epoch": 1.07,
"learning_rate": 1.4900559990341048e-05,
"loss": 0.0254,
"step": 203
},
{
"epoch": 1.08,
"learning_rate": 1.4850598461951963e-05,
"loss": 0.0231,
"step": 204
},
{
"epoch": 1.08,
"learning_rate": 1.4800478097469799e-05,
"loss": 0.023,
"step": 205
},
{
"epoch": 1.09,
"learning_rate": 1.4750200538119435e-05,
"loss": 0.025,
"step": 206
},
{
"epoch": 1.1,
"learning_rate": 1.4699767430273202e-05,
"loss": 0.0231,
"step": 207
},
{
"epoch": 1.1,
"learning_rate": 1.4649180425396972e-05,
"loss": 0.0239,
"step": 208
},
{
"epoch": 1.11,
"learning_rate": 1.4598441179996075e-05,
"loss": 0.0277,
"step": 209
},
{
"epoch": 1.11,
"learning_rate": 1.454755135556106e-05,
"loss": 0.0231,
"step": 210
},
{
"epoch": 1.12,
"learning_rate": 1.4496512618513289e-05,
"loss": 0.0225,
"step": 211
},
{
"epoch": 1.12,
"learning_rate": 1.444532664015037e-05,
"loss": 0.0266,
"step": 212
},
{
"epoch": 1.13,
"learning_rate": 1.4393995096591415e-05,
"loss": 0.0204,
"step": 213
},
{
"epoch": 1.13,
"learning_rate": 1.4342519668722184e-05,
"loss": 0.0256,
"step": 214
},
{
"epoch": 1.14,
"learning_rate": 1.4290902042140005e-05,
"loss": 0.0231,
"step": 215
},
{
"epoch": 1.14,
"learning_rate": 1.423914390709861e-05,
"loss": 0.0203,
"step": 216
},
{
"epoch": 1.15,
"learning_rate": 1.4187246958452772e-05,
"loss": 0.0259,
"step": 217
},
{
"epoch": 1.15,
"learning_rate": 1.413521289560281e-05,
"loss": 0.0242,
"step": 218
},
{
"epoch": 1.16,
"learning_rate": 1.4083043422438936e-05,
"loss": 0.0219,
"step": 219
},
{
"epoch": 1.16,
"learning_rate": 1.4030740247285466e-05,
"loss": 0.0295,
"step": 220
},
{
"epoch": 1.17,
"learning_rate": 1.3978305082844876e-05,
"loss": 0.0188,
"step": 221
},
{
"epoch": 1.17,
"learning_rate": 1.3925739646141721e-05,
"loss": 0.0253,
"step": 222
},
{
"epoch": 1.18,
"learning_rate": 1.3873045658466404e-05,
"loss": 0.0284,
"step": 223
},
{
"epoch": 1.19,
"learning_rate": 1.3820224845318821e-05,
"loss": 0.0229,
"step": 224
},
{
"epoch": 1.19,
"learning_rate": 1.3767278936351853e-05,
"loss": 0.0209,
"step": 225
},
{
"epoch": 1.2,
"learning_rate": 1.371420966531472e-05,
"loss": 0.0266,
"step": 226
},
{
"epoch": 1.2,
"learning_rate": 1.3661018769996228e-05,
"loss": 0.0253,
"step": 227
},
{
"epoch": 1.21,
"learning_rate": 1.3607707992167836e-05,
"loss": 0.0247,
"step": 228
},
{
"epoch": 1.21,
"learning_rate": 1.3554279077526648e-05,
"loss": 0.0213,
"step": 229
},
{
"epoch": 1.22,
"learning_rate": 1.3500733775638232e-05,
"loss": 0.0321,
"step": 230
},
{
"epoch": 1.22,
"learning_rate": 1.3447073839879339e-05,
"loss": 0.0238,
"step": 231
},
{
"epoch": 1.23,
"learning_rate": 1.3393301027380476e-05,
"loss": 0.0228,
"step": 232
},
{
"epoch": 1.23,
"learning_rate": 1.333941709896838e-05,
"loss": 0.0227,
"step": 233
},
{
"epoch": 1.24,
"learning_rate": 1.3285423819108349e-05,
"loss": 0.022,
"step": 234
},
{
"epoch": 1.24,
"learning_rate": 1.3231322955846469e-05,
"loss": 0.0261,
"step": 235
},
{
"epoch": 1.25,
"learning_rate": 1.3177116280751717e-05,
"loss": 0.0194,
"step": 236
},
{
"epoch": 1.25,
"learning_rate": 1.3122805568857948e-05,
"loss": 0.0227,
"step": 237
},
{
"epoch": 1.26,
"learning_rate": 1.3068392598605775e-05,
"loss": 0.0217,
"step": 238
},
{
"epoch": 1.26,
"learning_rate": 1.3013879151784321e-05,
"loss": 0.0261,
"step": 239
},
{
"epoch": 1.27,
"learning_rate": 1.2959267013472894e-05,
"loss": 0.0243,
"step": 240
},
{
"epoch": 1.28,
"learning_rate": 1.2904557971982514e-05,
"loss": 0.0257,
"step": 241
},
{
"epoch": 1.28,
"learning_rate": 1.2849753818797353e-05,
"loss": 0.0227,
"step": 242
},
{
"epoch": 1.29,
"learning_rate": 1.2794856348516095e-05,
"loss": 0.0187,
"step": 243
},
{
"epoch": 1.29,
"learning_rate": 1.2739867358793142e-05,
"loss": 0.021,
"step": 244
},
{
"epoch": 1.3,
"learning_rate": 1.2684788650279772e-05,
"loss": 0.0249,
"step": 245
},
{
"epoch": 1.3,
"learning_rate": 1.2629622026565147e-05,
"loss": 0.0238,
"step": 246
},
{
"epoch": 1.31,
"learning_rate": 1.2574369294117296e-05,
"loss": 0.0261,
"step": 247
},
{
"epoch": 1.31,
"learning_rate": 1.2519032262223913e-05,
"loss": 0.0242,
"step": 248
},
{
"epoch": 1.32,
"learning_rate": 1.2463612742933148e-05,
"loss": 0.0219,
"step": 249
},
{
"epoch": 1.32,
"learning_rate": 1.2408112550994253e-05,
"loss": 0.0273,
"step": 250
},
{
"epoch": 1.33,
"learning_rate": 1.2352533503798156e-05,
"loss": 0.0245,
"step": 251
},
{
"epoch": 1.33,
"learning_rate": 1.2296877421317958e-05,
"loss": 0.0186,
"step": 252
},
{
"epoch": 1.34,
"learning_rate": 1.2241146126049326e-05,
"loss": 0.0159,
"step": 253
},
{
"epoch": 1.34,
"learning_rate": 1.2185341442950829e-05,
"loss": 0.0225,
"step": 254
},
{
"epoch": 1.35,
"learning_rate": 1.2129465199384158e-05,
"loss": 0.024,
"step": 255
},
{
"epoch": 1.35,
"learning_rate": 1.2073519225054314e-05,
"loss": 0.0224,
"step": 256
},
{
"epoch": 1.36,
"learning_rate": 1.201750535194966e-05,
"loss": 0.0178,
"step": 257
},
{
"epoch": 1.37,
"learning_rate": 1.196142541428197e-05,
"loss": 0.0227,
"step": 258
},
{
"epoch": 1.37,
"learning_rate": 1.1905281248426333e-05,
"loss": 0.0248,
"step": 259
},
{
"epoch": 1.38,
"learning_rate": 1.1849074692861033e-05,
"loss": 0.0231,
"step": 260
},
{
"epoch": 1.38,
"learning_rate": 1.1792807588107358e-05,
"loss": 0.0221,
"step": 261
},
{
"epoch": 1.39,
"learning_rate": 1.1736481776669307e-05,
"loss": 0.0237,
"step": 262
},
{
"epoch": 1.39,
"learning_rate": 1.1680099102973271e-05,
"loss": 0.0205,
"step": 263
},
{
"epoch": 1.4,
"learning_rate": 1.1623661413307638e-05,
"loss": 0.0295,
"step": 264
},
{
"epoch": 1.4,
"learning_rate": 1.1567170555762335e-05,
"loss": 0.0224,
"step": 265
},
{
"epoch": 1.41,
"learning_rate": 1.1510628380168291e-05,
"loss": 0.023,
"step": 266
},
{
"epoch": 1.41,
"learning_rate": 1.14540367380369e-05,
"loss": 0.0218,
"step": 267
},
{
"epoch": 1.42,
"learning_rate": 1.1397397482499352e-05,
"loss": 0.0185,
"step": 268
},
{
"epoch": 1.42,
"learning_rate": 1.1340712468245977e-05,
"loss": 0.0261,
"step": 269
},
{
"epoch": 1.43,
"learning_rate": 1.1283983551465512e-05,
"loss": 0.024,
"step": 270
},
{
"epoch": 1.43,
"learning_rate": 1.1227212589784297e-05,
"loss": 0.0235,
"step": 271
},
{
"epoch": 1.44,
"learning_rate": 1.1170401442205475e-05,
"loss": 0.0214,
"step": 272
},
{
"epoch": 1.44,
"learning_rate": 1.1113551969048088e-05,
"loss": 0.0218,
"step": 273
},
{
"epoch": 1.45,
"learning_rate": 1.1056666031886193e-05,
"loss": 0.0185,
"step": 274
},
{
"epoch": 1.46,
"learning_rate": 1.099974549348787e-05,
"loss": 0.0211,
"step": 275
},
{
"epoch": 1.46,
"learning_rate": 1.0942792217754245e-05,
"loss": 0.0229,
"step": 276
},
{
"epoch": 1.47,
"learning_rate": 1.0885808069658452e-05,
"loss": 0.0232,
"step": 277
},
{
"epoch": 1.47,
"learning_rate": 1.0828794915184556e-05,
"loss": 0.0227,
"step": 278
},
{
"epoch": 1.48,
"learning_rate": 1.0771754621266466e-05,
"loss": 0.025,
"step": 279
},
{
"epoch": 1.48,
"learning_rate": 1.071468905572677e-05,
"loss": 0.02,
"step": 280
},
{
"epoch": 1.49,
"learning_rate": 1.0657600087215618e-05,
"loss": 0.0253,
"step": 281
},
{
"epoch": 1.49,
"learning_rate": 1.0600489585149485e-05,
"loss": 0.0218,
"step": 282
},
{
"epoch": 1.5,
"learning_rate": 1.0543359419649986e-05,
"loss": 0.0212,
"step": 283
},
{
"epoch": 1.5,
"learning_rate": 1.0486211461482625e-05,
"loss": 0.0225,
"step": 284
},
{
"epoch": 1.51,
"learning_rate": 1.0429047581995547e-05,
"loss": 0.0216,
"step": 285
},
{
"epoch": 1.51,
"learning_rate": 1.0371869653058235e-05,
"loss": 0.0209,
"step": 286
},
{
"epoch": 1.52,
"learning_rate": 1.0314679547000251e-05,
"loss": 0.0228,
"step": 287
},
{
"epoch": 1.52,
"learning_rate": 1.0257479136549889e-05,
"loss": 0.0187,
"step": 288
},
{
"epoch": 1.53,
"learning_rate": 1.0200270294772869e-05,
"loss": 0.0172,
"step": 289
},
{
"epoch": 1.53,
"learning_rate": 1.0143054895011011e-05,
"loss": 0.0247,
"step": 290
},
{
"epoch": 1.54,
"learning_rate": 1.0085834810820871e-05,
"loss": 0.016,
"step": 291
},
{
"epoch": 1.54,
"learning_rate": 1.0028611915912405e-05,
"loss": 0.0187,
"step": 292
},
{
"epoch": 1.55,
"learning_rate": 9.9713880840876e-06,
"loss": 0.0221,
"step": 293
},
{
"epoch": 1.56,
"learning_rate": 9.914165189179132e-06,
"loss": 0.0246,
"step": 294
},
{
"epoch": 1.56,
"learning_rate": 9.856945104988989e-06,
"loss": 0.0241,
"step": 295
},
{
"epoch": 1.57,
"learning_rate": 9.799729705227133e-06,
"loss": 0.0213,
"step": 296
},
{
"epoch": 1.57,
"learning_rate": 9.742520863450116e-06,
"loss": 0.025,
"step": 297
},
{
"epoch": 1.58,
"learning_rate": 9.68532045299975e-06,
"loss": 0.0232,
"step": 298
},
{
"epoch": 1.58,
"learning_rate": 9.628130346941767e-06,
"loss": 0.0286,
"step": 299
},
{
"epoch": 1.59,
"learning_rate": 9.570952418004455e-06,
"loss": 0.0247,
"step": 300
},
{
"epoch": 1.59,
"learning_rate": 9.513788538517375e-06,
"loss": 0.0209,
"step": 301
},
{
"epoch": 1.6,
"learning_rate": 9.456640580350019e-06,
"loss": 0.0223,
"step": 302
},
{
"epoch": 1.6,
"learning_rate": 9.399510414850518e-06,
"loss": 0.026,
"step": 303
},
{
"epoch": 1.61,
"learning_rate": 9.342399912784387e-06,
"loss": 0.0231,
"step": 304
},
{
"epoch": 1.61,
"learning_rate": 9.285310944273232e-06,
"loss": 0.0206,
"step": 305
},
{
"epoch": 1.62,
"learning_rate": 9.228245378733537e-06,
"loss": 0.0206,
"step": 306
},
{
"epoch": 1.62,
"learning_rate": 9.171205084815445e-06,
"loss": 0.0244,
"step": 307
},
{
"epoch": 1.63,
"learning_rate": 9.11419193034155e-06,
"loss": 0.0214,
"step": 308
},
{
"epoch": 1.63,
"learning_rate": 9.057207782245756e-06,
"loss": 0.0242,
"step": 309
},
{
"epoch": 1.64,
"learning_rate": 9.000254506512133e-06,
"loss": 0.0187,
"step": 310
},
{
"epoch": 1.65,
"learning_rate": 8.943333968113808e-06,
"loss": 0.0186,
"step": 311
},
{
"epoch": 1.65,
"learning_rate": 8.886448030951912e-06,
"loss": 0.0255,
"step": 312
},
{
"epoch": 1.66,
"learning_rate": 8.82959855779453e-06,
"loss": 0.0229,
"step": 313
},
{
"epoch": 1.66,
"learning_rate": 8.772787410215706e-06,
"loss": 0.0202,
"step": 314
},
{
"epoch": 1.67,
"learning_rate": 8.71601644853449e-06,
"loss": 0.0179,
"step": 315
},
{
"epoch": 1.67,
"learning_rate": 8.659287531754025e-06,
"loss": 0.0224,
"step": 316
},
{
"epoch": 1.68,
"learning_rate": 8.60260251750065e-06,
"loss": 0.0211,
"step": 317
},
{
"epoch": 1.68,
"learning_rate": 8.545963261963102e-06,
"loss": 0.021,
"step": 318
},
{
"epoch": 1.69,
"learning_rate": 8.48937161983171e-06,
"loss": 0.0203,
"step": 319
},
{
"epoch": 1.69,
"learning_rate": 8.432829444237667e-06,
"loss": 0.0226,
"step": 320
},
{
"epoch": 1.7,
"learning_rate": 8.376338586692367e-06,
"loss": 0.0224,
"step": 321
},
{
"epoch": 1.7,
"learning_rate": 8.319900897026734e-06,
"loss": 0.0194,
"step": 322
},
{
"epoch": 1.71,
"learning_rate": 8.263518223330698e-06,
"loss": 0.0191,
"step": 323
},
{
"epoch": 1.71,
"learning_rate": 8.207192411892645e-06,
"loss": 0.0196,
"step": 324
},
{
"epoch": 1.72,
"learning_rate": 8.150925307138968e-06,
"loss": 0.0269,
"step": 325
},
{
"epoch": 1.72,
"learning_rate": 8.094718751573669e-06,
"loss": 0.0221,
"step": 326
},
{
"epoch": 1.73,
"learning_rate": 8.038574585718032e-06,
"loss": 0.021,
"step": 327
},
{
"epoch": 1.74,
"learning_rate": 7.982494648050341e-06,
"loss": 0.0257,
"step": 328
},
{
"epoch": 1.74,
"learning_rate": 7.926480774945688e-06,
"loss": 0.0224,
"step": 329
},
{
"epoch": 1.75,
"learning_rate": 7.870534800615845e-06,
"loss": 0.0219,
"step": 330
},
{
"epoch": 1.75,
"learning_rate": 7.814658557049175e-06,
"loss": 0.0228,
"step": 331
},
{
"epoch": 1.76,
"learning_rate": 7.758853873950676e-06,
"loss": 0.0238,
"step": 332
},
{
"epoch": 1.76,
"learning_rate": 7.703122578682047e-06,
"loss": 0.0211,
"step": 333
},
{
"epoch": 1.77,
"learning_rate": 7.647466496201848e-06,
"loss": 0.018,
"step": 334
},
{
"epoch": 1.77,
"learning_rate": 7.591887449005748e-06,
"loss": 0.0223,
"step": 335
},
{
"epoch": 1.78,
"learning_rate": 7.536387257066854e-06,
"loss": 0.0243,
"step": 336
},
{
"epoch": 1.78,
"learning_rate": 7.480967737776089e-06,
"loss": 0.0214,
"step": 337
},
{
"epoch": 1.79,
"learning_rate": 7.425630705882707e-06,
"loss": 0.0199,
"step": 338
},
{
"epoch": 1.79,
"learning_rate": 7.370377973434854e-06,
"loss": 0.0246,
"step": 339
},
{
"epoch": 1.8,
"learning_rate": 7.315211349720231e-06,
"loss": 0.0192,
"step": 340
},
{
"epoch": 1.8,
"learning_rate": 7.260132641206861e-06,
"loss": 0.0179,
"step": 341
},
{
"epoch": 1.81,
"learning_rate": 7.2051436514839064e-06,
"loss": 0.0188,
"step": 342
},
{
"epoch": 1.81,
"learning_rate": 7.150246181202648e-06,
"loss": 0.0193,
"step": 343
},
{
"epoch": 1.82,
"learning_rate": 7.0954420280174915e-06,
"loss": 0.0177,
"step": 344
},
{
"epoch": 1.83,
"learning_rate": 7.040732986527108e-06,
"loss": 0.0197,
"step": 345
},
{
"epoch": 1.83,
"learning_rate": 6.9861208482156785e-06,
"loss": 0.0214,
"step": 346
},
{
"epoch": 1.84,
"learning_rate": 6.931607401394229e-06,
"loss": 0.0175,
"step": 347
},
{
"epoch": 1.84,
"learning_rate": 6.877194431142055e-06,
"loss": 0.0202,
"step": 348
},
{
"epoch": 1.85,
"learning_rate": 6.822883719248283e-06,
"loss": 0.0194,
"step": 349
},
{
"epoch": 1.85,
"learning_rate": 6.768677044153535e-06,
"loss": 0.0242,
"step": 350
},
{
"epoch": 1.86,
"learning_rate": 6.714576180891653e-06,
"loss": 0.0248,
"step": 351
},
{
"epoch": 1.86,
"learning_rate": 6.660582901031621e-06,
"loss": 0.025,
"step": 352
},
{
"epoch": 1.87,
"learning_rate": 6.6066989726195265e-06,
"loss": 0.0195,
"step": 353
},
{
"epoch": 1.87,
"learning_rate": 6.552926160120663e-06,
"loss": 0.0189,
"step": 354
},
{
"epoch": 1.88,
"learning_rate": 6.499266224361768e-06,
"loss": 0.0201,
"step": 355
},
{
"epoch": 1.88,
"learning_rate": 6.445720922473355e-06,
"loss": 0.0227,
"step": 356
},
{
"epoch": 1.89,
"learning_rate": 6.3922920078321685e-06,
"loss": 0.0196,
"step": 357
},
{
"epoch": 1.89,
"learning_rate": 6.3389812300037774e-06,
"loss": 0.0233,
"step": 358
},
{
"epoch": 1.9,
"learning_rate": 6.285790334685283e-06,
"loss": 0.02,
"step": 359
},
{
"epoch": 1.9,
"learning_rate": 6.232721063648148e-06,
"loss": 0.0173,
"step": 360
},
{
"epoch": 1.91,
"learning_rate": 6.179775154681184e-06,
"loss": 0.0198,
"step": 361
},
{
"epoch": 1.92,
"learning_rate": 6.1269543415336e-06,
"loss": 0.0224,
"step": 362
},
{
"epoch": 1.92,
"learning_rate": 6.074260353858283e-06,
"loss": 0.0186,
"step": 363
},
{
"epoch": 1.93,
"learning_rate": 6.0216949171551285e-06,
"loss": 0.0236,
"step": 364
},
{
"epoch": 1.93,
"learning_rate": 5.969259752714536e-06,
"loss": 0.0182,
"step": 365
},
{
"epoch": 1.94,
"learning_rate": 5.916956577561066e-06,
"loss": 0.0215,
"step": 366
},
{
"epoch": 1.94,
"learning_rate": 5.864787104397194e-06,
"loss": 0.0234,
"step": 367
},
{
"epoch": 1.95,
"learning_rate": 5.81275304154723e-06,
"loss": 0.0228,
"step": 368
},
{
"epoch": 1.95,
"learning_rate": 5.760856092901394e-06,
"loss": 0.0247,
"step": 369
},
{
"epoch": 1.96,
"learning_rate": 5.709097957860001e-06,
"loss": 0.0215,
"step": 370
},
{
"epoch": 1.96,
"learning_rate": 5.6574803312778196e-06,
"loss": 0.0184,
"step": 371
},
{
"epoch": 1.97,
"learning_rate": 5.6060049034085815e-06,
"loss": 0.0245,
"step": 372
},
{
"epoch": 1.97,
"learning_rate": 5.554673359849632e-06,
"loss": 0.0209,
"step": 373
},
{
"epoch": 1.98,
"learning_rate": 5.5034873814867125e-06,
"loss": 0.022,
"step": 374
},
{
"epoch": 1.98,
"learning_rate": 5.4524486444389455e-06,
"loss": 0.0236,
"step": 375
},
{
"epoch": 1.99,
"learning_rate": 5.40155882000393e-06,
"loss": 0.0192,
"step": 376
},
{
"epoch": 1.99,
"learning_rate": 5.3508195746030304e-06,
"loss": 0.0241,
"step": 377
},
{
"epoch": 2.0,
"learning_rate": 5.300232569726805e-06,
"loss": 0.0193,
"step": 378
},
{
"epoch": 2.01,
"learning_rate": 5.249799461880569e-06,
"loss": 0.0196,
"step": 379
},
{
"epoch": 2.01,
"learning_rate": 5.199521902530203e-06,
"loss": 0.0222,
"step": 380
},
{
"epoch": 2.02,
"learning_rate": 5.14940153804804e-06,
"loss": 0.0222,
"step": 381
},
{
"epoch": 2.02,
"learning_rate": 5.0994400096589565e-06,
"loss": 0.0213,
"step": 382
},
{
"epoch": 2.03,
"learning_rate": 5.049638953386635e-06,
"loss": 0.0192,
"step": 383
},
{
"epoch": 2.03,
"learning_rate": 5.000000000000003e-06,
"loss": 0.0201,
"step": 384
},
{
"epoch": 2.04,
"learning_rate": 4.950524774959801e-06,
"loss": 0.0228,
"step": 385
},
{
"epoch": 2.04,
"learning_rate": 4.901214898365396e-06,
"loss": 0.0189,
"step": 386
},
{
"epoch": 2.05,
"learning_rate": 4.852071984901696e-06,
"loss": 0.0134,
"step": 387
},
{
"epoch": 2.05,
"learning_rate": 4.803097643786289e-06,
"loss": 0.0208,
"step": 388
},
{
"epoch": 2.06,
"learning_rate": 4.754293478716755e-06,
"loss": 0.0173,
"step": 389
},
{
"epoch": 2.06,
"learning_rate": 4.705661087818149e-06,
"loss": 0.0174,
"step": 390
},
{
"epoch": 2.07,
"learning_rate": 4.6572020635906535e-06,
"loss": 0.0163,
"step": 391
},
{
"epoch": 2.07,
"learning_rate": 4.608917992857449e-06,
"loss": 0.019,
"step": 392
},
{
"epoch": 2.08,
"learning_rate": 4.560810456712754e-06,
"loss": 0.0162,
"step": 393
},
{
"epoch": 2.08,
"learning_rate": 4.51288103047003e-06,
"loss": 0.0188,
"step": 394
},
{
"epoch": 2.09,
"learning_rate": 4.465131283610425e-06,
"loss": 0.0201,
"step": 395
},
{
"epoch": 2.1,
"learning_rate": 4.417562779731355e-06,
"loss": 0.0143,
"step": 396
},
{
"epoch": 2.1,
"learning_rate": 4.370177076495312e-06,
"loss": 0.0183,
"step": 397
},
{
"epoch": 2.11,
"learning_rate": 4.322975725578871e-06,
"loss": 0.018,
"step": 398
},
{
"epoch": 2.11,
"learning_rate": 4.275960272621852e-06,
"loss": 0.0164,
"step": 399
},
{
"epoch": 2.12,
"learning_rate": 4.229132257176735e-06,
"loss": 0.0161,
"step": 400
},
{
"epoch": 2.12,
"learning_rate": 4.182493212658224e-06,
"loss": 0.0179,
"step": 401
},
{
"epoch": 2.13,
"learning_rate": 4.1360446662930445e-06,
"loss": 0.02,
"step": 402
},
{
"epoch": 2.13,
"learning_rate": 4.089788139069936e-06,
"loss": 0.0206,
"step": 403
},
{
"epoch": 2.14,
"learning_rate": 4.0437251456898465e-06,
"loss": 0.0157,
"step": 404
},
{
"epoch": 2.14,
"learning_rate": 3.997857194516321e-06,
"loss": 0.0205,
"step": 405
},
{
"epoch": 2.15,
"learning_rate": 3.952185787526112e-06,
"loss": 0.0177,
"step": 406
},
{
"epoch": 2.15,
"learning_rate": 3.90671242026002e-06,
"loss": 0.0164,
"step": 407
},
{
"epoch": 2.16,
"learning_rate": 3.86143858177388e-06,
"loss": 0.0208,
"step": 408
},
{
"epoch": 2.16,
"learning_rate": 3.816365754589845e-06,
"loss": 0.0205,
"step": 409
},
{
"epoch": 2.17,
"learning_rate": 3.7714954146478022e-06,
"loss": 0.0176,
"step": 410
},
{
"epoch": 2.17,
"learning_rate": 3.7268290312570622e-06,
"loss": 0.0172,
"step": 411
},
{
"epoch": 2.18,
"learning_rate": 3.6823680670482487e-06,
"loss": 0.0203,
"step": 412
},
{
"epoch": 2.19,
"learning_rate": 3.638113977925387e-06,
"loss": 0.0198,
"step": 413
},
{
"epoch": 2.19,
"learning_rate": 3.594068213018249e-06,
"loss": 0.0195,
"step": 414
},
{
"epoch": 2.2,
"learning_rate": 3.5502322146348843e-06,
"loss": 0.0174,
"step": 415
},
{
"epoch": 2.2,
"learning_rate": 3.506607418214395e-06,
"loss": 0.0197,
"step": 416
},
{
"epoch": 2.21,
"learning_rate": 3.4631952522799396e-06,
"loss": 0.0214,
"step": 417
},
{
"epoch": 2.21,
"learning_rate": 3.4199971383919538e-06,
"loss": 0.0183,
"step": 418
},
{
"epoch": 2.22,
"learning_rate": 3.377014491101577e-06,
"loss": 0.0174,
"step": 419
},
{
"epoch": 2.22,
"learning_rate": 3.334248717904368e-06,
"loss": 0.0246,
"step": 420
},
{
"epoch": 2.23,
"learning_rate": 3.2917012191941955e-06,
"loss": 0.0208,
"step": 421
},
{
"epoch": 2.23,
"learning_rate": 3.2493733882173794e-06,
"loss": 0.0194,
"step": 422
},
{
"epoch": 2.24,
"learning_rate": 3.207266611027069e-06,
"loss": 0.0192,
"step": 423
},
{
"epoch": 2.24,
"learning_rate": 3.165382266437874e-06,
"loss": 0.0249,
"step": 424
},
{
"epoch": 2.25,
"learning_rate": 3.123721725980683e-06,
"loss": 0.0205,
"step": 425
},
{
"epoch": 2.25,
"learning_rate": 3.082286353857782e-06,
"loss": 0.026,
"step": 426
},
{
"epoch": 2.26,
"learning_rate": 3.0410775068981615e-06,
"loss": 0.0207,
"step": 427
},
{
"epoch": 2.26,
"learning_rate": 3.0000965345130904e-06,
"loss": 0.0197,
"step": 428
},
{
"epoch": 2.27,
"learning_rate": 2.9593447786519424e-06,
"loss": 0.0186,
"step": 429
},
{
"epoch": 2.28,
"learning_rate": 2.9188235737582416e-06,
"loss": 0.0208,
"step": 430
},
{
"epoch": 2.28,
"learning_rate": 2.8785342467259568e-06,
"loss": 0.015,
"step": 431
},
{
"epoch": 2.29,
"learning_rate": 2.8384781168560693e-06,
"loss": 0.019,
"step": 432
},
{
"epoch": 2.29,
"learning_rate": 2.7986564958133564e-06,
"loss": 0.0151,
"step": 433
},
{
"epoch": 2.3,
"learning_rate": 2.7590706875834563e-06,
"loss": 0.0168,
"step": 434
},
{
"epoch": 2.3,
"learning_rate": 2.719721988430153e-06,
"loss": 0.0186,
"step": 435
},
{
"epoch": 2.31,
"learning_rate": 2.6806116868529364e-06,
"loss": 0.0193,
"step": 436
},
{
"epoch": 2.31,
"learning_rate": 2.6417410635448015e-06,
"loss": 0.0165,
"step": 437
},
{
"epoch": 2.32,
"learning_rate": 2.6031113913503337e-06,
"loss": 0.023,
"step": 438
},
{
"epoch": 2.32,
"learning_rate": 2.5647239352239948e-06,
"loss": 0.0153,
"step": 439
},
{
"epoch": 2.33,
"learning_rate": 2.526579952188735e-06,
"loss": 0.0182,
"step": 440
},
{
"epoch": 2.33,
"learning_rate": 2.4886806912948034e-06,
"loss": 0.0179,
"step": 441
},
{
"epoch": 2.34,
"learning_rate": 2.4510273935788632e-06,
"loss": 0.0173,
"step": 442
},
{
"epoch": 2.34,
"learning_rate": 2.413621292023349e-06,
"loss": 0.0151,
"step": 443
},
{
"epoch": 2.35,
"learning_rate": 2.376463611516098e-06,
"loss": 0.0167,
"step": 444
},
{
"epoch": 2.35,
"learning_rate": 2.339555568810221e-06,
"loss": 0.0219,
"step": 445
},
{
"epoch": 2.36,
"learning_rate": 2.302898372484278e-06,
"loss": 0.0185,
"step": 446
},
{
"epoch": 2.37,
"learning_rate": 2.2664932229027025e-06,
"loss": 0.0175,
"step": 447
},
{
"epoch": 2.37,
"learning_rate": 2.230341312176476e-06,
"loss": 0.0196,
"step": 448
},
{
"epoch": 2.38,
"learning_rate": 2.1944438241241185e-06,
"loss": 0.0213,
"step": 449
},
{
"epoch": 2.38,
"learning_rate": 2.158801934232897e-06,
"loss": 0.0226,
"step": 450
},
{
"epoch": 2.39,
"learning_rate": 2.123416809620351e-06,
"loss": 0.0184,
"step": 451
},
{
"epoch": 2.39,
"learning_rate": 2.0882896089960713e-06,
"loss": 0.0178,
"step": 452
},
{
"epoch": 2.4,
"learning_rate": 2.0534214826237486e-06,
"loss": 0.0144,
"step": 453
},
{
"epoch": 2.4,
"learning_rate": 2.0188135722835233e-06,
"loss": 0.019,
"step": 454
},
{
"epoch": 2.41,
"learning_rate": 1.9844670112345787e-06,
"loss": 0.0195,
"step": 455
},
{
"epoch": 2.41,
"learning_rate": 1.9503829241780416e-06,
"loss": 0.0179,
"step": 456
},
{
"epoch": 2.42,
"learning_rate": 1.9165624272201567e-06,
"loss": 0.0185,
"step": 457
},
{
"epoch": 2.42,
"learning_rate": 1.8830066278357395e-06,
"loss": 0.0179,
"step": 458
},
{
"epoch": 2.43,
"learning_rate": 1.8497166248318876e-06,
"loss": 0.0199,
"step": 459
},
{
"epoch": 2.43,
"learning_rate": 1.8166935083120351e-06,
"loss": 0.0183,
"step": 460
},
{
"epoch": 2.44,
"learning_rate": 1.7839383596402382e-06,
"loss": 0.0208,
"step": 461
},
{
"epoch": 2.44,
"learning_rate": 1.7514522514057552e-06,
"loss": 0.0165,
"step": 462
},
{
"epoch": 2.45,
"learning_rate": 1.719236247387951e-06,
"loss": 0.0257,
"step": 463
},
{
"epoch": 2.46,
"learning_rate": 1.6872914025214338e-06,
"loss": 0.0193,
"step": 464
},
{
"epoch": 2.46,
"learning_rate": 1.6556187628615273e-06,
"loss": 0.0209,
"step": 465
},
{
"epoch": 2.47,
"learning_rate": 1.6242193655500182e-06,
"loss": 0.0184,
"step": 466
},
{
"epoch": 2.47,
"learning_rate": 1.593094238781191e-06,
"loss": 0.0184,
"step": 467
},
{
"epoch": 2.48,
"learning_rate": 1.5622444017681438e-06,
"loss": 0.0141,
"step": 468
},
{
"epoch": 2.48,
"learning_rate": 1.5316708647094448e-06,
"loss": 0.0166,
"step": 469
},
{
"epoch": 2.49,
"learning_rate": 1.5013746287560227e-06,
"loss": 0.0181,
"step": 470
},
{
"epoch": 2.49,
"learning_rate": 1.4713566859784045e-06,
"loss": 0.0197,
"step": 471
},
{
"epoch": 2.5,
"learning_rate": 1.4416180193342143e-06,
"loss": 0.0203,
"step": 472
},
{
"epoch": 2.5,
"learning_rate": 1.4121596026359951e-06,
"loss": 0.0214,
"step": 473
},
{
"epoch": 2.51,
"learning_rate": 1.3829824005193183e-06,
"loss": 0.0174,
"step": 474
},
{
"epoch": 2.51,
"learning_rate": 1.3540873684111977e-06,
"loss": 0.0164,
"step": 475
},
{
"epoch": 2.52,
"learning_rate": 1.3254754524987956e-06,
"loss": 0.0165,
"step": 476
},
{
"epoch": 2.52,
"learning_rate": 1.2971475896984475e-06,
"loss": 0.0178,
"step": 477
},
{
"epoch": 2.53,
"learning_rate": 1.2691047076249852e-06,
"loss": 0.02,
"step": 478
},
{
"epoch": 2.53,
"learning_rate": 1.2413477245613438e-06,
"loss": 0.0163,
"step": 479
},
{
"epoch": 2.54,
"learning_rate": 1.2138775494285181e-06,
"loss": 0.0179,
"step": 480
},
{
"epoch": 2.54,
"learning_rate": 1.1866950817557743e-06,
"loss": 0.0181,
"step": 481
},
{
"epoch": 2.55,
"learning_rate": 1.1598012116512092e-06,
"loss": 0.0249,
"step": 482
},
{
"epoch": 2.56,
"learning_rate": 1.1331968197725985e-06,
"loss": 0.0205,
"step": 483
},
{
"epoch": 2.56,
"learning_rate": 1.1068827772985645e-06,
"loss": 0.0191,
"step": 484
},
{
"epoch": 2.57,
"learning_rate": 1.0808599459000368e-06,
"loss": 0.0184,
"step": 485
},
{
"epoch": 2.57,
"learning_rate": 1.0551291777120465e-06,
"loss": 0.0189,
"step": 486
},
{
"epoch": 2.58,
"learning_rate": 1.0296913153058164e-06,
"loss": 0.0166,
"step": 487
},
{
"epoch": 2.58,
"learning_rate": 1.004547191661178e-06,
"loss": 0.021,
"step": 488
},
{
"epoch": 2.59,
"learning_rate": 9.796976301392935e-07,
"loss": 0.0182,
"step": 489
},
{
"epoch": 2.59,
"learning_rate": 9.551434444556807e-07,
"loss": 0.0205,
"step": 490
},
{
"epoch": 2.6,
"learning_rate": 9.308854386535849e-07,
"loss": 0.0206,
"step": 491
},
{
"epoch": 2.6,
"learning_rate": 9.069244070776428e-07,
"loss": 0.0157,
"step": 492
},
{
"epoch": 2.61,
"learning_rate": 8.832611343478681e-07,
"loss": 0.0174,
"step": 493
},
{
"epoch": 2.61,
"learning_rate": 8.598963953339667e-07,
"loss": 0.0246,
"step": 494
},
{
"epoch": 2.62,
"learning_rate": 8.368309551299536e-07,
"loss": 0.0166,
"step": 495
},
{
"epoch": 2.62,
"learning_rate": 8.140655690291044e-07,
"loss": 0.0184,
"step": 496
},
{
"epoch": 2.63,
"learning_rate": 7.916009824992255e-07,
"loss": 0.0168,
"step": 497
},
{
"epoch": 2.63,
"learning_rate": 7.694379311582401e-07,
"loss": 0.0189,
"step": 498
},
{
"epoch": 2.64,
"learning_rate": 7.475771407500943e-07,
"loss": 0.0186,
"step": 499
},
{
"epoch": 2.65,
"learning_rate": 7.260193271210003e-07,
"loss": 0.0188,
"step": 500
},
{
"epoch": 2.65,
"learning_rate": 7.047651961959978e-07,
"loss": 0.0164,
"step": 501
},
{
"epoch": 2.66,
"learning_rate": 6.838154439558254e-07,
"loss": 0.0197,
"step": 502
},
{
"epoch": 2.66,
"learning_rate": 6.631707564141454e-07,
"loss": 0.0179,
"step": 503
},
{
"epoch": 2.67,
"learning_rate": 6.428318095950648e-07,
"loss": 0.0219,
"step": 504
},
{
"epoch": 2.67,
"learning_rate": 6.227992695110097e-07,
"loss": 0.0147,
"step": 505
},
{
"epoch": 2.68,
"learning_rate": 6.030737921409169e-07,
"loss": 0.0172,
"step": 506
},
{
"epoch": 2.68,
"learning_rate": 5.836560234087418e-07,
"loss": 0.0206,
"step": 507
},
{
"epoch": 2.69,
"learning_rate": 5.645465991623167e-07,
"loss": 0.0176,
"step": 508
},
{
"epoch": 2.69,
"learning_rate": 5.457461451525315e-07,
"loss": 0.0187,
"step": 509
},
{
"epoch": 2.7,
"learning_rate": 5.272552770128314e-07,
"loss": 0.0183,
"step": 510
},
{
"epoch": 2.7,
"learning_rate": 5.090746002390734e-07,
"loss": 0.018,
"step": 511
},
{
"epoch": 2.71,
"learning_rate": 4.912047101696848e-07,
"loss": 0.018,
"step": 512
},
{
"epoch": 2.71,
"learning_rate": 4.73646191966175e-07,
"loss": 0.0162,
"step": 513
},
{
"epoch": 2.72,
"learning_rate": 4.563996205939747e-07,
"loss": 0.0189,
"step": 514
},
{
"epoch": 2.72,
"learning_rate": 4.3946556080360916e-07,
"loss": 0.0156,
"step": 515
},
{
"epoch": 2.73,
"learning_rate": 4.2284456711219723e-07,
"loss": 0.017,
"step": 516
},
{
"epoch": 2.74,
"learning_rate": 4.065371837853016e-07,
"loss": 0.0181,
"step": 517
},
{
"epoch": 2.74,
"learning_rate": 3.9054394481910507e-07,
"loss": 0.0167,
"step": 518
},
{
"epoch": 2.75,
"learning_rate": 3.748653739229191e-07,
"loss": 0.0198,
"step": 519
},
{
"epoch": 2.75,
"learning_rate": 3.595019845020442e-07,
"loss": 0.0162,
"step": 520
},
{
"epoch": 2.76,
"learning_rate": 3.444542796409478e-07,
"loss": 0.0151,
"step": 521
},
{
"epoch": 2.76,
"learning_rate": 3.2972275208679625e-07,
"loss": 0.0237,
"step": 522
},
{
"epoch": 2.77,
"learning_rate": 3.1530788423332124e-07,
"loss": 0.0172,
"step": 523
},
{
"epoch": 2.77,
"learning_rate": 3.012101481050156e-07,
"loss": 0.0224,
"step": 524
},
{
"epoch": 2.78,
"learning_rate": 2.8743000534168673e-07,
"loss": 0.0172,
"step": 525
},
{
"epoch": 2.78,
"learning_rate": 2.739679071833301e-07,
"loss": 0.0207,
"step": 526
},
{
"epoch": 2.79,
"learning_rate": 2.608242944553607e-07,
"loss": 0.0183,
"step": 527
},
{
"epoch": 2.79,
"learning_rate": 2.479995975541749e-07,
"loss": 0.0219,
"step": 528
},
{
"epoch": 2.8,
"learning_rate": 2.354942364330559e-07,
"loss": 0.0169,
"step": 529
},
{
"epoch": 2.8,
"learning_rate": 2.2330862058842273e-07,
"loss": 0.023,
"step": 530
},
{
"epoch": 2.81,
"learning_rate": 2.1144314904642194e-07,
"loss": 0.0138,
"step": 531
},
{
"epoch": 2.81,
"learning_rate": 1.9989821034986034e-07,
"loss": 0.0168,
"step": 532
},
{
"epoch": 2.82,
"learning_rate": 1.8867418254548298e-07,
"loss": 0.0163,
"step": 533
},
{
"epoch": 2.83,
"learning_rate": 1.7777143317159407e-07,
"loss": 0.0175,
"step": 534
},
{
"epoch": 2.83,
"learning_rate": 1.6719031924601558e-07,
"loss": 0.0172,
"step": 535
},
{
"epoch": 2.84,
"learning_rate": 1.569311872544066e-07,
"loss": 0.017,
"step": 536
},
{
"epoch": 2.84,
"learning_rate": 1.4699437313891007e-07,
"loss": 0.0166,
"step": 537
},
{
"epoch": 2.85,
"learning_rate": 1.373802022871551e-07,
"loss": 0.0165,
"step": 538
},
{
"epoch": 2.85,
"learning_rate": 1.2808898952160198e-07,
"loss": 0.018,
"step": 539
},
{
"epoch": 2.86,
"learning_rate": 1.1912103908922945e-07,
"loss": 0.0187,
"step": 540
},
{
"epoch": 2.86,
"learning_rate": 1.1047664465157592e-07,
"loss": 0.017,
"step": 541
},
{
"epoch": 2.87,
"learning_rate": 1.02156089275125e-07,
"loss": 0.0166,
"step": 542
},
{
"epoch": 2.87,
"learning_rate": 9.415964542203059e-08,
"loss": 0.0166,
"step": 543
},
{
"epoch": 2.88,
"learning_rate": 8.648757494119752e-08,
"loss": 0.0169,
"step": 544
},
{
"epoch": 2.88,
"learning_rate": 7.914012905970936e-08,
"loss": 0.02,
"step": 545
},
{
"epoch": 2.89,
"learning_rate": 7.21175483745995e-08,
"loss": 0.016,
"step": 546
},
{
"epoch": 2.89,
"learning_rate": 6.542006284497304e-08,
"loss": 0.0163,
"step": 547
},
{
"epoch": 2.9,
"learning_rate": 5.90478917844739e-08,
"loss": 0.0228,
"step": 548
},
{
"epoch": 2.9,
"learning_rate": 5.300124385410943e-08,
"loss": 0.0191,
"step": 549
},
{
"epoch": 2.91,
"learning_rate": 4.728031705541369e-08,
"loss": 0.0151,
"step": 550
},
{
"epoch": 2.92,
"learning_rate": 4.188529872396374e-08,
"loss": 0.0164,
"step": 551
},
{
"epoch": 2.92,
"learning_rate": 3.681636552324452e-08,
"loss": 0.0179,
"step": 552
},
{
"epoch": 2.93,
"learning_rate": 3.2073683438866856e-08,
"loss": 0.0175,
"step": 553
},
{
"epoch": 2.93,
"learning_rate": 2.765740777313064e-08,
"loss": 0.02,
"step": 554
},
{
"epoch": 2.94,
"learning_rate": 2.3567683139936736e-08,
"loss": 0.0158,
"step": 555
},
{
"epoch": 2.94,
"learning_rate": 1.9804643460056284e-08,
"loss": 0.0229,
"step": 556
},
{
"epoch": 2.95,
"learning_rate": 1.636841195674199e-08,
"loss": 0.0166,
"step": 557
},
{
"epoch": 2.95,
"learning_rate": 1.325910115169471e-08,
"loss": 0.0188,
"step": 558
},
{
"epoch": 2.96,
"learning_rate": 1.0476812861377471e-08,
"loss": 0.0186,
"step": 559
},
{
"epoch": 2.96,
"learning_rate": 8.021638193682624e-09,
"loss": 0.0188,
"step": 560
},
{
"epoch": 2.97,
"learning_rate": 5.8936575449475284e-09,
"loss": 0.0214,
"step": 561
},
{
"epoch": 2.97,
"learning_rate": 4.092940597322237e-09,
"loss": 0.0174,
"step": 562
},
{
"epoch": 2.98,
"learning_rate": 2.6195463164901956e-09,
"loss": 0.0193,
"step": 563
},
{
"epoch": 2.98,
"learning_rate": 1.47352294973091e-09,
"loss": 0.0151,
"step": 564
},
{
"epoch": 2.99,
"learning_rate": 6.54908024348977e-10,
"loss": 0.0151,
"step": 565
},
{
"epoch": 2.99,
"learning_rate": 1.6372834643618718e-10,
"loss": 0.0218,
"step": 566
},
{
"epoch": 3.0,
"learning_rate": 0.0,
"loss": 0.018,
"step": 567
},
{
"epoch": 3.0,
"step": 567,
"total_flos": 0.0,
"train_loss": 0.030306552032524525,
"train_runtime": 9233.9434,
"train_samples_per_second": 7.819,
"train_steps_per_second": 0.061
}
],
"logging_steps": 1.0,
"max_steps": 567,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 500,
"total_flos": 0.0,
"train_batch_size": 16,
"trial_name": null,
"trial_params": null
}