ShapeLLM_7B_general_v1.0 / trainer_state.json
alexhafnerd's picture
upload checkpoints
84eb193
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"global_step": 356,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0,
"learning_rate": 1.8181818181818183e-06,
"loss": 2.8242,
"step": 1
},
{
"epoch": 0.01,
"learning_rate": 3.6363636363636366e-06,
"loss": 2.834,
"step": 2
},
{
"epoch": 0.01,
"learning_rate": 5.4545454545454545e-06,
"loss": 2.7539,
"step": 3
},
{
"epoch": 0.01,
"learning_rate": 7.272727272727273e-06,
"loss": 2.2236,
"step": 4
},
{
"epoch": 0.01,
"learning_rate": 9.090909090909091e-06,
"loss": 1.5918,
"step": 5
},
{
"epoch": 0.02,
"learning_rate": 1.0909090909090909e-05,
"loss": 1.3584,
"step": 6
},
{
"epoch": 0.02,
"learning_rate": 1.2727272727272728e-05,
"loss": 1.4824,
"step": 7
},
{
"epoch": 0.02,
"learning_rate": 1.4545454545454546e-05,
"loss": 1.2432,
"step": 8
},
{
"epoch": 0.03,
"learning_rate": 1.6363636363636366e-05,
"loss": 1.4717,
"step": 9
},
{
"epoch": 0.03,
"learning_rate": 1.8181818181818182e-05,
"loss": 1.3438,
"step": 10
},
{
"epoch": 0.03,
"learning_rate": 2e-05,
"loss": 1.1885,
"step": 11
},
{
"epoch": 0.03,
"learning_rate": 1.9999585400705652e-05,
"loss": 1.2236,
"step": 12
},
{
"epoch": 0.04,
"learning_rate": 1.9998341637201124e-05,
"loss": 1.1631,
"step": 13
},
{
"epoch": 0.04,
"learning_rate": 1.999626881261911e-05,
"loss": 1.2119,
"step": 14
},
{
"epoch": 0.04,
"learning_rate": 1.9993367098837927e-05,
"loss": 1.1982,
"step": 15
},
{
"epoch": 0.04,
"learning_rate": 1.9989636736467278e-05,
"loss": 1.1035,
"step": 16
},
{
"epoch": 0.05,
"learning_rate": 1.998507803482828e-05,
"loss": 1.1538,
"step": 17
},
{
"epoch": 0.05,
"learning_rate": 1.9979691371927834e-05,
"loss": 1.0952,
"step": 18
},
{
"epoch": 0.05,
"learning_rate": 1.9973477194427267e-05,
"loss": 1.1509,
"step": 19
},
{
"epoch": 0.06,
"learning_rate": 1.9966436017605296e-05,
"loss": 1.1001,
"step": 20
},
{
"epoch": 0.06,
"learning_rate": 1.9958568425315316e-05,
"loss": 1.2148,
"step": 21
},
{
"epoch": 0.06,
"learning_rate": 1.994987506993696e-05,
"loss": 1.0889,
"step": 22
},
{
"epoch": 0.06,
"learning_rate": 1.9940356672322037e-05,
"loss": 1.0859,
"step": 23
},
{
"epoch": 0.07,
"learning_rate": 1.9930014021734732e-05,
"loss": 1.0317,
"step": 24
},
{
"epoch": 0.07,
"learning_rate": 1.9918847975786172e-05,
"loss": 1.1777,
"step": 25
},
{
"epoch": 0.07,
"learning_rate": 1.9906859460363307e-05,
"loss": 1.0396,
"step": 26
},
{
"epoch": 0.08,
"learning_rate": 1.9894049469552152e-05,
"loss": 1.0405,
"step": 27
},
{
"epoch": 0.08,
"learning_rate": 1.988041906555533e-05,
"loss": 1.0356,
"step": 28
},
{
"epoch": 0.08,
"learning_rate": 1.9865969378604023e-05,
"loss": 1.0181,
"step": 29
},
{
"epoch": 0.08,
"learning_rate": 1.9850701606864225e-05,
"loss": 1.0444,
"step": 30
},
{
"epoch": 0.09,
"learning_rate": 1.9834617016337424e-05,
"loss": 1.0293,
"step": 31
},
{
"epoch": 0.09,
"learning_rate": 1.9817716940755586e-05,
"loss": 1.0566,
"step": 32
},
{
"epoch": 0.09,
"learning_rate": 1.98000027814706e-05,
"loss": 1.0029,
"step": 33
},
{
"epoch": 0.1,
"learning_rate": 1.9781476007338058e-05,
"loss": 1.0503,
"step": 34
},
{
"epoch": 0.1,
"learning_rate": 1.9762138154595448e-05,
"loss": 1.0342,
"step": 35
},
{
"epoch": 0.1,
"learning_rate": 1.9741990826734793e-05,
"loss": 1.0,
"step": 36
},
{
"epoch": 0.1,
"learning_rate": 1.9721035694369674e-05,
"loss": 1.0698,
"step": 37
},
{
"epoch": 0.11,
"learning_rate": 1.9699274495096712e-05,
"loss": 0.9609,
"step": 38
},
{
"epoch": 0.11,
"learning_rate": 1.967670903335148e-05,
"loss": 1.0439,
"step": 39
},
{
"epoch": 0.11,
"learning_rate": 1.9653341180258883e-05,
"loss": 0.958,
"step": 40
},
{
"epoch": 0.12,
"learning_rate": 1.9629172873477995e-05,
"loss": 0.9707,
"step": 41
},
{
"epoch": 0.12,
"learning_rate": 1.9604206117041408e-05,
"loss": 0.9385,
"step": 42
},
{
"epoch": 0.12,
"learning_rate": 1.957844298118904e-05,
"loss": 0.937,
"step": 43
},
{
"epoch": 0.12,
"learning_rate": 1.9551885602196482e-05,
"loss": 1.0112,
"step": 44
},
{
"epoch": 0.13,
"learning_rate": 1.9524536182197852e-05,
"loss": 1.0264,
"step": 45
},
{
"epoch": 0.13,
"learning_rate": 1.9496396989003195e-05,
"loss": 1.0469,
"step": 46
},
{
"epoch": 0.13,
"learning_rate": 1.9467470355910438e-05,
"loss": 0.9346,
"step": 47
},
{
"epoch": 0.13,
"learning_rate": 1.943775868151192e-05,
"loss": 0.9556,
"step": 48
},
{
"epoch": 0.14,
"learning_rate": 1.9407264429495484e-05,
"loss": 0.9302,
"step": 49
},
{
"epoch": 0.14,
"learning_rate": 1.9375990128440205e-05,
"loss": 0.9609,
"step": 50
},
{
"epoch": 0.14,
"learning_rate": 1.9343938371606714e-05,
"loss": 1.0024,
"step": 51
},
{
"epoch": 0.15,
"learning_rate": 1.931111181672216e-05,
"loss": 0.9839,
"step": 52
},
{
"epoch": 0.15,
"learning_rate": 1.9277513185759847e-05,
"loss": 0.9678,
"step": 53
},
{
"epoch": 0.15,
"learning_rate": 1.924314526471351e-05,
"loss": 0.9009,
"step": 54
},
{
"epoch": 0.15,
"learning_rate": 1.9208010903366307e-05,
"loss": 0.9878,
"step": 55
},
{
"epoch": 0.16,
"learning_rate": 1.917211301505453e-05,
"loss": 0.9521,
"step": 56
},
{
"epoch": 0.16,
"learning_rate": 1.913545457642601e-05,
"loss": 0.9897,
"step": 57
},
{
"epoch": 0.16,
"learning_rate": 1.9098038627193304e-05,
"loss": 0.9858,
"step": 58
},
{
"epoch": 0.17,
"learning_rate": 1.9059868269881637e-05,
"loss": 0.939,
"step": 59
},
{
"epoch": 0.17,
"learning_rate": 1.9020946669571656e-05,
"loss": 0.9585,
"step": 60
},
{
"epoch": 0.17,
"learning_rate": 1.8981277053636963e-05,
"loss": 0.9238,
"step": 61
},
{
"epoch": 0.17,
"learning_rate": 1.8940862711476515e-05,
"loss": 1.0298,
"step": 62
},
{
"epoch": 0.18,
"learning_rate": 1.8899706994241857e-05,
"loss": 1.0303,
"step": 63
},
{
"epoch": 0.18,
"learning_rate": 1.8857813314559258e-05,
"loss": 0.9834,
"step": 64
},
{
"epoch": 0.18,
"learning_rate": 1.8815185146246718e-05,
"loss": 1.0303,
"step": 65
},
{
"epoch": 0.19,
"learning_rate": 1.8771826024025944e-05,
"loss": 0.9111,
"step": 66
},
{
"epoch": 0.19,
"learning_rate": 1.872773954322923e-05,
"loss": 0.9678,
"step": 67
},
{
"epoch": 0.19,
"learning_rate": 1.8682929359501338e-05,
"loss": 1.0156,
"step": 68
},
{
"epoch": 0.19,
"learning_rate": 1.8637399188496384e-05,
"loss": 0.9521,
"step": 69
},
{
"epoch": 0.2,
"learning_rate": 1.8591152805569717e-05,
"loss": 0.9429,
"step": 70
},
{
"epoch": 0.2,
"learning_rate": 1.8544194045464888e-05,
"loss": 0.9609,
"step": 71
},
{
"epoch": 0.2,
"learning_rate": 1.8496526801995652e-05,
"loss": 0.9873,
"step": 72
},
{
"epoch": 0.21,
"learning_rate": 1.844815502772311e-05,
"loss": 1.0288,
"step": 73
},
{
"epoch": 0.21,
"learning_rate": 1.8399082733627967e-05,
"loss": 1.0054,
"step": 74
},
{
"epoch": 0.21,
"learning_rate": 1.8349313988777914e-05,
"loss": 1.019,
"step": 75
},
{
"epoch": 0.21,
"learning_rate": 1.8298852919990254e-05,
"loss": 0.9346,
"step": 76
},
{
"epoch": 0.22,
"learning_rate": 1.8247703711489684e-05,
"loss": 0.999,
"step": 77
},
{
"epoch": 0.22,
"learning_rate": 1.8195870604561365e-05,
"loss": 1.0156,
"step": 78
},
{
"epoch": 0.22,
"learning_rate": 1.81433578971992e-05,
"loss": 0.978,
"step": 79
},
{
"epoch": 0.22,
"learning_rate": 1.8090169943749477e-05,
"loss": 0.8535,
"step": 80
},
{
"epoch": 0.23,
"learning_rate": 1.8036311154549783e-05,
"loss": 0.9478,
"step": 81
},
{
"epoch": 0.23,
"learning_rate": 1.7981785995563322e-05,
"loss": 0.9556,
"step": 82
},
{
"epoch": 0.23,
"learning_rate": 1.7926598988008584e-05,
"loss": 0.9126,
"step": 83
},
{
"epoch": 0.24,
"learning_rate": 1.7870754707984445e-05,
"loss": 0.9321,
"step": 84
},
{
"epoch": 0.24,
"learning_rate": 1.7814257786090718e-05,
"loss": 0.9756,
"step": 85
},
{
"epoch": 0.24,
"learning_rate": 1.77571129070442e-05,
"loss": 0.9541,
"step": 86
},
{
"epoch": 0.24,
"learning_rate": 1.7699324809290193e-05,
"loss": 0.918,
"step": 87
},
{
"epoch": 0.25,
"learning_rate": 1.764089828460961e-05,
"loss": 0.8853,
"step": 88
},
{
"epoch": 0.25,
"learning_rate": 1.758183817772163e-05,
"loss": 0.9038,
"step": 89
},
{
"epoch": 0.25,
"learning_rate": 1.752214938588198e-05,
"loss": 0.9087,
"step": 90
},
{
"epoch": 0.26,
"learning_rate": 1.7461836858476858e-05,
"loss": 0.9287,
"step": 91
},
{
"epoch": 0.26,
"learning_rate": 1.740090559661252e-05,
"loss": 1.0015,
"step": 92
},
{
"epoch": 0.26,
"learning_rate": 1.7339360652700607e-05,
"loss": 0.9858,
"step": 93
},
{
"epoch": 0.26,
"learning_rate": 1.7277207130039175e-05,
"loss": 0.9595,
"step": 94
},
{
"epoch": 0.27,
"learning_rate": 1.721445018238956e-05,
"loss": 0.9302,
"step": 95
},
{
"epoch": 0.27,
"learning_rate": 1.7151095013548996e-05,
"loss": 0.938,
"step": 96
},
{
"epoch": 0.27,
"learning_rate": 1.7087146876919146e-05,
"loss": 0.9009,
"step": 97
},
{
"epoch": 0.28,
"learning_rate": 1.7022611075070476e-05,
"loss": 0.9883,
"step": 98
},
{
"epoch": 0.28,
"learning_rate": 1.695749295930256e-05,
"loss": 0.978,
"step": 99
},
{
"epoch": 0.28,
"learning_rate": 1.6891797929200376e-05,
"loss": 0.8745,
"step": 100
},
{
"epoch": 0.28,
"learning_rate": 1.6825531432186545e-05,
"loss": 0.9727,
"step": 101
},
{
"epoch": 0.29,
"learning_rate": 1.6758698963069645e-05,
"loss": 0.9131,
"step": 102
},
{
"epoch": 0.29,
"learning_rate": 1.6691306063588583e-05,
"loss": 0.9126,
"step": 103
},
{
"epoch": 0.29,
"learning_rate": 1.662335832195308e-05,
"loss": 0.9795,
"step": 104
},
{
"epoch": 0.29,
"learning_rate": 1.6554861372380274e-05,
"loss": 0.8843,
"step": 105
},
{
"epoch": 0.3,
"learning_rate": 1.648582089462756e-05,
"loss": 0.9243,
"step": 106
},
{
"epoch": 0.3,
"learning_rate": 1.6416242613521612e-05,
"loss": 0.9497,
"step": 107
},
{
"epoch": 0.3,
"learning_rate": 1.6346132298483677e-05,
"loss": 0.9829,
"step": 108
},
{
"epoch": 0.31,
"learning_rate": 1.6275495763051186e-05,
"loss": 0.9658,
"step": 109
},
{
"epoch": 0.31,
"learning_rate": 1.6204338864395683e-05,
"loss": 0.9199,
"step": 110
},
{
"epoch": 0.31,
"learning_rate": 1.6132667502837164e-05,
"loss": 0.9312,
"step": 111
},
{
"epoch": 0.31,
"learning_rate": 1.6060487621354816e-05,
"loss": 0.9243,
"step": 112
},
{
"epoch": 0.32,
"learning_rate": 1.5987805205094225e-05,
"loss": 0.9512,
"step": 113
},
{
"epoch": 0.32,
"learning_rate": 1.591462628087109e-05,
"loss": 0.8477,
"step": 114
},
{
"epoch": 0.32,
"learning_rate": 1.5840956916671476e-05,
"loss": 0.9165,
"step": 115
},
{
"epoch": 0.33,
"learning_rate": 1.5766803221148676e-05,
"loss": 0.9038,
"step": 116
},
{
"epoch": 0.33,
"learning_rate": 1.569217134311664e-05,
"loss": 0.9448,
"step": 117
},
{
"epoch": 0.33,
"learning_rate": 1.5617067471040175e-05,
"loss": 0.9141,
"step": 118
},
{
"epoch": 0.33,
"learning_rate": 1.554149783252175e-05,
"loss": 0.8706,
"step": 119
},
{
"epoch": 0.34,
"learning_rate": 1.5465468693785123e-05,
"loss": 0.9282,
"step": 120
},
{
"epoch": 0.34,
"learning_rate": 1.538898635915576e-05,
"loss": 0.8921,
"step": 121
},
{
"epoch": 0.34,
"learning_rate": 1.5312057170538033e-05,
"loss": 0.8599,
"step": 122
},
{
"epoch": 0.35,
"learning_rate": 1.5234687506889428e-05,
"loss": 0.9805,
"step": 123
},
{
"epoch": 0.35,
"learning_rate": 1.5156883783691521e-05,
"loss": 0.8926,
"step": 124
},
{
"epoch": 0.35,
"learning_rate": 1.5078652452418063e-05,
"loss": 0.9463,
"step": 125
},
{
"epoch": 0.35,
"learning_rate": 1.5000000000000002e-05,
"loss": 0.9111,
"step": 126
},
{
"epoch": 0.36,
"learning_rate": 1.4920932948287595e-05,
"loss": 0.9194,
"step": 127
},
{
"epoch": 0.36,
"learning_rate": 1.4841457853509606e-05,
"loss": 0.9331,
"step": 128
},
{
"epoch": 0.36,
"learning_rate": 1.4761581305729684e-05,
"loss": 0.8857,
"step": 129
},
{
"epoch": 0.37,
"learning_rate": 1.4681309928299895e-05,
"loss": 0.9297,
"step": 130
},
{
"epoch": 0.37,
"learning_rate": 1.4600650377311523e-05,
"loss": 0.915,
"step": 131
},
{
"epoch": 0.37,
"learning_rate": 1.4519609341043158e-05,
"loss": 1.0039,
"step": 132
},
{
"epoch": 0.37,
"learning_rate": 1.443819353940609e-05,
"loss": 0.9072,
"step": 133
},
{
"epoch": 0.38,
"learning_rate": 1.4356409723387092e-05,
"loss": 0.9746,
"step": 134
},
{
"epoch": 0.38,
"learning_rate": 1.4274264674488658e-05,
"loss": 0.8989,
"step": 135
},
{
"epoch": 0.38,
"learning_rate": 1.4191765204166643e-05,
"loss": 0.9282,
"step": 136
},
{
"epoch": 0.38,
"learning_rate": 1.4108918153265485e-05,
"loss": 0.9512,
"step": 137
},
{
"epoch": 0.39,
"learning_rate": 1.4025730391450946e-05,
"loss": 0.8979,
"step": 138
},
{
"epoch": 0.39,
"learning_rate": 1.3942208816640506e-05,
"loss": 0.833,
"step": 139
},
{
"epoch": 0.39,
"learning_rate": 1.3858360354431355e-05,
"loss": 0.8691,
"step": 140
},
{
"epoch": 0.4,
"learning_rate": 1.3774191957526144e-05,
"loss": 0.8701,
"step": 141
},
{
"epoch": 0.4,
"learning_rate": 1.3689710605156471e-05,
"loss": 0.8765,
"step": 142
},
{
"epoch": 0.4,
"learning_rate": 1.3604923302504146e-05,
"loss": 0.9688,
"step": 143
},
{
"epoch": 0.4,
"learning_rate": 1.3519837080120347e-05,
"loss": 0.9336,
"step": 144
},
{
"epoch": 0.41,
"learning_rate": 1.3434458993342613e-05,
"loss": 0.8784,
"step": 145
},
{
"epoch": 0.41,
"learning_rate": 1.3348796121709862e-05,
"loss": 0.8716,
"step": 146
},
{
"epoch": 0.41,
"learning_rate": 1.3262855568375316e-05,
"loss": 0.9146,
"step": 147
},
{
"epoch": 0.42,
"learning_rate": 1.3176644459517528e-05,
"loss": 0.8872,
"step": 148
},
{
"epoch": 0.42,
"learning_rate": 1.3090169943749475e-05,
"loss": 0.8408,
"step": 149
},
{
"epoch": 0.42,
"learning_rate": 1.3003439191525806e-05,
"loss": 0.9399,
"step": 150
},
{
"epoch": 0.42,
"learning_rate": 1.291645939454825e-05,
"loss": 0.9331,
"step": 151
},
{
"epoch": 0.43,
"learning_rate": 1.28292377651693e-05,
"loss": 0.9634,
"step": 152
},
{
"epoch": 0.43,
"learning_rate": 1.2741781535794155e-05,
"loss": 0.8423,
"step": 153
},
{
"epoch": 0.43,
"learning_rate": 1.2654097958281011e-05,
"loss": 0.9224,
"step": 154
},
{
"epoch": 0.44,
"learning_rate": 1.2566194303339738e-05,
"loss": 0.8804,
"step": 155
},
{
"epoch": 0.44,
"learning_rate": 1.2478077859929e-05,
"loss": 0.8657,
"step": 156
},
{
"epoch": 0.44,
"learning_rate": 1.2389755934651851e-05,
"loss": 0.8916,
"step": 157
},
{
"epoch": 0.44,
"learning_rate": 1.2301235851149867e-05,
"loss": 0.9146,
"step": 158
},
{
"epoch": 0.45,
"learning_rate": 1.2212524949495878e-05,
"loss": 0.9595,
"step": 159
},
{
"epoch": 0.45,
"learning_rate": 1.2123630585585333e-05,
"loss": 0.9043,
"step": 160
},
{
"epoch": 0.45,
"learning_rate": 1.2034560130526341e-05,
"loss": 0.8975,
"step": 161
},
{
"epoch": 0.46,
"learning_rate": 1.1945320970028462e-05,
"loss": 0.8799,
"step": 162
},
{
"epoch": 0.46,
"learning_rate": 1.1855920503790292e-05,
"loss": 0.8838,
"step": 163
},
{
"epoch": 0.46,
"learning_rate": 1.1766366144885877e-05,
"loss": 0.9072,
"step": 164
},
{
"epoch": 0.46,
"learning_rate": 1.167666531915001e-05,
"loss": 0.8843,
"step": 165
},
{
"epoch": 0.47,
"learning_rate": 1.1586825464562515e-05,
"loss": 0.9478,
"step": 166
},
{
"epoch": 0.47,
"learning_rate": 1.1496854030631443e-05,
"loss": 0.9404,
"step": 167
},
{
"epoch": 0.47,
"learning_rate": 1.1406758477775406e-05,
"loss": 0.8779,
"step": 168
},
{
"epoch": 0.47,
"learning_rate": 1.1316546276704925e-05,
"loss": 0.9023,
"step": 169
},
{
"epoch": 0.48,
"learning_rate": 1.1226224907802986e-05,
"loss": 0.8716,
"step": 170
},
{
"epoch": 0.48,
"learning_rate": 1.113580186050475e-05,
"loss": 0.8716,
"step": 171
},
{
"epoch": 0.48,
"learning_rate": 1.1045284632676535e-05,
"loss": 0.9492,
"step": 172
},
{
"epoch": 0.49,
"learning_rate": 1.0954680729994103e-05,
"loss": 0.8755,
"step": 173
},
{
"epoch": 0.49,
"learning_rate": 1.0863997665320272e-05,
"loss": 0.9668,
"step": 174
},
{
"epoch": 0.49,
"learning_rate": 1.077324295808197e-05,
"loss": 0.9199,
"step": 175
},
{
"epoch": 0.49,
"learning_rate": 1.0682424133646712e-05,
"loss": 0.8374,
"step": 176
},
{
"epoch": 0.5,
"learning_rate": 1.0591548722698599e-05,
"loss": 0.8936,
"step": 177
},
{
"epoch": 0.5,
"learning_rate": 1.0500624260613892e-05,
"loss": 0.7964,
"step": 178
},
{
"epoch": 0.5,
"learning_rate": 1.0409658286836144e-05,
"loss": 0.9468,
"step": 179
},
{
"epoch": 0.51,
"learning_rate": 1.0318658344251066e-05,
"loss": 0.8433,
"step": 180
},
{
"epoch": 0.51,
"learning_rate": 1.0227631978561057e-05,
"loss": 0.9321,
"step": 181
},
{
"epoch": 0.51,
"learning_rate": 1.013658673765951e-05,
"loss": 0.8774,
"step": 182
},
{
"epoch": 0.51,
"learning_rate": 1.0045530171004955e-05,
"loss": 0.8882,
"step": 183
},
{
"epoch": 0.52,
"learning_rate": 9.954469828995045e-06,
"loss": 0.915,
"step": 184
},
{
"epoch": 0.52,
"learning_rate": 9.863413262340491e-06,
"loss": 0.9229,
"step": 185
},
{
"epoch": 0.52,
"learning_rate": 9.772368021438943e-06,
"loss": 0.8555,
"step": 186
},
{
"epoch": 0.53,
"learning_rate": 9.681341655748935e-06,
"loss": 0.9282,
"step": 187
},
{
"epoch": 0.53,
"learning_rate": 9.590341713163858e-06,
"loss": 0.9014,
"step": 188
},
{
"epoch": 0.53,
"learning_rate": 9.499375739386113e-06,
"loss": 0.9365,
"step": 189
},
{
"epoch": 0.53,
"learning_rate": 9.4084512773014e-06,
"loss": 0.916,
"step": 190
},
{
"epoch": 0.54,
"learning_rate": 9.317575866353293e-06,
"loss": 0.9424,
"step": 191
},
{
"epoch": 0.54,
"learning_rate": 9.226757041918033e-06,
"loss": 0.8584,
"step": 192
},
{
"epoch": 0.54,
"learning_rate": 9.136002334679732e-06,
"loss": 0.9604,
"step": 193
},
{
"epoch": 0.54,
"learning_rate": 9.0453192700059e-06,
"loss": 0.8667,
"step": 194
},
{
"epoch": 0.55,
"learning_rate": 8.954715367323468e-06,
"loss": 0.8428,
"step": 195
},
{
"epoch": 0.55,
"learning_rate": 8.86419813949525e-06,
"loss": 0.835,
"step": 196
},
{
"epoch": 0.55,
"learning_rate": 8.773775092197018e-06,
"loss": 0.9126,
"step": 197
},
{
"epoch": 0.56,
"learning_rate": 8.683453723295075e-06,
"loss": 0.9072,
"step": 198
},
{
"epoch": 0.56,
"learning_rate": 8.593241522224597e-06,
"loss": 0.9316,
"step": 199
},
{
"epoch": 0.56,
"learning_rate": 8.503145969368562e-06,
"loss": 0.7871,
"step": 200
},
{
"epoch": 0.56,
"learning_rate": 8.413174535437486e-06,
"loss": 0.8418,
"step": 201
},
{
"epoch": 0.57,
"learning_rate": 8.323334680849992e-06,
"loss": 0.8926,
"step": 202
},
{
"epoch": 0.57,
"learning_rate": 8.233633855114127e-06,
"loss": 0.8696,
"step": 203
},
{
"epoch": 0.57,
"learning_rate": 8.144079496209712e-06,
"loss": 0.8535,
"step": 204
},
{
"epoch": 0.58,
"learning_rate": 8.054679029971541e-06,
"loss": 0.8398,
"step": 205
},
{
"epoch": 0.58,
"learning_rate": 7.965439869473664e-06,
"loss": 0.9497,
"step": 206
},
{
"epoch": 0.58,
"learning_rate": 7.876369414414669e-06,
"loss": 0.9463,
"step": 207
},
{
"epoch": 0.58,
"learning_rate": 7.787475050504126e-06,
"loss": 0.9673,
"step": 208
},
{
"epoch": 0.59,
"learning_rate": 7.698764148850138e-06,
"loss": 0.915,
"step": 209
},
{
"epoch": 0.59,
"learning_rate": 7.610244065348154e-06,
"loss": 0.8491,
"step": 210
},
{
"epoch": 0.59,
"learning_rate": 7.521922140071003e-06,
"loss": 0.8657,
"step": 211
},
{
"epoch": 0.6,
"learning_rate": 7.433805696660267e-06,
"loss": 0.8628,
"step": 212
},
{
"epoch": 0.6,
"learning_rate": 7.345902041718991e-06,
"loss": 0.8755,
"step": 213
},
{
"epoch": 0.6,
"learning_rate": 7.258218464205849e-06,
"loss": 0.8677,
"step": 214
},
{
"epoch": 0.6,
"learning_rate": 7.1707622348307e-06,
"loss": 0.8418,
"step": 215
},
{
"epoch": 0.61,
"learning_rate": 7.0835406054517505e-06,
"loss": 0.8892,
"step": 216
},
{
"epoch": 0.61,
"learning_rate": 6.9965608084741955e-06,
"loss": 0.9009,
"step": 217
},
{
"epoch": 0.61,
"learning_rate": 6.909830056250527e-06,
"loss": 0.9243,
"step": 218
},
{
"epoch": 0.62,
"learning_rate": 6.823355540482475e-06,
"loss": 0.8311,
"step": 219
},
{
"epoch": 0.62,
"learning_rate": 6.737144431624687e-06,
"loss": 0.8721,
"step": 220
},
{
"epoch": 0.62,
"learning_rate": 6.651203878290139e-06,
"loss": 0.8354,
"step": 221
},
{
"epoch": 0.62,
"learning_rate": 6.565541006657387e-06,
"loss": 0.8125,
"step": 222
},
{
"epoch": 0.63,
"learning_rate": 6.4801629198796565e-06,
"loss": 0.832,
"step": 223
},
{
"epoch": 0.63,
"learning_rate": 6.395076697495854e-06,
"loss": 0.8628,
"step": 224
},
{
"epoch": 0.63,
"learning_rate": 6.310289394843529e-06,
"loss": 0.8726,
"step": 225
},
{
"epoch": 0.63,
"learning_rate": 6.225808042473857e-06,
"loss": 0.8857,
"step": 226
},
{
"epoch": 0.64,
"learning_rate": 6.141639645568646e-06,
"loss": 0.9409,
"step": 227
},
{
"epoch": 0.64,
"learning_rate": 6.057791183359496e-06,
"loss": 0.877,
"step": 228
},
{
"epoch": 0.64,
"learning_rate": 5.974269608549052e-06,
"loss": 0.7935,
"step": 229
},
{
"epoch": 0.65,
"learning_rate": 5.891081846734519e-06,
"loss": 0.8608,
"step": 230
},
{
"epoch": 0.65,
"learning_rate": 5.8082347958333625e-06,
"loss": 0.9419,
"step": 231
},
{
"epoch": 0.65,
"learning_rate": 5.725735325511344e-06,
"loss": 0.8418,
"step": 232
},
{
"epoch": 0.65,
"learning_rate": 5.643590276612909e-06,
"loss": 0.8677,
"step": 233
},
{
"epoch": 0.66,
"learning_rate": 5.5618064605939175e-06,
"loss": 0.8506,
"step": 234
},
{
"epoch": 0.66,
"learning_rate": 5.480390658956848e-06,
"loss": 0.8325,
"step": 235
},
{
"epoch": 0.66,
"learning_rate": 5.399349622688479e-06,
"loss": 0.8467,
"step": 236
},
{
"epoch": 0.67,
"learning_rate": 5.31869007170011e-06,
"loss": 0.9424,
"step": 237
},
{
"epoch": 0.67,
"learning_rate": 5.238418694270318e-06,
"loss": 0.9546,
"step": 238
},
{
"epoch": 0.67,
"learning_rate": 5.1585421464904e-06,
"loss": 0.7983,
"step": 239
},
{
"epoch": 0.67,
"learning_rate": 5.07906705171241e-06,
"loss": 0.8784,
"step": 240
},
{
"epoch": 0.68,
"learning_rate": 5.000000000000003e-06,
"loss": 0.8843,
"step": 241
},
{
"epoch": 0.68,
"learning_rate": 4.921347547581939e-06,
"loss": 0.8843,
"step": 242
},
{
"epoch": 0.68,
"learning_rate": 4.843116216308483e-06,
"loss": 0.895,
"step": 243
},
{
"epoch": 0.69,
"learning_rate": 4.765312493110578e-06,
"loss": 0.8325,
"step": 244
},
{
"epoch": 0.69,
"learning_rate": 4.687942829461969e-06,
"loss": 0.7983,
"step": 245
},
{
"epoch": 0.69,
"learning_rate": 4.611013640844245e-06,
"loss": 0.8716,
"step": 246
},
{
"epoch": 0.69,
"learning_rate": 4.534531306214878e-06,
"loss": 0.8696,
"step": 247
},
{
"epoch": 0.7,
"learning_rate": 4.458502167478254e-06,
"loss": 0.8521,
"step": 248
},
{
"epoch": 0.7,
"learning_rate": 4.38293252895983e-06,
"loss": 0.8877,
"step": 249
},
{
"epoch": 0.7,
"learning_rate": 4.307828656883362e-06,
"loss": 0.855,
"step": 250
},
{
"epoch": 0.71,
"learning_rate": 4.2331967788513295e-06,
"loss": 0.9053,
"step": 251
},
{
"epoch": 0.71,
"learning_rate": 4.159043083328521e-06,
"loss": 0.8286,
"step": 252
},
{
"epoch": 0.71,
"learning_rate": 4.08537371912891e-06,
"loss": 0.9023,
"step": 253
},
{
"epoch": 0.71,
"learning_rate": 4.0121947949057745e-06,
"loss": 0.8418,
"step": 254
},
{
"epoch": 0.72,
"learning_rate": 3.939512378645185e-06,
"loss": 0.8481,
"step": 255
},
{
"epoch": 0.72,
"learning_rate": 3.867332497162836e-06,
"loss": 0.8662,
"step": 256
},
{
"epoch": 0.72,
"learning_rate": 3.7956611356043196e-06,
"loss": 0.8965,
"step": 257
},
{
"epoch": 0.72,
"learning_rate": 3.724504236948818e-06,
"loss": 0.7817,
"step": 258
},
{
"epoch": 0.73,
"learning_rate": 3.6538677015163258e-06,
"loss": 0.8335,
"step": 259
},
{
"epoch": 0.73,
"learning_rate": 3.583757386478389e-06,
"loss": 0.896,
"step": 260
},
{
"epoch": 0.73,
"learning_rate": 3.5141791053724405e-06,
"loss": 0.8198,
"step": 261
},
{
"epoch": 0.74,
"learning_rate": 3.4451386276197294e-06,
"loss": 0.9614,
"step": 262
},
{
"epoch": 0.74,
"learning_rate": 3.376641678046926e-06,
"loss": 0.8696,
"step": 263
},
{
"epoch": 0.74,
"learning_rate": 3.308693936411421e-06,
"loss": 0.876,
"step": 264
},
{
"epoch": 0.74,
"learning_rate": 3.2413010369303587e-06,
"loss": 0.8643,
"step": 265
},
{
"epoch": 0.75,
"learning_rate": 3.174468567813461e-06,
"loss": 0.8433,
"step": 266
},
{
"epoch": 0.75,
"learning_rate": 3.1082020707996265e-06,
"loss": 0.8481,
"step": 267
},
{
"epoch": 0.75,
"learning_rate": 3.0425070406974455e-06,
"loss": 0.9443,
"step": 268
},
{
"epoch": 0.76,
"learning_rate": 2.9773889249295294e-06,
"loss": 0.8853,
"step": 269
},
{
"epoch": 0.76,
"learning_rate": 2.9128531230808578e-06,
"loss": 0.8179,
"step": 270
},
{
"epoch": 0.76,
"learning_rate": 2.8489049864510053e-06,
"loss": 0.8252,
"step": 271
},
{
"epoch": 0.76,
"learning_rate": 2.7855498176104435e-06,
"loss": 0.8608,
"step": 272
},
{
"epoch": 0.77,
"learning_rate": 2.722792869960826e-06,
"loss": 0.8267,
"step": 273
},
{
"epoch": 0.77,
"learning_rate": 2.6606393472993974e-06,
"loss": 0.9116,
"step": 274
},
{
"epoch": 0.77,
"learning_rate": 2.599094403387481e-06,
"loss": 0.8682,
"step": 275
},
{
"epoch": 0.78,
"learning_rate": 2.5381631415231455e-06,
"loss": 0.8003,
"step": 276
},
{
"epoch": 0.78,
"learning_rate": 2.4778506141180237e-06,
"loss": 0.8501,
"step": 277
},
{
"epoch": 0.78,
"learning_rate": 2.418161822278374e-06,
"loss": 0.8364,
"step": 278
},
{
"epoch": 0.78,
"learning_rate": 2.3591017153903915e-06,
"loss": 0.8452,
"step": 279
},
{
"epoch": 0.79,
"learning_rate": 2.3006751907098092e-06,
"loss": 0.897,
"step": 280
},
{
"epoch": 0.79,
"learning_rate": 2.2428870929558012e-06,
"loss": 0.8662,
"step": 281
},
{
"epoch": 0.79,
"learning_rate": 2.1857422139092867e-06,
"loss": 0.7446,
"step": 282
},
{
"epoch": 0.79,
"learning_rate": 2.1292452920155594e-06,
"loss": 0.8442,
"step": 283
},
{
"epoch": 0.8,
"learning_rate": 2.0734010119914193e-06,
"loss": 0.9033,
"step": 284
},
{
"epoch": 0.8,
"learning_rate": 2.0182140044366772e-06,
"loss": 0.8413,
"step": 285
},
{
"epoch": 0.8,
"learning_rate": 1.963688845450218e-06,
"loss": 0.8911,
"step": 286
},
{
"epoch": 0.81,
"learning_rate": 1.9098300562505266e-06,
"loss": 0.8208,
"step": 287
},
{
"epoch": 0.81,
"learning_rate": 1.8566421028008019e-06,
"loss": 0.8823,
"step": 288
},
{
"epoch": 0.81,
"learning_rate": 1.8041293954386352e-06,
"loss": 0.9609,
"step": 289
},
{
"epoch": 0.81,
"learning_rate": 1.7522962885103145e-06,
"loss": 0.8926,
"step": 290
},
{
"epoch": 0.82,
"learning_rate": 1.7011470800097496e-06,
"loss": 0.8584,
"step": 291
},
{
"epoch": 0.82,
"learning_rate": 1.6506860112220902e-06,
"loss": 0.875,
"step": 292
},
{
"epoch": 0.82,
"learning_rate": 1.6009172663720352e-06,
"loss": 0.874,
"step": 293
},
{
"epoch": 0.83,
"learning_rate": 1.5518449722768891e-06,
"loss": 0.8799,
"step": 294
},
{
"epoch": 0.83,
"learning_rate": 1.5034731980043515e-06,
"loss": 0.8911,
"step": 295
},
{
"epoch": 0.83,
"learning_rate": 1.4558059545351144e-06,
"loss": 0.8784,
"step": 296
},
{
"epoch": 0.83,
"learning_rate": 1.4088471944302862e-06,
"loss": 0.8267,
"step": 297
},
{
"epoch": 0.84,
"learning_rate": 1.3626008115036182e-06,
"loss": 0.9331,
"step": 298
},
{
"epoch": 0.84,
"learning_rate": 1.3170706404986645e-06,
"loss": 0.9028,
"step": 299
},
{
"epoch": 0.84,
"learning_rate": 1.272260456770772e-06,
"loss": 0.9136,
"step": 300
},
{
"epoch": 0.85,
"learning_rate": 1.2281739759740575e-06,
"loss": 0.8936,
"step": 301
},
{
"epoch": 0.85,
"learning_rate": 1.1848148537532845e-06,
"loss": 0.8193,
"step": 302
},
{
"epoch": 0.85,
"learning_rate": 1.1421866854407472e-06,
"loss": 0.9473,
"step": 303
},
{
"epoch": 0.85,
"learning_rate": 1.1002930057581451e-06,
"loss": 0.8433,
"step": 304
},
{
"epoch": 0.86,
"learning_rate": 1.0591372885234885e-06,
"loss": 0.8237,
"step": 305
},
{
"epoch": 0.86,
"learning_rate": 1.01872294636304e-06,
"loss": 0.8652,
"step": 306
},
{
"epoch": 0.86,
"learning_rate": 9.79053330428348e-07,
"loss": 0.8267,
"step": 307
},
{
"epoch": 0.87,
"learning_rate": 9.401317301183655e-07,
"loss": 0.8828,
"step": 308
},
{
"epoch": 0.87,
"learning_rate": 9.019613728067001e-07,
"loss": 0.8657,
"step": 309
},
{
"epoch": 0.87,
"learning_rate": 8.645454235739903e-07,
"loss": 0.8096,
"step": 310
},
{
"epoch": 0.87,
"learning_rate": 8.278869849454718e-07,
"loss": 0.8589,
"step": 311
},
{
"epoch": 0.88,
"learning_rate": 7.919890966336929e-07,
"loss": 0.8657,
"step": 312
},
{
"epoch": 0.88,
"learning_rate": 7.568547352864941e-07,
"loss": 0.8804,
"step": 313
},
{
"epoch": 0.88,
"learning_rate": 7.224868142401542e-07,
"loss": 0.8921,
"step": 314
},
{
"epoch": 0.88,
"learning_rate": 6.888881832778415e-07,
"loss": 0.8496,
"step": 315
},
{
"epoch": 0.89,
"learning_rate": 6.560616283932897e-07,
"loss": 0.8892,
"step": 316
},
{
"epoch": 0.89,
"learning_rate": 6.240098715597975e-07,
"loss": 0.8198,
"step": 317
},
{
"epoch": 0.89,
"learning_rate": 5.927355705045179e-07,
"loss": 0.8169,
"step": 318
},
{
"epoch": 0.9,
"learning_rate": 5.622413184880815e-07,
"loss": 0.8013,
"step": 319
},
{
"epoch": 0.9,
"learning_rate": 5.325296440895622e-07,
"loss": 0.8574,
"step": 320
},
{
"epoch": 0.9,
"learning_rate": 5.036030109968082e-07,
"loss": 0.7627,
"step": 321
},
{
"epoch": 0.9,
"learning_rate": 4.754638178021498e-07,
"loss": 0.8037,
"step": 322
},
{
"epoch": 0.91,
"learning_rate": 4.481143978035196e-07,
"loss": 0.8857,
"step": 323
},
{
"epoch": 0.91,
"learning_rate": 4.215570188109608e-07,
"loss": 0.833,
"step": 324
},
{
"epoch": 0.91,
"learning_rate": 3.95793882958595e-07,
"loss": 0.8242,
"step": 325
},
{
"epoch": 0.92,
"learning_rate": 3.708271265220087e-07,
"loss": 0.9121,
"step": 326
},
{
"epoch": 0.92,
"learning_rate": 3.466588197411203e-07,
"loss": 0.9141,
"step": 327
},
{
"epoch": 0.92,
"learning_rate": 3.2329096664852065e-07,
"loss": 0.8745,
"step": 328
},
{
"epoch": 0.92,
"learning_rate": 3.0072550490328754e-07,
"loss": 0.8159,
"step": 329
},
{
"epoch": 0.93,
"learning_rate": 2.7896430563032704e-07,
"loss": 0.8335,
"step": 330
},
{
"epoch": 0.93,
"learning_rate": 2.5800917326521013e-07,
"loss": 0.8457,
"step": 331
},
{
"epoch": 0.93,
"learning_rate": 2.3786184540455449e-07,
"loss": 0.8525,
"step": 332
},
{
"epoch": 0.94,
"learning_rate": 2.1852399266194312e-07,
"loss": 0.9009,
"step": 333
},
{
"epoch": 0.94,
"learning_rate": 1.9999721852939857e-07,
"loss": 0.7441,
"step": 334
},
{
"epoch": 0.94,
"learning_rate": 1.8228305924441469e-07,
"loss": 0.8647,
"step": 335
},
{
"epoch": 0.94,
"learning_rate": 1.6538298366257975e-07,
"loss": 0.813,
"step": 336
},
{
"epoch": 0.95,
"learning_rate": 1.4929839313577609e-07,
"loss": 0.8867,
"step": 337
},
{
"epoch": 0.95,
"learning_rate": 1.3403062139598078e-07,
"loss": 0.8076,
"step": 338
},
{
"epoch": 0.95,
"learning_rate": 1.195809344446708e-07,
"loss": 0.8276,
"step": 339
},
{
"epoch": 0.96,
"learning_rate": 1.0595053044785031e-07,
"loss": 0.8813,
"step": 340
},
{
"epoch": 0.96,
"learning_rate": 9.314053963669245e-08,
"loss": 0.7749,
"step": 341
},
{
"epoch": 0.96,
"learning_rate": 8.115202421383084e-08,
"loss": 0.832,
"step": 342
},
{
"epoch": 0.96,
"learning_rate": 6.998597826526899e-08,
"loss": 0.8799,
"step": 343
},
{
"epoch": 0.97,
"learning_rate": 5.964332767796399e-08,
"loss": 0.8901,
"step": 344
},
{
"epoch": 0.97,
"learning_rate": 5.012493006304131e-08,
"loss": 0.8438,
"step": 345
},
{
"epoch": 0.97,
"learning_rate": 4.143157468468717e-08,
"loss": 0.8525,
"step": 346
},
{
"epoch": 0.97,
"learning_rate": 3.356398239470427e-08,
"loss": 0.8438,
"step": 347
},
{
"epoch": 0.98,
"learning_rate": 2.652280557273512e-08,
"loss": 0.7988,
"step": 348
},
{
"epoch": 0.98,
"learning_rate": 2.0308628072166493e-08,
"loss": 0.9102,
"step": 349
},
{
"epoch": 0.98,
"learning_rate": 1.4921965171720288e-08,
"loss": 0.9048,
"step": 350
},
{
"epoch": 0.99,
"learning_rate": 1.0363263532724433e-08,
"loss": 0.8325,
"step": 351
},
{
"epoch": 0.99,
"learning_rate": 6.6329011620747116e-09,
"loss": 0.8145,
"step": 352
},
{
"epoch": 0.99,
"learning_rate": 3.731187380893176e-09,
"loss": 0.8042,
"step": 353
},
{
"epoch": 0.99,
"learning_rate": 1.6583627988775441e-09,
"loss": 0.8091,
"step": 354
},
{
"epoch": 1.0,
"learning_rate": 4.145992943482835e-10,
"loss": 0.8618,
"step": 355
},
{
"epoch": 1.0,
"learning_rate": 0.0,
"loss": 0.8188,
"step": 356
},
{
"epoch": 1.0,
"step": 356,
"total_flos": 2.229675365414994e+17,
"train_loss": 0.9398124780547753,
"train_runtime": 2363.8849,
"train_samples_per_second": 19.243,
"train_steps_per_second": 0.151
}
],
"max_steps": 356,
"num_train_epochs": 1,
"total_flos": 2.229675365414994e+17,
"trial_name": null,
"trial_params": null
}