spider-llama160m / trainer_state.json
eqhylxx's picture
teacher fwd distill
3dbb1e9
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.9748571428571429,
"global_step": 108,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.02,
"learning_rate": 5e-06,
"loss": 5.2929,
"step": 1
},
{
"epoch": 0.04,
"learning_rate": 1e-05,
"loss": 5.1707,
"step": 2
},
{
"epoch": 0.05,
"learning_rate": 1.5000000000000002e-05,
"loss": 4.86,
"step": 3
},
{
"epoch": 0.07,
"learning_rate": 2e-05,
"loss": 4.909,
"step": 4
},
{
"epoch": 0.09,
"learning_rate": 1.9995437844895337e-05,
"loss": 4.4727,
"step": 5
},
{
"epoch": 0.11,
"learning_rate": 1.9981755542233175e-05,
"loss": 4.1336,
"step": 6
},
{
"epoch": 0.13,
"learning_rate": 1.995896557617091e-05,
"loss": 3.5284,
"step": 7
},
{
"epoch": 0.15,
"learning_rate": 1.992708874098054e-05,
"loss": 3.1435,
"step": 8
},
{
"epoch": 0.16,
"learning_rate": 1.9886154122075344e-05,
"loss": 2.9837,
"step": 9
},
{
"epoch": 0.18,
"learning_rate": 1.983619906947144e-05,
"loss": 2.6885,
"step": 10
},
{
"epoch": 0.2,
"learning_rate": 1.977726916370847e-05,
"loss": 2.6187,
"step": 11
},
{
"epoch": 0.22,
"learning_rate": 1.9709418174260523e-05,
"loss": 2.5325,
"step": 12
},
{
"epoch": 0.24,
"learning_rate": 1.9632708010475166e-05,
"loss": 2.2144,
"step": 13
},
{
"epoch": 0.26,
"learning_rate": 1.954720866508546e-05,
"loss": 2.1453,
"step": 14
},
{
"epoch": 0.27,
"learning_rate": 1.9452998150346403e-05,
"loss": 2.0979,
"step": 15
},
{
"epoch": 0.29,
"learning_rate": 1.9350162426854152e-05,
"loss": 2.1127,
"step": 16
},
{
"epoch": 0.31,
"learning_rate": 1.9238795325112867e-05,
"loss": 1.9363,
"step": 17
},
{
"epoch": 0.33,
"learning_rate": 1.91189984599209e-05,
"loss": 1.9072,
"step": 18
},
{
"epoch": 0.35,
"learning_rate": 1.899088113765426e-05,
"loss": 1.8657,
"step": 19
},
{
"epoch": 0.37,
"learning_rate": 1.8854560256532098e-05,
"loss": 1.9458,
"step": 20
},
{
"epoch": 0.38,
"learning_rate": 1.8710160199955158e-05,
"loss": 1.8255,
"step": 21
},
{
"epoch": 0.4,
"learning_rate": 1.8557812723014476e-05,
"loss": 1.9809,
"step": 22
},
{
"epoch": 0.42,
"learning_rate": 1.8397656832273982e-05,
"loss": 1.4541,
"step": 23
},
{
"epoch": 0.44,
"learning_rate": 1.8229838658936566e-05,
"loss": 1.496,
"step": 24
},
{
"epoch": 0.46,
"learning_rate": 1.805451132550946e-05,
"loss": 1.6486,
"step": 25
},
{
"epoch": 0.48,
"learning_rate": 1.7871834806090502e-05,
"loss": 1.4413,
"step": 26
},
{
"epoch": 0.49,
"learning_rate": 1.7681975780402807e-05,
"loss": 1.4627,
"step": 27
},
{
"epoch": 0.51,
"learning_rate": 1.7485107481711014e-05,
"loss": 1.548,
"step": 28
},
{
"epoch": 0.53,
"learning_rate": 1.7281409538757886e-05,
"loss": 1.3278,
"step": 29
},
{
"epoch": 0.55,
"learning_rate": 1.7071067811865477e-05,
"loss": 1.4126,
"step": 30
},
{
"epoch": 0.57,
"learning_rate": 1.68542742233504e-05,
"loss": 1.6469,
"step": 31
},
{
"epoch": 0.59,
"learning_rate": 1.6631226582407954e-05,
"loss": 1.2131,
"step": 32
},
{
"epoch": 0.6,
"learning_rate": 1.640212840462488e-05,
"loss": 1.4731,
"step": 33
},
{
"epoch": 0.62,
"learning_rate": 1.6167188726285433e-05,
"loss": 1.2919,
"step": 34
},
{
"epoch": 0.64,
"learning_rate": 1.592662191364017e-05,
"loss": 1.3771,
"step": 35
},
{
"epoch": 0.66,
"learning_rate": 1.568064746731156e-05,
"loss": 1.3238,
"step": 36
},
{
"epoch": 0.68,
"learning_rate": 1.542948982201479e-05,
"loss": 1.2339,
"step": 37
},
{
"epoch": 0.69,
"learning_rate": 1.5173378141776569e-05,
"loss": 1.2227,
"step": 38
},
{
"epoch": 0.71,
"learning_rate": 1.4912546110838775e-05,
"loss": 1.1856,
"step": 39
},
{
"epoch": 0.73,
"learning_rate": 1.4647231720437687e-05,
"loss": 1.2679,
"step": 40
},
{
"epoch": 0.75,
"learning_rate": 1.4377677051653404e-05,
"loss": 1.1049,
"step": 41
},
{
"epoch": 0.77,
"learning_rate": 1.410412805452757e-05,
"loss": 1.0227,
"step": 42
},
{
"epoch": 0.79,
"learning_rate": 1.3826834323650899e-05,
"loss": 1.2718,
"step": 43
},
{
"epoch": 0.8,
"learning_rate": 1.3546048870425356e-05,
"loss": 0.9942,
"step": 44
},
{
"epoch": 0.82,
"learning_rate": 1.3262027892208696e-05,
"loss": 1.2092,
"step": 45
},
{
"epoch": 0.84,
"learning_rate": 1.297503053855203e-05,
"loss": 1.0914,
"step": 46
},
{
"epoch": 0.86,
"learning_rate": 1.2685318674743769e-05,
"loss": 1.104,
"step": 47
},
{
"epoch": 0.88,
"learning_rate": 1.2393156642875579e-05,
"loss": 1.011,
"step": 48
},
{
"epoch": 0.9,
"learning_rate": 1.2098811020648475e-05,
"loss": 1.4294,
"step": 49
},
{
"epoch": 0.91,
"learning_rate": 1.180255037813906e-05,
"loss": 1.1338,
"step": 50
},
{
"epoch": 0.93,
"learning_rate": 1.1504645032747832e-05,
"loss": 1.154,
"step": 51
},
{
"epoch": 0.95,
"learning_rate": 1.1205366802553231e-05,
"loss": 1.0417,
"step": 52
},
{
"epoch": 0.97,
"learning_rate": 1.090498875829638e-05,
"loss": 0.89,
"step": 53
},
{
"epoch": 0.99,
"learning_rate": 1.0603784974222862e-05,
"loss": 0.8541,
"step": 54
},
{
"epoch": 1.01,
"learning_rate": 1.030203027800889e-05,
"loss": 1.0875,
"step": 55
},
{
"epoch": 1.02,
"learning_rate": 1e-05,
"loss": 1.1117,
"step": 56
},
{
"epoch": 1.04,
"learning_rate": 9.697969721991114e-06,
"loss": 1.0332,
"step": 57
},
{
"epoch": 1.06,
"learning_rate": 9.39621502577714e-06,
"loss": 0.9024,
"step": 58
},
{
"epoch": 1.08,
"learning_rate": 9.095011241703623e-06,
"loss": 0.8321,
"step": 59
},
{
"epoch": 1.1,
"learning_rate": 8.79463319744677e-06,
"loss": 0.817,
"step": 60
},
{
"epoch": 1.12,
"learning_rate": 8.49535496725217e-06,
"loss": 0.8457,
"step": 61
},
{
"epoch": 1.13,
"learning_rate": 8.197449621860944e-06,
"loss": 0.9017,
"step": 62
},
{
"epoch": 1.15,
"learning_rate": 7.901188979351527e-06,
"loss": 1.0759,
"step": 63
},
{
"epoch": 1.17,
"learning_rate": 7.606843357124426e-06,
"loss": 1.011,
"step": 64
},
{
"epoch": 1.19,
"learning_rate": 7.314681325256232e-06,
"loss": 0.8737,
"step": 65
},
{
"epoch": 1.21,
"learning_rate": 7.024969461447973e-06,
"loss": 0.9498,
"step": 66
},
{
"epoch": 1.23,
"learning_rate": 6.7379721077913095e-06,
"loss": 1.0125,
"step": 67
},
{
"epoch": 1.24,
"learning_rate": 6.453951129574644e-06,
"loss": 0.7371,
"step": 68
},
{
"epoch": 1.26,
"learning_rate": 6.173165676349103e-06,
"loss": 0.8445,
"step": 69
},
{
"epoch": 1.28,
"learning_rate": 5.895871945472434e-06,
"loss": 1.0347,
"step": 70
},
{
"epoch": 1.3,
"learning_rate": 5.622322948346595e-06,
"loss": 1.0074,
"step": 71
},
{
"epoch": 1.32,
"learning_rate": 5.352768279562315e-06,
"loss": 0.887,
"step": 72
},
{
"epoch": 1.33,
"learning_rate": 5.087453889161229e-06,
"loss": 0.9886,
"step": 73
},
{
"epoch": 1.35,
"learning_rate": 4.826621858223431e-06,
"loss": 0.9414,
"step": 74
},
{
"epoch": 1.37,
"learning_rate": 4.570510177985213e-06,
"loss": 1.098,
"step": 75
},
{
"epoch": 1.39,
"learning_rate": 4.319352532688444e-06,
"loss": 1.0022,
"step": 76
},
{
"epoch": 1.41,
"learning_rate": 4.073378086359834e-06,
"loss": 0.9925,
"step": 77
},
{
"epoch": 1.43,
"learning_rate": 3.832811273714569e-06,
"loss": 0.7212,
"step": 78
},
{
"epoch": 1.44,
"learning_rate": 3.5978715953751207e-06,
"loss": 0.7285,
"step": 79
},
{
"epoch": 1.46,
"learning_rate": 3.3687734175920505e-06,
"loss": 0.8824,
"step": 80
},
{
"epoch": 1.48,
"learning_rate": 3.145725776649602e-06,
"loss": 0.8606,
"step": 81
},
{
"epoch": 1.5,
"learning_rate": 2.9289321881345257e-06,
"loss": 0.7898,
"step": 82
},
{
"epoch": 1.52,
"learning_rate": 2.7185904612421177e-06,
"loss": 0.913,
"step": 83
},
{
"epoch": 1.54,
"learning_rate": 2.514892518288988e-06,
"loss": 0.9278,
"step": 84
},
{
"epoch": 1.55,
"learning_rate": 2.318024219597196e-06,
"loss": 0.9006,
"step": 85
},
{
"epoch": 1.57,
"learning_rate": 2.1281651939094996e-06,
"loss": 0.9705,
"step": 86
},
{
"epoch": 1.59,
"learning_rate": 1.94548867449054e-06,
"loss": 0.848,
"step": 87
},
{
"epoch": 1.61,
"learning_rate": 1.7701613410634367e-06,
"loss": 0.947,
"step": 88
},
{
"epoch": 1.63,
"learning_rate": 1.6023431677260215e-06,
"loss": 0.8865,
"step": 89
},
{
"epoch": 1.65,
"learning_rate": 1.4421872769855262e-06,
"loss": 0.9392,
"step": 90
},
{
"epoch": 1.66,
"learning_rate": 1.2898398000448441e-06,
"loss": 0.85,
"step": 91
},
{
"epoch": 1.68,
"learning_rate": 1.1454397434679022e-06,
"loss": 0.7694,
"step": 92
},
{
"epoch": 1.7,
"learning_rate": 1.0091188623457415e-06,
"loss": 0.8638,
"step": 93
},
{
"epoch": 1.72,
"learning_rate": 8.810015400790994e-07,
"loss": 0.8655,
"step": 94
},
{
"epoch": 1.74,
"learning_rate": 7.612046748871327e-07,
"loss": 0.8398,
"step": 95
},
{
"epoch": 1.76,
"learning_rate": 6.498375731458529e-07,
"loss": 0.7379,
"step": 96
},
{
"epoch": 1.77,
"learning_rate": 5.470018496535967e-07,
"loss": 0.8092,
"step": 97
},
{
"epoch": 1.79,
"learning_rate": 4.5279133491454406e-07,
"loss": 0.8303,
"step": 98
},
{
"epoch": 1.81,
"learning_rate": 3.6729198952483725e-07,
"loss": 0.7979,
"step": 99
},
{
"epoch": 1.83,
"learning_rate": 2.905818257394799e-07,
"loss": 0.834,
"step": 100
},
{
"epoch": 1.85,
"learning_rate": 2.2273083629153148e-07,
"loss": 0.8447,
"step": 101
},
{
"epoch": 1.87,
"learning_rate": 1.6380093052856482e-07,
"loss": 0.8207,
"step": 102
},
{
"epoch": 1.88,
"learning_rate": 1.1384587792465873e-07,
"loss": 0.8542,
"step": 103
},
{
"epoch": 1.9,
"learning_rate": 7.291125901946027e-08,
"loss": 1.091,
"step": 104
},
{
"epoch": 1.92,
"learning_rate": 4.103442382909051e-08,
"loss": 0.9051,
"step": 105
},
{
"epoch": 1.94,
"learning_rate": 1.824445776682504e-08,
"loss": 0.9512,
"step": 106
},
{
"epoch": 1.96,
"learning_rate": 4.562155104665955e-09,
"loss": 0.6725,
"step": 107
},
{
"epoch": 1.97,
"learning_rate": 0.0,
"loss": 0.7599,
"step": 108
},
{
"epoch": 1.97,
"step": 108,
"total_flos": 2.3415025142071296e+16,
"train_loss": 1.4361559649308522,
"train_runtime": 6959.0247,
"train_samples_per_second": 2.012,
"train_steps_per_second": 0.016
}
],
"max_steps": 108,
"num_train_epochs": 2,
"total_flos": 2.3415025142071296e+16,
"trial_name": null,
"trial_params": null
}