repo_name / trainer_state.json
doulujiyao's picture
7b model
55f1d12
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 5.0,
"eval_steps": 500,
"global_step": 130,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.04,
"learning_rate": 1e-05,
"loss": 3.9642,
"step": 1
},
{
"epoch": 0.08,
"learning_rate": 2e-05,
"loss": 4.071,
"step": 2
},
{
"epoch": 0.12,
"learning_rate": 3.0000000000000004e-05,
"loss": 3.8538,
"step": 3
},
{
"epoch": 0.15,
"learning_rate": 4e-05,
"loss": 3.9502,
"step": 4
},
{
"epoch": 0.19,
"learning_rate": 3.999378364001633e-05,
"loss": 3.7881,
"step": 5
},
{
"epoch": 0.23,
"learning_rate": 3.997513842437845e-05,
"loss": 3.4548,
"step": 6
},
{
"epoch": 0.27,
"learning_rate": 3.9944075943623605e-05,
"loss": 3.3921,
"step": 7
},
{
"epoch": 0.31,
"learning_rate": 3.990061550730803e-05,
"loss": 3.0533,
"step": 8
},
{
"epoch": 0.35,
"learning_rate": 3.984478413200345e-05,
"loss": 2.7676,
"step": 9
},
{
"epoch": 0.38,
"learning_rate": 3.977661652450257e-05,
"loss": 2.6457,
"step": 10
},
{
"epoch": 0.42,
"learning_rate": 3.9696155060244166e-05,
"loss": 2.3735,
"step": 11
},
{
"epoch": 0.46,
"learning_rate": 3.9603449756970877e-05,
"loss": 2.4271,
"step": 12
},
{
"epoch": 0.5,
"learning_rate": 3.949855824363647e-05,
"loss": 2.2207,
"step": 13
},
{
"epoch": 0.54,
"learning_rate": 3.938154572458156e-05,
"loss": 2.1449,
"step": 14
},
{
"epoch": 0.58,
"learning_rate": 3.925248493900024e-05,
"loss": 1.9469,
"step": 15
},
{
"epoch": 0.62,
"learning_rate": 3.911145611572282e-05,
"loss": 1.8312,
"step": 16
},
{
"epoch": 0.65,
"learning_rate": 3.895854692334264e-05,
"loss": 1.7768,
"step": 17
},
{
"epoch": 0.69,
"learning_rate": 3.879385241571817e-05,
"loss": 1.6802,
"step": 18
},
{
"epoch": 0.73,
"learning_rate": 3.861747497288409e-05,
"loss": 1.6101,
"step": 19
},
{
"epoch": 0.77,
"learning_rate": 3.842952423740816e-05,
"loss": 1.5403,
"step": 20
},
{
"epoch": 0.81,
"learning_rate": 3.823011704623347e-05,
"loss": 1.4114,
"step": 21
},
{
"epoch": 0.85,
"learning_rate": 3.801937735804838e-05,
"loss": 1.4065,
"step": 22
},
{
"epoch": 0.88,
"learning_rate": 3.7797436176229376e-05,
"loss": 1.3727,
"step": 23
},
{
"epoch": 0.92,
"learning_rate": 3.756443146740457e-05,
"loss": 1.3538,
"step": 24
},
{
"epoch": 0.96,
"learning_rate": 3.732050807568878e-05,
"loss": 1.4656,
"step": 25
},
{
"epoch": 1.0,
"learning_rate": 3.7065817632643115e-05,
"loss": 1.3587,
"step": 26
},
{
"epoch": 1.0,
"eval_loss": 1.2306509017944336,
"eval_runtime": 21.666,
"eval_samples_per_second": 2.123,
"eval_steps_per_second": 0.138,
"step": 26
},
{
"epoch": 1.04,
"learning_rate": 3.680051846301543e-05,
"loss": 1.2631,
"step": 27
},
{
"epoch": 1.08,
"learning_rate": 3.65247754863199e-05,
"loss": 1.2049,
"step": 28
},
{
"epoch": 1.12,
"learning_rate": 3.623876011431714e-05,
"loss": 1.2327,
"step": 29
},
{
"epoch": 1.15,
"learning_rate": 3.5942650144458454e-05,
"loss": 1.0479,
"step": 30
},
{
"epoch": 1.19,
"learning_rate": 3.56366296493606e-05,
"loss": 1.0442,
"step": 31
},
{
"epoch": 1.23,
"learning_rate": 3.532088886237956e-05,
"loss": 1.0004,
"step": 32
},
{
"epoch": 1.27,
"learning_rate": 3.499562405935469e-05,
"loss": 1.0744,
"step": 33
},
{
"epoch": 1.31,
"learning_rate": 3.4661037436596526e-05,
"loss": 0.8552,
"step": 34
},
{
"epoch": 1.35,
"learning_rate": 3.431733698519437e-05,
"loss": 1.0061,
"step": 35
},
{
"epoch": 1.38,
"learning_rate": 3.396473636172146e-05,
"loss": 0.8791,
"step": 36
},
{
"epoch": 1.42,
"learning_rate": 3.360345475541839e-05,
"loss": 0.8711,
"step": 37
},
{
"epoch": 1.46,
"learning_rate": 3.323371675193719e-05,
"loss": 1.1293,
"step": 38
},
{
"epoch": 1.5,
"learning_rate": 3.285575219373079e-05,
"loss": 0.7738,
"step": 39
},
{
"epoch": 1.54,
"learning_rate": 3.246979603717467e-05,
"loss": 0.7033,
"step": 40
},
{
"epoch": 1.58,
"learning_rate": 3.207608820650955e-05,
"loss": 0.8763,
"step": 41
},
{
"epoch": 1.62,
"learning_rate": 3.1674873444695804e-05,
"loss": 0.7544,
"step": 42
},
{
"epoch": 1.65,
"learning_rate": 3.126640116127244e-05,
"loss": 0.7928,
"step": 43
},
{
"epoch": 1.69,
"learning_rate": 3.0850925277315193e-05,
"loss": 0.8081,
"step": 44
},
{
"epoch": 1.73,
"learning_rate": 3.0428704067589963e-05,
"loss": 0.7112,
"step": 45
},
{
"epoch": 1.77,
"learning_rate": 3.0000000000000004e-05,
"loss": 0.7927,
"step": 46
},
{
"epoch": 1.81,
"learning_rate": 2.956507957242637e-05,
"loss": 0.6722,
"step": 47
},
{
"epoch": 1.85,
"learning_rate": 2.9124213147063263e-05,
"loss": 0.7657,
"step": 48
},
{
"epoch": 1.88,
"learning_rate": 2.8677674782351164e-05,
"loss": 0.7748,
"step": 49
},
{
"epoch": 1.92,
"learning_rate": 2.8225742062612236e-05,
"loss": 0.7922,
"step": 50
},
{
"epoch": 1.96,
"learning_rate": 2.7768695925493897e-05,
"loss": 0.8061,
"step": 51
},
{
"epoch": 2.0,
"learning_rate": 2.7306820487327906e-05,
"loss": 0.7763,
"step": 52
},
{
"epoch": 2.0,
"eval_loss": 0.6676424741744995,
"eval_runtime": 21.6655,
"eval_samples_per_second": 2.123,
"eval_steps_per_second": 0.138,
"step": 52
},
{
"epoch": 2.04,
"learning_rate": 2.684040286651338e-05,
"loss": 0.7045,
"step": 53
},
{
"epoch": 2.08,
"learning_rate": 2.6369733005033693e-05,
"loss": 0.7259,
"step": 54
},
{
"epoch": 2.12,
"learning_rate": 2.589510348821809e-05,
"loss": 0.684,
"step": 55
},
{
"epoch": 2.15,
"learning_rate": 2.5416809362860107e-05,
"loss": 0.6923,
"step": 56
},
{
"epoch": 2.19,
"learning_rate": 2.493514795380587e-05,
"loss": 0.6401,
"step": 57
},
{
"epoch": 2.23,
"learning_rate": 2.445041867912629e-05,
"loss": 0.7307,
"step": 58
},
{
"epoch": 2.27,
"learning_rate": 2.3962922863987956e-05,
"loss": 0.6738,
"step": 59
},
{
"epoch": 2.31,
"learning_rate": 2.3472963553338614e-05,
"loss": 0.6694,
"step": 60
},
{
"epoch": 2.35,
"learning_rate": 2.2980845323523487e-05,
"loss": 0.6664,
"step": 61
},
{
"epoch": 2.38,
"learning_rate": 2.2486874092949708e-05,
"loss": 0.6477,
"step": 62
},
{
"epoch": 2.42,
"learning_rate": 2.1991356931916335e-05,
"loss": 0.6324,
"step": 63
},
{
"epoch": 2.46,
"learning_rate": 2.149460187172849e-05,
"loss": 0.6167,
"step": 64
},
{
"epoch": 2.5,
"learning_rate": 2.0996917713213944e-05,
"loss": 0.6035,
"step": 65
},
{
"epoch": 2.54,
"learning_rate": 2.0498613834761462e-05,
"loss": 0.653,
"step": 66
},
{
"epoch": 2.58,
"learning_rate": 2e-05,
"loss": 0.6213,
"step": 67
},
{
"epoch": 2.62,
"learning_rate": 1.9501386165238548e-05,
"loss": 0.5961,
"step": 68
},
{
"epoch": 2.65,
"learning_rate": 1.9003082286786056e-05,
"loss": 0.6726,
"step": 69
},
{
"epoch": 2.69,
"learning_rate": 1.8505398128271517e-05,
"loss": 0.6147,
"step": 70
},
{
"epoch": 2.73,
"learning_rate": 1.800864306808367e-05,
"loss": 0.5905,
"step": 71
},
{
"epoch": 2.77,
"learning_rate": 1.7513125907050302e-05,
"loss": 0.671,
"step": 72
},
{
"epoch": 2.81,
"learning_rate": 1.701915467647651e-05,
"loss": 0.6069,
"step": 73
},
{
"epoch": 2.85,
"learning_rate": 1.6527036446661396e-05,
"loss": 0.6503,
"step": 74
},
{
"epoch": 2.88,
"learning_rate": 1.6037077136012054e-05,
"loss": 0.6333,
"step": 75
},
{
"epoch": 2.92,
"learning_rate": 1.5549581320873715e-05,
"loss": 0.6213,
"step": 76
},
{
"epoch": 2.96,
"learning_rate": 1.5064852046194127e-05,
"loss": 0.6124,
"step": 77
},
{
"epoch": 3.0,
"learning_rate": 1.4583190637139901e-05,
"loss": 0.6404,
"step": 78
},
{
"epoch": 3.0,
"eval_loss": 0.5793539881706238,
"eval_runtime": 21.5979,
"eval_samples_per_second": 2.13,
"eval_steps_per_second": 0.139,
"step": 78
},
{
"epoch": 3.04,
"learning_rate": 1.4104896511781916e-05,
"loss": 0.6378,
"step": 79
},
{
"epoch": 3.08,
"learning_rate": 1.3630266994966314e-05,
"loss": 0.629,
"step": 80
},
{
"epoch": 3.12,
"learning_rate": 1.3159597133486628e-05,
"loss": 0.5703,
"step": 81
},
{
"epoch": 3.15,
"learning_rate": 1.26931795126721e-05,
"loss": 0.5396,
"step": 82
},
{
"epoch": 3.19,
"learning_rate": 1.2231304074506108e-05,
"loss": 0.5786,
"step": 83
},
{
"epoch": 3.23,
"learning_rate": 1.1774257937387774e-05,
"loss": 0.638,
"step": 84
},
{
"epoch": 3.27,
"learning_rate": 1.132232521764884e-05,
"loss": 0.6425,
"step": 85
},
{
"epoch": 3.31,
"learning_rate": 1.087578685293674e-05,
"loss": 0.5446,
"step": 86
},
{
"epoch": 3.35,
"learning_rate": 1.0434920427573643e-05,
"loss": 0.5612,
"step": 87
},
{
"epoch": 3.38,
"learning_rate": 1.0000000000000006e-05,
"loss": 0.5518,
"step": 88
},
{
"epoch": 3.42,
"learning_rate": 9.57129593241004e-06,
"loss": 0.5649,
"step": 89
},
{
"epoch": 3.46,
"learning_rate": 9.149074722684815e-06,
"loss": 0.6007,
"step": 90
},
{
"epoch": 3.5,
"learning_rate": 8.733598838727559e-06,
"loss": 0.4963,
"step": 91
},
{
"epoch": 3.54,
"learning_rate": 8.325126555304208e-06,
"loss": 0.5296,
"step": 92
},
{
"epoch": 3.58,
"learning_rate": 7.923911793490449e-06,
"loss": 0.5389,
"step": 93
},
{
"epoch": 3.62,
"learning_rate": 7.530203962825331e-06,
"loss": 0.4734,
"step": 94
},
{
"epoch": 3.65,
"learning_rate": 7.1442478062692135e-06,
"loss": 0.5749,
"step": 95
},
{
"epoch": 3.69,
"learning_rate": 6.766283248062817e-06,
"loss": 0.558,
"step": 96
},
{
"epoch": 3.73,
"learning_rate": 6.396545244581609e-06,
"loss": 0.5729,
"step": 97
},
{
"epoch": 3.77,
"learning_rate": 6.035263638278546e-06,
"loss": 0.5351,
"step": 98
},
{
"epoch": 3.81,
"learning_rate": 5.682663014805631e-06,
"loss": 0.54,
"step": 99
},
{
"epoch": 3.85,
"learning_rate": 5.338962563403478e-06,
"loss": 0.454,
"step": 100
},
{
"epoch": 3.88,
"learning_rate": 5.004375940645314e-06,
"loss": 0.5863,
"step": 101
},
{
"epoch": 3.92,
"learning_rate": 4.679111137620442e-06,
"loss": 0.5254,
"step": 102
},
{
"epoch": 3.96,
"learning_rate": 4.363370350639405e-06,
"loss": 0.6083,
"step": 103
},
{
"epoch": 4.0,
"learning_rate": 4.057349855541557e-06,
"loss": 0.4868,
"step": 104
},
{
"epoch": 4.0,
"eval_loss": 0.5455546379089355,
"eval_runtime": 21.1145,
"eval_samples_per_second": 2.179,
"eval_steps_per_second": 0.142,
"step": 104
},
{
"epoch": 4.04,
"learning_rate": 3.76123988568287e-06,
"loss": 0.449,
"step": 105
},
{
"epoch": 4.08,
"learning_rate": 3.4752245136801065e-06,
"loss": 0.5543,
"step": 106
},
{
"epoch": 4.12,
"learning_rate": 3.199481536984572e-06,
"loss": 0.492,
"step": 107
},
{
"epoch": 4.15,
"learning_rate": 2.934182367356888e-06,
"loss": 0.561,
"step": 108
},
{
"epoch": 4.19,
"learning_rate": 2.679491924311226e-06,
"loss": 0.5317,
"step": 109
},
{
"epoch": 4.23,
"learning_rate": 2.435568532595427e-06,
"loss": 0.5508,
"step": 110
},
{
"epoch": 4.27,
"learning_rate": 2.2025638237706294e-06,
"loss": 0.4914,
"step": 111
},
{
"epoch": 4.31,
"learning_rate": 1.9806226419516195e-06,
"loss": 0.5875,
"step": 112
},
{
"epoch": 4.35,
"learning_rate": 1.7698829537665374e-06,
"loss": 0.4155,
"step": 113
},
{
"epoch": 4.38,
"learning_rate": 1.5704757625918454e-06,
"loss": 0.5517,
"step": 114
},
{
"epoch": 4.42,
"learning_rate": 1.3825250271159175e-06,
"loss": 0.5025,
"step": 115
},
{
"epoch": 4.46,
"learning_rate": 1.2061475842818337e-06,
"loss": 0.4756,
"step": 116
},
{
"epoch": 4.5,
"learning_rate": 1.0414530766573661e-06,
"loss": 0.5201,
"step": 117
},
{
"epoch": 4.54,
"learning_rate": 8.885438842771843e-07,
"loss": 0.5558,
"step": 118
},
{
"epoch": 4.58,
"learning_rate": 7.475150609997595e-07,
"loss": 0.4251,
"step": 119
},
{
"epoch": 4.62,
"learning_rate": 6.184542754184431e-07,
"loss": 0.6341,
"step": 120
},
{
"epoch": 4.65,
"learning_rate": 5.014417563635276e-07,
"loss": 0.4536,
"step": 121
},
{
"epoch": 4.69,
"learning_rate": 3.965502430291235e-07,
"loss": 0.5496,
"step": 122
},
{
"epoch": 4.73,
"learning_rate": 3.038449397558396e-07,
"loss": 0.5243,
"step": 123
},
{
"epoch": 4.77,
"learning_rate": 2.2338347549742956e-07,
"loss": 0.5546,
"step": 124
},
{
"epoch": 4.81,
"learning_rate": 1.5521586799655875e-07,
"loss": 0.4766,
"step": 125
},
{
"epoch": 4.85,
"learning_rate": 9.938449269197181e-08,
"loss": 0.5953,
"step": 126
},
{
"epoch": 4.88,
"learning_rate": 5.592405637639742e-08,
"loss": 0.5025,
"step": 127
},
{
"epoch": 4.92,
"learning_rate": 2.4861575621553112e-08,
"loss": 0.4808,
"step": 128
},
{
"epoch": 4.96,
"learning_rate": 6.216359983675091e-09,
"loss": 0.463,
"step": 129
},
{
"epoch": 5.0,
"learning_rate": 0.0,
"loss": 0.5275,
"step": 130
},
{
"epoch": 5.0,
"eval_loss": 0.5420305728912354,
"eval_runtime": 21.3644,
"eval_samples_per_second": 2.153,
"eval_steps_per_second": 0.14,
"step": 130
},
{
"epoch": 5.0,
"step": 130,
"total_flos": 1.4892268948732825e+20,
"train_loss": 1.008507801936223,
"train_runtime": 556.8684,
"train_samples_per_second": 3.735,
"train_steps_per_second": 0.233
}
],
"logging_steps": 1.0,
"max_steps": 130,
"num_input_tokens_seen": 0,
"num_train_epochs": 5,
"save_steps": 200,
"total_flos": 1.4892268948732825e+20,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}