Yi-6B-ruozhiba-1e-5-50 / trainer_state.json
yyx123's picture
Model save
2edae87 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 10.0,
"eval_steps": 500,
"global_step": 550,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.02,
"learning_rate": 1.8181818181818183e-07,
"loss": 2.7431,
"step": 1
},
{
"epoch": 0.07,
"learning_rate": 7.272727272727273e-07,
"loss": 2.89,
"step": 4
},
{
"epoch": 0.15,
"learning_rate": 1.4545454545454546e-06,
"loss": 2.7283,
"step": 8
},
{
"epoch": 0.22,
"learning_rate": 2.181818181818182e-06,
"loss": 2.6956,
"step": 12
},
{
"epoch": 0.29,
"learning_rate": 2.9090909090909093e-06,
"loss": 2.6684,
"step": 16
},
{
"epoch": 0.36,
"learning_rate": 3.6363636363636366e-06,
"loss": 2.7545,
"step": 20
},
{
"epoch": 0.44,
"learning_rate": 4.363636363636364e-06,
"loss": 2.8797,
"step": 24
},
{
"epoch": 0.51,
"learning_rate": 5.090909090909091e-06,
"loss": 3.0007,
"step": 28
},
{
"epoch": 0.58,
"learning_rate": 5.8181818181818185e-06,
"loss": 2.9172,
"step": 32
},
{
"epoch": 0.65,
"learning_rate": 6.545454545454546e-06,
"loss": 2.8011,
"step": 36
},
{
"epoch": 0.73,
"learning_rate": 7.272727272727273e-06,
"loss": 2.6401,
"step": 40
},
{
"epoch": 0.8,
"learning_rate": 8.000000000000001e-06,
"loss": 2.7162,
"step": 44
},
{
"epoch": 0.87,
"learning_rate": 8.727272727272728e-06,
"loss": 2.6036,
"step": 48
},
{
"epoch": 0.95,
"learning_rate": 9.454545454545456e-06,
"loss": 2.6749,
"step": 52
},
{
"epoch": 1.0,
"pls_score": 26.262626262626263,
"std": 2.1312214954012014,
"step": 55
},
{
"epoch": 1.0,
"eval_loss": 2.5386734008789062,
"eval_runtime": 4.9212,
"eval_samples_per_second": 4.674,
"eval_steps_per_second": 1.219,
"step": 55
},
{
"epoch": 1.02,
"learning_rate": 9.999899300364534e-06,
"loss": 2.6648,
"step": 56
},
{
"epoch": 1.09,
"learning_rate": 9.997482711915926e-06,
"loss": 2.5038,
"step": 60
},
{
"epoch": 1.16,
"learning_rate": 9.991845519630679e-06,
"loss": 2.424,
"step": 64
},
{
"epoch": 1.24,
"learning_rate": 9.982991356370404e-06,
"loss": 2.492,
"step": 68
},
{
"epoch": 1.31,
"learning_rate": 9.970925928158275e-06,
"loss": 2.5985,
"step": 72
},
{
"epoch": 1.38,
"learning_rate": 9.955657010501807e-06,
"loss": 2.2698,
"step": 76
},
{
"epoch": 1.45,
"learning_rate": 9.937194443381972e-06,
"loss": 2.3577,
"step": 80
},
{
"epoch": 1.53,
"learning_rate": 9.915550124911866e-06,
"loss": 2.4562,
"step": 84
},
{
"epoch": 1.6,
"learning_rate": 9.890738003669029e-06,
"loss": 2.266,
"step": 88
},
{
"epoch": 1.67,
"learning_rate": 9.862774069706346e-06,
"loss": 2.2448,
"step": 92
},
{
"epoch": 1.75,
"learning_rate": 9.831676344247343e-06,
"loss": 2.2374,
"step": 96
},
{
"epoch": 1.82,
"learning_rate": 9.797464868072489e-06,
"loss": 2.2036,
"step": 100
},
{
"epoch": 1.89,
"learning_rate": 9.760161688604008e-06,
"loss": 2.1167,
"step": 104
},
{
"epoch": 1.96,
"learning_rate": 9.719790845697534e-06,
"loss": 2.133,
"step": 108
},
{
"epoch": 2.04,
"learning_rate": 9.676378356149733e-06,
"loss": 2.2253,
"step": 112
},
{
"epoch": 2.11,
"learning_rate": 9.629952196931902e-06,
"loss": 2.1681,
"step": 116
},
{
"epoch": 2.18,
"learning_rate": 9.580542287160348e-06,
"loss": 2.3558,
"step": 120
},
{
"epoch": 2.25,
"learning_rate": 9.528180468815155e-06,
"loss": 2.1262,
"step": 124
},
{
"epoch": 2.33,
"learning_rate": 9.47290048621977e-06,
"loss": 2.0675,
"step": 128
},
{
"epoch": 2.4,
"learning_rate": 9.414737964294636e-06,
"loss": 2.0984,
"step": 132
},
{
"epoch": 2.47,
"learning_rate": 9.353730385598887e-06,
"loss": 2.0763,
"step": 136
},
{
"epoch": 2.55,
"learning_rate": 9.289917066174887e-06,
"loss": 2.0899,
"step": 140
},
{
"epoch": 2.62,
"learning_rate": 9.223339130211194e-06,
"loss": 1.9647,
"step": 144
},
{
"epoch": 2.69,
"learning_rate": 9.154039483540273e-06,
"loss": 2.0805,
"step": 148
},
{
"epoch": 2.76,
"learning_rate": 9.08206278598805e-06,
"loss": 1.996,
"step": 152
},
{
"epoch": 2.84,
"learning_rate": 9.007455422593077e-06,
"loss": 2.1735,
"step": 156
},
{
"epoch": 2.91,
"learning_rate": 8.930265473713939e-06,
"loss": 2.197,
"step": 160
},
{
"epoch": 2.98,
"learning_rate": 8.850542684044078e-06,
"loss": 2.0502,
"step": 164
},
{
"epoch": 3.0,
"pls_score": 60.4,
"std": 4.326291714621197,
"step": 165
},
{
"epoch": 3.0,
"eval_loss": 2.0924267768859863,
"eval_runtime": 4.9011,
"eval_samples_per_second": 4.693,
"eval_steps_per_second": 1.224,
"step": 165
},
{
"epoch": 3.05,
"learning_rate": 8.768338430554083e-06,
"loss": 1.8989,
"step": 168
},
{
"epoch": 3.13,
"learning_rate": 8.683705689382025e-06,
"loss": 2.149,
"step": 172
},
{
"epoch": 3.2,
"learning_rate": 8.596699001693257e-06,
"loss": 2.0518,
"step": 176
},
{
"epoch": 3.27,
"learning_rate": 8.507374438531606e-06,
"loss": 2.1007,
"step": 180
},
{
"epoch": 3.35,
"learning_rate": 8.415789564684673e-06,
"loss": 2.0187,
"step": 184
},
{
"epoch": 3.42,
"learning_rate": 8.322003401586463e-06,
"loss": 1.9728,
"step": 188
},
{
"epoch": 3.49,
"learning_rate": 8.226076389281316e-06,
"loss": 1.9981,
"step": 192
},
{
"epoch": 3.56,
"learning_rate": 8.128070347473609e-06,
"loss": 2.0167,
"step": 196
},
{
"epoch": 3.64,
"learning_rate": 8.028048435688333e-06,
"loss": 2.023,
"step": 200
},
{
"epoch": 3.71,
"learning_rate": 7.92607511256826e-06,
"loss": 2.0933,
"step": 204
},
{
"epoch": 3.78,
"learning_rate": 7.822216094333847e-06,
"loss": 2.0343,
"step": 208
},
{
"epoch": 3.85,
"learning_rate": 7.716538312432767e-06,
"loss": 2.0147,
"step": 212
},
{
"epoch": 3.93,
"learning_rate": 7.60910987040623e-06,
"loss": 2.0385,
"step": 216
},
{
"epoch": 4.0,
"learning_rate": 7.500000000000001e-06,
"loss": 1.953,
"step": 220
},
{
"epoch": 4.0,
"pls_score": 63.673469387755105,
"std": 4.56973253478639,
"step": 220
},
{
"epoch": 4.0,
"eval_loss": 2.038102626800537,
"eval_runtime": 4.9029,
"eval_samples_per_second": 4.691,
"eval_steps_per_second": 1.224,
"step": 220
},
{
"epoch": 4.07,
"learning_rate": 7.3892790165483164e-06,
"loss": 1.9949,
"step": 224
},
{
"epoch": 4.15,
"learning_rate": 7.2770182736595164e-06,
"loss": 2.0196,
"step": 228
},
{
"epoch": 4.22,
"learning_rate": 7.163290117232542e-06,
"loss": 1.8627,
"step": 232
},
{
"epoch": 4.29,
"learning_rate": 7.048167838833977e-06,
"loss": 2.051,
"step": 236
},
{
"epoch": 4.36,
"learning_rate": 6.931725628465643e-06,
"loss": 2.0636,
"step": 240
},
{
"epoch": 4.44,
"learning_rate": 6.814038526753205e-06,
"loss": 1.8729,
"step": 244
},
{
"epoch": 4.51,
"learning_rate": 6.695182376586603e-06,
"loss": 1.9095,
"step": 248
},
{
"epoch": 4.58,
"learning_rate": 6.5752337742434644e-06,
"loss": 1.9759,
"step": 252
},
{
"epoch": 4.65,
"learning_rate": 6.454270020026996e-06,
"loss": 1.9738,
"step": 256
},
{
"epoch": 4.73,
"learning_rate": 6.332369068450175e-06,
"loss": 1.9394,
"step": 260
},
{
"epoch": 4.8,
"learning_rate": 6.209609477998339e-06,
"loss": 1.9755,
"step": 264
},
{
"epoch": 4.87,
"learning_rate": 6.08607036050254e-06,
"loss": 1.8648,
"step": 268
},
{
"epoch": 4.95,
"learning_rate": 5.961831330156306e-06,
"loss": 1.9412,
"step": 272
},
{
"epoch": 5.0,
"pls_score": 57.6,
"std": 4.257323102608023,
"step": 275
},
{
"epoch": 5.0,
"eval_loss": 2.0098042488098145,
"eval_runtime": 4.9025,
"eval_samples_per_second": 4.692,
"eval_steps_per_second": 1.224,
"step": 275
},
{
"epoch": 5.02,
"learning_rate": 5.8369724522086545e-06,
"loss": 1.9483,
"step": 276
},
{
"epoch": 5.09,
"learning_rate": 5.711574191366427e-06,
"loss": 2.0268,
"step": 280
},
{
"epoch": 5.16,
"learning_rate": 5.585717359939192e-06,
"loss": 1.9855,
"step": 284
},
{
"epoch": 5.24,
"learning_rate": 5.459483065760138e-06,
"loss": 1.8754,
"step": 288
},
{
"epoch": 5.31,
"learning_rate": 5.33295265991652e-06,
"loss": 1.892,
"step": 292
},
{
"epoch": 5.38,
"learning_rate": 5.206207684323337e-06,
"loss": 2.016,
"step": 296
},
{
"epoch": 5.45,
"learning_rate": 5.07932981917404e-06,
"loss": 1.9611,
"step": 300
},
{
"epoch": 5.53,
"learning_rate": 4.952400830302117e-06,
"loss": 1.8741,
"step": 304
},
{
"epoch": 5.6,
"learning_rate": 4.825502516487497e-06,
"loss": 1.9852,
"step": 308
},
{
"epoch": 5.67,
"learning_rate": 4.6987166567417085e-06,
"loss": 1.8991,
"step": 312
},
{
"epoch": 5.75,
"learning_rate": 4.572124957605803e-06,
"loss": 1.8599,
"step": 316
},
{
"epoch": 5.82,
"learning_rate": 4.445809000494945e-06,
"loss": 1.9791,
"step": 320
},
{
"epoch": 5.89,
"learning_rate": 4.319850189123681e-06,
"loss": 1.8507,
"step": 324
},
{
"epoch": 5.96,
"learning_rate": 4.194329697045681e-06,
"loss": 1.9044,
"step": 328
},
{
"epoch": 6.0,
"pls_score": 61.2,
"std": 4.323332048316438,
"step": 330
},
{
"epoch": 6.0,
"eval_loss": 1.9948909282684326,
"eval_runtime": 4.9085,
"eval_samples_per_second": 4.686,
"eval_steps_per_second": 1.222,
"step": 330
},
{
"epoch": 6.04,
"learning_rate": 4.06932841534185e-06,
"loss": 1.9863,
"step": 332
},
{
"epoch": 6.11,
"learning_rate": 3.944926900490452e-06,
"loss": 1.9043,
"step": 336
},
{
"epoch": 6.18,
"learning_rate": 3.821205322452863e-06,
"loss": 1.8504,
"step": 340
},
{
"epoch": 6.25,
"learning_rate": 3.69824341300844e-06,
"loss": 1.912,
"step": 344
},
{
"epoch": 6.33,
"learning_rate": 3.5761204143717387e-06,
"loss": 1.9407,
"step": 348
},
{
"epoch": 6.4,
"learning_rate": 3.4549150281252635e-06,
"loss": 1.8498,
"step": 352
},
{
"epoch": 6.47,
"learning_rate": 3.3347053645005965e-06,
"loss": 1.9506,
"step": 356
},
{
"epoch": 6.55,
"learning_rate": 3.2155688920406415e-06,
"loss": 1.8538,
"step": 360
},
{
"epoch": 6.62,
"learning_rate": 3.097582387675385e-06,
"loss": 1.8497,
"step": 364
},
{
"epoch": 6.69,
"learning_rate": 2.980821887243377e-06,
"loss": 1.8615,
"step": 368
},
{
"epoch": 6.76,
"learning_rate": 2.8653626364907918e-06,
"loss": 1.9318,
"step": 372
},
{
"epoch": 6.84,
"learning_rate": 2.751279042579672e-06,
"loss": 1.8605,
"step": 376
},
{
"epoch": 6.91,
"learning_rate": 2.6386446261365874e-06,
"loss": 1.8682,
"step": 380
},
{
"epoch": 6.98,
"learning_rate": 2.527531973872617e-06,
"loss": 1.9567,
"step": 384
},
{
"epoch": 7.0,
"pls_score": 68.4,
"std": 4.003598381456362,
"step": 385
},
{
"epoch": 7.0,
"eval_loss": 1.9888746738433838,
"eval_runtime": 4.9083,
"eval_samples_per_second": 4.686,
"eval_steps_per_second": 1.222,
"step": 385
},
{
"epoch": 7.05,
"learning_rate": 2.418012691805191e-06,
"loss": 1.6192,
"step": 388
},
{
"epoch": 7.13,
"learning_rate": 2.310157359111938e-06,
"loss": 1.9044,
"step": 392
},
{
"epoch": 7.2,
"learning_rate": 2.204035482646267e-06,
"loss": 1.8584,
"step": 396
},
{
"epoch": 7.27,
"learning_rate": 2.09971545214401e-06,
"loss": 1.7605,
"step": 400
},
{
"epoch": 7.35,
"learning_rate": 1.9972644961499853e-06,
"loss": 1.9119,
"step": 404
},
{
"epoch": 7.42,
"learning_rate": 1.8967486386928819e-06,
"loss": 1.9704,
"step": 408
},
{
"epoch": 7.49,
"learning_rate": 1.798232656736389e-06,
"loss": 1.9329,
"step": 412
},
{
"epoch": 7.56,
"learning_rate": 1.7017800384339928e-06,
"loss": 1.8767,
"step": 416
},
{
"epoch": 7.64,
"learning_rate": 1.6074529422143398e-06,
"loss": 1.8444,
"step": 420
},
{
"epoch": 7.71,
"learning_rate": 1.5153121567235334e-06,
"loss": 1.8153,
"step": 424
},
{
"epoch": 7.78,
"learning_rate": 1.4254170616501828e-06,
"loss": 1.9326,
"step": 428
},
{
"epoch": 7.85,
"learning_rate": 1.3378255894584463e-06,
"loss": 1.9621,
"step": 432
},
{
"epoch": 7.93,
"learning_rate": 1.2525941880537307e-06,
"loss": 1.8408,
"step": 436
},
{
"epoch": 8.0,
"learning_rate": 1.1697777844051105e-06,
"loss": 1.9235,
"step": 440
},
{
"epoch": 8.0,
"pls_score": 62.4,
"std": 4.104241708281811,
"step": 440
},
{
"epoch": 8.0,
"eval_loss": 1.9856469631195068,
"eval_runtime": 4.9071,
"eval_samples_per_second": 4.687,
"eval_steps_per_second": 1.223,
"step": 440
},
{
"epoch": 8.07,
"learning_rate": 1.0894297491479044e-06,
"loss": 1.9699,
"step": 444
},
{
"epoch": 8.15,
"learning_rate": 1.0116018621892237e-06,
"loss": 1.9288,
"step": 448
},
{
"epoch": 8.22,
"learning_rate": 9.363442793386606e-07,
"loss": 1.7962,
"step": 452
},
{
"epoch": 8.29,
"learning_rate": 8.637054999856148e-07,
"loss": 1.8953,
"step": 456
},
{
"epoch": 8.36,
"learning_rate": 7.937323358440935e-07,
"loss": 1.8903,
"step": 460
},
{
"epoch": 8.44,
"learning_rate": 7.264698807851328e-07,
"loss": 1.8935,
"step": 464
},
{
"epoch": 8.51,
"learning_rate": 6.619614817762537e-07,
"loss": 1.7803,
"step": 468
},
{
"epoch": 8.58,
"learning_rate": 6.002487109467347e-07,
"loss": 1.8441,
"step": 472
},
{
"epoch": 8.65,
"learning_rate": 5.413713387966329e-07,
"loss": 1.8613,
"step": 476
},
{
"epoch": 8.73,
"learning_rate": 4.853673085668947e-07,
"loss": 1.8198,
"step": 480
},
{
"epoch": 8.8,
"learning_rate": 4.322727117869951e-07,
"loss": 1.8694,
"step": 484
},
{
"epoch": 8.87,
"learning_rate": 3.821217650159453e-07,
"loss": 1.9331,
"step": 488
},
{
"epoch": 8.95,
"learning_rate": 3.3494678779157464e-07,
"loss": 1.8121,
"step": 492
},
{
"epoch": 9.0,
"pls_score": 59.6,
"std": 4.435853920047414,
"step": 495
},
{
"epoch": 9.0,
"eval_loss": 1.9848822355270386,
"eval_runtime": 4.9078,
"eval_samples_per_second": 4.686,
"eval_steps_per_second": 1.223,
"step": 495
},
{
"epoch": 9.02,
"learning_rate": 2.9077818180237693e-07,
"loss": 1.8913,
"step": 496
},
{
"epoch": 9.09,
"learning_rate": 2.4964441129527337e-07,
"loss": 1.9085,
"step": 500
},
{
"epoch": 9.16,
"learning_rate": 2.1157198473197417e-07,
"loss": 1.8147,
"step": 504
},
{
"epoch": 9.24,
"learning_rate": 1.765854377057219e-07,
"loss": 1.801,
"step": 508
},
{
"epoch": 9.31,
"learning_rate": 1.4470731712944885e-07,
"loss": 1.971,
"step": 512
},
{
"epoch": 9.38,
"learning_rate": 1.1595816670552429e-07,
"loss": 1.785,
"step": 516
},
{
"epoch": 9.45,
"learning_rate": 9.035651368646647e-08,
"loss": 1.9679,
"step": 520
},
{
"epoch": 9.53,
"learning_rate": 6.791885693514134e-08,
"loss": 2.0259,
"step": 524
},
{
"epoch": 9.6,
"learning_rate": 4.865965629214819e-08,
"loss": 1.809,
"step": 528
},
{
"epoch": 9.67,
"learning_rate": 3.25913232572489e-08,
"loss": 1.8509,
"step": 532
},
{
"epoch": 9.75,
"learning_rate": 1.9724212990830938e-08,
"loss": 1.7555,
"step": 536
},
{
"epoch": 9.82,
"learning_rate": 1.006661764057837e-08,
"loss": 1.7928,
"step": 540
},
{
"epoch": 9.89,
"learning_rate": 3.6247609976319818e-09,
"loss": 1.8696,
"step": 544
},
{
"epoch": 9.96,
"learning_rate": 4.027944857032395e-10,
"loss": 1.874,
"step": 548
},
{
"epoch": 10.0,
"step": 550,
"total_flos": 1.8756405193949184e+16,
"train_loss": 0.0,
"train_runtime": 11.1151,
"train_samples_per_second": 195.23,
"train_steps_per_second": 49.482
}
],
"logging_steps": 4,
"max_steps": 550,
"num_input_tokens_seen": 0,
"num_train_epochs": 10,
"save_steps": 55,
"total_flos": 1.8756405193949184e+16,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}