autoprogrammer's picture
Add checkpoint-342
ba7684a verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"eval_steps": 500,
"global_step": 342,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.043859649122807015,
"grad_norm": 68.8519515991211,
"learning_rate": 3.6363636363636366e-06,
"loss": 3.8867,
"step": 5
},
{
"epoch": 0.08771929824561403,
"grad_norm": 21.211647033691406,
"learning_rate": 8.181818181818183e-06,
"loss": 1.8678,
"step": 10
},
{
"epoch": 0.13157894736842105,
"grad_norm": 19.21977424621582,
"learning_rate": 9.997973265157192e-06,
"loss": 0.5404,
"step": 15
},
{
"epoch": 0.17543859649122806,
"grad_norm": 8.171814918518066,
"learning_rate": 9.985593613021873e-06,
"loss": 0.3354,
"step": 20
},
{
"epoch": 0.21929824561403508,
"grad_norm": 6.866504669189453,
"learning_rate": 9.961988113473708e-06,
"loss": 0.2407,
"step": 25
},
{
"epoch": 0.2631578947368421,
"grad_norm": 4.359035968780518,
"learning_rate": 9.927209918004095e-06,
"loss": 0.1945,
"step": 30
},
{
"epoch": 0.30701754385964913,
"grad_norm": 3.8699371814727783,
"learning_rate": 9.881337335184879e-06,
"loss": 0.1668,
"step": 35
},
{
"epoch": 0.3508771929824561,
"grad_norm": 5.917017459869385,
"learning_rate": 9.824473654344297e-06,
"loss": 0.1493,
"step": 40
},
{
"epoch": 0.39473684210526316,
"grad_norm": 7.350202560424805,
"learning_rate": 9.756746912994832e-06,
"loss": 0.1837,
"step": 45
},
{
"epoch": 0.43859649122807015,
"grad_norm": 2.7234973907470703,
"learning_rate": 9.678309608536626e-06,
"loss": 0.1241,
"step": 50
},
{
"epoch": 0.4824561403508772,
"grad_norm": 4.651773452758789,
"learning_rate": 9.58933835488563e-06,
"loss": 0.1548,
"step": 55
},
{
"epoch": 0.5263157894736842,
"grad_norm": 6.711911201477051,
"learning_rate": 9.490033484799608e-06,
"loss": 0.1292,
"step": 60
},
{
"epoch": 0.5701754385964912,
"grad_norm": 6.491599082946777,
"learning_rate": 9.380618598797473e-06,
"loss": 0.138,
"step": 65
},
{
"epoch": 0.6140350877192983,
"grad_norm": 3.8013641834259033,
"learning_rate": 9.26134006168757e-06,
"loss": 0.1113,
"step": 70
},
{
"epoch": 0.6578947368421053,
"grad_norm": 3.2369492053985596,
"learning_rate": 9.132466447838598e-06,
"loss": 0.1128,
"step": 75
},
{
"epoch": 0.7017543859649122,
"grad_norm": 4.51101016998291,
"learning_rate": 8.994287936442226e-06,
"loss": 0.1093,
"step": 80
},
{
"epoch": 0.7456140350877193,
"grad_norm": 6.025557518005371,
"learning_rate": 8.84711565812904e-06,
"loss": 0.131,
"step": 85
},
{
"epoch": 0.7894736842105263,
"grad_norm": 3.6758241653442383,
"learning_rate": 8.691280994409044e-06,
"loss": 0.119,
"step": 90
},
{
"epoch": 0.8333333333333334,
"grad_norm": 7.3002214431762695,
"learning_rate": 8.527134831514116e-06,
"loss": 0.118,
"step": 95
},
{
"epoch": 0.8771929824561403,
"grad_norm": 3.0913829803466797,
"learning_rate": 8.355046770322528e-06,
"loss": 0.0927,
"step": 100
},
{
"epoch": 0.9210526315789473,
"grad_norm": 5.307352066040039,
"learning_rate": 8.175404294144482e-06,
"loss": 0.1152,
"step": 105
},
{
"epoch": 0.9649122807017544,
"grad_norm": 4.598392963409424,
"learning_rate": 7.98861189624256e-06,
"loss": 0.1091,
"step": 110
},
{
"epoch": 1.0087719298245614,
"grad_norm": 2.2277824878692627,
"learning_rate": 7.79509016905158e-06,
"loss": 0.1107,
"step": 115
},
{
"epoch": 1.0526315789473684,
"grad_norm": 4.761157989501953,
"learning_rate": 7.595274857148651e-06,
"loss": 0.0573,
"step": 120
},
{
"epoch": 1.0964912280701755,
"grad_norm": 2.9102399349212646,
"learning_rate": 7.389615876105773e-06,
"loss": 0.0453,
"step": 125
},
{
"epoch": 1.1403508771929824,
"grad_norm": 2.3797786235809326,
"learning_rate": 7.178576299434239e-06,
"loss": 0.0634,
"step": 130
},
{
"epoch": 1.1842105263157894,
"grad_norm": 3.0425777435302734,
"learning_rate": 6.962631315901861e-06,
"loss": 0.0643,
"step": 135
},
{
"epoch": 1.2280701754385965,
"grad_norm": 4.3089189529418945,
"learning_rate": 6.742267159570796e-06,
"loss": 0.0573,
"step": 140
},
{
"epoch": 1.2719298245614035,
"grad_norm": 1.6002540588378906,
"learning_rate": 6.51798001496514e-06,
"loss": 0.0591,
"step": 145
},
{
"epoch": 1.3157894736842106,
"grad_norm": 1.6954922676086426,
"learning_rate": 6.290274899833517e-06,
"loss": 0.0493,
"step": 150
},
{
"epoch": 1.3596491228070176,
"grad_norm": 2.1223790645599365,
"learning_rate": 6.059664528022267e-06,
"loss": 0.0514,
"step": 155
},
{
"epoch": 1.4035087719298245,
"grad_norm": 2.355539321899414,
"learning_rate": 5.82666815501964e-06,
"loss": 0.0552,
"step": 160
},
{
"epoch": 1.4473684210526316,
"grad_norm": 2.015369415283203,
"learning_rate": 5.5918104087704925e-06,
"loss": 0.0481,
"step": 165
},
{
"epoch": 1.4912280701754386,
"grad_norm": 1.41754949092865,
"learning_rate": 5.355620108394018e-06,
"loss": 0.0331,
"step": 170
},
{
"epoch": 1.5350877192982457,
"grad_norm": 2.061138868331909,
"learning_rate": 5.118629073464424e-06,
"loss": 0.0596,
"step": 175
},
{
"epoch": 1.5789473684210527,
"grad_norm": 2.276271343231201,
"learning_rate": 4.8813709265355766e-06,
"loss": 0.0738,
"step": 180
},
{
"epoch": 1.6228070175438596,
"grad_norm": 2.1886091232299805,
"learning_rate": 4.644379891605983e-06,
"loss": 0.0527,
"step": 185
},
{
"epoch": 1.6666666666666665,
"grad_norm": 3.1414003372192383,
"learning_rate": 4.40818959122951e-06,
"loss": 0.0774,
"step": 190
},
{
"epoch": 1.7105263157894737,
"grad_norm": 1.3887311220169067,
"learning_rate": 4.173331844980362e-06,
"loss": 0.0473,
"step": 195
},
{
"epoch": 1.7543859649122808,
"grad_norm": 2.8565640449523926,
"learning_rate": 3.940335471977733e-06,
"loss": 0.0408,
"step": 200
},
{
"epoch": 1.7982456140350878,
"grad_norm": 2.233248472213745,
"learning_rate": 3.7097251001664824e-06,
"loss": 0.055,
"step": 205
},
{
"epoch": 1.8421052631578947,
"grad_norm": 2.5530991554260254,
"learning_rate": 3.482019985034861e-06,
"loss": 0.048,
"step": 210
},
{
"epoch": 1.8859649122807016,
"grad_norm": 1.6235532760620117,
"learning_rate": 3.257732840429206e-06,
"loss": 0.0462,
"step": 215
},
{
"epoch": 1.9298245614035088,
"grad_norm": 1.505427598953247,
"learning_rate": 3.0373686840981396e-06,
"loss": 0.044,
"step": 220
},
{
"epoch": 1.973684210526316,
"grad_norm": 1.289949655532837,
"learning_rate": 2.821423700565763e-06,
"loss": 0.0362,
"step": 225
},
{
"epoch": 2.017543859649123,
"grad_norm": 1.6964356899261475,
"learning_rate": 2.610384123894229e-06,
"loss": 0.0255,
"step": 230
},
{
"epoch": 2.06140350877193,
"grad_norm": 1.0511788129806519,
"learning_rate": 2.4047251428513485e-06,
"loss": 0.0206,
"step": 235
},
{
"epoch": 2.1052631578947367,
"grad_norm": 0.7132877707481384,
"learning_rate": 2.2049098309484195e-06,
"loss": 0.0148,
"step": 240
},
{
"epoch": 2.1491228070175437,
"grad_norm": 0.5999829769134521,
"learning_rate": 2.0113881037574423e-06,
"loss": 0.01,
"step": 245
},
{
"epoch": 2.192982456140351,
"grad_norm": 1.44797682762146,
"learning_rate": 1.8245957058555203e-06,
"loss": 0.0158,
"step": 250
},
{
"epoch": 2.236842105263158,
"grad_norm": 1.1076303720474243,
"learning_rate": 1.6449532296774739e-06,
"loss": 0.0166,
"step": 255
},
{
"epoch": 2.280701754385965,
"grad_norm": 0.6465215682983398,
"learning_rate": 1.4728651684858835e-06,
"loss": 0.0098,
"step": 260
},
{
"epoch": 2.324561403508772,
"grad_norm": 0.8368536233901978,
"learning_rate": 1.3087190055909572e-06,
"loss": 0.008,
"step": 265
},
{
"epoch": 2.3684210526315788,
"grad_norm": 0.5175135135650635,
"learning_rate": 1.1528843418709623e-06,
"loss": 0.0117,
"step": 270
},
{
"epoch": 2.412280701754386,
"grad_norm": 1.0726501941680908,
"learning_rate": 1.005712063557776e-06,
"loss": 0.0093,
"step": 275
},
{
"epoch": 2.456140350877193,
"grad_norm": 0.8683268427848816,
"learning_rate": 8.675335521614036e-07,
"loss": 0.0103,
"step": 280
},
{
"epoch": 2.5,
"grad_norm": 0.5265653729438782,
"learning_rate": 7.386599383124321e-07,
"loss": 0.0064,
"step": 285
},
{
"epoch": 2.543859649122807,
"grad_norm": 2.1967923641204834,
"learning_rate": 6.193814012025278e-07,
"loss": 0.0146,
"step": 290
},
{
"epoch": 2.587719298245614,
"grad_norm": 2.777782440185547,
"learning_rate": 5.099665152003929e-07,
"loss": 0.0156,
"step": 295
},
{
"epoch": 2.6315789473684212,
"grad_norm": 1.3673884868621826,
"learning_rate": 4.106616451143719e-07,
"loss": 0.0125,
"step": 300
},
{
"epoch": 2.675438596491228,
"grad_norm": 0.8641520738601685,
"learning_rate": 3.2169039146337457e-07,
"loss": 0.0133,
"step": 305
},
{
"epoch": 2.719298245614035,
"grad_norm": 1.0037518739700317,
"learning_rate": 2.4325308700516805e-07,
"loss": 0.0091,
"step": 310
},
{
"epoch": 2.763157894736842,
"grad_norm": 0.5839938521385193,
"learning_rate": 1.7552634565570325e-07,
"loss": 0.0116,
"step": 315
},
{
"epoch": 2.807017543859649,
"grad_norm": 1.2752776145935059,
"learning_rate": 1.1866266481512234e-07,
"loss": 0.0113,
"step": 320
},
{
"epoch": 2.8508771929824563,
"grad_norm": 2.198166608810425,
"learning_rate": 7.279008199590543e-08,
"loss": 0.0167,
"step": 325
},
{
"epoch": 2.8947368421052633,
"grad_norm": 0.4964834153652191,
"learning_rate": 3.8011886526292394e-08,
"loss": 0.0135,
"step": 330
},
{
"epoch": 2.93859649122807,
"grad_norm": 1.2375760078430176,
"learning_rate": 1.4406386978128017e-08,
"loss": 0.0084,
"step": 335
},
{
"epoch": 2.982456140350877,
"grad_norm": 0.7600357532501221,
"learning_rate": 2.0267348428087974e-09,
"loss": 0.0093,
"step": 340
}
],
"logging_steps": 5,
"max_steps": 342,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 5.409409897081078e+17,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}