550 lines
15 KiB
JSON
550 lines
15 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 7.0,
|
|
"eval_steps": 500,
|
|
"global_step": 231,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"epoch": 0.15151515151515152,
|
|
"grad_norm": 3.5902750493970834,
|
|
"learning_rate": 6.666666666666667e-06,
|
|
"loss": 0.367,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.11483248323202133,
|
|
"step": 5,
|
|
"valid_targets_mean": 6069.3,
|
|
"valid_targets_min": 1102
|
|
},
|
|
{
|
|
"epoch": 0.30303030303030304,
|
|
"grad_norm": 1.523293973634436,
|
|
"learning_rate": 1.5000000000000002e-05,
|
|
"loss": 0.3295,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.09902024269104004,
|
|
"step": 10,
|
|
"valid_targets_mean": 5740.2,
|
|
"valid_targets_min": 975
|
|
},
|
|
{
|
|
"epoch": 0.45454545454545453,
|
|
"grad_norm": 0.5204685681128319,
|
|
"learning_rate": 2.3333333333333336e-05,
|
|
"loss": 0.2893,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.09341852366924286,
|
|
"step": 15,
|
|
"valid_targets_mean": 6076.2,
|
|
"valid_targets_min": 826
|
|
},
|
|
{
|
|
"epoch": 0.6060606060606061,
|
|
"grad_norm": 0.42120403433587034,
|
|
"learning_rate": 3.1666666666666666e-05,
|
|
"loss": 0.2586,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.09784922748804092,
|
|
"step": 20,
|
|
"valid_targets_mean": 6676.1,
|
|
"valid_targets_min": 1629
|
|
},
|
|
{
|
|
"epoch": 0.7575757575757576,
|
|
"grad_norm": 0.29716404096054466,
|
|
"learning_rate": 4e-05,
|
|
"loss": 0.2338,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.07702332735061646,
|
|
"step": 25,
|
|
"valid_targets_mean": 5522.2,
|
|
"valid_targets_min": 1847
|
|
},
|
|
{
|
|
"epoch": 0.9090909090909091,
|
|
"grad_norm": 0.2074554716015637,
|
|
"learning_rate": 3.994244399375679e-05,
|
|
"loss": 0.2068,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.08520618081092834,
|
|
"step": 30,
|
|
"valid_targets_mean": 7733.2,
|
|
"valid_targets_min": 1249
|
|
},
|
|
{
|
|
"epoch": 1.0606060606060606,
|
|
"grad_norm": 0.19383308521357237,
|
|
"learning_rate": 3.977010724441261e-05,
|
|
"loss": 0.1976,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.05378055199980736,
|
|
"step": 35,
|
|
"valid_targets_mean": 4624.0,
|
|
"valid_targets_min": 355
|
|
},
|
|
{
|
|
"epoch": 1.2121212121212122,
|
|
"grad_norm": 0.16410603240518978,
|
|
"learning_rate": 3.9483981653469586e-05,
|
|
"loss": 0.1883,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.06615366041660309,
|
|
"step": 40,
|
|
"valid_targets_mean": 6550.9,
|
|
"valid_targets_min": 1255
|
|
},
|
|
{
|
|
"epoch": 1.3636363636363638,
|
|
"grad_norm": 0.15449427864484538,
|
|
"learning_rate": 3.908571404555758e-05,
|
|
"loss": 0.1793,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.05769656226038933,
|
|
"step": 45,
|
|
"valid_targets_mean": 6499.9,
|
|
"valid_targets_min": 1751
|
|
},
|
|
{
|
|
"epoch": 1.5151515151515151,
|
|
"grad_norm": 0.14398778497253703,
|
|
"learning_rate": 3.8577596689969346e-05,
|
|
"loss": 0.18,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.056848011910915375,
|
|
"step": 50,
|
|
"valid_targets_mean": 7237.5,
|
|
"valid_targets_min": 372
|
|
},
|
|
{
|
|
"epoch": 1.6666666666666665,
|
|
"grad_norm": 0.13956070879698368,
|
|
"learning_rate": 3.7962554107273926e-05,
|
|
"loss": 0.1726,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.06066435948014259,
|
|
"step": 55,
|
|
"valid_targets_mean": 6926.7,
|
|
"valid_targets_min": 1357
|
|
},
|
|
{
|
|
"epoch": 1.8181818181818183,
|
|
"grad_norm": 0.1443322809922013,
|
|
"learning_rate": 3.724412623694427e-05,
|
|
"loss": 0.173,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.05192989856004715,
|
|
"step": 60,
|
|
"valid_targets_mean": 5459.8,
|
|
"valid_targets_min": 433
|
|
},
|
|
{
|
|
"epoch": 1.9696969696969697,
|
|
"grad_norm": 0.15138224844404377,
|
|
"learning_rate": 3.642644806287938e-05,
|
|
"loss": 0.178,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.0544576533138752,
|
|
"step": 65,
|
|
"valid_targets_mean": 6176.4,
|
|
"valid_targets_min": 1173
|
|
},
|
|
{
|
|
"epoch": 2.121212121212121,
|
|
"grad_norm": 0.14788211295786652,
|
|
"learning_rate": 3.55142258140884e-05,
|
|
"loss": 0.1706,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.050428979098796844,
|
|
"step": 70,
|
|
"valid_targets_mean": 6172.1,
|
|
"valid_targets_min": 1078
|
|
},
|
|
{
|
|
"epoch": 2.2727272727272725,
|
|
"grad_norm": 0.1489025346241851,
|
|
"learning_rate": 3.451270987751598e-05,
|
|
"loss": 0.1622,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.0515875369310379,
|
|
"step": 75,
|
|
"valid_targets_mean": 5412.4,
|
|
"valid_targets_min": 1256
|
|
},
|
|
{
|
|
"epoch": 2.4242424242424243,
|
|
"grad_norm": 0.15974193131852008,
|
|
"learning_rate": 3.342766457891194e-05,
|
|
"loss": 0.1596,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.052027106285095215,
|
|
"step": 80,
|
|
"valid_targets_mean": 5327.2,
|
|
"valid_targets_min": 1029
|
|
},
|
|
{
|
|
"epoch": 2.5757575757575757,
|
|
"grad_norm": 0.14890130460372,
|
|
"learning_rate": 3.226533500567433e-05,
|
|
"loss": 0.1565,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.05260023474693298,
|
|
"step": 85,
|
|
"valid_targets_mean": 6336.1,
|
|
"valid_targets_min": 826
|
|
},
|
|
{
|
|
"epoch": 2.7272727272727275,
|
|
"grad_norm": 0.15497597762513154,
|
|
"learning_rate": 3.1032411062620544e-05,
|
|
"loss": 0.1604,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.0602923147380352,
|
|
"step": 90,
|
|
"valid_targets_mean": 5714.4,
|
|
"valid_targets_min": 1409
|
|
},
|
|
{
|
|
"epoch": 2.878787878787879,
|
|
"grad_norm": 0.14743414796367607,
|
|
"learning_rate": 2.973598896756697e-05,
|
|
"loss": 0.162,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.04700557142496109,
|
|
"step": 95,
|
|
"valid_targets_mean": 5250.4,
|
|
"valid_targets_min": 978
|
|
},
|
|
{
|
|
"epoch": 3.0303030303030303,
|
|
"grad_norm": 0.14875208431462822,
|
|
"learning_rate": 2.8383530408333285e-05,
|
|
"loss": 0.1527,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.050968416035175323,
|
|
"step": 100,
|
|
"valid_targets_mean": 6015.5,
|
|
"valid_targets_min": 1215
|
|
},
|
|
{
|
|
"epoch": 3.1818181818181817,
|
|
"grad_norm": 0.14720059201307634,
|
|
"learning_rate": 2.6982819596247373e-05,
|
|
"loss": 0.1494,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.048501770943403244,
|
|
"step": 105,
|
|
"valid_targets_mean": 6575.0,
|
|
"valid_targets_min": 1490
|
|
},
|
|
{
|
|
"epoch": 3.3333333333333335,
|
|
"grad_norm": 0.14554091849093376,
|
|
"learning_rate": 2.554191846333378e-05,
|
|
"loss": 0.1517,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.04890170693397522,
|
|
"step": 110,
|
|
"valid_targets_mean": 6384.5,
|
|
"valid_targets_min": 1102
|
|
},
|
|
{
|
|
"epoch": 3.484848484848485,
|
|
"grad_norm": 0.14816303968679037,
|
|
"learning_rate": 2.4069120261052682e-05,
|
|
"loss": 0.1503,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.05088581144809723,
|
|
"step": 115,
|
|
"valid_targets_mean": 6863.3,
|
|
"valid_targets_min": 2033
|
|
},
|
|
{
|
|
"epoch": 3.6363636363636362,
|
|
"grad_norm": 0.18861192867251467,
|
|
"learning_rate": 2.2572901827656626e-05,
|
|
"loss": 0.1504,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.051915183663368225,
|
|
"step": 120,
|
|
"valid_targets_mean": 5318.8,
|
|
"valid_targets_min": 909
|
|
},
|
|
{
|
|
"epoch": 3.787878787878788,
|
|
"grad_norm": 0.17975884303700998,
|
|
"learning_rate": 2.1061874798894992e-05,
|
|
"loss": 0.1471,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.04352159798145294,
|
|
"step": 125,
|
|
"valid_targets_mean": 5009.9,
|
|
"valid_targets_min": 372
|
|
},
|
|
{
|
|
"epoch": 3.9393939393939394,
|
|
"grad_norm": 0.14992809475322363,
|
|
"learning_rate": 1.9544736042877886e-05,
|
|
"loss": 0.1483,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.06831197440624237,
|
|
"step": 130,
|
|
"valid_targets_mean": 6830.1,
|
|
"valid_targets_min": 947
|
|
},
|
|
{
|
|
"epoch": 4.090909090909091,
|
|
"grad_norm": 0.14832791011346455,
|
|
"learning_rate": 1.8030217604376628e-05,
|
|
"loss": 0.1394,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.04658203199505806,
|
|
"step": 135,
|
|
"valid_targets_mean": 5643.2,
|
|
"valid_targets_min": 355
|
|
},
|
|
{
|
|
"epoch": 4.242424242424242,
|
|
"grad_norm": 0.17266841542411188,
|
|
"learning_rate": 1.6527036446661396e-05,
|
|
"loss": 0.1412,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.046023882925510406,
|
|
"step": 140,
|
|
"valid_targets_mean": 6153.0,
|
|
"valid_targets_min": 510
|
|
},
|
|
{
|
|
"epoch": 4.393939393939394,
|
|
"grad_norm": 0.15432680905719748,
|
|
"learning_rate": 1.5043844280142005e-05,
|
|
"loss": 0.1369,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.04154321551322937,
|
|
"step": 145,
|
|
"valid_targets_mean": 6060.3,
|
|
"valid_targets_min": 563
|
|
},
|
|
{
|
|
"epoch": 4.545454545454545,
|
|
"grad_norm": 0.15852075197915375,
|
|
"learning_rate": 1.358917776657806e-05,
|
|
"loss": 0.1429,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.045568957924842834,
|
|
"step": 150,
|
|
"valid_targets_mean": 6214.1,
|
|
"valid_targets_min": 932
|
|
},
|
|
{
|
|
"epoch": 4.696969696969697,
|
|
"grad_norm": 0.15361105601774092,
|
|
"learning_rate": 1.2171409385463218e-05,
|
|
"loss": 0.1379,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.046185821294784546,
|
|
"step": 155,
|
|
"valid_targets_mean": 6136.9,
|
|
"valid_targets_min": 1255
|
|
},
|
|
{
|
|
"epoch": 4.848484848484849,
|
|
"grad_norm": 0.14910920681280979,
|
|
"learning_rate": 1.0798699245376959e-05,
|
|
"loss": 0.147,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.053926583379507065,
|
|
"step": 160,
|
|
"valid_targets_mean": 7605.3,
|
|
"valid_targets_min": 1808
|
|
},
|
|
{
|
|
"epoch": 5.0,
|
|
"grad_norm": 0.15956391595199854,
|
|
"learning_rate": 9.478948117658577e-06,
|
|
"loss": 0.1418,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.04224388301372528,
|
|
"step": 165,
|
|
"valid_targets_mean": 4956.4,
|
|
"valid_targets_min": 1186
|
|
},
|
|
{
|
|
"epoch": 5.151515151515151,
|
|
"grad_norm": 0.16166354923486542,
|
|
"learning_rate": 8.219751962722726e-06,
|
|
"loss": 0.1385,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.04815865680575371,
|
|
"step": 170,
|
|
"valid_targets_mean": 6351.8,
|
|
"valid_targets_min": 1414
|
|
},
|
|
{
|
|
"epoch": 5.303030303030303,
|
|
"grad_norm": 0.15713749630090504,
|
|
"learning_rate": 7.028358210744881e-06,
|
|
"loss": 0.1335,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.04832237958908081,
|
|
"step": 175,
|
|
"valid_targets_mean": 6324.0,
|
|
"valid_targets_min": 537
|
|
},
|
|
{
|
|
"epoch": 5.454545454545454,
|
|
"grad_norm": 0.16623362872731065,
|
|
"learning_rate": 5.911624048347757e-06,
|
|
"loss": 0.1368,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.04412321001291275,
|
|
"step": 180,
|
|
"valid_targets_mean": 5107.0,
|
|
"valid_targets_min": 975
|
|
},
|
|
{
|
|
"epoch": 5.606060606060606,
|
|
"grad_norm": 0.14675183366084796,
|
|
"learning_rate": 4.875976951373633e-06,
|
|
"loss": 0.1307,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.03708821162581444,
|
|
"step": 185,
|
|
"valid_targets_mean": 6408.9,
|
|
"valid_targets_min": 1162
|
|
},
|
|
{
|
|
"epoch": 5.757575757575758,
|
|
"grad_norm": 0.16118513545891452,
|
|
"learning_rate": 3.927377690900436e-06,
|
|
"loss": 0.1369,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.04341595247387886,
|
|
"step": 190,
|
|
"valid_targets_mean": 6496.9,
|
|
"valid_targets_min": 1027
|
|
},
|
|
{
|
|
"epoch": 5.909090909090909,
|
|
"grad_norm": 0.18904653293771853,
|
|
"learning_rate": 3.071286025423983e-06,
|
|
"loss": 0.1389,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.059906527400016785,
|
|
"step": 195,
|
|
"valid_targets_mean": 6184.4,
|
|
"valid_targets_min": 978
|
|
},
|
|
{
|
|
"epoch": 6.0606060606060606,
|
|
"grad_norm": 0.16058499044576713,
|
|
"learning_rate": 2.312629276668554e-06,
|
|
"loss": 0.1477,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.0467909500002861,
|
|
"step": 200,
|
|
"valid_targets_mean": 5470.1,
|
|
"valid_targets_min": 1107
|
|
},
|
|
{
|
|
"epoch": 6.212121212121212,
|
|
"grad_norm": 0.16458313236617705,
|
|
"learning_rate": 1.6557739698909436e-06,
|
|
"loss": 0.1363,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.039594992995262146,
|
|
"step": 205,
|
|
"valid_targets_mean": 6872.1,
|
|
"valid_targets_min": 1222
|
|
},
|
|
{
|
|
"epoch": 6.363636363636363,
|
|
"grad_norm": 0.1520783842215095,
|
|
"learning_rate": 1.1045007019049182e-06,
|
|
"loss": 0.1381,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.042891182005405426,
|
|
"step": 210,
|
|
"valid_targets_mean": 5591.5,
|
|
"valid_targets_min": 1134
|
|
},
|
|
{
|
|
"epoch": 6.515151515151516,
|
|
"grad_norm": 0.17583323971560555,
|
|
"learning_rate": 6.619823814758786e-07,
|
|
"loss": 0.1336,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.04007503390312195,
|
|
"step": 215,
|
|
"valid_targets_mean": 5927.2,
|
|
"valid_targets_min": 1202
|
|
},
|
|
{
|
|
"epoch": 6.666666666666667,
|
|
"grad_norm": 0.15107528231407247,
|
|
"learning_rate": 3.307659673251595e-07,
|
|
"loss": 0.1378,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.05574905872344971,
|
|
"step": 220,
|
|
"valid_targets_mean": 7710.5,
|
|
"valid_targets_min": 975
|
|
},
|
|
{
|
|
"epoch": 6.818181818181818,
|
|
"grad_norm": 0.1525570611116513,
|
|
"learning_rate": 1.1275780885282806e-07,
|
|
"loss": 0.1311,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.036159858107566833,
|
|
"step": 225,
|
|
"valid_targets_mean": 6852.6,
|
|
"valid_targets_min": 474
|
|
},
|
|
{
|
|
"epoch": 6.96969696969697,
|
|
"grad_norm": 0.1772473530958802,
|
|
"learning_rate": 9.212673951897177e-09,
|
|
"loss": 0.1337,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.048351939767599106,
|
|
"step": 230,
|
|
"valid_targets_mean": 7166.3,
|
|
"valid_targets_min": 1121
|
|
},
|
|
{
|
|
"epoch": 7.0,
|
|
"step": 231,
|
|
"total_flos": 6.648213778192138e+17,
|
|
"train_loss": 0.0,
|
|
"train_runtime": 0.8229,
|
|
"train_samples_per_second": 26879.554,
|
|
"train_steps_per_second": 280.704
|
|
}
|
|
],
|
|
"logging_steps": 5,
|
|
"max_steps": 231,
|
|
"num_input_tokens_seen": 0,
|
|
"num_train_epochs": 7,
|
|
"save_steps": 300,
|
|
"stateful_callbacks": {
|
|
"TrainerControl": {
|
|
"args": {
|
|
"should_epoch_stop": false,
|
|
"should_evaluate": false,
|
|
"should_log": false,
|
|
"should_save": true,
|
|
"should_training_stop": true
|
|
},
|
|
"attributes": {}
|
|
}
|
|
},
|
|
"total_flos": 6.648213778192138e+17,
|
|
"train_batch_size": 1,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|