435 lines
12 KiB
JSON
435 lines
12 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 1.0,
|
|
"eval_steps": 500,
|
|
"global_step": 2017,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"entropy": 1.466458451151848,
|
|
"epoch": 0.02479082739386427,
|
|
"grad_norm": 1.3046875,
|
|
"learning_rate": 7.277227722772277e-06,
|
|
"loss": 1.5758,
|
|
"mean_token_accuracy": 0.6734453928470612,
|
|
"num_tokens": 12987057.0,
|
|
"step": 50
|
|
},
|
|
{
|
|
"entropy": 1.3964170238375664,
|
|
"epoch": 0.04958165478772854,
|
|
"grad_norm": 0.8984375,
|
|
"learning_rate": 1.4702970297029704e-05,
|
|
"loss": 1.3806,
|
|
"mean_token_accuracy": 0.6944466464221477,
|
|
"num_tokens": 25982341.0,
|
|
"step": 100
|
|
},
|
|
{
|
|
"entropy": 1.2903039123117923,
|
|
"epoch": 0.07437248218159281,
|
|
"grad_norm": 0.6953125,
|
|
"learning_rate": 2.212871287128713e-05,
|
|
"loss": 1.2538,
|
|
"mean_token_accuracy": 0.7118277321755886,
|
|
"num_tokens": 38965652.0,
|
|
"step": 150
|
|
},
|
|
{
|
|
"entropy": 1.2123706243932246,
|
|
"epoch": 0.09916330957545708,
|
|
"grad_norm": 0.62109375,
|
|
"learning_rate": 2.9554455445544555e-05,
|
|
"loss": 1.191,
|
|
"mean_token_accuracy": 0.7229508677124977,
|
|
"num_tokens": 51955113.0,
|
|
"step": 200
|
|
},
|
|
{
|
|
"entropy": 1.1540314616262912,
|
|
"epoch": 0.12395413696932135,
|
|
"grad_norm": 0.60546875,
|
|
"learning_rate": 2.9950390604994976e-05,
|
|
"loss": 1.1391,
|
|
"mean_token_accuracy": 0.7324234773218632,
|
|
"num_tokens": 64940008.0,
|
|
"step": 250
|
|
},
|
|
{
|
|
"entropy": 1.121255104392767,
|
|
"epoch": 0.14874496436318563,
|
|
"grad_norm": 0.60546875,
|
|
"learning_rate": 2.978907368725783e-05,
|
|
"loss": 1.1099,
|
|
"mean_token_accuracy": 0.7375885286927223,
|
|
"num_tokens": 77929904.0,
|
|
"step": 300
|
|
},
|
|
{
|
|
"entropy": 1.1031529605388641,
|
|
"epoch": 0.1735357917570499,
|
|
"grad_norm": 0.61328125,
|
|
"learning_rate": 2.9517054537806585e-05,
|
|
"loss": 1.0939,
|
|
"mean_token_accuracy": 0.740147297680378,
|
|
"num_tokens": 90917917.0,
|
|
"step": 350
|
|
},
|
|
{
|
|
"entropy": 1.0829776644706726,
|
|
"epoch": 0.19832661915091415,
|
|
"grad_norm": 0.58203125,
|
|
"learning_rate": 2.9136369330613327e-05,
|
|
"loss": 1.0744,
|
|
"mean_token_accuracy": 0.7435841898620129,
|
|
"num_tokens": 103911070.0,
|
|
"step": 400
|
|
},
|
|
{
|
|
"entropy": 1.0623433598876,
|
|
"epoch": 0.22311744654477844,
|
|
"grad_norm": 0.6171875,
|
|
"learning_rate": 2.864986764929012e-05,
|
|
"loss": 1.0544,
|
|
"mean_token_accuracy": 0.7474137496948242,
|
|
"num_tokens": 116908048.0,
|
|
"step": 450
|
|
},
|
|
{
|
|
"entropy": 1.0478474554419517,
|
|
"epoch": 0.2479082739386427,
|
|
"grad_norm": 0.62109375,
|
|
"learning_rate": 2.8061191156796658e-05,
|
|
"loss": 1.0405,
|
|
"mean_token_accuracy": 0.7497137078642845,
|
|
"num_tokens": 129891292.0,
|
|
"step": 500
|
|
},
|
|
{
|
|
"entropy": 1.0365162779390813,
|
|
"epoch": 0.27269910133250697,
|
|
"grad_norm": 0.57421875,
|
|
"learning_rate": 2.737474633611185e-05,
|
|
"loss": 1.029,
|
|
"mean_token_accuracy": 0.7517231622338295,
|
|
"num_tokens": 142876112.0,
|
|
"step": 550
|
|
},
|
|
{
|
|
"entropy": 1.0335283496975898,
|
|
"epoch": 0.29748992872637126,
|
|
"grad_norm": 0.59765625,
|
|
"learning_rate": 2.6595671505916456e-05,
|
|
"loss": 1.0265,
|
|
"mean_token_accuracy": 0.7521103696525097,
|
|
"num_tokens": 155871785.0,
|
|
"step": 600
|
|
},
|
|
{
|
|
"entropy": 1.0292730098962783,
|
|
"epoch": 0.3222807561202355,
|
|
"grad_norm": 0.63671875,
|
|
"learning_rate": 2.5729798358187858e-05,
|
|
"loss": 1.0221,
|
|
"mean_token_accuracy": 0.7530479227006436,
|
|
"num_tokens": 168864657.0,
|
|
"step": 650
|
|
},
|
|
{
|
|
"entropy": 1.0121874098479748,
|
|
"epoch": 0.3470715835140998,
|
|
"grad_norm": 0.57421875,
|
|
"learning_rate": 2.478360830561368e-05,
|
|
"loss": 1.0054,
|
|
"mean_token_accuracy": 0.7560361577570438,
|
|
"num_tokens": 181847527.0,
|
|
"step": 700
|
|
},
|
|
{
|
|
"entropy": 1.0130389857292175,
|
|
"epoch": 0.37186241090796407,
|
|
"grad_norm": 0.546875,
|
|
"learning_rate": 2.376418396558165e-05,
|
|
"loss": 1.0063,
|
|
"mean_token_accuracy": 0.7556676927208901,
|
|
"num_tokens": 194832579.0,
|
|
"step": 750
|
|
},
|
|
{
|
|
"entropy": 1.0068794977664948,
|
|
"epoch": 0.3966532383018283,
|
|
"grad_norm": 0.62890625,
|
|
"learning_rate": 2.2679156143907717e-05,
|
|
"loss": 1.0002,
|
|
"mean_token_accuracy": 0.7569640880823135,
|
|
"num_tokens": 207819397.0,
|
|
"step": 800
|
|
},
|
|
{
|
|
"entropy": 1.0089694766700268,
|
|
"epoch": 0.4214440656956926,
|
|
"grad_norm": 0.55078125,
|
|
"learning_rate": 2.153664671515084e-05,
|
|
"loss": 1.0022,
|
|
"mean_token_accuracy": 0.7563263186812401,
|
|
"num_tokens": 220806448.0,
|
|
"step": 850
|
|
},
|
|
{
|
|
"entropy": 1.001319480240345,
|
|
"epoch": 0.4462348930895569,
|
|
"grad_norm": 0.546875,
|
|
"learning_rate": 2.0345207827078517e-05,
|
|
"loss": 0.9943,
|
|
"mean_token_accuracy": 0.7579582373797894,
|
|
"num_tokens": 233800554.0,
|
|
"step": 900
|
|
},
|
|
{
|
|
"entropy": 0.9973974451422691,
|
|
"epoch": 0.4710257204834211,
|
|
"grad_norm": 0.56640625,
|
|
"learning_rate": 1.9113757884362315e-05,
|
|
"loss": 0.9912,
|
|
"mean_token_accuracy": 0.7585813918709755,
|
|
"num_tokens": 246788996.0,
|
|
"step": 950
|
|
},
|
|
{
|
|
"entropy": 1.0010248126089574,
|
|
"epoch": 0.4958165478772854,
|
|
"grad_norm": 0.54296875,
|
|
"learning_rate": 1.7851514790691512e-05,
|
|
"loss": 0.9938,
|
|
"mean_token_accuracy": 0.7579035374522209,
|
|
"num_tokens": 259781274.0,
|
|
"step": 1000
|
|
},
|
|
{
|
|
"entropy": 0.9900908127427102,
|
|
"epoch": 0.5206073752711496,
|
|
"grad_norm": 0.546875,
|
|
"learning_rate": 1.6567926949014805e-05,
|
|
"loss": 0.9838,
|
|
"mean_token_accuracy": 0.759877897053957,
|
|
"num_tokens": 272766654.0,
|
|
"step": 1050
|
|
},
|
|
{
|
|
"entropy": 0.993307217657566,
|
|
"epoch": 0.5453982026650139,
|
|
"grad_norm": 0.55078125,
|
|
"learning_rate": 1.5272602536401258e-05,
|
|
"loss": 0.9867,
|
|
"mean_token_accuracy": 0.7593129487335681,
|
|
"num_tokens": 285761894.0,
|
|
"step": 1100
|
|
},
|
|
{
|
|
"entropy": 0.9955632282793522,
|
|
"epoch": 0.5701890300588782,
|
|
"grad_norm": 0.56640625,
|
|
"learning_rate": 1.3975237582927023e-05,
|
|
"loss": 0.9887,
|
|
"mean_token_accuracy": 0.7589140185713767,
|
|
"num_tokens": 298753069.0,
|
|
"step": 1150
|
|
},
|
|
{
|
|
"entropy": 0.9897587068378926,
|
|
"epoch": 0.5949798574527425,
|
|
"grad_norm": 0.53515625,
|
|
"learning_rate": 1.2685543392946551e-05,
|
|
"loss": 0.9833,
|
|
"mean_token_accuracy": 0.7598606558144092,
|
|
"num_tokens": 311750957.0,
|
|
"step": 1200
|
|
},
|
|
{
|
|
"entropy": 0.9926682394742966,
|
|
"epoch": 0.6197706848466068,
|
|
"grad_norm": 0.578125,
|
|
"learning_rate": 1.1413173852029593e-05,
|
|
"loss": 0.9857,
|
|
"mean_token_accuracy": 0.7594233611226082,
|
|
"num_tokens": 324739809.0,
|
|
"step": 1250
|
|
},
|
|
{
|
|
"entropy": 0.9897119259834289,
|
|
"epoch": 0.644561512240471,
|
|
"grad_norm": 0.5390625,
|
|
"learning_rate": 1.0167653163701218e-05,
|
|
"loss": 0.9829,
|
|
"mean_token_accuracy": 0.7600416775047779,
|
|
"num_tokens": 337733441.0,
|
|
"step": 1300
|
|
},
|
|
{
|
|
"entropy": 0.988048963546753,
|
|
"epoch": 0.6693523396343353,
|
|
"grad_norm": 0.546875,
|
|
"learning_rate": 8.958304556904608e-06,
|
|
"loss": 0.9812,
|
|
"mean_token_accuracy": 0.7602460590004921,
|
|
"num_tokens": 350729616.0,
|
|
"step": 1350
|
|
},
|
|
{
|
|
"entropy": 0.9850697210431099,
|
|
"epoch": 0.6941431670281996,
|
|
"grad_norm": 0.55078125,
|
|
"learning_rate": 7.794180497840417e-06,
|
|
"loss": 0.9782,
|
|
"mean_token_accuracy": 0.7609732107818127,
|
|
"num_tokens": 363726550.0,
|
|
"step": 1400
|
|
},
|
|
{
|
|
"entropy": 0.983981351107359,
|
|
"epoch": 0.7189339944220638,
|
|
"grad_norm": 0.546875,
|
|
"learning_rate": 6.683994928575307e-06,
|
|
"loss": 0.9773,
|
|
"mean_token_accuracy": 0.7610866691172123,
|
|
"num_tokens": 376715458.0,
|
|
"step": 1450
|
|
},
|
|
{
|
|
"entropy": 0.9889418958127498,
|
|
"epoch": 0.7437248218159281,
|
|
"grad_norm": 0.55859375,
|
|
"learning_rate": 5.636058039641382e-06,
|
|
"loss": 0.9818,
|
|
"mean_token_accuracy": 0.7599805426597596,
|
|
"num_tokens": 389703743.0,
|
|
"step": 1500
|
|
},
|
|
{
|
|
"entropy": 0.9846837730705738,
|
|
"epoch": 0.7685156492097924,
|
|
"grad_norm": 0.53515625,
|
|
"learning_rate": 4.6582140648800434e-06,
|
|
"loss": 0.9773,
|
|
"mean_token_accuracy": 0.7609700655937195,
|
|
"num_tokens": 402693509.0,
|
|
"step": 1550
|
|
},
|
|
{
|
|
"entropy": 0.9958163200318814,
|
|
"epoch": 0.7933064766036566,
|
|
"grad_norm": 0.52734375,
|
|
"learning_rate": 3.7577825641611918e-06,
|
|
"loss": 0.9896,
|
|
"mean_token_accuracy": 0.7585725654661656,
|
|
"num_tokens": 415676636.0,
|
|
"step": 1600
|
|
},
|
|
{
|
|
"entropy": 0.9893448100984097,
|
|
"epoch": 0.8180973039975209,
|
|
"grad_norm": 0.53125,
|
|
"learning_rate": 2.941503633500518e-06,
|
|
"loss": 0.9829,
|
|
"mean_token_accuracy": 0.7598716221749783,
|
|
"num_tokens": 428672727.0,
|
|
"step": 1650
|
|
},
|
|
{
|
|
"entropy": 0.9831250040233135,
|
|
"epoch": 0.8428881313913852,
|
|
"grad_norm": 0.53515625,
|
|
"learning_rate": 2.215487452699424e-06,
|
|
"loss": 0.976,
|
|
"mean_token_accuracy": 0.7612966455519199,
|
|
"num_tokens": 441658704.0,
|
|
"step": 1700
|
|
},
|
|
{
|
|
"entropy": 0.9900904747843743,
|
|
"epoch": 0.8676789587852495,
|
|
"grad_norm": 0.53125,
|
|
"learning_rate": 1.5851685481640143e-06,
|
|
"loss": 0.9827,
|
|
"mean_token_accuracy": 0.7600373357534409,
|
|
"num_tokens": 454644190.0,
|
|
"step": 1750
|
|
},
|
|
{
|
|
"entropy": 0.9897356040775775,
|
|
"epoch": 0.8924697861791138,
|
|
"grad_norm": 0.5546875,
|
|
"learning_rate": 1.0552651132645602e-06,
|
|
"loss": 0.983,
|
|
"mean_token_accuracy": 0.7598755882680416,
|
|
"num_tokens": 467637106.0,
|
|
"step": 1800
|
|
},
|
|
{
|
|
"entropy": 0.9918161721527576,
|
|
"epoch": 0.917260613572978,
|
|
"grad_norm": 0.5390625,
|
|
"learning_rate": 6.297436907390885e-07,
|
|
"loss": 0.9845,
|
|
"mean_token_accuracy": 0.7595949833095074,
|
|
"num_tokens": 480637699.0,
|
|
"step": 1850
|
|
},
|
|
{
|
|
"entropy": 0.9879055871069431,
|
|
"epoch": 0.9420514409668422,
|
|
"grad_norm": 0.546875,
|
|
"learning_rate": 3.117894815076755e-07,
|
|
"loss": 0.9816,
|
|
"mean_token_accuracy": 0.7603768594563007,
|
|
"num_tokens": 493631966.0,
|
|
"step": 1900
|
|
},
|
|
{
|
|
"entropy": 0.9933325570821762,
|
|
"epoch": 0.9668422683607065,
|
|
"grad_norm": 0.53515625,
|
|
"learning_rate": 1.0378250214803986e-07,
|
|
"loss": 0.9863,
|
|
"mean_token_accuracy": 0.7591915895044804,
|
|
"num_tokens": 506610108.0,
|
|
"step": 1950
|
|
},
|
|
{
|
|
"entropy": 0.9958015055954457,
|
|
"epoch": 0.9916330957545708,
|
|
"grad_norm": 0.54296875,
|
|
"learning_rate": 7.2797695034398924e-09,
|
|
"loss": 0.9891,
|
|
"mean_token_accuracy": 0.7586235144734382,
|
|
"num_tokens": 519598174.0,
|
|
"step": 2000
|
|
}
|
|
],
|
|
"logging_steps": 50,
|
|
"max_steps": 2017,
|
|
"num_input_tokens_seen": 0,
|
|
"num_train_epochs": 1,
|
|
"save_steps": 50000,
|
|
"stateful_callbacks": {
|
|
"TrainerControl": {
|
|
"args": {
|
|
"should_epoch_stop": false,
|
|
"should_evaluate": false,
|
|
"should_log": false,
|
|
"should_save": true,
|
|
"should_training_stop": true
|
|
},
|
|
"attributes": {}
|
|
}
|
|
},
|
|
"total_flos": 4.431062590428283e+18,
|
|
"train_batch_size": 2,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|