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": 1.268365060389042,
|
|
"epoch": 0.02479082739386427,
|
|
"grad_norm": 2.875,
|
|
"learning_rate": 7.277227722772277e-06,
|
|
"loss": 1.4445,
|
|
"mean_token_accuracy": 0.6872620263695717,
|
|
"num_tokens": 12987057.0,
|
|
"step": 50
|
|
},
|
|
{
|
|
"entropy": 1.240434679389,
|
|
"epoch": 0.04958165478772854,
|
|
"grad_norm": 0.578125,
|
|
"learning_rate": 1.4702970297029704e-05,
|
|
"loss": 1.2448,
|
|
"mean_token_accuracy": 0.7104562397301197,
|
|
"num_tokens": 25982341.0,
|
|
"step": 100
|
|
},
|
|
{
|
|
"entropy": 1.1758279597759247,
|
|
"epoch": 0.07437248218159281,
|
|
"grad_norm": 0.5625,
|
|
"learning_rate": 2.212871287128713e-05,
|
|
"loss": 1.1713,
|
|
"mean_token_accuracy": 0.7240416544675827,
|
|
"num_tokens": 38965652.0,
|
|
"step": 150
|
|
},
|
|
{
|
|
"entropy": 1.1409691953659058,
|
|
"epoch": 0.09916330957545708,
|
|
"grad_norm": 0.6875,
|
|
"learning_rate": 2.9554455445544555e-05,
|
|
"loss": 1.1349,
|
|
"mean_token_accuracy": 0.730599504262209,
|
|
"num_tokens": 51955113.0,
|
|
"step": 200
|
|
},
|
|
{
|
|
"entropy": 1.1036084993183612,
|
|
"epoch": 0.12395413696932135,
|
|
"grad_norm": 0.58984375,
|
|
"learning_rate": 2.9950390604994976e-05,
|
|
"loss": 1.098,
|
|
"mean_token_accuracy": 0.7369328507781029,
|
|
"num_tokens": 64940008.0,
|
|
"step": 250
|
|
},
|
|
{
|
|
"entropy": 1.0838019071519376,
|
|
"epoch": 0.14874496436318563,
|
|
"grad_norm": 0.57421875,
|
|
"learning_rate": 2.978907368725783e-05,
|
|
"loss": 1.0771,
|
|
"mean_token_accuracy": 0.7411529240012169,
|
|
"num_tokens": 77929904.0,
|
|
"step": 300
|
|
},
|
|
{
|
|
"entropy": 1.0728265669941903,
|
|
"epoch": 0.1735357917570499,
|
|
"grad_norm": 0.60546875,
|
|
"learning_rate": 2.9517054537806585e-05,
|
|
"loss": 1.0663,
|
|
"mean_token_accuracy": 0.7428552787005901,
|
|
"num_tokens": 90917917.0,
|
|
"step": 350
|
|
},
|
|
{
|
|
"entropy": 1.0577942258119584,
|
|
"epoch": 0.19832661915091415,
|
|
"grad_norm": 0.546875,
|
|
"learning_rate": 2.9136369330613327e-05,
|
|
"loss": 1.0517,
|
|
"mean_token_accuracy": 0.7456462162733078,
|
|
"num_tokens": 103911070.0,
|
|
"step": 400
|
|
},
|
|
{
|
|
"entropy": 1.0415315183997154,
|
|
"epoch": 0.22311744654477844,
|
|
"grad_norm": 0.5859375,
|
|
"learning_rate": 2.864986764929012e-05,
|
|
"loss": 1.0354,
|
|
"mean_token_accuracy": 0.7489144913852215,
|
|
"num_tokens": 116908048.0,
|
|
"step": 450
|
|
},
|
|
{
|
|
"entropy": 1.0311143945157528,
|
|
"epoch": 0.2479082739386427,
|
|
"grad_norm": 0.609375,
|
|
"learning_rate": 2.8061191156796658e-05,
|
|
"loss": 1.0249,
|
|
"mean_token_accuracy": 0.7507216791808605,
|
|
"num_tokens": 129891292.0,
|
|
"step": 500
|
|
},
|
|
{
|
|
"entropy": 1.0225630864500999,
|
|
"epoch": 0.27269910133250697,
|
|
"grad_norm": 0.57421875,
|
|
"learning_rate": 2.737474633611185e-05,
|
|
"loss": 1.0163,
|
|
"mean_token_accuracy": 0.7522288835048676,
|
|
"num_tokens": 142876112.0,
|
|
"step": 550
|
|
},
|
|
{
|
|
"entropy": 1.0209170418977738,
|
|
"epoch": 0.29748992872637126,
|
|
"grad_norm": 0.546875,
|
|
"learning_rate": 2.6595671505916456e-05,
|
|
"loss": 1.0149,
|
|
"mean_token_accuracy": 0.7525184072554112,
|
|
"num_tokens": 155871785.0,
|
|
"step": 600
|
|
},
|
|
{
|
|
"entropy": 1.0186570417881011,
|
|
"epoch": 0.3222807561202355,
|
|
"grad_norm": 0.5546875,
|
|
"learning_rate": 2.5729798358187858e-05,
|
|
"loss": 1.0122,
|
|
"mean_token_accuracy": 0.753164530545473,
|
|
"num_tokens": 168864657.0,
|
|
"step": 650
|
|
},
|
|
{
|
|
"entropy": 1.0038706909120083,
|
|
"epoch": 0.3470715835140998,
|
|
"grad_norm": 0.5390625,
|
|
"learning_rate": 2.478360830561368e-05,
|
|
"loss": 0.9977,
|
|
"mean_token_accuracy": 0.755882608294487,
|
|
"num_tokens": 181847527.0,
|
|
"step": 700
|
|
},
|
|
{
|
|
"entropy": 1.005766232162714,
|
|
"epoch": 0.37186241090796407,
|
|
"grad_norm": 0.5703125,
|
|
"learning_rate": 2.376418396558165e-05,
|
|
"loss": 0.9993,
|
|
"mean_token_accuracy": 0.7553357391059399,
|
|
"num_tokens": 194832579.0,
|
|
"step": 750
|
|
},
|
|
{
|
|
"entropy": 1.001170701533556,
|
|
"epoch": 0.3966532383018283,
|
|
"grad_norm": 0.5703125,
|
|
"learning_rate": 2.2679156143907717e-05,
|
|
"loss": 0.9949,
|
|
"mean_token_accuracy": 0.7563820435106754,
|
|
"num_tokens": 207819397.0,
|
|
"step": 800
|
|
},
|
|
{
|
|
"entropy": 1.0034910333156586,
|
|
"epoch": 0.4214440656956926,
|
|
"grad_norm": 0.53125,
|
|
"learning_rate": 2.153664671515084e-05,
|
|
"loss": 0.9973,
|
|
"mean_token_accuracy": 0.755660359710455,
|
|
"num_tokens": 220806448.0,
|
|
"step": 850
|
|
},
|
|
{
|
|
"entropy": 0.9967121610045433,
|
|
"epoch": 0.4462348930895569,
|
|
"grad_norm": 0.5625,
|
|
"learning_rate": 2.0345207827078517e-05,
|
|
"loss": 0.9902,
|
|
"mean_token_accuracy": 0.7571546165645122,
|
|
"num_tokens": 233800554.0,
|
|
"step": 900
|
|
},
|
|
{
|
|
"entropy": 0.9938886980712414,
|
|
"epoch": 0.4710257204834211,
|
|
"grad_norm": 0.57421875,
|
|
"learning_rate": 1.9113757884362315e-05,
|
|
"loss": 0.9881,
|
|
"mean_token_accuracy": 0.7577755935490131,
|
|
"num_tokens": 246788996.0,
|
|
"step": 950
|
|
},
|
|
{
|
|
"entropy": 0.9981730921566486,
|
|
"epoch": 0.4958165478772854,
|
|
"grad_norm": 0.5234375,
|
|
"learning_rate": 1.7851514790691512e-05,
|
|
"loss": 0.9914,
|
|
"mean_token_accuracy": 0.7567691704630852,
|
|
"num_tokens": 259781274.0,
|
|
"step": 1000
|
|
},
|
|
{
|
|
"entropy": 0.9877532237768173,
|
|
"epoch": 0.5206073752711496,
|
|
"grad_norm": 0.53125,
|
|
"learning_rate": 1.6567926949014805e-05,
|
|
"loss": 0.982,
|
|
"mean_token_accuracy": 0.7587743204832077,
|
|
"num_tokens": 272766654.0,
|
|
"step": 1050
|
|
},
|
|
{
|
|
"entropy": 0.9910286869108677,
|
|
"epoch": 0.5453982026650139,
|
|
"grad_norm": 0.5546875,
|
|
"learning_rate": 1.5272602536401258e-05,
|
|
"loss": 0.9848,
|
|
"mean_token_accuracy": 0.7581908470392227,
|
|
"num_tokens": 285761894.0,
|
|
"step": 1100
|
|
},
|
|
{
|
|
"entropy": 0.993671209961176,
|
|
"epoch": 0.5701890300588782,
|
|
"grad_norm": 0.51953125,
|
|
"learning_rate": 1.3975237582927023e-05,
|
|
"loss": 0.9873,
|
|
"mean_token_accuracy": 0.7577326658368111,
|
|
"num_tokens": 298753069.0,
|
|
"step": 1150
|
|
},
|
|
{
|
|
"entropy": 0.9884775359928608,
|
|
"epoch": 0.5949798574527425,
|
|
"grad_norm": 0.54296875,
|
|
"learning_rate": 1.2685543392946551e-05,
|
|
"loss": 0.9823,
|
|
"mean_token_accuracy": 0.758614943921566,
|
|
"num_tokens": 311750957.0,
|
|
"step": 1200
|
|
},
|
|
{
|
|
"entropy": 0.9911277142167091,
|
|
"epoch": 0.6197706848466068,
|
|
"grad_norm": 0.59375,
|
|
"learning_rate": 1.1413173852029593e-05,
|
|
"loss": 0.9848,
|
|
"mean_token_accuracy": 0.758147294819355,
|
|
"num_tokens": 324739809.0,
|
|
"step": 1250
|
|
},
|
|
{
|
|
"entropy": 0.9882946291565895,
|
|
"epoch": 0.644561512240471,
|
|
"grad_norm": 0.5078125,
|
|
"learning_rate": 1.0167653163701218e-05,
|
|
"loss": 0.9821,
|
|
"mean_token_accuracy": 0.7587668345868588,
|
|
"num_tokens": 337733441.0,
|
|
"step": 1300
|
|
},
|
|
{
|
|
"entropy": 0.986912748515606,
|
|
"epoch": 0.6693523396343353,
|
|
"grad_norm": 0.55078125,
|
|
"learning_rate": 8.958304556904608e-06,
|
|
"loss": 0.9808,
|
|
"mean_token_accuracy": 0.758950265944004,
|
|
"num_tokens": 350729616.0,
|
|
"step": 1350
|
|
},
|
|
{
|
|
"entropy": 0.9840033027529717,
|
|
"epoch": 0.6941431670281996,
|
|
"grad_norm": 0.515625,
|
|
"learning_rate": 7.794180497840417e-06,
|
|
"loss": 0.9778,
|
|
"mean_token_accuracy": 0.7596159276366233,
|
|
"num_tokens": 363726550.0,
|
|
"step": 1400
|
|
},
|
|
{
|
|
"entropy": 0.983212937861681,
|
|
"epoch": 0.7189339944220638,
|
|
"grad_norm": 0.53515625,
|
|
"learning_rate": 6.683994928575307e-06,
|
|
"loss": 0.9772,
|
|
"mean_token_accuracy": 0.7596646909415722,
|
|
"num_tokens": 376715458.0,
|
|
"step": 1450
|
|
},
|
|
{
|
|
"entropy": 0.9884254159033299,
|
|
"epoch": 0.7437248218159281,
|
|
"grad_norm": 0.5234375,
|
|
"learning_rate": 5.636058039641382e-06,
|
|
"loss": 0.9819,
|
|
"mean_token_accuracy": 0.758556911200285,
|
|
"num_tokens": 389703743.0,
|
|
"step": 1500
|
|
},
|
|
{
|
|
"entropy": 0.9844865338504314,
|
|
"epoch": 0.7685156492097924,
|
|
"grad_norm": 0.51171875,
|
|
"learning_rate": 4.6582140648800434e-06,
|
|
"loss": 0.9779,
|
|
"mean_token_accuracy": 0.7594577698409557,
|
|
"num_tokens": 402693509.0,
|
|
"step": 1550
|
|
},
|
|
{
|
|
"entropy": 0.9945496095716954,
|
|
"epoch": 0.7933064766036566,
|
|
"grad_norm": 0.515625,
|
|
"learning_rate": 3.7577825641611918e-06,
|
|
"loss": 0.9893,
|
|
"mean_token_accuracy": 0.757219355404377,
|
|
"num_tokens": 415676636.0,
|
|
"step": 1600
|
|
},
|
|
{
|
|
"entropy": 0.988469408750534,
|
|
"epoch": 0.8180973039975209,
|
|
"grad_norm": 0.51171875,
|
|
"learning_rate": 2.941503633500518e-06,
|
|
"loss": 0.9828,
|
|
"mean_token_accuracy": 0.7585532386600972,
|
|
"num_tokens": 428672727.0,
|
|
"step": 1650
|
|
},
|
|
{
|
|
"entropy": 0.98266371935606,
|
|
"epoch": 0.8428881313913852,
|
|
"grad_norm": 0.515625,
|
|
"learning_rate": 2.215487452699424e-06,
|
|
"loss": 0.9761,
|
|
"mean_token_accuracy": 0.759987825602293,
|
|
"num_tokens": 441658704.0,
|
|
"step": 1700
|
|
},
|
|
{
|
|
"entropy": 0.9894817672669888,
|
|
"epoch": 0.8676789587852495,
|
|
"grad_norm": 0.53125,
|
|
"learning_rate": 1.5851685481640143e-06,
|
|
"loss": 0.9826,
|
|
"mean_token_accuracy": 0.7585795709490776,
|
|
"num_tokens": 454644190.0,
|
|
"step": 1750
|
|
},
|
|
{
|
|
"entropy": 0.9887270894646645,
|
|
"epoch": 0.8924697861791138,
|
|
"grad_norm": 0.5234375,
|
|
"learning_rate": 1.0552651132645602e-06,
|
|
"loss": 0.9826,
|
|
"mean_token_accuracy": 0.7584862732887268,
|
|
"num_tokens": 467637106.0,
|
|
"step": 1800
|
|
},
|
|
{
|
|
"entropy": 0.9908865810930729,
|
|
"epoch": 0.917260613572978,
|
|
"grad_norm": 0.54296875,
|
|
"learning_rate": 6.297436907390885e-07,
|
|
"loss": 0.9841,
|
|
"mean_token_accuracy": 0.758246801495552,
|
|
"num_tokens": 480637699.0,
|
|
"step": 1850
|
|
},
|
|
{
|
|
"entropy": 0.987025645673275,
|
|
"epoch": 0.9420514409668422,
|
|
"grad_norm": 0.53125,
|
|
"learning_rate": 3.117894815076755e-07,
|
|
"loss": 0.9815,
|
|
"mean_token_accuracy": 0.7589400841295719,
|
|
"num_tokens": 493631966.0,
|
|
"step": 1900
|
|
},
|
|
{
|
|
"entropy": 0.9918687331676483,
|
|
"epoch": 0.9668422683607065,
|
|
"grad_norm": 0.50390625,
|
|
"learning_rate": 1.0378250214803986e-07,
|
|
"loss": 0.9859,
|
|
"mean_token_accuracy": 0.7579144223034382,
|
|
"num_tokens": 506610108.0,
|
|
"step": 1950
|
|
},
|
|
{
|
|
"entropy": 0.9944067592918873,
|
|
"epoch": 0.9916330957545708,
|
|
"grad_norm": 0.5234375,
|
|
"learning_rate": 7.2797695034398924e-09,
|
|
"loss": 0.9885,
|
|
"mean_token_accuracy": 0.7573176918923855,
|
|
"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
|
|
}
|