246 lines
6.7 KiB
JSON
246 lines
6.7 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 1.0,
|
|
"eval_steps": 500,
|
|
"global_step": 1048,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"entropy": 2.0116909193992614,
|
|
"epoch": 0.047744091668656,
|
|
"grad_norm": 3.390625,
|
|
"learning_rate": 1.0315789473684213e-05,
|
|
"loss": 2.0623187255859374,
|
|
"mean_token_accuracy": 0.589528747946024,
|
|
"num_tokens": 62415.0,
|
|
"step": 50
|
|
},
|
|
{
|
|
"entropy": 1.7590599358081818,
|
|
"epoch": 0.095488183337312,
|
|
"grad_norm": 3.078125,
|
|
"learning_rate": 1.9999915067429385e-05,
|
|
"loss": 1.754022216796875,
|
|
"mean_token_accuracy": 0.6332476472854615,
|
|
"num_tokens": 126004.0,
|
|
"step": 100
|
|
},
|
|
{
|
|
"entropy": 1.7224605268239974,
|
|
"epoch": 0.14323227500596802,
|
|
"grad_norm": 3.203125,
|
|
"learning_rate": 1.9984525009996967e-05,
|
|
"loss": 1.7351203918457032,
|
|
"mean_token_accuracy": 0.6340418252348899,
|
|
"num_tokens": 188614.0,
|
|
"step": 150
|
|
},
|
|
{
|
|
"entropy": 1.6983880323171616,
|
|
"epoch": 0.190976366674624,
|
|
"grad_norm": 2.984375,
|
|
"learning_rate": 1.9942640420376764e-05,
|
|
"loss": 1.7146597290039063,
|
|
"mean_token_accuracy": 0.6401417210698128,
|
|
"num_tokens": 251062.0,
|
|
"step": 200
|
|
},
|
|
{
|
|
"entropy": 1.6483143484592437,
|
|
"epoch": 0.23872045834328,
|
|
"grad_norm": 2.734375,
|
|
"learning_rate": 1.9874372441823628e-05,
|
|
"loss": 1.658107147216797,
|
|
"mean_token_accuracy": 0.6478794780373573,
|
|
"num_tokens": 313957.0,
|
|
"step": 250
|
|
},
|
|
{
|
|
"entropy": 1.6347757685184479,
|
|
"epoch": 0.28646455001193605,
|
|
"grad_norm": 2.703125,
|
|
"learning_rate": 1.9779902227487397e-05,
|
|
"loss": 1.6550674438476562,
|
|
"mean_token_accuracy": 0.6487274625897408,
|
|
"num_tokens": 376779.0,
|
|
"step": 300
|
|
},
|
|
{
|
|
"entropy": 1.5762060970067977,
|
|
"epoch": 0.334208641680592,
|
|
"grad_norm": 2.34375,
|
|
"learning_rate": 1.9659480459712203e-05,
|
|
"loss": 1.5960990905761718,
|
|
"mean_token_accuracy": 0.6587067684531211,
|
|
"num_tokens": 439472.0,
|
|
"step": 350
|
|
},
|
|
{
|
|
"entropy": 1.534512039422989,
|
|
"epoch": 0.381952733349248,
|
|
"grad_norm": 2.5625,
|
|
"learning_rate": 1.9513426684835922e-05,
|
|
"loss": 1.55486572265625,
|
|
"mean_token_accuracy": 0.6660523465275765,
|
|
"num_tokens": 501279.0,
|
|
"step": 400
|
|
},
|
|
{
|
|
"entropy": 1.5402596753835678,
|
|
"epoch": 0.429696825017904,
|
|
"grad_norm": 2.53125,
|
|
"learning_rate": 1.9342128465254943e-05,
|
|
"loss": 1.5535296630859374,
|
|
"mean_token_accuracy": 0.6673906499147415,
|
|
"num_tokens": 564442.0,
|
|
"step": 450
|
|
},
|
|
{
|
|
"entropy": 1.5164723497629167,
|
|
"epoch": 0.47744091668656,
|
|
"grad_norm": 2.34375,
|
|
"learning_rate": 1.9146040351004225e-05,
|
|
"loss": 1.5373231506347655,
|
|
"mean_token_accuracy": 0.670901304781437,
|
|
"num_tokens": 627102.0,
|
|
"step": 500
|
|
},
|
|
{
|
|
"entropy": 1.4891952139139175,
|
|
"epoch": 0.5251850083552161,
|
|
"grad_norm": 2.296875,
|
|
"learning_rate": 1.8925682673581706e-05,
|
|
"loss": 1.5036320495605469,
|
|
"mean_token_accuracy": 0.6744937270879745,
|
|
"num_tokens": 689421.0,
|
|
"step": 550
|
|
},
|
|
{
|
|
"entropy": 1.488253089785576,
|
|
"epoch": 0.5729291000238721,
|
|
"grad_norm": 2.15625,
|
|
"learning_rate": 1.868164016521766e-05,
|
|
"loss": 1.5044355773925782,
|
|
"mean_token_accuracy": 0.6755094349384307,
|
|
"num_tokens": 752388.0,
|
|
"step": 600
|
|
},
|
|
{
|
|
"entropy": 1.4446531629562378,
|
|
"epoch": 0.6206731916925281,
|
|
"grad_norm": 2.34375,
|
|
"learning_rate": 1.8414560407252874e-05,
|
|
"loss": 1.4617152404785156,
|
|
"mean_token_accuracy": 0.6805451279878616,
|
|
"num_tokens": 815803.0,
|
|
"step": 650
|
|
},
|
|
{
|
|
"entropy": 1.4555100780725478,
|
|
"epoch": 0.668417283361184,
|
|
"grad_norm": 2.109375,
|
|
"learning_rate": 1.8125152111742944e-05,
|
|
"loss": 1.469644775390625,
|
|
"mean_token_accuracy": 0.6796707132458687,
|
|
"num_tokens": 878686.0,
|
|
"step": 700
|
|
},
|
|
{
|
|
"entropy": 1.414481201171875,
|
|
"epoch": 0.71616137502984,
|
|
"grad_norm": 2.140625,
|
|
"learning_rate": 1.7814183240848635e-05,
|
|
"loss": 1.4264332580566406,
|
|
"mean_token_accuracy": 0.6893567860126495,
|
|
"num_tokens": 940629.0,
|
|
"step": 750
|
|
},
|
|
{
|
|
"entropy": 1.3958445823192596,
|
|
"epoch": 0.763905466698496,
|
|
"grad_norm": 2.34375,
|
|
"learning_rate": 1.7482478969002487e-05,
|
|
"loss": 1.4121397399902345,
|
|
"mean_token_accuracy": 0.6902168181538582,
|
|
"num_tokens": 1002810.0,
|
|
"step": 800
|
|
},
|
|
{
|
|
"entropy": 1.3906721925735475,
|
|
"epoch": 0.811649558367152,
|
|
"grad_norm": 2.421875,
|
|
"learning_rate": 1.7130919493259284e-05,
|
|
"loss": 1.3895584106445313,
|
|
"mean_token_accuracy": 0.6943912228941918,
|
|
"num_tokens": 1067041.0,
|
|
"step": 850
|
|
},
|
|
{
|
|
"entropy": 1.3663516825437545,
|
|
"epoch": 0.859393650035808,
|
|
"grad_norm": 2.375,
|
|
"learning_rate": 1.676043769764072e-05,
|
|
"loss": 1.3753732299804688,
|
|
"mean_token_accuracy": 0.6982124423980713,
|
|
"num_tokens": 1130309.0,
|
|
"step": 900
|
|
},
|
|
{
|
|
"entropy": 1.370076591670513,
|
|
"epoch": 0.907137741704464,
|
|
"grad_norm": 1.9296875,
|
|
"learning_rate": 1.637201667767204e-05,
|
|
"loss": 1.3839447021484375,
|
|
"mean_token_accuracy": 0.6986603051424026,
|
|
"num_tokens": 1192924.0,
|
|
"step": 950
|
|
},
|
|
{
|
|
"entropy": 1.3369020774960518,
|
|
"epoch": 0.95488183337312,
|
|
"grad_norm": 1.9609375,
|
|
"learning_rate": 1.5966687131679463e-05,
|
|
"loss": 1.3427734375,
|
|
"mean_token_accuracy": 0.7060227099061013,
|
|
"num_tokens": 1254418.0,
|
|
"step": 1000
|
|
},
|
|
{
|
|
"epoch": 1.0,
|
|
"eval_entropy": 1.2727481353174803,
|
|
"eval_loss": 1.291359543800354,
|
|
"eval_mean_token_accuracy": 0.7115251171210456,
|
|
"eval_num_tokens": 1314012.0,
|
|
"eval_runtime": 52.7795,
|
|
"eval_samples_per_second": 16.711,
|
|
"eval_steps_per_second": 16.711,
|
|
"step": 1048
|
|
}
|
|
],
|
|
"logging_steps": 50,
|
|
"max_steps": 3144,
|
|
"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": false
|
|
},
|
|
"attributes": {}
|
|
}
|
|
},
|
|
"total_flos": 7.962196730295091e+16,
|
|
"train_batch_size": 4,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|