345 lines
8.8 KiB
JSON
345 lines
8.8 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 1.0,
|
|
"eval_steps": 500,
|
|
"global_step": 313,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"entropy": 0.6765631157904863,
|
|
"epoch": 0.032,
|
|
"grad_norm": 2.875,
|
|
"learning_rate": 9.473684210526316e-05,
|
|
"loss": 3.33,
|
|
"mean_token_accuracy": 0.5463392566889524,
|
|
"num_tokens": 32197.0,
|
|
"step": 10
|
|
},
|
|
{
|
|
"entropy": 1.1614421313628553,
|
|
"epoch": 0.064,
|
|
"grad_norm": 1.7734375,
|
|
"learning_rate": 0.0002,
|
|
"loss": 1.6539,
|
|
"mean_token_accuracy": 0.6597556989639998,
|
|
"num_tokens": 64756.0,
|
|
"step": 20
|
|
},
|
|
{
|
|
"entropy": 0.939408042281866,
|
|
"epoch": 0.096,
|
|
"grad_norm": 0.60546875,
|
|
"learning_rate": 0.0001998660954234636,
|
|
"loss": 0.7808,
|
|
"mean_token_accuracy": 0.781595965847373,
|
|
"num_tokens": 95410.0,
|
|
"step": 30
|
|
},
|
|
{
|
|
"entropy": 0.9574495874345302,
|
|
"epoch": 0.128,
|
|
"grad_norm": 0.55078125,
|
|
"learning_rate": 0.00019946474030256675,
|
|
"loss": 0.859,
|
|
"mean_token_accuracy": 0.7595490481704473,
|
|
"num_tokens": 123619.0,
|
|
"step": 40
|
|
},
|
|
{
|
|
"entropy": 0.8130528301000595,
|
|
"epoch": 0.16,
|
|
"grad_norm": 0.498046875,
|
|
"learning_rate": 0.00019879700950305957,
|
|
"loss": 0.7411,
|
|
"mean_token_accuracy": 0.8014852415770293,
|
|
"num_tokens": 156066.0,
|
|
"step": 50
|
|
},
|
|
{
|
|
"entropy": 0.8076014310121536,
|
|
"epoch": 0.192,
|
|
"grad_norm": 0.67578125,
|
|
"learning_rate": 0.00019786469126914098,
|
|
"loss": 0.8028,
|
|
"mean_token_accuracy": 0.7921789325773716,
|
|
"num_tokens": 186345.0,
|
|
"step": 60
|
|
},
|
|
{
|
|
"entropy": 0.7697046212852001,
|
|
"epoch": 0.224,
|
|
"grad_norm": 0.58984375,
|
|
"learning_rate": 0.0001966702824343772,
|
|
"loss": 0.7332,
|
|
"mean_token_accuracy": 0.7981064647436142,
|
|
"num_tokens": 219449.0,
|
|
"step": 70
|
|
},
|
|
{
|
|
"entropy": 0.7365534473210573,
|
|
"epoch": 0.256,
|
|
"grad_norm": 0.66015625,
|
|
"learning_rate": 0.0001952169817349528,
|
|
"loss": 0.6818,
|
|
"mean_token_accuracy": 0.8138540856540203,
|
|
"num_tokens": 252398.0,
|
|
"step": 80
|
|
},
|
|
{
|
|
"entropy": 0.8327326748520136,
|
|
"epoch": 0.288,
|
|
"grad_norm": 0.447265625,
|
|
"learning_rate": 0.0001935086812431626,
|
|
"loss": 0.7288,
|
|
"mean_token_accuracy": 0.7888761077076196,
|
|
"num_tokens": 281833.0,
|
|
"step": 90
|
|
},
|
|
{
|
|
"entropy": 0.8304124176502228,
|
|
"epoch": 0.32,
|
|
"grad_norm": 0.57421875,
|
|
"learning_rate": 0.0001915499559440855,
|
|
"loss": 0.7666,
|
|
"mean_token_accuracy": 0.7884823590517044,
|
|
"num_tokens": 313194.0,
|
|
"step": 100
|
|
},
|
|
{
|
|
"entropy": 0.7173997653648257,
|
|
"epoch": 0.352,
|
|
"grad_norm": 0.609375,
|
|
"learning_rate": 0.00018934605148335608,
|
|
"loss": 0.6474,
|
|
"mean_token_accuracy": 0.8159435257315636,
|
|
"num_tokens": 345629.0,
|
|
"step": 110
|
|
},
|
|
{
|
|
"entropy": 0.7204551868140697,
|
|
"epoch": 0.384,
|
|
"grad_norm": 0.6640625,
|
|
"learning_rate": 0.00018690287011884508,
|
|
"loss": 0.674,
|
|
"mean_token_accuracy": 0.8059780243784189,
|
|
"num_tokens": 379281.0,
|
|
"step": 120
|
|
},
|
|
{
|
|
"entropy": 0.7626429490745068,
|
|
"epoch": 0.416,
|
|
"grad_norm": 0.470703125,
|
|
"learning_rate": 0.00018422695491387284,
|
|
"loss": 0.6938,
|
|
"mean_token_accuracy": 0.8045202873647213,
|
|
"num_tokens": 410187.0,
|
|
"step": 130
|
|
},
|
|
{
|
|
"entropy": 0.7266103811562061,
|
|
"epoch": 0.448,
|
|
"grad_norm": 0.484375,
|
|
"learning_rate": 0.00018132547221428675,
|
|
"loss": 0.7094,
|
|
"mean_token_accuracy": 0.8129661556333303,
|
|
"num_tokens": 442476.0,
|
|
"step": 140
|
|
},
|
|
{
|
|
"entropy": 0.7414866480976343,
|
|
"epoch": 0.48,
|
|
"grad_norm": 0.6875,
|
|
"learning_rate": 0.00017820619245633115,
|
|
"loss": 0.6545,
|
|
"mean_token_accuracy": 0.8073488272726536,
|
|
"num_tokens": 472224.0,
|
|
"step": 150
|
|
},
|
|
{
|
|
"entropy": 0.7822258766740561,
|
|
"epoch": 0.512,
|
|
"grad_norm": 0.54296875,
|
|
"learning_rate": 0.00017487746935670776,
|
|
"loss": 0.7011,
|
|
"mean_token_accuracy": 0.795686560869217,
|
|
"num_tokens": 501232.0,
|
|
"step": 160
|
|
},
|
|
{
|
|
"entropy": 0.745093465037644,
|
|
"epoch": 0.544,
|
|
"grad_norm": 0.55078125,
|
|
"learning_rate": 0.00017134821754055782,
|
|
"loss": 0.7028,
|
|
"mean_token_accuracy": 0.8048942796885967,
|
|
"num_tokens": 531476.0,
|
|
"step": 170
|
|
},
|
|
{
|
|
"entropy": 0.7522185202687979,
|
|
"epoch": 0.576,
|
|
"grad_norm": 0.55078125,
|
|
"learning_rate": 0.00016762788866728,
|
|
"loss": 0.7073,
|
|
"mean_token_accuracy": 0.8059838317334652,
|
|
"num_tokens": 564807.0,
|
|
"step": 180
|
|
},
|
|
{
|
|
"entropy": 0.7345199611037969,
|
|
"epoch": 0.608,
|
|
"grad_norm": 0.62890625,
|
|
"learning_rate": 0.0001637264461181213,
|
|
"loss": 0.6915,
|
|
"mean_token_accuracy": 0.80965431407094,
|
|
"num_tokens": 597440.0,
|
|
"step": 190
|
|
},
|
|
{
|
|
"entropy": 0.8049266938120127,
|
|
"epoch": 0.64,
|
|
"grad_norm": 0.486328125,
|
|
"learning_rate": 0.0001596543383133303,
|
|
"loss": 0.7237,
|
|
"mean_token_accuracy": 0.7935581188648939,
|
|
"num_tokens": 630050.0,
|
|
"step": 200
|
|
},
|
|
{
|
|
"entropy": 0.6652943708002568,
|
|
"epoch": 0.672,
|
|
"grad_norm": 0.52734375,
|
|
"learning_rate": 0.00015542247073033116,
|
|
"loss": 0.6553,
|
|
"mean_token_accuracy": 0.8136730916798115,
|
|
"num_tokens": 660867.0,
|
|
"step": 210
|
|
},
|
|
{
|
|
"entropy": 0.6704742344096303,
|
|
"epoch": 0.704,
|
|
"grad_norm": 0.55078125,
|
|
"learning_rate": 0.00015104217669785713,
|
|
"loss": 0.6572,
|
|
"mean_token_accuracy": 0.8189906783401966,
|
|
"num_tokens": 694781.0,
|
|
"step": 220
|
|
},
|
|
{
|
|
"entropy": 0.8146899126470089,
|
|
"epoch": 0.736,
|
|
"grad_norm": 0.6015625,
|
|
"learning_rate": 0.0001465251870442586,
|
|
"loss": 0.7881,
|
|
"mean_token_accuracy": 0.7911321885883809,
|
|
"num_tokens": 727207.0,
|
|
"step": 230
|
|
},
|
|
{
|
|
"entropy": 0.6836955199018121,
|
|
"epoch": 0.768,
|
|
"grad_norm": 0.53515625,
|
|
"learning_rate": 0.00014188359868127134,
|
|
"loss": 0.6743,
|
|
"mean_token_accuracy": 0.8147166304290294,
|
|
"num_tokens": 760836.0,
|
|
"step": 240
|
|
},
|
|
{
|
|
"entropy": 0.8043613195419311,
|
|
"epoch": 0.8,
|
|
"grad_norm": 0.48828125,
|
|
"learning_rate": 0.00013712984220737938,
|
|
"loss": 0.7463,
|
|
"mean_token_accuracy": 0.7957966670393943,
|
|
"num_tokens": 794412.0,
|
|
"step": 250
|
|
},
|
|
{
|
|
"entropy": 0.6497382273897528,
|
|
"epoch": 0.832,
|
|
"grad_norm": 0.5703125,
|
|
"learning_rate": 0.00013227664861753454,
|
|
"loss": 0.6431,
|
|
"mean_token_accuracy": 0.8244783975183964,
|
|
"num_tokens": 827099.0,
|
|
"step": 260
|
|
},
|
|
{
|
|
"entropy": 0.8308647315949201,
|
|
"epoch": 0.864,
|
|
"grad_norm": 0.609375,
|
|
"learning_rate": 0.00012733701520838681,
|
|
"loss": 0.7665,
|
|
"mean_token_accuracy": 0.7912245698273181,
|
|
"num_tokens": 858402.0,
|
|
"step": 270
|
|
},
|
|
{
|
|
"entropy": 0.6155038874596357,
|
|
"epoch": 0.896,
|
|
"grad_norm": 0.54296875,
|
|
"learning_rate": 0.00012232417077033414,
|
|
"loss": 0.6085,
|
|
"mean_token_accuracy": 0.8311799973249435,
|
|
"num_tokens": 887820.0,
|
|
"step": 280
|
|
},
|
|
{
|
|
"entropy": 0.7490381442010403,
|
|
"epoch": 0.928,
|
|
"grad_norm": 0.4921875,
|
|
"learning_rate": 0.00011725154015961088,
|
|
"loss": 0.6872,
|
|
"mean_token_accuracy": 0.8019387792795897,
|
|
"num_tokens": 918679.0,
|
|
"step": 290
|
|
},
|
|
{
|
|
"entropy": 0.7262276472523809,
|
|
"epoch": 0.96,
|
|
"grad_norm": 0.458984375,
|
|
"learning_rate": 0.0001121327083452942,
|
|
"loss": 0.7096,
|
|
"mean_token_accuracy": 0.8102037977427244,
|
|
"num_tokens": 951444.0,
|
|
"step": 300
|
|
},
|
|
{
|
|
"entropy": 0.6726594571024179,
|
|
"epoch": 0.992,
|
|
"grad_norm": 0.5625,
|
|
"learning_rate": 0.00010698138402751314,
|
|
"loss": 0.6115,
|
|
"mean_token_accuracy": 0.8189524404704571,
|
|
"num_tokens": 983476.0,
|
|
"step": 310
|
|
}
|
|
],
|
|
"logging_steps": 10,
|
|
"max_steps": 626,
|
|
"num_input_tokens_seen": 0,
|
|
"num_train_epochs": 2,
|
|
"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": 2.1729056081918976e+16,
|
|
"train_batch_size": 1,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|