227 lines
5.9 KiB
JSON
227 lines
5.9 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 0.4387082953997751,
|
|
"eval_steps": 99999999,
|
|
"global_step": 2048,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"epoch": 0.00021421303486317142,
|
|
"grad_norm": 174.29400634765625,
|
|
"learning_rate": 5e-05,
|
|
"loss": 175.3938,
|
|
"step": 1
|
|
},
|
|
{
|
|
"epoch": 0.00021421303486317142,
|
|
"eval_accuracy": 0.024215843286987115,
|
|
"eval_loss": 10.354011535644531,
|
|
"eval_runtime": 152.0177,
|
|
"eval_samples_per_second": 52.06,
|
|
"eval_steps_per_second": 6.512,
|
|
"step": 1
|
|
},
|
|
{
|
|
"epoch": 0.00042842606972634284,
|
|
"grad_norm": 83.63705444335938,
|
|
"learning_rate": 4.998929106875134e-05,
|
|
"loss": 164.501,
|
|
"step": 2
|
|
},
|
|
{
|
|
"epoch": 0.00042842606972634284,
|
|
"eval_accuracy": 0.024217696048750854,
|
|
"eval_loss": 10.13454532623291,
|
|
"eval_runtime": 137.1006,
|
|
"eval_samples_per_second": 57.724,
|
|
"eval_steps_per_second": 7.221,
|
|
"step": 2
|
|
},
|
|
{
|
|
"epoch": 0.0008568521394526857,
|
|
"grad_norm": 43.825008392333984,
|
|
"learning_rate": 4.996787320625402e-05,
|
|
"loss": 158.761,
|
|
"step": 4
|
|
},
|
|
{
|
|
"epoch": 0.0008568521394526857,
|
|
"eval_accuracy": 0.024217696048750854,
|
|
"eval_loss": 9.977794647216797,
|
|
"eval_runtime": 136.1039,
|
|
"eval_samples_per_second": 58.147,
|
|
"eval_steps_per_second": 7.274,
|
|
"step": 4
|
|
},
|
|
{
|
|
"epoch": 0.0017137042789053714,
|
|
"grad_norm": 35.26683044433594,
|
|
"learning_rate": 4.9925037481259376e-05,
|
|
"loss": 155.013,
|
|
"step": 8
|
|
},
|
|
{
|
|
"epoch": 0.0017137042789053714,
|
|
"eval_accuracy": 0.054267762612305155,
|
|
"eval_loss": 9.693761825561523,
|
|
"eval_runtime": 137.244,
|
|
"eval_samples_per_second": 57.664,
|
|
"eval_steps_per_second": 7.213,
|
|
"step": 8
|
|
},
|
|
{
|
|
"epoch": 0.0034274085578107428,
|
|
"grad_norm": 40.224300384521484,
|
|
"learning_rate": 4.983936603127008e-05,
|
|
"loss": 149.2168,
|
|
"step": 16
|
|
},
|
|
{
|
|
"epoch": 0.0034274085578107428,
|
|
"eval_accuracy": 0.016920902635887105,
|
|
"eval_loss": 11.783411026000977,
|
|
"eval_runtime": 146.7615,
|
|
"eval_samples_per_second": 53.924,
|
|
"eval_steps_per_second": 6.746,
|
|
"step": 16
|
|
},
|
|
{
|
|
"epoch": 0.0068548171156214855,
|
|
"grad_norm": 37.21358871459961,
|
|
"learning_rate": 4.96680231312915e-05,
|
|
"loss": 145.3381,
|
|
"step": 32
|
|
},
|
|
{
|
|
"epoch": 0.0068548171156214855,
|
|
"eval_accuracy": 0.0519994881436834,
|
|
"eval_loss": 8.588780403137207,
|
|
"eval_runtime": 142.279,
|
|
"eval_samples_per_second": 55.623,
|
|
"eval_steps_per_second": 6.958,
|
|
"step": 32
|
|
},
|
|
{
|
|
"epoch": 0.013709634231242971,
|
|
"grad_norm": 23.38722038269043,
|
|
"learning_rate": 4.932533733133434e-05,
|
|
"loss": 115.3768,
|
|
"step": 64
|
|
},
|
|
{
|
|
"epoch": 0.013709634231242971,
|
|
"eval_accuracy": 0.07924990816477524,
|
|
"eval_loss": 7.433769702911377,
|
|
"eval_runtime": 147.5767,
|
|
"eval_samples_per_second": 53.626,
|
|
"eval_steps_per_second": 6.708,
|
|
"step": 64
|
|
},
|
|
{
|
|
"epoch": 0.027419268462485942,
|
|
"grad_norm": 18.296302795410156,
|
|
"learning_rate": 4.8639965731420003e-05,
|
|
"loss": 92.7634,
|
|
"step": 128
|
|
},
|
|
{
|
|
"epoch": 0.027419268462485942,
|
|
"eval_accuracy": 0.09311190112872716,
|
|
"eval_loss": 6.83799409866333,
|
|
"eval_runtime": 143.065,
|
|
"eval_samples_per_second": 55.318,
|
|
"eval_steps_per_second": 6.92,
|
|
"step": 128
|
|
},
|
|
{
|
|
"epoch": 0.054838536924971884,
|
|
"grad_norm": 12.443135261535645,
|
|
"learning_rate": 4.726922253159135e-05,
|
|
"loss": 74.9485,
|
|
"step": 256
|
|
},
|
|
{
|
|
"epoch": 0.054838536924971884,
|
|
"eval_accuracy": 0.14537732234423276,
|
|
"eval_loss": 6.180610656738281,
|
|
"eval_runtime": 139.6377,
|
|
"eval_samples_per_second": 56.675,
|
|
"eval_steps_per_second": 7.09,
|
|
"step": 256
|
|
},
|
|
{
|
|
"epoch": 0.10967707384994377,
|
|
"grad_norm": 10.523282051086426,
|
|
"learning_rate": 4.452773613193403e-05,
|
|
"loss": 66.064,
|
|
"step": 512
|
|
},
|
|
{
|
|
"epoch": 0.10967707384994377,
|
|
"eval_accuracy": 0.14975638653155834,
|
|
"eval_loss": 6.066355228424072,
|
|
"eval_runtime": 143.9305,
|
|
"eval_samples_per_second": 54.985,
|
|
"eval_steps_per_second": 6.878,
|
|
"step": 512
|
|
},
|
|
{
|
|
"epoch": 0.21935414769988754,
|
|
"grad_norm": 11.639715194702148,
|
|
"learning_rate": 3.9044763332619404e-05,
|
|
"loss": 61.2244,
|
|
"step": 1024
|
|
},
|
|
{
|
|
"epoch": 0.21935414769988754,
|
|
"eval_accuracy": 0.1782677962090518,
|
|
"eval_loss": 5.892604351043701,
|
|
"eval_runtime": 158.4884,
|
|
"eval_samples_per_second": 49.934,
|
|
"eval_steps_per_second": 6.247,
|
|
"step": 1024
|
|
},
|
|
{
|
|
"epoch": 0.4387082953997751,
|
|
"grad_norm": 19.752676010131836,
|
|
"learning_rate": 2.807881773399015e-05,
|
|
"loss": 60.9252,
|
|
"step": 2048
|
|
},
|
|
{
|
|
"epoch": 0.4387082953997751,
|
|
"eval_accuracy": 0.25111369509618425,
|
|
"eval_loss": 4.6798505783081055,
|
|
"eval_runtime": 162.0054,
|
|
"eval_samples_per_second": 48.85,
|
|
"eval_steps_per_second": 6.111,
|
|
"step": 2048
|
|
}
|
|
],
|
|
"logging_steps": 99999999,
|
|
"max_steps": 4669,
|
|
"num_input_tokens_seen": 0,
|
|
"num_train_epochs": 1,
|
|
"save_steps": 99999999,
|
|
"stateful_callbacks": {
|
|
"TrainerControl": {
|
|
"args": {
|
|
"should_epoch_stop": false,
|
|
"should_evaluate": false,
|
|
"should_log": false,
|
|
"should_save": true,
|
|
"should_training_stop": false
|
|
},
|
|
"attributes": {}
|
|
}
|
|
},
|
|
"total_flos": 1.36992276873216e+17,
|
|
"train_batch_size": 8,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|