204 lines
5.1 KiB
JSON
204 lines
5.1 KiB
JSON
{
|
|
"best_metric": 0.8921161825726142,
|
|
"best_model_checkpoint": "UL_base_classification/checkpoint-123",
|
|
"epoch": 6.829268292682927,
|
|
"eval_steps": 500,
|
|
"global_step": 140,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"epoch": 0.4878048780487805,
|
|
"grad_norm": 4.009415149688721,
|
|
"learning_rate": 3.571428571428572e-05,
|
|
"loss": 1.4657,
|
|
"step": 10
|
|
},
|
|
{
|
|
"epoch": 0.975609756097561,
|
|
"grad_norm": 1.9595946073532104,
|
|
"learning_rate": 4.761904761904762e-05,
|
|
"loss": 0.8296,
|
|
"step": 20
|
|
},
|
|
{
|
|
"epoch": 0.975609756097561,
|
|
"eval_accuracy": 0.8229598893499308,
|
|
"eval_loss": 0.5682700872421265,
|
|
"eval_runtime": 28.9624,
|
|
"eval_samples_per_second": 24.963,
|
|
"eval_steps_per_second": 0.794,
|
|
"step": 20
|
|
},
|
|
{
|
|
"epoch": 1.4634146341463414,
|
|
"grad_norm": 1.4578920602798462,
|
|
"learning_rate": 4.3650793650793655e-05,
|
|
"loss": 0.5174,
|
|
"step": 30
|
|
},
|
|
{
|
|
"epoch": 1.951219512195122,
|
|
"grad_norm": 1.6582454442977905,
|
|
"learning_rate": 3.968253968253968e-05,
|
|
"loss": 0.4462,
|
|
"step": 40
|
|
},
|
|
{
|
|
"epoch": 2.0,
|
|
"eval_accuracy": 0.8603042876901799,
|
|
"eval_loss": 0.3948992192745209,
|
|
"eval_runtime": 28.607,
|
|
"eval_samples_per_second": 25.274,
|
|
"eval_steps_per_second": 0.804,
|
|
"step": 41
|
|
},
|
|
{
|
|
"epoch": 2.4390243902439024,
|
|
"grad_norm": 1.9573261737823486,
|
|
"learning_rate": 3.571428571428572e-05,
|
|
"loss": 0.3861,
|
|
"step": 50
|
|
},
|
|
{
|
|
"epoch": 2.926829268292683,
|
|
"grad_norm": 2.927914619445801,
|
|
"learning_rate": 3.1746031746031745e-05,
|
|
"loss": 0.3588,
|
|
"step": 60
|
|
},
|
|
{
|
|
"epoch": 2.975609756097561,
|
|
"eval_accuracy": 0.8575380359612724,
|
|
"eval_loss": 0.3632887601852417,
|
|
"eval_runtime": 27.4966,
|
|
"eval_samples_per_second": 26.294,
|
|
"eval_steps_per_second": 0.836,
|
|
"step": 61
|
|
},
|
|
{
|
|
"epoch": 3.4146341463414633,
|
|
"grad_norm": 2.270343065261841,
|
|
"learning_rate": 2.777777777777778e-05,
|
|
"loss": 0.2995,
|
|
"step": 70
|
|
},
|
|
{
|
|
"epoch": 3.902439024390244,
|
|
"grad_norm": 2.323387622833252,
|
|
"learning_rate": 2.380952380952381e-05,
|
|
"loss": 0.3196,
|
|
"step": 80
|
|
},
|
|
{
|
|
"epoch": 4.0,
|
|
"eval_accuracy": 0.8852005532503457,
|
|
"eval_loss": 0.3247418701648712,
|
|
"eval_runtime": 28.061,
|
|
"eval_samples_per_second": 25.765,
|
|
"eval_steps_per_second": 0.82,
|
|
"step": 82
|
|
},
|
|
{
|
|
"epoch": 4.390243902439025,
|
|
"grad_norm": 1.6208417415618896,
|
|
"learning_rate": 1.984126984126984e-05,
|
|
"loss": 0.2878,
|
|
"step": 90
|
|
},
|
|
{
|
|
"epoch": 4.878048780487805,
|
|
"grad_norm": 1.3482733964920044,
|
|
"learning_rate": 1.5873015873015872e-05,
|
|
"loss": 0.2921,
|
|
"step": 100
|
|
},
|
|
{
|
|
"epoch": 4.975609756097561,
|
|
"eval_accuracy": 0.8727524204702628,
|
|
"eval_loss": 0.3373877704143524,
|
|
"eval_runtime": 28.0046,
|
|
"eval_samples_per_second": 25.817,
|
|
"eval_steps_per_second": 0.821,
|
|
"step": 102
|
|
},
|
|
{
|
|
"epoch": 5.365853658536586,
|
|
"grad_norm": 1.1717610359191895,
|
|
"learning_rate": 1.1904761904761905e-05,
|
|
"loss": 0.2319,
|
|
"step": 110
|
|
},
|
|
{
|
|
"epoch": 5.853658536585366,
|
|
"grad_norm": 1.3873180150985718,
|
|
"learning_rate": 7.936507936507936e-06,
|
|
"loss": 0.2688,
|
|
"step": 120
|
|
},
|
|
{
|
|
"epoch": 6.0,
|
|
"eval_accuracy": 0.8921161825726142,
|
|
"eval_loss": 0.3124912977218628,
|
|
"eval_runtime": 28.2402,
|
|
"eval_samples_per_second": 25.602,
|
|
"eval_steps_per_second": 0.814,
|
|
"step": 123
|
|
},
|
|
{
|
|
"epoch": 6.341463414634147,
|
|
"grad_norm": 1.7034335136413574,
|
|
"learning_rate": 3.968253968253968e-06,
|
|
"loss": 0.2583,
|
|
"step": 130
|
|
},
|
|
{
|
|
"epoch": 6.829268292682927,
|
|
"grad_norm": 1.5103542804718018,
|
|
"learning_rate": 0.0,
|
|
"loss": 0.2366,
|
|
"step": 140
|
|
},
|
|
{
|
|
"epoch": 6.829268292682927,
|
|
"eval_accuracy": 0.8865836791147994,
|
|
"eval_loss": 0.3137405514717102,
|
|
"eval_runtime": 28.632,
|
|
"eval_samples_per_second": 25.251,
|
|
"eval_steps_per_second": 0.803,
|
|
"step": 140
|
|
},
|
|
{
|
|
"epoch": 6.829268292682927,
|
|
"step": 140,
|
|
"total_flos": 1.3849740766444585e+18,
|
|
"train_loss": 0.4427596058164324,
|
|
"train_runtime": 1849.9588,
|
|
"train_samples_per_second": 9.899,
|
|
"train_steps_per_second": 0.076
|
|
}
|
|
],
|
|
"logging_steps": 10,
|
|
"max_steps": 140,
|
|
"num_input_tokens_seen": 0,
|
|
"num_train_epochs": 7,
|
|
"save_steps": 500,
|
|
"stateful_callbacks": {
|
|
"TrainerControl": {
|
|
"args": {
|
|
"should_epoch_stop": false,
|
|
"should_evaluate": false,
|
|
"should_log": false,
|
|
"should_save": true,
|
|
"should_training_stop": true
|
|
},
|
|
"attributes": {}
|
|
}
|
|
},
|
|
"total_flos": 1.3849740766444585e+18,
|
|
"train_batch_size": 32,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|