gpt2-RMT-2 / trainer_state.json
KotshinZ's picture
Model save
eef26b5 verified
raw
history blame
2.16 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.005057992500218017,
"eval_steps": 100,
"global_step": 58,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0008720676724513823,
"grad_norm": 9.03160572052002,
"learning_rate": 1.9709418174260523e-05,
"loss": 3.068,
"mean_token_accuracy": 0.44124266058206557,
"step": 10
},
{
"epoch": 0.0017441353449027645,
"grad_norm": 15.666545867919922,
"learning_rate": 1.6631226582407954e-05,
"loss": 2.9078,
"mean_token_accuracy": 0.4430650696158409,
"step": 20
},
{
"epoch": 0.002616203017354147,
"grad_norm": 16.191038131713867,
"learning_rate": 1.1205366802553231e-05,
"loss": 2.8492,
"mean_token_accuracy": 0.45182240307331084,
"step": 30
},
{
"epoch": 0.003488270689805529,
"grad_norm": 10.86160945892334,
"learning_rate": 5.352768279562315e-06,
"loss": 2.7617,
"mean_token_accuracy": 0.4569960877299309,
"step": 40
},
{
"epoch": 0.004360338362256911,
"grad_norm": 14.749295234680176,
"learning_rate": 1.1454397434679022e-06,
"loss": 2.6656,
"mean_token_accuracy": 0.47067025154829023,
"step": 50
},
{
"epoch": 0.005057992500218017,
"mean_token_accuracy": 0.45992845110595226,
"step": 58,
"total_flos": 0.0,
"train_loss": 2.8374191810344827,
"train_runtime": 8.8944,
"train_samples_per_second": 103.136,
"train_steps_per_second": 6.521
}
],
"logging_steps": 10,
"max_steps": 58,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": false,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 0.0,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}