Tech-oriented's picture
Training in progress, epoch 4
0ac710f verified
raw
history blame
2.04 kB
{
"best_metric": 0.8371559633027523,
"best_model_checkpoint": "Tech-oriented/best_model_bert_uncasedbert-base-uncased-finetuned-sst2/run-2/checkpoint-8420",
"epoch": 4.0,
"eval_steps": 500,
"global_step": 8420,
"is_hyper_param_search": true,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 1.0,
"eval_accuracy": 0.5745412844036697,
"eval_loss": 0.658979058265686,
"eval_runtime": 2.4031,
"eval_samples_per_second": 362.862,
"eval_steps_per_second": 22.887,
"step": 2105
},
{
"epoch": 2.0,
"grad_norm": 11.47251033782959,
"learning_rate": 4.003340487387436e-07,
"loss": 0.6604,
"step": 4209
},
{
"epoch": 2.0,
"eval_accuracy": 0.7350917431192661,
"eval_loss": 0.5269513726234436,
"eval_runtime": 2.4016,
"eval_samples_per_second": 363.093,
"eval_steps_per_second": 22.901,
"step": 4210
},
{
"epoch": 3.0,
"eval_accuracy": 0.819954128440367,
"eval_loss": 0.4339826703071594,
"eval_runtime": 2.4278,
"eval_samples_per_second": 359.172,
"eval_steps_per_second": 22.654,
"step": 6315
},
{
"epoch": 4.0,
"grad_norm": 15.470611572265625,
"learning_rate": 1.9013728270659872e-10,
"loss": 0.5279,
"step": 8418
},
{
"epoch": 4.0,
"eval_accuracy": 0.8371559633027523,
"eval_loss": 0.4121007025241852,
"eval_runtime": 2.4148,
"eval_samples_per_second": 361.103,
"eval_steps_per_second": 22.776,
"step": 8420
}
],
"logging_steps": 4209,
"max_steps": 8420,
"num_input_tokens_seen": 0,
"num_train_epochs": 4,
"save_steps": 500,
"total_flos": 5552011212462480.0,
"train_batch_size": 32,
"trial_name": null,
"trial_params": {
"learning_rate": 8.004779601947806e-07,
"num_train_epochs": 4,
"per_device_train_batch_size": 32,
"seed": 37,
"weight_decay": 0.24080188545658526
}
}