vit-base-patch16-224-MSC-dmae / trainer_state.json
Augusto777's picture
End of training
51704cd
{
"best_metric": 0.95,
"best_model_checkpoint": "vit-base-patch16-224-MSC-dmae/checkpoint-7",
"epoch": 8.0,
"eval_steps": 500,
"global_step": 12,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.67,
"eval_accuracy": 0.5,
"eval_loss": 1.2257916927337646,
"eval_runtime": 0.3201,
"eval_samples_per_second": 62.477,
"eval_steps_per_second": 3.124,
"step": 1
},
{
"epoch": 2.0,
"eval_accuracy": 0.7,
"eval_loss": 1.0536372661590576,
"eval_runtime": 0.3238,
"eval_samples_per_second": 61.768,
"eval_steps_per_second": 3.088,
"step": 3
},
{
"epoch": 2.67,
"eval_accuracy": 0.75,
"eval_loss": 0.9143384099006653,
"eval_runtime": 0.3026,
"eval_samples_per_second": 66.089,
"eval_steps_per_second": 3.304,
"step": 4
},
{
"epoch": 4.0,
"eval_accuracy": 0.9,
"eval_loss": 0.6899412870407104,
"eval_runtime": 0.3297,
"eval_samples_per_second": 60.655,
"eval_steps_per_second": 3.033,
"step": 6
},
{
"epoch": 4.67,
"eval_accuracy": 0.95,
"eval_loss": 0.6299833059310913,
"eval_runtime": 0.286,
"eval_samples_per_second": 69.919,
"eval_steps_per_second": 3.496,
"step": 7
},
{
"epoch": 6.0,
"eval_accuracy": 0.9,
"eval_loss": 0.5069305896759033,
"eval_runtime": 0.3181,
"eval_samples_per_second": 62.873,
"eval_steps_per_second": 3.144,
"step": 9
},
{
"epoch": 6.67,
"learning_rate": 1e-05,
"loss": 0.8554,
"step": 10
},
{
"epoch": 6.67,
"eval_accuracy": 0.9,
"eval_loss": 0.46708640456199646,
"eval_runtime": 0.317,
"eval_samples_per_second": 63.085,
"eval_steps_per_second": 3.154,
"step": 10
},
{
"epoch": 8.0,
"eval_accuracy": 0.9,
"eval_loss": 0.43121227622032166,
"eval_runtime": 0.3145,
"eval_samples_per_second": 63.594,
"eval_steps_per_second": 3.18,
"step": 12
},
{
"epoch": 8.0,
"step": 12,
"total_flos": 1.1159046530924544e+17,
"train_loss": 0.795779287815094,
"train_runtime": 107.6532,
"train_samples_per_second": 20.064,
"train_steps_per_second": 0.111
}
],
"logging_steps": 10,
"max_steps": 12,
"num_train_epochs": 12,
"save_steps": 500,
"total_flos": 1.1159046530924544e+17,
"trial_name": null,
"trial_params": null
}