Model-J__ResNet__model_idx_0092 / trainer_state.json
Eliahu's picture
Upload folder using huggingface_hub
6528fe6 verified
{
"best_metric": 0.8482666666666666,
"best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0092/checkpoints/checkpoint-1332",
"epoch": 4.0,
"eval_steps": 500,
"global_step": 1332,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 1.0,
"grad_norm": 123.62455749511719,
"learning_rate": 8.140576474687264e-05,
"loss": 3.0021,
"step": 333
},
{
"epoch": 1.0,
"eval_accuracy": 0.6749333333333334,
"eval_loss": 1.3384606838226318,
"eval_runtime": 7.9503,
"eval_samples_per_second": 471.682,
"eval_steps_per_second": 7.421,
"step": 333
},
{
"epoch": 2.0,
"grad_norm": 84.5896224975586,
"learning_rate": 5.890576474687264e-05,
"loss": 0.9116,
"step": 666
},
{
"epoch": 2.0,
"eval_accuracy": 0.8050666666666667,
"eval_loss": 0.6513868570327759,
"eval_runtime": 7.8864,
"eval_samples_per_second": 475.501,
"eval_steps_per_second": 7.481,
"step": 666
},
{
"epoch": 3.0,
"grad_norm": 123.14508056640625,
"learning_rate": 3.1094235253127374e-05,
"loss": 0.5786,
"step": 999
},
{
"epoch": 3.0,
"eval_accuracy": 0.8341333333333333,
"eval_loss": 0.5302331447601318,
"eval_runtime": 7.9715,
"eval_samples_per_second": 470.425,
"eval_steps_per_second": 7.401,
"step": 999
},
{
"epoch": 4.0,
"grad_norm": 85.92764282226562,
"learning_rate": 8.59423525312737e-06,
"loss": 0.4812,
"step": 1332
},
{
"epoch": 4.0,
"eval_accuracy": 0.8482666666666666,
"eval_loss": 0.5042679309844971,
"eval_runtime": 7.9017,
"eval_samples_per_second": 474.581,
"eval_steps_per_second": 7.467,
"step": 1332
}
],
"logging_steps": 500,
"max_steps": 1665,
"num_input_tokens_seen": 0,
"num_train_epochs": 5,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 3.2705716958208e+18,
"train_batch_size": 64,
"trial_name": null,
"trial_params": null
}