Model-J__ResNet__model_idx_0394 / trainer_state.json
Eliahu's picture
Upload folder using huggingface_hub
2b2a5e7 verified
{
"best_metric": 0.8565333333333334,
"best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0394/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": 119.03367614746094,
"learning_rate": 9.045084971874738e-05,
"loss": 2.9136,
"step": 333
},
{
"epoch": 1.0,
"eval_accuracy": 0.6706666666666666,
"eval_loss": 1.2413145303726196,
"eval_runtime": 7.6226,
"eval_samples_per_second": 491.959,
"eval_steps_per_second": 7.74,
"step": 333
},
{
"epoch": 2.0,
"grad_norm": 154.16531372070312,
"learning_rate": 6.545084971874738e-05,
"loss": 0.8892,
"step": 666
},
{
"epoch": 2.0,
"eval_accuracy": 0.8077333333333333,
"eval_loss": 0.6499355435371399,
"eval_runtime": 7.9441,
"eval_samples_per_second": 472.05,
"eval_steps_per_second": 7.427,
"step": 666
},
{
"epoch": 3.0,
"grad_norm": 83.17750549316406,
"learning_rate": 3.4549150281252636e-05,
"loss": 0.5599,
"step": 999
},
{
"epoch": 3.0,
"eval_accuracy": 0.8517333333333333,
"eval_loss": 0.5091395974159241,
"eval_runtime": 7.5813,
"eval_samples_per_second": 494.637,
"eval_steps_per_second": 7.782,
"step": 999
},
{
"epoch": 4.0,
"grad_norm": 65.25487518310547,
"learning_rate": 9.549150281252633e-06,
"loss": 0.4549,
"step": 1332
},
{
"epoch": 4.0,
"eval_accuracy": 0.8565333333333334,
"eval_loss": 0.47455883026123047,
"eval_runtime": 7.8877,
"eval_samples_per_second": 475.423,
"eval_steps_per_second": 7.48,
"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
}