Model-J__ResNet__model_idx_0282 / trainer_state.json
Eliahu's picture
Upload folder using huggingface_hub
4019ade verified
{
"best_metric": 0.9010666666666667,
"best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0282/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": 70.92640686035156,
"learning_rate": 0.0002713525491562421,
"loss": 1.5244,
"step": 333
},
{
"epoch": 1.0,
"eval_accuracy": 0.8362666666666667,
"eval_loss": 0.5233182311058044,
"eval_runtime": 16.4868,
"eval_samples_per_second": 227.454,
"eval_steps_per_second": 3.579,
"step": 333
},
{
"epoch": 2.0,
"grad_norm": 38.3804817199707,
"learning_rate": 0.0001963525491562421,
"loss": 0.3904,
"step": 666
},
{
"epoch": 2.0,
"eval_accuracy": 0.8845333333333333,
"eval_loss": 0.37630128860473633,
"eval_runtime": 15.7009,
"eval_samples_per_second": 238.839,
"eval_steps_per_second": 3.758,
"step": 666
},
{
"epoch": 3.0,
"grad_norm": 51.94040298461914,
"learning_rate": 0.0001036474508437579,
"loss": 0.194,
"step": 999
},
{
"epoch": 3.0,
"eval_accuracy": 0.8914666666666666,
"eval_loss": 0.3653945326805115,
"eval_runtime": 15.6906,
"eval_samples_per_second": 238.996,
"eval_steps_per_second": 3.76,
"step": 999
},
{
"epoch": 4.0,
"grad_norm": 58.74272918701172,
"learning_rate": 2.8647450843757897e-05,
"loss": 0.0983,
"step": 1332
},
{
"epoch": 4.0,
"eval_accuracy": 0.9010666666666667,
"eval_loss": 0.3486242890357971,
"eval_runtime": 15.4705,
"eval_samples_per_second": 242.397,
"eval_steps_per_second": 3.814,
"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
}