Model-J__DINO__model_idx_0604 / trainer_state.json
Eliahu's picture
Upload folder using huggingface_hub
401a982 verified
Invalid JSON: Unexpected token 'I', ..."ad_norm": Infinity, "... is not valid JSON
{
"best_metric": 0.8744,
"best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0604/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": 17.982086181640625,
"learning_rate": 7e-05,
"loss": 0.8892,
"step": 333
},
{
"epoch": 1.0,
"eval_accuracy": 0.8416,
"eval_loss": 0.566001296043396,
"eval_runtime": 13.9974,
"eval_samples_per_second": 267.907,
"eval_steps_per_second": 4.215,
"step": 333
},
{
"epoch": 2.0,
"grad_norm": Infinity,
"learning_rate": 7e-05,
"loss": 0.3266,
"step": 666
},
{
"epoch": 2.0,
"eval_accuracy": 0.8434666666666667,
"eval_loss": 0.5411359667778015,
"eval_runtime": 14.3391,
"eval_samples_per_second": 261.522,
"eval_steps_per_second": 4.115,
"step": 666
},
{
"epoch": 3.0,
"grad_norm": 153.07705688476562,
"learning_rate": 7e-05,
"loss": 0.2033,
"step": 999
},
{
"epoch": 3.0,
"eval_accuracy": 0.8637333333333334,
"eval_loss": 0.4805768132209778,
"eval_runtime": 15.032,
"eval_samples_per_second": 249.467,
"eval_steps_per_second": 3.925,
"step": 999
},
{
"epoch": 4.0,
"grad_norm": 0.011496650986373425,
"learning_rate": 7e-05,
"loss": 0.1435,
"step": 1332
},
{
"epoch": 4.0,
"eval_accuracy": 0.8744,
"eval_loss": 0.4916727840900421,
"eval_runtime": 13.2436,
"eval_samples_per_second": 283.156,
"eval_steps_per_second": 4.455,
"step": 1332
}
],
"logging_steps": 500,
"max_steps": 1998,
"num_input_tokens_seen": 0,
"num_train_epochs": 6,
"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": 6.58965282490368e+18,
"train_batch_size": 64,
"trial_name": null,
"trial_params": null
}