trainer_state.json

{
  "best_metric": 0.4938666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0726/checkpoints/checkpoint-2331",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 2331,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0005,
      "loss": 3.2196,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.21093333333333333,
      "eval_loss": 2.9221014976501465,
      "eval_runtime": 6.3888,
      "eval_samples_per_second": 586.966,
      "eval_steps_per_second": 9.235,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0005,
      "loss": 2.7002,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.3216,
      "eval_loss": 2.478163480758667,
      "eval_runtime": 6.3789,
      "eval_samples_per_second": 587.878,
      "eval_steps_per_second": 9.249,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 12.79973316192627,
      "learning_rate": 0.0005,
      "loss": 2.3229,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.3589333333333333,
      "eval_loss": 2.299178123474121,
      "eval_runtime": 6.3882,
      "eval_samples_per_second": 587.024,
      "eval_steps_per_second": 9.236,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0005,
      "loss": 2.0544,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.41173333333333334,
      "eval_loss": 2.135622024536133,
      "eval_runtime": 7.6772,
      "eval_samples_per_second": 488.46,
      "eval_steps_per_second": 7.685,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0005,
      "loss": 1.8109,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.44693333333333335,
      "eval_loss": 1.9749958515167236,
      "eval_runtime": 6.2126,
      "eval_samples_per_second": 603.613,
      "eval_steps_per_second": 9.497,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 1.9810874462127686,
      "learning_rate": 0.0005,
      "loss": 1.5804,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.46853333333333336,
      "eval_loss": 1.911770224571228,
      "eval_runtime": 6.7125,
      "eval_samples_per_second": 558.659,
      "eval_steps_per_second": 8.79,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 11.697165489196777,
      "learning_rate": 0.0005,
      "loss": 1.3854,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.4938666666666667,
      "eval_loss": 1.85065495967865,
      "eval_runtime": 6.1359,
      "eval_samples_per_second": 611.153,
      "eval_steps_per_second": 9.615,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2331,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.153189244358144e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}