trainer_state.json

{
  "best_metric": 0.45653333333333335,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0028/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": 3.2817,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.19813333333333333,
      "eval_loss": 2.9370532035827637,
      "eval_runtime": 6.7392,
      "eval_samples_per_second": 556.448,
      "eval_steps_per_second": 8.755,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0005,
      "loss": 2.777,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.2824,
      "eval_loss": 2.5935006141662598,
      "eval_runtime": 7.0312,
      "eval_samples_per_second": 533.335,
      "eval_steps_per_second": 8.391,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 10.717896461486816,
      "learning_rate": 0.0005,
      "loss": 2.4666,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.3389333333333333,
      "eval_loss": 2.358398199081421,
      "eval_runtime": 6.6548,
      "eval_samples_per_second": 563.502,
      "eval_steps_per_second": 8.866,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 12.27419662475586,
      "learning_rate": 0.0005,
      "loss": 2.163,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.3744,
      "eval_loss": 2.1955957412719727,
      "eval_runtime": 6.7688,
      "eval_samples_per_second": 554.012,
      "eval_steps_per_second": 8.716,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 7.483421325683594,
      "learning_rate": 0.0005,
      "loss": 1.9388,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.4088,
      "eval_loss": 2.0623438358306885,
      "eval_runtime": 6.806,
      "eval_samples_per_second": 550.983,
      "eval_steps_per_second": 8.669,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 14.11381721496582,
      "learning_rate": 0.0005,
      "loss": 1.7376,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.41173333333333334,
      "eval_loss": 2.0828559398651123,
      "eval_runtime": 6.5952,
      "eval_samples_per_second": 568.595,
      "eval_steps_per_second": 8.946,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 15.765301704406738,
      "learning_rate": 0.0005,
      "loss": 1.5331,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.45653333333333335,
      "eval_loss": 1.9593614339828491,
      "eval_runtime": 6.7215,
      "eval_samples_per_second": 557.915,
      "eval_steps_per_second": 8.778,
      "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
}