trainer_state.json

{
  "best_metric": 0.6992,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0323/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": 0.0003,
      "loss": 3.048,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.30506666666666665,
      "eval_loss": 2.584726095199585,
      "eval_runtime": 12.4058,
      "eval_samples_per_second": 302.278,
      "eval_steps_per_second": 4.756,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0003,
      "loss": 2.1443,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.452,
      "eval_loss": 1.9340240955352783,
      "eval_runtime": 13.085,
      "eval_samples_per_second": 286.588,
      "eval_steps_per_second": 4.509,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0003,
      "loss": 1.4703,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6090666666666666,
      "eval_loss": 1.3267450332641602,
      "eval_runtime": 13.0522,
      "eval_samples_per_second": 287.309,
      "eval_steps_per_second": 4.52,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 12.949540138244629,
      "learning_rate": 0.0003,
      "loss": 0.9784,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6338666666666667,
      "eval_loss": 1.2399327754974365,
      "eval_runtime": 12.9872,
      "eval_samples_per_second": 288.745,
      "eval_steps_per_second": 4.543,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 8.240334510803223,
      "learning_rate": 0.0003,
      "loss": 0.6794,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.6754666666666667,
      "eval_loss": 1.081034779548645,
      "eval_runtime": 13.1717,
      "eval_samples_per_second": 284.702,
      "eval_steps_per_second": 4.479,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 13.333328247070312,
      "learning_rate": 0.0003,
      "loss": 0.4682,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.6861333333333334,
      "eval_loss": 1.0941795110702515,
      "eval_runtime": 13.0589,
      "eval_samples_per_second": 287.16,
      "eval_steps_per_second": 4.518,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 22.53818702697754,
      "learning_rate": 0.0003,
      "loss": 0.3342,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.6992,
      "eval_loss": 1.083463191986084,
      "eval_runtime": 13.1948,
      "eval_samples_per_second": 284.203,
      "eval_steps_per_second": 4.471,
      "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
}