trainer_state.json

{
  "best_metric": 0.8778666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0070/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": 3e-05,
      "loss": 2.1533,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7656,
      "eval_loss": 1.0234559774398804,
      "eval_runtime": 10.9045,
      "eval_samples_per_second": 343.896,
      "eval_steps_per_second": 5.411,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 3e-05,
      "loss": 0.7424,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8341333333333333,
      "eval_loss": 0.6377145051956177,
      "eval_runtime": 10.8855,
      "eval_samples_per_second": 344.495,
      "eval_steps_per_second": 5.42,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 3.001697063446045,
      "learning_rate": 3e-05,
      "loss": 0.4653,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8472,
      "eval_loss": 0.5499957203865051,
      "eval_runtime": 10.7847,
      "eval_samples_per_second": 347.716,
      "eval_steps_per_second": 5.471,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 37.45606231689453,
      "learning_rate": 3e-05,
      "loss": 0.3264,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8613333333333333,
      "eval_loss": 0.474644273519516,
      "eval_runtime": 10.7095,
      "eval_samples_per_second": 350.156,
      "eval_steps_per_second": 5.509,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 17.502492904663086,
      "learning_rate": 3e-05,
      "loss": 0.2458,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8709333333333333,
      "eval_loss": 0.4316202998161316,
      "eval_runtime": 10.6568,
      "eval_samples_per_second": 351.889,
      "eval_steps_per_second": 5.536,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.3035983741283417,
      "learning_rate": 3e-05,
      "loss": 0.18,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8752,
      "eval_loss": 0.4324495494365692,
      "eval_runtime": 11.2491,
      "eval_samples_per_second": 333.359,
      "eval_steps_per_second": 5.245,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.3797539472579956,
      "learning_rate": 3e-05,
      "loss": 0.1469,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8778666666666667,
      "eval_loss": 0.42924174666404724,
      "eval_runtime": 10.7996,
      "eval_samples_per_second": 347.236,
      "eval_steps_per_second": 5.463,
      "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
}