trainer_state.json

{
  "best_metric": 0.8848,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0499/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": 6.735155628520579e-05,
      "loss": 2.0152,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.76,
      "eval_loss": 0.8642678260803223,
      "eval_runtime": 7.1827,
      "eval_samples_per_second": 522.091,
      "eval_steps_per_second": 8.214,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 5.9807039661419336e-05,
      "loss": 0.5981,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8362666666666667,
      "eval_loss": 0.5373780131340027,
      "eval_runtime": 7.4543,
      "eval_samples_per_second": 503.064,
      "eval_steps_per_second": 7.915,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 9.252326011657715,
      "learning_rate": 4.8470148469326895e-05,
      "loss": 0.3269,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8570666666666666,
      "eval_loss": 0.47396984696388245,
      "eval_runtime": 7.0996,
      "eval_samples_per_second": 528.202,
      "eval_steps_per_second": 8.31,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 4.118952751159668,
      "learning_rate": 3.508254927997938e-05,
      "loss": 0.1829,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8645333333333334,
      "eval_loss": 0.45340394973754883,
      "eval_runtime": 7.1518,
      "eval_samples_per_second": 524.34,
      "eval_steps_per_second": 8.25,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 46.794593811035156,
      "learning_rate": 2.1682382711066092e-05,
      "loss": 0.0999,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8717333333333334,
      "eval_loss": 0.4455307424068451,
      "eval_runtime": 7.0204,
      "eval_samples_per_second": 534.159,
      "eval_steps_per_second": 8.404,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 2.2327558994293213,
      "learning_rate": 1.0309702649891636e-05,
      "loss": 0.0474,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8832,
      "eval_loss": 0.4144890606403351,
      "eval_runtime": 7.0328,
      "eval_samples_per_second": 533.215,
      "eval_steps_per_second": 8.389,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.2401283085346222,
      "learning_rate": 2.6958965421683553e-06,
      "loss": 0.0229,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8848,
      "eval_loss": 0.4172823429107666,
      "eval_runtime": 7.0987,
      "eval_samples_per_second": 528.264,
      "eval_steps_per_second": 8.311,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2664,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "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": 1.153189244358144e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}