trainer_state.json

{
  "best_metric": 0.5797333333333333,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0923/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": 0.0004287859287859288,
      "loss": 3.1398,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.22933333333333333,
      "eval_loss": 2.795700788497925,
      "eval_runtime": 4.2251,
      "eval_samples_per_second": 887.56,
      "eval_steps_per_second": 13.964,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 0.00035757185757185754,
      "loss": 2.5053,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.3512,
      "eval_loss": 2.3214452266693115,
      "eval_runtime": 4.0839,
      "eval_samples_per_second": 918.246,
      "eval_steps_per_second": 14.447,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 15.874137878417969,
      "learning_rate": 0.00028614328614328614,
      "loss": 2.0043,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.432,
      "eval_loss": 1.9879359006881714,
      "eval_runtime": 4.2525,
      "eval_samples_per_second": 881.841,
      "eval_steps_per_second": 13.874,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 7.880004405975342,
      "learning_rate": 0.0002147147147147147,
      "loss": 1.5633,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.5061333333333333,
      "eval_loss": 1.733163595199585,
      "eval_runtime": 4.0912,
      "eval_samples_per_second": 916.603,
      "eval_steps_per_second": 14.421,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": Infinity,
      "learning_rate": 0.00014350064350064352,
      "loss": 1.1111,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.5442666666666667,
      "eval_loss": 1.6177316904067993,
      "eval_runtime": 4.0358,
      "eval_samples_per_second": 929.192,
      "eval_steps_per_second": 14.619,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.9949679970741272,
      "learning_rate": 7.207207207207208e-05,
      "loss": 0.6201,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.568,
      "eval_loss": 1.566339373588562,
      "eval_runtime": 4.0886,
      "eval_samples_per_second": 917.174,
      "eval_steps_per_second": 14.43,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.5366959571838379,
      "learning_rate": 6.435006435006434e-07,
      "loss": 0.2268,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.5797333333333333,
      "eval_loss": 1.5865386724472046,
      "eval_runtime": 4.1849,
      "eval_samples_per_second": 896.071,
      "eval_steps_per_second": 14.098,
      "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
}