trainer_state.json

{
  "best_metric": 0.9464,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0053/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": 0.7236306667327881,
      "learning_rate": 7.875e-05,
      "loss": 0.613,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9261333333333334,
      "eval_loss": 0.25435394048690796,
      "eval_runtime": 10.6789,
      "eval_samples_per_second": 351.159,
      "eval_steps_per_second": 5.525,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 6.753378378378378e-05,
      "loss": 0.1208,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9386666666666666,
      "eval_loss": 0.22117750346660614,
      "eval_runtime": 10.9988,
      "eval_samples_per_second": 340.945,
      "eval_steps_per_second": 5.364,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.11881420761346817,
      "learning_rate": 5.628378378378378e-05,
      "loss": 0.0498,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9312,
      "eval_loss": 0.2583669126033783,
      "eval_runtime": 11.445,
      "eval_samples_per_second": 327.655,
      "eval_steps_per_second": 5.155,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.2163669615983963,
      "learning_rate": 4.503378378378379e-05,
      "loss": 0.0192,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9418666666666666,
      "eval_loss": 0.23416417837142944,
      "eval_runtime": 10.7753,
      "eval_samples_per_second": 348.019,
      "eval_steps_per_second": 5.475,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.030539333820343018,
      "learning_rate": 3.3783783783783784e-05,
      "loss": 0.0082,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.944,
      "eval_loss": 0.2338414490222931,
      "eval_runtime": 10.8879,
      "eval_samples_per_second": 344.418,
      "eval_steps_per_second": 5.419,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.032537102699279785,
      "learning_rate": 2.2533783783783782e-05,
      "loss": 0.0036,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.944,
      "eval_loss": 0.2368467003107071,
      "eval_runtime": 10.5369,
      "eval_samples_per_second": 355.891,
      "eval_steps_per_second": 5.599,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.010331105440855026,
      "learning_rate": 1.1283783783783785e-05,
      "loss": 0.0025,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9464,
      "eval_loss": 0.2364559918642044,
      "eval_runtime": 11.5982,
      "eval_samples_per_second": 323.325,
      "eval_steps_per_second": 5.087,
      "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
}