trainer_state.json

{
  "best_metric": 0.8488,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0609/checkpoints/checkpoint-1665",
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 1665,
  "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": 9e-05,
      "loss": 0.9053,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8114666666666667,
      "eval_loss": 0.5920436382293701,
      "eval_runtime": 11.3615,
      "eval_samples_per_second": 330.061,
      "eval_steps_per_second": 5.193,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 82.49163055419922,
      "learning_rate": 9e-05,
      "loss": 0.339,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8376,
      "eval_loss": 0.5626853108406067,
      "eval_runtime": 11.3823,
      "eval_samples_per_second": 329.458,
      "eval_steps_per_second": 5.183,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.14377477765083313,
      "learning_rate": 9e-05,
      "loss": 0.209,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8061333333333334,
      "eval_loss": 0.7162261605262756,
      "eval_runtime": 11.3981,
      "eval_samples_per_second": 329.003,
      "eval_steps_per_second": 5.176,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": Infinity,
      "learning_rate": 9e-05,
      "loss": 0.1688,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8376,
      "eval_loss": 0.6305705904960632,
      "eval_runtime": 11.2146,
      "eval_samples_per_second": 334.386,
      "eval_steps_per_second": 5.261,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.2679138779640198,
      "learning_rate": 9e-05,
      "loss": 0.1411,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8488,
      "eval_loss": 0.5748131275177002,
      "eval_runtime": 11.4045,
      "eval_samples_per_second": 328.818,
      "eval_steps_per_second": 5.173,
      "step": 1665
    }
  ],
  "logging_steps": 500,
  "max_steps": 1665,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "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": 8.2370660311296e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}