trainer_state.json

{
  "best_metric": 0.94,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0756/checkpoints/checkpoint-1998",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 1998,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.39246225357055664,
      "learning_rate": 7e-05,
      "loss": 0.7132,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9269333333333334,
      "eval_loss": 0.25721099972724915,
      "eval_runtime": 14.1272,
      "eval_samples_per_second": 265.445,
      "eval_steps_per_second": 4.176,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.11214681714773178,
      "learning_rate": 7e-05,
      "loss": 0.1437,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9384,
      "eval_loss": 0.2290843278169632,
      "eval_runtime": 11.9848,
      "eval_samples_per_second": 312.895,
      "eval_steps_per_second": 4.923,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.01654796116054058,
      "learning_rate": 7e-05,
      "loss": 0.0844,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9338666666666666,
      "eval_loss": 0.23592932522296906,
      "eval_runtime": 11.9851,
      "eval_samples_per_second": 312.888,
      "eval_steps_per_second": 4.923,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 10.901289939880371,
      "learning_rate": 7e-05,
      "loss": 0.0544,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9365333333333333,
      "eval_loss": 0.2592404782772064,
      "eval_runtime": 11.9348,
      "eval_samples_per_second": 314.207,
      "eval_steps_per_second": 4.944,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.036676280200481415,
      "learning_rate": 7e-05,
      "loss": 0.0407,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9274666666666667,
      "eval_loss": 0.2939024865627289,
      "eval_runtime": 11.8981,
      "eval_samples_per_second": 315.177,
      "eval_steps_per_second": 4.959,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.168839231133461,
      "learning_rate": 7e-05,
      "loss": 0.0323,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.94,
      "eval_loss": 0.25257807970046997,
      "eval_runtime": 11.9638,
      "eval_samples_per_second": 313.445,
      "eval_steps_per_second": 4.932,
      "step": 1998
    }
  ],
  "logging_steps": 500,
  "max_steps": 2997,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "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": 9.88447923735552e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}