trainer_state.json

{
  "best_metric": 0.9469333333333333,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0503/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": 24.614662170410156,
      "learning_rate": 6.125e-05,
      "loss": 0.6969,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9197333333333333,
      "eval_loss": 0.2783888876438141,
      "eval_runtime": 6.2139,
      "eval_samples_per_second": 603.481,
      "eval_steps_per_second": 9.495,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.9019562005996704,
      "learning_rate": 5.2499999999999995e-05,
      "loss": 0.1378,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9306666666666666,
      "eval_loss": 0.22865374386310577,
      "eval_runtime": 7.2123,
      "eval_samples_per_second": 519.942,
      "eval_steps_per_second": 8.18,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.0803641676902771,
      "learning_rate": 4.374999999999999e-05,
      "loss": 0.0615,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.936,
      "eval_loss": 0.22571013867855072,
      "eval_runtime": 6.3479,
      "eval_samples_per_second": 590.747,
      "eval_steps_per_second": 9.294,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.04575115814805031,
      "learning_rate": 3.5e-05,
      "loss": 0.029,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9378666666666666,
      "eval_loss": 0.23235498368740082,
      "eval_runtime": 6.2453,
      "eval_samples_per_second": 600.452,
      "eval_steps_per_second": 9.447,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.2334756702184677,
      "learning_rate": 2.6249999999999998e-05,
      "loss": 0.0118,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.94,
      "eval_loss": 0.23994942009449005,
      "eval_runtime": 6.1775,
      "eval_samples_per_second": 607.039,
      "eval_steps_per_second": 9.551,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.029063576832413673,
      "learning_rate": 1.75e-05,
      "loss": 0.0059,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9365333333333333,
      "eval_loss": 0.24738270044326782,
      "eval_runtime": 6.184,
      "eval_samples_per_second": 606.403,
      "eval_steps_per_second": 9.541,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": NaN,
      "learning_rate": 8.776276276276276e-06,
      "loss": 0.0041,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9469333333333333,
      "eval_loss": 0.22257943451404572,
      "eval_runtime": 6.0559,
      "eval_samples_per_second": 619.236,
      "eval_steps_per_second": 9.743,
      "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
}