trainer_state.json

{
  "best_metric": 0.9533333333333334,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0742/checkpoints/checkpoint-2664",
  "epoch": 8.0,
  "eval_steps": 500,
  "global_step": 2664,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.49540460109710693,
      "learning_rate": 8.657457896300791e-05,
      "loss": 0.6318,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9349333333333333,
      "eval_loss": 0.24303771555423737,
      "eval_runtime": 12.6458,
      "eval_samples_per_second": 296.541,
      "eval_steps_per_second": 4.666,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.3447047472000122,
      "learning_rate": 7.681980515339464e-05,
      "loss": 0.1179,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.944,
      "eval_loss": 0.20592908561229706,
      "eval_runtime": 12.4857,
      "eval_samples_per_second": 300.344,
      "eval_steps_per_second": 4.725,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.05165191739797592,
      "learning_rate": 6.222075445642905e-05,
      "loss": 0.0469,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9378666666666666,
      "eval_loss": 0.22994576394557953,
      "eval_runtime": 12.4303,
      "eval_samples_per_second": 301.683,
      "eval_steps_per_second": 4.746,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 4.704019546508789,
      "learning_rate": 4.5e-05,
      "loss": 0.0192,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9461333333333334,
      "eval_loss": 0.20835979282855988,
      "eval_runtime": 12.8254,
      "eval_samples_per_second": 292.389,
      "eval_steps_per_second": 4.6,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.05869955196976662,
      "learning_rate": 2.7779245543570963e-05,
      "loss": 0.0047,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9509333333333333,
      "eval_loss": 0.20348839461803436,
      "eval_runtime": 12.4942,
      "eval_samples_per_second": 300.14,
      "eval_steps_per_second": 4.722,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.0426449179649353,
      "learning_rate": 1.3180194846605365e-05,
      "loss": 0.0021,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9504,
      "eval_loss": 0.20653191208839417,
      "eval_runtime": 11.6776,
      "eval_samples_per_second": 321.126,
      "eval_steps_per_second": 5.052,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": NaN,
      "learning_rate": 3.4457579724370674e-06,
      "loss": 0.0014,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9514666666666667,
      "eval_loss": 0.2031608372926712,
      "eval_runtime": 12.5112,
      "eval_samples_per_second": 299.73,
      "eval_steps_per_second": 4.716,
      "step": 2331
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.016280239447951317,
      "learning_rate": 3.12905914479078e-11,
      "loss": 0.0012,
      "step": 2664
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9533333333333334,
      "eval_loss": 0.20442259311676025,
      "eval_runtime": 12.5656,
      "eval_samples_per_second": 298.434,
      "eval_steps_per_second": 4.695,
      "step": 2664
    }
  ],
  "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": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.317930564980736e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}