trainer_state.json

{
  "best_metric": 0.9472,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0233/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": 5.147052764892578,
      "learning_rate": 2.8514533018536286e-05,
      "loss": 1.179,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9248,
      "eval_loss": 0.32095175981521606,
      "eval_runtime": 6.4137,
      "eval_samples_per_second": 584.682,
      "eval_steps_per_second": 9.199,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.6127381324768066,
      "learning_rate": 2.4352347027881003e-05,
      "loss": 0.1918,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9381333333333334,
      "eval_loss": 0.22458773851394653,
      "eval_runtime": 6.4583,
      "eval_samples_per_second": 580.652,
      "eval_steps_per_second": 9.136,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.15513408184051514,
      "learning_rate": 1.8337814009344716e-05,
      "loss": 0.0885,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9402666666666667,
      "eval_loss": 0.21635761857032776,
      "eval_runtime": 6.2883,
      "eval_samples_per_second": 596.349,
      "eval_steps_per_second": 9.383,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 8.302488327026367,
      "learning_rate": 1.1662185990655285e-05,
      "loss": 0.0412,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9458666666666666,
      "eval_loss": 0.20724543929100037,
      "eval_runtime": 5.945,
      "eval_samples_per_second": 630.777,
      "eval_steps_per_second": 9.924,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.14174726605415344,
      "learning_rate": 5.647652972118998e-06,
      "loss": 0.0217,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9442666666666667,
      "eval_loss": 0.20968343317508698,
      "eval_runtime": 6.0685,
      "eval_samples_per_second": 617.946,
      "eval_steps_per_second": 9.722,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.06375926733016968,
      "learning_rate": 1.4854669814637145e-06,
      "loss": 0.0136,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9450666666666667,
      "eval_loss": 0.2111324816942215,
      "eval_runtime": 5.8935,
      "eval_samples_per_second": 636.296,
      "eval_steps_per_second": 10.011,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": NaN,
      "learning_rate": 1.362311416175377e-11,
      "loss": 0.0109,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9472,
      "eval_loss": 0.20160552859306335,
      "eval_runtime": 6.1191,
      "eval_samples_per_second": 612.838,
      "eval_steps_per_second": 9.642,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2331,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "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": 1.153189244358144e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}