trainer_state.json

{
  "best_metric": 0.8752,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0080/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": 179.4822998046875,
      "learning_rate": 6.733578363789503e-05,
      "loss": 3.3323,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6141333333333333,
      "eval_loss": 1.860074758529663,
      "eval_runtime": 7.4876,
      "eval_samples_per_second": 500.826,
      "eval_steps_per_second": 7.88,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 173.17893981933594,
      "learning_rate": 5.974873734152915e-05,
      "loss": 1.0986,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8034666666666667,
      "eval_loss": 0.7167912125587463,
      "eval_runtime": 7.7098,
      "eval_samples_per_second": 486.396,
      "eval_steps_per_second": 7.653,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 72.3719482421875,
      "learning_rate": 4.839392013277814e-05,
      "loss": 0.6353,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8346666666666667,
      "eval_loss": 0.5574302077293396,
      "eval_runtime": 7.7694,
      "eval_samples_per_second": 482.665,
      "eval_steps_per_second": 7.594,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 157.95555114746094,
      "learning_rate": 3.5e-05,
      "loss": 0.4706,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8570666666666666,
      "eval_loss": 0.48069775104522705,
      "eval_runtime": 7.8665,
      "eval_samples_per_second": 476.708,
      "eval_steps_per_second": 7.5,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 159.4249267578125,
      "learning_rate": 2.1606079867221858e-05,
      "loss": 0.3826,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.864,
      "eval_loss": 0.43981289863586426,
      "eval_runtime": 7.7605,
      "eval_samples_per_second": 483.215,
      "eval_steps_per_second": 7.603,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 63.015586853027344,
      "learning_rate": 1.0251262658470837e-05,
      "loss": 0.3383,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.872,
      "eval_loss": 0.42167553305625916,
      "eval_runtime": 7.7267,
      "eval_samples_per_second": 485.331,
      "eval_steps_per_second": 7.636,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 97.74658966064453,
      "learning_rate": 2.664216362104964e-06,
      "loss": 0.3157,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8752,
      "eval_loss": 0.4171558618545532,
      "eval_runtime": 7.3989,
      "eval_samples_per_second": 506.83,
      "eval_steps_per_second": 7.974,
      "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": 5.7235004676864e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}