trainer_state.json

{
  "best_metric": 0.8501333333333333,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0143/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": 92.87234497070312,
      "learning_rate": 4.375e-05,
      "loss": 3.636,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5581333333333334,
      "eval_loss": 3.01126766204834,
      "eval_runtime": 7.1295,
      "eval_samples_per_second": 525.986,
      "eval_steps_per_second": 8.276,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 125.26459503173828,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 1.7342,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7506666666666667,
      "eval_loss": 0.9619526863098145,
      "eval_runtime": 7.0943,
      "eval_samples_per_second": 528.592,
      "eval_steps_per_second": 8.317,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 97.52526092529297,
      "learning_rate": 3.125e-05,
      "loss": 0.7874,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8101333333333334,
      "eval_loss": 0.6372337341308594,
      "eval_runtime": 7.0971,
      "eval_samples_per_second": 528.388,
      "eval_steps_per_second": 8.313,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 100.095947265625,
      "learning_rate": 2.5e-05,
      "loss": 0.5734,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8328,
      "eval_loss": 0.5379827618598938,
      "eval_runtime": 7.0487,
      "eval_samples_per_second": 532.01,
      "eval_steps_per_second": 8.37,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 143.8428192138672,
      "learning_rate": 1.8750000000000002e-05,
      "loss": 0.4756,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8413333333333334,
      "eval_loss": 0.5080221891403198,
      "eval_runtime": 7.0417,
      "eval_samples_per_second": 532.542,
      "eval_steps_per_second": 8.379,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 248.130615234375,
      "learning_rate": 1.25e-05,
      "loss": 0.4243,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8466666666666667,
      "eval_loss": 0.48436662554740906,
      "eval_runtime": 7.0332,
      "eval_samples_per_second": 533.184,
      "eval_steps_per_second": 8.389,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 110.26295471191406,
      "learning_rate": 6.25e-06,
      "loss": 0.3867,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8501333333333333,
      "eval_loss": 0.4639977812767029,
      "eval_runtime": 7.0681,
      "eval_samples_per_second": 530.553,
      "eval_steps_per_second": 8.347,
      "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
}