trainer_state.json

{
  "best_metric": 0.892,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0473/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": 64.9155044555664,
      "learning_rate": 7e-05,
      "loss": 3.1803,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6741333333333334,
      "eval_loss": 1.432660698890686,
      "eval_runtime": 8.1944,
      "eval_samples_per_second": 457.631,
      "eval_steps_per_second": 7.2,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 51.79361343383789,
      "learning_rate": 7e-05,
      "loss": 0.8427,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8346666666666667,
      "eval_loss": 0.5580233335494995,
      "eval_runtime": 7.3968,
      "eval_samples_per_second": 506.973,
      "eval_steps_per_second": 7.976,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 147.0055389404297,
      "learning_rate": 7e-05,
      "loss": 0.4803,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8656,
      "eval_loss": 0.4396216869354248,
      "eval_runtime": 7.3839,
      "eval_samples_per_second": 507.864,
      "eval_steps_per_second": 7.99,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 66.55233001708984,
      "learning_rate": 7e-05,
      "loss": 0.3317,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8776,
      "eval_loss": 0.40195780992507935,
      "eval_runtime": 7.4143,
      "eval_samples_per_second": 505.779,
      "eval_steps_per_second": 7.958,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 160.28334045410156,
      "learning_rate": 7e-05,
      "loss": 0.2576,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8861333333333333,
      "eval_loss": 0.3800772726535797,
      "eval_runtime": 7.5175,
      "eval_samples_per_second": 498.834,
      "eval_steps_per_second": 7.848,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 91.19947052001953,
      "learning_rate": 7e-05,
      "loss": 0.1951,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8904,
      "eval_loss": 0.37971895933151245,
      "eval_runtime": 7.512,
      "eval_samples_per_second": 499.202,
      "eval_steps_per_second": 7.854,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 72.50609588623047,
      "learning_rate": 7e-05,
      "loss": 0.1578,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.892,
      "eval_loss": 0.36846038699150085,
      "eval_runtime": 7.2064,
      "eval_samples_per_second": 520.374,
      "eval_steps_per_second": 8.187,
      "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
}