trainer_state.json

{
  "best_metric": 0.9056,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0412/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": 78.97106170654297,
      "learning_rate": 7e-05,
      "loss": 3.1603,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6653333333333333,
      "eval_loss": 1.5846168994903564,
      "eval_runtime": 14.7889,
      "eval_samples_per_second": 253.568,
      "eval_steps_per_second": 3.989,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 184.65008544921875,
      "learning_rate": 7e-05,
      "loss": 0.8875,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8344,
      "eval_loss": 0.5694764256477356,
      "eval_runtime": 14.5801,
      "eval_samples_per_second": 257.199,
      "eval_steps_per_second": 4.047,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 343.2808837890625,
      "learning_rate": 7e-05,
      "loss": 0.502,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8642666666666666,
      "eval_loss": 0.44079604744911194,
      "eval_runtime": 14.3417,
      "eval_samples_per_second": 261.475,
      "eval_steps_per_second": 4.114,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 233.72239685058594,
      "learning_rate": 7e-05,
      "loss": 0.363,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8661333333333333,
      "eval_loss": 0.4130479097366333,
      "eval_runtime": 14.5452,
      "eval_samples_per_second": 257.817,
      "eval_steps_per_second": 4.056,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 49.652320861816406,
      "learning_rate": 7e-05,
      "loss": 0.2882,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8872,
      "eval_loss": 0.34735435247421265,
      "eval_runtime": 14.3704,
      "eval_samples_per_second": 260.953,
      "eval_steps_per_second": 4.106,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 107.62310791015625,
      "learning_rate": 7e-05,
      "loss": 0.2201,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8933333333333333,
      "eval_loss": 0.3406805992126465,
      "eval_runtime": 14.3948,
      "eval_samples_per_second": 260.511,
      "eval_steps_per_second": 4.099,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 121.18760681152344,
      "learning_rate": 7e-05,
      "loss": 0.1719,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9056,
      "eval_loss": 0.31824010610580444,
      "eval_runtime": 14.521,
      "eval_samples_per_second": 258.247,
      "eval_steps_per_second": 4.063,
      "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": 5.7235004676864e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}