trainer_state.json

{
  "best_metric": 0.8368,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0802/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": 188.6533203125,
      "learning_rate": 3e-05,
      "loss": 3.8475,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.29973333333333335,
      "eval_loss": 3.7053184509277344,
      "eval_runtime": 8.6972,
      "eval_samples_per_second": 431.171,
      "eval_steps_per_second": 6.784,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 292.2791442871094,
      "learning_rate": 3e-05,
      "loss": 3.1137,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5490666666666667,
      "eval_loss": 2.203028678894043,
      "eval_runtime": 8.8121,
      "eval_samples_per_second": 425.554,
      "eval_steps_per_second": 6.695,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 174.6547088623047,
      "learning_rate": 3e-05,
      "loss": 1.6225,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7152,
      "eval_loss": 1.1442999839782715,
      "eval_runtime": 8.3914,
      "eval_samples_per_second": 446.886,
      "eval_steps_per_second": 7.031,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 191.1522979736328,
      "learning_rate": 3e-05,
      "loss": 1.0003,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7682666666666667,
      "eval_loss": 0.8166203498840332,
      "eval_runtime": 8.4213,
      "eval_samples_per_second": 445.302,
      "eval_steps_per_second": 7.006,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 127.65345764160156,
      "learning_rate": 3e-05,
      "loss": 0.7538,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8013333333333333,
      "eval_loss": 0.6535776257514954,
      "eval_runtime": 8.5035,
      "eval_samples_per_second": 440.996,
      "eval_steps_per_second": 6.938,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 77.17083740234375,
      "learning_rate": 3e-05,
      "loss": 0.6167,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8176,
      "eval_loss": 0.5795504450798035,
      "eval_runtime": 8.4666,
      "eval_samples_per_second": 442.918,
      "eval_steps_per_second": 6.969,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 168.13217163085938,
      "learning_rate": 3e-05,
      "loss": 0.5298,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8368,
      "eval_loss": 0.5250616073608398,
      "eval_runtime": 8.3556,
      "eval_samples_per_second": 448.803,
      "eval_steps_per_second": 7.061,
      "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
}