trainer_state.json

{
  "best_metric": 0.8813333333333333,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0216/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": 171.2074737548828,
      "learning_rate": 8.657457896300791e-05,
      "loss": 2.8052,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7413333333333333,
      "eval_loss": 0.9977177977561951,
      "eval_runtime": 7.5641,
      "eval_samples_per_second": 495.761,
      "eval_steps_per_second": 7.8,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 131.61328125,
      "learning_rate": 7.681980515339464e-05,
      "loss": 0.752,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8285333333333333,
      "eval_loss": 0.5841368436813354,
      "eval_runtime": 7.6362,
      "eval_samples_per_second": 491.08,
      "eval_steps_per_second": 7.726,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 316.19610595703125,
      "learning_rate": 6.222075445642905e-05,
      "loss": 0.4492,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8610666666666666,
      "eval_loss": 0.4547383487224579,
      "eval_runtime": 7.5153,
      "eval_samples_per_second": 498.984,
      "eval_steps_per_second": 7.851,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 209.0740509033203,
      "learning_rate": 4.5e-05,
      "loss": 0.3108,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8621333333333333,
      "eval_loss": 0.47137895226478577,
      "eval_runtime": 7.9267,
      "eval_samples_per_second": 473.086,
      "eval_steps_per_second": 7.443,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 124.44168090820312,
      "learning_rate": 2.7779245543570963e-05,
      "loss": 0.2189,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8778666666666667,
      "eval_loss": 0.4185984432697296,
      "eval_runtime": 7.5352,
      "eval_samples_per_second": 497.661,
      "eval_steps_per_second": 7.83,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 124.99227905273438,
      "learning_rate": 1.3180194846605365e-05,
      "loss": 0.1819,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8770666666666667,
      "eval_loss": 0.41963496804237366,
      "eval_runtime": 7.4868,
      "eval_samples_per_second": 500.883,
      "eval_steps_per_second": 7.881,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 59.8276252746582,
      "learning_rate": 3.425421036992097e-06,
      "loss": 0.1519,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8813333333333333,
      "eval_loss": 0.40798714756965637,
      "eval_runtime": 7.4813,
      "eval_samples_per_second": 501.248,
      "eval_steps_per_second": 7.886,
      "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
}