trainer_state.json

{
  "best_metric": 0.8557333333333333,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0512/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": 219.49855041503906,
      "learning_rate": 3e-05,
      "loss": 3.8077,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.4824,
      "eval_loss": 3.565497875213623,
      "eval_runtime": 7.1821,
      "eval_samples_per_second": 522.131,
      "eval_steps_per_second": 8.215,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 189.3619384765625,
      "learning_rate": 3e-05,
      "loss": 2.8574,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6141333333333333,
      "eval_loss": 1.9694043397903442,
      "eval_runtime": 7.1326,
      "eval_samples_per_second": 525.756,
      "eval_steps_per_second": 8.272,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 241.11798095703125,
      "learning_rate": 3e-05,
      "loss": 1.4048,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7410666666666667,
      "eval_loss": 1.0141870975494385,
      "eval_runtime": 7.0688,
      "eval_samples_per_second": 530.499,
      "eval_steps_per_second": 8.347,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 61.817665100097656,
      "learning_rate": 3e-05,
      "loss": 0.851,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7922666666666667,
      "eval_loss": 0.7226144075393677,
      "eval_runtime": 7.2236,
      "eval_samples_per_second": 519.134,
      "eval_steps_per_second": 8.168,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 154.95501708984375,
      "learning_rate": 3e-05,
      "loss": 0.6468,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8218666666666666,
      "eval_loss": 0.604221761226654,
      "eval_runtime": 7.0647,
      "eval_samples_per_second": 530.81,
      "eval_steps_per_second": 8.351,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 216.7889404296875,
      "learning_rate": 3e-05,
      "loss": 0.5128,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8474666666666667,
      "eval_loss": 0.5239284634590149,
      "eval_runtime": 7.0599,
      "eval_samples_per_second": 531.17,
      "eval_steps_per_second": 8.357,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 126.57896423339844,
      "learning_rate": 3e-05,
      "loss": 0.4375,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8557333333333333,
      "eval_loss": 0.48482343554496765,
      "eval_runtime": 7.1202,
      "eval_samples_per_second": 526.669,
      "eval_steps_per_second": 8.286,
      "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
}