trainer_state.json

{
  "best_metric": 0.876,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0467/checkpoints/checkpoint-2997",
  "epoch": 9.0,
  "eval_steps": 500,
  "global_step": 2997,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 122.7616195678711,
      "learning_rate": 3e-05,
      "loss": 3.8168,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.4288,
      "eval_loss": 3.6064181327819824,
      "eval_runtime": 14.2061,
      "eval_samples_per_second": 263.972,
      "eval_steps_per_second": 4.153,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 305.166748046875,
      "learning_rate": 3e-05,
      "loss": 2.9433,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6288,
      "eval_loss": 2.1493144035339355,
      "eval_runtime": 14.0093,
      "eval_samples_per_second": 267.68,
      "eval_steps_per_second": 4.211,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 242.25543212890625,
      "learning_rate": 3e-05,
      "loss": 1.387,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7728,
      "eval_loss": 0.9492793083190918,
      "eval_runtime": 13.9957,
      "eval_samples_per_second": 267.939,
      "eval_steps_per_second": 4.216,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 163.9605255126953,
      "learning_rate": 3e-05,
      "loss": 0.7942,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8194666666666667,
      "eval_loss": 0.6805785298347473,
      "eval_runtime": 14.1837,
      "eval_samples_per_second": 264.388,
      "eval_steps_per_second": 4.16,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 188.47296142578125,
      "learning_rate": 3e-05,
      "loss": 0.563,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8381333333333333,
      "eval_loss": 0.5517776608467102,
      "eval_runtime": 14.0607,
      "eval_samples_per_second": 266.701,
      "eval_steps_per_second": 4.196,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 108.29222869873047,
      "learning_rate": 3e-05,
      "loss": 0.437,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8525333333333334,
      "eval_loss": 0.5053194761276245,
      "eval_runtime": 14.2808,
      "eval_samples_per_second": 262.591,
      "eval_steps_per_second": 4.131,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 153.24221801757812,
      "learning_rate": 3e-05,
      "loss": 0.3412,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8634666666666667,
      "eval_loss": 0.46413710713386536,
      "eval_runtime": 14.0575,
      "eval_samples_per_second": 266.761,
      "eval_steps_per_second": 4.197,
      "step": 2331
    },
    {
      "epoch": 8.0,
      "grad_norm": 265.21136474609375,
      "learning_rate": 3e-05,
      "loss": 0.2847,
      "step": 2664
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.8733333333333333,
      "eval_loss": 0.43790605664253235,
      "eval_runtime": 14.1295,
      "eval_samples_per_second": 265.402,
      "eval_steps_per_second": 4.176,
      "step": 2664
    },
    {
      "epoch": 9.0,
      "grad_norm": 187.478515625,
      "learning_rate": 3e-05,
      "loss": 0.2285,
      "step": 2997
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.876,
      "eval_loss": 0.42678597569465637,
      "eval_runtime": 14.0945,
      "eval_samples_per_second": 266.062,
      "eval_steps_per_second": 4.186,
      "step": 2997
    }
  ],
  "logging_steps": 500,
  "max_steps": 2997,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "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": 7.3587863155968e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}