trainer_state.json

{
  "best_metric": 0.868,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0925/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": 179.57933044433594,
      "learning_rate": 7e-05,
      "loss": 3.3109,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5874666666666667,
      "eval_loss": 2.086618423461914,
      "eval_runtime": 13.7309,
      "eval_samples_per_second": 273.106,
      "eval_steps_per_second": 4.297,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 92.38438415527344,
      "learning_rate": 7e-05,
      "loss": 1.1648,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7706666666666667,
      "eval_loss": 0.7661831378936768,
      "eval_runtime": 12.8691,
      "eval_samples_per_second": 291.396,
      "eval_steps_per_second": 4.585,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 78.4480209350586,
      "learning_rate": 7e-05,
      "loss": 0.6498,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8330666666666666,
      "eval_loss": 0.5602118372917175,
      "eval_runtime": 12.9075,
      "eval_samples_per_second": 290.529,
      "eval_steps_per_second": 4.571,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 93.47808837890625,
      "learning_rate": 7e-05,
      "loss": 0.4753,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8504,
      "eval_loss": 0.4936588406562805,
      "eval_runtime": 12.9497,
      "eval_samples_per_second": 289.582,
      "eval_steps_per_second": 4.556,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 111.59913635253906,
      "learning_rate": 7e-05,
      "loss": 0.3708,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8626666666666667,
      "eval_loss": 0.46356454491615295,
      "eval_runtime": 13.0795,
      "eval_samples_per_second": 286.708,
      "eval_steps_per_second": 4.511,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 127.00260162353516,
      "learning_rate": 7e-05,
      "loss": 0.3085,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8669333333333333,
      "eval_loss": 0.450529545545578,
      "eval_runtime": 12.9304,
      "eval_samples_per_second": 290.014,
      "eval_steps_per_second": 4.563,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 153.0213165283203,
      "learning_rate": 7e-05,
      "loss": 0.2389,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8674666666666667,
      "eval_loss": 0.44610899686813354,
      "eval_runtime": 12.8338,
      "eval_samples_per_second": 292.197,
      "eval_steps_per_second": 4.597,
      "step": 2331
    },
    {
      "epoch": 8.0,
      "grad_norm": 75.49369049072266,
      "learning_rate": 7e-05,
      "loss": 0.1981,
      "step": 2664
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.8666666666666667,
      "eval_loss": 0.4406563937664032,
      "eval_runtime": 12.831,
      "eval_samples_per_second": 292.262,
      "eval_steps_per_second": 4.598,
      "step": 2664
    },
    {
      "epoch": 9.0,
      "grad_norm": 126.18038940429688,
      "learning_rate": 7e-05,
      "loss": 0.1642,
      "step": 2997
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.868,
      "eval_loss": 0.4461146593093872,
      "eval_runtime": 12.8343,
      "eval_samples_per_second": 292.187,
      "eval_steps_per_second": 4.597,
      "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
}