trainer_state.json

{
  "best_metric": 0.9064,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0951/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": 31.164644241333008,
      "learning_rate": 0.0003,
      "loss": 1.461,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8565333333333334,
      "eval_loss": 0.4768185019493103,
      "eval_runtime": 12.5153,
      "eval_samples_per_second": 299.634,
      "eval_steps_per_second": 4.714,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 46.83688735961914,
      "learning_rate": 0.0003,
      "loss": 0.3959,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8842666666666666,
      "eval_loss": 0.40122222900390625,
      "eval_runtime": 12.4632,
      "eval_samples_per_second": 300.887,
      "eval_steps_per_second": 4.734,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 38.40022277832031,
      "learning_rate": 0.0003,
      "loss": 0.2399,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8938666666666667,
      "eval_loss": 0.37473446130752563,
      "eval_runtime": 12.4163,
      "eval_samples_per_second": 302.022,
      "eval_steps_per_second": 4.752,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 23.602340698242188,
      "learning_rate": 0.0003,
      "loss": 0.1584,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9034666666666666,
      "eval_loss": 0.34888261556625366,
      "eval_runtime": 12.2312,
      "eval_samples_per_second": 306.593,
      "eval_steps_per_second": 4.824,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 44.045257568359375,
      "learning_rate": 0.0003,
      "loss": 0.1146,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8986666666666666,
      "eval_loss": 0.3816640079021454,
      "eval_runtime": 12.0897,
      "eval_samples_per_second": 310.182,
      "eval_steps_per_second": 4.88,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 30.49802017211914,
      "learning_rate": 0.0003,
      "loss": 0.0924,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8938666666666667,
      "eval_loss": 0.44562095403671265,
      "eval_runtime": 12.1045,
      "eval_samples_per_second": 309.802,
      "eval_steps_per_second": 4.874,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 39.73857879638672,
      "learning_rate": 0.0003,
      "loss": 0.0708,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9005333333333333,
      "eval_loss": 0.44351622462272644,
      "eval_runtime": 12.105,
      "eval_samples_per_second": 309.79,
      "eval_steps_per_second": 4.874,
      "step": 2331
    },
    {
      "epoch": 8.0,
      "grad_norm": 32.05183792114258,
      "learning_rate": 0.0003,
      "loss": 0.0616,
      "step": 2664
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9026666666666666,
      "eval_loss": 0.4516476094722748,
      "eval_runtime": 12.0836,
      "eval_samples_per_second": 310.338,
      "eval_steps_per_second": 4.883,
      "step": 2664
    },
    {
      "epoch": 9.0,
      "grad_norm": 25.742677688598633,
      "learning_rate": 0.0003,
      "loss": 0.0536,
      "step": 2997
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.9064,
      "eval_loss": 0.43947356939315796,
      "eval_runtime": 12.2345,
      "eval_samples_per_second": 306.51,
      "eval_steps_per_second": 4.822,
      "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
}