trainer_state.json

{
  "best_metric": 0.8762666666666666,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0836/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": 84.18557739257812,
      "learning_rate": 7.875e-05,
      "loss": 3.1148,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6677333333333333,
      "eval_loss": 1.3214713335037231,
      "eval_runtime": 7.3898,
      "eval_samples_per_second": 507.454,
      "eval_steps_per_second": 7.984,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 195.1826629638672,
      "learning_rate": 6.75e-05,
      "loss": 0.8697,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8109333333333333,
      "eval_loss": 0.6215065121650696,
      "eval_runtime": 7.3228,
      "eval_samples_per_second": 512.101,
      "eval_steps_per_second": 8.057,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 68.16921997070312,
      "learning_rate": 5.6250000000000005e-05,
      "loss": 0.5273,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8410666666666666,
      "eval_loss": 0.5204702615737915,
      "eval_runtime": 7.3398,
      "eval_samples_per_second": 510.911,
      "eval_steps_per_second": 8.038,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 59.1320915222168,
      "learning_rate": 4.5e-05,
      "loss": 0.4001,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8546666666666667,
      "eval_loss": 0.4551571011543274,
      "eval_runtime": 7.3688,
      "eval_samples_per_second": 508.905,
      "eval_steps_per_second": 8.007,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 94.34404754638672,
      "learning_rate": 3.375e-05,
      "loss": 0.327,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8736,
      "eval_loss": 0.41850653290748596,
      "eval_runtime": 7.3426,
      "eval_samples_per_second": 510.719,
      "eval_steps_per_second": 8.035,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 107.09838104248047,
      "learning_rate": 2.25e-05,
      "loss": 0.2778,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8696,
      "eval_loss": 0.4221225380897522,
      "eval_runtime": 7.3155,
      "eval_samples_per_second": 512.609,
      "eval_steps_per_second": 8.065,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 93.8565444946289,
      "learning_rate": 1.125e-05,
      "loss": 0.2469,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8762666666666666,
      "eval_loss": 0.4010155200958252,
      "eval_runtime": 7.3651,
      "eval_samples_per_second": 509.155,
      "eval_steps_per_second": 8.011,
      "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
}