run-3/checkpoint-10552/trainer_state.json

{
  "best_metric": 0.9078847611827142,
  "best_model_checkpoint": "distilbert-base-uncased-finetuned-multiclass-classification/run-3/checkpoint-10552",
  "epoch": 4.0,
  "eval_steps": 500,
  "global_step": 10552,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.18953752843062927,
      "grad_norm": 9.870668411254883,
      "learning_rate": 1.490581283927063e-05,
      "loss": 5.8661,
      "step": 500
    },
    {
      "epoch": 0.37907505686125853,
      "grad_norm": 11.824650764465332,
      "learning_rate": 1.4318507368850198e-05,
      "loss": 4.7233,
      "step": 1000
    },
    {
      "epoch": 0.5686125852918877,
      "grad_norm": 11.662109375,
      "learning_rate": 1.3731201898429762e-05,
      "loss": 3.8044,
      "step": 1500
    },
    {
      "epoch": 0.7581501137225171,
      "grad_norm": 11.159646987915039,
      "learning_rate": 1.3143896428009328e-05,
      "loss": 3.0524,
      "step": 2000
    },
    {
      "epoch": 0.9476876421531463,
      "grad_norm": 13.040279388427734,
      "learning_rate": 1.2556590957588894e-05,
      "loss": 2.5575,
      "step": 2500
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7350265352539803,
      "eval_loss": 2.0373177528381348,
      "eval_runtime": 41.2753,
      "eval_samples_per_second": 63.912,
      "eval_steps_per_second": 3.998,
      "step": 2638
    },
    {
      "epoch": 1.1372251705837755,
      "grad_norm": 14.501738548278809,
      "learning_rate": 1.1969285487168458e-05,
      "loss": 2.0424,
      "step": 3000
    },
    {
      "epoch": 1.3267626990144048,
      "grad_norm": 8.37807846069336,
      "learning_rate": 1.1381980016748024e-05,
      "loss": 1.6845,
      "step": 3500
    },
    {
      "epoch": 1.5163002274450341,
      "grad_norm": 16.81246566772461,
      "learning_rate": 1.0794674546327588e-05,
      "loss": 1.4752,
      "step": 4000
    },
    {
      "epoch": 1.7058377558756632,
      "grad_norm": 9.910699844360352,
      "learning_rate": 1.0207369075907154e-05,
      "loss": 1.3347,
      "step": 4500
    },
    {
      "epoch": 1.8953752843062928,
      "grad_norm": 8.623641014099121,
      "learning_rate": 9.62006360548672e-06,
      "loss": 1.185,
      "step": 5000
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8442001516300227,
      "eval_loss": 1.0104457139968872,
      "eval_runtime": 41.2462,
      "eval_samples_per_second": 63.957,
      "eval_steps_per_second": 4.0,
      "step": 5276
    },
    {
      "epoch": 2.084912812736922,
      "grad_norm": 4.728893756866455,
      "learning_rate": 9.032758135066284e-06,
      "loss": 0.9927,
      "step": 5500
    },
    {
      "epoch": 2.274450341167551,
      "grad_norm": 2.276144027709961,
      "learning_rate": 8.445452664645852e-06,
      "loss": 0.8242,
      "step": 6000
    },
    {
      "epoch": 2.4639878695981805,
      "grad_norm": 2.264138698577881,
      "learning_rate": 7.858147194225416e-06,
      "loss": 0.8038,
      "step": 6500
    },
    {
      "epoch": 2.6535253980288096,
      "grad_norm": 3.794312000274658,
      "learning_rate": 7.270841723804981e-06,
      "loss": 0.8027,
      "step": 7000
    },
    {
      "epoch": 2.8430629264594387,
      "grad_norm": 8.422940254211426,
      "learning_rate": 6.683536253384547e-06,
      "loss": 0.6962,
      "step": 7500
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8847611827141774,
      "eval_loss": 0.7159162163734436,
      "eval_runtime": 45.9678,
      "eval_samples_per_second": 57.388,
      "eval_steps_per_second": 3.589,
      "step": 7914
    },
    {
      "epoch": 3.0326004548900682,
      "grad_norm": 2.8076231479644775,
      "learning_rate": 6.096230782964112e-06,
      "loss": 0.6665,
      "step": 8000
    },
    {
      "epoch": 3.2221379833206973,
      "grad_norm": 6.104276657104492,
      "learning_rate": 5.508925312543677e-06,
      "loss": 0.5994,
      "step": 8500
    },
    {
      "epoch": 3.411675511751327,
      "grad_norm": 2.379484176635742,
      "learning_rate": 4.921619842123243e-06,
      "loss": 0.5959,
      "step": 9000
    },
    {
      "epoch": 3.601213040181956,
      "grad_norm": 10.061728477478027,
      "learning_rate": 4.334314371702808e-06,
      "loss": 0.5272,
      "step": 9500
    },
    {
      "epoch": 3.7907505686125855,
      "grad_norm": 3.485311269760132,
      "learning_rate": 3.747008901282373e-06,
      "loss": 0.459,
      "step": 10000
    },
    {
      "epoch": 3.9802880970432146,
      "grad_norm": 2.9593420028686523,
      "learning_rate": 3.1597034308619386e-06,
      "loss": 0.507,
      "step": 10500
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9078847611827142,
      "eval_loss": 0.606600284576416,
      "eval_runtime": 40.7207,
      "eval_samples_per_second": 64.783,
      "eval_steps_per_second": 4.052,
      "step": 10552
    }
  ],
  "logging_steps": 500,
  "max_steps": 13190,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "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": 353736376451072.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": {
    "learning_rate": 1.5493118309691066e-05,
    "num_train_epochs": 5,
    "per_device_train_batch_size": 4,
    "seed": 29
  }
}