run-2/checkpoint-6414/trainer_state.json

{
  "best_metric": 0.42419730237002035,
  "best_model_checkpoint": "distilbert-base-uncased-finetuned-cola/run-2/checkpoint-6414",
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 6414,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.23386342376052385,
      "grad_norm": 3.80633282661438,
      "learning_rate": 1.4038087770081304e-06,
      "loss": 0.6287,
      "step": 500
    },
    {
      "epoch": 0.4677268475210477,
      "grad_norm": 5.112864017486572,
      "learning_rate": 1.3349270902069464e-06,
      "loss": 0.5844,
      "step": 1000
    },
    {
      "epoch": 0.7015902712815716,
      "grad_norm": 16.39091682434082,
      "learning_rate": 1.2660454034057625e-06,
      "loss": 0.5781,
      "step": 1500
    },
    {
      "epoch": 0.9354536950420954,
      "grad_norm": 7.839633464813232,
      "learning_rate": 1.1971637166045783e-06,
      "loss": 0.5482,
      "step": 2000
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.5578071475028992,
      "eval_matthews_correlation": 0.25376727839543856,
      "eval_runtime": 1.0111,
      "eval_samples_per_second": 1031.593,
      "eval_steps_per_second": 129.567,
      "step": 2138
    },
    {
      "epoch": 1.1693171188026192,
      "grad_norm": 17.102230072021484,
      "learning_rate": 1.1282820298033944e-06,
      "loss": 0.5349,
      "step": 2500
    },
    {
      "epoch": 1.4031805425631432,
      "grad_norm": 14.366960525512695,
      "learning_rate": 1.0594003430022104e-06,
      "loss": 0.5129,
      "step": 3000
    },
    {
      "epoch": 1.637043966323667,
      "grad_norm": 26.894271850585938,
      "learning_rate": 9.905186562010263e-07,
      "loss": 0.482,
      "step": 3500
    },
    {
      "epoch": 1.8709073900841908,
      "grad_norm": 9.88732624053955,
      "learning_rate": 9.216369693998422e-07,
      "loss": 0.4997,
      "step": 4000
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.5702487230300903,
      "eval_matthews_correlation": 0.4033061591161787,
      "eval_runtime": 1.0209,
      "eval_samples_per_second": 1021.619,
      "eval_steps_per_second": 128.315,
      "step": 4276
    },
    {
      "epoch": 2.1047708138447145,
      "grad_norm": 10.997356414794922,
      "learning_rate": 8.527552825986582e-07,
      "loss": 0.4925,
      "step": 4500
    },
    {
      "epoch": 2.3386342376052385,
      "grad_norm": 10.581222534179688,
      "learning_rate": 7.838735957974742e-07,
      "loss": 0.4627,
      "step": 5000
    },
    {
      "epoch": 2.5724976613657624,
      "grad_norm": 23.820741653442383,
      "learning_rate": 7.149919089962902e-07,
      "loss": 0.4931,
      "step": 5500
    },
    {
      "epoch": 2.8063610851262863,
      "grad_norm": 34.82048416137695,
      "learning_rate": 6.461102221951062e-07,
      "loss": 0.4764,
      "step": 6000
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.6053777933120728,
      "eval_matthews_correlation": 0.42419730237002035,
      "eval_runtime": 1.0601,
      "eval_samples_per_second": 983.872,
      "eval_steps_per_second": 123.574,
      "step": 6414
    }
  ],
  "logging_steps": 500,
  "max_steps": 10690,
  "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": 97760940208128.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": {
    "learning_rate": 1.4726904638093145e-06,
    "num_train_epochs": 5,
    "per_device_train_batch_size": 4,
    "seed": 37
  }
}