{
  "best_metric": 0.7326732673267327,
  "best_model_checkpoint": "distilhubert-finetuned-not-a-word2/run-5/checkpoint-24",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 144,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.21,
      "grad_norm": 1.6560618877410889,
      "learning_rate": 9.928194518007172e-07,
      "loss": 0.6997,
      "step": 5
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.6409028768539429,
      "learning_rate": 1.9856389036014343e-06,
      "loss": 0.6971,
      "step": 10
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.6359347105026245,
      "learning_rate": 2.978458355402151e-06,
      "loss": 0.6953,
      "step": 15
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.56245756149292,
      "learning_rate": 3.971277807202869e-06,
      "loss": 0.6889,
      "step": 20
    },
    {
      "epoch": 1.0,
      "eval_f1": 0.7326732673267327,
      "eval_loss": 0.6865768432617188,
      "eval_runtime": 1.3652,
      "eval_samples_per_second": 46.879,
      "eval_steps_per_second": 5.86,
      "step": 24
    },
    {
      "epoch": 1.04,
      "grad_norm": 0.6976048946380615,
      "learning_rate": 4.743470714158982e-06,
      "loss": 0.6865,
      "step": 25
    },
    {
      "epoch": 1.25,
      "grad_norm": 2.008615732192993,
      "learning_rate": 4.63315744173668e-06,
      "loss": 0.6681,
      "step": 30
    },
    {
      "epoch": 1.46,
      "grad_norm": 1.9885936975479126,
      "learning_rate": 4.522844169314378e-06,
      "loss": 0.6531,
      "step": 35
    },
    {
      "epoch": 1.67,
      "grad_norm": 0.5047381520271301,
      "learning_rate": 4.412530896892076e-06,
      "loss": 0.6664,
      "step": 40
    },
    {
      "epoch": 1.88,
      "grad_norm": 0.9776627421379089,
      "learning_rate": 4.302217624469774e-06,
      "loss": 0.6545,
      "step": 45
    },
    {
      "epoch": 2.0,
      "eval_f1": 0.7326732673267327,
      "eval_loss": 0.680206298828125,
      "eval_runtime": 1.3708,
      "eval_samples_per_second": 46.689,
      "eval_steps_per_second": 5.836,
      "step": 48
    },
    {
      "epoch": 2.08,
      "grad_norm": 0.46269264817237854,
      "learning_rate": 4.191904352047472e-06,
      "loss": 0.6518,
      "step": 50
    },
    {
      "epoch": 2.29,
      "grad_norm": 0.8560582995414734,
      "learning_rate": 4.081591079625171e-06,
      "loss": 0.6333,
      "step": 55
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.9135796427726746,
      "learning_rate": 3.971277807202869e-06,
      "loss": 0.6494,
      "step": 60
    },
    {
      "epoch": 2.71,
      "grad_norm": 0.71152263879776,
      "learning_rate": 3.8609645347805665e-06,
      "loss": 0.6444,
      "step": 65
    },
    {
      "epoch": 2.92,
      "grad_norm": 1.3949508666992188,
      "learning_rate": 3.750651262358265e-06,
      "loss": 0.5726,
      "step": 70
    },
    {
      "epoch": 3.0,
      "eval_f1": 0.7326732673267327,
      "eval_loss": 0.6824684143066406,
      "eval_runtime": 1.373,
      "eval_samples_per_second": 46.614,
      "eval_steps_per_second": 5.827,
      "step": 72
    },
    {
      "epoch": 3.12,
      "grad_norm": 0.7415557503700256,
      "learning_rate": 3.6403379899359627e-06,
      "loss": 0.6271,
      "step": 75
    },
    {
      "epoch": 3.33,
      "grad_norm": 0.6959993243217468,
      "learning_rate": 3.5300247175136605e-06,
      "loss": 0.6316,
      "step": 80
    },
    {
      "epoch": 3.54,
      "grad_norm": 1.2839587926864624,
      "learning_rate": 3.419711445091359e-06,
      "loss": 0.5607,
      "step": 85
    },
    {
      "epoch": 3.75,
      "grad_norm": 1.2485283613204956,
      "learning_rate": 3.309398172669057e-06,
      "loss": 0.6403,
      "step": 90
    },
    {
      "epoch": 3.96,
      "grad_norm": 1.1759371757507324,
      "learning_rate": 3.199084900246755e-06,
      "loss": 0.6242,
      "step": 95
    },
    {
      "epoch": 4.0,
      "eval_f1": 0.7326732673267327,
      "eval_loss": 0.6892623901367188,
      "eval_runtime": 1.3822,
      "eval_samples_per_second": 46.302,
      "eval_steps_per_second": 5.788,
      "step": 96
    },
    {
      "epoch": 4.17,
      "grad_norm": 0.6085842251777649,
      "learning_rate": 3.0887716278244532e-06,
      "loss": 0.564,
      "step": 100
    },
    {
      "epoch": 4.38,
      "grad_norm": 0.9830076098442078,
      "learning_rate": 2.978458355402151e-06,
      "loss": 0.641,
      "step": 105
    },
    {
      "epoch": 4.58,
      "grad_norm": 0.6557525992393494,
      "learning_rate": 2.8681450829798494e-06,
      "loss": 0.5943,
      "step": 110
    },
    {
      "epoch": 4.79,
      "grad_norm": 0.7436282634735107,
      "learning_rate": 2.7578318105575477e-06,
      "loss": 0.614,
      "step": 115
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.7114256620407104,
      "learning_rate": 2.6475185381352455e-06,
      "loss": 0.6081,
      "step": 120
    },
    {
      "epoch": 5.0,
      "eval_f1": 0.7326732673267327,
      "eval_loss": 0.6939811706542969,
      "eval_runtime": 1.3699,
      "eval_samples_per_second": 46.719,
      "eval_steps_per_second": 5.84,
      "step": 120
    },
    {
      "epoch": 5.21,
      "grad_norm": 0.6939758062362671,
      "learning_rate": 2.537205265712944e-06,
      "loss": 0.5683,
      "step": 125
    },
    {
      "epoch": 5.42,
      "grad_norm": 0.7150554656982422,
      "learning_rate": 2.426891993290642e-06,
      "loss": 0.6389,
      "step": 130
    },
    {
      "epoch": 5.62,
      "grad_norm": 0.5391159057617188,
      "learning_rate": 2.31657872086834e-06,
      "loss": 0.6185,
      "step": 135
    },
    {
      "epoch": 5.83,
      "grad_norm": 0.657724142074585,
      "learning_rate": 2.206265448446038e-06,
      "loss": 0.6216,
      "step": 140
    },
    {
      "epoch": 6.0,
      "eval_f1": 0.7326732673267327,
      "eval_loss": 0.69482421875,
      "eval_runtime": 1.3614,
      "eval_samples_per_second": 47.009,
      "eval_steps_per_second": 5.876,
      "step": 144
    }
  ],
  "logging_steps": 5,
  "max_steps": 240,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "total_flos": 4911381340990080.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": {
    "learning_rate": 4.765533368643442e-06,
    "per_device_train_batch_size": 8
  }
}