{
  "best_metric": 0.6666666666666666,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-3/checkpoint-214",
  "epoch": 18.0,
  "eval_steps": 500,
  "global_step": 3852,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.38364019989967346,
      "learning_rate": 3.063781435548824e-05,
      "loss": 0.0796,
      "step": 214
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6666666666666666,
      "eval_f1": 0.0,
      "eval_loss": 0.054877836257219315,
      "eval_mcc": 0.0,
      "eval_precision": 0.0,
      "eval_recall": 0.0,
      "eval_runtime": 3.1144,
      "eval_samples_per_second": 548.1,
      "eval_steps_per_second": 17.339,
      "step": 214
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.44893428683280945,
      "learning_rate": 2.883558998163599e-05,
      "loss": 0.054,
      "step": 428
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6666666666666666,
      "eval_f1": 0.0,
      "eval_loss": 0.050605349242687225,
      "eval_mcc": 0.0,
      "eval_precision": 0.0,
      "eval_recall": 0.0,
      "eval_runtime": 3.2339,
      "eval_samples_per_second": 527.845,
      "eval_steps_per_second": 16.698,
      "step": 428
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.47303804755210876,
      "learning_rate": 2.7033365607783743e-05,
      "loss": 0.0517,
      "step": 642
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6666666666666666,
      "eval_f1": 0.0,
      "eval_loss": 0.04936650022864342,
      "eval_mcc": 0.0,
      "eval_precision": 0.0,
      "eval_recall": 0.0,
      "eval_runtime": 3.1605,
      "eval_samples_per_second": 540.104,
      "eval_steps_per_second": 17.086,
      "step": 642
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.5809402465820312,
      "learning_rate": 2.523114123393149e-05,
      "loss": 0.0508,
      "step": 856
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6666666666666666,
      "eval_f1": 0.013864818024263433,
      "eval_loss": 0.04913894087076187,
      "eval_mcc": 0.024260699053001704,
      "eval_precision": 0.5,
      "eval_recall": 0.007029876977152899,
      "eval_runtime": 3.8427,
      "eval_samples_per_second": 444.219,
      "eval_steps_per_second": 14.053,
      "step": 856
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.7497197389602661,
      "learning_rate": 2.3428916860079242e-05,
      "loss": 0.0505,
      "step": 1070
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.6666666666666666,
      "eval_f1": 0.017271157167530225,
      "eval_loss": 0.04886631295084953,
      "eval_mcc": 0.027140265094376777,
      "eval_precision": 0.5,
      "eval_recall": 0.008787346221441126,
      "eval_runtime": 3.1197,
      "eval_samples_per_second": 547.164,
      "eval_steps_per_second": 17.309,
      "step": 1070
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.3793525993824005,
      "learning_rate": 2.162669248622699e-05,
      "loss": 0.0503,
      "step": 1284
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.664323374340949,
      "eval_f1": 0.017152658662092625,
      "eval_loss": 0.048080265522003174,
      "eval_mcc": 0.004592958330124466,
      "eval_precision": 0.35714285714285715,
      "eval_recall": 0.008787346221441126,
      "eval_runtime": 3.3992,
      "eval_samples_per_second": 502.172,
      "eval_steps_per_second": 15.886,
      "step": 1284
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.525617778301239,
      "learning_rate": 1.9824468112374745e-05,
      "loss": 0.05,
      "step": 1498
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.664323374340949,
      "eval_f1": 0.017152658662092625,
      "eval_loss": 0.04796423017978668,
      "eval_mcc": 0.004592958330124466,
      "eval_precision": 0.35714285714285715,
      "eval_recall": 0.008787346221441126,
      "eval_runtime": 3.1569,
      "eval_samples_per_second": 540.727,
      "eval_steps_per_second": 17.106,
      "step": 1498
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.456257700920105,
      "learning_rate": 1.8022243738522493e-05,
      "loss": 0.0497,
      "step": 1712
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.6654950205038078,
      "eval_f1": 0.017211703958691912,
      "eval_loss": 0.04803091287612915,
      "eval_mcc": 0.01487410293271824,
      "eval_precision": 0.4166666666666667,
      "eval_recall": 0.008787346221441126,
      "eval_runtime": 3.822,
      "eval_samples_per_second": 446.626,
      "eval_steps_per_second": 14.129,
      "step": 1712
    },
    {
      "epoch": 9.0,
      "grad_norm": 0.4184946119785309,
      "learning_rate": 1.6220019364670245e-05,
      "loss": 0.0498,
      "step": 1926
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.6666666666666666,
      "eval_f1": 0.017271157167530225,
      "eval_loss": 0.04849984124302864,
      "eval_mcc": 0.027140265094376777,
      "eval_precision": 0.5,
      "eval_recall": 0.008787346221441126,
      "eval_runtime": 3.1433,
      "eval_samples_per_second": 543.052,
      "eval_steps_per_second": 17.179,
      "step": 1926
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.32953304052352905,
      "learning_rate": 1.4417794990817994e-05,
      "loss": 0.0494,
      "step": 2140
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.664323374340949,
      "eval_f1": 0.017152658662092625,
      "eval_loss": 0.04773561656475067,
      "eval_mcc": 0.004592958330124466,
      "eval_precision": 0.35714285714285715,
      "eval_recall": 0.008787346221441126,
      "eval_runtime": 3.2081,
      "eval_samples_per_second": 532.094,
      "eval_steps_per_second": 16.833,
      "step": 2140
    },
    {
      "epoch": 11.0,
      "grad_norm": 0.2775495946407318,
      "learning_rate": 1.2615570616965746e-05,
      "loss": 0.0494,
      "step": 2354
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.664323374340949,
      "eval_f1": 0.017152658662092625,
      "eval_loss": 0.04799096658825874,
      "eval_mcc": 0.004592958330124466,
      "eval_precision": 0.35714285714285715,
      "eval_recall": 0.008787346221441126,
      "eval_runtime": 3.142,
      "eval_samples_per_second": 543.29,
      "eval_steps_per_second": 17.187,
      "step": 2354
    },
    {
      "epoch": 12.0,
      "grad_norm": 0.2784470319747925,
      "learning_rate": 1.0813346243113495e-05,
      "loss": 0.0494,
      "step": 2568
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.6660808435852372,
      "eval_f1": 0.01724137931034483,
      "eval_loss": 0.04788310080766678,
      "eval_mcc": 0.020707884164064556,
      "eval_precision": 0.45454545454545453,
      "eval_recall": 0.008787346221441126,
      "eval_runtime": 3.8964,
      "eval_samples_per_second": 438.099,
      "eval_steps_per_second": 13.859,
      "step": 2568
    },
    {
      "epoch": 13.0,
      "grad_norm": 0.7122122049331665,
      "learning_rate": 9.011121869261247e-06,
      "loss": 0.0493,
      "step": 2782
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.6654950205038078,
      "eval_f1": 0.017211703958691912,
      "eval_loss": 0.04763857275247574,
      "eval_mcc": 0.01487410293271824,
      "eval_precision": 0.4166666666666667,
      "eval_recall": 0.008787346221441126,
      "eval_runtime": 3.1513,
      "eval_samples_per_second": 541.682,
      "eval_steps_per_second": 17.136,
      "step": 2782
    },
    {
      "epoch": 14.0,
      "grad_norm": 0.3367031216621399,
      "learning_rate": 7.208897495408997e-06,
      "loss": 0.0491,
      "step": 2996
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.6649091974223784,
      "eval_f1": 0.01718213058419244,
      "eval_loss": 0.0474877767264843,
      "eval_mcc": 0.009529862152017439,
      "eval_precision": 0.38461538461538464,
      "eval_recall": 0.008787346221441126,
      "eval_runtime": 3.2218,
      "eval_samples_per_second": 529.831,
      "eval_steps_per_second": 16.761,
      "step": 2996
    },
    {
      "epoch": 15.0,
      "grad_norm": 0.6209991574287415,
      "learning_rate": 5.406673121556748e-06,
      "loss": 0.049,
      "step": 3210
    },
    {
      "epoch": 15.0,
      "eval_accuracy": 0.6649091974223784,
      "eval_f1": 0.01718213058419244,
      "eval_loss": 0.04748029261827469,
      "eval_mcc": 0.009529862152017439,
      "eval_precision": 0.38461538461538464,
      "eval_recall": 0.008787346221441126,
      "eval_runtime": 3.1605,
      "eval_samples_per_second": 540.101,
      "eval_steps_per_second": 17.086,
      "step": 3210
    },
    {
      "epoch": 16.0,
      "grad_norm": 0.8616418838500977,
      "learning_rate": 3.6044487477044986e-06,
      "loss": 0.0491,
      "step": 3424
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.664323374340949,
      "eval_f1": 0.017152658662092625,
      "eval_loss": 0.04751123487949371,
      "eval_mcc": 0.004592958330124466,
      "eval_precision": 0.35714285714285715,
      "eval_recall": 0.008787346221441126,
      "eval_runtime": 3.9087,
      "eval_samples_per_second": 436.72,
      "eval_steps_per_second": 13.815,
      "step": 3424
    },
    {
      "epoch": 17.0,
      "grad_norm": 0.38889187574386597,
      "learning_rate": 1.8022243738522493e-06,
      "loss": 0.0493,
      "step": 3638
    },
    {
      "epoch": 17.0,
      "eval_accuracy": 0.6649091974223784,
      "eval_f1": 0.01718213058419244,
      "eval_loss": 0.0475541353225708,
      "eval_mcc": 0.009529862152017439,
      "eval_precision": 0.38461538461538464,
      "eval_recall": 0.008787346221441126,
      "eval_runtime": 3.1517,
      "eval_samples_per_second": 541.62,
      "eval_steps_per_second": 17.134,
      "step": 3638
    },
    {
      "epoch": 18.0,
      "grad_norm": 0.3475455343723297,
      "learning_rate": 0.0,
      "loss": 0.0489,
      "step": 3852
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.6649091974223784,
      "eval_f1": 0.01718213058419244,
      "eval_loss": 0.047414712607860565,
      "eval_mcc": 0.009529862152017439,
      "eval_precision": 0.38461538461538464,
      "eval_recall": 0.008787346221441126,
      "eval_runtime": 3.2036,
      "eval_samples_per_second": 532.83,
      "eval_steps_per_second": 16.856,
      "step": 3852
    }
  ],
  "logging_steps": 500,
  "max_steps": 3852,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 18,
  "save_steps": 500,
  "total_flos": 9445961959920.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.06448750556013427,
    "learning_rate": 3.244003872934049e-05,
    "num_train_epochs": 18,
    "temperature": 34
  }
}