{
  "best_metric": 0.7217340363210311,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-5/checkpoint-2782",
  "epoch": 13.0,
  "eval_steps": 500,
  "global_step": 2782,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 2.161219835281372,
      "learning_rate": 0.0001946740481873714,
      "loss": 0.2896,
      "step": 214
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6666666666666666,
      "eval_f1": 0.0,
      "eval_loss": 0.27429890632629395,
      "eval_mcc": 0.0,
      "eval_precision": 0.0,
      "eval_recall": 0.0,
      "eval_runtime": 3.1537,
      "eval_samples_per_second": 541.276,
      "eval_steps_per_second": 17.123,
      "step": 214
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.0970500707626343,
      "learning_rate": 0.00017969912140372742,
      "loss": 0.2734,
      "step": 428
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6666666666666666,
      "eval_f1": 0.006980802792321117,
      "eval_loss": 0.2710207402706146,
      "eval_mcc": 0.01713474628469157,
      "eval_precision": 0.5,
      "eval_recall": 0.0035149384885764497,
      "eval_runtime": 3.9301,
      "eval_samples_per_second": 434.342,
      "eval_steps_per_second": 13.74,
      "step": 428
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.7171841859817505,
      "learning_rate": 0.0001647241946200835,
      "loss": 0.2685,
      "step": 642
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6678383128295254,
      "eval_f1": 0.010471204188481676,
      "eval_loss": 0.27606382966041565,
      "eval_mcc": 0.042836865711728934,
      "eval_precision": 0.75,
      "eval_recall": 0.005272407732864675,
      "eval_runtime": 3.1441,
      "eval_samples_per_second": 542.926,
      "eval_steps_per_second": 17.175,
      "step": 642
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.334978699684143,
      "learning_rate": 0.00014974926783643954,
      "loss": 0.266,
      "step": 856
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6795547744581136,
      "eval_f1": 0.10180623973727422,
      "eval_loss": 0.26485475897789,
      "eval_mcc": 0.14513196526792949,
      "eval_precision": 0.775,
      "eval_recall": 0.054481546572934976,
      "eval_runtime": 3.1938,
      "eval_samples_per_second": 534.478,
      "eval_steps_per_second": 16.908,
      "step": 856
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.377930998802185,
      "learning_rate": 0.0001347743410527956,
      "loss": 0.2643,
      "step": 1070
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.6918570591681312,
      "eval_f1": 0.18827160493827164,
      "eval_loss": 0.26378217339515686,
      "eval_mcc": 0.20505841470507494,
      "eval_precision": 0.7721518987341772,
      "eval_recall": 0.10720562390158173,
      "eval_runtime": 3.1292,
      "eval_samples_per_second": 545.505,
      "eval_steps_per_second": 17.257,
      "step": 1070
    },
    {
      "epoch": 6.0,
      "grad_norm": 1.2771140336990356,
      "learning_rate": 0.00011979941426915163,
      "loss": 0.263,
      "step": 1284
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.6936145284124194,
      "eval_f1": 0.18662519440124417,
      "eval_loss": 0.26149189472198486,
      "eval_mcc": 0.2156164618376391,
      "eval_precision": 0.8108108108108109,
      "eval_recall": 0.1054481546572935,
      "eval_runtime": 3.2914,
      "eval_samples_per_second": 518.626,
      "eval_steps_per_second": 16.406,
      "step": 1284
    },
    {
      "epoch": 7.0,
      "grad_norm": 1.073453426361084,
      "learning_rate": 0.00010482448748550767,
      "loss": 0.2612,
      "step": 1498
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.6994727592267135,
      "eval_f1": 0.21439509954058195,
      "eval_loss": 0.2620287537574768,
      "eval_mcc": 0.24129962353457945,
      "eval_precision": 0.8333333333333334,
      "eval_recall": 0.12302284710017575,
      "eval_runtime": 3.1567,
      "eval_samples_per_second": 540.751,
      "eval_steps_per_second": 17.106,
      "step": 1498
    },
    {
      "epoch": 8.0,
      "grad_norm": 1.2691621780395508,
      "learning_rate": 8.984956070186371e-05,
      "loss": 0.2597,
      "step": 1712
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.69302870533099,
      "eval_f1": 0.17088607594936708,
      "eval_loss": 0.2611652910709381,
      "eval_mcc": 0.21751991027491313,
      "eval_precision": 0.8571428571428571,
      "eval_recall": 0.09490333919156414,
      "eval_runtime": 3.2468,
      "eval_samples_per_second": 525.752,
      "eval_steps_per_second": 16.632,
      "step": 1712
    },
    {
      "epoch": 9.0,
      "grad_norm": 1.0226393938064575,
      "learning_rate": 7.487463391821977e-05,
      "loss": 0.2597,
      "step": 1926
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.6977152899824253,
      "eval_f1": 0.19626168224299068,
      "eval_loss": 0.2611730098724365,
      "eval_mcc": 0.2374955820778862,
      "eval_precision": 0.863013698630137,
      "eval_recall": 0.11072056239015818,
      "eval_runtime": 3.1639,
      "eval_samples_per_second": 539.53,
      "eval_steps_per_second": 17.068,
      "step": 1926
    },
    {
      "epoch": 10.0,
      "grad_norm": 1.0377492904663086,
      "learning_rate": 5.989970713457581e-05,
      "loss": 0.2565,
      "step": 2140
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.7193907439953134,
      "eval_f1": 0.32248939179632247,
      "eval_loss": 0.260220468044281,
      "eval_mcc": 0.310001756502818,
      "eval_precision": 0.8260869565217391,
      "eval_recall": 0.20035149384885764,
      "eval_runtime": 3.2066,
      "eval_samples_per_second": 532.341,
      "eval_steps_per_second": 16.84,
      "step": 2140
    },
    {
      "epoch": 11.0,
      "grad_norm": 1.2514437437057495,
      "learning_rate": 4.4924780350931855e-05,
      "loss": 0.2555,
      "step": 2354
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.700058582308143,
      "eval_f1": 0.20743034055727552,
      "eval_loss": 0.26075002551078796,
      "eval_mcc": 0.2474956228703306,
      "eval_precision": 0.8701298701298701,
      "eval_recall": 0.11775043936731107,
      "eval_runtime": 3.1394,
      "eval_samples_per_second": 543.734,
      "eval_steps_per_second": 17.201,
      "step": 2354
    },
    {
      "epoch": 12.0,
      "grad_norm": 0.8049026727676392,
      "learning_rate": 2.9949853567287906e-05,
      "loss": 0.2544,
      "step": 2568
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.715875805506737,
      "eval_f1": 0.31593794076163606,
      "eval_loss": 0.2587771415710449,
      "eval_mcc": 0.29589835954792404,
      "eval_precision": 0.8,
      "eval_recall": 0.1968365553602812,
      "eval_runtime": 3.3501,
      "eval_samples_per_second": 509.539,
      "eval_steps_per_second": 16.119,
      "step": 2568
    },
    {
      "epoch": 13.0,
      "grad_norm": 2.94110369682312,
      "learning_rate": 1.4974926783643953e-05,
      "loss": 0.2544,
      "step": 2782
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.7217340363210311,
      "eval_f1": 0.3356643356643356,
      "eval_loss": 0.2589167356491089,
      "eval_mcc": 0.31697199705587376,
      "eval_precision": 0.821917808219178,
      "eval_recall": 0.210896309314587,
      "eval_runtime": 3.165,
      "eval_samples_per_second": 539.337,
      "eval_steps_per_second": 17.062,
      "step": 2782
    }
  ],
  "logging_steps": 500,
  "max_steps": 2996,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 14,
  "save_steps": 500,
  "total_flos": 6822083637720.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.43581262355237016,
    "learning_rate": 0.00020964897497101535,
    "num_train_epochs": 14,
    "temperature": 35
  }
}