{
  "best_metric": 0.9408151870463428,
  "best_model_checkpoint": "/tmp/classification_phobertbase/checkpoint-5820",
  "epoch": 40.0,
  "eval_steps": 500,
  "global_step": 15520,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_f1": 0.9251814628699051,
      "eval_loss": 0.21701952815055847,
      "eval_runtime": 9.8556,
      "eval_samples_per_second": 363.447,
      "eval_steps_per_second": 5.682,
      "step": 388
    },
    {
      "epoch": 1.2886597938144329,
      "grad_norm": 4.1991682052612305,
      "learning_rate": 2.9033505154639176e-05,
      "loss": 0.3098,
      "step": 500
    },
    {
      "epoch": 2.0,
      "eval_f1": 0.9369067560022334,
      "eval_loss": 0.2119385302066803,
      "eval_runtime": 9.8764,
      "eval_samples_per_second": 362.683,
      "eval_steps_per_second": 5.67,
      "step": 776
    },
    {
      "epoch": 2.5773195876288657,
      "grad_norm": 6.048534870147705,
      "learning_rate": 2.806701030927835e-05,
      "loss": 0.1763,
      "step": 1000
    },
    {
      "epoch": 3.0,
      "eval_f1": 0.932998324958124,
      "eval_loss": 0.20233677327632904,
      "eval_runtime": 9.8515,
      "eval_samples_per_second": 363.598,
      "eval_steps_per_second": 5.684,
      "step": 1164
    },
    {
      "epoch": 3.865979381443299,
      "grad_norm": 3.9996299743652344,
      "learning_rate": 2.7100515463917526e-05,
      "loss": 0.1286,
      "step": 1500
    },
    {
      "epoch": 4.0,
      "eval_f1": 0.9343941931881632,
      "eval_loss": 0.26061517000198364,
      "eval_runtime": 9.8544,
      "eval_samples_per_second": 363.492,
      "eval_steps_per_second": 5.683,
      "step": 1552
    },
    {
      "epoch": 5.0,
      "eval_f1": 0.9391401451702959,
      "eval_loss": 0.23778466880321503,
      "eval_runtime": 9.8578,
      "eval_samples_per_second": 363.368,
      "eval_steps_per_second": 5.681,
      "step": 1940
    },
    {
      "epoch": 5.154639175257732,
      "grad_norm": 3.624183416366577,
      "learning_rate": 2.61340206185567e-05,
      "loss": 0.0943,
      "step": 2000
    },
    {
      "epoch": 6.0,
      "eval_f1": 0.9335566722501396,
      "eval_loss": 0.2650683522224426,
      "eval_runtime": 9.8594,
      "eval_samples_per_second": 363.308,
      "eval_steps_per_second": 5.68,
      "step": 2328
    },
    {
      "epoch": 6.443298969072165,
      "grad_norm": 7.170408248901367,
      "learning_rate": 2.5167525773195877e-05,
      "loss": 0.0706,
      "step": 2500
    },
    {
      "epoch": 7.0,
      "eval_f1": 0.9366275823562256,
      "eval_loss": 0.2985905408859253,
      "eval_runtime": 9.8626,
      "eval_samples_per_second": 363.191,
      "eval_steps_per_second": 5.678,
      "step": 2716
    },
    {
      "epoch": 7.731958762886598,
      "grad_norm": 2.5986905097961426,
      "learning_rate": 2.4201030927835052e-05,
      "loss": 0.0626,
      "step": 3000
    },
    {
      "epoch": 8.0,
      "eval_f1": 0.9304857621440537,
      "eval_loss": 0.31677183508872986,
      "eval_runtime": 9.879,
      "eval_samples_per_second": 362.589,
      "eval_steps_per_second": 5.669,
      "step": 3104
    },
    {
      "epoch": 9.0,
      "eval_f1": 0.9357900614182021,
      "eval_loss": 0.30201148986816406,
      "eval_runtime": 9.8829,
      "eval_samples_per_second": 362.444,
      "eval_steps_per_second": 5.666,
      "step": 3492
    },
    {
      "epoch": 9.02061855670103,
      "grad_norm": 0.1557622104883194,
      "learning_rate": 2.3234536082474227e-05,
      "loss": 0.0515,
      "step": 3500
    },
    {
      "epoch": 10.0,
      "eval_f1": 0.9366275823562256,
      "eval_loss": 0.30621805787086487,
      "eval_runtime": 9.8777,
      "eval_samples_per_second": 362.634,
      "eval_steps_per_second": 5.669,
      "step": 3880
    },
    {
      "epoch": 10.309278350515465,
      "grad_norm": 0.21442590653896332,
      "learning_rate": 2.2268041237113402e-05,
      "loss": 0.0397,
      "step": 4000
    },
    {
      "epoch": 11.0,
      "eval_f1": 0.9343941931881632,
      "eval_loss": 0.34871262311935425,
      "eval_runtime": 9.8759,
      "eval_samples_per_second": 362.7,
      "eval_steps_per_second": 5.67,
      "step": 4268
    },
    {
      "epoch": 11.597938144329897,
      "grad_norm": 13.442221641540527,
      "learning_rate": 2.1301546391752577e-05,
      "loss": 0.0337,
      "step": 4500
    },
    {
      "epoch": 12.0,
      "eval_f1": 0.9290898939140145,
      "eval_loss": 0.4043188989162445,
      "eval_runtime": 9.879,
      "eval_samples_per_second": 362.586,
      "eval_steps_per_second": 5.669,
      "step": 4656
    },
    {
      "epoch": 12.88659793814433,
      "grad_norm": 0.3749821186065674,
      "learning_rate": 2.0335051546391752e-05,
      "loss": 0.031,
      "step": 5000
    },
    {
      "epoch": 13.0,
      "eval_f1": 0.9366275823562256,
      "eval_loss": 0.37788987159729004,
      "eval_runtime": 9.8649,
      "eval_samples_per_second": 363.104,
      "eval_steps_per_second": 5.677,
      "step": 5044
    },
    {
      "epoch": 14.0,
      "eval_f1": 0.9293690675600224,
      "eval_loss": 0.39344093203544617,
      "eval_runtime": 9.8929,
      "eval_samples_per_second": 362.077,
      "eval_steps_per_second": 5.661,
      "step": 5432
    },
    {
      "epoch": 14.175257731958762,
      "grad_norm": 0.08398671448230743,
      "learning_rate": 1.936855670103093e-05,
      "loss": 0.0266,
      "step": 5500
    },
    {
      "epoch": 15.0,
      "eval_f1": 0.9408151870463428,
      "eval_loss": 0.36774319410324097,
      "eval_runtime": 9.8647,
      "eval_samples_per_second": 363.113,
      "eval_steps_per_second": 5.677,
      "step": 5820
    },
    {
      "epoch": 15.463917525773196,
      "grad_norm": 0.2722782790660858,
      "learning_rate": 1.8402061855670106e-05,
      "loss": 0.0246,
      "step": 6000
    },
    {
      "epoch": 16.0,
      "eval_f1": 0.9355108877721943,
      "eval_loss": 0.3874046206474304,
      "eval_runtime": 9.8701,
      "eval_samples_per_second": 362.913,
      "eval_steps_per_second": 5.674,
      "step": 6208
    },
    {
      "epoch": 16.75257731958763,
      "grad_norm": 8.601401329040527,
      "learning_rate": 1.743556701030928e-05,
      "loss": 0.0222,
      "step": 6500
    },
    {
      "epoch": 17.0,
      "eval_f1": 0.9343941931881632,
      "eval_loss": 0.42573627829551697,
      "eval_runtime": 9.8744,
      "eval_samples_per_second": 362.755,
      "eval_steps_per_second": 5.671,
      "step": 6596
    },
    {
      "epoch": 18.0,
      "eval_f1": 0.9369067560022334,
      "eval_loss": 0.4371758997440338,
      "eval_runtime": 9.8898,
      "eval_samples_per_second": 362.193,
      "eval_steps_per_second": 5.662,
      "step": 6984
    },
    {
      "epoch": 18.04123711340206,
      "grad_norm": 0.05734695494174957,
      "learning_rate": 1.6469072164948456e-05,
      "loss": 0.022,
      "step": 7000
    },
    {
      "epoch": 19.0,
      "eval_f1": 0.9363484087102177,
      "eval_loss": 0.44078370928764343,
      "eval_runtime": 9.8828,
      "eval_samples_per_second": 362.448,
      "eval_steps_per_second": 5.666,
      "step": 7372
    },
    {
      "epoch": 19.329896907216494,
      "grad_norm": 0.19072239100933075,
      "learning_rate": 1.550257731958763e-05,
      "loss": 0.0176,
      "step": 7500
    },
    {
      "epoch": 20.0,
      "eval_f1": 0.9357900614182021,
      "eval_loss": 0.4601401090621948,
      "eval_runtime": 9.8712,
      "eval_samples_per_second": 362.874,
      "eval_steps_per_second": 5.673,
      "step": 7760
    },
    {
      "epoch": 20.61855670103093,
      "grad_norm": 0.14792193472385406,
      "learning_rate": 1.4536082474226805e-05,
      "loss": 0.0142,
      "step": 8000
    },
    {
      "epoch": 21.0,
      "eval_f1": 0.9360692350642099,
      "eval_loss": 0.45033514499664307,
      "eval_runtime": 9.8715,
      "eval_samples_per_second": 362.861,
      "eval_steps_per_second": 5.673,
      "step": 8148
    },
    {
      "epoch": 21.90721649484536,
      "grad_norm": 0.03246806561946869,
      "learning_rate": 1.356958762886598e-05,
      "loss": 0.0134,
      "step": 8500
    },
    {
      "epoch": 22.0,
      "eval_f1": 0.9366275823562256,
      "eval_loss": 0.4835437834262848,
      "eval_runtime": 9.8869,
      "eval_samples_per_second": 362.298,
      "eval_steps_per_second": 5.664,
      "step": 8536
    },
    {
      "epoch": 23.0,
      "eval_f1": 0.9391401451702959,
      "eval_loss": 0.45941534638404846,
      "eval_runtime": 9.8785,
      "eval_samples_per_second": 362.605,
      "eval_steps_per_second": 5.669,
      "step": 8924
    },
    {
      "epoch": 23.195876288659793,
      "grad_norm": 0.08111796528100967,
      "learning_rate": 1.2603092783505155e-05,
      "loss": 0.0126,
      "step": 9000
    },
    {
      "epoch": 24.0,
      "eval_f1": 0.9366275823562256,
      "eval_loss": 0.48089537024497986,
      "eval_runtime": 9.8641,
      "eval_samples_per_second": 363.135,
      "eval_steps_per_second": 5.677,
      "step": 9312
    },
    {
      "epoch": 24.484536082474225,
      "grad_norm": 0.06357069313526154,
      "learning_rate": 1.163659793814433e-05,
      "loss": 0.0109,
      "step": 9500
    },
    {
      "epoch": 25.0,
      "eval_f1": 0.9369067560022334,
      "eval_loss": 0.48593300580978394,
      "eval_runtime": 9.8681,
      "eval_samples_per_second": 362.988,
      "eval_steps_per_second": 5.675,
      "step": 9700
    },
    {
      "epoch": 25.77319587628866,
      "grad_norm": 0.006769911386072636,
      "learning_rate": 1.0670103092783506e-05,
      "loss": 0.012,
      "step": 10000
    },
    {
      "epoch": 26.0,
      "eval_f1": 0.9385817978782803,
      "eval_loss": 0.48239514231681824,
      "eval_runtime": 9.8778,
      "eval_samples_per_second": 362.631,
      "eval_steps_per_second": 5.669,
      "step": 10088
    },
    {
      "epoch": 27.0,
      "eval_f1": 0.9360692350642099,
      "eval_loss": 0.5067496299743652,
      "eval_runtime": 9.8844,
      "eval_samples_per_second": 362.39,
      "eval_steps_per_second": 5.666,
      "step": 10476
    },
    {
      "epoch": 27.061855670103093,
      "grad_norm": 0.037592578679323196,
      "learning_rate": 9.703608247422681e-06,
      "loss": 0.0101,
      "step": 10500
    },
    {
      "epoch": 28.0,
      "eval_f1": 0.937465103294249,
      "eval_loss": 0.48702046275138855,
      "eval_runtime": 9.8977,
      "eval_samples_per_second": 361.901,
      "eval_steps_per_second": 5.658,
      "step": 10864
    },
    {
      "epoch": 28.350515463917525,
      "grad_norm": 0.12630492448806763,
      "learning_rate": 8.737113402061856e-06,
      "loss": 0.0102,
      "step": 11000
    },
    {
      "epoch": 29.0,
      "eval_f1": 0.9355108877721943,
      "eval_loss": 0.5301576256752014,
      "eval_runtime": 9.8877,
      "eval_samples_per_second": 362.268,
      "eval_steps_per_second": 5.664,
      "step": 11252
    },
    {
      "epoch": 29.63917525773196,
      "grad_norm": 0.2997748851776123,
      "learning_rate": 7.770618556701031e-06,
      "loss": 0.0088,
      "step": 11500
    },
    {
      "epoch": 30.0,
      "eval_f1": 0.9366275823562256,
      "eval_loss": 0.49529990553855896,
      "eval_runtime": 9.8876,
      "eval_samples_per_second": 362.27,
      "eval_steps_per_second": 5.664,
      "step": 11640
    },
    {
      "epoch": 30.927835051546392,
      "grad_norm": 0.33649611473083496,
      "learning_rate": 6.8041237113402065e-06,
      "loss": 0.0093,
      "step": 12000
    },
    {
      "epoch": 31.0,
      "eval_f1": 0.9360692350642099,
      "eval_loss": 0.4913821220397949,
      "eval_runtime": 9.8622,
      "eval_samples_per_second": 363.206,
      "eval_steps_per_second": 5.678,
      "step": 12028
    },
    {
      "epoch": 32.0,
      "eval_f1": 0.9388609715242882,
      "eval_loss": 0.5014358162879944,
      "eval_runtime": 9.8775,
      "eval_samples_per_second": 362.641,
      "eval_steps_per_second": 5.669,
      "step": 12416
    },
    {
      "epoch": 32.21649484536083,
      "grad_norm": 0.011193674989044666,
      "learning_rate": 5.837628865979382e-06,
      "loss": 0.0084,
      "step": 12500
    },
    {
      "epoch": 33.0,
      "eval_f1": 0.9383026242322725,
      "eval_loss": 0.5026176571846008,
      "eval_runtime": 9.8893,
      "eval_samples_per_second": 362.21,
      "eval_steps_per_second": 5.663,
      "step": 12804
    },
    {
      "epoch": 33.50515463917526,
      "grad_norm": 0.26669424772262573,
      "learning_rate": 4.871134020618557e-06,
      "loss": 0.0078,
      "step": 13000
    },
    {
      "epoch": 34.0,
      "eval_f1": 0.9380234505862647,
      "eval_loss": 0.5043131709098816,
      "eval_runtime": 9.8862,
      "eval_samples_per_second": 362.323,
      "eval_steps_per_second": 5.664,
      "step": 13192
    },
    {
      "epoch": 34.79381443298969,
      "grad_norm": 0.9884344339370728,
      "learning_rate": 3.904639175257732e-06,
      "loss": 0.0075,
      "step": 13500
    },
    {
      "epoch": 35.0,
      "eval_f1": 0.9377442769402569,
      "eval_loss": 0.5035226941108704,
      "eval_runtime": 9.8894,
      "eval_samples_per_second": 362.204,
      "eval_steps_per_second": 5.663,
      "step": 13580
    },
    {
      "epoch": 36.0,
      "eval_f1": 0.9377442769402569,
      "eval_loss": 0.5007095336914062,
      "eval_runtime": 9.9011,
      "eval_samples_per_second": 361.777,
      "eval_steps_per_second": 5.656,
      "step": 13968
    },
    {
      "epoch": 36.08247422680412,
      "grad_norm": 0.02359125204384327,
      "learning_rate": 2.938144329896907e-06,
      "loss": 0.0077,
      "step": 14000
    },
    {
      "epoch": 37.0,
      "eval_f1": 0.9377442769402569,
      "eval_loss": 0.5102269053459167,
      "eval_runtime": 9.8918,
      "eval_samples_per_second": 362.118,
      "eval_steps_per_second": 5.661,
      "step": 14356
    },
    {
      "epoch": 37.371134020618555,
      "grad_norm": 0.00852921698242426,
      "learning_rate": 1.9716494845360827e-06,
      "loss": 0.0076,
      "step": 14500
    },
    {
      "epoch": 38.0,
      "eval_f1": 0.9391401451702959,
      "eval_loss": 0.5068992376327515,
      "eval_runtime": 9.8982,
      "eval_samples_per_second": 361.883,
      "eval_steps_per_second": 5.658,
      "step": 14744
    },
    {
      "epoch": 38.65979381443299,
      "grad_norm": 0.01492872554808855,
      "learning_rate": 1.0051546391752577e-06,
      "loss": 0.0059,
      "step": 15000
    },
    {
      "epoch": 39.0,
      "eval_f1": 0.9385817978782803,
      "eval_loss": 0.5104745030403137,
      "eval_runtime": 9.9017,
      "eval_samples_per_second": 361.758,
      "eval_steps_per_second": 5.656,
      "step": 15132
    },
    {
      "epoch": 39.94845360824742,
      "grad_norm": 0.013034285046160221,
      "learning_rate": 3.865979381443299e-08,
      "loss": 0.0071,
      "step": 15500
    },
    {
      "epoch": 40.0,
      "eval_f1": 0.9383026242322725,
      "eval_loss": 0.5110998749732971,
      "eval_runtime": 9.8692,
      "eval_samples_per_second": 362.949,
      "eval_steps_per_second": 5.674,
      "step": 15520
    },
    {
      "epoch": 40.0,
      "step": 15520,
      "total_flos": 6.517582470540288e+16,
      "train_loss": 0.040758436324900574,
      "train_runtime": 8835.4698,
      "train_samples_per_second": 112.143,
      "train_steps_per_second": 1.757
    }
  ],
  "logging_steps": 500,
  "max_steps": 15520,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 40,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 6.517582470540288e+16,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}