{
  "best_metric": 0.7610168782310045,
  "best_model_checkpoint": "tiny-llama-lora-no-grad/checkpoint-1286",
  "epoch": 14.0,
  "eval_steps": 500,
  "global_step": 9002,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.78,
      "learning_rate": 4.740798341109383e-05,
      "loss": 1.1276,
      "step": 500
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8086754453911696,
      "eval_f1_macro": 0.6806109083601769,
      "eval_f1_micro": 0.8086754453911695,
      "eval_f1_weighted": 0.8034368965438116,
      "eval_loss": 0.6704964637756348,
      "eval_macro_fpr": 0.017184987145855975,
      "eval_macro_sensitivity": 0.6852607939619662,
      "eval_macro_specificity": 0.985525532858602,
      "eval_precision": 0.8055191978891799,
      "eval_precision_macro": 0.7052877886582329,
      "eval_recall": 0.8086754453911696,
      "eval_recall_macro": 0.6852607939619662,
      "eval_runtime": 225.0756,
      "eval_samples_per_second": 5.736,
      "eval_steps_per_second": 0.72,
      "eval_weighted_fpr": 0.016618448496265894,
      "eval_weighted_sensitivity": 0.8086754453911696,
      "eval_weighted_specificity": 0.9742075474878604,
      "step": 643
    },
    {
      "epoch": 1.56,
      "learning_rate": 4.481596682218767e-05,
      "loss": 0.503,
      "step": 1000
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8164213787761425,
      "eval_f1_macro": 0.7610168782310045,
      "eval_f1_micro": 0.8164213787761425,
      "eval_f1_weighted": 0.8154049040588042,
      "eval_loss": 0.7205804586410522,
      "eval_macro_fpr": 0.01631511703260385,
      "eval_macro_sensitivity": 0.7641451876525827,
      "eval_macro_specificity": 0.9862465922523334,
      "eval_precision": 0.8231306684180838,
      "eval_precision_macro": 0.7746203157296797,
      "eval_recall": 0.8164213787761425,
      "eval_recall_macro": 0.7641451876525827,
      "eval_runtime": 230.0342,
      "eval_samples_per_second": 5.612,
      "eval_steps_per_second": 0.704,
      "eval_weighted_fpr": 0.01580737677582872,
      "eval_weighted_sensitivity": 0.8164213787761425,
      "eval_weighted_specificity": 0.977277505008855,
      "step": 1286
    },
    {
      "epoch": 2.33,
      "learning_rate": 4.22239502332815e-05,
      "loss": 0.3617,
      "step": 1500
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8164213787761425,
      "eval_f1_macro": 0.7242203392667825,
      "eval_f1_micro": 0.8164213787761425,
      "eval_f1_weighted": 0.8123838748630571,
      "eval_loss": 0.8818831443786621,
      "eval_macro_fpr": 0.01641791191580845,
      "eval_macro_sensitivity": 0.7170126291644698,
      "eval_macro_specificity": 0.9861058068231509,
      "eval_precision": 0.8137334709054306,
      "eval_precision_macro": 0.7499445969412153,
      "eval_recall": 0.8164213787761425,
      "eval_recall_macro": 0.7170126291644698,
      "eval_runtime": 231.2189,
      "eval_samples_per_second": 5.583,
      "eval_steps_per_second": 0.701,
      "eval_weighted_fpr": 0.01580737677582872,
      "eval_weighted_sensitivity": 0.8164213787761425,
      "eval_weighted_specificity": 0.975165723571122,
      "step": 1929
    },
    {
      "epoch": 3.11,
      "learning_rate": 3.963193364437533e-05,
      "loss": 0.2293,
      "step": 2000
    },
    {
      "epoch": 3.89,
      "learning_rate": 3.7039917055469156e-05,
      "loss": 0.0618,
      "step": 2500
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8086754453911696,
      "eval_f1_macro": 0.740118070206755,
      "eval_f1_micro": 0.8086754453911695,
      "eval_f1_weighted": 0.8074036218868607,
      "eval_loss": 1.1433583498001099,
      "eval_macro_fpr": 0.017252929817186225,
      "eval_macro_sensitivity": 0.729298059210823,
      "eval_macro_specificity": 0.9854230690678942,
      "eval_precision": 0.8106816369031802,
      "eval_precision_macro": 0.7673360060599891,
      "eval_recall": 0.8086754453911696,
      "eval_recall_macro": 0.729298059210823,
      "eval_runtime": 230.3503,
      "eval_samples_per_second": 5.605,
      "eval_steps_per_second": 0.703,
      "eval_weighted_fpr": 0.016618448496265894,
      "eval_weighted_sensitivity": 0.8086754453911696,
      "eval_weighted_specificity": 0.972670590627246,
      "step": 2572
    },
    {
      "epoch": 4.67,
      "learning_rate": 3.4447900466562985e-05,
      "loss": 0.0243,
      "step": 3000
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8109992254066615,
      "eval_f1_macro": 0.7228040111858681,
      "eval_f1_micro": 0.8109992254066615,
      "eval_f1_weighted": 0.8085738170522284,
      "eval_loss": 1.2966285943984985,
      "eval_macro_fpr": 0.017057469386588052,
      "eval_macro_sensitivity": 0.7164376871359707,
      "eval_macro_specificity": 0.9857584427734511,
      "eval_precision": 0.8111777551016551,
      "eval_precision_macro": 0.748883500821637,
      "eval_recall": 0.8109992254066615,
      "eval_recall_macro": 0.7164376871359707,
      "eval_runtime": 230.5087,
      "eval_samples_per_second": 5.601,
      "eval_steps_per_second": 0.703,
      "eval_weighted_fpr": 0.016373641121997046,
      "eval_weighted_sensitivity": 0.8109992254066615,
      "eval_weighted_specificity": 0.9753774161951063,
      "step": 3215
    },
    {
      "epoch": 5.44,
      "learning_rate": 3.1855883877656815e-05,
      "loss": 0.0121,
      "step": 3500
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8195197521301317,
      "eval_f1_macro": 0.714317091817089,
      "eval_f1_micro": 0.8195197521301317,
      "eval_f1_weighted": 0.8169833922452127,
      "eval_loss": 1.2964715957641602,
      "eval_macro_fpr": 0.01621298630582096,
      "eval_macro_sensitivity": 0.7076912641211482,
      "eval_macro_specificity": 0.9863159800251362,
      "eval_precision": 0.8174973720287305,
      "eval_precision_macro": 0.7311701256978252,
      "eval_recall": 0.8195197521301317,
      "eval_recall_macro": 0.7076912641211482,
      "eval_runtime": 230.8554,
      "eval_samples_per_second": 5.592,
      "eval_steps_per_second": 0.702,
      "eval_weighted_fpr": 0.01548687271518777,
      "eval_weighted_sensitivity": 0.8195197521301317,
      "eval_weighted_specificity": 0.9752199482469127,
      "step": 3858
    },
    {
      "epoch": 6.22,
      "learning_rate": 2.926386728875065e-05,
      "loss": 0.003,
      "step": 4000
    },
    {
      "epoch": 7.0,
      "learning_rate": 2.667185069984448e-05,
      "loss": 0.0021,
      "step": 4500
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8187451587916343,
      "eval_f1_macro": 0.7165425493481726,
      "eval_f1_micro": 0.8187451587916342,
      "eval_f1_weighted": 0.8152226658384369,
      "eval_loss": 1.3710469007492065,
      "eval_macro_fpr": 0.01622856042065577,
      "eval_macro_sensitivity": 0.7112104764224284,
      "eval_macro_specificity": 0.9862882906960665,
      "eval_precision": 0.8168122443238953,
      "eval_precision_macro": 0.7519067055100973,
      "eval_recall": 0.8187451587916343,
      "eval_recall_macro": 0.7112104764224284,
      "eval_runtime": 231.4872,
      "eval_samples_per_second": 5.577,
      "eval_steps_per_second": 0.7,
      "eval_weighted_fpr": 0.015566790846194785,
      "eval_weighted_sensitivity": 0.8187451587916343,
      "eval_weighted_specificity": 0.975579201649363,
      "step": 4501
    },
    {
      "epoch": 7.78,
      "learning_rate": 2.4079834110938313e-05,
      "loss": 0.003,
      "step": 5000
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.820294345468629,
      "eval_f1_macro": 0.7158880722332298,
      "eval_f1_micro": 0.8202943454686291,
      "eval_f1_weighted": 0.8173423019940304,
      "eval_loss": 1.3348430395126343,
      "eval_macro_fpr": 0.0161589741015493,
      "eval_macro_sensitivity": 0.707324528826163,
      "eval_macro_specificity": 0.9863452414325693,
      "eval_precision": 0.817084329636889,
      "eval_precision_macro": 0.7416566900266442,
      "eval_recall": 0.820294345468629,
      "eval_recall_macro": 0.707324528826163,
      "eval_runtime": 231.7727,
      "eval_samples_per_second": 5.57,
      "eval_steps_per_second": 0.699,
      "eval_weighted_fpr": 0.015407092575375215,
      "eval_weighted_sensitivity": 0.820294345468629,
      "eval_weighted_specificity": 0.97488427601991,
      "step": 5144
    },
    {
      "epoch": 8.55,
      "learning_rate": 2.1487817522032143e-05,
      "loss": 0.0023,
      "step": 5500
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.8187451587916343,
      "eval_f1_macro": 0.7120780493516332,
      "eval_f1_micro": 0.8187451587916342,
      "eval_f1_weighted": 0.8141492515633661,
      "eval_loss": 1.4038118124008179,
      "eval_macro_fpr": 0.01630132436230029,
      "eval_macro_sensitivity": 0.7030162850725371,
      "eval_macro_specificity": 0.9861967008390129,
      "eval_precision": 0.8149366971498379,
      "eval_precision_macro": 0.7548435802187925,
      "eval_recall": 0.8187451587916343,
      "eval_recall_macro": 0.7030162850725371,
      "eval_runtime": 231.6565,
      "eval_samples_per_second": 5.573,
      "eval_steps_per_second": 0.699,
      "eval_weighted_fpr": 0.015566790846194785,
      "eval_weighted_sensitivity": 0.8187451587916343,
      "eval_weighted_specificity": 0.9742053537935573,
      "step": 5787
    },
    {
      "epoch": 9.33,
      "learning_rate": 1.8895800933125972e-05,
      "loss": 0.0033,
      "step": 6000
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.820294345468629,
      "eval_f1_macro": 0.7152055882276366,
      "eval_f1_micro": 0.8202943454686291,
      "eval_f1_weighted": 0.8163162253560097,
      "eval_loss": 1.4021339416503906,
      "eval_macro_fpr": 0.01619493341866027,
      "eval_macro_sensitivity": 0.7110217324819302,
      "eval_macro_specificity": 0.9863270327277461,
      "eval_precision": 0.8151084284815174,
      "eval_precision_macro": 0.7330251485200666,
      "eval_recall": 0.820294345468629,
      "eval_recall_macro": 0.7110217324819302,
      "eval_runtime": 231.6588,
      "eval_samples_per_second": 5.573,
      "eval_steps_per_second": 0.699,
      "eval_weighted_fpr": 0.015407092575375215,
      "eval_weighted_sensitivity": 0.820294345468629,
      "eval_weighted_specificity": 0.9746111454475601,
      "step": 6430
    },
    {
      "epoch": 10.11,
      "learning_rate": 1.6303784344219805e-05,
      "loss": 0.0014,
      "step": 6500
    },
    {
      "epoch": 10.89,
      "learning_rate": 1.3711767755313634e-05,
      "loss": 0.0017,
      "step": 7000
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.8210689388071263,
      "eval_f1_macro": 0.7155389432371102,
      "eval_f1_micro": 0.8210689388071263,
      "eval_f1_weighted": 0.8179308294442692,
      "eval_loss": 1.400085210800171,
      "eval_macro_fpr": 0.01600471707010785,
      "eval_macro_sensitivity": 0.7110426238860363,
      "eval_macro_specificity": 0.9864226161631717,
      "eval_precision": 0.8177607468470947,
      "eval_precision_macro": 0.736115805426312,
      "eval_recall": 0.8210689388071263,
      "eval_recall_macro": 0.7110426238860363,
      "eval_runtime": 233.5847,
      "eval_samples_per_second": 5.527,
      "eval_steps_per_second": 0.694,
      "eval_weighted_fpr": 0.015327450069670227,
      "eval_weighted_sensitivity": 0.8210689388071263,
      "eval_weighted_specificity": 0.975270303640449,
      "step": 7073
    },
    {
      "epoch": 11.66,
      "learning_rate": 1.1119751166407467e-05,
      "loss": 0.0023,
      "step": 7500
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.8226181254841208,
      "eval_f1_macro": 0.7177170991829454,
      "eval_f1_micro": 0.8226181254841208,
      "eval_f1_weighted": 0.8194785693417274,
      "eval_loss": 1.409959316253662,
      "eval_macro_fpr": 0.015849560370098895,
      "eval_macro_sensitivity": 0.7127452360706031,
      "eval_macro_specificity": 0.9865319770574401,
      "eval_precision": 0.8189327832635585,
      "eval_precision_macro": 0.7386070644276358,
      "eval_recall": 0.8226181254841208,
      "eval_recall_macro": 0.7127452360706031,
      "eval_runtime": 231.4046,
      "eval_samples_per_second": 5.579,
      "eval_steps_per_second": 0.7,
      "eval_weighted_fpr": 0.01516857653838511,
      "eval_weighted_sensitivity": 0.8226181254841208,
      "eval_weighted_specificity": 0.9753615303774794,
      "step": 7716
    },
    {
      "epoch": 12.44,
      "learning_rate": 8.527734577501296e-06,
      "loss": 0.0034,
      "step": 8000
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.8233927188226181,
      "eval_f1_macro": 0.7171044212306599,
      "eval_f1_micro": 0.8233927188226181,
      "eval_f1_weighted": 0.8200624957363772,
      "eval_loss": 1.4272679090499878,
      "eval_macro_fpr": 0.015762565967900863,
      "eval_macro_sensitivity": 0.7114968838472958,
      "eval_macro_specificity": 0.9866030749234862,
      "eval_precision": 0.8192001454232278,
      "eval_precision_macro": 0.7384958049410225,
      "eval_recall": 0.8233927188226181,
      "eval_recall_macro": 0.7114968838472958,
      "eval_runtime": 231.5396,
      "eval_samples_per_second": 5.576,
      "eval_steps_per_second": 0.7,
      "eval_weighted_fpr": 0.015089344804765056,
      "eval_weighted_sensitivity": 0.8233927188226181,
      "eval_weighted_specificity": 0.9756534050296731,
      "step": 8359
    },
    {
      "epoch": 13.22,
      "learning_rate": 5.935717988595127e-06,
      "loss": 0.0,
      "step": 8500
    },
    {
      "epoch": 14.0,
      "learning_rate": 3.343701399688958e-06,
      "loss": 0.0016,
      "step": 9000
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.8226181254841208,
      "eval_f1_macro": 0.7167677155595502,
      "eval_f1_micro": 0.8226181254841208,
      "eval_f1_weighted": 0.819203323139996,
      "eval_loss": 1.4322476387023926,
      "eval_macro_fpr": 0.015856078619640734,
      "eval_macro_sensitivity": 0.7111285597220655,
      "eval_macro_specificity": 0.9865346286538078,
      "eval_precision": 0.8182791715655534,
      "eval_precision_macro": 0.7381978217210116,
      "eval_recall": 0.8226181254841208,
      "eval_recall_macro": 0.7111285597220655,
      "eval_runtime": 230.5492,
      "eval_samples_per_second": 5.6,
      "eval_steps_per_second": 0.703,
      "eval_weighted_fpr": 0.01516857653838511,
      "eval_weighted_sensitivity": 0.8226181254841208,
      "eval_weighted_specificity": 0.975401304322996,
      "step": 9002
    }
  ],
  "logging_steps": 500,
  "max_steps": 9645,
  "num_train_epochs": 15,
  "save_steps": 500,
  "total_flos": 2.169209851383644e+17,
  "trial_name": null,
  "trial_params": null
}