{
  "best_metric": 0.8181082820112143,
  "best_model_checkpoint": "modernbert-medical-classifier/checkpoint-1196",
  "epoch": 25.0,
  "eval_steps": 500,
  "global_step": 2300,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_f1": 0.5993395707209687,
      "eval_loss": 0.6179381608963013,
      "eval_runtime": 7.1172,
      "eval_samples_per_second": 12.927,
      "eval_steps_per_second": 3.232,
      "step": 92
    },
    {
      "epoch": 1.0869565217391304,
      "grad_norm": 3.5607712268829346,
      "learning_rate": 4.782608695652174e-05,
      "loss": 0.7932,
      "step": 100
    },
    {
      "epoch": 2.0,
      "eval_f1": 0.5993395707209687,
      "eval_loss": 0.6113200187683105,
      "eval_runtime": 7.1729,
      "eval_samples_per_second": 12.826,
      "eval_steps_per_second": 3.206,
      "step": 184
    },
    {
      "epoch": 2.1739130434782608,
      "grad_norm": 1.3359662294387817,
      "learning_rate": 4.565217391304348e-05,
      "loss": 0.6589,
      "step": 200
    },
    {
      "epoch": 3.0,
      "eval_f1": 0.6304912478825522,
      "eval_loss": 0.6631842851638794,
      "eval_runtime": 7.1467,
      "eval_samples_per_second": 12.873,
      "eval_steps_per_second": 3.218,
      "step": 276
    },
    {
      "epoch": 3.260869565217391,
      "grad_norm": 3.4564907550811768,
      "learning_rate": 4.347826086956522e-05,
      "loss": 0.6418,
      "step": 300
    },
    {
      "epoch": 4.0,
      "eval_f1": 0.5993395707209687,
      "eval_loss": 0.6268433928489685,
      "eval_runtime": 7.1588,
      "eval_samples_per_second": 12.851,
      "eval_steps_per_second": 3.213,
      "step": 368
    },
    {
      "epoch": 4.3478260869565215,
      "grad_norm": 2.798123836517334,
      "learning_rate": 4.130434782608696e-05,
      "loss": 0.6315,
      "step": 400
    },
    {
      "epoch": 5.0,
      "eval_f1": 0.25077639751552794,
      "eval_loss": 0.9374740719795227,
      "eval_runtime": 7.2024,
      "eval_samples_per_second": 12.773,
      "eval_steps_per_second": 3.193,
      "step": 460
    },
    {
      "epoch": 5.434782608695652,
      "grad_norm": 3.6863114833831787,
      "learning_rate": 3.91304347826087e-05,
      "loss": 0.6522,
      "step": 500
    },
    {
      "epoch": 6.0,
      "eval_f1": 0.588628762541806,
      "eval_loss": 0.582027792930603,
      "eval_runtime": 7.2495,
      "eval_samples_per_second": 12.691,
      "eval_steps_per_second": 3.173,
      "step": 552
    },
    {
      "epoch": 6.521739130434782,
      "grad_norm": 5.186334133148193,
      "learning_rate": 3.695652173913043e-05,
      "loss": 0.6183,
      "step": 600
    },
    {
      "epoch": 7.0,
      "eval_f1": 0.6857289002557545,
      "eval_loss": 0.5537915825843811,
      "eval_runtime": 7.1555,
      "eval_samples_per_second": 12.857,
      "eval_steps_per_second": 3.214,
      "step": 644
    },
    {
      "epoch": 7.608695652173913,
      "grad_norm": 11.184236526489258,
      "learning_rate": 3.478260869565218e-05,
      "loss": 0.6136,
      "step": 700
    },
    {
      "epoch": 8.0,
      "eval_f1": 0.6735976418570376,
      "eval_loss": 0.5223021507263184,
      "eval_runtime": 7.2029,
      "eval_samples_per_second": 12.773,
      "eval_steps_per_second": 3.193,
      "step": 736
    },
    {
      "epoch": 8.695652173913043,
      "grad_norm": 7.914676666259766,
      "learning_rate": 3.260869565217392e-05,
      "loss": 0.496,
      "step": 800
    },
    {
      "epoch": 9.0,
      "eval_f1": 0.7776872198363687,
      "eval_loss": 0.7307997345924377,
      "eval_runtime": 7.1943,
      "eval_samples_per_second": 12.788,
      "eval_steps_per_second": 3.197,
      "step": 828
    },
    {
      "epoch": 9.782608695652174,
      "grad_norm": 1.3677582740783691,
      "learning_rate": 3.0434782608695656e-05,
      "loss": 0.4858,
      "step": 900
    },
    {
      "epoch": 10.0,
      "eval_f1": 0.7831521739130435,
      "eval_loss": 0.7452064156532288,
      "eval_runtime": 7.2905,
      "eval_samples_per_second": 12.619,
      "eval_steps_per_second": 3.155,
      "step": 920
    },
    {
      "epoch": 10.869565217391305,
      "grad_norm": 12.851914405822754,
      "learning_rate": 2.826086956521739e-05,
      "loss": 0.4181,
      "step": 1000
    },
    {
      "epoch": 11.0,
      "eval_f1": 0.7798831927319922,
      "eval_loss": 0.7523320913314819,
      "eval_runtime": 7.2341,
      "eval_samples_per_second": 12.718,
      "eval_steps_per_second": 3.179,
      "step": 1012
    },
    {
      "epoch": 11.956521739130435,
      "grad_norm": 12.312176704406738,
      "learning_rate": 2.608695652173913e-05,
      "loss": 0.3395,
      "step": 1100
    },
    {
      "epoch": 12.0,
      "eval_f1": 0.741650913673598,
      "eval_loss": 1.184059500694275,
      "eval_runtime": 7.2136,
      "eval_samples_per_second": 12.754,
      "eval_steps_per_second": 3.188,
      "step": 1104
    },
    {
      "epoch": 13.0,
      "eval_f1": 0.8181082820112143,
      "eval_loss": 0.7770065069198608,
      "eval_runtime": 7.1902,
      "eval_samples_per_second": 12.795,
      "eval_steps_per_second": 3.199,
      "step": 1196
    },
    {
      "epoch": 13.043478260869565,
      "grad_norm": 0.024596206843852997,
      "learning_rate": 2.391304347826087e-05,
      "loss": 0.2375,
      "step": 1200
    },
    {
      "epoch": 14.0,
      "eval_f1": 0.77725724787834,
      "eval_loss": 1.149732232093811,
      "eval_runtime": 7.2423,
      "eval_samples_per_second": 12.703,
      "eval_steps_per_second": 3.176,
      "step": 1288
    },
    {
      "epoch": 14.130434782608695,
      "grad_norm": 0.04395654425024986,
      "learning_rate": 2.173913043478261e-05,
      "loss": 0.1728,
      "step": 1300
    },
    {
      "epoch": 15.0,
      "eval_f1": 0.7634584417193113,
      "eval_loss": 1.518917202949524,
      "eval_runtime": 7.2313,
      "eval_samples_per_second": 12.722,
      "eval_steps_per_second": 3.181,
      "step": 1380
    },
    {
      "epoch": 15.217391304347826,
      "grad_norm": 0.007132470607757568,
      "learning_rate": 1.956521739130435e-05,
      "loss": 0.0324,
      "step": 1400
    },
    {
      "epoch": 16.0,
      "eval_f1": 0.7419546636937941,
      "eval_loss": 1.6500256061553955,
      "eval_runtime": 7.2475,
      "eval_samples_per_second": 12.694,
      "eval_steps_per_second": 3.174,
      "step": 1472
    },
    {
      "epoch": 16.304347826086957,
      "grad_norm": 0.004443590063601732,
      "learning_rate": 1.739130434782609e-05,
      "loss": 0.0764,
      "step": 1500
    },
    {
      "epoch": 17.0,
      "eval_f1": 0.7634584417193113,
      "eval_loss": 1.5010850429534912,
      "eval_runtime": 7.2861,
      "eval_samples_per_second": 12.627,
      "eval_steps_per_second": 3.157,
      "step": 1564
    },
    {
      "epoch": 17.391304347826086,
      "grad_norm": 8.947516441345215,
      "learning_rate": 1.5217391304347828e-05,
      "loss": 0.0244,
      "step": 1600
    },
    {
      "epoch": 18.0,
      "eval_f1": 0.75442242114237,
      "eval_loss": 1.5790338516235352,
      "eval_runtime": 7.2916,
      "eval_samples_per_second": 12.617,
      "eval_steps_per_second": 3.154,
      "step": 1656
    },
    {
      "epoch": 18.47826086956522,
      "grad_norm": 0.016934270039200783,
      "learning_rate": 1.3043478260869566e-05,
      "loss": 0.0002,
      "step": 1700
    },
    {
      "epoch": 19.0,
      "eval_f1": 0.7539112050739958,
      "eval_loss": 1.929887056350708,
      "eval_runtime": 7.2258,
      "eval_samples_per_second": 12.732,
      "eval_steps_per_second": 3.183,
      "step": 1748
    },
    {
      "epoch": 19.565217391304348,
      "grad_norm": 0.02545306272804737,
      "learning_rate": 1.0869565217391305e-05,
      "loss": 0.0004,
      "step": 1800
    },
    {
      "epoch": 20.0,
      "eval_f1": 0.7634584417193113,
      "eval_loss": 1.7885226011276245,
      "eval_runtime": 7.2042,
      "eval_samples_per_second": 12.77,
      "eval_steps_per_second": 3.193,
      "step": 1840
    },
    {
      "epoch": 20.652173913043477,
      "grad_norm": 0.0008984901360236108,
      "learning_rate": 8.695652173913044e-06,
      "loss": 0.0001,
      "step": 1900
    },
    {
      "epoch": 21.0,
      "eval_f1": 0.7634584417193113,
      "eval_loss": 1.82301926612854,
      "eval_runtime": 7.2146,
      "eval_samples_per_second": 12.752,
      "eval_steps_per_second": 3.188,
      "step": 1932
    },
    {
      "epoch": 21.73913043478261,
      "grad_norm": 0.0009025917970575392,
      "learning_rate": 6.521739130434783e-06,
      "loss": 0.0001,
      "step": 2000
    },
    {
      "epoch": 22.0,
      "eval_f1": 0.7634584417193113,
      "eval_loss": 1.8578201532363892,
      "eval_runtime": 7.1765,
      "eval_samples_per_second": 12.82,
      "eval_steps_per_second": 3.205,
      "step": 2024
    },
    {
      "epoch": 22.82608695652174,
      "grad_norm": 0.001884501543827355,
      "learning_rate": 4.347826086956522e-06,
      "loss": 0.0,
      "step": 2100
    },
    {
      "epoch": 23.0,
      "eval_f1": 0.7634584417193113,
      "eval_loss": 1.87649405002594,
      "eval_runtime": 7.1744,
      "eval_samples_per_second": 12.823,
      "eval_steps_per_second": 3.206,
      "step": 2116
    },
    {
      "epoch": 23.91304347826087,
      "grad_norm": 0.003843324724584818,
      "learning_rate": 2.173913043478261e-06,
      "loss": 0.0,
      "step": 2200
    },
    {
      "epoch": 24.0,
      "eval_f1": 0.7634584417193113,
      "eval_loss": 1.8845328092575073,
      "eval_runtime": 7.1501,
      "eval_samples_per_second": 12.867,
      "eval_steps_per_second": 3.217,
      "step": 2208
    },
    {
      "epoch": 25.0,
      "grad_norm": 0.0016703982837498188,
      "learning_rate": 0.0,
      "loss": 0.0,
      "step": 2300
    },
    {
      "epoch": 25.0,
      "eval_f1": 0.7634584417193113,
      "eval_loss": 1.8875935077667236,
      "eval_runtime": 7.3972,
      "eval_samples_per_second": 12.437,
      "eval_steps_per_second": 3.109,
      "step": 2300
    }
  ],
  "logging_steps": 100,
  "max_steps": 2300,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 25,
  "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": 1.57899997983e+16,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}