{
  "best_global_step": 539,
  "best_metric": 0.7652582159624414,
  "best_model_checkpoint": "/tmp/finetune/l2/checkpoint-539",
  "epoch": 11.0,
  "eval_steps": 500,
  "global_step": 539,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.20408163265306123,
      "grad_norm": 2.194807529449463,
      "learning_rate": 6.081081081081082e-07,
      "loss": 0.1998,
      "step": 10
    },
    {
      "epoch": 0.40816326530612246,
      "grad_norm": 5.4665727615356445,
      "learning_rate": 1.2837837837837838e-06,
      "loss": 0.2292,
      "step": 20
    },
    {
      "epoch": 0.6122448979591837,
      "grad_norm": 5.707193851470947,
      "learning_rate": 1.9594594594594595e-06,
      "loss": 0.1915,
      "step": 30
    },
    {
      "epoch": 0.8163265306122449,
      "grad_norm": 7.533806800842285,
      "learning_rate": 2.6351351351351353e-06,
      "loss": 0.226,
      "step": 40
    },
    {
      "epoch": 1.0,
      "eval_f1_macro": 0.7105381888127623,
      "eval_f1_micro": 0.7078384798099763,
      "eval_f1_weighted": 0.7004977849134466,
      "eval_loss": 0.18318259716033936,
      "eval_runtime": 0.7758,
      "eval_samples_per_second": 248.764,
      "eval_steps_per_second": 16.756,
      "step": 49
    },
    {
      "epoch": 1.0204081632653061,
      "grad_norm": 7.434163570404053,
      "learning_rate": 3.310810810810811e-06,
      "loss": 0.2259,
      "step": 50
    },
    {
      "epoch": 1.2244897959183674,
      "grad_norm": 4.361626148223877,
      "learning_rate": 3.986486486486487e-06,
      "loss": 0.1761,
      "step": 60
    },
    {
      "epoch": 1.4285714285714286,
      "grad_norm": 4.420626640319824,
      "learning_rate": 4.6621621621621625e-06,
      "loss": 0.1844,
      "step": 70
    },
    {
      "epoch": 1.6326530612244898,
      "grad_norm": 5.982855319976807,
      "learning_rate": 4.962178517397882e-06,
      "loss": 0.1538,
      "step": 80
    },
    {
      "epoch": 1.836734693877551,
      "grad_norm": 3.9699554443359375,
      "learning_rate": 4.886535552193646e-06,
      "loss": 0.171,
      "step": 90
    },
    {
      "epoch": 2.0,
      "eval_f1_macro": 0.7073515633787281,
      "eval_f1_micro": 0.7038834951456311,
      "eval_f1_weighted": 0.6978974909161666,
      "eval_loss": 0.17095711827278137,
      "eval_runtime": 0.6521,
      "eval_samples_per_second": 295.949,
      "eval_steps_per_second": 19.934,
      "step": 98
    },
    {
      "epoch": 2.0408163265306123,
      "grad_norm": 2.6549506187438965,
      "learning_rate": 4.81089258698941e-06,
      "loss": 0.1457,
      "step": 100
    },
    {
      "epoch": 2.2448979591836733,
      "grad_norm": 3.922772169113159,
      "learning_rate": 4.7352496217851745e-06,
      "loss": 0.1459,
      "step": 110
    },
    {
      "epoch": 2.4489795918367347,
      "grad_norm": 4.702055931091309,
      "learning_rate": 4.659606656580938e-06,
      "loss": 0.1306,
      "step": 120
    },
    {
      "epoch": 2.6530612244897958,
      "grad_norm": 4.254676818847656,
      "learning_rate": 4.583963691376702e-06,
      "loss": 0.1271,
      "step": 130
    },
    {
      "epoch": 2.857142857142857,
      "grad_norm": 5.3793182373046875,
      "learning_rate": 4.508320726172467e-06,
      "loss": 0.1322,
      "step": 140
    },
    {
      "epoch": 3.0,
      "eval_f1_macro": 0.7129027408610171,
      "eval_f1_micro": 0.711864406779661,
      "eval_f1_weighted": 0.7066061540411746,
      "eval_loss": 0.16003255546092987,
      "eval_runtime": 0.6419,
      "eval_samples_per_second": 300.689,
      "eval_steps_per_second": 20.254,
      "step": 147
    },
    {
      "epoch": 3.061224489795918,
      "grad_norm": 4.1884002685546875,
      "learning_rate": 4.43267776096823e-06,
      "loss": 0.1339,
      "step": 150
    },
    {
      "epoch": 3.2653061224489797,
      "grad_norm": 3.1676673889160156,
      "learning_rate": 4.357034795763994e-06,
      "loss": 0.1199,
      "step": 160
    },
    {
      "epoch": 3.4693877551020407,
      "grad_norm": 6.360077857971191,
      "learning_rate": 4.281391830559759e-06,
      "loss": 0.1056,
      "step": 170
    },
    {
      "epoch": 3.673469387755102,
      "grad_norm": 3.8555994033813477,
      "learning_rate": 4.2057488653555225e-06,
      "loss": 0.1199,
      "step": 180
    },
    {
      "epoch": 3.877551020408163,
      "grad_norm": 5.292940616607666,
      "learning_rate": 4.130105900151286e-06,
      "loss": 0.1269,
      "step": 190
    },
    {
      "epoch": 4.0,
      "eval_f1_macro": 0.739022809362487,
      "eval_f1_micro": 0.7380952380952381,
      "eval_f1_weighted": 0.7328483471625246,
      "eval_loss": 0.15181854367256165,
      "eval_runtime": 0.6743,
      "eval_samples_per_second": 286.242,
      "eval_steps_per_second": 19.281,
      "step": 196
    },
    {
      "epoch": 4.081632653061225,
      "grad_norm": 2.861799478530884,
      "learning_rate": 4.05446293494705e-06,
      "loss": 0.0965,
      "step": 200
    },
    {
      "epoch": 4.285714285714286,
      "grad_norm": 2.875753164291382,
      "learning_rate": 3.978819969742814e-06,
      "loss": 0.107,
      "step": 210
    },
    {
      "epoch": 4.489795918367347,
      "grad_norm": 3.6004979610443115,
      "learning_rate": 3.9031770045385785e-06,
      "loss": 0.1109,
      "step": 220
    },
    {
      "epoch": 4.6938775510204085,
      "grad_norm": 2.3077666759490967,
      "learning_rate": 3.827534039334342e-06,
      "loss": 0.1049,
      "step": 230
    },
    {
      "epoch": 4.8979591836734695,
      "grad_norm": 4.845459938049316,
      "learning_rate": 3.7518910741301064e-06,
      "loss": 0.0933,
      "step": 240
    },
    {
      "epoch": 5.0,
      "eval_f1_macro": 0.7462814517178719,
      "eval_f1_micro": 0.7469879518072289,
      "eval_f1_weighted": 0.7374498844458771,
      "eval_loss": 0.151344895362854,
      "eval_runtime": 0.6335,
      "eval_samples_per_second": 304.637,
      "eval_steps_per_second": 20.52,
      "step": 245
    },
    {
      "epoch": 5.1020408163265305,
      "grad_norm": 3.957547664642334,
      "learning_rate": 3.67624810892587e-06,
      "loss": 0.1143,
      "step": 250
    },
    {
      "epoch": 5.3061224489795915,
      "grad_norm": 2.440263509750366,
      "learning_rate": 3.6006051437216344e-06,
      "loss": 0.0842,
      "step": 260
    },
    {
      "epoch": 5.510204081632653,
      "grad_norm": 2.0533766746520996,
      "learning_rate": 3.524962178517398e-06,
      "loss": 0.0835,
      "step": 270
    },
    {
      "epoch": 5.714285714285714,
      "grad_norm": 2.234433174133301,
      "learning_rate": 3.449319213313162e-06,
      "loss": 0.0874,
      "step": 280
    },
    {
      "epoch": 5.918367346938775,
      "grad_norm": 2.01436448097229,
      "learning_rate": 3.3736762481089265e-06,
      "loss": 0.0775,
      "step": 290
    },
    {
      "epoch": 6.0,
      "eval_f1_macro": 0.7314792696872793,
      "eval_f1_micro": 0.7320574162679426,
      "eval_f1_weighted": 0.7229213437515356,
      "eval_loss": 0.15624862909317017,
      "eval_runtime": 0.745,
      "eval_samples_per_second": 259.059,
      "eval_steps_per_second": 17.45,
      "step": 294
    },
    {
      "epoch": 6.122448979591836,
      "grad_norm": 1.0920073986053467,
      "learning_rate": 3.2980332829046903e-06,
      "loss": 0.0948,
      "step": 300
    },
    {
      "epoch": 6.326530612244898,
      "grad_norm": 2.755589723587036,
      "learning_rate": 3.222390317700454e-06,
      "loss": 0.0893,
      "step": 310
    },
    {
      "epoch": 6.530612244897959,
      "grad_norm": 2.390260934829712,
      "learning_rate": 3.1467473524962183e-06,
      "loss": 0.0789,
      "step": 320
    },
    {
      "epoch": 6.73469387755102,
      "grad_norm": 1.349293828010559,
      "learning_rate": 3.071104387291982e-06,
      "loss": 0.0778,
      "step": 330
    },
    {
      "epoch": 6.938775510204081,
      "grad_norm": 3.054826498031616,
      "learning_rate": 2.995461422087746e-06,
      "loss": 0.0847,
      "step": 340
    },
    {
      "epoch": 7.0,
      "eval_f1_macro": 0.7618238296226221,
      "eval_f1_micro": 0.7612293144208038,
      "eval_f1_weighted": 0.7539287865439286,
      "eval_loss": 0.1503666639328003,
      "eval_runtime": 0.7057,
      "eval_samples_per_second": 273.478,
      "eval_steps_per_second": 18.421,
      "step": 343
    },
    {
      "epoch": 7.142857142857143,
      "grad_norm": 1.7866673469543457,
      "learning_rate": 2.9198184568835104e-06,
      "loss": 0.0691,
      "step": 350
    },
    {
      "epoch": 7.346938775510204,
      "grad_norm": 1.9781595468521118,
      "learning_rate": 2.844175491679274e-06,
      "loss": 0.07,
      "step": 360
    },
    {
      "epoch": 7.551020408163265,
      "grad_norm": 2.8111209869384766,
      "learning_rate": 2.768532526475038e-06,
      "loss": 0.0726,
      "step": 370
    },
    {
      "epoch": 7.755102040816326,
      "grad_norm": 1.4259343147277832,
      "learning_rate": 2.692889561270802e-06,
      "loss": 0.0741,
      "step": 380
    },
    {
      "epoch": 7.959183673469388,
      "grad_norm": 2.8336338996887207,
      "learning_rate": 2.617246596066566e-06,
      "loss": 0.0736,
      "step": 390
    },
    {
      "epoch": 8.0,
      "eval_f1_macro": 0.7534406853518505,
      "eval_f1_micro": 0.7553444180522565,
      "eval_f1_weighted": 0.7471700863864514,
      "eval_loss": 0.15129989385604858,
      "eval_runtime": 0.7101,
      "eval_samples_per_second": 271.785,
      "eval_steps_per_second": 18.307,
      "step": 392
    },
    {
      "epoch": 8.16326530612245,
      "grad_norm": 2.829455852508545,
      "learning_rate": 2.5416036308623297e-06,
      "loss": 0.0662,
      "step": 400
    },
    {
      "epoch": 8.36734693877551,
      "grad_norm": 1.9278897047042847,
      "learning_rate": 2.465960665658094e-06,
      "loss": 0.076,
      "step": 410
    },
    {
      "epoch": 8.571428571428571,
      "grad_norm": 2.8620712757110596,
      "learning_rate": 2.390317700453858e-06,
      "loss": 0.067,
      "step": 420
    },
    {
      "epoch": 8.775510204081632,
      "grad_norm": 1.9797635078430176,
      "learning_rate": 2.3146747352496223e-06,
      "loss": 0.0662,
      "step": 430
    },
    {
      "epoch": 8.979591836734693,
      "grad_norm": 1.4846246242523193,
      "learning_rate": 2.239031770045386e-06,
      "loss": 0.0634,
      "step": 440
    },
    {
      "epoch": 9.0,
      "eval_f1_macro": 0.7480204045030746,
      "eval_f1_micro": 0.7482352941176471,
      "eval_f1_weighted": 0.7407760140152096,
      "eval_loss": 0.1551072746515274,
      "eval_runtime": 0.6975,
      "eval_samples_per_second": 276.692,
      "eval_steps_per_second": 18.637,
      "step": 441
    },
    {
      "epoch": 9.183673469387756,
      "grad_norm": 3.946624755859375,
      "learning_rate": 2.16338880484115e-06,
      "loss": 0.0621,
      "step": 450
    },
    {
      "epoch": 9.387755102040817,
      "grad_norm": 2.2177817821502686,
      "learning_rate": 2.087745839636914e-06,
      "loss": 0.0599,
      "step": 460
    },
    {
      "epoch": 9.591836734693878,
      "grad_norm": 1.7603358030319214,
      "learning_rate": 2.012102874432678e-06,
      "loss": 0.0593,
      "step": 470
    },
    {
      "epoch": 9.795918367346939,
      "grad_norm": 2.6135215759277344,
      "learning_rate": 1.936459909228442e-06,
      "loss": 0.0751,
      "step": 480
    },
    {
      "epoch": 10.0,
      "grad_norm": 1.4198002815246582,
      "learning_rate": 1.8608169440242058e-06,
      "loss": 0.0598,
      "step": 490
    },
    {
      "epoch": 10.0,
      "eval_f1_macro": 0.7630328262826187,
      "eval_f1_micro": 0.7641509433962265,
      "eval_f1_weighted": 0.758002840665604,
      "eval_loss": 0.14575687050819397,
      "eval_runtime": 1.2622,
      "eval_samples_per_second": 152.91,
      "eval_steps_per_second": 10.3,
      "step": 490
    },
    {
      "epoch": 10.204081632653061,
      "grad_norm": 0.9668400883674622,
      "learning_rate": 1.78517397881997e-06,
      "loss": 0.0548,
      "step": 500
    },
    {
      "epoch": 10.408163265306122,
      "grad_norm": 2.780003547668457,
      "learning_rate": 1.709531013615734e-06,
      "loss": 0.0649,
      "step": 510
    },
    {
      "epoch": 10.612244897959183,
      "grad_norm": 1.962121844291687,
      "learning_rate": 1.6338880484114977e-06,
      "loss": 0.0601,
      "step": 520
    },
    {
      "epoch": 10.816326530612244,
      "grad_norm": 3.070321559906006,
      "learning_rate": 1.5582450832072619e-06,
      "loss": 0.0533,
      "step": 530
    },
    {
      "epoch": 11.0,
      "eval_f1_macro": 0.765227079277026,
      "eval_f1_micro": 0.7652582159624414,
      "eval_f1_weighted": 0.7586082650718015,
      "eval_loss": 0.14734649658203125,
      "eval_runtime": 0.6591,
      "eval_samples_per_second": 292.832,
      "eval_steps_per_second": 19.724,
      "step": 539
    }
  ],
  "logging_steps": 10,
  "max_steps": 735,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 15,
  "save_steps": 500,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 3,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 130948414108188.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}