{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 50.0,
  "eval_steps": 10000,
  "global_step": 126850,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 1.9999842333464724e-05,
      "loss": 0.1488,
      "step": 1
    },
    {
      "epoch": 0.39,
      "learning_rate": 1.9842333464722114e-05,
      "loss": 0.1737,
      "step": 1000
    },
    {
      "epoch": 0.79,
      "learning_rate": 1.968498226251478e-05,
      "loss": 0.1713,
      "step": 2000
    },
    {
      "epoch": 1.18,
      "learning_rate": 1.9527315727236894e-05,
      "loss": 0.1741,
      "step": 3000
    },
    {
      "epoch": 1.58,
      "learning_rate": 1.9369806858494287e-05,
      "loss": 0.1357,
      "step": 4000
    },
    {
      "epoch": 1.97,
      "learning_rate": 1.92121403232164e-05,
      "loss": 0.0985,
      "step": 5000
    },
    {
      "epoch": 2.36,
      "learning_rate": 1.9054473787938513e-05,
      "loss": 0.0746,
      "step": 6000
    },
    {
      "epoch": 2.76,
      "learning_rate": 1.8897437918801735e-05,
      "loss": 0.077,
      "step": 7000
    },
    {
      "epoch": 3.15,
      "learning_rate": 1.8739771383523847e-05,
      "loss": 0.0747,
      "step": 8000
    },
    {
      "epoch": 3.55,
      "learning_rate": 1.858210484824596e-05,
      "loss": 0.0607,
      "step": 9000
    },
    {
      "epoch": 3.94,
      "learning_rate": 1.8424438312968072e-05,
      "loss": 0.0648,
      "step": 10000
    },
    {
      "epoch": 3.94,
      "eval_accuracy": 0.971966859607309,
      "eval_f1": 0.8975528826213189,
      "eval_loss": 0.047781217843294144,
      "eval_precision": 0.9169491525423729,
      "eval_recall": 0.8789601949634444,
      "eval_runtime": 30.9282,
      "eval_samples_per_second": 284.886,
      "eval_steps_per_second": 35.631,
      "step": 10000
    },
    {
      "epoch": 4.34,
      "learning_rate": 1.8266771777690185e-05,
      "loss": 0.0604,
      "step": 11000
    },
    {
      "epoch": 4.73,
      "learning_rate": 1.8109105242412298e-05,
      "loss": 0.0587,
      "step": 12000
    },
    {
      "epoch": 5.12,
      "learning_rate": 1.795159637366969e-05,
      "loss": 0.0635,
      "step": 13000
    },
    {
      "epoch": 5.52,
      "learning_rate": 1.779408750492708e-05,
      "loss": 0.0558,
      "step": 14000
    },
    {
      "epoch": 5.91,
      "learning_rate": 1.7636420969649194e-05,
      "loss": 0.0525,
      "step": 15000
    },
    {
      "epoch": 6.31,
      "learning_rate": 1.7478754434371306e-05,
      "loss": 0.05,
      "step": 16000
    },
    {
      "epoch": 6.7,
      "learning_rate": 1.7321245565628696e-05,
      "loss": 0.0513,
      "step": 17000
    },
    {
      "epoch": 7.09,
      "learning_rate": 1.716357903035081e-05,
      "loss": 0.0423,
      "step": 18000
    },
    {
      "epoch": 7.49,
      "learning_rate": 1.70060701616082e-05,
      "loss": 0.0417,
      "step": 19000
    },
    {
      "epoch": 7.88,
      "learning_rate": 1.684840362633031e-05,
      "loss": 0.0432,
      "step": 20000
    },
    {
      "epoch": 7.88,
      "eval_accuracy": 0.9800249687890137,
      "eval_f1": 0.9293172690763052,
      "eval_loss": 0.03126443922519684,
      "eval_precision": 0.9189833200953137,
      "eval_recall": 0.9398862713241267,
      "eval_runtime": 30.5459,
      "eval_samples_per_second": 288.451,
      "eval_steps_per_second": 36.077,
      "step": 20000
    },
    {
      "epoch": 8.28,
      "learning_rate": 1.6690737091052424e-05,
      "loss": 0.0378,
      "step": 21000
    },
    {
      "epoch": 8.67,
      "learning_rate": 1.6533070555774536e-05,
      "loss": 0.041,
      "step": 22000
    },
    {
      "epoch": 9.07,
      "learning_rate": 1.637540402049665e-05,
      "loss": 0.0403,
      "step": 23000
    },
    {
      "epoch": 9.46,
      "learning_rate": 1.621805281828932e-05,
      "loss": 0.0338,
      "step": 24000
    },
    {
      "epoch": 9.85,
      "learning_rate": 1.6060386283011432e-05,
      "loss": 0.0352,
      "step": 25000
    },
    {
      "epoch": 10.25,
      "learning_rate": 1.5902877414268822e-05,
      "loss": 0.0356,
      "step": 26000
    },
    {
      "epoch": 10.64,
      "learning_rate": 1.5745210878990935e-05,
      "loss": 0.0343,
      "step": 27000
    },
    {
      "epoch": 11.04,
      "learning_rate": 1.5587702010248328e-05,
      "loss": 0.039,
      "step": 28000
    },
    {
      "epoch": 11.43,
      "learning_rate": 1.543003547497044e-05,
      "loss": 0.0336,
      "step": 29000
    },
    {
      "epoch": 11.82,
      "learning_rate": 1.5272368939692553e-05,
      "loss": 0.0294,
      "step": 30000
    },
    {
      "epoch": 11.82,
      "eval_accuracy": 0.9878560889796845,
      "eval_f1": 0.9563799429270281,
      "eval_loss": 0.023840567097067833,
      "eval_precision": 0.9599018003273322,
      "eval_recall": 0.9528838342810723,
      "eval_runtime": 29.3298,
      "eval_samples_per_second": 300.411,
      "eval_steps_per_second": 37.573,
      "step": 30000
    },
    {
      "epoch": 12.22,
      "learning_rate": 1.5114702404414664e-05,
      "loss": 0.0362,
      "step": 31000
    },
    {
      "epoch": 12.61,
      "learning_rate": 1.4957035869136779e-05,
      "loss": 0.0352,
      "step": 32000
    },
    {
      "epoch": 13.01,
      "learning_rate": 1.4799369333858891e-05,
      "loss": 0.0297,
      "step": 33000
    },
    {
      "epoch": 13.4,
      "learning_rate": 1.464186046511628e-05,
      "loss": 0.0316,
      "step": 34000
    },
    {
      "epoch": 13.8,
      "learning_rate": 1.4484193929838392e-05,
      "loss": 0.0312,
      "step": 35000
    },
    {
      "epoch": 14.19,
      "learning_rate": 1.4326527394560506e-05,
      "loss": 0.0315,
      "step": 36000
    },
    {
      "epoch": 14.58,
      "learning_rate": 1.4168860859282619e-05,
      "loss": 0.0275,
      "step": 37000
    },
    {
      "epoch": 14.98,
      "learning_rate": 1.401135199054001e-05,
      "loss": 0.0431,
      "step": 38000
    },
    {
      "epoch": 15.37,
      "learning_rate": 1.3853843121797399e-05,
      "loss": 0.034,
      "step": 39000
    },
    {
      "epoch": 15.77,
      "learning_rate": 1.3696176586519511e-05,
      "loss": 0.0275,
      "step": 40000
    },
    {
      "epoch": 15.77,
      "eval_accuracy": 0.9868346385200317,
      "eval_f1": 0.9522240527182867,
      "eval_loss": 0.022756921127438545,
      "eval_precision": 0.9657477025898078,
      "eval_recall": 0.9390739236393176,
      "eval_runtime": 29.4347,
      "eval_samples_per_second": 299.34,
      "eval_steps_per_second": 37.439,
      "step": 40000
    },
    {
      "epoch": 16.16,
      "learning_rate": 1.3538510051241624e-05,
      "loss": 0.0285,
      "step": 41000
    },
    {
      "epoch": 16.55,
      "learning_rate": 1.3381001182499016e-05,
      "loss": 0.0286,
      "step": 42000
    },
    {
      "epoch": 16.95,
      "learning_rate": 1.3223334647221128e-05,
      "loss": 0.0279,
      "step": 43000
    },
    {
      "epoch": 17.34,
      "learning_rate": 1.306566811194324e-05,
      "loss": 0.0254,
      "step": 44000
    },
    {
      "epoch": 17.74,
      "learning_rate": 1.2908001576665355e-05,
      "loss": 0.0239,
      "step": 45000
    },
    {
      "epoch": 18.13,
      "learning_rate": 1.2750335041387468e-05,
      "loss": 0.0269,
      "step": 46000
    },
    {
      "epoch": 18.53,
      "learning_rate": 1.259266850610958e-05,
      "loss": 0.0268,
      "step": 47000
    },
    {
      "epoch": 18.92,
      "learning_rate": 1.2435001970831693e-05,
      "loss": 0.0242,
      "step": 48000
    },
    {
      "epoch": 19.31,
      "learning_rate": 1.2277493102089081e-05,
      "loss": 0.0268,
      "step": 49000
    },
    {
      "epoch": 19.71,
      "learning_rate": 1.2119984233346473e-05,
      "loss": 0.0268,
      "step": 50000
    },
    {
      "epoch": 19.71,
      "eval_accuracy": 0.9872886165020996,
      "eval_f1": 0.9541734860883797,
      "eval_loss": 0.017343735322356224,
      "eval_precision": 0.9612530915086562,
      "eval_recall": 0.9471974004874086,
      "eval_runtime": 29.9772,
      "eval_samples_per_second": 293.923,
      "eval_steps_per_second": 36.761,
      "step": 50000
    },
    {
      "epoch": 20.1,
      "learning_rate": 1.1962317698068587e-05,
      "loss": 0.0238,
      "step": 51000
    },
    {
      "epoch": 20.5,
      "learning_rate": 1.1804808829325975e-05,
      "loss": 0.0279,
      "step": 52000
    },
    {
      "epoch": 20.89,
      "learning_rate": 1.1647142294048088e-05,
      "loss": 0.0251,
      "step": 53000
    },
    {
      "epoch": 21.28,
      "learning_rate": 1.148963342530548e-05,
      "loss": 0.0258,
      "step": 54000
    },
    {
      "epoch": 21.68,
      "learning_rate": 1.1331966890027592e-05,
      "loss": 0.0233,
      "step": 55000
    },
    {
      "epoch": 22.07,
      "learning_rate": 1.1174458021284984e-05,
      "loss": 0.0281,
      "step": 56000
    },
    {
      "epoch": 22.47,
      "learning_rate": 1.1016949152542374e-05,
      "loss": 0.0255,
      "step": 57000
    },
    {
      "epoch": 22.86,
      "learning_rate": 1.0859282617264486e-05,
      "loss": 0.0251,
      "step": 58000
    },
    {
      "epoch": 23.26,
      "learning_rate": 1.0701616081986599e-05,
      "loss": 0.0252,
      "step": 59000
    },
    {
      "epoch": 23.65,
      "learning_rate": 1.0543949546708712e-05,
      "loss": 0.0192,
      "step": 60000
    },
    {
      "epoch": 23.65,
      "eval_accuracy": 0.9854727045738282,
      "eval_f1": 0.9498825371965545,
      "eval_loss": 0.01524350605905056,
      "eval_precision": 0.9168556311413454,
      "eval_recall": 0.9853777416734363,
      "eval_runtime": 29.585,
      "eval_samples_per_second": 297.82,
      "eval_steps_per_second": 37.249,
      "step": 60000
    },
    {
      "epoch": 24.04,
      "learning_rate": 1.0386440677966103e-05,
      "loss": 0.0223,
      "step": 61000
    },
    {
      "epoch": 24.44,
      "learning_rate": 1.0228774142688216e-05,
      "loss": 0.0221,
      "step": 62000
    },
    {
      "epoch": 24.83,
      "learning_rate": 1.0071107607410328e-05,
      "loss": 0.0232,
      "step": 63000
    },
    {
      "epoch": 25.23,
      "learning_rate": 9.913441072132441e-06,
      "loss": 0.0215,
      "step": 64000
    },
    {
      "epoch": 25.62,
      "learning_rate": 9.755774536854554e-06,
      "loss": 0.0221,
      "step": 65000
    },
    {
      "epoch": 26.01,
      "learning_rate": 9.598265668111943e-06,
      "loss": 0.0216,
      "step": 66000
    },
    {
      "epoch": 26.41,
      "learning_rate": 9.440599132834056e-06,
      "loss": 0.0215,
      "step": 67000
    },
    {
      "epoch": 26.8,
      "learning_rate": 9.282932597556169e-06,
      "loss": 0.02,
      "step": 68000
    },
    {
      "epoch": 27.2,
      "learning_rate": 9.12542372881356e-06,
      "loss": 0.0227,
      "step": 69000
    },
    {
      "epoch": 27.59,
      "learning_rate": 8.967757193535673e-06,
      "loss": 0.018,
      "step": 70000
    },
    {
      "epoch": 27.59,
      "eval_accuracy": 0.9919418908182953,
      "eval_f1": 0.9716113554578168,
      "eval_loss": 0.011301785707473755,
      "eval_precision": 0.9566929133858267,
      "eval_recall": 0.9870024370430545,
      "eval_runtime": 29.5941,
      "eval_samples_per_second": 297.728,
      "eval_steps_per_second": 37.237,
      "step": 70000
    },
    {
      "epoch": 27.99,
      "learning_rate": 8.810248324793063e-06,
      "loss": 0.0211,
      "step": 71000
    },
    {
      "epoch": 28.38,
      "learning_rate": 8.652581789515177e-06,
      "loss": 0.0199,
      "step": 72000
    },
    {
      "epoch": 28.77,
      "learning_rate": 8.49491525423729e-06,
      "loss": 0.0197,
      "step": 73000
    },
    {
      "epoch": 29.17,
      "learning_rate": 8.337248718959402e-06,
      "loss": 0.0216,
      "step": 74000
    },
    {
      "epoch": 29.56,
      "learning_rate": 8.17989751675207e-06,
      "loss": 0.0177,
      "step": 75000
    },
    {
      "epoch": 29.96,
      "learning_rate": 8.022230981474182e-06,
      "loss": 0.0193,
      "step": 76000
    },
    {
      "epoch": 30.35,
      "learning_rate": 7.864564446196295e-06,
      "loss": 0.0167,
      "step": 77000
    },
    {
      "epoch": 30.74,
      "learning_rate": 7.706897910918409e-06,
      "loss": 0.0181,
      "step": 78000
    },
    {
      "epoch": 31.14,
      "learning_rate": 7.549231375640521e-06,
      "loss": 0.0194,
      "step": 79000
    },
    {
      "epoch": 31.53,
      "learning_rate": 7.3915648403626335e-06,
      "loss": 0.0155,
      "step": 80000
    },
    {
      "epoch": 31.53,
      "eval_accuracy": 0.9923958688003632,
      "eval_f1": 0.9729947601773479,
      "eval_loss": 0.012764379382133484,
      "eval_precision": 0.9656,
      "eval_recall": 0.9805036555645816,
      "eval_runtime": 29.6436,
      "eval_samples_per_second": 297.231,
      "eval_steps_per_second": 37.175,
      "step": 80000
    },
    {
      "epoch": 31.93,
      "learning_rate": 7.233898305084747e-06,
      "loss": 0.018,
      "step": 81000
    },
    {
      "epoch": 32.32,
      "learning_rate": 7.076389436342137e-06,
      "loss": 0.0184,
      "step": 82000
    },
    {
      "epoch": 32.72,
      "learning_rate": 6.9187229010642495e-06,
      "loss": 0.0169,
      "step": 83000
    },
    {
      "epoch": 33.11,
      "learning_rate": 6.76121403232164e-06,
      "loss": 0.0172,
      "step": 84000
    },
    {
      "epoch": 33.5,
      "learning_rate": 6.603547497043753e-06,
      "loss": 0.0191,
      "step": 85000
    },
    {
      "epoch": 33.9,
      "learning_rate": 6.446038628301144e-06,
      "loss": 0.0136,
      "step": 86000
    },
    {
      "epoch": 34.29,
      "learning_rate": 6.288372093023256e-06,
      "loss": 0.0173,
      "step": 87000
    },
    {
      "epoch": 34.69,
      "learning_rate": 6.130705557745369e-06,
      "loss": 0.0161,
      "step": 88000
    },
    {
      "epoch": 35.08,
      "learning_rate": 5.97319668900276e-06,
      "loss": 0.0143,
      "step": 89000
    },
    {
      "epoch": 35.47,
      "learning_rate": 5.815530153724872e-06,
      "loss": 0.0142,
      "step": 90000
    },
    {
      "epoch": 35.47,
      "eval_accuracy": 0.9943252752241516,
      "eval_f1": 0.9798549556809025,
      "eval_loss": 0.00998597126454115,
      "eval_precision": 0.9720223820943246,
      "eval_recall": 0.9878147847278635,
      "eval_runtime": 29.6518,
      "eval_samples_per_second": 297.149,
      "eval_steps_per_second": 37.165,
      "step": 90000
    },
    {
      "epoch": 35.87,
      "learning_rate": 5.657863618446985e-06,
      "loss": 0.0157,
      "step": 91000
    },
    {
      "epoch": 36.26,
      "learning_rate": 5.500354749704376e-06,
      "loss": 0.0164,
      "step": 92000
    },
    {
      "epoch": 36.66,
      "learning_rate": 5.342688214426488e-06,
      "loss": 0.0159,
      "step": 93000
    },
    {
      "epoch": 37.05,
      "learning_rate": 5.18517934568388e-06,
      "loss": 0.0134,
      "step": 94000
    },
    {
      "epoch": 37.45,
      "learning_rate": 5.0275128104059924e-06,
      "loss": 0.0148,
      "step": 95000
    },
    {
      "epoch": 37.84,
      "learning_rate": 4.870003941663382e-06,
      "loss": 0.0119,
      "step": 96000
    },
    {
      "epoch": 38.23,
      "learning_rate": 4.712337406385495e-06,
      "loss": 0.0116,
      "step": 97000
    },
    {
      "epoch": 38.63,
      "learning_rate": 4.554670871107608e-06,
      "loss": 0.0128,
      "step": 98000
    },
    {
      "epoch": 39.02,
      "learning_rate": 4.397162002364998e-06,
      "loss": 0.016,
      "step": 99000
    },
    {
      "epoch": 39.42,
      "learning_rate": 4.239495467087111e-06,
      "loss": 0.016,
      "step": 100000
    },
    {
      "epoch": 39.42,
      "eval_accuracy": 0.9950062421972534,
      "eval_f1": 0.9822437449556094,
      "eval_loss": 0.009744029492139816,
      "eval_precision": 0.9759422614274258,
      "eval_recall": 0.9886271324126726,
      "eval_runtime": 29.5778,
      "eval_samples_per_second": 297.892,
      "eval_steps_per_second": 37.258,
      "step": 100000
    },
    {
      "epoch": 39.81,
      "learning_rate": 4.081986598344502e-06,
      "loss": 0.0126,
      "step": 101000
    },
    {
      "epoch": 40.2,
      "learning_rate": 3.924320063066614e-06,
      "loss": 0.015,
      "step": 102000
    },
    {
      "epoch": 40.6,
      "learning_rate": 3.7666535277887274e-06,
      "loss": 0.0116,
      "step": 103000
    },
    {
      "epoch": 40.99,
      "learning_rate": 3.60898699251084e-06,
      "loss": 0.013,
      "step": 104000
    },
    {
      "epoch": 41.39,
      "learning_rate": 3.4514781237682304e-06,
      "loss": 0.0119,
      "step": 105000
    },
    {
      "epoch": 41.78,
      "learning_rate": 3.2938115884903434e-06,
      "loss": 0.0117,
      "step": 106000
    },
    {
      "epoch": 42.18,
      "learning_rate": 3.136145053212456e-06,
      "loss": 0.0107,
      "step": 107000
    },
    {
      "epoch": 42.57,
      "learning_rate": 2.9786361844698463e-06,
      "loss": 0.0103,
      "step": 108000
    },
    {
      "epoch": 42.96,
      "learning_rate": 2.8209696491919594e-06,
      "loss": 0.0137,
      "step": 109000
    },
    {
      "epoch": 43.36,
      "learning_rate": 2.663303113914072e-06,
      "loss": 0.0122,
      "step": 110000
    },
    {
      "epoch": 43.36,
      "eval_accuracy": 0.9962546816479401,
      "eval_f1": 0.9865689865689866,
      "eval_loss": 0.010858706198632717,
      "eval_precision": 0.9885807504078303,
      "eval_recall": 0.9845653939886271,
      "eval_runtime": 29.5587,
      "eval_samples_per_second": 298.085,
      "eval_steps_per_second": 37.282,
      "step": 110000
    },
    {
      "epoch": 43.75,
      "learning_rate": 2.5056365786361845e-06,
      "loss": 0.0104,
      "step": 111000
    },
    {
      "epoch": 44.15,
      "learning_rate": 2.3481277098935753e-06,
      "loss": 0.013,
      "step": 112000
    },
    {
      "epoch": 44.54,
      "learning_rate": 2.190461174615688e-06,
      "loss": 0.0095,
      "step": 113000
    },
    {
      "epoch": 44.93,
      "learning_rate": 2.0329523058730783e-06,
      "loss": 0.0114,
      "step": 114000
    },
    {
      "epoch": 45.33,
      "learning_rate": 1.8752857705951913e-06,
      "loss": 0.0116,
      "step": 115000
    },
    {
      "epoch": 45.72,
      "learning_rate": 1.7176192353173041e-06,
      "loss": 0.0084,
      "step": 116000
    },
    {
      "epoch": 46.12,
      "learning_rate": 1.559952700039417e-06,
      "loss": 0.0134,
      "step": 117000
    },
    {
      "epoch": 46.51,
      "learning_rate": 1.4024438312968075e-06,
      "loss": 0.0116,
      "step": 118000
    },
    {
      "epoch": 46.91,
      "learning_rate": 1.2449349625541981e-06,
      "loss": 0.0116,
      "step": 119000
    },
    {
      "epoch": 47.3,
      "learning_rate": 1.0872684272763107e-06,
      "loss": 0.0087,
      "step": 120000
    },
    {
      "epoch": 47.3,
      "eval_accuracy": 0.9962546816479401,
      "eval_f1": 0.9865799105327369,
      "eval_loss": 0.010331220924854279,
      "eval_precision": 0.987785016286645,
      "eval_recall": 0.9853777416734363,
      "eval_runtime": 29.6133,
      "eval_samples_per_second": 297.536,
      "eval_steps_per_second": 37.213,
      "step": 120000
    },
    {
      "epoch": 47.69,
      "learning_rate": 9.296018919984234e-07,
      "loss": 0.0127,
      "step": 121000
    },
    {
      "epoch": 48.09,
      "learning_rate": 7.719353567205361e-07,
      "loss": 0.0113,
      "step": 122000
    },
    {
      "epoch": 48.48,
      "learning_rate": 6.144264879779267e-07,
      "loss": 0.0103,
      "step": 123000
    },
    {
      "epoch": 48.88,
      "learning_rate": 4.5675995270003944e-07,
      "loss": 0.0078,
      "step": 124000
    },
    {
      "epoch": 49.27,
      "learning_rate": 2.992510839574301e-07,
      "loss": 0.0095,
      "step": 125000
    },
    {
      "epoch": 49.66,
      "learning_rate": 1.4158454867954278e-07,
      "loss": 0.0097,
      "step": 126000
    }
  ],
  "logging_steps": 1000,
  "max_steps": 126850,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 50,
  "save_steps": 500,
  "total_flos": 0.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}