{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 16.0,
  "eval_steps": 500,
  "global_step": 21008,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.38,
      "learning_rate": 0.0009873064229499874,
      "loss": 2.3443,
      "step": 500
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0009746128458999747,
      "loss": 1.6863,
      "step": 1000
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.575725640796642,
      "eval_loss": 1.955936312675476,
      "eval_runtime": 2249.8295,
      "eval_samples_per_second": 74.652,
      "eval_steps_per_second": 0.146,
      "step": 1313
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0009619192688499619,
      "loss": 1.4821,
      "step": 1500
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0009492256917999492,
      "loss": 1.3492,
      "step": 2000
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0009365321147499366,
      "loss": 1.3275,
      "step": 2500
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.606090917209967,
      "eval_loss": 1.8276444673538208,
      "eval_runtime": 1553.5911,
      "eval_samples_per_second": 108.108,
      "eval_steps_per_second": 0.212,
      "step": 2626
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0009238385376999238,
      "loss": 1.1767,
      "step": 3000
    },
    {
      "epoch": 2.67,
      "learning_rate": 0.0009111449606499112,
      "loss": 1.151,
      "step": 3500
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6130332529546605,
      "eval_loss": 1.885665774345398,
      "eval_runtime": 1611.9735,
      "eval_samples_per_second": 104.192,
      "eval_steps_per_second": 0.204,
      "step": 3939
    },
    {
      "epoch": 3.05,
      "learning_rate": 0.0008984513835998985,
      "loss": 1.1393,
      "step": 4000
    },
    {
      "epoch": 3.43,
      "learning_rate": 0.0008857578065498858,
      "loss": 1.0116,
      "step": 4500
    },
    {
      "epoch": 3.81,
      "learning_rate": 0.0008730642294998731,
      "loss": 1.0336,
      "step": 5000
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6322288708284958,
      "eval_loss": 1.8160221576690674,
      "eval_runtime": 1568.2238,
      "eval_samples_per_second": 107.099,
      "eval_steps_per_second": 0.21,
      "step": 5252
    },
    {
      "epoch": 4.19,
      "learning_rate": 0.0008603706524498604,
      "loss": 0.9722,
      "step": 5500
    },
    {
      "epoch": 4.57,
      "learning_rate": 0.0008476770753998477,
      "loss": 0.9309,
      "step": 6000
    },
    {
      "epoch": 4.95,
      "learning_rate": 0.000834983498349835,
      "loss": 0.947,
      "step": 6500
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.6317465987913429,
      "eval_loss": 1.8050605058670044,
      "eval_runtime": 1542.405,
      "eval_samples_per_second": 108.892,
      "eval_steps_per_second": 0.213,
      "step": 6565
    },
    {
      "epoch": 5.33,
      "learning_rate": 0.0008222899212998223,
      "loss": 0.8427,
      "step": 7000
    },
    {
      "epoch": 5.71,
      "learning_rate": 0.0008095963442498096,
      "loss": 0.8595,
      "step": 7500
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.6442737638057813,
      "eval_loss": 1.7996306419372559,
      "eval_runtime": 1561.8382,
      "eval_samples_per_second": 107.537,
      "eval_steps_per_second": 0.211,
      "step": 7878
    },
    {
      "epoch": 6.09,
      "learning_rate": 0.0007969027671997969,
      "loss": 0.8393,
      "step": 8000
    },
    {
      "epoch": 6.47,
      "learning_rate": 0.0007842091901497842,
      "loss": 0.7719,
      "step": 8500
    },
    {
      "epoch": 6.85,
      "learning_rate": 0.0007715156130997715,
      "loss": 0.801,
      "step": 9000
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.6533952546813134,
      "eval_loss": 1.7987133264541626,
      "eval_runtime": 1549.5067,
      "eval_samples_per_second": 108.393,
      "eval_steps_per_second": 0.212,
      "step": 9191
    },
    {
      "epoch": 7.24,
      "learning_rate": 0.0007588220360497588,
      "loss": 0.7443,
      "step": 9500
    },
    {
      "epoch": 7.62,
      "learning_rate": 0.0007461284589997462,
      "loss": 0.7296,
      "step": 10000
    },
    {
      "epoch": 8.0,
      "learning_rate": 0.0007334348819497334,
      "loss": 0.7508,
      "step": 10500
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.6521925515763151,
      "eval_loss": 1.7864270210266113,
      "eval_runtime": 1567.4501,
      "eval_samples_per_second": 107.152,
      "eval_steps_per_second": 0.21,
      "step": 10504
    },
    {
      "epoch": 8.38,
      "learning_rate": 0.0007207413048997208,
      "loss": 0.6659,
      "step": 11000
    },
    {
      "epoch": 8.76,
      "learning_rate": 0.0007080477278497081,
      "loss": 0.694,
      "step": 11500
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.6525795600011908,
      "eval_loss": 1.8871492147445679,
      "eval_runtime": 1630.371,
      "eval_samples_per_second": 103.016,
      "eval_steps_per_second": 0.202,
      "step": 11817
    },
    {
      "epoch": 9.14,
      "learning_rate": 0.0006953541507996953,
      "loss": 0.6611,
      "step": 12000
    },
    {
      "epoch": 9.52,
      "learning_rate": 0.0006826605737496827,
      "loss": 0.6326,
      "step": 12500
    },
    {
      "epoch": 9.9,
      "learning_rate": 0.00066996699669967,
      "loss": 0.6523,
      "step": 13000
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.6648149802030305,
      "eval_loss": 1.805672287940979,
      "eval_runtime": 1557.784,
      "eval_samples_per_second": 107.817,
      "eval_steps_per_second": 0.211,
      "step": 13130
    },
    {
      "epoch": 10.28,
      "learning_rate": 0.0006572734196496572,
      "loss": 0.602,
      "step": 13500
    },
    {
      "epoch": 10.66,
      "learning_rate": 0.0006445798425996446,
      "loss": 0.5976,
      "step": 14000
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.6706736923580721,
      "eval_loss": 1.8513981103897095,
      "eval_runtime": 1548.9015,
      "eval_samples_per_second": 108.435,
      "eval_steps_per_second": 0.212,
      "step": 14443
    },
    {
      "epoch": 11.04,
      "learning_rate": 0.0006318862655496319,
      "loss": 0.6049,
      "step": 14500
    },
    {
      "epoch": 11.42,
      "learning_rate": 0.0006191926884996192,
      "loss": 0.5476,
      "step": 15000
    },
    {
      "epoch": 11.81,
      "learning_rate": 0.0006064991114496065,
      "loss": 0.5743,
      "step": 15500
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.6628501681998155,
      "eval_loss": 1.927115559577942,
      "eval_runtime": 1525.7141,
      "eval_samples_per_second": 110.083,
      "eval_steps_per_second": 0.216,
      "step": 15756
    },
    {
      "epoch": 12.19,
      "learning_rate": 0.0005938055343995938,
      "loss": 0.5441,
      "step": 16000
    },
    {
      "epoch": 12.57,
      "learning_rate": 0.0005811119573495812,
      "loss": 0.5232,
      "step": 16500
    },
    {
      "epoch": 12.95,
      "learning_rate": 0.0005684183802995684,
      "loss": 0.5426,
      "step": 17000
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.6691911523920098,
      "eval_loss": 1.9220695495605469,
      "eval_runtime": 1530.5139,
      "eval_samples_per_second": 109.738,
      "eval_steps_per_second": 0.215,
      "step": 17069
    },
    {
      "epoch": 13.33,
      "learning_rate": 0.0005557248032495558,
      "loss": 0.5004,
      "step": 17500
    },
    {
      "epoch": 13.71,
      "learning_rate": 0.0005430312261995431,
      "loss": 0.5092,
      "step": 18000
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.6751987139412342,
      "eval_loss": 1.9163544178009033,
      "eval_runtime": 1663.4966,
      "eval_samples_per_second": 100.965,
      "eval_steps_per_second": 0.198,
      "step": 18382
    },
    {
      "epoch": 14.09,
      "learning_rate": 0.0005303376491495303,
      "loss": 0.5017,
      "step": 18500
    },
    {
      "epoch": 14.47,
      "learning_rate": 0.0005176440720995177,
      "loss": 0.4729,
      "step": 19000
    },
    {
      "epoch": 14.85,
      "learning_rate": 0.000504950495049505,
      "loss": 0.4808,
      "step": 19500
    },
    {
      "epoch": 15.0,
      "eval_accuracy": 0.6743115715519038,
      "eval_loss": 1.925864815711975,
      "eval_runtime": 1533.1814,
      "eval_samples_per_second": 109.547,
      "eval_steps_per_second": 0.215,
      "step": 19695
    },
    {
      "epoch": 15.23,
      "learning_rate": 0.0004922569179994923,
      "loss": 0.4595,
      "step": 20000
    },
    {
      "epoch": 15.61,
      "learning_rate": 0.00047956334094947955,
      "loss": 0.4489,
      "step": 20500
    },
    {
      "epoch": 15.99,
      "learning_rate": 0.00046686976389946687,
      "loss": 0.4611,
      "step": 21000
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.6784972165163288,
      "eval_loss": 1.9857261180877686,
      "eval_runtime": 1537.6389,
      "eval_samples_per_second": 109.229,
      "eval_steps_per_second": 0.214,
      "step": 21008
    },
    {
      "epoch": 16.0,
      "step": 21008,
      "total_flos": 7.331020593786202e+20,
      "train_loss": 0.0,
      "train_runtime": 372.6046,
      "train_samples_per_second": 27045.44,
      "train_steps_per_second": 52.858
    }
  ],
  "logging_steps": 500,
  "max_steps": 19695,
  "num_train_epochs": 15,
  "save_steps": 500,
  "total_flos": 7.331020593786202e+20,
  "trial_name": null,
  "trial_params": null
}