{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 20.0,
  "eval_steps": 500,
  "global_step": 1100,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02,
      "learning_rate": 4.545454545454545e-07,
      "loss": 2.7297,
      "step": 1
    },
    {
      "epoch": 0.36,
      "learning_rate": 9.090909090909091e-06,
      "loss": 2.7244,
      "step": 20
    },
    {
      "epoch": 0.73,
      "learning_rate": 1.8181818181818182e-05,
      "loss": 2.7672,
      "step": 40
    },
    {
      "epoch": 1.0,
      "gpt4_scores": 0.0,
      "step": 55
    },
    {
      "epoch": 1.0,
      "eval_loss": 2.312263250350952,
      "eval_runtime": 4.9535,
      "eval_samples_per_second": 4.643,
      "eval_steps_per_second": 1.211,
      "step": 55
    },
    {
      "epoch": 1.09,
      "learning_rate": 2.7272727272727273e-05,
      "loss": 2.4172,
      "step": 60
    },
    {
      "epoch": 1.45,
      "learning_rate": 3.6363636363636364e-05,
      "loss": 2.1557,
      "step": 80
    },
    {
      "epoch": 1.82,
      "learning_rate": 4.545454545454546e-05,
      "loss": 2.0304,
      "step": 100
    },
    {
      "epoch": 2.0,
      "gpt4_scores": 0.0,
      "step": 110
    },
    {
      "epoch": 2.0,
      "eval_loss": 1.96739661693573,
      "eval_runtime": 4.9491,
      "eval_samples_per_second": 4.647,
      "eval_steps_per_second": 1.212,
      "step": 110
    },
    {
      "epoch": 2.18,
      "learning_rate": 4.9987413559579636e-05,
      "loss": 1.8323,
      "step": 120
    },
    {
      "epoch": 2.55,
      "learning_rate": 4.988679806432712e-05,
      "loss": 1.8472,
      "step": 140
    },
    {
      "epoch": 2.91,
      "learning_rate": 4.968597221690986e-05,
      "loss": 1.7981,
      "step": 160
    },
    {
      "epoch": 3.0,
      "gpt4_scores": 0.0,
      "step": 165
    },
    {
      "epoch": 3.0,
      "eval_loss": 1.942185401916504,
      "eval_runtime": 4.9566,
      "eval_samples_per_second": 4.64,
      "eval_steps_per_second": 1.211,
      "step": 165
    },
    {
      "epoch": 3.27,
      "learning_rate": 4.938574467213518e-05,
      "loss": 1.6577,
      "step": 180
    },
    {
      "epoch": 3.64,
      "learning_rate": 4.898732434036244e-05,
      "loss": 1.6171,
      "step": 200
    },
    {
      "epoch": 4.0,
      "learning_rate": 4.849231551964771e-05,
      "loss": 1.5825,
      "step": 220
    },
    {
      "epoch": 4.0,
      "gpt4_scores": 0.0,
      "step": 220
    },
    {
      "epoch": 4.0,
      "eval_loss": 2.0124683380126953,
      "eval_runtime": 4.9519,
      "eval_samples_per_second": 4.645,
      "eval_steps_per_second": 1.212,
      "step": 220
    },
    {
      "epoch": 4.36,
      "learning_rate": 4.790271143580174e-05,
      "loss": 1.2929,
      "step": 240
    },
    {
      "epoch": 4.73,
      "learning_rate": 4.722088621637309e-05,
      "loss": 1.2811,
      "step": 260
    },
    {
      "epoch": 5.0,
      "gpt4_scores": 0.0,
      "step": 275
    },
    {
      "epoch": 5.0,
      "eval_loss": 2.265650510787964,
      "eval_runtime": 4.96,
      "eval_samples_per_second": 4.637,
      "eval_steps_per_second": 1.21,
      "step": 275
    },
    {
      "epoch": 5.09,
      "learning_rate": 4.644958533087443e-05,
      "loss": 1.1431,
      "step": 280
    },
    {
      "epoch": 5.45,
      "learning_rate": 4.559191453574582e-05,
      "loss": 0.9042,
      "step": 300
    },
    {
      "epoch": 5.82,
      "learning_rate": 4.465132736856969e-05,
      "loss": 0.928,
      "step": 320
    },
    {
      "epoch": 6.0,
      "gpt4_scores": 0.0,
      "step": 330
    },
    {
      "epoch": 6.0,
      "eval_loss": 2.5270919799804688,
      "eval_runtime": 4.9548,
      "eval_samples_per_second": 4.642,
      "eval_steps_per_second": 1.211,
      "step": 330
    },
    {
      "epoch": 6.18,
      "learning_rate": 4.3631611241893874e-05,
      "loss": 0.7725,
      "step": 340
    },
    {
      "epoch": 6.55,
      "learning_rate": 4.2536872192658036e-05,
      "loss": 0.6495,
      "step": 360
    },
    {
      "epoch": 6.91,
      "learning_rate": 4.137151834863213e-05,
      "loss": 0.6688,
      "step": 380
    },
    {
      "epoch": 7.0,
      "gpt4_scores": 0.0,
      "step": 385
    },
    {
      "epoch": 7.0,
      "eval_loss": 2.780010461807251,
      "eval_runtime": 5.0002,
      "eval_samples_per_second": 4.6,
      "eval_steps_per_second": 1.2,
      "step": 385
    },
    {
      "epoch": 7.27,
      "learning_rate": 4.014024217844167e-05,
      "loss": 0.5264,
      "step": 400
    },
    {
      "epoch": 7.64,
      "learning_rate": 3.884800159665276e-05,
      "loss": 0.4558,
      "step": 420
    },
    {
      "epoch": 8.0,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 0.4595,
      "step": 440
    },
    {
      "epoch": 8.0,
      "gpt4_scores": 0.0,
      "step": 440
    },
    {
      "epoch": 8.0,
      "eval_loss": 3.1072428226470947,
      "eval_runtime": 4.9459,
      "eval_samples_per_second": 4.65,
      "eval_steps_per_second": 1.213,
      "step": 440
    },
    {
      "epoch": 8.36,
      "learning_rate": 3.610166531514436e-05,
      "loss": 0.3119,
      "step": 460
    },
    {
      "epoch": 8.73,
      "learning_rate": 3.465862814232822e-05,
      "loss": 0.3117,
      "step": 480
    },
    {
      "epoch": 9.0,
      "gpt4_scores": 0.0,
      "step": 495
    },
    {
      "epoch": 9.0,
      "eval_loss": 3.326101541519165,
      "eval_runtime": 4.9468,
      "eval_samples_per_second": 4.649,
      "eval_steps_per_second": 1.213,
      "step": 495
    },
    {
      "epoch": 9.09,
      "learning_rate": 3.3176699082935545e-05,
      "loss": 0.3084,
      "step": 500
    },
    {
      "epoch": 9.45,
      "learning_rate": 3.166184534225087e-05,
      "loss": 0.2188,
      "step": 520
    },
    {
      "epoch": 9.82,
      "learning_rate": 3.012016670162977e-05,
      "loss": 0.242,
      "step": 540
    },
    {
      "epoch": 10.0,
      "gpt4_scores": 0.0,
      "step": 550
    },
    {
      "epoch": 10.0,
      "eval_loss": 3.5036461353302,
      "eval_runtime": 4.9552,
      "eval_samples_per_second": 4.642,
      "eval_steps_per_second": 1.211,
      "step": 550
    },
    {
      "epoch": 10.18,
      "learning_rate": 2.8557870956832132e-05,
      "loss": 0.198,
      "step": 560
    },
    {
      "epoch": 10.55,
      "learning_rate": 2.698124892141971e-05,
      "loss": 0.1773,
      "step": 580
    },
    {
      "epoch": 10.91,
      "learning_rate": 2.5396649095870202e-05,
      "loss": 0.1718,
      "step": 600
    },
    {
      "epoch": 11.0,
      "gpt4_scores": 0.0,
      "step": 605
    },
    {
      "epoch": 11.0,
      "eval_loss": 3.6741881370544434,
      "eval_runtime": 4.9476,
      "eval_samples_per_second": 4.649,
      "eval_steps_per_second": 1.213,
      "step": 605
    },
    {
      "epoch": 11.27,
      "learning_rate": 2.3810452104406444e-05,
      "loss": 0.14,
      "step": 620
    },
    {
      "epoch": 11.64,
      "learning_rate": 2.222904500247473e-05,
      "loss": 0.1337,
      "step": 640
    },
    {
      "epoch": 12.0,
      "learning_rate": 2.0658795558326743e-05,
      "loss": 0.157,
      "step": 660
    },
    {
      "epoch": 12.0,
      "gpt4_scores": 0.0,
      "step": 660
    },
    {
      "epoch": 12.0,
      "eval_loss": 3.789468288421631,
      "eval_runtime": 4.9943,
      "eval_samples_per_second": 4.605,
      "eval_steps_per_second": 1.201,
      "step": 660
    },
    {
      "epoch": 12.36,
      "learning_rate": 1.9106026612264316e-05,
      "loss": 0.1428,
      "step": 680
    },
    {
      "epoch": 12.73,
      "learning_rate": 1.7576990616793137e-05,
      "loss": 0.1188,
      "step": 700
    },
    {
      "epoch": 13.0,
      "gpt4_scores": 0.0,
      "step": 715
    },
    {
      "epoch": 13.0,
      "eval_loss": 3.9183552265167236,
      "eval_runtime": 4.9519,
      "eval_samples_per_second": 4.645,
      "eval_steps_per_second": 1.212,
      "step": 715
    },
    {
      "epoch": 13.09,
      "learning_rate": 1.6077844460203206e-05,
      "loss": 0.1176,
      "step": 720
    },
    {
      "epoch": 13.45,
      "learning_rate": 1.4614624674952842e-05,
      "loss": 0.108,
      "step": 740
    },
    {
      "epoch": 13.82,
      "learning_rate": 1.3193223130682936e-05,
      "loss": 0.1296,
      "step": 760
    },
    {
      "epoch": 14.0,
      "gpt4_scores": 0.0,
      "step": 770
    },
    {
      "epoch": 14.0,
      "eval_loss": 3.996537208557129,
      "eval_runtime": 4.9512,
      "eval_samples_per_second": 4.645,
      "eval_steps_per_second": 1.212,
      "step": 770
    },
    {
      "epoch": 14.18,
      "learning_rate": 1.181936330973744e-05,
      "loss": 0.1085,
      "step": 780
    },
    {
      "epoch": 14.55,
      "learning_rate": 1.049857726072005e-05,
      "loss": 0.1042,
      "step": 800
    },
    {
      "epoch": 14.91,
      "learning_rate": 9.236183322886945e-06,
      "loss": 0.1219,
      "step": 820
    },
    {
      "epoch": 15.0,
      "gpt4_scores": 0.0,
      "step": 825
    },
    {
      "epoch": 15.0,
      "eval_loss": 4.114721775054932,
      "eval_runtime": 4.9671,
      "eval_samples_per_second": 4.63,
      "eval_steps_per_second": 1.208,
      "step": 825
    },
    {
      "epoch": 15.27,
      "learning_rate": 8.0372647110717e-06,
      "loss": 0.101,
      "step": 840
    },
    {
      "epoch": 15.64,
      "learning_rate": 6.906649047373246e-06,
      "loss": 0.1184,
      "step": 860
    },
    {
      "epoch": 16.0,
      "learning_rate": 5.848888922025553e-06,
      "loss": 0.1,
      "step": 880
    },
    {
      "epoch": 16.0,
      "gpt4_scores": 0.0,
      "step": 880
    },
    {
      "epoch": 16.0,
      "eval_loss": 4.137594699859619,
      "eval_runtime": 4.9557,
      "eval_samples_per_second": 4.641,
      "eval_steps_per_second": 1.211,
      "step": 880
    },
    {
      "epoch": 16.36,
      "learning_rate": 4.868243561723535e-06,
      "loss": 0.0975,
      "step": 900
    },
    {
      "epoch": 16.73,
      "learning_rate": 3.968661679220468e-06,
      "loss": 0.1155,
      "step": 920
    },
    {
      "epoch": 17.0,
      "gpt4_scores": 0.0,
      "step": 935
    },
    {
      "epoch": 17.0,
      "eval_loss": 4.1758294105529785,
      "eval_runtime": 4.9475,
      "eval_samples_per_second": 4.649,
      "eval_steps_per_second": 1.213,
      "step": 935
    },
    {
      "epoch": 17.09,
      "learning_rate": 3.1537655732553768e-06,
      "loss": 0.1017,
      "step": 940
    },
    {
      "epoch": 17.45,
      "learning_rate": 2.4268365428344736e-06,
      "loss": 0.0972,
      "step": 960
    },
    {
      "epoch": 17.82,
      "learning_rate": 1.790801674598186e-06,
      "loss": 0.1083,
      "step": 980
    },
    {
      "epoch": 18.0,
      "gpt4_scores": 0.0,
      "step": 990
    },
    {
      "epoch": 18.0,
      "eval_loss": 4.201599597930908,
      "eval_runtime": 4.963,
      "eval_samples_per_second": 4.634,
      "eval_steps_per_second": 1.209,
      "step": 990
    },
    {
      "epoch": 18.18,
      "learning_rate": 1.248222056476367e-06,
      "loss": 0.1031,
      "step": 1000
    },
    {
      "epoch": 18.55,
      "learning_rate": 8.012824650910938e-07,
      "loss": 0.109,
      "step": 1020
    },
    {
      "epoch": 18.91,
      "learning_rate": 4.517825684323324e-07,
      "loss": 0.0961,
      "step": 1040
    },
    {
      "epoch": 19.0,
      "gpt4_scores": 0.0,
      "step": 1045
    },
    {
      "epoch": 19.0,
      "eval_loss": 4.204728603363037,
      "eval_runtime": 4.956,
      "eval_samples_per_second": 4.641,
      "eval_steps_per_second": 1.211,
      "step": 1045
    },
    {
      "epoch": 19.27,
      "learning_rate": 2.011296792301165e-07,
      "loss": 0.0981,
      "step": 1060
    },
    {
      "epoch": 19.64,
      "learning_rate": 5.033308820289184e-08,
      "loss": 0.1161,
      "step": 1080
    },
    {
      "epoch": 20.0,
      "learning_rate": 0.0,
      "loss": 0.0917,
      "step": 1100
    },
    {
      "epoch": 20.0,
      "gpt4_scores": 0.0,
      "step": 1100
    },
    {
      "epoch": 20.0,
      "eval_loss": 4.205177307128906,
      "eval_runtime": 4.9637,
      "eval_samples_per_second": 4.634,
      "eval_steps_per_second": 1.209,
      "step": 1100
    },
    {
      "epoch": 20.0,
      "step": 1100,
      "total_flos": 3.836907908090266e+16,
      "train_loss": 0.6597742699493061,
      "train_runtime": 3169.8576,
      "train_samples_per_second": 1.369,
      "train_steps_per_second": 0.347
    }
  ],
  "logging_steps": 20,
  "max_steps": 1100,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 20,
  "save_steps": 20,
  "total_flos": 3.836907908090266e+16,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}